Human malignant melanoma. A genetic disease?

BACKGROUND: Human hereditary malignant melanoma, comprising 5% of all cases of malignant melanoma, occurs in association with other malignancies, predominantly in families with dysplastic nevus syndrome. Additionally, higher incidences of malignant melanoma have been reported in individuals with genetic disorders such as ataxia telangiectasia and xeroderma pigmentosum. The results and observations as reported in the literature on the involvement of oncogenes and chromosomal aberrations in the development of malignant melanoma are reviewed and compared with the authors' own experimental and clinical experience.
RESULTS: Numerous chromosomal regions, as on chromosomes 1 and 9, were altered. The long arm of chromosome 6 was affected in 60% of melanomas. Introduction of a normal copy of chromosome 6 resulted in loss of tumorigenicity in vitro. True melanoma genes were evident in two animal models: the Sinclair swine and the teleost fish Xiphophorus. In the Xiphophorus system, the crossing-conditioned elimination of a tumor suppressor gene led to the uncontrolled activity of a dominantly acting oncogene in certain hybrids. The causative oncogene, Xmrk, encodes a receptor tyrosine kinase closely related to human epidermal growth factor receptor (EGFR). Among the numerous studied human oncogenes, mutations in the extensively investigated ras family are the result rather than the cause of malignant transformation. High expression of nuclear oncogenes simply may be a common feature of rapidly dividing cells. The receptor tyrosine kinase EGFR may be involved in late stage melanoma; the human exon with homology to Xmrk shows elevated transcription levels in 80% of human melanoma metastases. Deletions of the tumor suppressor gene MTS 1 may be important for melanoma formation, whereas deletions of p53 appear to be of minor relevance.
CONCLUSION: Scientific progress in treating and diagnosing malignant melanoma will largely depend on experimental approaches to define relevant genetic changes by functional analysis rather than descriptive phenomenology and correlative observations.