The archaeo-eukaryotic GINS proteins and the archaeal primase catalytic subunit PriS share a common domain

Abstract

Primase and GINS are essential factors for chromosomal DNA replication in eukaryotic and archaeal cells. Here we describe a previously undetected relationship between the C-terminal domain of the catalytic subunit (PriS) of archaeal primase and the B-domains of the archaeo-eukaryotic GINS proteins in the form of a conserved structural domain comprising a three-stranded antiparallel beta-sheet adjacent to an alpha-helix and a two-stranded beta-sheet or hairpin. The presence of a shared domain in archaeal PriS and GINS proteins, the genes for which are often found adjacent on the chromosome, suggests simple mechanisms for the evolution of these proteins.

Reviewers: This article was reviewed by Zvi Kelman (nominated by Michael Galperin) and Kira Makarova.

Figure 1

Multiple sequence alignment of archaeal primase CTDs and archaeal and eukaryotic GINS B-domains. The multiple sequence alignment of PriS CTD and archaeal GINS B-domains was generated using Clustal X 2.0 [28,29] with default parameters. Sequences are denoted by their species names (left) and numeric Genbank Identifiers (GI numbers, right). The positions of the first and last residues of the aligned region of the corresponding protein are indicated. The colouring is based on the consensus shown underneath the alignment. Hydrophobic positions (ACFILMVWYH) are indicated by the letter h and shaded yellow when present in 80% of the sequences shown; small residues (ACDGNPSTV) are indicated by the letter s and shaded green. The secondary structure of the CTD of the S. solfataricus PriS protein (PDB code 1TZ2) is shown underneath the alignment (with H, E and L indicating α-helix, β-strand and loop regions respectively, with α-helices shown in red and β-strands in blue), as are the primary sequences and secondary structures of the B-domains of three of the four human GINS proteins: Sld5, Psf2 and Psf3 (derived from PDB file 2E9X). The alignment of the human GINS and S. solfataricus PriS CTD sequences was generated by pairwise structure comparison (1ZT2 versus 2E9X with default parameters) using DaliLite [27]. The inverted triangles above the Sld5 and Psf3 sequences indicate that amino acids have been omitted at these positions; the number of amino acids omitted is shown.

Figure 2

Primase CTD and GINS B-domain structures share a common fold. A. Structure of the C-terminal domain (CTD, amino acids 274-329) of the S. solfataricus PriS protein (PDB code 1ZT2, chain A) with the five conserved β-strands β1-β5 and helix α1 indicated. B. Structure of the N-terminal B-domain (amino acids 12-61) of the human GINS subunit Psf2 (PDB code 2E9X, chain B). C. Structure of the C-terminal B-domain (amino acids 167-223) of human Sld5 (PDB code 2E9X, chain D). Note the presence of an additional α-helix between β-strands β2 and β3. D. Structure of the N-terminal B-domain (amino acids 30-87) of the human Psf3 (PDB code 2E9X, chain C). Amino acids 48-56 are missing from the structure (indicated by broken line).

Figure 3

Model for acquisition of the CTD by PriS. Tandem duplication (labelled 1) of a Gins51 ORF found adjacent to a Prim domain ORF in the last common archaeo-eukaryotic ancestor is followed by deletion (labelled 2) of Gins51 A-domain sequences resulting in fusion of Prim domain and B-domain sequences and creation of an ORF encoding a recognisable PriS protein in the last common archaeal ancestor. Subsequent archaeal evolution has seen loss of Gins23 (labelled 3) in many species and loss of the CTD (labelled 4) from PriS in the Thermococcales, including Pyrococcus and Thermococcus species, and the Methanobacteriales. Co-localisation and co-expression of ORFs is also absent in many extant species [13].