Repertoire of plant RING E3 ubiquitin ligases revisited: New groups counting gene families and single genes

Abstract

E3 ubiquitin ligases of the ubiquitin proteasome system (UPS) mediate recognition of substrates and later transfer the ubiquitin (Ub). They are the most expanded components of the system. The Really Interesting New Gene (RING) domain contains 40-60 residues that are highly represented among E3 ubiquitin ligases. The Arabidopsis thaliana E3 ubiquitin ligases with a RING finger primarily contain RING-HC or RING-H2 type domains or less frequently RING-v, RING-C2, RING-D, RING-S/T and RING-G type domains. Our previous work on three E3 ubiquitin ligase families with a RING-H2 type domain, ATL, BTL, and CTL, suggested that a phylogenetic distribution based on the RING domain allowed for the creation a catalog of known domains or unknown conserved motifs. This work provided a useful and comprehensive view of particular families of RING E3 ubiquitin ligases. We updated the annotation of A. thaliana RING proteins and surveyed RING proteins from 30 species across eukaryotes. Based on domain architecture profile of the A. thaliana proteins, we catalogued 4711 RING finger proteins into 107 groups, including 66 previously described gene families or single genes and 36 novel families or undescribed genes. Forty-four groups were specific to a plant lineage while 41 groups consisted of proteins found in all eukaryotic species. Our present study updates the current classification of plant RING finger proteins and reiterates the importance of these proteins in plant growth and adaptation.

Fig 1. Types of RING finger domains and number of RING finger subfamilies in A. thaliana.

A, Schematic representation of a canonical RING finger domain shows the eight residues involved in zinc coordination as cysteine (C) residues numbered from one to eight as well as the conventional number of amino acids represented by an x between them. Between the third and fourth zinc ligands, one to three residues were considered for all types of RING fingers. Between the fourth and fifth zinc ligands, two to three residues were considered for RING-H2, RING-HC, RING-D and RING-S/T, four to five residues for RING-C2 and seven residues for RING-v. Zinc ligand substitutions C>H, C>D, C>S, C>T or C>G were taken into account. For each RING-type, a consensus and the number of retrieved proteins are shown; data from Stone et al., 2005 was taken from S1 Table. B, Number of genes arranged in families or as a single gene is displayed. Described genes are shown in green and undescribed genes are shown in gray. Described families implied that there is a citation for at least one gene in the family (Table 2).

Fig 2. Phylogenetic distribution of A. thaliana RING finger proteins within seven RING domains.

The tree was generated with the RING finger domain from 508 RING finger polypeptide sequences of A. thaliana by the NJ (left) or ML (right) methods. RING-H2 and RING-HC domains are highlighted in gray and blue shades, respectively. Color codes highlighting RING-v, RING-C2, RING-D, RING-S/T, and RING-G are indicated. Misplaced sequences are denoted with arrowheads followed by the family code name. From proteins that contain two RING domains, the sequence from the domain that is present in all members of the family was used to build the tree (12 out the 38 RING sequences from FCH01, five out the six RING sequences from FCH07, one RING sequence from both LHC01 and LHC04; see S1 Table).

Fig 3. Phylogeny of A. thaliana RING-HC and RING-H2 finger domains.

Trees are displayed that were generated by the NJ method (A, C) and the ML method (B, D) from 258 A. thaliana RING-H2 (A, B) and 197 RING-HC domain sequences (C, D). A wide range of colors were used to distinguish each family. Misplaced sequence are denoted with arrowheads followed by the family name.

Fig 4. RING finger domains on A. thaliana groups.

Three regions of multiple consensus sequences alignment from the 41 RING-H2 or the 49 RING-HC groups that encompass the residues involved in zinc ligation are displayed. The circled numbers indicate the residues involved in zinc ligation. The amino acid length of three regions from the RING finger domain is shown: five residues encompassing ligands 1 and 2; 14 to 17 residues from the central region encompassing ligands 3, 4, 5 and 6; and five residues encompassing ligands 7 and 8. The shaded blocks indicate common residues to most RING fingers; J corresponds to any hydrophobic residue. The location of proline residues previously described in ATL and BTL families is enclosed by a rectangle. The symbol * denotes regions that deviate in length from the canonical sequence (FHC01, two extra residues; FHC02, two extra residues; FHC03, two extra residues in 50% of the proteins; FCH10, three proteins show the canonical distances and two proteins show an extra residue between the 4^th histidine and 5^th cysteine; FHC21, two extra residues). The symbol ** represents the alignment denoting a substitution of one of the zinc ligation residues in all members of the family (FHH11, C to D; FHC05, C to H). The underscore indicates a gap within the alignment.

Fig 5. GLD and GLD-like LOGO sequences in A. thaliana RING finger proteins.

An alignment of four sequence LOGOs that include the GLD and GLD-like conserved motifs obtained from RING-H2 or RING-D finger proteins (from top to bottom: LOGO #4, #11, #173, and #2 FD; see S3 Table). The location of sequence homology among LOGOs is highlighted by the blue square in the diagram representation. Except for Group LHH15, in all cases the LOGO sequences were located upstream and adjacent to the RING domain. The average distance between the LOGO and the start of the RING finger domain is presented in numbers of amino acids (aas). The symbol * depicts groups for which at least 50% of the members included sequence LOGO #4.

Fig 6. Number of retrieved RING-type proteins types across eukaryotes.

The phylogenetic relationship between thirteen viridiplantae (two green algae and eleven embryophyta), thirteen animal (six vertebrates, three basal animals, two basal vertebrates, two invertebrates), two fungal, and two protist genomes was based on data from the National Center of Biotechnology Information (NCBI) taxonomy server (http://www.ncbi.nlm.nih.gov/Taxonomy). The color code for a selected group of organisms is displayed at the branches. The identifiers for each RING-type are displayed in S1 Table and a list of retrieved genes in S4 Table. The total number of RING finger proteins is shown at the last column.

Fig 7. Number of retrieved RING proteins across eukaryotes categorized based on A. thaliana groups.

A heat map shows the number of proteins in each of the 107 RING groups. The fprequency of RING proteins is shown according to a pale green-red scale. The species tree is shown in Fig 6. A detailed distribution of RING identifiers within the groups is displayed in S4 Table.