Expression of the Embryonic Stem Cell Transcription Factor SOX2 in Human Skin
Alvaro C. Laga,Chiou-Yan Lai,Qian Zhan,Susan S. Huang,Elsa F. Velázquez,Qinghong Yang,Mei‐Yu Hsu,Gëorge F. Murphy
DOI: https://doi.org/10.2353/ajpath.2010.090495
IF: 5.77
2009-01-01
American Journal Of Pathology
Abstract:SOX2 is a gene located on chromosome 3q26.33 that encodes a transcription factor important to maintenance of embryonic neural crest stem cell pluripotency. We have identified rare SOX2-immunoreactive cells in normal human skin at or near the established stem cell niches. Three subsets of SOX2-positive cells were defined in these regions: those expressing only SOX2 and those that co-expressed SOX2 and either CK20 or microphthalmia-associated transcription factor, which are consistent with dichotomous differentiation of SOX2-expressing precursors along neuroendocrine (Merkel cell) or melanocytic lines, respectively. Examination of Merkel cell carcinomas confirmed nuclear SOX2 expression in this tumor type. In human patient melanoma, strong nuclear expression of SOX2 was noted in a subset of tumors, and the ability to detect SOX2 in lesional cells significantly correlated with primary tumor thickness in a survey cohort. To assess the potential role of SOX2 in melanoma growth, an in vivo tumorigenesis assay was used. Whereas SOX2 knockdown failed to influence proliferation of cultured melanoma cells in vitro, tumor xenografts generated with the SOX2-knockdown cell line showed significant decrease in mean tumor volume as compared with controls. In aggregate, these findings suggest that SOX2 is a novel biomarker for subpopulations of normal skin cells that reside in established stem cell niches and that might relate to Merkel cell and melanocyte ontogeny and tumorigenesis. SOX2 is a gene located on chromosome 3q26.33 that encodes a transcription factor important to maintenance of embryonic neural crest stem cell pluripotency. We have identified rare SOX2-immunoreactive cells in normal human skin at or near the established stem cell niches. Three subsets of SOX2-positive cells were defined in these regions: those expressing only SOX2 and those that co-expressed SOX2 and either CK20 or microphthalmia-associated transcription factor, which are consistent with dichotomous differentiation of SOX2-expressing precursors along neuroendocrine (Merkel cell) or melanocytic lines, respectively. Examination of Merkel cell carcinomas confirmed nuclear SOX2 expression in this tumor type. In human patient melanoma, strong nuclear expression of SOX2 was noted in a subset of tumors, and the ability to detect SOX2 in lesional cells significantly correlated with primary tumor thickness in a survey cohort. To assess the potential role of SOX2 in melanoma growth, an in vivo tumorigenesis assay was used. Whereas SOX2 knockdown failed to influence proliferation of cultured melanoma cells in vitro, tumor xenografts generated with the SOX2-knockdown cell line showed significant decrease in mean tumor volume as compared with controls. In aggregate, these findings suggest that SOX2 is a novel biomarker for subpopulations of normal skin cells that reside in established stem cell niches and that might relate to Merkel cell and melanocyte ontogeny and tumorigenesis. Biomarkers for tissue stem cell counterparts of Merkel cells and melanocytes are lacking. In related neoplasms such as melanoma, the identification of initiating/stem cells has relied in part on the observation that highly aggressive melanoma cell subsets are associated with molecular signatures that resemble those of pluripotent physiological stem cells.1Bittner M Meltzer P Chen Y Jiang Y Seftor E Hendrix M Radmacher M Simon Rm Yakhini Z Ben-Dor A Sampas N Dougherty E Wang E Marincola F Gooden C Lueders J Glatfelter A Pollock P Carpten J Gillanders E Leja D Dietrich K Beaudry C Berens M Alberts D Sondak V Molecular classification of cutaneous malignant melanoma by gene expression profiling.Nature. 2000; 406: 536-540Crossref PubMed Scopus (1655) Google Scholar, 2Hendrix MJ Seftor EA Hess AR Seftor RE Molecular plasticity of human melanoma cells.Oncogene. 2003; 22: 3070-3075Crossref PubMed Scopus (184) Google Scholar Several stem and progenitor cell-associated proteins have been detected in melanoma, such as cancer testis antigens,3Simpson AJ Caballero OL Jungbulth A Chen YT Old LJ Cancer/testis antigens, gametogenesis, and cancer.Nat Rev Cancer. 2005; 5: 615-625Crossref PubMed Scopus (1146) Google Scholar bone morphogenetic proteins,4Rothhammer T Wild PJ Meyer S Bataille F Pauer F Pauer A Klinkhammer-Schalke M Hein R Hofstaedter F Bosserhoff AK Bone morphogenic protein 7 (BMP7) expression is a potential novel prognostic marker for recurrence in patients with primary melanoma.Cancer Biomark. 2007; 3: 111-117PubMed Google Scholar, 5Hsu MY Rovinsky SA Lai CY Qasem S Liu X How J Engelhardt JF Murphy GF Aggressive melanoma cells escape from BMP7-mediated autocrine growth inhibition through coordinated Noggin upregulation.Lab Invest. 2008; 88: 842-855Crossref PubMed Scopus (32) Google Scholar, 6Hsu MY Rovinsky S Penmatcha S Herlyn M Muirhead D Bone morphogenetic proteins in melanoma: angel or devil?.Cancer Metastasis Rev. 2005; 24: 251-263Crossref PubMed Scopus (61) Google Scholar Notch receptors,7Balint K Xiao M Pinnix CC Soma A Veres I Juhasz I Brown EJ Capobianco AJ Herlyn M Liu ZJ Activation of Notch1 signaling is required for beta-catenin-mediated human primary melanoma progression.J Clin Invest. 2005; 115: 3166-3176Crossref PubMed Scopus (272) Google Scholar Wnt proteins,8Weeraratna AT Jiang Y Hostetter G Rossenblatt K Duray P Bittner M Trent JM Wnt5a signaling directly affects cell motility and invasion of metastatic melanoma.Cancer Cell. 2002; 1: 279-288Abstract Full Text Full Text PDF PubMed Scopus (739) Google Scholar or the ABCB5, CD133, CD166, CD34, nestin, or c-KIT antigens.2Hendrix MJ Seftor EA Hess AR Seftor RE Molecular plasticity of human melanoma cells.Oncogene. 2003; 22: 3070-3075Crossref PubMed Scopus (184) Google Scholar, 9Frank NY Margaryan A Huang Y Schatton T Waaga-Gasser AM Sayegh MH Sadee W Frank MH ABCB5-mediated doxorubicin transport and chemoresistance in human malignant melanoma.Cancer Res. 2005; 65: 4320-4333Crossref PubMed Scopus (442) Google Scholar, 10Klein WM Wu BP Zhao S Wu H Klein-Szanto AJ Tahan SR Increased expression of stem cell markers in malignant melanoma.Mod Pathol. 2007; 1: 102-107Crossref Scopus (249) Google Scholar, 11van Kempen LC van den Oord JJ van Muijen GN Weidle UH Bloemers HP Swart GW Activated leukocyte cell adhesion molecule/CD166, a marker of tumor progression in primary malignant melanoma of the skin.Am J Pathol. 2000; 156: 769-774Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar Although such studies suggest a link between physiological (tissue) and neoplastic stem cell markers, it remains unclear whether melanomas and Merkel cell carcinomas express genes and related proteins typical of embryonic and tissue progenitor stem cells. In the course of gene expression profiling of the rete-tip-associated stem cell niche by immunoguided laser capture microdissection and real time reverse transcription (RT)-PCR,12Zhan Q Signoretti S Whitaker-Menezes D Friedman TM Korngold R Murphy GF Cytokeratin 15-positive basal cells targeted in graft-versus-host disease express a constitutive antiapoptotic phenotype.J Invest Dermatol. 2007; 127: 106-115Crossref PubMed Scopus (23) Google Scholar one of us (Q.Z.) detected significantly increased mRNA expression for SOX2. The SOX2 protein is a transcription factor critical to maintenance of embryonic neural crest stem cell pluripotency and self-renewal. It belongs to the SRY-related high mobility group box proteins that are homologous to the sex-determining gene on chromosome Y.13Bylund M Anderson E Novitch BG Muhr J Vertebrate neurogenesis is counteracted by Sox1–3 activity.Nature Neurosci. 2003; 6: 1162-1168Crossref PubMed Scopus (605) Google Scholar The present study was therefore conducted to characterize the expression of SOX2 in human embryonic and adult skin, and to explore its potential relevance to cutaneous neoplasms of neural crest derivation. Paraffin-embedded sections were obtained from the Department of Pathology, Brigham and Women's Hospital, according to an approved Institutional Review Board protocol in all cases. Samples included discarded human fetal skin (n = 35), adult normal skin (n = 24), nevi (n = 19), Merkel cell carcinoma (n = 9), and melanoma (n = 26). Sections from discarded fetal tissue were obtained from conceptuses of 6 to 24 weeks' estimated gestational age (EGA), and consisted of scalp, acral, truncal, and extremity skin. Sections of adult human skin included acral skin, scalp, lip, trunk, earlobe, face, and thigh. Human melanoma tissue microarrays consisted of: 1) a melanoma progression microarray (Cat# ME1003, US Biomax Inc., Rockville, MD) containing 100 cores of which 94 were technically evaluable (21 banal nevi, 53 primary vertical growth melanomas, and 20 metastatic melanomas); and 2) an outcome-annotated array (Cat# IMH-369, Imgenex, San Diego, CA) containing 59 cores of which 52 were technically evaluable (52 vertical growth phase melanomas—at least 2 mm in Breslow thickness—where 41 patients succumbed to their disease [median follow-up 25 months] and 11 survived [median follow-up 80.5 months]). Human melanoma cell lines A2058 and SK-MEL-5 were obtained from American Type Culture Collection (Manassas, VA) and grown in Dulbecco's modified Eagle's medium (Sigma-Aldrich Inc, St Louis, MO) supplemented with 10% inactivated fetal bovine serum (Hyclone Laboratories Inc, Logan, UT), 200 mmol/L/L l-glutamine, 100 IU/ml penicillin, and 100 μg/ml streptomycin, and maintained at 37°C in a humidified atmosphere containing 5% CO2. Total RNA from human melanoma cell lines A2058 and SK-MEL-5 was extracted from cell cultures using the RNAeasy Mini Kit (Qiagen, Valencia, CA). Total RNA was reverse-transcribed to cDNA using SuperScript III First-strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA). cDNA from A2058 and SK-MEL-5 cells was surveyed for the expression of 84 human embryonic stem cell-related genes using the embryonic stem cell RT2Hendrix MJ Seftor EA Hess AR Seftor RE Molecular plasticity of human melanoma cells.Oncogene. 2003; 22: 3070-3075Crossref PubMed Scopus (184) Google Scholar profiler PCR array (Cat. no. PAHS-081, SA Biosciences, Frederic, MD). PCR was performed in a 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA). The glyceraldehyde-3-phosphate dehydrogenase housekeeping gene was used for normalization and data were analyzed using the 2−ΔΔCT method.14Livak KJ Schmittgen TD Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCT method.Methods. 2001; 25: 402-408Crossref PubMed Scopus (103415) Google Scholar Significant relative gene overexpression was defined as ≥fourfold difference in mRNA transcripts between the two cell lines. Recombinant lentiviral vectors were generated by co-transfecting pLKO.1-SOX2 (Sigma, St. Louis, MO), harboring short hairpin (sh)RNA for human SOX2, or non-target control shRNA (Sigma) with packaging plasmids VSVg and pCMV-ΔR8.2 (Sigma) into 293T packaging cells using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) according to manufacturer's instructions. Culture supernatants containing recombinant lentiviral particles were used to infect A2058 melanoma cells. Two days after infection, cells were selected with puromycin (1 μg/ml) for a period of 7 days. Subconfluent cultures were washed with PBS and extracted in lysis buffer containing 1% Triton X-100, 1% deoxycholic acid, 2 mmol/L CaCl2, and 1.8 mg/ml iodoacetamide, and 1 mmol/L phenylmethyl sulfonyl fluoride) in PBS. Cell lysates were quantified using a bicinchoninic acid protein assay kit (Pierce, Rockford, IL). An equal amount (50 μg) of total protein from each sample was subjected to electrophoresis on NuPAGE 4% to 12% Bis-Tris gels (Invitrogen), transblotted onto nitrocellulose membranes (Pierce), and probed with primary anti-SOX2 antibodies (Millipore, Billerica, MA), followed by a peroxidase-conjugated secondary antibody (Pierce). Subsequent re-probing using anti-β-actin (Abcam, Cambridge, MA) was also performed as internal loading control. Immunoreactive bands were detected using super-Signal West Femto Chemiluminescent substrate (Pierce), captured by a Syngene Chemi Genius Bio Imaging System (Syngene, Frederick, MD), and quantified by densitometry. Experiments were performed at least twice with consistency. Subconfluent cultures were trypsinized and seeded in 35-mm wells at 1 × 104 cells per well. Cells were re-fed twice weekly. At given intervals, cells in quadruplicate wells were harvested and counted in a Coulter counter (Coulter Electronics, Luton, UK). Statistical analyses were performed using the Mann-Whitney U-test. Experiments were repeated twice with similar results and representative data from one experiment is shown. Subconfluent melanoma cells were harvested and suspended in serum-free medium at a density of 108 cells/ml in PBS. One hundred μl of cell suspension (107 cells) were injected subcutaneously in the dorsal skin of each severe combined immunodeficient mouse (C.B-17; Taconic Laboratory, Germantown, NY; eight mice per condition). Tumor volume was monitored twice a week and determined as follows: (maximal dimension × minimal dimension)2Hendrix MJ Seftor EA Hess AR Seftor RE Molecular plasticity of human melanoma cells.Oncogene. 2003; 22: 3070-3075Crossref PubMed Scopus (184) Google Scholar/2. Statistical analyses were performed using analysis of variance following log transformation. The mice were sacrificed at the experimental end point and tumors were dissected, fixed in formalin, and subjected to immunohistochemistry for verification of SOX2 expression. Five-micron sections were deparaffinized in xylene, followed by treatment in 100% ethanol, and by serial hydration through 95% and 75% ethanol, and deionized H2O. Then, sections were placed in 1× target retrieval solution (Dako, Carpenteria, CA) and boiled in a Pascal pressure chamber (Dako) at 125°C for 30 seconds, 90°C for 10 seconds, and then allowed to cool down to room temperature. Immunohistochemistry was performed using a two-step horseradish peroxidase method. Briefly, the sections were first incubated with primary antibodies at 4°C overnight. The following primary antibodies were used: goat anti-human SOX2 (Neuromics, Edina, MN), mouse anti-microphthalmia-associated transcription factor (MITF) (Dako), and mouse anti-melanoma antigen recognized by T-cells (MART-1) (Signet Laboratories, Dedham, MA). After washing out unbound primary antibodies with Tris buffered saline-0.05% Tween 20, the tissue sections were incubated with peroxidase-conjugated secondary antibody at room temperature for 30 minutes, then washed with Tris buffered saline-0.05% Tween 20 three times for 5 minutes each. Immunoreactivity was detected using NovaRed peroxidase substrate (Vector Laboratories, Burlingame, CA). For double-label immunohistochemical staining, sections were deparaffinized and heat-induced antigen retrieval with Target Retrieval Solution (Dako) was used. Sections were incubated with SOX2 antibody (Neuromics) and one of the following: anti-CK20 (Dako), anti-CK15 (Lab Vision Corp., Fremont, CA), or anti-MITF (Dako) at room temperature for 1 hour and then with peroxidase-conjugated horse anti-goat antibody (Vector) and alkaline phosphatase-conjugated horse anti-mouse antibody at room temperature for 30 minutes. SOX2 and MITF staining was detected with NovaRed peroxidase substrate (Vector). CK20 and CK15 staining was detected with nitro blue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate, toluidine salt alkaline phosphatase substrate (Hoffman-La Roche, Nutley, NJ). Cell type(s) with strong nuclear immunoreactivity were recorded for embryonic-fetal skin, adult skin, and tumor samples. For patient tumor whole sections, SOX2 nuclear immunoreactivity was quantified using a semiquantitative scale (0 = negative, 1+ = ≤5% positive cells, 2+ = 6% to 25% positive cells, 3+ = 26% to 50%, and 4+ >50% positive cells). Tissue microarray cores were designated positive when unequivocal nuclear labeling was detected even in a minority of cells. Data were entered into an Excel file and imported into the SAS statistical software program. Descriptive statistics and univariate logistic regression were performed. Survival analysis was performed using the XLSTAT statistical package, using the Log-rank test for comparison of the survival distribution between two groups. For double-label immunofluorescence staining, sections were stained with anti-SOX2 (Neuromics) and one of the following antibodies: nestin (Chemicon, Temecula, CA), SOX10 (R&D, Minneapolis, MN), MITF (Dako), c-KIT (Dako), MART-1 (Signet Laboratories), CD31 (Bethyl Laboratories, Montgomery, TX), and CK20 (Dako). Isotype-matched antibodies (BD Pharmingen Inc., San Diego, CA) were included as controls. Alexa Fluor 594 donkey anti-goat IgG (Invitrogen, Eugene, OR), Alexa Fluor 488 donkey anti-mouse IgG (Invitrogen), and fluorescein isothiocyanate rat anti-mouse IgM (BD Biosciences, San Jose, CA) were used as secondary antibodies. Sections were deparaffinized and heat-induced antigen retrieval with Target Retrieval Solution (Dako) was performed. Slides were blocked for 30 minutes with horse or donkey sera, as appropriate, and incubated with the primary antibodies for 1 hour and then incubated with the appropriate secondary antibody in the dark for 1 hour. SOX2, MITF, and CK20 immunofluorescence triple labeling was performed by incubating sections with SOX2 (Neuromics), MITF (Dako), and CK20 (Dako) antibodies at room temperature for 1 hour and then with Alexa Fluor 594 donkey anti-goat IgG (Invitrogen) and Alexa Fluor 488 donkey anti-mouse IgG (Invitrogen) secondary antibodies. After washing to remove unbound secondary antibodies, slides were coverslipped with ProLong Gold Anti-Fade with DAPI (Invitrogen). Sections were analyzed with a BX51/BX52 microscope (Olympus America Inc, Melville, NY). Images were captured using the CytoVision 3.6 software (Applied Imaging, San Jose, CA). Figure 1 summarizes the localization of SOX2-positive cells at different stages of development in skin of conceptuses ranging from 6 to 24 weeks EGA. The early stages (eg, 6 to 12 weeks) demonstrated scattered single SOX2 immunoreactive cells within the dermis, often in a perivascular array (Figure 1A), and in the stratum germinativum of the epidermis (Figure 1B). Single cells exhibiting SOX2 nuclear staining were also noted in intimate association with forming buds of primary hair germ (Figure 1C), with the underlying mesenchymal condensation of cells also showing positivity that became progressively more prominent as the follicular papillae developed with downward descent to form differentiated lanugo follicles (Figure 1, D−H). SOX2-positive follicular papillae cells with progressive follicular maturation communicated with a thin mantle of positive cells within the adventitial sheath of the hair bulb and isthmus (Figure 1, D−H). A subpopulation of SOX2-positive cells (Figure 1I) within the follicular papilla co-expressed nestin (Figure 1, J−K). This was the only site of SOX2/nestin co-expression in the hair follicle. All SOX2-positive cells in this microdomain (Figure 1L) were negative for SOX10, MITF, c-Kit, and MART-1, all of which were clearly expressed by melanocytes within the basal layer of the adjacent hair bulb epithelium (eg, SOX10, Figure 1, M−N). Within the bulbous hair peg stage of trichogenesis (13 to 24 weeks), single SOX2-positive cells continued to be observed in the basal region of the epidermis as well as in the follicular infundibular epithelium in the vicinity of the forming bulge region. At this juncture, few if any dermal cells retained SOX2 immunoreactivity (Figure 1F). In adult skin, SOX2-positive cells also were present singly within the basal cell layer of the epidermis, with a marked tendency (>90%) to be localized at or near the tips of rete ridges (Figure 2, A−C), without evidence of staining within the underlying dermis (in contrast to embryonic skin, Figure 1A, inset). In the buccal mucosa of the lip, where akin to embryonic skin, rete ridges are not present, small clusters of SOX2-positive cells were present in the superficial submucosa and usually directly subjacent to singly disposed SOX2-positive cells within the epithelial basal layer (Figure 2D). The bulge region of hair follicles, which approximated the insertion site of arrector pili muscles, also consistently contained one to several SOX2-positive cells (Figure 2, E and F). The only mesenchymal localization of SOX2 in adult skin was the persistence of variable numbers of immunolabeled cells in follicular papillae, which otherwise retained a co-labeling profile similar to that of embryonic skin (see above). Specifically, nestin co-expression with SOX2 was restricted to the follicular papilla, as shown in Figure 1. The adventitial mantle at the tips of telogen follicles, presumably residua of follicular papillae, as well as the adjacent persistent follicular bulge regions, also contained positive cells (Figure 2, G−H). Three subsets of SOX2-positive intraepithelial cells were defined in both embryonic and adult skin based on co-expression of markers of melanocytic and Merkel cell lineage. In addition to SOX2-positive cells that failed to co-express nuclear MITF (Figure 2, I−K) or cytoplasmic CK20 (Figure 2, O and Q, arrow), CAM5.2, or the keratinocyte stem cell marker CK15,15Lyle S Christofidou-Solomidou M Liu Y Elder DE Albelda S Cotsarelis G The C8/144B monoclonal antibody recognizes cytokeratin 15 and defines the location of human hair follicle stem cells.J Cell Sci. 1998; 111: 3179-3188PubMed Google Scholar rare cells co-expressed SOX2 and MITF (Figure 2, L−N), and more abundant cells co-expressed SOX2 and either CAM5.2/CK20 (Figure 2, O−Q). Occasionally, SOX2-only and SOX2/CK20-positive cells were clustered in the same microdomain (Figure 2R), and in close proximity to either CK20-only cells (Figure 2S) or MITF-only cells (Figure 2T). Co-expression of SOX2 and MART-1, a marker for more mature melanocytes, was not identified, although numerous MART-1-positive, SOX2-negative melanocytes could be documented in the basal cell layer. Co-expression of SOX2 and c-KIT was not observed in either embryonic of adult skin. Although some Merkel cells had a CK20-positive, SOX2-negative phenotype, the precise ratio of SOX2-positive and negative cells also expressing CK20 could not be determined due to limitations in sampling. Based on the identification of SOX2 and CK20-coexpressing cells in embryonic and adult skin, we next sought to determine the potential presence and expression pattern of SOX2 in Merkel cell carcinoma. Nine samples of human primary and metastatic Merkel cell carcinoma interrogated by immunohistochemistry were all diffusely positive for SOX2, with >50% of cells exhibiting nuclear SOX2 immunopositivity (Figure 3, A−D). One of the cases studied contained areas of squamous differentiation (Figure 3, E and F); the neuroendocrine component of this tumor co-expressed SOX2 and CK20, whereas the malignant squamous component expressed neither (Figure 3, G−K). Double labeling confirmed restriction of CK20 expression to tumor cells that expressed SOX2, and SOX2-immunoreactive cells facilitated detection of intravascular invasion in one case (Figure 3, L−M). Because rare SOX2-positive cells in adult skin also co-expressed MITF, we also examined the immunoreactivity of SOX2 in a survey sample of nevi and melanomas. Table 1 summarizes the characteristics of patients and respective primary or metastatic melanomas and their SOX2 immunoreactivity patterns. Sixteen (62%) melanomas came from male patients and ten (38%) from female patients; eighteen tumors (69%) were primary and eight (31%) were metastatic. The majority of tumors were composed of epithelioid cells, with five (19%) having a prominent spindle cell component. In addition, two melanomas had a small cell component. Overall, zonal SOX2 expression was noted in 9/26 (35%) malignant melanomas. Of these, 1/9 showed 1+ immunopositivity, 3/9 showed 2+ positivity, 2/9 had 3+ positivity, and 3/9 tumors showed 4+ positivity. The remaining lesions had rare isolated SOX2-positive cells, but not zonal positivity as described above. The more intensely stained SOX2-positive cells tended to be concentrated at the interface of tumor and the surrounding stroma (Figure 4, A−C), where they were shown to be intimately associated with CD31-expressing peritumoral and intratumoral blood vessels (Figure 4, C and L−O). Melanomas with epithelioid, spindle, and small cell differentiation all harbored SOX2-positive cells (Figure 4, D−G), although occasional lesions with bimodal differentiation (eg, spindle cells or small cells) sharply juxtaposed with epithelioid cells tended to exhibit qualitatively more intense reactivity in the spindle and small cell components (Figure 4, H−K). In nevi, SOX2 immunoreactivity was present in 3/19 (16%) cases, although staining was qualitatively less intense and accentuations of staining at tumor-stromal interfaces or in association with differences in cell size and shape were not consistently observed (Figure 4C inset).Table 1Characteristics and Sox2 Immunoreactivity Pattern of Tumors from Patients with Primary and Metastatic Melanoma (MM)CaseAgeSexLocationMM typeCell type (s)Breslow (mm)ClarkMits/mm2Sox2166MFlankNevoidEpitheliod, nevoid1.05IV30263MShoulderResidualEpithelioid1.1IV40324MAbdomenSSEpithelioid1.2IV<10445FAnkleALEpithelioid0.85IV<10522MScalpSSEpithelioid1.2III<10662FGreat ToeSpindle cellEpithelioid, spindle6.0V243+777MShoulderMixedEpithelioid6.5IV170842FHeelALEpithelioid, spindle4.0IV23+954MBackSSEpithelioid0.8III201062FChinDesmoplasticSpindle10.0V22+1189MAbdomenNodularEpithelioid2.0IV501252FThighNodularEpithelioid, spindle4.0IV>62+1360MEarlobeUnclassifiedEpithelioid, small cell1.4IV104+1457FMandibleMucosal typeEpithelioid1.05N/AN/A01548FLungPrimary lungEpithelioidN/AN/AN/A01653MChestUnclassifiedEpithelioid0.25IIN/A01782MInguinal RegionMetastaticEpithelioidN/AN/AN/A01849FScalpMetastaticEpithelioidN/AN/AN/A01921MInner ThighNodularEpithelioid2.7IV>602065MPericavalMetastaticEpithelioidN/AN/AN/A1+2158MLymph NodeMetastaticEpithelioidN/AN/AN/A02264MSmall BowelMetastaticEpithelioidN/AN/AN/A02363MChest WallMetastaticEpithelioidN/AN/AN/A4+2459MLymph NodeMetastaticEpithelioid, small cellN/AN/AN/A2+2562FLymph NodeMetastaticEpithelioidN/AN/AN/A02652FShoulderSpindle cellSpindle5.0V124+N/A, not available or not applicable; Mits, mitosis.M, male; F, female; SS, superficial spreading type; AL, acral lentiginous type. Open table in a new tab N/A, not available or not applicable; Mits, mitosis. M, male; F, female; SS, superficial spreading type; AL, acral lentiginous type. An exploratory analysis was undertaken to investigate the clinical significance, if any, of SOX2 immunoreactivity in human melanomas. A univariate logistic regression analysis of the relatively small sample of melanomas evaluated indicated that Breslow thickness was significantly correlated with SOX2 immunopositivity. Via this analysis, melanomas from the cohort that differed by 1 mm in Breslow thickness were on average twice as likely to exhibit presence of SOX2 expression. Conversely stated, SOX2-positive melanomas were significantly associated with a greater likelihood (twice as likely on average) of being 1 mm thicker than their SOX2-negative comparators (OR = 2.01[1.04, 3.92], P = 0.039). To expand the clinical sample size, we also evaluated SOX2 expression in a melanoma progression tissue microarray and in a melanoma outcome-annotated tissue microarray. Of the additional 91 primary melanomas, 41 (45%) expressed SOX2, with positive staining also detected in 3 of 21 nevi (14%), and 13 of 34 metastases (38%). With regard to clinical outcome, 6 of 11 (55%) survivors, and 21 of 41 (51%) nonsurvivors (eg, death of melanoma) expressed SOX2. No statistically significant difference in survival time was found between SOX2-positive and SOX2-negative tumors in this outcome microarray (Log-rank test P = 0.27). Melanoma cell lines were selected based on SOX2 expression; A2058 melanoma cells are uniformly positive for SOX2, whereas SK-MEL-5 cells are largely negative by immunohistochemistry. Real time RT-PCR confirmed that A2058 cells had 247-fold higher expression of SOX2 than to SK-MEL-5 cells. Using a commercially available embryonic stem cell gene expression array (SA Biosciences), we compared the expression profiles in these SOX2-positive and SOX2-negative cells. Of 84 genes represented in the screening array, there was detectable signal for 41 genes in one or both cell lines and no signal for 43 genes in both cell lines. Of the 41 stem cell-related genes detected, 25 (61%) were significantly overexpressed in A2058 cells, and 5 genes (12%) were relatively overexpressed in SK-MEL-5 cells. The remaining 11 genes for which signal was detected showed no significant difference in expression between the two cell lines. Genes overexpressed in A2058 cells corresponded to transcription factors involved in the maintenance of "stemness" (eg, SOX2, FOXD3, GATA6, NANOG, nestin, CDH5 [vascular endothelial-cadherin]), and signaling molecules required for pluripotency and self-renewal (eg, fibroblast growth factors 4 and 5, interferon-induced transmembrane protein 1, telomerase reverse transcription, teratocarcinoma derived growth factor 1). The 5 genes overexpressed in SK-MEL-5 cells were growth differentiation factor 3; laminin-α1; noggin; nuclear receptor subfamily 5, group A, member 2; and transcription factor CP2-like 1. Three of these genes (noggin; nuclear receptor subfamily 5, group A, member 2; and transcription factor CP2-like 1) are related to maintenance of "stemness," and the other two are growth factors. In summary, of embryonic stem cell-related genes that w