Phosphorylation at Tyr-379 within HPK1 is essential for its interaction with BASH. (A) Tyr-379 within HPK1 is a Syk-phosphorylation site. pcDNA-HPK1:HA or pcDNA-HPK1(Y379F):HA (1.2 μg each) was cotransfected transiently with pME-Lyn, pME-Syk, or empty vector (mock) (2.4 μg each) into COS7 cells. HPK1 proteins were immunoprecipitated with anti-HA Ab and immunoblotted with anti-pY (top), then reprobed with anti-HPK1 #7 Ab (upper middle). Whole cell lysates were immunoblotted with anti-pY Ab (lower middle) and anti-Syk plus anti-Lyn Abs (bottom) to visualize expression of the PTKs. (B) Tyr-379 is essential for binding HPK1 to BASH. HPK1(WT) or HPK1(Y379F) was coexpressed transiently with T7-BASH and either Lyn or Syk in COS7 cells. BASH was immunoprecipitated, and coprecipitated HPK1 proteins were analyzed as described in Fig. 2 C.

BCR-mediated tyrosine-phosphorylation of HPK1 and its effect on interaction with BASH. (A) HPK1 is phosphorylated on tyrosine after BCR engagement. WEHI231 cells (10⁷) were stimulated with anti–IgM F(ab′)₂ fragment for the indicated periods of time (min) and HPK1 was immunoprecipitated with anti-HPK1 #2 Ab. The precipitates were immunoblotted with anti-pY Ab (top) and reprobed with anti-HPK1 #7 Ab (bottom). (B) BCR-mediated phosphorylation of HPK1 in PTK- or BASH-deficient DT40 cell lines. 1.5 × 10⁷ WT or mutant DT40 cells were transiently transfected with 75 μg pCAGGS-HPK1:HA. After 36 h, the cells were stimulated with anti–chicken IgM Ab for the indicated periods of time (min). HA-tagged HPK1 proteins were immunoprecipitated with anti-HA Ab and immunoblotted with anti-pY Ab (top), then reprobed with anti-HA Ab (bottom). (C) Tyrosine-phosphorylation of HPK1 is required for the association of HPK1 with BASH. pMT2-HPK1 (2 μg) and either (1 μg) of pAT7-BASH, pAT7-BASH(R373K), or empty vector pAT7 (mock) were cotransfected transiently with either (2 μg) of pME-Lyn, pME-Syk, or empty vector pME (mock) into COS7 cells. The cells were lysed and BASH proteins were immunoprecipitated using anti-T7 Ab. One-half of the precipitate was immunoblotted using anti-HPK1 #7 (top), while the other half was immunoblotted with anti-pY Ab (left, upper middle), then reprobed with anti-T7 Ab (left, lower middle; right, middle). Equal HPK1 expression levels were visualized by immunoblotting of the cell lysates with anti-HPK1 #7 Ab (bottom).

BCR-induced activation of HPK1 is largely dependent on its interaction with BASH. (A) HPK1 kinase is activated in response to BCR stimulation. WEHI231 cells (5 × 10⁶) were stimulated with anti–IgM F(ab′)₂ fragment for the indicated periods of time (min). Cells were lysed, HPK1 was immunoprecipitated using anti-HPK1 #2 Ab, and its in vitro kinase activity was determined. Reaction products were separated by SDS-PAGE and visualized by autoradiography. HPK1-autophosphorylation (top) and its kinase activity toward the exogenous substrate MBP (middle) are shown. Precipitated HPK1 proteins were visualized by immunoblotting with anti-HPK1 #7 Ab (bottom). (B) Activation of HPK1 in WT and mutant DT40 cells. WT or various mutant DT40 cell lines (1.5 × 10⁷ cells each) were transfected with pCAGGS-HPK1:HA (75 μg) and stimulated with anti–chicken IgM Ab. HA-tagged HPK1 proteins were immunoprecipitated using anti-HA Ab and their in vitro kinase activities toward MBP were determined as described in A (top). Precipitated HPK1 was visualized by immunoblotting with anti-HA Ab (bottom). (C) The BASH SH2 domain mediates BCR-induced HPK1 activation. WT and BASH-deficient (BASH⁻) DT40 cells were transiently transfected with pCAGGS-HPK1:HA (30 μg) in combination with 60 μg of either T7-BASH, T7-BASH (R373K), or empty vector (mock). Transfected cells were either left unstimulated (−) or stimulated with anti–chicken IgM Ab for 3 min (+) and lysed. HPK1 was immunoprecipitated using anti-HA Ab and subjected to an in vitro kinase assay (top) and immunoblotting (middle) as described in B. BASH protein expression levels were visualized by immunoblotting of the lysates with anti-T7 Ab (bottom). (D) Tyr-379 is critical for activation of HPK1 by BCR stimulation. DT40 cells were transfected transiently with pcDNA3-HPK1:HA or pcDNA3-HPK1(Y379F):HA vector (75 μg each). Cells were stimulated and HPK1 kinase activity was assayed as in B. (E) Spontaneous kinase activity of ectopically expressed HPK1(Y379F) is comparable to WT HPK1. pcDNA3-HPK1:HA, pcDNA3-HPK1(Y379F):HA, or pcDNA3-HPK1(K46E):HA DNA (3 μg each) was transfected into 293T cells and kinase activity of HPK1 was assayed as described in B.

HPK1 positively regulates BCR-induced IKKβ activation. Expression vector (40 μg) for HA-epitope tagged ERK2 (A) or Flag-epitope tagged IKKβ (B) and either 40 μg of pcDNA-HPK1:HA (HPK[WT]), pcDNA-HPK1(K46E):HA [HPK(K46E)], pCAT7-HPK1-ΔN (HPK1-ΔN) 41, or empty vector (mock) were cotransfected into DT40 cells (10⁷). After stimulation with anti–IgM Ab for the indicated periods of time (min), exogenous ERK2 (A) or IKKβ (B) were immunoprecipitated with antibodies directed against their corresponding tags and their catalytic activities were measured in vitro using the specific substrate proteins, GST-Elk1 (A) or GST-IκBα (B), respectively (top). The amount of the precipitated kinases was estimated by immunoblotting with anti-HA Ab (A, middle), anti-IKKβ Ab (B, middle, left), or anti-Flag Ab (B, middle, right). Expression levels of HPK1 proteins were visualized by immunoblotting with anti-HA Ab (bottom panels in A and B, left) or anti-HPK1 #7 Ab (B, bottom, right). In A, HA-epitope tagged HPK1 was also immunoprecipitated with ERK2. However, we confirmed that HPK1 is unable to phosphorylate GST-Elk1 as assessed in a separate in vitro kinase assay (data not shown). (C) Nuclear extracts were prepared from anti-IgM–stimulated (1 or 2 h) or unstimulated (0) cells of WEHI231, a WEHI231-stable transfectant highly expressing T7-tagged BASH (B7), B7-stable transfectants highly expressing either WT HPK1 (WT4-10 and WT5-32), or HPK1(K46E) (KE5-50 and KE5-58), and subjected to immunoblot analysis with anti-c-Rel antibody (top). A part of the blot was probed with anti-SP1 antibody to control for the amount and integrity of the nuclear extracts used (bottom).

Direct binding of tyrosine phosphorylated HPK1 to the SH2 domain of BASH in response to BCR stimulation. (A) BCR engagement induced binding to the BASH SH2 domain in vitro of phosphoproteins including HPK1. WEHI231.5 cells were stimulated for 3 min with anti–IgM F(ab′)₂ fragment (+) or left unstimulated (−). Lysates from these cells were precleared through binding to beads coupled with recombinant CBP fusion proteins of mutated (R373K) BASH SH2 domain, then incubated with the beads with the fusion proteins of the WT (lanes 1 and 2) or mutated (R373K) BASH SH2 domain (lanes 3 and 4). Bound material was eluted from the fusion protein, separated by SDS-PAGE, and subjected to immunoblotting with antiphosphotyrosine (anti-pY) Ab (top). Subsequently, the membrane was reprobed with anti-HPK1 #7 Ab (bottom). Whole cell lysates were included for visualization of all tyrosine-phosphorylated proteins (lanes 5 and 6). (B) Interaction of endogenous BASH with HPK1 in B cells. WEHI231.5 (1.5 × 10⁷) or splenic B cells (6.8 ×10⁷) were stimulated with anti–IgM F(ab′)₂ fragment for the indicated periods of time (min). Lysates from these cells were immunoprecipitated (IP) with rabbit anti-BASH serum (lanes 1–4, 13, and 14) or control antiserum against MIST peptide (unpublished data; lanes 5–8 and 15), separated on SDS-PAGE, and then immunoblotted with anti-HPK1 #7 Ab (top). The same membranes were reprobed with anti-BASH (C-19) Ab to visualize the immunoprecipitated BASH proteins (bottom). The presence of HPK1 in whole cell lysates was ascertained by immunoblotting with the anti-HPK1 #7 Ab (lanes 9–12, 16, and 17). (C) In vivo interaction of HPK1 and BASH is dependent on the BASH SH2 domain. Stably transfected WEHI231 cells (10⁷) expressing either T7 epitope–tagged WT BASH (T7-BASH) or a BASH SH2 domain truncation mutant (T7 BASHΔSH2) were stimulated with anti–IgM F(ab′)₂ fragment for the indicated periods of time (min). Lysates from these cells were immunoprecipitated with either anti-T7 Ab or control mouse IgG (mIgG), and the precipitates were immunoblotted with anti-HPK1 #7 Ab (top), then reprobed with anti-T7 Ab (middle). Anti-HPK1 immunoblotting of the lysates demonstrated comparable levels of HPK1 proteins between the transfectants (bottom). (D) Binding of recombinant HPK1 to the BASH SH2 domain depends on HPK1 tyrosine phosphorylation. GST and GST-HPK1-C proteins produced in bacterial strains containing (+) or not containing (−) an active PTK were affinity purified and incubated with immobilized CBP-BASH SH2 domain fusion proteins (as described in A). Bound proteins were eluted and immunoblotted with anti-GST (top, lanes 3–6), anti-HPK1 #7 (top, lanes 7–10), and anti-pY (bottom) Abs. One-twenty-seventh of the GST fusion proteins subjected to the binding reaction was loaded on the same gel (“input,” lanes 1, 2, 11, and 12). The electrophoretic mobility of GST and GST-HPK1-C proteins is indicated.