Hematopoietic progenitor kinase 1 (HPK1) negatively regulates prostaglandin E₂–induced fos gene transcription

Signals generated by the prostaglandin E₂ receptors (PGE₂Rs) are thought to be propagated via the classical, well-characterized G-protein signal transduction pathways, utilizing phospholipase B and adenylyl cyclase as effector molecules.1 We postulate that the PGE₂Rs may activate other pathways, perhaps, engaging one of the Ste20 family members, and utilize it to transmit or regulate the PGE₂-induced fos gene transcription. Data presented here support our contention that exposure to PGE₂ activates HPK1 kinase activity, which, in turn, negatively regulates PGE₂-induced fos gene transcription.

All cell lines used in this study are available from ATCC (Manassas, VA). The antihemagglutinin (HA) epitope antibody (12CA5) was purified from hybridoma supernatants in our laboratory. The anti-HPK1 polyclonal antibody no. 7,2 a generous gift from Dr F. Kiefer (Max-Planck Institute for Physiological and Clinical Research, Bad Nauheim, Germany) was used in Western blot analyses. Affinity-purified rabbit polyclonal antibody against the amino acid residues 341 to 366 of human HPK1 used for immunoprecipitations was custom produced for our laboratory by Cocalico Biologicals (Reamstown, PA). γ-[³²P]adenosine triphosphate (γ-[³²P]ATP) was obtained from Perkin Elmer Life Science (Boston, MA). Histone H2A, the exogenous substrate used in in vitro kinase reactions, was purchased from Roche Applied Science (Indianapolis, IN). PGE₂ was purchased from Calbiochem-Novabiochem (San Diego, CA).

The pcDNA3 vector carrying mouse cDNAs that encode the HA-tagged wild-type and kinase-inactive (K46E) mutant of murine HPK1 was a generous gift from Dr F. Kiefer. A construct encoding the firefly luciferase reporter gene under the control of human fospromoter (Hu-fos-Luc) was described.3 

Jurkat T cells (1.5 × 10⁷) were transfected as previously described.4 Whole cell lysates derived from resting or stimulated transfectants were subjected to immunoprecipitations by the indicated antibodies and were subjected to in vitro kinase reactions as described.2 Anti-HPK1 polyclonal antibody no. 7 ² was used in immunoblot assays to confirm the amounts of immunoprecipitated HPK1.

Jurkat cells were cotransfected with 2.5 μg of the Hu-fos-Luc reporter construct and 100 ng of pNullRenilla luciferase reporter construct along with 10 μg of the indicated experimental constructs. Dual luciferase assays were performed according to the manufacturer's direction (Promega, Madison, WI), and the relative luciferase activity was measured by the Monolight 2010 luminometer (Analytical Luminescence Laboratory, Ann Arbor, MI).

We hypothesized that the binding of PGE₂ to its cognate receptors would generate signals that activate the catalytic activities of mammalian Ste20 family members. As observed in many cell types,5 6 our in vitro kinase activity screen of Ste20 orthologs revealed that most are constitutively active, and are not responsive to PGE₂ stimulation, in Jurkat T cells (data not shown). However, HPK1, a hematopoietic cell–specific Ste20 family member, is robustly activated upon in vivo exposure to PGE₂, in a concentration-dependent manner (Figure 1A-B). Exposure to 1 nM PGE₂induced HPK1 kinase activity. Maximum levels of kinase activity were achieved when cells were stimulated with at least 10 nM PGE₂ (Figure 1A-B, lanes 4-6). The 2 lowest PGE₂ concentrations that activated HPK1, 1 and 10 nM, are at physiological levels found at the sites of inflammation.

It is possible that the kinase activity detected in our assay was due to a PGE₂-responsive kinase that coprecipitated with HPK1. To address this concern, we transfected either the wild-type or a kinase-defective HA-HPK1 K46E mutant into Jurkat cells and stimulated the transfectants with 10 nM PGE₂. The lack of kinase activity from the HA-HPK1 K46E mutant (Figure 1D-E, lanes 5-6) strengthened our conclusion that the observed kinase activities in Figure 1A-B were catalyzed by HPK1.

Furthermore, using anti-HPK1 antibody, we demonstrated that endogenous HPK1 responded to PGE₂ stimulation in all hematopoietic cell lines tested (Figure 1G, lanes 1-8). Kinase activity was not detected in the anti-HPK1 immunoprecipitates from the K562 cells, a chronic myelogenous line that does not express HPK1 (Figure 1G, lanes 9-10). Western blot analysis using anti-HPK1 antibody confirmed that comparable amounts of HPK1 were present in all kinase reactions (Figure1C,F,H). Taken together, these observations established PGE₂ as a potent activator of HPK1 kinase activity in hematopoietic cells. This is the first example of G protein–coupled receptor (GPCR) regulation of the catalytic activity of a Ste20 ortholog.

PGE₂ stimulation induces c-fos gene transcription in many cell types7,8 via a poorly characterized cyclic adenosine monophosphate (cAMP)–independent mechanism.9 Because Ste20 orthologs have been implicated as critical signaling molecules in many receptor systems,10 we postulated that HPK1 might function as a facilitator or a regulator of the PGE₂-induced Fos activation signal. To assess the effect of HPK1 on fos gene transcription, we cotransfected either a wild-type HPK1 or the catalytically inactive K46E mutant construct, along with afos promoter–regulated luciferase reporter construct, into the Jurkat T-cell line. Consistent with findings observed in other cell types, PGE₂ treatment increased fos promoter activity by approximately 3-fold in the Jurkat T-cell line (Figure2A). Ectopic expression of wild-type HA-HPK1 inhibited the PGE₂-induced fos promoter activity by approximately 60% to 70%. The K46E mutant failed to inhibit fos transcription, suggesting that the HPK1 kinase activity is required for the inhibition of fostranscription. This conclusion is consistent with a recent finding that, despite its ability to activate Jun N-terminal kinase (JNK) and c-Jun, HPK1 functions as a negative regulator of T-cell antigen receptor (TCR)– and B-cell antigen receptor (BCR)–mediated AP-1–dependent gene transcription.11 However, one must be cognizant of the possibility that the observed inhibition offos transcription may be due to kinase activity–dependent sequestration of critical signal transduction components by the overexpressed HPK1. It has recently been demonstrated that sequestration of SLP-76 family member(s) represents a possible mechanism underlying the HPK1-mediated inhibition of TCR and BCR signals.12 A better understanding of HPK1's role in biologic processes awaits the development of an HPK1-deficient animal model.

The mechanism by which PGE₂Rs couple activation signals to HPK1 is not known. While 4 PGE₂R genes have been identified (EP1 to EP4), the G_s-linked EP4 receptor is the predominant PGE₂R expressed in Jurkat T cells.13 Reverse transcriptase–polymerase chain reaction (RT-PCR) analysis revealed that all cell lines used in this study also express a high level of EP4 receptor (data not shown). The EP4 receptor links activation signals to G_sα, which, in turn, activates adenylyl cyclase, leading to elevation of cAMP concentrations. Although it is enticing to implicate cAMP as the second messenger that activates HPK1, administration of cell-permeable forms of cAMP (N⁶, O^2′-dibutyryl-cAMP, 8-bromo-cAMP, and 8-4-chlorophenylthio cAMP) to Jurkat cells failed to activate HPK1 (data not shown). These cAMP agents are biologically active as shown by their abilities to activate cAMP-dependent serine/threonine protein kinase A (PKA; data not shown). The inability of exogenous cAMP to activate HPK1 revealed that both the negative regulatory signal and the positive transcriptional signal9are not downstream of the cAMP signaling pathway. Further characterization of this pathway may reveal an insight into a novel mechanism for negative regulation.

We are grateful to Dr Joanne Pratt (Olin College, Needham, MA) for thorough reading of the manuscript. We thank Dr J. Pratt, Dr S. Baksh, Dr S. Pyarajan, and Dr Y.-J. Jin for thoughtful discussions. We are grateful to Dr F. Kiefer for providing the reagents.