Previous studies and clinical trials demonstrated that CDK4/6 inhibitors are more likely to be treated in a combination therapy rather than monotherapy [
19,
32‐
35]. Notably, CDK4/6 inhibitors are reported to exhibit good antitumor effects in combination with epigenetic regulatory drugs [
23]. Therefore, we screened the combination effects between ABE and 6 drugs targeting various regulators of epigenetic process, including JQ1 (target: BRD4), TAZ (EZH2), Vorinostat (HDAC), 5-azacitidine (DNMT), SETDB1-TTD-IN-1 (SETDB1), and UNC669 (L3MBTL1) (Fig.
2A). We performed a drug screening to identify targeted epigenetic inhibitors that potentiate response to CDK4/6 inhibitor in two GC cell lines to identify improved treatment strategies. The cells were treated 3 days against a 5-point dose range of the epigenetic compound combined with a 5-point dose concentration of CDK4/6 inhibitor and then tested for cell viability using the CellTiter-Glo Luminescent reagent. (Fig.
2A and
Materials and methods). We used Bliss model with SynergyFinder [
36] to calculate synergy score and reveal combinations that specifically induced higher cell death than single agents. Notably, the BRD4 inhibitor in the screen (JQ1) achieved the best synthetic lethal effect of ABE in two GC cell lines (synergy score, 24.67 in NUGC4 and 23.17 in NCI-N87), suggesting BRD4 as a candidate combination therapeutic target for ABE (Supplementary Fig. S
1A).
Pan-cancer analysis revealed that both CDK4 and BRD4 are super essential genes (Gene Effect score < -1) for gastric cancer cells (Fig.
2B) [
37]. BRD4 was screened out in more than 50% of gastric cancer cell lines (19/35). Furthermore, 14 out of 19 cell lines enriched for BRD4 are also enriched for genes related to the cell-cycle axis, which shows that BRD4 inhibition and cell cycle inhibitor are more likely to be associated with joint lethality (Supplementary Fig. S
1B). BRD4, a well-established reader for H3K27ac [
38] reported to promote gastric cancer progression and metastasis [
39], was elevated in gastric tumors (Fig.
2C and D) and was correlated to inferior prognoses of GC patients [
39]. To further explore the regulation relationship between CDK4 and BRD4, we performed Cleavage Under Target & Tagmentation sequencing (CUT&Tag-seq) for H3K27ac, a mark of active enhancers, in the presence and absence of ABE. As a result, more H3K27ac peaks were obtained in GC cells treated with ABE (Fig.
2E). Notably, the level of H3K27ac was elevated at BRD4 loci after CDK4/6 inhibition, implying an activation of BRD4 in response to ABE (Fig.
2G). Consistently, the protein level of BRD4 was significantly increased after ABE treatment (Fig.
2H and I). Further, over half of peaks in ABE-treated cells were located within distal intergenic regions or introns, suggesting that they might represent activated enhancers (Fig.
2F) [
40]. Given the vital role of BRD4 in super-enhancer (SE) organization [
41‐
43], we next want to define the expression-activated gene signature in ABE-treated GC cells. The enrichment analysis of these peaks in GO, KEGG and Reactome uncovered that cell junction and cell matrix adhesion regulation were significantly upregulated after ABE treatment (Fig.
2J). Notably, the level of H3K27ac was elevated at cell junction related gene loci including
SMAD3, VCL, RAC1 and
SNAI1 after CDK4/6 inhibition (Fig.
2K). Altogether, these results suggested that CDK4/6 inhibitors induce chromatin remodeling in gastric cancer cells characterized by the extensive enhancer activation and proposed that BRD4 was a promising combination target for ABE.