Expression Profiling Reveals Novel Pathways in the ......human melanocytes and melanoma cells. We...

[CANCER RESEARCH 64, 5270–5282, August 1, 2004]

Expression Profiling Reveals Novel Pathways in the Transformation ofMelanocytes to Melanomas

Keith Hoek,1 David L. Rimm,2 Kenneth R. Williams,1 Hongyu Zhao,3 Stephan Ariyan,4 Aiping Lin,1

Harriet M. Kluger,5 Aaron J. Berger,2 Elaine Cheng,6 E. Sergio Trombetta,7 Terence Wu,1 Michio Niinobe,8

Kazuaki Yoshikawa,8 Gregory E. Hannigan,9 and Ruth Halaban6

Departments of 1Molecular Biophysics and Biochemistry, 2Pathology, 3Epidemiology, 4Surgery, 5Internal Medicine, 6Dermatology, and 7Cell Biology, Yale University School ofMedicine, New Haven, Connecticut; 8Division of Regulation of Macromolecular Functions, Institute for Protein Research, Osaka University, Osaka, Japan; and 9LaboratoryMedicine and Pathobiology, Cancer Research Program, Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada

ABSTRACT

Affymetrix and spotted oligonucleotide microarrays were used to assessglobal differential gene expression comparing normal human melanocyteswith six independent melanoma cell strains from advanced lesions. Thedata, validated at the protein level for selected genes, confirmed theoverexpression in melanoma cells relative to normal melanocytes of sev-eral genes in the growth factor/receptor family that confer growth advan-tage and metastasis. In addition, novel pathways and patterns of associ-ated expression in melanoma cells not reported before emerged, includingthe following: (a) activation of the NOTCH pathway; (b) increased Twistexpression and altered expression of additional transcriptional regulatorsimplicated in embryonic development and epidermal/mesenchymal tran-sition; (c) coordinated activation of cancer/testis antigens; (d) coordinateddown-regulation of several immune modulation genes, in particular in theIFN pathways; (e) down-regulation of several genes implicated in mem-brane trafficking events; and (f) down-regulation of growth suppressors,such as the Prader-Willi gene NECDIN, whose function was confirmed byoverexpression of ectopic Flag-necdin. Validation of differential expres-sion using melanoma tissue microarrays showed that reduced ubiquitinCOOH-terminal esterase L1 in primary melanoma is associated withworse outcome and that increased expression of the basic helix-loop-helixprotein Twist is associated with worse outcome. Some differentially ex-pressed genes reside on chromosomal regions displaying common loss orgain in melanomas or are known to be regulated by CpG promotermethylation. These results provide a comprehensive view of changes inadvanced melanoma relative to normal melanocytes and reveal new tar-gets that can be used in assessing prognosis, staging, and therapy ofmelanoma patients.

INTRODUCTION

Wide-spread aberrations in gene expression due to losses and gainsin genetic material, suppressive or activating mutations, and dysregu-lation of transcription factors are rampant in cancer cells, includingmelanomas. Inactivation of p16INK4a, the cyclin-dependent kinaseinhibitor (cyclin-dependent kinase 4/6) that leads to suppression ofretinoblastoma protein, is an early event commonly observed in mel-anomas and other cancer cells as well (reviewed in Refs. 1 and 2).P16INK4a is eliminated in most melanomas not only by point muta-tions, but also by deletion (Refs. 3 and 4; reviewed in Ref. 5) or

silencing through promoter hypermethylation (6). The loss ofp16INK4a expression correlates with the invasive stage of melanomatumor progression (7–9), suggesting that its absence confers growthadvantage. Retinoblastoma protein is also likely to be inactivated bysignaling pathways that stimulate mitogen-activated protein kinase,such as activating mutation in B-Raf and N-Ras (10, 11), and consti-tutive stimulation of cell membrane receptors, such as the fibroblastgrowth factor (FGF) receptor 1 (12) and integrin pathway (13). One ofthe consequences of retinoblastoma protein inactivation is the releaseof E2F transcription factor, leading to aberrant expression of E2F-regulated genes that promote cell cycle progression (reviewed in Refs.14 and 15).

Several high-throughput approaches have been applied to melano-mas to assess chromosomal aberrations and gene expression patternsin an unbiased fashion. For example, comparative genome hybridiza-tion identified several chromosomal and genetic changes. These in-clude losses of chromosomes 6q, 8p, and 10 and gains in copy numberof chromosomes 1q, 6p, 7, and 8 in primary melanomas (16–22);amplification of the cyclin D1 locus in acral melanoma (17, 19);frequent deletion of chromosome 13 and 17p in melanomas arising inchronically sun-damaged skin; and frequent amplification of chromo-some 12q in mucosal melanomas that can affect metastatic potential(23, 24). Global gene expression profiling of several melanoma sub-types has implicated dysregulation of the Wnt pathway in contributingto melanoma invasion and motility (25, 26), which was validated byectopic expression of Wnt5a (26). More recently, serial analysis ofgene expression of three melanoma tumors indicated up-regulation ofintracellular calcium and G-protein signaling and increased expres-sion of CD74 (27).

Most gene expression analysis studies did not include normalhuman melanocytes. Therefore, we set out to monitor global geneexpression in normal melanocytes compared with melanoma cells. Asreported below, gene expression profiling, validated at the proteinlevel for selected gene products, provides a comprehensive view ofmolecular changes in malignant melanocytes and reveals possibleunderlying molecular pathogenesis not identified before.

MATERIALS AND METHODS

Cells. Normal human melanocytes from newborn foreskins were grown inHam’s F-12 medium (GibcoBRL, Invitrogen Corp., Grand Island, NY) sup-plemented with serum [7% fetal bovine serum from Gemini Bio-Products(Woodland, CA)], termed basal medium, which was further enriched withseveral ingredients required for optimal proliferation. They included 85 nM

12-O-tetradecanoyl phorbol-13-acetate (TPA), 0.1 mM 3-isobutyl-1-methyl-xanthine, 2.5 nM cholera toxin, 1 �M Na3VO4, and 0.1 mM N6,2�-O-dibutyry-ladenosine 3:5-cyclic monophosphate (all from Sigma-Aldrich, St. Louis,MO), termed TICVA (28). Early-passage melanocyte cultures (passage 1–2)pooled from three to six Caucasian donors were used to reduce individualdonor variations. The normal melanocytes proliferated at a population dou-bling time of of 3–4 days, were highly differentiated, and expressed all knownmelanocyte-specific genes (12).

The human primary melanoma WW165 (F, 2.25-mm lesion from the back)

Received 3/1/04; revised 5/10/04; accepted 5/21/04.Grant support: USPHS Grant CA44542 (R. Halaban), AR41942 Yale Skin Diseases

Research Center; R. E. Tigelaar, Program Investigator, the Patrick and Catherine WeldonDonaghue Investigator Award (D. L. Rimm); the C. J. Swebilius Award (H. M. Kluger);and NIH/National Institute of General Medical Sciences Medical Scientist TrainingProgram Grant GM07205 (A. J. Berger).

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Cancer Research Online(http://cancerres.aacrjournals.org) and at http://info.med.yale.edu/microarray/pub�serv.htm. K. Hoek is currently in the Department of Dermatology, University Hospital ofZurich, Zurich, Switzerland.

Requests for reprints: Ruth Halaban, Department of Dermatology, Yale UniversitySchool of Medicine, 15 York Street, P. O. Box 208059, New Haven, CT 06520-8059.Phone: (203) 785-4352; Fax: (203) 785-7637; E-mail: [email protected].

5270

Research. on July 23, 2021. © 2004 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

and metastatic YUHEIK melanoma cells (F, maxillary gingiva, lymph nodemetastasis) were grown in basal medium supplemented with 3-isobutyl-1-methylxanthine required for optimal proliferation. The metastatic melanomasMNT1 (M, lymph node metastasis; Ref. 29), YUMAC (M, recurrent metastasisat the site of excision), YUCAL (F, soft tissue metastasis), YUSAC2 (M, softtissue metastasis), YUSIT1 (site unknown), 501 mel (site unknown; Ref. 30),and YUGEN8 (F, brain) were maintained in basal medium as describedpreviously (28). The YUHEIK, YUMAC, and YUCAL melanoma cells wereused during early passage in culture (passage 3–8).

The normal and malignant melanocytes were harvested during the logarith-mic growth phase to avoid differences due to cell cycle distribution becausenormal melanocytes become arrested in the absence of specific growth factorsunder conditions that allow metastatic melanoma cells to proliferate; likewise,metastatic melanoma cells are arrested in the presence of TPA, N6,2�-O-dibutyryladenosine 3:5-cyclic monophosphate, and cholera toxin (31, 32).

The immortalized mouse melanocytes derived from a B10BR-black mousewere grown in Ham’s F-12 medium supplemented with antibiotics, 10% horseserum, and TPA (33).

RNA Isolation and Application to DNA Microarrays. Approximately20–30 million cells (normal human melanocytes and melanoma cells/each)were used for mRNA extraction per hybridization. High-quality total RNA wasprepared with the TRIzol reagent (Invitrogen Life Technologies, Inc., Invitro-gen Corp., Carlsbad, CA). The quality of RNA products was determinedvisually by 1% denaturing agarose gel, and RNA concentration was measuredusing a spectrophotometer. Poly(A) mRNA was further isolated from �300 �gtotal RNA/sample using the PolyATtract mRNA isolation system IV (Pro-mega, Madison, WI) following the manufacturer’s instructions. This additionalstep is required due to the presence of melanin in the total RNA preparation,especially in normal human melanocytes, that suppresses PCR hybridizationreactions (our experience as well as that of others; see, for example, Ref. 34).The quality and quantity of the poly(A) RNA products were determined byBioanalyzer (Agilent Technologies) and spectrophotometric analysis.

The human genome Affymetrix GeneChip system HG-U133A microarray(Affymetrix Inc., Santa Clara, CA) containing 11 pairs of match/mismatch25-mer oligonucleotide probes for each of �14,500 transcripts of known genesand the OHU16K human library oligonucleotide array comprising a single70-mer oligonucleotide probe in duplicate spots for each of 16,659 differentgene transcripts (printed by the Keck Microarray Resource at Yale) were used.A description of these slides can be found online.10

For the HG-U133A chip, cDNA templates were generated from purifiedmRNAs for in vitro transcription of biotin-labeled amplified RNAs. Protocolsfor hybridizing fragmented biotin-labeled cRNAs, washing unhybridized ma-terial from the chips, and scanning them with a GeneArray Scanner (AgilentTechnologies) were followed as recommended by Affymetrix Inc.

For the OHU16K microarray we used the two-color hybridization protocolsof Drs. Patrick O. Brown and Geoff Childs (35) optimized by the KeckResource. They are available online.11 The slides were scanned by the KeckMicroarray Resource using a GenePix 4000A scanner (Axon Instruments). Theresults appear as whole array images containing signal intensity informationfor both fluorescent Cy dyes. The primary data tables and the image files areavailable from the Yale Microarray Database.

Data Extraction and Statistical Analysis. The total HG-U133A signalwas normalized to an arbitrary signal intensity value of 500 using MAS 5.0software (Affymetrix Inc.). Subsequent analyses for both HG-U133A andOHU16K experiment data sets were performed using GeneSpring 6.0 (SiliconGenetics, Inc., Redwood City, CA). Comparisons were made between normalhuman melanocytes and melanoma cells. We first applied a fold change filterof 2.5 between average control and average melanoma cells, and only thoseprobesets with a minimal fold change of 2.5 combined with an average signalintensity of 500 or higher in either melanocyte or melanoma cells wereincluded in the analysis. The data were then further restricted using theGeneSpring confidence filter, which used the cross-gene error model to obtaina t test P statistic for each probeset and the Benjamini and Hochberg FalseDiscovery Rate for multiple testing correction (36). Probesets exceeding a t testP of 0.05 were not included. Genes were annotated with GenBank Accession

number, title, and chromosomal location using NetAffx, a web interfaceprogram from Affymetrix Inc. For analysis of the Affymetrix data, values inthe range of 2.5–5-fold change were considered moderate, values in the5–10-fold range were considered high, and values above the 10-fold rangewere termed extremely high.

Initial processing of the fluorescent hybridization signal from OHU16Kmicroarray slides was performed with QuantArray 2.0 (Perkin-Elmer, Boston,MA) and Qamerge12 (Hongyu Zhao; Laboratory of Statistical Genomics andProteomics, Yale Medical School, New Haven, CT). The OHU16K data werenormalized using the Intensity Dependent (Lowess) method. The values of theraw data were set to be 5-fold higher than the background level of 200. Genesthat were differentially expressed between normal human melanocytes andmelanoma cells were generated using t tests with a significance level of 0.05.

The two types of microarrays were compared to find the common genesusing UniGene as a common identifier and the KARMA program13 (KeckMicroarray Resource, Yale Center for Medical Informatics, Yale University).Only the overlapped UniGene clusters were considered for the comparisons.When a UniGene cluster ID has more than one gene, the ratio for that clusterwas calculated using the average value of the corresponding genes.

Hierarchical cluster analysis was performed by using the agglomerativeclustering algorithm QT_Clust (37) to determine associations between coex-pressed genes allowing for a correlation coefficient of 0.97 as describedonline.14

Western Blot Analysis. Normal melanocytes and melanoma cells werelysed in radioimmunoprecipitation assay buffer (PBS, 1% NP40, 0.5% sodiumdeoxycholate, and 0.1% SDS) supplemented with a mixture of protease (Com-plete Boehringer Mannheim Corp., Roche Molecular Biochemicals, Indianap-olis, IN) and phosphatase (100 mM NaF, 10 mM Na4P2O7, and 1 mM Na3VO4)inhibitors. Total cell extracts (35 �g proteins/lane), measured by the Bio-Radprotein assay (Bio-Rad Laboratories, Hercules, CA), were fractionated inprecast gels composed of 8% polyacrylamide Tris-glycine or 10% polyacryl-amide NuPAGE Bis-Tris (Novex, San Diego, CA) and Western blotted ac-cording to standard protocols (38).

Antibodies raised against the following proteins were used for Westernblotting: UCHL1 (ubiquitin COOH-terminal esterase L1, also known as pro-tein gene product 9.5; polyclonal antibody AB1761), Chemicon International,Inc.; fibronectin (ab299) and connective tissue growth factor [CTGF(ab6992)], abcam (Cambridge, United Kingdom); tenascin C (H-300), IGBP7(C-16), ErbB-3 (C-17), anti-Twist (sc-15393), and anti-ZEB/TCF8 (E-020),Santa Cruz Biotechnology (Santa Cruz, CA); ATPase Na�/K� transporting �1polypeptide, Upstate Biotechnology (Lake Placid, NY); integrin-linked kinase-associated serine/threonine phosphatase 2C (ILKAP) and Rab27A, BD Trans-duction Laboratories (San Diego, CA); IFN-inducible MX1/MxA (from Dr.Mark A. McNiven; Center for Basic Research in Digestive Diseases andDepartment of Biochemistry and Molecular Biology, Mayo Clinic, Rochester,MN); and ISG15 (from Dr. Ernest C. Borden; Departments of Microbiologyand Cancer Center, Medical College of Wisconsin, Milwaukee, WI). Antibod-ies against capping protein (gelsolin-like, CapG, anti-peptide WKGRKANEK-ERQAALQVAE) were raised in rabbits by Proteintech Group, Inc. (Chicago,IL) and against necdin (NC243) by Michio Niinobe (39); anti-RAB32 anti-bodies were from Dr. R. J. Haslam (Department of Pathology and MolecularMedicine, McMaster University, Ontario, Canada). Anti-tyrosinase (mono-clonal antibody T311) was used as a marker for differentiation, and anti-�-actin (A2066; polyclonal antibody; Sigma-Aldrich) was used as a measure forprotein loading in each well. In several cases, the same membrane wassuccessively blotted with two or three antibodies, thus sharing the �-actincontrol.

Melanoma Tissue Microarrays (TMAs). The TMA comprised 570 tissuecores (553 melanomas, 17 normal skin samples) measuring 0.6 mm in diam-eter, spaced 0.8 mm apart on a single glass slide. The cohort was constructedfrom paraffin-embedded formalin-fixed tissue blocks obtained from the ar-chives of Yale University Department of Pathology as described previously(40, 41). The specimens were resected between 1959 and 1994, with afollow-up period ranging between 2 months and 38 years. The TMA slideswere treated, probed, and scored as described previously (42). The primary

10 http://keck.med.yale.edu/affymetrix/ and http://keck.med.yale.edu/dna_arrays.htm.11 http://cmgm.Stanford.edu/pbrown/ and http://info.med.yale.edu/wmkeck/dnaarrays/

protocols.htm.

12 http://bioinformatics.med.yale.edu/.13 http://ymd.med.yale.edu/karma/cgi-bin/karma.pl.14 http://www.genome.org/cgi/content/full/9/11/1106.

5271

NOVEL PATHWAYS IN MELANOMAS



antibodies anti-UCHL1 and anti-Twist, validated by Western blotting, wereapplied at 1:500 and 1:100 dilution, respectively. The prognostic significanceof the parameters was assessed for predictive value using the Cox proportionalhazards model with overall survival as an end point. Survival curves werecalculated using the Kaplan-Meier method, with significance evaluated usingthe Mantel-Cox log-rank test.

Plasmid and Transfection. The mouse HEY1/HESR1-pIRES2-EGFPplasmid (43) was from Drs. C. C. Hughes and M. Henderson (Department ofMolecular Biology and Biochemistry, University of California Irvine, Irvine,CA), and pcDNA3.Flag-necdin (44) was from Drs. Timothy M. Wright and BoHu (Department of Medicine, University of Pittsburgh, Pittsburgh, PA). TheHEY1/HESR1-pIRES2-EGFP plasmid was transfected into immortalizedTPA-dependent normal mouse melanocytes (33), and the pcDNA3.Flag-necdin was transfected into human melanoma cells (501 mel) using Lipo-fectAMINE 2000 reagent (Invitrogen Life Technologies, Inc., Carlsbad, CA)following the manufacturer’s instructions. Parallel cultures were transfectedwith pcDNA.GFP as a control. Transfection efficiencies were in the order of30% as evaluated by sorting for green fluorescing cells. The level of theectopic protein was also assessed 1 day or 2 days after transfection by Westernblotting of cell lysates (35 �g protein/assay) with anti HEY1/HESR1 antibod-ies (AB5714; Chemicon International, Inc.) or anti-Flag M2 monoclonalantibody (Sigma-Aldrich). The HEY1/HESR1-transfected mouse melanocyteswere grown in medium deprived of growth factor (TPA) 4 days after trans-fection for over 3 months to evaluate changes in growth properties anddifferentiation. The growth properties of pcDNA3.Flag-necdin as comparedwith green fluorescent protein transfectants were evaluated by DNA synthesisassay (28) after 2 days of incubation in G418-supplemented medium (400�g/ml; 30,000 cells/well; 24-well plate in triplicates) applied 2 days aftertransfection and by cell proliferation after 8 days of selection in G418 startingwith an equal cell number (200,000 cells each, seeded in 55-cm2 Petri dishes).Each transfection was repeated three times with similar results.

Immunofluorescence Microscopy. Melanoma cells and normal humanmelanocytes were grown on coverslips and processed for immunofluorescenceanalysis with anti-necdin and anti-Flag antibodies as described previously (45).

Reverse Transcription-PCR. Poly(A) tracked mRNA was isolated fromtotal RNA using the PolyATtract mRNA isolation system IV (Promega)following the manufacturer’s instructions. Complementary DNA template wasgenerated from the poly(A)-selected RNA using Invitrogen SuperScript IIReverse Transcriptase according to the manufacturer’s instructions, and thelevels of Rab33A and FGF13 gene transcripts compared with �-actin controlwere assessed by PCR using the following primer pairs synthesized by theoligonucleotide synthesis facility at the Department of Pathology, Yale Uni-versity School of Medicine (Table 1).

DNA Sequencing. Sequencing of selected cDNA amplified from normalmelanocytes and melanoma cells with gene-specific primers by PCR wasperformed by the Keck DNA Sequencing Resource at Yale.

RESULTS

Global Changes in Gene Expression andHierarchical Clustering

Global probing for differences in gene expression between normalhuman melanocytes and melanoma cells showed statistically signifi-cant altered expression of 589 genes in the Affymetrix U133A data setusing a 2.5-fold change cutoff and of 186 genes in the OHU16Karrays using a 2-fold change cutoff. These cutoffs were determined asthe least stringent when compared with expression of known genes

and published information. Of these, 315 and 274 genes were up- anddown-regulated, respectively, in the Affymetrix U133A data set, and99 and 87 genes were up- and down-regulated, respectively, in theOHU16K data set.

A comparison between the gene expression data from the Af-fymetrix U133A and OHU16K microarrays using a P of �0.05without fold change cutoff restriction showed that the expressionlevels of only five genes were altered in the opposite direction in thetwo types of microarrays. Among these genes, ISG15 (IFN-�-induc-ible protein clone IFI-15K, also known as G1P2) showed significantlyreduced expression in five of six melanoma cell strains compared withnormal melanocytes in the Affymetrix data set (�21-fold) but wasslightly up-regulated in the OHU16K microarray (1.44-fold). Asshown in Fig. 5B, Western blot analysis revealed up-regulation ofISG15 in the melanoma cells, consistent with previously publisheddata (46). The change in expression of the other four genes was in therange of 1.3–2, below that used in our analysis.

Hierarchical clustering of the Affymetrix U133A data set indicatedthat the two independent cultures of normal human melanocytesgrown from different donors closely resembled each other (Fig. 1A,NM). Therefore, the average expression values for specific genes ofthe two data sets were used for further comparisons. In addition, thehierarchical clustering showed that among the six melanoma cellstrains, the MNT1 gene expression pattern was the least divergentfrom that of normal human melanocytes, followed by the primarymelanoma WW165 and the metastatic melanomas YUMAC,YUSIT1, YUHEIK, and YUCAL (Fig. 1A). The similarity betweenthe metastatic MNT1 melanoma cells and normal melanocytes couldnot be attributed solely to differentiated state, being highly pigmented,and expressing normal levels of melanocyte-specific proteins such astyrosinase (Fig. 1B) because YUHEIK and YUMAC melanoma cellshave similar characteristics (Fig. 1B).

The bioinformatic and QT_Clust analysis of the combined data setrevealed differential expression, in some cases in a coordinated man-ner, of genes representing several functional groups. We report herethe analysis derived mostly from the Affymetrix data set, as corrob-orated or not conflicted by the OHU16K data set when applicable. TheAffymetrix chip incorporates mismatch controls for cross-hybridiza-tion and was thus deemed more reliable.

Below is a description of changes in selected functional groupsfollowing, in most cases, the guidance set by Gene Ontology. The fullscope of differential gene expression can be accessed in the supple-mentary data.

Receptor Activity

The oligonucleotide microarray results showed changes in theexpression of receptors, ligands, cell surface and adhesion mole-cules that affect proliferation, angiogenesis, tissue remodeling,cell-cell communication, ion channels, attachment, motility, andinvasion (Fig. 2A; Supplementary Table 1, A and B). Some of thesechanges, as observed also at the protein level (Fig. 2B), confirmedpublished observations. These include integrin signaling (47),insulin-like growth factor (IGF1) receptor (48), FGF receptor 1(reviewed in Ref. 31), transforming growth factor � (49), ErbB-3(50), EphA3 (reviewed in Ref. 51), melanoma-inhibitory activity(52), N-cadherin and protocadherin, down-regulation of E andP-cadherins (53, 54), and the transmembrane protease dipeptidyl-peptidase IV (55). However, there were also changes in the ex-pression of additional genes, in some cases in a coordinatedmanner, that had not been observed previously.

We found increased expression of NOTCH2 and HEY1/HERS1, thetranscription factor activated by NOTCH signaling (Ref. 56; Fig. 2;

Table 1 Primer sequences

Gene Primer Product size (bp)

Rab33A Fa: 5�-GGAGATGGCGCAGCCCATCCT-3� 835B: 5�-TTTTATTTTGATATCAATTTGTATGGAAAC-3�

FGF13 F: 5�-CACCAAAGATGAGGACAGCACTTAC-3� 429B: 5�-TTCAGCACGCCAGAGACACTTC-3�

�-Actin F: 5�-GTGGGGCGCCCCAGGCACCA-3� 540B: 5�-CTCCTTAATGTCACGGCACGATTTC-3�

a F, forward; B, backward.

5272




Supplementary Table 1). NOTCH2 levels were coordinately up-regulated with genes belonging to the family of receptor activity,such as Tenascin C (TNC1), Fibronectin 1 (FN1), Tissue inhibitor ofmetalloproteinase 3 (TIMP3), and other adhesion/receptor molecules(Fig. 2A). However, overexpression of HEY1/HESR1-pIRES2-EGFPin immortalized mouse melanocytes did not facilitate release fromTPA-dependent mode of growth, suggesting that, by itself, HEY1/HERS1 does not confer growth advantage.

The immunoblotting confirmed the general trend of up-regulationof this class of genes (Fig. 2B). However, there was no consistentcorrelation between transcript levels, as reflected by hybridizationintensity and cellular proteins. TNC1, for example, was up-regulatedin all six melanoma cell strains examined with the Affymetrix mi-croarray (Fig. 2A), but the protein was not detected in YUHEIK andMNT1 melanoma cells (Fig. 2B).

The role of integrin signaling in melanoma progression was under-

scored by the overexpression of several integrins, in particular ITGA7,ITGA4, ITGA6, and ITGB3 (integrins �7, �4, �6, and �3; Fig. 2C); thefour integrin ligands TNC1, FN1, CTGF, and Cyr61 (cysteine-rich);and two tetraspanins, TM4SF13 and TM4SF1, known to complex withintegrins (Fig. 2D) and the underexpression of ILKAP (Fig. 2E;Supplementary Table 1, A and B). Here again we confirmed theoverexpression of CTGF and the underexpression of ILKAP at theprotein level, the latter correlated nicely with the Affymetrix genetranscript hybridization intensity (compare Fig. 2F, ILKAP withFig. 2E).

There was up-regulation of IGF1 (�4-fold, in four of six melanomacells) and several insulin-like growth factor-binding proteins (IG-FBPs), with IGFBP3 being the most prominent (Fig. 3). The expres-sion of IGFBPs appeared to be nonrandom, with higher levels inmetastatic cells, such as MNT1, YUHEIK, YUSIT1 and YUCAL, andto a lesser degree, YUMAC, compared with normal melanocytes andthe primary melanoma WW165 (Fig. 3). Interestingly, the cytokinesCyr61 and CTGF discussed above also belong to the family ofIGFBPs (IGFBP10 and IGFBP8, respectively), and their expression issimilarly coordinated with the other IGBPs (Fig. 3).

We show for the first time the up-regulation of FGF13/FHF2 (Fig.4A), confirmed by semiquantitative reverse transcription-PCR (Fig.4C). FGF13 maps to chromosomal region Xq26.3, and QT_Clustanalysis revealed that it is coordinately expressed in the differentmelanoma cell strains in a pattern similar to that of MAGE genefamily members located on Xq28 (Fig. 4A), suggesting a similarmechanism of activation. FGF2, another member of this family ofgrowth factors up-regulated in melanoma cells (57, 58) and known toconfer growth advantage (59, 60), was not identified here in themicroarray analysis probably due to low mRNA levels (57).

Cancer-Testis Antigens (CTAs)

The results not only confirmed the up-regulation of several genesencoding CTAs (reviewed in Ref. 61) but also unraveled differentpatterns of coexpression for MAGE and GAGE family members (Fig.4, A and B). The reexpression of at least some of these antigens intumors is driven by demethylation of selected CpG in the promoterregions (reviewed in Refs. 61–63), and the unique pattern of expres-sion for each group suggests a coordinated chromatin modification.Although the function of these antigens is not known, they arerecognized by autologous CTLs and are targets for CTL-directedimmunotherapy (reviewed in Ref. 61). The coordinated expression ofspecific CTA members revealed by our data may help in assessingantigen expression and selection of patients for immunotherapy.

Immune Modulation

In this category, we included the IFN-regulated genes, immunecomplements, and histocompatibility antigens, as well as genes en-coding proteins in the vesicular and ubiquitin pathways (Supplemen-tary Tables 2–4). The latter two categories may provide melanomacells with an immune escape mechanism by reducing the level ofantigenic peptides (see, for example, Ref. 64).

IFN-Regulated Genes. Genes that belong to the IFN pathwaywere predominantly down-regulated (21 of 22 genes), composing�8% of all down-regulated genes (Supplementary Table 2A). Fur-thermore, 14 of these genes were coordinately down-regulated, asrevealed by QT_Clust analysis (Fig. 5A). The down-regulation ofMX1/IFI78 was confirmed by Western blotting, showing nice corre-lation with the Affymetrix data set. MX1/IFI78 was expressed innormal melanocytes (NM) and YUHEIK (Fig. 5B), the only mela-noma cell strain with intensity levels similar to that in normal mela-nocytes (Fig. 5A, gray diamonds). On the other hand, the expression

Fig. 1. Hierarchical clustering and differentiation marker. A, cluster map and phylo-genetic tree resulting from a hierarchical cluster analysis of normalized intensity dataacross two control (melanocytes; NM) and six sample (melanoma, as indicated) data sets.Genes were selected for clustering according to a fold change and signal intensity cutoffprotocol described in the text. Clustering was performed using GeneSpring 6.0 (SiliconGenetics, Inc.). Because the data distribution is parametric, we used a standard correlationalgorithm for both gene and condition tree analyses. B, Western blot of cell extracts withanti-tyrosinase monoclonal antibody T311 and anti-�-actin as a control. The normalmelanocyte (NM) and melanoma cells (YUHEIK, YUMAC, and MNT1) with hightyrosinase levels were also highly pigmented. The other melanoma cells (WW165,YUCAL, YUSIT1, 501 mel, YUSAC, and YUGEN8) were amelanotic or very slightlypigmented.

5273




of ISG15, which was conflicted between the Affymetrix HG-U133Aand the Operon 70-mer 16K data sets (see above), was shown to beup-regulated (Fig. 5B), in agreement with previously published re-ports (46).

Leukemia Inhibitory Factor (LIF). The immune modulator LIF(also known as melanoma-derived LPL inhibitor) was up-regulated infour of six melanoma cell strains (Fig. 5C; Supplementary Table 2B).LIF is a cytokine that activates monocytes and macrophages andinduces chemotaxis in immune cells (reviewed in Ref. 65). Its pro-duction in melanomas was suggested to induce cancer cachexia (66,67). LIF also stimulated the human HLA-G promoter in JEG3 chori-ocarcinoma cells (68). In our QT_Clust analysis, LIF expressionstrongly correlated with the expression of genes encoding severalclass II major histocompatibility complexes, HLA-DQB1, HLA-DRA,HLA-DMA, and HLA-DRB4 (Fig. 5C), suggesting a common mech-anism by which this group of genes is regulated.

Vesicular Pathway. Several genes in the membrane transportpathway were down-regulated. These include VAMP8, TRS20, HPS1,and several Ras-related GTP-binding proteins (Fig. 5D; Supplemen-tary Table 3). The products of some of these genes participate in thetransport of proteins into melanosomes and/or the transport of mela-nosomes to the plasma membrane (reviewed in Refs. 69–71). Based

on the hybridization intensity data, Rab38, Rab27A, and Rab32 werethe most abundant, and Rab31 and Rab33A were the least abundantfamily members in normal human melanocytes (Fig. 5D). As con-firmed by Western blotting, Rab38 and Rab27A were down-regulatedin several melanomas but remained at high levels in the melanoticMNT1 melanoma cells, the line least divergent from normal humanmelanocytes (Fig. 5B). In contrast, reverse transcription-PCR analysisrevealed that Rab33A was easily detected in normal melanocytesbut not in the five melanoma cell strains tested, including MNT1(Fig. 5E).

RAGA (Ras-related GTP-binding protein, also known as FIP-1) wasalso about 3-fold down-regulated in the six melanoma cell strainstested (Supplementary Table 3). RAGA is involved in transport ofviruses and cytoplasmic organelles and in chromosome segregationthrough its binding to the light chain of cytoplasmic dynein (72), andtherefore its down-regulation may abrogate these processes. BecauseRAGA localizes to 9p21.3, a region commonly lost in melanomas, wesequenced the cDNA from normal melanocytes and several melanomacell strains (501 mel, YUMAC, and YUSIT1), but we ruled out thepresence of any mutation.

Ubiquitin Pathway/Proteolysis. The ubiquitin pathway encom-passes an additional group of genes that was predominately down-

Fig. 2. Differential expression of genes in thecategory of receptor activity. The figure showsdifferences in the expression of a wide spectrum ofgenes meeting fold change and intensity minimumrequirements and identified as belonging to recep-tor activity. To accommodate the large span inhybridization intensity values, they are plotted hereand in most subsequent figures in log scale. A,coordinated regulation of receptor activity genes.Data were subjected to QT_Clust analysis in Gene-Spring 6.0, as described in the Fig. 1 legend. Min-imum cluster size used was 2, minimum correla-tion was 0.98, and a standard correlation similaritymeasure was used. FN1, fibronectin 1; TNC, tena-scin C1 (hexabrachion); TNFRSF12A, tumor ne-crosis factor receptor superfamily member 12A;TIMP3, tissue inhibitor of metalloproteinase 3(Sorsby fundus dystrophy, pseudoinflammatory);BPAG1, bullous pemphigoid antigen 1, 230/240kDa; PCOLCE2, procollagen C-endopeptidase en-hancer 2; SDC2, syndecan 2 (heparan sulfate pro-teoglycan 1, cell surface-associated, fibroglycan);ITGA6, integrin, �6; COL9A3, collagen, type IX,�3. B, Western blots of cell extracts. ATP1B1,ATPase, Na�/K� transporting �1 polypeptide;ErbB3, v-erb-B2 erythroblastic leukemia viral on-cogene homolog 3; CapG, capping protein (actinfilament), gelsolin-like. Others terms are as definedin A. �-Actin was used as a control. Normal humanmelanocytes (NM) were grown continuously in thepresence of growth factor (�) or overnight withoutgrowth factors. The same membrane was used toblot successively for ATP1B1 and FN1 and forintegrin-linked kinase-associated serine/threoninephosphatase 2C and CapG. C�E, expression ofgenes belonging to the integrin pathway. ITGA7,ITGA4, ITGA6, and ITGB3, integrin �7, �4, �6, and�3, respectively; CTGF, connective tissue growthfactor; Cyr61, cysteine-rich; TM4SF1 andTM4SF13, tetraspanin 1 and 13, respectively;ILKAP, integrin-linked kinase-associated serine/threonine phosphatase 2C. F, Western blots of cellextracts using antibodies against CTGF, integrin-linked kinase-associated serine/threonine phospha-tase 2C, and �-actin as a control.

5274




regulated in melanoma cells compared with normal melanocytes (12of 14 genes; Supplementary Table 4).

GRAIL (also known as RNF128), an E3 ubiquitin ligase, wasup-regulated in five of the seven melanoma cell strains tested (Fig.6A). Western blot analysis also showed up-regulation at the proteinlevel in five of nine melanoma cell strains (Fig. 6B). The two data setswere in general agreement, with the exception of YUSIT1 (Fig. 6,compare hybridization intensity levels in A and protein level in B).

Down-regulation of UCHL1 at the gene transcription level, asreflected by the Affymetrix hybridization intensity values, correlatednicely with the protein level (compare NM, YUHEIK, and YUCALexpression in Fig. 6, A and B, panel marked UCHL1). Assessment ofUCHL1 expression in the 553 case melanoma TMA showed a trendtoward decreased survival with decreased levels that was not statisti-cally significant (Fig. 6C; P � 0.0671). However, in primary mela-nomas, there was a significant association between poor survival anddecreased UCHL1 expression (Fig. 6D; P � 0.0361). High UCHL1expression was more abundant in primary lesions compared withmetastases (P � 0.01) and in lesions thinner than 2 mm (P � 0.04),suggesting that it is lost as malignancy progresses (see Fig. 6E forhistospots representing high and low UCHL1 expression). There wasno significant association between UCHL1 expression and Clarklevel, presence of ulceration, patient age, and gender. On multivariateanalysis, UCHL1 was not an independent predictor of survival amongprimary melanomas (P � 0.1), and the only two independent predic-tors were Breslow depth and Clark level.

Transcription Factors and Growth Suppressors

There were changes in the expression of genes encoding highlyconserved families of transcriptional regulators critical for develop-mental programs, mesoderm/ectoderm transition, and differentiation(Supplementary Table 5). In addition to HEY1/HERS1 mentionedabove, ASXL1, SHOX2, and FOXD1 were persistently up-regulated inmelanoma cells (Fig. 7A; Supplementary Table 5). FOXD1 transcrip-tion factor regulates the expression of growth factors secreted bystromal cells that modulate the differentiation of neighboring epithe-lial cells, such as vascular endothelial growth factor and placentalgrowth factor (73). In our QT_Clust analysis, FOXD1 expressionclearly correlated with vascular endothelial growth factor and heparin-binding epidermal growth factor-like growth factor (Fig. 7A, VEGFand HBEGF), implying a cause and effect. On the other hand, fourHOX genes were down-regulated (Supplementary Table 4). The HOX

genes are sequentially activated during embryogenesis, setting inmotion correct embryonic patterning (see, for example, Ref. 74),suggesting that the down-regulation in melanoma may be the reversalof this process.

The Affymetrix data showed that Twist was up-regulated in five ofsix melanoma cell strains and displayed a pattern of expression similarto that of TBX3 (Fig. 7B). Western blot analysis revealed that TWIST

Fig. 3. Up-regulation of insulin-like growth fac-tor-binding protein (IGFBs). The histogram showsthat six members of the IGFBP group (MolecularFunction ontology identifier 5520) were signifi-cantly up-regulated. Inset, higher magnification tobetter demonstrate the differences in the expressionof IGFBs.

Fig. 4. Differential expression of cancer-testis antigen group. Signal intensity datacorresponding to the cancer-testis antigens were subject to QT_Clust analysis in Gene-Spring 6.0. The results show different patterns of expression for MAGE (A) and GAGE(B) family members. Notice that FGF13 genes expression pattern is clustered within theMAGE group. C, expression of FGF13 gene transcript compared with �-actin using 30cycles of reverse transcription-PCR. Numbers on the left indicate size markers of a 1-kbDNA ladder.

5275




was up-regulated in five of nine melanoma cell strains (Fig. 7C). Hereagain, there was no direct correlation between the Affymetrix inten-sity values and the levels of the protein in two melanoma cell strains(YUHEIK and YUSIT1). Probing of the melanoma TMA with anti-Twist antibodies indicated that Twist overexpression was associatedwith worse patient survival in the entire cohort (P � 0.0037) andwithin the subset of primary melanomas (P � 0.0028; Fig. 7, D andE), highlighting the significance of its involvement in tumorigenesisin vivo. Intense Twist staining was more prominent in metastaticlesions (including nodal and distant metastases) than in primarylesions (P � 0.0053; Fig. 7E). Among the primary lesions, there wasno significant association between high Twist expression and Clarklevel, Breslow depth, presence of ulceration in the lesion, patient age,or gender. On multivariate analysis using the Cox model, high Twistexpression was an independent predictor of poor outcome (hazardratio, 3.2; 95% confidence interval, 1.16–8.99; P � 0.024), as wereBreslow depth of �1 mm and a Clark level of IV or V. Other variablesincluded in this analysis that were not of significant predictive valuewere presence of ulceration on primary tumors, patient age, andgender.

Three genes encoding growth suppressors, NECDIN (neurally dif-ferentiated embryonal carcinoma cell-derived factor), NBL1 (neuro-blastoma suppressor of tumorigenicity 1 precursor), and CIRBP (coldinducible RNA-binding protein) were also down-regulated in mela-noma cells (Supplementary Table 5), but in an uncoordinated manner(Fig. 8A). Reduced Necdin expression was confirmed by Westernblotting in seven of nine melanoma cell strains (Fig. 8, B and D), butattempts to stain melanoma TMA failed due to the nonspecific bind-

ing of the anti-Necdin antibodies to the tissues on the microarray (datanot shown). Overexpression of Flag-tagged Necdin in melanoma cells(501 mel cells, Fig. 8C, inset) suppressed growth, as observed by a30% reduction in [3H]thymidine incorporation 4 days after transfec-tion (data not shown) and by a cell count determined 10 days aftertransfection (Fig. 8C). Immunofluorescence analysis revealed thatendogenous necdin was cytoplasmic in normal and malignant cells butdisplayed a patchy and aggregated pattern in melanoma comparedwith even distribution in normal pigment cells (Fig. 8E). However,sequencing NECDIN in 501 mel cells did not indicate any mutation,suggesting that the different pattern was due to other factors.

DISCUSSION

We report here a comprehensive analysis of changes in geneexpression in human melanoma cells compared with normal melano-cytes at a scope not demonstrated previously. These changes wereobserved by analyzing proliferating melanoma cells from six ad-vanced lesions and normal human melanocytes grown in culture. Themelanoma cells were selected to represent strains with high and lowlevels of pigmentation expressing a wide range of tyrosinase, themelanocyte-specific gene product, as a measure of differentiation.Several of the differences between normal and malignant melanocytesdetected by our analysis have already been reported in cells andtumors, supporting the notion that most of them were not generatedduring growth in culture and do not reflect a response to ingredientsin the culture medium. The use of proliferating normal and malignantmelanocytes proved that several critical changes in gene expression

Fig. 5. Differential expression of genes in theimmune modulation pathway. A, QT_Clust analy-sis of the insulin-responsive genes. B, Westernblots of cell extracts with antibodies against MX1/IFI78, ISG15, and Rab32, using �-actin as a con-trol. NT, not tested. C, coordinated expression ofleukemia inhibitory factor (LIF) and HLA (MHCclass II types as indicated) genes as determined byQT_Clust analysis. D, coordinated down-regula-tion of genes in the vesicular pathway. VAMP8,vesicle-associated membrane protein 8 (endobre-vin). TRS20 is also known as LOC51693; RAB38,RAB27A, RAB32, RAB33A, and RAB31, Ras-related proteins. E, expression of Rab33A genetranscript compared with �-actin using 35 cycles ofreverse transcription-PCR. Numbers on the left in-dicate size markers of a 1-kb DNA ladder.

5276




characteristic to the malignant state are independent of proliferationstatus.

Hierarchical clustering showed that the primary melanoma cellstrain WW165 isolated from an advanced lesion was more similar tothe metastatic melanoma cells than to normal melanocytes, confirm-ing an observation made almost two decades ago (75). Indeed, theWW165 cells, unlike primary melanoma cells from superficial spread-ing or thin lesions, proliferate in culture independently of most of thegrowth factors required for normal human melanocytes (76). TheMNT1 melanoma cells, on the other hand, isolated from lymph nodemetastasis, were the least divergent from normal human melanocytes.The similarity included not only high pigmentation, which was ob-served in at least two other melanoma cell strains (YUHEIK andYUMAC), but also genes in the receptor activity group, such asTNC1, TNFRSF12A, TIMP3, ITGA6, SDC2, and in the Rab family(Rab38 and Rab27A). MNT1 is a melanoma cell strain that wasestablished in culture more than a decade ago (29), and informationregarding the patient’s clinical outcome is not available.

Although extensive activation of cell surface receptors in mela-noma cells has been reported previously, our analysis revealed severalnovel details. We show, for example, the increased expression ofFGF13 as a possible mediator of FGF receptor 1 stimulation. FGF13was originally cloned from an ovarian cancer cell library, and itsexpression is particularly high in the brain. The growth factor is likelyto be mitogenic for the melanoma cells because it stimulates themitogen-activated protein kinase pathway (Ref. 77 and the referencestherein).

We also demonstrated that several components of the integrinpathway are overexpressed in a synergistic manner, which include

integrins (ITGA7, ITGA4, ITGA6, ITGB3), integrin ligands (Cyr61,TNC1, FN1, CTGF), and two tetraspanins (TM4SF13 and TM4SF1),and demonstrated down-regulation of the serine/threonine phospha-tase ILKAP. ILKAP selectively suppresses the activity of integrin-linked kinase 1 (78), a mediator of integrin signal transduction thatpromotes angiogenesis (see, for example, Ref. 79). More recently, itwas demonstrated that integrin-linked kinase 1 protein expressionincreases as melanoma progresses, with strong positive correlationbetween expression and lymph node invasion and with decreased5-year survival (80). Suppression of ILKAP by small interfering RNAincreased entry of cells into S phase, and overexpression of ILKAPinhibited anchorage-independent growth of LNCaP prostate carci-noma cells (81). Reduced ILKAP expression, therefore, in combina-tion with high integrin-linked kinase expression, is likely to enhancethe aggressive state of melanomas.

Other changes in the integrin pathway shown here have beenobserved previously in melanoma tumors, such as Tenascin C (82–84). Tenascin C is also present in the serum of patients with stage IVmelanoma, but not in healthy donors (85), suggesting that it can beused as a marker for early transformation and metastatic load. Cyr61,on the other hand, is expressed in uveal melanomas (86). It is anangiogenic factor that induces migration and stimulates mitogenesisby interacting with integrins (87). Cyr61 confers resistance to apo-ptosis by increasing nuclear factor-�B activity (88), and the expres-sion of Cyr61 and CTGF, in addition to vascular endothelial growthfactor, is likely to contribute to the angiogenic activity of melanomacells. These three factors are produced by other cancer cells as welland are targets for tumor therapy [see, for example, Ref. 89; Lester F.Lau (University of Illinois, Chicago, IL and FibroGen, South San

Fig. 6. Differential expression of genes in theproteasomal pathway. A, Affymetrix hybridizationintensity pattern for grail (E) and UCHL1 (F).Intensity scales on the left and right are for grailand UCHL1, respectively. B, Western blots of cellextracts with anti-grail and anti-UCHL1 comparedwith �-actin control. NT, not tested. C and D,UCHL1 Kaplan-Meier survival curves over a 30-year period. Results are based on immunostainingof melanoma tissue microarray with anti-UCHL1polyclonal antibodies. The entire cohort of mela-noma patients (C) and primary melanoma lesionsonly (D) are displayed separately. E, histospotsshowing cytoplasmic staining of UCHL1 with high(score of 3) and low (score of 0) expression scores.Pictures were taken at �10 magnification (insets,�60).

5277




Francisco, CA)]. TM4SF1 is also being evaluated as a diagnosticmarker for colorectal and gastric cancers and as a target for therapywith monoclonal antibodies (90, 91).

We demonstrated overexpression at the gene transcript level ofNOTCH2 and HEY1, the latter of which is a downstream effector ofthe NOTCH receptor. The increased HEY1/HERS1 expression in theleast divergent melanocytic lesions, i.e., MNT1 and the primarymelanoma WW165, suggests a role in early transformation. Based ona limited microarray analysis, Seykora et al. (92) also found increasedNOTCH2 expression in two advanced primary melanoma tumorscompared with benign nevi. However, ectopic expression of HEY1/HERS1 in immortalized benign mouse melanocytes did not releasethe cells from their dependence on the external growth factor TPA,suggesting that HEY1/HERS1, by itself, is not sufficient to confermalignancy. This observation is in agreement with previously pub-lished results showing that collaborating genetic events are requiredfor tumorigenic transformation by NOTCH, in particular Ras-medi-ated activation of Erk/mitogen-activated protein kinase and phosphati-dylinositol 3�-kinase (Ref. 93; reviewed in Ref. 94).

A surprising finding was the extent of changes in genes whose

products affect immune response. In particular, we observed ex-tensive suppression of IFN-regulated genes (21 of 22 genes; 14were suppressed in a coordinated manner). Among the down-regulated genes was OAS-1. OAS-1 activity is required for RNaseendoribonuclease activation and the antiviral and apoptotic activityof IFNs (Refs. 95 and 96; reviewed in Refs. 97 and 98). Severalmutations in RNase endoribonuclease are associated with prostatecancer (98), and IFN-responsive genes are down-regulated in colonpolyps and tumors as compared with normal colon tissue (99, 100).The repression of IFN signaling genes might reflect immortaliza-tion because the predominant genes silenced by DNA methylationin spontaneously immortal Li-Fraumeni fibroblasts belonged tothis pathway (46% of 85 genes; Ref. 101). The coordinated down-regulation of IFN-responsive genes is of particular importancebecause IFN �2� is frequently administered as an adjuvant inimmunotherapy for melanoma (102, 103). The effectiveness of thistreatment is likely to be conditional on the ability to induce theexpression of at least some of these genes.

We also demonstrated that genes in the ubiquitin pathway weremostly down-regulated (12 of 14 genes). One of them, UCHL1, was

Fig. 7. Expression pattern of transcriptional reg-ulators and their affected genes. A and B, QT_Clustanalysis. FOXD1, forkhead box D1; HBEGF,heparin-binding epidermal growth factor-likegrowth factor (also known as diphtheria toxin re-ceptor); VEGF, vascular endothelial growth factor;TBX3, T-box 3 (ulnar mammary syndrome);TWIST, twist homolog (acrocephalosyndactyly 3;Saethre-Chotzen syndrome). C, anti-Twist Westernblot of cell extracts, using �-actin as a control. Dand E, Twist Kaplan-Meier survival curves basedon immunostaining of melanoma tissue microarraywith anti-Twist polyclonal antibodies. The entirecohort of melanoma patients (D) and primary mel-anoma lesions only (E) are displayed separately. F,histospots showing cytoplasmic staining of Twistwith high (score of 3) and low (score of 0) expres-sion scores. Pictures were taken at �10 magnifi-cation (insets, �60).

5278




suppressed in all of the melanoma strains tested, and its expressionwas also reduced in metastatic melanoma lesions, as seen by probingthe melanoma TMA. Reduction in UCHL1 in primary melanomas wassignificantly associated with poor survival. UCHL1 has a dual func-tion, a hydrolase activity that removes small COOH-terminal ubiq-uitin to generate the ubiquitin monomer and a dimerization-dependentubiquityl ligase activity. Several UCHL1 variants associated withParkinson disease may play roles in proteasomal protein degradation(reviewed in Ref. 104). The UCHL1 enzymatic activity most conse-quential for melanocyte tumorigenesis is not yet clear.

Twist was up-regulated in �55% of the melanoma cell strainstested, and its overexpression in melanoma tumors correlated withpoor survival. Twist is a basic helix-loop-helix transcription factorthat is required for mesodermal cell fates, and its pattern of expressionin melanomas was similar to that of TBX3, a transcriptional repressorrequired for normal breast development, with mutations in the TBX3gene linked to ulnar-mammary syndrome (Ref. 105 and the referencestherein). TBX3 and Twist suppress senescence by down-regulatingARF, a protein that inhibits MDM2-mediated proteolysis of p53(106–108). Twist also blocks myogenic differentiation and is inap-propriately expressed in 50% of rhabdomyosarcomas (109). In oursamples, there was no correlation between Twist expression anddifferentiation, as determined by the levels of tyrosinase and pigmen-tation. In another report, increased Twist protein levels were associ-ated with resistance to Taxol and vincristine in nasopharyngeal, blad-der, ovarian, and prostate cancers (110). We showed that Twist

expression correlated with poor survival, and we can only speculatethat it contributes to melanoma resistance to this type of drug.

Necdin was one of the growth suppressors down-regulated in six ofeight melanoma cell strains tested. Low expression in melanomas canstill confer advantage for tumor growth, as shown for other tumorsuppressors. For example, tumors developed in animals lacking onecopy of p27Kip1 expressing half the levels of functional wild typep27Kip1 (111). Similar observations were gleaned from p53, and thephenomenon of gene dosage sensitivity is termed haploinsufficiency(Ref. 112 and the references therein). Furthermore, Necdin in mela-noma cells is localized to a different cytoplasmic compartment com-pared with normal melanocytes, suggesting that it is inactivated bysequestration. Its growth-suppressive function was validated by ec-topic overexpression that led to rapid and efficient arrest of melanomacell proliferation. Necdin is expressed predominantly in postmitoticcells, such as neurons and muscle cells (Ref. 113 and the referencestherein), and in melanocytes, as demonstrated here for the first time.NECDIN is on chromosomal region 15q11.2-q12, and is one of fourgenes encoding the Prader-Willi syndrome. This is an autosomaldominant neurodevelopmental disorder caused by mutations or dele-tions because the respective genes on the maternal chromosome in thisregion are inactive through imprinting. In the case of melanoma, theexpressed necdin was not mutated in the coding region. The exactfunction of necdin has not yet been determined. It binds to thetranscription factors E2F1 and E2F4 and to p75 neurotrophin receptor(113). Ectopic coexpression of necdin with E2F1 in U2OS osteosar-coma cells suppressed E2F1-stimulated transcription. On the otherhand, coexpression of necdin with E2F1 or p75 neurotrophin receptorin N1E-115 neuroblastoma cells sequestered the protein in the nucleusor the cell membrane, respectively (113). Therefore, identifying thenecdin-binding proteins in melanoma cells and normal melanocytes islikely to shed light on the molecular basis for the growth-arrestingfunction of necdin in this cellular system.

Some of the down-regulated genes are localized at chromosomalregions known to be altered in melanomas. For example, NBL1 is on1p36.3-p36.2, and GBP1 (guanylate-binding protein 1, IFN-inducible,67 kDa) is on 1p22.2, regions harboring melanoma suppressor genes(114). Another example is RRAGA (Ras-related GTP-binding pro-tein) on 9p21.3, a region frequently deleted in melanomas that maypossess more then one tumor suppressor gene (115). However, se-quencing of RRAGA in normal melanocytes and several melanomacell strains failed to identify any mutation, suggesting that if down-regulation is due to loss of chromosomal material, then the remaininggene is normal.

Expression or reexpression of some genes might be driven bydemethylation/methylation of selected CpG in the promoter regions,processes common in cancer cells (reviewed in Ref. 116). Althoughthe majority of the activated genes on the X chromosome encodedCTAs known to be reexpressed due to demethylation (reviewed inRefs. 61–63), other X-linked genes were also activated, such asTRAG3 (Taxol resistance-associated gene 3; Xq28) and FGF13(Xq26.3), and might be affected by the same process. DNA methyl-ation/demethylation processes of promoter CpG-rich regions mightcause the change in expression of HOXB genes (HOXB2, HOXB7,and HOXB13; 17q21.2; Ref. 117), MX1/IFI78 (12q24.2; Ref. 101),and dipeptidylpeptidase IV (2q24.3; Ref. 99; reviewed in Refs. 100and 118–120). Indeed, preliminary sequence analysis of the promoterregion and exon 1 using the MethPrime program of Li and Dahiya(121) available online15 revealed that 64% (of 35 genes) and 70% (of48 genes) of the top down- and up-regulated genes, respectively,

15 http://itsa.ucsf.edu/�urolab/methprimer/index1.html.

Fig. 8. Expression patterns of transcriptional repressors and functional validation ofnecdin. A, Affymetrix intensity data. NDN, necdin; NBL1, neuroblastoma, suppression oftumorigenicity 1; CIRBP, cold inducible RNA-binding protein. B, anti-necdin Westernblot of cell extracts, using �-actin as a control. C, a histogram showing the number ofG418-resistant pcDNAFlag.Necdin- and pcDNA.GFP-transfected cells assessed 10 daysafter transfection. Inset, anti-Flag Western blot of cell lysates derived from pcDNA.GFP-and Flag.Necdin-expressing cells 1 day after transfection. D, anti-necdin Western blot ofnormal melanocytes and 501 mel melanoma cell extracts. E, necdin immunofluorescentanalysis of normal human melanocytes and 501 mel melanoma cells using anti-necdinantibodies.

5279




contain at least one CpG-rich island (data not shown). MonitoringCpG methylation patterns and the genes they affect can be used forassessing tumor progression, drug resistance, and immunological re-sponses. This information is particularly needed to assess drug effi-cacy of novel DNA demethylating agents, such as 5-aza-2-deoxycy-tidine (commercially known as decitabine), antisense to DNAmethyltransferases such as DNMT1 (MG98), and histone deacetylaseinhibitor depsipeptide FR901228, which are in Phase II trials (see, forexample, Refs. 122 and 123; reviewed in Ref. 124).

In summary, we have identified a number of novel genes andpathways that are up- or down-regulated in metastatic melanoma cellstrains compared with normal melanocytes. The data provide newinsights into immune modulation and possible causes for failed IFN orchemotherapy treatments common in melanomas. We identifiedFGF13 as a new autocrine factor, Twist as clinically valuable prog-nostic markers for patients with melanomas, and Necdin as a potentmelanoma growth suppressor. We demonstrated coordinated expres-sion of genes representing several functional groups, including recep-tor activity, IFN-responsive pathway, CTAs, Rab family members,and genes involved in embryonic tissue remodeling. These data canserve as a basis for further evaluation of pathways leading to malig-nant transformation, prognostic markers, and/or drug targets.

ACKNOWLEDGMENTS

We thank Dr. Mark A. McNiven for the anti-MX1/MxA antibodies; Dr.Ernest C. Borden for the anti-ISG15 antibodies; Dr. R. J. Haslam for theanti-RAB32 antibodies; Drs. Timothy M. Wright and Bo Hu for the HEY1/HESR1-pIRES2-EGFP plasmid; Dr. Shrikant Mane (Affymetrix Resource) forhelp in conducting the microarray experiments; Dr. Janet Hager (MicroarrayResource) for critical review of the raw OHU16K data; Donna LaCivita, whois in charge of the Cell Culture Core facility at the Yale Skin Disease ResearchCenter, for the normal human melanocytes and melanoma cells; and the YaleMelanoma Unit for discussions and support.

REFERENCES

1. Cannon-Albright LA, Kamb A, Skolnick M. A review of inherited predisposition tomelanoma. Semin Oncol 1996;23:667–72.

2. Dracopoli NC, Fountain JW. CDKN2 mutations in melanoma. Cancer Surv 1996;26:115–32.

3. Kamb A, Gruis NA, Weaver-Feldhaus J, et al. A cell cycle regulator potentiallyinvolved in genesis of many tumor types. Science (Wash DC) 1994;264:436–40.

4. Kamb A, Shattuck-Eidens D, Eeles R, et al. Analysis of the p16 gene (CDKN2) asa candidate for the chromosome 9p melanoma susceptibility locus. Nat Genet1994;8:23–6.

5. Tucker MA, Goldstein AM. Melanoma etiology: where are we? Oncogene 2003;22:3042–52.

6. van der Velden PA, Metzelaar-Blok JA, Bergman W, et al. Promoter hypermethy-lation: a common cause of reduced p16INK4a expression in uveal melanoma. CancerRes 2001;61:5303–6.

7. Reed JA, Loganzo F, Shea CR, et al. Loss of expression of the p16/cyclin-dependentkinase inhibitor 2 tumor suppressor gene in melanocytic lesions correlates withinvasive stage of tumor progression. Cancer Res 1995;55:2713–8.

8. Piccinin S, Doglioni C, Maestro R, et al. p16/CDKN2 and CDK4 gene mutations insporadic melanoma development and progression. Int J Cancer 1997;74:26–30.

9. Alonso SR, Ortiz P, Pollan M, et al. Progression in cutaneous malignant melanomais associated with distinct expression profiles: a tissue microarray-based study. Am JPathol 2004;164:193–203.

10. Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in human cancer.Nature (Lond) 2002;417:949–54.

11. Alsina J, Gorsk DH, Germino FJ, et al. Detection of mutations in the mitogen-activated protein kinase pathway in human melanoma. Clin Cancer Res 2003;9:6419–25.

12. Halaban R, Cheng E, Smicun Y, Germino J. Deregulated E2F transcriptional activityin autonomously growing melanoma cells. J Exp Med 2000;191:1005–15.

13. Seftor RE, Seftor EA, Hendrix MJ. Molecular role(s) for integrins in humanmelanoma invasion. Cancer Metastasis Rev 1999;18:359–75.

14. Kaelin WG Jr. E2F1 as a target: promoter-driven suicide and small moleculemodulators. Cancer Biol Ther 2003;2:S48–54.

15. Ma D, Zhou P, Harbour JW. Distinct mechanisms for regulating the tumor suppres-sor and antiapoptotic functions of Rb. J Biol Chem 2003;278:19358–66.

16. Bastian BC, LeBoit PE, Hamm H, Brocker EB, Pinkel D. Chromosomal gains andlosses in primary cutaneous melanomas detected by comparative genomic hybrid-ization. Cancer Res 1998;58:2170–5.

17. Bastian BC, Kashani-Sabet M, Hamm H, et al. Gene amplifications characterizeacral melanoma and permit the detection of occult tumor cells in the surroundingskin. Cancer Res 2000;60:1968–73.

18. Harvell JD, Bastian BC, LeBoit PE. Persistent (recurrent) spitz nevi: a histopatho-logic, immunohistochemical, and molecular pathologic study of 22 cases. Am J SurgPathol 2002;26:654–61.

19. Sauter ER, Yeo UC, von Stemm A, et al. Cyclin D1 is a candidate oncogene incutaneous melanoma. Cancer Res 2002;62:3200–6.

20. Bastian BC, Xiong J, Frieden IJ, et al. Genetic changes in neoplasms arising incongenital melanocytic nevi: differences between nodular proliferations and mela-nomas. Am J Pathol 2002;161:1163–9.

21. Balazs M, Adam Z, Treszl A, et al. Chromosomal imbalances in primary andmetastatic melanomas revealed by comparative genomic hybridization. Cytometry2001;46:222–32.

22. Bastian BC, Olshen AB, LeBoit PE, Pinkel D. Classifying melanocytic tumors basedon DNA copy number changes. Am J Pathol 2003;163:1765–70.

23. Bastian BC. Understanding the progression of melanocytic neoplasia using genomicanalysis: from fields to cancer. Oncogene 2003;22:3081–6.

24. Takata M, Maruo K, Kageshita T, et al. Two cases of unusual acral melanocytictumors: illustration of molecular cytogenetics as a diagnostic tool. Hum Pathol2003;34:89–92.

25. Bittner M, Meltzer P, Chen Y, et al. Molecular classification of cutaneous malignantmelanoma by gene expression profiling. Nature (Lond) 2000;406:536–40.

26. Weeraratna AT, Jiang Y, Hostetter G, et al. Wnt5a signaling directly affects cellmotility and invasion of metastatic melanoma. Cancer Cell 2002;1:279–88.

27. Weeraratna AT, Becker D, Carr KM, et al. Generation and analysis of melanomaSAGE libraries: SAGE advice on the melanoma transcriptome. Oncogene 2004;23:2264–74.

28. von Willebrand M, Zacksenhaus E, Cheng E, Glazer P, Halaban R. The tyrphostinAG1024 accelerates the degradation of phosphorylated forms of retinoblastomaprotein (pRb) and restores pRb tumor suppressive function in melanoma cells.Cancer Res 2003;63:1420–9.

29. Cuomo M, Nicotra MR, Apollonj C, et al. Production and characterization of themurine monoclonal antibody 2G10 to a human T4-tyrosinase epitope. J InvestigDermatol 1991;96:446–51.

30. Zakut R, Perlis R, Eliyahu S, et al. KIT ligand (mast cell growth factor) inhibits thegrowth of KIT-expressing melanoma cells. Oncogene 1993;8:2221–9.

31. Halaban R, Miglarese MR, Smicun Y, Puig S. Melanomas, from the cell cycle pointof view. Int J Mol Med 1998;1:419–25.

32. Halaban R. The regulation of normal melanocyte proliferation. Pigment Cell Res2000;13:4–14.

33. Tamura A, Halaban R, Moellmann G, et al. Normal murine melanocytes in culture.In Vitro Cell Dev Biol 1987;23:519–22.

34. Giambernardi TA, Rodeck U, Klebe RJ. Bovine serum albumin reverses inhibitionof RT-PCR by melanin. Biotechniques 1998;25:564–6.

35. Eisen MB, Brown PO. DNA arrays for analysis of gene expression. MethodsEnzymol 1999;303:179–205.

36. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical andpowerful approach to multiple testing. J R Stat Soc B 1995;57:289–300.

37. Heyer LJ, Kruglyak S, Yooseph S. Exploring expression data: identification andanalysis of coexpressed genes. Genome Res 1999;9:1106–15.

38. Ausubel FM, Brent R, Kingston R, Moore RE, Moore DD, Seidman JG, Smith JA,Struhl K. Current protocols in molecular biology. New York: John Wiley & Sons;1995.

39. Niinobe M, Koyama K, Yoshikawa K. Cellular and subcellular localization ofnecdin in fetal and adult mouse brain. Dev Neurosci 2000;22:310–9.

40. Camp RL, Chung GG, Rimm DL. Automated subcellular localization and quanti-fication of protein expression in tissue microarrays. Nat Med 2002;8:1323–8.

41. Kielhorn E, Provost E, Olsen D, et al. Tissue microarray-based analysis showsphospho-beta-catenin expression in malignant melanoma is associated with pooroutcome. Int J Cancer 2003;103:652–6.

42. Berger AJ, Kluger HM, Li N, et al. Subcellular localization of activating transcrip-tion factor 2 (ATF2) in melanoma specimens predicts patient survival. Cancer Res2003;63:8103–7.

43. Henderson AM, Wang SJ, Taylor AC, Aitkenhead M, Hughes CC. The basichelix-loop-helix transcription factor HESR1 regulates endothelial cell tube forma-tion. J Biol Chem 2001;276:6169–76.

44. Hu B, Wang S, Zhang Y, et al. A nuclear target for interleukin-1alpha: interactionwith the growth suppressor necdin modulates proliferation and collagen expression.Proc Natl Acad Sci USA 2003;100:10008–13.

45. Halaban R, Patton RS, Cheng E, et al. Abnormal acidification of melanoma cellsinduces tyrosinase retention in the early secretory pathway. J Biol Chem 2002;277:14821–8.

46. Padovan E, Terracciano L, Certa U, et al. Interferon stimulated gene 15 constitu-tively produced by melanoma cells induces e-cadherin expression on human den-dritic cells. Cancer Res 2002;62:3453–8.

47. Van Belle PA, Elenitsas R, Satyamoorthy K, et al. Progression-related expression ofbeta3 integrin in melanomas and nevi. Hum Pathol 1999;30:562–7.

48. Easty DJ, Herlyn M, Bennett DC. Abnormal protein tyrosine kinase gene expressionduring melanoma progression and metastasis. Int J Cancer 1995;60:129–36.

49. Reed JA, McNutt NS, Prieto VG, Albino AP. Expression of transforming growthfactor-beta 2 in malignant melanoma correlates with the depth of tumor invasion.Implications for tumor progression. Am J Pathol 1994;145:97–104.

50. Bodey B, Kaiser HE, Goldfarb RH. Immunophenotypically varied cell subpopula-tions in primary and metastatic human melanomas. Monoclonal antibodies for

5280




diagnosis, detection of neoplastic progression and receptor directed immunotherapy.Anticancer Res 1996;16:517–31.

51. Easty DJ, Bennett DC. Protein tyrosine kinases in malignant melanoma. MelanomaRes 2000;10:401–11.

52. Garbe C, Leiter U, Ellwanger U, et al. Diagnostic value and prognostic significanceof protein S-100beta, melanoma-inhibitory activity, and tyrosinase/MART-1 reversetranscription-polymerase chain reaction in the follow-up of high-risk melanomapatients. Cancer (Phila) 2003;97:1737–45.

53. Li G, Herlyn M. Dynamics of intercellular communication during melanoma devel-opment. Mol Med Today 2000;6:163–9.

54. Herlyn M, Berking C, Li G, Satyamoorthy K. Lessons from melanocyte develop-ment for understanding the biological events in naevus and melanoma formation.Melanoma Res 2000;10:303–12.

55. Wesley UV, Albino AP, Tiwari S, Houghton AN. A role for dipeptidyl peptidase IVin suppressing the malignant phenotype of melanocytic cells. J Exp Med 1999;190:311–22.

56. Maier MM, Gessler M. Comparative analysis of the human and mouse Hey1promoter: Hey genes are new Notch target genes. Biochem Biophys Res Commun2000;275:652–60.

57. Halaban R, Kwon BS, Ghosh S, Delli Bovi P, Baird A. bFGF as an autocrine growthfactor for human melanomas. Oncogene Res 1988;3:177–86.

58. Albino AP, Davis BM, Nanus D. Induction of growth factor RNA expression inhuman malignant melanoma: markers of transformation. Cancer Res 1991;51:4815–20.

59. Yayon A, Ma YS, Safran M, Klagsbrun M, Halaban R. Suppression of autocrine cellproliferation and tumorigenesis of human melanoma cells and fibroblast growthfactor transformed fibroblasts by a kinase-deficient FGF receptor 1: evidence for theinvolvement of Src- family kinases. Oncogene 1997;14:2999–3009.

60. Becker D, Lee PL, Rodeck U, Herlyn M. Inhibition of the fibroblast growth factorreceptor 1 (FGFR-1) gene in human melanocytes and malignant melanomas leads toinhibition of proliferation and signs indicative of differentiation. Oncogene 1992;7:2303–13.

61. Scanlan MJ, Gure AO, Jungbluth AA, Old LJ, Chen YT. Cancer/testis antigens: anexpanding family of targets for cancer immunotherapy. Immunol Rev 2002;188:22–32.

62. Sigalotti L, Coral S, Nardi G, et al. Promoter methylation controls the expression ofMAGE2, 3 and 4 genes in human cutaneous melanoma. J Immunother 2002;25:16–26.

63. Coral S, Sigalotti L, Altomonte M, et al. 5-Aza-2�-deoxycytidine-induced expressionof functional cancer testis antigens in human renal cell carcinoma: immunothera-peutic implications. Clin Cancer Res 2002;8:2690–5.

64. Marincola FM, Wang E, Herlyn M, Seliger B, Ferrone S. Tumors as elusive targetsof T-cell-based active immunotherapy. Trends Immunol 2003;24:335–42.

65. Metcalf D. The unsolved enigmas of leukemia inhibitory factor. Stem Cells 2003;21:5–14.

66. Mori M, Yamaguchi K, Honda S, et al. Cancer cachexia syndrome developed innude mice bearing melanoma cells producing leukemia-inhibitory factor. CancerRes 1991;51:6656–9.

67. Paglia D, Oran A, Lu C, et al. Expression of leukemia inhibitory factor andinterleukin-11 by human melanoma cell lines: LIF, IL-6, and IL-11 are not coregu-lated. J Interferon Cytokine Res 1995;15:455–60.

68. Bamberger AM, Jenatschke S, Schulte HM, Loning T, Bamberger MC. Leukemiainhibitory factor (LIF) stimulates the human HLA-G promoter in JEG3 choriocar-cinoma cells. J Clin Endocrinol Metab 2000;85:3932–6.

69. Dell’Angelica EC. Melanosome biogenesis: shedding light on the origin of anobscure organelle. Trends Cell Biol 2003;13:503–6.

70. Seabra MC, Mules EH, Hume AN. Rab GTPases, intracellular traffic and disease.Trends Mol Med 2002;8:23–30.

71. Hammer JA III, Wu XS. Rabs grab motors: defining the connections between RabGTPases and motor proteins Curr Opin Cell Biol 2002;14:69–75.

72. Lukashok SA, Tarassishin L, Li Y, Horwitz MS. An adenovirus inhibitor of tumornecrosis factor alpha-induced apoptosis complexes with dynein and a small GTPase.J Virol 2000;74:4705–9.

73. Zhang H, Palmer R, Gao X, et al. Transcriptional activation of placental growthfactor by the forkhead/winged helix transcription factor FoxD1. Curr Biol 2003;13:1625–9.

74. Forlani S, Lawson KA, Deschamps J. Acquisition of Hox codes during gastrulationand axial elongation in the mouse embryo. Development (Cambr) 2003;130:3807–19.

75. Herlyn M, Balaban G, Bennicelli J, et al. Primary melanoma cells of the verticalgrowth phase: similarities to metastatic cells. J Natl Cancer Inst (Bethesda) 1985;74:283–9.

76. Halaban R, Ghosh S, Duray P, Kirkwood JM, Lerner AB. Human melanocytescultured from nevi and melanomas. J Investig Dermatol 1986;87:95–101.

77. Greene JM, Li YL, Yourey PA, et al. Identification and characterization of anovel member of the fibroblast growth factor family. Eur J Neurosci 1998;10:1911–25.

78. Leung-Hagesteijn C, Mahendra A, Naruszewicz I, Hannigan GE. Modulation ofintegrin signal transduction by ILKAP, a protein phosphatase 2C associating withthe integrin-linked kinase, ILK1. EMBO J 2001;20:2160–70.

79. Tan C, Cruet-Hennequart S, Troussard A, et al. Regulation of tumor angiogenesis byintegrin-linked kinase (ILK). Cancer Cell 2004;5:79–90.

80. Dai DL, Makretsov N, Campos EI, et al. Increased expression of integrin-linkedkinase is correlated with melanoma progression and poor patient survival. ClinCancer Res 2003;9:4409–14.

81. Kumar AS, Naruszewicz I, Wang P, Leung-Hagesteijn C, Hannigan GE. ILKAPregulates ILK signaling and inhibits anchorage independent growth. Oncogene2004;23:3454–61.

82. Tuominen H, Kallioinen M. Increased tenascin expression in melanocytic tumors. JCutan Pathol 1994;21:424–9.

83. Tuominen H, Pollanen R, Kallioinen M. Multicellular origin of tenascin in skintumors–an in situ hybridization study. J Cutan Pathol 1997;24:590–6.

84. Clark EA, Golub TR, Lander ES, Hynes RO. Genomic analysis of metastasis revealsan essential role for RhoC. Nature (Lond) 2000;406:532–5.

85. Burchardt ER, Hein R, Bosserhoff AK. Laminin, hyaluronan, tenascin-C and typeVI collagen levels in sera from patients with malignant melanoma. Clin ExpDermatol 2003;28:515–20.

86. Walker TM, Van Ginkel PR, Gee RL, et al. Expression of angiogenic factors Cyr61and tissue factor in uveal melanoma. Arch Ophthalmol 2002;120:1719–25.

87. Kunz M, Moeller S, Koczan D, et al. Mechanisms of hypoxic gene regulation ofangiogenesis factor Cyr61 in melanoma cells. J Biol Chem 2003;25:45651–60.

88. Lin MT, Chang CC, Chen ST, et al. Cyr61 expression confers resistance to apoptosisin breast cancer MCF-7 cells by a mechanism of NF-kappaB-dependent XIAPup-regulation. J Biol Chem 2004;279:24015–23.

89. Huang J, Frischer JS, Serur A, et al. Regression of established tumors and metastasesby potent vascular endothelial growth factor blockade. Proc Natl Acad Sci USA2003;100:7785–90.

90. Schiedeck TH, Wellm C, Roblick UJ, Broll R, Bruch HP. Diagnosis and monitoringof colorectal cancer by L6 blood serum polymerase chain reaction is superior tocarcinoembryonic antigen-enzyme-linked immunosorbent assay. Dis Colon Rectum2003;46:818–25.

91. Kaneko R, Tsuji N, Kamagata C, et al. Amount of expression of the tumor-associated antigen L6 gene and transmembrane 4 superfamily member 5 gene ingastric cancers and gastric mucosa. Am J Gastroenterol 2001;96:3457–8.

92. Seykora JT, Jih D, Elenitsas R, Horng WH, Elder DE. Gene expression profiling ofmelanocytic lesions. Am J Dermatopathol 2003;25:6–11.

93. Fitzgerald K, Harrington A, Leder P. Ras pathway signals are required for notch-mediated oncogenesis. Oncogene 2000;19:4191–8.

94. Nam Y, Aster JC, Blacklow SC. Notch signaling as a therapeutic target. Curr OpinChem Biol 2002;6:501–9.

95. Carr DJ, Al-khatib K, James CM, Silverman R. Interferon-beta suppresses herpessimplex virus type 1 replication in trigeminal ganglion cells through an RNaseL-dependent pathway. J Neuroimmunol 2003;141:40–6.

96. Al-khatib K, Williams BR, Silverman RH, Halford W, Carr DJ. The murinedouble-stranded RNA-dependent protein kinase PKR and the murine 2�,5�-oligoad-enylate synthetase-dependent RNase L are required for IFN-beta-mediated resist-ance against herpes simplex virus type 1 in primary trigeminal ganglion culture.Virology 2003;313:126–35.

97. Samuel CE. Antiviral actions of interferons. Clin Microbiol Rev 2001;14:778–809.98. Silverman RH. Implications for RNase L in prostate cancer biology. Biochemistry

2003;42:1805–12.99. Karpf AR, Jones DA. Reactivating the expression of methylation silenced genes in

human cancer. Oncogene 2002;21:5496–503.100. Chawla-Sarkar M, Lindner DJ, Liu YF, et al. Apoptosis and interferons: role of

interferon-stimulated genes as mediators of apoptosis. Apoptosis 2003;8:237–49.

101. Kulaeva OI, Draghici S, Tang L, et al. Epigenetic silencing of multiple interferonpathway genes after cellular immortalization. Oncogene 2003;22:4118–27.

102. Gray RJ, Pockaj BA, Kirkwood JM. An update on adjuvant interferon for melanoma.Cancer Control 2002;9:16–21.

103. Kirkwood JM, Richards T, Zarour HM, et al. Immunomodulatory effects of high-dose and low-dose interferon alpha2b in patients with high-risk resected melanoma:the E2690 laboratory corollary of intergroup adjuvant trial E1690. Cancer (Phila)2002;95:1101–12.

104. Chung KK, Dawson VL, Dawson TM. New insights into Parkinson’s disease.J Neurol 2003;250:III15–24.

105. Davenport TG, Jerome-Majewska LA, Papaioannou VE. Mammary gland, limb andyolk sac defects in mice lacking Tbx3, the gene mutated in human ulnar mammarysyndrome. Development (Camb) 2003;130:2263–73.

106. Lingbeek ME, Jacobs JJ, van Lohuizen M. The T-box repressors TBX2 and TBX3specifically regulate the tumor suppressor gene p14ARF via a variant T-site in theinitiator. J Biol Chem 2002;277:26120–7.

107. Maestro R, Dei Tos AP, Hamamori Y, et al. Twist is a potential oncogene thatinhibits apoptosis. Genes Dev 1999;13:2207–17.

108. Brummelkamp TR, Kortlever RM, Lingbeek M, et al. TBX-3, the gene mutated inulnar-mammary syndrome, is a negative regulator of p19ARF and inhibits senes-cence. J Biol Chem 2002;277:6567–72.

109. Carlson H, Ota S, Song Y, Chen Y, Hurlin PJ. Tbx3 impinges on the p53 pathwayto suppress apoptosis, facilitate cell transformation and block myogenic differenti-ation. Oncogene 2002;21:3827–35.

110. Wang X, Ling MT, Guan XY, et al. Identification of a novel function of TWIST, abHLH protein, in the development of acquired Taxol resistance in human cancercells. Oncogene 2004;23:474–82.

111. Fero ML, Randel E, Gurley KE, Roberts JM, Kemp CJ. The murine gene p27Kip1is haplo-insufficient for tumour suppression. Nature (Lond) 1998;396:177–80.

112. Sherr CJ. Principles of tumor suppression. Cell 2004;116:235–46.

5281




113. Kuwako KI, Taniura H, Yoshikawa K. Necdin-related MAGE proteins differentiallyinteract with the E2F1 transcription factor and the p75 neurotrophin receptor. J BiolChem 2004;279:1703–12.

114. Gillanders E, Hank Juo SH, Holland EA, et al. Localization of a novel melanomasusceptibility locus to 1p22. Am J Hum Genet 2003;73:301–13.

115. Pollock PM, Welch J, Hayward NK. Evidence for three tumor suppressor loci onchromosome 9p involved in melanoma development. Cancer Res 2001;61:1154–61.

116. Esteller M, Herman JG. Cancer as an epigenetic disease: DNA methylation andchromatin alterations in human tumours. J Pathol 2002;196:1–7.

117. Flagiello D, Poupon MF, Cillo C, Dutrillaux B, Malfoy B. Relationship betweenDNA methylation and gene expression of the HOXB gene cluster in small cell lungcancers. FEBS Lett 1996;380:103–7.

118. Karpf AR, Lasek AW, Ririe TO, et al. Limited gene activation in tumor and normalepithelial cells treated with the DNA methyltransferase inhibitor 5-aza-2�-deoxycy-tidine. Mol Pharmacol 2004;65:18–27.

119. Ballestar E, Paz MF, Valle L, et al. Methyl-CpG binding proteins identify novel sitesof epigenetic inactivation in human cancer. EMBO J 2003;22:6335–45.

120. Lu R, Au WC, Yeow WS, Hageman N, Pitha PM. Regulation of the promoteractivity of interferon regulatory factor-7 gene. Activation by interferon and silencingby hypermethylation. J Biol Chem 2000;275:31805–12.

121. Li LC, Dahiya R. MethPrimer: designing primers for methylation PCRs. Bioinfor-matics 2002;18:1427–31.

122. Robert MF, Morin S, Beaulieu N, et al. DNMT1 is required to maintain CpGmethylation and aberrant gene silencing in human cancer cells. Nat Genet 2003;33:61–5.

123. Reid GK, Besterman JM, MacLeod AR. Selective inhibition of DNA methyltrans-ferase enzymes as a novel strategy for cancer treatment. Curr Opin Mol Ther2002;4:130–7.

124. Goffin J, Eisenhauer E. DNA methyltransferase inhibitors-state of the art. AnnOncol 2002;13:1699–716.

5282




2004;64:5270-5282. Cancer Res Keith Hoek, David L. Rimm, Kenneth R. Williams, et al. Transformation of Melanocytes to MelanomasExpression Profiling Reveals Novel Pathways in the

Updated version

http://cancerres.aacrjournals.org/content/64/15/5270

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2004/10/15/64.15.5270.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/64/15/5270.full#ref-list-1

This article cites 123 articles, 35 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/64/15/5270.full#related-urls

This article has been cited by 57 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

SubscriptionsReprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/64/15/5270To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/suppl/2004/10/15/64.15.5270.DC1

http://cancerres.aacrjournals.org/content/64/15/5270.full#ref-list-1

http://cancerres.aacrjournals.org/content/64/15/5270.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Expression Profiling Reveals Novel Pathways in the ......human melanocytes and melanoma cells. We...

Documents

Transcript of Expression Profiling Reveals Novel Pathways in the ......human melanocytes and melanoma cells. We...