Comparative analysis of drug-like EP300/CREBBP acetyltransferase inhibitors

Abstract

The human acetyltransferase paralogs EP300 and CREBBP are master regulators of lysine acetylation whose activity has been implicated in various cancers. In the half-decade since the first drug-like inhibitors of these proteins were reported, three unique molecular scaffolds have taken precedent: an indane spiro-oxazolidinedione (A-485), a spiro-hydantoin (iP300w), and an aminopyridine (CPI-1612). Despite increasing use of these molecules to study lysine acetylation, the dearth of data regarding their relative biochemical and biological potencies makes their application as chemical probes a challenge. To address this gap, here we present a comparative study of drug-like EP300/CREBBP acetyltransferase inhibitors. First, we determine the biochemical and biological potencies of A-485, iP300w, and CPI-1612, highlighting the increased potency of the latter two compounds at physiological acetyl-CoA concentrations. Cellular evaluation shows that inhibition of histone acetylation and cell growth closely aligns with the biochemical potencies of these molecules, consistent with an on-target mechanism. Finally, we demonstrate the utility of comparative pharmacology by using it to investigate the hypothesis that increased CoA synthesis caused by knockout of PANK4 can competitively antagonize binding of EP300/CREBBP inhibitors and demonstrate proof-of-concept photorelease of a potent inhibitor molecule. Overall, our study demonstrates how knowledge of relative inhibitor potency can guide the study of EP300/CREBBP-dependent mechanisms and suggests new approaches to target delivery, thus broadening the therapeutic window of these preclinical epigenetic drug candidates.

Figure 1.

(a) Domain architecture of EP300 and CREBBP transcriptional coactivators. (b) Structure and primary reference for drug-like acetyltransferase inhibitors 1-3 comparatively assessed in this study.

Figure 2.

(a) TR-FRET assay for EP300-catalyzed acetylation of a histone H3 peptide. (b) Biochemical inhibition of EP300 by drug-like acetyltransferase inhibitors 1-3. (c) Tabulated half-maximal inhibition concentrations for inhibitors. Fold-change represents the IC₅₀ at high acetyl-CoA concentrations relative to the IC₅₀ at high acetyl-CoA concentrations. Values represent the average of n=2 technical replicates.

Figure 3.

(a) Schematic for cell-based analysis of 1-3. (b) Western blot analysis of histones extracted from MCF-7 cells after treatment with 1-3. The left lane of each gel corresponds to vehicle-treated control, followed by cells incubated with five-fold increasing concentrations of 1, 2, or 3 (8, 40, 200, 1000, 5000 nM). Exposure times are the same within each histone acetylation mark. (c) Heat-map depiction of growth caused by treatment of cells with 1-3 in the NCI-60 cell line screen. (d) Cross correlation analysis of NCI-60 inhibition by 1-3 and bromodomain inhibitors (e) Sequence of histone H3 N-terminus indicating sites of acetylation and the KAT responsible for modification. EP300/CREBBP-regulated K18Ac is colored in red. (f) Quantitative LC-MS analysis of histone H3 acetylation. Values correspond to the normalized intensities observed for histone peptides containing the specified modifications isolated from cells treated with 3, with signal for untreated cells set equal to 100%. The intensity of singly modified peptides was used, in all cases with the exception of K14Ac, whose value was derived from the summed intensity of singly and doubly modified K14Ac peptides (2^nd modification: K9me1, K9me2, or K9me3). Values represent the average of n=3 biological replicates.

Figure 4.

(a) PANK4 inhibits the production of acetyl-CoA from vitamin B5 by consuming a metabolic intermediate (4’-phosphopantetheine, not shown). Knockout of PANK4 can increase acetyl-CoA levels and cause resistance to inhibitors of EP300. (b) Rationale for how inhibitors with distinct potencies can be used to differentiate direct and indirect acetyl-CoA-dependent resistance mechanisms. (c) Dose-response curves of MCF-10A cells (PANK4 KO+WT and PANK4 KO+EV) grown in the presence of 1 or 3. (d) Relative proliferation of MCF-10A cells (PANK4 KO+WT and KO+EV) grown in the presence of 1 μM 1 or 3. (e) Effects of 3 h treatment of 1 and 3 ((8, 40, 200, 1000, 5000 nM) on H3K18Ac and (f) H3K23Ac in MCF-10A cells (PANK4 KO+WT and PANK4 KO+EV). Exposure times are the same within each histone acetylation mark.

Figure 5.

(a) Photocaged CPI-1612 analogue 4 and active inhibitor 5. (b) Rationale for caging CPI-1612 analogues via derivatization of the secondary amine. (c) Relative EP300 biochemical inhibitor potency of 4 and 5 as assessed by TR-FRET histone acetylation assay (n=2 technical replicates). (d) Irradiation of 4 with UV light (302 nm) produces parent compound 5 in vitro. (e) Schematic for cellular photouncaging studies. (f) Light-dependent inhibition of EP300/CREBBP-dependent histone acetylation in living MCF-7 cells.