October 1, 1997;
Xenopus HDm, a maternally expressed histone deacetylase, belongs to an ancient family of acetyl-metabolizing enzymes.
Modification of core histones can alter chromatin structure, facilitating the activation and repression of genes. A key example is the acetylation of N-terminal lysines of the core histones. Recently, the mammalian histone deacetylase HD1
was cloned from Jurkat T cells, and shown to be 60% identical to the yeast global gene regulator Rpd3
(Taunton et al., 1996). Here we report the cloning of HDm
, a maternally expressed putative deposition histone deacetylase from Xenopus laevis. Comparison of the amino acid sequences of histone deacetylases from diverse eukaryotes shows high levels of identity within a putative enzyme core region. Further alignment with other types of protein: acetoin-utilizing enzymes from eubacteria; acetylpolyamine hydrolases from mycoplasma and cyanobacteria; and a protein of unknown function from an archaebacterium, reveals an apparently conserved core, and suggests that histone deacetylases belong to an ancient family of enzymes with related functions.
[+] show captions
Molecular cloning of HDm and strategy for studying expression. (A) cDNA sequence and deduced amino acid sequence of clone AB21 (EMBL data bank accession number: X78454). The cytoplasmic polyadenylation element and the conventional polyadenylation element are underlined. Putative instability elements are shown in italics. (B) Restriction map of AB21 and locations of the antisense hybridization probe, the open reading frame and synthesized polypeptides. B, BglII; D, DraI; E, EaeI; F, FspI; H, HindIII; P, PstI; R, EcoRI; V, PvuII. The open reading frame (ORF) consists of a conserved N-terminal region (cross-hatched) and a highly charged C-terminal region (filled). Subclones were used to produce glutathione S-transferase (GST) fusions as shown. ΔR/ΔH and ΔV were used to raise antibodies in rabbits. AB21 was isolated from an expression library constructed in λZapII (Stratagene) from poly(A+)RNA of previtellogenic oocytes of Xenopus laevis. Selection was made by immunoscreening using antiserum HP, directed against a subset of mRNA-associated proteins. The riboprobe was synthesized from 1 μg of a HindIII–PstI subclone in pBSII KS (Stratagene) using T3 RNA polymerase (Pharmacia) in the presence of 2 mCi/ml of [α-32P]CTP (800 Ci/mmol, Amersham). GST fusion proteins were expressed from pGEX (Pharmacia) vectors and isolated on glutathione–Sepharose 4B according to the manufacturer's instructions.
Expression of HDm during early development at the level of RNA. (A) Autoradiograph of Northern blot of total RNA from oocytes (O, stage VI); and from embryos at: cleavage (C, stage 4); blastula (B, stage 8); gastrula (G, stage 12); neurula (N, stage 16); tail-bud (TB, stage 26); swimming tadpole (ST, stage 40). Transfers were hybridized with the riboprobe described in Fig. 1. (B) As in (A) but hybridized with an antisense riboprobe covering the length (476 nucleotides) of the cDNA encoding the X. laevis ribosomal protein L22 (EMBL data bank accession number: X94243). Total RNA equivalent to three oocytes or embryos was separated on denaturing formaldehyde/formamide gels transferred to nylon membranes (Zeta-Probe, Bio-Rad) and hybridized for 18 h with the HDm and L22 riboprobes labelled to similar specific activity (approx. 2 mCi/mg). After washing to high stringency (0.1×SSC at 65°C), blots were exposed for 18 h in contact with X-ray film (Agfa).
Immunoblot of proteins extracted from oocyte stages III and VI (mid- and late-oogenesis) and the same embryo stages as described in Fig. 2. Total protein minus pigment and yolk, equivalent to two oocytes or embryos, was separated by SDS-PAGE and transferred to nitrocellulose membranes (Schleicher and Schuell). Blots were incubated for 2 h at 20°C with anti-δV (1:2000) then with peroxidase-conjugated anti-rabbit IgG (1:2000, Chemicon) and developed with 3,3′-diaminobenzidine/H2O2. Reaction with the GST-δR fusion protein is also shown.
Multiple sequence alignment of HDm with related proteins, identified using the BLITZ algorithm. Accession numbers and references: Xenopus laevis HDm (X78454); human HD1 (U50079, Taunton et al., 1996); Drosophila melanogaster HD/Rpd3 (Y09258, De Rubertis et al., 1996) Caenorhabditis elegans hypothetical protein Yp82 (Q09440, Waterston et al., 1992); Saccharomyces cerevisiae Rpd3 (P32561, Vidal and Gaber, 1991); Bacillus subtilis acuC (L17309, Grundy et al., 1993); Staphylococcus xylosus acuC (X95439); Agmenellum quadruplicatum hypothetical protein Ygla (P28606, Wagner et al., 1993); Mycoplasma bullata Aph (D10463, residues 84–104 omitted for formating, Sakurada et al., 1996) Methanococcus jannaschii 0535 (U67502). Positions identical in six or more of the ten proteins are boxed black. The alignment was obtained using the PILEUP algorithm. Residues of MbAph suggested to be involved in zinc co-ordination ( Sakurada et al., 1996) are indicated (closed circles) as are Asp and His residues conserved within all ten proteins (open circles).