Expression profiling of t(12;22) positive clear cell sarcoma of soft tissue cell lines reveals characteristic up-regulation of potential new marker genes including ERBB3

Clear cell sarcoma of soft tissue (CCSST), also known as malignant melanoma of soft parts, represents a rare lesion of the musculoskeletal system usually affecting adolescents and young adults. CCSST is typified by a chromosomal t(12;22)(q13;q12) translocation resulting in a fusion between the Ewing sarcoma gene (EWSR1) and activating transcription factor 1 (ATF1), of which the activity in nontransformed cells is regulated by cyclic AMP. Our aim was to identify critical differentially expressed genes in CCSST tumor cells in comparison with other solid tumors affecting children and young adults to better understand signaling pathways regulating specific features of the development and progression of this tumor entity. We applied Affymetrix Human Genome U95Av2 oligonucleotide microarrays representing approximately 12,000 genes to generate the expression profiles of the CCSST cell lines GG-62, DTC-1, KAO, MST2, MST3, and Su-CC-S1 in comparison with 8 neuroblastoma, 7 Ewing tumor, and 6 osteosarcoma cell lines. Subsequent hierarchical clustering of microarray data clearly separated all four of the tumor types from each other and identified differentially expressed transcripts, which are characteristically up-regulated in CCSST. Statistical analysis revealed a group of 331 probe sets, representing approximately 300 significant (P < 0.001) differentially regulated genes, which clearly discriminated between the CCSST and other tumor samples. Besides genes that were already known to be highly expressed in CCSST, like S100A11 (S100 protein) or MITF (microphthalmia-associated transcription factor), this group shows an obvious portion of genes that are involved in cyclic AMP response or regulation, in pigmentation processes, or in neuronal structure and signaling. Comparison with other expression profile analyses on neuroectodermal childhood tumors confirms the high robustness of this strategy to characterize tumor entities based on their gene expression. We found the avian erythroblastic leukemia viral oncogene homologue 3 (ERBB3) to be one of the most dramatically up-regulated genes in CCSST. Quantitative real-time PCR and Northern blot analysis verified the mRNA abundance and confirmed the absence of the inhibitory transcript variant of this gene. The protein product of the member of the epidermal growth factor receptor family ERBB3 could be shown to be highly present in all of the CCSST cell lines investigated, as well as in 18 of 20 primary tumor biopsies. In co