Direct inhibition of the NOTCH transcription factor complex

Abstract

Direct inhibition of transcription factor complexes remains a central challenge in the discipline of ligand discovery. In general, these proteins lack surface involutions suitable for high-affinity binding by small molecules. Here we report the design of synthetic, cell-permeable, stabilized alpha-helical peptides that target a critical protein-protein interface in the NOTCH transactivation complex. We demonstrate that direct, high-affinity binding of the hydrocarbon-stapled peptide SAHM1 prevents assembly of the active transcriptional complex. Inappropriate NOTCH activation is directly implicated in the pathogenesis of several disease states, including T-cell acute lymphoblastic leukaemia (T-ALL). The treatment of leukaemic cells with SAHM1 results in genome-wide suppression of NOTCH-activated genes. Direct antagonism of the NOTCH transcriptional program causes potent, NOTCH-specific anti-proliferative effects in cultured cells and in a mouse model of NOTCH1-driven T-ALL.

Figure 1. Design of MAML1-derived stapled peptides targeting NOTCH1-CSL

a, Structure of the NOTCH1 ternary complex (Protein Data Bank (PDB) accession 2F8X): CSL (tan), DNA (grey), dnMAML1 (green) and ICN1 (blue). The 16-amino-acid stretch of MAML1 targeting ICN1 and CSL is shown in red, and was used to design the stapled peptide SAHM1. b, Schematic of peptide stapling. A non-natural alkenyl amino acid (S₅) is incorporated at two positions in the peptide chain and then cross-linked by ring-closing olefin metathesis. c, Schematic, sequences and helical character of MAML1-derived SAHM peptides. βAla denotes a β-alanine spacer. Asterisks denote the location of S₅ residues, which are cross-linked in all SAHM peptides except SAHM1-UN. d, Circular dichroism spectroscopy of four MAML1 derived peptides illustrating the incremental effects of synthetic modification.

Figure 2. SAHM1 specifically engages the NOTCH1 transactivation complex

a, In vitro assembly of the NOTCH1 complex. Bead-immobilized RAMANK protein was incubated as indicated with CSL (0.5 µM), dnMAML1 (0.5 µM for lanes 3 and 5 (from left); 2.5 µM for lanes 4 and 6; 5 µM for lanes 7 and 8), and SAHM1 (10 µM). Bound proteins were washed, eluted and resolved by gel electrophoresis (Coomassie). b, Fluorescence polarization of FITC–SAHM peptides binding to RAMANK–CSL. c, Direct competition between unlabelled dnMAML1 and FITC–SAHM1. Concentrations of FITC–SAHM1 (15 nM) and RAMANK–CSL (0.6 µM) were held constant. dnMAML1 IC₅₀ = 3.9 ± 0.9 µM. d, CSL binding to immobilized RAMANK by SPR. Black curves represent sensogram data and the red curve denotes fit to a two-step kinetic model. Binding constants are shown. k_on, association rate; k_off, dissociation rate; RU, response units. e, f, Binding of RAMANK–CSL complexes to immobilized bioSAHM1 (e) and bioSAHM1-D1 (f). g, bioSAHM1 and bioSAHM1-D1 pull-down assays in KOPT-K1 lysates. Bound protein fractions were probed with antibodies specific for ICN1 (top) and CSL (bottom). h, Competitive co-immunoprecipitation of endogenous ICN1 by MAML1 in the presence of vehicle, SAHM1 (0.5, 1 and 10 µM from left to right) or SAHM1-D1 (10 µM). Unless noted otherwise data represent the mean ± s.d. (n = 3).

Figure 3. SAHM1 represses NOTCH1 target gene expression

a, Inhibition of a NOTCH1-dependent luciferase reporter by SAHM peptides. Signal was normalized to Renilla luciferase control. b, Dose-dependent effects of SAHM1 and SAHM1-D1 in the dual-luciferase assay, using threefold dilutions (0.55–45 µM) of ligand compared to vehicle alone. c, qRT–PCR analysis of the HES1, MYC and DTX1 mRNA levels in KOPT-K1 cells treated for 24 h with SAHM1 or SAHM1-D1 (20 µM) relative to dimethylsulphoxide (DMSO) control. d, qRT–PCR analysis of DTX1 mRNA levels in a panel of human T-ALL cell lines. e, Heat map representation of the top 50 downregulated genes (P < 0.001), induced by SAHM1 in KOPT-K1 and HPB-ALL cells. f, Quantitative comparison of genes downregulated by GSI (GSI-NOTCH gene set) with the SAHM1 gene expression profile in KOPT-K1 and HPB-ALL cells by GSEA. g, Comparison of all transcription factor target gene sets in the Molecular Signatures Database to the GSI-NOTCH gene set for enrichment in the SAHM1 expression profile by GSEA. Data are plotted as the family-wise error rate (FWER) P value versus the NES. GSI-NOTCH is marked as the most enriched gene set. h, GSEA of the second most enriched gene set (MYC/MAX), applied to the SAHM1 expression profile. Unless noted otherwise data represent the mean ± s.d. (n = 3). *P < 0.05, **P < 0.01.

Figure 4. SAHM1 reduces T-ALL proliferation and leukaemic initiation potential

a, Growth effects of SAHMs on a panel of human T-ALL cell lines of known mutational status. Cells were incubated with 15 µM DAPT (blue triangles), SAHM1 (red squares), SAHM1-D1 (green diamonds) or DMSO (black circles) and monitored for proliferation after 3 and 6 days of culture. Data points are mean ± s.d. (n = 3). b, Effects of SAHMs on apoptosis of T-ALL cells monitored using Capase-glo 3/7 (Promega) in cultures carried out as in a. Error bars, s.d. c, d, Ex-vivo treatment of L1601PΔP cells with SAHM1 (5 µM, 12 h) limits leukaemia initiation in secondary murine recipients. Reduction of spleen weight in the SAHM1 cohort (n = 6) compared to vehicle (n = 6) at the first sign of disease toxicity (23 days) (P = 0.001) (c). Circulating GFP-positive cell count in the blood is reduced by ~100-fold in the SAHM1 cohort (n = 6) relative to vehicle (n = 6) (P = 0.0026) (d). Error bars, s.d. Statistical analyses performed with a two-tailed t-test. *P < 0.01, **P < 0.005. e, An immunohistochemical stain for GFP that imparts a brown colour shows extensive leukaemic infiltration of bone marrow (BM) and spleen (SPL) in representative mice receiving vehicle-treated transplants relative to SAHM1. SPL, spleen. Scale bars,

Figure 5. SAHM1 treatment inhibits NOTCH signalling and leukaemic progression in vivo

a, Flow cytometric analysis of cells isolated from C57BL/6-Tyr^C/C mice reconstituted with luciferase-expressing haematopoietic stem cells transduced with an L1601PΔP NOTCH1 allele or empty vector (MIG). BL, blood. b, Bioluminescence imaging of primary recipients of Luc-L1601PΔP cells or control cells 2 months postreconstitution. c, Bioluminescence quantification of tumour burden in mice with established disease treated with vehicle, 35 mg kg⁻¹ SAHM1 per day (QD; P = 0.17), or 30 mg kg⁻¹ SAHM1 twice daily (BID; P = 0.02). A Fisher’s Exact test was used to compare disease progression between the cohorts. d, qRT–PCR analysis reveals repression of Hesl (P = 0.0187), Myc (P = 0.023), Nrarp (P = 0.001), Heyl (P = 0.0006) and Dtxl (P = 0.0006) mRNA levels in blood collected at day five from vehicle (n = 3) and SAHM1 (30 mgkg⁻¹ BID, n = 3)-treated mice. e, Enrichment of GSI and dnMAML1 downregulated transcripts in the gene expression profile of leukocytes isolated from SAHM1-treated mice. Data in d represent the mean ± s.d. of triplicate measurements. Unless otherwise noted statistical analyses were performed with a two-tailed t-test. *P < 0.05, **P < 0.005.