Repression of LSD1 potentiates homologous recombination-proficient ovarian cancer to PARP inhibitors through down-regulation of BRCA1/2 and RAD51

Abstract

Poly (ADP-ribose) polymerase inhibitors (PARPi) are selectively active in ovarian cancer (OC) with homologous recombination (HR) deficiency (HRD) caused by mutations in BRCA1/2 and other DNA repair pathway members. We sought molecular targeted therapy that induce HRD in HR-proficient cells to induce synthetic lethality with PARPi and extend the utility of PARPi. Here, we demonstrate that lysine-specific demethylase 1 (LSD1) is an important regulator for OC. Importantly, genetic depletion or pharmacological inhibition of LSD1 induces HRD and sensitizes HR-proficient OC cells to PARPi in vitro and in multiple in vivo models. Mechanistically, LSD1 inhibition directly impairs transcription of BRCA1/2 and RAD51, three genes essential for HR, dependently of its canonical demethylase function. Collectively, our work indicates combination with LSD1 inhibitor could greatly expand the utility of PARPi to patients with HR-proficient tumor, warranting assessment in human clinical trials.

Fig. 1. LSD1 is key mediator in OC and promotes OC progress.

a–c Representative IHC images (a) and quantification (b, c) of LSD1 protein level in human OC tissues and NATs from TMA (n = 45 paired samples). The quantification analyses were based on staining density scores of IHC (paired two-tailed t test). Scale bar, 400 μm. d, e Representative IHC images (d) and quantification (e) of LSD1 protein level in human FTE (n = 30 samples), HOSE (n = 10 samples) and OC (n = 39 samples) tissues. The quantification analyses were based on staining density scores of IHC (unpaired two-tailed Student’s t test). Scale bar, 50 μm. f Kaplan–Meier plot depicting overall survival of OC patients with tumors expressing high (red) or low (black) levels of LSD1 using TMA (log-rank test). g Cell growth of LSD1 knockdown cells detected by CCK8 assay compared with their control (shCtrl). Data represent mean ± SEM of three biologically independent experiments (two-way ANOVA). h, i Representative images (h) and quantification (i) of colony formation assay for OC cells with LSD1 knockdown. Samples were normalized to shCtrl. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). j–o Tumor volume and tumor weight in shLSD1-expressing and shCtrl-expressing A2780 (j, k), SKOV3 (l, m) and ES2 (n, o) subcutaneous xenografts in nude mice. Data represent mean ± SEM (n = 8 mice per group for A2780 and SKOV3 xenograft models, n = 7 mice per group for ES2 xenograft models; two-way ANOVA for panels j, l, n, and unpaired two-tailed Student’s t test for panels k, m, o). p–s Representative living luminescence images (p, r) and quantification of the luciferase fluorescence signal intensity (q, s) of shLSD1-expressing and shCtrl-expressing A2780 and SKOV3 intraperitoneal xenografts in nude mice. Data represent mean ± SEM (n = 6 mice per group for A2780 xenograft models, n = 7 mice per group for SKOV3 xenograft models; unpaired two-tailed Student’s t test). *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided in the Source Data file.

Fig. 2. LSD1 pharmacological inhibition has therapeutic potential in OC.

a Cell viability in the indicated OC cell lines after ZY0511 treatment for 72 h. Data represent mean ± SEM of three biologically independent experiments. b CCK8 assay at varied concentrations of ZY0511 in shCtrl-expressing and shLSD1-expressing A2780, SKOV3 and ES2 cells. Data represent mean ± SEM of three biologically independent experiments). c Quantification of colony formation assay. Data represent mean ± SEM of three biologically independent experiments; two-way ANOVA). d, e Tumor volume (d) and tumor weight (e) of mice bearing A2780, SKOV3 and ES2 subcutaneous xenografts. Data represent mean ± SEM (n = 6 mice per group; two-way ANOVA for panel d and one-way ANOVA for panel e). f The luciferase fluorescence signal intensity of mice bearing A2780-Luc and SKOV3-Luc intraperitoneal xenografts. Data represent mean ± SEM (n = 6 mice per group for A2780 xenograft models, n = 5 mice per group for SKOV3 xenograft models; one-way ANOVA). g CETSA performed in harvested tissues. Data represent mean ± SEM (n = 6 mice per group; unpaired two-tailed Student’s t test). h, i Quantification of IHC of the indicated proteins in tumor tissues from mice. Data represent mean ± SEM of from three different mice (one-way ANOVA). CC3 Cleaved Caspase 3. j Pharmacokinetics of LSD1 inhibitor (ZY0511). SD rats were administered 5 mg/kg ZY0511 intravenously (n = 5) or 30 mg/kg ZY0511 intraperitoneally (n = 4). Data represent mean ± SEM. k Distribution of ZY0511 in main organs and plasma. Data represent mean ± SEM (n = 6 animals per group). l Representative H&E staining images of the heart, liver, spleen, lung, and kidney at the end of the dosing. Scale bar, 200 μm. m, n Blood routine assay (m) and blood biochemical assay (n) performed at the end of treatment. Data represent mean ± SEM (n = 6 mice per group; unpaired two-tailed Student’s t test). ns not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided in the Source Data file.

Fig. 3. LSD1 inhibition activates DDR and suppresses expression of HR proteins.

a, b Ingenuity Pathway Analysis of the representative twenty significantly regulated pathways of shLSD1-expressing versus non-targeting control shCtrl-expressing ES2 cells (a) and LSD1i (ZY0511)-treated versus untreated ES2 cells (b). Upregulated pathways are presented in orange and downregulated pathways are in blue. p values generated by right-tailed Fisher’s exact test. c GSEA enrichment score curves of LSD1 knockdown or LSD1i (ZY0511) treatment regulated genes of A2780 and ES2 cells. ES, enrichment score; NES, normalized enrichment score; FDR, false discovery rate. d Heatmap showing gene expression changes between shLSD1-expressing and non-targeting control shCtrl-expressing ES2 cells with respect to genes contained in the “Reactome_DNA_Double_Strand_Break_Repair” gene set. e Heatmap (left) and HRD scores (right) from unsupervised clustering of HRD gene signatures using the RNA-seq dataset of A2780 and ES2 treated by LSD1i (ZY0511) or LSD1 knockdown (shLSD1). Higher scores represent defective HR. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). f, g RT-qPCR analysis of indicated gene expression in A2780, SKOV3 and ES2 cells treated with different concentrations of LSD1i (ZY0511) for 36 h (f) or with 1 μM LSD1i (ZY0511) for different time periods (g). GAPDH was used as the loading control. Data represent mean ± SEM of three biologically independent experiments (one-way ANOVA). h Western blot analysis of indicated proteins in A2780, SKOV3 and ES2 cells treated with the indicated dose of ZY0511 for 48 h or in ES2 cells treated with shRNAs targeting LSD1 (shLSD1 #1 and shLSD1 #2) and non-targeting control (shCtrl). α-Tubulin was used as the loading control. ns, not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided in the Source Data file.

Fig. 4. LSD1 inhibition suppresses HR and increases DNA DSBs.

a–d Representative images (a, c) and quantification (b, d) of neutral comet assays in A2780, SKOV3 and ES2 cells treated with indicated ZY0511 for 48 h or LSD1 knockdown (shLSD1) treatment after 5 Gy ionizing radiation (IR) treatment. Scale bar, 100 μm. Data represent mean ± SEM (unpaired two-tailed Student’s t test). The experiments were repeated three times. e, f Representative images (e) and quantification (f) of γH2AX-foci staining performed in A2780, SKOV3 and ES2 cells with or without 1 μM LSD1i (ZY0511 or SP2577) for 48 h or LSD1 knockdown (shLSD1) treatment. Green, γH2AX; blue, DAPI. Scale bar, 10 μm. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). g, h Representative images (g) and quantification (h) and of RAD51 nuclear foci in A2780, SKOV3 and ES2 cells with or without 1 μM LSD1i (ZY0511 or SP2577) for 48 h or LSD1 knockdown (shLSD1) treatment at 4 h after 2 Gy IR treatment. Green, RAD51; blue, DAPI. Scale bar, 10 μm. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). i–l Quantification of γH2AX foci and RAD51 foci per nucleus at the indicated time points after 2 Gy IR treatment in ES2 cells treated with or without 1 μM LSD1i (ZY0511 or SP2577) for 48 h (i, k), or shRNA suppression of LSD1 (j, l). Data represent mean ± SEM of three biologically independent experiments (two-way ANOVA). m Schematic illustration of the GFP-based HR reporter assay (DR-GFP) and NHEJ reporter assay (EJ5-GFP). iGFP, internal GFP repeat. n, o Quantification of HR and NHEJ using DR-GFP and EJ5-GFP reporter assay, respectively. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). ns, not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided as the Source Data file.

Fig. 5. LSD1 binds BRCA1, BRCA2, and RAD51 gene promoter, regulating these gene transcription dependently of its canonical demethylase function.

a Pie chart showing the genomic distribution of LSD1 peaks based on RefSeq. b Levels of LSD1, H3K9me2, and H3K4me2 bound at the TSS of peaks in ES2 cells, as measured by CUT&Tag-seq analysis. Transcription start site, TSS. c Levels of ATAC bound at the TSS in ES2 cells, as measured by ATAC-seq analysis. d IGV plot showing the distributions of LSD1, H3K9me2, H3K4me2, and ATAC-seq peaks binding in the promoters of BRCA1, BRCA2, and RAD51 in ES2 cells. e ChIP-qPCR analysis showing the enrichment levels of LSD1, H3K9me2 and H3K4me2 at the BRCA1, BRCA2 and RAD51 gene promoter in A2780, SKOV3 and ES2 cells. Data represent the percent of total chromatin input ±SEM of three biologically independent experiments; unpaired two-tailed Student’s t test; ns, not significant). f, g ChIP-qPCR analysis showing the enrichment levels of H3K9me2 and H3K4me2 at the BRCA1, BRCA2 and RAD51 gene promoter. Data represent the percent of total chromatin input ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test; ns, not significant). h, i RT-qPCR analysis of indicated gene expression. GAPDH was used as the loading control. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). j Western blot analysis of indicated proteins. α-Tubulin was used as the loading control. Numbers below western blot panels represent relative quantification of the respective bands normalized to loading control by densitometry. k Quantification of RAD51 nuclear foci at 4 h after 2 Gy IR treatment in A2780 and ES2 cells. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). l Cell viability in response to olaparib in A2780 and ES2 cells. Data represent mean ± SEM of three biologically independent experiments (two-way ANOVA). ns, not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided as the Source Data file.

Fig. 6. LSD1 inhibition enhances PARPi sensitivity in OC cells.

a Cell viability in response to the PARPi (olaparib, niraparib, or rucaparib) in A2780 and ES2 cells with or without knockdown of LSD1. The IC₅₀ values were calculated using GraphPad software. Data represent mean ± SEM of three biologically independent experiments. b Dose-response curves of ZY0511 or PARPi (olaparib, niraparib, or rucaparib) alone or combined in A2780, SKOV3, and ES2 cell lines or in normal ovarian epithelial HOSEpiC and IOSE80 cells lines treated with varying concentrations of ZY0511 and PARPi for 72 h. Combination index (CI) was calculated using CompuSyn software with the Chou-Talalay equation. Data represent mean ± SEM of three biologically independent experiments. c Representative images of colony formation assay for A2780 cells treated with LSD1i (ZY0511), PARPi (olaparib, niraparib, or rucaparib), or their combination as indicated. d Percentage inhibition at each concentration of LSD1i (ZY0511), PARPi (olaparib, niraparib, or rucaparib), or their combination in A2780 cells. Data represent mean ± SEM of three biologically independent experiments. e Combination index (CI) scores for A2780 cells treated with LSD1i (ZY0511) in combination with PARPi (olaparib, niraparib, or rucaparib) at the indicated concentrations. Each CI score represents data from three biologically independent experiments. ns, not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data are provided as the Source Data file.

Fig. 7. LSD1 inhibition enhance PARPi-induced DNA damage and apoptosis.

a, b Quantification of neutral comet assays in A2780 cells (a) treated with vehicle, LSD1i (1 μM ZY0511), olaparib (4 μM), niraparib (4 μM), rucaparib (4 μM) alone or combined for 48 h, and in ES2 cells (b) treated with vehicle, LSD1i (1 μM ZY0511), olaparib (20 μM), niraparib (10 μM), rucaparib (20 μM) alone or combined for 48 h. Data represent mean ± SEM (unpaired two-tailed Student’s t test). The experiments were repeated three times. c, d Quantification of γH2AX foci in A2780 cells (c) and ES2 cells (d) treated with same concentration for 48 h as shown in Fig. 7a, b. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). e, f Representative images (e) and quantification (f) of cell apoptosis analysis in A2780 and ES2 cells treated with same concentration for 48 h as shown in Fig. 7a, b. Annexin V-positive cells were analyzed by flow cytometry after treatment. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). g, h Representative images (g) and quantification (h) of cell apoptosis analysis in shLSD1-expressing and shCtrl-expressing A2780 and ES2 cells. A2780 cells were treated with olaparib (4 μM), niraparib (4 μM), or rucaparib (4 μM), while ES2 cells were treated with olaparib (20 μM), niraparib (10 μM), or rucaparib (20 μM). Annexin V-positive cells were analyzed by flow cytometry at 48 h after treatment. Data represent mean ± SEM of three biologically independent experiments (unpaired two-tailed Student’s t test). i Western blot analysis of indicated proteins in A2780 cells treated with vehicle, 1 μM ZY0511, 4 μM PARPi (olaparib, niraparib or rucaparib), or a combination for 48 h. Numbers below western blot panels represent relative quantification of the respective bands normalized to loading control by densitometry. ns, not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided as the Source Data file.

Fig. 8. LSD1 inhibition sensitizes HR-proficient tumors to PARPi treatment in vivo.

a Schematic diagram of A2780, SKOV3 and ES2 subcutaneous tumor model and drug delivery. Mice were treated with vehicle (30 mg/mL PEG4000 plus 12 mg/mL Tween 20 in water and 5% DMSO plus 30% PEG300 in water intraperitoneally), ZY0511 (intraperitoneally 30 mg/kg), PARPi (olaparib 30 mg/kg, niraparib 10 mg/kg and rucaparib 10 mg/kg intraperitoneally), or a combination of ZY0511 and PARPi as indicated. Schematic diagram was created with BioRender.com. b–d Tumor volume of mice bearing A2780 (b), SKOV3 (c) and ES2 (d) subcutaneous xenografts and treated with vehicle or drugs showed in a. Data represent mean ± SEM (n = 6 mice per group; two-way ANOVA). e Body weight curves of mice treated with vehicle or drugs showed in a. Data represent mean ± SEM. (n = 18 mice per group from A2780, SKOV3 and ES2 subcutaneous xenografts; two-way ANOVA). f, g Tumor volume of mice bearing SKOV3 subcutaneous xenografts (f) and ES2 subcutaneous xenografts (g). Data represent mean ± SEM (n = 6 mice per group; two-way ANOVA). h Schematic diagram of PDX model and drug delivery. Created with BioRender.com. i, j Tumor volume (i) and body weight curves (j) of mice bearing PDX and treated with vehicle or drugs showed in h. Data represent mean ± SEM (n = 6 mice per group; two-way ANOVA). k–m Western blot analysis (k) and representative images (l) and quantification (m) of IHC of indicated proteins in tumor tissues from PDX subcutaneous xenografts. α-Tubulin was used as loading control. Numbers below western blot panels represent relative quantification of the respective bands normalized to loading control by densitometry. CC3, Cleaved Caspase 3. Scale bars, 100 μm (black), 40 μm (green). Data represent mean ± SEM of three random fields of view from three different mice (one-way ANOVA). ns not significant, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001. Source data and exact p values are provided as the Source Data file.