Evidence for Clonal Selection of -y/b T Cells inResponse to a Human PathogenBy Koichi Uyemura," Robert J. Deans,§ Hamid Band,1Jeffrey Ohmen," Govindaswamy Panchamoorthy,1Craig T. Morita,l Thomas H . Rea,11 and Robert L. Modlin"$

From the *Division ofDermatology and #Department of Microbiology and Immunology,UCLA School ofMedicine, Los Angeles, California 90024; the DDppartment of Neurologyand Il Section of Dermatology, USC School ofMedicine, Los Angeles, California 90033; andthe (Laboratory of Immunochemistry, Dana Farber Cancer Institute, and Department ofRheumatology and Immunology, Harvard Medical School, Boston, Massachusetts, 02115

Summary

T cells bearing y/b antigen receptors comprise a resident population ofintraepithelial lymphocytesin organs such as skin, gut, and lungs, where they are strategically located to contribute to theinitial defense against infection . An important unsolved question about antigen-driven y/b Tcell responses regards the breadth of their T cell receptor (TCR) repertoire, since many specificepithelial compartments in mice display limited diversity. We have examined the diversity ofTCR S gene expression among human y/b T cells from skin lesions induced by intradermalchallenge with Mycobacterium leprae . We show that the vast majority of'y/b cells from M. lepraelesions use either V61-J61 or V62-J61 gene rearrangements and, within a given region of thelesion, display limited junctional diversity. This contrasts markedly with the extensive diversityof y/b T cells from peripheral blood of these same individuals, as well as skin from normaldonors . These results indicate that they/b response toM. leprae involves the selection of a limitednumber ofclones from among a diverse repertoire, probably in response to specific mycobacterialand/or host antigens .

One of the fundamental aspects of cellular immunologyis the selection and clonal expression of T cells bearing

specific receptors by antigens. While the selection of ci/(3T cells has been studied in detail, little is known about theselection ofy/b T cells by antigen . The genetic diversity ofthe TCR provides a measure of the scope of the T cell reper-toire. In contrast to the a/a TCR, the germline gene seg-ment diversity for both the TCR y and S chains is small .The further limitation of this diversity by the preferentialusage of only a few variable (V) genes or V gene pairs atspecific anatomical locations (1-3) suggests a particularlynarrow TCR repertoire, likely reflecting recognition of alimited number of ligands . On the other hand, there is un-precedented junctional diversity, particular in the b chainsof this receptor (4-6) . We reasoned that analysis of the TCRy/b repertoire of a specific immune response would provideclues about the set of antigens recognized : whether diverseor limited, conventional, or superantigen-like. The presenceof antigen-reactive TCR -y/6-bearing cells in leprosy skinlesions (7) provides a unique in vivo model to examine they/b receptor repertoire at the site of immunopathologicreaction .

Much experimental evidence indicates that y/b T cells con-tribute to the granuloma formation in response to mycobac-terial infection (7) . First, TCR y/b cells comprise a strik-ingly high percentage of the T cell population in infectiousdisease lesions that contain recently formed granulomas. Theseinclude lepromin skin tests (Mitsuda reactions), which areexperimental DTH reactions induced by intradermal injec-tion ofMycobacterium lepme, reversal reactions in leprosy, whichrepresent a naturally occurring DTH response to M. leprae,and localized American cutaneous leishmaniasis. Second, they/b T cells from these infectious lesions proliferate to mycobac-terial antigens in vitro. Third, the y/b T cells from lesionsappear to release a lymphokine(s) that synergizes with gran-ulocyte/macrophage (GM)-CSF to induce macrophage adhe-sion, aggregation, and proliferation, cellular events that wouldbe necessary for the granulomatous response.The present study was undertaken to determine the distri-

bution and TCR diversity of y/b T cell subpopulations inlepromin skin tests . This was accomplished by immunopatho-logic analysis according to VS chain expression and molec-ular analysis of Vb, Jb, and junctional elements. The reper-toire in these infectious lesions was compared with the

683

J. Exp. Med. ® The Rockefeller University Press " 0022-1007/91/09/0683/10 $2 .00Volume 174 September 1991 683-692

on February 13, 2018

jem.rupress.org

Dow

nloaded from

684

Clonal Selection of ,y/S T Cells in Response to a Human Pathogen

on February 13, 2018

jem.rupress.org

Dow

nloaded from

Figure 1 A-F.

Immunohistochemical analysis of VS phenotypes in lesions and peripheral blood. Both V61 (A) and V62 (B) -bearing T cells werepresent in the dermal granulomas of lepromin skin tests. In contrast, V61 cells Pwere present in the epidermis in greater numbers than V62 cells(D), which were virtually absent . A more extensive view of the epidermis (E) shows V61 cells throughout . The percentages of V61- and V62-bearingT cells (± SD) are shown in F.

peripheral blood of the same individual and the repertoirein normal human skin .

Materials and MethodsImmunoperoxidaseStaining ofBiopsies.

Lepromin skin tests (3-wkMitsuda reactions) were studied from 14 tuberculoid leprosy pa-tients (8) . All specimens were obtained with informed consent.The patients were distributed among the different diagnostic groupsshowing no segregation according to sex, race, or age. Skin biopsyspecimens were obtained by punch or scalpel technique at the timeof diagnosis, embedded in OCT medium (Ames Co., Elkhart, IN),and snap frozen in liquid nitrogen . The tissues were stored at -70°Cuntil sectioning. Sections (3-5 P.m) were acetone fixed before un-dergoing incubations with the mAbs for 45 min. mAbs includedthe Cb-specific antibody antiTCR61 (9) (1:100, T-cell Sciences,Boston, MA), the V61-J61-specific S antibody, 6TCS1 (1 :10, TcellSciences), and the V62-specific antibody, BB3 (1:3,000 ; a generousgift ofDr. L. Moretta, Istituto Nazionale per la Ricerca sul Cancro,Genova, Italy) . The results obtained with 6TCS1 were corrobo-rated with the V61-reactive antibody A13 (1:1,000 ; kindly providedby Dr. S. Ferrini, Istituto Nazionale per la Ricerca sul Cancro,Genova, Italy) . The primary antibody was visualized using the alka-line phosphatase anti-alkaline phosphatase technique (Dako Corp.,Carpinteria, CA). Slides were sequentially incubated with rabbitanti-mouse antibody for 30 min, alkaline phosphatase mouse anti-alkaline phosphatase immune complex for 30 min, and then devel-oped with Vector Red substrate for 30 min (Vector Laboratories,Burlingame, CA). Slides were washed in Tris buffer between incu-bations . Endogenous alkaline phosphatase was blocked using 125mM levamisole (Vector Laboratories). Slides were counterstainedwith hematoxylin and then mounted in glycerine-gelatin . Someslides were stained using theABC Elite system (Vector Laborato-ries). The numbers ofVSpositive cells in the epidermis and dermis

685

Uyemura et al .

of an entire tissue section (-5 x 5 mm) were enumerated andexpressed as the percentage of total TCR61' cells in each micro-au.atomic location . Peripheral blood was obtained from the sameindividuals and analyzed by flow cytometry.PCR Analysis . 20 5-P.m.sections from each of five lepromin

skin tests were placed in 0.5 ml of4M guanidinium isothiocyanatebuffer. Additionally, 106 PBMC were obtained by Ficoll-Hypaquedensity centrifugation . GenomicDNAwas isolatedby phenol/chlo-roform extraction and ethanol precipitation with yields of 1-3 wgof DNA. For each PCR reaction, -100 ng of DNA was used ina 25-w1 reaction . By calculating the numbers of y/b cells/mm2 ofgranuloma as detailed previously (10), and knowing the volumeoftissue utilized, we determined that each PCRreaction oflesionscontained DNA from -500 y/b T cells . Similar calculations forperipheral blood indicated identical numbers of y/bTcells per eachreaction . Titration ofVBJS containing plasmid confirmed that thesensitivity of our PCR amplification was on the order of 102 or103 copies. PCR amplification mixtures contained 10 pmol of eacholigonucleotide primer, 2.5 U Taq polymerase (Promega Corp.,Madison, WI), and 2.5 MM M9C12 in PCR buffer (PromegaCorp.) . Oligonucleotide primers were tested on a panel ofy/6 Tcell clones with known rearrangements. Sequences ofoligonucleo-tide primers were selected to yield -300 by products (11) . 30 cyclesof PCR amplification were performed using the following condi-tions: 1 min ofdenaturation at 94°C and 2 min of annealing/exten-sion at 65°C . Amplified products were subjected to electropho-resis on 1.5% agarose gels and visualized as single bands byethidiumbromide. Verification of product was accomplished by nucleotidesequencing. Two skin biopsy specimens were obtained from normaldonors after informed consent. DNA was extracted from the en-tire 4-mm punch biopsy and 100 ng used for PCR amplificationfollowed by sequencing of the V-J junctions.

Nucleotide Sequence Analysis of TLCR S VJJunctions.

To facili-tate cloning into sequencing vector, oligonucleotide primers for

on February 13, 2018

jem.rupress.org

Dow

nloaded from

PCR amplification were designed to contain a Sall or EcoRl re-striction site in VS andJb primers, respectively. Amplified productwas digested with both restriction enzymes and gel purified. Purifiedproducts were ligated into pUC18 vector and used to transformDH5a (BRL Laboratories, Gaithersburg, MD) and selected for am-picillin resistance. Plasmid DNA from insert containing colonieswas sequenced (Sequenase 2.0 Kit ; U.S. Biochemicals, Cleveland,OH) using the method of Sanger et al. (12) . Precautions taken toavoid PCR contamination artifact included the use ofpositive dis-placement pipettes, assembling ofPCR reactions in a laminar flowhood in a separate room from plasmid preparations, and inclusionof negative and well-characterized controls in the PCR reactions.

ResultsDistinct Microanatomic Localization of Vb T Cell Popula-

tions . The microanatomic distribution of 'y/b T cell sub-populations within lepromin skin tests was investigated byimmunohistologic staining of frozen sections with mAbsdirected against Vb-encoded determinants (Fig . 1) . Withinthe dermal granulomas, Vbl- and V62-bearing cells accountedfor the majority of infiltrating y/b cells, with a V62/V61ratio -2:1 compared to 9:1 in the peripheral blood of thesesame individuals . Unexpectedly, in contrast to the predomi-nance of V62-bearing cells in the peripheral blood and in thedermis, TCR6 * cells infiltrating the epidermis primarily ex-pressed the V61-encoded TCRs. V82+ cells were rarely de-tected in the epidermal layer.

Preferential Usage ofJ61 by y/b T Cells in Lesional Skin.

Tofurther characterize the repertoire of the y/b T cells in re-sponse to mycobacterial stimulation in vivo, we performedPCR amplification oftissue-derived vs . peripheral blood-de-rived DNA using paired V and joining (J) oligonucleotideprimers. PCR conditions were established such that each PCRreaction included DNA derived from -500 y/b T cells. PCRanalysis confirmed that Vbl and Vb2 were used by y/b Tcells in these lesions, as amplified product was obtained fromall five lesions for V61-J61 and V62-J61 rearrangements (Fig .2) . In contrast, V62-J63 rearrangements were detected in 1/5lesions, but 5/5 blood samples (Figure 2) . These data sug-gest a selective localization of the V62-J61 subpopulation ofVS2+ cells to lesions . PCR using other VS and Jb combi-nations did not reveal additional gene rearrangements (datanot shown) .

Limited functional Diversity of y/6 TCRs in Leprosy SkinLesions. Thejunctional diversity of y/b T cells in leprominskin tests was determined by cloning and sequencing PCRamplified products . We chose to focus on TCR b chain diver-sity for two reasons : it encodes the greatest amount of diver-sity in the y/b receptor ; and it is specific for y/b T cells sincethe b locus is deleted during rearrangement of the TCR achain . By using DNA rather than mRNA, bias towards ac-tivated lymphoblasts containing abundant mRNA is elimi-nated as each y/b T cell contains only one copy of its func-tionally rearranged b gene. Strikingly, in each of the threelepromin skin tests subjected to nucleotide sequencing anal-ysis, the majority ofVblJb1 and V62-J61junctional sequenceswere found to be identical, but distinct for each patient (Fig .3) . This was clearly different from peripheral blood of these

686

Clonal Selection of y/b T Cells in Response to a Human Pathogen

Figure 2 .

PCR analysis of S chain V-J gene segment usage in lesionsand peripheral blood . VJ junctions were PCR amplified from DNA ex-tracted from five lepromin skin tests (lanes A-E) and the peripheral bloodof the same individuals. PCR-amplified product was detected in all fiveskin lesions and blood for VblJbl (A) and VS2JSl (B) . In all five bloodsamples, but only one of five skin lesions, V62-J63 rearrangements couldbe amplified (C) . No rearrangements were detected for any other VBJbpairs (data not shown) .

on February 13, 2018

jem.rupress.org

Dow

nloaded from

A

B

D

V61-J61 JUNCTIONAL NUCLEOTIDE SEQUENCES FROM BIOPSY OF LEPROMIN SKIN TESTS

687

Uyernura et al .

V62-J61 JUNCTIONAL NUCLEOTIDE SEQUENCES FROM BIOPSY OF LEPRONIN SKIM TESTS

V62-J61 JUNCTIONAL NUCLEOTIDE SEQUENCES FROM PATIENT PEMC

V61-JSI JUNCTIONAL NUCLEOTIDE SEQUENCES FROM PATIENT PBMC

Figure 3 .

Nucleotide sequenceanalysis ofTCR6 VJjunctions ofskin test lesions and blood . Se-quencing of PCR products wasperformed to analyze the diversityat the VJjunction . Limited diver-sity was found in both the V61-J61(A) andV62JSl (B) junctionsof y/6 T cells from lesions inpatients A, B, and D. In contrast,there was marked genetic diversityin both theV61-J61 (C) and V62-J61(D)junctions from the bloodof the same individuals . Nonfunc-tional sequences are marked withan asterisk .

P/N D_61 P/N D62 P/N D63 P/N J61 FREQUENCYgermline GTGACACC GAAATAGT CCTTCCTAC ACTGGGGGATACG ACACCCATAA

PT A GTGACACC G ATAG G CCTT GCAAT ACCGATAA (1/13)GTGAC CCGGCC CTGGGG CCCACAT ACACCGATAA (1/13)GTGACAC CCGGCCCG ACTGGGG CCCACATCACT CGATAA (1/13) "GTGACACC TGAC AAT GC TCCT CCT ACACCGATAA (1/13)GTGACACC C TACT GAAGG CTT ACC ACACCGATAA (1/13)GTGAC CT ACTGGGGGAT T ACACCGATAA (1/13)GTGA ATCGCTT TGGGGG T ACACCGATAA (1/13)GTGACACC GTG CTGGGG CC ACACCCATAA (1/13)-GTGAC CCCTTA GAAATA CGCT ACACCGATAA (1/13)GTGAC TCAA TGGGGGATACG CAGG CCGATAA (1/13)GTGAC CC ACTGGGGGAC CCAG ACACCGATAA (1/13)GTGAC CAG CTGGGGG TTACTGAGG ACACCGATAA 11/13)GTGACACC CTTTCGAGGG TCC CAG ACACCCATAA (1/13)

PT B GTGAC CCCTGT ACTGGGGG TACGTGT ACACCGATAA (2/12)GTGACACC GCAACCACC GGGGG T ACACCGATAA (1/12)GTGAC C CCT CTCGGTTG TGGGG TC ACACCGATAA (1/121GTGAC CTAG TGGGGGA CC ACACCGATAA (1/12)ETGAC GCGT ACTGGGGG GGAAAG ACACCGATAA (1/12)+GTGACACC G TCCTAC G TGGGGG CCCTC ACACCGATAA (1/12)GTGACAC COG ATAG AAGCACAA GGGGGAT ACACCGATAA (1/12)GTGACACC TT ACTGGGGG CACCGA T ACACCGATAA (1/12) "GTGAC GGC CTGGGGG GTATAAAGCGTG ACACCGATAA (1/12)GTGACA A ACTGGGGG GTGCCTTT ACACCGATAA (1/12) "OTC TGAC GGGGG TTCAGGAGTTTTT AA (1/12) "

PT D GTGAC CCCCG TCC CT ACTGGGGG CTACCGGGCG CACCGATAA (1/9)"GT CTGGTGCATGCCCTG TACT CCCAG CTAC TC ACACCGATAA (1/9)GT CATCC ACTGGGGG GCCCGT ACACCGATAA (1/9)GTGAC GGCGC ACTGGGGG CGAAGA ACACCGATAA (1/9)GTGAC CCTTGCTCCGGGGC ACT ACTGGGGGAT GCCGT ACACCGATAA (1/91 .GTGAC CCAC ACTGG AGGCGAGCCCCAG ACACCGATAA (1/9)GTGACACC GTC GGGGA CA ACACCGATAA (1/9)GTGAC CAGGTC CTGGGGGATACG CGGGGAC CCGATAA (1/91GTGACACC ACT ACTGGGGGAT TCCTCGT ACACCGATAA (1/9) .

VVal P/N 26-1 P/N 62 P/N D6 P/N J61 FREQUENCY

germline CTCTTGGGGAC GAAATAGT CCTTCCTAC ACTGGGGGATACG ACACCGATAA

PT A CTCTTGGGG CC GGGATAC GAG ACACCGATAA (1/5)CTCTTGGGGA TCA TGGGG ACT ACACCGATAA (1/5)CTCTTGGGGA T PLAT CTGGGGG CCA ACACCGATAA (1/5)CTCTTGGGGAC CTT ATGA GGGGAT T ACACCGATAA (1/5) "CTCTTGGGGAC TGG CCTAC TGGGGG T ACACCGATAA (1/5) "

PT B CTCTTGGGGAC A TCCT T ACTGGGGG GT ACACCGATAA (1/7)CTCTTGGGGAC GA TCCT TAGCC ACACCGATAA (1/7)*CTCTTGGGGAC TCC ACTGGGGG CCGT ACACCGATAA (1/7)wCTCTTGGGGAC CACG TGGGGG G ACACCGATAA (1/7)CTCTTGGGGAC ACGCGC GGGGA GA ACACCGATAA (1/7)CTCTTGGGGAC TGCG TCCT TGTG TGGGG CC ACACCGATAA (1/7)CTCTTGGGGA TTGCGT TGGGGG TO ACACCGATAA (1/7)"

PT D CTCTTGGGGAC CCCGGTG TGGGGG CATCC ACCGATAA (1/6)CTCTTGGGGAC TCGAGGG TTCC A CTGGG AT ACACCGATAA (1/6)

CTCTTGGGGAC GGGACGAGG ACTGGG ACACCGATAA (1/6) "CTCTTGGGGAC AACCTGTCG TCCT GCT ACACCGATAA (1/6)CTCTTGGGGAC G TGGGG C ACACCGATAA (1/6)CTC AAT CT ACTG CCAGCTGC ACACCGATAA (1/6)

Val P/N D61 P/N "6 P/N Qf3 P/N JJ61 FREQUENCYgermline CTCTTGGGGAC GAAATAGT CCTTCCTAC ACTGGGGGATACG ACACCGATAA

PT A CTCTTGGGGAC AT CTA ACCCCGACA GGGAT TA ACCGATAA (5/8)-CTCT C TCCTA AGGGTAG CTGGGG TTCCC ACACCGATAA 13/8)

PT B CTCTTGGGGA TGCG CTA TTTGA GGGGGATACG CGAAATA ACACCGATAA (9/9)

PT D CTCTTGG TGACCCC GAA GTGAC TCC AG ACACCGATAA (8/8)

P/N 261 P/N 62 P/N D63_ P/N J61 FREQUENCYgermline GTGACACC GAAATAGT CCTTCCTAC ACTGGGGGATACG ACACCGATAA

PT A GTGACACC TTGC ACTGGGGGA GACGGG ACACCGATAA (6/6)

PT B GTGA TCG ACTGGGGG GCCCGT ACACCGATAA (B/B)

PT D GTGACACC G TGGGGG GAACCTCGT ACACCGATAA (11/13)GTGACA GACC CCTT GAC CTGGGGGA GACGGGGAGG AA (1/13)GTGACAC CC "CTGGGG CCCACAT ACACCGATAA (1/13)

on February 13, 2018

jem.rupress.org

Dow

nloaded from

same individuals, which exhibited extensive diversity (Fig.3) . The data indicate clonal selection by antigen witholigoclonal expansion within the lesion. The occurrence ofidentical sequences that are nonproductive in a given sampleis likely to represent the nonfunctional allele in the cell un-dergoing clonal expansion .To determine the extent of oligoclonal expansion through-

out the lesion, spatially separated regions ofone biopsy spec-imen were studied . Within each region of the biopsy spec-imen, there was limited diversity of the VJjunctions, althoughthe predominant sequence differed from site to site (Fig. 4) .Given the limited diversity observed in these lesions, we ad-dressed the possibility of a PCR artifact producing the sur-prising results. Extensive precautions were undertaken to avoidPCR contaminations as outlined in the Materials and Methods.Simultaneous PCR and cloning was performed on lesionsand blood, so that finding of limited diversity in lesions wasaccompanied by finding of expected diversity in blood. Thefinding of different predominant sequences in separate areasof the biopsy argues against sample to sample contamina-tion. The finding of limited diversity was obtained by com-plete workup of tissue samples by three separate investigatorsin three separate laboratories . We are confident that the PCRreaction was performed on DNA from -500 cells and noton a single cell based upon: (a) calculation of the numberof y1b cells in the biopsy specimen ; (b) comparison of PCRproduct to PCR titration ofknown amounts of plasmid DNAcontaining 6 gene segments; (c) appearance ofseveral uniquesequences along with the predominant sequence in many ofthe clonings of lesional skin ; and (d) derivation of the samesequence by separate PCR reactions and cloning of the tissuesample.

Extensive Diversity of y/b TCRs in Normal Human Skin.The resident y/b population in normal human skin is small,with few cells identified by immunohistologic techniques (13) .

V6-J6 JUNCTIONAL NUCLEOTIDE SEQUENCES FROM PATIENT B BIOPSY OF LEPROMIN SKIN TEST

688

We wanted to determine whether the limited TCR diversityamong -y1b T cells from leprosy lesions reflected a limiteddiversity within the resident y/b population in skin or ratherclonal expansion of specific TCRs in response to antigenicselection . The y/b TCR repertoire analysis of normal skinindicated the presence of V62-J61 rearrangements, with minorV61Jb1 products and undetectable V62-J63 gene rearrange-ments (Fig. 5) . This indicates that in diseased or normal skin,there is selective localization of y/b T cells expressing theJ61 gene product . Lack ofb chain-bearing J63 rearrangementsin DNA derived from normal skin further supports our con-clusion that the PCR products are derived from a residenty/b population, not from T cells within blood vessel lumina.In contrast to diseased skin, the analysis of TCRjunctionalsequences derived from normal skin indicates extensive diversityequivalent to peripheral blood (Fig. 5) .

Conserved Motif in Predicted Amino Acid Sequences of V,62-J61 Junctions. The predicted amino acid sequences of pre-dominant lesional V62-J61 gene junctions indicated con-servation of amino acids from site to site within a biopsyand from patient to patient (Fig . 6) . The conserved motifcomprised the 5' end of the VJ junction : T .L/V.G.G/D,beginning in the third amino acid position after the secondcysteine of V62. This motif results primarily from the con-tribution of V62, P, and D63 gene segments. A similar motifis seen in the peripheral blood of these individuals, but onlyin those clones that solely use the D63 gene segment (Fig.6) . This motif was rarely observed in normal skin (Fig . 6) .In addition, the length of the VJjunction was relatively con-served in the lesional sequences, as compared to the periph-eral blood .To ascertain whether the conserved motif we observed in

the V62-J61 junctions of leprosy lesions was present in othergranulomatous diseases, we translated the V62-J61 nucleo-tide sequences of y/6 T cells from the lungs of patients with

Clonal Selection of y/b T Cells in Response to a Human Pathogen

Figure 4.

Microheterogeneityofy/8 T cell receptors in lesions.Within different regions of a bi-opsy specimen, distinct clonaljunctional nucleotide sequenceswere identified . Nonfunctionalsequences are marked with anasterisk . Sequences marked bya pound sign were obtained bythree separate PCR reactions andcloning .

Vb2 P/N D61 P/N Db2 P/N D63 P/N J61 FREQUENCY

'germline GTGACACC GAAATAGT CCTTCCTAC ACTGGGGGATACG ACACCGATAA

AREA 1 GTGA TCG ACTGGGGG GCCCGT ACACCGATAA (8/e)

AREA 2 GTGAC GOO CTGGGGGATAC TA ACACCGATAA (10/17)GTGACACC GTA GGGGG TCGGT ACACCGATAA (6/17)

GTGAC CCCC CTGGGG CCCACAT ACACCGATAA (1/17)-

GTGAC TCC ATT ACTGGGGGA GAACAGT ACACCGATAA (1/17)

AREA 4 GTGACACC GT CTGG A ACACCGATAA (15/30) f

GTGACACC A TGGGGGA CCCACGG GATAA (8/30)

GTGACACC GTGAC ACTGGGGG GTACGCTGAG CGATAA (1/30)

GTGACACC OTC CCTA CTGGGGG GGCCCT CACCGATAA (1/30)GTGA TCC C CTGGGGG GGAGGG ACCGATAA (1/30)'

GTGAC GG ACTGGG CACCGATAA (1/30)

GTGACACC TGTA GGGGG G ACACCGATAA (1/30)GTCACA TO CCTT G GGGGGATACG G ACACCOATAA (1/30)

GTGAC CG CCT GGGGG GCG ACACCGATAA (1/30)

Vdl-J61

V61 P/N D61 P/N Db2 P/N D63 P/N J61 FREQUENCY

germline CTCTTGGGGAC GAAATAGT CCTTCCTAC ACTGGGGGATACG ACACCGATAA

AREA 1 CTCTTGGGGA TGCG CIA TTTGA GGGGGATACG CGAAATA ACACCGATAA (9/9)

AREA 3 CTCTTGGG ACGGGGGTGA GGGGA CTCCAGG ACACCGATAA (12/12)

V62-J61

on February 13, 2018

jem.rupress.org

Dow

nloaded from

sarcoidosis (14) . This comparison revealed a relatively highfrequency of the T-L/V.G.G/D motifin V62-J61 junctionalsequences (data not shown) . This motifoccurred in the sameamino acid positions as in sequences from leprosy lesions,but only when the D63 segment was the only D segmentused . The length of the VJjunction in the repeated sequencesof the sarcoidosis patients was conserved, at a length similarto that observed in the leprosy lesion sequences. This wasin contrast to the variable length of the V62-J61 junctionsequences of normal blood y/6 T cells reported (14) .

DiscussionDistinct Microanatomic Patterns of V8 Populations in Leprosy

Lesions. Immunohistologic identification of VS T cell popu-lations in leprosy lesions reveals distinct microanatomic pat-terns for V61- and V62-bearing cells : the dermal infiltratecontaining both populations, but the epidermal y/6 T cellspreferentially expressing the V61 chain . The localization ofV61' cells to epidermis in these mycobacterial lesions sug-gests either homing, retention, and/or in situ expansion ofV61 + cells during the immunopathogenic reaction . This lo-calization of V61+ cells to an epithelial surface may be char-acteristic of the inflammatory process . The gut epithelia of

689

Uyemura et al .

Figure 5 . PCR amplificationand 'PCR V-J junctional sequencesin normal skin . Analysis ofV6 andJ6 gene usage by PCR amplifica-tion revealed product for bothV61-J61 and V62-J61 but notother Vb-J5 combinations (Ashows the results for normaldonor F; donor G gave identicalresults) . Nucleotide sequences ofVb2J61 junctions revealed exten-sive nucleotide diversity for spec-imens from normal donors F andG (B) . Nonfunctional sequencesare marked with an asterisk .

patients with Celiac disease, a gluten sensitive enteropathy,was found to contain V61 + cells (15) .The experimental evidence indicates that V61 and V62

populations may recognize distinct antigens and/or presen-tation molecules. Some human V61 + cells appear to havecytotoxic activity against allogeneic targets ; this cytotoxicityis blocked by a mAb that recognizes a 43-kD protein termedTCT1 (16) . Also, V61 + cells have been reported to recog-nize CD1 molecules (17, 18) . Langerhans cells express CD1molecules, can present antigen, and are increased in theepidermis and dermis of these lesions (8) . Perhaps, the ex-pression of the TCT1 molecule and/or CD1 molecules bydiseased epithelia results in the stimulation of the V61 popu-lation . In contrast, human VS2+ cells expand in responseto M. tuberculosis antigens in vitro. The ability of these sameV62+ cells to proliferate to Daudi cell lines is blocked by apolyclonal serum to a human HSP 58-kD protein (19) .

Limited Diversity ofy/S TCRs in Lesional MicroenvironmentsSuggests Clonal Selection by Antigen . The most striking datapresented here are the limited diversity of the VJjunctionalelements of y/6 T cells in leprosy lesions. The junctionalsequences of y/6 T cells from lesions were not found in therespective individual's peripheral blood, indicating that thelesional sequences do not represent an already existing clonal

on February 13, 2018

jem.rupress.org

Dow

nloaded from

LEPROSY SKIN BIOPSY SPECIMENS

PT A

C D T L H W G R R D K

PT B

C D R L G G P Y T D KC D G L G D T N T D KC D T V G G R Y T D KC D T W N T D KC D T M G D P R T D K

PT D

C D T V G G T S Y T D K

PBMC FROM LEPROSY PATIENTS

PT A

C D T D T P C N T D KC D S M G D T Q A D KC D P L G D P D T D KC D Q L G V T E D T D KC D T Q C S S Y T D KC E S L W G Y T D KC D P L E I R Y T D KC D T H W G I T P T D K

PT B

C D P C T G G T C T D KC D P S R L W H G T D KC D L V G D H T D KC D T V L R G G P H T D KC D G L G G I K R D T D K

PT D

C D P R P Y W G L P G A P KC A G A C P V V P A T H T D KC H R L G G P Y T D KC D G A L G A K N T D KC D P C S G A V L G D A V D KC D P H W R R A P D T D KC D T V G E N T D KC D Q V L G D T R G P D K

NORMAL SKIN

NL F

C D T D P W K T D KC D H I V V T D KC D S L L Y T D KC D T S L L Y T D KC E M G D W G I Y T D KC D T P S L H D T D K

NL G

C D T R V T G G C T D KC D R L G G P Y T D K

Figure 6 .

Predicted amino acid sequences based on V62J61 nucleotidesequences. The predicted amino acid sequences for the predominant le-sional V62-J61 junctions are shown . The conservation in amino acid se-quences, the T-L/V-G-G/D motif beginning in the third amino acid po-sition after the second cysteine of V62, is primarily encoded by the V62,P, and the D63 gene segments. Although D63 gene segment usage is fre-quent in the peripheral blood, the conserved motifdoes not appear in thesame location within the VJjunction with the high frequency of leprosylesional sequences. In addition, the junctional amino acid lengths of pe-ripheral blood V62* cells are more variable, occasionally showing a greaterlength than the lesion-derived sequences. The amino acid sequences ofnormal skin show even greater heterogeneity.

expansion in the peripheral repertoire. It is noteworthy thatin each spatially separated region of the biopsy specimen, alimited number ofjunctional nucleotide sequences were rep-resented multiple times, although the predominant sequencediffered from site to site. Since there was conservation in thepredicted amino acid sequence in various areas of the biopsyand among individuals, we hypothesize that they/6 T cellexpansion in lesions is selected by a limited set of antigenswithin the tissue microenvironment . The limited micro-

690

Clonal Selection of y/6 T Cells in Response to a Human Pathogen

heterogeneity implies that a very small number of y/6 clonesinitiate the oligoclonal expansion and that their progeny donot disperse homogeneously throughout the lesion . Studiesof murine -y/b T cells indicate the positive selection of y/6T cells in the periphery (20) . The study of leprosy lesionsindicates that the clonal selection by foreign antigen occursat the site ofinfection . Furthermore, y/b T cells in the lungsand blood ofpatients with sarcoidosis (14) may similarly re-sult from the oligoclonal expansion to antigen .The nucleotide sequence analysis of lesional y/6 T cells

revealed that while V61J61 T cells utilize multiple Db seg-ments at the VJ junction, the V62-J61 T cells use only theD63 gene segment in 98% of the isolates. This is in contrastto the 67% usage of D63 alone in sequences derived fromthe blood of the same individuals, or the 33% usage in normalskin . The differences in D6 usage suggest a fundamental dis-tinction in the V61 and V62 populations in skin lesions . Theexclusive use of the D63 gene segment with minimal use ofN segments is a feature of y/6 T cells representing an earlyfetal thymic wave (11) . Since V62-D63-J61 cells in leprosyskin lesions were found to contain N segments, they are likelyderived from a later developmental stage when the levels ofterminal deoxynucleotide transferase are higher. The findingof conserved amino acid residues and length of the predomi-nant lesion-derived V62-J61junctions suggests that thejunc-tional region, and specifically the D63-encoded sequence, par-ticipates in antigen recognition (Fig. 4) . This is supportedby murine studies indicating that the junctional sequencesinfluence the specificity of the ylb TCR (21) . When com-pared to the limited TCR diversity displayed by y/b T cellsfrom leprosy lesions, the normal skin data suggests that they/6 T cells in leprosy lesions represent clonal selection fromamong a genetically diverse peripheral blood or resident skinpopulation . The reported diversity of VJ junctions inmycobacterial-reactive murine y/6 T cells derived fromantigen-unselected neonatal thymus indicates that the breadthof the antigen-reactive repertoire is large (22) .Human vs. Murine Skin y/b T Cells.

The normal murineepidermis contains a dendritic cell population that is Thy-1+ and bears y/6 TCRs. This resident population is of ex-tremely limited genetic diversity, with virtually exclusive usageof a single Vy gene and single V6 gene and no junctionaldiversity (1) . Similar to the murine intraepidermal T cells,we find the normal human TCR y/6 skin population to beencoded by a restricted set of V and J genes (V62-J61) . Inresponse to infection, the epidermis is found to contain aspecific TCR 6 population (V61-J61), and all TCRs are oflimited genetic diversity.Our findings indicate clear differences between human versus

murine skin y/6 T cells. The resident human skin popula-tion, known to be significantly smaller than the murine resi-dent population (13), displays extensive, rather than limited,genetic diversity ofjunctional gene elements . Even in humanskin lesions of infectious etiology, the number of epidermaly/6 T cells was relatively few as compared to the residentmurine population . However, limited genetic diversity occursin y/6 T cells responding to infection and appears to be due

on February 13, 2018

jem.rupress.org

Dow

nloaded from

to the oligoclonal expansion of specific TCR-bearing cellsfrom a diverse pool.

Implications.

y/ST cells appear to function as a first lineofdefense against infectious pathogens (23) . This hypothesisis based on the finding of large numbers of y/b T cells atperipheral interfaces, including normal murine skin (24), gut(25, 26), and lungs (3). y/S T cells derived from antigen-unselected murine neonatal thymocytes (27) respond tomycobacterial antigens in vitro. Furthermore, the initial im-mune response to M. tuberculosis (28), Listeria monocytogenes(29), and Trypanosoma cruzi (30) infection is characterized byexpansion of y/S T cells . In humans, y/b T cells accumulate

References

2.

3.

4.

5.

6.

7.

8.

in early lesions of leprosy and leishmaniasis (7). Studies ofhuman y/S have indicated reactivity to a number of infec-tious agents, including mycobacteria (7, 31-33), listeria (34),streptococcus (34, 35), staphylococcus (34-36), plasmodium(37, 38), and herpes viruses (39) . The present data indicatethat the y/b T cell response to infection involves clonal se-lection from a diverse TCRrepertoire and further expansionby antigen in situ . The limitedjunctional diversity in leprosylesions is not consistent with current knowledge of superan-tigen responses (40) . Our data suggest that VJjunctions arecritical for antigen recognition .

We are grateful to Drs. Michael B. Brenner, Mitchell Kronenberg, and Barry R. Bloom for scientificdiscussions and encouragement. We are indebted to Drs. Shuzhuang Lu, Erica Nelson, and Florence Hofmanfor technical expertise.

This work was supported by grants from the National Institutes of Health (AI-22553, AR-40312, AI-28508), the UNDP/World Bank/World Health Organization Special Programme for Research and Trainingin Tropical Diseases (IMMLEP and THELEP), the Heiser Trust, and The Dermatologic Research Founda-tion ofCalifornia, Inc. H. Band is a Leukemia Society ofAmerica Special Fellow and an Arthritis Founda-tion Investigator.

Address correspondence to Robert L. Modlin, Division of Dermatology, 52-121 CHS, UCLA Schoolof Medicine, 10833 Le Conte Avenue, Los Angeles, CA 90024.

Received for publication 8 April 1991 .

Asarnow, D.M., W.A . Kuziel, M. Bonyhadi, R.E . Tigelaar,P.W. Tucker, andJ.P. Allison. 1988 . Limiteddiversity ofgammadelta antigen receptor genes of Thy-1+ dendritic epidermalcells. Cell. 55:837 .Takagaki, Y, A. DeCloux, M. Bonneville, and S. Tonegawa .1989 . Diversity ofgamma delta T-cell receptors on murine in-testinal intra-epithelial lymphocytes . Nature (Loud.). 339:712 .Augustin, A., R.T. Kubo, and G.K . Sim. 1989 . Resident pul-monary lymphocytes expressing the gamma/delta Tcell re-ceptor. Nature (Loud.). 340:239 .Hata, S., K. Satyanarayana, P. Devlin, H. Band, J. McLean,J.L . Strominger, M.B. Brenner, and M.S. Krangel. 1988 . Ex-tensivejunctional diversity ofrearranged humanTcell receptordelta genes. Science (Wash. DC. 240:1541 .Loh, E.Y., S. Cwirla, AT Serafini, J.H . Phillips, and L.L. Lanier.1988 . Human Tcell-receptor delta chain: genomic organiza-tion, diversity, and expression in populations of cells . Proc Nad.Acad. Sci. USA. 85:9714.Takihara, Y, J. Reimann, E. Michalopoulos, E. Ciccone, L.Moretta, and T.W. Mak. 1990. Diversity and structure ofhuman Tcell receptor S chain genes in peripheral blood 7/6-bearing T lymphocytes. J. Exp Med. 169:393 .Modlin, R.L ., C. Pirmez, F.M . Hofman, V Torigian, K. Uye-mura, TH. Rea, B.R . Bloom, andM.B . Brenner. 1989. Lym-phocytes bearing antigen-specific gamma/delta Tcell receptorsin human infectious disease lesions . Nature (Lond.). 339:544 .Dugan, E., R.L . Modlin, and TH. Rea. 1985 . An in situ im-

691

Uyemura et al .

9.munohistological study ofMitsuda reactions. IntJ Lepr 53:404 .Band, H., F. Hochstenbach, J. McLean, S. Hata, M.S. Krangel,andM.B . Brenner. 1987. Immunochemical proof that a novelrearranging gene encodes the T cell receptor delta subunit.Science (Wash. DC). 238:682 .Cooper, C.L., C. Mueller, T.-A. Sinchaisri, C. Pirmez, J. Chan,G. Kaplan, S.M.M . Young, I.L . Weissman, B.R. Bloom, TH.Rea, and R.L . Modlin . 1989 . Analysis of naturally occurringdelayed-type hypersensitivity reactions in leprosy by in situ hy-bridization . J. Exp Med. 169:1565 .Krangel, M.S., H. Yssel, C. Brocklehurst, and H. Spits. 1990.A distinct wave of human Tcell receptor gamma/delta lym-phocytes in the early fetal thymus : evidence for controlled generearrangement and cytokine production.J Exp Med. 172:847 .Sanger, F., S. Nicklen, and A.R . Coulson. 1977 . DNA se-quencing with chain-terminating inhibitors. Proc. Nad. Acad.Sci. USA. 74:5463.Groh, V., S. Porcelli, M. Fabbi, L. Lanier, L.J. Picker, T An-derson, R.A . Warnke, A.K. Bhan, J.L . Strominger, andM.B.Brenner. 1989 . Human lymphocytes bearing T cell receptorgamma/delta are phenotypically diverse and evenly distributedthroughout the lymphoid system . J. Exp Med. 169:1277.Tamura, N., K.J . Holroyd, T Banks, M. Kirby, H. Okayama,and R.G. Crystal . 1990. Diversity in junctional sequences as-sociated with the common human V-y9 and V62 gene seg-ments in normal blood and lung compared with the limiteddiversity in a granulomatous disease. J Exp Med. 172:169 .

on February 13, 2018

jem.rupress.org

Dow

nloaded from

15 . Yoshikai, Y, G. Matsuzaki, T Inoue, and K. Nomoto. 1990.An increase in number of T-cell receptor gamma/delta-bearingT cells in athymic nude mice treated with complete Freund'sadjuvants . Immunology. 70:61 .

16 . Maori-Chouaib, F., C. Miossec, P. Del Porto, C. Flament, F.Triebel, andT Hercend. 1990 . T cell target 1(TCT1) : a noveltarget molecule for human non-major histocompatibilitycomplex-restricted T lymphocytes . J. Exp. Med. 172:1071.

17 . Porcelli, S.A ., M.B. Brenner, J.L . Greenstein, S.P. Balk, C.Terhorst, and P.A . Bleicher. 1989 . Recognition of cluster ofdifferentiation 1 antigens by human CD4-CD8- cytotoxic Tlymphocytes. Nature (Lon4 341:447.

18 . Porcelli, S., M.B. Brenner, and H. Band. 1991 . Biology of thehuman gamma/delta Tcell receptor. Immunol. Rev. In press.

19 . Fisch, P., M. Malkovsky, S. Kovats, E. Sturm, E. Braakman,B.S. Klein, S.D. Voss, L.W. Morrissey, R. DeMars, W.J . Welch,R.L.H. Bolhuis, andPM. Sondel.1990. Recognition by humanVgamma9/Vdelta2 Tcells of a groEL homolog on Daudi Bur-kitt's lymphoma cells. Science (Wash. DC). 250:1269.

20 . Sim, G.K ., and A. Augustin . 1990 . Dominantly inherited ex-pression ofBID, an invariant undiversified T cell receptor deltachain. Cell. 61:397 .

21 . Rellahan, B.L., J.A . Bluestone, B.A. Houlden, M.M . Cot-terman, and L.A . Matis. 1991 . Junctional sequences influencethe specificity of y/S T cell receptors. J. Exp. Med. 173:503 .

22 . Happ, M.P, R.T. Kubo, E. Palmer, W.K . Born, and R.L .OBrien. 1989 . Limited receptor repertoire in a mycobacteriasubset ofgamma/delta Tlymphocytes . Nature (Ivnd). 342:696 .

23 . Janeway, C.A . 1988 . Frontiers of the immune system (newsand views) . Nature (Lond.). 333:804 .

24 . Kuziel, WA., A. Takashima, M. Bonyhadi, P.R. Bergstresser,J.P. Allison, R.E. Tigelaar, and P.W. Tucker. 1987. Regula-tion ofT-cell receptor gamma-chain RNA expression in mu-rine Thy-1+ dendritic epidermal cells . Nature (Loud.). 328:263 .

25 . Goodman, T, and L. Lefrancois. 1988 . Expression of thegamma-delta T-cell receptor on intestinal CD8+ intraepitheliallymphocytes. Nature (Lon4 333:855 .

26 . Bonneville, M., C.A. Janeway, Jr., K. Ito, W. Haser, I . Ishida,N. Nakanishi, and S. Tonegawa . 1988 . Intestinal intraepitheliallymphocytes are a distinct set of gamma delta T cells. Nature(Loud.) . 336:479 .

27 . OBrien, R.L ., M.P. Happ, A. Dallas, E. Palmer, R. Kubo,and WK. Born . 1989. Stimulation of a major subset of lym-phocytes expressing T cell receptor gamma deltaby an antigenderived from Mycobacterium tuberculosis . Cell. 57:667.

28 . Janis, E.M., S.H . Kaufmann, R.H . Schwartz, andD.M. Pardoll .1989 . Activation of gamma delta T cells in the primary im-

692

mune response to Mycobacterium tuberculosis . Science(Wash. DC).244:713 .

29 . Ohga, S., Y Yoshikai, Y. Takeda, K. Hiromatsu, and K.Nomoto . 1990 . Sequential appearance of gamma/delta- andalpha/beta-bearing T cells in the peritoneal cavity during ani.p. infection with Listeria monocytogenes . Eur. J. Immunol.20:533 .

30 . Minoprio, P., A. Bandeira, TM. Santos, and A. Coutinho.1989 . Preferential expansion of Ly-1 B and CD4- CD8- Tcells in the polyclonal lymphocyte responses to murine Tcruziinfection. Int. Immunol. 1:176 .

31 . Holoshitz, J., F. Koning, J.E . Coligan, J. De Bruyn, and S.Strober. 1989 . Isolation of CD4- CD8- mycobacteria-reactiveTlymphocyte clones from rheumatoid arthritis synovial fluid .Nature (Lond.). 339:226 .

32 . Haregewoin, A., G. Soman, R.C . Hom, and R.W. Finberg.1989 . Human gamma delta+ T cells respond to mycobacterialheat-shock protein . Nature (Loud.) . 340:309 .

33 . Kabelitz, D., A. Bender, S. Schondelmaier, B. Schoel, and S.H .Kaufmann . 1990 . A large fraction of human peripheral blood-y/S+ Tcells is activated by Mycobacterium tuberculosis but notby its 65-kD heat shock protein . J. Exp. Med. 171:667 .

34 . Munk,M.E ., A.J. Gatrill, and S.H.E. Kaufmann.1990. Targetcell lysis and 11,2 secretion by gamma/delta T lymphocytesafter activation with bacteria . J. Immunol. 145:2434.

35 . Abo, T, S. Sugawara, S. Seki, M. Fujii, H. Rikilshi, K. Takeda,and K. Kumagai. 1990 . Induction of human TCR gammadelta + and TCR gamma delta-CD2+CD3 - double negativelymphocytes by bacterial stimulation . Int. Immunol. 2:775 .

36 . Rust, CJ., F. Verreck, H. Vietor, and F. Koning. 1990 . Specificrecognition ofstaphylococcal enterotoxin Aby human Tcellsbearing receptors with the Vgamma9 region. Nature (Load.).346:572 .

37 . Roussilhon, C., M. Agrapart, J.J . Ballet, and A. Bensussan.1990 . Tlymphocytes bearing the gamma delta T cell receptorin patients with acute Plasmodium falciparum malaria [letter] .J. Infect. Dis. 162:283 .

38 . Ho, M., H.K. Webster, P. Tongtawe, K. Pattanpanyasat, andW.P. Weidanz. 1990 . Increase d gamma delta T cells in acutePlasmodium falciparum malaria. Immunol. Left. 25:139 .

39 . De Paoli, P., D. Gennari, P. Martelli, V Cavarzerani, R.Comoretto, andG. Santini. 1990 . Gamma delta T cell receptor-bearing lymphocytes during Epstein-Barr virus infection. J.Infect. Dis. 161:1013.

40 . Blackman, M., J. Kappler, and P. Marrack. 1990 . The role ofthe T cell receptor in positive and negative selection of de-veloping T cells . Science (Wash. DC). 248:1335.

Clonal Selection of .y/S T Cells in Response to a Human Pathogen

on February 13, 2018

jem.rupress.org

Dow

nloaded from