SKY1 is involved in cisplatin-induced cell kill in Saccharomyces cerevisiae, and inactivation of its human homologue, SRPK1, induces cisplatin resistance in a human ovarian carcinoma cell line.

The therapeutic potential of cisplatin, one of the most active and widely used anticancer drugs, is severely limited by the occurrence of cellular resistance. In this study, using budding yeast Saccharomyces cerevisiae as a model organism to identify novel drug resistance genes, we found that disruption of the yeast gene SKY1 (serine/arginine-rich protein-specific kinase from budding yeast) by either transposon insertion or one-step gene replacement conferred cellular resistance to cisplatin. Heterologous expression of the human SKY1 homologue SRPK1 (serine/arginine-rich protein-specific kinase) in SKY1 deletion mutant yeast cells restored cisplatin sensitivity, suggesting that SRPK1 is a cisplatin sensitivity gene, the inactivation of which could lead to cisplatin resistance. Subsequently, we investigated the role of SRPK1 in cisplatin sensitivity and resistance in human ovarian carcinoma A2780 cells using antisense oligodeoxynucleotides. Treatment of A2780 cells with antisense oligodeoxynucleotides directed against the translation initiation site of SRPK1 led to down-regulation of SRPK1 protein and conferred a 4-fold resistance to cisplatin. The human SRPK1 gene has not been associated with drug resistance before. Our new findings strongly suggest that SRPK1 is involved in cisplatin-induced cell kill and indicate that SRPK1 might potentially be of importance for studying clinical drug resistance.

Characterization of RAD52 homologs in the fission yeast Schizosaccharomyces pombe.

The RAD52 gene of Saccharomyces cerevisiae is essential for repair of DNA double-strand breaks (DSBs) by homologous recombination. Inactivation of this gene confers hypersensitivity to DSB-inducing agents and defects in most forms of recombination. The rad22+ gene in Schizosaccharomyces pombe (here referred to as rad22A+) has been characterized as a homolog of RAD52 in fission yeast. Here, we report the identification of a second RAD52 homolog in Schizosaccharomyces pombe, called rad22B+. The amino acid sequences of Rad22A and Rad22B show significant conservation (38% identity). Deletion mutants of respectively, rad22A and rad22B, show different phenotypes with respect to sensitivity to X-rays and the ability to perform homologous recombination as measured by the integration of plasmid DNA. Inactivation of rad22A+ leads to a severe sensitivity to X-rays and a strong decrease in recombination (13-fold), while the rad22B mutation does not result in a decrease in homologous recombination or a change in radiation sensitivity. In a rad22A-rad22B double mutant the radiation sensitivity is further enhanced in comparison with the rad22A single mutant. Overexpression of the rad22B+ gene results in partial suppression of the DNA repair defects of the rad22A mutant strain. Meiotic recombination and spore viability are only slightly affected in either single mutant, but outgrowth of viable spores is almost 31-fold reduced in the rad22A-rad22B double mutant. The results obtained imply a crucial role for rad22A+ in repair and recombination in vegetative cells just like RAD52 in S. cerevisiae. The rad22B+ gene presumably has an auxiliary role in the repair of DSBs. The drastic reduced spore viability in the double mutant suggests that meiosis in S. pombe is dependent on the presence of either rad22A+ or rad22B+.

Spt4 modulates Rad26 requirement in transcription-coupled nucleotide excision repair.

The nucleotide excision repair machinery can be targeted preferentially to lesions in transcribed sequences. This mode of DNA repair is referred to as transcription-coupled repair (TCR). In yeast, the Rad26 protein, which is the counterpart of the human Cockayne syndrome B protein, is implicated specifically in TCR. In a yeast strain genetically deprived of global genome repair, a deletion of RAD26 renders cells UV sensitive and displays a defect in TCR. Using a genome-wide mutagenesis approach, we found that deletion of the SPT4 gene suppresses the rad26 defect. We show that suppression by the absence of Spt4 is specific for a rad26 defect and is caused by reactivation of TCR in a Rad26-independent manner. Spt4 is involved in the regulation of transcription elongation. The absence of this regulation leads to transcription that is intrinsically competent for TCR. Our findings suggest that Rad26 acts as an elongation factor rendering transcription TCR competent and that its requirement can be modulated by Spt4.

Identification and characterization of the rhp23(+) DNA repair gene in Schizosaccharomyces pombe.

We have identified rhp23(+), the ortholog of the Saccharomyces cerevisiae RAD23 and human HHR23A and HHR23B genes, in Schizosaccharomyces pombe and examined its role in cell survival and DNA repair. In S. pombe two repair mechanisms are operative on UV-induced photoproducts, i.e., UV damage repair (UVDR) and nucleotide excision repair (NER). Here we show that Rhp23 is solely involved in NER and study its role in DNA repair in the absence of the UVDR pathway. S. pombe rhp23-deficient cells are sensitive toward UV irradiation, although not as sensitive as complete NER-deficient cells. Furthermore we demonstrate that the residual survival observed in rhp23-deficient cells is NER dependent. Despite this NER-dependent survival, uvde rhp23 double mutants are unable to repair cyclobutane pyrimidine dimers. The inability to remove these photolesions from both DNA strands clearly demonstrates that rhp23(+) is involved in transcription coupled repair as well as global genome repair.

Characterization of the rhp7(+) and rhp16(+) genes in Schizosaccharomyces pombe.

The global genome repair (GGR) subpathway of nucleotide excision repair (NER) is capable of removing lesions throughout the genome. In Saccharomyces cerevisiae the RAD7 and RAD16 genes are essential for GGR. Here we identify rhp7 (+), the RAD7 homolog in Schizosaccharomyces pombe. Surprisingly, rhp7 (+)and the previously cloned rhp16 (+)are located very close together and are transcribed in opposite directions. Upon UV irradiation both genes are induced, reaching a maximum level after 45-60 min. These observations suggest that the genes are co-regulated. Schizo-saccharomyces pombe rhp7 or rhp16 deficient cells are, in contrast to S.cerevisiae rad7 and rad16 mutants, not sensitive to UV irradiation. In S.pombe an alternative repair mechanism, UV damage repair (UVDR), is capable of efficiently removing photolesions from DNA. In the absence of this UVDR pathway both rhp7 and rhp16 deficient cells display an enhanced UV sensitivity. Epistatic analyses show that rhp7 (+)and rhp16 (+)are only involved in NER. Repair analyses at nucleotide resolution demonstrate that both Rhp7 and Rhp16, probably acting in a complex, are essential for GGR in S.pombe.

Removal of cyclobutane pyrimidine dimers by the UV damage repair and nucleotide excision repair pathways of Schizosaccharomyces pombe at nucleotide resolution.

In Schizosaccharomyces pombe two different repair mechanisms remove UV-induced lesions from DNA, i.e. nucleotide excision repair (NER) and UV damage repair (UVDR). Here, the kinetics of removal of cyclobutane pyrimidine dimers (CPDs) by both pathways is determined at base resolution in the transcribed strand (TS) and the non-transcribed strand (NTS) of the sprpb2 +gene. UVDR does not remove lesions in a strand-specific manner, indicating that UVDR is neither stimulated nor inhibited by RNA polymerase II transcription. In contrast, in a UVDR-deficient strain the TS is repaired preferentially. This strong strand bias suggests that in S.pombe, as in other species, NER is coupled to transcription. In repair-proficient S.pombe the TS is repaired very rapidly, as a consequence of two efficiently operating pathways, while the NTS is repaired more slowly, mainly by UVDR. Furthermore, we demonstrate that UVDR is not always faster than NER.

Cloning of Schizosaccharomyces pombe rph16+, a gene homologous to the Saccharomyces cerevisiae RAD16 gene.

The RAD16 gene is involved in the nucleotide excision repair of UV damage in the transcriptional silenced mating type loci (Terleth et al., 1990 and Bang et al., 1992) and in non-transcribed stands of active genes in Saccharomyces cerevisiae (Verhage et al., 1994). Using touchdown-PCR with primers derived from various domains of the S. cerevisiae Rad 16 protein, a specific Schizosaccharomyces pombe probe was isolated. This probe was used to obtain the complete RAD16 homologous gene from a S. pombe chromosomal bank. DNA sequence analysis of the rph16+ gene revealed an open reading frame of 854 amino acids. Comparison of the amino acid sequences of the Rhp16 and Rad16 proteins showed a high level of conservation: 68% similarity. The Rhp16 protein sequence contains the two Zn-finger motifs and the putative helicase domains as found in the Rad16 protein. Like the RAD16, the rph16+ gene is UV-inducible (Bang et al., 1995). In analogy with the rad16 mutant, the rhp16 disruption mutant is viable and grows normally, indicating that the gene does not have an essential function. The rhp16 disruption mutant is not sensitive for UV but is sensitive for cisplatin. The rhp16+ gene cloned behind the GAI 1 promoter partially complements the UV sensitivity and the defect in the non-transcribed strand DNA repair of a S. cerevisiae rad16 mutant, indicating functional homology between the rhp16+ and RAD16 genes. The structural and functional homology between the two genes suggests that the RAD16 dependent subpathway of NER for the repair of non-transcribed DNA is evolutionary conserved.

The in vitro more efficiently repaired cisplatin adduct cis-Pt.GG is in vivo a more mutagenic lesion than the relative slowly repaired cis-Pt.GCG adduct.

The toxic effect and the mutagenicity of two differentially repaired site-specific cis-diamminedichloroplatinum(II) (cis-DDP) lesions were investigated. Detailed analysis of the UvrABC-dependent repair of the two lesions in vitro showed a more efficient repair of the cis-Pt.GG adduct compared to that of the cis-Pt.GCG adduct (Visse et al., 1994). Furthermore, previously, a dependency of cis-DDP mutagenesis on UvrA and UvrB, but not on UvrC was found (Brouwer et al., 1988). To possibly relate survival and mutagenesis to repair, plasmids containing the same site-specific cis-DDP lesions as those that were used in the detailed repair studies were transformed into Escherichia coli. The results indicate that both lesions are very efficiently bypassed in vivo. Mutation analysis was performed using a denaturing gradient gel electrophoresis technique, which allows identification of mutations without previous selection. Although the cis-Pt.GG adduct is in vitro more efficiently repaired than the cis-Pt.GCG adduct, it appeared to be more mutagenic. We present a model in which this result is related to the previously observed dependency of the mutagenicity of cis-DDP lesions on the Uvr A and B proteins.

The first zinc-binding domain of UvrA is not essential for UvrABC-mediated DNA excision repair.

Specific mutations in uvrA were introduced to analyze the role of the zinc-binding domains of the protein in DNA excision repair. Zinc-coordinating cysteines were substituted into non-coordinating serine or glycine residues. Mutations leading to changes in the second zinc-binding domain had a profound effect on UV survival in vivo; however these mutant proteins could not be isolated for in vitro analyses. Amino acid substitutions in the first zinc-binding domain had very little effect on UV survival in vivo. In vitro analyses showed that although this domain no longer coordinates zinc, ATPase activity, helicase activity, DNA binding, incision of damaged DNA and DNA repair synthesis appeared to be normal. Therefore it seems that the first zinc-binding domain of UvrA is not essential for DNA excision repair.

Uvr excision repair protein complex of Escherichia coli binds to the convex side of a cisplatin-induced kink in the DNA.

The Escherichia coli UvrABC endonuclease is capable of initiating the repair of a wide variety of DNA damages. To study the binding of the UvrAB complex to the DNA at the site of a lesion we have constructed a synthetic DNA fragment with a defined cis-diamminedichloroplatinum(II) (cis-Pt).GG adduct. The cis-Pt.GG is the major adduct after treatment of DNA with the antitumor agent cisplatin. Binding to the DNA at the site of the defined lesion was studied with DNase I and MPE.Fe(II) hydroxyl radical footprinting. The results indicate that the UvrAB complex binds to the convex side of the kink in the DNA caused by the cis-Pt.GG adduct. Concerted incisions of the damaged strand by the UvrABC endonuclease were at the 8th phosphodiester bond 5' to and at the 4th bond 3' of the adjacent guanines. An additional incision was found at the 15th phosphodiester bond 5' to the damaged site. This extra incision was stimulated by a high concentration of UvrC.

Identification of the uvrA6 mutation of Escherichia coli.

The uvrA6 mutation has been cloned on a multicopy plasmid by using a chloramphenicol resistance marker introduced next to the uvrA gene in the Escherichia coli chromosome. The mutation was shown to reside in the N-terminal part of the uvrA gene. Sequencing part of this region of the mutant gene revealed a frameshift mutation at positions 207 to 209, which leads to a stop codon at position 262. A marker rescue experiment showed that this frameshift is the only mutation responsible for the UV-sensitive phenotype of the UvrA6 mutant. The method presented is suitable for the cloning of every chromosomal uvrA mutation and can be useful for the study of the functional domains of the UvrA protein.

Location and cloning of the ultraviolet-sensitizing function from the chromosomally associated IncJ group plasmid, R391.

The IncJ plasmid R391, which specifies a uv-sensitizing function, has been shown to be associated with chromosomal DNA. Deletions originating from Tn10 insertion into the kanamycin-resistance determinant of plasmid R391 gave rise to uv-resistant derivatives. This apparent linkage between the kanamycin-resistance determinant and the uv-sensitizing gene(s) was used to clone the uv-sensitizing function from plasmid R391 into pUR222. A recombinant plasmid containing both functions (KanR and Uvs+) was obtained. The uv-sensitizing function was mapped to a 4-kb EcoRI fragment.

Analysis of the regulatory region of the ssb gene of Escherichia coli.

The regulation of the ssb gene of E. coli has been studied. We reported earlier that the SOS box of the neighbouring uvrA gene also controls the transcription of the ssb gene. Detailed analysis of the upstream region of ssb by S1 mapping reveals the existence of three in vivo functional promoters of which the most upstream one (PI) is inducible by DNA damage. Measurement of galactokinase synthesis using galK fusion plasmids indicates that the uninduced level of transcription from the PI promoter is low. Ssb multicopy plasmids lacking the PI promoter still complement the UV sensitivity of an Ssb mutant. The role of the three promoters in the regulation of the level of Ssb protein in the cell, is discussed.

A common regulatory region shared by divergently transcribed genes of the Escherichia coli SOS system.

The Escherichia coli single-stranded DNA binding protein (SSB) is implicated in DNA replication, recombination and repair. On the chromosome, the ssb gene is located adjacent to the excision repair gene uvrA, but the two genes are transcribed in opposite directions. uvrA has been shown to be part of the E. coli SOS system by introducing Mud(Ap, lac) insertions distal to the regulatory region of the gene in the chromosome. Recent investigations suggest that SSB is also involved in the SOS response. However, because the SSB protein is essential to the cell, the inducibility of the ssb gene cannot be investigated by the insertion method. Therefore, we used plasmids harbouring the regulatory region of ssb fused to the galK structural gene, while leaving an intact ssb gene in the chromosome. We show here that expression of the ssb gene is dependent on two promoters of which one is damage inducible. Evidence is presented that the divergently transcribed ssb and uvrA genes are controlled by a common LexA binding site.

In vivo regulation of the uvrA gene: role of the "-10" and "-35" promoter regions.

The effect of increasing deletions in the uvrA promoter region on the transcriptional efficiency was quantitatively analysed by fusion to the galK structural gene. A physical analysis of uvrA messenger RNA synthesis from the different deletion plasmids was performed using the S1 mapping technique. Both methods indicate that the uvrA "-10" promoter sequence is sufficient to trigger uvrA transcription. Although not essential, the "-35" region, which is overlapping with the LexA binding site, is shown to have an enhancing function, as the exposure of this region after SOS induction results in a 3- to 4-fold increase in uvrA transcription. A model is presented which accounts both for the observed basal and induced expression of the uvrA gene on a molecular level.

Effect of lexA and ssb genes, present on a uvrA recombinant plasmid, on the UV survival of Escherichia coli K-12.

The recombinant plasmid pJA01 contains, besides the uvrA gene, the genes lexA, ubiA and ssb. This plasmid does not fully complement a uvrA mutation in a Rec+ background. Plasmids which contain the uvrA and ssg genes, but not the lexA gene, show a higher but still only partial complementation. Full complementation achieved when the ssb gene us inactivated by insertion of Tn5. Furthermore, it appears that the presence of the ssb gene on a multicopy plasmid sensitizes wild-type cells to UV light. The effect of Ssb (single-strand DNA binding protein) overproduction on UV survival is discussed.

Use of transposons in cloning poorly selectable genes of Escherichia coli: cloning of uvrA and adjacent genes.

A transposon was introduced close to a poorly selectable gene. This gene could be cloned by using selection for the antibiotic resistance marker of the transposon.