Biotinylation of lysine 16 in histone H4 contributes toward nucleosome condensation

Abstract

Holocarboxylase synthetase (HLCS) is part of a multiprotein gene repression complex and catalyzes the covalent binding of biotin to lysines (K) in histones H3 and H4, thereby creating rare gene repression marks such as K16-biotinylated histone H4 (H4K16bio). We tested the hypothesis that H4K16bio contributes toward nucleosome condensation and gene repression by HLCS. We used recombinant histone H4 in which K16 was mutated to a cysteine (H4K16C) for subsequent chemical biotinylation of the sulfhydryl group to create H4K16Cbio. Nucleosomes were assembled by using H4K16Cbio and the ‘Widom 601’ nucleosomal DNA position sequence; biotin-free histone H4 and H4K16C were used as controls. Nucleosomal compaction was analyzed using atomic force microscopy (AFM). The length of DNA per nucleosome was ∼30% greater in H4K16Cbio-containing histone octamers (61.14 ± 10.92 nm) compared with native H4 (46.89 ± 12.6 nm) and H4K16C (47.26 ± 10.32 nm), suggesting biotin-dependent chromatin condensation (P < 0.001). Likewise, the number of DNA turns around histone core octamers was ∼17.2% greater in in H4K16Cbio-containing octamers (1.78 ± 0.16) compared with native H4 (1.52 ± 0.21) and H4K16C (1.52 ± 0.17), judged by the rotation angle (P < 0.001; N = 150). We conclude that biotinylation of K16 in histone H4 contributes toward chromatin condensation.

Introduction

Nucleosomal core particles (NCP) are composed of a histone H3/H3/H4/H4 tetramer and two histone H2A/H2B dimers, around which ∼147 bp of DNA is wrapped in ∼1.7 turns [1], [2]. Nucleosomal (chromatin) condensation hinders the access of proteins such as RNA polymerases to DNA. A number of strategies have evolved to regulate access to nucleosomal DNA, including posttranslational modification of histones, the incorporation of histone variants, and ATP-dependent chromatin remodeling [3]. Histones are modified post-translationally by covalent attachment of a variety of predominantly small molecules [4], [5], [6], [7], [8]. These modifications provide an important regulatory mechanism for regulating gene expression, replication, and DNA repair [9], [10]. For example, acetylation of lysines (K) 5, 8, 12, and 16 in histone H4 is associated with transcriptionally active chromatin, whereas histone deacetylation is associated with inactive chromatin [11], [12]. Acetylation of K16 in histone H4 (H4K16ac) is a frequent epigenetic mark with major implications in chromatin structure [13], [14], [15].

Histones H3 and H4 are also modified by covalent binding of biotin to lysine residues. Biotinylated histones have been detected using radiotracers, streptavidin, anti-biotin, and biotinylation site-specific antibodies as probes [5], [7], [8], [16], [17], [18]. Biotinylation sites include K12 and K16 in histone H4 (H4K12bio and H4K16bio) [7], [8], [17], [18].

Biotinylation of histones is catalyzed by holocarboxylase synthetase (HLCS) [19], [20]. HLCS knockdown produces strong phenotypes including heat stress susceptibility and short life span in Drosophila melanogaster [21], and de-repression of long terminal repeats in humans, mice, and flies [22]. The abundance of biotinylation marks in the human epigenome depends on dietary biotin in humans and fruit fly [22], [23] and on the concentration of biotin in cell culture media [17], [18], [22], [24]. Histone biotinylation marks are enriched in repressed loci [17], [18], [22], [24], [25], where they co-localize with the well-established H3K9me gene repression mark [12]. The putative role of biotin in gene repression has been corroborated by atomic force microscopy (AFM¹) studies of H4K12bio, suggesting that biotinylation of histones causes condensation of nucleosomes [26].

Biotinylation of histones is a rare event and only <0.001% of histones H3 and H4 are biotinylated [5], [8], [27]. The rarity of biotinylation marks raises concern as to how this modification can elicit major biological effects and precipitate severe phenotypes. HLCS catalyzes the binding of biotin to both histones and carboxylases [28], rendering it difficult to assign phenotypes solely and unambiguously to changes in histone biotinylation. Recent studies in transgenic Drosophila suggest for the first time that phenotypes are different for HLCS knockdown flies (decreased heat stress survival, short life span) compared with carboxylase mutants (increased stress survival), thereby implying histone biotinylation in gene regulation by mechanisms distinct from those due to changes in carboxylase biotinylation [21], [29].

Due to remaining uncertainties regarding histone biotinylation we integrated previous findings by us and others into the following novel model. We propose that the roles of biotin and HLCS in gene regulation are mediated primarily by HLCS-dependent assembly of a multiprotein gene repression complex in human chromatin [8], and that biotinylation marks are markers for HLCS docking sites in chromatin in addition to contributing toward chromatin condensation [26]. Given the pre-eminent role of K16 marks in altering chromatin condensation states [13], [14] we hypothesized that the newly discovered H4K16bio mark [18] has a greater effect on mediating chromatin condensation than that described before for H4K12bio [26]. Here we quantified H4K16bio-dependent chromatin condensation by AFM [26], [30], [31], [32].