Discovery of an MLLT1/3 YEATS Domain Chemical Probe

Abstract

YEATS domain (YD) containing proteins are an emerging class of epigenetic targets in drug discovery. Dysregulation of these modified lysine-binding proteins has been linked to the onset and progression of cancers. We herein report the discovery and characterisation of the first small-molecule chemical probe, SGC-iMLLT, for the YD of MLLT1 (ENL/YEATS1) and MLLT3 (AF9/YEATS3). SGC-iMLLT is a potent and selective inhibitor of MLLT1/3-histone interactions. Excellent selectivity over other human YD proteins (YEATS2/4) and bromodomains was observed. Furthermore, our probe displays cellular target engagement of MLLT1 and MLLT3. The first small-molecule X-ray co-crystal structures with the MLLT1 YD are also reported. This first-in-class probe molecule can be used to understand MLLT1/3-associated biology and the therapeutic potential of small-molecule YD inhibitors.

Keywords: MLLT1; MLLT3; YEATS; chemical probes; epigenetics.

Figure 1

Identified hit MLLT1/3 inhibitor compound 1.

Figure 2

A) Co‐crystal structure of MLLT1 YD with H3Kac27 peptide bound (PDB ID 5J9S). B) Docking studies of compound 1 bound to MLLT1. YD from the co‐crystal structure with H3Kac27 forms similar hydrogen bonds with the backbone NH moieties of Y78 and A79, structural water HOH312, the S58 side chain, and π‐π‐π stacking interactions between the side chains of F28, F59, Y78, and H56.

Figure 3

Synthetic disconnection of hit compound 1 into simple building blocks for the rapid construction of potential MLLT1/3 YD inhibitors.

Scheme 1

Synthesis of MLLT1/3 benzimidazole based inhibitors. Reagents and conditions: a) ethyl 2‐chloroacetate, ethyl 2‐chloropropanoate (1.2 equiv), 4 n HCl (0.6 m), 16 h, 100 °C, quant.; b) amine (1.2–1.5 equiv), Na₂CO₃ (1.5 equiv), 23 °C, 3–82 %; c) H₂ Pd/C (10 %), MeOH, RT, 16 h, 17–88 %; d) sulfonyl/acid chloride (1.2 equiv), PS‐DIPEA (2 equiv), CH₂Cl₂ (0.2 m), 16 h, 11–100 %; e) acid (1.2 equiv), PS‐DIPEA (2 equiv), 16 h, 6–100 %.

Figure 4

A) Detailed interactions of compound 92 (orange sticks) with MLLT1 YD (loop 1 in cyan, loop 4 in magenta, loop 6 in yellow; PDB ID 6HT1). B) Detailed interactions of compound 92 (orange sticks) with MLLT1 YD (cyan sticks) overlaid with a co‐crystal structure of Kcr (yellow sticks):MLLT3 YD (yellow sticks) (PDB ID 5HJB).

Figure 5

A) Western blot analysis of dose‐dependent thermal stabilization of endogenous MLLT1 in MV4;11 cells with compounds 92 and 91. B) Stabilisation of MLLT1 induced by 92 and 91 normalised to DMSO. C) FRAP assay showing the half‐life recovery time in U2OS cells transfected with GFP‐tagged MLLT1 wild‐type (WT) and mutant (MUT) and D) MLLT3 WT after treatment with compound 92 or 91 (24 h) in the presence of 2.5 μm SAHA. Graphs represent n=3 biological replicates, with n≥10 cells per treatment group. Mean±SD, one‐way ANOVA with Tukey–Kramer correction for multiple comparisons **** P<0.0001, ** P=0.0025. E) NanoBRET dose response for 91 and 92 after 24 h treatment using N‐terminal nanoLuc‐MLLT3 and C‐terminal HaloTag‐H3.3 in HEK293 cells, in the presence of 2.5 μm SAHA. ▪ compound 92, ▾ compound 91, • DMSO, ○ DMSO+SAHA. Graphs represent n=6 biological replicates, with n>4 technical replicates. Mean±SEM, mBU: BRET units. F) qPCR for MYC, MPO, DDN, CTSG, and CD86 expression in MV4;11 cell line after 72 h incubation with compound 91 (1 μm), 92 (1 μm), or JQ1 (50 nm, positive control for MYC downregulation). Graph represents n=3 biological replicates with n=2 technical replicates. Mean±SEM; a two‐way ANOVA with post‐hoc Dunnett's multiple comparisons test was used for statistical analyses compared to DMSO (ns not significant, * P<0.05, ** P<0.005, *** P<0.001, **** P<0.0001).