Investigating synthetic lethality and PARP inhibitor resistance in pancreatic cancer through enantiomer differential activity

Abstract

The RAD51-BRCA2 interaction is central to DNA repair through homologous recombination. Emerging evidence indicates RAD51 overexpression and its correlation with chemoresistance in various cancers, suggesting RAD51-BRCA2 inhibition as a compelling avenue for intervention. We previously showed that combining olaparib (a PARP inhibitor (PARPi)) with RS-35d (a BRCA2-RAD51 inhibitor) was efficient in killing pancreatic ductal adenocarcinoma (PDAC) cells. However, RS-35d impaired cell viability even when administered alone, suggesting potential off-target effects. Here, through multiple, integrated orthogonal biological approaches in different 2D and 3D PDAC cultures, we characterised RS-35d enantiomers, in terms of mode of action and single contributions. By differentially inhibiting both RAD51-BRCA2 interaction and sensor kinases ATM, ATR and DNA-PK, RS-35d enantiomers exhibit a 'within-pathway synthetic lethality' profile. To the best of our knowledge, this is the first reported proof-of-concept single small molecule capable of demonstrating this built-in synergism. In addition, RS-35d effect on BRCA2-mutated, olaparib-resistant PDAC cells suggests that this compound may be effective as an anticancer agent possibly capable of overcoming PARPi resistance. Our results demonstrate the potential of synthetic lethality, with its diversified applications, to propose new and concrete opportunities to effectively kill cancer cells while limiting side effects and potentially overcoming emerging drug resistance.

Fig. 1. RS-35d, S-35d and R-35d differential effects on HR, RAD51 nuclear activity and DNA damage.

a Effect on HR caused by RS-35d, S-35d and R-35d in BxPC-3 cells. Results are expressed as mean ± SD (n = 3). b mClover-Lamin assay for DNA repair by HR performed in HEK293 cells transfected with Lamin A sgRNA and mClover-Lamin A donor and treated with RS-35d, S-35d and R-35d. c Representative microscopic merged images of mClover-Lamin A fusion protein and DNA (DAPI staining) fluorescence after exposure to increasing doses of RS-35d, S-35d or R-35d (scale bar, 30 μm). d Analysis of HR-positive (mClover-Lamin A-positive) cells after administration of increasing doses of RS-35d, S-35d or R-35d (the three compounds were tested in independent experiments and their effects were compared to their respective vehicle control). Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with one-way ANOVA followed by Dunnett’s multiple comparison test, with p < 0.01 or p < 0.001 vs CTRL. e Immunofluorescence detection of nuclear RAD51 in BxPC-3 exposed to 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or after pre-treatment with 50 µM cisplatin (CDDP). Representative images of DAPI-stained nuclei and the corresponding nuclear localisation of RAD51 immune-labelling (scale bar, 30 μm) and analysis of RAD51-positive nuclei (%). Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with one-way ANOVA followed by Tuckey’s multiple comparison test, with ^§§§p < 0.001 vs CTRL and ^##p < 0.01 vs 50 µM CDDP. f Evaluation of DNA damage via nuclear γH2AX foci immune detection in BxPC-3 cells treated for 48 h with 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or in combination with 10 µM olaparib (OLA). Representative images of DAPI-stained nuclei and the corresponding γH2AX immune-labelling (scale bar, 50 μm) and analysis of γH2AX-positive nuclei (%). g Micronuclei evaluation in BxPC-3 cells treated for 72 h with 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or in combination with 10 µM OLA. Representative images of DAPI-stained nuclei (white asterisks indicate micronuclei presence) (scale bar, 15 μm) and analysis of micronuclei-bearing cells (%). Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with one-way ANOVA followed by Tuckey’s multiple comparison test, with ^§p < 0.05 or ^§§p < 0.01 vs CTRL; ^#p < 0.05, ^##p < 0.01 or ^###p < 0.001 vs 10 µM OLA; *p < 0.05, **p < 0.01 or ***p < 0.001 vs RAD51-BRCA2 inhibitor alone.

Fig. 2. RS-35d, S-35d and R-35d differential synergism with olaparib and their diverse impact on cell viability.

a Cell viability measured after 72 h exposure to 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or in combination with 10 µM olaparib (OLA) in BxPC-3, HPAC, Capan-1 and H-6037 cells. b Gene expression profile of the main HR genes in the employed three PDAC cell lines. Relative gene expression values were obtained from the Cancer Dependency Map portal Depmap (https://depmap.org/portal/) and are inferred from RNA-seq data using the RSEM tool and are reported after log₂ transformation, using a pseudo-count of 1; log₂(TPM + 1). c Interaction index (i. index) of RS-35d, S-35d or R-35d association with olaparib in BxPC-3, HPAC, Capan-1 and H-6037 cell lines. I. index values were made explicit in their respective graphs when the drug association showed synergism (i. index < 0.8). Cell death assessment by CellTox Green (d) and vital dyes staining (e) measured after 72 h exposure to 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or in combination with 10 µM OLA in BxPC-3 cells; d RS-35d, S-35d or R-35d alone or in combination with 10 µM OLA in BxPC-3 cells; e Representative images of DAPI-stained nuclei and the corresponding PI-stained cells (scale bar, 50 μm) and analysis of PI-positive cells (%). Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with two-way ANOVA followed by Tuckey’s multiple comparison test, with ^§p < 0.05, ^§§p < 0.01 or ^§§§p < 0.001 vs CTRL; ^#p < 0.05, ^##p < 0.01 or ^###p < 0.001 vs 10 µM OLA; *p < 0.05, **p < 0.01 or ***p < 0.001 vs RAD51-BRCA2 inhibitor alone.

Fig. 3. RS-35d, S-35d and R-35d differentially influence cancer hallmarks of tumour aggressiveness.

Evaluation of BxPC-3 cell proliferation (a), migration (b) and apoptosis (c) after exposure to 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or in combination with 10 µM OLA. a Representative images of BxPC-3 colonies and analysis of the corresponding colony rate (expressed as % of CTRL). b Representative images of BxPC-3 wound-healing scratches and analysis of the corresponding migration rate (expressed as % of the wound area at t = 0 h). c Representative dot-plots of PI- and FITC-Annexin V-stained BxPC-3 cells analysed through flow cytometry and analysis of the measured apoptotic cells (%). Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with one-way ANOVA followed by Tuckey’s multiple comparison test, ^§p < 0.05 or ^§§§p < 0.001 vs CTRL; ^#p < 0.05, ^##p < 0.01 or ^###p < 0.001 vs 10 µM OLA; *p < 0.05, **p < 0.01 or ***p < 0.001 vs RAD51-BRCA2 inhibitor alone.

Fig. 4. RS-35d, S-35d and R-35d differential effect in PDAC 3D cultures.

a Cell viability measured after 72 h exposure to 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d alone or in combination with 10 µM olaparib (OLA) in BxPC-3 3D spheroids. b I. index of RS-35d, S-35d or R-35d association with olaparib in BxPC-3 3D spheroids. I. index values were made explicit when the drug association showed synergism (i. index < 0.8). c Evaluation of the effect of the 72 h treatment with 40 µM RS-35d alone or in combination with 10 µM OLA on BxPC-3 3D spheroid volume and cell death; time-course brightfield and fluorescence images of BxPC-3 3D spheroids treated as previously described (scale bar, 200 μm); time-course analysis of the corresponding 3D spheroid volume (expressed as % of the spheroid volume at t = 0 h); analysis of cell death in 3D spheroid at 72 h (expressed as PI/Calcein-AM ratio). Results are expressed as mean ± SD (n = 6). Evaluation of cell viability (d) and apoptosis (e) upon 120 h treatment with 40 µM RS-35d alone or in combination with 10 µM OLA in human PDAC organoids; time-course representative images of RS-35d- and RS-35d/olaparib-induced apoptosis via caspase 3/7 cleavage (scale bar, 400 μm) and analysis of the corresponding caspase 3/7 cleavage-induced green fluorescence signal. Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with two-way ANOVA (a, c, d, e) or one-way ANOVA (c) followed by Tuckey’s multiple comparison test, with ^§p < 0.05, ^§§p < 0.01 or ^§§§p < 0.001 vs CTRL; ^##p < 0.01 or ^###p < 0.001 vs 10 µM OLA; *p < 0.05 or **p < 0.01 vs RAD51-BRCA2 inhibitor alone.

Fig. 5. RS-35d, S-35d and R-35d inhibition of DDR kinases and dissection of RS-35d mechanism of action.

a 2D and 3D i. indices of the reconstituted racemic mixture (20 µM S-35d + 20 µM R-35d) and their association with 10 µM olaparib (OLA). b RS-35d, S-35d and R-35d inhibitory effect, expressed as residual activity (% of controls), on a panel of DDR cascade kinases. Results are expressed as mean ± SD (n = 2). c Western blot analysis of p-CHK1, p-CHK2, p-Akt (S473) levels in BxPC-3 cells treated with 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d for 24 h (DMSO 0.6% as vehicle control, CTRL). The images are representative Western blots. Effect of 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d treatment for 24 h on cell cycle phases (d), ROS production (e) and ATP levels (f). Unsupervised PCA score plot (g) and relative metabolites consumption and release (h) of culture media of BxPC-3 cells exposed to 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d for 24 h (DMSO 0.6% as vehicle control, CTRL); g 95% confidence ellipses are displayed for each group; h heatmap of the significative consumption/release (relative to CTRL values) of metabolites identified from univariate analysis. Results are expressed as mean ± SD (n = 3). Statistical analysis was performed with one-way (b, e, h) or two-way ANOVA (c, d, f) followed by Tuckey’s (b–f) or Bonferroni’s (h) multiple comparison test, with *p < 0.05, **p < 0.01 or ***p < 0.001 vs CTRL.

Fig. 6. RS-35d proteomic profile in BxPC-3 cells.

a Volcano plot of control vs RS-35d up- (red) and downregulated (green) proteins in BxPC-3 cells, p-value < 0.05 and log₂(fold change) > 1 or < −1 were used as significance cut-off. b Venn diagram of RS-35d up- and downregulated common proteins with RAD51 interactome and ATM/ATR/DNA-PK common interactomes; Western blot analysis of FANCD2, FANCI and RPA3 expression in BxPC-3 cells treated with 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d for 24 h. The images are representative Western blots. Results are normalised over β-actin expression and expressed as mean ± SD (n = 3). Statistical analysis was performed with two-way ANOVA followed by Dunnett’s multiple comparison test, with **p < 0.01 vs CTRL. c Gene Ontology (GO) cellular component, molecular functions and biological processes enriched from the analyses of the up- (red) and downregulated (green) proteins by RS-35d. d STRING functional analysis of up-regulated proteins by RS-35d. Proteins were clustered using Markov clustering (MCL), with default inflation parameter = 3; the thickness of lines representing protein-protein interactions indicates the strength of data support.

Fig. 7. RS-35d exerts intrinsic synthetic lethality in PARPi-resistant, BRCA2-mutated Capan-1/OP 2D and 3D cultures.

Cell viability after 72 h exposure to 10 µM olaparib (OLA), 2 µM talazoparib (TAL), 10 µM rucaparib (RUC) or 10 µM AZD2461 (AZD), alone or in combination with 40 µM RS-35d, 20 µM S-35d or 20 µM R-35d in Capan-1/OP 2D (a) and 3D (b) cultures. c Evaluation of the effect of the 72 h treatment with 40 µM RS-35d on Capan-1/OP 3D spheroid volume and cell death. Time-course representative brightfield and fluorescence images of Capan-1/OP 3D spheroids treated as previously described (scale bar, 200 μm), analysis of the corresponding 3D spheroid volume (expressed as % of the spheroid volume at t = 0 h) and cell death in 3D spheroid at 72 h (expressed as PI/Calcein-AM ratio). Results are expressed as mean ± SD (n = 6). Statistical analysis was performed with Student’s t-test (c right) or two-way ANOVA (a, b, c middle), followed by Tuckey’s multiple comparison test, with ^§p < 0.05, ^§§p < 0.01 or ^§§§p < 0.001 vs CTRL; ^##p < 0.01 or ^###p < 0.001 vs PARPi alone.