Preferential DNA repair of (6-4) photoproducts in the dihydrofolate reductase gene of Chinese hamster ovary cells

UHPLC-Q-TOF/MS detection of UV-induced TpT dimeric lesions in genomic DNA

Ultraviolet (UV) radiation induces mutagenicity and cytotoxicity in human cells by the formation of DNA lesions, including cyclobutane pyrimidine dimers (CPDs) and pyrimidine (6-4) pyrimidone photoproducts (6-4PPs), mainly on thymine-thymine (TpT) dinucleotides. Here, we firstly synthesized the two TpT dimeric lesions with satisfactory yields using a unique UV-irradiated water droplet approach followed by HPLC purification. By the use of purified TpT lesions as standards, we further developed and optimized a quantitative UHPLC-Q-TOF/MS method for the detection of CPDs and 6-4PPs. After the optimization of the enzyme composition and the pH values of hydrolysis solution, a combination of snake venom phosphodiesterase, nuclease P1, and calf intestine alkaline phosphatase can be used for one-step enzymatic digestion to efficiently release the dimeric lesions (CPDs and 6-4PPs) from the genomic DNA. By the use of the one-step digestion and UHPLC-Q-TOF/MS assay for scanning all dimeric lesions, we demonstrate that only are TpT dimeric lesions detectable in genomic DNA of HCT116 cells upon UVC irradiation. The estimated frequency of the CPD of TpT increases from 28.7 to 409 per 10⁶ bases with increasing UVC dosage from 40 J/m² to 1200 J/m², while the 6-4PP of TpT increases from 3.7 to 54 per 10⁶ bases. The proposed UHPLC-Q-TOF/MS method is promising for accurate identification and quantitative detection of UV-induced dimeric lesions in cellular DNA.

Photobiological Origins of the Field of Genomic Maintenance

Although sunlight is essential for life on earth, the ultraviolet (UV) wavelengths in its spectrum constitute a major threat to life. Various cellular responses have evolved to deal with the damage inflicted in DNA by UV, and the study of these responses in model systems has spawned the burgeoning field of DNA repair. Although we now know of many types of deleterious alterations in DNA, the approaches for studying them and the early mechanistic insights have come in large part from pioneering research on the processing of UV-induced bipyrimidine photoproducts in bacteria. It is also notable that UV was one of the first DNA damaging agents for which exposure was directly linked to cancer; the sun-sensitive syndrome, xeroderma pigmentosum, was the first example of a cancer-prone hereditary disease involving a defect in DNA repair. We provide a short history of advances in the broad field of genomic maintenance as they have emerged from research in photochemistry and photobiology.

Rescue of cells from apoptosis increases DNA repair in UVB exposed cells: implications for the DNA damage response

Classically, the nucleotide excision repair (NER) of cyclobutane pyrimidine dimers (CPD) is a lengthy process (t_1/2 > 48 h).

Photoinduced damage to cellular DNA: direct and photosensitized reactions

The survey focuses on recent aspects of photochemical reactions to cellular DNA that are implicated through the predominant formation of mostly bipyrimidine photoproducts in deleterious effects of human exposure to sunlight. Recent developments in analytical methods have allowed accurate and quantitative measurements of the main DNA photoproducts in cells and human skin. Highly mutagenic CC and CT bipyrimidine photoproducts, including cyclobutane pyrimidine dimers and pyrimidine (6-4) pyrimidone photoproducts (6-4PPs) are generated in low yields with respect to TT and TC photoproducts. Another striking finding deals with the formation of Dewar valence isomers, the third class of bipyrimidine photoproducts that is accounted for by UVA-mediated isomerization of initially UVB generated 6-4PPs. Cyclobutadithymine (T<>T) has been unambiguously shown to be involved in the genotoxicity of UVA radiation. Thus, T<>T is formed in UVA-irradiated cellular DNA according to a direct excitation mechanism with a higher efficiency than oxidatively generated DNA damage that arises mostly through the Type II photosensitization mechanism. C<>C and C<>T are repaired at rates intermediate between those of T<>T and 6-4TT. Evidence has been also provided for the occurrence of photosensitized reactions mediated by exogenous agents that act either in an independent way or through photodynamic effects.

Nucleotide excision repair proteins rapidly accumulate but fail to persist in human XP-E (DDB2 mutant) cells

The xeroderma pigmentosum (XP-E) DNA damage binding protein (DDB2) is involved in early recognition of global genome DNA damage during DNA nucleotide excision repair (NER). We found that skin fibroblasts from four newly reported XP-E patients with numerous skin cancers and DDB2 mutations had slow repair of 6-4 photoproducts (6-4PP) and markedly reduced repair of cyclobutane pyrimidine dimers (CPD). NER proteins (XPC, XPB, XPG, XPA and XPF) colocalized to CPD and 6-4PP positive regions immediately (<0.1 h) after localized UV irradiation in cells from the XP-E patients and normal controls. While these proteins persist in normal cells, surprisingly, within 0.5 h these repair proteins were no longer detectable at the sites of DNA damage in XP-E cells. Our results indicate that DDB2 is not required for the rapid recruitment of NER proteins to sites of UV photoproducts or for partial repair of 6-4PP but is essential for normal persistence of these proteins for CPD photoproduct removal.

Rapid repair of UVA-induced oxidized purines and persistence of UVB-induced dipyrimidine lesions determine the mutagenicity of sunlight in mouse cells

Despite the predominance of ultraviolet A (UVA) relative to UVB in terrestrial sunlight, solar mutagenesis in humans and rodents is characterized by mutations specific for UVB. We have investigated the kinetics of repair of UVA- and UVB-induced DNA lesions in relation to mutagenicity in transgenic mouse fibroblasts irradiated with equilethal doses of UVA and UVB in comparison to simulated-sunlight UV (SSL). We have also analyzed mutagenesis-derived carcinogenesis in sunlight-associated human skin cancers by compiling the published data on mutation types found in crucial genes in nonmelanoma and melanoma skin cancers. Here, we demonstrate a resistance to repair of UVB-induced cis-syn cyclobutane pyrimidine-dimers (CPDs) together with rapid removal of UVA-induced oxidized purines in the genome overall and in the cII transgene of SSL-irradiated cells. The spectra of mutation induced by both UVB and SSL irradiation in this experimental system are characterized by significant increases in relative frequency of C-->T transitions at dipyrimidines, which are the established signature mutation of CPDs. This type of mutation is also the predominant mutation found in human nonmelanoma and melanoma tumor samples in the TP53, CDKN2, PTCH, and protein kinase genes. The prevailing role of UVB over UVA in solar mutagenesis in our test system can be ascribed to different kinetics of repair for lesions induced by the respective UV irradiation.

Genomic heterogeneity of nucleotide excision repair

Nucleotide excision repair (NER) is one of the major cellular pathways that removes bulky DNA adducts and helix-distorting lesions. The biological consequences of defective NER in humans include UV-light-induced skin carcinogenesis and extensive neurodegeneration. Understanding the mechanism of the NER process is of great importance as the number of individuals diagnosed with skin cancer has increased considerably in recent years, particularly in the United States. Rapid progress made in the DNA repair field since the early 1980s has revealed the complexity of NER, which operates differently in different genomic regions. The genomic heterogeneity of repair seems to be governed by the functional compartmentalization of chromatin into transcriptionally active and inactive domains in the nucleus. Two sub-pathways of NER remove UV-induced photolesions: (I) Global Genome Repair (GGR) and (II) Transcription Coupled Repair (TCR). GGR is a random process that occurs slowly, while the TCR, which is tightly linked to RNA polymerase II transcription, is highly specific and efficient. The efficiency of these pathways is important in avoiding cancer and genomic instability. Studies with cell lines derived from Cockayne syndrome (CS) and Xeroderma pigmentosum (XP) group C patients, that are defective in the NER sub-pathways, have yielded valuable information regarding the genomic heterogeneity of DNA repair. This review deals with the complexity of repair heterogeneity, its mechanism and interacting molecular pathways as well as its relevance in the maintenance of genomic integrity.

Formation of the main UV-induced thymine dimeric lesions within isolated and cellular DNA as measured by high performance liquid chromatography-tandem mass spectrometry

UVB radiation-induced formation of dimeric photoproducts at bipyrimidine sites within DNA has been unambiguously associated with the lethal and mutagenic properties of sunlight. The main lesions include the cyclobutane pyrimidine dimers and the pyrimidine (6-4) pyrimidone adducts. The latter compounds have been shown in model systems to be converted into their Dewar valence isomers upon exposure to UVB light. A new direct assay, based on the use of liquid chromatography coupled to tandem mass spectrometry, is now available to simultaneously detect each of the thymine photoproducts. It was applied to the determination of the yields of formation of the thymine lesions within both isolated and cellular DNA exposed to either UVC or UVB radiation. The cis-syn cyclobutane thymine dimer was found to be the major photoproduct within cellular DNA, whereas the related (6-4) adduct was produced in an approximately 8-fold lower yield. Interestingly, the corresponding Dewar valence isomer could not be detected upon exposure of human cells to biologically relevant doses of UVB radiation.

Electrospray-mass spectrometry characterization and measurement of far-UV-induced thymine photoproducts

Far-UV-induced formation of dimeric pyrimidine photoproducts within DNA is a major cause of the carcinogenic effects of solar light. The chemical structure of this class of lesion has been mostly determined by studies on model compounds. The present work is aimed at providing mass spectrometry data on the thymine-thymine photoproducts, including the diastereoisomers of the cyclobutane dimer, the (6-4) adduct, the related Dewar valence isomer and the spore photoproduct. Fragmentation mass spectra of the modified bases, nucleosides, dinucleoside monophosphates and dinucleotides were recorded following electrospray ionization with either triple-quadrupolar or ion-trap detection. The results showed differences in fragmentation pattern between the different types of photoproducts. In addition, a drastic effect of the diastereoisometry was observed for the cyclobutane dimers. A sensitive detection technique has been developed for the analysis of dinucleoside monophosphate photoproducts by high-performance liquid chromatography associated with mass spectrometry in the negative mode with multiple reaction-monitoring detection.

DNA repair and chromatin structure in genetic diseases

Interaction of DNA repair proteins with damaged DNA in eukaryotic cells is influenced by the packaging of DNA into chromatin. The basic repeating unit of chromatin, the nucleosome, plays an important role in regulating accessibility of repair proteins to sites of damage in DNA. There are a number of different pathways fundamental to the DNA repair process. Elucidation of the proteins involved in these pathways and the mechanisms they utilize for interacting with damaged nucleosomal and nonnucleosomal DNA has been aided by studies of genetic diseases where there are defects in the DNA repair process. Two of these diseases are xeroderma pigmentosum (XP) and Fanconi anemia (FA). Cells from patients with these disorders are similar in that they have defects in the initial steps of the repair process. However, there are a number of important differences in the nature of these defects. One of these is in the ability of repair proteins from XP and FA cells to interact with damaged nucleosomal DNA. In XP complementation group A (XPA) cells, for example, endonucleases present in a chromatin-associated protein complex involved in the initial steps in the repair process are defective in their ability to incise damaged nucleosomal DNA, but, like the normal complexes, can incise damaged naked DNA. In contrast, in FA complementation group A (FA-A) cells, these complexes are equally deficient in their ability to incise damaged naked and similarly damaged nucleosomal DNA. This ability to interact with damaged nucleosomal DNA correlates with the mechanism of action these endonucleases use for locating sites of damage. Whereas the FA-A and normal endonucleases act by a processive mechanism of action, the XPA endonucleases locate sites of damage distributively. Thus the mechanism of action utilized by a DNA repair enzyme may be of critical importance in its ability to interact with damaged nucleosomal DNA.

Inhibition of cyclobutane pyrimidine dimer formation in epidermal p53 gene of UV-irradiated mice by alpha-tocopherol

Mutations or alterations in the p53 gene have been observed in 50-100% of ultraviolet light (UV)-induced squamous cell carcinoma in humans and animals. Most of the mutations occurred at dipyrimidine sequences, suggesting that pyrimidine dimers in the p53 gene play a role in the pathogenesis of cutaneous squamous cell carcinoma. We previously showed that topical alpha-tocopherol prevents UV-induced skin carcinogenesis in the mouse. In the present study we asked whether topical alpha-tocopherol reduces the level of UV-induced cyclobutane pyrimidine dimers in the murine epidermal p53 gene. Mice received six dorsal applications of 25 mg each of alpha-tocopherol, on alternate days, before exposure to 500 J/m2 of UV-B irradiation. Mice were killed at selected times after irradiation. The level of dimers in the epidermal p53 gene was measured using the T4 endonuclease V assay with quantitative Southern hybridization. Topical alpha-tocopherol caused a 55% reduction in the formation of cyclobutane pyrimidine dimers in the epidermal p53 gene. The rate of reduction of pyrimidine dimers between 1 and 10 hours after irradiation was similar in UV-irradiated mice, regardless of alpha-tocopherol treatment. Therefore, the lower level of cyclobutane pyrimidine dimers in UV-irradiated mice treated with alpha-tocopherol than in control UV-irradiated mice resulted from the prevention of formation of the dimers, and not from enhanced repair of these lesions. Our results indicate that alpha-tocopherol acts as an effective sunscreen in vivo, preventing the formation of premutagenic DNA lesions in a gene known to be important in skin carcinogenesis.

Three-dimensional visualization of ultraviolet-induced DNA damage and its repair in human cell nuclei

The two major forms of DNA damage produced by 254 nm UV light are cyclobutane pyrimidine dimer (CPD) and (6-4) photoproduct (6-4PP). Both photolesions are repaired in normal human cells by nucleotide excision repair; however, little is known about where CPD or 6-4PP are repaired in relation to the various subnuclear structures. This study aimed to produce a three-dimensional demonstration of UV-induced DNA damage and its repair in human cell nuclei. We first investigated the repair kinetics of CPD and 6-4PP using an enzyme-linked immunosorbent assay with damage-specific monoclonal antibodies in normal human and xeroderma pigmentosum complementation group C cells. We also examined the kinetics of repair DNA synthesis (unscheduled DNA synthesis) using a quantitative immunofluorescence method with anti-5-bromo-2'-deoxyuridine antibodies. We confirmed the normal repair in normal human cells and the impaired repair in xeroderma pigmentosum complementation group C cells. Then, using laser scanning confocal microscopy, we succeeded in forming a three-dimensional visualization of the nuclear localization of CPD, 6-4PP, and unscheduled DNA synthesis in individual human cells. The typical three-dimensional images of photolesions or unscheduled DNA synthesis at various repair times reflected the repair kinetics obtained by enzyme-linked immunosorbent assay or immunofluorescence very well. The important finding is that the punctate, not diffusely spread, nuclear localization of unrepaired 6-4PP was found 2 h after irradiation. Similarly, the focal nuclear localization of unscheduled DNA synthesis was observed during both the first and the second 3 h repair periods. The present results suggest that both 6-4PP and CPD are nonrandomly repaired from nuclei in normal human cells.

Far-UV-induced dimeric photoproducts in short oligonucleotides: sequence effects

Cyclobutane pyrimidine dimers and pyrimidine(6-4)pyrimidone adducts represent the two major classes of far-UV-induced DNA photoproducts. Because of the lack of appropriate detection methods for each individual photoproduct, little is known about the effect of the sequence on their formation. In the present work, the photoproduct distribution obtained upon exposure of a series of dinucleoside monophosphates to 254 nm light was determined. In the latter model compounds, the presence of a cytosine, located at either the 5'- or the 3'-side of a thymine moiety, led to the preferential formation of (6-4) adducts, whereas the cis-syn cyclobutane dimer was the main thymine-thymine photoproduct. In contrast, the yield of dimeric photoproducts, and particularly of (6-4) photoadducts, was very low upon irradiation of the cytosine-cytosine dinucleoside monophosphate. However, substitution of cytosine by uracil led to an increase in the yield of (6-4) photoproduct. It was also shown that the presence of a phosphate group at the 5'- end of a thymine-thymine dinucleoside monophosphate does not modify the photoproduct distribution. As an extension of the studies on dinucleoside monophosphates, the trinucleotide TpdCpT was used as a more relevant DNA model. The yields of formation of the thymine-cytosine and cytosine-thymine (6-4) photoproducts were in a 5:1 ratio, very close to the value obtained upon photolysis of the related dinucleoside monophosphates. The characterization of the two TpdCpT (6-4) adducts was based on 1H NMR, UV and mass spectroscopy analyses. Additional evidence for the structures was inferred from the analysis of the enzymatic digestion products of the (6-4) adducts of TpdCpT with phosphodiesterases. The latter enzymes were shown to induce the quantitative release of the photoproduct as a modified dinucleoside monophosphate in a highly sequence-specific manner.

A modified quantitative polymerase chain reaction assay for measuring gene-specific repair of UV photoproducts in human cells

Methods for measuring the induction and repair of ultraviolet (UV) induced modifications in short DNA fragments are essential for the study of gene-specific DNA repair. Measurements in genomic fragments of 14 kilobases (kb) or larger can be obtained using the enzyme-sensitive site (ESS) assay introduced by Hanawalt and Bohr (Bohr et al., 1985). More recently, several assays based on variations of the polymerase chain reaction (PCR) technique have been developed which have increased sensitivity (Govan et al., 1990; Kalinowski et al., 1992; Jennerwein and Eastman, 1991), even nucleotide resolution (Pfeifer et al., 1993). However, examination of these reports indicates that the PCR based DNA repair assays lack precision (Govan et al., 1990; Kalinowski et al., 1992; Tornaletti and Pfeifer, 1994; Jennerwein and Eastman, 1991). We report here, the development of a highly precise QPCR DNA repair assay. The assay was used to measure the induction and repair of UV photoproducts in a 2.7 kb genomic fragment containing the human growth hormone (hGH) gene in SL89 (wild-type) fibroblasts. The assay was exceedingly reproducible with an overall coefficient of variation from the mean of about 2.5%. This level of precision enabled the apparent simultaneous resolution of cyclobutane dimer (CPD) and (6-4) photoproduct (6-4PP) induction and repair.

Comparison of the rate of excision of major UV photoproducts in the strands of the human HPRT gene of normal and xeroderma pigmentosum variant cells

Xeroderma pigmentosum (XP) variant patients are genetically predisposed to sunlight-induced skin cancer. Fibroblasts from such patients are extremely sensitive to mutations induced by UV radiation, and the spectrum of mutations induced in their hypoxanthine phosphoribosyltransferase (HPRT) gene differs significantly from that seen in normal cells. To determine if this UV hypermutability reflects abnormally slow excision repair of cyclobutane pyrimidine dimers (CPD) or 6-4 pyrimidine-pyrimidones (6-4s) in that gene, we synchronized XP variant and normal fibroblasts, irradiated them in early G1-phase, 12 or more hours prior to the scheduled onset of S phase, harvested them immediately or after allowing various times for repair, and analyzed the DNA for photoproducts in the HPRT gene, using quantitative Southern blotting. To incise the DNA at CPD, we used T4 endonuclease V; to incise at 6-4s, we first used photolyase and UV365nm to reverse CPD and then UvrABC excinuclease. Excision of CPD was rapid, preferential, and strand-specific, but there was no significant difference in rate between the two kinds of cells. The half life was 4 h in the transcribed strand of the gene and 6.5 h in the nontranscribed strand. For excision of CPD in the genome overall, this value is 12 h. Excision of 6-4s from either strand of the HPRT gene was extremely rapid and preferential in both kinds of cells, with a half life of approximately 30 min. The results indicate that the UV hypermutability of the XP variant cells cannot be caused by slower rates of repair of CPD and/or 6-4s in the target gene for mutagenesis.

Induction and repair of DNA damage in UV-irradiated human lymphocytes. Spectral differences and repair kinetics

The alkaline elution assay has been employed to study the induction and repair kinetics of DNA damage in human lymphocytes after irradiation with biologically relevant doses of UVB (297 and 302 nm) or UVA (365 nm) radiation. At 365 nm, when the predominant lesions are single-strand breaks, the rate of lesion induction was 1.5 x 10(-3) per 10(8) Da per kJ m-2. The number of breaks decayed with a half-life of about 50 min after a dose of 20 kJ m-2. In the UVB region, cyclobutyl pyrimidine dimers and 6-4 photoproducts are formed, both of which are repairable via the nucleotide excision repair pathway. By using repair inhibitors, the rate of induction of such lesions at 297 and 302 nm was found to be 0.07 per 10(8) Da per J m-2. Lesions were removed with a half-life of about 100 min. Mathematical modelling of the excision repair process revealed a time-dependent polymerization-ligation rate: after an initial lag phase the polymerization-ligation rate increased, reaching 50% of its maximum rate at 80-100 min after the start of repair incubation. This course of development might be due to a damage-associated regulation of DNA precursors synthesis.

Strand bias of ultraviolet light-induced mutations in a transcriptionally active gene in human cells

Ultraviolet (UV)-induced repair and mutational spectra were analyzed in an inducible marker gene, the metallothionein-l/guamine-xanthine phosphoribosyl transferase (gpt) fusion gene, carried by an Epstein-Barr virus-derived shuttle vector episomically maintained in human cells. The repair rate of UV photodimers from the shuttle-vector molecules was typical of transcriptionally active sequences, 70% of the dimers being removed within 8 h after irradiation. The spectrum obtained under basal gene transcription was compared with that obtained under induced transcription. In both cases, base substitutions at dipyrimidine sequences predominated. Multiple mutations and deletions probably due to recombinational events induced by UV damage were also observed. Most of the UV-mutated dipyrimidine sites were located in the transcribed strand and were independent of the transcriptional activity of the target gene. In contrast, the distribution of mutations throughout the coding region of the gpt gene was affected by transcription, with a preferential clustering of mutations occurring in the 3' half of the gene after transcription induction. The strand bias observed in the UV spectra most likely reflects selection for nonfunctional gpt protein.

Repair of UV-induced pyrimidine(6-4)pyrimidone photoproducts is selectively inhibited in transcriptionally active genes after heat treatment of human fibroblasts

In normal human fibroblasts, repair of (6-4)PP in the active adenosine deaminase (ADA) gene occurs with similar rate in the transcribed and non-transcribed strand of the ADA gene, and removal of (6-4)PP from the active ADA gene is faster than from the inactive X-chromosomal 754 locus. Heat shock decreased the rate of repair of the active ADA gene down to the level of inactive genes, whereas the rate of repair of the inactive 754 locus was not affected.

Repair of 6-4 photoproducts and cyclobutane pyrimidine dimers in rad mutants of Saccharomyces cerevisiae

Repair rates of both pyrimidine-pyrimidone (6-4) photoproducts and cyclobutane pyrimidine dimers have been measured in the UV-sensitive mutants of Saccharomyces cerevisiae: rad1 to rad12 and rad14 to rad24. A dot blot immunoassay for UV photoproducts was used which measures lesions in the genome as a whole and which distinguishes 6-4 photoproducts from cyclobutane dimers. The principal findings are: (1) Wild-type yeast cells, like normal mammalian cells, repair 6-4 photoproducts more rapidly than cyclobutane dimers. (2) All mutants that are defective in repair are defective in repair of both lesions. (3) The most sensitive alleles of rad1, rad2, rad3, rad4 and rad10 show no repair of either lesion. (4) Leaky alleles of rad1, rad3 and rad14 show a very marked difference in repair rates of the two lesions, rather like the human XPA revertant cell line XP129 and the Chinese hamster mutants UV61 and V-H1. (5) No mutant repairs cyclobutane dimers more rapidly than 6-4 photoproducts.

Gene-specific DNA repair of UV-induced cyclobutane pyrimidine dimers in some cancer-prone and premature-aging human syndromes

We have examined the gene-specific DNA repair of UV-induced cyclobutane pyrimidine dimers (CPDs) in fibroblasts from the following cancer prone syndromes: familial dysplastic nevus syndrome (DNS), Gardner's syndrome (GS), and Bloom's syndrome (BS). These heritable human syndromes are associated with DNA damage hypersensitivity and have been considered as potentially DNA repair deficient. Previous determinations of DNA repair in these cell strains have been done solely at the level of the overall genome. That approach is not sensitive enough to detect deficiencies in repair at the level of the gene. Defective preferential repair of active genes may impair survival and affect genomic stability. This is exemplified by the disorder Cockayne's syndrome (CS) which is associated with a selective deficiency in the preferential repair of active genes. In this study, we have used a Cockayne's syndrome cell strain and also a normal human fibroblast cell line as a control. Repair was studied in the transcriptionally active gene dihydrofolate reductase (DHFR), the inactive delta globin gene, and in the c-myc protooncogene. In the DNS, GS and BS cell lines, we find preferential repair similar to that in normal cells. In Cockayne's syndrome cells, there is no preferential repair of the DHFR gene.

A UV-sensitive mutant of Arabidopsis defective in the repair of pyrimidine-pyrimidinone(6-4) dimers

Plants are continually subjected to ultraviolet-B (UV-B) irradiation (290 to 320 nanometers) as a component of sunlight, which induces a variety of types of damage to the plant DNA. Repair of the two major DNA photoproducts was analyzed in wild-type Arabidopsis thaliana and in a mutant derivative whose growth was sensitive to UV-B radiation. In wild-type seedlings, repair of cyclobutane pyrimidine dimers occurred more slowly in the dark than in the light; repair of this photoproduct was not affected in the mutant. Repair, in the dark, of pyrimidine-pyrimidinone(6-4) dimers was defective in the UV-sensitive mutant.

UV-induced photolesions, their repair and mutations

UV-induced cyclobutane pyrimidine dimers (CPD) are selectively removed from the transcribed strand of transcriptionally active genes in V79 Chinese hamster cells. This strand specificity of repair corresponds well with the observation that UV-induced mutations in the HPRT gene are primarily generated by DNA photolesions in the non-transcribed strand. This strand bias for mutations is, however, much more pronounced at 2 J/m2 than at the higher dose of 12 J/m2. An alternative explanation for strand specificity of mutations would be that most of the mutations are caused by pyrimidone 6-4 pyrimidine photoproducts (6-4 PP). Indeed experiments with a V79-derived cell line capable of repairing 6-4 PP but not CPD have revealed direct evidence for 6-4 PP as the mutagenic lesions in UV-irradiated hamster cells. This implies that 6-4 PP are also preferentially repaired in the transcribed strand. We have investigated the repair of DNA photolesions in the HPRT gene by measuring the distribution of bromodeoxyuridine-labeled repair patches in the transcribed and non-transcribed strands of genes employing a newly developed immunoextraction procedure. Three cell lines with different capacities to remove CPD and 6-4 PP from the HPRT gene and from the genome overall were used. We found no evidence for preferential repair of 6-4 PP in the transcribed strand of the HPRT gene in cells exposed to 10 J/m2. These data are in favor of a lack of strand-specific repair of 6-4 PP underlying the much less pronounced strand bias for induced mutations at high UV dose. However, the conclusive test would be the demonstration of preferential repair of 6-4 PP in the transcribed strand of transcriptionally active genes in cells exposed to 2 J/m2.

Nucleotide excision repair

Nucleotide excision repair is the major DNA repair mechanism in all species tested. This repair system is the sole mechanism for removing bulky adducts from DNA, but it repairs essentially all DNA lesions, and thus, in addition to its main function, it plays a back-up role for other repair systems. In both pro- and eukaryotes nucleotide excision is accomplished by a multisubunit ATP-dependent nuclease. The excision nuclease of prokaryotes incises the eighth phosphodiester bond 5' and the fourth or fifth phosphodiester bond 3' to the modified nucleotide and thus excises a 12-13-mer. The excision nuclease of eukaryotes incises the 22nd, 23rd, or 24th phosphodiester bond 5' and the fifth phosphodiester bond 3' to the lesion and thus removes the adduct in a 27-29-mer. A transcription repair coupling factor encoded by the mfd gene in Escherichia coli and the ERCC6 gene in humans directs the excision nuclease to RNA polymerase stalled at a lesion in the transcribed strand and thus ensures preferential repair of this strand compared to the nontranscribed strand.

New patterns of bulk DNA repair in ultraviolet irradiated mouse embryo carcinoma cells following differentiation

Mouse embryocarcinoma stem cells differentiate in culture, given the appropriate induction. We examined whether these cells could provide information about the regulation of nucleotide excision repair in relation to differentiation by measuring the rate-limiting incision step, the removal of cyclobutane dimers and (6-4) photoproducts from the genome as a whole and the effect of the bacteriophage T4 endonuclease (denV) gene on repair in differentiated cells. It was found that differentiation is accompanied by a marked decline in the early incision ability after UV irradiation (sixfold for P19, fourfold for PCC7 and twofold for F9), and we measured, in parallel, the loss of two common UV photoproducts [cyclobutane dimers and (6-4) photoproducts] from P19 cells. After differentiation, the excellent overall cyclobutane dimer repair capacity of proliferating cells (84% removal in 24 h) is lost (no removal in 24 h), while removal of (6-4) photoproducts, although normal at 24 h (94%), is much slower than in undifferentiated P19 at 3 h (no removal versus 64%). The presence of the denV gene greatly stimulates the repair of cyclobutane dimers in undifferentiated P19 cells (94% removal at 3 h versus 40%) and also in differentiated cells (50% removal at 24 h versus no removal). The denV gene also stimulates the early repair of (6-4) photoproducts in both differentiated and undifferentiated cells.

The Rad3 protein from Saccharomyces cerevisiae: a DNA and DNA:RNA helicase with putative RNA helicase activity

The Rad3 protein from Saccharomyces cerevisiae is a DNA helicase which participates in the repair of ultraviolet-irradiated DNA and is inhibited in the presence of DNA containing thymine dimers. This protein is also involved in mitotic recombination and spontaneous mutagenesis and is essential for cell viability in the absence of DNA damage. Furthermore, the Rad3 protein also exhibits a DNA:RNA helicase activity in which there is a significant preference for a partial DNA:RNA hybrid rather than a partial duplex DNA substrate, which suggests that this protein might be involved in DNA repair within transcriptionally active genes. Finally, the Rad3 protein contains the DEAH motif and shares high amino acid sequence similarity with the DEAD family of RNA helicase proteins, suggesting that it might also possess an RNA helicase activity.

Fine tuning of DNA repair in transcribed genes: mechanisms, prevalence and consequences

Cells fine-tune their DNA repair, selecting some regions of the genome in preference to others. In the paradigm case, excision of UV-induced pyrimidine dimers in mammalian cells, repair is concentrated in transcribed genes, especially in the transcribed strand. This is due both to chromatin structure being looser in transcribing domains, allowing more rapid repair, and to repair enzymes being coupled to RNA polymerases stalled at damage sites; possibly other factors are also involved. Some repair-defective diseases may involve repair-transcription coupling: three candidate genes have been suggested. However, preferential excision of pyrimidine dimers is not uniformly linked to transcription. In mammals it varies with species, and with cell differentiation. In Drosophila embryo cells it is absent, and in yeast, the determining factor is nucleosome stability rather than transcription. Repair of other damage departs further from the paradigm, even in some UV-mimetic lesions. No selectivity is known for repair of the very frequent minor forms of base damage. And the most interesting consequence of selective repair, selective mutagenesis, normally occurs for UV-induced, but not for spontaneous mutations. The temptation to extrapolate from mammalian UV repair should be resisted.

Repair of 6-4 photoproducts in Saccharomyces cerevisiae

We have developed a dot blot immunoassay for UV photoproducts which distinguishes 6-4 photoproducts from cyclobutane dimers. The assay uses a polyclonal antiserum that is specific for UV-irradiated DNA. Cyclobutane dimers are measured in DNA samples which have been treated with hot alkali to destroy 6-4 photoproducts. 6-4 Photoproducts are measured using blots that have been incubated in photoreactivating enzyme to eliminate cyclobutane dimers. A combination of the two treatments leaves no detectable antigenic lesions. Wild-type S. cerevisiae repairs 6-4 photoproducts, in the genome overall, more rapidly than cyclobutane dimers. The most sensitive alleles of rad1, rad2, rad3 and rad4 are completely unable to repair either kind of photoproduct. We conclude that 6-4 photoproducts are repaired by essentially the same mechanism as are cyclobutane dimers.

Quantitation of 2-amino-3-methylimidazo[4,5-f]quinoline and 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline DNA adducts in specific sequences using alkali or uvrABC excinuclease

2-Amino-3-methylimidazo[4,5-f]quinoline (IQ) and 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MelQx) are carcinogens found in cooked meats that form DNA adducts upon metabolic activation. Purified DNA from Chinese hamster ovary (CHO) cells was reacted in vitro with the active metabolites N-acetoxy-IQ or N-acetoxy-MelQx, and the adduct levels in the 5' dihydrofolate reductase (DHFR) gene and downstream region were quantitated by Southern hybridization. Adducted and restricted DNA was treated with Escherichia coli uvrABC excinuclease or alkali (0.1 N NaOH, 37 degrees C, 60 min) to incise DNA at IQ and MelQx adduct sites. The DNA was then denatured with formamide, electrophoresed on a neutral agarose gel, transferred to a support membrane, and hybridized with sequence-specific DNA probes. Both uvrABC and alkali reduced the intensity of Southern hybridization in proportion to the number of IQ or MelQx adducts in DNA, indicating that these adducts are substrates for uvrABC and that they form alkali-labile lesions in DNA. IQ and MelQx adduct levels were the same in the 5' DHFR gene and in the downstream region. Southern hybridization analysis of pBR322 containing known levels of IQ or MelQx adducts showed that the efficiency of cutting IQ or MelQx adducts by uvrABC excinuclease and alkali was approximately 30% and 15%, respectively. 32P-postlabeling studies examining adduct level in bulk DNA further showed that the adduct profiles were identical in pBR322, CHO DNA, and cultured CHO cells exposed to the reactive metabolites of IQ or MelQx. The results indicate that IQ and MelQx adducts can be quantitated in specific genomic sequences and that this method should be directly applicable to studies of gene-specific repair of these adducts in cultured cells.

Photochemistry of nucleic acids in cells

A survey of the recent aspects of the main photoreactions induced by far-UV radiation in cellular DNA is reported. This mostly includes the formation of cyclobutadipyrimidines, pyrimidine(6-4)pyrimidone photoadducts and related Dewar valence isomers in various eukaryotic and prokaryotic cells, as monitored by using either specific or more general assays. Information is also provided on mechanistic aspects regarding the formation of 5,6-dihydro-5-(alpha-thyminyl) thymine, the so-called "spore photoproduct" within far-UV-irradiated bacterial spores. The second major topic of the review deals with the effects of near-UV radiation and visible light on cellular DNA which are mostly mediated by photosensitizers. The main photoreactions of furocoumarins with DNA, one major class of photosensitizers used in the phototherapy of skin diseases, involve a [2 + 2] cycloaddition to the thymine bases according to an oxygen-independent mechanism. In contrast a second type of photosensitized reaction which appears to play a major role in the genotoxic effects of both near-UV and visible light requires the presence of oxygen. The photodynamic effects which are mediated by either still unidentified endogenous photosensitizers or defined exogenous photosensitizers lead to the formation of a wide spectrum of DNA modifications including base damage, oligonucleotide strand breaks and DNA-protein cross-links.

Excision repair influences the site and strand specificity of sunlight mutagenesis in yeast

A collection of 384 mutations recovered in a tRNA gene (SUP4-o) following exposure of isogenic excision-repair-proficient (RAD1) or deficient (rad1) strains of the yeast Saccharomyces cerevisiae to sunlight was characterized by DNA sequencing. In each case, greater than 90% of the mutations were single base-pair substitutions with events at G.C pairs constituting most of the changes. However, more than half of these substitutions were transversions in the RAD1 strain whereas transitions predominated in the rad1 strain. Tandem double substitutions were recovered in both strains and the individual changes were exclusively G.C----A.T transitions. The majority of single substitutions, and all tandem double changes, were at base-pairs where the pyrimidine(s) was part of a dipyrimidine sequence and the site specificities were consistent with cyclobutane dimers and/or pyrimidine (6-4) pyrimidone photoproducts contributing to sunlight mutagenesis. Yet, the data also pointed to an important role for lesions that form at G.C pairs and give rise to transversions. Analysis of the strand specificity of sunlight mutagenesis indicated that transitions or transversions at G.C pairs occurred preferentially in SUP4-o at sites where a dipyrimidine or a guanine, respectively, was on the transcribed strand. These biases required a functional excision-repair system.

The XPD complementation group. Insights into xeroderma pigmentosum, Cockayne's syndrome and trichothiodystrophy

The xeroderma pigmentosum complementation group D is defined by more than 30 unrelated individuals of whom less than half show major abnormalities of the central nervous system, once considered to be the hallmark of the group. Fibroblasts from the great majority of these individuals show very considerable sensitivity to UV light in vitro despite the fact that the cells carry out what appears to be substantial excision repair, as judged from repair synthesis and incision activity. This article reviews the XPD group and the defects in cellular DNA repair and examines the lack of correlation between repair and the appearance of neurological abnormalities. The article also discusses the recent awareness that at least some members of two other inherited conditions, trichothiodystrophy and Cockayne's Syndrome, carry mutations in the XPD gene.

The Dewar valence isomer of the (6-4) photoadduct of thymidylyl-(3'-5')-thymidine monophosphate: formation, alkaline lability and conformational properties

The formation of the Dewar valence isomer of the pyrimidine(6-4)pyrimidone photoadduct of thymidylyl-(3'-5')-thymidine monophosphate (TpT) was investigated under different irradiation conditions. This photoproduct was generated on exposure of TpT to far-UV radiation. However, no detectable amount of the Dewar isomer or its precursor (pyrimidine(6-4)pyrimidone photoadduct) was observed following acetone photosensitization of TpT. The Dewar valence isomer was much more unstable than the pyrimidine(6-4)pyrimidone photoproduct when treated with hot piperidine. A detailed conformational analysis of the TpT Dewar isomer photoproduct is reported as inferred from extensive one- and two-dimensional 300 and 620 MHz proton nuclear magnetic resonance (1H NMR) measurements and molecular mechanics calculations.

(6-4) photoproducts and not cyclobutane pyrimidine dimers are the main UV-induced mutagenic lesions in Chinese hamster cells

A partial revertant (RH1-26) of the UV-sensitive Chinese hamster V79 cell mutant V-H1 (complementation group 2) was isolated and characterized. It was used to analyze the mutagenic potency of the 2 major UV-induced lesions, cyclobutane pyrimidine dimers and (6-4) photoproducts. Both V-H1 and RH1-26 did not repair pyrimidine dimers measured in the genome overall as well as in the active hprt gene. Repair of (6-4) photoproducts from the genome overall was slower in V-H1 than in wild-type V79 cells, but was restored to normal in RH1-26. Although V-H1 cells have a 7-fold enhanced mutagenicity, RH1-26 cells, despite the absence of pyrimidine dimer repair, have a slightly lower level of UV-induced mutagenesis than observed in wild-type V79 cells. The molecular nature of hprt mutations and the DNA-strand specificity were similar in V79 and RH1-26 cells but different from that of V-H1 cells. Since in RH1-26 as well as in V79 cells most hprt mutations were induced by lesions in the non-transcribed DNA strand, in contrast to the transcribed DNA strand in V-H1, the observed mutation-strand bias suggests that normally (6-4) photoproducts are preferentially repaired in the transcribed DNA strand. The dramatic influence of the impaired (6-4) photoproduct repair in V-H1 on UV-induced mutability and the molecular nature of hprt mutations indicate that the (6-4) photoproduct is the main UV-induced mutagenic lesion.

Immunoanalysis of ultraviolet radiation induced DNA damage and repair within specific gene segments of plasmid DNA

The region-specific heterogeneity of repairing DNA damage has been established in several biological systems. A flexible and sensitive approach, based upon DNA damage specific antibodies, is described to monitor the repair of specific lesions within discrete genomic segments. Membrane transblotted DNA restriction fragments are immunoanalyzed for the initial formation and repair of 254 nm radiation induced pyrimidine dimers. Sensitivity of dimer immunodetection increases proportional to fragment concentration and size. Antibody binding was detectable in a 0.5 kb fragment obtained from approx. 100 ng of restriction digested phage lambda DNA irradiated with 50 J m-2. Dimers within larger fragments (greater than 5 kb) could be detected at ultraviolet doses as low as 1 to 2 J m-2. To determine the occurrence of preferential repair in prokaryotic cells, damage was assessed in DNA sequences established in various Escherichia coli strains. In vivo repair of 8.9 kb vector and 6.4 and 3.2 kb gene inserts occurred with an approximate t1/2 of 45 min in UvrABC excision repair-proficient strains. Antibody binding sites were retained by DNA within repair-deficient strains. Compared to UvrABC, the repair of DNA fragments mediated by T4 endonuclease V was rapid and complete within 30 min of cellular irradiation. The efficient repair in DenV+ strain is attributable to a highly processive repair enzyme rather than to selective repair of actively replicating target genes. The results demonstrate the exceptional ability of antibodies specific for altered biomolecular lesions to map damage and repair in gene segments episomally established within cells.

Cyclobutane dimers and (6-4) photoproducts in human cells are mended with the same patch sizes

The size of excision repair patches corresponding to excision of (6-4) pyrimidine-pyrimidone photoproducts and (5-5, 6-6) cyclobutane dimers have been independently determined by using bromodeoxyuridine substitution and density increases in isopycnic gradients of small DNA fragments. The two classes of photoproducts were distinguished by using (a) a xeroderma pigmentosum (XP) revertant cell line that excises (6-4) photoproducts normally, but does not excise cyclobutane dimers from bulk DNA or from an actively transcribed sequence; (b) an XP cell line containing the denV gene of bacteriophage T4, which repairs only cyclobutane dimers by a unique glycosylase mechanism, and (c) normal cells analyzed during time intervals in which cyclobutane dimer repair is the main repair process in action. The patch sizes for the two lesions were similar under all conditions and were estimated to be approximately 30-40 bases. These values are slightly large than corresponding estimates for Escherichia coli and Saccharomyces cerevisiae but close to estimates from in vitro experiments with human cell extracts. The size of 30 bases may consequently be very close to the actual distance between cleavage sites made on either side of a photoproduct during repair.