3,227 1,593 59MB
Pages 698 Page size 336 x 511.68 pts Year 2007
Foodborne Disease Handbook Second Edition, Revised and Expanded Volume 1: Bacterial Pathogens
edited by
Y. H. Hui Science TechnologySystem West Sacramento, California
Merle D. Pierson Virginia PolytechnicInstitute and State University Blacksburg, Virginia
J. Richard Gorham Uniformed Services University of the Health Sciences Bethesda, Maryland
m M A R C E L
D E K K E R
MARCEL DEKKER, INC.
NEWYORK BASEL
ISBN: 0-8247-0337-5 This book is printed on acid-free paper.
Headquarters Marcel Dekker, Inc. 270 Madison Avenue. New York, NY 10016 tel: 212-696-9000; fax: 212-685-4540 Eastern Hemisphere Distribution Marcel Drkker AG Hutgasse 4. Postfach 812, CH4001 Basel, Switzerland tel: 4 1-6 1-261-8482; fax: 41 -6 1 -26 1-8896 World Wide Web http://www.dekker.com The publisher offers discounts on this book when ordered in bulk quantities. For more information, write to Special Sales/Professional Marketing at the headquarters address above.
Copyright 0 2001 by Marcel Dekker, Inc. All Rights Reserved. Neither this book nor any part may be reproduced or transmitted in any form or by any means. electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage and retrieval system, without permission in writing from the publisher. Current printing (last digit): l 0 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA
Introduction to the Handbook
The Foodboine Disease Handbook,Second Edition,Revised and Expanded, could not be appearing at a more auspicious time. Never before has the campaign for food safety been pursued so intensely on so many fronts in virtually every country around the world. This new edition reflects at least one of the many aspects of that intense and multifaceted campaign: namely, that research on food safety has been very productive in the years since the first edition appeared. The Hcmdbook is now presented in four volumes instead of the three of the 1994 edition. The four volumes are composed of 86 chapters, a 22% increase over the 67 chapters of the first edition. Much of the information in the first edition has been carried forward to this new edition because that information is still as reliable and pertinent as it was in 1994. This integration of the older data with the latest research findings gives the reader a secure scientific foundation on whichto base important decisions affecting the public's health. We are not so naive as to think that only scientific facts influence decisions affecting food safety. Political and economic factors and cotnpelling national interests may carry greater weight in the minds of decisionmakers than the scientific findings offered in this new edition. However, if persons in the higher levels of national governments and international agencies, such as the Codex Alinlentarius Commission, the World Trade Organization, the World Health Organization, and the Food and Agriculture Organization, who must bear the burden of decision-making need and are willing to entertain scientific findings, then the information in these four volumes will serve them well indeed. During the last decade of the previous century, we witnessed an unprecedentedly intense and varied program of research on food safety, as we have already noted. There are compelling forces driving these research efforts. The traditional food-associated pathogens, parasites, and toxins of forty years ago still continue to cause problems today, and newer or less well-known species and strains present extraordinary challenges to human health. These newer threats may be serious even for the immunocompetent, but for the immunocompromised they can be devastating. The relative numbers of the immunocompromised in the world population are increasing daily. We include here not just those affected by the humanimmunodeficiency virus (HIV), but also the elderly: the veryyoung; the recipients of radiation treatments, chemotherapy, and immunosuppressive drugs; paiii
Handbook iv
the
to
Introduction
tients undergoing major invasive diagnostic or surgical procedures; and sufferers of debilitating diseases such as diabetes. To this daunting list of challenges must be added numerous instances of microbial resistance to antibiotics. Moreover, it is not yet clear how the great HACCP experiment will play out on the worldwide stage of food safety. Altruism and profit motivation have always made strange bedfellows in the food industry. It remains to be seen whether HACCP will succeed in wedding these two disparate tnotives into a unifying force for the benefit of all concerned-producers, manufacturers, retailers, and consumers. That HACCP shows great promise is thoroughly discussed in Volume 2, with an emphasis on sanitation in a public eating place. All the foregoing factors lend a sense of urgency to the task of rapidly identifying toxins, species, and strains of pathogens and parasites as etiologic agents, and of determining their roles in the epidemiology and epizootiology of disease outbreaks, which are described in detail throughout the Foodborne Disease Hmdbook. It is very fortunate for the consumer that there exists in the food industry a dedicated cadre of scientific specialists who scrutinize all aspects of food production and bring their expertise to bear on the potential hazards they know best. A good sampling of the kinds of work they do iscontained in these four new volumes of the Handbook. And the benefits of their research are obvious to the scientific specialist who wants to learn even more about food hazards, to the scientific generalist who is curious about everything and who will be delighted to find a good source of accurate, up-to-date information, and to consumers who care about what they eat. We are confident that these four volumes will provide competent, trustworthy, and timely information to inquiring readers, no matter what roles they may play in the global campaign to achieve food safety.
I
Y. H. Hui J. Richard Gorham Dnvid Kitts K. D. Murre11 Wai-Kit Nip Merle D. Piersorl Syed A. Sattnr R. A. Smith David G. Spoerke, Jr. Peggy S. Starzfield
Preface
The first volume of the Foodborne Disease Handbook, Second Edition, Revised m d Expanded, focuses on bacterial pathogens. Although great strides have been made in food sanitation in general and in the application of HACCP principles in particular, hardly a week goes by without a report in the media of some outbreak of foodborne bacterial disease. It is evident that there is a gap between the principles and facts recorded in the first volume of the Handbook and the application of these facts and principles to the protection of the public’s health. Outbreaks of foodborne bacterial disease do not happen spontaneously-the principle of cause-and-effect remains fully operational. Moreover, people do not usually get sick by consuming wholesome food. Further, those who produce, manufacture, retail, and serve food have no wish to make people sick. Thus, when a foodborne disease outbreak occurs, it suggests that some (probably preventable) circumstance or set of circumstances was allowed to occur that permitted the proliferation of bacteria or the production of their toxins. As HACCP principles emphasize, prevention is the key to averting foodborne bacterial disease outbreaks. The editors and contributors to this volume of the Foodborne Disease Harzdbook have worked diligently to ensure that readers representing all aspects of the food industry now have readily at hand all the facts and principles needed to prevent foodborne outbreaks of bacterial diseases.
Y. H. Hui Merle D. Piersol? J. Richard Gorhnm
V
This Page Intentionally Left Blank
Contents
Introduction to the Harldbook Prefkce Contributors Contents of Other Volurzres
...
111
17
xi AV
I. Poison Centers 1. The Role of U.S. Poison Centers in Bacterial Exposures David G. Spoerke, Jr.
1
11. Bacterial Pathogens 2.
Bacterial Biota (Flora) in Foods James M. Jay
23
3. Aerornonas hydrophilu Carlos Abeyta, Jr., Sarrruel A. Palrmbo. arrd Gerard N. Stelma, Jr.
35
4. Update: Food Poisoning and Other Diseases Induced by Bacilluscereus Kathleen T. Rajkowski and James L. Smith
61
5. Brucella Shirley M. Halling and Edward J. Young
77
6. Campylobacter jejuni Don A. Franco a ~ Charles d E. Willinrns
83
vii
Contents
viii
7.
Clostridium botulinum John W. Austh and Karen L. Dodds
107
8. Clostridium perfringens Dorothy M. Wrigley
139
9. Escherichia coli Mcrrguerite A. Neill, Phillip I. Tarr, David N. Tcrvlor, and Marcia Worf
169
10. Listeria nzonocytogenes Catherine W. Donnelly
213
11. Bacteriology of Salmonella Robin C. Anderson and Richcrrd L. Ziprin
247
12. Salmonellosis in Animals David J. Nisbet and Richard L. Ziprin
265
13. Human Salmonellosis: General Medical Aspects Richard L. Ziprin and Michael H. Hume
285
14. Shigella Anthony T. Maurelli and Keith A. Lmnpel
323
15. Staphylococcusaureus Scott E. Martin, Eric R. Mvers, and John J. Iandolo
345
16. Vibrio cholerae Charles A. Kaysner and June H. Wetherirlgton
383
17. Vibrio parahaenzolyticus Tuu-jyi Chai and John L. Pace
407
18. Vibrio vuln@cus Anders Dalsgaard, Lise H@i, DebiLinttous: and James D. Oliver
439
19. Yersinin Scott A. Minnich, Michael J. Smith, Steven D. Weagant, crnd Peter Feng
47 l
111. Disease Surveillance, Investigation, andIndicatorOrganisms 20.
Surveillance of Foodborne Disease Ewer1 C. D. Todd
515
Contents
ix
21. Investigating Foodborne Disease Dale L. Morse, Guthrie S. Birkhead, and Jack J. Guzewich
587
22. Indicator Organisms in Foods James M. Juy
645
Index
655
This Page Intentionally Left Blank
Contributors
Carlos Abeyta, Jr.
U.S. Food and Drug Administration, Bothell, Washington
Robin C. Anderson Food Animal Protection Research Laboratory, Agricultural Research Service, U.S. Department of Agriculture, College Station, Texas John W. Austin Bureau of Microbial Hazards, Health Protection Branch, Health Canada, Ottawa, Ontario, Canada Guthrie S. Birkhead New York State Department of Health, Albany, New York Tuu-jyi Chai Department of Food Science, National Taiwan Ocean University, Keelung, Taiwan, Republic of China Anders Dalsgaard Department of Veterinary Microbiology, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark Karen L. Dodds Health Canada, Ottawa, Ontario, Canada Catherine W. Donnelly Department of Nutrition and Food Sciences, University of Vermont, Burlington, Vermont Peter Feng Division of Microbiological Studies, U.S. Food and Drug Administration, Washington, D.C. Don A. Franco
Animal Protein Producers Industry, Huntsville, Missouri
Jack J. Guzewich U.S. Food and Drug Administration, Washington, D.C. Shirley M. Halling U.S. Department of Agriculture, Ames, Iowa Lise HGi Department of Veterinary Microbiology, The Royal Veterinary and Agricultural University, Frederiksberg, Denmark Michael H. Hume Food Animal Protection Research Laboratory, Agricultural Research Service, U.S. Department of Agriculture, College Station, Texas JohnJ. Iandolo Department of Microbiology and Immunology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma xi
xii
James M. Jay
Contributors
University of Nevada Las Vegas, Las Vegas, Nevada
Charles A. Kaysner Seafood Products Research Center, U.S. Food and Drug Administration, Bothell, Washington Keith A. Lampel Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, Washington, D.C. Debi Linkous
Burroughs Wellcome Research Fund, Raleigh, North Carolina
Scott E. Martin Department of Food Science and Human Nutrition, University of Illinois at Urbana-Champaign, Urbana, Illinois Anthony T. Maurelli Department of Microbiology and Immunology, Uniformed Services University of the Health Sciences, Bethesda, Maryland Scott A. Minnich Department of Microbiology, University of Idaho, Moscow, Idaho Dale L. Morse Wadsworth Center, New York State Department of Health, Albany, New York Eric R. Myers Nalco Chemical Company, Naperville, Illinois Marguerite A. Neil1 Department of Medicine, Division of Infectious Disease, Brown University School of Medicine, Providence, Rhode Island David J. Nisbet Food and Feed Safety Research Unit, Food Animal Protection Research Laboratory, Agricultural Research Service, U.S. Department of Agriculture, College Station, Texas James D. Oliver University of North Carolina at Charlotte, Charlotte, North Carolina JohnL.Pace Biochenlistry Department, Advanced Medicine, Inc., South San Francisco, California Samuel A. Palumbo U.S. Department of Agriculture, Philadelphia, Pennsylvania KathleenT.Rajkowski Department of Food Safety, Agricultural Research Service, U.S. Department of Agriculture, Wyndmoor, Pennsylvania James L. Smith Microbial Food Safety Lab, Agricultural Research Service, U.S. Department of Agriculture, Wyndmoor, Pennsylvania Michael J. Smith Department of Microbiology, University of Idaho, Moscow, Idaho David G. Spoerke, Jr. Bristlecone Enterprises, Denver, Colorado Gerard N. Stelma, Jr. U.S. Environmental Protection Agency, Cincinnati, Ohio Phillip I. Tarr University of Washington and Children’s Hospital and Medical Center, Seattle, Washington David N. Taylor Walter Reed Army Institute of Research, Washington, D.C. Ewen C. D. Todd Health Protection Branch, Bureau of Microbial Hazards, Health Canada, Ottawa, Ontario, Canada Steven D. Weagant U.S. Food and Drug Administration, Bothell, Washington
I
Contributors
Xiii
June H. Wetherington U S . Food and Drug Administration, Bothell, Washington Charles E. Williams Consultant, Arlington, Virginia Marcia Wolf Walter Reed Army Institute of Research, Washington, D.C. Dorothy M. Wrigley Mankato, Minnesota
Department of Biological Sciences, Minnesota State University,
Edward J. Young VA Medical Center and Baylor College of Medicine, Houston, Texas Richard L. Ziprin Food and Feed Safety Research Unit, Food Animal Protection Research Laboratory, Agricultural Research Service, U.S. Department of Agriculture, College Station, Texas
This Page Intentionally Left Blank
Contents of Other Volumes
VOLUME 2: VIRUSES, PARASITES, PATHOGENS, AND HACCP I. PoisonCenters 1. The Role of Poison Centers in the United States David G. Spoerke, Jr.
11. Viruses 2. Hepatitis A and E Viruses Theresa L. Cron.leans, Michael 0. Fmol-ov, Omana V. Nainan, and Harold S. Margolis
3. Norwalk Virus and the Small Round Viruses Causing Foodborne Gastroenteritis Hazel Appleton 4. Rotavirus Syed A. Snttnr, V. S u s m Springthorpe, and Jason A. Tetro
5. Other Foodborne Viruses Syed A. Sattar and Jason A. Tetl-o 6. Detection of Human Enteric Viruses in Foods Lee-Ann Jaykus 7. Medical Management of Foodborne Viral Gastroenteritis and Hepatitis SuZarwe M. Mcrtsui and Rarnsey C. Cheung xv
xwi
Contents of Other Volumes
8. Epidemiology of Foodborne Viral Infections Thornns M. Liithi
9. Environmental Considerations in Preventing the Foodborne Spread of Hepatitis A Syed A. Sattar and Sabah Bidawid
111. Parasites 10. Taeniasis and Cysticercosis Zbigniew S. Pawlowski and K. D. Murre11
11. Meatborne Helminth Infections: Trichinellosis William C.Campbell 12. Fish- and Invertebrate-Borne Helminths John H. Cross
13. Waterborne and Foodborne Protozoa Ronald Foyer
14. Medical Management Pm1 Prociv 15. Immunodiagnosis of Infections with Cestodes Bruno Gottstein
16. Immunodiagnosis: Nematodes H. Ray Gamble 17. Diagnosis of Toxoplasmosis Alan M. Johnson and J. P. Dubey 18. Seafood Parasites: Prevention, Inspection, and HACCP Ann M. Admm and Debra D. DeVlieger
IV. HACCP and the Foodservice Industries 19. Foodservice Operations: HACCP Principles 0. Peter S q d e r , Jr. 20. Foodservice Operations: HACCP Control Programs 0. Peter Snyder, Jr. Index
VOLUME 3: PLANT TOXICANTS I. PoisonCenters 1. U.S. Poison Centers for Exposures to Plant and Mushroom Toxins David G. Spoerke, Jr.
Contents of Other Volumes
11. Selected Plant Toxicants 2. Toxicology of Naturally Occurring Chemicals in Food Ross C. Beier and Herbert N. Nigg
3. Poisonous Higher Plants Doreen Grace Lang and R. A. Smith 4. Alkaloids R. A. Smith 5. Antinutritional Factors Related to Proteins and Amino Acids Iwin E. Liener 6. Glycosides Walter Majak and Miclznel H. Berm
7. Analytical Methodology for Plant Toxicants Alister David Muir
8. Medical Management and Plant Poisoning Robert H. Poppenga 9. Plant Toxicants and Livestock: Prevention and Management Michcrel H. R a l p h
111. Fungal Toxicants 10. Aspergillus Zojia Kozcrkiewicz 11. Claviceps and Related Fungi Gretchen A. Kuldau and Charles W. Bacon
12. Fusariunl Walter F. 0. Marmas 13. Penicillium Johr~I. Pitt 14. Foodborne Disease and Mycotoxin Epidemiology Sara Hale Herlly and F. Xcwier Bosch 15. Mycotoxicoses: The Effects of Interactions with Mycotoxins Heather A. Koshinsky, Adrieme Woytowich, and George G. Khachatourians
16. Analytical Methodology for Mycotoxins James K. Porter 17. Mycotoxin Analysis: Immunological Techniques Fur1 S. Chrr
xvii
xviii
Volumes Contents of Other
18. Mushroom Biology: General Identification Features Dmid G. Spoerke, Jr. 19. Identification of Mushroom Poisoning (Mycetismus), Epidemiology, and Medical Management David G. Spoerke, Jr. 20. Fungi in Folk Medicine and Society Dnvid G. Syoerke, Jr.
VOLUME 4: SEAFOOD AND ENVIRONMENTAL TOXINS I. Poison Centers 1. Seafood and Environmental Toxicant Exposures: The Role of Poison Centers Dnvid G. Spoerke, Jr.
11. Seafood Toxins 2. Fish Toxins Bruce W. Halstencl 3. Other Poisonous Marine Animals Bruce W. Hulsteacl
4. Shellfish Chemical Poisoning Lyndon E. Llewellyn 5. Pathogens Transmitted by Seafood Russell P. Hemlig
6. Laboratory Methodology for Shellfish Toxins David Kitts 7. Ciguatera Fish Poisoning Yoshitsugi Hokema and Jocmne S. M. Yoshikuwa-Ebesu 8. Tetrodotoxin Joame S. M. Yoslzikablta-Ebesu,Yoslzitsugi Hokamn, nnd TUIIKIO Noguchi
9. Epidemiology of Seafood Poisoning Lorn E. Flerning, Dolores Kat:, Judy A. Bean, and Robertl! Hnnurlond
Contents of Other Volumes
10. The Medical Management of Seafood Poisoning Donm Glad Blythe, Eileen Hack, Giavarmi Washirzgtofl,and Lorn E. Fleming 11. The U.S. National Shellfish Sanitation Program Rebecca A. Reid a d Timothy D. Durmce
12. HACCP, Seafood, and the U.S. Food and Drug Administration Kim R. Yomg, Miguel Rodrigues Krrnzrst, arid George Per)? Hoskin 111. Environmental Toxins
13. Toxicology and Risk Assessment D o r ~ ~J.l lEcobichorz 14. Nutritional Toxicology David Kitts 15. Food Additives Lasclo P. Sornogyi
16. Analysis of Aquatic Contaminants Joe W. Kiceniuk 17. Agricultural Chemicals Debra L. Browning and Car1 K. Winter 18. Radioactivity in Food and Water Hank Kocol
19. Food Irradiation Hark Kocol 20. Drug Residues in Foods of Animal Origin Austin R. Long and Jose E. Roybal 21. Migratory Chemicals from Food Containers and Preparation Utensils Yvonne V. Y ~ r m 22. Food and Hard Foreign Objects: A Review J. Richard Gol-ham 23. Food, Filth, and Disease: A Review J. Richard Gorham 24. Food Filth and Analytical Methodology: A Synopsis J. Richard Gol-hnm
xix
This Page Intentionally Left Blank
1 The Role of U S . Poison Centers in Bacterial Exposures David G. Spoerke, Jr. Bristlecorle Enterprises, Denver, Colorado
I. Epidemiology A. B. C. D. E. F. G. H. I. J.
1
AAPCC 2 Who staffs a poison center'? 3 What types of calls are received? 4 How calls are handled 5 What references are used? 6 How poison centers are monitored for quality 6 Professional and public education programs 7 Related professional toxicology organizations 7 International affiliations 9 Toxicology and poison center Web sites 10
11. U.S. Poison Information Centers
10
References 2 1
1.
EPIDEMIOLOGY
Epidemiological studies aid treatment facilities in determining risk factors, determining who becomes exposed, and establishing the probable outcomes with various treatments. A few organizations have attempted to gather such information and organize it into yearly reports. The American Association of Poison Control Centers and some federal agencies work toward obtaining epidemiological information, while the AAPCC has an active role in assisting with the treatment of exposures. Epidemiological studies assist government and industry in determining package safety, effective treatment measures, conditions of exposure, and frequency of exposure. Studies on bacterial exposures provide information on the type of people most commonly involved. (e.g., children, adults at home, outdoorsmen, industrial workers, or bluecollar workers). Studies can also tell us which bacterial species are most commonly involved. The symptoms first seen, the onset of symptoms, and the sequelae may also be determined and compared to current norms. l
Spoerke
2
A.
AAPCC
1. What Are Poison Centers and the AAPCC? The group in the United States most concerned on a daily basis with potential poisonings due to household agents, industrial agents, and biologics (including plants and mushrooms) is the American Association of Poison Control Centers (AAPCC). This is an affiliation of local and regional centers that provides information concerning all aspects of poisoning and often refers patients to treatment centers. This group of affiliated centers is often supported by both government, private funds, and industrial sources. Poison centers were started in the late 1950s, with the first center thought to be in the Chicago area. The idea caught on quickly, and at the peak of the movement there were hundreds of centers throughout the United States. Unfortunately there were few or no standards regarding what might be called a poison center, the type of staff, hours of operation, or information resources. One center may have had a dedicated staff of doctors, pharmacists, and nurses trained specifically in handling poison cases; the next center may just have had a book on toxicology in the emergency room or hospital library. In 1993 the Health and Safety Code (Sec. 777.002) specified that a poison center must provide a 24-hour service for public and health care professionals and meet requirements established by the AAPCC. This action helped to standardize the activities and the staffs of the various poison centers. The federal government does not fund poison centers, even though for every dollar spent on poison centers there is a savings of $2 to $9 in unnecessary expenses (1,2). The federal agency responsible for the Poison Prevention Packaging Act is the U.S. Consumer Product Safety Commission (CPSC). The National Clearinghouse for Poison Control Centers initially collected data on poisonings, information on commercial product ingredients, and biological toxic agents. For several years the National Clearinghouse provided product and treatment information to the poison centers who handled day-to-day calls. At first, most poison centers were funded by the hospital in which they were located. As the centers grew in size and number of calls being handled, both city and state governments took on the responsibility of contributing funds. In recent years the local governments have found it very difficult to fund such operations and centers have had to look to private industry for additional funding. Government funding may take several forms, as a line item on a state’s budget, as a direct grant, or as moneys distributed on a percall basis. Some states with fewer residents may contract with a neighboring state to provide services to its residents. Some states are so populous that more than one center is funded by the state. Industrial funding also varies, sometimes as a grant, sometimes as payment for handling the company’s poison or drug information-related calls, sometimes as payment for collection of data regarding exposure to the company’s product. Every year the AAPCC issues a summary of all kinds of exposures. A few bacterial exposures are listed in this log, most of which have to do with food poisoning.
2. Regional Centers The number of listed centers has dropped significantly since its peak of more than 600. Many centers have been combined into regional organizations. These regional poison centers provide poison information and telephone management and consultation, collect pertinent data, and delivery professional and public education. Cooperation between regional poison centers and poison treatment facilities is crucial. The regional poison information center, in cooperation with local hospitals, should determine the treatment capabilities of
Poison Centers and Bacterial Exposure
3
the treatment facilities of the region and identify and have a working relationship with their analytical toxicology laboratories, emergency departments, critical care wards, medical transportation systems, and extracorporeal elimination capabilities. This should be done for both adults and children. A “region” is usually determined by state authorities in conjunction with local health agencies and health care providers. Documentation of these state designations must be in writing unless a state chooses (in writing) not to designate any poison center or accepts a designation by other political or health jurisdictions. Regional poison information centers should serve a population base of greater than one million people and must receive at least 10,000 human exposure calls per year. The number of certified regional centers in the United States is now under 50. Certification as a regional center requires the following. 1. Maintenance of a 24 hours per day, 365 days per year service. 2. Providing service to both health care professionals and the public. 3. Having available at least one specialist in poison information in the center at all times. 4. Having a medical director or qualified designee, on call by telephone, at all times. 5. Service should be readily accessible by telephone from all areas within the region. 6. Comprehensive poison information resources and comprehensive toxicology information covering both general and specific aspects of acute and chronic poisoning should be available. 7. The center is required to have a list of on-call poison center specialty consultants. 8. Written operational guidelines, which provide a consistent approach to evaluation, follow-up, and management of toxic exposures should be obtained and maintained. These guidelines must be approved in writing by the medical director of the program. 9. There should be a staff of certified professionals manning the phones (at least one has to be a pharmacist or nurse with 2000 hours and 2000 cases of supervised experience). 10. There should be a 24-hour physician (Board Certified) consultation service. 11. The Regional Poison Center shall have an ongoing quality assurance program. 12. Other criteria, determined by the AAPCC, may be established with membership approval. 13. The regional poison information center must be an institutional member in good standing of the AAPCC. Many hospital emergency rooms still maintain a toxicology reference such as the POISINDEX@system to handle routine exposure cases, but rely on regional centers to handle most of the calls in their area.
B. Who Staffs a PoisonCenter? The staffing of poison centers varies considerably. The three professional groups most often involved are physicians, nurses, and pharmacists. Who answers the phones is somewhat dependent on the local labor pool, moneys available, and the types of calls being
Spoerke
4
received. Others used have included students in medically related fields, toxicologists, and biologists. Persons responsible for answering the phones are either certified by the AAPCC or are in the process of obtaining certification. Passage of an extensive examination on toxicology is required for initial certification, with periodic recertification required. Regardless of who takes the initial call, a medical director and the other physician back-up are available. These physicians have specialized training or experience in toxicology and are able to provide in-depth consultations for health care professionals calling a center. 1. Medical Director A poison center medical director should be board certified in medical toxicology or in internal medicine, pediatrics, family medicine, or emergency medicine. The medical director should be able to demonstrate ongoing interest and expertise in toxicology as evidenced by publications, research, and meeting attendance. The medical director must have a medical staff appointment at a comprehensive poison treatment facility and must be involved in the management of poisoned patients. 2. Managing Director The managing director must be a registered nurse, pharmacist, or physician or hold a degree in a health science discipline. The individual should be certified by the American Board of Medical Toxicology (for physicians) or by the American Board of Applied Toxicology (for nonphysicians). They must be able to demonstrate ongoing interest and expertise in toxicology. 3. Specialistsin Poison Information These individuals must be registered nurses, pharmacists, or physicians or be currently certified by the AAPCC as a specialist in poison information. Specialists in poison information must complete a training program approved by the medical director and must be certified by the AAPCC as a specialist in poison information within two examination administrations of their initial eligibility. Specialists not currently certified by the Association must spend an annual average of no less than 16 hours per week in poison centerrelated activities. Specialists currently certified by the AAPCC must spend an annual average of no less than 8 hours per week. Other poison information providers must have sufficient background to understand and interpret standard poison information resources and to transmit that information understandably to both health professionals and the public. 4. Consultants In addition to physicians specializing in toxicology, most centers also have lists of experts in many other fields as well. Poison center specialty consultants should be qualified by training or experience to provide sophisticated toxicology or patient care information in their area(s) of expertise. An infectious disease consult would be appropriate for bacterial infections.
C.What
Types of Calls Are Received?
All types of calls are received by poison centers, most of which are handled immediately; others referred to more appropriate agencies. Which calls are referred depends on the center, its expertise, and the appropriateness of a referral. Below are lists of calls that generally fall into each group. Remember, there is considerable variation between poison
Poison Centers and Bacterial Exposure
5
centers, and if there is any doubt, call the poison center and they will tell you if your case is more appropriately referred. Poison centers do best on calls regarding acute exposures. Complicated calls regarding exposure to several agents over a long period of time that produce nonspecific symptoms are often referred to other medical specialists, to the toxicologist associated with the center, or to an appropriate government agency. The poison center will often follow up on these cases to track the outcome and type of service given. Types of Culls Usually Accepted Drug identification Actual acute exposure to a drug or chemical Actual acute exposure to a biological agent (plants, mushrooms, various animals) Information regarding the toxic potential of an agent Possible food poisonings Types of Calls Often Referred Questions regarding treatment of a medical condition (not poisoning) Questions on common bacterial, viral, or parasitic infections General psychiatric questions Questions regarding proper disposal of household agents, such as batteries, bleach, insecticides Questions regarding use of insecticides (e.g., which insecticide to use, how to use it) unless related to a health issue, for example, a person allergic to pyrethrins wanting to know which product does not contain pryrethrins Records of all calls/cases handled by the center shall be kept in a form that is acceptable as a medical record. The regional poison information center should submit all its human exposure data to the Association’s National Data Collection System. The regional poison information center shall tabulate its experience for regional evaluation on at least an annual basis. It 1983 the AAPCC formed the Toxic Exposure Surveillance System (TESS) from the former National Data Collection System. Currently TESS contains nearly 16.2 million human poison exposure cases. Sixty-five poison centers, representing 181.3 million people, participate in the data collection. The information has various uses to both governmental agencies and industry, providing data for product reformulations, repackaging, recall, bans, injury potential, and epidemiology. The summation of each year’s surveillance is published in the Anlerican Jorlrnal sf Emergency Medicine in late summer or fall.
D. HowCallsAreHandled Most poison centers receive requests for information via the telephone. Calls come from both health care professionals and consumers. Only a few requests are received by mail or in person. These often involve medico-legal or complex cases. Most centers can be reached by a toll-free phone number in the areas they serve, as well as by a local number. Busy centers will have a single number that will ring on several lines. Calls are often direct referrals from the 911 system. In most cases poison center specialists are unable to determine the exact bacterial species, so it is difficult to give specific information. When the species is known (as in some food poisonings), specific information is available. Information about culture sensitivities, rates of infection, or potential sources is usually not available, and such calls are often referred to a regional epidemiologist or regional infectious disease agency.
6
Spoerke
Poison information specialists listen to the caller, recording the history of the case on a standardized form developed by AAPCC. Basic information such as the agent involved, the amount of agent, time of ingestion, symptoms, previous treatment, and current condition are recorded, as well as patient information such as sex, age, phone number, who is with the patient, relevant medical history, and sometimes patient address. All information is considered a medical record, and is therefore confidential. The case is categorized (using various references) as (1) information only, no patient involved, (2) harmless and not requiring follow-up, (3) slightly toxic, no treatment necessary but a follow-up call is given, (4) potentially toxic, treatment given at home and followup given to case resolution, ( 5 ) potentially toxic, treatment may or may not be given at home, but it is necessary for the patient to be referred to a medical facility, or (6) emergency-an ambulance and/or paramedics are dispatched to the scene. Cases are usually followed until symptoms have resolved. In cases where the patient is referred to a health care facility, the receiving agency is notified. This history is relayed, toxic potential discussed, and suggestions for treatment given.
E. WhatReferencesAreUsed? References used also vary from center to center, but virtually all centers use a toxicology system called POISINDEX, which contains lists of products, their ingredients, and suggestions for treatment. The system is compiled using medical literature and medical specialists throughout the world. Only a few bacterial infections (such as food poisonings) are listed in this resource. After accessing the individual bacterial or food-poisoning entry in the database, the physician or poison information specialist is then referred to a treatment protocol that may be for a general class of agents. An unknown skin irritation or potential infection would deserve a consult with an infectious disease specialist. POISINDEX is available on microfiche, as a CD ROM, over a network, or on a mainframe. It is updated every 3 months. Various texts may also be used. It is very difficult to identify bacterial infections based on infortnation given over the phone, so often the assistance of an epidemiologist and infectious disease specialist is used. Some poison centers have more experience with certain types of poisonings than others, so often one center will consult another on an interesting case. These are often more complex cases or cases involving areas within both centers’ regions. A recent trend has been for various manufacturers not to provide product information to all centers via POISINDEX, but to contract with one poison center to provide for poison information services for the whole country. Product information is given to that center only, and cases throughout the country can only be handled by that one center.
F. How Poison Centers Are Monitored for Quality Most poison centers have a system of peer review in place. One person takes a call, another reviews it. Periodic spot review is done by supervisors and physician staff. General competence is assured by certification and re-certification via examination of physicians and poison information specialists.
Poison Centers and Bacterial Exposure
7
G. ProfessionalandPublicEducationPrograms The regional poison information center is required to provide information to the health professionals throughout the region who care for poisoned patients on the management of poisoning. Public education progranx aimed at educating both children and adults about poisoning dangers and other necessary concepts related to poison control should be provided. In the past, several centers provided stickers or logos, such as Officer Ugh, Safety Sadie, and Mr. Yuck, that could be placed on or near potentially toxic substances. While the intent was to identify potentially toxic substances that children should keep away from, the practice has been much curtailed on the new assumption that in some cases the stickers actually attracted children to the products. In the spring of every year there is a poison prevention week during which national attention is focused on the problem of potentially toxic exposures. Many centers run special programs for the public, including lectures on prevention, potentially toxic agents in the home, potentially toxic biological agents, or general first aid methods. Although this week is an important time for poison centers, public and professional education is a yearround commitment. Physicians are involved in medical toxicology rounds, journal clubs, and lectures by specialty consultants. Health fairs, school programs, and various women’s clubs are used to educate the public. The extent of these activities is often determined by the amount of funding received from government, private organizations, and public donations.
H. RelatedProfessionalToxicologyOrganizations ACGIH American Conference of Governmental and Industrial Hygienists Address: Kemper Woods Center; Cincinnati, OH, 45240 Phone: 5 13-742-2020 FAX: 5 13-742-3355 ABAT American Board of Applied Toxicology Address: Truman Medical Center, West; 2301 Holmes St.: Kansas City, MO, 64 108 Phone: 8 16-556-3112 FAX: 8 16-881-6282 AACT American Association of Clinical Toxicologists Address: c/o Medical Toxicology Consultants; Four Columbia Drive; Suite 810: Tampa, FL, 33606 AAPCC American Association of Poison Control Centers Address: 3201 New Mexico Avenue NW: Washington, DC, 20016 Phone: 202-362-7217 FAX: 202-362-8377 ABEM American Board of Emergency Medicine Address: 300 Coolidge Road: East Lansing, MI, 48823 Phone: 5 17-332-4800 FAX: 5 17-333-2234 ACEP American College of Emergency Physicians (Toxicology Section) Address: P.O. Box 61991 1; Dallas, TX, 75261-991 1 Phone: 800-798- 1822 FAX: 214-580-2816
8
Spoerke ACMT American College of Medical Toxicology (formerly ABMT) Address: 777 E. Park Drive; P.O. Box 8820; Harrisburg, PA, 17105-8820. Phone: 717-558-7846 FAX: 717-558-7841 e-mail: [email protected] (Linda L. Koval) ACOEM American College of Occupational and Environmental Medicine Address: 55 West Seegers Road; Arlington Heights, IL, 60005 Phone: 708-228-6850 FAX: 708-228-1856 ACS Association of Clinical Scientists Address: Dept. of Laboratory Medicine; University of Connecticut Medical School; 263 Farmington Ave.; Farmington, CT, 06030-2225 Phone: 203-679-2328 FAX: 203-679-2328 ACT American College of Toxicology Address: 9650 Rockville Pike; Bethesda, MD, 20814 Phone: 30 1-571- l 840 FAX: 30 1-571- 1 852 AOEC Association of Occupational and Environmental Clinics Address: 1010 Vermont Ave., N W , #513; Washington, DC, 20005 Phone: 202-347-4976 FAX: 202-347-4950 e-mail: [email protected] ASCEPT Australian Society of Clinical and Experimental Pharmacologists and Toxicologists Address: 145 Macquarie St.; Sydney N.S.W. 2000, Australia Phone: 61-2-256-5456 FAX: 61-2-252-3310 BTS British Toxicology Society Address: MJ Tucker, Zeneca Pharmaceuticals, 22B 11 Mareside; Alderley Park, Macclesfield; Cheshire, SKlO 4TG, United Kingdom Phone: 0428 65 5041 CAPCC Canadian Association of Poison Control Centers Address: Hopital Sainte-Justine; 3 175 Cote Sainte-Catherine; Montreal, Quebec, Canada H3T1C5 Phone: 5 14-345-4675 FAX: 5 14-345-4822 CSVVA (CEVAP) Center for the Study of Venoms and Venomous Animals Address: UNESP; Alameda Santos; N 647; CEP 01419-901; Sao Paulo, SP, Brazil Phone: 55 011 252 0233 FAX: 55 01 1 252 0200 EAPCCT European Association of Poison Control Centers Address: J. Vale: National Poisons Information Centre: P.O. Box 81898 Dep; N-0034 Oslo, Norway Phone: 47-260-8460
Y-
Poison Centers and Bacterial Exposure
9
HPS Hungarian Pharmacological Society Address: Centeral Research Insttitute for Chemistry; Hungarian Academy of Sciences; H-l525 Budapest; P.O. Box 17; Pusztaszeri ut 59-67, Hungary Phone: 36-1-135-21 12 ISOMT International Society of Occupational Medicine and Toxicology Address: USC School of Medicine; 222 Oceanview Ave., Suite 100; Los Angeles, CA, 90057 Phone: 213-365-4000 JSTS Japanese Society of Toxicological Sciences Address: Gakkai Center Building; 4-16, Yayoi 2-chome; Bunkyo-ku; Tokyo 113, Japan Phone: 3-3812-3093 FAX: 3-3812-3552 SOT Society of Toxicology Address: 1101 14th Street, Suite 1100; Washington, DC., 20005-5601 Phone: 202-37 1- 1393 FAX: 202-37 1-1090 e-mail: [email protected] SOTC Society of Toxicology of Canada Address: P.O. Box 517: Beaconsfield, Quebec; H9W 5V1, Canada Phone: 5 14-428-2676 FAX: 5 14-482-8648 STP Society of Toxicologic Pathologists Address: 875 Kings Highway, Suite 200; Woodbury, NJ, 08096-3172 Phone: 609-845-7220 FAX: 609-853-041 1 SSPT Swiss Society of Pharmacology and Toxicology Address: Peter Donatsch; Sandoz Phmna AG; Toxicologtie 88 1/130; CH4132 Muttenz, Switzerland Phone: 41-6 1-469-5371 FAX: 4 l -6 1-469-6565 WFCT World Federation of Associations of Clinical Toxicology Centers and Poison Control Centers Address: Centre Anti-Poisons; Hopital Edonard Herriot; 5 p1 d'Arsonva1; 69003 Lyon, France Phone: 33 72 54 80 22 FAX: 33 72 34 55 67 1.
International Affiliations
The AAPCC and its members attend various world conferences to learn of toxicology problems and new methods used by these agencies. An especially close relationship has formed between the American and Canadian poison center associations. Once a year the AAPCC and CAPCC hold a joint scientific meeting and invite speakers and other toxicology specialists from throughout the world to attend. Some international affiliated organizations are listed with the North American groups above.
Spoerke
10
J. Toxicology and Poison Center Web Sites
Association of OccupationalandEnvironmentalClinics This group is dedicated to higher standards of patient-centered, multidisciplinary care emphasizing prevention and total health through information sharing, quality service, and collaborative research. Address: [email protected] FingerLakesRegionalPoisonCenter Address: [email protected] Medical/Clinical/Occupational Toxicology Professional Groups A list of primarily U.S. professional groups interested in toxicology, including a description of each group, their address, phone numbers, and contact names. Keyword: poison centers, toxicology Address: http://www.pitt.edu/-martint/pages/motoxorg.htr-t1 PoisonNet A mail list dedicated to sharing information, problem solving, and networking in the areas of poisoning, poison control centers, hazardous materials, and related topics. The list is intended for health care professionals, not the lay public. The moderators do not encourage responses to individual poisoning cases from the public. Keyword(sj: poisoning, poison control centers II. U.S. POISON INFORMATION CENTERS The Poison Control Center telephone numbers and addresses listed below are thought to be accurate as of the date of publication. Poison Control Center telephone numbers or addresses may change. The address and phone number of the Poison Control Center nearest you should be frequently checked. If the number listed does not reach the poison center, contact the nearest emergency service, such as 91 1 or local hospital emergency rooms. The author disclaims any liability resulting from or relating to any inaccuracies or changes in the phone numbers provided below. This information should NOT be used as a substitute for seeking professional medical diagnosis, treatment, and care. ("Indicates a Regional Center designated by the American Association of Poison Control Centers.)
ALABAMA Birmingham Regional Poison Control Center* Children's Hospital of Alabama 1600 Seventh Avenue, South Birmingham, AL 35233- 171 1 (800) 292-6678 (AL only) (205 933-4050
Tuscaloosrr Alabama Poison Control System, Inc. 408 A Paul Bryant Drive, East Tuscaloosa, AL 35401
(800) 462-0800 (AL only) (205) 345-0600
ALASKA Anchorage Anchorage Poison Center Providence Hospital P.O. Box 196603 3200 Providence Drive Anchorage, AK 995 19-6604 (800) 478-3193 (AK only)
Poison Centers and Bacterial Exposure FairbaA-s Fairbanks Poison Center Fairbanks Memorial Hospital 1650 Cowles St. Fairbanks, AK 99701 (907) 456-7 182
Los Angeles Los Angeles County University of Southern California Regional Poison Center" 1200 North State, Room 1107 Los Angeles, CA 90033 (800) 825-2722 (213) 222-3212
ARIZONA
Orange University of California Irvine Medical Center Regional Poison Center" 101 The City Drive, South Route 78 Orange, CA 92668-3298 (800) 544-4404 (CA only) (7 14) 634-5988
Phoenix Samaritan Regional Poison Center" Good Samaritan Medical Center l 130 East McDowell Road, Suite A-5 Phoenix, AZ 85006 (602) 253-3334 Tucson Arizona Poison and Drug Information Center" Arizona Health Sciences Center, Room 1156 1501 N. Campbell Ave. Tucson, AZ 85724 (800) 362-0101 (AZ Only) (602) 626-6016
Richnlond Chevron Emergency Information Center 15299 San Pablo Avenue P.O. Box 4054 Richmond, CA 94804-0054 (800) 457-2202 (510) 233-3737 or 3738 Sacmrnento Regional Poison Control Center" University of California at Davis Medical Center 23 15 Stockton Boulevard Rm HSF-124 Sacramento, CA 95817 (800) 342-3293 (northern CA only) (9 16) 734-3692
ARKANSAS Little Rock Arkansas Poison & Drug Information Center University of Arkansas College of Pharmacy 4301 West Markham, Slot 522 Little Rock, AR 77205 (800) 482-8948 (AR only) (501) 661-6161
CALIFORNIA
11
,
Fresno Fresno Regional Poison Control Center:': Fresno Community Hospital & Medical Center 2823 Fresno Street Fresno, CA 93721 (800) 346-5922 (CA only) (209) 445-1222
Sal1 Diego San Diego Regional Poison Center" University of California at San Diego Medical Center 225 West Dickinson Street San Diego. CA 92013-8925 (800) 876-4766 (.CA only) (619) 543-6000
Sail Francisco San Francisco Bay Area Poison Center" San Francisco General Hospital 1001 Potrero Avenue Rm 1E86 San Francisco, CA 94122 (800) 523-2222 (315) 476-6600
Spoerke
12 Sun Jose Regional Poison Center Santa Clara Valley Medical Center 751 South Bascom Avenue San Jose, CA 95128 (800) 662-9886, 9887 (CA only) (408) 299-5112, 5113, 5114
COLORADO Denver Rocky Mountain Poison Center* 1010 Yosemite Circle Denver. CO 80230 (800) 332-3073 (CO only) (303) 629-1 123
CONNECTICUT Furmington Connecticut Poison Control Center University of Connecticut Health Center 263 Farmington Avenue Farmington, CT 06030 (800) 343-2722 (CT only) (203) 679-3456
DELAWARE Wilmington Poison Information Center Medical Center of Delaware Wilmington Hospital 501 West 14th Street Wilmington, DE 19899 (302) 655-3389
DISTRICT OF COLUMBIA Wushingtorz National Capital Poison Center* Georgetown University Hospital 3800 Reservoir Road, North West Washington, DC 20007 (202) 625-3333
FLORIDA Jucksonville Florida Poison Information Center University Medical Center 655 West Eighth Street Jacksonville, FL 32209 (904) 549-4465 or 764-7667 Tduhussee Tallahassee Memorial Regional Medical Center 1300 Miccosukk Road Tallahassee, FL 32308 (904) 68 1-54 1 1
Tampa Tampa Poison Information Center* Tampa General Hospital Davis Islands P.O. Box 1289 Tampa, FL 33601 (800) 282-3171 (FL only) (813) 253-4444
GEORGIA Atluntu Georgia Regional Poison Control Center* Cerady Memorial Hospital 80 Butler Street South East Box 26066 Atlanta, GA 30335-3801 (800) 282-5846 (GA only) (404) 616-9000 Macon Regional Poison Control Center Medical Center of Central Georgia 777 Hemlock Street Macon, GA 3 1208 (912) 744-1 146, 1100or 1427 Sclvunnnlz Savannah Regional Poison Control Center Memorial Medical Center Inc. 4700 Waters Avenue Savannah, GA 31403 (912) 355-5228 or 356-5228
Poison Centers and Bacterial Exposure HAWAII Honolulu Kapiolani Women’s and Children’s Medical Center 1319 Punahou Street Honolulu, HI 96826 (800) 362-3585, 3586 (HI only) (808) 941-4411
IDAHO Boise Idaho Poison Center St. Alphonsus Regional Medical Center 1055 North Curtis Road Boise, ID 83706 (800) 632-8000 (ID only) (208) 378-2707
ILLINOIS Chicago Chicago and NE Illinois Regional Poison Control Center Rush Presbyterian-St. Luke’s Medical Center 1653 West Congress Parkway Chicago, IL 60612 (800) 942-5969 (Northeast IL only) (312) 942-5969
Normal Bromenn Hospital Poison Center Virginia at Franklin Normal, IL 61761 (309) 454-6666 Springfield Center and Southern Illinois Poison Resource Center St. John’s Hospital 800 East Carpenter Street Springfield, IL 62769 (800) 252-2022 (IL only) (217) 753-3330 Urbana National Animal Poison Control Center
13 University of Illinois Department of Veterinary Biosciences 2001 South Lincoln Avenue, 1220 VMBSB Urbana, IL 61801 (800) 548-2423 (subscribers only) (217) 333-2053
INDIANA Indianapolis Indiana Poison Center* Methodist Hospital 1701 North Senate Boulevard Indianapolis, IN 46202-1367 (800) 382-9097 (317) 929-2323
IOWA Des Moines Variety Club Drug and Poison Information Center Iowa Methodist Medical Center 1200 Pleasant Street Des Moines, IA 50309 (800) 362-2327 (5 15) 24 1-6254
Iowa City University of Iowa Hospitals and Clinics 200 Hawkins Drive Iowa City, IA 52246 (800) 272-6477 or (800) 362-2327 (IA only) (319) 356-2922 Sioux City St. Luke’s Poison Center St. Luke’s Regional Medical Center 2720 Stone Park Boulevard Sioux City, IA 51 104 (800) 352-2222 (IA, NE, SD) (712) 277-2222
KANSAS Kansas City Mid America Poison Center
Spoerke
14
Kansas University Medical Center 39th and Rainbow Boulevard Room B-400 Kansas City, KS 66160-7231 (800) 332-6633 (KS only) (913) 588-6633 Topeka Stormont Vail Regional Medical Center Emergency Department 1500 West 10th Topeka, KS 66604 (913) 354-6100
Wichita Wesley Medical Center 550 North Hillside Avenue Wichita, KS 67214 (316) 688-2222
KENTUCKY Ft. Tllor?m Northern Kentucky Poison Information Center St. Luke Hospital 85 North Grand Avenue Ft. Thomas, KY 41075 (513) 872-51 11 Louislille Kentucky Poison Control Center of Kosair Children’s Hospital 3 15 East Broadway P.O. Box 35070 Louisville, KY 40232 (800) 722-5725 (KY only) (502) 589-8222
LOUISIANA Houma Terrebonne General Medical Center Drug and Poison Infommation Center 936 East Main Street Hourna. LA 70360 (504) 873-4069
Monroe Louisiana Drug and Poison Information Center
Northeast Louisiana University School of Pharmacy. Sugar Hall Monroe, LA 7 1209-6430 (800) 256-9823 (LA only) (3 18) 362-5393
MAINE Portland Maine Poison Control Center Maine Medical Center 22 Bramhall Stree t Portland. ME 04102 (800) 442-6305 (ME only) (207) 87 1-2950
MARYLAND Baltimore Maryland Poison Center* University of Maryland School of Pharmacy 20 North Pine Street Baltimore, MD 21201 (800) 492-2414 (MD only) (410) 528-7701
MASSACHUSETTS Bostoit Massachusetts Poison Control System* The Children’s Hospital 300 Longwood Avenue Boston, MA 031 15 (800) 682-9211 (MA only) (617) 232-2120 or 735-6607
MICHIGAN Ahian Bixby Hospital Poison Center Emma L. Bixby Hospital 8 1 Riverside 8 Avenue Adrian, MI 49331 (5 17) 263-2412
Poison Centers and Bacterial Exposure Detroit Poison Control Center Children's Hospital of Michigan 3901 Beaubien Boulevard Detroit, MI 48201 Outside metropolitan Detroit; (800) 4626642 (MI only) (3 13) 745-57 11 G r a d Rapids Blodgett Regional Poison Center 1840 Wealthy Street, South East Grand Rapids, MI 49506 Within MI: (SOO) 632-2727
Kalmmcoo Bronson Poison Information Center 252 East Love11 Street Kalamazoo, MI 49007 (800) 442-41 12 616 (MI only) (6 16) 34 1-6409
15
Hattiesburg Forrest General Hospital 400 S. 28th Avenue Hattiesburg, MS 39402 (601) 288-3235
MISSOURI Kansas City Poison Control Center Children's Mercy Hospital 2401 Gillham Road Kansas City, MO 64108-9898 (8 16) 234-3000 or 234-3430 St. Louis Regional Poison Center* Cardinal Glennon Children's Hospital 1465 South Grand Boulevard St. Louis, MO 63104 (800) 392-9 111 (MO only) (800) 366-8888 (MO, west IL) (314) 772-5200
MINNESOTA MONTANA Minrleapolis Hennepin Regional Poison Center* 701 Park Avenue South Minneapolis, MN 554 15 (612) 347-3144 (612) 347-3141 (Petline) St. Paul Minnesota1 Regional Poison Center* St. Paul-Ramsey Medical Center 640 Jackson Street St. Paul, MN 55101 (800) 222-1222 (MN only) (612) 221-2113
MISSISSIPPI
Jackson University of Mississippi Medical Center 2500 North State Street Jackson, MS 39216 (601) 354-7660
Dernver Rocky Mountain Poison and Drug Center Denver, CO 80204 (800) 525-5042 (MT only)
NEBRASKA Omaha The Poison Center* Children's Memorial Hospital 8301 Dodge Street Omaha, NE 681 14 (800) 955-9119 (WY, NE) (402) 390-5400, 5555
NEVADA Las Vegas Humana Hospital-Sunrise* 3186 Maryland Parkway Las Vegas, NV 89109 (800) 446-6179 (NV only)
Spoerke
16
Reno Washoe Medical Center 77 Pringle Way Reno, NV 89520 (702) 328-4144
NEW HAMPSHIRE Lebanon New Hampshire Poison Center Dartmouth-Hitchcock Medical Center 1 Medical Center Drive Lebanon, NH 03756 (800) 562-8236 (NH only) (603) 650-5000
NEW JERSEY Newark New Jersey Poison Information and Education Systems" 201 Lyons Avenue Newark, NJ 071 12 (800) 962-1253 (NJ only) (973) 923-0764 Plzillipsburg Warren Hospital Poison Control Center 185 Rosberg Street Phillipsburg, NJ 08865 (800) 962-1253 (908) 859-6768
NEW MEXICO Albuquerque New Mexico Poison and Drug Information Center* University of New Mexico Albuquerque, NM 87 131 (800) 432-6866 (NM only) (505) 843-251 1
NEW YORK Buffalo Western New York Poison Control Center Children's Hospital of Buffalo 219 Bryant Street Buffalo, NY 14222 (800) 888-7655 (NY only) (7 16) 878-7654 Mineola Long Island Regional Poison Control Center" Winthrop University Hospital 259 First Street Mineola, NY 1 1501 (516) 542-2323, 2324, 2325 New York City New York City Poison Control Center* 455 First Avenue, Room 123 New York, NY 10016 (212) 340-4494 (213) 764-7667 Nyack Hudson Valley Regional Poison Center Nyack Hospital 160 North Midland Avenue Nyack, NY 10920 (800) 336-6997 (NY only) (914) 353-1000 Rochester Finger Lakes Regional Poison Control Center University of Rochester Medical Center 601 Elmwood Avenue Rochester, NY 14642 (800) 333-0542 (NY only) (716) 275-5151 Syracuse Central New York Poison Control Center SUNY Health Science Center 750 E. Adams Street Syracuse, NY 13210 (800) 252-5655 (3 15)476-4766
NORTH CAROLINA Ashville Western North Carolina Poison Control Center
c
Poison Centers and Bacterial Exposure Memorial Mission Hospital 509 Biltmore Avenue Ashville, NC 28801 (800) 542-4225 (NC only) (704) 255-4490 or 258-9907 Charlotte Carolinas Poison Center Carolinas Medical Center 100 Blythe Boulevard Charlotte, NC 28232-2861 (800) 848-6946 (704) 355-4000 Durham Duke Regional Poison Control Center P.O. Box 3007 Durham, NC 277 I O (800) 672-1697 (NC only) (919) 684-8111 Greensboro Triad Poison Center Moses H. Cone Memorial Hospital 1200 North Elm Street Greensboro, NC 2740 1- 1020 (800) 953-4001 (NC only) (919) 574-8105 Hickon Catawba Memorial Hospital Poison Control Center 810 Fairgrove Church Road, South East Hickory, NC 28602 (704) 322-6649
NORTH DAKOTA Fargo North Dakota Poison Center St. Luke’s Hospital 720 North 4th Street Fargo, ND 58 122 (800) 732-2200 (ND only) (701) 234-5575
OHIO Akron Akron Regional Poison Center
17 281 Locust Street Akron, OH 44308 (800) 362-9922 (OH only) (2 16) 379-8562 Canton Stark County Poison Control Center Timken Mercy Medical Center 1320 Tinlken Mercy Drive, North West Canton, OH 44667 (800) 722-8662 (OH only) (216) 489-1304 Cincinnati South West Ohio Regional Poison Control System and Cincinnati Drug and Poison Information Center* University of Cincinnati College of Medicine 231 Bethesda Avenue ML #l44 Cincinnati, OH 45267-0144 (800) 872-5 11 1(Southwest OH only) (513) 558-51 11 Cleveland Greater Cleveland Poison Control Center 2074 Abington Road Cleveland, OH 44106 (2 16)23 1-4455 Col~u~b~w Central Ohio Poison Center* 700 Children’s Drive Columbus. OH 43205 (800) 682-7625 (OH only) (614) 228-1323 Dayton West Ohio Regional Poison And Drug Information Center Children’s Medical Center One Children’s Plaza Dayton, OH 45404- 18 15 (800) 762-0727 (OH only) (5 13) 222-2227
Lorain County Poison Control Center Lorain Community Hospital 3700 Kolbe Road Lorain, OH 44053 (800) 821-8972 (OH only) (216) 282-2220
Spoerke
18
Sandusky Firelands Community Hospital Poison Information Center 1101 Decatur Street Sandusky. OH 44870 (419) 626-7423
J
PENNSYLVANIA
Toledo Poison Information Center of Northwest Ohio Medical College of Ohio Hospital 3000 Arlington Avenue Toledo, OH 49614 (800) 589-3897 (OH only) (419) 381-3897
Hershey Central Pennsylvania Poison Center* Milton Hershey Medical Center Pennsylvania State University P.O. Box 850 Hershey, PA 17033 (800) 521-61 10 (717) 531-6111
Youngstown Mahoning Valley Poison Center St. Elizabeth Hospital Medical Center 1044 Belmont Avenue Youngstown, OH 44501 (800) 426-2348 (OH only) (2 16) 746-2222
Lrrncnster Poison Control Center St. Joseph Hospital and Health Care Center 250 College Avenue Lancaster, PA 17604 (717) 299-4546
Znnesville Bethesda Poison Control Center Bethesda Hospital 2951 Maple Ave Zanesville, OH 4370 1 (800) 686-4221 (OH only) (614) 454-4221
OKLAHOMA
l
(800) 452-7165 (OR only) (503) 494-8968
Oklalzom City Oklahoma Poison Control Center Children's Memorial Hospital 940 Northeast 13th Street Oklahoma City, OK 73104 (800) 522-4611 (OK only) (405) 27 1-5454
OREGON Portland Oregon Poison Center Oregon Health Sciences University 3 181 South West Sam Jackson Park Road Portland, OR 97201
Philadelphia Philadelphia Poison Control Center" One Children's Center 34th and Civic Center Boulevard Philadelphia, PA 19104 (215) 386-2100 Pittsbtqh Pittsburgh Poison Center* One Children's Place 3705 Fifth Avenue at DeSoto Street Pittsburgh, PA 15213 (4 12) 68 1-6669 Willinrmport The Williamsport Hospital Poison Control Center 777 Rural Avenue Williamsport, PA 17701 (717) 321-2000
RHODE ISLAND Providence Rhode Island Poison Center* 593 Eddy Street Providence, RI 02903 (401) 444-5727
i
19
Poison Centers and Bacterial Exposure SOUTH CAROLINA Chrrrlotte Carolinas Poison Center Carolinas Medical Center 1000 Blythe Boulevard Charlotte. NC 28232-2861 (800) 848-6946
Columbirt Palmetto Poison Center University of South Carolina College of Pharmacy Columbia, SC 29208 (800) 922-1 117 (SC only) (803) 765-7359
SOUTH DAKOTA Aberdeen Poison Control Center St. Luke's Midland Regional Medical Center 305 S. State Street Aberdeen, SD 57401 (800) 592-1889 (SD. MN, ND. WY) (605) 622-5678
Rapid City Rapid City Regional Poison Control Center 835 Faimont Boulevard P.O. Box 6000 Rapid City, SD 57709 (605) 341-3333
Sioux Falls McKennan Poison Center McKennan Hospital 800 East 21st Street P.O. Box 5045 Sioux Falls, SD 571 17-5045 (800) 952-0123 (SD only) (800) 843-0505 (IA, MN, NE) (605) 336-3894
TENNESSEE Knoxville Knoxville Poison Control Center
University of Tennessee Memorial Research Center and Hospital 1924 Alcoa Highway Knoxville, TN 37920 (6 15) 544-9400 hlerrrphis Southern Poison Center, Inc. Lebanheur Children's Medical Center 848 Adams Avenue Memphis, TN 38103-2821 (901) 528-6048
Nashville Middle Tennessee Regional Poison Center, Inc. 501 Oxford House 1161 21st Avenue South B-101VUII Nashville, TN 37232-4632 (800) 288-9999 (TN only) (615) 322-6435
TEXAS Conroe Montgomery County Poison Information Center Medical Center Hospital 504 Medical Center Blvd.
TX 77304 (409) 539-7700 Dallas North Central Texas Poison Center* Parkland Memorial Hospital 5201 Harry Hines Boulevard P.O. Box 35926 Dallas, TX 75235 (.goo) 441 -0040 (TX only) (214) 590-5000 El Pnso El Paso Poison Control Center Thomas General Hospital 48 15 Alameda Avenue El Paso, TX 79905 (915) 533-1244
Galveston Texas State Poison Control Center University of Texas Medical Branch 8th and Mechanic Street
!
Spoerke
20 Galveston, TX 77550-2780 (800) 392-8548 (TX only) (7 13) 654-1701 (Houston) (409) 765-1420 (Galveston)
Lubbock Methodist Hospital Poison Control 3615 19th Street Lubbock, TX 7941 3 (806) 793-4366
WASHINGTON Seattle Washington Poison Center P.O. Box 5371 Seattle, WA 98105-0371 (800) 732-6985 (within WA) (206) 526-2121
WEST VIRGINIA UTAH Salt Lake City Utah Poison Control Center* Intermountain Regional Poison Control Center 410 Chipeta Way, Suite 230 Salt Lake City, UT 84108 (800) 456-7707 (UT only) (801) 581-2151
Charleston West Virginia Poison Center’# West Virginia University 3 110MacCorkle Avenue, South East Charleston, WV 25304 (304) 348-4211 (800) 632-3625 (WV only) Parkersburg St. Joseph’s Hospital Center 19th Street and Murdoch Avenue Parkersburg, WV 26101 (304) 424-4222
VERMONT WISCONSIN Burlington Vermont Poison Center Medical Center Hospital of Vermont 1 1 1 Colchester Avenue Burlington, VT 05401 (802) 658-3456
VIRGINIA Charlottesville Blue Ridge Poison Center* University of Virginia Health Sciences Center Box 67 Charlottesville, VA 22901 (800) 451-1428 (VA only) (804) 924-5543 Richmond Virginia Poison Center Virginia Commonwealth University MCV Station Box 522 Richmond, VA 23298-0522 (800) 552-6337 (VA only) (804) 786-9123
Madison Regional Poison Control Center University of Wisconsin Hospital 600 Highland Avenue Madison, W1 53792 (608) 262-3702 Milwaukee Poison Center of Eastern Wisconsin Children’s Hospital of Wisconsin 9000 West Wisconsin Avenue P.O. Box 1997 Milwaukee, W1 53201 (414) 266-2222
WYOMING Omaha The Poison Center* Children’s Memorial Hospital 8301 Dodge Street Omaha, NE 681 14 (800) 955-9119 (WY, NE) (402) 390-5400, 5555
Poison Centers and Bacterial Exposure
21
REFERENCES 1. Harrison, D. L., Draugalis, J. R., Slack, M. K.. and Langly, P. C. (1996). Cost effectiveness of Regional Poison Control Centers. Arch. Intern. Med. 156:2601-2608. 2. CPSC. CPSC Chairman Ann Brown Suggests Information Technology Studyto Support Work of Poison Centers. News Release #94-047, Tuesday March 15, 1994.
This Page Intentionally Left Blank
2 Bacterial Biota (Flora) in Foods James M. Jay Uitirlersityoj-Nevada Las Vegas. Las Vegm, NeIlacIa
I. Introduction 23 A. Soils 24 B. Waters 24 C. Human and animal sources 11. CommonFoodborneGenera A. Gram positives B. Gram negatives 28
24 24
24
111. DetectionandEnumeration 28 A. Choice of incubationtimeandtemperature B. Choice of culture medium 28 C. Selective and differential media 39 D. Ratio of numbers 29 IV. PrevalenceinFoods 29 A. Ready-to-use vegetables B. Spices 30 C. Meat products 31 References
1.
28
29
32
INTRODUCTION
Because natural foods come from either a soil or water environment, they may beexpected to contain at least some of the bacteria that are common to these environments unless steps have beentaken to effect their destruction. In addition, some, especially human pathogens, enter the food supply from infected animals and human handlers. Fortunately, only a relatively small percentage of known soil bacteria can be found on foods of plant or animal origin. In the case of foods from fresh and ocean waters, a higher percentage of the bacterial biota from these environments may be associated with such foods due in large part to a less diverse biota in contrast to that of soils.
23
24
Jay
A. Soils The number of bacteria in the top layer of a rich farm soil is typically around one billion per gram. Numbers differ for soils under active cultivation or where cover crops have been plowed under in contrast to those that have not been disturbed for several years. In the former, the bacterial biota is dominated by zymogeneous types that are active degraders of simple plant constituents, and they consist of both gram-positive and gram-negative heterotrophic genera. These bacteria are typically on the surface of plants when they are plowed under, and they are the ones most likely to be found on plant-based foods. Foodborne pathogens such as the proteolytic strains of Clostridium botulinum and Bacillus cereus are soil bacteria. The bacterial biota of stable soils is dominated by the autochthonous or indigenous types that consist of humus feeders. This group is dominated by gram positives such as the mycobacteria and nocardioforms, and they are slow growers. Also, stable soils are populated by autotrophic bacteria that are antagonized by utilizable organic matter. While any or all of these types may appear transiently on plant products, they lack the capacity to adhere, and importantly, they are slow growers or are outcompeted by the zymogeneous biota noted above.
B. Waters The bacterial biota of aquatic-based foods is less diverse than that of soils because of the general lack of humus feeders and chemoautotrophic types in the water column. In the case of fresh waters, the nonphotosynthetic biota is quite similar to that of the surrounding soils due to rain run-off. The significant difference between the biota of fresh and sea waters is that the latter consists essentially only of gram-negative bacteria. Further, except for marine coastal zones, where numbers of nonphotosynthetic bacteria may be high, the biota of the open oceans is quite low. The salinity and lower temperatures of marine waters allow for the existence and growth of halophilic bacteria such as Photobacterium spp. Also, true psychrophilic bacteria are found along with larger populations of psychrotrophic types. Foodborne pathogens such as Vibrio cholerae and Vibrio paralzaemolyticus are found in marine and estuarine waters.
C. Human andAnimalSources Although soil and water may be viewed as the primary and original habitats of all bacteria, some species have lost the capacity to grow and multiply in either of these environments (e.g., whooping cough and syphilis agents). Others maintain the capacity to live in these environments while acquiring the capacity to live as parasites or conlmensuals on human or animal hosts (e.g., the two Vibrio spp. noted above). The largest number of foodborne pathogens are most often found in or on human and animal hosts. Shigella spp. are found only in humans, whereas Salmonella spp. are found in humans and other vertebrate animals.
II. COMMON FOODBORNEGENERA A.
Gram Positives
The gram-positive bacteria most often recovered from foods are summarized in Table 1. The 40 genera listed represent those that may be recovered from fresh, fermented, and
Biota
Bacterial Foods
(Flora) in
25
Table 1 The Most Common Genera of Gram-positive Bacteria Found in Fresh and Processed Foods ~
Group/genus Spore-formers Alicyclobacillus Aneurinibacillus Bacillus Brevibacillus Clostridium Paenibacillus Sporolactobacilhrs Lactics Carnobacterium Enterococcus Lactococcus Lactobacillus Luctospltaem Leuconostoc Oenococcus Pediococcus Streptococcus Tetragenococcus Vagococcro Weissella Corynefomx & related Arthrobacter Colvnebacteriurrr Bifidobncterium Brevibacterium Caseobacter Cellulonlorms Microbacterium Propionibacterium Miscellaneous groups Aerococclrs Brochothrix Deirlococcus Ensipe1othri.x Halobacterium Hdococcus Kocuria Listeria Micrococcus Mycobacterium Plnnococc1rs Rubrobacter Staphylococclts
Morphology
~~
~
~~~
~
Common sources/conunents
R R R R R R R
Canned fruits, juices Same as Bacillus Soils, air, utensils Same as Bacillus Soil, air, utensils Same as Bacillus Soil, chicken feed
R C C R C C C C C C C R
Meats, poultry, vegetation Feces, water, vegetation Raw milk Gastrointestinal tract, vegetation Rumen, anaerobic sludge Vegetation, sugar refineries Grapes, wines Vegetation Raw milk Pickling brines Water, fish, feces Processed meats, vegetation
R R R R R R R R
Soils (rare in foods) Decaying organic matter Raw milk, feces Certain cheeses Meat carcasses Vegetation Vegetation Vegetation, cheeses
C R C R R C C R C R C R C
Raw milk Processed meats Extremely radiation resistant Cattle, raw milk Salt water Salt water Same as Micrococcus Vegetation, zoonotic transmission Air, dust, utensils, handlers Raw milk of infected dairy herds Sea water, seafoods (a halophile) Extremely radiation resistant Nasals/Sl;in of handlers; hides
26
Jay
processed foods. In no single edible food product would one find all of the bacteria listed in Table 1. Fresh foods such as ground meats may be expected to contain 10-20 of the genera noted. The gram positives are listed under three groups based on morphological and physiological features. The endospore-forming bacteria are widespread in nature, and some of these may be found in any fresh food product. As soil organisms, their primary entry into foods is via dust and soil contamination of ingredients and utensils. The genus Alicyclobacillus is of significance as a cause of spoilage of canned fruits and juices. Foodborne pathogens are found in the genera Bacillus and Clostridium. The lactic acid bacteria are very common in nature on plants and plant products, especially fruits. They are responsible for a number of fermented food products including sauerkraut, pickles, and cheeses. Although common on vegetation, the enterococci are also common in the gastrointestinal tract of mammals. The presence and numbers of enterococci in human feces led to their use as indicators of fecal pollution of waters and also to their use as indicators of sanitary quality of some fresh foods. In general, the lactics are as a group the most beneficial of all bacteria to humans relative to the food supply. In addition to the fermented foods they produce, their production of bacteriocins and their overall antagonism of foodborne pathogens have led to their designation and use as “protective cultures” (1). On the other hand, some lactics have been incrinlinated in human infections, with the lactobacilli, pediococci, and enterococci being most often reported (2,3). Coryneform bacteria are a somewhat loosely defined group of organisms that are common in soils. The term “coryneform” refers to a microscopic picture of young cells that display a club or wedge shape and undergo postsnapping division that leads to palisade formations. The cells of older cultures of coryneforms shorten and appear as coccoids, and the genera do not produce spores. As originally defined, they consisted of four genera: Arthrobacter, Cellulonzor~as,Co~vnebacterirrm,and Microbacterirrm (4). The other coryneform and coryneform-like genera in Table 1 share microscopic features similar to the four genera noted. Some of the coryneform genera are involved in food fermentations (e.g., Brevibncterizm and Propiorzibcrcterium), and the genus Bijidobacterium is very important in the gastrointestinal tract of infants. Some bifido strains are used as food starter cultures (5). Because of their association more with human feces than with those of other animals, the bidifobacteria have been suggested as indicator organisms for human fecal pollution of waters (for review, see Ref. 6). The miscellaneous group in Table 1 consists of 13 genera of bacteria that are phylogenetically diverse. The genus Brochotlzrix is significant on processed meats, especially those stored under CO., atmospheres. More information on the gram-positive biota of meats can be obtained from Holzapfel (7). The deinococci and rubrobacteria are important only in irradiated foods because of their extreme radiation resistance. The micrococci are widespread in nature, and some species along with Kocuricr may be found in a number of food products. Significant foodborne pathogens are found among the staphylococci, and they are discussed further in a later chapter. Listeria nzonocytogerzes causes listeriosis in animals and humans and is closely related to Ensipelothri.~rhusioynthiae, which causes erysipeloid in animals. Mycobncterimz pcrratlrberc.clrlosisis of concern in pasteurized milk as the possible cause of Crohn’s disease.
Bacterial Biota (Flora) in Foods
27
Table 2 Most Common Gram-Negative Bacteria Found on Fresh and Processed Foods Group/genus
Morphology
Enterobacteriaceae Arizona Cedecea Citrobacter Edwardsiella Emir1 in Enterobacter Escherichicl Hajk ia Kluyvera Klebsiella Morganella Obsermbacterium Parltoea Proteus Providencin Salmonella Serratia Shigella Yersinia Miscellaneous Acetobacter Acirzetobacter Aerontorras Alcaligenes Alteromonrts Arcobacter Bacteroides Burkholderia Brucella Caityylobncter Chrorlzobacteritlill Deiilobacter Desu~otomaculurrl Flavobucteriunl Glucoizobucter Megasphaera Moraxella Pectinatus Photobacterium Plesiornonas Pseudonlonas Psycltrobncter Ralstonin Shewanella Vibrio Xnnthonronas Zymophilrrs ~~
~~
~~
C = Coccus; R = rod; S = spiral.
Common sources/comments
R R R R R R R R R R R R R R R R R R R
Feces, water Feces, water Vegetation, water, feces Feces, water Vegetation. feces, water Vegetation, water Feces, water Water, feces, meats Feces Vegetation, feces, water Feces, vegetation, water Beer wort Feces, spoiled meats Feces, water, fresh foods Similar to Proteus Poultry, feeds, other animals Vegetation, waters, feces Human feces Water, fresh meats, feces
R R R R R R R R R R R R R R R C R R R R R R R R S R R
Fresh apples, cider mills Water, vegetation, refrig. foods Water, feces. seafoods Water, vegetation, feces Marine waters. seafoods Hogs, cattle, sheep, vegetation Feces, polluted waters (anaerobic) Plant pathogens, vegetation Cows, sheep, hogs; raw milk Raw poultry, milk, bovines Vegetation, raw milk Extremely radiation resistant Canneries, canned foods Vegetation Grapes, honey bees, beerdwines Spoiled beer Soil, water, refrig. foods Spoiled beer Sea water, spoiled fish Water, feces Soil, water, refrig. fresh foods Meat, fish, poultry Tomato wilt disease Rancid butter, marine/fresh waters Waters, seafoods, feces Contain numerous plant pathogens In beer yeasts
28
Jay
B. Gram Negatives Members of the family Enterobacteriaceae constitute the largest phylogenetic group of foodborne gram-negative bacteria, and 19 genera are listed in Table 2. Any or all of these organisms may be found in animal feces under appropriate conditions, hence the common designation “enterics.” All of the salmonellae and shigellae are human pathogens, and some species/strains of Escherichia, Klebsiella, and Yersinia are human pathogens. The four coliform genera are among the enterics (Citrobacter,Enterobacter, Escherichia, and Klebsiella). The genera Citrobcrcter,Enterobacter, and Emjinia are well-known plant inhabitants. Common as food spoilage organisms are some Citrobacter, Hafizia, Puntoetr, Proteus, and Serrcrtia species since they contain psychrotrophic strains. The miscellaneous group in Table 2 includes some that cause human illness (Brucella, Campylobncter, and Vibrio) and others that are well-known plant pathogens (Pseudomonas, Ralstonia, and Xmthomonas). The bacteria that are most conspicuous on refrigerator-spoiled fresh foods such as ground meats are in the genera Acinetobacter, Alcaligenes, Alteronzonns, Moraxella, Psychrobncter, and Pseudornonas, with the latter genus being the single most important of the psychrotrophic spoilage bacteria. More information on the foodborne bacterial pathogens can be obtained from the ICMSF monograph (8) and from the specific chapters that follow in this volume.
111.
DETECTION ANDENUMERATION
On the surface, the detection and enumeration of bacteria in foods could be carried out in the same way as for nonfood products. However, the experiences of many over the years indicate that it is not that simple. Factors that set a food analysis apart from that of, say, a blood or spinal fluid analysis are discussed in the following sections.
A.
Choice of Incubation Time and Temperature
A typical fresh food product contains some bacteria that grow well between 4 and 7°C and some that can grow between 42 and 46°C. Those that grow at the lower temperatures generally require longer incubation times. When colony-forming units (CFU) are to be enumerated, the long incubation can lead to overgrowth of the slow growers by some of the fast growers and spreaders such as Bacillus and Proteus spp. All too often, individuals who are not experienced food microbiologists incubate plates at 37°C and expect results in 18-24 hours. With the exception of Canlpylobacter spp., all foodborne bacteria of significance grow well at 30-32”C, and incubation time should be extended to 48 hours.
B. Choice of Culture Medium Because of the high heterogeneity of the foodborne bacterial biota, electing the proper culture medium is not easy. Those without experience with foodborne bacteria tend to use media that are too rich or complex. Such media tend to allow for overgrowth of some bacteria during prolonged incubations. Before carrying out bacterial determinations on foods, it is strongly recommended that a standard reference method be used: the two recommended reference works are the FDA Bacteriological Amlytical Mar2ual (commonly referred to as BAM) (9) and the Compendium of Methods-for the Microbiological
Bacterial Foods Biota (Flora) in
29
Examination of Foods, referred to as the Compendium (10). Both works contain recommended culture methods for all bacteria of significance in foods in addition to some bioassay, molecular, and genetic methods of analysis for cells and/or their products.
C.SelectiveandDifferentialMedia The use of a selective plating medium such as Macconkey agar to recover gram-negative bacteria from, say, spinal fluid is distinctly different from its use to recover gram-negative organisms from a food product such as fresh ground beef. In the case of spinal fluid, the bacterial biota is neither as heterogeneous nor as large in quantity as for ground beef. Further, when low dilutions of meat homogenates are planted onto the surface of selective media, meat particles and constituents are known to exert a neutralizing effect on the selective agents. The small food particles can serve as microniches for the growth of some bacteria that would otherwise be inhibited by the selective agent.
D. Ratio of Numbers Recovering bacterial pathogens from foods is often like trying to find the proverbial needle in a haystack. When human stool cultures are examined for salmonellae where salmonellosis is suspected, typically these organisms are at levels of 10s-108 CFU/g of stool, although lower numbers are sometimes seen. In this case, the ratio of salmonellae to nonsalmonellae is such as to allow for direct enumeration and isolation using appropriate selective media. The ratio of salmonellae to the background bacterial biota of a product such as ground beef is such that direct enumeration and isolation are all but impossible since the usually low numbers of salmonellae are typically overwhelmed by the much larger background biota. Thus, nonselective enrichments are necessary, and they are widely used in the food microbiology laboratory. The foregoing deals with the determination of viable numbers of bacteria, variously referred to as aerobic plate count (APC) or standard plate count (SPC) of organisms per gram or CFU. Rapid nonculture methods have been developed for all significant foodborne pathogens, and the references noted above are excellent sources. Although these methods have proven to beinvaluable, there continues to be a valuable place for the classical culture methods. Nonculture methods typically detect both viable and nonviable cells, and when it is desirable to know if viable cells are present, a culture method is required. The impact of molecular and other nonculture detection methods on bacteriological assessments of foods has been reviewed and discussed by Feng (1 l).
IV. PREVALENCE IN FOODS A.
Ready-to-UseVegetables
In consideration of the high incidence and prevalence of bacteria in the soils and waters that basic food products grow in, it should not besurprising to find relatively high numbers on products that have not been subjected to bactericidal treatments. The prevalence of
30
Jay
Table 3 Log,, Mean APC/g of a Variety of Ready-to-Use Vegetable Products from Four Different Countries Product
N
Means
Location
Ref.
Salad mix Parsley. fresh Parsely, frozen Lettuce Salad mix Carrot sticks Celery Chopped lettuce Cole slaw Radishes Bean sprouts Lettuce (hydroponic)
2 11 14 24 3 15 4 4 1 1 2 34
7.85 6.67 5.26 7.82. 7.92 5.60 4.50 52 0 7 .OO 6.04 7.76 7.35
England Germany Germany Italy Italy United States United States United States United States United States United States United States
12 13 13 14 15 16 16 16 17 17 18 19
One sample/month over a 23-month period.
bacteria in a variety of ready-to-use vegetable products is presented in Table 3, and it can be seen that numbers between 1 and 10 million/g are common. Additional ready-to-use vegetables are presented in Table 4 where on day 0 the APC means/g were between 10,000 and 1 million. However, after these products were held for 4 days at 4"C, the APCs of most increased to > 1 milliodg (20). Typically, the bacterial biota of products of this type consists of organisms on the plant products during their growth (e.g., many of the lactic genera and gram-negative bacteria such as Envinin and Enterobacter). Also, a significant part of the biota would be expected to come from handlers, cutting and processing equipment, storage containers, and the air.
Spices B. Numbers of bacteria per gram of retail-store spices are known to be high, and some examples are presented in Table 5 of spices examined in Austria (21). Of the 10 products presented, rosemary had the lowest APC, with a mean of 4.83 log/g, and black pepper Table 4 Log,,, Aerobic Plate Counts/g on Fresh-Cut, Ready-to-Use Vegetables Stored at 4"Ca Vegetables Chopped lettuce Salad mix Cauliflower florets Sliced celery Cole slaw mix Carrot sticks Broccoli florets Green peppers
APC day 0
APC day 4
4.85 5.35 4.82 5.67 5.13 5.13 5.58 5.99
5.63 6.05 5.45 6.59 6.95 6.27 6.59 7.22
The products had a recommended shelf life of 7 days at 4°C. Source: Ref. 20. a
Bacterial Biota (Flora) in Foods
31
Table 5 Log,,, Numbers/g of Microorganisms Found in Some Spices in Vienna, Austria Allspice Caraway Chili powder China spice curry Ginger Nutmeg Black pepper Rosemary Thyme
6.89 6.04 5.91 7.42. 6.79 7.00 5.53 7.34 4.83 6.7 1
Source: Ref. 21
and China spice were the highest, with 7.34 and 7.42/g, respectively. Typically, products of this type contain large numbers of molds, especially black pepper. Among the bacteria, sporeformers are usually abundant as well as gram-positive cocci such as the micrococci. Gram-negative bacteria are either absent or quite low in numbers in products of this type if they have been properly dried and packaged. Some spices (e.g., rosemary and thyme) are also known to possess some antibacterial properties.
C. Meat Products The bacterial biota of fresh ground meats is reflective of carcass contamination during slaughtering, processing, and storage. Since such products are kept at refrigerator temperatures, it is not surprising to find large numbers of psychrotrophic bacteria. Mean log,, APCs/g of ground beef tested in 1914 and in 1994 are presented in Table 6. Retail store fresh ground beefmay be expected to contain >lo5 CFU/g. This is true for the data presented in Table 6 prior to 1994 where the USDA reported a mean of only 3.90 CFU/g for563 nationwide samples of fresh ground beef (29). This surprisingly low numTable 6 Examples of Log,, Aerobic Plate Count (APC) Numbers of Bacteria/g of Fresh Ground Beef Year
Number
APC mean
Ref.
1914 1936 1957 1964 1975
44 41 96 26 5 5 140 140 563
7.23 6.49 7.5 1 6.72 5.30J 7.95b 6.30 6.81 3.90
22 23 24 25 26 26 27 28 29
1976 1977 1994 ~~~~~
On day 0. After 5 days at 4.5"C.
32
Jay
ber could be a reflection of the downward trend for appearance of bacteria in such products due in part to more attention being paid to sanitation throughout the production cycle or in part to the methodology employed in gathering these data. For example, the mean of 3.90 is the result of APC determinations being made at 35°C for 24 hours. For fresh meat products, this number would be higher if incubation was between 25 and 32°C for at least 48 hours. It was shown by Rogers and McCleskey (24) that the APC of retail store ground beef was considerably higher when plates were incubated at 7°C for 7 days than when incubated for 48hours at 30°C. From Table 6 it may be noted that >2.5 log cycle increase occurred after a 5-day incubation at 4.5"C in the meats sampled by Goepfert and Kim (26). Regarding the history of our knowledge of bacteria in fresh ground meats, the earliest proposed APC standard for acceptable product was made by Marxer in 1903 (30), and the limit was 106/g. The numbers in Table 6 for samples tested in 1914 and 1936 are reflective of this general level. The only U.S. state ever to adopt bacteriological standards for ground beef was Oregon, and it allowed up to 5 X 106/g (see Ref. 31). The Oregon law was in effect in 1973-1977. The APC limit in the Oregon standard is a reflection of what was considered safe and achievable under good production practices during that time. With APCs of approximately 106/g, fresh ground beef was rarely the source of foodborne outbreaks in the United States prior to the late 1980s. Whether the increased incidence of foodborne outbreaks from ground meats are related to lower numbers of background organisms is of possible concern, and it has been addressed (32). Of the genera of bacteria listed in Tables l and 2, most investigators have found around 20 in beef, poultry, fish, and sausage products. Ayres (33) identified 19 genera from refrigerated beef and 30 genera from the surfaces of beef, sausage, fish, and chicken (34). In a study of pork sausage, Sulzbacher and McLean (35) identified 16 genera. In a Canadian study, 19 genera were identified from fresh ground beef (36). In a study of the number of genera in fresh and spoiled ground beef, nine genera were identified from 69 isolates from the beef when fresh, and following the frank refrigerator spoilage of samples from the same batch, only four genera were found, with 16 of the 19 isolates being pseudomonads (37).
REFERENCES 1. Holzapfel. W. H., Geisen, R., and Schillinger, U. (1995). Biological preservation of foods with reference to protective cultures, bacteriocins and food-grade enzymes.Znt. J. Food Microbiol.. 24:343-362. 2. Aguirre, M., and Collins. M. D., (1993). Lactic acid bacteriaand human clinical infection. J. Appl. Bacteriol., 75:95-107. 3. Jett, B. D., Huycke,M. M., and Gilmore. M.S . (1994). Virulenceof enterococci. Clin. Microbiol. Rev., 7:462-478. 4. Jones, D. (1975). A numerical taxonomic study of coryneform and related bacteria. J. Gen. Microbiol., 8752-96. 5. Hughes, D. B., and Hoover, D. G. (1991). Bifidobacteria: Their potential for use in American dairy products. Food Teclznol., 45(4):74, 76, 78-80. 82. 6. Jay, J. M. (2000). Modern Food Microbiology, 6th ed., Aspen Publishers, Gaithersburg,MD. 7. Holzapfel, W. H. (1998). The gram-positive bacteria associated with meat and meat products. (1998). In The Microbiology of Meat and Poultn (A. Davies and R. Board, eds.). Aspen Publishers, Gaithersburg, MD, 1998.
Bacterial Biofa (Flora) in Foods
33
8. ICMSF. (1996). Microorganisms in Foods 5. Microbiological specifications of Food Pnthogem, Aspen Publishers, Gaithersburg, MD. 9. (1995). FDA Bacteriological Analytical Manual, 8th ed. AOAC Int., Gaithersburg, MD. 10. Vanderzant, C., and Splittstoesser, D. F., eds. (1992).Cornpendiurn of Methods for the Microbiological Examination of Foods. American Public Health Association, Washington,DC. 11. Feng, P. (1997). Impact of molecular biology on the detection of foodborne pathogens.Mol. Biotechnol., 7:267-278. 12. Brocklehurst, T.F., Zaman-Wong, C.M., and Lund, B. M. (1987). A note on the microbiology of retail packs of prepared salad vegetables. J. Appl. Bacteriol., 63:409-415. 13. Kaferstein, F. K. (1976). The microflora of parsley. J. Milk Food Technol., 39:837-840. 14. Ercolani, G.L. (1976). Bacteriological quality assessment of fresh marketed lettuce and fennel. Appl. Environ. Microbiol., 31:847-852. (1976). 15. Vescovo, M., Orsi,C., Scolari, G., and Torriani, S . (1995). Inhibitory effectof selected lactic acid bacteria onmicroflora associated with ready-to-eat vegetables.Lett. Appl. Microbiol., 21: 121-125. 16. Garg, S . , Churey, J. J., and Splittstoesser, D. F. (1990). Effect of processing conditions on the microflora of fresh-cut vegetables. J. Food Prot., 53:701-703. 17. Rafil, F., Holland, M.A., Hill, W. E., and Cerniglia, C. E. (1995). Survivalof Shigellaflerneri on vegetables and detection by polymerase chain reaction. J. Food Prot., 58:727-732. 18. Jinneman, K. C., Trost, P. A., Hill, W. E., Weagant, S . D., Bryant, J. L., Kaysner, C. A., and Wekell, M. M.(1995). Comparisonof template preparation methods from foods for amplification of Escherichia coli 0157 Shiga-like toxins type I and I1 DNA by multiplex polymerase chain reaction. J. Food Prot., 58:722-726. 19. Riser, E. C., Grabowski, J., and Glenn, E. P. (1984). Microbiology of hydroponically-grown lettuce. J. Food Prot., 47:765-769. 20. Odumeru, J. A., Mitchell, S . J., Alves, D. M., Lynch,J. A., Yee, A. J., Wang, S . L., Styliadis, S., and Farber,J. M. (1997). Assessment of the microbiological qualityof ready-to-use vegetables for health-care food services. J. Food Prot., 60:954-960. 21. Kneifel, W., and Berger, E. (1994). Microbiologicalcriteria of random samples of spices and herbs retailed on the Austrian market. J. Food Prot., 57393-901. 22. Weinzirl, L., and Newton, E. B. (1914). Bacteriological analyses of hamburger steak with reference to sanitary standards. Am. J. Public Health, 4:413-416. 23. Elford, W. C. (1936). Bacterial limitations in ground fresh meat. Am. J. Public Health, 26: 1204-1206. 24. Rogers, E. R., and McCleskey, C. S . (1957). Bacteriological quality of ground beef in retail markets. Food Technol., 11:318-320. 25. Jay, J. M. (1964). Beef microbial quality determinedby extract-release volume (ERV).Food Technol.. 18(10):133-137. 26. Goepfert, J. M., and Kim, H. U. (1975). Behavior of selected foodborne pathogens in raw ground beef. J. Milk Food Technol., 38:449-452. 27. Westhoff, D., and Feldstein, F. (1976). Bacteriological analysisof ground beef. J. Milk Food Prot., 39:401-404. 28. Foster, J. F., Fowler. J. L., and Ladiges, W. C. (1977). A bacteriological surveyof raw ground beef. J. Food Prot., 40:790-794. 29. U.S. Department of Agriculture, Food Safety and Inspection Service. (1996). Nationwide federal plant raw ground beef microbiological survey,August 1993-March 1994, USDA, Washington, DC. und der Haltbarkeit des Fleisches 30. Marxer, A. (1903). Beitrag zur Frage des Bakteriengehaltes bei gewohnlicher Aufbewahrung. For-tsckr. Vet.-Hyg., 1:328. 31. Carl, K. E., (1975). Oregon’s experience with microbiological standards for meat. J. Milk Food Technol., 38:483-486. 32. Jay, J. M. (1997). Do background microorganisms play a role in the safety of fresh foods? Trends Food Sci. Teclmol., 8:421-424.
34
Jay
33. Ayres, J. C. (1960). Temperature relationshipsand some other characteristicsof the microbial flora developing on refrigerated beef. Food Res.. 25:l-18. 34. Ayres, J. C. (1960). The relationship of organisms of the genus Psezdonzorzns to the spoilage of meat, poultry and eggs. J. Appl. Bacteriol. 23:471-486. 35. Sulzbacher, W. L., and McLean, R. A. (1951). The bacterialflora of fresh pork sausage.Food Techno].,5:7-8. 36. Lefebvre, N., Thibault, C., and Charbonneau, R. (1992). Improvementof shelf-life and wholesomeness of ground beef by irradiation. 1. Microbial aspects. Meat Sci., 32203-213. 37. Jay, J. M. (1967). Nature, characteristics, and proteolytic properties of beef spoilage bacteria at low and high temperatures. Appl. Microbial.. 15:943-944.
Aeromonas hydrophila Carlos Abeyta, Jr. U.S. Food and Drug Administration, Borhell, Wadzington
Samuel A. Palumbo U.S. Department of Agriculture, Philadelphia, Pennsylvania
Gerard N. Stelma, Jr. U.S. Emironmentrrl Protection Agewcy, Cincinnati, Ohio
I. Introduction 36 11. Classification and Characteristics 37
A. Classification B. Characteristics
37 40
111. Pathogenicity 40 A. Aeromonas infections B. Enterotoxins 41 C. Other virulence factors
40 43
IV. Control 44
A. Low temperature 44 44 B. Modified atmosphere C. pH and NaCl 45 D. Nitrite 45 E. Bacteriocins 46 46 F. Water activity 46 G. Chemical food additives H. Multiple barrier approach 47 I. Inactivation 47 V. Detection and Isolation 50 A. General considerations 50 B. Procedures 50 C. Sampling, enrichment, and isolation procedures References 52
50
Abeyta et al.
36
I.
INTRODUCTION
Aeromonas hydrophila translated simply means a "water-loving, gas-producing' ' bacterium. It was first reported by Zimmerman (1) and then Sandrelli (2), who isolated the organism and demonstrated its pathogenicity in frogs. Since that time, Aeromoncrs species have been isolated from other aquatic animals such as finfish, shellfish, crustaceans, and amphibians (3-5). This group of organisms is pathogenic to many aquatic species and causes hemorrhagic septicemia (red sore disease) in many fresh water pond cultured and wild native fish (6,7). It is well established that Aeronzonas is a component of the intestinal flora of healthy fish (8,9) and is widely distributed in nature. A. hydrophila is commonly found in watersheds (5,10,11). The prevalence and distribution of Aeronzonns spp. in fresh water habitats and the marine environment are well documented. It appears that Aeronzonas spp. may not be truly indigenous to the marine environment but may have a transient existence after entering salt water via rivers or sewage inputs. Aeromonas that are shed in sewage can multiply in sewage lines to significant numbers prior to discharge into receiving waters (12). Aeromonas is also present in terrestrial warm-blooded animals (13), including humans (14), and have been associated with three types of human illnesses; extraintestinal, wound, and gastrointestinal infections (14-20). Extraintestinal diseases are the most common, with a mortality rate as high as 61% in septicemic patients that are immunosuppressed. Wound infections involving Aeromonas usually are linked with injuries incurred during recreation or other activities in the aquatic environment (21-23). For gastrointestinal infections, a common source of A. hydrophila outbreaks is such water supplies as natural mineral springs, marine water environments, chlorinated and unchlorinated domestic supplies, and watersheds polluted by sewage effluents (24,25). There are statistical relationships between Aerolnonas species in water sources and health risks. However, there is no firm documentation of serious health effects after exposure to domestic waters containing Aeromonas (24). A serious health risk may be present for immunosuppressed patients with underlying malignancies after exposure to domestic and recreational waters containing Aeromonns species (14,26,27). The presence of Aeromonas in the food chain is well documented. This organism can be readily isolated from seafoods, foods of terrestrial animal origin such as meats, dairy products, poultry, and vegetables (28-34). Its presence in the food chain is of great concern because of its capability of growth at refrigerated temperatures (35). Seafood products are the most common food source of Aeromorzas. Both finfish and shellfish are known reservoirs of this bacterium. In 1986, A. hydrophila was the causative agent of an oysterborne outbreak in Florida (36). An attack rate of 100% (N = 7) was reported. A. hydrophila was isolated from the remaining uneaten oysters and from stool specimens from patients. In cases of foodborne bacterial illnesses in which shellfish are implicated, laboratories have included Aeronzonns in the general screening for causative microorganisms. Controversy concerning pathogenicity still remains. At present there are great difficulties assessing the regulatory significance of Aeromonas species in foods (37). From a clinical standpoint, the organism is no doubt of great concern to immunocompromised patients with underlying malignancies. Foods containing high levels of Aeromonns species destined to these individuals should be regarded hazardous. On the other hand, the role of foods containing high levels of Aeromonas spp. destined for healthy individuals is
Aeromonas hydrophila
37
uncertain. Although Aeromonas has been implicated in cases of gastrointestinal illness from foods, the exact mechanism by which Aerornonas species cause disease is not understood fully. Confusion concerning the assessment of pathogenicity remains unsolved. This was made evident in humanfeeding studies of virulent viable strains of Aeromonas species (high doses ranging from lo4 to lolo) involving 57 volunteers; only 2 developed mild diarrhea (38). One explanation for the failure of these strains to elicit diarrhea was presented. Kirov et al. observed the inability of Aerornonas species to adhere to the intestinal mucosa (39). This observation was noted in the shift of pilated environmental isolates toward nonpilated forms once the intestinal tract was infected. The nonpilated forms isolated from stools would be noninfective when used in challenge studies. The objectives of this chapter are to review the available information concerning this emerging recognizable pathogen. The reader is given information needed to assess further epidemiology, pathogenicity, management, detection, and isolation.
II. CLASSIFICATION AND CHARACTERISTICS A.
Classification
The genus Aeromonas was first proposed by Kluyver and Van Niel in 1936, as described by Popoff in Bergey’s Manual of Systematic Bacteriology (40). Various investigators divided the motile aeromonads into species including A. hydrophila, A. punctnda, A. formicans, A.liquefuciens, A. anaerogenes,A. proteolytica, and A. cnviae andA. sobria (41,42). The complexity of the problem was shown by Popoff et al., who found that A. hydrophila could be divided into distinct groups based on their divergence by DNA/DNA hybridization (43). The mesophilic A. hydroplzilngroup is collectively referred to as motile aeromonads. The genus Aeromonas consists of two well-separated groups of organisms: (a) the psychrophilic nonmotile aeromonads known as Aeromonas sahonicidn subspecies (pathogenic to fish but not to humans) and (b) the mesophilic motile aeromonads (the A. hydrophila group), which are divided into the species A. hydrophila, A. sobria, and A. caviae. The A. hydrophila group is associated with human illnesses. Most recent, a fourth motile Aeromonas spp., A. veronii, has been added to this list (44); it is distinguished from the other motile aeromonads by a positive ornithine decarboxylase test. Motile aeromonads belong to the family Vibrionaceae. Separation of the genus Aeromonas from related members of the family can be accomplished by biochemical reactions (Table 1). Aeromonas species are easily confused with group F Vibrios (V. fluvialis). Vibrios and motile aeromonads can be differentiated by their phenotypical reactions: salt requirement for growth and 0/129 sensitivity. Motile aeromonads are facultative anaerobic, gram-negative straight rods with rounded ends, measuring approximately 0.3- 1.O pm in diameter and 1.O-3.5 pm in length. They are motile by means of a single polar flagellum in liquid medium, and their metabolism is both respiratory and fermentative. They break down carbohydrates to acid or acid and gas (CO? and H?). Nitrate is reduced to nitrite, and oxidase and catalase are produced. Optimum growth temperatures range from 22 to 28°C. NaCl tolerance ranges from 0 to 4%, and tolerance to pH ranges from 5.2 to 9.8. They produce exoenzymes such as amylase, protease, phospholipase, and DNase and are resistant to vibriostatic agent 01129 (2,4-diamino-6,7-diisopropylpteridine). The mol% G + C of the DNA is 57-63 (Bd, T,,,).
Abeyfa et a/.
38
Table 1 Differentiation of the Genus Aeromonas from Other Genera of the Family Vibrionaceae Test Plesiowonas
Vibrio
Inhibition by 0/129 10 P8 150 P8 Na' requirement for growth Oxidase Gas from glucose Fermentation of inositol Fermentation of mannitol Ornithine decarboxylasea Growth on thiosulfate-citrate-bile salt-sucrose agar
Aeronzonns R
R S
R R
S
+ +
-
+ + +
-
+ + +
-
-
-
+ D +
-
-
R. Resistant: S . sensitive; D, differs among biotypes. verorlii is positive (+).
Table 2 Differential Phenotypic Characteristics salmorlicidu
of Motile Aeronlonas Species and Aeromonas b
Motile Characteristics
Aerorttonns spp.
1
,..I
.
. L
A. snlnlonicida
"
Motility Oxidase Ornithine decarboxylasea Arginine dihydrolase Indole production Starch, gelatin, DNA and RNA hydrolysis Citrate (Simmons') Citrate (Christensen's) ONPG Brown water-soluble pigment Growth without NaCl Fermentation of sucrose Fermentation of maltose, galactose, and trehalose Fermentation of cellobiose, lactose. and sorbitol Fermentation of glycerol, dulcitol, rhamnose, inositol, xylose, raffinose, and adonitol Rods in singles and pairs Coccobacilli in pairs, chains, and clumps Growth in nutrient broth at 37°C Aeronzonas varonii is ODC Differs among strains. c Aberrant strains occur. Source: Ref. 30.
i
+ + + + + db d
+ +
+ +"
+ -
+ + + -
d
+ + +c
d
-
-
-
+
-
-
+
+ -
+.
t
I
Aeromonas hydrophila 39
Abeyta et al.
40 Table 4 DifferentiationAmong the Aeromonas Izydrophia Group ~
~~
~~
~
Biochemical test“
A. hydrophila
Motility Esculin hydrolysis Growth in KCN L-Arginine utilization L-Lysine utilization L-Arabinose utilization Fermentation of salicin Fermentation of sucrose Fermentation of mannitol Breakdown of inositol Acetoin from glucose (Voges-Proskauer) Gas from glucose Indole production Oxidase P-Hemolysis H2S from cysteine
A. sobria
A. cavine
+
+ + + + + + + +
+b
+ + + + + + + + -
-
+
+
+ + -
+, Typically positive; -, typically negative. Source: Ref. 45. B. Characteristics The differential characteristics of motile aeromonads from nonmotile Aeromonns species are presented in Table 2. The common characteristic that distinguishes the A. hydrophila group (motile aeromonads) from A. snlnlonicicla subspecies of course is motility, but other similarities include lack of pigmentation, growth in nutrient broth at 37”C, colonial morphology, and monotrichous flagellation in liquid medium. In contrast, A. sdnzonicida does not grow at 37°C (optimum growth is at 22-25°C) and are nonmotile. Aeromonads can be easily grown in most routinely bacteriological selective and differential media (Table 3). On nonselective media such as trypticase soy agar, they are difficult to distinguish morphologically among members of the Enterobacteriaceae. The leading factors to consider in choosing the optimal culture medium is the type of sample matrix and the selective agent that will eliminate other competing organisms. Speciation of the motile aeromonads has been proposed by numerous investigators (Table 4). Biochemical reactions that are important for speciating among motile Aeromonns are based on esculin hydrolysis, growth in potassium cyanide (KCN) broth, salicin fermentation at 26”C, gas from glucose, and H2S from cysteine but not from thiosulfate.
111.
PATHOGENICITY
A. Aeromonas Infections Gastroenteritis is the most common foodborne illness attributed to the A. hydrophila group. About three quarters of Aerornorzns gastroenteritis cases are “choleralike,” characterized by watery stools and a mild fever; vomiting may also occur in children under 2 years of age. The other one quarter of cases are “dysenterylike,’ ’ characterized by blood and mucus
Aeromonas
41
in the stools (17). Aeromonns-associated diarrhea is normally mild and self-limiting (19); however, severe cases of both types of diarrhea have been observed (46-50). Aeromonads also have been implicated as the cause of localized wound infections, pneumonia, and such disseminating infections as bacteremia or septicemia and meningitis (16,18,19,5155). Although disseminating Aeromorlns infections may originate in infected wounds (16), the patient's gastrointestinal tract is considered to be the source of Aeromonns species in those infections (56). Therefore, disseminating infections should also be considered as potentially foodborne. One of several controversial issues regarding Aeromonas pathogenicity is the question of whether the aeromonads are all opportunistic pathogens, capable only of attacking hosts with impaired body defenses, or whether some of these organisms are sufficiently virulent to pose a threat to normal hosts as well. The fact that Aeromonas species are associated most frequently with diarrhea in the very young or in older adults (17,57-59) and the unusually high rate of isolation of aeromonads from patients with hematological malignancies (60) suggest that theyare opportunists. On the other hand, there is accumulating evidence that Aeromoms spp. also infect normal adults. George et al. identified Aeromonas isolates as the primary infectious agents responsible for gastroenteritis in a number of adult patients with no underlying disorders (20), and several literature reports describe severe cases of bothcholeralike (46,6 1) and dysenterylike (48) diarrhea in individuals who were otherwise healthy. Surprisingly, even bacteremia (51) and septicemia (53), which are usually associated with immunocompromised hosts (14,56), have been observed in immunocompetent young adults. It can be argued that immunocompetent individuals infected by aeromonads were affected by unrecognized predisposing conditions; however, it is quite likely that some highly virulent strains exist among the members of this diverse genus, which consists of at least 9 validated and/or proposed species (62) and 13 known hybridization groups that are extremely heterogeneous in biochemical structural and genetic properties (63). B. Enterotoxins The most controversial issue related to Aeromonns pathogenicity concerns the relative roles of various putative enterotoxins in causing diarrhea. There are reports of both heatstable (56"C, 10 min) and heat-labile cytotonic enterotoxins and cytotoxic enterotoxins, some related and others unrelated to cholera toxin. The controversy is not over the variety, which is not unusual for a diverse group of organisms, of enterotoxins reported, but over the fact that researchers who reported evidence for one of these toxins usually were unable to observe any of the others. The various putative enterotoxins produced by Aeromonas species are described below. Ljungh et al. reported partially purifying an enterotoxin with a molecular weight of 15 kD that was stable after treatment at 56°C for 10 minutes (64,65). This molecule, which was serologically unrelated to cholera toxin (CT), caused fluid accumulation in the permanently ligated rabbit ileal loop (RIL), rounding of Y-l cells without death, and stimulation of cyclic AMP (adenosine monophosphate) synthesis. Further evidence for a heat-stable cytotonic enterotoxin was provided by Chakraborty et al., who reported cloning the cytotonic enterotoxin gene into Escherichia coli (66). Culture filtrates of the clone caused elongation of Chinese hamster ovary (CHO) cells and fluid accumulation in the RIL after treatment at 56°C for 20 minutes. These activities also were unrelated to CT. Both of these groups reported that the P-hemolysin was inactive in the RIL.
42
Abeyfa et al.
Potomski et al. (67) reported using affinity chromatography to isolate an A. sobria toxin that cross-reacted with CT, caused rounding of Y-l cells and fluid accumulation in both RILs and infant mice after heating at 56°C for 20 minutes. All of these activities were reportedly neutralized by antiserum to CT. Schultz and McCardell (68) also reported rounding of Y-l cells and stimulation of cyclic AMP synthesis by Aeronzorzns culture filtrates that were heated at 56°C for 20 minutes. These activities were partially neutralized by antiserum to CT. They also reported that DNA from strains producing the cytotonic enterotoxin reacted with one or more synthetic oligonucleotide probes coding for CT. Chopra and Houston (69) reported purifying a cytotonic enterotoxin that caused fluid accumulation in the RIL, stimulated cyclic AMP synthesis, and caused elongation of C H 0 cells. The purified toxin had a molecular weight of 44 kDa, was free of hemolytic and cytotoxic activities, and was not cross-reactive with antiserum to CT. The biological activities of the purified toxin were heat labile at 56°C. The first evidence for a cytotoxic enterotoxin was provided by Cumberbatch et al. (70), who reported a correlation between cytotoxic and RIL activities. All of their cytotoxic isolates also were hemolytic. They found no evidence for a separate cytotonic activity in any of their isolates. Turnbull et al. (71) and Burke et al. (72) also observed a strong correlation between enterotoxin production and hemolytic activity. Asao et al. provided the first direct evidence for a cytotoxic enterotoxin (73). They purified a hemolysin from A. hydrophila strain AH-l to electro-phoretic homogeneity and observed that it was cytotoxic to Vero cells and had enterotoxic activity in both the RIL and suckling mouse assays. The hemolysin had a molecular weight of 50 kDa and was heat labile at 56°C for 5 minutes. Stelma et al. (74) showed that the Asao hemolysin was P-hemolysin and that antiserum to the purified hemolysin completely neutralized the RIL activities of filtrates of P-hemolytic Aeronzonas isolates, indicating that P-hemolysin alone can cause the changes in intestinal permeability associated with diarrhea. These researcher also used polyclonal antiserum to show serological cross-reactions between the hemolysin purified from the Japanese strain, AH- 1, and hemolysins produced by isolates from diverse geographic origins. Several additional studies provided evidence that the Aeromonns P-hemolysins are a family of molecules related to the "aerolysin" originally described by Bernheimer and Avigad (75). Rose et al. (76,77) purified a 52 kDa protein toxin that possessed hemolytic, cytotoxic, and enterotoxic activities as well as serological cross-reactivities to both CT and the Asao (AH-l) hemolysin. The biological activities of their toxin were neutralized by homologous antiserum and antiserum against the AH-l hemolysin but not by antiserum against cholera toxin. Asao et al. (78) demonstrated that the hemolysin produced by A. hydrophila CA1 1, a U.S. Gulf Coast isolate, was related immunologically to AH-l hemolysin but also possessed unique antigenic determinants. Millership et al. (79), Kozaki et al. (80), and Stelma et al. (81) provided evidence that hemolysins from A. sobt-in and A. veronii also possessed enterotoxic activities and were serologically related to AH-l hemolysin. Evidence from several surveys suggests that the cytotoxic enterotoxins are the most common Aerornorzas enterotoxins. Cumberbatch et al. found no cytotonic activity in the filtrates of 96 isolates (70). Likewise, Johnson and Lior found no cytotonic activity in the filtrates of 73 isolates (82), and Seidler et al. found cytotonic activity in the filtrates of only 20 of 330 isolates (6%) (83). Stelma et al. found evidence for cytotonic activity in one of 24 isolates (74), but this was later shown to be caused by sublethal doses of partially denatured cytotoxic enterotoxin after heating at 56°C (84).
Aeromonas hydrophila
43
The absence of significant enterotoxic activity in the P-hemolysin purified by Ljungh et al. (64) may be due to differences in the purification procedures. Ljungh et al. used a six-step procedure that recovered only 0.6% of the original hemolytic units (64). In contrast, Asao et al. used a two-step procedure that recovered 65% of the original hemolytic units and probably yielded a product closer to the native molecule in all of its properties (73). The failure of the E. coZi clones carrying the A. hydrophila hemolysin gene to cause fluid accumulation in the RIL (66) was later shown to be due to the inability of the E. coli to release the hemolysin from the cells (85). Although the heat-labile cytotoxic enterotoxins appear to be the mostcotnmon Aeromonas enterotoxins, the relative roles of these toxins and the cytotonic e,nterotoxins in Aerornonas diarrhea are not known. Determination of the relative roles of these toxins will require the development of better animal models and either the use of strains that produce only one toxin or the use of transposon mutagenesis to inactivate various toxins one at a time. It has been agreed generally that A. cctvirre isolates were noncytotoxigenic and were not enteric pathogens (17,86). However, Namdari and Bottone recently reported detecting a heat-stable cytotoxin in culture filtrates of isolates grown in double-strength trypticase soy broth (TSB) (87). They, like earlier investigators, did not detect enterotoxin in filtrates of A. caviae grown in single-strength TSB. Several recent studies have also linked A. caviae to diarrhea in very young children and individuals aged 50 years or more (8890).
C. OtherVirulenceFactors The negative results of the human feeding study performed by Morgan et al. provided evidence that strains producing enterotoxin active in the RIL are not always diarrheagenic and that multiple virulence factors are involved (38). The strains used in that study did not possess either adhesion factors for colonization of the intestine or ability to invade tissues. Several studies have provided evidence that some aeromonads possess adhesion factors that correlate with the possession of pili (39,91,92). The most interesting study was that of Kirov et al. (39), who observed that environmental enterotoxigenic isolates possessed nutnerous pili, which appeared to be lost once infection was established. This suggests that in future studies environmental isolates may be more appropriate for human feedings than clinical isolates. The relationship between virulence factors and ability of Aeronzor~zsstrains to cause disseminating infections has not been established. Lethality tests with both normal mice and mice immunosuppressed by x-irradiation have focused primarily on relating virulence to biotype or phenospecies (93,94). The results of those studies and one of invasiveness to mammalian cells (95) indicated that A. hydrophila and A. sobria were inherently more virulent than A. caviae, but with significant strain-to-strain variation within a species. One property linked to virulence is a surface array protein (S layer) commonly found in human isolates from extraintestinal infections. The role of the S layer in human and animal infections is not clear, but it appears to be substantially different from the role of the S layer of A. salmonicida in fishdisease (96,97). Other properties of motile aeromonads that have been linked to virulence in other bacterial species include the ability to invade mammalian cells (95,98), resistance to serum bactericidal effects (99,100), production of a siderophore capable of removing iron from transferrin (101,102), a mechanism for utilization of iron from heme compounds (102), and production of proteases (103). Loss of
et
44
Abeyta
al.
viability after growth in broth containing 0.5% glucose (suicide phenomenon) has been associated with both enteropathogenicity and virulence (104), but the significance of this phenomenon is not known. The relative significance of each of these putative virulence characteristics cannot be determined until appropriate animal models are developed, preferably models in which the animals are compromised in a way that mimics the condition of the susceptible human host. i
IV. CONTROL The presence, survival, and growth of bacteria in foods can be controlled by the applying the three Ks: keep them out, keep them from growing, and/or kill them. The A. hydrophila group occurs widely in the aquatic environment (105), and this undoubtedly represents the major source of these bacteria in foods. Their widespread occurrence in virtually all fresh and processed foods surveyed attests to their ubiquitous presence in foods (105). It would thus appear that it is difficult if not impossible to keep them out of foods. Based on this, control of their levels would then depend on extrinsic and intrinsic factors such as low temperature, modified atmosphere, pH/acid, salt and water activity, and the use of food preservatives to keep them from growing and on measures such as high temperatures (55°C and above), irradiation, and low pH, especially when adjusted with organic acids, to kill them.
A.
Low Temperature
Low-temperature holding (5°C) or refrigeration of fresh and processed foods has traditionally been relied on by the food industry to prevent the growth of foodborne pathogens. However, A. hydrophila, along with other foodborne pathogens, has been observed to grow at holding temperatures typically used for fresh and processed foods (106). It has long been observed that A. hydrophila can grow at 5°C or below (35,107,108). Palumbo et al. (32) determined that naturally occurring A. hydrophila could grow competitively in various retail foods of animal origin during storage for one week at 5°C. Callister and Agger (34) made a similar observation for fresh vegetables held at 5°C. Using ground pork inoculated with A. hydrophila, Palumbo (109) verified the low-temperature competitive growth of this bacterium in a food and determined that additional factors such as NaC1, vacuum packaging, and pH can also affect its growth. Since A. hvdrophila can readily and competitively grow in foods held at 5"C, factors other than low refrigeration assume greater importance in controlling its growth in foods.
B. Modified Atmosphere Members of the family Vibrionaceae are facultative anaerobes (42). Broth studies (1 10) and surveys of red meats stored under various atmospheres (30,3 1,35,111-113) have verified this. Based on broth studies, Palumbo et al. (1 10) determined that overall A. hydrophila grew as well anaerobically as it did aerobically. The observations from the red meat storage surveys indicated that the meats' microflora and/or storage temperature rather than the packaging atmosphere appears to control the numbers and types of the bacteria that develop and are detected. When conditions favor the development of Pseudomonas
.
.
I~
Aeromonas hydrophila
45
species or lactic acid bacteria, these bacteria will dominate the Aeromonas species (31,109,113).
C. pH andNaCl After temperature, pH and NaCl are the next most important food parameters controlling the growth of most bacteria in foods. As a typical gram-negative bacterium, A. hydrophila is fairly sensitive to acid or low pH (4.5%) (35); in addition, as shown in Table 5, the limiting pH and NaCl levels are temperature dependent. While the observations in Table 5 were generated in brain-heart infusion (BHI) broth, similar responses were seen during studies of the bacterium in ground pork in which pH and NaCl interacted to restrict the growth of A. hydrophila at lower levels than with either factor individually (109). Using a multifactorial approach, Palumbo et al. (110,115) studied the individual effects and interactions of temperatures, pH, NaC1, and sodium nitrite on the growth kinetics of the bacterium in BHI broth. In these two studies, all factors interacted to increase lag and generation times as pH decreased and the NaCl level increased. For most studies described, HC1 was the acidulant; however, Palumbo and Williams (1 14) used different acidulants and observed differences in the response of A. hydrophila based on the acidulant used, with acetic and lactic acids being the most restrictive and H2S04and HC1 being the least so. Overall, the order of effectiveness for the acids tested are (listed from most restrictive to the most permissive): acetic, lactic, tartaric, citric, H2S04,and HCl. Their activity appeared to be related to their pK,s. D. Nitrite Sodium nitrite in combination with NaCl acts as the curing agent in cured meat products. Its continued permitted use in cured meat products is based on its anticlostridial activity. While members of the A. hydrophila group generally cannot grow in cured meats, they can be isolated from these products (1 16,117). The growth-inhibiting mechanism is sodium nitrite combined with their brine content, pH (about 6.0), vacuum packaging, and low storage temperatures (1 10). These conditions provide an environment in which A.
Table 5 Influence of Temperature on pH and NaCl Limits for A. hydrophila K144 Grown Aerobically in Brain-Heart Infusion Broth Limits of Temperature (C)
28
4
PH G at 5.5 NG at 4.5 NG at 5.5
G. Growth: NG, no growth. Highest level tested. Source: Ref. 35.
NaCl (9%) G at 4 3 G at 3.5 NG at 4.5
46
Abeyfa et al.
Izydrophiln cannot grow readily or compete with the normal flora of lactobacilli, micrococci, and yeasts found in these products.
E. Bacteriocins Bacteriocins are polypeptide antimicrobial compounds produced by various strains of lactic acid bacteria (LAB). There is increasing interest today in the “natural” preservation of foods, and the use of lactic acid cultures and/or the bacteriocin(s) isolated from them have been studied for their effectiveness against spoilage and foodborne pathogens. Lewus et al. (1 18) reported on a study in culture broth, which investigated the activity of various bacteriocin-producing LAB against A. hydrophila. They observed that LAB isolated from meats as well as known LAB cultures would inhibit the bacterium. Santos et al. (1 19) observed that the bacteriocin(s) of a dairy starter culture inhibited A. hydrophila in both skim and ewe’s milk. However, in studies by Kalchayanand et al. (120), neither nisin (4000 activity units/mL) nor pediocin (4000 activity units/mL) was individually effective against the bacterium in broth, but when combined with a preliminary heating or freezing treatment (stresses to damage the cell membrane and cell wall), A. hydrophila became sensitive. At this time, bacteriocins would appear to have some potential uses against this bacterium, but conditions for their optimal activity and effectiveness need to be developed.
F. WaterActivity Water activity (a,) is ameasure of the amount of free water available for microbial growth. Since all microbes need free water for growth, any food substance that binds water can control microbial growth. Some common food substances that can bind water include various salts including NaCl (its activity is discussed above), sugars and carbohydrates, and amino acid and proteins. In perhaps the only study of its kind currently available, Santos et al. (121) studied the effect of a, (adjusted with NaC1, glycerol, and polyethylene glycol) on the growth of three strains (food isolates) of A. hyc/rophiln at 28, 10, and 33°C. The minimum a, for growth varied with strain, temperature, and type of humectant, with NaCl being the most inhibitory and glycerol being the least at comparable a,$s.
G. ChemicalFoodAdditives Various miscellaneous chemicals (many are GRAS [generally regarded as safe]) are added to foods, and in addition to their other functions, they can control the growth of foodborne pathogens, including A. lzydoyhilu. Some of these include essential oils, liquid smoke, food preservatives, and polyphosphates. Stecchini et al. (122) studied the effects of the essential oils of clove, coriander, nutmeg, and pepper in botha model system and noncured cooked pork and observed that the bacterium was inhibited. Liquid smoke prepared from several species of wood also inhibited A . h,dr-ophi/a (123). When the food preservatives methyl p-hydroxybenzoate and potassium sorbate were tested against four psychrotrophic foodborne bacteria, A . hydrophila was the most sensitive (134). When Venugapal et al. (125) testedthe food preservatives butylated hydroxyanisole, propylhydroxy parabenzoate, and sodium tripolyphosphate against A. hydrophila, they observed that protease secretion was more sensitive than growth. When Palumbo et al. (126) tested the food polyphosphates Sodaphos, Hexaphos, sodium pyrophosphate, and sodium tripolyphosphate against the bacterium individually in culture broth, they observed only small changes
4
I
Aeromonas hydrophila
47
in growth kinetics (lag and generation times). However, when combined with 3.5% NaCl, the number of viable A. hydrophila declined to an undetectable level. This effect was also noted when the combination was tested in ground pork.
H. MultipleBarrierApproach As is apparent from the above discussion on the influence of specific factors in controlling the growth of A. hydrophila, most factors (NaCl [expressed both as a percent and as a,], sodium nitrite, and pH), at the levels generally encountered in foods, individually cannot restrict the growth of the bacterium in foods. Research from this laboratory has indicated that multiple barriers (factors) or the multifactorial approach (two or more factors at less than maximum inhibitory levels) can successfully be used to inhibit foodborne pathogens such as Listeria ~~lonocytogenes (127), Shigella fiexl1eri (128), Yersinia enterocolitica (129), E. coli 0157:H7 (130), and A. h,,drOphih (110,115). From these studies, predictive models or equations have been developed to allow description of how changes in such culture (food) parameters as temperature, pH, NaCl level, and atmosphere can alter a bacterium's growth kinetics (e.g., lag and generation times). Often, small changes in a single parameter can bring about dramatic increases in lag and generation times and a decreased hazard frotn that particular pathogen. These equations have been incorporated into a user-friendly, Windows-based computer software program (Pathogen Modeling Program, version 5.1, available from the Microbial Food Safety Research Unit (Eastern Regional Research Center [ERRC], U. S. Department of Agriculture, 600 E. Mermaid Lane, Wyndmoor, PA 19038 or from the ERRC website (WWW.ARSERRC.GOV). I. Inactivation There are several food-processing operations that can inactivate bacteria and other microorganisms. These include heating (cooking), irradiation, sanitizing and disinfecting, and acidification. In addition, treatments such as the lactoperoxidase system have been shown to inactivate bacteria in foods. 1. Heating Heating is one of the primary treatments for the destruction of pathogenic and spoilage bacteria in foods. As withother bacteria, the thermal resistance of A. hydrophila is affected by factors such as growth temperature, age of the culture, and heating menstruum [buffer or different foods] (13 1,132). Nishikawa et al. (133) determined that A. hydrophila was more heat sensitive than Salmonella typhiwzuiurn and E. coli 0157:H7 when heated in either hamburger or egg yolk. Condon et al. (13 1) calculated a DS5"C= 0.17 minute and a Z = 5.1 1°C for a single strain when heated in buffer. Using both clinical and food strains, Palumbo et al. (132) determined DlsJcof 5.2 and 4.3 minutes, respectively, when heated in saline or raw milk and a Z = 6.2"C. These few studies suggest that this bacterium's thermal resistance is similar to that of other gram-negative bacteria found in foods and the bacterium should be inactivated by the heat treatments given many food products during their normal processing
2. Irradiation Irradiation along with heating represents a means of putting energy into a system (culture broth or food) to inactivate microorganisms. Though there are very few studies on the
Abeyta et al.
48
radiation resistance of A. hydrophila (134,135), it does appear to be relatively radiation sensitive and pasteurizing doses aimed at eliminating other foodborne pathogens such as SuZmoneZZa and E. coli 0157 :H7 should also eliminate A. hydrophila (136). 3. Chlorine and Other Sanitizers Cattabiani evaluated the influence of various food plant sanitizers (disinfectants), including chlorine, on four strains of A. hydrophila (137). The results of his study are presented in Table 6; in general, the bacterium appears to have susceptibilities similar to other gramnegative bacteria found in foods. Knochel studied the chlorine resistance of motile Aeromonas species using two different methods and observed that Aeromonas species were more susceptible to chlorine that other gram-negative bacteria such as E. coli, Klebsiella species, and Pseudomonas aeruginosa (138). As can be seen from the data in Table 6 , A. hydrophila is inactivated readily by chlorine at the levels used for the treatment of drinking water. However, the bacterium can be isolated from chlorinated water supplies (24,139,140), even when the test for E. coli is negative (24,140). The recovery of the organism from chlorinated water may be explained as posttreatment recontamination, the presence of unusually large numbers of A. hydrophila, or the presence of organic matter that can inactivate the added chlorine. An alternate explanation may be offered: A. hydrophila is known to be injured by sanitizer treatments (141) and, when selective media are used to isolate the bacterium after treatment, the bacterium is not recovered and thought to be absent. Thus, the presence of the
,
Table 6 Sensitivity of A. hydrophila to Disinfectants Time of exposure at 25°C (min) Concentration
5
10
+ + + +
-
-
3+/l3+/1-
1+/31+/3-
-
-
-
3+/1-
2+/2-
1
Compound Sodium hypochlorite
Quaternary ammonium compound
Iodoform
2-Chlorophenol
Glutaraldehyde
5 PPm 2.5 ppm 1.25 ppm 0.625 ppm 0.31 ppm 1:12,500 1:25,000 1:50,000 1:100,000 10 PPm 5 PPIn 1 PP" 0.2% 0.1% 0.05% 0.125% 0.0625% 0.03 % 1
-, Sensitive. reduction of four or more log cycles in viable count; Source: Ref. 137.
-
+ + 3+/1+ + + + -
3+/1-
+
+. resistant,
+
+ + 1+/3+ + -
+
Aeromonas hydrophila
49
bacterium in such foods as poultry carcasses and vegetables may represent contamination via the potable water supply. 4. Organic Acids In the study by Palumbo and Williams (1 14) cited above in the section on pH and NaC1, the influence of organic acids in combination with temperature and NaCl on the growth kinetics (lag and generation times) of A. hydrophila was investigated. They also observed that certain combinations were lethal to the bacterium, These findings are presented in Fig. 1. Again, the form of the acidulant was important, with acetic and lactic acids being the most toxic and the two inorganic acids being the least toxic. While these findings suggest that A. hydrophila should not be a problem in various fermented and pickled foods, actual experimental studies and food surveys gave mixed observations. Aytac and Ozbas (142) determined that A. hydrophila inoculated into yogurt mix decreased to undetectable during the fermentation; however, Knochel and Jeppesen (117) were able to isolate the bacterium from 10% of the mayonnaise-based salads, often at levels of >105/g.
5. Lactoperoxidase System The enzyme lactoperoxidase (LP) in the presence of thiocyanate and hydrogen peroxide can produce an active antimicrobial system in milk; this system is bactericidal for gramnegative bacteria including A, hydrophila. Santos et al. (143) observed that A. hydrophila decreased to undetected in broth, skim milk, and ewes’ milk and indicated that the LP system, when used in combination with low temperatures, could be useful in controlling the presence of this bacterium in fluid dairy products. Santos et al. (144) determined that the LP system could effectively reduce the levels of A. hydrophila during the manufacture of the Spanish fresh sheep’s cheese Villalon. It would appear that the LP system could provide an adjunct to good manufacturing practices in controlling A. hydrophila in dairy foods in general.
4.4 4.0
S
p
3.6
W
,$ 3.2
2
2.8
S 2.4
8 2.0
l4
1.6
1.2 0
50
100
150
200
250
300
Hours at 5°C Fig. 1 Effect of different acids (HC1, sulfuric, tartaric, citric, acetic, and lactic) on the decline in viable count of A. hydrophila in BH1 broth at pH 5.0 and 5°C (0.5% NaCl). Dashed line-lower limit of detection (log,, = 1.33).
Abeyta et al.
50
V.
DETECTIONANDISOLATION
A.
GeneralConsiderations
In handling of suspect samples, immediate analysis upon arrival in the laboratory is preferred. Motile aeromonads have the capability of proliferating at refrigeration temperatures. If samples are to be analyzed within a few days, store at refrigeration temperatures. Samples kept longer than one week should be stored at -20 or -72°C. An enrichment procedure is necessary when analyzing frozen foods and foods that contained injured aeromonads or low levels of motile aeromonads.
B. Procedures Procedures suggested for isolating and enumerating motile aeromonads from foods are found in the Bacteriological AnalyticalMarzual (BAM) (45). These procedures have been used in Food and Drug Administration laboratories for the analysis of various foods samples. Since the publication of these methods, other procedures have been developed and are recommended in conjunction with BAM methodology for isolation and enumeration of motile aeromonads. These procedures are discussed in the next section.
C. Sampling,Enrichment,andIsolationProcedures The following procedures are based on analysis of a 25 mL or 50 g analytical unit at 1:9 (sample/diluent) ratio. For samples containing less that 25 g, add enough diluent to maintain a 1:9 ratio. For samples requiring enrichment techniques, the following procedures should be followed. Aseptically weigh a 25 g sample into a sterile wide-mouthed, screw-cap jar (500 mL) or other appropriate container (i.e., stomacher bags). Add 225 mL of sterile trypticase soy broth with ampicillin (TSBA) and blend for 2 minutes Loosen jar cap about 1/4 turn and incubate 24 t 2 hours at 35 ? 2°C. After incubation of enrichment broth for 18-24 hours at35"C, transfer 3 mm loopfuls of inoculum onto Macconkey agar (MAC), peptone-beef extract-glycogen (PBG) agar, and Yersinia selective agar (YSA) base supplemented with cefsulodin and novobiocin as described in BAM to yield isolated colonies. A recent recommendation is to use starch ampicillin (SA) agar instead of PBG and YSA. For enumerating aeromonads in TSBA, inoculate a 3- or 5- tube most probable number (MPN) series containing TSBA from serial dilutions. Incubate broth tubes for 18-24 hours at 35°C. Follow procedures as described above for isolation of aeromonads. For samples not requiring an enrichment procedure, aseptically weigh 25 g of sample into an appropriate container and add 225 mL of sterile 0.1% peptone water. Make appropriate serial dilutions and surface plate 0.1 mL portions onto selective media, as described below, with a sterile bent glass rods to distribute the entire inoculum evenly over the surface of the media. Starch-ampicillin (SA) agar developed by Palumbo et al. (32) is recommended for direct surface plating of samples. Starch hydrolysis was selected since this enzyme activity is largely restricted to Aerornorzas and Vibrio species. The use of ampicillin effectively suppresses growth of coliforms and members of the family Enterobacteriaceae. SA has been used effectively for red meats, chicken, raw milk, seafoods, and fresh vegetables. Most recent, marine environmental samples including shellstock oysters, sediment, and marine waters were evaluated by SA direct plating for enumeration of motile aeromonads (see Table 7). Results indicated that direct plating with SA was
Aeromonas
51
Table 7 Evaluation of Starch-Ampicillin Agar in Recovering Motile Aeromonads from Shellfish-Growing Waters in Humbolt Bay, Eureka, California Media Sample lo)
MCA/TSBA ( MPN)a
SAA/TSBA (MPN)b
MCA & SAA/TSBA (CFU)' (MPN)'
SAA
Oysters Water Sediment
1.17'(60)' 2.17 (90) 3.69 (100)
1.40 (70) 2.23 (90) 3.76 (100)
1.47 (70) 2.34 (90) 4.11 (100)
0.15 (10) 0.91 (70) 3.37 (100)
(I2 =
Most probable number(MPN) determinations in tryptic soy brothwith ampicillin (TSBA) plated onto MacConkey agar (MCA). h Plated onto starch-ampicillin agar (SAA). c Combined MCA and SAA. Colony forming unit determinations in direct plating onto SAA. e Log,(, perg or mL. Percent positive of motile aerornonads detected.
ENRICHMENT OR DIRECT PLATING
c SELECTIVE AGARS .1 OXIDASE (
c
A . HYDROPHlL 5.
+) MEDIUM(AHM)
SALT TOLERANCE OR VlBRlOSTATlC 01129
c DIFFERENTIATION AND CONFIRMATION .1 CONVENTIONAL TEST OR RAPID DIAGNOSTIC TEST (i.e., API) Hemolysin Gas from glucose H,S from cysteine Esculin hydrolyis Acetoin from glucose KCN Arginine Arabinose Salicin Sucrose
Fig. 2 Schematic diagram for the isolation of motile aeromonads. (From Ref. 45.)
52
Abeyta et al.
effective with sediment samples, but not for shellstock oysters and marine water, in which the enrichment procedure showed an advantage over SA direct plating. Identification and confirmation of motile aeromonads can be accomplished by following the schematic diagram of Fig. 2. Aeromonas hydrophila medium (AHM) described by Kaper et al. (145) is useful for rapid presumptive identification and differentiation from coliforms and enterics. AHM is a single tube medium testing for fermentation of mannitol and inositol, ornithine decarboxylation, indole production, motility, and H,S production from sodium thiosulfate and cysteine. Studies by Abeyta et al. found this medium useful in identifying environmental marine isolates (1 1). The efficacy of the multitest screening AHM was determined. Of the 1396 strains positive for oxidase, 76% (1,065) gave typical reactions in AHM. Of the 1065 isolates, 95% were confirmed as motile aeromonads. Other rapid test systems such as API 20E, VITEK, and MICRO-ID are useful in rapid identification of typical motile aeromonads, however, these systems have their limitations. For example A. hydrophila and V.JEuvialisare related closely biochemically. Such additional tests as salt tolerance are recommended before a final identification can be reached.
REFERENCES 1. Zimmerman, 0. E. R. (1890). Die Bakterien unserer Tink und Nutzwassers, part 1. Ber. Naturw. Ges. Chemnits p. 38-39. 2. Sandrelli, G . (1891). Uber einen neuen Mikroorganismus des Wassers, welcher fiir Tiere mit veraenderlicher und konstantert Temperatur Pathogen 1st. Zentral. Bakteriol. Parasitenk. Irfektionskr. Hyg. Abt., 192:222-228. 3. Hird, D. W., Diesch, S. L., McKinnell, R. G., Gorham, E., Martin, F. B., Meadows, C. A., and Gasiorowski, M. (1983). Enterobacteriaceae and Aeromonas hydrophila in Minnesota frogs and tadpoles (Rana pipiens).Appl. Environ. Microbiol., 46: 1423-1325. 4. Food and Drug Administration. (1985). Pathogen surveillance sampling of Aerornonas hydrophila in foods. Food Safety Compliance Program 7303-030. 5. Abeyta, C., Jr., Kaysner, C. A., Wekell, M. M., and Stott, R. F. (1986). Incidence of motile aeromonads from United States west coast shellfish growing estuaries. J. Food Prot., 53: 849-855. 6. Snieszko, S. F. and Bullock, G. L. (1976). Disease of freshwater fishes caused by bacteria of the genera Aerornorm, Pseudomonas, and Vibrio. FDL-40. USK1, FWS. 7. Rigney, M. M., Zilinsky, J. W., and Roug, M. A. (1978). Pathogenicity of Aeromonas hydrophila in red leg disease of frogs. Current Microbiol., 1:175-179. 8. Trust, T. J. and Sparrow, R. A. H. (1974). The bacterial flora in the alimentary tract of freshwater salmonid fishes. Can. J. Microbiol., 20:1219-1228. 9. Wolke, R. E. (1975). Pathology of bacterial and fungal disease affecting fish. In The Pcrtkology of Fishes (W. E. Ribelin and G. Migaki, eds.), University of Wisconsin Press, Madison, W, pp. 33-116. 10. Hazen, T. C., Fliermans, C. B., Hirsch, R. P., and Esch, G. W. (1978). Prevalenceand distribution of Aerornonas hydrophila in the United States. Appl. Environ. Microbiol., 36:731738. 11. Abeyta, C., Jr., Weagant, S. D., Kaysner, C. A., Wekell, M. M., Krane, M. H., and Peeler, J. T. (1988).Aeromonas hydrophilain shellfish growing waters: Incidence and media evaluation. J. Food Prot., 52:7-12. 12. Rippey, S. R., and Cabelli, V. J. (1979). Membrane filter procedure forenumeration of Aerumonas hydrophila. Appl. Environ. Microbiol., 38:108-113. 13. Myers, B. R., Marshall, R. T., Edmondson, J. E., and Stringer, W. C. (1982). Isolation of
Aeromonas
14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
24.
25.
26. 27.
28. 29. 30. 31. 32.
33. 34.
35. 36.
53
pectinolyticAeromonas hydrophilaand Yersinia enterocoliticafrom vacuum-packaged pork. J. Food Prot., 45:33-37. Davis 11, W. A., Kane, J. G., and Garagusi, V. F. (1978). Human Aeromonas infections, a review of the literature and a case report of endocarditis. Medicine, 57:267-277. Qadri, S. M. H., Gordon, L. P., Wende, R. D. and Williams, R. C. (1976). Meningitis due to Aeromonas hydrophila. J. Clin. Microbiol., 3: 102-104. Wolff. R. L., Wiseman, S. L., and Kitchens, S. C. (1980).Aerontonas hydrophila bacteremia in ambulatory immunocompromised hosts. A m J. Med., 68:238-242. Gracey, M., Burke, V., and Robinson, J. (1982). Aeronronns-associated gastroenteritis. Lam cet, 2: 1304- 1306. Ellison, R. T., and Mostow, S. R. (1984). Pyogenic meningitis manifesting during therapy for Aerontonas hydroplziln sepsis. Arch. Intern. Med., 144:2078-2079. Holmberg, S. D., and Farmer 111, J. J. (1984). Aerontonas hydrophilaand Plesionzonas shigelEoides as causes of intestinal infections. Rev. Infect. Dis., 6:633-639. George, W. L., Nakata, M. M., Thompson,J., and White, M. L. (1985). Aeromonas-related diarrhea in adults. Arch. Intern. Med., 145:2207-2211. Rosenthal, S. G., Bernhardt, H. E., and Phillips 111, J. A. (1974). Aeromonas hydrophila wound infection. Plast. Reconstruct. Surg., 53:77-79. Hanson, P. G., Standridge, J., Jarrett, F., and Maki, D. G. (1977). Freshwater wound infection due to Aerontonas hydrophila. JAMA., 238: 1053-1054. Joseph, S. W., Daily, 0. P., Hunt, W. S., Seidler, R. J., Allen, D. A., and Colwell, R. R. (1979).Aeronzonas primary wound infection of a diver inpolluted waters. J. Clin. Microbiol., 10146-49. Burke, V., Robinson, J., Gracey, M., Peterson, D. and Partridge, K. (1984). Isolation of Aerontonas hydrophila from a metropolitanwater supply: Seasonal correlationwith clinical isolates. Appl. Environ. Microbiol., 48:36 1-366. Holmberg, S. D., Schell,W. L., Fanning,G. R., Wachsmuth, I. K., Hickman-Brenner, F. W., Blake, P. A., Brenner, D. J., and Farmer, 111, J. J. (1986). Aeromonas intestinal infections in the United States. Ann. Int. Med., 105:683-689. von Graevenitz, A., and Zinterhofer, L. (1970). The detection of Aeronzonas hydrophila in stool specimens. Health Lab. Sci., 7:124-126. Ramsay, A. M., Rosenbaum, B. J., Yardbrough, C. L., and Hotz, J. A. (1978). Aerontonas hydrophila sepsis in a patient undergoing hemodialysis therapy. J. Am. Med. Assoc., 239: 128-129. Nagel, C. W.,Simnpson, K. L., Ng, H., Vaughn, R.H., and Stewart, G.F. (1959). Microorganisms associated with spoilage of refrigerated poultry. Food Technol., 14:21-23. Kielwein, G., Gerlach, R., and Johne, H. (1969). Prevalence of A. hydrophila in raw milk. Arch. Lebensmittel Hyg., 20:34-38 (in German). Enfors, S. O., Molin, G.,and Ternstrom, A. (1979). Effect of packaging under carbon dioxide, nitrogen or air on the nlicrobialflora of pack stored at 4°C. J. Appl. Bacteriol., 47:197-208. Simard, R. E., Zee, J. and L’Heureux, L. (1984). Microbial growth in carcasses and boxed beef during storage. J. Food Prot., 47:773-777. Palumbo, S. A.. Maxino, F., Williams, A. W., Buchanan, R. L., and Thayer, D. W. (1985). Starch-ampicillin agar for the quantitative detection of Aeromonas hydrophila. Appl. Environ. Microbiol., 50: 1027-1030. Abeyta, C., Jr., and Wekell, M. M. (1988). Potential sources of Aerontonns hydrop?da. J. Food Safety, 9: 11-22. Callister, S. M., and Agger, W. A. (1987). Enumeration and characterization of Aeromonas hydrophila and Aerontonas cnviaeisolated fromgrocery store produce.Appl. Environ. Microbiol., 53:249-253. Palumbo, S. A., Morgan, D. R.,and Buchanan, R. L. (1985). Influenceof temperature, NaCl, and pH on the growth of Aeromonas hydrophila. J. Food Sci., 50:1417-1421. Abeyta, C., Jr., Kaysner, C. A., Wekell, M. M., Sullivan, J. J., and Stelma, G. N. (1986).
54
37. 38.
39.
40. 41. 42. 43. 44.
45.
46. 47. 48. 49.
50. 51.
52. 53.
53.
55. 56. 57. 58.
Abeyta et al. Recovery of Aerovnonas hydrophila from oysters implicated in an outbreak of foodborne illness. J. Food Prot., 49:643-646. Archer, D. L., and Kvenberg, J. E. (1988). Regulatory significance of Aerornonas in foods. J. Food Safely, 9:53-58. Morgan, D. R., Johnson, P. C., Dupont, H. L., Sattenvhite, T. K. and Wood, L. V. (1985). Lack of correlation between known virulence properties of Aeromonas hydrophila and enteropathogenicity for humans. Infect. Irnmut., 50:62-65. Kirov, S. M., Rees, B., Wellock, R. C.. Goldsmid, J. M., andvan Galen, A.D. (1986). Virulence characteristics of Aerornonns spp. in relation to source and biotype. J. Clin. Microbiol., 24:827-834. Popoff, M. (1984). Genus 111. In Bergey 'S Manual o f Systevnatic Bncteriology, Vol. 1, Williams and Wilkins, Baltimore/London, pp. 545-548. Schubert, R. (1969). Infrasubspecific taxonomy of Aerolzlonns hydrophiln (Chester 1901) Stainer 1943. Zentral. Bnkteriol. Parasitenk. Infektionsk. Hyg. Orig. Abt. 1, 31 1:406-408. Popoff, M., and Veron. M. (1976). A taxonomic study of the Aeronlonns ltydropltilu-Aeromonas purzctata group. J. Gen. Microbiol., 94: 11-22. Popoff, M., Coynault, C., Kiredjian, M., and Lemelin, M. (1981). Polynucleotide sequence relatedness among motile Aeromonns species. Curr. Microbiol., 5: 109-1 14. Hickman-Brenner, F. W., MacDonald. K. L., Steigerwalt. A. G., Fanning, G. R., Brenner, D. J., and Farmer 111, J. J. (1987). Aeronronns tleronii, a new ornithine-decarboxylase positive species that may cause diarrhea. J. Clin. Micl-obiol., 25:900-906. Abeyta. C., Jr., and Stelma, G., Jr. (1984). Aerontonas hydrophila group. In Bncteriological Analytical Manual, 6th ed., Association of Official Analytical Chemists, Arlington, VA, pp. 30.01-30.10. Champsaur, H., Andremont, A., Mathieu, D., Rottman, E., and Auzepy, P. (1982). Choleralike illness due to Aeromonns sobria. J. Infect. Dis., 145:248-254. Marsik, F., and Werlin, S. L. (1984). Aerontonas lzydroplzilu colitis in a child. J. Pediatr. Gastroenterol. Ntrtr., 3:808-8 11. Rhaman, A. F. M. S . , and Willoughby, J. M. T. (1980). Dysentery-like syndrome associated with Aeromonns hydrophila. Br. J. Med., 281:976. Rosner, R. (1964). Aeromonas hydrophila as the etiologic agent in a case of severe gastroenteritis. Am. J. Clin. Prttlzol. 42:402-404. Sawle, G. V., Das, B. C., Acland, P. R.and Heath, D. A. (1986). Fatal infection with Aeromonos sobria and Pleisiomonas shigelloides. Br. Med. J., 393525-526. Golik, A., Leonov. Y., Schlaeffer, F., Gluskin, I., and Lewinsohn, G. (1990). Aeromonas species bacteremia in immunocompromised hosts. Two case reports and a review of the literature. Israel J. Med. Sci., 26:87-90. Reines, H. D., and Cook, F. V. (1981). Pneumonia and bacteremia due to Aeronzolzas hydrophila. Chest., 80264-266. Scott, E. G., Russell, C. M., Noell, K. T., and Sproul, A. E. (1978). Aeromonas lzydrophila sepis in a previously healthy man. J. Ant. Med. Assoc., 239:1742. von Gravenitz, A. (1984). Aerornonas and Pleisiornonas as agents of diarrhea. In Irlfectiolrs Diarrheal Diseases. Current Concepts and Laboraton1 Procedures (P. D. Ellner, ed.), Marcel Dekker, New York, pp. 59-75. von Gravenitz, A., and Mensch, A. H. (1968). The genus Aeromonas in human bacteriology. Report of 30 cases and review of the literature. N. Engl. J. Med., 278245-249. Harris, R. L., Fainstein, V., Elting, L. E., Hopfer, R. L., and Bodey, G. D. (1985). Bacteremia caused by Aeromonas species in hospitalized cancer patients. Rev. Infect. Dis., 7:314-320. Agger, W. A., McCormick, J. D., and Gurwith, M. J. (1985). Clinical and microbiological features of Aeromonas hydrophila-associated diarrhea. J. Clin. Microbiol., 21 :909-9 13. Janda, J. M., Bottone, E. J., and Reitano, M. (1983). Aeronlorzns species in clinical microbiology: Significance epidemiology and speciation. Diagn. Microbiol. Infect. Dis.. 1:22 1-238.
Aeromonas hydrophila
55
59. Millership, S. E., Cumow. S. R.. and Chattopadhyay. B. (.1983). Faecalcarriage rate ofAeromonas hydrophila. J. Clin. Patlzol., 36:920-923. 60. Sherlock, C. H., Burdge, D. R., and Smith, J. A. (1987). Does Aeronzorzas lrydroplzilapreferentially colonize the bowels of patients with hematologic malignancies? Dirrgn. Microbiol. Irzfect. Dis., 7:63-68. 61. Gelbart. S. M., Prabhudesai, M., and Magee, S. M. (1985). A case report: Aeronronas sobrin gastroenteritis in an adult. J. Clin. Pathol., 83:389-391. 62. Camahan, A. M., Chakraborty, T., Fanning, G. R., Venna. D., Ali, A., Janda. J. M., and Joseph, S. W. (1991). Aeromorzns trotn sp. nov. an ampicillin-susceptible species isolated from clinical specimens. J. Clin. Microbiol., 29: 1206-1210. 63. Janda, J. M. (1991). Recent advances in the study of the taxonomy, pathogenicity, and infectious syndromes associated with the genus Aerontonas. Clin. Microbiol. Re)?.,4:397-410. 64. Ljungh, A., Wretlind, B., and Mollby, R. (1981). Separation and characterization of enterotoxin and two hemolysins from Aerontortns hydsophiln. ActaPcrtlzol. Micr-obiol. Scund. Sect B, 89:387-397. 65. Ljungh, A., Enroth. P., and Wadstrom, T. (1982). Cytotonic enterotoxin from Aeronzonas hydrophila. Toxicon, 20:787-794. 66. Chakraborty, T.. Montenegro, M. A.. Sanyal, S. C.. Helmuth, R., Bulling, E., and Tinmis, K. N. (1984). Cloning of enterotoxin gene from Aeronzonas hydrophila provides conclusive evidence of production of a cytotonic enterotoxin. Infect. Inzmuz., 46:435-441. 67. Potomski, J., Burke, V., Robinson, J., Fumarola. D., and Miragliotta, G. (1987). Aeromorzas cytotonic enterotoxin cross reactive with cholera toxin. J. Med. Microbiol., 23:179-186. 68. Schultz, A. J.. and McCardell. B. A. ( 1988). DNA homology and immunological cross-reactivity betweenAei-ontonashydrophilrr cytotonic enterotoxin andcholera toxin. J. Clin. Microbiol., 26:57-61. of a 69. Chopra, A. K.,andHouston. C. W. (1989). Purificationandpartialcharacterization hydrophila. Cart. J. Microbiol., 35:719-727. cytotonic enterotoxin produced by Aeromo~~ns 70. Cumberbatch, N., Gurwith, M. J., Langston, C., Sack, R. B., and Brunton, J. (1979). Cytotoxic enterotoxin produced by Aerornorras ltydrophiln: Relationship of toxigenic isolates to diarrheal disease. Infect. I m n z l n . , 23:829-837. 71. Turnbull, P. C. B., Lee, J. V.. Miliotis, M. D., Van De Walle, S.. Kornhof, H. J., Jeffrey. L., and Bryant, T. N. (1984). Enterotoxin production in relation to taxonomic grouping and source of isolation of Aeromonas species. J. Clin. Microbiol., 19: 175-180. Beaman, J.,andBundell, C. (1983). The 72. Burke,V..Gracey, M., Robinson,J..Peck,D., microbiology of childhood gastroenteritis: Aeromonus species and other infective agents. J. Infect. Dis., 148:68-74. 73. Asao, T., Kinoshita, Y., Kozaki. S., Uemura, T.. and Sakaguchi, G. (1984). Purification and some properties of Aerornonas hydrophilu hemolysin. Irlfect. Imnun., 46: 122- 127. 74. Stelma, G. N., Jr., Johnson, C. H.. and Spaulding, P. (1986). Evidence for the direct involvement of P-hemolysin inAeromonns hydrophila entero-pathogenicity. Cztrre~ttMicr-obiol.,14: 7 1-77. 75. Bernheimer, A. W.. and Avigad. L. S. (1974). Partial characterization of aerolysin, a lytic I I 1016-1021. I., exotoxin from Aeromonas hydrophila. Irgect. I ~ I Z Y ~ L9: 76. Rose, J. M., Houston, C. W.. Coppenhaver. D. H., Dixon, J. D., and Kurosky, A. ( 1989). Purification and chemical characterization of a cholera toxin-cross-reactive cytolytic entero. . 1165-1 169. toxin produced bya human isolate of Aerornorzrrs hydropltiln. Iizfecr. I Y ~ I N ~ M57: 77. Rose, J. M., Houston, C. W., and Kurosky, A.(1989). Bioactivity and inmunological characterization of a cholera toxin-cross-reactive cytolytic enterotoxin from Aero~~tu~rns hydruplzilr. Infect. Irnnrun.. 57: 1170-1 176. 78. Asao, T.. Kozaki, S., Kato. K., Kinoshita, Y . ,Otsu, K., Uemura, T., and Sakaguchi. G. (1986). Purification and characterization of anAerornonrrs hydrophila hemolysin. J. Clin. Microbiol., 23:228-232.
56
i
Abeyta et al.
79. Millership, S. E., Barer, M. R., Mulla, R. J., and Maneck, S. (1992). Enterotoxic effects of Aeromonas sobria hemolysin in a rat jejunal perfusion system identified by a specific neutralization with a monoclonal antibody. J. Gen. Microbiol., 138:261-267. 80. Kozaki, S., Asao, T., Kamata, Y., and Sakaguchi, G. (1989). Characterization of Aeromonas sobria hemolysin by use of monoclonal antibodies against Aeromonas hydrophila hemolysins. J. Clin. Microbiol., 27:1782-1786. 81. Stelma, G. N.. Jr., Johnson, C. H., and Spaulding, P. L. (1988). Experimental evidence for enteropathogenicity in Aeromonns veronii. Can. J. Microbiol., 34:877-880. 82. Johnson, W. M., and Lior, H. (1981). Cytotoxicity and suckling mouse reactivity of Aeronzonas hydrophila isolated from human sources. Can. J. Microbiol., 27:1019-1027. 83. Seidler, R. J., Allen, D. A., Lockman, H.,Colwell, R. R., Joseph,S. W., and Daily. 0. P. (1980).Isolation enumeration and characterization of Aeronzonas from polluted waters encountered in diving operations. Appl. Environ. Microbiol., 39:lOlO-1018. Kaylor, L. O., and Johnson, C. H. ( 1986). 84. Bunning, V. K., Crawford, R. G., Stelma, G. N., Jr., Melanogenesis in murine B 16 cellsexposed to Aerontonas hydrophilu cytotoxic entertoxin. Can. J. Microbiol., 32:815-819. 85. Chakraborty, T., Huhle, B., Bergbauer, H., and Goebel, W. (1986). Cloning expression and mapping of the Aeronlonas hydrophila aerolysin gene determinant in Escherichiacoli K12. J. Bncteriol., 167:368-374. 86. Janda, J. M., Reitano, M., and Bottone, E. J. (1984). Biotyping of Aeromonas isolates as correlate to delineating a species-associated disease spectrum. J. Clin. Microbiol.. 19:4447. 87. Namdari, H.,and Bottone, E. J. (1990). Cytotoxin and enterotoxin production as factors delineating enteropathogenicity of Aerontonas caviae. J. Clin. Microbiol., 28: 1796-1798. 88. Kuijper, E. J., Zanen, H. C. and Peeters. M. F. (1987). Aerontorzas-associated diarrhea in the Netherlands. Ann. Intern. Med., 106:640-641. 89. Kuijper, E. J., and Peeters, M. F. (1991). Bacteriologicaland clinical aspects of Aeromonasassociated diarrhea in the Netherlands. Experientia, 47:432-434. 90. Namdari, H., and Bottone, E. J. (1990). Microbiologic and clinical evidence supporting the role of Aerornonns caviae as a pediatric enteric pathogen. J. Clin. Microbiol., 28:837-840. 91. Clark, R. B., Knoop, F. C., Padgitt, P. J., Hu, D. H., Wong, J. D., and Janda, J. M. (1989). Attachment of mesophilic aeromonads to cultured mammalian cells. Curr. Microbiol., 19: 97- 102. 92. Hokama, A., Honma, Y., and Nakasone, N. (1990). Pili of an Aerontonas hydrophila strain as a possible colonization factor. Microbiol. Intmz~nol.,34:901-915. 93. Brenden, R. A., and Huizinga, H. W. (1986). Susceptibilityof normal and X-irradiated animals to Aeromonas hydrophila infections. Curr. Microbiol., 13:129-132. 94. Janda, J. M., Clark, R. B., and Brenden, R. (1985). Virulence of Aeromonas species as assessed through mouse lethality studies. Curr. Microbiol., 12:163-168. 95. Watson, I. M., Robinson, J. O., Burke, V., and Gracey, M. (1985). Invasiveness ofrleromonas spp in relation to biotype virulence factors and clinical features. J. Clin. Microbiol., 22:4851. 96. Janda, J. M., Oshiro, L. S., Abbott, S. L., and Duffey, P. S. (1987). Virulence markers of mesophilic aeromonads: Association of the autoagglutination phenomenon with mouse pathogenicity and the presence of a peripheral cell-associated layer.Irlfect. Inmun., 55:30703077. 97. Kokka, R.P.. Vedros, N. A., and Janda, J.M. (1991). Characterization of classic and atypical serogroup 0 : l l Aerornonns: Evidence that the surface array protein is not directly involved in mouse pathogenicity. Microb. Patlzog., 10:71-79. 98. Pazzoglia, G., Sack, R. B., Bourgeois, A. L., Froehlich, J., and Eckstein, J. (1990). Diarrhea and intestinal invasiveness of Aerontonas strainsin the removable intestinal tie rabbit model. Ilfect. Immun., 58:1924-1931.
Aeromonas
57
99. Janda, J. M., Brenden, R., and Bottone, E. J. (1984). Differential susceptibility to human serum by Aeromonas spp. Curr. Microbiol.. 11:325-328. 100. Palumbo, S. A., Bencivengo, M. M., Del Corral, F., Williams, A. C., and Buchanan, R. L. (1989). Characterization of the Aeromonas hydrophila group isolated from retail foods of animal origin. J. Clin. Microbiol., 27:854-859. 101. Barghouthi, S., Young, R., Olson, M. 0. J., Arceneaux, J. E. L., Clem, L. W., and Byers, B. R. (1989). Amonobactin, anovel tryptophan- or phenylalamine-containingphenolate siderophore in Aeromonas hydrophila. J. Bacteriol., 171:1811-1816. 102. Massad, G., Arceneaux, J. E. L., and Byers, B. R. (1991). Acquisition of iron from host sources by mesophilic Aeromonas species. J. Gen. Microbiol., 137:237-241. and distribution of extracellu103. Leung, K.-Y.,and Stevenson, R. M. W. (1988). Characteristics lar proteases from Aeromonas hydrophila. J. Gen. Microbiol., 134:151- 160. 104. Namdari, H., and Bottone, E.J. (1988). Correlationof the suicide phenomenon in Aerolnonas species with virulence and enteropathogenicity. J. Clin. Microbiol., 262615-2619. 105. Palumbo, S . , Abeyta, C., Jr., and Stelma, G. N., Jr. (1992). Aeromonas hydrophila Group. In Comperlditun of Methods for the Microbiological Examinationof Foods, 3rd ed., APHA, Washington, D.C., pp. 497-515. 106. Palumbo, S. A. (1986). Is refrigeration enough to restrain foodborne pathogens? J. Food Prot., 49:1003-1009. 107. Eddy, B. P., and Kitchell, A. G. (1959). Cold-tolerant fermentative gram-negative organisms from meat and other sources. J. Appl. Bacteriol., 2257-63. and temperature characteristics 108. Rouf, M. A., andRigney, M.M. (1971). Growth temperatures of Aeronzonas. Appl. Microbiol., 22:503-506. 109. Palumbo, S . A. (1988). The growth of Aeromonns hydrophila K144 in ground pork at 5°C. Int. J. Food Microbiol., 7:41-48. 110. Palumbo, S. A., Williams, A. C., Buchanan, R. L., and Phillips, J. G. (1992). Model for the anaerobic growth of Aeromonas hydrophila K144. J. Food Prot. 55:260-265. 111. Gill, C. O., and Reichel, M. P. (1989). Growthof cold-tolerant pathogensYersinia enterocolitica, Aeromonns hydrophila and Listeria monocytogenes on high-pH beef packaged under vacuum or carbon dioxide. Food Microbiol., 6:223-230. 112. Lee, B. H., Simard, R. E., and Holley, R. A. (1985). Effectof temperature and storage duration on the microflora, physicochemical and sensory changesor vacuum- or nitrogen-packed pork. Meat Sci., 13:99-112. 113. Grau, F. H., Eustace, I. J.. and Bill, B. A. (1985). Microbial flora of lamb carcasses stored at 0°C in packs flushed with nitrogen or filled with carbon dioxide. J. Food Sci., 50:482485, 491. 114. Palumbo, S. A., and Williams, A. C. (1992). Growth of Aeronlonas hydrophila as affected by organic acids. J. Food Sci., 57:233-235. 115. Palumbo, S. A., Williams, A. C., Buchanan, R. L., and Phillips, J. G. (1991). Model for the aerobic growth of Aerontonas hydrophila K144. J. Food Prot., 54:429-435. 116. Gobat, P.-F., and Jemmi, T. (1993). Distribution of mesophilic Aeromonas species in raw and ready-to-eat fish and meat products in Switzerland. Znt. J. Food Microbiol., 20:117120. 117. Knochel, S., and Jeppesen, C. (1990). Distributionand characteristics of Aeromonas in food and drinking water in Denmark. Int. J. Food Microbiol., 10:317-322. 118. Lewus, C. B., Kaiser, A., and Montville, T. J. (1991). inhibitionof foodborne bacterial pathogens by bacteriocins fromlactic acid bacteria isolated from meat. Appl. Environ. Microbiol., 57:1683-1688. 119. Santos. J. A., Lopez-Diaz, T. M., Garcia-Fernandez,M. C., Garcia-Lopez, M.-L., and Otero, A. (1996). Effectof a lactic starter culture on the growth and protease activityof Aeromonas hydroihila. J. Appl. Bncteriol., 80:13-18. 120. Kalchayanand, N., Hanlin, M. B., and Ray, B. (1992). Sublethal injurymakes gram-negative
58
121.
122. 123.
124.
125. 126. 127.
128.
129.
130.
131.
132. 133.
134. 135.
136. 137. 138. 139.
140.
Abeyta et al. and gram-positive bacteria sensitive to the bacteriocins, pediocin AcH and nisin. Lett. Appl. Microbiol., 15:239-243. Santos, J.. Lopez-Diaz, T.-M., Garcia-Lopez, M.-L., Garcia-Fernandez, M.-G., and Otero, A. (1994). Minimum water activity for the growth of Aerornonas hydrophila as affected by strain, temperature and humectant. Lett. Appl. Microbiol., 19:76-78. Stecchini. M. L., Sarais, I., and Giavedoni, P. (1993). Effect of essential oils on Aeromonas a medium and cooked in pork. J. Food Prot., 56:406-409. hydrophila inculture Sofos, J. N., Maga. J. A., and Boyle, B. L. (1988). Effect of ether extracts from condensed wood smokes on the growth of Aeromonas hydrophila and Staphylococcus aureus. J. Food Sci., 53:1840-1843. Moir, C.J., and Eyles.M. J. (1992).Inhibition, injury, and inactivation offour psychrotrophic foodborne bacteria by the preservatives methyl p-hydroxybenzoate and potassium sorbate. J. Food Prot.. 55:360-366. Venugapal, V., Pansare, A. C., and Lewis. F. N. (1984). Inhibitory effect of food preservatives on protease secretion by Aero~nonas hydrophila J. Food Sci..49:1078-1081. Palumbo, S. A., Call, J. E., Cooke, P. H., and Williams, A. C. (1995). Effect of polyphosphates and NaCl on Aerornonus hydrophila. J. Food SafeQ, 15:77-87. Buchanan. R. L., and Phillips, J. G. (1990). Response surface model for predicting the effects of temperature, pH, sodium chloride content, sodium nitrite concentration, and atmosphere on the growth of Listeria ntonocytogenes. J. Food Prot., 53:370-376. Zaika, L. L.. Phillips, J. G., and Buchanan, R.L. (1992).Model for aerobic growth of Shigella flexneri under various conditions of temperature, pH, sodium chloride and sodium nitrite concentrations. J. Food Prot., 55:509-513. Bhaduri, S.. Turner-Jones, C. O., Buchanan, R. L., and Phillips, J. G. (1994). Response surface model of the effect of pH, sodium chloride and sodiumnitrite on the growth of Yersinin enterocolitica at low temperatures. Int. J. Food Microbiol., 23:333-343. Buchanan, R. L., andBagi, L. K. (1994). Expansion of responsesurfacemodels for the growth of Escherichia coli to include sodium nitrite as a variable. Int. J. Food Microbiol., 23:317-332. Condon, S., Garcia,M. L., Otero,A.,andSala, F. J. (1992).Effectofcultureage,preincubation at low temperature and pH on the thermal resistance of Aeromonas hydrophila. J. Appl. Bacteriol., 72322-326. Palumbo, S. A., Williams, A. C., Buchanan. R. L., and Phillips, J. G. (1987). Thermal resistance of Aerornonas hydroplziln. J. Food Prot., 50:761-764. Nishikawa, Y., Ogasawara, J., and Kimura, T. (1993). Heat and acid sensitivity of motile Aer-ontonas:A comparison with other food-poisoning bacteria. Irzt. J. Food Microbiol., 18: 27 1-278. Palumbo. S. A., Jenkins, R. K., Buchanan, R. L., and Thayer, D. W. (1986). Determination of irradiation D-values for Aer-omonas hydrophila. J. Food Prot., 49: 189-191. Rashid, H. O., Ito, H., and Ishigaki, I. (1992). Distribution of pathogenic vibrios and other bacteria in imported frozen shrimps and their decontamination by gamma-irradiation. World J. Microbiol. Biotechnol., 8:494-499. Monk, J. D.. Beuchat, L. R., and Doyle. M. P. (1995). Irradiation inactivation of food-borne microorganisms. J. Food Prot., 58:197-208. Cattabiani, F. (1986). Sensibilita ai disinfecttani di Aeromonns hydrophila e di Vibriofluvialis. Arch. Vet. Ital., 37:65-72. Knochel, S. (1991 ). Chlorine resistance of motile Aerornonas spp. Water Sci. Technol., 24: 327-330. LeChevialier,M. W., Evans, T. M., Seidler, R. J., Daily, 0. P., Merrell,B.R.,Rollins, D. M., and Joseph, S. W. (1982). Aeromonns sobria in chlorinated drinking water supplies. Microb. Ecol., 8:325-333. Slade. P. J., Falah, M. A., and Al-Ghady, A. M. R. (1986). Isolation ofAeromonns hydrophila
b
Aeromonas hydrophila
141. 142. 143.
144.
145. 146. 147. 148.
149.
150.
151.
152. 153.
154.
155. 156.
157.
158. 159.
59
from bottle waters and domestic water supplies in Saudi Arabia. J. Food Prof., 49:471476. Cattabiani, F., and Brindani, F. (1988). Valutazione del Danneggiamento di Tip0 Structurale da Disinfecttanti in Aeronzonas, Vibrio, e Plesiomonas. Arch. Vet. Ital., 39:245-253. Aytac, S. A., and Ozbas, Z. Y. (1994). Survey of the growth and survival of Yersinia enterocolitica and Aerontonas hydrophila in yogurt. Milchwissenschafi, 49:321-324. Santos, J. A., Gonzalez. C., Garcia-Lopez, M.-L., Garcia-Fernandez, M.-C., Otero, A. (1994). Antibacterial activity of the lactoperoxidase system against Aerorrronns hydrophila in broth, skim milk and ewes' milk. Lett. Appl. Micr-obiol., 19:161-164. Santos, J. A., Lopez-Diaz, T. M., Garcia-Fernandez, M. C., Garcia-Lopez, M. L., and Otero, A. (1995). Antibacterial effect of the lactoperoxidase system against Aeronzorzns hydrophila and psychrotrophs during the manufacturing of the Spanish sheep fresh cheese Villalon. Milchwissenschnj?, 50:690-692. Kaper, J. B., Lockman, H., and Colwell, R. R. (1981). Aer-ortzoitns hydrophila, ecology and toxigenicity of isolates from an estuary. J. Appl. Bacteriol.. 50:359-377. Joseph, S. W., Janda, M., and Carnahan, A. (1988). Isolation, enumeration and identification of Aeromonas sp. J. Food Safety, 9:23-25. McCoy, R. H., and Pilcher, K. S. (1974). Peptone beef extract glycogen agar, a selective and differential Aeronzonns medium. J. Fish. Res. Board Can.. 3 1:1153- 1155. Havelaar, A. H., During, M.. and Versteegh, J. F. M. (1987). Ampicillin-dextrin agar medium for the enumeration of Aero.omonns species in water by membrane filtration. J. Appl. Bncteriol.. 62:279-287. Von Graevenitz, A., and Bucher, C. (1983). Evaluation of differential and selective media for isolation of Aeronlonas and Plesiomonas sp. from human feces. J. Clin. Microbiol., 17: 16-21. Janda, J. M., Dixon, A., Raucher, B., Clark, R. B., and Bottone, E. J. (1984). Value of blood agar for primary isolation and clinical implication of simultaneous isolation of Aeromonas hydrophila and Aerornonas ctlvine from a patient with gastroenteritis. Clin. Microbiol., 20: 1221-1222. Schubert, R. H. W. (1977). Uber den Nachweis von Plesionlonns shigelloides Habs and Schubert, 1962, und ein Elektivmedium, den Inositol-Brillantgriin-Gallesalz-agar. Ernst-Rodenwaldt-Arch., 4:97- 103. Schubert, R. H. W. (1967). Das vorkommen der Aeromonaden in oberirdische Gewassern. Arch. Hyg., 1501688-708. Roland, F. P., Salt-starch-xylose-lysine deoxycholate agar. A single medium for the isolation of sodium and non-sodium dependent enteric gram negative bacilli. Med. Microbiol. Irnmunol., 1631241-249. Rogol, M., Sechter, I., Grinberg, L., and Gerichter, C. B. (1979). Pril-xylose-ampicillin agar, a new selective medium for the isolation of Aeronlonas hydrophila. J. Med. Microbiol., 12: 229-23 1. Shread, P., Donovan, F., and Lee, J. (1981). A survey of the incidence of Aeromonas in human feces. Soc. Gen. Microbiol. Quart., 3:184. Altorfer, R., Alwegg, M., Zollinger-Iten, J.. and von Graevenitz, A. (1985). Growth of Aeromonas spp. on cefsulodin-irgasan-novobiocin agar selective for Yersinia enterocolitica. J. Clirt. Microbiol., 22:478-480. Hoban, D., Forsyth, W., Gratton, G., and William, T. (1981). Detection of Aeromonas hq" drophila from diarrhea stools using Macconkey Tween 80 agar. Abst. Ann. Mtg. Ant. Soc. Microbiol. C4. Hoban, D. (1983). Survey of diarrheal illness associated with Aeronzonas hyd?-oplzilain Manitoba. Abst. Ann. Mtg. Am. Soc. Microbiol. C151. Shotts. E. B., Jr., and Rinder, R. (1973). Medium for the isolation of Aeronzonas hydrophila. Appl. Microbiol., 26:550-553.
This Page Intentionally Left Blank
4 Update: Food Poisoning and Other Diseases induced by Bacillus cereus Kathleen T. Rajkowski and James L. Smith U.S. Department of Agriculture, Wyndnoor, Pennsylvania *
I. Introduction 61 11. Foodborne Bucillus cereus Outbreaks62
111.
IV. V. VI. VII. VIII.
IX.
1.
A. Guidelinesforidentifying B. cereus outbreaks63 B. B.cereus foodborneincidentsreported in 1992-199763 Characteristics of the Food Poisoning Toxins Produced by Bacillus cereus A. Emetic toxin 65 B. Diarrheic toxin(s) 66 Nongastrointestinal Disease Induced by Bacillus cereus Infection 67 Occurrence inFood 67 Growth and Survival 70 Inhibition and Inactivation7 1 Isolation, Identifications and Characterization 7 1 FutureResearchNeeds72 References 73
65
INTRODUCTION
The endospores of the genus Bacillus are ubiquitous in the environment and are capable of surviving adverse conditions. To date, Norris et al. (1) provide the most comprehensive list of conditions and environments from which Bucillus spores can be isolated. Since the spores are air-, soil-, and/or waterborne, it is only natural that various Bacillus spp. are isolated from foods and areas where food is produced. Since its first isolation in 1948
* Mention of brand or firm name does not constitute an endorsement by the U.S. Department of Agriculture over others of a similar nature not mentioned. 61
62
Rajkowski and Smith
associated with a foodborne disease outbreak, Bacillus cereus is considered the most common pathogen in the genus. Toxins produced by the vegetative cells lead to gastrointestinal upset and/or vomiting. In the United States, B. cereus foodborne diseases are a relatively minor concern: however, they are still a major concern worldwide. Recent reviews on various aspects of B. cereus and its disease potential have been published by Drobniewski (2), Fermanian et al. (3), Granum ( 4 3 , Granum and Lund (6), and Schultz and Smith (7). This chapter will summarize the reports of the more recent B. cerem foodborne outbreaks, toxin characterization, nongastrointestinal diseases caused by B. cereus infection, and the psychrotropic characteristics of B. cereus strains.
II. FOODBORNEBACILLUSCEREUS OUTBREAKS Data from the surveillance for foodborne disease outbreaks in the United States indicate that B. cereus is not a major concern. For the 10-year period 1983-1992, there were a total of 1396 bacterial foodborne outbreaks, with only 37 of those outbreaks due to B. cereus. Also, during that time period, there were a total of 83,5 10 cases of bacterial foodborne diseases, with 694 cases occurring in the B. cer-eus outbreaks. Thus, B. cereus was the cause of 2.7% of the bacterial foodborne outbreaks and 0.83% of the outbreaks cases. Death does not appear to be a consequence of foodborne B. cereus outbreaks (8,9). While foodborne disease due to B. cereus is not commonly reported, it is probable that the disease is underreported because most cases are sporadic. The ingestion of contaminated Chinese food and fried rice was responsible for almost half (45.9%) of the B. cereus foodborne outbreaks in the United States. Most outbreaks (67.6%) occulred during the months of May to October (8,9). During the 10-year reporting period, 43.2% of the B. cereus outbreaks occurred in commercial food establishments (cafeterias, delicatessens, restaurants) and 21.6% occurred in homes. In 28 of the 37 outbreaks, the factors (or errors) contributing to B. cereus outbreaks were reported. In 27 of 28 outbreaks (96.4%), improper holding temperature of the food was a contributing factor. The use of contaminated equipment, inadequate cooking, poor personal hygiene, and obtaining food from an unsafe source were contributing errors in 17.9, 17.9, 14.3, and 3.6%, respectively, of foodborne B. cereus outbreaks (8,9). Wilson et al. (10) indicated that Bacillus food poisoning has increased in Northern Ireland. During the 1 9 8 0 ~only ~ one Bncillus-induced foodborne incident was reported; however, nine incidents occurred in 1991 and four in 1992. B. cereus was isolated in most incidents. Bacillus subtilis was isolated in four incidents, however, B. cereus also was isolated from the food in two of those incidents. The foods involved were generally rice and/or chicken and in 10 of 14 incidents, the food was eaten in or taken out from Chinese restaurants. Clinically, the symptoms in almost all of the cases involved in the incidents were predominately vomiting and nausea with an incubation period of 1-6 hours. Thus, it appears that the emetic toxin was responsible for the majority of incidents occurring in Northern Ireland (10). Most of the incidents occurred during the months of April to August. It is probable that inadequate temperature control of cooked rice was the food-processing error that contributed to most of the Bncillrrs-induced food poisoning in Northern Ireland. The mean number of foodborne outbreaks in Taiwan for the period 1987-1993 was 81 (range: 57-93); however, in 1994 the number of outbreaks reported totaled 102 (1 1). A bacterial agent was identified in 74 of the 1994 outbreaks. Eleven of the outbreaks
Bacillus
cereus
63
were due to B. cereus (14.9%) exceeded only by Vibrio parahaemolyticus (56.7%) and Staphylococcus aureus (20.3%). Snln?onella spp. accounted for only 8.1% of the foodborne outbreaks of bacterial origin (l 1). Food poisoning incidents involving B. cereus are common in the Netherlands. For the four year period 1991-1994, there were 124 incidents with 789 cases of foodborne disease with a bacterial etiology. B. cereus was responsible for 40 of 124 (32.3%) incidents with 172 of 789 (21.7%) cases. Chinese-Indonesian food accounted for 17 of 40 (42%) of the B. cereus-induced food poisoning incidents (12). Salmonella spp. caused fewer incidents (31/124) than B. cereus; however, there were more cases (290/789) with Salmonelln.
A.
Guidelinesfor Identifying B. cereus Foodborne Outbreaks
Two or more persons must become ill for an incident to be called a foodborne disease outbreak (9). In order to confirm that an outbreak is due to B. cereus, organisms must be isolated from the stool of two or more ill persons and not from the stool of controls; alternatively, an outbreak can be confirmed by isolation of 2 10’ B. cereleMs organisms from epidemiologically implicated food (provided that the food specimen has been properly handled). The incubation period for illness induced by emetic toxin is 1-6 hours. Vomiting is the major symptom seen in patients; some patients have diarrhea, but fever is uncommon. The incubation period for persons ingesting diarrheal toxin is 6-24 hours, with patients having diarrhea and abdominal cramps. A few patients show vomiting; fever is uncommon (9).
B. B. cereus Foodborne Incidents Reported in 1992-1 997 1. Two individuals developed gastrointestinal symptoms after eating spaghetti with homemade pesto. Both emetic and diarrheic toxin were involved in this incident. A large batch of the food had been prepared 4 days before the incident and although stored refrigerated, the food had been permitted to remain at room temperature for lengthy periods during the 4-day period. The preparation of a large batch of food which was intermittently subjected to temperature abuse over a 4-day period contributed to the outbreak (13). The outbreak was interesting because one patient died of fulminant liver failure induced by emetic toxin. 2. B. cereus was present in vegetable pakora (4.4 X lo5 CFU/g) and fried rice (2.0 X lo7 CFU/g) in two foodborne incidents involving five people who had eaten in an Indian restaurant in Scotland. All isolates belonged to the same enterotoxigenic serotype. There were a number of processing errors: rice was prepared too far in advance and in greater quantities than needed for a given safe time period. The cooked rice was not properly cooled and was at improper holding temperatures for long periods. The presence of B. cereus in foods other than rice indicated cross-contamination, and a high coliform count in foods indicated handling by person(s) with poor personal hygiene (14).
64
Rajkowski and Smith
3. At a college sport dayin Thailand, 470 individuals were ill with vomiting, nausea, and abdominal pain; approximately half ofthe individuals reported diarrhea. The ingestion of cream-filled eclairs was significantly associated with illness. The mean incubation period was 3.2 hours, which suggested a preformed toxin in thefood. Initial laboratory investigation indicated presence of B. cereus in the eclairs. A few days later, a different laboratory reported the presence of enterotoxin-producing S. aureus. A dual etiology for the outbreak may have been possible; however, before the eclairs had been sent to second laboratory, they had been at room temperature for 12 hours. In addition, the phage types of S. nureus isolated from the cooks were not similar to the phage type isolated from eclairs. Therefore, it is probable that the outbreak was due to B. cereus emetic toxin. The failure to refrigerate the eclairs after preparation and during serving led to the outbreak (15). 4. A stew ingested by 142 competitors at a Norwegian Ski Championship was the source of B. cereus-induced diarrhea. B. cereus was present in the stew at 2 X los CFU/g. Theincubation time was 12-36 hours, and most patients recovered within 24 hours. The kitchen in which the food was prepared was contaminated profusely with B. cereus. The problem was eliminated by thorough cleansing and decontamination of the kitchen with hypochlorite (16). 5. An outbreak of diarrheic foodborne disease involved 139 individuals attending a university field day barbecue in South Carolina. B. cereus was isolated from barbecued pork at levels > lo5 CFU/g. At least half of the pork was cooked 2 days in advance of the barbecue and held unrefrigerated for 18 hours before boning. The boned barbecued pork was then stacked 20 cm deep in metal trays and refrigerated. On the day of the barbecue, the pork was heated; however, the heating unit was not working properly. A hot temperature was not maintained during transportation to the site and while serving the pork to attendees. Preparation of the pork barbecue too far in advance and failure to maintain safe storage and serving temperatures were the food-processing errors that led to the outbreak (17). 6. Emetic toxin from B. cereus was responsible for food poisoning in 14 individuals (12 children and 2 adults) from two child care centers located in Virginia. The ill individuals had eaten chicken fried rice at a catered luncheon prepared by a local restaurant. Leftover chicken fried rice contained B. cereus at > lo6 cfu/g. Improper cooling of the cooked rice and improper holding temperatures contributed to the outbreak (18). 7. A catered wedding reception in California was the site of an outbreak of B. cereus food poisoning due to diarrheic toxin. The outbreak involved 55 individuals, with Cornish hen as the suspect food. The birds contained B. cereus at a level of 1.4 X lo6 CFU/g. There were several errors committed by the caterer which played a role in the outbreak. He was licensed to serve only 29 restaurant customers, yet on this occasion he was serving over 300 customers at two different locations on the same day. It was probable that the Cornish hens were not properly thawed, resulting in undercooked birds. In addition, the cooked Cornish hens were transported and held in an unrefrigerated van for several hours. That the caterer lacked facilities for large-scale food preparation was the chief reason for this outbreak (19).
Bacillus
cereus
65
Symptoms found in the above outbreaks included diarrhea without vomiting or vomiting that was sometimes followed by diarrhea. The diarrheal toxin does not appear to induce vomiting, but on occasion the emetic toxin appears to 'give both vomiting and diarrhea. It would appear that under some conditions, emetic toxin-producing strains may also produce a diarrheal-type toxin. In the outbreaks listed above, the error in food preparation appeared to be improper holding of the food at temperatures high enough to allow growth of B. cereus. Prompt cooling of the food with refrigerated storage would have prevented the outbreaks. An interesting outbreak of B. cereus intoxication occurred in Germany, which involved Rottweiler puppies; B. cereus-contaminated milk powder was the food agent. Puppies fed formula containing milk powder (in addition to receiving mother's milk) became ill with vomiting and diarrhea. Five puppies, whose mother was not producing milk, were fed exclusively on formula containing the contaminated milk powder and died. B. cereus was present at a level of 2.7-3.5 > lo4 CFU/g in the milk powder. The presence of the organism in the powdered milk and symptoms suffered by the puppies indicated that this outbreak was due to intoxication by B. cereus (20).
111.
CHARACTERISTICSOFTHE FOOD POISONING TOXINS PRODUCED BY Bacillus cereus
A. Emetic Toxin Emetic toxin is preformed in foods, i.e., B. cereus at levels of 105-108 CFU/g of food produce toxin, which causes vomiting when the food is eaten. Symptoms-vomiting, nausea and malaise-appear within 0.5 to 5 hours. Duration of illness is generally 6-24 hours (5,6). Foods implicated in B. cereus emetic food poisoning include fried and cooked rice, pasta, noodles, and pastry. Agata et al. (21) found a strong association of emetic toxin production with the H1 serovar phenotype of B. cereus. Isolates of B. cereus that produce enterotoxin activity or hydrolyzed starch did not produce emetic toxin. The structure of the emetic toxin has been elucidated, and the compound has been named cereulide. Cereulide is a 36-membered cyclic dodecadepsipeptide with the sequence of cycb(D-O-Leu-D-Ala-L-o-val-L-Val)3with the formula, C57H9601SN6 and a molecular weight of l 19l .6. Cereulide is a potassium ionophore closely related to valinomycin (22-24). Cereulide induces vomiting in Suncus murinus (house musk shrew) and rhesus monkeys (25,26). The emetic toxin is not antigenic, does not act in the rabbit or mouse ileal loop test, produces vacuolation in HEp-2 cells but is not cytotoxic, is stable for 90 minute to 121"C, is stable to pH 2.0 and pH 11, is not hemolytic, and does not induce vascular permeability in the skin of the rabbit (5,26). Cereulide induces vomiting when it stimulates the vagus nerve by binding to the 5-HT3receptor. Using the S. murinus emetic model, Agata et al. (25) found that vagotomy eliminated the emetic action of cereulide. The 5-HT3receptor antagonist, ondansetron, also abolished cereulide-induced vomiting in Suncus. In addition, cereulide is a mitochondrial poison. Using rat liver mitochondria, Mahler et al. (13) showed that emetic toxin inhibited electron transport and uncoupled oxidative phosphorylation.
,
66
Rajkowski and Smith
Nothing is known about the synthesis of cereulide. It is probable that the cyclic peptide is enzymatically synthesized from its component amino acids and is not a product of a gene.
B. Diarrheic Toxin(s) The toxin(s) responsible for the diarrheal syndrome are formed in the small intestine of the host. Unlike the emetic toxin, which is preformed in food, any preformed diarrheic toxin, due to its loss of activity at pH 3 and its sensitivity to trypsin, would be destroyed as it passes through the gastrointestinal system (27). The infectious dose ranges from lo5 to lo7 organisms per host. Symptoms include abdominal pain, watery diarrhea, and occasionally nausea, which generally appear within 8-16 hours after ingestion of contaminated food. Duration of illness is generally 12-24 hours (6). Foods that are frequently implicated in B. cereus diarrheic food poisoning include meat products, soups, vegetables, puddings, sauces, milk and milk products. Enterotoxin activity is decreased or lost if thetoxin is subjected to a pH of 3, trypsin, chymotrypsin, pepsin, or heating at 80°C for 10 minute (4,27). Baker and Griffiths (27) demonstrated that the immunological activity of the enterotoxin was approximately four times more stable when heated in milk as compared to broth. However, since diarrheic activity was not tested, it is not clear that the toxin still retained the ability to cause disease. Granum and Lund (28) studied 85 strains of B. cerem isolated from dairy products and found that 50 (58.8%) were enterotoxigenic, 13 (15.3%) were psychrophilic, and 6 (7. l %) produced toxin at 6°C. The ability to grow and produce enterotoxin at refrigerated temperatures indicates that refrigeration may not be enough to protect consumers from B. cereus food poisoning. Hemolysin BL (HBL), a tripartite entity consisting of three protein components, possesses hemolytic, cytotoxic, dermonecrotic, and vascular permeability activity (29,30). In addition, HBL induces fluid accumulation in rabbit ileal loops, thereby indicating that it is an enterotoxin. HBL consists of three protein components: L? (46 kDa), L, (38 kDa), and B (37 kDa), and maximal expression of all HBL activities requires all three protein entities (30). Heinrichs et al. (31) and Suwan et al. (32) have cloned the genes for HBL (hbZC, hblD, IzblA, hZaB), determined the nucleotide sequence of the genes, and deduced the amino acid sequences of the three protein components. Lund and Granum (33,34) characterized a nonhemolytic enterotoxin complex from strains of B. cereus responsible for outbreaks of diarrheal food poisoning. The nonhemolytic enterotoxin (NHE) consisted of three protein moieties: B (105 m a ) , L2 (54 kDa), and L, (39 m a ) , and all components were needed for maximum cytotoxicity. Apparently, NHE was not tested in the ileal loop assay. The sequencing of NHE is underway (34). Diarrhea-inducing strains of B. cereus may produce either HBL or NHE or both (34). It would appear that there are at least two tripartite enterotoxins, HBL and NHE, in B. cereus. The role of “enterotoxin T” as an agent of diarrhea in B. cereus foodborne illness is not clear. Enterotoxin T consists of a single protein that is cytotoxic, positive in the mouse ileal loop assay and possesses vascular permeability activity (35). Enterotoxin T is a product of the gene bceT, which has been cloned (35). Granum and Lund (34) state that there is no real evidence that enterotoxin T causes food poisoning. Two commercial kits are available for the detection of B. cereus enterotoxins: the Oxoid BCET-RPLA and Tecra ELISA kits. The Oxoid kit measures L? of HBL, whereas t
c
. .
, J
. ~
. ..
Bacillus cereus
67
the other kit measures L: of NHE: thus, the kits do not assay the same enterotoxin (34). But it would appear that both kits are necessary to detect enterotoxin production in food poisoning due to B. cereus. Beecher and Wong (36) developed a simple and inexpensive method for detecting HBL-producing strains of B. cereus. Their assay is based on a distinctive discontinuous hemolytic pattern produced by HBL-producing strains on blood agar plates.
IV. NONGASTROINTESTINALDISEASEINDUCED Bacillus cereus INFECTION
BY
In addition to causing classical food poisoning symptoms affecting the gastrointestinal system, B. cereus may cause disease in other areas of the body. Nongastrointestinal diseases that may result from B. cereus infections include (a) local infections such as ocular and wound infections, (b) bacteremia and septicemia, (c) central nervous system infections, (d) respiratory tract infections, and (e) endocarditis (2). Details from a number of case reports (1995-1997) of B. cereus-induced nongastrointestinal diseases are given in Table 1. Drobniewski (2) has a table listing cases of B. cereus-induced ocular infections that occurred during the period 1951-1990; another table lists other nongastrointestinal B. cereus infections that occurred during 1965-1992 (2). A table of clinical features that appeared in cases of B. cereus infections occurring in patients with leukemia during 19711995 is presented by Akiyama et al. (41). The cases of nongastrointestinal B. cereus infections presented in Table 1 and in the tables of Drobniecwski (2) and Akiyama et al. (41) indicate that individuals immunocompromised either by illness or medication are at particular risk for disease induced by this organism. Using both in vitro and in vivo methods, Beecher et al. (45) demonstrated that the B. cereus HBL enterotoxin had retinal toxicity and may be involved in endophthalmitis (inflammation of the inner eye). B. cereus is one of the major infective agents responsible for blindness in humans (45). Antibody neutralization of HBL factor in cell-free crude concentrated and dialyzed culture filtrates eliminated approximately 50% of the in vitro retinal toxicity. Purified HBL wasless toxic than culture filtrates containing the same level of HBL. These results indicate that B. cereus ocular virulence may be multifactorial and requires other toxic moieties in addition to HBL (45). Fulminant liver failure ending in death was apparently induced by the B. cereus emetic toxin cereulide (13). Purified emetic toxin from the disease causing strain inhibited electron transport and uncoupled oxidative phosphorylation in rat liver mitochondria (13). Many liver failure syndromes have been linked to mitochondrial injury and disruption of mitochondrial activity (40). Thus, cereulide acting as a mitochondrial poison led to liver failure in the patient described by Mahler et al. (13). The apparent correlation of nongastrointestinal disease syndromes with B. cereus diarrheic and emetic toxins warrants further study.
V.
OCCURRENCE IN FOOD
Some food sources from which B. cereus has been isolated are summarized in Table 2. In a more recent report, Hatakka (58) examined the microbiological quality of airline hot
68 3
d.
0
.3 Y
h
?j s
"
a
8
c .,-
c
0
.3 . 3
m
.n
Y
8
U
Q
2
4
Rajkowski and Smith
Bacillus cereus
69
Table 2 Survey of Food Products from Which Bacillus cereus was Isolated and % Incidence Food product
5% Incidence
Rice and rice products Boiled rice Spices Milk products Chicken and meat products Buffalo meat Ready-to-serve moist food Powdered infant formula Nonfat dry milk Skim milk powder Buesco Blanco cheese soft hard Various food samples from a Spanish retail market Seed samples (alfalfa, mung bean, wheat or seed mixtures) Poultry products Pork carcasses Beef carcasses Chicken carcasses Raw beef Pork Chicken Ham Sausage Uncooked hamburger Cooked hamburger Meat product additives Tempeh samples
28.5 100 30 100 80 35 83 75 100 10.3 69 63 14.7 cells 82.4 spores 67.3
6.9 7 11 0 7.9 4.4 7.2 3.8 14.9 45.5 23.7 39.1 16
Country
Ref.
India
47
India United States
48 49
Japan
50
Venezuela
51
Spain
52
United States
53
England Sweden
54 55
Japan
56
Netherlands
57
meals over a 3-year period and found that 3% of the meals contained pathogenic bacteria, of which B. cereus was the most common pathogen. Using the fatty acid profiles for the identification of 229 B. cereus from environmental sources (dairy farm to processing plant), Lin et al. (59) determined that B. cereus spores in raw milk were the major source of B. cereus in pasteurized milk based on the comparison of the profiles from environmental and pasteurized milk isolates. The report of teGiffel and Beumer (60) confirmed that the presence of psychrotrophic B. cereus spores in raw milk resulted in contaminated pasteurized milk products. The vegetative cells in raw milk are rapidly killed during pasteurization at 65OC, whereas the endospores of B. cereus are heat-resistant (61). Properly pasteurized dairy products, when stored at recommended refrigerated conditions (maximum 7°C) and used within the expiration date usually cause no problems for healthy adults. More information can be found in the bulletins of the International Dairy Federation devoted to B. cereus in milk and milk products (62,63).
Rajkowski and Smith
70
VI.GROWTH
AND SURVIVAL
The Agricultural Research Service of the U.S. Department of Agriculture has developed a growth model for B. cereus, which can be found on the Internet at http:// www.arserrc.gov/mfs. The Pathogen Modeling Program Version 5.1 forWindows can be entered through the home page. The growth characteristics of B. cereus were reviewed by Schultz and Smith (7). The incidence of B. cereus in raw and processed foods, especially when held at low temperatures, has received considerable attention. The research data indicate that few foods are free of the organism and that growth and toxin fortnation can occur at psychrotropic temperatures (28). Bergann (64), reporting on the growth temperature requirements of 50 B. cereus strains, showed that more than half of the strains grew at 10°C, six at 8"C, and one at 6°C. Garcia-Armesto and Sutherland (65), after characterizing both psychrotropic and nlesophilic Bacillus species isolated from milk, found one B. cereus isolate that was clearly psychrotropic (grew at 6.5"C but not at 40°C in 2 days). The cold adaptation response was demonstrated by Foegeding and Berry (66) after determining the growth characteristics of 27 B. cereus clinical and food isolates. This cold adaptation was also reported by Christiansson et al. (67). They found that some strains of B. cereus were acclimated to grow at 8"C, and when the relationship between cytotoxicity and growth of these isolates was compared, most became cytotoxic at approximately lo* c f d n i l . The dairy isolates that produced cytotoxicity in milk at 8°C under aerated conditions also gave positive vascular permeability reaction with rabbits, which is indicative of B. cereus diarrheal toxin (67). In the study by Granum et al. (28), they found of the 85 B. cereus strains isolated from dairy products, 6 were both psychrotropic and enterotoxigenic and exhibited good growth at 6°C. While B. cereus is commonly found in milk and other dairy products and milk is a suitable substrate for toxin production under aerated conditions even at low temperatures, the incidence of B. cereus food poisoning in milk is low probably due to insufficient aeration during normal storage (67). Van Netten et al. (68) found that B. cereus isolates involved in two foodborne outbreaks of diarrheal illness were able to grow and produce toxin at 4-7°C. Strains of B. cereus of the emetic type, which were isolated during an outbreak involving pasteurized milk, were able to grow at 4-7"C, but enterotoxin production was not detected using the Oxoid test kit for diarrheal toxin. It is quite probable that the emetic toxin can beproduced at low incubation temperatures, although they did not test for it (68). Fermanian et al. (69) reported that a foodborne and clinical isolate grew at 10°C and that both isolates were toxin positive using the Oxoid kit. The presence of psychrotropic B. cereus strains in dried infant foods is a concern. Rowan and Anderson (70) reported that of the B. cereus isolated from reconstituted milkbased infant fornlulas (MIF), 1,4, and 16 strains grew at 4, 6, and 8"C, respectively, after 15 days. They were tested for diarrhea enterotoxin production, and 9 of the 21 isolates produced enterotoxin. The authors recommended brief refrigeration (4°C) storage of the reconstituted MIF forthis product to remain safe. They also tested the effect of maltodextrin on the growth and synthesis of the diarrheal enterotoxin in MIF and found that supplementing with 20.1% maltodextrin supported both growth and diarrheal toxin production when incubated for 14 hours or more at 25°C (71). Jaquette and Beuchat (72) reported that psychrotrophic B. cereus can survive and grow in reconstituted infant rice cereal. They recommended immediate consumption or holding the food for 99% of the population was injured (137). When either catalase or pyruvate was added to TSAS, however, enumeration of the heat-injured cells increased 14,000-fold, to approximately 76% of that found on TSA. Catalase addition to various other staphylococcal selective agar media was also effective in increasing enumeration ( 138). Addition of catalase concurrent with the presence of phosphatidylcholine and beef extract in Vogel and Johnson agar resulted in a medium that gave equal or better enumeration than Baird-Parker agar (164). Baird-Parker agar is currently recommended by the Food and Drug Administration (10) for the enumeration of S. nurez4s. Brewer et al. (163) developed a modified most probable number procedure employing either catalase or pyruvate addition. This method frequently detected the presence of low numbers of S. aureus missed by the currently accepted procedure. The observation that catalase or pyruvate increased the enumeration of heat-stressed S. nurezrs suggested that both ofthese agents were acting through the degradation of H701. Bucker et al. (165) allowed heat-injured cells to recover both aerobically and anaerobically. Cells allowed to recover anaerobically showed no sensitivity to NaC1, suggesting that although catalase levels were depressed, little or no hydrogen peroxide was formed. The optimal pH for S. cu4reus MF-3 1 catalase was 5-6 (130,166). The enzyme was stable over the pH range of 4-9. The apparent isoelectric point was 5.3 +- 0.1 pH units. With respect to temperature stability, purified S. awezrs MF-3 1 catalase was stable at 52°C after 45 minutes of heating. The enzyme was inactivated at 60°C for 10 minutes. However, this was dependent upon the concentration of the enzyme and the presence of protectants. The apparent subunit molecular weight was 64,000 2 1,000 daltons, whereas the apparent native molecular weight was 235,000 +- 5,000 daltons. Amino acid analysis revealed that S. aweus MF-31 catalase was similar to amino acid analyses of catalase from other sources. The iron content of S. mrezrs MF-31 catalase was found tobe 0.098%. The enzyme was inhibited by millimolar concentrations of sodium cyanide, sodium azide, and hydroxylamine, indicating the presence of a heavy-metal catalyst. The effects of salt, pH, and salt concentration also influenced activity. The chloride anion inhibited catalase activity at low pH, but this inhibition was influenced by the cation. Sodium chloride was more inhibitory at low pH than either potassium chloride or magnesium chloride. Superoxide dismutase activity has been shown to be lowered following sublethal heating of S. nzueus MF-31 (52"C, phosphate buffer) (167). When cells were heated for 90 minutes, SOD levels dropped approximately 20%, during which time >99% of the cells were killed. The observed loss in activity was not great, but this decrease in concert with a decrease in catalase levels may be significant. S. uurezls cells were sublethally heated and the SOD levels assayed both during the injury and the recovery periods. There was a significant (-85%) drop in SOD activity at the start of the recovery period. This decrease closely approximated the decrease observed in catalase activity following a heat
Staphylococcus aureus
367
stress. The levels remained depressed for approxitnately 2.5 hours until the cells began to multiply. After 2.5 hours, the SOD levels increased concurrently with increases in cell numbers.
VI.
DETECTIONANDENUMERATION
There are four reasons why a food or any ingredient is examined to determine the presence of S. aureus (130): 1. To confirm the presence of S. aweus following a food-poisoning outbreak 2. To determine whether or not a food is a potential source of staphylococcal food poisoning 3. To demonstrate postprocessing contamination 4. As part of a routine quality-control program It has been observed that when staphylococcal cells are sublethally stressed, many are no longer able to grow on selective media. The reasons for this failure, and the methods of overcoming it, have been extensively studied in S. c u u m s . The choice of the method for the detection of S. aureus depends upon the purpose for conducting the test and the product involved. When food is suspected as the source of a staphylococcal food-poisoning outbreak, large numbers of S. nureus are frequently present and sensitive methods may not berequired. More sensitive methods are required to detect small populations of S. aweus, which may bepresent as the result of postprocessing contamination. In many cases, S. aureus is not the sole or even the predominant organism present in a sample. For this reason, selective inhibitory media are employed for isolation and enutneration. Selective media utilize a number of different toxic chemicals to achieve selectivity. Some of the ingredients used include sodium chloride, tellurite, lithium chloride, and various antibiotics. A number of media have been suggested for the isolation of S. aureus from food when more than 100 per gram may be present. Some of these include Staphylococcal Medium 110, Vogel-Johnson Agar, egg yolk-sodium azide agar, tellurite-polymyxin-egg yolk agar, and Baird-Parker Agar (Table 7). Composition of typical staphylococcal selective media is found in Table 8.
Table 7 Examples of Selective Media for Staphylococcus nureus Agar medium Staphylococcus Medium 110 Vogel-Johnson
Egg yolk-sodium azide Baird-Parker
Selective agent Sodium chloride Lithium chloride Potassium tellurite Glycine Lithium chloride Potassium tellurite Polymyxin B sulfate Lithium chloride Potassium tellurite
Diagnostic agent Mannitol Gelatin Mannitol Tellurite Phenol red Egg yolk Tellurite
Martin et al.
368 Table 8 Staphylococcal Selective Media Composition Component Peptone Yeast extract Beef extract LiCl Glycine Sodium pyruvate Potassium phosphate dibasic Mannitol Phenol red Phosphatidyl choline DNA Agar Tellurite ( 1%) Egg yolk (50%) Catalase
B-P(g/L) PCVJ(g/L) VJ(g/L) pH 7.0 10 1 5 5 12 10
pH 7.2
pH 7.2
10 5 5 5 10
10 5
5 10 -
5
-
1s 10 mL 53 mL
10 0.025 2 2 16 10 mL
10 0.025 -
16 20 mL
-
780 units
B-P = Baird-Parker agar; PCVJ = phosphatidyl choline-Vogel and Johnson agar; VJ = Vogel and Johnson agar.
Most selective media are suitable for the enumeration of nomlal or unstressed S. aureus. However, due to processing, preservation, or other adverse conditions, sublethal stress may occur, resulting in the increased sensitivity of S. aur-eus to the selective agents. Because injured cells exhibit an increased sensitivity to selective agents, S. nureus may go undetected in conventional selective enumeration procedures. Baird-Parker and Davenport (168) demonstrated that the recovery of heated or dried cells of S. aureus may be lost or its activity reduced by heating or drying and that blood, which contains catalase, or the addition of pyruvate, helped in the enumeration by destroying H202produced by recovering cells. It has been found that Baird-Parker (B-P) agar is most satisfactory in enumerating injured cells when compared with other staphylococcal selective media (169-171). The addition of catalase to tryptic soy agar plus 7% NaCl (TSAS) (Table 9) and other selective media can increase the enumeration of themlally stressed S. aureus cells, while the addition of heat-inactivated catalase had little effect on enumeration (130,137,138). The addition of catalase to other selective media has been found to increase the enumeration of both heat-stressed and unstressed cells (138). Carlsson et al. (172) suggested that when phosphate and glucose were autoclaved together in a culture medium at neutral or alkaline pH, products resulted that rapidly auto-oxidized, forming these reactive species. Attempts have been made to develop a staphylococcal-selective medium that gives an enumeration equal to that of B-P agar while overcoming difficulties in the use of BP agar (e.g., with milk products) and its expense as described by Martin (130). Andrews and Martin (164) modified Vogel and Johnson agar by the addition of 0.5% beef extract, 0.2% DNA, 0.2% phosphatidylcholine (lecithin), and 780 units of catalase spread on the agar surface prior to inoculation. They found this phosphatidylcholine-Vogel and Johnson
Staphylococcus aureus
369
Table 9 Catalase and Enumeration of Thermally Stressed S. uureus MF-3 1 Cells3 Unstressed cells
Stressed cells %
%
Medium
CFU/mL
B-P VJ VJ + cat' TSA TSA + cat' TS AS TSAS + catc MSA MSA + cat' S1 10 S1 10 + cat' TPEY TPEY + catc
3.9 3.5 4.2 3.4 3.9 3.1 3.5 3.0 3.4 2.2 2.5 4.2 4.3
x lo9 X
X
x X X
x X X
109 lo9 lo9 109 109 lo9 109 109 109
X X 109 X 10' X
109
enumeration' 100 90 108 87 100 79 90 77 87 56 64 108 110
CFU/mL 2.4 X 6.2 X 2.9 X 1.9 X 2.5 X 1.8 X 1.2 X 2.5 X 7.2 X 2.3 X 3.7 x 8.0 X 2.5 X
109 10' lo9 109 109 107 109 lo6 lo8 10' los los 109
enumerationh 100 26 121 79 104 0.8 50 0.1 30 0.1 15 33 104
CFU = Colony-forming units; B-P = Baird-Parker agar; VJ = Vogel and Johnson agar; TSA = Tryptic soy agar; TSAS= Tryptlc soy agar 7% NaCl; MSA = mannitol salt agar; S1 10 = staphylococcal 1 10 agar; TPEY = tellurite polymyxin egg-yolk agar. Cells were heated in 100 mM potassium phosphate buffer (pH 7.2) at 52°C for 15 minutes. h Percentage of enumeration was calculated by dividing CFU/mL on the various media by CFU/mL on B-P and nlultiplying by 100. c Catalase (cat) activity was about 780 units per plate.
+
agar (PCVJ) gave an equivalent enumeration of stressed S. nureus cells and staphylococci from naturally contaminated food samples to that found for B-P enumeration. Enumeration with this medium was easier than with B-P and allowed ascertaining the production of DNase. Idziak and Mossel (173) modified the B-P agar formulation by replacing the egg yolk with pig plasma (B-PP) and found that, based on selectivity, diagnostic characterization, and increased sensitivity, B-PP agar was superior to B-P agar. Lachia (174) described a simplified method for the enumeration of S. nureus from food. He replaced the egg yolk in B-P agar with Tween 80 (0.05% wthol) and MgC12 (0.1%). When these compounds were added to the egg yolk-free B-P agar, the recovery of stressed cells were comparable to recovery on complete B-P agar. Mentzer-Morgenstern and Katzenelson (175) developed a single-step staphylococcal selective medium identified as 4-S agar. This medium permitted the isolation and identification of staphylococci and was achieved in a single step. Coagulase-positive staphylococci form small, grey to dark grey colonies surrounded by a dense, white opacity. Unfortunately, heat-stressed cells were inhibited by this medium, and a 3-hour preincubation period in brain heart infusion was required to enumerate these injured cells. This medium is reported to be very selective for S. uureus. When low numbers of S. nureus are expected in a food sample, a most-probablenumber (MPN) procedure is generally employed. The MPN technique is considered more efficient in the enumeration of low numbers of organisms or when high levels of competing
Martin et al.
370
organisms are present (176-178). An MPN value is an estimate of the population and not a precise enumeration of viable organisms. Microbiological counts are reported as “number of microorganisms per quantity of sample by MPN method.” Strict interpretation of the confidence limits for a MPN value of 2O/g, for example, asserts that the true population density lies between 7 and 89/g in 95% of all samples (179). Because of its selectivity, NaCl (10%) has been incorporated into tryptic soy broth (TSBS) in a MPN procedure. After 48 hours in TSBS, suspected tubes must be streaked onto B-P agar for an additional 48 hours for confirmation (180). The enumeration of injured S. c~ureuscells in TSBS has been shown to be greatly depressed (163,180). Brewer et al. (163) found that the addition of 1% pyruvate or catalase significantly increased the enumeration of stressed cells. Because of the requirement to add catalase after autoclaving, pyruvate is considered more desirable. Other staphylococcal MPN media have been examined (181,182).
A.
DetectionMethods
1. High Numbers When a sample is thought to contain 21000 S. c~ureuscells per gram, the most widely recommended enumeration medium is B-P agar (183). The samples are suspended in sterile diluent and 1.0 mL is spread-plated in triplicate on B-P agar; plates are dried, inverted, and incubated for 48 hours at 35°C. Typical S. uureus colonies are black to dark grey, circular, smooth, convex, moist, frequently with a light-colored margin, surrounded by an opaque zone of precipitation, and frequently with an outer clear zone. The colonies have a buttery to gummy consistency (184). Unfortunately, the zones are not always apparent, causing some S. aurezls cells to be missed. 2. Low Numbers The Association of Official Analytical Chemists (AOAC) procedure for the detection of low numbers of S. nureus is a MPN procedure that utilizes MPN tubes containing TSBS and 1% sodium pyruvate (185). In a collaborative study, this method was shown to significantly increase the enumeration of low, middle, and high levels of S. nureus from naturally contaminated products and is highly selective. The MPN tubes are inoculated from appropriate dilutions and are incubated at 35°C for 48 hours. The tubes are confirmed positive by streaking on B-P agar incubated at 35-37°C for 48 hours. Lancette et al. ( l 85) reported that the addition of 1% pyruvate to TSBS in a MPN procedure gave a significantly better enumeration of both artificially and naturally contaminated foods than did TSBS alone. They also found that the addition of pyruvate to TSBS increased the recovery of heat-stressed and nonstressed staphylococci. Their results are similar to those of Brewer et al. (163), with both catalase and pyruvate giving improved recovery. For confirmation, the positive MPN tubes should be used to inoculate B-P agar, followed by biochemical tests.
B. Confirmatory Tests Confirmatory tests for the positive identification of S. aureus include gram-positive cocci, production of P-hemolysis on blood agar, the coagulase test, the catalase test (production of 0, gas during the degradation of H202),anaerobic utilization of glucose and mannitol, and the production of thermostable nuclease. An additional test frequently performed is
Staphylococcus aureus
371
to determine the sensitivity of the suspect isolate to the enzyme lysostaphin. Lysostaphin specifically lyses cells of thegenus StnphyZococcus.Hawcroft and Geary (186) have found that restriction length polymorphism analysis of genes for ribosomal RNA (' 'ribotyping' ') is a useful tool for staphylococcal identification.
VII.
,
FOODPOISONING
Any food that provides S. uureus with the proper nutritional and environmental requirements is a potential source of foodborne illness. In the United States, frequently implicated foods include baked ham, pork, salads (meat, potato), pastries with cream or custard fillings, dried milk, and whey. In manyinstances, the source of contamination with S. nureus is from food handlers who have open lesions due to enterotoxin-producing S. aureus or who are asymptomatic carriers. Contaminated food-handling equipment and milk from cows with staphylococcal mastitis are also potential sources of the organism. Frequently, S. ~ u r e u sis introduced into the food as a result of post-processing contamination. The food is then improperly stored at temperatures, which allow S. nureus to grow and produce enterotoxin. One exception to postprocessing contamination can occur in the production of naturally fermented sausage. In this case, there may be sufficient growth of S. nureus for enterotoxin to be produced before it is inhibited by lactic acid, produced during sausage fermentation. The most common symptoms of staphylococcal food poisoning are nausea, vomiting, retching, abdominal cramps, and diarrhea. These symptoms may vary such that there may be vomiting but no diarrhea, or diarrhea but no vomiting. When severe cases occur, there may be, in addition to the symptoms described above, headache, muscle cramping, and prostration. Occasionally, low fever may be present in some victims. The mechanisms of action of SE are not entirely understood, but it is postulated that the toxin acts directly in the abdominal viscera. The onset of the symptoms, typical for food intoxication, generally occurs within 1-6 hours following consumption of the contaminated food. The average time for onset is 2-3 hours, but symptoms may develop in less than 1 or after 6 hours. The development of symptoms is determined both by the sensitivity of the victim and the amount of toxin consumed. Recovery is generally complete within 1-3 days in otherwise healthy individuals; however, the more severe the symptoms, the longer the recovery period. The mortality rate of staphylococcal food poisoning is extremely low. although death does occur, usually among the elderly or very young.
VIII.
PHAGECLASSIFICATION
Strains of S. aureus are routinely identified by bacteriophage typing. The first demonstration of a bacteriophage was in England by Twort in 1915; this phage had the ability to lyse cells of S. nureus. Since that time a set of typing bacteriophages has been compiled for the typing of S. cw-eus. Five groups, divided into host-range specificities, each containing a varying number of phage species, are differentiated by the response of S. nureus cells to the set of typing bacteriophages (Table 10). One drop of each standardized phage suspension (routine test dilution or RTD) is placed on an agar surface previously inoculated with a heavy inoculum of the staphylococcal strain to be tested. The pattern of the zones of lysis is recorded following overnight incubation at 30°C. The use of phage typing is a
" . ""
" P "
"
-_..
,..
.,
, .
.. -
-.
--I__.-.__"-
Martin et al.
372 Table 10 StaphylococcalTypingPhageGroups Phages group Phage Group I Group I1 Group I11 Group IV Miscellaneous
group within 29,52,52A,79,80 3A,3B,3C,55,71 6,7,42E,47,53,54,75,77.83A,84.85 42D 8 1,94,95,96
valuable epidemiological tool that can be used in determining the source of an enterotoxin producing strain. Types of S. nureus found in lytic group I11 are most frequently implicated in foodborne disease outbreaks. Not all cultures are typable by this procedure, and the susceptibility patterns of circulating strains vary in time and locality.
IX. SUMMARY S. aureus is a gram-positive bacterium that is capable of causing serious infections in susceptible individuals and is a leading cause of bacterial illness. Some of the important characteristics of S. aureus include its ability to survive and grow at reduced water activity and its ability to produce a wide variety of extracellular enzymes and toxins. Some of the more important include coagulase, heat-stable nucleases, penicillinase, hemolysins, toxic shock syndrome toxin, and staphylococcal enterotoxins. Staphylococcal enterotoxins are produced during the growth of S. nureus in foods, and < l pg can cause staphylococcal food poisoning in humans. These enterotoxins are heat-resistant and are produced under a variety of growth conditions. They are detected using either animal models or by immunological techniques. Cells of S. aureus can be sublethally injured by a number of treatments (heating, freezing, etc.) and sublethally injured cells become susceptible to environmental challenges to which unstressed cells show no effect. Oxygen and its toxic reduction products are involved in the failure of many staphylococcal selective media to enumerate and/or detect injured cells. Baird-Parker agar and a most-probable-number procedure using 10% NaCl and 1% pyruvate are recommended for the enumeration of samples containing high and low numbers, respectively, of S. cw-eus. Staphylococcal food poisoning is among the leading causes of foodborne disease outbreaks in the world. Symptoms include nausea, vomiting, diarrhea, and prostration.
REFERENCES 1. Sneath. P. A., Mair, N. S., Sharpe. M. E., and Holt, J. G. (1986). Section 12 gram-positive cocci. Bergey 'S Manual of SystenrtcticBacteriology, Vol2, Williams and Wilkins, Baltimore, MD, pp. 999-1103. 2. (1982). American Type Culture Collection.Catalogue of Strains I 15th ed., Rockville,MD. 3. Longree, K. (1972). Qualit?, Food Sanitation, Wiley-Science, New York. 4. Polledo, J. J. F., Garcia, M. L., Moreno B.,and Menes, I. (1985). Importanceof food handlers as a source of enterotoxigenic staphylococci. Zbl. Bakt. H y g . , I. Abt. Orig. B., 181:364. 5. Bergdoll, M. S. (1989). StapltylococcLls aurens. In Fooclbonle Bacterial Pathogens (M. P. Doyle, ed.), Marcel Dekker, Inc., New York, p. 463.
Staphylococcus
373
6. Youmans, G. P., Palerson, P. Y., and Sommers, H. M. (1980). The Biologic and Clinical Basis of hzfectiotrs Diseases 2nd ed., W. B. Saunders Co., Philadelphia, pp. 639-648. 7. Amin, U. S., Lash, T. D., and Wilkinson, B. J. (1995). Proline betaine is a highly effective osmoprotectant for Staphylococcus aweus. Arch. Microbiol., 163:138-142. 8. Iandolo, J. J. (1989). Genetic analysis of extracellular toxins of Staphylococcus aureus. Annu. Rely. Microbiol., 431375. 9. Bjorklind, A., and Arvidson, S. (1980). Mutants of Staphylococcus aureus affected in the regulation of exoprotein synthesis. FEMS Microbiol. Lett. 7:203. 10. (1984). Bacteriological Analytical Munuul, 6th ed. Division of Microbiology, Center for Food Safety and Applied Nutrition, U.S. Food and Drug Administration, Assoc. Off. Anal. Chem., Arlington, VA. 11. Weers-Pothoff, G., Moolhuijzen, C. E. M., and Bongaerts. G. P. A. (1987). Comparison of seven coagulase tests for identification of Staphylococcus nureus. Eur. J. Clin. Microbiol., 6(5):589. 12. Selepak, S. T., and Witebsky, F. G. (1985). Inoculum size and lot-to-lot variation as significant variables in the tube coagulase test for Staphylococcus aureus.J. Clin. Microbiol., 22(5): 835. 13. Shivram, S., and Kelkar, S. S.(1985). Comparative study of markers of pathogenic staphylococci. Indiun J. Med. Res., 82:194. 14. Gudding, R. (1983). Differentiation of staphylococci on the basis of nuclease properties. J. Clin. Microbiol., 18:1098. 15. Ratner, H. B., and Stratton, C. W. (1985). Thermonuclease test for samz-day identification of Staphylococcus aureus in blood cultures. J. Clin. Microbiol., 21:995. 16. Adekeye, J. D. (1984). Reliability of thermonuclease production for the identification of human and animal Staphylococcus aureus. Vet. Microbiol., 9:271. 17. Otero. A., Garcia, M. L., Garcia, M. C., Moreno, B., and Bergdoll, M. S. (1990). Production of staphylococcal enterotoxins C, and C2 and thermonuclease throughout the growth cycle. Appl. Emiron. Microbiol., 56(2):555. 18. Dack, G. M., Cary, W. E., Woolpert. O., and Wiggers, H. (1930). An outbreak of food poisoning proved to be due to a yellow hemolytic Staphylococcus. J. Prev. Med., 4:167. 19. Reiser, R., Conaway, D., and Bergdoll, M. S. (1974). Detection of staphylococcal enterotoxins in foods. Appl. Microbiol., 27(1):83. 20. Bergdoll, M. S. (1983). Enterotoxins, In Staphylococci and Staphylococcal Infections., Vol. 1 (C. S. F. Easman and C. Adlam, eds.). pp. 559-598. 21. Sugiyama, H., and Hayama, T. ( 1965). Abdominal viscera as site of emetic action for staphylococcal enterotoxin in the monkey. J. Infect. Dis., 115:330-336. 22. Marrack, P., and Kappler, J. (1990). The staphylococcal enterotoxins and their relatives. Science, 248:705-711. 23. Bohach, G. H., Fast, D. J., Nelson, R. D., and Schlievert, P. M. (1990). Staphylococcal and streptococcal pyrogenic toxins involved in toxic shock syndrome and related illnesses. CRC Crit. Rev. Microbiol., 17251-272. 24. Smith, B. G., and Johnson, H. H. (1975). The effect of staphylococcal enterotoxins on the primary in vitro immune response. J. Irnmmol., 115575-578. 25. Sugiyama. H., McKissic, E. M., Bergdoll, M. S., and Heller, B. (1964). Enhancement of bacterial endotoxin lethality by staphylococcal enterotoxin. J. Zrlfect. Dis., 114:lll-118. 26. Betley, M. J., Borst, D. W., and Regassa, L. B. (1992). Staphylococcal enterotoxins toxic shock syndrome toxin and streptococcal pyrogenic exotoxins: A comparative study of their molecular biology. Chem. Imnunol., 55: 1-35. 27. Ren, K., Bannan, J. D., Pancholi, V., Cheung, A. L., Robbins, J. C., Fischetti, V. A., and Zabriskie, J. B. (1994). Characterization and biological properties of a new staphylococcal exotoxin. Genbank Accession U1 1702. 28. Soltis, M. T., Mekalanos, J. J., and Betley, M. J. (1990). Identification of a bacteriophage
374
29. 30.
31. 32. 33.
34. 35. 36. 37. 38. 39. 40.
41. 42. 43. 44. 45.
46.
47.
48.
49. 50. 51.
Martin et al. containinga silent staphylococcalvariantenterotoxingene (sez4'). Infect.. Immuz., 58: 1614-1619. Tweten, R. K., and Iandolo, J. J. (1983). Transport and processing of staphylococcal enterotoxin B. J. Bacteriol, 153:297-303. Christianson, K., Tweten, R. K., and Iandolo, J. J. (1985). Transport and processingof staphylococcal enterotoxin A. Appl. Emiron. Microbiol., 50:696-697. Banwert, J. B. (1989). Basic FoodMicrobiology, 2nd ed. (J. B. Banwert, ed.). Von Nostrand Reinhold, New York, pp. 203-219. Bayles, K. W., and Iandolo,J. J. (1989). Genetic and molecular analysisof the gene encoding Stnplzylococcus enterotoxin D, J. Bncteriol., 171(9):4799. Couch, J. L., and Betley, M. S. (1989). Nucleotide sequence of the type C3 staphylococcal enterotoxin gene suggests that intergenic recombination causes antigenic variation, J. Bacterial., 171(8):4507. Betley, M. J., and Mekalanos, J. J. (1988). Nucleotide sequence of the type A staphylococcal enterotoxin gene. J. Bncteriol., 170(1):34. Couch, J. L.. Soltis, M. T.. and Betley. M. J. (1988). Cloning and nucleotide sequence of the type E staphylococcal enterotoxin gene. J. Bacteriol., 170(7):2954. Bohach, G. A.,andSchlievert, P. M. (1987). Nucleotidesequenceofthestaphylococcal Mol. Gen. Genet., 209:15. enterotoxin C, gene and relatedness to other pyrogenic toxins. Jones, C. L., and Khan, S. A. (1986). Nucleotide sequence of the enterotoxin B gene from Staphylococcr~smreus. J. Bncteriol., 166(1):39. Huang, I-Y., Hughes, J. L., Bergdoll. M. S., and Schantz, E. J. (1987). Complete amino acid sequence of staphylococcal enterotoxin A. J. Biol. Chenz., 262(15):7006. Schmidt, J. L., and Spero, L. (1983). The complete anlino acid sequence of staphylococcal enterotoxin C. J. Biol. Chent.. 258(10):6300. Huang, I.-Y., and Bergdoll. M.S. (1970). The primary structure of staphylococcal enterotoxin B. 111. The cyanogen bromide peptides of reduced and anlinoethylated enterotoxin B and the complete amino acid sequence. J. Biol. Chent., 245(14):3518. Bergdoll, M. S., and Schlievert, P. M. (1984). Toxic-shock syndrome toxin. Lancet. ii:691. Bergdoll, M. S. (1983). Enterotoxins. In Staphylococci crnd Staphylococcal I?$ections, Vol. 2. (C. S. F. Easman and C. Adlam, eds.), Academic Press. New York, p. 559. Alouf, J. E. (1980). Streptococcal toxins (streptolysin 0, streptolysin S, erythrogenic toxin). Phrrrnz. Ther., 11:661. Schlievert, P. M. (1980). Purification and characterization of staphylococcal pyrogenic exotoxin type B. Biocltenzistry, 19:6204. Schlievert, P. M., Schoettle, D. J., and Watson, D. W. (1979). Purification and physiochemical and biological characterization of a staphylococcal pyrogenic exotoxin. Itlfect. Imnzzm., 23(3):609. D. W. (1978). Barsunlian. E. L., Cunningham, C. M.,Schlievert, P. M..andWatson, Heterogeneity of group A streptococcal pyrogenic exotoxin type B. Injkt. Irrzmurt., 20(2): 5 12. Kim, Y. B., and Watson,D. W. (1970). A purified groupA streptococcal pyrogenic exotoxin. Physiochemical and biological properties including the enhancement of susceptibility to endotoxin lethal shock. J. Exp. Med., 131:611. Johnson, H. M., Russell, J. K., and Pontzer, C. H. (1991). Staphylococcal enterotoxin superantigens, Minireview. Microb. Superantigem. Proc. Soc. Exp. Biol. and Med. 198:765-771. Marrack, P.. andKappler, J. (1990). The staphylococcalenterotoxinsandtheir relatives. Scierzce, 238:705. Johnson, L. P., and Schlievert, P. M. (1984). Group A streptococcal phage T12 carries the structural gene for pyrogenic exotoxin type A. Mol. Gen. Genet., 19453. Spero. L., Morlock, B. A., and Metzger,J. F. (1978). On the cross-reactivity of staphylococcal enterotoxins A. B, and C. J. Inz~nunol.,120( 1):86.
Staphylococcus
375
52. Smith, B. G., and Johnson, H. M. (1975). The effect of staphylococcal enterotoxin on the primary in vitro immune response. J. Zmmuzol., 115(2):575. 53. Brunson, K. W., and Watson, D. W. (1974). Pyrogenic specificity of streptococcal exotoxins, staphylococcal enterotoxins, and gram-negative endotoxin. Zrfect. Imnzmz., 10(2):347. 54. Sugiyama, H., McKissic, E. M., Bergdoll. M. S., and Heller. B. (1964). Enhancement of bacterial endotoxin lethality by staphylococcal enterotoxin. J. Infect. Dis.. 114:111. 55. Spero, L., Griffin,B. Y., Middlebrook,J. L., and Metzger, J. F. (1976). Effect of single and double peptide bond scission by trypsin on the structure and activity of staphylococcal enterotoxin C. J. Biol. Clzem., 251(18):5580. 56. Spero, L., Warren, J. R., and Metzger, J. F. (1973). Effect of single and double peptide bond scission by trypsin on the structure and activity of staphylococcal enterotoxin B. J. Biol. Clzem., 248(21):7289. 57. Spero, L., and Morlock, B. A. (1979). Cross reactions between tryptic polypeptides of staphylococcal enterotoxin B and C. J. Zrwzzozol.. 122(3):1285. 58. Huang, I.-Y., Schantz, E. J., and Bergdoll, M. S. (1975). The amino acid sequence of the staphylococcal enterotoxins. Jprz. J. Med. Sci. Biol., 28:73. 59. Bloomster-Hautamaa, D. A., Kreiswirth, B. N., Kornblum, J. S., Novick, R. P., and Schlievert. P. M. (1986). The nucleotide and partial amino acid sequence of toxic-shock syndrome toxin-l. J. Biol. Chem., 261(33):15783. 60. Johnson, L. P.. L'Italien, J. J., andSchlievert, P. M. (1986). Streptococcal pyrogenic exotoxin type A (scarlet fever toxin) is related to Stcqdzylococcus allreus enterotoxin B. Mol. Gen. Genet., 203:345. 61. Weeks, C. R., and Ferrietti. J. J. (1986). Nucleotide sequence of the type A streptococcal exotoxin (erythrogenic toxin) gene from Streptococcus pyogenes bacteriophage T12. Zrfect. Z r n m u n . , 52(1): 144. 62. Bohach, G. A.. and Schlievert. P. M. (1989). Conservation of the biologically active portions of staphylococcal enterotoxins C , and C,. Irzfect. Z m m w . , 57(7):2249. (1989). Comparative structural analysis of staphylococcal 63. Singh. B. R.. and Betley. M. J. enterotoxins A and E. J. Biol. Clzern., 264(8):4404. 64. Bergdoll, M. S. (1979). Staphylococcal intoxications. In Foodboi-m Zrzfections ami hzto.Yicatiorzs (H. Riemann and H. L. Bryan. eds.), Academic Press, New York, pp. 443-494. 65. Schantz, E. J., Roessler, W. G., Wagman, J.. Spero, L., Dunnery, D. A., and Bergdoll, M. S. (1965). Purification of staphylococcal enterotoxin B. Biochenzistn, 4(6): 1011. 66. Borja, C. R., and Bergdoll. M. S. (1967). Purification and partial characterization of enterotoxin C produced by Staplzylococczrs azweus strain 137. Biochemistlv, 6(15): 1457. 67. Bergdoll, M.S. (1967). The staphylococcal enterotoxins. In Biocherttistry sf Some Foodborne Microbial Tosim (R. I. Mateles and G. N. Wogan, eds.), MIT Press, Cambridge, MA, pp. 125. 68. Tibana, A., Rayman, K., Akhtar, M.. andSzabo, R. (1987). Thermal stability of staphylococcal enterotoxins A, B, and C in a buffered system. J. Food Prot., 50(3):239. 69. Denny, C. B.. Humber, J. Y., and Bohrer. C. W. (1971). Effect of toxin concentration on the heat inactivationof staphylococcal enterotoxin Ain beef bouillon andin phosphate buffer. Appl. Microbiol., 21(6):1064. 70. Fung, D. Y. C.. Steinberg. D. H., Miller, R. D., Kurantnick, M. J., andMurphy, T. F. ( 1973). Thermal inactivation of staphylococcal enterotoxin B and C. Appl. Mici-obiol., 26:938. 71. Bennett, R. W., and Berry, Jr., M. R. (1987). Serological reactivity and in vivo toxicity of Stcrylzylococcrls azlreus enterotoxin A and D in selected canned foods. J. Food Sci.. 52:416. 72. Read, Jr., R.B., and Bradshaw, J. G. (1967). Gamma irradiation of staphylococcal enterotoxin B. Appl. Mici-obiol., 15:603. 73. Modi. N. K.,Rose, S. A., and Tranter. H.S. (1990). Theeffects of irradiation and temperature ontheimmunologicalactivity of staphylococcalenterotoxinA. Zrzt. J. Food Mici-ohiol., 11(1):85.
376 74. 75. 76. 77.
78.
79.
80. 81. 82. 83. 84. 85. 86.
87.
88. 89.
90. 91. 92.
93. 94. 95. 96.
Martin et al. Smith, J. L, Buchanan. R. L.. and Palumbo, S. A. (1983). Effect of food environment on staphylococcal enterotoxin synthesis: A review, J. Food Prot., 46(6):545. Genigeorgis, C. A. (1989). Present state of knowledge on staphylococcal intoxication. Zrzt. J. Food Microbiol., 9:327. Halpin-Dohnalek, M. I., and Marth, E. H. (1989). Strrplrylococczls aweus and production of extracellular compounds and behavior in foods-a review, J. Food Prot., 52(4):367. Bryan, F. L. (1976). Staphylococcus nureus.In Food Microbiology: Public Health and Spoilage Aspects (M. P. Defigueiredo and D. F. Splittstoesser, eds.), AV1 Publishing, Westport, CT, pp. 12-128. Nychas, G. J., Tranter, H. S., Brehm. R. D., and Board, R. G. (1991). Staphylococczts mrezts S-6: Factors affecting its growth, enterotoxin B production and exoprotein formation.J. Appl. Bacteriol., 70:344. Marth, E. H. (1973). Behavior of food microorganisms during freeze-preservation. In LO” Tenzpemture Preservation of Foods and Living Matter (0.R. Fennema, W. D. Powrie, and E. H. Marth. eds.), Marcel Dekker, New York, pp. 386-435. Morse, S. A., Mah, R. A.. and Dobrogosz, W. (1969). Regulation of staphylococcal enterotoxin B. J. Bacteriol., 98:4. McLean, R. A.,Lilly, H. D., and Alford,J. A. (1968). Effects of meat-curingsalts and temperature on production of staphylococcal enterotoxin B. J. Brrcteriol., 95(4):1207. Markus, Z. H., and Silverman, G. J. (1970). Factors affecting the secretion of staphylococcal enterotoxin A. J. Appl. Microbiol., 70(3):492. Markus, Z. H., and Silverman, G. J. (1969). Enterotoxin B synthesis by replicating and nonreplicating cells of Staphylococcus nureus. J. Bacteriol., 97(2):506. Taylor, D., and Holland, K. T. (1989). Amino acid requirements for the growth and production of some extracellular products of Stcrphylococczts aweus. J. Appl. Bacteriol., 66:319. Bjorklind,A..andArvidson, S. (1978).Influenceofaminoacids on thesynthesisofan s . Gen. Microbiol., 107:367. extracellular proteinase from Staphylococcus ~ ~ l r e l lJ. Jarvis, A. W., Lawrence,R. C., and Pritchard, G. G. (1975). Glucose repression of enterotoxin A, B, and C and other extracellular proteins in staphylococci in batch and continuous culture. J. Gen. Microbiol., 86175. Miller, R. D., and Fung, D. Y. C. (1973). Amino acid requirements for the production of enterotoxin B by Staphylococcus nureus S-6 in a chemically defined medium. Appl. Microbiol., 25:800. Wu, C.H.. and Bergdoll. M. S. (1971). Stimulation of enterotoxin B production TI. Synthetic IZ., medium for staphylococcal growth and enterotoxin B production. hfect. Z ~ Z ~ I Z ~ 3(6):784. Coleman, G., Aboshkiwa, M., and Al-Ani, B. (1989).The effect of glucose on the expression of extracellular protein genes by Staphylococcus nureus strain V8. FEMS Microbiol. Lett., 61:247. Coleman, G. (1983). The effect of glucose on the differential rates of extracellular protein s 46). Arch. Microbiol., 134:208. and a-toxin formation by Staphylococcus a ~ ~ r e u(Wood Morse, S. A., and Baldwin, J. N. (1971). Regulation of staphylococcal enterotoxin B: Effect of thiamine starvation. Appl. Microbiol., 23(2):242. Coleman, G. (1985). A comparison of the patterns of extracellular proteins produced by the high a-toxin secreting organism Staphjllococcsls aweus (Wood 46) during aerobic and anaerobic growth. J. G e x Microbiol., 131:405. Hallis, B. A., Thurston, C. F.. and Mason, J. R. (1991). Glucose control of staphylococcal enterotoxin A synthesis and location is mediated by c-AMP. FEMS Microbiol. Lett., 80:247. Jay, J. M. (1992). Staphylococcw gastroenteritis. In Modern Food Microbiology (J. M. Jay, ed.), Von Nostrand Reinhold, New York, pp. 455-478. Notermans, S.. and Van Offerdijk, R.L. M. (1985). Production of enterotoxinA by StaplzyloC O C C I ~ Saweus in food. h t . J. Food Microbiol., 2:145. Gourama, H.. Tsai. W. Y.J.. and Bullerman, L. B. (1991). Growth and production of entero-
Staphylococcus
97.
98. 99. 100.
101. 102.
103.
104. 105.
106.
107. 108. 109.
110.
111.
112. 113.
113. 115.
377
toxins A and D by Staphylococcus allreus in salad bar ingredients and clam chowder. J. Food Prot., 54( 11):844. Daoud, S. M., and Debevere,J. M. (1984). Growth and thermonuclease production Staplzyby lococcus aureus in vegetables. In Microbial Association and Interactions irz Food (I. Kiss, T. Deak, and K. Incze, eds.), D. Reidel Pub. Co., Dordrecht, pp. 165- 170. Betley, M.J., and Mekalanos,J. J. (1985). Staphylococcal enterotoxin Ais encoded by phage. Science, 229: 185- 187. Smeltzer, M. S., Hart, M. E., and Iandolo, J. J. (1994). Theeffect of lysogeny on the genomic organization of Staphylococcus aureus. Gem, 138:5 1-57. Coleman, D. C., Arbuthnott. H. P.. Pomeroy, H. M., and Birkbeck, T. H. (1986). Cloning and expression in Escherichia coli and Staphylococczls nureus of the beta-lysin determinant from Staphylococcus aureus. Evidence that bacteriophage conversion of beta-lysin activity is caused by insertional inactivationof the beta-lysin determinant.Microbial Pathog., 1:549564. Johns Jr., M. B., and Khan, S. A. (1988). Staphylococcal enterotoxin B gene is associated with a discrete genetic element. J. Bacteriol. 171:4799-4806. Lindsay, J. A., Ruzin A., Ross, H. F., Kurepina, N., and Novick, R. P., (1998). The gene for toxic shock toxin is carried by a family of mobile pathogenicity islands in Staphylococcus aureus. Mol. Microbiol., 29:527-543. Recsei, P.. Kreiswirth, B., O’Reilly, M., Schlievert, P., Gruss, A., and Novick, R. P. (1986). Stclphylococcus aureus by agr. Mol. Gen. Regulation of exoproteingeneexpressionin Genet., 20258-61. Ocasio, W., and Martin, S. E. (1990).19. New methodsfor staphylococcal enterotoxin detection. Devel. Ind. Microbiol., 3 1: 175. Pereira, J. L., Salzberg, S. P., andBergdoll, M. S. (1991).Production of staphylococcal enterotoxin D in foods by low-enterotoxin producing staphylococci. Int. J. Food Microbiol., 14(1):19. Ewald, S., and Christensen,S. (1987). Detectionof enterotoxin production byStaph~lococcus aureus from aviation catering meals by the ELISA and the micro-slide immunodiffusiontest. Int. J. Food Microbiol., 5:87. Johnson, H. M., Hall, H. E., and Simon, M. (1967). Enterotoxin B: Serological assays in cultures by passive hemagglutination. Appl. Microbiol., 15(4):815. Mores, S. A., and Mah,R. A. (1967). Microtiter hemagglutination-inhibition assayfor staphylococcal enterotoxin B. Appl. Microbiol., 15:58. Igarashi, H., Shingaki, M., Fujikawa,H.. Ushioda, S. H., and Terayama, T. (1985).Detection of Staphylococcus enterotoxins in food poisoning outbreaks by reversed passive latex agglutination. Zerztralbl. Bakteriol. Mikrobiol. Hyg. Abt. 1. Suppl., 14255. Thorpe, G. H. G., Kricka, L. J., Moseley, S. B., and Whitehead, T. P. (1985). Phenols as enhancers of the chemiluminescent horseradish-peroxidase-luminol-hydrogenperoxide reaction: Application in luminescence monitored enzyme immunoassays. Clin. Chenz., 3 l(8): 1335. test for the detection of Fujikawa, H., and Igarashi, H. (1988). Rapid latex agglutination staphylococcalenterotoxins A to E thatuseshighdensitylatex particles. Appl. Environ. Microbiol., 54(10):2345. Dickie, N., and Akhtar,S. M. (1982). Improved radioimmunoassay of staphylococcal enterotoxin A. J. Assoc. 08 Anal. Chem., 65(1):180. Miller, B. A., Reiser, R. F., and Bergdoll, M. S. (1978). Detection of staphylococcal enterotoxins A, B, C. D, and E in food by radioimmunoassay using staphylococcalcells containing protein A as immunosorbent. Appl. Emiron. Microbiol., 36(3):421. Orth, D. S. (1977). Statistical analysis and quality control in radioimmunoassays for staphylococcal enterotoxins A, B, and C. Aypl. Envil-on. Microbiol., 34:710. Saunders, G. C.. andBartlett, M. L. (1977). Double-antibody solid-phase enzyme immunoas-
378
116.
117.
118.
119. 120. 121. 122.
123. 124.
125.
126. 127.
128.
129. 130.
131.
132. 133. 134.
Martin et al. say for thedetection of staphylococcalenterotoxinA. Appl. Enviro~l.Microbiol., 34(5): 518. Freed, R.C..Evenson. M. L.. Reiser, R. F., and Bergdoll, M. S. (1982). Enzyme-linked immunosorbent assay for detection of staphylococcal enterotoxins in foods. Appl. Ent9irorz. Microbiol., 44:1349. Gomez-Lucia. E.. Goyache. J., Orden, J. A.. Blanco, J. L.. Ruizsanta-Quiteria, J. A., Dominguez, L., and Suirez, G. ( 1989). Production of enterotoxin A by supposedly non-enterotoxigenic Staphylococcus mreus strains. Appl. Erzviroiz. Microbiol., 55(6): 1447. 0. (1984). Comparative evaluation of different enzymeFey, H..Pfister,H.,andRiiegg, linked immunosorbent assay systems for the detection of staphylococcal enterotoxins A, B. C, and D. J. Clin. Microbiol., 19(1):34. Simpson, J. S. A., Campbell, A. K., Ryall, M. E. T., and Woodhead, J. S. (1979). A stable chemiluminescent-labelled antibody for immunological assays. Nature, 279:646. Cheng, P. J., Hemmil& I., and Lovgren, T. (1983). Development of solid-phase immunoassay using chemiluminescent IgG conjugates. J. Irnmunol. Methods., 48: 159. Patel. A., and Campbell, A. K. (1983). Homogeneous immunoassay based on chemiluminescence energy transfer. Clin. Chenz., 29(9): 1604. Schroeder, H. R., Hines, C. M., Osborn, D. D., Moore, R. P., Hurtle, R. L., Wogoman, F. F., Rogers, R. W., andVogelhut, P. 0. (1981). Immunochemi-lunlinometric assay for hepatitis B surface antigen. Clin. Chenl., 27(8):1378. Kricka, L. J.. and Whitehead. T. P. (1984). Luminescent immunoassays: new labels for an established technique. Diagnostic Med., (May):45. Barnard, G. J.. Kim. J. B., Brockelbank, J. L.. Collins, W. P.. Gaier. B., and Kohen. F. (1984). Measurement of choriogonadotropin by chemiluminescence immunoassay and immunochemi-luminometric assay; 1. Use of isoluminol derivatives. Clin. Chenl., 30(4):538. Gadow. A., Fricke, H., Strasburger. C. J., and Wood, W. G. (1983). Synthesis and evaluation of luminescent tracers and hapten-protein conjugates for use in luminescence immunoassays with immobilized antibodies and antigens. A critical study of macro solid phases for use in immunoassay systems, Part 11. J. Clin. Chenl. Biochem., 32337. Lohneis, M.. Jaschke, K. H., and Terplan, G. (1987). An immunoluminometric assay (ILMA) for the detection of staphylococcal enterotoxins. Int. J. Food Microbiol., 5:117-127. Arakawa, H.. Maeda, M.. Tsuji. A., and Kambegawa. A.(1981). Chemiluminescence enzyme immunoassay of dehydroepiandrosterone and its sulfate using peroxidase as labels. Steroids, 38(4):453. Thompson, N. E., Razdan, M., Kuntsmann, G., Aschenbach, J. M.. Evenson, M.L..and Bergdoll, M. S. (1986). Detection of staphylococcal enterotoxins by enzyme-linked itnmunosorbent assays and radioimmunoassays: comparison of monoclonal and polyclonal antibody systems. Appl. Emiron. Microbiol., 51(5):885. Ocasio, W.. and Martin, S. E. (1988). Enhanced chemiluminescent immunoassay for staphylococcal enterotoxin B. Abstr.. Annu. Meeting Am. Sco. Microbiol., p. 282. Martin, S. E. ( 1989). Detection of injured Stapl~ylococcusaureus from foods. In Irzjzcred I11de.rand Pathogenic Bacteria: Occurrence m d Detectioil in Foods, Water, and Feeds (B. Ray, ed.). CRC Press. Boca Raton, FL, pp. 133-136. Russell, A. D. (1984). Potential sites of damage in microorganisms exposed to chemical and physical agents. In The Revilvd of Injzrred Microbes (M. H. E. Andrews and A. D. Russell, eds.), Academic Press, Orlando, FL, p. 1. Iandolo, J. J.. and Ordal, Z. J. (1966). Repair of thermal injury of Staphylococcus azu-ezls. J. Bacteriol.. 91:134. Hurst, A. (1977). Bacterial injury: A review. Cm. J. Micl-obiol.. 23:934. Macky, B. M. (1983). Lethal and sublethal effect of refrigeration, freezing, and freeze-drying of microorganisms. In The Revival sf InjlrrclMicrobes (M. H. E. Andrews and A.D. Russell, eds.), Academic Press, Orlando. FL, p. 45.
Staphylococcus aureus
379
135. Frey, H. E.. and Pollard, E. C. (1968). The action of ganlmay-ray-irradiated medium in bacteria: relation to the electron transport system. Rndiat. Res., 35:59. 136. Hurst,A.,Hendry, G. S., Hughes,A.,andPaley, B. (1976).Enumeration of sublethally injured staphylococci is some foods. Cm. J. Bacteriol., 22677. 137. Martin, S. E., Flowers. R. S., and Ordal,Z. J. (1976). Catalase: its effect on microbial enumeration. Appl. Erniron. Microbiol., 32731. 138. Flowers, R. S., Martin, S. E., Brewer, D. G., and OrdalZ. J. (1977). Catalase and enumeration of stressed Staplzylococcus uureus cells. Appl. Erzvirorz. Microbiol., 33:1112. 139. Allwood. M., and Russell, A. D. (1970j. Mechanisms of thermal injury in nonsporulating bacteria. Ah,. Appl. Microbiol., 1289. 140. Witter, L. D., and Ordal. Z. J. (1977). Stress effects and food microbiology. In Antibiotics and Antibiosis in Agriculture (M. Woodbine, ed.), Buttenvorths, Reading, MA, pp. 102112. 141. Rosenthal, L. J., and Iandolo, J. J. (1970). Thermally induced intracellular alteration of ribosomal ribonucleic acid. J. Bacteriol., 103:833. 142. Witter, L. D. (1981). Thermal injury and recoveryof selected microorganisms. J. Daily Sci., 64: 174. 143. Bluhm, L., and Ordal, Z. J. (1969). Effect of sublethal heat on the metabolic activity of Stuplzylococcus cmreus. J. Bucteriol., 97: 140. 144. Sogin, S. J., and Ordal, Z. J. (1967). Regeneration of ribosomes and ribosomal ribonucleic acid during repair of thermal injury in Stuphylococcus aureus. J. Bucteriol., 94:1082. 145. Rosenthal, L. J., Martin, S. E., Pariza, M. W., and Iandolo, J. J. (1972). Ribosomal synthesis in thermally shocked cells of Staphylococcus aureus. J. Bacteriol.. 109:243. 146. Miller, L. L., and Ordal. Z. J. (1972).Thermal injury and recovery of Bacillus subtilis. Appl. Microbiol., 24:878. 147. Chababurtty, K., and Burma, D. P. (1973).The purification and properties of a ribonuclease from Sublrorrella fyphinzurizan. J. Biol. Chern., 243:671. 148. Flowers, R. S., and Martin, S. E. (1980). Ribosome assembly during recoveryof heat-injured Stuplzylococcus aureus cells, J. Bacteriol., 141:645. 149. Fridovich, I. (1982) Superoxide dismutase in biology and medicine. InPathology of Oxygen (P. Autor, ed.), Academic Press, New York, p. 1. 150. Morris, J. G. (1976). Oxygen and the obligate anaerobe. J. Appl. Bucteriol., 40229. 151. Koppenol. W. H., and Butler. J. (1977). Mechanism of reactions involving singlet oxygen and the superoxide anion. FEBS Lett., 83: 1. 152. Wardle, M.. and Renninger, G. (1975). Bactericidal effects of hydrogen peroxide on spacecraft isolates. Appl. Microbiol., 30:710. 153. Edsall, G.. and Ley, H. L. (1965). The prevention of infection. In Bacterial and Mycotic hfections of M m (R. J. Dubos and J. Hirsh. eds.), Lippincott, Philadelphia, p. 913. 153. Grump, W. (1979). Disinfectants and antiseptics. In Emyclopediu of Chemical Technology (Kirk, R. E. and Othmer, D. F., eds.), John Wiley & Sons, New York, p. 807. 155. Amin, V. M., and Olson, N. F. (1968). Selective increase in hydrogen peroxide resistance of a coagulase-positive Staplzyloc.occus. J. Bacteriol., 95: 1604. 156. Kleppe, K. (-1966).The effect of hydrogen peroxide on glucose oxidase from Aspergillus niger. Biochemistry, 5: 139. 157. Kong, S., and Davison, A. J. (1981). The relative effectiveness of OH, H202,O?,and reducing free radicals in causing damage to biomembranes: A study of radiation damage to erythrocyte ghosts using free radical scavengers. Bioclzim. Biophys. Acta, 640:313. 158. Massie, H.. Samis, H., and Baird, M. (1972). The kinetics of degradation of DNA and DNA by H202.Biochint. Biophys. Acta, 272539. 159. Rhaese, H., Freese, J. E., and Meizer, M. (1968). Chemical analysis of DNA alterations. 11. Alteration and liberation of bases of deoxynucleotides induced by hydrogen peroxide and hydroxylamine. Biochinl. Biophvs. Actu, 115:491.
380
Martin et al.
160. Yamamoto, N. (1969). Damage, repair and recombination. 11. Effect of hydrogen peroxide on the bacteriophage genome. Virology, 38:457. 161. Uchida, Y., Shigematsu, H., and Yamfugi, K. (1965). The mode of action of hydrogen peroxide on deoxyribonucleic acid. Enzymologia, 29:369. 162. Beauchamp, C., and Fridovich, I. (1970). A mechanism for the production of ethylene from methional. J. Biol. Cllent., 245:4641. 163. Brewer, D. G., Martin, S. E., and Ordal, Z. J. (1977). Beneficial effect of catalase or pyruvate in a most-probable-number technique for the detection of Staphylococcus aurem Appl. Environ. Microbiol., 34:797. 164. Andrews, G.P., and Martin, S. E. (1978). Modified Vogel and Johnson agar for Staplzylococcus aweus. J. Food Prot., 41:530. 165. Bucker, E. R. Martin, S. E., Andrews, G. P., and Ordal, Z. J. (1979). Effect of hydrogen peroxide and sodium chloride on enumeration of thermally stressed cells of Staphplococcus aureus. J. Food Prot., 42(12):961. 166. Martin, S. E., and Barrier, W. A. (1990). Influence of salt, pH and temperature on Staphylococcus nureus MF-31 catalase. Food Microbiol., 7:121. 167. Bucker, E. R., and Martin, S. E. (1981). Superoxide dismutase activity in thermally stressed Staphylococcus culre'ells.Appl. Environ. Mici-obiol., 41 (2):449. 168. Baird-Parker, A. C., and Davenport, E. (1965). The effect of recovery medium on the isolation of Staphylococcus nzweus after heat treatment and after the storage of frozen or dried cells. J. Appl. Bacteriol., 28:390. 169. Collins-Thompson. D. L., Hurst, A., and Aris, B. (1974). Comparison of selective media for the enumeration of sublethally heated food-poisoning strains of Stnplzylococcztsaureus. Can. J. Microbiol., 20:1072. 170. Rayman, M. K., Deyovod, J. J., Purvis, U., Kusch, D., Lanier, J., Gilbert, R. J., Till, D. G., and Jarvis, G. A. (1978). ICMSF methods studies. X. An international comparative study of four media for the enumneration of Staphylococcus aureus in foods. Can. J. Microbiol., 24: 274. 171. Stiles, M. E., and Clark, P. C. (1974). The reliability of selective media for the enumeration of unheated and heated staphylococci. Can. J. Microbiol., 20:1735. 172. Carlsson, J., Nyberg, G., and Wrethen, J. (1978). Hydrogen peroxide and superoxide radical formation in anaerobic broth media exposed to atmospheric oxygen. Appl. EnvirorI. Micro biol., 36:223. 173. Idziak, E. S., and Mossel, D. A. A. (1980). Enumeration of vital and thermally stressed Stnplzylococcus aureus in foods using Baird-Parker pig plasma agar (BPP). J. Appl. Bacteriol. 48:lOl. 174. Lachia, R. V. (1984). Egg yolk-free Baird-Parker medium for the accelerated enumeration of foodborne Staphylococcus aureus. Appl. Emiron. Microbiol., 482370. 175. Mentzer-Morgenstern, L., and Katzenelson, E. (1982). A simple medium for isolation of coagulase-positive staphylococci in a single medium. J. Food Prot., 45218. 176. Baer, E. F., Gilden, M., Wienke. C., and Mellitz, M. (1971). Comparative efficiency of two enrichment and four plating media for isolation of Staphylococcus mreus. J. Assoc. 08 A m l . Cltenl., 54:736. 177. Giolitti, G., and Cantoni, C. (1966). A medium for the isolation of staphylococci from foods. J. Appl. Bacteriol., 29:395. 175. Patterson, J. T. (1973). Comparison of plating and the most probable number techniques for the isolation of staphylococci from foods. J. Appl. Bacteriol.. 36:273. 179. Oblinger, J. L., and Koburger, J. A. (1984). The most probable number technique. In Conzpend i m of Methods for the Microbiological Examination of Foods (M. L. Speck, ed.), American Public Health Association, Washington, DC. pp. 99-1 11. 180. Lancette, G. A. (1986). Current resuscitation methods for recovery of stressed Staphylococcus nureus cells from food. J. Food Prot., 49:477.
Staphylococcus aureus
381
181. Van Dorne, H., Baird, R. M., Hendriksz, D. T.. Van Der JSreck, D. M., and Pauwels. H. P. (198 1). Liquid modification of Baird Barker’s medium for the selective enumeration of Staphylococcus aweus. Antonie van Leeuwenhoek; J. Microbiol, Serol., 47:267. 182. Chopin, A., Malcom, S., Jarvis, G., Asperger, H., Beckers, H. J., Berbona, A. M., Cominazzini, C., Carini, S., Lodi, R., Hahn, G., Heechen, W., Jans. J. A., Jervis, D. I., Lanier, J. M., O’Connor, F. O., Rea, M., Rossi, J., Seligmann, R., Tesone, S., Waes, G., Macquot, G., and Pivnick. H. (1985). CMSF methods studies. XV. Comparison of four media and methods for enumerating Stnplzylococcus nureus in powdered milk. J. Food Prot., 4821. 183. (1980). OfJicialMethods of Annlysis of the A.O.A.C., 13th ed., Association of Official Analytical Chemists, Washington, DC, Sec. 46. 184. Lancette, G. A., and Lanier, J. (1987). Most probable number method for isolation and enumeration of Staphylococcus nureus in foods: collaborative study. J. Assoc. Ofs Anal. Chent., 70:35. 185. Lancette, G. A., Peeler, J. T.. and Lanier, J. M. (1986). Evaluation of an improved MPN medium for recovery of stressed and nonstressed Staphylococcus aureus.J. Assoc. Ofs Anal. Clzem., 69:44. 186. Hawcroft, D., and Geary, C. (1996). The use of nonradioactively labeled probe system in an electrophoretic ribotyping method for the differentiation of strains of coagulase-negative staphylococci. Electrophoresis, 1755-57.
This Page Intentionally Left Blank
16 Vibrio cholerae Charles A. Kaysner and June H. Wetherington U.S. Food and Drug Aduzinistratiorz, Bothell, Washington
I. Introduction 384 11. ClassificationandCharacterization
385
Classification A. 385 B. Characterization 386 Serological C. classification 386 111. DistributionandEcology387 Distribution A. 387 Ecology 388 B. C. The viable but nonculturable JV.
state
389
Relationship to WaterandFood389
Water 389 A. B. Food 390 V. Pathogenicity39 1 A. Characteristics of disease 391 Mechanisms B. of pathogenicity 393 C. Adhesion and colonization 394 D. Genetic regulation 394 E. Virulence of non-01/0139 V. cholerae VI. BiologicalandPhysicalControls
of Growth395
Temperature A. 395 B. Moisture and pH 396 C. Antimicrobials, disinfectants, and preservatives 397 VII. Principles of Detection in Water and Food 397 General A. considerations 397 Isolation B. procedures 398 Identification C. 398 D. Genetic-based procedures VIII.
Control of V. cholerae References 401
399
399
395
Kaysner and Wetherington
384
1.
INTRODUCTION
Vibrio cholerae is a species that contains both harmless aquatic strains as well as strains responsible for the gastrointestinal disease cholera in the form of epidemics and global pandemics. A disease called cholera was mentioned in ancient writings nearly 2500 years ago (57). Cholera was the term given to any type of gastrointestinal disorder, particularly those with resultant diarrhea. The disease is also called Asiatic cholera, since the organism was subsequently determined to beendemic in that region and responsible for vast epidemics in both the Ganges basin of India and in Bangladesh. In recent times the population of the Bengal region has notbeen free of the disease. This intestinal infection is transmitted primarily by contaminated, untreated drinking water as well as by contaminated food. Cholera is rapidly spread among individuals after drinking water and foods have become contaminated due to inadequate sewage disposal and poor sanitation. The lack of potable water sources has been associated with cholera epidemics occurring primarily in regions of underprivileged, low socioeconomic populations, Crowded living conditions aid the rapid spread of the disease. In addition, pandemic waves of cholera have occurred along trade routes and in association with pilgrimage and human migration. This life-threatening but easily preventable and treatable disease still causes an estimated 150,000 cases each year on several continents (70,90,91), and in 1997, a nearly 20% mortality rate occurred in several countries in Africa. The Italian F. Pacini first described the causative agent of cholera in 1854 (67). He discovered that the intestinal contents of cholera victims contained large numbers of a curved bacterium that he called Vibrio cholera. His discovery, however, was overshadowed by that of Robert Koch, who studied the disease in Egypt during the 1880s. Koch demonstrated that cholera was caused by a comma-shaped organism, which he called Konlrmbnzillen. For several decades the name Vibrio comnzn was used. Pacini's work was finally recognized and Vibrio clzolercre, Pacini 1854 was designated as the type species of the genus. We now know that V. cholerae is a Gram-negative, facultative anaerobe that is a common organism throughout the temperate climates of the world. Koch first proposed that a toxin produced by V. comnm was responsible for the disease. But it was not until 1959 that a toxin was demonstrated by two groups working independently (17,23). Ten years later, the toxin was purified (29), which allowed for further investigations of its structure and mode of action. It was further demonstrated that two important characteristics were necessary for a strain of V. cholerae to cause cholera. One was the ability to produce cholera toxin. The second was the somatic antigenic type, which was designated 01. Biotypes of the 0 1 serogroup, Classical and El Tor, were found to be responsible for the epidemics and pandemic waves of the disease. The El Tor biotype currently is the more prevalent biotype and is endemic in several areas of the world. Classical strains have been isolated infrequently in Bangladesh since the advent of the seventh pandemic, seemingly being replaced by the El Tor biotype. Seven pandemics have been recorded since the early nineteenth century (Table 1). The first six originated on the Indian subcontinent and ended about 1925. They are believed to have been caused by the Classical biotype. The seventh pandemic originated in Indonesia in 1961 and has endured for nearly 40 years. This pandemic, which was the first to be attributed to the El Tor biotype, spread throughout Southeast Asia, into Asia and the Middle East, and reached the African continent in 1970. Subsequently, in early 1991 cholera appeared on the South American continent, which had been free of the disease for over a century. The initial cases were reported in Peru, from which theepidemic rapidly
Vibrio cholerae
385
Table 1 CholeraPandemics Dates
Pandemic I I1 I11 IV V VI VI1 VI11
1817-1823 1829-1851 1852- 1859 1863- 1879 1881-1896 1899-1923/5 1961 -presenth El 1992-presenth
Indian subcontinent3 Indian subcontinenta Indian subcontinenta Indian subcontinenta Indian subcontinenta Classical Indian subcontinent3 Classical Indonesia Indian subcontinent
Tor 0139 Bengal
From Ref. 67. From Refs. 40, 63.
spread to bordering countries and into Central America. Over one million cases of cholera in 20 South and Central American countries resulting in more than 10,000 deaths have occurred (84,91). Through 1998, cases have been reported on a continual basis in Peru, Central America (El Salvador, Guatemala, Nicaragua, Belize), China (Guangzhou Province), Sri Lanka, Africa (Uganda, Kenya, Mozambique, Somalia), and India (70,91). In the fall of 1992, another explosive epidemic originating in the Bengal region of India occurred that was determined to be caused by a previously unknown serogroup of V. cholerae. This new serogroup was designated as 0139 and the strain as Bengal (1,26). This epidemic was similar in all aspects to those caused by the 0 1 serogroup, afflicting thousands of individuals and causing many deaths. The Bengal epidemic spread through India and subsequently reached Bangladesh, Nepal, Burma, Pakistan, Thailand, China, Malaysia, and Saudi Arabia. Imported cases were also reported in the United Kingdom and the United States. The Bengal strain is reported to share many virulence features with that of theEl Tor biotype of 0 1 V. cholerae of the seventh pandemic. Due to the similarity of disease, 0139 Bengal is considered to be the second etiological agent of cholera. Since this epidemic rapidly spread to other countries within a 2-year period, it is considered by some to be the eighth cholera pandemic (40,63,82). Several recent reviews are recommended to the reader that cover in more depth the information summarized here and below ( l .40,63,87) and recent electronic-based information (3 1).
II. CLASSIFICATION AND CHARACTERIZATION A.
Classification
In 1965, Veron (3) first proposed the family Vibrionaceae to group organisms into genera that were primarily oxidase-positive and motile by means of a single polar flagellum. This grouping was to differentiate these organisms from the Enterobacteriaceae and the recognized human pathogens in that family. Veron’s grouping was not intended to imply any ancestral relationship. Oxidase activity is the important character in differentiation of the Vibrionaceae from the common enteric pathogens. The Vibrionaceae currently contains four genera: Vibrio, Aeromonns, Plesiomonns, and Photobcrcteriurn. Vibrio species are Gram-negative, nonsporeforming, straight or curved rods and are facultative anaerobes by both a respiratory and fermentative metabolism (4). All Vibrios utilize D-glucose as a
386
Kaysner and Wetherington
sole source of carbon and energy, and most utilize ammonium salts as a sole nitrogen source. Several of these species cause disease in humans as well as in marine vertebrates and invertebrates, such as fish, eels, and mollusks. Of the 34 recognized Vibrio species, V. clzolerae is probably the most notable. Within the past two decades, a closely related species, V. nzi~~zicz~s, was identified (16). This species has phenotypic characteristics similar to V. cholerne. One notable difference is the fermentation of sucrose. V. rninzicus does not ferment sucrose, and this biochemical test can be used for differentiation. Tolerance to salt is similar between the two species. Of notable importance is that some strains of V. mirrliczrs have been reported to produce an enterotoxin identical to cholera toxin (78). Human illnesses have been reported from the consumption of raw or undercooked shellfish in which the causative agent has been identified as V. minzicus (74). Although these infections are considered to be rare, clinicians should be aware of an organism similar to the disease spectrum of V. cholerne. Growth of all vibrios is stimulated by sodium ions in concentrations of 5 to 700 M’. Tolerance of various levels of NaCl in laboratory medium is used as a basis for species identification. Vibrios can be divided into two groups based on this requirement: those not requiring the addition of NaCl to laboratory media (V. cholerne, V. nzinzicrts) and the remaining halophilic species that require media supplemented with NaC1.
B. Characterization V. cholerae tolerates moderately alkaline conditions, with growth in medium at pH 9. They do not tolerate acidic conditions and rapidly decline in numbers at pH 5 or less. Most V. cholerm strains are also sensitive to the vibriostatic agent 0/129 (2,4-diamino6,7-diisopropylpteridine), although there are recent reports of epidemic strains resistant to this vibriocidal agent, including 0139 Bengal strains (40,63). The sensitivity patterns, however, can generally be used to aid in the differentiation of species. The biochemical properties of the clinically important species of Vibrios are published in a number of reference manuals (4,24,45,56). These properties can help to differentiate V. cholerne from another common environmental and food-associated species in the family Vibrionaceae, Aeronzonas hvdrophiln. Table 2 presents a list of the minimal features considered necessary for the identification of V. choZerae.
C.SerologicalClassification In addition to the production of cholera toxin (CT), the somatic serogroup determination is an important property to identify epidemic V. cholerne. Currently, over 150 0 serogroups have been identified antigenically, the majority of environmental strains encountered belong in serogroups other than 0 1 and 0139. The 0 1 serogroup, as mentioned above, had until recently been exclusively linked with epidemic and pandemic cholera. Until the emergence of the epidemic 0139 serogroup, isolates identified as V. cholerne that would not agglutinate in 0 1 antiserum were reported as non-01 V. cholerae. The recent literature now refers to these as V. clzolerae non-01/0139 strains. The routine determination of serogroups, other than 0 1 and 0139, in the laboratory is impractical, however. The 0 1 serogroup can be subdivided into the Ogawa, the Inaba, and the very rare Hikojima serotypes. The serotype distinction has been used for epidemiological purposes. However, recent reports indicate that serotype conversion can occur, so this distinction
Vibrio cholerae
387
Table 2 Minimal Number of Characteristics Needed to Identify V. cholerne Percent Reaction positive Gram-negative, asporogenous rod Production of oxidase Glucose, acid under a petrolatum seal Glucose, gas D-Mannitol, acid meso-Inositol. acid Hydrogen sulfide (black butt on TSIj L-Lysine decarboxylase L-Arginine dihydrolase L-Ornithine decarboxylase Growth in 1% tryptone broth1
100 100 100 0 99.8 0 0 100 0 98.9 99.1
No sodium chloride added.
may be of limited value, whereas being of the 0 1 serogroup is of importance. V. cholerne 0 1 can also be divided into two biotypes, Classical and El Tor, based on hemolytic ability. Classical strains are nonhemolytic, whereas El Tor strains produce a P-hemolysin detectable on sheep blood agar plates. The 0139 serogroup is also hemolytic, resembling El Tor strains. Several other laboratory tests can be used to differentiate the 0 1 biotypes besides hemolytic ability, including a Voges-Proskauer reaction, inhibition by polymyxin B, agglutination of chicken erythrocytes, and a phage typing system (4,24,45,56). Of these differentiation tests, determination of the P-hemolysin and tolerance or susceptibility to polymyxin B are the most practical.
111.
DISTRIBUTIONAND ECOLOGY
A.
Distribution
Environmental investigations have determined V. cholerne to be a common bacterium in the temperate environments of the world. The majority of the bacterial species encountered in this environment are species of Vibrio. V. cholerae is part of the normal, free-living microflora of aquatic environments. Coastal areas with brackish waters and estuarine regions are niches for many species, including strains of toxigenic 0 1 V. cholerae. Epidemic cholera strains are endemic in several regions including Australia and the U.S. Gulf Coast and are sporadically involved in illnesses in those regions. V. cholerne 0 1 strains are occasionally encountered in the environment of nonendemic areas, but they are normally nontoxigenic and considered to be nonpathogenic (40). The predominant strains of V. cholerne encountered in these environments are of the non-01/0139 serogroups and are generally nonpathogenic to humans, although human infections resulting from these strains are occasionally reported. During the warmer months of the year, environmental samples frequently contain high levels of these non-01 /0139 strains. In regions of endemic human infection, the epidemic strains also appear in a regular seasonal pattern. Environmental factors trigger the dormant pathogen to multiply, which results in outbreaks.
388
Kaysner and Wetherington
B. Ecology Ecological studies of V. cholerne (41) and also V. nzi~nicus(13) identified the physical factors that enhanced their incidence and distribution in the environment. Water salinity is an important physical parameter in the ecology of these organisms. V. cholerae is more prevalent and found in higher numbers in areas where water has a general salinity range of 2-25 parts per thousand (ppt). The lower salinity range of 2-5 ppt in areas with significant fresh water inclusion favors V. cholerne and its cousin, V. mimicrrs. These two species are less prevalent or rarely recovered in water of salinity 30 ppt or greater found in the open ocean. Many estuarine areas fluctuate in salinity due to the amount of rainfall and to fresh water inclusion from streams or rivers and tidal action, which will influence the incidence of these species. V. cholerne is seldom, if ever, isolated from water of a temperature below 10°C but can be frequently isolated from water when the temperature ranges between 15 and 35°C. Thus, a seasonal occurrence is typical in estuarine areas, with frequent isolations during the warmer summer months. Toxigenic V. cholerae was recovered from sea salt solution preparations for up to 70 days stored at 25°C (58), indicating the influence of temperature on the incidence of the organism. During colder or winter months, recoveries of this species can occasionally be made from the top layer of sediment, where they may beinsulated from the lower temperatures and remain dormant until the next summer season. The effect of environmental temperature changes also correlates with the frequency of reports of infections. In regions where the water temperature is warmer year-round, the seasonal incidence is not as dramatic. The environmental presence of pathogenic vibrios does not distinctly correlate with the presence of human sewage, although nutrients may be contributed by sewage influx that may enhance survival of the organism. In regions where epidemics have occurred and there has been aninflux of untreated human waste into the aquatic environment, endemic 01, and now 0139, V. choleme strains are more frequently encountered. Extreme climatic events have resulted inan increase inthe number of cases of cholera in many regions, along with other human diseases (10). These events resulted in increased rainfall and hurricanes causing flooding and also droughts. Flooding affects the availability of drinking water by contamination of sources and droughts effect hygiene by limiting sewage disposal. These weather conditions are related to a phenomenon termed El Nifio, resulting in warmer-than-normal ocean water temperatures, which has caused many regions to experience unusual weather this past decade. As an example, Tanzania reported only 1.424 cholera cases and 35 deaths in 1996. After heavy rains and flooding during 1997, over 40,000 cases and 2,200 deaths were recorded. Dramatic increases in cases have also occurred in South America since 1991 with assistance from El Nifio. A relationship exists between phytoplankton and zooplankton in the water and V. cholerne. This species is chitinoclastic, and its ability to digest chitin may play a role in its persistence in the sediment. V. cholerae are able to colonize copepods and crustacea that have chitinous exoskeletons. It is suggested that attachment to planktonic forms enhances survival during adverse environmental conditions. Similarly, V. cholerae was found to colonize other aquatic biota, such as water hyacinths in Bangladesh, probably aiding its persistence in fresh water environments (40). Due to the ability of V. cholerne to attach to material suspended in thewater, animals that reside in estuaries can be expected to pick up these vibrios during feeding. No particular animal reservoir has beenidentified; however, bivalve mollusks and finfish may contain
Vibrio cholerae
389
Table 3 Recorded Outbreaks of Cholera Year
Location
No. cases
Source
Ref.
Philippines Italy Portugal Portugal Louisiana Gilbert Islands Texas Singapore Thailand Thailand Maryland Colorado
330 278 2467 136 11 572 15 37 24 71 3 1
Raw shrimp Raw seafood Raw/Undercooked cockles Bottled water Boiled/Steamed shrimp Raw clams, sardines Cooked rice Seafood Uncooked pork Uncooked beef Frozen coconut milk Cooked blue crab
39 2 6 7 9 55 38 33 81 80 85 30
~~
1961 1973 1974 1974 1977 1978 1981 1982 1987 1988 1991 1998
the organism on the surface or in the intestinal contents. Thus, seafood harvested from coastal and in-shore areas in endemic regions may frequently contain toxigenic V. C h d erue, which may ultimately be consumed by humans. Seafood commodities implicated in cholera infections include several species of finfish, squid, crustacea (shrimp, crab, lobster), and mollusca (oysters, clams, cockles, mussels) (Table 3).
C. The Viable but Nonculturable State V. cholerue and several other Gram-negative bacteria have been reported to survive in a viable but nonculturable state (VBNC) (14). This dormant state is induced by extended exposure to saline water and nutrient deprivation. The bacterial cells retain metabolic function but are not culturable on routinely employed nonselective bacteriological media. During this phenomenon the cells exhibit the starvation response, become ovoid in shape and reduced in size, and are detectable by epifluorescent microscopy. Cells can be induced into the VBNC state in the laboratory by inoculating sea-salt formulations adjusted between 0.5 and 2.5% (w/v) with NaCl and then storing at 5°C. When environmental conditions such as temperature and presence of nutrients become favorable, cells transform into a normal size and become culturable. VBNC cells of V. cholerae were reported to induce fluid accumulation upon injection into ligated rabbit ileal loops, thus showing a retention of virulence characteristics. The VBNC state is a plausible explanation for the persistence of this pathogen in the environment during adverse conditions, such as the colder winter water temperatures in temperate regions, while remaining undetected by laboratory examination. This dormant state may explain, at least in part, why endemic cholera strains cannot be recovered from the environment for periods of time, only to reemerge and once again infect humans.
W. RELATIONSHIP TO WATERAND FOOD A.
Water
Water plays a critical role in the transmission of cholera. During the second pandemic. Snow determined the role of water in the spread of the disease in England in the mid-
390
Kaysner and Wetherington
1800s (67). A central supply of public drinking water in London, the “Broadstreet Pump,” became infamous. The drinking water had been contaminated with the epidemic strain by sewage discharge into the Thames River. Since Snow’s discovery, numerous investigations into cholera outbreaks have consistently identified water as the source of the organism causing rapid, epidemic spread of disease (44,59). In countries with recorded epidemics, the lack of potable water sources was the fundamental reason for spread of the disease. Household water supplies, i.e., cisterns used for collection and storage, often become contaminated by family members when washing their hands. Bottled, contaminated spring water was responsible for a cholera outbreak of over 100 cases in Portugal (7). The contaminated, untreated public water system in Trujillo, Peru, the drinking of unboiled water, and drinking from a household water storage container in whichhands had been introduced were all highly associated with the spread of cholera in that country (83). In addition, ice made from Contaminated water was also incriminated in transmission during this 1991 epidemic (73). Interestingly, cargo ships may have been responsible for transmitting the Latin American epidemic cholera strain to another country’s coastal waters (18) via contaminated bilge water (53). In underdeveloped countries, residents are urged to boil water prior to drinking and before use for food preparation. Water is most often collected from untreated sources, mainly due to the lack of public water systems. Because of the lack of public sewage collection and treatment systems, human waste is disposed of in a manner that contributes to the contamination of drinking water sources. Beverages that were prepared from contaminated water and sold by street vendors were associated with epidemic cholera cases in Ecuador (89). During outbreaks and epidemics, health officials struggle to assist in providing potable drinking water; this usually requires educating the public to boil any water collected and sometimes involves providing tablets that chemically treat drinking supplies. In more developed countries, cholera outbreaks have been virtually eliminated by the establishment of treated public water distribution systems.
B. Food Contaminated food also plays a major role in thespread of cholera (21,44,59,69).A variety of foods have been implicated in outbreaks of cholera, and a few of these reports are summarized in Table 3. Often it is difficult to determine whether the food became contaminated by direct environmental contact with the organism, by an infected food handler, or by the water used in food preparation. Water contaminated by sewage used to prepare food for workers on a U.S. off-shore oil platform (38) and likewise during the preparation of rice in Peru (73) is incriminated i n the spread of the disease. Similarly. vegetables irrigated with untreated sewage were associated with the transmission of cholera during the South American epidemic (83). In many countries, this led to monitoring of vegetables imported from South America to prevent the possible spread of cholera outside of Latin America. During 1992- 1994, numerous samples including fruits, vegetables, and frozen seafood were examined for the presence of V. cholerae by several countries. Foods investigated during epidemics include cooked rice and legumes, millet gruel, vegetables, and fruit-usually foods that are prepared using water or are normally washed prior to consumption. Generally, nonacidic foods are most frequently vectors of the disease because V. clzolerne does not tolerate acidic conditions. Other food products that have been linked to outbreaks in several countries are potatoes, gelatin, chopped eggs, frozen coconut milk, and contaminated meat preparations (i.e. pork, beef, and seafood). Implicated seafood is
Vibrio cholerae
391
usually associated with harvesting from estuarine areas that have been contaminated by sewage during epidemics or from regions where the epidemic strains of V. clzolerae have become endemic. Products implicated include both raw and processed seafood such as raw marinated, salted, or dried fish, shrimp, crab, and the filter-feeding mollusks oysters, clams, mussels, and cockles. Ceviche is raw marinated fish, which was suspected of cholera transmission in South America, but this was not substantiated. Fish used in its preparation were harvested from coastal waters found to be contaminated with the epidemic strain. In Ecuador, eating raw seafood and cooked crab was highly associated with the spread of the epidemic (89). Food items obtained from street vendors were highly incriminated in many Latin American countries (84) and also in other countries during earlier epidetnics (51). In the United States, the major domestic food vehicle for cholera in the past 25 years has been crabmeat (88). Cooked blue crabs harvested and processed in Louisiana were recently responsible for an illness in Colorado (30). Inspection of the processor found that cooked crab was allowed to contact surfaces that were also used for raw crab. In addition to other poor sanitation conditions, whole crab were not completely submerged when boiled. Investigations into the U.S. outbreaks have most often found that the implicated food had been processed (cooked in the case of seafood) but had become contaminated after processing. In a 2-year study of illnesses associated with raw oyster consumption in the state of Florida, the reported disease agents in decreasing order of frequency were V. ynml?ael71olE,ticrrs,non-01 V. cholerne, V. ~~ulrziJcus, V. hollisne, V. mirniczfs, V. jhvialis, and 0 1 V. clzolerne (46). Similar findings were also reported for infections reported in one year in four Gulf Coast states (49). In general, consumption of molluscan shellfish accounts for the majority of gastroenteritis cases from seafood in the United States (74). Because of the increase in travel and the transport of food products between countries, the potential for transmission of V. choleme has increased. An example is the outbreak of cholera from coconut milk imported into the United States (85). Food served on airlines has likewise been implicated in cholera infections (1 1,SS). Travelers returning from countries with current epidemics or areas of endemic cholera have become infected from food consumed prior to departure or served on the airplane. In a few instances, travelers transported Contaminated products in their baggage to another country. U.S. illnesses reported to have been caused by the non-01/0139 V. cholerae and also V. rzlirnicus have predominantly arisen from the consumption ofraw oysters (60,7 1,74). These reports have for the most part been gastroenteritis cases, however, septicemia and peritonitis cases have also resulted. They differ in disease manifestation from toxigenic V. choler-ne in that the gastrointestinal illnesses are not as severe and 0 1 V. clzolerne have not been demonstrated in extraintestinal infections.
V.
PATHOGENICITY
A.
Characteristics of Disease
Cholera, recently reviewed in detail (40), starts with an incubation period from several hours to 5 days after ingestion of water or food contaminated with toxigenic V. cholerae. Time of onset is dependent in part on the inoculum level ingested. The infective dose of V. choleme is believed to be about one million cells. In volunteers, approximately 10” CFU in buffered saline were necessary to induce diarrhea in volunteers (50), while only
392 Wetherington
and
Kaysner
lo6 CFU ingested in a sodium bicarbonate solution to neutralize stomach acid induced diarrhea. The ingestion of lo6CFU in food such as rice and fish resulted in a 100% attack rate in a separate volunteer feeding study. This suggests that the ingested food can buffer the gastric acid, enabling the organism to reach and colonize the intestinal tract. Once colonized in the small intestine, V. choleme multiplies rapidly and produces cholera toxin. One known predisposition to severe infections is hypochlorhydria, a low level of stomach acid. Typical cholera is characterized by the sudden onset of vomiting and painless diarrhea, with the characteristic “rice-water” stools caused by the presence of mucous, developing as the disease progresses. Suppressed renal function, thirst, leg and abdominal cramping, and collapse due to marked dehydration and the resultant electrolyte imbalance follows in the severe diarrhetic episodes. Vomiting is common and occurs a few hours after the onset of diarrhea. Profuse secretory diarrhea is the main symptom, with resulting life-threatening dehydration. The fluid loss is so dramatic that an infected person can die within hours. These severe cases, “cholera gravis,” are reported to afflict about 15% of those infected by the classical biotype and but only about 2% of those infected by the El Tor biotype during an epidemic. Mild cases and asymptomatic infections occur most often. Moderate cases requiring medical attention occur in about 15% of those infected with the classical strains and only in about 5% of those with El Tor infections. An infected person sheds millions of these organisms in their feces, thus the disease can spread rapidly where poor sanitation and poor hygienic practices occur. The spread of the disease by personto-person contact has not been substantiated. Immediate treatment by rapid infusion of intravenous fluids is usually successful for the advanced cases, for replacement of the vital fluidslost. Oral rehydration is normally used for the milder cases and has been described as one of the most important therapeutic interventions developed in the twentieth century. Antimicrobial therapy has also been effective in shortening the duration of infection and reducing the carrier state; tetracycline is the drug of choice. However, widespread use of antimicrobials is strongly discouraged due to development of resistant epidemic strains, which are appearing more frequently. Cholera initiates an immune response in humans, and evidence exists for substantial infection-derived immunity. Several vaccines have been developed, some of which show promise for prevention of future epidemics. An easily administered vaccine, delivered orally, is desired. Currently live, attenuated bacteria, killed whole-cell preparations, wholecell toxoids, and a Salmonella-based carrier of V. cholercre antigens are preparations being evaluated in clinical trials. Non-01 /0139V. cholet-ae strains also cause human diarrheal illness and other clinical manifestations (46,60,61). Gastroenteritis is the most common of the illnesses reported, however, these strains have also caused wound, ear, septic, and peritoneal infections. Raw oyster consumption is highly correlated with the gastrointestinal infections. The vast majority of non-01/0139 strains do not produce cholera toxin and are not associated with epidemics. Two foodborne outbreaks of gastroenteritis have been reported as caused by non-01/0139 V. cholerme (61), and there are numerous individual reports of illness implicating consumption of rawshellfish (46,74). There is a seasonal occurrence of these gastroenteritis cases, accompanying the increased presence of these strains in the environment during summer and fall months. Generally, non-01/0139 gastroenteritis is reported to be mild or moderate in severity, however, occasionally a severe cholera-like disease has been reported. Symptoms reported include diarrhea, abdominal cramps, fever, and less frequently nausea and vomiting. Severe dehydration has been reported in a few patients.
.
Vibrio cholerae
393
Duration of illness can be from one to 10 days, and hospitalization and rehydration have been required for severe cases. There is no distinguishing symptom that would differentiate these infections from other gastrointestinal infections. Non-0 1/O 139 strains have also been isolated from cases of septicemia and peritonitis (10,65,7 1,75). The source of these infections was not identified in many of these cases, although some patients reported a recent history of seafood consumption or an association with seawater. A 60% mortality rate is reported for these infections. Epidemiological evidence from some cases suggests that septicemia could be acquired by invasion through the intestinal tract, similar to infections of V. vulnificus, a highly invasive species commonly encountered in seafood. The 0 1 epidemic strains have not been demonstrated to be invasive, however. Several of these patients had underlying medical conditions, such as liver disorders, that probably made them more susceptible to an invasive infection by these rare strains of V. cholerue. Isolates from septicemic patients are reported to be heavily encapsulated, similar to patient strains of V. vuZniJcz~~.
B. Mechanisms of Pathogenicity 1. Cholera Toxin The major determinant of virulence is the ability of V. cholerae to produce cholera toxin (CT) or choleragen. Strains of 0 1 and 0139 producing CT are considered fully virulent and capable of causing epidemics. Cholera toxin is a heat-labile protein toxin composed of a single A and five B subunits with an approximate molecular weight of 85,000 daltons (40). The initial action of CT is binding of the B subunit to the receptor, ganglioside G,,, on the cell membrane of epithelial cells. This binding is enhanced by the enzyme neuraminidase produced by the organism. The A subunit, after proteolytic cleavage to two further subunits, stimulates adenylate cyclase, resulting in increased cellular levels of cyclic AMP. Increased CAMPconcentration leads to increased chloride ion secretion by intestinal crypt cells and decreased NaCl absorption by villus cells, resulting in electrolyte movement into the lumen. The resultant osmotic gradient causes a massive water flow to the lumen, which overwhelms the absorptive capacity of the intestine, resulting in diarrhea. The gene for CT, ct.x-, was sequenced some time ago, which has allowed for molecular based studies of its action. The heat-labile enterotoxin of Escherichia coli was subsequently found to have similar homology to CT and to be similar in mode of action. Additional mechanisms may also be involved in the secretory effects of CT. Prostaglandins and the enteric nervous system both respond to CT and may add to the secretion that occurs with this disease. 2. Other Toxins In addition to CT, V. cholerae 0 1 and 0139 strains produce other toxins (40). Zonal occludens toxin (ZOT) increases permeability of the small intestinal mucosa by decreasing tissue resistance of the intercellular tight junction (zonal occludens). It is hypothesized that ZOT causes diarrhea as a result of leakage of water and electrolytes into the lumen by hydrostatic pressure due to the increase in permeability. Sequences of the zot gene have subsequently been found in non-01 /0139 strains. Accessory cholera toxin (ACE) causes fluid accumulation in ligated rabbit ileal loops and is suspected of forming an ion channel after its insertion into the epithelial cell membrane. The genes encoding these two toxins, Zot and m e , are located immediately upstream of the ctx genes encoding CT in what has been termed a “virulence cassette.’’
394
Kaysner and Wetherington
The hemolysin that differentiates the El Tor from classical strains is cytolytic to c? variety of erythrocytes and mammalian cells in culture, is lethal to mice, and has been demonstrated to cause fluid accumulation in the ligated rabbit ileal loop assay. This hemolysin/cytolysin is also produced by non-0 1/O 139 strains and may play anaccessory role in diaqhea produced by those strains. Other reported toxins produced by 0 1 strains include a “new cholera toxin,’‘ a sodium channel inhibitor, and a Shiga-like toxin. Volunteer feeding studies have shown that ctx gene-deleted 0 1 strains can still elicit mild to moderate diarrhea in some individuals, thus these other toxins may be responsible for the diarrheal cases from which CT-negative strains were recovered. Additionally, some strains of non-01/0139 V. cholerne produce a heat-stable toxin, referred to as Nag-ST, similar to that produced by the enterotoxigenic E. coli. Although all of these toxins may contribute in part to diarrhea caused by strains that do not produce CT, their role in human pathogenesis needs further study.
C. AdhesionandColonization V. cholerne must first colonize the small intestine to begin the disease process. The toxin co-regulated pilus (TCP) is the most characterized colonization factor of this organism. These pili consist of long filaments that are laterally associated in bundles. Expression of the pilus is correlated with the expression of CT, hence the name. Another potential colonization factor, accessory colonization factor, has been described that may be involved in colonization via motility and/or chemotaxis, but its exact nature is still under study. The polar flagellum of V. cholerae, besides being involved in motility, may serve as an adhesin and an important virulence factor. In response to chemotaxins, motile V. choZerne are drawn to the mucosal surface of the intestine. Nonmotile but fully toxigenic strains show a markedly diminished virulence in animal models, indicating that motility is an important factor in establishing disease. V. cholerne has been reported to produce at least four hemagglutinins that may be involved in cellular adherence in the gut. Additional adherence factors that have been described and are potentially involved include several outer metnbrane proteins (OMP), the lipopolysaccharide (LPS) of the 0 1 strains, a polysaccharide capsule of non-01/01 39 strains, and the presence of other pili. Some strains of 0139 V. choZer-aeproduce a polysaccharide capsule, which is not found in strains of the 01 serogroup. V. cholerne also produces a siderophore, vibriobactin, seemingly not necessary for secretory disease, but which may be involved in establishing bacteremia by some 01/0139 strains. D. GeneticRegulation Several systems for the regulation of virulence genes in V. cholerne have been identified. The most extensively characterized regulon is ToxR, which controls expression of CT, the TCP colonization factor, the accessory colonization factor, two OMPs, and three lipoproteins. The control of seventeen distinct genes by toxR has been reported, and this gene is called the “master switch,” regulating the “virulence cassette.” This was demonstrated during a study of volunteers fed a V. cholerne strain with mutated toxR that did not illicit diarrheal symptoms, although it contained other virulence genes. Recently, a genetic pathogenicity island, PAI, was identified in epidemic strains (42). This chromosomal PA1 is present in epidemic and pandemic strains but not in nonpathogenic strains of the organism, including some of the 0 1 serogroup. It was proposed
Vibrio cholerae
395
to call this the vibrio pathogenicity island (VPI) to differentiate it from those of other bacterial species. The VPI may represent the initial genetic factor required for the emergence of epidemic and pandemic strains. VPI is 39.5 kilobases in size and contains the genes important in establishing disease, including those having a direct role in disease or an indirect role in the transfer and mobility of the VPI. It has been hypothesized that transducing phages may transfer a PA1 and that horizontal transfer may occur resulting in the emergence of new pathogenic strains. The VPI was demonstrated in two clinical non-0 l /O 139 strains associated with outbreaks, suggesting acquisition of the virulence package.
E. Virulence of Non-01/0139 V. cholerae Diverse variability of virulence exists among the non-Ol/O139 strains encountered from clinical specimens and foods. When fed to volunteers, tnost strains failed to produce illness. Inoculum levels of clinical non-01/0139 strains greater than lob were necessary to induce diarrhea in volunteers (60), seemingly a greater number than required for the toxigenic 0 1 strains. No toxins unique to these serogroups have been described, although some produced by 0 1 strains and other species have been reported, for example, the hemolysin/cytolysin of the El Tor biotype and the thermostable direct hemolysin, TDH, of K pnrahaemulyticus. A heat-stable enterotoxin, Nag-ST, is produced by some strains and a cholera-like toxin by others, which results in a less severe diarrheal disease than true cholera. In addition, the genes encoding CT, ZOT, and ACE have been detected in some strains. Other possible virulence factors reported are a Shiga-like toxin, various cellassociated hemagglutinins, and a polysaccharide capsule that could facilitate production of septicemia in susceptible hosts. The production of multiple virulence factors by non01/0139 V. choler-ne appears to be necessary to elicit a diarrheal illness, although host susceptibility may play an important role. However, case reports indicate that these bacteria have also caused gastroenteritis in healthy individuals as well as people with preexisting diseases. Obviously, the virulence mechanisms of the occasional non-01/0139 strains of V. cholerae associated with human infections remain unidentified and need further study.
VI.
BIOLOGICALANDPHYSICALCONTROLS OF GROWTH
A. Temperature V. ckolercre is a mesophilic organism that thrives in the temperature range of 15-40°C, with optimum growth at 37°C. The ability of this species to grow at 43°C has allowed for a temperature-selective approach to isolation (20) and is also used for identification (24.45). V. cholerae grow rapidly at temperatures in the range of 30-43°C. They do not grow at temperatures less than 10°C, consequently, storage of food at recommended refrigeration temperatures (29"C (38). Radu et al. (39) found that V. vuZniJicus failed to multiply in oysters kept at 13°C or below for 30 hours, whereas the numbers of the bacterium were significantly higher when oysters were held at 18°C or higher (see also Sec. V.C). A similar time-dependent decrease in number of recoverable V. vulnijicus cells in oysters during cold storage was found by Cook and Ruple (40). This indicates that endogenous V. vulrlificus can multiply in unchilled shellstock oysters. Similar increases in numbers of V. vulnijicus were found in a later study when the number of V. vulnijicus in summer-harvested oysters held without refrigeration was followed over a 14-hour postharvest period (37). It is therefore clear , that a reduction of the time oyster shellstock remains outside refrigeration can decrease consumer exposure to high numbers of V. vul~ificus,but shellstock must be cooled immediately after harvest to eliminate postharvest growth of V. vulnijicus (37).
Dalsgaard et al.
442
111.
TAXONOMY,ISOLATION,ANDIDENTIFICATION
A. Taxonomy Like other members of the genus Vibrio (family Vibrionaceae), V. 1-~zrZrz(ficus is a gramnegative rod, aerobic and facultatively anaerobic, motile by means of a polar sheathed flagellum, and is oxidase and catalase positive (41). The taxonomy of V. 1~1niJicuswasfirst investigated by Baumann et al. (42). V. vr~lnificzrswas described as a group of gram-negative, fermentative marine organisms, which were assigned to group C-2. Hollis et al. (43) designated the same organism as “lactose-positive Vibrio” or “L+Vibrio” since the ability to ferment lactose was one characteristic that could distinguish this species from Vibrio yarcdmenzolyticus and Vibrio dgirzolyticus. Today it is known that lactose fermentation is negative in up to 25% of the V. vulr.ziJicusisolates (44). Strains within group C-2 were found to be genetically related based on DNA/DNA hybridization and were assigned as a new species designation Berzeckecr vulrzijiccr (vrrlrzificn = wound in Latin) (45). Similar studies performed on the L+Vibrio concluded that this group was a species separate from V. parnhcrenzol?)ticusand V. crlgi~zolyticrrs(46). In 1979, the transfer of Berzeckecr vuln(fica (synonym = L+Vibrio) to the genus Vibrio was proposed and its name became V. 1~1rziJiczrs (47). The name V. 1~1rziJiczrs was given official taxonomic status in 1980 (48). V. wlnificus formerly was often misidentified as V. ycrmlzaelnol?lticus (43). The species V. ~?ulrziJicrrs comprises two biogroups, which in the original definition differed phenotypically, serologically, and in host range (49). V. ~~ulrziJicus biogroup 1 is ubiquitous in estuarine environments and is an opportunistic human pathogen (1,5,3 1). Biogroup 2 is typically recovered from diseased eels butis also reported to cause wound infections in humans after handling eels (5,7). Taxonomic traits for V. vrrlrzificus are shown in Tables 1 and 2. The division into biogroups has recently been questioned, and a division into serovars hasbeen suggested (29). This chapter discusses mainly the originally described biogroup, 1, which is the major foodborne human pathogen (49). B. IsolationUsingPreenrichmentBrothsand Selective Media The choice of including a preenrichment step in isolation of V. ~~ulrzijicrrs depends on four factors: (a) the expected concentration of V. wlnijiczrs in the samples, (b) how precise the results should be, (c) the conditions of the cells, and (d) the level and composition of background flora. Any preenrichment step should improve the ratio of target to background flora before a selective plating step. Preenrichment procedures often give improved recovery of V. ~~ul~zijicrrs compared to plating on selective media, although the choice of procedure should always be dependent on the sample type (24,50-52). The isolation of pathogenic Vibrio spp. is usually accomplished by culture methods that start with preenrichment in alkaline peptone water (APW) (1% peptone, pH 8.6, with 1% NaC1) to recover sublethally injured organisms, followed by plating onto thiosulfatecitrate-bile salts-sucrose (TCBS) agar (53). Early studies of the environmental distribution of V. vulnificzrs, as well as clinical investigations, used this protocol, which was developed for other Vibrio spp. and not optimized for the isolation of V. ~~ulrziJicus (18). Various enrichment broths have been tested for their capability in isolation of V. wlnijicus, includ-
Vibrio vulnificus
443
ing APW with various salt concentrations, marine broth, salt-polymyxin B broth, Horie's broth, Monsur's broth, and glucose-salt-teepol broth (24,50,54,55). Overnight preenrichment in APW with 1% NaCl at 35-37°Cgenerally gives the best recovery of V. vuZniJicus, and this procedure is recommended in the Bacteriological Analytical Munual of the U.S. Food and Drug Administration (FDA) (56). The use of APW in combination with cellobiose-polymyxin B-colistin (CPC) agar and modified CPC (mCPC) agar has been reported to be effective in recovering V. vulrzijicus from oyster and water samples (19,57-61). Sun and Oliver (61) found 82% (with correct morphology) of over 1000 colonies probed with a hemolysin gene probe to be V. vulnificus. Figure 1 show a flow diagram for the isolation of v. vulniJicus. Overnight preenrichment in APW withpolymyxin B (20 U/mL; APWP) gave higher recovery rates than preenrichment in regular APW in combination with mCPC agar when analyzing samples of coastal water and sediment in Denmark (22). APWP and mCPC agar was subsequently used with success for isolation of V. vulnijkus from fresh and frozen seafood (3 1,62). Other studies have also reported that different sample types require different isolation strategies for V. vulnijicus (24,51). Arias et al. (50) reported that 3-hour preenrichment in APW with 3% NaCl followed by streaking onto CPC agar was optimal for recovering V. vulrzijicus from seawater and shellfish samples from the western Mediterranean coast and that this culture technique gave more positive results than detection by direct PCR (50). The high salt concentration in the preenrichment may favor isolation of V. vulnijicus cells adapted to the high salinity of the Mediterranean (around 35%). Arguments for using both polymyxin B and colistin in a V. vuln@cus-selective agar have not been provided (58,63). Colistin and polymyxin B are both fatty acyl decapeptide antibiotics with bactericidal activity against most gram-negative bacteria and are known by the name "polymyxins" (64). The chemical composition of colistin and polymyxin B differs only in a single amino acid, and their mode of action and microbiological activity are identical (64). Hoi et al. (65) examined a collection of V. vulniJcus strains for their sensitivity to colistin and recommended a new medium termed cellobiose colistin (CC) agar. CC agar gave a better V. vulrziJicus recovery than TCBS, CPC, and mCPC agar in laboratory studies with pure cultures andwith Danish water and sediment samples. The recovery rate on CC agar was significantly better than on mCPC agar (65). The confirmation rate of presumptive isolates from CC agar was as high as previously reported for mCPC (approximately 95%) when taking into consideration the typical colony morphology of V. wlnijicus on this medium (flat, yellow colonies -2 mm in diameter) (31,65). However, further research is needed to determine whether CC agar enhances the recovery of V. wlrzijicus from oysters compared to current methods. TCBS agar gave a very low plating efficiency (1%) of both clinical and environmental V. vulnificus strains and cannot be recommended for the isolation of V. vulnificus (65). This is in agreement with other reports of low recovery of V. vulnificus on TCBS (66,67). V. vulrzijicus canbe isolated from blood or other clinical samples by culture on several media including blood agar and other nonselective agars. The use of TCBS was recommended when patients present with a compatible diarrheal illness and a history of eating raw seafood (68). However, the recent findings of low recovery rates of clinical V. vulnijicus strains on TCBS agar suggest that this medium should not be used for direct plating of clinical specimens (65).
I
444
I
Dalsgaard et al.
Vibrio vulnificus
22
m
4 4 4
zcnr"
Pft2m-e
ooooooooggoooooo oooooo 0 0 0 mm
*S 00
Q
POm000om0000m00000000om
2
22
ooomoooomQoooooom~oooo
2
000000000000m0000000om
t-zftm-'cn
" 4
3 4 4 3
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
ocno0000 o o 00 m cnftP I f t
I
ft*s
o g ~ o o o0 0 0
c n m o o o o o o o * o o ~ o 0 0 o o m P o o ~ cn cncn
445
446
I
I
l
+++
+++
+++
+++
l + S +
+++
I
I
I
I
I
I
I
I +
I +
I +
I +
I
I
+++
I
I
11
I
I
l
Dalsgaard et al.
I
I
I + + + I
+ll
I S + + +
l
I
I
c '3
Vibrio vulnificus
447
Overnight preenrichrnent in alkaline peptone water with polymyxin B for 18-24 h at 37OC
Subculture onto CC agar followed by incubation for 18-24 h at 40°C
Identification by colony hybridization with DNA probe
Fig. 1 Flow diagram for the isolation of V. vztlnijiczo.
C. IdentificationUsingSerologicalandMolecular Met hods FDA is currently using an enzyme immunoassay (EIA) in an ELISA format to identify presumptive V. vu/rziJicussubcultured from mCPC agar (56). The assay uses a V. vu/tz(ficus-specific monoclonal antibody (mAb) directed against an intracellular epitope of V. vubzijicus (56,58). No cross-reactions with other Vibrio species or non-Vibrio species have been described, and the ELISA format reduces assay time and facilitates handling of large numbers of test samples (58). The cell line producing the V. vulntficus-specific mAb is available at the American Type Culture Collection (ATCC HB 10393). V. vulnijiclrs can be identified one step beyond primary isolation with antiflagellar (anti-H) antibody (69). The agglutination reaction is fast and reliable and has been optimized to include the use of mAbs. However, the anti-core H antibodies are not available commercially, so the technique includes purification of flagellar protein and immunization of rabbits, which is time-consuming (69). Molecular techniques, particularly specific oligonucleotide probes, constitute a very sensitive andspecific tool for detecting V. vulnijicus. An alkaline phosphatase-labeled oligonucleotide probe directed toward the cytolysin gene of V. vulnijicrls was constructed by Wright et al. (70,71) (probe sequence: GAGCTGTCACGGCAGTTGGAACCA). This probe, termed VVAP, demonstrated 100% specificity and sensitivity for clinical and environmental isolates of V. vzh@cus, and numerous investigators have shown that cytolysin is produced by all V. tJulniJicusstrains, including both biogroups, and is species-specific (15,7%74). The sequence of the cytolysin gene has also been used for constructing primers for PCR identification (75,76). Two studies have argued that the use of the cytolysin
Dalsgaard et al.
448
gene as a target region in PCR amplification or as a target for an oligonucleotide probe is not suitable for the detection of V. vulniJcus (77,781. A nonessential gene, such as the cytolysin gene, could theoretically be lost or rearranged without affecting the viability of the bacteria. Instead, it has been suggested that primerdprobes directed against rRNA genes should be used since rRNA molecules are essential constituents of all living organisms and are present in growing cells in very high numbers (77,78).
IV. PATHOGENICITY A.
Infections with V. vulnificus (Biogroup 1)
V. vulniJcus causes both foodborne and wound infections throughout the world, and in the United States it carries the highest death rate of any foodborne disease agent (79). There are approximately 50 foodborne cases per year in the United States that require hospitalization. This bacterium is highly invasive, causing fulminant primary septicemia in persons at risk for infection, with mortality rates of approximately 60% (1). Infection resulting in primary septicemia is associated with consumption of raw shellfish containing the bacteria, especially raw oysters, with symptoms typically developing within 24 hours of ingestion. Death may occur within hours of hospital admission. Individuals who are immunocompromised or who have elevated serum iron levels, typically a result of liver disease (such as cirrhosis or viral hepatitis), are at highest risk for infection by this organism (2). Infections most frequently occur in males [82% of the cases reviewed by Oliver (l)], whose average age exceeds 50 years. The most common symptoms in the primary septicemia form of infection include fever (94%), chills (86%), nausea (60%), and hypotension (systolic pressure < 85% mm; 43%). These values are similar to those reported by Hlady and Klontz (2) in a recent study of 333 patients with Vibrio infections associated with raw oyster consumption in Florida. Hlady and Klontz (2) also found that 94% of patients were hospitalized for up to 43 days (mean of >8 days). An unusual symptom is the development (in 69% of patients) of secondary lesions, typically of the extremities, which often require surgical debridement and/or result in amputation (1). In addition to the primary septicemia that follows ingestion, V. vulnificus is known to infect wounds of otherwise healthy individuals, although the majority of patients with serious wound infections have an underlying disease (1$0). These occur most often as a result of contamination of preexisting wounds with seawater or after contact with fish or shellfish. Wound infection symptoms include localized pain, edema, erythema, with possible severe necrosis of the surrounding tissue requiring surgical debridement or amputation (l). Mortality rates following wound infection are approximately 25% (1,SO). In a review of 11 patients infected with V. ~~ulrzijicus during an unusually warm summer in 1994 in Denmark, Dalsgaard et al. ( 5 ) reported that 4 developed bacteremia, one of whom died, and 9 developed skin lesions. These infections and additional wound infections in 1995 were reported when water temperatures were above 20°C ( 5 ) (Fig. 2).
B. Biogroup 2 Strains While V. vulr~ijic~~s is a pathogen for humans, Tison et al. (49) reported that certain strains isolated from locations in Japan were pathogenic for eels. Recently, biogroup 2 strains have caused major disease problems in Danish eel culture (29,30). This subset of V. vul-
Vibrio vulnificus
449
22 n
0 20 & \
\
g 12 IO
8
8
May
-
June -v-
1993
-+
-
August
1994
m
1995
September
1996
Fig. 2 Mean water temperatures from 1993to 1996 at three popular recreational beaches inDenmark. The bar shows the time period when the human V. vulrz~ficusinfections occurred in1994 and 1995.
nijkus strains was termed biogroup 2, based on phenotypic differences from the human pathogens that comprise biogroup 1. Biogroup 2 strains have been shown to possess similar virulence factors as biogroup 1, including production of exoproteins, uptake of various iron sources via phenolate and hydroxamate siderophores, and both LPS and capsule expression (81). However, the lipopolysaccharides of biogroup 2 strains are homologous, unlike those of biogroup l, which are heterologous (see Sec. 1V.D). Other differences between these biogroups include the fact that biogroup 2 strains are virulent for eels in either the encapsulated or nonencapsulated form, and nonencapsulated biogroup 2 isolates are able to utilize transferrin-bound iron (81). Recently, Amaro and Biosca (82) reported a V. vu1nZiJicu.sstrain isolated from a human leg wound that wasdetermined phenotypically to be biogroup 2, suggesting that at least some strains within biogroup 2 may also be opportunistic pathogens for humans. Biogroup 2 strains have also been associated with two cases of wound infections in Denmark (A. Dalsgaard and L. H@i,unpublished). There have been no human cases of V. vul~ziJicz~s associated with the consumption of infected eels, and the risks of such an infection appear to be very low. C. Virulence Factors A variety of factors have been implicated as possible virulence determinants for V. vulrzificus, including an extracellular hemolysin/cytolysin, an elastolytic protease, the ability to acquire iron from transferrin, the presence of a polysaccharide capsule and an endotoxic lipopolysaccharide, and resistance to the bactericidal effects of sera (for a review of these putative virulence factors, see Refs. 12 and 13). In addition, V. vulnijicus strains demonstrate a variation between virulent and avirulent forms, with virulent forms being encapsulated, serum resistant, and possessing the ability to acquire iron from iron-saturated transferrin, while avirulent variants lack these characteristics (83).
1. Exoenzymes A large number of extracellular compounds are produced by this bacterium, including hemolysin, elastase, collagenase, DNase, lipase, phospholipase, mucinase, chondroitin sul-
450
Dalsgaard et al.
fatase, hyaluronidase, fibrinolysin, and albuminase (84). An elastolytic protease has been purified and shown to be toxic for mice regardless of the injection route. Recent studies have also implicated V. vulnijicus proteases in the production of bradykinin ( S ) , which is an inflammatory mediator known to increase vascular permeability, cause vasodilation, and induce both pain and contraction of smooth muscle. Such proteases may be important in the intravascular dissemination of this pathogen, allowing septicemia to develop. The hemolysin produced by V. ~~uZ~zijicus strains has been isolated, purified, and shown to be lethal to mice when administered intravenously at concentrations as low as 3 pg/kg. While the hemolysin has been shown to be produced in vivo, a lack of correlation between hemolysin production and virulence has been demonstrated (86).
2. iron Utilization Elevated serum iron levels appear to be a critical element in the pathogenesis of V. vulrzijicus infections, with successful infection apparently requiring an increase in transferrin saturation. Indeed, Wright et al. (87) directly correlated virulence with host iron availability (Fig. 3). V. vdrzijicus does not appear able to grow in normal human serum (Fig. 4), while iron injections into mice prior to the injection of bacterial cells significantly lowers the LD50(Table 3). V. ~ ~ u l ~ z ~ fsimultaneously icus produces both phenolate and hydroxamate siderophores, with the phenolate siderophore enabling virulent isolates to acquire iron from highly saturated transferrin (88,89). Similar results have been reported for other ironbinding proteins such as lactofen-in and ferritin (89). It appears that cleavage of these proteins is caused by an exoprotease, resulting in iron release to the siderophore(s). Confirming that iron acquisition is required for V. vulrzijicus virulence, Litwin et al. (88) showed that a mutagenized virulent strain that lost phenolate siderophore production exhibited reduced virulence.
0
24
48
72
Time of Vibrio injection (h) Fig. 3 Effectofelevated serum iron levels on V. 1~1rziJicusinfection after chloroform (CC14) treatment (produces transient liver damage). Inocula of V. vulrtijiczrs (lo3,lo5, loqcfu) were injected for inocula of 10' and lo5 at 0, 24, 48, and 72 h after the injection of CCl.,. Increased mortality cfu correlated directly with increased serum iron (SI) levels, which were monitored over the same time period. Inocula of lo9 cfu always produced 100% mortality. (Adapted from Ref. 87.)
Vibrio vulnificus
451
10%
0
107J
3E
106-
105-
104-
0 -"- HS
1034
-0-
HS+O.Ol mg/mL
102-
+"
HS+O.I mg/mL
O I '
0
I
1
2
4
-
1
6
1
8
Time (h) Fig. 4 Growth of V. vulrzijicus in human serum. Pooled human serum (HS) was inoculated with V. vz~lniJicuswith and without iron additions (ferric ammonium citrate). (Adapted from Ref. 87.)
3. Capsule The presence of a capsular polysaccharide (CPS) is the best studied virulence factor of V. vuZmj?cus and is essential to its ability to cause human infection. Kreger et al. (90) demonstrated that an "antiphagocytic surface antigen" allows virulent V. 1~uZr2iJicusstrains to resist phagocytosis by human polymorphonuclear leukocytes. This antigen was subsequently shown by electron microscopy and ruthenium red staining to be an acidic polysaccharide capsule.
Table 3 Effect of Iron on LD50Determinations" Number dead Log I O inoculum
PBS
Fe"
8 7
6 5 4 3 2
5 5 5 5 4 4 0
1
0 -1
LD,,
=
1
X
10'
V. vulrtificus was inoculated into groups of mice (n = 5 ) at levels varying from lo8 to 10" cfu, concurrently with either phosphatebuffered saline (PBS) or iron (ferric ammonium sulfate at 4 pg/ g body weight). Source: Adapted from Ref. 87.
452
Daisgaard et ai.
Studies of virulent and avirulent strains have found a correlation between virulence and colony opacity (Fig. 5). All virulent strains are of the encapsulated, or opaque, colony type, whereas nonencapsulated, or translucent, cells are avirulent (83), supporting the presence of capsule as an explanation for resistance to phagocytic activity. Interestingly, we have observed that encapsulated cells mutate at a very high rate (typically 10-2-10-3) to produce nonencapsulated cells, with the loss of capsule correlating with loss of virulence. Whether such mutations are common in natural communities or are a consequence of laboratory manipulation is not known at this time. Reversion of translucent cells to opaque cells has also been reported in some strains, but at a very low (20°C) and intermediate salinities (1 5-25%). Environmental and clinical strains both show a high degree of heterogeneity in phenotypic and genotyping tests, with few differences being found in their pathogenic potential using experimental animals. However, the low number of reported cases compared with the persons exposed to high numbers of V. v u l ~ ~ ~ f ithrough cus consumption of oysters suggest that all strains of V. ~~r~lrzificus are not equally pathogenic or that not all individuals in the defined risk groups are equally susceptible.
Vibrio vulnificus
463
Virulence is clearly dependent on the presence of an antiphagocytic capsule and most likely the endotoxin. Successful infection also generally requires certain underlying host syndromes, which predispose to this pathogen. Infection of such high-risk individuals carries an extremely high fatality rate, however. Numerous questions remain concerning this organism, including the role of the viable but nonculturable state in the epidemiology of the infection. The response of this organism to environmental stress may also prove to be a factor in disease production. Finally, the genetic heterogeneity of isolates of V. vulnificus should be noted. How can such variation exist for a single species? What is the role, if any, of the gene rearrangements that appear to be commonplace in this bacterium? And possibly most important, what tnechanisms of control can be applied to oysters such that the highly fatal primary septicemias produced by V. vuZniJcus can be reduced or eliminated?
REFERENCES 1. Oliver, J. D. (1989). Vibrio vulnijicus.In Foodborne Bacterial Pathogens(M. P. Doyle, ed.), Marcel Dekker, Inc., New York, pp. 569-600. 2 . Hlady, W. G., and Klontz, K. C. (1996). The epidemiology of Vibrio infections in Florida, 1981-1993. J. Infect. Dis., 173:1176-1183. 3. Baethge, B. A., and West, B. C. (1988). Vibrio vulniJicm: Did Hippocrates describe a fatal case? Rev. hfect. Dis., 10:614-615. 4. Bisharat. N., and Raz, R. (1997). Vibrio infection in Israel due to changes in fish marketing. Lancet, 348:1585-1586. 5. Dalsgaard, A., Frimodt-Mdler, N., Bruun, B.. H@i,L., and Larsen, J. L. (1996). Clinical manifestations and epidemiology of Vibrio vulnijcus infections in Denmark. Eur. J. Clin Microbiol. hfect. Dis., 15:227-231. 6. Veenstra, J., Rietra, P. J., Stoutenbeek, C. P., Coster, J. M., de Gier, H. H., and Dirks Go, S. (1992). Infection by an indole-negative variant of Vibrio vulnijicus transmitted by eels. J. Iizfect. Dis., 166:209-210. 7. Mertens, A., Nagler, J., Hansen, W., and Gepts-Friedenreich, E. (1979). Halophilic, lactosepositive Vibrio in a case of fatal septicemia. J. Clin. Microbiol., 9:233-235. Homldahl, T., and Tjernderg, I. (1995). First documented case of bacteremia 8. Melhus, with Vibrio vulnijiczu in Sweden. Scand. J. Infect. Dis., 27:81-82. 9. Hoyer, J., Engelmann, E., Liehr, R."., Distler, A., Hahn, H., and Shimada, T. (1995). Septic serogroup 0 4 wound infection acquired from the Baltic sea. shock due to Vibrio ~~ulnijicus Eur. J. Clin. Microbiol. hlfect. Dis., 14:1016-1018. from raw oysters. 10. Hlady, W. G., Mullen, R. C., and Hopkin, R. S. (1993). Vibrio vulnijc~~s Leading cause of reported deaths from foodbome illness in Florida. J. Fln. Med. Assoc., 80: 536-538. 11. Dalsgaard, A., Dalsgaard, I., Hgi, L., and Larsen, J. L. (1995). Forekomst og betydning af Vibrio vralnijicus i kystnaere omrider. Dansk Vet. Tidsskr., 78:496-501. 12. Oliver, J. D., and Kaper, J. (1997). Vibrio species. In Food Microbiology Fundanlentals and Frorztiers (M. P. Doyle, L. R. Beauchat, T. J. Montville, eds.). ASM Press, Washington, DC, pp. 228-364. FEMS Microbiol. 13. Linkous, D. A., and Oliver, J. D. (1 999). Pathogenesis of Vibrio vu~niJicus. Lett., 174:207-214. 14. Motes, M. L., DePaola, A., Cook, D. W., Veazey, J. E., Hunsucker, J. C., Garthright, W. E., Blodgett, R. J., and Chritel, S. J. (1998). Influence of water temperature and salinity on Vibrio vulnijicus in Northern Gulf and Atlantic Coast oysters (Crassostrea virginin). Appl. Ensiron. Microbiol., 64: 1459-1465.
A.,
464
Dalsgaard et al.
15. Kaysner, C. A., Abeyta, C. J., Wekell, M. M., DePaola, A., Stott. R. F., and Leitch, J. M. (1987). Virulentstrains of Vibrio lwlnijicusisolated from estuariesof the United States West Coast. Appl. Environ. Microbiol., 53: 1349-135 1. and characteriza16. Tamplin, M. L., Rodrick, G.E., Blake, N.J., and Cuba, T. (1982). Isolation tion of Vibrio vulrzificus from two Florida estuaries. Appl. Emiron. Microbiol., 44:14661470. 17. Kelly, M.T. (1982). Effectof temperature and salinity on Vibrio (Beneckea)vulnijicus occurrence in a Gulf coast environment. Appl. Environ. Microbiol., 44:820-824. 18. Oliver, J. D., Warner, R. A., and Cleland, D. R. (1982). Distribution and ecology of Vibrio vzhificus and other lactose-fermenting marine vibriosin coastal waters of the southeastern United States. Appl. Environ. Microbiol., 44: 1404-1414. 19. O'Neill, K. R., Jones, S . H. and Grimes, D. J. (1992). Seasonal incidenceof Vibrio vulnificus in the Great Bay estuary of New Hampshire and Maine. Appl. Environ. Microbiol.,58:32573262. 20. Oliver, J. D., Warner, R.A., and Cleland, D. R. (1983). Distributionof Vibrio vulnificus and other lactose fermenting vibrios in the marine environment. Appl. Environ. Microbiol.. 45: 985-998. 21. Veenstra, J., Rietra, P. J., Coster, J. M., Slaats, E., and Dirks-Go, S . (1994). Seasonal variations in the occurrence of Vibrio vulnificus along the Dutch coast. Epidenziol. Infect., 112: 285-290. 22. Dalsgaard, A., Dalsgaard, I., Hoi, L., and Larsen, J. L. (1996). Comparisonof a commercial biochemical kit and an oligonucleotide probe for identificationof environmental isolatesof Vibrio vulnificus. Lett. Appl. Microbiol., 22: 184- 188. 23. Arias, C. R., Pujalte, M. J., Garay, E., and Aznar, R. (1998). Genetic relatedness among environmental, clinical, and diseased-eel Vibrio vul~zijicusisolates from different geographic regions by ribotyping and randomly amplified polymorphic DNA PCR. Appl. Envirorl. Microbiol., 64:3403-3410. 24. Biosca, E. G., Marco-Noales, E., Amaro, C., and Alcaide, E. (1997). An enzyme-linked immunosorbent assay for rapid detection of V. vulniJicus biotype 2: Development and field studies. Appl. Environ. Microbiol., 63537-542. 25. Wright, A. C., Hill, R. T., Johnson, J. A., Roghman, M.-C., Colwell, R. R., and Morris, J. G. (1996). Distribution of Vibrio vulnijiczrs in Chesapeake Bay. Appl. Emiron. Microbiol., 621717-724. 26. DePaola, A., Capers, G. M., and Alexander, D. (1994). Densitiesof Vibrio vulnijicus in the intestines of fish from the U. S . Gulf Coast. Appl. Environ. Microbiol., 60:984-988. 27. Biosca, E. G., Amaro, C., Esteve, C.. Fouz, B., and Toranzo, A. E. (1991). First record of Vibrio vulnijicusbiotype 2 from diseased Europeaneels, Anguilla mlguilln. J. Fish Dis., 14: 103-109. 28. Muroga, K., Jo, Y., and Nishibuchi, M. (1976). Pathogenic Vibrio isolated from cultured eels. I. Characteristics and taxonomic status. Fish Pathol., 11:141-145. 29. Hoi, L., Dalsgaard, I., DePaola, A., Siebeling, R. J., and Dalsgaard, A. (1998). Heterogeneity among isolates of Vibrio vulnificus recovered from eels (Anguilla ctnguilla) in Denmark. Appl. Emiron. Microbiol., 64:4676-4682. 30. Dalsgaard, I., Hoi,L., Siebeling, R. J., and Dalsgaard, A. (1999). Indole-positive Vibrio vulnificus isolatedfromdiseaseoutbreaksinaDanisheelfarm. Dis.Aqunt.Org., 35: 187-194. 31. Hoi, L., Larsen, J. L., Dalsgaard, I., and Dalsgaard, A. (1998). Occurrenceof Vibrio IwlniJicus biotypes in Danish marine environments. Appl. Environ. Microbiol., 64:7-13. 32. Kaspar, C. W., and Tamplin, M. L. (1993). Effectsof temperature and salinity on the survival of Vibrio vulnificus in seawater and shellfish. Appl. Envirorz. Microbiol., 59:2425-2429. 33. Barcina, I., Lebaron, P., and Vives-Rego, J. (1997). Survival of allochthonous bacteria in aquatic systems: A biological approach. FEMS Microbiol. Ecol., 23:l-9.
Vibrio vulnificus
465
34. DePaola, A., Mcleroy, S . , and Mcmanus, G. (1997). Distributionof Vibrio vulnijicus phage in oyster tissue and other estuarine habitats. Appl. Environ. Microbiol., 63:2464-2467. 35. DePaola, A., Motes, M. L., Chan, A. M., and Suttle, C. A. (1998). Phages infecting Vibrio vulnijicus are abundantand diverse in oysters(Crassostrea virginica)collected from the Gulf of Mexico. Appl. Environ. Microbiol., 64:346-35 1. 36. Dalsgaard, A., Huss H. H., H-Kittikun, A., and Larsen, J. L. (1995). Prevalence of Vibrio cholera and Saln~onellain a major shrimp production area in Thailand. Int. J. Food Microbiol., 28:lOl-113. 37. Cook. D. W. (1997). Refrigerationof oyster shellstock: Conditionswhich minimize the outgrowth of Vibrio vulnijicus. J. Food Prot., 60:349-352. 38. Shapiro, R. L., Altekruse, S . , Hutwagner, L., Bishop, R., Hammond, R., Wilson, S . , Ray, B., Thompson, S . , Tauxe, R. V., Griffin, P. M., and Vibrio working group. (1998). The role of Gulf Coastoysters harvested in warmer months in Vibrio vulnjfjcus infections in the United States, 1988-1996. J. Infect. Dis., 178:752-759. 39. Radu, S . , Elhadi, N., Hassan, Z., Rusul, G., Lihan, S . , Fidara. N., Yuherman, and Purwati, E. (1998). Characterizationof Vibrio vulnijkus isolated from cockles(Andara granosa):Antimicrobial resistance, plasmid profiles and randomly amplification of polymorphic DNA analysis. FEMS Microbiol. Lett., 165:139-143. 40. Cook, D. W., and Ruple, A. D. (1992). Cold storage and mild heat treatment as processing aids to reduce the numbers of Vibrio vulnijicus in raw oysters. J. Food Prot., 55:985-989. 41. Baumann, P., and Schubert, R. H. W. (1984). Family 11. Vibrionaceae. In Bergey’s Manual of Systematic Bacteriology (N. R. Krieg and J. G. Holg, eds.), Williams & Wilkins Co., Baltimore, pp. 5 16-550. 42. Baumann, P., Baumann, L., and Reichelt, J. L. (1973). Taxonomy of marine bacteria: Beneckea parahaernolytica and Beneckea alginolytica. J. Bacteriol., 113:1144-1 155. 43. Hollis, D. G., Weaver, R. E., Baker, C. N., and Thornsberry, C. (1976). Halophilic Vibrio species isolated from blood cultures. J. Clin. Microbiol., 3:425-431. 44. Baumann, P., Baumann, L., and Mandel, M. (1971). Taxonomy of marine bacteria: The genus Beneckea. J. Bacteriol., 107:268-294. 45. Reichelt, J. L., Baumann, P., and Baumann, L. (1976). Studyof genetic relationship among marine species of the genera Beneckea and Photobacterium by means of in vitro DNAIDNA hybridization. Arch. Microbiol., 11O:lOl-120. 46. Clark, W. A., and Steigerwalt, A. G. (1977). Deoxyribonucleic acid reassociation experiments with a halophilic, lactose-fermentingVibrio isolated from blood cultures.Znt. J. Syst. Bacteriol., 27: 194- 199. with sepsis, 47. Farmer 111, J. J. (1979).Vibrio (“Beneckea”) vulnijicus,the bacterium associated septicaemia, and the sea. Lancet, 27:903. 48. Farmer 111, J. J. (1980). Revival of the name Vibrio vulniJcus. h . , J. Sys. Bacteriol., 30: 656. 49. Tison, D. L., Nishibuchi, M., Greenwood, J. D., and Seidler, R. J. (1982). Vibrio vulnijicus biogroup 2, a new biogroup pathogenic for eels. Appl. Environ. Microbiol., 44:640-646. 50. Arias, C. R., Aznar, R., Pujalte, M. J., and Garay, E. (1998). A comparisonof strategies for of the western Meditenathe detectionand recovery of Vibrio vulnijicusfrom marine samples nean coast. System. Appl. Microbiol., 2 1:128- 134. 51. Kaysner, C. A., Tamplin, M. L., Wekell, M. M., Stott, R. F., and Colburn, K. G. (1989), Survival of Vibrio vulnificusin shellstockand shucked oysters(Crassostrea gigas and Crassostrea virginica) and effects of isolation medium on recovery. Appl. Environ. Microbiol., 55~3072-3079. 52. DePaola, A., Motes, M. L., Cook, D. W., Veazey, J., Garthright, W. E. and Blodgett, R. (1997). Evaluationof an alkaline phosphatase-labeledDNA probe for enumerationof Vibrio vullrijicus in Gulf Coast oysters. J. Microbiol. Methods, 29:115-120. 53. Colwell, R. R., ed. (1984). Vibrios in the Environment, John Wiley & Sons, New York.
466
Dalsgaard et al.
54. Sloan, E. M., Hagen. C. J., Lancette, G. A., Peeler, J. T., and Sofos, J. N. (1992).Comparison of five selective enrichment broths and two selective agars for recovery of Vibrio vulnificus from oysters. J. Food Prot., 55:356-359. 55. Hagen, C. J., Sloan, E. M., Lancette, G. A., Peeler, J. T.,and Sofos, J. N. (1994). Enumeration of Vibrio paralzaemolyticrcs and Vibrio vztlnificrts in various seafoods with two enrichment broths. J. Food Prot., 57:403-409. 56. U.S. Food and Drug Administration. ( 1995). Bacteriological AnalyticalManual. Association of Official Analytical Chemists, Arlington, VA. 57. Kaysner, C. A., Abeyta. C. J., Stott, R. F., Krane, M. H., and Wekell, M. M. (1 990). Enumeration of Vibrio species, including V. cholera from samples of an oyster growing area, Grays Harbor, Washington. J. Food Prot., 53:300-302. 58. Tamplin, M. L., Martin, A. L., Ruple, A. D., Cook, D. W., and Kaspar, C. W. (1991). Enzyme immunoassay for identification of Vibrio lwlnijicusin seawater, sediment, and oysters. Appl. Environ. Microbiol.. 57: 1235-1240. 59. Bryant, R. G., Jarvis, J., and Janda, J. M. (1987). Use of sodium dodecyl sulfate-polymyxin B-sucrose medium for isolation of Vibrio vulnijicusfrom shellfish. Appl. Emfiron. Microbiol., 53:1556-1559. 60. Oliver, J. D., Guthrie. K., Preyer, J., Wright, A. C., Simpson, L. M., Siebeling, R. J., and Morris, J. G. (1992). Use of colistin-polymyxin B-cellobiose agar for isolation of Vibrio vulnijicus from the environment. Appl. Environ. Microbiol., 58:737-739. 61. Sun, Y., and Oliver J. D. (1995). Value of cellobiose-polymyxin B-colistin agar for isolation from oysters. J. Food Prot., 58:439-440. of Vibrio v~~lrr(ficus 62. Dalsgaard, A., and Hgi, L. (1997). Prevalence and characterization of Vibrio vulrzjficus isolated from shrimp products imported into Denmark. J. Food Prot., 60: 1132-1 135. 63. Massad, G., and Oliver, J. D. (1987). New selective and differential medium for Vibrio clzolercre and Vibrio vulnijcus. Appl. Enviorn. Microbiol., 532262-2264. 64. SGgaard, H. 1982. The pharmacodynamics of polymyxin antibiotics with special reference to drug resistance liability. J. Vet. Pharrnacol. Tlzer., 5219-23 1. 65. Hgi, L., Dalsgaard, I., and Dalsgaard, A. (1998). Improved isolation of Vibrio v Z t h i f i C r 4 S from seawater and sediment with cellobiose-colistin agar. Appl. Emiron. Microbiol., 64: 17211724. 66. Beazley, W. A., and Palmer, G. G. (1992). TCI-a new bile free medium for the isolation of Vibrio species. Aust. J. Med. Sci., 13:25-27. 67. Brayton, P. R., West, P. A.. Russek, E., and Colwell, R. R. (1983). New selective plating medium for isolation of Vibrio vulnijicus biogroup 1. J. Clin. Microbiol., 17:1039-1044. 68. Mclaughlin, J. C. (1995). Vibrio. In Marrual of Clinical Microbiology (P. R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, and R. H. Yolken. eds.), ASM Press, Washington, DC. pp. 465-476. 69. Simonson, J., and Siebeling, R. J. (1986). Rapid serological identification of Vibrio ~wlnijicus by anti-H coagglutination. Appl. Environ. Microbiol., 52: 1299-1304. 70. Wright. A. C., Morris, J. G., Maneval, D. R., Richardson, K., and Kaper, J. B. (1 985). Cloning of the cytotoxin-hemolysin gene of Vibrio vulnificus. Infect. Irmnun.. 50:922-924. 71. Wright, A. C., Miceli, G. A., Landry, W. L., Christy, J. B., Watkins. W. D., and Morris, J. G. (1993). Rapid identification of Vibrio vzrlnijicus on nonselective media with an alkaline phosphatase-labeled oligonucleotide probe. Appl. Environ. Microbiol., 59541-546. 72. Biosca, E. G., Oliver, J. D., and Amaro, C. (1996). Phenotypic characterization of Vibrio vulnijcus biotype 2, a lipopolysaccharide-based homogenous 0 serogroup within Vibrio vulnijiczrs. Appl. Envirorl. Microbiol., 62:918-927. 73. Parker, R. W., and Lewis, D. H. (1995). Sandwich enzyme-linked imnlunosorbent assay for Vibrio vulnijicus hemolysin to detect V. wlnificzcs in environmental specimens. Appl. Enlyir o ~Microbiol., . 61:476-480. 74. Morris. J. G., Wright, A. C.. Roberts, D. M., Wood, P. K., Simpson L. M., and Oliver, J.
Vibrio vulnificus
75.
76.
77.
78.
79.
80. 81.
82. 83. 84.
85.
86.
87.
88. 89. 90.
91.
92.
93.
467
D. (1987). Identification of environmental Vibrio ~wlnijcusisolates with a DNA probe for the cytotoxin-hemolysin gene. Appl. Enlliron. Microbiol. 53: 193-195. Brauns, L. A., Hudson, M. C., and Oliver, J. D. (1991). Use of the polymerase chain reaction in detection of culturable and nonculturable Vibrio vhiJicus cells. Appl. Erzviron. Microbiol.. 571265 1-2655. Coleman, S. S., Melanson, D. M., Biosca, E. G,, and Oliver J. D. (1996). Detection of Vibrio wlnijicus biotypes 1 and 2 in eels and oysters by PCR amplification. Appl. Envirorz. Microbiol., 62:1378-1382. Aznar, R., Ludwig, W., Amann, R. I., and Schleifer, K. H. (1994). Sequence determination of rRNA genes of pathogenic Vibrio species and whole-cell identificationof Vibrio wlaijicus with rRNA-targeted oligonucleotide probes. Int. J. Syst. Bncteriol., 44:330-337. Arias, C. R., Garay, E., and Aznar, R. (1995). Nested PCR method for rapid and sensitive detection of Vibrio vulrlifcus in fish, sediments, and water. Appl. Emiron. Microbiol., 61: 3476-3478. Todd, E. C. D. (1989). Preliminary estimates of costs of foodborne disease in the U.S. J. Food Prot., 52:595-601. Hlady, W. G. (1997). Vibrio infections associated with raw oyster consumption in Florida, 1981-1994. J. Food Prot., 60:353-357. Amaro. C., Biosca, E. G., Fouz, B., Toranzo, A. E., and Garay, E. (1994). Role of iron, capsule, and toxins in the pathogenicity of Vibrio vulrzificus biotype 2 for mice. Infect. 1171ft1llil,., 62:759-763. Amaro, C., and Biosca, E. G. (1996). Vibrio wdnificus biotype 2, pathogenic for eels, is also an opportunistic pathogen for humans. Appl. Envirorx. Microbiol., 62: 1454-1457. Simpson, L. M., White, V. K., Zane, S. F., and Oliver, J. D. (1987). Correlation between virulence and colony morphology in Vibrio vzhificus. Iflfect In~mrm,55:269-272. Oliver, J. D., Wear, J. E., Thomas, M. B., Warner, M., and Linder, K. (1986). Production of extracellular enzymes and cytotoxicity byVibrio vulrrijiczds.Dingrz. Microbiol. Infect. Dis., 5199-1 11. Maruo, K., Akaike, T., Ono, T.,and Maeda, H. ( 1998). Involvement of bradikinin generation in intravascular dissemination of Vibrio rdrzificus and prevention of invasion by bradikinin antagonist. Irfect. Irnmzm., 66:866-869. Morris, J. G., Wright. A. C., Simpson, L. M.. Wood, P. K., Johnson, D. E., and Oliver, J. D. (1987). Virulence of Vibrio vulnifcus: Association with utilization of transferrin-bound iron, and lackof correlation with levels of cytotoxin or protease production. FEMSMicrobiol. Lett., 40:55-59. Wright, A. C., Simpson, L. M., and Oliver, J. D. (1981). Role of iron in the pathogenesis infections. ITfect. I n ~ ~ r w n34:503-507. ., of Vibrio ~wl~ziJicus Litwin, C. M.,Rayback, T. W.,and Skinner, J. (1996). Role of catechol siderophore synthesis in Vibrio 17ulnijicusvirulence. Infect. Imnun., 642834-2838. Simpson, L. M., and Oliver, J. D. (1987). Ability of Vibrio vzrlnificus to obtain iron from transferrin and other iron-binding proteins. C u m Microbiol., 15: 155-157. Kreger, A., DeChatelet,L., and Shirley,P. (1981).Interaction of Vibrio vzrlrzijicus with human polymorphonuclear leukocytes: Association of virulence with resistance to phagocytosis. J. Infect. Dis., 144244-248. McPherson,V. L., Watts, J. A..Simpson. L. M., andOliver, J. D. (1991).Physiological effects of the lipopolysaccharide of Vibrio vu1nificu.s on mice and rats. Microbios., 67:141149. Espat, N. J., Auffenberg, T., Abouhamze, A., Baumhofer, J., Moldawer, L. L.. and Howard, R. J. (1996). A role for the tumor necrosis factor-alfa in the increased mortality associated with Vibrio ~.~ulrzificus infection in the presence of hepatic dysfunction.A m . Surg., 233:428433. Elmore, S. P., Watts, J. A., Simpson, L. M.. and Oliver, J.D. (1992). Reversal of hypotension
468
94. 95.
96.
97.
98. 99.
100.
101. 102.
103.
104.
105.
106.
107. 108. 109.
110.
111. 112.
Dalsgaard et al. induced by Vibrio vulnificus lipopolysaccharide in the rat by inhibition of nitric oxide synthase. Microbiol. Pathog., 13:391-397. Bahrani, K., and Oliver, J. D. (1990). Studies on the lipopolysaccharide of a virulent and an avirulent strain of Vibrio vulnijicus. Biochem. Cell. Biol., 68:547-55 1. Bahrani, K. F., and Oliver, J. D. (1991). Electrophoretic analysisof lipopolysaccharide isolated from opaque and translucent colony variantsof Vibrio vzrlnifcus using various extraction methods. Microbios., 66:83-93. Linkous, D. A., Simpson, L. M., and Oliver, J. D. (1997). Comparison of pathogenicity among Vibrio vulnijiczls strains based on capsular and LPS serotypes. 97th General Meeting of the American Society of Microbiology, Miami Beach, FL, abstract B-210. Hayat, U., Reddy, G.P., Bush, C.A., Johnson, J. A., Wright,A. C., and Morris, J. G. (1993). Capsular types of Vibrio vulnijicus: An analysis of strains from clinical and environmental sources. J. Infect. Dis., 168:758-762. Simonson, J. G., and Siebeling, R. J. (1993). Immunogenicity of Vibrio vzclnifcus capsular polysaccharides and polysaccharide-protein conjugates.Infect. Imntun., 612053-2058. Buchrieser, C., Gangar, V. V., Murphree.R. L., Tamplin, M. L., and Kaspar, C. W. (1995). Multiple Vibrio vzhijiczrsstrains in oystersas demonstrated by clamped homogenouselectric field gel electrophoresis. Appl. Environ. Microbiol., 61:1163-1168. Jackson, J. K., Murphree, R. L., and Tamplin, M. L. (1997). Evidence that mortality from Vibrio vukzijicus infection results from single strains among heterogeneous populations in shellfish. J. Clin. Microbiol., 35:2098-2101. Martin, S. J., and Siebeling, R. J. (1991). Identificationof Vibrio vulnijicus 0 serovars with antilipopolysaccharide monoclonal antibody.J. Clin. Microbiol., 29: 1684- 1688. Warner, J. M., and Oliver, J. D. (1999). Randomly amplified polymorphic DNA analysis of clinical and environmental isolatesof Vibrio vu111iJicusand other Vibrio species. Appl. EnviYon. Microbiol., 65:1141-1144. Hgi, L., Dalsgaard, A., Larsen, J. L., Warner, J. M., and Oliver, J. D. (1997). Comparison of ribotyping and randomly amplified polymorphic DNA PCR for characterizationof Vibrio vulnijicus. Appl. Environ. Microbiol.. 63: 1674-1678. Arias, C.R., Verdonck, L., Swings, J., Garay, E., and Aznar, R. (1997). Intraspecific differenand ribotyptiation of Vibrio vulnijicusbiotypes by amplified fragment length polymorphism ing. Appl. Environ. Microbiol., 63:2600-2606. Warner, J. M., and Oliver, J. D. (1998). Randomly amplified polymorphic DNA analysis of starved and viable but nonculturable Vibrio vulnijicus cells. Appl. Environ. Microbiol., 64: 3025-3028. Desenclos, J. A., Klontz, K. C . , Wolfe, L. E., and Hoecheri, S. (1991). The risk of Vibrio illness in the Floridaraw oyster eating population, 1981-1988.Ant. J. Epidentiol., 134290297. Hijarrubia, M. J., Lazaro,B., Sunen, E., and Fernandez-Astorga,A. (1998). Survivalof Vibrio vulnifcus under pH,salinity and temperature combinedstress. Food Microbiol, 13:193-199. Cook, D. W. (1994). Effect of time and temperature on multiplication of Vibrio vulnijicus in postharvest Gulf Coast shellstock oysters. Appl. Environ. Microbiol., 60:3483-3484. Parker, R. W., Maurer, E. M., Childers, A. B., and Lewis, D. H. (1994). Effect of frozen storage and vaccum-packaging on survival of Vibrio vulnijicus in Gulf Coast oysters (Crrtssostrea Iirginica). J. Food Prot., 57:604-606. Dixon, W. D. (1992). The effects of gamma radiation (6nCo) upon shellstock oysters in terms of shelf life and bacterial reduction, including Vibrio vulnjficus levels. M.S. thesis, Univ. Florida. Sun, Y . , and Oliver, J. D. (1994). Effects of GRAS compounds on natural Vibrio vulnijicus populations in oysters. J. Food Prot., 57:921-923. Sun, Y., and Oliver, J. D. (1995). Hot sauce: No elimination of Vibrio ~~ulnijicus in oysters. J. Food Prot., 58:441-442.
Vibrio vulnificus
469
113. Farmer 111, J. J., Hickman-Brenner, F. W., and Kelly, M. T. (1985). Vibrio. In Manual of Clinical Microbiology (E. H. Lenette, A. Balows, W. J. Hausler, and H. J. Shadomy, eds.), ASM Press, Washington, DC, pp. 282-301. 114. Hoge, C. W., Watsky, D., Peeler, R. N., Libonati, J. P., Israel, E., and Morris, J. G. (1989). Epidemiology and spectrum of Vibrio infections in a Chesapeake Bay community. J. Inject. Dis., 160:985-993. 115. Fang, F. C. (1992). Use of tetracycline for treatment of Vibrio vulnijicus infections [published erratum appears in Clin. Infect. Dis. 1993; 16(2):346]. Clin. Zrzfect. Dis., 15:1071-1072. 116. Morris, J. G., and Tenney, J. (1985). Antibiotic therapy for Vibrio vulnificusinfection. JAMA, 253~1121-1122. 117. Bowdre. J. H., Hull, J. H., and Cocchetto, D. M. (1983). Antibiotic efficacy against Vibrio vulnificus in the mouse: Superiority of tetracycline. J. Pharmtlcol. Exp. Therp., 225595598. 118. Chuang. Y. C., Yuan, C. Y., Liu. C. Y., Lan, C. K., and Huang, A. H. (1992). Vibrio awhzificus infection in Taiwan: Report of 28 cases and review of clinical manifestations and treatment. Clin. Infect. Dis., 15:271-276. 119. Nystrom, T., Olsen, R. M.. and Kjelleberg, S. (1992). Survival, stress resistance, and alternations in protein expression in the marine Vibrio sp. strain S14 during the starvation for different individual nutrients. Appl. Environ. Microbiol., 58:55-56. 120. Koga, T., and Kawata, T. (1986). Composition of major outer membrane proteins of Vibrio vulniJcus isolates: Effect of different growth media and iron deficiency. Microbiol. Zrnmmol., 30: 193-201. 121. Morton, D. S . , and Oliver, J. D. (1994). Induction of carbon starvation-induced proteins in Vibrio vulnificus. Appl. Environ. Microbiol.. 60:3653-3659. 122. Oliver, J. D. (1995). The viable but non-culturable state in the human pathogen Vibrio vulnificus. FEMS Microbiol. Lett., 133:203-208. 123. Oliver, J. D. (1993). Formation of viable but nonculturable cells. In Stunntion in Bacteria (S. Kjelleberg, ed.), Plenum Press, New York, pp. 239-272. 124. Oliver, J. D. (1999). Public health significance of viable but nonculturable bacteria. In Nonczrlturable Microorganisms in the Environment(R.R. Colwell. and D. J. Grimes, eds.), New York: Chapman and Hall Publ. 125. Whitesides, M. D., and Oliver, J. D. (1 997). Resucitation of Vibrio vulnificusfrom the viable but nonculturable state. Appl. Environ. Microbiol., 63:1002-1005. 126. Oliver, J. D. (1995). The viable but nonculturable state. In Proceedings of the 1994 Vibrio wlnificus workshop, U.S. Food and Drug Administration. Washington, DC, pp. 63-72. 127. Oliver, J. D., Hite. F., McDougald, D., Andon, N. L., and Simpson, L. M. (1996). Entry into, and resusitation from, the viable but nonculturable state by Vibrio vulnificus in an eustarine environment. Appl. Environ. Microbiol., 61:2624-2630. 128. Oliver, J. D., and Bockian, R. (1995). In vivo resuscitation, and virulence towards mice, of viable but nonculturable cells of Vibrio vulnificus. Appl. Environ. Microbiol.,61:2620-2623. 129. Lefkowitz, A., Fout, G. S . , Losonsky, G., Wasserman, S . S . , Israel, E., and Morris, J. G. (1992). A serosurvey of pathogens associated with shellfish: prevalence of antibodies to Vibrio species and Norwalk virus in the Chesapeake Bay region. Anz. J. Epidemiol.. 135:369380. 130. Firth, J. R., Diaper, J. P,, and Edwards, C. (1994). Survival and viability of Vibrio vulniJicus in seawater tnonitored by flow cytometry. Lett. Appl. Microbiol., 18:268-271. from 131. Nilsson. L., Oliver, J. D., and Kjelleberg, S . (1991). Resuscitation of Vibrio ~.~ulnificus the viable but nonculturable state. J. Bncteriol., 173:5054-5059. 132. Weichart, D., and Kjelleberg, S . (1996). Stress resistance and recovery potential of culturable and viable but nonculturable cells of Vibro vulnificus. Microbiology, 142:845-853. 133. Bloomfield, S . F., Stewart, S . A. B., Dodd, C. E. R., Booth, I. R..and Power, E. G. M. (1998). The viable but non-culturable phenomenon explained? Microbiology, 144:1-2.
470
Dalsgaard et al.
134. Dixon, B. (1998). Viable but nonculturable. ASM News,64:372-373. 135. Weichart, D., McDougald, D., Jacobs, D., and Kjelleberg, S. (1997). In situ analysis of nucleic acids in cold-induced nonculturable Vibrio vtrlniJicrts. Appl. Em7I'ron. Microbiol., 63: 2754-2758. 136. McGovern, V. P., and Oliver, J. D. (1995). Induction of cold-responsive proteins in Vibrio vulr$cus. J. Bncteriol., 177:4131-4333.
19 Yersinia Scott A. Minnich and Michael J. Smith Urliversiv of Idrrho, Moscou: Idaho
Steven D. Weagant U.S. Food m d Drug Administration, Bothell, Washington
Peter Feng U.S. Food a i d Drug Admirlistmtiorz, Wrrshington, D. C.
I. Introduction
472
11. Yersirzirr Species 473 A. Characteristics of the organisms B. Serotypes 473 Other C. classification systems
473 474
111. Disease 475 A. Clinicalmanifestations B. Foodborneoutbreaks
475 476
IV.ReservoirandDistribution477 V.Growth,Isolation,andIdentification A. Growth and survival B. Isolation methods Identification C. methods
479
479 480 482
VI. Pathogenicity 483 A. Endotoxin 483 B. Enterotoxin 484 C. Cellular invasion 485 Iron-regulated D. proteins 486 E. Plasmid-associated virulence 487 F. Pathogenicity testing 496 VII.PreventionandControl498 References
499
471
472
1.
Minnich et al.
INTRODUCTION
Based on DNA homology and biochemical profiles, there are 11 recognized species in the genus Yersinicr (1-4), but only 3 species, Y. pestis, K enterocolitica, and Y.yseudotuberculosis, are known to be pathogenic to humans. Of these, Y. pestis, best known for causing the great pandemics of black plague in medieval Europe, is considered to be a zoonotic organism and is transmitted by direct contact with infected animals or humans or by the bites of flea vectors from infected animals (5). Y. pestis today appears only sporadically as epizootic plague in a number of rodent species, but human infections still occur on occasion as a result of contact with diseased animals or flea vectors. The Y. pestis complete genomic DNA sequence has been completed and is accessible on the Internet (ftp://ftp.scrnger.rrc.u~/~ub/~~yy). The other 2 pathogenic species, Y. enterocoliticn and Y.pseudotuberculosis, are known as enteropathogenic Yersinia and primarily cause gastroenteritis-like illness, known as yersiniosis. Both species are recognized as foodborne pathogens, because infections are usually transmitted by the consumption of contaminated food and water. In the United States, there are estimated tobe only 3,000-20,000 cases of yersiniosis per year (6,7); hence, cotnpared to Snlrnonella or Ccrr7zpylobacter, Yersinin species are not frequent causes of foodborne illness. But still, 5 outbreaks of gastroenteritis caused by Y. enterocoliticn have been reported and were traced to the consumption of contaminated water, food, and a variety of dairy products (8,9). Elsewhere, sporadic foodborne infections by Y. enterocoliticcr continue to be a common problem in northern Europe and Scandinavia (10,l l), and incidences of food and waterborne illness caused by Y. enterocoliticn and Y. pseudotuberculosis are also fairly prevalent in Japan (12). Although unrelated to foods, Y. erzterocolitica have also been implicated worldwide in cases of bacterial sepsis and endotoxic shock resulting from transfusion of contaminated blood or blood products (13). Yersiuin species are ubiquitous and maybe isolated from many environmental sources. They are also isolated frequently from foods, especially from a wide variety of meats, but pork remains to be the only known reservoir for pathogenic Y. enterocoliticcr (14-16). Because of the organism's ability to proliferate in cold temperatures, microbiological procedures to isolate Y. enterocoliticn and Y. pselldotuberclrlosis from foods usually include a lengthy cold enrichment followed by alkali treatment before plating onto selective media (17). The identification of Yersinia species, especially the biotypes of Y. enterocoliticn, requires extensive biochemical tests followed by serology to determine the many serotypes that exist in the species. The presence of Yersirlier in foods is not always associated with disease, because most yersiniae are not pathogenic (10,123).It is essential, therefore, to test all Yersi~zia isolates for virulence factors. The pathogenic mechanism of Yersinin spp. has not been fully determined, but it includes the production of an enterotoxin, invasion of intestinal cells, and a host of plasmid-encoded virulence factors regulated by a complex system of intracellular and extracellular signals. In this chapter we will specifically review the two enteropathogenic Yersirzia species that are important in foodborne illness with respect to the organism, its characteristics, classification, disease, outbreaks, growth and isolation procedures as well as the extensive amount of information that has been published on its virulence factors and the regulation of these factors.
Yersinia
II.
YERSINIA SPECIES
A.
Characteristics of theOrganisms
473
Yersinia are gram-negative rod-shaped bacteria that are facultatively anaerobic. Like other genera of Enterobacteriaceae, Yersinin are oxidase negative, ferment glucose, and are motile by peritrichous flagella (19). Yersinia species are distinguished from other members of this family by their inability to ferment lactose, by their coccobacillary shapes, and by their loss of motility at temperatures above 30°C. Yersinia species are negative for lysine decarboxylase and arginine dihydrolase, and most, except for Y. pestis and Y. ruckeri, hydrolyze urea. Of the two enteropathogenic species, Y. erzterocoliticnand Y. pseudotuberculosis, that are important in foodborne illness, Y. pseudotuberculosisspecies do not decarboxylate ornithine, and both species produce little or no gas from the fermentation of carbohydrates. These two species also display a temperature-sensitive motility phenotype, becoming nonmotile at 37°C. Y. pestis is nonmotile. Y. enterocoliticn can be further subgrouped biochemically into seven biotypes designated as lA, lB, 2, 3, 4, 5, and 6. Currently, only strains in the biotypes 1B,2,3,4, and 5 are known to be pathogenic. These pathogenic biotypes and Y. erzterocolitica biotype 6 (now reclassified as Y. rnolleretti or Y. bercovierii and Y. kristerzsenii do not hydrolyze esculin rapidly (within 24 h) or ferment salicin (20). However, Y. enterocolitica biotype 6 and Y. kristerzsenii are relatively rare in the environment and can be distinguished from the pathogenic biotypes by their inability to ferment sucrose; they are also positive for pyrazinamidase (21). One note of caution, however, is that although most pathogenic Y. enterocolitica are positive for sucrose fermentation, some fully virulent, atypical phenotypic variants that are unable to ferment sucrose have been isolated from clinical samples (22). Characteristically, isolates of Y. pseudotubercrrlosis are negative for fermentation of ornithine, sorbitol, and sucrose. Otherwise, the species is fairly homogeneous biochemically, except for minor variations in the production of acid from melibiose, raffinose, and salicin (20). Also, unlike Y. erzterocolitica, isolates of Y. pseudotzrberculosis, especially those that are invasive for tissue culture cells, are esculin-positive,
B. Serotypes Strains of Y. erzterocoliticn and related species can be grouped serologically according to heat-stable somatic antigens. Wauters et al. (23) originally devised a scheme of 30 serogroups based on 34 antigenic factors. This was later expanded to add 20 more serogroups (24); however, many Y. erzterocoliticn-like organisms were later reclassified into separate species (35). as a result, Aleksic and Bockemuhl (1) proposed simplifying the serological scheme of Y. eilterocolitica into 8 serogroups based on 20 somatic factors. Not all Y. erzterocolitica are pathogens; however, pathogenic strains are widely distributed among serogroups and have been identified in serogroups 0:1,2a,3; 0:2a,3; 0:3; 0:8; 0:9; 0:4,32; 0:5,27;0:12,25; 0:13a,13b; 0:19; 0:20;and 0:21 (26). Those that predominate in human illness belong to serogroups 0:3, 0:8, 0:9, and 0:5,27. Clonal analysis of Y. erzterocolitica isolates by multilocus enzyme electrophoresis showed that the serotypes tend to cluster into two groups (27), and the pathogenic serotypes in these groups appeared to be distributed according to geographical niches (27,28). For example, the 0 : 9 serotype was found only in northern Europe and the 0:8 serotype was most frequently, and almost
474
Minnich et al.
exclusively, isolated in the United States (27). Serotypes 0:20, 0:21, 0:13a,b, 0:18,and 0:4,32, which tended to be more virulent than the other serotypes, were also considered to be “American” serotypes. Based on recent reports, however, these geographical boundaries for serotype distribution may no longer be valid. Y. elzterocoZitica serotype 0:9 has now been isolated in the United States (29), and isolations of serotype 0 : 8 have also been reported from Europe and Asia (30). In Japan, the first case of gastroenteritis caused by Y. erzterocolitica serotype 0 : 8 was reported in 1991, and raw pork was the suspect vehicle (31). Also, serotype 0:3, which is supposed to be commonly found worldwide, was not a frequent isolate in the United States. Now, however, there is a marked increase in incidences of 0 : 3 infections in the United States (29,32), and it is the predominant Yersinia serotype isolated from stools (3334). The current broader dissemination of Yersirzia serotypes worldwide maybe attributed in part to increases in international importation and exportation of meat products (31 3 ) . However, infringement into natural habitats may also be a factor, as wild rodents in Japan have been found to carry Y. erzterocolitica serotype 0 : 8 and are suspected to be a source of this pathogen in human infections (36). The Y. pseudotllberculosis species is subgrouped based on a heat-stable somatic antigen, and at present there are six serogroups, designated by Roman numerals I-VI. Serogroups I, 11,111, and IV also have subtypes, but they are not as easily determined, because the antiserum to the serogroup will often cross-react with the different subtype strains and vice versa. Each of the six serogroups are known to contain pathogenic strains (37).
C.OtherClassificationSystems Aside from the conventional biochemical and serological typing schemes, other techniques are also used for epidemiological typing and classification of Yers-sinin.Phage typing has been used to some extent (38), but the limited availability of serotype-specific phages has precluded reliable classification. Genetically, Yersinia is ahighly diverse group; therefore, methods comparing DNA homologies of Yersinin have been useful for epidemiological studies. Y. etlterocolitica DNA is only 3-1696 homologous to DNA from other genera in Enterobacteriaceae (39) and 40-6096 related to DNA from Y. pseudotuberculosis and Y. pestis (40). The latter two species are virtually identical, showing greater than 90% DNA homology overall (41). Within species, yersiniae are highly conserved genetically; isolates representing nine different serotypes of Y. erlterocolitica showed 80-100% DNA homology (39). Multilocus enzyme electrophoresis has also been used as a technique to subtype Yersinia strains. This technique looks at the electrophoretic mobility of a panel of up to 21 enzyme proteins. These patterns provide distinctive profiles that can group by species, by serotype, and by epidemiologically related groups within serotypes (42,43). Genetic techniques have been used extensively to compare the relatedness of plasmids in Yersinin species. Restriction endonuclease analyses of plasmid DNA (REAP) showed no relatedness between the plasmids of avirulent Y. r~ckeriand that of virulent Y. enterocoliticn (44). However, the virulence plasmids from Y. et1terocoZiticn,Y. pestis, and Y. pseuclotubercrrlosis shared several restriction fragments (45). Restriction analysis of plasmids may also be used to determine differences in serotypes, because virulence plasmids from Y. yseudotl,fberculosisand Y. enterocolitica showed serogroup-specific restriction fragmentation patterns (46,47). In a study of 123 strains of Y. enterocoliticcr,Fukushima et al. found
Yersinia
475
a close correlation between REAP patterns and geographic distribution (48). REAP was also used to subtype 678 strains of Y. pseudotuberculosis resulting in 16 distinct REAP patterns that were used in epidemiological investigations. Kaneko et al. (49) found REAP useful in subtyping Y. pseudotuberculosis strains within each serotype. Restriction analysis of chromosomal DNA (REAC) may also be used for epidemiological typing or identification of pathogenic Yersinia (50). Kapperud et al. compared the use of REAC to REAP and conventional phenotypic tests for differentiating strains of Y. erzterocoliticn and found it to be an effective supplement to other tests (51). However, the complexity and the large numbers of fragments obtained from the digestion of genomic DNA makes interpretation difficult. Analysis of genomic DNA may be facilitated by using probes to select small subsets of restriction fragments for comparison. This technique, known as restriction fragment length polymorphism (RFLP), has been used in typing Yersinia. Using RFLP and ribosomal RNA probes (ribotyping), several serogroups that were indistinguishable by other means can be effectively identified (52,53). Ribotyping may be used to trace reservoirs and the routes of infection in epidemiological studies of foodborne Yersirlia infections (52,53). DNA amplification using the polymerase chain reaction (PCR) has also been a useful method to differentiate serotypes of pathogenic Y. enterocoliticn. Using primers directed to the St gene that encodes for the enterotoxin, different size DNA fragments were amplified from pathogenic European serotypes as compared to American serotypes of Y. enterocoliticn (54). PCR has also been used to confirm the identity of suspect Yersinia cultures by amplifying genes that are limited to potentially virulent Yersinicr strains. (These PCR tests are explained in greater detail in Sec. V.B.)
111.
DISEASE
A. ClinicalManifestations The most prevalent symptoms of Yersinin infections are abdominal pain and fever (10). However, other gastrointestinal disorders such as diarrhea, nausea, headache, and vomiting may also be associated with the illness (10,5547). The minimal infective dose of Yersinia for humans has not been determined (14). The incubation period is about 24-36 hours (57), but periods of up to 11 days have been reported (14). The illness usually lasts 1-3 days (57); however, in some cases it may persist for 5-14 days or longer (14). Yersiniosis can manifest as diarrhea and abdominal pain, resembling gastroenteritis caused by other enteric pathogens (58). But when Yersinin invades the lymph system, the symptoms of fever and abdominal pain may closely mimic characteristics of acute appendicitis (58). Incidences of appendectomies being performed on yersiniosis patients have occurred in past outbreaks (10,55,58). Persistent Yersinia infections may lead to secondary complications such as erythema nodosum, septicemia (see Sec. VI.A), Reiter's syndrome, and reactive arthritis (lo,%). The latter illness is an autoimmune disease of the joints, presumed to be caused by host immune responses to Yersinia antigens (59). Presence of Yersinin antigens has been detected in the synovial fluid cells of affected joints (60). Patients prone to developing reactive arthritis from yersiniosis also have histocompatibility antigen HLA-B27 (59). Proteins that cross-react with HLA-B27 have been found on Y. yseudotubercrrlosis, suggesting that the arthritic condition may be induced by the molecular mimicry of bacterial antigens with HLA-B27 (59,61). Graves' disease or hyperthyroidism may be another complication associated with yersiniosis. Persons af-
476
Minnich et al.
flicted with this autoimmune thyroid disorder have high titers of antibodies to Y. enterocolitica (62). Presence of thyrotropin binding sites on Y. enterocoliticn have been identified (63), suggesting that Graves’ disease may be caused by the cross-reactivity of these sites with receptors of thyroid-stimulating hormones. 1. Mechanism of Infection The mechanism of infection by yersiniae has not been fully established. However, based on animal infection studies and the types of virulence factors encoded by the pathogenic Yersinin species, a hypothetical mode of infection may be formulated. The pathogen, typically ingested as contaminants in foods and dairy products, is transported to the small intestine, where it adheres to the epithelial cell linings, via perhaps an adhesin produced by the virulence plasmid (see Sec. V1.E). Invasive factors encoded by these organisms (see Sec. V1.C) enable the bacteria to penetrate epithelial cell barriers and disseminate into the lamina propia (64), where the various plasmid-encoded virulence factors allow yersiniae to resist the bactericidal effects of serum and the phagocytic activity of macrophages (see Sec. V1.E). The pathogen may then be transported to the lymph nodes, enter the blood stream, and spread to the rest of host (65). A few incidences of direct infection into the blood via transfusion of Yersinia-contaminated blood products have also been reported (see Sec. V1.A).
2. Patient Susceptibility Age and physical condition of the host can influence the severity of Yersinia infections (55). The population most susceptible are children and the elderly, but infants less than 1 year old are most seriously affected (14,55,57,58). In European countries, Y. enterocoliticn infections most frequently involve children 1-3 years of age, and likewise, Y. pseudotuber” culosis infections also tend to occur in young adults between the ages of 1 and 16 (56). Although the sex of the host has not been established as a factor, the number of males infected by Y. pseudotuberculosis tend to outnumber females (56). Iron is a required growth factor for Yersinia; hence, increases in free iron in the mammalian tissues can also increase the susceptibility of hosts to Yersinia infections (66,67). Patients in South Africa suffering from Bantu siderosis, caused by excess consumption of alcoholic beverages that contain a large amount of iron, have been shown to be very susceptible to Y. enterocoliticn infection (68). Most iron in mammalian tissue is complexed with iron-binding proteins resulting in low free iron levels. Hence, the absence of free iron provides an antimicrobial effect for the host (65). To overcome this barrier, pathogenic Yersinia produce iron-binding proteins to sequester iron from the host (see Sec. V1.D). B. FoodborneOutbreaks The foodborne nature of yersiniosis is well established, and numerous outbreaks have occurred worldwide. Two outbreaks in Quebec, Canada, in the mid 1970s affected 138 schoolchildren and were traced to the consumption of raw milk (69). Five outbreaks of Y. enterocoliticn serotype 0 : 3 in Japan involved over 1100 school children, and although never confirmed, foods were the suspected sources of infection (70). In the United States, an outbreak in New York State in 1976 affected 217 students, of which 38 were culture positive. Pasteurized chocolate milk was the implicated source, and Y. enterocoZitica serotype 0:8 was isolated from the milk (71). In 1980, an outbreak in Washington State, also
Yersinia
477
caused by serotype 0:8, affected 87 people. The source of contamination was traced to unchlorinated spring water used in the packaging of tofu, a soybean-based product (9). More outbreaks occurred in 1983 in southeastern United States, and again, pasteurized milk was implicated. A total of 172 cases were identified, and the causative organism was Y. enterocolitica serotype 0:13a, 13b (72). The most recent Yersinia outbreak occurred in 1989 in Atlanta, and the contamination, which affected 14 infants in the same household, was traced to the caretaker, who was preparing pork chitterlings (73). The unusual aspect about this outbreak was that it was caused by serotype 0:3, which is rare in the United States. Furthermore, two different phage types of serotype 0:3, as well as serotype 0: 1,2,3 were also implicated (74). Since then, only sporadic cases of Yersinicr infections have been reported from several New England states (75). It is interesting to note that although the primary reservoir of pathogenic Y. enterocoliticcr is swine, many outbreaks have implicated milk or milk products. Several microbiological surveys of milk showed that Yersinia are present in raw milk from cows and goats (71,76-79), but only in one instance was a potentially pathogenic strain isolated from pasteurized milk (71). Since Yersinin will not survive pasteurization, contamination of milk by Y. enterocolitica is probably due to deficient pasteurization or a breakdown in postprocess sanitation (76,78). Isolates of Y. enterocoliticn are also frequently found in streams, lakes, springs, and wells (80-84); hence, unchlorinated surface water has also been implicated as a vehicle in yersiniosis. In 1972, a 75-year-old man hunting in the woods in upstate New York was afflicted with Y. erzterocoliticn septicemia after drinking from a stream. Analysis of water taken from the mountain stream identified Y. enterocolitica of the same serotype and biotype as those isolated from the patient (85). The man recovered after 67 days of hospitalization and antibiotic therapy. Water was also implicated but not confirmed as the source of a yersiniosis outbreak at a ski resort in Montana (86). In the United States, Y. pseudotuberculosis is less commonly found than Y. enterocolitica, and although it is frequently associated with animals (including swine, birds, rodents, and hare), Y. pseudotuberculosis has not been implicated in foodborne illness and has only rarely been isolated from soil, water, and foods (87). In Japan on the other hand, several foodborne and waterborne outbreaks have been linked to Y. pseudotuberculosis (12,88). In one case, children were infected after drinking water from a garden pond that was contaminated with feces from a stray cat that carried Y. pseudotuberculosis (88). In Europe, a study done at a hospital in Ireland (89) found that 28% of patients presenting with symptoms of appendicitis symptoms and 11% ofpatients presenting with nonspecific abdominal pain had serum titer to Y. pseudotubercuzosis compared to 1% of controls.
IV.RESERVOIRANDDISTRIBUTION Swine are generally recognized as the primary reservoir for pathogenic Yersinia around the world. Y. enterocolitica have been isolated from tongues, throats, cecal contents, and feces of pigs; however, the isolation rate can vary greatly. Interestingly, Neyt et al. (90) recently determined that the pYV plasmids of Y. enterocoliticn low-virulence serotypes (0:1,2,3; 0: 1,2; 0:3; 0:9; and 0:5,27) contain a class I1 transposon, Tn2502, conferring arsenite and arsenate resistance. As these strains are of worldwide distribution, it suggests a clonal origin. Arsenicals were used for chemotherapy in Europe prior to World War I1 to treat swine for Serpulirla hyodysenteriae (formerly Treponema hyodvsenteriae)infections.
478
Minnich et al.
Thus, these authors speculate that such arsenical treatments, and consequent gain of resistance, may have contributed to the establishment, or maintenance, of Yersinia in pigs. One study of 50 slaughter pigs at an abattoir found that only 2 (4%) carried the pathogenic 0:5,27 serotype (91). Other reports, however, showed that as many as 25 and 58% of the pigs sampled in Belgium and Denmark, respectively, carried pathogenic serotypes of Y. enterocoliticn (92). A more recent survey of 3375 slaughter pigs showed that Y. enterocoliticn was present in 808 (23.9%) of pigs and of the 107 pathogenic isolates obtained, mostly serotype 0 : 5 and a few of serotype 0 : 3 were identified (93). Although it is evident that pathogenic Yersinia are prevalent in swine, not all studies seem to support the hypothesis that swine is the reservoir for human infections. A study from China conducted over a period of 11 years showed no correlation between the serotypes of animal isolates versus those that caused infections in humans; hence, it suggested that swine may not be the source of pathogenic Y. enterocolitica in humans (94). Contaminated pork or pork products account for a large portion of yersiniosis infections worldwide; however, this organism is fairly prevalent and can be isolated from a variety of other foods. A survey of various products in France showed Yersinin to be present in raw vegetables, milk, ice cream, cakes, and pork products (95). Of the 666 samples examined, an average of 33% were contaminated, but only one Y. enterocoliticn out of the 180 isolated was potentially pathogenic. A similar survey in Denmark found Yersinirr to be present in 40-80% of pork products, l-17% of dairy products, 43% of raw vegetables, 8-20% of soy products, 22% of seafoods, and 9% of salads. In that analysis, 71% of the isolates were Y. enterocolitica, ,but only one strain isolated from pork tongues was found to be potentially pathogenic (96). Aside from pork, other meat products such as poultry can also be contaminated with potentially pathogenic serotypes of Y. enterocoliticn (97). More recent surveys from other parts of the world are also consistent with above findings. Studies from New Zealand and Australia showed that Y. enterocolitica was present in 3.4% of the cooked processed meats and seafood (98), in 18% of beef, 10% of lamb, and 12% of pork samples examined, but none of the isolates were pathogenic (99). Similarly, a survey from Argentina of 450 samples of cold, ready-to-eat foods, such as ham, salami, and cheese, showed that only 1-2% were contaminated with Y. enterocolitica. Although all of these isolates were serotype 0:9, a documented serotype known to be pathogenic, none of them harbored virulence factors (100). These studies consistently show that Yersinia are prevalent in foods, however-and fortuitously-the presence of pathogenic serotypes in foods seems to be rare. Finally, Y. enterocolitica have been found in samples of raw milk, but not in pasteurized milk (99); hence, they are not very resistant to heating. But as the above surveys showed, Yersinin species are present in many cooked foods, and some past outbreaks have also implicated pasteurized milk, thus indicating that postprocessing contamination may bea common problem. Environmental analysis of dairy plants in Vermont showed that Y. erzterocoliticn was present in 10.5% of the plant sites surveyed ( 101). Y. enterocoliticcr are also ubiquitous in the environment and may be found in lakes, streams, soil, and vegetation. The organism has been isolated from the feces of dogs, cats, goats, cattle, chinchilla, mink, and primates (80,102).Wild animals such as deer frequently harbor Y. enterocolitica (85), and in Japan (103) and the United States (104) it has been isolated from rodents as well as from flies and other insects. In the estuarine environment, many birds such as waterfowl and sea gulls may be important carriers, as Yersinin has also been isolated from oysters, clams, and shrimp (84). Consistent with those findings, Y. enterocolitica has been found to survive well in stream water at temperatures of 6-
Yersinia
479
16"C, suggesting that surface water may serve as a persistent vehicle for this pathogen between animals and humans (105). The distribution of Y. enterocolitica is worldwide, but it tends to present greater problems in cooler climates. Yersiniosis is endemic to northern Europe where sporadic infections are common, especially in the fall and winter months. The seasonal distribution of Yersinia in the estuarine environment was also evident in Washington State, where peak levels occurred during winter months (unpublished data). Y. pseudotuberculosis is also found in tnany environmental sources and animals, and, like Y. enterocolitica, its primary reservoir also appears to be swine. Y. pseudotuberculosis is rarely found in the United States, but it seems to be fairly common in Japan. Pathogenic strains of Y. yse~~dotz~berculosis has been isolated from the throats of healthy swine and from retail pork samples surveyed in Japan (106), as well as from 20.6% of the freshwater samples analyzed (107). In Italy, a large survey of over 30,000 samples showed that Y. pseu~otuberculosiswas isolated from few clinical samples and animals examined but was not found in any of the environmental samples or from the dairy and raw meat products examined (108).
V.
GROWTH,ISOLATION,ANDIDENTIFICATION
A.
Growth and Survival
Yersinia can grow at temperatures of 0-44°C. In a rich medium, maximum growth is at 32°C with a doubling time of 34 minutes. Generation time rises to 40 minutes when temperature is increased to 40°C or decreased to 28"C, and at room temperature it is about 1 hour (109). Further decreases in temperature to 10°C increases doubling time to 5 hours (1 lo), and at 1°C generation time is close to 40 hours (1 11). Y. erzterocolitica shows a remarkable tolerance to low temperatures, and this attribute has been used effectively in isolation methods to selectively enrich for Yersirzin (see Sec. V.B). The optimal pH for growth of Yersinia is 7.6-7.9, but the organism will grow in pH ranges of 4.6-9.0 (109). The nature of the acidulent can also have an effect on minimum pH for growth. Karapinar and Gonul (1 12) demonstrated that acetic acid is more effective than citric acid in inhibiting growth. Y. enterocolitica is a versatile organism that can survive long periods in cool, dilute environments such as well water (1 13). Y. erzterocolitica can tolerate up to 5% sodium chloride in its growth medium (114). Growth rates of Y. enterocoliticn in pork decreased as the levels of NaCl, KC1, and CaCl? increased. CaCll was found to be more inhibitory than equivalent levels of NaCl or KC1. Concentrations of 2.2% CaClz (w/w)were enough to prevent outgrowth of Y. ejzterocoliticn in ground pork (1 15). However, Yersirzia are not very heat resistant. High-temperature, short-time pasteurization conditions of 71.8"C for 18 seconds easily kills Y. enterocolitica ( l 16,117). When some bacteria are exposed to elevated sublethal temperatures, they respond by synthesizing a group of heat-shock or stress proteins. Shenoy and Murano (1 18) found that Y. enterocolitica, when exposed to 45°C for 60 minutes, produced heat-shock proteins and were more resistant to subsequent lethal heat treatment. Evaluation of survival of control cells and heat-shocked cells in heated ground pork revealed that D values at 60°C increased from 1.7 to 6.7 minutes, respectively. They also found that heat-shocked cells grew as well as control cells when held in ground pork at 4 or 25°C (119). Gamma-irradiation has also been investigated as a means of eliminating Y. enterocolitica in pork products. The strains investigated were
480
Minnich et al.
found to be sensitive to gamma-irradiation, with a D value of 0.25 kGy at 0°C. When samples of cooked ham, salami, and raw pork were artificially contaminated with Y. enterocoliticn at 106 cfu/g, cooked ham and salami were decontaminated with doses of 3 and 4 kGy, respectively. Viable cells could not be eliminated at a dose of 6 kGy. The dose of 1 kGy at -40°C was efficient in eradicating low numbers of naturally occurring Y. enterocolitica (120). Yersinia can withstand freezing and survive for extended periods in frozen food, even after repeated freezing and thawing (1 16). Destruction of viable cells under freezingthawing and constant freezing conditions at -20°C was more rapid in distilled water than in milk. Presumably proteins and fats in the food matrix provide a protective effect for Yersirzia (1 16). Several studies have examined the survival and growth of Yersinin in foods. On cooked beef or pork, Y. enterocolitica can increase by 6 log within 10 days at 7°C. At 25"C, the growth rate is even more rapid, attaining similar increases within 24 hours. Growth is much slower on raw beef or pork at both temperatures (1 11). Yersinia seeded into boiled eggs and boiled fish grows rapidly at 4°C (121). Y. enterocoliticn inoculated into pasteurized liquid eggs also grew well at low temperatures, but it was inhibited by the addition of 5% salt (122). Intofu (soybean curd) and pasteurized whole milk, Yersinia grows well at temperature ranges of 3-25°C (1 lo), however, growth may be slowed by the presence of psychrophilic microflora (1 14). Yersinin can also proliferate in seafoods, but at slower rates. Yersinin seeded in shucked oysters at 0-2°C or 5-7°C showed slow increases in numbers over 14 days (123). In raw shrimp and cooked crab meat stored at 5"C, Yersinia grew rapidly during the first week but declined in number with additional weeks of storage (123). Oxygen-free vacuum packaging and saturated carbon dioxide (CO,) controlled-atmosphere packaging (CAP) have been used to extend the shelf life of refrigerated foods. Gill and Reichel found that Y. enterocoliticn could grow in saturated CO, CAP beef at 50°C (124). Hudson et al. found that Y. enterocoliticn could grow slowly (generation time of 13.9 h) in vacuumpackaged sliced roast beef held at 30°C and even more slowly (generation time of 77 h) when held under saturated CO2 CAP. When held at - 1.5"C growth was slowed with a generation time of 32.1 hours in vacuum packaging and did not multiply under CO, CAP (125). B. Isolation Methods Yersinia spp. have been isolated from clinical samples by direct plating on a variety of differential and selective agars used for enteric bacteria. These include Snlmonelln-Shigelln agar with (126) or without deoxycholate, bismuth sulfite agar, and Macconkey agar with (127) or without Tween 80. The use of these selective media at 37"C, however, is not recommended for Yersinin isolation since the organism grows slowly and may be outgrown by other enterics. Furthermore, the virulence plasmid of Yersinin, pYV, is unstable and may be spontaneously lost during overnight growth at 37°C. Cefsulodin-IrgasanNovobiocin (CIN) agar, an agar specifically designed for isolation of Yersinin spp. at 32°C is useful for isolation (128). Methods to isolate Yersinia from foods are problematic due to competition and overgrowth by food microflora. Direct plating is rarely successful unless Yersinia numbers are very high. Thus, isolation of Yersinin from foods necessitates the use of enrichment techniques. Many enrichment techniques take advantage of the tolerance of Yersinia to growth temperatures as low as 4°C. As the enrichment temperature is lowered, the rate
Yersinia
481
of growth is decreased and the time of enrichment increases. Cold temperature enrichment of samples in phosphate buffered saline for 2-4 weeks at 4°C improved recovery (129). Better results were obtained by including sorbitol and bile salts (130) and peptone in the enrichment medium as well as shortening the enrichment period to 10 days by increasing incubation to 10°C (13 1). There are also methods to selectively enrich for virulent Yersinin strains in foods. Modified Rappaport broth (126) as well as Irgasan-tarcacillin-potassium chlorate broth (132) are effective for Y. enterocoliticn 0 : 3 and 0:9, but not as effective in the enrichment of other bio/serotypes. Toora et al. proposed the use of modified tryptic soy broth (MTSB) supplemented with Irgasan (133). Bhaduri et al. modified this procedure by enrichment in MTSB without Irgasan at 12°C for 24 hours with shaking followed by addition of Irgasan and further selective enrichment at 12°C for 24-48 hours (134). Pathogenic Yersinin species in foods may also be detected using genetic methods. One such method is DNA colony hybridization, where DNA fragments containing sequences encoding traits known to be present in the target bacteria are labeled with a marker. These fragments then can be used to detect the hotnologous sequence in situ in suspect colonies on agar plates. Excised DNA fragments from the conserved calciumdependent region of the pYV virulence plasmid (45) were radiolabeled and used in DNA colony hybridization analyses of Y. enterocoliticn seeded into foods (135-137). The probes effectively distinguished potentially virulent Yersinin from normal flora in foods without the need for enrichment if present in high numbers. Synthetic oligonucleotide sequences derived from known sequences of genes determining pathogenic traits have been radiolabeled and used as probes for DNA colony hybridization. Such probes specific for the virulence plasmid have been evaluated, including one directed to the plasmid region that encodes for cytotoxicity to HEp-2 cells (138,139) and another that is specific to the yopA gene of the plasmid that encodes for the temperature-inducible outermembrane protein YadA (see Section V1.E) (140). Analyses for Y. enterocolitica in several types of seeded foods showed the detection efficiencies of these probes to range from 33 to 100% (135,138,140). Variations in food matrices and differences in the concentrations of normal flora in the foods affected probe sensitivity. Several probes specific for the invasion genes on the Yersinia chromosome have also been developed, but the potential of these probes to identify Yersinin species in foods have not been tested. The invasion-specific probes and other assays used for testing virulence of Yersinin are discussed in Section VI. F. Recently, nonradioactively labeled DNA probes have been prepared by polymerase chain reaction to Yersinia virulence genes and tested to identify plasmid-bearing Yersinia. They were also used successfully to detect Yersinia from artificially contaminated foods (141). The polymerase chain reaction (PCR) has also had numerous applications to the detection of potentially virulent Y. enterocolitica and Y. pseudotuberculosis from foods. It was first used by Wren and Tabaqchali (142) to detect potentially virulent Yersirzia species by amplification of the virF transcriptional regulator gene for outermembrane proteins important to the virulence mechanism. Kwaga et al. (143) targeted primers to the ail (attachment and invasion locus) gene of Y. enterocoliticn in a PCR assay. Nakajima et al. (144) combined the Wren and Tabaqchali primers with those the inv gene specific to potentially virulent Y. pseudotuberculosis and with primers for ail specific for potentially virulent Y. enterocoliticn. These primers were mixed in a multiplex PCR to simultaneously detect and differentiate potentially virulent Yersinin species. To this multiplex PCR, Weynants et al. (145) added the twist of adding primers specific to the gene encoding the 0: 3 antigen, a serogroup often pathogenic for humans. This allowed not only speciation
482
Minnich et al.
and potential virulence of suspected isolates, but also serotype identification in the same reaction. One technical hurdle to the use of PCR for detecting bacterial DNA in environmental samples was the inhibition of the reaction by sample matrix components. To detect the presence of potentially virulent Y. enterocoliticn in fecal samples, Ibrihim et al. (146) used chemical extraction of DNA with PCR targeting yst (heat-stable toxin gene) of Y. enterocolitica. Harnett et al. (147) combined extraction of DNA with multiplex PCR to detect the presence of Y. enterocoliticn in artificially contaminated fecal samples using primers for ail, yst,and virF. Rasnlussen et al. (148) devised a method for PCR detection of Y. enterocoliticn cells from naturally contaminated pig tonsils. Swabs of tonsils were cold enriched for 7-10 days before treatment via immunomagnetic separation (IMS) to selectively remove target cells of Y. enterocoliticn from the enrichment broth. After IMS, Yersinia were detected by PCR with primers to ail. By using this IMS-PCR method, 80% of pig tonsil samples were positive compared to 68% by cultural techniques. Bhaduri and Cottrell(l49) have expanded this approach by swabbing various artificially Contaminated foods using a short selective enrichment and centrifugation and treatment with proteinase K prior to multiplex PCR for Y. enterocoliticn virF and uil to detect target bacteria. Despite the promise of these exciting developments in molecular methods, PCRbased detection methods have an inherent disadvantage of not isolating the contaminating culture for confirmatory tests. For this reason, they have not been adopted in routine testing for Yersinia in foods. However, PCR-based methods may have a place as screening tests for focusing culture-based techniques on those samples likely to be contaminated. Both Y. pseudotuberculosis and Y. enterocolitica are fairly resistant to alkaline conditions; therefore, pretreating enrichment cultures for a few seconds in 0.5% KOH prior to streaking onto selective agars can reduce the numbers of competing microflora (17). The selective media used for clinical sample analysis are also applicable for isolation of Yersirzinfrom foods. Among the more effective are Macconkey and Bismuth sulfite agars, but CIN agar has also proved to be reliable. CIN may be the preferred isolation medium because it is more differential for Yersirzia species, however, it isalso slightly more inhibitory to some Y. pseudotuberculosis strains (150). Combining the use of CIN and MacConkey’s agar with incubation at 30°C for 24 hours was found more successful than either alone (151). Another selective agar medium designed specifically for isolating Yersirlia is virulent Yersinin (VYE) agar. This medium has antibiotic supplements as well as esculin and was developed to differentiate potentially virulent biotypes from environmental strains (152). Recently, Bhaduri and Cottrell have suggested differentiating potentially virulent strains on Congo red brain heart infusion agarose medium after enrichment (153).
C. Identification Methods Conventional methods for the identification of Yersirzin species require a battery of biochemical tests. These are reviewed in detail elsewhere (10,151). Several commercially available miniaturized biochemical test kits, as well as automated microbial identification systems, have also been suggested for identifying Yersinia and other enterics (154,155). Many of these systems have been evaluated and are generally good for presumptive identification of Yersinia to the genus level but are not as reliable for speciation or biotyping (154,156-159). Conventional biochemical tests can be used to supplement the kits for speciating Yersiniu isolates and must be used for complete biotyping of Y. enterocolitica isolates.
Yersinia
VI.
483
PATHOGENICITY
As stated in the introduction of this chapter, the Yersinia are not large contributors to foodborne disease relative to other enteric pathogens. However, the disease-associated yersiniae have contributed tremendously to our understanding of molecular pathogenicity. Studies on the pathogenic Yersinia spp. have provided a common thread that has led to the unraveling of a pattern of common strategies for virulence in not only mammalian pathogens but plant pathogens as well. Indeed, many of the basic principles in molecular pathogenesis were first established in the yersiniae. Hallmarks of these discoveries include the first association of plasmids with virulence, the role of iron in the infectious process, the isolation of the first bacterial proteins involved with the invasion of eukaryotic cells, elucidation of the type I11 protein secretion system, and the importance of temperature as a key environmental cue for virulence gene activation. How did the yersiniae become so important in dissection of such key host-parasite interactions? In part, this is attributable to the common physiology among the pathogenic yersiniae, findings in one Yersinia species commonly extrapolated to the others. More importantly, many of the phenotypes associated with specific virulence attributes are pronounced, lending themselves well to genetic analyses. For example, all pathogenic yersiniae are calcium dependent at 37"C, and suppressors of this phenotype (i.e., calcium independent) were shown to be nonvirulent (159). Thus, from the early days, researchers had easily scorable phenotypes with which to work. Finally, as recombinant DNA and molecular genetics methodologies were developed for analyses of Escherichia coli and Salmonella typhimurium, these techniques were easily applied to the related yersiniae. There are a number of recent and excellent reviews on Yersinia virulence and particularly type I11 secretion systems (for recent reviews, see Refs. 160-163). In fact, over the past 8 years there has been such an increase in information that a book may be the only means to cover every detail. That being said, this section will focus on the essence of our present understanding of Y. erzterocolitica virulence, emphasizing an historical perspective as to how the Yersinin have been used so successfully as a model system. The pathogenic mechanism of Yersinia is highly complex and not fully understood, however, it involves multiple factors encoded by both chromosomal and plasmid genes. These virulence factors include endotoxin, an enterotoxin, several invasion genes, ironregulated proteins, and a large number of plasmid-encoded factors that are important in the process of virulence. These factors as well as motility are under a highly complex system of genetic regulation affected by temperature, DNA structure, histone-like proteins and other as-yet unidentified parameters.
A.
Endotoxin
Yersinin endotoxin is not a major factor in foodborne illnesses, but it is significant in clinical infections. Between 1987 and 1991, Y. enterocolitica was implicated in seven of the eight fatalities due to transfusions of bacteria contaminated red blood cells in the United States (164), and similar incidences of bacteremia and endotoxic shock have been reported around the world (74,164,165). In most instances, the source of Yersinia contamination appeared to be asymptomatic blood donors (1 64,165), and the blood used for transfusion showed no gross evidence of contamination (166-168). Other bacterial species have also caused sepsis (170), but Y. enterocoliticn is most commonly implicated, and it
Minnich et al.
484
is estimated that so far it has caused about 40 cases of transfusion-related infections worldwide, with a mortality rate of greater than 60% (171). The higher frequency of transfusion-mediated Yersinin endotoxemia as compared to other bacteria may be attributed to several characteristics of this organism. First, Y. enterocolitica has the ability to proliferate in cold temperatures (166,168,179). At blood storage temperature of 4OC, 1 cfu/mL of Y. enterocoliticn inoculated into blood will undergo a lag phase of 10-20 days, then rapidly proliferate to lo8 or lo9cfu/mL (166,168). During log-phase growth, endotoxins produced may reach concentrations of 240-600 ngl mL (166). Second, iron affects the in vitro growth response of Y. enterocolitica (172). Hemolysis of red blood cells during storage may provide a readily available source of exogenous iron to stimulate growth of yersiniae (169,173). Third, Y. enterocoliticn produces plasmid-encoded outer proteins, designated Yops (discussed below), that enable the organism to resist the bactericidal effects of human serum and phagocytosis by macrophages and polymorphonuclear leukocytes. These abilities promote the survival of this organism in stored blood (169,174,175). A number of solutions have been proposed to prevent transfusion sepsis by Y. enterocoliticn, however, none of these seems to be viable (170,171). Screening and eliminating donors with gastrointestinal symptoms is only partially effective, because a third of the donors implicated in Yeminicl transmission did not exhibit gastrointestinal illness. Since Yersinin undergoes a 3-week lag phase before exponential growth in blood, it was proposed that the existing storage time for blood cells be lowered to 3 weeks. This, however, would have a devastating effect on blood supplies worldwide. Similarly, to include antibiotics in blood units to control Y. enterocoliticct may result in greater problems due to side effects or allergic reactions to the drug (170,171). Screening blood units is another alternative, but it is labor intensive and logistically difficult. The presence of Y. enterocolitica in blood may be detected by microbiological plate counts or by the use of hematology stains such as acridine-orange, Wright, or Wright-Giemsa stains (164). The level of Yersinia endotoxins in blood inay be determined by the Linzulus amebocyte lysate assay (165,166). A polymerase chain reaction assay may also be potentially applicable for the specific identification of Y. enterocolitica contamination in blood (176). All these methods, however, lack sensitivity and will not beable to detect the low numbers of Y. enterocoliticn cells that may be present at the time the units of blood are collected. B.
Enterotoxin
Pathogenicity of Yersinin may also involve the production of a heat-stable enterotoxin (ST) that is detectable by intragastric injection of culture filtrates into suckling mouse or by the rabbit ileal loop assay (10,177.178). The production of ST has been observed mainly in isolates of Y. enterocolitica but not in Y. pestis or Y. pseudotuberculosis (179). The ST of Yersirzia closely resembles the heat-stable ST toxin of enterotoxigenic Escherichia coli (10,180) in terms of heat resistance and pH stability (181,182). The molecular weight of 9700 estimated for the ST of Y. enterocoliticct is slightly larger than that of E. coli (178), but the toxins are immunologically cross-reactive (180,183). Both toxins also share common mode of action in stimulating cyclic guanosine monophosphate levels in the suckling mouse (182,183). In spite of these similarities, however, the implication of Yersinia enterotoxin in foodborne diseases is not fully established and remains controversial. For example, the Yersirzia ST toxin is not produced in vivo (10,184) and not detected in vitro at temperatures above 30°C (178), hence cannot be produced in the mammalian host. Also, there
Yersinia
485
wasno correlation between production of ST and pathogenicity in mice (185-187); not all pathogenic serotypes of Y. enterocolitica produced ST (177,178), and other Yersinia species that were generally regarded as non pathogenic were also found to produce ST (188,189). Therefore, these factors tended to suggest that ST wasnot involved in virulence. Although ST is not produced in the host, intoxication due to ingestion of prefomled ST in foods was raised as a possibility for public health concern (188,190). Heat and pH stability of ST (1 8 1,182), as well as findings that 32-100% of Y. enterocoliticn isolated from milk were toxigenic (188), tended to support this concern. Other studies, however, showed that Yersinia produced ST in media only at 25°C (187,188), but not at 4°C (190). Also, despite good bacterial cell growth, no ST was produced in foods, milk, or milk-like media at 4 or 22°C (188,190). The absence of ST production at these commonly used food storage temperatures, plus the fact that no cases of foodborne intoxications due to Yersinia ST have been reported, would dispute the likelihood that performed Yersinia ST is a factor in foodborne illness. The S toxin is encoded by the yst gene on the chromosome of Y. enterocoliticn. Unlike physiological studies, genetic analysis of yst tended to suggest that there is a correlation between the presence of vst gene and pathogenicity of Y. enterocolitica (191). For example, infection of rabbits with isogenic ,st+ and yst- mutants confirmed that only yst+ strains of Y. erzterocolitica were able to cause diarrhea in rabbits (179). Similarly, DNA hybridization studies of pathogenic and nonpathogenic strains showed that yst-homologous gene sequences were present only in pathogenic Y. erzterocolitica and a few isolates of Y. kristensenii (191). A PCR assay for the yst gene also confirmed that a yst gene-like product was amplified from Y. kristensenii (192). Since Y. kristensenii are generally regarded as nonpathogenic, the presence of yst genes in this species would seem to contradict the role of ST in virulence. But subsequent hybridization studies using a DNA probe specific for the internal regions of the yst gene showed that only pathogenic isolates of Y. enterocolitica reacted with the probe (191). Although Y. kristensenii do not appear to produce ST, this species was found to be pathogenic to iron-loaded mice; hence, it may be carrying some other virulence factors (193). Finally, although ST is not produced by Y. enterocolitica at temperatures above 3OoC, transfer of the yst gene into E. coli resulted in the production of active toxin at 37°C (179). This suggests that yst gene expression in Y. erzterocolitica may be thermoregulated, analogous to the expression of the invasion gene, inv, in Y. pseudotuberculosis (194). More recent studies seem to provide stronger evidence that ST is a virulence factor in Yersinia. Production of ST may be dependent on the age of cultures, as fresh isolates of pathogenic Y. erzterocolitica were found to produce ST, but in strains of older collections yst had become silent, resulting in no toxin production (191). The absence of yst gene expression would account for the lack of ST production by some pathogenic serotypes, observed in earlier studies. Finally, genetic analysis of yst regulation showed that, even though the gene was not expressed at 37"C, low pH and increased osmolarity conditions, similar to that in the lumen, can induce yst to express at 37°C (195); hence, it is possible for Yersinin to produce the ST toxin in the infected host.
C. CellularInvasion Pathogenicity of Yersinia is closely associated with the presence of a virulence plasmid. However, isogenic strains of Y. enterocolitica and Y. pseudotuberculosis, with or without
486
Minnich et al.
the virulence plasmid, showed equal efficiency in invading tissue culture cells (196-198). Therefore, the ability of Yersirzia to invade mammalian cells is chromosomally encoded. An invasion gene, designated inv, was first identified on the chromosome of Y. pseudotuberculosis (199,200). Transfer of the i m gene into E. coli conferred the invasive phenotype to E. coli (200). Expression of the im?gene in Y. yseudotube~-culosis is thermoregulated (194). The gene encodes a 103 kDa outer membrane protein (invasin) that binds to integrin receptors on mammalian cells, which stimulates Yersinia uptake (301,202). A similar i m gene and another gene, ail (attachment invasion locus), that also encodes for cellular invasion were identified on the chromosome of Y. erzterocoliticn (1 34,135). Although the inttasin protein of Y. enterocoliticu is slightly smaller than that of Y. pseudotrrberculosis, the inv genes from these species are 73% identical at the DNA level, 77% identical at the protein level, and the proteins share similar antigenic epitopes (136-1 38). The nil gene encodes for a 17 kDa Ail protein that also promotes attachment and invasion of tissue cells (133,138). Interestingly, Ail confers varying degrees of invasiveness to various pathogenic Y. erzterocolitica serotypes (133). Analyses of Ail proteins showed slightly different amino acid sequences, which may account for the differences in invasiveness among these serotypes (133). The invasion phenotype encoded by inv enables invasion of several cell lines (130,139), while that encoded by nil exhibits tissue cell specificity (134,139). The ability of Yersinia species to invade mammalian tissue cells appears to be correlated with pathogenicity because many avirulent yersiniae are not invasive (5,127,139). Analyses of Y. elzterocolitica isolates showed that inv gene sequences were present in both pathogenic and nonpathogenic strains (96,139). However, further analysis showed that the i12v sequences of avirulent strains were nonfunctional and these isolates were noninvasive (140). In contrast, genetic studies showed that the presence of ail is closely correlated with pathogenicity. DNA probing of disease-causing Y. erzterocolitica strains showed that only pathogenic isolates that were tissue cell invasive carried the nil gene (139). The presence of nil, therefore, appears to be closely associated with the potential for virulence in Y. enterocolitica (134,139). While im? and ail are the primary invasion genes in Yeninin, there is evidence that the enteropathogenic serotypes also have a virulence plasmid-mediated pathway for adhesion and low-level entry into mammalian cells (133,141,142). Although the precise function of invasin and ail proteins in cellular invasion has not been determined, studies showed that mice orally challenged with imp mutants of Y. pseudotuberczdosis exhibited slower rates of infection (143). This is consistent with the hypothesis that these invasion gene products enable Yersiiliu to penetrate intestinal cell walls to allow colonization and that the contribution of the virulence plasmid to pathogenesis probably occurs after penetration of the cell membrane (127,128,144,145). D. Iron-RegulatedProteins Iron acquisition is an important factor in Yersinin pathogenicity. Isolates of Y. kristensenii (215) and low-virulence Y. enterocolitica of serotypes 0 : 3 and 0:9 are lethal in the mouse model only when allowed to infect iron-loaded mice (172). In contrast, highly pathogenic Y. pseudotut7erculosis, Y. pestis, and Y. enterocolitica produce a set of conserved chromosomally encoded proteins to synthesize and uptake the Yersirlin siderophore yersiniabactin. Yersiniabactin's iron-complexing property is due to a phenolic group and three fivemembered heterocyclic thiazole moieties that serve as iron ligands (216). Thegene cluster
Yersinia
487
encoding this iron uptake system is genetically unstable. Its spontaneous loss, or rearrangement, has been associated with large chromosomal deletions (102 kb high-pathogenicity island), loss of pigmentation, and virulence attenuation in Y. pestis (reviewed in Ref. 217). In Yersinin enterocolitica biotype 1B strain 8081, these genes are clustered on a 45 kb high-pathogenicity island (HPI) inserted into a resident tRNA-Asn gene (218). Pelludat et al. (219) examined this HP1 in biotype 1B strain WA-314 and identified five iron protein genes (Irps) within a 13 kb stretch of this region. These identified genes encode Irpl and Irp2, encoding previously designated HMWPl and HMWP2 (high molecular weight proteins 1 and 2) (220,221). These two proteins are antigenically cross-reactive (222). The predicted amino acid sequence of Irp 1 (3161 amino acids specifying a protein of predicted mass of 384.6 kDa) shows the C-terminal region to be highly homologous to Irp2 (HMWP2), accounting for immunological cross-reactivity. HMWP2 is a nonribosoma1 peptide synthetase, and recent biochemical studies show it to have the capacity for thiazoline formation (heterocyclization) (223). More recently, Gehring et al. (216) reported the complete sequence of the yersiniabactin region of Y. pestis. Their biochemical studies on Y. pestis yersiniabactin synthesis suggest that it is assembled in a modular fashion by HMWPl and HMWP2. Intermediates are passed from the amino terminus of HMWP2 to the carboxyl terminus of HMWP1. Because the genes for yersiniabactin are conserved, it is presumed that Y. enterocoliticn and Y. pseudotuberculosiswill have similar pathways. In summary, the genes for sequestering iron from the host have been identified, and the biochemistry of yersiniabactin assembly should be known in detail in the near future.
E. Plasmid-AssociatedVirulence 1. Yersinia Virulence Plasmids All three Yersinia pathogenic species require 2.5 mM calcium ion at 37°C. Without calcium at elevated temperature, the cells can only proceed through approximately two generations followed by growth arrest. Calcium dependence is, however, an unstable phenotype. That is to say, spontaneous suppressors can be isolated at frequencies of 1 X lops.This phenomenon, discovered in the 1950s (159), is similar to mutation rates associated with loss of plasmids and prophage or later identified transposition-associated events. Indeed, in 1980 a large virulence plasmid of 40 megadaltons (70 kb) correlated with calciumdependent growth and virulence in each species (186,224,225). Loss of calcium dependence at 37°C was associated with the loss of the virulence plasmid. Production of V (protein) and W (lipoprotein) antigens have long been recognized as virulence-associated properties of Y. pestis and Y. pseudotuberculosis (226-228). However, when pathogenic Y. enterocoliticn were also found to produce immunologically identical antigens (229), it suggested that all disease causing Yersiniae shared common mechanisms of pathogenicity. Studies on virulence of Y. enterocolitica serotype 0:8 showed that the production of V and W antigens was closely associated with calcium-dependent growth at 37”C, which in turn wasdependent on thepresence of the plasmid (186). Genetic studies and mouse models (186,197,226,229-232) verified the correlation of this plasmid withthe production of virulence antigens and with pathogenicity of Y. erlterocolitica. This plasmid, originally known as Vwa due to its association with V and W antigens, is genotypically designated as pYV. The related plasmid of Y. pseudotuberculosis is commonly designated as pIB, but in subsequent discussion the virulence plasmid from different pathogenic Yersinia species will be collectively referred to as pYV.
488
Minnich et al.
The pYV plasmids of Yersinia species are similar in function (230,231) and size (233). However, DNA hybridization studies showed that these plasmids are not identical. The pYV of Y. enterocolitica and Y. pestis are only 55% homologous in DNA sequences (233), and homologies ranging from 58 to 87% have been reported even within species (197). For example, among pathogenic Y. enterocolitica, pYV of serotypes 0 : 3 and 0:9 are 90% homologous, but they shared only 75% homology with pYV of serotype 0:8 (234). In spite of these differences, the extent of sequence homologies among pYV of Yersirzia would suggest a common origin for this plasmid. A key observation in the early 1980s involved the correlation of the virulence plasmid with additional specific proteins. Portnoy et al. (197,235) showed that the outermembrane proteins of pathogenic yersiniae changed with temperature, but only in pYV' cells. Furthermore, mutations in specific regions of the pYV plasmid resulted in virulence attenuation or loss, and these changes were reflected by specific changes in outermembrane protein profiles. In addition to encoding calcium dependency and V and W antigen production, the pYV was shown to be responsible for other virulence-related phenotypes. These included auto-agglutination, serum resistance, uptake of crystal violet and Congo red dyes, cytotoxicity for tissue culture cells, production of Yersinin outer membrane proteins (Yops), guinea pig conjunctivitis (Sereny reaction), and lethality for mice and other animals (10,55,235). Two key observations were made with regard to pYV-encoded proteins in the late 1980s. First, Heesemann's laboratory showed that the outer membrane proteins induced at 37°C were actually secreted into the extracellular environment under calcium-limiting conditions (236). This was unusual because, at the time, the general dogma was that grampositive organisms secrete extracellular proteins, whereas gram-negative organisms, with few exceptions (e.g., E. coli hemolysin), export proteins to the periplasnl or the outermembrane, not the extracellular milieu. Second, Michiels et al. (237) reported that these proteins were secreted in an unprocessed form, i.e., secretion did not require proteolytic cleavage of a signal sequence. Hence, the Yersir~inouter membrane proteins were unique in this respect and appeared to be sec-independent. This property was confirmed. Originally referred to as Yops (for Yersinicr outer membrane proteins), with these two observations the Yop moniker for this set of proteins was modified; Yop was retained but now refers to Yersiizia outer proteins as opposed to the original context of outer membrane proteins. The secretion of Yops can readily be displayed in the laboratory. Y. erzterocoliticn is grown to mid to late exponential phase in calcium-depleted tryptic soy or BHI broth (Le., supplemented with 20 mM sodium oxalate and 20 mM MgSO, or some other chelation agent such as EGTA) at 25-30°C. The cells are then shifted to 37OC, and after 24 hours strands of proteineous material (insoluble Yops) can be spooled from the culture or precipitated from culture supernates with solvent or salt. SDS-PAGE gels can then be used to resolve individual secreted proteins. By the early 1990s the focus of work was well defined. Significant data indicated secreted Yops used a unique export pathway, they comprised a set of proteins essential for virulence, and their synthesis was induced by temperature. Hence, the stage was set to define each of these aspects: What is the nature of the export pathway? What is the function of each individual Yop? and How is the system regulated at the genetic level? In the ensuing 8 years, immense progress has been made in each of these areas. In particular the elucidation of the Yop secretion pathway led to the realization that this is acommon mechanism employed by numerous gram-negative pathogens for protein export (both plant and animal pathogens). Surprisingly, it was also determined that the mechanism of Yop
Yersinia
489
secretion is related to flagellum biosynthesis. These secretion mechanisms have now been termed the type 111 system. 2. Type Ill ProteinSecretion The type 111protein-secretion system defines a set of dedicated secretory ports, or channels, on the surface of a gram-negative organism. About 20 conserved proteins are required to form the protein export channel. For Yersinia, this set of conserved proteins is denoted as Ysc and Lcr proteins (for Yersinia secretion and low calcium response). In contrast, the secreted proteins using this export pathway can vary considerably. That is, one finds the secretion machinery to be highly conserved from Yersinia to Xanthomonns, but the virulence proteins secreted are directed toward specific attributes of the host. Conserving the secretion apparatus in this manner, yet varying secreted proteins, has undoubtedly played a significant role in conferring gram-negative bacteria the ability to parasitize a wide spectrum of hosts ranging from mammals to plants. Other gram-negative human and animal pathogens employing type 111 protein secretion as a component of virulence expression include Scdmonella spp., Shigella spp., enteropathogenic E. coli (EPEC), and Pseudomonas neruginosn. A number of additional pathogens are suspected to operate type I11 systems based on preliminary sequence data. Plant pathogens with characterized type I11 systems, comprising HRP proteins involved in the hypersensitivity response, include Pseudomonas syringne, Xnnthomonns cmnpestris, Erwinin anzylovorn, and Rnlstonia solanacenarum (for recent reviews, see Refs. 160-163). For Y. enterocolitica, it is estimated that there are about 9-12 type I11 portals per organism. This is equivalent to the number of flagella made at 530°C (the relation of type I11 secretion and flagellum biosynthesis is discussed below). Uniform secretion of Yops from all export sites occurs in vitro when calcium ion is limiting, but secretion is polar and contact dependent in vivo when the organism encounters a mammalian host cell, as discussed below. In hindsight, however, in vitro Yop secretion in limiting-calcium medium may be a fortuitous artifact that led to the elucidation of this remarkable system. The milestones in the dissection of this system follow. The first insight into the mechanism of Yop secretion was made serendipitously. Two independent laboratories working on the developmental regulation of flagellum biosynthesis in the dimorphic organism Caulobacter crescentus sequenced the flagellar gene JEbF (now designatedJEhA)(238,239). When this sequence was compared to the GenBank, the predicted amino acid sequence was strikingly similar to the predicted protein encoded by Y. pseudotuberculosis 1cr.D. The IcrD gene maps within the highly conserved region of all pYV plasmids. Because the predicted amino acid sequence similarity between FlbF and LcrD was approximately 7096, it suggested a common function. Yet why a flagellar gene in the unrelated innocuous aquatic bacterium C. crescentus was related to a gene essential for virulence in Yersinin was enigmatic. Using phoA fusions, it was shown that both C. crescentus FlbF and Y. enterocolitica LcrD proteins were localized to the inner membrane and showed similar membrane-spanning domains (238,240). Mutations in lcrD abolish Yop secretion (240) and JEbF mutants fail to assemble flagella (238,239). These observations were the first hint that flagellum biosynthesis and Yop export were somehow related. As more DNA sequences became available in both flagellar and Yop secretion genes, the similarity between the two systems became more evident. Other flagellar homologs of Yersinia type I11 secretory components are listed in Table 1. All of these genes have the common feature of being required in either flagellar protein or Yop export. Hence,
Minnich et al.
490 Table 1 Yersinia Type I11 Secretion: Lcr/Ysc and Flagellar Homologous Proteins
1. 2. 3. 4.
5. 6. 7. 8. 9. 10.
Yop secretion
Flagellar assembly
LcrD YscN YscQ YscR Yscs YscT Ys c u YscD YscJ YSCL
FlhA FliI FliY Flip FliQ FliR FlhB FliG FliF FliH
Source: Ref. 163.
the bacterial flagellum, generally regarded solely as an organelle for motility, was now further appreciated in a new venue-its more subtle function as a dedicated secretory portal for flagellar-specificproteins. Flagellin subunits, like Yops, are secreted in an unprocessed (sec-independent) manner. Furthermore, extracellular flagellar components, such as the hook and flagellin subunits, are translocated through the core of the basal body (241). When one considers the amount of flagellin protein comprising the flagellar filaments, one can conclude that the flagellar basal body is an exquisite secretion apparatus as well as a fine motor. Likewise, the amount of Yops secreted under pemlissive conditions is striking (up to 20% if total cellular protein) (160), being visually evident in culture supernates. Until these parallels were recognized, secretion of Yops was generally assumed to be firstdirected to the outermembrane and then uniformly released across the surface upon calcium ion chelation. As the number of virulence proteins showing similarity to proteins required for flagellar protein secretion increased, this view changed. This semblance suggested that Yops may be secreted from a dedicated portal analogous to the basal body of the flagellum. This connection between Yop and flagellar protein export was experimentally reinforced by the following observations. Genetic evidence had previously determined that YopE was responsible for cytotoxicity of eukaryotic host cells (242). Similarly, YopH was known to inhibit phagocytosis by macrophages (243), and Guan and Dixon (244) determined that YopH was a protein tyrosine phosphatase (PTPase). However, pure preparations of Yops had no cytotoxic effect when added to tissue culture cells (245). Furthermore, the possible role of YopH as an extracellular PTPase was not understood. Rosqvist et al. (246) determined that cytotoxicity of purified YopE required direct microinjection into Hela cells. Further, they showed that YopE cytotoxicity imparted by Yeninin cells in tissue culture was dependent upon YopD. That is to say that yopD- mutants could secrete YopE, but such mutants were not cytotoxic to host cells. Hence, the logical implication drawn by these authors from these experiments was that, in vivo, the Yop proteins are delivered across the eukaryotic membrane into the target cell cytoplasm and that delivery required functional YopD. This point was elegantly proven by two separate approaches. In what will undoubtedly be viewed as a classic experiment, Rosqvist et al. (247), using confocal laser microscopy with fluorescent-tagged anti-YopE antibodies, showed that the Yops were vectorally secreted into the host cell from the side of the bacterium in direct '
Yersinia
491
contact with the host cell membrane. Additionally, these experiments showed that Yops emanated from distinct sites on the surface of the organism as opposed to a uniform release. Similarly, Sory and Cornelis (248) and Sory et al. (249) showed this same effect employing YopE- and YopH-adenylcyclase reporter fusions. This strategy employed fusing the calmodulin-dependent adenylcyclase domain of Bordetelkr pertussis cyclolysin to the secretion competent N-terminus of YopE. Y. enterocolitica harboring these fusion genes were allowed to infect tissue culture cells, and cAMP levels were monitored in the cells and in the outside medium. Because the fusion protein adenylcylase activity is dependent upon calmodulin, which is only present within eukaryotic cells, increased cAMP levels reflected internalization of the hybrid protein. These authors were also able to show that adenylcylase activity of the fusion protein was dependent upon YopB and YopD. With these two important studies, Yops were importantly recognized as two categorical groups: host cell effector Yops (e.g., YopE and YopH) and Yops required to translocate these effectors across the eukaryotic membrane (e.g., YopD and YopB). In summary, between 1990 and 1995, it was shown that Yop secretion is related to secretion of flagellar proteins, that Yop secretory sites are distinct on the cell surface, and that Yops are directionally delivered, or injected, into the cytosol of target host cells. These type I11 secretion organelles have recently been visualized under the electron microscope in SaZmorzeZlu typhinzuriurrz.Termed "needle-like" structures, they are strikingly similar to flagellar basal bodies in form (250).
3. Yop Function Below is a brief summary of effector and translocation individual Yop functions. It should be noted that each Yop has a cognate chaperone, designated Syc. These are covered in detail in the recent review of Cornelis et al. (160). Yop Effectors YopE. This 23 kDa protein accumulates in the perinuclear region of target host cells. It has cytotoxic activity causing host cells to round up due to disruption of the cytoskeleton. It has sequence similarity to P. creruginosn Exoenzyme S and the SptP protein of Scdmorzelln (160). YopP/J. YopP (Y. enterocolitica)and YopJ (Y. pseudotuberculosis) induce apotosis of the host cell. The protein has a molecular weight of 30 kDa and appears to remain localized in the cytoplasm of the host cell (160). YopT. YopT is the most recently described effector Yop (251). Mutants deleted for other effector Yops but secreting functional YopT disrupt the cytoskeleton of target cells. Secretion requires an intact type I11 apparatus and the YopT cognate cytosolic chaperone, Syc. The molecular weight of the protein is 35.5 kDa. YopM. Mutations in YopMgreatly attenuate virulence in mouse studies. This protein shows significant predicted atnino acid similarity to the SIzigelZcr IpaH protein. Early studies showed that this protein-mediated thrombin binding and in vitro studies showed it inhibited thrombin-induced platelet activation (160). Although this suggested an extracellular function, more recent studies show that Y. enterocolitica YopM is secreted into host cells (252). More detailed studies on Y. pestis YopM have shown it is targeted to the host cell nucleus (253). YopH. YopH is the best characterized Yop, with biochemical analyses and crystal structure confirming its function as a protein tyrosine phosphatase (PTPase)
a.
492
Minnich et al.
(244). It is localized to the cytoplasm of the host cell, where it has been shown to dephosphorylate p13OCas and FAK (focal adhesion kinase) by Persson et al. (254). This activity prevents signal transduction between integrin receptors and intracellular signaling transducers. In macrophages, this activity inhibits phagocytosis. YopO/YpkA. This Yop has a molecular weight of 84 kDa. Designated YopO in Y. erzterocolitica, but in Y. pseudotubercuZosis it was designated YpkA when the predicted amino acid sequence was shown to have high similarity to serine/ threonine kinases of eukaryotes (Yersirlia protein kinase A). Mutant strains lacking YopO function show greatly reduced virulence in mice. In the mouse model, a y o p 0 - mutant passes normally through early stages of infection, but it does not colonize the spleen. The target protein of YopO/YpkA has not been reported (160). b. Yop Translocation-Associcrted Proteins Less is known at this time about the proteins involved in contact delivery of the Yop effector proteins described above. A brief synopsis is given below in the order of the protein molecular weights.
LUG. The LcrG protein (ca. 11 kDa) forms dimers and contains a heparin-binding domain. Mutations in IcrG abolishes directed translocation of all Yop effectors even though lcrG- strains are deregulated for Yop expression. That is, mutations in ZcrG confer the calcium-blind phenotype (160). YopK/YopQ. This protein has a molecular weight of ca. 21 kDa, and preliminary analysis of the role of this protein implicates it in the control of Yop translocation by modulation of the YopB pore. YopD. As mentioned above, YopD mutants were essential in demonstrating that Yop effectors are secreted into the host target cell cytoplasm (246,249). This protein, (33.3 kDa) is hypothesized to form the entry pore in the host cell membrane in association with YopB (160). LcrV. This protein is described as the protective antigen against plague and one of the early Yeniuia proteins described. Mutations in LcrV diminish YopB and YopD export. It is hypothesized that LcrV forms an integral component of the secretiodinjection apparatus ( l 60). YopB. This protein (41.8 kDa) is also essential for delivery of Yop effector molecules into the host cell (249). It associates with YopD and demonstrates homology with the RTX family of a-hemolysins and leukotoxins (pore-forming toxins). As such it is implicated in disrupting the host cell membrane. Purified YopB can cause lysis of sheep erythrocytes and disrupt lipid bilayers. YopB also displays sequence similarity to IpaB or ShigeZZa JEexneri and PopB of Pseudomonos aerugirlosa ( 160). Two additional proteins are implicated in the control ofYop effector secretion. YopN (32.6 kDa) is secreted in large quantities at 37°C in calcium-depleted medium. It is also synthesized at 37°C in the presence of calcium, where it is localized to the outer membrane. YopN is the hypothetical "lid" to the secretion channel and interacts with TyeA. This latter protein, recently reported (255), has a molecular mass of 10.8kDa
Yersinia
493
and maps immediately downstream of yopN. Mutations in both yopN and QeA, like the aforementioned ZcrG, give a calcium-blind phenotype. In summary, the pYV-encoded Yops are induced by host temperature (37°C). Upon host cell contact, the translocation Yops form a pore in the host cell membrane through which the effector Yops are passed into the cytoplasm of this target cell. The contactdependent signal during this interaction stimulating Yop transfer is not known. However, it is tempting to speculate that activation of host cell receptors stimulates a flushing of calcium ion from the eukaryotic cytoplasmic membrane uncapping the export channel. 4. Virulence Gene Regulation
a. Secretion-Driver? Feedback Circuit on Transcription The regulation of flagellar genes served as an additional paradigm for investigating the control of Yop synthesis and secretion. Transcription of both flagellar and yop regulons is enhanced once protein secretion initiates. In other words, both systems canbe viewed as "secretion-driven" expression systems. An understanding of this "feedback" system was elucidated first for the flagellar regulon. Until the early 1990s it was unknown why mutations in any flagellar basal body gene (flagellar class I1 gene) resulted in transcriptional repression of flagellin, chemotaxis, and motor genes, all of which are positioned below basal body genes in the regulatory hierarchy (class I11 flagellar genes) (241). Hughes et al. showed the mechanism explaining this phenomenon (256), and it has a direct parallel in Yop synthesis. During flagellum biosynthesis, S. typhirnurium expresses an anti-sigma factor, FlgM, that directly antagonizes the sigma factor for flagellin gene transcription, FliA (oF). This interaction holds FliA in check until basal body synthesis is complete and competent for secretion of flagellin. Once flagellar hook assembly is complete, the anti-sigma factor, FlgM, is secreted from the cell through the basal-body hook complex. This dilution of FlgM from the bacterial cytoplasm allows FliA to then activate the next tier (class 111)of structural genes (flagellin, chemotaxis, and motor genes). Mutations in basal body assembly prevent FlgM secretion and, therefore, prevent FliA from promoting class I11 gene transcription (256). The parallel to the FliA-FlgM interaction for Yersinia yop gene expression was readily recognized based on published genetic studies. In fact, Rimpilainen et al. (257) speculated that Y. pseudotuberculosis LcrQ might play the equivalent role of FlgM several years before it was proven based on this genetic evidence. What are these similarities in regulation? First, reporter fusions to a number of yop genes showed they are induced upon a temperature shift to 37°C. However, yop gene expression remains at a low level when calcium ion is present (258-261). Chelation of calcium ion promotes secretion of Yops, which in turn results in a dramatic increase in yop gene expression levels. Therefore, calcium, in a not-yet-understood manner, blocks secretion and gene expression. Mutations in Y. pseudotuberculosispYV-encoded 1crQ phenotypically express and secrete Yops even in the presence of calcium ion, implying its role as a negative regulator comparable to FlgM. Secretion of LcrQ was shown in 1996 and fit the predicted model nicely (262). Interestingly, these findings could not be initially reproduced in Y. enterocolitica, which secretes a homolog of LcrQ termed YscM. This was because the Y. erzterocolitica pYV plasmid contains two copies of yscM. Mutations in bothyscM genes phenotypically resemble 1crQ mutants (263). The target of LcrQ (YscM)-negative regulation has notbeen determined.
494
Minnich et al.
b. Temperature arzd Gene Regulation Temperature has a profound effect on Yersirzicr spp. within the narrow range of 30-37°C. Shifting cells to 37°C includes the following changes in cell phenotypes; motility is lost, porin synthesis is repressed in pYV+ cells, calcium-dependent growth initiates, the vop regulon is induced, and the cells morphologically change from single cells to chains that undergo autoagglutination (264). The Y. er~terocoliticcryop genes are generally dispersed around the pYV plasmid organized as monocistronic units (reviewed in Ref. 265). Transcription of Y. ellterocoliticct vop genes is dependent upon the positive activator VirF and temperature (261). VirF is a DNA-binding protein with sequence similarity to the AraC class of helix-turn-helix proteins (266). DNA footprinting experiments have identified the consensus VirF-binding site, TTTTaGYcTtTat (Y indicates C or T), for yopE, yopH, virC, and the 1crGVsycD.v opBD operon (267). The Y. enterocoliticcr virF gene is induced by temperature, even when placed in E. coli (266.268). In contrast, the homolog of VirF in Y. pestis, LcrF, is constitutively expressed, but yop gene transcription still requires elevated temperature (37°C) (269). When Cornelis et al. (270) fused Y. erzterocoliticn virF to the inducible tcrc promoter, artificial VirF overproduction at 30°C did not induce yop gene expression. Hence, both the natural constitutive expression of LcrF in Y. pestis and artificial manipulation of virF in Y. enterocolitica show that temperature is an integral signal in yop gene induction in addition to the positive activator. Several lines of evidence have implicated temperature-dependent changes in DNA topology as a factor in vop induction. Cornelis et al. (271) isolated constitutive yop mutants by transposon mutagenesis. DNA sequence identified the insertionally inactivated gene, ymoA (for Yersirzin tnodulator of virulence), to encode a histone-like protein. This protein is highly similar to the regulator of E. coli hemolysin expression, M z n , which has been shown to affect DNA supercoiling (272). YmoA mutants are not only constitutive for yop expression, but are VirF-independent. Yet temperature still enhances yop transcription in y r ~ o A -mutants (271), suggesting that at 30°C the yop promoters are structurally constrained and that temperature somehow alleviates this constraint. Rohde et al. (268) determined that induction of yop genes and repression of flagellar genes are coincident with changes in DNA supercoiling upon a temperature upshift from 25 to 37°C. Moreover, low levels of the DNA gyrase inhibitor novobiocin were shown to induce supercoiling, yop gene expression, and repress flagellin gene transcription at 25°C. As such, pharmacologically inducing supercoiling atlow temperature elicited a high-temperature phenotype. These authors also isolated a class of novobiocin-resistant mutant (DNA gyrase) that was constitutive for yop expression, like the aforementioned VmoA mutants, and determined that these mutants were also nonmotile (268). These data suggested that the reciprocal regulation of the flagellar and yop type I11 regulons was coordinate. Examination of the temperature requirement in yop expression by both groups suggested temperature-induced changes in DNA topology. mediated by YmoA and DNA supercoiling, appear to be required in the phenotypic temperature phasing of virulence genes between 30 and 37°C. Another factor in thermoregulation also indicates DNA topology is critical. The pYV plasmid is enriched for regions of intrinsic DNA bending (268) (J. R. Rohde, X. Luan, H. Rohde, J. M. Fox, and S. A. Minnich, in press). Furthermore, these intrinsic DNA bends can influence DNA supercoiling. Most striking is the fact that these bends are maintained by temperature until 37OC, whereupon they ‘melt’ (J. R. Rohde, X. Luan, H. Rohde, J. M. Fox, and S. A. Minnich, unpublished). Based on these facts, Rohde et
Yersinia
495
al. (268) proposed the following model to account for thermoregulation in Y. enterocolitica. At low temperature (35 pg/L in four urine specimens from the cases. Although botulism from dairy products is considered rare, some recent outbreaks show that this does occur. In 1989, the largest outbreak in theUnited Kingdom took place after commercially prepared hazelnut yogurt containing contaminated nut paste was consutned (30). In Italy in August and September 1996, at least 8 persons and one death resulted after tiramisu made with mascarpone cheese was eaten (3 l). Most of those stricken were children. The cheese contained spores of C. botuZi~zum,but it is uncertain how the contamination took place. The widely exported product was recalled on a massive scale. Overall, in Italy there were 33 cases of botulism in 1995 and 58 in 1996. Typical foods were vegetables preserved at home in oil or salty water or in sauces. In 1998 two cases of botulism, one fatal, occurred in a family in southern Italy after they ate home-preserved mushrooms bottled in oil (32). In the United Kingdom the link between human Creutzfeldt-Jakob disease (CJD) and consumption of beef that originated from cows with bovine spongiform encephalitis (BSE) was considered close enough to warrant bans by other countries on importing British beef. This type of CJD is supposed to be caused by a variant form of prion (V-CJD) that is found in bovine organ tissues. V-CJD is characterized by early age of onset, more prolonged duration, predominantly psychiatric presentation, and distinct brain pathology. In the United Kingdom, 25 persons diagnosed with V-CJD have died. The economic loss in 1996 to the European Community was $2.8 billion in subsidies to the beef industry for herd slaughtering and lost business (33). 4. Foodborne Disease Reports in Specific Countries The following are examples of some European countries that have reported additional foodborne disease data.
528
Todd
a. France. In France, the number of outbreaks rose from 594 in 1990 to 732 in 1992. Where the agent was identified, Saknzorzella caused 83-8796 of outbreaks. Eggs and egg products were associated with many outbreaks, particularly after they were contaminated with S. enteritidis (1). Meat and meat products, as well as mixed foods, were also important vehicles for causing Salmonellcl, C. pelfrifzgens, and S. nureus outbreaks. Most fish- and shellfishborne outbreaks were caused by histamine and diarrhetic shellfish poison, respectively. Outbreaks involving dairy products such as cheese and ice cream and muchhandled prepared foods most likely arose from S. aureus intoxication. Outbreaks most frequently occurred after people had eaten at homes, schools, restaurants, canteens, hospitals and homes for the aged, holiday resorts, prisons, and religious gatherings. The most frequent contributing factors were contaminated equipment, faults in processing, inadequate cooling, contaminated raw ingredients, preparation too far in advance, and contamination through personnel. Botulism has been reported in France for many years. Of the 108 cases seen in a Poitiers hospital between 1965 and 1990, 83% had consumed home-cured ham contaminated with C. bofulinurntoxin, mainly type B (34). Raw goat milk cheese contaminated with Salmonelka parntyphi B infections affected 273 persons throughout the country in 1993 (35). Contaminated pork tongues in aspic sold in delicatessens was responsible for a major listeriosis outbreak with 279 cases, 22 abortions, and 63 deaths in 1992 (36). Consumption of wildboar meat led to two family outbreaks of trichinosis in 1993 and 1995 in the Languedoc region (37). Horse meat has been implicated in trichinosis outbreaks in France. The latest occurred in 1998, when 128 persons who ate horse meat imported from the Federal Republic of Yugoslavia were infected (38). b. Poland. The number of outbreaks in Poland steadily decreased from 991 in 1988 to 640 in 1992 (1). Salmonellosis fell from 35,268 in 1988 to 24,573 in 1992, botulism from 357 to 165, other bacteria from 4232 to 2633, and mushroom poisoning from 489 to 284 over the same years. S. aureus intoxications, however, remained much the same (517-491). From 1988 to 1992, most illnesses were associated with contaminated cakes, sweets, and ice cream (39.4%), meat dishes, including raw minced meat and eggs, poultry, and venison (25.8%), ready-to-serve food, e.g., croquettes, pancakes, dumplings, mayonnaise, salads (9.3%), eggs and egg products (4.0%), and milk and milk products (2.5%). Dishes prepared with uncooked eggs are common especially for fillings, e.g., custards, cheese and egg crepes, and pastries. Most outbreaks occurred at homes, with the most frequent places of contamination being farms, homes, schools, canteens, and vacation resorts.
c. Former U.S.S.R. Countries. The rate of salmonellosis per 100,000 in former U.S.S.R. countries (some in Europe and some in Asia) in 1990 were 16.7 (Georgia), 25 (Belarus for 1991-1992), 28.0 (Azerbaijan), 28.9 (Tajikistan), 32.0 (Ukraine), 37.2 (Uzbekistan), 37.7 (Turkmenistan), 38.2 (Armenia), 42.1 (Kyrgystan), 5 1.7 (Kazakhstan), 70.4 (Russian Federation), and 83.1 (Moldova) (1). S. enteritidis was the most frequently isolated serovar in Belarus, Moldova, and the Russian Federation and was often associated with poultry and eggs. In Russia, up to 85% of foodborne salmonellosis was caused by S. enteritidis (39). In the Russian Federation, salmonellosis rates rose from 70.4 in 1990 to 80.1 in 1992. In Russia, only 400-500 cases of campylobacteriosis are reported each year, probably because of lack of laboratory media and suitable incubators. There were 369 cases of botulism on average each year from 1986 to 1991 (death rate, 7.9) compared
"
Surveillance of Foodborne Disease
529
with 661 cases between 1992 and 1994 (death rate, 8.1%); the increase was due to the changing economic conditions and an increase in home canning and curing of mushrooms, vegetables, fruits, and fish (39). In Belarus, the last outbreak from brucellosis was in 1969 with 27 cases and 3,689 infected animals. However, yersiniosis occurs every year with a rate of 7.0 per 100,000 in 1991. In Moldova, most salmonellosis cases occur during the warmer months when large quantities of food are prepared and poorly stored. Moldova is reported to be free from trichinosis, the last outbreaks being in 1982. In July 1998, Kazakhstan officials have temporarily banned shashlik kebabs in Almaty, the country's capital, after five people were hospitalized as a result of eating meat from animals infected with anthrax (40). No deaths were reported. Over 100 restaurants and cafes in the capital had to find alternative dishes to serve their customers for over a week.
V.
SURVEILLANCEPROGRAMS IN AUSTRALIAyNEW ZEALAND, AND OCEANIA
A.
Australia
1. Surveillance In Australia, some trends in notifications of enteric diseases are apparent in 1991-95 data. Notifications for Campvlobacter, Listeria, and Salmonella isolations are increasing, and those for Yersinia are decreasing (Table 3). In 1995 there were 10,933 cases of campylobacteriosis (91.6 per 100,000 population), up over the previous 4 years; high regional rates were recorded from South Australia and Northern Territory (41). In the same year there were 5,895 salmonellosis cases (32.7 per 100,000population), with more cases recorded in the warmer months. There were also 29 cases of brucellosis, 58 of listeriosis, 734 of shigellosis, 69 of typhoid, and 305 of yersiniosis, but no cases of botulism. A study of campylobacteriosis cases in Tasmania showed a decrease in notifications after 1992, which coincided with the introduction of an infection control program in commercial chicken farms (42). 2. FoodborneDiseaseOutbreaks A summary of foodborne disease annual reports for the years 1980-1995 was published by Crerar et al. (43) (Table 4). Although the types of agents responsible for illness are similar to those in other industrialized countries, with Salmonella being the predominant cause of morbidity and mortality, the figures are relatively low for the 15-year period, and many more outbreaks probably occurred. From the data available, S. enteritidis does
Table 3 Notifications of Selected Enteric Diseases in Australia, 1991-1995 Year
Campylobacteriosis
Salmonellosis
1991 1992 1993 1994 1995
515 8,672 9,135 450 8,311 10,117 10,933
44 5,440 4,6 14 534,73 l 5,283 5,895
Source: Ref. 41.
Shigellosis 902 567 894 708 724 306 734
Listeriosis
Yersiniosis
38 34 58
414
Todd
530 Table 4 Foodborne Disease Outbreaks in Australia, 1980-1995, by Etiological Agent ~~
Agent Salmonella C. pelfrirtgews S. aureus Ccrnlpylobucter B. cereus V. parallrreiilol~ticus L. monocytogenes E. coli 0111 C. botulimm SRSV Rotavirus Hepatitis A virus Toxoplasnrn Scombrotoxin Ciguatera Mushroom poison Total known Total unknown Total
No. of outbreaks
Percent of outbreaks
No. of cases
Percent of cases
No. of deaths
27 14 9 5 5 4 2 1 1 11 1 1 1 2 1 1 86 42 128
53 11 7 4 4 3 2 0.8 0.8 9 0.8 0.8 0.8 2 0.8 0.8 67
2,053 280 99 106 27 181 13 23 1 2.267 55 7 13 8 30 5 4,438 1.5 14 5,932
35 5 2 2 1 3 0.2 0.4 0.02 38 0.9 0.1 0.2 0.1 0.5 0.08 75
5 0 1 0 0 2. 0 1 0 0 0 0 1
33 100
2 100
0 0 0 6 0 6
Source: Ref. 43.
not seem to be a major problem in Australia. Vibrio parahuemolyticus infections were more common than in North America or Europe, but this probably reflects the fact that most of Australia's population are close to the sea and consume seafood regularly. Scombrotoxin and viral outbreaks were also associated with seafood. S. snbandaka, which caused at least 54 cases of illness in Victoria and South Australia in 1996, arose from consumption of one brand of peanut butter (44). Individuals infected with this rare serotype occurred in other states and territories, and about half of the cases were younger than 5 years; links to peanut butter were only established in three Western Australian cases. One case also occurred from the same brand of peanut butter in New Zealand. The product was recalled in both countries. Another salmonellosis outbreak in the same year affected 52 persons-patients, staff, and spouses-from sandwiches served in a Brisbane hospital (45): the source was not identified. An outbreak in Melbourne in 1997 affected more than 700 people, about 3% of whom developed arthritis and half of whom are expected to have long-term problems (46). Food poisoning from tropical fish containing ciguatoxin and related toxins originating from benthic dinoflagellates is regularly reported from Queensland and Northern Territory coasts originating from a variety of fish including barracuda, Spanish mackerel, kingfish, coral trout, grouper, and reef cod. However, underreporting is suspected, and the rate is probably higher than the documented 1.6 cases per 100,000 population. In one outbreak in 1995, a family of four who ate a coral trout and developed typical gastrointestinal and neurological symptoms were given intravenous mannitol within 18 hours of the poisoning (47). The mother was 11 weeks pregnant, and the resultant child was observed for 2 years and suffered no ill effects.
"_
'
Surveillance of Foodborne Disease
531
Although sporadic cases of HUS have been associated with E. coli 0 1 11 and other VTEC in Australia, the first'outbreak was in 1995 when mettwurst produced by a small manufacturer infected many people and caused HUS in 23 children with one death (48). The manufacturing process was criticized for its lack of control procedures (no starter culture but the use of "backslop" inoculation of the meat ingredients, no monitoring of pH or water activity, no pasteurization) (49). This outbreak stimulated considerable interest in VTEC infections and HUS. In a study of 55 cases of Queensland children with HUS, most of those preceded by diarrhea were under 5 years of age ( S % ) , had reduced or no urine output (71"m), experienced hypertension (3 1"m), and developed seizures (29%). Eighty-five percent required transfusion, 56% antihypertensive therapy, 56% peritoneal dialysis, and 2% hemodialysis, which lasted 3-29 days, and 10% ventilation (50). One child died. In two other HUS incidents in Adelaide, E. coli 048:H21 and Enterobncter cloacae OR:H9, respectively, seem to have been responsible (51,52);both strains produced SLT-I1 toxins. In February and March 1996, the first 0157 outbreak occurred in Australia when 6 persons were infected from food served in a delicatessen on the Gold Coast in southeast Queensland (53). A food handler was the index case and may have contracted the infection from her pet dog, which had bloody diarrhea the week before she developed symptoms.
B. New Zealand 1. Surveillance In 1995 in New Zealand, the rates for campylobacteriosis, salmonellosis, hepatitis A, shigellosis, listeriosis, and VTEC disease were 223.0, 40.4, 10.3, 5.6, 0.4, and 0.2 per 100,000, respectively (54). The estimated rates have also been determined for giardiasis (158.1), rotavirus infection (141.4), yersiniosis (87.4), and cryptosporidiosis (66.6). There are probably about 300,000 cases of foodborne illness each year. For a New Zealand population of about 3.6 million, this represents one illness per 12 persons each year. Surveillance of listeriosis was conducted in 1995. There were 15 cases compared with 1l in 1994 (55). Three of the cases were perinatal, and two of the infants died. Of the remaining 12 cases, 11 had underlying disease or were elderly. None of the cases occurred in clusters, and there were no links to food. Risk factors for campylobacteriosis were determined through a case-control study from June 1994 to February 1995 (56). The main factors were consumption of raw or undercooked foods (especially poultry and unpasteurized dairy products) and untreated drinking water, overseas travel, and contact with animals. Thorough cooking of chicken could significantly reduce the incidence of campylobacteriosis. 2. Foodborne Disease Outbreaks In New Zealand, illnesses have been documented from C. botulinurrz in home-bottled meat and watercress, C. pe~fiingensin meat, chicken, and bean dip, Salmonella in pork, Canzpylobacter in chicken livers, and B. cereus in rice (57). More recent outbreaks include three separate episodes of S. Qplzirmriunz phage type 35 associated with consumption of bakery products in Christchurch in 1993 (58), two incidents in 1994, with people ill after eating curry probably contaminated with Clostridizm perfringens, one with 31 persons at an Auckland wedding reception in March (59) and the other with 59 attendees at a fashion show in April (60). The two C. per@ingerzsoutbreaks probably involved the same supplier, although this is not stated in the reports, and in one of them the practice for years had
Todd
532
been for large pots of meat to be left at room temperature because they were too big for the cold room. Hepatitis A occurred in Wellington in 1996 from delicatessen food contaminated by the owner/operator, who was the index case (61). In 1998, 64 people attending a local Maori hui outside Auckland and eating lunch consisting of roast pork suffered from gastroenteritis 1-3 1 hours later (mean incubation period, 12 hours) (62). The pig was home-killed and not inspected before being roasted without a meat thermometer. It was cooled for 90 minutes before serving. C. perfringerzs was the agent responsible. Outbreaks following huis have occurred before in 1997 and 1998. Illnesses associated with traditional Maori foods, however, are rarely reported.
C. Oceania Relatively few South Pacific countries reported diseases like salmonellosis (7 of 22 countries with a total of 221 ill), hepatitis A (10 countries, 121 cases), shigellosis (1 1 countries, 271 cases); however, practically every country recorded diarrheal illnesses (a total of 1,352/100,000 population) (63). Also, fish poisoning was significant (a total of 4,707 cases), with most poisonings in Fiji (1,653) and French Polynesia (856). On a population basis, these were highest in the Solomon Islands (2,125/100,000) and Kiribati (1,152/ 100,000). Although the details of these poisonings are not given, most would probably be ciguatera (64), PSP (65), or scombroid (histamine) poisoning (66). In many Pacific Islands seaweed is served as a side dish at meals. In 1994, boiled seaweed (Grncilurin coronopifolin) served at a picnic in Hawaii affected 7 persons with a burning sensation in the mouth and throat, headache, and gastrointestinal symptoms (67). Extracts of the seaweed killed mice, and therefore the etiological agent was probably a heat-resistant toxin. Illnesses from seaweeds Gracilarin and Grncilariopsishad been previously reported from Guam, Japan, and California with deaths in two of the episodes.
VI.
SURVEILLANCEPROGRAMS IN ASIA
Relatively little in the way of surveillance of foodbome disease is carried out in Asian countries. Most information is gleaned from specific but limited investigations and studies. Some studies cover several countries, e.g., parasitic diseases are widespread arising from consumption of raw or undercooked freshwater fish, shellfish, snails, frogs, and tadpoles. For instance, echinostomiasis, caused by the intestinal trematode fluke Echinostornu that encysts in freshwater mollusks as a secondary host, is prevalent throughout eastern Asia (e.g., 3% in the Philippines, 11-65% in Taiwan, 5% in China, 1% in Indonesia, and up to 50% in some parts of Korea and northern Thailand) (68). The disease is caused by at least 16 species transmitted by snails, the primary host. The disease is most prevalent in remote rural locations among low wage earners and women of childbearing age. Risk factors are promiscuous defecation and the use of night soil for fertilization of fish ponds. In addition, the likelihood of fungal intoxication is greater in areas where grain is stored over long periods of time. For instance, the cost of premature death from primary liver cancer and disability and morbidity to liver cancer attributable to aflatoxins in maize was estimated to be $A32.0 million in Indonesia, $A36.8 million in the Philippines, and $A7.7 billion in Thailand (69). The same estimates for peanuts were $A96.5 million in Indonesia, $A2.2 million in the Philippines, and $A16.0 million for Thailand. The difference in costs
Surveillance of Foodborne Disease
533
is related to the amount of aflatoxin ingested and is greatest in Indonesia and least in Thailand.
A. China Foodborne disease outbreaks are documented annually in China, and data for the years 1993-1995 are presented in Table 5. Overall, there were 1,365 outbreaks, 33,979 cases, and 246 deaths in 1993, 1167 outbreaks, 37,018 cases, and 264 deaths in 1994, and 947 outbreaks, 23,556 cases, and 244 deaths in 1995 (70-72). The decrease in 1995 was because of fewer outbreaks of microbiological and natural toxin etiology. The most frequently occurring outbreaks (in order) were caused by natural toxins, organophosphorous compounds, Salmonella, nitrite, Vibrio parahnemolyticus, S. nureus, Proteus, pathogenic or toxigenic coli-bacillus (probably E. coli), and histamine. Botulism was a relatively infrequent disease (7-1 1 outbreaks and 41-161 cases); the death rate for cases was 6.314.6%. A few illnesses resulting from ingestion of fungal toxins (mycotoxins and gossypol) were documented and, like botulism, had a high death rate (1.3-5.2%). The natural toxins are not defined but probably include seafood toxins such as ciguatoxin, paralytic shellfish toxin, and tetrodotoxin (the last since balloonfish are specifically mentioned in a list of foods). One disease limited to China is deteriorated sugar cane poisoning caused by Arthrinium spp. (molds) that grow on stored sugar cane in the winter months (73). The disease is characterized by sudden onset of gastroenteritis followed by toxic encephalopathy and a delayed dystonia in severe cases. It particularly affects children who like to chew the canes. The mold produces the toxin, 3-nitroproprionic acid, when the cane becomes mildewed because of poor storage. Another unusual disease is caused by Burkholderin cocoverlenans (Pseudomonas cocovenenms subsp. ferinofermentans) in fermented corn flour, potato products, and stored Tremella mushrooms. Poisoning may arise from ingestion of bongkrekic acid produced by the Burkholderia, leading to high mortality rates. Tremella is grown on cottonseed shells in homes and by small industries. Many outbreaks of chemical origin resulted in acute illness and deaths, unlike industrialized countries. The vibrio- and histamine-associated illnesses were probably mainly marine in origin. However, the most frequently implicated foods were meat and mushrooms. More illnesses were recorded in urban communities than in rural environments. The specific responsible groups listed were caterers, other food-service operators, and street vendors. A very large outbreak of hepatitis A occurred in Shanghai in January and February 1988, with 292,301 cases and 32 deaths. The virus was transmitted through clams contanlinated with sewage water (74,75). Hepatitis E can be transmitted by fecally contaminated water or possibly food, and there are endemic areas including China. For instance, in the Xinjiang region more than 100,000 cases were reported between 1986 and 1988 (76). From information available up to 1989 there were 746 outbreaks and 2,866 cases attributed to botulism in China with a 14.7% fatality rate. Most of the outbreaks (71.8%) were from home-fermented bean or cereal products (77). As more commercial products become available to the population, it is expected that fewer home-made products will be consumed and there will be correspondingly fewer botulism outbreaks.
B. Taiwan From 1986 to 1995 the number of foodborne disease outbreaks reported in Taiwan ranged from 57 to 123 (78). The main etiological agents were V. parahaemolyticus (197 out-
534
9 M
U
6
Surveillance of Foodborne Disease
535
breaks), S. aureus, mainly enterotoxin A-producing (169 outbreaks), and B. cereus (104 outbreaks). Over this period there were 31 outbreaks of salmonellosis (1,038 cases) and 10 of botulism, mainly A and B toxin types (19 cases). The most frequent serovar was S. typhimurium, and S. virchow had caused a number of large outbreaks. In recent years S. enteritidis was emerging as an important pathogen. In 1995, there were 123 outbreaks and 4,950 cases, the highest numbers since before 1986. It would appear from this study and that of Lee et al. (79) that the characteristics of outbreaks in Taiwan are more similar to those in Japan than in Korea. One Shigella outbreak in that year affected 646 persons (80). The Vibrio outbreaks occurred in the warmer months (April to October), and some were associated with seafood in lunch boxes delivered to schools or in catered food (78), Vibrio vulniJicrrs incidents also occur every year, and of the 28 cases between 1985 and 1990, most had ingested seafood or had exposed abraded skin to seawater (81). In two unusual outbreaks (1992 and 1994) schoolchildren mistook tung nuts from the tung oil tree (Alezlrites) for chestnuts and ate them; within 2 hours they developed gastrointestinal symptoms, fatigue, and headache (82).
C.
HongKong
From the number of enteric pathogens isolated from stools of patients in a major hospital in Hong Kong(83), it would seem that thedistribution of these pathogens in the population is similar to that in Japan, except that the proportion caused by Shigellu is higher and that of Vibrio is lower (Snlnlonella 52.5%, Campylobcrcter 16.6%, Shigella 11.3%. Vibrio 5.3% and EPEC 4.4%). Hepatitis A and E viruses have been implicated in shellfish outbreaks (84). Episodes of Vibrio yaralzner~zol~ticus and ~ ? u l ~ ~ i Jhave c u s been documented, mainly through consumption of shellfish. Dipping shellfish into hot water (“hot-pot”), a frequent practice, is not sufficient to destroy these pathogens. Seafood toxins, paralytic shellfish poison and ciguatoxin are regularly reported as causing illness in Hong Kong.
D. Korea The characteristics of foodborne outbreaks in Korea and Japan between 1971 and 1990 were compared (79) (Tables 6-8). There were considerable differences in the morbidity (3.0% in Korea, 29.2% in Japan) and mortality rates (2.48% in Korea, 0.07% in Japan), as well as agents involved (Vibrio spp. important in both countries, but Salmorzella was more a cause of outbreaks in Korea than S. c w m s , and vice versa for Japan). Most incidents occurred in the workplace and the home in Korea, whereas it was more in restaurants and hotels in Japan. Seafood was often implicated in both countries, but food of animal origin was much more frequently associated with outbreaks in Korea. The differences Table 6 Foodborne Disease Outbreaks by Etiology in Korea and Japan
Korea Japan
3.O 29.2
2.4s 0.07
23.1 14.s
14.9 47.3 24.8
0.5 0.2
37.6
6.8 3.5
17.1 9.6
536
Todd
Table 7 Foodborne Disease Outbreaks by Place of Eating in Korea and Japan
Country Korea Japan
Home Restaurant Workplace HotelSchool (9%) (%) (%)
38.8 17.2
10.6 32.7
5.3 11.5
(5%)
(%)
Retail store m )
19.1 c. 3
2.5 c. 4
0.5 c. 3
Other Unknown (%)
(%)
11.9 c. 16
1.3 c. 15
Solrrce: Ref. 82.
may have been due to types of food eaten, food-handling customs, and the fact that Koreans typically do not seek help at hospitals. Norwalk-like viruses are a common cause of gastroenteritis in young children, but it is not known how much a role food and water playin their transmission (85). Toxoplasmosis occurred after three men ate raw boar viscera and pork at a farmhouse in 1994 and also after five soldiers consumed raw liver and uncooked meat from a domestic pig in 1995 (86j. Three experienced chorioretinitis (two permanently blind in one eye), and five had lymphadenopathy.
E. Japan In 1987,40.6% of isolates were Snlnlorzelln, 19.9% were Vibrio spp., 16.7% were Campylobncter, 12.4% were E. coli, and 5.6% were Shigella. By 1990, S. enteritidis had become the dominant Snl1nor?elln serovar, and by 1992 salmonellosis was the leading cause of foodborne disease in Japan. In 1997, 11,000 people in 499 incidents of foodborne disease suffered from salmonellosis, three times greater than in 1992 (87). In the first 3 months of 1998, the number of salmonellosis cases increased dramatically with 1,770 people ill, almost five times the average number for the same period during the last 3 years. The figure included a massive outbreak of 1,100 elementary school students in Tokyo and the nearby prefectures of Kanagawa and Iwate. The import of foreign-hatched chicks for both poultry farming and egg production was a possible entry route into Japan for the Salmonella. The rise in Salrnonella incidents has been a trend over the last 5 years. Foodborne disease statistics have been kept by Japan for many decades, and the number of foodborne illnesses occurring each year was highest in the 1950s and 1960s, with up to 2,000 outbreaks reported annually. Little change, however, has been noted since 1982 (in 1982, 923 outbreaks and 35,536 cases; in 1990, 926 outbreaks and 37,561 cases. Slightly fewer outbreaks occurred in 1987 and 1988 (840 and 724, respectively). However, the proportion caused by V.pnrahnemolyticus increased from 33.4% of bacterial Table 8 Foodborne Disease Outbreaks by Implicated Food Grain/
Country Korea Japan
Vegetable Animal Seafood products mushrooms Confectionery (96) 31.8 21.7
+
in Korea and Japan Multiple foods
Unknown Other
m’o)
(%l
(%)
(%b)
(5%)
(%)
25 .O 3.6
17.5 14.6
2.9 1.2
18.3 9.6
1.9 1
2.6 48.3
Surveillance of Foodborne Disease
537
outbreaks in 1982 to 53.2% in 1990. More outbreaks occurred in the summedfall months (June-October), with a peak in August/September. After Vibrio spp., SaImotdla and S . aurezds were the dominant etiological agents. Poisoning from natural toxins (mushrooms, PSP, DSP, tetrodotoxin) accounted for 13.5% and 13.6% ofthe reported outbreaks in 1982 and 1990, respectively. The most serious was pufferfish poisoning (from tetrodotoxin); in 1982 there were 26 episodes and 8 deaths, but only 1 death in 1990. In fact, the reduction in deaths has been consistent over the years, from 41 1 in 1949 to 1 in 1990. Restaurant/ hotels and homes were the places where most food was mishandled that led to outbreaks. Infant botulism occurred for the first time in 1984, with C. botulimm type A spores being found in the infant’s stool, honey fed to the infant, and soil and dust specimens (89). In 1991, 21 persons in an inn near Tokyo suffered from cholera and one died after they consumed contaminated imported Korean clams (90). There had been a cholera outbreak in Korea at that time. Since 1984, there have been episodes of EHEC affecting hundreds of nursery or school children. In the largest outbreak (3 19 cases) there were 2 deaths. For example, one in 1990 affected kindergarten children who suffered neurological problems (e.g., stupor, deep coma, convulsions, tremors, and incontinence) arising from the action of verotoxin (91). A far greater problem occurred in 1996 from May to October, when the same pathogen caused a series of 24 outbreaks, with over 9,000 cases and 12 deaths. Several molecular subtypes of the E. coli were isolated (92). In the largest outbreak, in July and August in Sakai City (Osaka), a total of 6,309 schoolchildren and 92 staff from 62 elementary schools were affected and a further 160 secondary infections developed, mainly in family members of the schoolchildren (93). The high number of severely affected children may be related to an unbalanced diet resulting in protein malnutrition (94). In Habikino City (Osaka) another outbreak affecting 98 persons in a home for the elderly and three other small outbreaks in the same region occurred. All of the strains from these five outbreaks had identical DNA patterns, and radish sprouts from one farm were consumed by those ill. However, no isolates could be obtained from samples of soil, water, or sprouts on the farm. It has been shown, however, that E. coli 0157 can penetrate into the inner tissues of radish sprouts and disinfection will not remove them (95). In 1998, 39 persons suffered from E. coli 0157 infection after they consumed sushi made with soy sauce-seasoned ikzlra in restaurants in five Japanese cities (96). The ikura was produced from roe of domestically caught salmon.
F. Viet Nam In Viet Nan1 it was estimated that 30-57% of students in university hostels between 1984 and 1988 suffered from diarrhea, mainly because of pathogens being in poorly prepared and stored food (97). Street-vended meals in Hanoi in a 1990/91 survey often contained E. coli and C. perfringens. Most of the 5,714 documented illnesses between 1983 and 1988 were caused by Salmonella, E. coli, and S. aureus, and include 156 deaths (fatality rate of 2.7%). Because anthrax-infected cattle may be used for meat, hundreds of persons develop bacteremia, and 3-7 die each year (97). Chemical residues in foods are not adequately controlled, and manyillnesses are thought to be due to chemical poisonings. Deliberate illegal additives in alcohol, candies, and sweet products have caused intoxications; one such adulterated liquor caused 14 deaths. Foodborne disease is the most widespread public health problem in Viet Nam and the second leading cause of illness and death (98), even though cases are highly underreported. Infected food handlers and pesticide residues
538
Todd
in foods were considered to be important risk factors for foodborne disease in Viet Nam, although no direct links were made between these and investigated foodborne illnesses.
G. Thailand Although there has been national foodborne disease reporting since 1970, the focus appears to be mainly on chemical poisonings, particularly from insecticides, although these accounted for only 0.33% of the reported 207,580 cases of food poisoning between 1981 and 2986 (99). More details on poisonings were given in a later study (100). Between 1981 and 1987, insecticides accounted for 27.4% of outbreaks and 58.4% of cases. Because of the widespread use of insecticides, some of these have accidentally Contaminated desserts, beverages, fruits, and other foods. Methomyl, which looks like sugar or flour and with little odor, was responsible for 15 of the 18 insecticide-related outbreaks. Since 1987 this chemical was sold blue-colored in an attempt to reduce these poisonings. In addition, an alcoholic beverage containing methanol affected 10 males during a party: 5 died and one had permanent visual impairment. Poisonous plants, such as mushrooms (Awzcrnitn and CMoropl~yZlumspp.), cassava (raw roots), and wild plant seeds (Jcrtroplm spp.), caused 58.9% of outbreaks and 34% of cases, and poisonous animals, 11.0% of outbreaks and 6.5% of cases. Among the poisonous animals group, there was one episode of PSP with 63 cases (1 death) following consumption of green mussels containing 465-714 mouse units PSP toxins/g. In addition, there were four outbreaks associated with meals made with horseshoe crab meat and three outbreaks resulting from ingestion of pufferfish, presumably containing tetrodotoxin, although no analysis was done. Only serious intoxications seemed to have been documented, but many others undoubtedly occurred, such as the 2,207 reported cases of mild mushroom poisons, some of which could have been of bacterial origin, but no follow-up was done. Shigellosis is an important cause of diarrheal disease in Thailand, especially in children under 5 years of age (101). Whereas most of these are spread from person to person, one waterborne outbreak was identified in a rural region. A case-control study showed an association between the 242 cases of ShigeZ1aJlesner.i and drinking unboiled piped water. The water came from a river and had been unchlorinated for 6 weeks prior to the outbreak because of a lack of disinfection chemicals. Food handlers may have become infected through this source. Salmonellae were present extensively in chicken flocks and environmental samples in a 1991-92 study (102). They were found in all 7 parent breeder flocks and 13 broiler flocks and in 13 of 15 layer flocks examined. Of the 21 serovars isolated, no one was dominant. The shells and contents of eggs from layers were contaminated (4% and 2%, respectively), with S. enteritidis found only on one egg shell. Chicken is an important protein source in southeast Asian countries, and Thailand exports birds to several of these. Therefore, poultry could be a major vehicle for transmission of ScrZmor~elluto infect a large number of people. If virulent serovars such as S. enteritidis become established in flocks, the situation could become worse. H. Cambodia Diarrheal disease is a problem for very young children, since children less than 5 years of age with diarrhea constituted 15%of all outpatient consultations and 17% of those hospitalized. In the 7 most populated provinces in 1993, there were 906 cases of cholera (72 deaths), 13,323 cases of dysentery (2 deaths), 2,242 cases of typhoid (2 deaths), and 33,939 persons with diarrhea (32 deaths) (103). In the whole country in 1996, there were
Surveillance of Foodborne Disease
539
762 cases of cholera with 20 deaths, all from one province, but none of the cases was confirmed. Ice and bottled water is generally unacceptable according to French standards (93.3% for bacterial and 50% for chemical analyses) (103). A survey of restaurants showed that many were not clean or had some evidence of improper personnel hygiene; 75% had foods unacceptable by bacteriological analysis.
1.
New Guinea (Irian Jaya and Papua New Guinea)
Pigs play an important economic and cultural role in the tribes that live in the central highlands of New Guinea. In the Indonesian part of the island (Irian Jaya), the prevalence of cysticercosis, resulting from ingestion of Taenia solium, is the highest in the world according to Muller et al. (104), with a rate of over 30%. Pork is often consumed insufficiently cooked to destroy these parasites. Pig-bel, a severe form of Clostridium pe$ringens enteritis, is associated with ritual feasts involving consumption of roast pork in Papua New Guinea (105). After the feasts, portions of the pork are taken away for later consumption at ambient temperatures that allow germination of the spores and production of enterotoxin. There is of recent interest in kuru because of its possible similarity with V-CJD in the United Kingdom (106). Kuru was a fatal disease contracted though ritual consumption of the brains of dead persons by the Fore tribe in Papua New Guinea. Although this practice was banned in 1959, there are still some kuru-affected people, and questions are being asked about the amount of brain that was eaten and the length of the incubation period. J.
Indonesia
In a 5-month period in 1994, 408 children were monitored for diarrheal disease in Jakarta (107). Of the 36% of children with diarrhea during that time, 19.6% had ETEC isolated from rectal swabs: most of these children were under 2 years of age. These are similar results to those found in a rural area in central Java: ETEC were isolated from 19% of diarrhetic stools from 340 ill children (108). The incidence of typhoid fever in 1989 was high in southern Sulawesi, an island in Indonesia (3.1 per 100,000 population and a 5.1% case-fatality rate) (109). To account for this, a hospital-based case-control study was conducted in the city of Ujung Pandang (1.3 million population) on the southwestern tip of Sulawesi. Those at most risk were single, unemployed, or students or recent graduates of a university, whose lifestyle did not allow many home-prepared meals. It was also found that those hospitalized with the fever tended to eat food from street stalls and were not likely to use soap when washing their hands. Since typhoid fever occurred most frequently during a period when there was little rain, the authors postulated that only stagnant water was available, and any S. qphi present from fecal contamination could survive in it. Street vendors and households generally did not refrigerate food. Those hospitalized were severely ill, most likely because the dose was high as a result of growth of the pathogen in the food. It is possible that those that were not referred to a hospital could have had lower doses arising from water or another source.
K. Malaysia The incidence of food poisoning in Malaysia was 9.62/100,000 in 1981, with the most frequent etiological agents being S. aweus, V. paruhnemolyticus, and Salmonellu (1 10). In 1983 at a school canteen, 48 students eating meehoon (fried rice noodles) developed
540
Todd
S. aureus intoxication. The organism was isolated from the vomitus and nasal swabs of three food handlers (1 10). Although food poisonings are frequent in Malaysia, they are rarely investigated by acceptable epidemiological procedures (1 10). Even in the meehoon outbreak the source of the S. nureus was not identified, since only scanty growth of the organism was obtained from the positive food handlers. The sources of these were not determined, but the V. pnrahnemol~~ticus infections probably arose from consumption of contaminated shellfish. ETEC seem to be a major cause of childhood watery and bloody diarrhea (66% of cases) in Malaysia, based on a study of 107 children with acute gastroenteritis (1 11). Salmonella was present in 10% of cases. Pathogens have been found in domestic foods, including B. cereus and S. nureus; even if there was a final heating stage, heat-resistant toxins could be produced (1 12). Snbnorzella has also been isolated from raw market foods and ready-to-eat items (1 13).An outbreak of Vibrio choleme 0139 with 52 cases in 1993 from a contaminated street food, rojak, led to a study of growth of the vibrios in a variety of street foods (1 14). It was found that strain 0139 could grow in cendol and tofu but did not grow in rojak gravy or noodles. In 1988, 25 ethnic Chinese were ill in a northwestern Malaysian state after eating a Chinese noodle, loh see fun (1 15). Seventeen of these were children who were admitted to hospital with acute hepatic encephalopathy, 13 of whom died. High levels of aflatoxins were found in the blood, brain, kidney, lung, and other postmortem tissues. The mean incubation period was very short for acute aflatoxicosis (8 hours). The cases were from six towns in two districts and had obtained their noodles from food stalls that were supplied by the same family-run factory. The noodles were made from rice, corn flour, tapioca flour, and wheat starch. None of these ingredients was found to contain aflatoxins at the time of testing, but it was assumed that one of the ingredients for one batch of noodles had to be highly contaminated to have such an effect on the children.
L. Singapore The main bacterial pathogens isolated from 7334 patients with diarrhea in Singapore were Salmonella ( 10.1%), Canzpylobncter (1.2%), Shigella (1.1%), Vibrio pnrtrlzaemol~ticus (0.8%), and V. ckolerae 0 1 (0.2%) (1 16). Salmonellosis outbreaks were associated with a variety of foods, such as coconut rice (nasi lemak), mutton curry, and Malay pancake (roti jala). Snlmonellu enteritidis was the main serovar, replacing S. typhinzurium in 1994. In 1995, 188 inmates in a penal institution were infected with S. enteritidis from canned sliced pork luncheon meat served at a lunch on March 26 (1 17). Although the specific cause of this outbreak was not determined, in previous outbreaks at Singapore institutions there were lapses in food hygiene. Another outbreak of salmonellosis, this time caused by serovar weltevreclen in 1996, affected at least 116 workers at a shipyard (118). Various cut fruits and vegetables including watermelon, pineapple, papaya, and vegetables in oyster sauce eaten over a 2-day period were implicated. These were sold from food stalls whose owners used industrial water derived from sewage effluent. Vibrio vul~~ificus was first diagnosed in Singapore in 1985. Since this country is surrounded by tropical waters and has the highest per capita consumption of seafood in the world, the risks of V. vulm!jicus infection are high. It is, however, not a notifiable disease in Singapore, and the true incidence of the disease is not known. Three hospitalized cases examined by Lee et al. (1 19) revealed a history of eating seafood-raw fish, crabs, and steamed cockles. None
Surveillance Diseaseof Foodborne
541
of them had evidence of wound infections. One of them, a fishmonger, died, and another had a leg amputated.
M. India, Pakistan, Bangladesh, Sri Lanka, and the Maldive Islands 1. Epidemiology Cholera is widespread in the Indian subcontinent and seems to be increasing, especially since the emergence of V. cholerae 0139 and drug-resistant strains (120- 122). Contamination of water is a typical problem, often from inadequate sewerage systems, common to many parts of Asia. In 1991-92, in Dhaka, Bangladesh, 451 children with acute diarrhea were compared with 602 matched controls (123). ETEC and EPEC were the only two of six pathogenic groups of E. coli looked for to be significantly associated with diarrhea. ETEC infections tended to be in the wet warm month of August, and EPEC infections peaked in the dry months of February to May. Flies and water samples were reservoirs for Cmzpylobacter, ETEC, Shigella, and Salnzonella (124). However, Canzpylobacter was the pathogen found most frequently in flies (12- 13% in urbadperiurban slums, 4% in hospitals, 2.6% in upper middle class homes, and 2% in markets) and water samples (30% in hospitals, 28% in markets, 12.5% in a village, 5.5% in riverdcanals, 4% in urban/ periurban slums). Therefore, Campylobacter may be a more important cause of morbidity than is currently recognized. This is particularly important now that a link has been made between Campylobacter and Guillain-Barr6 syndrome (GBS) in Kerala State, India (125). From serum samples taken from patients, 26% showed high antibody titers to Campylobacter jejurzi and 38% of stool specimens of new GBS cases were positive for C. j e j u d coli. Cooks and waiters in hotels in Poona had fecally contaminated hands (73% and 44%, respectively) based on coliform counts, and plates, spoons, forks, and kitchen towels were similarly contaminated (126). Apart from coliforms, these contained Salmonella, Proteus, and Pseudomonas spp. Hand-washing practices had not seemed to have improved from a more recent study in rural Bangladesh, partly because soap was too expensive to buy and partly because of religious and cultural practices (127). Similarly, mothers’ hands in rural communities in West Bengal are typically fecally contaminated (40%) compared with those of children (17%) (128). E. coli was also found in leftover food and water (59%) and utensils (27-32%). The contamination was highest in the wet summer season. Breastfed children were less likely to have E. coli isolated from their hands than partially breastfed children. There were 721 outbreaks and 1199 sporadic cases in the twin cities of Hyderabad/ Secunderabad between 1984 and 1989 (129). The majority of outbreaks affected 2-10 persons and occurred between February and June. The main vehicles of transmission were “stale” food (36.5%), rice dishes (23.5%), sweets (12.6%), and curry (9.8%). “Stale” food is probably food that has been left over from a previous meal at room temperature for a lengthy period of time, usually overnight (130). Chicken, pork, goat, and fish were the most frequent components of curries and rice dishes and were most often eaten at parties or in homes. S. aureus and Bacillus spp. were the most likely causative agents since they are often present in these foods. S. aureus was implicated in one outbreak in Hyderabad in which more than 100 persons fell ill after eating a sweet porridge (13 1). Home-prepared foods in small communities also contained pathogens, such as S. aureus,
542
Todd
C.pe@-irzgens,and B. cereus but not Sdmonella. As in India, sweet dishes are vehicles for S. u w e m intoxication in Pakistan (133). One such outbreak from khoa, a confectionery with concentrated buffalo milk, caused eight persons to be hospitalized in the early 1980s. More recently, some samples of khoa obtained from manufacturers in a large Pakistani city contained up to los S. nureuslg. Salmonella was also found in khoa and in cheesebased confectioneries (132). Pulses, ground meat dishes, and chickpeas sold at bus and train stations in the same city contained 104-107 C. pegrirzgenslg, when the holding temperature was not hot enough (38-46°C) (133). Home-prepared foods in small communities also contained pathogens, such as S. uzmus, C. yerfiingens, and B. cereus, but not Salmor2elln. The main hazard identified was holding foods for long periods of time (e.g., overnight) at ambient temperatures (134). In India and many other Asian countries, cysticercosis from Taertin soliurn is a major public health problem because of the widespread consumption of insufficiently cooked pork (135). 2. Foodborne Disease Outbreaks The incidence of salmonellosis has increased in India in recent years, and some outbreaks have been investigated. In 1995 33 persons were admitted to hospital in Maharashtra after consuming vegetarian food served at a party (136). Twenty-three developed high-grade fever 24 hours after eating, one died, and one had irreversible cerebral damage with neurological complications. S. pnratyphi was isolated from 12 stool samples. In the Maldive Islands, a small outbreak of five persons eating salad containing carrots and peas in a cream base was caused by two Salrnomlla serovars: S. stanley and S. ormienburg were both isolated from the stools and the food (137). These isolates were resistant to the same antibiotics. Water was a probable source of an outbreak of EAggEC (EAEC) infection in south India (138). In one village, 69 of 451 residents experienced diarrhea, mostly watery but 24% with blood. The most frequent pathogen isolated from ill but not well persons was EAggEC. Following a feast in a village in Tamil Nadu, 25 of the 48 villagers developed diarrhea, abdominal cramps, and fever. An investigation showed that Yersiniu errterocolitica serotype 3, biotype 4, was the causative agent, being transmitted through buttermilk (139). Although all leftover foods had been discarded, this pathogen was isolated from a patient’s stool and water used to dilute the buttermilk. In addition, high anti-Y. enterocolitica antibody titers were found in sera of two patients, and toxin, determined by suckling mouse bioassay, was found in inoculated buttermilk kept for 6 hours at 4°C and room temperature. Soy milk is being encouraged as a substitute for animal milk in India. Unfortunately, an outbreak involving this product affected 35 of 263 schoolchildren drinking this at a midday meal in Dehli (140). The 30-minute incubation period and mild symptoms were indicative of S. uut-ells or B. cereus intoxication, although only E. coli at > 105/mLwas found in the milk. In a Sri Lankan village the carcass of a freshly dead monkey was made into a curry, and nine persons who ate this were subsequently infected with S. enteritidis PT 8; one 12-year old boy died (141). The curry containing monkey entrails and meat was insufficiently cooked. Mycotoxins have been responsible for illnesses in 1974 from aflatoxin in maize (15.6 ppm) and in 1987 from deoxynivalenol and other trichothecenes in wheat (142-144). In Bombay, 132 persons became ill and 4 died after eating fish that wereharvested from algae-rich water, and an algal toxin was believed to be responsible for their symptoms (145). In Kerala State, 500 persons were hospitalized, 7 of whom died after consumption of Perna mussels in September 1997 (146). PSP toxin levels exceeded 10,000 mouse units/100 g. The subsequent temporary ban on shellfish sales affected the jobs of 1000 families. A similar outbreak had occurred in 1984. Illnesses
borne
of
Surveillance
543
from Vibrio vzrln$cus have occurred from consumption of seafood in southern India based on the finding of up to 107/g in intestines of a variety of market fish (147).
N. Nepal In Nepal 30,000-40,000 people die from gastroenteritis each year. The annual incidence of diarrheal diseases is 3.1-3.3 episodes per child, and 25% of childhood deaths are associated with diarrhea (148). Cholera outbreaks occurred in 1991 (1,800 deaths), 1992 (1,049 deaths), and 1995/96. In the latter outbreak, 1,017 children with acute diarrhea were studied by Pokhrel and Kubo (149), who found that the inadequate water sanitation system in Nepal contributed to the spread of enteric diseases. A WHO evaluation of food-safety practices showed that food for home preparation is generally purchased from markets where the hygiene is poor or from street-vended foods where sanitation is equally bad and the operators have no sanitation training (148). Proper food-safety education was thought to be essential in reducing future enteric illnesses.
VII. SURVEILLANCEPROGRAMSINTHEMIDDLEEAST
A.
Israel
Salmonellosis cases rose from 3,469 in 1989 to 4,542 in 1992, in contrast to a decrease of foodborne and waterborne diseases (all agents including those of unknown origin) from 2377 in 1989 to 1304 in 1992 (1). Campylobacteriosis cases decreased slightly over this period (from 1869 in 1989 to 1439 in 1992). In 1992, there were 5 C. perfringens, 3 Salmonella, 2 Shigellcr, and 2 S. aure-eus outbreaks; the remaining 28 outbreaks were of other or unknown etiology. Cheese, beef, and vegetables were the foods most frequently associated with outbreaks, and mass catering, restaurants, food-processing establishments, and homes were where the contamination was likely to occur. Some of the population, particularly non-Jews and Jews of Afro-Asian origin, were more likely to suffer from bacterial food poisoning; e.g., non-Jews (110/10,000) were twice as likely to be hospitalized as Jews (52/10,000) for bacterial food poisoning. This was probably because of different socioeconomic conditions and traditional food-preparation habits. Many of the non-Jewish communities had primitive water systems, inadequate sanitation, and poor food hygiene (150). This was borne out by a study of 399 hospitalized Arab infants in the West Bank, where there are villages and refugee camps with poor sanitation systems (151). They showed that 44% were infected with ETEC, and dehydration occurred in 58.3% of them and 28.5% failed to thrive. Religious differences are also reflected in the incidence rate per 100,000of the Israeli Arab population for echinococcosis (7.0 for Muslims, 22.5 for Christians, and 45.9 for Druze) (152). Home-slaughter of sheep, hunting of wildpig, and keeping of dogs were factors in thetransmission of theEchirzococcz4s tapeworm within these groups. In 1992, 197 air force personnel presented themselves to the base clinic with pharyngitis (153). Group A Streptococcr4s type 8/25 was isolated from the patients and a food handler. Processed white cheese was the vehicle served at a lunch. The asymptomatic handler mixed the cheese with his hands 24 hours before the meal and placed it in the refrigerator overnight before he went on vacation. It was put on the table for 6 hours at room temperature (28°C) before it was served. There were some secondary cases. No cheese was available for testing. Since this pathogen was shown to be able to grow in
544
Todd
cheese, it was assumed that the organism from the handler mixed in with the white cheese multiplied during the 6 hours on the table. From December 1994 to February 1995, a peanut-flavored savory kosher snack item imported from Israel containing S. ngona infected 27 people in England and Wales and 10 in the United States. Information relayed to Israel helped identify the cause of more than 2,200 phage type 15 S. clgona infections in that country during the same time period (8,9) (see also Sec. IV). B. Jordan The isolation rate of enteric pathogens from healthy Jordanese food handlers was 6% for Salmonella, 1.4% for Shigella, and between 0.4 and 4.9% for eight types of intestinal parasite (154). Although this would appear to be a high risk factor for foodborne illnesses in food-service operations, there was no evidence that foodborne disease in Jordan was directly caused through contamination of food by infected food handlers. Of more significance was the incidence of brucellosis, with 33.2 and 46.2 persons infected with Brucella melitensis per 100,000 population in 1986 and 1987, respectively (155). The source of the Brucella was sheep and goats (infection rates >lo%) and possibly cattle (infection rate 2%). Unpasteurized milk and raw-milk cheese were frequently identified as vehicles of infection. In 17 cases of diarrhea admitted to a hospital in the Jordan Valley, only rota virus, EIEC, and EPEC were detected; six other pathogens looked for were not found (156). Sabnonella is widespread in poultry farms, broiler, layer, and breeder flocks, with 70% of birds having evidence of infection by serological testing (157). The serotypes most frequently isolated were S. gallinarum, S. euteritidis, and S. typhinzzu-irm.The yield of eggplants irrigated with treated effluent was twice that under conventional fertilizer application. However, the wastewater used for drip irrigation was shown to contain high levels of fecal coliforms, e.g., in July, 1.8 X 105/100 mL before chlorination, 4.6 X 10' mL after chlorination, and 8.1 X lo3 mL at the irrigation site (158). The fecal coliform count was over 100 times higher in irrigated soil than in dry soil. All fruits and leaves were negative for Snlmorzella and Shigella and had very low coliform counts.
C. Lebanon Salmonella outbreaks have also been documented from Lebanon (159). Raw meat and poultry products, milk and milk products, vegetables, and foods with multiple ingredients have been vehicles of transmission.
D. SaudiArabia Diarrhea in British troops deployed to Saudi Arabia was partly caused by ETEC (160). This pathogen is probably also present in the local population. The rate of brucellosis is high (617 per 100,000), particularly among the local Bedouin population, which is dependent on the raising of sheep, goats, and camels (161). There were 1.3 incidents and 22.4 cases of foodborne disease per 100,000 persons between 1982 and 1985 in the eastern province of Saudi Arabia (162). S. aureus was isolated more frequently than Salmo12ella from implicated foods, such as milk, fermented milk, cheese, meat, chicken, vegetables, and rice. Workers of Indian or Southeast Asian origin were the groups most frequently affected by outbreaks through mishandling in their work camps. Insufficient cooking and improper storage of food were the main factors contributing to the incidents. In one partic-
Surveillance of Foodborne Disease
545
ular outbreak in 1985, 168 of 419 Filipino workers at a workers’ camp in Damman, Saudi Arabia, contracted salmonellosis (163). A rice/meat/vegetable dish was the suspected food. One food handler and 57 cases were positive for S. rninnesota. Only those served at one meal were infected, and the source of the Salmonella was thought to be raw meat. In one town in central Saudi Arabia, an outbreak of typhoid fever was attributed to a cake (164). The cake containing cream had been kept overnight and served to schoolgirls on a bus the next day. There was a total of 19 ill, both those eating the cake and secondary cases.
E. Iran For those persons seeking medical aid in Tehran after experiencing an apparent attack of food poisoning, the most common etiological agents identified were C. pe~fringens (66.5%), Salmonella (17.8%), S. nureus (6.9%), E. coli (5.5%), B. cereus (1.6%), and C. botuhzum (1.0%). Most of these people (96%) had eaten in restaurants or mass-catering establishments (165).
F. Yemen An epidemic in Hodeida, Yemen, of 149 cases of diphtheria affected mostly children under 5 years of age. Those who had been previously vaccinated against the disease with at least three doses were generally protected (166). There were associations between those ill and consumption of water from a wheeled cart (possibly from direct contact with the infected driver) or locally made yogurt. It has been previously shown that milk can be a vehicle in the spread of diphtheria.
G. Bahrain Several large foodborne disease outbreaks have occurred in Bahrain including salmonellosis from improper handling of food in a hotel. Every school has a cafeteria, but the quality of the food is often questioned as being unappetizing, unhygienic, and containing foreign objects (167). The food workers had no training, and there was a potential for contamination by workers’ utensils, cloths, sponges, etc., growth of pathogens in food left at ambient temperature, and no reheating of items. Education and a HACCP plan were suggested solutions.
VIII.SURVEILLANCEPROGRAMS A.
IN AFRICA
Cholera in Africa
Cholera outbreaks have continued to occur in Africa since the seventh pandemic began in 1970. In a 1984 outbreak in Mali, millet gruel was associated with the illness. Typically, this gruel had curdled goat milk added to it, but at the time of the outbreak there was little milk available because of a drought. Laboratory studies showed that V. cholerae survived less than 6 hours in gruel with curdled milk compared with >24 hours in gruel alone (168). Other foods implicated in African outbreaks are peanut sauces (Guinea), leftover crabs (Guinea-Bissau), and rice meals prepared by persons preparing cholera victims for burial (Guinea and Guinea-Bissau). Direct person-to-person transmission is proba-
546
Todd
bly rare because of the low rate of infection by care workers. However, in areas of crowding cholera spreads rapidly, and refugees are an increasing concern in Africa as a result of civil wars and international conflicts. It has caused severe morbidity and mortality in refugee camps in Malawi, Somalia, Ethiopia, and the Sudan (169). For instance, in April 1997 there was a cholera outbreak affecting 90,000 Rwandan refugees in three temporary camps in the Democratic Republic of Congo, with a high death rate. Between March 30 and April 20, a total of 1521 deaths was recorded, with a crude mortality rate estimated to be 9.9/10,000/day (170). The refugees were severely malnourished, and access by relief workers to the camps was difficult. How much of the cholera was due to food and water is not known, but both of these can be major vehicles of enteric pathogens in such situations. A study of one outbreak in Mozambican refugees in Malawi in 1988 indicated that there was a common source, which could have been food or water, but no environmental Vibrio isolates were found. Heavy rains destroyed some latrines 15 days before the outbreak, probably contaminating the local water table (169). In another outbreak in Malawi, three factors were identified with cholera: drinking river water, placing hands into drinking water in storage containers, and eating cooked peas kept overnight (166). B. EHEC Infections in Malawi and the Central African Republic During 1992, in the Malawan Lisungwi refugee camp holding 60,000 Mozambicans, 772 cases of abdominal cramps and bloody diarrhea were documented (171). The case fatality rate was 4.7%. The major factor contributing to illness was consumption of cooked food from the market. Based on analysis of stool cultures and the presence of the VT1 gene in some of these, the authors concluded that most of the cases were caused by E. coZi 0157:H7 and some by Shigella dysenteriae type 1. In the town of Zemio in the Central African Republic in 1996, 108 cases of bloody diarrhea were documented, with several HUS cases and 4 deaths (172). Viral hemorrhagic fever was originally suspected, and stool samples were negative for enteric pathogens. However, EHEC was eventually diagnosed based on the presence of virulence factors for this group of pathogens through PCR analysis of stools (presence of genes for SLTl and ene in 80% of specimens). In later hospital examinations of patients, E. coli 0157:H7 was isolated from two fatal cases of bloody diauhea. In other patients non-0157 EHEC were suspected. A case-control study showed that consumption of locally made meat pies (kanda) was associated with bloody diarrhea. Kanda is made by soaking smoked zebu cow meat in water for several hours and then mixing it with cooked marrow squash, wrapped in a banana leaf and steamed. It is then displayed in markets or roadside stands for up to several days at ambient temperatures until it is sold. Contaminated zebu meat was subsequently suspected as the cause of the 1996 outbreak.
C. Aflatoxicosis in Mozambique and Other Countries A consequence of consumption of moldy food is hepatocellular carcinoma (HCC); levels of aflatoxin B I have been found as high as 1.5 mg/kg of food. There was a strong association with mutations of the p53 gene in HCC and dietary aflatoxin intake in an international study of patients in 14 countries (173). The incidence of HCC is higher in Mozambique than any other country. Moldy grain, which contains aflatoxins, is often consumed under drought conditions. In Kenya and the Sudan, aflatoxins were present in the sera of children
Surveillance of Foodborne Disease
547
suffering from kwashiorkor but were metabolized in a different way than in children with other forms of lnalnutrition or in normally nourished children (174). The aflatoxins, therefore, may bea contributory factor to kwashiorkor. This is supported by more recent studies of Hendrickse (175). Because of the high humidity in tropical Africa, maize and cassava are often contaminated with aflatoxins, which are also found in up to 40% samples of breast milk, occasionally in high concentrations. Kwashiorkor-affected children have had as much as 4 pg aflatoxindkg body weight. The aflatoxin-induced immunosuppression may explain why the human immunodeficiency virus (HIV) is spreading so rapidly and aggressively in African infants.
D. Toxoplasmosis and Brucellosis in the Sudan and Other Countries Exposure to Toxoplasma gondii is relatively common in Africa, with a 18.2-61% prevalence of antibodies in the population of eight countries (176). In the Sudan, both Toxoplusma and Brucellu infections are probably associated with consumption of raw liver and intestines (177).
Foodborne disease increased in Egypt between 1989 (25 outbreaks, 115 cases, 3 deaths) and 1991 (146 outbreaks, 551 cases, 24 deaths) (178,179). The highest-risk foods in 1985 and 1986 were white cheese, fermented cream, meat/chicken, floudbutter oil, cabbage/ rice, and potatoes, and many illnesses occurred in homes (179). Street-vended food has been shown to contain pathogens, and many foods awaiting sale were at temperatures favorable for microbial growth (15-44°C) (180). Occasionally, acute illnesses directly associated with a food are documented. Between 1983 and 1985 in Egypt, three outbreaks from white cheese and two from cream/fermented cream were caused by EPEC. In 1991, the first major botulism outbreak with 20 deaths arose from ingestion of locally made faseikh (uneviscerated fish) (181). In 1994, three children died and six others suffered from severe diarrhea caused by E. coli 0157:H7 after they ate hamburgers, koshari, and dairy products in Egypt. As a follow-up to this, a survey of 175 foods obtained from slaughterhouses, supermarkets, and farmers' homes was conducted for E. coli 0157. This pathogen was detected in 6% ofunpasteurized milk, 6% of fresh retail beef, 4% of boneless chicken, and 4% of lamb meat samples (182).
F. Algeria An outbreak of botulism in the eastern provinces of Setif and Constantine killed 17 people in July 1998 and made another 100 persons ill after they ate rotten poultry and kashir, a processed meat (183). In addition, during the first halfof 1998, 1400 persons were reported to be ill from foodborne disease of unreported origin.
G. Tanzania,Ethiopia,andKenya In Tanzania, children los enterobacteria CFUl100 g. Infant food such as formula, baby cereal, and traditional “rice water’’ were even more contaminated (56% with >10’ CFU/100 g) because they had been stored at room temperature for up to 8 hours. In the urban slum community it was a common
Surveillance of Foodborne Disease
549
practice to prepare food in advance of a meal and store it at ambient temperatures for up to 24 hours. It was only occasionally reheated, partly because of lack of fuel. Also, women only had time to make food once a day because they worked far from their homes. Unfortunately, no foods were examined for pathogens. These observations indicate that food- and waterborne diseases can be easily spread in such communities and cause a high morbidity. In Abidjan, C6te d’Ivoire (Ivory Coast), drinking water sold in bags to school pupils had up to 10’ E. coli/lOO mL and no residual chlorine present. This had the potential to cause illness ( 194).
J.
Nigeria
When strains of E. coli isolated from children with diarrhea in southwest Nigeria were examined, ETEC, EPEC, EIEC, EAggEC, and VTEC were found (195). The type of foods eaten may determine if they will be infected: e.g., ogi, a fermented maize porridge used for weaning infants in Nigeria, has a low enough pH to prevent growth of Salrzzomlln and EPEC (196). A case-control study in Nigeria showed that diarrhea in households was less related to poor food hygiene practices and more to improper disposal of feces (197). Human outbreaks of brucellosis have occurred in the past, and the disease is still endemic in Nigeria with a 3.1% seropositive rate in cattle (198). Of equal concern is the finding that Mycobacterium tuberculosis is present in slaughtered pigs (in lungs and lymph node), M. bovis in cattle, and M. avian in several animals (199), since human tuberculosis has a fairly high prevalence in the country. Because of limited laboratory procedures, it is not known how much tuberculosis comes from animal sources, but this study suggests that it could be one source of the disease. A 64-year-old male ate fried, prefrozen edible land snails collected locally in 1980. He developed diarrhea and abdominal cramps. His stool yielded a heavy growth of Aeromonas hydrophila, and thawed land snails had 1 X lo9 A. hydrophilalg (200). A 27-year-old male who ate with him did not suffer the same symptoms. A. hydrophila, Salmonella,or Shigella were isolated from about 40% of land snails in eastern Nigeria, indicating that this food source is a potential source of pathogens.
K.South
Africa and KwaZulu-Natal
In South Africa there are an estimated 12 million people without adequate potable water supplies and 21 million without safe sanitation systems (201). In 1995 and 1996, a Shigella dysenteriae type 1 in KwaZulu-Natal epidemic was responsible for thousands of cases and many hundreds of deaths. The impact of diarrhea for these two countries was estimated at 24 million cases in South Africa and 5.4 million (and an additional 76,000 of Shigella dysenterine type 1) in KwaZulu-Natal each year. For both countries, there would be a total of 54,000 deaths, and lost productivity was over 26 million days annually. The healthcare system would be burdened each year with over 5 million visits and 6.23 million hospitalization days. The total costs would be 3,375 million Rand (380 R/household) in South Africa and 905 million Rand (430 R for diarrhea and 67 R for dysentery/household) in KwaZulu-Natal. In South Africa, these figures represent 15% of the annual health budget spent on treating diarrheal disease, which costs 1% of the South African GDP (gross domestic product). These are the direct costs. The true costs may be 2.4 times higher. If these figures are extrapolated to all developing countries, the economic impact of diarrheal disease spread through contaminated water and food is enormous.
Todd
550 L. Madagascar
Shark is frequently caught and eaten in Madagascar, but one incident of ciguatera poisoning affected over 500 people who ate a shark in Novetnber 1993 on the east coast of Madagascar (202,203). The shark, 2 meters long and weighing 100-200 kg, was caught in a net and had been sold to six wholesalers for distribution to several villages. About 200 persons were hospitalized and 98 died, many of whom went into a deep coma preceding death. This high mortality rate is not typical of ciguatera. Some shark remains were available for analysis, and two new potent heat-stable lipid-soluble toxins, carchatoxin A and B, were isolated that were the likely cause of the severe illnesses. It is possible that this was similar to another incident of shark poisoning with cardiovascular symptoms, called selachian (pertaining to sharks and rays) ciguatera on Reunion Island, close to Madagascar (204).
IX. SURVEILLANCEPROGRAMS IN THECARIBBEAN AND CENTRAL AND SOUTH AMERICA A.
Surveillance of FoodborneDisease
All Central American, South American, and Caribbean countries have some form of notifiable disease system, with diarrheal diseases being one of the main causes of death in young children; e.g., the mortality rate per 100,000 for diarrheal diseases in children less than 5 years of age in 1994 was 9.8 in Nicaragua, 0.18 in Cuba, and 0.13 in Trinidad and Tobago (205). The causes of these are not generally known, but amebic dysentery, trichinosis, giardiasis, shigellosis, brucellosis, typhoid fever, E. coli, and hepatitis infections are all documented from Latin America and the Caribbean. There was little evidence, however, to link these with specific foods. In 1996, however, a more systematic approach to surveillance and reporting of foodborne disease has been attempted by the PanAmerican Institute for Food Protection and Zoonoses and the Pan American Health Organization (INPPAZ/PAHO). Information to date is limited, although many countries are contributing to the surveillance program. Between 1995 and 1997 there was a total of 2,236 foodborne outbreaks and 68,868 cases reported (205). The range of outbreaks for each of the 19 contributing countries was from