Search

B-type eukaryotic polymerases contain a [4Fe-4S] cluster in their C-terminus domain, whose role is not fully understood yet. Among them, DNA polymerase delta (Polδ) plays an essential role in chromosomal DNA replication, mostly during lagging strand synthesis. Previous in vitro work suggested that the Fe-S cluster in Polδ is required for efficient binding of the Pol31 subunit, ensuring stability of the Polδ complex. Here, we analyzed the in vivo consequences resulting from an impaired coordination of the Fe-S cluster in Polδ. We show that a single substitution of the very last cysteine coordinating the cluster by a serine is responsible for the generation of massive DNA damage during S phase, leading to checkpoint activation, requirement of homologous recombination for repair, and ultimately to cell death when the repair capacities of the cells are overwhelmed. These data indicate that impaired Fe-S cluster coordination in Polδ is responsible for aberrant replication. More generally, Fe-S in Polδ may be compromised by various stress including anti-cancer drugs. Possible in vivo Polδ Fe-S cluster oxidation and collapse may thus occur, and we speculate this could contribute to induced genomic instability and cell death, comparable to that observed in pol3-13 cells., (© The Author(s) 2021. Published by Oxford University Press on behalf of Genetics Society of America.)

Human acute promyelocytic leukemia (APL) is characterized by a specific balanced translocation t(15;17)(q22;q21) involving the PML and RARA genes. In both de novo and therapy-related APL, the most frequent PML breakpoints are located within intron 6, and less frequently in intron 3; the precise mechanisms by which these breakpoints arise and preferentially in PML intron 6 remain unsolved. To investigate the intrinsic properties of the PML intron sequences in vivo, we designed Saccharomyces cerevisiae strains containing human PML intron 6 or intron 3 sequences inserted in yeast chromosome V and measured gross chromosomal rearrangements (GCR). This approach provided evidence that intron 6 had a superior instability over intron 3 due to an intrinsic property of the sequence and identified the 3' end of intron 6 as the most susceptible to break. Using yeast strains invalidated for genes that control DNA replication, we show that this differential instability depended at least upon Rrm3, a DNA helicase, and Mrc1, the human claspin homolog. GCR induction by hydrogen peroxide, a general genotoxic agent, was also dependent on genetic context. We conclude that: 1) this yeast system provides an alternative approach to study in detail the properties of human sequences in a genetically controlled situation and 2) the different susceptibility to produce DNA breaks in intron 6 versus intron 3 of the human PML gene is likely due to an intrinsic property of the sequence and is under replication fork genetic control.

The kinetic and spatial separation of redox systems renders redox biology studies a particularly challenging field. Genetically encoded biosensors including the glutathione-specific redox-sensitive yellow fluorescent protein (rxYFP) provide an alternative way to overcome the limitations of conventional glutathione/glutathione disulfide (GSH/GSSG) redox measurements. In this study, the plasmids expressing respectively cytosol-, nucleus-, and mitochondrial matrix- targeted rxYFP were created and introduced to human cervical carcinoma (HeLa) cells. The rxYFP redox states were monitored by direct assessment of the oxidized to reduced rxYFP ratio via redox protein extraction, redox Western blot and signal quantification. RxYFP proteins expressed in the cytosol, nucleus or mitochondrial matrix of HeLa cells were responsive to the intracellular redox state changes induced by reducing as well as oxidizing agents. Compartment-targeted rxYFP sensors were able to detect different steady-state redox conditions between the cytosol, nucleus and mitochondrial matrix. Furthermore, rxYFP sensors were able to sense dynamic and compartment-specific redox changes caused by 100µM hydrogen peroxide (H2O2). Mitochondrial matrix-targeted rxYFP displayed a greater dynamics of oxidation in response to a H2O2 challenge than the cytosol- and nucleus-targeted sensors, largely due to a more alkaline local pH environment. Our data provide direct evidence that mitochondrial glutathione redox state is maintained and regulated independently from that of the cytosol and nucleus. Complementary to existing redox sensors and conventional redox measurements, compartment-targeted rxYFP sensors provide a novel tool for examining mammalian cell redox homeostasis, permitting high resolution readout of steady glutathione state and dynamics of redox changes., (Copyright © 2014. Published by Elsevier Inc.)

The absence of Tsa1, a key peroxiredoxin that scavenges H2O2 in Saccharomyces cerevisiae, causes the accumulation of a broad spectrum of mutations. Deletion of TSA1 also causes synthetic lethality in combination with mutations in RAD51 or several key genes involved in DNA double-strand break repair. In the present study, we propose that the accumulation of reactive oxygen species (ROS) is the primary cause of genome instability of tsa1Δ cells. In searching for spontaneous suppressors of synthetic lethality of tsa1Δ rad51Δ double mutants, we identified that the loss of thioredoxin reductase Trr1 rescues their viability. The trr1Δ mutant displayed a Can(R) mutation rate 5-fold lower than wild-type cells. Additional deletion of TRR1 in tsa1Δ mutant reduced substantially the Can(R) mutation rate of tsa1Δ strain (33-fold), and to a lesser extent, of rad51Δ strain (4-fold). Loss of Trr1 induced Yap1 nuclear accumulation and over-expression of a set of Yap1-regulated oxido-reductases with antioxidant properties that ultimately re-equilibrate intracellular redox environment, reducing substantially ROS-associated DNA damages. This trr1Δ -induced effect was largely thioredoxin-dependent, probably mediated by oxidized forms of thioredoxins, the primary substrates of Trr1. Thioredoxin Trx1 and Trx2 were constitutively and strongly oxidized in the absence of Trr1. In trx1Δ trx2Δ cells, Yap1 was only moderately activated; consistently, the trx1Δ trx2Δ double deletion failed to efficiently rescue the viability of tsa1Δ rad51Δ. Finally, we showed that modulation of the dNTP pool size also influences the formation of spontaneous mutation in trr1Δ and trx1Δ trx2Δ strains. We present a tentative model that helps to estimate the respective impact of ROS level and dNTP concentration in the generation of spontaneous mutations.

Intracellular redox homeostasis is crucial for many cellular functions but accurate measurements of cellular compartment-specific redox states remain technically challenging. Genetically encoded biosensors including the glutathione-specific redox-sensitive yellow fluorescent protein (rxYFP) may provide an alternative way to overcome the limitations of conventional glutathione/glutathione disulfide (GSH/GSSG) redox measurements. This study describes the use of rxYFP sensors for investigating compartment-specific steady redox state and their dynamics in response to stress in human cells. RxYFP expressed in the cytosol, nucleus, or mitochondrial matrix of HeLa cells was responsive to the intracellular redox state changes induced by reducing as well as oxidizing agents. Compartment-targeted rxYFP sensors were able to detect different steady-state redox conditions among the cytosol, nucleus, and mitochondrial matrix. These sensors expressed in human epidermal keratinocytes HEK001 responded to stress induced by ultraviolet A radiation in a dose-dependent manner. Furthermore, rxYFP sensors were able to sense dynamic and compartment-specific redox changes caused by 100 μM hydrogen peroxide (H2O2). Mitochondrial matrix-targeted rxYFP displayed a greater dynamics of oxidation in response to a H2O2 challenge than the cytosol- and nucleus-targeted sensors, largely due to a more alkaline local pH environment. These observations support the view that mitochondrial glutathione redox state is maintained and regulated independently from that of the cytosol and nucleus. Taken together, our data show the robustness of the rxYFP sensors to measure compartmental redox changes in human cells. Complementary to existing redox sensors and conventional redox measurements, compartment-targeted rxYFP sensors provide a novel tool for examining mammalian cell redox homeostasis, permitting high-resolution readout of steady glutathione state and dynamics of redox changes., (Copyright © 2013 Elsevier Inc. All rights reserved.)

Peroxiredoxin (Prx) 1 is a member of the thiol-specific peroxidases family and plays diverse roles such as H2O2 scavenger, redox signal transducer and molecular chaperone. Prx1 has been reported to be involved in protecting cancer cells against various therapeutic challenges. We investigated how modulations of intracellular redox system affect cancer cell sensitivity to reactive oxygen species (ROS)-generating drugs. We observed that stable and transient Prx1 knockdown significantly enhanced HeLa cell sensitivity to β-lapachone (β-lap), a potential anticancer agent. Prx1 knockdown markedly potentiated 2 µM β-lap-induced cytotoxicity through ROS accumulation. This effect was largely NAD(P)H:quinone oxidoreductase 1 dependent and associated with a decrease in poly(ADP-ribose) polymerase 1 protein levels, phosphorylation of JNK, p38 and Erk proteins in mitogen-activated protein kinase (MAPK) pathways and a decrease in thioredoxin 1 (Trx1) protein levels. Trx1 serves as an electron donor for Prx1 and is overexpressed in Prx1 knockdown cells. Based on the fact that Prx1 is a major ROS scavenger and a partner of at least ASK1 and JNK, two key components of MAPK pathways, we propose that Prx1 knockdown-induced sensitization to β-lap is achieved through combined action of accumulation of ROS and enhancement of MAPK pathway activation, leading to cell apoptosis. These data support the view that modulation of intracellular redox state could be an alternative approach to enhance cancer cell sensitivity to ROS-generating drugs or to overcome some types of drug resistance.

Intracellular redox homeostasis is crucial for many cellular functions but accurate measurements of cellular compartment-specific redox states remain technically challenging. To better characterize redox control in the nucleus, we targeted a yellow fluorescent protein-based redox sensor (rxYFP) to the nucleus of the yeast Saccharomyces cerevisiae. Parallel analyses of the redox state of nucleus-rxYFP and cytosol-rxYFP allowed us to monitor distinctively dynamic glutathione (GSH) redox changes within these two compartments under a given condition. We observed that the nuclear GSH redox environment is highly reducing and similar to the cytosol under steady-state conditions. Furthermore, these sensors are able to detect redox variations specific for their respective compartments in glutathione reductase (Glr1) and thioredoxin pathway (Trr1, Trx1, Trx2) mutants that have altered subcellular redox environments. Our mutant redox data provide in vivo evidence that glutathione and the thioredoxin redox systems have distinct but overlapping functions in controlling subcellular redox environments. We also monitored the dynamic response of nucleus-rxYFP and cytosol-rxYFP to GSH depletion and to exogenous low and high doses of H₂O₂ bursts. These observations indicate a rapid and almost simultaneous oxidation of both nucleus-rxYFP and cytosol-rxYFP, highlighting the robustness of the rxYFP sensors in measuring real-time compartmental redox changes. Taken together, our data suggest that the highly reduced yeast nuclear and cytosolic redox states are maintained independently to some extent and under distinct but subtle redox regulation. Nucleus- and cytosol-rxYFP register compartment-specific localized redox fluctuations that may involve exchange of reduced and/or oxidized glutathione between these two compartments. Finally, we confirmed that GSH depletion has profound effects on mitochondrial genome stability but little effect on nuclear genome stability, thereby emphasizing that the critical requirement for GSH during growth is linked to a mitochondria-dependent process., (Copyright © 2012 Elsevier Inc. All rights reserved.)

Template switching induced by stalled replication forks has recently been proposed to underlie complex genomic rearrangements. However, the resulting models are not supported by robust physical evidence. Here, we analyzed replication and recombination intermediates in a well-defined fission yeast system that blocks replication forks. We show that, in response to fork arrest, chromosomal rearrangements result from Rad52-dependent nascent strand template exchange occurring during fork restart. This template exchange occurs by both Rad51-dependent and -independent mechanisms. We demonstrate that Rqh1, the BLM homolog, limits Rad51-dependent template exchange without affecting fork restart. In contrast, we report that the Srs2 helicase promotes both fork restart and template exchange. Our data demonstrate that template exchange occurs during recombination-dependent fork restart at the expense of genome rearrangements., (Copyright 2010 Elsevier Inc. All rights reserved.)

A mutated allele of the essential gene TAH18 was previously identified in our laboratory in a genetic screen for new proteins interacting with the DNA polymerase delta in yeast [1]. The present work shows that Tah18 plays a role in response to oxidative stress. After exposure to lethal doses of H(2)O(2), GFP-Tah18 relocalizes to the mitochondria and controls mitochondria integrity and cell death. Dre2, an essential Fe/S cluster protein and homologue of human anti-apoptotic Ciapin1, was identified as a molecular partner of Tah18 in the absence of stress. Moreover, Ciapin1 is able to replace yeast Dre2 in vivo and physically interacts with Tah18. Our results are in favour of an oxidative stress-induced cell death in yeast that involves mitochondria and is controlled by the newly identified Dre2-Tah18 complex.

The Drosophila (fruit fly) model system has been instrumental in our current understanding of human biology, development, and diseases. Here, we used a high-throughput yeast two-hybrid (Y2H)-based technology to screen 102 bait proteins from Drosophila melanogaster, most of them orthologous to human cancer-related and/or signaling proteins, against high-complexity fly cDNA libraries. More than 2300 protein-protein interactions (PPI) were identified, of which 710 are of high confidence. The computation of a reliability score for each protein-protein interaction and the systematic identification of the interacting domain combined with a prediction of structural/functional motifs allow the elaboration of known complexes and the identification of new ones. The full data set can be visualized using a graphical Web interface, the PIMRider (http://pim.hybrigenics.com), and is also accessible in the PSI standard Molecular Interaction data format. Our fly Protein Interaction Map (PIM) is surprisingly different from the one recently proposed by Giot et al. with little overlap between the two data sets. Analysis of the differences in data sets and methods suggests alternative strategies to enhance the accuracy and comprehensiveness of the post-genomic generation of broad-scale protein interaction maps.

The Saccharomyces cerevisiae Srs2 protein is involved in DNA repair and recombination. In order to gain better insight into the roles of Srs2, we performed a screen to identify mutations that are synthetically lethal with an srs2 deletion. One of them is a mutated allele of the ULP1 gene that encodes a protease specifically cleaving Smt3-protein conjugates. This allele, ulp1-I615N, is responsible for an accumulation of Smt3-conjugated proteins. The mutant is unable to grow at 37 degrees C. At permissive temperatures, it still shows severe growth defects together with a strong hyperrecombination phenotype and is impaired in meiosis. Genetic interactions between ulp1 and mutations that affect different repair pathways indicated that the RAD51-dependent homologous recombination mechanism, but not excision resynthesis, translesion synthesis, or nonhomologous end-joining processes, is required for the viability of the mutant. Thus, both Srs2, believed to negatively control homologous recombination, and the process of recombination per se are essential for the viability of the ulp1 mutant. Upon replication, mutant cells accumulate single-stranded DNA interruptions. These structures are believed to generate different recombination intermediates. Some of them are fixed by recombination, and others require Srs2 to be reversed and fixed by an alternate pathway.

The pol3-13 mutation is located in the C-terminal end of POL3, the gene encoding the catalytic subunit of polymerase delta, and confers thermosensitivity onto the Saccharomyces cerevisiae mutant strain. To get insight about DNA replication control, we performed a genetic screen to identify genes that are synthetic lethal with pol3-13. Mutations in genes encoding the two other subunits of DNA polymerase delta (HYS2, POL32) were identified. Mutations in two recombination genes (RAD50, RAD51) were also identified, confirming that homologous recombination is necessary for pol3-13 mutant strain survival. Other mutations were identified in genes involved in repair and genome stability (MET18/ MMS19), in the control of origin-firing and/or transcription (ABF1, SRB7), in the S/G2 checkpoint (RAD53), in the Ras-cAMP signal transduction pathway (MKS1), in nuclear pore metabolism (SEH1), in protein degradation (DOC1) and in folding (YDJ1). Finally, mutations in three genes of unknown function were isolated (NBP35, DRE2, TAH18). Synthetic lethality between pol3-13 and each of the three mutants pol32, mms19 and doc1 could be suppressed by a rad18 deletion, suggesting an important role of ubiquitination in DNA replication control. We propose that the pol3-13 mutant generates replicative problems that need both homologous recombination and an intact checkpoint machinery to be overcome.

The POL3 encoded catalytic subunit of DNA polymerase delta possesses a highly conserved C-terminal cysteine-rich domain in Saccharomyces cerevisiae. Mutations in some of its cysteine codons display a lethal phenotype, which demonstrates an essential function of this domain. The thermosensitive mutant pol3-13, in which a serine replaces a cysteine of this domain, exhibits a range of defects in DNA repair, such as hypersensitivity to different DNA-damaging agents and deficiency for induced mutagenesis and for recombination. These phenotypes are observed at 24 degrees, a temperature at which DNA replication is almost normal; this differentiates the functions of POL3 in DNA repair and DNA replication. Since spontaneous mutagenesis and spontaneous recombination are efficient in pol3-13, we propose that POL3 plays an important role in DNA repair after irradiation, particularly in the error-prone and recombinational pathways. Extragenic suppressors of pol3-13 are allelic to sdp5-1, previously identified as an extragenic suppressor of pol3-11. SDP5, which is identical to HYS2, encodes a protein homologous to the p50 subunit of bovine and human DNA polymerase delta. SDP5 is most probably the p55 subunit of Pol delta of S. cerevisiae and seems to be associated with the catalytic subunit for both DNA replication and DNA repair.

Suppressors of the methyl methanesulfonate sensitivity of Saccharomyces cerevisiae diploids lacking the Srs2 helicase turned out to contain semidominant mutations in Rad5l, a homolog of the bacterial RecA protein. The nature of these mutations was determined by direct sequencing. The 26 mutations characterized were single base substitutions leading to amino acid replacements at 18 different sites. The great majority of these sites (75%) are conserved in the family of RecA-like proteins, and 10 of them affect sites corresponding to amino acids in RecA that are probably directly involved in ATP reactions, binding, and/or hydrolysis. Six mutations are in domains thought to be involved in interaction between monomers; they may also affect ATP reactions. By themselves, all the alleles confer a rad5l null phenotype. When heterozygous, however, they are, to varying degrees, negative semidominant for radiation sensitivity; presumably the mutant proteins are coassembled with wild-type Rad51 and poison the resulting nucleofilaments or recombination complexes. This negative effect is partially suppressed by an SRS2 deletion, which supports the hypothesis that Srs2 reverses recombination structures that contain either mutated proteins or numerous DNA lesions.

The expression of the SRS2 gene, which encodes a DNA helicase involved in DNA repair in Saccharomyces cerevisiae, was studied using an SRS2-lacZ fusion integrated at the chromosomal SRS2 locus. It is shown here that this gene is expressed at a low level and is tightly regulated. It is cell-cycle regulated, with induction probably being coordinated with that of the DNA-synthesis genes, which are transcribed at the G1-S boundary. It is also induced by DNA-damaging agents, but only during the G2 phase of the cell cycle; this distinguishes it from a number of other repair genes, which are inducible throughout the cycle. During meiosis, the expression of SRS2 rises at a time nearly coincident with commitment to recombination. Since srs2 null mutants are radiation sensitive essentially when treated in G1, the mitotic regulation pattern described here leads us to postulate that either secondary regulatory events limit Srs2 activity of G1 cells or Srs2 functions in a repair mechanism associated with replication.

A 3.6 kb DNA fragment from Saccharomyces douglasii, containing the ARG4 gene, has been cloned, sequenced and compared to the corresponding region from Saccharomyces cerevisiae. The organization of this region is identical in both yeasts. It contains besides the ARG4 gene, another complete open reading frame (ORF) (YSD83) and a third incomplete one (DED81). The ARG4 and the YSD83 coding regions differ from their S. cerevisiae homologs by 8.1% and 12.5%, respectively, of base substitutions. The encoded proteins have evolved differently: amino acid replacements are significantly less frequent in Arg4 (2.8%) than in Ysc83 (12.4%) and most of the changes in Arg4 are conservative, which is not the case for Ysc83. The non-coding regions are less conserved, with small AT-rich insertions/deletions and 20% base substitutions. However, the level of divergence is smaller in the aligned sequences of these regions than in silent sites of the ORFs, probably revealing a higher degree of constraints. The Gcn4 binding site and the region where meiotic double-strand breaks occur, are fully conserved. The data confirm that these two yeasts are evolutionarily closely related and that comparisons of their sequences might reveal conserved protein and DNA domains not expected to be found in sequence comparisons between more diverged organisms.

Transposition of Saccharomyces cerevisiae Ty1 retroelements has been shown to involve reverse transcription in intracytoplasmic virus-like particles (Ty-VLPs). Ty DNA present in the particles specified by Ty1-H3 element was found to consist of the full-length genomic DNA as well as incomplete cDNAs mainly of plus polarity. Our results indicate that identical sequences (TGGGTGGTA) are used as primers for the synthesis of plus strand cDNA, generating cDNAs of 0.345 kb (analogous to the retroviral strong-stop plus cDNA) and of 2.1 kb. Electron microscopic analyses of Ty1-VLP DNA revealed two distinct classes, one full-length and the other corresponding to 0.34 kbp molecules, the size of a LTR sequence. The full-length molecules are either completely double-stranded or only partially double- stranded at one end or at both ends. These double-stranded regions are of a length corresponding to those of incomplete plus strands detected by biochemical techniques. Double-stranded circular molecules mainly of a length corresponding to that of two-LTR circles were also detected on electron micrographs. These analyses allowed us to propose a scheme for reverse transcription in Ty particles.

Eleven suppressors of the radiation sensitivity of Saccharomyces cerevisiae diploids lacking the Srs2 helicase were analyzed and found to contain codominant mutations in the RAD51 gene known to be involved in recombinational repair and in genetic recombination. These mutant alleles confer an almost complete block in recombinational repair, as does deletion of RAD51, but heterozygous mutant alleles suppress the defects of srs2::LEU2 cells and are semidominant in Srs2+ cells. The results of this study are interpreted to mean that wild-type Rad51 protein binds to single-stranded DNA and that the semidominant mutations do not prevent this binding. The cloning and sequencing of RAD51 indicated that the gene encodes a predicted 400-amino-acid protein with a molecular mass of 43 kDa. Sequence comparisons revealed homologies to domains of Escherichia coli RecA protein predicted to be involved in DNA binding, ATP binding, and ATP hydrolysis. The expression of RAD51, measured with a RAD51-lacZ gene fusion, was found to be UV- and gamma-ray-inducible, with dose-dependent responses.

A 7965 bp DNA segment from the right arm of chromosome III of Saccharomyces cerevisiae, encompassing the sup61 and RAD18 genes, was sequenced. Four new open reading frames were found in this DNA fragment. One of them, YCR103, is 51% homologous with the G10 gene product of Xenopus laevis.

A single base pair mismatch (G:T or A:C) in the CYC1 gene of the integrative plasmid pAB218 stimulates up to a five-fold integration into the yeast chromosome. Analysis of chromosomal sites of plasmid integration suggests that the mismatch-stimulated integration is not targeted as would be expected if crossovers, localised in the region of the mismatch, were a necessary step in mismatch repair. Instead, the observed mismatch-stimulated plasmid integration could be due to potentially recombinogenic structures formed during mismatch repair, such as single-stranded gaps or denatured DNA regions extending around the plasmid molecule.

Reverse and forward mutation, induced by photoaddition of 8-methoxypsoralen (8-MOP) and 3-carbethoxypsoralen (3-CPs) or ultraviolet light (UV), are reduced in three pso mutants of Saccharomyces cerevisiae. The pso1-1 strain exhibits a lower frequency of spontaneous reversion (anti-mutator) and is almost entirely unaffected by the three agents in both the haploid and diploid states. The pso2-1 strain demonstrates very reduced frequencies of 8-MOP and 3-CPs plus 365 nm radiation-induced mutations in haploid and diploid cells. UV-induced mutation are slightly reduced, whereas survival is almost normal. The pso3-1 strain is mutable by 8-MOP and 3-CPs photoaddition only in the low-dose range. After UV treatment, survival of pso3-1 is nearly normal, whereas the frequencies of induced mutants are diminished as compared to the normal PSO+. An analogue of adenine, 6-N-hydroxyaminopurine, is capable of inducing reversions in wild type, as well as in pso and rad6-1 mutant strains, indicating that this drug may act as a direct mutagen in yeast. The comparison of photoaddition of the bifunctional agent (8-MOP) to that of the monofunctional one (3-CPs) confirms that cross-links, as well as monoadditions, are mutagenic in S. cerevisiae. Repair, of the recombinational type, taking place in diploid cells or in haploid cells in G2 phase leads to higher survival, but appears to be error-free.

The RAD18 gene of Saccharomyces cerevisiae is involved in the error-prone DNA repair. Its nucleotide sequence, as reported here, predicts an open reading frame of 1461 nt which corresponds to a protein of 487 amino acids, with an Mr of 55,237. This protein has three putative zinc fingers, two acidic regions and a nucleotide-binding domain, suggesting that it is a nucleic acid-binding protein with a possible regulatory role.

A new type of radiation-sensitive mutant of S. cerevisiae is described. The recessive radH mutation sensitizes to the lethal effect of UV radiations haploids in the G1 but not in the G2 mitotic phase. Homozygous diploids are as sensitive as G1 haploids. The UV-induced mutagenesis is depressed, while the induction of gene conversion is increased. The mutation is believed to channel the repair of lesions engaged in the mutagenic pathway into a recombination process, successful if the events involve sister-chromatids but lethal if they involve homologous chromosomes. The sequence of the RADH gene reveals that it may code for a DNA helicase, with a Mr of 134 kDa. All the consensus domains of known DNA helicases are present. Besides these consensus regions, strong homologies with the Rep and UvrD helicases of E. coli were found. The RadH putative helicase appears to belong to the set of proteins involved in the error-prone repair mechanism, at least for UV-induced lesions, and could act in coordination with the Rev3 error-prone DNA polymerase.

In Saccharomyces cerevisiae, after 8-methoxypsoralen [8-(OMe)Ps] photoaddition, more crosslinks are induced per unit dose in mitochondrial DNA than in nuclear DNA. In wild-type cells treated in the exponential phase of growth, single- and double-strand breaks are produced during crosslink removal and then are rejoined upon postexposure incubation. The incision step is almost blocked in the rad 3-2 mutant, which is also defective in excision-repair of UV-induced (254 nm) pyrimidine dimers. The cutting of crosslinks from nuclear DNA is depressed in wild-type stationary-phase cells. This is correlated with a higher sensitivity of such cells to 8-(OMe)Ps photoinduced cell killing. The incision of crosslinks is dramatically reduced in mitochondrial DNA. The rejoining of single- and double-strand breaks is not only dependent on the product of the RAD51 gene (as shown by others) but also of the PSO2 gene. A correlation was found between the ability to recombine and strand rejoining. Therefore, as in bacteria, both the excision and the recombinational repair systems are involved in crosslink repair in yeast. However, double-strand breaks in yeast constitute repair intermediates which are not detected in Escherichia coli. The LD37 (dose necessary to induce a mean of one lethal hit per cell) corresponds to about 120 crosslinks per genome in exponential-phase cells of the wild type and to 1-2 crosslinks in the pso2-1 mutant.

The RAD18 gene of Saccharomyces cerevisiae is involved in mutagenic DNA repair. We describe its isolation from a yeast library introduced into the centromeric YCp50 vector, a low copy number plasmid. The insert was subcloned into YCp50 and into the multicopy YRp7 plasmid. RAD18 is not toxic when present in multiple copies but the UV survival response indicates an heterogeneity in the cell population, a fraction of it being more sensitive. A DNA segment, close to RAD18, is toxic on the multicopy plasmid and may correspond to the tRNA sup61 known to be tightly linked to RAD18. Chromosomal deletions of RAD18 were constructed. The gene is not essential and the deleted strains have the properties of single site mutants. Thus, RAD18 appears to be essentially involved in DNA repair metabolism.