A STUDY ON THE AAODA STEON 0A AAAAA A AAA 8 - .‘ ANAEROAAXLN REAALAmA AAAM THE AAAW‘AA AF SAAAAALococws AUAAAA AA CHEESE ANA BRAIN HEART MAL-AAA AA AAA Thesis for AAA Degree of M. a A‘EICHECM STATE UMVERSEIY LARRY LEE H000, £963 [.1812 A R Y Michigan 5‘3” UmVCISltY A 1/ 9 LIE/[1&7 WI? A” A ABSTRACT PRODUCTION OF TYPES A AND B ENTEROTOXIN BY STAPHYLOCOCCUS AUREUS IN CHEESE AND IN BRAIN HEART INFUSION BROTH by Larry Lee Hood Cheddar and colby cheese were manufactured from normal pasteur- ized milk with and without inoculations of strains of Staphylococcus aureus producing types A and B enterotoxin. The above varieties of cheese were also made from inoculated milk to which penicillin had been added at the rate of approximately 0.49 to 0.54 Units per milliliter of milk. One of the Cheddar cheeses and one of the colby cheeses manufac- tured from milk treated with penicillin and inoculated with a strain of §. gggggs producing type A enterotoxin was found to contain enterotoxin during ripening at 48 F. The toxic Cheddar cheese had a maximum §..gg£§g§ population of llO million organisms/g of cheese after 30 days of curing. Toxin was formed in this Cheddar cheese between 20 and 30 days of ripening. The maximum amount of toxin de- tected in the cheese was 4 pg/IOO g of cheese after 30 days of curing. The toxic colby cheese had a maximum staphylococcal population of 6.5 million organisms/g of cheese after 20 days of curing. Toxin was formed in the colby cheese between I and I0 daws of ripening. The maximum amount of toxin detected in the cheese was 2 ug/IOO g of cheese. No toxin was detected in the cheese manufactured from milk inoc- ulated with the strain of §. aureus producing type B enterotoxin. Larry Lee Hood However, one of the Cheddar cheeses manufactured from milk containing penicillin and inoculated with the §._§g£§g§ producing type B toxin, contained 480 million staphylococci per gram of cheese after 60 days of curing. The type A producing strain of §._ag£eg§ produced maxima of 8, 8 and 6 pg of toxin/ml of supernatant in aerated brain heart infusion (BHI) with respective pH values of 6.0, 7.4, and 8.0. Toxin was first detected in the BHI broth at pH 7.4 when the §. gurggs popu- lation was 270 million organisms/ml. This population coincides reasonably well with the maximum §.-gg£§g§ population that occurred during ripening of the toxic Cheddar cheese, but is substantially greater than the 6.5 million staphylococci/g in the toxic colby cheese. The type B toxin producing strain of_§..§g£gg§ produced maxima of 382, 382, and 32 pg of toxin/ml of supernatant in aerated BHI broth with respective pH values of 6.0, 7.4 and 8.0. Toxin was first detected in the BHI broth at pH 7.4 when the §.'gg£§g§ population was 40 million organism/ml. No explanation is offered as to why toxin was not detected in the Cheddar cheese made from milk inoculated with the type B enterotoxin producing strain of §._§g£gg§ and which attained 480 million staphylococci/g during ripening. A STUDY ON THE PRODUCTION OF TYPES A AND B ENTEROTOXIN RESULTING FROM THE GROWTH OF STAPHYLOCOCCUS AUREUS IN CHEESE AND BRAIN HEART INFUSION BROTH BY LARRY LEE HOOD A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science I968 ACKNOWLEDGMENTS The author most sincerely thanks Dr. L. G. Harmon for his guidance and patience during this graduate program and for his aid in preparing this manuscript. Thanks are continued to Dr. E. P. Casman for his generous attitude in allowing the author to visit his laboratory and for suppling the gel-diffusion reagents used in this study. The author wishes to acknowledge Dr. R. V. Lechowich of the Food Science Department and Dr. H. A. Lillevik of the Biochemistry Department who served on the committee and for their suggestions in editing this manuscript. Appreciation is also extended to Dr. B. S. Schweigert, Chairman, Food Science Department, for his interest in this program and for providing the Public Health Services traineeship which financially supported the project. Finally, I shall be forever indebted to my fiancee, Miss Catherine Harris, for her understanding and encouragement during these past few months. This thesis is dedicated to Mom and Dad, whose sacrifices the author has truly appreciated. L.L.H. iii CONTENTS Page INTRODUCTION ............................... . ................... I REVIEW OF THE LITERATURE ........................ . ............. . 3 History .. ..... . .............................. ..... ..... ... 3 Production of Enterotoxin ........ ............. ...... ...... II Effect of temperature .................. ........ ...... II Effect of pH ... ..... . ........ ...... ...... ............ I2 Time of incubation and toxin production in food ..... . l3 Competition of staphylococci with other bacteria ..... l3 Amount of enterotoxin and number of staphylococci necessary to induce food poisoning ................. l5 Staphylococcal food poisoning from cheese ............ I6 Detection of Enterotoxin . ..... .. ...... .................... I8 Biological methods ..... .............................. l8 Immunological methods ................ ... ...... ....... 2] Indirect bacteriological methods ... ........... . ...... 26 MATERIALS AND METHODS ........ . .............. . ...... . ......... .. 28 Gel-diffusion Reagents ..... .............. . ............. ... 28 Staphylococcal Strains ............................... ..... 28 Culture Media ........................ ..... ......... ...... . 28 Pilot Plant EXPeriments ................................... 29 Enumeration of Organisms .................................. 32 Analysis of Cheese for Enterotoxin , ,,,,,,,,,, ,,,,,,,,,,,,, 32 Recovery of Enterotoxin Added to Non-toxic Cheese ooooooooo Relationship of Growth and Toxin Production by Toxigenic Strains of §. aureus in 3.7% BHI with Different pH Values ..... RESULTS AND DISCUSSION .............. . ........ . ........ ......... Recovery of Enterotoxin Added to Non-toxic Cheese ......... Pilot Plant Experiments ooooooooooooooooo o ooooooo a... ccccc 0 Effect of Penicillin Upon pH, Lactic Acid Bacteria and Staphylococcal Populations in Cheddar Cheese ...-.... Effect of Penicillin Upon pH, Lactic Acid Bacteria and Staphylococcal P0pulations in Colby Cheese ..... ..... Production of Enterotoxin in Cheddar and Colby Cheese Manufactured from Milk Inoculated with TOXIgenIC Strains Of §o aureus 00000000000000.0000ooooooo Production of Enterotoxin in BHI Broth at Differing pH Values 0000000000000000 o 000000000000 oooooooooooooooooo SUMMARY AND CONCLUSIONS ........................... . .......... .. LITERATURE CITED ........ 42 43 46 46 49 50 58 66 74 89 9I TABLES Table Page 1 Amounts of penicillin and toxigenic strains of Staphylococcus aureus added to vats of milk made into Cheddar and colby cheese. . . . . . . . . . . . . . . . . . 31 2 Recovery of types A and B enterotoxin added to 100 3 samples of non-toxic colby cheese . . . . . . . . . . . . . 47 3 Percent acidity of the whey at milling during the manufacture of Cheddar cheese . . . . . . . . . . . . . . . 51 4 Percent acidity of the whey at dipping during the manufacture of colby cheese . . . . . . . . . . . . . . . . 60 5 Enterotoxin produced during curing at 48 F of Cheddar cheese manufactured from pasteurized milk inoculated with toxigenic strains of §, aureus and containing penicillin as indicated. . . . . . . . . . . . . . . . . . . . . . . . 67 6 Enterotoxin produced during curing at 48 F of colby cheese manufactured from pasteurized milk inoculated with toxigenic strains of §, aureus and containing penicillin as indicated. 0 0 O O O O O O O O O O O O O O O O O O O O O 70 7 Time, optical density at 620mg, viable count as determined on S-110 medium, pH of the supernatant, and enterotoxin concentration when §, aureus producing type A enterotoxin was grown in 3.7% BHI broth at pH 7.4 (non-buffered) and incubated at 37 C. No= 51,000/m1 . . . . . . . . . . . . . 76 8 Time, optical density at 620mg, estimated viable count, pH of the supernatant and enterotoxin concentration when §, aureus producing type A enterotoxin was grown in 3.7% BHI broth, buffered at pH 6.0 with 0.2 M sodium ph08phate and incubated at 37 C. No= 8,000/m1. . . . . . . . . . . . . . 78 9 Time, optical density at 620mg, estimated viable count, pH of the supernatant and enterotoxin concentration when g, aureus producing type A enterotoxin was grown in 3.7% BHI broth, buffered at pH 8.0 with 0.2 M tris and incubated at37C.No=8,000/m1.................o. 8]- 10 Time, optical density at 620mg, viable count as determined on S-llO medium, pH of the supernatant, and enterotoxin concentration when.§, aureus producing type B enterotoxin was grown in 3.7% BHI broth at pH 7.4 (non-buffered) and incubated at 37 C. No= 31,000/m1 . . . . . . . . . . . . . 83 vi 11 12 vii Time, optical density at 620mg, estimated viable count, pH of the supernatant, and enterotoxin concentration when §, aureus producing type B enterotoxin was grown in 3.7% BHI broth, buffered at pH 6.0 with 0.2M sodium phOSphate and incubated at 37 C. NO: 270,000/ml . . . . . . . . . . . . .. 86 Time, optical density at 620mu, estimated viable count, pH of the supernatant, and enterotoxin concentration when §, aureus producing type B enterotoxin was grown in 3.7% BHI broth, buffered at pH 8.0 with 0.2 M tris and incubated at 37 CO No: 31,000/m1 O O O I 0 O O O I O O O O O O O O O O O 88 Figure 10 FIGURES Extraction of enterotoxin from cheese. Diagram of slide and accessories for gel-diffusion . Total and §, aureus populations and pH of Cheddar cheese manufactured from pasteurized milk containing (A) no penicillin or (B) 0.54 Units of penicillin/m1 (cured at 48 F) . O O O O C O O O O O O O O O O I O O I O O O O O 0 Total and §, aureus populations and pH of Cheddar cheese manufactured from pasteurized milk (A) inoculated with 480,000 type A enterotoxin producing fi, aureus cells/m1 or (B) inoculated as above and containing 0.54 Units of penicillin/ml (cured at 48 F). . . . . . . . . . . . . . Total and §, aureus populations and pH of Cheddar cheese manufactured from pasteurized milk (A) inoculated with 450,000 type B enterotoxin producing §, aureus cells/m1 or (B) inoculated as above and containing 0.49 Units of penicillin/ml (cured at 48 F). . . . . . . . . . . . . . Total and g, aureus populations and pH of colby cheese manufactured from pasteurized milk containing (A) no penicillin or (B) 0.49 Units of penicillin/m1 (cured at 48 F) O O O O 0 O O O O O O O O I O O O O O O O O O O I 0 Total and §, aureus populations and pH of colby cheese manufactured from pasteurized milk (A) inoculated with 440,000 type A enterotoxin producing g, aureus cells/ml or (B) inoculated as above and containing 0.49 Units of penicillin/m1 (cured at 48 F). . . . . . . . . . . . . . Total and g, aureus populations and pH of colby cheese manufactured from pasteurized milk (A) inoculated with 450,000 type B enterotoxin producing §, aureus cells/m1 or (B) inoculated as above and containing 0.49 Units of penicillin/m1 (cured at 48 F). . . . Relationship between optical density when cells of §, aureus that produce grown in non-buffered 3.7% BHI broth incubated at 37 C. . . . . . . . . . Relationship between optical density when cells of g, aureus that produce grown in non-buffered 3.7% BHI broth incubated at 37 C. . . . . . . . . . viii and population type A toxin are at pH 7.4 and and population type B toxin are at pH 7.4 and Page 34 39 52 55 57 59 63 65 79 84 INTRODUCTION The occurrence of staphylococci as common residents of the skin and in the suppuration of wounds and open sores on the body is unquestioned. Also the wideSpread relationship between food poisoning and the growth of certain strains of Staphylococcus aureus has been thoroughly investigated by many workers (Barber, 1914; Dolman, 1934; Dack, 1956; and Casman and Bennett, 1963, 1965). Early investigations on the nature of the toxic principal resulting from the growth of §, aureus by Dolman g; al. (1936) has lead to the purification of several distinct toxic proteins, which in turn has resulted in the application of serological methods in detection of staphylococcal enterotoxin. The amount of enterotoxin necessary to cause illness in humans is not known, nor has much information been gathered on the amount of enterotoxin that can result from the growth of toxigenic strains of g, aureus in foods. Very little is known presently of the re- lationship between time and the population necessary for the appearance of enterotoxin in food. Dairy products have not generally been implicated in staphylococcal food poisoning incidents, but recently ice cream, milk powder and both raw and pasteurized milk cheese have been found to contain enterotoxin. The purpose of this research was to enlarge our knowledge of the growth and production of toxin by certain strains of g, aureus in Cheddar and colby cheese. A supporting investigation was conducted to see what effects pH of the medium had upon growth and toXin production by g, aureus 1 in aerated brain heart infusion broth. One of the strains used in this research produces type A toxin while the other produces type B toxin. REVIEW OF THE LITERATURE History Partially because of the ubiquity of staphylococci, their role in food poisoning was not recognized until l9l4. However, descriptions of food poisoning cases given by medical workers, even before Pasteur found staphylococci in pus in I880, are evidence that staphylococci undoubtedly had caused food poisoning for centuries (Dack, I956). Barber in l9l4 (Dack, I956) associated the etiology of an explosive, violently nauseating, selectively occurring food poisoning with the growth of a single strain of staphylococcus. Barber's report was overlooked until Dack_§tial. (I930) rediscovered the role of staphylococci in food poisoning. These workers not only isolated staphylococci from incriminated food, but also reproduced the illness by feeding culture filtrates to human volunteers. Jordan £3.El° (l93l) demonstrated the heat resistance of the toxic principle resulting from the growth of staphylococci, by boiling culture filtrates for 30 minutes and subsequently inducing food poisoning symptoms in humans by feeding them the filtrates. These early studies indicated that the toxic agent was present in culture supernatants as an exotoxin and because the predominant symptoms of the food poisoning illness were gastroenteric, the toxin was called enterotoxin. Since the ”modern era” of staphylococcal food poisoning investi- gations began in I930, the search for an adequate enterotoxin assay -3- has turned from human volunteers to animal tests and more recently, to biological and chemical methods. The earliest of the animal assays was the monkey feeding test of Jordan and McBroom (l93l). However, Dolman_gt_21. (I936) sought to circumvent the difficulty of handling and cost of the young monkeys used by Jordan and McBroom and so developed the ”Dolman Kitten Test“ for enterotoxin. During the development of the kitten test, Dolman found evidence that enterotoxin might be antigenic in nature, suggesting the possibility of an antigen-antibody assay for the toxin. Further proof of the antigenic nature of enterotoxin was offered by Dolman and Wilson (I938) when they demonstrated a specific flocculation reaction which appeared to involve enterotoxin and its homologous antibody. However, development of an antigen-antibody assay for enterotoxin was delayed until enterotoxin could be purified. Therefore, because of the lack of a better assay, workers were forced to depend upon animal tests for enterotoxin for many years. Such tests were inaccurate due to varying levels of susceptibility to the enterotoxin within a group of test animals, non-Specific emetic responses during a given test, and increasing resistance to the enterotoxin by repeated exposure. Scientists in the area of enterotoxin research always recognized the need for a pure enterotoxin preparation around which to build a chemical assay and for various study purposes. Early attempts by Jordan and Burrows (I933) to concentrate the toxin by dialysis were successful. Following the suggestion of Dolman that the enterotoxin was antigenic and therefore a protein, many people turned to the application of protein fractionation techniques in order to purify the toxin. Davidson and Dack (I939) attempted to purify enterotoxin by ammonium sulfate precipitation. Hammon (l94l) suggested the use of a combination of ammonium sulfate and ethanol precipitation to concen- trate the toxin. An acid precipitation technique was used by Bergdoll £3 21' (I9SI) in a successful attempt to concentrate the toxin from bacterial filtrates. While these attempts were being made to purify the toxin, bacterial physiologists were working on methods to rapidly indicate the presence of enterotoxin in food by bacteriolog- ical methods rather than depend upon inferior animal assays. Information gathered on the physiology of the staphylococcal organism, revealed that staphylococci have the ability to ferment mannitol with acid production, produce a gelatinase, and grow well in the presence of high sodium chloride concentrations. But of greater importance was the finding that certain strains of Staphylococcus aureus associated with food poisoning incidents could also produce the enzyme coagulase. The work of Evans and Niven (I950), and Evans, Buettner, and Niven (I950) showed that the enterotoxigenic strains of .§'.§ELEE§ appeared to be a very homogeneous group. 0f the strains tested, they found that the enterotoxigenic strains produced coagulase, but not all coagulase-positive strains produced enterotoxin. 0n the other hand, they reported that all coagulase-negative cultures studied failed to produce enterotoxin. They suggested that the coagulase test could be a rapid indication of staphylococcal food poisoning when coagulase-positive strains could be isolated from a suspect food. The coagulase test has remained an important indirect determination for enterotoxin since the work by Evans, Buettner, and Niven until recently when Bergdoll (I967) proved the existence of several coagulase-negative, enterotoxin producing strains Of.§'.§££22§° The test has now assumed a somewhat less important role in the identification of enterotoxin producing strains of_§._§3£eg§ in favor of the more specific serological tests. The production of heat labile hemolysins by the growth of ‘§. aureus was recognized (Surgalla and Hite, I945) but not until I953 did peOpIe learn that the organism might also produce more than one distinct enterotoxin (Surgalla 95 21' I954; Matheson and Thatcher, I955). Through the use of gel-diffusion analysis Surgalla £3.21' (I954) attempted to demonstrate conclusively that staphylococcal enterotoxin was a specific antigen. Although they failed to detect the specific antigen they were able to neutralize toxic material completely. However, Casman (I958) was able to demonstrate the antigenicity of enterotoxin and also lg 3139 protection of cats against the toxin with both heterologous and homologous strains. A second aspect of the pathogenicity of enterotoxin other than as a causative agent in food poisoning was investigated by Surgalla and Dack in I955. These workers were able to isolate several entero- toxigenic strains of_§. aureus from fecal cultures of a few indiv- iduals experiencing gastroenteritis while on antibiotic therapy. Improved fractionation procedures for enterotoxin from strain 5-6 of §. aureus were reported by Bergdoll £5 21. (I959) and Hibnick and Bergdoll (I959). Included in the procedure was a combination of acid precipitation, alumina absorption, alcohol precipitation, Amberlite lRC-SO column chromatography, and finally zone starch electrophoresis to determine purity. Single and double gel-diffusion tests were used to follow the purification process. The chemical properties of the purified preparation were also reported. They found that the toxin was a protein with no indication of the presence of carbohydrate or lipid, had a molecular weight of 24,000 i 3,000, and an isoelectric point of 8.6. An earlier publication by Bergdoll (I956) gave evidence that the toxin was resistant to trypsin and contained a high percentage of lysine. Bergdoll 25 21' also in I959, using gel- diffusion techniques and highly purified enterotoxin (presently called type B), demonstrated the presence of a specific precipitating antibody with an enterotoxin neutralizing property. The identification of this antibody provided a basis, for the first time, for an_ig_li££9 assay of staphylococcal enterotoxin. Using the more relaible serological method for studying entero- toxin, Sugiyama gt 21. (I960) reported a difference in the levels of toxin produced by different subcultures of a single §. agrggs strain and also gave evidence for at least one more antigenically distinct enterotoxin resulting from the growth of strain l96-E and differing from that produced by strain S-6. Casman (I958, I960) prepared antiserum for the new type toxin (l96-E) using a technique of specific absorption with certain non-enterotoxigenic cultures known to share antigens with the new type. Casman also reported that there were strains of_§. aureus which (a) reacted with both S-6 and I96-E antisera, (b) reacted only with 5-6 antiserum-strain 243, (c) reacted only with I96-E antiserum and finally, (d) showed no reaction with either antiserum but were toxic to cats. This toxicity was lost however, when the culture filtrates were heated in boiling water for 30 minutes or when dialyzed through cellophane. Strains producing the type l96-E, heat-resistant enterotoxin were of food origin and Casman designated such enterotoxin as type ”F“. The heat-resistant type S-6 enterotoxin produced by strain 243 was related to strains associated with enteritis from antibiotic therapy and was designated as type “E“. Bergdoll_§t_§l. (I962) reported that possibly a third type of enterotoxin existed, resulting from the growth of strain I37, since its toxin did not cross react with antisera from either strain S-6 or I96-E. He also reported an increase in the recovery and purification of type E toxin and a 70% purification of type F toxin by using Amberlite XE-64 chromatography. Dack (I962) reviewed the staphylococcal enterotoxins with respect to types, production, resistance to heat, chemistry and mode of action in the human and animal bodies. He recognized three types of enterotoxin: one associated with food poisoning (type F); one produced by strains of §. EEEEEE associated with enteritis (type E); one immunologically unrelated to either of the above (type I37). These were described as distinctive proteins, and unlike Iysins, resistant to pepsin and trypsin. A new method of designating the staphylococcal toxins was sug- gested by Casman gt 21' (I963) using the capital letters of the alpha- bet. With this nomenclature the previous type l96-E or type F would be designated as type A with the prototype strain being I96-E. The type 5-6 or type E became known as type B with the prototype strain being 243. Recently, the serological identification of type C enterotoxin from strain I37 (Bergdoll gt 21' (I966)) and type D enterotoxin from strain 494 (Casman £3 21' (I967)) has also been reported. The designation of additional enterotoxins will follow alphabetically when they are established as immunologically distinct entities. Casman and Bennett (I963) reported that, at least in culture media, type B strains usually produced more toxin than type A strains. Because type A strains produced very low amounts of enterotoxin, they suggested growing the cultures in dialysis tubing placed in contact with the media for pure culture studies. The growth and production of enterotoxin by the type A strain in raw and cooked meat was studied by Casman 2; 21° in I963. To detect the low amounts of enterotoxin produced they used carboxymethyl cellulose in column chromatography for the extraction and concentration of the toxin. The concentrated toxin was then identified with a microscope slide version of the Ouchterlony gel-diffusion assay. Other workers have since reported on the extraction and identification of either type A or B enterotoxin from foods (Hall §;_gl., I963; Read EE.31°’ I964; Genigeorgis, I966; and Zehren and Zehren, I968) with varying success. Most methods used to detect enterotoxin in food have included blending to separate the solid material from the water soluble toxins, concentration of the toxin by dialysis, and ion-exchange chromatography and gel-diffusion methods for quantitation of the toxin -10- levels. Casman (I966) also reported on the frequency with which the different types of enterotoxins were produced by coagulase-positive staphylococci isolated from several different sources. 0f the staphylococci isolated from foods implicated in seventy-five different food poisoning incidents, 49% produced enterotoxin A alone and 28.5% produced enterotoxin A together with other enterotoxins. Enterotoxin D was produced alone by I0% of these strains and together with enterotoxin A by approximately 25% of the strains. The incidence for each of types B and C was approximately 4%. Thermal stability of staphylococcal enterotoxin was discussed by Jordan in I93l, by Denny_§t_al. (I966), and Read and Bradshaw (I966). Denny reported F values (2 = 48 F) of II minutes and 8 250 minutes for the inactivation of type A toxin as assayed by intraper- itoneal injection of cats and the monkey feeding test, re5pectively. Read and Bradshaw reported F250 values (2 = 32.4 F) of 9.9 minutes for type B enterotoxin as assayed by both gel-diffusion and cat emetic responses. They also gave some evidence for an increase in heat stability of the toxins if they were in a crude protein preparation rather than highly purified. The increase in use of gamma radiation preservation of foods in recent years has also prompted a study of the stability of type B enterotoxin to radiation. Read and Bradshaw (I967) determined D values for inactivation of toxin in Veronal buffer and milk to be 2.7 and 9.7 Mrad, respectively. End points for inactivation were deter- mined by parallel titrations with gel-diffusion methods and cat emetic -1]- responses. It can be concluded on the basis of the data presented in these studies on the thermal and radiation treatments of entero- toxin that the current thermal processing times and temperatures are sufficient to inactivate any existing type A or type B enterotoxins, unless they are afforded some unknown protection by the food or occur in extremely high concentrations. However, Read and Bradshaw in- dicated that the irradiation processes now used for pasteurization or sterilization of foods cannot be expected to inactivate type B toxin if it were present in food before processing. Production of Enterotoxin Very little has been reported on the conditions under which toxigenic strains of staphylococci produce enterotoxin. Due to the lack of an easy assay for this toxin, the bulk of the reported studies deal with the growth of staphylococci in culture media or food, under a variety of environmental conditions. Appreciable levels of entero- toxin are produced only after considerable growth of the staphy- lococcus. Usually a population of at least several million per milliter or gram must be attained. Therefore, the conditions that favor toxin production are those best for growth of the staphylococcus (Frazier, I967). The effect of temperature, pH, competing organisms, and the number of staphylococci needed to initiate growth and toxin production will be reviewed. EEEEEE.QI Temperature. Segalove and Deck (I94I), under laboratory conditions, demonstrated enterotoxin production in a culture grown for 3 days at I8 C and in one grown for l2 hours at 37 C. Cultures grown -12- for shorter periods did not contain enterotoxin as assayed by the kitten test. They also found that enterotoxin was not demonstratable in cultures grown 3 and 7 days at 9 and IS C, respectively. According to Frazier (I967), toxin production is best at 2I.l to 36.] C, a temperature range which easily falls within that required for the best growth of_§._ag£gg§. Casman and Bennett (I963) reported maximum toxin production for type A and B enterotoxigenic strains of §._§g£§g§ when incubated at 37 C for 48 hours on Brain Heart Infusion (BHI) agar at pH 5.3. Casman gt 21. (I963) have also reported the production of type A enterotoxin when organisms were incubated on disks of ham held at 30 C for 72 hours. Dack (I956), Hendricks _t _1. (I958) and Zehren and Zehren (l968b) all have identified cheddar cheese in food poisoning outbreaks, indicating the development of toxin from the growth of §-.22£22§ during cheese manufacture and curing. During manufacture and processing cheddar cheese is at a temperature of 80 to l02 F and during ripening or curing the cheese is usually stored at 40 to 55 F for 2 to I2 months. EUZEEEHQI.EU' Lechowich gt 31' (I956) found staphylococci able to grow vigorously in ground pork muscle containing any combination of curing ingredients permissible and palatable. However, if the pH of the meat was lowered to 4.8 to 5.0, anaerobic growth of staphylococci was prevented. In a broth medium, growth was inhibited at pH 5.6 anaerobically, and aerobically growth was inhibited at pH 5.6 and prevented at pH 4.8. Deck and Lippitz (I962) reported that staphylococci inoculated -13- into slurries of frozen pot-pies grew well at pH 5.0 and higher. They grew slightly at pH 4.5 but not at pH 4.3 to 4.0. Casman and Bennett (I963) found that a semisolid BHI agar at pH 5.3 to 5.5 gave higher yields of type A toxin than BHI agar at pH 6.0, 6.5, 7.0 and 7.5. With 3.7% BHI broth, they found maximum toxin when the initial pH was 6.5 to 7.0. This seems to indicate that toxin production depends upon more than one environmental factor. For instance, the effect of sodium chloride and pH upon the production of type B enterotoxin was studied by Genigeorgis (I966). He found that with the same initial pH, the growth rate and toxin production by .§°.EE£EE§ decreased as the sodium chloride concentration increased. Enterotoxin B was detected in six day old BHI broth cultures in which the concentration of salt was I0%. Time 2f incubation and toxin production 13 food. The shortest period of time necessary for toxin production in foods in not known. Only a small number of experiments have been conducted in the past (Dack, I956) because of a lack of an easy assay method. Experiments with human volunteers showed that cream became toxic after 5 hours at room temperature, mashed potatoes and milk after 6.5 hours at room temperature, canned oysters after 72 hours at 37 C, canned corn after 96 hours at 37 C, and home-baked ham and lettuce sandwich after 72 hours at 37 C. Competition 9f staphylococci with other bacteria. The poor ability of staphylococci to compete with other organisms present in foods has long been recognized (Casman and Bennett, I963). Bacterial -14- inhibition of staphylococcal growth and toxin production is mainly due to a competition for essential nutrients, or production of antibiotic or other inhibitory substances. Dack and Lippitz (I962) inoculated slurries of frozen pot-pies with varying numbers of staphylococci. In the presence of the natural flora, staphylococci usually did not grow sufficiently to suggest a hazard in such foods from staphylococcal food poisoning. The predominant organism, a Iactobacillus, produced considerable acid in the medium. Such was also the case in a recent, rather large outbreak of staphylococcal food poisoning occurring in cheddar cheese. Epidemiology of the cheese revealed toxin production only in those cases where there was low acid production during cheese manufacture (Zehren and Zehren, l968b). Normal acid production in cheese manufacture is the result of good growth of the starter organisms in the milk. Staphylococcal food poisoning appears to occur in foods which have been treated to drastically reduce the bacterial population or in foods which selectively favor the growth of staphylococci; for example, cured hams which are pasteurized and con- tain a high salt concentration. With such a reduction of natural flora and inhibition of non-salt tolerant bacteria, staphylococci grow luxuriously and produce toxin. This may explain why ham is the most commonly implicated food item in staphylococcal food poisoning (Brandly, I965). Casman and Bennett (I963), working with raw and cooked meat inoculated with a type A enterotoxin producing strain of .§'.2££SE§: have determined that there was no growth of_§._gg£gg§ in the raw meat but good growth and toxin production in the cooked. This Shows -15- again the inhibitory effect of the competitive bacteria in the raw meat. Finally, staphylococcal enteritis which appears in patients under oral antibiotic treatment can result from a lack of microbial com- petition. Antibiotic resistant staphylococci in this case survive the antibiotic treatment and grow and produce toxin after the antibiotic sensitive competitors have been eliminated from the digestive tract. .Amgggt_gf enterotoxin and number 2f staphylococci necessary £9 induce food poisoning. The amount of enterotoxin necessary to induce illness in man is not known. Purified (50%) type A enterotoxin was found by Casman and Bennett (I965) to have a toxicity level of l to 2 pg in one cat test. A highly purified type B enterotoxin (Bergdoll, I961) produced emesis in rhesus monkeys when administered orally in amounts of I pg of nitrogen. Casman and Bennett (I965) reported the following foods and numbers of staphylococci responsible for food poisoning out breaks: coconut cream pie and banana cream pie with 200 million and 72 to 73 million staphylococci/g producing type A enterotoxin, respectively; and macaroni salad with l million organisms/g producing type A toxin. Experimentally, they found the production of 0.2, l4 and 8-9 pg of type A toxin per gram of food in coconut custard, shrimp paste, and turkey paste with 2.5, l5 and l3.5 billion staphylococci/g of food, respectively. Hall gt 21' (I963) found detectable levels of type B enterotoxin (4 to 68 pg/ml) were produced consistently in slurries of shrimp, -I6- scallops, lobster, and crabmeat containing IIO million to l billion organisms/ml. This number was reached after 24 to 48 hours incubation at 35 C following inoculation with one milliter of an I8 hour culture. Staphylococcal food pgisoning from cheese. Staphylococcal food poisoning is presently considered to be the most common type of food-bourne disease in the United States (Dauer, I961). Even though food poisoning in general is poorly reported, many different types of food have been incriminated in staphylococcal food intoxications. Very few food poisoning cases have been attributed to milk and dairy products in recent years due to improved supervision in sanitary production and processing of milk and dairy products. Furthermore, of the cases due to dairy products, those reported as being due to cheese are uncommon (Hendricks gt 31., I959). There have been a few, however, and some of these will be reviewed. Vaughn (I884) reviewed 300 cases of poisoning that supposedly resulted from Michigan cheese of a cheddar type that was eaten by many persons over a six month period. Staphylococcal food poisoning was not recognized at that time and he concluded that the causative agent was a chemical poison; however, he did state that the chemical poison might have been generated by the agency of bacteria. Also in I884, Dr. Sternberg at the Johns H0pkins University was able to recover micrococci from samples of the su5pect Michigan cheese investigated by Dr. Vaughn. Sternberg concluded that the food poisonings were probably the result of the production of a poisonous ”ptomaine” as the result of the growth of the micrococci. Dack (I956) -17- describes the work of a man named Barber who, in I9l4, attributed several acute gastro—intestinal upsets on a certain farm in the Philippine Islands to the growth of a white staphylococcus in one of the udders of the family cow. In I930, three separate outbreaks of food poisoning involving l8 persons who had eaten cheese occurred in Puerto Rico. Staphylococci were recovered from the cheese in each outbreak. Filtrates of the organisms isolated from two of the outbreaks were given to human volunteers and produced symptoms similar to those of the original cases (Hendricks, 23.21., I959). MacDonald (I945) described four severe cases of staphylococcal food poisoning in Great Britain from home made goat’s milk cheese. .§‘.EEEEH§ was recovered from the cheese and from freshly drawn milk from one of the goats. Hendricks g3 a1. (I959) reported an outbreak of food poisoning occurring in 200 persons as the result of eating Cheddar cheese manufactured from raw milk. Coagulase-positive §°.EELEE§ organisms were isolated from both the cheese and from the milk of some of the herds that supplied milk to the factory that produced the cheese. The most recent instance of staphylococcal poisoning resulting from cheese is the case discussed by Zehren and Zehren (l968a) in which 2II2 vats of Cheddar, Monterey and Kuminost cheese were implicated in an unknown number of widespread food poisonings. Examination showed 56 vats of Cheddar, two vats of Monterey and one vat of Kuminost contained type A enterotoxin and were destroyed. The remainder of the cheese implicated was released for sale and no reports of additional illnesses have occurred. -13- Detection of Enterotoxin Because staphylococcal food poisoning is an intoxication, researchers have worked to develop a reliable, sensitive and inexpensive test for the detection of enterotoxin in food poisoning outhEaks and in culture media. The first attempts were directed toward methods involving the use of sensitive animals which re3pond in a characteristic and reproducible way to the enterotoxin. When the nature of enterotoxin was further elucidated and more data was obtained abouts its properties, a new area in the methodology of enterotoxin detection opened. This was mainly due to the demonstration of its antigenicity and pre- cipitation by Specific antibodies. The techniques tried and developed through the years can be divided into three major groups. The first group involves purely biolobical methods in which live animals are used. The second group involves immunological methods based on the antigenicity of enterotoxin and the reaction with its antibody. The third group involves methods which demonstrate certain staphylococcal characteristics which later are correlated with enterotoxicity. These three major groups involve biological, immunological and indirect bacteriological methods. Biglogical_methods. Testing with human volunteers is considered best because humans are highly sensitive to the enterotoxin. However, variations in the response of different individuals, naturally occur- ring resistance against enterotoxin, and the difficulty in recruiting volunteers complicates tests using humans. -19- The monkey feeding test was developed by Jordan and McBroom (I93l) and is still in use. Culture filtrates or food extracts are made to 50 ml volume and are fed by stomach tube to young Rhesus monkeys (Macaca mullata). The animals are then observed continuously for 5 hours. Of the symptoms caused by ingestion of enterotoxin, only vomiting is accepted as a positive reaction. Because monkeys develop tolerance to enterotoxin, it is recommended that six animals be used per sample of which at least two should react to consider the sample positive for enterotoxin. Although alpha- and beta-hemolysins have been shown to provoke emesis when injected into monkeys and other animals, both toxins are sensitive to the proteolytic enzymes of the gastro-intestinal tract and no special treatment is needed to eliminate these toxins from the sample. However, should the test be performed by intravenous injection it is necessary to destroy the hemolysins by boiling the filtrates from the.§. aureus culture for 30 minutes, or by neutralization with specific antiserum or by adding a proteolytic enzyme such as trypsin which does not affect enterotoxin. The kitten test was developed by Dolman 35 al. (I936) who suggested the use of kittens from 6 weeks to 3 months of age and weighing between 350-700 grams. The toxic material, culture filtrate, or food extract is boiled for 30 minutes to destroy hemolysins and centrifuged to remove precipitates. The clear supernatant is injected intraperitoneally in amounts of l to 3 ml. For reliable results at least 3 kittens per sample should be used. Strong peristaltic move- ments may be noted which, after l5 minutes to l.5 hours, culminate in the first of a series of attacks of retching and vomiting. The same -20- kitten may be used for several tests within a 7 to ID day period provided the injections are so Spaced as to permit complete recovery from each preceding dose. Hammon (I94I) has described a kitten test wherein the toxic material is injected intravenously. Usually 0.5 to 5.0 ml of culture filtrate that has been properly treated to remove hemolysins, is injected into a kitten weighing about 800 grams. With this intra- venous injection, only vomiting occurring after l5 minutes should be considered as a positive reaction. A moderate meal eaten shortly before the inoculation of the toxic material increases the effect- iveness of the vomiting stimulus. All toxic samples should be first inoculated into cats not previously used. Positive samples can then be confirmed on twice-used or thrice-used kittens. The parenteral administration of enterotoxin to cats is probably less desirable than intravenous administration, since it may be complicated by the hemolysins, rapid development of increased tolerance to the enterotoxin, and by a considerable variation in susceptibility of test animals (Casman, I958). The kitten test, however, is more sensitive, less expensive and more convenient than the monkey feeding test. The frog test developed by Robinton (I950), involves feeding toxic filtrates to frogs (Rana pipiens) and observing for reverse peristaltic waves in the stomach and intestine. It was found however, that the reaction of frogs to enterotoxin was inconsistent and apparently unrelated to dosage. Positive reactions (reverse peristaltic motion) have been seen in frogs fed saline and other nonenterotoxic -2]- materials (Surgalla, I953). Nematodes were once thought to show certain coiling reactions when in contact with solutions containing enterotoxin. However, it is now known that such coiling reactions were not specific for enterotoxin, and the test is invalid (Bergdoll, I962, I963). Milone (I96l) reported the cytopathogenic effects on tissue cultures by the A and B enterotoxins in various states of purity, in an attempt to adapt tissue culture methods to a bioassay of enterotoxin. This test was not reliable because not all of the presently known enterotoxins have cytopathogenic effects on tissue, and because impurities may cause non-specific effects (Bergdoll, I963). Tropical fish of Tilapia spp. were used by Raj and Liston (I962) to demonstrate enterotoxin. Bergdoll (I963) concluded from tests performed with purified enterotoxin that the observed reaction of distress and death of fish was due to some substance other than enterotoxin. Immunological methods. The demonstration of the antigenicity of enterotoxin and the identification by gel-diffusion methods of a specific precipitating antibody against enterotoxin opened a new area for which to study the properties of the toxin. Dolman and Wilson (I938) using horse antiserum and crude toxins reported 3 flocculation zones in Ouchterlony plates. The Oudin (I952) tube gel diffusion test has also been used in the study of entero- toxin homogeniety, purification and production. Surgalla 3;.21. (I952) applied this technique to follow the purification of staphylococcal -22- enterotoxin. Immunological techniques, are considered more Specific than the biological tests since they are based on the reaction of entero- toxin (antigen) with its specific antisera. The main disadvantage of these techniques is that antisera prepared against highly purified enterotoxin are necessary and thus only the known antigenic types of enterotoxins can be detected. Crowle (l96l) described the principles and details of all the tests based on gel-diffusion. A brief description of such tests is included here. Gel-diffusion can either be single or double: Single gel-diffusion, also known as Oudin's technique, involves the layering of an agar column containing antiserum with a solution . of antigen. Under favorable conditions, a band of precipitate forms in the agar for each antigen-antibody system present. In the case of enterotoxin, as it diffuses the front of the band formed moves down the agar column at a rate corresponding to the concentration of entero- toxin and the concentration of antibody. This method has been adapted for the determination of the concentration of type B enterotoxin, at levels as low as l pg/ml (Bergdoll, I962). In double gel-diffusion, a thin glass tube is prepared by placing a layer of agar containing the antisera in the bottom, then a layer of neutral agar, and finally a layer of antigen, also in agar, over the neutral zone. As both the reagents diffuse into the neutral zone, a band of precipitate is formed where the two reagents meet in Optimal proportions. The test as described above is known as -23- Oakley's double diffusion tube method (Oakley and Fulthorpe, I955) and has been used by Hall at al. (I963) to detect enterotoxin concen- trations as low as 0.05 pg/ml. Double diffusion can also be performed in a petri dish layered with agar. This technique is known as Ouchterlony‘s method (Ouchterlony, I949, I953) and is an excellent method for detecting soluble antigens and antibodies. To conduct the test, one first layers a petri dish with a buffered agar and cuts a central hole in the agar after solidification. Several more holes are also cut so that they form a circle around the center one. Antiserum is added to the centrally located well in the agar and antigen is added to the peripheral wells. Antigens and antibodies diffuse through the agar to form zones or lines of precipitation where they combine in optimal proportion. The method permits separation of multiple precipitating systems into their individual components and, in addition, permits comparison of two antigens or antibodies for identification. A more sensitive modification of the above method was developed by Wadsworth (I957) and described in detail by Crowle (l96l). This method involves layering a microscope slide with a thin film of agar and then placing a plastic template in complete contact with the agar. Reagents are placed in funnel-like wells drilled in the plastic template. This method, known as the slide gel-diffusion test or microdiffusion test, involves the same principles and method of interpretation as the Ouchterlony test. Measurement of the amount of enterotoxin consists in the determination of the highest dilution giving a line of precipitation which could -24- be identified through its coalescence with a reference line of precipitation produced by known samples of pure enterotoxin. The product of the volume of toxic food extract or toxic broth used and the reciprocal of its dilution giving such a line of precipitation is used to indicate the amount, in pg, of enterotoxin since it is estimated that a concentration of l pg/ml represents the limit of sensitivity of the test (Casman and Bennett, I965). Hopper (I963) used a flotation system to extract and detect staphylococcal enterotoxin from food. Ham salad was chosen as a food to which enterotoxin type B was added. The toxin was extracted by grinding and centrifugation. Rhodamine conjugated antiserum was added to the clear supernatant liquid. The treated supernatant was then passed through Sephadex C-25 and the eluent was mixed with a wetting agent. Through the use of compressed air, foam was formed and the conjugate-toxin complex, located on the layer of foam was separated to give a distinctly red-colored fraction. The toxin was identified by mixing one dr0p of the foam fraction, one drop of phosphate buffer at pH 4.5 and one drop of latex polyester suspension. Within one minute, a heavy agglutination occurred. Concentrations of l pg/ml from aqueous solutions were detected by this technique within 2 to 3 hours. Robinson and Thatcher (I965) have reported an indirect hemagglutination inhibition procedure for detecting staphylococcal enterotoxin. Erythrocytes were sensitized with purified enterotoxin agglutinated in the presence of specific antiserum. The hemag- -25- glutination reaction was inhibited by prior incubation of a specific dilution of immune serum with graded amounts of enterotoxin in the presence or absence of impurities. By comparing the inhibitory effect of this reaction of known and unknown preparation of enterotoxin, Robinson and Thatcher were able to detect as little as 0.04 pg/ml of enterotoxin type A in 2 to 3 hours. Since other than enterotoxic substances may nonspecifically agglutinate sensitized red blood cells it seems that only highly purified food extracts will be identified as containing enterotoxin. Therefore, the problem of adequate extraction and purification methods must be solved first. Labeling of antibodies by coupling with fluorescent dyes to permit their detection by direct microscopic examination was suggested by Coons §£_al. in l94l. Briefly, the method consists of attaching a compound such as fluorescein to the antibody molecule. The fluore- scein-antibody conjugate is then mixed with antigen and the excess conjugate washed away, leaving behind an antigen-antibody complex which will fluoresce under ultra-violet light. The detection of small amounts of antigen-antibody complex is possible when the test is conducted on a microscope slide and the slide viewed under a U.V. microscope. Such an immunofluorescent technique has been applied to the detection of type B enterotoxin in foods by Genigeorgis (I966). Specific staphylococcal enterotoxin B antiserum was conjugated with fluorescein isothiocynate and used to detect amounts of enterotoxin less than 0.05 pg/ml of food extract without involving any purifi- cation procedures. ,. . . '\ , _ I I I ‘ I « ‘ . f . . V .. . l ' . I I i I _ , . - . 0 1| -26- The demonstration of enterotoxin in food involves two main problems, extraction and concentration when small quantities are present. Experiments have been conducted by adding known amounts of toxin to foods and then recovering them by extraction. The amount of recovery varies. For example, Casman and Bennett (I965) reported recovery of 68% of enterotoxin type A and 48 to 72% of enterotoxin type B in such experiments. Hall__t__l. (I963, I965) reported recoveries of over 90% of type B. Zehren and Zehren (l968a) have reported recoveries of l6 to 35% for type A enterotoxin added to Cheddar cheese. The strain of S. aureus producing enterotoxin necessitates some consideration of the method of analysis used. For instance, those producing type B toxin characteristically do so in large amounts, often over IOO pg/ml in broth cultures. Conversely, those producing type A toxin do so in very low amounts, only 2 to 4 pg/ml in aerated culture as reported by Casman and Bennett (I965). Much lower amounts of toxin are expected to occur in food, and a concentration procedure is required when gel-diffusion techniques are used to detect the toxin. Indirect bacteriological methods. As a result of intensive studies during the past 30 years, a variety of metabolic character- istics of staphylococci have been studied and correlated with entero- toxin production. The production of the enzyme coagulase is the most notable of those studied. The presence of coagulase may be demonstrated directly by mixing about 0.l ml of a broth culture of_§. aureus with 0.5 to l.O ml of a A la ‘7 -\ Q‘ I. w... n .- r - A A l a I" 4 1 I I -27- l/l0 dilution of citrated human or rabbit plasma and incubating at 37 C. Clotting of the plasma in 3 to 6 hours is a positive indi- cation of coagulase. Many indirect assays for coagulase production are also available. The use of tellurite-glycine agar as described by Zebovitz §t_al. (I955) is a common method. To conduct this test one streaks a suspension of §. aureus onto a petri plate of tellurite—glycine agar and incubates it 24 hours at 37 C. The production of black colonies on the plate is considered a positive indication of coagulase production. Other characteristics of S. aureus that have been studied as possible indicators of enterotoxin production include phosphatase, lipase, hemolysins, gelatinase, golden pigment and phage typing. However, it appears that no single one of these characteristics can be correlated completely with enterotoxicity since most of them have also been found to be characteristic of nonenterotoxic strains (Genigeorgis, I966). MATERIALS AND METHODS Gel-diffusion reagents. Antisera for types A and B enterotoxin and highly purified A and B reference toxins were obtained through the courtesy of Dr. E. P. Casman, U. S. Food and Drug Administration Laboratories, Washington, D. C. The antisera and reference toxins were held at -30 C until used. Staphylococcal strains. Two different strains of Staphylococcus aureus have been used throughout this study. They are identified as (I) strain 265-l which produces type A enterotoxin and (2) strain 243 which produces type B enterotoxin. The toxin producing strains of S. aureus were also provided by Dr. Casman and were maintained on porcelain beads held at 4 C until reactivated by rolling the beads onto nutrient agar slants and incubating at 37 C. The cultures were maintained in an actively growing condition by transferring them to a fresh nutrient agar slant 24 hours before they were needed for inoculation into brain heart infusion broth used in toxin production investigations or into milk manufactured into cheese. Culture media. Nutrient agar slants were used for reactivation of the §. aureus strains as mentioned. Nutrient broth at pH 7.4 (Difco), was used for growing the strains of §. aureus prior to inoc- ulation into milk used to make the cheese in this investigation. Brain heart infusion (BHI) broth (Difco) was used for growing the strains to obtain enterotoxin. Casman (I967) has reported this is an excellent medium for obtaining maximum toxin production for several strains of S. aureus. -29- Staphylococcus IIO (Difco) and Standard Plate Count agar (Cudahy) have been used for enumeration of bacterial populations. The Staphylococcus ll0 (S-llO) medium was used as a selective growth substrate for the determination of the number of staphylococci occurring during the manufacture and ripening of the cheese made in this investigation and in the cultures of §°.EE£2!§ grown in BHI broth. Plate Count agar (PCA) was used to determine total populations of microorganisms during the manufacture and ripening of the cheese. Pilot plant experiments. Approximately 2500 pounds of pasteurized milk was made into Cheddar and colby cheese by a standard commercial process (Van Slyke and Price, I952) in the Michigan State University dairy plant. A total of twelve vats of cheese were manufactured - six Cheddar and six colby. Each vat initially contained 2l0-230 pounds of milk which was made into a cheese weighing 20-25 pounds. Fol- lowing manufacture the cheese was stored in a 48 F curing room until termination of the study at six months. After 60 days of ripening, the cheese was removed from its paper wrapping material because of mold growth on the cheese surface. The moldy areas were trimmed off and the remaining cheese coated with wax and returned to the curing room. The problem of mold on the cheese did not occur again during the remainder of the ripening period. During manufacture of the cheese, certain vats of milk were treated in the following manner: (a) inoculated with 24 hour cultures of type A or type B toxigenic strains of s. EELSEE: (b) added penicil- lin to a concentration of 0.49 to 0.54 U/ml of milk or (c) added both -30- penicillin and the §'.§E£EE§ inoculum. Other vats of milk received no special additions during manufacture and served as control vats. The staphylococcal inoculum and the penicillin were added 45 minutes after the addition of a l% lactic starter culture to the milk. A summary of the treatments of the twelve vats during manufacture of the Cheddar and colby cheese used in this research is illustrated by the data in Table I. Samples for determination of total and staphylococcal populations, pH and of enterotoxin were taken from, (a) milk before addition of the starter, (b) milk after inoculation with_§._32£§E§, (c) whey-at-cutting, (d) curd-at-cutting, (e) curd-at-pressing, and (f) the cheese after I, I0, 20, 30, 60, 90, and I80 days of ripening at 48 F. Whey and curd samples were taken as aseptically as possible using a clean ladle and held at 4 C in sterile milk dilution bottles until bacterial populations could be determined. The curd-at-pressing and cheese samples, were held at 4 C in sterile glass beakers until bacterial populations could be determined. The I, I0, and 20 day cheese samples were mistakenly frozen at -30 C before populations were determined. Freezing of these samples probably accounts for some of the erratic populations observed in the cheeses during the first 20 days of ripening. The pH of the samples was determined by inserting the electrodes of a Corning Model 7 pH meter directly into the whey and curd. For the curd-at-pressing and cheese samples, pH was determined on a 50% (w/V) slurry made with distilled water. -31- Table I: Amounts of penicillin and toxibenic strains of Staphylococcus aureus added to vats of milk made into Cheddar and colby cheese. Vat number Penicillin added Staphylococcus aureus (U/ml milk) (organisms/ml milk) Cheddar l none none 2 0.54 none none 480,000 type A 4 0.54 480,000 type A 5 none 450,000 type B 6 0.49 450,000 type B Colby none none 8 0.49 none 9 none 440,000 type A l0 0.49 440,000 type A ll none 450,000 type A -32- Following the enumeration of bacteria and measurement of pH, all samples were frozen at -30 C until enterotoxin assays could be conducted. Enumeration 2f organisms. Total bacterial populations were determined with PCA medium, and the plates were incubated at 32 C for 72 hours. Staphylococcal populations were determined by surface streaking on pre-poured S-llO medium. A separate, sterile glass rod, bent into the shape of an ”L? was used to spread the inoculum over the surface of the medium. For zero dilutions one ml of inoculum was spread over the surface of the medium in three plates while for other dilutions one-tenth ml of the appropriate dilution was used to inoculate each plate. Incubation was at 37 C for 48 hours. Dilutions for both total and staphylococcal populations were plated in triplicate. The populations of each sample were recorded as the average number of organisms in the three plates at each dilution. Whey samples required no special treatment before enumeration. Samples of curd and cheese required blending of II grams of the sample with 99 ml of sterile 0.2% sodium citrate before dilutions could be made. Analysis 3f cheese for enterotoxin. Casman (I966) has reported an extraction procedure which selectively removes and concentrates trace amounts of enterotoxin resulting from the growth of toxigenic strains of_§..ag£§g§ in food. Casman's procedure employs two prin- ciples in the recovery of enterotoxin from food. These are, (a) the use of physical and chemical procedures for the selective removal of -33- enterotoxin from food constituents, leaving the toxin in solution and (b) the selective absorption of the enterotoxin from the soluble food extractives onto a carboxymethyl cellulose column. A description of the technique used to detect enterotoxin in the cheese manufactured in this study is illustrated in Figure I. To conduct the assay, IOO grams of cheese were blended with 500 ml of 0.2 M NaCl at high speed in a Waring blender for 3 minutes and the slurry adjusted to pH 7.5 with l.0 N NaOH. The slurry was then allowed to stand I5 minutes after which the pH was checked and re-adjusted to 7.5 if necessary. The slurry was then centrifuged at 29,700 x g for IS minutes in a Sorvall RC-ZB refrigerated centrifuge. The supernatant was strained through a 4 x 4 inch piece of screen-type wire to remove any floating lipoidal material and held at 4 C during which time the sediment was re-extracted with I25 ml of 0.2 M NaCl by blending as before and centrifuging for IS minutes at 29,700 x g. The two super- natants were pooled, placed in dialysis tubing and concentrated to 40-50 ml by immersion in 30% polyethylene glycol (PEG) 20,000 (Matheson, Coleman and Bell). This step was very conveniently ac- complished overnight since usually 7-9 hours were required to obtain the desired reduction in volume. Following dialysis, the tubing was washed in tap water to remove the PEG and rehydrated momentarily in distilled water and then placed in a beaker of 0.2 M NaCl for 3 minutes, to further rehydrate the contents. The dialysis tubing was then emptied of its contents and each section of tubing carefully washed with 0.2 M NaCl to remove any traces of extract containing -34- Blend 3 min 100 g cheese 500 m1 and 125 ml 0.2 M NaCl 15 min 29,700 x g ""' Adjust Concentratg to pH 7.5 to 40-50 m1 with 30% PEG Adjust to pH 7.5 10 min 37,000 x g ~ Add area 40 vols. éfd 1/4 volt 0.005 M P04 CHCl3 buffer, pH 6 *— --ouom x ommmno w ooH\vouoouow w: omooco w ooH\wmwpm mi waouououao m omhu mo unsoe< H.HH do own n.0H om ohm n.0a om ohm 0.0m «ma Odo 0.03 onm oqo o.n¢ 0mm Odo xpo>ooom N omooso m ooH\wouooumv m1 omoono w ooH\poppm m1 saxouououno < mama mo unnoad .omoono mnfioo owxouuaoa mo monEMw m 00H ou poppm saxououmuao m can < mommu wo muo>ooom "N oHan -48- weighing I00 grams, and the toxin extracted as previously described. During the extraction procedure several of the samples did not Iyophilize properly and it is possible that some of the toxin was lost when the samples foamed over into the vacuum chamber. One of the samples was lost in a laboratory accident prior to lyophilization. The data from the remaining five samples is presented in Table 2. The type B toxin recovered ranged from I0.0 to 66.9% and averaged 36.l%, which agrees with a report by Casman and Bennett (I965) where 30% of type B toxin added to Feta cheese was recovered. Several people working with enterotoxin extraction have evaluated their procedures experimentally to determine the lower limit of toxin detectable in food extracts. For instance, Zehren and Zehren (I968a) found the technique they used could detect as little as 0.003 pg of type A toxin added per gram of cheese. Read, £3 21' (I964) reported that 0.02 pg of type A and 0.05 pg of type B toxin per gram of cheese was the lower limit of their extraction procedure. In another inves- tigation, Read, at 21' (I964) reported the minimum limit of detection for types A and B toxin added to milk was 0.l5 and 0.03 pg/ml, respec- tively. In the work reported herein there was no attempt to determine the lower limit of the amount of toxin detectable with the extraction procedure used. The lower limit of toxin that is detectable by any procedure depends upon the average percentage recovery that is possible by that procedure. For instance, assuming a requirement of l pg of toxin in the final concentrate for a positive gel-diffusion test and a recovery of 27% of type A toxin, then 3.7 pg of toxin must be present -49- in the original IOO gram sample of food for detection. This corres- ponds to a minimum detection of type A toxin for the extraction procedure used in this investigation of 0.037 pg/g of sample. By the same reasoning the minimum amount of type B toxin that could be de- tected is 0.027 pg/g of sample. Pilot plant experiments. Cheese manufacture depends upon good growth and acid production by the lactic starter organisms (Van Slyke and Price, I956). On certain occasions the starter culture may be destroyed by bacteriophage or inhibited by antibiotics in milk produced by cows receiving therapy for mastitis. Such milk, when used in cheese manufacture, results in ”dead vats” and staphylococci, present either as part of the flora in the raw milk or as post-pasteurization contaminants, are able to grow because of lack of competition (Harmon, I967). These strains of staphylococci, if toxigenic, could produce toxin during manufacture or curing of the cheese. To approximate a commercial manufacturing situation cheese was prepared as described previously and aqueous suspensions of penicillin added to certain vats of milk to produce dead vats. The lactic acid bacteria normally used in making cultured dairy products are usually inhibited by penicillin at a concentration of 0.5 Units/ml of milk. 0n the other hand most strains of staphylococci, including those that are toxigenic, secrete a penicillinase which permits them to over-grow the lactic acid bacteria in low concentrations of penicillin. The concen- tration of penicillin added to the vats ranged from 0.49 to 0.54 U/ml of milk as shown in Table I. Some of the vats containing the penicil- lin were also inoculated with toxigenic strains of S. aureus. Other -50- vats of milk lacking the penicillin were also inoculated with these strains of staphylococci and served as standards for observing the effect of lactic acid bacteria growth and acid production upon staphylococcal growth and toxin formation. Effect 9f penicillin upon EU: lactic acid bacteria and staphy- lococcal populations 19 Cheddar cheese. Figures 3, 4 and 5 show the total and staphylococcal populations and pH of Cheddar cheese during storage at 48 F for I80 days. Samples of the curd at pressing represent zero time. The data in Figure 3A are from samples of cheese made from milk (vat I) that contained neither penicillin nor a staphylococcus inoculum and represents the production of normal Cheddar cheese. The percent acidity of the whey-at-milling during Cheddar cheese manufacture is indicative of lactic acid microorganism development and should be between 0.40 and 0.60% to facilitate normal pH and organism devel0pment during ripening. The results of tests made on the Cheddar cheese made in this investigation are shown in Table 3. The control Cheddar cheese made in vat I had a whey-acid of 0.50% when the curd was milled. The pH of normal Cheddar cheese should be about 5.2 when first made and should gradually increase to about 5.6 as soluble nitrogenous compounds resulting from proteolysis accumulate during ripening. Figure 3A (vat I) shows that the pH of this cheese varied from 5.2 to 5.6 during I80 days of ripening, which is normal. The curve representing total population of the cheese manufactured in vat I (Figure 3A) indicates that the maximum population was attained after 90 days of ripening. All cheese samples taken in this investi- -51.. Table 3: Percent acidity of the whey-at-milling during the manufacture of Cheddar cheese. Vat number Acidity (% lactic acid) I 0.500 2 0.l65 3 0.630 4 0.l25 5 0.5l0 6 0.l25 _52- Log of count per g g”, ( \ \ 1' / \ \ \ \ Curing Time in Days at 48F 17 on 9L\ /\\ \/ h 8" \ ————— §§ pfl_-—-I q: M_____ --_——' C1 7L oa q 6 c; 6 3 . C) . 0 SF /\0 —/ ~0— 4 ' o \. 3 \5 00 3 2. I B I iii 1 l 1 I l 30 60 90 I20 ISO IBO Curing Time in Days at 48F Figure 3: Total and S, aureus populations and pH of Cheddar cheese manufactured from pasteurized milk containing (A) no penicillin or (B) 0.54 Units of penicillin/m1 (cured at 48 F). pH pH -53- gation at l, l0 and 20 days of curing were frozen before enumeration of microorganisms, which accounts for erratic counts on these samples. The cheese manufactured in vat I also contained a low population of .§‘.E££EE§ which attained a maximum of 500 organisms/g of curd at pressing and gradually declined to less than I organism/g at l20 dayS. The milk used to manufacture the cheese was not inoculated with staphylococci but was probably contaminated by equipment used in adjacent vats of milk which were inoculated with staphylococci. There are other instances where contamination of non-inoculated vats of milk was suspected. The data in Figure 3B are from samples of cheese made from milk in vat 2 which contained 0.54 Units of penicillin per milliliter of milk. Table 3 indicates that the acidity of the whey-at-milling for the cheese from vat 2 was 0.l65%, suggesting little acid production by the lactic acid starter microorganisms. The pH of this cheese was abnormally high and ranged from 6.3 to 6.8 during ripening (Figure 3B). The maximum total population was 50 million organisms/g after 90 days of curing. The staphylococcal population, probably from contamination as suggested above, was l300 organisms/g on the first day and attained a maximum of ll0,000/g after 60 days of curing. Numerous workers have suggested that food with a staphylococcal population of 500,000 organisms/g may contain emetic levels of toxin. The staphylococcal population that occurred in the cheese from vat 2 probably did not present any hazard, but obviously staphylococci can grow well in vats of milk lacking proper acid development. -54- Figure 4A contains data from the analysis of cheese manufactured from milk inoculated with 480,000 cells of S. aureus/ml (vat 3) which produce type A enterotoxin. The acidity of the whey-at-milling (Table 3) was 0.63% suggesting good lactic starter development. The pH range of the cheese during ripening was normal, varying from 4.9 to 5.4 (Figure 4A). The apparent decrease in total population of the cheese from vat 3 (Figure 4A) during the first 20 days of ripening probably reflects the effect of freezing the samples as mentioned before. The staphylococcal population was 8,000/g on the first day of curing and increased until a maximum population of l30,000 organisms/g was reached after 60 days of ripening. The S. agsggg population then decreased gradually until termination of ripening. Data in Figure 4B represent the results of analysis of samples of Cheddar cheese manufactured from milk inoculated with 480,000 cells of type A toxin producing_§._ag£gg§/ml of milk (vat 4). Penicillin was also added to the milk at a concentration of 0.54 U/ml. The percent acidity of the whey-at-milling was 0.l25 and indicates that this was a dead vat during manufacture. The pH of this cheese during curing was higher and ranged from 5.9 to 6.4. The total population in the cheese from vat 4 (Figure 4B) reached a peak of l20 million organisms/g after 30 days. The staphylococcal population was 5,000 organisms/g on the first day after manufacture and increased to a maximum of ll0 million organisms/g after 30 days of curing. The total and staphylococcal populations then decreased slightly during the remainder of the ripening period. The high total -55- .s h- 0) Q *3 C 3 O O t.- 0 no 3 2 I A l— L J l l l L 30 60 90 I20 l50 IOO Curing Time in Days at 48F 17 In Than A 'M _______ u —— -\ .6 g 'N O. U o.— O «5 00 O ..I j l l L L I; L 30 60 90 I20 l50 IIO Curing Time in Days at 48F Figure 4: Total andi§_° aureus populations and pH of Cheddar cheese manufactured from pasteurized milk (A) inoculated with 480,000 type A enterotoxin producing §, aureus cells/m1 or (B) inoculated as above and containing 0.54 Units of penicillin/m1 (cured at 48 F). pH -56- population probably results from the high staphylococcal count, since the low acidity of the whey-at-milling and the high pH of this cheese indicates poor lactic acid starter development. Comparing the two cheeses from vats 3 and 4 (Figure 4A and 48) one is also able to see very dramatically, the influence of the presence of penicillin in reducing lactic culture activity, causing poor acid development and permitting staphylococci to grow well in the milk and cheese. Data in Figure 5A indicate the results from the analysis of samples of Cheddar cheese manufactured from milk inoculated with 450,000 cells of S. EEEEEE/ml which produce type B toxin (vat 5). The percent acidity of the whey-at-milling was 0.5l (Table 3) which is indicative of good lactic starter growth. The pH of the cheese during ripening was normal and varied from 5.2 to 5.6. The total and staphylococcal population in the cheese from vat 5 (Figure 5A) seem to follow the same trend throughout curing. The curd contained over I million staphylococci/g when pressed, reached a peak of approximately l8 million/g after 60 days and gradually decreased to about 2,500 organisms/g after l80 days. The high population of .§°.22£§E§ was unexpected since normal lactic starter development occurred in vat 5 during manufacture. The data in Figure SB are from samples of cheese made from milk in vat 6 which contained 450,000 cells of type B producing §‘.EE£EE§ and 0.49 Units of penicillin per milliliter of milk. The percent acidity of the whey-at-milling was 0.l25 and indicates that this was a dead vat. The cheese from vat 6 (Figure SB) had a pH range during -57- l7 s: h 0) Q‘ IQIEL_———» O .6 H c \ 3 — O 8 ,I’ T‘\~~ ’- . ‘~‘~~pi+ ~0— ~.‘ 0 .8. as 3!- on O _l 2" 't A 5 l g L 4 1 a 30 60 90 I20 150 ISO Curing Time in‘ Days at 48F v- 17 I 9- no . 'Rnal “a o __ - .E.// """".' ‘-W 3. ’- /'\ ' ’ \\‘-—"'-—-——-—_—T*—~p.li ~\. «I6 to ‘rr- 3 o 5' u- 4- O 3" 45 00 O _I 2r l- B i .l l l l l l 30 60 90 I20 l50 IOO Curing Time in Days at 48F Figure 5: Total and §, aureus populations and pH of Cheddar cheese manufactured from pasteurized milk (A) inoculated with 450,000 type B enterotoxin producing §, aureus cells/ml or (B) inoculated as above and containing 0.49 Units of penicillin/m1 (cured at 48 F)° pH pH -53- curing which was higher than normal, varying from 5.9 to 6.4. The total p0pulation reached a maximum of about I billion organisms/g 20 days after manufacture and was relatively constant for the remainder of the curing period. The staphylococcal population in vat 6 was nearly maximum l0 days after manufacture, stationary until 90 days, and then decreased slowly during the remainder of the ripening period. It is evident that the staphylococci contribute substantially to the total population throughout curing. The cheeses from vats 5 and 6 (Figure 5A and 5B) could possibly be hazardous since their respective staphylococcal populations at the time the curd was pressed were I.l million and l.9 million organisms/g. _§£:§g£.gf penicillin upon EU: lactic acid bacteria and staphy- lococcal pgpulations jg colby cheese. Colby is a lower acid cheese than Cheddar because the curd is washed after the whey is drained, thus removing some of the lactic acid. The initial pH of the cheese is usually about 5.3 or 5.4 and increases to 5.7 or 5.8 during ripening. Figures 6, 7, and 8 show the total and staphylococcal populations and pH of samples of colby cheese during ripening at 48 F for I80 days. Samples of the curd at pressing represent zero time. Data in Figure 6A show the results of analysis of samples of colby cheese manufactured in vat 7 which contained milk that lacked the addition of penicillin or staphylococci. The acidity of the whey-at- dipping is an indication of acid development during colby cheese manufacture and normally should be 0.l3 to 0.l5%. A summary of the acidity of the whey-at-dipping for the colby cheese made in this investigation is shown in Table 4. The percent acidity in the colby -59- -7 d: L 3 J Total '6 H _________ DH 5 -------- "i 0 U % O 35 a” \. u _. W. \ l A . l l l l l l k 30 60 90 I20 l50 I80 Curing Time in Days at 48F W7 °° Tm: . 33 o. H .. I E ___________ Q.——-- fl6 3 O U a. o . u) 3- ‘T‘T“- is 3 \. 2r \ II- B I L L L J l I so so 90 no I50 no Curing Time in Days at 48F Figure 6: Total and _S__.’ aureus populations and pH of colby cheese manufactured from pasteurized milk containing (A) no penicillin or (B) 0.49 Units of penicillin/m1 (cured at 48 F). pH pH -60- Table 4: Percent acidity of the whet-at-dipping during the manufacture of colby cheese. Vat number Acidity (% lactic acid) 7 0.l30 8 0.l20 9 0.l30 IO 0.l20 ll 0.l35 I2 0.l20 -6I- cheese from vat 7 was 0.l3 which is within the normal range. Data presented in Figure 6A (vat 7) show that the pH of the control colby cheese during ripening varied from 5.0 to 5.7. The apparent initial decrease in total population of this cheese is attributed to loss of viability of organisms during freezing. A total population of ap- proximately l million organisms/g prevailed during the latter half of the curing period. The cheese from vat 7 also contained approximately 800.§°.22£22§ per gram when pressed, probably introduced during the manufacturing process from adjacent vats. The staphylococcal population increased to a maximum of l000 organisms/g at 60 days and decreased to less than l0 organisms/g at l80 days of ripening. Figure 6B shows data from the analysis of samples of colby cheese manufactured from milk in vat 8 that contained 0.49 Units of penicillin per milliliter of milk. The acidity of the whey-at-dipping in vat 8 was 0.l2% (Table 4) which indicates that this was a "slow” vat. The pH of this cheese during curing as shown in Figure 6B was higher than normal, varying from 5.7 to 5.9. The high pH during curing and abnormal dipping acidity applied to all colby cheeses in this investigation which were manufactured from milk that contained penicil- lin. However, the inhibition of the lactic acid starter bacteria may be less in colby than in Cheddar because some penicillin may be washed out of the colby during manufacture. The cheese made from milk in vat 8 (Figure 6B) contained a low population of staphylococci, probably caused by contamination as previously mentioned. The staphylococci reached a maximum of 50,000 -62- organisms/g in the cheese after 30 days of curing and gradually decreased to 3 organisms/g after I80 days. Figure 7A contains data from the analysis of samples of cheese manufactured from milk inoculated with 440,000 cells of S. aggggg/ml, which produce type A toxin (vat 9). Table 4 shows the acidity of the whey-at-dipping was 0.l3% suggesting good lactic starter devel0pment. Figure 7A shows the pH range of the cheese during ripening was normal at 5.2 to 5.6. The total population attained in the cheese from vat 9 was maximum after 90 days of curing with 300 million organisms/g. The staphylococci attained a maximum of 300,000 organisms/g after 20 days of ripening. Both the §'.22£22§ and total populations were fairly constant throughout the remainder of the curing period. Data in Figure 7B represent the results of analysis of samples of colby cheese manufactured from milk in vat I0 which was inoculated with 440,000 cells of type A toxin producing S. aggggg/ml and contained penicillin at a concentration of 0.49 U/ml. Table 4 shows that the acidity of the whey-at-dipping was 0.l2% and indicates that vat l0 was a slow vat. During ripening, the pH of this cheese varied from 5.l to 5.9. Figure 7B (vat l0) reveals that the staphylococcal population attained a maximum of 6.5 million organisms/g 20 days after manufacture, decreased to about 56,000/g after 90 days and increased slightly until the termination of the investigation. The two cheeses made from vats 9 and IQ (Figure 7A and 73) contained sufficient numbers of staphylococci during ripening to make them potentially hazardous after only I day of curing. -63- '17 cm 4T final 8.. 0 O. -6 H S 8 W;___.._._____-FLI'I____ o /’ .No “5 45 on o .4 1. 1 L I L l '1_ 30 60 90 I20 ISO IOO Curing Time in Days at 48F . *7 do 9- d) 8" O. 7" . ‘E —"‘::T“~. ________QH_______:6 3 ' \\ \.. "I ------ . , \ X 0 \ s I ‘4 . “I“ [I \_____._______ W e. I o 4 DD 3 15 o ..l 2 I B 1 L l L J L L 30 60 90 I20 ISO 180 Curing Time in Days at 48F Figure 7: Total and §, aureus populations and pH of colby cheese manufactured from pasteurized milk (A) inoculated with 440,000 type A enterotoxin producing §, aureus cells/ml or (B) inoculated as above and containing 0.49 Units of penicillin/ml (cured at 48 F). pH pH -64- Data in Figure 8A show the results obtained from samples of colby cheese made from milk in vat ll inoculated with 450,000 cells of -§..ggggg§/ml, which produce type B toxin. Table 4 indicates that the percent acidity of the whey-at-dipping for the cheese made from vat ll was 0.1351and suggests the lactic acid bacteria developed normally during manufacture. The pH of the cheese during ripening was normal and ranged from 5.3 to 5.8. The total population was relatively uniform throughout the ripening period except for a slight decrease during the first 30 days. The staphylococci inoculated into the milk during manufacture failed to grow to any significant population in the cheese and reached a maximum of only 2,400 organisms/g 30 days after the beginning of ripening. Figure 88 shows data obtained by analysis of samples of colby cheese manufactured from milk in vat 12 inoculated with 450,000 cells of §.'§££gg§/ml, producing type B enterotoxin. The milk also contained 0.49 U/ml of penicillin during manufacture. The dipping-acidity of the whey from vat l2 (Table 4) was 0.12% and indicates that this was a slow vat. Figure 88 indicates that the pH of this cheese was high and varied from 5.8 to 6.5 during curing. Figure 88 (vat l2) shows that the staphylococci attained a maximum population of 43 million organisms/g 90 days after manufacture and decreased steadily until l,600 organisms/g remained after 180 days of ripening. The total population in the cheese gradually increased throughout the entire storage period and reflects the presence of a high staphylococcal population along with the lactic acid bacteria. 0f the per g; count Log of per g. Log of count -65- -7 RHaL__ 16 ~ ~ ____________________ EtL- ...— ‘n— —-—-—’ —---I- “‘ M.-.$ films .. o.—___5 A 30 6.0 910 I20 1&0 100 Curing Time in Days at 48F -7 45 n n 1 1 1 30 60 90 I20 I50 I80 Curing Time in Days- at 48F Figure 8: Total and g, aureus populations and pH of colby cheese manufactured from pasteurized milk (A) inoculated with 450,000 type B enterotoxin producing §, aureus cells/m1 or (B) inoculated as above and containing 0.49 Units of penicillin/m1 (cured at 48 F)o pH pH -66- two vats of milk inoculated with the type B toxin producing staphylo- cocci, only the cheese manufactured from vat l2, a typical slow vat as indicated by its low dipping acidity during manufacture, contained staphylococci in a sufficiently high number to suggest a hazard. Production 2f enterotoxin lg Cheddar and colby cheese manufactured from milk inoculated with toxigenic strains 2f_§.-au£§g§. One hundred gram samples of Cheddar and colby cheese manufactured in the Michigan State University dairy plant were removed from storage at -30 C and analyzed for enterotoxin by the procedure already discussed. Cheese manufactured from non-inoculated milk was examined after curing at 48 F for 180 days. Cheese manufactured from inoculated milk was examined after curing at 48 F for 60, 90, and l80 days; and if toxin was present then analyses were performed on additional samples taken earlier in the ripening period to determine the age of the cheese when toxin first appeared. Tables 5 and 6 present data obtained by performing slide gel-diffusion titrations on the cheese extracts. Table 5 shows the summary of the assays for toxin as conducted on samples of Cheddar cheese made for this research. There was no toxin found in either of the non-inoculated control Cheddar cheeses during curing. The cheese from vat l was made from milk that did not contain penicillin. The cheese from vat 2 was made from milk that contained 0.54 Units of penicillin per milliliter of milk. Both of these cheeses contained low §..ag£§g§ populations during ripening which did not reach a concentration high enough to suggest a hazard (Figure 3A and 3B). The Cheddar cheese made from vat 3 was manufactured from milk inoculated with a strain of_§. aureus which produces type A entero- -67- 000050 mcoc m omhu 000.0mq 00.0 00 0 25: m 2:3 000.03 3.0 00 0 98a 0 3.3 000.03 3.0 03 0 macs m omhu ooo.om¢ mac: owH m 28 < 25 08.03 and ON a e a 25 08.23 $5 cm a m a 25 25.03 and 8 s a. a 25 08.03 and 0m a macs < om%u ooonomd ofiofi owe m macs macs qm.o owH N oaoa 0:0: 0:0: 00H H $22 #523533 9:? H53 A933 w 00H\:wxou w: @0000 msousm am 00000 aHHHwoaaom omooso mo ow< gonad: .wouwofipcfi mm awHHaowaom mawcwmuaoo can msousm aw mo mdwmuum owcowwxou cuw3 pouwHSUOGH xHHE pouquoummm scum wounuumm inane omooco powwono mo m we no wawuno mcwunw pooswoum saxouououcm um manna um> -68- toxin. The staphylococci attained a maximum population of l30,000 cells/g after 60 days of curing (Figure 4A) and the cheese did not contain toxin when analyzed after l80 days at 48 F. Figure 4B (vat 4) indicates that the §. EELEEE population of the cheese was approximately l million organisms/g after l0 days of curing and reached a maximum of llO million organisms/g after 30 days. The cheese sample obtained after 30 days of ripening contained 4 pg of type A toxin/100 g of cheese. The cheese sample obtained after 20 days of curing gave no indication of toxin, which means that toxin was formed in the cheese from vat 4 between 20 and 30 days of ripening at 48 F. Table 5 indicates that samples of the cheese from vat 4 after 30 and 90 days of ripening contained 4 pg of toxin/lOO 9 cheese. The cheese sample obtained after 60 days of curing contained only 3 pg of toxin/IOO 9 cheese. The fluctuation in toxin content probably reflects variation in recovery since the toxin does not disappear after pro- duction. Evidence will be presented later that associates toxin production with an actively growing §'.EELEE§ population and not a stationary or declining population. One would not have expected an increase in the amount of toxin contained in the cheese from vat 4 after 30 days of curing since the §. EEEEEE population began to decline after this time. Casman and Bennett (I965) have reported type A enterotoxin in food poisoning incidents from coconut cream pie which contained 200 million staphylococci/g and banana cream pie which contained 72 million staphylococci/g. With such populations as indices, one would certainly -69- have expected toxin to be present in the cheese from vat 4 which contained an §..§g£§g§ population of llO million organisms/g in one sample during ripening. Toxin was not detected in the Cheddar cheese manufactured from the remaining vats of milk inoculated with a strain of §°.22£EE§ that produces type B toxin. The milk in vat 6 contained 0.49 Units of penicillin per milliliter of milk but the milk in vat 5 lacked the addition of any antibiotic. During curing, the cheese made from vat 5 (Figure 5A) contained a maximum of 18 million staphylococci/g after 60 days. Lactic acid bacteria developed normally during cheese manufacture in vat 5 and the pH of the cheese varied from 5.2 to 5.6 during curing. The cheese made from vat 6 (Figure 58) contained a maximum of 48 million staphylococci/g after 90 days of ripening. Lac- tic acid bacteria developed poorly in vat 6 during manufacture and the pH of the cheese during ripening varied from 5.8 to 6.3. The population of g. EELEEE required to produce measurable amounts of type B toxin in food has not been reported. No conclusions can be made regarding the absence of toxin and the maximum p0pulation of staphylococci in the cheese produced frcm vats 5 and 6. Table 6 shows the summary of the assays for toxin conducted on samples of colby cheese made for this research. No toxin was detected in either of the non-inoculated colby cheeses during curing. The cheese from vat 8 (Figure 6B) was manufactured from milk that contained 0.49 Units of penicillin per milliliter of milk. The cheese from vat 7 (Figure 6A) was made from milk that did not contain penicillin. Both of these cheeses contained low §. aureus populations throughout ripening. -70- 0505 m mmhu ooo.om0 00.0 00 0000 0 0000 000.000 00.0 000 0aoa m 00%0 000.0m0 0aoa 00 0505 m mahu ooonom0 0505 owH 0000 0 0000 000. 000 00 .0 H 0 0 0000 000.000 00.0 00 0 0 0000 000.000 00.0 00 N 0 0000 000.000 00.0 00 0 0 0000 000.000 00.0 00 N 0 0000 000.000 00.0 00 0 0 0000 000.000 00.0 000 0000 0 0000 000.000 0000 000 0505 0505 00.0 owH QSOG wfiofi mfiOfi owH 00000 00\0000000000 00000 Ha\00 000000 000030 w 00H\aonu mi 00000 050050 .w. @0000 :HHHHoHc0m 000050 mo 0m¢ .v0u00Hwaw 00 GHHHH0050Q wGHGHmuaoo 0H0 050050 .m.mo 0500000 0050waou £003 p0umH5ooaH xHHE u0uwu5mumma Eoum 00u5uomm5cma 000050 mnHoo mo m wq 00 wGHu5o w5H050 00050000 saxouou0uam NH NH HH HH OH OH OH oH oH oH oH umna5n u0> .0 00000 -7]- The colby cheese produced from vat 9 (Figure 7A) was made from milk inoculated with a strain of §. EEEEEE that produces type A enterotoxin. No toxin was detected in this cheese after l80 days of curing although the staphylococcal population was 850,000 organisms/g on the first day of ripening. The colby cheese produced from vat l0 (Figure 78) during manu- facture contained l pg of type A enterotoxin/loo g of cheese after l0 days of curing. The maximum amount of toxin detected in this cheese was 2 pg/IOO g of cheese and occurred after 30 days of ripening. The measurable toxin content fluctuated between I and 2 pg/lOO g of cheese throughout the remainder of the curing period. This fluctuation is attributed to the lack of precision inherent in the extraction procedure because the toxin is stable. Figure 7B shows an_§._2g£gg§ population in the cheese from vat 10 of 3 million organisms/g after one day of curing. This cheese was potentially hazardous after only one day yet no toxin was detected. The number of organisms and the amount of toxin necessary to induce illness in humans is not known. It is possible that emetic quantities of toxin are below the sensitivity of the extraction method used in this investigation. Toxin was not detected in the cheese from vat 10 (Figure 73) until after l0 days of curing when the staphylococcal population was approximately l.4 million cells/g. The maximum staphylo- coccal population in the cheese from vat ID was 6.5 million organisms/g after 20 days of curing. The staphylococcal population in the cheese was declining after 30 days of curing when the maximum concentration of toxin was indicated. Production of toxin by a declining population -72- is unlikely, as will be shown later. Inconsistencies in the extraction procedure probably explain the reason maximum toxin concentration in the cheese was not indicated at 20 days. Casman and Bennett (1965) have reported type A enterotoxin in a food poisoning incident involving macaroni salad which contained one million staphylococci/g. If such a population can produce measurable amounts of toxin, one would have expected toxin in the cheese from vat 10 which contained 6.5 million §._§g£§g§ organisms/g during ripening. However, the staphylococcal population occurring in the cheese from vat 10 is several times less than the 110 million organisms/g in the toxic Cheddar cheese from vat 4 (Figure 48). Data in Table 6 indicate that no toxin was detected in either of the colby cheeses manufactured from milk inoculated with a strain of .§.'22£§g§ that produces type B toxin (vats 11 and 12). The milk in vat 12 contained 0.49 Units of penicillin per milliliter of milk but the milk in vat ll lacked the addition of any antibiotic. During ripening, the cheese made from vat 11 (Figure 8A) exhibited a maximum staphylococcal p0pulation of approximately 2,500 organisms/g after 20 days. The development of the lactic acid bacteria in vat 11 was normal during cheese manufacture and the pH of the cheese varied from 5.3 to 5.8, which is normal for colby cheese. The cheese manufactured from vat 12 (Figure BB) exhibited an §..gg£§g§ population exceeding one million organisms/g after one day and a maximum of 33 million/g after 90 days of ripening. Lactic acid starter deve10pment in vat 12 during manufacture was above normal with the pH of the cheese varying from 5.8 to 6.5 during curing. -73- The results of this pilot plant experiment support the suggestion of Casman (1963), Brandly (1965), Harmon (1967) and the observation by Zehren and Zehren (l968b) that staphylococci can grow well and reach hazardous populations in foods receiving treatments to lower the natural bacterial population. Zehren and Zehren (l968a) examined a large quantity of cheese for type A enterotoxin and found that toxic cheese resulted only from those vats of milk failing to show proper lactic starter growth and acid development, which is partially supported by the results of this investigation. Two cheeses manufactured from milk receiving penicillin to artificially inhibit the lactic bacteria and containing a type A toxigenic strain of_§._gg£§g§, contained toxin during ripening. However, two cheeses also manufactured from milk treated with penicillin and receiving a type B toxigenic strain of _§..§g£§g§, failed to indicate the presence of toxin even though extremely high populations of the organisms occurred during ripening. The particular type B toxin producing strain of §. aureus used in this research produces very high amounts of toxin in broth cultures but has never been associated with a food poisoning incident. Other workers have investigated the growth and persistence of toxigenic strains of §..gg£§g§ inoculated into milk that was manufactured into cheese (Takahaski and Johns, 1959; Walker 95 31., 1960). These workers found that staphylococci concentrated in the curd during cheese manufacture and that the curd at pressing and cheese after one day of curing often contained potentially hazardous numbers of staphylo- cocci. No pattern of staphylococcal growth was evidenced by the work -74- reported herein, however. Potentially hazardous numbers were present for the first time after I, 10, 20 and 30 days of ripening for both Cheddar and colby cheeses manufactured from milk inoculated with staphylococci. Production 2f enterotoxin lg Brain Heart Infusion (BHI) broth at different BE values. Non-buffered BHI broth was suggested by Casman (1967) as the best medium to use for production of enterotoxin resulting from the growth of toxigenic strains of S. aureus. This medium served as the control for observing growth and toxin production in a series of experiments to determine if pH influenced toxin formation. Cheddar cheese is a highly buffered food material, normally having a pH of 5.2 to 5.6. Cheddar cheese produced from dead vats of milk has a pH of 6.0 to 6.8 as reported previously herein. To evaluate the effect of normal and low acid conditions as they can exist in Cheddar cheese, BHI broth was prepared from 0.2 M stock buffer solutions of pH 5.4 and 6.0. The BHI prepared from the pH 5.4 buffer corresponded to the pH of normal Cheddar cheese, while the BHI prepared from pH 6.0 buffer corresponded to the pH of low acid Cheddar cheese. Both type A and type B toxin producing strains of §. aureus failed to grow after 14 days in BHI at pH 5.4. This was unexpected because both strains grew to some extent in Cheddar cheese that had normal acid development during manufacture. However, in making cheese the or- ganism may become well established before the pH decreases enough to be inhibitory. The strain of_§. aureus producing type A toxin was inoc- ulated into the milk in vat 3 and reached a maximum population in the cheese of 130,000 organisms/g after 60 days of curing. The pH of the -75- cheese from vat 3 varied from 4.9 to 5.4 during that 60 day period. The strain of §._§g£§g§ producing type B toxin was inoculated into the milk in vat S and reached a maximum population in the cheese of 18 million organisms/g after 60 days of curing. The pH of the cheese from vat 5 varied from 5.2 to 5.6 during the first 60 days of ripening. Casman and Bennett (1963) have also reported that semisolid BHI agar with pH 5.3 is a suitable medium for the production of types A and B enterotoxin from several different strains of §..2£££2§. Casman and Bennett did not use a buffered BHI, but simply adjusted the medium to pH 5.3 with HCl before sterilization. Both of the toxigenic strains used in the research described herein grew well in non-buffered BHl at pH 5.4. During growth, both strains of §..§g£§g§ raised the pH of the medium to 8.6 after 48 hours. The supernatant from the culture of the toxigenic strains of §. EELEEE in non-buffered BHI at pH 5.4 was not examined for toxin. Data in Tables 7, 8, and 9 show the toxin concentration, optical density at 620 mp, viable count, and pH of the supernatant from the extended growth of the type A toxin producing §._gg£§!§ in BHl broth at various pH values. The data in Table 7 indicate that toxin was first detected in the BHI broth at pH 7.4 after 5.0 hours of growth when the 0.0. was 0.59 and the staphylococcal population was 270 million cells/ml. The maximum toxin concentration attained was 8 pg/ml after 12.0 hours of growth when the population was 4.3 billion organisms/ml. The pH of the supernatant (Table 7) decreased from 7.4 to 6.5 between 5.0 and 6.0 -76- 05000500050 00 0HQE00 000000000500 0Homi0m Au 05000000050 mo 000800 000000500000 0H00iooH 00 000000 002 A0 00.0 00.0 000.000.~ 00.0 0.00 00.0 00.0 000.000.0 00.0 0.00 00.0 00.0 000.000.m 00.0 0.00 00.0 00.0 000.000.0 00.0 0.00 02 00.0 000.000.~ 00.0 0.0 00.0 00.0 000.000.~ 00.0 0.0 00.0 00.0 000.000.0 00.0 0.0 00.0 00.0 000.000.0 00.0 0.0 00.0 00.0 000.000 00.0 0.0 0000. 00.0 000.000 00.0 0.0 00-- 00.0 000.00 00.0 0.0 Hz 00.0 000.00 00.0 0.0 02 00.0 000.0 00.0 0.0 0002 00.0 00 00.0 0.0 00000000 0000 0000000 0000 HE\G«xou m1 $0 05500 0Hn0H> .n.o 0609 0 .00000 000000.00 u z .0 0m 00 000000000 000 0000000000000 0.0 00 00 00000 Ham Nn.m 50 a3o0w 003 00x00000050 < 0000 w50050000 050050 0W 0003 5000000500000 00x00000000 000 .05000000050 0:0 00 :0 .850008 oHHum 00 0000800000 00 05500 0H00H> .18 0N0 00 h000500 H000000 .0809 “n 0H00H -77- hours and then increased throughout the remainder of incubation to reach a maximum of 8.9 after 48.0 hours. Additional toxin was not produced after l2.0 hours of growth which is near the time the pop- ulation reached the stationary phase. Toxin production was not observed when the staphylococcal population was decreasing. Data in Table 8 show that in the BHI broth at pH 6.0, toxin first appeared after 24.5 hours when the 0.0. was 0.58 and the population was approximately 260 million cells/ml. Estimates of the §.'22£§2§ populations occurring in samples of BHI broth at pH 6.0 and 8.0 (Tables 8 and 9) were obtained by comparing the sample 0.0. to the standard population ver5us optical density curve in Figure 9. Figure 9 was obtained by plotting the log of population versus optical density of culture samples taken during the growth of the type A toxigenic strain of §. aureus in non-buffered BHI broth at pH 7.4. Presumably, the population occurring in the non-buffered BHI broth and the BHI broth at pH 6.0 and 8.0 were the same when the optical densities were the same. The é..gg£gg§ population in the Cheddar cheese made from vat 4 (Figure 48) when toxin was first detected was llO million cells/g. The staphylococcal population in the cheese from vat 4 compares reasonably well to the staphylococcal population of 260 million organisms/ml when toxin was first detected in the BHI broth at pH 6.0. The toxin concentration in the BHI broth at this time was 0.50 ug/ml. When the §'.EH£EE§ population was llO million organisms/g in the cheese from vat 4, approximately 0.l5 pg of toxin/g of cheese was present. -73- 00omuom 00000050o5oo 05000500050 00 000000 0oz A0 00.0 00.0 000.000.0 00.0 00.0 00.0 000.000.0 00.0 00.0 00.0 000.000 00.0 00.0 00.0 000.000 00.0 00.0 00.0 000.000 00.0 00.0 00.0 000.000 00.0 0000.0 00.0 000.000 00.0 00 -- 00.0 000.000 00.0 02 00.0 000.000 00.0 02 00.0 00.0 Hz 00.0 00.0 0002 00.0 00.0 00000050 0000 018 0000 08\50x00 01 00 05500 00000> 00008000m .a.o £0000 08\ooo.w u 02 .0 mm 00 000005050 050 000:00oc0 8500O0 z «.0 0003 0.0 30 00 00000050 .00050 Hmm Nu.m 50 5300w 003 50%00000050 0 0000 w50050ou0 050050 .w.5003 5000000500500 50xo0ou0050 050 05000500050 000 mo $0 .05500 0000H> 000080000 .18 0N0 00 0000500 000000o .0808 H0 00000 -79- .0 mm 00 000005050 050 0.5 $5 00 50000 Hmm Nmom 00000050u505 50 5300w 000 50x00000050 < 05%0 w50050005 050050 mm 00 00000 5053 5000005505 050 5000500 0000050 5003000 50:05000000m "m 005m05 1t» . god 0 o.— 0.0 «A 0.. ad 06 v.0 «.0 u u q d 0 u u u m 301 405 ‘o h: JO IL 9' Q |uJ/1unoo lOd— -80- This amount of toxin is the result of correcting the amount detected in the cheese, 0.04 ug/g, for the percent recovery known for the extraction procedure used in this investigation. Lower concentrations of toxin in cheese as compared to broth cultures of g. aureus, even with nearly equal populations, was expected since aeration in the cheese is lower than in the broth. Increased aeration has been cor- related with increased toxin production by type A toxin producing strains of §. aureus (Kato _t _l., I966). The toxic colby cheese from vat 10 (Figure 78) had a maximum _§. aureus population of 6.5 million organisms/g during curing. Toxin was never detected in any of the BHI broths containing less than 260 million organisms/m of the type A producing strain of §. aureus. Unless there is some unknown difference between toxin development in cheese and in broth, two things are evident about the toxic colby cheese from vat ID: (i) the §. aureus population in the cheese during manufacture and ripening was higher than indicated, possibly because of enumeration errors or (2) false-positives were obtained in the test for toxin. The former is more tenable than the latter because positive tests for toxin were obtained in 6 different samples. Data in Table 9 represent results of analysis when the type A toxigenic strain of §. aureus was grown in BHI broth at pH 8.0. Toxin was not detected until after 12.5 hours of growth and the population was approximately 530 million organisms/ml. This is approximately two times the p0pulation present when toxin was first detected in the BHI at pH 7.4 and 6.0. The maximum toxin concentration observed in the BHI broth at pH 8.0 was 6 ug/ml as compared to a maximum of 8 ug/ml -81- 00.0 0m.w 00.0 0N.w 00.0 00.5 00.m 05.5 00.m 05.5 00.0 05.5 A0mN.0 05.5 An— ll Ohofi 82 05.5 Hz 00.5 Hz mw.5 Hz mm.5 Hz 00.0 A082 00.0 HE\50000 w: 05 00000 0a\000.0 u 0 2 0000-00 000000500500 05000500550 A0 0000:000 000000500500 05000500550 00 000000 002 A0 000.000.0 00.0 0.00 000.000.0 00.0 0.00 000.000.0 00.0 0.00 000.000.0 00.0 0.00 000.000 00.0 0.00 000.000 00.0 0.00 000.000 00.0 0.00 000.000 00.0 0.00 000.000 00.0 0.00 000.000 00.0 0.00 000.00 00.0 0.00 00.0 0.0 00.0 0.0 00.0 0.0 00000050 0000 015 0000 0000 05500 00000> 000050000 .5.o 0E0H .0 00 00 000000000 000 0000 z 0.0 0003 0.0 05 00 00000050 .00000 000 N5.m 50 53000 003 50000000050 < 0500 050050005 050050 .w.5003 5000000500500 50x00000050 050 05000500550 000 00 $5 .05500 00000> 000080000 «18 000 00 0000500 0000050 .0809 no 00009 -32- observed in the control BHI broth at pH 7.4 and the BHI broth at pH 6.0. Growth of the type A producing strain 0f.§°.22£22§ in BHI at pH 8.0 was not significantly inhibited but toxin production was less than when growth occurred in BHI broth at pH 6.0 and 7.4. Similar observations of reduced toxin production but little inhibition of growth for type B toxin producing strains 0f.§°.EELEE§ have been reported. Genigeorgis (I966) observed that strain S-6 0f.§°.22£22§ grew normally in BHI broth with lOZ NaCl added, but failed to produce any toxin. Tables l0, ll, and l2 contain data showing the toxin concentration, optical density at 620 mp, viable counts and pH of the supernatant from the extended growth of the type B toxin producing strain of .§°.22L§E§ in BHI broth with various pH values. Estimates of_§..ag£§!§ populations occurring in samples of BHI broth at pH 6.0 and 8.0 (Tables ll and l2) were obtained by comparing the sample O-D- to the standard population verses optical density curve in Figure 10. The data in Table 10 show the result of growing the type B toxin producing strain 0f.§'.EELEE§ in non-buffered BHI broth at pH 7.4. The first indication of toxin occurred after 6.0 hours when the optical density was 0.36 and the population was 47 million organisms/ml which is lower than the population present in the BHI broth at pH 7.4 when type A toxin was first detected. The type B strain produced a maximum of 382 pg of toxin/ml of supernatant in the BHI at pH 7.4 at which time the population of §._gg£gg§ was 6.6 billion cells/ml. When the type A toxin producing strain of §. aureus was grown in the BHI broth at pH 7.4, the maximum toxin concentration occurred when the population wHomnom wmumuuamocou uaMumfinmmnm An wmummu uoz Am oo.~wm om.m Hz oo.H o.- oo.~wm om.w ooo.oo~.fi oo.H o.m¢ Hz mh.w Hz oo.H 0.0m oo.~mm mm.m ooo.ooo.o oo.H o.q~ oo.~HH om.“ ooo.ooo.o om.H o.~H oo.qo oo.o Hz o¢.H m.oH oo.qN 00.0 ooo.oo¢“~ mm.H m.m Hz mm.o ooo.o- om.H o.m Hz om.o ooo.oo¢ oH.H m.“ x4 Hz mm.o ooo.om~ mh.o 0.5 .w Hz co.“ ooonfio oq.o m.o oo.H 0H.“ ooo.~q om.o 0.0 An -- o~.n ooo.- mH.o o.m Hz oq.~ ooq oo.o o.m Awaz cc.“ ooN 00.0 0.0 AcmuuMEO oo0v his omov Aggy HE\c«xou m1 mm unsou mHan> .Q.o mafia spoun He 0 \ooo.Hm u z .0 Km um umumnaoafi yam Awmummmsnnoav «.5 ma umlnuoun Ham Nn.m aw :3ouw was awxououmufim m mmmu wawosuoua mamuam .m amnB aoHumnuawoaoo saxououmucm van ucmumaummsm mnu mo mm .Esfiume oHHnm so woawanmumv mm uasoo manww> .18 omo um muHmec Hmowuao .meH "OH maan .0 um um vmumnnocfi vcm «.5 ma um spawn Hmm Nu.m vmummwsauco: a“ CBOHw mun aflxOuoumucm m mmhu wcwosvoum msmusm .m.mo mHHmo cm£3 coflumasmom vdm muHmCmv Hmofluao awmauma awzmcowumamm "OH mudwwm £5» o«o.D .O o; v;_ u.- 0;. ex. 020 «so «Au 1 . fl . . . . . 4 3. -84- 301 ;o uu/1unoo 6.0— -35- was 4.3 billion cells/ml. The relationship between maximum population and maximum toxin production by toxigenic strains 0f.§°.EE£EE§ is not clear. Genigeorgis (1966), working with strain S-6 and McLean_§t_§l. (1968) working with strain 243, both of which produce type B enterotoxin, reported the first indication of toxin occurred as the population entered the stationary phase of growth. Because of this observation, Genigeorgis suggested that toxin production was the result of the accumulation of inhibitory materials in the growth medium. The observations reported herein, suggest that toxin production is as- sociated with active growth of the organisms and toxin production ceases as the population stops active division. Table II contain data which show that the growth of the type B toxin producing strain of §. EELEEE caused the pH of the BHI broth buffered at pH 6.0 to decrease after 9.0 hours until a minimum of 5.75 was reached at 12.0 hours. The pH then increased until a maximum of 8.9 was attained after 48.0 hours of growth. Type B toxin was not detected in cheese manufactured from milk having poor acid development and a high p0pulation of a strain of .§°.§£L£2§ producing type B toxin during ripening. Maximum staphylococcal populations of 20, 38, and 480 million organisms/g of cheese were observed for three separate cheeses during ripening. When the type B toxin producing strain of §..g2£gg§ was grown in BHI broth at pH 6.0, toxin was detected at a population of l80 million organisms/ml. Theoretically then, the Cheddar cheese from vat 6 which contained 480 million staphylococci/g at one time during curing should have been vaomuooH vmumuuamocoo unaumnuwasm An oo.~wm 00.5 ooo.ooo.o ow.H Hz om.o ooo.ooo.o om.H oo.~mm om.o ooo.ooo.o om.u oo.¢o oo.e ooo.ooo.m oe.H oo.Nm mk.m ooo.ooq oH.H Hz ow.m ooo.o¢m mm.o ,m oo.q om.m ooo.omm Nk.o ,4 An -- 00.0 ooo.oo~ an.o Hz oo.o coo.m¢ em.o Hz oo.o ooo.a~ mH.o 92 00.0 mo.o Amaz oo o oo o Aemuusao 000v A15 ONOV HE\conu m: we assoc wanmw> woumawumm .n.o o guess He\ooo.ok~ u z .0 um um woumnsoafi can oumnamonm esficom z «.0 nufla 0.0 um um wonwmmsn .nuoun Ham Nm.m fig aBOHw mmB awxououmudm m mama wawosuoum mamusm 2W amn3 cowumuuamunoo GHxOuoumuao paw .uamumaummsm mnu mo mm .unsoo maan> vmumafiumm .15 owe um huwmamv Hmowumo .oEHH wmumou uoz Am "HH manna -37- toxic. No explanation as to why toxin was not detected in the cheese from vat 6 is offered, but apparently environmental factors are in- volved. Growth and toxin production by the type B toxin producing strain of_§._gg£§g§ in BHI broth at pH 8.0 is indicated by data in Table l2. Toxin was first detected after 9.5 hours of growth when the population was approximately 350 million cells/ml which is higher than when toxin was first detected in the BHI broth at pH 7.4 and 6.0. At the time when toxin was first detected in the BHI at pH 7.4, the staphylococcal population was 47 million organisms/ml. The staphylococcal population in the BHI at pH 6.0 was 180 million/ml when toxin was first detected. The maximum toxin concentration observed in the BHI broth at pH 8.0 was 32 ug/ml after 24 hours of growth which is considerably lower than the 382 ug/ml produced in the control BHI and BHI at pH 6.0. The maximum population occurring in the BHI broth at pH 8.0 was approxi- mately 5 billion organisms/ml, which corresponds to a maximum population of 6 billion organisms/ml in the control BHI at pH 7.4. Apparently the BHI broth at pH 8.0 is inhibitory to toxin production by the strain 0f.§°.i!£22§ producing type B toxin but is only slightly inhibitory to cell growth. Cellular growth but reduced toxin production were also observed when the type A toxin producing strain of §°.2!£EE§ was grown in BHI broth at pH 8.0. No explanation of why this occurred can be made, but the phenomenon merits investigation. waomnooH wmumuuamoaoo undumcuomam An woummu uoz Am oo.~m oh.w ooo.ooH.o om.H o.Nk oo.~m om.k ooo.ooo.m me.H o.mq oo.~m om.“ ooo.ooo.m me.H o.em oo.w om.k ooo.ooo.m o¢.H o.~H . oo.H om.k ooo.oon.~ om.H o.oH 00 .w Ho.o no.“ ooo.omm om.o m.m An -- ck.» ooo.om~ Nk.o o.m Hz oh.“ ooo.mm Ne.o m.m Hz om.“ coo.om o~.o o.m 92 cm.“ ooo.m~ es.o m.~ Amaz oo.w oo.o o.o Aeouuaao 000v Ana omov Auev HE\Gflxou w: mm unsoo manmfi> vmumawumm .n.o mafia O :uoun Ha\ooo.um u z .0 km um amumnsonu was menu 2 ~.o an“; o.m mm um wmummmsn .cuoun Hmm xn.m aw :3ouw mwa saxououmudm m mmhu wawosvoua magnum aw smcz aofluwuuamoaoo saxouououco can .uamumaummsm mnu mo mm .uasoo manww> woumafiumm .18 omo um hufimcoc Hwowumo .mafla "NH wanma SUMMARY AND CONCLUSIONS Milk treated with 0.49 to 0.54 Units of penicillin per milliliter of milk and inoculated with toxigenic strains of Staphylococcus aureus, yielded cheese which contained high p0pulations of staphylococci during ripening. One of the Cheddar and one of the colby cheeses manufactured from milk treated with penicillin and inoculated with a strain of .§°.22£22§ producing type A enterotoxin was found to contain enterotoxin during ripening at 48 F. The toxic Cheddar cheese had a maximum_§._§!£§u§ population of 110 million organisms/g of cheese after 30 days of curing. Toxin was formed in this Cheddar cheese between 20 and 30 days of ripening. The maximum amount of toxin detected in the cheese was 4 ug/lOO g of cheese after 30 days of curing. The toxic colby cheese had a maximum staphylococcal population of 6.5 million organisms/g of cheese after 20 days of curing. Toxin was formed in the colby cheese between i and 10 days of ripening. The maximum amount of toxin detected in the cheese was 2 pg/lOO g of cheese. No toxin was detected in the cheese manufactured from milk inocu- lated with the strain 0f.§'.22L22§ producing type B enterotoxin. However, one of the Cheddar cheeses manufactured from milk containing penicillin and inoculated with the §._gg£§u§ producing type B toxin, contained 480 million staphylococci per gram of cheese after 60 days of curing. Growth of the types A and B enterotoxin producing strains of_§. _§urgg§ was inhibited in buffered BHI broth at pH 5.4. Growth of the strains in non-buffered BHI broth at pH 5.4 appeared to be uninhibited. -39- -90- The type A producing strain of S. aureus produced maxima of 8, 8 and 6 pg of toxin/ml of supernatant in aerated BHI broth with respective pH values of 6.0, 7.4 and 8.0. These maximum toxin con- centrations were present after 12, 48, and 32 hours of growth at 37 C in BHI broth with the same reSpective pH values. None of the BHI broths inoculated with the type A producing strain 0f.§°.22£22§ con- tained toxin before at least 270 million organisms/ml were present in the culture. This population relates reasonably well to the maximum of llO million §..gu£gu§ organisms/g that occurred during ripening of the toxic Cheddar cheese. The toxic colby cheese contained only 6.5 million staphylococci/g when the maximum population occurred. The type B toxin producing strain 0f.§°.22£22§ produced maxima of 382, 382, and 32 pg of toxin/ml of supernatant in aerated BHI broth with respective pH values of 6.0, 7.4 and 8.0. The toxin concentration was at a maximum after 24 hours of growth at 37 C in BHI broth with these same respective pH values. Toxin was first indicated in the BHI broth at pH 7.4 when the g. EELEEE population of 40 million organisms/ml was present in the culture. No explanation is offered as to why toxin was not detected in the Cheddar cheese made from milk inoculated with the type B enterotoxin producing strain of §'.§2£EE§ and which attained 480 million staphylococci/g during ripening. 10. 11. 12. 13. LITERATURE CITED Barber, M. A. 1914. Milk poisoning due to a type of Staphylo- coccus aureus. Phillipine J. of Sci., 2g515-519. Bergdoll, M. S., J. L. Kadavy, M. J. Surgalla and G. M. Dack. 1951. Partial purification of staphylococcal enterotoxin. Arch. Biochem.,ggz259-262. Bergdoll, M. S. 1956. The chemistry of staphylococcal entero- toxin. Ann. New Ybrk Acad. Sci., 65:139-143. Bergdoll, M. S., M. J. Surgalla and G. M. Dack. 1959. Staphylo- coccal enterotoxin. Identification of a Specific precipitating antibody with enterotoxin neutralizing prOperty. J. Immunol., _8_3_:334-338. Bergdoll, M. S., H. Sugiyama, and G. M- Dack. 1959. Staphyloco- ccal enterotoxin. 1. Purification. Arch. Biochem. BiOph., .§§:62-69. Bergdoll, M. S., H. Sugiyama, and G. M. Dack. 1961. The recovery of staphylococcal enterotoxin from bacterial culture super- natants by Ion Exchange. J. Biochem. Microbiol. Techn. Engin., 3:41-50. Bergdoll, M. S. 1962. The chemistry and detection of staphylo- coccal enterotoxin. Am. Meat Inst. Circular No. 70, p. 47. Bergdoll, M. S. 1963. Microbiological quality of foods. Academic Press, New YOrk, pp. 54-58. Bergdoll, M. S., C. A. Borja, and R. M. Avena. 1965. Identification of a new enterotoxin as Enterotoxin C. J. Bacteriol., 29;1481- 1485. Bergdoll, M. S. 1967. In Biochemistry of Some Foodbourne Microbial Toxins. The M.I.T. Press, Cambridge, Massachusetts. Brandly, P. J. 1965. Trichinosis, botulism, staphylococcal food poisoning: prevention and control. WOrld Symp. Vet. Food Hygienists. Casman, E. P. 1958. Serological studies of staphylococcal entero- toxin. Pub. H. Rep.,,1§:599-609. Casman, E. P. 1960. Further serological studies of staphylococcal enterotoxin. J. Bacteriol.,'12:890-856. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. -92- Casman, E. P., M. S. Bergdoll, and J. Robinson. 1963. Designa- tion of staphylococcal enterotoxins. J. Bacteriol., §§: 715-716. Casman, E. P. and R. W. Bennett. 1963. Culture medium for the production of staphylococcal enterotoxin A. J. Bacteriol., 86:18-23. Casman, E. P., D. W. McCoy and P. J. Brandly. 1963. Staphylo- coccal growth and enterotoxin production in meat. Appl. Microbiol., 11:498-500. Casman, E. P. and R. W. Bennett. 1965. Detection of staphylo- coccal enterotoxin in food. Appl. Microbiol., 12:181-189. Casman, E. P. 1966. Recent advances in the microbiology of food- borne diseases. Staphylococcal food poisoning. Address presented Nov. 1, 1966, at Ann. Meeting of Am. Pub. H. Assn. Casman, E. P. 1967. Personal communication. Casman, E. P., R. W. Bennett, A. E. Dorsey, and J. A. Issa. 1967. Identification of a fourth staphylococcal enterotoxin as Enterotoxin D. J. Bacteriol., 23:1875-1882. Coons, A. H., J. H. Creech, and R. N. Jones. 1941. Immunological prOperties of an antibody containing A fluorescent group. Proc. Soc. Exper. Biol. and Med., 41:200-202. Crowle, A. J. 1958. A simplified microdouble-diffusion agar precipitin technique. J. Lab. Clin. Med., 22:784-787. Crowle, A. J. 1961. Immunodiffusion. Academic Press, Inc. Dack, G. M., W. E. Cary, 0. Woolpert and H. Wiggers. 1930. Out- break of food poisoning proved to be due to a yellow hemolytic staphylococci. J. Prev. Med.,‘4:167-175. Dack, G. M. 1956. Food Poisoning. University of Chicago Press, Chicago. Third Edition. Dack, G. M. 1962. Staphylococcal enterotoxin. In Chemical and Biological Hazards in Foods. Ia. State Univ. Press, Ames, pp. 320-329. Dack, G. M. and G. Lippitz. 1962. Fate of staphylococci and enteric microorganisms introduced into slurries of frozen pot pies. Appl. Microbiol., 19:472-479. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. -93- Dauer, C. C. 1961. Summary of disease outbreaks. Pub. H. Rep., 16:915-922. Davison, E. and G. M. Dack. 1939. Some chemical and physical studies of staphylococcus enterotoxin. J. Infect. Dis., 66:302-306. Denny, C. B., P. L. Tan, and C. W. Bohrer. 1966. Heat inactiva- tion of staphylococcal enterotoxin A. J. Food Sci., 615762- 767. Dolman, C. E., R. J. Wilson, and W. H. Cockroft. 1936. A new method for detecting staphylococcus enterotoxin. Canad. Pub. H. Journ., 61:68-71. Dolman, C. E. and R. J. Wilson. 1938. Experiments with staphylo- coccal enterotoxin. J. Immun., 26:13-30. Dolman, C. E. 1939. Staphylococcus enterotoxin. Address presented at the Sixth Pacific Science Congress, held at Berkeley, Stanford, and San Francisco. Dolman, C. E. and R. J. Wilson. 1940. The kitten test for staphylococcus enterotoxin. Canad. Pub. H. Jour., 21:68-71. Donnelly, C. B., J. E. Leslie, and L. A. Black. 1968. Production of enterotoxin A in milk. Appl. Microbiol., 16:917-924. Evans, J. B. and C. F. Niven. 1950. A comparative study of known food-poisoning staphylococci and related varieties. J. Bacteriol., 62:545-550. Evans, J. B., L. G. Buettner, and C. F. Niven, Jr. 1950. Evalu- ation of the coagulase test in the study of staphylococci associated with food poisoning. J. Bacteriol., 66:481-484. Frazier, W. C. 1967. Food Microbiology. McGraw-Hill Book Co., Inc., New Ybrk. Second Edition. Genigeorgis, C. 1966. Studies on the production and identification of staphylococcal enterotoxin in food. Ph.D. Thesis, Univ. Calif. Genigeorgis, C. and W. W. Sadler. 1966. Immunofluorescent detection of staphylococcal enterotoxin B II. Detection in foods. J. Food Sci., 11:605-609. 41. 42. 43. 44. 45. 46. 47. 48. 4'9. 50. 51. 52. -94- Hall, H. E., R. Angelotti, and K. H. Lewis. 1963. Quantitative detection of staphylococcal enterotoxin B in food by gel- diffusion methods. Pub. H. Rep.,‘12:1089-1098. Hall, H. E., R. Angelotti, and K. H. Lewis. 1965. Detection of the staphylococcal enterotoxins in food. H. Lab. Sci., 23179-191. Hammon, W. McD. 1941. Staphylococcus enterotoxin. An improved cat test, chemical and immunological studies. Am. J. Publ. H., 2131191-1198. Harmon, L. G. 1967. Significance and importance of staphylococci in the dairy and food industry. Milk Dealer, 56:pp. 16, 18, 30. Hausler, W. J., Jr., E. J. Byers, Jr., L. C. Scarborough, Jr., and S. L. Hendricks. 1960. Staphylococcal food intoxication due to Cheddar cheese II. Laboratory evaluation. J. Milk and Food Tech., 2231-6. Haynes, W. C. and G. J. Hucker. 1946. A review of micrococcus enterotoxin food poisoning. Food Res. .11:281-297. Hendricks, S. L., R. A. Belknap, W. J. Hausler, Jr. 1959. Staphylococcal food intoxication due to Cheddar cheese. I. Epidemiology. J. Milk and Food Tech., 223313-317. Hibnick, H. E. and M. S. Bergdoll. 1959. Staphylococcal entero- toxin. II. Chemistry. Arch. Biochem. BiOph., 66370-73. HOpper, S. H. 1963. Detection of staphylococcus enterotoxin. I. Flotation antigen-antibody system. J. Food Sci., 26:572-577. Jordan, E. 0., G. M. Back and 0. J. Woolpert. 1931. Effect of heat, storage and chlorination on toxicity of staphylococcus filtrates. J. Prev. Med., 6:383-386. Jordan, E. 0. and J. McBroom. 1931. Results of feeding staphylo- coccal filtrates to monkeys. Proc. Soc. Exper. Biol. Med., 29:161-162. Jordan, E. 0. and W. Burrows. 1933. Nature of the substance causing staphylococcus food poisoning. Proc. Soc. Exper. Biol. Med., 293448-449. 53. 54'. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. -95- Kato, E., Mahmood Khan, L. Kujovich, and M. S. Bergdoll. 1966. Production of enterotoxin A. Appl. Microbiol., 16:966-972. Lechowich, R. V., J. B. Evans, and C. F. Niven, Jr. 1956. Effect of curing ingredients and procedures on the survival and growth of staphylococci in an on cured meats. Appl. Microbiol.,‘6:360-363. MacDonald, A. 1944. Staphylococcal food poisoning caused by cheese. Monthly Bull. Ministry of Health and Emergency Pub. H. Lab. Ser. Med. Res. Council 33121. Matheson, B. H. and F. S. Thatcher. 1955. Studies with staphylo- coccal toxin. I. A reappraisal of the validity of the kitten test as an indication of staphylococcal enterotoxin. Can. J. Microbiol.,.1:372-38l. McLean, R. A., H. D. Lilly and J. A. Alford. 1968. Effects of meat-curing salts and temperature on production of staphylo- coccal enterotoxin B. J. Bacteriol., 26:1207-1211. Milone, N. A. 1961. On the use of tissue cultures for bioassay of staphylococcal enterotoxin. Am. Pub. H. Assn. 89th Ann. Meeting, Detroit, Mich., Nov. 16, 1961. Oakley, C. L. and A. J. Fulthorpe. 1953. Antigenic analysis by diffusion. J. Path. Bacteriol., 66349-60. Ouchterlonly, 0. 1953. Antigen-antibody reactions in gels. Acta. Path. et Microbiol. Scandin., 263507-515. Oudin, J. 1952. Specific precipitation in gels. Methods Med. Res.,.6:335-378. Raj, H. and J. Listen. 1962. Fish bioassay for thermostable toxins of staphylococci. Bact. Proc. Abstr. 82nd Ann. Meeting p. 65. Read, R. B., Jr., J. Bradshaw, W. L. Pritchamd and L. A. Black. 1965. Assay of staphylococcal enterotoxin from cheese. J. Dairy Sci.,‘§:420-424. Read, R. B., Jr., and J. G. Bradshaw. 1966. Thermal inactivation of staphylococcal enterotoxin B in veronal buffer. Appl. Microbiol.,l16:130-l32. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. -96- Read, R. B., Jr., and J. G. Bradshaw. 1967. Gamma irradiation of staphylococcal enterotoxin B. Appl. Microbiol., 16: 603-605. Robinson, J. and F. S. Thatcher. 1965. Determination of staphylo- coccal enterotoxin by indirect hemagglutination inhibition procedure. Bacteriol. Proc. Abstr., 65th Ann. Meeting, p. 72. Robinton, E. D. 1950. A rapid method for demonstrating the action of staphylococcus enterotoxin upon Rana pipiens. Yale J. Biol. Med., 22394-98. Segalove, M. and G. M. Dack. 1941. Relation of time and tempera- ture to growth and enterotoxin production of staphylococci. Food Res., 6:127-133. Sugiyama, H., M. S. Bergdoll and G. M. Dack. 1960. In vitro studies on staphylococcal enterotoxin production. J. Bacteriol. 66:265-270. Surgalla, M. J. and G. A. Hite. 1945. A study of enterotoxin and alpha and beta lysins production by certain staphylo- coccal cultures. J. Inf. Dis.,‘Z6z78-82. Surgalla, M. J., M. S. Bergdoll, and G. M. Back. 1952. Use of antigen-antibody reactions in agar to follow the progress of fractionation of antigenic mixtures: Application to purification of staphylococcal enterotoxin. J. Immunol., 62:357-365. Surgalla, M. J., M. S. Bergdoll, and G. M. Dack. 1953. Some observations on the assay of staphylococcal enterotoxin by the monkey-feeding test. J. Lab. Clin. Med., 61:782-788. Surgalla, M. J., M. S. Bergdoll, and G. M. Back. 1954. Staphylo- coccal enterotoxin: Neutralization by rabbit antiserum. J. Immunol.,l22:398-403. Surgalla, M. J. and G. M. Back. 1955. Enterotoxin produced by micrococci from cases of enteritis after antibiotic therapy. J. Am. Med. Assn., 158:649-650. Takahashi, I. and C. K. Johns. 1959. Staphylococcus aureus in Cheddar cheese. J. Dairy Sci., 62:1032-1042. Thatcher, F. S. and B. H. Matheson. 1955. Studies with staphylo- coccal toxin. II. The specificity of enterotoxin. Can. J. Microbiol.,‘1:382-400. 77. 78. 79. 80. 81. 82. 83. 84. -97- Thatcher, F. S. and J. Robinson. 1962. Food poisoning: An analysis of staphylococcal toxins. J. Appl. Bacteriol, 26: 387-388. Van Slyke, L. L. and W. V. Price. 1952. Cheese. Orange Judd Publishing Co., Inc. New York. Vaughn, A. C. 1884. Poisonous or sick cheese. Pub. H. Papers and Rep. Am. Pub. H. Assn., 193241-245. Wadsworth, C. A. 1957. Slide microtechnique for the analysis of immune precipitates in gel. Internat. Arch. Allergy. Appl. Walker, G. C., L. G. Harmon, and C. M. Stine. 1961. Staphylococci in colby cheese. J. Dairy Sci., 99:1272-1282. Zebovitz, E., J. B. Evans, and C. F. Niven, Jr., 1955. Tellurite- glycine agar: A selective plating medium for the quantitative detection of coagulase-positive staphylococci. J. Bacteriol., 193686-688. Zehren, V. L. and V. F. Zehren. 1968a. Examination of large quantities of cheese for staphylococcal enterotoxin.A. J. Dairy Sci., 61:635-644. Zehren, V. L. and V. F. Zehren. l968b. Relation of acid deve10p- ment during cheesemaking to deve10pment of staphylococcal enterotoxin A. J. Dairy Sci., 613645-649. i‘lICHIGQN STATE UNIV. LIBRQRIES I III II lllllll l 31293008090387