Proc. Natl. Acad. Sci. USAVol. 84, pp. 1679-1683, March 1987Medical Sciences

Genes for the major protein antigens of Mycobacterium tuberculosis:The etiologic agents of tuberculosis and leprosy share animmunodominant antigen

(Mycobacterium leprae/immunity)

ROBERT N. HUSSON*tt AND RICHARD A. YOUNG*t*Whitehead Institute for Biomedical Research, Nine Cambridge Center, Cambridge, MA 02142; tDepartment of Biology, Massachusetts Institute ofTechnology, Cambridge, MA 02139; and tDivision of Infectious Diseases, Department of Medicine, The Children's Hospital, Boston, MA 02115

Communicated by Maclyn McCarty, November 7, 1986 (received for review July 18, 1986)

ABSTRACT Mycobacterium tuberculosis genes encodingimmunologically relevant proteins were isolated by systemat-ically screening a Xgtll recombinant DNA expression librarywith a collection of murine monoclonal antibodies directedagainst protein antigens of this pathogen. These antibodies,previously characterized by a World Health Organizationworkshop on monoclonal antibodies against mycobacteria,were used to isolate DNA sequences encoding five major proteinantigens of this pathogen. To evaluate the extent of crossreac-tivity between these M. tuberculosis antigens and antigens ofMycobacterium leprae, recombinant antigens were probed withmonoclonal antibodies directed against the protein antigens ofthese bacilli. One of the antigens, a 65-kDa protein, hasdeterminants common to M. tuberculosis and M. leprae. Wefind not only that this antigen is recognized by mouse mono-clonal antibodies but that it is the major protein recognized byanti-M. tuberculosis rabbit sera. The 65-kDa proteins of M.tuberculosis and M. leprae appear to play a role in the humoraland cell-mediated immune response to these pathogens.

Tuberculosis, recognized as the major cause of infectiousmortality in Europe and the United States in the 19th andearly 20th centuries (1), remains a significant global healthproblem. In the United States there are over 20,000 new casesof tuberculosis annually, and the steadily declining incidenceof tuberculosis established in preceding decades appears tohave changed course, reaching a plateau in 1985 and showingan actual increase in 1986 (2, 3). Worldwide, tuberculosisremains widespread and constitutes a health problem ofmajor proportions, particularly in developing countries. TheWorld Health Organization estimates that there are tenmillion new cases of active tuberculosis per year with anannual mortality of approximately three million (4).

Tuberculosis is caused by Mycobacterium tuberculosis,M. africanum, or M. bovis, the "tubercle bacilli" of thefamily Mycobacteriaceae. The mycobacteria are aerobic,acid-fast, non-spore-forming, non-motile bacilli with highlipid contents and slow generation times. Among the othermycobacteria, M. leprae, the etiologic agent of leprosy, is theonly major pathogen (5). Other species, however, are capableof causing disease (6). Members of the M. avium-intracel-lularae complex are important pathogens among individualswith the acquired immunodeficiency syndrome (AIDS). Cer-tain groups of individuals with AIDS have a markedlyincreased incidence of tuberculosis as well (7).

Diagnostic and immunoprophylactic measures for myco-bacterial diseases have changed little in the past half century.Tuberculin, developed by Koch as a cure for tuberculosis inthe late 1800s and subsequently used as a skin-test reagent,

is a crude mixture of antigens from M. tuberculosis cultures(8). Enrichment of the protein fraction of this material in the1930s produced the purified protein derivative (PPD) (9),which is still used to diagnose exposure to tuberculosis.Bacillus Calmette-Guerin (BCG), an avirulent strain of M.bovis (10), has been used as a vaccine against tuberculosis formore than 50 years. Though numerous studies found that ithad protective efficacy against tuberculosis (reviewed in ref.11), a major trial of BCG in India indicated that this vaccinewas not protective against tuberculosis in this setting (12).Isolation and characterization of major antigens of M. tuber-culosis and of other mycobacteria should lead to the devel-opment of more effective tools for the diagnosis and preven-tion of tuberculosis and other mycobacterial diseases.

In 1984 and 1985 the World Health Organization (WHO)sponsored two workshops to characterize murine monoclo-nal antibodies against mycobacteria (13,,14). Fifty-six mono-clonal antibodies developed by several independent investi-gators were analyzed in multiple laboratories, and the reac-tivity of the monoclonal antibodies with several species ofmycobacteria was determined. Some of these antibodiesrecognized determinants unique to a single mycobacterium;however, many were found to crossreact with antigens frommultiple species. Ten of the anti-M. leprae antibodies rec-ognized four protein antigens, of 12, 18, 36, and 65 kDa.These, and additional monoclonal antibodies against a 28-kDa antigen (15), were used to isolate M. leprae genesencoding the five antigens (16). Nineteen of the anti-M.tuberculosis antibodies bound to seven protein antigens, of12, 14, 19, 23, 38, 65, and 71 kDa. We have used these anti-M.tuberculosis monoclonal antibodies to screen a Xgtll recom-binant DNA expression library ofM. tuberculosis in order toisolate genes for the protein antigens of this bacillus. Wereport here the isolation of genes for five major proteinantigens of M. tuberculosis and their use to investigate theantigenic relationship between M. tuberculosis and M.leprae.

MATERIALS AND METHODSPhage and Bacterial Strains. Bacteriophage Xgtll and

Escherichia coli strain Y1090 have been described (17, 18).M. tuberculosis Recombinant DNA Library. Construction of

the M. tuberculosis recombinant DNA library in Xgtll hasbeen described (19). The library has a titer of 2 x 1010plaque-forming units per ml and contains approximately 40%recombinants with an average insert size of 4 kilobases.

Monoclonal Antibody Probes. Murine monoclonal antibod-ies to protein antigens of M. tuberculosis, M. leprae, andother mycobacteria were generously supplied by the World

Abbreviations: BCG, bacillus Calmette-Guerin; PPD, purified pro-tein derivative.

1679

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 12

, 202

1

1680 Medical Sciences: Husson and Young

Health Organization and independent investigators. Amongthese antibodies, those used in the studies reported here wereIT-3 (20), IT-10 (21), IT-li (14), IT-12 (22), IT-13 (23), IT-15(23), IT-16 (14), IT-17 (15), IT-19 (23), IT-20 (23), IT-21 (22),IT-23 (23), IT-27 (14), IT-29 (14), IT-31 (14), IT-33 (24),ML06-A13 (25), L7-15 (26), SA1.D2D (identical to IT-17),SAL.B11H (15), F47-9-1 (13), ML04-A (25), C1.1 (13), IIH9(identical to IT-33), IIIE9 (24), IIC8 (24), IIIC8 (24), T2.3 (13),Y1-2 (13), SA2.D7C (gift of D. B. Young) and ML30A (25).All monoclonal antibodies were used at approximately 1:200to 1:300 dilution in 50 niM TrisHCl, pH 8/150 mMNaCl/0.05% Tween 20.

Rabbit Anti-M. tuberculosis Serum. Anti-M. tuberculosishyperimmune serum, produced by repeated immunization ofrabbits with M. tuberculosis H37Rv culture filtrate, wasgenerously provided by J. Bennedsen (Statens Seruminstitut,Copenhagen). This serum was used at 1:100 dilution.

Screening of the Xgtll M. tuberculosis Library with Anti-body Probes. Screening was as described (19), except that 1%bovine serum albumin was used in place of 20% fetal bovineserum to decrease background. Positive plaques were de-tected with a biotinylated secondary antibody system(Vectastain, Vector Laboratories, Burlingame, CA) or withan alkaline phosphatase-conjugated secondary antibody sys-

tem (Protoblot, Promega Biotec, Madison, WI).Probing of Arrays of Xgtll DNA Clones with Antibody

Probes. Two hundred microliters of a saturated culture of E.coli Y1090 was added to 2.5 ml of molten Luria-Bertani (LB)soft agar, poured onto 100-mm plates containing 1.5% LBagar, and allowed to harden at room temperature for 10 min.One hundred microliters of phage plate stock containing=1010 of the Xgtll DNA clones of interest was placed intoalternate wells of 96-well tissue culture plates. A multi-pronged transfer device was placed briefly in the wellscontaining phage and then touched lightly to the surface ofthe plate. Plates were then incubated at 42°C for =3 hr, atwhich point clear plaques -5 mm in diameter were visible.Plates were then overlayed with nitrocellulose filters satu-rated with 10 mM isopropyl,-D-thiogalactopyranoside andincubated at 37°C for 3.5 hr. Processing of filters for detectionof antigen was as described for screening the Xgtll library.Recombinant DNA Manipulation. DNA from recombinant

Xgtll clones was isolated and mapped with restrictionendonucleases by standard techniques (27).

Filter Hybridization of Insert DNA. Arrays of Xgtll cloneswere created as described above and incubated at 42°C for 5hr. The plates were then overlayed with nitrocellulose filtersand placed at4°C for 1 hr. Probe DNA was labeled with32pby nick-translation. Filter hybridization was performed asdescribed (27). Hybridization conditions were 50% (vol/vol)formamide/5x SSPE (lx SSPE is 0.18 M NaCl/10 mMNaj.5Hj.5PO4/1 mM Na2EDTA, pH 7.0)/i x Denhardt'ssolution [0.02% (wt/vol) Ficoll/0.02% (wt/vol) polyvinylpyr-rolidone/0.02% (wt/vol) bovine serum albumin]/0.3%NaDodSO4 at 42°C for -16 hr, followed by washing in 2xSSPE/0.2% NaDodSO4 at 45°C.

RESULTS

Gene Isolation and Characterization. Monoclonal antibod-ies directed against protein antigens of M. tuberculosis were

used individually to probe a Xgtll M. tuberculosis recombi-nant DNA library. The specific antibody probes used and thesize of antigen recognized are as follows: IT-3 (12 kDa); IT-20(14 kDa); IT-19 and IT-27 (19 kDa); IT-17 and IT-29 (23 kDa);IT-15, IT-21, and IT-23 (38 kDa); IT-13 (65 kDa); and IT-11(71 kDa) (14). This DNA library was previously constructedwith M. tuberculosis genomic DNA fragments such that allprotein-coding sequences would be represented and ex-pressed (19). Signal-producing clones were isolated using

antibodies directed against protein antigens of 12, 14, 19, 65,and 71 kDa. In each case similar numbers of clones wereisolated in screens of 105 recombinant plaques. DNA clonesencoding the 23-kDa and 38-kDa antigens could not bedetected with these antibodies.The insert DNAs of the recombinant DNA clones were

mapped with restriction endonucleases. Fig. 1 shows thegenomic DNA restriction maps deduced for the genes en-coding each of the five M. tuberculosis antigens and illus-trates how each of the cloned DNAs aligns with that map. Allclones isolated with monoclonal antibodies directed againstany single antigen align with a single genomic DNA segment.This result indicates that all clones were isolated becausethey express the protein of interest rather than an unrelatedpolypeptide containing a similar or identical epitope. Inaddition, this result suggests that each antigen is the productof a single gene.

4Kbh

12 kDa

Sol

L

14 kDa

B HX BXI I I Il

R Y3275

E B XSI I II

*--a.w a

L

L

L

L

19 kDa

*. -

L

65 kDa * *e.XiFNr

' Y3144- F Y3145. '(3248- It Y3155i Y3151

KB K P PP11I II

-one

- A Y3147

L IR Y3148

KAA EI I1I I

A

R L Y3141R L Y3143

L 'Y3150L R Y3253

R L Y3262

71 kDaK X 6KG PBE EII 1 1 1a 1l

L A

.

L

Y3268P Y3271

Y3272Y3273

FIG. 1. Restriction maps of M. tuberculosis DNA. RecombinantDNA clones isolated with monoclonal antibodies directed against the12-, 14-, 19-, 65- and 71-kDa protein antigens were mapped withrestriction endonucleases. The insert DNA endpoints are designatedleft (L) or right (R) in relation to lacZ transcripts that traverse theinsert from right to left. Restriction enzymes: A, SalI; B, BamHI;E, EcoRI; G, BglII; K, Kpn I; P, Pvu I;S, Sac I; X, Xho I. Asterisksindicate restriction endonuclease cleavage sites that are shared bythe M. tuberculosis and M. leprae genes encoding the 65-kDaantigen. kb, Kilobase.

Proc. Natl. Acad. Sci. USA 84 (1987)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 12

, 202

1

Proc. Natl. Acad. Sci. USA 84 (1987) 1681

The orientation of each DNA insert in the recombinantclones was determined by restriction analysis. Only amongthe clones for the 65-kDa antigen were the inserts found inboth possible orientations relative to the direction of lacZtranscription in Xgtll. This result suggests that this proteincan be expressed in E. coli from signals independent of thoseprovided by lacZ. Similar results have been obtained forrecombinant DNA clones encoding the 65-kDa antigens ofM.bovis (28) and M. leprae (16).Recombinant Antigens Recognized by Rabbit Serum. To

assess the response ofanother experimental animal to antigenpreparations of M. tuberculosis, we examined the reactivityof rabbit anti-M. tuberculosis hyperimmune serum withrecombinant antigens. This serum produced positive signalswith Xgtll clones encoding each of the five M. tuberculosisepitopes that had been isolated with murine monoclonalantibodies (Fig. 2). Particularly strong signals were observedwith the 65-kDa and 71-kDa antigens (Fig. 2). These resultsdemonstrate that mice and rabbits can mount an antibodyresponse to the same protein antigens of M. tuberculosis.We find that clones for the five M. tuberculosis antigens are

detected at similar frequencies in the Xgtll recombinantDNA library, using monoclonal antibodies. Thus, the relativenumber of each of the five genes isolated with polyclonalserum antibodies should reflect the relative titer and diversityof the serum antibodies to each of the five antigens.To determine whether any of the five M. tuberculosis

antigens are relatively immunodominant in the rabbit hu-moral immune response to M. tuberculosis, we performed thefollowing experiment. The M. tuberculosis Xgtll recombi-nant DNA library was screened with the rabbit serum, and 40signal-producing clones were isolated. These 40 clones werearrayed and probed with monoclonal antibodies directedagainst each of the five recombinant M. tuberculosis proteinantigens. Remarkably, 17 of the 40 clones (43%) reactedstrongly with at least one of the four anti-65-kDa monoclonalantibodies tested. An additional six clones (15%) reactedstrongly with the anti-71-kDa monoclonal antibody IT-11.These results indicate that a large proportion of the anti-M.tuberculosis antibody present in this rabbit serum is directedagainst the 65-kDa antigen of M. tuberculosis and suggestthat it is a dominant antigen for the rabbit humoral immuneresponse. Seventeen of the clones did not react with any ofthe monoclonal antibodies tested, suggesting that the rabbit

?

e0

1 2

3 4 5

6 7 8 910 11 12

13 14 15 1617

FIG. 2. Arrays ofantigens from mycobacterial recombinant DNAclones probed with rabbit hyperimmune serum. Sixteen cloned XgtIirecombinants were arrayed on lawns of E. coli Y1090. The phagewere grown, antigen expression was induced, and the antigens wereblotted and probed with serum as described in the text. The code ofthe recombinant DNA clones shown on the numbered filter is asfollows: 1, Y3275; 2, Y3274; 3, Y3279; 4, Y3277; 5, Y3247; 6, Y3272;7, Y3150; 8, Y3254; 9, Y3147; 10, Y3163; 11, Y3179; 12, Y3191; 13,Y3252; 14, Y3178; 15, Y3180; 16, Y3143; 17, Xgtll. Clones 1, 5, 6,7, 9, and 16 are M. tuberculosis recombinants described in the text.Clones 10, 11, 14, and 15 are M. leprae recombinants expressingepitopes of the 18-, 28-, 36-, and 65-kDa antigens, respectively (28).Clones 2, 3, 4, 8, 12, and 13 are uncharacterized recombinants fromthe Xgtii M. tuberculosis and M. leprae libraries. Clone 17 is anonrecombinant Xgtll control.

serum may identify M. tuberculosis proteins not recognizedby the murine antibodies.

Antigenic Relatedness of M. tuberculosis and M. lepraeProteins. Evidence that M. tuberculosis and M. leprae shareimmunologically important antigens led us to investigate theexact nature of the immunological relatedness among recom-binant protein antigens ofM. tuberculosis and M. leprae. Foreach of five M. tuberculosis and four M. leprae proteinantigens, a single recombinant DNA clone containing most orall of the gene of interest was used to express antigen in thefollowing experiment. The recombinant phage clones werearrayed on a laiwn ofE. coli Y1090, which was then grown andinduced for antigen expression. Antigen immobilized onnitrocellulose filters was then probed with 26 individualanti-M. tuberculosis and anti-M. leprae monoclonal antibod-ies. Fig. 3 shows the array ofDNA clones used and the resultsobtained with the anti-M. tuberculosis antibodies IT-1, IT-3,IT-li, IT-16, and IT-31, which recognize proteins of 14, 12,71, 19, and 65 kDa, respectively. Table 1 details the resultsof these experiments, showing the reactivity of antigenexpressed from individual recombinant DNA clones witheach of the monoclonal antibodies.

Several conclusions can be drawn from the results shownin Table 1. Among the 11 monoclonal antibodies that recog-nize a 65-kDa antigen, 7 react with the 65-kDa protein fromboth mycobacteria [IT-31, C1.1, IIH9 (identical to IT-33),

0~~~~~~~~~~

21

43

1 23 4 5

6 7 8 910 11 12

13 14 15 1617

5

FIG. 3. Arrays of recombinant mycobacterial antigens probedwith monoclonal antibodies to assess antigen crossreactivity. ClonedXgtll recombinants were arrayed on lawns ofE. coli and probed withmonoclonal antibodies as described in the text. The array of clonesis identical to that shown in Fig. 2. Antibody probes and the antigensizes recognized are as follows: 1, IT-li (71 kDa); 2, IT-31 (65 kDa);3, IT-16 (19 kDa); 4, IT-1 (14 kDa); 5, IT-3 (12 kDa).

Medical Sciences: Husson and Young

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 12

, 202

1

1682 Medical Sciences: Husson and Young

11C8, T2.3, Y1-2, SA2.D7C], one antibody reacts only withthe M. tuberculosis protein (IT-13), and two antibodies reactonly with the M. leprae protein (IIIE9 and IIIC8). Oneantibody, ML30A, crossreacts with an antigen in E. coli andcould not specifically identify antigen-producing clones.These results indicate that the 65-kDa protein antigens of M.tuberculosis and M. leprae are hpmologues and share anumber of epitopes. In addition to these shared epitopes,however, both 65-kDa antigens have epitopes that are spe-cific for one species relative to the other.No crossreactivity was observed between other antigens of

these two mycobacterial species. Because monoclonal anti-

Proc. Natl. Acad. Sci. USA 84 (1987)

bodies recognize a single epitope and because only one or afew antibodies were available for each antigen, it is not clearwhether the 65-kDa proteins are the only homologous proteinantigens of M. tuberculosis and M. leprae. Among the otherantigens for which Xgtll clones have been isolated, however,only the 18-kDa antigen of M. leprae and the 19-kDa antigenof M. tuberculosis are of similar size. To determine whetherthese two antigens are related, we examined the homology ofthe DNA sequences that encode these antigens. At condi-tions of moderate stringency, where the DNAs encoding theM. tuberculosis and M. leprae 65-lDa antigens hybridizedstrongly, no hybridization was observed between the insert

Table 1. Reactivity of monoclonal antibodies with recombinant protein antigens

H. leprae

12kD 14kD

Y3275 Y3247

l9kD 6CkD 71AD

Y3147 Y3150 Y3272

- l8kD

Agtll Y3179

28kD

Y3163

36kD 65kD

Y3180 Y3178

H. tuberculosis

12kD IT-3

14kD IT-1

IT-4

IT-20

l9kD IT-10

IT-12

IT-16

IT-19

0D_ )-0-0-0@

_ -0@_ -0@_ -0@

e0_ _

- 0 (

- 0@_ _ _-0

- - - - - 0@- - - - - 0@

-0

e-

28kD SA1.D2D - -

SA1.BIIH - -

36kD F47-9-1 - -

_ _

0

0ML04 -A - - - - - -

65kD CI.I - - - 0

111j9 - - - 0iIIIE9 - - -

IIcs - - - 0

iiics - - - -

T2.3 - - 0(3Y1-2 - - (0SA2.D7C - - - (ML30A 0) (0 )0 i 000DE

--0Q- 0(- 0(-0(3000~~G

65kD IT-13

IT-31

IT-33

71kD IT-il

%. leprae

18kD L7-15

Arrays ofrecombinant antigens ofM. tuberculosis and M. leprae were probed with monoclonal antibodies directed againstantigens of these species. Clones were scored as positive only if the signal produced was clearly greater than the backgroundsignal produced by the nonrecombinant Xgtll clone included in each array. kD, kDa.

M. tuberculosis

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 12

, 202

1

MedicalSciences:HussonandYoungProc.Nati. Acad. Sci. USA 84 (1987) 1683

DNA of Y3147 (an M. tuberculosis 19-kDa clone) and Y3179

(an M. leprae 18-kDa clone), indicating no significant homol-

ogy between the DNA sequences of the insert DNAs of these

two clones. This result suggests that the M. tuberculosis

19-kDa protein and the M. leprae 18-kDa protein are unlikelyto be homologous.

DISCUSSION

Recombinant DNA clones encoding five major protein anti-

gens of M. tuberculosis were isolated with an extensive

collection of well-characterized murine monoclonal antibod-

ies. These five proteins were also found to be major antigensin the rabbit humoral immune response to M. tuberculosis.

One of these antigens, the 65-kDa protein, is shared with the

other mycobacterial pathogen M. leprae.Several lines of evidence indicate that the 65-kDa antigen

is an immunodominant protein antigen of M. tuberculosis.

Eleven of the 25 different M. tuberculosis and M. lepraemonoclonal antibodies examined in this study recognized the

65-kDa recombinant antigen from one or both mycobacteria.In addition, almost half of the recombinant DNA clones

isolated with rabbit polyclonal anti-M. tuberculosis serum

express the 65-kDa antigen, reflecting the predominance of

antibody to this antigen in this antiserum.

Considerable evidence indicates that the 65-kDa antigen

plays a role in the human immune response to tuberculosis.

Antibodies directed against this protein can be detected in the

serum of patients with tuberculosis (28). The 65-kDa antigenis present in PPDs of M. tuberculosis, M. bovis, and other

mycobacteria (28). Finally, helper-T-cell clones reactive with

recombinant 65-kDa antigen have been isolated from patientswith tuberculosis (29, 30), indicating that this antigen is

involved in the cell-mediated as well as the humoral immune

response to tuberculosis. Because a large number of T-cell

clones were isolated that do not appear to react with the

65-kDa antigen, it is possible that antigens equally or more

important to T-cell immunity remain to be identified.

Among the major antigens of the leprosy bacillus, the

65-kDa antigen appears to elicit antibody and T-cell respons-

es similar to those observed for the M. tuberculosis antigen.Both serum antibodies (26) and T cells (31) directed againstthe 65-kDa M. leprae antigen have been observed in patientswith leprosy. In addition, T-cell. clones from leprosy patientshave been found to respond to recombinant 65-kDa protein of

M. bovis (32), as well as to PPDs from both M. bovis BCG and

M. leprae (33). It is interesting to note that in vaccine trials

in Asia and Africa, BCG provided significant protection

against leprosy, ranging from 20% to 80% (reviewed in ref.

34). An intriguing possibility is that theM. bovis BCG 65-kDa

antigen is involved in engendering the immune protection

provided by this vaccine against M. leprae as well as againstM. tuberculosis.

There is evidence that the 19-kDa and 71-kDa antigens of

M. tuberculosis, in addition to the 65-kDa antigen, are

involved in the immune response to this bacillus. Helper-T-cell clones from tuberculosis patients have been isolated that

respond to the recombinant 19-kDa protein (29). The 71-kDa

antigen is recognized by the humoral immune system of both

mice and rabbits, and we have shown that antibody to this

antigen is a prominent component of hyperimmune anti-M.

tuberculosis rabbit serum.

The isolation of genes for major protein antigens of M.

tuberculosis should permit the development of improved

reagents for diagnosis and immunoprophylaxis of tuberculo-

sis. For example, proteins encoded by some of these DNA

sequences may be a source of specific serodiagnostic and

skin-test antigens, reagents that would be valuable for mon-

itoring disease transmission and the effectiveness of vacci-nation or therapy. Finally, it is now possible to introduce intoviruses, and possibly into cultivable mycobacteria, genes thatspecify polypeptides that may provide immunological pro-tection in order to produce more specific and effectivevaccines.

We thank Barry Bloom, Vijay Mehra, Tore Godal, DouglasYoung, Juraj Ivanyi, Thomas Shinnick, Steven Shoemaker, JimWatson, Doug Sweetser, and Bobby Cherayil for advice and stim-ulating discussion. We are also grateful to the World Health Orga-nization and the individual investigators for their gifts of monoclonalantibodies. This work was supported by grants from the NationalInstitutes of Health (A123545), the World Health OrganizationProgram for Vaccine Development, the World Health Organization/World Bank/United Nations Development Program Special Programfor Research and Training in Tropical Diseases, and the HeiserProgram for Research in Leprosy.

1. Dubos, R. & Dubos, J. (1952) The White Plague: Tuberculosis, Man andSociety (Little, Brown, Boston).

2. Centers for Disease Control (1986) Morbid. Mortal. Wkly. Rep. 35,699-703.

3. Centers for Disease Control (1987) Morbid. Mortal. Wkly. Rep. 35, 810.4. Joint International Union Against Tuberculosis and World Health Orga-

nization Study Group (1982) Tubercle 63, 157-169.5. Bloom, B. R. & Godal, T. (1983) Rev. Infect. Dis. 5, 765-780.6. Wallace, R. J., Swenson, J. M., Silcox, V. A., Good, R. C., Tschen,

J. A. & Stone, M. 5. (1984) Rev. Infect. Dis. 5, 657-679.7. Pitchenik, A. E., Cole, C., Russell, B. W., Fischl, M. A., Spira, F. J.,

Snider, D. E., Jr. (1984) Ann. Intern. Med. 101, 641-645.8. Koch, R. (1891) Dtsch. Med. Wochenschr. 17, 101-102.9. Seibert, F. M., Aronson, J. D., Reichel, J., Clark, L. T. & Long, E. R.

(1934) Am. Rev. Tuberc. 30, Suppl. 705-778.10. Calmette, A., Guerin, C. & Weill-Halle, B. (1924) Bull. Acad. Nat. Med.

Paris 91, 787-796.11. Luelmo, F. (1982) Am. Rev. Respir. Dis. 125, 70-72.12. World Health Organization (1980) WHO Tech. Rep. Ser. 651, 1-15.13. Engers, H. D. (1985) Infect. Immun. 48, 603-605.14. Engers, H. D. (1986) Infect. Immun. 51, 718-720.15. Young, D. B., Fohn, M. J., Khanolkar, S. R. & Buchanan, T. M. (1985)

Clin. Exp. Immunol. 60, 546-552.16. Young, R. A., Mehra, V., Sweetser, D., Buchanan, T., Clark-Curtiss,

J., Davis, R. W. & Bloom, B. R. (1985) Nature (London) 316, 450-452.17. Young, R. A. & Davis, R. W. (1983) Proc. Natl. Acad. Sci. USA 80,

1194-1198.18. Young, R. A. & Davis, R. W. (1983) Science 222, 778-782.19. Young, R. A., Bloom, B. R., Grosskinsky, C. M., Ivanyi, J., Thomas,

D. & Davis, R. W. (1985) Proc. Natl. Acad. Sci. USA 82, 2583-2587.20. Minden, P., Kelleher, J. H., Freed, L. D., Nielsen, L. D., Brennan,

P. J., McPheron, L. & McClatchy, J. K. (1984) Infect. Immun. 46,519-525.

21. Kolk, A. H. J., Minh, L. H., Klatser, P. R., Eggelte, T. A., Kuijper, S.,de Jonge, S. & von Leeuwen, J. (1984) Clin. Exp. Immunol. 58, 511-521.

22. Andersen, A. B., Yuan, Z., Haslov, K., Vergmann, B. & Bennedsen, J.(1986) J. Clin. Microbiol. 23, 446-451.

23. Coates, A. R. M., Hewitt, J., Allen, B. W., Ivanyi, J. & Mitchison,D. A. (1981) Lancet Hi, 167-169.

24. Gillis, T. P. & Buchanan, T. M. (1982) Infect. Immun. 37, 172-178.25. Ivanyi, J., Sinha, S., Aston, R., Cussel, D., Deen, M. & Sengupta, U.

(1983) Clin. Exp. Immunol. 52, 528-536.26. Britton, W. J., Hellqvist, L., Basten, A. & Raison, R. L. (1985) J.

Immunol. 135, 4171-4177.27. Davis, R. W., Botstein, D. & Roth, J. R. (1980) Advanced Bacterial

Genetics: A Manual for Genetic Engineering (Cold Spring HarborLaboratory, Cold Spring Harbor, NY).

28. Thole, J. E. R., Dauwerse, H. G., Das, P. K., Groothuis, D. G.,Schouls, L. M. & van Embden, J. D. A. (1985) Infect. Immun. 50,800-806.

29. Oftung, F., Mustafa, A. S., Husson, R. N., Young, R. A. & Godal, T.(1987) J. Immunol., in press.

30. Lamb, J. R., Ivanyi, J., Rees, A., Young, R. A. & Young, D. B. (1987)Lepr. Rev., in press.

31. Mustafa, A. S., Gill, H. K., Nerland, A., Britton, W. J., Mehra, V.,Bloom, B. R., Young, R. A. & Godal, T. (1986) Nature (London) 319,63-66.

32. Emmrich, F., Thole, J., van Embden, J. & Kaufmann, S. H. E. (1986) J.Exp. Med. 163, 1024-1029.

33 OSakar, P. Agis, F., Walach, D.FTageul,-., CotentF. Augier, JT& Bach, M. (1986) J. Immunol. 136, 4255-4263.

34. Fine, P. (1984) Tubercle 65, 137-153.

Medical Sciences: Husson and Young

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 12

, 202

1