Differential gene expression in premalignant human epidermis revealed by cluster analysis of serial analysis of gene expression (SAGE) libraries

Mechanistic attributes of S100A7 (psoriasin) in resistance of anoikis resulting tumor progression in squamous cell carcinoma of the oral cavity

Background

Squamous cell carcinoma of the oral cavity (SCCOC) is the dominant origin of cancer associated mortality. Previous findings by our study reported that acquisition of anoikis resistance has a significant role in tumor progression of oral cavity. Several genes were over-expressed in anoikis-resistant cells under detached conditions which we confirmed earlier by microarray. Normal oral squamous epithelia grow adherent to a basement membrane, and when detached from the extracellular matrix, undergoes programmed cell death. The acquisition of anoikis-resistance is crucial phenomena in oral tumor advancement. In the current study, we have identified S100A7 expression as contributing factor for anoikis resistance and tumorigenicity in human oral cancer cells. Further, we have explored that elevated S100A7 expression in anoikis-sensitive oral keratinocytes and cancer cells reshape them more resistant to anoikis and apoptosis inducers via activation of cellular intrinsic and extrinsic avenue.

Methods

A subset of human cancer cell lines TU167, JMAR, JMARC39, JMARC42 and MDA-MB-468 were utilized for the generation of resistant stable cell lines. Further, immunohistochemistry, western blot and immunoprecipitation, assays of apoptosis, soft agar assay, orthotopic animal model and signaling elucidation were performed to establish our hypothesis.

Results

S100A7 gene is found to be responsible for anoikis resistance and tumorigenicity in human oral cancer cells. We have observed up-regulation of S100A7 in anoikis resistant cell lines, orthotropic model and patients samples with head and neck cancer. It is also noticed that secretion of S100A7 protein in conditioned medium by anoikis resistant head & neck cancer cell and in saliva of head and neck cancer patients. Up-regulation of S100A7 expression has triggered enhanced tumorigenicity and anchorage-independent growth of cancer cells through Akt phosphorylation leading to development of aniokis resistance in head and neck cancer cells.

Conclusions

These data have led us to conclude that S100A7 is the major contributing factor in mediating anoikis-resistance of oral cancer cells and local tumor progression, and S100A7 might be useful as diagnostic marker for early detection of primary and recurrent squamous cell cancer.

Identification of a gene signature of a pre-transformation process by senescence evasion in normal human epidermal keratinocytes

BACKGROUND

Epidemiological data show that the incidence of carcinomas in humans is highly dependent on age. However, the initial steps of the age-related molecular oncogenic processes by which the switch towards the neoplastic state occurs remain poorly understood, mostly due to the absence of powerful models. In a previous study, we showed that normal human epidermal keratinocytes (NHEKs) spontaneously and systematically escape from senescence to give rise to pre-neoplastic emerging cells.

METHODS

Here, this model was used to analyze the gene expression profile associated with the early steps of age-related cell transformation. We compared the gene expression profiles of growing or senescent NHEKs to post-senescent emerging cells. Data analyses were performed by using the linear modeling features of the limma package, resulting in a two-sided t test or F-test based on moderated statistics. The p-values were adjusted for multiple testing by controlling the false discovery rate according to Benjamini Hochberg method.The common gene set resulting of differential gene expression profiles from these two comparisons revealed a post-senescence neoplastic emergence (PSNE) gene signature of 286 genes.

RESULTS

About half of these genes were already reported as involved in cancer or premalignant skin diseases. However, bioinformatics analyses did not highlight inside this signature canonical cancer pathways but metabolic pathways, including in first line the metabolism of xenobiotics by cytochrome P450. In order to validate the relevance of this signature as a signature of pretransformation by senescence evasion, we invalidated two components of the metabolism of xenobiotics by cytochrome P450, AKR1C2 and AKR1C3. When performed at the beginning of the senescence plateau, this invalidation did not alter the senescent state itself but significantly decreased the frequency of PSNE. Conversely, overexpression of AKR1C2 but not AKR1C3 increased the frequency of PSNE.

CONCLUSIONS

To our knowledge, this study is the first to identify reprogrammation of metabolic pathways in normal keratinocytes as a potential determinant of the switch from senescence to pre-transformation.

A large-scale multi-technique approach identifies forty-nine new players of keratinocyte terminal differentiation in human epidermis

At the latest stage of terminal differentiation in the epidermis, granular keratinocytes (GKs) undergo cornification, a programmed cell death required for the establishment of a functional skin barrier. A complex genetic regulatory network orchestrates the underlying biochemical modifications, but very few transcription factors specific to this programme have been identified to date. Here, we describe a large-scale, multi-technique approach performed on cells purified from normal human epidermis, primarily focusing on the identification of regulators. We combined data from microarray analysis of cell fractions enriched in GKs or basal keratinocytes, from an expressed sequence tag (EST) library built from GKs and from an in silico promoter analysis of 52 differentiation-associated genes. Among 3576 genes potentially expressed in GK, 298 candidates were selected, and half were directly profiled for the first time in the different layers of the epidermis by quantitative real-time PCR. Forty-nine genes upregulated during terminal differentiation, associated with numerous function of GK including lipid synthesis and secretion, were identified. Of 94 transcription factors detected, 37 were found to be either positively or negatively regulated, suggesting their involvement as regulators of gene expression in the GKs. These results largely extend the number of genes known as involved in the latest step of the terminal differentiation of human epidermis as well as the number of transcription factors known to control the expression of these genes.

Microarray analysis of cutaneous squamous cell carcinomas reveals enhanced expression of epidermal differentiation complex genes

Gene expression profiles were determined for 12 cutaneous squamous cell carcinomas (SCC) removed from sun-exposed sites on nonimmunosuppressed patients. Gene expression in each SCC was compared to that in sun-exposed skin from the same patient using the Affymetrix HGU133 2.0 PlusGeneChip. We identified 440 genes with increased expression in SCC and 738 with decreased expression; overall we identified a large number of small changes in gene expression rather than a few marked changes that distinguished SCC from sun-exposed skin. Analyzing this robust data set according to biofunctional pathways using DAVID, transcriptional control elements using oPOSSUM, and chromosomal location using GSEA suggested genetic and epigenetic mechanisms of gene expression regulation in SCC. Some altered patterns of gene expression in SCC were consistent with regulation of spatially separated genes by a number of developmentally important transcription factors (forkhead, HMG, and homeo factors) that negatively regulated gene expression and to a few factors that positively regulated expression (Creb-1, NFkappaB, RelA, and Sp-1). We also found that coordinately enhanced expression of epidermal differentiation complex genes on chromosome 1q21 was a hallmark of SCC. A novel finding in our study was enhanced expression of keratin 13 in SCC, a result validated by immunohistochemical staining of an SCC tumor tissue array.

Expression of the vanin gene family in normal and inflamed human skin: induction by proinflammatory cytokines

The vanin gene family encodes secreted and membrane-bound ectoenzymes that convert pantetheine into pantothenic acid and cysteamine. Recent studies in a mouse colitis model indicated that vanin-1 has proinflammatory activity and suggest that pantetheinases are potential therapeutic targets in inflammatory diseases. In a microarray analysis of epidermal gene expression of psoriasis and atopic dermatitis lesions, we identified vanin-3 as the gene showing the highest differential expression of all annotated genes that we studied (19-fold upregulation in psoriasis). Quantitative real-time PCR analysis confirmed the microarray data on vanin-3 and showed similar induction of vanin-1, but not of vanin-2, in psoriatic epidermis. Immunohistochemistry showed that vanin-3 is expressed in the differentiated epidermal layers. Using submerged and organotypic keratinocyte cultures, we found that vanin-1 and vanin-3 are induced at the mRNA and protein level by psoriasis-associated proinflammatory cytokines (Th17/Th1) but not by Th2 cytokines. We hypothesize that increased levels of pantetheinase activity are part of the inflammatory-regenerative epidermal differentiation program, and may contribute to the phenotype observed in psoriasis.

The biology of cystatin M/E and its cognate target proteases

Cystatin M/E is a member of a superfamily of evolutionarily-related cysteine protease inhibitors that provide regulatory and protective functions against uncontrolled proteolysis by cysteine proteases. Although most cystatins are ubiquitously expressed, high levels of cystatin M/E expression are mainly restricted to the epithelia of the skin (epidermis, hair follicles, sebaceous glands, and sweat glands) and to a few extracutaneous tissues. The identification of its physiological targets and the localization of these proteases in skin have suggested a regulatory role for cystatin M/E in epidermal differentiation. In vitro biochemical approaches as well as the use of in vivo mouse models have revealed that cystatin M/E is a key molecule in a biochemical pathway that controls skin barrier formation by the regulation of both crosslinking and desquamation of the stratum corneum. Cystatin M/E directly controls the activity of cathepsin V, cathepsin L, and legumain, thereby regulating the processing of transglutaminases. Misregulation of this pathway by unrestrained protease activity, as seen in cystatin M/E-deficient mice, leads to abnormal stratum corneum and hair follicle formation, as well as to severe disturbance of skin barrier function. Here, we review the current knowledge on cystatin M/E in skin barrier formation and its potential role as a tumor suppressor gene.

Clustering analysis of SAGE transcription profiles using a Poisson approach

To gain insights into the biological function and relevance of genes using serial analysis of gene expression (SAGE) transcription profiles, one essential method is to perform clustering analysis on genes. A successful clustering analysis depends on the use of effective distance or similarity measures. For this purpose, by considering the specific properties of SAGE technology, we modeled the SAGE data by Poisson statistics and developed two Poisson-based measures to assess similarity of gene expression profiles. By employing these two distances into a K-means clustering procedure, we further developed a software package to perform clustering analysis on SAGE data. The software implementing our Poisson-based algorithms can be downloaded from http://genome.dfci.harvard.edu/sager. Our algorithm is guaranteed to converge to a local maximum when Poisson likelihood-based measure is used. The results from simulation and experimental mouse retina data demonstrate that the Poisson-based distances are more appropriate and reliable for analyzing SAGE data compared to other commonly used distances or similarity measures.

Characterization of the global profile of genes expressed in cervical epithelium by Serial Analysis of Gene Expression (SAGE)

BACKGROUND

Serial Analysis of Gene Expression (SAGE) is a new technique that allows a detailed and profound quantitative and qualitative knowledge of gene expression profile, without previous knowledge of sequence of analyzed genes. We carried out a modification of SAGE methodology (microSAGE), useful for the analysis of limited quantities of tissue samples, on normal human cervical tissue obtained from a donor without histopathological lesions. Cervical epithelium is constituted mainly by cervical keratinocytes which are the targets of human papilloma virus (HPV), where persistent HPV infection of cervical epithelium is associated with an increase risk for developing cervical carcinomas (CC).

RESULTS

We report here a transcriptome analysis of cervical tissue by SAGE, derived from 30,418 sequenced tags that provide a wealth of information about the gene products involved in normal cervical epithelium physiology, as well as genes not previously found in uterine cervix tissue involved in the process of epidermal differentiation.

CONCLUSION

This first comprehensive and profound analysis of uterine cervix transcriptome, should be useful for the identification of genes involved in normal cervix uterine function, and candidate genes associated with cervical carcinoma.

The S100A7-c-Jun Activation Domain Binding Protein 1 Pathway Enhances Prosurvival Pathways in Breast Cancer

S100A7 is among the most highly expressed genes in preinvasive breast cancer, is a marker of poor survival when expressed in invasive disease, and promotes breast tumor progression in experimental models. To explore the mechanism of action, we examined the role of S100A7 in cell survival and found that overexpression of S100A7 in MDA-MB-231 cell lines promotes survival under conditions of anchorage-independent growth. This effect is paralleled by increased activity of nuclear factor-κB (3-fold) and phospho-Akt (4-fold), which are known to mediate prosurvival pathways. S100A7 and phospho-Akt are also correlated in breast tumors examined by immunohistochemistry (n = 142; P < 0.0001; r = 0.34). To explore the underlying mechanism, we examined the role of a putative c-Jun activation domain-binding protein 1 (Jab1)–binding domain within S100A7 using a panel of MDA-MB-231 breast cell lines stably transfected with either S100A7 or S100A7 mutated at the Jab1 domain. Structural analysis by three-dimensional protein modeling, immunoprecipitation, and yeast two-hybrid assay and functional analysis using transfected reporter gene and Western blot assays revealed that the in vitro effects of S100A7 on phospho-Akt and the nuclear factor-κB pathway are dependent on the Jab1-binding site and the interaction with Jab1. Enhanced epidermal growth factor receptor signaling was also found to correlate with the increased phospho-Akt. Furthermore, the Jab1-binding domain is also necessary for the enhanced tumorigenicity conferred by S100A7 expression in murine xenograft tumors in vivo. We conclude that the S100A7-Jab1 pathway acts to enhance survival under conditions of cellular stress, such as anoikis, which may promote progression of breast cancer.

Cancer target discovery using SAGE

Cancer is a genetic disease. Genetic events including mutations, chromosomal gains, losses and rearrangements, along with epigenetic alterations, lead to significant transcriptional changes in cancer cells. Changes in the expression of many genes associated with the onset and progression of cancer likely contribute to the cancerous phenotype. SAGE (Serial Analysis of Gene Expression) is an expression profiling method that allows for global, unbiased and quantitative characterisation of transcriptomes. The expression of thousands of genes can be analysed simultaneously without prior knowledge of their sequence, thus leading to the discovery of novel transcripts. In addition to characterising normal and malignant gene expression patterns, SAGE can be used to identify downstream targets of tumour suppressors and oncogenes and further annotate genomes. Comprehensive analyses of expression profiles using SAGE will yield many new diagnostic and prognostic markers as well as therapeutic targets in cancer.

Expression patterns of S100A7 (psoriasin) and S100A9 (calgranulin-B) in keratinocyte differentiation

S100 proteins are involved in many biological processes. S100A7 and S100A9 have been shown to be markedly upregulated both in ductal carcinoma in situ of the breast and in psoriasis. We have examined the relationship between keratinocyte differentiation and the expression of the two proteins. Using Western blot analysis and quantitative reverse transcriptase-polymerase chain reaction (RT-PCR), both S100A7 and S100A9 were shown to be induced in normal primary keratinocytes (HEKn), when differentiation was promoted by high extracellular calcium, loss of contact with extracellular matrix and confluent conditions, as previously reported for S100A7 in mammary epithelial cells. Differentiation was confirmed by using RT-PCR for the differentiation marker keratin-1. Using immunohistochemistry with monoclonal antibodies, we compared the expression of the two proteins in a spectrum of conditions of dysregulated keratinocyte differentiation. We found a strikingly similar distribution of the proteins. Their expression correlated with the degree of keratinocyte differentiation. They were both absent in undifferentiated basalioma and strongly expressed in carcinoma in situ, as well as in keratoacanthoma and differentiated squamous cell carcinoma. In normal epithelium, they were expressed in the superficial, differentiated region of the epithelium rather than in the basal region. These findings support the hypothesis that these two S100 proteins are involved in keratinocyte differentiation.

SAGE profiling of UV-induced mouse skin squamous cell carcinomas, comparison with acute UV irradiation effects

Ultraviolet (UV) irradiation is the primary environmental insult responsible for the development of most common skin cancers. To better understand the multiple molecular events that contribute to the development of UV-induced skin cancer, in a first study, serial analysis of gene expression (SAGE) was used to compare the global gene expression profiles of normal SKH-1 mice epidermis with that of UV-induced squamous cell carcinomas (SCCs) from SKH-1 mice. More than 200 genes were found to be differentially expressed in SCCs compared to normal skin (P < 0.0005 level of significance). As expected, genes related to epidermal proliferation and differentiation were deregulated in SCCs relative to normal skin. However, various novel genes, not previously associated with skin carcinogenesis, were also identified as deregulated in SCCs. Northern blot analyses on various selected genes validated the SAGE findings: caspase-14 (reduced 8.5-fold in SCCs); cathepsins D and S (reduced 3-fold and increased 11.3-fold, respectively, in SCCs); decorin, glutathione S-transferase omega-1, hypoxia-inducible factor 1 alpha, insulin-like growth factor binding protein-7, and matrix metalloproteinase-13 (increased 18-, 12-, 12-, 18.3-, and 11-folds, respectively, in SCCs). Chemokine (C-C motif), ligand 27 (CCL27), which was found downregulated 12.7-fold in SCCs by SAGE, was also observed to be strongly downregulated 6-24 h after a single and multiple UV treatments. In a second independent study we compared the expression profile of UV-irradiated versus sham-treated SKH-1 epidermis. Interestingly, numerous genes determined to be deregulated 8 h after a single UV dose were also deregulated in SCCs. For instance, genes whose expression was upregulated both after acute UV-treated skin and SCCs included keratins 6 and 16, small proline-rich proteins, and S100 calcium binding protein A9. Studies like those described here do not only provide insights into genes and pathways involved in skin carcinogenesis but also allow us to identify early UV irradiation deregulated surrogate biomarkers of potential use in chemoprevention studies.

Microarray and serial analysis of gene expression analyses identify known and novel transcripts overexpressed in hematopoietic stem cells

The human CD34(+)/CD38(-)/Lin(-) cell subset, comprising approximately 1-10% of the CD34(+) cell population, contains few of the less primitive hematopoietic (lineage-committed) progenitor cells (HPCs) but most of the primitive in vivo engrafting (lympho-)hematopoietic stem cells (HSCs). We analyzed gene expression in CD34(+)/CD38(-)/Lin(-) cell populations isolated from normal human adult donor bone marrow, neonatal placental/umbilical cord blood, and mobilized adult donor peripheral blood stem-progenitor cells. As measured by Affymetrix microarrays, 4746 genes were expressed in CD34(+)/CD38(-)/Lin(-) cells from all three tissues. We also determined the transcriptomes of the stem cell-depleted, HPC-enriched CD34(+)/[CD38/Lin](++) cell population from each tissue. Comparison of CD34(+)/CD38(-)/Lin(-) (HSC-enriched) versus CD34(+)/[CD38/Lin](++) (HPC-enriched, HSC-depleted) cells from each tissue yielded 81 genes overrepresented and 90 genes underrepresented, common to all three of the CD34(+)/CD38(-)/Lin(-) cell populations. These transcripts, which are selectively expressed in HSCs from all three tissues, include a number of known genes (e.g., transcription factors, receptors, and signaling molecules) that might play roles in key functions (e.g., survival, self-renewal, differentiation, and/or migration/adhesion) of human HSCs. Many genes/transcripts of unknown function were also detected by microarray analysis. Serial analysis of gene expression of the bone marrow HSC and HPC populations confirmed expression of most of the overrepresented transcripts for which reliable serial analysis of gene expression tags were detected and additionally suggested that current microarrays do not detect as many as 30% of the transcripts expressed in HSCs, including a number of previously unknown transcripts. This work is a step toward full definition of the transcriptome of normal human HSCs and may identify new genes involved in leukemogenesis and cancer stem cells.

Clustering analysis of SAGE data using a Poisson approach

Two Poisson-based distances were developed for SAGE data; their application to simulated and experimental mouse retina data show that they are more appropriate and reliable for analyzing SAGE data than other commonly used distances or similarity measures.

Serial analysis of gene expression (SAGE) data have been poorly exploited by clustering analysis owing to the lack of appropriate statistical methods that consider their specific properties. We modeled SAGE data by Poisson statistics and developed two Poisson-based distances. Their application to simulated and experimental mouse retina data show that the Poisson-based distances are more appropriate and reliable for analyzing SAGE data compared to other commonly used distances or similarity measures such as Pearson correlation or Euclidean distance.

Serial Analysis of Gene Expression: Applications in Human Studies

Serial analysis of gene expression (SAGE) is a powerful tool, which provides quantitative and comprehensive expression profile of genes in a given cell population. It works by isolating short fragments of genetic information from the expressed genes that are present in the cell being studied. These short sequences, called SAGE tags, are linked together for efficient sequencing. The frequency of each SAGE tag in the cloned multimers directly reflects the transcript abundance. Therefore, SAGE results in an accurate picture of gene expression at both the qualitative and the quantitative levels. It does not require a hybridization probe for each transcript and allows new genes to be discovered. This technique has been applied widely in human studies and various SAGE tags/SAGE libraries have been generated from different cells/tissues such as dendritic cells, lung fibroblast cells, oocytes, thyroid tissue, B-cell lymphoma, cultured keratinocytes, muscles, brain tissues, sciatic nerve, cultured Schwann cells, cord blood-derived mast cells, retina, macula, retinal pigment epithelial cells, skin cells, and so forth. In this review we present the updated information on the applications of SAGE technology mainly to human studies.

DNA expression analysis: serial analysis of gene expression, microarrays and kidney disease

PURPOSE OF REVIEW

Expression profiling using serial analysis of gene expression and microarray technologies allows global description of expressed genes present in biological systems. Although relatively new technologies, each having been developed in the mid-1990s, both have become established and widely used tools for identification of gene networks and gene function.

RECENT FINDINGS

This review highlights DNA expression analyses published in 2002, emphasizing primarily serial analysis of gene expression and microarray technologies. We focus on the applicability of DNA expression analysis to renal disease, especially as some investigators have developed custom serial analysis of gene expression kidney libraries and kidney disease-specific 'designer chip' microarrays. Data analysis techniques and statistics are also discussed, since the challenge is generation of accurate messenger RNA profiles and interpretation of data in a manner that is both coherent and reproducible.

SUMMARY

Because kidney disease pathophysiology is complex, expression analysis can identify candidate nephropathy pathogenesis genes and gene networks, which eventually could become targets for therapeutic intervention.

Serial analysis of gene expression (SAGE): advances, analysis and applications to pigment cell research

As cells progress from normal to diseased states, they may undergo a series of gene expression changes. Advances in molecular biology allow us to examine a host of these changes at once, in a high throughput fashion. Serial analysis of gene expression (SAGE) allows for the expression profiling of the complete transcriptome of a given cell, and has the potential for identifying novel genes as well as those in low abundance. In this review, we will outline the technique, how one analyzes the massive amounts of data generated, and describe pigment cell libraries currently in the making.

Profile of gene expression induced by the tumour promotor TPA in murine epithelial cells

Malignant transformation of mouse skin by chemical carcinogens and tumour promoters, such as the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA), is a multistage process that leads to squamous cell carcinoma (SCC) formation. In an effort to identify tumour-associated genes, we studied the influence of short-term TPA-treatment on the gene expression profile of murine skin. A comprehensive microarray with some 5,000 murine gene specific cDNA fragments was established and hybridised with pooled RNA derived from control and TPA-treated dorsal skin samples. Of these genes, 54 were up- and 35 were down-regulated upon TPA application. Additionally, we performed suppression subtractive hybridisation (SSH) with respective RNA pools to generate and analyse a cDNA library enriched for TPA-inducible genes. Expression data of selected genes were confirmed by quantitative real-time PCR and Northern blot analysis. Comparison of microarray and SSH data revealed that 26% of up-regulated genes identified by expression profiling matched with those present in the SSH library. Besides numerous known genes, we identified a large set of unknown cDNAs that represent previously unrecognised TPA-regulated genes in murine skin with potential function in tumour promotion. Additionally, some TPA-induced genes, such as Sprr1A, Saa3, JunB, Il4ralpha, Gp38, RalGDS and Slpi exhibit high basal level in advanced stages of skin carcinogenesis, suggesting that at least a subgroup of the identified TPA-regulated genes may contribute to tumour progression and metastasis.

Psoriasin (S100A7) expression is altered during skin tumorigenesis

BACKGROUND

Psoriasin (S100A7) expression has previously been associated with psoriasiform hyperplasia as well as with tumor progression in breast cancer. Its expression profile for different stages of skin lesions is unknown. The aim of this study was to determine the relationship between psoriasin (S100A7) and tumor progression in skin.

METHODS

Psoriasin was assessed by immunohistochemistry and levels of expression determined by semi-quantitative scoring in skin biopsies from 50 patients. The cohort included normal skin, actinic keratosis, squamous carcinoma in-situ, invasive squamous cell carcinoma, and basal cell carcinoma.

RESULTS

In normal skin, psoriasin was rarely detected in epidermis but was expressed in underlying adnexae. In abnormal epidermis psoriasin was frequently expressed in abnormal keratinocytes in actinic keratosis, in-situ and invasive squamous cell carcinoma, but was rarely observed in the basal epidermal layer or in superficial or invasive basal cell carcinoma. The highest levels of expression were seen within squamous carcinoma in-situ. Significantly reduced levels of expression were observed in both unmatched (p = 0.0001) and matched (p < 0.004) invasive squamous cell carcinoma. Psoriasin expression within abnormal squamous lesions correlated with mitotic count (r = 0.54, p = 0.0036), however no significant relation was found with the intensity of dermal inflammatory cell infiltrates assessed within each pathology.

CONCLUSION

These results suggest that altered psoriasin expression occurs in abnormal epidermis and that downregulation may be related to the onset of invasion in squamous cell carcinoma in skin.