Mislocalization of XPF-ERCC1 nuclease contributes to reduced DNA repair in XP-F patients

Abstract

Xeroderma pigmentosum (XP) is caused by defects in the nucleotide excision repair (NER) pathway. NER removes helix-distorting DNA lesions, such as UV-induced photodimers, from the genome. Patients suffering from XP exhibit exquisite sun sensitivity, high incidence of skin cancer, and in some cases neurodegeneration. The severity of XP varies tremendously depending upon which NER gene is mutated and how severely the mutation affects DNA repair capacity. XPF-ERCC1 is a structure-specific endonuclease essential for incising the damaged strand of DNA in NER. Missense mutations in XPF can result not only in XP, but also XPF-ERCC1 (XFE) progeroid syndrome, a disease of accelerated aging. In an attempt to determine how mutations in XPF can lead to such diverse symptoms, the effects of a progeria-causing mutation (XPF(R153P)) were compared to an XP-causing mutation (XPF(R799W)) in vitro and in vivo. Recombinant XPF harboring either mutation was purified in a complex with ERCC1 and tested for its ability to incise a stem-loop structure in vitro. Both mutant complexes nicked the substrate indicating that neither mutation obviates catalytic activity of the nuclease. Surprisingly, differential immunostaining and fractionation of cells from an XFE progeroid patient revealed that XPF-ERCC1 is abundant in the cytoplasm. This was confirmed by fluorescent detection of XPF(R153P)-YFP expressed in Xpf mutant cells. In addition, microinjection of XPF(R153P)-ERCC1 into the nucleus of XPF-deficient human cells restored nucleotide excision repair of UV-induced DNA damage. Intriguingly, in all XPF mutant cell lines examined, XPF-ERCC1 was detected in the cytoplasm of a fraction of cells. This demonstrates that at least part of the DNA repair defect and symptoms associated with mutations in XPF are due to mislocalization of XPF-ERCC1 into the cytoplasm of cells, likely due to protein misfolding. Analysis of these patient cells therefore reveals a novel mechanism to potentially regulate a cell's capacity for DNA repair: by manipulating nuclear localization of XPF-ERCC1.

Figure 1. Biochemical characterization of XPF^R153P-ERCC1 and XPF^R799W-ERCC1 mutants.

(A) Gel filtration profiles from the purification of recombinant XPF-ERCC1, XPF ^R153P-ERCC1 and XPF ^R799W-ERCC1 from baculovirus-infected Sf9 insect cells using a His₆ tag on ERCC1. Aggregated proteins elute at ∼45 ml in the void volume of the column; heterodimeric XPF-ERCC1 elutes at ∼65 ml, corresponding to ∼200 kD, and monomeric ERCC1 elutes at ∼78 ml (∼50kD). (B) SDS PAGE analysis of purified protein complexes. Lane 1, 3 and 5 (D): XPF-ERCC1, XPF ^R153P-ERCC1 and XPF ^R799W-ERCC1, respectively, after purification over NTA-agarose, gel filtration and heparin columns. Lanes 2 and 4 (A) show the proteins present in the fractions eluting at 45 ml in the gel filtration column step of XPF ^R153P-ERCC1 and XPF ^R799W-ERCC1, respectively. (C) Immunodetection of XPF in normal (C5RO) and XPF mutant cells. The star indicates the migration of a cross-reactive band demonstrating equal loading . (D) Incision activities of XPF-ERCC1, XPF ^R153P-ERCC1 and XPF ^R799W-ERCC1 (200 fmol) on a 5′-³²P-labeled stem-loop DNA substrate (100 fmol) in the presence of either 0.4 mM MnCl₂ (lanes 2, 4 and 6) or 2 mM MgCl₂ (lanes 3, 5 and 7). Reactions were analyzed on a 15% denaturing polyacrylamide gel. The 46-mer substrate and 9–10-mer products are indicated.

Figure 2. Differential immunofluorescence of cells from patients with XPF mutations.

Fibroblasts from patients with mutations in XPF and a normal control were grown in the presence of different size beads. After 24 hr the cultures were washed to remove extracellular beads, mixed and co-plated on glass coverslips. The next day, the cells were fixed and immunostained as indicated. Cells were stained with Dapi to identify nuclei and examined by phase contrast microscopy to identify the cell type by their bead content and by fluorescence microscopy for immunodetection of XPF or ERCC1. (A) Analysis of XPF protein sub-cellular localization. Cells from an unaffected individual were labeled with 2 µM beads; XPF mutant cells were labeled with 0.8 µM beads. (B) Analysis of ERCC1 subcellular localization in patients with mutations in XPF. (C) Immunodetection of XPF and ERCC1 in nuclear and cytoplasmic fractions of normal fibroblasts (C5RO) and XPF mutant cells (XP51RO). Tubulin is used as a loading control of the cytoplasmic fraction. Nucleophosmin is used as a loading control for the nuclear fraction. (D) Quantitation of the fraction of cells containing exclusively nuclear XPF-ERCC1, XPF-ERCC1 in the nucleus and cytoplasm (pancellular) or exclusively cytoplasmic complex, as determined from immunofluorescence images (n≥100 cells per cell line).

Figure 3. Characterization of XPF-YFP and XPF¹⁵³-YFP in CHO cells.

(A) Western blot analysis of XPF-YFP expressed in Xpf mutant cells. XPF-deficient hamster cell line, UV41, was transiently transfected with wild type XPF-YFP or XPF¹⁵³-YFP and the fusion proteins were detected using an antibody against XPF or GFP. C5RO was used as positive control for the XPF blot and as a negative control for the GFP blot. UV41 cells transfected with YFP alone was used as a negative control for XPF blot and as a positive control for GFP blot. (B) Clonogenic survival of wild-type (wt), XPF-deficient CHO cell line UV41, and UV41 transfected with wild type XPF-YFP and XPF¹⁵³-YFP after UV and MMC treatment. Colonies were counted 7–10 days after treatment and results are plotted as mean 3 independent experiments. (C) Subcellular localization of wild type XPF-YFP and XPF¹⁵³-YFP after transient transfection in XPF-deficient the CHO cell line UV41 detected by fluorescence microscopy.

Figure 4. Correction of XPF mutant cell NER defect by microinjection of XPF-ERCC1.

Primary fibroblasts from XFE progeroid patient XP51RO were fused to create homopolykaryons by treatment with inactivated Sendai virus then plated on glass coverslips. Only homopolykaryons were injected with recombinant XPF-ERCC1 protein complex (A) wild-type (B) XPF^R799W-ERCC1 (C) XPF^R153P-ERCC1. The cultures were irradiated with 10 J/m² UV-C and ³H-thymidine was added to the culture. UV-induced unscheduled DNA synthesis was detected by autoradiography. Homopolykaryons are indicated with arrows. (D) Histogram indicating the average number of radiographic grains in nuclei injected with each of the recombinant protein complexes and uninjected cells in the same sample. Error bars indicate the standard deviation. N indicates the number of nuclei analyzed in each population. P values for the comparison between injected and uninjected cells were calculated using an unpaired two-tailed Student's t-test.