Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Mar 16;19(1):37.
doi: 10.1186/s13059-018-1401-9.

Mutational Signatures Reveal the Role of RAD52 in p53-independent p21-driven Genomic Instability

Affiliations
Free PMC article

Mutational Signatures Reveal the Role of RAD52 in p53-independent p21-driven Genomic Instability

Panagiotis Galanos et al. Genome Biol. .
Free PMC article

Abstract

Background: Genomic instability promotes evolution and heterogeneity of tumors. Unraveling its mechanistic basis is essential for the design of appropriate therapeutic strategies. In a previous study, we reported an unexpected oncogenic property of p21WAF1/Cip1, showing that its chronic expression in a p53-deficient environment causes genomic instability by deregulation of the replication licensing machinery.

Results: We now demonstrate that p21WAF1/Cip1 can further fuel genomic instability by suppressing the repair capacity of low- and high-fidelity pathways that deal with nucleotide abnormalities. Consequently, fewer single nucleotide substitutions (SNSs) occur, while formation of highly deleterious DNA double-strand breaks (DSBs) is enhanced, crafting a characteristic mutational signature landscape. Guided by the mutational signatures formed, we find that the DSBs are repaired by Rad52-dependent break-induced replication (BIR) and single-strand annealing (SSA) repair pathways. Conversely, the error-free synthesis-dependent strand annealing (SDSA) repair route is deficient. Surprisingly, Rad52 is activated transcriptionally in an E2F1-dependent manner, rather than post-translationally as is common for DNA repair factor activation.

Conclusions: Our results signify the importance of mutational signatures as guides to disclose the repair history leading to genomic instability. We unveil how chronic p21WAF1/Cip1 expression rewires the repair process and identifies Rad52 as a source of genomic instability and a candidate therapeutic target.

Keywords: Break-induced replication (BIR); Genomic instability; Rad52; Single nucleotide substitution (SNS); Single strand annealing (SSA); Translesion DNA synthesis (TLS); p21WAF1/Cip1.

Conflict of interest statement

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors are aware of the content and agree with the submission.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Figures

Fig. 1
Fig. 1
Reduction of single nucleotide substitution (SNS) and malfunction of the translesion DNA synthesis and repair (TLS) process upon protracted p21WAF1/Cip1 expression. a Chronic p21WAF1/Cip1 expression, in a p53-deficient environment, leads to the emergence of a subpopulation of p21WAF1/Cip1 aggressive and chemo-resistant (escaped (ESC)) cells, after bypassing an initial senescence-like phase, that carry a lower SNS “load” [9]. SNS identification and filtering were performed with the use of Samtools and VCFtools in non-induced and escaped (Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON) cells (see also Additional file 1: Figure S1), depicted in accompanying histograms (*p < 0.05 (Saos2 and Li-Fraumeni), OFF vs ESC, Welch’s t-test) (for details see “Methods” section). b TLS pathway function. TLS is a DNA damage tolerance process enabling the DNA replication machinery to replicate over DNA lesions. Upon DNA damage PCNA is mono-ubiquitinated, followed by polymerase switch from normal high-fidelity DNA replication polymerases to TLS ones. TLS polymerase Polη, bound to PCNA, inserts a nucleotide opposite to the lesion and, assisted or not by an additional TLS polymerase like Polκ or Polζ, extends beyond the insertion. Finally, a second polymerase switch takes place by substituting TLS polymerases with high-fidelity ones. c Sustained p21WAF1/Cip1 expression results in decreased mono-ubiquitination of PCNA (mono-Ub). Immunoblots (IBs) in 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells (n = 3 experiments). d Reduced binding of Polη to chromatin in cells with protracted p21WAF1/Cip1 expression. IBs after cell fractionation (described in scheme) depicting lower levels of Polη in chromatin extracts from 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells (n = 3 experiments). e Immunofluorescent confocal microscopy (top panel) showing reduced Polκ loading on regions of damaged chromatin after UV-laser ablation in 96-h induced Saos2-p21WAF1/Cip1 Tet-ON cells transfected with a GFP-Polκ vector. Plots (lower panel) depict recruitment kinetics of Polκ in Saos2- and Li-Fraumeni- p21WAF1/Cip1 Tet-ON cells, respectively (see also Additional files 2, 3, 4 and 5). The average intensity of fluorescence at the site of damage and the total cell fluorescence with respect to time were quantified and plotted. Five cells in each condition of three independent experiments were processed. Time frames for obtaining IFs and recruitment plots are depicted in middle panel. f A specific p21WAF1/Cip1 mutant (p21PCNA; harboring Q144, M147, F150 substitutions to A in its PCNA-interacting-protein (PIP) degron motif) with an abrogated interaction with PCNA [9]. IBs depict mono-ubiquitination of PCNA (mono-Ub) in 96-h induced Saos2- and Li-Fraumeni-p21PCNA Tet-ON cells (n = 3 experiments). g Overexpression of p21WAF1/Cip1 decreases TLS repair efficiency across a site-specific lesion in a gapped plasmid TLS assay (i). Induced Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells were assayed for TLS efficiency (ii) and accuracy of repair (iii) with a gap-lesion vector carrying a site-specific benzo[a]pyrene-guanine (BP-G) adduct (i) (Additional file 6: Table S1) (n = 3 experiments). Actin and lamin B serve as loading control (* p < 0.05, error bars indicate SDs). MQ mapping quality, AF allele frequency, DP sequencing depth, MCM Mini-Chromosome Maintenance protein complex, RPA Replication Protein A, WCE whole cell extract, S2 soluble cytosolic, S3 soluble nuclear, P3 chromatin-nuclear matrix
Fig. 2
Fig. 2
Decreased activity of the base excision repair (BER) pathway in cells with sustained p21WAF1/Cip1 expression. a Increased reactive species (RS) levels were assessed with a DCFH-DA assay in Saos2 (i) and Li-Fraumeni (ii) cells with protracted p21WAF1/Cip1 expression (*p < 0.05 (Saos2), *p = 0.05 (Li-Fraumeni), t-test; error bars indicate standard deviation; n = 5 experiments). As shown in the middle panel RS production can lead to generation of base/nucleotide oxidative lesions. b RNAseq analysis showed that essential factors of the BER pathway were statistically significantly down-regulated (p ≤ 0.05) in 96-h induced Saos2- (i) and Li-Fraumeni- (ii) p21WAF1/Cip1 Tet-ON cells (see also Additional file 1: Figure S3 for specific real time RT-PCR validation). Note that although in Saos2- p21WAF1/Cip1 Tet-ON cells OGG1 expression was not found by RNAseq analysis, specific real-time RT-PCR and microarray analysis (see also Additional file 1: Figure S3) [9] confirmed its decreased expression. Selective immunoblots for APEX1, LIG3, TDG, and MUTY confirmed the specificity of the RNA analysis results. Note that LIG3 participates also in mismatch repair (MMR; Additional file 1: Figure S3). α-Tubulin served as loading control. c Modified alkaline Comet assay demonstrated the presence of oxidized purines like 8-oxo-dG in 96-h induced Saos2- (i) and Li-Fraumeni- (ii) p21WAF1/Cip1 Tet-ON cells, using 8-oxoguanine glycosylase (OGG1) (*p < 0.05 (Saos2), *p = 0.05 (Li-Fraumeni), t-test; error bars indicate standard deviation; n = 5 experiments). Comet data were corroborated by an 8-oxo-dG-specific assay measuring DNA incorporation of 8-oxo-dG in p21WAF1/Cip1-expressing cells [24], which indicated lower OGG1 activity (*p < 0.05 (Saos2), *p = 0.05 (Li-Fraumeni), t-test; error bars indicate standard deviation; n = 5 experiments). Consequently, as depicted in the model in the middle panel, recognition and excision of the affected nucleotide lesion is impaired in the BER process. The middle panel depicts the components and steps during BER. The BER pathway is responsible for removal of small lesions from DNA, especially oxidized, alkylated, deaminated bases and abasic sites. BER can be induced by oxidative stress and various genotoxic insults. Its specificity relies on the excision of base damage by glycosylases. In humans, the mechanism of BER involves the initial action of DNA glycosylases followed by the processing of the resulting abasic site either by the AP-lyase activity of the glycosylases or by the apurinic/apyrimidic endonucleases APE1/APE2, which incise the DNA strand. The resulting single-strand break can be processed by two BER subpathways. Either the short-patch branch is engaged, if a single nucleotide is replaced, or the long-patch branch, if 2–10 new nucleotides are synthesized. OGG1 8-oxoguanine DNA glycosylase, UNG uracil DNA glycosylase, TDG thymine DNA glycosylase, SMUG1 single-strand-selective monofunctional uracil-DNA glycosylase 1, NTH DNA glycosylase and apyrimidinic (AP) lyase (endonuclease III), MBD4 methyl-CpG binding domain 4, DNA glycosylase, MPG N-methylpurine DNA glycosylase, MUTY adenine DNA glycosylase, NEIL1/2/3 Nei-like DNA glycosylase 1/ 2/ 3, APEX1/2 apurinic/apyrimidinic endodeoxyribonuclease 1/2, POLB/POLD, DNA polymerase beta/ delta, PCNA proliferating cell nuclear antigen, RFC replication factor C, FEN1 flap structure-specific endonuclease 1, LIG1/LIG3 DNA ligase 1/3, PARP1 poly(ADP-ribose) polymerase 1, XRCC1 X-ray repair cross complementing 1
Fig. 3
Fig. 3
Decreased activity of the nucleotide excision repair (NER) pathway in cells with prolonged p21WAF1/Cip1 expression. a RNAseq analysis showed that essential factors of the NER pathway were statistically significantly down-regulated (p ≤ 0.05; see also Additional file 1: Figure S3 for specific real-time RT-PCR validation) in 96-h induced Saos2- (i) and Li-Fraumeni- (ii) p21WAF1/Cip1 Tet-ON cells. Selective immunoblots for DDB1, ERCC1, ERCC4-XPF, XPC, and ERCC5-XPG, key factors in NER [22], confirming the specificity of the RT-PCR results. α-Tubulin served as loading control; the same protein extracts were used as in Fig. 2. b Decreased repair capacity of N-alkylpurine monoadducts in induced Saos2- (i) and Li-Fraumeni- (ii) p21WAF1/Cip1 Tet-ON cells and treated with monohydroxymelphalan, an inducer of specific NER substrates [–71]. The data shown are based on five independent experiments with at least two analyses for independent experiment/independent experiment experiment (* p < 0.05, error bars indicate SDs). The middle panel depicts the components and function of NER. The NER pathway is responsible for repair of bulky lesions, especially UV-induced thymine dimers and 6,4-photoproducts, as well as non-bulky ones. Following DNA damage recognition, a short single-stranded DNA fragment that contains the lesion is removed. The remaining undamaged single-stranded DNA segment is used by DNA polymerase as a template to synthesize the complementary sequence. Final ligation to complete NER and formation of a double-stranded DNA is carried out by DNA ligase. Depending on how the DNA damage is recognized, NER can be divided into two subpathways: transcription coupled NER (TC-NER) and global genome NER (GG-NER). While the two subpathways differ in how they recognize DNA damage, they share the same process for lesion incision, repair, and ligation. RBX1 Ring-box 1, Cul4 Cullin 4, DDB1/2 Damage specific DNA binding protein 1/2, ERCC8 (CSA) ERCC excision repair 8, CSA ubiquitin ligase complex subunit, ERCC6 (CSB) ERCC excision repair 6, chromatin remodeling factor, USP7 Ubiquitin-specific peptidase 7; ERCC4-XPF Excision repair 4, endonuclease, ERCC5-XPG ERCC excision repair 5, endonuclease, XPA XPA, DNA damage recognition and repair factor, XAB2 XPA binding protein 2, RPA Replication protein A, HMGN1 High mobility group nucleosome binding domain 1; XPC XPC complex subunit, DNA damage recognition and repair factor, RAD23B RAD23 homolog B, CETN2 Centrin 2, CDK7 Cyclin-dependent kinase 7, MNAT1 CDK activating kinase assembly factor, CCNH Cyclin H, TFIIH1–4 Transcription/repair factor IIH 1–4, ERCC3 ERCC excision repair 3, TFIIH core complex helicase subunit, ERCC2 ERCC excision repair 2, TFIIH core complex helicase subunit, TTDA (GTF2H5/TFB5) General transcription factor IIH subunit 5
Fig. 4
Fig. 4
Extended p21WAF1/Cip1 over-expression shapes the mutational signature landscape. a Escaped (30 days induced) Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells exhibit specific patterns of single nucleotide substitution (SNS). SNSs with mapping quality above 30 found only in the escaped cells were filtered based on sequencing depth and scored as ESC-specific (see also Additional file 1: Figure S1). Those SNSs were used to calculate the mutational signature of ESC versus OFF cells (for details see “Methods” section and Additional file 1: Figure S1). Heat map shows the number of mutation type at each mutation context, which was corrected for the frequency of each triplet in the human genome (hg19). Histograms present the mutation-type frequency at each mutation context from two biological replicates of escaped Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells, respectively. Both presentations show reproducible patterns of the mutational signatures 6, 15, 3 [8]. b Heat map showing the association of SNSs, nucleotide insertions (INS), and nucleotide deletions (DEL) with the observed chromosomal breakpoints (±50 kb around the breakpoint) versus the remaining genome in escaped Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells. c Real-time RT-PCR assessment of BRCA1 and BRCA2 mRNA expression in induced and non-induced Saos2 and Li-Fraumeni p21WAF1/Cip1 Tet-ON cells (*p < 0.05 (Saos2), *p = 0.05 (Li-Fraumeni), t-test; error bars indicate standard deviation; n = 3 experiments). Loss of heterozygosity at the q arm of chromosome 13 (which hosts the BRCA2 locus (q13.10)) in induced Li-Fraumeni-p21WAF1/Cip1Tet-ON cells [9]. d Immunoblots depict reduced BRCA1 and BRCA2 expression in induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells at the indicated time points. α-Tubulin served as loading control. MQ mapping quality, AF allele frequency, DP sequencing depth
Fig. 5
Fig. 5
Rad52 increased expression, recruitment, and foci formation at DSBs upon prolonged p21WAF1/Cip1 expression. a Immunofluorescent (IF) analysis showing increased Rad52 foci formation in 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells. Immunoblot (IB) and real-time RT-PCR assessment of Rad 52 expression levels in induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells at the indicated time point (*p < 0.05 (Saos2), *p = 0.05 (Li-Fraumeni), t-test; error bars indicate standard deviation; n = 3 experiments). b IF analysis showing Rad52 and RPA foci formation and their co-localization in 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells. Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells were pre-extracted with ice-cold PBS containing 0.2% Triton X-100 for 2 min on ice before fixation as previously described [52]. c Timelapse microscopy showing Rad52 loading on regions of damaged chromatin after UV-laser ablation in 96-h induced Saos2- and Li-Fraumeni p21WAF1/Cip1 Tet-ON cells transfected with the YFP-Rad52 vector. Plots depict Rad52 recruitment kinetics at sites of DNA damage in the same cells, respectively. The average intensity of fluorescence at the site of damage and the total cell fluorescence in respect to time were quantified and plotted. Five cells in each condition of two independent experiments were processed. d Rad52 promoter is occupied by E2F1 upon p21WAF1/Cip1 induction in Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells, respectively, as assessed by chromatin immunoprecipitation (ChIP; *p < 0.05, t-test; error bars indicate standard deviation; n = 3 experiments; see also Additional file 1: Figure S8). e Silencing of E2F1 resulted in decreased Rad52 levels as assessed by immunoblot analysis in induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells, respectively (*p < 0.01, t-test; error bars indicate standard deviation; n = 3 experiments). Actin serves as loading control; (Ctl) siRNA; arrows indicate Rad52
Fig. 6
Fig. 6
Single nucleotide substitutions (SNSs) cluster around chromosomal breakpoints in cells with continuous p21WAF1/Cip1 expression. a, b Diagrams depict a dense distribution of SNSs (purple triangles) in genome areas surrounding chromosomal breakpoints (green triangles), suggestive of the kataegis phenomenon (intense vertically lined piles of SNSs denoted by the purple triangles), relative to disparate distribution of SNSs in the remaining genome of escaped (Esc-30 days induced), Saos2- (a) and Li-Fraumeni- (b) p21WAF1/Cip1 Tet-ON cells. Note that the total number of chromosomal breakpoints (green triangles) depicted is double as each side of a break corresponds to a different chromosomal arm (Additional file 6: Table S2). The dashed genome areas are depicted as magnifications of representative breakpoints. Green colored dashed areas signify representative breakpoints that show a high positional conservation in all experimental (biological) repetitions. Position of breaks and distribution of SNSs, nucleotide insertions (INS) and nucleotide deletions (DEL) in these examples is depicted in the corresponding subchromosomal magnifications. c, d Histograms depict the clustering frequency of SNSs (c) as well as INS and DEL (d) over all breakpoints in the genome of escaped (Esc-30 days induced) Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells (see also Additional file 6: Table S2)
Fig. 7
Fig. 7
Prolonged p21WAF1/Cip1 expression promotes Rad52-dependent break-induced replication (BIR) and single strand annealing (SSA) repair of DNA double strand breaks (DSBs). a Reduced synthesis-dependent strand annealing (SDSA) in 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1Tet-ON cells. Flow cytometry analysis (FACS) after p21WAF1/Cip1 induction in cells stably expressing a DR-GFP report vector and following I-SceI-induced DSBs shows decreased SDSA activity (*p < 0.05, t-test; error bars indicate standard deviation; n = 5 experiments), regardless of Rad52 silencing. Similar manipulations in Saos2- and Li-Fraumeni-p21PCNA Tet-ON cells showed no differences in SDSA in these cells. b Increased BIR activity in 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells. FACS after p21WAF1/Cip1 induction in cells stably expressing a BIR-GFP report vector and following I-SceI-induced DSBs shows increased BIR activity (*p < 0.05, t-test; error bars indicate standard deviation; n = 5 experiments) that is suppressed upon Rad52 silencing. Similar experiment in Saos2- and Li-Fraumeni-p21PCNA Tet-ON cells showed no effect on BIR function in these cells. c Increased SSA activity in 96-h induced Saos2- and Li-Fraumeni-p21WAF1/Cip1 Tet-ON cells. FACS after p21WAF1/Cip1 induction in cells stably expressing an SA-GFP report vector and following I-SceI-induced DSBs shows increased SSA activity (*p < 0.05, t-test; error bars indicate standard deviation; n = 5 experiments) that is dependent on Rad52. A similar experiment in Saos2- and Li-Fraumeni-p21PCNA Tet-ON cells showed no effect on SSA function in these cells
Fig. 8
Fig. 8
Proposed model depicting how p53-independent p21WAF1/Cip1 expression fuels Rad52-dependent error-prone double strand break repair promoting genomic instability. Sustained p53-independent p21WAF1/Cip1 induction leads to increased levels of nucleotide lesions mediated by elevated reactive oxygen species (ROS). Given the negative impact exerted by p21WAF1/Cip1 on the error free nucleotide repair mechanisms (BER and NER), a significant proportion of such base lesions escape unrepaired. This creates an additional repair “load” to the error prone repair mechanism of TLS, which is further compromised by p21WAF1/Cip1 overexpression, leading to a decreased SNS load and in favor of DSBs. In turn, this further increases the DSB burden generated also through re-replication [9]. As components of SDSA are down-regulated, a shift to Rad52-mediated error prone DNA repair takes place by invoking the BIR and SSA repair routes, fueling genomic instability. This repair switch is mediated by a shift in the balance between Rad51 and Rad52 levels as the former is suppressed by E2F4 [9] and the latter is induced by E2F1 (present study). DSB DNA double strand break, BER base excision repair, NER nucleotide excision repair, TLS translesion DNA synthesis and repair, SDSA synthesis-dependent strand annealing, BIR break-induced repair, SSA single strand annealing

Similar articles

See all similar articles

Cited by 15 articles

See all "Cited by" articles

References

    1. Gorgoulis VG, Vassiliou LV, Karakaidos P, Zacharatos P, Kotsinas A, Liloglou T, et al. Activation of the DNA damage checkpoint and genomic instability in human precancerous lesions. Nature. 2005;434:907–913. doi: 10.1038/nature03485. - DOI - PubMed
    1. Bartkova J, Horejsí Z, Koed K, Krämer A, Tort F, Zieger K, et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis. Nature. 2005;434:864–870. doi: 10.1038/nature03482. - DOI - PubMed
    1. Bartkova J, Rezaei N, Liontos M, Karakaidos P, Kletsas D, Issaeva N, et al. Oncogene-induced senescence is part of the tumorigenesis barrier imposed by DNA damage checkpoints. Nature. 2006;444:633–637. doi: 10.1038/nature05268. - DOI - PubMed
    1. Halazonetis TD, Gorgoulis VG, Bartek J. An oncogene-induced DNA damage model for cancer development. Science. 2008;319:1352–1355. doi: 10.1126/science.1140735. - DOI - PubMed
    1. Negrini S, Gorgoulis VG, Halazonetis TD. Genomic instability--an evolving hallmark of cancer. Nat Rev Mol Cell Biol. 2010;11:220–228. doi: 10.1038/nrm2858. - DOI - PubMed

Publication types

Substances

Feedback