We have developed a procedure for purifying highly specific polyclonal antibodies against 5-methylcytosine. These antibodies were used to probe the distribution of 5-methylcytosine among fractionated nucleosomes of mouse cell chromatin. Our results demonstrate that at least 80% of the 5-methylcytosine is localized in nucleosomes that contain histone HI. Native nucleosomes that lack histone HI or possess high mobility group proteins package DNA that is 1.6-to 2.3-fold undermethylated. We suggest that the preferential association of methylated sequences with histone HI has functional significance because DNA methylation has been linked to gene inactivation and histone HI is known to promote chromatin condensation.Although DNA methylation has been implicated in the epigenetic repression of genetic information in higher eukaryotes (see refs. 1 and 2 for reviews), the mechanism of this repression has not yet been elucidated. In prokaryotes, DNA methylation has been shown to alter the binding affinities of specific proteins for specific DNA sequences (3,4). If a generally similar (but not necessarily identical) mechanism exists in eukaryotes, then one might predict that methylation either reduces or promotes the binding of specific chromosomal proteins to DNA.Previous studies on chromatin have suggested that sequences that contain 5-methylcytosine (m5C) are packaged into nucleosomes (5-7) and may be primarily localized within core particles (7). However, nucleosomes are chemically heterogeneous because of the association of different accessory proteins with histone octamer-DNA complexes. The highly abundant nonhistone proteins, high mobility group (HMG) 14 and 17, are believed to be bound to nucleosomes in transcribed regions of chromatin (8), whereas some evidence suggests that histone Hi may be enriched in inactive chromatin (9, 10). Most nucleosomes that possess HMG proteins seem to lack histone Hi and vice versa (11), indicating that these proteins are nonrandomly distributed, although not every nucleosome that lacks HMG proteins possesses histone Hi (12). Based on stoichiometry estimates, at least 40% of the nucleosomes of a cultured mouse cell line lack histone Hi (13), but Hi content varies between biological systems (13,14), and estimates of the fraction of nucleosomes that lack HI in vivo are dependent on whether 1 or 2 mol of this protein are bound per nucleosome (15).The proteins associated with nucleosomes that package methylated DNA have not been identified. We have addressed this question in the present study by fractionating nucleosomes into histone Hl-depleted (fraction SI) and histone Hil-enriched (fraction S2) components and have further resolved these fractions by gel electrophoresis into either DNA components or discrete nucleosome subsets that possess well-defined protein compositions. By modifying the immunological techniques of Sager and co-workers (16) to probe the m5C content among these species, we have found that m5C is localized primarily in nucleosomes that contain histone H1 and ...

The minor histone H2A subtype, H2A.Z, has been purified to homogeneity from calf thymus and subjected to automated Edman degradation. The sequence of the first 30 amino acids possesses only 60% homology with major H2A subtypes of the same tissue. This sequence difference is more extreme than that exhibited between evolutionarily distant major H2A subtypes. However, an analysis of secondary structure reveals that H2A.Z and major H2A subtypes exhibit the same general topographical features within their N‐terminal domains.

Electrophoresis fractionates nucleosomes which possess different protein compositions. We report here a procedure for transferring the DNA components of electrophoretically resolved nucleosomes to diazobenzyloxymethyl cellulose (DBM-paper). Histones are first removed from nucleosome components by electrophoresis in the presence of cetyltrimethylammonium bromide (CTAB), leaving DNA fragments fixed within the original gel as the CTAB salts. The DNA is then converted to the sodium salt, denatured, and electrophoretically transferred to DBM-paper. The overall pattern of DNA on the resulting blot is visualized either by fluorography or by immunoautoradiography. This DNA pattern is then compared with autoradiograms obtained after hybridizing the same blot with specific 32P-labeled probes. Using mouse satellite DNA as a hybridization probe, we illustrate the above techniques and demonstrate that nucleosomes carrying satellite sequences are compositionally heterogeneous. The procedures described here should also be useful in the analysis of the nucleic acid components associated with other nucleoprotein complexes.

No abstract

The 2',3'-dialdehyde of adenosine 5'-diphosphate, oADP, exhibited the properties of an affinity label with Escherichia coli succinyl-CoA synthetase. Inactivation of this synthetase by oADP followed pseudo-first-order kinetics and was competitively blocked by ADP. The stoichiometry of labeling of the synthetase was 1 mol/mol ufl or, extrapolated, 2 mol/mol inactive a2& molecule. oADP also exhibited the properties of a substrate, bringing about rapid dephosphorylation of the enzyme. Further specificity of oADP was demonstrated in partially inactivated succinyl-CoA synthetase by selective inhibition of the succinate t* succinyl-CoA exchange reaction, in comparison to the CoA tt succinyl-CoA exchange reaction. Modification of the synthetase by oADP resulted in cross-linking of the enzyme, casting uncertainty over the subunit binding site for ADP. Modification of the synthetase by ADP-2'-semialdehyde occurred at a faster rate than that by oADP but exhibited biphasic inhibitor concentration dependence and did not exhibit saturability.Succinyl-CoA synthetase catalyzes the reaction shown in Eqn (1)( 1 1 where NDP and NTP represent purine ribonucleoside diphosphate and triphosphate, respectively. Partial reactions that reflect the catalytic steps of the synthetase activity are given in Eqns (2a -c).The enzymes from pig heart and rat liver have an aj?quaternary structure [I, 21, whereas the Escherichia coli enzyme can be described as an u& protein [3]. The u subunit is phosphorylated during the reaction [1,3]. The subunit contains an important sulfhydryl group that is labeled by CoA thiosulfonate, an affinity labeling agent of the enzyme [4,5]. These data are consistent with the view that the active site of the enzyme is located at the point of contact between subunits. Recently, it has been hypothesized that E. coli succinylCoA synthetase exhibits alternating sites cooperativity [6]. In the model, the two active sites are viewed as being alternatively productive with substrate (ATP or succinyl-CoA) binding at the nonproductive site bringing about release of product (succinvl-CoA or ATP) from the oroductive site. The phenomena associated with the alternating sites concept were not exhibited by the up pig heart enzyme. Other data that are consistent with the hypothesis have also been publishedWe have studied the NTP(NDP) binding site of the UP rat liver enzyme with oGDP as the probe [2]. This compound brough about dephosphorylation of the enzyme and subsequent rapid inactivation of the enzyme. The dephosphorylated enzyme was also much more rapidly inactivated by oGDP than the phosphorylated enzyme. Data indicative of binding to both u and p subunits was obtained, suggesting participation of both polypeptide chains in the make-up of the NTP(NDP) binding site. It was thus logical to extend the study to the E. coli enzyme with its two active sites that appear to interact. In the present study, the effects of oADP have been examined as an affinity label of the E. coli enzyme.