Reconstructing ordinal relationships in the Demospongiae using mitochondrial genomic data

https://doi.org/10.1016/j.ympev.2008.05.014Get rights and content

Abstract

Class Demospongiae (phylum Porifera) encompasses most of sponges’ morphological and species diversity. It also represents one of the most challenging and understudied groups in animal phylogenetics, with many higher-level relationships still being unresolved. Among the unanswered questions are the most fundamental, including those about the monophyly of the Demospongiae and the relationships among the 14 recognized orders within the class. The lack of resolved phylogeny hampers progress in studies of demosponge biology, evolution and biodiversity and may interfere with the efficient conservation and economic use of this group. We addressed the question of demosponge relationships using mitochondrial genomic data. We assembled a mitochondrial genomic dataset comprising all orders of demosponges that includes 17 new and five previously published complete demosponge mitochondrial genomes. To test for the congruence between mtDNA-based and nuclear rRNA-based phylogenies, we also determined and analyzed 18S rRNA sequences for the same set of species. Our results provide strong support for five major clades within the Demospongiae: Homoscleromorpha = G0 (order Homosclerophorida), Keratosa = G1 (orders Dendroceratida, Dictyoceratida, and Verticillitida), Myxospongiae = G2 (orders Chondrosida, Halisarcida, and Verongida), marine Haplosclerida = G3 and the rest of demosponges = G4 (orders Agelasida, Astrophorida, Hadromerida, Halichondrida, Poecilosclerida, Spirophorida, and freshwater Haploscerida), and for the (G0((G1 + G2)(G3 + G4)) relationships among these clades. Conversely, mitochondrial genomic data do not support the monophylies of traditional subclasses Ceractinomorpha and Tetractinomorpha as well as several currently recognized orders of demosponges. Furthermore, we demonstrate that mitochondrial gene arrangements can also be informative for the inference of order-level demosponge relationships and propose a modified method for the analysis of gene order data that works well when translocation of tRNA genes are more frequent than other rearrangements.

Introduction

Class Demospongiae (Sollas, 1885) is the largest (>85% of species) and the most morphologically diverse in the phylum Porifera. It includes sponges with discrete cellular elements and skeletons made of siliceous spicules and/or spongin fibers (occasionally the skeleton is hypercalcified or is completely absent). Within the extant Demospongiae three subclasses have been traditionally recognized that encompass 14 orders, 88 families, 500 genera, and more than 8000 described extant species (Hooper and van Soest, 2002a, Hooper and van Soest, 2002b, van Soest et al., 2005). The relationships among the higher taxa of demosponges (suborders, orders, and above) have been the subject of extensive research since the second half of the 19th century but still remain mostly unresolved (Hooper and van Soest, 2002a, Hooper and van Soest, 2002b, Hooper et al., 2002, Boury-Esnault, 2006). In fact, it has been repeatedly stated that the Demospongiae is one of the few groups of animals, in which ordinal relationships are still unsettled (Minchin, 1900, Lévi, 1957, Boury-Esnault, 2006).

Traditionally, sponge systematics has been based mostly on skeletal morphology and spicule geometry and diversity. In particular, the shape and size of the large structural spicules (megascleres) and/or of the small reinforcing or packing spicules (microscleres) have been used as important taxonomic characters for demosponge classification (Sollas, 1888, Lendenfeld, 1889, Dendy, 1921, Hentschel, 1923, Topsent, 1928, De Laubenfels, 1936). In the 1950s Lévi added the mode of reproduction as an additional taxonomic character and used it to divide all demosponge taxa identified by earlier authors into two subclasses: Ceractinomorpha (viviparous taxa) and Tetractinomorpha (oviparous taxa) (Lévi, 1957). Subsequently, the order Homosclerophorida was isolated from Tetractinomorpha into its own subclass, Homoscleromorpha (Lévi, 1973). Lévi’s classification was elaborated by Bergquist (Bergquist, 1980) and Hartman (Hartman, 1982) and formed the backbone of demosponge systematics in the second half of the 20th century. At the beginning of this century, the Systema Porifera was published in which available morphological data on sponges was compiled and reanalyzed (Hooper and van Soest, 2002a, Hooper and van Soest, 2002b). This massive work (the volume on Demospongiae has 1101 pages and involves 29 authors) provides the modern-day authority on sponge systematics, but still leaves many questions of demosponge relationships open. In fact, the editors of Systema Porifera specifically emphasize that “the systematics of Porifera is still largely unresolved at higher levels of classification” and that “resolving the higher systematics of sponges is clearly beyond the capabilities of this present book or morphometric data alone” (Hooper and van Soest, 2002a, Hooper and van Soest, 2002b, Hooper et al., 2002).

The advent of molecular studies brought new ideas and new controversies to the field of sponge systematics. Since the early work by Kelly-Borges et al. (1991) and Lafay et al. (1992), nuclear ribosomal RNA genes have become the molecules of choice in demosponge phylogenetics, and both the number of studies utilizing these genes and the number of sampled species have been steadily increasing (reviewed in Boury-Esnault, 2006). The results of these studies challenge several established notions in sponge systematics. First, they do not support the monophyly of either the phylum Porifera (Borchiellini et al., 2001, Medina et al., 2001) or the class Demospongiae (Borchiellini et al., 2004, Nichols, 2005). Second, within the Demospongiae, they reject the monophyly of subclasses Ceractinomorpha and Tetractinomorpha, several orders (e.g., Halichondrida, Hadromerida, and Haplosclerida), families (e.g., Axinellidae) and genera (e.g., Axinella) (Borchiellini et al., 2004, Erpenbeck et al., 2005a, Nichols, 2005, Redmond et al., 2007). Finally, some unexpected associations have emerged from these studies, culminating in a new classification scheme proposed by Borchiellini and coworkers (2004) that groups all demosponges into four major clades: Keratosa (G1) (Dictyoceratida + Dendroceratida), Myxospongiae (G2) (Chondrosida, Halisarcida, and Verongida), Marine Haplosclerida (G3), and all the remaining groups (G4). However, the relationships among these clades, as well as many relationships within them, remain unresolved or poorly supported (see also Redmond et al., 2007). To make the matter worse, there is little congruence between the results of rRNA-based studies and those that used alternative molecular markers: mitochondrial cox1 (Erpenbeck et al., 2002, Nichols, 2005), nuclear Hsp70 (Borchiellini et al., 1998), tubulin (Schroder et al., 2003), and EF-1 (Erpenbeck et al., 2005b), although a recent study of animal relationships based on concatenated dataset of seven nuclear genes that included 10 demosponge taxa showed some support for the G1 + G2 vs. G3 + G4 division within the demosponges (Sperling et al., 2007).

We addressed the question of demosponge relationships using mitochondrial genomic data. There are several reasons to favor such an approach for demosponge phylogenetics. First, demosponge mtDNA has several distinct features, including extra genes, a minimally modified genetic code, and Metazoa-specific synapomorphies (Lavrov, 2007) that can help to detect contamination by foreign DNA—a significant problem in any molecular work on sponges. Second, although mtDNA is often associated with high rates of sequence evolution as is typical in mammalian species, this is not the case in demosponges. Our previous study has shown that the rate of mitochondrial sequence evolution in demosponges is several-fold lower than in mammals and is comparable to the rates of evolution in non-animal mtDNA (Lavrov et al., 2005). Third, a large amount of sequence data (∼19 kbp/genome) and three different types of genes (rRNA, tRNA, and protein-coding) minimize the sampling error in phylogenetic analysis. Fourth, most of the mitochondrial genes are clearly orthologous among all eukaryotes, the only exception being some tRNA genes, which evolved by gene recruitment (Lavrov and Lang, 2005b). Fifth, horizontal gene transfer is extremely rare in the mtDNA of animals and other eukaryotes, with a possible exception of plants (Bergthorsson et al., 2003). A recently proposed example of horizontal intron transfer in the demosponge Tetilla sp. (Rot et al., 2006) may actually represent a rare case of intron retention (Wang and Lavrov, in press). Sixth, in addition to sequence data, some rare genome changes can be used for phylogenetic inference (Rokas and Holland, 2000), in particular mitochondrial gene rearrangements (Boore and Brown, 1998, Lavrov and Lang, 2005a). Our previous analysis revealed that some gene arrangements in demosponge mtDNA are conserved from the time of their divergence with other animals, but also that multiple rearrangements occurred among demosponges (Lavrov and Lang, 2005a). Finally, from a technical point of view, the gene-rich nature of animal mtDNA, its stable and nearly identical gene content among most animal groups, and the presence of multiple mtDNA copies per cell make mitochondrial genomic research both efficient and cost-effective.

For this project we determined the complete mitochondrial genome sequences of 17 demosponges (Table 1) and analyzed them with five previously published (Lavrov and Lang, 2005b, Lavrov et al., 2005, Erpenbeck et al., 2007a, Wang and Lavrov, 2007). The resulting dataset represents all recognized orders within the class Demospongiae, and phylogenetic analyses based on this dataset provide a well supported phylogenetic hypothesis for the class Demospongiae.

Section snippets

Taxon sampling

Species used in this study were chosen to cover all recognized orders of demosponges (Table 1). Most of the specimens were collected off the Florida Keys and an effort has been made to use more common, and/or better-studied species from each group. Amphimedon compressa Duchassaing & Michelotti, 1864, Ectyoplasia ferox (Duchassaing & Michelotti, 1864), Hippospongia lachne de Laubenfels, 1936, Plakortis angulospiculatus (Carter, 1882), and Xestospongia muta (Schmidt, 1870) were collected by DL

Mitochondrial genomes of demosponges

We determined the complete mitochondrial genome sequences of 17 species of demosponges and by doing this completed our sampling of all 14 currently recognized orders in the Demospongiae (Table 1). The majority of the new mitochondrial genomes fit well with our previous description of demosponge mtDNA (Lavrov et al., 2005). These genomes are between 18 and 20 kbp in size, about 70% AT-rich and contain 44 or 45 genes, including genes for subunit 9 of ATPase synthase, tRNACAUIle, and tRNACAUIle,

Implications for demosponge systematics and biology

The results of our study based on 22 complete mtDNA sequences reinforces the notion that the current higher-level classification of demosponges does not reflect their phylogenetic relationships and thus is in need of revision. First, in congruence with previous studies (van Soest, 1991, Boury-Esnault, 2006), our results show that subclasses Tetractinomorpha and Ceractinomorpha are not monophyletic and should be abandoned. Given that these groups were originally defined based on the mode of

Acknowledgments

We thank Alexander Ereskovsky, Franz Lang, Sally Leys, Antonio Solé-Cava, Robert Thacker, and Gert Wörheide for samples contributed to this project; Shirley Pomponi for help with species collection and identification; Natalia Frishman, Megan Rasmuson, Brandi Sigmon and Zhiyong Shao, for assistance in laboratory work. We are grateful to Nicole Boury-Esnault and Rob van Soest for insightful discussions and Frank Anderson, Karri Haen, and five anonymous reviewers for their help with earlier

References (107)

  • L. Medlin et al.

    The characterization of enzymatically amplified eukaryotic 16S-like rRNA-coding regions

    Gene

    (1988)
  • S.A. Nichols

    An evaluation of support for order-level monophyly and interrelationships within the class Demospongiae using partial data from the large subunit rDNA and cytochrome oxidase subunit I

    Mol. Phylogenet.

    (2005)
  • N.E. Redmond et al.

    Reassessment of the classification of the order Haplosclerida (class Demospongiae, phylum Porifera) using 18S rRNA gene sequence data

    Mol. Phylogenet. Evol.

    (2007)
  • D.F. Robinson et al.

    Comparison of phylogenetic trees

    Math. Biosci.

    (1981)
  • A. Rokas et al.

    Rare genomic changes as a tool for phylogenetics

    Trends Ecol. Evol.

    (2000)
  • Z. Shao et al.

    Mitochondrial genome of the moon jelly Aurelia aurita (Cnidaria, Scyphozoa): a linear DNA molecule encoding a putative DNA-dependent DNA polymerase

    Gene

    (2006)
  • F. Abascal et al.

    MtArt: a new model of amino acid replacement for Arthropoda

    Mol. Biol. Evol.

    (2006)
  • J. Adachi et al.

    Plastid genome phylogeny and a model of amino acid substitution for proteins encoded by chloroplast DNA

    J. Mol. Evol.

    (2000)
  • J.C. Avise et al.

    Mitochondrial DNA differentiation during the speciation process in Peromyscus

    Mol. Biol. Evol.

    (1983)
  • J.C. Avise

    Molecular Markers, Natural History, and Evolution

    (2004)
  • E. Bazin et al.

    Population size does not influence mitochondrial genetic diversity in animals

    Science

    (2006)
  • D.A. Benson et al.

    GenBank

    Nucleic Acids Res.

    (2008)
  • P.R. Bergquist

    The ordinal and subclass classification of the Demospongiae (Porifera), appraisal of the present arrangement, and proposal of a new order

    NZ J. Zool.

    (1980)
  • P.R. Bergquist

    Sponges

    (1978)
  • U. Bergthorsson et al.

    Widespread horizontal transfer of mitochondrial genes in flowering plants

    Nature

    (2003)
  • J.L. Boore et al.

    Deducing the pattern of arthropod phylogeny from mitochondrial DNA rearrangements

    Nature

    (1995)
  • C. Borchiellini et al.

    Phylogenetic analysis of the Hsp70 sequences reveals the monophyly of Metazoa and specific phylogenetic relationships between animals and fungi

    Mol. Biol. Evol.

    (1998)
  • C. Borchiellini et al.

    Sponge paraphyly and the origin of Metazoa

    J. Evol. Biol.

    (2001)
  • G. Bourque et al.

    Genome-scale evolution: reconstructing gene orders in the ancestral species

    Genome Res.

    (2002)
  • N. Boury-Esnault et al.

    Larval development in the Homoscleromorpha (Porifera, Demospongiae)

    Invertebr. Biol.

    (2003)
  • N. Boury-Esnault

    Systematics and evolution of Demospongiae

    Can. J. Zool.

    (2006)
  • G. Burger et al.

    Sequencing complete mitochondrial and plastid genomes

    Nat. Protoc.

    (2007)
  • B.L. Cantarel et al.

    Exploring the relationship between sequence similarity and accurate phylogenetic trees

    Mol. Biol. Evol.

    (2006)
  • J. Castresana

    Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis

    Mol. Biol. Evol.

    (2000)
  • C. Chombard et al.

    Good congruence between morphology and molecular phylogeny of Hadromerida, or how to bother sponge taxonomists

    Mem. Queensl. Mus.

    (1999)
  • M.W. De Laubenfels

    A discussion of the sponge fauna of the Dry Tortugas in particular and the West Indies in general

    (1936)
  • A. Dendy

    The tetraxonid sponge spicule: a study in evolution

    Acta Zool.

    (1921)
  • C.W. Dunn et al.

    Broad phylogenomic sampling improves resolution of the animal tree of life

    Nature

    (2008)
  • S. Duran et al.

    Low levels of genetic variation in mtDNA sequences over the western Mediterranean and Atlantic range of the sponge Crambe crambe (Poecilosclerida)

    Mar. Biol.

    (2004)
  • D. Erpenbeck et al.

    Unravelling host and symbiont phylogenies of halichondrid sponges (Demospongiae, Porifera) using a mitochondrial marker

    Mar. Biol.

    (2002)
  • D. Erpenbeck et al.

    Mitochondrial diversity of early-branching Metazoa is revealed by the complete mt genome of a haplosclerid demosponge

    Mol. Biol. Evol.

    (2007)
  • D. Erpenbeck et al.

    Towards a DNA taxonomy of Caribbean demosponges: a gene tree reconstructed from partial mitochondrial CO1 gene sequences supports previous rDNA phylogenies and provides a new perspective on the systematics of Demospongiae

    J. Mar. Biolog. Assoc. UK

    (2007)
  • D. Erpenbeck et al.

    Implications from a 28S rRNA gene fragment for the phylogenetic relationships of halichondrid sponges (Porifera: Demospongiae)

    J. Zoolog. Syst. Evol. Res.

    (2005)
  • D. Erpenbeck et al.

    Identification, characterization and phylogenetic signal of an elongation factor-1 alpha fragment in demosponges (Metazoa, Porifera, Demospongiae)

    Zool. Scr.

    (2005)
  • B. Ewing et al.

    Base-calling of automated sequencer traces using phred. II. Error probabilities

    Genome Res.

    (1998)
  • B. Ewing et al.

    Base-calling of automated sequencer traces using phred. I. Accuracy assessment

    Genome Res.

    (1998)
  • J. Felsenstein

    Cases in which parsimony or compatibility methods will be positively misleading

    Syst. Zool.

    (1978)
  • Felsenstein, J., 2005. PHYLIP (Phylogeny Inference Package). Version 3.65. Distributed by the author. Department of...
  • A. Gelman et al.

    Inference from iterative simulation using multiple sequences

    Stat. Sci.

    (1992)
  • J.H. Gillespie

    Genetic drift in an infinite population. The pseudohitchhiking model

    Genetics

    (2000)
  • Cited by (106)

    View all citing articles on Scopus
    View full text