Deoxyribonucleic Acid Base Composition and Biochemical Properties of Certain Coagulase-Negative Enterotoxigenic

Transcription

1 Appum MICROBIOLOGY, Feb. 1975, p American Society for Microbiology Vol. 29, No. 2 Printed in U.SA. Deoxyribonucleic Acid Base Composition and Biochemical Properties of Certain Coagulase-Negative Enterotoxigenic Cocci LEONARD P. LOTTER AND CONSTANTIN A. GENIGEORGIS* Institut fur Lebensmittelhygiene der Frein Universitat Berlin, 1 Berlin 33, and Department of Epidemiology and Preventive Medicine, Veterinary School, University of California, Davis, California 95616* Received for publication 17 September 1974 Eight coagulase-negative, enterotoxigenic strains of cocci and one weakly coagulase-positive strain isolated from a number of different sources, including cases of food poisoning incidents, were evaluated for their relationship to Staphylococcus aureus on the basis of deoxyribonucleic acid (DNA) buoyant density and physiological studies. One strain of cocci produced enterotoxins A and C, two strains produced types B and C, four strains produced only type C, and one strain only type D. The enterotoxin produced by one strain of cocci was serologically untypable. None of the test organisms produced detectable amounts of enterotoxin in broth cultures. The test strains of cocci exhibited the following profile: all produced catalase; all grew anaerobically and fermented glucose; five were sensitive to lysostaphin; the percentage of guanine plus cytosine content of their DNA varied from 32.7 to 37.6; five produced acid from mannitol both aerobically and anaerobically; two formed 6-hemolysin; five produced phosphatase and acetoin; and all produced heat-stable nuclease. None of the organisms exhibited typical characteristics of S. aureus, S. epidermidis, or S. saprophyticus. On the basis of the present data and data reported elsewhere, these organisms should be considered as variants or mutants of S. aureus. The food industry has routinely concerned itself with coagulase-positive staphylococci because these microorganisms are presumed to produce one or more enterotoxins. The early studies of Evans et al. (13) suggested that coagulase-negative strains are non-enterotoxigenic and, hence, could be neglected when isolated from foods. Enterotoxin production by micrococci and coagulase-negative staphylococci has been reported in earlier studies (19, 28). In 1959, Omori and Kato (32) reported a food poisoning outbreak caused by coagulase-negative staphylococci among Japanese high school students. Several other coagulase-negative, enterotoxigenic cocci were recovered from normal foods, foods suspected to contain enterotoxin, from a fatal case of human enteritis (M. S. Bergdoll, K. F. Weiss, and M. J. Muster, Bacteriol. Proc., p. 12, 1967; Bergdoll, personal communication), and from cases of food-borne gastroenteritis (9). All of these strains produced low levels of enterotoxin which could be detected only when 30- to 50-fold concentrated culture filtrates were fed to monkeys or when tested by the micro- Ouchterlony slide technique (10). This result [52 may indicate why Evans et al. (13) did not detect enterotoxins among their coagulasenegative staphylococci, presumably because these strains might have formed low amounts of enterotoxin. These reports have aroused extensive interest in the public health significance (enterotoxigenesis) of coagulase-negative staphylococci. An attempt has been made to classify coagulasenegative enterotoxigenic staphylococci (Bergdoll et al., Bacteriol. Proc., p. 12, 1967), but their data are not sufficient to classify these strains according to the presently acceptable schemes for the Micrococcaceae (3, 4). It would be interesting to know how coagulase-negative, enterotoxin-positive strains relate to classical enterotoxigenic strains of Staphylococcus aureus. Should species other than S. aureus form enterotoxin, changes will be necessary in the current approaches for microbiological evaluation of normal foods or foods implicated in food poisoning, thus allowing appropriate classification of the isolated cocci. In this paper, we will report the deoxyribonucleic acid (DNA) base composition and biochemical properties of eight coagulase-negative

2 VOL. 29, 1975 COAGULASE-NEGATIVE ENTEROTOXIGENIC COCCI strains isolated originally as coagulase negative and one weakly coagulase-positive strain of enterotoxigenic cocci. We suggest that they should be considered as mutants or variants of S. aureus. We also confirm and extend the observation made by Bergdoll and co-workers concerning enterotoxin production and other characteristics of these strains. MATERIALS AND METHODS Bacteria. All the coagulase-negative, enterotoxigenic cocci (Table 1) and S. aureus strain 137 (ATCC 19095) were furnished by M. S. Bergdoll, Food Research Institute, Madison, Wisc.; S. epidermidis BP23 and Micrococcus lysodeikticus (ATCC ISU-1) were supplied by R. V. F. Lachica of this department; and S. aureus strains 243 (ATCC 14458) and Wood 46 were originally obtained from the late E. P. Casman of the Food and Drug Administration, Washington, D.C. Enterotoxin production in sac cultures. The sac culture assembly method of Donnelly et al. (12) was used to obtain maximal yields of enterotoxin. Several sac culture assemblies after inoculation were incubated on a 37 C reciprocating shaker for 24 h. The liquid culture surrounding the sac was removed from the flask and was centrifuged at 23,000 x g for 15 min. The supernatant fluid was frozen until used. Enterotoxin production in broth cultures. A loopful of a culture on a brain heart infusion (BHI) agar slant (Difco Laboratories, Detroit, Mich.) was added to 8 ml of BHI broth contained in a 15-ml micro-fernbach flask. All inoculated flasks were incubated on a 37 C reciprocating water bath shaker for 24 h. The liquid culture was pipetted from the flask and was centrifuged at 23,000 x g for 15 min. The supernatant fluid was frozen until used. Supernatant fluids from sac and broth cultures were analyzed by the micro-ouchterlony slide technique (10) or by the single gel-diffusion tube test (40). Reference enterotoxins and antisera. Purified enterotoxins A and B were supplied by E. H. Schantz TABEi 1. and M. S. Bergdoll, both of the Food Research Institute, University of Wisconsin, Madison. Reference enterotoxin D (strain 315) and anti-enterotoxin D serum were supplied by R. W. Bennett, Division of Microbiology, Food and Drug Administration, Washington, D. C. Specific type A, B, and C rabbit antisera and purified enterotoxin C were prepared in this laboratory (17). Immunodiffusion tests. The micro-slide technique (10) and the single gel-diffusion tube method (40) were employed. Lyophilization. Sac culture supernatant fluids were lyophilized and reconstituted with standard phosphate-buffered saline (ph 7.2, 0.02 M) plus 2% NaCl for gel-diffusion studies. Additional characteristics of coagulase-negative enterotoxigenic cocci. Hemolysins were determined by the method of Nakagawa (29). Egg yolk lipase was observed on the egg yolk agar medium of Garber et al. (16). Proteinase was assayed by the modified agar plate method of Sandvik (34) including 1% (wt/vol) sodium caseinate. Penicillinase was detected by means of the overlay plate technique of Adams et al. (1) using dicloxacillin-induced colonies on Mueller- Hinton agar, Novick's reagent (N-phenyl-1-naphthylamine-azo-O-carboxybenzene), and aqueous 10% (wt/vol) penicillin as substrate. Nitrate reductase was determined by the technique of Conn et al. (11). Lysostaphin sensitivity was assayed by the tube method of Lachica et al. (25) and lysozyme sensitivity by the tube method of Marmur (28). Heat-stable nuclease detection included precipitation of unhydrolyzed DNA by 1 N HCO in the micro-slide DNA agar gel technique of Jarvis and Lawrence (20). The coagulase tube test included coagulase plasmaethylenediaminetetraacetic acid (Difco), and the procedure recommended by the manufacturer was followed. Phosphaase was assayed on the phenolphthalein diphosphate agar medium of Barber and Kuper (5). Acetoin production was determined by Barritt's (6) modification of the Voges-Proskauer reaction. Acid production from glucose and mannitol, aerobically and anaerobically, was measured by means of the techniques of Baird-Parker (3). Isolation of DNA. The methods of Marmur (28) and Klesius and Schuhardt (22) were employed to isolate DNA of test and control strains during the Source, place, and date of isolation of coagulase-negative, enterotoxigenic cocci Strain Source Place of isolation Date of isolation 217 Potato salad Chicago Health Dept., Chicago, Ill. 10 February Cheese Kraft Foods Co., Chicago, Ill. NKa 248 Intestinal tract of boy Children's Hospital, Buffalo, N.Y. 24 August 1953" dead from enteritis 304 Frozen chicken pie Food Research Institute, Madison, Wisc. November NK Laboratory of G. Omori, Osaka, Japan 18 January 1961" 492 Roast beef Connecticut July 1967" 698 Sputum Mayo Clinic, Rochester, Minn. 17 September 1968" 724 Skewered beef kabob Center for Disease Control, Denver, Colo. October 1969 W990-1 NK Laboratory of H. Blobel, Univ. of Wis- 16 November 1960" consin, Madison a NK, Not known. " Date received at Food Research Institute, Madison, Wisc. 153

3 154 LOTTER AND GENIGEORGIS APPL. MICROBIOL. logarithmic phase of growth. Depending on the respective sensitivity of cells, either lysozyme (Nutritional Biochemical Corp., Cleveland, Ohio) or lysostaphin (Mann Research Laboratories, Orangeburg, N.Y.), each at a final concentration of 25 ug/ml, was added to lyse cells prewashed twice by 0.15 M NaCl and 0.1 M disodium ethylenediaminetetraacetic acid for 30 to 60 min at 37 C (22). In the case of strains 248 or 352, which were lysozyme and lysostaphin resistant, the cells were disrupted by a Braun MSK Mechanical Cell Homogenizer (Bronwill Scientific Inc., Rochester, N.Y.) at 3,000 rpm for 1 min at 4 C. DNA determination. The DNA concentration was estimated by ultraviolet absorption of 260 nm with a DU spectrophotometer (Beckman Instruments, Inc., Fullerton, Calif.). The optical density at 260 nm is related to DNA concentration in such a manner that an optical density of 1 equals 50 ug of DNA per ml. The DNA content of samples ranged from 10 to 50 Mg/Ml. Percentage of guanine plus cytosine content of DNA. The mole percentage of guanine plus cytosine present in DNA was determined by buoyant density in cesium chloride (CsCl) using the method and equation of Schildkraut et al. (35). Micrococcus Lysodeikticus DNA (Miles Laboratories, Elkhart, Ind.) with a density of g/cm2 was selected as density reference for the test DNA. The CsCl gradient centrifugation was done at 25 C in a Spinco model E analytical centrifuge (Beckman Instruments, Inc., Fullerton,. Calif.) at 140,000 x g for 19 to 21 h. Tracings of the ultraviolet absorption bands of DNA at equilibrium were provided by a recorder attachment to the ultracentrifuge. Protein A determination. Protein A, on the surface of the coccal cells, was detected by means of a fluorescent antibody test where protein A-positive cells bind immunoglobulin G antibody non-specifically (27). Preparation of fluorescent normal human immunoglobulin G conjugates and staining of coccal strains was done according to methods reported previously (18; C. A. Genigeorgis and W. W. Sadler, Bacteriol. Proc., p. 1, 1969). RESULTS- Enterotoxm production by coagulase-negative cocci. Table 2 indicates the types of enterotoxins produced by the test strains as determined by this laboratory and by Bergdoll and co-workers. In repeated immunodiffusion tests, strain 492 did not react with antienterotoxin A serum, but did react with antienterotoxin C serum. Strain 230 reacted with anti-enterotoxin B and C sera. It may be noted that in the case of reference coagulase-positive strains 243 and 137, about four times as much enterotoxin was formed by sac as compared to broth cultures. Sac cultures of strains 217, 230, 248, 492, and W990-1 produced measurable amounts of enterotoxin as shown in the single gel-diffusion test. Demonstration of enterotox- TABLE 2. Strain Enterotoxigenic types of coagulasenegative cocci Enterotoxin Entertosintype type Concn of enter- otoxin (;&g/ml) la ][[b Sac Broth culturec cultured 217 A A 8.0e C C 25.0d C B 6.8d 0 C 1.4d 248 C C 1.4d C C 9.6' C C 6.O A,B B 4.0d 0 C? C 2.3d D? D 4.0' 0 724?? 0.0' 0' W990-1 C? C 1.4d 0 S. aureus (ATCC B B 230.0d ) 243 S. aureus (ATCC C C 62.0d ) 137 IEnterotoxin type determined at Food Research Institute, Madison, Wisc. b Enterotoxin type determined in our laboratory. c Sac culture technique of Donnelly et al. (12). d Enterotoxin concentration determined by gel-diffusion tube procedure (40). ' Enterotoxin concentration determined by the micro-slide technique (10). ' Same as c, but sac culture supernatant fluids were lyophilized and reconstituted with 2 ml of standard phosphate-buffered saline (ph 7.2, 0.02 M) plus 2% NaCl. ' Negative for enterotoxins -A, B, C, and D. ins in sac cultures of strains 304 and 352 required concentration of sac supernatant fluids by lyophilization. In repeated immunodiffusion tests of strain 724 sac supernatant fluids, no enterotoxins A, B, C, or D were detected. Additional characteristics of coagulasenegative, enterotoxigenic cocci. Of the nine strains of cocci examined, one strain, 217, formerly reported to be coagulase-negative, was found to be weakly coagulase-positive in the rabbit plasma standard tube test, both in our laboratory and in M. S. Bergdoll's laboratory (personal communication). The rest of the strains showed no coagulasq activity in sac cultures. Table 3 illustrates some additional characteristics of the coagulase-negative, enterotoxigenic cocci studied and their classification according to the 1966 diagnostic scheme of Baird-Parker (3) and by percent guanine plus cytosine content. All test strains and control S. aureus strains were catalase-positive; they did not indicate any change in their utilization of

4 VOL. 29, 1975 COAGULASE-NEGATIVE ENTEROTOXIGENIC COCCI mannitol under aerobic or anaerobic conditions, and they formed acid from glucose, both aerobically and anaerobically. Control S. epidermidis BP23 utilized mannitol aerobically, but not anaerobically. M. lysodeikticus ISU-1 yielded acid from glucose only aerobically (3). The percent guanine plus cytosine content of staphylococci was found to range from 30 to 40%, whereas that of micrococci ranged from 65 to 75% (2, 22, 23, 39). The fluorescent antibody technique, used in the detection of protein A, indicated that strains 217 and 698 were strongly positive, whereas the remaining strains were weakly or questionably positive. Control strains S. aureus Wood 46, S. epidermidis BP23, BP47, and M. lysodeikticus ISU-1 were negative and S. aureus strains 137, 243, S-6, and 196-E were positive as expected. DISCUSSION Enterotoxin production by the eight coagulase-negative strains and the weakly coagulasepositive strain, 217, was considerably enhanced when the sac assembly culture technique of Donnelly et al. (12) was used. Eight of the nine test strains (excluding strain 724) formed low levels of enterotoxin. The detection of enterotoxin types for these strains in our laboratory supports and extends the findings of Bergdoll (personal communication). The minor variations in the enterotoxin patterns (Table 2), as determined by the two laboratories, might be due to differences in methods used, or loss of the ability to produce a certain type of enterotoxin by a test organism. Strain 724 produced no enterotoxin of type A, B, C, or D; possibly, it may form enterotoxin E, which has recently been purified (8), or some other unknown type of enterotoxin. Breckinridge and Bergdoll (9) reported that strain 724, when grown in concentrated culture and fed to six monkeys, caused vomiting in four of the six monkeys 2 to 3 h later. The enterotoxicity of this strain cannot be attributed to jb-hemolysin, which has been associated with an emetic response (38), because 724 was found to lack a-,,b-, and 6-hemolysins (Table 3). On the basis of the anaerobic growth and fermentation of glucose, all strains should be considered as staphylococci (3, 4). Sensitivity to lysis by lysostaphin has been reported to be a characteristic of staphylococci, but not micrococci (4, 25), with S. aureus being more sensitive to lysostaphin than S. epidermidis (37). On the basis of this criterion, five of the strains should be considered as staphylococci. Analysis of cell walls for ribitol of glycerol teichoic acids which are present in staphylococci, but not micrococci (4), was not performed. The DNA base composition of the test strains identifies them as belonging to the genus Staphylococcus, and not to the genus Micrococcus (2, 4, 22, 23, 36). Yet, no DNA base composition difference has been demonstrated between S. TABLE 3. Additional characteristics and classification of coagulase-negative enterotoxigenic COCCia Characteristic Strains W BP 23 ISU-2 Hemolysis _ - - Egg yolk lipase Proteinase Penicillinase Nitrate reductase... + _ Lysis by lysostaphin Lysis by lysozyme Heat-stable nuclease _ Coagulase... + (Wk) _ Phosphatase Acetoin Acid from: Arabinose... + Lactose _ Maltose _ Mannitol... _ % Guanine plus cytosine ND 72.5 Content protein A A ± Baird-Parker class (3)... I VI II IV VI VI II II II I II 7M a Wk, weak; ND, not determined. 155

5 156 LOTTER AND GENIGEORGIS APPL. MICROBIOL. aureus and S. epidermidis (39). A DNA-DNA hybridization study has indicated that S. aureus strains were related to, but distinct from, S. epidermidis (N. H. Nielsen, Bacteriol. Proc., p. 45, 1970), but this method was not used in the present study. According to the 1966 classification scheme of Baird-Parker (3) for staphylococci and micrococci, the eight coagulase-negative, enterotoxigenic cocci should be considered as strains of S. epidermidis and the weakly coagulase-positive as S. aureus. The presence of ribitol or glycerol teichoic acids in the cell walls, the sensitivity to Novobiocin, and the requirement of biotin for growth have not been used as criteria of classification (4) of the test organisms in this study. The other characteristics like production of coagulase, heat-resistant nucleases and a-toxin, presence of protein A, and finally,. utilization of glucose and mannitol suggested for distinguishing the species of the genus Staphylococcus in the most recent scheme for the classification of Micrococcaceae (4) were evaluated. On the basis of anaerobic growth and fermentation of glucose none of the strains is S. saprophyticus. On the basis of aerobic and anaerobic production of acid from mannitol, five strains behaved -like S. aureus (4). Lack of a-hemolysin production suggests that none of the strains is S. aureus (4), whereas production of heat-resistant nuclease suggests that all strains are S. aureus (4, 5, 21, 22, 26). Oeding (30) reported the absence of protein A in S. epidermidis, whereas Forsgren (14) described its presence in 98% of human strains of S. aureus. Recent findings by our group (R. V. F. Lachica et al., unpublished data) indicate that S. aureus strains, isolated from clinical cases in dogs, produce, by fluorescent antibody test, no detectable protein A. On the basis of protein A production, several of the strains in the present study should be classified as S. aureus strains (4). In summary, on the basis of DNA base composition, glucose utilization, and lysostaphin sensitivity, the data suggest that at least five of the strains used in the present study should be considered as staphylococci. The study failed to identify any of the strains as typical S. aureus, S. epidermidis, or S. saprophyticus. The fact that production of heat-resistant nuclease has been shown so far to be produced only by S. aureus supports the idea that all the entertoxigenic strains studied are variants or mutants of S. aureus. Storage of staphylococci isolated from cases of food poisoning resulted in phenotypic or genetic changes in characteristics such as coagulase production and enterotoxigenesis. H. E. Hall (Bacteriol. Proc., p. 77, 1968) studied 806 strains selected from 305 original cultures of S. aureus from cases of food poisoning and from foods. Enterotoxins A, B, and C were detected in 294 coagulase-positive and 31 coagulasenegative strains. Z. H. Markus (Bacteriol. Proc., p. 114, 1970) isolated non-toxigenic variants of S. aureus S-6, a type A and B enterotoxin producer, by surface plating on medium with type B antiserum and selection of colonies devoid of precipitation halos. He observed that these variants differed from the parental culture in formation of nuclease, coagulase, and hemolysin. The variants were not phage typable. Several investigators have isolated mutants of S. aureus, many of which had lost their ability to form coagulase as well as other physiological characteristics but still were able to revert to their original form (15, 31). Omenn and Friedman (31) observed that, in contrast to the wild-type culture of S. aureus, mutants, isolated by nitrosoguanidine treatment, lacked nuclease, coagulase, fl-hemolysin, and lysostaphin sensitivity, but all these characteristics were regained after treatment with ethyl methane sulfonate. Recently, Forsgren et al. (15) obtained protein A-deficient mutants of S. aureus by treatment with either ethyl methane sulfonate or nitrosoguanidine. More than 50% of all protein A-deficient mutants also lacked coagulase, nuclease, a-hemolysin, fibrinolysin, mannitol utilization, and a phage type pattern. Spontaneous reversion, as well as induced reversion of ethyl methane sulfonate mutants with nitrosoguanidine and of nitrosoguanidine mutants with ethyl methane sulfonate, resulted in the regain of nearly all the previously missing characteristics. Korman (24), in an earlier study of coagulase-negative mutants of S. aureus, described the pleiotropic loss of coagulase production. The present data as well as the above mentioned reports indicate that we may mislead ourselves in expecting that all strains of a species of bacteria should exhibit the same characteristics at all times. Richmond (33) suggests that it is only reasonable to expect production of a constant character under constant environmental conditions. It seems true that single or pleiotropic mutations will always complicate the identification of S. aureus even if many different criteria are used. In additional studies (L. P. Lotter, Ph.D.

6 VOL. 29, 1975 COAGULASE-NEGATIVE ENTEROTOXIGENIC COCCI thesis, Univ. of California, Davis, 1974), several attempts were made to demonstrate production of coagulase by the test strains using a number of approaches. Production of coagulase could not be demonstrated by the use of different animal and human plasmas, fibrinolysin studies, or induction of coagulase release by bovine serum albumin. Coagulase-positive variants, without change in enterotoxigenesis, were isolated repeatedly from broth and sac cultures of the corresponding coagulase-negative wild types in the case of all nine test strains. The coagulase-positive variants were obtained from sac cultures after at least 10 monthly transfers from BHI agar to BHI broth to BHI agar and storage in the refrigerator, or from BHI broth after at least 24 monthly transfers similar to the above. All selected coagulase-positive variants remained stable in regard to coagulase. The biochemical profile of the coagulase-positive variants based on a large number of characteristics was compared with that obtained from the coagulase-negative, enterotoxigenic wild type. Despite inconsistencies, a pattern emerged: the coagulase-positive variants usually showed a more rapid growth rate, produced more heatstable nuclease and enterotoxin, and gave 6- hemolysis. There is enough evidence to suggest that staphylococcal-type enterotoxins are produced only by strains, variants, or mutants of S. aureus. Even though enterotoxins have not been shown to be produced by all S. aureus strains, we suggest that the ability to produce them should be considered as a major taxonomic criterion. ACKNOWLEDGMENTS This investigation was sponsored by the Food Protection and Toxicology Center of the University of California and supported by Public Health Service research grant FD from the Food and Drug Administration. We wish to acknowledge the cooperation of Stephen A. Douglass of the Department of Food Science and Technology, University of California at Davis, for carrying out studies on buoyant density in CsCl. Furthermore, we wish to thank M. S. Bergdoll, E. J. Schantz, and R. W. Bennett for having contributed test strains and reagants used during our study. LITERATURE CITED 1. Adams, A. P., A. L. Barry, and E. J. Benner A simple rapid test to differentiate penicillin-susceptible from penicillin-resistant Staphylococcus aureus. J. Infect. Dis. 122: Auletta, A. E. and E. R. Kennedy Deoxyribonucleic acid base composition of some members of the Micrococcaceae. J. Bacteriol. 92: Baird-Parker, A. C Methods for classifying staphylococci and micrococci, p Identification methods for microbiologists. Academic Press Inc., New York Baird-Parker, A. C The basis for the present classification of staphylococci and micrococci. Ann. N.Y. Acad. Sci. 236: Barber, M., and S. W. A. Kuper Identification of Staphylococcus pyogenes by the phosphatase reaction. J. Pathol. Bacteriol. 63: Baritt, M. M The intensification of the Voges-Proskauer reaction by the addition of a-naphthol. J. Pathol. Bacteriol. 42: Borja, C. R., E. Fanning, L.-Y. Huang, and M. S. Bergdoll Purification and some physiochemical properties of staphylococcal enterotoxin E. J. Biol. Chem. 247: Barry, A. L., R. V. F. Lachica, and F. W. Atchison Identification of Staphylococcus aureus by simultaneous use of tube coagulase and thermonuclease tests. Appl. Microbiol. 25: Breckinridge, J. C., and M. S. Bergdoll Outbreak of food-borne gastroenteritis due to a coagulase-negative enterotoxin-producing staphylococcus. N. Engl. J. Med. 284: Casman, E. P., R. W. Bennett, A. E. Dorsey, and J. E. Stone The micro-slide gel double diffusion test for the detection and assay of staphylococcal enterotoxins. Health Lab. Sci. 6: Conn, H. J., M. W. Jennison, and 0. B. Weeks Routine tests for the identification of bacteria, p In H. J. Conn (ed.), Manual of microbiological methods. McGraw-Hill, New York. 12. Donnelly, C. B., J. E. Leslie, L. A. Black, and K. H. Lewis Serological identification of enterotoxigenic staphylococci from cheese. Appl. Microbiol. 15: Evans, J. B., L. G. Beuttner, and C. F. Niven, Jr Evaluation of the coagulase test in the study of staphylococci associated with food poisoning. J. Bacteriol. 60: Forsgren, A Significance of protein A production by staphylococci. Infect. Immun. 2: Forsgren, A., K. Nordstrom, L. Philipson, and J. Sjoquist Protein A mutants of Staphylococcus aureus. J. Bacteriol. 107: Garber, C. D., R. Latta, J. P. Fairchild, and E. H. Vogel, Jr Production of opalescence by staphyococci in egg yolk medium, as index to bacteriophage typability. Am. J. Clin. Pathol. 30: Genigeorgis, C., M. Foda, A. Mantis, and W. W. Sadler Effect of sodium chloride and ph on enterotoxin C production. Appl. Microbiol. 21: Genigeorgis, C., and W. W. Sadler Immunofluorescent detection of staphylococcal enterotoxin B. I. Detection in culture media. J. Food Sci. 31: Haynes, W. M. C., and G. J. Hucker Studies on Coccaceae. XVIII. The enterotoxin-producing micrococci. N.Y. State Agric. Exp. Stn. Technol. Bull. 275: Jarvis, A. W., and R. C. Lawrence A micro-method for the quantitative determination of deoxyribonuclease. Can. J. Biochem. 47: Jasper, D. E Thermostable nuclease production by staphylococci in milk samples from bovine mastitis. Am. J. Vet. Res. 34: Klesius, P. H., and V. T. Schuhardt Use of lysostaphin in the isolation of highly polymerized deoxyribonucleic acid and in the taxonomy of aerobic Micrococcaceae. J. Bacteriol. 95: Kocur, M., T. Bergan, and N. Mortensen DNA base composition of Gram-positive cocci. J. Gen. Microbiol. 69: Korman, R. Z Coagulase-negative mutants of Staphylococcus aureus: genetic studies. J. Bacteriol. 86:

7 158 LOTTER AND GENIGEORGIS APPL. MICROBIOL. 25. Lachica, R. V. F., P. D. H. Hoeprich, and C. Genigeorgis Nuclease production and lysostaphin susceptibility of Staphylococcus aureus and other catalase-positive cocci. Appl. Microbiol. 21: Lachica, R. V. F., K. F. Weiss, and R. H. Deibel Relationships among coagulase, enterotoxin, and heatstable deoxyribonuclease production by Staphylococcur aureus. Appl. Microbiol. 18: Lind, I Non-specific adsorption of FITC-labelled serum globulins to Staphylococcus aureus. Acta Pathol. Microbiol. Scand. 73: Marmur, J Procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 3: Nakagawa, M Studies on staphylococci from the bovine udder. I. Biological characteristics of staphylococci and some observations on the pathogenic strains. Jap. J. Vet. Res. 6: Qeding, P Antigenic properties of staphylococci. Ann. N.Y. Acad. Sci. 128: Omenn, G. S., and J. Friedman Isolation of mutants of Staphylococcus aureus lacking extracellular nuclease activity. J. Bacteriol. 101: Omori, G., and Y. Kato A staphylococcal food-poi- soning caused by a coagulase-negative strain. Biken J. 2: Richmond, M. H Plasmids and extrachromosomal genetics in Staphylococcus aureus, p In Jay 0. Cohen (ed.), The staphylococci. John Wiley and Sons, Inc., New York. 34. Sandvik, Studies on casein-precipitating enzymes of aerobic and facultatively anaerobic bacteria, p In Publication from the Department of Food Hygiene and Microbiology. Veterinary College of Norway, Oslo. 35. Schildkraut, C. L., J. Marmur, and P. Doty Determination of the base composition of deoxyribonucleic acid from its buoyant density in CsCl. J. Mol. Biol. 4: Silvestri, L. G., and L. R. Hill Agreement between deoxyribonucleic acid base composition and taxometric classification of gram-positive cocci. J. Bacteriol. 90: Smith, H. B. H., and H. Farkas-Himsley The relationship of pathogenic coagulase-negative staphylococci to Staphylococcus aureus. Can J. Microbiol. 15: