INFECTION AND IMMUNITY, May 1985, p. 474-4790019-9567/85/050474-06$02.00/0Copyright C) 1985, American Society for Microbiology

Vol. 48, No. 2

Antibody Response to Phenolic Glycolipid I in Inbred MiceImmunized with Mycobacterium leprae

CORY TEUSCHER,'t DONNA YANAGIHARA,2 PATRICK J. BRENNAN,2 FREDERICK T. KOSTER,' AND

KENNETH S. K. TUNG'*Departments of Pathology and Medicine, The University ofNew Mexico, Albuquerque, New Mexico 87131,1 and

Department of Microbiology, Colorado State University, Fort Collins, Colorado 805232

Received 19 October 1984/Accepted 23 January 1985

The level of circulating antibody to phenolic glycolipid I of Mycobacterium leprae was determined in 18inbred strains of mice after immunization with M. leprae organisms. By usinlg a solid-phase radioimmunoassaywith phenolic glycolipid I as test antigen, a continuous distribution of antibody levels ranging from high to lowwas observed. The level was found to be controlled by multiple genes, including both H-2 complex- and Ighallotype complex-linked genes. Low antibody response to phenolic glycolipid I was shown to be inherited as a

dominant trait in three combinations of high x low responder F1 progeny.

Mycobacterium leprae, the causative agent of leprosy, is aremarkably nontoxic intracellular parasite capable of reach-ing extremely high numbers in various tissues without re-sulting in damage or inciting an inflammatory response (28,30). In human infection, polar tuberculoid leprosy, thehigh-resistance form, is characterized by few lesions, lownumbers of bacilli, high lymphocyte transformation tests,and low titers of anti-M. leprae antibodies (11, 12, 24, 40). Atthe other end of the clinical spectrum, polar lepromatousleprosy, the low-resistance form, is exemplified by largenumbers of disseminated bacilli, reduced reactions in lym-phocyte transformation tests, and high levels of anti-M.leprae antibodies. Between the two ends of this spectrum isa wide range of clinical-pathological symptoms with variouscutaneous and immunological manifestations known as bor-derline leprosy. The results of host immune reactions tobacillary antigens (30), rather than variations in the M.leprae organism itself (27, 32), are believed to be responsiblefor most of the clinical-pathological spectrum of the disease.Furthermore, the expression of the clinical spectrum may bedue in part to human major histocompatibility complex(HLA) and non-HLA controlled host immune responses toM. leprae (31, 38).

Studies examining immune responses to whole M. lepraeand M. Ieprae antigens in humans may not discriminatebetween naive responses and altered responses secondary tothe infectious process. On the other hand, studies in nonin-fected inbred mice, with well-defined genetic compositions,allow meaningful analysis of the host genetic factors in-volved in governing immune responsiveness to antigens ofM. leprae. Comparisons between the magnitude of geneti-cally controlled humoral and cellular immune responses indifferent infected and uninfected mouse strains also providedata pertinent to the hypothesis that the different clinical-pathological expressions of leprosy result from differences inthe magnitude of the two types of immune response, aconsequence of imbalanced T-cell immunoregulation (21-23,36).

* Corresponding author.t Present address: Division of Reproductive Biology and Endo-

crinology, Department of Obstetrics and Gynecology, University ofPennsylvania, Philadelphia, PA 19104.

An M. leprae-specific glycolipid antigen, triglycosylphenolic diacyl phthiocerol (phenolic glycolipid I [PGL-I]),has recently been isolated, purified, and structurally charac-terized (13, 14). It is distinguished by a species-specifictrisaccharide [3,6-di-O-methyl-3-D-glucopyranosyl(1->4)2,3-di-O-methyl-o-L-rhamnopyranosil (1--2) 3 -0- methyl-a-L-rhamnopyranose] glycosidically linked to a genus-specificphenolic diacylphthiocerol. PGL-I has been shown to beantigenic in animals (1, 13, 14) and leprosy patients (2, 4, 14,25, 43). By using solid-phase immunoassays, immunoglobu-lin G (IgG) and IgM antibodies to PGL-I were demonstratedin the majority of untreated leprosy patients of all clinical-pathological classifications. Patients with lepromatous lep-rosy had higher levels of anti-PGL-I antibodies than didpatients with tuberculoid leprosy (2, 4, 43). PGL-I was alsofound to induce nonspecific suppressor T-cell function invitro (21).PGL-I represents ca. 2% of the organism mass and is

present in large quantities in infected tissues from which thebacilli have been removed (13). Since it may represent themajor interface between host defense mechanisms and themicroorganisms, we have undertaken a series of studies toexamine the genetic control of immune responses to PGL-Iin mice immunized with M. leprae organisms.

MATERIALS AND METHODS

Mice. The following strains of inbred female mice andF, hybrids were used: A/J, A.SW/SnJ, B10.BRISgSnJ,BALB/cByJ, CBA/J, C57BL/6J, C57BL/10J, DBA/1J,DBA/2J, SJL/J, SWR/J, (C57BL/6J x A/J)Fl, (SWR/J xA/J)Fl, (SJL/J x CBA/J)F1 (Jackson Laboratory, Bar Har-bor, Maine) and BALB.10, BALB.K, BALB.Igc (N-20),BALB.1ge (N-10), BALB.Igf (N-18), BALB.Igg (N-11,F-3),BALB.Ig" (N-11) (Noel Warner and Ed Walker, Universityof New Mexico School of Medicine, Albuquerque, N.M.).All mice were 8 to 20 weeks of age at the start of theexperiment.

Immunization. M. leprae was isolated from irradiatedinfected armadillo livers (13). Purified M. Ieprae was emul-sified in incomplete Freund adjuvant (Marcol:Arlacel, 4:1).Primary and secondary immunizations with 0.1 ml of emul-sion containing 50 p.g of M. leprae were carried out intraperi-

474

ANTIBODY RESPONSE TO M. LEPRAE PHENOLIC GLYCOLIPID I

H-2 IgStrain Type Allotype

SWR/J q/q c /c

No.Animals

6

C57BL/6J b/b b/b 7

CBA/JBALB.KDBA/IJ

C57BL/IOJ

BIO.BR/SgSnJ

k/k a/a 7

k/k

q/qb/bk/k

a/a

c/c

b/b

b/b

5

5

7

7

A/J a/a e/e 5

A.SW/SnJ s/s e/e 6

DBA/2J d/d c/c 6

BALB/cByJ d/d a/a 9SJL/J s/s b/b 6

BALB. BIO b/b a/a 7

Acpm I125 SaMk (x103)

2 3 4 5

FIG. 1. Levels of circulating anti-PGL-I antibodies detected by a solid-phase radioimmunobinding assay with 1251I-SaMK in differentinbred strains of mice immunized with 100 ,u.g of nonviable M. leprae organisms. A continuous distribution in the level of circulatinganti-PGL-I antibodies was observed. Mouse strains are arbitrarily grouped into high, intermediate, and low responders. Data expressed asmeans + standard error.

toneally on day 0 and day 21. Serum was obtained by tailbleeding 2 weeks before and 30 days after primary immuni-zation. Sera were stored at -70°C until tested.

Solid-phase radioimmunoassay for anti-PGL-I antibodies.The solid-phase radioimmunoassay used to detect antibodiesto PGL-I in the serum of immunized mice was essentiallythat of Smolarsky (35). Briefly, 50-plI samples of PGL-Isuspended in 100% ethanol at a concentration of 0.04 mg/mlwere coated onto flat-bottomed 96-well polyvinyl microtiterplates (Dynatech Laboratories, Inc., Alexandria, Va.) bydrying. The plates were washed three times, 5 min eachtime, with 150 plI of 0.3% gelatin (gel) dissolved in phos-phate-buffered saline (PBS; pH 7.2) containing 1 mM EDTA(gel-EDTA-PBS). Samples of 50 pA of preimmune and day-30postimmunization sera diluted 1:5 in gel-EDTA-PBS wereadded to duplicate wells and incubated at room temperature

Strain

C57BL/IOJBIO. BR/SgSnJ

H-2Iypeb/b

k/k

No.Animals

7

7

BALB/cByJ d/d 9

BALB.BIOBALB.K

b/b 7

k/k 5

for 3 h. Plates were then washed rapidly three times withgel-EDTA-PBS, followed by a 5-min wash with gel-EDTA-PBS and a 5-min wash with 1% gamma globulin-free bovineserum albumin (BSA) in PBS (BSA-PBS). An amount of100,000 cpm of 125I-sheep anti-mouse kappa (125I-SaMK) or125I-protein A (125I-PA) diluted in 50 ,ul of BSA-PBS wasadded and incubated for 3 h at room temperature. Each wellwas washed three times, for 5 min each time, with 150 ,ul ofBSA-PBS, cut from the plate with a hot wire, and counted inan a-spectrometer. All sera were also studied in dilutions of1:50 to ensure that the differences observed in uptake of 1251at 1:5 dilutions was not due to a pro-zone effect. The pairedt-test was used to determine the statistical significance ofantibody present in serum on day 30 after immunization. Thespecific antibody response for each mouse represents thedifference in counts per minute obtained by subtracting

Acpm I125 SaMk (xI03)2 3 4 paI I

Value~-A

. II

NS

<0.005

FIG. 2. The levels of circulating anti-PGL-I antibodies detected by a solid-phase radioimmunobinding assay with '251I-SaMK in H-2congenic mouse strains. The H-2k haplotype was associated with a high antibody response on the BALB/c background. NS, Not statisticallysignificant. Data expressed as means + standard error.

475VOL. 48, 1985

476 TEUSCHER ET AL.

Ig No.Strain Allotyp. Animals

BALB/cBy a/a 9

BALB.lgn

BALB.IgcBALB.Ig9

BALB.lgf

BALB Ige

n/n 10

c/c 9

g/g 10f/f 9e/e 9

Acpm I125 SaMk (x103)1 2 3 4 5

FIG. 3. Levels of circulating anti-PGL-I antibodies detected by a solid-phase radioimmunobinding assay with 'l25-SaMK in BALB/c Ighallotype congenic mice. BALB/c mice possessing the Igg, Igf, and Ige allotypes produced significantly higher levels of anti-PGL-I antibodies.

NS, Not statistically significant. Data expressed as means standard error.

preimmune counts per minute from postimmunization countsper minute. The mean ± standard error of the mean of theantibody responses for each mouse strain were calculated,and the data analyzed by the Student t-test for comparisonsbetween strains.

RESULTS

Strain distribution of anti-PGL-I antibody response. Allmouse strains tested produced significant anti-PGL-I anti-bodies (P < 0.05) after two intraperitoneal injections of 50,ug of nonviable M. Ieprae emulsified in incomplete Freundadjuvant (Fig. 1). Similar results were obtained with 1:5dilutions (data not shown). A continuous distribution in thelevel of antibody, indicative of polygenic control (7), wasobserved. Nevertheless, the animals could be arbitrarilygrouped as high-, intermediate-, and low-responder strains(Fig. 1).

Role of H-2 in controlling the anti-PGL-I antibody response.Possible H-2 influence in controlling the level of anti-PGL-Iantibody response was suggested by the finding that micepossessing the H-2q, H-2b, and H-2k haplotypes were gener-ally high to intermediate responders (BALB.B10 mice beingan exception), whereas mice possessing the H-2' and H-2d

haplotypes were low responders (Fig. 1). Additional supportcame from studies with the H-2 congenic resistant strains ofBALB/c (H-2d): BALB.K (H-2k) and BALB.B10 (H-2b) (Fig.2). The H-2k haplotype was associated with a high antibodyresponse when present on the BALB/c background (P <0.005), whereas the H-2b and the H-2d haplotypes were bothlow responders. This was not observed with C57BL/10 H-2congenic mice, however (Fig. 2).

Role of Igh allotype complex-linked genes in anti-PGL-Iantibody response. To examine the influence of Igh allotypecomplex-linked genes on the anti-PGL-I antibody response,BALB/c immunoglobulin allotype congenic mice were stud-ied. BALB/c mice possessing the Igg, Igf, and Ige allotypesdemonstrated significantly higher levels of anti-PGL-I anti-bodies (P < 0.002, < 0.005, and < 0.001, respectively) thanBALB/c mice possessing the Iga, Ign, and Igc allotypes (Fig.3).

Inheritance of anti-PGL-I antibody response. The mode ofinheritance of the anti-PGL-I antibody response was studiedin three sets of F1 progeny of high or intermediate x lowparental phenotypes: (CBA/J x SJL/J)F1, (C57BL/6J xA/J)F1, and (SWR/J x A/J)F1. In all three, the anti-PGL-Iantibody response was below that of the low-responder

Strain

CBA/J

SJL/J

(SJL/J x CBA/J) Fl

SWR/J

A/J

(SWR/J x A/J) Fl

C57BL/6J

A/J

(C57BL/6J x AJ)FI

H-2 Ig

Type Allotyp5k/k a/a

No.Animal

7

s/s b/b 6

s/k b/a 10

q/q c/c 6

a/a e/e 5

q/a c/e I0

b/b b/b 7

a/a e/e 5b/a b/e I0

iLcpm I125 SaMk (xI03)2 3 4 5

FIG. 4. Levels of circulating anti-PGL-I antibodies detected by a solid-phase radioimmunobinding assay with 125I-SaMK in parentalstrains and F1 progeny of high x low responder strains of mice. In all three groups of F1 progeny the anti-PGL-I antibody response was belowthat of the low-responder parental strain, suggesting that low anti-PGL-1 antibody responsiveness is inherited as a dominant genetic trait. Dataexpressed as means ± standard error.

paValue

NS

NS<0.002<0.005<0.001

INFECT. IMMUN.

II

ANTIBODY RESPONSE TO M. LEPRAE PHENOLIC GLYCOLIPID I

H-2 Ig No.Strain Type Allotyp8 Animals

BALB.K k/k a/a 5

C57BL/6J b/b b/b 7

C57BL/IOJ b/b b/b 7

DBA/IJ q/q c/c 5

SWR/J q/q c/c 6

CBA/J k/k a/a 7

BIO.BR/SgSnJ k/k b/b 7

tscpm (103)

2 3 4 5

PA I

SaMk l

_+PA

SaMk +ftPA

[SaMk - I

± PA

ISaMk II

SaMk I|

IPA A

P

DBA/1 strains was composed mainly of non-PA-bindingimmunoglobulins (Fig. 5), whereas PA-binding antibodiesconstituted a significant portion of the total response inSWR, CBA, B1O.BR, and BALB.K strains (Fig. 5). In theother strains, including the F1 progeny, the level of antibodyresponse determined by PA correlated with the level deter-mined by SaMK.We next examined the PA-binding anti-PGL-I antibody

responses in the H-2 and Igh allotype congenic mice men-tioned above. The results again clearly demonstrate that theH-2k haplotype is associated with high responsiveness (Fig.6). Unlike the results from the study on antibodies detectedwith SaMK, the PA-binding antibody response of B1O.BRwas significantly higher than that of C57BL/10, indicatingthat the PA-binding anti-PGL-I antibody response is gov-erned by H-2k haplotype linked genes (Fig. 6). These resultssuggest that the anti-PGL-I antibody response may be iso-type restricted, which in turn is under the control of H-2. Byidentifying the isotype(s) of the anti-PGL-I antibody re-sponse in all mice tested, it should be possible in futurestudies to determine whether or not the anti-PGL-I antibodyresponse is indeed isotype restricted (26).

FIG. 5. Comparison of the levels of circulating anti-PGL-I anti-bodies detected by a solid-phase radioimmunobinding assay with251I-SaMK and 125I-PA in high- and intermediate-responder strainsof mice. SWR, B1O.BR, CBA, and BALB.K mice responded withsignificant levels of PA-binding anti-PGL-I antibodies, whereas theanti-PGL-I antibody response of C57BL/6, C57BL/10, and DBA/1mice consisted mainly of non-PA-binding immunoglobulins. Dataexpressed as means ± standard error.

parental strain (Fig. 4). The results suggest that low anti-PGL-I antibody response is inherited as a dominant genetictrait.

Anti-PGL-I antibody isotype. To study the antibody iso-type of the anti-PGL-I antibody response, sera were exam-ined by using 1251I-PA in place of 1251-SaMK in the solid-phase radioimmunoassay. PA reacts predominantly withmurine IgG2 and to a lesser degree with IgG3 (18); thus,PA-binding antibodies represent mainly antibodies of IgG2or IgG3 isotypes, or both. 1251-PA counts per minute werehigher than 12 I-SaMK counts per minute in only two strainsof mice, BALB.K and CBA/J (Fig. 5). Taking into consid-eration the finding that PA is more sensitive than SaMk indetecting anti-PGL-I antibodies (data not shown), it appearsthat the antibody response of C57BL/6, C57BL/10, and

H-2 No.Strain Type Animals

C57BL/IOJ b/b 7 L

BIO.BR/SgSnJ k/k 7

BALB.BIO b/b 7

BALB.K k/k 5

DISCUSSIONIn this study, the antibody response to an M. leprae-spe-

cific glycolipid antigen, PGL-I, was examined in inbredmouse strains after immunization with nonviable organisms.The results of this study demonstrate a continuous distribu-tion of antibody response, which is presumably linked to thegenetic characteristics of each mouse strain. Similar resultshave been observed in mice immunized with syntheticantigens poly-L-(Tyr, Glu)-10ly-D,L-Ala-poly-L-Lys (20), andterpolymer poly(Glus7 Lys 8 Ala5) (17) with group A strepto-coccal polysaccharide (5), staphylococcal nuclease (16),sperm whale myoglobin (41) and ferritin (42). Continuousdistributions in antibody response may indicate multigenecontrol, with Ir genes operating at the various levels ofcellular interactions (7). An alternative explanation is basedon the cross-tolerance hypothesis (8, 41, 42): a low-re-sponder strain would be the result of self-antigens cross-re-acting with an epitope(s) on PGL-I, whereas a high-re-sponder strain is one with minimal cross-reactivity betweenself-antigens and PGL-I. In this case, the continuous spec-trum of magnitude simply reflects varying degrees of cross-reactivity between self-antigens and PGL-I. A similar modelfor molecular mimicry or antigenic cross-reactivity has beenhypothesized to account for the mechanism by which HLA-

cpm I125 PA (x103)1 2 3 4 5 p

Value

<0.05

FIG. 6. Levels of circulating anti-PGL-I antibodies in H-2 congenic strains of mice detected by a solid-phase radioimmunobinding assaywith 1251I-PA. Higher levels of PA-binding anti-PGL-I antibody are associated with the H-2k haplotype on both the BALB/c and C57BL/10background. Data expressed as means + standard error.

477VOL. 48, 1985

=1 <0.002

i I

478 TEUSCHER ET AL.

DR2 serves as a marker for susceptibility to tuberculoidleprosy (38). An extrapolation of the cross-tolerance hypoth-esis would predict that varying degrees of cross-reactivitybetween HLA and M. Ieprae antigens dictate whether thehost fails to eliminate the infectious agent or proceeds toabnormal immune reactivity (37). However, available infor-mation on the immunochemistry of the antigenic determi-nant of PGL-I would argue against the cross-tolerancemechanism.

Studies with pooled sera from lepromatous leprosy pa-tients and hyperimmune anti-M. Ieprae rabbit antiserum in asolid-phase enzyme-linked immunosorbent assay have led tothe elucidation of the primary epitope of PGL-I (4). Antige-nic activity was associated only with the moieties containingthe trisaccharide entity. Since the essential epitope of PGL-Irequired for antigen antibody interaction was found to residein the terminal, 3,6-di-O-Me-glucopyranosyl residue (9), andsince this dimethyl glycose has not previously been recog-nized in nature, it would seem unlikely that cross-reactivitybetween self-antigens and PGL-I could be responsible forthe continuous distribution of anti-PGL-I antibody re-sponses. Multigene control as an explanation for the spec-trum of anti-PGL-I responses is therefore more plausible.

Since the animals in this study were immunized with thewhole M. leprae organisms, M. leprae antigens other thanPGL-I can function as carrier antigenic determinants andinfluence the anti-PGL-I antibody response. Thus, the strainvariations observed in this study could reflect differentresponses to the carrier determinants in M. leprae ratherthan to PGL-I per se. Although this is expected to lead tomore complex results, the experimental design more closelymimics a study on the control of immune responses to PGL-Iin leprosy patients. In fact, similar studies have been re-ported for immune response to streptococcal group A car-bohydrate antigens after injection of whole organisms (3).

Strain distribution studies suggest an H-2-linked influenceon the level of anti-PGL-I antibody response. In general,mice possessing the H-2q, H-2b, and H-2k haplotypes werehigh to intermediate responders, whereas mice possessingthe H-2S and the H-2d haplotypes were consistently lowresponders. The use of H-2 congenic resistant strains alsosupport the role of H-2-linked genes in controlling theantibody response to PGL-I. However, a discrepancy in theantibody response detectable by SaMk between the BALB/cand the C57BL/10 congenic mice indicate that gene comple-mentation between H-2-linked genes and non-H-2-linkedgenes may be operative.

Control of antibody response has also been shown toinvolve Igh allotype complex linked genes (3, 7, 34). BALB/callotype congenic mice were therefore examined more closelyfor the possible role of Igh-linked genes in controlling thelevel of the anti-PGL-I antibody response. This was done inspite of the lack of apparent correlation between Igh allo-types and anti-PGL-I responses in the strain distributionstudies. The use of BALB/c Igh allotype congenic miceclearly demonstrate significantly higher levels of anti-PGL-Iantibodies in mice with some immunoglobulin allotypes (Igg,Ig , and Ige). In addition, gene complementation betweenimmunoglobulin allotype-linked Ir genes and other genesmay play a role in governing the level of anti-PGL-I antibodyresponse. This is exemplified by the fact that mice possess-ing the Ige allotype are higher responders on the BALB/cbackground than they are on the A/J or A.SW background(compare the BALB.Ige response in Fig. 3 with the A/J andA.SW responses in Fig. 1). Genes linked to the Igh allotypecomplex have also been shown to influence the antibody

responses to group A streptococcal carbohydrates (3), p-ami-nobenzoic acid and sulphanilic acid coupled to bovine gammaglobulin (29), and to sheep erythrocytes (19).The apparent dominant inheritance of low antibody re-

sponse to PGL-I in F1 hybrid progeny of high- and low-re-sponder parental phenotypes has been demonstrated in micewith other antigen systems. Examples include the antibodyresponse to trinitrophenylated-mouse serum albumin (39),mouse IgG2a (15), Ea-1 erythrocyte antigens (10), the ran-dom L-glutamic acid50-L-tyrosine50 polymer (6), and theautoimmune response to the liver antigen F (33). Whendominant nonresponsiveness or low responsiveness is ob-served, the explanation has frequently been that the nonr-esponding strain, as well as the resultant F1 hybrids, possesscross-reactive epitopes with the test antigen. In the case ofPGL-I this seems unlikely for the reasons previously dis-cussed.There are several alternative explanations for dominant

inheritance of low responsiveness, and they are not mutuallyexclusive. The interaction of molecules involved in antigenpresentation between high and low responders in the F1progeny could result in the expression of a low-responderphenotype. High- and low-responder interactions occurringafter antigen processing at the genetic or the cellular level, orboth, could also play a role. Gene products in the F1 hybridsmay be defective at the cellular level, leading to inefficientcellular cooperation, which is involved in the anti-PGL-Iantibody response. Finally, preferential stimulation of sup-pressor activity of the immune response is characteristic oflow-responder phenotypes. Thus, in F1 hybrids, as well aslow-responder parental stains, the low-responder phenotypecould be the expression of genetically controlled suppres-sion of anti-PGL-I immune responses. The fact that in twoout of three F1 progeny (SJL/J x CBA/J) and (SWR/J x A/J)the response was significantly below that of the low-re-sponder parental phenotype (P < 0.05) is suggestive of anactive suppressor effect by genes of the low-responderparent. Future studies in the mouse system, including thedetermination of the number of genes involved in dominantlow responsiveness by backcross analysis, may help toelucidate the potential role of genetic influences on thecomplex immunoregulatory mechanisms involved in lep-rosy.

ACKNOWLEDGMENTSWe thank Noel Warner and Ed Walker for providing the BALB/c

congenic mice and Shirley Hunter for preparing the purified PGL-I.We are particularly grateful to Ellen Goldberg and Craig Spellmanfor helpful discussions during this study and to Joanne Migneault andDanna Richards for their help in preparation of the manuscript.

This work was supported by Public Health Service programproject A121057-01 and grant AI-22628 from the National Institutesof Health. C.T. was the recipient of a Heiser Postdoctoral Fellow-ship.

LITERATURE CITED1. Brennan, P. J., and W. W. Barrow. 1981. Evidence for species-

specific lipid antigens in Mycobacterium leprae. Int. J. Lepr.48:382-387.

2. Brett, S. J., P. Draper, S. N. Payne, and R. J. W. Rees. 1983.Serological activity of a characteristic phenolic glycolipid fromMycobacterium leprae in sera from patients with leprosy andtuberculosis. Clin. Exp. Immunol. 52:271-279.

3. Briles, D. E., R. M. Krause, and J. M. Davie. 1977. Immuneresponse deficiency of BSVS mice. I. Identification of Ir genedifferences between A/J and BSVS mice in the antistreptococcalgroup A carbohydrate response. Immunogenetics 4:381-392.

4. Cho, S. N., D. L. Yanagihara, S. W. Hunter, R. H. Gelber, and

INFECT. IMMUN.

ANTIBODY RESPONSE TO M. LEPRAE PHENOLIC GLYCOLIPID I

P. J. Brennan. 1983. Serological specificity of phenolic glyco-lipid I from Mycobacterium leprae and use of serodiagnosis ofleprosy. Infect. Immun. 41:1077-1083.

5. Cramer, M., and D. G. Braun. 1974. Genetics of restrictedantibodies to streptococcal group polysaccharides in mice. I.

Strain differences of isoelectric focusing spectra of group Ahyperimmune sera. J. Exp. Med. 139:1513-1528.

6. Debre, P., J. A. Kapp. M. E. Dorf, and B. Benacerraf. 1975.Genetic control of specific immune suppression. II. H-2 linkeddominant genetic control of immune suppression by the randomcopolymer L-glutamic acid50-L-tyrosine50(GT). J. Exp. Med.142:1447-1454.

7. Dorf, M. E., E. K. Dunham, J. P. Johnson, and B. Benacerraf.1974. Genetic control of the immune response: the effect ofnon-H-2 linked genes on antibody production. J. Immunol. 112:1329-1336.

8. Ebringer, A., and D. A. L. Davies. 1973. Cross reactivitybetween synthetic (T,G)-A-L and transplantation antigens inCBA mice. Nature (London) New Biol. 241:144-147.

9. Fujiwara, T., S. W. Hunter, S.-N. Cho, G. 0. Aspinall, and P. J.Brennan. 1984. Chemical synthesis and serology of disacchar-ides and trisaccharides of phenolic glycolipid antigens from theleprosy bacillus and preparation of a disaccharide protein con-

jugate for serodiagnosis of leprosy. Infect. Immun. 43:245-252.10. Gasser, D. L. 1969. Genetic control of the immune response in

mice. I. Segregation data and localization to the fifth linkagegroup of a gene affecting antibody production. J. Immunol.103:66-70.

11. Harboe, M. 1982. Significance of antibody studies in leprosy andexperimental models of the disease. Int. J. Lepr. 50:342-350.

12. Harboe, M., 0. Closs, G. Bjune, G. Kronvall, and N. H. Axelsen.1978. Mycobacterium leprae specific antibodies detectable byradioimmunoassay. Scand. J. Immunol. 7:111-120.

13. Hunter, S. W., and P. J. Brennan. 1981. A novel phenolicglycolipid from Mycobacterium leprae possibly involved inimmunogenicity and pathogenicity. J. Bacteriol. 147:728-735.

14. Hunter, S. W., T. Fujiwara, and P. J. Brennan. 1982. Structureand antigenicity of the major specific glycolipid antigen ofMycobacterium leprae. J. Biol. Chem. 257:15072-15078.

15. Kindred, B., and E. Weiler. 1971. Recessive inheritance of rapidanti-allotype antibody production. J. Immunol. 107:389-393.

16. Lozner, E. C., D. H. Sachs, and G. M. Shearer. 1974. Geneticcontrol of the immune response to staphylococcal nuclease. I.

Ir-Nase: control of the antibody response to nuclease by theIr-region of the mouse H-2 complex. J. Exp. Med. 139:1204-1214.

17. Maurer, P. H., and C. F. Merryman. 1974. Genetic control ofimmune responses of inbred mice. Response against terpoly-mers poly(Glu51 Lys38 Ala5). Immunogenetics 1:174-183.

18. Mackenzie, M. R., N. L. Warner, and G. F. Mitchell. 1978. Thebinding of murine immunoglobulins to staphylococcal proteinA. J. Immunol. 120:1493-1496.

19. McCarthy, M. M., and R. W. Dutton. 1975. Humoral response

of mouse spleen cells to two types of sheep erythrocytes. III. Anew VH region marker. J. Immunol. 115:1327-1329.

20. McDevitt, H. O., and A. Chinitz. 1969. Genetic control of the

antibody response: relationship between immune response andhistocompatibility (H-2) type. Science 163:1207-1208.

21. Mehra, V., P. J. Brennan, E. Rada, J. Convit, and B. R. Bloom.

1984. Lymphocyte suppression in leprosy induced by unique M.

leprae glycolipid. Nature (London) 308:194-196.22. Mehra, V., L. H. Mason, J. Fields, and B. R. Bloom. 1979.

Lepromin-induced suppressor cells in patients with leprosy. J.

Immunol. 123:1813-1817.23. Mehra, V., L. H. Mason, W. Rothman, E. Reinherz, S. F.

Schlossman, and B. R. Bloom. 1980. Delineation of a humanT-cell subset responsible for lepromin-induced suppression inleprosy patients. J. Immunol. 125:1183-1188.

24. Myrvang, B., T. Godal, D. S. Ridley, S. S. Froland, and Y. K.Song. 1973. Immune responsiveness to Mycobacterium lepraeand other mycobacterial antigens throughout the clinical andhistopathological spectrum of leprosy. Clin. Exp. Immunol.14:541-553.

25. Payne, S. N., P. Draper, and R. J. W. Rees. 1982. Serologicalactivity of purified glycolipid from Mycobacterium leprae. Int.J. Lepr. 50:220-1.

26. Perlmutter, R. M., D. Hansburg, D. E. Briles, R. A. Nicolotti,and J. M. Davie. 1978. Subclass restriction of murine anti-car-bohydrate antibodies. J. Immunol. 121:566-572.

27. Rees, R. J. W. 1969. New prospects for the study of leprosy inthe laboratory. Bull. W.H.O. 40:785-800.

28. Ridley, D. S. 1978. The pathology of leprosy. Southeast Asian J.Trop. Med. Public Health 9:205-208.

29. Rihova-Skarova, B., and I. Riha. 1974. Genetic regulation of theimmune response to haptens. Ann. Immunol. (Paris) 125C:195-198.

30. Sansonetti, P., and P. H. Lagrange. 1981. The immunology ofleprosy: speculations on the leprosy spectrum. Rev. Infect. Dis.3:422-469.

31. Serjeantson, S. W. 1983. HLA and susceptibility to leprosy.Immunol. Rev. 70:89-112.

32. Shepard, C. C., and D. H. McRae. 1971. Hereditary character-istic that varies among isolates of Mycobacterium lyprae. In-fect. Immun. 3:121-126.

33. Silver, D. M., and D. P. Lane. 1975. Dominant nonresponsive-ness in the induction of autoimmunity to liver-specific F antigen.J. Exp. Med. 142:1455-1461.

34. Smith, S. M., D. B. Ness, J. A. Talcott, and F. C. Grumet. 1977.Genetic control of IgM responses to (T,G)-A-L. H-2 and Ig-1linkage. Immunogenetics 4:221-232.

35. Smolarsky, M. 1980. A simple radioimmunoassay to determinebinding of antibodies to lipid antigens. J. Immunol. Methods38:85-93.

36. Stoner, G. L. 1981. Hypothesis: do phases of immunosuppres-sion during a Mycobacterium leprae infection determine theleprosy spectrum? Lepr. Rev. 52:1-10.

37. Svejgaard, A., P. Platz, L. P. Ryder, L. Staub-Nielsen, and M.Thomsen. 1975. HLA and disease association-a survey. Trans-plant. Rev. 22:3-43.

38. van Eden, W., B. G. Elferink, R. R. P. de Vries, D. L. Leiker,and J. J. van Rood. 1984. Low T lymphocyte responsiveness toMycobacterium leprae antigens in association with HLA-DR3.Clin. Exp. Immunol. 55:140-148.

39. Wicker, L. S., W. J. Urba, and W. H. Hildemann. 1980.Hierarchy of H-2 haplotypes governs inheritance of immuneresponsiveness to TNP-MSA. Immunogenetics 10:235-246.

40. Yoder, L., B. Naafs, M. Harboe, and G. Bjune. 1979. Antibodyactivity against Mycobacterium leprae antigen 7 in leprosy:studies on variation in antibody content throughout the spec-trum and on the effects of DDS treatment and relapse in BTleprosy. Lepr. Rev. 50:113-131.

41. Young, C., and A. Ebringer. 1976. Genetic control of theimmune response to sperm whale myoglobin in inbred mice.Immunogenetics 3:299-304.

42. Young, C. R., N. J. Deacon, A. Ebringer, and D. A. L. Davies.1976. Genetic control of the immune response to ferritin in mice.J. Immunogenet. 3:199-205.

43. Young, D. B., and T. M. Buchanan. 1983. A serological test forleprosy with a glycolipid specific for Mycobacterium leprae.Science 221:1057-1059.

479VOL. 48, 1985