1NTERNATIONAL JOURNAL OF LEPROSY^ Volume 63, Number

Printed in the

Identification and Purification ofArmadillo (Dasypus novemcinctus) Immunoglobulins:

Preparation of Specific Antisera to Evaluate theimmune Response in These Animais'Leopoldo Santos-Argumedo, Fernando Guerra-Infante,Fausto Quesada-Pascual, and Sergio Estrada-Parra2

In 1971 Kirchheimer and Storrs accom-plished experimental infection of the nine-banded armadillo (Da.ripus novenicinetusLinn.) with /1/jrobacterium frprae(7. "). Thepath to the study ofleprosy was cleared withthe availability of large quantities of bacillito characterize useful antigens in immu-nodiagnosis and immunotherapy (4. s. 21). Todate, however, there are only a few reportswhich throw light on the host-parasite re-lationship established between armadillosand Al. leprae. This is due to the almosttotal lack of knowledge of the immunologyof these animais (5' '2. 16).

In this work, we purified and character-ized IgM and IgG from armadillos. Mono-specific antiserums were raised in rabbitsand used to evaluate immunological param-eters in healthy and Aí. /eprae-experimen-tally infected armadillos.

MATERIAIS AND METHODSAnimais

Armadillos were captured in the state ofNuevo Leon and in the high part of Mi-choacan state, México, and were kept incaptivity as dcscribed elsewhere (")• After1-3 months of adaptation, 20 ml of bloodwas withdrawn from each animal by cardiacpuncture. Some animais were inoculated

Received for publication on 21 January 1994; ac-cepted for publication in revised form on 11 October1994.

= L. Santos-Argumedo, Ph.D., Department of CellBiology, Centro de Investigacion y Estudios Avanza-dos, I.P.N., Apartado Postal 14-740, Mexico, D.F.07000, Mexico. F. Guerra-Infante, M.Sc.; F. Quesada-Pascual, M.Sc.; S. Estrada-Parra, Ph.D., Departmentof Immunology, Escuda Nacional de Ciencias Biolo-gicas, I.P.N., Prol. Carpio y Plan de Ayala, Mexico,D.E. 11340, Mexico.

with Al. leprae from human lepromas, con-trol animais remained untouched, and onearmadillo was inoculated with bovine se-rum albumin (BSA) (Sigma Chcmical Co.,St. Louis, Missouri, U.S.A.) in completeFreund's adjuvant in order to obtain spe-cific antibodies by allinity chromatography.

Isolation of immunoglohulins from serumIgM. Serum (50 ml) was extensively di-

alyzed against tap water. The content of thedialysis bag, primarily IgM and traces ofIgG, was washed three times with chilleddistiilcd water by centrifugation, and thendissolved in 5 ml of 0.1 M phosphate buf-fered saline (PBS) pH 5.3. IgM was furtherpurified by filtration through SephadexG-200 (Pharmacia) and void volume frac-tions were further characterized (2).

IgG. Serum (50 ml) was precipitated threetimes with ammonium sulfate (33% of sat-uration). The precipitate was extensively di-alyzed against 0.01 M phosphate buffcr pH7.5 and then fractionatcd by DEAE-cellu-lose ion exchange chromatography (Sigma).Nonabsorbed fractions were further char-acterized (2).

Characterization of immunoglobulinsThe amount of protein was quantified by

the method described by Lowry, et al. (i").The proteins were characterized by immu-noelcctrophoresis (2), and electrophoresis inpolyacrylamide gels as was described byLaemmli (8).

Preparation of rabbit antiserum againstarmadillo proteins

Rabbits were inoculated, subcutaneouslyand intramuscularly, with 3-5 mg of ar-madillo whole scrum or purified IgM or IgG

56

63, 1^Santos-Arginnedo, et al.: Artnaddlo I intnunoglobidins^57

--

0.2

1.5^

20hactan

' e2.01

1

102 •

5^1.0^15^20froction

FIG. I. l'urification ofarmadillo immunoglobulins.

A = IgM was precipitated in a low ionic strength so-

lution from whole serum, dissolved in 0.1 M phosphatebuffered saline pH 5.3, and applied orno a Sephadex

G-200 column. The first peak contained only IgM, the

second contained other euglobulins. B = IgG was pre-cipitated from sera using a 33% final concentration of

ammonium sulfate. The precipitate was dissolved in

0.01 M phosphate buffer p11 7.5 and then loaded onto

a DEAE-cellulose column. The only peak containedpure IgG.

in complete Freund's adjuvant. Fifteen clayslater, the doses were repeated. On days 30—33 the doses were repeated again, exceptthat those doses were inoculated withoutadjuvant and, finally, the rabbits were bledon day 39.

Coupling of armadillo immunoglobulins toSepharose 4B

Scpharose 4B (Pharmacia) was cyanogen-bromide-activated as described by John-ston and Thorpc (6), and 15 mg ofarmadilloIgM or IgG per ml of activatcd Sepharosewas coupled. The remaining active moietieswerc blocked with ethanolamine. Immu-noabsorbent was washed and stored in PBS-0.3% sodium azide.

Purification of rabbit antiserumHyperimmune sera against armadillo IgM

and IgG were purified by affinity chroma-tography using the immunosorbents pre-pared as above. Each rabbit antiserum wasfirst absorbed with an heterologous isotype

THE TABLE. Percentage rosette-form-ing and surfàce IgM-bearing lynzphocytes inhealthy arniaddlos.

Armadillo no. Rosette-formingcells

IgM-bearingcells

BOI 10A35 6A73 1 15A77 4A87 3 9A79 24 9A81 7 18A82 11A68 9A9I 16 20A93 17 171502 I^11301 5 241305 15 14Mean (± S.D.),

N = 14 9.6 (± 6.6) 16.7 (± 6.1)

immunoglobulin, thus any crossreactive an-tibody (i.e., anti-heavy or -light chain) wasbound to the column. Then the unboundmaterial was purified through a specific ho-mologous isotype immunoglobulin.

ConjugatesThe monospecific antibodies were la-

beled with Iluorescein isothiocyanate (FITC)(Sigma) (()• Horseradish peroxidase (Sigma)conjugates were also obtained using a two-step glutaraldchyde reaction (20)• Both con-jugates were titrated for optimal dilution inimmunofluorescence (on peripheral bloodarmadillo lymphocytes) (6) or ELISA usingBCG sonicate as antigen (4. 14).

Preparation of armadillo peripheralblood lymphocytes

The mononuclear cells were purified fromperipheral blood using Histopaque 1077(Sigma), and then stained with the FITCconjugatc using standard methods ("), orevaluated for their capacity to form rosetteswith sheep red blood edis (SRBC). Briefly,armadillo lymphocytes (1 x 106) were mixedwith a suspcnsion of SRBC (1 x 108). Themixture was incubatcd for 15 min at 37°C,centrifugated at 400 x g x 3 min and thenincubatcd overnight at 4°C. Enumeration ofrosette-forming edis was done by countingthe lymphocytes with three or more SR BCattached to thcir surface (().

58^ International Journal of Leprosy^ 1995

A A—

B --B

c—2

C 3

DO ^ 4

E

E FIG. 4.^Analysis of rabbit antiserum to IgM andIgG from armadillo. The samples analyzed were: A =armadillo serum; B = armadillo proteins after precip-itation with ammonium sulfate; C = armadillo 1gM;I) = armadillo 1gM and IgG; E = armadillo Iga 1m-munoelectrophoresis was developed with: 1 = rabbitantiserum to armadillo whole serum; 2 = rabbit an-tiserum to armadillo IgG; 3 = rabbit antiserum to ar-madillo 1gM; 4 = rabbit antiserum to armadillo IgMand IgG.

c

— -

FIG. 2. Immunoelectrophoresis of serum proteinsfrom armadillo. The samples analyzed were: A = ar-madillo serum; B = armadillo proteins after precipi-tation with ammonium sulfate; C = fraction puriliedin DEAE-cellulose; I) = first peak from Sephadex G-200column; E = a mixture ofsamples C and D. Ali sampleswere developed with a rabbit antiserum to armadillowhole serum.

RESULTSFigure lA shows the elution proffie of IgM

from the Sephadex G-200 column. The firstpeak corresponds to IgM; in peak two, wefound other euglobulins. Figure 1B showsthe elution profile of IgG from the DEAE-cellulose column. Ali of the fractions fromthe only peak eluted contained pure IgG.

KI) Ko

7055 *Não 4•1. N.O^4.1.

66

-----45---34

28 mileo--24

••••••• 14

I 2 3 4 5 6 789

FIO. 3. Electrophoresis in polyacrylamide of thepurified immunoglobulins from armadillo. A reducing10% SDS-PAGE was loaded with the following sam-ples: 1 = armadillo IgM purified from the SephadexG-200 column; 2 and 5 = human immunoglobulins(mixture of IgM and IgG); 3 = armadillo IgG purifiedfrom DEAE-cellulose column; 4 = armadillo IgG di-gested with papain; 6 = affinity purified anti-BSA an-tibodies from armadillo; 7 = a mixture of samples 1and 3; 8 = armadillo proteins after precipitation withammonium sulfate; 9 = molecular weight markerslactalbumin 14 kDa, trypsinogen 24 kDa, pepsin 34kDa, egg albumin 45 kDa and bovine albumin 66 kDa).

The purity of the fractions from both col-umns was tested by immunoelectrophore-sis, and the results are shown in Figure 2.As can bc observed, the rabbit antiserum(anti-whole serum proteins from the ar-madillo) recognizes between 13-15 anti-gens. However, this antiserum recognizesonly one protein in the first peak from theSephadex G-200 column or in the peak fromthe DEAE-cellulose column as well. In or-der to know the purity of the fractions, theywere analyzed by 10% polyacrylamide gelelectrophoresis in the presence of sodiumdodecyl sulfate and 2-mercaptoethanol. Theresults are shown in Figure 3. IgM is sep-arated in two fractions, one correspondingto the light chains of 28-kDa and one of 70-kDa corresponding to the g heavy chains.Similarly, IgG separates in 28-kDa lightchains and 55-kDa for the heavy -y chains.In the same gel we compared the immu-noglobulins of the armadillos with humangammaglobulins. Armadillo antibodiesagainst BSA, purified by immunoaffinity,also were compared in the same gel. Theresults showed an equivalent molecular mo-bility for ali the proteins.

The purified fractions were used to im-munize rabbits, and the resulting hyper-immune sera were purified by immunoaf-finity using columns of armadillo IgM andarmadillo IgG coupled to Sepharose 4B.Figure 4 shows the monospecific antibodies.

63, I^Santos-Argumedo, et al.: Armaddlo 1111171unoglobulins^59

As can bc seen, crossreactivity was not de-tected.

The lymphocytes bearing surface immu-noglobulin M (sIgM +) were quantified byimmunofluorescence using the rabbit anti-bodies against armadillo IgM coupled toFITC. The rosette-forming capacity of theselymphocytes with SRBC was also evaluat-ed. The results (The Tablc) show that 17%of peripheral blood lymphocytes from 14armadillos are sIgM+, and 10% of the lym-phocytes from the same armadillos formrosettes with SRI3C.

Additionally, rabbit anti-armadillo IgMor IgG antibodies were peroxidase labeledin order to mcasure the isotypc-specific, hu-moral immune response of healthy arma-dillos in comparison with those experimen-tally infected with M. leprae. As shown inFigure 5, animais with active infection rec-ognized BCG crossreacting antigens withboth IgM and IgG isotypes.

DISCUSSIONInoculation of BSA into armadillo in-

duces the appearance of specific antibodies.Thesc antibodies were purified by immu-noaffinity, and were used as markers for theidentification and characterization of im-munoglobulins from whole serum of ar-madillos. Other authors have described theappearance of specific antibodies after in-oculation of diversc antigcns, but they havenot characterized their molecules (5).

The characterization of the i m munoglob-Mins purified from the armadillo using im-munoelectrophoresis and a rabbit antise-rum against armadillo whole serum pro-teins demonstrated the presence ofonly oneantigen; whereas this antiserum recognizedbetween 13 and 15 antigcns from the wholeserum of the animais in the same gel. Inexperiments done by Lewis and Doylc (9),five armadillo plasma fractions were de-tccted by electrophoresis corresponding toalbumin, ai, a2, 6' and -y globulins. Nagassiand co-workers (") detected 20 antigens bytwo-dimensional immunoelectrophoresis.The diffcrences with our results can be cx-plained by the use of diffcrent methods andalso by the quality of the antiserum used todevelop the antigens in the later work.

The presence of light as well as g and -yheavy chains was revcaled by scparation inpolyacrylamide gel electrophoresis under

FIG. 5. Humoral immune response of M. leprac-infected versus not-inoculated armadillos. The iso-

type-specific humoral immune response was evaluatedhy ELISA using a sonicated IICG as antigen. Groups

of armadillos were: Not inoculated (NI) (N = II); in-

oculated but without infection (lw/o1) (N = 6); inoc-ulated with poor infection (1w/P1) (N = 2); inoculated

with a clearly developing infection (1w/C1) (N = 3).

Infection was evaluated by macroscopic explorationand looking for bacilli in biopsies as was described ('').

Animais with poor infection did not present lepromas

and the amount of bacilli was very poor at the time of

the evaluation, in contrast with the clearly visible lep-

romas and numerous bacilli found in the other group.

Ali animais were evaluated ai the same time, li monthsafter inoculation with .11. leprac.

reducing conditions. In this gel we were un-able to identify J chains. The specific anti-bodies to BSA purified by immunoaffmity,and the IgM and IgG purified from Se-phadex G-200 and DEAE-cellulosc, whencompared to human IgM and IgG, showedthe same molecular weight for both heavyand light chains. These results correlate withthose reported by Vadiec, et al. ("); theyfound that armadillo IgG have approxi-mately the same electrophoretical mobilityas human IgG. Compared with our results,they reported the same molecular weightsfor light chains and slightly lower molecularweights for the -y heavy chains. The differ-ence in the molecular weights of the -y hcavychains is not significam (only 4-kDa differ-ence) and can be easily explained by theexperimental error of the method. We foundthat the rabbit antiscrums to IgM had cross-reactivity with IgG and vice versa (data notshown). Thus, in order to work with mono-specific antibodies, immunoabsorption withthcir heterologous and homologous im-munoglobulins coupled to Scpharose 4B wasperformed. Using this mcthod we preparedhighly specific reagcnts to analyze the cel-

I.6-t

-6 0^ IgG

04

w/o l^

I w/Pl.^I w/CL

60^ International fournal of Leprosy^ 1995

lular and humoral immune responses in thearmadillo.

The identification of armadillo lympho-cyte markers is important for the descrip-tion and quantification of lymphoid popu-lations in these animais. This is pertinentinformation for scanning the possible alter-ations in these cell populations during in-fection. In this work we identified and quan-tified lymphocytes bearing surface IgM (acharacteristic B-Iymphocyte marker), androsette-forming lymphocytes with SIZ BC (ahuman T-Iymphocyte marker), as well. Bal-lila (') reported 27.13% for sIgM + popu-lations in D. hybridus, while we found only16.7% in D. novemeinctus. The differentvalues can be explained since we dealt witha different species and because we ()multi-fied only the IgM-bearing lymphocytes.

For a rosette-ffirming lymphocyte popu-lation, Bali fia (1) reported 7.27% in D. hy-bridas and Escobar, et al. (`) found valuesless than 49% in D. novemeinetus; our datashow values of 9.6% for this cell type. Theselow values are very far from those reportedfor man, thus "a priori" it may be specu-lated that such a method is not good enoughto identify armadillo T lymphocytes.

In this work, the humoral response ofhealthy and experimentally Aí. leprae-in-fected armadillos against a crossreacting an-ligen (sonicated BCG) was measured. De-termination of isotype-specific antibodiesagainst BCG epitopes demonstrated a risingresponse of specific IgM and IgG in the ar-madillos when the infection has been estab-lished. In dose correlation with other re-ports this work shows that these an-imais are capable of recognizing and react-ing against crossreactive antigens onmycobacterial species.

We conclude that in armadillos it is pos-sible to evaluate the humoral immune re-sponse as the M. lepra(' infection progresses.On the other hand, more work is needed toidentify the lymphoid populations of D.novemeinetus.

SUMMARYIn this work we desenhe the purification

and characterization of armadillo immu-noglobulins. The IgM was precipitated us-ing low-strength ionic solution and furtherpurified by filtration through SephadexG-200. The IgG was obtained in pure form

by precipitation of serum with ammoniumsulfate and DEAE-cellulose ion exchangechromatography. The purity of these im-munoglobulins was evaluated by polyacryl-amide gel electrophoresis. The resultsshowed 28-kDa light chains and 55-kDa and70-kDa heavy chains for IgG and IgM, re-spectively. The rabbit antibodies againstthese molecules were used to prepare fluo-rescein (FITC) and peroxidase conjugates.The FITC conjugate was used to quantifyIgM-bearing lymphocyles. An average of17% of peripheral blood lymphocytes weresIgM+ from 14 healthy animais. Addition-ally, in the same animais we quantified lym-phocytes with the capacity to form rosetteswith sheep red-blood edis; the average forthis marker Nva s 1 0')/0. Also, the productionof crossreacting antibodies to BCG wasevaluated in healthy and Myeobacteriumhprac-inoculated animais using the perox-idase conjugates. Ali animais with activeinfection recognized BCG antigens.

RESUMENEn este trabajo se describe la purificación y carac-

terización de las inmunoglobulinas de armadillo. La

1gM fite precipitada usando tina solución de haja fuerza

jónica y posteriormente purificada por filtración enSephadex G-2()0. La IgG se (Maly° CO forma pura por

precipitación del suero con sulfato de amonio y cro-matogra lia de intercambio jónico en DEAE celulosa.

La pureza de las inmunoglobulinas se evaluó por elec-

troforesis en gel de poliacrilamida. Los resultados mos-

traron cadenas ligeras de 28 kD y cadenas pesadas de55 kD y 70 kl) para IgG e 1gM, respectivamente. Tam-

bién se prepararon anticuerpos de conejo contra estasmoléculas. Los anticuerpos se marcaron con fluores-

ceina (F1TC) y peroxidasa. El conjugado con F1TC setis() para cuantificar linfocitos con 1gM membranal. En

14 animales saltos estudiados, un promedio de 17% de

los linfbcitos periféricos fueron sIgM+. Adicional-

mente, en los mismos animales se cuantificó la capa-cidad de los linfocitos para formar rosetas coo eritro-

citos de carnero; cl promedio para este marcador fuedel 10%. También se evaluó la producción de anti-

cueropos reactivos con IICG en los animales sanos y

en los inoculados con Mycobaaerium leproc, usandolos conjugados con peroxidasa. Todos los ani males coninfcción activa reconocieron a los antígenos de BCG.

RÉSUMÉDans ce trztvail, nous décrivons la purification et la

caractérisation des immunoglobulincs de tatou. L'IgMa été précipitée en utilisant une solution ionique faible

et purifiée par filtration à travers du Sepbadex G-200.

L'IgG a été obtenue sous forme pure par précipitation

63, 1^Santos-Argumedo, et al.: Armadillo bizmunoglobulins^61

de serum avec du sulfate d'ammonium et chromato-graphic échangeuse d'ions sur cellulose DEAE. La pu-reté de ces immunoglobulines a été évaluée par élec-

trophorèse en gel de polyacrylamicie. Les résultats ont

montré des chaines légères de 28 kDa et 55 kDa et des

chaines lourdes de 70 kDa, respectivement pour lesIgG et IgM. Les anticorps de lapin contre ces molécules

ont été utilisés pour préparer des conjugués de fluo-rescéinc (FITC) et de peroxydase. Le FITC a été utilisé

pour quantifier les lymphocytes porteurs d'IgN1. Unemoyenne de 17% des lymphoeytes du sang périphé-

rique de 14 animaux en bonne sante était positive pour

les IgNI. De plus, chez les mimes animaux, nous avonsquantifié la capacité des lymphocytes à former des ro-

settes avec des gobules muges de mouton; la moyennepour ce marquem était de 10%. On a aussi évalué, avec

les conjugués de peroxydase, la production d'anticorps

réagissant de manière croisée au 11CG chez des ani-

maux en bonnc sante et des ani maus chez qui ont a vaitinoculé do lycobacterium hprae. Tous les ani mausavec une infection active ont reconnu les antigènes duBCG.

Acknowledgment. This work was supported bygrantsfrom the Consejo Nacional de Ciencia y Tecnologia,

México, D.F., México, and Dirección de EstudiosPostgrado e Investigación, 1.I'.N., México. The authorshold fellowships from COFAA, and/or SNI,México.

REFERENCES1. IIALISA, L. M., CARDAMA, J. E., GATTI, J. C.,

VALDEZ, R. P. and BIANum, O. Research on ar-

rnadillos in Argentina. In the armadillo as ao ex-

perimental model in biomedical research. l'anAmerican I lealth Organization, Scientific Publi-

cation 366 (1978) 103-106.2. CAMPBELL, D. H., GARVEN, J. S., CREMER, N. E.

and SUSSDORE, D. H., eds. A I ethOdS

London: W. A. Benjamin Inc., 1970, pp.193-201.

3. ESCOI3AR-GUTIERREZ, A. and AMEZCUA, M. E. El

armadillo: un nuevo animal de experimentación

para cl estudio de las zoonosis. Ciencia Veterinaria3(1981) 200-224.

4. ESTRADA-GAgei A, 1., QuEsADA-PAscuAL, F.,

SANTOS-ARGUMEDO, L., FLORES-ROMO, L.,

ESTRADA-PARRA, S. and BUCHANAN, T. M. Theearly serodiagnosis of leprosy. I. The use of coun-ter-immunoelectrophoresis and enzyme-linkedimmunosorbent assay. Rev. Latino-am. Micro-biol. 26 (1984) 267-272.

5. HARBOE, M. Significance on antibody studies inleprosy and experimental models of the disease.

Int. J. Lepr. 59 (1982) 342-350.

6. JOHNSTONE, A. and THORPE, R. Inununoehem-

istry in Pructice. Oxford: Illackwell Scientific Pub-

lication, 1982, pp. 150-162; 202-209; 256-259.

7. KIRCHHEIMER, W. F. and STORRS, E. E. Attemptsto establish the armadillo (Dusypus noremeinetus

Linn) as a model for study ofleprosy. Int. J. Lepr.39 (1971) 693-702.

8. LAEMMLI, U. K. Cleavage of structural proteins

during the assembly of the head of bacteriophageT4. Nature 227 (1970) 680-685.

9. LEWIS, J. H. and DOYLE, A. P. Coagulation pro-

teins and cellular studies on armadillo blood.Comp. Biochem. Physiol. 12 (1964) 61-66.

10. LOWRY, H. O., ROSEBROUGH, N. 1., EARR, A. L.

and RANDALL, R. J. Protein measurement withthe folin phenol reagent. J. Biol. Chem. 193 (1951)

265-275.I I. NEGASSI, K., CLOSS, O. and HARBOE, M. CrOSS-

reaction between serum proteins and water soluble

liver tissue antigens in the nine-banded armadillo(Dastpus /toren/eine/tu Linn) and man. Clin. Exp.Immunol. 38 (1979) 135-147.

12. PuR-mo, G. P. and S-roRRs, E. E. Impact of cooltemperatures on transformation of human and ar-maddlo lymphocytes (Dastpus novenicinc(us Linn)

as related to leprosy. Nature 248 (1974) 450-452.13. QUESADA-PASCUAL, R., ROJAS-ESPINOSA, O.,

SANTOS-ARGUMEDO, L. and ESTRADA-PARRA, S.

A Mexican armadillo (Dasipus novemeinctus) col-

ony for leprosy research. Int. J. Lepr. 55 (1987)716-718.

14. SAN.ros-Antatmox), L., HERNÁNDEZ-LOPEZ, J.,

QUESADA-PASCUAL, F. and ESTRADA-PARRA, S.Estandarización del ininunoanálisis enzimático.Infectologia 6 (1986) 18-23.

15. STORRS, E. E. The nine-banded armadillo: newmodel for biomedical research. Int. J. Lepr. 39

(1971) 703-713.

16. Utittot, M., CONVIT, J. and RAPETTI, M. IM-

munological characteristics of the armadillo Das-

tpus sabanicola. Clin. Exp. Immunol. 25 (1976)

170-176.17. VADIEE, A. R., HARRIS, E. and SHANNON, E. J. The

evolution of antibody response in armadillos in-oculated with Alycobacteritun hprae. Lepr. Rev.61 (1990) 215-226.

18. VADIEE, A. R., SHANNON, E. J., GILLIS, T. P. andHASTINGS, R. C. Partial characterization of an-tigens from ,11. hprae evoking IgG and IgM an-

tibodies in armaclillos. Int. J. Lepr. 56 (1988) 274-281.

19. VADIEE, A. R., SIIANNON, E. J., Giws, T. P.,MSHANA, R. N. and HASTINGS, R. C. Armadillo

IgG and 1gM antibody responses to phenolicglycolipid-I during experimental infection with M.

hprae. Int. J. Lepr. 56 (1988) 422-427.

20. VOLLER, A., Ilipww, D. E. and BARTLETT, A.

The enzyme-linked immunosorbent assay (ELISA).

London: Dynatech Laboratories Inc.

21. YOUNG, D. R. and BUCHANAN, T. M. A serolog-

ical test for leprosy with a glycolipid specific forMycohacterium hprae. Science 221 (1983) 1057-1059.