SY‘JIDY OF THE EFFECT GF' ?R@?Eii‘i CQMPONENTS EN 35.00%) SEE‘QUEfi 259994 THE GEGWTH OFF ERU CELLAA§3§1T§£S ”(basis For flu Dogma of M. S. MimiGAN STATE EfiNB‘a’EREIW Ram Virginia Baitzwr €955 THESIS STUDY OF THE EFFECT or PROTEIN COL'EPONENTS IN BLOOD SERUM UPON THE GROWTH OF BRUCELLA ABORTUS Betty Virginia ggltzer AN ABSTRACT Submitted to the School of Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1955 \, / , App r OVe a\v./' vf/ JC/JL/ /,.-[/¢ 7/. {(12 2? ‘ém - , g ABSTRACT Inhibition, both of colonial development on tryptose agar and of multiplication in tryptose broth, when seeded with cells of the fastidi- ous, cog-dependent Brucella abortus Type II (hfllson), hereafter referred to as strain 119, was shown to be corrected by the addition of whole bovine serum to either medium. This response was assayed by the filter paper disc method on agar, and by the resultant degree of turbidity in broth. The plasma protein fraction responsible for effecting growth was found to be albumin. Neither bovine plasma fraction II (gamma globulin), nor fraction III-l (beta globulin), effected any growth response. Attempts were made to improve the growth-enhancing action of albumin both by heat and chemical denaturation, and by digestion with three proteolytio enzymes -- pepsin, trypsin, and papain. However, each of these procedures depressed albumin activity. A milder heat treatment, performed while protecting albumin with saturated sugar solutions, did not alter the growth-stimulatory capacity of albumin. Beside trial of the pkmnm.protein fractions, incorporation in tryp- tose agar of two synthetic agents, Tween 80 and Span 80, showed that only the former, water-miscible agent enhanced growth, and the response was nearly equal to that obtained from the addition of bovine plasma albumin. Incorporation in tryptose agar of the sodium salts of two fatty acids gave the following results: lactic acid inhibited strain 119 at - ‘p m”? x 5r. (”rt awn-J ‘v-l H-u- n. '- - SCI??? '31 Clsc VIC: & ~58 are: 1.07 x 10-1 M concentration, but did not inhibit the typical, aerobic “gg. abortus strain 2308; oleic acid inhibited strain 119 at 3.28 x 10-615 concentration, and also inhibited strain 2308 at 6.6 x 10-8 M, a higher concentration. The technique of incorporating inhibiting agent (oleic acid) in tryptose agar and of adding enhancing agent (albumin or Tween 80) to the paper disc, revealed that these latter agents reversed the toxicity of fatty acid. That is, the diameter of the growth zone around the paper disc widened as weaker concentrations of inhibitory agent were mixed in the agar. Furthermore, growth-stimulation, better even than that obtained from enhancing agent alone, resulted from interacting the threshold inhibitory concentration of oleic acid (in agar) with either albumin or Tween 80 (on paper disc). The growth-enhancing action of albumin (as well as Tween 80) was attributed, not to a nutritive effect, but rather to a protective, detoxifying action, since the binding of ions by this protein, especially fatty acid anions, notably surpassed that of any other protein so examined (1). The protective action of albumin was established by treating tryp- tose broth with an adsorbing agent known to lack nutritive value, Norite-A. Whether Norite-A was removed before inoculation, or was present through- out the growth period, Norite-A-treated broth subsequently supported as ready multiplication of a minimum number of bacterial cells of strain 119 as did tryptose broth containing 0.005 percent albumin. Thus, prior to treatment with either binding agent, the tryptose constituent used to prepare broth and agar apparently contained some agent free to inhibit growth of strain 119. This inhibitor was likened to a fatty acid. Apparently some strains of Brucella are far more sensitive to fatty acids than others. Yet, if firmly bound or detoxified, the amount of fatty acid present in the medium was of no consequence. Thus, the concen— tration as well as the time of adding the binding agent determined whether the action of the inhibitor would be bactericidal, bacteriostatic, without effect, or perhaps even beneficial. RstRmics l. Klotz, I. M. (l9h9). The nature of some ion-protein complexes. Cold Spring Harbor Symposia On Quantitative Biology, Long Is- land Biological Association, Cold Spring Harbor, N. Y., 1h, 97" 1120 STUDY OF THE EFFET OF PROTEIN CCMPWDITS IN BLOOD SERUM UPON THE GROWTH OF BRUCELLA ABORTUS By Betty Virginia Baltzer A THESIS \‘w \ \_ Submitted to the School of Graduate Studies of MEchigan : State University of Agriculture and Applied Science in partial fulfillment of the requirementEi for the degree of “' \ MASTER OF SCIENCE Department of Microbiology and Public Health 1955 Jews INTRODUCTION . 0 TABLE OF CONTENTS 6“ Em PROC EDURm O O 0 O O O O O O O 0 O O O O O O O O O 0 O A. Materials B. Methods . mRmst AND mmJTs O O O O O O O O O O O O O O O O O O O The effect The effect The effect The effect of whole bovine serum on colonial growth . . of blood plasma proteins on colonial growth . of egg albumin on colonial growth . . . . . . of denaturation and digestion of albumin on oolonialgrowth..o............... The effect on colonial development of albumin heated in the presence of sugars . . . . . . . . . . . . . . The effect The effect The effect Tween The effect DISCUSSION . . . SUMMARY . . . . LITERATURE CITED ACKNOWLEDGEHENTS PLATES . . . . . of two wetting agents on colonial growth . . of two fatty acids on colonial growth . . . . of crystalline bevine plasma albumin and 80 as nutrients e e e e e e e e e e e e e o e of starch, and activated carbon in media . . Page \‘lxlxlwmm ll 11 18 25 27 bl h} as 55 57 62 63 STUDY OF THE EFFECT OF PROTEIN COMPONENTS IN BLOOD SERUM UPON THE GROWTH OF BRUCELLA ABORTUS INTRODUCTION - One of the most challenging problems that continues to confront study of all microbial life, has been that of promoting growth artificialbm extracellularly. The more closely allied the life cycle of the microorgan- ism has been to intracellular development, the more difficult has been the solution of this problem. Yet, again and again reports have come of the successful growth of fastidious bacteria following the incorporation of blood serum or plasma in culture mediums. Though not obligate intracellular parasites, members of the genus Brucella frequently resist primary isolation, and some strains, more fastidious than others, characteristically do not grow well on most culture mediums. The result of employing blood serum in a culture medium for the growth of one such fastidious Brucella abortus,-and the serum protein component responsible for growth is reported here, as well as the probable explanation of the role of action of this component. GENERAL PROCEDURES A. MATERIALS Cultures. An S-type strain of an atypical, cog-dependent 31;. abortus, 119, classified as Type II (Wilson)(l) and known to be fastidious, was used in all investigative studies. Two. other S-type strains were used for comparative purposes: a typical, cog-dependent _B_1;. abortus, 3038, and a non-COg-dependent BL. abortus, 2308. Agar medium. One basic agar medium, tryptose, was used throughout the investigation. It was prepared in the laboratory from the constituents recommended by the manufacturer. One lot of tryptose was used in all preparations. Since an alkalinity higher than that of the unadjusted medium was always desired, the final pH was controlled by a predetermined quantity of 5 percent Naacom added before autoclaving. Unless indicated, all mediums prepared for the cog-dependent strains were adjusted to obtain a final pH of 7.5; mediums for the aerobic strainwehe adjusted to pH 7.1 - 7.3. These initial pH values were determined as optimum in a previous study by Huddleson (2). ’ Tryptose agar fortified with killed brucella cells and variously used either experimentally or as a control type medium, was prepared by the addition before autoclaving of heat-killed cells of 21;. abortus 2308 09 to a final concentration of 2 x 1 cells per m1. of agar. Before using, all agar plates were dried I48 hours at 37'C. -9- Liquid mediumg, Tryptose and trypticase soy mediums were prepared in the laboratory as follows: Tryptose broth medium: BQOtO'tryptOSO ....................2.0 EMSe DUXtrOCO eeeeeeeeooeoeeeoeeeeeeeeeeOel nge SOdium Chloride ooeesseeOOOeeonOOOOOS gmse Distilled‘water eeeeeeeeeeeeeeeeeeeloo m189 Dissolve ingredients at room temperature by stirring. Adjust pH to 6.8 with .1 N HCl. , Trypticase soy broth medium: Tryptioaso eeeeeeeeeeeeeeeeeeeeeeeele7 EMSe Phyton. eeeeeeeeeeeeeeeeeeeeeeeeeee0e3 gnu. Sodium.ohloride eeeeeoeeeeeeeeeeeeeOe5 gmfle DOXtPOUC eeeeeeeeeeeeeeeeeeeeeeeeeeOeesgmae Distilled water ...................100 mls. Dissolve ingredients at room.temperature by stirring. pH not adjusted. Both mediums were dispensed in 20 m1. amounts into cotton plugged 50 ml. serum bottles, and autoclaved at 120.0. for twenty minutes. B. 'HRTHODS Inoculation of agar mediums. ‘ng abortus cultures were maintained on slants of Albimi brucella agar. Slant cultures of 119 and 3038 were incubated 36 hours under 5 percent 00,. Slant cultures of 2308 were incubated 2h hours aerobically. Cells were suspended in tryptose diluting fluid (0.5 percent NaCl and 0.05 percent tryptose) to a concentration of 2 x 109 total cells per;nl. as measured by the Libby Photronreflectometer. ‘flhen agar plates were inoculated with either of the Gog-dependent strains, a 1:375,000 dilution of the cell suspension was made. One ml. of this dilution, containing from 2,500 to 3,000 viable cells, was flooded over the surface of the agar and the excess liquid poured off within two -4- minutes. Inoculation with E5. abortus 2308 was done similarly except that the one m1. inoculum was made to contain approximately 2,000 viable cells. Since plate count reproducibility was not the objective, this procedure proved easy, and a successful one for the even distribution of approximately 1,200 isolated bacterial cells. The inoculated plates were air dried upon a level surface in the incubator for at least two hours following inoculation. The filter-paper disc method, first described by Vincent and Vincent (3), was used to assay the effect of an agent upon colonial development. A sterile disc was placed in the center of an inoculated plate, followed in five to ten seconds by the addition of 0.1 m1. of a liquid suspension of the agent. The plates were then air dried for an hour on a level surface in the incubator. This ensured uniform diffusion of the material into the agar medium. Plates inoculated with the cog-dependent BL. abortus strains were incubated, inverted, for 72 hours in a 5 percent C03 atmosphere and partial vacuum, in glass Jars or in dessicators, in which was also a beaker con- taining five grams of .Dehydrite crystals. The latter prevented excessive condensation of moisture within the plates. Plates inoculated with strain 2308 were incubated aerobically for 72 hours. Measurement of colonial development. At the end of this time a dissecting-type, binocular stage microscOpe (x 12) with a micrometer eye- piece was employed to measure the diameters of colonies present on each plate, and changes in the growth of colonies around the filter paper disc to which an agent had been added. On the plates without filter paper discs, the colonial diameter appearing most often was recorded. Also, a plate colony count was made. When colonial growth was enhanced by addition of agent to the filter paper disc, the diameters of the colonies were of maximum size nearestthe paper disc and became progressively smaller as the amount of the agent decreased by diffusion through the agar. The diameters of the two size extremes were recorded. Also measured was the radius from the disc over which the enhanced growth extended. If, however, the added agent caused inhibition, the radius from the disc of depressed colonial development was likewise recorded. In the latter case, the colonies bordering the zone of inhibition were smaller than those farther away. Inoculation and measurement of growth in liquid mediums. The essence of the procedure followed here has been described by Huddleson (b). Because evaluation of the growth promoting properties of broth mediums is more decisive when inoculated with a minimum number of bacterial cells, the final dilution of the 2 x 109 cell suspension (described previously) was such that twenty to twenty-five cells of strain 119, or ten to fifteen cells of strain 3038 (determined from agar plate colony count) were placed in each bottle of broth. Next, the agent (agents) to be tested for growth-promoting effect was added in a suitable concentration so that dilution of the medium was not appreciable. Last, the substance added to all bottles, was 0.5 ml. of a freshly prepared, aqueous, 8 percent NaHCOa solution, sterilized by filtering through a D-B Hormann pad in a Seitz filter. This served to adjust the initial pH, as well as to main- tain a favorable pH during incubation by controlling the quantity of C02 within the bottle. Within the ensuing two minutes after addition of the NaflCOa, the cotton plugs were replaced with rubber diaphragm stOppers, and the bottles then incubated for 72 hours at 37°C. Surveillance was kept for the initial appearance of turbidity indicating growth, and the increase in visible turbidity‘was recorded at h8 and at 72 hours. As a check, the pH level was determined by glass electrode immediately after unstoppering each bottle at completion of incubation. -I- EXPERDMENTS AND RESULTS 1. The effect of whole bovine serum on colonial growth. During the course of previous studies by Huddleson (2), from which stems the report that presence of brucella cells in mediums accelerate the reproduction rate of Brucella abortus, especially when smll inoculums are used, another agent, aged blood serum, was noted to have a similar influence on this species. Therefore, to make a more thorough study of this lat- ter agent, various saline dilutions of sterile whole bovine serum (Armour and Company) were assayed for their growth-promoting effect on the colo- nial develOpment of the atypical strain 119. This was done by the filter paper disc method already described (plate 1). The results (table 1) show that the concentration of whole serum influences the maximum size attained by colonies, and the distance that colonial development extends from the disc. The latter factor affects the colony count around the disc. 2. The effect of blood plasma Jroteins on colonial growth. In order to determine which of the protein constituents of bovine blood plasma was responsible for the growth-promoting effect of whole serum on the £5. abortus under study, the following experiment was designed to assay the effect of beta and gamma globulin, and crystalline bovine plasma albumin, as well as whole bovine senxm. The filter paper disc method was used. These fractions (Armour's) were easily reconstituted in physio- logical saline, to a concentration per ml. near that found in whole bovine serum. No pH adjustment was made on these solutions. Alpha globulin, TABLE 1 m WT or CONCENTRATIWS OF WHOLE BOVINE SEW ON COLONIAL GRNTH (Paper disc method) __— 6:? 3:312:11“ 33:13:11“. Colony Dianst er‘ 2:22: 32:? (percent) frown Dist! 8.0 15 1'3 " °'1 510 11.0 12 1.2 - 0.1 17° 0.1. s 1.0 - 0.1 18° 0.16 5 °"‘ ' 0’1 90 0.06 2 0.2 - 0.1 1° 0.008 ° 0 ‘In no -8- designated Fraction IV (Armour's), however, would not dissolve readily, even at a higher dilution, and was therefore omitted. Each resuspended fraction, as well as the whole serum was sterilized by filtration through a DbB Hormenn pad in a Seitz filter. The results (table 2) show that neither of the globulin fractions had any growth stimulatory properties, whereas the enhancement effected by albumin was as marked as that of whole serum (plate 2). Also, the two control plates strikingly contrast one another, and re-smphasize the need of an added agent in tryptose agar to permit growth of this brucella strain. Further proof of the distinctive,fastidious character of this strain was obtained in a parallel experiment, when the size attained by by the colonies of the typical, cog-dependent brucella strain 3038 showed no evidence of any growth-promoting effect by crystalline bovine plasma albumin, or by whole bovine serum. Furthermore, no zone of enhanced growth resulted. The similar ineffectiveness of added, killed brucella cells on the colonial deve10pment of typical strains has already been reported by Huddleson (2). Other data not included in table 2,'wherein the listed blood frac- tions were recombined in pairs and added to agar by paper disc, caused no enhancement, but demonstrated that a decrease in the concentration of albumin, reduced both the size of the colony and the zone of enhancement (plate 3). Because of this fact, the minimum.concentration of albumin that would Just support enhanced colonial deve10pment was then detenmined. To do this, the basic tryptose mediumnwas prepared as usual, autoclaved, and the pr0per dilution of a sterile albumin suspension added to the melted, sterile agar Just before pouring plates. The data obtained appear in table 3. Unfortunately, at this time the culture dissociated to a TABLE 2 mm OF MINE BLOOD PLASMA PRACTI 038 ON COLONIAL cam (Paper disc method) “L _: mu. .1 “‘2‘ Agent Una of rmm 3...... mm 601m a (”n”) from m_"a Diameter Count Crystallised bovine plasma albuinb 11.8 16 1.11 - 0.1 200 Can globulin. 1.6 0 Beta glebulind 1.6 0 Bovine serumb 8.0 15 1.3 - 0.1 178 Contrel. 0 Central": 1.1 - 0.5 1200 Av.’0.95 ‘In ‘e 5m and Cupany “Fraction II, Amour and quany arrestion III-l, Arlmur and Oapany 'Io agent added 9 tunes pg abortus «11- (2 x 10 /-1.) in agar type which on tryptose agar of pH 7.5 displayed a low incidence of micro- cclony development. Efforts to regain the previous phase type were of no avail. This, henceforth, necessitated use of a control plate of tryptose agar, pH 7.5, with each experiment in order to determine the significant minimum colony diameter on the test plates. At no time did the maximum diameter of the dissociated type colony exceed 0.6 mm. on a control plate. Unchanged, still, was the fact that no colonies developed on agar of pH 7.1 - 7.2. Colonies of the typical, cog-dependent 3038 strain show no such response. Reducing the pH from 7.5 to 7.1 decreases colony size no more than 0.h mm, and has no effect on plate colony count. Table 3 verifies the observation that colonial size decreased as albumin concentration was reduced. Even so, the lowest concentration tried, 0.0001 percent, enhanced the size of the colonies over those of the control. Since the colonial diameter of a strahi of typical brucella develops in 72 hours to 1.2 - 1.3 mm., then from table 3, the 0.01 to 0.02 percent concentration is shown to be the lowest concentration of albumin that will produce normal size colonies. As the albumin concentration exceeds the optimum 0.01 to 0.02 percent range, there is no corresponding increase in colony size, as one might expect. A further observation gained from table 3 concerns pH. As intimated in a previous paragraph, increasing the acidity of non-fortified tryptose agar diminishes the colonial diameter of the strain 119 2:, abortus until no colonies arise at all on media of pH 7.1 - 7.2; 'but addition of album- in to agar changes this picture. If an excess of albumin is present, little or no variation in colony size results, even though the pH be reduced from 7.5 to 7.1. A mmall decrease in colony size does occur, however, if the albumin concentration is less than the optimum 0.01 to TABLB3 ms or communion 0r cammszso Bonus PLASIA mom n 1013 011 «2an mm Col “'62::223“ v3 755?; 22:.” 0.0001 7.5 0.7 moderate. 0.0005 " 0.8 ' 0.001 ' 1.0 ' 0.005 " 1.1 ' " 7.2 0.8 " 0.01 7.5 1.2 highb 0.02 " 1.3 " 0.05 9 1.1 1 moderate " 7.1 1.1 high 0.1 7.5 1.1 low. " 7.1 1.0 " 0.21. 7.5 1.3 ' " " 7.1 1.2 ' contra“ 7.5 0.5 " ecdzrold'. " 1.0 high .700 to 1200 colonies per plate 111200 to 2000 " " " °300 to 700 '.' '.' " dllc added alumin OKilled .BLZ. abortus cells (2 2 109/111.) in agar -10- 0.02 percent. The weaker the concentration of albumin, the greater the decrease in colonial diameter when the medium pH is 7.1. Similarly, though not shown in the table, albumin concentration is again a factor when comparing the procedure of adding albumin to the agar medium before autoclaving, with that of adding after autoclaving. When the albumin concentration is in excess of the optimum, the order of pro- cedure causes no resultant variation in colony size, but when the albumin concentration is below the Optimum, size variation occurs. For example, a colonial diameter of 1.0 mm. upon 0.001 percent albumin agar, diminishes to 0.6 mm. when this same amount of albumin is present during autoclaving. Also of interest in table 3 is the unmistakable trend in colony count. This likewise bears a relationship to the concentration of albumin in the medium. Due to the inoculation procedure employed, the counts are not presented as absolute values, but because of the surprising degree of reproducibility that was apparent when the results were reworked because of the dissociation of the culture already mentioned, it was felt the singular trend deserved mention. To determine, if possible, whether the enhancement activity of whole bovine serum could be entirely accounted for in the albumin fraction, whole bovine serum was added to agar medium in the same manner as that described for albumin. The results recorded in table l1 compare favorably with those of table 3. The lowest concentration of bovine serum required for normal colonial development of 1.2 mm. lies within the broad 0.1 to 1.0 percent range. More precisely, 0.5 percent whole serum would be the median amount of choice. This is equivalent to 0.025 percent concentration of albumin. Essentially, then, this reaffirms the Optimum range set forth in table 3 and thereby strengthen the likelihood that the growth TABLE N THE EFFECT OF CONCENTRATION OF WHOLE BOVINE SERUM IN AGAR ON COLONIAL GROWTH =; Serum Added pH Colony Colony Diameter Count 0 ncentration Albumin ° Concentration (mm.) (percent) (percent) . 0.001 0.00005 7.5 0.3 ion“ 0.01 0.0005 " 0.0 " 0.1 0.005 " 1.0 " Des 0e025 1! 1e 2 highb 1.0 0.05 " 1.2 moderatec 2.0 0e]. " 1e3 10V! 5.0 0.25 7 1.3 " Controld * 0.5 " Controld’o " 1.0 high a B 300 to 700 colonies per plate b 8 1200 to 2000 colonies per plate 0 3 700 to 1200 d 8 no added bovine serum_ 9 c 8 killed Br. abortus cells (2 x 10 /ml) in agar -11- promoting factor in serum is albumin. Data analogous to thatin table 3, isthe effect of concentration upon colonial development, and also the trend in plate colony count. Though the effect of a more acid pH was not determined, a trend like that indicated in table 3 is predicted here. One can note that the two lowest serum concentrations used have no enhancement effect. 3. The effect of gggralbumin on colonia1_growth. Although it is well known that egg albumin is a protein distinct from serum.albumin, its trial, in the substance of fresh egg white, was included as a matter of interest. Accordingly, the white from a fresh egg was added either from saline sus- pension, or directly to the basic ingredients of the agar medium.bef0re heating to dissolve. Egg albumin also, depending upon concentration was found to promote enhanced colonial deve10pment, as table 5 illustrates. h. The effect of denaturation and_d£gggtion of albwmhn on colonial growth. Subsequent to the conclusion that the growth—promoting action of whole bovine serum lay in the albumin fraction, attempts were made to improve the stimulatory capacity of this fraction. Since the characteristics of a pro- tein are determined by the physical configuration or arrangement of polypep- tide chains within the molecule, it seemed probable that steric rearrange- ment, likened to 'unwinding,” or disruption of these chains, would alter the growth-promoting action of albumin -- possibly intensify it. Such molecular modification, or denaturation, is not typified by an "all or none” behavior (5). That is, degrees of structural change evolve depending upon the nature of the protein and the denaturhng agent used. Two methods of denaturation 1 (physical, and chemical) and three of digestion (biological) were studied. 1 The term, denaturation, has fallen into rather loose usage. Gener- ally speaking the term denotes ill-defined changes in protein properties. Therefore, Neurath, Siufllf (5) have defined denaturation as "any non- proteolytic modification of the unique structure of native proteins, giv- TABLE 5 EFFECT OF CONCENTRATION OF 300 IHITE IN AGAR ON COLONIAL GROITH Com entrat ion Colony Colony (percent) :23 Diameter Count (in an.) 0.001 7.5 0.1 (low 0.01 " 0.11 low.‘ 0.1 '_' 0.8 ' 1.0 . 1.2 nab 5.0 '3 1.2 ' 20.0 8.2. 1.3 '3 OOl‘l‘Gl‘Olo 7‘. 5 Gas 10' o ,d control ' 1.0 high .. ‘300 to 700 colonies per plate b1200 to 2000 :80 added egg shite Killed Br__._ _al:_c_____rtus cells (2 x 109/1111.) in agar -12- The initial attempt consisted of heating sterile, saline solutions of albumin (h.8 percent) at 60°C. and at 80°C. for 30 minutes in a water bath. A colloidal, "cloud-point” suspension developed at 60°C., but due to extending the degree of denaturation, the higher temperature caused the second solution to coagulate. Though centrifugation separated supernatant from precipitate, only soluble material lends itself to assay by the filter paper disc method, so that it was impossible to learn, on a comparable basis, whether activity, if any, may have been retained, or, possibly concentrated, in the precipitate portion of the coagulated protein. Be that as it may, assay of the solution heated at 60°C. showed activity reduced by approximately one-third (plate h), while none remained in the supernatant liquid from the 80°C.-treated specimen. The second means of denaturing (chemical) utilized the action of hydrogen and hydroxyl ions outside the stable range of pH. This hydrolytic procedure seemed worthy of trial primarily because the solubil- ity of bovine albumin could be maintained even though heating to a temper- ature higher than 80°C. Such is possible when performed in the presence of sufficient acid or alkali (pH 2 or pH 11). Although solubilizing is apparently the main effect occurring at pH 2, the soluble "sodium album- inate", after extended heating, is characterized by visible hydrolytic changes, such as an increase in viscosity and an amber coloration (6). For this reason, only a trace of alkali (pH 8) was employed. Thus, two ing rise to definite changes in chemical, physical or biological proper- ties." (Omission of any reaction whereby protons are removed from or added to the native protein, is understood.) In accord with this defini— tion, enzymic digestion is set apart from denaturation. -13. aqueous solutions of bovine plasma albumin (2 percent), one adjusted to pH 8, the other to pH 2, were autoclaved 30 minutes at 121.0. The pH of the alkaline solution remained constant during autoclaving and was not lowered before testing. The solution of acid pH was tested unadjusted, as well as after adjusting pH to neutrality. The results obtained from testing the three solutions showed activity reduced by nearly one-thi rd. For example, the albumin sample autoclaved at pH 2 and tested withoit pH adjustment showed maximum colony size of 1.2 mm., and a radius of enhance- ment zone equal to 10 mm., compared with the respective values for untreat— ed albumin (2 percent) ofa 1.5 m. colonial size, and a 15 mm. radius. Such reduced activity may be caused by one or several changes char- acteristically wrought by denaturing agents. One change of a physical nature that is likely to tnve occurred, is that of a lowered diffusibility. When heat is involved as a denaturation agent, a decrease in diffusibility is explained by an increase in molecular size, due to a polymerization that follows intramolecular rearrangement. Polymerization that has gone to completion is visible in the fom of coagulum. Yet, whereas a coagu- lated protein is always a denatured protein, it of course does not follow that one remaining soluble has not undergone denaturation. COOper and Neurath (7), by diffusion, viscosity and electrophoretio measurements, showed that polymerization occurred in crystalline horse serum albumin which they had heated at the pH that maintained solubility. There are also changes of a chemical nature within the molecule that should be considered as causes of reduced activity. In the many recent studies of Klotz and his associates concerning the chemical binding power of proteins, and especially of serum albumin, they postulate the chemical change that occurs within the albumin molecule when heat denatured, and which is manifested by a diminished binding power, to be due to a decrease in net number of free, cationic residue groups, particularly sNH+ residues, remaining on the albumin molecule (8). They believe these cati- ons responsible for the strong binding of anions to albumin. If such binding action by albumin favors growth enhancement of 2:. abortus 119, then, 22$ 22, any process that reduces the available cations on the albu— min molecule would also reduce the growth-promoting ability of albumin. The third method of treating albumin, digestim by biological agents, involved the use of three proteolytic enzymes: papain, trypsin, and pepsin. Examination of the albumin fraction by digestive means'was based on the principle that enzyme action -- fragmentation -- might release the physio- logically active segment or segments responsible for the growth-enhancing activity of the intact albumin. Procedural modifications that will not be elaborated upon were tried before concluding the appraisal of this treatment. The initial examination cf’both trypsin and papain, using a saline suspension of the crystalline bovine plasma albumin (u.8 percent), follow- ed the procedures recommended by the mmnufacturer. The initial digestion procedure using pepsin, was that adopted by Huddleson and Pennell (9) to purify Brucella antiserum. Assurance that digestion had proceeded tctxmmfle- tion was signified by failure of”the albumin-enzyme mixture to coagulate when heated at isoelectric conditions for 30 minutes at 80°C. Yet, no matter how the papaic digestion procedure'was modified, at no time did the reaction go to completion as determined by this criterion if begun with native protein. It was found that the effect of the resultant enzyme-albumin digests on the growth of the fastidious E5. abortus also was a measure of the -15- completeness of digestion. That is, when the 30-minute heating at 80°C. caused no protein to coagulate, neither was the digest growth stimulatory, either on solid or in liquid media, as the data in table 6 and 7 indicate. 0n the other hand, if any coagulation occurred before tenmination of the heating (incomplete digestion), then the extent of digestion was reflected in the extent of the growth stimulated by the agent. Though progressive loss of growth activity as digestion proceeded to completion should have been ample evidence that this method of treating albumin‘was ineffectual, yet known variability in end-products obtained by use of different enzymes provoked trial of some modified procedures, especially some involving the use of papain, as already implied. The results as set forth in table 6 are representative of data from these various modified procedures. Because pepsin repeatedly failed to elicit complete digestion, this enzyme received closest inspection. Beside trials with non-activated pepsin, both thioglycollic acid-activated and EON-activated papain preparations were tested. The report by Rice :3 =1, (10) that heat denaturation of protein increased the rate of papaic diges- tion of human serum.albumin, suggested trial of a similar procedure using bovine plasma albumin. Accordingly, digestion was initiated on a soluble albumin sample (2 percent), treated previously by autoclaving at pH 8. This routine actualized complete digestion as indicated by lack of coagu- lum at 80°C. (table 6, agent 5). To establish that the enzymes, alone, had no inhibitory effects on the growth of the BE. abortus strain, each one was tested by the paper disc method on agar fortified with killed brucella cells, in the concen- tration used for digestion. Finally, to avoi d any inhibitory effects from reagents other than enzymes, these, too, were tested by placing on TABLE 6 EFFECT OF ENZYME DIGESTE LLBUHIN ON COLONIAL mom (Paper disc method) Radius of Cone. Colony ent Growth ‘8 (percent) fro Diso‘ Diameter! 1. Albmin, whole 14.8 16 1J4 - 0.1 2. Alhamin, after 8 hr. trypsin treatmalt (so'c. heat-coagulated) L.s 13 1.2 - 0.1 3. Albumin, after 20 hr. trypsin 1: treatment (not heat-coagulated) the o 14. Album, after 20 hr. papain treatment (80°C. heat-coagulated) h.8 16 1.5 - 0.1 5. Albuin, autoclaved at pH 8, followed by 11 hr. papain treatment (not heat-coagulated) 2.0 --° 0.6 - 0.1 6. Albmin, ai‘ter’l 1/2 hr. pepsin treatment (slight coagulation .1; so'c.) 5.0 10 1.1 - 0.1 7. Albmin, after 2 hr. pepsin o treatment (not heat coagulated) 5.0 -- 0.5 - 0.1 8. Trypsin «- saline solution‘I 0.1 -—° 1.2 - 0.8 9. Pepsin - saline solutiond 0.1 -—° 1.2 - 0.8 is. Control° --° 0.6 - 0.1 “In In. “No detectable colonies anywhere on agar airface °Coloniea distributed evenly over entire agar surface lIo zone of inhibition or enhancement “cu agar fortified with killed cells ‘10 agent added TABLE 7 THE EFFECT OF ENZYME DIGESTED ALBUMIN ON GROWTH IN BROTH W Agent Added to Broth Cono. Turbidity . (percent) he hrs. 72 hrs. A Albumin, whole . O-Eh 2* 5* Albumin, after 20 hr. trypsin ‘ treatment (not heat—coagulated) 0.1 - - Albumin, after 20 hr. papain treatment (80°C. heat-coagulated) 0.1 1+ 5+ Controlb - - Control0 2. 5. - degree of growth by turbidity - tryptose - trypticase soy no turbidity ioa‘fl -16- paper discs. In view of the frequency with which the h.8 percent solution of native crystalline bovine albumin was used as a control when test- ing samples by the filter paper disc method, four observations deserve mention here. An albumin solution (b.8 percent) retains its growth-enhanc- ing potency after storage for as long as six months at h°c. And activi- uyis not influenced by a minor change in pH (that is, from the usual norm of 5.7 - 6.0 to a neutral 7.0 - 7.3), as ascertained by the disc method. Yet,increasing atmospheric 003 content to 10 percent reduces colonial diameters slightly, but has little or no effect on the spread of enhance- ment from the disc; 'while repassage through a Hormann DB filtration pad understandably decreases growth-stimulatory activity because the adsorb- ing action of the filter pad reduces albumin concentration. It seemed possible that by enzyme fragmentation of the albumin mole— cule, some nutritionally stimulating factor might be released that could later be isolated and identified. 0n the other hand, if potency of album- in were due to some other physiological influence, digestion probably would not be worthy of trial. However, there have been a few important instances where loss in biological activity of proteins has not followed digestion. An exemplary case of this has been the preparation of antitox— ins for therapeutic use —- especially diphtheria antitoxin. Such apossibility further encouraged trial of digestion. Choice of enzymes'was based both on the fact thst trypsin, pepsin, and pepsin are noted for their proteolytic activity, and on the fact that these three enzymes have been used successfully in the above-mentioned preparation of diphtheria antitoxin. Also adOpted from this latter diges- tion procedure, in which non-antitoxic protein is digested away from active -17- antitoxin in serum globulin from the immunized horse, was the criterion of complete digestion, i.e.,failure of protein to heat coagulate (ll). leasuring the extent of digestion by this arbitrary standard by no means discloses what split products have resulted from enzyme action. Because the present concept of the serum.a1bumin molecule, either human or bovine, is that of a single polypeptide chain containing one readily available sulfhydryl group per mole (12), and one N-teminal alpha-amino group (always an aspartic acid residue) (13), a mixture of split products is reasonably expected. (0n the other hand, several -SH groups per mole of egg albumin are indicated, but because these are not near the surface of the molecule they are not readily available.) Furthermore, the mixture proportions and extent of fragmentation are variable. Both the enzyme used and experimental conditions, even though only slightly modified, are known to affect final outcome. For this reason, it would seem that if any products of digestion were to have displayed growth stimulation, then the three enzymes used, and the several procedural methods of digestion tried should have provided adequate Opportunity to show this, unless the final concentration of the fragmented product was too dilute. Though interest in what end-products might be present was of second- ary moment, conjecture arose as to their possible identity'when indication of growth inactivity became more certain. Actually the amino acid compo- sition of bovine plasma albumin has been well characterized from numerous studies. For instance, the four most abundant amino acids present, in decreasing order by weight, are: glutamic acid, leucine, lysine, and aspartio acid (1h). Search of the literature revealed a rather surprising report by Riesen and Blvehjem (6). They report that there was only a negli- gible release of free amino acids from native,or heated, soluble bovine plasma albumin when digested either by pepsin, trypsin, trypsin-pepsin, or pepsin-trypsin. Instead of amino acids, enzymatic hydrolysis resulted mostly in release of peptides, the average size of'which depended upon the enzyme used. For example, pepsin-freed peptides had fewer amino acid resi- dues than these released by trypsin. Likewise, the enzyme determined the principal amino acids contained in these respective peptides. Further proof of the individuality of enzymatic action for a given substrate was the mention that pepsin was twenty times as active as trypsin (on a weight basis) in effecting digestion of native albumin. Analogous to this find- ing is the notation in table 6 of the time required fer digestion by pepsin compared to that needed by trypsin. Incidentally, the improved ease with which papain attacked the solu- ble, heat-denatured albumin is the result of better accessibility to pep- tide bonds, the characteristic structural point of attack of proteases. Such predenaturation of protein, that is, intramolecular 'unwindings whether by heat or other enzymes, improves the rate and ease of attack of most proteases, except pepsin (6) (10) (15). 5, The effect on colonial development of albumingheated in the presence of sugars. Since, from the above results, it is clearly apparent that degradation of the albumin molecule, either by denaturation or enzymatic digestion, does not secure the improvement sought in the growth—stimulatory capacity, there remained one alternative approach to the problem, that of heating albumin while, at the same time, by means of saturated sugar solutions, protecting agahnst denaturation. The promise held forth by this method lay in the fact that, whereas degradation was provoked by the previous procedures, now the albumin molecule would remain intact, or nearly so, as evidenced by electrophoretic patterns made before and after H a. -19- heat treatment (16). Besides this, possible formation of the colloidal 'C' component, characteristically noted in electrophoretic analyses of heated serums and plasmas, would be averted. The investigation was confined to trial of three sugars: glucose, lactose, and sucrose. The procedure followed was pattenned after that used by Hardt _e_t_:_1_._. (16, 17). Three saline solutions of albumin (hi8 percent) were each saturated at room temperature with a sugar. Aliquots from these stock sugar-albumin solutions were heated for one hour, at 65°C. and at 80°C., in a constant temperature water bath. The appearance of each solution was noted after heating. One sample from each stock sugar-albumin solution was not heated. These served as controls. A fourth control was an albumin solution (14.8 percent), containing no sugar, heated at 65°C. for one hour. In this lat- ter control a very fine, nearly colloidal ccagulum formed which was later removed by centrifugation. All heated samples and all controls were dia- lyzed 72 hours against physiological saline. The final volume and pH of each solution was recorded after dialysis. In the case of glucose, assur- ance of complete sugar removal was confirmed by Benedict's test for reduc- ing sugar. Finally, each solution was sterilized by filtration through a Hermann DB pad in a Seitz filter, and tested for growthrstimulatcry activity by the paper disc method. The characteristic appearance of each sample after heating apparently depended on the saturating sugar. All non-heated samples were water-clear, and there was no change from this initial appearance for either glucose or sucrose samples heated at 65°C. But at 80°C., both samples slowly developed an opalescence and a slight viscosity and gel. However, the lactose saturated sample heated at 65°C., developed an opalescence almost -20- at once, while the 80°C. temperature further proved lactose a poor heat- protecting sugar. That is, a dense, white coagulum precipitated immedi- ately and rendered further procedures on this sample useless. The pH of the samples varied from.5.7h.fcr the albumin control heated at 65°C. in the absence of sugar, to 7.0 for the lactose-albumin sample heated at 65°C. No adjustment of pH was made. No adjustment of protein concentration was made following dialysis. The‘weakest albumin concentrations occurred in the samples previously saturated with sucrose. These concentration values ranged from 0.72 per- cent to 0.78 percent. In the specimens from which glucose and lactose had been dialyzed, albumin concentrations varied from 1.78 percent to 1.92 percent. Yet when testing these various samples, it was necessary to con- sider only the greater dilution of albumin in the sucrose-treated samples, since it was known from previous trial by the paper disc method, that a content reduction from.h.8 percent to 2.0 percent albumin elicited no modification in total growth enhancement. The inclusive results obtained by this heating procedure showed that the process neither augmented nor diminished the growth-enhancement action of albumin on 25. abortus 119. (table 8). Failure of albumin to show an altered capacity in its growth stimula- tion of the fastidious brucella strain subsequent to heat treatment in the presence of saturated sugars, may be accounted for by any one, or, combina— tion, of the following three suppositions: 1) heat exposure evoked no appreciable intramolecular change in the albumin molecule, which is to say, that protection by sugars against denaturation was sufficiently com- Plete to avert any responsive structural change that in turn*would reflect filtered growth promotion of 8:. abortus 119; 2) the extent of regeneration TABLE 8 THE EFFECT ON COLONIAL GROWTH OF ALBUMIN HEATED IN SOLUTIONS SATURATED WITH SUGAR (Paper disc method) .1 Albumin Plus TTreatmznt Colonial Growth 3 I' U d cm are ure uga se 5°C.) No sugar 65° slight decrease in colony size and enhancement area Glucose rm. temp. a like response R 650 n I! n " 80° colony size same, enhancement area smaller Sucrose rm. temp. decrease in colony sige and enhancement area ” 65° slight decrease in colonyb size and enhancement area " 80° colony size same, b enhancement area smaller Lactose 65° a like response " 80° not tested a ' results expressed on comparative basis, i.e., with the colonial growth obtained using h.8% unheated albumin b a lowered albumin concentration (percent) -21- of albumin from the denatured state was adequate 'tc show an almost unchanged growth response; 3) by the suppression of '0' component forma- tion, there was no concomitant reduction in the amount of free albumin. As mentioned before, denaturation consists of two consecutive reac- tions: structural unfolding within the molecule, followed by aggregation or polymerization. Of these, unfolding, by far the more important action, is the essence of denaturation. But the term, thermal stability, here per- tains only to the visible manifestation of denaturation, and«especially to the initial stages noted as opalescence or cloud formation -- Tyndall effects. The degree to which sugars raise the thermal stability of pro- teins depends upon both the concentration of sugar and concentration of protein. A sugar at half-saturation permdts more extensive denaturation (16). Likewise, an increase in protein concentration Lessens the efficien- cy of the sugar action. Other factors, too, affect the protective action of sugars length of heating, temperature, pH, sugar utilized, and the protein in question (16, 18). Sugars are not equally efficient in their protection against denatur- ation. In the results, this was visually noted. Hardt 23_=&, (l6), prov- ed such a difference electrophoretically, while Ball gtngl. (l9), demon- strated differences in sugar efficiency by measuring the content of sulf- hydryl groups in egg albumin heated in the presence of sugars. Apparently glucose, fructose, and sucrose are the most effectual. The reason for this seems related to the solubility and osmotic property of the added sugar. For instance, using equal volumes of a single concentration of albumin, saturation with lactose required 3 gms., as compared with the need for 19 gms. and 21 gms. of glucose, and sucrose, respectively. Similarly, comparison of saturation concentrations of two sugars and a agar alcohol -22- used to protect egg albumin from heat denaturation (19), indicate a high degree of protection from glucose (b.6 M at saturation) and fructose (5.0 M at saturation), but poor protection from mannitol (0.8 M at saturation). Supposition #2 preSUpposes that denaturation has occurred to some extent. Denatured protein that regains partial or complete solubility in water or saline, is termed regenerated protein. This process of rever- sion, or restoration, proceeds more readily during cooling or dialysis, and the extent of recovery of many of the properties characterizing the native protein, particularly solubility, depends partly on the extent of denaturation. In general, the milder the denaturation conditions, the more likely will reversion occur (20). The treatment adapted here, of heating while protecting against denaturation, would be classed as the mildest of denaturation conditions, so that regeneration is not unlikely, both at 65°C., and also at 80°C. where opalescence was noticeable in both sucrose and glucose saturated samples. Furthermore, the study of revers- ibility by electrOphoresis has in many instances disclosed that regenerat- ed protein may exhibit a mobility identical with that of the native protein (21). Such was found true for bovine serum albumin denatured ‘with 8 M guanidine hydrochloride, a treatment more drastic than the heat denaturation used in this investigation. The 'C' component mentioned in supposition #5 was discovered by also- trophoretic study of heat denatured protein (22). Instead of a pattern that revealed greater heterogeneity than did native protein, as is the usual result of denaturation, the formation of 'C' component is an example Of increased electrOphoretic homogeneity accompanying heat denaturation. That is, component proteins within a heated protein mixture, may, at their own expense, form into a colloidal aggregation-complex. This is the 'C' component (22) of the electrophoretic pattern. The mixed proteins taking part in this aggregate formation may be unrelated in nature, such as the alpha casein and gamma globulin that Krecji (25) heated together, or they may be the component fractions of a single serum (22), or plasma (16). Actual confirmation of '0' component development in a heated, pure protein,such as the boVine albumin fraction used in this study;has not been established. The purest fraction of bovine plasma that Hardt 33.:1. (16) studied consisted of albumin and alpha globulin, and exhibited 'C' component formation when heated one hour at 65°C. 'lhen van Der Scheer and associates (22) heated a single protein, ‘the albumin fraction from horse serum, no such complex was elicited. But it should also be pointed out that two forces in the solutions, pH and ionic strength, control the nature and magnitude of change in the electrophoretic picture of the dena- tured protein molecule, with the result that both the proportion of compo- nents formed, and their mobilities may be varied (7). Nevertheless, wheth- or or not 'C' component specifically forms, suppression of heat denatura- tion by sugar serves to minimize formation of heat-denatured products. As a result, the concentration of the intact albumin molecule remains unchanged. The fundamental chemistry of the protective action by‘which sugars, neutral solutes, suppress heat denaturation of albumin is still not satis- factorily explained (21;). Ball 35.3.13 (19), discounted formatim of a sugar albumin complex. But recently a report by Mohammad £33}: (25) suggesteithat glucose adds reversibly to the amino groups of bovine serum albumin, since, if the protein is first acetylated so as to mask the amino groups, and then treated with glucose, no 'browning' reaction occurs. Ordinarily the 'browning' that appears in a heated glucose-protein solution is apparently dependent on free -NH3 groups. Similarly, in his extensive studies of albumin, Klotz (26) revealed that fatty acid anions add to albumin at the ENH+ sites. But, if first denatured, albumin does not bind fatty acids. Also, like sugars, when fatty acids are bound to albumin they exert a protective action against denaturation, in fact are more efficiently protective than sugars. For instance, sodium caprylate (0.5 M) prevented coagulation of albumin at 100°C., whereas glucose (0.3 M) could not raise the thermal stability of albumin higher than 65.1°C. (27). The binding of fatty acids by albumin has been recognized, too, from a biological standpoint by Davis and Dubos (28). They asserted that in the liquid media developed for growth of tubercle bacilli, the function of albumin is protective rather than nutritive. That is, because oleic acid contaminates the synthetic, commercially impure Tween 80 utilized in the media, a bacteriostatic effect on the growth of small inocula results unless albumin is included in the medium. Similarly, they too, found that the protection afforded a small inoculum by the presence of albumin was lost if the albumin was first denatured. Cohn (29) reported that one of the several functions of albumin in blood is its interaction with a variety of smaller molecules such as fatty acid anions. For these it acts as transporting agent, and also serves to reduce any hemolytic action toward erythrocytes (30). If than, this binding of fatty acid is one of the natural functions fulfilled by albumin, and relevant to the explanation of stimulated bacterial growth in one report already mentioned, why could not albumin be functioning similarly in each instance in this study in which growth- cnhancement occurred in the presence of albumin? -25- 6. The effect of two wetting agents on colonial growth. Before directly seeking the answer to this question, interest was aroused over the possi- bility of whether the above-mmntioned Tween 80, utilised by Dubos in the liquid mmdia supporting growth of tubercle bacilli, would similarly display a growth-promoting effect on the fastidious 131. abortus 119. This possi- bility came to mdnd since,if the added albumin was functioning protective- ly rather than nutritively in Dubos' medium, TWeen 80 might either be act- ing in a manner similar to albumin (i.e., protectively), or else replacing some natural growth factor in the oft-used egg-base media in which egg albumin is a major component. Accordingly, Tween BO (the polyoxyethylene derivative of sorbitan mono- oleatc), with a viscosity and color similar to that of honey, was diluted in distilled water,'with which it is readily miscible, and calculated quantities added to tryptose agar before autoclaving, as indicated in table 9. From this table it may be seen that colonial development improv- ed favorably in a manner analogous to that resulting from addition to agar_ of bovine albumin (table 3), or egg albumin (table 5), and, likewise, depending on agent concentration. Also available for an analogous trial was the wetting agent, Span 80 (sorbitol esterified with oleic acid), and structurally different from Tween 80. The Tween products are ester-others, and derived from the Span agents by the addition of polyoxyethylene chains at the sites of non-ester- ified hydroxyl groups. Substitution at these ~03 sites yield water-soluble products, whereas the Span—type materials tend to be water insoluble. Due to this characteristic water insolubility, those dilutions added to tryPtose media before autoclaving were prepared from a warmed emulsion of water-Span 80. Tiny globules of Span appeared on the hardened surfaces TABLE 9 a THE EFFECT OF CONCENTRATION OF TWEEN 80 IN AGAR ON COLONIAL GROWTH Concentration pH Colony Colony (percent) Diameter Count (in mm.) 0.001 7.h 0.5 moderateb 0.005 7.35 0.7 highc 0.05 7.1; 1.0 " 0.1 7.h 1.2 moderate contrcle 7.5 0.5 lewd controla'f 7.5 1.0 ' high a - polyoxyethylene derivative of sorbitan monooleate, Atlas Powder Co. - 700 to 1200 colonies per plate 1200 to 2000 " 300 to 700 " ” " no added Tween 80 killed Br. abortus cells (2 x 10 9/51. ) in agar ”0000‘ I -26- of those plates containing the higher concentrations. As may be seen in table 10, colonies of the fastidious brucella strain were inhibited by all concentrations of Span except the lowest, which permitted colonial development to be no greater than 0.2 mm. in diameter, a size even smaller than those colonies arising on the control plate. In sharp contrast, Span had no inhibitory effect on either number or size of the aerobic B5, abortus 2308 colonies. The inhibitory effect of Span just noted, recurs in the data recorded in table 11, in which the effect of the three agents, albumin, Tween 80, and Span 80, on the colonial growth of both the aerobic and the Cog-depend- ent brucella strains was comparatively evaluated by the paper disc method. By means of colonial development, the results point out the contrast in the wetting agents, Tween 80 and Span 80, as well as in the two 2:. abortus strains. Deve10pment of the aerobic strain 2308 responded similarly to i all three agents. In each case colonies were evenly distributed over the surface of the agar, and were quite constant in size, whether adjacent to the disc pad or at the edge of the plate. 0n the other hand, strain 119 was again inhibited by Span 80, but the growth enhancement action of both albumin and Tween 80 was reaffirmed; the action of albumin.was apparently more effective, but only by a margin that might readily yield to use of an increased concentration of Tween 80. These three agents were next placed on agar fortified with killed brucella cells, and known to support normal colonial deve10pment of the fastidious 119 strain (table 12). By this means it was thought possible to demonstrate that the fastidious strain would develop colonies in the presence of these agents just as had the aerobic strain on the non—fortified agar. The outcome was that colonies were distributed evenly over the plate, TABLE 10 THE EFFECT OF CONCENTRATION OF SPAN 80 IN AGAR 0N COLONIAL GROWTH OF COZ-DEPENDENT AND AEROBIC‘QE. ABORTUS Concentrationa pH Colony Count ___ (percent) . Strain 119 Strain 2305‘ 0.0001 7.20 very lowb highc 0.001 7.23 0 " 0.01 7.3 0 " 0.1 7.2h 0 moderated Control 7.5 very lowb highc 8 expressed on a liquid basis ' 70 to 100 colonies per plate = lhOO to 2000 colonies per plate = 1000 colonies per plate. 00 0" TABLE 11 THE EFFECT OF THREE AGENTS ON THE COLONIAL GROUTH OF A Goa-DEPENDENT AND AN AEROBIC‘EQ. ABORTUS (Paper disc method) Colonial Agent - Zone Zonala 001°?! Diameterfk Strain on Paper Type Radius Next to At Platels Disc from Disc Disc Rim 119 Albumin Enhance- 18 1.14 - 0.1 0b (h.8%) ‘ment Tween 80 Enhancc- 1h 1.2 - 0.1 0 (10%) ment Span 80 Inhibition 35 0 0 (10%) (complete) Control 0.6 - 0.1 2308 Albumin * lel '8 1.0 1e]. 8' 0e (aerobic) (h.8%) ' Tween 80 a 1.2 - 0.9 1.3 - 1.0 (10%) 3 an 30 * 1.0 - 0.3 ,1.2 - 1.0 10%) COntrOI leg 8 Deg a 8 in.mm. b 8 no detectable colonies i 8 colonies distributed evenly over agar surface TABLE 12 THE EFFECT OF THREE AGENTS ON COLONIAL GROWTH UPON TRYPTOSE AGAR FORTIFIED WITH KILLED BRUCELLA CELLS (Paper disc method) Agent on Zone Zonala Colonial Diameters Paper Disc Type rRadiB; Next to At Plato's 1' °“ ’° Disc Rim Albumin * ~ l.h - 1.0 1.2 - 0.9 (14.8%) Tween 80 31. enhancement -S.0 1.2 - 1.0 1.0 - 0.7 (10%) b Span 80 Inhibition 5.0 0 0.1 - 0.9 (10%) Control 1.1 - 0.7 a 8 in.mm. b 8 no detectable colonies a 8 colonies distributed evenly over agar surface although those adjacent to the pad containing albumin, showed a very slight enlargement. In table 12, the notation, slight enhancement, for Tween 80, indicates that colonies tended to cluster more densely in the region of the pad. Span 80, however, still inhibited colonial development, though on this medium such action was neutralized and ultimately overcome as diffusion in agar reduced the Span concentration. Since both albumin and Tween 80 have been shown to promote colonial development, it is not surprising that on fortified medium, this effect was masked. Yet, regard- less of whether the fortified agar permitted normal colonial deve10pment or not, inhibitory action of Span persisted. This result is further evidence of a fUndamental difference underlying strains 119 and 2308. 7. The effect of two fatty acids on colonial growth. In the extensive studies on the synthetic Tween and Span products by Dubos and cedworkers, reference is frequently made to the fact that fatty acids gradually form in these agents by hydrolysis, and as a result, manifest an inhibitory effect on small inocula of tubercle bacilli. Because of this report, and the unanswered question concerning the essence of growth-enhancement action by albumin, determination of the effect of fatty acids, alone, on the colonial growth of 25! abortus 119 and 2308 was next undertaken. From aqueous stock solutions, the sodium salts of the organic acids chosen for test, lactic and oleic, were added to the agar medium.before autoclaving. For comparative purposes, the final concentration of the two acids is presented in terms of molarity (table 15), since at room temperature one salt is a liquid, the other a solid. Again in table 13, a comparison is made of the two brucella strains, and a colony count fluctuating in response to variation in pH is again TABLE 13 THE EFFECT OF TWO ORGANIC ACIDS IN AGAR 0N COLONIAL GROWTH OF A Goa-DEPENDENT AND AN AEROBIC BR. ABORTUS Colony’Count Agent Concentration pH Strain 119 Strain 2300 sodium 1.07 x 10'114. 7.06 0 highd lactate _ b 1.07 x 10 214. His low " 1.07 x 10'2M. 7.06 very lowa " 1.07 x 10’3M. 7.38 " " " 1.07 x 10"”14. " " " Sodium 6.6 x 10'3M. 0 oleate _3 c 3.28 x 10 M. moderate 3.28 x 10"“14. 7.1;8 0 highd 1.61. x 10-184. 0 " 3.28 x 10-614. 7.06 0 " (l micro-gm./ml) 3.28 x 10-711 ‘ 7.1;8 lowb " 3.28 x lO'BM. 7.2 very low‘ " Control 7,5 u u u a 8 70 to 300 colonies per plate b 8 300 to 500 " " " c 8 725 " " V d 8 1&00 to 2000 " " " -28- noted. The aerobic strain was not inhibited by any test concentration of lactic acid. Though the cog-dependent strain was finally inhibited by a 0.107 M lactic acid concentration, colonies may even have developed at this molarity had the growth-favoring pH of 7.5 also been tried, as may be concluded from the pH-variable colony count recorded for the 1.07 x 10-8 M concentration. 0n the contrary, oleic acid, while still very dilute, was inhibitory to both brucella strains. Still, the aerobic culture tolerated a concentration of this fatty acid exceeding by approx- imately 2000 times that dilution effecting inhibition of the fastidious strain. And for this acid also, a more alkaline pH favored a numerical increase in colony count of strain 119 even though oleic acid concentra- tion simultaneously increased ten-fold, from 3.28 x 10‘.8 M, to 3.28 x 10-7M. Tables lb and 15 report the effect on strain 119 of oleic and lactic acid, respectively, when assayed by means of the paper disc method. Dilution both by absorption on the paper disc and by diffusion into agar caused this method of assay to show toleration of a higher concentration of both these acids than was recorded for the method in which each acid was incorporated directly in the medium (table 13). The effect of pH is apparent in these two tables, also. 'lhen either fatty acid was at the threshold of the inhibitory concentration, the degree of inhibition was modified by pH, so that an alkalinity increase from pH 7.2 to pH 7.5 promoted both enlargement of colonial diameter and deve10pment of colonies where none would otherwise have appeared. Accordingly, in table 15, complete inhibition of strain 119 by undiluted lactic acid is probable had the medium also been tested at pH 7.1. TABLE 1h THE EFFECT OF MEDIUM COMPOSITION AND OF CONCENTRATION OF SODIUM OLEATE UPON COLONIAL GROWTH (Paper disc method) Sodium Tryptose Agar Tryptose Agar Fortified Oleate, Agar with Killed Brucella Cells Concentration PH Zone Colonial Diameter Zone Colonial Diameter Placed on DISC Type (in m.) Type (in mm.) 3.28 x 10-6Me 705 * 0.1 X 006 (1d) 3028 x 1.0-SM. 7e5 * Oel ' OeS * OeS "' 1.0 (104) 6.56 1d 10-5n. 7.2 a 0.1 - 0.5 a 0.5 - 1.1 (20(1 ) 1.31 zdlo-hn. 7.5 t 0.1 - 0.3a * 0.6 - 0.9 (NO ) and slight inhibition 1.W6( 1 lo’hn. 7.2 complete 0b SOd inhibition 1'67d 1 lo'hn. 7.5 slight 0.1 - 0.u‘ Sod inhibition° 1.97 x 10-hh. 7.5 a 0.6 - 0.78 (605) and slight inhibition 2.30 xle'uH. 7.5 moderate 0.1 - 0.281 a 0.2 - 0.7a (70 ) inhibition? and slight inhibition 3. 28 1610.454 7.5 e 0.2 .. 0.1.“ (100 ) and moderate inhibition Control 7.5 * 0.1 - 0.6 a 0.5 - 1.2 a 8 in.3 mm. radial son:e from.disc 1 B : 38r§° ggtgg gnigsogeygnd the 3 mm. radius from.disc d 8 micro rams m1. a 8 colon es distributed evenly over agar surface TABLE 15 THE EFFECT OF pH AND CONCENTRATION OF SODIUM LACTATE ON COLONIAL GROWTH (Paper disc method) Sodium. Zonal Colonial Diametersa Lactate Agar Zone Radiusa Concentration pH Type From Next t0' At Plato's Placed on Disc Disc Disc Rim - 1007 I 10-3Me 7e; * 0e? “(Del 0e? '<0e1 1.07 M. 7.5 Slight 0.2 -