The field of DNA methylation in mammalian cells was pioneered by the discovery of inverse correlations between the extent of segmental DNA methylation and the genetic activity of these segments. In integrated Ad12 DNA in Ad12-transformed cells, the early viral genes are transcribed; the late Ad12 genes responsible for the synthesis of virion capsid proteins are permanently silenced. Hence, in Ad12-transformed cells or in Ad12-induced tumor cells, late viral gene products and mature virions are not synthesized [74]. Thus, Ad12-transformed cells provide a suitable tool to study the levels of DNA methylation in distinct sections of the viral genome and to document inverse correlations to gene transcription for the first time [15, 33, 75, 76]. A particularly clear example has been offered by the adenovirus type 2 (Ad2)-transformed cell lines HE1, HE2, and HE3 [77]. The E2A region of Ad2 DNA is not expressed in cell lines HE2 and HE3, whereas cell line HE1 does express the E2A region of the integrated Ad2 DNA [78]. Accordingly, the 5´-CCGG-3´ dinucleotides (HpaII sites) in the promoter region of this gene in cell line HE1 are unmethylated and methylated in cell lines HE2 and HE3 [33].
At the time, one had to rely on the analyses of a limited number of 5´-CG-3´ dinucleotides whose state of methylation could be assessed only by the use of methylation-sensitive restriction endonucleases, in this example of HpaII and MspI (5´-CCGG-3´). These seminal observations started a burst of similar studies with numerous viral and cellular genes and confirmed almost without exceptions the initial observations which suggested that specific promoter methylation patterns are instrumental in the long-term silencing of eukaryotic genes [2, 72]. To this day however, we do not understand whether there have to be a specific number or pattern of m5C residues in a promoter to assure its long-term silencing.
Productive versus abortive infection of cells with Ad12. Ad12 interacts with human cells in a productive infection leading to the synthesis of a large number of progeny virions. In contrast, the replication of Ad12 in hamster cells is completely blocked [79, 80], possibly because only few copies of Ad12 DNA are capable of reaching the nucleus of the hamster cells [81]. However, the expression of the Ad12 genome in hamster cells appears to be suppressed in several steps of the normal replication cycle. Nevertheless, viral DNA can be traced in the nucleus of the abortively infected hamster cells and limited transcription of the early genes of non-integrated Ad12 has been documented [74, 81]. There is, however, no methylation of the intranuclear Ad12 DNA in productively or abortively infected cells [33]. Obviously, methylation of free viral DNA, even in the abortive system, does not serve to regulate or to inactivate the late Ad12 viral genes, which are not transcribed in abortively infected hamster cells.
The actively transcribed genome of frog virus 3 (FV3) is completely 5´-CG-3´ methylated. The hypermethylated state of the virion-encapsidated or of the intracellular FV3 genomes [29] in fish or mammalian cells has taught us that the biological significance of DNA methylation cannot be schematically interpreted and depends entirely on the biological system studied. While the inverse correlations described above hold true for most systems investigated so far, there are notable exceptions to this “rule” and, of course, no stringent dogmas in biology. It has been documented that the viral L1140 gene is actively transcribed late after infection of fish cells with FV3, although it is methylated in all 5´-CG-3´ dinucleotides [82]. In FV3-infected fish or hamster cells, a transfected L1140 promoter-indicator gene construct is active in the unmethylated or fully 5´-CG-3´-methylated form. When the same construct is methylated only in the 5´-CCGG-3´ (HpaII) sequences, its activity is reduced. Compatibility between the methylation of an immediate-early FV3 promoter with its active transcription has also been reported [83]. These data confirm the special methylation requirements of this promoter in FV3 DNA. Special properties of the FV3 DNA-protein interactions may account for these unexpected activity patterns. It would be interesting to study in greater biochemical detail the transcription of FV3 genes and the proteins involved in their regulation.
Answers to all your Biology Questions
