TET2 Regulates Stability of Mismatch Repair Protein MSH6 and Helps Maintain Genomic Stability Independent of Its Dioxygenase Function

Abstract

Loss of function TET2 mutations (TET2^MT) are one of the most common somatic mutations in myeloid neoplasia (MN). TET enzymes, encoded by TET1, TET2 and TET3, are Fe²⁺- and α-ketoglutarate (αKG)-dependent DNA-dioxygenases that oxidize 5-methylcytosine (mC), resulting in the demethylation of promoter and enhancer mCpG. Of all TET enzymes, TET2 accounts for more than half of the DNA TET activity in hematopoietic cells. TET2’s net effect is the demethylation of the promoters and enhancers, which is crucial for efficient transcription required for lineage commitment, differentiation, proliferation, and survival. Thus, functional inactivation of TET2 in hematopoietic stem and progenitor cells (HSPCs) leads to the abnormal accumulation of mC, causing loss of lineage plasticity and transcriptional repression of tumor suppressor and differentiation genes, leading to a volatile pre-neoplastic state. The loss of TET2 is associated with a higher mutational burden around 5hmC locus (Pan, F et al., Nature communications 2017), although the mechanism is largely unknown. Here, we report that non-catalytic N-terminal domain of TET2 recruits the mismatch repair (MMR) complex protein MHS6 to the site of 5hmC, facilitating G to T mismatch repair and maintaining genomic stability.

Methods and Results

Analysis of the murine whole genome sequencing reveals that C to T mutations are consistently associated with regions of higher 5hmC, not 5mC, in TET2 mutant cells. This pattern is not observed in high 5hmC loci in cells with wild-type (WT) TET2. This suggests a role for TET2 in regulating genomic stability beyond the canonical deamination of 5mC. To investigate whether TET2^MT are associated with higher mutations rates, we analyzed whole exome sequencing (WES) data from 435 MN cases. TET2^MT cases (n=82/435; 19%) showed nearly a 2-fold increase in median single nucleotide variants (SNVs) compared to WT (p<0.0001), including a 2.1-fold increase in truncations (p=0.03). There was a linear correlation (r=0.7) between SNVs and clonality, consistent with a gene-dose effect. ShRNA knockdown of TET2 in HeLa cells and CRISPR knockout in HEK293 cells resulted in a 24- and 30-fold increase in spontaneous mutations in the HPRT assay, respectively. This heightened mutability in TET2 knockout/knockdown cells was reversed by ectopic expression of TET2.

Co-immunoprecipitation of TET2 followed by tandem-affinity purification and LCMS/MS characterization in various systems (K562 and CMK, murine ESC and MEL cells) identified all major components of the DNA mismatch repair (MMR) complex as part of the TET2 interactome. MSH6, a key component of the MMR complex, was abundantly found to interact with TET2. This interaction was further confirmed by reverse co-immunoprecipitation using anti-MSH6 antibodies. Using truncating deletion constructs of TET2, the MSH6 binding site was mapped to the TET2 N-terminal amino acid segment 460-942. Loss of TET2 compromised the assembly of MMR, which is assessed by the co-IP assay on the MSH2 and MSH6 in TET2 knockout K562 and THP1. Moreover, examining MSH6 protein levels showed a 2-fold decrease (p=0.008) in TET2 knockout cells. The increased MSH6-associated ubiquitination in TET2 knockout cells, which was reversed by proteasome inhibitor MG132 and ectopic expression of N-terminal TET2, is necessary and sufficient to stabilize MSH6 and the MMR complex, implicating that the stability of MSH6-MMR complex is regulated by the non-catalytic N-terminal domain of TET2.

Consistent with our mechanistic findings, in vivo exposure to ENU or ionizing radiation led to the quicker evolution of MN in Tet2^-/- mice. WT and Tet2^-/- mice were treated with ENU (100 mg/kg) or vehicle showed significant differences in disease progression. Leukemia developed in 89.5% Tet2^-/- mice (47.4% AML and 16.8% with MDS/MPN), compared to 11.1% in WT mice, who mostly remained healthy (88.95) with only a small fraction developing lymphoid leukemia, MDS, or MPN. In the irradiation group, 100% of WT mice remained healthy, while 40% of the Tet2^-/- mice developed leukemia (25% AML and 15% other neoplasms). These findings suggest that genomic instability in TET2^MT mutant cells contributes to leukemia development. In summary, we conclude that TET2 directly binds to MSH6 and stabilize the MMR complex. Loss of TET2 impairs this process, leading to the accumulation of mutations progressive malignant evolution.