| SSL PRAMS RE TTL PTT 7 THe TMA) PA Pe ee a ee "HEAL ; % viet’ f Aa ' ' a AiaeaRy ‘ Examination of Foods for ENTEROPATHOGENIC and INDICATOR BACTERIA wieview of Methodology sna ‘Marual of Selected Procedures /,,,, bd CAT. FOR PUBLIC HEALTH Ke S&p | Binders label ‘US. PAS, ay U. S$. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE } ~ Ee : USSD. US Public Health Service A ip Ki. pub # Won nA wots . } \ + 4.2 es Pee eT a OE OS ee SP ee ee ee ee es ee eee Pee ee ek. ee oe Ps ee ee eee Fee '' ''Examination of Foods for ENTEROPATHOGENIC and INDICATOR BACTERIA ihe of Methodology and anual of Selected Procedures Edited by KeirH H. Lewis and Ropert ANGELOTTI Robert A. Taft Sanitary Engineering Center Cincinnati, Ohio U. S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE US Public Health Service, / A: pbx H¥ a Division of Environmental Engineering and Food Protection Milk and Food Branch Washington, D. C. 20201 ''cat. FO puBLic HEALTH ''WRus Ly YMEALTH PREFACE LBRARY Members of the Milk and Food Research staff of the Public Health Serv- ice’s Robert A. Taft Sanitary Engineering Center in Cincinnati, Ohio, and the Veterinary Public Health Laboratory of the Public Health Service’s Communicable Disease Center in Atlanta, Georgia, are individually respon- sible for the chapters in Part I of this publication. Each has reviewed areas of bacterial methodology in which he is working and has a continuing interest, and together their chapters emphasize the development and current status in the United States of methods for the isolation and identification of bacteria which have special importance as “indicator organisms” or which cause “food poisoning.” Part II of this book presents a manual of laboratory methods commonly used in Public Health Service laboratories to examine for samples implicated in disease outbreaks, or collected for survey purposes. Both agar plating and liquid enrichment methods for selective isolation and enumeration of most groups are described. Each procedure was selected from the profusion of methods reviewed in Part I because experience demonstrated the use- fulness of that particular procedure for experimental work and for routine examinations. Although no procedure may be termed perfect, most listed are regarded as superior to other methods now available for purposes of the public health laboratory. Research now being conducted may permit the improvement or replacement of methods shown because of the develop- ment of techniques having greater simplicity, selectivity, sensitivity, and precision. In addition, important advances in methodology will continue to be made. Until such new developments have been proven by experience, however, the methods described in Part II are recommended as practical tools, to be used for the bacteriological examination of foods by laboratories having public health responsibilities, In no sense does this publication replace “Recommended Methods for the Microbiological Examination of Foods” published by the American Public Health Association, because the latter is wider in scope and contains more detailed information on nonpathogenic bacteria causing food spoilage. The editors do hope, however, to draw attention to methodological prob- lems now limiting the role of the laboratory, insofar as public health evalu- ation of foods is concerned, so that standard methods may eventually be agreed upon both nationally and perhaps internationally. The original request for information contained in this report came from the members of the Committee on Food Microbiology and Hygiene, International Associa- tion of Microbiological Societies, who organized an international conference 336 iii ''on microbiological standards for frozen foods at Montreal, Canada, in 1962. The minutes of that conference noted the pronounced lack of uni- formity in methods presently used by different laboratories and stressed “the need for an improved system to expedite comparative assay of, and authoritative decisions for selection of microbiological methods.” Subsequent correspondence from the committee urged the publication of our suggestions on methodology to provide a point of departure for the collaborative studies needed to evolve acceptable uniform procedures. Although our choice of methods may be controversial, we shall be amply rewarded if the selections prove helpful as reference procedures for the evaluation of other methods by individual laboratories and professional organizations in this country or abroad. Keiru H. Lewis AnD RoBERT ANGELOTTI, Editors. iv '' CONTRIBUTORS Robert Angelotti—Chief, Food Microbiology, Milk and Food Research! John R. Boring, I1J]—Assistant Chief,” Francis D, Crisley—Microbiologist, Food Microbiology, Milk and Food Research + Mildred M. Galton—Chief,? Herbert E. Hall—Microbiologist, Food Microbiology, Milk and Food Re- search ? Keith H. Lewis—Chief, Milk and Food Research } ‘U.S. Department of Health, Education, and Welfare, Public Health Service, Division of Environmental Engineering and Food Protection, Robert A. Taft Sanitary Engineer- ing Center, Cincinnati, Ohio, 45226. *Veterinary Public Health Laboratory, U.S. Department of Health, Education, and Welfare, Public Health Service, Communicable Disease Center, Epidemiological Branch, Atlanta, Georgia. '''' II meena Natta ananacennpepeaiiguieae = ACCOR AG, i teen mene si pag reese Part One: Review of Methodology I I Ii IV Vv VI VII VHT Part II II IV VI Vil VIII Ix CONTENTS Introduction—Keith H. Lewis _------------ Significance of “Total Counts” in Bacteriological Examination of Foods— Robert Angelotti -----------e Methods for Isolation and Enumeration of Coliform Organisms— Herbert: FE, Hall 2. ce Methods for Isolation and Enumeration of Enterococci—Herbert E. Hall __ Methods for Isolation of Salmonellae and Shigellae From Food Products— Mildred M. Galton and John R. Boring, IIT _--------- Methods for Isolation and Enumeration of Staphylococci—Francis_ D. Crisley ~--_____.o---- ie nnuu-_______-_________ Methods for Isolation and Enumeration of Clostridium Perfringens— Robert Angelotti —----------- Demonstration of Botulinus Toxins and Clostridium Botulinum in Foods— Francis D. Crisley ~--------_---- Two: Manual of Laboratory Methods Introduction ____--------- Preparation of Food Homogenate ____._-...---_- Microscopic Examination __-___------- Agar Plate Colony Count —--------- Coliform Bacteria _----___-_---e Most probable number method ____--------- Coliform organisms of fecal origin _-------- Fecal Streptococci —-----_----- Most probable number method __------------ Staphylococci ------_-__-----ee Most probable number method __----------- Salmonellae-Shigellae _-------- Enrichment method ~------------ Enumeration of salmonellae in food specimens ________________ Alternate method _---------- Clostridium Perfringens ---_--------- Plate count method _---_------ Enrichment method ~---_-__------- Clostridium Botulinum ___-----_-----8 13 20 26 33 50 55 81 84. 85 87 88 88 89 90 94 '' '' PART ONE: REVIEW OF METHODOLOGY I. Introduction Keith H. Lewis Selection of a bacteriological method for the determination of an actual or potential health hazard of food is, at best, a very difficult problem for which no perfect solution is likely to be found in the near future. The available techniques are largely of two kinds, those adapted from the diagnostic laboratory to detect specific pathogens in foods implicated in disease outbreaks, and those intended to serve as indicators of sanitary quality when used in conjunction with inspection programs, The first group often depends on primary isolation in a selective environ- ment that is less inhibitory to the pathogen than to other organisms in the food. Procedures that are adequate for the qualitative detection of an infectious agent in high concentrations may be unreliable for the quantitative determination of that species, especially at low levels or in the presence of other, more abundant microorganisms. Any one or a combination of any of the following factors may affect adversely the re- sults obtained from routine examination of foods: 1. Presence of food constituents that interfere with the selective prop- erties of the system; 2. Presence of microorganisms that overgrow, antagonize, or sim- ulate the appearance of the pathogen; 3. Failure of certain cells to initiate growth, especially when injured by such factors as freezing, chemical preservatives, heating or drying; 4. Sensitivity of some strains to the selective agent intended to facili- tate their isolation. Experience has shown that isolation procedures must be chosen carefully and modified to meet the conditions encountered in different foods. For example, Tellurite-Glycine medium was developed especially for the selective isolation of coagulase-positive Staphylococcus aureus from meat products (Zebovitz, el al., 1955) (1), but when applied to dairy products it proved inferior to Staphylococcus-110 agar foredetection of this organism in low concentrations (Donnelly, et al., 1962) (2). Numerous modifications have also been found necessary in the procedures for isolation of salmonellae ''Review of Methodology from different foods (See Part One, Ch. V). They range from pre-incuba- tion in nonselective media to addition of a wetting agent and even to new formulations of selective media (Galton, Lowery and Hardy, 1954 (3) ; Hajna and Damon, 1956 (4); and North, 1961) (5). Unfortunately, there is considerable disagreement among laboratory workers on the relative merits of these modifications. In most instances, the available enrichment and isolation procedures are not sufficiently selective to permit identification of a particular species without confirmatory evidence. Additional methods are usually required for characterization of the cultures isolated from selective media. Fre- quently these procedures are so time-consuming that the food may have been consumed before the results are available and therefore, their value is nullified. These procedures are costly to perform and tend to interfere with the normal flow of trade. Major improvements in the specificity, sensitivity, and rapidity of procedures for detection of pathogens in foods are needed to permit the establishment of precise standards for general use. For products that present a demonstrable public health hazard, the methods may allow a qualitative determination that certain pathogens are or are not demonstrable in a given amount of sample. Proof of the total absence of pathogens or their quantitative enumeration are gen- erally beyond the practical capabilities of control laboratories. Centralized processing and widespread distribution of commercially prepared foods has led receiving areas to consider the use of microbio- logical standards as indicators of quality for products produced outside their jurisdictions (Abrahamson, et al., 1959 (6); Massachusetts Depart- ment of Public Health, 1959) (7). Under these circumstances, the relation of plate counts, numbers of coliform bacteria, or other indicator organisms to the sanitary conditions in the factory cannot be verified by inspection. So many interacting factors affect the bacterial content of the finished prod- uct that interpretation of the bacteriological data is, at best, very difficult in the absence of a detailed history of the product. High counts may indicate inferior raw materials, unsanitary processing, normal flora of wholesome foods, or exposure to growth temperatures during storage or transportation, as well as other, less obvious possibilities. Bacteria not associated with health hazards may grow in certain foods, while organisms of human or animal origin may be destroyed in other products by freezing, drying, warming, or other procedures that do not eliminate all pathogens. On the basis of problems encountered in the evaluation of bacteriological data obtained from the examination of dairy products (Donnelly, et al., 1960) (8) and of shellfish (Kelly, 1960) (9), it is evident that much de- tailed information must be collected and analyzed before meaningful stand- ards can be evolved for any kind of food. In each instance, however, a be- ginning was made by agreement among experienced microbiologists on 2 '' Introduction tentative guides to sampling, laboratory examination, and indices of good practice, These steps were taken in the face of imperfect methods and limited knowledge about the relation of the results to public health protection. Nevertheless, these indices have proved helpful to industry in improving the quality of its products, and to the consumer by reducing the risk of contrac- ting illness from foods prepared in compliance with them. Undoubtedly, similar benefits can be achieved for other segments of the food industry and the general public, provided the necessary effort and cooperation are focused on specific products that have the potential for disease trans- mission, This review does not consider in detail problems of collecting, transporting, or storing food specimens intended for bacteriological examination, Apart from the general requirement for aseptic technique, methods of collection may differ considerably with the kind of food, batch size, method of pres- ervation, type of container, and especially with the need for obtaining sta- tistically representative samples. Once a specimen has been placed in a sterilized container, it should, of course, be treated so that the bacterial content does not change. Ideally, the food should be examined within an hour ‘after sampling, but, where longer periods are unavoidable, prompt refrigeration is essential. Freezing tends to injure or kill many of the pathogenic bacteria and is to be avoided if the specimens can be kept cold by less drastic means such as packing the containers in crushed ice. It is axiomatic that the best laboratory procedures cannot compensate for de- fects in sampling. Full discussion of this important topic might well be the subject of a separate review. Part I of this report reviews the methodology pertaining to the more common “indicator” and “food-poisoning” bacteria. It includes chapters on plate counts, coliform organisms, fecal streptococci, salmonellae, shigellae, Clostridium botulinum, and Clostridium perfringens. Part II presents a manual of bacteriological methods currently used by the Public Health Service to examine foods implicated in disease outbreaks or col- lected for survey purposes. Wherever practical, quantification has been emphasized. The choice of methods is based on our own experience and recently published data from United States and Canadian sources. Undoubt- edly, much additional information on methodology of interest to food micro- biologists could be derived from a more comprehensive review of the world’s literature. REFERENCES (1) Zebovitz, E., Evans, J. B., and Niven, C. G. 1955. Tellurite-glycine agar: a selective plating medium for the quantitative detection of coagulase-positive staphylococci. J. Bacteriol. 70: 686-690. 3 ''Review of Methodology (2) Donnelly, C. B., Black, L. A., and Lewis, K. H. 1962. The occurrence of coagulase-positive staphylococci in cheddar cheese—a preliminary report. Bacteriol. Proc. p. 23. (3) Galton, M. M., Lowery, W. D., and Hardy, A. V. 1954. Salmonella in fresh and smoked pork sausage. J. Infectious Diseases 94: 232-235. (4) Hajna, A. A. and Damon, S. R. 1956. New enrichment and plating media for the isolation of Salmonella and Shigella organisms. Appl. Microbiol. 4: 391- 395. (5) North, W. R. 1961. Lactose pre-enrichment method for isolation of Salmonella from dried egg albumin. Appl. Microbiol. 9: 188-195. (6) Abrahamson, A. E., Buchbinder, L., Guenkel, J., and Hiller, M. 1959. A study of frozen precooked foods: Their sanitary quality and microbiological standards for control. Assoc. Food and Drug Officials of the United States, 23: 63-72. (7) Massachusetts Department of Public Health, Bureau of Consumer Products Pro- tection, Division of Food and Drugs, Commonwealth of Massachusetts, Acts of 1959. Rules and regulations relative to storage and distribution of frozen foods. (8) Donnelly, C. B., Harris, E. K., Black, L. A., and Lewis, K. H. 1960. Statistical analysis of standard plate counts of milk samples split with state laboratories. J. Milk and Food Technol. 23: 315-319. (9) Kelly, C. B. 1960. Bacteriological criteria for market oysters. Technical Report F60-2, Robert A. Taft Sanitary Engineering Center, Cincinnati, Ohio. '' II. Significance of ‘‘Total Counts” in Bacteriological Examination of Foods Robert Angelotti As generally employed, total counts are intended to indicate the level of bacterial contamination in a product. Because the growth response of bac- teria is affected by environmental factors, such as incubation time and temperature, type of nutrients present, the availability or absence of oxygen, competitive association among organisms, etc., the number and types of bacteria that develop are determined to a large extent by the counting method employed. Nevertheless, there appears to be an inverse relation- ship, at least in the case of perishable foods, between the level of contam- ination and shelf life. Carelessness in avoiding contamination, or exposure to time-temperature conditions conducive to growth, usually results in higher bacterial counts than those obtained in foods handled under sanitary conditions, Perishable foods with high counts undergo spoilage more readily. Because high counts reflect, to a certain degree, inadequate sanitation prac- tices, it is felt that the same situations that permit entry or development of large numbers of bacteria could conceivably allow entry and development of pathogenic types as well. Thus, counts of foods are usually employed for the following purposes: to evaluate general sanitation practices prevail- ing during handling; to reasonably predict shelf life; and, to a certain degree, to evaluate safety. Unfortunately, it is generally not possible to obtain meaningful informa- tion on these three objectives by means of a single total-count procedure applicable to a variety of foods. Consequently, the total-count method chosen is often a reflection of the need for specific information concerning one of these objectives and results in the use of methods tailored to the specific need. Because of this practice, variation in methodology related to total-count procedures for foods is common. Some standardization of total count procedures has, however, been developed for a limited number of food items. The greatest strides toward development of standardized bacteriological procedures have been made with milk and dairy products. Through the efforts of the American Public Health Association and the publication of eleven editions of “Standard Methods for the Examination of Dairy Products” (APHA, 1960”) (7), dairy products are examined by uniform procedures and held to common standards. The total aerobic count is obtained for all dairy products, with the exception of dry milk and eggs used in frozen desserts, by means of the “standard plate count” procedure, 5 ''Review of Methodology which calls for the use of Standard Methods agar and incubation at 32° (+ 0.5 to 1.0°) C, or 35° (+ 0.5 to 1.0°) C, for 48 (+ 3) hours. The medium and incubation temperatures are the same for dry milk and eggs; the recommended time of incubation is, however, extended to 72 hours. More recently, efforts on the part of the American Public Health Asso- ciation have been directed toward developing standard methods for the microbiological examination of foods other than dairy products (APHA, 1946 (2) and 1958 (3)). “Recommended Methods for the Microbiological Examination of Foods” (APHA, 1958) (3), an association publication has centered a great deal of attention on methods useful in food analysis. The procedure recommended for total aerobic counts on most foods calls for the use of Total Plate Count agar (two formulations: tryptone, glucose, yeast extract agar; or milk protein hydrolysate, glucose agar) and incu- bation at 32° C for 3 days. Certain classes of foods, such as pickled veg- etables, sugars, syrups, dried fruits and vegetables, and fresh meat, require, however, the use of specific media or incubation conditions that differ from those described above. These modifications are necessary to permit growth of the predominant flora associated with these products. The most recent publication of the American Public Health Association relative to bacteriological procedures for analysis of foods is the “Rec- ommended Procedures for the Bacteriological Examination of Sea Water and Shellfish” (APHA, 1962) (4). This document recommends that a standard plate count of shellfish be obtained through the use of Standard Methods agar with incubation at 35° (+ 0.5°) C for 48 (+ 3) hours. The procedures for total counts outlined above in the three recommenda- tions differ in the recommended temperature for incubation (32 or 35° C for dairy products, 32° C for most foods, and 35° C for shellfish) ; length of incubation (48 hours for dairy products and shellfish versus 72 for most other foods) ; and in the choice of medium (digest of casein, yeast extract, glucose agar for dairy products and shellfish versus tryptone, yeast extract, glucose agar, or milk protein hydrolysate, glucose agar for most other foods). This apparent trend toward development of minor variations in total-count procedures for specific food items may not be significant, for the “Recommended Methods for the Microbiological Examination of Foods” is presently undergoing revision that may result in the selection of a total plate count method comparable to that for dairy products and shellfish. Here again, however, the question arises as to the purpose of a total count in foods. Are total counts to be performed to reflect keeping quality, safety, or sanitation history? The choice of medium and conditions of incubation depends upon which of these objectives is sought. The public health worker is primarily oriented toward evaluating the sanitation history and safety of a food. He is prone to use a technique that permits the growth of pathogenic bacteria or closely related types, if they are present. 6 '' Total Counts In an attempt to gain some insight concerning the relationship of total counts to sanitary condition and safety of foods, a review of published papers was undertaken that dealt with the microbiological quality of frozen foods. Frozen foods were selected for review because of the availability of recently published papers describing their microbiological quality and be- cause as a group, they contain in one preparation or another almost every major food item in the diet. Papers published in the United States recently revealed an unusual una- nimity of choice of method for obtaining “total counts” of frozen foods. Direct plating of serial dilutions of food in Tryptone Glucose Extract agar (TGE agar) (American Public Health Association, 1960*) (5) followed by aerobic incubation at 35 to 37° C for 48 hours is reported most commonly. This conformity apparently is related to the original suggestion by Fitz- gerald (1947) (6) that a total of 100,000 organisms per gram be set for frozen foods and to the application of this standard by the Quartermaster Food and Container Institute in its specifications for precooked frozen meals (Rayman, et al., 1955) (7). Subsequent investigators have centered their attention on the microbiological quality of frozen foods in relation to this standard. Litsky and coworkers (1957) (8) conducted a bacteriological survey of commercially frozen tuna, chicken, turkey, and beef pies on a nationwide basis. These foods were examined for total count, as well as coliform, enterococcal, and salmonellae types of organisms. Total count was per- formed by plating in TGE agar and incubating at 37° C for 48 hours, Of the 132 foods examined, 84. percent yielded total counts of less than 100,000 per gram. Sixty percent yielded counts of less than 25,000 per gram. Cor- relations between coliform count, enterococcus, most probable number (MPN), and total count were poor in all four groups of foods. Canale-Parola and Ordal (1957) (9) examined frozen chicken and turkey pies produced by five different manufacturers. Total count determinations were made in TGE agar plates incubated at 37° C for 48 hours. Coliform, enterococcus, coagulase-positive staphylococci, and salmonellae-para- colon MPN counts were also performed. Of the 40 unbaked pies that were tested, 20 had total counts above 100,000 per gram and 18 had coliform counts above 10 per gram. Enterococci were present in all of the samples, and coagulase-positive staphylococci were detected in 37 of the 40 pies tested. Five salmonellae cultures were isolated from the pies. These data may suggest there is a correlation between high total count and presence of detectable levels of staphylococci and salmonellae in frozen foods. Ross and Thatcher (1958) (10) conducted a survey on 117 meat, fish, poultry, and dessert frozen pies representing products of nine manufacturers in both the United States and Canada. Total counts were made in TGE agar, followed by incubation at 37° C for 48 hours. Determinations were a ''Review of Methodology also made of E. coli, enterococci, coagulase-positive staphylococci, and sal- monellae. Their data revealed that 75 percent of all the samples conformed to the following numerical values: total count, 50,000 per gram; E. coli, 15 per gram; enterococci, 256 per gram, and staphylococci, 1,000 per gram. Salmonellae-type organisms were not confirmed in any of the 117 samples. Foods that contained the fewest number of organisms and were categor- ized in the first quartile conformed to the following values: total count, 50 per gram; E. coli, 0 per gram; enterococci, 0 per gram, and staphylococci, 0 per gram, These data reveal an excellent correlation between low total count and absence of indicators of fecal pollution and salmonellae and staphylococci. Kereluk and Gunderson (1959) (11) sampled 188 frozen meat pies pro- cured at the retail level and examined them for total count, coliform bac- teria, enterococci, staphylococci, and salmonellae-paracolon bacteria. Total counts were made in TGE agar, but the incubation period and temperature were not specified. It is assumed that common practice was followed (37° C for 48 hours). Of the 188 pies, 75 percent gave total counts of 25,000 per gram or less, which is in agreement with findings of Ross and Thatcher (10). About 83 percent had total counts under 50,000 per gram and only 14 pies (7 percent) had total counts of 100,000 per gram, or more. These authors did not attempt to categorize their findings in terms of relationship of total count to presence of pathogens, but their data indicate that poor correlation exists between total count and staphylococcal count. In almost every sample of turkey, beef, and tuna pies in which a staphylococcal count in the range of 100 to 10,000 or more per gram was obtained, a total count of 25,000, or less, was observed. Oddly, good correlation between the two was recorded for chicken pies. Shelton and coworkers (1958-1959) (12), of the Food and Drug Ad- ministration, conducted a bacteriological survey of the frozen, precooked food industry during the period from March 1958 to June 1959, in which 63 frozen food plants in 18 States were inspected. Some 3,000 samples, repre- senting 81 food items, were collected and examined. Determinations of total aerobic count, coliform most probable number (MPN), E. coli MPN, and staphylococci were performed. Total counts were obtained in Tryptone Glucose Yeast Extract agar (American Public Health Association, 1960*) (5) incubated at 35° C for 48 hours. The results of this survey indicated, generally, that the sanitary conditions and operating practice in the plants were considerably below the level desired for a nonsterile food product that does not receive a final heat process and that may be consumed without sufficient heating or cooking to destroy microorganisms. In general, the total count and coliform and E, coli counts were well correlated. The total counts and coliform counts were found to be related to equipment cleanli- ness and time-temperature holding periods of the foods, whereas the E. coli 8 '' Total Counts counts were related principally to contamination by food handlers. In products containing raw natural cheeses or uncooked eggs, however, coli- form organisms, including E. coli, occurred frequently. Little correlation was noted between staphylococcal counts and total counts, though the find- ing of coagulase-positive staphylococci was usually related to employee con- tamination of the product during an operation requiring a great deal of handling. Nickerson, et al. (1962) (13), performed microbial analyses of 78 samples of frozen fish sticks for total count, coliform and coagulase-positive sta- phylococci counts, and presence of salmonellae-shigellae type organisms. Total counts were made in plate count agar (American Public Health Asso- ciation, 1960") (5) incubation at 35° C for 24 hours. The majority of fish sticks examined were of acceptable bacteriological quality by the standards of the Commonwealth of Massachusetts (total plate count: 50,000 organisms per gram, 10 coliform organisms per gram). A total of 20 samples distrib- uted among 12 processors did not, however, meet this sanitary standard. Of fish sticks obtained from 5 producers, which yielded coliform counts of more than 10 per gram, only products of two producers had total counts of more than 50,000 per gram. In the majority of samples, a good correlation existed between low total count and counts of less than 10 coliform organ- isms per gram. The data presented in the references cited above indicate that precooked frozen foods are being manufactured with total counts of considerably less than 100,000 organisms per gram. Very little published data is, however, available that reports on the effects of media, incubation time and temper- ature, blending, diluents, etc., on the results of total counts. Zaborowski, et al. (1958) (14), performed a limited evaluation of some of these factors. They were specifically interested in effects on total counts of incubation temperature and blending versus shaking of food samples. The tentative methods for the microbiological examination of frozen foods, published by the American Public Health Association (1946) (2), recommended incuba- tion of TGE agar plates for 4 days at 35° C. Because minimal incubation times are generally desired, comparisons of plate counts as influenced by times and temperatures were made, using TGE agar. Plate counts were made on 114 components of precooked frozen meals (19 meats and 95 vege- tables) to compare the effect of 48- and 72-hour incubation periods. The majority of the counts obtained with vegetables (92%) were similar at 48 and 72 hours, whereas 11, or 60 percent, of the meat samples gave higher counts after 72 hours than after 48 hours. Comparisons of incubation at 32 and 35° C revealed that, from a total of 94 samples of meats and vegetables, there was little difference between counts incubated at 32 or 35° C for 72 hours. A “difference” was arbitrarily set as an average colony count of 9 ''Review of Methodology five or more. Results of experiments comparing shaking versus blending of dilutions revealed that blending yielded higher counts. At present, only New York City and the Commonwealth of Massachusetts have applied bacteriological standards to frozen foods, Both contain a numerical limit for a total count. The former states: “An acceptable quality of frozen precooked foods are: a total plate count of less than 100,000 colonies per gram and the absence of Staphylococcus aureus.” (Abraham- son, et al., 1959) (15). The latter standard states: “Precooked or partially cooked frozen foods shall not have a bacteria count in excess of the follow- ing: Standard plate count—50,000 colonies per gram; Coliform plate count —10 colonies per gram.” (Massachusetts Department of Health, 1959) (16). Both of these standards are applied in terms of a standardized methodology proposed as adequate by the agencies employing them, but, in view of the experience of those attempting to evaluate and certify media and methods for use in analyzing dairy products, it is interesting that so little considera- tion has gone into the choice of medium and incubation time and tempera- ture for total counts of frozen foods. It is common experience that media vary among manufacturers and lots in their ability to support microbial growth, and that incubation temperature has a profound effect on numbers and types of organisms recovered. In spite of the present recommendation of the APHA for use of Total Plate Count agar and incubation at 35° C for 72 hours for total counts, there appears to be a number of points that should be clarified before a method is considered adequate for food analysis. These may be listed as: (a) medium and substantiation for its use, including some control to insure minimum variation among manufacturers and lots; (b) incubation time and tempera- ture comparative evaluations among cooperating agency laboratories to determine those conditions yielding maximum recovery with minimum variability; (c) sampling method studies related to sample collection and handling, including such factors as frozen versus thawed sampling, size of sample, composite or single item sampling; (d) type of blending—speed, duration, and volume: (e) diluents. In relation to these basic questions, the recent publications of Hartman and Huntsberger (1960 (17) and 1961 (18) ) are of interest, for they specifically attempted to measure the significance of variables associated with total plate counts of frozen foods. Their basic methodology consisted of sampling 100 grams of a representative portion of the contents of frozen pies and plating in Trypticase Soy agar at 32° C for 2 days. They observed that total counts were higher during the warm months than during the cool months; seasonal influences were. however, minimized when very low counts were constantly maintained. Samples ob- tained on Mondays and on latter days of the work week were generally higher in counts than samples obtained on Tuesdays and Wednesdays. Sampling at hourly intervals during processing was necessary to detect 10 '' Total Counts the majority of high-count products. Bacterial counts obtained on one prod- uct were not necessarily indicative of the level of contamination to be ex- pected in another product prepared in the same plant. In their second study (1961) (18), they reported on the influence of the following 11 variables on plate count values: (1) sample size, (2) size of initial dilution, (3) addi- tion of diluent, (4) duration of homogenization, (5) holding in minutes af- ter homogenization in a liter container, (6) holding in minutes after homog- enization in a gallon container, (7) initial diluent, (8) subsequent diluent, (9) shaking, (10) type of blender unit, (11) condition of blender. Of the 11 variables, 3 contributed significantly to the results obtained. The size of the initial dilution was important, and evidence indicated that initial dilu- tions other than 1:5 or 1:10 led to uncontrollable variation in results. These two initial dilutions yielded equally satisfactory results. The type of diluent also was shown to affect results, and it was noted that phosphate-buffered dilution water (American Public Health Association, 1960”) (7) was satis- factory as an initial diluent and in the preparation of subsequent dilutions. The third variable of importance, the degree of shaking received by serial dilution blanks of food homogenates, revealed less variation in results when the blanks were shaken 5 times through a 1-foot arc than when shaken 25 times through the same arc. A less vigorous shaking of dilution blanks subsequent to mechanical homogenization is indicated, because of the ex- tensive breaking up of most clumps during homogenization. These studies point up the need for the establishment of a standardized methodology, and additional studies by others directed specifically toward choice of medium and incubation time and temperature are indicated. REFERENCES (1) American Public Health Association. 1960". Standard methods for the ex- amination of dairy products, lith Ed., p. 60. Am. Public Health Assoc., Inc., New York 19, N. Y. (2) American Public Health Association. 1946. Report of the Committee on Standard Methods for the | microbiological examination of foods: tentative methods for the microbiological examination of frozen foods. Am. J. Public Health, 36: 332-335. (3) American Public Health Association. 1958. Recommended methods for the microbiological examination of foods. Am. Public Health Assoc., Inc., New York 19, N. Y. (4) American Public Health Association. 1962. Recommended Procedures for the Bacteriological Examination of Sea Water and Shellfish, 3rd ed., pp. 22 and 32. Am. Public Health Assoc., Inc., New York 19, N. Y. (5) American Public Health Association. 1960." Standard methods for the examina- tion of water and wastewater, 11th Ed., p. 487. Am. Public Health Assoc., Inc., New York 19, N. Y. (6) Fitzgerald, G. A. 1947. How to control the quality of frozen cooked foods. Food Industry, 19: 623-625. Il ''Review of Methodology (7) Rayman, M. M., Huber, D. A., and Zaborowski, H. 1955. Current micro- biological standards of quality for precooked foods and their basis. Precooked Frozen Foods—A Symposium. Advisory Board on Quartermaster Research and Development Committee on Foods, National Academy of Science, National Research Council, Washington, D. C., 55-67. (8) Litsky, W., Fagerson, I. S., and Fellers, C. R. 1957. A bacteriological survey of commercially frozen beef, poultry, and tuna pies. J. Milk and Food Technol. 20: 216-219. (9) Canale-Parola, E. and Ordal, Z. J. 1957. A survey of the bacteriological quality of frozen poultry pies. Food Technol. 11: 578-582. (10) Ross, A. D. and Thatcher, F. S. 1958. Bacteriological content of marketed pre- cooked frozen foods in relation to public health. Food Technol. 12; 369-371. (11) Kereluk, K. and Gunderson, M. F. 1959. Studies on the bacteriological quality of frozen meat pies. I. Bacteriological survey of some commercially frozen meat pies. Appl. Microbiol. 7: 320-323. (12) Shelton, L. R., Leininger, V. H., Surkiewicz, B. F., Baer, E. F., Elliot, R. P., Hyndman, J. B., and Kramer, N. 1958-1959. A bacteriological survey of the frozen precooked food industry, U. S. Department of Health, Education, and Welfare, Food and Drug Administration, U. S. Government Printing Office Pub. No. 905826, Washington 25, D. C. (13) Nickerson, J. T. R., Silverman, G. J., Salberg, M., Duncan, D. W., and Joselow, M. M. 1962. Microbial analysis of commercial frozen fish sticks. J. Milk and Food Technol. 25: 45-47. (14) Zaborowski, H., Huber, D. A., and Rayman, M. M. 1958. Evaluation of microbiol. methods used for the examination of precooked frozen foods. Appl. Microbiol. 6: 97-104. (15) Abrahamson, A. E., Buchbinder, L., Guenkel, J., and Hiller, M. 1959. A study of frozen precooked foods: Their sanitary quality and microbiological standards for control. Assoc. Food and Drug Officials of United States, 23: 63-72. (16) Massachusetts Department of Public Health, Bureau of Consumer Products Pro- tection, Division of Food and Drugs, Commonwealth of Massachusetts, Acts of 1959. Rules and regulations relative to the storage and distribution of frozen foods. (17) Hartman, P. A. and Huntsberger, D. V. 1960. Sampling procedures for bacterial analysis of prepared frozen foods. Appl. Microbiol. 8: 382-386. (18) Hartman, P. A. and Huntsberger, D. V. 1961. Influence of subtle differences in plating procedure on bacterial counts of prepared frozen foods. Appl. Microbiol. 9: 32-38. 12 '' Ill. Methods for Isolation and Enumeration of Coliform Organisms Herbert E. Hall Of the many media developed for the isolation and enumeration of coli- form organisms from water and sewage, only a few have found practical application in the examination of foods. The simplest, least selective, and least inhibitory medium is lactose broth (American Public Health Asso- ciation, 1960) (J). Because it does lack inhibitory or stimulatory sub- stances, gas production in this medium may be markedly influenced by the composition of the food added with the inoculum. Workers who have used it in obtaining a presumptive MPN (Most Probable Number) have had the experience of Larken, Litsky, and Fuller (1955) (2) that the number of con- firmed coliform organisms is very low in comparison to the presumptive MPN. Numerous studies indicate that Lauryl Sulphate Tryptose broth (Mallman and Darby, 1941) (3) is much better adapted to use as a presumptive medium with food (Mundt, et al., 1954 (4); Raj and Liston, 1961 (5); Fanelli and Ayres, 1959 (6); Wilkerson, et al., 1961 (7); Kelly, 1960 (8) ; and Shelton, et al., 1961) (9) than is plain lactose broth. Confirmation may be made in Brilliant Green Lactose, 2 percent Bile broth, or by streaking on Eosin Methylene Blue agar (EMB). The proportion of confirmed tubes is much higher with this medium, and the effect of added food constituents appears to be much less. Mundt, et al. (1954) (4), however, found that the carbohydrate from strawberries affected the rate of gas production, although it did not lead to a significant number of false-positive reactions. This medium has been used by the Public Health Service for the examination of oysters (Kelly, 1960) (8) and by the Food and Drug Administration (Shelton, e¢ al., 1961) (9) in the examination of a very broad sampling of frozen, precooked foods. This medium also serves well as the preliminary growth medium for the isolation and enumeration of EF. coli Types I and I (Table 1). Because of their definite association with human fecal material, the identification of these organisms has a real place in the study of some types of foods. In the examination of oysters, Kelly (1960) (8) recommends the inoculation of the modified Eijkman medium of Hajna and Perry (1943) (10) (EC Medium), from the gas-positive Lauryl Sulfate Tryptose tubes, incubation at 44.5° C+0.2° C; and confirmation as £. coli on EMB and by IMViC pattern de- terminations. Similar identification and confirmation procedures were carried 13 ''Review of Methodology out by Shelton, et al. (1961) (9), on frozen foods. Raj and Liston (1961) (5), in a study of frozen sea foods, found that only 33 percent of the EC- positive tubes were confirmed as E£. coli Types I and I. Kelly (1960) (8) observed that a reduction in temperature of as little as 0.5° C caused a signi- ficant rise in false-positive results. Recent unpublished work from Kelly’s laboratory (Presnell, 1961) (11) on the use of the EC test with oysters showed that 53 to 84 percent of the EC-positive tubes were confirmed as E. coli, Type I, when the incubation temperature was 44.5° C, The varia- tion noted in confirmation depended upon the degree of contamination with coliform organisms of types other than Type I, such as A. aerogenes, Type I; E. coli, Type II; and intermediate Types I and IH. Attempts to increase the specificity of the test by raising the incubation temperature were success- ful in that all EC-positive tubes confirmed as E. coli, Type I, at 45 or 46° C, but false negatives were obtained in from 8 to 25 percent of the EC tubes, depending upon the incubation temperature and the nature of the specimen tested. This emphasizes the critical nature of the temperature of incubation in the use of the EC medium to determine levels of FE. coli, and appears to be applicable also to the use of the Boric Acid medium of Vaughn and Levine (1935) (12) for the same purpose. Although Walford and Berry (1948) (13) and Beisel and Troy (1949) (14) found this medium to give good differentiation of E. coli from other coliforms in the study of orange juices, both the critical incubation temperature requirements (Lavine, et al., 1955) (15) and the effect upon its efficiency of the source of the or- ganisms (Clark, et al., 1957) (16) make it less desirable than the EC medium. It is obvious from the work so far done that it is not safe to de- pend upon the results of EC gas positives to indicate E. coli, but that careful confirmation studies must accompany these findings. In recent years many workers have used direct plating techniques for the determination of coliform densities in foods. Of the various media available, two, Violet Red Bile agar (VRB) and Desoxycholate Lactose agar, have been employed most frequently. Studies by Litsky, et al. (1957) (17); Ross and Thatcher (1957) (18); Canale-Parola and Ordal (1957) (19); Kachikian, et al. (1959) (20); Fanelli and Ayers (1959) (6); Hartman (1960) (22) ; Wilkerson, et al. (1961) (7); and Machala (1961) (22), among others, in- dicate that Violet Red Bile agar may be successfully used to isolate coliform organisms from foods. The studies of Huber, e¢ al. (1958) (23); Kereluk and Gunderson (1959) (24); Silverman, et al. (1961) (25); and Nicker- son, et al. (1962) (26), indicate Desoxycholate Lactose agar may be used with equal success. In reviewing the work of these and other authors, a number of interesting facts concerning the quantitative aspects of recovery of coliform organisms with these solid media come to light. Many workers appear to have accepted the presence of red colonies on these media as a 14 '' Coliform Organisms reliable indication of coliform organisms, although there is considerable evidence to the contrary. Ross and Thatcher (1957) (18), using VRB, found that if care were taken to record only typical red colonies surrounded by a precipitate of bile and to adhere to a 24-hour incubation period, confirmation by means of the IMViC reactions was very good. Other organisms (Flavo-bacterium, Proteus, Aerobacter, and Pseudomonas) produced atypical colonies. Furthermore, Gunderson and Rose (1948) (27) showed that VRB would recover no more than 60 percent of the viable coliform organisms from foods and that after frozen storage for 9 or more days the percentage of recovery was as low as 12 to 25 percent. This is confirmed to some extent by Wilkerson, et al. (1961) (7), who found that better recovery levels were obtained in Lauryl Tryptose broth than in VRB from stored poultry samples. In addition, Hartman (1960) (21) found that the percentage of confirmed coliform bacteria varied markedly with the food examined. With cream pies, 20 percent confirmation was obtained; while from frozen chicken, turkey, or beef pies, about 70 percent confirmation was obtained. This author also studied the effect of autoclaving VRB and found that heat sterilization had little or no effect upon the specificity (Hartman, 1958) (28). In either case, Proteus, intermediates, Alcaligenes, and Achromobacter species were able to produce red colonies. In most instances, these were small or “atypical” colonies, but in the 1960 study he reported that many colonies of this type were confirmed as coliform organisms and that an appreciable number of typical large (1.0 mm) colonies failed to be confirmed. In contrast to this, Fanelli and Ayres (1959) (6), and Wilkerson, et al. (1961) (7), obtained higher coliform counts with VRB than with Lactose broth or Lauryl Sulfate broth MPN techniques, and in these studies careful confirmation work was carried out. Kereluk and Gunderson (1959) (24) in preliminary studies found that VRB and Desoxycholate Lactose agar gave comparable results. In their ex- amination of frozen meat pies they found MPN determinations using Lactose broth gave higher values, but that the use of Desoxycholate Lactose was so much faster and less cumbersome that its use appeared to be the method of choice. Silverman, et al. (1961) (25), and Nickerson, et al. (1962) (26), found that Desoxycholate Lactose agar produced excellent results in the ex- amination of frozen raw and cooked shrimp and fish sticks. They experi- enced the same difficulty described by Hartman (1960) (21) in correlating levels obtained from 1 milliliter of a 1:10 dilution in small plates with levels from 10 milliliters of this dilution in large plates. It has been our own experience that excellent recovery of coliform organ- isms can be obtained with VRB agar. In a series of examinations of labo- ratory prepared meat pie fillings (beef or chicken with potatoes, carrots, and peas in gravy) inoculated with known numbers of E, coli strains iso- 15 ''Review of Methodology lated from feces, quantitative recovery in a range from approximately 10 to several million organisms per gram of food has been obtained. It would appear from these findings that the use of Bile Salts No. 3 (Difco) or Bile Salts Mixture (BBL) in place of the original Bile Salts has improved the medium so that the deficiencies noted by Gunderson and Rose (1948) (27) are no longer applicable. The only interfering organisms of consequence that we have encountered, to date, are members of the genus Proteus. When P. vulgaris or P. mirabilis are present in appreciable numbers, they interfere with the coliform counts in two ways. They form small red atypi- cal colonies, and they appear to inhibit the precipitation of bile by coliform organisms concomitantly present. This results in too high counts if all red colonies are counted and too low counts if only typical colonies are counted. Furthermore, we have found that when pure cultures of E. coli, Types I or II, are added to food and plated in VRB, both large and small colonies will arise, which would appear to confirm Hartman’s (1960) (21) findings and indicate that Ross and Thatcher (1957) (18), by their very care in selecting only typical colonies, were eliminating from their counts some coliform organisms. In confirmation of the work of Kereluk and Gunderson (1959) (24), we have found that in addition to E. coli I and II, Aerobacter J and II and E. freundii may produce typical colonies on VRB. We have found that the Paracolobactrum, Alcaligenes, and Erwinia species tested did not grow in VRB, while Pseudomonas and Serratia species did grow but produced white or colorless colonies. On the whole, our experience with VRB has been in accord with that of other workers who have employed the medium suc- cessfully. Confirmation in Brilliant Green Lactose 2 percent Bile broth readily eliminates the few species of noncoliform organisms that may grow out and produce red colonies. The use of a plating medium such as VRB appears to give highly reproducible results, and quantitative isolation of ten or more organisms per gram appears possible. A greater degree of variability has occurred when liquid media were used in place of VRB agar. From the above publications, it is evident that the workers in Canada and the United States have selected two major avenues of approach to the isolation and enumeration of coliform organisms from foods. Those who prefer the determination of an MPN appear to favor the use of a somewhat selective medium, such as the Lauryl Sulfate Tryptose broth of Mallman and Darby (1941) (3); while others prefer the less cumbersome pour plate methods using Violet Red Bile or Desoxycholate Lactose agar. In either case, accurate results require confirmation in more selective or differential media. Whichever of these approaches is used, it appears that satisfactory results can be obtained with most foods, but it also appears questionable whether direct comparisons of results obtained by the two methods can be made. That one is more accurate than the other is doubtful, but the 16 '' Coliform Organisms inaccuracies of each are different and in some respects inherent in the methods In those cases where it is particularly desirable to know not only the coli- form density, but actual number of E. coli specifically, additional tests have to be made. The findings to date are in agreement that the use of the modified Eijkman method with either the EC medium or Boric Acid broth is unsatisfactory without cultural confirmation on the bases of the IMViC reactions, but it does appear that the number of confirmations can be greatly decreased by the use of this technique. The use of two con- firmatory media, EC broth for E. Coli, and Brilliant Green Lactose 2 percent Bile broth for coliforms, as suggested by Kelly (1960) (8) and Shelton, et al. (1961) (9), allows the worker to obtain an indication of the relative levels of contamination by the two groups. If the E. coli confirmation is to have significant meaning, strict attention must be paid to the incubation temperature, and the fact that some strains of E. coli will not grow quantita- tively at 44.5° C must be kept in mind. Table 1.—Types or varieties of coliform organisms! (Standard Methods for the Examination of Water and Wastewater. 11th Edition. American Public Health Association, Inc., New York, N. Y., 1960, p. 518.) | | Voges Organism Indole | Methyl Red | Proskauer Citrate t Escherichia coli: | | Veristy: Ti seceeecescceeoeceseeas + 1 | _ — Vetlety Ul scceeec eee cee — | me — Escherichia freundii: | Variety) mice nnn mene _ | — fa nl Mariety TE cnc cece enemas + | | _— + Aerobacter aerogenes: | | Wate Ear eee _ — | = 54 Wariety: UY) season eceeoaeccc es = | _ | + 1 Reference. 2 Also referred to as Type. (1) American Public Health Association. 1960. Standard Methods for the Examina- tion of Water and Wastewater. 11th Ed. Am. Public Health Assoc., Inc., New York 19, N. Y. (2) Larkin, E. P., Litsky, W., and Fuller, J. E. 1955. Fecal Streptococci in: frozen foods. I. A bacteriological survey of some commercially frozen foods. Appl. Microbiol. 3: 98-101. (3) Mallman, W. L. and Darby, C. W. 1941. Uses of lauryl sulfate tryptose broth for the detection of coliform organisms. Am. J. Public Health 3]; 127-134. (4) Mundt, J. O., Shury, G. A., and McGarty, I. C. 1954. The coliform bacteria on strawberries. J. Milk and Food Technol. 17: 362-365. (5) Raj, H. and Liston, J. 1961. Detection and enumeration of fecal indicator organisms in frozen sea foods. I. Escherichia coli. Appl. Microbiol. 9: 171-174. (6) Fanelli, M. J. and Ayres, J. C. 1959. Methods of detection and effect of freezing on the microflora of chicken pies. Food Technol. 13: 294-300. ue ''Review of Methodology (7) Wilkerson, W. B., Ayers, J. C., and Kraft, A. A. 1961. Occurrence of enterococci and coliform organisms on fresh and stored poultry. Food Technol. 15: 286-292. (8) Kelly, C. B. 1960. Bacteriological criteria for market oysters. Robert A. Taft Sanitary Engineering Center (Cincinnati, Ohio) Technical Report F60-2. (9) Shelton, L. R., Leininger, H. V., Surkiewicz, B. F., Baer, E. F., Elliott, R. P., Hyndman, J. B., and Kramer, N. 1961. A bacteriological survey of the frozen precooked food industry. Division of Microbiology, Food and Drug Adminis- tration, U. S. Department of Health, Education, and Welfare, pp. 1-30. (10) Hajna, A. A. and Perry, C. A. 1943. Comparative study of presumptive and confirmative media for bacteria of the coliform group and for fecal streptococci. Am. J. of Public Health, 33: 550-556. (11) Presnell, M. W. 1961. An evaluation of the EC method in enumerating E. coli type I. Milk and Food Research Progress Report, May, pp. 4-5; August, pp. 6-7. (Preliminary data—unpublished. Robert A. Taft Sanitary Engineering Center.) (12) Vaughn, R. and Levine, M. 1935. Effect of temperature and boric acid on gas production in the colon group. J. of Bacteriol. 29: 24-25. (13) Walford, E. R., and Berry, J. A. 1948. Condition of oranges as affecting bacterial content of frozen juice with emphasis on coliform organisms. Food Research 13: 172-178. (14) Beisel, C. G. and Troy, V. S. 1949. The Vaughn-Levine boric acid medium as a screening presumptive test in the examination of frozen concentrated orange juice. Fruit Prod. J. 28: 356-357. (15) Lavine, M., Tanimoto, R. H., Minett, H., Arokaki, J., Fernandes, G. B. 1955. Simultaneous determination of coliform and Escherichia coli indices. Appl. Microbiol. 3: 310-314. (16) Clark, H. F., Geldreich, E. E., Kabler, P. W., Bordner, R. H., and Huff, C. B. 1957. The coliform group 1: The boric acid lactose broth reaction of coliform IMViC types. Appl. Microbiol. 5: 396-400. (17) Litsky, W., Fagerson, I. S., and Fellers, C. R. 1957. A bacteriological survey of commercially frozen beef, poultry, and tuna pies. J. of Milk and Food Technol. 20: 216-219. (18) Ross, A. D. and Thatcher, F. S. 1957. Bacteriological content of marketed pre- cooked frozen foods in relation to public health. Food Technol. 12: 369-371. (19) Canale-Parola, E. and Ordal, Z. J. 1957. A survey of the bacteriological quality of frozen poultry pies. Food Technol. 11: 578-582. (20) Kachikian, R., Fellers, C. R., and Litsky, W. 1959. A bacterial survey of frozen breaded shrimp. J. Milk and Food Technol. 22: 310-312. (21) Hartman, P. A. 1960. Further studies on the selectivity of violet red bile agar. J. of Milk and Food Technol. 23: 45-48. (22) Machala, W. E. 1961. A bacteriological investigation of frozen foods in the Oklahoma area. J. of Milk and Food Technol. 24: 323-327. (23) Huber, D. A., Zaborowski, H., and Rayman, M. M. 1958. Studies on the micro- bial quality of precooked frozen meals. Food Technol. 12: 190-194. (24) Kereluk, K. and Gunderson, M. F. 1959. Studies on the bacteriological quality of frozen meat pies. II. A comparison of the methods for the enumeration of coliforms. J. of Milk and Food Technol. 22: 176-178. (25) Silverman, G. J., Nickerson, J. T. R., Duncan, D. W., Davis, N. S., Schlachter, J. S., and Jaselow, M. M. 1961. Microbial analysis of frozen raw and cooked shrimp. 1. General results. Food Technol. 15: 455-458. 18 '' Coliform Organisms (26) Nickerson, J. T. R., Silverman, G. J., Salberg, M., Duncan, D. W., and Joselow, M. M. 1962. Microbial analysis of commercial frozen fish sticks. J. of Milk and Food Technol. 25: 45-47. (27) Gunderson, M. F., Rose, K. D. 1948. Survival of bacteria in a precooked fresh frozen food. Food Research 13; 256-263. (28) Hartman, P. A. 1958. The selectivity of autoclave-sterilized violet red bile agar. Food Research 23; 532-535. 19 ''IV. Methods for Isolation and Enumeration of Enterococci Herbert E. Hall There are two streptococcal groups which are frequently considered as common food contaminants. The first group is the enterococci which in- clude the species Streptococcus faecalis, S. faecalis var. zymogenes, S. faecalis var. liquifaciens, and S. durans. The second and larger group is the fecal streptococci which include the four enterococcal species listed above, in addition to S. bovis, S. equinus, S. acidominimus, and, on occasion, the S. mitis-salivarius species. For the purpose of this discussion, however, the term enterococci is used to designate the four species in the enterococcal group and the S. faecalis biotype referred to by many authors as S. faecium. A review of the cultural and fermentation reactions frequent- ly employed to differentiate the enterococci is offered in the table at the end of this chapter. Most media for the isolation of enterococci were devised for the exami- nation of water or sewage, but many have since been applied to problems of food contamination. Three types of media have been suggested: liquid media for use in isolation and most probable number (MPN) determina- tion; solid media for pour plating techniques; and solid media for spread- ing techniques. Liquid media that-have been used include the Streptococcus faecalis (SF) medium of Hajna and Perry (1943) (1); Enterococcus Presumptive (EP) broth of Winter and Sandholzer (1946) (2); Azide Dextrose (AD) broth of Mallman and Seligmann (1950) (3); Buffered Azide Glucose-Glycerol (BAGG) broth of Hajna (1951) (4); Ethyl Violet Sodium Azide (EVA) medium of Litsky, Mallman, and Fifield (1953) (5); Thallous Acetate broth of Barnes (1956*) (6); and the KF Streptococcus (KF) broth of Kenner, Clark, and Kabler (1961) (7). Solid media include the Enterococcus Confirmatory (EC) agar of Winter and Sandholzer (1946) (2); Mitis-Salivarius(MS) agar of Chapman (1946) (8); Enterococcus Medium of Reinhold, Swern, and Hussong (1953) (9) ; Thallous Acetate-Tetrazolium Glucose (TLTG) agar of Barnes (1956*) (6) and the KF Streptococcus (KF) agar of Kenner, Clark, and Kabler (1961) (7) Of these various media, the Azide Dextrose broth appears to be the least inhibitory as well as the least selective, but has been employed by a num- ber of workers as a presumptive medium with excellent results. Confirma- 20 '' Enterococci tion by means of a more selective medium such as the EVA broth, or by isolation and study of the characteristics of the isolates, is necessary when this medium is employed. The combination of AD and EVA media has been used by a number of workers in the study of enterococcal contamination of foods, Larkin, Litsky, and Fuller (1955) (JO) used this combination in the examination of 64 samples of frozen foods, including green beans, spinach, corn, lima beans, mashed potatoes, orange juice, lemon juice, limeade, grape juice, grapefruit juice, orangeade, melon balls, and sliced peaches. They reported enterococcal levels of 1.8 to 180,000 per 100 grams of food. No indication is given that studies were made of the isolates, so it must be assumed that they accepted growth in EVA as indicative of the presence of enterococci only. This was the indication given by Litsky, Mallman, and Fifield (1953) (5), who found that of 30 species of organisms tested, including 14 species of streptococci, only the enterococci grew in the EVA medium. Zaborowski, et al. (1958) (11), employed the AD-EVA combination with a large number of precooked frozen foods, such as beef pot pies, Mexican corn, peas, lima beans, cauliflower, chicken pot pies, roast turkey and dressing, and buttered potatoes. The authors considered the AD-EVA combination as satisfac- tory for the examination of frozen foods, as also was the EP medium cf Winter and Sandholzer (2) when used with EVA. Levels of organisms be- tween 15 and 45,000 per gram were reported, and slightly higher MPN’s were obtained with the AD medium than with EP broth. In this study, pre- liminary testing indicated that a positive result in EVA broth was indicative of the presence of enterococci. Kenner, et al. (1961) (7), in developing their KF media found that, in the EVA medium, Lactobacillus plantarum, Pediococcus cerevisiae, and S. mitis gave results identical to those of the enterococci, and growth without the typical color reaction was produced by S. salivarius, S. lactus, S. py- ogenes, and S. uberis. In confirmation of these findings, Splittstoesser, et al. (1961) (12), found that a study of their isolates from frozen vegetables, such as beans, corn, broccoli, and spinach, indicated that false-positive confirmations in EVA broth occurred in every trial and that the actual num- bers of enterococci were only a fraction of that indicated by growth in EVA broth. For example, in green beans in EVA, MPN was 16,000 per gram, but the revised MPN, based on the study of the isolates, was only 700 per gram. The nonconfirmable organisms appeared to be a species of Leu- conostoc, viridans streptococci, and other streptococci. These authors recommend further confirmation by culturing in 6.5 percent NaCl broth and by incubating at 45° C. In justice to the originators of the EVA medium, it should be added that the studies indicating a lack of specificity far greater than the original reported findings were made with commercial products, as compared to their laboratory-formulated medium, Furthermore, the 2) ''Review of Methodology ethyl violet dye is not a certified dye and may vary, therefore, in its compo- sition. From the above mentioned studies and others, it appears that AD- EVA will quantitate enterococci, but that some further confirmation is necessary, possibly that suggested by Raj, et al. (1961) (13), in which the Thallous Acetate Tetrazolium Glucose agar of Barnes (19567) (6) is used. A number of studies have been made on frozen foods in which the SF medium of Hajna and Perry (1943) (1) has been employed (Zaborowski, et al. (1958) (11), Splittstoesser, et al. (1961) (12), and Burton (1949) (14). Most of these have been comparative studies in which a number of media were employed, and in most instances the SF medium was less satis- factory than some of the others. If the incubation temperature of 45.5° C is accurately maintained, a high degree of specificity appears’ to be possible, but some workers (Splittstoesser, e¢ al., 1961) (12) find that a larger inoculum is required to initiate growth. This suggests that failure to detect low levels with this medium is a definite possibility. The enterococcus presumptive medium of Winter and Sandholzer (1946) (2) has been successfully employed by several workers (Splittstoesser, et al., 1961 (12); Zaborowski, et al., 1958 (11)). There does not, however, appear to be any significant advantage of this medium over AD broth, and confirmation is necessary. The Enterococcus Confirmatory medium of these authors appears satisfactory but not significantly better than EVA broth. The Thallous Acetate Tetrazolium Glucose medium of Barnes et al. (1956) (15) was successfully employed by her in studies of enterococcal populations in becon factories and by Raj, e¢ al. (1961) (13), as a con- firmatory medium. Fanelli and Ayres (1959 (16) obtained greater numbers of enterococci from commercially prepared frozen and unfrozen chicken pies with this medium than with AD or BAGG broth, and growth in these liquid media had to be confirmed either in EVA or by streaking on the Tetrazolium agar. It should be noted that in none of the above mentioned studies was quanti- tative recovery of known numbers of added organisms attempted. Most of the workers used several methods of isolation and quantitation and accepted the higher levels as indicating the actual enterococcal population. That many of the recoveries are actually quantitative is not seriously questioned, but this point is not definitely proven. Our own experience (Hall, Brown, and Angelotti, 1963) (17) with the KF Streptococcus agar of Kenner, et al. (1961) (7), indicates that quantitative recovery in the range of less than 10 to over 10° organisms per gram of food is possible. The specificity of this medium corresponds to that described by the authors for water and sewage when used to examine such foods as frozen and unfrozen meat pies, vegeta- bles, and meats (potatoes, carrots, peas, beef, chicken, turkey, and ham). The only growths on this medium that have required confirmation have been pink colonies produced by some streptococci other than Streptococcus 22 . | ''&% Table 1.—Characteristics of the Enterococci (Sherman, 1937 (18); Lancefield 1941 (19); Barnes, et al. 1956” (6); Lake, et al., 1957 (20); Bartley and Slanetz, 1960 (21)) Growth, Beta Growth, 0.1% Growth, Reduction Species hemolysis | Liquifaction | Growth, Growth, 6.5% meth 0.04% of tetra- Lancefield of blood of gelatin 10° G 45° C NaCl blue tellurite zolium Mannitol Sorbital group Streptococcus faecalis _ — + + + a + + + D Streptococcus faecalis var. liquifaciens ~_-_ = + + + + + + + + + D Streptococcus faecalis var. zymogenes ~----- | + —_— + + + + + + + + D Siveptococeus darais 22. nee + + + + _ — — — = D Streptococcus faecium ~-~--------------- Strong —_— + + + + — _— + — D alpha Or very hemolysis. poor | | uy 19909019 ''Review of Methodology durans, In addition to the organisms described by the authors, some Boccillus species may grow; but their growth characteristics differentiate them from streptococci. The KF Streptococcus broth of Kenner, et al. (1961) (7), offers the ad- vantage of a single-series test for obtaining a presumptive MPN test with more accuracy than the AD broth. Using the same foods as previously de- scribed for the KF agar, it has given us excellent results with both large and small inocula. Confirmation is required, however, since some strains of S. bovis, S. equinus, S. mitis, and S. salivarius will grow and produce an acid reaction. It appears that a number of methods are available that will satisfactorily isolate and, in some cases, quantitate the enterococci in foods. Agreement on one method as a base line for comparative work would do much to sim- plify the interpretation of results obtained from various foods, by other methods, and in different laboratories. When this is done it will be possible to determine the significance of the various levels and types of enterococci in foods as indicators of fecal contamination. REFERENCES (1) Hajna, A. A., and Perry, C. A. 1943. Comparative study of presumptive and confirmative media for bacteria of the coliform group and for fecal streptococci. Am. J. Public Health 33: 550-556. (2) Winter, C. E. and Sandholzer, L. A. 1946. Recommended procedure for detect- ing the presence of enterococci. Commercial Fisheries TL2, U. S. Department of Interior, Fish and Wildlife Service (November 1946). (3) Mallman, W. L. and Seligmann, E. B. 1950. A comparative study of media for the detection of streptococci in water and sewage. Am. J. Public Health 40: 286-289. (4) Hajna, A. A. 1951. A buffered azide glucose-glycerol broth for the presumptive and confirmatory tests for fecal streptococci. Public Health Laboratories 9: 80-81. (5) Litsky, W., Mallman, W. L., and Fifield, C. W. 1953. A new medium for the detection of enterococci in water. Am. J. Public Health 43: 873-879. (6) Barnes, E. M. 1956%. Methods for the isolation of faecal streptococci (Lancefield Group D) from bacon factories. J. Appl. Bacteriol. 19: 193-203. Barnes, E. M. 1956." Tetrazolium reduction as a means of differentiating Streptococcus faecalis from Streptococcus faecium. J. Gen. Microbiol. 14: 57-68. (7) Kenner, B. A., Clark, H. F., and Kabler, P. W. 1961. Fecal streptococci. I. Cultivation and enumeration of streptococci in surface waters. Appl. Microbiol. 9: 15-20. (8) Chapman, G. H. 1946. The isolation and testing of fecal streptococci. Am. J. Digest. Dis. 13: 105-108. Reinhold, G. W., Swern, M., and Hussong, R. V. 1953. A plating medium for the isolation and enumeration of enterococci. J. Dairy Sci. 36: 1-6. (10) Larkin, E. P., Litsky, W., and Fuller, J. E. 1955. Fecal streptococci in frozen foods. I. A bacteriological survey of some commercially frozen foods. J. Appl. Microbiol. 3: 98-101. (9 ~~ 24 '' Enterococci (11) Zaborowski, H., Huber, D. A., and Rayman, M. M. 1958. Evaluation of micro- bial methods used for the examination of precooked frozen foods. Appl. Microbiol. 6: 97-104. (12) Splittstoesser, D. F., Wright, R., and Hucker, G. J. 1961. Studies on media for enumerating enterococci in frozen vegetables. J. Appl. Microbiol. 9: 303-308. (13) Raj, H., Wiehe, W. J., and Liston, J. 1961. Detection and enumeration of fecal indicator organisms in frozen sea foods. II. Enterococci. Appl. Microbiol. 9: 295-303. (14) Burton, M. O. 1949, Comparison of coliform and enterococcus organisms as indices of pollution in frozen foods. Food Research 14; 434-438. (15) Barnes, E. M., Ingram, M., and Ingram, G. C. 1956. The distribution and significance of different species of faecal streptococci in bacon factories. J. Appl. Bacteriol. 19: 204-211. (16) Fanelli, M. J., and Ayres, J. C. 1959. Methods of detection and effect of freez- ing on the microflora of chicken pies. Food Technol. 13: 294-300. (17) Hall, H. E., Brown, D. F., and Angelotti, R. 1963. The recovery of enterococci from food using KF streptococcus media. J. Food Science. 28: 1-6. (18) Sherman, J. M. 1937. The Streptococci Bacteriological Reviews 1: 1-97. (19) Lancefield, R. 1941. Specific relationships of cell composition to biological activity of hemolytic streptococci. The Harvey Lectures, Sect. XXXVI, pp. 251-290. (20) Lake, D. E., Diebel, R. H., and Niven, C. F, 1957. The identity of Streptococcus faecium. Bacteriol. Proc. 57th Gen. Mtg., Soc. Am. Bacteriol., p. 13. (21) Bartley, C. H. and Slanetz, L. W. 1960. Types and sanitary significance of fecal streptococci isolated from feces, sewage, and water. Am. J. Public Health 50: 1545-1552. 20 ''V. Methods for Isolation of Salmonellae and Shigellae from Food Products Bildred M. Galton and John R. Boring, III Since members of the genus Salmonella were first recognized in the late 19th Century and were established as a widespread cause of human and animal disease, numerous methods for the isolation of these organisms from clinical specimens have been developed. As the epidemiology of salmonel- losis began to unfold, it became apparent that foods played an important part in the chain of infection. With the need for detection of salmonellae in foods, microbiologists began by applying methods that had been found satisfactory for recovery from human and animal fecal samples, but found in many instances that these procedures were inadequate. Many bacteriological problems have been raised by the rapid develop- ment of food processing. The commonly used media for the isolation of salmonellae were designed primarily for feces or other clinical specimens (Dack, 1955) (1) and they are not necessarily adequate for the detection of salmonellae in prepared foods. For example, the addition of comparative- ly large amounts of organic material, such as food samples, may have an adverse effect on the selectivity and enrichment quality of these media. Liquid Media for Salmonellae The two selective liquid enrichment media that have been used most widely are the Tetrathionate broth of Mueller as modified by Kauffman (1935( (2)and the Selenite broth of Leifson (1963 (3). However, Thom- son (1955) (4) found that ordinary nutrient broth was often as effective as, or superior to Selenite broth for the enri hment ef salmonellae from feces. Smith (1952) (5) reported that direct plating of intestinal contents on a selective or differential agar was superior to selenite and tetrathionate enrichments for the isolation of S. choleraesuis and S. abortusovis, and later recommended Brilliant Green MacConkey broth for isolation of the former type (Smith, 1959) (6).He concluded that Selenite and Tetrathi- onate broths were toxic for these types. Similarly, Banwart and Ayres (1953) (7) found that Tetrathionate broth inhibited the growth of S. para- typhi A, Several modifications of Tetrathionate broth have been recommended to suppress the growth of Proteus. Galton, et al. (1952) (8), found that the addition of 0.125 milligrams of sodium sulfathiazole per 100 milliliters of tetrathionate suppressed the multiplication of Proteus from canine fecal 26 '' Salmonellae and Shizellae swabs. Recently, Jameson (1961) (9) reported that the addition of 1 per- cent sodium lauryl sulphate and a bismuth sulphite solution to tetrathionate broth favored the selective inhibition of Proteus and allowed active multi- plication of salmonellae from sewer swabs. In recent years, other workers have made considerable progress in the development of methods and media particularly adapted to the recovery of salmonellae from foods. During a study to detect salmonellae in fresh pork sausage, Galton, Lowery, and Hardy (1954) (10) found that the addi- tion of a wetting agent, Tergitol No. 7 (Sodium heptadecyl sulphate, Union Carbide and Carbon Corporation) , to Tetrathionate Enrichment broth aided in dispersion and emulsification of the heavy layer of fat on the broth. Later (Galton, et al., 1955) (11) this agent, which appears to enhance salmonellae growth within a wide range of dilutions, was found to be helpful in the rehy- dration of dried dog meals. Selenite broth has been modified by the addition of cystine (North and Bartram, 1953) (12), brilliant green (Stokes and Os- borne, 1955) (13), and sulfapyridine (Osborne and Stokes, 1955) (14). Byrne, et al. (1955) (15), found that rehydration of dried eggs and ege products in distilled water and inoculation of this diluted suspension into multiple tubes of Selenite-Cystine broth favored the isolation of salmonellae. Similar enhancement of salmonellae growth was obtained from dried yeast when an aqueous suspension was incubated at 30° C for 24 hours prior to inoculation of Selenite or Tetrathionate broths. Slocum (cited in Dack, 1955) (1) observed that direct pre-enrichment procedures to enhance development of salmonellae prior to selective enrich- ment were more feasible than modification of the composition of the selec- tive medium for optimum selectivity with each of a wide variety of foods. The value of a lactose pre-enrichment for detecting small numbers of sal- monellae in dried egg products has recently been reported by North (1961) (16). He attributed the success of a pre-enrichment method to the restora- tion of a larger number of salmonellae to a state of active growth from a state of reduced viability or inanition caused by drying or freezing of the products during processing. Earlier, Silliker and Taylor (1958) (17) noted the inhibitory effect of egg albumen on the performance of selective enrich- ment broth. The value of nutrient broth as a primary medium was observed earlier by Thomson (1955) (4). He was able to isolate S. paratyphi B from a sample of flour over a period of nearly a year when it was inoculated into this medium; whereas, cultures from the same sample in Selenite broth were negative after 5 months. Thomson (1954) (18) also observed that salmonel- lae could usually be recovered more frequently from small rather than from large inocula. In this connection, he noted that, as a sample of feces was diluted up to 1:1000, coliform organisms progressively decreased, which made it easier to observe salmonellae colonies on selective plating media. He was of the opinion that the advantage of Selenite broth and Tetrathionate 27 ''Review of Methodology broth was due more to the fact that they are fluid media than to their composition. McCoy (1961) (19) was also able to recover salmonellae from higher dilutions of bone meal in Tetrathionate Enrichment broth when the organ- isms appeared to be absent from concentrated dilutions containing the larger inoculum. He attributed this to the probable presence of salmonellae in clumps of such material and the dispersion of these clumps during dilu- tion. Sugiyama, Dack, and Lippitz (1960) (20) found that salmonellae could be recovered effectively from egg albumen by inoculation into a non- inhibitory broth (lauryl tryptose or nutrient) containing polyvalent Salmo- nella “H” antiserum. After incubation and centrifugation, the sediment was then incubated in Selenite-Cystine broth. Plating Media Numerous plating media have been developed for the isolation of sal- monellae and other enteric pathogens. These include the differential media, such as Endo’s agar and Eosin-Methylene Blue agar, on which lac- tose- and non-lactose-fermenting organisms may be distinguished and Gram-positive organisms are inhibited. In addition, certain differential media, such as MacConkey’s agar (MacConkey, 1908) (21) and Leifson’s Desoxycholate agar (1935) (25), contain bile salts which makes the media selective and usually suppresses the spreading of Proteus. However, the highly selective plating media are more effective for the isolation of salmo- nellae from enrichment broths or for direct plating, since they are designed to inhibit many strains of the normal intestinal flora. The most widely used selective media are Bismuth Sulfite agar (Wilson and Blair, 1927 (22) ; Wil- son and Blair, 1931 (23); and de Loureiro, 1942) (24), Desoxycholate Citrate agar (Leifson, 1935) (25), SS agar, MacConkey Brilliant Green agar (Smith, 1959) (6), and the Brilliant Green-Phenol Red agar (BG) of Kristensen, Lester and Jurgens (1925) (26). Although salmonellae may be isolated on all these media, the BG agar, when properly prepared, is more inhibitive for enteric organisms other than salmonellae and salmonella colonies are detected with greater ease than on other selective media. In the experience of numerous investigators (Banwart and Ayres, 1953 (7); Galton, et al., 1952 (8); Galton, et al., 1954 (10); Galton, et al., 1955 (11); Osborne and Stokes, 1955 (14); and Edwards and Ewing, 1962 (27)), the use of BG agar, after enrichment from clinical specimens and foods, is indicated for maximum yield of salmonellae, other than S. typhi. In the examination of meats and other materials in which pseudomonads were present in large numbers, Galton, et al. (1954) (10), noted that the addition of 8 to 16 milligrams of sodium sulfadiazine per 100 milliliters of BG agar gave excellent results. While pseudomonads were effectively inhibited, there was no evidence of inhibition of salmonellae. Os- 28 '' Salmonellae and Shizellae borne and Stokes (1955) (14) found in the examination of egg products that the addition of 0.1 percent sulfapyridine to BG agar enhanced its selectivity for salmonellae. Enumeration of Salmonellae It is often desirable to estimate the total number of salmonellae present in samples of food harboring these microorganisms. For the enumeration of salmonellae in various foods, the method of choice must always reflect a consideration of the type of food being examined. Such factors as acidity, fat content, and the presence of inhibitory substances must be taken into account. One method used for this enumeration has been a modification of the most probable number (MPN) technique (Hoskins, 1934) (28). This method involves the enrichment of the bacteria from a known quantity of food and the subsequent isolation of the salmonellae on a selective medium. In general, the foods that have been examined for salmonellae with this method have been dried specimens, including egg albumen and animal feeds. North (1961) (16) has made a careful study of the techniques neces- sary to determine the MPN of salmonellae in dried egg albumen, and he has indicated several points of critical importance. These include the use of a pre-enrichment medium, such as lactose broth, and the ) (54), and Schmidt (1955) (55) are useful. For good recovery of heated spores, Wynne and Foster (1948) (58) found Pork Infusion Thioglycollate agar with 0.1 percent starch supplement in Prickett-tubes to be superior to 6 other media tested. Other recovery media that have been successfully used are Pork-Pea-Starch Thioglycollate agar (Andersen, 1951) (87) yeast extract starch (Wynne, et al., 1955) (83) and peptone supplemented with actively growing yeast suspension (Adams and Ayres, 1955) (88). Carbon dioxide, found to be necessary for adequate botulinal spore germination has been provided by an atmosphere of natural gas containing 5 percent CO, (Wynne and Foster, 1948) (58) and the addition of bicarbonate to media (Andersen, 1951 (87); Wynne, et al., 1955) (83). In general, an incubation temperature for spore germi- nation of 30° C has been found to be superior to 37° C and strong evi- dence, already discussed in the section on enrichment cultures, has been presented to support optimal incubation temperatures of about 24 to 27° C. There is need for additional work on the effects of incubation temperature on germination of all of the types of C. botulinum spores. Anaerobic methods A number of methods for development of anaerobic conditions are avail- able in the “Manual of Microbiological Methods” (Soc. Am. Bacteriol- ogists, 1957) (89) and will not be detailed here. A note of caution is in- serted, however, that certain of the methods may result in explosion or other danger to personnel if the proper precautions are not observed. Requiring emphasis, however, is the removal of CO, by alkaline-pyrogallol method, which has been shown to be necessary for. spore germination in some media (Wynne and Foster, 1948) (58). This method also produces some CO dur- ing oxidation of pyrogallate, which is inhibitory to some strains. Anaerobiosis under the required CO, environment can be accompanied most easily by evacuation of the anaerobic jar and replacement with pre- pared gas mixtures containing CO, (Angelotti, et al., 1962) (81). Also use- ful are systems such as the reaction of excess acid with a calculated amount of solid NaHCO, (Wynne and Foster, 1948) (58) ; incorporation of bicar- bonate into the medium (Andersen, 1951) (87); Wynne, et al., 1955) (83); or the addition to the anaerobic vessel of saturated solutions of car- bonates, bicarbonates, or mixtures of these as needed to supply the desired CO, vapor pressures (Parker, 1955) (90). Liquid media are easily rendered anaerobic by placing tubes or flasks of media in a boiling water bath to drive off oxygen and by subsequent rapid cooling in a water bath. With careful handling and inoculation of the organism into the bottom of the vessel, growth will usually occur in 72 ''Clostridium botulinum media devoid of thioglycollate or layer of oil or vapor over the liquid. The addition of sodium thioglycollate and agar in small amounts (Crisley, 1960) (33) to broth media is helpful in overcoming extreme oxygen sensi- tivity in some botulinal strains, particularly in frozen cultures, If it is necessary to avoid heating the medium, the inoculated vessels can be placed in an anaerobic vessel or vacuum desiccator and evacuated. During evacuation, the container should be bumped or jarred intermittently to effect bubbling in the inoculated medium to drive off dissolved oxygen (Crisley, 1960) (33). Anaerobic methods for isolation or spore-counting procedures are more involved. Wynne and Foster (1948) (58) and Wynne, et al., (1955) (83), employed agar cultures in oval Prickett-tubes for spore germination work. In these cultures an effective anaerobic seal was obtained by covering the inoculated, solidified agar in the tubes with 3 to 4 milliliters of 2 percent agar containing thioglycollate supplement. Andersen (1951) (87) applied the agar overlay principle to Petri plate counts. The basal recovery medium, spore inoculum, and sodium bicarbonate were mixed on a large, sterile Petri dish as for a pour plate. After solidification, 9 milliliters of sterile medium was added and a sterile glass plate was laid in it to exclude air. Above the plate, 30.0 milliliters of plain agar containing 0.1 percent sodium thioglycollate was added. C. botulinum has been successfully cultured on plates and in liquid cul- tures (Crisley, 1960 (33); Crisley and Helz, 1961) (34) by employing vacuum desiccators and the “activated iron wool” method of Parker (1955) (90) for removal of oxygen. The method is successful for plate cultures with- out preliminary evacuation, although it is slower, but some preliminary evacuation is required for broth cultures. Grade 00 steel wool was em- ployed as “iron wool.” Activation was accomplished by soaking the steel wool in acidified copper sulfate before addition to the vacuum desiccator containing the inoculated culture vessels. C. botulinum has also been successfully cultured (Crisley, 1962) (82) in Case-Anaero jars containing a mixture of 90 percent nitrogen and 10 percent CO., which has been used also for culture of C. perfringens (Angelotti, et al., 1962) (81). Indicators of anaerobiosis that change color upon exposure to air have been described “Manual of Microbiological Methods”, Soc. Am. Bacteriol: ogists, 1957 (89); Ulrich and Larsen, 1948 (91); Parker, 1955) (90). They are usually placed in the anaerobic vessel with the cultures before the anaerobic vessels are sealed. Summary The reported incidence of botulism is probably lower than the actual number of cases. The recent discovery of a new toxin type, type-F, also points up the possibility of the existence of other types presently unknown. 73 ''Review of Methodology Several reasons appear to account for gaps in the present state of knowl- edge of botulism. Included are the ubiquitous nature of C. botulinum and the tendency to overlook botulism as a possibility in cases of the disease, because its symptoms mimic those of several other diseases, and also because of the currently accepted belief in its rarity. The disease may also be missed in laboratory diagnoses as a result of the inadequacy of present routine methods, which may fail to detect low concentrations of toxin or botulinal strains whose environmental requirements differ from those clas- sically described. There has also been an increase in the number of en- vironments that the investigator must consider as sources of botulism. Foods processed by newer methods that do not involve heat sterilization, meats into which the organism has probably gained entrance during the life of the animal, and foods (particularly marine types) that are held at refrigerator temperatures are all possibilities calling for renewed vigilance on the part of investigators concerned with botulism. A number of measures can be taken to improve botulinal methodology in the laboratory. The sensitivity of routine animal tests for botulinus toxin in foods and cultures can probably be improved by the use of an extracting and diluting fluid composed of 0.2 percent gelatin solution in 0.1 M phosphate buffer at pH 6.5. For enrichment cultures of suspected samples, cooked meat medium, now widely used, should definitely be improved by the addition of 0.2 per- cent soluble starch, which has been shown to adsorb natural inhibitors of spore germination that are almost universally present in complex natural media. The use of heat as the selective ecological agent for destruction of concomitant organisms present in enrichments together with C. botulinum can also be made more critical. Because of widely varying thermal stabil- ities among the various known strains of botulinum spores, it is suggested that the following triplicate enrichments be employed: (a) Unheated, (b) heated at 60° C for 15 minutes, and (c) heated at 80° C for 30 minutes. The incubation of the enrichment cultures should be carried out at 30° C, which appears to be closer to the optimum for spore germination and toxin production than 37° C, now commonly employed. Whenever feasible it would be wise to employ multiple cultures incubated at 15 to 20° C, 30° C, and 37° C to increase the possibility of detection of new strains with varying optimal growth temperatures. Plate cultures for the isolation of C. botulinum from enrichment cultures are usually successful if made during the phase of vigorous growth in the enrichments. Blood or brain heart agars, or one of the agar recipes useful for the study of spore germination, can be usefully employed for this purpose. The tendency toward spreading of C. botulinum growth on plates can be minimized by several methods, the simplest of which is to minimize the condensation collecting on the plates in anaerobic jars. Carbon diox- 74 ''Clostridium botulinum ide should be provided by gas mixtures containing CO., saturated solutions of carbonates and bicarbonates placed in containers within the anaerobic jars, or bicarbonates incorporated into the medium. The existence of suitable anaerobic conditions is best determined by the use of oxidation- reduction indicators placed in the anaerobic jars. The minimal evidence for the incrimination of C. botulinum as the cause of a food intoxication is the demonstration of the toxin in the food consumed or in the contents of the alimentary tract of the patient. To accomplish this, animal protection tests should be carried out as described for types A, B, and E at the very least. Demonstration of the toxin and the isolation of the organisms from enrichment cultures bolsters the evidence of toxin in the incriminated food and yields material for further investigations of toxicity, thermal resistance and other properties of C. botulinum that are related to the potential food-poisoning hazard presented by the organ- ism. REFERENCES (1) Dolman, C. E. 1954. Additional botulism episodes in Canada. Can. Med. Assoc. J. 71: 245-249. (2) Dolman, C. E. and Murakami, L. 1961. Clostridium botulinum type F with recent observations on other types. J. Infectious Diseases 109: 107-128. (3) Dolman, C. E., Tomsich, M., Campbell, C. C. R., and Laing, W. B. 1960. Fish eggs as a cause of human botulism. Two outbreaks in British Columbia due to types E and B botulinus toxins. J. Infectious Diseases 106: 5-19. (4) Mackey-Scollay, E. M. 1958. Two cases of botulism. J. Pathol, and Bacteriol. 75: 482-485. (5) Meyer, K. F. 1956. The status of botulism as a world health problem. Bull. World Health Organization 15; 281-298. (6) Dolman, C. E, 1960. Type E botulism: A hazard of the North. Arctic 13: 230-256. Dolman, C. E. 1957.” Type E (fish-borne) botulism. A review. Jap. J. Med. Sci. and Biol. 10; 383-395. (8) Schmidt, C. F., Leckowich, R. V. and Folinazzo, J. F. 1961. Growth and toxin production by type E Clostridium botulinum below 40° F. J. Food Sci. 26: 15. (9) Meyer, K. F. and Dubovsky, B. J. 1922." The distribution of the spores of B. botulinus in California. II, J. Infectious Diseases 31: 541-555. (10) Easton, E, J. and Meyer, K. F. 1924. Occurrence of Bacillus botulinus in human and animal excreta. XXI. J. Infectious Diseases 35: 207-212. (11) Orr, P. F. 1920.* Some observations on the biological characteristics of Bacil- lus botulinus. Proc, Soc. Exp. Biol. and Med. 17; 47. (12) Orr, P. F. 1920. Some observations on the biological characteristics of Bacil- . lus botulinus and its toxins. Abstr. Bacteriol. 4: 10. (13) Coleman, G. E. and Meyer, K. F. 1922. Some observations on the patho- genicity of B. botulinus. J. Inf. Dis. 31: 622-649. (7 ~ 75 ''(14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) Review of Methodology Starin, W. A. and Dack, G. M. 1923. Aggdutination studies of Clostridium botulinum. J. Infectious Diseases 33: 169-183. Starin, W. A. and Dack, G. M. 1925. Pathogenicity of Clostridium botulinum. J. Infectious Diseases 36: 383-412. Wallace, G. I. and Park, S. E. 1933. Microbiology of frozen foods. V. The behavior of Clostridium botulinum in frozen fruits and in vegetables. J. In- fectious Diseases 5z: 150-156. Mundt, J. O., Mayhew, C. J., and Stewart, G. 1954. Germination of spores in meats during cure. Food Technol. 7: 435-436. Topley and Wilson’s Principles of Bacteriology and Immunity 4th ed., Volumes I, IT, 1955, Williams and Wilkins Company, Baltimore, Maryland. Smith, L. D. S. 1955. Introduction to the Pathogenic Anaerobes. Univ. of Chicago Press, Chicago, Illinois. Littauer, U. 1951. Observations on the type A toxin of Cl. botulinum. Nature 167: 994-995. Boor, A. K., Tresselt, H. B., and Schantz, E. J. 1955. Effects of salts and colloids on potency of botulinum toxin. Proc. Soc. Exp. Biol. and Med. 89: 270-272. Bronfenbrenner, J. and Schlesinger, M. J. 1922. Some of the factors contrib- uting to toxicity of botulinus toxin by mouth. J. Am. Med. Assoc. 78: 1519- 1521. Abrams, A., Kegeles, G., and Hottle, G. A. 1946. The purification of toxin from Clostridium botulinum type A. J. Biol. Chem. 164; 63-79. Lewis, K. H. and Hill, E. V. 1947. Practical media and control measures for producing highly toxic cultures of Clostridium botulinum type A. J. Bacteriol. 53: 213-230. Mager, J., Kindler, S. H., and Grossowicz, N. 1954. Nutritional studies with Clostridium parabotulinum type A. J. Gen. Microbiol. 10: 130-141. Appleton, G. S. and White, P. G. 1957. Clostridium botulinum type C. I. Studies on laboratory variables affecting toxin production and toxoid evaluation. Am. J. Vet. Res. 18: 942-946. Stevenson, J. W., Helson, V. A., and Reed, G. B. 1947. Preparation of Clostri- dium parabotulinum toxins. Canadian J. Res. 25E: 14-24. Cartwright, T. E. and Lauffer, M. A. 1952. Assay of botulinum A toxin with goldfish. Proc. Soc. Exp. Biol. Med. 8]: 508-511. Cartwright, T. E. and Lauffer, M. A. 1958. Temperature effects on botulinum A toxin. Proc. Soc. Exp. Biol. Med. 98: 327-330. Lamanna, C., Jensen, W. I., and Brass, I. D. J. 1955. Body weight as a factor in the response of mice to botulinal toxins. Am. J. Hyg. 62; 21-28. Duff, J. T., Wright, G. G., Klerer, J., Moore, D. E., and Bibler, R. H. 1957." Studies on immunity to toxins of Clostridium botulinum. I, A simplified proce- dure for isolation of type A toxin. J. Bacteriol. 73: 42-47. Duff, J. T., Klerer, J., Bibler, R. H., Moore, D. E., Gottfield, C., and Wright, G. G. 1957.” Studies on immunity to toxins of Clostridium botulinum. II. Produc- tion and purification of type B toxin for toxoid. J. Bacteriol. 73: 597-601. Crisley, F. D. 1960. Routine method for the goldfish assay of toxin in crude culture centrifugates of Clostridium botulinum type A. Appl. Microbiol. 8: 282-285. Crisley, F. D. and Helz, G. E. 1961. Some observations of the effect of filtrates of several concomitant representative bacteria on Clostridium botulainum type- A. Can. J. Microbiol. 7: 633-639. 76 ''(35) (36) (37) (38) (39) (40) (41) (42) (43) (44) (45) (46) (47) (48) (49) (50) (51) (52) (53) (54) (55) (56) Clostridium botulinum Bonventre, P. F. and Kempe, L. L. 1960. Physiology of toxin production by Clostridium botulinum types A and B. I. Growth, autolysis, and toxin produc- tion. J. Bacteriol. 79: 18-23. Dolman, C. E. 1961. Further outbreaks of botulism in Canada. Can. Med. Assoc. J. 84: 191-200. Bronfenbrenner, J., Schlesinger, M. J., and Orr, P. F. 1921. Concerning toxic byproducts of Bacillus botulinus. Proc. Soc. Exp. Biol. and Med. 18: 181-182. Prevot, A. R. and Brygoo, E. R. 1953. Nouvelles recherches sur le botulisme et ces cing types toxiniques. Ann. Inst. Pasteur 85: 544-575. Duff, J. T., Wright, G. G., and Yarinsky, A. 1956. Activation of Clostridium botulinum type E toxin by trypsin. J. Bacteriol. 72: 455-460. Sakaguchi, G. and Tokyama, Y. 1955." Studies on the toxin production of Clostridium botulinum type E. I. A strain of genus Clostridium having the ac- tion to promote type E botulinal toxin production in a mixed culture. Jap. J. Med. Sci. and Biol. 8: 247-253. Sakaguchi, G. and Tokyama, Y. 1955.” Studies on the toxin production of Clostridium botulinum type E. II. The mode of action of the contaminant organisms to promote toxin production of type E organisms. Jap. J. Med. Sci. and Biol. 8: 255-262. Batson, H. C. 1956. An introduction to statistics in the medical sciences. Bur- gress Publishing Co., Minneapolis, Minnesota. Schiibel, K. 1923. Uber das Botulinus toxin. Arch. Exp. Pathol. Pharmak. 96: 193-259. Coleman, I. W. 1954. Studies on the oral toxicity of Clostridium botulinum type A. Can. J. Biochem. and Physiol. 32: 27-34. Dack, G. M. 1956. Food Poisoning. 3d ed. University of Chicago Press, Chicago, Illinois. Dack, G. M. 1926." Behavior of botulinum toxin in alimentary tract of rats and guinea pigs. J. Infectious Diseases 38: 174-181. Meyer, K. F. and Dubovsky, B. J. 1922.” The distribution of the spores of B. botulinus in the United States. IV. J. Infectious Diseases 3]: 559-594. Dolman, C. E. and Kerr, D. E. 1947. Botulism in Canada, with report of a type E outbreak at Nanaimo, B. C. Can. J. Public Health 38: 48-57. Hall, I. C. and Peterson, E. 1923. The effect of certain bacteria upon the toxin production of Bacillus botulinus in vitro. J. Bacteriol. 8: 319-341. Francillon, M. 1925. Einfluss der aeroben Mischinfektion auf Entwicklung and Toxinbildung des Bacillus botulinus. Arch. F. Hyg., Munich 95: 121-139. Dack, G. M. 1926. Influences of some anaerobic species on toxin of Cl. botu- linum with special reference to Cl. sporogenes. J. Infectious Diseases 38: 165— 173. Stark, C. N., Sherman, J. M., and Stark, P. 1928. The influence of the filtrates of Clostridium botulinum and Clostridium sporogenes upon the growth of each of these organisms. J. Bacteriol. 15: 38. Stark, C. N., Sherman, J. M., and Stark, P. 1929. Destruction of botulinum toxin by Bacillus subtilis. Proc. Soc. Exp. Biol. and Med. 26: 343-344. Foster, J. W. and Wynne, E. S. 1948." The problem of “dormancy” in bacterial spores. J. Bacteriol. 55: 623-625. Schmidt, C. F. 1955. The resistance of bacterial spores with reference to spore germination and its inhibition. Ann, Rey. Microbiol. 9: 387-400. Foster, J. W. and Wynne, E. S. 1948." Physiological studies on spore germina- tion, with special reference to Clostridium botulinum. IV. Inhibition of germi- nation by unsaturated C18 fatty acids. J. Bacteriol. 55: 495-501. 77 ''(57) (58) (59) (60) (61) (62) (63) (64) (65) (66) (67) (68) (69) (70) (71) (72) (73) (74) (75) (76) Review of Methodology Olsen, A. M. and Scott, W. J. 1946. Influence of starch in media used for the detection of heated bacterial spores. Nature 157: 337. Wynne, E. S. and Foster, J. W. 1948. Physiological studies on spore germina- tion with special reference to Clostridium botulinum. I. Development of a quantitative method. J. Bacteriol. 55: 61-68. Olsen, A. M. and Scott, W. J. 1950. The enumeration of heated bacterial spores. I. Experiments with Clostridium botulinum and other species of Clostridium. Australian J. Sci. Res. Series B 3: 219-233. Dolman, C. E. 1957." Recent observations on type E botulism. Can. J. Pub. Health 48: 187-198. Cardella, M. A., Duff, J. T., Gottfried, C., and Begal, J. S. 1958. Studies on immunity to toxins of Clostridium botulinum. IV. Production and purification of type C toxin for conversion to toxoid. J. Bacteriol. 75: 360-365. Sterne, M. and Wentzel, L. M. 1950. A new method for the large-scale produc- tion of high-titre botulinum formol-toxoid types C and D. J. Immunol. 65: 175-183. Polson, A. and Sterne, M. 1946. Production of potent botulinum toxins and formal-toxoids. Nature 158: 238. Boroff, D. A. 1955. Study of toxins of Clostridium botulinum. III. Relation of autolysis to toxin production. J. Bacteriol. 70: 363-367. Gendon, Iu. Z. 1957. Growth and toxin formation of Cl. botulinum type A in cellophane sacs. J. Microbiol. Epidemiol. and Immunobiol., U.S.S.R. 28: 373- 377. Cardella, M. A., Duff, J. T., Wingfield, B. H., and Gottfried, C. 1960. Studies on immunity to toxins of Clostridium botulinum. VI. Purification and detoxifi- cation of type D toxin and the immunological response to toxoid. J. Bacteriol. 79: 372-378. Dubovsky, B. J. and Meyer, K. F. 1922. An experimental study of the methods available for the enrichment, demonstration, and isolation of B. botulinus in specimens of soil and its products, in suspected food, clinical and in necropsy material. I. J. Infectious Diseases 31: 501-540. Pederson, H. O. 1955. On type E botulism. J. Appl. Bacteriol. 18: 619-628. Ohye, D. F. and Scott, W. Je 1957. Studies in the physiology of Clostridium botulinum type E. Australian J. Biol. Sci. 10: 85-94. Tanner, F. W. and Oglesby, E. W. 1936. Influence of temperance on growth and toxin production by Clostridium botulinum. Food Res. 1: 481-494, Ohye, D. F. and Scott, W. J. 1953. The temperature relations of Clostridium botulinum types A and B. Australian J. Biol. Sci. 6: 178-189. Williams, O. B. and Reed, J. M. 1942. The significance of the incubation tem- perature of recovery cultures in determining spore resistance to heat. J. Infee- tious Diseases 71; 225-227. Sugiyama, H. 1951. Studies on factors affecting the heat resistance of spores of Clostridium botulinum. J. Bacteriol. 62: 81-96. Prevot, A. R. and Huet, M. 1951. Existence en France du botulisme humain d’ origine pisciaire et de Cl. botulinum E. Bull. Acad. Nat. Med. 135; 432-435. Dolman, C. E., Chang, H., Kerr, D. E., and Shearer, A. R. 1950. Fish-borne and type E botulism: two cases due to home-pickled herring. Can. J. Public Health 41; 215-229. Kindler, S. H., Mager, J., and Grossowiez, N. 1955. Production of toxin by resting cells of Cl. parabotulinum type A. Science 122: 926-927. 78 ''(77) (78) (79) (80) (81) (82) (83) (84) (85) (86) (87) (88) (89) (90) (91) Clostridium botulinum Eales, C. E. and Turner, A. W. 1952. Description of Clostridium botulinum type D recovered from soil in South Australia. Australian J. Exp. Biol. Med. Sci. 30: 295-300. Dolman, C. E. 1953. Clostridium botulinum, type E. Atti. Del. VI. Cong. Intern. Microbiol. 4, Sec. 12: 130-132. McClung, L. S. and Toabe, R. 1947. The egg yolk plate reaction for the pre- sumptive diagnosis of Clostridium sporogenes and certain species of the gan- grene and botulinum groups. J. Bacteriol. 53: 139-147. Willis, A. T. and Hobbs, B. C. 1958. A medium for the identification of clostridia producing opalescence in egg-yolk emulsions. J. Pathol. Bacteriol. 75: 299-305. Angelotti, R., Hall, H. E., Foter, M. J., and Lewis, K. H. 1962. Quantitation of Clostridium perfringens in foods. Appl. Microbiol. 10: 193-199. Crisley, F. D. 1962. Unpublished data. Wynne, E. S., Schmieding, W. R., and Daye, G. T., Jr. 1955. A simplified me- dium for counting Clostridium spores. Food Research 20: 9-12. Riemann, H. 1957. Some observations on the germination of Clostridium spores and the subsequent delay before the commencement of vegetative growth. J. Appl. Bacteriol. 20: 404-412. Tsuji, K. and Perkins, W. E. 1962. Sporulation of Clostridium botulinum. 1. Selection of an aparticulate sporulation medium. J. Bacteriol. 84: 81-85. Halvorson, H. O. 1959. Spores Symposium. Am. Inst. Biol. Sci. Pub. No. 5, Washington, D.C. Andersen, A. A. 1951. 2 0 49 3 2 i 150 27 17 4 2 1 94 26 5 2 1 70 3 2 2 210 33 20 4 2 2 130 32 5 2 2 95 3 2 3 290 40 24 4 2 3 170 38 5 2 3 120 3 2 4 47 27, 4 2 4 210 44 5 2 4 150 3 2 5 31 4 2 5 50 5 2 5 180 3 3 0 240 28 17 4 3 0 110 27 2 3 0 79 3 3 1 460 34 21 4 3 1 160 33 5 3 1 110 3 3 2 1,100 41 24 4 3 2 220 39 5 3 2 140 3 3 3 48 28 4 3 3 280 45 5 3 3 180 3 3 4 56 31 4 3 4 360 52 5 3 4 210 3 3 5 35 4 3 5 59 5 3 S 250 3 4 0 35 21 4 4 0 240 34 5 4 0 130 3 4 1 43 24 4 4 1 390 40 5 4 1 170 3 4 2 50 28 4 4 2 700 47 5 4 2 220 3 4 3 59 32 4 4 = 1,400 54 5 4 3 280 a 4 4 67 36 4 4 4 62 5 4 4 350 3 4 5 40 4 4 5 69 s 4 5 430 3 5 0 25 4 5 0 41 5 5 0 240 3 5 1 29 4 5 1 48 5 5 1 350 3 5 2 32 4 5 2 56 5 5 2 540 3 5 3 37 4 5 3 64 5 5 3 920 3 5 4 41 4 § 4 72 5 5 4 1,600 3 5 5 45 4 5 5 81 Adapted from: Hoskine, J: K, “Most Probable Numbers for Evaluation of Coli-Aerogenes Tests by Fermenation Tube Method,” Public Health Reports, Vol. 49, 393-405 (1934). miuiajopg w10fyo-y ''Laboratory Manual VI. Fecal Streptococci A. Plate Count Method (suitable for analyzing foods in which large num- bers of fecal streptococci may be expected ) : 1. Methods, materials and culture media: a. Same as Agar Plate Colony Count (IV, A, steps 1 through 8). b. KF Streptococcus agar—approximately 300 ml for each food spe- cimen examined. . 2. Procedure: a. Pipette aseptically 1 ml of each dilution of the food-homoge- nate, prepared in II above to each of appropriately marked du- plicate culture dishes. b. Pour each plate with 10 to 15 ml of KF Streptococcus agar. c. Thoroughly mix inoculum and agar immediately after pouring plates and allow to solidify. Pour an agar control plate on each bottle of agar. d. Invert plates and incubate at 35° C for 48 hours. e. Remove plates from incubator and observe macroscopically for type of growth, colony size, and color. f. Select plates showing an estimated 30 to 300 colonies and, using a dissecting microscope, count dark red colonies or those having ; a red or pink center. g. Calculate number of organisms per g of food by multiplying col- ony counts by dilution factor. Record results. B. Most Probable Number Method (recommended for use in the routine surveillance of foods for sanitary quality in which small numbers of fecal streptococci may be expected) : 1. Apparatus, materials and culture media: , a. Same as Agar Plate Colony Count (IV, A, steps 3, 4, 5, and 6). b. 1.1 ml pipettes with 0.1-ml and 1-ml graduations—one pipette for each diln. 7 c. KF Streptococcus broth. Dispense 10-ml amounts into 150 x 15 mm screw-capped or aluminum capped tubes. Sterilize by autoclaving for 10 minutes at 121° C. Five tubes of broth for each diln. 2. Procedure: a. Pipette aseptically 1 ml of each of the decimal dilns of food-hom- ogenate prepared in II above to each of 5 separate tubes of KF Streptococcus broth. . b. Incubate inoculated tubes at 35° C for 48 hours. oA. '' Fecal Streptococci c. Following incubation, observe tubes for color change to yellow and for turbidity. d. Tubes that have developed yellow color are considered positive for fecal streptococci. Select the highest diln at which all 5 tubes are yellow and the next 2 higher dilns. e. Determine the MPN of fecal streptococci per gram following instructions under V, B, 2, e and h above (Coliform bacteria, Most Probable Number Method.) 95 ''Laboratory Manual VII. Staphylococci A. Plate Count Method (suitable for analyzing foods in which large num- bers of staphylococci may be expected). 1. Apparatus, materials and culture media: a. Same as Agar Plate Colony Count (IV, A, 1, 3 through 8). b. Pipettes, 1.1 ml with 0.1 ml and 1.0 ml graduations—one pipette for each diln. c. Sterile glass spreader bar (Glass rod bent into hockey stick form and fire polished). Six spreader bars for each food specimen examined. d. Tellurite-Polymyxin-Egg Yolk Agar (TPEY )—Approximately 200 ml for each food specimen examined: Tryptone ~---------------------------------------- 10 g Yeast extract _____----_---------------------------- 5 g déManritdl —— oa aeecee eee eee 5 g Sodium chloride _--__------------------------------- 20 g Lithium chloride _-__-_----------------------------- 2¢ 1% soln of Polymyxin B (Seitz filtered) ~--------------- 0.4 ml 1% solution of potassium tellurite (sterilized at 121° C for 15. minutes). __---__----asep renee emcees 10 ml Agar —...---------------------------------------- 18 g Egg yolk emulsion ~-------------------------------- 100 ml Distilled Watér == 2s e eee ee 900 ml Place the first 5 ingredients in 900 ml cold distilled water and dissolve completely by warming. Adjust to pH 7.2 to 7.3 and add the agar. Heat to boiling to dissolve the agar. Sterilize by autoclaving at 121° C for 15 minutes. Cool to 50 to 55° C ina water bath. Aseptically add the Polymyxin B and potassium tellurite solns and egg yolk emulsion. Mix and pour 15 to 17 ml per plate. Permit plates to dry thoroughly overnight. (Egg yolk emulsion is a 30% volume per volume mixture in physiological saline. Fresh eggs are soaked for one min in a 1:1000 diln of saturated mercuric chloride soln to sterilize the shells. The eggs are cracked and the yolks removed aseptically. Thirty ml of egg yolk is added to 70 ml of sterile physiological saline in a Waring Blendor and homogenized for approximately 5 seconds at slow speed). e. Trypticase Soy Agar: Dispense in 5- to 7-ml volumes in screw- capped, 150 x 15-mm tubes and sterilize at 121° C for 15 96 ''g. h. Staphylococci minutes. Tilt tubes to make agar slants. Five slants for each countable plate. Trypticase Soy Broth: Dispense in 5-ml volumes in 150 x 15-mm screw-capped tubes and sterilize at 121° C for 15 min. One tube of broth is sufficient for approximately 8 coagulase tests. Coagulase Plasma (available commercially). Agglutination tubes, 10 x 100 mm—one tube for each coagulase test and one plasma control and one coagulase-positive control. 2. Procedure: a. ek} Pipette aseptically 0.1 ml of each diln of the food-homogenate prepared in II above to the surface of each of appropriately marked, duplicate plates of TPEY medium. In surface plating techniques 0.1 ml of each dilution is plated. Because only 0.1 ml is used, 0.1 ml of each diln is always plated on the agar plate marked with the next higher diln. (For example, 0.1 ml of the 10~ diln onto the surface of the plate marked 10-2.) Continue inoculating aliquots from all serial dilns through 10-°. Using a separate and sterile glass spreader bar for each diln, smear the inoculum evenly and completely over the agar surface. Place all diln plates in the incubator in the inverted position and incubate for 24 hours at 35° C. Following incubation, observe plates for the development of jet black circular, convex colonies 1.0 to 1.5 mm in diameter. Coag- ulase-positive staphylococci usually show one or more of the fol- lowing reactions (a) zone of precipitation around colony (b) clear zone (halo) around colony and white precipitate beneath (c) no zone or halo around colony but precipitation beneath the colony. If the colonies are too small or reactions indistinct, rein- cubate plates for an additional 12 to 24 hours at 35° C. Count the number of such colonies on plates with 30 to 300 colo- nies wherever possible and record as the presumptive number of coagulase-positive staphylococci per g. From countable plates with well isolated colonies select and mark 5 colonies for coagulase testing. With the tip of a straight needle touch the selected colony and inoculate a slant of Trypticase Soy Agar. Incubate for 24 hours at 35°C. With a loop remove the remainder of the colony and emulsify in 0.2 ml of Trypticase Soy Broth in an agglutination tube. Add to the tube 0.5 ml of coagulase plasma, mix and incubate in a 35° C water bath for 4 hours. Following incubation, observe plasma tubes for coagulation. Any oF ''Laboratory Manual degree of coagulation, even a fibrin strand, is considered posi- tive. k. Note those tubes which are negative and obtain the 24-hour Tryp- lL. ticase Soy Agar slant culture made from the original colony. Emulsify a loopful of growth from 24-hour slant culture in 0.2 ml Trypticase Soy Broth and repeat coagulase test as in (i) and (j) above. B. Most Probable Number Method (recommended for use in the routine surveillance of foods for sanitary quality in which small numbers of staphylococci may be expected) : 1. Apparatus, materials and culture media: a. Same as Agar Plate Colony Count (IV, A, 3 through 6). 1.1-ml pipettes marked in 0.1- and 1.0-ml graduations—one pipette for each diln. Inoculating needle, nichrome or platinum-iridium wire B & S guage No. 25. TPEY agar (See VII, A, 1, d above)—one plate for each broth tube inoculated. Cooked Meat 10% Salt (NaCl) Broth. The basal medium described above is available commercially in the dehydrated form and is used as the stock medium for pre- paring Cooked Meat 10% Salt (NaCl) Broth. Suspend 125 g of Cooked Meat Medium in 1,000 ml of cold distilled water. Add 95 g of sodium chloride, mix thoroughly and allow to stand for at least 15 min or until all meat particles are saturated. Dispense the medium in 150 x 15 mm screw-capped tubes. Attempt to maintain an approximate ratio in each tube of 14 meat particles and 2% supernatant broth for a total volume of approximately 10 ml. Sterilize at 121° C for 30 min. Five tubes of medium for each diln. f. Trypticase Soy Agar (See VII, A, 1, e above) o SB One slant tube of agar for each colony to be coagulase tested. Trypticase Soy Broth (See VII, A, 1, f above) Four tubes of broth. 2. Procedure: a. b. Cc Pipette aseptically 1 ml of each of the decimal dilns of food-hom- ogenate prepared in II above to each of 5 separate tubes of Cooked Meat 10% Salt (NaCl) Broth. Take care to introduce the inoculum well down into the meat particles. Incubate inoculated tubes at 35° C for 24 hours. Following incubation, transfer a loopful of material from each inoculated tube to the surface of dried plates of TPEY agar to 98 ''Staphylococci obtain isolated colonies. Incubate plates for 24 hours at 35° C and select typical staphylococcal colonies (VII, A, 2, e above) for coagulase testing. d. Test colonies for coagulase production (as described in VII, A, 2, h through k above). e. Record the number of tubes in each diln from which coagulase- positive staphylococci were isolated on TPEY agar. f. Determine the MPN per gram of confirmed coagulase-positive staphylococci according to the instructions under V, B, 2, e and h above (Coliform bacteria, Most Probable Number Method). 99 ''Laboratory Manual VIII. Salmonellae-Shigellae A. Enrichment method suitable for the isolation of salmonellae and shigellae from foods in which they are present in low concentrations: 1. Materials: a. Sterile, 16-ounce, screw-capped jars or sterile plastic bags similar to those used for the collection of milk samples. One container for each food specimen to be examined. b. Sterile tongue depressors or similar instruments for filling con- tainers with food specimens. c. Balance, with weights, 500 g capacity, sensitivity 100 mg. d. Mechanical blender capable of operation at approximately 8,000 rpm. e. Inoculating loop, 5 to 6 mm, platinum-iridium or nichrome, B & S gauge, No. 24. f. Culture dishes (100 x 15 mm) glass or plastic. g. Pipettes, 1-ml capacity, with 0.01-ml graduations and 5 and 10 ml capacity with 0.l-ml graduations and 0.2-ml capacity with 0.01-ml graduations. h. Incubator 35° C to 37° C. 2. Culture media and reagents: a. Dehydrated products to be prepared as directed on bottle: (1) (4) (5) (6) (7) (8) Brilliant Green agar (BG). Nove: Sterilization of this medium is critical. A longer period tends to decrease its selectivity and insufficient sterilization results in inhibi- tory action against salmonellae. Brilliant Green Sulfadiazine agar (BGS). Prepare a 1.6% soln of sodium sulfadiazine in distilled water. Boil for 10 min to sterilize. Add aseptically 5 ml of this solution per liter of Brilliant Green agar just prior to pouring plates. Triple Sugar Iron agar (TSI). Slant in such a manner to obtain a 1- to 14 in butt. Bismuth Sulfite agar (WB). SS agar (Shigella-Salmonella agar). MacConkey agar. Lysine Iron agar. Slant tubes so as to obtain approximately a l-in butt and a 1% inch slant. Nutrient broth. 100 ''Salmonellae-Shigella (9) Lactose broth, (10) Simmons Citrate agar. (11) Urea agar (for 24 hour test). (12) Tetrathionate enrichment broth. Cool the base broth and add 1 ml of a 1:1000 aqueous soln of brilliant green to 100 ml of broth and 2 ml of iodine soln just prior to tubing. The iodine soln is prepared by dissolving 5 g of potassium iodide (CP) in 20 ml of distilled water and adding 6 g of iodine crystals (CP-resublimed). Comparative studies have shown that the iodine soln may be added when the medium is prepared if used within 8 days after preparation (Galton, et al., J. Infectious Disease 9]: 1-18, 1952). b. Tergitol No. 7—10% aqueous soln. Sterilize at 121° C for 20 Cc. minutes, Rapid urea test medium: Urea ~------- 2.0 ¢ Monopotassium phosphate, dihydrogen ~----____________ 0.1 g Dibasic sodium phosphate _---__---_------§ 0.1 g Sodium chloride ~--~-__-_-- 0.5 g Ethyl alcohol ~--~----_- 1.0 ml Distilled water ~----_-____-- 99.0 ml The reagent is adjusted to pH 7.0 with NaOH and 0.5 ml of a 0.2% aqueous soln of phenol red added. Do not sterilize except by filtration. Sterility is not necessary, however, as the medium may be stored in the refrigerator for several weeks unsterilized without spoilage. If the medium becomes pink (alkaline to phe- nol red), it should be discarded. To perform test, dispense broth into 3- or 4-inch clean tubes in 0.2 ml amounts. Inoculate with a generous amount of growth from the Triple Sugar Iron agar slant using a wire loop. Incubate for 30 min at 37° C, then read. A change in color from the the pale yellow of the fresh medium to an intense pink or fuchsia indicates the hydrolysis of urea or a positive test. d. Rapid indol test soln Tryptone’ sess2-2—----— ee ge 2.0 ¢g Dibasic sodium phosphate (anhydrous) ~~-~__________ 0.2 g Monopotassium phosphate (anhydrous), dihydrogen —___ 0.1 g Sodium chloride ~~~~--- 0.8 ¢ Distilled water _--------- 100.9 ml The pH is adjusted to 7.0 to 7.2. The soln should be sterilized at 121° C for 20 minutes. It may be kept in the refrigerator in 101 ''Laboratory Manual a flask and dispensed as needed with a sterile pipette. This soln should be diluted with an equal volume of normal saline before use. The tubes in which the tests are performed need not be sterile. To perform test, dispense the diluted medium into 3- or 4- inch clean tubes in 0.2-ml amounts. Inoculate a generous amount of growth from the Triple Sugar Iron agar (TSI) slant using a wire loop. Incubate 2 hours at 37° C and add 0.2 to 0.3 ml of Kovac’s reagent. This reagent forms a layer over the sur- face of the broth culture. The liberation of indol is indicated by a color change of this layer from yellow to red. Tryptone Broth (for 24-hour test). Prepare a 1% soln of tryptone, dispense in 2- to 3-ml amounts in 4-inch tubes and sterilize at 121° C for 20 minutes. e. Kovac’s reagent: Para-dimethylaminobenzaldehyde —~—~—~-_--.—-...----~-- 5 g Amyl alcoho] ~---~------~------------------------- 75 ml Concentrated hydrochloric acid ~----_-----------~------- 25 ml Mix the alcohol and aldehyde and heat in water bath or incu- bator at 50 to 60° C until the aldehyde is dissolved. When cool, add the hydrochloric acid slowly. Store in the dark in a brown bottle with a glass or rubber stopper. f. Beef extract broth base for fermentation tests: Beef @Xttaet, f. McClung-Toabe Egg Yolk agar—formula: : Proteose peptone —---_--_-- 40 g NaHPO: . 7 H2O ______ 5 g Ky POY cece ee lg NaC] _~--___--_-_-- 28 i MSO a ee 0.1 g Glucose ~----_-_-- ee 2 £ POE se 25 g Distilled water ~~~ ------ 1,000 ml Adjust to pH 7.6 and sterilize in the autoclave at 121° C for 20 min. Add 10 ml of sterile egg yolk suspension to each 90 ml of sterile, cooled (50° C) medium. Pour 15 ml per plate and dry plates overnight in the incubator before using. g. Egg yolk suspension: Scrub and sterilize shells by immersing fresh eggs in concen- trated HgCl. solution. Aseptically remove yolks from eggs and place in sterile graduated cylinder. Add an equal volume of sterile saline, mix well. Add 10 ml of egg yolk suspension to each 90 ml of sterile, cooled (50° C) medium. : 2. Procedure: a. Aseptically place duplicate 25-¢ amounts of food into each of 117 ''Laboratory Manual two tubes containing 25 ml of Fluid Thioglycollate broth. Do not fully tighten screw-caps. Incubate the inoculated broth tubes in the 46° C water bath for 4 to 6 hrs. The level of water in the bath should extend above the level of fluid in the tube by at least 1 in. Note: Extended incubation (12 to 24 hours) at 46° C usually results in a reduction of numbers of C. perfringens and overgrowth by concomi- tant flora capable of growing at 46° C. Following incubation, examine tubes for growth. Rapid growth of C. perfringens occurs at 46° C and is accompanied by pro- fuse gas production. Using a loop, remove a portion of turbid broth and streak to the dry surface of McClung-Toabe Egg Yolk plates to obtain well-isolated colonies. Incubate inoculated egg yolk plates anaerobically in Case-Anaero jars at 35° C for 24 hours. See IX, A, 1, b and ¢ and IX, A, 2, c). Observe plates after incubation for presence of circular, slightly raised colonies surrounded by a zone of precipitated egg indica- tive of lecithinase production. Pick representative lecithinase-positive colonies to motility-ni- trate medium and proceed to confirm as C. perfringens as de- scribed in IX, A, 2, f through p. 118 ''Clostridium botulinum X. , Clostridium Botulinum A. Detection of Toxin in Food: 1. Methods, materials, and reagents: a. Standard toxin diluent (Phosphate buffer, 0.1 M, pH 6.5 with 0.2% gelatin). Gelatin and buffer solutions must be prepared separately to prevent flocculation, (1) (2) (3) Gelatin solution (0.4%) add 2.0 g of gelatin to 500 ml dis- tilled water and dissolve by heating. Buffer solution—mix 200 ml of 0.2 M Na,PO, solution with 300 ml of 0.2 M KH.PO, solution. Determine pH of the buffer mixture and adjust to pH 6.5, if necessary, by dropwise addition of dilute HCl or NaOH solutions. Sterilize solutions (1) and (2) separately at 121° C for 15 minutes. As needed, mix aseptically equal volumes of the gelatin and buffer solutions to obtain the standard toxin diluent. b. Mortar and pestle—small size, one for each food specimen ex- amined. c. Sterile sand (approximately 20 g for each food specimen exam- ined). Sm mo Crushed ice. Mice (approximately 20 g weight). Sterile physiological saline solution (0.85% NaCl) —100 ml. Antitoxin for C. botulinum toxins A, B, and E. Refrigerator. i. Pipettes, 1.0 and 10 ml capacities. j. Refrigerated centrifuge. . 2. Procedures: a. Extraction of toxin from food: (1) Suspected sample is aseptically macereted with sterile, chilled sand and as small a quantity of toxin diluent as is practicable in a sterile mortar immersed in ice-water mixture. If possi- ble, the volume of extracting fluid should be no greater than the volume of sample. Overnight storage in the cold of the macerated sample diluent mixture is recommended for opti- mum extraction of toxin. This is not practical, however, where speed of diagnosis is important. 119 ''(2) Laboratory Manual The macerated sample—diluent mixture is centrifuged thor- oughly in the cold (refrigerated centrifuge installed in a cold room) and the clear supernatant is drawn off aseptically. The extract is then diluted 1:10 (1 part extract to 9 parts toxin diluent) for injection and held in the cold until used. If type E toxin is suspected, much of the toxic material may be present in the form of “protoxin” and the extract must be treated with proteolytic enzyme in order to manifest toxicity. To accomplish activation, a 1.0% solution of crude trypsin (Difco 1:250 or equivalent) is added to the extract to give a final trypsin concentration of 0.1%. The mixture is then incu- bated for 60-75 minutes at 37° C. before assays are made. Ac- tivation of both the undiluted centrifugal food extract and a 1:10 extract in gelatin-phosphate buffer have been used with success. b. Mouse protection test: (1) (2) (3) Three to 4 hours before injection of toxic extract, mice of about 20 g weight are divided into groups of at least two mice each. There should be one group for each toxin type, plus two control groups. Each experimental group is “protected” by injection intraperitoneally with no more than 0.2 ml of undi- luted known antitoxin to C. botulinum toxin of types A, B, and E. The two control groups receive the same volume of sterile physiological saline solution in place of antitoxin. Note: Most human cases of botulism are caused by ingestion of Types A, B, or E toxins. Botulism in animals is usually caused by Types C and D. Consequently, when investigating a human case of botulism, it is advisable to test for at least Types A, B, and E toxins. If antisera to Types C and D are available, these may be included. About 3 or 4 hours later, the protected groups and one control group are “challenged” by intraperitoneal injection with 0.5 ml of the 1:10 dilution of food extract supernate. The second control group is challenged with 0.5 ml of the same superna- tant dilution which has been heated at 100° C for 10 minutes to destroy the toxin. These serve to indicate the presence of heat stable, nonbutulinal poisons. If speed is essential, anti- toxin and the supernatant dilution can be mixed and injected together. Animals are observed for 96 hrs following injection. Symp- toms of paralysis, respiratory difficulty, and finally death in- 120 ''Clostridium betulinum dicate botulinus toxicity. Symptoms may appear in hours. Severe symptoms in mice without subsequent death may indi- cate a low level of toxin. (4) Presence of botulinal toxin is determined by deaths in the unprotected group which received unheated toxic supernatant and by survival of the unprotected mice which received heated toxin. The type of toxin is indicated by survival of one of the protected groups. For example, if the group of mice injected with Type A antitoxin survives and the two groups injected with Type B and Type E antitoxins die, then the type toxin produced in the original food was Type A. B. Isolation and Cultivation From Food: 1. Methods, materials, culture media: a. Pipettes, 1-ml capacity. b. Culture dishes (100 x 15 mm) glass. O° (1) Glazed porcelain covers for Petri dishes—obtainable from Fisher Scientific Company, Pittsburgh, Pennsylvania. (2) Brewer aluminum Petri dish covers with absorbent discs— obtainable from Fisher Scientific Company, Pittsburgh, Penn- sylvania. . Water bath, 60° C. . Water bath, 80° C. Inoculating loop, nichrome or platinum-iridium wire B & S gauge No. 26. Case-Anaero jar (See IX, A, 1, b and c¢ above). . Cooked meat enrichment medium: (1) Add 1.25 g of commercially available, dehydrated cooked meat medium to each of a series of 150 x 15mm screw-capped tubes. (2) Prepare a solution containing 0.2% soluble starch and 0.5% glucose in distilled water. (3) Add 10 ml of the solution to each tube of medium and allow to stand until meat particles are thoroughly wetted. (4) Sterilize in the autoclave at 121° C for 30 minutes. Blood agar (Trypticase Soy agar base). Dispense into flasks and sterilize at 121° C for 15 min. Cool to about 45° C and add 5% sterile defibrinated blood. Pour into plates. i. Reinforced Clostridium medium (RCM, Oxoid). This medium is an excellent medium for cultivation of clostridia and can be obtained in the prepared form from Consolidated Laboratories, Inc., Chicago Heights, Ill. T21 ''Laboratory Manual 2. Procedure: a. b. c. og Heat nine tubes of cooked meat enrichment medium in boiling water to drive off oxygen, and cool rapidly in a water bath. Place three tubes of the medium in a 60° C water bath and three in an 80° C water bath and allow to equilibrate to bath tempera- ture. Three tubes are left unheated. Inoculate approximately 1 g (or 1 ml) of the suspected sample into each of the nine tubes below the surface of the cooked meat layer and heat as follows: Those tubes that were preheated to 60° C before inoculation are returned to the 60° C water bath and heated at this temperature for 15 min; those tubes that were preheated to 80° C before inoculation are returned to the 80° C water bath and heated at 80° C for 30 minutes, After heat- treatment, cool tubes rapidly in running tap water. Incubate inoculated tubes at 30° C until vigorous growth is vis- ible, as evidenced by gas production, turbidity, and in some instances partial hydrolysis of the meat particles.* Cultures should be incubated for at least 10 days before they are consid- ered negative for growth. Toxin is usually detectable within 48 to 96 hours and as a rule is in the highest concentration after the period of active growth and gas production. Following incubation, inoculate the surface of blood agar and/or RCM agar plates with culture material from the tubes of cooked meat medium. Attempt to obtain well-isolated colonies.** Incubate plates at 30° C in a Case-Anaero jar, under an atmos- phere of 90% N. and 10% CO,. Colonies usually appear within 1 to 3 days, depending upon the strain and medium employed. Though growth may be more rapid at 35° C for most strains, some Type E strains will not multiply at this temperature. For this reason, 30° C is recommended. Pick isolated colonies from plates and inoculate tubes of cooked meat medium. If desired, these isolates may be tested for tox- icity as already described. *Note: If pure-culture isolates are desired, streak plates should be made at this time, according to instructions in steps (e), (f), and (g) below. If not, proceed immediately to step (h). **Nore: C. botulinum displays a tendency to yield a “spreading’ ’type of growth on agar plates. To reduce spreading, the following techniques have been used successfully: (a) Increase the agar concentration of solid media to 6%; (b) thoroughly dry plates before inoculation (not reliable if much condensation develops in anaerobic jars); (c) use of absorbant type Petri dish covers as described in (X, B, 1, b) above. 122 ''Clostridium botulinum h. Test cooked meat medium tubes inoculated with food sample for toxicity, as described under Mouse Protection Test (X, A, 2, b) above. If heated cultures (60° C for 15 minutes and 80° C for 30 minutes) are negative for toxin, test unheated cultures for toxicity. As a rule, if C. botulinum spores were present in the food, the heated tubes inoculated with this food will show toxic- ity as a result of germination and reproduction, If spores were not present, however (highly improbable), the heat treatments would be sufficient to destroy vegetative cells of C. botulinum and these tubes would be negative for toxin by the mouse test. For this reason, unheated tubes are also employed. Providing the vegetative cells of C. botulinum are not overgrown by con- comitant bacteria, the unheated cultures usually develop toxicity. 123 ''''ae =< ila “ i. aU a “ ''This publication names representative products for identification only, and listing does not imply endorse- ment by the Public Health Service and the U.S. De- partment of Health, Education, and Welfare. Public Health Service Publication No. 1142 March 1964 For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C., 20402. Price 50 cents . ''028730564 ''''