Effect of in vitro histone phosphorylation on template activity of rat liver chromatin

Abstract

Conditions for maximal phosphorylation of rat liver chromatin and free histones by a cyclic 3′,5′-AMP-dependent protein kinase preparation from calf liver were determined, and template activities of phosphorylated and non-phosphorylated chromatin for Escherichia coli RNA polymerase were compared. The phosphorylation of chromatin by the protein kinase preparation is stimulated by cyclic AMP, and both histones and non-histone proteins in chromatin are phosphorylated. The 32P in histones from phosphorylated chromatin is principally in very lysine-rich histones (F1). In contrast, rat liver histones phosphorylated in the free state contain 54 % of the 32P in slightly lysine-rich histones (F2b) and 14 and 18 % of the 32P in F1 histones and glycine-arginine-rich histones (F2a1), respectively. The F1 histones bound to native DNA are poor substrates for the protein kinase preparation in comparison with free histones. However, F1 histones bound to denatured DNA are phosphorylated as effectively as free F1 histones. The inhibition of phosphorylation of F1 histones bound to native DNA increases as the ratio of histones to DNA increases, suggesting that F1 histones are bound to the DNA in compact clusters resulting in mutual interference with the phosphorylation sites. Maximally phosphorylated chromatin and the chromatin reconstituted with maximally phosphorylated histones show the same template activity for E. coli RNA polymerase as corresponding non-phosphorylated chromatin in terms of rate of incorporation of labeled nucleotides into RNA. Contrary to expectation, the chromatin reconstituted with histones from regenerating rat liver exhibits a slightly smaller template activity for E. coli RNA polymerase than the chromatin reconstituted with histones from normal rat liver.