NUTRITIONAL REQUIREMENTS FOR THE GROWTH AND ARTHROSPORE

NUTRITIONAL REQUIREMENTS FOR THE GROWTH AND ARTHROSPORE FORMATION OF COCCIDIOIDES IMMITIS E. P. GOLDSCHMIDT AND G. W. TAYLOR Fort Detrick, Frederick, Maryland Received for publication August 30, 1957 Coccidioides immitis, a dimorphic pathogenic fungus, when grown in vitro on solid or in liquid media, usually forms mycelium with septate hyphae which, upon continued incubation, fragment into numerous arthrospores approximately 2 to 4 A in diameter. No definitive studies on the production of arthrospores in chemically defined liquid media have appeared since the work of Roessler et al. (1946), which suggested that additional study of the optimal nutritional conditions might lead to improved growth, sporulation, and fragmentation of this organism. The experiments and results presented here have led to the development of an improved liquid synthetic medium for the production of arthrospores. Both total growth (cellular protein) and viable count have been increased approximately 10-fold above the levels previously reported (Roessler et al., 1946). The requirements for each component of the medium have been studied and the composition reported is believed to be optimal for the growth of C. immitis strain Cash. table 1, unless stated otherwise. The culture media were dispensed in 25-ml volumes in 500-ml Erlenmeyer flasks and incubated on a reciprocating shaker operating through a 3-in stroke at 96 excursions per min in a 34 C incubator adjusted to approximately 50 per cent relative humidity. Analytical methods. All cultures were analyzed after incubation for 7 days for cellular protein (total growth), viable count, and final ph. Cellular protein was determined on aliquots of cells washed with trichloracetic acid by the biuret method of Stickland (1951) using Pro-Sol (Standard Scientific Supply Corp.) as the protein standard. Cultures grown in the medium shown in table 1 were found to have less than 0.2 mg of extracellular protein per ml when analyzed by the biuret reaction for heat coagulable protein or by the turbidimetric procedure of Stadtman et al. (1951). Viable counts were obtained by a standard pour-plate technique. Aliquots of the culture, diluted in a solution containing 0.1 per cent tryptose (Difco) and 0.5 per cent NaCl, were plated in a medium consisting of 2 per cent glucose, 1 per cent peptone (Difco), 0.1 per cent METHODS Culture conditions. The organism used in this yeast autolysate (Basamin Busch) and 2 per cent investigation was the Cash strain of C. immtitis. agar. The plates were incubated 2 days at 34 C The culture was maintained as refrigerated (4 C), before counting. liquid stocks which were renewed at monthly Growth of the mycelial phase of C. immitis intervals. The stock cultures served as inocula for cannot be adequately defined by measuring either seed cultures which were used within a 3-day cellular protein or viable count alone. When an period. The seed inocula were grown in shake inoculum of arthrospores is used these germinate flasks for 7 days exactly as described for experimental cultures except when noted otherwise, incubation and subsequently sporulate to form to form hyphae which increase in length during The composition of the seed medium was altered, long chains of arthrospores. Further incubation as indicated below, in several experiments in order results in fragmentation of these chains to short to produce nutritional deficiencies with respect chains or single arthrospores. Measurement of to a specific constituent and thereby reduce the increase in protein alone does not distinguish carry-over into the experimental growth medium. between the different morphological forms of The experimental cultures were inoculated with growth. A small increase in viable count may 1 ml of the seed culture, except for the experiment result from large increases in cellular protein with dealing with the effect of inoculum volume on little fragmentation or small increases in cellular growth. All growth experiments were performed protein with extensive fragmentation. The ratio with modifications of the medium presented in of the viable count per mg cellular protein may be 265

266 GOLDSCHMIDT AND TAYLOR [VOL. 75 TABLE 1 Composition of the glucose, ammonium lactate medium for the growth of Coccidioides immitis Component g Amount per 1000 ml mm Glucose*...100.0 555.0 Lactic acid... 20.0 222.0 NH40H, concentrated... (14.75 ml) 222.0 KH2PO4... 3. 0 22.0 K2HPO4... 1.0 5.75 K2SO4... 2.0 11.5 - MgSO4 7H20... 0.5 2.0 - ZnSO4 7H20... 0.04 0.139 FeSO4-7H20... 1 5 X 10-3 5.4 X 10-3 MnSO4-H20.:... 0.77 X 10-3 4.5 X 10-3 CaCl2-2H20... 0.73 X 10-3 5.0 X 10-3 * Glucose sterilized as 40 per cent solution by autoclaving and added aseptically to the mediuim. ph of medium after auitoclaving = 5.3 to 5.5. of some value in measuring the predominant morphological form in the culture. The fragmentation index is defined here as the ratio: viable count X 10-7 per mg cellular protein. High fragmentation indices indicate that the cultures contain primarily single arthrospores. Low fragmentation indices suggest that the major viable particles are hyphae or long chains of arthrospores. It is realized that factors other than the degree of fragmentation may affect this index, such as the percentage of viable cells to total cells and variations in the size of the arthrospores under different nutritional conditions. In general, however, correlation of fragmentation indices with microscopic appearances of the cultures has been good and this has proven to be a valuable means for studying factors affecting the degree of fragmentation of cultures. The requirements for zinc, iron, and other trace elements were determined in the basal medium (table 1) which had been treated previously with alumina according to the procedure of Donald et al. (1952). The glucose solution (40 per cent) was treated with alumina and sterilized by autoclaving. Metal contamination of the growth flasks was reduced before addition of the medium by filling the flasks with a 0.5 per cent solution of disodium ethylenediamine tetraacetate and autoclaving for 10 to 15 min at 120 C. The flasks were rinsed with distilled water and then with deionized water (distilled water treated with a mixedbed resin consisting of Amberlites IRA-400 and IR-120) prior to the addition of the medium. All ph values of the media before and after growth were determined with a Beckman glass electrode ph meter. RESULTS Preliminary experiments. Growth experiments employing a basal medium similar to that of Roessler et al. (1946), failed to reveal any compound superior to glucose as a carbon source. Cellobiose and mannitol supported arthrospore formation equally as well as glucose. Fructose yielded only 50 per cent of the viable count obtained with glucose. Glycerol was a poor carbon source; and galactose, sodium gluconate, sorbitol, arabinose, xylose, lactose, and maltose failed tg support growth. A preliminary examination of various nitrogen sources revealed that proline would replace the ammonium acetate in Roessler's medium. Glycine, alanine, asparagine, and sodium glutamate, although supporting growth, reduced the viable count significantly. Ammonium acetate could be replaced by ammonium lactate or ammonium succinate. Ammonium sulfate gave poor growth and resulted in a marked decrease in the ph (3.0) of the culture; probably due to the accumulation of the sulfate anion. Ammonium salts of citric and tartaric acids also failed to support good growth. Urea, sterilized by filtration, supported good growth and arthrospore formation when added to the medium which had an initial ph of 4.5. Growth was negligible with potassium nitrate as a source of nitrogen. A complex nitrogen source, deionized casein partial hydrolyzate (Higuchi and Carlin, 1957), used at a concentration of 200,umoles total nitrogen per ml in a basal medium similar to that reported in table 1, supported growth and arthrospore formation comparable to those observed with ammonium lactate. WVhen the casein hydrolyzate was increased to levels of 300,umoles total nitrogen per ml, the growth increased to 16 mg cellular protein per ml with no increase in viable count. Preliminary experiments suggested that the nutrients were limiting in a modified Roessler's medium when increased aeration was employed. Therefore the nutrient concentration was increased to give a medium composition similar to that in table 1 and the effect of aeration was determined again. The results indicated that the

1958] NUTRITION FOR ARTHROSPORE FORMATION OF C. IMMITIS 267 viable count and cellular protein increased with increasing aeration rates up to a rate of approximately 1 mm 02 per L per min (highest level tested). The data suggested that aeration rates greater than 1 mm 02 per L per min may result in even greater rates of growth and fragmentation than reported here. Effect of medium constituents. The preliminary data suggested that a synthetic medium consisting of glucose, ammonium lactate, and salts would be suitable, with adequate aeration, for improved growth and arthrospore formation of C. immitis. The constituents described below were increased initially to concentration levels which were believed to be nonlimiting. Each constituent then was varied independently and the glucose and ammonium lactate concentration determined in factorial experiments. The constituents were added finally at optimum concentrations or in excess when preliminary experiments indicated that higher concentrations were not inhibitory. The results presented below were determined after the concentrations of the constituents were readjusted to give optimum growth. Glucose. The results in figure 1 indicate that glucose concentrations below approximately 450,umoles per ml were not adequate for maximum growth. The final ph was above 8.0 with low glucose concentrations whereas, with concentrations of glucose above 450 Amoles per ml, the final ph was approximately 6.0. These differences in terminal ph may be due to the greater utilization of lactic acid than of ammonia in the presence of low concentrations of glucose. Ammonium lactate. The results in figure 2 show that the concentration of ammonium lactate was critical for maximum viable count. Concentrations above approximately 222,umoles per ml resulted in a marked decrease in viable count. However, higher concentrations of ammonium lactate were necessary for maximum growth (cellular protein). These results indicate that the fragmentation of hyphae decreased when the ammonium lactate concentration was higher than 222 j,moles per ml. A maximum fragmentation index of approximately 35 was obtained with concentrations of ammonium lactate which were insufficient for maximum growth (140 to 222,Amoles per ml). Cultures deficient in ammonium lactate had final ph values of less than 6.0. This may be due to a greater utilization of ammonia 5 100 300 500 700 AM GLUCOSE PER ML Figure 1. Effect of glucose concentration growth of Coccidioides immitis strain Cash. on AM AMWONtIU ULTATE PER ML Figure 2. Effect of ammonium lactate concentration on growth of Coccidioides immitis strain Cash. than of lactate under these conditions. Substitution of ammonium acetate for ammonium lactate or the addition of ammonium acetate (2 g per L) to the medium completely inhibited growth. Effect of inorganic ions. (1) Phosphate:-Preliminary experiments suggested that a 3 to 1 ratio of monobasic to dibasic potassium phosphate was superior to 1:1, 1:3, 2:0, or 0:2 ratios of these salts. The effect of phosphate concentration was determined with concentrations varying from 0 to 55,uatoms phosphorus per ml. Maximum viable counts were obtained in the concentration range of 15 to 35,uatoms phosphorus per ml. z i-

268 [VOL. 75 GOLDSCHMIDT AND TAYLOR (2) Sulfate:-Seed inocula were grown in a medium containing only 0.5 g of potassium sulfate per L. The growth medium contained magnesium and zinc as the chloride salts at 3.0 and 0.139,umoles per ml respectively, and varying concentrations of potassium sulfate. Growth and viable count increased linearly with increasing levels of potassium sulfate in the range of 0 to 4,moles per ml. No differences in growth and viable counts were observed with concentrations from 4 to 50 Amoles potassium sulfate per ml. (3) Metallic cations:-the various inocula for testing the effect of metallic cations on the growth and viability of C. immitis were prepared as follows: the inocula for testing potassium and iron were grown in seed media containing no additional potassium or iron salts; the seed medium for testing magnesium and zinc contained 250 mg MgSO4 7H20 and 8 mg - ZnSO4.7H20 per L, respectively. All cations were added to the final growth medium in varying concentrations as sulfate salts except potassium, which was added as the chloride salt. The results presented in figures 3, 4, and 5 reveal that a higher concentration of potassium, magnesium, and zinc was required for maximum viable count than for maximum growth. The increased requirement for these cations for maximum viable count as compared with the requirements for growth suggest that these cations may play a role in the process of fragmentation. The NATOMS POhSSIUM PER ML Figure S. Effect of potassium concentration on growth of Coccidioides immitis strain Cash. The inoculum was grown in a medium containing no added potassium salts. The potassium salts in the growth medium were replaced by sodium salts at the same molarity. Potassium was added as the chloride salt. LAToMS M"NERUM MR ML Figure 4. The effect of magnesium concentration on the growth of Coccidioides immitis strain Cash. The inoculum was grown in a medium containing only 0.25 g magnesium sulfate heptahydrate per L. ZIO3AATOMS ZINC PER ML Figure 5. The effect of zinc concentration on the growth of Coccidioides immitis strain Cash. The inoculum was grown in a medium containing only 8 mg zinc sulfate heptahydrate per L. The growth medium was treated with alumina. concentration of iron required for maximum viable count was very critical. Optimum yields were obtained with approximately 5 to 10 x 10-3,Aatoms iron per ml. When the iron concentration was increased to approximately 55 x 103,uatoms per ml, the viable count decreased to about 35 per cent of the maximum count, thus suggesting that excess iron may inhibit fragmentation. On the other hand, no effect on growth was noted with ferrous sulfate concentrations ranging from 5 to 55 X 103,umoles per ml. (4) Other trace elements:-seed inocula lacking

1958] NUTRITION FOR ARTHROSPORE FORMATION OF C. IMMITIS added trace elements, and growth media which had been treated with alumina, were used to test the effect of various trace elements on growth and fragmentation. No significant effects were noted for manganese, calcium, copper, molybdenum, cobalt, or boron salts. It is probable that more effective methods for removing trace elements might reveal a requirement for some of these elements. Effect of ph. Media were prepared containing approximately 87 per cent of the standard ammonium hydroxide concentration. A sterile solution of 5 per cent KOH was added aseptically to sterile media to obtain the various initial ph values reported in figure 7. The results show that the initial ph is very critical with respect to fragmentation, but has little effect on growth within the range of ph 4.7 to 7.7. The optimum ph range for maximum viable counts was 5.3 to 5.5. Inoculum volume and age. No growth was obtained when an inoculum containing 2.5 X 107 viable particles was used. NVhen the inoculum contained 2.5 X 101 or 2.5 X 109 viable particles, optimum growth was obtained. Growth was initiated with less than 20 viable particles in a diluted medium when lower rates of aeration were used. Growth of small inocula may be inxmo3/ ATOS IRON PER ML Figure 6. The effect of iron concentration on growth. The inoculum was grown in a medium containing no added iron salt. The growth medium was treated with alumina and supplemented with varying concentrations of ferrous sulfate heptahydrate. INITIAL pm 269 Figure 7. The effect of initial ph on the growth of Coccidioides immitis strain Cash. The ammonium hydroxide concentration was reduced to 13 ml per L. A sterile 5 per cent KOH solution was added to adjust the media to the different initial ph values. hibited by the high concentration of medium constituents in the medium listed in table 1. No significant differences in growth were observed when seed inocula were grown 2, 4, and 7 days prior to inoculation of the growth medium. The 2-day-old inoculum had hyphae, the 4-dayold inoculum contained short chains of arthrospores, and the 7-day-old inoculum contained single arthrospores or short chains as the predominant morphological form. Growth of different strains of C. immitis. Several strains which showed large differences in viable count yields when grown in the medium of Roessler et al. (1946) were selected for comparison. These strains were grown first in diluted media with low rates of aeration. The final seed inocula were grown in the standard glucose, ammonium lactate, salts medium, as described in the Methods section. The results in table 2 show the growth of the different strains in the standard glucose ammonium lactate medium compared with the growth of the same strains in Roessler's medium as determined by Roessler, Geating, and Converse (1952, unpublished data). The results indicated that the Cash strain gave the best growth, particularly in the standard medium. Some of the strains revealed quite different order of rank when compared in the two media. Several of the strains grew better in

270 TABLE 2 A comparison of the growth of Coccidioides immitis strains in Roessler's medium and in the glucose, ammonium lactate medium. Strain Cash Norman Warren Dog Blood Soil 3A Perry Venezuela Euphrates Smith Mouser Brooker M-11 Silviera Final ph 6.65 6.85 6.7 6.6 6.7 6.1 5.8 6.2 6.2 6.2 6.4 8.0 6.1 GAL Medium* Cell protein mg/mtil 10.0 10.5 9.1 7.3 7.6 6.8 9.7 7.0 9.4 8.4 3.1 3.3 2.8 Viable count X 107/ 318 244 141 126 100 55 55 38 38 20 1 <0.2 <0.05 GOLDSCHMIDT AND TAYLOR Fragmentation index4 31.6 23.3 15.5 17.2 13.2 8.1 5.7 5.5 4.0 2.4 0.4 <0.06 <0.02 * Composition of GAL medium given in table 1. Viable count after 7 days incubation. t For composition of Roessler's medium (basal medium no. 2 + Zn, 2 ppm) see Roessler et al., 1946. Viable counts determined by Roessler, Geating, and Converse, 1952 (unpublished data). t Fragmentation index = viable count X 10-7 per mg cell protein. Roessler's medium than in the glucose, ammonium lactate, and salts medium. DISCUSSION The synthetic medium developed during this investigation for the growth of C. immitis strain Cash, has supported yields of 2.5 to 3.0 X 109 viable particles per ml and 8 to 9 mg of cellular protein per ml. These yields are approximately 10 times higher than those obtained with the synthetic medium reported by Roessler et al. (1946). The yield obtained with this improved medium corresponds to approximately 40 g dry weight of cellular material per L, which compares favorably with yields reported for other fungi (Foster, 1949). AMoreover, experiments revealed that the cellular protein, which is approximately 20 per cent of the cell dry weight, showed little variation for cells grown in media containing varying concentrations of glucose or ammonium lactate. [VOL. 75 Higher yields of total growth are possible in complex media containing a deionized, partial hydrolyzate of casein as a source of nitrogen. In this medium, cell yields corresponding to 16 mg cellular protein per ml have been obtained. The results of experiments showing the effect of aeration on growth and fragmentation suggest that aeration rates higher than those employed here might increase the y-ields and rate of growth above those now obtained. With the techniques used in this study one cannot distinguish between factors resulting in poor fragmentation per se and factors affecting adversely some stage prior to fragmentation such as sporulation, etc. Some of the factors affecting fragmentation were very critical such as ph, iron concentration and ammonium lactate concentration. High ph, or high concentrations of ammonium lactate, or iron resulted in reduced viable counts without inhibition of growth; in fact, higher concentrations of ammonium lactate increased total growth significantly. The requirements for potassium, zinc, and magnesium were found to be higher for fragmentation than for growth. Additional study is necessary to determinie if these nutrients affect the process of fragmentation. Preliminary findings have suggested- that fragmentation might be studied with nonproliferating cultures. Experiments with this type of system should aid in clarifying the role of various nutrients which appeared to inhibit or stimulate the fragmentation process. ACKNOWLEDGMENTS The authors wish to thank Dr. N. D. Gary and Dr. L. H. Graf for their support and encouragement during this investigation. SUMMARY The nutritional requirements for growth and arthrospore formation of Coccidioides immitis strain Cash have been studied. Yields of approximately 3 billion viable particles per ml and approximately 9 mg cellular protein per ml (approximately 40 g dry weight per L) were obtained after 7 days incubation in a synthetic medium consisting of glucose, ammonium lactate, and inorganic salts. The Cash strain gave higher yields than 12 other strains of C. immitis in this medium. The results suggested that high concentrations

1958] NUTRITION FOR ARTHROSPORE FORMATION OF C. IMMITIS 271 of ammonium lactate and iron and high ph were inhibitory for fragmentation but not for growth. There were higher requirements for potassium, zinc, and magnesium for maximum viable count (fragmentation) than for maximum growth. REFERENCES DONALD, C., PASSEY, B. I., AND SWABY, R. J. 1952 A comparison of methods for removing trace metals from microbiological media. J. Gen. Microbiol., 7, 211-220. FOSTER, J. W. 1949 Chemical activities of fungi, pp. 151-152. Academic Press, Inc., New York. HIGUCHI, K. AND CARLIN, C. E. 1957 Studies on the nutrition and physiology of Pasteurella pestis. I. A casein hydrolyzate medium for the growth of Pasteurella pestis. J. Bacteriol., 73, 122-129. ROESSLER, W. G., HERBST, E. J., MCCULLOUGH, W. G., MILLS, R. C., AND BREWER, C. R. 1946 Studies with Coccidioides immitis. I. Submerged growth in liquid mediums. J. Infectious Diseases, 79, 12-22. STADTMAN, E. R., NOVELLI, G. D., AND LIPMANN, F. 1951 Coenzyme A function in and acetyl transfer by the phosphotransacetylase system. J. Biol. Chem., 191, 365-376. STICKLAND, L. H. 1951 The determination of small quantities of bacteria by means of the biuret reaction. J. Gen. Microbiol., 5, 698-703. Downloaded from http://jb.asm.org/ on October 5, 2018 by guest