De-JAKing resistance with CDK7 inhibitors in post-MPN sAML

The prognosis of secondary acute myeloid leukemia (sAML) cases that arise from advanced myeloproliferative neoplasms (MPNs) is poor. In this issue of Blood, Fiskus et al¹ demonstrate that cyclin-dependent kinase 7 (CDK7) is a vulnerability in JAK2 inhibitor–resistant sAML cells in vitro and suggest that treatment with CDK7 inhibitors should be explored as a potential therapeutic intervention.

Patients with MPNs, including high-risk myelofibrosis and polycythemia vera, have mutations in components of the JAK2 pathway, including the thrombopoietin receptor MPL, JAK2, and calreticulin. The development of the potent JAK1/2 inhibitor ruxolitinib was a pivotal advance in MPN therapy, resulting in rapid and durable response with significant improvement in disease-related symptoms, although its use is limited by anemia and thrombocytopenia in some cases.^2,3 However, a fraction of patients develop resistance, accumulating other mutations and culminating in post-MPN sAML, a disease with dismal prognosis. Therefore, the development of new effective therapies for post-MPN sAML is an unmet need.

CDK7 is a critical component of gene transcription. As a subunit of the transcription factor complex TFIIH, it promotes transcription initiation through phosphorylation of the C-terminal domain of RNA Pol II and transcription elongation by phosphorylation of CDK9. Its kinase activity is also essential in cell cycle progression as it influences every phase of the cell cycle by phosphorylating the cell cycle checkpoint CDK1, CDK2, CDK4, and CDK6.⁴ 

The growth-promoting function of CDK7 places it at the center of oncogenic function through multiple mechanisms. For example, CDK7 acts as a hub in the oncogenic RB-E2F pathway and in MYC-dependent glycolytic cascade in multiple myelomas.⁵ In T-cell lymphoblastic leukemia, the CDK7 kinase activity seems to drive expansion of leukemia stem cells by regulating the activity of the RUNX1 superenhancer element.⁶ In addition, when given with the B-cell lymphoma 2 inhibitor venetoclax, treatment with CDK7 inhibitors may target cycling AML stem cells.⁷ As a consequence of the accumulating evidence indicating that CDK7 activity is central to core dependencies in cancer, a number of CDK7 inhibitors have been developed with various degrees of specificity, including a few which have progressed to phase 1/2 clinical trials for the treatment of solid tumors. Among these is the oral, noncovalent small molecule SY-5609, which has high specificity for CDK7.⁴ 

In this study, Fiskus et al identified that CDK7 expression was increased in post-MPN sAML patient samples with JAK2 mutations and used pharmacologic and gene-editing approaches to demonstrate that CDK7 function is critical for the viability of post-MPN sAML stem and progenitor cells. The specificity of CDK7 activity was functionally confirmed as they found a reduction in cell cycle and an increase in cell death in vitro, thereby suggesting that CDK7 is a potential vulnerability in post-MPN sAML.

Mechanistically, the authors combined transcriptomic and proteomic profiling to show that SY-5609 treatment triggered a significant negative enrichment of MYC/E2F signatures, as well as of cell cycle and cell death pathways. These results validate the specificity of CDK7 function in known target pathways in post-MPN sAML and suggest that it may control specific enhancer elements that activate expression of known oncogenes, such as c-MYC, c-MYB, and E2F. As the authors focused on defining targeted pathways in the study, additional studies will be necessary to refine our understanding on how CDK7 inhibition may affect other functions that control the expansion of high-risk MPN stem cells. For example, MYC activity regulates the glycolytic cascades in multiple myelomas,⁵ which could be a vulnerability for post-MPN sAML stem cell expansion. In addition, in vivo studies should focus on understanding the impact of CDK7 inhibition in the bone marrow microenvironment and immune response.

Surprisingly, the authors found that SY-5609 has a synergistic response when combined with ruxolitinib in post-MPN sAML cells, indicating that CDK7 inhibition resensitizes MPN cells to JAK2 inhibitors. Although the study falls short in uncovering the mechanism by which JAK2 resistance relates with CDK7 dependency, an open question is whether the upregulation in CDK7 expression in MPN cells causes JAK2 resistance. If this is the case, the understanding of the molecular base driving the transition of JAK2 to CDK7 dependency may have critical mechanistic and clinical implications. In addition, these results support further studies to evaluate the combined treatment with ruxolitinib and SY-5609 for high-risk MPN cases to preempt possible JAK2 resistance.

Finally, the authors used an elegant protein domain-specific CRISPR screen technique to identify candidate resistance factors, such as the chromatin components CREB binding protein and E1A associated p300 protein in SET2 and BRD4 in HEL92.1.7 cells, as candidate codependencies. These results highlight the central role of transcription regulation by CDK7 in sAML and suggest that combined treatment with bromodomain and extraterminal protein (BET) inhibitors, such as OTX015, may have synergistic effects to effectively eliminate the post-MPN sAML stem cells. These options would only gain therapeutic strength after the careful study of possible toxicity issues associated with the combined use of these drugs in in vivo models.

Overall, the study by Fiskus et al identifies CDK7 as a novel candidate target for therapeutic intervention in post-MPN sAML. Considering the poor prognosis for patients with this disease and that CDK7 inhibitors are in clinical trials, the study supports its use as a single agent or in combination protocols with ruxolitinib or BET inhibitors against advanced MPNs.

Conflict-of-interest disclosure: The author declares no competing financial interests.