Identification of a New Pyk2 Isoform Implicated in Chemokine and Antigen Receptor Signaling*

Pyk2 is a protein tyrosine kinase that links G-protein-coupled receptors, inflammatory cytokines, and extracellular stimuli that elevate intracellular calcium concentration with activation of the mitogen-activated protein kinase pathways and regulation of ion channel functions. Here we describe the identification, cloning, and characterization of a new isoform of Pyk2 (Pyk2-H) that is generated by alternative RNA splicing. Pyk2-H is mainly expressed in hematopoietic cells including T-cells, B-cells, and natural killer cells. Engagement of T-cell or B-cell antigen receptors leads to rapid tyrosine phosphorylation of Pyk2-H. Pyk2-H is also activated in response to the chemokines RANTES and macrophage inflammatory protein-1β in T cells. In addition, we show that glutathione S-transferase fusion proteins containing the carboxyl termini of Pyk2 and Pyk2-H bind to a different set of tyrosine-phosphorylated proteins in thymus lysates. Specific expression of Pyk2-H and its activation by antigens or chemokines in hematopoietic cells may contribute toward the generation of cell type-specific signals involved in host immune responses.

Pyk2 together with focal adhesion kinase (FAK) 1 constitute a distinct family of non-receptor protein tyrosine kinases that are regulated by extracellular stimuli (1,2). Pyk2 is predominantly expressed in the central nervous system and in cells and tissues derived from hematopoietic lineages (2)(3)(4). In neuronal cells, Pyk2 links various extracellular stimuli that lead to elevation of intracellular calcium concentration, such as activation of the G-protein-coupled receptors, nicotinic acetylcholine receptors, and membrane depolarization, with ion channel and mitogen-activated protein kinase functions (2,5). Furthermore, we have shown that Pyk2 acts in concert with the Src protein tyrosine kinase to activate the mitogen-activated protein kinase signaling pathway in response to activation of G i and G q coupled receptors in PC12 cells (5). Pyk2 was implicated as an upstream regulator of the JNK/SAPK signaling pathway in response to stress stimuli, inflammatory cytokines, or activation of angiotensin II receptors (6,7). It has also been demonstrated that Pyk2 plays a role in signaling by lymphocyte antigen receptors, integrin receptors, and chemokine receptors in hematopoietic cells (9 -12).
We and others have observed that in some cells of hematopoietic origin Pyk2 protein migrates in SDS-PAGE gels as a doublet (9,11). This observation suggested that different Pyk2 isoforms may be expressed in hematopoietic cells. In this report we describe the identification, cloning, and initial characterization of an isoform of Pyk2, that is generated by alternative RNA splicing, and is mainly expressed in hematopoietic cells (Pyk2-H). Using isoform-specific anti-Pyk2 antibodies we show that Pyk2-H is expressed in primary T and B lymphocytes and in natural killer cells. We demonstrate that Pyk2-H and not Pyk2 is rapidly activated in response to stimulation with chemokines and following activation of the T-cell or B-cell antigen receptors suggesting a potential role in host immune responses.

EXPERIMENTAL PROCEDURES
Reagents-All tissue culture media and antibiotics were obtained from Life Technologies, Inc. The In vitro mutagenesis kit was from CLONTECH, GeneAmp XL PCR Kit from Perkin-Elmer, TA cloning kit from Invitrogen, anti-mouse CD3 antibodies from PharMingen, goat anti-hamster IgG from Boehringer Mannheim, affinity purified goat anti-mouse IgM F(abЈ) 2 fragment, and rabbit anti-goat IgG antibodies (F(abЈ) 2 specific) were from Jackson ImmunoResearch Laboratories, LipofectAMINE was from Life Technologies, Inc., RANTES and macrophage inflammatory protein-1␤ were from PeproTech. Murine brain and thymus cDNAs were provided by Zoran Bogdanovic (New York University, New York). A murine genomic clone containing Pyk2 sequences (phage clone P1) was purchased from Genome Systems, Inc. BALB/c mice were obtained from the Laboratory Animal Facility of the NYU Medical Center.
Cell Culture-Human promyelocytic leukemia HL-60 cells, human T cell leukemia Jurkat cells, human erythroleukemia K562 cells, rat mast/basophilic leukemia RBL cells, human erythroleukemic Mo7 cells, and human premegakaryocyte leukemia CHRF cells were maintained in RPMI containing 10% fetal calf serum (FCS). Rat pheochromocytoma PC12 cells were maintained in Dulbecco's modified Eagle's medium containing 7% FCS and 7% horse serum. Human DU6 cell line was obtained from Derya Untumaz (Skirball Institute, New York) and were cultured in RPMI containing 10% FCS and supplemented with 100 units/ml human IL2 (Chiron) as described previously (12). Lysates of human platelets were obtained from Carl-Henrik Heldin (Ludwig Institute, Uppsala, Sweden). Lysates of purified natural killer cell population were provided by Angela Gismondi (University of Rome, Rome, Italy). Natural killer cells were obtained from human peripheral mononuclears isolated from buffy coats by gradient centrifugation and passage through nylon wool columns. Nylon non-adherent cells were further purified by negative cell sorting using a mixture of anti-CD5 and anti-CD14 monoclonal antibodies (13). Murine thymocytes were isolated as described previously (9). Briefly, single cell suspensions were collected from mice at 6 weeks of age and maintained in RPMI medium supplemented with 5% FCS. B cells were depleted by addition of sheep anti-mouse immunoglobulin-coated magnetic beads (Dynabeads, Dynal Inc.). For each experiment, 10 8 cells/ml were stimulated for the indicated times by addition of anti-CD3 antibodies (50 g/ml). Following stimulation the cells were rapidly centrifuged using Picofuge (Stratagene) and washed once in cold phosphate-buffered saline and the pellet was lysed in lysis buffer. Enriched populations of B cells were obtained from murine splenocytes by depletion of T cells with anti-Thy1, anti-CD4, and anti-CD8 antibodies as described elsewhere (14). For stimulation, B cells (approximately 5 ϫ 10 8 /ml) were incubated in serum-free RPMI containing goat anti-mouse IgM (50 g/ml) for 5 min and then added with anti-goat IgG antibodies for 15 min at 37°C. DU6 T cells were starved in RPMI containing 0.5% FCS and then stimulated with 0.5 M chemokines for indicated time periods, spun down with Picofuge and immediately lysed, essentially as described previously (12). For transient transfections, human embryonic kidney 293T cells were grown to 50% confluency on 10-cm tissue culture dishes in Dulbecco's modified Eagle's medium supplemented with 10% FCS and transfected with 10 g of pRK5-PYK2, 10 g of pRK5-Pyk2-H, 10 g of pRK5-Pyk2-Y402F, or 10 g of pRK5-Pyk2-H-Y402F expression vectors, using LipofectAMINE reagent.
Fusion Proteins and Antibodies-Glutathione S-transferase (GST) fusion proteins were generated by subcloning PCR products containing amino acid residues 684 to 1009 (GST-Pyk2-CT), amino acid residues 739 to 780 (GST-Pyk2-AS) of Pyk2, or amino acid residues 684 to 1009 deleted for 42 amino acids that are spliced out (GST-Pyk2H-CT) of Pyk2-H in the BamHI and EcoRI site of pGex2T vector (Pharmacia Biotech). GST fusion proteins were purified on glutathione-agarose as described previously (14). Polyclonal antibodies against Pyk2 were raised in rabbits injected with the GST fusion protein containing the following amino acids of human Pyk2: residues 684 to 1009 (rabbit number 598), residues 684 to 762 (rabbit number 600), residues 739 to 780 (rabbit number 638), or with the peptide corresponding to the last 17 residues of Pyk2 (rabbit number 623). Anti-Pyk2 antibodies (KD and N-18) were previously described (2). Mouse monoclonal anti-paxillin and anti-Cas antibodies were obtained from Transduction Laboratories. Mouse monoclonal anti-phosphotyrosine antibodies (4G10) were obtained from Dr. Craig Davis or rabbit polyclonal anti-phosphotyrosine antibodies (rabbit number 72) were used as described previously (12,15,16).
In Vitro Kinase Reaction and Identification of Autophosphorylation Sites in Pyk2 and Pyk2-H-For exogenous substrate phosphorylation, equal amounts of Pyk2 proteins were immunoprecipitated from 293T cells transiently transfected with vector alone (pRK5) or vector containing Pyk2 or Pyk2-H. The immunoprecipitates were washed three times with lysis buffer, once with kinase buffer (50 mM Tris-HCl, pH 7.5, 5 mM MnCl 2 , 5 mM MgCl 2 ), and incubated with 50 l of kinase buffer containing 20 g of poly(Glu-Tyr) (4:1) and 5 Ci of [␥-32 P]ATP (NEN Life Science Products Inc.) for 10 min at room temperature. The reaction was stopped by addition of 3 ϫ sample buffer (Bio-Rad) and resolved on 7% SDS-PAGE gels. Fold increase in phosphorylation of poly(Glu-Tyr) was determined by quantitation with a Molecular Dynamics Phospho-Imager and ImageQuant software (Molecular Dynamics). For identification of autophosphorylation sites, 293T cells were transiently transfected with pRK5 containing Pyk2, Pyk2-H, or mutant forms in which tyrosine 402 was replaced with phenylalanine (Pyk2-Y402F and Pyk2-H-Y402). The immunoprecipitated proteins were incubated in 50 l of kinase buffer containing 15 Ci of [␥-32 P]ATP for 7 min at room temperature. 32 P-Labeled autophosphorylated Pyk2 or Pyk2-H proteins were digested with trypsin and labeled peptides were separated on a reverse-phase HPLC column as described previously (17). Fractions of 0.5 ml were collected and 32 P radioactivity was analyzed with a ␤-counter.
Cloning of Pyk2-H, in Vitro Mutagenesis, and PCR Analysis-cDNAs of murine brain and thymus tissues were used to amplify PCR products with pairs of oligonucleotides corresponding to the amino-terminal or carboxyl-terminal regions of Pyk2 (ATGTCTGGGGTGTCCGAGC and ATCTTCACGGGCAATGCCA for the amino terminus; AGTGACGTT-TATCAGATG and TCACTCTGCAGGTGGGTG for the carboxyl-terminal domain of Pyk2). The PCR was performed under following conditions: 30 s at 94°C, 30 s at 55°C, 1 min at 72°C for 35 cycles and extension at 72°C for 10 min. The PCR products were subcloned in the TA vector and analyzed by DNA sequencing. The mutagenic oligonucleotide (CCAAAGCTGCAGTTCCAGGAGGAGGACTTCATCCAAC) and the trans-oligonucleotide (GAATTCGATATCACGCGTGGCCGC-CATGGC) were used to delete 126 base pairs of Pyk2 using a CLON-TECH kit. The deletion was confirmed by DNA sequencing. In order to obtain exon-intron boundaries around the spliced exon, a mouse genomic clone of Pyk2 (P1) was subjected to the long range PCR using the following oligonucleotides (CCAAAGCTGCAGTTCCAG and ACA-CAGACCCTCAGGAAC for cloning of the preceding intron; AAACGC-CACAGCATGCGG and TTGGATGAAGTCCTCCTC for cloning of the succeeding intron) according to the manufacturer's recommendation (Perkin-Elmer). The mutagenic oligonucleotide (GAGTCAGACATTCT-TCGCAGAGATTCC) and the trans-oligonucleotide (GAATTCGATAT-CACGCGTTGGCCGCCATGGC) were used to convert tyrosine 402 to phenylalanine of Pyk2-H using the In Vitro Mutagenesis CLONTECH kit. The mutation was confirmed by DNA sequencing.

Different Forms of Pyk2 Are Expressed in Murine Tissues-
Site-specific rabbit polyclonal antibodies against synthetic peptides from Pyk2 or against GST fusion proteins containing different domains of Pyk2 have been generated (Fig. 1A). All our anti-Pyk2 antibodies were specific for Pyk2 and did not recognize the related protein tyrosine kinase FAK (data not shown). In our attempt to study Pyk2 functions, Pyk2 was subjected to immunoprecipitation from lysates of mouse brain, spleen, thymus, and other tissues with anti-Pyk2 antibodies (number 600) followed by SDS-PAGE analysis, Commassie Brilliant Blue staining (Fig. 1B), or by immunoblotting with antibodies that recognize the kinase domain of Pyk2 (Fig. 1, C and D). Pyk2 immunoprecipitated from hippocampus, olfactory bulb, cortex, diencephalon, as well as other brain tissues migrates in SDS-PAGE gels with an apparent molecular weight of 110,000 (Fig. 1B). However, Pyk2 immunoprecipitated from lysates from spleen or thymus tissues migrates in the SDS gel as a doublet in which the major band has an apparent molecular weight of 106,000 and the minor band of 110,000 (Fig. 1C). In addition, the presence of p106 and p110 Pyk2 forms was also observed in liver (two forms of Pyk2 in liver and thymus are clearly seen in the shorter exposure) and to a lesser extend in heart and lung (Fig. 1D). These results suggest the presence of different Pyk2 isoforms, differential phosphorylation of Pyk2, or degradation products in the spleen and thymus, as compared with Pyk2 expressed in neuronal cells. In order to avoid possible proteolysis and to test whether the two forms of Pyk2 present in spleen and thymus are due to differences in phosphorylation state, we have included several inhibitors against the broad spectrum of proteases (Protease inhibitors set, Boehringer Mannheim) in our lysis buffer and have subjected immunoprecipitated Pyk2 from spleen and thymus to dephosphorylation with either alkaline or potato phosphatases, respectively. However, these treatments did not affect the pattern of Pyk2 migration in SDS-PAGE gels (data not shown), suggesting that the two forms of Pyk2 are probably different Pyk2 isoforms present in spleen and thymus.
Cloning of an Alternative Splice Isoform of Pyk2 from Murine Thymus-In order to identify the nature of the hematopoietic Pyk2 isoform, cDNAs were prepared from murine brain or thymus mRNAs and used in PCR with pairs of oligonucleotides corresponding to the non-catalytic amino-and carboxyl-terminal regions of Pyk2 ( Fig. 2A). The primers corresponding to the amino-terminal region of Pyk2 gave an identical PCR product when the reaction was performed with cDNA from thymus or brain tissues ( Fig. 2A). However, the major PCR product observed using the carboxyl-terminal primers and cDNA from thymus as a template was smaller by approximately 120 base pairs from the PCR reaction products obtained with identical primers on brain cDNA or when Pyk2 cDNA was used as a template ( Fig. 2A). In addition, a small amount of full size product was also detected in the PCR reaction with 3Ј primers with thymus cDNA (Fig. 2A). This is consistent with the rela-tive abundance of the p106 Pyk2 isoform versus the full size p110 Pyk2 that were detected by immunoprecipitation analysis of lysates from thymus (Fig. 1C). Deduced amino acid sequences of the reaction products indicates that the short PCR product represents an in-frame deletion of 126 base pairs which eliminates 42 amino acid from the proline-rich region of Pyk2. Fig. 2B shows a comparison of the nucleotide and amino acid sequences of the proline-rich region of Pyk2 and of the shorter proline-rich region of the Pyk2 isoform cloned from thymus cDNA. To confirm that the shorter form of Pyk2 is generated by alternative RNA splicing, we have cloned the corresponding genomic regions of murine Pyk2. The exon that is spliced out (126 base pairs) in the shorter form of Pyk2 is preceded by a 2.2-kilobase intron and followed by a 4.8-kilobase intron in the mouse Pyk2 gene (Fig. 2C). Since the shorter form of Pyk2 was cloned from and is predominantly expressed in hematopoietic tissues this form was named Pyk2-H.
Isoform-specific Anti-Pyk2 Antibodies (Number 638) Recognize Pyk2 and Not Pyk2-H-In order to compare and analyze the properties of Pyk2 and Pyk2-H proteins we have raised rabbit polyclonal antisera against a GST fusion protein containing the amino acid sequence corresponding to the exon that is spliced out from Pyk2 to generate Pyk2-H (number 638). 293T cells were transiently transfected with mammalian vectors containing the Pyk2 or Pyk2-H genes in human embryonic 293T cells. Lysates from the transfected cells were subjected to immunoprecipitation with antisera numbers 638 or 600 followed by SDS-PAGE and immunoblotting with antibodies number 623, that recognize both Pyk2 and Pyk2-H (Fig. 3A). Antisera number 638 recognizes Pyk2, but not Pyk2-H, while antisera 600 immunoprecipitated both Pyk2 and Pyk2-H from transfected 293 cells, indicating that these antisera recognize the Pyk2, but not the Pyk2-H isoform.
Expression Pattern of Pyk2 and Pyk2-H in Hematopoietic and Neuronal Cells-We next compared the pattern of Pyk2 and Pyk2-H protein expression in the thymus and spleen by using isoform-specific 638 antibodies that could distinguish Pyk2 from Pyk2-H, or number 600 antibodies that recognize both isoforms. Lysates from murine brain, spleen, and thymus were subjected to immunoprecipitation with antisera 638 or with antisera 600 followed by SDS-PAGE and immunoblotting with antibodies number 623 (Fig. 3B). Similar amounts of Pyk2 were immunoprecipitated from brain lysates with either antibodies 600 or 638, while expression of Pyk2-H was approximately 10-fold higher than expression of Pyk2 in the spleen and thymus (Fig. 3B). To test whether Pyk2 and Pyk2-H are co-expressed in the same cell or whether each isoform is present in a different cell type present in these tissues, we explored the expression pattern of Pyk2 and Pyk2-H in various hematopoietic cell types by using the same immunoprecipitation and immunoblotting strategy. We found that Pyk2-H is expressed in T cells, B cells, and natural killer cells, while Pyk2 is expressed in platelets, mast/basophilic cell line (RBL), megakaryocytic (CHRF), and erythroleukemic cell lines (Mo7, K562) (Fig.  3C). Both Pyk2 and Pyk2-H are present in human T-cells leukemia Jurkat cells and human promyelocytic leukemia HL-60 cells (Fig. 3C).

Pyk2-H Is Activated by Chemokine and Antigen Receptors in T and B Cells-Given the fact that Pyk2-H is present in primary T and B cells we further tested whether
Pyk2-H participates in signaling pathways that are part of the host immune response, such as antigen and chemokine receptors signaling pathways. We have therefore examined the status of Pyk2-H phosphorylation in response to antigens and inflammatory cytokines in primary hematopoietic cells. Murine B cells were stimulated with goat anti-mouse IgM (50 g/ml) and anti-goat IgG antibodies. The stimulated cells were lysed, subjected to immunoprecipitation with anti-Pyk2 antibodies (numbers 600 or 638) followed by immunoblotting with antibodies against phosphotyrosine or anti-Pyk2 (number 623). Murine thymocytes were stimulated with anti-CD3 monoclonal antibodies for 5 min at 37°C, lysed, and subjected to immunoprecipitation with anti-Pyk2 antibodies (numbers 600 and 638). The precipitates were resolved on 7% SDS-PAGE gels and immunoblotted with anti-phosphotyrosine antibodies or anti-Pyk2 (number 623) antibodies. The experiment presented in Fig. 4A shows that activation of T-cell or B-cell antigen receptors leads to a significant increase in tyrosine phosphorylation of Pyk2-H (Fig. 4A). This experiment also confirms that Pyk2-H and not Pyk2 is FIG. 1. Expression pattern of Pyk2 isoforms in brain and hematopoietic tissues. A, schematic diagram of antibodies against Pyk2 used in these studies; B, approximately 1.5 mg of total tissue lysates of mouse brain and thymus were immunoprecipitated with anti-Pyk2 antibodies (number 600) or preimmune antisera and stained with Commassie Brilliant Blue. The arrows indicate two isoforms of Pyk2 present in lysates of spleen and thymus and heavy chains of IgG. C, lysates of different sections of rat brains, or lysates of mouse brain, spleen, and thymus were immunoprecipitated with anti-Pyk2 antibodies (number 600) followed by immunoblotting with antibodies against Pyk2 (KD). D, lysates of different murine tissues were immunoprecipitated with anti-Pyk2 antibodies (number 600) followed by immunoblotting with antibodies against Pyk2 (KD). The arrows indicate the p106 and p110 forms of Pyk2. The separation of p106 Pyk2 from the p110 isoform in liver and thymus is better seen in the shorter exposure. exclusively expressed in murine thymocytes and B cells (Fig. 4A).
We have previously observed that stimulation of chemokine receptors by HIV envelope proteins or chemokines leads to rapid tyrosine phosphorylation of Pyk2 in several hematopoietic cells (12). We therefore examined whether activation of chemokine receptors in the human T cell line (DU6) leads to tyrosine phosphorylation and activation of Pyk2-H. DU6 cells were stimulated with 0.5 M RANTES and macrophage inflammatory protein-1␤ for 30 s, lysed, and subjected to the same immunoprecipitation/immunoblotting procedures. Immunoprecipitated Pyk2-H is rapidly tyrosine phosphorylated following stimulation of DU6 cells by chemokines (Fig. 4B). Taken together, these experiments demonstrate that in primary T and B cells Pyk2-H is activated upon engagement of chemokine and antigen receptors (Fig. 4, A and B).
Splicing Event in Pyk2-H Does Not Change Its Catalytic Activity-Since it has been observed that deletions in the carboxyl-terminal domain of Pyk2 might lead to changes in enzyme activity (18), we further examined whether structural changes in the carboxyl terminus of Pyk2-H might have altered activation of this kinase as compared with kinase activity of Pyk2. We have compared in vitro kinase activities of Pyk2 and Pyk2-H toward the exogenous substrate (poly(Glu-Tyr)) or by measuring phosphorylation of tyrosine 402, a major autophosphorylation site of Pyk2. Immunoprecipitated Pyk2-H and Pyk2 were subjected to in vitro kinase reaction, run to SDS-PAGE gels, and intensity of 32 P-labeled poly(Glu-Tyr) were compared. The experiments show that Pyk2-H and Pyk2 ex- 2. Cloning of an alternatively spliced isoform of Pyk2. A, PCR was performed using cDNA of mouse brain and thymus and oligonucleotides (arrows) corresponding to the amino-or carboxyl-terminal regions of Pyk2, respectively. Pyk2 cDNA subcloned in pRK5 (2) was used as a positive control; B, comparison of the amino acid sequences of the carboxyl-terminal region of Pyk2 from brain and thymus. Nucleotide sequence of Pyk2-H reveals consensus sequences for RNA splicing in higher eukaryotes (bold). C, XL PCR was performed using genomic clone of Pyk2 as a template and oligonucleotides corresponding to the adjacent exons in the mouse Pyk2 gene. Schematic presentation of the spliced exon and the adjacent introns in a murine genomic Pyk2 clone. hibit similar in vitro catalytic activities, measured by in vitro phosphorylation of poly(Glu-Tyr) (Fig. 5A). In order to compare the autophosphorylation sites of Pyk2 and Pyk2-H we have subjected the 32 P-labeled autophosphorylated Pyk2-H, Pyk2, or their mutant forms where tyrosine 402 was mutated to phenylalanine (Pyk2-H-Y402F and Pyk2-Y402F) to a tryptic digest followed by separation using reverse-phase HPLC. Three phosphorylated peptides were detected in preparations of Pyk2-H and Pyk2. Phosphorylated peptides obtained from Pyk2-H-Y402F and Pyk2-Y402F proteins have only two peaks, suggesting that the first peak observed in Pyk2-H or Pyk2 separations corresponds to peptide containing autophosphorylated tyrosine 402. Given the fact that more than 80% of labeled 32 P was incorporated in that single peptide in both Pyk2 and Pyk2-H preparations, we concluded that tyrosine 402 is a major autophosphorylation site of Pyk2-H and Pyk2. Together, these results suggest that structural changes in the carboxyl terminus of Pyk2 did not have a major affect on the kinase activity of Pyk2-H.
The Carboxyl Termini of Pyk2-H and Pyk2 Bind Different Tyrosine-phosphorylated Proteins-The carboxyl-terminal domain of Pyk2 contains two proline-rich regions that could be responsible for binding to proteins containing SH3 or WW domains (2). Given the fact that half of the first proline-rich region is deleted by a splicing event in Pyk2-H ( Fig. 2A), we further examined whether this region could mediate specific interactions of Pyk2 with different cellular proteins. In order to screen for proteins that could bind differently to Pyk2 or Pyk2-H, the carboxyl termini of Pyk2 and Pyk2-H or the exon that is spliced out in Pyk2-H were expressed as GST fusion proteins (Fig. 6A) and used to precipitate proteins from thymus lysates. The precipitates were separated on 6% SDS-PAGE gels, transferred to nitrocellulose membranes, and immunoblotted with the indicated antibodies. The region encoded by the exon that is spliced out in Pyk2-H recognized similar pattern of tyrosine-phosphorylated proteins to those recognized by GST alone (Fig. 6B). Analysis of precipitated proteins with GST fusion proteins containing the entire carboxyl termini of Pyk2 or Pyk2-H has revealed several common tyrosine-phosphorylated proteins with molecular mass of 62-70 kDa and a tyrosine-phosphorylated protein of approximately 115 kDa that bind specifically to the carboxyl terminus of Pyk2 but not to the carboxyl terminus of Pyk2-H (Fig. 6B). Using specific antibodies we demonstrate that paxillin is the 62-70-kDa associated protein with Pyk2 and Pyk2-H (Fig. 6B). A weak association of p130 Cas with Pyk2 was previously shown in B cells (11). We therefore examined the possibility whether tyrosine-phosphorylated p115 could be p105 Hef1 or p130 Cas . By using anti-Cas antibodies that recognize both p105 Hef1 and p130 Cas (Fig. 6B) we could not detect p130 Cas or p105 Hef-1 that co-precipitated with the carboxyl termini of both Pyk2 or Pyk2-H in thymus lysates, suggesting that tyrosine-phosphorylated associated p115 protein is distinct from p105 Hef1 or p130 Cas (Fig. 6C). We therefore conclude that the carboxyl termini of Pyk2-H and Pyk2 bind both common and unique target proteins in lysates from activated thymus cells.

DISCUSSION
In this report, we describe identification and cloning of a novel Pyk2 isoform that is generated by alternative RNA splicing in the carboxyl terminus of Pyk2. Pyk2-H is mainly expressed in hematopoietic tissues and cells, with the highest expression in lysates of the spleen and thymus, where immunoprecipitated Pyk2-H could be detected by Commassie Brilliant Blue staining (Fig. 1B). In addition, expression of Pyk2-H appears to be restricted to specific hematopoietic cells, such as thymocytes, B cells, and natural killer cells, while Pyk2 is present in platelets, megakaryocytes, mast cells, and neuronal cells. We have also observed the presence of Pyk2-H in the heart, liver, and lung, although to a much lower level than in the thymus and spleen. Since the liver, heart, and lung contain blood and since several cell lines derived from these organs express either only Pyk2 or neither of the Pyk2 isoforms (data not shown) we believe that Pyk2-H detected in immunoprecipitates from these organs is most likely due to circulating hematopoietic cells, rather than specific cells of these tissues.
Since the expression of Pyk2-H is restricted to specific hematopoietic cell lineages, whereas Pyk2 is expressed mainly in neuronal cells, an alternative splicing event may provide specificity in signaling between Pyk2 and Pyk2-H in hematopoietic versus neuronal cells. The role of different spliced isoforms of tyrosine kinases and phosphatases, such as Fyn and CD45, in hematopoietic cells have been previously reported (19,20). Alternatively spliced isoform of Fyn in cells of hematopoietic lineages (FynT) was shown to have different signaling properties in T lymphocytes as compared with the brain Fyn (FynB) isoform (19). It has been demonstrated that structural changes in the amino-terminal part of kinase domains of FynT and FynB lead to differences in their abilities to promote antigen receptor-triggered calcium fluxes (21). In addition, a specific role for different spliced isoforms of FAK, a Pyk2-related tyrosine kinase, has also been demonstrated (22)(23)(24). Activation of an alternatively spliced form of FAK by anandamide appears to link cannabinoid G-protein-coupled receptors and tyrosine phophorylation in hippocampal slices (22). Furthermore, the presence of the carboxyl-terminal isoform of FAK, called FRNK, was suggested to act as an endogenous regulator of FAK functions and formation of focal adhesions (23). In addition, a putative homologue of FAK, FakB, has been identified and shown to be differentially regulated from FAK in T cell and B cells (24). FakB was shown to participate in signaling by antigen receptors and associate with T-cell receptor-linked protein tyrosine kinase ZAP 70 in lymphocytes (24).
We and others have previously demonstrated that Pyk2 isoform present in T cells is selectively phosphorylated by Fyn, but not Lck, following engagement of T cell receptors (8,9) and that stimulation of chemokine receptors by chemokines or HIV FIG. 5. Structural changes in the carboxyl terminus of Pyk2-H does not influence its catalytic activity. A, Pyk2 isoforms were immunoprecipitated from lysates of 293 cells transfected with vector alone (pRK5), or pRK5 containing Pyk2 or Pyk2-H with anti-Pyk2 antibodies (number 600). Immunoprecipitations were split such that one-third was subjected to immunoblotting with anti-Pyk2 623 antibodies (left panel), while twothirds of the same immunoprecipitations were subjected to in vitro kinase reaction using poly(Glu-Tyr) as an exogenous substrate (right panel). Fold increase in phosphorylation of poly(Glu-Tyr) is indicated. B, autophosphorylation profile of in vitro phosphorylated Pyk2 isoforms. Pyk2 isoforms or their mutant forms, Pyk2-H-Y402F and Pyk2-Y402F, were in vitro phosphorylated and subjected to tryptic digestion. The phosphopeptides were separated by reverse-phase HPLC and 32 P content was analyzed with a ␤-counter. Radioactivity is shown in counts/min of each HPLC fraction. envelope proteins leads to rapid tyrosine phosphorylation of Pyk2 isoform in several hematopoietic cells (12). Here we demonstrate that Pyk2-H, and not Pyk2, is the main isoform that is activated by antigen or chemokine receptors in primary hematopoietic cells. We have demonstrated that several cell lines such as Jurkat or HL-60 cells express both isoforms. However, the functional differences between Pyk2 and Pyk2-H in these cells are not well understood. For example, we have observed that activation of T-cell receptors in Jurkat T-cells or chemokine receptors in HL-60 cells lead to tyrosine phosphorylation of both Pyk2 isoforms (data not shown). Since Jurkat and HL-60 cells are neoplastic cell lines and since Pyk2 and Pyk2-H are normally present in different primary cells the biological importance of co-activation of Pyk2 and Pyk2-H in Jurkat and HL-60 cells is rendered less conclusive.
In addition, a little is known about the molecular mechanisms by which Pyk2 and Pyk2-H may signal differently. The spliced exon of Pyk2 contains 42 amino acids enriched in proline, serine, and threonine residues (7 proline, 6 serine, and 3 threonine), suggesting that this region may mediate proteinprotein interactions, may be a target of proline-directed serine/ threonine kinases, or might lead to changes in the intrinsic tyrosine kinase of Pyk2. In this report we show that structural changes in the carboxyl terminus of Pyk2-H did not have a major effect on Pyk2-H kinase activity toward exogenous substrate or toward autophosphorylated tyrosine 402. However, the carboxyl termini of Pyk2 and Pyk2-H may mediate both common and differential interactions with cellular proteins. GST fusion proteins containing the carboxyl termini of Pyk2-H or Pyk2 bind several common tyrosine-phosphorylated proteins in lysates of the thymus (Fig. 6B). An identified protein that binds to GST fusion proteins containing the carboxyl termini of Pyk2 and Pyk2-H, but not to the GST fusion protein containing the region encoded by the exon that is spliced out in Pyk2-H is the protein paxillin. It has been documented that Pyk2 constitutively associates with paxillin in several hematopoietic cell lines and our results show that the splicing event in the proline-rich domain of Pyk2-H did not lead to structural changes that could affect binding of Pyk2-H to paxillin (13,25). In addition, we have observed a p115 tyrosine-phosphorylated protein that preferentially binds to GST fusion protein contain-ing the carboxyl terminus of Pyk2 and not Pyk2-H (Fig. 6B). A weak co-precipitation of p130 Cas with Pyk2 was shown in B cell lines (11) and Pyk2 was shown to phosphorylate p130 Cas and p105 Hef1 (26). However, we have not been able to show coprecipitation of the carboxyl termini of Pyk2 or Pyk2-H with p105 Hef-1 or p130 Cas in thymus lysates (Fig. 6B). Although the identity and function of p115 tyrosine-phosphorylated protein that is associated with the carboxyl terminus of Pyk2 is currently unknown, the identification of the proteins that associate specifically with the residues of the spliced exon of Pyk2 or that bind specifically to the shorter carboxyl terminus of Pyk2-H will shed light on the role this region plays in the control of Pyk2 functions.