Cancer Research Donn Young Telomeres
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 67, 5649, June 15, 2007. doi: 10.1158/0008-5472.CAN-06-3806
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yang, G.
Right arrow Articles by Tsao, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yang, G.
Right arrow Articles by Tsao, H.

Molecular Biology, Pathobiology, and Genetics

Loss of Xeroderma Pigmentosum C (Xpc) Enhances Melanoma Photocarcinogenesis in Ink4a-Arf–Deficient Mice

Guang Yang1, David Curley4, Marcus W. Bosenberg4 and Hensin Tsao1,2,3

1 Wellman Center for Photomedicine, 2 Department of Dermatology, 3 Massachusetts General Hospital Cancer Center, Massachusetts General Hospital, Boston, Massachusetts; and 4 Department of Pathology, University of Vermont, Burlington, Vermont

Requests for reprints: Hensin Tsao, Department of Dermatology, Massachusetts General Hospital, Bartlett 622, 48 Blossom Street, Boston, MA 02114. Phone: 617-726-9569; Fax: 617-724-2745 or 617-726-1206; E-mail: htsao{at}partners.org.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Despite an extensive body of evidence linking UV radiation and melanoma tumorigenesis, a clear mechanistic understanding of this process is still lacking. Because heritable mutations in both INK4a and the nucleotide excision repair (NER) pathway predispose individuals to melanoma development, we set out to test the hypothesis that abrogation of NER, by deletion of the xeroderma pigmentosum C (Xpc) gene, will heighten melanoma photocarcinogenesis in an Ink4a-Arf–deficient background. Experimentally, we generated a strain of mice doubly deficient in Xpc and Ink4a-Arf and subjected wild-type, Xpc–/–Ink4a-Arf+/+, Xpc–/–Ink4a-Arf–/–, and Xpc+/+Ink4a-Arf–/– mice to a single neonatal (day P3) dose of UVB without additional chemical promotion. Indeed, there was a significant increase in the development of dermal spindle/epithelioid cell melanomas in Xpc–/–Ink4a-Arf–/– mice when compared with Xpc+/+Ink4a-Arf–/– mice (P = 0.005); wild-type and Xpc–/–Ink4a-Arf+/+ mice failed to develop tumors. These neoplasms bore a striking histologic resemblance to melanomas that arise in the Tyr-vHRAS/Ink4a-Arf–/– context and often expressed melanocyte differentiation marker Tyrp1, thus supporting their melanocytic origination. All strains, except wild-type mice, developed pigmented and non-pigmented epidermal-derived keratinocytic cysts, whereas Xpc+/+Ink4a-Arf–/– mice exhibited the greatest propensity for squamous cell carcinoma development. We then screened for NRas, HRas, Kras, and BRaf mutations in tumor tissue and detected a higher frequency of rare KrasQ61 alterations in tumors from Xpc–/–Ink4a-Arf–/– mice compared with Xpc+/+Ink4a-Arf–/– mice (50% versus 7%, P = 0.033). Taken together, results from this novel UV-inducible melanoma model suggest that NER loss, in conjunction with Ink4a-Arf inactivation, can drive melanoma photocarcinogenesis possibly through signature Kras mutagenesis. [Cancer Res 2007;67(12):5649–57]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Several lines of evidence exist to support a role for UVB in the induction of skin cancer, including melanoma. Epidemiologic studies have revealed an increased risk of melanoma in case-control analyses (1). Despite the vast clinical data, there is limited genetic data to directly impute a mechanism. Patients with the rare disorder xeroderma pigmentosum (XP) exhibit a defect in nucleotide excision repair (NER) and a 600- to 8,000-fold increased risk in melanoma (2); individuals from XP complementation group C (XPC) seem to be at particular risk for melanoma (3). Other heritable causes of melanoma include germ-line CDKN2A (Ink4a-Arf) mutations (4). Melanoma penetrance among CDKN2A mutation carriers seems to correlate with ambient melanoma incidences suggesting that environmental UV radiation may modulate tumor formation even in the context of this high-risk, high-penetrance gene (5). Moreover, recent evidence suggests that murine cells lacking Ink4a-Arf may be impaired in UV lesion processing (6). Taken together, there is clear environmental participation (i.e., UV radiation), even in the presence of heritable high-risk melanoma alleles. A genetically faithful and experimentally tractable model that recapitulates melanoma ontogeny would be an ideal setting to dissect apart these various inputs and to potentially test novel preventive agents.

In response to combinations of topical 7,12-dimethylbenz(a)anthracene (DMBA) and UVB, hairless mice have been shown to develop benign and malignant melanocytic tumors (7, 8). Furthermore, several genetically engineered strains of mice have been shown to develop melanocytic tumors after neonatal UV irradiation. In one model, enforced global overexpression of hepatocyte growth factor/steel factor (HGF/SF) leads to the induction of melanomas (9), a process that is further enhanced by UV radiation (10) and loss of Ink4a-Arf (11). Thus, the Tg(MT-HGF/SF) mouse model elegantly shows that ectopic mitogenic signaling can cooperate with UV radiation insult to drive melanoma development. However, the effect of NER loss, a known risk factor for melanoma, has not been fully addressed by extant models.

Mice deficient in Xpa (12), Xpc (13), and Ddb2/Xpe (14) all recapitulate the human UV cancer phenotype; however, these NER-deficient mice do not seem to develop melanomas in response to UV irradiation. A recent report described a low rate of melanoma tumor formation in Xpa–/–Ink4a-Arf–/––deficient pigmented hairless mice treated with various combinations of topical DMBA, neonatal UV irradiation, and adult UV irradiation (15). Multiple melanocytic nevi, which have been described in chemical and UV-treated hairless mice, and six spindle-cell melanomas developed in 316 mice, although the melanomas were not associated with any specific treatment or genotype.

To create a mouse model of UV melanoma carcinogenesis faithful to known human susceptibility loci, we engineered a line of mice deficient in both Xpc and Ink4a-Arf. Because neonatal UV irradiation alone has been productive in creating melanomas in other genetically modified backgrounds, we adopted this paradigm for our UV radiation exposure treatment. Moreover, because oncogenic vHRAS (16) or NRAS (17) mutations can cooperate with Ink4a-Arf loss to form melanomas, we hypothesized that Xpc loss may create a genetic background that permits UV radiation mutagenesis and activation of RAS pathway genes and thus drive carcinogenesis in Ink4a-Arf–deficient mice.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Mice. Wild-type B6/129SF1 and Xpc null mice were purchased from Taconic Animal Models. Xpc null mice were crossbred with Ink4a-Arf null (Xpc+/+Ink4a-Arf–/–; courtesy of L. Chin, Dana Farber Cancer Institute, Boston, MA; ref. 16) mice. The obligate Xpc+/–Ink4a-Arf+/– F1 generation (50% B6/129SF1/50% FVB) were then intercrossed to generate the Xpc–/–Ink4a-Arf–/– founders. Wild-type B6/129SF1 controls and single knockout mice were backcrossed to generate the approximately 50% C6/129/50% FVB strain composition. Matched Xpc+/+Ink4a-Arf–/– null mice were generated by crossing wild-type B6/129SF1 mice with Xpc+/+Ink4a-Arf–/– null mice on a 100% FVB background, and the F1 generation was intercrossed to generate the necessary Xpc+/+Ink4a-Arf–/– mice. All mice were maintained and euthanized in accordance with husbandry protocols defined by the Massachusetts General Hospital (MGH) Subcommittee on Research Animal Care.

Tg(Tyr-vHRAS)-Xpc–/–Ink4a-Arf–/– mice were derived from crossing Tg(Tyr-vHRAS);Xpc+/+Ink4a-Arf–/– mice into Xpc–/–Ink4a-Arf–/– mice to yield the obligate Tg(Tyr-vHRAS)-Xpc+/–Ink4a-Arf–/– F1 generation. Because the Tg(Tyr-vHRAS) allele is on the Y chromosome (16), F1 mice were then intercrossed, and the male animals with Xpc+/+Ink4a-Arf–/– and Xpc–/–Ink4a-Arf–/– genotypes were selected and observed.

Mouse genomic DNA isolation and genotyping. Mouse genomic DNA was isolated from tail clip samples using the DNeasy Tissue kit from QIAGEN, Inc. Xpc genotyping was carried out using primers: F-PGK-864, 5'-CGCACGCTTCAAAAGCGCACGTCTGCCGCG-3' (Xpc-KO F); R-HPRT988, 5'-CGACGCTGGGACTGCGGGTCGGCATGACGG-3' (Xpc-KO R); mXPC-IN445F, 5'-ACAGCCTACCCAGGGGATGTGATGCTTCCT-3' (Xpc-WT F); mXPC-IN670R, 5'-TGGTCTGTGGACAGAGAACACTGGCTGTGC-3' (Xpc-WT R).

Ink4a genotyping was done using published protocols (16). Amplification products were detected on 1.5% or 2% agarose gel stained with ethidium bromide.

Mouse UVB radiation, tumor development surveillance, and tissue collection and storage. A single dose of UVB irradiation (1,000 mJ/cm2) was delivered to P3-4 neonatal pups in accordance with an animal protocol (legacy 2004N000171/1) approved by the MGH. Close follow-up after irradiation did not reveal any toxicity. No fatalities or skin injuries (e.g., blistering) were observed in any of the irradiated mice. Mice were monitored two to three times per week for evidence of tumor formation or other health issues. As per our Institutional Review Board protocol, mice were sacrificed if any lesion exceeded 14 mm, or if the tumors created a functional deficit. Otherwise, mice were euthanized after 1 year of observation.

Clinically detectable lesions were then split, and half of the lesion was immediately frozen in liquid nitrogen for RNA, whereas the other half was fixed in formalin and processed in formalin. Paraffin-embedded, 5-µm sections were stained with H&E and evaluated by two independent observers (H. Tsao and M. Bosenberg).

Mutational detection. Nras, Kras, Hras, and Braf mutations were assessed by direct sequencing of reverse transcription-PCR (RT-PCR)–amplified products using the following primers using standard PCR conditions: Nras (forward, 5'-GGAGTTTGAGGTTTTTGCTGGTGTG-3'; reverse, 5-GCCAGTTCGTGGGCTTGCTTT-3'), Kras (forward, 5'-CGCGGCGCGGAGAGAG-3'; reverse, 5-CCTTGCTAACTCCTGAGCCTGTTTC-3'), Hras (forward, 5'-GATTGGCAGCCGCTGTAGAAGCT-3'; reverse, 5-GGTCCTGGGCCTGCCGA-3'), Braf (forward, 5'-TCATGGGCTATTCTACAAAGCCACAAC-3'; reverse, 5-CGTCTGACTGAAAGCTATACGGGTTTTTA-3').

Murine embryo fibroblast studies. Day 13.5 embryos were digested with 0.25% Trypsin-EDTA (Invitrogen) and cultured in DMEM (Sigma-Aldrich), supplemented with 10% FCS (Invitrogen) and 1% penicillin and streptomycin (Invitrogen). Cells were expanded for two additional passages and then frozen down in liquid nitrogen.

One day before UVB irradiation, 100,000 murine embryo fibroblast (MEF) cells were seeded in six-well plates. Cells were then irradiated with UVB at the indicated doses. Survival was then analyzed by either cell count or 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma-Aldrich) assay. For this latter assay, a 20 x MTT stock solution was made in a final concentration of 5 mg/mL in PBS. MTT medium was made by diluting the stock solution to 1 x with DMEM growth medium immediately before use. MTT assay was carried out by standard protocol with spectrometric measurements at 560 to 700 nm.

RNA isolation, reverse transcription, and quantitative PCR. Total RNA was isolated by RNeasy mini kit from QIAGEN. Reverse transcription was carried out by SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen). The first-strand cDNA synthesizing was carried out in 40-µL reaction volume with 5 µg of total RNA. Real-time PCR was carried out in 50 µL reaction volume on MX 4000 instrument (Stratagene) with TaKaRa Taq Polymerase (Fisher Scientific) to a final concentration of 10 mmol/L Tris-HCl (pH 8.3), 50 mmol/L KCl, and 1.5 mmol/L MgCl2.

Pre-designed Taqman primers were obtained from Applied Biosystems, and ß-actin was used for normalization (ß-actin, 4352341E; Tyrp1, Mm00453201_m1). For the quantitative PCR reaction, 4 µL of 1:8 diluted RT products (the RT product of 62.5 ng RNA) was used in each reaction. Quantitative PCR was carried out by the protocol from Applied Biosystems (50°C for 2 min, 95°C for 10 min, 95°C for 15 s, 60°C for 1 min x 40 cycles). Serial dilutions for each primer set was used to determine efficiency, and quantitation of target gene product relative to endogenous control was obtained using the Pfaffl equation (18), with Etarget and Eref being the efficiencies (E = 2.0 at 100% efficiency) of the target and endogenous control, respectively.

For immunohistochemistry, an anti-Tyrp1 antibody (PEP1 polyclonal antiserum from Dr. Vincent Hearing, NIH, Bethesda, MD) was used on formalin-fixed, paraffin-embedded sections of mice tumors. Standard immunohistochemical techniques were employed with 3,3'-diaminobenzidine staining after the secondary antibody.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
UVB-induced tumors. Upon UVB exposure, there were two large classes of tumors observed in the experimental animals (Supplementary Table S1): cysts and invasive cancers. An overall tabulation of the tumor phenotypes by genotype is shown in Supplementary Table S1.

UVB-induced cysts. With loss of either Xpc or Ink4a-Arf, mice developed cysts, often pigmented and frequently oversized. Figure 1A and B shows the range in sizes of these lesions (~2 mm to >10 mm), which are usually, but not exclusively, located on the tail. Within any one animal, the number of lesions ranged from 1 to >10 (data not shown). Figure 1C and D shows the histologic pattern of these cysts. In general, they were dermal nodules with a relatively thick wall composed of multiple layers of bland keratinocytes. Within these cysts, there was substantial keratotic debris with occasional "ghosting" of cells. The keratinocytes surrounding the debris were frequently pigmented suggestive of melanocytic participation, although the cells themselves do not show evidence of atypia.


Figure 1
View larger version (84K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Epidermal inclusion cysts after UV radiation exposure. A, gross view of a small ~3-mm cystic papule located on tail. B, a 1-cm cystic lesion also located on tail. Bar, 1 cm. C, epidermal inclusion cyst (low power). The sections show tail skin with an underlying squamous epidermal inclusion cyst. The lesion is well circumscribed and exhibits central keratinization. D, epidermal inclusion cyst (high power). The sections show a circumscribed cystic squamous proliferation with central keratinization.

 
Figure 2A shows the Kaplan-Meier cyst-free survival for the four genotypes of interest (Xpc+/+Ink4a-Arf+/+, Xpc–/–Ink4a-Arf+/+, Xpc–/–Ink4a-Arf–/–, and Xpc+/+Ink4a-Arf–/–). Deletion of either Xpc or Ink4a-Arf accelerated UV-induced cyst formation, although the combined loss of both Xpc and Ink4a-Arf did not enhance this effect. These cysts were not observed in unirradiated mice of any genotype. Thus, loss of either Xpc or Ink4a-Arf is rate limiting for cyst formation.


Figure 2
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Kaplan-Meier curves of (A) cyst-free survival and (B) cancer-free survival after (C) melanoma-free survival and (D) SCC-free survival after UV radiation exposure. WT, wild type.

 
UVB-induced cancers. With the single neonatal UVB irradiation protocol, Xpc loss in isolation did not lead to a strong cancer phenotype (Kaplan-Meier cancer-free survival shown in Fig. 2B). One of 27 Xpc–/–Ink4a-Arf+/+ mice developed squamous cell carcinoma (SCC) approximately 1 year after irradiation. Although there was the suggestion that Xpc–/–Ink4a-Arf–/– mice were more susceptible to cancers than Xpc+/+Ink4a-Arf–/– mice, the difference was not statistically significant (P = 0.199, log-rank test). Supplementary Table S1 also delineates the cancer pathologies from the tumor-susceptible animals. The presence of cysts and cancers seemed to behave independently; that is, all possible combinations (cyst alone, cyst and cancer, and cancer alone) occurred with similar frequencies in the Xpc–/–Ink4a-Arf+/+ and Xpc+/+Ink4a-Arf–/– genotypes.

Many tumors in the irradiated mice exhibited spindle- and epithelioid-cell morphologies arranged in intersecting cords. This cellular and architectural pattern is reminiscent of melanomas arising in the Ras/Ink4a-Arf model (Fig. 3A–D ). Although the histologic similarities between the spindle/epithelioid melanomas from the Ras/Ink4a-Arf and spindle/epithelioid tumors from UV-irradiated Xpc–/–Ink4a-Arf–/– mice were striking, we further assessed for expression of Tyrp1, a known melanocytic antigen, in many tumors. In our quantitative PCR reaction, we normalized Tyrp1 mRNA levels to ß-actin (Tyrp1N) and measured relative levels to a negative biological control (MEFs) and a positive biological control [B16 murine melanoma cell line; i.e., normalized, relative Tyrp1 levels (Tyrp1NR) where Tyrp1NR = 1.0 for MEFs]. To establish arbitrary tissue levels that can be used to discriminate melanomas from non-melanomas and to define the range of Tyrp1NR expression in tissue specimens, we used a set of UV radiation–induced SCCs (negative tissue control) and Ras/Ink4a-Arf melanomas (positive tissue control). The Tyrp1NR levels ranged from 29 to 109 for SCC tumor specimens (i.e., 29- to 109-fold more than normalized MEF levels), whereas the Tyrp1NR levels for Ras/Ink4a-Arf melanomas ranged from 188 to 4,528. Thus, for tumors with some histologic ambiguity and for which we had mRNA, we considered the lesion a melanoma only if the relative Tyrp1NR exceeded the lowest level observed for Ras/Ink4a-Arf melanomas (i.e., 188-fold; listed as Tyrp1+ in Supplementary Table S1; others lower than this cutoff are listed as Tyrp1–; NR, no RNA available for analysis). As shown in Fig. 3E to H, there is correlation between the protein levels of Tyrp1, as judged by immunohistochemistry, and Tyrp1 mRNA levels, as determined by quantitative PCR. Given the pathologic and molecular features of these tumors, we classified them as spindle-cell or spindle/epithelioid-cell melanomas depending on the predominant cellular morphology. As shown in Fig. 2C, loss of Xpc enhanced the development of these melanomas in Ink4a-Arf null mice (Xpc+/+Ink4a-Arf–/– versus Xpc–/–Ink4a-Arf–/–: P = 0.005, log-rank test).


Figure 3
View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Malignant melanomas. A and B, malignant melanoma, spindle and epithelioid type, arising in a Ras/Ink4a-Arf–/– mouse. A, low-power view of the spindle and epithelioid melanoma arising in ear skin. B, high-power view showing the cytologic features of the spindle and epithelioid cells. C and D, malignant melanoma, spindle and epithelioid type, arising in a UV-irradiated Xpc–/–Ink4a-Arf–/– mouse. C, low-power view of the spindle and epithelioid melanoma invading skeletal muscle. D, high-power view showing the cytologic features of the spindle and epithelioid cells. Note the similar histologic features of the melanomas arising in these two genetic backgrounds. Tyrp1 staining in spindle/epithelioid tumors. E and G, H&E staining of two spindle/epithelioid tumors. F and H, Tyrp1 expression levels for tumors shown in E and G, respectively, as determined by immunostaining with the PEP1 polyclonal anti-Tyrp1 antibody (courtesy Vincent Hearing). F, clear anti-Tyrp1 staining. H, lacks staining. By quantitative PCR, the tumor in (E) had 60-fold greater expression of Tyrp1 mRNA than the tumor in (G). The tumor in (E) was diagnosed as a melanoma, whereas the tumor in (G) was designated a fibrosarcoma.

 
There were also tumors that were keratotic in clinical appearance and on histology, represented SCCs with typical keratin whirls (Fig. 4 ). There is the suggestion that loss of Xpc actually decreased the formation of SCCs in Xpc+/+Ink4a-Arf–/– mice (Fig. 1D; Xpc+/+Ink4a-Arf–/– versus Xpc–/–Ink4a-Arf–/–: P = 0.048, log-rank test). Thus, the effect of Xpc loss on UV radiation–induced cancers in the Ink4a-Arf–/– context seems to be cell type specific, although it is possible that our UV irradiation protocol itself is not inductive of keratinocytic malignancies.


Figure 4
View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Invasive SCC. A, low-power view showing a moderately differentiated squamous neoplasm that is composed of islands of squamous epithelium showing partial keratinization. B, high-power view shows cellular pleomorphism and invasion.

 
Tumor mutations. We were able to screen 28 tumors from UV radiation–exposed animals (Table 1 ) for mutations in Kras, Nras, Hras, and Braf; there were no identifiable mutations in Hras, Nras, or Braf. However, 7 of 14 tumors from Xpc–/–Ink4a-Arf–/– mice and 1 of 14 tumors from Xpc+/+Ink4a-Arf–/– animals harbored mutations at codon 61 of Kras (50% versus 7%: P = 0.033, two-tail Fisher exact test). These changes were not limited to the typical dipyrimidine C -> T alterations commonly seen with direct UVB damage; however, the most common oncogenic mutation in human melanomas (i.e., BRAFT1799A) is also not a UVB signature change. Thus, the relative, but not absolute, selectivity of KrasQ61 alterations for Xpc null mice suggests that mutational events at this codon may represent a signature change for tumors deficient in Xpc function.


View this table:
[in this window]
[in a new window]

  		Table 1. Kras mutations detected in tumors

 
Loss of Xpc in Ras/Ink4a-Arf model of melanoma tumorigenesis. We also explored the possibility that deletion of Xpc would increase the formation of Ras-driven melanomas in Ink4a-Arf null mice without UVB exposure. We found no difference in the rate of melanoma tumor development between Tg(Tyr-vHRAS);Xpc–/–Ink4a-Arf–/– and Tg(Tyr-vHRAS);Xpc+/+Ink4a-Arf–/– mice (P = 0.449; data not shown). Thus, inactivation of Xpc does not increase tumorigenesis once oncogenic RAS activity is present.

MEF sensitivity. One diagnostic hallmark of XP is the exquisite sensitivity of XP fibroblasts to UV radiation. We determined the UV sensitivity of MEFs from each of the individual genotypes. As expected, Xpc–/–Ink4a-Arf+/+ MEFs were significantly more sensitive to UVB irradiation than wild-type or Xpc+/+Ink4a-Arf–/– MEFs (Fig. 5 ). Inactivation of Ink4a-Arf in Xpc deficient MEFs, however, does not seem to restore UVB resistance. Rather, Xpc loss seems dominant to Ink4a-Arf loss as Xpc–/–Ink4a-Arf+/+ and Xpc–/–Ink4a-Arf–/– MEFs showed similar UV sensitivity. Despite comparable cellular sensitivities, however, there is a substantial clinical difference between Xpc–/–Ink4a-Arf+/+ and Xpc–/–Ink4a-Arf–/– mice in terms of cancer susceptibility. Xpc deficiency initiates a mutagenic pathway that stochastically increases the probability of oncogenic hits (such as the KrasQ61 alteration) and/or tumor suppressor gene inactivation. The low rate of cancers observed in Xpc–/–Ink4a-Arf+/+ mice after a single UVB dose may reflect the low statistical chance of these multiple genetic changes occurring in the same cancer-originating cell.


Figure 5
View larger version (9K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Sensitivity of MEFs to UV radiation. Day E13.5 MEFs were exposed to increasing doses of UVB. Survival was determined by either cell count or MTT assay. Points, % survival from four independent experiments; bars, SD.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
With this study, we set out to test the hypothesis that UV radiation in conjunction with NER loss cooperates with Ink4a-Arf deletion to recapitulate melanoma tumorigenesis. With the observation of enhanced melanoma photocarcinogenesis and the identification of recurrent KrasQ61 mutations in the tumor specimens, this model represents the first UV-responsive melanoma model that does not enforce specific oncogenic alleles [i.e., Tg(Tyr-vHRAS) (16) or Tg(Tyr- NRASQ61K) (17)] or mitogenic stimuli [i.e., Tg(MT-HGF/SF) (9)]. The model is permissive in allowing an environmental carcinogen (i.e., UV radiation) to specify the requisite oncogenic mutations once NER is abrogated.

Given the design of this experiment, we were able to assess for the effects of either Xpc loss in an Ink4a-Arf–deficient context (i.e., Xpc+/+Ink4a-Arf–/– versus Xpc–/–Ink4a-Arf–/–) or Ink4a-Arf inactivation in the context of Xpc deficiency (Xpc–/–Ink4a-Arf+/+ and Xpc–/–Ink4a-Arf–/–). After a single high-dose UVB irradiation, two phenotypes were observed. Benign, proliferative cysts, which has been previously described (15), occurred with high penetrance when either Xpc or Ink4a-Arf were genetically ablated; concurrent deletion of both loci did not seem to enhance cyst formation. A second stochastically independent phenotype is that of cancer formation. In contrast to chronic UV irradiation, neonatal UV exposure was only weakly inductive of skin cancers in Xpc–/–Ink4a-Arf+/+ mice. In our UV treatment protocol, loss of Ink4a-Arf was necessary to induce the cancer phenotype. It is entirely possible that a more rigorous UV regimen is needed to undergo biallelic inactivation of the Ink4a-Arf locus.

One commonly observed cancer in our model exhibited morphologic features and Typr1 expression levels comparable to melanomas from the Ras/Ink4a-Arf model. The UV radiation induction of these tumors in Ink4a-Arf null animals was significantly enhanced by Xpc loss. These findings have several implications. First, loss of Ink4a-Arf is not rate limiting for melanoma photocarcinogenesis, but rather, additional genetic lesions (e.g., Xpc deletion) can cooperate to further drive melanocytic transformation. This suggests that in humans, repair efficiency may dictate melanoma risk and potentially modulate melanoma development among CDKN2A mutation carriers. Second, the high frequency of KrasQ61 mutations in the Xpc–/–Ink4a-Arf–/– tumors suggests that Xpc abrogation may indeed define a novel Kras hotspot. To date, RAS-driven melanomas in Ink4a-Arf–/– mice [vHRAS (16) or NRAS (17)] have relied upon enforced expression of activated human RAS alleles in melanocytes; thus, a mutagenic mechanism is not needed. Our model is a permissive model whereby global NER is compromised, thereby allowing UV radiation to generate melanoma-causative mutations in murine genes. It is interestingly to note that direct UVB signatures were not prevalent in our tumors consistent with human melanoma oncogenes that are predominated by transversion events, especially BRAFT1799A. In XP patients, the frequency of RAS mutations in skin cancers is ~50%, which is roughly twice that of non-XP skin tumors (19). Most of these mutations are on codon 12, with a preponderance of NRAS mutations, although there was a single KrasQ61 mutation in 30 XP tumors but none in non-XP tumors. Additional analysis of human XP melanomas would certainly be useful. In human non-XP melanomas, mutations occur almost exclusively on NRASQ61, although the frequency is still less than that observed for BRAFV600 (20). In contrast, KRAS mutations are common in colon, pancreatic, and lung cancers (21) with codon 12 alterations occurring almost exclusively with only rare exceptions. In terms of environmental carcinogenesis, KRASG12 hotspot mutations have been reported in connection with benzo[a]pyrene diol epoxide exposure in vitro (22), whereas a KrasY64 hotspot has been described in murine hepatocellular carcinomas induced by i.p. injections of bleomycin and 1-nitropyrene (23). The KrasQ61 mutations share codon specificity with melanoma (i.e., codon 61 mutations of NRAS) but isoform selectivity (i.e., KRAS) with other visceral malignancies. To date, the activation of KrasQ61 has been a rare phenomenon, and the full functional consequences of this change are still unknown. The complex genetic interaction between RAS isoform, codon mutation, and the tissue of origin remains a central enigma in cancer biology.

Loss of Xpc had neither a positive nor negative effect on vHRAS/Ink4a-Arf melanoma tumorigenesis. In contrast, Ink4a-Arf null mice deficient in DNA repair through disruption of the Cockayne-Syndrome B (Csb) gene exhibited impaired spontaneous tumor formation (24). This suggests that ongoing removal of non-UV bulky DNA adducts by transcriptional coupled NER, as executed by Csb, may be required for Ink4a-Arf–driven tumorigenesis. However, our data suggest that global genome NER, which is functionally linked to Xpc protein, is unnecessary if a primary mutagenic event (such as vHRas) has already occurred. Our data lend further support to the view that Xpc loss, at least in part, plays a larger role in initiation by UV radiation but a less important role in progression.

A somewhat paradoxical observation was that loss of Xpc seemed to mitigate against SCC formation (P = 0.048, log-rank test). This may be related to our technical protocol using the single neonatal irradiation rather than chronic UV irradiation. However, there is also emerging evidence that different cells may have differential sensitivities to UV radiation. Recently, Stout et al. showed that Xpc–/– keratinocytes responded to UVB irradiation by arresting in late S phase without any significant apoptosis (25). Because Ink4a-Arf loss promotes G1 progression, our observation is compatible with a model whereby increased numbers of cells transit G1 to S in Ink4a-Arf–/– keratinocytes; however, they are arrested in late S phase because of Xpc loss and eventually shed during epidermal turnover (25). Although consistent with in vitro data, this mechanistic hypothesis is at odds with in vivo data where both mice and humans deficient in Xpc are found to be highly susceptible to UV-induced SCCs.

It is also instructive to compare and contrast our results with those published by van Schanke et al. (15). The Xpa–/–/Ink4a-Arf–/– mice that were used in that study were generated on a Hairless (Hr) background. Although there are many advantages to using the Hr strain for skin carcinogenesis studies, the effect of an additional genetic alteration (i.e., Hr loss) makes direct comparison between the two studies difficult. Van Schanke et al. did observe an augmentation of cyst and melanoma formation with neonatal UVB irradiation, although they were predominantly in the Xpa+/+ mice; we also observed cysts in Xpc+/+Ink4a-Arf–/– mice, although cyst formation was also quite brisk in the Xpc–/–Ink4a-Arf+/+ population. The melanomas observed after irradiation also exhibited spindle-cell morphology comparable to our tumors. The various chemical and UV treatment regimens between the two studies were also quite different. We did not use topical DMBA and chronic adult UVB treatment, which have been shown to induce melanocytic nevi and tumors in regular Hr mice without loss of Xpa (7, 8). We also did not observe pigmented nevi despite performing these experiments on pigmented animals. The loss of Hr may render the animal more prone to nevus formation, and thus in our Hr+/+ background, we would not observe these lesions. Alternatively, the exposure protocol may require additional insults to generate nevi. Of note, among all 15 experimental conditions outlined in the van Schanken et al. study, the yield of nevi was lowest in the single neonatal irradiation protocol. Thus, experimental conditions may also limit our ability to detect nevi. Genetically, none of the melanomas reported by van Schanken harbored mutations in Kras, Hras, Kras, or Braf; however, two pigmented nevi had demonstrable changes: one at codon 60 in Nras and the other at codon 13 in Kras.

Another UV-inducible mouse melanoma model is that of transgenic mice overexpressing HGF/SF (9, 10); furthermore, inactivation of Ink4a-Arf in the Tg(MT-HGF/SF) mice accelerated melanomagenesis (11). The histology of melanomas arising on the HGF/SF background is very different than the ones we observed or the tumors that arise on the vHRAS/Ink4a-Arf background. The Xpc/Ink4a-Arf melanomas, like the vHRAS/Ink4a-Arf lesions (16), are primarily dermal with spindle and epithelioid cellular morphology, whereas the HGF/SF melanomas exhibit a high degree of junctional activity along with Pagetoid spread (9). Normal truncal skin from Tg(MT-HGF/SF) mice show melanocytes outside of their normal niche (i.e., hair follicles); melanocytes can be detected in the epidermis, dermis and junction in these animals (9). Moreover, melanomas can arise spontaneously in the absence of UV exposure in the HGF/SF mice. Thus, the mitogenic stimulus conferred by ectopic and broad expression of Tg(MT-HGF/SF) seems to define melanocyte localization and, as a consequence, may dictate the histology of melanoma. Similarly, in Xpa–/–Tg(Scf) mice, Yamazaki et al. (26) also reported melanocyte retention in the epidermis comparable to human skin; upon high-dose UVB irradiation (150 J/cm2 over 10 weeks), the transgenic mice also developed melanomas with features suggestive of humanized architecture (i.e., lentigo maligna melanoma and nodular melanomas). Because the genetic bases of melanomas from these models have not been fully clarified, the molecular correlates of these histologic features are not known. Moreover, because UVB exposure also leads to immunosuppression, and because immune surveillance is known to play a critical role in melanoma biology (27), these models also serve as opportunities to understand the immune system in the context of melanomagenesis.

In conclusion, Xpc loss seems to sensitize Ink4a-Arf null mice to UV-induced melanomas and to specify a unique KrasQ61 mutational hotspot. This model integrates many known components of human melanoma carcinogenesis (i.e., UV exposure, RAS pathway activation, Ink4a-arf inactivation, and NER compromise) and may be very useful in dissecting apart the various genetic and environmental contributions to melanoma formation; moreover, it can serve as a potential model for chemoprevention.


   Acknowledgments
 
Grant support: Dermatology Foundation, American Skin Association, and NIH grants K08CA095532 (H. Tsao) and R01CA112054-02 (M.W. Bosenberg).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


   Footnotes
 
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Received 10/16/06. Revised 3/22/07. Accepted 4/20/07.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Gandini S, Sera F, Cattaruzza MS, et al. Meta-analysis of risk factors for cutaneous melanoma: II. Sun exposure. Eur J Cancer 2005;41:45–60.[CrossRef][Medline]
  2. Kraemer KH, Lee MM, Andrews AD, Lambert WC. The role of sunlight and DNA repair in melanoma and nonmelanoma skin cancer. The xeroderma pigmentosum paradigm. Arch Dermatol 1994;130:1018–21.[Abstract/Free Full Text]
  3. Lynch HT, Fusaro RM, Johnson JA. Xeroderma pigmentosum. complementation group C and malignant melanoma. Arch Dermatol 1984;120:175–9.[Abstract/Free Full Text]
  4. Hussussian CJ, Struewing JP, Goldstein AM, et al. Germline p16 mutations in familial melanoma. Nat Genet 1994;8:15–21.[Medline]
  5. Bishop DT, Demenais F, Goldstein AM, et al. Geographical variation in the penetrance of CDKN2A mutations for melanoma. J Natl Cancer Inst 2002;94:894–903.[Abstract/Free Full Text]
  6. Sarkar-Agrawal P, Vergilis I, Sharpless NE, DePinho RA, Runger TM. Impaired processing of DNA photoproducts and ultraviolet hypermutability with loss of p16INK4a or p19ARF. J Natl Cancer Inst 2004;96:1790–3.[Abstract/Free Full Text]
  7. Husain Z, Pathak MA, Flotte T, Wick MM. Role of ultraviolet radiation in the induction of melanocytic tumors in hairless mice following 7,12-dimethylbenz(a)anthracene application and ultraviolet irradiation. Cancer Res 1991;51:4964–70.[Abstract/Free Full Text]
  8. Epstein JH, Epstein WL, Nakai T. Production of melanomas from DMBA-induced "blue nevi" in hairless mice with ultraviolet light. J Natl Cancer Inst 1967;38:19–30.[Medline]
  9. Noonan FP, Otsuka T, Bang S, Anver MR, Merlino G. Accelerated ultraviolet radiation-induced carcinogenesis in hepatocyte growth factor/scatter factor transgenic mice. Cancer Res 2000;60:3738–43.[Abstract/Free Full Text]
  10. Noonan FP, Recio JA, Takayama H, et al. Neonatal sunburn and melanoma in mice. Nature 2001;413:271–2.[CrossRef][Medline]
  11. Recio JA, Noonan FP, Takayama H, et al. Ink4a/arf deficiency promotes ultraviolet radiation-induced melanomagenesis. Cancer Res 2002;62:6724–30.[Abstract/Free Full Text]
  12. de Vries A, van Oostrom CT, Hofhuis FM, et al. Increased susceptibility to ultraviolet-B and carcinogens of mice lacking the DNA excision repair gene XPA. Nature 1995;377:169–73.[CrossRef][Medline]
  13. Sands AT, Abuin A, Sanchez A, Conti CJ, Bradley A. High susceptibility to ultraviolet-induced carcinogenesis in mice lacking XPC. Nature 1995;377:162–5.[CrossRef][Medline]
  14. Itoh T, Cado D, Kamide R, Linn S. DDB2 gene disruption leads to skin tumors and resistance to apoptosis after exposure to ultraviolet light but not a chemical carcinogen. Proc Natl Acad Sci U S A 2004;101:2052–7.[Abstract/Free Full Text]
  15. van Schanke A, van Venrooij GM, Jongsma MJ, et al. Induction of nevi and skin tumors in Ink4a/Arf Xpa knockout mice by neonatal, intermittent, or chronic UVB exposures. Cancer Res 2006;66:2608–15.[Abstract/Free Full Text]
  16. Chin L, Pomerantz J, Polsky D, et al. Cooperative effects of INK4a and ras in melanoma susceptibility in vivo. Genes Devel 1997;11:2822–34.[Abstract/Free Full Text]
  17. Ackermann J, Frutschi M, Kaloulis K, McKee T, Trumpp A, Beermann F. Metastasizing melanoma formation caused by expression of activated N-RasQ61K on an INK4a-deficient background. Cancer Res 2005;65:4005–11.[Abstract/Free Full Text]
  18. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001;29:e45.[Abstract/Free Full Text]
  19. Daya-Grosjean L, Robert C, Drougard C, Suarez H, Sarasin A. High mutation frequency in ras genes of skin tumors isolated from DNA repair deficient xeroderma pigmentosum patients. Cancer Res 1993;53:1625–9.[Abstract/Free Full Text]
  20. Tsao H, Goel V, Wu H, Yang G, Haluska FG. Genetic interaction between NRAS and BRAF mutations and PTEN/MMAC1 inactivation in melanoma. J Invest Dermatol 2004;122:337–41.[CrossRef][Medline]
  21. Minamoto T, Mai M, Ronai Z. K-ras mutation: early detection in molecular diagnosis and risk assessment of colorectal, pancreas, and lung cancers: a review. Cancer Detect Prev 2000;24:1–12.[Medline]
  22. Feng Z, Hu W, Chen JX, et al. Preferential DNA damage and poor repair determine ras gene mutational hotspot in human cancer. J Natl Cancer Inst 2002;94:1527–36.[Abstract/Free Full Text]
  23. Bai F, Nakanishi Y, Takayama K, et al. Codon 64 of K-ras gene mutation pattern in hepatocellular carcinomas induced by bleomycin and 1-nitropyrene in A/J mice. Teratog Carcinog Mutagen 2003;Suppl 1:161–70.[Medline]
  24. Lu Y, Lian H, Sharma P, et al. Disruption of the Cockayne syndrome B gene impairs spontaneous tumorigenesis in cancer-predisposed Ink4a/ARF knockout mice. Mol Cell Biol 2001;21:1810–8.[Abstract/Free Full Text]
  25. Stout GJ, Westdijk D, Calkhoven DM, et al. Epidermal transit of replication-arrested, undifferentiated keratinocytes in UV-exposed XPC mice: an alternative to in situ apoptosis. Proc Natl Acad Sci U S A 2005;102:18980–5.[Abstract/Free Full Text]
  26. Yamazaki F, Okamoto H, Miyauchi-Hashimoto H, et al. XPA gene-deficient, SCF-transgenic mice with epidermal melanin are resistant to UV-induced carcinogenesis. J Invest Dermatol 2004;123:220–8.[CrossRef][Medline]
  27. Hanneman KK, Cooper KD, Baron ED. Ultraviolet immunosuppression: mechanisms and consequences. Dermatol Clin 2006;24:19–25.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yang, G.
Right arrow Articles by Tsao, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yang, G.
Right arrow Articles by Tsao, H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online