Yilmaz O et al. (2020), Unravelling the Complex Duplication History of ...

ECB-ART-49741

Cells 2020 Jul 10;97:. doi: 10.3390/cells9071663.

Show Gene links Show Anatomy links

Unravelling the Complex Duplication History of Deuterostome Glycerol Transporters.

Yilmaz O , Chauvigné F , Ferré A , Nilsen F , Fjelldal PG , Cerdà J , Finn RN .

???displayArticle.abstract???
Transmembrane glycerol transport is an ancient biophysical property that evolved in selected subfamilies of water channel (aquaporin) proteins. Here, we conducted broad level genome (>550) and transcriptome (>300) analyses to unravel the duplication history of the glycerol-transporting channels (glps) in Deuterostomia. We found that tandem duplication (TD) was the major mechanism of gene expansion in echinoderms and hemichordates, which, together with whole genome duplications (WGD) in the chordate lineage, continued to shape the genomic repertoires in craniates. Molecular phylogenies indicated that aqp3-like and aqp13-like channels were the probable stem subfamilies in craniates, with WGD generating aqp9 and aqp10 in gnathostomes but aqp7 arising through TD in Osteichthyes. We uncovered separate examples of gene translocations, gene conversion, and concerted evolution in humans, teleosts, and starfishes, with DNA transposons the likely drivers of gene rearrangements in paleotetraploid salmonids. Currently, gene copy numbers and BLAST are poor predictors of orthologous relationships due to asymmetric glp gene evolution in the different lineages. Such asymmetries can impact estimations of divergence times by millions of years. Experimental investigations of the salmonid channels demonstrated that approximately half of the 20 ancestral paralogs are functional, with neofunctionalization occurring at the transcriptional level rather than the protein transport properties. The combined findings resolve the origins and diversification of glps over >800 million years old and thus form the novel basis for proposing a pandeuterostome glp gene nomenclature.

???displayArticle.pubmedLink??? 32664262
???displayArticle.pmcLink??? PMC7408487
???displayArticle.link??? Cells

???attribute.lit??? ???displayArticles.show???

	Figure 1. Molecular phylogeny of deuterostome glycerol transporters. The tree is mid-point rooted and inferred via maximum likelihood (RAxML: generalized time-reversible (GTR) gamma model with 3000 bootstraps and search for the best-scoring maximum likelihood (ML) tree) and Bayesian analysis (15 million Markov chain Monte Carlo (MCMC) generations) of 315,034 nucleotide sites of 249 taxa partitioned by codon. Numbers within or external to each collapsed triangular branch refer to the number of taxa with Bayesian posterior probabilities annotated at each node. The scale bar indicates the expected rate of substitutions per site. Modes of gene duplication are shown on respective branches with additional duplications annotated for Teleostei and Homininae. Circled branches within the AQP10 channel cluster illustrate the origin of the polygenic aqp10 subfamily in actinopterygian fishes. (See Supplementary Figure S1 for the fully annotated tree including RAxML bootstrap support values).
	Figure 2. Synteny and molecular phylogeny of non-vertebrate chordate and Ambulacrarian glycerol-transporting channels (glps). (A) Syntenies are assembled from genome sequences and arranged according to the taxonomic position of the selected organisms. Gene coding direction is indicated by the pointed end of each gene symbol with linkages represented by solid lines between paralogs. A circular arrow indicates that the chromosome (Chr) or scaffold is flipped in relation to the European starfish super 8 scaffold. Orange arrows above gene symbols indicate tandem duplication. (B) Mid-point rooted Bayesian majority rule consensus tree (5 million MCMC generations) of 97,721 nucleotide sites of 116 taxa partitioned by codon. Bayesian posterior probabilities are annotated at each node with tandem “T” or unspecified “D” modes of duplication shown on selected nodes. The scale bar indicates the expected rate of substitutions per site. (See Supplementary Figure S3 for the fully annotated tree). (C) Mid-point rooted FastTree using GTR model optimized with gamma20 likelihood showing support values at each node. (D) Mid-point rooted maximum likelihood tree (PAUP: heuristic search optimized for parsimony GTR model with NST = 6; partitioned by codon) with RAxML support values shown at each node.
	Figure 3. Molecular phylogenies of actinopterygian glps. (A) Mid-point rooted Bayesian majority rule consensus tree of aqp3 channels (30 million MCMC generations) of 365,143 nucleotide sites of 382 taxa partitioned by codon. (See Supplementary Figure S5 for the fully annotated tree). (B) Bayesian majority rule consensus tree of aqp7 channels (15 million MCMC generations) of 225,618 nucleotide sites of 232 taxa partitioned by codon. The tree is rooted with reedfish (Erpetoichthys calabaricus) aqp7. (See Supplementary Figure S6 for the fully annotated tree). (C) Mid-point rooted Bayesian majority rule consensus tree of aqp9 channels (40 million MCMC generations) of 440,522 nucleotide sites of 460 taxa partitioned by codon. (See Supplementary Figure S7 for the fully annotated tree). (D) Mid-point rooted Bayesian majority rule consensus tree of aqp10 channels (50 million MCMC generations) of 457,471 nucleotide sites of 431 taxa partitioned by codon. (See Supplementary Figure S8 for the fully annotated tree). Bayesian posterior probabilities are indicated with color nodes as specified in the key. Six non-WGD lineage-dependent duplications are highlighted together with modes of gene duplication or loss in each tree.
	Figure 4. Evolution and expression of glps in paleotetraploid salmonids. (A) Schematic representation of the evolution, coding direction, and diversification of glps in a non-teleostean bichir and protocanthopterygian teleosts (Esociformes and Salmoniformes). Whole genome (R3 and R4) and tandem (T) duplication events are annotated on relevant nodes. (B) glp tissue expression profiles (RT-PCR) from four male and four female Atlantic salmon using elongation initiation factor 1a (eif1a) as a reference. Negative controls: cDNA without reverse transcriptase (RT) or no cDNA (NT).
	Figure 5. Functional characterization of Atlantic salmon Glps. Osmotic water permeability (Pf) and glycerol uptake (Pgly) of X. laevis oocytes injected with water (control) or cRNAs encoding the Atlantic salmon Glps. Data are box and whisker plots with the number of biologically independent oocytes indicated beside each plot. *** p < 0.001, statistically different (unpaired Student’s t-test), with respect to control oocytes. NS: not statistically significant with respect to control oocytes.
	Figure 6. A timeline for the evolution of vertebrate glps. (A) A schematic representation of historical events that occurred during the evolution of vertebrate glps. Data are arranged in accordance with the divergence of each lineage through geological time [61,77]. Ma: millions of years ago. The presence of single copy aqp3, -7, -9, -10, or -13 orthologs in Sarcopterygii is indicated with a “+” with extra copies or loss within each subfamily superscripted “+” (blue) or “-” (red). Four rounds (R1–R4) of whole genome duplication, together with three interchromosomal duplications/translocations and >20 tandem duplication and gene loss events, are shown. Numbers in square brackets refer to the number of taxa sourced for coding sequence (CDS) assembly. (B) Proposed aquaporin (aqp) nomenclature in non-sarcopterygian vertebrates, with the genes colored according to their origins. Numbers in parentheses refer to gene copy numbers. (C) Bayesian divergence time estimation of aqp10aa1 and -10aa2 duplication in Cyprinodontoidei. (D) Bayesian divergence time estimation of aqp3a1 and -3a2 duplication in African cichlids.
	Figure 7. The origin of deuterostome Glp subfamilies. A proposed scheme for the origin of glp gene subfamilies in Deuterostomia as a function of geological time. Modes of gene expansion are indicated as T: tandem duplication; or R1 and R2: whole genome duplication. Mz: Mesozoic; Cz: Cenozoic; and Ma: millions of years ago.

References [+] :

Abascal, Diversity and evolution of membrane intrinsic proteins. 2014, Pubmed