[NM23, an example of a metastasis suppressor gene].

Metastasis suppressor genes - unlike tumor suppressor genes - are defined by their capacity to control metastatic dissemination in vivo without affecting growth of the primary tumor. The first of these metastasis suppressor genes, NM23, was identified in 1988. Since then, expression of NM23 has been studied widely in human tumor cohorts, often with contradictory results. Not only is NM23 overexpressed in most human solid tumors when compared to healthy tissues, but also low expression of NM23 correlates with metastasis and poor clinical prognosis in the advanced stages of a number of epithelial cancer types, including melanoma, breast, colon, and liver carcinoma. This does not hold true, however, for other cancer types such as neuroblastoma and hematological malignancies, in which high NM23 expression correlates with more aggressive disease. Genetic alterations in the NM23 gene - loss of heterozygosity, spontaneous mutations and polymorphisms - are rarely found in tumors; thus, the metastatic potential of tumor cells is probably affected by NM23 protein levels. Three lines of evidence demonstrate the anti-metastatic activity of NM23: first, overexpression of NM23 in metastatic cell lines reduces their metastatic potential in xenograft models; second, the incidence of lung metastases is elevated in NM23 knockout mice prone to develop hepatocellular carcinoma, and, third, silencing NM23 by RNA interference confers a "metastatic phenotype" on non-invasive human epithelial liver and colon cancer cell lines. It appears that NM23 is crucial for inhibiting invasive migration, so acting at early stages of metastatic dissemination. The mechanistic basis of the metastasis suppressor function of NM23 and its regulated expression still remains obscure, however. Reactivation of expression of the endogenous NM23 gene in tumor cells, or stimulation of the pathways it controls, constitutes a promising avenue for anti-metastatic therapy.