5-Methylcytosine in eukaryotic DNA

Enzymatic methylation of chicken erythrocyte DNA modified by two carcinogens, 2-(N-acetoxyacetylamino) fluorene and methylnitrosourea

Both DNA-AAF and MNU-alkylated DNA are methylated less than nonmodified DNA by rat brain nuclei cytosine 5-methyltransferase purified either by chromatography on DEAE cellulose or by Dyematrex. The inhibition of methylation is proportional to the modification of the DNA, and DNA having a given percentage of bases modified with MNU is less methylated than DNA modified to the same extent with AAF. Moreover, DNA-AAF irreversibly inhibits the methylation of native DNA, whereas MNU-alkylated DNA does not inhibit the methylation of native DNA. The AAF-substituted DNA has a higher affinity for the enzyme than native DNA. However, this is probably not due to the AAF-induced local destabilization of the DNA helix, since heat-denatured DNA shows a lower affinity for the enzyme than double-stranded DNA. Addition of DNA-AAF to the enzyme preincubated with native DNA inhibits methylation, but only after a lag period. This agrees with the model in which the methylase walks along the strand to methylate cytosine residues before being detached from the DNA. AAF bound to guanine residues may block the movement of the enzyme along the helix. The in vitro hypomethylation of DNA, caused by carcinogens, could explain the in vivo observations made by several authors and could have significance in gene activity, cellular differentiation, and oncogenesis.

Inhibition of DNA methylation by chemical carcinogens in vitro

A diverse range of ultimate chemical carcinogens inhibited the transfer of methyl groups from S-adenosylmethionine to hemimethylated DNA in a reaction catalyzed by mouse spleen methyltransferase. The formation of alkali-labile sites in DNA lessened its ability to accept methyl groups in vitro, but the methylation reaction was much less sensitive to thymine dimers or double-strand breaks. Carcinogens induced the formation of alkali-labile DNA lesions, but the degree of methyltransferase inhibition observed was greater than that expected for this damage alone. Certain carcinogens were also capable of direct modification and inactivation of the methyltransferase enzyme. Benzo(a)pyrene treatment of living BALB/3T3 A31 clone 1-13 but not C3H/10T1/2 clone 8 cells resulted in a 12% decrease in total 5-methylcytosine content of cellular DNA. Carcinogenic agents may therefore cause heritable changes in 5-methylcytosine patterns in certain cell types by a variety of mechanisms, including adduct formation, induction of apurinic sites and single-strand breaks and direct inactivation of DNA methyltransferase.

Demethylation and expression of murine mammary tumor proviruses in mouse thymoma cell lines

Murine mammary tumor virus (MMTV) expression is analyzed in a T-lymphoid cell line (T1M1) sensitive to the killing effect of glucocorticoids and in two of its variants, one resistant (T1M1r) and one supersensitive (T1M1ss) to glucocorticoid-induced lymphocytolysis. In the T1M1 line, MMTV is expressed and induced approximately 10-fold by short treatment with dexamethasone. Southern blot analyses of restriction enzyme digests of DNA from T1M1 cells reveal three proviruses similar to those of normal C57BL mouse tissue. In the T1M1ss line, which has retained functional glucocorticoid receptors, MMTV mRNA is inducible by glucocorticoids, while induction is reduced in the T1M1r line defective in glucocorticoid receptors. Moreover, the T1M1r line expresses a strikingly elevated basal level of MMTV mRNA in the absence of hormone. No rearrangements or superinfection have occurred in the variants, but all the regions containing 5'-long terminal repeats are demethylated in the T1M1r variant although other sites of the provirus remain methylated. Because this variant was selected by prolonged treatment with dexamethasone, these observations raise the possibility that the continuous transcription of MMTV that occurred during this selection can result in glucocorticoid-induced demethylation of long-terminal-repeat sequences.

Cytoplasmic modification of nuclear gene expression

A review is presented on 1) the autonomous nature of mammalian cell cytoplasm and 2) the cytoplasmic modification of nuclear gene expression. Topics include a discussion of cytoplasmic suppression of tumorigenicity. It is proposed that alterations in DNA methylation patterns may be a possible mechanism to explain cytoplasmic modification of nuclear gene expression.

Asymmetrical distribution of CpG in an 'average' mammalian gene

The frequency and distribution of the rare dinucleotide CpG was examined in 15 mammalian genes. CpG is highly methylated at cytosine in mammalian DNA (1,2) and 5-methylcytosine (5mC) is thought to undergo a transition mutation via deamination to produce thymine (3). This would result in the accumulation of TpG and CpA and depletion of CpG during evolution (4). Consistent with this hypothesis, the gene sample of 26,541 dinucleotides contained CpG at 40% the frequency expected by base composition and the CpG transition products, TpG+CpA, were significantly elevated at 124% of expected random frequency. However, because CpG occurs at only 25% of expected random frequency in the genome, the sampled genes were considerably enriched in this dinucleotide. CpGs were asymmetrically distributed in sequences flanking the genes. 5'-flanking sequences were enriched in CpG at 135% of the frequency expected assuming a symmetrical distribution of all the CpGs in the sampled genes (p less than 0.01), while 3'-flanking regions were depleted in CpG at 40% of expected values (p less than 0.0001). This asymmetry may reflect the role of 5-methylcytosine in gene expression. In contrast the frequencies of GpC and GpT+ ApC did not differ significantly from that predicted by base composition and these dinucleotides were not asymmetrically distributed.

Molecular structure of (m5 dC-dG)3: the role of the methyl group on 5-methyl cytosine in stabilizing Z-DNA

The hexamer (m5 dC-dG)3 has been synthesized and its three-dimensional structure determined by a single crystal X-ray diffraction analysis. The structure has been refined to a final R value of 15.6% at 1.3 A resolution. The molecule forms a left-handed Z-DNA helix which is similar to the unmethylated Z-DNA structure. The presence of the methyl group has resulted in slight changes in the twist angle between successive base pairs and modification of some of the interatomic contacts. Methylation of cytosine in the C5 position is associated with a relative destabilization of the B-DNA structure and a stabilization through hydrophobic bonding of the Z-DNA structure.

Structure of nonintegrated, circular Herpesvirus saimiri and Herpesvirus ateles genomes in tumor cell lines and in vitro-transformed cells

Nonintegrated, circular DNA molecules of Herpesvirus saimiri and Herpesvirus ateles were found in five lymphoid cell lines originating from tumor tissues or established by in vitro immortalization of T lymphocytes. The arrangement of unique (L) and repetitive (H) DNA sequences in circular viral genomes was analyzed by partial denaturation mapping followed by visualization with an electron microscope. Three types of circular viral DNA structures were found. (i) The virus-producing cell line RLC, which is derived from an H. ateles-induced rabbit lymphoma, contains circular viral genomes which consist of a single L-DNA and a single H-DNA region, both the same length as in virion DNA. (ii) The circular viral genomes of the nonproducer cell lines H1591 and A1601, in vitro transformed by H. saimiri and H. ateles, respectively, have deletions in the unique L-DNA region and larger H-DNA regions. Cell line A1601 lacks about 8% of virion L-DNA, and H1591 cells lack about 40% of viral L-DNA information. (iii) The nonproducing H. saimiri tumor cell lines 1670 and 70N2 harbor viral genomes with two L-DNA and two H-DNA regions, respectively. Both types of circular molecules have a long and a short L-segment. The sequence arrangements of circular DNA molecules from H. saimiri-transformed cell lines were compared with those of linear virion DNA by computer alignment of partial denaturation histograms. The L-DNA deletion in cell line H1591 was found to map in the right half of the virion DNA. Comparison of the denaturation patterns of both L regions of cell lines 1670 and 70N2 identified the short L regions as subsets of the long L regions. Thus, circular viral DNA molecules of all four nonproducer cell lines represent defective genomes.

A decrease in S-adenosylmethionine synthetase activity increases the probability of spontaneous sporulation

Starting with a relaxed (relA) strain, mutants with reduced activity of adenosine triphosphate:L-methionine S-adenosyl transferase (EC 2.5.1.6; SAM synthetase) were isolated in Bacillus subtilis. One such mutant (gene symbol metE1) had only 3% of the normal SAM synthetase activity but grew almost as well as the parent strain. Another mutant was isolated (gene symbol spdC1) as being able to sporulate continually at a high frequency; it had one-half the normal SAM synthetase activity at 33 degrees C. Both mutants continually and spontaneously entered spore development at a higher frequency than the parent strain in a medium containing excess glucose, ammonium ions, and phosphate. Sporulation was prevented by a high concentration of SAM (1 mM or more) or by the combination of adenosine and methionine (0.5 mM or more each), both of which are precursors of SAM. In contrast to this continual increase in the spore titer, addition of decoyinine, an inhibitor of GMP synthetase, rapidly initiated massive sporulation. Various amino acid analogs also induced sporulation in the relA strain, the methionine analogs ethionine and selenomethionine being most effective.

Expression of transferred thymidine kinase genes is controlled by methylation

Plasmid pTKx-1, containing the herpes simplex virus gene for thymidine kinase (TK) inserted into the BamHI site of plasmid pBR322, was introduced into Ltk- cells by calcium phosphate precipitation in the absence of carrier DNA. Line 101 is a TK+ derivative of Ltk- that contains multiple copies of pTKx-1 in a multimeric structure. A derivative of 101 that retained but no longer expressed the herpes simplex TK genes (termed 101BU1) and derivatives of line 101BU1 that reexpressed the genes (termed 101H1, 101HC, and 101HG) were selected. The TK genes in 101BU1 were hypermethylated relative to those in the TK+ parent and derivatives. Growth of 101BU1 in the presence of the methylation inhibitor 5-azacytidine resulted in an average 13-fold increase in the number of TK+ reexpressors, DNA from 101BU1 was inactive in secondary gene transfer, whereas DNA from 101 and from TK+ reexpressors was active. These data support a causative relationship between DNA methylation and decreased gene expression. All TK+ reexpressors examined had DNA rearrangements involving TK DNA.

DNA methylation affecting the expression of murine leukemia proviruses

The endogenous, vertically transmitted proviral DNAs of the ecotropic murine leukemia virus in AKR embryo fibroblasts were found to be hypermethylated relative to exogenous AKR murine leukemia virus proviral DNAs acquired by infection of the same cells. The hypermethylated state of the endogenous AKR murine leukemia virus proviruses in these cells correlated with the failure to express AKR murine leukemia virus and the lack of infectivity of cellular DNA. Induction of the endogenous AKR murine leukemia virus proviruses with the methylation antagonist 5-azacytidine suggested a causal connection between DNA methylation and provirus expression. Also found to be relatively hypermethylated and noninfectious were three of six Moloney murine leukemia virus proviral DNAs in an unusual clone of infected rat cells. Recombinant DNA clones which derived from a methylated, noninfectious Moloney provirus of this cell line were found to be highly active upon transfection, suggesting that a potentially active proviral genome can be rendered inactive by cellular DNA methylation. In contrast, in vitro methylation with the bacterial methylases MHpaII and MHhaI only slightly reduced the infectivity of the biologically active cloned proviral DNA. Recombinant DNA clones which derived from a second Moloney provirus of this cell line were noninfectious. An in vitro recombination method was utilized in mapping studies to show that this lack of infectivity was governed by mechanisms other than methylation.

Hypomethylation of Epstein-Barr virus DNA in the nonproducer B-cell line EBR

The conversion of the Epstein-Barr virus-negative Ramos cell line has previously been shown to result in an Epstein-Barr virus-positive non-virus-producer cell line, EBR. We report here that Epstein-Barr virus DNA from EBR alone among several cell lines examined was totally unmethylated at three of four sites containing guanine plus cytosine which were tested. This is in direct contrast to reports of high degrees of methylation in the DNAs of other animal viruses, including herpesviruses, isolated from cells in which the viral genome is expressed at a low level.

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.

Different levels of DNA modification at 5'CCGG in murine erythroleukemia cells and the tissues of normal mouse spleen

The DNA of a Friend erythroleukemic cell line (clone DS-19) and that of mouse spleen from which it was derived were compared with respect to modification at 5'CCGG. Methylation patterns in clone DS-19 DNA were very different from those in normal spleen DNA. Our results suggest that the mechanism by which the internal cytosine in the 5'CCGG sequence is modified has been partially inactivated in clone DS-19. We observed a general decrease of about two-fold in total modification at this site in DS-19 DNA. This general decrease was shown to extend to 5'CCGG sites in specific classes of repeated sequences with two-dimensional displays of restriction fragments. Interspersed repeated sequences which are concertedly modified (i.e., modified at many different chromosomal locations) in spleen, were found to be unmodified at these locations in the cell line. However, we did not detect a difference in DNA modification at 5'CCGG with these techniques when uninduced and hexamethylene-bis-acetamide (HMBA) induced cells of clone DS-19 were compared. In other experiments, we obtained further evidence that selected classes of repeated sequences in Physarum polycephalum are methylated to the same extent independent of chromosomal location.

5-Azacytidine stimulates fetal hemoglobin synthesis in anemic baboons

In an attempt to stimulate Hb F synthesis in baboons by means other than erythropoietic stress, we considered the possibility that an agent that inhibits methylation of CpG sequences in DNA may be effective. 5-Azacytidine, a cytosine analogue that cannot be methylated, is such an agent. Animals whose packed red cell volume was maintained at approximately 20% by bleeding were given 10 daily intravenous injections of the drug (6 mg/kg) in 12 days. Hb F levels in these animals started to increase on day 5 of this regimen and peak levels, which were 6-30 times higher than those produced by bleeding alone, occurred 5-7 days after the last dose of the drug. In animals previously identified as genetically "high" or "low" Hb F responders, the maximal Hb F levels were 70-85% and 35-40% respectively. In dose-response studies 5-azacytidine given daily at 3-4 mg/kg produced maximal Hb F increases. The drug did not correlate the percentage (number) of Hb F-containing cells (F cells) beyond the maximal number achieved by bleeding alone and thus its main effect was to increase Hb F per F cell. The finding that Hb F synthesis can be modulated to such a high degree by a drug may have therapeutic implications--e.g., in sickle cell anemia, in which stimulation of Hb F synthesis may prevent sickling.

Estrogen induces a demethylation at the 5' end region of the chicken vitellogenin gene

The two isoschizomeric restriction endonucleases Msp I and Hpa II have been used to study changes in the methylation pattern of the vitellogenin gene after estrogen treatment. In the liver of the both estrogen-treated and nontreated chickens the vitellogenin gene is heavily methylated. However, estradiol causes a demethylation of a Hpa II site(s) at the 5' end region of the gene. The same Hpa II restriction site(s) is also unmethylated in the liver of egg laying hens but is methylated in the liver of roosters or in erythrocyte DNA. In the mature oviduct where no vitellogenin is synthesized but where several egg white proteins are under the control of estrogen, there is also a demethylation of the 5'end region of the vitellogenin gene. This demethylation does not precede the initiation of transcription of the gene but persists, and increases, even after vitellogenin transcription has ceased in the liver. Taken together, the results strongly suggest that in estrogen-responsive tissues the vitellogenin gene becomes demethylated at its 5' end region after estrogen stimulation whether the gene is expressed or not.

Complete sequence analysis of cDNA clones encoding rat whey phosphoprotein: homology to a protease inhibitor

Lactoprotein clones have been isolated from a rat mammary gland recombinant library of cDNA plasmids. Clones p-Wp 52 and p-Wp 47 were shown by hybrid selection, in vitro translation, and immunoprecipitation to represent a cloned DNA sequence encoding rat whey phosphoprotein. We report here the nucleotide sequence of the cDNA insert of p-Wp 52 and shows that it encodes the complete whey phosphoprotein sequence. The encoded sequence shows a high content of half-cystine, glutamic acid, aspartic acid, and serine but an absence of tyrosine. The half-cystines appear in unique arrangements and are repeated in two domains of the protein. The second domain has striking similarities with the second domain of the red sea turtle protease inhibitor. Clone p-Wp 52 has allowed the study of expression of whey phosphoprotein mRNA during functional differentiation of rat mammary gland and in mammary tumors. The whey phosphoprotein mRNA is detected during midpregnancy and lactation in the rat mammary gland but is barely detected in mammary tumors in which other milk protein mRNAs are expressed. The whey phosphoprotein gene in these tumors is hypermethylated, correlating with the reduced expression of this gene.

Heat- and alkali-induced deamination of 5-methylcytosine and cytosine residues in DNA

5-methylcytosine residues in DNA underwent deamination at high temperatures. Furthermore, their rate of deamination at neutral or alkaline pH was greater than that of cytosine residues in DNA. As sources of [14C]-5-methylcytosine-containing DNA, we used bacteriophage XP-12 DNA, in which 5-methylcytosine residues completely replace C residues, and calf thymus DNA experimentally substituted with [14C] 5-methylcytosine residues. Upon incubation at 95 degrees C in a physiological buffer or at 60 degrees C in 1 M NaOH, the respective rates of deamination of 5-methylcytosine residues were about 3- and 1.5-times those on cytosine residues. Under the same conditions, the free 5-methyldeoxycytidine was converted to thymidine more rapidly than deoxycytidine was converted to deoxyuridine. The reactions at physiological pH and elevated temperature suggest that deamination of 5-methylcytosine residues may yield a significant portion of spontaneous mutations in vivo, especially in view of the lack of thymine-specific mismatch repair systems with specificity and efficiency comparable to that of uracil excision repair systems.

Relative levels of methylation in human growth hormone and chorionic somatomammotropin genes in expressing and non-expressing tissues

It has been shown that the extent of methylation of cytosine in vertebrate DNA is inversely correlated with gene expression. We studied cytosine methylation in and around the homologous human growth hormone (GH) and chorionic somatomammotropin (CS) genes to determine if these genes are undermethylated in DNA from tissues in which they are expressed (pituitary and placenta, respectively) compared to other tissues. Hpa II and Hha I (which cleave only unmethylated 5' CCGG 3' and 5' GCGC 3' respectively) and Msp I (which cleaves CCGG and CmeCGG) were used to digest DNA samples followed by gel electrophoresis, Southern transfer and hybridization with a GH cDNA probe. The extent of methylation of Hpa II and Hha I sites in the GH and CS genes was leukocyte much greater than pituitary greater than placenta = hydatidiform mole. Taken as a whole, our data support the hypothesis that undermethylation is a necessary but not sufficient condition for gene expression since placental and pituitary DNAs are less methylated than leukocyte DNA in this region. However, the correlation between gene expression and undermethylation is imperfect since (1) hydatiform mole DNA has a very similar methylation pattern compared to placental DNA even though moles make little or no CS and (2) the level of methylation of the GH gene compared to the CS gene does not vary in a tissue-specific manner.

Tissue specificity and clustering of methylated cystosines in bovine satellite I DNA

The positions of all 5-methylcytosine (mC) residues in bovine satellite I DNA were determined by sequence analysis of native purified satellite I DNAs from three bovine tissues as well as from cloned DNA. The EcoRI cleavage units from thymus and liver were found to contain 1,402 residues; that from brain contained 1,401 residues. Satellite I DNA from thymus contained a total of 5.0% mC, whereas that from liver and brain contained 4.4% and 2.6% mC, respectively. Thus, the extent of methylation of this DNA is tissue-specific. So is the location. In each tissue, the location of mCs is nonrandom, consisting of three clusters of heavily methylated regions, each of about 200 bases. However, the extent of methylation within each cluster is tissue-specific. The mCs are located entirely in C-G doublets and primarily in palindromic sequences, C-C-G-G sequences (10 methylatable sites) are almost completely methylated in all tissues examined, but T-G-G-A sequences (16 methylated in all tissues examined, but T-G-G-A sequences (16 metylatable sites) are methylated to different extents in each tissue. Neither the tissue specificity of methylation nor the clustering pattern is detectable by examining only G-C-G-G sites, leading us to emphasize the importance of total sequence determination for genomic DNAs in studies of methylation. The clustering pattern, which is preserved despite a 2-fold difference in mC content between brain and thymus, may indicate a role for DNA methylation in chromatin structure.

Unusual methylation pattern of the alpha 2 (l) collagen gene

We studied the methylation pattern of the alpha 2 (type 1) collagen gene in DNA from five cell types with varying rates of type I collagen synthesis: chick embryo fibroblast (CEF), CEF transformed by Rous sarcoma virus, erythrocyte, brain and sperm. The methylation-sensitive restriction enzymes, Msp I, Hpa II, Ava I and Sma I were used to detect methylation in three regions of the alpha 2 (type I) collagen gene: a 5.7 kb region, which includes the start site of transcription and the first two exons of the collagen gene; a 5.2 kb region containing exons in the middle of the gene; and a 3.5 kb region containing exons in the 3' portion of the gene. The DNA around the start site of transcription is not methylated whether or not the cells synthesize collagen. In contrast, the DNA from the central and 3' region of the gene is methylated to about the same extent whether or not the cells make collagen. Our data indicate that a gene that is methylated can be actively transcribed and that the level of expression of the alpha 2 (type I) collagen gene seems to be independent of methylation.

The anatomy of A-, B-, and Z-DNA

Recent advances in DNA synthesis methods have made it possible to carry out single-crystal x-ray analyses of double-stranded DNA molecules of predetermined sequence, with 4 to 12 base pairs. At least one example has been examined from each of the three known families of DNA helix: A, B, and Z. Each family has its own intrinsic restrictions on chain folding and structure. The observed solvent positions in these crystal structures have confirmed earlier fiber and solution measurements, and have led to proposals explaining the transitions from B to A and from B to Z helices. Prospects are improving for an understanding of the mode of bending of DNA in chromatin, and the way in which specific DNA sequences are recognized by drug molecules and repressor proteins.

Amount and distribution of 5-methylcytosine in human DNA from different types of tissues of cells

Analysis of the total base composition of DNA from seven different normal human tissues and eight different types of homogeneous human cell populations revealed considerable tissue-specific and cell-specific differences in the extent of methylation of cytosine residues. The two most highly methylated DNAs were from thymus and brain with 1.00 and 0.98 mole percent 5-methylcytosine (m5C), respectively. The two least methylated DNAs from in vivo sources were placental DNA and sperm DNA, which had 0.76 and 0.84 mole percent m5C, respectively. The differences between these two groups of samples were significant with p less than 0.01. The m5C content of DNA from six human cell lines or strains ranged from 0.57 to 0.85 mole percent. The major and minor base composition of DNA fractionated by reassociation kinetics was also determined. The distribution of m5C among these fractions showed little or no variation with tissue or cell type with the possible exception of sperm DNA. In each case, nonrepetitive DNA sequences were hypomethylated compared to unfractionated DNA.

Methylation of the viral genome in an in vitro model of herpes simplex virus latency

An in vitro model of latency of herpes simplex virus type 1 (HSV-1) in a lymphoid cell line has been developed recently. CEM cells persistently infected with HSV-1 transiently ceased to produce virus for 24 days. This nonproductive state could either be reversed with phytohemagglutinin or maintained with concanavalin A. This system was used to study the relationship between DNA methylation and HSV-1 latency. DNA was probed for methylation by comparing the cleavage patterns generated by two pairs of restriction endonucleases (Sma I vs. Xma I and Hpa II vs. MspI); these enzymes show differential activity reflecting methylation of the recognition sequences. Viral DNA in the concanavalin A-treated cells (not producing virus) was found to be extensively methylated. By contrast, no methylated copies were detected in viral DNA from producer cells. About 800 days after the initial infection, the productive culture once again became nonproductive. Viral sequences in the latter cells were also methylated. Reconstitution experiments revealed 1-2 copies of viral DNA in cells from the latent stages and 40-80 copies in cells from productive stages. Most (if not all) of the viral genome is present in cells from various productive and latent stages. No differences in sequence arrangement were detected (although a terminal fragment of intracellular HSV-1 DNA appeared to be under-represented in latent cells). These results suggest a role for DNA methylation in the mechanism of HSV-1 latency in this system.

Transformation with DNA from 5-azacytidine-reactivated X chromosomes

It has been shown that 5-azacytidine (5-Aza-Cyd) can reactivate genes on the inactive human X chromosome. It is assumed that the 5-Aza-Cyd acts by causing demethylation of the DNA at specific sites, but this cannot be demonstrated directly without a cloned probe. Instead, we have utilized the technique of DNA-mediated transformation to show that the 5-Aza-Cyd-induced reactivation occurs at the DNA level. DNAs from various mouse-human or hamster-human hybrid cell lines, deficient for mouse or hamster hypoxanthine phosphoribosyltransferase (HPRT, EC 2.4.2.8) and varying in whether they contained either an active or inactive human X chromosome, were used in transformation of HPRT- cells. DNA from the active human X chromosome-containing cell lines yielded HPRT+ transformants, whereas DNA from the inactive X chromosome-containing cells lines did not. The inactive X chromosomal DNA was able to transform thymidine kinase-deficient mouse cells, indicating that the DNA solution was normal. These results confirm that inactivation of the X chromosome involves a DNA modification. Furthermore, DNAs from three cell lines with a 5-Aza-Cyd-reactivated X chromosome also transform HPRT- cells, demonstrating that the 5-Aza-Cyd has altered the DNA structure and supporting the idea that methylation plays a role in X chromosome inactivation.

5-methyl-dCTP deaminase induced by bacteriophage XP-12

Bacteriophage XP-12, whose DNA contains 34 mol% 5-methylcytosine, induces the synthesis of a unique enzyme, 5-methyl-dCTP deaminase. The substrate for this enzyme, 5-methyl-dCTP, is produced by reactions catalyzed in part by other phage-induced enzymes, and the product of the reaction is dTTP. The deaminase therefore provides a novel pathway for biosynthesis of thymine residues for phage XP-12 DNA. Evidence is presented that this pathway is used for dTTP synthesis in XP-12-infected Xanthomonas oryzae.

Digestion of highly modified bacteriophage DNA by restriction endonucleases

The ability of thirty Type II restriction endonucleases to cleave five different types of highly modified DNA has been examined. The DNA substrates were derived from relatively large bacteriophage genomes which contain all or most of the cytosine or thymine residues substituted at the 5-position. These substituents were a proton (PBS1 DNA), a hydroxymethyl group (SP01 DNA), a methyl group (XP12 DNA), a glucosylated hydroxymethyl group (T4 DNA), or a phosphoglucuronated, glucosylated 4,5-dihydroxypentyl group (SP15 DNA). Although PBS1 DNA and SP01 DNA were digested by most of the enzymes, they were cleaved much more slowly than was normal DNA by many of them. 5-Methylcytosine-rich XP12 DNA and the multiply modified T4 and SP15 DNAs were resistant to most of these endonucleases. The only enzyme that cleaved all five of these DNAs was TaqI, which fragmented them extensively.

Psoralen-deoxyribonucleic acid photoreaction. Characterization of the monoaddition products from 8-methoxypsoralen and 4,5'8-trimethylpsoralen

The isolation and structural characterization are described of the major monoaddition products formed in the photoreaction of two naturally occurring psoralens, 8-methoxypsoralen and 4,5',8-trimethylpsoralen, with high molecular weight, double-stranded DNA. Hydrolysis of the psoralen-modified DNA and subsequent chromatography resulted in the isolation of four modified nucleosides from each psoralen. Structural characterization was accomplished by mass spectrometry and 1H NMR analysis. The major products, accounting for 44-52% of the covalently bound psoralen, are two diastereomeric thymidine adducts formed by cycloaddition between the 5,6 double bond of the pyrimidine and the 4',5' (furan) double bond of the psoralen. A minor product, less than 2% of the covalently bound psoralen, is a furan-side adduct to deoxyuridine, derived from an initially formed deoxycytidine adduct by hydrolytic deamination. A fourth product is a thymidine adduct where cycloaddition has taken place between the 5,6 double bond of the pyrimidine and the 3,4 (pyrone) double bond of the psoralen. This pyrone-side adduct accounts for 19% of the covalently bound 8-methoxypsoralen but for less than 3% of the covalently bound 4,5'8-trimethylpsoralen. All of the isolated adducts have cis-syn stereochemistry. The stereochemistry and product distribution of the adducts are determined in part by the constraints imposed by the DNA helix on the geometry of the noncovalent intercalation complex formed by psoralen and DNA prior to irradiation.

Studies of X chromosome DNA methylation in normal human cells

Studies of methylation along 28 kilobases of X-chromosome DNA, assayed by Southern blot analysis using cloned X-chromosome-specific probes, indicates that X DNA methylation in normal human cells changes with replication, is not correlated with number of X chromosomes or transcriptional activity, and is less stable and more prevalent than when the human X is in the foreign environment of mouse-human hybrid cells. In contrast with observations of others in heteroploid cells, we observed no derepression of the inactive X in clonal populations of normal human fibroblasts treated with 5-azacytidine. This may reflect the differences in stability of the methylation of the human X in a foreign environment. Our observations preclude ubiquitous methylation differences as the molecular basis for X-chromosome inactivation.

Expression of a cloned adenovirus gene is inhibited by in vitro methylation

In many viral and nonviral eukaryotic systems, an inverse correlation has been observed between the extent of DNA methylation at 5'-C-C-G-G-3' sites and the extent of expression of specific genes as mRNA. The E2a region of adenovirus serotype 2 (Ad2) DNA encodes the Ad2-specific DNA binding protein required for viral DNA replication. In three lines of Ad2 transformed hamster cells (HE1, HE2, and HE3), multiple copies of the major part of the Ad2 genome persist in an integrated state. Cell lines HE2 and HE3 do not express the DNA-binding protein whereas line HE1 does so. It has been shown that, in cell line HE1, all 5'-C-C-G-G-3' (Hpa II/MspI) sites in the E2a region remain unmethylated. Conversely, in lines HE2 and HE3 lacking expression of the E2a region all Hpa II sites are methylated. The cloned E2a region of Ad2 DNA, the HindIII A fragment in pBR322, was methylated in vitro by using Hpa II DNA methyltransferase (5'-C-C*G-G-3') or was left unmethylated. In vitro methylation did not break or nick supercoiled circular DNA. Methylated or unmethylated DNA was then microinjected into the nuclei of Xenopus laevis oocytes, and the subsequent synthesis of Ad2-specific RNA was monitored. In vitro-methylated DNA remained in the methylated state for 24 hr on microinjection into nuclei of xenopus oocytes; unmethylated DNA remained unmethylated. When the injected DNA had been methylated by using Hpa H DNA methyltransferase, Ad2-specific RNA was not synthesized as late as 24 hr after microinjection. Unmethylated DNA was readily expressed into Ad2-specific RNA. As an internal control, unmethylated histone genes (h22 DNA) from sea urchin were microinjected together with methylated E2a DNA from Ad2. Ad2-specific RNA was not found; h22 DNA-specific RNA was readily detected. This finding ruled out nonspecific inhibitory effects in the methylated DNA preparation. Ir was also shown that transcription of the unmethylated HindIII A fragment of Ad2 DNA in Xenopus oocytes was initiated on the late promoter of the E2a region. The same promoter was used in productively infected KB cells. Methylation by BsuRI methylse (5'-G'G-C*C-3') did not inactivate the HindIII A fragment. These results provide evidence for the notion that methylated sequences at highly specific sites are involved in the regulation of gene expression. The actual nature of the regulatory signal is not yet understood.

Modifications of chromatin structure in control of gene expression

Repetitive sequences in intron and spacer DNA could be sites for binding of chromosomal proteins which maintain chromatin structure and control gene activity. Methylation of DNA guides the binding of acidic nonhistone proteins and maintains the differentiation state during DNA replication. Differentiation inducers modify repressor proteins permitting unfolding of chromatin. Histone H 1 must be removed for gene activity. Phosphorylation of nonhistone proteins probably induces allosteric modifications which permit unfolding of chromatin. Acetylation of nucleosomal histones is necessary to permit passage of RNA polymerase. Deacetylation quickly returns the gene to a normal histone repressed state. Chromosomal RNA attached to nonhistone proteins aids the binding of RNA polymerase to the DNA template. Carcinogens can disrupt normal gene control leading to circumvention of normal cell cycle controls.

Effect of methylation on expression of microinjected genes

The cloned genes for the simian virus 40 large tumor antigen and for herpes simplex virus (HSV) thymidine kinase (TK) were methylated with EcoRI methylase. The genes were microinjected into the nuclei of TK-deficient (tk-) cells, and expression of the genes was determined by immunofluorescence staining for the simian virus 40 large tumor antigen and by [3H]thymidine incorporation followed by autoradiography for HSV TK. We found that methylation of the simian virus 40 gene, under EcoRI or EcoRI* conditions, resulting in methylation at sites within the gene and in the surrounding sequences, has no effect on expression of the large tumor antigen when the gene is manually microinjected into mammalian nuclei. However, methylation of the HSV tk gene at the two EcoRI sites markedly reduces or abolishes the expression of this gene. One of the EcoRI sites of HSV tk is approximately 1.1 kilobases downstream from the 3' end of the gene and is believed to have no regulatory function in the expression of the tk gene. The other EcoRI site is 79 base pairs upstream from the 5' end of the gene and has considerable homology to the regulatory sequence proposed by [Benoist C., O'Hare, K., Breathnach, R., & Chambon, P. (1980) Nucleic Acids Res. 8, 127-142]. Our results are direct proof that methylation can alter gene expression and also that the effect depends strictly on the sites that are methylated.

A bacteriophage-induced 5-methyldeoxycytidine 5'-monophosphate kinase

Bacteriophage XP-12-infected Xanthomonas oryzae have been found to be a source of a kinase preparation which converts m5dCMP to m5dCDP and then to m5dCTP using ATP as the phosphate donor. Optimal formation of the triphosphate required the presence of creatine phosphate and creatine kinase. In the presence of dGTP, dTTP and dATP, Escherichia coli DNA polymerase I and T4 DNA polymerase catalyzed the incorporation of m5dCTP into DNA just as efficiently as that of dCTP. Neither dTMP nor dCMP served as substrate for the m5dCMP monophosphate kinase. Analogous preparations from uninfected X. oryzae were unable to phosphorylate m5dCMP.

Pre-adipocyte determination either by insulin or by 5-azacytidine

CHEF/18 is a diploid Chinese hamster cell line of embryonic origin, which is fibroblastic in structure, but behaves like a mesenchymal stem cell line in its ability to differentiate into adipocytes, myoblasts, and chondrocytes. With these cells, adipocyte formation has been divided experimentally into two stages: (i) determination of pre-adipocytes, which have lost the ability to form other cell types while retaining their fibroblast structure; and (ii) commitment or terminal differentiation, in which lipids accumulate, adipocyte structure develops, and cells lose the ability to divide. This paper reports that the first stage can be induced by exposure to 5-azacytidine or 2'-deoxy-5-azacytidine, drugs that also induce CHEF cells to form other mesenchymal cell types, or by growth with added insulin. Pre-adipocytes are distinguished from CHEF stem cells by (i) their inability to form other mesenchymal cell types; and (ii) their rapid accumulation of lipid in response to added insulin. The possibility is discussed that both insulin and the cytidine analogs promote differentiation by the same mechanism, namely changes in the pattern of DNA methylation.

Cytochrome P1-450 structural gene in mouse, rat, and rabbit: differences in DNA methylation and developmental expression of mRNA

Clone 46 previously was shown to represent an 1100-bp cDNA clone of the mouse cytochrome P1-450 structural gene. Clone 46 [32P]DNA was hybridized to DNA and mRNA from mouse, rat, and rabbit of different ages. In Hpa II digests of DNA, two hybridizable fragments of less than 0.5 kb exist in adult "Ah-responsive" C57BL/6N mouse liver but not in C57BL/6N sperm or embryo or in adult "Ah-nonresponsive" DBA/2N mouse liver. The reason for this hypomethylation of adult C57BL/6N liver DNA is not known but might be related to the high degree of expressivity of this gene in adult C57BL/6N liver, compared with adult DBA/2N liver. No differences in Hpa II- or Msp I-digested DNA are seen in C57BL/6N or DBA/2N inbred strains treated with the P1-450 inducer, 3-methylcholanthrene (3-MC), versus untreated controls. The ontogenetic expression of 3-MC-induced P1-450 mRNA (23S) from mouse or rat liver corresponds well to previous developmental studies from this laboratory involving 3-MC-inducible aryl hydrocarbon (benzo[a]pyrene) hydroxylase (AHH) (EC 1.14.14.1) activity. P1-450 mRNA--induced transplacentally by 3-MC given to the mother--is readily detectable by clone 46 as early as gestational day 15. The cloned cDNA probe hybridizes to rat and rabbit DNA fragments of different sizes, and with less intensity, when compared with hybridization to mouse DNA. No hybridization of this DNA is observed with rabbit mRNA of all ages ranging from neonate to adult. These data suggest that sequence homology exists among the mouse, rat, and rabbit P1-450 structural genes and between mouse and rat P1-450 mRNA. The mouse cDNA probe is believed to hybridize to a segment of the rabbit P1-450 gene that is not transcribed.

DNA methylation in eukaryotes

The DNA of higher eukaryotes contains one minor base, namely 5-methylcytosine. The distribution of this minor base between different species and different DNA fractions will be considered together with the actual sequences methylated. The properties of the enzyme responsible for DNA modification will be reviewed, particular note being paid to the efficiency of methylation of different DNA substrates. Various possible functions of the 5-methylcytosine in DNA will be considered and particular attention will be paid to the finding that specific modified bases present in DNA not undergoing transcription are absent in the same genes when these are being actively transcribed.