T-cell receptor Vβ selectivity in T-cell clones alloreactive to HLA-Dw14

T-Cell Receptor Vfl Selectivity in T-Cell Clones Alloreactive to HLA-Dwl4 Katsuo Yamanaka, William W. Kwok, Eric M. Mickelson, Susan Masewicz, and Gerald T. Nepom

ABSTRACT: The HLA-DR4 subtypes Dw14 and Dw4 are T-cell-defined allospecificities encoded by the DRBI*0404 and DRBI*0401 genes, respectively. Although these allelic subtypes differ in only two amino acids, allorecognition between Dw14 and Dw4-positive individuals is brisk. This provides an opportunity to analyze T-cell receptor (TCR) usage in a very limited and specifically targeted case, namely the Dw4 anti-Dw14 allogeneic T-cell response. The variable (V), diversity (D), and joining (J) region sequences of the TCR/3 chain from two different Dwl4-specific alloreactive T-cell clones derived from a Dw4 donor were examined. Clone EMO25 recognized the Dw14.1, Dw14.2, and Dw15 subtypes, which share a DRB1 polymorphism at codon 71 on a

DR4 background, while clone EMO36 reacted with only the Dwl4.1 subtype associated with polymorphisms at codons 71 and 86. TCR/3 cDNA from each clone was amplified using an anchored polymerase chain reaction (PCR) and subsequently expanded with V/3- and C/3-specific primers for asymmetric PCR and direct DNA sequencing. Both clones were found to express the same TCR V/38.2 gene segment; however, they have several different residues within the V/3-D/3-J/3 junctional regions. V/38 usage was also enriched in polyclonal cells obtained from mixed lymphocyte cultures performed between the Dw4 and Dwl4 responder-stimulator combination from which EMO25 and EMO36 were derived. Human Immunology 33, 57-64 (I992)

ABBREVIATIONS

CDR LCL MHC MLC

complementarity determining regions lymphoblastoid cell line major histocompatibility complex mixed lymphocyte culture

PBMC PCR RA TCR

peripheral blood mononuclear cells polymerase chain reaction rheumatoid arthritis T-cell receptor

INTRODUCTION The allorecognition repertoire o f the ~/3 T-cell receptor (TCR) is very heterogeneous. Studies of both clonal and polyclonal responses have documented a diverse array of fine specificities among alloreactive T cells; recent analyses o f V/3 gene segments used by such cells also show marked diversity [1]. Contributing to this heterogeneity are at least two genetically controlled molecular variables: polymorphisms within the major histocompatibility complex ( M H C ) target molecule, and T C R gene rearrangement. Sequence variation within the H L A - D R 4 family of alFrom the Virginia Mason Research Center (K.Y.; W.W.K.; G.T.N.), Seattle, and the Fred Hutchinson Cancer Research Center (E.M.M. ,"S.M.), Seattle, Washington. Address reprint requests to Gerald T. Nepom, Virginia Mason Research Center, 1000 Seneca Street, Seattle, WA 98101. ReceivedJune 10, 1991; accepted October 18, 1991. Human Immunology 33, 57-64 (1992) © American Society for Histocompatibility and lmmunogenetics, 1992

Ides provides an opportunity to limit these variables and evaluate H L A - T C R interactions recognizing a specific class II sequence element. Two of the H L A - D R 4 subtypes, alleles termed Dw14.1 (DRBI*0404) and Dw14.2 ( D R B I * 0 4 0 8 ) , encode a distinct allospecificity known as Dw14. T cells respond to Dw14, even when such T cells derived from individuals carrying other subtypes of H L A - D R 4 , such as the Dw4 (DRBI*0401) allele. Since Dw4 differs from Dw14.1 at only two codons, encoding amino acids 71 and 86 of the D R B 1 first domain, and differs from Dw14.2 at only codon 71, the allorecognition between Dw4 and Dw14 is likely to be structurally quite restricted to determinants at or near these residues (Table 1). Indeed, site-directed mutagenesis of the D R B I * 0 4 0 4 gene has demonstrated the loss of anti-Dw14 alloreactivity with a single amino acid change at codon 71 [2]. 57 0198-8859/92/$5.00

58

K.

Yamanaka

et al.

TABLE 1 Amino acid sequences for residues 20-90 of the DR/31 alleles 20

CONSENSUS:

*

30

............

DR4 0 ~ I 0

.

DR4 O w l 3 DR4 O w l 4 .

*

0R4

* DR4

I

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.YH .

.

.

.

.

. YH

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Y.H

.

.

.

.

.

.

.

.

.

.

.

.Y H .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

DR

subtypes

.

.

. L - E - C

......

*ORwl4,Owl6

L-E-C

........

associated

60

70

YH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

Dwl5

DR I , D w 2 0

* HLA

.

0w14.2

*ORl,Owl

SO

80

qO

GTERVRFLDRYFHNQEEYVRFOSDVGEYRAVTELGRPDAEYWNSOKDLLEQRRAAVDTYERHNYGVVESFT

DR4 Ow4

*

40

with

Y.H

IY

.

.

.

IY ....

E ........

.

.

.

.

.

.

.

.

.

.

S

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

S ..

.

.

.

.

..

.

.

.

.

.

..

.

.

.

.

.

.

.

.

.

.

K ..............

I .-

.

.

.

.

.

.

.

.

.

.

.

.

-. D. E. . . .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

E . ..

.

.

.

.

.

.

.

..

.

.

.

.

.. . . . . . . .

.

.

.

g .... .

.

.

.

.

.

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

S...............................................

N................................................

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. G. . ..

.

.

. ..

g ..

.

.. . .

G.

.

.

.

.

.

.

A .....

G. . . .

RA.

Allorecognition of the DR/3 residues around codon 71 may proceed via direct recognition of the class II molecule, via recognition of bound peptide, or both [3]. Evidence that some of the predominant allorecognition is directed to the class II sequence alone comes from studies of naturally occurring allelic variation in the population. The DRBI*0404 (Dwl4) sequence around codon 71 (the Dwl4 epitope) is identical to the sequence in four other class II genes, DRBI*0408(Dwl4.2), DRBI*0405(Dwl5), DRBI* 0101(Dwl), and DRBI*1402(Dwl6). Remarkably, many alloreactive clones that recognize Dwl4 cross react with these other alleles as well [4, 5]. This implies that other residues within DR/3, including multiple substitutions that differ among these alleles and appear to influence peptide binding, have little influence on recognition of the "Dwl4 epitope." The sequence encoding the Dwl4 epitope is of interest for studies of HLA and disease, as well. Each of the alleles that shares this sequence is associated with rheumatoid arthritis (RA) [6-9], implying both a common structural basis for the HLA association with RA and also suggesting the possibility that T-cell activation in this disease may be influenced by corecognition of this discrete DR/3 sequence. In structural models of TCR-class II interactions, it has been proposed that parts of the TCR variable region play the major role in directly contacting the MHC molecule and that junctional regions are likely to contact peptides bound by the MHC molecule [10-12]. This raises the possibility that requirements for recognition

of precise MHC or peptide residues might constrain the selection of Vc~ or V/3 elements in the TCR. Indeed, recent data suggest limited human TCR V/3 gene usage in studies of the antigen-specific recognition of myelin basic protein and in the allorecognition of HLA-B27 [13, 14]. In this study we analyzed structural features of TCR V/3 that are associated with recognition of HLA-Dwl4 in a very restricted donor-responder combination. TCR/3 chain mRNA from Dw14 restricted T-cell clones was analyzed by specific amplification and sequencing, and a cDNA amplification procedure was used to estimate the frequency of V/3 usage in polyclonal mixed lymphocyte cultures (MLC) directed against Dw14. MATERIALS A N D METHODS Characterization of Dwl4-specific Alloreactive T-Cell Clones HLA-Dw14 reactive T-cell clones were generated by priming responder peripheral blood mononuclear cells (PBMC) (HLA-A1,2; B8,44; DQw2,7; DR3,4; Dw3,4) with stimulator PBMC (HLA-A1,24; B8, w60; DQw2,8; DR3,4; Dw3,14) and cloned by limiting dilution in the presence ofinterleukin 2. The fine specificity of the DR-directed clones was determined by panel analysis with HLA-DR homozygous lymphoblastoid cell lines (LCL) and homozygous or heterozygous PBMC. Clone EM025 recognized Dw14.1, Dw14.2, and Dw15 on both LCL and PBMC, while EMO36 responded to only Dw14.1 on PBMC. T-cell clones EMO25 and

.

.

.

.

.

TCR V/3 Selectivity

59

EMO36 were screened for specific responses by assaying 1 x 104 cells for proliferative activity after incubation with 1 × l0 ~ PBMC or 2.3 × 10 4 LCL irradiated stimulator cells in U bottom microtiter plates for 66 hours. Cultures were pulsed with 2 ttCi of [-~H] thymidine (specific activity 6.7 Ci/mmol) 18 hours before harvest, and significant restimulation values were determined comparing counts per minute (cpm) from a target stimulator cell with cpm from the positive control (original priming cell) using a centroid cluster analysis. The detailed fine specificity of these clones has been previously reported [15] and is summarized in Fig. 1. cDNA Synthesis and Amplification Oligonucleotides were synthesized from phosphoramidites using an automated DNA synthesizer (Applied Biosystems, Foster City, CA). The oligonucleotides used in these experiments were: CBN:3 'CCGCGGCCGCGCCTGTGGCCAGGCACACCA CBBN: 5 ' C C G C G G C C G C T G T G G G

AG ATCTCTG

3' CTT3

'

agarose gel (NuSieve GTG, FMC, Rockland, ME) and the DNA was phenol extracted and ethanol precipitated [17]. One fifth of the sample was then reamplified with CBBN and AN2 primers, reprecipitated, and labeled with 3ep by random priming (USB, Cleveland, OH); unincorporated deoxynucleoside triphosphate were removed by G-50 Select-D chromatography (3 prime ---, 3 prime, Inc., Boulder, CO). Slot-blot Analysis V/~ cDNAs from different V/~/ families ligated into Bluescript (Stratagene, La Jolla, CA) were kindly provided by Patrick Concannon and nomenclature was described in R.K. Wilson et al. [18]. The V/~ cDNAs were denatured with 0.3 M sodium hydroxide for 30 minutes at 65°C and neutralized by 2 M ammonium acetate. To nitrocellulose paper (Genetrans, Plasco, Woburn, MA) were applied 0.1, 0.3, and 0.5 /.tg of DNA and were fixed with baking for 30 minutes at 80°C. The slot-blot was prehybridized in a solution consisting of 50~, for-

ANI:5'CCGCGGCCGCAAGCTTATCGATCCCCCCCCCCCCCC3' AN2:5 'CCGCGGCCGCAAGCTTATCGAT3' V/38.1 H N : 5 ' T F G C G G C C G C A A G C T T G A A T A T T C C A C A T C T G C T C T 3 '

V/98.1 HN is identical to, while CBN and CBBN are complementary to, the mRNA sequence for TCR cDNA. The consensus sequences of the V/3 and C/3 genes are underlined. Total cellular RNA was isolated from each T-cell clone or from cells of MLC reactions by guanidium thiocyanate-phenol-chloroform extraction method [ 16]. To synthesize the first strand cDNA, 5/.tg of total RNA was heated at 65°C for 5 minutes and reverse-transcribed using 1 /*M CBN oligonucleotide complementary to conserved C/~ sequences (cDNA synthesis system plus, Amersham, Arlington, IL). Centricon 100 (Amicon, Beverly, MA) was used to remove dNTPs and the cDNA was precipitated with ethanol. A 5' poly dG sequence was added by incubation with terminal deoxynucleotidyl transferase (TdT) (BRL; Gaithersburg, MD) in 100 m potassium cacodylate (pH 7.2), 2 mM cobalt chloride, 0.2 mM DTT, and 5 /.tM dGTP (Pharmacia, Piscataway, NJ) for 30 minutes at 37°C. The reaction was stopped by heating 65°C for 5 minutes, and the product was filtered by Centricon 100, then ethanol precipitated. Amplification was performed using 1 ttM CBBN as an internal C/3 primer, and a mixture of upstream primers: 0.1 /*M AN1 (poly dC primer) and 0.9 /.tM AN2 primer (without the poly (dC) tail), which contains several restriction sites (SaclI, NotI, HindlII, ClaI) in 30 amplification cycles of 1 minute at 94°C, 2 minutes at 55°C, and 3 minutes at 72°C. A fraction with size from approximately 400 bps to 800 bps was recovered after electrophoresis of the polymerase chain reaction (PCR) product in a 1.4% low melting temperature

FIGURE 1 Reactivityof anti-Dwl4 T-cell clones EMO23 and EMO36 to LCL and PBMC.

DR

Dw

14,1 14.2 4 15 4 4 I 4 10 4 13 1 1 I 1 20 w14 16 4 4

/

EM025 I

I

I

!

I

i

4 14.1 :i:~:ii i!)!i~:i~i:i~!:!:!~:i:i:i~i !:Z~:i~i i:~:F:i:~:i:!ii~:::::~:i~!:!:~:i:i:!:!:i:i~i~:~F~ii:i:i:!~:i:!:i:!~ • LCL 4 14.2 [] PBMC 4 15 4 4 4 10 4 13 1 1 20 1 EM036 wl 4 16 ] I

0

5

,

I

I

I

li

I

10

15

20

25

30

CPM x 103

35

60

mamide, 5 × SSC, 1 x Denhardt's, 20 mM sodium pyrophosphate, 10% dextran sulfate, 1% SDS, and 100 /*g/ml of denatured salmon sperm D N A at 42°C for 30 minutes, then hybridized overnight with ~P-labeled amplified T C R c D N A or Bluescript D N A for standardization of the results. After hybridization the blot was washed twice in 2 × SSC and 0.1% SDS at room temperature for 15 minutes, then with 0.3 × SSC and 0.1% SDS at 65°C for 30 minutes. Band densities were measured using a densitometer (Bio Image Co., Ann Arbor, MI) and a phosphorimager (Molecular Dynamics Co., Sunnyvale, CA). DNA Sequencing Three micrograms of total R N A from clones EMO25 or EMO36 was reverse-transcribed as mentioned above and amplified with PCR primers V/~8.1 H N and CBBN. The initial double-stranded PCR product was gel purified and then used as a template for a second PCR in which asymmetric PCR amplification was done using 1 /.tM V/~ 8.1 H N plus 0.1 /~M C B B N for generating a positive strand template and using 0.1 /,M V/~ 8.1 H N plus 1/*M C B B N for a minus strand template. Direct sequencing [19] was carried out using VB8.1 H N oligonucleotide or C B B N as primers for sequencing the minus or the plus strand, respectively.

Northern Blot Analysis Total R N A was isolated as described above, and 10/.~g samples were electrophoresed on a 1.2% formaldehyde-agarose gel and transferred to a nitrocellulose membrane (Schleicher and Schuell, Keene, NH). After prehybridizing in 50% formamide, 5 × SSC, 1 x Denhardt's, 20 mM sodium pyrophosphate, 10% dextran sulfate, 1% SDS, and 100 /~g of salmon sperm D N A per milliter, hybridization was performed with V~88 or Cfl c D N A probes at 42°C overnight. After hybridization the filter was washed in 2 × SSC and 0.1% SDS at room temperature, at 42°C for 20 minutes, and three times at 60°C for 20 minutes. Autoradiography was performed with Kodak X-Omat AR film (Rochester, NY). To remove a probe the filter was washed in 1 × SSC plus 0.1% SDS at 80°C for 1 hour, and then was exposed to film for at least 5 days at - 70°C to verify removal of the probe before rehybridization. Band densities were quantitated by using a densitometer (Bio Image Co., Ann Arbor, MI).

K. Yamanaka et al.

nized only DR4 subtypes that carry a Dw14 epitope. The fine specificity of the clones EMO25 and EMO36 was established by panel analysis with LCL and PBMC, as summarized in Fig. 1. EMO25 proliferated when stimulated by the Dwl4.1, Dw14.2, and D w l 5 subtypes of DR4 on both LCL and PBMC. Clone EMO36, by contrast, recognized only those expressing Dwl4.1 on PBMC, and not on LCL whose DRB1 molecule might be presenting with Epstein-Barr virus-derived peptide. TCR/3 Gene Structure Total cellular R N A from EMO25 and EMO36 was reverse-transcribed, and the c D N A was then amplified by anchored PCR, as described above [20]. The amplified c D N A was labeled with 32p and used to probe a slotblot filter with 15 Vf3-containing plasmids. A hybridization signal indicates the V3 family amplified from each T-cell clone. Figure 2 shows V/38 gene usage in both EMO25 and EMO36. Hybridization to C/3 c D N A on the slot-blot is used as an internal positive control. A Vf38 specific primer corresponding to the 5' untranslated region [21] and C/3-specific C B B N primer were then used to amplify full-length cDNA. V/38 c D N A were amplified from each T-cell clone and subjected to asymmetric PCR for D N A sequencing. The complete V~, D/~, and J/3 regions were sequenced. Table 2 shows the D/3, J/3, and part of the Vf3, C3 segment sequences obtained. TCR from EMO25 and EMO36 contain the same V/38.2 gene segment, and both utilize C/32. However, the V/3-Df3-Jf3 junctional regions differ: EMO25 uses J/32.2 and EMO36 uses J/32.3; D-region and N-segment diversity exists as well, so that 10 con-

FIGURE 2 Slot-blot analyses of EMO25 and EMO36 T-cell clones. V/~ cDNAs from different V/~ families ligated into Bluescript were blotted on a filter at 0.1, 0.3, and 0.5 gg. Each filter was hybridized with PCR-amplified V/~ sequences derived from EMO25 and EMO36. 0.5 0.3

EM025

U')

.~ 0.1 .~ 0.5

Lii.!~ii~.~i~ . . . . . . .

~i!~#ii~#i!{iiii~!iilJi ~ii ................. i~ ~i~!ii#ii:iii!: ~i~ ..... ~J~

z

RESULTS T-Cell Clones E M O 2 5 and E M O 3 6 H a v e D i f f e r e n t

Fine Specificities Two alloreactive T-cell clones were obtained from a Dw4-anti-Dw14 priming combination MLC, and recog-

0.3

EM056

0.1 c#

i

VB

'Bluescript

TCR V/3 Selectivity

TABLE 2

61

Nucleotide and amino acid sequence of human TCR3 chains used by Dwl4-specific T-cell clones

vp8.2 EM 025 EM035

GATTCTCAGCTAAGATGCCTAATGCATCATTCTCCACTCTGAAGATCCAG

EMO25 EM035

CCCT~GAACCCAGGGACT~GCTGTGTACTTCT~GCC

Junctional Regions DE Ser Ser EM025

AC;CAG

Ser Gin Gly Pro Trp CTCTCAGGC.ACCATGGG

Ser. Ser. Leu EM036

.....

TTT

CCG~-A3AGCTGTTTg82.2)

Thr Ser Gly Thr

Thr~-s~ Thr Gin Tyr.

-- : T - ( ; C - : AAC

CA-A-ATAC-- ; - - A-

1/3 EM025 EM036

Nonpolar AminoAcids

Ala Gly ~---~ Leu Phe

V3 Usage in Bulk MLC Intrigued by the fact that both EMO25 and EMO36 Tcell clones use members of the Vfl8 family, we studied the V/3 usage in cells obtained from the nonclonal MLC T-cell population from which EMO25 and EMO36 were derived. Total RNA was extracted from responder PBMC without stimulation, and three days after stimulation with irradiated Dwl4 + PBMC, RNA was then reverse-transcribed with the CBN primer and amplified as described above. Slot-blots containing immobilized

Uncharged AminoAcids

c/32 GAGGACCTGAAAAA

TTTGGAGAAGGCTCTAGGCTGACCGTACTG ..... CCC . . . . A-CC . . . . . . . A--G--C

tiguous amino acids in complementarity determining regions (CDR)3 differ. Most of the amino acid sequences of this junctional segment consist of either nonpolar or unchanged polar amino acids in both EMO25 and EMO36. In models of TCR interactions, this CDR3 region is thought likely to play a critical role in contacting MHC class II-bound peptides, whereas the CDR1 and CDR2 Vfl segments are thought to be potential contact sites for the MHC molecule itself [10-12]. We also detected the surface expression of V/38 on both EM025 and EM036 by monoclonal antibody-specific for V38 determinants (T-cell sciences, Cambridge, MA) [data not shown].

(182.a)

[ ] Acidic AminoAcids

Vfi families were hybridized to each ~2P-labeled amplified cDNA, then labeled with Bluescript DNA after stripping the probes. The band densities were quantitared using both a densitometer and a phosphoimager, and density of each Vfi slot probed with Bluescript that represented the quantity of DNA was used to standardize the density of the slot probed with TCR cDNA. The percentage of each V/3 family usage was calculated with the following equation: relative V/3n usage (%) = 100 x

density of V/3n slot probed with TCR cDNA (A) density of V/~n slot probed with Bluescript (B)

X - -

1

N=I ~

As shown in Fig. 3 there is heterogeneity in V/3 segment usage in the starting T-cell populations before allostimulation, with highest usage of V35.1, 8, 15, and 17. After an MLC in the same donor-responder combination as above, in which the HLA-DR typing difference between the responder and the stimulator is Dw4-

62

K. Yamanaka et al.

anti-Dw14, we predicted that most of the activated T cells in this MLC should represent an anti-Dwl4 response [22]. V/38 was significantly increased in this MLC population, and the V/35.1 family represented in the unstimulated starting material was no longer significantly represented. Since the magnitude of the apparent increase in V/38 usage could be influenced by the amplification efficiency, N o r t h e r n blot analysis was also performed on total R N A derived from the day 0 and day 3 MLC cultures. The density of the C/3-specific band that represents total T C R ~ was used to standardize the V38-specific band density. Tlae frequency of V/38 usage was calculated with the following formula: density of the V/38 band frequency of V/38 usage = density of the C/3 band As expected, CB transcripts were increased after stimulation. Hybridization with a V/38 c D N A probe showed that Vfi8 transcripts were also increased (Fig. 4). Though the band in lane 1 is a little weak to accurately quantitate density using a densitometer, V/38 usage was about three times increased after the stimulation, consistent with the slot-blot hybridization analysis.

FIGURE 3 The percentage of V/3 family usage before and after Dw14 stimulation. Each photograph shows the slot-blot filter probed with amplified TCR cDNA from MLC responder cells before and after Dw14 stimulation. The filter was then hybridized with Bluescript to measure the quantity of the DNA in each slot. The percentage was calculated with an equation described in Results.

3O

% 20 1o o

i

i i

|

II

I

1 2 3 4 5.15.267.17.28 9101112 1.3141517C/3 Bluescript

t

VB

3O %

20 10

o 1 '

2 3 4 5.1 5.2 6 7.1 7,2 8 9 10 11 12 13 14 15 17 C/3 Bluescript VB

'

V/38 1

C/3

2

3

2.42

3.56

4

kb 2.4-

1.4-

Density:

0.16

Frequency of V/38 usage density of V/38 band density of C3 band

16.1

Dwl 4 Stimulation Before

After

0.045

0.150 i

FIGURE 4 Northern blot hybridization of MLC responder cells with (lanes 2 and 4) and without (lanes 1 and 3) Dwl4 stimulation. The same filter was probed with V/38 (lanes 1 and 2) and C/3 cDNA (lanes 3 and 4). The density of each band was measured and frequency of V3 usage before and after Dwl4 stimulation was calculated with the formula on the left.

DISCUSSION Allorecognition in vivo, and in MLC, represents a heterogeneous immune response. In a broad sense allorecognition probably represents a composite ofT-cell recognition dependent on peptide interactions as well as on recognition of M H C molecules directly. In a more limited sense the recognition of M H C molecules directly is likely to involve multiple different epitopes characteristic o f the target molecule. We have adopted a "reductionist" experimental approach to analyze one example of such allorecognition, namely the characterization of TCRs used in a focused allorecognition event involving an MLC between a Dw4 responder and a Dw14 stimulator. Allorecognition in this responder-stimulator combination is predicted to be directed against amino acid residues at codon 71 and 86, or at peptides influenced by these residues. It was our hypothesis that T C R V/3 usage in such combination might be correspondingly limited. After finding a suitable class II matched donor-responder combination mismatched for Dw4 versus Dw14, bulk MLC and alloreactive T-cell clones derived from restimulation with Dw14 were analyzed for V/3

TCR V~ Selectivity

usage patterns. The TCR in both clones used members of the V~38.2 gene family and were correspondingly very similar in primary sequence. However, the CDR3 junctional regions in the T C R contributed by VDJ junctional diversity were markedly different in the two clones. V~8 m R N A was also predominant in the bulk MLC in the same donor-responder combination, suggesting that the clonal analysis is representative of the primary T-cell stimulated population. It is not yet known how representative V/38.2 usage is in other anti-Dwl4 allorecognition responses, and it remains to be seen how other donor-responder combinations, particularly in more complex MLC responses, will relate to these data. Nevertheless, this example may help clarify some of the fine structural details relating to the interaction between TCR and M H C class II molecules by comparing the T C R / 3 sequence, the class II sequence, and the recognition allospecificity of these clones. The fine specificity of T-cell recognition for the EMO25 and EMO36 clones is not identical: EMO25 recognizes, in addition to the Dwl4.1 (DRBI~0404) allele, molecules encoded by Dw14.2 (DRBI*0408) and D w l 5 (DRBI~0405). Each of these alleles differs from the Dw4 present in the responder individual by the presence of an arginine at codon 71 (Table 1). Dw14.2 has an additional polymorphism at codon 86, and D w l 5 at codon 57 as well. Since Arg-71 is the principal feature in common to these alleles, it is likely to be the primary stimulating determinant recognized by the EM025 clone. Clone EM036, on the other hand, recognizes only the D w l 4 . 1 stimulator specificity. Thus, its recognition is restricted by residues at both codon 71 and 86. The observation that the TCR for these two clones is identical at V/3 and different throughout the CDR3 junctional sequences can be interpreted to reflect these differences in fine structural specificity for the class II molecule, although the TCRc~ chain might also have an important role. One possible interpretation is that the clones recognize the class II molecule directly and utilize the shared V~8.2 segment for recognition o f the key D w l 4 epitope residues centered on Arg-71, predicted to lie on the c~-helical loop of the class II/3 chain. In this model junctional residues that comprise CDR3 are apparently influenced by the codon 86 polymorphism and may represent altered structural recognition of the class II molecule in this region, which is predicted to lie near the end of the chain c~-helical loop. An alternate interpretation is that these junctional residues primarily contact antigen, in which case the polymorphism at codon 86 may play a critical role in the positioning of the bound peptide. In this case, the V/38.2 selectivity would reflect a preference for the D w l 4 epitope on the class II molecule and

63

the CDR3 residues would reflect peptide variation dependent on constraints imposed by the polymorphism at codon 86. In previous analyses of TCRs generated for allorecognition in the mouse, preferential V segment usage has been described [23, 24], and recently Bill et al. [1] have reported that the V/35 gene family was expressed less frequently, while the V/314, V/315, and V/316 genes were used more frequently, in the response to the murine mutant class I alloantigen A bm~e. Bragado et al. [13] have also found a nonrandom use of V~ gene diversity in an HLA class II alloantigen B27 response. Though there are very few V gene sequences in this study, our findings are compatible with the notion that TCR genes in well-characterized HLA class II restricted alloreactive responses may be similarly selective, at least within an individual responder. The D w l 4 epitope is not only an allospecificity, but also has a strong association with RA. While the most highly associated HLA specificity in RA is HLA Dw4, four other DRB 1 alleles are also highly associated with this disease in different populations. Among Caucasians, the non-Dw4 patients are predominantly Dwl4. In populations with low prevalence of Dw4 and Dwl4, such as the Native American Indians, D w l 6 predominates; in the Japanese, D w l 5 predominates; and in Ashkenazi Jews, DR1 predominates. What all these alleles have in common, and what distinguishes them from all other class II genes, is the D w l 4 epitope, encoded by nucleotide identity bounded by codons 67 and 74. Limited heterogeneity in recognition of Dw 14 as an allospecificity, as suggested in this report, indicates that it may be worthwhile to look for similar patterns of V/3 usage in disease-associated T-cell activation events in D w l 4 positive patients with RA.

ACKNOWLEDGMENTS

We thank Patrick Concannon for useful advice, Jocyndra Wright and Phillip Thurtle for excellent technical assistance, and Nicolette Ducommun and Holly Chase for manuscript preparation. This study was supported by grant AR37296 from the National Institutes of Health.

REFERENCES 1. Bill J, Yag6e J, Appel V, White J, Horn G, Erlich H, Palmer E: Molecular genetic analysis of 178 I-A(bml2)reactive T cells. J Exp Med 169:115, 1989. 2. Hiraiwa A, Yamanaka K, Kwok WW, Mickelson EM, Masewicz S, Hansen JA, Radka SF, Nepom GT: Structural requirements for recognition of the HLA-Dwl4 class II epitope--a key HLA determinant associated with rheumatoid arthritis. Proc Natl Acad Sci USA 87:8051, 1990.

64

K. Yamanaka et al.

3. Lombardi G, Sidhu S, Batchelor JR, Lechler RI: Allorecognition of DR1 by T cells from a DR4/DRwl3 responder mimics self-restricted recognition of endogenous peptides. Proc Natl Acad Sci USA 86:4190, 1989.

14. Wucherpfennig KW, Ota K, Endo N, Seidman JG, Rosenzweig A, Weiner HL, Hailer DA: Shared human T cell receptor V~8 usage to immunodominant regions of myelin basic protein. Science 248:1016, 1990.

4. Seyfried CE, Mickelson E, Hansen JA, Nepom GT: A specific nucleotide sequence defines a functional T-cell recognition epitope shared by diverse HLA-DR specificities. Hum Immunol 21:289, 1988.

15. Mickelson EM, Masewicz S, Nelson J, PetersdorfE, Martin PJ, Hansen JA: T-cell allorecognition sites associated with HLA DR4 alleles. Hum Immunol 1990, Abstracts.

5. Weyand CM, Goronzy JJ: Mapping of allospecific T-cell recognition sites encoded by the HLA-DR4 beta(subscript 1)-chain. Hum Immunol 24:133, 1989. 6. Nepom GT, Byers P, Seyfried C, Healey LA, Wilske KR, Stage D, Nepom BS: HLA genes associated with rheumatoid arthritis. Arthritis Rheum 32:15, 1989. 7. Willkens RF, Nepom GT, Marks CR, Nettles JW, Nepom BS: The association of HLA-Dwl6 with rheumatoid arthritis in Yakima Indians: Further evidence for the "shared epitope" hypothesis. Arthritis Rheum 34:43, 1991. 8. Gregersen PK, Silver J, Winchester RJ: The shared epitope hypothesis: An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 30:1205, 1987. 9. Winchester RJ: The HLA system and susceptibility to diseases: an interpretation. Clin Aspects Autoimmun 1:9, 1986. 10. Claverie JM, Chalufour AP, Bougueleret L: Implications of a Fab-like structure for the T cell receptor. Immunol Today Jan 10:10, 1989. 11. Davis MM, Bjorkman P: T-cell antigen receptor genes and T-cell recognition. Nature 334:395, 1988. 12. Chothia C, Boswell DR, Lesk AM: The outline structure of the T cell a/3 receptor. EMBO J 17:3745, 1988. 13. Bragado R, Lauzurica P, Lopez D, Lopez de Castro JA: T cell receptor V/3 gene usage in a human alloreactive response. J Exp Med 171:1189, 1990.

16, Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156, 1987. 17. Sambrook JF, Fritsch E, Maniatis T. Molecular Cloning. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 6.30, 1989. 18. Wilson RK, Lai E, Concannon P, Barth RK, Hood LE: Structure, organization and polymorphism of murine and human T cell receptor c~ and/3 chain gene families. Immunol Rev 101:149, 1988. 19. Gyllensten UB, Erlich HA: Generation of singlestranded D N A by the polymerase chain reaction and its application to direct sequencing of the HLA-DQA locus. Proc Natl Acad Sci USA 85:7652, 1988. 20. Loh EY, Elliott JF, Cwirla S, Lanier LL, Davis MM: Polymerase chain reaction with single-sided specificity: Analysis of T cell receptor delta chain. Science 243:2 l 7, 1989. 21. Siu G, Strauss EC, Lai E, Hood LE: Analysis of a human V~ gene subfamily. J Exp Med 164:1600, 1986. 22. Reinsmoen N, Bach F: T-cell clonal analysis of HLADR2 haplotypes. Hum Immunol 20:13, 1987. 23. Reimann J, Bellan A: Use of V{3.8 genes in splenic Lyt2 + cytotoxic lymphocyte precursors reactive to bml or bm 14 alloantigen in individual C57BL/6 mice. Eur J Immunol 16:1597, 1986. 24. Kappler JW, Wade T, White J, Kushner E, Blackman M, Bill J, Roehm N, Marrack P: A T cell receptor V/3 segment that imparts reactivity to a Class II major histocompatibility complex product. Cell 49:263, 1987.