|Home | About | Journals | Submit | Contact Us | Français|
Archaeal histone-encoding genes have been identified in marine Crenarchaea. The protein encoded by a representative of these genes, synthesized in vitro and expressed in Escherichia coli, binds DNA and forms complexes with properties typical of an archaeal histone. The discovery of histones in Crenarchaea supports the argument that histones evolved before the divergence of Archaea and Eukarya.
Histones bind and compact nuclear DNA in virtually all eukaryotes (17). Histones were therefore presumably a feature of the first nucleus and also of its immediate ancestor, but the origin of histones has been an unresolved issue. Bacteria do not have histones but Archaea do, a fact which argues that histones evolved in the archaeal lineage after the Bacteria-Archaea divergence (Fig. (Fig.1A)1A) (10, 21). Histones have, however, been documented only in Euryarchaea, and their conspicuous absence in Crenarchaea is often cited and noted as a difference between the two major archaeal lineages (1, 14). If valid, this would argue that histones evolved relatively late, after the Euryarchaea-Crenarchaea divergence (Fig. (Fig.1A),1A), consistent with proposals, such as the hydrogen hypothesis (8), that posit a euryarchaeal origin for the eukaryotic nucleus. Alternatively, histones could have been lost in the lineages leading to all the thermophilic Crenarchaea studied to date. Support for this alternative was recently provided by the identification of Nanoarchaeum equitans, claimed as the first member of a third, deep-branching archaeal lineage (Fig. (Fig.1A)1A) (20), and the presence of two histone-encoding genes in the N. equitans genome (Fig. (Fig.1B).1B). Only thermophilic and hyperthermophilic Crenarchaea have been grown and studied in pure culture, but very large numbers of as-yet-uncultured Crenarchaea are known to be present in cold marine environments (2, 5, 6, 9). We therefore screened the DNA sequences obtained from microorganisms filtered from the Sargasso Sea (18) for histone genes. Genes that encode three closely related proteins (EAG39378 [hereafter designated EAG3], EAJ42364, and EAK74163) that have sequences, key residues, and basic pIs consistent with being archaeal histones are present in this database (Fig. (Fig.1B).1B). These proteins do not, however, have two prolines near the N terminus as found in almost all other archaeal histones (e.g., P4 and P7 in HMfB) (Fig. (Fig.1B)1B) (14). In an archaeal histone dimer, these residues form a proline tetrad motif that has been proposed to position surface-located residues appropriately for DNA binding (4).
Microorganisms were not cultured from the Sargasso Sea, and therefore, to pursue this observation, a gene encoding the EAG3 sequence was synthesized with codon usage optimized for expression in Escherichia coli (http://www.kazusa.or.jp/codon) and NdeI, SacI, SphI, and BamHI sites incorporated to facilitate gene assembly and cloning. Oligonucleotides (Operon Biotechnologies Inc., Huntsville, AL) were mixed in complementary pairs (Table (Table1),1), phosphorylated using T4 kinase (Invitrogen, Carlsbad, CA), purified through Sephadex G-50 spin columns (Amersham Biosciences, Piscataway, NJ), heated to 70°C, and allowed to cool to 25°C. The resulting double-stranded DNA molecules had single-strand extensions that facilitated their cloning into pUC19. The DNA molecules generated in oligonucleotide mixtures 1, 2, and 3 (Table (Table1)1) were ligated together and cloned into NdeI- and SacI-digested pUC19. The DNA molecules generated in mixtures 4 and 5 and in mixtures 6 and 7 were similarly ligated together and then cloned into SacI- and SphI-digested pUC19 and into SphI- and BamHI-digested pUC19, respectively. Preparations of the resulting plasmids, pLC1A, pLC2A, and pLC3A, were isolated after transformation and amplification in E. coli, and the EAG3-encoding DNA sequences were released by digestion with NdeI plus SacI, SacI plus SphI, and SphI plus BamHI, respectively. These DNAs were ligated together and cloned into SphI- and BamHI-digested pT7-7, generating pLC16B that was transformed into E. coli BL21 (DE3). Cultures of E. coli BL21 (DE3; pLC16B) were grown in LB medium to an optical density at 600 nm of ~0.7, IPTG (isopropyl-β-d-thiogalactopyranoside) was added (1 mM final concentration), and incubation continued for 3 h at 37°C. The cells were harvested by centrifugation, resuspended in 1 M NaCl and 50 mM Tris-HCl (pH 8), and lysed and a clarified extract of soluble protein prepared as previously described (15). The polypeptides present in aliquots of each fraction generated from the extract by gel filtration chromatography (HiPrep16/60 Sephacryl S-100 HR column; Amersham Biosciences) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) through 15% (wt/vol) polyacrylamide gels (Bio-Rad, Hercules, CA) and visualized by GelCode blue staining (Pierce Biotechnology, Rockford, IL). Fractions that contained large amounts of a polypeptide with the electrophoretic mobility predicted for EAG3 were pooled, dialyzed against 0.1 M NaCl and 50 mM Tris-HCl (pH 8), and subjected to cation exchange chromatography (HiTrap CM Sepharose FF column; Amersham Biosciences). Polypeptides in fractions that eluted between 0.1 and 0.5 M NaCl were visualized by GelCode blue staining after SDS-PAGE. The identity of the material in a stained band was confirmed as EAG3 by electrospray mass analysis using a Micromass QTOF II instrument, and fractions that contained only EAG were pooled and stored at −70°C.
Dimerization is required for native histone fold formation (17), and all histones are dimers or higher-order oligomers in solution. Consistent with this, formaldehyde cross-linking of EAG3 resulted predominantly in molecules with the mobility of a dimer and, to a lesser extent, the mobility of tetramers (Fig. (Fig.1C).1C). EAG3 bound spontaneously to DNA and formed complexes with linear DNA molecules that migrated faster during agarose gel electrophoresis than the original histone-free DNA (Fig. (Fig.1D).1D). Binding to closed circular DNAs introduced positive supercoils (Fig. (Fig.1E).1E). These are both unique and characteristic features of complexes formed by archaeal histones (12, 14, 15) and argue convincingly that EAG3 is an archaeal histone.
Most of the archaeal genes in the Sargasso Sea database appear to be derived from Archaea closely related to previously documented and abundant planktonic crenarchaeal groups (2, 5, 6, 9, 18). The origin of EAG3 was nevertheless uncertain until a gene encoding a close homologue (designated HCs [Fig. [Fig.1B])1B]) was identified in the genome of Cenarchaeum symbiosum, a nonthermophilic crenarchaeon found in high abundance in association with a marine sponge (Fig. (Fig.1A)1A) (13, 16) (GenBank accession number DQ090941). The genome sequence of C. symbiosum is now almost complete (S. J. Hallam and E. F. DeLong, unpublished results), and as previously reported (13, 16), this verifies that C. symbiosum has a deep affiliation within the crenarchaeal lineage and occupies a basal phylogenetic position relative to other known Crenarchaea and Euryarchaea. The Sargasso Sea and C. symbiosum DNA sequences have substantially different G+C contents, approximately 38% and 58% G+C, respectively, but the EAG3 and HCs amino acid sequences are ~70% identical. The EAG3 and HCs amino acid sequences are also 35% to 45% identical to the sequences of all known euryarchaeal and nanoarchaeal histones. Consistent with being archaeal histones, they lack the N-terminal extensions present in all eukaryotic histones (Fig. (Fig.1B).1B). It remains formally possible that EAG3 and HCs have evolved from an ancestor that entered their common crenarchaeal lineage by lateral gene transfer, but given the very different G+C contents of the genomic DNAs, if this transfer occurred, it was not a recent event. In phylogenetic tree constructions, EAG3 and HCs consistently group outside the euryarchaeal histones (unpublished data), also indicating that the crenarchaeal histones were not acquired by lateral gene transfer from a euryarchaeon.
With the presence of histones in Crenarchaea, there is strong parsimonious support for the argument that histones evolved very early, after the divergence of Bacteria-Archaea but before the separation of the archaeal and eukaryotic lineages. This also adds support to the proposal that it was the existence of the efficient histone-based mechanism of DNA packaging that allowed the genome expansion needed for eukaryotic evolution (11). The machineries to replicate and transcribe histone-bound DNA must presumably have evolved in parallel, and these machineries do also have many features in common in Archaea and the eukaryotic nucleus that differ from their bacterial counterparts (7, 14). Apparently, histones were lost and replaced by structurally unrelated chromatin proteins in modern hyperthermophilic Crenarchaea but, intriguingly, posttranslation acetylation may also regulate the interactions of these nonhistone proteins with DNA (3, 19).
This research was supported by grants from the Department of Energy, National Science Foundation, Gordon and Betty Moore Foundation, and National Institutes of Health.