Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jul 18;114(29):7671-7676.
doi: 10.1073/pnas.1620208114. Epub 2017 Jul 3.

MRG15-mediated Tethering of PALB2 to Unperturbed Chromatin Protects Active Genes From Genotoxic Stress

Affiliations
Free PMC article

MRG15-mediated Tethering of PALB2 to Unperturbed Chromatin Protects Active Genes From Genotoxic Stress

Jean-Yves Bleuyard et al. Proc Natl Acad Sci U S A. .
Free PMC article

Abstract

The partner and localiser of BRCA2 (PALB2) plays important roles in the maintenance of genome integrity and protection against cancer. Although PALB2 is commonly described as a repair factor recruited to sites of DNA breaks, recent studies provide evidence that PALB2 also associates with unperturbed chromatin. Here, we investigated the previously poorly described role of chromatin-associated PALB2 in undamaged cells. We found that PALB2 associates with active genes through its major binding partner, MRG15, which recognizes histone H3 trimethylated at lysine 36 (H3K36me3) by the SETD2 methyltransferase. Missense mutations that ablate PALB2 binding to MRG15 confer elevated sensitivity to the topoisomerase inhibitor camptothecin (CPT) and increased levels of aberrant metaphase chromosomes and DNA stress in gene bodies, which were suppressed by preventing DNA replication. Remarkably, the level of PALB2 at genic regions was frequently decreased, rather than increased, upon CPT treatment. We propose that the steady-state presence of PALB2 at active genes, mediated through the SETD2/H3K36me3/MRG15 axis, ensures an immediate response to DNA stress and therefore effective protection of these regions during DNA replication. This study provides a conceptual advance in demonstrating that the constitutive chromatin association of repair factors plays a key role in the maintenance of genome stability and furthers our understanding of why PALB2 defects lead to human genome instability syndromes.

Keywords: DNA replication; MRG15; PALB2; SETD2; transcription.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.
Fig. 1.
MRG15, SETD2, and H3K36me3 are required for PALB2 chromatin association in unperturbed cells. (A) Volcano plot showing FLAG-EGFP-PALB2 (FE-PALB2)-associated proteins identified by MS. The log2 of the fold change (FC) over a FLAG-EGFP control was plotted against negative log10 P values (n = 3). (B) Schematic of the pathway linking transcription and PALB2. (C) WB analysis of PALB2 in the chromatin fraction of U2OS cells with CRISPR SETD2 knockout (Left), KDM4A overexpression (OE) (Middle), and wild type (WT) or mutant (K36M) histone variant H3.3 overexpression (Right). Asterisks indicate endogenous H3 (black) and exogenous H3.3 (red). (D) Venn diagram showing the overlap of H3K36me3-containing and PALB2-bound genes. (E) The averaged profile of H3K36me3 and PALB2 around gene loci with H3K36me3 peaks. Read density (y axis) is normalized for total number of mapped reads. (F) ChIP-qPCR quantification of FLAG-PALB2 (Left) and H3K36me3 (Right). Mean values ± SD (n = 3, with triplicate qPCR reactions). Where indicated, cells were treated with α-Am. Asterisks indicate two-tailed paired Student's t test; *P < 0.05, **P < 0.01, ***P < 0.001.
Fig. 2.
Fig. 2.
Ablation of PALB2 binding to MRG15 confers an increased level of DNA stress in unperturbed cells. (A) Schematic of PALB2 variants depicting point mutations (red bars) disrupting each domain. (B) Anti-FLAG IPs from EUFA1341 cells stably expressing FLAG-PALB2 variants. Lamin A was used as a negative control. (C) WB of indicated proteins in the chromatin fraction of EUFA1341 cells stably expressing FLAG-PALB2 variants. (D) The levels of PALB2 in the chromatin fraction of cells expressing FLAG-PALB2 variants were quantified and, following normalization against their respective levels in whole-cell extract, expressed as % of WT. Mean values ± SD (n = 3). (E) ChIP-qPCR quantification of FLAG-PALB2 variants at the ACTB gene. (F) WB showing the levels of γH2A.X in untreated EUFA1341 cells complemented with FLAG-PALB2 variants. (G) The number of γH2A.X foci in U2OS Flp-In T-REx cells stably expressing FLAG-PALB2 variants. Where indicated, cells were treated with 2 μg/mL doxycycline (Dox) for 5 d to induce shRNA targeting endogenous PALB2. Dots represent individual cells and bars mean values ± SD. Statistical significance was determined using two-tailed paired Student’s t test (D) or Mann–Whitney U test (G). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. EV, empty vector; ns, nonsignificant.
Fig. S1.
Fig. S1.
(Related to Fig. 1) (A) WB analysis of chromatin-associated FLAG-PALB2 following treatment of U2OS cells with siRNA targeting BRCA1. (B) WB showing chromatin-associated PALB2 in HT-1080 cells treated with siRNA targeting indicated genes. Arrowheads denote specific bands. (C) Quantification of PALB2 and BRCA2 levels in the chromatin fraction of HT-1080 cells treated with the indicated siRNA. Mean values ± SD (n = 2). (D) Schematic diagram of the MRG15 and MRGX proteins. The chromodomain (green), MRG domain (red), rubella capsid-like domain (brown), and point mutations (red bars) are depicted. (E) Nucleosome pull-down assay using affinity-purified GFP-fusion of ChAM or MRG15 variants and partially purified human nucleosomes. The presence in the pull-down samples of histone H3, H3K36me3, and monoubiquitinated H2B (H2BmUB) was examined by WB. (F) Genome-wide PALB2 occupancy in S-phase HeLa cells was determined by ChIP-seq, and PALB2 peaks were categorized as upstream (≤5 kb upstream of gene bodies), downstream (≤5 kb downstream of gene bodies), and intergenic. (G) Snapshots of PALB2 ChIP-seq aligned with the expression levels of genes (Caltech RNA-seq; HeLa-S3 whole cells) and histone modifications (Broad HeLa-S3; H3K4me3, H3K27me3, H3K36me3) at the UCSC Genome Browser on Human Feb. 2009 (GRCh37/hg19) assembly.
Fig. S2.
Fig. S2.
(Related to Fig. 1) (A) WB analysis of chromatin-associated FLAG-PALB2 WT in EUFA1341 cells following 17 h of treatment with the indicated concentrations of α-Am. The levels of FLAG-PALB2, BRCA2, and Rad51 in the chromatin fraction were quantified and, following normalization against their respective levels in whole-cell extract, expressed as % of untreated cells. (B) The 2,050 genes containing PALB2-binding sites were categorized by GSEA, and the proportion of periodic and nonperiodic genes, as identified by Dominguez et al. (30), is shown in each category. (C) A strip-plot showing the average expression level of PALB2-bound genes during the cell cycle. Red and blue dots, respectively, indicate periodic and nonperiodic genes. (D) As in B, but the ratio of k (number of genes containing PALB2-binding sites) against K (number of genes consisting of a given hallmark gene set) was plotted against negative log10 P values. The size of each circle indicates the respective k value. (E) GSEA of 1,183 periodic genes as identified by Dominguez et al. (30). (F) The GSEA of the top 25% highly expressed genes (2,204 genes), as detected by RNA-seq of synchronized HeLa cells. The level of average gene expression was calculated as mean of FPKM (fragments per kilobase of transcript per million mapped reads) at each time point during the cell cycle. (G) The GSEA of the top 25% highly expressed genes in S phase (2,204 genes), as in F but at 3 h after the thymidine release.
Fig. S3.
Fig. S3.
(Related to Fig. 2) (A) Schematic diagram of the PALB2 protein depicting the position of the coiled-coil domain (blue), the ChAM (orange), the MBD (green), and the WD40-repeat domain (purple). Protein sequence alignments for ChAM and conserved regions MBD-I and MBD-II are also presented. Identical and conserved residues are, respectively, highlighted in green and blue. Asterisks indicate residues mutated in this study. ChAM helical secondary structure as predicted by JPred 2 and the FxLP motif within MBD-I are indicated at the top and bottom, respectively. (B) FLAG-PALB2 variants with a MBD-I deletion or indicated point mutations were expressed in HEK293T cells and immunoprecipitated from whole-cell lysates. WB detection of the indicated proteins was performed to assess co-IP. (C) The indicated FLAG-PALB2 variants were expressed in HEK293T cells and immunoprecipitated from whole-cell lysates. WB detection of MRG15 was performed to assess co-IP. (D) GFP-ChAM peptides, with the indicated mutations, were transiently expressed in HEK293T cells and their chromatin association was determined following chemical cell fractionation. Lamin A and histone H3 are markers for extraction of nuclear and chromatin-associated proteins, respectively. (E) Pull-down assay using affinity-purified GFP-ChAM and recombinant mononucleosomes. Salmon sperm DNA was included to outcompete nonspecific interactions. (F) WB showing the levels of γH2A.X in U2OS Flp-In T-REx cells stably expressing FLAG-PALB2 variants, following down-regulation of endogenous PALB2 using shRNA induced by 2 µg/mL doxycycline (Dox). Arrowhead indicates PALB2-specific band. (G) Levels of γH2A.X on day 4 were quantified and expressed as % of untreated. Mean values ± SD (n = 3). EV, empty vector.
Fig. 3.
Fig. 3.
Ablation of PALB2 binding to MRG15 confers hypersensitivity to CPT. (A) The IC50 for MMC, CPT, APH, and HU were determined by WST-1 assay. n = 3, with two technical replicates. Error bars indicate 95% CI. (B) Quantification of RAD51-positive cells (% of cells with ≥10 foci per nucleus). Mean values ± SD (n ≥ 3, >210 nuclei scored per repeat). (C) Quantification of RAD51 and PALB2 colocalization (% of total cells with ≥5 overlapping foci per nucleus). Mean values ± SD (n ≥ 3, >210 nuclei scored per repeat). (D) Quantification of the number of RAD51 foci per cell. Where indicated, cells were treated with CPT or α-Am. Dots represent individual cells and bars mean values ± SD. Statistical significance was determined using the extra sum-of-squares f test (A), two-tailed paired Student’s t test (B and C), or Mann–Whitney U test (D). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, nonsignificant.
Fig. S4.
Fig. S4.
(Related to Fig. 3) Survival curves for EUFA1341 cells complemented with FLAG-PALB2 WT or -MBD mutant following treatment with indicated concentrations of MMC (A) or CPT (B). Mean values ± SD (n = 3). (C) EUFA1341 cells complemented with FLAG-PALB2 variants were stained for RAD51 and FLAG-PALB2 foci following 17 h of treatment with DMSO or 10 nM CPT. For each cell line used in this study, representative pictures of RAD51 (purple) and FLAG-PALB2 (yellow) foci are shown. (D) Representative pictures of RAD51 foci (green) in EUFA1341 cells complemented with WT FLAG-PALB2. Where indicated, cells were treated for 17 h with DMSO, 4 µg/mL α-Am, or 10 nM CPT. (Scale bars, 10 μm.)
Fig. 4.
Fig. 4.
PALB2 interaction with MRG15 protects active genes during DNA replication. (A) WB showing γH2A.X levels in EUFA1341 cells expressing FLAG-PALB2 variants. (B) Workflow of metaphase chromosome spread preparation and examples of chromosomal aberrations. (C) Aberrant chromosomes per metaphase cell. Dots represent individual cells and bars mean values ± SD (n = 40). P values are for Mann–Whitney U test. (D) Diagram depicting the mechanism of TOP1 poisoning by CPT, stabilizing TOP1 cleavage complexes (TOP1cc) at active genes. (E) Genome-wide distribution of γH2A.X in EUFA1341 cells expressing PALB2 WT or -MBD mutant, treated with CPT, was determined by ChIP-seq. γH2A.X peaks were categorized as upstream (≤5 kb upstream of gene bodies), downstream (≤5 kb downstream of gene bodies), or intergenic. (F) ChIP-qPCR quantification of γH2A.X (γH2A.X/H2A.X ratio). Where indicated, EUFA1341 cells expressing FLAG-PALB2 variants were treated with CPT alone or in combination with APH. Mean values ± SD (n = 3, with triplicate qPCR reactions).
Fig. S5.
Fig. S5.
(Related to Fig. 4) (A) HeLa Kyoto cells were synchronized by double thymidine block and released for the indicated time. The chromatin association of PALB2, BRCA2, BRCA1, RAD51, MRG15, and MRGX and the level of H3K36me3 were examined following fractionation of a synchronized cell population. PCNA and H3-S10p are markers for S phase and mitosis, respectively. Arrowheads indicate the specific band. (B) Venn diagram showing genes containing γH2A.X peaks in WT–expressing cells or in -MBD–expressing cells and PALB2-bound genes. (C–E) GSEA of PALB2-bound genes containing γH2A.X peaks only in -MBD–expressing cells (cat. A) (C), in both WT and -MBD expression cells (cat. B) (D), or only in WT-expressing cells (cat. C) (E). The ratio of k (number of genes containing PALB2-binding sites) against K (number of genes consisting of a given hallmark gene set) was plotted against negative log10 P values. The size of each circle indicates respective k value. (F) Schematic diagram of the beta actin (ACTB) and TCOF1 genes. Boxes and blue bars indicate exons and pairs of primers, respectively. (G–I) γH2A.X levels at the ACTB (G), TCOF1 (H), and IRF2BP2 (I) loci were analyzed by ChIP-qPCR. Where indicated, EUFA1341 cells stably expressing FLAG-PALB2 WT or -MBD mutant were treated for 17 h with 10 nM CPT alone or in combination with 0.5 µM APH. γH2A.X intensity is shown as the γH2A.X/H2A.X ChIP-qPCR signal ratio. Mean values ± SD (n = 3, with triplicate qPCR reactions). P values are for the two-tailed paired Student’s t test. (J–M) Same as G–I, except γH2A.X levels at the ACTB (J), IRF2BP2 (K), TCOF1 (L), and WEE1 (M) loci were analyzed by ChIP-qPCR following 17 h of treatment with 10 nM CPT alone or in combination with 4 µg/mL α-Am. Mean values ± SD (n = 3, with triplicate qPCR reactions).
Fig. 5.
Fig. 5.
Chromatin-associated PALB2 is mobilized in response to CPT treatment. (A) ChIP-qPCR quantification of FLAG-PALB2 variants. Mean values ± SD (n = 3, with triplicate qPCR reactions). Where indicated, EUFA1341 cells expressing FLAG-PALB2 variants were treated with CPT. Asterisks indicate two-tailed paired Student's t test; *P < 0.05, **P < 0.01. (B) Pull-down experiment showing the interaction between chromatin-associated PALB2 and MRG15 upon CPT treatment. (C) Quantification of BRCA2, MRG15, MRGX, and RAD51 levels in FLAG-PALB2 WT IP. Mean values ± SD (n = 2). (D) Proposed function of constitutive PALB2 chromatin association.
Fig. S6.
Fig. S6.
(Related to Fig. 5) (A–C) Occupancy of FLAG-PALB2 at ACTB (A), TCOF1 (B), and IRF2BP2 (C) loci was analyzed by FLAG ChIP-qPCR. FLAG-PALB2 signal is shown as the FC over the IgG control. Where indicated, EUFA1341 cells stably expressing FLAG-PALB2 WT or -MBD mutant were treated for 17 h with 10 nM CPT. Mean values ± SD (n = 3, with triplicate qPCR reactions). Statistical significance was determined using the two-tailed paired Student’s t test. *P < 0.05, **P < 0.01. (D) The chromatin association of the indicated proteins and the levels of H3K36me3 were examined following fractionation of CPT-treated EUFA1341 cells complemented with WT FLAG-PALB2. (E) MRG15 occupancy at the ACTB, IRF2BP2, TCOF1, and WEE1 loci was determined by ChIP-qPCR. MRG15 signal is shown as the FC over the IgG control. Mean values ± SD (n = 3, with triplicate qPCR reactions). Where indicated, EUFA1341 cells stably expressing WT FLAG-PALB2 were treated for 17 h with 10 nM CPT.

Similar articles

  • MRG15 binds directly to PALB2 and stimulates homology-directed repair of chromosomal breaks.
    Hayakawa T, Zhang F, Hayakawa N, Ohtani Y, Shinmyozu K, Nakayama J, Andreassen PR. Hayakawa T, et al. J Cell Sci. 2010 Apr 1;123(Pt 7):1124-30. doi: 10.1242/jcs.060178. J Cell Sci. 2010. PMID: 20332121 Free PMC article.
  • Exploring the link between MORF4L1 and risk of breast cancer.
    Martrat G, Maxwell CM, Tominaga E, Porta-de-la-Riva M, Bonifaci N, Gómez-Baldó L, Bogliolo M, Lázaro C, Blanco I, Brunet J, Aguilar H, Fernández-Rodríguez J, Seal S, Renwick A, Rahman N, Kühl J, Neveling K, Schindler D, Ramírez MJ, Castellà M, Hernández G; EMBRACE, Easton DF, Peock S, Cook M, Oliver CT, Frost D, Platte R, Evans DG, Lalloo F, Eeles R, Izatt L, Chu C, Davidson R, Ong KR, Cook J, Douglas F, Hodgson S, Brewer C, Morrison PJ, Porteous M, Peterlongo P, Manoukian S, Peissel B, Zaffaroni D, Roversi G, Barile M, Viel A, Pasini B, Ottini L, Putignano AL, Savarese A, Bernard L, Radice P, Healey S, Spurdle A, Chen X, Beesley J; kConFab, Rookus MA, Verhoef S, Tilanus-Linthorst MA, Vreeswijk MP, Asperen CJ, Bodmer D, Ausems MG, van Os TA, Blok MJ, Meijers-Heijboer HE, Hogervorst FB; HEBON, Goldgar DE, Buys S, John EM, Miron A, Southey M, Daly MB; BCFR; SWE-BRCA, Harbst K, Borg A, Rantala J, Barbany-Bustinza G, Ehrencrona H, Stenmark-Askmalm M, Kaufman B, Laitman Y, Milgrom R, Friedman E, Domchek SM, Nathanson KL, Rebbeck TR, Johannsson OT, Couch FJ, Wang X, Fredericksen Z, Cuadras D, Moreno V, Pientka FK, Depping R, Caldés T, Osorio A, Benítez J, Bueren J, Heikkinen T, Nevanlinna H, Hamann U, Torres D, Caligo MA, Godwin AK, Imyanitov EN, Janavicius R; GEMO Study Collaborators, Sinilnikova OM, Stoppa-Lyonnet D, Mazoyer S, Verny-Pierre C, Castera L, de Pauw A, Bignon YJ, Uhrhammer N, Peyrat JP, Vennin P, Ferrer SF, Collonge-Rame MA, Mortemousque I, McGuffog L, Chenevix-Trench G, Pereira-Smith OM, Antoniou AC, Cerón J, Tominaga K, Surrallés J, Pujana MA. Martrat G, et al. Breast Cancer Res. 2011 Apr 5;13(2):R40. doi: 10.1186/bcr2862. Breast Cancer Res. 2011. PMID: 21466675 Free PMC article.
  • Breast cancer-associated missense mutants of the PALB2 WD40 domain, which directly binds RAD51C, RAD51 and BRCA2, disrupt DNA repair.
    Park JY, Singh TR, Nassar N, Zhang F, Freund M, Hanenberg H, Meetei AR, Andreassen PR. Park JY, et al. Oncogene. 2014 Oct 2;33(40):4803-12. doi: 10.1038/onc.2013.421. Epub 2013 Oct 21. Oncogene. 2014. PMID: 24141787 Free PMC article.
  • PALB2: the hub of a network of tumor suppressors involved in DNA damage responses.
    Park JY, Zhang F, Andreassen PR. Park JY, et al. Biochim Biophys Acta. 2014 Aug;1846(1):263-75. doi: 10.1016/j.bbcan.2014.06.003. Epub 2014 Jul 3. Biochim Biophys Acta. 2014. PMID: 24998779 Free PMC article. Review.
  • The Role of PALB2 in the DNA Damage Response and Cancer Predisposition.
    Nepomuceno TC, De Gregoriis G, de Oliveira FMB, Suarez-Kurtz G, Monteiro AN, Carvalho MA. Nepomuceno TC, et al. Int J Mol Sci. 2017 Aug 31;18(9):1886. doi: 10.3390/ijms18091886. Int J Mol Sci. 2017. PMID: 28858227 Free PMC article. Review.
See all similar articles

Cited by 11 articles

See all "Cited by" articles

Publication types

MeSH terms

LinkOut - more resources

Feedback