INFECTION AND IMMUNITY, JUlY 1970, p. 29-37Copyright © 1970 American Society for Microbiology

Vol. 2, No. 1Prinzted in U.S.A.

Antibody Responses in Serum, Secretions, and Urineof Man After Parenteral Administration of Vaccines

STITAYA SIRISINHA AND CHINDA CHARUPATANA

Department of Microbiology, Faculty of Science, Mahidol University, Banlgkok 4, Thlailanld

Received for publication 6 April 1970

The nature of antibody activities associated with purified immunoglobulin frac-tions of serum, secretions (whole saliva, parotid secretion, and intestinal secretion),and urine of a volunteer after subcutaneous booster injections with rabies virus,poliovirus, diphtheria and tetanus toxoids, and typhoid-paratyphoid-cholera vac-cines was investigated. The results showed that the pattern of antibody responses inthese fluids differed from one antigen to another. Serum-antibody responses tokilled-bacterial vaccine were associated mainly with the immunoglobulin M (1gM)component, slight activities were detected in the IgG, and only traces of activities, ifany, were found in the IgA. These antibodies were primarily of the secretory IgA typein whole saliva and parotid secretion. Slight activities were also observed in theurinary IgG fraction. Responses to inactivated viral vaccine and toxoids were

almost exclusively associated with the serum IgG component. Some antitoxicactivities to diphtheria and tetanus toxins were noted in a low-molecular-weighturinary immunoglobulin component.

Resistance to certain types of infections of bodysurfaces is associated with antibodies in thesecretions bathing the surfaces rather than withserum antibodies (15, 28, 31). It is generallyaccepted that these antibodies [secretory anti-bodies or secretory immunoglobulin A (IgA) ]differ from and are independent of serum anti-body (17, 21, 22, 27, 31, 32). Although recentobservations suggest that local administration ofantigen may be superior to parenteral immuniza-tion in stimulating a local immune response(9-11, 17, 21-23, 29, 31, 37), appropriate sys-temic administration of antigen can stimulate theproduction of local antibody (3, 25, 29, 31). Thepresent investigation was undertaken to analyzethe nature of immune responses in the serum,whole saliva, parotid secretion, intestinal fluid,and urine of a volunteer who received injectionsof several vaccines by the routine parenteralroute. The results suggest that secretory IgA isprobably synthesized locally and that certaintypes of antigenic preparations are superior toothers in stimulating a local immune responsewhen given systemically.

MATERIALS AND METHODSImmunization of a human. A healthy adult male

volunteer received full courses of vaccination againsttyphoid, paratyphoid, cholera, poliomyelitis, rabies,and tetanus approximately 15 months before thepresent vaccination schedule. He also had been

immunized against diphtheria during childhood. Forthe present investigation, the subject received boosterinjections of the above vaccines subcutaneously bythe following schedule: day 1, rabies (duck embryo;Eli Lilly & Co., Indianapolis, Ind.); day 5, poliovirustypes 1, 2, and 3 (Swiss Serum and Vaccine Institute,Berne, Switzerland); day 9, diphtheria-tetanus toxoid(Connaught Medical Research Laboratories, Toronto,Canada); and day 15, typhoid-paratyphoid A andB-cholera (TAB-cholera; The Government Pharma-ceutical Organization, Bangkok, Thailand).

Collection and fractionation of specimens. Preboosterblood, whole saliva, and urine were obtained 1 to 2weeks before the present series of immunizations. Thepostbooster specimens, i.e., blood, parotid secretion,whole saliva, intestinal secretion, and urine, werecollected 7 to 22 days after the last booster. Thecombined fluids of each specimen were clarified bycentrifugation, kept frozen at -20 C until they werefractionated, and analyzed for antibody activities.Fractionation procedures were done at 4 C unlessindicated otherwise. Purified immunoglobulin frac-tions were sterilized by filtration through a 0.22 j,mmembrane filter (Millipore Corp., Bedford, Mass.).

Procedures for the purification of serum IgG,IgA, and IgM were slightly modified from the methodoriginally described by Vaerman et al. ( 36). Theeuglobulin fraction, which precipitated when serumwas dialyzed in water, was dissolved, applied to aSephadex G-200 column (2.5 by 90 cm), and elutedwith 51% NaCl. The fractions that contained IgM, asdetected by immunodiffusion and immunoelectro-phoresis, were eluted at a position close to the excludedvolume of the column. Appropriate fractions were

29

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

SIRISINHA AND CHARUPATANA

pooled, concentrated by ultrafiltration, and rechro-matographed twice on the same Sephadex G-200column. IgG was isolated and purified from thewater-soluble pseudoglobulin fraction by ion-exchangechromatography on a diethylaminoethyl (DEAE)-cellulose (Whatman DE 52) column equilibrated with0.01 M sodium phosphate buffer (pH 8.0). The IgA-containing fractions, which were eluted from theabove DEAE-cellulose column with 0.1 M phosphatebuffer (pH 8.0), were pooled, dialyzed in water,concentrated, and further purified by the zinc precipi-tation method (36). The final concentration of zincsulfate was 0.025 M instead of the 0.05 M; more IgAwas recovered at the lower zinc concentration.

Parotid secretion, stimulated by lemon-flavoredcandies, was collected from Stensen's duct withLashley cups. Whole saliva was stimulated by chewingparaffin. Fractionation procedures for salivary im-munoglobulins have been described (Sirisinha, ArchOral Biol., in press). Briefly, the proteins were pre-cipitated with 650 g of solid ammonium sulfate perliter of saliva. The precipitate was allowed to standovernight at 4 C, dissolved in 0.85% NaCl, dialyzedfree from sulfate, and, finally, equilibrated with 0.005M phosphate buffer (pH 8.0). The concentrated pro-tein solution wasapplied to a DEAE-cellulose column,previously equilibrated with the above buffer. Theproteins were eluted by stepwise application of pH8.0 buffer of increasing molarity. The small amountof IgG present in whole saliva was eluted with thestarting buffer. Most of the salivary IgA was elutedonly with a buffer with a molarity of 0.04 M or higher(Fig. 1). Appropriate immunolobulin-containingfractions were pooled, concentrated, and furtherpurified by gel-filtration. The partially purified salivaryIgG and IgA pools were applied to Sephadex G-150and G-200 columns, respectively, and the fractionsthat eluted close to the excluded volume of the columnwere pooled, concentrated, and rechromatographedtwice on the same Sephadex column.

Intestinal fluid was collected from the jejunum,proteolytic activity was destroyed, and the fluid wasconcentrated as described by Plaut and Keonil (24).

o0.o5o PO Buffer pN aLo

E 20o

Ck2

The IgG component was isolated and purified by thechromatographic method described for saliva.

Twenty-four hour urine collections were accumu-lated in bottles containing a small amount of toluene.The fractionation and purification of urinary IgG andIgA were essentially the same as described for saliva.In addition, a small amount of a low-molecular-weight component that precipitated with anti-K andanti-X sera was eluted from the DEAE column infractions that also contained IgG (Fig. 2). This low-molecular-weight component, referred to as yL inthis communication (see below), was freed fromdetectable contamination with other urinary com-ponents by gel-filtration on a Sephadex G-150 column(Fig. 3).

Antisera. Rabbits received in the footpad antigenin complete Freund's adjuvant and, 2 weeks later,multiple subcutaneous injections of the same antigenicpreparation. Sera obtained 3 weeks after the lastinjection were appropriately absorbed to render themmonospecific. After absorption, anti-IgG, anti-IgA,and anti-IgM sera were specific for the respectiveimmunoglobulin class, as determined by immuno-diffusion and immunoelectrophoretic analyses. Anti-sera to human type K (anti-K) and type L (anti-X)light-chain determinants were produced by immuniza-tion of rabbits with purified Bence-Jones protein oftype K and type L, respectively. Antiserum to humanserum proteins (anti-NHS) was obtained from rabbitsimmunized with pooled normal human serum; theanti-NHS gave at least 15 bands of precipitation withhuman serum in immunoelectrophoretic analysis.Antiserum specific for both serum IgA and secretoryIgA was collected from animals injected with purifiedsalivary IgA (Sirisinha, Arch. Oral Biol., in press).The serum was absorbed with pure serum IgG untilno precipitate formed upon subsequent additions.Multivalent antiserum specific for salivary com-ponents, including secretory piece, was obtainedfrom an animal immunized with concentrated, pooledwhole human saliva; the antiserum was absorbedwith pooled human serum until no precipitate formedupon subsequent additions of human serum. This

-.020 PO Suffer PN6.0- -0.043 PG.ff.e PN .30-

Effluent Vel.me,ml

FIG. 1. DEAE column chromatography ofconcentrated whole saliva. The fractions taht contained immunoglobu-lins are shaded; IgG was eluted by 0.005 M buffer, and IgA was eluted by 0.04M buffer.

30 INFEC. IMMUN.

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

HUMAN ANTIBODY RESPONSES 31

X m0 SffON pH 8.0- . o-.02M PI0 Suff r pH O..004M P0 b,ffor ,Ha0-0.06M P0 Bwf.

FIG. 2. DEAE columni chromatography ofconcentrated urine. Thle fractions that contained immunoglobulins areshiaded; IgG and yL were elutedby O.OOSM buffer, and IgA was eluted by 0.04 u buffer.

B.D. IgG HSA B-J

1I, 1 t,

200

Effluent volume. ml

400

FIG. 3. Gel filtration of the mixture of uriniary IgGanzd -yL on a Seplhadex G-150 column. IgG eluted inth2e first and yL in the second peak. Arrows indicatethe elutionz position of the markers used to calibratethe columnt (B.D., blue dextran; IgG, immunoglobulinG; HSA, human serum albumin; and B-J, Bence-Jonesproteinfrom a myeloma patient).

antiserum (antisaliva) gave at least eight bands ofprecipitation in the immunoelectrophoretic patternwhen reacted against concentrated saliva.

Passive hemagglutination test. Antibodies to somaticO antigens of Salmonella typhosa, S. para-typhosa (Para A), S. schottmulleri (Para B), andVibrio cholerae (Inaba CRL 22463 and Ogawa CRL465) were determined by passive hemagglutinationreaction, employing the microtitration technique. The

O antigens were prepared for sensitization of sheeperythrocytes (SRBC) as previously described (14).Those immunoglobulin preparations that reacted withunsensitized SRBC were adsorbed with packed,washed SRBC before the titration for antibodyactivities. Equal volumes (0.025 ml) of twofold dilu-tions of samples and of 1% sensitized SRBC suspen-sion were incubated at 37 C for 1 hr before resultswere recorded. To detect traces of antibody activityin immunoglobulin preparations that did not producevisible agglutination, excess specific (for the testimmunoglobulin) rabbit antiserum was added to thereaction mixtures, and after thorough mixing theplates were reincubated.

Vibriocidal titration. The vibriocidal activity to V.cholerae was determined by a microtitration technique(4).

Toxin neutralization tests. Known amounts ofimmunoglobulin fractions were mixed with a cali-brated quantity of toxin and the mixtures wereinjected into suitable animals. A standard antitoxinwas titrated along with each analysis so that the unitsof antitoxin per milligram of immunoglobulin beinganalyzed could be compared and estimated.

In the tetanus-toxin neutralization test, twofolddilutions of either standard tetanus antitoxin (Bur-roughs Wellcome and Co., London, England) or thepurified immunoglobulin samples under investigationwere mixed with equal volumes of a toxin solutioncontaining 10 MLD/ml (The Government Pharma-ceutical Organization). (One MLD is that amount oftetanus toxin which, when injected intramuscularlyinto the base of the tail, will uniformly kill adult malealbino mice within 48 hr after injection.) The mixtureswere allowed to stand at room temperature for 1 hr,and 0.2-ml portions were injected into animals. Theend point was taken as the dilution of antitoxin whichprotected all test mice.

Antitoxic activity against diphtheria toxin wasestimated by intracutaneous injection of 0.1 ml oftoxin-antitoxin mixture into rabbit skin (8). The con-centration of diphtheria toxin (Connaught MedicalResearch Laboratories) was five times the minimaldose which gave a positive reaction within 72 hr after

VOL. 2, 1970

1.0

E

CJ .c0

6

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

SIRISINHA AND CHARUPATANA

injection. Shick-test control, standard diphtheria anti-toxin (Connaught Medical Research Laboratories),and the immunoglobulin preparation being titratedwere also inoculated into each rabbit as controls.

Poliovirus neutralization test. Neutralizing anti-bodies to the three types of poliovirus were determinedin an LLC-MK2 monkey kidney cell suspension,employing 100 TCID5o of virus (26).

Rabies antibody. Antibody to rabies was determinedby the indirect fluorescent-antibody technique (30).

Other methods. The concentration of the variousimmunoglobulin preparations was determined by theradial immunodiffusion technique (18), employingcommercial Immunoplates (Hyland Laboratories, LosAngeles, Calif.). Because serum immunoglobulinstandards were used for all calculations, the valuesobtained for secretory IgA, which has a higher molec-ular weight than serum IgA (31), represented minimalestimates only. The concentration of the low-molecular-weight component isolated from urine(yL) was estimated by its optical density at 280 nm[E28nm = 4.47 (20)]. Immunodiffusion and im-munoelectrophoretic analyses were performed asdescribed (1, 12).

RESULTSIsolation and purification of immunoglobulins

from serum, secretions, and urine. The serum IgGfraction was free from contamination with otherserum proteins that were detectable by immuno-diffusion and immunoelectrophoretic analysesemploying anti-NHS. Although the IgA fractionwas free from detectable IgG or IgM by thesemethods, this immunoglobulin preparation con-tained two nonimmunoglobulin components thatmigrated in the electrophoretic field as a2- and,-globulins. The IgM preparation contained a

trace of material that migrated to the a2 regionin the immunoelectrophoretic pattern.Although all three major classes of immuno-

globulins (IgG, IgA, and IgM) have been de-tected in saliva (31-33), only IgA from parotidsecretion and IgA and IgG from whole salivacould be isolated and purified in sufficient quan-tity for further investigation. The final IgA prep-arations from both of these oral fluids eluted froma Sephadex G-200 column as a single componentwith a characteristic elution position of moleculeswith a molecular weight greater than 200,000.Both preparations gave one band of precipitationin the expected position for secretory IgA in theimmunoelectrophoretic pattern when reactedwith either anti-NHS, antisaliva, or antisecretoryIgA serum. They also gave a reaction of partialidentity with serum IgA when tested with anti-secretory IgA serum in the Ouchterlony analysis,suggesting that serum IgA was antigenically de-ficient with respect to these IgA preparations.These results indicated that the IgA componentisolated from both parotid secretion and whole

saliva had characteristics of secretory immuno-globulin described previously by Tomasi et al.(32). The purified IgG preparation isolated fromwhole saliva was found to be antigenically indis-tinguishable from serum or urinary IgG.Two immunoglobulin preparations were iso-

lated from intestinal secretion. The protein com-ponent that eluted from the DEAE-cellulose col-umn with 0.005 M phosphate buffer (pH 8.0)yielded a reaction of identity with serum IgGwhen reacted with anti-IgG or anti-NHS serum.The other preparation eluted in the 0.06 M frac-tion was a mixture of IgG and IgA. The amountof proteins recovered in this latter fraction wasinsufficient for further purification and character-ization of the IgA component.

Three immunoglobulin fractions were isolatedand purified from postbooster urine. IgG, themain immunoglobulin fraction, gave a single bandof precipitation which showed a reaction of iden-tity with serum IgG in immunodiffusion analysiswith specific anti-IgG serum. The IgA prepara-tion was of the secretory IgA type and gave a re-action of identity with salivary IgA preparationsin the Ouchterlony analysis. A low-molecular-weight immunoglobulin, 'yL, eluted from theDEAE column together with IgG; this componentreacted with anti-light-chain sera but not withspecific anti-IgG, anti-IgA, or anti-IgM. Thiscomponent was freed from detectable contami-nation with IgG by gel-filtration on a SephadexG-150 column. Its elution position from the col-umn coincided with that of the Bence-Jones pro-tein marker (Fig. 3). It is suggested, therefore,that the low-molecular-weight urinary immuno-globulin component was either normal immuno-globulin light chain or a small immunoglobulinfragment containing light-chain determinants as amajor part of its structure.Serum antibodies. Table 1 presents the results

of tests for antibody activity in the three serumimmunoglobulin fractions, IgG, IgA, and IgM.The IgM contained the highest concentration ofserum antibodies to Salmonella 0 antigens, as de-tected by the passive hemagglutination technique.The IgG had only a small amount and the IgAhad but a trace amount, if any, of antibody tothese bacterial antigens. Vibriocidal antibodies toboth strains of V. cholerae were mainly associatedwith the IgM; slight activity was also present inthe IgG. Poliovirus-neutralizing antibody (type 1)was associated with the IgG and the IgA fractionsbut not with the IgM. Antibody to rabies virus,on the other hand, was detected mainly i-n IgMand IgA fractions; the IgG contained only a trace,if any, of such antibody. Antitoxic activities toboth tetanus and diphtheria toxins were almostexclusively in the IgG fraction; trace activities

32 INFEC. IMMUN.

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

HUMAN ANTIBODY RESPONSES

TABLE 1. Anitibody activities of purified serum immunoglobulins before and afterparenteral admintistration of vaccines

Minimal concn (,ug/ml) required for positive reaction

Method ofAntigen tested antibody IgG A Ig__

determination' ItgA6(after)

Before After Before After

S. typhosa............... PHA 259 516 440 46 12S. paratyplhosa ........... PHA 64 64 > 880 > 350c 98S. schottmulleri........... PHA 259 129 440 >350 98V. cholerae (Inaba) ...... PHA >2,000 > 16, 500 >880 >350 >380V. cholerae (Ogawa) ... PHA >2,000 > 16, 500 >880 >350 >380V. cholerae (Inaba) V >1,000 129 NT 22 6V. cholerae (Ogawa). V >1,000 258 NT 11 2Poliovirus type I ....... VN NT 20 18 NT >76Rabies virus ......... FAT NT 325 18 NT 8Tetanus toxin ........ TN 0.290d 1.212 0.011 NT 0.029Diphtheria toxin. .TN 0.150 0.480 0.009 NT 0.012

a PHA, passive hemagglutination; V, vibriocidal; VN, virus neutralization; FAT, fluorescent anti-body; TN, toxin neutralization; and NT, not tested.

Prebooster IgA fraction was not available for antibody titrations.No activity detected at concentration indicated.

d International unit of antitoxin per milligram of immunoglobulin tested.

were also detected in the IgA. The small amountof antitoxic activities in the IgM fraction couldbe accounted for by contamination with the IgGcomponent.Comparison of antibody activities in the corre-

sponding immunoglobulin fractions from pre- andpostbooster serum indicated that the increase inserum antibody to the booster injection of killedbacterial vaccine was mainly associated with theincrease of antibodies in the IgM component ofthe serum (Table 1). Some increase of the vibrio-cidal IgG antibody was also observed. Serumantibody responses to toxoid injection were al-most exclusively in the IgG fraction. Since thepostbooster serum IgA and IgM fractions con-tained only trace amounts of antitoxic activities,an increase over those of the prebooster levels,if any, was minimal.

Salivary antibodies. The immunoglobulins ofboth parotid secretion and whole saliva had slightbut definite antibody activity to bacterial antigens(Table 2). These antibodies were primarily of thesecretory IgA type. The IgG concentration of thepurified fraction from whole saliva was too lowfor accurate measurement of antibody activity; areaction of IgG with some of the bacterial anti-gens could, however, be demonstrated by theantiglobulin test. The salivary IgG componentalso had vibriocidal activity for both strains ofV. cholerae. The only salivary immunoglobulinfraction that contained detectable antiviral anti-body was the parotid IgA; this secretory immuno-globulin preparation neutralized poliovirus type 1.

Because the concentrations of the other two sali-vary immunoglobulin fractions were lower thanparotid IgA, the presence of traces of antiviralantibody in these preparations was not excluded.Slight antitoxic activities, to both diphtheria andtetanus toxins, were demonstrated only in theIgG fraction.

Booster immunization promoted a slight, butsignificant increase of antibacterial antibodies inthe secretory IgA fraction of whole saliva. Thevibriocidal and antitoxic activities in the IgGfraction also increased slightly after the booster.

Intestinal antibodies. The only antibody thatcould be demonstrated in the intestinal IgG frac-tion was against V. cholerae; it was detected bythe passive hemagglutination method (Table 2).No vibriocidal antibody to either strain of V.cholerae used could be detected in this same intes-tinal immunaglobulin preparation.

Urinary antibodies. The antibodies of urinewere primarily IgG antibodies (Table 3). Tracesof antibody to bacterial 0 antigens were detectedin the secretory IgA fraction by the antiglobulintest. Neither the passive hemagglutination nor theantiglobulin methods disclosed such antibody ac-tivity in the low-molecular-weight component.No antiviral antibodies were detecteid in the threeimmunoglobulins at the concentrations used. Onthe other hand, low levels of antitoxic activitieswere present in all three urinary immunoglobulinpreparation,. The presence of such activities inthe yL component was unexpected, but the re-sults were reproducible. Adding an equal volume

VOL. 2, 1970 33

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

SIRISINHA AND CHARUPATANA

TABLE 2. Antibody activities of purified immunoglobulins from whole saliva, parotid secretion, andintestinial secretion before and after parenteral administration of vaccines

Antigen tested

S. typhosa..............S. paratyphosa.........S. schottmulleri.........V. cholerae (Inaba)....V. cholerae (Ogawa)...V. cholerae (Inaba)....V. cholerae (Ogawa).Poliovirus type I.......Rabies virus............Tetanus toxin..........Diphtheria toxin.......

Method ofantibody

determination'

PHAPHAPHAPHAPHAVVVNFATTNTN

Minimal concn (Ag/ml) required for positive reaction

Parotid IgA

Before

>140c>140>140>140>140NTNT>28NTNTNT

After

175175175

>310>310NTNT

62>310

00

Whole saliva IgA

Before

>200>200>200>200>200NTNTNTNTNTNT

After

(150)3838

>150>150NTNT>30>150

00

Whole saliva IgG

Before After

NTNTNTNTNT>60>60NTNT

00

(200)d(200)(200)(200)(200)

5050

>40>200

0.052e0.020

IntestinalIgG (after)b

> 200> 200> 200

10025

>100>100>40NTNTNT

a See footnote a, Table 1.b See footnote b, Table 1.c See footnote c, Table 1.d Positive with antiglobulin test at concentration enclosed in parentheses.See footnote d, Table 1.

TABLE 3. Antibody activities of purified urinary immunoglobulins before antdafter parenteral administration of vaccines

Minimal concn (pug/ml) required for positive reaction

Methods ofAntigen tested antibody IgG

determination" IgA yL(after)b (af ter)

Before After

S. typhosa .................. PHA 225 157 (I10)c >1,000S. paratyphosa .............. PHA 56 78 (110) >1,000S. schottmulleri ............. PHA 225 157 (110) > 1,000V. cholerae (Inaba) ......... PHA > 1,800d (1,100) >110 >1,000V. cholerae (Ogawa) ........ PHA > 1, 800 (1,100) >110 > 1,000V. cholerae (Inaba) ...... V 900 312 NT NTV. cholerae (Ogawa) ...... V 225 156 NT NTPoliovirus type I ............ VN NT > 220 >22 > 200Rabies virus ................ FAT NT > 1, 100 >110 > 1,000Tetanus toxin ............... TN NT 0.090e 0 0.030Diphtheraia toxin ........... TN NT 0.018 0.090 0.030

a See footnote a, Table 1.b See footnote b, Table 1.c See footnote d, Table 2.d See footnote c, Table 1.e See footnote d, Table 1.

of specific antisera to either IgG, IgA, or light-chain proteins to this urinary immunoglobulinpreparation failed to inhibit significantly anti-toxic activities associated with this fraction.

DISCUSSIONThe present results show that the antibody re-

sponses in serum, secretions, and excretion from

a human who received parenteral booster immu-nizations differed from one antigen to another(Table 4). The predominant serum antibodiesformed in response to systemic stimulation withkilled-bacterial vaccines resided in the IgM com-ponent, with little activity in the IgG and only atrace, if any, in the IgA (Tables 1 and 4). Al-though on a weight basis IgM contained more

34 INFEC. IMMUN.

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

HUMAN ANTIBODY RESPONSES

TABLE 4. Anttibody activities detected in the various immunoglobulinfractionisfrom serum,whole saliva, parotid secretion, intestinal secretion, and urine

Serum Saliva Urine

Antigen tested IntestineAntigentested Nhole NWhole Parotid IgG

IgG IgA IgMI saliva saliv IgA IgG IgA -yL

S. typhosa.± + +++ + -4i + + + -

S. paratyphosa.+ _ +++ ± + + + -_S. schottmulleri.+ + +++ i + + + i -V. cholerae (Inaba).. + +++ ± _ - 4 - _ +V. cholerae (Ogawa) + - +-+++ 4 4- ± - - +

Poliovirus type I ........ +++ +++ - + - _ _Poliovirus type 2..........Poliovirus type 3 ........Rabies virus ............. + + +++ - _ _ - - - NTa

Tetanus toxin ........... +++ 4 i + + + NTDiphtheria toxin. +++I_+ __ + + - + + NT

a Not tested.

antibacterial antibodies than IgG, the contribu-tion of IgG antibodies to the unfractionatedserum cannot be minimized because IgG is themain immunoglobulin component of serum. Onthe other hand, the serum IgA antibody responsesto these antigenic stimulations were minimal and,because of the low IgA concentration in serum,probably represented an insignificant fraction ofthe total antibody content of the serum. Thiswas in good agreement with previous observa-tions concerning the IgM response in children toS. typhosa vaccine (2).

In contrast to serum antibodies, salivary anti-bodies to bacterial antigens were almost exclu-sively of the secretory IgA type (Tables 2 and 4).Because secretory IgA is the predominant classof immunoglobulin in saliva (6, 31), the anti-bodies of this immunoglobulin fraction contrib-ute significantly to the total antibody content ofthis secretion. It is not unreasonable to expectthat antibacterial antibodies in this location, to-gether with lysozyme and other salivary com-ponents, may effectively serve as the importantfirst line of defense against infections in the oralcavity. The present observation confirms and ex-tends previous findings of the antibody againstS. typhosa in the partially purified IgA fractionsof parotid secretions from normal healthy sub-jects (7).

Urinary antibodies to bacterial antigens werepresent mainly in the IgG fraction rather than inthe secretory IgA; traces of antibodies, however,could be demonstrated in the IgA fraction by theantiglobulin test (Table 3). Since there is moreIgG than secretory IgA present in normal hu-

man urine, these IgG antibodies could play agreater role than secretory IgA antibodies at thislocality. This agrees with previous observations(13) which showed that most of the urinary anti-bodies formed in response to systemic antigenicstimulation resided in the 7S 'y-globulin com-ponent. Turner and Rowe (34), however, foundthe bulk of antibacterial antibody in urine to beassociated with IgA. The difference might be dueto the use of different subjects as well as to the useof different techniques for isolation of immuno-globulins.Antibody responses to systemic stimulations

with protein antigens (diphtheria and tetanustoxoids) were almost exclusively associated withthe IgG fractions of these fluids (Tables 1 to 3).However, on a weight basis, serum IgG appearedto possess more toxin-neutralizing activities thanthe IgG isolated from other fluids. No antitoxicactivities were observed in the secretory IgApreparations, at least at the concentrations used.The possibility exists, however, that some anti-toxic activities might have been detected in thesecretory IgA if more concentrated preparationshad been available, more sensitive tests had beenused, or samples had been obtained from othersubjects. Turner and Rowe (35) demonstratedantibody to tetanus toxoid in the urinary IgAfraction by the passive hemagglutination tech-nique.Although antiviral antibodies were distributed

in the various serum immunoglobulin fractions,with the exception of parotid IgA, no other ex-travascular immunoglobulin preparations pos-sessed antiviral activity. Other investigators pre-

VOL. 2, 1970 35

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

SIRISINHA AND CHARUPATANA

viously reported poliovirus antibodies in the IgAcomponent of intestinal secretion (5, 17, 23),stool (16), urine (5), nasal secretion (23), andsaliva (5).

It is interesting that parenteral immunizationwith killed bacterial vaccine can stimulate theproduction of salivary IgA when no IgA anti-body is detected in serum. This indicates thatsome antigens given parenterally must reach theantibody-producing cells that are involved in thesynthesis of secretory IgA antibodies. Further,these findings suggest that secretory IgA antibodyis synthesized both locally and independent ofserum antibody. Moreover, results obtained fromcomparing antibody activities in the various IgGfractions also suggest that, in addition to thepresence of a local IgA system, there must also bea local immune system for the production of IgGantibody at these sites. Previous observations byothers (21) also suggested that, although a sizableportion of IgG in saliva and nasal fluid appearedto derive from serum, some of these might never-theless be produced and secreted locally, espe-cially after local antigenic stimulation.

Antitoxic activities for both diphtheria andtetanus toxins which were found associated withthe urinary yL fraction were not due to IgG con-tamination because activities remained unchangedafter addition of anti-IgG serum to the prepara-tion. Whether these activities were truly associ-ated with this low-molecular-weight immuno-globulin component or, possibly, with other uri-nary protein(s) which remained undetected couldnot be satisfactorily answered. The quantities ofantisera specific for either type of light-chain pro-teins that could be added to absorb out -yLwithout too much dilution of the preparationfailed to alter any antitoxic activities associatedwith this urinary fraction. Other investigators(13, 19, 20) have also demonstrated the presenceof antibody activities in the low-molecular-weightimmunoglobulin component of human urine afterparenteral administration of vaccines.

ACKNOWLEDGMENTSWe thank P. Z. Allen for Bence-Jones proteins, T. Savanat for

V. cholerae, and H. E. Noyes for salmonellae. We are also grate-ful to R. Snitbhan for the poliovirus-neutralization test, K. Law-haswasdi for the fluorescent-antibody titration of anti-rabies, T.Savanat for vibriocidal titration, and W. D. Sawyer for his valu-able suggestions and criticism during the preparation of the manu-script.

This investigation was supported by a research grant from theDivision of Medical Sciences of the National Research Council,Bangkok, Thailand.

LITERATURE CITED

1. Allen, P. Z., and M. Morrison. 1963. Lactoperoxidase. IV.Immunological analysis of bovine lactoperoxidase prepara-tions obtained by a simplified fractionation procedure. Arch.Biochem. Biophys. 102:106-113.

2. Altemeier, W. A., III, J. A. Bellanti, and E. L. Buescher. 1969.The IgM response of children to Salmonella typhosa vac-cine. I. Measurements of anti-typhoid concentrations andtheir proportions of total serum IgM globulins. J. Immunol.103:917-923.

3. Bellanti, J. A., R. L. Sanga, B. Klutinis, B. Brandt, and M. S.Artenstein. 1969. Antibody responses in serum and nasalsecretions of children immunized with inactivated and at-tenuated measles-virus vaccines. New Engl. J. Med. 280:628-633.

4. Benenson, A. S., A. Saad, and W. H. Mosley. 1968. Serologi-cal studies in cholera. Bull. W.H.O. 38:277-285.

5. Berger, R., E., Ainbender, H. L. Hodes, H. D. Zepp, and M. M.Hevizy. 1967. Demonstration of IgA polioantibody insaliva, duodenal fluid and urine. Nature (London) 214:420-422.

6. Chodirker, W. B., and T. B. Tomasi, Jr. 1963. Gamma-globu-lins. Quantitative relationships in human serum and non-vascular fluids. Science 142:1080-1081.

7. Evans, R. T., and S. E. Mergenhagen. 1965. Occurrence ofnatural antibacterial antibody in human parotid fluid. Proc.Soc. Exp. Biol. Med. 119:815-819.

8. Fraser, D. T. 1931. Technique of method for quantitative de-termination of diphtheria antitoxin by skin test in rabbits.Trans. Royal Soc. Can. 25:175-181.

9. Freter, R. 1965. Coproantibody and oral vaccines, p. 222-228.In Proceedings Cholera Research Symposium. U.S. Depart-ment of Health, Education, and Welfare, Washington,D.C.

10. Fulk, R. V., D. S. Fedson, M. A. Huber, J. R. Fitzpatrick, B.F. Howar, and J. A. Kasel. 1969. Antibody responses inchildren and elderly persons following local or parenteraladministration of an inactivated influenza virus vaccine, A2/Hong Kong/68 variant. J. Immunol. 102:1102-1105.

11. Genco, R. J., and M. A. Taubman. 1969. Secretory 7A anti-bodies induced by local immunization. Nature (London)221:679-681.

12. Grabar, P., and P. Burtin. 1960. L'analyse immuno-electro-phoretique, p. 6-14, Masson et Cie., Paris.

13. Hanson, L. A., and E. M. Tan. 1965. Characterization ofantibodies in human urine. J. Clin. Invest. 44:703-715.

14. Kabat, E. A., and M. M. Mayer. 1967. Experimental im-munochemistry, 2nd ed., p. 119-120. Charles C ThomasPublisher, Springfield, Ill.

15. Kasel, J. A., R. D. Rossen, R. V. Fulk, D. S. Fedson, R. B.Couch, and P. Brown. 1969. Human influenza: aspects of theimmune response to vaccination. Ann. Intern. Med. 71:369-398.

16. Keller, R., and J. E. Dwyer. 1968. Neutralization of poliovirusby IgA coproantibodies. J. Immunol. 101:192-202.

17. Keller, R., J. E. Dwyer, W. Oh, and M. D'Amodio. 1969. In-testinal IgA neutralizing antibodies in newborn infants fol-lowing poliovirus immunization. Pediatrics 43:330-338.

18. Mancini, G., A. 0. Carbonara, and J. F. Heremans. 1965. Im-munochemical quantitation of antigens by single radialimmunodiffusion. Immunochemistry 2:235-254.

19. Merler, E. 1966. The properties of isolated serum and urinaryantibodies to a single antigen. Immunology 10:249-258.

20. Merler, E., J. S. Remington, M. Finland, and D. Gitlin. 1963.Characterization of antibodies in normal human urine. J.Clin. Invest. 42:1340-1352.

21. Newcomb, R. W., K. Ishizaka, and B. L. DeVald. 1969. Hu-man IgG and IgA diphtheria antitoxins in serum, nasalfluids and saliva. J. Immunol. 103:215-224.

22. Ogra, P. L., and D. T. Karzon. 1969. Distribution of polio-virus antibody inserum, nasopharynx and alimentary tractfollowing segmental immunization of lower alimentarytract with poliovaccine. J. Immunol. 102:1423-1430.

23. Ogra, P. L., D. T. Karzon, F. Righthand, and M. MacGillivray.1968. Immunoglobulin response in serum and secretionsafter immunization with live and inactivated poliovaccineand natural infection. New Engl. J. Med. 279:893-900.

36 INFEC. IMMUN.

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from

HUMAN ANTIBODY RESPONSES

24. Plaut, A. G., and P. Keonil. 1969. Immunoglobutins in humansmall intestinal fluid. Gastroenterology 56:522-530.

25. Rossen, R. D., S. M. Wolff, and W. T. Butler. 1967. The anti-body response in nasal washings and serum to S. typhosaendotoxin administered intravenously. J. Immunol. 99:246-254.

26. Schmidt, N. J. 1964. Tissue culture methods and proceduresfor diagnostic virology, p. 78-176. In E. H. Lennette and N.J. Schmidt (ed.), Diagnostic procedures for viral and rickett-sial diseases, 3rd ed. American Public Health Association,Inc., New York.

27. Smith, C. B., J. A. Bellanti, and R. M. Chanock. 1967. Im-munoglobutins in serum and nasal secretions following in-fection with type 1 parainfluenza virus and injection of in-activated vaccines. J. Immunol. 99:133-141.

28. Smith, C. B., R. H. Purcell, and J. A. Bellanti. 1966. Protectiveeffect of antibody to parainfluenza type 1 virus. New Engl.J. Med. 275:1145-1152.

29. Straus, E. K. 1961. Occurrence of antibody in human vaginalmucus. Proc. Soc. Exp. Biol. Med. 106:617-621.

30. Thomas, J. B., R. K. Sikes, and A. S. Ricker. 1963. Evaluationof indirect fluorescent antibody technique for detection ofrabies antibody in human sera. J. Immunol. 91:721-723.

31. Tomasi, T. B., Jr., and J. Bienenstock. 1968. Secretory im-munoglobulins. Advan. Immunol. 9:1-96.

32. Tomasi, T. B., Jr., E. M. Tan, A. Solomon, and R. A. Prender-gast. 1965. Characteristics ofan immune system common tocertain external secretions. J. Exp. Med. 121:101-124.

33. Tourville, D., J. Bienenstock, and T. B. Tomasi, Jr. 1968.Natural antibodies of human serum, saliva, and urine reac-

tive with Escherichia coli. Proc. Soc. Exp. Biol. Med. 128:722-727.

34. Tumer, M. W., and D. S. Rowe. 1964. Characterization ofhuman antibodies to Salmonella typhi by gel-filtration andantigenic analysis. Immunology 7:639-656.

35. Tumer, M. W., and D. S. Rowe. 1967. Antibodies of IgA andIgG class in normal human urine. Immunology 12:689-699.

36. Vaerman, J.-P., J. F. Heremans, and C. Vaerman. 1963.Studies of the immune globulins of human serum. 1. Amethod for the simultaneous isolation of the three immuneglobulins (iySs, y1M and y1A) from individual small serum

samples. J. Immunol. 91:7-10.37. Wigley, F. M., S. H. Wood, and R. H. Waldman. 1969. Aero-

sol immunization of humans with tetanus toxoid. J. Im-munol. 103:1096-1098.

VOL. 2, 1970 37

on July 24, 2019 by guesthttp://iai.asm

.org/D

ownloaded from