Immunochemical Studies of Human Serum Gamma Globulins

IMMUNOCHEMICAL STUDIES OF HUMAN SERUM GAMMA GLOBULINS

Solomon J. Zak, Robert A. Good

J Clin Invest. 1959;38(3):579-586. https://doi.org/10.1172/JCI103835.

Research Article

Find the latest version: https://jci.me/103835/pdf IMMUNOCHEMICAL STUDIES OF HUMAN SERUM GAMMA GLOBULINS * By SOLOMON J. ZAK t AND ROBERT A. GOOD t (From the Minneapolis Veterans Administration Hospital, and the University of Minnesota Medical School, Minneapolis, Minn.) (Submitted for publication September 15, 1958; accepted October 10, 1958) The establishment of the quantitative precipi- the zone of antibody excess, has been well docu- tin reaction as an analytical-chemical procedure mented (7, 8). For the purposes of this study by Heidelberger and Kendall (1) was followed in it was felt that a physically or chemically homo- 1937 by Kendall's original use of the method in geneous gamma globulin preparation, possessing the estimation of human serum proteins in health immunologic homogeneity in the zone of anti- and disease (2, 3). A major drawback to the body excess, might be employed for preparation of use of this immunochemical method for obtaining antibody to be used in a quantitative precipitin meaningful human serum protein values has con- test for estimation of gamma-reactive globulin in cerned the purity or homogeneity of the specific human fluids. Because of the inadequacy of other protein used as the antigen (4). In particular, methods, it was considered likely that this tech- the heterogeneity of the human gamma globulin nique might be especially useful when the con- fraction, as defined electrophoretically, has been centration of the gamma globulin to be measured pointed out (5). Cohn, Deutsch and Wetter have is very low. demonstrated distinctive immunochemical be- The physical, chemical and immunological simi- havior in various subfractions of the human se- larities of the human gamma globulins and most rum gamma globulins (6). In these studies, specific antibodies are well documented (9, 10). however, the principal differences in immuno- On the biological level, the clinical entity agam- chemical behavior of the subfractions were noted maglobulinemia provides strong evidence in sup- in zones of antigen excess; this zone is not em- port of the concept that gamma globulin and ployed in the studies reported herein. most antibodies arise from a common source (11, Although numerous physical, chemical and im- 12). Accordingly, immunochemical estimations, munological criteria have been set forth for the patterned directly after the methods of Heidel- establishment of protein purity, the lack of in- berger and Kendall (1), have been applied to the formation regarding protein structure does not study of agammaglobulinemic and normal human permit a clear definition of a single human pro- sera, and reference comparisons to electrophoretic tein species. In general, the proteins comprising observations have been accomplished. The pur- the human serum gamma globulins separated by poses of the present study were several-fold and ethanol fractionation have an isoelectric distribu- included the following: 1. To establish the ef- tion of pH 5.7 to 7.5; their electrophoretic mo- ficacy of the quantitative precipitin reaction for bilities range from 0.97 to 2.64. The marked the accurate estimation of human serum gamma antigenic similarity, indeed almost antigenic iden- globulins. 2. To measure the serum concentra- tity, of the members of this family of proteins in tions of gamma-reactive globulin in normal humans. 3. To describe the quantitative gamma *Aided by grants from the Minnesota Heart Associ- globulin deficit in patients with both congenital ation, the American Heart Association, the Helen Hay estab- Whitney Foundation, the Minnesota Division of the and acquired agammaglobulinemia. 4. To American Cancer Society, and the United States Public lish the biological survival time of parenterally Health Service (RG E-798 and H-2085). administered gamma globulin in humans by im- t Resident, Department of Internal Medicine, Minne- munochemical means. 5. To study the appearance apolis Veterans Administration Hospital. of gamma-reactive globulins in the newborn baby tAmerican Legion Memorial Heart Research Profes- sor of Pediatrics, University of Minnesota Medical of a mother with acquired agammaglobulinemia, School. and to relate the appearance of gamma globulin 579 580 SOLOMON J. ZAK Al.ND ROBERT A. GOOD to the time of onset of specific immune globulin performed as outlined by Kabat and Mayer (16). A synthesis and release; to study the disposition of highly purified gamma globulin preparation obtained by transmitted gamma globulin by the ethanol fractionation was kindly supplied by Dr. J. T. maternally Edsall, and served as antigen in these studies. The agammaglobulinemic offspring of an immunologi- method of Kendall (2) for alum precipitation of the cally normal female who had previously borne a antigen and that of Jager, Smith, Nickerson and Brown son afflicted with the congenital form of the di- (17) for the immunization schedule were adopted. Anti- sease. 6. To compare maternal and cord serum human gamma globulin was produced in albino mixed hy- concentrations of gamma-reactive globulin. brid rabbits weighing 4 to 5 Kg. The alum precipitated antigen was administered intravenously only and in a concentration of 1 mg. per ml. over a three week period. MATERIALS AND METHODS Each animal received a total of 12 mg. of the alum precipitated material. On the eighth day following the Sixty immunologically normal children ranging in age last injection of antigen, each animal was exsanguinated from one day to 15 years: These children were patients by cardiac puncture under sterile conditions. The bloods at the University of Minnesota Hospitals, Minneapolis obtained were allowed to clot at room temperatures for General Hospital and the Ancker Hospital of St. Paul, one hour, placed at 2° C. overnight, the clots then who had been admitted for study or treatment of con- rimmed and the tubes centrifuged at 1,500 rpm for 20 genital anomalies or of trauma. None had been troubled minutes in an International PR-1 centrifuge; the sera with recurrent infections, and in no case was there rea- were decanted and pooled after preliminary testing for son to suspect an abnormality of protein synthesis or an antibody content. The pool was then recentrifuged at immunological deficit. Twenty-four immunologically 2,000 rpm for 20 minutes, decanted, and merthiolate was normal, healthy adults: These subjects were laboratory added to a final concentration of 1 part in 10,000. The and hospital personnel, ranging in age from 19 to 34 years. pool was then heated at 560 C. for 30 minutes, allowed Thirty-one maternal cord serum pairs: In all cases, the to come to room temperature, dispensed into sterile glass maternal blood was drawn within two hours of parturi- tubes, quick-frozen in an alcohol-acetone dry ice bath, and tion. Sera from 16 patients with agammaglobulinemia: stored at -26° C. until ready for study. Prior to use in Nine of these were patients with the congenital form of these and all subsequent determinations, sera and anti- the disease, and seven with the acquired form. The sub- sera were clarified by centrifugation for one hour in the jects chosen for study of gamma globulin half-life deter- cold at 3,000 rpm. minations included four congenitally agammaglobulinemic Absorption-precipitin study. A solution of the gamma patients and a normal newborn who was bled frequently globulin was prepared in nitrogen-free diluent, con- during the neonatal period. All subjects were free of sisting of 0.01 molar phosphate at pH 7.4, and brought clinical infections at the time these studies were per- to ionic strength 0.145 with sodium chloride. The con- formed. centration of gamma globulin in this solution was estab- After preinjection bleedings, each subject in the half- lished by repeated micro-Kjeldahl analyses; volumetric life study was given gamma globulin in various dos- dilutions of this standard were then prepared. Quanti- ages intramuscularly. Venous blood was drawn there- tative precipitin determinations were carried out by add- after at varying intervals. The bloods were allowed to ing 1 ml. of each antigen dilution to 1 ml. of antiserum clot at room temperatures for one hour, placed at 20 C. in standard heavy-walled precipitin tubes. After thor- overnight, the sera separated by centrifugation in the ough mixing by agitation, the tubes were capped with cold, quick-frozen at -70° C., and stored at -260 C. un- parafilm, and allowed to incubate for one hour at 370 C., til ready for use. then placed at 0 to 40 C. for eight days. The tubes The decline of maternally transmitted gamma globulin were agitated daily. They were then centrifuged for one was assessed by studying the male offspring of an ap- hour in the cold at 3,000 rpm, and the supernates poured parently normal mother who had previously borne a off and tested for the presence of antigen and antibody. congenitally agammaglobulinemic male. In this study, All procedures were carried out in the cold. The tubes blood samples were drawn from the jugular vein of the were allowed to drain over filter paper for 15 minutes, and baby in the neonatal period and at intervals through- the precipitates were washed by resuspension in 0.145 N out the first year. nitrogen-free NaCl. This washing procedure was re- of in the newborn and The appearance gamma globulin peated twice. The washed specific precipitates were its quantitative characteristics during the first year of ml. solution and for life were determined in the unique situation of a nor- digested with 0.5 digestion analyzed mal female offspring of a patient with acquired agam- nitrogen by the micro-Kjeldahl method. Repeated analy- maglobulinemia, reported upon in detail elsewhere (13, ses, using the same as well as different lots of antiserum, 14). Frequent venous blood samples were obtained from yielded results agreeing within less than 5 per cent in all the baby's jugular vein throughout the first year of life. instances, and in most instances, the agreement was Total serum proteins were determined by the biuret within less than 2 per cent. Reference was thenceforth method of Weichselbaum (15). Quantitative precipitin made to the zone of antibody excess in the precipitin determinations for human gamma-reactive globulin were curves thus constructed. One hundred-fifty ml. of rab- IMMUNOLOGICAL STUDIES OF HUMAN SERUM GAMMA GLOBULINS 581

TABLE I Serum gamma globulin concentrations in normal humans: 0) Immunochemical estimations 0'

Gamma Group Number globulin Range ONM2400- mg. % Normal infants 13 663.19 417.90-1140.62 C. 1 day-7 wks. Normal infants 9 398.3 214.84- 621.09 3-6 mo. 200 9#4 Normal infants 14 731.2 279.69-1125.00 E 6-15 mo. E0 Normal children 18 882.37 440.62-1541.60 0 _ / * Unabsorbed 19-54 mo. u / Absorbed 11 Normal children 6 1151.9 833.40-1312.50 * Absorbed I 0{,, ,,, O 7-15 yrs. 0 30 60 90 120 Normal adults 24 1124.3 734.00-1687.50 Micrograms antigen nitrogen 19-34 yrs. FIG. 1. ABSORPTION-PRECIPITIN STUDY OF RABBIT ANTI-HUMAN GAMMA GLOBULIN SYSTEM bit anti-human gamma globulin serum was treated with 75 mg. of lyophilized agammaglobulinemic serum which Studies performed with Oudin (18), double diffusion, had been found to contain 2.5 mg. per cent of gamma and Ouchterlony (18) methods revealed a single band of globulin, incubated for one hour at 370 C. and then stored precipitate at equivalence when the antiserum was tested at 0 to 40 C. for eight days, with daily agitation. Sterile against the standard human gamma globulin antigen, procedure was used throughout. The treated antiserum normal human sera and agammaglobulinemic sera. was then centrifuged at 3,000 rpm for one hour in the cold and decanted. Seventy-five ml. of this once ab- RESULTS AND DISCUSSION sorbed antiserum was again absorbed by the same technique. Finally, aliquots of the unabsorbed, once and Summarized in Table I are the gamma globulin twice absorbed antisera were decomplemented with concentrations as determined immunochemically crystalline BSA-rabbit-anti-BSA.' The latter reagents were added in precise proportions corresponding to their on normal infants, children and adults. The dis- ,equivalence point as determined by previous quantita- tribution of gamma globulin levels is also shown tive precipitin studies. Separate quantitative precipitin as the stippled area on Figure 3. determinations were then performed using each of the Table II describes the quantitative immunologi- antisera with dilutions of the standard gamma globulin cal deficit in 16 agammaglobulinemic patients. All solution. Figure 1 demonstrates the close agreement among agammaglobulinemic patients have some gamma the amounts of gamma globulin precipitated in the zone globulin as measured in this way; the serum of antibody excess with the three antisera, and serves gamma globulin concentrations in patients with to demonstrate that even the unabsorbed antiserum pos- congenital agammaglobulinemia ranged from 2.5 sessed sufficient immunological homogeneity for use in to 16.0 mg. per cent while those of patients with determinations of gamma globulin concentrations. As may be seen from the figure, no decrease in the amount acquired agammaglobulinemia ranged from 12.0 to of precipitate occurring in the zone of antibody excess 97.0 mg. per cent. Although some overlap between -took place when absorbed serum was used. These find- the two groups was observed, it will be seen that ings also suggest that the gamma globulin preparation the adult agammaglobulinemics generally possess used as antigen did not contain other protein constituents higher concentrations of gamma globulin in their in significant amounts. Free electrophoresis was carried out in a Klett Model serum than do the children with the congenital E Tiselius apparatus at 4° C. against veronal buffer, pH form of the disease. 8.6, ionic strength 0.1; the gamma globulin used as anti- Figure 2 summarizes the data obtained in the gen was found to be symmetrically monophoretic. Paper immunochemical estimation of the biologic half- electrophoresis was carried out in a Spinco Model R series life of parenterally administered gamma globulin B paper electrophoresis apparatus, and the patterns ob- In this -tained in the Spinco Analytrol. This study, too, re- in patients with agammaglobulinemia. vealed that the gamma globulin fraction employed as anti- study, serum gamma globulin concentrations were gen was homogeneous. plotted as a semilogarithmic function of time fol- 1 Armour and Co., bovine plasma albumin, crystal- lowing the injection of various doses of gamma lized. globulin. These patients ranged in age from one 582 SOLOMON J. ZAK AND ROBERT A. GOOD

TABLE 11 900 Immunochemical estimations of serum gamma globulin c 800 concentrations in patients with agammaglobulinemia 3 700 Gamma 0 Form of disease, globulin °'600 Subject Sex Age concentration 9 500 yrs. mg. % E Congenital a`4000 agammaglobulinemia E T.A. M 1 14.0 J.S. M 2 2.5 W.A. M 6 4.4 B. H. M 6 10.0 C E.S. M 7 11.5 R. J. M 3 16.5 0 R. D. M 3 13.4 ._ .O- S. E. M 11I 8.8 E J. Si. M 1 10.0 0 Mean: 10.1 .2 W.A. I Range: 2.5-16.0 33.5 Tk-: Acquired agammaglobulinemia L. L. F 30 12.0 20 40 60 F. H. M 58 40.0 Days after injection of gamma globulin K. N. M 35 97.0 A. L. M 20 65.5 FIG. 2. BIOLOGIC HALF-LIFE OF HUMAN GAMMA GLOBU- L. Z. F 56 30.0 LIN IN PATIENTS WITH AGAMMAGLOBULINEMIA M. O'B. F 63 19.0 B. A. F 67 74.0 Mean: 48.2 to seven years. The biologic half-life of gamma Range: 12.0-97.0 globulin in these patients ranged from 30 to 35 days and appeared to be independent of age, con- protein was nearly 200 mg. per cent, antibodies centration of the protein, or time after injection. to H antigen were first demonstrated in signifi- Figure 3 describes the appearance of gamma cant titer. Thereafter, the baby formed agglutin- globulin in the normal female offspring of a pa- ins to 0 and B antigens, formed diphtheria anti- tient with acquired agammaglobulinemia, and its toxin, and became Schick negative at four months decline in the agammaglobulinemic male offspring of age. It is of interest that appreciable quan- of an apparently normal female. In the first in- tities of gamma globulin were present for a con- stance, the gamma globulin concentration of the siderable period preceding the appearance of dem- maternal blood did not differ significantly from onstrable levels of circulating antibodies.2 that of the cord blood, nor were the isohemag- The decline of maternally transmitted gamma glutinins or antibodies to any of the antigens with globulin in the agammaglobulinemic son of a nor- which the mother had been stimulated observed in mal female emphasizes several points of interest. the baby's serum: Antigenic stimulation was pro- First, this child who cannot form gamma globulins vided the baby by weekly injections of typhoid- begins life with at least a normal concentration of paratyphoid vaccine, and diptheria, pertussis and this protein in the serum. The logarithmic de- tetanus vaccine during the first two months and cline of gamma globulin in this baby was uni- monthly thereafter. The baby was born agam- formly linear throughout the first 10 months of maglobulinemic and remained so throughout the 2 Since this study has been completed, another female first 40 days of life. On approximately the for- child was born to the same agammaglobulinemic mother. tieth day of life, the serum gamma globulin con- Similar studies of gamma globulin concentration re- centration was found to be 115 mg. per cent and vealed a curve almost identical to that of the baby de- it rose rapidly thereafter. Eighteen days follow- scribed here except that the initial increment in circulating gamma globulin concentration occurred between 20 ing the first increase in gamma globulin concentra- and 28 days after birth rather than between 35 and 42 tion and at a time when the concentration of this days as was the case with this baby. IMMUNOLOGICAL STUDIES OF HUMAN SERUM GAMMA GLOBULINS 583 life, at which time the serum concentrations fell I0 II I I I within the range observed among other agam- 9 B. E,- NORMAL NEWBORN 8 B. H.- 6 YR. OLD AGAMMAGLOBULIN - maglobulinemic children. The half-life of gamma c EMIC PATIENT globulin in this child throughout the first 10 7 to months of life is comparable to that seen in agam- 6 maglobulinemic patients given gamma globulin 0 B.E. r/2 ' 38.3 days parenterally. S In Figure 4 are shown semilogarithmic plots of 21 4 the decline of gamma globulin concentration of a to normal offspring of a patient who had previ- 00 ously borne a son found to be agammaglobuli- E 3 nemic. In this graph, the decline of gamma globu- to lin concentration is compared to the decline ob- B.H. r/2' 37 days served following injection of gamma globulin in 2 I a six year old agammaglobulinemic boy. During 0 20 40 60 80 100 the first 80 days of life, the curve in this patient Days after birth (BE.) or injection of Gamma Globulin (B.H.) half-life was similar in slope to that defining the FIG. 4. BIOLOGIC SURVIVAL TIMES OF GAMMA GLOBU- of injected gamma globulin in the agammaglobuli- LIN IN A CONGENITAL AGAMMAGLOBULINEMIC PATIENT nemic patients and virtually identical to that de- COMPARED WITH DECLINE IN A NORMAL NEONATE fining the decline in gamma globulin concentration in an agammaglobulinemic baby born of an ap- cant. These observations, although consistent parently normal mother (Figure 3). with their interpretations, provide little support Table III summarizes the serum values on 31 for the conclusions of Moore, Martin du Pan and maternal blood-cord blood pairs taken at parturi- Buxton, and others who suggest that active trans- tion. The differences between the two groups of port and/or local placental production of gamma sera with respect to total protein concentra- globulins are responsible for elevated cord levels tion and gamma globulin concentration estimated noted with electrophoretic methods (19). both electrophoretically and immunologically were small. Although the gamma globulin con- DISCUSSION the immunochemical centrations determined by The immunoelectrophoretic observations of Gra- method were slightly higher in the cord serum the difference be- bar and Williams (20, 21) indicate that human than in the maternal serum, gamma globulins may be immunologically homo- was not tween the two groups statistically signifi- geneous and physico-chemically heterogeneous.

1765.6 The immunological studies in the agar precipitin Normal offspring of agammaglobulinemic a°>406.3mOm.otther systems performed here support this contention. .~i4O6.3 m.m-o Agommaglabulinemic baby of 0 q. onomal mother EE NcrMnal humon range Furthermore, the absence of significant amounts c 1200 Anfibody to H antigen .2 "B of serum proteins other than gamma globulins Iso-agglutinin titer 0 I 1000 X from the material used as the antigen in these 0 '1 studies, as well as the specificity of the antibodies C 800 1.1280 produced in the rabbit against this human gamma .0 globulin, is suggested by the absorption-precipitin 6600 E study. E 400 1:.640 Previous studies comparing results of the gamma globulin levels obtained with the immunochemi- LO 200 1.;:320 cal method and gamma globulin levels obtained by I I t {i IV I*1 1.80 electrophoretic analyses have revealed that higher B 40 80 120160 200 240 280 320 Days after 13irth concentrations are obtained with the former tech- FIG. 3. APPEARANCE OF GAMMA GLOBULIN AND ANTI- nique. This finding has been variously attributed BODY SYNTHESIS IN THE NEWBORN PERIOD to the presence of gamma-reactive globulin under 584 SOLOMON J. ZAK- AND ROBERT A. 6OOD

TABLE III Serum values on 31 maternal-cord samples taken at delivery

Electrophoretic Immunochemical Total serum protein gamma globulin gamma globulin Maternal Cord Maternal Cord Maternal Cord Gm. % mg. % mg. % Mean 6.965 6.529 940.4 1139.7 978.7 1081.8 S. D. 0.81 1.28 264 320.8 296.4 285.9 Range 5.6- 4.85- 580- 510- 509.4- 645.3- 9.3 10.2 1826 2070 1762.5 2187.5 p Value 0.1096 0.0076 0.1645 the alpha and beta globulin peaks of the electro- and salt precipitation techniques, respectively phoretic spectrum. When paper electrophoretic (23-25). methods are used, albumin tailing, higher affinity The appearance in the newborn of apparently of the gamma globulins for the dyes used, and the nonimmune, gamma-reactive globulin in appreci- application artifact introduce errors which pro- able quantities well before the appearance of hibit strict quantitation. The data presented here, specific serum antibody might suggest that only however, in agreement with those of Goodman some gamma globulin is antibody. However, the and co-workers (22), suggest that immunochem- existence of this apparently nonimmune globulin ical and electrophoretic values are of the same may well represent a response to antigens which order of magnitude within the usual ranges of are ubiquitous and for which testing was not ac- concentrations observed in human sera. The im- complished. The failure of this child to syn- munochemical method, being more sensitive and thesize gamma-reactive globulin during the first more precise, affords clearer insight into subtle 40 days of life is also apparent from these data. differences in protein concentrations, and is par- Further data collected in this laboratory on a ticularly valuable in assessing very low concen- similar case revealed initial appearance of serum trations of protein. gamma globulin between 21 and 28 days of life. Among the sera of patients suffering from the The immunochemical studies reported herein severe handicap in protein synthesis, agamma- are in essential agreement with- other studies per- globulinemia, we have observed no instance in formed using different methods to reveal the which gamma globulin was not detectable by pre- variation in serum gamma globulin concentrations cipitin methods in liquid-liquid or in semisolid throughout infancy and childhood. Normal in- media. The finding of extreme hypogammaglobu- fants display a variable interval, lasting between linemia rather than complete absence of the pro- two and six months, prior to the assumption of tein in the sera of these patients by immunochemi- full immunological responsibility. During this cal means was reported upon in the original de- interval, the gamma globulin concentrations may scription of the disease. Our observations sup- decline normally to values slightly above 200 mg. port this conclusion and indicate that neither im- per cent. Certainly most children show no ap- purities in antigens used nor cross reactivity parent ill effects of the immunological inade- among the serum proteins account for the small quacy or the low gamma globulin levels. Failure amounts of gamma globulin measured in each to assume active gamma globulin synthesis, or case. delay in development of this function, is followed The mean biologic half-life of parenterally ad- by further decline in gamma globulin concentra- ministered gamma globulin, established by the tion. This state has been referred to as the tran- im-munochemical method, reveals a value which is sient hypogammaglobulinemia of infancy and is comparable to that reported by Bruton, Apt, characterized clinically, just as is clinical agam- Gitlin and Janeway (11) in agammaglobuli- maglobulinemia, by extreme susceptibility to bac- nemics by a similar technique, and by others in terial infection. Similarly, a child subsequently normal humans using immunologic electrophoretic proved to be agammaglobulinemic showed con- IMMUNOLOGICAL STUDIES OF HUMAN SERUM GAMMA GLOBULINS 585 tinued decline in gamma globulin concentration globulinemic patients was generously provided by Dr. at an almost identical rate to that observed early Sam T. Gibson of the American Red Cross. in the neonatal period in children subsequently shown to be immunologically normal. In this in- REFERENCES stance, however, the decline continued until levels 1. Heidelberger, M., and Kendall, F. E. A quantitative under 10 mg. per cent were reached. This child theory of the precipitin reaction. III. The reaction between crystalline egg albumin and its developed increased susceptibility to infection homologous antibody. J. exp. Med. 1935, 62, 697. when the gamma globulin level descended below 2. Kendall, F. E. Studies on serum proteins. I. Iden- 100 mg. per cent (Figure 3). From these and tification of a single serum globulin by immunologi- other observations, it would appear that the criti- cal means. Its distribution in the sera of normal cal level of serum gamma globulin concentration individuals and of patients with cirrhosis of the liver and with chronic glomerulonephritis. J. is of the order of 100 mg. per cent. clin. Invest. 1937, 16, 921. 3. Kendall, F. E. The use of immuno-chemical meth- SUMMARY ods for the identification and determination of human serum proteins. Cold Spr. Harb. Symp. Procedures for the quantitative estimation of quant. Biol. 1938, 6, 376. gamma globulins in human sera used in this labo- 4. Luck, J. M. On the turnover of plasma proteins. ratory are presented. Data suggesting the ade- Amer. J. Med. 1956, 20, 317. 5. Slater, R. J. The serum gamma globulins defined by quate homogeneity of a gamma globulin prepara- electrophoretic and immunologic analyses. Arch. tion for use in immunochemical estimations are Biochem. 1955, 59, 33. presented. Normal values for the gamma globu- 6. Cohn, M., Deutsch, HI F., and Wetter, L. R. Bio- lin concentration in newborns, infants, children physical studies on blood plasma proteins. XII. and adults, as determined by immunochemical Analysis of immunological heterogeneity of human gamma globulin fractions. J. Immunol. 1950, methods, are presented and compared with electro- 64, 381. phoretic values. Gamma globulin concentrations 7. Deutsch, H. F. The immunochemical reactions of in 16 patients with agammaglobulinemia are human sera with gammar-globulin antisera. Arch. shown. The half-life of injected gamma globulin Biochem. 1954, 49, 356. in four congenitally agammaglobulinemic pa- 8. Kunkel, H. G., Slater, R. J., and Good, R. A. Re- as determined the lation between certain myeloma proteins and nor- tients, by immunochemical mal gamma globulin. Proc. Soc. exp. Biol. (N. Y.) method, is documented. The assumption of 1951, 76, 190. gamma globulin synthesis and immunological re- 9. Cann, J. R., Brown, R. A., Kirkwood, J. G., and sponsibility in the unique instance of a normal off- Hink, J. H., Jr. Fractionation of human immune spring of an agammaglobulinemic patient is de- gamma globulin. J. biol. Chem. 1950, 185, 663. scribed and a 10. Oncley, J. L., Melin, M., Richert, D. A., Cameron, compared with "reciprocal" in- J. W., and Gross, P. M., Jr. The separation of stance of an agammaglobulinemic offspring of an the antibodies, isoagglutinins, prothrombin, plas- immunologically normal female. The rate of de- minogen and beta1-lipoprotein into subfractions of cline of gamma globulin concentration during the human plasma. J. Amer. chem. Soc. 1949, 71, 541. first two and one-half months of life observed in 11. Bruton, 0. C., Apt. L., Gitlin, D., and Janeway, C. A. an immunologically normal child was found to be Absence of serum gamma globulins. A. M. A. J. Dis. Child. 1952, 84, 632. similar to that of parenterally injected gamma 12. Good, R. A., and Zak, S. J. Disturbances in gamma globulin in agammaglobulinemic patients and simi- globulin synthesis as experiments of nature. lar to the decline of gamma globulin concentration Pediatrics 1956, 18, 109. observed during the neonatal period in an agam- 13. Prasad, A. S., and Koza, D. W. Agammaglobuline- maglobulinemic child. mia. Ann. intern. Med. 1954, 41, 629. 14. Good, R. A., Varco, R. L., Aust, J. B., and Zak, S. J. Transplantation studies in patients with agam- ACKNOWLEDGMENTS maglobulinemia. Ann. N. Y. Acad. Sci. 1957, 64, 882. The authors wish to acknowledge the technical con- 15. Weichselbaum, T. E. An accurate and rapid method tributions of Mrs. Noorbibi Day and Mr. Arthur Loerzel. for the determination of proteins in small amounts The gamma globulin used for the half-life studies and of blood serum and plasma. Amer. J. clin. Path., subsequently in the clinical management of the agamma- Tech. Sect. 1946, 10, 40. 586 SOLOMON J. ZAK AND ROBERT A. GOOD

16. Kabat, E. A., and Mayer, M. M. Experimental Im- 21. Williams, C. A., Jr., and Grabar, P. Immunoelectro- munochemistry. Springfield, Ill., C. C Thomas, phoretic studies on serum proteins. III. Human 1948. gamma globulin. J. Immunol. 1955, 74, 404. 17. Jager, B. V., Smith, E. L., Nickerson, M., and 22. Goodman, M., Ramsey, D. S., Simpson, W. L., Remp, Brown, D. M. Immunological and electrophoretic D. G., Basinski, D. H., and Brennan, M. J. The studies on human gamma globulins. J. biol. Chem. use of chicken antiserum for the rapid determina- 1948, 176, 1177. tion of plasma protein components. I. The assay 18. Oudin, J. Specific precipitation in gels and its ap- of human serum albumin and gamma globulin. plication to immunochemical analysis in Methods J. Lab. clin. Med. 1957, 49, 151. in Medical Research 5, A. C. Corcoran, Ed. Chi- 23. Wiener, A. S. The half-life of passively acquired cago, Year Book Publishers, 1952, p. 335. antibody globulin molecules in infants. J. exp. 19. Moore, D. H., Martin du Pan, R., and Buxton, C. L. Med. 1951, 94, 213. An electrophoretic study of maternal, fetal and 24. Barr, M., Glenny, A. T., and Randall, K. J. Con- infant sera. Amer. J. Obstet. Gynec. 1949, 57, 312. centration of diphtheria antitoxin in cord blood 20. Grabar, P., and Williams, C. A., Jr. Methode im- and rate of loss in babies. Lancet 1949, 2, 324. muno-electrophoretique d'analyse de melanges de 25. Orlandini, T. O., Sass-Kortsak, A., and Ebbs, J. H. substances antigeniques. Biochim. Biophys. Acta Serum gamma globulin levels in normal infants. 1955, 17, 67. Pediatrics 1955, 16, 575.