A role for sunlight in skin cancer: UV-induced p53 mutations in squamous cell carcinoma

Sunlight is a carcinogen to which everyone is exposed. Its UV component is the major epidemiologic risk factor for squamous cell carcinoma of the skin. Of the multiple steps in tumor progression, those that are sunlight-related would be revealed if they contained mutations specific to UV. In a series of New England and Swedish patients, we find that 14/24 (58%) of invasive squamous cell carcinomas of the skin contain mutations in the p53 tumor suppressor gene, each altering the amino acid sequence. Involvement of UV light in these p53 mutations is indicated by the presence in three of the tumors of a CC----TT double-base change, which is only known to be induced by UV. UV is also implicated by a UV-like occurrence of mutations exclusively at dipyrimidine sites, including a high frequency of C----T substitutions. p53 mutations in internal malignancies do not show these UV-specific mutations. The dipyrimidine specificity also implicates dipyrimidine photoproducts containing cytosine as oncogenic photoproducts. We believe these results identify a carcinogen-related step in a gene involved in the subsequent human cancer.

Sunburn and p53 in the onset of skin cancer

Squamous cell carcinoma of the skin (SCC) can progress by stages: sun-damaged epidermis, with individual disordered keratinocytes; actinic keratosis (AK), spontaneously regressing keratinized patches having aberrant cell differentiation and proliferation; carcinoma in situ; SCC and metastasis. To understand how sunlight acts as a carcinogen, we determined the stage at which sunlight mutates the p53 tumour-suppressor gene and identified a function for p53 in skin. The p53 mutations induced by ultraviolet radiation and found in > 90% of human SCCs were present in AKs. Inactivating p53 in mouse skin reduced the appearance of sunburn cells, apoptotic keratinocytes generated by overexposure to ultraviolet. Skin thus appears to possess a p53-dependent 'guardian-of-the-tissue' response to DNA damage which aborts precancerous cells. If this response is reduced in a single cell by a prior p53 mutation, sunburn can select for clonal expansion of the p53-mutated cell into the AK. Sunlight can act twice: as tumour initiator and tumour promoter.

The role of the human homologue of Drosophila patched in sporadic basal cell carcinomas

Basal cell carcinoma (BCC) is the most common cancer in humans. The majority of sporadic BCCs have allele loss on chromosome 9q22 implying that inactivation of a tumour suppressor in this region is an important step in BCC formation. The gene for nevoid basal cell carcinoma syndrome (NBCCS), an autosomal dominant disorder characterized by multiple BCCs, maps to the same region and is presumed to be the tumour suppressor inactivated at this site. NBCCS has been identified recently and encodes a protein with strong homology to the Drosophila segment polarity gene, patched. Analysis of Drosophila mutants indicates that patched interacts with the hedgehog signalling pathway, repressing the expression of various hedgehog target genes including wingless, decapentaplegic and patched itself. Using single strand conformational polymorphism (SSCP) to screen human patched in 37 sporadic BCCs, we detected mutations in one-third of the tumours. Direct sequencing of two BCCs without SSCP variants revealed mutations in those tumours as well suggesting that inactivation of patched is probably a necessary step in BCC development. Northern blots and RNA in situ hybridization showed that patched is expressed at high levels in tumour cells but not normal skin suggesting that mutational inactivation of the gene leads to overexpression of mutant transcript owing to failure of a negative feedback mechanism.

Frequent clones of p53-mutated keratinocytes in normal human skin

The multiple genetic hit model of cancer predicts that normal individuals should have stable populations of cancer-prone, but noncancerous, mutant cells awaiting further genetic hits. We report that whole-mount preparations of human skin contain clonal patches of p53-mutated keratinocytes, arising from the dermal-epidermal junction and from hair follicles. These clones, 60-3000 cells in size, are present at frequencies exceeding 40 cells per cm2 and together involve as much as 4% of the epidermis. In sun-exposed skin, clones are both more frequent and larger than in sun-shielded skin. We conclude that, in addition to being a tumorigenic mutagen, sunlight acts as a tumor promoter by favoring the clonal expansion of p53-mutated cells. These combined actions of sunlight result in normal individuals carrying a substantial burden of keratinocytes predisposed to cancer.

Sunlight and the onset of skin cancer

The photons of sunlight precipitate a series of genetic events in skin leading to cancer. These events involve somatic mutations as well as inherited alleles. Competition between cell populations ensues, as a single mutated cell expands into a clone. Thus cancer involves both a single-cell problem and a many-cell problem; in skin cancer, sunlight appears to drive both.

Distribution of UV light-induced damage in a defined sequence of human DNA: detection of alkaline-sensitive lesions at pyrimidine nucleoside-cytidine sequences

The distribution of UV light-induced damage to the highly reiterated alpha sequence of human DNA was investigated. The results show that the distribution of UV light-induced cyclobutane dimers within a defined sequence is similar whether the DNA is exposed to UV light as part of the chromosome of intact cells or as naked DNA. However, the cellular environment shields the nuclear DNA, resulting in about 50% decrease in apparent dose. A new type of UV photodamage was detected. Treatment of UV light-irradiated DNA with hot alkali results in strand breaks at positions of cytidine located 3' to pyrimidine nucleosides. The chemical nature and biological significance of the pyrimidine nucleoside-cytidine lesion is discussed.

Photoproduct frequency is not the major determinant of UV base substitution hot spots or cold spots in human cells

The role of UV radiation-induced photoproducts in initiating base substitution mutations in human cells was examined by measuring photoproduct frequency distributions and mutations in a supF tRNA gene on a shuttle vector plasmid transfected into DNA repair-deficient cells (xeroderma pigmentosum, complementation group A) and into normal cells. Frequencies of cyclobutane dimers and pyrimidine-pyrimidone (6-4) photoproducts varied by as much as 80-fold at different dipyrimidine sites within the gene. All transition mutations occurred at dipyrimidine sites, predominantly at cytosine, with a 17-fold variation in mutation frequency between different sites. Removal of greater than 99% of the cyclobutane dimers by in vitro photoreactivation before transfection reduced the mutation frequency while preserving the mutation distribution, indicating that (i) cytosine-containing cyclobutane dimers were the major mutagenic lesions at these sites and (ii) cytosine-containing non-cyclobutane dimer photoproducts were also mutagenic lesions. However, at individual dipyrimidine sites neither the frequency of cyclobutane dimers nor the frequency of pyrimidine-pyrimidone (6-4) photoproducts correlated with the mutation frequency, even in the absence of excision repair. Mutation hot spots occurred at sites with low or high frequency of photoproduct formation and mutation cold spots occurred at sites with many photoproducts. These results suggest that although photoproducts are required for UV mutagenesis, the prominence of most mutation hot spots and cold spots is primarily determined by DNA structural features rather than by the frequency of DNA photoproducts.

Transgenic expression of survivin in keratinocytes counteracts UVB-induced apoptosis and cooperates with loss of p53

The inhibitor of apoptosis protein survivin has been implicated in both cell cycle control and apoptosis resistance. To discriminate between these different roles, we used transgenic expression of survivin in the skin as a model for cell proliferation, differentiation, and apoptosis. Transgenic mice expressing survivin under the control of a keratin-14 promoter developed normally, without histologic abnormalities of the skin or hair, epidermal hyperplasia, or developmental abnormalities of basal or suprabasal epidermis. Keratinocyte proliferation assessed under basal conditions, or after ultraviolet-B (UVB) irradiation, or phorbol ester stimulation was unchanged in survivin transgenic mice. In contrast, survivin expression inhibited UVB-induced apoptosis in vitro and in vivo (i.e., sunburn cell formation), whereas it did not affect Fas-induced cell death. When crossed with p53 knockout mice, transgenic expression of survivin in a p53(+/-) background substituted for the loss of a second p53 allele and further inhibited UVB-induced apoptosis. These data provide the first in vivo evidence that survivin inhibits apoptosis and suggest that this pathway may oppose the elimination of cancerous cells by p53.

Relationship between sunlight exposure and a key genetic alteration in basal cell carcinoma

BACKGROUND

Basal cell carcinoma (BCC) of the skin is the most common cancer in humans. Epidemiologic studies implicate sunlight exposure as one risk factor, but the limited association between BCCs and UVB radiation (i.e., UV radiation of a wavelength of 280-320 nm) suggests that additional factors must be involved. At the molecular level, not much is known about the role of specific environmental agents in the pathogenesis of BCCs. Point mutations of the types produced by UVB radiation are seen in the p53 gene (also known as TP53; chromosome 17p) of 40%-56% of BCCs. Loss of heterozygosity (LOH) on chromosome 9q22, however, is the most frequent genetic alteration in these tumors, and its causative agent is unknown.

PURPOSE

We investigated whether the genetic alteration in chromosome 9 is common to all clinical subtypes of BCCs and whether inactivation of this putative tumor suppressor is related to sunlight exposure. The presence of UVB radiation-related point mutations in the p53 gene was used as an internal control for sunlight exposure to the precursor cells.

METHODS

Tumor and blood samples were obtained from skin cancer patients by a surgeon who used Mohs' micrographic surgical technique. Clinical information on each tumor included location, size, histologic, subtype and whether it was primary or recurrent and sporadic or hereditary. Sixty BCCs from 58 patients were evaluated for LOH with 12 polymorphic markers that span chromosome 9. A subset of 18 tumors was evaluated for point mutations in exons 2-11 of the p53 gene, and a subset of 26 tumors was evaluated for LOH by use of a polymorphism in exon 4 of the p53 gene. Associations between tumor characteristics and molecular alterations were tested by a two-tailed chi-squared analysis or a two-tailed Fisher's exact test, depending on sample size.

RESULTS

In a clinically diverse series of 47 informative tumors, 32 (68%) showed LOH for chromosome 9q markers, irrespective of histologic characteristics or clinical behavior. Forty-four (94%) of the 47 tumors were from sun-exposed areas of the body, defined as the head and neck in both sexes, shoulders or chest in males, and legs in females. No association was found between chromosome 9q LOH and sunlight exposure, as assessed by either the location of tumors on the body or the presence of UVB radiation-related p53 mutations. Of note, there was a striking difference between the frequency of LOH on chromosome 17p (two [12.5%] of 16 informative tumors) and on chromosome 9q (32 [68%] of 47 informative tumors; P < .001).

CONCLUSIONS

Inactivation of a gene on chromosome 9q22 may be a necessary event for basal cell carcinogenesis. The pathogenesis of mutations in this gene may involve factors other than sunlight in a large proportion of tumors.

IMPLICATIONS

The limited association between sunlight exposure and BCC incidence may reflect an etiologic contribution of additional environmental agents.

Escaping the stem cell compartment: sustained UVB exposure allows p53-mutant keratinocytes to colonize adjacent epidermal proliferating units without incurring additional mutations

Once mutated, a single cell must expand into a clone before becoming significant for carcinogenesis. The forces driving clonal expansion and the obstacles that must be overcome are poorly understood. In a genetic mechanism, acquiring a second mutation conferring a proliferative advantage would enable the cell to expand autonomously. If carcinogen exposure instead induced a physiological change, clonal expansion would require the carcinogen's continued presence. To determine which is the case, we studied microscopic clones of keratinocytes mutated in the p53 tumor suppressor gene. Carcinogen exposure was controlled by irradiating mice with 280-320 nm UV radiation (UVB), sunlight's principal carcinogenic component; expansion of mutant clones was observed in epidermal sheets. p53-mutant clones grew only during chronic UVB exposure. Therefore, clonal expansion was not triggered by a proliferative mutation but was instead continually driven by UVB. Unexpectedly, the clone size distribution showed periodicity with maxima at estimated intervals of 16 +/- 6 cells, the size of the epidermal proliferating unit in murine dorsal skin. In the absence of UVB, rare "imprisoned clones" increased in cell number without increasing in area. We conclude that: stem cell compartments act as physical barriers to clonal expansion of a p53-mutant keratinocyte; a rate-limiting step in clonal expansion is the colonization of an adjacent compartment; and sustained UVB enables the p53-mutant keratinocyte to colonize without incurring an additional mutation.

Ultraviolet radiation induced signature mutations in photocarcinogenesis

The photons of sunlight begin a series of genetic events in skin leading to cancer. UV signature mutations provide an alternative to inherited mutations as a way of identifying genes that are involved in cancer development. They augment epidemiologic and clinical data by serving as molecular evidence for the role of UV radiation in skin carcinogenesis. Signature mutations are present in TP53 and PTCH, two tumor suppressor genes responsible for non-melanoma skin cancer. We review evidence that clones of TP53-mutated cells are present in normal human and murine epidermis exposed to UVB and conclude that, in addition to being a tumorigenic mutagen, sunlight acts as a tumor promoter by favoring the clonal expansion of TP53 mutated cells. These combined actions of sunlight result in normal individuals' carrying a substantial burden of keratinocytes predisposed to cancer. Thus cancer involves both a single-cell problem and a multi-cell problem; in skin cancer, sunlight appears to drive both.

Local recurrence versus new primary: clinical analysis of 82 breast relapses and potential applications for genetic fingerprinting

PURPOSE

The purpose of this study was to perform a detailed clinical pathological analysis of breast relapses in patients treated with conservative surgery and radiation therapy in an effort to classify those relapses as true local recurrences or second primary tumors, and to assess the prognostic and therapeutic implications of such a classification system.

METHODS AND MATERIALS

Of 990 patients treated with conservative surgery and radiation therapy at our facilities prior to December 1987, 82 patients have experienced a relapse in the conservatively treated breast as the primary site of failure. Patients were classified as having new primary tumors if they fulfilled any one of the following criteria: a) breast relapse occurring at a site distinctly removed from the original tumor; b) histology of the breast relapse compared with the original tumor consistent with a new primary; or c) DNA flow cytometry converting from an aneuploid primary to a diploid relapse.

RESULTS

As of 2/92, with a median follow-up of 5.4 years from the time of breast relapse, the overall 5-year survival rate following breast relapse was 55%. Forty-seven patients were classified as true recurrences and 33 patients were classified as new primaries. Patients classified as true recurrences had a shorter median time to breast relapse than patients classified as new primaries (3.16 years vs. 5.42 years, p < .05) and an inferior post breast recurrence survival rate compared to patients classified as new primaries (36% vs. 89%, p < .05). Residual disease outside of the recurrent tumor bed was also noted to be more frequent in patients classified as true recurrences compared to patients classified as new primaries (48% vs. 16%, p < .05).

CONCLUSION

Based on the clinical and pathological criteria outlined, it appears that a significant portion of patients experiencing a relapse in the conservatively treated breast may have new primary tumors as opposed to true local relapses. Distinction between a true recurrence and a new primary tumor may have significant prognostic implications. Uncertainties associated with the clinical and pathological criteria are presented and further investigations with genetic fingerprinting techniques to establish the clonality of breast relapses are presented and discussed.

The C-C (6-4) UV photoproduct is mutagenic in Escherichia coli

Mutation induced by ultraviolet light is predominantly targeted by UV photoproducts. Two primary candidates for the premutagenic lesion are the cyclobutane pyrimidine dimer and the less frequent (by a factor of 10) pyrimidine-pyrimidone (6-4) photoproduct. Methylation of the 3'-cytosine in the sequence 5' CCAGG 3' reduces the yield of (6-4) lesions, but not of cyclobutane dimers, at these sites. By taking advantage of mutants deficient in cytosine methylation, we show here that at the three sites in the lacI gene of Escherichia coli having this sequence, the specific increase in the formation of the (6-4) photoproducts is accompanied by a concomitant increase in mutation. At each site, a G X C to A X T transition results in an amber mutation. In the unmethylated state, these sites become among the most frequent nonsense mutations recovered. We conclude that the (6-4) photoproduct constitutes a major premutagenic lesion in E. coli.

Rapid repair kinetics of pyrimidine(6-4)pyrimidone photoproducts in human cells are due to excision rather than conformational change

UV-induced pyrimidine(6-4)pyrimidone photoproducts in DNA of mammalian cells are apparently repaired much more rapidly than cyclobutane dimers. Since only immunological assays for (6-4) photoproducts have been sensitive enough for repair measurements, it was possible that these apparently rapid repair kinetics reflected a change in physical conformation of antibody-binding sites, resulting in epitope loss rather than excision. To discriminate between these possibilities, we developed a procedure to photochemically convert (6-4) photoproducts to single-strand breaks in UV-irradiated DNA with a background low enough to permit repair measurements. Analysis of a specific DNA sequence indicated that photoinduced alkali-labile sites (PALS) were induced with the same site-specificity as (6-4) photoproducts. Normal human and xeroderma pigmentosum (XP) variant cells rapidly excised (6-4) photoproducts measured as PALS, but little repair was seen in cells from XP complementation group A. These repair kinetics corresponded to those determined in the same samples by radioimmunoassay of (6-4) photoproducts. Thus we conclude that the rapid repair of (6-4) photoproducts observed in UV-irradiated human cells is not the result of a conformational change resulting in epitope loss, but reflects excision of this photoproduct from DNA.

Roles of the transcription factor p53 in keratinocyte carcinomas

The transcription factor p53 is mutated in most keratinocyte carcinomas (nonmelanoma skin cancers). In these tumours, the gene bears the trace of its mutagen, sunlight. Sunlight-induced p53 mutations are also seen in skin precancers and even sun-exposed skin, which harbours thousands of p53-mutant keratinocyte clones. Normal p53 is upregulated by sunlight exposure, after which it acts as a tumour suppressor in several ways: increasing DNA repair, arresting the cell cycle and inducing apoptosis of badly damaged keratinocytes. This UV-induced upregulation has been used as an assay for assessing the effectiveness of sunscreens. Once mutated, however, p53 renders cells apoptosis-resistant and therefore less sensitive to sunlight overexposure than normal cells. This reversal of roles drives clonal expansion of precancerous keratinocytes.

Excision repair at individual bases of the Escherichia coli lacI gene: relation to mutation hot spots and transcription coupling activity

To determine whether base-to-base variations in the rate of excision repair influence the distribution of mutations, we have developed a method to measure UV photoproducts at individual nucleotides in the Escherichia coli chromosome. Specific gene fragments are 3' end-labeled using a sequence-specific oligonucleotide to direct the site of labeling, and photoproducts are identified by enzymatic incision. On the nontranscribed strand of the E. coli lacI gene, the cyclobutane pyrimidine dimer frequency was 2- to 8-fold higher in chromosomal DNA than in a cloned DNA fragment. The chromosomal lesion frequency corresponded to the frequency of UV-induced mutations at mutation hot spots reported in the literature. Only 0-30% of cyclobutane dimers at various sites on this strand were excised in 20 min. In contrast, repair on the transcribed strand was 80-90% complete in 20 min. However, the transcribed strand contained an excision repair "slow spot" at the site of its single mutation hot spot: At this site, no repair occurred for the first 10 min, after which repair proceeded more slowly than typical of that strand. In an mfd strain, deficient in a factor that couples repair to transcription in cell extracts, the excision rate at individual nucleotides on the transcribed strand was minimal at most sites for at least 30 min. Wild-type E. coli's bias for producing mutations at photoproducts on the nontranscribed strand, reported to require the mfd gene, therefore appears to be due to an excision repair system specific for the transcribed strand of chromosomal DNA.

Induction of cyclin-dependent kinase inhibitors and G(1) prolongation by the chemopreventive agent N-acetylcysteine

Cyclin-dependent kinase (cdk) inhibitors, such as p16(INK4a) and p21(WAF1/CIP1), often inhibit G(1) cyclin kinases and result in G(1) arrest. It has been suggested that p21(WAF1/CIP1) may also play a role in other chemopreventive activities such as DNA repair, slowdown of DNA replication and induction of cellular differentiation. In this report we demonstrate that the antioxidant N-acetylcysteine (NAC), a well-known chemopreventive agent, induces p16(INK4a) and p21(WAF1/CIP1) gene expression and prolongs cell-cycle transition through G(1) phase. A portion of the G(1) arrest by NAC is governed by p16(INK4a); it is independent of p53. NAC's usual mechanism of increasing intracellular glutathione level is not required for the G(1) arrest. An antioxidant whose action is limited to scavenging radicals, Trolox, does not induce G(1) arrest. Taken together, these results suggest a potential novel molecular basis for chemoprevention by NAC.

Regulation of TNFalpha production and release in human and mouse keratinocytes and mouse skin after UV-B irradiation

TNFalpha is a primary cytokine responsible for inflammatory and immunosuppressive responses in skin. After UV-B irradiation of cultured human keratinocytes, we found that TNFalpha was released into the media, as monitored by ELISA, and was bound to cells, as observed by immunofluorescence microscopy. The release of TNFalpha into cell culture supernatant during the 24 h after UV-B irradiation was augmented by the addition of IL-1alpha to the cells. Further, we found this secretion was unaffected by rapamycin, and therefore independent of FRAP DNA-protein kinase mediated signal transduction. However, UV-B also induced expression of membrane-bound TNFalpha, and this was dependent on FRAP signaling. In wild type mice, TNFalpha bound to skin increased immediately after irradiation, declined at 6 h, and then rose again at 12 h before falling by 24 h. This pattern of induction was confirmed by RT-PCR of TNFalpha mRNA message in cultured epidermal cells. Induction of membrane-bound TNFalpha was also found in c-fos gene knockout mice deficient in the AP-1 transcription factor, suggesting that, although AP-1 containing c-fos signaling is required for some UV responses, AP-1 containing c-fos is not required for this TNFalpha activation. However, in homozygous p53 knockout mice the basal level of TNFalpha bound to the epidermis was greatly elevated without UV irradiation. This level declined and remained constant following irradiation. This implies that p53 directly or indirectly represses TNFalpha gene expression and that modification of p53 mRNA stability or phosphorylation of p53 protein after UV may be responsible for TNFalpha induction in the membrane. Overexpression of the immunosuppressive cytokine TNFalpha in this locale may contribute to the carcinogen-susceptibility of p53 knockout mice.

The DNA damage signal for Mdm2 regulation, Trp53 induction, and sunburn cell formation in vivo originates from actively transcribed genes

The stratum corneum and DNA repair do not completely protect keratinocytes from ultraviolet B. A third defense prevents cells with DNA photoproducts from becoming precancerous mutant cells: apoptosis of ultraviolet-damaged keratinocytes ("sunburn cells"). As signals for ultraviolet-induced apoptosis, some studies implicate DNA photoproducts in actively transcribed genes; other studies implicate non-nuclear signals. We traced and quantitated the in vivo DNA signal through several steps in the apoptosis-signaling pathway in haired mice. Homozygous inactivation of Xpa, Csb, or Xpc nucleotide excision repair genes directed the accumulation of DNA photoproducts to specific genome regions. Repair-defective Xpa-/- mice were 7-10-fold more sensitive to sunburn cell induction than wild-type mice, indicating that 86-90% of the ultraviolet B signal for keratinocyte apoptosis involved repairable photoproducts in DNA; the remainder involves unrepaired DNA lesions or nongenomic targets. Csb-/- mice, defective only in excising photoproducts from actively transcribed genes, were as sensitive as Xpa-/-, indicating that virtually all of the DNA signal originates from photoproducts in active genes. Conversely, Xpc-/- mice, defective in repairing the untranscribed majority of the genome, were as resistant to apoptosis as wild type. Sunburn cell formation requires the Trp53 tumor suppressor protein; 90-96% of the signal for its induction in vivo involved transcribed genes. Mdm2, which regulates the stability of Trp53 through degradation, was induced in vivo by low ultraviolet B doses but was suppressed at erythemal doses. DNA photoproducts in actively transcribed genes were involved in approximately 89% of the Mdm2 response.