Cell, Vol. 70, 585-593,
August 21, 1992, Copyright
0 1992 by Cell Press
Genetic Evidence for the Involvement of the Ick Tyrosine Kinase in Signal Transduction through the T Cell Antigen Receptor David B. Straus and Arthur Weiss Howard Hughes Medical Institute Department of Medicine Department of Microbiology and Immunology University of California, San Francisco San Francisco, California 94143
Summary Signaling through the T cell antigen receptor (TCR) results both in rapid increases in tyrosine phosphorylation on a number of proteins and in the activation of the phosphatidylinositol pathway. It is not clear how stimulation of the TCR leads to these signaling events. Mutants of the Jurkat T cell line have been previously isolated that fail to show increases in calcium following receptor stimulation. Analysis of one of these mutants, JCaMl, which is defective in the induction of tyrosine phosphorylation, revealed a defect in the expression of functional Ick tyrosine kinase. The lack of Ickactivity was caused in part by a splicing defect. Expression of the /c/r cDNA in JCaMl restores the ability of the cell to respond to TCR stimulation. These results indicate that Ick is required for normal signal transduction through the TCR. Introduction T cells are induced to proliferate and carry out differentiated functions following recognition of an antigen that has been processed and presented on the surface of other cells (Weiss, 1989; Crabtree, 1989). This activation event is initiated as a consequence of a number of ligand-receptor interactions, but its specificity relies on the recognition of antigen by the variable regions of the a and 8 chains of the T cell receptor (TCR) (Dembic et al., 1986; Saito et al., 1987). Efficient expression of the TCRa and 8 polypeptides on the surface of Tcells requires association with the CD3 chains (y, S, and E) and the < chain homodimer (Weiss and Stobo, 1984; Weissman et al., 1989). These associated proteins are responsible for the signal transduction functions of the receptor. Recent work has shown that both the 1; and E polypeptides have the capacity to transduce signals independently when their cytoplasmic domains are expressed as chimeras with heterologous transmembrane proteins (Irving and Weiss, 1991; Romeo and Seed, 1991; Letourneur and Klausner, 1992). The signaling capability of these proteins appears to reside within a short sequence of amino acids that is shared with a number of other putative signaling proteins (Letourneur and Klausner, 1992; Romeo et al., 1992). However, the mechanism by which these proteins couple the TCR to downstream signaling pathways is unknown. Several biochemical signaling events that follow TCR
stimulation have been identified. The earliest identified event is the induction of tyrosine phosphorylation of a number of proteins immediately following TCR stimulation (June et al., 1990a). Inhibition of tyrosine kinase activity prevents T cell activation (June et al., 1990b). It is not known how ligation of the TCR results in the induction of tyrosine phosphorylation, and it is also unknown which kinase(s) is involved. However, two src family kinases have been proposed as candidates. The fyn kinase can be coimmunoprecipitated with the TCR (Samelson et al., 1990), and overproduction of fyn (or expression of an activated form) appears to make Tcells hypersensitive to stimulation through the receptor (Cooke et al., 1991; Davidson et al., 1992). Similarly, the Ick kinase has been reported to immunoprecipitate with the TCR in one cell line (Burgess et al., 1991), and expression of an activated form of this kinase also causes hypersensitivity to stimulation through the TCR (Abraham et al., 1991). Stimulation of the receptor also leads to the induction of a phosphatidylinositol-specific phospholipase C activity (Imboden and Stobo, 1985). This increased activity results in the accumulation of inositol phosphates and diacylglycerol, causing the release of internal calcium stores and the activation of protein kinase C (Berridge, 1987). Treatment of T cells with phorbol esters and calcium ionophores mimics these effects and is sufficient to induce downstream activation events, suggesting that these effects are physiologically important (Truneh et al., 1985; Weisset al., 1984). Phospholipase C-r1 has been recently identified as one of the substrates associated with the elevation of tyrosine kinase activity after receptor stimulation (Park et al., 1991; Weiss et al., 1991). The increased tyrosine phosphorylation of phospholipase C is likely to be responsible for its increased catalytic activity, as has been described for its activation by growth factor receptors (Ullrich and Schlessinger, 1990). To help analyze the complex events of T cell activation, we have taken a genetic approach. Three mutants derived from the Jurkat T cell line were isolated by virtue of their failure to respond to stimulation through the TCR (Goldsmith and Weiss, 1987; Goldsmith et al., 1988, 1989). Despite normal expression of the receptor, these mutants fail to show induction of inositol phosphates and intracellular calcium following stimulation of the TCR. Heterokaryon analysis demonstrated that these mutants belong to separate complementation groups. We have continued the characterization of one of these mutants, JCaMl. Our results indicate that this mutant is defective in early signaling events that follow TCR stimulation owing to a loss of Ick tyrosine kinase function. Results JCaMl Is Defective Very Early in the Signaling Pathway The earliest detected biochemical signaling event that follows TCR stimulation is the induction of tyrosine phosphor-
Jurkat
Ick
JCaMl
fyn
/7/
106
-106
80 -
-
80
-
32.5
-
27.5
49.5 -
32.5 27.5 1
18.5-
L
Figure 1. Induction of Protein Tyrosine Phosphorylation Anti-TCR Stimulation in Jurkat and JCaMl
2
3
4
5
6
Figure 2. Activity of the Ick and fyn Kinases from Jurkat, JCaMl, and JCaM3 following
Cells were left unstimulated (or stimulated) for 2 min at 37OC with the anti-T@ MAb C305 or the anti-CD3 MAb 235, both at 1500 dilution of ascites. After lysis in 1% NP-40, equal amounts of protein from postnuclear supernatants equivalent to - 2 x 108cells were analyzed by 11% SDS-PAGE and immunoblotted with anti-phosphotyrosine antibody as described in Experimental Procedures. The positions of molecular size markers, in kilodaltons, are indicated.
ylation on a number of proteins (June et al., 199Oa). The JCaMl mutant is severely defective in this induction (Figure 1; Weiss et al., 1991). Analysisof cell lysates by immunoblotting with anti-phosphotyrosine antibody showed that there was no induction of phosphotyrosine on proteins following stimulation of JCaMl with an antibody against the p chain of the TCR (MAb C305; Figure 1, lane 5). This is in contrast to the response of the parental Jurkat cell line (Figure 1, lane 2). Either stimulation with the anti-CD3 antibody 235 or cross-linking of anti-receptor antibodies with secondary antibodies can induce a small increase in inositol phosphates and a calcium increase in JCaMl, although the increase is not maintained and does not result in the production of interleukin 2 (Goldsmith and Weiss, 1988). Similarly, stimulation of the mutant with MAb 235 resulted in an induction of tyrosine phosphorylation, although the response was weak and appeared to involve only a subset of tyrosine phosphoproteins. These findings illustrate that the JCaMl mutant is severely defective in the earliest steps of signal transduction through the TCR, though weak signaling events can be detected under some circumstances. Although the defect in JCaMl is very early in the signal transduction pathway, it is apparently not a defect in the TCR itself. Previous work demonstrated that fusion of either TCRa- or P-negative cells with JCaMl could complement the signaling defect (Goldsmith and Weiss, 1987). Furthermore, we have found that a CD8-c chain chimera, which is capable of providing all the signaling functions of the TCR (Irving and Weiss, 1991), is unable to bypass the signaling defect in JCaMl. The mutant was transfected
7 in lmmunoprecipitates
The Ick and fyn kinases were immunoprecipitated from Triton X-100 lysates containing equal amounts of protein (- 50 x 1 Oe cell equivalents)from the indicatedcell lines. An immunoprecipitation with normal rabbit serum (NW) from the Jurkat cell lysate was included as a negative control (lane 7). Kinase assays were performed using [y-“P]ATP, and the products were analyzed by 10% SDS-PAGE.
with the CDS-C chimera and assayed for the ability to respond to stimulation with antibodies against the CD8 domain. In two independent clones, there was no increase in calcium level following antibody stimulation (data not shown). This observation indicates that the defect in receptor-mediated signal transduction observed in JCaMl is downstream of the TCR itself. Ick Kinase Function Is Defective in JCaMl It is possible that a defect in a kinase that mediates signaling from the TCR could explain the JCaMl phenotype. Although the tyrosine kinase(s) required for the increase in tyrosine phosphorylation following T cell activation has not been identified, two src family kinases, fyn and Ick, have been proposed as candidates. We examined the activity of these two kinases in lysates of Jurkat, JCaMl , and JCaM3, another signal transduction mutant with a similar phenotype. Figure 2 shows the autophosphorylation capability of the Ick and fyn kinases in immunoprecipitates formed with antisera against amino-terminal peptides. We found that Ick function was severely defective in JCaMl lysates (Figure 2, lane 2). In contrast to Ick, all the lysates contained fyn kinase activity, although the level may be reduced somewhat in JCaM3 lysates. To determine whether the Ick protein was present in JCaMl, we immunoblotted lysates with a MAb directed against an epitope in the carboxy-terminal 384 aa of Ick (Ansotegui et al., 1991; Figure 3). lmmunoblots showed no detectable wild-type Ick protein in JCaMl compared with Jurkat. lmmunoblotting with an anti-lck antiserum directed against an amino-terminal peptide also failed to detect any wild-type Ick, although a low level of a - 49 kd protein specific to JCaMl lysates was detectable (data not shown). This smaller protein may represent a mutant form
Involvement 507
of Ick in TCR Signaling
of the Ick protein that is missing the epitope recognized by the MAb and lacks the ability to autophosphorylate (see below). The data presented in Figures 2 and 3 indicate that JCaMl is severely deficient in Ick kinase function owing to a defect in its expression.
106Structure of /c/r Transcripts in JCaMl Initial examination of Ick RNA by Northern blot analysis indicated that there were Ick transcripts present in JCaMl (data not shown). To examine the mRNA in more detail, the polymerase chain reaction (PCR) was used to amplify segments of Ick transcripts from Jurkat and JCaMl RNA.% Products of the expected size were amplified from the 5’ and S’ends of /ckcDNA (Figure 4A). However, using primers to amplify the middle of the Ick cDNA, the major product from JCaMl was - 150 bp shorter than expected (Figure 4A, lane 4). The same result was obtained when a different pair of primers was used to amplify this region of the Ick transcript (data not shown). Sl nuclease protection analysis of JCaMl RNA using end-labeled probes from this region of the Ick cDNA mapped the deletion to the 5’ and 3’ boundaries of exon 7 (Figure 4B; Ftouer et al., 1989). Sequencing of the PCR product amplified from JCaMl RNA confirmed that the Ick transcript from the mutant lacked exon 7 and instead contained exon 6 directly
80-
49.5 -
32.5 Figure 3. Ick Levels in Lysates from Jurkat and JCaMl Equal amounts of protein from NP-40 lysates of Jurkat and JCaMl (- 2.5 x lo8 cell equivalents) were analyzed by 8% SDS-PAGE and immunoblotted with the anti-lck MAb 183. The blot was developed with alkaline phosphatase-conjugated secondary antibody.
B.
A. Middle r--L--T
5’ probe III J JC
A
-
J
JC
672 1 Cm 872 603
-
603 -
we (I, -.1
% 234 -
310 $$I = 234 -
194-
123456
123456
-
Ku*--------A
(5’)
Sl analysis probes
Figure 4. Analysis of /c/r Transcripts kat and JCaMl
3’ probe
PCR primers
II,
-
from Jur-
(A) PCR productsamplifiedfrom /ckcDNAfrom Jurkat (lanes 1, 3, and 5) and JCaMl (lanes 2, 4, and 6). Ick cDNA was synthesized from Jurkat and JCaMl RNA using the 3’ primer, and segments of the cDNA corresponding to the 5, middle, and 3’ parts of the Ick cDNA were amplified with the pairs of primers shown in the diagram. The products were analyzed on an agarose gel containing ethidium bromide. (B) Sl nuclease mapping of the deletion end points in Ick transcripts from JCaMl RNA from Jurkat (lanes 1 and 4) and JCaMl (lanes 2 and 5) or yeast RNA (lanes 3 and 6) was hybridized with 5’ or 3’ end-labeled double-stranded probesasindicated.The hybridsweredigested with Sl nuclease and analyzed on a 5% ureaacrylamide gel. (C) Diagram of the Ick cDNA indicating the approximate locations of the primers used for PCR analysis and the probes used for Sl mapping.
Cdl 566
JCaMl/ vector
Jurkat III 10680-
+
-
JCaMl/lck +
-
+ ....‘
32.5 32.5 -
27.5 -
27.5-
Figure 5. Ick Level and Kinase Activity in JCaMl with the Ick cDNA or Vector
Cells Transfected
(A) lmmunoblot analysis of Ick level in Jurkat or the JCaMl transfectants. NP-40 lysates of the indicated cell lines were analyzed by 6% SDS-PAGE (2.5 x 1 O6 cell equivalents per sample) and immunoblotted with anti-lck as in Figure 3. (6) Ick kinase activity in immunoprecipitates from lysates of Jurkat or the JCaMl transfectants. Ick was immunoprecipitated from NP-40 lysates of the indicated cell lines (5 x 108cell equivalents per sample) and assayed for autophosphorylation activity as described in Experimental Procedures.
spliced to exon 8 (data not shown). The missing exon in Ick transcripts from JCaMl encodes 51 aa, including the putative ATP-binding site based on homology to other protein kinases (Hanks et al., 1988). The correct reading frame is maintained in the mutant transcript and its translation would generate a -50 kd protein. It is possible that the - 49 kd protein found in JCaMl lysates that is reactive with the anti-peptide antiserum may represent this truncated form of Ick. A small fraction of Icktranscripts from Jurkat generated an Si nuclease-resistant fragment of the same size as seen in the mutant RNA (Figure 48). This may reflect the presence of unspliced Ick transcripts in the RNA sample. However, the shorter fragment is observed using cytoplasmic RNA preparations from Jurkat. In addition, upon longer exposure, PCR amplification of the middle segment of Ick cDNA from Jurkat does show a band of the same size as seen in the mutant (data not shown). A low level of mutant Ick transcripts in Jurkat may indicate that one of the Ick alleles present in the parental cell is already defective, but poorly expressed. Ick Expression Can Rescue the Signaling Defect of JCaMl To determine whether the defect in Ick function was in fact responsible for the defect in signaling in JCaMl, we attempted to restore the wild-type phenotype by transfection with acDNAclone of Ick. The murine IckcDNA, which is - 98% identical to human at the amino acid level, was cloned into an Epstein-Barr virus-based episomal expression vector (Hambor et al., 1988) and transfected into JCaMl A polyclonal population of transfectants was se-
Figure 6. Induction of Protein Tyrosine Phosphorylation following Anti-Ti Stimulation of Jurkat and JCaMl Transfected with the /c/r cDNA or the Vector Alone Cells from the indicated lines were either left unstimulated (-) or stimulated with anti-Ti MAb C305 (+) for 2 min at 37OC. NP-40 lysates (2.5 x 10’ cell equivalents per sample) were analyzed by 10% SDSPAGE and immunoblotted with anti-phosphotyrosine antibody. The -26 kd protein present in the stimulated cell lysates from all the cell lines is observed inconsistently and is likely to be the light chain of the anti-receptor antibody.
lected for growth in medium with antibiotic and analyzed for Ick expression. lmmunoblot analysis showed that transfection with the cDNA, but not with the vector, restores Ick to levels comparable with those seen in the parental cell line (Figure 5A). In addition, kinase activity was present in Ick immunoprecipitates from JCaMl transfected with the Ick cDNA, but not from JCaMl transfected with the vector (Figure 58). Next, we examined the ability of the JCaMl transfectants to transduce signaling events following stimulation through the antigen receptor. Jurkat and the JCaMl transfectants were stimulated with an antibody against the TCR, and the induction of tyrosine phosphoproteins was examined by immunoblotting with anti-phosphotyrosine antibody (Figure 6). Transfection of the Ick cDNA into JCaMl restores both a basal level of tyrosine phosphorylation and the inducibility of tyrosine phosphoproteins following stimulation of the TCR. The activation of phospholipase C, as measured by the elevation of internal calcium levels following receptor stimulation, was also restored in the Ick transfectants (Figure 7). Jurkat and the JCaMl transfectants were loaded with the calcium binding dye indo-1, then stimulated with anti-TCR antibody. Changes in calcium were followed spectrofluorometrically. The JCaMl cells transfected with Ick have recovered the ability to increase calcium levels in response to TCR stimulation. Stimulation with the anti-CD3 antibody 235 results in increased levels of calcium in all of the cell lines. These results indicate that transfection of the Ick cDNA into JCaMl restores the ability of the mutant to induce early signaling events that normally follow stimulation through the TCR. We have also examined late signaling events in the JCaMl transfectants by analyzing the ability of phytohe-
Involvement 589
of Ick in TCR Signaling
32.5 -
27.5 -
18.3F 1KK
7
P--J-
2w
50
-l 100 200
14--PO ,
Figure 8. r-Associated In Vitro Kinase Activity in Lysates from Jurkat and the JCaMl Transfectants c Was immunoprecipitated from NP-40 lysates of the indicated cell lines, 50 x 10” cell equivalents each, using anti-peptide antiserum raised against a peptide derived from the cytoplasmic domain of i. lmmunoblotting showed that equivalent amounts of rare immunoprecipitated from Jurkat and JCaMl under these conditions (data not shown). lmmunoprecipitates were assayed for kinase activity as described in Experimental Procedures.
TIME
Figure 7. Induction of Calcium Levels in Jurkat or the JCaMl Transfectantsfollowing Stimulation with Anti-Ti orAntiCD3Antibodies Cells from Jurkat (A and B), JCaMl transfected with vector only (C and D), or JCaMl transfected with the /c/r cDNA (E and F) were loaded with the calcium binding indo-1. After equilibration at 37% in a spectrofluorimeter. the samples (1 x 106cells/ml) were stimulated with saturating amounts of anti-Ti MAb C305 (A, C, and E) or anti-CD3 MAb 235 (Et, D, and F). For the sample shown in (C), ionomycin was added at the indicated time to 1 uM to show that the cells were properly loaded with indo-1.
magglutinin to induce the activation antigen CD69 (Testi et al., 1989). Phytohemagglutinin is apolyclonal stimulator that requires TCR function to activate Tcells. The transfection of the Ick cDNA into JCaMl was able to restore fully the inducibility of CD69 in >90% of the transfectants (data not shown). This finding indicates both that Ick is expressed in almost all of the transfectants and that it is able to reconstitute late as well as early signaling events in JCaMl. TCR-Associated Kinase Activity in JCaMl A tyrosine kinase activity associated with immunoprecipitates of the c chain of the TCR from Jurkat cells has recently been described (Chan et al., 1991). We examined the kinase activity of < immunoprecipitates from Jurkat and the JCaMl transfectants to determine if Ick is required for this activity. Figure 8 shows that the phosphorylation of the c chain is greatly reduced in immunoprecipitates from
JCaMl transfected with the vector alone, but full activity was restored by transfection with the Ick cDNA. Phosphoamino acid analysis showed that the residual 5 phosphorylation in immunoprecipitates from the JCaMl-vector transfectant was almost entirely limited to threonine and serine residues. This is in contrast to the JCaMl -/cktransfectant, in which 5 phosphorylation in immunoprecipitates was predominantly on tyrosine residues (data not shown). Although Ick has not been found in 5 chain immunoprecipitates, these results suggest that Ick function is required for the tyrosine kinase activity associated with the < chain. Discussion The results described above identify the defect in a mutant T cell line that cannot respond to stimulation through the TCR. Initial studies indicated that the defect in JCaMl was very early in the signal transduction pathway, prior to the induction of phospholipase C activity or tyrosine phosphorylation. However the defect appeared to be downstream of the TCR since a CD84 chimera, which can bypass the TCR when stimulated through the CD8 extracellular domain (Irving and Weiss, 1991) was unable to provide signaling function when transfected into JCaMl. To determine if the defect was in a kinase that might mediate signaling from the TCR to downstream effectors, we examined the activity of the fyn and Ick kinases in the mutant. Our results demonstrated a severe loss of Ick kinase activ-
Cdl 590
ity and a lack of wild-type Ick protein in the JCaMl mutant. Analysis of Ick transcripts from the mutant indicated that there is a splicing defect resulting in the loss of exon 7 that encodes the putative ATP-binding site. The homologous exon of the fyn gene can be alternatively spliced to generate two forms of this kinase that are expressed in a tissuespecific manner (Cooke and Perlmutter, 1989). However, there is no evidence for an alternate exon 7 in the human Ick gene (Rouer et al., 1989). The molecular basis for the altered splicing has not yet been determined. This defect may exist in the parental cell line, since a low level of the mutant RNA form (
August 21, 1992, Copyright
0 1992 by Cell Press
Genetic Evidence for the Involvement of the Ick Tyrosine Kinase in Signal Transduction through the T Cell Antigen Receptor David B. Straus and Arthur Weiss Howard Hughes Medical Institute Department of Medicine Department of Microbiology and Immunology University of California, San Francisco San Francisco, California 94143
Summary Signaling through the T cell antigen receptor (TCR) results both in rapid increases in tyrosine phosphorylation on a number of proteins and in the activation of the phosphatidylinositol pathway. It is not clear how stimulation of the TCR leads to these signaling events. Mutants of the Jurkat T cell line have been previously isolated that fail to show increases in calcium following receptor stimulation. Analysis of one of these mutants, JCaMl, which is defective in the induction of tyrosine phosphorylation, revealed a defect in the expression of functional Ick tyrosine kinase. The lack of Ickactivity was caused in part by a splicing defect. Expression of the /c/r cDNA in JCaMl restores the ability of the cell to respond to TCR stimulation. These results indicate that Ick is required for normal signal transduction through the TCR. Introduction T cells are induced to proliferate and carry out differentiated functions following recognition of an antigen that has been processed and presented on the surface of other cells (Weiss, 1989; Crabtree, 1989). This activation event is initiated as a consequence of a number of ligand-receptor interactions, but its specificity relies on the recognition of antigen by the variable regions of the a and 8 chains of the T cell receptor (TCR) (Dembic et al., 1986; Saito et al., 1987). Efficient expression of the TCRa and 8 polypeptides on the surface of Tcells requires association with the CD3 chains (y, S, and E) and the < chain homodimer (Weiss and Stobo, 1984; Weissman et al., 1989). These associated proteins are responsible for the signal transduction functions of the receptor. Recent work has shown that both the 1; and E polypeptides have the capacity to transduce signals independently when their cytoplasmic domains are expressed as chimeras with heterologous transmembrane proteins (Irving and Weiss, 1991; Romeo and Seed, 1991; Letourneur and Klausner, 1992). The signaling capability of these proteins appears to reside within a short sequence of amino acids that is shared with a number of other putative signaling proteins (Letourneur and Klausner, 1992; Romeo et al., 1992). However, the mechanism by which these proteins couple the TCR to downstream signaling pathways is unknown. Several biochemical signaling events that follow TCR
stimulation have been identified. The earliest identified event is the induction of tyrosine phosphorylation of a number of proteins immediately following TCR stimulation (June et al., 1990a). Inhibition of tyrosine kinase activity prevents T cell activation (June et al., 1990b). It is not known how ligation of the TCR results in the induction of tyrosine phosphorylation, and it is also unknown which kinase(s) is involved. However, two src family kinases have been proposed as candidates. The fyn kinase can be coimmunoprecipitated with the TCR (Samelson et al., 1990), and overproduction of fyn (or expression of an activated form) appears to make Tcells hypersensitive to stimulation through the receptor (Cooke et al., 1991; Davidson et al., 1992). Similarly, the Ick kinase has been reported to immunoprecipitate with the TCR in one cell line (Burgess et al., 1991), and expression of an activated form of this kinase also causes hypersensitivity to stimulation through the TCR (Abraham et al., 1991). Stimulation of the receptor also leads to the induction of a phosphatidylinositol-specific phospholipase C activity (Imboden and Stobo, 1985). This increased activity results in the accumulation of inositol phosphates and diacylglycerol, causing the release of internal calcium stores and the activation of protein kinase C (Berridge, 1987). Treatment of T cells with phorbol esters and calcium ionophores mimics these effects and is sufficient to induce downstream activation events, suggesting that these effects are physiologically important (Truneh et al., 1985; Weisset al., 1984). Phospholipase C-r1 has been recently identified as one of the substrates associated with the elevation of tyrosine kinase activity after receptor stimulation (Park et al., 1991; Weiss et al., 1991). The increased tyrosine phosphorylation of phospholipase C is likely to be responsible for its increased catalytic activity, as has been described for its activation by growth factor receptors (Ullrich and Schlessinger, 1990). To help analyze the complex events of T cell activation, we have taken a genetic approach. Three mutants derived from the Jurkat T cell line were isolated by virtue of their failure to respond to stimulation through the TCR (Goldsmith and Weiss, 1987; Goldsmith et al., 1988, 1989). Despite normal expression of the receptor, these mutants fail to show induction of inositol phosphates and intracellular calcium following stimulation of the TCR. Heterokaryon analysis demonstrated that these mutants belong to separate complementation groups. We have continued the characterization of one of these mutants, JCaMl. Our results indicate that this mutant is defective in early signaling events that follow TCR stimulation owing to a loss of Ick tyrosine kinase function. Results JCaMl Is Defective Very Early in the Signaling Pathway The earliest detected biochemical signaling event that follows TCR stimulation is the induction of tyrosine phosphor-
Jurkat
Ick
JCaMl
fyn
/7/
106
-106
80 -
-
80
-
32.5
-
27.5
49.5 -
32.5 27.5 1
18.5-
L
Figure 1. Induction of Protein Tyrosine Phosphorylation Anti-TCR Stimulation in Jurkat and JCaMl
2
3
4
5
6
Figure 2. Activity of the Ick and fyn Kinases from Jurkat, JCaMl, and JCaM3 following
Cells were left unstimulated (or stimulated) for 2 min at 37OC with the anti-T@ MAb C305 or the anti-CD3 MAb 235, both at 1500 dilution of ascites. After lysis in 1% NP-40, equal amounts of protein from postnuclear supernatants equivalent to - 2 x 108cells were analyzed by 11% SDS-PAGE and immunoblotted with anti-phosphotyrosine antibody as described in Experimental Procedures. The positions of molecular size markers, in kilodaltons, are indicated.
ylation on a number of proteins (June et al., 199Oa). The JCaMl mutant is severely defective in this induction (Figure 1; Weiss et al., 1991). Analysisof cell lysates by immunoblotting with anti-phosphotyrosine antibody showed that there was no induction of phosphotyrosine on proteins following stimulation of JCaMl with an antibody against the p chain of the TCR (MAb C305; Figure 1, lane 5). This is in contrast to the response of the parental Jurkat cell line (Figure 1, lane 2). Either stimulation with the anti-CD3 antibody 235 or cross-linking of anti-receptor antibodies with secondary antibodies can induce a small increase in inositol phosphates and a calcium increase in JCaMl, although the increase is not maintained and does not result in the production of interleukin 2 (Goldsmith and Weiss, 1988). Similarly, stimulation of the mutant with MAb 235 resulted in an induction of tyrosine phosphorylation, although the response was weak and appeared to involve only a subset of tyrosine phosphoproteins. These findings illustrate that the JCaMl mutant is severely defective in the earliest steps of signal transduction through the TCR, though weak signaling events can be detected under some circumstances. Although the defect in JCaMl is very early in the signal transduction pathway, it is apparently not a defect in the TCR itself. Previous work demonstrated that fusion of either TCRa- or P-negative cells with JCaMl could complement the signaling defect (Goldsmith and Weiss, 1987). Furthermore, we have found that a CD8-c chain chimera, which is capable of providing all the signaling functions of the TCR (Irving and Weiss, 1991), is unable to bypass the signaling defect in JCaMl. The mutant was transfected
7 in lmmunoprecipitates
The Ick and fyn kinases were immunoprecipitated from Triton X-100 lysates containing equal amounts of protein (- 50 x 1 Oe cell equivalents)from the indicatedcell lines. An immunoprecipitation with normal rabbit serum (NW) from the Jurkat cell lysate was included as a negative control (lane 7). Kinase assays were performed using [y-“P]ATP, and the products were analyzed by 10% SDS-PAGE.
with the CDS-C chimera and assayed for the ability to respond to stimulation with antibodies against the CD8 domain. In two independent clones, there was no increase in calcium level following antibody stimulation (data not shown). This observation indicates that the defect in receptor-mediated signal transduction observed in JCaMl is downstream of the TCR itself. Ick Kinase Function Is Defective in JCaMl It is possible that a defect in a kinase that mediates signaling from the TCR could explain the JCaMl phenotype. Although the tyrosine kinase(s) required for the increase in tyrosine phosphorylation following T cell activation has not been identified, two src family kinases, fyn and Ick, have been proposed as candidates. We examined the activity of these two kinases in lysates of Jurkat, JCaMl , and JCaM3, another signal transduction mutant with a similar phenotype. Figure 2 shows the autophosphorylation capability of the Ick and fyn kinases in immunoprecipitates formed with antisera against amino-terminal peptides. We found that Ick function was severely defective in JCaMl lysates (Figure 2, lane 2). In contrast to Ick, all the lysates contained fyn kinase activity, although the level may be reduced somewhat in JCaM3 lysates. To determine whether the Ick protein was present in JCaMl, we immunoblotted lysates with a MAb directed against an epitope in the carboxy-terminal 384 aa of Ick (Ansotegui et al., 1991; Figure 3). lmmunoblots showed no detectable wild-type Ick protein in JCaMl compared with Jurkat. lmmunoblotting with an anti-lck antiserum directed against an amino-terminal peptide also failed to detect any wild-type Ick, although a low level of a - 49 kd protein specific to JCaMl lysates was detectable (data not shown). This smaller protein may represent a mutant form
Involvement 507
of Ick in TCR Signaling
of the Ick protein that is missing the epitope recognized by the MAb and lacks the ability to autophosphorylate (see below). The data presented in Figures 2 and 3 indicate that JCaMl is severely deficient in Ick kinase function owing to a defect in its expression.
106Structure of /c/r Transcripts in JCaMl Initial examination of Ick RNA by Northern blot analysis indicated that there were Ick transcripts present in JCaMl (data not shown). To examine the mRNA in more detail, the polymerase chain reaction (PCR) was used to amplify segments of Ick transcripts from Jurkat and JCaMl RNA.% Products of the expected size were amplified from the 5’ and S’ends of /ckcDNA (Figure 4A). However, using primers to amplify the middle of the Ick cDNA, the major product from JCaMl was - 150 bp shorter than expected (Figure 4A, lane 4). The same result was obtained when a different pair of primers was used to amplify this region of the Ick transcript (data not shown). Sl nuclease protection analysis of JCaMl RNA using end-labeled probes from this region of the Ick cDNA mapped the deletion to the 5’ and 3’ boundaries of exon 7 (Figure 4B; Ftouer et al., 1989). Sequencing of the PCR product amplified from JCaMl RNA confirmed that the Ick transcript from the mutant lacked exon 7 and instead contained exon 6 directly
80-
49.5 -
32.5 Figure 3. Ick Levels in Lysates from Jurkat and JCaMl Equal amounts of protein from NP-40 lysates of Jurkat and JCaMl (- 2.5 x lo8 cell equivalents) were analyzed by 8% SDS-PAGE and immunoblotted with the anti-lck MAb 183. The blot was developed with alkaline phosphatase-conjugated secondary antibody.
B.
A. Middle r--L--T
5’ probe III J JC
A
-
J
JC
672 1 Cm 872 603
-
603 -
we (I, -.1
% 234 -
310 $$I = 234 -
194-
123456
123456
-
Ku*--------A
(5’)
Sl analysis probes
Figure 4. Analysis of /c/r Transcripts kat and JCaMl
3’ probe
PCR primers
II,
-
from Jur-
(A) PCR productsamplifiedfrom /ckcDNAfrom Jurkat (lanes 1, 3, and 5) and JCaMl (lanes 2, 4, and 6). Ick cDNA was synthesized from Jurkat and JCaMl RNA using the 3’ primer, and segments of the cDNA corresponding to the 5, middle, and 3’ parts of the Ick cDNA were amplified with the pairs of primers shown in the diagram. The products were analyzed on an agarose gel containing ethidium bromide. (B) Sl nuclease mapping of the deletion end points in Ick transcripts from JCaMl RNA from Jurkat (lanes 1 and 4) and JCaMl (lanes 2 and 5) or yeast RNA (lanes 3 and 6) was hybridized with 5’ or 3’ end-labeled double-stranded probesasindicated.The hybridsweredigested with Sl nuclease and analyzed on a 5% ureaacrylamide gel. (C) Diagram of the Ick cDNA indicating the approximate locations of the primers used for PCR analysis and the probes used for Sl mapping.
Cdl 566
JCaMl/ vector
Jurkat III 10680-
+
-
JCaMl/lck +
-
+ ....‘
32.5 32.5 -
27.5 -
27.5-
Figure 5. Ick Level and Kinase Activity in JCaMl with the Ick cDNA or Vector
Cells Transfected
(A) lmmunoblot analysis of Ick level in Jurkat or the JCaMl transfectants. NP-40 lysates of the indicated cell lines were analyzed by 6% SDS-PAGE (2.5 x 1 O6 cell equivalents per sample) and immunoblotted with anti-lck as in Figure 3. (6) Ick kinase activity in immunoprecipitates from lysates of Jurkat or the JCaMl transfectants. Ick was immunoprecipitated from NP-40 lysates of the indicated cell lines (5 x 108cell equivalents per sample) and assayed for autophosphorylation activity as described in Experimental Procedures.
spliced to exon 8 (data not shown). The missing exon in Ick transcripts from JCaMl encodes 51 aa, including the putative ATP-binding site based on homology to other protein kinases (Hanks et al., 1988). The correct reading frame is maintained in the mutant transcript and its translation would generate a -50 kd protein. It is possible that the - 49 kd protein found in JCaMl lysates that is reactive with the anti-peptide antiserum may represent this truncated form of Ick. A small fraction of Icktranscripts from Jurkat generated an Si nuclease-resistant fragment of the same size as seen in the mutant RNA (Figure 48). This may reflect the presence of unspliced Ick transcripts in the RNA sample. However, the shorter fragment is observed using cytoplasmic RNA preparations from Jurkat. In addition, upon longer exposure, PCR amplification of the middle segment of Ick cDNA from Jurkat does show a band of the same size as seen in the mutant (data not shown). A low level of mutant Ick transcripts in Jurkat may indicate that one of the Ick alleles present in the parental cell is already defective, but poorly expressed. Ick Expression Can Rescue the Signaling Defect of JCaMl To determine whether the defect in Ick function was in fact responsible for the defect in signaling in JCaMl, we attempted to restore the wild-type phenotype by transfection with acDNAclone of Ick. The murine IckcDNA, which is - 98% identical to human at the amino acid level, was cloned into an Epstein-Barr virus-based episomal expression vector (Hambor et al., 1988) and transfected into JCaMl A polyclonal population of transfectants was se-
Figure 6. Induction of Protein Tyrosine Phosphorylation following Anti-Ti Stimulation of Jurkat and JCaMl Transfected with the /c/r cDNA or the Vector Alone Cells from the indicated lines were either left unstimulated (-) or stimulated with anti-Ti MAb C305 (+) for 2 min at 37OC. NP-40 lysates (2.5 x 10’ cell equivalents per sample) were analyzed by 10% SDSPAGE and immunoblotted with anti-phosphotyrosine antibody. The -26 kd protein present in the stimulated cell lysates from all the cell lines is observed inconsistently and is likely to be the light chain of the anti-receptor antibody.
lected for growth in medium with antibiotic and analyzed for Ick expression. lmmunoblot analysis showed that transfection with the cDNA, but not with the vector, restores Ick to levels comparable with those seen in the parental cell line (Figure 5A). In addition, kinase activity was present in Ick immunoprecipitates from JCaMl transfected with the Ick cDNA, but not from JCaMl transfected with the vector (Figure 58). Next, we examined the ability of the JCaMl transfectants to transduce signaling events following stimulation through the antigen receptor. Jurkat and the JCaMl transfectants were stimulated with an antibody against the TCR, and the induction of tyrosine phosphoproteins was examined by immunoblotting with anti-phosphotyrosine antibody (Figure 6). Transfection of the Ick cDNA into JCaMl restores both a basal level of tyrosine phosphorylation and the inducibility of tyrosine phosphoproteins following stimulation of the TCR. The activation of phospholipase C, as measured by the elevation of internal calcium levels following receptor stimulation, was also restored in the Ick transfectants (Figure 7). Jurkat and the JCaMl transfectants were loaded with the calcium binding dye indo-1, then stimulated with anti-TCR antibody. Changes in calcium were followed spectrofluorometrically. The JCaMl cells transfected with Ick have recovered the ability to increase calcium levels in response to TCR stimulation. Stimulation with the anti-CD3 antibody 235 results in increased levels of calcium in all of the cell lines. These results indicate that transfection of the Ick cDNA into JCaMl restores the ability of the mutant to induce early signaling events that normally follow stimulation through the TCR. We have also examined late signaling events in the JCaMl transfectants by analyzing the ability of phytohe-
Involvement 589
of Ick in TCR Signaling
32.5 -
27.5 -
18.3F 1KK
7
P--J-
2w
50
-l 100 200
14--PO ,
Figure 8. r-Associated In Vitro Kinase Activity in Lysates from Jurkat and the JCaMl Transfectants c Was immunoprecipitated from NP-40 lysates of the indicated cell lines, 50 x 10” cell equivalents each, using anti-peptide antiserum raised against a peptide derived from the cytoplasmic domain of i. lmmunoblotting showed that equivalent amounts of rare immunoprecipitated from Jurkat and JCaMl under these conditions (data not shown). lmmunoprecipitates were assayed for kinase activity as described in Experimental Procedures.
TIME
Figure 7. Induction of Calcium Levels in Jurkat or the JCaMl Transfectantsfollowing Stimulation with Anti-Ti orAntiCD3Antibodies Cells from Jurkat (A and B), JCaMl transfected with vector only (C and D), or JCaMl transfected with the /c/r cDNA (E and F) were loaded with the calcium binding indo-1. After equilibration at 37% in a spectrofluorimeter. the samples (1 x 106cells/ml) were stimulated with saturating amounts of anti-Ti MAb C305 (A, C, and E) or anti-CD3 MAb 235 (Et, D, and F). For the sample shown in (C), ionomycin was added at the indicated time to 1 uM to show that the cells were properly loaded with indo-1.
magglutinin to induce the activation antigen CD69 (Testi et al., 1989). Phytohemagglutinin is apolyclonal stimulator that requires TCR function to activate Tcells. The transfection of the Ick cDNA into JCaMl was able to restore fully the inducibility of CD69 in >90% of the transfectants (data not shown). This finding indicates both that Ick is expressed in almost all of the transfectants and that it is able to reconstitute late as well as early signaling events in JCaMl. TCR-Associated Kinase Activity in JCaMl A tyrosine kinase activity associated with immunoprecipitates of the c chain of the TCR from Jurkat cells has recently been described (Chan et al., 1991). We examined the kinase activity of < immunoprecipitates from Jurkat and the JCaMl transfectants to determine if Ick is required for this activity. Figure 8 shows that the phosphorylation of the c chain is greatly reduced in immunoprecipitates from
JCaMl transfected with the vector alone, but full activity was restored by transfection with the Ick cDNA. Phosphoamino acid analysis showed that the residual 5 phosphorylation in immunoprecipitates from the JCaMl-vector transfectant was almost entirely limited to threonine and serine residues. This is in contrast to the JCaMl -/cktransfectant, in which 5 phosphorylation in immunoprecipitates was predominantly on tyrosine residues (data not shown). Although Ick has not been found in 5 chain immunoprecipitates, these results suggest that Ick function is required for the tyrosine kinase activity associated with the < chain. Discussion The results described above identify the defect in a mutant T cell line that cannot respond to stimulation through the TCR. Initial studies indicated that the defect in JCaMl was very early in the signal transduction pathway, prior to the induction of phospholipase C activity or tyrosine phosphorylation. However the defect appeared to be downstream of the TCR since a CD84 chimera, which can bypass the TCR when stimulated through the CD8 extracellular domain (Irving and Weiss, 1991) was unable to provide signaling function when transfected into JCaMl. To determine if the defect was in a kinase that might mediate signaling from the TCR to downstream effectors, we examined the activity of the fyn and Ick kinases in the mutant. Our results demonstrated a severe loss of Ick kinase activ-
Cdl 590
ity and a lack of wild-type Ick protein in the JCaMl mutant. Analysis of Ick transcripts from the mutant indicated that there is a splicing defect resulting in the loss of exon 7 that encodes the putative ATP-binding site. The homologous exon of the fyn gene can be alternatively spliced to generate two forms of this kinase that are expressed in a tissuespecific manner (Cooke and Perlmutter, 1989). However, there is no evidence for an alternate exon 7 in the human Ick gene (Rouer et al., 1989). The molecular basis for the altered splicing has not yet been determined. This defect may exist in the parental cell line, since a low level of the mutant RNA form (
