\_ WWW \ 3%: WWWWWNW A METHOD FOR ADJUSTiNG THE RAWO OF ANTIGEN TO ANTIBODY 1N ANTIBODY NETROGEN DETERMENATlONS OF ANTEPNEUMOCOCCEC SERA AND CONCENTRATES Thesis fcr Lhe Degree of M. S. MICHIGAN STATE COLLEGE C! Dede Barrett 1939 A METHOD FOR ADJUSTING THE RATIO 0? ANTIGENI TO ANTIBODY IN ANTIBODY NITROGEN DHEWMIONS OI ANI‘IPNEUMOCOCCIC SERA AND (DNCESTRATES 3/ \ ‘. ,{‘ --’ . {\- f . V,» .\_ by C. Dale Barrett ’- A THESIS Submitted to the Graduate School of Michigan State College of Agriculture and Applied Science in partial ful- filment of the requirements for the degree of mm 01' SCIENCE Department of Beet eriology 1939 1115515 A MOD 10R ADJUSI'DIG THE RATIO OF ANTIGEN T0 ANTIBODY IN ANTIBODY NITROGM DETERMINATIONS OI AN'I‘IPNEUMOCOCCIG SERA AND GONG ENTRATB Thesis for degree or M. S. Michigan State College C. Dale Barrett 1939 The quantitative evaluation or the antibody content of anti- pneumococcic sera by analysis of the antigen-antibody precipitates has been reported by Heidelberger (15). According to this author the nitrogen content of the specific precipitates is determined by means or the Pregl mim-KJaldahl procedure. It has been shown by Heidelberger and Kendall (18), and Heidelbergor and Kabat (20) , that this method is accurate only then the preper ratio of antigen to antibody is established. The principal: difficulty in the routine use or this procedure for ascertaining the potency of antipneumococcic sera is the es- tablishment or the prOper ratio of antigen to antibody. ‘Ihis problem is particularly important when it is necessary to perform a large number ct analyses with samples of widely differing anti- body content. The method reported in this paper is designed to provide a rapid means for adjusting the ratio of antigen to anti- 136’2‘19 2. how in the appraisanent of antipneumoccccic sera and concentrates. hr the sake of clarity the nitrogen determination will be re- viewed very briefly. The antibody is removed from solution as a specific precipitate upon the addition of either the homologous purified carbohydrate or a saline suspension of homologous pneumo- cocci. From the standpoint of laboratory practice the carbohydrate has no advantage over a suspension of the organisms (20) , and the latter is much more economical and far less difficult to prepare. The antigen referred to in this report is a suspension of the organ- imns. The preportions of pneumococci and serum must be chosen so as to leave an excess of cells. The nitrogen content of an accurately measured veins of the pneumococcie suspension is deducted from the total nitrogen found in the specific agglutinate. The difference is the amount of agglutinin nitrogen ranoved by the bacterial cells (82). When analyzing an unknown serum for antibocb' nitrogen it has been customary to select a 0.5 ml or a 1.0 ml aliquot, depending upon the potency of the serun. Using amounts less than 0.5 ml involves unavoidable errors in pipetting. These aliquots may prove satisfac- tory for original ears, but for concentrated products difficult com-_ plications arise. With such products the amount of nitrogen in even a 0.5 ml aliquot is usually far too great for micro methods. In addition to this factor, enormous amounts of antigen are involved in removing the antibow from these solutions. Obviously such concen- trated products can be arbitrarily diluted, e.g., ten times; than a 3. 1.0 ml aliquot can be analyzed with accuracy. When confined to the analysis of only a few samples a day, or 1when the ample. all contain nearly the same concentration of anti- body, this simple method of haphazard dilution can be used, but with doubtful satisfaction. A more reliable method must be employed there there are a large number of samples in which the mount of anti- body is an entirely unknonn factor, or at best only rmJghly known. An example of this situation would be the simultaneous analysis of a large number of routine concentrations of antipneumococcic sera, where it is desired to know the antibody nitrogen content of the original sore, fractions obtained during refinement, and the final concentrates. In such instances the antibody my vary from a mere trace to over 10 mg of nitrOgen per ml. A preliminary procedure for the estimation of the potencies would simplify the analysis. Agglutinin titers indicate approximately the amount of anti- body present, and, when properly interpreted, this titer has a semi-quantitative significance. It is on the basis of thistiter that an unknown pneumococeic antibody solution is classified as to its relative antibocw concentration. Ihen this is known, the volume of antigen required to precipitate completely the antibody in a given aliquot can be calculated, thereby eliminating the necessity of adding the antigen in successive small amounts until no anti- body remains in the supernatant. 4. kperimental Detemination of the Agglutipin Titer. The mknown antibody was diluted with physiological saline solution in the ratio of a geometrical progression starting with a 1:2 dilution. It was selibm found necessary to go beyond a series of ten tubes for each sample. For the sake of convenience, 0.5 ml of saline was pipetted into each tube. Into the first was placed 0.5 ml of the smple, which was mixed, and transferred in like manner to each mcceeding tube. One drop of homologous antigen was added to each tube excepting the tenth which was reserved in case the titer went beyond this point. The antigen used for this purpose was the same smpension that was pre- pared for the quantitative detenninati on. The tubes were then placed in a mechanical shaker and observed after tn minutes of agitation. Unless the titer was zero, the ”equivalence zone“ was readily apparent. In this zone there was always one tube in which. occurred an optimum agglutination of the organisms. Considerable care was exercised in selecting this tube. Often there were two or three tubes very much alike in appearance. In this sitmtion the tube having the larger individual flakes, which tend to stick to- gether, has been designated as the ”Optimal" titer. When there was any doubt, it was found advisable to select the tube having the higher dilution. Later on it will be dicwn how on the basis of this titer the prepcr dilution for a s atisfactory nitrogen determination was calculated. 5. Alow potency horse serum was selected which contained from 0.3 mg to 0.5 mg of antibow nitrogen per ml. A stronger serum could have been used provided it was diluted to meet these requirements. It is this serum which is referred to as the control antiserum, because it was to this concmtration of nitrogen that all of the unknown antibody solutions were adjusted. the agglutinin titer was ascertained according to the method just described, and was rechecked about every two weeks. A 500 ml lot of Type 1 antipneumococcic serum was prepared in this manner, preserved with merthiolate and stored at 4°C. we supply has lasted for nearly six months, and its potency has not appreciably dropped as indicated by its agglutinin titer or by quantitative analysis. The purpose in selecting a control serum of low potency, i.e., between 0.3 mg and 0.5 mg antibody nitrOgen per ml, was to restrict, within practical limitations, the amount of standardized antigen re0 quired for complete precipitation of the antibody. This resulted in a more economical use of the antigen than could have been obtained with a control scrim of higher potency. It was assumed that any un- known antibody solution adjusted to this concentration would require the same volume of antigen as indicated for the control serum. cient pnetmcoccus culture was prepared to make about 500 ml of ultimate bacterial suspension. It was found inadvisable to prepare 6. more than a month's supply of antigen. After this period of time the agglutinatingactivity of— the antigen was unreliable. ne antigen used for this test was a formalin-killed, 12-hour culture of pnemiococci of maxim virulence. 'Ihe cells were centri- fuged and washed until the supernatant was nitrogen 12.... Every precaution was taken to prevent autolysis of the organisms. The final suspension was diluted with saline solution, to which had been added formalin to a concentration of 2 per cent, so as to con- tain between 0.20 mg and 0.25 mg of total nitrogen per ml. the procedure for standardizing the antigen as to its agglutin- ating ability consists in adding varying mounts to a constant volume of the control some One ml of this serum was pipetted into each of 14 pyrex test tubes 28 an x 150 ms. These were the same tubes whidi were used to carry out the digestion and distilla- tion in the Kjeldshl apparatus. Diana saline solution was added to each tube tomake the ultimate volume 10 ml. Duplicate quanti- ties‘of the antigen were added to the serum in the following amounts: 0.25 ml, 0.5 ml, 1.0 ml, 2.0 ml, 3.0 ml, 4.0 m1, and 5.0 m1. 'flle antigen was measured out in a calibrated pipette. The saline solu- tion was present to insure the optima proportions of antigen to antibody in all tubes. The tube which shows maxim precipitation of the antibody, with‘a minim. amount of antigen, indicates the 0p- timum volume of antigen to add to any solution which has the same concentration of antibody as the control serum. Prgmgtion of Smnples for Quantitative Anglysis. “me concen- tration of the unknown antibody solution in relation to the central serun was calculated from the ratio of their agglutinin titers. amples stronger than the control serum were diluted accordingly; the weaker samples were used undiluted and in larger ammts. for ample, if the control serum 0 had a titer of 1:64, and antibody solution A bad a titer of 1:512, than A was approximtely 8 times more concentrated than 0. Therefore, a 1:8 dilution of serum A was accurately prepared using physiological saline solution as the diluent, and a 1 ml aliquot of this was used for the actual assaw. If antiserum B had a titer of 1:16, then 4 ml of B would be used undiluted. Since it was found impractical to use serun aliquots larger than 8 ml, this amount was not exceeded in any instance. The were were measured in duplicate quantities with cali- brated pipettes into the aforementioned pyrex test tubes (22 m x 150 m). Saline solution was then added to each tube followed by the calculated amount of antigen, to make a total volume of 10 ml. 1 calibrated burette was used to measure out the antigen. 'Ihe contents were then thoroughly mixed by placing the tubes in a mechanical shaking machine for 5 minutes. The tubes were sealed with parafiln and allowed to stand overnight at 4° C. ‘ihey were centrifuged from 30 to 60 minutes at 2000 r.p.m. The period of centrifuging depended upon the nature of the agglutinate. 8. As a precautionary measure the supernatant of each sample was checked.ror the presence of nonprecipitated antibody. The agglutin- in.tewt was used for this purpose, as it was found to be more satis- factory for routine use than the precipitin test. An undiluted aliquot of each supernatant was tested with one drOp of the antigen; the desired reaction was the complete absence or agglutination, which indicated that the antibody had been entirely precipitated. From the assay of over two hundred samples the criterionwas estab- lished that not more than 10 per cent of a series of samples should show tree antibody in the supernatant in accordance with this test. If this situation did not exist then it was assumed that either the amount of calculated antigen had been incorrectly detmined, or the antiserum was not sufficiently diluted. As a tmporary measure more antigen was arbitrarily added to the supernatant in 1.0 m1 quantities until the agglutinin test was negative. The supernatant was then renoved by means of suction applied to a capillary pipette, care being taken not to lose any of the agglutinate. Each agglutinate was washed twice with ice cold water, and centrifuged. The nitrogen was quantitatively determined by a modification of the Pregl micro-KJeldahl procedure. A measur- ed amount of antigen was run through with each series of samples in order to check the nitrOgen content of the cells. The differ- ence between.this value and that obtained fer the specific agglup tinate represents the agglutinin nitrogen of the sample. A.comparison of methods: Table 1 results obtained (1) when the antigen-antibody ratio was not properly adjusted. Tube ‘vfiolume of volume of anti- -Qualitative Mg of anti- No. antibody gen test on sup- body nitro- Average ernatant for gen per*ml Dilution nu. .Ireehantibody mg. 1m; 1 1 m1 of 1:10 0.25 initial + 4.0 2 4.3 4'2 3 ' 0.5 " ++ 5.7 4 6.7 6'2 5 ” 1.0 ” +++ 8.0 6 7.9 8.0 7 < " 2.0 " - 9.5 a __ 9.1 9'3 9 H 400 ” - 11.04 10 _ _ 11.9 11:7 11 " 6.0 " - 9.3 12 9.1 9'2 0.25 " + 13 ” 0.5 added * ++ 14 0.5 " +++ 005 " "' 1.75 Total 11.4 ~_J 11.4 11.4 _W“ ‘ ’L fi'.’ ‘11:" w — "" I2) when the antigen—antibody ratio was properly adjusted. 15 1 m1 of 1:3 3.0 initial ** - 11.4 11 2 16 11.1 ‘ * 0.5 ml quantities were added in successive amounts until the supernatant tested negative fOr antibody. ** Calculated amount. Scheme for interpreting the qualitative test: + slight ++ moderate +++ heavy. no agglutination 10. Table 2 Importance of selecting‘lie correct titer and the correct quantity of antigen. 1'an Igglutin-F voime or venue of anti- Qualitative n3 antibody No. in titer antibody gen test on sup- nitrogen ernatant for per ml __ mm ‘ 2.0 m1 initial + 1.0 " added + 1.0 ” ' ++ 1 1:128 1 m1 of 1:2 1.0 " " ++ -. 1.0 ” " +++ 1.0 * " +++ 1.0 ' ” " 8.0 " total - 11.2 2 13256 1 ml or 1:4 2.0 " initial +++ 1.0 ' added +++ 1.0 Q R " 4.0 " total - 10.8 _§_ 1:512 1 m1 of 1:8 3.0 " initial - 11.3 5 1:1924 1 m1 of 1:16 2.0 " initial - 9.? 6 1:2048 1 m1 of 1:32 2.0 . initial - 10.2 7 1:40“ 1 m1 of 1:64 2.0 * initial - 3.5 11. Table '3 standardization of the antigen as to its agglutinative activity. Volume of antigen added to 1.0 ml of the control serum Lot No. m1 m1 m1 m1 m1 m1 m1 Initial vol- of anti 0.25 0.50 1.0 2.0 3.0 4.0 5.0 ms of anti- gen _ gen required llg antibody nitrogen per m1 of control serum as calculated Type 1 0.18 0.26 0.36 0.40 0.42 0.42 0.40 3.0 m1 5a - :7 Type 1 0.17 0.31 0.35 0.42 0.39 0.41 0.41 2.0 ml 51: ,, Type 1 - - 0.45 0.47 0.47 0.47 0.45 2.0 m1 6 m. 1 0014 0023 0.32 0035 0.38 001‘ . 300 m1 7a Type 1 0.11 0.23 0.32 0.39 0.40 - 0.32 3.0 m1 7b m. 2 " 0e51 0062 0066 0077 0078 0.75 3.0 m1 2a 2.0 m1 Note: highest and lowest values. lOt'e It is evident that regardless of the quantity of antigen required for complete precipitation or the antibody in a given control serum, the net nitrogen value should be the same, lith the various lots of Type 1 antigen shown there is an unexplainable error of 0.09 mg between the This error, however, does not have any effect on selecting the initial calculated volume of antigen for the individual Table 4- 12. Data showing the reliance of the calculated mount of antiga: as being sufficient fa'complete~ precipitation of the antion in an unknown serun. __ lint- inin titer Volume of anti- body solution used Sam No. Lot No. of an- tigen Calculate ed mount ant igen addad 8 m1 undiluted 5'7 99 0000' 00 Qualitative test on su- pernatant fer free £21393! Additi 07:3 antigen required 0 (10‘ r 1:8 +IO'O 3:16 0866' +011, 1:32 IOO'P. 1:64 O'U'O’ 1:128 1 m1 of a 1:2 dilution 4—; onto IOO.+ 1:256 1 ml of a 1:4 dilution 105g 0900““ 0000010 ENGGG 95005009 00000 OOOGO O DONNIFU NNIFNM NWNMN {010N301 NIFNNN ”£0513? NIPFr‘ .0000 OOOOH 0000“ 00000 000100 POOOO HOOOO Table 4 (Cont' d.) Agglut- Volume of anti- Serum Lot No.1 Calculat- Qualitative! Additional inin body solution No. of an- ed amount test on su- antigen titer used " tigen ant igen pernat ant required added for free ml km 1:512 1 m1 of a 1:8 22 5 c 4 - 0 dilution 32 5 a 3 - 0 ‘38 5 a 3 - 0 42 5 c 4 -- 0 78 6 2 - 0 1:1024 1 ml of a 1:16 14 5 b 2 - 0 dilution 44 5 c 4 - 0 59 6 2 - 0 64 6 2 - 0 71 6 2 - 0 1:2048 1 ml of a 1:32 2 5 b 2 — 0 dilution '3 5 b 2 - 0 5 5 b 2 - 0 6 5 b 2 - 0 23 5 b 2 - 0 1:4096 1 m1 of a 1:64 dilution 4 ' 5 b 2 - 0 10 5 b 2 - 0 17 2 a 3 - 0 18 2 a 3 - 0 iL 26 l 5 b l 2 «- l 0 14. Table 5 Showing results of agglutinin titers and their interpretation. said 1:2 1:52 1:64 1:1 1:256 1:512 1:10 1:2048 Titer + + + ! v ‘00 1 l i 0 Ya + 6 + 6 t '3 + .3 + '.§.._..... t '64 + . . + 28 + + + + + + + + + + + + + + + + + + + + + + - None + Slight ++ Heavy +++ Optimum * Control serum Table 6 15. Correlation between agglutinin titer and antibody content of Type 1 antipneumococcic sera. m-~.-_—. .1-.. .p— 4“ _.~_~..a own~-—._ _...__. _. ‘_ __ _ __.—..~...r ...—.. —- 1...... .. “‘“-..’-~ _..._-_.._. —. _._-_l- _. _. I - Serum Mg. Anti - Agglutinin L. M T . t e la 0.09 9. 41 0.01 1:4 43 0.05 45 0.02 47 0.01 57 0.09 62 0.08 22...... ALL 86 0.20 1:8 97 0.22 10: 0.17. 24 0.06 1:6 27 0.27 30 0.24 35 0.29 34 0.28 80 0.24 LQQ 0,37 55 0.48 1:32 50 0.50 67 0.11 76 0.50 75 0.47 79 0.58 98 0.54 89 0.41 91 0.58 93 0.30 95 0,42 53 0.70 1:64 81 0.49 84 0.79 195...... L.___.I__o 65 52 0.80 1:128 48 1.38 56 1.60 107 1.24 13 1.26 58 1.56 60 1.92 61 1.12 66 1.44 70 1.28 85 0.98 87 1. 26 100 2.06 92 0.86 96 1.54 F k W““_ _.~W-—.—‘-"-o—- --- —. ,. . .— .— r... .. Serum Mg. Anti- Agglutinin ,9 bog: Nfiyl T110: 25 2.24 1:256 37 5.04 22 3.40 65 I.“ 75 2.68 77 4.08 103 3.20 ': 3 35 :38 9.26 1:512 46 8.56 40 7.52 44 8435 49 8.88 15 9.75 54 8.25 64 8.00 69 8.01 78 9.52 83 8.00 88 7.69 108 6.40 74 7.76 32 11.28 1:1024 33 11.00 29 11.28 31 10.09 28 10.02 42 11.60 51 10.89 59 9.14 71 11.07 ' .JQJE CONTENT OF. VARIOUS AN Tl PN. SERA 82 SAMPLES CORRELATION BETWEEN AGGLUTININ TlTER AND ANTIBODY 16. CURVE REPRESENTING MAXIMUM —-> VALUES VALUES I: 2 4 8 IO 32 64 I28 256 5|! l024 AGGLUTININ TITER \ 17. Discussion The difficulties encountered when the antibody content of a serum is ascertained without adjusting the antigen—antibody ratio are shown in Table l. A Type 1 concentrated antipneumococcic serum (No. 71) was chosen at random. Since it was known that this was a concentrated product, it was arbitrarily diluted ten times with saline. Under actual conditions only one of the six quantities of antigen shcwn in the table would have been added to the 1 ml aliquot of serum. Supposing the initial amount had been 0.5 m1 an erroneous result would have been obtained. 0n the other hand, if it had been 6.0 ml the nitrogen value would be low, by virtue of the fact that a large excess of antigen inhibits complete precipitation of antibody. Obviously, the only accurate procedurais to begin with a small ini- tial value of antigen, and add successive small amounts to the supernatant until the Optimal prOportions of the components are reached. This is impractical and time'conmnning as is evidenced by the one instance cited in this table when it required an hour and a half to follow this procedure. In tubes N0. 15 and 16 this same sample was determined according to the method present ed in this paper and the results are self-explanatory. In Table 2 the importance of interpreting the correct titer from which to make the subsequent dilution of the antibody is an- phasized. Also, the importance of correctly determining the ini- tial volume of antigen is pointed out. It is evident in this table 18. and the one preceding that the qualitative agglutinin test does not detect sinificant traces of antibody in the supernatant. This would explain how the situation in tubes No. 3 and 4 could arise. 0n the basis of the criterion (24) that the antigen must be in slight excess for the complete precipitation of antibom', it is evident that this condition does not exist in tube No. .3. When the antibody is diluted until the antibody concentration is quite small aid the antigen is held constant, the results become less reliable. In tube No. 7 there was a large excess of cells in propor- tion to the antibody. Table .3 illustrates how the initial quantity of antigen was cal- culated in order to cause complete precipitation of the antibody in any antiserum adjusted to the concentration of the control sari... Several different lots are shown including two Type 2 pneumococcus suspensions. Whenever there was a question regarding the choice of two quantities of antigen, the larger volume was always selected in order to increase the margin of safety. The reliability of employing this calculated amount of antigen is town in Table 4. There are cited fifty-five samples s elected at random from this study. It was from these data that the criterion was established that not over ten per cent of a series of samples should show free antibochr in the supernatant after the initial suspension of pneumococci had been added. ‘ Unavoidable errors may arise due to the interpretation of the 19. agglutinin titer by different persons. For this reason Table 5 has been included. Here an attanpt has been made to depict the many different appearances the dilutions may have throughout the ”equiva- lance zone”. Two antiserums are cited for each titer. The writer's interpretation is recorded in the last column. The reliability of the agglutinin titer as a means of approx- imating the antibody content of antisera and concentrates is veri- fied in Table 6 and the accompanying graph. There is a definite correlation between the agglutinin titer and the subsequent anti- body nitrogen value. As would be anticipated with dilutions organ- ized in a geometrical progression, the points plotted for any given titer are m1. especially grouped at one value but are distributed throughout a definite area. It is also evident that for titers below 1:4 the antibody nitrogen is so low as to be considered in- significant. No attempt has been made to prove that this serological test could ever replace the nitrogen analysis as a quantitative deter- mination. However, it is conceivable that a closer relationship could be worked out between the agglutinin titer and the correspond- ing antibody nitrogen value. In Table 4 it is shown that this ree lationship is close enough for all practical purposes. It was found important in performing the agglutinin titer test that the quantity of antigen added to the serun dilutions should not beexcessive; e.g., rather than mixing 0.5 ml of the serum dilutions with an equal volume of antigen, one dr0p was sufficient. With ex- cessive amounts of antigen their agglutination was inhibited, which resulted in a false interpretation of the titer and inconsistent results. The amount of agglutinns:fonned was shown by Heidelberger (20) to be independent of the concentration of antibody but dependent upon the relative proportions in Which the components are mixed. The agglutinin content is given in absolute terms when the proportions of pneumococci and serum are such as to leave the cells in slight excess. In the method presented in this paper, the correct amount of cells is selected experhmentally for a given serum.and by means of the agglutinin titer the prOportion of antibody to antigen is brought to the proper ratio regardless of the potency of the serum. The experimental procedure for the develOpment of this method was done almost entirely with Type 1 pneumococci. Enough work was done with Types 2 and 3, however, to justify the application of this method to those types. Summary The proper antigen-antibody ratio of antipneumococcic sera must be established prior to an accurate nitrogen determination. A practical method for obtaining this ratio, when a large number of samples of widely differing antibody content are to be assayed, is presented in this paper. 21. There was shown to be a close enough correlation between the agglutinin titer of an antiserum.and its specific nitrogen content to enable a preliminary estimation of its potency to be made. The antigen used in this study was a formalin killed, saline suspension of pneumococci standardized as to its nitrogen content and agEIUP tinative activity. The correct preportion of cells to antibody was first ascer- tained for a control serum.of low potency. The agglutinin titer of the unknown serum was obtained.and.compared with the titer of the control scrim. The antibody concentration of the unknown was then adjusted to that of the control, and the same volume of cells ‘Ias added as was indicated for the control serum. It was seldom found.necessary to add further antigen. This method is applicable to any solution of Type 1, 2 and 3 pneumococcic sera regardless of their respective antibody concen- tration. . ”i. ifl| “in I . Idlljiil“ \ 1. 2. 3. 4. 5. 6. '7. 8. 9. 10. 22. Bibliography Avery, O.T., and Heidelberger, M. Imm01031ca1 relationships of cell constituents of pneumococcus. J'. Exper. Med., gym-cs (1923) Avery, O.T., and Heidelberger, M.. Immunological relationships of cell constituents of pneumococcus. Paper II. J’. Erper. Med., $367-$576 (1925) ‘Barnes, L.A., Clarke, 0.11., and flight, LC. Comparisons of various methods for the routine titration of Types 1 and II antipneumococcic horse some. I. Imunol., _39_:127-138 (1956) Boyd, 1.0., and Hooker, 8.3. The influence of molecular weight antigen on the preportion of antibomr to antigen in precipitates. J’. Gen. Physiol., _l_1:34l-548 (1934) Brown, R.A. A specific precipitation test for the standardization of Type I antipneumococcus serum. I. Immmol., 32:149-154 (1933) Chou, B.F., and Goebel, 13‘. The purification of the antibodies in Type I antipneumococcus serum, and the chemical nature of the type-specific precipitin reaction. 3'. Exper. Med., §_2_:179-202 (1935) Dean, H.R., and lebb, R.A. The influence of Optimal proportions of antigen and antibody in the serum precipitation reaction. I. Path. 8: Bact., 33:473-492 (1926) Duncan, LT. Optimal agglutination: The significance of the different ratios of serum to bacteria. Brit. J. Exper. Path., fizfiS-M (1934) Duncan, 191'. The specific character of the stage of aggregation in agglutina- tion and precipitation of antibody-antigen compounds. Brit. J’. Exper. Path., £25286838 (1938) Francis, T., Jr. The identity of the mechanisms of type-specific agglutinin and precipitation reactions with pneumococcus. J'. Exper. Med., _5_5_:55-62 (1932) 11. 12. 1’3. 14. 15. 16. 17. 18. 19. Eudora, 1.2!. A note on the specific agglutination of pneumococcus Types I, II and III. I. kper. Med., {33191-202 (1%2) Goodner, L, and Horstall, 1.1... Ir. Properties of the type specific proteins of antipneumococcus sera. I. The mouse protective value of Type I sera with re- ference to the precipitin content. I. kper. Med. 99413-424 (1937) Ibid. II. Imunological fractionation of Type I antipneumococcus horse and rabbit ears. I. Exper. Med., §_§_:425-435 (1937) Eeidelberger, 11., and Kendall, LII. A qmntitative study of the precipitin reaction between Type III pneumococcus polysaccharide and purified homologous anti body. I. hper. Med., 203809-825 (1929) Heidelberger, M., Sia, R.H.P., and Kendall, 1.1. Specific precipitation and mouse protection in Type I antipneumo- 00cm am. I. hper. lied., 2:477-483 (1930) Heidelberger, M. , and Kabat, E.H. Ohmical studies on bacterial agglutination. I. A method. J. Exper. lied” Qz643~653 (.1934) Kai delberger, 11., and Kendall, RE. The precipitin reaction between Type III pneumococcus polysac- eheride and homologous antibccw. II. Conditions for quantita- tive precipitation of antibody in horse sera. J. kper. Med" 313559-562 (1935) Heidelberger, 1L, and Kendall, LE. The precipitin reaction betveen Type III pneumococcus polysac- charide and homologous antibody. III. A quantitative study and a theory of the reaction mechanism. Jo kper. “Ode, 25563.59]. (1.935) Heidelberger, M., and Kabat, LA. musical studies on bacterial agglutination. II. The identity of precipitin and agglutinin. J’. Exper. lied., gz737-744 (1936) so. 21. 22. 2'7. 24. Heidelberger, 11., and Kabat, 13.13.. Chemical studies on bacterial agglutination. III. A reaction mechanism and a quantitative theory. I. kper. Med., §_§:885-902 (1937) Heidelberger, l. , and Kabat, E.A. Chemical studies on bacterial agglutination. IV. The reaction of pneumococcus specific polysaccharide with homologous rabbit antisera. J'. kper. lied., fig:64'7-660 (1937) Heidelberger, M. , and Kabat, E.A. Chancel studies on bacterial agglutination. 7. Quantitative data on pneumococcus R (Dawson 3) anti-R (8) systems. J’. hper. Med., 91:545-550 (1938) Heidelberger, M. Chanical aspects of the precipitin and agglutinin reactions. Chm. Reviews, fizSZS—S‘is (1939) Heidelberger, H. Quantitative absolute methods in the study of antigen-antibody ' reactions. Beet. Reviews, _3_:49-95 (1939) Hooker, 8.13., and Boyd, EC. The non-specificity of the flocculative phase of serologic ag- mgatione J". Imunol. yam-351 (19:57) Harrack, 1., and Snith, 1.0. (a) Quantitative aspects of the immunity reactions; the precip- itin reaction. Brit. J. Exper. Path., 13:30-35 (1930) (b) Composition of the precipitate in precipitin reactions. Ibid, pause-190 (1931) Halkiel, 8., and Boyd, 1.0. The composition of specific precipitates in the region of anti- gen 810088. J’. Exper. Med., §_§_:383-395 (1937) Morgan, 3.3. The inhibition zone in precipitin reactions with the soluble specific substance of pneumococcus. J. Immunol., 93449-456 (1923) 51. 52. 83. 25. Northrop, 1.11., and DeKruif, P.H. The stability of bacterial suspensions. III. Agglutination in the presence of protein, normal serum and immune serum. J’. Gen. Physiol” 33655-667 (1922) anith, I. The titration of antipneumococcus serum. J'. Path. and Bact., 315:509-526 (1932) Sobotka, H., and Friedlander, M. The precipitin reaction of antipneumococcus sera. III. The ratio of precipitin to protective antibody in Type II. J'. Exper. Med., _5_§:299-301 (1931) Topley' 'OWOC’ Wilmn’ IO. and Duncan, JCT. The mode of formation of aggregation in bacterial agglutination. White, B. The biology of pneumococcus. The Commnwealth Fund, (1938), Oxford University Press. WOW his sort was financed by a grant from the Con-onsealth land of new York City to the Michigan Department of Health Laboratories. The author is indebted to Dr. lard Giltner of liaigen State College for his advice and criticise, and to Dr. I. T. Tripp of the Michigan Deparhent of Health Labor- atoriee for directing the investigations carried out in this thesis. The assistance of Dr. P. A. Iolf and Ire. J’. Glasses, also of the Michigan Department of Health Laboratories, in conducting'the laboratory analyses is gratefully acknowledged. ROOM use ONLY ( J v 1 . ( - t e I . \> \.