suis. The multiple amino acid media devised by these workers (KBD and MMHRB) contained cystine and methionine as organic sources of sulfur.

THE CULTIVATION OF BRUCELLAE ON CHEMICALLY DEFINED MEDIA L. J. RODE, GLENDA OGLESBY, AND V. T. SCHUHARDT The Brucellosis Research Laboratory of the Clayton Foundation and the Department of Bacteriology, The University of Texas, Austin, Texas Received for publication August 7, 950 The selection of a chemically defined medium for nutritional or physiological studies depends upon the objectives of the investigator. The four chemically defined media for the cultivation of brucellae that have been reported during the past decade have tended toward one or the other of two objectives. Either multiple amino acid nitrogen source media have been devised and used to establish the essential or optimal growth requirements of selected strains of these organisms, or the objective has been to establish the minimal nutritional requirements of selected strains. Koser, Breslove, and Dorfman (9) reported the vitamins that are generally accepted as essential or stimulatory in Brucella nutrition. McCullough, Mills, Herbst, Roessler, and Brewer (97) reported a comprehensive study of the optimal growth conditions for a strain of Brucella suis. The multiple amino acid media devised by these workers (KBD and MMHRB) contained cystine and methionine as organic sources of sulfur. Each medium contained additional inorganic sulfur, since the required magnesium was added in the form of magnesium sulfate and utilized glucose as an added carbon source. McCullough and Dick (9), in this laboratory, developed a mineral base medium (MD) containing ammonium sulfate as the sole source of nitrogen, sodium thiosulfate as the primary source of sulfur, and glucose as the carbon source. Gerhardt and Wilson (98) described a medium (GW) in which either DLasparagine, L-glutamic acid, or L-histidine could serve as the nitrogen source. These workers chose glycerol and lactic acid as their major carbon sources and utilized sodium thiosulfate as the major source of sulfur. The composition of each of these four media is tabulated along with that of our formulation (ROS) in table. McCullough and Dick (9a,b) used the KBD medium in studying the growth response of a number of strains of brucellae and reported failure in their efforts to grow C0-requiring strains of Brucella abortus on this medium. McCullough et al. (97) and Gerhardt and Wilson (98) reported failure in their efforts to grow selected strains of brucellae on the MD single nitrogen source medium. We have found both the MD and the GW single nitrogen source media inadequate for the cultivation of many strains of brucellae from small inocula. In our cystine toxicity studies (Schuhardt, Rode, and Oglesby, 99) we found both the KBD and the MMHRB media toxic to certain strains of B. abortus and were By the term "C0-requiring" we mean those strains that require the addition of CO, to the growth environment. 66 Downloaded from http://jb.asm.org/ on April, 08 by guest

66 L. J. RODE, G. OGLESBY, AND V. T. SCHUHARDT [VOL. 60 TABLE Composition of chemically defined media formulated for the growth of brucellae COMPOSITION MD GW KBD jmherb ROS Per cent Sodium chloride... 0.75 0.75 0.80 0.80 0.75 Dipotassium phosphate... 0.0.00 0.0 0.0 0.0 Magnesium sulfate.....0 0.0 0.0 0.0 0.0 Sodium thiosulfate... 0.0 0.0 0.0 Glucose... 0.0-0.0.00 0.0 Glycerol....00 0.0 Lactic acid... 0.50 Micrograms per ml Fe.-. 0.0.00 0.0 Mn-... _ 0.0 0.0 0.0 Thiamine... 0.0 0.0 0.0 0.0 0.0 Nicotinic acid... 0.0 0.0 0.0 0.80 0.0 Pantothenate... I 0.0 0.0 0.0 0.0 0.0 Biotin... 0.00 0.00 0. 50t 0.00 0.00 Ammonium sulfate... 500 DL-Asparagine.-.- 000$ - Glycine... 00 7 00 DL-Alpha-alanine... - 500-00 DL-Valine... 00 00 DL-Leucine.-.-.00-50 DL-Isoleucine... 59 50 DL-Aspartic acid id.- - 665 00 L-Glutamic acidi.- 500 (DL) 000 DL-Serine.. - 5 6 5 DL-Threonine.-.. 5 7 5 L-Proline.-.- 00-00 L-Hydroxyprolmne... 00 L-Cystine... 50 9 5 DL-Methionine...... 00 0 50 DL-Phenylalanine... 00 99 5 i..tyrosine... 50 9 5 L-Tryptophan... 00 0 5 L-Arginine...... 00 8 00 L-Lysine... 00 6 (DL) 00 L-Histidine... 00 6 00 MD-McCullough and Dick (9); GW-Gerhardt and Wilson (98); KBD-Koser, Breslove, and Dorfman (9); MMHfRB-McCullough, Mills, Herbst, Roessler, and Brewer (97); ROS-Rode, Oglesby, and Schuhardt. * Reported that nucleic acid or any of its component purines and pyrimidines stimulated early growth. t A crude biotin concentrate. Koser and Wright later (9) reported 0.000 micrograms per ml crystalline biotin gave 90 per cent maximum growth. t Glutamic acid or histidine could substitute singly for asparagine. able to correlate some instances of this toxicity with the high cystine contents of these two media. Downloaded from http://jb.asm.org/ on April, 08 by guest

950] CULTIVATION OF BRUCELLAE 66 These reports indicate that no synthetic medium thus far described meets the nutritional requirements of all strains of brucellae tested. We have attempted to formulate a chemically defined medium that would promote growth from small inocula of all available strains of brucellae for use in determining the possible taxonomic significance of the relative toxicity for B. abortus and B. suis of cystine, and as a preliminary step in studies on the sulfur requirements of the brucellae. EXPERIMENTAL RESULTS In our preliminary observation of a toxicity differential between B. abortus and B. suis in media containing added cystine, we used 0.5 to.0 per cent tryptose TABLE The effect of the basal medium on the inhibitory level of cystine for Brucella abortus 57 BASAL MEDIA INHIBITORY LEVELS OF CYSTINE* micrograms per ml Chemically defined media MvD... GW... KBD... 6 NMHRB... 56 Digest media ( per cent) "Casamino acids" (Difco)... Tryptose lot A (Difco)... Tryptose lot B (Difco)... Tryptose lot C (Difco)... 6 Tryptose lot D (Difco)... 8 Tryptose lot E (Difco)... 8 "N-Z-Case"... 8 Albimi Brucella brotht... 5 * Tested at twofold concentration differences in tubes inoculated with 0. ml of a 0- dilution of a 8-hour tryptose broth culture. f Contains glucose and sodium bisulfite, both effective in neutralizing toxicity of cystine for Brucella abortus. as the basal medium. However, when different lots of tryptose and other peptones were used, the neutralization, or supplementation, of the cystine-induced toxicity varied considerably for the same strain of B. abortus (table ). Consequently we were convinced that a reproducible, synthetic medium might be necessary to test the taxonomic significance of the cystine-induced toxicity for these organisms. Formulation of our synthetic medium. In the experimental formulation of our synthetic medium (ROS) we chose test cultures of Brucella that, in our experience, grew poorly or not at all from small inocula in the previously described chemically defined media. Stock cultures were carried on tryptose agar, and seed cultures generally were derived from 8-hour tryptose broth cultures. Inocula consisted of 0.-ml amounts of diluted broth cultures. All glassware was chemically cleaned and rinsed in distilled water. Media were tested at ph 6.8 to 7.0 Downloaded from http://jb.asm.org/ on April, 08 by guest

66 L. J. RODE, G. OGLESBY, AND V. T. SCHUHARDT [VOL. 60 in 5-ml amounts in 0-by-50-mm pyrex tubes. All sterilization was by autoclaving at C for 0 minutes. Comparative growth tests were incubated for 5 days at 7 C, with or without 0 per cent added C0, as required. Growth was observed visually and recorded as +, +, +, +, [, or -. The amount of turbidity developed by our B. abortus 57 culture in per cent tryptose broth was selected as the + growth standard for comparison. In preliminary studies with single nitrogen source media, or with casein hydrolyzate media treated with hydrogen peroxide to destroy the organic sulfur, we observed strain differences in the sulfur requirements of brucellae. Some strains did not utilize methionine well; certain others failed to grow with thiosulfate or cystine as the sole sulfur source. The sulfur requirements were more exacting on the single nitrogen source media than on the casein hydrolyzate media. Because of these observations and the fact that casein digest media have proved to be adequate nonsynthetic media for the cultivation of brucellae, we chose to incorporate the mino acids of the casein digest series in our synthetic media. The 8 amino acids used were incorporated in concentrations that have proved nontoxic to brucellae in our amino acid toxicity studies. This amino acid base was tested with various concentrations of glucose, glycerol, lactic acid, mineral supplements, and the growth factors previously described as essential or beneficial in Brucella nutrition. The current formulation of our med ROS) is recorded in table. Efforts to simplify it have not proved beneficial when growth of all available strains of brucellae is used as the index of adequacy of the test medium. For other objectives, modifications or selection of other chemically defined media will prove essential. Cutivation of C0-requirtng strains of BruceUa abortus. Since McCullough and Dick (9b) reported failure in their efforts to grow recently isolated C00- requiring strains of B. abortus on the KBD medium and since none of the other authors have reported successful growth of these organisms on synthetic media, we directed our first efforts at the cultivation of C0-requiring strains. Four such strains of B. abortus were carried through 0 serial transfers in our medium using as inocula 0. ml of 0- dilutions of the - to -day growth. The normal atmosphere controls remained negative throughout the experiment, whereas + to + growth developed in each transplant incubated in increased C0. Four additional COrrequiring strains of B. abortus were tested by a decimal dilution inoculation procedure. The 8-hour tryptose broth culture of each organism was diluted decimally through 0-9. Duplicate tubes of our synthetic medium were inoculated with 0. ml of the undiluted cultures and of each dilution of each culture and incubated with and without added C0,. Growth occurred in each of the 0 tubes of each culture incubated in added C00. The 5- day growth in the tenth tube of each culture was diluted decimally through 09, Since the submission of this manuscript Gerhardt, Tucker, and Wilson (J. Bact., 59, 777-78,950) have reported the cultivation of two CO-requiring strains of B. abortus on their medium utilizing glutamic acid as the nitrogen source. Downloaded from http://jb.asm.org/ on April, 08 by guest

950] CULTIVATION OF BRUCELLAE 665 and the previous inoculation and incubation procedure was repeated. This process was repeated to a third series of tubes of our synthetic medium, and in each case a minimum of + to + growth developed in all tubes incubated in added C0. No growth occurred in the controls incubated in normal atmosphere. Convinced that our medium would grow C0-requiring strains of B. abortus, we proceeded to test 0 additional recently isolated strains of this organism by the decimal dilution technique. In the case of 8 of these, the colonies isolated from the milk of infected cows on tryptose agar plates were fished, with a straight bacteriological needle, directly to tubes of our synthetic medium, and the 8- hour growth in these cultures was used for preparing decimal dilutions for the inoculation of additional tubes of our medium. Again minimal + to + growth developed in all tubes incubated in added C0. The growth of each of these 0 freshly isolated strains approximated that in similarly inoculated tubes of tryptose broth. Comparative growth potentialities of five chemically defined media for Brucella spp. Having convinced ourselves that we could grow freshly isolated, C0-requiring strains of B. abortus from small inocula on our synthetic medium, we undertook a comparative growth study of this and the other four chemically defined Brucella media. For this study we used the decimal dilution inoculation technique described above except that only a single-passage series of decimal dilution inocula from tryptose seed cultures was used for each medium and organism tested. This method is subject to possible criticism because of the amounts of the tryptose and metabolites (0. ml of undiluted to 0-9) of the seed culture inoculated into the tubes of test medium. However, since each test medium receives the same inocula, and since we repeatedly have noted failure to obtain growth in the single nitrogen source media (MD and GW) receiving the 09 or larger inocula, the growth-potentiating factors generally are diluted out before the 0-9 dilution is reached. Furthermore the decimal dilution inoculation technique has the added advantage of not introducing the selective adaptation tendencies involved in the serial passage technique of testing a medium for the growth of a particular microorganism. Table illustrates the types of growth responses observed in this study. With the largest inoculum (0. ml of undiluted 8-hour tryptose broth seed culture) growth is abundant in all media for each organism tested. This growth in any given medium might be expected to be influenced by the number of organisms inoculated, or by the amount of tryptose and metabolites in the inoculum, or by both. When the growth response plays out gradually, as it does in each instance in which the single nitrogen source media (MD and GW) are employed, we are inclined to interpret this as evidence of an inadequacy of the medium, compensated for in the first few tubes by the contents of the inocula. When the growth stops abruptly, as observed in the case of B. abortus lob in the KBD medium, we are inclined to attribute this situation to an unneutralized toxicity of the medium. In this particular case the toxicity was proved to be due to the cystine content of the KBD medium. The fact that the MMHRB medium, which contains more cystine than the KBD medium, was not toxic for this strain of B. Downloaded from http://jb.asm.org/ on April, 08 by guest

666 L. J. RODE, G. OGLESBY, AND V. T. SCHUHARDT [VOL. 60 abortus can be explained in terms of a greater cystine-neutralizing capacity on the part of the MMHRB medi (see table ). In the case of the relatively abrupt termination of growth illustrated by B. meliten8s 59 in the MMHRB medium TABLE Comparative growth of selected strains of Brucella on five chemically defined media 5-DAY GROWTH RZADINGS ORGANISMS Brucella 8uis 9 Brucella melitensis 59 Brucella abortus 0B MZDIA MD GW KBD MMH&lRB ROS MD GW KBD MDMHRB ROS MD GW KBD MMdHRB ROS Tubes * Tubes to 0 inoculated with 0. ml of undiluted to 0-' dilutions respectively of 8- hour tryptose broth cultures of the Brucella strains. The (+) signs are omitted from the readings; i.e., = +, etc. TABLE The effect of sodium thiosulfate in our amino acid medium on the inhibitory levels of cystine for B. abortus and B. suis TEST ORGAISM 0. ML 0 DIL. INOCULUM 5 6 7 8 INHIBITORY LEVELS OF CYSTINE (MICROGRAMS PER Ml) With NaStO* 9 Without NaiSt0s B. abortus 57... 0 50 B. abortus 79... 0 50 B. 8ui8 P... 0 500 B. sui8 7... 500 * One hundred micrograms per ml. 0 Downloaded from http://jb.asm.org/ on April, 08 by guest (table ), the toxicity has been correlated with the high histidine content of this medium. Seven strains each of B. sui8 and B. melitenis and strains of B. abortu8 (6 requiring added CO) were tested by the decimal dilution inoculation technique for growth on each of the five chemically defined media. The results of this experiment confirmed our prior observation that the multiple amino acid media

950] CULTIVATION OF BRUCELLAE 667 generally are superior to the single nitrogen source media for the cultivation of the Brucella species tested. No significant superiority was observed in the growth potential of the three multiple amino acid media for the strains of B. suis or for the C0-requiring strains of B. abortus tested. Our medium (ROS) was significantly superior to the MMHRB medium with of the 7 strains of B. melitensis and with of the 8 acclimated strains of B. abortus. Our medium proved superior to the KBD medium for the cultivation of of 7 strains of B. melitensis and of of the 8 acclimated strains of B. abortus tested. Our medium was not significantly inferior to the other chemically defined media in the cultivation of any of the 8 strains tested. The effect of sodium thiosulfate on cystine toxicity for B. abortus and B. suis. Having formulated a suitable synthetic medium for cultivating our available strains of brucellae, we proceeded to use it to test the possibility of taxonomic significance of cystine toxicity for differentiating B. abortus and B. suis strains. Much to our surprise the strains tested showed no significant differences in susceptibility to cystine toxicity. Since they showed marked differences in tryptose, we began a search for the cause of the apparent increased susceptibility of B. suis to cystine-induced toxicity in our medium. Table gives the inhibitory levels of cystine for two strains each of B. abortus and B. suis in our medium with and without sodium thiosulfate. The deletion of sodium thiosulfate from our synthetic medium increases the tolerance of both species for cystine-induced toxicity, but the increase in tolerance is markedly greater for B. suis than for B. abortus. We are currently studying the significance of this observation, as well as the taxonomic and other significance of cystine toxicity for brucellae. SUMMARY AND CONCLUSIONS A multiple amino acid synthetic medium has been formulated which yields sustained growth from small inocula of all 6 strains of brucellae tested. These included 7 C0-requiring strains of Brucella abortus. None of the four previously described synthetic media for the cultivation of brucellae yielded comparable growth of all of these strains. Marked differences were observed in the nutritional requirements of different Brucella strains. Since these differences frequently are correlated with inoculum size, a decimal dilution inocula-testing technique is preferable to a single inoculum technique for Brucella nutritional studies. A significant toxicity relationship between sodium thiosulfate and cystine was noted in our synthetic medium. REFERENCES GERHARDT, P., AND WILSON, J. B. 98 The nutrition of brucellae: growth in simple chemically defined media. J. Bact., 56, 7-. KOSER, S. A., BRESLOVE, B. B., AND DORFMAN, A. 9 Accessory growth factor requirements of some representatives of the Brucella group. J. Infectious Diseases, 69, -. KOSER, S. A., AND WRIGHT, M. H. 9 Further experiments on accessory growth factor requirements of the Brucella group. J. Infectious Diseases, 7, 86-88. Downloaded from http://jb.asm.org/ on April, 08 by guest

668 L. J. RODE, G. OGLESBY, AND V. T. SCHUHARDT [VOL. 60 MCCULLOUGH, N. B., AND DICK, L. A. 9a Physiological studies of Brucella. I. Quantitative accessory growth factor requirement of certain strains of Brucella. J. Infectious Diseases, 7, 9-97. MCCULLOUGH, N. B., AND DICK, L. A. 9b Physiological studies of Brucella. II. Accessory growth factor requirement of recently isolated strains of Brucella abortus. J. Infectious Diseases, 7, 98-00. MCCULLOUGH, N. B., AND DICK, L. A. 9 Growth of Brucella in a simple chemically defined medium. Proc. Soc. Exptl. Biol. Med., 5, 0-. MCCULLOUGH, W. G., MILLS, R. C., HERBST, E. J., RoEssLrR, W. G., AND BREWER, C. R. 97 Studies on the nutritional requirements of Brucella 8Ui8. J. Bact., 5, 5-5. SCHUIJRDT, V. T., RODE, L. J., AND OGLESBY, G. 99 The toxicity of certain amino acids for brucellae. J. Bact., 58, 665-67. Downloaded from http://jb.asm.org/ on April, 08 by guest