MLLT1/3 targeted protein degradation as the best-in-class approach for targeting the super elongation complex (SEC) in acute leukemia, including menin inhibitor refractory disease

Abstract Background: MLLT1 and 3 are highly homologous histone reader and scaffolding proteins, with sole known roles as critical regulators of the Super Elongation Complex (SEC), a key developmental transcription regulator. The SEC is frequently hijacked in cancer, especially acute leukemias, where it localizes via MLLT1 and 3 to the transcription start sites (TSSs) of major tumour drivers and oncogenes e.g. MYC, HOXA9 and MEIS1. Genetic studies have shown the SEC to be a key dependency in acute leukemias. Targeting the SEC has been clinically validated with menin inhibitors in KMT2Ar and mNPM1 acute leukemias. However, many patients are refractory to treatment and those that do respond rapidly progress. Consequently, a high unmet need remains. Unlike MLLT1 and 3, menin is not a constitutive component of the SEC, instead it acts in some SECs by binding directly to the rearranged KMT2A oncogenic driver fusion, stabilizing the SEC on specific TSSs. Given the unique and central role MLLT1 and 3 play in SEC biology, an MLLT1/3 targeted protein degrader (MLLT-TPD) that targets the histone reader and scaffolding properties could provide a best-in-class approach, with potential for broad and durable activity across multiple AML and ALL sub-populations. Methods: Degradation of MLLT1/3 was assessed by JESS Simple Western, Western blotting and flow cytometry. MLLT target gene expression was measured by qPCR. RNAseq and CUT&Tag analysis were performed to assess SEC-dependent biology and gene transcription. Viability was assessed by CellTiter-Glo assay. Compounds were assessed in stromal co-culture experiments using freshly collected AML cells from patients. In vivo activity was assessed in mouse xenograft models of AML. Results: We have developed a selective and potent first-in-class MLLT-TPD. This compound drives rapid and sustained degradation of MLLT1 in all cell lines tested (4h DC50 <1nM). This in turn leads to rapid and broad decreases in SEC dependent gene transcription, in contrast to menin inhibition, which has a slower and narrower impact. There are notable examples of genes where the MLLT-TPD but not menin inhibitor blocks transcription e.g. MYC. Gene set enrichment analysis of MLLT-TPD treated KMT2Ar cells shows rapid downregulation of a previously identified leukemia stem cell signature associated with KMT2Ar leukemias, as well as HOXA9/10/11 and MEIS1 signatures, and upregulation of mature myeloid differentiation signatures. This is consistent with the MLLT-TPD targeting a core leukemia sustaining oncogenic mechanism. Furthermore, in several non-KMT2Ar acute leukemias, menin inhibition has no impact on SEC gene transcription in stark contrast to MLLT degradation. Notably, in a RUNX1-RUNX1T1 cell line, MLLT-TPD but not the menin inhibitor downregulates SEC gene transcription including the tumor driver RUNX1-fusion transcript. Together, this supports the hypothesis that all SECs are dependent on MLLT but not menin. Against a broad panel of AML and ALL cell lines the MLLT-TPD is highly active in about 80% of the cell lines. Notably, a number of these MLLT-TPD sensitive lines were insensitive or markedly less sensitive to menin inhibition, including cell lines expressing the RUNX1 driver fusion. In addition, the MLLT-TPD retains full activity in KMT2Ar cell lines engineered to express clinically relevant menin inhibitor resistance mutations. Furthermore, in stromal co-culture experiments with fresh primary AML cells, the MLLT-TPD has a profound impact on cell viability and growth across a range of molecular subtypes, several of which do not respond to menin inhibition. Finally, we tested the MLLT-TPD in several mouse models of AML, including models which are refractory to menin inhibition. In all cases we observe marked anticancer activity from low oral, well tolerated, doses. Conclusion: Our in vitro and in vivo characterization of MLLT-TPD vs menin inhibition reveals that MLLT degradation has a broader activity than menin inhibition, including against AML and ALL cells insensitive to menin inhibitors, and AML cells that carry clinically relevant menin inhibitor resistance mutations. Together these data demonstrate that targeting MLLT1/3, the master regulator of the SEC, is a best-in-class approach with potential for broad, deep and durable clinical responses, including in patients that are resistant to menin inhibition.