To extend our knowledge of the excision repair system in mammalian cells we have focussed on the isolation of genes and proteins involved in this process. For the purification and characterization of human repair proteins the microneedle injection assay technique is utilized. This system is based on the transient correction of the excision repair defect of xeroderma pigmentosum (XP) fibroblasts (scored as increase of ultraviolet (u.v.)-induced unscheduled DNA synthesis (UDS)) upon microinjection of crude extracts from complementing XP or normal cells. Specific correction is observed in fibroblasts of all (9) excision-deficient XP complementation groups. The XP-A and G correcting factors were found to be proteins and several purification steps (including (NH4)2SO4 fractionation, chromatography of phosphocellulose, heparin and u.v.- irradiated DNA-cellulose) have been worked out for the XP-A correcting protein.

The microinjection system was also used for the introduction of (partially) purified repair enzymes of lower organisms. Micrococcus luteus endonuclease and bacteriophage T4 endonuclease V were able to correct all XP complementation groups tested, in marked contrast to the more sophisticated Escherichia coli uvrABC complex injected with uvrD. Photoreversal of dimers could be registered after introduction of the yeast photoreactivating enzyme in repair-competent, XP- variant, XP-C and XP-I fibroblasts (monitored as decrease of (residual) UDS). Remarkably, no effect was noticed in XP-A, D, E and H, suggesting that something prevents dimers in these cells from being monomerized by the injected enzyme.

Using DNA-mediated gene transfer we have cloned a human gene (designated ERCC-1) that compensates for the excision defect of the u.v. and mitomycin C-sensitive Chinese hamster ovary cell (CHO) mutant 43-3B (complementation group 2). Characterization of this gene and its cDNA revealed the following features:

(1) ERCC-1 corrects the full spectrum of repair deficiencies in mutants of complementation group 2. No correction is observed in mutants of the other CHO complementation groups.

(2) The ERCC-l gene has a size of 15 ×101 base-pairs (bp) and consists of 10 exons, one of which appears to be differentially spliced.

(3) It encodes two largely identical mRNAs, which differ in the presence or absence of a 72bp coding exon, situated in the 3 ′ half of the mRNA. Only the cDNA of the large transcript is able to confer repair proficiency to 43-3B cells. No effect of u.v. treatment is found at the level of ERCC-1 transcription in HeLa cells.

(4) Sequence analysis of full-length cDNA copies of the two ERCC-1 mRNAs revealed open reading frames for proteins of 297 and 273 amino acids, respectively. Significant amino acid sequence homology was found between portions of the putative ERCC-1 product and the protein encoded by the yeast excision-repair gene RAD10. Regional homology was also discovered between a part of ERCC-1 and uvrA. On the basis of homology with functional protein domains a tentative nuclear location signal, DNA binding domain and ADP-ribosylation site could be identified in the ERCC-1 aa sequence.

The implication of DNA repair systems in vital processes such as mutagenesis, carcinogenesis and maintenance of primary DNA structure stresses the importance of investigating how these systems operate at the molecular level. Availability of genes and proteins involved in DNA repair is of basic interest to such studies. The isolation of genes, and subsequent purification of the products, functioning in repair systems of Escherichia coli (particularly in the excision repair pathway and the removal of alkylation lesions) has led to a considerable extension of our knowledge about mechanisms and genetic control of these processes in prokaryotes (Walker, 1985; Teo et al. 1986; for an extensive review on DNA repair in general, see Friedberg, 1985). For lower eukaryotes the phase of cloning repair genes has been reached with the recent isolation of a series of genes involved in three major repair pathways in the yeast Saccharomyces cerevisiae (e.g. see Yasui & Chevallier, 1983; Adzuma et al. 1984; Naumovski et al. 1985; Nicolet et al. 1985; Reynolds et al. 1985a,b’, Perozzi & Prakash, 1986). As far as higher eukaryotes are concerned, valuable tools for the isolation of repair genes and proteins are available in the form of naturally occurring and laboratory-induced mutant cells. These are mainly of human and rodent origin. The most extensively characterized class of mutants is disturbed in the excision repair process, one of the major repair systems in the cell. Fibroblasts from many patients suffering from the autosomal recessive human syndrome xeroderma pigmentosum (XP) fall into this category (Cleaver, 1968). Genetic studies involving cell hybridization have disclosed the existence of nine complementation groups within the excision-deficient class of XP individuals (de Weerd-Kastelein et al. 1972; Fischer et al. 1985) and at least five within excision-deficient mutants generated from Chinese hamster ovary (CHO) cells (Thompson et al. 1981, 1982; Thompson & Carrano, 1983). It is not known whether some of these complementation groups are the same in both species; however, none of the mutants is able to perform efficiently the first step postulated in the excision pathway, i.e. the incision of the damaged DNA strand at or near the photolesion. This suggests the participation of at least nine, and possibly more than 13, genes and proteins in nucleotide excision. To date, progress with regard to the isolation of the components involved is limited, due to the complexity of the system and the experimental limitations of the organisms. Recently, we have cloned the first human gene (designated ERCC-1), implicated in the excision of lesions induced by ultraviolet (u.v.) light and cross-linking agents (Westerveld et al. 1984). Using microneedle injection we are in the process of purifying and characterizing proteins deficient in XP. Furthermore, we are investigating whether well-characterized repair proteins of heterologous organisms can interact with mammalian repair systems. In this chapter we will summarize results obtained on these topics. Part of the work reviewed here has been published in detail elsewhere (de Jonge et al. 1983, 1985; Zwetsloot et al. 1985; Hoeijmakerset al. 1987a,b; Van Duinet al. 1986; Vermeulen et al. 1986; Zwetsloot et al. 1986a,b).

The microinjection system

Notwithstanding many attempts the establishment of a reliable mammalian in vitro repair system has not been a very rewarding enterprise. As an alternative we have developed an in vivo test system suitable for at least some of the factors involved in excision repair. This system utilizes living human (XP) fibroblasts as test-tubes in which repair components are introduced by microneedle injection (de Jonge et al. 1983). u.v.-induced unscheduled DNA synthesis (UDS) is used as a single cell parameter to determine the effect on the excision repair of the injected cells. This test-system has a wide range of possibilities and applications to the study of nucleotide excision.

(1) Crude protein extracts from various cells can be screened for the presence of proteins that transiently compensate for the defect in one of the XP complementation groups (see e.g. Fig. 1A). On the basis of this property these polypeptides can then be isolated and characterized.

Fig. 1.

A. Photomicrograph of XP25RO(XP-A) homopolykaryons (arrowheads) after microinjection of a crude HeLa cell extract, assay for u.v.-induced UDS and autoradiography. The monokaryons (one of which in S-phase) have not been injected. B. Photomicrograph of a C5R0 (repair-proficient) homodikaryon (arrowhead) after injection of purified photoreactivating enzyme from S. cerevisiae. The monokaryons have not been injected. After microinjection the fibroblasts were u.v.-irridiated (10 J m −2), illuminated with photoreactivating light, assayed for UDS, followed by autoradiography. Homopolykaryons are used for microinjection for reasons specified by de Jonge et al. (1983). The micrographs are of different magnification.

Fig. 1.

A. Photomicrograph of XP25RO(XP-A) homopolykaryons (arrowheads) after microinjection of a crude HeLa cell extract, assay for u.v.-induced UDS and autoradiography. The monokaryons (one of which in S-phase) have not been injected. B. Photomicrograph of a C5R0 (repair-proficient) homodikaryon (arrowhead) after injection of purified photoreactivating enzyme from S. cerevisiae. The monokaryons have not been injected. After microinjection the fibroblasts were u.v.-irridiated (10 J m −2), illuminated with photoreactivating light, assayed for UDS, followed by autoradiography. Homopolykaryons are used for microinjection for reasons specified by de Jonge et al. (1983). The micrographs are of different magnification.

(2) As shown by Legerski et al. (1984), microinjection of mRNA can, after in vivo translation into protein, also temporarily alleviate some of the XP deficiencies. In principle this opens the possibility of isolating the responsible gene by cDNA cloning of the correcting RNA.

(3) Purified repair proteins from various prokaryotes and eukaryotes can be tested individually or in specific combinations for their effect on human excision repair or repair defects.

(4) The role of endogenous polypeptides that might be implicated in the excision pathway can be directly assessed in vivo by injection of (monoclonal) antibodies into human fibroblasts (e.g. monoclonal antibodies against DNA polymerase α).

(5) Finally, microinjection might be used to introduce certain repair substrates (in the form of DNA molecules carrying specific lesions) or inhibitors of specific steps in the excision process.

There are also obvious limitations inherent in the microinjection assay technique. Only a small number of samples can be tested in one experiment; it is a laborious, time-consuming procedure and yields only semi-quantitative results. We have concentrated mainly on the first and third application listed above and this section focusses on some of the results obtained thus far.

Injection of repair proteins of prokaryotes and eukaryotes into human fibroblasts

The effect of microinjection of defined prokaryotic proteins involved in excision repair on UDS of wild-type or XP fibroblasts of various complementation groups is summarized in Table 1 (for quantitative data see de Jonge et al. 1985). Micrococcus luteus endonuclease and bacteriophage T4 endonuclease V (which both catalyse the incision of damaged DNA at the site of cyclobutane dimers by a combined glycosylase and apyrimidinic endonuclease activity; Haseltine et al. 1980; Grafstrom et al. 1982) correct the defect of all XP complementation groups tested as judged by a significant increase in u.v.-induced UDS. This aspecific bypass of different excision defects in XP, which was noted earlier for the T4 enzyme by Tanaka et al. (1975, 1977) and in the case of XPA-E and XP-F by Hayakawa et al. (1981), indicates that these prokaryotic enzymes are able to act on pyrimidine lesions in mammalian chromatin and that the cellular repair machinery can recognize the resulting incision products and use them as substrates for repair synthesis.

Table 1.

Effect of various injected repair enzymes on XP UDS

Effect of various injected repair enzymes on XP UDS
Effect of various injected repair enzymes on XP UDS

In contrast to the two u.v. endonucleases examined above, microinjection of the E. coli UvrABC and D proteins into XP-A, C and H fibroblasts did not result in restoration of u.v.-induced UDS (Table 1; Zwetsloot et al. 1986<z). The E. coli excinuclease was assayed under conditions in which a T4 endonuclease preparation with the same in vitro incising activity was able to induce significant repair synthesis in the three XP complementation groups. To exclude the possibility that the negative results with the E. coli proteins could be due to potential problems in transport of this bulky enzyme complex (total mass ∼350 ×101Mr, compared to 16 ×101Mr for T4 endonuclease) through the nuclear membrane both endonucleases were also tested by direct microinjection into the nucleus of XP-A and C fibroblasts. The same results were obtained (see Table 1). Although various explanations of a negative result are possible (as discussed elsewhere; Zwetsloot et al. 1986a) the most likely interpretation is that the A. coli excinuclease does not act on DNA photoproducts in a mammalian chromatin context.

A different manner of removing dimer lesions induced by u.v. is exemplified by enzymic photoreactivation accomplished by a single protein: the photoreactivating enzyme (PRE). This flavoprotein binds specifically to dimers and after absorption of a quantum of the visible to near-u.v. spectrum it simply reverses the dimerization regenerating two unlinked pyrimidine bases. This reaction does not involve repair synthesis and can compete with the excision repair of cyclobutane dimers. To examine whether PRE is active on pyrimidine dimers in chromatin of mammalian cells purified photolyase from yeast was injected into repair-proficient fibroblasts. Since photoreactivation of these lesions should be at the expense of repair synthesis a reduction of the UDS level is expected in the injected cells if photoreactivation occurs. Fig. IB and Table 1 show that indeed a strong decrease in UDS of injected wild-type fibroblasts (to 20 % of non-injected control cells) was found. The fact that this reduction is photoreactivating light-dependent, proteinase JC-sensitive and not observed with extracts prepared from phr yeast mutants (in contrast to extracts from phr+ cells) argues in favour of the idea that it is due to photoreactivation of dimers by the injected enzyme (Zwetsloot et al. 1985). To determine whether lesions in XP cells also can be photoreactivated by the exogenous photolyase, the yeast protein was injected into fibroblasts from excision-deficient XP strains representing different complementation groups (all displaying more than 10% of the UDS of wild-type cells) and fibroblasts from two excision-proficient XP variant strains. In XP variant cells UDS was reduced to the same extent as in normal fibroblasts. Injected cells from XP complementation groups C, I and to a lesser extent XP-F also exhibited a reduced level of UDS (up to 50% of non-injected XP cells) indicating that pyrimidine dimers in these cells are accessible to and can be monomerized by the injected enzyme. In contrast, fibroblasts from groups A, D, E and H (with residual UDS activities up to 50% of wild-type cells) did not show any reduction in UDS upon PRE-injection and illumination with photoreactivating light (Zwetsloot et al. 19866). These findings strongly suggest that the photoproducts are not equally well accessible in both sets of XP complementation groups. In the latter XP dimers might be ‘protected’ against injected PRE by a defective repair complex. At present, we are characterizing CHO transformants containing the yeast and E. coli PRE genes inserted in a mammalian expression vector with the eventual aim of studying the long-term effects of selective dimer removal on survival, mutability and induction of chromosomal aberrations.

Purification and characterization ofXP correcting proteins

As mentioned above, crude extracts of human cells upon injection transiently compensate for the excision defect in XP fibroblasts. Table 2 shows that temporary correction is found for all excision-deficient XP complementation groups as indicated by an increase in UDS. In the complementation groups examined extracts prepared from complementing XP (as well as normal) cells induce correction, in contrast to extracts from non-complementing XP fibroblasts. We conclude, therefore, that in each instance specific factors are responsible for the observed phenotypic correction. The data in Table 2 display a wide variation in the level of correction between different complementation groups. Undoubtedly, this reflects at least in part differences in concentration of the various correcting factors in the injected extract, particularly in those XP groups in which the correcting component is rate-limiting. Another factor that influences the level of UDS is the kinetics of complementation, which is found to vary between different complementation groups. Cell fusion and microinjection experiments have demonstrated that, e.g. the XP-A defect is fully corrected within 30 min after introduction of the correcting factor (e.g. see Vermeulen et al. 1986). In contrast, complete compensation for the XP-C and D deficiencies is not achieved within 16 –24 h after fusion with repair-competent fibroblasts (Giannelli et al. 1982; Keijzer et al. 1982). Characterization of the XP-A and XP-G factors established that both of them are sensitive to proteolytic degradation, suggesting that they are proteins (de Jonge et al. 1983; Vermeulen et al. 1986). Both proteins were found to precipitate at between 25% and 40% (NH4)2SO4 saturation. Testing of subsequent purification steps revealed that the XP-A correcting factor is retained on phosphocellulose and heparin columns and that it has high affinity for single-stranded DNA as well as u.v.-irradiated doublestranded DNA attached to cellulose. A weaker affinity was found for non-irradiated double-stranded DNA (preliminary results). These findings support the idea that the XP-A correcting protein exerts its function in the nucleus presumably by involving binding to the u.v.-induced DNA lesions. At present, we are attempting large-scale purification of this polypeptide by a combination of the steps described above. Successful combinations of (NH4)2SO4 fractionation and phosphocellulose chromatography using a total HeUa cell extract, and phosphocellulose followed by u.v.-irradiated DNA-cellulose chromatography using a placenta extract have been achieved. However, one of the problems (in addition to the laborious assay procedure) is that the protein seems to be more labile upon further purification.

Table 2.

Levels of u.v.-induced UDS in homopolykaryons of various XP complementation groups after microinjection of human cell extracts

Levels of u.v.-induced UDS in homopolykaryons of various XP complementation groups after microinjection of human cell extracts
Levels of u.v.-induced UDS in homopolykaryons of various XP complementation groups after microinjection of human cell extracts

Isolation of human repair genes

Although simian virus 40 (SV40)-transformed XP fibroblasts seem to be the ideal source for the cloning of complementing excision repair genes they have thus far failed to yield genuine repair-proficient transformants upon genomic DNA transfections (Lehmann, 1985). In a number of laboratories, u.v.-resistant repair- competent revertants have been obtained instead at a low frequency (e.g. see Royer- Pokora & Haseltine, 1984). The revertants probably arise in the course of long- lasting selection protocols using low doses of u.v. light. We think that the reason for the failure to obtain genuine transformants originates at least in part from the fact that XP cells (and also other human cells tested) are very restricted in the amount of DNA that becomes stably integrated in their genome after genomic DNA transfection. This problem may be overlooked because the relative ease and high efficiency by which transformants containing dominant marker genes are obtained in certain XP cell lines suggests that gene transfer is very efficient. However, dominant marker genes are present in the transfected DNA in vast molar excess and they are in general very small (less than 2 kilobases (kb)), increasing the chance that the gene remains intact during the transfection process. Relying on the transfection efficiency of dominant markers alone is deceptive. We have found that on average a 20- to 100fold smaller amount of exogenous DNA is present per human transformant compared with e.g. CHO, Ltk and NIH3T3 transformants (Hoeijmakers et al. 1987 α). It is, therefore, not surprising that to date most promising results have been obtained using CHO repair mutants induced in the laboratory. Transformants corrected by the uptake of human genes have been isolated for a number of excision-deficient CHO complementation groups (Rubin et al. 1983; Maclnnes et al. 1984; Thompson et al. 1985; our unpublished results). The first human repair gene has been cloned using one such mutant (Westerveld et al. 1984) and more genes are on the way. The repair gene isolated was obtained by DNA-mediated gene transfer of the excision-deficient CHO mutant 43-3B (constructed by Wood & Burki, 1982) that falls in the u.v. and mitomycin C (MM-C)-sensitive CHO complementation group 2 described by Thompson. The fact that linked transfer of the correcting human gene and a covalently attached dominant marker gene could be achieved in secondary transformations greatly facilitated the eventual cloning of the gene (Westerveld et al. 1984). The gene (designated ERCC-1) induces concomitant u.v. and MM-C resistance after transfection to 43-3B cells and other mutants of the same complementation group. Mutants of the other four excision-deficient CHO complementation groups are not corrected by this gene (results obtained in collaboration with Dr L. Thompson).

Phenotypic characterization of ERCC-1 correction

ERCC-1 corrects the repair defect of the 43-3B mutant for all repair parameters tested: u.v. and MM-C survival, sensitivity to 4NQO, NAcAAF and alkylating agents, u.v.-induced UDS, dimer removal as measured by the T4 endonuclease assay, mutability and induced chromosomal aberrations (Zdzienicka et al. 1987; Daroudi, unpublished results). Although the human repair gene restores the full spectrum of impaired repair functions of 43-3 B, the complementation is not for all end-points to the wild-type level. In particular, u.v. and MM-C survival is generally lagging behind that of the parental line, even when multiple ERCC-1 gene copies are integrated in the 43-3B genome. This suggests that it is not due to a gene dose effect. A plausible explanation might be that the human gene product is unable to substitute fully for the CHO counterpart in the excision process.

Molecular characterization of ERCC-1

ERCC-1 was cloned on cosmid 43-34. Subsequent analysis of the insert by various methods revealed thatERCC-1 resided in a 15 –17 kb region of cosmid 43-34. Unique probes derived from this area were used to probe Northern filters and to screen cDNA libraries. On Northern blots hybridization was found mainly with an RNA species of 1 ·0 –1 ·1 kb, present in HeLa and a number of other human cell lines. A similarly sized transcript is found in CHO cells (including the 43-3B mutant), mouse testis and brain (unpublished results). The fact that the CHO mutant shows a hybridization pattern indistinguishable from that of the parental wild-type cells, suggests that the gene is not inactivated by gross deletions or promotor and processing mutations. (This was confirmed by Southern blot analysis of the mutant and normal gene.) To investigate whether ERCC-1 gene expression is induced by DNA-damaging treatments, HeLa cells were irradiated with u.v. (1 and 10 J m −2). At various times after irradiation poly(A)+ RNA was extracted and analysed by Northern hybridization. No significant quantitative differences were observed, suggesting that ERCC-1 does not belong to a u.v.-inducible set of genes, like the SOS genes in E. coli and some of the yeast repair genes. Analysis of the ERCC-1 promotor region is in progress to obtain more information on the regulation of ERCC-1 transcription.

Screening of the Okayama/Berg (1983) cDNA library with genomic ERCC-1 probes resulted in the isolation of three cDNA clones, all of which appeared to be incomplete. However, by recombining different parts of each clone we succeeded in producing complete cDNA versions of the ERCC-1 transcript, which permitted analysis of the ERCC-1 gene structure and its gene product (van Duin et al. 1986). Comparison of the cDNA sequence with the genomic DNA revealed that the gene is split over 10 exons. Sequence analysis of different cDNAs and nuclease Si digestion experiments indicated that one exon of 72 base-pairs (bp) (encoding an internal 24 amino acid portion of the predicted amino acid sequence of the ERCC-1 gene product) is alternatively spliced. Hence, the 1 ·0 –1 ·1 kb mRNA band, visible on Northern blots is in fact made up of two unresolved mRNA species, differing by 72 bp in size. To investigate the function of the two cDNAs, transfection experiments were carried out to 43-3B cells. The cDNA derived from the larger transcript (pcDE) inserted into an Okayama/Berg (1983) mammalian expression vector was able to confer both u.v. and MM-C resistance on the mutant, in contrast to the cDNA of the small transcript (pcDE-72). Apparently, the presence of the 72bp exon is essential for the excision repair process. The function of the small RNA (if any) remains obscure. Using the Maxam & Gilbert (1980) sequencing technique the nucleotide sequences of both the 1098 and 1026 bp ERCC-1 cDNAs were determined and are presented in Fig. 2 along with the deduced amino acid sequence of the only likely open reading frame. The alternatively spliced 72bp exon, absent in pcDE-72 is underlined. The predicted ERCC-1 gene products have sizes of 297 and 274 amino acids (Mr 32 562 and 29 993, respectively). At present, we are utilizing E. coli expression systems and gene amplification in CHO cells to produce large quantities of the ERCC-1 gene products for isolation and functional characterization.

Fig. 2.

Nucleotide and deduced amino acid sequence of the ERCC-1 cDNA clone pcDE. The alternatively spliced 72 bp exon, absent in pcDE-72, is underlined. Regions exhibiting homology with functional protein domains are boxed. NLS, nuclear location signal; helix-turn-helix, DNA binding domain; ADP-rs, ADP-ribosylation site. The arrowhead points to the arginine résidu that by homology with other ADP-ribosylation sites is suggested to be the actual site for mono-ADP-ribosylation, e.g. by cholera toxin (see Hoeijmakers et al. 1987&, for details). Asterisk, stop codon. Interrupted underlining, polyadenylation signal.

Fig. 2.

Nucleotide and deduced amino acid sequence of the ERCC-1 cDNA clone pcDE. The alternatively spliced 72 bp exon, absent in pcDE-72, is underlined. Regions exhibiting homology with functional protein domains are boxed. NLS, nuclear location signal; helix-turn-helix, DNA binding domain; ADP-rs, ADP-ribosylation site. The arrowhead points to the arginine résidu that by homology with other ADP-ribosylation sites is suggested to be the actual site for mono-ADP-ribosylation, e.g. by cholera toxin (see Hoeijmakers et al. 1987&, for details). Asterisk, stop codon. Interrupted underlining, polyadenylation signal.

Homology of ERCC-1 gene products with other polypeptides

To obtain information on the role of ERCC-1 in excision repair the deduced ERCC-1 amino acid sequence was compared with those of other proteins and functional protein domains to search for homology. Within the class of repair polypeptides extensive amino acid homology was detected with the yeast excision repair protein RADIO, which is predicted to have a size of 210 amino acid residues (Reynolds et al. 1985a). The level of sequence identity is highest between the C- terminal half of RADIO and the middle 120 amino acids of ERCC-1 (34% identity). However, detectable homology exists also in the N-terminal parts of both proteins (Fig. 3). This finding supports the idea that ERCC-1 and RADIO are descendants of the same primordial gene and that they fulfil similar functions. The apparent C- terminal extension of the ERCC-1 gene product compared to the deduced RADIO protein might suggest that ERCC-1 has gained (or retained) additional functions not encoded by RADIO. In this respect it is worth noting that the 72 bp alternatively spliced exon, which is essential for excision repair in 43-3B, is located in the ‘extra’ part of ERCC-1. Transfection experiments of ERCC-1, RAD 10 and ERCC-1 /RAD 10 hybrid genes to the yeast and CHO repair mutants will have to be done to reveal to what extent the two proteins are functionally related.

Fig. 3.

Homology of the predicted protein sequence of ERCC-1 to that of RADIO and uvrA. Sequence identities are indicated by thick boxes. Physicochemically closely related amino acids (K, R; D, E; I, L, V) by thin boxes and weakly related amino acids (see Schwartz & Dayhoff, 1978, for group classification) by thin underlining. The standard one-letter amino acid abbreviations are used. Numbering corresponds with the residue in the respective protein. The RADIO, and uvrA sequences are from Reynolds et al. (1985a), and Husain et al. (1986), respectively.

Fig. 3.

Homology of the predicted protein sequence of ERCC-1 to that of RADIO and uvrA. Sequence identities are indicated by thick boxes. Physicochemically closely related amino acids (K, R; D, E; I, L, V) by thin boxes and weakly related amino acids (see Schwartz & Dayhoff, 1978, for group classification) by thin underlining. The standard one-letter amino acid abbreviations are used. Numbering corresponds with the residue in the respective protein. The RADIO, and uvrA sequences are from Reynolds et al. (1985a), and Husain et al. (1986), respectively.

In addition to homology with RAD10, computer analysis indicated the existence of a 42 amino acid homologous region between ERCC-1 and uvrA (31% identity) (Fig. 3). In ERCC-1 this stretch includes the point where the alignment with RAD10 stops. If this region represents a functional domain shared by uvrA and ERCC-1 it is remarkable that it is only for the first part present in RAD 10, since uvrA and ERCC-1 are from such distant organisms and RAD10 should be closer to both of them. Further analysis has to be done to clarify the significance of this observation.

A search for homology of ERCC-1 amino acid sequences with identified functional protein domains revealed homology with nuclear location signals (NLS), DNA binding domains (helix-turn-helix motive) and adenosine phosphate (ADP) ribosylation sites found in other polypeptides. The position of these tentative functional domains in ERCC-1 is indicated in Fig. 2. Details on the amino acid sequence comparisons have been published elsewhere (Hoeijmakers et al. 1987b). The NLS is situated in the N terminus of the ERCC-1 protein and suggests that the ERCC-1 gene product is actively transported into the nucleus. Transfection experiments with a ERCC-1 cDNA clone lacking the first 53 amino acids have shown that this part of the ERCC-1 protein, which includes the tentatively identified NLS, is not essential for correction of the 43-3 B cells (Van Duin et al. 1986). It is possible that passive diffusion of the protein through the nuclear pore complex is sufficient to allow the transformants to survive our u.v. and MM-C selection conditions. Alternatively, the ERCC-1 protein molecules might take the opportunity to reach (and possibly bind to) the DNA when the nuclear membrane is temporarily absent during mitosis. The proposed DNA binding property of ERCC-1 could be relevant in this respect. The potential DNA binding domain is located in the middle part of ERCC-1. It coincides with the region that exhibits the highest level of sequence conservation with RADIO (Van Duin et al. 1986). Finally, the tentative ADP ribosylation site is located in the C-terminal portion. Intriguingly, it is positioned precisely at the border between exon 7 and the alternatively spliced exon 8. Transcripts lacking this exon would specify a gene product that lacks amino acid sequences thought to be essential for the ADP-ribosyl acceptor function. Although this and the other suggested properties of the ERCC-1 gene product(s) are not unexpected for repair proteins, definite proof awaits experiments with the isolated protein. Production and purification of the ERCC-1 protein(s) therefore have a high priority in our research programme, as well as the isolation of additional repair genes. It is hoped that this will lead to a detailed understanding of the mechanism of excision repair in higher organisms, as in E. coli.

The following persons were involved in the work presented here: MrM. Van Duin, Dr A. Yasui, Mr W. Vermeulen, Mrs J. Zwetsloot, Mrs H. Odijk, Mr J. de Wit, Dr A. Westerveld and Dr D. Bootsma.

We are very grateful to Mr M. Koken and P. ten Dyke for help in some of the experiments, Dr R. van Gorcom (Medical Biological Laboratory, TNO, Rijswijk, The Netherlands) for valuable assistance in operating the computer and Dr H. Okayama for the generous gift of the cDNA library. Drs Eker (Delft), van Zeeland and Backendorf (Leiden) are thanked for the gift of enzymes. Furthermore, we thank Mrs R. Boucke for skilful typing of the manuscript and Mr T. van Os for photography. This work was supported by FUNGO (Foundation of Medical Scientific Research in the Netherlands) and EURATOM contract no. B16-141-NL.

Adzuma
,
K.
,
Ogawa
,
T.
&
Ogawa
,
H.
(
1984
).
Primary structure of the RAD52 gene in Saccharomyces cerevisiae.
Molec. cell. Biol.
4
,
2735
2744
.
Cleaver
,
J. E.
(
1968
).
Defective repair replication of DNA in xeroderma pigmentosum
.
Nature, Lond.
218
,
652
656
.
de Jonge
,
A. J. R.
,
Vermeulen
,
W.
,
Keuzer
,
W.
,
Hoeijmakers
,
J. H. J.
&
Bootsma
,
D.
(
1985
).
Microinjection of Micrococcus luteus UV-endonuclease restores UV-induced unscheduled DNA synthesis in cells of 9 xeroderma pigmentosum complementation groups
.
Mutat. Res.
150
,
99
105
.
de Jonge
,
A. J. R.
,
Vermeulen
,
W.
,
Klein
,
B.
&
Hoeijmakers
,
J. H. J.
(
1983
).
Microinjection of human cell extracts corrects xeroderma pigmentosum defect
.
EMBOJ.
2
,
637
641
.
de Weerd-Kastelein
,
E. A.
,
Keuzer
,
W.
&
Bootsma
,
D.
(
1972
).
Genetic heterogeneity of xeroderma pigmentosum demonstrated by somatic cell hybridization
.
Nature, new Biol.
238
,
80
83
.
Fischer
,
E.
,
Keuzer
,
W.
,
Thielmann
,
H. W.
,
Popanda
,
O.
,
Bohnert
,
E.
,
Edler
,
L.
,
Jung
,
E. G.
&
Bootsma
,
D.
(
1985
).
A ninth complementation group in xeroderma pigmentosum, XP-I
.
Mutat. Res.
145
,
217
225
.
Friedberg
,
E. C.
(
1985
).
DNARepair. San Francisco: Freeman & Company
.
Giannelli
,
F.
,
Pawsey
,
S. A.
&
Avery
,
J. A.
(
1982
).
Differences in patterns of complementation of the more common groups of xeroderma pigmentosum: possible implications
.
Cell
29
,
451
458
.
GRAFSTROM
,
R. H.
,
Park
,
L.
&
Grossman
,
L.
(
1982
).
Enzymatic repair of pyrimidine dimercontaining DNA: a 5’-dimer DNA glycosylase-3’ apyrimidinic endonuclease mechanism from Micrococcus luteus.
J. biol. Chem.
257
,
3465
3473
.
Haseltine
,
W. A.
,
Gordon
,
L. K.
,
Lindan
,
C. P.
,
Grafstrom
,
R. H.
,
Shaper
,
N. L.
&
Grossman
,
L.
(
1980
).
Cleavage of pyrimidine dimers in specific DNA sequences by a pyrimidine dimer DNA-glycosylase of M. luteus. Nature, Lond.
285
,
634
641
.
Hayakawa
,
H.
,
Ishizaki
,
K.
,
Inoue
,
M.
,
Yagi
,
T.
,
Sekiguchi
,
M.
&
Takebe
,
H.
(
1981
).
Repair of ultraviolet radiation damage in xeroderma pigmentosum cells belonging to complementation group F
.
Mutat. Res.
80
,
381
388
.
Hoeijmakers
,
J. H. J.
,
Oduk
,
H.
&
Westerveld
,
A.
(
1987a
).
Differences between rodent and human cell lines in the amount of integrated DNA after transfection
.
Expl Cell Res. (in press)
.
Hoeijmakers
,
J. H. J.
,
van Duin
,
M.
,
Westerveld
,
A.
,
Yasui
,
A.
&
Bootsma
,
D.
(
1987b
).
Identification of DNA-repair genes in the human genome
.
Cold Spring Harbor. Symp. quant. Biol, (in press)
.
Husain
,
I.
,
van Houten
,
B.
,
Thomas
,
D. C.
&
Sancar
,
A.
(
1986
).
Sequence of Escherichia coliuvr A gene and protein reveal two potential ATP binding sites
.
J. biol. Chem.
261
,
4895
4901
.
Keuzer
,
W.
,
Verkerk
,
A.
&
Bootsma
,
D.
(
1982
).
Phenotypic correction of the defect in xeroderma pigmentosum after fusion with isolated cytoplasts
.
Expl Cell Res.
140
,
119
125
.
Legerski
,
R. J.
,
Brown
,
D. B.
,
Peterson
,
C. A.
&
Robberson
,
D. L.
(
1984
).
Transient complementation of xeroderma pigmentosum cells by microinjection of poly(A) + RNA
.
Proc, natn. Acad. Sci. U.S A.
81
,
5676
5679
.
Lehmann
,
A. R.
(
1985
).
Use of recombinant DNA techniques in cloning DNA repair genes and in the study of mutagenesis in human cells
.
Mutat. Res.
150
,
61
67
.
MacInnes
,
M. A.
,
Bingham
,
J. D.
,
Thompson
,
L. H.
&
Strniste
,
G. F.
(
1984
).
DNA-mediated cotransfer of excision repair capacity and drug resistance into Chinese hamster ovary cell line UV-135
.
Molec. cell. Biol.
4
,
1152
1158
.
Maxam
,
A. M.
&
Gilbert
,
W.
(
1980
).
Sequence end-labeled DNA with base-specific chemical cleavages
.
Meth. Enzym.
65
,
499
560
.
Naumovski
,
L.
,
Chu
,
G.
,
Berg
,
P.
&
Friedberg
,
E. C.
(
1985
).
RAD3 gene of Saccharomyces cerevisiae: nucleotide sequence of wild type and mutant alleles, transcript mapping, and aspects of gene regulation
.
Molec. cell. Biol.
5
,
17
26
.
Nicolet
,
C. M.
,
Chenevert
,
J. M.
&
Friedberg
,
E. C.
(
1985
).
The RAD2 gene of Saccharomyces cerevisiae-. nucleotide sequence and transcript mapping
.
Gene
36
,
225
234
.
Okayama
,
H.
&
Berg
,
P.
(
1983
).
A cDNA cloning vector that permits expression of cDNA inserts in mammalian cell
.
Molec. cell. Biol.
3
,
280
289
.
Perozzi
,
G.
&
Prakash
,
S.
(
1986
).
RAD7 gene of Saccharomyces cerevisae-. Transcripts, nucleotide sequence analysis, and functional relationship between the RAD7 and RAD23 gene products
.
Molec. cell. Biol.
6
,
1497
1507
.
Reynolds
,
P.
,
Prakash
,
L.
,
Dumais
,
D.
,
Perozzi
,
G.
&
Prakash
,
S.
(
1985a
).
Nucleotide sequence of the RADIO gene of Saccharomyces cerevisiae.
EMBOy.
4
,
3549
3552
.
Reynolds
,
P.
,
Weber
,
S.
&
Prakash
,
L.
(
1985b
).
RAD6 gene of Saccharomyces cerevisiaeencodes a protein containing a tract of 13 consecutive aspartates
.
Proc. natn. Acad. Sci. U.SA.
82
,
168
172
.
Royer-Pokora
,
B.
&
Haseltine
,
W. A.
(
1984
).
Isolation of UV-resistant revertants from a xeroderma pigmentosum complementation group A cell line
.
Nature, Lond.
311
,
390
392
.
Rubin
,
J. S.
,
Joyner
,
A. I.
,
Bernstein
,
A.
&
Whitmore
,
G. F.
(
1983
).
Molecular identification of a human repair gene following DNA-mediated gene transfer
.
Nature, Lond.
306
,
206
208
.
Schwartz
,
R. M.
&
Dayhoff
,
M. D.
(
1978
). Matrices for detecting distant relationships.
In Atlas of Protein Sequence and Structure
(ed.
M. D.
Dayhoff
), vol.
5
, uppl. 3, pp.
353
358
.
Washington, DC
:
National Biomedical Research Foundation
.
Tanaka
,
K.
,
Hayakawa
,
H.
,
Sekiguchi
,
M.
&
Okada
,
Y.
(
1977
).
Specific action of T4 endonuclease V on damaged DNA in xeroderma pigmentosum cells in vivo.
Proc. natn. Acad. Sci. U.SA.
74
,
2958
2962
.
Tanaka
,
K.
,
Sekiguchi
,
M.
&
Okada
,
Y.
(
1975
).
Restoration of ultraviolet-induced unscheduled DNA synthesis of xeroderma pigmentosum cells by the concomittant treatment with bacteriophage T4 endonuclease V and HVJ (Sendai virus)
.
Proc. natn. Acad. Sci. U.SA.
72
,
4071
4075
.
Teo
,
I.
,
Sedgwick
,
B.
,
Kilpatrick
,
M. W.
,
McCarthy
,
T. V.
&
Lindahl
,
T.
(
1986
).
The intracellular signal for induction of resistance to alkylating agents
in
E. coli. Cell
48
,
315
324
.
Thompson
,
L. H.
,
Brookman
,
K. W.
,
Dillehay
,
L. E.
,
Mooney
,
C. L.
&
Carrano
,
A. V.
(
1982
).
Hypersensitivity to mutation and sister-chromatid exchange induction in CHO cell mutants defective in incising DNA containing UV-lesions
.
Somat. Cell Genet.
8
,
759
—773.
Thompson
,
L. H.
,
Brookman
,
K. W.
,
Minkler
,
J. L.
,
Fuscoe
,
J. C.
,
Henning
,
K. A.
&
Carrano
,
A. V.
(
1985
).
DNA-mediated transfer of a human DNA repair gene that controls sister chromatid exchange
.
Molec. cell. Biol.
5
,
881
884
.
Thompson
,
L. H.
,
Busch
,
D. B.
,
Brookman
,
K. W.
,
Mooney
,
C. L.
&
Glaser
,
P. A.
(
1981
).
Genetic diversity of UV-sensitive DNA-repair mutants of Chinese hamster ovary cells
.
Proc, natn. Acad. Sci. U.SA.
78
,
3734
3737
.
Thompson
,
L. H.
&
Carrano
,
A. V.
(
1983
). Analysis of mammalian cell mutagenesis and DNA repair using in vitro selected CHO cell mutants. In
Cellular Responses to DNA Damage
(ed.
E. C.
Friedberg
&
B. R.
Bridges
), U.C.L.A. Symp. Molec. Cell. Biol. New Series, vol.
11
, pp.
125
143
.
New York
:
Alan R. Liss
.
Van Duin
,
M.
,
de Wit
,
J.
,
Odhk
,
H.
,
Westerveld
,
A.
,
Yasui
,
A.
,
Koken
,
M. H. M.
,
Hoeiimakers
,
J. H. J.
&
Bootsma
,
D.
(
1986
).
Molecular characterization of the human excision repair gene ERCC-1: cDNA cloning and amino acid homology with the yeast DNA repair gene RADIO. Cell
44
,
913
923
.
Vermeulen
,
W.
,
Osseweijer
,
P.
, de
Jonge
,
A. J. R.
&
Hoeijmakers
,
J. H. J.
(
1986
).
Transient correction of excision repair defects in fibroblasts of 9 xeroderma pigmentosum complementation groups by microinjection of crude human cell extracts
.
Mutat. Res.
165
,
199
206
.
Walker
,
G. C.
(
1985
).
Inducible DNA repair systems
.
A. Rev. Biochem.
54
,
425
457
.
Westerveld
,
A.
,
Hoeijmakers
,
J. H. J.
,
Van Duin
,
M.
,
de Wit
,
J.
,
Odijk
,
H.
,
Pastink
,
A.
&
Bootsma
,
D.
(
1984
).
Molecular cloning of a human DNA repair gene
.
Nature, Lond.
310
,
425
429
.
Wood
,
R. D.
&
Burki
,
H. J.
(
1982
).
Repair capability and the cellular age response for killing and mutation induction after UV
.
Mutat. Res.
95
,
505
514
.
Yasui
,
A.
&
Chevallier
,
M.-R.
(
1983
).
Cloning of photoreactivation repair gene and excision repair gene of the yeast Saccharomyces cerevisiae.
Curr. Genet.
7
,
191
194
.
Zdzienicka
,
M. Z.
,
Roza
,
L.
,
Westerveld
,
A.
,
Bootsma
,
D.
&
Simons
,
J. W. I. M.
(
1987
).
Biological and biochemical consequences of the human ERCC-1 gene after transfection into a repair-deficient CHO cell line
.
Mutat. Res. (in press)
.
Zwetsloot
,
J. C. M.
,
Barbeiro
,
A. P.
,
Vermeulen
,
W.
,
Arthur
,
H. M.
,
Hoeijmakers
,
J. H. J.
&
Backendorf
,
C.
(
1986a
).
Microinjection of Escherichia coli uvrA,B,C and D proteins into fibroblasts of xeroderma pigmentosum groups A and C does not result in restoration of UV- induced unscheduled DNA synthesis
.
Mutat. Res.
166
,
89
98
.
Zwetsloot
,
J. C. M.
,
Hoeijmakers
,
J. H. J.
,
Vermeulen
,
W.
,
Eker
,
A. P. M.
&
Bootsma
,
D.
(
1986b
).
Unscheduled DNA synthesis in xeroderma pigmentosum cells after microinjection of yeast photoreactivating enzyme
.
Mutat. Res.
165
,
109
115
.
Zwetsloot
,
J. C. M.
,
Vermeulen
,
W.
,
Hoeijmakers
,
J. H. J.
,
Yasui
,
A.
,
Eker
,
A. P. M.
&
Bootsma
,
D.
(
1985
).
Microinjected photoreactivating enzymes from Anacystis and Saccharomyces monomerize dimers in chromatin of human cells
.
Mutat. Res.
146
,
71
77
.