Involvement of Class II Chain Amino Acid Residues 85 and 86 in T-Cell Allorecognition David D. Eckels, Mary J. Geiger, Thomas W. Sell, and Jack A. Gorski

A B S T R A C T : Alloreactive T-cell clones were derived by limiting dilution following priming to allogeneic

cells bearing HLA-DR I alloantigens. Clonal specificities were determined by extensive testing on a panel ofallogeneic lymphoblastoid cell lines and by blocking studies with monoclonal antibodies specificfor HLA-DR, -DQ, and -DP class 1I molecules. Out of nine DRi-positive cell lines, three failed to stimulate a subset of the T-cell clones in conventional proliferation assays. Proliferation by all of the clones was blocked by anti-DR antibodies, not by anti-DQ or anti-DP, which was consistent with the conclusion that the HLA-DR molecule was recognized, This DRl-associated polymorphism has been identified as Dw20 by the Tenth International Histocompatibility Workshop. The molecular basis for this altered recognition of the DRI molecule was determined by allele-specific oligonucleotide hybridization and by D N A sequencing studies. The first, second, and third hypervariable regions of all nine DRl-positive cell lines were identical. Valine and glycine were found at positions 85 and 86 of the DR1BI chain in DRI moleculesfrom six of the nine lymphoblastoid celllines, whereas alanine and valine werefound in the three z,ariant (Dw20) DR l-positive cells. By analogy with class I structure, residues 85 and 86 would be located at the extreme C-terminal end of the B-chain ~ helix. Together or separately, these amino acid differences may define a T-cell recognition element on the DR1 molecule serving to contact allospecific T-tell receptors. Alternatively, if allorecognition involves recognition of a self peptide complexed with an allogeneic MHC molecule, then it is possible that the differences T cells recognize on DRI class II proteins arise from peptide-specific interactions with residues 85 and 86. ABBREVIATIONS

LCL MHC MLR MoAb

lymphoblastoid cell line major histocompatibility complex mixed leukocyte reaction monoclonal antibody

PBL TCGF TcR TLC

peripheral blood lymphocyte T-cell growth factor T-cell receptor T-lymphocyte clone

INTRODUCTION The rejection of transplanted foreign tissue is mediated by alloreactive T cells that recognize alloantigens encoded by the donor's major histocompatibility complex (MHC). As an in vitro correlate of donor and recipient incompatibility, when peripheral blood mononuclear cells from allogeneic individuals are combined, vigorous proliferation ensues [1]. This mixed leukocyte reaction (MLR) is controlled by class II molecules encoded in the MHC. Because an exogenous antigen From the Immunogenetics Research Section, The Blood Center of Southeastern Wisconsin. Milwaukee, Wisconsin. Address reprint requeststo Dr. David D. Eckels, ImmunogeneticsResearchSection. The Blood Center of Southeastern Wisconsin, 1701 West Wisconsin Avenue, Milwaukee. WI 53233. ReceivedAugust 28, 1989; acceptedOctober20. 1989.

240 0198-8859/90/$3.50

HumanImmunology27, 240-253 (1990) © AmericanSocietyfor Histocompatibilityand Immunogenetics,1990

T-Cell Allorecognition

241

is not required to elicit a response, the MLR appears to involve recognition of class II alloantigens directly, an apparent violation of the dogma of self-restricted immune recognition. Therefore, how do T cells, which normally recognize antigens in the context of autologous (self) MHC molecules, recognize allogeneic (nonself) MHC molecules? Furthermore, the frequency of allospecific T cells in the MLR is 100-1000 times higher than that of antigen-specific T cells [2]. Therefore, based on numbers, this somewhat paradoxical response appears to be biologically important. In this regard, early studies with alloreactive T cells, cloned from the MLR, revealed an unexpectedly large array of alloantigenic determinants that could not be attributed solely to the extreme genetic complexity of the HLA-D region [3]. However, recent evidence stemming from biochemical studies and the crystallographic structure of class l MHC molecules suggests that MHC molecules may contain endogenous peptides [4-7]. Thus, what appears as additional genetic variation may, in fact, derive from the diversity of MHC-bound, endogenous peptide fragments. Models of allorecognition must therefore account for not only the T-cell interactions with primary structural differences in allogeneic molecules but also the endogenous peptides that may play a role in the formation of alloantigenic epitopes. In attempting to "map" potential sites ofT-cell interaction with class I1 MHC molecules, T-lymphocyte clones (TLCs) were generated against HLA-DR1 alloantigens. HLA-DR1 was selected as a priming antigen because, unlike those encoded by other alleles, only a single DR/31 chain is expressed, along with a nonpolymorphic o~chain [8,9], making it possible to study the diversity ofT-cell responses to a single c~//3 heterodimer. Clones specific for the DQ or DP isotypes were not analyzed since ~ and/3 chains from these isotypes are both polymorphic. Alloreactive T-cell response patterns were analyzed on a panel of DRl-positive and DR 1negative lymphoblastoid B-cell lines (LCLs) and in blocking studies using isotypespecific monoclonal antibodies (MoAbs). A DRl-associated polymorphism was observed and further characterized by sequencing studies and by allele-specific oligonucleotide probes. Changes in amino acids 85 and 86 of the DR13 chain correlated with the alterations in TLC proliferative patterns and thus may be involved in T-cell allorecognition, either by serving as a T-cell receptor (TcR) contact site or by influencing alloantigenic peptides which may bind to class II MHC molecules. MATERIALS AND METHODS

T-eel~ clones. TLCs were generated as previously described [10]. Three series of alloreactive TLCs were derived (Table 1): series 61 (DR2,DRw13 anti-DR1); series 62 (DR2,DRw14 anti-DR1,DR2); and series 63 (DR2,DRw13 antiDR1,DR2). Therefore, DRl-associated allodeterminants should have been primarily recognized. Briefly, peripheral blood lymphocytes (PBLs) were isolated by density gradient centrifugation over Ficoll-Hypaque and primed in the presence of an optimized concentration of irradiated allogeneic PBLs. After 6 days the primed lymphoblasts were fractionated on a 38% Percoll gradient and cloned by limiting dilution in the presence of T-cell growth factor (TCGF) and a fresh alloantigenic challenge. TLCs were expanded by serial exposure to TCGF and allogeneic feeder cells and frozen at - 180°C before being thawed and screened in proliferation assays for their specificities on panels of allogeneic PBLs or LCLs. Characterization of allogeneic stimulator panels. PBLs were HLA-typed using standard protocols for NIH microcytotoxicity, as previously described [3]. HLADP typing was also performed as described elsewhere [11,12] using specifically primed, polyclonal T-cell lines. Homozygous stimulator LCLs and their respective

242

D.D. Eckels et al.

TABLE 1

HLA typing of TLC donors and LCL stimulators

Priming series Series 61 Responder 960 Stimulator 816 Series 62 Responder 485 Stimulator 128 Series 63 Responder 960 Stimulator 128 Sequenced cell lines LCL 9005 (HOM-2) LCL9078 (PMG075)

HLA type A1,25; B18,57; w4,w6;C - , - ; DR2,13; DRw52,-; DQwl,-; DPw4,All,29; Bw56,w62; w6,w6; Cwl,w3; DRI,-; DQwl,-; DPw4,A3,11; BT,w55; w6,w6; Cw3,w7; DR2,14; DRw52,-; DQwl,-; DPw4,A3,-; B35,w44; w4,w6; Cw4,w7; DRl,16; DQwl,-; DPw2,w4 See above See above A3,-; A3,-;

B27,-; B14,-;

w4,w4;Cwl,-; DR1,-; Dwl,-; w6,w6;Cw8,-; DR1,-; Dw20,-;

DQwl,-; DPw4,DQwl,-; DPw3,4

abilities to restimulate the alloreactive TLCs were characterized as part of the Tenth International Histocompatibility Workshop [13].

Proliferation and inhibition assays. Stimulator cells (PBLs or LCLs) were irradiated (3000 or 10,000 rads, respectively) and resuspended at the indicated concentrations in RPMI-1640 tissue culture medium supplemented with 10% human plasma, 2 mM L-glutamine, 25 mM HEPES buffer, 50/.~g/ml gentamicin, 100 ~g/ ml streptomycin, 100 IU/ml penicillin, and 25 IU/ml sodium heparin. Responder TLCs were plated at 1 x 104 cells/well in supplemented medium. Triplicate cultures (200 ~l) were incubated at 37°C in 5% CO2/air and then pulsed overnight with 1.0/~Ci tritiated thymidine. Proliferation, as correlated with incorporation of radiolabel, was measured by liquid scintillation spectroscopy and expressed as the mean ± SEM. Results were evaluated statistically using a maximized T test (Tm~) that has been described elsewhere [14].

Monoclonalantibody (MoAb) blocking assays. MoAbs at varying concentrations were added to cultures to assess the nature of the product recognized by the TLCs. MoAb L243 recognizes a public epitope on HLA-DR molecules [15] and was obtained from the American Type Culture Collection (ATCC), Rockville, Maryland. The MoAb Genox 3.53 recognizes DQwl molecules [16] and was also obtained from the ATCC, as was MoAb W6/32, which recognizes HLA class I molecules [17]. MoAb B7/21 recognizes a monomorphic HLA-DP allodeterminant [18] and was obtained from Dr. Ian Trowbridge through the Salk Institute Cell Bank, LaJolla, California. For blocking studies, stimulator LCLs were irradiated 10,000 rads and plated at 2.5 x 104 cells/well in supplemented medium. MoAbs were titrated starting with 50 k~g/ml followed by the addition of TLCs at 1 x 104 cells/well. The conditions of incubation and assessment of proliferation were as described above.

cDNA synthesis and amplification. Total RNA was isolated from 107 DR1;Dwl (Workshop #9005, HOM-2) and DR1;Dw20 (Workshop #9078, PGM075) B-LCLs by homogenization in 4 M guanidinium isothiocyanate buffer and ultracentrifugation through a discontinuous cesium chloride gradient [19]. cDNA specific for either the DR3 chain or the DRo~ chain was synthesized using 20 ~g of total RNA dissolved in water, 500 units Maloney murine leukemia virus reverse transcriptase (Bethesda Research Laboratories, Gaithersburg, MD), and 10 ~M

T-Cell AUorecognition

243

oligonucleotide primer in 50 mM Tris-HCL, pH 8.3, 75 mM KCI, 3 mM MgC12, 10 mM dithiothreitol, and 500/zM each dATP, dCTP, dGTP, and dTTP. The reaction mixture was incubated for 1 hr at 42°C. The cDNA was transferred to Taq polymerase buffer (10 mM Tris-HC1 pH 8.3, 50 mM KC1, 1.5 mM MgCl2, 0.01% gelatin, 200 mM each dNTP, 1/zM of each primer.) After denaturation (2 min, 100°C), five units ofTaq polymerase (Perkin Elmer Cetus, Norwalk, CT) were added [20]. Thirty cycles of amplification were run in a D N A Thermal Cycler (Perkin Elmer Cetus) with 5-min extensions at 72°C, 1.5-min denaturations at 94°C, and 2-min annealings at 42°C.

DNA sequencing. Amplified D N A was purified by electroelution from a 1% SeaKem GTG agarose gel (FMC BioProducts, Rockland, ME) with 10 × TBE buffer. D N A from one Dwl and one Dw20 LCL was sequenced directly using the dideoxy chain termination method [21] with T7 polymerase (Pharmacia, Piscataway, NJ) and oe-35S-dATP (NEN Research Products, Boston, MA) according to the manufacturer's recommendations for double-stranded templates. Sequences were determined from both strands using the original primers from the polymerase chain reaction and internal primers. D N A from at least four independent rounds of amplification was sequenced.

Hybridization analysis using oligonucleotide probes. Total RNA was isolated from DRl-positive LCLs as described above. Ten micrograms of each RNA preparation in 1 M ammonium acetate was spotted onto a Gene Screen Plus membrane (NEN Research Products) using a slot blot apparatus and baked at 80°C for 20 min. The filters were soaked for 3 hr at 68°C with agitation in 5X Denhardt's solution, 5X SSC, 1% SDS, 10 mM sodium phosphate, pH 7.0, and 5 mM EDTA. Afterwards they were hybridized overnight at 50°C in 10X Denhardts, 5X SSC, 7% SDS, 20 mM sodium phosphate, pH 7.0, and 100/zg/ml salmon sperm D N A with the appropriate radiolabeled oligonucleotide probe (106 cpm/ml): A85V86 (Dw20), V85Gs6 (Dw 1), or VssVs6 (negative control). The filters were washed with agitation once at room temperature and twice at 50°C, 30 min each, in 3X SSC, 10X Denhardt's, 5% SDS, and 70 mM sodium phosphate, pH 7.0 followed by two washes at 50°C, 30 min each, in 1X SSC and 1% SDS. A85V86 and V85Gs6 were washed one additional time at 55°C and 58°C, respectively. The filters were exposed to film with intensifying screens at - 8 0 ° C for 21 hr. The oligonucleotide probes were labeled by incubation for 30 min at 37°C in a reaction mixture consisting of 50 ng of oligonucleotide, 10X kinase buffer, ~/32p_ ATP (NEN Research Products) and 5 units of polynucleotide kinase (Pharmacia). The enzyme was inactivated by the addition of 0.1 M NaC1-0.5% SDS and salmon sperm DNA. The labeled oligonucleotides were purified by passage through a Sephadex G-25 column equilibrated with TE-0.1% SDS. RESULTS TLC Specificities Twenty-six alloreactive TLCs were studied as part of the Tenth International Histocompatibility Workshop. All clones recognized DRl-associated specificities although some TLCs also recognized additional homozygous LCLs expressing different DR molecules. In standard 2-day proliferation assays, TLC proliferative responses to HLA-DR1 could be divided into two groups, those that recognized all DRl-positive LCLs and those that recognized only 6/9 LCLs, consistently failing to respond to LCLs 9002, 9078, and 9079, which were nevertheless DR1-

244

D.D. Eckels et al. DR1+

Lymphoblastoid Cell Lines

Ai.61.288imiiiiimi ~14.71 i m i l l l i m i ~e3.14 m m m m m m m m m ~aSl.142mmmmmmmmm

O

i

E-

,,J_~3.~v6mmmmmmmmm ~173mmmmmmmmm ,~al~9mmmmmmmmm ,~63.,8m m m m m m m m m AL6~.~mmmmmmmmm ~z~4mmmmmmmmm ~1.. mmnmmmmmm ~63~smmmmmmmmm ~z~4mmmmmmmmm ~,~.~3.~ m m n m m m m m m A~6~ m m m m m m m m m ~z~osmmmmmmmmm A~6~.~54mNNNNNNNN ~z178mmmmmmmmm ~26mmmmmmnD~ ~6zyo2mmmmmm~::

,~62.24mmmmmm~: ~.:7ommmmmm~D~ ~z~77mmmmmm~:~ A~3.73 m m n m m n ~ : ~ ~8o mmmmmm~~ ~.267mmmmmm~~ Dwl

Dw20

FIGURE 1 TLC reactivity patterns on DRl-positive allogeneic LCL panel from Tenth International Histocompatibility Workshop. ( i ) Positive responses. (E]) Negative responses. TLCs and LCLs highlighted in italics are emphasized in the text for either characterization or sequencing studies.

positive by serologic typing criteria (Figure 1). Because HLA-Dw specificities are defined by T-cell proliferative patterns, such results form the basis of the newly described Dw20 specificity [13]. Two TLCs typify these observations: TLC AL63.75 recognized DRl-positive LCLs of either the Dwl or Dw20 subtypes, but no additional specificities. TLC AL61.102 recognized D w l but not Dw20 stimulator LCLs (Figure 2). The molecular specificities of these alloreactive TLCs were determined by blocking experiments with MoAbs specific for the DR, DQ, or DP class II isotypes (Figure 3). When stimulated by a DR1;Dw I LCL stimulator, all TLCs were blocked by the L243 MoAb (anti-DR) and not by Genox 3.53 (antiDQ), B7/21 (anti-DP), or w6/32 (anti-class I).

T-Cell Allorecognition

245

S e q u e n c e Analysis o f t h e H L A G e n e s f r o m a D w l and a D w 2 0 Cell T o determine if sequence differences in the H L A - D R 1 product could explain the different T-cell proliferative patterns, the gene sequences were analyzed. As an initial level of analysis, R N A from DR1 LCLs was probed with allele-specific oligonucleotide probes corresponding to the hypervariable regions of the D R 1 3 chain found at amino acid positions 9 - 1 4 , 27-33, and 70-75. N o variations were observed (data not shown), suggesting that sequence differences between D w l and D w 2 0 LCLs, if present, were located outside these regions. Therefore, D R FIGURE 2 Representative TLC specificity patterns on LCL panel cells. Results (mean cpm ± SEM) are averaged for LCLs within a given DR category. Values in parentheses indicate numbers of different stimulators included in each group. Only responses to DR1positive LCLs are considered positive by statistical criteria established previously [ 10,13 ].

DR1;Dwl(6) DR1 ;Dw20(5) m m DR2(15) m m t8

S DRS(~O) m o DR4(1 S) m : DR6(1)

~09

o~

~

DR7(11 ) DR8(9) m DRg(6) m m

DRy1(11) m DR12(1) m OR~ s(9) m !

DR1 4(6)

AL61.102

m L

0

5

1 1~,

10 i

DR1;Dwl(6) DR1;Dw20(5) ~o DR2(15)

I

III

DRS(~ O) S DR4(15) S DR6(]) m E

+~ DRT(1 1) m if) 0Rs(9) m o ~ DR9(6) m DRl1(1t) DR12(1) m DR~ S(9) m DR14(6) m 0

AL65.75

5

10

15

CPM (xlO -'7}

20

25

0

5

10

Oenox 3.53 GDQ

GDR

I

L243

AL61.102 ---IAL63.75

GDP

B7/21

T

GCIoss I

w6/32

T

vledium

III

FIGURE 3 MoAb blocking patterns with representative TLCs. MoAbs consisted of L243 (anti-DR), Genox 3.53 [anti-DQ), B7/21 (anti-DP), and w6/32 (anti-class I). Results reflect mean cpm -+ SEM of triplicate values.

G_ (D

x

w---

O

I

if3

15

Na 4~

F-Cell Allorecognition

247

I

2

:3

9001 9003 9004 9005 9006 9080 9002 9078 9079

I

FIGURE 4 Hybridization analysis of RNA from DRl-positive Workshop LCLs using allele-specific oligonucleotide probes. Lane 1:AssV86 (Dw20). Lane 2:VssG86 (Dw 1). Lane 3:VssV86 (nonhybridizing control). The hybridization patterns correlate with the TLC proliferative patterns shown in Figure 1.

gene sequences were determined for a D w l (LCL 9005) and a Dw20 (LCL 9078) cell line, respectively. The first domain of the D w l sequence was identical to that previously reported [8], however the Dw20/3-chain sequence differed from the D w l sequence at four positions (Table 2). These nucleotide variations result in two amino acid changes, an alanine for a valine at position 85 and a valine for a glycine at position 86 (Dwl = VssG86; Dw20 -- A85V86). No additional polymorphisms were found in the second domain of the/3 chain or in the first domain of the ee-chain gene from the Dw20 cell line. To determine if this sequence polymorphism (AssV86) was common to all the Dw20 cells, oligonucleotide probes corresponding to the VssG86 and AssV86 sequences were used to hybridize with the DR1 B-LCL stimulator panel. R N A from all of the Dw20 cells hybridized with the AssV86 probe, whereas that from D w l cells bound only the VssG86 probe, as expected (Figure 4). A control probe (V85Vs6) failed to hybridize. Thus, the altered TLC response pattern correlated exactly with the change from ValssGlys6 in DR1;Dwl molecules to AlassVa186 in DR1;Dw20/3 chains. It is interesting to note that the AssV86 polymorphism is also found in HLADR2;Dw21 and DR2;Dw22/3-chain genes [2 2 ], however, B-LCLs expressing such molecules are unable to stimulate those T-cells such as AL63.75 that respond to the HLA-DR 1 ;Dw20 alloantigens [ 13 ].

Position Amino acids Consensus DR1 Dwl Dw20 Substitutions

Position Amino acids Consensus DR1 Dwl Dw20

Position Amino acids Consensus DR1 Dwl Dw20

Position Amino acids Consensus DR1 Dwl Dw20

TABLE 2

V GTG

D GAC

L CTC

L CTG

G E GGG GAG

R CGG .

1 G GGG --

E GAG

Y TAC

R CGG .

T ACC

.

70 Q CAG

R CGG

L TTG

R CGA

R AGG

A GCG

L CTG

P CCA

.

.

R CGG

50 V GTG

E GAA

R CGT

.

A GCC

T ACG

R AGA

F TTC

A GCG

E GAG

30 C TGC .

L TTG

.

V GTG

L CTG

1 ATC

W TGG

D w l and D w 2 0 nucleotide and amino acid sequences

D GAC

G GGG

Y TAT

10 Q CAG

T ACC

R CGG

N AAC

L CTT

T

Y TAC

P CCT

Q CAA

K AAG

C TGC

D GAT

E GAG

F TTT

--

80 R AGA

A GCC

E GAG

E GAA

H CAC

E GAG

S TCC

C TGT

N AAC

60 Y TAC

V GTG

H CAT

Y TAC

W TGG

R CGC

F TFC

G GGG

N AAC

40 F ~FC

F TTC

CA

85 V GTT

S AGC

D GAC

-TG V

86 G GGT

Q CAG

S AGC

20 N G AAT G G G

E GAG

K AAG

D GAC

T ACG

I,J

S AGC

D GAC

V GTG

E GAG

o~

H CAC

F TYC --

Position Amino acids Consensus DR1 Dwl Dw20

Position Amino acids Consensus DR1 Dwl Dw20

H CAC ---

F TTC .

Position Amino acids R Consensus DR1, C G G Dwl Dw20

Position Amino acids Consensus DR1 Dwl Dw20

Position Amino acids Consensus DR1 Dwl Dw20

P CCA

Q CAG .

N AAC

H CAC

90 T ACA --

S AGT ---

T ACC

G GGC

N AAC

V GTG

180 V GTG

L CTG .

Q CAG

L CTC

Q CAG

V GTC

94 R CGA

T ACG

S AGC

160 V M G T G ATG .

E E GAA GAG

L CTG

R CGG

P CCT

L CTG

K AAG

C TGC

V GTT

L CTC

E GAA

140 A GCT

S TCT

E GAG

T ACA

T ACA

G GGG

V GTG

P CCT

V GTG

V GTY

V GTG

120 S AGT

K AAG

E GAA

P CCT

V GTG

G GGT

V GTG

W TGG

R CGG

S TCC

F TI'C

100 T ACT

S AGT

T ACA

Y TAT

V GTG

L CTG

G GGC

P CCT

G E GGA GAG

G GGC

P CCA

Y TAT

170 V GTT

1 ATC

S AGC

S TCA

Y TAC

Q CAG

1 ATT

K AAG

T ACC

150 N AAT

E GAA

T ACC

C TGC

G GGA

V GTC

Q CAG

Q CAA

D GAT

130 R AGG

P CCC

v GTG

W TGG

W TGG

L CTG

IxO

E GAG

T ACC

F TTC

110 Q CAG

250

D.D. Eckels et al.

DISCUSSION We have shown that T-cell allorecognition of two nearly identical class II molecules is profoundly influenced by conservative changes in the DRfl chain at positions 85 and 86. The new allelic product described is identical to the previously published HLA-DRlfl chain except that an alanine and valine are found in place of a valine and glycine at amino acid positions 85 and 86, respectively. The cells from which these sequences have been derived were provisionally typed as Dwl and Dw20 based on TLC reactivity patterns in the Tenth International Histocompatibility Workshop [13]. The A85V86 sequence is associated with BLCLs expressing the Dw20 polymorphism, whereas the V86G86 polymorphism is found on cells bearing Dwl. Hurley and colleagues have described an identical DR1B gene present in American blacks [23]. A85 and Vs6 as well as the flanking sequences are also found in DR2;Dw21 and DR2;Dw22flj chains which probably arose from a gene conversion event [24]. The T-cell-defined Dw20 polymorphism, which fails to stimulate some TLCs, is likely to reside on the DR molecule rather than on other class II isotypes because clonal proliferation was blocked by anti-DR MoAbs and not by anti-DQ or anti-DP. This is important since some Dw subtypes have been shown to derive from DQ-based polymorphisms [25]. Since T-cell recognition of nominal antigen can be influenced by either the MHC molecule recognized, the peptide bound to the MHC or a combination of the two, the polymorphisms at positions 85 and 86 in the DR 3 chain may serve as contact residues for the TcR or endogenous peptides. If class II molecules can be correctly understood in relation to the class I structural paradigm as proposed by Brown and colleagues [4], then residues 85 and 86 would be located at the extreme carboxy terminus of the 3-chain o~helix where an indirect effect on TcR recognition may be localized [7,26]. Therefore, if amino acids at positions 85 and 86 do not interact directly with TcR, then they may affect either the orientation of endogenous peptides or exclude other peptides necessary for T-cell allorecognition. This is consistent with the notion that allorecognitionn may involve more than TcR interactions with allogeneic MHC molecules and may also include a peptide component, as conceptualized by Matzinger and Bevan [2]. In support of this hypothesis, we have found that Dw20 molecules bind a radiolabeled influenza hemagglutinin peptide much less efficiently than Dwl (Eckels, manuscript in preparation). However, we cannot formally exclude the possibility that either or both residues at positions 85 and 86 interact directly, if not exclusively, with TcR. If an endogenous peptide is part of the alloantigenic complex recognized by T cells, then the way it interacts with TcR and MHC molecules could have important consequences for T-cell activation. Where DRl-specific T cells fail to recognize Dw20 molecules (Figure 1), binding of endogenous peptide to Dw20 molecules would have to inhibit T-cell recognition, perhaps by replacing another peptide that is permissive or required for allorecognition. We have demonstrated this possibility by showing that an exogenous influenza hemagglutinin peptide can inhibit allorecognition of HLA-DR 1 by DR 1-specific TLCs [27]. Intriguingly, the Dw20 specificity identified in the Workshop is highly associated with HLA-B14 suggesting the possibility that presentation of class I, B 14-derived peptides on Dw20 molecules may block T-cell recognition. Similar observations have been made by Davis and coworkers [28], who have described a variant DR1 molecule. derived from B14-positive cells, that fails to stimulate some T cells. Although class II-restricted presentation of class I peptides has not been reported, class I peptides have been used to inhibit class I-restricted cytotoxicity [29]. Finally,

T-Cell Allorecognition

251

Marrack and Kappler have described alloreactive T cells that are capable of distinguishing isogeneic M H C molecules found on cells derived from different tissues [30] and attribute this p h e n o m e n o n to tissue specific peptides which occupy the MHC's binding site. Further support for peptide involvement stems from the reactivity pattern of AL63.75. Since AL63.75 is responsive to both D w l and Dw20, the changes at amino acid positions 85 and 86 appear to have little effect. However, ifallorecognition involves only nonself M H C contact sites, then AL63.75 should also be stimulated by DR4;Dw14 and DRw6;Dw16 molecules because these alleles share identical sequences with D R 1 ; D w l in the fl-chain c~ helix except at residues 85 and 86 [31,32]. TLC AL63.75, however, does not respond to D w l 4 or D w l 6 LCLs, presumably because DR1, DR4, and DRw6 are very different in the/3sheet regions, which are thought to influence the binding o f antigenic peptides. Therefore, it appears that AL63.75 may require both a specific c~-helical sequence (histotope) and a peptide (epitope) which is bound only by DR1 and not by DR4 or DRw6. Thus allorecognition may require a tripartite structure consisting of TcR, allogeneic MHC, and endogenous peptides, which is similar to that required for nominal antigenic recognition.

ACKNOWLEDGMENTS

We thank Michael Bull, Suzanne Goodacre, and Roxanne Pinzl for expert technical assistance. This work was supported by grants AI22832 (D.D.E.) and AI26085 (J.A.G.) and by the Blood Center Research Foundation. D.D.E. is recipient of Research Career Development Award AI00799.

REFERENCES 1. Bach F, Hirschorn K: Lymphocyte interaction: A potential histocompatibility test in vitro. Science 143:813, 1964. 2. Matzinger P, Bevan MJ: Why do so many lymphocytes respond to major histocompatibility antigens? Cell Immunol 29:1, 1977. 3. Eckels DD, Hartzman RJ: Characterization of human T-lymphocyte clones (TLCs) specific for HLA-region gene products. Immunogenetics 16:117, 1982. 4. BrownJH,Jardetzky T, Saper MA, Samraoui B, Bjorkman PJ, Wiley DC: A hypothetical model of the foreign antigen binding site of class II histocompatibility molecules. Nature 332:845, 1987. 5. Buus S, Sette A, Colon SM, Grey HM: Autologous peptides constitutively occupy the antigen binding site on Ia. Science 242:1045, 1988. 6. Bjorkman PJ, Saper MA, Samraoui B, Bennett WS, Strominger JL, Wiley DC: Structure of the human class I histocompatibility antigen, HLA-A2. Nature 329:506, 1987. 7. Bjorkman PJ, Saper MA, Samraoui B, Bennett WS, Strominger JL, Wiley DC: The foreign antigen binding site and T cell recognition regions of class I histocompatibility antigens. Nature 329:512, 1987. 8. Tonnelle C, DeMars R, Long EO: DO/3: A new/3 chain gene in HLA-D with a distinct regulation of expression. EMBO J 4:2839, 1985. 9. Bohme J, Andersson M, Andersson G, Moiler E, Peterson P, Rask L: HLA-DR

252

D . D . Eckels et al.

genes vary in number between different DR specificities, whereas the number of DQ /3 genes is constant. J Immunol 135:2149, 1985. 10. Rosen-Bronson S, Johnson AH, Hartzman RJ, Eckels DD: Human allospecific TLCs generated against HLA antigens associated with DR1 through DRw8. I. Growth and specificity analysis. Immunogenetics 23:368, 1986. 1 l. Shaw S, Johnson AH, Shearer GM: Evidence for a new segregant series of B cell antigens that are encoded in the HLA-D region and that stimulate secondary allogeneic proliferative and cytotoxic responses. J Exp Med 152:565, 1980. 12. Shaw S, Pollack MS, Payne SM,Johnson AH: HLA-linked B cell alloantigens of a new segregant series: Population and family studies of the SB antigens. Hum Immunol 1:177, 1980. 13. Eckels D, Sell T, Eiermann T, Nikaein A: DR1: Antigen report of the cellular studies of the Tenth International Histocompatibility Workshop. In Dupont B (ed): Immunobiology of HLA, volume I. New York, Springer-Verlag, 1989, p 502. 14. Rosen-Bronson S, Johnson AH, Hartzman RJ, Eckels DD: Human allospecific TLCs generated against HLA antigens associated with DR1 through DRw8. II. Populations analyses and blocking studies with monoclonal antibodies. Immunogenetics 24:286, 1986. 15. Lampson LA, Levy R: Two populations of IaJlike molecules on a human B-cell line. J Immunol 125:293, 1980. 16. Brodsky FM: A matrix approach to human class I1 histocompatibility antigens: Reactions four monoclonal antibodies with the products of nine haplotypes. Immunogenetics 19:179, 1984. 17. Brodsky FM, Parham P, Barnstable CJ, Crumpton MJ, Bodmer WF: Monoclonal antibodies for analysis of the HLA system. Immunol Rev 47:3, 1979. 18. Robbins PA, Evans EL, Ding AH, Warner NL, Brodsky FM: Monoclonal antibodies that distinguish between class II antigens (HLA-DP, DQ and DR) in 14 haplotypes. Hum Immunol 18:301, 1987. 19. Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ: Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294, 1979. 20. Saiki RK, ScharfS, Faloona F, Mullis KG, Horn GT, Erlich HA, Arnheim N: Enzymatic amplification of 3-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230:1350, 1985. 21. Sanger F, Nicklen S, Coulson AR: DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 80:6972, 1977. 22. Wu S, Saunders T, Bach FH: Polymorphism of human Ia antigens generated by reciprocal intergenic exchange between two DR 3 loci. Nature 324:676, 1986. 23. Hurley CK, ZiffBL, Silver J, Gregersen PK, Hartzman R, Johnson AJ: Polymorphism of the HLA-DR1 haplotype in American blacks. Identification of a DR1 /3-chain determinant recognized in the mixed lymphocyte reaction. J Immunol 140:4019, 1988. 24. Gorski J, Mach B: Polymorphism of human Ia antigens: Gene conversion between two DR/3 loci results in a HLA-D/DR specificity. Nature 322:67, 1986. 25. Bach FH, Reinsmoen NL: The role of HLA-DR and HLA-DQ products in T lymphocyte activation and in contributing to "Dw specificities." Hum Immunol 16:271, 1986. 26. Nathenson SG, GeliebterJ, Pfaffenbach GM, ZeffRA: Murine major histocompatibility complex class-I mutants: Molecular analysis and structure-function implications. Ann Rev Immunol 4:471, 1986.

F-Cell Allorecognition

253

27. Eckels DD, GorskiJ, RothbardJ, Lamb JR: Peptide mediated modulation ofallorecognition. Proc Natl Acad Sci USA 85:8191, 1988. 28. DavisJE, Rich RR, Van M, Le HV, Polack MS, Cook RG: Defective antigen presentation and novel structural properties of DR1 from an HLA haplotype associated with 21-hydroxylase deficiency. J Clin Invest 80:998, 1987. 29. Clayberger C, Parham P, Rothbard J, Ludwig DS, Schoolnik GK, Krensky AM: HLA-A2 peptides can regulate cytolysis by human allogeneic T lymphocytes. Nature 330:763, 1987. 30. Marrack P, KapplerJ: T cells can distinguish between allogeneic major histocompatibility complex products on different cell types. Nature 332:840, 1988. 31. CairnsJS, CurtsingerJM, Dahl CA, Freeman S, Alter BJ, Bach FH: Sequence polymorphism of HLA DRBI alleles relating to T-cell recognized determinants. Nature 317:166, 1985. 32. Gorski J: First domain sequence of the HLA DRB1 chain from two HLA DRw14 homozygous typing cell lines: TEM (Dw9) and AMALA (Dw16). Hum Immunol 24:145, 1989.