et al. (1954), in their studies on the flora of the a significant difference in the pathogenicity of A. bovis for mice as compared to "anaerobic

Transcription

1 METABOLIC AND SEROLOGIC DIFFERENTIATION OF ACTINOMYCES BOVIS AND "ANAEROBIC DIPHTHEROIDS" SYLVIA KING' AND ESTHER MEYER Department of Bacteriology, College of Medicine, University of Illinois, Chicago, Illinois Received for publication February 11, 1957 It is frequently a difficult task to differentiate Actinomyces bovis from "anaerobic diphtheroids" of questionable pathogenicity. Diphtheroids, isolated from normal and pathological tissues, resemble A. bovis morphologically and in some cultural characteristics. These similarities and lack of suitable criteria for their definitive identification have been responsible for the errors in separating "anaerobic diphtheroids" from A. bovis and vice versa. Taxonomists have been divided in their opinions regarding the classification of the genus Actinomyces. Erikson (1940) and Thompson (1950) believed the genus Actinomyces could be separated into 2 species depending upon their source, A. bovis (bovine) and Actinomyces israeli (human), by means of morphological and- biochemical criteria. Rosebury et al. (1944) felt there was insufficient evidence for this separation. Slack (1955, personal communication), using an agglutination-adsorption technique, found that separation of the genus Actinomyces into antigenic groups could not be associated with their host source. Erikson and Porteous (1955) expressed some doubt regarding their earlier conclusions when they discovered some smooth strains previously identified as A. bovis of bovine origin were actually "anaerobic diphtheroids." There is a limited amount of literature available on the classification of the "anaerobic diphtheroids," and many investigators note the present classification scheme is inadequate. Differences in opinion have been expressed regarding the pathogenicity of these microorganisms. Fleisher (1952), Prevot and Tardieux (1953), Seeliger (1953), and Linzenmeier (1954) believed these diphtheroids may be responsible for certain pathological conditions in man; however, Kassel and Rottino (1955) commented that most of these organisms could be eliminated if more careful techniques were used in collecting specimens and 1 Present address: Hektoen Institute for Medical Research, Chicago, Illinois. preparing cultures. "Anaerobic diptheroids" were isolated by Evans et al. (1950) and Pachtman et al. (1954), in their studies on the flora of the normal skin and in cases of seborrheic dermatitis. A few reports have appeared in the literature comparing A. bovis with organisms of similar morphology. Meyer and Verges (1950) reported a significant difference in the pathogenicity of A. bovis for mice as compared to "anaerobic diphtheroids." Frank and Skinner (1954) compared the morphology of these diphtheroids with A. bovis and Lactobacillus bifidus and concluded that the diphtheroids were more closely related to A. bovis than to L. bifidus. More recently, Suter (1956), in his studies of A. boats and "anaerobic diphtheroids," expressed the opinion that the catalase reaction is a good screening test for the separation of these troublesome organisms. MATERIALS AND METHODS Organisms. The organisms used were obtained from the University of Illinois Research and Education Hospital, Cook County Hospital, and Michael Reese Hospital, all located in Chicago, Illinois, as well as the United States Public Health Service in Chamblee, Georgia, and the American Type Culture Collection. Cultures of A. bovis isolated from human and bovine sources, and "anaerobic diphtheroids" from human sources were grown in Brewer's liquid thioglycolate-methylene blue medium and stored at -20 C. To maintain viability and uniformity, they were thawed as needed and subcultured in the same medium. Since previous investigators had compared A. boats to other similar anaerobic microorganisms, 1 isolate each of L. bitfdus and Propionibacterium freudenreichii and 5 isolates of Corynebacterium acnes were included in this study. Standardization of inocula. Growth from a 5 day old Brewer's thioglycolate broth culture was drawn up in a pipette, transferred to sterile normal saline, washed, centrifuged, and resus- 234

2 19571 DIFFERENTIATION OF A. BOVIS AND DIPHTHEROIDS 235 pended in saline. Two drops of inoculum from each suspension, standardized by nephelometry, were introduced into eac-h tube of culture media. Carbohydrate fermentations. A comparison of different basal media indicated a semisolid trypticase agar base (no. 151, Baltimore Biological Laboratory) was preferable. This medium was sterilized by autoclaving at 15 lb pressure, 121 C for 20 min. Carbohydrates in 20 per cent concentrations, sterilized separately in the same manner, were added aseptically to this medium, making the final concentration of sugar in each tube 1 per cent. To our knowledge, this medium has not been used previously in studies covering A. bovis or "anaerobic diphtheroids." Gelatin liquefaction. Nutrient gelatin having been found inadequate to support the growth of the fastidious A. bovis, an enriched medium, Thiogel (no. 293, Baltimore Biological Laboratory), was employed. Although most of the organisms which liquified gelatin did so within 15 days, we kept all of our cultures for 30 days before recording results as negative. Nitrate and indole tests. To determine nitrate reduction and the presence of indole, a buffered thioglycolate medium described by Douglas and Gunter (1946) was used to which nitrate or tryptone was added as necessary. It is advisable to test for nitrate reduction within 5 to 7 days. Occasionally, with the "anaerobic diphtheroids," nitrite had disappeared after this interval. The tryptone broths were examined for the presence of indole within 5 to 7 days. Litmus milk reactions. Reactions in litmus milk medium, inoculated and overlayed with 12 in of Vaspar (50 per cent Vaseline and 50 per cent paraffin), were slow; generally an incubation period of 30 days was required. Catalase test. The presence of the constitutive enzyme catalase was determined as follows: A loopful of culture from each organism grown anaerobically on a brain heart infusion agar slant was emulsified in a drop of saline on a glass slide after which a drop of hydrogen peroxide was added. Liberation of oxygen was shown by bubbling, indicating the presence of catalase. Organisms too hard to emulsify, were either broken up by shaking with glass beads or ground in a mortar. In order to verify the validity of this test for the presence of catalase, a more sensitive standard titration method described by Sumner and Somers (1947) was used for measuring catalase activity as follows: The organisms were weighed (wet wt), triturated in a mortar, suspended in a phosphate buffer of ph 6.8, and stored at -20 C until ready for use. Then the suspension was thawed and centrifuged, the supernatant decanted, and the cells resuspended in 1 ml of the same buffer. This was added to 25 ml of 0.01 N H202 held in an ice bath. At 2 min intervals, from 0 to 10 min, 5 ml of this solution were transferred into 5 ml of 2 N H2SO4 to stop the reaction. At the end of 10 min, the H202 in each flask was titrated with N KMnO4. If catalase were present, the quantity of H202 decreased with time as indicated by subsequent titrations with KMnO4. Immunological technique. The organisms used for immunization were grown on brain heart infusion agar, washed off with 0.3 per cent formalized normal saline, shaken with glass beads, and standardized to the density of a no. 3 Mc- Farland nephelometer tube. Rabbits 4 to 6 mo old, previously trial bled and tested, were given an intravenous and intracutaneous inoculation using the following protocol: The animals were given 1 ml intracutaneously every other day for 7 inoculations. On the injection days 0.1 ml was administered intravenously, and this dose was doubled every other day through 7 inoculations. The rabbits were bled 7 to 9 days after the final injection, allowed to rest for 2 weeks, and then given another series of injections in the same manner and bled again 7 to 9 days later. All sera were preserved with Merthiolate in a final concentration of 1:10,000 and stored at 0 to 5 C. Serological methods. Organisms used as antigens were grown on brain heart infusion agar and harvested in formalized 0.4 per cent saline. Then they were ground in a mortar or a Tembroecke grinder, standardized to the density of a no. 1 McFarland nephelometer tube, and stored at 0 to 5 C. For agglutination studies, serial dilutions of serum from 1:10 to 1:5,120 were prepared. To each 0.5 ml of diluted serum 0.5 ml of antigen was added. These were then incubated for 4 hr at 55 C, refrigerated overnight, and read the following morning. RESULTS As shown in table 1, all organisms studied produced acid without gas. The carbohydrates

3 236 KING AND MEYER [VOL. 74 TABLE 1 Carbohydrate fermentations* of organisms studied Organism No.of,, S. Isolates 4) tj 4) 4) 4) 04) ~~~ ~~ Actinomyces bovis (human) v Actinomyces bovis (bovine) Anaerobic diphtheroids v v + + v Corynebacterium acnes v Lactobacillus bifidus Propionibacterium freudenreichii * + = fermentation;-= no fermentation; v = variable. TABLE 2 Metabolic activity* of organisms studied No. of Gela- Ni- I-ICaa Organism Isolates tin trate dole Litmus Milk Catse Actinomyces bovis (human) v v Acid only Actinomyces bovis (bovine) v v Acid only Anaerobic diphtheroids v v v Acid with curd slight reduction + Corynebacterium acnes Acid with curd slight reduction + Lactobacillus bifidus Acid with curd slight reduction + Propionibacterium freudenreichii Acid with curd slight reduction + * + = activity;- = no activity; v = variable. TABLE 3 Agglutination tests of organisms studied Antigens (Organisms and No. of Isolates) Antisera (Titer Range 640-S120) A bovis 2 Anaerobic L. P. (un( ediphthe- C..esfrdes Actinomyces bovis (human) # * A. bovis (human) # A. bovis (human) # A. bovis (bovine) A Anaerobic diphtheroid M Anaerobic diphtheroid #D Anaerobic diphtheroid S Anaerobic diphtheroid SS Corynebacterium acnes ATCC # C.acnes#A, Lactobacillus bifidus ATCC #4962.0O Propionibacterium freudenreichii ATCC # * Number of isolates agglutinated by antiserum. a fording a distinction between A. bovis and the Table 2 shows that the A. bovis strains, whether "anaerobic diphtheroids" were xylose, salicin, and of bovine or human origin, differed consistently raffinose. The first 2 sugars distinguish A. bovis from the "anaerobic diphtheroids," C. acnes, L. from all of the other organisms included in this bifidus, and P. freudenreichii in respect to their study. action on litmus milk and their catalase activity.

4 1957] DIFFERENTIATION OF A. BOVIS AND DIPHTHEROIDS 237 The catalase activity of 2 isolates of A. bovis and 3 of the "anaerobic diphtheroids" was compared by the standard titration method referred to earlier. In a 10 min period flasks containing H202 to which equal weights of organisms had been added showed a reduction in H202 averaging 50 per cent for the "anaerobic diphtheroids" and revealed no decomposition of H202 by A. bovis. The results in table 3 reveal several important distinctions between the organisms in question. A. bovis isolates of human and bovine origin did not cross-react. There was no cross-agglutination between A. bovis and the "anaerobic diphtheroids." Aside from cross-reactions between the "anaerobic diphtheroids" and C. acnes, and between one of the "anaerobic diphtheroids" and P. freudenreichii, all agglutination results required homologous antiserum. DISCUSSION Collecting authentic cultures of A. bovis, particularly those of animal origin, was difficult. Nevertheless, several recent isolates were obtained from human sources which were classical examples of this organism. The large group of "anaerobic diphtheroids" studied, however, were consistent both in their similarities to each other and in the manner in which they differed from the microorganisms classified as A. bovis. Using our new criteria for differentiation, it was found that a few of our stock cultures, previously classified as A. bovis, were actually diphtheroids. This might have been due originally to earlier difficulties in identifying these organisms. It is also possible that an "anaerobic diphtheroid" was a contaminant in the original culture and remained viable through many subcultures, while the A. bovis failed to survive (see also Erikson and Porteous, 1955). Controversial reports in the literature regarding the variability of carbohydrate fermentations may be due to failures to observe the sugars daily for initial acid production; reversion to alkaline reaction frequently occurred. The necessity for standardized procedures applies to other differential tests as well. It is generally believed that catalase is present in almost all aerobic organisms and some facultative anaerobes, but under usual conditions of culture it is not found in strict anaerobes. We obtained conclusive evidence to indicate that the enzyme is present in "anaerobic diphtheroids" but not in A. bovis cultures. During the early stages of our investigation, we were in communication with Suter who was likewise engaged in studying these organisms. His report (Suter, 1956) also stresses the value of the catalase reaction in separating these morphologically similar organisms. Since various investigators discussed the possibility of spontaneous agglutination with A. bovis cultures, we thought it desirable to verify our agglutination reactions with complement fixation tests. The A. bovis antisera prepared against human and bovine strains of A. bovis fixed complement in the presence of their homologous antigens. The ratio of complement fixing antibodies to agglutinating antibodies was approximately 1:5. SUMMARY These studies indicate that the "anaerobic diphtheroids" are more closely related to the genus Corynebacterium than to Actinomyces as the latter is interpreted in the 1943 classification of Waksman and Henrici. Although morphological similarities do exist between Actinomyces bovis and the "anaerobic diphtheroids," selected differential criteria such as the catalase test, litmus milk reactions, and the fermentation of xylose, salicin, and raffinose can be used to implement proper identification of Actinomyces bovis. REFERENCES DOUGLAS, H. C. AND GUNTER, SHIRLEY E The taxonomic position of Corynebacterium acnes. J. Bacteriol., 52, ERIKSON, D Pathogenic anaerobic organisms of the Actinomyces group. Med. Research Council, Great Britain, Special Report Series #240, ERIKSON, D. AND PORTEOUS, J. W Commensalism in pathogenic anaerobic Actinomyces cultures. J. Gen. Microbiol., 13, EVANS, C. A., SMITH, W. M., JOHNSTON, E. A., AND GIBLETT, E. R The bacterial flora of the normal human skin. J. Invest. Dermatol., 15, FLEISHER, M. S Significance of diphtheroid microorganisms in blood cultures from human beings. Am. J. Med. Sci., 224, FRANK, H. A. AND SKINNER, C. E The relationship between Actinomyces bovis and Lactobacillus bifidus. Mycologia, 46,

5 238 KING AND MEYER [VOL. 74 KASSEL, R. AND ROTTINO, A The significance of diphtheroids in malignant disease studied by germ-free technics. A. M. A. Arch. Internal Med., 98, LINZENMEIER, G Thtude serologique des Corynebacteries anaerobies par la methode des agglutinines. Ann. inst. Pasteur, 87, MEYER, E. AND VERGES, P Mouse pathogenicity as a diagnostic aid in the identification of Actinomyces bovis. J. Lab. Clin. Med., 36, PACHTMAN, E. A., VICHER, E. E., AND BRUNNER, M. J The bacteriologic flora in seborrheic dermatitis. J. Invest. Dermatol., 22, PREVOT, A. R. AND TARDIEUX, P Pathogbne des Corynebacteries anaerobies. Ann. inst. Pasteur, 84, ROSEBURY, E., Epps, L. J., AND. CLARK, A. R A study of the isolation, cultivation, and pathogenicity of Actinomyces israeli recovered from the human mouth and from actinomycosis in man. J. Infectious Disseases, 74, SEELIGER, H Ein Beitrag zur Bacteriologie anerober Corynebakterien und deren Vorkommen in pathologischem Material. Arch. Hyg. u. Bakteriol., 137, SUMNER, J. B. AND SOMERS, G. F Enzymes, pp Academic Press, New York, N. Y. SUTER, LOYAL S Evaluation of criteria used in the identification of Actinomyces bovis with particular reference to the catalase reaction. Mycopathol. et Mycol. Appl., 7, THOMPsON, L Isolation and comparison of Actinomyces from human and bovine infections. Proc. Staff Meetings Mayo Clinic, 25, WAKSMAN, S. A. AND HENRICI, A. T The nomenclature and classification of the Actinomycetes. J. Bacteriol., 46, Downloaded from on October 2, 2018 by guest