PAXX and its paralogs synergistically direct DNA polymerase λ activity in DNA repair

Abstract

PAXX is a recently identified component of the nonhomologous end joining (NHEJ) DNA repair pathway. The molecular mechanisms of PAXX action remain largely unclear. Here we characterise the interactomes of PAXX and its paralogs, XLF and XRCC4, to show that these factors share the ability to interact with DNA polymerase λ (Pol λ), stimulate its activity and are required for recruitment of Pol λ to laser-induced DNA damage sites. Stimulation of Pol λ activity by XRCC4 paralogs requires a direct interaction between the SP/8 kDa domain of Pol λ and their N-terminal head domains to facilitate recognition of the 5' end of substrate gaps. Furthermore, PAXX and XLF collaborate with Pol λ to promote joining of incompatible DNA ends and are redundant in supporting Pol λ function in vivo. Our findings identify Pol λ as a novel downstream effector of PAXX function and show XRCC4 paralogs act in synergy to regulate polymerase activity in NHEJ.

Fig. 1

Comparative Analysis of the Proteomes of XRCC4 Family Proteins isolated from Soluble Chromatin. Interactome analysis using FLAG-tagged PAXX, -XLF, -XRCC4 and -DNA-PKcs as bait from the benzonase-treated soluble chromatin fraction of HEK293F cells was performed using Cytoscape. Proteins highlighted by shaded yellow boxes indicate proteins which interact with the indicated bait protein e.g. PAXX, XLF, XRCC4 or DNA-PKcs. Shaded grey boxes depict proteins identified in combined proteomes of XRCC4 family proteins and DNA-PKcs (PRKDC)

Fig. 2

Comparative Analysis of the Nucleoplasmic Proteomes of XRCC4 Family Proteins. As described in Fig. 1, except the nucleoplasmic fraction was analysed

Fig. 3

Pol λ interacts with XRCC4 family proteins via its BRCT domain in cells. a HEK293F cells were irradiated with 10 Gy X-ray or left untreated. Soluble nuclear extracts were isolated following 0–60 min post-irradiation recovery time at 37 °C. Following IP with anti-Pol λ or rabbit IgG (rIgG), Pol λ and associated proteins were resolved by SDS-PAGE and immunoblotted for the indicated NHEJ factors. b HEK293F cell nucleoplasmic (NP) or soluble chromatin (sol. Chr) extracts were immunoprecipitated with rIgG, anti-PAXX or -XLF or mouse IgG (mIgG) or anti-XRCC4. Immunoprecipitated proteins were resolved by SDS-PAGE and immunoblotted for the indicated NHEJ factors. c As described in Panel A, except that soluble nuclear extracts were incubated with 0-200 μg/ml EtBr for 1 h prior to IP with anti-Pol λ or rIgG. d EMSA showing that interaction of Pol λ with DNA-bound Ku requires R57 and L60 in the BRCT domain of Pol λ. Reactions were performed with IRDye® 700-labelled 5nt-gapped dsDNA (33-mer) in the presence or absence of FLAG-Ku70/80 (20 nM) and either FLAG-Pol λ-WT or a R57A/L60A mutant (50 nM). e HEK293F cells were transiently transfected with either pCMX-LacZ (control) or pCMX-FLAG-Pol λ-WT, -ΔBRCT or a R57A/L60A mutant and anti-FLAG IPs performed

Fig. 4

Interaction of PAXX and XLF with Pol λ requires C-terminal Ku-binding regions. a EMSA showing that interaction of PAXX with Pol λ requires DNA-bound Ku. Reactions were performed with 10 or 20 nM IRDye^® 700-labelled 5nt-gapped dsDNA (90-mer) and the following concentrations of FLAG-Ku70/80 (20 nM), FLAG-Pol λ (40 nM) or cleaved PAXX (100 nM). b Binding of PAXX to DNA-bound Ku requires C-terminal residues V199 and F201 of PAXX. Reactions were performed with 20 nM IRDye^® 700-labelled 5nt-gapped dsDNA (90-mer) and the indicated concentrations of FLAG-PAXX-WT or a FLAG-PAXX-V199A/F201A mutant and FLAG-Ku70/80 (20 nM). c As described in Panel A, except that reactions contained FLAG-Ku (20 nM), FLAG-Pol λ (100 nM), FLAG-PAXX-WT (2.5 μM, left panel) or a V199A/F201A mutant (2.5 μM, right panel). d As described in Panel A, except that reactions contained FLAG-Ku (20 nM), FLAG-Pol λ (200 nM), FLAG-XLF-WT (2.5 μM, left panel) or a C-terminal FLAG-XLF (aa1-233) deletion mutant (2.5 μM, right panel)

Fig. 5

Role of XRCC4 family members in the recruitment of Pol λ to laser microirradiation-induced DNA damage sites. a Upper, Schematic figure showing N-terminal EGFP- and mCherry-Pol λ fusion proteins; Lower, Representative immunofluorescence images showing that N-terminal EGFP- and mCherry-Pol λ fusion proteins are localised to nuclei in U2OS cells. b Recruitment of N-terminal EGFP-Pol λ to laser-induced DNA damage sites in U2OS cells. c Time course of N-EGFP-Pol λ recruitment to laser-induced DNA damage sites in U2OS cells. Data shown are the mean and SEM from 16 individual cells. d Immunoblot analysis of N-EGFP-Pol λ expressing U2OS cells deficient in PAXX, XLF or XRCC4 generated by CRISPR-Cas9. WCL were resolved by SDS-PAGE and the indicated proteins detected by immunoblotting. e Localisation of N-mCherry-Pol λ in U2OS WT, PAXX-, XLF- and XRCC4-deficient cells. Representative immunofluorescence images showing that N-terminal mCherry-Pol λ fusion protein localises to nuclei in PAXX-, XLF- or XRCC4-deficient U2OS cells. Cells were co-stained with DAPI or dynamin-2, a perinuclear-enriched protein. f Time course of N-EGFP-Pol λ recruitment to laser-induced DNA damage sites in U2OS-WT cells and cells deficient in PAXX, XLF or XRCC4. Data shown are the mean and SEM from WT, PAXX, XLF and XRCC4 knockout cells. Graphs shown are for the following cell numbers: WT: 8 cells; PAXX KO, 10 cells; XLF KO 17 cells; XRCC4 KO 18 cells. g Time course of N-EGFP-FLAG-Ku70 recruitment to laser-induced DNA damage sites in U2OS-WT cells and PAXX KO cells. Data shown are the mean and SEM from WT and PAXX knockout cells. Graphs shown are for the following cell numbers: WT: 17 cells; PAXX KO, 19 cells

Fig. 6

XRCC4 family proteins stimulate gap filling synthesis activity of Pol λ. a Gap filling activity of Pol λ-WT and a catalytically inactive Pol λ-D427A/D429A/D490A mutant. b PAXX, XLF and XRCC4 stimulate gap-filling synthesis activity of Pol λ with an IRDye^® 700-labelled 5nt-gapped dsDNA (33-mer) substrate. c As described in Panel B, except that some reactions also contained either FLAG-PAXX or –XLF alone d Gap filling synthesis assays were performed as described in Panel B with Pol λ immunoprecipitated from RPE-1 PAXX^+/+ or PAXX KO cells incubated with or without XLF or XRCC4 siRNA

Fig. 7

Head domains of XRCC4 Family Proteins interact with and stimulate Pol λ-dependent gap filling synthesis. a Schematic representation of XRCC4 family proteins. b PAXX head domain, but not the CC-CTR region, stimulates gap filling synthesis activity of Pol λ. c Silver stain and immunoblot analysis of purified XRCC4 family protein head domains. d Head domain of XRCC4 family proteins stimulate Pol λ-dependent gap filling synthesis activity. e Far-Western blot analysis showing that Pol λ-WT interacts with the head domain of XRCC4 family proteins. f Stimulation of Pol λ-dependent gap filling synthesis activity by PAXX-WT but not PAXX-VF, a non-Ku binding C-terminal mutant. g PAXX head domain stimulates comparable gap filling synthesis activity of Pol λ-WT and –R57A/L60A

Fig. 8

The Pol λ 8 kDa domain is required for stimulation of Pol λ-dependent gap filling activity via interaction with the head domain of XRCC4 family proteins. a Schematic representation of N-terminal Pol λ deletion mutants. b PAXX, XLF and XRCC4 stimulate gap filling synthesis activity of ΔBRCT- and ΔBRCT-Ser-Pro-Pol λ but not ΔBRCT-Ser-Pro-8kDa-Pol λ with an IRDye^® 700-labelled 5nt-gapped dsDNA (33-mer) substrate. c PAXX head domain promotes gap filling synthesis activity of ΔBRCT- and ΔBRCT-Ser-Pro-Pol λ but not ΔBRCT-Ser-Pro-8kDa-Pol λ. d Far-Western blot analysis showing that the head domain of XRCC4 family proteins interacts with Pol λ-WT and -ΔBRCT but not -ΔBRCT-Ser-Pro-8kDa

Fig. 9

PAXX and XLF together with Pol λ to promote ligation of noncohesive DNA ends which requires gap filling activity of Pol λ. a–e Linear DNA substrates as shown were incubated with the indicated combinations of XRCC4/Lig IV, Ku70/80, PAXX, XLF and Pol λ and the joining efficiency quantified by qPCR with a TaqMan probe using a standard curve of log₁₀ % joining efficiency versus C_t value generated using prejoined DNA fragments. DNA substrates were as follows: (a) EcoRV-PvuI blunt-2nt 3’ overhang; (b) EcoRV-KpnI blunt-4nt 3′ overhang; (c) EcoRV-EcoRV blunt-blunt ends; (d) EcoRI-KpnI 4nt 5’ overhang-4nt 3′ overhang, (e) EcoRV-BstEII blunt-5nt 5′ overhang; (f) As described in Panel (e), except that ligation assays contained either Pol λ-WT or a catalytically inactive Pol λ-3D mutant. Results shown are the mean ± SEM from 2–3 experiments performed in triplicate

Fig. 10

Pol λ, PAXX and XLF function in common and parallel pathways. a Immunoblot analysis of control- or Pol λ siRNA-depleted U2OS PAXX KO, XLF KO and PAXX/XLF DKO cells. WCL were resolved by SDS-PAGE and indicated proteins detected by immunoblotting. b Clonogenic survival assays following IR (0-4 Gy) for U2OS WT, -PAXX KO, -XLF KO and -PAXX/XLF DKO cells with or without depletion of Pol λ Mean and SD from three independent experiments are shown. Statistical analysis was performed using a two-tailed paired t-test to compare cells incubated with Pol λ siRNA with control siRNA: 1 Gy - WT p = 0.91, PAXX KO p = 0.38, XLF KO p = 0.10, PAXX/XLF DKO p = 0.10; 2 Gy - WT p = 0.0003, PAXX KO p = 0.0007, XLF KO p = 0.36, PAXX/XLF DKO p = 0.82; 4 Gy - WT = 0.02, PAXX KO = 0.002, XLF KO p = not determined, PAXX/XLF DKO p = not determined. c Cartoon showing a model for regulation of Pol λ by XRCC4 family proteins. At DSBs that are positioned proximal to a Pol λ substrate gap XRCC4 family proteins strongly interact with Ku heterodimers via their C-terminal regions; their head domains promote gap filling synthesis activity via comparatively weakly binding to the 8 kDa domain of Pol λ, which interacts with the 5′ end of the gap. Pol λ strongly interacts with Ku heterodimers via its N-terminal BRCT domain