A STUDY OF THE BRUCELLA GROWTH-INHIBITING FACTOR IN BOVINE SERUM AND COLOSTRUM By Marvis Anne Richardson A THESIS Submitted to the Sohool of Graduate Studies of Michigan State College of Agrioulture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Baoteriology and Public Health 1950 TABLE OF CONTENTS I INTRODUCTION....................................................1 II GENERAL PROCEDURES........................... j.................. b III EXPERIMENTS AND RESULTS A. Study of the Agents Used for Determining the Brucella Growth-Inhibiting Action of Normal Bovine Serum and Colostrum......... 7 1. Normal rabbit serum as a source of complement...... ....7 2. The growth-inhibiting aotion of normajL rabbit serum and its effeot on the action of normal bovine serum....12 3. The growth-inhibiting action of normajL bovine serum on five different growth phases of Bruoella abortus... .11; U. Potentiating aotion of sodium sulfadiazine on the growth-inhibiting action of normal bovine serum and plasma....... 17 B. The Brucella Growth-Inhibiting Aotion of Normal Bovine Serum.................................... 1. Serum from cows.................... 2. Serum from newborn calves........... 20 20 21 i C. The Brucella Growth-Inhibiting Action of Normal Bovine Colostrum.......................... 23 1. The growth-inhibiting action of colosbrum.......... ...25 2. A colostral growth-inhibiting aotion independent of complement....................... 27 D. The Brucella Growth-Inhibiting Factor of Bovine Serum Fractions Prepared by Sodium Sulfate Precipitation.......... 30 1. Fraotionation of specific antiserum.................. 31 a. Preparation of fractions...... ...31 b. Growth-inhibiting aotion...... 33 c. Electrophoretio analyses...... 3U 2. Fractionation of normal serum ••»3b a. Preparation of fractions .............. 3U b. Growth-inhibiting action....................... 3& c. Electrophoretio analyses...... 39 3. Comparison of specific and normal serumfraotions......I4.I E. The Effeot of Certain Physical Foroes on the Brucella Growth-Inhibiting Aotion of Bovine Serum, Plasma, and Colostrum................................................. i|6 1. The effect of heat.................. ...1*6 a. Normal serum and serum fractions................1*6 b. Specific antiserum............................. 50 2. 3. F. c. Plasma fractions....... ........... 50 d. Normal oolostrum................................ 33 The effeot of storage....... ....*5U a. Normal serum........ .................. ..... b. Normal serum fractions.................. 3& c. Normal oolostrum................................ 37 The effect of filtration............................. 60 a. Normal serum.................................. *.60 b. Normal oolostrum................................ 63 The Effeot of Normal Bovine Serum and Colostrum on the Prozone Growth-Inhibiting Aotion of Serum from a BrucellaInfeoted Cow............................. ......... ....... 63 1. Prozone-inhibiting aotion............................. 63 a. Normal serum.................................... 66 b. Normal serum fractions.......................... 68 c. Normal oolostrum............................... .70 2. The effeot of oertain physical forces on the prozone-inhibitingaotion of serum..................... 70 a. Effect of heat.................................. 71 b. Effect of storage...............................71 c. Effect of filtration.................... J2 IV DISCUSSION.................................................... 73 V SUMMARY....................................................... 79 VI LITERATURE CITED.............. ............................... 81 VII ACKNOWLEDGEMENTS.............................................. 85 < A STUDY OF THE BRUCELLA GROWTH-INHIBITING FACTOR IN BOVINE SERUM AND COLOSTRUM In view of the faot that previous studies had demonstrated a po­ tentiating aotion of sulfadiazine on the specific brucella antibody| complement system and the irole played by this combination of agents in infection (l), it seemed desirable to investigate in more detail the ! in vitro growth-inhibiting action of normal bovine serum and colostrum against Brucella with and without the presence of sodium sulfadiazine. I The growth-inhibiting aotiyity of normal serum and colostrum in the presence of sodium sulfadiazine was studied in the hope that it might provide a new avenue of approach to the problem of the resistance of oalve3 to brucellosis and clarify the significance of the natural baoj terioidal antibody. The resistance of calves up to 6 months of age and its laok of de­ pendence on the presence of speoific antibody (2) are well established i faots. Inasmuoh as accumulative evidenoe (3), (h), (5) indicates a rela­ tionship between the speoific baoterioidins and the resistance of cattle to brucellosis, investigators have sought to demonstrate a correlation bei tween the natural bactericidins and the high resistance of calves. Irwin and Bell (6), who bred a strain of rabbits of high resistance to Brucella suis and another of low resistance, found that a correlation existed between the bactericidal activity of the blood and the resistance of the animal. Huddleson et al. (3) found no direct proof that the normal plasma baoterioidins of calves acoounted for their high resistance to brucella infeotion. However, these authors demonstrated that plasma from newborn oalves showed iitoie, or no, bactericidal action until after the ingestion of colostrum and argued against ignoring the immunological significance of the natural antibody. It is now well established that oolostrum, the first milk after par­ turition, is the sole means of specific antibody transfer from the rumi­ nant to its offspring (7 ). Within recent years Smith (8), (9 ) has isolated the oomponent of oolostrum associated with the passive transfer of speoific antibodies to calves. This author (10) and Smith and Holm (11) have shown that this component and certain specific antibodies per­ sist in the blood of the calf for many months. If one accepts the natural immunity phenomenon as analogous to that of the immune state, oolostrum may be assumed to be the bearer of those substances imparting natural immunity to the calf. This is substantiated by the appearance of normal baoterioidins in the blood of the calf after the ingestion of oolostrum from normal oows. The investigation of naturail colostral antibodies against Brucella appears to have been neglected. Since a dairy herd known to have been free from brucellosis for more than 15 years was available, it afforded an opportunity to study serum and oolostrum normal with respect to Brucella. In the present study the sulfadiazine-antibody-complement system has been used to demonstrate the presence of brucella growth-inhibition in colostrum and in the serum of calves at birth and to compare the growth-inhibiting aotivity of serum from nev/born calves after the ingestion of colostrum with that of serum from adult oows. Although the original purpose was to study the growth-inhibiting activity of serum from newborn calves and adult cows, certain observations on the effeot of heat on this activity directed attention to character­ istics of the normal growth-inhibiting factor which had not been reported and to fundamental differences between the normal and specific growthinhibitora. Fractions of specific and normal bovine serum were prepared and examined in an attempt to characterize the normal antibody as compared with the specific antibody. In early experiments designed to study the bacterioidal activity of plasma from animals having different statuses with respect to brucellosis, the brucella-infected animal was shown to be distinct in that its plasma failed to inhibit growth of Brucella (3)» Huddleson (l) later found that this phenomenon was a manifestation of the well known prozone reaction and that prozone growth oould be inhibited In vitro and in vivo in the guniea pig by the addition of fresh, normal rabbit serum and sodium sul­ fadiazine. In the follo’wing experiments normal bovine serum and colostrum have been shown to be capable of prozone-inhibition in the presenoe of sodium sulfadiazine, and the possible identity of this factor and the growth-inhibiting factor has been investigated. GENERAL PROCEDURES The method of Huddleson (l) for the in vitro estimation of growthinhibiting action against Brucella was used throughout the experiments. With the exceptions noted in individual experiments the procedure was as follows t Nature and source of blood specimens. — Since it had been demon­ strated that citrated plasma exerted a greater bactericidal action than serum and that the enhanced aotion was proportional to the amount of sodium citrate used as anticoagulant, serum was used in this study. How­ ever, some of the first speoimens were plasma and certain plasma growthinhibition tests oontain comparative data that seem to permit their inolusion. All other samples were serums. Blood speoimens were collected from cow3 and oalves normal with re­ spect to infection with brucella organisms and from one brucella-infected cow. The normal animals were in a herd known to have been free from brucellosis for more than 15 years. ever been treated with vaccine. None of the animals in this herd had Colostrum was obtained from cows in the same herd. Preparation of blood serum. — The blood was collected aseptically and allowed to clot at room temperature. If the bactericidal test wa 3 carried out on the day of bleeding, unfiltered serum was used. Bovine serums to be stored and tested at intervals were filtered through a Hormann D8 pad in a Seitz filter and stored at I40 C. Normal rabbit serum collected on the day of use served as the source of added complement. Preparation of colostral whey. — Following the separation of fat and its removal, the oolostrum was heated to 37* C. Sufficient commer­ cial ronnin to precipitate the casein within 1 hour was added and the whole incubated for 1 hour at 37° C. The whey was sterilized by filtra­ tion through a Hormann D8 pad in a Seitz filter and stored at Lj0 C. Preparation of sodium sulfadiazine. — A solution containing 0.1 mg. of sodium sulfadiazine per ml. was prepared by adding 1.0 ml. of a 1.0percent solution, which had been sterilized by filtration through a Hormann D6 pad in a Seitz filter, to 99 ml. of sterile distilled water. The solutions were prepared on the day of use. Preparation of bacterial suspension. — A smooth strain of Bruoella abortus was used in all titration experiments. The organisms were grown on beef liver infusion agar slants for 21; hours at 37° C. The growth was removed, suspended in sterile dilution fluid, and the suspension stand­ ardized by a Libby photronrefleotometer to a viable colony count of 1 x 10 per ml. Titration of the growth-inhibiting aotion. — All the titration experiments were made in liquid oulture medium containing 1.5-percent Difco's "Baoto-Tryptose", 0.5-percent glucose, 0.5-percent sodium ohloride and 0.5 mg.-percent thiamine hydrochloride. The medium was filtered through paper and the pH adjusted to 6.7 with phosphoric acid. It was then dispensed in 5~ or 10-ml. amounts and sterilized at 115° C. for 15 minutes. The growth-inhibiting aotion of the sample was determined as followsj Serial twofold dilutions of the sample were made in 5 ml. of the culture medium. The other agents, 0.2 ml. fresh rabbit serum, 0.1 ml. of sodium sulfadiazine solution, and 0.1 ml. of the bacterial suspension (10^ viable organisms), were added to each tube; the organisms were always added last. After thorough shaking, the cultures were incubated at 37° C. for 72 hours The degree of visible growth was estimated at 2l|-hour intervals and re­ corded by the use of + and - signs. The highest serum dilution showing no visible growth at 72 hours is referred to as the growth-inhibiti titer. EXPERIMENTS AND RESULTS A. Study of the Agents Used for Determining the Brucella Growth-Inhibiting Aotion of Normal Bovine Serum and Colostrum Experiments were conducted to examine (l) the effioaoy of normal rab­ bit serum as a source of complement for measuring the grovrth-inhibiting activity of normal bovine serum in the presence of sodium sulfadiazine, (2) the effect of the growth-inhibit or in normal rabbit serum on the growth-inhibiting activity of normal bovine serum, and (3) the growthinhibiting aotion of normal bovine serum and sodium sulfadiazine on dif­ ferent growth phases of Br. abortus. It was felt that the latter two effects might constitute significant souroes of error in the determina­ tion of the aotivity of normal bovine serum, which is of relatively low order. In addition, while examining the reagents of the test, the mutual enhancement of the growth-inhibiting action of sodium sulfadiazine and normal bovine serum was demonstrated. 1, Normal rabbit serum as a source of complement It is an established faot that complement is required in bactericidal systems, but the status of the interohangeability of complement from dif­ ferent species remains uncertain. Mackie and Finkelstein (12) found that absorbed serum from one species was capable of acting as complement with the natural bactericidal antibody from another speoies irrespective of the animal and type of typhoid-paratyphoid organism studied. rabbit serums were among those apparently interchangeable. Bovine and Irwin et al. (13) reported that unabsorbed guinea pig serum was unable to restore the bactericidal action of stored or heated bovine plasma against Br. abortus. Pursuing this investigation, Shrigley and Irwin (llj.) showed that the absorbed serum complements from different animal souroes were not always interchangeable in normal bactericidal action against Br, suis. ticular, rabbit complement activated only heated rabbit serum. In par­ Recently Coombs and Hale (15) have demonstrated the importance of the choice of complement when examining antiserums for the presence of complementfixing antibodies, Huddleson (l) found complement essential to the growth-inhibiting system acting in the presence of sodium sulfadiazine. In his work fresh, normal rabbit serum served as an adequate source of complement for the action of serum from bruoella-infected guinea pigs, cows, and rabbits. Before proceeding with this study it seemed desirable to demonstrate the efficacy of fresh, normal rabbit serum as a source of complement for the bruoella growth-inhibitor of normal bovine serum in the presence of sodium sulfadiazine. The natural complement of three fresh bovine serums was inactivated by heating samples at 56 ° C. for 30 minutes. The activity of the heated serums was determined in the presence of sodium sulfadia­ zine both with from the data and without fresh, normal in table 1 that the heated no growth-inhibiting aotion. rabbitserum. It may be seen serumswithout complement showed However, the original aotivity was not re­ produced upon the addition of rabbit complement. Sample 1, which origi­ nally possessed a prozone reaction, was less active, A prozone reaction and lowered aotivity appeared after heat inactivation of sample 2, and no aotivity was evidenced by the heated sample 3« These examples oited are representative of numerous similar results, many of which are inoluded in the section on Since the the effect of heat on the normalgrowth-inhibitor. modified action of heat-inactivated serum might result Table 1 Effect of Fresh Normal Rabbit Serum on the Growth-Inhibiting Action of Heat-Inactivated Normal Bovine Serum Degree of growth, turbidity8 Serum No. 1 2 3 Serum Agents added to serum dil. Dil. of serum, 1: 10 20 40 80 160 Bacterial control Unheated NaSD 3+ + - 2+ 4+ 3+b Heated NaSD 5+ 4+ 4+ 4+ 4+ 3+b Heated NRS and NaSD 5+ 4+ 2+ 3+ 4+ 4+c Unheated NaSD - - - + 4+ 4+b Heated NaSD 4+ 4+ 4+ 4+ 4+ 4+b Heated NRS and NaSD 4+ - - 4+ 4+ 4+ o Unheated NaSD - - 1+ 3+ 4+ 4+b Heated NaSD 4+ 4+ 4+ 4+ 4+ 4+b Heated NRS and NaSD 4+ 4+ 4+ 4+ 4+c a10° Br. abortus added to each dilution and to con­ trols (5 m l . ). “Control tube of medium also contains 0.01 mg. sodium sulfadiazine. “Control tube of medium also contains 0.2 ml. of fresh, normal rabbit serum and 0.01 mg. sodium sulfadiazine. NRS = 0.2 ml. of fresh, normal rabbit serum. NaSD = 0.01 mg. sodium sulfadiazine. Incubation period, 72 hours. Serum heated at 56° C. for 30 minutes. - = no visible growth. + = degree of growth. from thermolability of the antibody or inadequacy of the complement, the effect of fresh, normal rabbit serum on the grov/th-inhibitor of normal bovine serum and serum and plasma fractions which had been rendered com­ plement-free by means other than heat was investigated. were stored at I*0 C. for 5 and 6 weeks respectively. Serums 1 and 2 Fraction 3 was that portion firecipitated from serum, as described elsewhere, by 20-percent sodium sulfate after removal of the material precipitated at 15-percent concentration of the salt. Two v/eeks intervened between bleeding of the animal and the determination of growth-inhibiting activity, and, in view of the manipulation in fractionation, dialysis, etc., it is unlikely that active complement remained. Fraction 1+ was precipitated from serum by dialysis for 21| hours against a pH 5«U phosphate buffer of 0.02 ionio strength. Although this method of isolation produces a minimum of protein denaturation, guinea pig and human complement are knoy/n to be inactivated by its use. The grov;th-inhibiting activity of this serum fraction was tested both 1 week and 8 weeks after bleeding of the animal. The data set forth in table 2 show that in the presence of sodium sulfadiazine fresh, normal rabbit serum functioned as complement with serums whose natural complement had deteriorated during storage and with fraotions of normal serums devoid of complement as a result of the teohnique of prepa­ ration. Certain dilutions of a 0.2-percent solution of Armour and Company's fraction III-l from bovine plasma v:ere found to inhibit grovrth in the presence of fresh, normal rabbit serum and sodium sulfadiazine. While it is highly improbable that such a pooled plasma represented only animals normal v/ith respect to Brucella, the data in table 2 support the evidence that fresh, normal rabbit serum serves as an adequate source of complement for the sulfadiazine-potentiated action of the bovine groy/th-inhibitor Table 2 Effect of Fresh Normal Rabbit Serum on the Growth-Inhibiting Action of Normal Bovine Serum, Serum Fractions, and a Plasma Fraction Sample No. Substance Time between bleeding and testing (weeks) Agents added to sample dil. Degree of growth, turbidity8Dilutions of sample, 1: H H cl ri Ot R Ct o o o o o o o o o rH rH rH iH rH iH rH 1— 1 X X X X X X X! * o o o o uj oj ^ CO . . . . . . . • Bacterial 1—1 0] 20 H to :-0 1—1 control 1 Serum, whole 5 NaSD NRS and NaSD 5+ 5+ 4+ 4+ 4+ 2+ 1+ 4+ 4+ 3+b 3+c 2 Serum, whole 6 NaSD NRS and NaSD 4+ 4+ 4+ 4+ 4+ 2+ 1+ 2+ 3+ 4+ 4+b 4+c 3 Serum, fraction:. Pptd. at 20 % Ma2S04 2 NaSD NRS and NaSD 4+ 4+ 4+ 4+ 4+ 4+ 4+b 4+c 4 Serum, fraction: Pptd. at pH 5.4-0.02p 1 NaSD NRS and NaSD 4+ 4+ 4+ 1+ 4+b 4+c 4 Repeated 8 NaSD NRS and NaSD 4+ 4+ 4+ 2+ 4+b 4+c 5 Plasma, fraction: Armourrs III-l (0.2%) - NaSD NRS and NaSD 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 5+ 4+ 4+ 4+b 4+c ?10B Br. abortus added to each dilution and to controls (5 ml.). Control tube of medium also contains 0.01 mg. sodium sulfadiazine. cControl tube of medium also contains 0.2 ml. of fresh, normal rabbit serum and 0.01 mg. sodium sulfadiazine. NRS = 0.2 ml. of fresh, normal rabbit serum. NaSD = 0.01 mg. sodium sulfadiazine. Incubation period, 72 hours. - = no visible growth. + = degree of growth. against Br. abortus. 2. The growth-inhibiting action of normal rabbit serum and its effect on the action of normal bovine serum Since the agent, normal rabbit serum, used as a source of complement was known to possess slight bactericidal activity (l), (li;), (l6), the growth-inhibiting activity per se was investigated. In order to determine the activity, nine fresh, normal rabbit serums were examined on the day of bleeding, thus utilizing the complement of the serum sample instead of added complement. Although fresh, normal rabbit serum in the presence of sodium sulfadiazine proved capable of inhibiting growth of Br. abortus when diluted 1:5, and often at 1:10, in no instance did it show any ac­ tivity at a dilution of 1:20 (table 3)» That this action was not limited by dilution of the complement is evident from data offered elsewhere in which normal rabbit serum in a dilution of 1:25 invariably sufficed for added complement. ■While normal rabbit serum was unable to inhibit growth in a dilution of 1 :25, its concentration when used as a reagent in the test, the pos­ sibility of an additive action was investigated. The growth-inhibiting ability of nine fresh, normal bovine serums and one plasma was determined within 2I4 hours of bleeding. D8 pads in Seitz filters. Five samples we re filtered through Hormann The growth-inhibiting activity of the five filtered and eight unfiltered samples in the presence of sodium sulfadia­ zine, both with and without fresh, normal rabbit serum, is shown in the data of table ij.. Of the 13 samples tested, nine exhibited a slight in­ crease in activity in the presence of normal rabbit serum, and four in­ hibited growth at the same dilution with and without normal rabbit serum. The effect was not apparent beyond one twofold dilution. It may be noted that serums 6 and 7 failed to inhibit growth completely in dilutions of -13A- Table 3 Growth-Inhibiting Action of Fresh Normal Rabbit Serum Degree of growth, turbiditya Dilutions of serum, 1: Rabbit No. 10 20 1 - 4+ 4+ 2 - - 4+ 4+ 3+ 3 - 1+ 4+ 4+ 3+ 4 - 4+ 4+ 3+ 5 - 4+ 4+ 4+ 4+ 6 - 2+ 4+ 4+ 4+ 7 2+ 2+ 4+ 4+ 4+ 8 - - 4+ 4+ 4+ 9 - 4+ 4+ 4+ 4+ - 40 Bacterial control 5 3+ a10° Br. abortus and 0.01 mg. sodium sulfadia­ zine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. - = no visible growth. + = degree of grovfth. i -13B- Table 4 Effect of Fresh Normal Rabbit Serum on the Growth-Inhibiting Action of Normal Bovine Serum and Plasma Degree of growth, turbidity8 Sample No. Agents added to sample dil. 1. Plasma,unfiltered NaSD NRS + NaSD 2. Serum,unfiltered NaSD NRS + NaSD 3. Serum,unfiltered NaSD NRS + NaSD unfiltered NaSD NRS + NaSD NaSD NRS + NaSD 4. Serum, filtered unfiltered 5. Serum, filtered NaSD NRS + NaSD NaSD NRS + NaSD filtered NaSD NRS + NaSD NaSD NRS + NaSD 7. Serum,filtered NaSD NRS + NaSD 8. Serum, filtered NaSD NRS + NaSD unfiltered 6. Serum, Dil . of sample, 1: 10 20 40 80 160 320 Bacterial control - - 2+ 3+ 3+ - - - - 3+ ± 3+ 4+ 3+b 4+c - C 1+ - C 3+ 3+ 4+c mm mm 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+b 4+c 4+b 4+c 1+ 4+ 4+ 4+ 4+ 4+b 4+c 4+b 4+c mm mm - - 3+b 4+c 4+ 4+ 3+b 4+b 1+ 2+ 2+ 4+ 2+ 4+ 4+ 3+ 4+ 1+ 1+ 4+ 4+ 4+ 4+c 4+b 4+ 4+ 4+ 4+ 4+ 4+ 4+b 3+ 4+ 4+ 4+ 4+ 4+b 4+c 2+ 4+ dt 5+ 4+ 5+ 1+ 1+ 1+ 1+ - mm 9. Serum,unfiltered NaSD NRS + NaSD - - 10. Serum,unfiltered NaSD NRS + NaSD - - ± - 4+ 3+ . 4+c 4+c 4+b 4+c 4+b 4+ 4+c a106 Br. abortus added to each dilution and to controls (5 ml.). Incubation period, 72 hours. ^Control tube of medium also contains 0.01 mg. sodium sul­ fadiazine. cControl tube of medium also contains 0.2 ml. of fresh, normal rabbit serum and 0.01 mg. sodium sulfadiazine. NRS = 0.2 ml. of fresh, normal rabbit serum. NaSD = 0.01 mg. sodium sulfadiazine. C = contamination. - - no visible growth. + = degree of growth. lrlO and 1:20 when normal rabbit serum was present; no prozone growth was evident without it. The normal rabbit serum used as a source of complement was incapable of growth-inhibiting action in the dilution at which it was used as a reagent in the sulfadiazine-antibody-complement test. However, it appeared to contribute slightly to the activity of normal bovine serum. Although the additive effect did not exceed one twofold dilution, which is within the error of the method, variations in the aotion of normal rabbit serum indicated that tests conducted with the same fresh rabbit serum would be more strictly comparable. Therefore, with the exception of samples tested after intervals of storage, all comparative data presented in the experi­ ments were obtained from tests conducted with the same normal rabbit serum. 3. The growth-inhibiting action of normal bovine serum on five different growth phases of Br. abortus Differences in sensitivity of diverse phases of organisms to various agents is generally recognized. Huddleson et al.(3) demonstrated that the plasma baoterioidins affect smooth and rough phases of Br. abortus to a different extent and emphasized the importance of an undissooiated cell suspension for testing growth-inhibition. The question arose as to whether other dissociated phases, in particular those not so readily distinguished from the smooth phase, might evidence differences in sensitivity and con­ tribute errors to the results of the test. Five growth phases of Br. abortus were compared as to sensitivity to the growth-inhibiting aotion of normal bovine serum in the presenoe of sodium sulfadiazine. Colonial characteristics of the phases were dis­ tinguished, in the manner described by Huddleson (17)» by examination of oolonies incubated for i| days at 37° C. on tryptose agar medium using a low power stereoscopic microscope with an oblique light source. Of the five growth phases studied, the smooth-intermediate SI^ most closely resembled the smooth S phase. Isolated SI-^ colonies could not be distinguished by sight from those of the S phase, but adjacent S and SI^ oolonies showed differences in opacity? the SI-^ oolony was the more opaque. Colonies of the smooth-intermediate SIg phase had more opaque oenters and oadmium yellow color than those of the S and SI^ phases. Colonies of the mucoid phase appeared very similar to those of the S and SI^ phases except for their slight mucoid oonsistenoy. phase studied was sector SI-^. The fifth Colonies of this growth phase were dis­ tinguished from the S phase by sectors extending from the oenters to the borders of the colonies. Each succeeding transfer from a single sector oolony on agar medium resulted in SI^ and sector SI^ oolonies. Data which demonstrate the comparative sensitivity of four growth phases of one Br. abortus strain to the growth-inhibiting action of a fresh and a stored serum are presented in table 5 « those obtained with other serums. The results typify It may be noted that, in the prosenoe of sodium sulfadiazine, growth of the M 1# S, SI^ and sector SI^ phase was inhibited by fresh serum 1 in dilutions of 1*10, lil+O, lil+Of and I 18O respectively. Although the action of stored serum 2 was less pronounoed, the same relative sensitivity of the four growth phases was apparent. Therefore, of the four phases, proved the least sensitive, S and SI1 of equal sensitivity and exoeeding that of M^, and the sector phase the most sensitive to the growth-inhibiting action of normal bovine serum under the conditions of the test. An inspeoticn of the results in the bacterial control tubes reveals that the relative degrees of sensitivity of these four phases of the same Br. abortus strain are reflected therein. -16- Table 5 Comparative Sensitivity of the Smooth Phase and Four Dissociated Phases of Brucella Abortus to the Growth-Inhibiting Action of Normal Bovine Serum Degree of growth, turbidity8- Dilution of serum, 1 : Serum No. n X 2 3 u A 4 Phase Smooth, S Intermediate, Six Mucoid, Mx Sector, Six Smooth, S Intermediate, Six Mucoid, Mx Sector, Six 10 - 20 2+ 80 160 _ 5+ 4+ 6+ 5+ 5+ C 2+ 5+ *• 2+ 2+ 4+ 2+ 3+ 5+ Smooth, S Intermediate, SIa Smooth, S Intermediate, SIa 40 4+ 4+ 5+ 1+ 4+ 4+ 6+ Bacterial control 5+ 6+ 3+ 5+ 5+ 6+ 3+ 2+ 4+ 4+ 6+ 3+ 2+ 4+ 4+ 6+ 2+ 3+ 4+ 4+ 1+ 4+ 4+ 1+ 4+ 1+ 3+ ** * 320 4+ * a106 J3r. abortus, 0.2 ml. fresh, normal rabbit serum, and 0.01 mg. sodium sulfadiazine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. C = contamination. - = no visible growth. + = degree of growth. 4 The phases thus varied in sensitivity to normal "bovine serum acting in the presence of sodium sulfadiazine and to sodium sulfadiazine alone. The data also show a comparison of the sensitivity of an S and a smooth-intermediate SIg phase to the growth-inhibiting aotivity of two normal "bovine serums. Growth of the SIg phase was inhibited by serums 3 and 1; in dilutions of lrl60 and 1*80 respectively; growth of the S phase was inhibited by these serums only in dilutions of lil+O or less. sensitivities of S and SIg phases from another strain were of the same relative order as those above. phase is more The The results demonstrate that the SIg sensitive than the S phase and, accordingly, more than the SI^ phase to the growth-inhibiting aotion of normal bovine serum. Inspection of the data in table 5 reveals the wide range of bovine serum titers obtainable with bacterial suspensions of five pure growth phases. Differences that might be found in a series of tests due to changes in the growth phase of the organism used are readily appreciated. Not so readily known and considered is the similarity in the appearance of the smooth, mucoid M p to the untrained observer). and smooth-intermediate SI^ phases (even SIg The neoessity for constant vigilance to obviate such error is apparent. Ij. Potentiating aotion of sodium sulfadiazine on action the growth-inhibiting of normal bovine serum and plasma It has been shown by Huddleson (l) that oertain ooncentrations of sodium sulfadiazine and specific ariti-bruc ella serums which are ineffec­ tive by themselves exert growth-inhibiting aotion in a zone of dilutions when acting together. This author demonstrated that fresh, normal bovine serum enhanced the temporary growth-inhibiting effeot of sulfonamides to a bactericidal effect. In order to determine whether the presence of sodium sulfadiazine increased the activity of the growth-inhibitor in normal bovine serum against Br. abortus, the activity of serums and plasmas from animals of different age groups was determined with and without sodium sulfadiazine in the presence of fresh, normal rabbit serum. are typical of all the samples tested. The results in table 6 Both serum and plasma in the pres­ ence of complement alone proved unable to inhibit the growth of Br. abortus for 72 hours under the conditions of the test (several serums at low di­ lutions had shown a transient growth-inhibiting action at 2l+ hours). "When sodium sulfadiazine plus complement were present, the serum or plas­ ma of adult animals in dilutions of I 1I4.O and 1:80 inhibited visible growth. The sulfonamide showed a similar but less pronounced action with serum from newborn calves. That this effeot was not due to the action of sodium sulfadiazine alone was evidenced by the growth of Br, abortus in its presence in the control tube. The foregoing results show that the growth-inhibiting action of normal bovine serum in the presenoe of sodium sulfadiazine represents a mutual enhancement of temporary growth-inhibiting properties inherent in the two substances. Table 6 Potentiating Action of Sodium Sulfadiazine on the Growth-Inhibiting Action of Normal Bovine Serum and Plasma Degree of growth, turbiditya Sample No. Sample Dilution of sample, 1: Agents added to sample dil. 10 20 40 80 160 Bacte­ rial control 1 Serum, calfb NRS NRS + NaSD 6+ 6+ 6+ 6+ 6+ 3+ 4+ 4+ 6+b 4+c 2 Serum, calf^ NRS NRS + NaSD 6+ 6+ 6+ 6+ 6+ 2+ 2+ 4+ 5+ 5+ 6+b 5+c 3 Plasma, heifer NRS NRS + NaSD 6+ 6+ 6+ 6+ 6+ 3+ 3+ 6+b 5+° 4 Plasma, NRS NRS + NaSD 4+ 5+ 5+ 5+ 5+ mm 3+ 6+b 4+° 5 Serum, cow NRS NRS + NaSD 5+ 6+ 6+ 6+ 6+ 4+ 4+ 1+ 1+ 6+b 4+c 6 Serum, cow NRS NRS + NaSD 6+ 6+ 6+ 6+ 6+ 3+ 4+ 6+b 4+c 7 Serum, cow NRS NRS + NaSD 4+ 5+ 6+ 6+ e+ + + 5+ 6+b 4+c 8 Serum, cow NRS NRS + NaSD 4+ 5+ 5+ 5+ 5+ 3+ 6+b 4+c cow 10 Br. abortus added to each dilution and to controls (5 ml.). Incubation period, 72 hours. ^Control tube of medium also contains 0.2 ml. fresh, normal rabbit serum. cControl tube of medium also contains 0.2 ml. fresh, normal rabbit serum and 0.01 mg. sodium sulfadiazine. ^Newborn calf, before colostrum. NRS = 0.2 ml. of fresh, normal rabbit serum. NaSD = 0.01 mg. sodium sulfadiazine. - = no visible growth. + = degree of growth. B. The Brucella Growth-Inhibiting Action of Normal Bovine Serum 1. Serum from oows Many methods for estimating bactericidal activity against Brucella have been employed since Mackie and Finkelstein (16) demonstrated its presence in normal serum. These authors (12) devised an index method which involved the sterilization of a cell suspension using a oonstant amount of undiluted serum and decimal dilutions of a bacterial suspension. Using this method, Irwin and his associates (13) found that normal bovine Q serum destroyed 10 Br. abortus organisms per ml. in 2l+ hours. Irwin and Ferguson (U) extended the index method by adding complement to dilutions of serum in order to determine the highest dilution of serum capable of bactericidal action. Some activity was usually displayed by normal serum in a dilution of 1*25 and rarely in the next dilution, 1*125. Irwin and Beach (5) found that normal serum in the presence of added complement was active in dilutions of lsi+O or 1*80 and sometimes l:l6o. The bactericidin c in 0.3 ml. of undiluted serum from a normal cow killed 10^ organisms of a virulent strain in 2/4. hours. When diluted 1*10, 1*20, 1*^0 and 1*80, 2 3 1 1 0.3 ml. of the diluted serum killed 10 , 10 , 10 , and 10 organisms respectively. Huddleson and his associates (3) found that serums from 10 normal 5 6 animals destroyed 10 to 10 Br. abortus organisms per ml. in i;8 hours when the serum was in a dilution of 1*2 and various numbers of organisms were added. In establishing a procedure to reveal the maximum bactericidal activity of bovine serum or plasma, these authors, using both the number of organisms and amount of antibody as variables, came to the conclusion that the antibody dilution method showed more distinct differences in bactericidal action. The discovery that sodium sulfadiazine enhanced the bactericidal action of serum led to the development of a method em­ ploying this salt (l). In this method, which proved far more sensitive for measuring the action of specific antibody than those cited, the anti­ body was diluted in liquid culture medium and a oonstant number of organ­ isms added. The absence of, or suppression of, visible growth was termed growth-inhibiting action rather than bactericidal action. In order to determine the brucella growth-inhibiting activity of serum from normal cows, fresh, unfiltered serums from 12 cows were exam­ ined. Three cows possessed a grcwth-inhibition titer of lil60, eight of I 18O, and one of lil+O. Certain of these animals were bled on several occasions; the titer was reproducible in eaoh case. Normal bovine serum in the presence of sodium sulfadiazine and added complement thus was 5 capable of inhibiting the growth of 10 Br. abortus for 72 hours when in an average dilution of 1 *80. 2. Serum from newborn calves The high degree of resistance of young calves from normal dams to brucellosis led Huddleson and his associates (3) to investigate the bac­ tericidal aotivity of blood from young oalves as oompared to that of the adult animal. It was observed that plasma from newborn oalves exerted only slight, or no, killing action against a minimum number of organisms but that plasma from the same calves after the ingestion of colostrum killed large numbers of Br. abortus. However, the age at whioh calves showed maximum baoterioidal aotion varied and the aotivity did not exoeed that of the adult animal. Although no correlation between demonstrable bactericidins and resistance had "been shown, the discovery of a more sensitive method for detecting the normal antibody prompted re-examination of the bactericidal activity of serum from calves. Many investigators (7) have found the transmission of specific antibodies to the calf to depend on the ingestion of colostrum during the first day of life and shown that specific anti­ bodies reach a maximum concentration in the serum of the calf within 2i| hours after the ingestion of colostrum. If the substance responsible for the protection of young oalves against Brucella is transmitted by colostrum and demonstrable in the blood, it, too, should be present in maximum con­ centration immediately following the ingestion of oolostrum. To determine the growth-inhibiting activity in the presence of sodium sulfadiazine, serum samples were collected from four calves before and from three of the four oalves after the ingestion of colostrum from their dams. The oalves were allowed to remain with their dams from 2l| to IjB hours after birth and had free acoess to colostrum during the time. An examination of the data in table 7 discloses, in the serums taken before the ingestion of oolostrum, complete growth-inhibit!on at J2 hours by the serum of oalf 1 in a dilution of 1:20 and partial inhibition by the serums of calves 2 and 3 in this dilution. The serum of calf U, taken before colostrum but net shown in the comparisons of table 7» completely inhibited growth in a dilution of 1:20. Serum from calf 1 taken 2I4 hours after the ingestion of colostrum showed no significant increase in growth-inhibiting action. The activity of serums taken from oalves 2 and 3 forty-eight hours after the ingestion of oolostrum was definitely increased sinoe they completely inhibited growth for 72 hours in dilutions of 1:1*0 and 1:20 respectively as compared to partial inhibition in dilutions of 1:20 at birth. Table 7 Growth-Inhibiting Action of Serum from Newborn Calves Before and After the Ingestion of Colostrum Degree of growth, turbidity8Ingestion of colostrum Calf No. Incu­ bation period (hr.) Dil. of serum, 1: 10 20 40 60 Bacterial control 1+ 3+ 2+ 4+ 3+ 4+ 3+ 4+ Before 48 72 — — 24 hr. after 48 72 - - 2+ 2+ 4+ Before 48 72 2+ 2+ 1+ 4+ 2+ 5+ 4+ 5+ 48 hr. after 48 72 mm - - 2+ 4+ 2+ 4+ Before 48 72 2+ 2+ 3+ 3+ 4+ 3+ 4+ 48 72 — mm 48 hr. after - 3+ 2+ 4+ 3+ 4+ 1 2 4+ 3 Q B - 10 Br. abortus, 0.2 ml. fresh, normal rabbit serum, and 0.01 mg. sodium sulfadiazine added to each dilution and to controls (5 ml.). - = no visible growth. + = degree of growth. Although calves up to 6 months of age have been oonsidered resistant to infeotion, this study did not reveal that their serums show a higher growth-inhibiting action against Brucella than do those from older ani­ mals. Serums taken from calves I4.8 hours after the ingestion of colostrum, although more active than before the ingestion of colostrum, were less active than serums from normal adult animals. In addition to confirming the results of Huddleson et al (3) as to the increase in, and relative concentration of, normal growth-inhibiting entibody after the ingestion of colostrum, the sulfadiazine-antibody-complement test also revealed that eaoh calf possessed considerable growth-inhibiting activity prior to the ingestion of colostrum. Inasmuch as numerous workers (7 ) had failed to demonstrate specifio antibodies in calf serum before the in­ gestion of oolostrum, it was surprising to find growth-inhibiting anti­ body for Brucella in oalf serum at birth. C. The Brucella Growth-Inhibiting Action of Normal Bovine Colostrum Appreciation of the significance of oolostrum as a vehicle of anti­ body transfer to the young of certain species arose from the demonstration of speoific antibodies in the oolostrum and the appearance of these anti­ bodies in the blood of the young only after the ingestion of colostrum. Presumably oolostrum plays a similar role in the passage of natural anti­ bodies. Although bacterioidins have been shown to appear in the blood of the calf after the ingestion of oolostrum from normal cows, no inves­ tigations of the normal brucella antibody of colostrum itself have been noted. 1. The growth-inhibiting action of oolostrum In order to determine the brucella growth-inhibiting activity of colostrum from normal cows in the presenoe of sodium sulfadiazine, colostrums from eight different animals were examined. All the colostrums were either the "first milking" or taken within one day post partum since this is the period when colostrum has been shown to be particularly high in antibody content. In table 8 are shown the results of growth-inhibition determinations on the colostral wheys, fresh and heated at 56® C. for 30 minutes, with and without complement. The growth-inhibiting aotivity of the unheated colostrums, in the presenoe of sodium sulfadiazine and added complement, varied from slight inhibition in a dilution of 1:10 to complete inhibition in a dilution of 1:80. Seven of the eight colostrums possessed growth- inhibition titers, one of 1:10, two of 1:20, three of l:i|0, and one of 1:80. Several oolostrums exhibited partial growth-inhibition in higher dilutions. -26- Table 8 Growth-Inhibiting Action of Colostral Whey Irora No Degree of growth, turbicitya Heated colost Unheated colostral whey Colos­ trum No. NaSD NRS and NaSD NaSD Dilution of whey, 1: Dilution of whey, 1: Dilution of whey, 1: 10 20 40 80 1 - 2+ 4+ 2 ± 2+ 3 - 4 - 5 80 160 1 20 - 4+ I 4+ 4+ 4+ I 2+ - 1+ 2+ 3+ 1+ 3+ 4+ 4+ 4+ ± 4+ 4+ 4+ I 4+ 4+ 4+ - 4+ 4+ 4+ 2 - - ± 2+ 4+ 4+ 4+ 4+ 4+ 4 - - - 4+ 4+ 4+ 4+ 4+ 4+ 20 40 80 5+ - - 2+ 4+ 4+ 4+ - - - 4+ 2+ 4+ - - - 2+ 4+ 4+ - - 2+ 1+ 4+ 4H- 4+ 1+ 2+ 6 - 3+ 4+ 4+ - 7 4+ 4+ 4+ 4+ 8 - - 4+ 4+ - 40 10 10 160 D - 160 4+ 1 a10B Br, abortus added to each dilution (5 ml.). Incubation period, 72 4+ growth. “Whey heated at 56° G. for 30 minutes. “Colostral whey added to a 1:100 dilution of serum from a brucolla-infec which also contains 10B Br. abortus. 0.2 ml. of fresh, normal rabbit serum, NRS = 0.2 ml. fresh, normal rabbit serum added to each dilution. NaSD = 0.01 mg. sodium sulfadiazine added to each dilution. - = no visible growth. + = degree of growth. Table 8 Iting Action of Colostral Whey Irora Normal Cows Degree of growth, turbicitya NaSD NRS and NaSD Dilution of whey, 1: Dilution of whey, 1: Colostral whey added to test prozone growth inhibition0 (ml.) 10 20 10 0.5 - 4+ 1+ •» 1+ 2+ 2+ 1+ + 2+ 1+ 1+ Heated colostral wheyk rhey and NaSD >n of whey, 1: 160 40 40 80 2+ 4+ - 4+ 4+ 4+ 4+ - 2+ - 1+ 2+ 2+ 3+ 1+ 3+ 4+ 4+ 4+ ± 4+ 4+ 4+ 4+ - 4+ ± 2+ 4+ 4+ - 4+ 4+ 4+ .ution (5 ml.). 80 160 20 40 4+ 80 160 - 3+ 0.25 1+ 0.1 2+ 1+ - 2+ 3+ 4+ 1+ 3+ 4+ 1+ 4+ 4+ 4+ 4+ 1+ 2+ 2+ 4+ 4+ 2+ 4+ 4+ 4+ 4+ 1+ 4+ 4+ 4+ 4+ 3+ 2+ 3+ 4+ 4+ 4+ 4+ — — 3+ 4+ 4+ 3+ 4+ Incubation period, 72 hours. 2+ Bacterial controls, ites. lilution of serum from a brucella-infected cow in 5 ml. of culture medium 0.2 ml. of fresh, normal rabbit serum, and 0.01 mg. sodium sulfadiazine. ; serum added to each dilution, le added to each dilution. } of growth. The colostrums thus possessed a definitely lower growth-inhibiting activity in the presence of sodium sulfadiazine and complement than did blood serums, which had an average titer of 1:80. In contrast to the results with this normal growth-inhibitor, it is a well established fact that the specific antibodies of colostrum generally exceed their concen­ tration in the blood serum. 2. A colostral growth-inhibiting action independent of complement In the course of demonstrating the effect of added fresh, normal rabbit serum alone and sodium sulfadiazine alone on colostral growthinhibiting activity, an apparent departure from the classical bactericidal behavior of serum was noted. The colostrum alone exerted no activity under the conditions of this test nor did the colostrum plus fresh, normal rabbit serum as a source of complement. However, colostrum in the presence of sodium sulfadiazine did evidence activity. An examination of table 8 discloses that two wheys inhibited growth at a 1:20 dilution and four at a 1:10 dilution while one was capable of retarding growth and one showed no activity at a 1:10 dilution when tested with sodium sulfadiazine but without fresh, normal rabbit serum. Since milk had been reported devoid of complement, no activity was anticipated in fresh, colostral wheys without the presence of added com­ plement. Inasmuch as complement had reoently been detected in fluids previously considered lacking in this complex, the colostrums wore heated at ^6 ° C. for JO minutes to render any complement present inactive, and they were again tested for activity with sodium sulfadiazine. As may be observed in the data presented in table 8, the activity of samples 2 and 8 disappeared entirely on heating, but five samples, in the presenoe of sodium sulfadiazine alone, exhibited growth-inhibition although the -28- activity of four had been somewhat reduced. Therefore, some colostrums appeared capable of inhibiting growth without the presence of complement when sodium sulfadiazine was present. No analogous growth-inhibiting behavior was noted in blood serum. If eight colostrums may be considered representative, the lack of correlation between the complement-free growth-inhibiting action and the growth-inhibiting action in the presenoe of complement suggests two sepa­ rate growth-inhibiting entities in oolostrum neither of whioh is entirely destroyed at 56° C. for JO minutes. For example, the titer of unheated colostrums 5 and 6 remained unaltered by the addition of complement, seemingly indicating an antibody not requiring or enhanced by complement. On the other hand, sample 7 possessed no aotivity with sodium sulfadiazine alone but was found to be active in a dilution of I 1I4.O in the presenoe of complement and the salt. The growth-inhibiting activity of the remaining oolostrums was apparent without complement and enhanced by its presenoe, more so than by any additive effect demonstrated with rabbit serum. From an extensive study of the baoterioidal reactions of normal serum, Maokie and Finkelstein (16) oonoludod that the thermostable prin­ ciple of serum confined itself to activity only against the Gram-positive organisms whereas the thermolabile principle (so defined beoause of the complement requirement of the system) reacted against only the Gramnegative organisms. Previous workers had drawn attention to the existence of two different bacterioidal mechanisms in normal serum but did not make this sharp differentiation as to Gram reaction. Fleming (18) found that the lysozyme of body tissues and secretions lysed both Gram-positive and Gram-negative organisms. He was able to show some lysis of Br. abortus but not of Br. melitensis. No attempt has been made here to characterize -29- the growth-inhibiting principle in colostrum which acts without comple­ ment, but its presence is notable. -30- D. The Brucella Growth-Inhibiting Factor of Bovine Serum Fractions Prepared by Sodium Sulfate Precipitation No attempts to fractionate serum in order to study the natural bactericidal properties of the components appear to have been undertaken. While the isolation and purification of a single component from a system as complex as serum constitutes a major problem, it was felt that oertain rudimentary information as to the nature of the growth-inhibiting anti­ body might be obtained by an examination of serum fractions. Antibodies engendered by the injection of antigens have been shown to be associated with serum globulins and it is to be expected that nor­ mal antibodies might also prove to be of globulin nature. San Clemente and Huddleson (19) demonstrated by adsorption of specific antiserum that the brucella agglutinin of bovine serum was associated with the Y-globulin. Hess and Deutsch (20) also found specific brucella antibodies in bovine serum associated with theYx - and Y a-globulins with a predominance in Y xglobulin. The antibody activity of bovine anti serums against several other antigens was found in the T- and Y-components by Smith (8). That serum component migrating between theP- and Y-globulins has been variously termed T, P a and Y x. Limited data concerning the globulin nature of some natural anti­ bodies resulted from studies of natural agglutinins and complement-fixing antibodies (12), (21), (22), (23), (2I4). These authors examined the activity of fractions prepared by carbon dioxide precipitation of euglobulin from a dilute serum solution. In certain species the euglobulin alone contained the antibody while serum from other species yielded frac­ tions of equal and unequal activity. In this experiment a modification of the Howe (25) sodium sulfate method was employed to separate three fractions from serum. siderations led to modification of the Howe method. Several con­ Precipitation from serum in a dilution of 1j2 to 1*3, instead of Howe’s 1*20, was based on the observation of Cohn and his associates (26) that this allowed effec­ tive precipitation of serum fractions by ammonium sulfate. Furthermore, successful precipitation of specific brucella antibody by ammonium sul- ' fate had been effected at a serum dilution of 1*2 (19). It seemed advisa­ ble to preoipitate the globulin at room temperature, instead of 37° C., as the thermolability of the normal growth-inhibitor was apparent by that time. The salt concentrations established as optimum in the Howe method were not strictly applicable to these preparations in view of the varia­ tions used in dilution and temperature (25). It seemed possible that the natural growth-inhibitor might preoipitate with the same fraction as the specifio antibody and that a method which satisfactorily separated the more readily detected specific antibody might serve to fractionate the natural one. On the other hand, failure of the natural and specific antibodies to precipitate under the same conditions would indioate dissimilar properties of the two. With this goal, the salting-out of serum from a brucella-infected cow was studied, and the fractions obtained from normal and specific serums were examined and com­ pared. 1. Fractionation of specific antiserum a. Preparation of fractions. -- Serum 991> which possessed unusually high growth-inhibition and agglutination titers, was obtained from a brucella-infected cow. This serum was separated into three fractions. One volume of serum was diluted with one volume of 25-percent sodium sulfate solution (at 37° C.) and allowed to stand at room temperature overnight. The resultant precipitate, termed the 12.5-peroent fraction, was centrifuged, washed once with 12.5-porcent sodium sulfate solution, and made up to one-half the original volume with water. Following dialysis for 3 days against 0 .85-peroent sodium chloride solution at J4.0 C., the solution was sterilized by filtration through a Hormann D8 pad in a Seitz filter and stored at I40 C. After removal of the 12.5- percent precipitate, sufficient dry sodium sulfate to bring the con­ centration of the supernatant to 15 peroent was added slowly with me­ chanical stirring. After standing at room temperature overnight, the preoipitate was washed once with 15-percent sodium sulfate solution, made up to one-third the original volume, dialyzed, and sterilized. ond fraction was designated as the 15-percent one. The seo- The supernatant obtained from preoipitation at 15-peroent salt concentration was dialyzed against running tap water for I4.O hours, concentrated to original volume by evaporation for 10 hours at 37° C., dialyzed against 0 .85-percent sodium chloride solution at I40 C. for an additional 2 days, and sterilized. After restoring the concentrated fractions to their original serum volume, the growth-inhibiting activity was determined in the customary manner. Nitrogen was determined by the Kjeldahl method. When analyses revealed appreciable aotivity both in the 12.5- and 15-percent fractions, a single preoipitation of serum 991 was made at 15-percent salt concentration to determine whether it would effect as complete a separation of the specifio antibody as the two separate pre­ cipitations. A second single precipitation of this serum wa 3 made under slightly different conditions. It was thought that precipitation in the presence of normal serum woula provide a more normal albumin-globulin ratio and approximate the conditions encountered in normal serum fractionation. One part of serum 991 was diluted with two parts of nor­ mal "bovine serum and precipitated at 15-percent concentration of sodium sulfate. Otherwise the procedure was as previously outlined. As deter­ mined from the growth-inhibiting activity, in both instances the same percent of total brucella growth-inhibiting antibody was precipitated by one preoipitation as by the sum of the two separate precipitations. b. Growth-inhibiting action. — Serum 991 and its 12.5-percent fraction (a) inhibited growth in a dilution of li512,000 while the 15percent fraction (B) and supernatant from the 15-pcroent preoipitate (c) inhibited growth in dilutions of 1*61;, 000 and 1 i 2,560 respectively when in the presence of sodium sulfadiazine. On the basis of information con­ cerning the (3- and y-glotulin character of specific antibodies, the activity of this serum and its fractions might have been referred to these components as determined electrophoretically. However, sinoe the primary purpose was one of comparison with the activity obtained in normal serum fractions and since the normal growth-inhibitor remains to be identified with particular serum protein components, reference of activity to the total protein seemed preferable. Therefore, the relative effectiveness of serum and its fraotions was expressed as the minimum amount of total protein necessary for demonstrable action. _7 1.8 x 10 _7 , 0.8 x 10 It required a minimum of _ _rj , 1.1 x 10 , and 165 x 10 7 gm. of total protein in the whole serum and fractions A, B, and C respectively to evidence growth-inhibiting action. Fractions A, B, and C contained 512,000, 614,000 and 2,560 units respectively capable of growth-inhibiting action or 88.5, 11.0, and 0.5 percent of the antibody. Thus over 99 percent of the specific growth-inhibitor in this serum was precipitated at 15percent sodium sulfate concentration, 89 percent of which precipitated out at 12.5-percent sodium sulfate concentration. c. Electrophoretio analyses. — Electrophoretic analyses of serum 991 and its three fractions were made by Dr. R. E. Sanders from patterns obtained in the Tiselius apparatus by the Longsworth scanning method. Electrophoresis was conducted at 0.5° C. in a barbital buffer of 0.2 ionio strength and pH- 8.6 at a potential gradient of 3»5 volts per cm. The time was 20,000 seconds. The protein concentration was 1.5 percent. Certain data from the ascending patterns are presented in table 9» A correlation was noted between the Y"gl°6ulin content and the activity of fractions A and C. The lack of resolution of the globulin components in the eleotrophoretic patterns of fraction B and of the whole serum did not permit calculation of the y-globulin content. The 12.5- percent fraction (A) consisted of almost pure y-globulin as indioated by the symmetry of the single peak; fraction C contained some of each of the components present in the original serum. Since 89 percent of the growth-inhibiting activity was associated with fraction A, or y-globulin, the specifio growth-inhibiting factor would appear to migrate with the y-globulin. Fraction A possessed O.Q^lij. gm. of y-globulin and 512,000 growth-inhibiting units per ml. as compared to O.OOllj. gm. of y-globulin and 2,560 growth-inhibiting units per ml. in fraction C. 0.8 x 10 and 0.6 x 10 A minimum of gm. of y-globulin in fractions A and C respec­ tively produced demonstrable growth-inhibiting action. Hess and Deutsch (20) have recently reported the specifio bovine bactericidin against Brucella to be in Y x“ and y a-globulin fractions prepared by the ethanol method. 2. Fractionation of normal serum a. Preparation of fractions. — By analogy with the specific serum -35- Table 9 Growth-Inhibiting Action and Electrophoretic Analyses of Serum and Its Pi Electrophi Substance Serum 991 Proteina gm./lOO ml. Percent of original protein Growth inhi­ bition titera 1: Protein per unit growth inhi­ bition gm. 9.28 100 5.12x10® 1.8xl0-7 Relative c< Perce] A 33.6 a 12.6 Fraction A 4.14 45 5.12x10® 0.8x10 7 0 Fraction B 0.73 8 6.40x10* l.lxio”7 1.7 10.3 Fraction C 4.23 46 2.56x10® 165xl0-7 64.1 16.8 0 These electrophoretic analyses were made from photographs takei alter apparatus at 0.5° C. in barbital buffer of pH 8.6 at an ionic strength o sample was 1.5 percent. A = fraction precipitated at 12.5 percent sodium sulfate conceitratio: B = fraction precipitated at 15 percent sodium sulfate concentration C = supernatant of fraction B. aFractions diluted to original serum volume. Table 9 itic Analyses of Serum and Its Fractions from a Brucella-Infected Cow Electrophoretic analyses, ascending pattern rrowth ,nhitition ;itera 1: Protein per unit growth inhi­ bition gm. 12x10® 1.8xl0-7 12x10® 0.8x10 7 0 ,40x10* l.lxio”7 1.7 10.3 56x10® 165x10“7 64.1 16.8 Relative concentration Protein Percent area gm./lOO ml. A 33.6 a (3 y 53.8 — > 12.6 100.0 0 <-88.1-> 13.7 3.4 A a 3.12 1.17 0 P <-4.99 -> 4.14 0 0.01 0.08 2.72 0.80 Y <-0. 64— » 0.58 0.14 ide from photographs takei alter 20,000 seconds in the Tiselius )f pH 8.6 at an ionic strength of 0.2. The protein content of the sent sodium sulfate conceitration. it sodium sulfate concentration after removal of A. rolume. it was anticipated that the growth-inhibitor of normal serum would be precipitated at 15-peroent sodium sulfate concentration. Preliminary precipitation of a normal bovine serum at this salt concentration revealed that its activity lay in the supernatant fraction, oontrary to the results obtained with the specifio serum. Upon reprecipitation of the supernatant at 20-percent salt concentration, the activity was found to lie in the fraction preoipitated at 20 peroent. No activity was apparent in the supernatant from the 20-percent preoipitate. Subsequently, six normal bovine serums were preoipitated at 15percent and, after removal of the preoipitate, reprecipitated at 20percent concentration of sodium sulfate. Serums 1, 2, and 3 were taken from the same animal at the fourth and fifth months of pregnancy and i; months after parturition; serums i+, 5» and 6 were from different animals. The normal serum fractions were prepared as were the specific ones with the following exceptions* the serums were processed within 2 k hours of bleeding; an interval of 1 hour was allowed between the addition of sodium sulfate and removal of the precipitate; 30-psrcent sodium sulfate solution was used to precipitate the 15-peroent fraction and to raise the salt concentration from 15 percent to 20 peroent. Concentration of the supernatant from the 20-percent precipitate by evaporation at 37° C. required 10 to 2 k hours depending on the degree of concentration attained. b. Growth-inhibiting action. — The results of the growth-inhibition determinations conduoted immediately after processing the normal serum fractions are presented in table 10. In order to approximate any loss of activity in the fractions due to filtration, the whole serums had also been filtered. All of the 15-peroent fractions in dilutions of 1*10 exhibited partial or oomplete growth-inhibition and those of serums Table 10 Growth-Inhibiting Action of Normal Bovine Ssrum and I Dil. of serum, 1: 10 20 40 80 160 10 20 40 80 (%) ] 28.5 1+ 4+ 4+ 4+ - ] 2+ 3+ 4+ 4+ 10.3 1+ 1+ 4+ 4+ 10.0 2 7.02 1 + 26.9 3 8.46 - - - - 4+ 30.4 4 8.47 - ±2+4+ 46.4 2+ 4+ 4+ 5+ 9.4 5 7.62 - - 35.0 1+ 4+ 4+ 4+ S.l 6 8.49 -3+4+ + + i i 7.34 i 1 1 1+ 1+ 1+ 4+ 4+ Original protein H- Protein gm./lOO ml. Dil. of A, 1: Original protein (*) H- Serum No. Fracti Fraction Aa Serum, whole 45.2 - -2+4+ 6.5 Di i t Serum3 1, 2, and 3 were obtained from the same cow at the 4th and 5th r after parturition. Serums 4, 5, and 6 were from different cows. Each dilution in 5 ml. of culture medium also contains 108 Br. abortus serum, and 0.01 mg. 3odium sulfadiazine. Incubation period, 72 lours. B* A = fraction precipitated at 15-percent sodium sulfate concentration. B = fraction precipitated at 20-percent sodium sulfate concentration a: C = supernatant of fraction B. aFraction diluted to original serum volume. - = no visible growth. + = degree of growth, turbidity. Table 10 ion of Normal Bovine Sarum and Its Fractions Fraction Ba Fraction A a Fraction Ca Dll. of A, 1: Original protein Dil. of B, 1: Original protein Dil. of C, 1: %) 10 20 40 80 (70 10 20 40 80 (70 10 20 40 80 8.5 1+ 4+ 4+ 4+ - - 4+ 4+ 53.3 4+ 4+ 4+ 6.9 2+ 3+ 4+ 4+ 10.3 mm - 4+ 4+ 58.0 4+ 4+ 4+ 0.4 1+ 1+ 4+ 4+ 10.0 - - 2+ 4+ 53.7 4+ 4+ 4+ 6.4 2+ 4+ 4+ 5+ 9.4 4+ 4+ 4+ 4+ 32.5 3+ 3+ 4+ 5.0 1+ 4+ 4+ 4+ S.l - 3+ 4+ 42.5 mm 3+ 4+ 2+ 4+ 6.5 4+ 4+ 4+ 4+ 42.2 5+ 5+ 5+ Inal tein 5.2 - - 1+ - he same cow at the 4th and 5th month of pregnancy and 4 months e from different cows. an also contains 108 Br. abortus. 0.2 ml. of fresh, normal rabbit Incubation period, 72 lours. Bacterial controls, 4+ growth. ; sodium sulfate concentration. . sodium sulfate concentration after removal of A. ,ume. ;rowth, turbidity. i 3 and 6 were slightly active in dilutions of 1*20 and 1 *1+0 respectively. Of the 20-percent fractions, four completely inhibited growth in a 1*20 dilution hut two showed no activity. One 20-peroent supernatant, No. 5» inhibited growth at a 1*10 dilution and evidenced a very slight, or no, aoticn at 1*20 as did supernatant 1+ in dilutions of 1*10 and 1*20. other supernatants were inactive. tions at 3 The Tests on the 15- and 20-percent frac­ or 1+ times the original oonoentration confirmed the degree of activity in table 10 and failed to reveal any in the 20-peroent frac­ tion of serum I4. although that of serum 6 showed a slight activity. The relative effectiveness of normal serum and its fractions, expressed as the minimum of total protein neoessary for demonstrable action, was 0.1-0.2 gm. in the whole serum, 0 .2-0 .1+ gm. in the 15-peroent fraotion, 0 .03-0.01+ gm. in the 20-percent fraction, and 0.3 gm* in the 20-percent supernatant. These are approximations* for example, only partial inhibition of growth was shown by the 1*10 dilutions of some 15-peroent fractions but a titer of 1*10 was assumed in the above cal­ culations. No attempt was made to estimate the percent of total aotivity preoipitated with each fraction of normal serum. The small differences in activities did not warrant an evaluation of purification on this basis. However, since the protein necessary for demonstrable activity by the 20-peroent fraction of four out of six serums was one-third to one-fifth that of the whole serum and other fractions, a certain degree of puri­ fication had been effected in these oases. If one considers both the percent of original protein and growthinhibiting activity of fractions from serums 1, 2, and 3 (from the same animal), the salting-out procedure appears to have given fairly repro­ ducible fractions. The protein content of the 15-peroent fraction from different animals varied from 27 to U6 percent of the original protein. The difference is not surprising in view of reports of individual varia­ tions in the normal bovine albumin component alone of values frcm 35 b 0 53 percent (27). However, it may be noted that the two serums, Nos. /+ and 6, which showed no activity in the 20-peroent fraction were those from which U 5 bo I(.6 percent of the original protein precipitated with the 15-percent fraction. The significance of this is not clear. The protein content of the 20-percent fractions varied from 6 to 10 peroent of the original protein. The normal brucella growth-inhibiting factor was irregularly dis­ tributed over the serum fractions obtained by precipitation at 15-percent and reprecipitatian at 20-percent concentrations of sodium sulfate. distribution varied with serum from different animals. The All the demon­ strable growth-inhibiting factor of serums from two animals precipitated at 15-percent sodium sulfate concentration and was associated with i;5 to J4.6 peroent of the original protein. The major part of the factor in serum from three out of five animals precipitated with the 20-percent fraction whioh contained 10 percent, or less, of the original protein. 0 . Electrophoretio analyses. — Identification of the major part of the activity with a fraction low in protein led to the electrophoretic examination of serum 3 and its fractions. Analyses were made by Dr. R. E. Sanders from patterns obtained in the Tiselius apparatus. Elec­ trophoresis was oonducted under the same conditions as for the specific serum. Certain data from the ascending pattern are presented in table 11. Resolution of the boundaries on the ascending side permitted calculation of an additional component in one fraction; in other respects the asoending and descending patterns were in close agreement. Table 11 Growth-Inhibiting Action and Electrophoretic Analyses of a formal Bov Ele ctropbore ti c Substance Protein8gm./lOO ml. Percent of original protein Serum 3 8.46 Fraction A 2.57 30 Fraction B 0.85 1 0 Fraction C 4.54 54 Growthinhibition titer*3 1 : Protein per unit grow thinhibition gm. 80 1 0 0 < 2 0 2 0 None > Relative corcentrat Percent area A a Pi 8.4 0 . 1 1 51.2 0 . 1 2 4.8 9.9 17.8 42.9 8.4 78.1 13.5 8.5 0.04 1 2 . 0 P2 <-2i 0.12 CO . 8.4 0.11 CO CO Protein per unit grow thinhibition gm. 0.04 17.8 42.9 8.4 78.1 13.5 8.5 • 1 2 . 0 1 2 . 8 1.02 0.71 <— 2.40 -* <--- 2 . 1 9 ---- * 0.15 from photographs take l after 20,000 seconds in the Tiselius pH 8 . 6 at an ionic strsngth of 0.2. The protein content of the sodium sulfate concentration. sodium sulfate concentration after removal of A. ume. 0.11 Examination of the data reveals that the inactive 20-percent super­ natant fraction (C) contained a higher concentration of albumin and P^globulin than the active 20-percent fraction (B). and a- Fraction C con­ tained 3.55, 0.6l, and 0.39 gm. per 100 ml. of albumin, a-,and Pj-globulin respectively as compared to 0.15, O.36, and 0.07 gra* Per 100 components in fraction B. possessed neither active fraction B. these On the other hand, the inaotive fraction C nor y- globulins both of which were present in the Assuming no qualitative differences in the common components, one is lead to infer that the growth-inhibitor may migrate with the P 2- or Y”gl°bulins wr.ich were absent in the inaotive and present in the active fractions. As the broad y'globulin peak of the whole serum and that of the 15-percent fraotion did not permit estimation of the individual P2- and Y-areas, no oaloulaticn of the percent precipitated in the 20-peroent fraction was possible. appear to be a function of However, the activity does not total y-globulin as the 20-percent fraction was slightly more active than the 15-peroent fraotion which contained more than 90 peroent of the total Y~gl°bulin. 3. Comparison of specific and normal serum fractions The failure of the normal growth-inhibiting factor of different serums to precipitate uniformly prevented acomparison of the salt solu­ bility of the specific and normal factor as originally intended. It would appear, from the one specific serum examined, that 99 percent of the growth-inhibitor was preoipitated at 15-percent sodium sulfate con­ centration, whereas the major part of the activity in three out of five normal serums was not precipitated at this salt concentration. The low growth-inhibition titer encountered in normal serums does not permit calculation of the amount present when the factor appears in more than - I |X- one fraction. Additional investigation might establish a solubility differential of the factors. The differences observed in serums from only five normal animals indioate that this would necessitate a con­ siderable number of determinations for significant data and entail a study in itself. It has been assumed that the antibody of speoifio serums would precipitate in a uniform manner, a conclusion whioh pos­ sibly is not justified although the antibacterial antibodies have gener­ ally been found to be insoluble at 15-percent sodium sulfate concentra­ tion. The data in tables 9 and 11 show that the concentration of serum oomponent3 in the normal and specifio serums was vastly different; the former oontained I4..3 gm. of albumin and 1+.1 gm. of globulin per 100 ml. and the latter 3»1 gm. of albumin end 6.2 gm. of globulin per 100 ml. Different amounts of protein migfrt be expected to precipitate from these two serums at identical salt concentrations. Nevertheless, the wide range in concentration of normal bovine serum components (27) does not permit a comparison of fraotions from relatively few serums on this basis. The electrophoretic and growth-inhibiting data in tables 9 and 11 indicate that the specific antibody is a function of total y-globulin while the normal growth-inhibiting factor is not. The 12.5-percent fraction of specific serum 991 with which 89 peroent of the activity was associated comprised more than 80 percent of the total y-globulin. The major part of the activity of normal serum 3 was present in the 20-percent fraotion which contained 5 percent of the total yglobulin. Since development of the fractionation procedure resulted in frac­ tions of specific and normal serum preoipitated at different limits, the electrophoretic patterns were not strictly oomparahle. cided to compare identical fractions. It was de­ A comparison of fractions pre­ cipitated at, or below, 15-percent salt concentration seemed of limited value as the P- and Y“gl°l>ulin boundaries of these speoific serum frac­ tions had not been resolved. Therefore, a fraction of specifio serum 991 was precipitated at 20-peroent sodium sulfate concentration after removal of the material precipitated at 15-peroent and compared with the corresponding fraction of normal serum 3. Electrophoresis was oonduoted under the same conditions as previously stated. The data presented in table 12 show that 0.85 gm. and O .38 gm. of protein or 10 percent and i* peroent of the original protein preoipitated from serums 3 and 991 respectively. As may be noted in table 10, from 6 to 10 percent of the original protein preoipitated from five fresh, normal serums under these same conditions. One year had elapsed between the preparation of the original fractions from serum 991 and this par­ ticular fraction; during storage of the whole serum a small amount of material precipitated and the protein fell from 9*28 percent to 8.95 percent. In view of the variation in protein content of the 20-peroent fraction from normal serums and in view of the loss of undefined protein during storage of serum 991» ‘th® significance of the low protein content of the speoific serum is uncertain. No correlation existed between the absolute concentration of serum components and the growth-inhibiting activity. The normal serum fraction was approximately twice as concen­ trated in each oomponent as the specific serum fraction although it con­ tained only 20 growth-inhibiting units per ml. as oompared with 2,560 units per ml. in the specific serum fraction. The albumin-globulin ratios of the whole serums were 0.5 and 1.0. Table 12 Growth-Inhibiting Action and Electrophoretic Analyses of from a Normal Cow and from a Brucella-Infecte Electrophore Serum No. 3 991 Serum source Protein^ gm./lOO ml. N ormal cow 0.85 Infected cow 0.38 Percent of original protein 1 0 4 Growthinhibition titer 1 : 2 0 2560 Relative conce Percent a A a 17.8 42.9 2c 17.0 39.4 3( 1The fraction of serum insoluble in 20 percent sodium sulfat 9 but soli aFraction diluted to original serum volume. Table 12 n d Electrophoretic Analyses of a Serum Fraction1 Cow and from a Brucella-Infected Cow Electrophoretic analyses, ascending pattern LU ial Ln Growthinhibition titer Is. 2 0 2560 Relative concentration Protein Percent area gm./lOO ml. A a (3 Y A a p Y 17.8 42.9 26.5 1 2 . 8 0.15 0.36 0.22 0 . 1 1 17.0 39.4 30.9 12.7 0.06 0.15 0.12 0.05 percent sodium sulfat 9 but soluble in 15 percent sodium sulfate lume. Nevertheless, the relative concentrations of components in the 20-percent fraction from the normal and specific serum were in very good agreement. That from normal serum 3 contained 17.8 percent albumin and ij2.9, 26.5, and 12.8 percent of a-, (3-, and yglobulin respectively. That from spe­ cific serum 991 contained 17.0 percent albumin and 39»U» 30.9»and 12.7 percent of X) O o o o• CO. CM• • CM. • • • CO rH to cO rH CJ iH (0 o I —i X CO to . CM «0 o 1 —i X CM i —1 • in Bacterial control + 4+ 4+ 4+ 3+ 3+ 4+ 4+ 0.2 4+ 3+ 2+ 4+ 4+ IV, unheated 1.0 ± IV, heated^ 30 min. 1.0 II, unheated 0.2 4+ II, heated * 5 30 min. 0.2 II, heated *5 150 min. 2 - + C - 2 - 1+ 4+ 4+ 4+ — 3+ 4+ 4+ 4+ a105 Br. abortus, 0.2 ml. fresh, normal rabbit serum, and 0.01 mg. sodium sulfa­ diazine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. ^Fraction heated at 56° C. C = contamination. - = no visible growth. + = degree of growth. 150 minutes failed to alter the growth-inhibition titer or the prozone reaction of the plasma fraction. Data on the effect of heat on the activity of fraction IV are included in table 16. It was not possible to obtain a clear 1-peroent solution of fraction IV in 0.85-porcent sodium ohloride solution. The oloudy suspension, therefore, was filtered through a Hormann D8 pad in a Seitz filter which resulted in a clear filtrate of undetermined con­ centration. This sample failed to show agglutinins in a dilution of 1»20 or to inhibit growth in dilutions greater than 1*80. Although the growth-inhibition titer of fraction IV was the same as that of the average normal serum, heating at $6* C. for 30 minutes did not signifi­ cantly reduce its activity. The growth-inhibiting factor of bovine plasma fractions II and IV thus resembled that of specifio antiserum in its thermostability, d. Normal colostrum, — Comparative data on the growth-inhibiting activity of heated and unheated colostrum in the presenoe of complement and sodium sulfadiazine are set forth in table 8 , The results show that heating produced little effect on the activity of colostrums 1, 3> U, and 5, slightly reduced the activity of oolostruros 2, 6, and 8, and markedly lowered that of colostrum 7» Colostrums 1, 2, lj, said 7 evi­ denced a slight prozone reaction after heating. Heating produced a similar varied reduotion in the potency of the oolostral principle oapable of growth-inhibition in the presence of sodium sulfadiazine without the presenoe of complement (table 8 ), A consideration of the pronounoed difference in activity between heated and unheated normal serums, in the presenoe of sodium sulfadia­ zine and complement, and the comparatively slight effect upon exposing colostrum to the same conditions indicates a difference in thermosta­ bility between the normal brucella growth-inhibitor of serum and that of colostrum. 2. The effect of storage a. Normal serum — In view of the marked thermolability of the normal growth-inhibitor of serum, the effect of storage cn its activity was investigated. Irwin et al. (13) found the bactericidal activity of bovine serums stored at 3* C., as determined without sulfonamides and without added complement, to decrease until at 28 days little or none was apparent. complement. The activity was restored by the addition of bovine serum Likewise, Huddleson et al. (3) reported a loss in bac­ tericidal action in 30 days, presumably due to deterioration of the complement. A survey of the literature revealed no investigation of the effeot of extended storage on the natural antibody independent of com­ plement . Seven normal bovine serums were stored at ij.* C., and the growthinhibiting action, in the presence of sodium sulfadiazine and added complement, was estimated at various intervals. Serums 2| and 5 were from the same cow at the fourth and fifth months of pregnancy; the other serums were from different animals. The data in table 17 illus­ trate the effect of storage on the natural brucella growth-inhibiting factor. Serums 1 and 2 failed to inhibit growth significantly in dilu­ tions of 1*10 or greater after storage for 9 and 8 weeks respectively. At 10 weeks serum 3 WQs active in a dilution of 1*20 as compared with an original activity in a dilution of liUO. When tested at 20 weeks, no growth-inhibition was evidenced by serum 3 although it exhibited partial inhibition in dilutions of 1*10 and 1*20 at 22 weeks. Serum 1+ Table 17 Effect of Storage on the Growth-Inhibiting Action of Normal Bovine Serum Degree of gr c m th, turb id i ty a Serum No. Storage at 4°c. (weeks) Dilution of serum, 1 0 2 0 + 4+ 4+ 4+ 4+ C 4+ 4+ _ _ _ - - 4+ 2 + 4+ 1 + 3+ 4+ 4+ 4+ 4+ 4+ 4+ _ mm _ - 4+ 4+ 4+ 3+ 5+ 4+ 3+ 4+ 4+ 4+ 4+ 4+ 4+ + 3+ 4+ 3+ 4+ 4+ _ - - - ± ± - - - 4+ 4+ 4+ 4+ 4+ 4+ + 4+ 4+ _ _ _ 4+ 4+ 16 ± - 3+ 17 2+ 3+ - 4+ 4+ 5+ 4+ 4+ 4+ 4+ 4+ 8 1 0 2 0 2 2 3+ 4+ 4+ 1 ± - 2 2 4+ 4+ 0 8 18 2 0 0 6 2 0 1 7 1 9 13 4 rj 4+ 4+ 4+ + 3+ 4+ 4+ 0 6 4+ 5+ 4+ • 0 5 mm Bacterial control 4+ 4+ 4+ 2 1 4 160 4+ 4+ 1 2 % 80 4+ 3+ 9 23 0 o 40 : 4+ 4+ 1 1 1 28 + 3+ 1 + 3+ 4+ 1 1 2 + + 2 + 2 - 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4 a10 8 Br. abortus. 0.2 ml. of fresh, normal rabbit serum, and 0 . 0 1 mg. sodium sulfadiazine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. - - no visible growth. + = degree of growth. C = contamination. exhibited activity in a dilution of 1:20 at 13 -weeks as compared with original activity in a IjIj.0 dilution; this serum failed to inhibit growth in a dilution of 1:10 or greater after 22 weeks. Serum 5 showed reduced activity during storage with evidence of the original prozone reaction throughout the time; no significant activity remained at 20 weeks. Serums 6 and 7 were active in a dilution of 1:20 at 20 and 28 weeks respectively although the latter showed prozone growth in a dilu­ tion of 1:10 and the original activity of both serums was reduced. It must be stressed that the growth-inhibiting activities deter­ mined at different times are not strictly oomparable. Analyses were necessarily conducted with different rabbit serums as a source of com­ plement and different bacterial suspensions. The discrepancy in the growth-inhibiting activity of serum 3 at 20 and 22 weeks is obvious. The culture used for the determination at 20 weeks was suspected of dissociation. A second determination at 22 weeks was made with a cul­ ture derived from a selected colony. Serum 5 also presented a pioture of less activity after 1; weeks than after 8 weeks of storage. These irregularities emphasize the fact that interpretation of the effect of storage on growth-inhibiting activity should be based on general trends and not on absolute values. The foregoing results show that some normal bovine serums lost their growth-inhibiting activity, as determined in the presence of sodium sulfadiazine and added complement, within 2 months and others within 5 months of storage. Some serums remained'active after 5 and 7 months of storage although the growth-inhibition titer was reduced. b. Normal serum fractions. — The effect of storage for 5 months at i|° C. on the growth-inhibiting activity of normal serums and their fractions was determined. The activities of one representative serum and its fractions are recorded in table 18. Data were obtained on samples of serum fractions diluted to original serum volume and also on concentrated samples so that any residual aotivity might be apparent. The growth-inhibition titer of the whole serum declined from 1*80 to 1*20 in 8 weeks, and that of fraction B declined from 1*20 to 1*10; the activity of fraction A remained essentially unohanged. However, after 5 months of storage the fractions failed to exhibit activity although a very slight zone action was evident in the whole serum. Storage at li® C. affected three other serums and their fractions in a similar manner. At 3 months a diminished activity still remained in the fractions but at 5 months this had disappeared. Two of the three whole serums were still slightly active at 5 months. o. Normal colostrum. — Table 19 illustrates the effect of storage at different temperatures on the growth-inhibiting activity of a repre­ sentative colostral whey in the presence of sodium sulfadiazine. wheys were stored at room temperature, Five C., and -10° C. and examined for growth-inhibiting activity after 6, 11, and 22 weeks. After 6 weeks at room temperature some reduction in the growth-inhibiting activity of three of the wheys was apparent. A definite loss in activity was evident in four of the oolostrums held at room temperature for 11 weeks; the oolostrum which had the highest original titer retained its activity. Unfortunately, the appearance of mold in the samples held at room temperature prevented analyses at 22 weeks. Although slight deviations from the original activity oocurred in samples stored for 22 weeks at h° C. or -10° C., they did not exoeed the variables of the test. It may be concluded that the bruoella growth-inhibiting aotivity of colostrum, Table 18 Effect of Storage on the Growth-Inhibiting Action of a Normal Bovine Serum and Its Fractions Degree of growth, turbidity8Serum, whole Cone.,3 times Cone., orig. Cone.,4 times Dil. of serum, 1: Dil. of A, 1: Dil. of Dil. of 3, 1: Dil. of 3, 10 20 40 80 160 10 20 40 80 10 20 40 80 2+ 3+ 4+ 4+ 2+ 1+ 4+ 4+ 2+ 4+ 4+ 4+ 2+ 2 1+ — 3+ 4+ 5 4+ 3+ 4+ 4+ + t i -H -H 0 Fraction B Cone., orig. Cone., orig. Stor­ age'3 (mo.) Fraction A 4+ a , 1: - 2+ 4+ 4 + 4 + 4 + 10 20 40 80 - -4+4+ - 3+ 4+ 4+ 1 : 10 20 40 80 160 - 4+ - - - 4+ - - 3+ 4+ 4+ 4+ 4+ A = fraction precipitated at 15-percent sodium sulfate concentration. B - fraction precipitated at 20-percent sodium sulfate concentration after removal of A. a106 Br. abortus, 0.2 ml. of fresh, normal rabbit serum, and 0.01 mg. sodium sulfa­ diazine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. Bacterial controls, 4+ growth. ^Stored at 4° c. - = no visible growth. + = degree of growth. Table 19 Effect of Storage at Different Temperatures on the Growth-Inhibiting Action of Colostral Whey in the Presence of Complement Degree of growth , turbidity0Temp., 4° C. Temp., room Storage time (weeks) Dil. of whey, 1 0 2 0 40 80 1 10 20 0 6 1 1 2 2 + 3+ 3+ 4+ 4+ 4+ 4+ 4+ 1 4+ 1+ 0 ° C. Dil. o f whey, Dil. o f whey, 1: : 160 Temp., -1 10 40 60 160 2+ 2+ 3+ 3+ 4+ 4+ 4+ 4+ - 4+ 5+ 1+ 20 40 80 3+ 4 — 4+ 4+ 1+ 3+ 4+ + 1 : 160 4+ a10° Br. abortus. 0.2 ml. of fresh, normal rabbit serum, and 0.01 mg. sodium sulfadiazine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. Bacterial controls, 4+ grov/th. - = no visible growth. + = degree of growth. -60- in the presence of sodium sulfadiazine and complement, was lowered by exposure to room temperature for several weeks but was not significantly reduced by storage at 4° C. or -10° C. for 5 months. It proved to be less affected by storage than the factor in serum or serum fractions, which parallels the effect of heat on their respective activities. Since growth-inhibition ocourred in colostrum without the presence of complement, the ability to inhibit Brucella in the presenoe of sodium sulfadiazine alone (without complement) was examined after storage at J4.0 C. for 5 months. Inspection of table 20 shows that, with the excep­ tion of whey 2 , the complement-independent growth-inhibiting factor retained a large proportion of its activity through 5 months of storage. 3 . The effect of filtration a. Normal serum. — Repeated experiments by Huddloson et al. (3) revealed that filtration through a Hormann pad brought about an increase in the bactericidal action of normal bovine plasma. In thi3 study filtration was found to reduoe the growth-inhibiting action of normal serum in the presence of sodium sulfadiazine. Samples of serum, or plasma, from 11 different animals were filtered through Hormann DO pads in Seitz filters and their growth-inhibiting activities were compared with those of the unfiltered speoimens. 1 through 9 were filtered once. Samples Samples 10 and 11 were subjected to seven successive filtrations through fresh filter pads each time. Three samples wore examined on the day of bleeding of the animal and eight after a short storage of the filtered and unfiltered samples under iden­ tical conditions. The results are recorded in table 21. It is apparent that in each case the filtered serums and plasma possessed less aotivity than the Table 20 Effect of Storage on the Colostral Growth-inhibitor Not Requiring Complement Degree of growth, turbiditya Presh Dil. of whey, 1: Colos trum No. Stored13, 5 mo. Dil. o f whey, 1: Bacterial control 10 20 40 80 10 20 40 80 1 - 2+ 4+ 5+ — 4+ 4+ 2 + 2+ 4+ 4+ 5+ 5+ 4+ 3 - - - 2 + 4+ 4 - 2+ 4+ - 4+ 4+ 5 1+ 4+ 4+ 4+ - 5+ 4+ 2+4+ a10e Br. abortus and 0.01 mg. sodium sulfadiazine added to each dilution and to controls (5 ml.). Incuba­ tion period, 72 hours. ^Stored at 4° C. - = no visible growth. + = degree of growth. Table 21 Effect of Filtration on the Growth-Inhibiting Action of Normal Bovine Serum and Plasma Degree of growth, turbidity8Dilution of sample, Sample No. 1 Sample Plasma Filtration^* None 1 0 2 0 40 mm - wm 1 1 2 3 Serum Serum None Serum 5 Serum 6 7 Serum Serum Serum - None 1 Serum Serum + None 1 None None - - «• 4+ - 4+ 4+ 4+ 4+ 4+ 4+ 4+ 4* + 1 1 - mm + + 1 7 4+ - 1 + 4+ 5+ 5+ + + 4+ 4+ 2+ 1+ 1 + Serum 4+ 4+ 2+ 4+ 3+ 4+ 5+ 5+ 5+ mm «■ 4+ 2+ 3+ 4+ 4+ 4+ 4+ 1+ 2+ 5+ 2+ 5+ 5+ + 1 7 3+ 5+ 3+ None 1 - None 11 4+ 4+ - 2+ 320 2+ - 2 : 3+ 4+ - 1 None - None 10 2 - 1 9 + 1 + 160 None 1 8 •• 1 1 4 80 1 2+ 2+ a10B Br. abortus. 0.2 ml. fresh, normal rabbit serum and 0 . 0 1 mg. sodium sulfadiazine added to each dilution and to controls (5 ml.). Incubation period, 72 hours. Bacterial controls, 4+ growth. ^Filtration through a Hormann D 8 pad in a Seitz filter. - = no visible growtn. + = degree of growth. unfiltered ones. Although the growth-inhibiting activities of plasma 1 and serums 5 and 6 were only slightly reduced by one filtration, the other samples required 2 to 1+ times as great a concentration of oncefiltered serum as of unfiltered serum to inhibit the growth of Br. abortus. While seven filtrations of serums 10 and 11 removed more growth-inhibitor than did one filtration, the activity was not reduced in proportion to the number of filtrations. It is possible that this indioates the presence of one antibody faotor whioh was removed by filtration and another which was relatively unaffected by filtration. b. Normal colostrum. — Investigation of the effect of filtration on the growth-inhibiting action of colostrum was limited to one sample and, as such, can only be presented as indicative. Following seven oonseoutive filtrations of the whey through a Hormann H9 pad in a Seitz filter, the aotivity of the filtrate was determined in the presenoe of sodium sulfadiazine alone and with complement. The growth-inhibiting substanoe oapable of action without comple­ ment showed, in the presence of sodium sulfadiazine, a titer of 1:20 both before and after filtration. Repeated filtration did not impair the activity of the growth-inhibiting substanoe oapable of action without complement. In the presence of complement and sodium sulfadiazine, the unfil­ tered colostrum displayed a titer of 1:1+0 as compared with only partial inhibition in eaoh dilution through 1:1+0 by the filtered colostral whey. The significance of these results is not d e a r since complete inhibition was evidenced in dilutions of 1:10 and 1:20 without the rabbit serum agent. It is conceivable that the added rabbit serum provided nutri­ tional enrichment thereby allowing growth in dilutions showing inhibition with sodium sulfadiazine alone. Additional data would be desirabl F. The Effect of Normal Bovine Serum and Colostrum on the Prozone GrowthInhibiting Action of Serum from a Brucella-Infected Cow The recognition of the prozone antibody phenomenon in the blood serum of brucella-infected animals and the ability of fresh, normal rabbit serum in the presence of sodium sulfadiazine to inhibit it in vitro and in vivo in guinea pigs (l) disolosed a potential therapeutio agent against brucellosis. It was shown that the maintainance of a bactericidal state in brucella-infected guinea pigs by the above agents for a sufficient length of time brought about a rapid and oomplete ter­ mination of the infection. The normal serum faotor responsible for prozone-inhibition in the presence of sodium sulfadiazine has not been identified. It was thought of interest to examine bovine serum and oolostrum for in vitro prozoneinhibiting aotivity and, if demonstrable, to determine its correlation with the growth-inhibiting aotivity. 1. Prozone-inhibiting action Repeated experiments had invariably shown serum 991 from a brucella- infected oow to be capable of growth-inhibition in high dilutions but not in relatively low ones (under 11/4.,000) when in the presence of sodium sulfadiazine. In order to facilitate determinations, a 1*100 dilution of this serum wa3 chosen as representative, and the prozone-inhibiting activity of normal serum, fractions from serum, and oolostrum was esti­ mated with a 1-tube test. The sample to be tested was added in the stated amounts to a 1*100 dilution of specifio serum 991 ln 5 ml. of liquid culture medium. The reagents of the sulfadiazine-antibody- complement test were added, and the degree of grov/th estimated at 72 hours. Growth in a IjIOO dilution of the specific serum had been shown to be 1++ or 5+ on innumerable occasions; therefore, less than 1++ growth in the presence of added substances indioated prozone-inhibition. a. Normal serum. — A preliminary examination demonstrated that normal bovine serum in a dilution of 1:10 was able to inhibit prozone growth when in the presence of sodium sulfadiazine. Additional infor­ mation was sought as to the minimum concentration of normal serum re­ quired to exhibit this action. The 1-tube prozone-inhibition test was performed with the addition of 0.5 ml., 0.25 ml., and 0.1 ml. amounts of normal serum which resulted in dilutions of approximately 1 :10, Ii20, and 1*50 respectively of the serum. The data set forth in table 22 show that five of the six fresh serums prevented the prozone growth when present in dilutions of 1:10. Serum 2 proved capable of inhibiting prozone growth in dilutions through 1:50# Serum 6 completely inhibited prozone growth in a dilution of 1:50 but not in dilutions of 1:10 or 1:20. Serums J>, I;, and 5 did not completely inhibit prozone growth when diluted more than 1:10. ■While the results showed fresh normal bovine serum to possess prozone-inhibiting action in dilutions of 1:10 or greater, this activity and the ability to inhibit growth per se oould not be examined for correlation. Unfortunately, an appreciation of the effect of filtra­ tion and storage on the growth-inhibiting action was lacking at the time this experiment was started, and each serum was not examined on the day of bleeding nor was a distinction made between once-filtered and un­ filtered speoimens. Although exactly comparable activities can not bo presented, certain faots relating to the identity of the two factors are apparent. Table 22 Effect of Normal Bovine Serum on the Prozone Growth-Inhibiting Action of Serum Prom a Brucella-Infected Cow N ormal bovine serum, ml. a Serum No. Age of serum** (days) Heated 0 Unheated 0.5 0.25 0 . 1 0.5 0.25 0 . 1 Bacterial control15 0 6 1 1 0 30 + 1 + 4+ 2 + 4+ 4+ 2 mm 1 + 5+ 4+ 4+ 4+ — 4+ 19 56 1 + 4+ 2 4 41 71 + 4+ 3+ 4+ 4+ 5+ 3 4 5+ 4+ 5+ 5+ 28 5 4 - 2 + 4+ 6 1 + 2 + - 2 2 6 2 5+ 2 + 3+ 4+ 4+ 5+ 4+ 5+ 5+ 5+ 5+ 3+ 3+ 3+ 5+ 4+ aNormal serum added to a 1:100 test dilution of serum from a brucella-infected cow in culture medium (5 ml.). Test dilution also contains 10B Br. abortus, 0.2 ml. of fresh, normal rabbit serum and 0 . 0 1 mg. sodium sulfadia­ zine. Incubation period, 72 hours. ^Control tubes of medium (5 ml.) contain a 1:100 dilu tion of serum from the brucella-infected cow, 10B Br. abortus, 0 . 2 ml. of fresh, normal rabbit serum, and 0 . 0 1 rag. sodium sulfadiazine. cSerum heated at 56° C. for 30 minutes. ^Stored at 4° C. - = no visible growth. + = degree of growth. -68- A relationship was noted "between a prozone effect in the growthinhibiting activity of normal serum prozone-inhibiting aotivity. table 13) 6 per se and a similar effeot in its Of the six serums, only serum 6 (animal 5> showed a prozone in its growth-inhibiting aotion, and it was the only serum to evidence complete prozone-inhibiting aotion when pre­ sent in a high dilution but not in a low dilution. On the other hand, three normal bovine serums, less than a week old and capable of growth-inhibition at dilutions of I 1/4O or greater during that time, failed to exhibit prozone-inhibition in a dilution of 1 »20. This indioates either a modification of the growth-inhibiting aotivity when acting in the presence of specific antibody or an independence of the two factors. b. Normal serum fractions. — The ability of fraotions from normal serum, in the presence of sodium sulfadiazine, to affeot the prozone reaction of serum from a brucella-infected cow is demonstrated by the data in table 23. This property was determined by the 1-tube method using the amounts of serum fractions indicated. The preparation of the fraotions is described elsewhere, and their growth-inhibiting activities are presented in table 10. Examination of the data reveals that none of the 15-peroent frao­ tions evidenced any inhibitory effect on the prozone growth, and all the 20-percent fraotions proved active in this respect. The 20-percent supernatant fraction of serum 5 inhibited prozone growth and that of serum I4 retarded but did not completely inhibit it. The 3+ growth in tubes containing 0.5 ml. of supernatants 1 and 2 might represent a slight activity but the significance is doubtful. The foregoing results show that certain serum fractions, in the - 69- Table 23 Effect of Normal Bovine Serum and Its Fractions on the Prozone Growth-Inhibiting Action of Serum From a Brucella-Infected Cow Serum No. 1 2 A 4 5 Substance Serum, whole Fraction Ad Fraction B^ Fraction C Serum, whole Fraction A c Fraction Bd Fraction Cb Age of serum (days) 5 N ormal bovine serum and fractions8. 0.5 ml. + 4+ 2 0.25 ml. 0 . 1 ml. + 4+ 4+ 5+ 1 + 4+ 5+ 5+ 3+ + + 4+ 6 + 4+ 4+ 2 - 3+ 6 5+ - 3+ Serum, whole Fraction A® Fraction Bd Fraction C^ 19 Serum, whole Fraction A c Fraction B^ Fraction Cb 4 + + ± + 2+ 5+ 3+ 5+ 1 6 2 2 6 - % 2 + 5+ 5+ 1 + 4+ A = fraction precipitated at 15-percent sodium sul­ fate concentration. B = fraction precipitated at 20-percent sodium sul­ fate concentration after removal of A. C = supernatant of fraction B. aNormal serum or fraction added to a 1:100 dilution of serum from a brucella-infected cow in 5 ml. of cul­ ture medium which also contains 10B Br. abortus. 0.2 ml. of fresh, normal rabbit serum and 0 . 0 1 mg. sodium sul­ fadiazine. Incubation period, 72 hours. Bacterial con­ trols as in Table 22. ^Concentrated to one-half the original serum volume. cConcentrated to one-third the original serum volume. ^Concentrated to one-quarter the original serum vol­ ume. - = no visible growth. + = degree of growth, tur­ bidity. presence of sodium sulfadiazine, were able to inhibit prozone growth. ■Whether any relationship exists between the growth-inhibiting activity per se and the prozone-inhibiting aotivity is not clear. The four 15- peroent fractions showed a low growth-inhibiting aotivity at dilutions of lilO and no effect against the prozone growth. Although three of the four 20-peroent fractions were capable of completely inhibiting growth in dilutions of 1*20, the fourth possessed no activity. all four samples caused prozone inhibition. However, Of the four supernatants, one had shown b definite and one a very slight growth-inhibition; only these two were effective against prozone growth. c. Normal colostrum. — Seven oolostral wheys were tested for prozone-inhibiting action by the 1-tube method. The results presented in the last column of table 8 demonstrate that colostrum resembles nor­ mal serum in its ability to prevent or retard prozone growth in the presence of sodium sulfadiazine. The seven oolostrums evidenced dif­ ferent degrees of prozone-inhibiting action which showed no apparent correlation with their independent growth-inhibiting aotion, in the presenoe of sodium sulfadiazine and complement, although both tests were done on identical colostral whey samples. 2. The effect of certain physioal forces on the prozone-inhibiting aotion of serum Sinoe inconclusive, or contradictory, results concerning the cor­ relation of growth-inhibiting and prozone-inhibiting activity had been obtained, the effect of physioal forces on the prozone-inhibiting activity was investigated. Heating at 56° C. for JO minutes, storage, and fil­ tration had been shown to reduoe the growth-inhibiting activity of nor­ mal bovine serum. The following data demonstrate a similar effect on the prozone-inhibiting aotivity of bovine serum. a. Effect of heat. — Three normal serums were heated at 5^° C. for 30 minutes and their effect on grcwth in the 1-tube prozone test ■was compared to that of unheated serum. Data in table 22 on serums 2, I4, and 6 show that the ability of normal serum to abolish prozone growth was destroyed, or reduced, at 5 6 ° C. for 30 minutes. Some residual activity remained in serum 2 after heating for while the lilO dilution of serum did not prevent growth it did retard it. To determine whether heating altered the prozone-inhibiting aotivity of fractions from serum as it did whole serum, two active 20-peroent fraotions were heated at 5&° C. for 30 minutes, and their pro zoneinhibiting aotivity was compared with that of the unheated fractions. As with whole serum, heating destroyed the ability of the two fractions, in the presenoe of sodium sulfadiazine, to function against the prozone phenomenon in serum from a brucella-infected cow. b. at Effeot of storage. — Four serums were stored at U° C. and tested intervals by the 1-tube method. The data in table 22 oonceming storage indioates that the prozone-inhibiting aotion disappeared within 1 or 2 months. Serums 1 and ij. showed no aotivity at I4.weeks. exhibited a reduced activity at 3 weeks and none at 8 weeks. Serum 2 Serum 3 remained somewhat aotive against the prozone reaction for 6 weeks but not for 10 weeks. Data, not presented in this report, on the effect of storage at U° C. on the ability of fractions from serum to prevent prozone growth indicate that this activity diminished with time and disappeared between 1 and 2 months. Serum fractions thus retained this property for approximately the same length of time as whole serum. -72- c. Effeot of filtration. — The reduction of the prozone reaotion effected by two fresh, normal serums before filtration and after pas­ sage through seven Hormann D8 pads was compared in the presence of sodium sulfadiazine. Serum from the brucella-infected cow 991 v/as diluted 1:10 and thence through four twofold dilutions. To eaoh tube were added 0.5 ml. of fresh or filtered normal serum and the growth-inhibition test reagents. After an incubation period of J2 hours the dilutions contain­ ing unfiltered serum showed reduced growth, represented by 1+ or 2+, as compared to a 5+ growth in the presenoe of serum filtered seven times. From these results it appears that the amount of substance capable of preventing prozone growth in serum from a bruoella-infected oow, in the presenoe of sodium sulfadiazine, was reduced by filtration. % DISCUSSION An examination of the in vitro growth-inhibiting activity of serum from calves after the ingestion of colostrum, as compared to that of adult animals, leads one to question the significance of normal growthinhibiting activity as an index of resistance. At this time, when any passive protection conferred by colostrum should be at its maximum, the growth-inhibiting aotivity of serum from the calf is less than that from the adult animal. Furthermore, if a relationship exists between resist­ ance and growth-inhibition, serum from the pregnant cow should exhibit a drop in activity for this is the period when the animal is most sus­ ceptible to brucellosis. However, data not presented in the report showed no significant variation in the growth-inhibition titer of serum from an animal taken at intervals before and during the pregnancy. An unidentified factor which is responsible for the resistance of the calf to brucellosis may well be transmitted by colostrum. A complement-independent principle capable of brucella growth-inhibition in the presence of sodium sulfadiazine has been detected in seven out of eight colostrums although a similar mechanism has not been shown in the blood of oalves or adult animals. It is we11 established that the concentration of specific antibodies in colostrum exceeds that of the maternal blood. Possibly the concentration of this principle in colostrum places it just within the range of our present means of de­ tection. If the substance were selectively absorbed from the intestine of the calf, as the specific antibodies have been shown to be, subse­ quent dilution in the blood would reduce it beyond the limits of our method although its concentration at that period might far exceed that of the adult animal. On the other hand, this factor may be peculiar to colostrum for the mammary gland synthesir.es proteins not found in serum and even imparts its ovm characteristics to those v/hich are closely related to serum proteins (10). In this study the normal growth-inhibiting factor against Brucella has been demonstrated in the serum of calves prior to the ingestion of oolostrum. Hansen and Phillips (30) recently reported that the serum of the newborn calf, before the ingestion of colostrum, contains small amounts of proteins immunologicaliy similar to colostrum ''immune” pro­ teins. These were detected by preoipitin methods using antiserums to purified pseudoglobulin from bovine colostrum. According to the present concept, the relative importance of oolostrum to the newborn animal is correlated with the placental structure of the species. If this is the case, the proteins present in the serum of the calf at birth must be of such a nature as to pass the placental barrier and the normal growthinhibiting antibody, in distinction to the specific, must be of this class. Although the experiments described in this paper failed to clarify the significance of the natural brucella growth-inhibiting antibody, certain of its oharacteristios in the presence of sodium sulfadiazine have been made apparent. The normal growth-inhibiting factor shows a marked lability at 56° C.; in some instances the heated serum was unable to inhibit growth in a dilution of 1*10 or greater, and in other instances it exhibited a less pronounced reduction in activity whioh was aocompaniea by prozone growth. These results are in contrast to those of Mackie and Finkelstein who found the natural bactericidins generally stable at 55° c» Although concluding that baotericidins were not destroyed at this temperature, they noted that an inhibiting substance from unwashed organisms was unable in some cases to inactivate the bacterioidin in unheated serum but inactivated it in heated serum. This difference in sensitivity of the bacterioidin led them to suggest that some additional factor, possible thermolabile, was concerned in the bactericidal phenomenon. It is presumed that the increased sensitivity of the sulfadiazine-antibodycomplement system aocounts for the discrepancy between their results and ours. On the other hand, it may be due to differences in the individual bactericidins and organisms studied. The other natural complement- fixing antibodies which react with bacterial antigens are labile at 55° C., and it is not surprising to find the natural brucella growthinhibiting faotor in this class. Natural opsonins have long been rec­ ognized as thermolabile at 5&° C. Ginsberg and H0rsfall (jl) found the component of normal serum which neutralized the infectivity of several viruses to be thermolabile at 56° C. Likewise, Chang (i52) re­ ported the spermicidal substance of fresh normal serums, including bovine, to be in this category. However, in the latter investigation the lability does not appear to be distinguished, with certainty, from a complement requirement of the system. From studies of the effect of heat on specifio antiserums, re­ viewed and extended by Follensby and Hooker (33)* it would appear that the reduced aotivity and prozone growth of heated normal serum might result from a similar meohanism taking place at lower temperatures. Snidey (^h) showed that heating induced prozone agglutination of anti­ serums, and he confirmed the finding of Jones (p5) that heated antiserums inhibited the action of unheated antiservuns. Van der Scheer et al. (3&) and Hardt et aJL. (37) have demonstrated electrophoretically that heating of normal serum at 56° or 65° C. causes the formation of a colloidal complex at the expense of various serum components. In examining the effect of heat on the formation of protein oomplexes, Kleczkowski (38)» (39) and Bawden and Kleczkowski (1+0) demonstrated that antiserums which were partly denatured by heat behaved like mixtures of particles with different properties, all able to combine with antigen but only some able to cause precipitation. by the regular antibody. The mixed complex inhibited precipitation Jennings and Smith (1+1) showed that heating of antipneumocoocal antiserum with non-specific protein reduced its precipitability and the concomitant formation of a new component was confirmed electrophoretically by Krejci et jal. (1+2). While the formation of a new component could account for the reduced aotivity and prozone reaction of heated normal serum, it fails to explain the difference in sensitivity of the specific and normal growth-inhibitors to heat. In view of the demonstration by Hardt et al. (37) of the pro­ tective action of certain sugars against heat denaturation of serum proteins, it seems possible that the relatively high resistance of oolostrum to heat and storage, as compared to serum, may be a funotion of its high sugar concentration. It has been shown that storage of serums and serum fractions at I40 C. also resulted in a reduction, and in some cases disappearance, of their growth-inhibiting activity in the presence of sodium sulfadia­ zine. As the effect of heating and storage on the activity of serum is similar, it may result from the same reaction proceeding at different rates. It would be of interest to attempt to confirm this electrophoretically and to correlate the formation of a new component with the reduction in growth-inhibiting activity. In these experiments filtration of normal serum through asbestos pads invariably resulted in a slight reduction of the grovrth-inhibiting activity. Huddleson et al. obtained quite the opposite results as fil­ tration inoreased the bactericidal activity of normal plasma acting without the presence of sodium sulfadiazine. They failed in attempts to elute any inhibitor of the reaction from the filter. The reduction in growth-inhibiting activity demonstrated in this study may have been due either to the retention of material on the filter or to the addi­ tion of ions from the asbestos pad to the filtrate. The presence of the magnesium ion and other cations in the filtrate after passage through asbestos filters is well known. Mayer and his associates (1*3) have demonstrated that the magnesium ion has a stimulatory or inhibitory effect on complement-fixation, the effect depending on its concentra­ tion and ratio to other ions such as caloium. Possibly the magnesium ion concentration following filtration favors the bactericidal aotion of citrated plasma and retards that of serum in the presence of sodium sulfadiazine. Successive filtration through seven asbestos filter pads progressively reduced the activity of the normal sulfadiazine-antibodycomplement system. However, each filtration reduced a smaller proportion of the residual activity. Whether this represents the removal of one of two growth-inhibiting factors in the serum or approach to maximum inhibitory effect of ions added to the filtrate is not known. The sulfadiazine-antibocy-complement system proved to be very sen­ sitive for detecting normal grov.'th-inhibiting aotion against Bruoella. For example, normal bovine serum in an average dilution of 1:80 and in the presence of sodium sulfadiazine and complement inhibited the growth of 10® Br. abortus organisms. The method itself is subject to errors inherent in the use of biological reagents. While fresh, normal rabbit serum was not able to inhibit growth in the concentration at which it was used as a reagent, it exhibited a slight but variable effect on the growth-inhibiting action of normal bovine serum. Although evidence of the additive effect did not extend beyond one twofold dilution, com­ parative results were obtained from tests conducted with the same rabbit serum with the one exception noted. A more significant error was the variation encountered in the effect of the normal growth-inhibitor on dissooiated growth phases of Br. abortus which were not discernible by gross examination of the culture. This can be avoided only by placing proper emphasis on the cultural phase of the organism used for the bac­ terial suspension. No explanation can be offered for the discrepancy between our re­ sults and those of Irwin and his associates who found normal rabbit serum inadequate as a source of complement for detecting bovine serum bacteri­ cidins against Brucella. Possibly sodium sulfadiazine is capable of augmenting some deficiency in normal rabbit serum, or growth-inhibiting action in the sulfadiazine-antibody-complement system may be the mani­ festation of a different bacterioidal meohanism. seem likely to the author. The latter does not Results not presented here indicate that the transient growth-inhibition titer evident without the presence of sodium sulfadiazine and in the early hours of incubation coincides with that observed far more distinctly after 72 hours of incubation in the presence of sodium sulfadiazine. SUMMARY Normal bovine serum, in the presence of sodium sulfadiazine and added complement, was capable of inhibiting the growth of 10 jBr. abortus for 72 hours when in an average dilution of li30. The growth-inhibiting factor resembled other natural complement-fixing antibodies in being thermolabile at 56° C., as compared with the specific antibody which was stable at this temperature. Both storage at I4.0 C. and filtration re- duoed the growth-inhibiting action of normal bovine serum. A brucella growth-inhibiting factor was demonstrated in the serum of newborn calves before the ingestion of colostrum. Serum taken from calves J48 hours after the ingestion of colostrum inhibited growth in higher dilutions than serum taken from calves at birth, but its activity did not exceed that of the adult animal. Colostral whey was found to possess two growth-inhibiting factors, one whioh did not require complement and a second dependent on the presence of complement to inhibit growth of Br. abortus in the presenoe of sodium sulfadiazine. The effect of heat, storage, and filtration on the colostral growth-inhibiting activity was less than on the cor­ responding serum factor. Fraotions of serum from one brucella-infeoted cow and five normal cows were separated by precipitation with different concentrations of sodium sulfate and examined with respeot to growth-inhibiting activity and electrophoretic components. The growth-inhibiting factor of the specific antiserum was insoluble in 15 percent sodium sulfate; the major part of the factor in three of the five normal serums was soluble in 15 percent but insoluble in 20 percent sodium sulfate. A correlation was noted between the y-globulin content and the growth-inhibiting activity of fraotions from specific antiserum. No correlation existed between the y-globulin content and the activity of fraotions from normal serum. In the presence of sodium sulfadiazine fresh, normal bovine serum and colostrum were able to inhibit prozone growth in serum from a brucella-infected cow. The identity of this faotor with the growth- inhibiting factor was examined. Fresh, normal rabbit serum served as an adequate source of comple­ ment for the brucella growth-inhibiting action of normal bovine serum in the presence of sodium sulfadiazine. It was demonstrated that the use of different, but closely related growth phases of Br. abortus influenced the results of the sulfadiazine antibody-complement titrations. error in comparative studies. This constitutes a potential souroe of LITERATURE CITED Huddleson, I. P.* The potentiating action of sulfonamides on the brucella antibody-complement system. Amerioan Journal of Veterinary Research, 9* (1948), 277-285. Huddleson, I. F.* Immunity in brucellosis. 6, (19I+1), 111-142. Bacteriological Reviews, Huddleson, I. P.* The bactericidal action of bovine blood for Brucella and its possible significance. Journal of Bacteriology, 50/ Cl9k5), 261-277. Irwin, M. R., and Ferguson, L. C.i Increase of bactericidins in the serum of cattle following recovery from infection with Brucella abortus. Proceedings of the Society for Experimental Biology and Medicine, 38, (1938), 451-452. Irwin, M. R., and Beach, B. A .: Differential bactericidal activity of bovine serum toward strains of Brucella abortus of high and low virulence. Journal of Agricultural Research, 72, (1946)* 83-91. Irwin, M. R., and Bell, F. N.t On natural antibodies in the rabbit and hereditary resistance to infeotions of Brucella suis. Journal of Infectious Diseases, 57* (1935)* 74-77* McGirr, J. L.j Colostral transmission of antibody substances from mother to offspring. Ifeterinary Journal, 103, (1947), 345-356* Smith, E. L .1 The immune proteins of bovine colostrum and plasma. Journal of Biological Chemistry, I6L4 ., (1946), 345-358. Smith, E. L.» Isolation and properties of immune lactoglobulins from bovine whey. Journal of Biological Chemistry, lt>5» (1946), 665-676. Smith, E. L .1 The isolation and properties of the immune proteins of bovine milk and oolostrum and their role in immunityj a review. Journal of Dairy Science, 31* (1948), 127-138. Smith, E. L., and Holm, A.j The transfer of immunity to the new­ born calf from colostrum. Journal of Biologioal Chemistry, 175, (1948), 349-357. Mackie, T. J., and Finkelstein, M. H.s Natural baoterioidal anti­ bodies j observations on the bactericidal mechanism of normal serum. Journal of Hygiene, 31* (1931), 35-55* -82- 13. Irwin, M. R., Beaoh, B. A., and Bell, F. N . : Studies on the baoterioidal action of bovine whole blood and serum towards Brucella abortus and Brucella suis. Journal of Infectious Disea3e3, $8, (T936), 15-^2 <, llj.. Shrigley, E. W., and Irwin, M, R, i On the differences in activity of serum complement from various animal speoios. Journal of Immunology, 32, (1937)* 281-290, 15. Coombs, R. R. A,, and Hale, N, H. 1 The conglutination phenomenon, IV, The importance of the ohoioe of complement when examining antisera for the presenoe of oomplement-fixing or complemoitabsorbing antibodies. Journal of Hygiene, J46, (l9l*B), 296-310• 16 . Mackie, T. J,, and Finkelstein, M. H, 1 The bactericidins of normal serum: their characters, occurrence in various animals and the susceptibility of different bacteria to their aotion. Journal of Hygiene, 32, (1932), 1-21+. 17. Huddleson, I. F.: The mucoid phases of the genus Brucella, Amerioan. Journal of Veterinary Research, J, (191*6), 5"10." 18. Fleming, A.: On a remarkable baoteriolytic element found in tissues and secretions. Proceedings of the Royal Sooiety (London). Series B, 93, (1922), 306-317. 19. San Clemente, C. L., and Huddleson, I. F . : Eleotrophoretic studies of the proteins of bovine serums with respect to Brucella. Michigan Agricultural Experiment Station Teohnioal" Bu'lletin, 182, (191*3), 3-kk. 20. Hes 3, E. L., and Deutsoh, H. F. 1 Biophysical studies of the blood plasma proteins. IX. Separation and properties of the immune globulins of the sera of hyper immunized cows. Journal of the American Chemical Sooiety, 71# (191*9), 1376-1382. 21. Maokie, T. J., and Watson, H. F.: On the immunological nature of the principle in serum responsible for the Wassermann r e a c t i o n , with reference also to the flooculation reaction of Saohs and Georgi. Journal of Hygiene, 25, (192b), 176-208. 22. Maokie, T. J., and Finkelstein, M. H . : A study of non-speoifio 00mplement-fixation with particular reference to the i n t e r a c t i o n of normal serum and certain non-antigenic substances. J o u r n a l of Hygiene, 28, (1928), 172-197. 23. Mackie, T. J., and Finkelstein, M. H.i Complement-fixation by the interaction of normal serum and bacterial suspensions— a con­ tribution to the study of natural immunity phenomena. Journal of Hygiene, 50, (1950), 1-2U. 2I4.. Gibson, H. J.: Observations on the ooourrenoe, characteristics and specificity of natural agglutinins. Journal of Hygiene, % (1930), 337-356. -83- 25. Howe, P. E.i The use of sodium sulfate as the globulin precipitant in the determination of proteins in blood. Journal of Biologioal Chemistry, 49, (1921), 93-10?. 26. Cohn, E. J., MoMeekin, T. L., Oncley, J. L., Newell, J. M., and Hughes, W. L.j Preparation and properties of serum and plasma proteins. I. Size and oharge of proteins separating upon equilibration aoross membranes with ammonium sulfate solutions of controlled pH, ionio strength and temperature. Journal of the Amerioan Chemioal Sooiety, 62, (1940), 3387"3393* 27. Svensson, H.i Electrophoresis by the moving boundary methodi a theoretical and experimental study. Arkiv for Kemi, Minerologi och Geologi, 22A. No. 10, (19^6), 1-154* 28. Gordon, J., and Wormall, A.s The relationship between the baoterioidal power of normal guinea-pig serum and complenent activity. Journal of Pathology and Bacteriology, 31, (1928), 753-768. 29. Landsteiner, K.i The specificity of serologioal reaotions. 2d ed. Harvard University Press, Cambridge, (1945), P* 132, li+X• 30. Hansen, R. G., and Phillips, P. H. » Studies on proteins from bovine colostrum. III. The homologous and heterologous transfer of ingested protein to the blood stream of the young animal. Journal of Biological Chemistry, 179, (1949), 523-527* 31. Ginsberg, H. S., and Horsfall, F. L.j A labile component of normal serum which combines with various viruses. Neutralization of infoctivity and inhibition of hemagglutination by the component. Journal of Experimental Medicine, 90, (19^+9)> 4-75“495* 32. Chang, M. C.i The effeots of serum on spermatazoa. General Physiology, 30, (1947), 321-335* 33. Follensby, E. M., and Hooker, S. B.: The effect of heat upon antihemocyanin. Journal of Immunology, 55* (1947), 205-218. 34. Shibley, G. S.j Studies in agglutination. IV. The agglutination inhibition zone. Journal of Experimental Medicine, 50, (1929), Journal of 825-841. 35. Jones, F. S.i The effect on subsequent agglutination of the expo­ sure of bacteria to heated antiserum. Journal of Experimental Medioine, 47, (1928), 245-25U* 36. van der Soheer, J., Wyckoff, R. W. G., and Clarke, F. L .1 An eleotrophorotio study of heated horse-sorum. nology, 40, (1941), 39-^5* 37. Journal of Immu- Hardt, C. R., Huddleson, I. F., and Ball, C. D.: An eleotrophoretio analysis of ohanges produced in blood serum and plasma proteins by heat in the presence of sugars. Journal of Biological Chemistry, 163, (1946), 211-220. i Kleozkowski, A,i The formation of protein complexes in heated solutions of rabbit serum proteins. British Journal of Experimental Pathology, 22, (194.1), 188-192. Kleczkowski, A.J Effect of heat on flocculating antibodies of rabbit antisera. British Journal of Experimental Pathology, 22, (1941). 192-208. Bawden, F. C., and Kleozkowski, A.i The effect of heat on the serological reactions of antisera. British Journal of Experimental Pathology, 25, (1942), 178-188. Jennings, R. K., and Smith, L. D.i The influenoe of non-speoifio protein on the heat inaotivation of antibody to pneumocoooal polysaccharide. I. Effect of various proteins on the heat stability of antibody. Journal of Immunology, 45, (1942), 105-109. Krejoi, L., Jennings, R., and Smith, L. D.: The influence of non­ specific protein on the heat inactivation of antibody to pneumoooooal polysacoharide. II. The eleotrophoretio inves­ tigation of the heat inactivation of the antibody in the presence of oasein. Journal of Immunology, 45» (1942), 111-122. Mayer, M. M., Osier, A. G., Bier, 0. G., and Heidelberger, M.i The activating effect of magnesium and other cations on the hemolytio function of complement. Journal of Experimental Medioine, 84, (1946), 535-548. ACKNOWLEDGEMENTS The author wishes to express her appreciation to Dr. I. Forest Huddleson both for guidance in this study and for the opportunity to observe his dynamic research' in the problem of brucellosis and to Dr. R. Evelyn Sanders for her unfailing assistance and advice.