PARP-inhibitor synthetic lethality screen with protein kinase siRNA library. (A) HTS method. CAL51 cells plated in 96-well plates were transfected with siRNA. Each transfection plate contained 80 experimental siRNAs (SMARTPools of four different siRNA targeting the same gene) supplemented with four wells of non-targeting siCON, and two wells of siRNA directed against BRCA1 (positive control). Transfected cells were divided into six replica plates, half treated with DMSO vehicle alone and half with PARP inhibitor KU0058948 at 1 μM, the SF₈₀ of CAL51. Cell viability was assessed after 5 days of KU0058948 exposure using CellTiter-Glo Luminescent Cell Viability Assay (Promega). (B) Reproducibility of HTS method. Correlation of the effect of siRNA on cell growth in vehicle-treated plates from two replicates of the entire screen. Spearman correlation coefficient, r=0.83. (C) Correlation of KU0058948 sensitivity Z-scores from two replicates of the entire screen. r=0.54. (D) Scatter plot of averaged Z-scores from PARP inhibitor sensitivity screen carried out in duplicate with KU0058948. Red line indicates −3 averaged Z-score significance threshold. Black, siRNA (SMARTPools) targeting 779 protein kinase genes; red, siCON and blue, siBRCA1. Reflecting the reproducibility and sensitivity of the screen, siCON and siBRCA1 Z-scores were widely separated, with a screen Z′-factor (Zhang et al, 1999) of 0.34.

CDK5 silencing sensitises cells to DNA-damaging agents and induces a DNA-damage response. (A) Clonogenic survival assays in CAL51 cells transfected with siCDK5 or siCON and treated with camptothecin (SF₅₀ siCON 33.5 nM, siCDK5 11.3 nM, threefold more sensitive) or cisplatin (SF₅₀ siCON 13.8 μM, siCDK5 6.4 μM, 2.2-fold more sensitive). (B) Sensitivity to DNA damage in CDK5-silenced cells is characterised by an increase in apoptosis. FACS plots showing Annexin V and propidium iodide (PI) staining after treatment with hydrogen peroxide and either siCON or siCDK5 transfection. Percentages of Annexin V-positive/PI-negative cells were as follows: siCON-transfected cells, 3.4% (−H₂O₂) and 4.5% (+H₂O₂); siCDK5-transfected cells, 6.2% (−H₂O₂) and 10.4% (+H₂O₂). (C) CDK5 silencing induces γH2AX foci formation. The panels indicate confocal microscopic images from siCON- and siCDK5-transfected cells, with red indicating γH2AX and blue indicating ToPro3 DNA staining. Quantification of cells with ⩾5 γH2AX foci from three independent experiments. Error bars represent s.e.m.; silencing of CDK5 P=0.031 relative to siCON (Student's t-test). Silencing of BRCA2 acted as a control, P>0.05 relative to siCON. (D) CDK5 silencing induces basal RAD51 foci formation. Quantification of basal RAD51 foci, and RAD51 foci 5 h after 8-Gy irradiation (non-irradiated, siCDK5 transfected 17% versus siCON 7%). Error bars represent s.e.m. of three independent experiments; ^*P=0.009 relative to siCON (Student's t-test). Silencing of BRCA2 acted as a control for impairment of the radiation-induced foci.

CDK5 is required for cell-cycle DNA-damage checkpoints. (A) CDK5 is required for the intra-S-phase checkpoint. ³H-labelled thymidine incorporation was assessed by scintillation counting in an assay of radiation resistant DNA synthesis. DNA synthesis in cells irradiated 1 hour earlier with 10-Gy expressed relative to non-irradiated cells. Error bars represent s.e.m. of three independent experiments. Relative DNA synthesis after irradiation 45% siCON versus 66% siCDK5, ^*P=0.033 (Student's t-test). Right panel, DNA synthesis in untransfected CAL51 cells and CAL51 cells exposed to 10 μM KU0055933 (ATM inhibitor), relative DNA synthesis 45 versus 72%, respectively P<0.01. (B, C) CDK5 is required for G₂/M checkpoint function. (B) FACS plots of cells stained with propidium iodide and an antibody directed against the mitosis marker phospho-histone H3. Mitotic cells are highlighted in the box. (C) Quantification of three independent experiments, with proportion of mitotic cells after irradiation expressed relative to non-irradiated cells. Error bars represent s.e.m. of three independent experiments. Relative mitotic entry after irradiation 12% siCON versus 42% siCDK5; ^*P=0.017 (Student's t-test). Right panel, mitotic entry in untransfected CAL51 cells and CAL51 cells exposed to 10 μM KU0055933 (ATM inhibitor), relative mitotic entry 11 versus 97%, respectively; P<0.001.

Validation of PARP-inhibitor synthetic lethality screen. (A) Validation of hits from the PARP inhibitor HTS. PARP inhibitor sensitivity assay repeated in triplicate with the four different siRNA originally included in each SMARTPool. The surviving fractions following PARP inhibition are shown, including those after transfection with siBRCA1 (blue) and siCON (red). CDK5 revalidated with all four siRNA; PNKP revalidated with three siRNAs; and MAPK12, PLK3, STK36 and STK22C revalidated with two siRNA. As an example of a probable off-target hit, ADRA1A silencing sensitises to PARP inhibitor due to an off-target effect of the fourth siRNA. ^*P<0.0227 compared with siCON (Student's t-test). Error bars represent the s.e.m. (B) PARP-inhibitor sensitivity assessed by clonogenic assay in the six novel validated hits from the screen. Error bars represent s.e.m. of three independent experiments.

CDK5 is not required for early DNA DSB signalling or DNA DSB repair. (A) Expression of dominant-negative CDK5 sensitises to PARP inhibitor. CAL51 cells transfected with empty vector (pcDNA3.1empty), CDK5 expression vector (pCDK5) or kinase-dead D145N CDK5 expression vector, and 24 h following transfection plated for clonogenic assay with KU0058948. (B) CDK5 is not required for ATM and ATR activation. CAL51 cells transfected 72 h earlier with siRNA where treated with 10-Gy irradiation, 50 J/m² ultraviolet light or not treated. Lystates where made 1 h following treatment, and a western blot was probed with antibodies against ATM Ser1981 autophosphorylation site and CHK1 Ser317 ATM/ATR phosphorylation site, with total ATM, CHK1, TP53 and a β-Tubulin loading control. (C) CDK5 does not affect homologous recombination by gene conversion. Gene conversion was measured using an adapted single-copy, chromosomally integrated HR reporter construct in a 293 human embryonic kidney cell line (Tutt et al, 2001). Silencing of BRCA1 acted as a positive control for reduced gene conversion. Details of this assay are outlined in Supplementary Figure 4. (D) CDK5 silencing does not affect DNA DSB end joining. FIGE assay in CAL51 cells transfected with siCDK5 or siCON. Time points represent minutes after 30 Gy irradiation and non-irradiated control (NIR).