Nucleotide Excision Repair and Cancer

Cancer ranks as one of the most frequent causes of death worldwide and in Western society it is competing with cardiovascular disease as the number one killer. This high frequency in Western countries can be attributed to lifestyle and environmental factors, only 5-10% of all cancers are directly due to heredity. Common environmental factors leading to cancer include: tobacco (25-30%), diet and obesity (30-35%), infections (15-20%), radiation, lack of physical activity and environmental pollutants or chemicals (Anand et al.,2008). Exposure to these environmental factors cause or enhance abnormalities in the genetic material of cells (Kinzler KW et al.,2002). These changes in the DNA or hereditary predisposition can result in respectively uncontrolled cell growth, invasion and metastasis. Cancer cells can damage tissue and disturb homeostasis leading to dysfunctions in the body that can eventually lead to death. Under normal conditions cell growth is under strict conditions and control. Hereditary dysfunctions or introduced DNA damage in tumor suppressor genes, oncogenes or DNA repair genes can create an imbalance that may lead to cancer development. DNA repair and cell cycle arrest pathways are essential cellular mechanisms to prevent or repair substantial DNA damage which, if left unattended, can cause diseases. Here, one of the most important and versatile DNA repair pathways, the Nucleotide Excision Repair (NER) pathway, will be discussed in relation to DNA damage accumulation and carcinogenesis together with its mechanistic mode of action.


Nucleotide excision repair
The abundant targeting of bases and nucleotides in the genome makes the Nucleotide Excision Repair one of the most essential repair pathways. NER can restore the correct genomic information, but also replication and transcription after these types of damage. The pathway can deal with a broad spectrum of (mostly) structurally unrelated bulky DNA lesions, arisen from either endogenous or exogenous agents. NER for example removes DNA lesions from the genome such as photolesions, crosslinks, bulky aromatic hydrocarbon and alkylation adducts (Figure 1). Nucleotide excision repair is a multistep pathway using over 30 proteins that eliminate the helix-distorting lesions. As mentioned, lesions of this matter can originate upon exposures to several damaging agents. For instance, UV radiation (sunshine) is a physical DNA damaging agent that mainly produces cyclobutane pyrimidine dimers (CPDs) and pyrimidine-(6,4)-pyrimidone products (6-4PP) but is also believed to induce oxidative DNA damage (Lo et al.,2005). Exposure to numerous chemicals can result in helix-distorting bulky adducts, for example polycyclic aromatic hydrocarbons (present in cigarette smoke or charcoaled meat) (de Boer et al.,2000). Interstrand crosslinks, alkylation adducts and oxygen free-radical induced minor base damage can trigger NER (Friedberg et al.,2006b).

Global genome-NER and transcription coupled-NER
NER is divided into two subpathways which mechanistically initiate in a divergent manner, but after damage recognition both pathways proceed along the same molecular route (see Figure 2). The subpathways are designated Global Genome NER (GG-NER) and Transcription Coupled NER (TC-NER). GG-NER recognizes and removes lesions throughout the entire genome, and is considered to be a relatively slow and somewhat more inefficient process, since it scans the whole genome for DNA damage (Guarente et al.,2008). However, UV induced helix-distorting lesions like 6-4PPs, are rapidly cleared by GG-NER (Garinis et al.,2006). TC-NER is responsible for eliminating lesions in the transcribed strand of active genes. This repair process takes care of lesions blocking the transcription machinery and otherwise possible resulting dysfunctions. Since TC-NER is directly coupled to the transcription machinery it is considered to be faster acting and more efficient than GG-NER, but is only initiated when transcription of a gene is blocked.

DNA damage recognition
The difference between the two sub pathways is the initial damage recognition step ( Figure 2). As mentioned previously, a helical distortion and alteration of DNA chemistry appears to be the first structural element that is recognized. For GG-NER, the XPC/hHR23B complex (including centrin2), together with the UV-Damaged DNA Binding (UV-DDB) protein (assembled by the DDB1 (p127) and DDB2/XPE (p48) subunits), are involved in lesion recognition (Dip et al.,2004). The XPC/hHR23B complex is additionally essential for recruitment of the consecutive components of the NER machinery to the damaged site, also known as the preincision complex (Yokoi et al.,2000;Araujo et al.,2001). Schematic overview of the Nucleotide Excision Repair (NER) pathway. Damaged DNA is recognized by either initial factors of the Global Genome Repair (e.g. Xpc) or Transcription Coupled Repair (CSA and CSB), which constitute the two different repair pathways in NER. After DNA damage recognition the repair route progresses along the same way. After helix unwinding and verification of the damage incisions are made to remove the faulty stretch of DNA. Finally, DNA synthesis and subsequent ligation reproduce the correct DNA sequence. It has been shown that XPC itself has affinity for DNA and can initiate GG-NER in vitro, but its functionality is enhanced when hHR23b and centrin2 are added Araki et al.,2001). The latter two are not able to bind to DNA themselves. Centrin2 as well as hHR23B stabilize the heterotrimer complex, putatively by inhibiting polyubiquitination of XPC and hence preventing subsequent degradation by the 26S proteasome . XPC recognizes various helix-distorting base lesions that do not share a common chemical structure. Biochemical studies have revealed that XPC recognizes a specific secondary DNA structure rather than the lesions themselves and the presence of singlewww.intechopen.com Nucleotide Excision Repair and Cancer 125 stranded DNA seems a crucial factor Sugasawa et al.,2002;Min et al.,2007). XPC appears to scan the DNA for distortions by migrating over the DNA, repeatedly binding and dissociating from the helix. When XPC encounters a lesion the protein changes its conformation and aromatic amino acid residues stack with unpaired nucleotides opposite the lesion, thereby increasing its affinity and creating a conformation which makes it possible to interact with other NER factors (Hoogstraten et al.,2008). Binding affinity of XPC to the DNA seems to correlate with the extent of helical distortion. 6-4PP products substantially distort the DNA structure and are therefore more easily recognized than CPDs, which only induce a minimal helical distortion . More recent studies have indicated that UV-DDB facilitates recognition of lesions that are less well-recognized by the XPC-hHR23B complex, like CPDs, via ubiquitylation of XPC (Fitch et al.,2003). The UV-DDB is able to recognize UV-induced photoproducts in the DNA and is now believed to precede the binding of XPC-hHR23B to the UV-damaged site. CPD repair is UV-DDB dependent (Fitch et al.,2003;Tang et al.,2000). Since affinity of the XPC-hHR23B to CPD sites is low, DDB2 is needed for efficient binding (Tang et al.,2000). Affinity of DDB2 for 6-4PP is also extremely high and the protein is furthermore able to bind to DNA lesions such as apurinic/apyrimidinic (AP) sites and mismatches (Nichols et al.,2000;Wittschieben et al.,2005). DDB2 is also part of the E3 ubiquitin ligase complex which is further comprised of CUL4A, ROC1/RBX1, COP9 signalosome (CSN) and DDB1 (Groisman et al.,2003). Live cell imaging studies show prompt recruitment of DDB1, DDB2 and Cul4a to UV induced lesions (Alekseev et al.,2008). CUL4A displays ubiquitin ligase activity and was shown to ubiquitylate DDB2 (Chen et al.,2001;Nag et al.,2001;Matsuda et al.,2005). The CSN subunit contains deubiquitylation capacities. This interactive mechanism is thought to be responsible for (poly)ubiquitylation of XPC and DDB2, but not in a similar fashion and result. Upon ubiquitylation DDB2 is degraded by the 26S proteasome (Fitch et al.,2003;Rapic-Otrin et al.,2002). XPC is not degraded after UV DNA damage, hereby increasing its binding affinity to the DNA as well as stimulating the interaction with hHR23B (Araki et al.,2001;Ortolan et al.,2004;Ng et al.,2003). Degradation of UV-DDB enhances the binding of XPC-hHR23B to the DNA in vitro . Timing of the programmed degradation of DDB2 determines the recruitment of XPC-hHR23B to the UV-damaged site (El Mahdy et al.,2006). The XPC protein contains several binding domains, a DNA binding domain, a hHR23B binding domain, centrin2 binding domain and a TFIIH binding domain (Sugasawa,2008). TFIIH is a multifunctional transcription factor and NER complex and amongst others contains the helicases XPB and XPD ( Figure 2). This complex is essential for the continuation of the NER pathway and is responsible for unwinding the DNA helix after damage recognition by XPC/hHR23B. XPC has been shown to physically interact with TFIIH and in vivo and in vitro studies have shown the recruitment of the NER complex to unwind the DNA is executed in a XPC-dependent manner (Sugasawa,2008;Friedberg et al.,2006b). The XPC protein is redundant in TC-NER. Here a stalled RNA polymerase II (RNA polII) is the onset of the NER machinery. The proteins CSA and CSB play a crucial role in setting the transcription coupled repair in motion, but are also implicated in RNA polII transcription functions. The CSB protein interacts with RNA polII (Tantin et al.,1997), while CSA does not (Tantin,1998). CSA mainly interacts with CSB, XAB2 (XPA binding protein 2) and the p44 subunit of the TFIIH complex (Henning et al.,1995;Nakatsu et al.,2000). The function of CSA remains to be elucidated but is implicated to be required for TC-NER during elongation of the transcription process (Groisman et al.,2003;Kamiuchi et al.,2002). CSA is also part of an E3 ubiquitin ligase complex in which CUL4A, CSN and DDB1 are involved. Both CSA and CSB are part of RNA PolII associated complexes, but for CSB additional functions are assigned outside NER (Sunesen et al.,2002). In TC-NER, CSB is thought to be responsible for displacement of the stalled RNA polymerase. Additionally, as with XPC in GG-NER, the preincision complex of NER is recruited in a CSB-dependent manner (Fousteri et al.,2008;Fousteri et al.,2006). But first, as in GG-NER, the TFIIH complex is recruited after damage recognition.

DNA helix unwinding
From DNA damage recognition and subsequent recruitment of TFIIH on, GG-NER and TC-NER converge into the same pathway. The TFIIH complex consists of 10 proteins: XPB, XPD, p62, pP52, p44, p34, p8 and the CDK-activating kinase (CAK) complex: MAT1, CDK7 and Cyclin H. TFIIH forms an open bubble structure in the DNA helix (Giglia-Mari et al.,2004;Goosen,2010). The DNA helicases XPB and XPD facilitate the partial unwinding of the DNA duplex in an ATP-dependent manner, allowing the preincision complex to enter the site of the lesion ( Figure 2) (Oksenych et al.,2010). The preincision complex consists of the XPA, RPA and XPG proteins and is assembled at the damage site (Zotter et al.,2006) ( Figure 2). The function of XPA is verification of the lesion; in addition, XPA acts, together with the single strand DNA binding complex RPA, as an organizational orchestrator, so that the repair machinery is positioned around the lesion. The arrival of XPA and RPA leads to complete opening of the damaged DNA and catalyzes the release of the CAK complex from the TFIIH complex. Data suggest this step is essential for the initiation of incision/excision of the damaged DNA (Andressoo et al.,2006;Coin et al.,2008). XPA and RPA in the preincision complex bind to the damaged DNA. Specific binding properties for certain structural DNA distortions have been reported for XPA, which suggests the protein recognizes conformations of the DNA and is able to verify the damage (Lao et al.,2000;Krasikova et al.,2010). XPA has no enzymatic activity to attribute to the incision step, but nevertheless is indispensable for DNA incision (Miyamoto et al.,1992). Through zinc finger motifs of XPA interaction with RPA is established (Ikegami et al.,1998). RPA consists of 3 subunits and with high affinity binds to the undamaged strand (de Laat et al.,1998b) (Figure 2). The protein roughly covers 30 nucleotides (corresponding in size to the excised product later in NER) and acts as a wedge between the DNA strands. It is also believed to protect the undamaged intact strand from inappropriate nuclease activity (de Laat et al.,1998a;Hermanson-Miller et al.,2002). RPA is concerned as the major component in the preincision complex, thereby also protecting the native template strand. Furthermore, RPA interacts with several other factors of the nucleotide excision repair pathway, like the endonucleases XPG and the ERCC1-XPF dimer, which are required for the dual incision of the damaged strand ( Figure 2). RPA hereby facilitates the correct positioning of the endonucleases and orchestrates the open complex formation (Krasikova et al.,2010;Park et al.,2006). Besides that, RPA later plays a role in the DNA repair synthesis and ligation steps (Shivji et al.,1995). XPA exhibits a moderate preference for binding to damaged DNA, as do other NER factors. Individually, the proteins XPA and RPA do not show sufficient selectivity to explain the high efficiency of NER lesion removal, which is most likely due to a discrimination cascade of recognition and verification steps (Lin et al.,1992). XPA is thought to play a major role in verification, possibly by acting as a molecular sensor of aberrant electrostatic potential along the DNA substrate (Camenisch et al.,2008). Structural changes, like kinked backbones, in damaged DNA can induce deviant electrostatic potentials. Bulky adducts often result in sharply bent backbones due to the decrease of rigidity. Normally, base stacking supplies the DNA helix with strength and structure, but bulky lesions and hence loss of base stacking can weaken the sturdiness of the backbone. It is known that XPA has a higher affinity for kinked backbones. However how the exact recognition of the lesion is executed remains to be elucidated. But it is clear that the XPA-RPA complex is indispensable for the high efficiency of NER (Camenisch et al.,2008).

Incision, DNA repair synthesis and ligation
When the preincision complex is accurately positioned in relation to the damaged site by the XPA-RPA complex, single strand breaks are introduced by XPG and ERCC1-XPF ( Figure 2). Several mechanistic theories were postulated over the years. A general consensus is that the combined actions of XPG and ERCC1-XPF result in excision of a 24-32 nucleotide long single strand fragment including the damaged site (Hess et al.,1997). XPG is responsible for the 3' incision and is putatively recruited by the TFIIH complex (Zotter et al.,2006). According to some studies its presence appears to be necessary for ERCC1-XPF activity, which is responsible for carrying out the 5' incision (Friedberg et al.,2006b;Wakasugi et al.,1997). Others propose a 'cut-patch-cut-patch' mechanism for the incision and resynthesis process within NER, where the 5' incision possibly precedes the 3' incision (Staresincic et al.,2009). XPG is expected to have additional stabilization features, because of its ability to interact with XPB, XPD and several other subunits of the TFIIH complex (Friedberg et al.,2006b). Since loss of XPG results in very early death (Wijnhoven et al.,2007) the protein might be involved in systemic and important additional mechanisms, like transcription (Bessho,1999;Lee et al.,2002). Furthermore, XPG is suggested to have a role in oxidative damage removal (Dianov et al.,2000). The ERCC1-XPF seems to be a multifunctional complex as well, since it is also involved in interstand crosslink repair and homologous recombination (Niedernhofer et al.,2001;Al Minawi et al.,2009). The excision of the damaged fragment is restored in original (undamaged) state by DNA synthesis and ligation steps (either by cut-patch-cut-patch mechanism or full excision followed by resynthesis and ligation). Both XPG and RPA are thought to be required for the transition between (pre)incision and post-incision events . XPG is thought to be involved in the recruitment of PCNA (Staresincic et al.,2009;. Resynthesis of DNA requires PCNA because of its ability to interact with DNA polymerases . The mechanism of involvement of these polymerases in DNA resynthesis is not yet fully elucidated. Recent studies show at least three DNA polymerases are involved. Pol , Pol κ and Pol are recruited to damage sites ( Figure 2). Recent in vivo studies show Pol β most likely plays no major role in NER (Ogi et al.,2010;Moser et al.,2007). To complete the repair of the damaged DNA site the resynthesized strand needs to be ligated. The primary participant in the subsequent ligation process of NER appears to be the XRCC1-Ligase 3 complex, which is shown to accumulate in both quiescent as well as proliferating cells after local UV irradiation (Moser et al.,2007). Ligase 1 appears to be involved in the ligation step in proliferating cells only (Moser et al.,2007). To date, the cross play of over 30 proteins in total is involved in NER to counteract DNA damage in the error free manner described above.

NER in cancer
DNA repair is vital to humans and other organisms and a defect in one of the genes can result in some severe syndromes or diseases by loss of genomic stability. Essential consequences of genomic instability can be cancer and other age-related diseases, such as neurological disorders as Huntington's disease and ataxias (Friedberg et al.,2006b). DNA damage for example can cause mutations that trigger (pre-)oncogenes, inactivate tumor suppressor genes or other indispensable genes which cause loss of homeostasis. Defects in the DNA repair machinery will inflate the mutational load, since DNA damage will be left unattended and subsequently gene mutations will accumulate. Therefore, organisms that harbor defective DNA repair are often more prone to develop cancer or (segmental) agerelated diseases. In humans, several syndromes have been identified which are the result of an impaired nucleotide excision repair pathway, of which Xeroderma Pigmentosum (XP), Cockayne syndrome (CS) and Trichothiodystrophy (TTD) are the most well-known. Since NER is the major defense against UV-induced DNA damage, all three syndromes are hallmarked by an extreme UV-sensitivity, of which XP ensues a highly elevated risk of developing skin cancer (Friedberg et al.,2006b;Cleaver et al.,2009). The involvement of NER genes in rare and severe syndromes underscores the vital importance of this repair pathway. It is known that accumulative DNA damage is one of the most important causes in cancer development and loss of homeostasis in organisms (Mullaart et al.,1990;Lindahl,1993;Friedberg et al.,2006b;de Boer et al.,2000;Cleaver et al.,2009). Defects in DNA repair pathways are therefore also considered to accelerate aging and tumorigenesis. In defective NER both types of endpoints occur, XP patients are predisposed to cancer development while CS and TTD patients are not. The latter exhibit premature aging features which XP patients lack (Friedberg et al.,2006b;de Boer et al.,2000;Cleaver et al.,2009). Reason for this might be the involvement of several NER proteins in other significant cellular mechanisms. CSB is believed to be involved in (TC-)BER, while XPD is also associated with replication and transcription. Some of these affected mechanisms could overshadow the NER deficiency and ever increasing mutational load eventually predisposing an individual to cancer. Severely affected developmental and neurological systems could be more life threatening on the shorter term than tumor development is. This could be the rationale behind the fact that CS and TTD patients are extremely short-lived and not cancer prone.

Xeroderma pigmentosum
Xeroderma pigmentosum (XP), meaning parchment pigmented skin, was the first human causal NER-deficient disease identified (Cleaver et al.,2009). It is a rare, autosomal inherited neurodegenerative and skin disease in which exposure to sunlight (UV) can lead to skin cancer. In Western Europe and the USA the incidence frequency is approximately 1:250,000, rates are higher in Japan (1:40,000). XP-C and XP-A are the most common complementation forms of XP (Bhutto et al.,2008). Early malignancies (from 1-2 years of age) in parts of the skin, eyes and the tip of the tongue develop due to sun-exposure (Table 1). Additionally, benign lesions like blistering, hyperpigmented spots and freckles are abundant (Figure 3). XP is associated with a more than 1,000-fold increase in risk of developing skin cancer. These cancers are mainly basal  (Kraemer et al.,1984). The disease is mostly symptomatic during childhood. The mean latency time for cutaneous neoplasms is 8 years, this is in comparison to the general population in which the mean latency time is 50 years later (Kraemer,1997). Progressive neurological degeneration occurs in approximately 20% of the XP cases and can be correlated to deficiencies in specific XP genes (XPA, XPB, XPD and XPG) (Cleaver et al.,2009). XP-C and XP-F patients rarely develop neurological degeneration and if so with a later onset when compared for example to XP-A and XP-D patients (Kraemer,1997;Friedberg et al.,2006b). The heterogeneity in exhibited symptoms is correlated to the genetic heterogeneity in XP patients. XP-A, XP-B, XP-D and XP-G patients are in general the most severely affected and all these patients are defective in both GG-NER and TC-NER. Solely GG-NER is defective in XP-C and XP-E patients. XP-C and XP-E cells show higher survival rate after UV exposure than XP-A and XP-D cells for example (Friedberg et al.,2006b). This could be the reason that XP-C patients suffer less from sunburn. Most abundant XP variants in human are XP-A and XP-C (~50% of all XP cases) (Zeng et al.,1997). Photo of a 19 year old Xeroderma pigmentosum patient suffering from hyperpigment skin lesions and a tongue carcinoma (IARC).

NER mouse models in cancer research
To investigate the role of the proteins involved in NER on survival and cancer development several transgenic mouse models were created, mimicking the existing NER mutations or deletions in humans. Table 2 shows an overview of NER mouse models and their accompanying spontaneous phenotypes. Selected knockout mouse models (Xpa, Xpc and Xpe) are described in more detail further below. These three models show a decreased lifespan in comparison to their concurrent wild type controls, but not as extreme as several other NER-deficient mouse models in Table 2. Therefore the mouse models survive long enough to study the effect of impaired NER on cancer development. Others, like Xpb, Xpf, Xpg and Ercc1 deficient models are too short-lived to study carcinogenesis.

Xpa deficient mouse model
The first DNA repair defective models were the Xpa-deficient mouse models, generated by de Vries et al. (de Vries et al.,1995) and independently by Nakane et al (Nakane et al.,1995). Xpa-deficient mice appeared more cancer prone compared to their heterozygous and wild type littermates when exposed to carcinogenic and genotoxic compounds (de Vries et al.,1997b;Takahashi et al.,2002;Ide et al.,2001;Hoogervorst et al.,2005;Hoogervorst et al.,2004). As in humans, the mouse model exhibited a marked predisposition to skin cancer upon UV treatment of shaved dorsal skin (de Vries et al.,1995). Survival studies without directed exposure were performed initially but always in a mixed genetic background, C57BL/6J/Ola129 (de Vries et al.,1997b) and C3H/heN strains (Takahashi et al.,2002) and fairly small numbers. However, both studies indicated that Xpa -/-mice (from here mentioned as Xpa mice) developed a significant number of spontaneous liver tumors. The C3H/heN strain wild type mice already showed 47% liver tumor incidence in the male mice within 16 months. The C57BL/6J/Ola129 mice were more resilient, no enhanced mortality was observed until the age of 1.5 years. The Xpa mice showed a 15% hepatocellular adenoma tumor incidence after 20 months, while there were no tumors in the wild type and heterozygous littermates. The lack of a pure genetic

132
C57BL/6J background more recently was investigated (Melis et al.,2008). C57BL/6J mice showed a low baseline tumor response and appear therefore suitable for studying mutagenesis and tumorigenesis. In a pure genetic background a significant increase in liver tumors was observed (10%). A small (but not significant) increase in lung tumors was also observed (6.6% of the Xpa mice) (Melis et al.,2008). Correspondingly, mutation accumulation in the C57BL/6J Xpa mice was significantly increased during survival compared to wild type mice in liver, implicating an Xpa repair defect and subsequent mutation induction in carcinogenesis (Melis et al.,2008). Like human XP-A patients, Xpa mice appeared predisposed to skin cancer after UV light exposure to shaved dorsal skin of the mice (de Vries et al.,1995;Tanaka et al.,2001). Heterozygous Xpa mice did not show this cancer prone phenotype after UV exposure, not even when the Xpa mutation was crossed in in hairless mice . Skin cancer predisposition in XP mice might not only involve NER deficiency, but several reports indicate enhanced immunosuppression and impaired natural killer cell function are involved (Gaspari et al.,1993;Horio et al.,2001;Miyauchi-Hashimoto et al.,2001). Xpa mice were also predisposed to tumors of the cornea when exposed to UV radiation, see Table 3  (de Vries et al.,1998). Chemical exposure of Xpa mice to 7,12-dimethyl-1,2-benz[a]anthracene (DMBA) also resulted in skin cancer (de Vries et al.,1995). Several chemical exposures in Xpa mice however shed some more light on the cancer development other than skin cancer, which in humans is the predominant tumor phenotype (Table 3). For example, oral treatment of Xpa deficient mice with genotoxic carcinogens like benzo[a]pyrene (B[a]P), 2acetylaminofluorene (2-AAF), and 2-amino-1-methyl-6-phenylimidazo [4,5-b]-pyridine (PhIP) resulted in lung tumors and lymphomas (B[a]P), liver and bladder tumors (2-AAF) and intestinal adenomas plus lymphomas (PhIP) (de Vries et al.,1997b;van Steeg et al.,1998;van Steeg et al.,2000;Ide et al.,2000). Other human carcinogens like cyclosporin A (CsA) and diethylstilbestrol (DES), although not directly mutagenic, showed to be carcinogenic in Xpa mice after 39 week exposure, but in contrary the low potent human carcinogen phenacetin did not result in a significant increase in tumors. LacZ and Hprt mutation measurements in Xpa mice after B[a]P and 2-AAF treatment showed a 2-3 fold increase in mutations compared to wild type mice after only 12-13 weeks of exposure (Hoogervorst et al.,2005;van Oostrom et al.,1999;Bol et al.,1998;de Vries et al.,1997b). This increase in mutational load in comparison to wild type indicates Xpa mice are more sensitive to mutation accumulation, which consequently corresponds to the increased cancer susceptibility of Xpa mice. The increased sensitivity towards cancer development of Xpa mice made it possible to identify genotoxic carcinogens even more accurate and faster when combined with heterozygosity for p53. This latter mouse model could be beneficial in reducing and refining in vivo carcinogenicity testing of compounds.

Xpc deficient mouse model
Two independent Xpc-deficient mouse models were also created in the mid-nineties (Cheo et al.,1997;Sands et al.,1995). As the Xpa mouse model, this model is informative for human XP and cancer development in general. The model is especially interesting since it is only defective for GG-NER and not for TC-NER. Hereby, differences between pathways can be investigated.

Mouse model Treatment Target
Enhanced tumor response*

Xpe/ DDB2
UV-B radiation Skin Yes (Itoh et al.,2004) DMBA paint Skin No (Itoh et al.,2004) * in comparison to the untreated controls DEHP = Di(2-ethylhexyl) phthalate AFB1 = Aflatoxin B1 4NQO = 4-Nitroquinoline 1-oxide WY = Wyeth-14643 p-cres = p-cresidine NOH-AAF = N-hydroxyacetylaminofluorene As in human XP-C patients, Xpc mice are highly predisposed to UV radiation-induced skin cancer (Table 3) Cheo et al.,1996;Cheo et al.,2000;Friedberg et al.,1999;Sands et al.,1995). Contrasting to Xpa +/-mice the heterozygous Xpc mice are more susceptible to UVinduced skin cancer (but only at approximately from 1 year old) when compared to their wild type littermates. The haploinsufficient sensitivity could mean that XPC is a rate limiting factor in NER and since XPC is involved in damage recognition might explain the difference with Xpa heterozygous mice. Exposure studies with 2-AAF using Xpc mice showed a significant predisposition to liver and lung tumors compared to the heterozygous Xpc and wild type mice (Table 3) Friedberg et al.,2006b). Internal tumor incidence is higher in XP mice than in human XP, since patients normally develop skin cancer at a faster rate and die of resulting metastatic complications. NER is believed to be the sole pathway to remove CPD and 6-4PP lesions, while for chemical carcinogenic exposure other repair mechanisms are also present in the cell. In human, other types of cancer generally do not develop fast enough and are possibly overshadowed by skin cancers in XP. In a mixed genetic background (C57BL/6J/129) no decrease in survival was found in relation to wild type mice, even though Xpc mice showed an extremely high and significantly increased lung tumor incidence (100%). However, in this study the wild type mice were not genetically related to the Xpc mice (Hollander et al.,2005). The spontaneous survival characteristics of Xpc mice in a pure genetic C57BL/6J background together with their related wild type littermates were also investigated. Xpc mice showed a significant decrease in survival, again exhibited a significant increase in lung and liver tumors and an increased mutation accumulation in these tissues compared to wild type mice (Melis et al., 2008). Here, Xpc mice showed a divergent tumor spectrum from Xpa mice in the same genetic C57BL/6J background. The additional increase in lung tumor development in two independent spontaneous survival studies indicate XPC is involved in other pathways besides NER. A corresponding strong increase in mutational load during aging was found in lungs of the C57BL/6J Xpc mice, which was not the case in Xpa mice (Melis et al.,2008). Uehara et al. have shown that enhanced spontaneous age-related mutation accumulation in Xpc mice is tissue dependent. Liver, lung, heart and spleen exhibited an increase in mutant frequency compared to wild type, while this difference was not visible in brain and small intestine. Mutant frequencies of liver, lung and spleen are higher in Xpc mice compared to Xpa mice, just as the tumor incidence in this study (Melis et al.,2008). The additional increase in mutational load in Xpc mice might be caused by increased sensitivity towards oxidative DNA damage. XPC functioning has been implied in other DNA repair pathways like base excision repair and non-homologous end joining or might be involved in redox homeostasis (D'Errico et al.,2006;Despras et al.,2007;Liu et al.,2010;Okamoto et al.,2008;Rezvani et al.,2010;Shimizu et al.,2003;Uehara et al.,2009). Chemical exposures to B[a]P (Wickliffe et al.,2006), 3,4-epoxy-1-butene (EB) (Wickliffe et al.,2006), DMBA (Wijnhoven et al.,2001) and UV-B (Ikehata et al.,2007) also showed significantly enhanced mutant frequencies compared to wild type mice in several tissues. Direct comparisons to Xpa mice in these studies have not been made, however when Xpa and Xpc mice were exposed to pro-oxidants (DEHP and paraquat) for 39 weeks, Xpc again exhibited a higher mutant frequency than Xpa.

Xpe deficient mouse model
In 2004 (Itoh,2006;Itoh et al.,2004;Yoon et al.,2005). The latter group reported that DDB2 -/-mice show a decrease in spontaneous survival (n=10) compared to wild type (Yoon et al.,2005). Also the heterozygous DDB2 +/-mice showed a decreased lifespan, although not as severe as the DDB2 -/-mice. Six out of 10 DDB2 -/-mice harbored tumors at the end of life, while 3 out of 10 DDB2 +/-mice were tumor bearing (Yoon et al.,2005). DDB2 -/-mice additionally showed to be cancer prone upon UV-B exposure, resulting in a significant increase in skin tumors (Table 3) (Itoh,2006;Itoh et al.,2004;Yoon et al.,2005). DMBA treatment however did not enhance tumor incidence compared to wild type (Table 3) (Itoh,2006;Itoh et al.,2004). DDB2 deficiency is, due to these and other studies, since being classified as a XPE phenotype. Besides UV-B and DMBA exposure, other in vivo carcinogen exposures have not been reported in these models so far. DDB2 is well conserved between humans and mice and appears to function as a tumor suppressor, at least in part, by controlling p53-mediated apoptosis after UV-irradiation (Itoh et al.,2004).

Conclusion
DNA repair has proven to be of vital importance and protects or at least delays cancer development and several age-related diseases. DNA damage accumulation and consequent mutation accumulation is considered pathogenic. NER has been shown to be a highly versatile and important DNA repair pathway, removing helix distorting DNA damages. Mutations in the XP genes of NER in human can result in the severe syndrome Xeroderma pigmentosum, which is accompanied by a cancer predisposition and severe UV sensitivity. Mouse models mimicking this human syndrome are important to study cancer development and the consequences of persistent DNA damage. Novel functionality of DNA repair proteins and implications of their deficiency in mutagenesis, cell cycle regulation, carcinogenesis and aging were discovered using NER-deficient models. Besides mechanistic insight these models can be used as a refined model in carcinogenicity testing, especially in combination with p53 heterozygosity. The increased cancer susceptibility can be beneficial towards a decrease in the number of animals used and the duration of carcinogenicity testing.

References
Al