Feedback regulation of DNA methyltransferase gene expression by methylation.

This paper tests the hypothesis that expression of the DNA methyltransferase, dnmt1, gene is regulated by a methylation-sensitive DNA element. Methylation of DNA is an attractive system for feedback regulation of DNA methyltransferase as the final product of the reaction, methylated DNA, can regulate gene expression in cis. We show that an AP-1-dependent regulatory element of dnmt1 is heavily methylated in most somatic tissues and in the mouse embryonal cell line, P19, and completely unmethylated in a mouse adrenal carcinoma cell line, Y1. dnmt1 is highly over expressed in Y1 relative to P19 cell lines. Global inhibition of DNA methylation in P19 cells by 5-azadeoxycytidine results in demethylation of the AP-1 regulatory region and induction of dnmt1 expression in P19cells, but not Y1 cells. We propose that this regulatory region of dnmt1 acts as a sensor of the DNA methylation capacity of the cell. These results provide an explanation for the documented coexistence of global hypomethylation and high levels of DNA methyltransferase activity in many cancer cells and for the carcinogenic effect of hypomethylating diets.     

Bacteriophage P4 DNA replication. Nucleotide sequence of the P4 replication gene and the cis replication region

A 3100 base piece of DNA from the 11,500 base genome of bacteriophage P4 was analyzed for its nucleotide sequence. This segment of DNA contains two open reading frames of 106 and 777 amino acid residues; the latter of which is the coding sequence for the Mr 84,841 alpha protein, which is necessary for P4 DNA replication and is thought to act as a P4-specific DNA primase. A region of about 300 base-pairs localized just beyond the alpha gene and about 4500 bases from the origin of replication (ori), was defined as the locus for P4's cis replication region (crr). This region is required for replication both in vivo and in vitro, and consists of two directly repeated sequences of 120 base-pairs that match one another at 98 positions. These directly repeated sequences are separated by 60 base-pairs, which are not necessary for replication. Each repeat in crr contains three copies of the octamer TGTTCACC that is found six times in ori. Either of the 120 base-pair repeat sequences in crr is sufficient for replication, and the entire crr can function in an inverted orientation. crr is also active at a distance of 1800 bases from the P4 origin of replication.     

Evidence for methylation as a regulatory mechanism in HLA-DR x gene expression

We examined the possibility that one mechanism for controlling HLA-DR gene expression occurs through DNA hypomethylation. We employed the restriction enzyme Hpa II, which recognizes the sequence 5CCGG3 but not 5C^mCGG3, to study DNA methylation. We first compared a DR-positive B lymphoblastoid cell line (LCL) with an isogenic DR-negative T-LCL. Using a genomic probe for the DR gene, we showed that an Hpa II digestion of DNA from the B-LCL resulted in bands of lower molecular weight than that of the T-LCL. This indicates that the B-LCL DR gene is hypomethylated relative to the T-LCL gene. Demethylation of the gene from the B-LCL is incomplete, suggesting that complete demethylation is not required for its expression. We also examined somatic cell hybrids of T-LCL and B-LCL since the DR gene, which is inactive in the T-LCL, is expressed in the hybrids, providing a system to study DR gene induction. We examined the hybrid line 174 × CEM.T1, which contains and expresses solely the DR gene from the T-LCL parent since both copies of the DR gene from the B-LCL parent, 174, are deleted. The expressed DR gene from the hybrid was compared with the unexpressed gene from the T-LCL parental line, and again an association between DR gene expression and DNA hypomethylation was observed. In contrast to the DR gene from B-LCL, which is not completely demethylated, the DR gene in this hybrid line is not methylated at either of the Msp I sites covered by our probe. This suggests that different regulatory mechanisms operating through DNA methylation may be involved in the expression of DR genes from T-LCL and B-LCL. Examination of another hybrid line which has DR genes from both parental lines supports this contention. The implications of these findings are discussed.     

Differential regulation of DNA methylation in rat testis and its regulation by gonadotropic hormones

Eukaryotic DNA methylation occurs exclusively at the 5'-position of cytosine and has been implicated in the regulation of gene expression. Using high-performance liquid chromatography, the methylation of testis DNA during its development, in different cell populations and during regulation by gonadotropic hormones, were studied. The 5-mC content of testis DNA increased significantly from days 30 to days 150, while in 2-yr-old testis 5-mC content decreased significantly. Among various populations of testicular cells, pachytene spermatocyte DNA contained a significantly high amount of 5-mC when compared to spermatogonia, spermatids and mature sperm DNA. However, the 5-mC content of elongated spermatids was significantly less when compared to the above four fractions. Administration of follicle stimulating hormone to immature rats caused hypomethylation of seminiferous tubular DNA while luteinizing hormone caused similar effects in Leydig cells. These results indicate that in testis, DNA methylation is differentially regulated during development and is controlled by gonadotropic hormones.     

DNA methylation as a regulatory mechanism in rat gamma-crystallin gene expression.

We have investigated the methylation state of the rat gamma-crystallin genes in DNA from lens cells at different developmental stages as well as from kidney and heart cells. A clear correlation between the extent of demethylation of the promoter and 5' gene regions and the expression of these genes was observed. No change in the methylation state of the far upstream or 3' regions of the genes was seen. The demethylation of the promoter region was shown to occur during the differentiation from the lens epithelial to the lens fiber cell. The effect of cytosine methylation on gamma-crystallin promoter activity was tested by measuring gamma-crystallin promoter/chloramphenicol acetyltransferase fusion gene expression after in vitro primed repair synthesis of the promoter region in the presence of either dCTP or 5mdCTP. The hemimethylated promoter was no longer capable of promoting high CAT activity after introduction into lens-like cells. Taken together, our data suggest that DNA demethylation may be the determining step in the developmental stage-specific expression of the rat gamma-crystallin genes.     

Evidence of two levels of control of P1 oriR and host oriC replication origins by DNA adenine methylation.

A mutant mini-P1 plasmid with increased copy number can be established in Dam- strains of Escherichia coli, where mini-P1 plasmid replication is normally blocked. Comparison of this plasmid and a plasmid driven by the host oriC replication origin showed that both origins are subject to control by methylation at two different levels. First, both origins appear to be subject to negative regulation acting at the level of hemimethylation. This probably involves the sequestration of the hemimethylated DNA produced by replication, as has been previously described for oriC. Second, both origins show a positive requirement for adenine methylation for efficient function in vivo. This conclusion is supported by the behavior of the P1 origin in an improved in vitro replication system. In vitro, where sequestration of hemimethylated DNA is not expected to occur, the hemimethylated P1 origin DNA was fully functional as a template. However, the activity of fully unmethylated DNA was severely restricted in comparison with that of either of the methylated forms. This in vitro uncoupling of the two effects of origin methylation suggests that two separate mechanisms are involved.     

The hormone-sensitive hepatic Na+-pump. Evidence for regulation by diacylglycerol and tumor promoters

Ouabain-sensitive 86Rb+ uptake by isolated rat hepatocytes was studied to elucidate how Ca2+-mobilizing hormones stimulate the Na+-pump. Stimulation of this uptake was observed with concentrations of vasopressin ([8-arginine]vasopressin, AVP), angiotensin II, and norepinephrine which elicited Ca2+ mobilization and phosphorylase activation. These results suggested that changes in cytosolic Ca2+, mediated by inositol trisphosphate, might trigger sodium pump stimulation by AVP. However, in hepatocytes incubated in Ca2+-free Krebs-Henseleit buffer, Na+-pump activity was not altered over 15 min by either 1.5 mM EGTA or 1.5 mM Ca2+. Furthermore, incubation of cells in 5 mM EGTA for 15-30 min drastically impaired the ability of AVP to increase cytosolic Ca2+, but only modestly attenuated AVP-stimulated Na+-pump activity. Two tumor promoters, phorbol myristate acetate (PMA) and mezerein, stimulated Na+/K+-ATPase-mediated transport activity. Similarly, addition of synthetic diacylglycerols or of exogenous phospholipase C from Clostridium perfringens to increase endogenous diacylglycerol levels also resulted in a stimulation of the Na+-pump in the absence of changes in cytosolic or total cellular Ca2+ levels. Stimulation of the Na+-pump by the combination of maximal concentrations of PMA and AVP did not produce an additive response, and both agents displayed a transient time course, suggesting that the two agents share a common mechanism. Stimulation of the Na+-pump by AVP and PMA was not blocked by amiloride analogs which inhibit Na+/H+ exchange, but these compounds blocked the action of insulin. These data suggest that the elevated Na+/K+-ATPase-mediated transport activity observed in hepatocytes following exposure to Ca2+-mobilizing hormones is a consequence of stimulated diacylglycerol formation and may involve protein kinase C.