Dual targeting of JAK2 and ERK1/2 enhances clone control and therapeutic efficacy in MPN

The activation of JAK2 signaling triggers the overproduction of mature blood cells of the myeloid lineage, ultimately leading to myeloproliferative neoplasms (MPN) and increasing the risk of leukemic transformation. Treatment with JAK2 inhibitors can offer clinical benefits, however the compensatory activation of MAPK pathway signaling with the sequential RAF, MEK, and ERK kinases, hinders their effectiveness. Therefore, MAPK pathway components such as ERK1/2 have been suggested as potential targets for the optimal treatment of MPN.

Brkic, et al. postulated that dual targeting of JAK2 and ERK1/2 could enhance clone control and therapeutic efficacy. The preclinical study was recently published in Leukemia, and here we report the key findings.¹

Methods

Genetic and pharmacologic targeting of ERK1/2 was performed in JAK2^V617F MPN mice, cells, and patient clinical isolates.

Mouse models

Genetic studies were used to assess the role of ERK1/2 for hematopoiesis, JAK2^V617F mice expressing Mx-1-Cre recombinase C57BL/6 were crossed with ERK1^−/−ERK2^fl/fl C57BL/6. JAK2^V617F ERK1^−/−ERK2^fl/fl Mx-1-Cre CD45.2 mice were induced by poly I:C (pIpC), resulting in ERK1/2 deficiency of hematopoietic cells. Bone marrow (BM) was mixed 1:1 with BM from JAK2 wildtype (WT) CD45.1 mice and a total of 2 × 10⁶ cells were transplanted into lethally irradiated CD45.1 C57BL/6 female mice. For the secondary transplantation, 2 × 10⁶ pooled BM cells from primary recipients were injected into lethally irradiated CD45.1 secondary recipients.

Inhibitor studies were also used, such as CD45.1 C57BL/6 female recipient mice being competitively transplanted with JAK2^V617F Vav-Cre CD45.2 BM mixed 1:1 with JAK2^WT CD45.1 BM were randomized post-transplant according to blood counts and treated by gavage for 1–4 weeks. An MPLW515L model was used to evaluate fibrosis, by reticulin staining 30 weeks after transplantation.

Studies of combined genetic targeting of ERK1/2 and inhibition of JAK2 were carried out: CD45.1 C57BL/6 female recipients of JAK2^V617F ERK1^−/−ERK2^fl/fl Mx-1-Cre CD45.2 BM mixed 1:1 with JAK2^WT CD45.1 BM were treated with pIpC 5 weeks after transplantation and ruxolitinib for 2 weeks. LTT462, an ERK1/2 inhibitor, was administered by gavage at 75 mg/kg qd, ruxolitinib at 60 mg/kg bid and MK-8353 at 30/40 mg/kg bid.

Cell lines

Ba/F3 cells stably expressing JAK2^V617F along with erythropoietin receptor were cultured in RPMI1640/10%FCS, supplemented with 10 U/ml erythropoietin receptor if expressing JAK2^WT. Proliferation was assessed upon ERK1/2 knockdown and/or exposure to inhibitors for 48 hours using cell viability luminescent assay.

Cells were exposed to an inhibitor for 4 or 24 hours and lysed in presence of protease arrest and phosphatase inhibitor for signaling analyses.

Patient samples

Blood, BM samples, and clinical data from MPN patients were collected and diagnoses of MPN were recorded according to the revised World Health Organization (WHO) criteria. Colony formation assays involved CD34⁺ peripheral blood mononuclear cells (PBMC) or BM cells being plated at 3,000 or 150,000 cells/well, respectively, into MethoCult with 0.25 μM ruxolitinib and/or 0.25–2.5 μM LTT462. Colony number and subtypes including erythroid, granulocyte-macrophage, and granulocyte-erythroid-macrophage-megakaryocyte were scored after 10 days.

For signaling analyses, PBMC were serum-starved in αMEM/1% BSA and exposed to 0.25 μM ruxolitinib and/or 2.5 μM LTT462 for 16 hours.

Results

Mouse models

Genetic targeting of ERK1/2 mitigates the MPN phenotype and impairs the fitness of the JAK2^V617F clone.

• Erythrocytosis reflected by increased hematocrit (Hct) and reticulocytes was moderated by ERK1/2 deficiency in JAK2^V617F settings (Hct, p ≤ 0.01; reticulocytes, p ≤ 0.05 in week 16 post-transplant), without inducing cytopenia.
• CD45.2/CD45 chimerism reflecting JAK2^V617F allele burden was significantly reduced by ERK1/2 deficiency in peripheral blood (PB) (p ≤ 0.001) and BM (p ≤ 0.01) 16 weeks after transplantation.
• BM fibrosis was not evident in ERK1/2 deficient JAK2V617F mice, whereas fibrosis was detected in JAK2^V617F mice with intact ERK1/2.
• Hematopoietic progenitor cells including Lin⁻Sca1⁺Kit⁺ and Lin⁻Sca1⁻Kit⁺ multipotent myeloid progenitors were substantially reduced with impaired myeloid colony formation from BM and spleen cells ex vivo, without compromising BM cellularity and hematopoietic stem/ progenitor cell frequencies.

Dual JAK2 and ERK1/2 inhibition by ruxolitinib/LTT462 in a JAK2^V617F MPN preclinical model.

• Hct was significantly corrected by combined JAK2/ERK1/2 inhibition with LTT462 and ruxolitinib (p ≤ 0.001).
• Splenomegaly was moderated by treatment with LTT462 or ruxolitinib (p ≤ 0.001).
• LTT462 enhanced ruxolitinib effects when both agents were combined.
• JAK2^V617F clone control was significantly improved, reflected by CD45.2/CD45 chimerism in BM as compared to ruxolitinib monotherapy (p ≤ 0.05 at Week 4 of treatment).
• The activation of ERK1/2 downstream targets RSK3, DUSP6, and ETV5 was effectively inhibited as shown by reduced phosphorylated RSK3 (pRSK3) and significantly lower DUSP6 and ETV5 expression. However, these ERK1/2 effectors were not reduced by ruxolitinib highlighting that ERK1/2 kinase activity needs to be targeted to increase therapeutic efficacy with respect to MAPK signaling.
• Combined ruxolitinib/LTT462 was tolerable with restoration of BM and spleen architecture and corrective effects were maintained upon prolonged therapy.

Dual JAK2 and ERK1/2 inhibition by ruxolitinib/LTT462 in a MPLW515L MPN preclinical model.

• Improved splenomegaly control (p ≤ 0.0001 at Week 4 of treatment) and corrected thrombocytosis in MPLW515L mice.

• Enhanced reduction of leukocytosis vs ruxolitinib monotherapy (p ≤ 0.001 at 2 weeks of treatment).
• MPLW515L mice developed pronounced BM fibrosis, which was moderated by ruxolitinib and more effectively reduced by combined ruxolitinib/LTT462 (p ≤ 0.0001 at 4 weeks of treatment), with consistent decrease of fibrosis by 2 grades.
• Ruxolitinib reduced the expanded BM megakaryocytes, while combined ruxolitinib/ LTT462 improved correction of splenic architecture and clearance of extramedullary hematopoiesis from the liver, while BM cellularity was maintained.
• JAK2/ERK1/2 inhibition was able to reduce mutant allele burden reflected by GFP⁺ cells in blood, BM, and spleen with consistent clone reductions in animals on combination treatment. Dual inhibition was also able to prolong survival as compared to vehicle-treated animals after 2 weeks of treatment.
• Corrective effects were maintained upon extended treatment despite the aggressive disease dynamics. Thrombocytopenia, which progressively developed in the longer course limited survival benefit in this model.
• Analyses of serum concentrations in steady state excluded accumulation of inhibitors when administered as combination in MPLW515L or JAK2^V617F settings suggesting combining JAK2 and ERK1/2 inhibition was safe.

Combined pharmacologic JAK2/ERK1/2 inhibition with ruxolitinib and ERK inhibitors decreased proliferation of JAK2V617F cells, corrected erythrocytosis, and normalized splenomegaly of JAK2^V617F MPN mice along with effects on MPN clone size.

• Erythrocytosis was reduced by ERK1/2 deficiency or by treatment with ruxolitinib.
• ERK1/2 deficient settings enhanced ruxolitinib effects as indicated by near-normalized Hct values upon combined treatment (p ≤ 0.05).
• Ruxolitinib corrected splenomegaly without additional benefit by targeting ERK1/2.
• CD45.2/CD45 chimerism reflecting JAK2^V617F allele burden was significantly reduced by targeting ERK1/2 in PB (p ≤ 0.05) and BM (p ≤ 0.0001) and reductions were most profound with combined ERK1/2 deficiency and ruxolitinib (PB, p ≤ 0.01; BM, p ≤ 0.0001). These outcomes in BM and in myeloid and erythroid progenitor subsets were also seen in the spleen.
• Expression of 28 ERK1/2 downstream targets and 36 cytokines in BM was most effectively reduced by combined targeting of ERK1/2 and ruxolitinib.

Cell lines

Pharmacologic ERK1/2 inhibition enhances sensitivity to JAK2 inhibition in MPN cells. Proliferation dynamics of JAK2^V617F cells was significantly reduced in ERK1/2 deficient settings induced by shRNA-mediated genetic targeting of ERK1/2 with two different hairpins #1 and #2, as indicated by reduced increase of cell count over 4 days. ERK1/2 deficient JAK2^V617F cells were more susceptible to JAK2 inhibition by ruxolitinib at increasing concentrations with 4- to 5-fold reduced half-maximal inhibitory concentration (IC₅₀) (shERK1/2 #1, p ≤ 0.0001; shERK1/2 #2, p ≤ 0.001).

Pharmacologic ERK1/2 inhibition by LTT462 increased susceptibility of JAK2^V617F cells to JAK2 inhibition with ruxolitinib with decreased IC₅₀ (p ≤ 0.0001) and dose-dependently suppressed ERK1/2 downstream targets including pRSK3, DUSP6, and c-MYC expression; thus, enhancing the effects mediated by ruxolitinib when used in combination.

Patient samples

Dual JAK2 and ERK1/2 inhibition by ruxolitinib/LTT462 enhances suppression of myeloid colony outgrowth and ERK1/2 target activation from primary JAK2^V617F patient cells.

ERK1/2 inhibition by LTT462 at 0.25 µM (p ≤ 0.0001), 1 µM (p ≤ 0.0001) and 2.5 µM (p ≤ 0.0001) improved control of myeloid colony formation from BM cells seen with ruxolitinib at 0.25 µM in a dose-dependent manner. Enhanced suppression of colony outgrowth was not restricted to a specific subtype of MPN but seen in primary cells from patients with primary myelofibrosis, polycythemia vera, and essential thrombocythemia.

Erythroid as well as granulocytic-macrophage colony subtypes were affected by enhanced ruxolitinib/LTT462 effects and analysis of signaling in freshly isolated PBMC from a patient with JAK2^V617F mutated MPN exposed to inhibitors ex vivo for 16 hours showed improved inhibition of ERK1/2 downstream target RSK by ruxolitinib/LTT462 as reflected by pRSK3.

Conclusion

Analyses of paired blood and BM isolates from patients with JAK2^V617F MPN provide a first indication of enhanced corrective effects with dual JAK2/ERK1/2 inhibition compared with ruxolitinib monotherapy. Overall, the study showed targeting of JAK2 and ERK1/2 effectively addresses ERK1/2 kinases as a second node of oncogenic signaling, which enhances the therapeutic efficacy in a number of MPN settings, suggesting a novel treatment approach for patients with MPN. Future clinical trials should focus on dosage and treatment schedules as well as on specific vulnerabilities of JAK2^V617F mutant and WT settings to ERK1/2 inhibition.