- Email: [email protected]
DNA Mismatch Repair Proteins: Potential Guardians Against Genomic Instability and Tumorigenesis Induced by Ultraviolet Photoproducts
REVIEW ARTICLE
DNA Mismatch Repair Proteins: Potential Guardians Against Genomic Instability and Tumorigenesis Induced by Ultraviolet Photoproducts Leah C. Young, John B. Hays,w Victor A. Tron,zy and Susan E. Andrewy
Departments of Medical Genetics and yExperimental Oncology, and zLaboratory of Medicine and Pathology, University of Alberta, Edmonton, Canada; wDepartment of Environmental and Molecular Toxicology, Oregon State University, Corvallis, Oregon, USA
In addition to their established role in repairing postreplicative DNA errors, DNA mismatch repair proteins contribute to cell cycle arrest and apoptosis in response to a wide range of exogenous DNA damage (e.g., alkylation-induced lesions). The role of DNA mismatch repair in response to ultraviolet-induced DNA damage has been historically controversial. Recent data, however,
suggest that DNA mismatch repair proteins probably do not contribute to the removal of ultraviolet-induced DNA damage, but may be important in suppressing mutagenesis, e¡ecting apoptosis, and suppressing tumorigenesis following exposure to ultraviolet radiation. Key words: apoptosis/DNA mismatch repair/mutagenesis/survival/ultraviolet. J Invest Dermatol 121:435 ^440, 2003
T
mismatches and short insertion-deletion loops in the DNA (typically one to two nucleotides). MutSb, comprising Msh2 and Msh3, recognizes larger loops (42 nucleotides). In the absence of Msh3, the MutSa heterodimer can partially compensate for the lack of MutSb. However, MutSb can only minimally compensate for the lack of MutSa, when Msh6 is absent (reviewed by Aquilina and Bignami, 2001; Marti et al, 2002). MutL also is represented in the eukaryotes by several heterodimers. Eukaryotic Mlh1 and Pms2 bind to form a MutL heterodimer (or MutLa) and, more recently, orthologs Mlh3 and Pms1 also have been identi¢ed and shown to dimerize with Mlh1 (reviewed by Marti et al, 2002). Compared to Mlh1 and Pms2, little is known about the role of Pms1 and Mlh3 in response to postreplicative or exogenous DNA damage. Unlike MutS and MutL, eukaryotic orthologs for MutH and MutU have not been identi¢ed and the processes of strand recognition in eukaryotes have not been resolved. However, MMR proteins, bind to, or are associated with, proteins that are required for excision of lesions and resynthesis of the DNA strand (e.g., ExoI and proliferating cell nuclear antigen). In addition to a critical role in correcting base mismatches, there is evidence to suggest that MMR is involved in the cellular response to DNA lesions formed by some exogenous or endogenous substances. For example, MMR is thought to contribute to recognition of some oxidation-induced lesions in a replication context and levels of mutations caused by oxidized bases are increased in MMR-defective cells (Earley and Crouse, 1998; Jackson et al, 1998; Ni et al, 1999; Mazurek et al, 2002). MMR-de¢cient cells are also resistant to the cytotoxic e¡ects of oxidizing agents (Hardman et al, 2001). Additionally, MMR proteins bind O6 -meG DNA adducts formed by some alkylating agents, speci¢cally SN1 methylating agents (e.g., N-methyl-N0 -nitro-N-nitrosoguanidine, MNNG), and loss of MMR confers strong resistance to these agents (de Wind et al, 1995; Duckett et al, 1996; Hickman and Samson, 1999; Karran, 2001). Similarly, loss of MMR confers partial resistance to cisplatin (cis diamminedichloroplatinum), and MMR proteins bind to DNA containing cisplatin adducts (Duckett et al, 1996; Mu et al, 1997; Reitmair et al, 1997; Gong
he highly conserved DNA mismatch repair (MMR) proteins contribute to DNA replication ¢delity by removing insertion/deletion loops that result from template primer slippage at repetitive DNA sequences and correcting single base mismatches that escape polymerase proofreading. If left unrepaired, such errors will give rise to permanent mutations in the genome of the affected cell. Hereditary nonpolyposis colon cancer is associated with the loss of function of one or more of the MMR proteins and underlines the importance of MMR to the maintenance of genomic stability and the prevention of cancer. The process of MMR is best characterized in Escherichia coli and is thought to occur in ¢ve steps (recently reviewed by Marti et al, 2002): (1) recognition of the DNA lesion by the MutS homodimer; (2) interaction of the mismatch-bound MutS homodimer with both ATP and the MutL homodimer, resulting in activation of the MutS-MutL complex; (3) identi¢cation of the nascent DNA by MutH endonuclease through the recognition of hemimethylation at GATC sequences; (4) removal of up to 1000 bases in the nascent DNA surrounding the lesion by processes involving MutH endonuclease, MutU (also known as UvrD and DNA helicase II), and exonucleases speci¢c for single strand DNA; and (5) resynthesis of DNA by the replicative PolIII holo-enzyme. In eukaryotes, MutS and MutL are represented by multiple orthologs (Table I). The MutS homodimer is represented in eukaryotes by Muta and MutSb heterodimers. MutSa, comprising Msh2 and Msh6, primarily recognizes single base Manuscript received March 6, 2003; revised April 8, 2003; accepted for publication June 2, 2003 Reprint requests to: Leah C. Young, Department of Medical Genetics, 8-33 MSB, University of Alberta, Edmonton, Alberta, Canada T6G 2H7; Email: [email protected] Abbreviations: BrdUrd, Bromodeoxyurdine; CPD, cyclobutane pyrimidine dimer; FdUrd, 5-£urodeoxyurdine; MEF, mouse embryonic ¢broblast; MMR, DNA mismatch repair; MNNG, N-methyl-N0 -nitro-Nnitrosoguanidine; NER, nucleotide excision repair; T[CPD]T, thymine dimer cyclobutane pyrimidine dimer; TCR, transcription coupled repair.
0022-202X/03/$15.00 . Copyright r 2003 by The Society for Investigative Dermatology, Inc. 435
436
YOUNG ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Table I. Nomenclature of MMR proteins involved in the repair of postreplicative and exogenous DNA damage Eukaryote E. coli
S. cerevisiae
Mouse
Human
MutS
Msh2p Msh3p Msh6p Mlh1p Mlh2p ^ Mlh3p Pms1p ^ ^
Msh2 Msh3 Msh6 Mlh1 Pms1 Mlh3 Pms2 ^ ^
MSH2 MSH3 MSH6 MLH1 PMS1 MLH3 PMS2 ^ ^
MutL
MutH MutU/UvrD/EvrE/ecL
Proteins involved in recombination are not addressed. ^, no ortholog identi¢ed.
et al, 1999). Such partial or complete resistance conferred by MMR de¢ciency often corresponds with a failure to apoptose in response to DNA damage (Gong et al, 1999; Hickman and Samson, 1999; Hardman et al, 2001; Meyers et al, 2001). Also, MMR has been shown to facilitate cell cycle arrest in the G2M transition in response to DNA damage (Hawn et al, 1995; Davis et al, 1998; Lan et al, 2002). Therefore, depending on the type of DNA damage, loss of MMR may result in loss of cell cycle control and/ or resistance to apoptosis, both of which promote neoplastic transformation. The ability of MMR to contribute to the postreplicative repair of single base mismatches and loops, to contribute to the removal of some endogenous and exogenous DNA lesions, as well as to in£uence cell cycle arrest and apoptosis in response to some exogenous DNA damage illustrates the versatility of MMR for the maintenance of genomic integrity. Interaction of the MMR system with ultraviolet (UV) induced DNA damage was observed as early as 20 y ago in E. coli, and since then has been directly investigated in E. coli, Saccharomyces cerevisiae, and mammals. However, likely due to the diversity of model organisms, cell lines, and UV dosages employed the role of MMR in post-UV responses has been controversial. A number of studies with mammalian cell lines have found no e¡ect of MMR de¢ciency on UV-induced toxicity or mutagenesis. Other studies, however (some unpublished), have revealed a potentially important role for MMR in response to UV-induced DNA damage. By integrating this wide body of literature, we intend to provide a more complete picture of the role of MMR in post-UV response. We will address the ability of MMR proteins to bind photoproducts, a¡ect the removal of these photoproducts, contribute to maintenance of post-UV genetic stability, and e¡ect UV-induced apoptosis. In addition, we will summarize the responses of MMR-de¢cient mice to UV radiation and data pertaining to the MMR status of human skin cancers. BINDING OF MMR PROTEINS TO PHOTOPRODUCTS
Although UVC radiation (200^290 nm) primarily induces cyclobutane pyrimidine dimers (CPDs) and [6^4]photoproducts, UVB (290^320 nm) and especially UVA (320^400 nm) produce reactive oxygen species in addition to photoproducts. Reactive oxygen species generate a wide variety of oxidized bases in DNA (e.g., thymine glycols and 8-oxoguanine) and MMR is important to the processing of oxidative DNA damage (Leadon and Avrutskaya, 1998; Hardman et al, 2001; Mazurek et al, 2002). However, mutation spectra from human skin tumors and the increased incidence of UVB-induced skin cancers in nucleotide excision repair (NER) de¢cient patients both indicate, that DNA damage generated by UVB radiation is more mutagenic than that generated by reactive oxygen species. Therefore, this review will focus
on the role of MMR in response to UV-induced DNA photoproducts. The two primary DNA adducts caused by UV radiation are cissyn CPDs and (6^4) pyrimidine-pyrimidinones (or [6^4]photoproducts). The adjacent pyrimidines can be either thymine or cytosine, with some preferences at 50 and 30 positions (e.g., thymine dimer CPDs (T[CPD]T) is more abundant than T[6^4]T, but T[6^4]C is more abundant than T[CPD]C (Ravanat et al, 2001)). CPDs are considered more mutagenic than [6^4]photoproducts. However, T[CPD]T, which is the most abundant CPD, does not give rise to the majority of permanent UV-induced mutations (Ravanat et al, 2001). UV mutation spectra are dominated by C-T transitions, with CC-TT and T-C also occurring (Ravanat et al, 2001). Regardless of whether MMR participates directly in the repair of photoproducts or acts as a general sensor of DNA damage modulating post-UV responses, MMR proteins (MutSa and/or MutSb) would be predicted to interact with UV-induced photoproducts. When studying the relative binding a⁄nities of MutS orthologs for oligoduplexes representing undamaged, mismatched, and/or photodamaged DNA, there are two relevant situations for examination: (1) an undamaged oligoduplex compared to an oligoduplex containing a single photoproduct lesion (simple lesion) and (2) an oligoduplex containing a photoproduct lesion compared to an oligoduplex with an incorrect nucleotide inserted opposite the photoproduct (compound lesion). Comparing an undamaged oligoduplex to an oligoduplex containing either a CPD or a [6^4]photoproduct, the addition of a photoproduct to an undamaged oligoduplex may decrease slightly the binding a⁄nity of hMutSa (Mu et al, 1997; Wang et al, 1999). In the compound lesion case, the insertion of a mismatch opposite a preexisting photoproduct (e.g. T[CPD]T/AA T[CPD]T/AG) is an expected result of error-prone translesion synthesis across a photoproduct (Wang et al, 1999). Mu et al (1997) tested the e¡ect of such mismatched nucleotides opposite the 50 nucleotide of a CPD and found that the addition of a 50 mismatch increased the a⁄nity of hMutSa for these photoproducts. Wang et al (1999) tested the e¡ect of mismatched nucleotides opposite the 30 nucleotide of both [6^4]photoproducts and CPDs, typically the sites of most UV-induced mutations, and found that the insertion of a mismatch increased the a⁄nity of hMutSa for the photoproducts. Relative to matched counterparts, a⁄nities of MutSa for photoproduct/base mismatches were somewhat less than for base/base mismatches, but were within the a⁄nity range apparently compatible with base/base mismatch correction (Wang et al, 1999). Moreover, recent unpublished work (P. Ho¡man and J.B. Hays) demonstrates that the binding of E. coli MutS to mismatched photoproducts is conserved, but suggests some complexities in recognition of these mismatches that may be related to the degree of distortion, in£uenced by both the type of photoproduct and the position of the mismatch. These data clearly support the concept of the MutS heterodimer binding to UV-induced DNA damage, as has been observed for other MMR-associated lesions including oxidized and alkylated bases (Duckett et al, 1996; Mazurek et al, 2002). In the cases of CPDs and [6^4]photoproducts formed by UV radiation, and 8-oxoguanine formed by oxidizing agents, MutS heterodimers show preferences for those lesions containing a mismatched base, providing an opportunity for MMR to antagonize UV- and reactive-oxygen-species-induced mutations. Therefore, MMR proteins are perfectly poised to modulate cellular responses to such DNA damage. DNA REPAIR
Bulky lesions such as UV-induced photoproducts are removed from the mammalian genome by NER, and speci¢cally from template strands of transcribed genes by transcription coupled repair (TCR) (de Gruijl et al, 2001). Although yeast MMR proteins
VOL. 121, NO. 3 SEPTEMBER 2003
have been found to physically interact with NER proteins (Bertrand et al, 1998) and hMutSa to bind to photoproductmismatch compound lesions (Mu et al, 1997; Wang et al, 1999), these interactions do not necessarily indicate a role for MMR in the process of photoproduct repair. The TCR of CPDs has been investigated in E. coli, yeast, and mammalian cells with varied conclusions. In some instances opposing results have been found using similar experimental approaches. One common repair assay used to measure repair of CPDs exploits the ability of T4 endonuclease V to cleave DNA speci¢cally at CPD. DNA samples are digested with T4 endonuclease V and levels of the transcribed strand or the nontranscribed strand of a speci¢c gene are determined by Southern blot analysis. Reappearance of the transcript is indicative of repair and TCR is characterized by a more rapid repair of the transcribed strand. Using this assay, two separate research groups have investigated TCR in MMR-de¢cient E. coli with opposing results; both groups compared MMR-de¢cient strains to wild-type E. coli and E. coli lacking the TCR protein Mfd. The ¢rst group, Mellon and Champe (1996), found that TCR is lost completely in E. coli strains lacking MutL and MutS, but not MutH, following UVC exposure. Moreover, the repair levels in these MMR-de¢cient strains were similar to those observed for a TCR-defective E. coli strain. The second research group, Li and Bockrath (1995a), found that in comparison to wild-type strains UVC-irradiated MMR-de¢cient E. coli demonstrated no decrease or only a minimal decrease (depending on parental strain) in the repair of the lactose operon, and that any decrease was never to the same degree as observed in TCR-defective E. coli. Using a di¡erent repair assay, Selby and Sancar (1995) observed that cell-free extracts from MutS-de¢cient E. coli were pro¢cient at TCR of a UVC-irradiated plasmid and that supplementation with puri¢ed MutS protein did not increase TCR e¡ectiveness. In contrast, additional repair assays performed in yeast clearly support the concept that MMR is not directly involved in the repair of UV-induced photoproducts. S. cerevisiae strains defective in Msh2p, Msh3p, Mlh1p, Pms1p, Msh3p/Msh2p, or Mlh1p/ Pms1p do not have alterations in levels of either TCR or global genomic repair of UVC-induced CPDs, compared to wild-type controls (Sweder et al, 1996; Leadon and Avrutskaya, 1998).n Using isogenic transformed mouse embryonic ¢broblasts (MEFs), immortalized kidney cells, and primary mouse dermal ¢broblasts, it has been demonstrated that there is no reduction in TCR of CPD in Msh2-null and Mlh1-null cells over wild-type control cells (Sonneveld et al, 2001; Shin et al, 2002). Other data supporting MMR-independent TCR of CPDs in mammals include (1) pro¢cient TCR of UVB-induced CPDs in the MSH6de¢cient DLD1 adenocarcinoma cell line (Therrien et al, 2001); (2) unaltered excision rates of mismatchphotoproduct compound lesions after restoration of MMR protein levels in cell extracts from the PMS2-de¢cient HEC-1-A and MSH2-de¢cient LoVo cell lines (Mu et al, 1997); (3) pro¢cient TCR of UVBand UVC-induced CPDs in the MLH1-de¢cient HCT-116 colon cancer cell line (Adimoolam et al, 2001; Rochette et al, 2002); (4) e¡ective RNA synthesis in the HCT-116 cell line following either UVB or UVC irradiation (Rochette et al, 2002); and (5) MMR-independent degradation of RNA polymerase II (McKay et al, 2001). Conversely, Mellon et al (1996) demonstrated a decrease in TCR of UVC-induced CPDs in MMR-de¢cient human cancer cell lines using the endonuclease-V-based TCR assay. Also using tumor-derived cell lines, Leadon and Avrutskaya (1997) demonstrated that MMR-de¢cient cells have reduced levels of rapid TCR following exposure to ionizing radiation or UVC radiation. In these experiments the cells were labeled with BrdUrd and FdUrd prior to irradiation in order to separate replicated and nonreplicated DNA. However, recent research has shown that MMR in£uences the cellular responses to these compounds
MMR AND UV-INDUCED DNA DAMAGE
(cytotoxicity, cell cycle arrest, removal of BrdUrd from genome) and BrdUrd sensitizes MMR-de¢cient cells to ionizing-radiationinduced cytotoxicity (Berry et al, 1999; 2000; Meyers et al, 2001). In the light of this more recent knowledge, the data of Leadon and Avrutskaya are di⁄cult to interpret.n n Surveying the data pertaining to the role of MMR in the TCR of CPDs, it is di⁄cult to make de¢nite conclusions. In E. coli and mammalian cells there is evidence both for and against the hypothesis that MMR is involved in the repair of CPDs, whereas in S. cerevisiae MMR does not appear to be necessary for the repair of CPDs. The balance of the data for both the yeast and mammalian cells, however, favors MMR-independent repair of UV-induced photoproducts. MUTAGENESIS AND GENOMIC INSTABILITY
MMR still has a potential role in counteracting UV-induced genomic instability, even if MMR does not participate in the excision and/or repair of photoproducts. Incorrect bases may be inserted opposite a photoproduct during error-prone translesion synthesis, thus creating a compound lesion. This misincorporated base, as well as the corresponding photoproduct, must be repaired to avoid generating permanent mutations. Although initial investigations into the role of MMR in UV-induced mutagenesis and toxicity began some years prior (McGraw and Marinus, 1980; Caillet-Fauquet and MaenhautMichel, 1988), it was not until 1991 that Feng, Lee, and Hays (Feng et al, 1991) were able to demonstrate direct interaction between MMR and UV-damaged DNA in E. coli. Their data suggest that UV photoproducts provoke MMR-dependent nicks through both strands of nonreplicating DNA at unmethylated GATC sequences, thus producing recombinagenic double-strand breaks; MMR-de¢cient strains have reduced levels of recombination (Feng et al, 1991; Feng and Hays, 1995). More recently, MMRde¢cient strains of S. cerevisiae and E. coli have been found to show a higher mutation frequency after UVC irradiation than wild-type controls (Li and Bockrath, 1995b; Durant et al, 1999; Liu et al, 2000). In mammalian cells, the contribution of MMR to the suppression of UV-induced mutagenesis has been less clearly de¢ned. Using spontaneously immortalized mouse kidney cells generated from Mlh1- and Pms2-null mice, Shin et al (2002) did not detect an MMR-dependent di¡erence in overall UV-induced mutation rates. However, they did observe, that following UVC irradiation the MMR-de¢cient cells, compared to wild-type controls, had increased numbers of UV-signature C-T mutations (primarily on the nontranscribed strand), increased numbers of mutations arising from two or three base pair substitutions, and greater loss of heterozygosity (Shin et al, 2002). Additionally, another study showed that loss of MMR from NER-de¢cient cells increased UVC-induced mutations several hundred-fold (Nara et al, 2001), although a separate study found no increase in levels of sister chromatid exchange (O’Driscoll et al, 1999). Moreover, observations that embryonic stem cells from Msh2-null mice are hypermutable following UVC irradiation1 and two di¡erent strains of Msh2-null mice develop UVB-induced skin tumors earlier than wild-type control mice2 (Meira et al, 2001; Yoshino et al, 2002) are consistent with the theory that mammalian MMR suppresses UV-induced mutagenesis. Thus, loss of MMR clearly in£uences UV-induced mutagenesis and genomic instability in both E. coli and S. cerevisiae. For mammalian cells there is not yet consensus on the relationship between loss of MMR and increased mutagenesis post-UV, but some data strongly suggest that loss of MMR results in increased UV-induced mutagenicity. 1
V. Borgdor¡ and N. de Wind, personal communication, 2002. Also L.C. Young et al, unpublished data.
2
Manuscript retracted, Leadon. SA, DNA Repair 2003.
437
Manuscript retracted, Cancer Research, 2003.
438
YOUNG ET AL
POST-UV CELLULAR RESPONSES
The observation that MutS orthologs bind to photoproducts, taken with the proposal that MutS orthologs in£uence UVinduced mutability without directly removing these adducts, infers that MMR acts as a sensor or mediator of UV-induced DNA damage. In addition to its role in error correction, MMR is required for apoptosis (Gong et al, 1999; Hickman and Samson, 1999; Hardman et al, 2001; Meyers et al, 2001) and G2M cell cycle arrest (Hawn et al, 1995; Davis et al, 1998; Lan et al, 2002) in response to some chemical carcinogens. Furthermore, overexpression of MMR proteins is su⁄cient to induce apoptosis (Zhang et al, 1999). The association of MMR de¢ciency with increased UV-induced mutagenesis, but not removal of photoproducts from DNA, suggests that, independent of the possibility that MMR removes incorrect nucleotides opposite photoproducts, MMR may be involved in the elimination of mutated cells by in£uencing UV-induced apoptosis and cell cycle arrest. Both E. coli MMR-de¢cient mutants (Caillet-Fauquet et al, 1984; Mellon and Champe, 1996; Qiu et al, 1998) and S. cerevisiae MMR mutants (Bertrand et al, 1998; Durant et al, 1999) have similar or increased levels of UVC-induced cell death compared to wild-type control strains. Neither bacteria nor yeast, however, are known to actively complete a cell death program to eliminate overly damaged cells, probably because they are individual entities and are not programmed to protect a complex multicellular organism. In mammalian cells, studies of cell survival after UV irradiation are contradictory. Comparing MMR-pro¢cient and -de¢cient colon cancer cell lines, Mellon et al (1996) demonstrated that the MMR-de¢cient cells were sensitized to UVC-induced cell death. Using the same cell lines plus two additional colon cancer cell lines, Leadon and Avrutskaya (1997) did not observe an MMR-dependent di¡erence in survival following UVC irradiation. Fritzell et al (1997) compared Pms2-null and wild-type MEFs immortalized with SV40 T antigen and demonstrated that the MMR-de¢cient cells were marginally more sensitive to UVC-induced cell death. Also using SV40 transformed cell lines, O’Driscoll et al (1999) found that ¢broblasts de¢cient in both NER and MMR did not have altered UVC sensitivity compared to a cell ¢broblast line that was NER de¢cient and MMR pro¢cient. Similar results were found in NER-de¢cient Chinese hamster ovary cell lines that had lost MMR through selection with the alkylating agent MNNG (Nara et al, 2001). However, Ichikawa et al (2000), found that MMR function was lost in some UVB-induced skin tumors from xeroderma pigmentoosum group A (XPA) null mice and that loss of MMR was associated with decreased UV sensitivity of the XPA-null cells. These con£icting studies have in common that they were performed using tumor-derived cell lines or transformed cells, which may have acquired additional genetic alterations. An interesting study by Reitmair et al (1997) compared Msh2-null and wild-type MEFs that had been immortalized with E7/Ras with MEFs that had undergone spontaneous immortalization. In comparison with the appropriate wild-type control, E7/Rastransformed Msh2-null MEFs were more sensitive to UVC, whereas spontaneously immortalized Msh2-null MEFs were more resistant to UVC (Reitmair et al, 1997). Moreover, di¡erent methods of immortalization have been shown to alter UV-induced cytotoxicity and mutagenesis, as well as removal of CPDs (Parris and Kraemer, 1992; Cleaver et al, 1999; Bowman et al, 2000). These data demonstrate that transformation a¡ects postUV survival and complicates the comparison of MMR-pro¢cient and -de¢cient cells. The same proteins targeted by immortalization are often key components in the pathways signaling cell cycle arrest and apoptosis. We have found that primary Msh2- and Msh6-null MEFs, as well as SV40-transformed Msh2-null MEFs, demonstrate increased survival and viability post-UVB compared to appropriate isogenic wild-type control cells (Young et al, 2003; Peters et al,
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
2003). Speci¢cally, these Msh2- and Msh6-null MEFs experience lower levels of apoptosis 48 h after UVB exposure. Biologic consequences of UVB radiation arise from both UV-induced photoproducts and oxidative DNA damage. Therefore, it is not clear which component of the UVB-induced DNA damage was processed di¡erently by these MMR-de¢cient MEFs. However, de Wind et al have shown, that ES cells derived from Msh2-null mice are tolerant to UVC radiation, which forms primarily UV photoproducts, and exhibit reduced levels of UVC-induced apoptosis compared to wild-type ES cells.3 The e¡ect of MMR de¢ciency on UV-induced cell cycle arrest has not been examined. However, Msh2-null MEFs (Peters et al, 2003) and an MSH6-de¢cient human lymphoblastoid cell line (Duckett et al, 1999) both exhibit reduced UV-induced p53 activation, compared to wild-type controls. In summary, loss of MMR has a minimal e¡ect on UV-induced cytotoxicity in transformed and tumor-derived cell lines. However, in nontransformed primary cells loss of MMR is associated with a decrease in apoptosis, suggesting that these proteins are an important component in the initiation of apoptosis following UV irradiation, thus protecting against UV-induced mutagenesis and tumorigenesis. MMR AND UV: PHYSIOLOGIC AND CLINICAL DATA
To date, there are few studies addressing MMR protein levels in human skin cancer samples. Varying levels of microsatellite instability are observed in melanoma and basal cell carcinoma (Sardi et al, 2000; Hussein et al, 2001a; 2001b; Kroiss et al, 2001; 2002; Alvino et al, 2002); however, this indicates that MMR function has probably been lost, but not that loss of MMR is involved in the etiology of the disease. Loss of MMR, however, has been found to be associated with a more aggressive basal cell carcinoma (Staibano et al, 2001) and decreased survival in patients with melanoma (Korabiowska et al, 2000). It has been demonstrated that levels of MMR proteins are increased in premalignant skin lesions over levels in normal skin, and that levels sharply decrease with transition into squamous cell carcinoma or melanoma (Hussein et al, 2001b; 2002; Liang et al, 2001). It has been proposed that the increased levels of MMR proteins observed in premalignant human skin lesions re£ect induction of MMR as a response to either continued sun exposure and/or accumulated DNA damage during the development of squamous cell carcinoma (Liang et al, 2001). Conversely, other studies have found that MMR levels are increased in the skin tumor over that of the adjacent normal tissue (Rass et al, 2000; 2001). Studies performed using cell-free extracts and cell culture suggest a role for MMR proteins in communicating the presence of UV-induced DNA damage to the cell and e¡ecting protective cellular response mechanisms (described in previous sections). The strongest data supporting the concept that MMR could protect against UV-induced tumorigenesisby in£uencing protective cellularresponses are studies demonstrating that Msh2-null mice chronically exposed to UVB have an increased propensity for developing skin cancer4 (Meira et al, 2001; Yoshino et al, 2002). CONCLUSION
MMR is necessary for the repair of postreplicative DNA errors. Additionally, MMR has been shown to be involved in the response of cells to a wide range of exogenous DNA damage. Prior to most studies speci¢cally addressing the role of MMR in post-UV responses, Siede and Eckardt-Schupp (1986) postulated that MMR may process UV-induced DNA adducts as 3 4
V. Borgdor¡ and N. de Wind, personal communication, 2002. L.C. Young et al, unpublished data.
VOL. 121, NO. 3 SEPTEMBER 2003
MMR AND UV-INDUCED DNA DAMAGE
439
Table II. Summary of the e¡ects of MMR de¢ciency on the cellular responses to UV exposure
E. coli S. cerevisiae Mammalian
Binding
Repair
Mutation rates/frequencies
Apoptosis
Yesa Not determined Yes
Possible loss of TCR No change in TCR Probably little to no loss of TCR
Increased Increased Possibly increased; Msh2-null mice are predisposed to UV-induced tumors
^ ^ Decreased
Details are discussed in the text. a John B. Hays, unpublished data. ^, not applicable.
is an Alberta Cancer Board Fellow. S.E.A. is a CIHR, AHFMR, and CGDN Scholar. J.B.H. gratefully acknowledges research support from the US National Institute of Environmental Health Sciences (NIH) grant ES09848.The authors gratefully acknowledge the contribution of unpublished data by Dr Niels de Wind (Department of Toxicogenetics, Leiden University Medical Center,The Netherlands).
REFERENCES
Figure 1. The proposed roles of MMR in the protective cellular response of primary mammalian cells to UV radiation. The MutS orthologs bind the compound lesion containing a mismatch, created by translesion synthesis past either a CPD or a [6^4]photoproduct. Based on the observations that MMR-de¢cient cells have increased mutation rates and decreased apoptosis following exposure to UV radiation, it can be proposed that MMR is required for the suppression of UV-induced mutagenesis through the removal of the acquired mismatches by canonical MMR error correction and/or the elimination of mutated cells by signaling apoptosis. It has not been determined whether MMR in£uences UV-induced cell cycle arrest.
replicative mismatches. Subsequent research has since supported the model put forth by Wang et al (1999), however, that errorprone translesion synthesis inserts an incorrect nucleotide opposite the photoproduct, thus producing a photoproductmismatch compound lesion. Human MutSa binds to these lesions and directs removal of the nucleotides by canonical MMR processing (Wang et al, 1999). More recent studies demonstrate a role for MMR proteins in cellular responses to a variety of DNA adducts, and we have shown that MMR-de¢cient cells are resistant to UVB-induced apoptosis (Peters et al, 2003; Young et al, 2003), which would eliminate mutated cells. Additionally, although studies indicate that MMR does not have a direct role in the TCR of CPD photoproducts, MMR may indirectly facilitate removal of UV-induced adducts through cell cycle arrest,5 thus suppressing mutagenesis. We propose that the Hays model can be expanded to include post-UV protective cellular responses such as cell arrest and apoptosis (Fig 1) and that an MMR-de¢cient cell is unable to respond fully to UV-induced DNA damage and is predisposed to malignant transformation. In whole-organism models, this predisposition is ampli¢ed by the mutator phenotype inherent to MMR-de¢cient cells; genes encoding proteins important to damage response contain microsatellites and are targets for mutation in MMR-de¢cient cells. The body of literature reviewed here is consistent with MMR being an important component in the cellular responses to UV and hence in the prevention of UVinduced skin cancer (Table II). This research was funded by the Canadian Institutes of Health Research (CIHR MOP-42395) and the National Cancer Institute of Canada (NCIC 11231). L.C.Y. 5
S.E. Andrew et al, unpublished data.
Adimoolam S, Lin CX, Ford JM: The p53 -regulated cyclin-dependent kinase inhibitor, p21 (cip1, waf1, sdi1), is not required for global genomic and transcription-coupled nucleotide excision repair of UV-induced DNA photoproducts. J Biol Chem 276:25813^25822, 2001 Alvino E, Marra G, Pagani E, et al: High-frequency microsatellite instability is associated with defective DNA mismatch repair in human melanoma. J Invest Dermatol 118:79^86, 2002 Aquilina G, Bignami M: Mismatch repair in correction of replication errors and processing of DNA damage. J Cell Physiol 187:145^154, 2001 Berry SE, Garces C, Hwang HS, et al: The mismatch repair protein, hMLH1, mediates 5-substituted halogenated thymidine analogue cytotoxicity, DNA incorporation, and radiosensitization in human colon cancer cells. Cancer Res 59:1840^1845, 1999 Berry SE, Davis TW, Schupp JE, Hwang HS, de Wind N, Kinsella TJ: Selective radiosensitization of drug-resistant MutS homologue-2 (MSH2) mismatch repair-de¢cient cells by halogenated thymidine (dThd) analogues: Msh2 mediates dThd analogue DNA levels and the di¡erential cytotoxicity and cell cycle e¡ects of the dThd analogues and 6 -thioguanine. Cancer Res 60:5773^5780, 2000 Bertrand P, Tishko¡ DX, Filosi N, Dasgupta R, Kolodner RD: Physical interaction between components of DNA mismatch repair and nucleotide excision repair. Proc Natl Acad Sci USA 95:14278^14283, 1998 Bowman KK, Sicard DM, Ford JM, Hanawalt PC: Reduced global genomic repair of ultraviolet light-induced cyclobutane pyrimidine dimers in simian virus 40transformed human cells. Mol Carcinog 29:17^24, 2000 Caillet-Fauquet P, Maenhaut-Michel G: Nature of the SOS mutator activity: Genetic characterization of untargeted mutagenesis in Escherichia coli. Mol Gen Genet 213:491^498, 1988 Caillet-Fauquet P, Maenhaut-Michel G, Radman M: SOS mutator e¡ect in E. coli mutants de¢cient in mismatch correction. EMBO J 3:707^712, 1984 Cleaver JE, Afzal V, Feeney L, et al: Increased ultraviolet sensitivity and chromosomal instability related to P53 function in the xeroderma pigmentosum variant. Cancer Res 59:1102^1108, 1999 Davis TW, Wilson-Van Patten C, Meyers M, et al: Defective expression of the DNA mismatch repair protein, MLH1, alters G2-M cell cycle checkpoint arrest following ionizing radiation. Cancer Res 58:767^778, 1998 Duckett DR, Drummond JT, Murchie AI, Reardon JT, Sancar A, Lilley DM, Modrich P: Human MutSa recognizes damaged DNA base pairs containing O6 -methylguanine, O4 -methylthymine, or the cisplatin-d (GpG) adduct. Proc Natl Acad Sci USA 93:6443^6447, 1996 Duckett DR, Bronstein SM, Taya Y, Modrich P: hMutSa- and hMutLa-dependent phosphorylation of p53 in response to DNA methylator damage. Proc Natl Acad Sci USA 96:12384^12388, 1999 Durant ST, Morris MM, Illand M, et al: Dependence on RAD52 and RAD1 for anticancer drug resistance mediated by inactivation of mismatch repair genes. Curr Biol 9:51^54, 1999 Earley MC, Crouse GF: The role of mismatch repair in the prevention of base pair mutations in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 95:15487^15491, 1998 Feng WY, Hays JB: DNA structures generated during recombination initiated by mismatch repair of UV-irradiated nonreplicating phage DNA in Escherichia coli: Requirements for helicase, exonucleases, and RecF and RecBCD functions. Genetics 140:1175^1186, 1995 Feng WY, Lee EH, Hays JB: Recombinagenic processing of UV-light photoproducts in nonreplicating phage DNA by the Escherichia coli methyl-directed mismatch repair system. Genetics 129:1007^1020, 1991 Fritzell JA, Narayanan L, Baker SM, et al: Role of DNA mismatch repair in the cytotoxicity of ionizing radiation. Cancer Res 57:5143^5147, 1997
440
YOUNG ET AL
Gong JG, Costanzo A, Yang H-Q, Melino G, Kaelin WG, Levrero M, Wang JYJ: The tyrosine kinase c-Abl regulates p73 in apoptotic response to cisplatin-induced DNA damage. Nature 399:806^809, 1999 de Gruijl FR, van Kranen HJ, Mullenders LH: UV-induced DNA damage, repair, mutations and oncogenic pathways in skin cancer. J Photochem Photobiol B 63:19^27, 2001 Hardman RA, Afshari CA, Barrett JC: Involvement of mammalian MLH1 in the apoptotic response to peroxide-induced oxidative stress. Cancer Res 61:1392^1397, 2001 Hawn MT, Umar A, Carethers JM, Marra G, Kunkel TA, Boland CR, Koi M: Evidence for a connection between the mismatch repair system and the G2 cell cycle checkpoint. Cancer Res 55:3721^3725, 1995 Hickman MJ, Samson LD: Role of DNA mismatch repair and p53 in signaling induction of apoptosis by alkylating agents. Proc Natl Acad Sci USA 96:10764^ 10769, 1999 Hussein MR, Roggero E, Sudilovsky EC, Tuthill RJ, Wood GS, Sudilovsky O: Alterations of mismatch repair protein expression in benign melanocytic nevi, melanocytic dysplastic nevi, and cutaneous malignant melanomas. Am J Dermatopathol 23:308^314, 2001a Hussein MR, Sun M, Tuthill RJ, et al: Comprehensive analysis of 112 melanocytic skin lesions demonstrates microsatellite instability in melanomas and dysplastic nevi, but not in benign nevi. J Cutan Pathol 28:343^350, 2001b Hussein MR, Sun M, Roggero E, Sudilovsky EC, Tuthill RJ, Wood GS, Sudilovsky O: Loss of heterozygosity, microsatellite instability, and mismatch repair protein alterations in the radial growth phase of cutaneous malignant melanomas. Mol Carcinog 34:35^44, 2002 Ichikawa M, Nakane H, Marra G, et al: Decreased UV sensitivity, mismatch repair activity and abnormal cell cycle chechpoints in skin cancer cell lines derived from UVB-irradiated XPA-de¢cient mice. Mutat Res 459:285^298, 2000 Jackson AL, Chen R, Loeb LA: Induction of microsatellite instability by oxidative DNA damage. Proc Natl Acad Sci USA 95:12468^12473, 1998 Karran P: Mechanisms of tolerance to DNA damaging therapeutic drugs. Carcinogenesis 22:1931^1937, 2001 Korabiowska M, Brinck U, Dengler H, Stachura J, Schauer A, Droese M: Analysis of the DNA mismatch repair proteins expression in malignant melanomas. Anticancer Res 20:4499^4505, 2000 Kroiss MM,Vogt TM, Schlegel J, Landthaler M, Stolz W: Microsatellite instability in malignant melanomas. Acta Derm Venereol 81:242^245, 2001 Lan Z, Sever-Chroneos Z, Strobeck MW, et al: DNA-damage invokes mismatch repair dependent cyclin D1 attenuation and RB signaling pathways to inhibit CDK2. J Biol Chem 277:8372^8381, 2002 Leadon SA, Avrutskaya AV: Di¡erential involvement of the human mismatch repair proteins, hMLH1 and hMSH2, in transcription-coupled repair. Cancer Res 57:3784^3791, 1997 Leadon SA, Avrutskaya AV: Requirement for DNA mismatch repair proteins in the transcription-coupled repair of thymine glycols in Saccharomyces cerevisiae. Mutat Res 407:177^187, 1998 Li BH, Bockrath R: Bene¢t of transcription-coupled nucleotide excision repair for gene expression in UV-damaged Escherichia coli. Mol Microbiol 18:615^622, 1995a Li BH, Bockrath R: Mutation frequency decline in Escherichia coli. I. E¡ects of defects in mismatch repair. Mol Gen Genet 249:585^590, 1995b Liang SB, Furihata M,Takeuchi T, Sonobe H, Ohtsuki Y: Reduced human mismatch repair protein expression in the development of precancerous skin lesions to squamous cell carcinoma. Virchows Arch 439:622^627, 2001 Liu H, Hewitt SR, Hays JB: Antagonism of ultraviolet-light mutagenesis by the methyl-directed mismatch-repair system of Escherichia coli. Genetics 154:503^ 512, 2000 Marti TM, Kunz C, Fleck O: DNA mismatch repair and mutation avoidance pathways. J Cell Physiol 191:28^41, 2002 Mazurek A, Berardini M, Fishel R: Activation of human MutS homologs by 8-oxoguanine DNA damage. J Biol Chem 277:8260^8266, 2002 McGraw BR, Marinus MG: Isolation and characterization of Dam þ revertants and suppressor mutations that modify secondary phenotypes of dam-3 strains of Escherichia coli K-12. Mol Gen Genet 178:309^315, 1980 McKay BC, Chen F, Clarke ST, Wiggin HE, Harley LM, Ljungman M: UV lightinduced degradation of RNA polymerase II is dependent on the Cockayne’s syndrome A and B proteins but not p53 or MLH1. Mutat Res 485:93^105, 2001 Meira LB, Reis AM, Cheo DL, Nahari D, Burns DK, Friedberg EC: Cancer predisposition in mutant mice defective in multiple genetic pathways: Uncovering important genetic interactions. Mutat Res 477:51^58, 2001 Mellon I, Champe GN: Products of DNA mismatch repair genes mutS and mutL are required for transcription-coupled nucleotide-excision repair of the lactose operon in Escherichia coli. Proc Natl Acad Sci USA 93:1292^1297, 1996 Mellon I, Rajpal DK, Koi M, Boland CR, Champe GN: Transcription-coupled repair de¢ciency and mutations in human mismatch repair genes. Science 272:557^560, 1996
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Meyers M, Wagner MW, Hwang HS, Kinsella TJ, Boothman DA: Role of the hMLH1 DNA mismatch repair protein in £uoropyrimidine-mediated cell death and cell cycle responses. Cancer Res 61:5193^5201, 2001 Mu D, Tursun M, Duckett DR, Drummond JT, Modrich P, Sancar A: Recognition and repair of compound DNA lesions (base damage and mismatch) by human mismatch repair and excision repair systems. Mol Cell Biol 17:760^769, 1997 Nara K, Nagashima F, Yasui A: Highly elevated ultraviolet-induced mutation frequency in isolated Chinese hamster cell lines defective in nucleotide excision repair and mismatch repair proteins. Cancer Res 61:50^52, 2001 Ni TT, Marsischky GT, Kolodner RD: MSH2 and MSH6 are required for removal of adenine misincorporated opposite 8 -oxo-guanine in S. cerevisiae. Mol Cell 4:439^444, 1999 O’Driscoll M, Martinelli S, Ciotta C, Karran P: Combined mismatch and nucleotide excision repair defects in a human cell line: Mismatch repair processes methylation but not UV- or ionizing radiation-induced DNA damage. Carcinogenesis 20:799^804, 1999 Parris CN, Kraemer KH: Ultraviolet mutagenesis in human lymphocytes: The e¡ect of cellular transformation. Exp Cell Res 201:462^469, 1992 Peters AC, Young LC, Maeda T, Tron V, Andrew SE: Mammalian DNA mismatch repair protects cells from UVB-induced DNA damage by facilitating apoptosis and p53 activation. DNA Repair (Amst) 138:1^9, 2003 Qiu J, Guan MX, Bailis AM, Shen B: Saccharomyces cerevisiae exonuclease-1 plays a role in UV resistance that is distinct from nucleotide excision repair. Nucl Acids Res 26:3077^3083, 1998 Rass K, Gutwein P, Muller SM, et al: Immunohistochemical analysis of DNA mismatch repair enzyme hMSH-2 in normal human skin and basal cell carcinomas. Histochem J 32:93^97, 2000 Rass K, Gutwein P, Welter C, Meineke V, Tilgen W, Reichrath J: DNA mismatch repair enzyme hMSH2 in malignant melanoma: Increased immunoreactivity as compared to acquired melanocytic nevi and strong mRNA expression in melanoma cell lines. Histochem J 33:459^467, 2001 Ravanat JL, Douki T, Cadet J: Direct and indirect e¡ects of UV radiation on DNA and its components. J Photochem Photobiol B 63:88^102, 2001 Reitmair AH, Risley R, Bristow RG, et al: Mutator phenotype in Msh2-de¢cient murine embryonic ¢broblasts. Cancer Res 57:3765^3771, 1997 Rochette PJ, Bastien N, McKay BC, Therrien JP, Drobetsky EA, Drouin R: Human cells bearing homozygous mutations in the DNA mismatch repair genes hMLH1 or hMSH2 are fully pro¢cient in transcription-coupled nucleotide excision repair. Oncogene 21:5743^5752, 2002 Sardi I, Piazzini M, Palleschi G, et al: Molecular detection of microsatellite instability in basal cell carcinoma. Oncol Report 7:1119^1122, 2000 Selby CP, Sancar A: Structure and function of transcription-repair coupling factor. II. Catalytic properties. J Biol Chem 270:4890^4895, 1995 Shin CY, Mellon I, Turker MS: Multiple mutations are common at mouse Aprt in genotoxin-exposed mismatch repair de¢cient cells. Oncogene 21:1768^1776, 2002 Siede W, Eckardt-Schupp F: A mismatch repair-based model can explain some features of UV mutagenesis in yeast. Mutagenesis 1:471^474, 1986 Sonneveld E, Vrieling H, Mullenders LH, van Ho¡en A: Mouse mismatch repair gene Msh2 is not essential for transcription-coupled repair of UV-induced cyclobutane pyrimidine dimers. Oncogene 20:538^541, 2001 Staibano S, Lo Muzio L, Pannone G, et al: P53 and hMSH2 expression in basal cell carcinomas and malignant melanomas from photoexposed areas of head and neck region. Int J Oncol 19:551^559, 2001 Sweder KS, Verhage RA, Crowley DJ, Crouse GF, Brouwer J, Hanawalt PC: Mismatch repair mutants in yeast are not defective in transcription-coupled DNA repair of UV-induced DNA damage. Genetics 143:1127^1135, 1996 Therrien JP, Loignon M, Drouin R, Drobetsky EA: Ablation of p21waf1cip1 expression enhances the capacity of p53 -de¢cient human tumor cells to repair UVBinduced DNA damage. Cancer Res 61:3781^3786, 2001 Wang H, Lawrence CW, Li GM, Hays JB: Speci¢c binding of human MSH2.MSH6 mismatch-repair protein heterodimers to DNA incorporating thymine- or uracil-containing UV light photoproducts opposite mismatched bases. J Biol Chem 274:16894^16900, 1999 de Wind N, Dekker M, Berns A, Radman M, te Riele H: Inactivation of the mouse Msh2 gene results in mismatch repair de¢ciency, methylation tolerance, hyperrecombination, and predisposition to cancer. Cell 82:321^330, 1995 Yoshino M, Nakatsu Y, Riele H, Hirota S, Kitamura Y, Tanaka K: Additive roles of XPA and MSH2 genes in UVB-induced skin tumorigenesis in mice. DNA Repair (Amst) 1:935^940, 2002 Young LC, Peters AC, Maeda T, Edelmann W, Kucherlapati R, Tron V, Andrew SE: DNA mismatch repair protein Msh6 is required for e¡ective UVB-induced apoptosis in primary mouse ¢broblasts. J Invest Dermatol, in press Zhang H, Richards B, Wilson T, et al: Apoptosis induced by overexpression of hMSH2 or hMLH1. Cancer Res 59:3021^3027, 1999
DNA Mismatch Repair Proteins: Potential Guardians Against Genomic Instability and Tumorigenesis Induced by Ultraviolet Photoproducts Leah C. Young, John B. Hays,w Victor A. Tron,zy and Susan E. Andrewy
Departments of Medical Genetics and yExperimental Oncology, and zLaboratory of Medicine and Pathology, University of Alberta, Edmonton, Canada; wDepartment of Environmental and Molecular Toxicology, Oregon State University, Corvallis, Oregon, USA
In addition to their established role in repairing postreplicative DNA errors, DNA mismatch repair proteins contribute to cell cycle arrest and apoptosis in response to a wide range of exogenous DNA damage (e.g., alkylation-induced lesions). The role of DNA mismatch repair in response to ultraviolet-induced DNA damage has been historically controversial. Recent data, however,
suggest that DNA mismatch repair proteins probably do not contribute to the removal of ultraviolet-induced DNA damage, but may be important in suppressing mutagenesis, e¡ecting apoptosis, and suppressing tumorigenesis following exposure to ultraviolet radiation. Key words: apoptosis/DNA mismatch repair/mutagenesis/survival/ultraviolet. J Invest Dermatol 121:435 ^440, 2003
T
mismatches and short insertion-deletion loops in the DNA (typically one to two nucleotides). MutSb, comprising Msh2 and Msh3, recognizes larger loops (42 nucleotides). In the absence of Msh3, the MutSa heterodimer can partially compensate for the lack of MutSb. However, MutSb can only minimally compensate for the lack of MutSa, when Msh6 is absent (reviewed by Aquilina and Bignami, 2001; Marti et al, 2002). MutL also is represented in the eukaryotes by several heterodimers. Eukaryotic Mlh1 and Pms2 bind to form a MutL heterodimer (or MutLa) and, more recently, orthologs Mlh3 and Pms1 also have been identi¢ed and shown to dimerize with Mlh1 (reviewed by Marti et al, 2002). Compared to Mlh1 and Pms2, little is known about the role of Pms1 and Mlh3 in response to postreplicative or exogenous DNA damage. Unlike MutS and MutL, eukaryotic orthologs for MutH and MutU have not been identi¢ed and the processes of strand recognition in eukaryotes have not been resolved. However, MMR proteins, bind to, or are associated with, proteins that are required for excision of lesions and resynthesis of the DNA strand (e.g., ExoI and proliferating cell nuclear antigen). In addition to a critical role in correcting base mismatches, there is evidence to suggest that MMR is involved in the cellular response to DNA lesions formed by some exogenous or endogenous substances. For example, MMR is thought to contribute to recognition of some oxidation-induced lesions in a replication context and levels of mutations caused by oxidized bases are increased in MMR-defective cells (Earley and Crouse, 1998; Jackson et al, 1998; Ni et al, 1999; Mazurek et al, 2002). MMR-de¢cient cells are also resistant to the cytotoxic e¡ects of oxidizing agents (Hardman et al, 2001). Additionally, MMR proteins bind O6 -meG DNA adducts formed by some alkylating agents, speci¢cally SN1 methylating agents (e.g., N-methyl-N0 -nitro-N-nitrosoguanidine, MNNG), and loss of MMR confers strong resistance to these agents (de Wind et al, 1995; Duckett et al, 1996; Hickman and Samson, 1999; Karran, 2001). Similarly, loss of MMR confers partial resistance to cisplatin (cis diamminedichloroplatinum), and MMR proteins bind to DNA containing cisplatin adducts (Duckett et al, 1996; Mu et al, 1997; Reitmair et al, 1997; Gong
he highly conserved DNA mismatch repair (MMR) proteins contribute to DNA replication ¢delity by removing insertion/deletion loops that result from template primer slippage at repetitive DNA sequences and correcting single base mismatches that escape polymerase proofreading. If left unrepaired, such errors will give rise to permanent mutations in the genome of the affected cell. Hereditary nonpolyposis colon cancer is associated with the loss of function of one or more of the MMR proteins and underlines the importance of MMR to the maintenance of genomic stability and the prevention of cancer. The process of MMR is best characterized in Escherichia coli and is thought to occur in ¢ve steps (recently reviewed by Marti et al, 2002): (1) recognition of the DNA lesion by the MutS homodimer; (2) interaction of the mismatch-bound MutS homodimer with both ATP and the MutL homodimer, resulting in activation of the MutS-MutL complex; (3) identi¢cation of the nascent DNA by MutH endonuclease through the recognition of hemimethylation at GATC sequences; (4) removal of up to 1000 bases in the nascent DNA surrounding the lesion by processes involving MutH endonuclease, MutU (also known as UvrD and DNA helicase II), and exonucleases speci¢c for single strand DNA; and (5) resynthesis of DNA by the replicative PolIII holo-enzyme. In eukaryotes, MutS and MutL are represented by multiple orthologs (Table I). The MutS homodimer is represented in eukaryotes by Muta and MutSb heterodimers. MutSa, comprising Msh2 and Msh6, primarily recognizes single base Manuscript received March 6, 2003; revised April 8, 2003; accepted for publication June 2, 2003 Reprint requests to: Leah C. Young, Department of Medical Genetics, 8-33 MSB, University of Alberta, Edmonton, Alberta, Canada T6G 2H7; Email: [email protected] Abbreviations: BrdUrd, Bromodeoxyurdine; CPD, cyclobutane pyrimidine dimer; FdUrd, 5-£urodeoxyurdine; MEF, mouse embryonic ¢broblast; MMR, DNA mismatch repair; MNNG, N-methyl-N0 -nitro-Nnitrosoguanidine; NER, nucleotide excision repair; T[CPD]T, thymine dimer cyclobutane pyrimidine dimer; TCR, transcription coupled repair.
0022-202X/03/$15.00 . Copyright r 2003 by The Society for Investigative Dermatology, Inc. 435
436
YOUNG ET AL
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Table I. Nomenclature of MMR proteins involved in the repair of postreplicative and exogenous DNA damage Eukaryote E. coli
S. cerevisiae
Mouse
Human
MutS
Msh2p Msh3p Msh6p Mlh1p Mlh2p ^ Mlh3p Pms1p ^ ^
Msh2 Msh3 Msh6 Mlh1 Pms1 Mlh3 Pms2 ^ ^
MSH2 MSH3 MSH6 MLH1 PMS1 MLH3 PMS2 ^ ^
MutL
MutH MutU/UvrD/EvrE/ecL
Proteins involved in recombination are not addressed. ^, no ortholog identi¢ed.
et al, 1999). Such partial or complete resistance conferred by MMR de¢ciency often corresponds with a failure to apoptose in response to DNA damage (Gong et al, 1999; Hickman and Samson, 1999; Hardman et al, 2001; Meyers et al, 2001). Also, MMR has been shown to facilitate cell cycle arrest in the G2M transition in response to DNA damage (Hawn et al, 1995; Davis et al, 1998; Lan et al, 2002). Therefore, depending on the type of DNA damage, loss of MMR may result in loss of cell cycle control and/ or resistance to apoptosis, both of which promote neoplastic transformation. The ability of MMR to contribute to the postreplicative repair of single base mismatches and loops, to contribute to the removal of some endogenous and exogenous DNA lesions, as well as to in£uence cell cycle arrest and apoptosis in response to some exogenous DNA damage illustrates the versatility of MMR for the maintenance of genomic integrity. Interaction of the MMR system with ultraviolet (UV) induced DNA damage was observed as early as 20 y ago in E. coli, and since then has been directly investigated in E. coli, Saccharomyces cerevisiae, and mammals. However, likely due to the diversity of model organisms, cell lines, and UV dosages employed the role of MMR in post-UV responses has been controversial. A number of studies with mammalian cell lines have found no e¡ect of MMR de¢ciency on UV-induced toxicity or mutagenesis. Other studies, however (some unpublished), have revealed a potentially important role for MMR in response to UV-induced DNA damage. By integrating this wide body of literature, we intend to provide a more complete picture of the role of MMR in post-UV response. We will address the ability of MMR proteins to bind photoproducts, a¡ect the removal of these photoproducts, contribute to maintenance of post-UV genetic stability, and e¡ect UV-induced apoptosis. In addition, we will summarize the responses of MMR-de¢cient mice to UV radiation and data pertaining to the MMR status of human skin cancers. BINDING OF MMR PROTEINS TO PHOTOPRODUCTS
Although UVC radiation (200^290 nm) primarily induces cyclobutane pyrimidine dimers (CPDs) and [6^4]photoproducts, UVB (290^320 nm) and especially UVA (320^400 nm) produce reactive oxygen species in addition to photoproducts. Reactive oxygen species generate a wide variety of oxidized bases in DNA (e.g., thymine glycols and 8-oxoguanine) and MMR is important to the processing of oxidative DNA damage (Leadon and Avrutskaya, 1998; Hardman et al, 2001; Mazurek et al, 2002). However, mutation spectra from human skin tumors and the increased incidence of UVB-induced skin cancers in nucleotide excision repair (NER) de¢cient patients both indicate, that DNA damage generated by UVB radiation is more mutagenic than that generated by reactive oxygen species. Therefore, this review will focus
on the role of MMR in response to UV-induced DNA photoproducts. The two primary DNA adducts caused by UV radiation are cissyn CPDs and (6^4) pyrimidine-pyrimidinones (or [6^4]photoproducts). The adjacent pyrimidines can be either thymine or cytosine, with some preferences at 50 and 30 positions (e.g., thymine dimer CPDs (T[CPD]T) is more abundant than T[6^4]T, but T[6^4]C is more abundant than T[CPD]C (Ravanat et al, 2001)). CPDs are considered more mutagenic than [6^4]photoproducts. However, T[CPD]T, which is the most abundant CPD, does not give rise to the majority of permanent UV-induced mutations (Ravanat et al, 2001). UV mutation spectra are dominated by C-T transitions, with CC-TT and T-C also occurring (Ravanat et al, 2001). Regardless of whether MMR participates directly in the repair of photoproducts or acts as a general sensor of DNA damage modulating post-UV responses, MMR proteins (MutSa and/or MutSb) would be predicted to interact with UV-induced photoproducts. When studying the relative binding a⁄nities of MutS orthologs for oligoduplexes representing undamaged, mismatched, and/or photodamaged DNA, there are two relevant situations for examination: (1) an undamaged oligoduplex compared to an oligoduplex containing a single photoproduct lesion (simple lesion) and (2) an oligoduplex containing a photoproduct lesion compared to an oligoduplex with an incorrect nucleotide inserted opposite the photoproduct (compound lesion). Comparing an undamaged oligoduplex to an oligoduplex containing either a CPD or a [6^4]photoproduct, the addition of a photoproduct to an undamaged oligoduplex may decrease slightly the binding a⁄nity of hMutSa (Mu et al, 1997; Wang et al, 1999). In the compound lesion case, the insertion of a mismatch opposite a preexisting photoproduct (e.g. T[CPD]T/AA T[CPD]T/AG) is an expected result of error-prone translesion synthesis across a photoproduct (Wang et al, 1999). Mu et al (1997) tested the e¡ect of such mismatched nucleotides opposite the 50 nucleotide of a CPD and found that the addition of a 50 mismatch increased the a⁄nity of hMutSa for these photoproducts. Wang et al (1999) tested the e¡ect of mismatched nucleotides opposite the 30 nucleotide of both [6^4]photoproducts and CPDs, typically the sites of most UV-induced mutations, and found that the insertion of a mismatch increased the a⁄nity of hMutSa for the photoproducts. Relative to matched counterparts, a⁄nities of MutSa for photoproduct/base mismatches were somewhat less than for base/base mismatches, but were within the a⁄nity range apparently compatible with base/base mismatch correction (Wang et al, 1999). Moreover, recent unpublished work (P. Ho¡man and J.B. Hays) demonstrates that the binding of E. coli MutS to mismatched photoproducts is conserved, but suggests some complexities in recognition of these mismatches that may be related to the degree of distortion, in£uenced by both the type of photoproduct and the position of the mismatch. These data clearly support the concept of the MutS heterodimer binding to UV-induced DNA damage, as has been observed for other MMR-associated lesions including oxidized and alkylated bases (Duckett et al, 1996; Mazurek et al, 2002). In the cases of CPDs and [6^4]photoproducts formed by UV radiation, and 8-oxoguanine formed by oxidizing agents, MutS heterodimers show preferences for those lesions containing a mismatched base, providing an opportunity for MMR to antagonize UV- and reactive-oxygen-species-induced mutations. Therefore, MMR proteins are perfectly poised to modulate cellular responses to such DNA damage. DNA REPAIR
Bulky lesions such as UV-induced photoproducts are removed from the mammalian genome by NER, and speci¢cally from template strands of transcribed genes by transcription coupled repair (TCR) (de Gruijl et al, 2001). Although yeast MMR proteins
VOL. 121, NO. 3 SEPTEMBER 2003
have been found to physically interact with NER proteins (Bertrand et al, 1998) and hMutSa to bind to photoproductmismatch compound lesions (Mu et al, 1997; Wang et al, 1999), these interactions do not necessarily indicate a role for MMR in the process of photoproduct repair. The TCR of CPDs has been investigated in E. coli, yeast, and mammalian cells with varied conclusions. In some instances opposing results have been found using similar experimental approaches. One common repair assay used to measure repair of CPDs exploits the ability of T4 endonuclease V to cleave DNA speci¢cally at CPD. DNA samples are digested with T4 endonuclease V and levels of the transcribed strand or the nontranscribed strand of a speci¢c gene are determined by Southern blot analysis. Reappearance of the transcript is indicative of repair and TCR is characterized by a more rapid repair of the transcribed strand. Using this assay, two separate research groups have investigated TCR in MMR-de¢cient E. coli with opposing results; both groups compared MMR-de¢cient strains to wild-type E. coli and E. coli lacking the TCR protein Mfd. The ¢rst group, Mellon and Champe (1996), found that TCR is lost completely in E. coli strains lacking MutL and MutS, but not MutH, following UVC exposure. Moreover, the repair levels in these MMR-de¢cient strains were similar to those observed for a TCR-defective E. coli strain. The second research group, Li and Bockrath (1995a), found that in comparison to wild-type strains UVC-irradiated MMR-de¢cient E. coli demonstrated no decrease or only a minimal decrease (depending on parental strain) in the repair of the lactose operon, and that any decrease was never to the same degree as observed in TCR-defective E. coli. Using a di¡erent repair assay, Selby and Sancar (1995) observed that cell-free extracts from MutS-de¢cient E. coli were pro¢cient at TCR of a UVC-irradiated plasmid and that supplementation with puri¢ed MutS protein did not increase TCR e¡ectiveness. In contrast, additional repair assays performed in yeast clearly support the concept that MMR is not directly involved in the repair of UV-induced photoproducts. S. cerevisiae strains defective in Msh2p, Msh3p, Mlh1p, Pms1p, Msh3p/Msh2p, or Mlh1p/ Pms1p do not have alterations in levels of either TCR or global genomic repair of UVC-induced CPDs, compared to wild-type controls (Sweder et al, 1996; Leadon and Avrutskaya, 1998).n Using isogenic transformed mouse embryonic ¢broblasts (MEFs), immortalized kidney cells, and primary mouse dermal ¢broblasts, it has been demonstrated that there is no reduction in TCR of CPD in Msh2-null and Mlh1-null cells over wild-type control cells (Sonneveld et al, 2001; Shin et al, 2002). Other data supporting MMR-independent TCR of CPDs in mammals include (1) pro¢cient TCR of UVB-induced CPDs in the MSH6de¢cient DLD1 adenocarcinoma cell line (Therrien et al, 2001); (2) unaltered excision rates of mismatchphotoproduct compound lesions after restoration of MMR protein levels in cell extracts from the PMS2-de¢cient HEC-1-A and MSH2-de¢cient LoVo cell lines (Mu et al, 1997); (3) pro¢cient TCR of UVBand UVC-induced CPDs in the MLH1-de¢cient HCT-116 colon cancer cell line (Adimoolam et al, 2001; Rochette et al, 2002); (4) e¡ective RNA synthesis in the HCT-116 cell line following either UVB or UVC irradiation (Rochette et al, 2002); and (5) MMR-independent degradation of RNA polymerase II (McKay et al, 2001). Conversely, Mellon et al (1996) demonstrated a decrease in TCR of UVC-induced CPDs in MMR-de¢cient human cancer cell lines using the endonuclease-V-based TCR assay. Also using tumor-derived cell lines, Leadon and Avrutskaya (1997) demonstrated that MMR-de¢cient cells have reduced levels of rapid TCR following exposure to ionizing radiation or UVC radiation. In these experiments the cells were labeled with BrdUrd and FdUrd prior to irradiation in order to separate replicated and nonreplicated DNA. However, recent research has shown that MMR in£uences the cellular responses to these compounds
MMR AND UV-INDUCED DNA DAMAGE
(cytotoxicity, cell cycle arrest, removal of BrdUrd from genome) and BrdUrd sensitizes MMR-de¢cient cells to ionizing-radiationinduced cytotoxicity (Berry et al, 1999; 2000; Meyers et al, 2001). In the light of this more recent knowledge, the data of Leadon and Avrutskaya are di⁄cult to interpret.n n Surveying the data pertaining to the role of MMR in the TCR of CPDs, it is di⁄cult to make de¢nite conclusions. In E. coli and mammalian cells there is evidence both for and against the hypothesis that MMR is involved in the repair of CPDs, whereas in S. cerevisiae MMR does not appear to be necessary for the repair of CPDs. The balance of the data for both the yeast and mammalian cells, however, favors MMR-independent repair of UV-induced photoproducts. MUTAGENESIS AND GENOMIC INSTABILITY
MMR still has a potential role in counteracting UV-induced genomic instability, even if MMR does not participate in the excision and/or repair of photoproducts. Incorrect bases may be inserted opposite a photoproduct during error-prone translesion synthesis, thus creating a compound lesion. This misincorporated base, as well as the corresponding photoproduct, must be repaired to avoid generating permanent mutations. Although initial investigations into the role of MMR in UV-induced mutagenesis and toxicity began some years prior (McGraw and Marinus, 1980; Caillet-Fauquet and MaenhautMichel, 1988), it was not until 1991 that Feng, Lee, and Hays (Feng et al, 1991) were able to demonstrate direct interaction between MMR and UV-damaged DNA in E. coli. Their data suggest that UV photoproducts provoke MMR-dependent nicks through both strands of nonreplicating DNA at unmethylated GATC sequences, thus producing recombinagenic double-strand breaks; MMR-de¢cient strains have reduced levels of recombination (Feng et al, 1991; Feng and Hays, 1995). More recently, MMRde¢cient strains of S. cerevisiae and E. coli have been found to show a higher mutation frequency after UVC irradiation than wild-type controls (Li and Bockrath, 1995b; Durant et al, 1999; Liu et al, 2000). In mammalian cells, the contribution of MMR to the suppression of UV-induced mutagenesis has been less clearly de¢ned. Using spontaneously immortalized mouse kidney cells generated from Mlh1- and Pms2-null mice, Shin et al (2002) did not detect an MMR-dependent di¡erence in overall UV-induced mutation rates. However, they did observe, that following UVC irradiation the MMR-de¢cient cells, compared to wild-type controls, had increased numbers of UV-signature C-T mutations (primarily on the nontranscribed strand), increased numbers of mutations arising from two or three base pair substitutions, and greater loss of heterozygosity (Shin et al, 2002). Additionally, another study showed that loss of MMR from NER-de¢cient cells increased UVC-induced mutations several hundred-fold (Nara et al, 2001), although a separate study found no increase in levels of sister chromatid exchange (O’Driscoll et al, 1999). Moreover, observations that embryonic stem cells from Msh2-null mice are hypermutable following UVC irradiation1 and two di¡erent strains of Msh2-null mice develop UVB-induced skin tumors earlier than wild-type control mice2 (Meira et al, 2001; Yoshino et al, 2002) are consistent with the theory that mammalian MMR suppresses UV-induced mutagenesis. Thus, loss of MMR clearly in£uences UV-induced mutagenesis and genomic instability in both E. coli and S. cerevisiae. For mammalian cells there is not yet consensus on the relationship between loss of MMR and increased mutagenesis post-UV, but some data strongly suggest that loss of MMR results in increased UV-induced mutagenicity. 1
V. Borgdor¡ and N. de Wind, personal communication, 2002. Also L.C. Young et al, unpublished data.
2
Manuscript retracted, Leadon. SA, DNA Repair 2003.
437
Manuscript retracted, Cancer Research, 2003.
438
YOUNG ET AL
POST-UV CELLULAR RESPONSES
The observation that MutS orthologs bind to photoproducts, taken with the proposal that MutS orthologs in£uence UVinduced mutability without directly removing these adducts, infers that MMR acts as a sensor or mediator of UV-induced DNA damage. In addition to its role in error correction, MMR is required for apoptosis (Gong et al, 1999; Hickman and Samson, 1999; Hardman et al, 2001; Meyers et al, 2001) and G2M cell cycle arrest (Hawn et al, 1995; Davis et al, 1998; Lan et al, 2002) in response to some chemical carcinogens. Furthermore, overexpression of MMR proteins is su⁄cient to induce apoptosis (Zhang et al, 1999). The association of MMR de¢ciency with increased UV-induced mutagenesis, but not removal of photoproducts from DNA, suggests that, independent of the possibility that MMR removes incorrect nucleotides opposite photoproducts, MMR may be involved in the elimination of mutated cells by in£uencing UV-induced apoptosis and cell cycle arrest. Both E. coli MMR-de¢cient mutants (Caillet-Fauquet et al, 1984; Mellon and Champe, 1996; Qiu et al, 1998) and S. cerevisiae MMR mutants (Bertrand et al, 1998; Durant et al, 1999) have similar or increased levels of UVC-induced cell death compared to wild-type control strains. Neither bacteria nor yeast, however, are known to actively complete a cell death program to eliminate overly damaged cells, probably because they are individual entities and are not programmed to protect a complex multicellular organism. In mammalian cells, studies of cell survival after UV irradiation are contradictory. Comparing MMR-pro¢cient and -de¢cient colon cancer cell lines, Mellon et al (1996) demonstrated that the MMR-de¢cient cells were sensitized to UVC-induced cell death. Using the same cell lines plus two additional colon cancer cell lines, Leadon and Avrutskaya (1997) did not observe an MMR-dependent di¡erence in survival following UVC irradiation. Fritzell et al (1997) compared Pms2-null and wild-type MEFs immortalized with SV40 T antigen and demonstrated that the MMR-de¢cient cells were marginally more sensitive to UVC-induced cell death. Also using SV40 transformed cell lines, O’Driscoll et al (1999) found that ¢broblasts de¢cient in both NER and MMR did not have altered UVC sensitivity compared to a cell ¢broblast line that was NER de¢cient and MMR pro¢cient. Similar results were found in NER-de¢cient Chinese hamster ovary cell lines that had lost MMR through selection with the alkylating agent MNNG (Nara et al, 2001). However, Ichikawa et al (2000), found that MMR function was lost in some UVB-induced skin tumors from xeroderma pigmentoosum group A (XPA) null mice and that loss of MMR was associated with decreased UV sensitivity of the XPA-null cells. These con£icting studies have in common that they were performed using tumor-derived cell lines or transformed cells, which may have acquired additional genetic alterations. An interesting study by Reitmair et al (1997) compared Msh2-null and wild-type MEFs that had been immortalized with E7/Ras with MEFs that had undergone spontaneous immortalization. In comparison with the appropriate wild-type control, E7/Rastransformed Msh2-null MEFs were more sensitive to UVC, whereas spontaneously immortalized Msh2-null MEFs were more resistant to UVC (Reitmair et al, 1997). Moreover, di¡erent methods of immortalization have been shown to alter UV-induced cytotoxicity and mutagenesis, as well as removal of CPDs (Parris and Kraemer, 1992; Cleaver et al, 1999; Bowman et al, 2000). These data demonstrate that transformation a¡ects postUV survival and complicates the comparison of MMR-pro¢cient and -de¢cient cells. The same proteins targeted by immortalization are often key components in the pathways signaling cell cycle arrest and apoptosis. We have found that primary Msh2- and Msh6-null MEFs, as well as SV40-transformed Msh2-null MEFs, demonstrate increased survival and viability post-UVB compared to appropriate isogenic wild-type control cells (Young et al, 2003; Peters et al,
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
2003). Speci¢cally, these Msh2- and Msh6-null MEFs experience lower levels of apoptosis 48 h after UVB exposure. Biologic consequences of UVB radiation arise from both UV-induced photoproducts and oxidative DNA damage. Therefore, it is not clear which component of the UVB-induced DNA damage was processed di¡erently by these MMR-de¢cient MEFs. However, de Wind et al have shown, that ES cells derived from Msh2-null mice are tolerant to UVC radiation, which forms primarily UV photoproducts, and exhibit reduced levels of UVC-induced apoptosis compared to wild-type ES cells.3 The e¡ect of MMR de¢ciency on UV-induced cell cycle arrest has not been examined. However, Msh2-null MEFs (Peters et al, 2003) and an MSH6-de¢cient human lymphoblastoid cell line (Duckett et al, 1999) both exhibit reduced UV-induced p53 activation, compared to wild-type controls. In summary, loss of MMR has a minimal e¡ect on UV-induced cytotoxicity in transformed and tumor-derived cell lines. However, in nontransformed primary cells loss of MMR is associated with a decrease in apoptosis, suggesting that these proteins are an important component in the initiation of apoptosis following UV irradiation, thus protecting against UV-induced mutagenesis and tumorigenesis. MMR AND UV: PHYSIOLOGIC AND CLINICAL DATA
To date, there are few studies addressing MMR protein levels in human skin cancer samples. Varying levels of microsatellite instability are observed in melanoma and basal cell carcinoma (Sardi et al, 2000; Hussein et al, 2001a; 2001b; Kroiss et al, 2001; 2002; Alvino et al, 2002); however, this indicates that MMR function has probably been lost, but not that loss of MMR is involved in the etiology of the disease. Loss of MMR, however, has been found to be associated with a more aggressive basal cell carcinoma (Staibano et al, 2001) and decreased survival in patients with melanoma (Korabiowska et al, 2000). It has been demonstrated that levels of MMR proteins are increased in premalignant skin lesions over levels in normal skin, and that levels sharply decrease with transition into squamous cell carcinoma or melanoma (Hussein et al, 2001b; 2002; Liang et al, 2001). It has been proposed that the increased levels of MMR proteins observed in premalignant human skin lesions re£ect induction of MMR as a response to either continued sun exposure and/or accumulated DNA damage during the development of squamous cell carcinoma (Liang et al, 2001). Conversely, other studies have found that MMR levels are increased in the skin tumor over that of the adjacent normal tissue (Rass et al, 2000; 2001). Studies performed using cell-free extracts and cell culture suggest a role for MMR proteins in communicating the presence of UV-induced DNA damage to the cell and e¡ecting protective cellular response mechanisms (described in previous sections). The strongest data supporting the concept that MMR could protect against UV-induced tumorigenesisby in£uencing protective cellularresponses are studies demonstrating that Msh2-null mice chronically exposed to UVB have an increased propensity for developing skin cancer4 (Meira et al, 2001; Yoshino et al, 2002). CONCLUSION
MMR is necessary for the repair of postreplicative DNA errors. Additionally, MMR has been shown to be involved in the response of cells to a wide range of exogenous DNA damage. Prior to most studies speci¢cally addressing the role of MMR in post-UV responses, Siede and Eckardt-Schupp (1986) postulated that MMR may process UV-induced DNA adducts as 3 4
V. Borgdor¡ and N. de Wind, personal communication, 2002. L.C. Young et al, unpublished data.
VOL. 121, NO. 3 SEPTEMBER 2003
MMR AND UV-INDUCED DNA DAMAGE
439
Table II. Summary of the e¡ects of MMR de¢ciency on the cellular responses to UV exposure
E. coli S. cerevisiae Mammalian
Binding
Repair
Mutation rates/frequencies
Apoptosis
Yesa Not determined Yes
Possible loss of TCR No change in TCR Probably little to no loss of TCR
Increased Increased Possibly increased; Msh2-null mice are predisposed to UV-induced tumors
^ ^ Decreased
Details are discussed in the text. a John B. Hays, unpublished data. ^, not applicable.
is an Alberta Cancer Board Fellow. S.E.A. is a CIHR, AHFMR, and CGDN Scholar. J.B.H. gratefully acknowledges research support from the US National Institute of Environmental Health Sciences (NIH) grant ES09848.The authors gratefully acknowledge the contribution of unpublished data by Dr Niels de Wind (Department of Toxicogenetics, Leiden University Medical Center,The Netherlands).
REFERENCES
Figure 1. The proposed roles of MMR in the protective cellular response of primary mammalian cells to UV radiation. The MutS orthologs bind the compound lesion containing a mismatch, created by translesion synthesis past either a CPD or a [6^4]photoproduct. Based on the observations that MMR-de¢cient cells have increased mutation rates and decreased apoptosis following exposure to UV radiation, it can be proposed that MMR is required for the suppression of UV-induced mutagenesis through the removal of the acquired mismatches by canonical MMR error correction and/or the elimination of mutated cells by signaling apoptosis. It has not been determined whether MMR in£uences UV-induced cell cycle arrest.
replicative mismatches. Subsequent research has since supported the model put forth by Wang et al (1999), however, that errorprone translesion synthesis inserts an incorrect nucleotide opposite the photoproduct, thus producing a photoproductmismatch compound lesion. Human MutSa binds to these lesions and directs removal of the nucleotides by canonical MMR processing (Wang et al, 1999). More recent studies demonstrate a role for MMR proteins in cellular responses to a variety of DNA adducts, and we have shown that MMR-de¢cient cells are resistant to UVB-induced apoptosis (Peters et al, 2003; Young et al, 2003), which would eliminate mutated cells. Additionally, although studies indicate that MMR does not have a direct role in the TCR of CPD photoproducts, MMR may indirectly facilitate removal of UV-induced adducts through cell cycle arrest,5 thus suppressing mutagenesis. We propose that the Hays model can be expanded to include post-UV protective cellular responses such as cell arrest and apoptosis (Fig 1) and that an MMR-de¢cient cell is unable to respond fully to UV-induced DNA damage and is predisposed to malignant transformation. In whole-organism models, this predisposition is ampli¢ed by the mutator phenotype inherent to MMR-de¢cient cells; genes encoding proteins important to damage response contain microsatellites and are targets for mutation in MMR-de¢cient cells. The body of literature reviewed here is consistent with MMR being an important component in the cellular responses to UV and hence in the prevention of UVinduced skin cancer (Table II). This research was funded by the Canadian Institutes of Health Research (CIHR MOP-42395) and the National Cancer Institute of Canada (NCIC 11231). L.C.Y. 5
S.E. Andrew et al, unpublished data.
Adimoolam S, Lin CX, Ford JM: The p53 -regulated cyclin-dependent kinase inhibitor, p21 (cip1, waf1, sdi1), is not required for global genomic and transcription-coupled nucleotide excision repair of UV-induced DNA photoproducts. J Biol Chem 276:25813^25822, 2001 Alvino E, Marra G, Pagani E, et al: High-frequency microsatellite instability is associated with defective DNA mismatch repair in human melanoma. J Invest Dermatol 118:79^86, 2002 Aquilina G, Bignami M: Mismatch repair in correction of replication errors and processing of DNA damage. J Cell Physiol 187:145^154, 2001 Berry SE, Garces C, Hwang HS, et al: The mismatch repair protein, hMLH1, mediates 5-substituted halogenated thymidine analogue cytotoxicity, DNA incorporation, and radiosensitization in human colon cancer cells. Cancer Res 59:1840^1845, 1999 Berry SE, Davis TW, Schupp JE, Hwang HS, de Wind N, Kinsella TJ: Selective radiosensitization of drug-resistant MutS homologue-2 (MSH2) mismatch repair-de¢cient cells by halogenated thymidine (dThd) analogues: Msh2 mediates dThd analogue DNA levels and the di¡erential cytotoxicity and cell cycle e¡ects of the dThd analogues and 6 -thioguanine. Cancer Res 60:5773^5780, 2000 Bertrand P, Tishko¡ DX, Filosi N, Dasgupta R, Kolodner RD: Physical interaction between components of DNA mismatch repair and nucleotide excision repair. Proc Natl Acad Sci USA 95:14278^14283, 1998 Bowman KK, Sicard DM, Ford JM, Hanawalt PC: Reduced global genomic repair of ultraviolet light-induced cyclobutane pyrimidine dimers in simian virus 40transformed human cells. Mol Carcinog 29:17^24, 2000 Caillet-Fauquet P, Maenhaut-Michel G: Nature of the SOS mutator activity: Genetic characterization of untargeted mutagenesis in Escherichia coli. Mol Gen Genet 213:491^498, 1988 Caillet-Fauquet P, Maenhaut-Michel G, Radman M: SOS mutator e¡ect in E. coli mutants de¢cient in mismatch correction. EMBO J 3:707^712, 1984 Cleaver JE, Afzal V, Feeney L, et al: Increased ultraviolet sensitivity and chromosomal instability related to P53 function in the xeroderma pigmentosum variant. Cancer Res 59:1102^1108, 1999 Davis TW, Wilson-Van Patten C, Meyers M, et al: Defective expression of the DNA mismatch repair protein, MLH1, alters G2-M cell cycle checkpoint arrest following ionizing radiation. Cancer Res 58:767^778, 1998 Duckett DR, Drummond JT, Murchie AI, Reardon JT, Sancar A, Lilley DM, Modrich P: Human MutSa recognizes damaged DNA base pairs containing O6 -methylguanine, O4 -methylthymine, or the cisplatin-d (GpG) adduct. Proc Natl Acad Sci USA 93:6443^6447, 1996 Duckett DR, Bronstein SM, Taya Y, Modrich P: hMutSa- and hMutLa-dependent phosphorylation of p53 in response to DNA methylator damage. Proc Natl Acad Sci USA 96:12384^12388, 1999 Durant ST, Morris MM, Illand M, et al: Dependence on RAD52 and RAD1 for anticancer drug resistance mediated by inactivation of mismatch repair genes. Curr Biol 9:51^54, 1999 Earley MC, Crouse GF: The role of mismatch repair in the prevention of base pair mutations in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 95:15487^15491, 1998 Feng WY, Hays JB: DNA structures generated during recombination initiated by mismatch repair of UV-irradiated nonreplicating phage DNA in Escherichia coli: Requirements for helicase, exonucleases, and RecF and RecBCD functions. Genetics 140:1175^1186, 1995 Feng WY, Lee EH, Hays JB: Recombinagenic processing of UV-light photoproducts in nonreplicating phage DNA by the Escherichia coli methyl-directed mismatch repair system. Genetics 129:1007^1020, 1991 Fritzell JA, Narayanan L, Baker SM, et al: Role of DNA mismatch repair in the cytotoxicity of ionizing radiation. Cancer Res 57:5143^5147, 1997
440
YOUNG ET AL
Gong JG, Costanzo A, Yang H-Q, Melino G, Kaelin WG, Levrero M, Wang JYJ: The tyrosine kinase c-Abl regulates p73 in apoptotic response to cisplatin-induced DNA damage. Nature 399:806^809, 1999 de Gruijl FR, van Kranen HJ, Mullenders LH: UV-induced DNA damage, repair, mutations and oncogenic pathways in skin cancer. J Photochem Photobiol B 63:19^27, 2001 Hardman RA, Afshari CA, Barrett JC: Involvement of mammalian MLH1 in the apoptotic response to peroxide-induced oxidative stress. Cancer Res 61:1392^1397, 2001 Hawn MT, Umar A, Carethers JM, Marra G, Kunkel TA, Boland CR, Koi M: Evidence for a connection between the mismatch repair system and the G2 cell cycle checkpoint. Cancer Res 55:3721^3725, 1995 Hickman MJ, Samson LD: Role of DNA mismatch repair and p53 in signaling induction of apoptosis by alkylating agents. Proc Natl Acad Sci USA 96:10764^ 10769, 1999 Hussein MR, Roggero E, Sudilovsky EC, Tuthill RJ, Wood GS, Sudilovsky O: Alterations of mismatch repair protein expression in benign melanocytic nevi, melanocytic dysplastic nevi, and cutaneous malignant melanomas. Am J Dermatopathol 23:308^314, 2001a Hussein MR, Sun M, Tuthill RJ, et al: Comprehensive analysis of 112 melanocytic skin lesions demonstrates microsatellite instability in melanomas and dysplastic nevi, but not in benign nevi. J Cutan Pathol 28:343^350, 2001b Hussein MR, Sun M, Roggero E, Sudilovsky EC, Tuthill RJ, Wood GS, Sudilovsky O: Loss of heterozygosity, microsatellite instability, and mismatch repair protein alterations in the radial growth phase of cutaneous malignant melanomas. Mol Carcinog 34:35^44, 2002 Ichikawa M, Nakane H, Marra G, et al: Decreased UV sensitivity, mismatch repair activity and abnormal cell cycle chechpoints in skin cancer cell lines derived from UVB-irradiated XPA-de¢cient mice. Mutat Res 459:285^298, 2000 Jackson AL, Chen R, Loeb LA: Induction of microsatellite instability by oxidative DNA damage. Proc Natl Acad Sci USA 95:12468^12473, 1998 Karran P: Mechanisms of tolerance to DNA damaging therapeutic drugs. Carcinogenesis 22:1931^1937, 2001 Korabiowska M, Brinck U, Dengler H, Stachura J, Schauer A, Droese M: Analysis of the DNA mismatch repair proteins expression in malignant melanomas. Anticancer Res 20:4499^4505, 2000 Kroiss MM,Vogt TM, Schlegel J, Landthaler M, Stolz W: Microsatellite instability in malignant melanomas. Acta Derm Venereol 81:242^245, 2001 Lan Z, Sever-Chroneos Z, Strobeck MW, et al: DNA-damage invokes mismatch repair dependent cyclin D1 attenuation and RB signaling pathways to inhibit CDK2. J Biol Chem 277:8372^8381, 2002 Leadon SA, Avrutskaya AV: Di¡erential involvement of the human mismatch repair proteins, hMLH1 and hMSH2, in transcription-coupled repair. Cancer Res 57:3784^3791, 1997 Leadon SA, Avrutskaya AV: Requirement for DNA mismatch repair proteins in the transcription-coupled repair of thymine glycols in Saccharomyces cerevisiae. Mutat Res 407:177^187, 1998 Li BH, Bockrath R: Bene¢t of transcription-coupled nucleotide excision repair for gene expression in UV-damaged Escherichia coli. Mol Microbiol 18:615^622, 1995a Li BH, Bockrath R: Mutation frequency decline in Escherichia coli. I. E¡ects of defects in mismatch repair. Mol Gen Genet 249:585^590, 1995b Liang SB, Furihata M,Takeuchi T, Sonobe H, Ohtsuki Y: Reduced human mismatch repair protein expression in the development of precancerous skin lesions to squamous cell carcinoma. Virchows Arch 439:622^627, 2001 Liu H, Hewitt SR, Hays JB: Antagonism of ultraviolet-light mutagenesis by the methyl-directed mismatch-repair system of Escherichia coli. Genetics 154:503^ 512, 2000 Marti TM, Kunz C, Fleck O: DNA mismatch repair and mutation avoidance pathways. J Cell Physiol 191:28^41, 2002 Mazurek A, Berardini M, Fishel R: Activation of human MutS homologs by 8-oxoguanine DNA damage. J Biol Chem 277:8260^8266, 2002 McGraw BR, Marinus MG: Isolation and characterization of Dam þ revertants and suppressor mutations that modify secondary phenotypes of dam-3 strains of Escherichia coli K-12. Mol Gen Genet 178:309^315, 1980 McKay BC, Chen F, Clarke ST, Wiggin HE, Harley LM, Ljungman M: UV lightinduced degradation of RNA polymerase II is dependent on the Cockayne’s syndrome A and B proteins but not p53 or MLH1. Mutat Res 485:93^105, 2001 Meira LB, Reis AM, Cheo DL, Nahari D, Burns DK, Friedberg EC: Cancer predisposition in mutant mice defective in multiple genetic pathways: Uncovering important genetic interactions. Mutat Res 477:51^58, 2001 Mellon I, Champe GN: Products of DNA mismatch repair genes mutS and mutL are required for transcription-coupled nucleotide-excision repair of the lactose operon in Escherichia coli. Proc Natl Acad Sci USA 93:1292^1297, 1996 Mellon I, Rajpal DK, Koi M, Boland CR, Champe GN: Transcription-coupled repair de¢ciency and mutations in human mismatch repair genes. Science 272:557^560, 1996
THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Meyers M, Wagner MW, Hwang HS, Kinsella TJ, Boothman DA: Role of the hMLH1 DNA mismatch repair protein in £uoropyrimidine-mediated cell death and cell cycle responses. Cancer Res 61:5193^5201, 2001 Mu D, Tursun M, Duckett DR, Drummond JT, Modrich P, Sancar A: Recognition and repair of compound DNA lesions (base damage and mismatch) by human mismatch repair and excision repair systems. Mol Cell Biol 17:760^769, 1997 Nara K, Nagashima F, Yasui A: Highly elevated ultraviolet-induced mutation frequency in isolated Chinese hamster cell lines defective in nucleotide excision repair and mismatch repair proteins. Cancer Res 61:50^52, 2001 Ni TT, Marsischky GT, Kolodner RD: MSH2 and MSH6 are required for removal of adenine misincorporated opposite 8 -oxo-guanine in S. cerevisiae. Mol Cell 4:439^444, 1999 O’Driscoll M, Martinelli S, Ciotta C, Karran P: Combined mismatch and nucleotide excision repair defects in a human cell line: Mismatch repair processes methylation but not UV- or ionizing radiation-induced DNA damage. Carcinogenesis 20:799^804, 1999 Parris CN, Kraemer KH: Ultraviolet mutagenesis in human lymphocytes: The e¡ect of cellular transformation. Exp Cell Res 201:462^469, 1992 Peters AC, Young LC, Maeda T, Tron V, Andrew SE: Mammalian DNA mismatch repair protects cells from UVB-induced DNA damage by facilitating apoptosis and p53 activation. DNA Repair (Amst) 138:1^9, 2003 Qiu J, Guan MX, Bailis AM, Shen B: Saccharomyces cerevisiae exonuclease-1 plays a role in UV resistance that is distinct from nucleotide excision repair. Nucl Acids Res 26:3077^3083, 1998 Rass K, Gutwein P, Muller SM, et al: Immunohistochemical analysis of DNA mismatch repair enzyme hMSH-2 in normal human skin and basal cell carcinomas. Histochem J 32:93^97, 2000 Rass K, Gutwein P, Welter C, Meineke V, Tilgen W, Reichrath J: DNA mismatch repair enzyme hMSH2 in malignant melanoma: Increased immunoreactivity as compared to acquired melanocytic nevi and strong mRNA expression in melanoma cell lines. Histochem J 33:459^467, 2001 Ravanat JL, Douki T, Cadet J: Direct and indirect e¡ects of UV radiation on DNA and its components. J Photochem Photobiol B 63:88^102, 2001 Reitmair AH, Risley R, Bristow RG, et al: Mutator phenotype in Msh2-de¢cient murine embryonic ¢broblasts. Cancer Res 57:3765^3771, 1997 Rochette PJ, Bastien N, McKay BC, Therrien JP, Drobetsky EA, Drouin R: Human cells bearing homozygous mutations in the DNA mismatch repair genes hMLH1 or hMSH2 are fully pro¢cient in transcription-coupled nucleotide excision repair. Oncogene 21:5743^5752, 2002 Sardi I, Piazzini M, Palleschi G, et al: Molecular detection of microsatellite instability in basal cell carcinoma. Oncol Report 7:1119^1122, 2000 Selby CP, Sancar A: Structure and function of transcription-repair coupling factor. II. Catalytic properties. J Biol Chem 270:4890^4895, 1995 Shin CY, Mellon I, Turker MS: Multiple mutations are common at mouse Aprt in genotoxin-exposed mismatch repair de¢cient cells. Oncogene 21:1768^1776, 2002 Siede W, Eckardt-Schupp F: A mismatch repair-based model can explain some features of UV mutagenesis in yeast. Mutagenesis 1:471^474, 1986 Sonneveld E, Vrieling H, Mullenders LH, van Ho¡en A: Mouse mismatch repair gene Msh2 is not essential for transcription-coupled repair of UV-induced cyclobutane pyrimidine dimers. Oncogene 20:538^541, 2001 Staibano S, Lo Muzio L, Pannone G, et al: P53 and hMSH2 expression in basal cell carcinomas and malignant melanomas from photoexposed areas of head and neck region. Int J Oncol 19:551^559, 2001 Sweder KS, Verhage RA, Crowley DJ, Crouse GF, Brouwer J, Hanawalt PC: Mismatch repair mutants in yeast are not defective in transcription-coupled DNA repair of UV-induced DNA damage. Genetics 143:1127^1135, 1996 Therrien JP, Loignon M, Drouin R, Drobetsky EA: Ablation of p21waf1cip1 expression enhances the capacity of p53 -de¢cient human tumor cells to repair UVBinduced DNA damage. Cancer Res 61:3781^3786, 2001 Wang H, Lawrence CW, Li GM, Hays JB: Speci¢c binding of human MSH2.MSH6 mismatch-repair protein heterodimers to DNA incorporating thymine- or uracil-containing UV light photoproducts opposite mismatched bases. J Biol Chem 274:16894^16900, 1999 de Wind N, Dekker M, Berns A, Radman M, te Riele H: Inactivation of the mouse Msh2 gene results in mismatch repair de¢ciency, methylation tolerance, hyperrecombination, and predisposition to cancer. Cell 82:321^330, 1995 Yoshino M, Nakatsu Y, Riele H, Hirota S, Kitamura Y, Tanaka K: Additive roles of XPA and MSH2 genes in UVB-induced skin tumorigenesis in mice. DNA Repair (Amst) 1:935^940, 2002 Young LC, Peters AC, Maeda T, Edelmann W, Kucherlapati R, Tron V, Andrew SE: DNA mismatch repair protein Msh6 is required for e¡ective UVB-induced apoptosis in primary mouse ¢broblasts. J Invest Dermatol, in press Zhang H, Richards B, Wilson T, et al: Apoptosis induced by overexpression of hMSH2 or hMLH1. Cancer Res 59:3021^3027, 1999