XPD localizes in mitochondria and protects the mitochondrial genome from oxidative DNA damage

Abstract

Xeroderma pigmentosum group D (XPD/ERCC2) encodes an ATP-dependent helicase that plays essential roles in both transcription and nucleotide excision repair of nuclear DNA, however, whether or not XPD exerts similar functions in mitochondria remains elusive. In this study, we provide the first evidence that XPD is localized in the inner membrane of mitochondria, and cells under oxidative stress showed an enhanced recruitment of XPD into mitochondrial compartment. Furthermore, mitochondrial reactive oxygen species production and levels of oxidative stress-induced mitochondrial DNA (mtDNA) common deletion were significantly elevated, whereas capacity for oxidative damage repair of mtDNA was markedly reduced in both XPD-suppressed human osteosarcoma (U2OS) cells and XPD-deficient human fibroblasts. Immunoprecipitation-mass spectrometry analysis was used to identify interacting factor(s) with XPD and TUFM, a mitochondrial Tu translation elongation factor was detected to be physically interacted with XPD. Similar to the findings in XPD-deficient cells, mitochondrial common deletion and oxidative damage repair capacity in U2OS cells were found to be significantly altered after TUFM knock-down. Our findings clearly demonstrate that XPD plays crucial role(s) in protecting mitochondrial genome stability by facilitating an efficient repair of oxidative DNA damage in mitochondria.

Figure 1.

Co-localization of XPD protein with mitochondria in U2OS cells by an indirect immunofluorescence approach. The cells were transfected with pcDNA3.0-Flag-XPD expression plasmid, and stained by 5 μM of Mitotracker Red CM-H2Xros for 30 min at 24 h post transfection. After fixation in 4% paraformaldehyde for 10 min and 0.1% Triton X-100 for 5 min, the cells were incubated with the first Flag antibody (Sigma) overnight at 4°C followed by fluorescein goat anti-mouse IgG secondary antibody (Vector laboratories) at RT for 2 h. The staining result was visualized under the Leica Confocal microscope.

Figure 2.

Localization of XPD in mitochondria examined by western blotting coupled with protease treatment or alkali extraction on isolated mitochondrial fraction. (A) Endogenous level of XPD protein in both nucleic and mitochondrial fractions of U2OS and HEK293 cells determined by western blotting. Presence of XPD protein in mitochondrial fraction of both cell types was observed. XPD antibody was purchased from Cell Signaling. Anti-VDAC (mitochondrial marker), Lamin B (nucleic marker) and GAPDH (cytoplasmic marker) were ordered from Cell Signaling, Santa Cruz, Millipore, respectively. (B) Protease treatment was performed on intact mitochondria fraction isolated from HEK293 cells by digestion with 100 ng/ml proteinase K (PK) in the presence or absence of 1% TritonX-100 on ice followed by western blotting analysis. VDAC serves as the marker of mitochondrial outer membrane protein, and SMAC as the marker of mitochondrial intermembrane space protein. Anti-SMAC antibody was purchased from Cell Signaling. Alkali extraction was carried out by using Na₂CO₃ to treat the mitochondrial fraction isolated from HEK293 cells. Both the soluble protein fraction (S) and integral membrane protein fraction (P) were isolated and used for western blotting analysis. ‘M’ means untreated control mitochondrial sample. Results demonstrate the localization of XPD in the mitochondrial inner matrix compartment and inner mitochondrial membrane.

Figure 3.

Mitochondrial ROS production in XPD silenced U2OS and XPD patient skin fibroblasts. (A) XPD levels in whole cell extracts (WCE) and mitochondrial fraction of U2OS determined by western blotting. A markedly decreased level of XPD was observed in both samples. Mitochondrial ROS level in shControl and shXPD U2OS cells was detected by MitoSOX^TM Red. *P < 0.05. (B) Mitochondrial ROS level in normal human dermal fibroblasts (NHDF) and XPD-deficient patient skin fibroblasts. Cells were stained with MitoSOX^TM Red for 30 min and then analyzed by flow cytometry. Relative ROS production was calculated by PE-A means from three independent experiments. Error bars represent standard deviations. **P < 0.01.

Figure 4.

Level of mtDNA CD post oxidative stress in XPD-wild type and deficient cells. (A) An increased mitochondrial distribution of XPD protein in HEK293 cells after H₂O₂ treatment. Cells were treated with 0.5 mM H₂O₂ for 1 h, and samples were collected at 45 min post treatment. Fifty micrograms of protein for each of nuclear, cytosolic and mitochondrial fractions were loaded and separated on SDS-polyacrylamide gel for western blotting analysis. Mitochondrial proportion of XPD protein was quantified using the Image J software (http://rsbweb.nih.gov/ij/) and normalized to mitochondrial loading control (VDAC). (B) The relative quantities of CD were measured by real-time PCR in XPD knock-down U2OS cells, normal NHDF versus XPD patient cells at 48 h post 0.5 mM of H₂O₂ treatment in the presence or absence of 5000 U/ml of Catalase. Levels of CD were assessed by Bio-Rad CFX Manager 2.1. Samples were collected at 0 h (Con) and 48 h recovery with or without catalase (H₂O₂ or H₂O₂+catalase) post treatment. Data were from three independent experiments. Error bars represent standard deviations. *P < 0.05.

Figure 5.

Long-range QPCR results of full-length mitochondrial DNA (mt-FL) in shControl and shXPD U2OS cells (A) and in NHDF and XPD patient fibroblasts (B) at recovery time-points of 0, 6, 12, 24 and 48 h post 1 mM H₂O₂ treatment for 1 h. Equal amount of mtDNA template in each sample normalized by mtND1 gene was used for long-range QPCR reactions. Nucleic β-globin gene was used as the repair efficiency marker for oxidative nuclear DNA damage.

Figure 6.

XPD helicase activity is critical in XPD-regulated mtDNA repair. (A) Reconstitution of HA-WT XPD and HA-XPD/K48R mutant in XPD-silenced U2OS cells examined by western blotting. (B) Level of CD in XPD-WT or mutant reconstituted XPD-silenced U2OS cells quantified by real-time PCR at 0 h (Con), 48 h (H₂O₂ or H₂O₂+catalase) post H₂O₂ treatment. Cells were treated with 0.5 mM H₂O₂ for 1 h. Data from three independent experiments. (C) Long-range QPCR results of full-length mitochondrial DNA (mt-FL) in XPD-WT versus helicase mutant-reconstituted XPD-silenced U2OS cells at recovery time-points of 0, 6, 12, 24 and 48 h post H₂O₂ treatment.

Figure 7.

XPD associates with mitochondrial TUFM. (A) Silver-staining of XPD-associated proteins. Mitochondrial fraction isolated from HEK293 cells stably expressing Flag-XPD were immunoprecipitated with antibody against Flag M2 beads and analyzed by a Mass-Spectrometer (lane 1). Lane 2 is mock control without transfection with Flag-XPD expression vector. TUFM (arrow) was identified to be one of the partners of XPD. (B) Endogenous XPD was immunoprecipitated with Flag-TUFM recognized by an anti-Flag antibody from cell extracts of 10⁶ HEK293 cells. The immunoprecipitated proteins were detected with antibody against XPD (Cell Signaling). Conversely, endogenous TUFM was immunoprecipitated with Flag-XPD recognized by an anti-Flag antibody from mitochondrial extracts of 10⁶ HEK293 cells. The immunoprecipitated proteins were visualized by western blotting analysis with antibody against TUFM (Sigma). Five percent of the lysate was used for the loading control (Input) and the remaining 95% for co-IP. (C) Endogenous TUFM was immunoprecipitated with anti-XPD antibody (Cell Signaling) from cell extracts of 10⁶ HEK293 cells. TUFM in the pull-down complex was visualized by western blotting analysis using antibody against TUFM (Sigma). (D) In the upper panel, schematic diagram of XPD deletion constructs used for co-IP studies is shown. In the lower panel, HEK293 cells were transfected with either Flag-GFP or one of the Flag-tagged truncated XPD expressing plasmids. TUFM was strongly immunoprecipitated with XPD N-terminal domain, while weakly with XPD C-terminal (CT) and ARCH domains.

Figure 8.

CD and oxidative damage repair in mitochondria after TUFM knock-down. (A) U2OS cells were transfected with control siRNA (siControl) and TUFM siRNA (siTUFM), respectively. Levels of TUFM in whole cell lysate and mitochondrial fraction were determined by western blotting. (B) CD of mtDNA was quantified by real-time PCR in siControl and siTUFM transfected U2OS cells. Cells were treated with H₂O₂ for 1 h in the presence or absence of catalase, and collected at 0 h and 48 h post treatment. Three independent experiments were carried out. Error bars represent standard deviations. **P < 0.01. (C) Result of long-range QPCR result of full-length mitochondrial DNA in siControl and siTUFM U2OS cells treated with 1 mM H₂O₂ for 1 h. Equal loading of DNA was verified by monitoring the level of mt-ND1 gene.