SMARCA4/2 loss inhibits chemotherapy-induced apoptosis by restricting IP3R3-mediated Ca2+ flux to mitochondria

Abstract

Inactivating mutations in SMARCA4 and concurrent epigenetic silencing of SMARCA2 characterize subsets of ovarian and lung cancers. Concomitant loss of these key subunits of SWI/SNF chromatin remodeling complexes in both cancers is associated with chemotherapy resistance and poor prognosis. Here, we discover that SMARCA4/2 loss inhibits chemotherapy-induced apoptosis through disrupting intracellular organelle calcium ion (Ca²⁺) release in these cancers. By restricting chromatin accessibility to ITPR3, encoding Ca²⁺ channel IP3R3, SMARCA4/2 deficiency causes reduced IP3R3 expression leading to impaired Ca²⁺ transfer from the endoplasmic reticulum to mitochondria required for apoptosis induction. Reactivation of SMARCA2 by a histone deacetylase inhibitor rescues IP3R3 expression and enhances cisplatin response in SMARCA4/2-deficient cancer cells both in vitro and in vivo. Our findings elucidate the contribution of SMARCA4/2 to Ca²⁺-dependent apoptosis induction, which may be exploited to enhance chemotherapy response in SMARCA4/2-deficient cancers.

Fig. 1. SMARCA4/2 loss causes resistance to chemotherapeutics in ovarian and lung cancers.

The half maximal inhibitory concentration (IC₅₀) of cisplatin in pan cancer (a) and lung cancer (b) cell lines with differential mRNA expression for SMARCA4 and SMARCA2 (see Supplementary Fig. 2b for stratification). A4^H: SMARCA4^High; A4^L: SMARCA4^Low; A2^H: SMARCA2^High; A2^L: SMARCA2^Low. Cell line numbers are indicated in gray below each group. One-way ANOVA Kruskal–Wallis test followed by Dunn’s test for multiple comparisons to A4^HA2^H group, p values (p): a A4^HA2^L—0.5338, A4^LA2^H—0.0035, A4^LA2^L < 0.0001; b A4^HA2^L—0.4615, A4^LA2^H—0.0517, A4^LA2^L—0.0019. c Schematic outline of a pooled CRISPR screen with a sgRNA knockout library against epigenetic regulators to identify genes required for cisplatin response in OVCAR4 cells. d MAGeCK analysis^,, for screen in c. Genes were ranked by robust rank aggregation (RRA). Immunoblots (e), annexin V⁺/PI⁻ apoptotic cell population determined by flow cytometry (f), and representative phase-contrast images (g) of OVCAR4 cells with indicated SMARCA4/2 perturbations and cisplatin treatments (e, f 48 h). Immunoblots (h), annexin V⁺/PI⁻ apoptotic cell population (i), and representative phase-contrast images (j) of H1703 cells with indicated SMARCA4/2 perturbations and cisplatin treatments (h, i 72 h). e–j Ctrl Control, A4^KO SMARCA4 knockout, shA2 shRNA targeting SMARCA2, cl. PARP cleaved PARP, cl. caspase 3 cleaved caspase 3, A4 SMARCA4, A2 SMARCA2. Scale bar, 150 μm. Mean ± SD, n = 3 independent experiments, one-way ANOVA followed by Dunnett’s test for multiple comparisons, p values (p): f all (<0.0001); i A2 (0 μM)—0.0032, A4 (0 μM)—0.0023, A2 or A4 (3 μM) < 0.0001. **p < 0.01, ****p < 0.0001.

Fig. 2. SMARCA4 modulates Ca²⁺ flux from the ER to mitochondria.

Gene set enrichment analysis plots of indicated Gene Ontology terms in SCCOHT-1 (a) and BIN-67 (b) cells ± SMARCA4 (A4) restoration. Ctrl control, FDR false discovery rate. c Immunoblots of indicated proteins in SCCOHT-1 cells ± A4 restoration. Changes of cytosolic (d) and mitochondrial (e) Ca²⁺ contents in SCCOHT-1 cells ± A4 restoration upon histamine stimulation. d 28 Ctrl and 21 A4 cells from n = 4 independent experiments were analyzed. e 44 Ctrl and 20 A4 cells from n = 4 independent experiments were analyzed. f Immunoblots of indicated proteins in H1703 cells ± A4 restoration. Changes of cytosolic (g) and mitochondrial (h) Ca²⁺ contents in H1703 cells ± A4 restoration upon histamine stimulation. g 41 Ctrl and 74 A4 cells from n = 3 independent experiments were analyzed. h 45 Ctrl and 63 A4 cells from n = 3 independent experiments were analyzed. i Immunoblots of indicated proteins in OVCAR4 cells ± SMARCA4 knockout (A4^KO). Changes of cytosolic (j) and mitochondrial (k) Ca²⁺ contents in OVCAR4 cells with ± A4^KO upon histamine stimulation. j 60 Ctrl and 53 A4^KO cells from n = 3 independent experiments were analyzed. k 41 Ctrl and 40 A4^KO cells from n = 3 independent experiments were analyzed. l Immunoblots of indicated proteins in H1437 cells ± A4^KO. Changes of cytosolic (m) and mitochondrial (n) Ca²⁺ contents in H1437 cells ± A4^KO upon histamine stimulation. m 39 Ctrl and 37 A4^KO cells from n = 3 independent experiments were analyzed. n 38 Ctrl and 42 A4^KO cells from n = 3 independent experiments were analyzed. d, e, g, h, j, k, m, n Left: traces of cytosolic and mitochondrial Ca²⁺ contents upon 100 μM histamine stimulation (mean ± SEM). Right: quantification of the maximal Ca²⁺ signal peaks induced by histamine stimulation (mean ± SD). The Ca²⁺ probes R-GECO (R-GECO F/F0) and CEPIA-2mt (CEPIA-2mt F/F0) were used to monitor cytosolic and mitochondrial Ca²⁺, respectively. Arb. units arbitrary units. Two-tailed t-test, p values (p): d 0.0003, e 0.0016, g 0.0012, h 0.0084, j 0.0088, k 0.0182, m 0.0004, n 0.0084. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 3. SMARCA4/2 regulate ITPR3 transcription through remodeling chromatin accessibility at its gene locus.

a Venn diagram of Ca²⁺-related genes from Fig. 2a, b that are enriched in SCCOHT-1 and BIN-67 cells with SMARCA4 restoration. b Heatmap of Ca²⁺-related genes bound by SMARCA4 (n = 69) in indicated SCCOHT (SCCOHT-1 and BIN-67) and NSCLC (H1299) cell lines with SMARCA4/2 restoration. Left: normalized reads from RNA-seq data of BIN-67 and SCCOHT-1 cells with SMARCA4 restoration. Middle: normalized reads from RNA-seq data of BIN-67 cells with SMARCA4/2 restoration. Right: normalized signal from microarray data of H1299 cells with SMARCA4 restoration. Row scaling was used to generate the heatmap. The last column represents changes of genes in H1299 cells ± SMARCA4 restoration: ns not significant, up upregulated, down downregulated. c RT-qPCR measurements of ITPR3 mRNA expression in indicated SCCOHT and NSCLC cell lines with SMARCA4/2 restoration. GAPDH was used for normalization. Mean ± SD, n = 3 (BIN-67, SCCOHT-1, H1299) or 4 (H1703) independent experiments, one-way ANOVA followed by Dunnett’s test for multiple comparisons to the control group (BIN-67, SCCOHT-1, H1703) or two-tailed t-test (H1299), p values (p): BIN-67, all <0.0001; SCCOHT-1, A2—0.0004, A4—0.0001; H1299—0.0318; H1703, A2—0.0097, A4—0.0013. d Immunoblots of indicated proteins in indicated SCCOHT and NSCLC cell lines ± SMARCA4/2 restoration. e SMARCA4 occupancy in vicinity of the ITPR3 locus assessed by chromatin immunoprecipitation sequencing (ChIP-seq) in indicated SCCOHT and lung cancer cell lines ± SMARCA4 restoration. SMARCA4 in H1299 cells was induced by doxycycline (Dox). Track height is normalized to relative number of mapped reads. f Chromatin structure changes in vicinity of the ITPR3 locus assessed by H3K27Ac ChIP-seq and assay for transposase-accessible chromatin sequencing (ATAC-seq) in indicated SCCOHT and lung cancer cell lines ± SMARCA4/2 restoration. Track height is normalized to relative number of mapped reads. a–f Ctrl control, A4 SMARCA4, A2 SMARCA2. *p < 0.05, **p < 0.01, ****p < 0.0001.

Fig. 4. SMARCA4/2 loss inhibits apoptosis by constricting IP3R3-mediated Ca²⁺ flux.

a Immunoblots of SCCOHT-1 cells ± SMARCA4 (A4) and ITPR3 (R3) perturbations. shR3: shRNA targeting ITPR3. Changes of cytosolic (b) and mitochondrial (c) Ca²⁺ contents in SCCOHT-1 cells ± A4 and R3 perturbations upon histamine stimulation. b 43 control (Ctrl), 30 A4, 51 A4 ShR3♯1, and 50 A4 ShR3♯2 cells from n = 4 independent experiments were analyzed. c 31 Ctrl, 30 A4, 50 A4 ShR3♯1, and 50 A4 ShR3♯2 cells from n = 4 independent experiments were analyzed. d Immunoblots of H1703 cells ± A4 and R3 perturbations. siR3 siRNA targeting ITPR3. Changes of cytosolic (e) and mitochondrial (f) Ca²⁺ contents in H1703 cells ± A4 and R3 perturbations upon histamine stimulation. e 64 Ctrl, 70 A4, and 64 A4 siR3 cells from n = 3 independent experiments were analyzed. f 53 Ctrl, 53 A4, and 50 A4 siR3 cells from n = 3 independent experiments were analyzed. g Immunoblots of OVCAR4 cells with ITPR3 knockdown 48 h post cisplatin treatment. cl cleaved. h Annexin V⁺/PI⁻ apoptotic cell population determined by flow cytometry in OVCAR4 cells described in g. Immunoblots of H1703 cells ± R3 overexpression (i) or ±A4 and R3 perturbations (j). Cells were collected 72 h after the treatment. k Annexin V⁺/PI⁻ apoptotic cell population determined by flow cytometer in H1703 cells described in j. b, c, e, f Left: traces of cytosolic or mitochondrial Ca²⁺ contents upon 100 μM histamine stimulation (mean ± SEM). Right: quantification of the maximal Ca²⁺ signal peaks induced by histamine (mean ± SD). The Ca²⁺ probes R-GECO (R-GECO F/F0) and CEPIA-2mt (CEPIA-2mt F/F0) were used to monitor cytosolic and mitochondrial Ca²⁺, respectively. Arb. units arbitrary units. One-way ANOVA followed by Dunnett’s test for multiple comparisons to Ctrl, p values (p): b A4—0.0003, shR3#1—0.6223, shR3#2—0.9866; c A4 < 0.0001, shR3#1—0.9109, shR3#2—0.9845; e A4—0.0038, siR3—0.4232; f A4—0.0007, siR3—0.1620. h, k Mean ± SD, n = 3 independent experiments, one-way ANOVA followed by Dunnett’s test for multiple comparisons, p values (p): all <0.0001. *p < 0.01, ***p < 0.001, ****p < 0.0001; ns not significant.

Fig. 5. IP3R3 expression is reduced in SMARCA4/2-deficient cancers.

a Correlation of ITPR3 and SMARCA2 (A2) mRNA in ovarian (n = 11) and lung (n = 48) cancer cell lines with low expression of SMARCA4 (A4). Expression data were obtained from Cancer Cell Line Encyclopedia (CCLE) in reads per kilobase million (RPKM). A4^Low, bottom quartile. r: Pearson correlation; p: p value (two-tailed). b Immunoblots of lung cancer cell lines with indicated SMARCA4/2 (A4/2) status. Pro proficient, def deficient; * KRAS mutant. c Correlation of ITPR3 and SMARCA2 mRNA in ovarian cancer (n = 89) and lung adenocarcinoma (LUAD, n = 128) patient tumors with low expression of SMARCA4. Gene expression data were obtained from UCSC Xena in fragments per kilobase million (FPKM). A4^Low, bottom quartile. r, Pearson correlation; p, p value (two-tailed). d ITPR3 mRNA expression in SCCOHT and ovarian cancer patient tumors. TCGA ovarian cancers (OV) (n = 379) were stratified based on SMARCA4/2 expression as indicated in Supplementary Fig. 2b. ITPR3 expression in FPKM was normalized to ACTB. H: high; L: low. One-way ANOVA Brown–Forsythe and Welch tests followed by Dunnett’s test for multiple comparisons to A4^HA2^H, p values (p): A4^HA2^L—0.3830, A4^LA2^H—0.0009, A4^LA2^L < 0.0001, SCCOHT—0.0108, or two-tailed t-test between A4^LA2^L group and SCCOHT, p = 0.4953. Representative images (e, g) and H-score (f, h) of immunohistochemistry analysis for IP3R3 and SMARCA4 expression in patient tumors. e, f HGSC (n = 49) and SCCOHT (n = 45). g, h NSCLC (n = 59). Scale bar, 100 μm. Mann–Whitney test (two-tailed), p values (p): f 0.0066, h 0.0270. i Tumor growth in H1703 xenograft models ± exogenous SMARCA4 expression. Doxycycline (Dox) was given daily starting on day 21 to induce SMARCA4. Upper, tumor size; lower, endpoint tumor weight. Mean ± SEM, − Dox (n = 4 animals), + Dox (n = 6 animals), two-way ANOVA (upper), two-tailed t-test (lower), p values (p): upper—0.0014, lower—0.0008. Representative images (j) and digital quantification (k) of immunohistochemistry analysis in endpoint tumors described in i. Scale bar, 100 μm. Mean ± SD, − Dox (n = 4), + Dox (n = 6), two-tailed t-test (lower), p values (p): upper, middle <0.0001; lower—0.0041. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns not significant.

Fig. 6. The histone deacetylase inhibitor quisinostat rescues IP3R3 expression and enhances cisplatin response in SMARCA4/2-deficient cancer cells.

a SMARCA2 and ITPR3 mRNA expression in H1703 cells treated with quisinostat for 48 h. Mean ± SD, n = 3 independent experiments, one-way ANOVA followed by Dunnett’s tests for multiple comparisons to untreated, p values (p): left, 10 nM—0.0003, 40 nM < 0.0001; right, 10 nM—0.0027, 40 nM—0.0002. Immunoblots (b) and annexin V⁺/PI⁻ apoptotic population (c) in H1703 cells treated with cisplatin (3 μM) and quisinostat (10 nM) for 72 h. cl cleaved. Mean ± SD, n = 3 independent experiments, one-way ANOVA followed by Dunnett’s test for multiple comparisons, p values: all <0.0001. d Colony formation assay for H1703 cells treated with cisplatin and quisinostat for 12 days. Immunoblots of H1703 cells ± SMARCA2 knockout (e) or ±IP3R3 knockdown (f) treated with cisplatin (3 μM) and quisinostat (10 nM) for 72 h. g Representative images from confocal live-cell imaging of H1703 cells overexpressing the mitochondrial Ca²⁺ probe CEPIA-2mt treated with quisinostat or/and cisplatin and stained with the mitochondrial marker Mitotracker deep red. Cisplatin: 2 μM, 24 h; quisinostat: 40 nM, 72 h. Scale bar, 25 μm. h Quantification of basal mitochondrial Ca²⁺ levels from g, showing the ratio of CEPIA-2mt/Mitotracker fluorescence intensities compared to control. 43 control (ctrl), 42 cisplatin, 38 quisinostat, and 46 quisinostat/cisplatin cells from n = 4 independent experiments were analyzed. Mean ± SD, one-way ANOVA followed by Dunnett’s test for multiple comparisons to ctrl, p values: cisplatin—0.2105, quisinostat—0.2985, combination—0.0032. i Tumor growth in H1703 xenograft models treated with cisplatin (4 mg kg⁻¹) and quisinostat (10 mg kg⁻¹). Vehicle (n = 6 animals), all other groups (n = 5 animals); upper, tumor volume, mean ± SEM, two-way ANOVA; lower, endpoint tumor weight, one-way ANOVA followed by Dunnett’s tests for multiple comparisons to the combination, p values: upper, cisplatin < 0.0001, quisinostat—0.0054; lower, cisplatin—0.0008, quisinostat—0.0330. Representative images (j) and digital quantification (k) of immunohistochemistry analysis in endpoint tumors described in i. Scale bar, 100 μm. Mean ± SD, all groups (n = 5), one-way ANOVA followed by Dunnett’s test for multiple comparisons, p values: all <0.0001. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001; ns not significant.