The Evolutionary Landscape of Dbl-Like RhoGEF Families: Adapting Eukaryotic Cells to Environmental Signals

Abstract

The dynamics of cell morphology in eukaryotes is largely controlled by small GTPases of the Rho family. Rho GTPases are activated by guanine nucleotide exchange factors (RhoGEFs), of which diffuse B-cell lymphoma (Dbl)-like members form the largest family. Here, we surveyed Dbl-like sequences from 175 eukaryotic genomes and illuminate how the Dbl family evolved in all eukaryotic supergroups. By combining probabilistic phylogenetic approaches and functional domain analysis, we show that the human Dbl-like family is made of 71 members, structured into 20 subfamilies. The 71 members were already present in ancestral jawed vertebrates, but several members were subsequently lost in specific clades, up to 12% in birds. The jawed vertebrate repertoire was established from two rounds of duplications that occurred between tunicates, cyclostomes, and jawed vertebrates. Duplicated members showed distinct tissue distributions, conserved at least in Amniotes. All 20 subfamilies have members in Deuterostomes and Protostomes. Nineteen subfamilies are present in Porifera, the first phylum that diverged in Metazoa, 14 in Choanoflagellida and Filasterea, single-celled organisms closely related to Metazoa and three in Fungi, the sister clade to Metazoa. Other eukaryotic supergroups show an extraordinary variability of Dbl-like repertoires as a result of repeated and independent gain and loss events. Last, we observed that in Metazoa, the number of Dbl-like RhoGEFs varies in proportion of cell signaling complexity. Overall, our analysis supports the conclusion that Dbl-like RhoGEFs were present at the origin of eukaryotes and evolved as highly adaptive cell signaling mediators.

Keywords: Dbl; Rho GTPases; cell signaling; guanine nucleotide exchange factors.

Fig. 1.

—The 71 human Dbl-like RhoGEFs cluster into 20 structurally related subfamilies. (A) Phylogenetic cladogram of DH domains. The tree was deduced from multiple sequence alignment and processed by PhyML and MrBayes analysis. RhoGEFs subfamilies were delineated from node supports (PP: Posterior probability, BS: Bootstrap proportion). Only highly supported nodes (PP > 0.95 and BS > 75) are indicated (red circle). Numbers indicate subfamilies. (B) Subfamilies cluster members with similar functional domain organization. For each subfamily, the structural domains associated with the catalytic DH domain are indicated. All proteins are drawn at the same scale, except OBSCN. Domains typical of subfamilies are in red. BAR: Bin, Amphiphysin, Rvs; SH3: Src homology 3; CH: Calponin homology; PH: Pleckstrin homology; C1: N-terminal region of PKC; SH2: Src homology 2; UBQ: Ubiquitin; DEP: SEC14: Dishevelled, Egl10, Pleckstrin; SPEC: Spectrin homology; IG: Immunoglobulin-like; FN: Fibronectin-like; EH: Eps15 homology; C2: Ca²⁺ binding domain of PKC; PDZ: PSD95, Dlg1, Zo-1; RGS: Regulator of G protein Signaling; FYVE: Fab1, YOTB, ZK632.12, Vac1, EEA1; FERM: Four-point-one, Ezrin, Radixin, Moesin; FA: FERM Adjacent; BRCT: BRCA1 C-Terminus; RCC1: Regulator of Chromosome Condensation 1; MORN: Membrane Occupation and Recognition Nexus.

Fig. 2.

—Conservation of the human Dbl-like RhoGEF repertoire across Metazoa. Human RhoGEF orthologs were searched in genomes of species covering the major Metazoa clades, as indicated on the top. In most clades, three or more species were examined and orthology was deduced from reciprocal BLAST scores (“Vertebrates”, from mammals to bony fish) and by phylogenetic analysis (shark, lampreys, Ambulacraria, Cnidaria, Porifera, and nonmetazoan Filozoa). Members that were not found are indicated by an x box. Vertical bars show duplications that were deduced from phylogenetic analyses. Dashed vertical bars indicate duplications that cannot be precisely dated. The color code for Dbl-like subfamilies is the same as in figure 1A.

Fig. 3.

—Phylogenetic analyses of Dbl-like RhoGEFs in Metazoa. (A) Clustering of human and early vertebrate RhoGEFs. Hs: Homo sapiens; Cartilaginous fishes: Cm: Callorhinchus milii; Le: Leucoraja erinacea; Rt: Rhincodon typus; Lamprey: Pm: Petromyzon marinus. Lamprey members (names in orange and colored branches) at the roots of subfamilies or clusters are figured by an orange dot. No ortholog to ECT2L (blue dot) was found in Cm, Le, Rt or Pm genomes. (B) Clustering of human and prochordate RhoGEFs. Tunicates (names in green, blue branches): Ci: Ciona intestinalis; Hr: Halocynthia roretzi; Pm: Phallusia mammillata; Cephalochordate (names in orange, red branches) Bf: Branchiostoma floridae. (C) Clustering of human and early metazoan RhoGEFs. Cnidaria (green): Hv: Hydra vulgaris; Nv: Nematostella vectensis. Porifera (orange): Lc: Leucosolenia complicata. (D) Clustering of human and nonmetazoan Filozoa RhoGEFs. Choanoflagellida (green): Mb: Monosiga brevicollis; Sr: Salpingoeca rosetta. Filasterea (orange): Co: Capsaspora owczarzaki. DH domain amino acid sequences were aligned and analyzed by PhyML and MrBayes methods. Highly supported nodes are figured by a red circle. Names and accessions are listed in supplementary tables S1 and S3, Supplementary Material online. Color codes for RhoGEF subfamilies are as in figure 1A.

Fig. 4.

—Conservation of tissue specific expression of Dbl-like RhoGEFs in vertebrates. (A) Heatmap of RhoGEF mRNA expression in six tissues across nine vertebrate species. Orthologous RhoGEF values were extracted from global RNA-seq data. Mean centered log₁₀(RPKM) values were normalized and hierarchically clustered (Euclidian distance, complete method). Colors of RhoGEF names and sub-families correspond to those delineated in figure 1A. Red and green frames illustrate closely related members expressed in same or distinct tissues, respectively. Red and green dots indicate clusters with Pearson correlations of > 0.7 and >0.5, respectively. B: Brain; C: Cerebellum; K: Kidney. Hsap: Homo sapiens; Ptro: Pan troglodytes; Ggor: Gorilla gorilla; Mmul: Macaca mulatta; Pabe: Pongo abelii; Mmus: Mus musculus; Mdom: Monodelphis domesticus; Oana: Ornithorhynchus anatinus; Ggal: Gallus gallus. (B) Symmetrical heat map of Pearson correlations from RhoGEF gene mRNA expression (log RPKM values) for the six tissues and nine species examined.

Fig. 5.

—Sizes of Dbl-like families in Metazoa correlate with numbers of cell types. Dot plot showing the relationship between the numbers of cell types in different metazoan species (Schad et al. 2011; Hedges et al. 2004; Valentine et al. 1994) and the numbers of RhoGEFs (circles) and their target Rho GTPases (squares) (calculated from Metazoa data in supplementary tables S1 and S3, Supplementary Material online). Data from the following species were used: Primates: H. sapiens; Squamates: A. carolinensis; Amphibia: X. tropicalis; Fishes: Danio rerio; Agnatha: P. marinus; Tunicates: C. intestinalis; Arthropods: D. melanogaster, A. gambiae; Cnidaria: H. magnipapillata; Porifera: A. queenslandica; Choanoflagellida: M. brevocollis. S indicates the slopes of regression lines, whose confidence intervals are indicated by dashed lines.

Fig. 6.

—Summary of Dbl-like RhoGEF gain and loss events across eukaryote supergroups. A global view of the phylogeny of supergroups and taxa examined is presented, with the timeline of eukaryote emergence. Gain (in blue) and loss (in red) events that have built the repertoires of extant species are indicated. ¹: In Holometabola and Paraneoptera, and ²: In Chlamydomonales. The number of subfamilies, as defined in figure 1, is indicated for Metazoa. Green dots indicate nodes important for inferring the expansion of the Dbl family in Metazoa.

Fig. 7.

—The Dbl-like RhoGEF family in Fungi. (A) RhoGEFs were searched in genomes of species (in italics) distributed in the various Fungi phyla (in bold). RhoGEFs were classified according to their family (Dbl-like or DOCK) and their structural domain organization (see B). Ascomycota: Pezizomycotina: PE: Eurotiomycetes; PL: Leotiomycetes; PO: Orbiliomycetes; PP: Pezizomycetes; PS: Sordariomycetes. Ascomycota Saccharomycetales: SD: Debaryomycetaceae; SS: Saccharomycetaceae. T: Taphrinomycotina. Basisiomycota A: Agaricomycotina Agaricales; C: Agaricomycotina Corticiales; M: Pucciniomycotina Microbotryomycetes; P: Pucciniomycotina Pucciniomycetes; U: Ustilaginomycotina. (B) Eight types of RhoGEFs were identified in Fungi, based on the presence of functional domains. CH: Calponin homology, PB1: Phox/Bem1, DEP: Dishevelled/Egl10/Pleckstrin, CNH: Citron/Nik1 homology, BAR: Bin/Amphiphysin/Rvs, FYVE: Fab1/YOTB/ZK632.12/Vac1/EEA1, EH: Eps15 homology, SH3: Src homology 3, LRR: Leucine Rich Repeats. Three fungal RhoGEFs (DNMBP, FGD, ITSN) share similar functional domain organization with human RhoGEFs. (C) PhyML and MrBayes phylogenetic analysis of DH domains from Fungi of different clades, as indicated by the color code on the top left, excluding (left tree) or including (right tree) human DH sequences (in black). Nodes supported by posterior probabilities above 0.95 are indicated by red and yellow circles, with bootstrap BS values > 60 or >40, respectively. The domain organization of each cluster is shown.