MicroRNAs down-regulate homologous recombination in the G1 phase of cycling cells to maintain genomic stability

  1. Young Eun Choi
  2. Yunfeng Pan
  3. Eunmi Park
  4. Panagiotis Konstantinopoulos
  5. Subhajyoti De
  6. Alan D'Andrea
  7. Dipanjan Chowdhury  Is a corresponding author
  1. Dana-Farber Cancer Institute, Harvard Medical School, United States
  2. University of Colorado School of Medicine, United States
8 figures

Figures

Figure 1 with 1 supplement
miRNA screen for PARP inhibitor sensitivity.

(A) Schematic of a gain-of-function screen using miRNA mimic libraries from Applied Biosystems and Qiagen to identify miRNAs that sensitize cells to the PARP inhibitor, ABT888. (B) Scatter plot (wells/plate) of luminescence (y-axis) as a read-out for viability of each miRNA-transfected well (grey circle) in the presence of ABT888 (20 µM). The plates are numbered in the x-axis. Positive control (BRCA2 siRNA, blue circles) and negative controls (control mimics, pink circles) are shown. Scatter plot for untreated samples is shown in Figure 1—figure supplement 1. (C) List of top miRNAs from the screen displayed in the order of % control viability along with Z-score. (D) Clonogenic survival assay to validate the impact of selected miRNAs on sensitivity to ABT888. MDAMB231 cells were transfected with control miRNA mimics, indicated miRNA mimics, BRCA1 siRNA, or BRCA2 siRNA and treated with vehicle or ABT888, before measuring colony formation. Curves were generated from three independent experiments and a representative image of colony formation with 1 µM ABT888 is shown in the inset. (E and F) Luminascence-based viability assay was performed in MDAMB231 cells with PARP inhibitor, olaparib (E) or in 21NT cells with ABT888 (F). Cells were transfected with control miRNA, indicated miRNA mimics, BRCA1 siRNA, or BRCA2 siRNA and treated with vehicle or PARP inhibitor before ATP quantification. Curves were generated from three independent experiments.

https://doi.org/10.7554/eLife.02445.003
Figure 1—figure supplement 1
miRNAs screen for PARP inhibitor sensitivity.

Scatter plot (wells/plate) of luminescence (y-axis) as a read-out for viability of each miRNA transfected well (grey circle) in the absence of ABT888. The plates are numbered in the x-axis. Positive control (BRCA2 siRNA, blue circles) and negative controls (control mimics, pink circles) are shown.

https://doi.org/10.7554/eLife.02445.004
miRNAs sensitize cells to PARP inhibitors by targeting HR-mediated DSB repair.

(A and B) Measurement of HR-mediated repair of an I-SceI induced site-specific DSB. Cells carrying a single copy of the recombination substrate (DR-GFP) were transfected with control miRNA mimic, indicated miRNA mimics, or BRCA2 siRNA before transfection with I-SceI or control vector. GFP positive cells were analyzed 48 hr later by flow cytometry (FACS). Representative images of the FACS profile are shown in (A), and the mean ± SD of six independent experiments is graphically represented in (B). The dotted line represents the cut-off which was set at 70% of the control. (C) Analysis of HR-mediated repair by RAD51 focus formation. MDA-MB231 cells were transfected with control miRNA mimic, indicated miRNA mimics, or BRCA2 siRNA, stained for RAD51 (green) and 4′,6-diamidino-2-phenylindole (DAPI) (blue) 6 hr after exposure to IR. The images were captured by fluorescence microscopy and RAD51 focus-positive cells (with >5 foci) were quantified by comparing 100 cells. Mean ± SD of three independent experiments is shown in right panel. * indicates p<0.05. (D) γ-H2AX accumulation after treatment with ABT888. Cells were transfected with control miRNA mimic, indicated miRNA mimics, or BRCA2 siRNA and treated with ABT888 (100 μM) before evaluation of γ-H2AX by immunoblotting at indicated time points. Total H2AX served as loading control for these experiments. Images were quantified by ImageJ software and the mean ± SD of three independent experiments is graphically shown.

https://doi.org/10.7554/eLife.02445.005
Figure 3 with 2 supplements
miR-1255b, miR-193b*, and miR-148b* regulate PARP inhibitor sensitivity by regulating expression of HR factors in TNBCs.

(A) miRNA expression profile in a panel of breast cancer lines. Endogenous expression of indicated miRNAs was quantified by qRT-PCR (normalized to 5srRNA) and represented relative to non-tumorigenic breast epithelial cell, HMEC. Expression of miR-1255b, miR-193b*, and miR-148b* were detected in these lines. Mean ± SD of four independent experiments is shown. (BD) Expression of DDR genes is impacted by miR-1255b, miR-193b*, and miR-148b*. MDA-MB231 cells were transfected with control mimic or mimics for miR-1255b, miR-193b*, and miR-148b* and mRNA levels of predicted and prioritized DDR genes were analyzed by qRT-PCR using gene-specific primers and normalized to GAPDH. Mean ± SD of three independent experiments is shown (B). (C and D) Cell lysates were then analyzed by immunoblot for factors which had ≥50% reduction in mRNA in cells transfected with miR-1255b, miR-193b*, and miR-148b*. Images were quantified by ImageJ software and the mean ± SD of three independent experiments is graphically shown, * indicates p<0.05. (EG) Interaction of target transcripts with miR-1255b, miR-193b*, and miR-148b*. MDA-MB231 cells were transfected with biotinylated-control mimics or biotinylated mimics for miR-1255b, miR-193b*, and miR-148b* as a single (F) or a combination (G). The immunoprecipitated RNA was analyzed by qRT-PCR using gene-specific primers and normalized to GAPDH. Mean ± SD of three independent experiments is shown and statistical significance of enrichment of specific gene transcripts is indicated by * (p<0.05). The principle steps of the method are illustrated in Figure 3E.

https://doi.org/10.7554/eLife.02445.006
Figure 3—figure supplement 1
Expression of the excluded miRNAs.

Endogenous expression of the indicated miRNAs that were excluded from selection was quantified by qRT-PCR (normalized to 5srRNA) and represented as relative expression. Mean ± SD of three independent experiments is shown.

https://doi.org/10.7554/eLife.02445.007
Figure 3—figure supplement 2
The effect of miRNAs on cell cycle.

Cell cycle profile was examined after transfection of control, miRNA-1255b, miR-193b*, or miR-148b* and cell population of each phase is graphically represented.

https://doi.org/10.7554/eLife.02445.008
Figure 4 with 1 supplement
Predicted miRNA recognition sites (MREs) of miRNAs and their impact on targets.

(A) Predicted MREs were obtained from PITA (http://genie.weizmann.ac.il/pubs/mir07/mir07_prediction.html) and their mutants were generated by mutating nucleotides providing complementarity and G-U wobble to corresponding miRNAs. The region where MRE is located in the gene is indicated in the parentheses. CDS: coding sequence, 3′UTR: 3′ untraslated region. (B) Luciferase reporter assay to assess direct interaction of miR-1255b, miR-193b*, and miR-148b* with BRCA1, BRCA2, and RAD51. Combinations of predicted miRNA recognition sites (MREs) for each putative target transcript of miR-1255b, miR-193b*, and miR-148b* were cloned into the luciferase reporter vector and transfected in MDA-MB231 cells along with the indicated miRNA mimics. Renilla luciferase activity of the reporter was measured 48 hr after transfection by normalization to an internal firefly luciferase control. Mean ± SD of three independent experiments is shown and statistical significance is indicated by * (p<0.05). (C) Luciferase reporter assay for individual MREs for each target of miRNAs was performed in the same way as described in Figure 4B. Mean ± SD of three independent experiments is shown and statistical significance is indicated by *(p<0.05). (D) Luciferase reporter assay with miR-1255b, miR-193b*, and miR-148b* ANTs. Combinations of predicted miRNA recognition sites (MREs) in the luciferase vector for each putative target transcript of miR-1255b, miR-193b*, and miR-148b* were transfected in MDA-MB231 cells along with the indicated miRNA ANTs. Renilla luciferase activity of the reporter was measured 48 hr after transfection by normalization to an internal firefly luciferase control. Mean ± SD of three independent experiments is shown and statistical significance is indicated by *(p<0.05).

https://doi.org/10.7554/eLife.02445.009
Figure 4—figure supplement 1
Conservation of predicted miRNA recognition sites (MREs) of miRNAs.

Predicted MRE sequences in each miRNA target genes were aligned across different species.

https://doi.org/10.7554/eLife.02445.010
Figure 5 with 1 supplement
Impact of miRNAs on DSB repair in different phases of the cell cycle.

(A) Rescue of the impact of miRNAs on ABT888 sensitivity. MDAMB231 cells were transfected with control miRNA or indicated miRNA mimics with or without target gene cDNAs (lacking MREs) and treated with vehicle or ABT888, before viability assay by ATP quantification. Expression of each target protein is examined by immune blot. (B and C) Expression of miRNAs and target transcripts in synchronized cells. MDAMB231 (B) or MCF10A (C) cells were synchronized with mimosine and the relative amount of miR-1255b, miR-193b*, and miR-148b* or BRCA1, BRCA2, and RAD51 mRNA for G1- or S-phase was determined by qRT-PCR (normalized to RNU1 or GAPDH, respectively). Mean ± SD of three independent experiments is shown and statistical significance is indicated by *(p<0.05). (DF) Impact of inhibiting miRNAs on targets in G1 cells. MDAMB231 cells were transfected with control ANT or ANTs for miR-1255b, miR-193b*, and miR-148b* as a single (D) or a combination (F). Subsequently, the cells were synchronized with mimosine and BRCA1, BRCA2, and RAD51 mRNA was assessed by qRT-PCR (normalized to GAPDH) in the G1 and/or S-phase (D and F). Cell lysates from G1 cells were analyzed by immunblot for BRCA1, BRCA2, and RAD51 (E). Images were quantified by ImageJ software and the mean ± SD of three independent experiments is shown, * indicates p<0.05.

https://doi.org/10.7554/eLife.02445.011
Figure 5—figure supplement 1
The impact of miRNA antagomirs (ANTs) on cell cycle progression.

MDAMB231 cells or MCF10A cell (data not shown) were transfected with control antagomir or antagomirs for miR-1255b, miR-193b*, and miR-148b* for 48 hr. Subsequently, the cells were synchronized with 500 µM mimosine for 24 hr and collected at indicated times after release into growth media. The cell cycle distribution was analyzed by FloJo and representative images of the cell cycle profile from three independent experiments are shown.

https://doi.org/10.7554/eLife.02445.012
Figure 6 with 1 supplement
Impact of inhibiting miRNAs on DSB repair.

(AD) Impact of inhibiting miRNAs on DSB repair in the G1 phase of MDA-MB231 cells. Cells were transfected with control ANT or ANTs for miR-1255b, miR-193b*, and miR-148b* with or without 20 nM CtIP siRNA, exposed to IR (5 Gy) and stained for γ-H2AX (green) (A) or RPA2 (green) (C), cyclin A (red) and 4′,6-diamidino-2-phenylindole (DAPI) (blue). The images were captured by fluorescence microscopy and the number of γ-H2AX foci (A) or RPA2 foci (C) was calculated from 100 cells. Mean ± SD of three independent experiments is graphically represented (B and D). * indicates p<0.05. (EI) Impact of inhibiting miRNAs on DSB repair in different phase of RPE-1 cells. RPE-1 cells expressing the Fucci system (illustrated in Figure 6E) were transfected with control ANT or ANTs for miR-1255b, miR-193b*, and miR-148b* with or without 20 nM CtIP siRNA, exposed to IR (5 Gy) and stained for γ-H2AX (green) and 4′,6-diamidino-2-phenylindole (DAPI) (blue). The images were captured by fluorescence microscopy and the number of γ-H2AX foci in G1 cells (red, mKO2-Cdt1) was calculated from 100 cells (F and G). The images were captured by fluorescence microscopy and the number of γ-H2AX foci (green foci) in S/G2/M cells (green background, mAG-Geminin) was calculated from 100 cells (H). Mean ± SD of three independent experiments is graphically represented (I). * indicates p<0.05.

https://doi.org/10.7554/eLife.02445.013
Figure 6—figure supplement 1
The impact of miRNAs on DNA repair during cell cycle.

(A) MDAMB231 cells were transfected with control ANT or ANTs for miR-1255b, miR-193b*, and miR-148b* and stained for γ-H2AX (green), cyclin A (red) and 4′,6-diamidino-2-phenylindole (DAPI) (blue). The images were captured by fluorescence microscopy and the number of γ-H2AX foci was calculated from 100 cells each in G1 phase and S/G2 phase. Mean ± SD of three independent experiments is shown in the right panel. * indicates p<0.05. (B) MDAMB231 cells were transfected with control ANT or ANTs for miR-1255b, miR-193b*, and miR-148b* with or without 20 nM CtIP siRNA. CtIP mRNA expression was assessed in control ANT ± CtIP siRNA transfected MDAMB231 cells by qPCR and normalized to GAPDH. * indicates p<0.05.

https://doi.org/10.7554/eLife.02445.014
Figure 7 with 1 supplement
Correlation of LOH with loss of miRNAs.

(A) miRNA expression profile in a panel of ovarian cancer lines. Endogenous expression of indicated miRNAs was quantified by qRT-PCR (normalized to 5srRNA) and represented relative to non-tumorigenic ovarian epithelial cell, HIO-80. Expression of miR-1255b, miR-193b*, and miR-148b* were detected in these lines. (B and C) Correlation of LOH with deletion of miRNAs in TCGA data set. Box plots show the frequency of (B) LOH or (C) somatic copy number amplification or deletion (SCNΔ) in the 418 high-grade serous ovarian tumors from TCGA that have no amplifications or deletions of any of these 3 miRNAs (WT), against those with deletion of 1255b (either −1 or −2), 148b* or 193b*. The LOH events are >1 Mb. (D) Correlation of LOH with deletion of miRNAs in DF/HCC data set. Box plot shows the frequency of LOH in 47 high-grade serous ovarian tumors that have no amplifications or deletions of any of these 3 miRNAs (WT), against those with deletion of miR-1255b (either 1 or 2), miR-148b* or miR-193b*. The LOH events are >1 Mb. (E) Correlation of BRCA1 expression with deletion of miRNAs in TCGA data set. Box plot shows expression levels of BRCA1 in the 418 high-grade serous ovarian tumors from TCGA that have no amplifications or deletions of any of these 3 miRNAs (WT), against those with deletion of miR-1255b (either 1 or 2), miR-148b* or miR-193b*. In (BE), statistical significance was calculated using one tailed Mann Whitney's U test and significant differences (p<0.05) indicated with asterisk. Width of the bars indicates the number of samples. The horizontal line indicates the median of the WT set.

https://doi.org/10.7554/eLife.02445.015
Figure 7—figure supplement 1
miRNA dependent regulation of DSB repair during cell cycle.

Rescue of the impact of miRNA inhibition on DNA repair in ovarian cancer cells. UWB1.289 BRCA1-deficient ovarian cancer cells were transfected with control ANT or ANTs for miR-1255b, miR-193b*, and miR-148b* and stained for γ-H2AX (green), cyclin A (red) and 4′,6-diamidino-2-phenylindole (DAPI) (blue). The images were captured by fluorescence microscopy and the number of γ-H2AX foci was calculated from 100 cells. Mean ±SD of 3 independent experiments is shown in the right panel. * indicates p <0.05.

https://doi.org/10.7554/eLife.02445.016
A model of miRNA dependent regulation of DSB repair during cell cycle.

Model of miRNA-dependent regulation of DSB repair during cell cycle. The balance of HR and NHEJ in dividing cells is crucial for efficient DSB repair. NHEJ is the preferred pathway in the G1 phase with 53BP1 and the Ku complex binding the broken DNA end. miRNAs (such as miR-1255b and miR-193b*) suppress HR factors, particularly BRCA1, preventing end resection of the DNA lesions. However, when these miRNAs are inhibited or deleted it may disrupt the correct choice of DSB repair pathway. Ectopic over expression of BRCA1 will allow CtIP-mediated resection in G1 cells, preventing NHEJ. Furthermore, HR-mediated repair in G1 is detrimental to cell health as it would lead to LOH. In S-phase, DSBs are predominantly repaired by HR and down-regulation of miRNAs targeting BRCA1, BRCA2 and RAD51 may be important in ensuring efficient HR-mediated DSB repair. Over-expression of miRNAs (such as miR-1255b, miR-193b*, and miR-148b*) targeting BRCA1, BRCA2 and RAD51 in the S phase will impede various steps of HR, and the HR deficiency will sensitize these cells to PARP inhibitors.

https://doi.org/10.7554/eLife.02445.017

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Young Eun Choi
  2. Yunfeng Pan
  3. Eunmi Park
  4. Panagiotis Konstantinopoulos
  5. Subhajyoti De
  6. Alan D'Andrea
  7. Dipanjan Chowdhury
(2014)
MicroRNAs down-regulate homologous recombination in the G1 phase of cycling cells to maintain genomic stability
eLife 3:e02445.
https://doi.org/10.7554/eLife.02445