Transfected murine cells expressing HLA class II can be used to generate alloreactive human T cell clones

ELSEVIER

JOURNALOF IMMUNOLOGICAL METHODS Journal of Immunological Methods 169 (1994) 25-33

Transfected murine cells expressing H L A class II can be used to generate alloreactive human T cell clones Anthony N. Warrens *, Tricia Heaton, Sid Sidhu, Giovanna Lombardi, Robert I. Lechler Department of Immunology, Royal Postgraduate Medical School, Hammersmith Hospital, DuCane Road, London W12 ONN, UK

(Received 13 July 1993, revised received 8 October 1993, accepted 11 October 1993)

Abstract Alloreactive human T cells are conventionally generated in vitro using peripheral blood mononuclear cells (PBMCs). The disadvantage of such an approach is that PBMCs express multiple H L A class II molecules and, as a consequence, it is difficult to generate T cells specific for an individual H L A alloantigen. This paper describes a technique in which T cell clones can be generated using stimulators which do express only one alloantigen. This has permitted the generation of HLA-DR-specific T cell clones and will be applied to produce T cell clones specific for other isotypes which cannot easily be obtained using other techniques. Murine DAP.3 cells were transfected with cDNAs encoding human class II molecules and used to stimulate primary alloresponses by purified human CD4 ÷ T cells. The cloning of these T cells provided a good yield of cells allospecific for the class II molecule expressed by the transfected cells. A large percentage of the T cell clones were able to recognise human cells, suggesting that specificity for DR-bound peptides of mouse origin does not limit the applicability of this approach. Despite having been raised against mouse stimulators cells, the responses of the T cell clones to alloantigen-expressing human B cell lines were profoundly inhibited by anti-human LFA-3 monoclonal antibody. The possible mechanisms responsible for these results are discussed. Key words." T cell clone; Alloreactivity; HLA; Transfected murine fibroblast

I. Introduction T h e g e n e r a t i o n o f allospecific h u m a n T cells in vitro c o n v e n t i o n a l l y relies o n t h e use of p e r i p h -

* Corresponding author. Tel.: 081 743 8349; Fax: 081 743 8602. Abbreviations: B-LCL, B lymphoblastoid cell lines; FCS, foetal calf serum; ICAM, intercellular adhesion molecule; LFA, lymphocyte function associated antigen; PBMCs, peripheral blood mononuclear cells; PHA, phytohaemagglutinin; rlL-2, recombinant interleukin-2.

eral b l o o d m o n o n u c l e a r cells (PBMCs). T h e disa d v a n t a g e of such an a p p r o a c h is t h a t P B M C s express m u l t i p l e H L A class II m o l e c u l e s a n d , as a c o n s e q u e n c e , it is difficult to g e n e r a t e T cells specific for an i n d i v i d u a l H L A a l l o a n t i g e n . By m a t c h i n g host H L A t y p e with d o n o r as closely as possible, t h e n u m b e r o f a l l o g e n e i c m o l e c u l e s can b e m i n i m i s e d . H o w e v e r , g e n e t i c l i n k a g e is so close (e.g., b e t w e e n H L A - D R a n d - D Q ) (Trowsd a l e et al., 1991), t h a t it is very difficult to find c o m b i n a t i o n s in w h i c h t h e s t i m u l a t o r e x p r e s s e s

0022-1759/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0022-1759(93)E0271-I

26

A.N. Warrenset al. / Journal of lmmunological Methods 169 (1994) 25-33

only a single HLA alloantigen. An alternative approach is to stimulate alloreactive T cells in the presence of antibodies specific for one (or more) alloantigen in order to bias the response towards others. There is a particular problem in the generation of human T cell clones alloreactive against HLA class II molecules, since the response to DR dominates over that to DQ and DP, just as it dominates in the restriction of antigen-specific responses (Altmann et al., 1991). Techniques which might permit the generation of clones specific for DQ or DP are limited (Merkenschlager et al., 1991). We have developed a system in which transfected murine fibroblasts expressing a single MHC molecule are used as allostimulators of human T cells in vitro. In this way a population of T cells specific for a single HLA-DR class II molecule can be generated with no requirement for further selection. It should be possible to employ this technique to generate a response against molecules which, using conventional techniques, is overshadowed by anti-DR reactivity. 2. Materials and methods 2.1. Transfection of murine fibroblasts

cDNAs of DRA (gift of Dan Denney, Stanford University, USA), DRBI*0406 (gift of Peter Gregerson, New York University Medical Center, USA), and DRB1*1101 (gift of Ben Schwartz, National Institutes of Health, Rockville, MD, USA), in the pJ4fl, pcEXV-l-neo and pH/3APr1-neo expression vectors respectively, were transfected into the L cell, DAP.3 using the standard calcium phosphate precipitation technique. Geneticin (Sigma, Poole, UK) selection was undertaken after 48 h. Colonies of surviving cells were pooled and analysed by flow cytometry (EPICS Profile, Coulter Electronics, Hialeah, FL, USA). The fluorescence activated cell sorter (Coulter EPICS Multiparameter Sensor System, Coulter Electronics, Hialeah, FL, USA) was used to select for high levels of expression; cloning by limiting dilution was then used to obtain cells with a stable level of expression. Cells were maintained by frequent subculture in DMEM supplemented

with sodium bicarbonate, L-glutamine 2 mM, penicillin 50 IU/ml, streptomycin 50 tzg/ml (all: Flow, Irvine, Scotland) and 5% foetal calf serum (FCS) (Globepharm, Esher, UK). 2.2. B cell lines

EBV-transformed B lymphoblastoid cell lines (B-LCLs) from the panel of the 10th International Histocompatibility Workshop were cultured in RPMI 1640 medium supplemented with sodium bicarbonate, L-glutamine 2 mM, penicillin 50 IU/ml, streptomycin 50/zg/ml (all: Flow, Irvine, Scotland) and 10% FCS (Globepharm, Esher, UK) and were passaged regularly. 2.3. Antibodies

The antibodies used in this study were goat anti-human immunoglobulin (originally prepared by Nick Davey, RPMS, UK), L243, a mouse antiD R a monoclonal (ATCC, Rockville, MD, USA), mouse anti-human CD8 (ATCC, Rockville, MD, USA), mouse anti-human Leu 19 (Becton Dickinson, Oxford, UK), mouse anti-human immunoglobulin (Sigma, Poole, UK), mouse antihuman CD4 (Dako, High Wycombe, UK), and mouse anti-human LFA-3 (ATCC, Rockville, MD, USA). The antibodies used to isolate CD4 + T cells (see below) were affinity-purified. Those used in proliferation assays were concentrated hybridoma supernatants. 2.4. Generation of alloreactive T cell clones using L cell transfectants 2.4.1. Purification of CD4 + T cells

Peripheral blood mononuclear cells (PBMCs) were separated from freshly collected heparinised whole blood using Ficoll-Hypaque gradient centrifugation. After washing, the PBMCs were layered on to tissue culture dishes pre-coated with goat anti-human immunoglobulin antibody and the mouse anti-DRa monoclonal, L243. In this way the cells were depleted of adherent cells, B cells, and class II positive cells. The remaining cells were incubated with a mixture of the following purified antibodies (at the final concentra-

27

A.N. Warrenset al. /Journal of lmmunological Methods 169 (1994) 25-33

tions indicated): anti-CD8 (10 ~ g / m l ) , anti-Leu 19 (12.5/zg/ml), mouse anti-human immunoglobulin (5 /.tg/ml) and L243 (10 /~g/ml) and depleted twice using sheep anti-mouse immunoglobulin-coated magnetic beads (50 /zl beads/107 cells per ml) (Dynal, Wirral, UK).

restimulated weekly on three occasions by the addition of a further 3 × 105 mitomycin C-treated transfected DAP.3 cells for every 106 T cells. Activated T cells were expanded by the addition of recombinant IL-2 (rlL-2) 20 U / m l (Boehringer Mannheim. Lewes, UK) 3 days after each stimulation, and viable cells were enriched by FicollHypaque density centrifugation when required. After 3 weeks, the ceils were cloned by limiting dilution, at a cell concentration of one cell/well in a volume of 20/.tl in a Terasaki plate (Nunc, Roskilde, Denmark) containing either 103 mitomycin C-treated transfectants or 104 irradiated allogeneic PBMCs with purified P H A (2 /zg/ml) (Wellcome, Dartford, UK) and rlL-2 (20 U / m l ) . The number of wells producing clones was counted, the clones expanded and their specificity confirmed using the proliferation assay

2.4.2. Preparation o f alloreactive T cell clones.

Transfected DAP.3 cells were treated with mitomycin-C (68 /zg/ml) (Martindale Pharmaceuticals, Romford, UK) for 45 min and washed in serum-free medium three times. The level of class II expression of transfected DAP.3 cells is illustrated in Fig. 1. 3 × 105 such cells were cultured with 106 purified CD4 ÷ T cells in one well of a 24 well plate in RPMI 1640 medium supplemented as for B ceils, but with 10% pooled human AB serum rather than FCS. They were

~. Ill l I | i~.~ 'I il

11 ii llJ l iI, t

! ! !

b' I

I I I

E ! ! I I LI

l O

f'

¢

! kl

'II" I| I

I; r I

Hf

'~.~. ~.'~.

I i ! I I I I l

I 10 Relative

i 100 Fluorescence

\~/,/, I 1000

Intensity

Fig, 1. Surface expressionof transfected human class II molecules on murine DAP.3 cells as determined by indirect immunofluorescence and flow cytometry. Each cell line is represented as follows: DAP.3 transfected with DRA and DRBI*0406 (light uninterrupted line), DAP.3 transfected with DRA and DRBI*ll01 (dashes), and untransfected DAP.3 cells (dotted line). For comparison, the staining of DR1 +-B-LCLs is shown (heavy uninterrupted line).

A.N. Warrens et al. /Journal of Immunological Methods 169 (1994) 25-33

28

the cultures harvested 18 h later on to glass fibre mats. Proliferation was measured as [3H]thymidine incorporation determined by liquid scintillation spectroscopy and results were expressed as Acpm.

described below. Clones were restimulated by the addition of PBMCs and rlL-2 every seven days, and rlL-2 alone was added 3 - 4 days after restimulation.

2.5. T cell proliferation assay 3. Results

T cell clones (10 4 cells/well) were cultured in the presence of either irradiated B-LCLs or mitomycin C-treated DAP.3 cells (transfected or untransfected) in fiat bottomed 96-well plates, in a final volume of 200 /xl. Stimulator ceils were added in numbers between 104 and 105, as indicated in the figure legends. Antibodies were added as indicated in the results section. After 48 h [3H]thymidine was added (1 p, Ci/well) and

3.1. Anti-DR alloreactive human T cell clones can be generated using DR-expressing DAP.3 transfectants CD4 ÷ T cells from two donors, one homozygous at the D R locus (DR3) and the other heterozygous ( D R w l l {DRA; DRBI*ll04}, DR4

Table 1 Cloning efficiency of various combinations of responder and stimulator cells using two different conditions for cloning Series

Responder DR phenotype

Stimulator DR phenotype

Accessory cells for T cell cloning

Number of wells seeded

Number of wells with growing clones

Cloning efficiency

A B E F

DR3 (homozygous) DR3 (homozygous) DR3 (homozygous) DRwll [DRBI*ll04] DR4 DwKT2

DR4 DwKT2 DR4 DwKT2 DRwll [DRBI*ll01] DRwll [DRBI*ll01]

Transfectants PBMCs + PHA PBMCs + PHA PBMCs + PHA

1,200 800 800 800

25 122 13 95

2% 15% 2% 12%

MHC class II phenotype is given on the basis of serological and cellular typing and direct DNA sequencing of the exon coding for the/31 domain. In the case of the D R w l l serotype, the DRB1 allele is also given in square brackets. Table 2 Initial screening of clones using the same transfectants that had originally been used to stimulate the bulk cultures Series

Responder DR phenotype

Stimulator DR phenotype

Accessory cells for T cell cloning

Number screened

Number specific for stimulating class II

A B

DR3 (homozygous) DR3 (homozygous)

DR4 DwKT2 DR4 DwKT2

Transfectants PBMCs + PHA

7 16

5 (71%) 6 (38%)

MHC class II phenotype is given on the basis of serological and cellular typing. Table 3 Screening of clones with murine fibroblasts and human B cells Series

Responder DR pbenotype

Stimulator DR phenotype

Accessory cells for T cell cloning

Number screened

Specific for stimulating class II

Recognize stimulating class II on B-LCLs

A B E F

DR3 (homozygous) DR3 (homozygous) DR3 (homozygous) DRwll [DRBI*ll04] DR4 DwKT2

DR4 DwKT2 DR4 DwKT2 DRwll [DRBI*ll01] D R w l l [DRBI*ll01]

Transfectants PBMCs + PHA PBMCs + PHA PBMCs + PHA

2 23 13 95

2 10 9 56

2 (100%) 8 (80%) 6 (67%) 36 (64%)

MHC class II phenotype is presented as described in the legend to Table 1.

A.N. Warrens et aL / Journal o f lmmunological Methods 169 (1994) 25-33

DwKT2 {DRA; DRBI*0406}), were purified and bulk cultures set up with one of two cloned, transfected murine fibroblastoid (DAP.3) cells, one expressing DR4 DwKT2 {DRA; DRBI*0406},

80

29

the other D R w l l {DRA; DRBI*ll01}. Having demonstrated the specificity of the bulk cultures, the T ceils were cloned in the presence of either the stimulating transfectant or peripheral blood

A6

60 40 20

1:1

1:3

1:10

50

'O X

40 30

E tj

20 10 ¸ D

111

1:3

80

1:10

B25 100

B2

'O T-

80

60

X

60

E Q.

40

<1

40

20

20 D

0 1:1

®

1:3

1:10

@ Responder:Stimulator ratio

[L t~

d~

r~

tm

r~

Fig. 2. Responses of T cell clones to various stimulators show specificity for the stimulating class II molecule (a) both on transfected murine fibroblasts, with which the T cells had been originally stimulated, and on h u m a n B-LCLs and (b) on transfected murine fibroblasts, but not on h u m a n B-LCLs. Each graph illustrates a representative experiment. In b, the responder:stimulator ratio was 1 : 10. All four clones illustrated had been raised against transfectants expressing D R 4 DwKT2. Responses of the same n u m b e r of T cells to rlL-2 ( + ) is also shown on the graphs. T h e stimulator cells were: untransfected DAP.3 (o), DAP.3-DRw! 1 [ D R B I * l l 0 1 ] (D), D A P . 3 - D R 4 DwKT2 (o), heterozygous B-LCL D R w l l [ D R B I * l l 0 4 ] / D R 4 DwKT2 (11), and homozygous B-LCLs D R w l l [ D R B I * l l 0 1 ] ( z~), D R 4 DwKT2 ( • ), and DR1 ( x ).

30

A.N. Warrens et aL / Journal o f lmmunological Methods 169 (1994) 25-33

mononuclear cells (PBMCs) with added phytohaemagglutinin (PHA). Because the cloning efficiency was much greater following cloning with PBMCs and P H A in the first pair of r e s p o n d e r / stimulator combinations (Table 1, compare series A and B), subsequent bulk cultures were cloned only with PBMCs and P H A (series E and F). Although cloning efficiency was higher using PBMCs and PHA, comparison of the specificities of series A and B suggested that cloning on transfectants yields a higher fraction of clones with the desired specificity (Tables 2 and 3, series A and B), although this did not achieve statistical significance using the X 2 test. Between 2% and 15% of Terasaki plate wells produced growth from an apparently single cell (Table 1). Samples of these were screened against DAP.3 transfectants expressing various H L A - D R molecules. For each of the populations studied, over a third appeared to recognise the D R molecule expressed by the transfectant which had been used as the stimulator (Table 2). A number of patterns of response was noted amongst the others. Some failed to proliferate in response to any DAP.3 cell, transfected or untransfected, despite a vigorous response to rlL-2. Others responded similarly to all DAP.3 cells, irrespective of whether or not they expressed class II. It was assumed that they were responding not to the product of the transfected cDNAs, but to a murine product of the DAP.3 cells which was being recognised as a xenoantigen.

3.2. A significant fraction of the anti-DR alloreactire clones raised against murine transfectants were able to recognize human B cells expressing the appropriate DR molecule Despite the initial success in raising anti-DR alloreactive T cell clones, one concern was that the human class II molecule would be recognised only in the context of a murine cell, due to specificity for a DR-bound peptide displayed by mouse DAP.3 but not human ceils. Such a pattern of reactivity would reduce the applicability of this approach. For this reason, the responses of a number of these clones were tested against transfectants and human B-LCLs bearing the

80-

~,o

60-

x E

40-

20-

0-

1:1

1:3

Responder:Stimulator

1:10

ratio

Fig. 3. Response of clone F9 to B-LCL D R w l l [ D R B I * l l 0 1 ] is inhibited by the presence of the anti h u m a n class-II mouse monoclonal antibody, L243, in a dose dependent fashion. T h e effects of adding no antibody (o), or antibody at a titre of 1/160 (zx) or 1/320 ([3) are shown.

same H L A - D R molecule. The results are presented in Table 3 and show that in each of the four series, over 60% of the anti-DR allospecific clones responded to the D R alloantigen expressed by murine and by human cells. Fig. 2a illustrates examples of the specificity assays of a representative sample of three of these clones. In dose-response titrations, some clones were more sensitive to transfectant than B-LCL stimulation, as illustrated by A6 and A7 (Fig. 2a) although this was not true for all clones, as shown by B2. Fig. 2b illustrates the response of a clone that appeared to be specific for the stimulating class II molecule, but only on a murine presenting cell. The specificity of the clones for H L A - D R was confirmed by successfully inhibiting proliferation using the L243 anti-DRa monoclonal antibody. A representative example of this is illustrated in Fig. 3.

3.3. Proliferation of T cell clones generated against DAP.3 transfectants was consistently and profoundly inhibited by anti-LFA-3, but only variably inhibited by anti-CD4 antibodies Given that the murine accessory molecules ICAM-1 and LFA-3 either fail to bind to the human ligands LFA-1 and CD2, or do so with decreased affinity, it might be predicted that this technique of generating alloreaetive cells on

31

A.N. Warrens et al. /Journal of Immunological Methods 169 (1994) 25-33

xenogeneic transfectants would favour T cells bearing antigen receptors with high specific affinity, and relative i n d e p e n d e n c e of accessory molecule interactions. T h e above was addressed by examining the effects of a n t i - h u m a n C D 4 and L F A - 3 m o n o clonal antibodies on the responses of several of these T cell clones to h u m a n B - L C L stimulators. Some clones were refractory to anti-CD4 inhibition as shown in Fig. 4a; others such as A 7 and B3 were significantly inhibited (Fig. 4b). M o r e consistent inhibition was seen with a n t i - h u m a n LFA-3, as illustrated in Fig. 4a. This inhibition was seen for all clones tested. W h e n stimulator and r e s p o n d e r cells were p r e i n c u b a t e d separately with a n t i - h u m a n L F A - 3 and thoroughly washed before the assay was set up, it was clear that the

inhibition o c c u r r e d as a result of the antibody interacting with the stimulators (Fig. 5). F u r t h e r studies are u n d e r w a y to test this antibody as a tool for assessing the affinity of T cell clones for their specific ligand by c o m p a r i n g its effects on clones g e n e r a t e d by conventional techniques.

4. D i s c u s s i o n

M o u s e cell lines expressing the products of transfected wild type and m u t a t e d H L A genes have b e e n extremely valuable tools in the analysis of h u m a n T cell restriction and allospecificity (Lechler et al., 1988), ( L o m b a r d i et al., 1991). Here, we illustrate the value of these cellular reagents in the p r o d u c t i o n of allospecific T cells

80

F9 60 40

20 'o

O"

x

E

100

,

1:1

1:3

,

r----"l

1:10

IL2

<1

80

30 ¸

60

20

40

'o K

20

J 1:1

(~

E= o <3

1:'3

1:i0

Responder:Stimulator ratio

10

~IL2 ~

0

A7

Clones

B3

Fig. 4. The alloproliferative responses of clones were markedly inhibited by anti-LEA-3 antibody (a) but variably inhibited by anti CD4 antibody (a and b). a: the responses of clones F9 and F19 to B-LCL DRwll were profoundly inhibited by anti-LFA-3 but much less affected by anti-CD4, b: the responses of clones A7 and B3 to B-LCL DR4DwKT2 were significantly inhibited by anti-CD4. This assay was performed at a responder:stimulator ratio of 1 : 10. The following amounts of antibody were added: (a) none (©), anti-LFA-3 7/zl ([:]), anti-LFA-3 20/zl ( I ) , anti-CD4 1/zl (zx), and anti-CD4 5/~1 (A); (b) none (m) and anti-CD4 5/L1 (D).

A.N. Warrens et aL /Journal of Immunological Methods 169 (1994) 25-33

32 3O 03 IO ~.=

x 20 E o

10

None

llJ I

7~ I

20 I~ I

P r e t r e a t m e n t of T cells A m o u n t of a n t i - h u m a n LFA-3 added

Fig. 5. Inhibition of the alloproliferative response resulted from the interaction of anti-human LFA-3 with stimulator cells rather than the responding T cells. Each cell type was preincubated with various amounts of antibody (marked along the abcissa for responding T cells and as follows for stimulators: none (©), 1 /zl (zx), 7/xl (rq), and 20/zl ( I ) ) . Following washing, proliferation assays were set up as before.

with a particular specificity. In order to determine whether such an approach was possible, we used murine cells expressing D R molecules on their surface to generate anti-DR allospecific T cells. This technique will be particularly valuable in the generation of T cells specific for D Q and DP molecules, using both haplotype-matched and mismatched o~-/3 combinations, and class II molecules of mixed isotype, against which it has been difficult or impossible to generate T cells using conventional techniques. The ability of a significant fraction of T cells generated in this way to recognise human cells expressing the alloantigen corresponds to the frequency with which clones generated by more conventional protocols recognise D R molecules expressed by DAP.3 transfectants. This is likely to reflect the presentation of peptides derived from human serum a n d / o r cells in the proliferation assays and the conservation of some cellular proteins between mouse and man (Lombardi et al., 1989a,b). The accessory cell function of DAP.3 cells for human T cells is greater than might be expected of a mouse cell. They have been used extensively to stimulate established human T cell clones, are able to stimulate a primary proliferative response by purified human T cells (Lombardi et al., 1991),

and are effective in restoring the PHA reactivity of accessory cell-free human T cells (G. Lombardi, unpublished results). The results reported here offer further evidence of their ability to provide the costimulatory signals required by human T cells. Of note in this regard is the interaction between CD28 and CTLA4 on the T cell with their ligand B7 on the stimulating cell (Norton et al., 1992; Linsley et al., 1991). Surprisingly, it has recently been shown that DAP.3 cells express substantial amounts of mouse B7 (Razi-Wolf et al., 1992; our own observations). It may be that mouse B7 interacts with human CD28 with sufficient affinity to provide the costimulatory signals that allow activation. It is well documented that the molecular interactions between LFA-1 and the family of ICAMs, and between CD2 and LFA-3 are important in promoting cell interactions (Dustin and Springer, 1991). The T cell clones described in this paper were generated independently of human ICAM and LFA-3 on the stimulator cells and might therefore be expected to be of high specific affinity. This is difficult to address for allospecific clones because the ligand density, in the form of M H C : p e p t i d e complexes, cannot usually be defined. Sensitivity to anti-CD4 mAb has been used to assess T cell affinity (Lombardi et al., 1991), although this is complicated by the more complex role of CD4 in T cell signalling (Altman et al., 1990). The use of anti-LFA-3 antibodies may be more reliable, although even this is complicated by the possible transduction of signals through LFA-3 (Altman et al., 1990; Ohno et al., 1991). Paradoxically the T cell clones generated using transfectants were very sensitive to antibodies against human LFA-3. Assuming that human CD2 does not interact effectively with mouse LFA-3 expressed on DAP.3 cells, the responses of these T cell clones to the DAP.3 transfectants must be independent of LFA-3 on the stimulator cell. The inhibition may be explained in two ways. One possibility is that the density of the D R molecules expressing the peptide for which these T cells are specific is lower on human than on mouse stimulator cells. In this case, the anti-human response could be more LFA-3 dependent. Alternatively, this may be another example of the inhibitory

A.N. Warrens et al. /Journal of lmmunological Methods 169 (1994) 25-33

effect of anti-accessory molecule antibodies when the accessory molecule :ligand pair is present, despite the fact that the T cell does not require the presence of this interaction (Greenlaw et al., 1992). The further value of the strategy described in this paper will be established by experiments in progress using transfectants expressing other MHC class II molecules.

5. Acknowledgements We should like to thank Mrs. Puspa Batten for purifying the CD4 ÷ T cells.

6. References Altman, A., Coggeshill, K.M. and Mustelin, T. (1990) Molecular events mediating T cell activation. Adv. Immunol. 48, 227. Altmann, D.M., Sansom, D. and Marsh, S.G.E. (1991) What is the basis of HLA-DQ association with autoimmune disease? Immunol. Today 12, 267. Dustin, M.L. and Springer, T.A. (1991) Role of lymphocyte adhesion receptors in transient interactions and cell locomotion. Annu. Rev. Immunol. 9, 27. Greenlaw, R., Robinson, P., Heaton, T., Lombardi, G. and Lechler, R. (1992) Transfection of HLA-DR-expressing DAP.3 cells with a cDNA clone encoding the glycosyl phosphatidyl-linked form of lymphocyte function associated antigen-3: biochemical features and functional consequences. Int. Immunol. 4, 673. Lechler, R.I., Bal, V., Rothbard, J.B., Germain, R.N., Sekaly, R., Long, E.O. and Lamb, J. (1988) Structural and functional studies of HLA-DR restricted antigen recognition

33

by human helper T lymphocyte clones by using transfected murine cells. J. Immunol. 141, 3003. Linsley, P.S., Brady, W., Urnes, M., Grosmaire, L.S., Damle, N.K. and Ledbetter, J.A. (1991) CTLA-4 is a second receptor for the B cell activation antigen B7. J. Exp. Med. 174, 561. Lombardi, G., Sidhu, S., Daly, M., Batchelor, J.R., Makoba, W. and Lechler, R.I. (1989a) Are primary alloresponses truly primary? Int. Immunol. 2, 9. Lombardi, G., Sidhu, S., Lamb, J.R., Batchelor, J.R. and Lechler, R.I. (1989b) Co-recognition of endogenous antigens with HLA-DR1 by alloreactive human T cell clones. J. Immunol. 142, 753. Lombardi, G., Barber, L., Aichinger, G., Heaton, T., Sidhu, S., Batchelor, J.R. and Lechler, R.I. (1991a) Structural analysis of anti-DR1 allorecognition by using DR1/H-2E k hybrid molecules. J. Immunol. 147, 2034. Lombardi, G., Barber, L., Sidhu, S., Batchelor, J.R. and Lechler, R.I. (1991b) The specificity of alloreactive T cells is determined by MHC polymorphisms which contact the T cell receptor and which influence peptide binding. Int. Immunol. 3, 769. Merkenschlager, M., Ikeda, H., Wilkinson, D., Beverley, P.C.L., Trowsdale, J., Fisher, A.G. and Altmann, D.M. (1991) Allorecognition of HLA-DR and -DQ transfectants by human CD45RA and CD45RO T cells: repertoire analysis and activation requirements. Eur. J. Immunol. 21, 79. Norton, S.D., Zuckerman, L., Urdahl, K.B., Shefner, R., Miller, J. and Jenkins, M.K. (1992) The CD28 ligand, B7, enhances IL-2 production by providing a costimulatory signal to T cells. J. Immunol. 149, 1556. Ohno, H., Nakamura, T., Yagita, H., Okumura, K., Tanigushi, M. and Saito, T. (1991) Induction of negative signal through CD2 during antigen-specific T cell activation. J. Immunol. 147. Razi-Wolf, Z., Freeman, G., Galvin, F., Benacerraf, B., Nader, L. and Reiser, H. (1992) Expression and function of the murine B7 antigen, the major costimulatory molecule expressed by peritoneal exudate cells. Proc. Natl. Acad. Sci. USA 89, 4210. Trowsdale, J., Ragoussis, J. and Campbell, R.D. (1991) Map of the human MHC. Immunol. Today 12, 443.