hnRNP K: A Driver of Myeloid Malignancy

Acute myeloid leukemia (AML) is increasingly defined as a heterogeneous group of malignancies divided largely by recurrent genetic and/or chromosomal aberrations. With this knowledge, development of targeted therapeutics has improved outcomes for some AML subsets. Increasing our understanding of alterations that drive disease development and progression will expand our armamentarium of therapeutic opportunities. To this end, our laboratory has identified a subset of AML that harbors increased expression of heterogeneous nuclear ribonucleoprotein K (HNRNPK; protein- hnRNP K).

In a sizeable cohort of patients with AML, we have demonstrated HNRNPK gene amplification, RNA overexpression, and protein overexpression using FISH, qRT-PCR, and reverse phase protein array analyses. Importantly, the patient population with high protein expression of hnRNP K exhibited decreased median overall survival, increased WBC, and increased blast percentage in both bone marrow and peripheral blood. Together, these data suggest that hnRNP K may function as an oncogene in AML when overexpressed.

To formally test this hypothesis, we generated a Vav-Cre;Hnrnpk^Tg mouse model. Indeed, these mice overexpress hnRNP K in comparison to Vav-Cre mice or unrecombined littermates. Critically, Vav-Cre;Hnrnpk^Tg mice develop myeloid malignancies. When transplanted, these malignancies have a decreased latency period and are rendered more aggressive. Currently, we are evaluating HSPC proliferation, differentiation potential, and using time of flight mass cytometry (CyTOF) to extensively characterize these transplanted leukemias.

Functionally, hnRNP K is an RNA and ssDNA binding protein that can shuttle bidirectionally between the nucleus and cytoplasm. To further understand how hnRNP K can cause AML, we are evaluating subcellular localization of hnRNP K in the context of overexpression. As our preliminary experiments suggest a predominant cytoplasmic localization when hnRNP K is overexpressed, we are performing hnRNP K RNA-immunoprecipitation (RIP) followed by next generation sequencing in human cell lines with inducible overexpression of hnRNP K. In addition to confirming our previous observations of hnRNP K binding to the MYC transcript, we have identified several other putative targets for hnRNP K binding that may potentiate these malignant phenotypes.

Additionally, we have performed RNA-Seq on HNRNPK- overexpressing human cell lines. These data suggest a significant role for hnRNP K in the regulation of innate immune response-observations that are consistent with phenotypes seen in our transplant models. Together, these studies provide insight into the pathways most critically altered in hnRNP K-overexpressing AML and will allow for appropriate utilization of therapeutic modalities.