Development And Characterization of Irreversible NSD1 And NSD3 Histone Methyltransferase Inhibitors

Abstract

The Nuclear SET domain containing (NSD) family of histone methyltransferases plays a role in the regulation of gene expression by catalyzing methylation of histone H3 at lysine 36 (H3K36). The three large, multidomain members of this family are implicated in various forms of cancer. NSD1 is part of a translocation that results in the fusion protein NUP98-NSD1 found in aggressive, rare cases of pediatric leukemia. The histone methyltransferase activity of this fusion is necessary for leukemic transformation. NSD3 is overexpressed in multiple solid cancers including breast, lung, and head and neck cancers. Overexpression of NSD3 is associated with poor prognosis in breast cancer, and expression of the catalytic isoform can rescue breast cancer tumor growth. The importance in cancer and complex structure of NSD1 and NSD3 create a need for small molecule inhibitors of the histone methyltransferase activity to parse the role of NSD family mediated H3K36 methylation in cancer. We have developed three series of covalent inhibitors with varying degrees of selectivity between NSD1 and NSD3. Our NSD1 inhibitors are developed from a fragment screening hit and target a cysteine in the autoinhibitory loop. The lead compound from this series, BT5,selectively inhibits the NSD family of histone methyltransferases over other histone methyltransferases. BT5 binds to the NSD1 SET domain in cells and has antiproliferative activity against NUP98-NSD1 containing leukemia cells. Additionally, BT5 inhibits colony formation and NUP98-NSD1 target gene expression in a patient sample of a NUP98-NSD1 leukemia. BT5 was an excellent starting point for developing NSD family member inhibitors, and we were able to develop a series of NSD1/NSD3 inhibitors after an extensive medicinal chemistry campaign. Co-crystal structures of NSD1 and NSD3 in complex with a lead compound from this series, SZ881, confirm that SZ881 covalently engages with a conserved zinc binding cysteine in both proteins. This engagement confers potent engagement to NSD1/3 and low micro-molar inhibition of NSD1/3. SZ881 also exhibits anti-proliferative activity against breast cancer cells, and SZ1817, an analog of SZ881, pulls down NSD3 from breast cancer cells. This pull-down validates the on-target binding of SZ1817 in breast cancer cells. From SZ881, we used structure-based design to develop NSD3-selective inhibitors that covalently engages with a non-conserved cysteine in NSD3. The lead inhibitor from this series, SLN479, binds potently and selectively to NSD3. SLN479 potently inhibits NSD3 histone methyltransferase activity and binds to the NSD3 SET domain in cells. Additionally, SLN479 has anti-proliferative activity against a NSD3-dependent breast cancer cell line. In conclusion, we have developed three series of covalent inhibitors with varying degrees of selectivity between NSD1 and NSD3. Compounds from each series exhibit potent anti-proliferative activity against cancer cells in vitro, illustrating the power of using covalent compounds to inhibit challenging therapeutic targets. These inhibitors represent first-in-class covalent inhibitors of NSD1/3 and will facilitate the design of optimized inhibitors and investigation into the role of NSD1/3 in a variety of cancers.