SMARCA4 loss is synthetic lethal with CDK4/6 inhibition in non-small cell lung cancer

Abstract

Tumor suppressor SMARCA4 (BRG1), a key SWI/SNF chromatin remodeling gene, is frequently inactivated in cancers and is not directly druggable. We recently uncovered that SMARCA4 loss in an ovarian cancer subtype causes cyclin D1 deficiency leading to susceptibility to CDK4/6 inhibition. Here, we show that this vulnerability is conserved in non-small cell lung cancer (NSCLC), where SMARCA4 loss also results in reduced cyclin D1 expression and selective sensitivity to CDK4/6 inhibitors. In addition, SMARCA2, another SWI/SNF subunit lost in a subset of NSCLCs, also regulates cyclin D1 and drug response when SMARCA4 is absent. Mechanistically, SMARCA4/2 loss reduces cyclin D1 expression by a combination of restricting CCND1 chromatin accessibility and suppressing c-Jun, a transcription activator of CCND1. Furthermore, SMARCA4 loss is synthetic lethal with CDK4/6 inhibition both in vitro and in vivo, suggesting that FDA-approved CDK4/6 inhibitors could be effective to treat this significant subgroup of NSCLCs.

Fig. 1

Reduced cyclin D1 in SMARCA4-deficient non-small cell lung cancer (NSCLC) cells causessensitivities to cyclin-dependent kinase 4/6 (CDK4/6) inhibitors. a, b SMARCA4-deficient NSCLC cell lines express reduced cyclin D1 levels. Western blot analysis for the indicated proteins (a) and CCND1 messenger RNA (mRNA) expression (b) of a panel of NSCLC cell lines. HSP90 was used as a loading control. Relative CCND1 mRNA expression (relative to GAPDH) was measured by real-time quantitative reverse transcription PCR (RT-qPCR). A4: SMARCA4, A4/2: SMARCA4/2, Pro: proficient, Def: deficient, K: KRAS mutation. Empty triangles indicate RB-deficient cell lines. Turquoise color indicates cell lines with KRAS mutation. Error bars: mean ± standard deviation (s.d.) of biological replicates (n = 3); two-tailed t test, *p < 0.05. c, d SMARCA4-deficient NSCLC cells are highly sensitive to palbociclib treatment, similar to KRAS mutation cells. c Half-maximal inhibitory concentration (IC₅₀) of palbociclib in the above cell line panel was determined by measuring cell viability using CellTiter-Blue assay. Error bars: mean ± s.d. of biological replicates (n = 4); two-tailed t test, *p < 0.05, **p < 0.01. d Colony formation assays of the representative cell lines. Cells were cultured in the absence or presence of palbociclib at the indicated concentrations for 10–14 days. For each cell line, all dishes were fixed at the same time. e, f Palbociclib treatment in SMARCA4-deficient NSCLC cells induces strong G1 cell cycle arrest. H1299 (e) and H1703 (f) cells treated with palbociclib for 24 h were fixed, stained with propidium iodide and analyzed by flow cytometry using the Guava easyCyte HT System. g, h Ectopic expression of cyclin D1 confers drug resistance to palbociclib in H1299 (g) and H1703 (h) cells. Upper, colony formation assays; lower, immunoblot of cells with stable ectopic expression of GFP or CCND1 and treated with palbociclib (H1299, 300 nM; H1703, 33 nM). i, j Cyclin D1 knockdown sensitizes HCC827 (i) and PC9 (j) cells to palbociclib. Upper, colony formation assays in the absence or presence of 300 nM palbociclib; lower, immunoblot of cells expressing pLKO control or short hairpin RNAs (shRNAs) targeting CCND1

Fig. 2

Palbociclib is effective against SMARCA4-deficient non-small cell lung cancer (NSCLC) tumor growth in vivo. Palbociclib inhibits tumor growth in xenograft models of H1299 (a, b, e, f) and H1703 (c, d, g, h). a, c Tumor size from day 0 of treatment in H1299 (a, n = 4 per group) and H1703 (c, n = 8 for vehicle, n = 7 for palbociclib; 150 mg kg⁻¹) models. Error bars represent mean ± standard error of mean (s.e.m.); two-way analysis of variance (ANOVA), ****p < 0.0001. b, d Final tumor weight measured after surgery in H1299 (b) and H1703 (d) models. Two-tailed t-test, **p < 0.01, ****p < 0.0001. e–h Palbociclib treatment resulted in suppression of RB phosphorylation, Ki67 expression and mitotic index in xenograft tumors of the trial endpoints. Representative images of Immunohistochemistry (IHC) (p-RB, Ki67) and hematoxylin and eosin (H&E) analysis of H1299 (e) and H1703 (g) xenograft tumor tissues. Bar 50 µm; black arrows point to mitotic active cells as examples. f, h Quantifications of p-RB, Ki67 and mitotic count of H1299 (f, n = 3) and H1703 (h, n = 4). Two-tailed t-test, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 3

SMARCA4/2 loss causes reduced cyclin D1 expression in non-small cell lung cancer (NSCLC). a–d SMARCA4/2 regulate cyclin D1 expression in NSCLC. a, b SMARCA4 restoration upregulates cyclin D1 protein (left) and messenger RNA (mRNA) (right) expression in H1299 (a) and H1703 (b) cells. c SMARCA2 knockdown in H1299 cells suppresses cyclin D1 protein (left) and mRNA (right) expression. d SMARCA2 restoration in H1703 cells elevates cyclin D1 protein (left) and mRNA (right) expression. Relative CCND1 mRNA expression (relative to GAPDH) was measured by real-time quantitative reverse transcription PCR (RT-qPCR). Error bars: mean ± s.d. of biological replicates (n = 3, two-tailed t-test, *p < 0.05, **p < 0.01, ***p < 0.001). e, f Correlation of CCND1 and SMARCA4 mRNA expression in two cohorts of lung adenocarcinomas (LUADs) from BC Cancer Agency (BCCA; n = 83, e) and The Cancer Genome Atlas (TCGA; n = 230, f). g Correlation of CCND1 and SMARCA2 mRNA expression in SMARCA4-mutated LUADs (n = 13) in the TCGA cohort. r, Pearson's correlation coefficient. h–j Immunohistochemistry (IHC) analysis of cyclin D1 protein expression in two cohorts of NSCLC patient tumors: NCT (n = 93; h, i) and McGill University Health Center (MUHC; n = 91; j). Representative IHC images of SMARCA4 IHC-negative tumors are shown; a SMARCA4/2 IHC-positive tumor served as staining control (h). Cyclin D1 in IHC analysis was quantified with H-score and analyzed by Wilcoxon rank sum test (i, j)

Fig. 4

Extensive opening of regulatory elements by induction of SMARCA4/2. a Venn diagram showing overlap of open chromatin sites in control infected H1703 cells with or without SMARCA4 overexpression. Note the dramatic increase in open chromatin sites upon SMARCA4 overexpression. b Distribution of SMARCA4-dependent and -independent open chromatin sites relative to nearest gene transcriptional start site. Note that SMARCA4-dependent sites are much less likely to be promoters (within 1 kb of transcription start site (TSS)). c, d Metaplot (c) and heatmap (d) of assay for transposase-accessible chromatin sequencing (ATAC-seq) read data from control-infected, SMARCA4-infected and SMARCA2-infected cells over the 62,878 SMARCA4-dependent ATAC peaks. Note similar effect of SMARCA2 and SMARCA4. e, f Metaplot (e) and heatmap (f) of H3K27Ac chromatin immunoprecipitation (ChIP) data from control-transfected, SMARCA4-infected and SMARCA2-infected cells over the 62,878 SMARCA4-dependent ATAC peaks

Fig. 5

SMARCA4/2 regulate CCND1 via controlling chromatin accessibility and upregulating JUN. a Assay for transposase-accessible chromatin sequencing (ATAC-seq) and chromatin immunoprecipitation sequencing (ChIP-seq) data in vicinity of the CCND1 locus indicate enhanced chromatin accessibility upon SMARCA4/2 restoration. Note SMARCA4 at CCND1 promoter and formation of new putative enhancer ~50 kb upstream of CCND1 promoter. All data were generated in H1703 cells before and after restoration of SMARCA4 or SMARCA2, except the publicly available SMARCA4 ChIP data in H1299 cells expressing doxycycline (Dox)-inducible SMARCA4. Track height is normalized to relative number of mapped reads. b Zoomed-in view of the putative CCND1 enhancer region. Shown are ATAC-seq peaks in H1703 cells before and after SMARCA4/2 restoration and the publicly available c-Fos/c-Jun ChIP data of endothelial cell line, human umbilical vein endothelial cell (HUVEC) (GSM935585, GSM935278). Location of canonical adaptor protein-1 (AP-1) motifs are indicated. c ATAC and ChIP-seq data in vicinity of JUN locus as described in a. Note SMARCA4 at JUN promoter and extensive opening of nearby putative enhancers. d–i Restoration of SMARCA4 in H1703 (d, e) and H1299 (f, g) or SMARCA2 restoration in H1703 (h, i) cells upregulate c-Jun messenger RNA (mRNA) (d, f, h) and protein (e, g, i). j, k Knockdown of JUN partially abrogated SMARCA4-mediated induction of cyclin D1 mRNA (j) and protein (k) expression in H1703 cells. l Proposed model showing that SMARCA4 directly regulates CCND1 and also upregulates JUN which positively regulates CCND1. Two-tailed t-test. Error bars represent mean ± s.d., ***p < 0.001, ****p < 0.0001

Fig. 6

SMARCA4 loss is synthetic lethal with cyclin-dependent kinase 4/6 (CDK4/6) inhibition in non-small cell lung cancer (NSCLC). a, b SMARCA4 restoration in SMARCA4-deficient cell lines confers drug resistance to palbociclib. Colony formation assays of H1299 (a) and H1703 (b) cells expressing vector control or SMARCA4 and treated with palbociclib (H1299, 300 nM; H1703, 100 nM). c SMARCA2 knockdown in SMARCA4-deficient H1299 cells sensitizes cells to palbociclib treatment. Colony formation assay of H1299 cells expressing pLKO control or SMARCA2 short hairpin RNAs (shRNAs) and treated with palbociclib. d SMARCA2 restoration in SMARCA4/2-dual deficient cells H1703 confers resistance to palbociclib. Colony formation assay of H1703 cells expressing vector control or SMARCA2 and treated with palbociclib. e–g Resistance to palbociclib after restoration of SMARCA4 is also observed in mouse xenograft models using an isogenic cell pair of H1299 cells expressing vector control or SMARCA4. e Tumor volume evolution during the course of the experiment in H1299 xenograft models expressing vector control (left) or SMARCA4 (right). f Tumor volume fold change during the establishment phase (left) and during palbociclib treatment (right) in the same models. g Immunohistochemistry (IHC) analysis of SMARCA4 in the representative endpoint tumors of H1703 control or SMARCA4-restored from above. Bar 50 µm. Error bars represent mean ± standard error of mean (s.e.m.); two-way analysis of variance (ANOVA), ****p < 0.0001