|
Fig. 1. Phylogenetic analysis of relaxin/insulin/IGF-type peptide precursors. The tree was generated in IQ-tree v 2.3.6 using the maximum likelihood method (SH-aLRT) and rooted using a clade comprising insulin-, IGF-, and bombyxin-type precursors. The clade that includes relaxin/Dilp8-type precursors and Dilp7-type precursors is shown without concatenation and the branches in this clade are highlighted according to the phyla that species belong to, following the color code shown in the key. Other clades are concatenated. The A. rubens RGP1 and RGP2 precursors are labeled with a green arrowhead and the A. cf. solaris RGP1 and RGP2 precursors are labeled with a yellow arrowhead. Bootstrap support values are color-coded with stars as indicated in the key. The accession numbers and sequences of the precursor proteins included in this phylogenetic tree are listed in Additional file 10: Dataset S3 and Additional file 11: Dataset S4, respectively
|
|
Fig. 2. Comparison of the exon/intron structure of genes encoding relaxin/Dilp8-type precursors and Dilp7-type precursors in selected bilaterian species. Schematic representations of the gene structures are shown, with protein-coding exons displayed as rectangles and introns shown as lines (with intron length stated above). The phase of introns is stated at the start of each intron above each diagram. Protein-coding exons are color-coded to show regions that encode the N-terminal signal peptide (blue), the B-chain (pink), monobasic or dibasic cleavage sites (green), the C-chain and other regions of the precursors (black), and the A-chain (orange). The residue numbers of the first and last amino acids are labeled underneath each diagram. A Genes encoding relaxin-type precursors in chordates, the starfish A. cf. solaris and A. rubens (RGP1 and RGP2), the annelid O. fusiformis and D. melanogaster (Dilp8) have a conserved phase 1 intron that interrupts the coding sequence for the C-chain region. B Genes encoding Dilp7-type precursors in the hemichordate S. kowalevskii, the starfish A. cf. solaris and A. rubens, the annelid O. fusiformis, the mollusk C. virensis, and the arthropod D. melanogaster (Dilp7) also have a conserved phase 1 intron that interrupts the coding sequence for the C-chain region, like that found in relaxin/Dilp8-type precursor genes. However, in addition, the Dilp7-type precursor genes have a second intron (phase 2) located between exons that encode the signal peptide and B-chain and this feature distinguishes Dilp7-type precursor genes from relaxin/Dilp8-type precursor genes. Accordingly, unlike in relaxin/Dilp8-type precursors where the B-chain is located immediately after the signal peptide, in Dilp7-type precursors the signal peptide and B-chain are separated by a polypeptide sequence (shown in black). The accession numbers and sequences of precursor proteins included in this figure are listed in Additional file 14: Dataset S7
|
|
Fig. 3. Phylogenetic analysis of leucine-rich repeat type G-protein coupled receptors (LRR-GPCRs). LRR-GPCRs are positioned in three distinct clades: type A includes glycoprotein hormone-type receptors (concatenated in this tree), type B includes bursicon-type receptors (concatenated in this tree), and type C includes relaxin-type receptors (type C1; not concatenated in this tree) and the GRL101-type receptors (type C2; concatenated in this tree). The type C1 receptors in the starfish A. cf. solaris include two receptors that are labeled with yellow arrows: Firstly, AsolRXFP/LGR3, which is an ortholog of vertebrate RXFP1/RXFP2-type relaxin receptors and Drosophila LGR3. Secondly, AsolLGR4, which is an ortholog of Drosophila LGR4. Likewise, the type C1 receptors include two receptors in the starfish A. rubens that are labeled with green arrows: Firstly, ArubRXFP/LGR3, which is an ortholog of vertebrate RXFP1/RXFP2-type relaxin receptors and Drosophila LGR3. Secondly, ArubLGR4, which is an ortholog of Drosophila LGR4. Based on this phylogenetic analysis, we identified AsolRXFP/LGR3 and ArubRXFP/LGR3 as candidate receptors for RGP1 and RGP2 in A. cf. solaris and A. rubens. The tree was generated in IQ-tree v 2.3.6 using the maximum likelihood method (SH-aLRT), with the clade comprising type A and type B receptor sequences selected as an outgroup to root the tree. The colored stars represent bootstrap support (1000 replicates, see key) and the colored backgrounds represent different taxonomic groups as also shown in the key. The accession numbers and sequences of the receptors included in this phylogenetic tree are listed in Additional file 12: Dataset S5 and Additional file 13: Dataset S6, respectively
|
|
Fig. 4. Comparison of the exon/intron structure of genes encoding RXFP/LGR3-type receptors and LGR4-type receptors in selected bilaterian taxa. Exons are shown as colored boxes, introns are shown as horizontal black and white lines, numbers above the colored boxes show the intron phase and numbers below the colored boxes show the positions of the first and last amino acid residues of each protein. The red rectangles show regions of the genes that encode the predicted seven transmembrane domains of each receptor and the more detailed diagrams on the right show the positions of introns with respect to the seven transmembrane domains. A phase 1 intron interrupting the coding sequence for TM domain 2 (orange) and a phase 1 intron interrupting the coding sequence for TM domain 5 (purple) are unique features of some genes encoding RXFP/LGR3-type receptors that distinguish these genes from those encoding LGR4-type receptors. Conversely, a phase 0 intron located after the exon that encodes the TM domain 7 (brown) is a unique feature of some genes encoding LGR4-type receptors that distinguish these genes from those encoding RXFP/LGR3-type receptors. The key shows the color coding for different regions/domains of the receptors. Species names are as follows: Hsap (H. sapiens), Asol (A. cf. solaris), Arub (A. rubens), Ofus (O. fusiformis), Dmel (D. melanogaster), Cvir (C. virginica). See detailed sequence analysis and positions of the introns in Additional file 15; Dataset S8
|
|
Fig. 5. A. cf. solaris RGP1 is a ligand for A. cf. solaris RXFP/LGR3 but not for A. cf. solaris LGR4. A AsolRGP1 induces a dose-dependent increase in CRE reporter gene activation in CHO-AsolRXFP/LGR3 cells. The potency of AsolRGP1 at AsolRXFP/LGR3 was ~ 16 nM (pEC50 = 7.80 ± 0.18). AsolRGP2 induced a small but significant response but no clear dose–response, while the negative control H2 relaxin induced no response. B, C Neither AsolRGP1 nor AsolRGP2 induced CRE reporter gene activation in HEK-293 T cells transfected with AsolLGR4 (long) or AsolLGR4 (short). The graphs show the pooled data from at least three assays performed in triplicate within each assay (n ≥ 3) expressed as mean values with error bars (S.E.M.). The raw data from all three experiments are shown in Additional file 17: Dataset S10
|
|
Fig. 6. A. rubens RGP1 and RGP2 act as a ligand for A. rubens RXFP/LGR3, but not for A. rubens LGR4. A, B ArubRGP1 and ArubRGP2 trigger dose-dependent luminescence in CHO-K1 cells transfected with ArubRXFP/LGR3 and the chimeric G-protein GqS5 and stably expressing the calcium-sensitive luminescent GFP-apoaequorin fusion protein G5A. The EC50 values for ArubRGP1 and ArubRGP2 are 2.5 × 10−8 M and 3.8 × 10−7 M, respectively. C, D ArubRGP1 and ArubRGP2 trigger dose-dependent luminescence in CHO-K1 cells transfected with a modified form of ArubRXFP/LGR3 containing a mammalian N-terminal signal peptide (from the prolactin precursor) and the chimeric G-protein GqS5 and stably expressing the calcium-sensitive luminescent GFP-apoaequorin fusion protein G5A. The EC50 values for ArubRGP1 and ArubRGP2 are 1.1 × 10−7 M and ~ 5.1 × 10.−7 M, respectively. E, F ArubRGP1 and ArubRGP2 do not trigger luminescence in CHO-K1 cells transfected with ArubLGR4 and the chimeric G-protein GqS5 and stably expressing the calcium-sensitive luminescent GFP-apoaequorin fusion protein G5A. G, H ArubRGP1 and ArubRGP2 do not trigger luminescence in CHO-K1 cells transfected with an empty pcDNA3.1 plasmid, the chimeric G-protein GqS5 and stably expressing the calcium-sensitive luminescent GFP-apoaequorin fusion protein G5A. The absence of ArubRGP1 and ArubRGP2 induced luminescence in E–H demonstrates that the ArubRGP1 and ArubRGP2 induced luminescence in A–D can be specifically attributed to transfection of cells with ArubRFXP/LGR3 or modified ArubRFXP/LGR3. The graphs show results from a single experiment with each point showing mean values (n = 3) with error bars (S.E.M.). The data shown are representative of three independent experiments and the raw data from all three experiments are shown in Additional file 19: Dataset S12
|
|
Fig. 7. Diagram showing the phylogenetic distribution of RXFP/LGR3-type proteins in the Bilateria, with representative species from phyla/sub-phyla illustrated. Taxa in which genes encoding RXFP/LGR3-type proteins have not been found are shown with an empty square. Taxa in which genes encoding RXFP/LGR3-type proteins have been found but without experimental identification of a cognate ligand(s) are shown with a filled square without an asterisk. Taxa in which genes encoding RXFP/LGR3-type proteins have been found and with experimental identification of a cognate ligand(s) are shown with a filled square and an asterisk together with the name of the ligand(s). The identification of an RXFP/LGR3-type protein as the receptor for RGP1 and RGP2 in starfish (A. rubens, A. cf solaris), as reported in this study and highlighted here in red, is the first experimental characterization of a relaxin-type signaling system in a deuterostome invertebrate. The findings of this study indicate that the evolutionary origin of relaxin-RXFP1,2/LGR3-type signaling can be traced back to the common ancestor of the Bilateria, but with subsequent loss in some taxa. Abbreviations: D, Deuterostomia; P, Protostomia; C, Chordata; A, Ambulacraria; S, Spiralia; E, Ecdysozoa. Images of representative animals from each phylum/sub-phylum were obtained from http://phylopic.org or were created by the authors and their collaborators
|