Interaction of Hematopoietic Progenitor Kinase 1 and c-Abl Tyrosine Kinase in Response to Genotoxic Stress*

The c-Abl protein tyrosine kinase is activated by certain DNA-damaging agents and regulates induction of the stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK). The hematopoietic progenitor kinase 1 (HPK1) has also been shown to act upstream to the SAPK/JNK signaling pathway. We report here that exposure of hematopoietic Jurkat T cells to genotoxic agents is associated with activation of HPK1. The results demonstrate that exposure of Jurkat cells to DNA-damaging agents is associated with translocation of active c-Abl from nuclei to cytoplasm and binding of c-Abl to HPK1. Our findings also demonstrate that c-Abl phosphorylates HPK1 in cytoplasm and stimulates HPK1 activity. The functional significance of the c-Abl-HPK1 interaction is supported by the demonstration that this complex regulates SAPK/JNK activation. Overexpression of c-Abl(K-R) inhibits HPK1-induced activation of SAPK/JNK. Conversely, the dominant negative mutant of HPK1 blocks c-Abl-mediated induction of SAPK/JNK. These findings indicate that activation of HPK1 and formation of HPK1/c-Abl complexes are functionally important in the stress response of hematopoietic cells to genotoxic agents.

The cellular response to ionizing radiation (IR) 1 and other genotoxic agents includes cell cycle arrest, activation of DNA repair, and apoptosis or programmed cell death. However, the intracellular signals that control these events are largely unclear. The available evidence supports a role for the c-Abl protein tyrosine kinase in the induction of apoptosis (1,2).
Transient transfection studies with wild-type but not kinaseinactive c-Abl have demonstrated induction of an apoptotic response (2). Also, cells that stably express the dominant negative c-Abl(K-R) mutant exhibit resistance to induction of apoptosis by IR and other DNA-damaging agents (2,3). Similar results have been obtained in Abl Ϫ/Ϫ fibroblasts (2,3). The apoptosis-resistant phenotype is more pronounced in cells expressing c-Abl(K-R) than in c-Abl null cells. In addition, a proapoptotic role for c-Abl is supported by c-Abl-dependent induction of the stress-activated protein kinase/c-Jun N-terminal kinase (SAPK/JNK) in the response to genotoxic stress (4 -6).
HPK1, a 97-kDa serine/threonine kinase, is restricted to hematopoietic tissues in adults (23,24). Studies have shown that HPK1 interacts with MEKK-1 (23), MLK-3 (24), and TAK1 (32), which, in turn, can activate MKK4/SEK1 and thereby result in activation of the SAPK signaling pathway. Previous studies have demonstrated that four proline-rich motifs in HPK1 are potential binding sites for SH3 domain-containing proteins. HPK1 interacts with the SH2/SH3 domain-containing adaptor proteins Crk and CrkL (33). Using yeast two-hybrid analysis, HPK1 has also been shown to associate with the c-Abl SH3 domain (24). The demonstration that Abl Ϫ/Ϫ cells exhibit a defective SAPK response in response to certain DNA-damaging agents has provided support for c-Abl as an upstream effector in the SAPK pathway (5,6) and has raised the possibility of a functional interaction between c-Abl and HPK1.
The present studies demonstrate that exposure of Jurkat cells to IR is associated with activation of HPK1. Similar re-sults were obtained with another genotoxic agent, 1-␤-Darabinofuranosylcytosine (ara-C). The results also demonstrate that activated HPK1 forms a complex with cytoplasmic c-Abl in the cellular response to genotoxic agents. The functional significance of the c-Abl/HPK1 interaction is supported by the finding that HPK-1-induced activation of SAPK is inhibited by a dominant negative c-Abl and that kinase-inactive mutants of HPK1 block c-Abl-mediated induction of SAPK activity.

MATERIALS AND METHODS
Cell Culture and Reagents-Human Jurkat T cells (American Type Culture Collection, Manassas, VA) were grown in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum (Sigma), 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM L-glutamine. Human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal bovine serum and antibiotics. MCF-7/neo and MCF-7/c-Abl(K-R) (34) cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, antibiotics, and 500 g/ml Geneticin sulfate (Life Technologies, Inc.). Cells were seeded at a density of 1 ϫ 10 6 cells/100-mm culture dish for 24 h before treatment with 20 Gy of IR or 10 M ara-C (Sigma). Irradiation was performed at room temperature with a ␥-ray source (Cs 173 ; Gamma Cell 1000; Atomic Energy of Canada, Ontario, Canada) at a fixed dose rate of 0.76 Gy/min.
Isolation of the Cytosolic Fraction-Cytosolic fractions were prepared as described previously (35). Cells were washed twice with phosphatebuffered saline and then suspended in ice-cold buffer (20 mM HEPES, pH 7.5, 1.5 mM MgCl 2 , 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, and 10 g/ml leupeptin and aprotinin) containing 250 mM sucrose. The cells were disrupted by five strokes in a Dounce homogenizer. After centrifugation of the lysate at 10,000 ϫ g for 5 min at 4°C, the supernatant fraction was centrifuged at 105,000 ϫ g for 30 min at 4°C. The resulting supernatant was used as the soluble cytosolic fraction.
Isolation of the Nuclear Fraction-Nuclear proteins were isolated as described previously (36). In brief, cells were washed three times with phosphate-buffered saline and resuspended in 4 cell volumes of hypotonic lysis buffer (10 mM HEPES, pH 7.5, 2 mM MgCl 2 , 10 mM KCl, 10 g/ml leupeptin, and 10 g/ml aprotinin). After incubation on ice for 30 min to allow swelling, the cells were disrupted in a Dounce homogenizer (15-20 strokes). The homogenate was layered on a cushion of 1 M sucrose in hypotonic solution and subjected to centrifugation for 10 min. The nuclei were then suspended in lysis buffer containing 0.5% Nonidet P-40. After incubation at 4°C for 30 min, the suspension was centrifuged at 12,000 ϫ g, and the supernatant was used as the nuclear fraction.
Immunoprecipitation and Immunoblot Analysis-Total cell lysates were prepared as described in lysis buffer containing 1% Nonidet P-40 (37). Equal amounts of total, cytosolic, or nuclear proteins were subjected to immunoprecipitation with anti-c-Abl (K-12; Santa Cruz Biotechnology) or anti-HPK1 (32). Immune complexes were recovered by incubation with protein A-Sepharose for 1 h at 4°C, washed three times with lysis buffer, separated by SDS-PAGE, and then transferred to nitrocellulose filters. After blocking with 5% dried milk in phosphatebuffered saline-Tween, the filters were incubated with anti-HPK1, antic-Abl (Ab-3; Oncogene Research Products), anti-Flag M2 (Sigma), anti-P-Tyr (4G10; Upstate Biotechnology), anti-Lamin A (Santa Cruz Biotechnology), anti-␤-actin (Santa Cruz Biotechnology), or anti-SAPK (Santa Cruz Biotechnology). The filters were analyzed by ECL (Amersham Pharmacia Biotech).
Fusion Protein Binding Assays and Peptide Competition Assays-GST and GST-Abl SH3 (39) were purified by affinity chromatography using glutathione-Sepharose beads and equilibrated in lysis buffer. Cell lysates were incubated with 5 g of immobilized GST or GST-c-Abl SH3 for 2 h at 4°C. The resulting protein complexes were washed three times with lysis buffer and boiled for 5 min in SDS sample buffer. The complexes were then separated by SDS-PAGE and subjected to immunoblot analysis with anti-HPK1. GST-c-Abl SH3 fusion protein was incubated with PR2 (33), PR3, or PR4. The fusion protein-peptide mixtures were incubated separately with cell lysates for 30 min at room temperature. After washing, bound proteins were analyzed by immunoblotting.
c-Abl and HPK1 Kinase Assays-Cell lysates were subjected to immunoprecipitation with anti-HPK1 or anti-c-Abl as described previously (35). The protein complexes were washed and incubated in kinase buffer (20 mM HEPES, pH 7.4, and 10 mM MgCl 2 ) containing 2.5 Ci of [␥-32 P]ATP and either GST-Crk(120 -225) (40) or myelin basic protein (MBP; Sigma) as substrates for 15 min at 30°C. The reaction products were analyzed by SDS-PAGE and autoradiography.
c-Jun Kinase Assays-293T cells were transfected with pEBG-SAPK, pEBG-SEK-1, Flag-HPK1, Flag-HPK1(M46), and c-Abl or c-Abl(K-R). After 12 h of incubation at 37°C, the medium was replaced, and the cells were incubated for another 24 h. Cell lysates were prepared as described, and 200 -250 g of soluble proteins were incubated with 5 g of immobilized GST for 30 min at 4°C. The protein complexes were washed with lysis buffer and then incubated in kinase buffer containing [␥-32 P]ATP and GST-c-Jun(2-100) (38) for 15 min at 30°C. Reactions were terminated by the addition of SDS-PAGE sample buffer and boiling. Phosphorylated proteins were resolved by SDS-PAGE and analyzed by autoradiography. Cell lysates were also subjected to immunoblotting with anti-GST (Santa Cruz Biotechnology).
Transient Transfections and Immunoprecipitations-293T cells were cotransfected by the calcium phosphate method with HA-c-Abl and Flag-HPK1. After incubation for 36 h, the cells were lysed in lysis buffer containing 1% Nonidet P-40 and then subjected to immunoprecipitation with anti-HA (Boehringer Mannheim), and the immunoprecipitates were analyzed by immunoblotting with anti-Flag. 293T cells were also transiently transfected by the calcium phosphate method with Flag-HPK1 or Flag-HPK1(M46) in the presence of c-Abl or c-Abl(K-R). After incubation for 36 h, the cells were lysed in lysis buffer containing 1% Nonidet P-40 and then subjected to HPK1 kinase assay or immunoblot analysis with anti-Flag. MCF-7/neo or MCF-7/c-Abl(K-R) cells were transiently transfected with Flag-HPK1 by LipofectAMINE (Life Technologies, Inc.). Total cell lysates were subjected to immunoprecipitation with anti-Flag and then subjected to immunoblot analysis with anti-P-Tyr. Autoradiograms were scanned by laser densitometry, and the intensity of the signals was quantitated with the ImageQuant program (Molecular Dynamics, Sunnyvale, CA).
In Vitro Phosphorylation of HPK1-Recombinant c-Abl protein was incubated with GST-HPK1-KD or GST-HPK1-CD fusion proteins in the presence of [␥-32 P]ATP for 30 min at 30°C. Phosphorylation of the reaction products was assessed by SDS-PAGE and autoradiography. 293T cells were transiently transfected with pCIneo-Flag-HPK1. Cell lysates were subjected to immunoprecipitation with anti-Flag, and the precipitates were incubated with recombinant purified c-Abl or kinaseinactive c-Abl(K-R) in the presence of [␥-32 P]ATP for 30 min at 30°C. Phosphorylation of the reaction products was assessed by SDS-PAGE and autoradiography.

Genotoxic Stress Induces the Interaction of c-Abl and HPK1
in Jurkat Cells-To assess whether c-Abl and HPK1 associate in cells, lysates from human Jurkat T cells were subjected to immunoprecipitation with anti-HPK1, and the protein precipitates were analyzed by immunoblotting with anti-c-Abl. Immunoblot analysis of precipitates using a control antibody or preimmune rabbit serum demonstrated little, if any, detection of c-Abl ( Fig. 1a; data not shown). However, a similar analysis of anti-HPK1 immunoprecipitates demonstrated the coprecipitation of HPK1 and c-Abl (Fig. 1a). To assess interactions between c-Abl and HPK1 in response to genotoxic agents, Jurkat cells were treated with 10 M ara-C and harvested at 3 h. Analysis of anti-HPK1 immunoprecipitates by immunoblotting with anti-c-Abl demonstrated induction of HPK1-c-Abl complexes (Fig. 1b). Whereas ara-C incorporates into DNA and inhibits DNA replication (41), IR induces single-and doublestrand DNA breaks. The finding that exposure of Jurkat cells to IR is also associated with increased binding of c-Abl and HPK1 indicated that this response is induced by diverse types of genotoxic stress (Fig. 1b).
To further determine the interaction of c-Abl with HPK1, we transiently overexpressed Flag-HPK1 with HA-c-Abl in 293T cells and analyzed anti-HA immunoprecipitates with anti-Flag. Reactivity of anti-Flag with a 100-kDa protein supported the coprecipitation of HPK1 with c-Abl (Fig. 2a). In the reciprocal experiment, anti-Flag immunoprecipitates were subjected to immunoblot analysis with anti-HA. The results confirmed the detection of complexes containing HPK1 and c-Abl (data not shown). Lysates from transfected 293T cells were also subjected to immunoprecipitation with anti-HA. Analysis of protein precipitates by immunoblotting with anti-HA demonstrated equal levels of c-Abl (Fig. 2a). Taken together, these findings indicate that c-Abl associates with HPK1 in cells. To assess whether interaction between c-Abl and HPK1 is induced by genotoxic stress under conditions overexpressing c-Abl and HPK1, 36 h after the transfection with Flag-HPK1 and HA-c-Abl, cells were treated with 10 M ara-C for 3 h. Analysis of anti-HA immunoprecipitates by immunoblotting with anti-Flag demonstrated a significant induction of HPK1-c-Abl complex in response to ara-C (Fig. 2b). Four proline-rich sequences (PR1-PR4) are present in the C-terminal region of HPK1 (33). One of these proline-rich sequences (SGPPPNSPRPGPPPS; aa 430 -444) displays homology with motifs located in the C-terminal domains of 3BP1, 3BP2, and ST5 that bind c-Abl SH3. To determine whether the c-Abl SH3 domain binds to HPK1, lysates from irradiated Jurkat cells were incubated with GST or GST-c-Abl SH3, and the resulting precipitates were analyzed by immunoblotting with anti-HPK1. The results demonstrate that in contrast to GST, HPK1 was detectable in the adsorbates to GST-c-Abl SH3 (Fig. 2c). Because the HPK1 proline-rich motif PR3 (but not PR1, PR2, or PR4) matches the c-Abl SH3binding consensus motif (PXXXXPXPP), we examined the ability of HPK1 proline-rich peptides to block the formation of c-Abl/HPK1 complexes. The results demonstrate that the PR3 proline-rich peptide efficiently blocks the interaction of HPK1 with c-Abl, whereas PR2 and PR4 had at best a marginal effect on c-Abl/HPK1 complex (Fig. 2d). These findings collectively indicate that the interaction between HPK1 and c-Abl likely involves c-Abl-SH3 and HPK1-Pro. It is possible that the coprecipitation of HPK1 and c-Abl is due to interactions of each kinase with other molecules.
HPK1 is localized primarily in the cytoplasm (23). To define the subcellular localization of the interaction between c-Abl and HPK1, we subjected nuclear and cytoplasmic lysates from control and ara-C-treated cells to immunoprecipitation with anti-HPK1. The immunoprecipitates were then analyzed by immunoblotting with anti-c-Abl. Signal intensities from the anti-c-Abl immunoblotting experiments (n ϭ 3) were analyzed by densitometric scanning. Immunoblot analysis of the immunoprecipitates from control and ara-C-treated cells demonstrated little if any reactivity with anti-c-Abl in the nuclear fraction (Fig. 3a). Formation of HPK1/c-Abl complexes was significantly increased in the cytoplasm but not in the nucleus of ara-C-treated cells (Fig. 3a). Studies have shown that although c-Abl contains three nuclear localization signals, it is not localized exclusively to the nucleus (42,43). c-Abl contains a functional nuclear export signal, and the subcellular localization of c-Abl is determined by a balance of nuclear import and export (44). To assess whether c-Abl translocates to the cytoplasm in response to genotoxic stress, Jurkat cells were treated with ara-C for different intervals of time. Nuclear and cytoplasmic fractions were isolated and analyzed by immunoblotting with anti-c-Abl. As controls, nuclear and cytoplasmic fractions were also analyzed by immunoblotting with anti-Lamin A and anti-␤-actin, respectively. The results demonstrate that treatment of Jurkat cells with ara-C was associated with significant decreases in nuclear c-Abl levels (Fig. 3b, left panel). Moreover, levels of cytoplasmic c-Abl were increased in response to ara-C (Fig. 3b, right panel). Translocation of c-Abl from nucleus in the response to genotoxic stress may initiate formation of com- FIG. 1. c-Abl associates with HPK1. a, total cell lysates from Jurkat cells were subjected to immunoprecipitation with anti-c-Abl, anti-HPK1, or preimmune rabbit serum (PIRS). The protein precipitates were separated by SDS-PAGE and transferred to nitrocellulose filters. The filters were analyzed by immunoblotting with anti-c-Abl antibody. b, Jurkat cells were treated with 10 M ara-C or exposed to 20 Gy of IR and harvested after 3 h. Cell lysates (approximately 150 g of total protein) were subjected to immunoprecipitation with anti-HPK1 and analyzed by immunoblotting with anti-c-Abl (top panel). As a control, anti-HPK1 immunoprecipitates were analyzed by immunoblotting with anti-HPK1 (bottom panel). Total cell lysates (10 g of total protein; ϩve) were also analyzed by immunoblotting with anti-c-Abl or anti-HPK1.
FIG. 2. c-Abl associates with HPK1 via its SH3 domain. a, 293T cells were transiently transfected with Flag-HPK1 in the presence of HA-c-Abl or empty vector. Lysates were subjected to immunoprecipitation with anti-HA, and the precipitates were analyzed by immunoblotting with anti-Flag (top panel). As a positive control (ϩve), total cell lysate was also analyzed by immunoblotting with anti-Flag. As controls, anti-HA and anti-Flag immunoprecipitates were analyzed by immunoblotting with anti-HA (middle panel) and anti-Flag (bottom panel), respectively. IgH, immunoglobulin heavy chain. b, 293T cells were transiently transfected with Flag-HPK1 and HA-c-Abl. Thirty-six h after the transfection, cells were treated with 10 M ara-C for 3 h. Total cell lysates were subjected to immunoprecipitation with anti-HA. The precipitates and lysate were analyzed by immunoblotting with anti-Flag (top panel) or anti-HA (middle panel). As a control, anti-Flag immunoprecipitates were also analyzed by immunoblotting with anti-Flag (bottom panel). c, Jurkat cell lysates (150 g of total protein) were incubated with GST or GST-c-Abl SH3 fusion proteins. The protein adsorbates and lysate (10 g of total protein; ϩve) were analyzed by immunoblotting with anti-HPK1. d, 293T cells were transiently transfected with Flag-HPK1, and total lysates were divided in three equal portions. GST-c-Abl SH3 fusion protein was incubated with PR2 (32), PR3, or PR4. The fusion protein-peptide mixtures were incubated separately with lysates for 1 h at 4°C. After washing, bound proteins were analyzed by immunoblotting with anti-Flag.
plexes with multiple molecules in cytoplasm. Indeed, densitometric scanning of autoradiograms and quantitative analysis demonstrate that the formation of c-Abl complexes with HPK1 in cytoplasm is significantly less than the translocation of c-Abl from the nucleus to the cytoplasm. These findings support a model in which c-Abl is activated in the nucleus in response to genotoxic stress, translocates to the cytoplasm, and thereby interacts with HPK1.
To determine whether the kinase function of HPK1 is necessary for the interaction with c-Abl, we transiently cotransfected Flag-HPK1 or a kinase-inactive Flag-HPK1(M46) mutant with c-Abl in 293T cells. After treatment with ara-C, anti-c-Abl immunoprecipitates were analyzed by immunoblotting with anti-Flag. The results demonstrate that in contrast to HPK1(M46), overexpression of wild-type HPK1 is associated with an increase in binding with c-Abl (Fig. 3c). Because c-Abl is also activated by ara-C and IR (11), we asked whether the association of HPK1 with c-Abl is also dependent on the c-Abl kinase function. To address this issue, 293T cells were transiently transfected with Flag-HPK1 and c-Abl or dominant negative c-Abl(K-R) mutant and then treated with ara-C. Antic-Abl immunoprecipitates were analyzed by immunoblotting with anti-Flag. The results demonstrate that the interaction between HPK1 and c-Abl is significantly increased in cells overexpressing wild-type c-Abl (Fig. 3d). Taken together, these findings suggest that the kinase function of c-Abl and HPK1 may be necessary for their interaction. However, our data do not rule out the possibility that the loss of interaction between c-Abl and HPK1 might also be due to improper folding of these mutants. Additional studies using purified recombinant c-Abl and HPK1 proteins are required to delineate this issue.
To assess in part the functional significance of the interaction of c-Abl and HPK1, we incubated purified recombinant c-Abl with GST-HPK1-KD (HPK1 kinase domain; aa 1-291) or GST-HPK1-CD (C-terminal domain; aa 292-833) (Fig. 4a) fusion proteins in the presence of [␥-32 P]ATP. Analysis of the reaction products demonstrated little, if any, phosphorylation of either GST-HPK1-KD or GST-HPK1-CD (Fig. 4b, left panel). The unavailability of full-length GST-HPK1 protein precluded us from using it as a substrate. The potential HPK1 binding sequence for the c-Abl SH3 domain (SGPPPNSPRPGPPPS; aa 430 -444) is present in the HPK1-CD, whereas the c-Abl phosphorylation site (YXXP) (39) in HPK1 (Y 232 QPP; aa 232-235) is localized in HPK1-KD. To determine whether full-length HPK1 acts as a substrate for c-Abl, 293T cells were transiently transfected with full-length Flag-HPK1. Cell lysates were subjected to immunoprecipitation with anti-Flag, and the precipitates were incubated with recombinant c-Abl in the presence of [␥-32 P]ATP. As control, anti-Flag immunoprecipitates were incubated separately with recombinant c-Abl(K-R) protein in the presence of [␥-32 P]ATP. Analysis of the reaction products demonstrated c-Abl-mediated phosphorylation of full-length HPK1 (Fig. 4b, right panel). As a control, anti-Flag immunoprecipitates were analyzed separately by immunoblotting with anti-Flag. The results demonstrate equal expression of Flag-HPK1 in multiple transfections (data not shown). To confirm c-Ablmediated tyrosine phosphorylation of HPK1, 293T cells were cotransfected with Flag-HPK1 with increasing amounts of c-Abl. Lysates were subjected to immunoprecipitation with anti-Flag and analyzed by immunoblotting with anti-P-Tyr. Signal intensities from anti-P-Tyr immunoblotting experiments were analyzed by densitometric scanning. The results from quantitation of signal intensities demonstrate 3.2 Ϯ 0.4-fold induction in tyrosine phosphorylation of HPK1 by c-Abl (Fig. 4c).
Because c-Abl phosphorylates HPK1 in vitro, we asked whether HPK1 is tyrosine-phosphorylated in the cellular response to genotoxic stress. Jurkat cells were treated with ara-C or exposed to IR. Total cell lysates were subjected to immunoprecipitation with anti-HPK1, and the precipitates were analyzed by immunoblotting with anti-P-Tyr. As a control, anti-HPK1 immunoprecipitates were also analyzed by immunoblotting with anti-HPK1. The results demonstrate a 2-3-fold induction in tyrosine phosphorylation of HPK1 in response to genotoxic agents ( Fig. 5a; data not shown). Exposure of cells to genotoxic agents is associated with activation of c-Abl (4 -6). We therefore investigated whether genotoxic stress affects c-Abl-mediated tyrosine phosphorylation of HPK1. MCF-7 cells expressing neo cassette (MCF-7/neo) and MCF-7 cells expressing c-Abl(K-R) (MCF-7/c-Abl(K-R)) were transiently transfected with Flag-HPK1. After transfection, cells were exposed to IR or treated with ara-C and harvested after 3 h. Cytoplasmic lysates were subjected to immunoprecipitation with anti-Flag antibody, and the protein precipitates were analyzed by immunoblotting with anti-P-Tyr. As a control, anti-Flag immunoprecipitates were also analyzed by immunoblotting with anti-Flag. Exposure of MCF-7/neo cells to IR was associated with increases (3 Ϯ 0.5-fold; Fig. 5b, bottom panel) in tyrosine phosphorylation of HPK1 (Fig. 5b). Moreover, IR had no detectable effect on tyrosine phosphorylation of HPK1 in MCF-7/c-Abl(K-R) cells (Fig. 5b). Similar results were obtained when MCF-7/neo or MCF-7/c-Abl(K-R) cells were treated with ara-C (Fig. 5b). Because MCF-7/c-Abl(K-R) cells stably overexpress c-Abl(K-R), and because of the very minimal basal activity of c-Abl(K-R), the background level of tyrosine phosphorylation of c-Abl(K-R) is high compared with the endogenous levels seen in MCF-7/neo. Collectively, these findings demonstrate that genotoxic stress induces tyrosine phosphorylation of HPK1 by a c-Abl-dependent mechanism.
c-Abl Activates HPK1 in Vitro and in the Response to Genotoxic Stress-To further assess the functional significance of the interaction between c-Abl and HPK1, we investigated whether c-Abl affects HPK1 activity. 293T cells were cotransfected with Flag-HPK1 and empty vector or c-Abl. Anti-Flag immunoprecipitates were assayed for HPK1 kinase activity using MBP as a substrate. Analysis of the reaction products by autoradiography demonstrated that overexpression of c-Abl is associated with an increase (ϳ3-fold) in the kinase activity of HPK1 (Fig. 6a). Because HPK1 is a serine/threonine kinase, we next assessed whether HPK1 activates c-Abl. To address this issue, 293T cells were transfected with HA-c-Abl and HPK1 or empty vector, and anti-HA immunoprecipitates were assayed for phosphorylation of GST-Crk(120 -225) (40). The results demonstrate that coexpression of HPK1 and c-Abl had no detectable effect on c-Abl kinase activity (Fig. 6b).
To determine whether HPK1 is activated in the response to genotoxic stress, lysates from Jurkat cells treated with ara-C were subjected to immunoprecipitation with anti-HPK1. The immunoprecipitates were assayed for phosphorylation of MBP. The results demonstrate an increase (5 Ϯ 1.1-fold) in phosphorylation of MBP by HPK1 in ara-C-treated cells as compared with control cells (Fig. 7a). Similar results were obtained when Jurkat cells were exposed to IR (Fig. 7b). To define the subcellular localization of HPK1 activation, we assayed nuclear and cytoplasmic lysates from control and ara-C-treated cells for HPK1 activity. The results demonstrate increased activation of HPK1 in cytoplasmic but not nuclear lysates of the ara-Ctreated cells (Fig. 7c). Similar results were obtained when Jurkat cells were exposed to IR (data not shown). We next determined the role of c-Abl in the regulation of HPK1 activity in response to genotoxic agents. Studies have shown that HPK1 is expressed in Jurkat cells and other cell types that are predominantly hematopoietic. Due to the extremely low transfection efficiency of Jurkat cells, we transiently transfected 293T cells with Flag-HPK1 and empty vector, c-Abl or c-Abl(K-R). After transfections, cells were treated with 10 M ara-C and harvested after 3 h. Anti-Flag immunoprecipitates were analyzed for phosphorylation of MBP. As controls, anti-Flag immunoprecipitates and total cell lysates were analyzed separately by immunoblotting with anti-Flag and anti-c-Abl, respectively. The results demonstrate increased phosphorylation of MBP in cells overexpressing wild-type c-Abl but not Abl(K-R) (Fig. 7d). Moreover, to demonstrate whether the kinase function of HPK1 is necessary for HPK1 activation in response to genotoxic stress, we transiently overexpressed empty vector, Flag-HPK1, or Flag-HPK1(M46) and treated with ara-C. Anti-Flag immunoprecipitates were analyzed for phosphorylation of MBP. The results demonstrate increased phosphorylation of MBP in cells overexpressing wild-type HPK1 but not HPK1(M46) (Fig. 7e). To further determine activation of HPK1 in response to IR, we transiently overexpressed Flag-HPK1 in 293T cells and then exposed cells to IR. Anti-Flag immunoprecipitates were analyzed for phosphorylation of MBP. The results demonstrate that exposure of cells to IR is associated with an increase in activation of HPK1 (Fig. 7f).
c-Abl and HPK1 Synergistically Activate SAPK/JNK-Treatment of cells with diverse genotoxic agents activates SAPK (5,(7)(8)(9)(10)(11)(12)(13)(14)(15). We next asked whether SAPK is activated in Jurkat cells in the response to genotoxic stress. To assess SAPK activation, Jurkat cells were either treated with 10 M ara-C or exposed to 20 Gy of IR and harvested after different times. Total cell lysates were subjected to immunoprecipitation with anti-SAPK and assayed for phosphorylation of GST-c-Jun. In concert with previous findings, treatment of Jurkat cells with ara-C or IR was also associated with activation of SAPK (Fig. 8, a and b). Studies have shown that both HPK1 and c-Abl are upstream activators of SAPK (5,6,23,24,33,(45)(46)(47). To assess whether c-Abl and HPK1 cooperate in the activation of SAPK, 293T cells were transiently cotransfected with pEBG-SAPK and SEK1 in the presence and absence of HPK1 and/or c-Abl. Lysates were incubated with glutathione beads, and the precipitates were assayed for GST-c-Jun phosphorylation. Transient expression of c-Abl induced a 4 -5-fold increase in FIG. 6. c-Abl activates HPK1. a, 293T cells were transfected with Flag-HPK1 (1 g) in the presence of different concentrations (0, 1, 2, and 5 g) of c-Abl or empty vector. Total cell lysates were subjected to immunoprecipitation with anti-Flag and assayed for phosphorylation of MBP in the presence of [␥-32 P]ATP. The reaction products were separated by SDS-PAGE and analyzed by autoradiography (top panel). The fold increase in MBP phosphorylation is described as the mean of three independent experiments. As a control, total lysates were also analyzed by immunoblotting with anti-Flag (middle panel) or anti-c-Abl (bottom panel). b, 293T cells were transiently cotransfected with HA-c-Abl (1 g) in the presence of increasing concentrations (0, 1, 2, and 5 g) of HPK1 or empty vector. After transfection, anti-HA immunoprecipitates were assayed for phosphorylation of GST-Crk(120 -225) in the presence of [␥-32 P]ATP. The reaction products were separated by SDS-PAGE and analyzed by autoradiography (top panel). As a control, total lysates were also analyzed by immunoblotting with anti-Flag (middle panel) or anti-HA (bottom panel).

FIG. 7.
Genotoxic agents induced activation of HPK1. a, Jurkat cells were treated with 10 M ara-C and harvested at 3 h. Total cell lysates were subjected to immunoprecipitation with anti-HPK1. The immunoprecipitates were analyzed in immune complex kinase assays using MBP as a substrate (top panel). As a control, total cell lysates were analyzed by immunoblotting with anti-HPK1 (bottom panel). The fold increase in MBP phosphorylation is described as the mean of two independent experiments. b, Jurkat cells were exposed to 20 Gy of IR and harvested after 3 h. Total cell lysates were subjected to immunoprecipitation with anti-HPK1 and assayed as described above. c, Jurkat cells were treated with 10 M ara-C and harvested after 3 h. Nuclear (left panels) and cytoplasmic (right panels) lysates were subjected to immunoprecipitation with anti-HPK1. The immunoprecipitates were analyzed in immune complex kinase assay using MBP as a substrate (top panels). As a control, total cell lysates were analyzed by immunoblotting with anti-HPK1 (bottom panels). d, 293T cells were transiently transfected with empty vector, c-Abl, or c-Abl(K-R) in the presence of Flag-HPK1. After treatment of cells with ara-C for 3 h, anti-Flag immunoprecipitates were assayed for phosphorylation of MBP (top panel). As a control, anti-Flag immunoprecipitates were analyzed by immunoblotting with anti-Flag (middle panel). Total cell lysates were also analyzed by immunoblotting with anti-c-Abl (bottom panel). e, 293T cells were transiently transfected with empty vector, Flag-HPK1, or Flag-HPK1(M46). After treatment of cells with ara-C, anti-Flag immunoprecipitates were assayed for phosphorylation of MBP (top panel). As a control, total cell lysates were analyzed by immunoblotting with anti-Flag (bottom panel). f, 293T cells were transiently transfected with Flag-HPK1. After exposure of cells to IR for different time intervals, anti-Flag immunoprecipitates were assayed for phosphorylation of MBP (top panel). As a control, total cell lysates were analyzed separately by immunoblotting with anti-Flag (bottom panel).
SAPK activity as compared with empty vector. Overexpression of HPK1 with SEK1 and SAPK was associated with 13-15-fold activation of SAPK. Moreover, HPK1 and c-Abl together induced SAPK activity 18 -20-fold over basal level (Fig. 8c). These findings suggest that c-Abl interacts with HPK1 to activate SAPK. To determine whether a dominant negative mutant of c-Abl (c-Abl(K-R)) affects HPK1-induced SAPK activation, 293T cells were transiently transfected with HPK1 and c-Abl(K-R). Cells were also cotransfected with SEK1 and pEBG-SAPK. Lysates were subjected to protein precipitation with glutathione beads, and the precipitates were assayed for phosphorylation of GST-c-Jun. The results demonstrate that c-Abl(K-R) significantly inhibits HPK1-induced SAPK activation (Fig. 8c). Conversely, overexpression of HPK1(M46) also inhibited c-Abl-induced activation of SAPK (Fig. 8c). Taken together, these findings indicated that c-Abl and HPK1 synergize for activation of SAPK.
Studies have shown that MLK-3 acts as a substrate for HPK1 and that HPK1-induced SAPK activation is inhibited by overexpression of a dominant negative mutant of MLK-3 (24). Our recent studies have shown that c-Abl phosphorylates and activates MEKK-1 in nuclei (48). Moreover, c-Abl-induced activation of SAPK is inhibited by overexpression of a dominant negative mutant of MEKK-1 (48). MEKK-1 stimulates SEK1/ MKK4, which in turn activates SAPK (8,18,49). The finding that MEKK-1(K-R) fails to completely block c-Abl-induced SAPK activation further indicates that c-Abl also stimulates the SAPK/JNK pathway by MEKK-1-independent mechanisms (48). To determine whether c-Abl and HPK1 function upstream to MLK-3, 293T cells were transiently cotransfected with c-Abl or HPK1 and MLK-3(K-R). Cells were also cotransfected with SEK1 and pEBG-SAPK. Total cell lysates were subjected to protein precipitation with glutathione beads and assayed for GST-c-Jun phosphorylation. The results demonstrate that transfection of MLK-3(K-R) also inhibits c-Abl-induced activation of SAPK (data not shown). The results further demonstrate that overexpression of MLK-3(K-R) blocks HPK-1-induced activation of SAPK (data not shown). Taken together, these findings indicate that the c-Abl/HPK1 complex functions upstream to MLK-3 and induces SAPK activation in the response to genotoxic stress, at least in Jurkat cells. DISCUSSION Eukaryotic cells respond to DNA damage with cell cycle arrest, activation of DNA repair, and, in the event of irreparable damage, the induction of apoptosis. The signals that determine cell fate, that is, repair of DNA damage and survival or activation of cell death mechanisms, remain unclear. The c-Abl tyrosine kinase is activated in the cellular response to certain DNA-damaging agents (5,6,14,47,50,51). Previous studies have also demonstrated that c-Abl functions upstream to activation of the SAPK/JNK pathway in the response of cells to genotoxic stress (5,6,14,47,51). The exposure of diverse types of mammalian cells to genotoxic agents is associated with SAPK activation (5-7, 11, 48, 52-54). Other studies have demonstrated that activation of SAPK in the DNA damage response is associated with the induction of apoptosis (7,17,52,53). Whereas c-Abl has also been linked to DNA damage-induced apoptosis, the precise role for c-Abl as an upstream effector of the SAPK pathway has been controversial. In this context, other work has indicated that c-Abl is not required for the activation of SAPK by genotoxic agents (51). The discrepancy between findings may be related to the demonstration that c-Abl is necessary for activation of SAPK in proliferating cells but not growth-arrested cells (1). As further support for c-Abl involvement, recent work has shown that c-Abl directly activates MEKK-1, an upstream effector in the SEK13 SAPK cascade, in the DNA damage response (48). The present findings extend the role of c-Abl in the activation of SAPK signaling by demonstrating that c-Abl interacts with HPK1 in transducing signals to SEK1 and SAPK.
HPK1 is a mammalian Ste20/PAK-like serine/threonine kinase that is primarily found in hematopoietic cells (23,24). Whereas little is known about the signals responsible for activation of HPK1, studies have shown that HPK1 activity is induced in cells treated with transforming growth factor ␤ (55). Notably, HPK1 functions as an upstream effector of transforming growth factor ␤-induced activation of SAPK (55). Other studies have demonstrated that HPK1 is phosphorylated by the epidermal growth factor receptor (56). Epidermal growth factor stimulation induces the binding of HPK1 with the Grb2 adaptor protein and recruitment of these complexes to the autophosphorylated epidermal growth factor receptor (56). HPK1 also associates with the Crk and CrkL adaptor proteins in signaling that results in activation of the SAPK pathway (33). The present results demonstrate that HPK1 is activated in the response of cells to DNA-damaging agents. In this context, IR exposure is associated with the accumulation of singleand double-strand DNA breaks (57). By contrast, ara-C is a nucleoside analog that incorporates into DNA and causes arrest of DNA replication by functioning as a relative chain FIG. 8. c-Abl and HPK1 cooperate to synergistically activate JNK. a and b, Jurkat cells were either treated with 10 M ara-C (a) or exposed to 20 Gy of IR (b) and harvested after the indicated times. Total cell lysates were subjected to immunoprecipitation with anti-SAPK. The immunoprecipitates were analyzed in immune complex kinase assay using GST-c-Jun as a substrate (top panels). As control, total cell lysates were analyzed by immunoblotting with anti-SAPK (bottom panels). c, 293T cells were transiently transfected in duplicate with the indicated cDNAs. Total cell lysates were subjected to protein precipitation with glutathione beads. The adsorbates were assayed in immune complex kinase assay using GST-c-Jun as a substrate (top panel). As a control, total lysates were analyzed separately by immunoblotting with anti-GST (middle panel). The bottom panel depicts the fold increase in GST-c-Jun phosphorylation expressed as the mean Ϯ S.D of two independent experiments performed in duplicate.
terminator (41, 58 -60). The finding that both IR and ara-C activate HPK1 indicates that this response is induced by diverse types of DNA damage. In addition, the findings that IR and ara-C induce the activation of SAPK (5-7, 11, 48, 52-54) suggest that HPK1 could contribute to SAPK-mediated signals induced by these genotoxic agents. Indeed, expression of a kinase-inactive HPK1 mutant partially abrogated IR-and ara-C-induced SAPK activation.
The present findings provide further support for an interaction between c-Abl and HPK1 in the DNA damage response. The results demonstrate that genotoxic stress induces the association of c-Abl and HPK1. In vitro studies indicate that the c-Abl SH3 domain interacts directly with a proline-rich motif in the HPK1 C-terminal region. Moreover, studies in cells cotransfected to express c-Abl and HPK1 demonstrate that c-Abl phosphorylates HPK1. The finding that HPK1 is phosphorylated on tyrosine in IR-or ara-C-treated cells expressing wildtype c-Abl, but not in cells expressing c-Abl(K-R), further indicates that HPK1 is phosphorylated by a c-Abl-dependent mechanism in the DNA damage response. The functional significance of the c-Abl-HPK1 interaction is supported by the finding that c-Abl activates HPK1. Conversely, the results indicate that HPK1 has no apparent effect on c-Abl activity. These findings support a model in which HPK1 is a downstream effector of the c-Abl response to genotoxic stress. The recent demonstration that c-Abl is also activated in cells exposed to hydrogen peroxide has supported a role for c-Abl in the response to diverse types of stress (61). The available evidence, however, indicates that the interaction between c-Abl and HPK1 is induced by genotoxic stress and not by oxidative stress (data not shown).
Previous work has shown that nuclear c-Abl functions as an upstream effector of the MEKK13 SEK13 SAPK pathway in the response of cells to DNA damage in nonhematopoietic cells (48). However, the incomplete abrogation of DNA damageinduced SAPK activation by the kinase-inactive MEKK1(K-R) mutant indicates that c-Abl can also stimulate the SAPK pathway by a MEKK1-independent mechanism (48). HPK1 is predominantly a cytoplasmic kinase, and DNA damage-induced formation of c-Abl/HPK1 complexes is detectable in the cytoplasm and not in the nucleus. In concert with these findings, the results demonstrate that in response to DNA damage, nuclear c-Abl translocates to the cytoplasm. Therefore, whereas nuclear c-Abl interacts with MEKK-1 (48), after translocation to the cytoplasm, c-Abl associates with HPK1. Coexpression of c-Abl and HPK1 was associated with a synergistic effect on SAPK activation, and that this activation of SAPK is sensitive to MLK-3. Moreover, expression of either c-Abl(K-R) or HPK1(M46) blocked DNA damage-induced SAPK activation. Thus, the kinase functions of both cytoplasmic c-Abl and HPK1 are required in the induction of SAPK activity, at least in hematopoietic cells. These findings support a model in which c-Abl mediates activation of HPK1 and confers an additional signal that is also necessary for SAPK activation.