Thrombopoietin Receptor Signaling Promotes the Competitive Advantage of TET2-Mutated Hematopoietic Stem and Progenitor Cells in Clonal Hematopoiesis

Clonal Hematopoiesis of Indeterminate Potential (CHIP) is an age-related condition that predisposes individuals to blood cancers. The time interval from acquisition of CHIP to overt neoplasia can span many years, providing an opportunity to delay or prevent malignant transformation. A key to achieving this goal is identifying differential dependencies between mutant and normal hematopoietic stem and progenitor cells (HSPCs). Targeting such dependencies may prevent against CHIP-related health risks.

One of the most commonly mutated genes in CHIP is TET2, which encodes a methylcytosine dioxygenase. In mouse models, Tet2 knockout (KO) HSPCs have a competitive advantage over wild-type (WT) HSPCs. In this study, we aimed to identify differential dependencies between Tet2^KO and Tet2^WT HSPCs. We first generated isogenic murine cell lines by overexpressing HOXB4, a transcription factor involved in self-renewal, in HSPCs derived from a Tet2^KO mouse and Tet2^WT littermate. Using the isogenic cell lines termed HPC^HOXB4, we conducted an epigenetic drug screen and found SGC0946, a selective inhibitor of H3K79 methyltransferase DOT1L, reduced the advantage of Tet2^KO HPC^HOXB4 cells in competition assays by preferentially reducing their proliferation and viability. SGC0946 also reduced the competitive advantage of unmodified Tet2^KO HSPCs in competition assays, indicating that the dependency was not a consequence of HOXB4 overexpression.

To decipher the mechanism behind this differential dependency, we performed RNA-seq and ChIP-seq for H3K79me2 on Tet2^KO and Tet2^WT HPC^HOXB4 cells at baseline and after treatment with SGC0946. Expression of Mpl, which encodes the thrombopoietin receptor TPOR, was 20-fold higher in Tet2^KO HPC^HOXB4 than in Tet2^WT cells at baseline and decreased after drug treatment. Mpl expression at the mRNA and protein level was also elevated in unmodified Tet2^KO HSPCs compared with Tet2^WT cells. Importantly, ChIP-seq analysis revealed that the level of H3K79me2 which associates with active transcription at the Mpl gene locus was higher in Tet2^KO than in Tet2^WT HPC^HOXB4 cells at baseline, a difference that was erased with SGC0946 treatment.

Based on the above findings, we hypothesized that Tet2^KO HSPCs are more dependent on TPOR signaling than Tet2^WT cells. Consistent with our hypothesis, a higher concentration of TPO in the culture medium increased the competitive advantage of Tet2^KO HPC^HOXB4 cells. Moreover, overexpression of Mpl in Tet2^KO HPC^HOXB4 cells rescued the effects of DOT1L inhibition, and shRNA-mediated suppression of Mpl expression was sufficient to decrease the competitive advantage of Tet2^KO HPC^HOXB4 cells and unmodified Tet2^KO HSPCs over their corresponding Tet2^WT counterparts.

To explore the translational relevance of our findings, we tested if inhibition of JAK2, a critical mediator of TPOR intracellular signaling, would also reduce the competitive advantage of Tet2^KO HSPCs. In line with our hypothesis, treatment with JAK2 inhibitors (fedratinib and AZ960), but not JAK1 inhibitors, suppressed the competitive advantage of Tet2^KO HPC^HOXB4 cells in vitro. Importantly, competitive repopulation studies showed that treatment with ruxolitinib, a clinically approved JAK1/2 inhibitor, also reduced the competitive advantage of untransformed Tet2^KO HSPCs in vivo.

To determine if our findings apply to humans, we generated two TET2^KO cell pools from cord blood HSPCs using CRISPR/Cas9. The TET2^KO cells expressed higher levels of TPOR and were more sensitive to treatment with SGC0946 and ruxolitinib compared with control cells. In addition, we expanded HSPCs from TET2-mutated mobilized peripheral blood samples in short-term cultures and monitored the size of the mutant clones by measuring the allele frequency of the sample-specific TET2 variants. In untreated cultures, the relative size of the mutant clones increased over time, an effect that was diminished by treatment with either SGC0946 or ruxolitinib.

In summary, our findings demonstrate that TET2-mutated HSPCs are more dependent on TPOR signaling than TET2^WT HSPCs. This differential dependency is mediated through increased H3K79 methylation at the MPL gene locus and enhanced expression of MPL. Furthermore, inhibition of the TPOR/JAK2 signaling axis is effective in suppressing the competitive advantage of TET2-mutated HSPCs, thus identifying a tractable therapeutic target for TET2-mutated CHIP.

No relevant conflicts of interest to declare.