THE JOURNAL OF B~XOGICAL CHEMISTRY 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 265. No. 23, Issue of August 15, pp. 14036..14043,19’30 Printed in U.S.A.

Assembly and Function of the T Cell Antigen Receptor REQUIREMENT OF EITHER THE LYSINE OR ARGININE RESIDUES IN THE TRANSMEMBRANE REGION OF THE 01 CHAIN*

(Received for publication, December 20, 1989)

Richard S. BlumbergSQlI, Balbino AlarconQ 11, Jaime Sancho8 **, Francis V. McDermott+& Peter Lopez& James Breitmeyer$#$, and Cox Terhorstg From the Laboratories of §Molecular Immunology ana’ $$Tunwr Immunology, Dana Farber Cancer Institute, and the $Ga.stroenterology Division, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115

The T cell receptor (TCR) for antigen consists, on the majority of peripheral lymphocytes, of an immu- noglobulin-like, disulfide-linked heterodimeric glyco- protein: the a and fi chain. These proteins are nonco- valently linked to at least four nonvariant proteins which comprise the CD3 complex: CD3 y, 6, t, and {. Whereas the TCR a and /3 proteins have positively charged residues in the transmembrane region, all the CD3 proteins have similarly placed negatively charged amino acid residues. It has been suggested that these basic and acidic amino acid residues may play an im- portant role in TCR.CD3 complex assembly and/or function. In this paper, the structural and functional role of the lysine and arginine residues of the TCR (Y chain was addressed using oligonucleotide mediated site directed mutagenesis. The Argzee and LysZ61 resi- dues of the TCR a cDNA of the HPB-ALL cell line were mutated to either G1y266 and/or IleZ81. The altered cDNAs were transfected into a TCR (Y negative recip- ient mutant cell line of REX, clone 20A. Metabolic labeling of the T cell transfectants showed that muta- tion of either the Args6’ or Lys2’l amino acid residues had no effect on the ability of the TCR (Y chain to form either a heterodimer with the TCR j3 chain or a complex with the CD3 y, 6, and c proteins. Consequently, the Artis to G1y258 and Lyszel to Ile2’l mutations did not prevent the formation of a mature, functional TCR. CD3 complex on the cell surface as determined by immunofluorescence, cell surface radioiodination, and the ability of the transfectants to mobilize intracellular calcium after stimulation with a mitogenic anti-CD3 c monoclonal antibody. In contrast, a mutant cDNA in which both the Arg2” and Lys281 residues were mu- tated to G1y256 and Ile2’l, respectively, failed to recon- stitute the cell surface expression of the TCReCD3 complex and, consequently, the ability to respond to mitogenic stimuli. In the absence of both the Arg2” and Lys”’ residues, TCR aB heterodimer formation

* This research project was supported in part by National Institute of Health Grants Al 15066 and Al 17651. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ll Supported by National Institute of Diabetes and Digestive and Kidnev Diseases Grant 1 K08 DK01886-01. To whom corresnondence should be addressed: Laboratory of Molecular Immunology, Dana Farber Cancer Inst., 44 Binney St., Boston, MA 02115. Tel.: 617-732- 3463.

I( Special Fellow of the Leukemia Society of America. ** Recipient of a Doctres y Tecnologes fellowship (Ministerio de

Education y Ciencia, Spain). $5 Supported by National Institute of Allergy and Infectious Dis-

eases Grant Al 12609 and National Cancer Institute Grant CA 25369.

was not observed. Cotransfection studies in COS cells showed that the failure of assembly of a heterodimer was likely due to an inability of the mutated TCR a chain to form a subcomplex with either the CD3 y, 6, c, or { proteins. Therefore, either one of the basic amino acid residues in the transmembrane region of the TCR a chain was sufficient to maintain the interactions of the TCR a chain and the CD3 complex. In the absence of this interaction, TCR a/3 heterodimer formation and, consequently, cell surface expression and functional competence of the cell was impaired.

The T cell receptor (TCR),’ which subserves both antigen and major histocompatibility complex (MHC) restricted rec- ognition on human thymus derived lymphocytes (T cells) consists of a heterodimer of immunoglobulin-like glycopro- teins (Clevers et al., 1988). On the majority of peripheral lymphocytes, this receptor consists of an a: and @ chain which are disulfide linked (Marrack and Kappler, 1987). On a smaller proportion of the peripheral lymphocytes, the T cell receptor consists of TCR y and 6 disulfide or nondisulfide- linked heterodimeric glycoproteins (Raulet, 1989). The func- tion and specific recognition structures of these latter TCR forms are not well known.

Intimately associated with the TCR proteins on the surface of the cell are at least four nonvariant proteins which comprise the CD3 protein complex: CD3 y, 6, t, and { (Krissansen et al., 1986; van den Elsen et al., 1984; Gold et al., 1986; Weiss- man et al., 1988a). A fifth chain, CD3 7, has been detected in murine and human T cells and forms disulfide-linked dimers with CD3 { (Baniyash et al., 1988; Orloff et al., 1989). The physical association between the TCR and these proteins is supported by several lines of evidence. These include the coprecipitation of the CD3 proteins with anti-TCR antibodies and vice versa utilizing the appropriate detergent conditions (Borst et al., 1983), the ability to chemically cross-link the CD3 y and TCR /3 chains (Brenner et al., 1987), and the comodulation of both structures from the cell surface after incubation of T cell clones with either anti-TCR or anti CD3 antibodies (Meurer et al., 1983). In addition, the ability of T cell clones to interact with antigen correlates with cell surface

1 The abbreviations used are: TCR, T cell antigen receptor; MHC, major histocompatibility complex; mAb, monoclonal antibody; DMEM, Dulbecco’s modified Eagle’s medium; neo’, G418 sulfate- resistant; PBS, phosphate-buffered saline; HEPES, 4-(2-hydroxy- ethyl)-1-piperazineethanesulfonic acid; bp, base pair(s); NEPHGE, nonequilibrium pH gradient electrophoresis; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; ds, double- stranded; ss, single-stranded.

14036

Assembly and Function of the T Cell Receptor a Chain 14037

CD3 expression (Zanders et al., 1983) and anti-CD3 antibodies can, depending on the experimental conditions, be either mitogenic or inhibitory to antigen specific T cell activation (Reinherz et aZ., 1980, Van Wauwe et al., 1986). These findings also suggest that the CD3 protein complex is likely to be involved in the processes of signal transduction that follow the interactions of the TCR with antigen-MHC on the anti- gen-presenting cell (Weiss et al., 1986).

For a T cell to respond to an antigen-MHC complex the T cell must correctly assemble and display the complete TCR. CD3 complex on the cell surface. Model studies utilizing pulse chase metabolic labeling in normal T cells and mutant T cells lacking specific chains of the complex have suggested the existence of an ordered hierarchy of protein interactions that precede the transport of a functional TCR. CD3 complex to the cell surface (Alarcon et al., 1988a; Koning et al., 1988; Bonifacino et al., 1988). In the absence of certain members of the TCR. CD3 complex, surface expression and, consequently, immune responsiveness is impaired (Weiss and Stobo, 1984). In many cases, it has been shown that replacement of the missing component by gene transfer will restore surface expression and, consequently, the functional capacity of the cell (Ohashi et al, 1985; Saito et al., 1987).

The specific structural features of the proteins of the TCR. CD3 complex that are important in the formation and func- tion of the receptor are just beginning to be elucidated. A review of the protein sequences of the TCR and CD3 proteins, as deduced from the nucleotide sequences of their respective cDNAs, reveals 1 or 2 charged residues in each of their transmembrane regions. Whereas the TCR proteins have positively charged residues; lysine and arginine in TCR cx and 6, and lysine in TCR p and y, the CD3 proteins have negatively charges residues; glutamic acid in CD3 y and aspartic acid in CD3 6, t, and { (Clevers et al., 1988; Raulet, 1985; Weissman et al., 1988a). By analogy with other systems, it is possible that these charged residues may play a role in receptor assem- bly and/or function (Adams and Rose 1985; Kaback, 1988). In order to determine the potential structural/functional role of the charged transmembrane residues of the TCR LY chain, we have used site-directed mutagenesis to remove one or both charges from the TCR (Y chain and transfected the mutated cDNAs into a T cell recipient that lacks this chain. In addi- tion, we have used a non-T cell recipient and observed the effects of the mutation on the assembly of the TCR. CD3 complex. In these studies we have shown that either the lysine or arginine residue of the TCR (Y chain was adequate for receptor complex assembly and the early biochemical events associated with T cell activation as determined by the ability to mobilize intracellular calcium. Only in the absence of both charged residues was receptor assembly and, consequently, cell surface receptor presentation and functional competence impaired.

MATERIALS AND METHODS

Cell Lines-The cell line used as a recipient for all transfections was clone ZOA, a T cell receptor (Y negative variant of the REX T cell line (Breitmever et al.. 1987a). The wild tvpe REX T cell line was also utilized. All T cell lines were maintainedin RPM1 1640 (GIBCO) supplemented with 10% fetal calf serum (GIBCO), 100 units/ml penicillin and 100 pg/ml streptomycin (complete medium) in 5% CO,. COS cells were utilized for transient transfections and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented as above and maintained in 10% CO,.

Antibodies-The monoclonal antibody WT31, a mouse IgGl, anti- human T cell receptor framework monoclonal antibody (mAb) was kindly provided by WJM Tax (University Hospital, Nijmegen, The Netherlands) and used for immunofluorescence assays (Tax et al., 1983). SP34, a murine IgG3 mAb which recognizes the CD3 c chain

(Pessano et al., 1985), IdenTity” cyF1 (aF1) and IdenTity” @Fl (BFl), which recognize a common framework determinant on the TCR a and fl proteins, respectively (kindly provided by T Cell Sciences, Cambridge, MA), were used in all immunoprecipitations. 2TSF4, a mitogenic IgGl murine anti-human CD3 t mAb was used in the studies of calcium mobilization (Transy et aZ., 1989).

Expression Vectors-The wild-tvpe HPB-ALL TCR 01 expression vector was constructed with a 1266-bp BglII fragment of the HPB- ALL TCR (Y cDNA. This cDNA had oreviouslv been isolated in our laboratory from a XgtlO cDNA libra& using poly(A) RNA derived from the HPB-ALL cell line (Berkhout et al.,-1988a). This was subcloned from DCD HPB-ALL TCR LY into the BamHI site of the pSR (Y Neo expression vector (Takebe et al., 1988) as described previously (DeWaal Malefyt et al., 1989).

To generate site-directed mutants, the full length TCR a cDNA from the T cell line HPB-ALL was isolated from the pGEM Blue plasmid as a SacI-PstI fragment (Sac1 is from the polylinker, PstI is in the a/-untranslated region of the TCR (Y cDNA) and ligated into the SacI-PstI sites of Ml3 mp18 (Ml3 mp18 a) (Fig. 1). Ml3 mp18 a containing viruses were isolated and site-directed mutants were created using a single synthetic oligonucleotide primer with the modification described by Kunkel (1985). Briefly, Ml3 virus contain- ing the TCR (Y cDNA were used to infect the ung- (uracil N- glycosylase-) dut- (dUTPase-) strain of Escherichia cob, CJ236, in the presence of chloramphenicol(20 rg/ml) and uridine (12.5 fig/ml). Virus was isolated by polyethylene glycol precipitation and ssDNA- containing uracils extracted with successive phenol extractions. The isolated ssDNA was annealed to the appropriate antisense oligonu- cleotide that was phosphorylated with polynucleotide kinase and mismatched at a single nucleotide to induce a point mutation. The complementary DNA sequence encoded by each mutant antisense oligonucleotide is shown in Fig. 1. The annealed oligonucleotide was extended with Klenow enzyme, ligated with T4 DNA ligase to form a dsDNA plasmid, and the dsDNA plasmid used to transform the dut+ ung+ strain of E. coli, DHF, which will preferentially replicate the mutant strand. Plaques were isolated and mutant clones identified by sequencing with the dideoxynucleotide method (Sequenase, United States Biochemical Corp.) using an 18-mer recognizing the cDNA approximately 100 bp downstream from the mutated &don(s). The mutant cDNAs were subcloned into the XhoI-BamHI sites of the pSR ~1 Neo expression vector via pBluescript KS in the following manner. The EcoRI-PstI fragment of the mutated cDNAs in Ml3 mp18 were ligated into pBluescript KS (pBluescript a). The &II- BamHI fragment of pBluescript a containing the full length cDNA of HPB-ALL (Sal1 is in the pBluescript polylinker, BumHI is in the 3’untranslated region of the TCR a cDNA) was ligated into the XhoI-AamHI sites of pSR (Y Neo creating pSR cr Neo --a (see results for nomenclature).

Additional expression vectors used in COS cell transfections in- cluded pSR 01 Neo-+y (pSR y) (800-bp XhoI-BamHI fragment of pSVL-yh’- (Krissansen et al., 1986) into the XhoI-BamHI sites of pSR (Y Neo), pSR a Neo-t (pSRc) (1600-bp XhoI-XhoI fragment of pCD chUman (Gold et al., 1986) into the XhoI site of pSR-ol Neo), MNC8S (625-bp KnnI-XbaI fraament of DCD dhume” (van den Elsen et al., 1984) via-Ml3 mp 19 into-the SocI-*PstI sites of MNC8 (Seed, 1987)) and sH3M { (isolated from a human HPB-ALL cDNA library made in plasmid xH3M (Aruffo and Seed, 1987) by probing with a murine CD3 < cDNA2 (kindly provided by Dr. Allen Weissman, National Institutes of Health) (Weissman et al., 1988b).

DNA Transfer into T Cells-Transfection was accomplished utiliz- ing electroporation. 5 X lo6 cells were washed with RPM1 1640 at room temperature and placed in a sterile disposable cuvette (Bio- Rad) in RPM1 1640 in a total volume of 0.8 ml. 40 rg of circular DNA was added, the cells mixed and electroporation performed with a Bio-Rad gene pulser unit using a capacitance extender at 960 rF and 250 volts at room temperature. After 2 min incubation, the cells were placed in 10 ml of Iscove’s modified Dulbecco’s medium (GIBCO) containing 2.5% heat-inactivated pooled human AB serum (Pel-Freez Biologicals, Rogers, AZ), 100 units/ml penicillin, and 100 fig/ml gentamicin for 48 h at 37 “C in 5% CO,. After 48 h, cells were selected in 96-well flat-bottomed plates (Nunc, Denmark) at 2 x lo6 cells/ml in Iscove’s modified Dulbecco’s medium containing G418 sulfate (Geneticin) (GIBCO) at 1.25 mg/ml (specific activity, 500 pg/ mg). After 2-3 weeks of selection, G418 resistant (neo’) cells were expanded and maintained in complete medium containing 1.0 mg/ml G418 sulfate.

*J. Sancho and C. Terhorst, unpublished observations.

14038 Assembly and Function of the T Cell Receptor LY Chain

FIG. 1. Construction of site-di- rected mutants of TCR a cDNA of the HPB-ALL cell line (a”). The TCR a” cDNA (El) was subcloned from pGEM (IBP) into Ml3 mp 18 (Omn). Site-directed mutants were created using a single syn- thetic antisense oligonucleotide primer containing either one or two nucleotide mismatches resulting in the indicated complementary DNA sequences. The mutant cDNAs were shuttled via pBluescript (f@ into the pSR a-neo expression vector which contains the SR (Y promotor (HTLV-I/SVIO fusion pro- motor), the SV40 poly(A) adenylation site and the neoR gene. Scl, SacI; E, EcoRI; Bg, BglII, D, DraIII; B, BarnHI; P, PstI; Sl, SalI; X, XhoI.

G ATT GGG TTC .HArp2”“,G,ps 5’ - - - - 3’

.HLYP -II&G 5’ TC CTC CTG OTC GCC GGG 3’ - - - - - -

.HAr.$5s- Gly-: 5’ ATC CTC CTC CTG -- - -

LYP - I l&G’

a* E C> C> Cl0 IlC LC” LC” Ll”

COS Cell Transfections-COS cell transfections were performed using the DEAE-dextran method of Aruffo and Seed (1987). Briefly, subconfluent cell monolayers in loo-mm Petri dishes (Falcon) were washed with 10 ml of phosphate-buffered saline (PBS) and covered with 2 ml of transfection medium containing DMEM with 10% NuSerum (Collaborative Research, Lexington, MA) and 100 NM chlo- roquine diphosphate (Sigma). Subsequently, 2 fig of each plasmid DNA was added per plate followed by the slow addition of 0.5 ml of DEAE-dextran, 2 pg/ml (Pharmacia LKB Biotechnology Inc.) in transfection medium. After 4 h at 37 “C in 10% COs, the medium was removed and the cells were incubated for 2 min with 2 ml of 10% dimethyl sulfoxide (Fisher) in PBS. The cells were then washed with PBS and 10 ml of DMEM containing 10% fetal calf serum were added and the cells incubated for 48-72 h at 37 “C in 10% CO, prior to metabolic labeling.

Zmmunofluorescence-Surface expression of antigens were detected with biotin/phycoerythrin-conjugated streptavidin. Approximately 1 x lo6 cells were washed with Dulbecco’s phosphate-buffered saline containing 1% bovine serum albumin (Sigma) and 0.02% sodium azide (wash buffer) and incubated for 30 min with either normal mouse serum as a control or one microgram of the WT-31 mAb at 4 “C. The cells were washed 3 times with wash buffer and then sequentially incubated with 1 pg of biotinylated goat anti-mouse immunoglobulin (Southern Biotechnology, Birmingham, AL) and phycoerythrin-conjugated streptavidin (Becton Dickinson, Mountain View, CA) for 30 min, each on ice with washing after each step. Cells were fixed in 1% paraformaldyhyde and analyzed on a Coulter Epics 750 Series flow cytometer (Hialeah, FL) using standard optical phy- coerythrin filters.

Isolation of RNA-Total cellular RNA was isolated by the proce- dure of Chomczynski and Sacchi (1987). Briefly, approximately 5-10 x lo6 cells were washed with PBS and transferred to an Eppendorf tube. One volume of a solution containing 4 M guanidinium thiocya- nate, 25 mM sodium citrate, pH 7.0, 0.5% Sarcosyl, and 0.1 M p2- mercaptoethanol was added and the cells gently mixed. The following solutions were added sequentially: 0.1 volume of 2 M sodium acetate, pH 4.0, 1 volume of phenol saturated with water, and 0.2 volume of chloroform isoamylalcohol (49:l) with gentle mixing. After addition of the last solution, the mixture was shaken vigorously and placed on ice for 15 min. The solution was centrifuged in a Microfuge for 20 min at 4 “C and the aqueous phase removed. One volume of isopro- panol was added and incubated at -20 ‘C for at least 1 h. The RNA precipitate was isolated by centrifugation at 4 ‘C for 20 min in a microcentrifuge. The supernatant was decanted and the pellet resus- pended in 25 ~1 of diethyl pyrocarbamate-treated water after drying in a Savant Speed-Vat.

Northern Blotting-Twenty micrograms of cellular RNA was elec- trophoresed in a 1.2% agarose gel under denaturing conditions using formaldehyde, blotted, and hybridized as described previously (Ber- khout et al., 1988b). DNA probes were labeled to a specific activity of 2 X lOa cpm/rg using [a-32P]dGTP with Klenow DNA polymerase primed by the random hexanucleotide priming method (Berkhout et al., 1988b). All probes were inserts from cDNA clones purified from

low melting agarose. The following probes were used: TCR a, either a P83-bp EcoRI-DraIII fragment of pGEM-HPB-ALL 01 or a 1200-bp BglII fragment of pCD-HPB-ALL oi (Berkhout et al., 1988a); TCRB, 400-bp BglII C@2 fragment of JUR@Z (Yoshikai et al., 1984); CD3 y, 900-bp XhoI fragment of pJ6 T3y (Krissansen et al, 1986); CD3b, 900-bp XhoI fragment of pBG BC9 (van den Elsen et al., 1984); CD3 e, 1500-bp BamHI fragment of pDJ4 (Gold et al., 1986); and @ cytoplasmic actin (kindly provided by Dr. Steven Balk).3

Radiolabeling-Cell surface iodination with carrier-free Nalz61 (Du Pont-New England Nuclear) was performed by the lactoperoxidase method. 2 X 10’ cells were washed with PBS, centrifuged, and resuspended in 200 ~1 of PBS. 2 mCi of carrier-free Na’*‘I and 30 ~1 of a 140 IU/ml solution of lactoperoxidase (Sigma) was added. Sub- sequently, lo-p1 aliquots of a 0.06% solution of hydrogen peroxide was added four times at 5-min intervals. The reaction was stopped by diluting the cells in 10 ml of 20 mM potassium iodide in PBS containing a saturating concentration of tyrosine.

Metabolic labeling of T cells was performed in the following manner. lo-20 X lo6 cells were washed with PBS and resuspended at 5-10 X lo6 cells/ml in methionine/cysteine-free DMEM (Flow labo- ratories, McLean, VA) supplemented with 20 mM HEPES (DMEM- S). After a l-h period of starvation, the cells were labeled for 6 h with 0.5 mCi of trans-[35S]methionine (ICN Biomedicals, Irvine, CA). Metabolic labeling of COS cells was performed in the following manner. The transfected COS cell monolayers were washed twice with PBS and labeled with 0.5 mCi of trans-[35S]methionine for 1 h after a starvation of 1 h in DMEM-S.

Zmmunoprecipitation and Electrophoresis-Labeled T cells or COS cell monolayers were washed once with PBS and lysed on ice with either 0.5 or 1.0 ml, respectively, of immunoprecipitation buffer containing 20 mM Tris-HCI, 0.15 M sodium chloride, 10 mM iodoa- cetamide, 1 mM phenylmethylsulfonyl fluoride, and 1 mg/ml each of leupeptin, pepstatin, antipain, and chymostatin (small peptidase in- hibitors) at pH 7.6 with 1.0% digitonin and 0.12% Triton X-100 as detergents (DTX buffer) (Oettgen et al, 1986). After a 30-min incu- bation, the cell lysates were centrifuged for 15 min at 14,000 X g followed by centrifugation at 100,000 X g for 30 min in a Beckman Airfuge at 4 “C. The lysates were precleared overnight at 4 “C with lo-20 ~1 of protein A-Sepharose beads (Pharmacia LKB Biotechnol- ogy Inc.) followed by three l-h incubations with 5 ~1 of protein A- Sepharose beads that had been preabsorbed with an equal volume of normal mouse serum. The beads from the last preclear were used as the normal control immunoprecipitate. The precleared lysates were incubated for 4-24 h at 4 “C with 5.0 ~1 of protein A-Sepharose beads which had been preabsorbed with the specific mAb. Control and specific immunoprecipitates were then washed 7-9 times with DTX buffer. The immunoprecipitates were subjected to either one-dimen- sional nonreducing, two-dimensional nonreducing/reducing, or two- dimensional nonequilibrium pH gradient (NEPHGE)/reducing SDS- polyacrylamide gel electrophoresis (PAGE) as described previously (Alarcon et al., 1988a; Pettey et al., 1987; Jones, 1984).

’ S. Balk, unpublished observations.

Assembly and Function of the T Cell Receptor (Y Chain 14039

Measurement of Zntracellular Calcium Concentration-Cells were washed with complete medium and resuspended at 10 x 10” cells/ml in l-2 fig/ml indo-l (Molecular Probes Inc, Tulatuan, OR) for 30 min at 37 “C, 5% CO, with shaking every 5 min. The cells were washed and resuspended at 1 x lo6 cells/ml in complete medium and maintained at room temperature. Cells were stimulated with the 2T8F4 mitogenic mAb at 10 rg/ml. Intracellular calcium concentra- tion was calculated from the indo-l fluorescence ratio using an Epics V flow cytometer as described previously (Breitmeyer et al., 1987b).

RESULTS

Absence of TCR a Chain Expression in the REX Mutant, Clone 20A-T3.Nl is a mutant of REX that was obtained by negative selection with anti-CD3 antibodies which inhibit colony formation of REX wild-type cells in soft agar (Breit- meyer et al., 1987a, 1987b). T3.Nl and a number of other variants were isolated on the basis of their inability to be inhibited by anti-CD3 antibodies. Flow cytometric analysis of T3.Nl has previously shown that the lack of anti-CD3-in- duced growth inhibition is associated with the absence of TCR. CD3 complex expression on the cell surface. Clone 20A is a subclone of T3.Nl which was obtained after four succes- sive rounds of cloning by limiting dilution.

Total cellular RNA was obtained from clone 20A and REX and Northern blotting performed. As can be seen in Fig. 2, when cDNAs of TCR a, TCR @, CD3 y, 6, and c were used as probes, clone 20A expressed amounts of TCR p, CD3 y, 6, and t mRNA that were comparable to REX. However, clone 20A did not express any TCR 01 mRNA.

To analyze intracellular production of TCR o( protein, REX and clone 20A were metabolically labeled with [35S]methio- nine for 6 h and lysates prepared with a detergent buffer containing 1% digitonin and 0.12% Triton X-100 (DTX buffer) which had previously been shown to maintain the integrity of the TCR.CD3 complex (Oettgen et al., 1986; Alarcon et al., 1988a). The lysates were immunoprecipitated with the CD3 t-specific mAb SP34, the TCRcu-specific mAb aF1, and the TCR @specific mAb PFl and analyzed by one- dimensional SDS-PAGE under nonreducing conditions (Fig. 3). Using these antibodies, the CD3 y, b, and e and TCR (Y and p proteins could be detected in REX (Fig. 3A). In contrast, although normal amounts of CD3 y, 6, and t and TCR 0 protein could be discerned in clone 20A (Fig. 3B, lanes 2 and 4), there was no evidence of TCR (Y protein synthesis as delineated by the absence of the mature (m) and immature (i) TCR cup heterodimer (Fig. 3B, lane 2) and free TCR LY protein (Fig. 3B, lane 3). CD3 [ could be detected in clone 20A by immunoprecipitation with a CD3 <-specific antiserum (data not shown).4 Therefore, clone 20A did not express TCR LY mRNA and, consequently, intracellular protein and was used as the recipient in all subsequent transfections.

Intracellular Expression of TCR 01 Proteins That Lack the Arginine and/or Lysine Residues-To determine the role of the positively charge transmembrane residues in the TCR o( chain, oligonucleotide-mediated, site-directed mutagenesis was utilized to generate mutants of the TCR LY chain of HPB- ALL for use in transfection studies in clone 20A. Antisense oligonucleotides were used to change ArgZ56 and Ly?l of the HPB-ALL TCR (Y cDNA (aH) to either G1y256 and/or Ilez6’ (Fig. 1). The mutant TCR aH cDNAs were subsequently subcloned into the pSRa expression vector to create pSRa Neo-aH (Arg256-Gly256) (pSR&), pSRa Neo-tiH (Lys26’-Ile261) (pSRaK-) and pSRa Neo aH (Arg2”6-Gly256: Lys26’-Ile261) (~=a RK-) for transfection into clone 20A. Total cellular RNA was obtained from the bulk population derived from lo-12 wells of each transfection and analyzed by Northern blotting.

‘J. Sancho and C. Terhorst, unpublished observations.

20A R 20A R 20A R 20A R 20A R

fir

.* hd -3.5

w -1.6 -1.4 -1.3 4

i (; @ -0.7

TCRaC TCR/3 CDBY CD36 CD3c

FIG. 2. Northern blot analysis of REX and clone 20A. Total cellular RNA was isolated from the REX (R) and clone 20A and 20 pg of each were analyzed under denaturing conditions in the presence of formaldehyde. The filters were hybridized with a TCR a probe, stripped, and sequentially hybridized with a TCR @, CD3 y, 6, and c probe. The published sizes of each transcript are indicated.

93w

66,

45,

31*

14,

ml

il 8 -Y -s/c

REX 20A

FIG. 3. Immunoprecipitation of REX and 20A. REX (A) and 20A (B) were labeled for 6 h with [35S]methionine and [35S]cysteine, lysed with immunoprecipitation buffer containing 1% digitonin and 0.12% Triton X-100 and immunoprecipitated with the SP34 (34), o( Fl ((u), and fl Fl (p) mabs, and normal mouse serum (M. The immunoprecipitates were analyzed on a 12.5% SDS-polyacrylamide gel. Mature (m) and immature (i) TCR ~ufl heterodimer, single TCR LY and p proteins, and the CD3 proteins, y, 6, and c, are indicated on the right. The molecular mass markers in kilodaltons are indicated on the left.

The Northern blots were probed with a 283-bp EcoRI-DraIII fragment of pGEM aH representing sequences derived from the variable region of the HBP-ALL TCR (Y cDNA (Berkhout et al., 1988a) (VaH). This probe was used to distinguish between the transfected TCR CY message (Va12.1) and any potential endogenously reactivated message (Votl.2). A rep- resentative population of cells from each transfection is shown in Fig. 4 (top). Whereas none of the cell lines derived from the mock-transfection hybridized with the VaH probe (lane I), probing of the Northern blot with the VaH probe revealed a specific message in 3 of 12 of the neo’ cell lines using pSRaH (lane 2), 6 of 12 of the analyzed clones using pSRaR- (lane 3), 2 of 10 wells using pSRc? (lane 4), and 5 of 12 wells using pSRolRK- (lane 5) as the expression vectors. Analysis with a ,&actin cDNA probe (Fig. 4, bottom) revealed significant

14040 Assembly and Function of the T Cell Receptor LY Chain

12345

firm-b PACTIN

L FIG. 4. Northern blot analysis of clone 20A transfectants.

Total cellular RNA was isolated from transfectants of clone 20A and 20 fig analyzed. Lane I, pSRa-neo vector; lane 2, pSRa” vector; lane 3, pSR crR- vector; lane 4, pSR #- vector; and lane 5, pSRaRK- vector. The blot was hybridized with a 283-bp EcoRI-DraIII fragment of pGEM c? representing variable region sequences of the HPB-ALL TCR 01 cDNA (VaH) (top) and rehybridized with a p-actin probe (bottom).

amounts of RNA in all lanes. The Northern blot analysis, therefore, indicated the successful transfection of clone 20A with the heterologous HPB-ALL TCR o( cDNAs.

The 20A CY~, 20A aR-, 20A &, and 20A @- cell lines were analyzed further after metabolic labeling with [““S]methio- nine for 6 h. The lysates in DTX buffer were used for immunoprecipitation with the CD3 t-specific mAb SP34 and subsequently analyzed by two-dimensional SDS-PAGE under nonreducing/reducing conditions (Fig. 5, A and B). The anti- CD3 t-specific mAb coprecipitated the mature (m) and im- mature (i) forms of the TCR cub heterodimer in conjunction with the CD3 y, 6, and t proteins (CD3) (Fig. 5, A and B, top panels) from the REX, 20A c?, 20A &, and 20A oK- cell lines. Precipitation of the 20A aRK- cell line and untransfected clone 20A with the SP34 mAb, however, revealed only the CD3 y, 6, and t proteins in association with a TCR /3@ homodimer (&). Neither free TCR N protein nor a TCR cup heterodimer could be detected in association with the CD3 proteins in either clone 20A or the 20A aRK- cell line (Fig. 5, A and 5B, top).

To determine whether the 20A aRK- cell line contained any TCR LY protein, the cell lines were also analyzed with the anti- TCR 01 mAb (Fig. 5, A and B, bottom panels). In each of the lysates prepared from REX, 20A a”, 20A c?, and 20A c?, aF1 precipitated both free TCR LY protein and immature TCR m/3 heterodimer. In contrast, no TCR afi heterodimer but only free TCR N protein could be detected in the 20A mRK- cell line (Fig. 5B, bottom). As expected, clone 20A did not contain any free or complexed TCR cy protein as judged by immuno- precipitation with the ~vF1 mAb (Fig. 5B, bottom).

Therefore, substitution of both positively charged trans- membrane residues in the TCR 01 protein for noncharged amino acid residues prevented the formation of a heterodimer with the TCR /3 chain. In contrast, changing either the trans- membrane lysine or arginine residues of the TCR LY chain alone had no effect on the intracellular formation of the TCR c@ heterodimer and on the association of the TCR N chain with the CD3 y, 6, and t proteins.

Ability of TCR o( Chain Mutants to Interact with the CD3 Proteins-To determine whether the mutated TCR cy chains interact with the CD3 proteins, COS cell transfections were performed. COS cells were cotransfected with pSRt, pSRr, MNC8-6, KHBM 5; and either pSR aH, pSR OIL-, or pSR aRK-. The transfected cells were metabolically labeled for 1 h with [?S]methionine and lysed with DTX buffer. The immunopre- cipitates prepared with the SP34 and cuF1 mAbs were analyzed in the first dimension by NEPHGE and the second dimension

A

\ f t& \

CD3

am Y

FIG. 5. ?3 metabolic labeling of clone 20A transfectants. REX, 20A and the transfected lines, 20A a”, 20A c?-, 20A OIL- and 20A aRK- were labelled for 6 h with [““Slmethionine and [?S]cysteine, lysed with immunoprecipitation buffer containing 1% digitonin and 0.12% Triton X-100 and immunoprecipitated with either SP34 (top panel) or aF1 (bottom panel). The immunoprecipitates were analyzed on an 11% SDS-polyacrylamide gel under two-dimensional nonre- ducing/reducing conditions. CD3 { is poorly visualized on a gel of this concentration. The mature (m) and immature (i) TCR cup het- erodimer, single TCR a protein, and the CD3 y, b, and c proteins (CD3) are indicated. @2 indicates a TCR BP homodimer. The molec- ular mass markers in kilodaltons are shown on the right.

by SDS-PAGE under reducing conditions using a 12.5% poly- acrylamide gel (Fig. 6).

When the anti-CD3 t (SP34) immunoprecipitates were analyzed, no TCR (Y protein derived from the pSRaRK- vector could be shown to be coprecipitated with the CD3 proteins (Fig. 6, bottom). In contrast, the pSRa” and pSRaK- derived TCR N proteins (Fig. 6, bottom, arrow) could be coprecipitated with the CD3 proteins (Fig. 6, bottom, double arrow indicates position of CD3 t). When analyzed with the aF1 mAb, equiv- alent amounts of TCR 01 protein could be detected in all the transfections as a large single spot at the acidic end of the gel indicating that the TCR 01 protein was present (Fig. 6, top, arrow). The results shown in this section demonstrated that

Assembly and Function of the T Cell Receptor LY Chain 14041

-NEPHGE-

$ uF1

-31

psd WiLYK- pSRP-

FIG. 6. ‘%3 metabolic labeling of COS cell transfectants. COS cells were transiently transfected with the pSRc, pSR y, MNC8 6, and H3M [ vectors and either the pSR aH, pSR aK-, or pSR c?“- expression vectors. The transfected COS cells were metabolically labeled for 1 h with [35S]methionine and @]cysteine, lysed as described above, immunoprecipitated with either the aF1 (top) or SP34 (bottom) mabs, and the immunoprecipitates analyzed by NEPHGE in the first dimension and on a 12.5% SDS-polyacrylamide gel under reducing conditions in the second dimension. The acidic and basic direction of the NEPHGE are indicated. The molecular mass markers in kilodaltons are shown on the right. The TCR 01 protein (single arrow) and CD3 t (double UFFOW) are indicated.

either one of the basic amino acid residues in the transmem- brane region of TCR cy was necessary and sufficient to main- tain the interaction of the mutated chain with the CD3 proteins.

Cell Surface Expression of the TCR a Chain Mutants- Previous studies in T cell mutants lacking either the TCR (Y or /3 proteins have shown that, in the absence of either of these proteins, surface expression of the remaining TCR. CD3 complex was not observed (Weiss et al., 1984). Transfection of either the TCR cy or /3 cDNA into TCR (Y- or 6- T cell lines, respectively, resulted in reconstitution of the surface expression of the TCR. CD3 complex (Ohashi et al., 1985; Saito et al., 1987). We, therefore, determined whether wild- type and mutant HPB-ALL-derived TCR cy protein could restore surface expression of the TCR.CD3 complex in the REX mutant, clone 20A. Cell lines from each transfection expressing the highest amounts of mRNA by Northern blot- ting were stained with either an anti-TCR framework mAb or normal mouse serum followed by incubation with a biotin- ylated goat anti-mouse antibody and, subsequently, phycoer- ythrin-conjugated streptavidin (Table I). Significant amounts of the TCR.CD3 complex were detected on the surface of the 20A aH, 20A c?, and 20A aK- cell lines (54.7,49.8, and 41.0%, respectively) (Table I). There was, however, no evidence of TCR.CD3 complex surface expression on all five of the 20A aRK- cell lines derived from the transfection using the pSRaRK- vector. Similarly, two of the cell lines mock-trans- fected with the pSRa Neo vector alone were tested and were negative (20A-pSR (Y neo).

TABLE I

Flow cytometric analysis of 20A trunsfectants REX, 20A and transfectants of 20A, 20A a”, 20A &, 20A aK-,

20A aRK- (l-5) and 20A pSRa neo (l-2) (mock), were stained with the WT-31 mAb. Bound antibodies were detected with a biotinylated second antibody followed by phycoerythrin-conjugated streptavidin. Nonspecific immunofluorescence was analyzed by using normal mouse serum during the first incubation. The proportion of cells specifically binding to the WT-31 antibody and mean fluorescence intensitv (MFI) are indicated.

Cell line WT-31 MFI

REX 20A

% 91.9

0.8 153.2

22.3

20A c?’ 54.7 118.8 20A LY~- 49.8 112.1 20A c?- 41.0 116.3

20A aRK-’ 3.2 21.0 20A aRK-* 1.3 26.5 20A aRK-’ 0.9 26.5 20A aRK-’ 1.4 23.8 20A aRK-’ 2.8 21.4

20A pSR (Y neo 2.0 31.6 20A pSR (Y neo 2.4 29.9

R 20H R- K-RK-

45*

14*

FIG. 7. Radioiodination of clone 20A transfectants. REX (I?), 20A (20), and the 20A transfectants, aH (H), a’- (R-), aK- (K-), and aRK- (RK-), were iodinated with Na”‘I by the lactoperoxidase- catalyzed method and lysed with DTX buffer as outlined above. The lysates were immunoprecipitated with the SP34 mab and the immu- noprecipitates analyzed on a 12.5% polyacrylamide gel under nonre- ducing conditions. The mature TCR cup (c&J and CD3 proteins (y, 6, and c) are indicated. The molecular mass markers in kilodaltons are shown on the left.

The surface expression of the TCR c@ heterodimer was confirmed by radiolabeling the mutant cell lines with Naiz51 by the lactoperoxidase-catalyzed method followed by immu- noprecipitation with the anti-CD3 E mAb (Fig. 7) from DTX- prepared lysates. As judged by SDS-PAGE under nonreducing conditions, mature (m) TCRcvP heterodimer could be isolated from the REX, 20A aH, 20A aR- and 20A aK- cell lines (lanes 1,3,4 and 5) but not clone 20A (lane 2) and the 20A aRK- cell line (lane 6). Partial complexes consisting of the CD3 y, 6, and/or c chains could be observed on the 20A CY~~- cell line, as well as clone 20A, which is consistent with recent obser- vations of Ley et al. (1989) in murine T cells. Therefore, transfection of clone 20A with the heterologous TCR (Y chain

Assembly and Function of the T Cell Receptor LY Chain 14042

1234567

TIME (MINUTES)

FIG. 8. Anti-CD3 c-mediated calcium mobilization of 20A transfectants. Antibody-stimulated changes in plasma calcium lev- els were determined by loading the cells with indo-l and monitoring the UV-stimulated indo- florescence ratio (515/405 nm) of individual cells on an Epics V flow cytometer. The intracellular calcium concen- tration as determined by the indo-l ratio is indicated on the ordinate in nanomoles (nM). The time (minutes) after stimulation with the 2T8F4 mab (arrow) and ionomycin (asterisk) are indicated on the a4wcissa. A, REX, B, 204 C, 20A (Us; D, 20A (uR-; E, 20A &; F, 20A P‘-. The proportion of responding cells were 86, 4, 37, 37, 68, and 3%, respectively.

derived from the HPB-ALL cell line was able to restore the surface expression of the TCR. CD3 complex. Only the ab- sence of both positively charged transmembrane residues prevented the formation of TCR a/? heterodimer and, conse- quently, the surface expression of the mature TCR. CD3 complex.

Signal Transduction Is Not Dependent on Either the Argi- nine or Lysine Residues-The functional integrity of the TCR. CD3 complexes containing the mutant TCR (Y chains of the HPB-ALL cell line was analyzed by the ability of these complexes to induce an increase in intracellular calcium. The transfected and control REX cells were loaded with indo-l and stimulated with a mitogenic anti-CD3 t mAb, 2T8F4 (Fig. 8). After stimulation (arrow) REX (7A) and the 20A aH, 20A aR-, and 20A aK- cell lines (Fig. 7 C, D, and E, respectively) responded with an increase in intracellular calcium concen- tration that ranged from approximately 600 to 900 nM. Evi- dence of calcium mobilization could not be discerned in either clone 20A (Fig. 7B) or the 20A aRK- cell line (Fig. 71;3. Evidence of calcium mobilization was observed after the ad- dition of ionomycin in clone 20A (asterisk) and the 20A aRK- cell line (data not shown) indicating that they were effectively loaded with indo-l and were able to mobilize calcium. The proportion of responding cells generally reflected the level of expression of the TCR.CD3 complex on the cell surface (Table I) and are shown in the figure legend.

Therefore, mutation of either Arg256 or LysZ61 to G1y256 or IleZ61, respectively, had no effect on not only receptor complex assembly but also on function as determined by the ability to mobilize intracellular calcium after stimulation of the CD3 complex. Although partial complexes of CD3 proteins could be observed on the cell surface of clone 20A and the 20A aRK- cell line (Fig. 7), stimulation of these complexes with a CD3 e-specific mAb showed them not to be functional (Fig. 8) and is consistent with previous observations in murine T lympho- cytes (Ley et al., 1989).

DISCUSSION

In this paper we have shown that, in the absence of both the lysine and arginine residues in its transmembrane region, the TCR a chain could not assemble with the CD3 y, 6, t, and {proteins or form a heterodimer with the TCR /3 protein. In

addition, we have shown that in the absence of both the arginine and the lysine residues, TCR. CD3 complex forma- tion and, consequently, cell surface expression of the complex was abrogated. However, in the presence of either the lysine or arginine residues, the mutant TCR cr chain was able to interact with the CD3 proteins and form a mature heterodimer with the TCR p chain that was transported to the cell surface (Figs. 5 and 7). These findings are in agreement with previous studies on pulse-chase labeled human T cells which indicated that the association of the TCR ab heterodimer occurred after the interaction of single TCR chains with the CD3 complex (Alarcon et al., 1988a). It would seem, therefore, that the inability of the mutant TCR cy protein lacking both the lysine and arginine residues to form a TCR cup heterodimer would be due to its inability to form a TCR e-CD3 subcomplex (Figs. 5 and 6). It was difficult to fully test this hypothesis by pulse-chase metabolic labeling studies due to the low amounts of TCR (Y protein present in the pSRaRK- transfectants which, in turn, was likely related to the instability of unassociated TCR LY protein (Bonifacino et al., 1989; Alarcon et al., 1988a).

Although we show that the pSR aH, pSR aR-, and pSR aK- derived proteins were able to form a heterodimer with the heterologous TCR @ protein of the recipient TCR a negative REX cell line, the levels of cell surface expression observed were significantly less than that of the wild-type REX T cell line (Table I). These results would suggest that the HPB- ALL TCR LY chain does not pair efficiently with the TCR @ chain of REX. This unfavorable match may preclude the efficient assembly and, consequently, cell surface expression of this heterodimer pair. Recent studies by Saito and Germain (1989) have suggested that this might be a common occurrence in T cells, thus limiting the functional repertoir. Our unpub- lished data agrees with this observation. We have transfected the homologous Jurkat TCR (IL cDNA clone, PY14 (Yanagi et al., 1985), into clone 20A and have shown that surface expres- sion, as determined by immunofluorescence and radioiodina- tion, was similar to the wild-type T cell line (data not shown).

The presence of charged amino acid residues in the nor- mally hydrophobic environment of the transmembrane region is very unusual. Model studies using site-directed mutagenesis and observations of protein sequence on a number of resident transmembrane proteins have suggested that such residues may play a role in either assembly or function. Specifically, insertion of a charged residue within the transmembrane region of the vesicular stomatitis virus G protein (Adams and Rose, 1985) and the Rous sarcoma virus envelope glycoprotein (Davis and Hunter, 1987) prevents the cell surface expression of the proteins and, in the latter case, appears to direct the altered protein to an intracellular degradative pathway. In addition, a number of transmembrane proteins that appear to function as ion channels including the lac permease of E. coli (Kaback, 1988) and the &adrenergic receptors (Kobilka et al., 1988) possess several charged residues in the membrane spanning region.

Our study, as well as others (Morley et al., 1988; John et al., 1989), suggest an important role of the basic amino acid residues of the TCR proteins in assembly. Morley et al. (1988) changed the lysine 290 residue of the TCR /I chain of Jurkat to either arginine, glutamic acid, glutamine, serine, or leucine. They found that only a TCR p chain containing a lysine residue could restore the cell surface expression of a TCR /3- deficient mutant of Jurkat. More recently, John et al. (1989) have confirmed, in part, our observations. They mutated the TCR a chain of Jurkat by simultaneously changing the trans- membrane arginine 121 and lysine 126 residues to leucine and isoleucine, respectively. In comparison to the wild type Jurkat

Assembly and Function of the T Cell Receptor (Y Chain 14043

TCR 01 chain, removal of both charged residues prevented the reconstitution of cell surface expression of the TCR.CD3 complex in a TCR a-negative T cell line, MOLT-4. In this paper, we provide a potential reason for this defect in TCR. CD3 complex expression on the cell surface and show that the double mutation but not the single mutation prevents the assembly of TCR (Y into a functional TCR. CD3 complex.

Our studies suggest that in the absence of the transmem- brane lysine and arginine residues, TCR (Y f CD3 subcomplex and, consequently, TCR a@ heterodimer formation can not occur. In view of the observation that cell surface expression of the TCR. CD3 complex is, in part, dependent upon CD3 { (Alarcon et al., 1988b; Sussman et al., 1988; Sancho et al., 1989; Geisler et al., 1989), it is possible that in the absence of TCR a@ heterodimer formation the association of CD3 {and, consequently, the transport of the mature TCR. CD3 complex to the cell surface is impaired. Studies in mutants of the HPB- ALL cell line generated by chemical mutagenesis have shown that TCR c@ heterodimer formation is required for associa- tion of CD3 { into a functionally competent TCR. CD3 com- plex. The analysis of several mutants of HPB-ALL with variable expression of the TCR (Y chain, suggested a direct correlation between the amounts of TCR c@ heterodimer and CD3 2 within the TCR.CD3 complex (Sancho et al., 1989). Bonifacino et al. (1988) have also noted an association be- tween CD3 { and the TCR cup heterodimer in studies of subcomplexes in T cells. Finally, it has been shown, in a human T cell line, that the pentameric complex TCR cup. CD3 y, 6, t forms in the endoplasmic reticulum in the absence of CD3 { but is unable to be processed to maturity and be presented on the cell surface (Geisler et al., 1989).

Surprisingly, we also observed that expression on the cell surface of a complex containing either one or the other of the charged transmembrane residues had little effect on the func- tion of the complex as determined by the ability to mobilize intracellular calcium after stimulation with a mitogenic mono- clonal antibody directed against the CD3 e chain. This does not rule out the possibility that these amino acid substitutions may have other effects on the intracellular biochemical ma- chinery associated with T cell activation.

In summary, the arginine and lysine residues in the trans- membrane region of the TCR (Y chain appear to play an important role in assembly and cell surface presentation of the TCR. CD3 complex. In their absence, assembly of the TCR LY chain with either the TCR @ chain or the CD3 proteins is not observed. Consequently, cell surface expression of a functional complex does not occur. In contrast, the individual lysine and arginine residues can provide stability to the for- mation of functional complexes on the cell surface.

Acknowledgments-We are grateful to Karen Bentley for her expert secretarial assistance. We thank Dr. Tom Wileman for his critical review of the manuscript.

REFERENCES

Adams, G. A., and Rose, J. K. (1985) Mol. Cell Biol. 5.14421448 Alarcon, B., Berkhout, B., Breitmeyer, J., and Terhorst, C. (1988a) J. Biol.

Chcm. 263,2953-2961 Alarcon, B., Regueiro, J. R., Amaiz-Villena, A., and Terhorst, C. (1988b) N.

En&. J. Med. 319,1203-1208 Aruffo, A., and Seed, B. (1987) Proc. Natl. Acad. Sci. (I. S. A. 84, 8573-8577

Baniyash, M., Garcia-Morales, P., Bonafacino, J. S., Samelson, L. E., and Klausner, R. D. (1988) J. Biol. Chem. 263,9874-9878

Berkhout, B., Hall, C., and Terhorst, C. (1988a) Nucleic Acid Res. 11, 5209 Berkhout, B., Alarcon, B., and Terhorst, C. (1988b) J. Biol. Chm. 263,8528-

8536 Bonifacino, J. S., Chen, C., Lippincott-Schwartz, J., Ashwell, J. D., and Klaus-

ner, R. D. (1988) Proc. N&l. Acad. Sci. U. S. A. 85, 6929-6933 Bonifacino, J. S., Suzuki, C. K., Lippincott-Schwartz, J., Weissman, A. M., and

Klausner, R. D. (1989) J. Cell Biol. 109.73-83 Borst. J.. Alexander. S.. Elder. J.. and Terhorst. C. (1983) J. Biol. Chem. 258.

5135-5141 Breitmeyer, J. B., Oppenheim, S. O., Daley, J. F., Levine, H. B., and Schlossman,

S. F. (1987a) J. Immunol. 138,726-731 Breitmeyer, J. B., Daley, J. F., Levine, H. B., and Schlossman, S. F. (1987b) J.

Immuml. 139.2899-2905 Brenner, M. B., Trowbridge I. S., and Strominger, J. (1987) Cell 40, 183-190 Chomczynski, P., and Sac&i N. (1987) Anal. Biochm. 162, 156-159 Clevers, H., Alarcon, B., Wileman, T., and Terhorst, C. (1988) Annu. Reo.

Immuml. 6,629-662 Davis, G. L., and Hunter, E. (1987) J. Cell Biol. 105,1191-1203 De Waal Malefyt, R., Alarcon, B., Yssel, H., Sancho, J., Miyajima, A., Terhorst,

C. P., Spits, H., and De Vries, J. E. (1989) J. Immunol. 142, 3634-3642 Geisler, C., Kuhlmann, J., and Rubin, B. (1989) J. Immunol. 143,4069-4077 Gold, D. P., Puck, J. M., Pettey, C. L., Cho, M., Coligan, T. E., Woody, J. N.,

and Terhorst, C. (1986) Nature 321,431-434 John, S., Banting, G. S., Goodfellow, P. N., and Owen, M. J. (1989) Eur. J.

Immunol. 19.335-339 Jones, P. K. (1984) Methods Enzymol. 108,452-466 Kaback, H. R. (1988) Annu. Reu. Physiol. 50, 243-256 Kobilka, B. K., Kobilka, T. S., Daniel, K. Regan, J. W., Caron, M. G., and

Lefkowitz, R. J. (1988) Science 240,1310-1316 Koning, F., Lew, A. M., Maloy, W. L. Valas, R., and Coligan, J. E. (1988) J.

Immuml. 140,3126-3134 Krissansen, G. W., Owen, M. J., Verbi, W., and Crumpton, M. J. (1986) EMBO

J. 5.1799-1809 Kunkel, T. A. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 488-492 Ley, S. L., Tan, K. N., Sancho, J., Rosa, F., Kubo, R., Sy, M. S., and Terhorst,

C. (1989) Eur. J. Immunol., in press Marrack, P., and Kappler, J. (1987) Science 238,1073-1078 Meurer, S. C., Fitzgerald, K. A., Hussey, R. E., Hodgdon, J. C., Schlossman, S.

F., and Reinhers, E. L. (1983) J. Exp. Med. 157,705-719 Morley, B. J., Chin, K. N., Newton, M. E., and Weiss, A. (1988) J. Exp. Med.

168, 1971-1978 Oettgen, H. C., Pettey, C. L., Maloy, W. L., and Terhorst, C. (1986) Nature

320,272-275 Ohashi, P. S., Mak, T. W., van den Elsen, P., Yanagi, Y., Yoshikai, Y., Calman,

A. F., Terhorst, C., Stobo, J. D., and Weiss, A. W. (1985) Nature 316,606- mu4 - “ ”

Orloff, D. G., Frank, S. J., Robey, F. A., Weissman, A. M., and Klausner, R. D. (1989) J. Bill. Chem. 264.14812-14817

P&and, S., Oettgen, H., Bhan, A. K., and Terhorst, C. (1985) EMBO J. 4, 337-344

Pettey, C. L., Alarcon, B., Malin, R., Weinberg, K., and Terhorst, C. (1987) J. Biol. Chem. 262,4854-4859

Raulet, D. H. (1985) Annu. Reu. Immunol. 7,175-207 Reinhers, E. L., Hussey, R. E., and Schlossman, S. F. (1980) Eur. J. Immunol.

10,758-762 Saito, T., Weiss A., Gunter, K. C., Shevach, E. M., and Germain, R. N. (1987)

J. Immunol. 139.625-628 Saito, T., and Germain, R. N. (1989) J. Immunol. 143, 3379-3384 Sancho, J., Chatila, T., Wong, R. C. K., Hall, C., Blumberg, R., Alarcon, B.,

Geha, R. S., and Terhorst, C. (1989) J. Biol. C&em. 264, 20760-20769 Seed, B. (1987) Nature 329, 840-846 Sussman, J. J., Bonifacino, J. S., Lippincott-Schwartz, J., Weissman, A. M.,

Saito, T., Klausner, R. D., and Ashwell, J. D. (1988) Cell 52, 85-95 Takebe, Y., Seiki, M., Fujisawa, J. I., Hoy, I., Yokota, K., Arai, K. I., Yoshida,

M., and Arai, N. (1988) Mol. Cell. Biol. 8, 460-472 Tax, W. J. M., Willems, H. W., Reekess, P. P. M., Capel, P. J. A., and Koene,

R. A. P. (1983) Nature 304,445-447 Transy, C., Mingeon, P. E., Marshall, B., Stebbins, C., and Reinherz, E. L.

(1989) Eur. J. Immunol. 19,947-950 van den Elsen, P., Shepley, B. A., Borst, J., Coligan, J. E., Markhan, A. F.,

Orkin, S., and Terhorst, C. (1984) Nature 312, 413-418 Van Wauwe, J. P., de Mey, J. R., and Goosens, J. G. (1980) J. Immunol. 124,

2708-2713 Weiss, A., and Stobo, J. D. (1984) J. Exp. Med. 160, 1284-1299 Weiss, A., Imboden, J., Hardy, K., Manger, B., Terhorst, C., and Stobo, J.

(1986) Annu. Reu. Immonol. 4,593-619 Weissman, A. M., Hay, D., Orloff, D. G., Modi, W., Sevanez, H., O’Brien, S. J.,

and Klausner, R. D. (1988a) Proc. N&l. Acad. Sci. U. S. A. 85.9709-9713 Weissman, A. M., Baniyash, M., Hous, D., Samelson, LA., Burgess, W. H., and

Klausner, R. D. (1988b) Science 239, 1018-1020 Yanagi, Y., Ghan, A., Chin, B., Minden, M., and Mak, T. W. (1985) Proc. Natl.

Acad. Sci. U. S. A. 82,3430-3434 Yoshikai, Y., Anatoniou, D., Clark, D. P., Yanagi, Y., Snagster, R., van den

Elsen, P., Terhorst, C., and Mark, T. W. (1984) Nature 312, 521-524 Zanders, E. D., Lamb, J. R., Feldman, M., Green, N., and Beverly, P. C. L.

(1983) Nature 303, 625-627