The ERK2-DBP domain opposes pathogenesis of a mouse JAK2V617F-driven myeloproliferative neoplasm

Abstract

Although Ras/mitogen-activated protein kinase (MAPK) signaling is activated in most human cancers, attempts to target this pathway using kinase-active site inhibitors have not typically led to durable clinical benefit. To address this shortcoming, we sought to test the feasibility of an alternative targeting strategy, focused on the ERK2 substrate binding domains, D and DEF binding pocket (DBP). Disabling the ERK2-DBP domain in mice caused baseline erythrocytosis. Consequently, we investigated the role of the ERK2-D and -DBP domains in disease, using a JAK2-dependent model of polycythemia vera (PV). Of note, inactivation of the ERK2-DBP domain promoted the progression of disease from PV to myelofibrosis, suggesting that the ERK2-DBP domain normally opposes progression. ERK2-DBP inactivation also prevented oncogenic JAK2 kinase (JAK2V617F) from promoting oncogene-induced senescence in vitro. The ERK2-DBP mutation attenuated JAK2-mediated oncogene-induced senescence by preventing the physical interaction of ERK2 with the transcription factor Egr1. Because inactivation of the ERK2-DBP created a functional ERK2 kinase limited to binding substrates through its D domain, these data suggested that the D domain substrates were responsible for promoting oncogene-induced progenitor growth and tumor progression and that pharmacologic targeting of the ERK2-D domain may attenuate cancer cell growth. Indeed, pharmacologic agents targeting the ERK2-D domain were effective in attenuating the growth of JAK2-dependent myeloproliferative neoplasm cell lines. Taken together, these data indicate that the ERK-D and -DBP domains can play distinct roles in the progression of neoplasms and that the D domain has the potential to be a potent therapeutic target in Ras/MAPK-dependent cancers.

Graphical abstract

Figure 1.

Effect of disabling the ERK2-DBP domain on erythropoiesis and JAK2V617F-driven PV progression. (A) Photograph of representative spleens from Erk2-WT and ERK2^Y261A/Y261A mice. (B) Flow cytometry analysis on the effect of disabling the ERK2-DBP domain on expansion of red blood cell precursors (CD44⁺/Ter119⁺) in mice of the indicated genotypes. (C) Adoptive transfer model of PV. Lin⁻ HSPCs were harvested from the BM of ERK2-WT, -KO or -Y261A (DBP) mice; transduced with JAK2V617F-expressing retrovirus; and IV injected into sublethally irradiated immunodeficient recipient mice (n = 7 per group). (D) Peripheral blood smears from ERK2-WT, -KO, and –DBP-Y261A HSPC recipient mice at 12 weeks after transfer. Scale bars, 20 µm, Wright-Giemsa stain. (E-F) Spleen size and weight of mice transferred with JAK2V617F-transduced ERK-WT–, -KO–, and -DBP–mutant HSPCs. Results are expressed as the mean ± standard deviation (SD). *P < .05. (G) BM cell counts of JAK2V617F-transduced ERK-WT–, -KO–, or -DBP–mutant mice. The data are expressed as the mean ± SD. *P < .05; n.s., not significant.

Figure 2.

Morphologic evidence of MF in the BM and spleen of mice receiving ERK2-DBP–mutant HSPCs. (A-B) Hematoxylin and eosin and reticulin staining of representative BM sections from mice that received ERK2-WT–, -KO– and -DBP–mutant HSPCs, at 12 weeks after BM transplant. (B′) Enlarged images delimited by the blue inset in panel B. Scale bars, 20 µm. (C-D) Representative images of sections of spleen from the mice above, stained with hematoxylin and eosin (C) or reticulin (D) 12 weeks after BMT. Scale bars, 50 µm.

Figure 3.

Preferential accumulation of myeloid progenitors in the BM of mice that received ERK2-DBP–mutant HSPCs. (A-B) Percentage and absolute number of GFP⁺/Gr1⁺ or GFP⁺/Mac1⁺ myeloid cells per 1 × 10⁶ cells in the BM and spleen of mice receiving ERK2-WT–, -KO–, and -DBP–mutant HSPCs. (C) Percentage and absolute number of GFP⁺/Ter119⁺ erythroid cells per 1 × 10⁶ cells in BM and spleen in the mice in panels A and B. (D) Percentage and absolute number of GFP⁺/CD41⁺ platelets per 1 × 10⁶ cells in BM and spleen in the same mice as in panels A and B. Data are expressed as the mean ± standard error of the mean (n = 5). *P < .05; **P < .01; ***P < .001.

Figure 4.

JAK2V617F-induced senescence was alleviated by inactivation of the ERK2-DBP domain. (A-B) Control and JAK2V617F-transduced HSPCs (3 × 10³) derived from ERK2-WT–, -KO–, or -DBP–mutant mice were sorted and plated on M3434 (A), or M3534 methylcellulose medium (B). The number of colonies was scored 7 days after plating. The transduction of the pMIG empty vector was used as the control. Data are expressed as the mean ± standard deviation (SD). ***P < .001. (C) Control and JAK2V617F-transduced HSPCs from mice with the same genotype as in panels A and B were sorted and then cultured with IMDM supplemented with IL-3, IL-6, and stem-cell factor cytokines for 3 days. Representative images of SA-β-gal staining are depicted on the left. Percentages of SA-β-gal⁺ cells were determined by counting 100 cells in randomly selected fields from triplicate cultures. Data are expressed as the mean ± SD.***P < .001. (D-E) pMIG and JAK2V617F-transduced HSPCs from the mice above were cultured for 3 days after transduction, and then GFP⁺ cells were sorted into Trizol. The relative expression of senescence-associated and profibrotic genes was measured by real-time quantitative polymerase chain reaction. Data are expressed as the mean ± SD. All results represent ≥3 independent experiments. **P < .01; ***P < .001; ****P < .0001. IMDM, Iscove modified Dulbecco medium.

Figure 5.

ERK-DBP induction of senescence depends upon interaction with Egr1. (A) Real-time quantitative polymerase chain reaction analysis of Egr1 expression in pMIG or pMIG-JAK2V617F–transduced HSPCs from ERK2-WT–, -KO–, and -DBP–mutant mice that were cultured for 3 days. (B) The physical interaction between ERK2-DBP domain and the DEF motif of Egr1. (C) Physical association of ERK2 and Egr1 after PMA stimuli. Anti-FLAG anti-ERK immunoblots were performed on input and anti-FLAG immunoprecipitates (immppt) from protein extracts of SCID.adh cells transduced with pMICherry empty vector (EV), FLAG-Egr1 (WT), and FLAG-Egr1^Y253A mutant. (D-E) Colony-forming assays were performed on WT and Egr1^−/− HSPCs transduced with the indicated constructs (pMIG, EV; pMIG-JAK2V617F; pMICherry, EV; pMICherry-Egr1; pMICherry-Egr1Y253A). Sorted cells (1 × 10³) were plated on M3434 medium and scored after 7 days in culture. (D) Representative images of cultures and a graphic depiction of the mean ± standard deviation (SD) colonies are shown. **P < .01; ***P < .001. (F-G) Representative images of SA-β-gal staining and the percentages of SA-β-gal⁺ cells were determined in cultures equivalent to those in panel D. All results are from ≥3 independent experiments. Data are expressed as the mean ± SD. **P < .01.

Figure 6.

ERK-D domain targeting inhibited cellular proliferation of JAK2V617F-dependent cells. (A) Space filling model of ERK2 and the D docking domain (orange pocket region) complexed with a MAPK docking peptide (MKNK1, blue colored structure). (B) Selective inhibition of ERK2-mediated phosphorylation of RSK by D-domain inhibitor (#76). SCID.adh cells were pretreated with U0126 or #76 inhibitor and then stimulated with PMA. Immunoblot analysis of protein extracts was performed, using the following antibodies: anti-EGR1, anti–phospho-RSK, anti-RSK, anti–phospho-ERK, anti- ERK, and anti–β-actin, which served as a loading control. (C-D) Colony forming analysis was performed on JAK2V617F-transduced HSPCs treated with vehicle control, U0126 or #76. Cells (3 × 10³) were plated on M3434 medium and cultured for 7 days. The number of colonies was scored and compared. Data are expressed as the mean ± standard deviation (SD). **P < .01. (E) Human SET-2 cells expressing JAK2V617F were treated with vehicle control, U0126 (10 µM), or #76 at the indicated doses. Cell viability and proliferation were measured each day for 3 days and normalized to untreated cells at day 0 and are expressed as the mean ± SD. **P < .01; ***P < .001. (F) Human SET-2 cells were implanted subcutaneously into Il2rg^−/−Rag2^−/− mice and left to reach a size of 40 mm³ before treatment every other day by intraperitoneal injection with vehicle control (dimethyl sulfoxide) or #76 at 10 mg/kg. Tumor size and cell recovery upon tumor disaggregation are expressed as the mean ± SD (n = 6 per condition). Results in panels B-E were derived from at least 3 independent experiments.

Figure 7.

Derivatives of ERK-D domain inhibitor #76 exhibit greater potency in inhibiting the growth of JAK-dependent cells. (A) Selective inhibition of ERK2-mediated phosphorylation of RSK by the D-domain inhibitors #76, 30 g, and 30 h. SCID.adh cells were pretreated with the indicated inhibitors and then stimulated with PMA. Immunoblot analysis of protein extracts was performed with the following antibodies: anti-EGR1, anti–phospho-RSK, anti-RSK, anti–phospho-ERK, anti- ERK, and anti–β-actin, which served as a loading control. (B) Cell lines expressing JAK2V617F (SET-2, UKE-1, KU812, and BAF/3-JAK2) were treated with vehicle control, U0126, #76, 30 g, and 30 h at the indicated doses. All results are from ≥3 independent experiments. Cell viability and proliferation were measured after 1 day of culture and normalized to untreated cells at day 0. Data are expressed as the mean ± standard deviation. *P < .05; **P < .01; ***P < .001; ns, not significant.