Methylation of the Cyclin A1 Promoter Correlates with Gene Silencing in Somatic Cell Lines, while Tissue-Specific Expression of Cyclin A1 Is Methylation Independent

Abstract

Gene expression in mammalian organisms is regulated at multiple levels, including DNA accessibility for transcription factors and chromatin structure. Methylation of CpG dinucleotides is thought to be involved in imprinting and in the pathogenesis of cancer. However, the relevance of methylation for directing tissue-specific gene expression is highly controversial. The cyclin A1 gene is expressed in very few tissues, with high levels restricted to spermatogenesis and leukemic blasts. Here, we show that methylation of the CpG island of the human cyclin A1 promoter was correlated with nonexpression in cell lines, and the methyl-CpG binding protein MeCP2 suppressed transcription from the methylated cyclin A1 promoter. Repression could be relieved by trichostatin A. Silencing of a cyclin A1 promoter-enhanced green fluorescent protein (EGFP) transgene in stable transfected MG63 osteosarcoma cells was also closely associated with de novo promoter methylation. Cyclin A1 could be strongly induced in nonexpressing cell lines by trichostatin A but not by 5-aza-cytidine. The cyclin A1 promoter-EGFP construct directed tissue-specific expression in male germ cells of transgenic mice. Expression in the testes of these mice was independent of promoter methylation, and even strong promoter methylation did not suppress promoter activity. MeCP2 expression was notably absent in EGFP-expressing cells. Transcription from the transgenic cyclin A1 promoter was repressed in most organs outside the testis, even when the promoter was not methylated. These data show the association of methylation with silencing of the cyclin A1 gene in cancer cell lines. However, appropriate tissue-specific repression of the cyclin A1 promoter occurs independently of CpG methylation.

ACKNOWLEDGMENTS

We are grateful to S. Kudo, Hokkaido Institute, Sapporo, Japan, for the MeCP2 expression vector and to A. Bird, University of Edinburgh, Edinburgh, United Kingdom, for antibodies against MeCP2. We thank Ian Williamson, Danlin Chen, Stephan Rust, Jochen Kremerskothen, Omeni Osian, Xin Min Yan, and Christina Bornemann for technical assistance and Kamlesh Asotra, Scott Fraser, and Mary Dickinson for their assistance with confocal laser scanning microscopy.

This work is supported by grants from NIH and by U.S. Defense Grants as well as grants from the Parker Hughes and C. and H. Koeffler Funds and the Gladys Lichtenstein Trust. C. Müller's work is supported by grants from the Deutsche Forschungsgemeinschaft (Mu 1328/2-1) and the Deutsche Krebshilfe (10-1539-Mü1). G. Idos was supported by the Howard Hughes Institute Undergraduate Program. H. P. Koeffler holds the Mark Goodson Chair in Oncology Research and is a member of the Jonsson Cancer Center. C. Readhead's work is supported by NIH grant 1RO1RR12406.