Ineffective erythropoiesis of TET2 deficiency

(De)methylation of DNA at specific loci and elements can guide hematopoietic progenitor cell growth and differentiation.²  If imbalanced, hypermethylation can repress tumor suppressors, whereas hypomethylation can contribute to genomic instability.^2,3  Proteins TET-1, TET-2, and TET-3 function as Fe(II), 2-oxoglutarate- dependent dioxygenases to mediate DNA demethylation. This involves 5mC oxidation to 5hmC, 5fC, and 5caC. Seminal clinical studies have associated deletions and/or mutations within TET2 (4q24) with myelodysplastic syndrome (MDS), myeloproliferative neoplasm, chronic myelomonocytic leukemia, and acute myeloid leukemia.⁴  TET2 mutations commonly cluster within catalytic regions (residues 1104-1478, and 1845-2002) with missense mutations typically decreasing catalytic activity, 5hmC levels, and DNA methylation profiles. Such TET2 mutations additionally are associated with B-cell lymphogenesis^2,3  and certain T-cell leukemias.^2,3  TET-1 and TET-3, in contrast, are rarely mutated in hematological malignancies.

MDS, myeloproliferative neoplasm, and chronic myelomonocytic leukemia also commonly exhibit anemia because of ineffective erythropoiesis,^5,6  a feature that raises central questions regarding effects exerted by TET2 during normal erythropoiesis. Short hairpin RNA (shRNA)-mediated loss-of-function (LOF) studies now reported by Qu et al¹  focus on this question and contribute several novel findings. Prior studies using Tet2 knockout (KO) mice, and human erythroid progenitor cells (EPCs) ex vivo established that TET2 loss-of-function (LOF) impairs erythroid differentiation.^2,3  Qu et al now reveal that shRNA-mediated TET2 LOF selectively promotes the hyperexpansion of a CFU-E–like cohort of EPCs. Such hyperexpansion occurs among ∼20% of EPCs; depends sharply upon stem cell factor (SCF) and erythropoietin; includes a glycophorin A^low phenotype; and persists for up to 1 month. By comparison, no similar hyperproliferative effects of TET2 LOF were observed among later stage erythroblastic progenitors. As correlated with the SCF dependency of these dysregulated CFU-E–like progenitors, TET2 LOF led to an apparent approximate twofold increase in KIT levels, together with parallel decreases in the levels of SHP1 tyrosine phosphatase, a negative effector of KIT. SHP1 activation, however, was not assessed. Furthermore, inhibition of KIT by ST157 reversed TET2 LOF EPC phenotypes and promoted late erythroblast differentiation.

As isolated from ex vivo cultures, d6 CFU-E and d13 TET2 LOF CFU-E also were compared. In d13 TET2 LOF CFU-E, RNA sequencing revealed >10-fold elevated expression of the receptor tyrosine kinase AXL. AXL together with TYRO3 and MER are TAM receptors that share GAS6 as a common ligand.⁷  AXL is also an acute myeloid leukemia drug target.⁷  Interestingly, and as studied by Qu et al, the hyperproliferative and attenuated differentiation phenotypes of TET2 LOF CFU-E were each partially reversed by the AXL inhibitor R428. In compound Axl^−/−Mer^−/− KO mice, attenuated development of (pro)erythroblasts similarly has been observed.⁸  This effect, however, was not exhibited by single Axl^−/− KO mice. In normal d6 CFU-E vs d13 TET2 LOF CFU-E, Qu et al also analyzed effects of shRNA-mediated TET2 knockdown on DNA methylation. In TET2 LOF CFU-E, 5mC levels were heightened and 5hmc levels were decreased. In addition, region −414 of the AXL promoter was shown to be differentially methylated because of TET2 knockdown, but for 5hmC (and not 5mC). This 5hmc phenomenon, nonetheless, is not without precedent and has been observed for select DNA loci in TET2 KO embryonic stem cells. In future studies, possible alternate mechanisms underlying heightened AXL expression from TET2 LOF can now be considered, including possible effects on genes involved in AXL activation. In MDS patients with TET2 mutations, possible increases in AXL expression within EPCs also will be meaningful to assess.

Overall, these recent investigations by Qu et al provide new insight into several specific effects of TET2 deficiency on ineffective erythropoiesis and raise several questions of significant interest. Does SCF-dependent hyperproliferation of TET2 LOF CFU-E mechanistically involve elevated KIT expression, or is heightened KIT expression perhaps an indirect consequence of the actions of other TET2 targets that regulate CFU-E self-renewal? Beyond SHP1, what signal transduction factors might be affected by TET2 LOF that contribute to increased KIT signaling? What TET2-dependent mechanisms upregulate AXL expression, and is AXL similarly upregulated by inhibitors of DNA 5mC methylation? To what extent is R428 specific for AXL (and at the dosing used)? And what are the candidate endogenous sources of AXL ligands (eg, blood island macrophage)? The insight, and EPC stage-specificity model systems generated by Qu et al, should serve well in providing tractable inroads to critically assess these problems.