Differential processing of UV mimetic and interstrand crosslink damage by XPF cell extracts

Abstract

We have recently developed a mammalian cell free assay in which interstrand crosslinks induce DNA synthesis in both damaged and undamaged plasmids co-incubated in the same extract. We have also shown using hamster mutants that both ERCC1 and XPF are required for the observed incorporation. Here, we show that extracts from an XPF patient cell line differentially process UV mimetic damage and interstrand crosslinks in vitro. XPF extracts are highly defective in the stimulation of repair synthesis by N-acetoxy-N- acetylaminofluorene, but are proficient in the stimulation of DNA synthesis by psoralen interstrand crosslinks. In addition, we show that extracts from the hamster UV140 mutant, which has high UV sensitivity, but moderate mitomycin C sensitivity, are similar in both assays to XPF cell extracts. These findings support the hypothesis that the activities of XPF in nucleotide excision repair (NER) and crosslink repair are separable, and that mutations in XPF patients result in the abolition of NER, but not recombinational repair pathways, which are likely to be essential as has been observed in ERCC1 homozygous –/– mice.

INTRODUCTION

Xeroderma pigmentosum (XP) consists of eight complementation groups (A–G), seven of which are deficient in nucleotide excision repair (NER), and an eighth group (variant) which is deficient in the translesion polymerase η (1–3). All eight of these genes have been cloned, and various functions have been ascribed to the encoded proteins. XPF encodes a 115 kDa protein that in a complex with ERCC1 forms a junction-specific endonuclease which incises the 5′-side of the lesion during NER (4,5). XPF patients exhibit mild clinical symptoms in that sun sensitivity is only slightly elevated, the average age of onset of skin cancers is later compared to other complementation groups, and most XPF patients do not present with neurological abnormalities. Also, XPF-derived cell lines show a lesser degree of sensitivity to UV irradiation, and somewhat higher levels of unscheduled DNA synthesis than cells from most other complementation groups (6).

In addition to their hypersensitivity to UV irradiation, cells deficient in either ERCC1 or XPF also show a highly increased sensitivity to agents that produce interstrand crosslinks in DNA, which is in contrast to other mammalian complementation groups involved in NER (7,8). These findings indicate that ERCC1 and XPF are involved in recombinational repair pathways in addition to their role in NER, a conjecture that is supported by results in Saccharomyces cerevisiae which demonstrate that the yeast homologs of these genes, RAD1 and RAD10, are involved in removing regions of non-homology during either the single-strand annealing or homologous recombination pathways of double-strand break repair (reviewed in 9). RAD1 and RAD10 have also been implicated in some pathways of mismatch repair in yeast (10). Finally, the Drosophila homolog of XPF, mei-9, has been shown to be involved in both meiotic and mitotic recombination pathways (11).

Homozygous disruption of the ERCC1 gene in mice leads to severe runting and death before weaning due to an unusual liver pathology exhibiting enlarged nuclei with both polyploidy and aneuploidy (12,13). In addition, these mice exhibited an absence of subcutaneous fat, early onset of ferritin deposition in the spleen, kidney malfunction, cytoplasmic invaginations in nuclei of liver and kidney, and defects in NER and interstrand crosslink repair. Weeda et al. (13) have proposed that the phenotype of the ERCC1 –/– mice can be attributed to a deficiency in the repair of interstrand crosslinks by means of mitotic recombinational pathways since XPA –/– or XPC –/– mice (defective only in NER) are not similarly affected (14,15). In addition, hamster cell mutants, deficient in ERCC1, exhibit a recombination-dependent phenotype that results in an ~50-fold increase in deletions of a tandemly duplicated APRT gene, suggesting that ERCC1 is involved in processing of heteroduplex intermediates during homologous recombination (16).

Recently, we developed an in vitro assay (referred to as CRS) in which DNA synthesis is stimulated by the presence of a single defined interstrand crosslink in plasmid DNAs (17). Interestingly, this stimulation of DNA synthesis is also observed in undamaged plasmids co-incubated in the same extract suggesting that the crosslink induces a recombinational repair processing pathway. Hamster mutants deficient in either ERCC1 or XPF were found to be highly deficient in the CRS assay. Because of the observed lethality associated with the absence of ERCC1 in mice, we were prompted to examine whether or not extracts from XPF patient cell extracts would also be deficient in recombinational repair processing of crosslinks. Our findings show that XPF patient cell extracts, while defective in NER, are proficient in the CRS assay, suggesting that mutations in XPF patients affect only NER and that essential recombination pathways involving XPF are intact.

MATERIALS AND METHODS

Cell lines and biochemicals

The SV40 transformed XPF cell line XP2YO was derived by Yagi and Takebe (18), and was obtained from Errol Friedberg. The hamster cell lines UV20, UV140, AA8 and CHO9 were obtained from Rodney Nairn, and UVS1 was derived by Hata et al. (19), and provided by David Busch. Cells were grown in minimal essential medium plus 10% calf serum. Purified ERCC1/XPF protein complex was generously supplied by Aziz Sancar.

Substrates and extracts

Plasmids and psoralen interstrand crosslinked substrates for the CRS assay were prepared as previously described (17). Briefly, 100 µg of a 21 bp duplex oligonucleotide was added to 4,5′,8-trimethylpsoralen at 5 µg/ml in 10 mM Tris pH 7.5, 0.5 mM EDTA, 25 mM NaCl, and the sample was irradiated with 365 nm UV light (10 min at 9 mW/cm²) to effect formation of interstrand crosslinks. The crosslinked oligonucleotide was purified from uncrosslinked DNA by denaturing polyacrylamide gel electrophoresis, and subsequently ligated to linearized plasmid DNA. After ligation, covalently closed plasmid DNA termed CLT, and containing a single defined crosslink, was purified by CsCl/ethidium bromide gradient centrifugation. The DT and CT plasmids were prepared by alkaline lysis of host cells and purified by CsCl/ethidium bromide gradient centrifugation.

Treatment of plasmid substrates with N-acetoxy-N-2-acetylaminofluorene (AAAF) for the NER in vitro assay was performed as previously described (20). pGEX plasmid used as the control for the NER assay was prepared by standard procedures.

Mammalian whole cell extracts for use in both the NER and CRS assays were prepared by the method of Manley et al. (21).

In vitro repair assays

The NER (22) and CRS (17) in vitro assays were performed as previously described. Briefly, 50 µl reactions contained (final concentration) 100 of µg extract, 30 ng of CLT or CT, 100 ng of DT, 45 mM HEPES–KOH pH 7.8, 75 mM KCl, 7.4 mM MgCl₂, 0.9 mM dithiothreitol, 0.4 mM EDTA, 2 mM ATP, 20 µM each of dATP, dGTP and TTP, and 8 µM dCTP, 2 µCi [α-³²P]dCTP (3000 Ci/mmol), 40 mM phosphocreatine, 2.5 µg creatine phosphokinase, ~3% glycerol and 18 µg bovine serum albumin. The reactions were incubated for 3 h at 22°C. The NER assays were performed by essentially the same procedure except that 150 µg extract, and 150 ng of AAAF-damaged and undamaged plasmid DNAs each, were added, and the reaction was incubated at 30°C. After incubation, the reactions were stopped by addition of EDTA and the samples were processed to remove RNA and protein. Plasmids were sequentially digested with the appropriate restriction enzymes and analyzed by agarose gel electrophoresis.

Colony survival assay

Colony survival of mammalian cells after exposure to mitomycin C were performed as previously described (23).

RESULTS

XPF cell extracts are proficient in the CRS assay

We prepared whole cell extracts from the SV40 transformed XPF cell line XP2YO and examined these extracts in the NER in vitro assay (22). As is typical of XPF cells, these extracts are deficient in the NER assay, but activity could be restored by the addition of purified ERCC1/XPF (Fig. 1a). Surprisingly, however, these extracts were as proficient in the CRS assay as were extracts prepared from HeLa cells (Fig. 1b), indicating that XPF cell extracts are competent in the initial stages of processing of interstrand crosslinks.

Figure 1.

Activity of XP2YO and HeLa cell extracts in the (a) NER and (b) CRS in vitro assays. Assays were performed as described in Materials and Methods. In each part the upper panel represents the autoradiogram, and the lower panel represents the ethidium bromide-stained gel.

XPF is required for repair processing of interstrand crosslinks

UV41 and UVS1 are hamster mutants that are defective in XPF (4). These cell lines are hypersensitive to UV irradiation and extracts from these cells are deficient in the in vitro NER assay. As shown in Figure 2a, extracts prepared from UV41 cells were found to be highly deficient in the CRS assay in comparison to the parental cell line AA8. In addition, complementation of these extracts was achieved by the addition of purified ERCC1/XPF protein complex (Fig. 2b). We also examined extracts prepared from UVS1 cells, and as shown (Fig. 3a and b) these extracts exhibited greatly reduced activity in both the NER and CRS assays in comparison to the parental cell line CHO9. As was found for the UV41 cells, activity of the UVS1 extracts in the CRS assay was restored by the addition of purified ERCC1/XPF.

Figure 2.

Activity of UV41 and AA8 cell extracts in the CRS in vitro assay (a) without or (b) with the addition of ERCC1/XPF purified protein.

Figure 3.

Activity of UVS1 and CHO9 cell extracts in the (a) NER, (b) CRS or (c) CRS assays with the addition of purified ERCC1/XPF protein.

The finding that UVS1 extracts were highly deficient in the CRS assay was somewhat surprising since as shown in Figure 4, these cells exhibited only a slight sensitivity to the crosslinking agent mitomycin C (MMC), whereas XPF cells showed an intermediate level of survival to this agent. Also, as shown, UV41 cells exhibited an extreme degree of sensitivity to MMC.

Figure 4.

Survival of various mammalian mutants and wild-type cell lines after treatment with MMC.

Since the findings with UVS1 were unexpected, we assayed extracts from a third XPF hamster mutant cell line, UV140. This mutant, like UVS1, has been shown in colony survival assays to be highly sensitive to UV, but only mildly sensitive to MMC (24). It also exhibits extremely low levels of unscheduled DNA synthesis after UV exposure. We assayed extracts prepared from UV140 cells in both the NER and CRS in vitro assays (Fig. 5). Consistent with previous results, UV140 extracts were highly deficient in NER (24), but were complemented by purified ERCC1/XPF. More importantly, similar to the findings with XPF extracts described above, UV140 extracts exhibited near wild-type levels of activity in the CRS assay.

Figure 5.

Activity of UV140 and AA8 cell extracts in the (a) NER assay with and without the addition of purified ERCC1/XPF, and (b) CRS assay.

DISCUSSION

We have shown that extracts from XPF cells, which are highly deficient in response to UV-mimetic damage, nevertheless, exhibited wild-type levels of activity in response to interstrand crosslinks as determined by the CRS assay. The hamster mutant cell lines UV41 and UVS1 were found to be deficient in the CRS assay indicating that XPF is an essential element of crosslink repair. The finding that UVS1 cells were deficient in the CRS assay was surprising in light of the moderate level of sensitivity to MMC observed in vivo (Fig. 4). One possibility is that the particular XPF mutation in UVS1 cells renders the protein highly labile to extraction procedures. A third hamster XPF mutant, UV140, which exhibits a phenotype similar to UVS1 in survival assays after exposure to UV or MMC, was, however, similar to extracts from XPF patient cells in the NER and CRS in vitro assays. These results thus agree with previous conclusions that the activities of XPF in NER and crosslink repair are separable (24). In addition, we have shown previously that cell extracts from XP complementation groups A, C and G exhibited normal levels of activity in the CRS assay (17), which is consistent with the very moderate degree of hypersensitivity to MMC found in these complementation groups (7,8,25). Thus, our in vitro results are consistent with the in vivo findings indicating that XPF and ERCC1 play an integral role in the recombinational repair of crosslinks that is not observed with other members of the NER complementation groups.

Recently, Matsumura et al. (26) characterized the mutations in seven XPF Japanese patients. Interestingly, in six of the seven patients at least one allele contained either a point mutation or a small in-frame deletion. An additional patient described by Sijbers et al. (4) also contained a point mutation in one allele indicating that full-length or nearly full-length XPF is present in seven of eight XPF patient cell lines examined. In fact, Yagi et al. (27) found XPF protein, albeit at reduced levels, in five of these cell lines that were examined by western analysis. In the eighth patient cell line, XP23OS, a single mutation was identified that is predicted to result in truncation of the protein at residue 482 which would presumably abrogate both NER and recombination pathways. However, levels of XPF transcript in XP23OS were found to be comparable to that observed in normal cells (26). Premature termination codons often result in active degradation of mRNA by a surveillance mechanism (28), thus, XP23OS may possess a second unidentified allele that is active in recombinational repair, but possessing a mutation that inactivates NER. Another interesting aspect to these identified mutations is that they tend to be clustered in the middle segment of the XPF protein (Fig. 6). In the six Japanese patients examined (excluding XP23OS) at least one allele was found with a point mutation or in-frame deletion between amino acid residues 443 and 600. The caucasian patient XP126LO contains a point mutation at residue 788. None of these mutations occur in the ERCC1 binding domain located at residues 814–905 (29), or in either of the two leucine zipper motifs that occur in the region from residues 222 to 287 (26). Thus, conceivably, these identified mutations in the middle third of the protein may affect the function of XPF in NER, but not in recombinational repair of interstrand crosslinks. An alternative possibility is that the low levels of ERCC1/XPF protein found in these mutants affect NER more profoundly than the pathway of interstrand crosslink repair.

Figure 6.

Clustering of mutations in XPF as observed in patient cell lines.

Recently, Bessho et al. (30) have demonstrated that purified NER components recognize and incise DNA at dual sites 5′ to interstrand crosslinks. It was postulated in that report that this mechanism of incision could lead in vivo to repair of these lesions. However, in a subsequent publication this same group reported that NER-mediated processing of crosslinks in mammalian cell extracts results in a futile cycle of excision and gap filling that apparently does not lead to repair of the damage (31). Our results described here, however, clearly demonstrate that a pathway of crosslink-induced DNA synthesis separate from NER exists in mammalian cell extracts. XPF cell extracts from XP2YO cells while deficient in NER are, nevertheless, highly proficient in the CRS assay indicating that this assay represents a distinct in vitro response to interstrand crosslinks. In addition, unlike NER, interstrand crosslinks, as observed in the CRS assay, induce DNA synthesis in undamaged plasmids co-incubated in the extract possibly suggesting a recombinational mode of damage processing. Taken together these results suggest that the CRS assay represents the initial stages of a recombinational pathway that represents the major mode of crosslink repair in mammalian cells.