Histone modification in constitutive heterochromatin versus unexpressed euchromatin in human cells

Histone modifications are implicated in regulating chromatin condensation but it is unclear how they differ between constitutive heterochromatin and unexpressed euchromatin. Chromatin immunoprecipitation (ChIP) assays were done on various human cell populations using antibodies specific for acetylated or methylated forms of histone H3 or H4. Analysis of the immunoprecipitates was by quantitative real-time PCR or semi-quantitative PCR (SQ-PCR). Of eight tested antibodies, the one for histone H4 acetylated at lysine 4, 8, 12, or 16 was best for distinguishing constitutive heterochromatin from unexpressed euchromatin, but differences in the extent of immunoprecipitation of these two types of chromatin were only modest, although highly reproducible. With this antibody, there was an average of 2.5-fold less immunoprecipitation of three constitutive heterochromatin regions than of four unexpressed euchromatic gene regions and about 15-fold less immunoprecipitation of these heterochromatin standards than of two constitutively expressed gene standards (P < 0.001). We also analyzed histone acetylation and methylation by immunocytochemistry with antibodies to H4 acetylated at lysine 8, H3 trimethylated at lysine 9, and H3 methylated at lysine 4. In addition, immunocytochemical analysis was done with an antibody to heterochromatin protein 1alpha (HP1alpha), whose preferential binding to heterochromatin has been linked to trimethylation of H3 at lysine 9. Our combined ChIP and immunocytochemical results suggest that factors other than hypoacetylation of the N-terminal tails of H4 and hypermethylation of H3 at lysine 9 can play an important role in determining whether a chromatin sequence in mammalian cells is constitutively heterochromatic. (C) 2004 Wiley-Liss, Inc.