Active X chromosome DNA is unmethylated at eight CCGG sites clustered in a guanine-plus-cytosine-rich island at the 5'' end of the gene for phosphoglycerate kinase.

The 5' control region and first exon for human X-chromosome-linked phosphoglycerate kinase is contained in a G + C-rich island. We measured methylation at all HpaII sites in this 5' region of leukocyte DNA. By use of a blotting procedure that allows analysis of small DNA fragments, we found that the HpaII sites are entirely methylated when from an inactive X chromosome and entirely unmethylated when from an active one. In contrast, methylation of HpaII sites in more downstream regions of the gene is essentially the same in active and inactive X chromosomes.     

The effects of 5-azacytidine on the expression of the rat growth hormone gene. Methylation modulates but does not control growth hormone gene activity

The role of DNA methylation in the expression of the rat growth hormone (rGH) gene was assessed by using a hypomethylating agent, 5-azacytidine, and the iso-schizomeric restriction enzymes MspI and HpaII. 5-Azacytidine increased rGH mRNA 3-8-fold in GH3D6 cells, a subclone of rat pituitary tumor cell lines that expresses one-tenth to one-fifteenth the GH expressed by two other clones, GH3 and GC. The effect was also detected at the level of pre-mRNA. The effect was independent of glucocorticoids and thyroid hormones and was found to be inheritable. The DNA methylation pattern generated by the isoschizomeric restriction enzymes indicated that the HpaII sites in the rGH gene were mostly methylated in GH3D6 cells but mostly unmethylated in GC cells. After treatment with 5-azacytidine, about 22% of these HpaII sites in GH3D6 cells became unmethylated. Thus, DNA methylation correlates inversely with the expression of the rGH gene in these cell lines. However, three other observations indicate that factors in addition to DNA methylation control rGH expression. First, in GC cells, even though most of the HpaII sites are unmethylated, the gene is not fully expressed. Second, in rat hepatoma cells, which do not express GH at all, the GH gene is less methylated than that in GH3D6 cells. Third, within the sensitivities of the assay methods, 5-azacytidine has no effect on the GH gene when it is completely silent. Taken together, the findings indicate that DNA methylation modulates but does not control GH gene expression. It is tempting to speculate that DNA methylation can influence expression only when the gene is committed to express.     

Methylation of adenovirus genes in transformed cells and in vitro: influence on the regulation of gene expression?

An inverse correlation has been described between the levels of DNA methylation in specific segments of adenovirus DNA integrated into the genomes of transformed and tumor cells and the extent to which these segments are expressed as messenger RNA. In the adenovirus type 2 (Ad2)-transformed hamster cell lines HE2 and HE3, the virus-specific DNA binding protein (DBP) is not expressed, and the DNA in the DBP gene is completely methylated in all 5'-CCGG-3' sites. At least part of the late promoter/leader sequence of the DBP gene is present in cell lines HE2 and HE3. In line HE1, on the other hand, the DBP is expressed, and the DNA in the DBP gene is unmethylated at the 5'-CCGG-3' (HpaII) sites. The late promotor/leader sequence of the DBP gene is expressed in cytoplasmic RNA isolated from line HE1. The effect of DNA methylation has also been tested in vitro in a microinjection system using Xenopus laevis oocytes. Unmethylated DNA fragments of Ad2 (E2a region) have been found to serve as active templates. When the same fragments are methylated at the 5'-CCGG-3' sites by the HpaII DNA-methyltransferase, viral RNA synthesis is inhibited upon microinjection into oocyte nuclei. These results provide direct evidence for the notion that DNA methylated at highly specific sites is somehow involved in the regulation of gene expression.     

In vitro methylation of bovine papillomavirus alters its ability to transform mouse cells.

Bovine papillomavirus (BPV) was methylated in vitro at either the 29 HpaII sites, the 27 HhaI sites, or both. Methylation of the HpaII sites reduced transformation by the virus two- to sixfold, while methylation at HhaI sites increased transformation two- to fourfold. DNA methylated at both HpaII and HhaI sites did not differ detectably from unmethylated DNA in its efficiency of transformation. These results indicate that specific methylation sites, rather than the absolute level of methylated cytosine residues, are important in determining the effects on transformation and that the negative effects of methylation at some sites can be compensated for by methylation at other sites. BPV molecules in cells transformed by methylated BPV DNA contained little or no methylation, indicating that the pattern of methylation was not faithfully retained in these extrachromosomally replicating molecules. Methylation at the HpaII sites (but not the HhaI sites) in the cloned BPV plasmid or in pBR322 also inhibited transformation of the plasmids into Escherichia coli HB101 cells.     

Specifically unmethylated cytidylic-guanylate sites in Herpesvirus saimiri DNA in tumor cells.

The restriction endonucleases MspI (CCGG), HpaII (CCGG), FnuDII (CGCG), and HaeIII (GGCC) were used to study the methylation of Herpesvirus saimiri DNA in tumor cells taken directly from tumor-bearing animals. No evidence was found for methylation of the 5' terminal C in the sequence CCGG or of the internal C in the sequence GGCC, but extensive methylation of CG was detected. Fifteen HpaII sites and 17 FnuDII sites were detected in the unique DNA region of the H. saimiri strain used. Twenty-eight of the 32 sites were methylated in greater than 90% of the viral DNA molecules in tumor cells, but the remaining 4 sites were unmethylated in greater than 95% of the viral DNA molecules in tumor cells. The locations of the four specifically unmethylated sites were mapped and appeared to be identical in the four different induced leukemias examined (one owl monkey and three white-lipped marmosets). The nonproducer 1670 tumor cell line, in continuous passage for over 7 years, contained four similar specifically unmethylated sites. Possibilities for the physiological significance of the unmethylated sites are discussed.     

Methylation of the mouse hprt gene differs on the active and inactive X chromosomes.

It has been proposed that DNA methylation is involved in the mechanism of X inactivation, the process by which equivalence of levels of X-linked gene products is achieved in female (XX) and male (XY) mammals. In this study, Southern blots of female and male DNA digested with methylation-sensitive restriction endonucleases and hybridized to various portions of the cloned mouse hprt gene were compared, and sites within the mouse hprt gene were identified that are differentially methylated in female and male cells. The extent to which these sites are methylated when carried on the active and inactive X chromosomes was directly determined in a similar analysis of DNA from clonal cell lines established from a female embryo derived from a mating of two species of mouse, Mus musculus and Mus caroli. The results revealed two regions of differential methylation in the mouse hprt gene. One region, in the first intron of the gene, includes four sites that are completely unmethylated when carried on the active X and extensively methylated when carried on the inactive X. These same sites are extensively demethylated in hprt genes reactivated either spontaneously or after 5-azacytidine treatment. The second region includes several sites in the 3' 20kilobases of the gene extending from exon 3 to exon 9 that show the converse pattern; i.e., they are completely methylated when carried on the active X and completely unmethylated when carried on the inactive X. At least one of these sites does not become methylated after reactivation of the gene. The results of this study, together with the results of previous studies by others of the human hprt gene, indicate that these regions of differential methylation on the active and inactive X are conserved between mammalian species. Furthermore, the data described here are consistent with the idea that at least the sites in the 5' region of the gene play a role in the X inactivation phenomenon and regulation of expression of the mouse hprt gene.     

The unmethylated state of the promoter/leader and 5'-regions of integrated adenovirus genes correlates with gene expression.

An inverse correlation has been established between the levels of DNA methylation at 5'-CCGG-3' (MspI/HpaII) sites in specific genes of integrated viral DNA in adenovirus type 12 (Ad12)-transformed hamster cell lines and the extent to which these genes are expressed ( Sutter and Doerfler , 1979, 1980). In general, early genes are transcribed into mRNA, while late genes are permanently switched off in these cell lines. Adenovirus type 2 genes methylated in vitro at 5'-CCGG-3' sites are not transcribed upon microinjection into nuclei of Xenopus laevis oocytes - unmethylated genes are expressed ( Vardimon et al., 1982a ). The MspI sites in the early and in some of the late Ad12 genes in cell lines HA12 /7, T637 , and A2497 -3 have now been precisely mapped. The data presented here reveal that the promoter/leader and 5'-regions of the early genes are unmethylated both at MspI sites and at 5'-GCGC-3' (HhaI) sites. In some instances, e.g., in the E2a regions in all three lines, the main parts of the early genes are partly methylated, even though the genes are expressed. In cell line HA12 /7, the early region E3 is not expressed, and the promoter/leader and 5'-regions of this segment are fully methylated. All late regions are completely methylated. The results suggest that the state of methylation in the promoter/leader and 5'-regions of integrated adenovirus genes is important in the control of gene expression.     

Developmental changes in the methylation of the rat albumin and alpha-fetoprotein genes

We have analyzed methylation of the rat albumin and alpha-fetoprotein (AFP) genes by hydridizing labeled cDNA clones to HpaII and MspI digests of DNA from different stages of development. These CCGG-cutting enzymes distinguish 5-methylcystosine in mCCGG (sensitive to HpaII) and CmCGG (sensitive to MspI). In the liver, the albumin gene is heavily methylated at 18 days gestation and uniformly demethylated in the adult. The AFP gene is also heavily methylated at 18 days gestation, and develops demethylated regions at the 3' half of the gene in the adult. These methylation changes are not observed in other embryonic or adult tissues. We also evaluated expression of these genes by measuring their corresponding mRNAs. The albumin gene is actively transcribed in 18-day fetal liver, when it is heavily methylated, as well as in adult liver, when it is unmethylated. In contrast, the AFP gene is transcribed only in fetal liver, even though it is less methylated in adult liver. These findings suggest that specific methylation changes are associated with changes in gene expression, but that this association is not adequately described by the simple hypothesis that methylation turns genes off.     

Determination of DNA sequences containing methylcytosine in Bacillus subtilis Marburg.

The methylcytosine-containing sequences in the DNA of Bacillus subtilis 168 Marburg (restriction-modification type BsuM) were determined by three different methods: (i) examination of in vivo-methylated DNA by restriction enzyme digestion and, whenever possible, analysis for methylcytosine at the 5' end; (ii) methylation in vitro of unmethylated DNA with B. subtilis DNA methyltransferase and determination of the methylated sites; and (iii) the methylatability of unmethylated DNA by B. subtilis methyltransferase after potential sites have been destroyed by digestion with restriction endonucleases. The results obtained by these methods, taken together, show that methylcytosine was present only within the sequence 5'-TCGA-3'. The presence of methylcytosine at the 5' end of the DNA fragments generated by restriction endonuclease AsuII digestion and the fact that in vivo-methylated DNA could not be digested by the enzyme XhoI showed that the recognition sequences of these two enzymes contained methylcytosine. As these two enzymes recognized a similar sequence containing a 5' pyrimidine (Py) and a 3' purine (Pu), 5'-PyTCGAPu-3', the possibility that methylcytosine is present in the complementary sequences 5'-TTCGAG-3' and 5'-CTCGAA-3' was postulated. This was verified by the methylation in vitro, with B. subtilis enzyme, of a 2.6-kilobase fragment of lambda DNA containing two such sites and devoid of AsuII or XhoI recognition sequences. By analyzing the methylatable sites, it was found that in one of the two PyTCGAPu sequences, cytosine was methylated in vitro in both DNA strands. It is concluded that the sequence 5'-PyTCGAPu-3' is methylated by the DNA methyltransferase (of cytosine) of B. subtilis Marburg.     

CG sites and their nucleotide environment in the human gamma-delta-beta-globin gene cluster: distribution, frequency and possible frame for differential methylation

Available restriction endonucleases including CG dinucleotides in their target sequences (most of them being unable to cut the DNA when the cytosine of the CG sequence is methylated) have been used to map cloned DNA covering the human gamma-delta-beta globin gene cluster. Since the human DNA fragments were cloned in Escherichia coli, only the internal cytosine in the sequence CCAT GG could be methylated. Thus, any recognized "CG enzyme" site can be detected since they are unmethylated. Results show that frequencies of "CG enzyme" sites regularly decrease from the gamma-globin region to the beta-globin region, the latter being very poor in "CG enzyme"' sites. The array of enzymes used here detects 4 times more CG sites than the classical MspI/HpaII system. Examination of previously sequenced parts of the gamma-delta-beta globin gene cluster shows that CG dinucleotides correspond to an average frequency of 1 out of 104 nucleotides in the gamma-globin region and 1 out of 138 nucleotides in the beta-globin region. In the gamma-globin region, 1 CG out of 4 or 5 may be detected by the enzymes used; the detected frequency is less than 1 out of 10 CG in the beta-region. Analysis of nucleotide environment around CG dinucleotides shows occurrence of local differences, the main sequences being CGG in the 5' side flanking the gamma genes and ACG in the corresponding area of the beta gene. The results presented introduce some new considerations about analysis of cytosine methylation which has been previously proposed as playing a role in the control of the activity of gamma, delta and beta genes respectively.     

The somatic replication of DNA methylation

We have tested the hypothesis that DNA methylation patterns are replicated in the somatic cells of vertebrates. Using M-Hpa II, the modification enzyme from Haemophilus parainfluenzae which methylates the internal cytosine residues in the sequence 5'CCGG 3' GGCC, we methylated bacteriophage phi X174 RF DNA and the cloned chicken thymidine kinase (tk) gene in vitro and then introduced these DNAs and unmethylated controls into tk- cultured mouse cells by DNA-mediated transformation. Twenty-five cell generations later, the state of methylation of transferred DNA was examined by restriction endonuclease analysis and blot hybridization. We conclude that methylation at Hpa II sites is replicated by these cultured cells but not with 100% fidelity. We have also noted that methylation of the cloned chicken tk gene decreases its apparent transformation efficiency relative to unmethylated molecules.     

Distribution of 5-methylcytosine in the DNA of somatic and germline cells from bovine tissues.

Genomic DNA of calf thymus contains 1.5 times as much 5-methylcytosine as similar sperm DNA, but the major EcoRI repeat fragment from satellite I of thymus contains ten times as much 5-methylcytosine as the corresponding fragment from sperm DNA. Restriction enzyme analyses of the total DNA and the satellite I fragment show that three HpaII sites in the fragment are completely unmethylated in sperm but fully methylated in thymus DNA. Under-methylation of many sites in the satellite DNAs can probably account for the lower level of methylation of sperm DNA rather than hemimethylation as previously suggested. These results are also discussed in relation to maintenance and de novo (initiation-type) methylases.     

The hypervariable DXS255 locus contains a LINE-1 repetitive element with a CpG island that is extensively methylated only on the active X chromosome.

The DXS255 locus at Xp11.22 is highly polymorphic due to a 26-bp variable number of tandem repeats (VNTR) motif. In previous studies, one of the MspI sites flanking the VNTR manifested a correlation between methylation and X chromosome inactivation. Here we show, by DNA sequence analysis, that this MspI site is located within the CpG island at the 5' end of a LINE-1 element, which is 2.5 kb from the VNTR. The methylation status of the CpG island was assessed in Southern blotting experiments using the methylation-sensitive enzymes HpaII, HhaI, and BssHII. All these sites were completely methylated on active X chromosomes, consistent with previously reported findings of full methylation of LINE-1 elements throughout the genome. However, on inactive X chromosomes these sites were predominantly unmethylated, although patterns were found to be heterogeneous. The results suggest that LINE-1 elements on the inactive X chromosome are not suppressed by full methylation of their CpG islands. The differential methylation of the DXS255 CpG island provides the basis for a highly informative X inactivation analysis system.     

Expression of the G6PD locus on the human X chromosome is associated with demethylation of three CpG islands within 100 kb of DNA.

We have previously reported that expression of the G6PD locus is correlated with the methylation status of two islands of CpG dinucleotides which are 3' to the locus and in the 5' region of two adjacent genes of unknown function, P3 and GdX. We have now examined the methylation of a third CpG island in the promoter region of the G6PD gene itself in DNA from males, females and reactivants that express G6PD on the inactive X chromosome. Our results show that expression of the G6PD gene is associated with concordant demethylation of all three CpG islands in this 100-kb region of DNA.