Peptidomimetic blockade of MYB in acute myeloid leukemia

Abstract

Aberrant gene expression is a hallmark of acute leukemias. MYB-driven transcriptional coactivation with CREB-binding protein (CBP)/P300 is required for acute lymphoblastic and myeloid leukemias, including refractory MLL-rearranged leukemias. Using structure-guided molecular design, we developed a peptidomimetic inhibitor MYBMIM that interferes with the assembly of the molecular MYB:CBP/P300 complex and rapidly accumulates in the nuclei of AML cells. Treatment of AML cells with MYBMIM led to the dissociation of the MYB:CBP/P300 complex in cells, its displacement from oncogenic enhancers enriched for MYB binding sites, and downregulation of MYB-dependent gene expression, including of MYC and BCL2 oncogenes. AML cells underwent mitochondrial apoptosis in response to MYBMIM, which was partially rescued by ectopic expression of BCL2. MYBMIM impeded leukemia growth and extended survival of immunodeficient mice engrafted with primary patient-derived MLL-rearranged leukemia cells. These findings elucidate the dependence of human AML on aberrant transcriptional coactivation, and establish a pharmacologic approach for its therapeutic blockade.

Fig. 1

MYBMIM disrupts the MYB:CBP complex in AML cells. a Molecular structure of the complex of the transactivation domain of MYB (blue) with the KIX domain of CBP (gray) assembled in Maestro (Schrödinger). MYB residues making contacts with CBP are labeled as indicated. (PDB: 1SB0). b Molecular structures of the transactivation domain of MYB (left, green) and MYBMIM (right) in complex the KIX domain of CBP (gray), as modeled using replica exchange molecular dynamics. Both MYBMIM and MYB retain E308 and R294 salt bridge and L302 hydrophobic interactions, as marked by sidechain representation. c Binding of FITC-conjugated MYB (green), MYBMIM (red), compared to control TG1, TG2, TG3, and TAT (black), as measured using microscale thermophoresis (K_d = 4.2 ± 0.5 µM and 21.3 ± 2.9 µM for MYB and MYBMIM, respectively, 59.2 ± 12.4 µM for TG1, 75.1 ± 12.5 µM for TG2 and 113.5 ± 36.6 µM for TG3). Error bars represent standard error mean of three biological replicates. d Live cell confocal fluorescence microscopy photographs of MV-411 cells treated with 50 nM FITC-MYBMIM (green) for 1 h, as visualized using Mitotracker (red) and Hoechst 33342 (blue). Scale bar indicates 20 μm, with z-stack of 1.5 μm. e Western blot showing comparable binding of cellular CBP/P300 to streptavidin bead-immobilized BIO-MYBMIM, specifically competed by 20-fold excess free retro-inverso TAT (RI-TAT) and MYBMIM peptides, as indicated by + signs. f Representative western blot of MYB:CBP/P300 complex immunoprecipitated from MV411 cells disrupted 20 μM MYBMIM and TG3, as indicated

Fig. 2

MYBMIM regulates MYB enhancers and promoters and MYB-dependent target genes. a Heatmap of changes in normalized gene expression of MOLM13 cells treated with 20 µM MYBMIM versus TG3 control for 6 h, as analyzed by RNA-seq of three biological replicates. b Volcano plot of normalized gene expression, with BCL2, IKZF1, MYC and MTL5 as indicated. c, d Gene set enrichment analysis of downregulated c and upregulated d genes with respect to MYB target genes, as defined by. NES = −2.47 and 2.09, and q = 0 and 0, respectively

Fig. 3

MYBMIM suppresses the assembly of chromatin complexes. a Volcano plot of MYB occupancy in MV-411 cells treated with 20 µM MYBMIM versus TG3 control for 6 h, as analyzed by MYB ChIP-seq. p-values denote t-test statistical significance of 3 biological replicates. b Heatmap of changes in H3K27Ac occupancy of MV411 treated with 20 µM MYBMIM versus TG3 control for 24 h, as analyzed by ChIP-seq of three biological replicates. c Genome track of the BCL2 locus showing elimination of the MYB-bound enhancer (star) upon treatment with MYBMIM, but not control or TG3 treatment. d Genome track of the BCL2 locus showing elimination of the H3K27Ac-bound enhancer (star) upon treatment with MYBMIM, but not control or TG3 treatment. e Magnified boxed area of MYB-bound enhancer peak shown in 3c. f Magnified boxed area of H3K27Ac-bound enhancer peak shown in (d). g Analysis of relative enrichment of H3K27Ac at the BCL2 enhancer locus compared to NANOG, as measured by ChIP-PCR upon treatment with control PBS (black), 20 μM TG3 (blue), and 20 μM MYBMIM (red) for 24 h. Error bars represent standard deviations of three biological replicates. * p = 8.6e–3 when compared to untreated control

Fig. 4

MYBMIM induces apoptosis and downregulates MYB-regulated genes. a Viability of MOLM-13, MV-411, ML2, and HL60 cells, treated for 6 days with control PBS (black), 20 μM TG1, TG2, or TG3 (blue), and 10 μM MYBMIM (orange) and 20 μM MYBMIM (red), with peptide replacement every 48 h. Error bars represent standard deviations of three biological replicates. *p = 7.2e–7; **p = 9e–6; ^p = 1.7e–6; ^^p = 3.3e–6 when compared to untreated control. b Representative photographs of Giemsa-stained MV-411 cells after 6 h treatment, as indicated. Scale bar corresponds to 20 μm. c Flow cytometry analysis of apoptosis of MV-411 cells upon peptide treatment at 20 μM for 24 h, as indicated. Numbers denote percentage of cells that are both Annexin V and DAPI positive. d Analysis of BCL2 and MYC mRNA expression in MV411 cells as measured by qRT-PCR, upon treatment with control PBS (black), 20 μM TG3 (blue), and 20 μM MYBMIM (red) for 6 h. Error bars represent standard deviations of three biological replicates. *p = 0.0046; **p = 0.008 when compared to untreated control. e MV-411 cells expressing MSCV-IRES-GFP (MIG) BCL2 but not empty MIG or wild-type cells are protected from treatment with 20 μM MYBMIM (red) as compared to control PBS (black) and 20 μM TG3 (blue) peptides. Error bars represent standard deviations of three biological replicates. *p = 1.4e–5; **p = 4.2e–7; ^p = 0.0005 MYBMIM treatment compared to respective untreated controls

Fig. 5

MYBMIM exhibits anti-leukemia efficacy in vivo. a Activity of burst forming units-erythroid (BFU-E, red) and colony forming units-granulocyte/monocyte (CFU-GM, black) of CD34⁺ human umbilical cord progenitor cells treated with control PBS, or 20 μM TG3 or MYBMIM for 14 days. Error bars represent the standard deviation of 3 biologic replicates. b Representative phase photographs of CFU-GM colonies treated as indicated. c–f Peripheral blood count analysis of C57BL/6 J mice treated for 7 days with MYBMIM (25 mg/kg IP daily), as compared to control PBS. Bars indicate the mean and standard deviation of individual mice. g Plasma concentration of BIO-MYBMIM after one-time IP injection of 25 mg/kg in C57BL/6 J mice. Plasma was collected at 30 min, 2 h, 4 h, and 24 h post-injection and the concentration of BIO-MYBMIM was determined by spectrophotometric avidin assay. h Representative fluorescent micrographs of human-specific CD45 staining (green) and DAPI staining (blue) in femur sections of NSG mice engrafted with primary patient-derived MLL-rearranged leukemia cells and treated with MYBMIM (25 mg/kg IP daily) as compared to control PBS for 21 days upon development of peripheral leukemia, quantified in (i). Error bars represent standard deviation of 6 individual mice. *p = 1.2e–4, log-transformed t-test. j Images of fluorescent micrographs of human BCL2 (green) and DAPI (blue), quantified in (k). Error bars represent standard deviation of 6 individual mice. *p = 0.3, log-transformed t-test. l Kaplan–Meier survival analysis of NSG mice engrafted with primary patient-derived MLL-rearranged leukemia cells and treated 3 days post transplantation with MYBMIM (red, 25 mg/kg IP twice daily) as compared to control PBS (black) for 14 days. n = 15 mice per group. p = 0.0038, log-rank test. m Kaplan–Meier survival analysis of NSG mice serially transplanted with bone marrow collected from moribund mice engrafted with primary patient-derived MLL-rearranged leukemia cells treated with MYBMIM for 14 days. n = 10 mice per group. p < 0.0001, log-rank test