A Likely Ancient Genome Duplication in the Speciose Reef-Building Coral Genus Acropora

摘要

class="head no_bottom_margin" id="sec1title">IntroductionReef-building corals contribute to tropical marine ecosystems that support innumerable marine organisms, but reefs are increasingly threatened because of recent increases in seawater temperatures, pollution, and other stressors (, ). The Acroporidae is a family of reef-building corals within the phylum Cnidaria, one of the basal phyla of the animal clade (, , ). Astreopora (Anthozoa: Acroporidae) is the sister genus in the acroporid lineage according to fossil records and molecular phylogenetic evidence (, , ). Importantly, Acropora (Anthozoa: Acroporidae), one of the most diverse genera of reef-building corals, including more than 150 species in the Indo-Pacific Ocean, is thought to have originated from Astreopora ∼60 mya with several species turnovers (, , , ). Investigating the evolutionary history of this group importantly contributes to our understanding of coral reef biodiversity and conservation. Hybridization among Acropora species has been observed in the wild (), and variable chromosome numbers have been determined in different Acropora lineages (). In addition, gene duplications have been shown in several Acropora gene families (, href="#bib23" rid="bib23" class=" bibr popnode">Hamada et al., 2013). Based on their unique lifestyle, variable chromosome numbers, and complicated reticular evolutionary history, Indo-Pacific Acropora likely originated via polyploidy (href="#bib20" rid="bib20" class=" bibr popnode">Gacesa et al., 2015, href="#bib23" rid="bib23" class=" bibr popnode">Hamada et al., 2013, href="#bib30" rid="bib30" class=" bibr popnode">Kenyon, 1997, href="#bib42" rid="bib42" class=" bibr popnode">Richards and Hobbs, 2015, href="#bib58" rid="bib58" class=" bibr popnode">Van Oppen et al., 2001, href="#bib62" rid="bib62" class=" bibr popnode">Vollmer and Palumbi, 2002, href="#bib68" rid="bib68" class=" bibr popnode">Willis et al., 2006). However, there is no direct molecular and genetic evidence to support this hypothesis.Ancient whole (large-scale)-gene/genome duplication (WGD), or paleopolyploidy, has shaped the genomes of vertebrates, green plants, and other organisms, and is usually regarded as an evolutionary landmark in the origin and diversification of organisms (href="#bib50" rid="bib50" class=" bibr popnode">Soltis et al., 2015, href="#bib56" rid="bib56" class=" bibr popnode">Van de Peer et al., 2009, href="#bib57" rid="bib57" class=" bibr popnode">Van De Peer et al., 2017) (href="#mmc1" rid="mmc1" class=" supplementary-material">Figure S1). Two separate WGD events have been documented in the common ancestors of vertebrates (two rounds of WGD) (href="#bib15" rid="bib15" class=" bibr popnode">Dehal and Boore, 2005) and another major WGD has been reported in the last common ancestor of teleost fish (href="#bib13" rid="bib13" class=" bibr popnode">Christoffels et al., 2004, href="#bib21" rid="bib21" class=" bibr popnode">Glasauer and Neuhauss, 2014). Meanwhile, living angiosperms share an ancient WGD event (href="#bib27" rid="bib27" class=" bibr popnode">Jiao et al., 2011, href="#bib55" rid="bib55" class=" bibr popnode">Tiley et al., 2016), and many other WGD events have been reported in major clades of angiosperms (href="#bib49" rid="bib49" class=" bibr popnode">Soltis et al., 2009, href="#bib59" rid="bib59" class=" bibr popnode">Vanneste et al., 2014, href="#bib66" rid="bib66" class=" bibr popnode">Wang et al., 2018). In addition, two rounds of WGDs in the vertebrates are suggested to have occurred during the Cambrian Period, and some WGDs in plants are believed to have occurred during Cretaceous-Tertiary (href="#bib48" rid="bib48" class=" bibr popnode">Smith et al., 2013, href="#bib57" rid="bib57" class=" bibr popnode">Van De Peer et al., 2017, href="#bib59" rid="bib59" class=" bibr popnode">Vanneste et al., 2014). Thus, WGD is regarded as an important evolutionary way to reduce the risk of extinction (href="#bib56" rid="bib56" class=" bibr popnode">Van de Peer et al., 2009, href="#bib57" rid="bib57" class=" bibr popnode">Van De Peer et al., 2017, href="#bib59" rid="bib59" class=" bibr popnode">Vanneste et al., 2014). However, the study of WGD in Cnidaria has received less attention (href="#bib29" rid="bib29" class=" bibr popnode">Kenny et al., 2016, href="#bib33" rid="bib33" class=" bibr popnode">Li et al., 2018, href="#bib44" rid="bib44" class=" bibr popnode">Schwager et al., 2017, href="#bib56" rid="bib56" class=" bibr popnode">Van de Peer et al., 2009, href="#bib57" rid="bib57" class=" bibr popnode">Van De Peer et al., 2017).Duplicated genes created by WGD have complex fates during diploidization (href="#bib46" rid="bib46" class=" bibr popnode">Sémon and Wolfe, 2007, href="#bib56" rid="bib56" class=" bibr popnode">Van de Peer et al., 2009). Usually, one of the duplicated genes is silenced or lost due to redundancy of gene functions, termed “nonfunctionalization.” However, retained duplicated genes provide important sources of biological complexity and evolutionary novelty due to subfunctionalization, neofunctionalization, and dosage effects (href="#bib14" rid="bib14" class=" bibr popnode">Conant et al., 2014, href="#bib27" rid="bib27" class=" bibr popnode">Jiao et al., 2011). Duplicated genes may develop complementary gene functions via subfunctionalization, evolve new functions through neofunctionalization, or are retained in complicated regulatory networks with different gene expressions due to dosage effects. For instance, duplicated MADS-Box genes are crucial for flower development and the origin of phenotypic novelty in plants (href="#bib56" rid="bib56" class=" bibr popnode">Van de Peer et al., 2009, href="#bib61" rid="bib61" class=" bibr popnode">Veron et al., 2006). Duplicated homeobox genes provide raw genetic material for vertebrate development (href="#bib11" rid="bib11" class=" bibr popnode">Canestro et al., 2013, href="#bib21" rid="bib21" class=" bibr popnode">Glasauer and Neuhauss, 2014). In addition, toxin diversification following gene duplications has been recognized as a mechanism to enhance adaptation in animals (href="#bib31" rid="bib31" class=" bibr popnode">Kondrashov, 2012, href="#bib32" rid="bib32" class=" bibr popnode">Kordiš and Gubenšek, 2000), especially in snake venoms (href="#bib25" rid="bib25" class=" bibr popnode">Hargreaves et al., 2014, href="#bib63" rid="bib63" class=" bibr popnode">Vonk et al., 2013). Interestingly, toxic proteins are involved in various important processes in corals, including prey capture, protection from predators, wound healing, etc (href="#bib2" rid="bib2" class=" bibr popnode">Armoza-Zvuloni et al., 2016, href="#bib4" rid="bib4" class=" bibr popnode">Ben-Ari et al., 2018), but it is still unclear how gene duplications of toxic proteins evolved in corals.Isozyme electrophoresis and restriction fragment length polymorphism were used to identify gene duplications in polyploids a few decades ago (href="#bib19" rid="bib19" class=" bibr popnode">Fürthauer et al., 1999, href="#bib53" rid="bib53" class=" bibr popnode">Stuber and Goodman, 1983). In the past 10 years, next-generation sequencing has generated a wealth of genomic data at vastly decreased cost and reduced efforts (href="#bib22" rid="bib22" class=" bibr popnode">Goodwin et al., 2016, href="#bib24" rid="bib24" class=" bibr popnode">Hardwick et al., 2017). Three main methods were developed to identify WGD, based on (1) analysis of the rate of synonymous substitutions per synonymous site (dS) of duplicated genes within a genome (dS-based method) (href="#bib8" rid="bib8" class=" bibr popnode">Blanc et al., 2003, href="#bib34" rid="bib34" class=" bibr popnode">Lynch and Conery, 2000, href="#bib59" rid="bib59" class=" bibr popnode">Vanneste et al., 2014, href="#bib72" rid="bib72" class=" bibr popnode">Mao, 2019), (2) phylogenetic analysis of gene families among multiple genomes (phylogenomic analysis) (href="#bib9" rid="bib9" class=" bibr popnode">Blomme et al., 2006, href="#bib27" rid="bib27" class=" bibr popnode">Jiao et al., 2011), and (3) synteny block identification compared with sister lineages without WGD (synteny analysis) (href="#bib10" rid="bib10" class=" bibr popnode">Bowers et al., 2003, href="#bib15" rid="bib15" class=" bibr popnode">Dehal and Boore, 2005, href="#bib71" rid="bib71" class=" bibr popnode">Zhang et al., 2017). The dS-based method and phylogenomic analysis only require gene family information, without genome assembly. However, too ancient WGD cannot be detected by the dS-based method, while gene tree uncertainty usually causes bias in the phylogenomic analysis. Both methods rely heavily on gene family estimation and clustering. Inaccurate gene predictions (gene models) and rough gene family cluster algorithms can easily fail to detect WGD using either method. In contrast, the synteny analysis relies heavily on genome assembly quality. Poor assembly quality can hide the WGD signals, and some genomes with huge rearrangements cannot be used to detect WGD using synteny block identification. Therefore, the most credible conclusions depend on complementary evidence from different methods (href="#bib12" rid="bib12" class=" bibr popnode">Chen and Birchler, 2013, href="#bib52" rid="bib52" class=" bibr popnode">Soltis and Soltis, 2012, href="#bib55" rid="bib55" class=" bibr popnode">Tiley et al., 2016).Here, using all three methods, we analyzed a genome of Astreopora (Astreopora sp1) as an outgroup, and five Acropora genomes (Acropora digitifera, Acropora gemmifera, Acropora subglabra, Acropora echinata, and Acropora tenuis) to address the following questions: (1) whether and when WGD occurred in Acropora, (2) what is the fate of duplicated genes in Acropora after the event, (3) what are the gene expression patterns of duplicated genes across five developmental stages in A. digitifera, and (4) what is the role of WGD in diversification of toxic proteins in Acropora (href="#mmc1" rid="mmc1" class=" supplementary-material">Figure S2).