Molecular Biology and Evolution

Takahashi, Hirokazu; Kamiya, Akiko; Ishiguro, Akira; Suzuki, Atsushi C.; Saitou, Naruya; Toyoda, Atsushi; Aruga, Jun

doi: 10.1093/molbev/msm228pmid: 17940209

Abstract Msx (/msh) family genes encode homeodomain (HD) proteins that control ontogeny in many animal species. We compared the structures of Msx genes from a wide range of Metazoa (Porifera, Cnidaria, Nematoda, Arthropoda, Tardigrada, Platyhelminthes, Mollusca, Brachiopoda, Annelida, Echiura, Echinodermata, Hemichordata, and Chordata) to gain an understanding of the role of these genes in phylogeny. Exon–intron boundary analysis suggested that the position of the intron located N-terminally to the HDs was widely conserved in all the genes examined, including those of cnidarians. Amino acid (aa) sequence comparison revealed 3 new evolutionarily conserved domains, as well as very strong conservation of the HDs. Two of the three domains were associated with Groucho-like protein binding in both a vertebrate and a cnidarian Msx homolog, suggesting that the interaction between Groucho-like proteins and Msx proteins was established in eumetazoan ancestors. Pairwise comparison among the collected HDs and their C-flanking aa sequences revealed that the degree of sequence conservation varied depending on the animal taxa from which the sequences were derived. Highly conserved Msx genes were identified in the Vertebrata, Cephalochordata, Hemichordata, Echinodermata, Mollusca, Brachiopoda, and Anthozoa. The wide distribution of the conserved sequences in the animal phylogenetic tree suggested that metazoan ancestors had already acquired a set of conserved domains of the current Msx family genes. Interestingly, although strongly conserved sequences were recovered from the Vertebrata, Cephalochordata, and Anthozoa, the sequences from the Urochordata and Hydrozoa showed weak conservation. Because the Vertebrata–Cephalochordata–Urochordata and Anthozoa–Hydrozoa represent sister groups in the Chordata and Cnidaria, respectively, Msx sequence diversification may have occurred differentially in the course of evolution. We speculate that selective loss of the conserved domains in Msx family proteins contributed to the diversification of animal body organization. Msx, homeodomain, protein conserved domain, Groucho, protein–protein interaction, exon–intron boundary Introduction Msx family proteins are critical regulators of metazoan ontogeny, as has been revealed in many studies (reviewed in Davidson 1995; Bendall and Abate-Shen 2000; Ramos and Robert 2005). In the Ecdysozoa, Drosophila muscle segment homeobox (msh) has a role in the regional specification of neuroectoderm and muscle progenitors (Isshiki et al. 1997; Nose et al. 1998), and Caenorhabditis Vab-15 has a role in touch receptor neuron development (Du and Chalfie 2001). (Here, we use the term “Msx family” to denote a gene family orthologous to those encoding vertebrate Msx, including msh and vab-15.) In the Echinodermata, Strongylocentrotus and Heliocidaris Msx homologs may control gastrulation and skeletal patterning (Tan et al. 1998; Wilson et al. 2005). In the Vertebrata, Msx1 or Msx2 plays roles in dorsoventral patterning of embryos, regionalization of the neural tube, establishment of neural crest tissue, and the genesis of various types of organs, including the palate, tooth, skull, middle ear, jaw, hair, mammary gland, and limbs, both in mammals (Satokata and Maas 1994; Bendall and Abate-Shen 2000; Satokata et al. 2000; Lallemand et al. 2005; Ramos and Robert 2005) and amphibians (Suzuki et al. 1997; Monsoro-Burq et al. 2005; Khadka et al. 2006). Mutations in human MSX1 are responsible for selective tooth agenesis (STA) (Vastardis et al. 1996), cleft palate, cleft lip (van den Boogaard et al. 2000), and nail dysplasia (Jumlongras et al. 2001), and those in MSX2 are responsible for parietal foramina (PFM) (Wuyts et al. 2000) and craniosynostosis (CSO) (Wilkie et al. 2000). These facts enlighten us about not only the importance of Msx family genes in each animal but also the similarities and differences of their roles in different animal groups. In phylogenetic terms, the structural features and/or expression profiles of Msx homologs have been described in insects (Walldorf et al. 1989), ascidians (Holland 1991), sponges (Seimiya et al. 1994), and leeches (Master et al. 1996). Some researchers have noticed that the role of the Msx family in neuroectodermal patterning is similar in the fruit fly and vertebrates (Isshiki et al. 1997; Arendt and Nubler-Jung 1999); Msx family genes specify lateral longitudinal columns of neuroectoderm in both types of animals, raising the possibility that the bilaterian ancestors had already used Msx family genes in establishing their nervous systems. The same expression pattern is essentially conserved in another insect, Tribolium castaneum (Wheeler et al. 2005). However, this hypothesis is still uncertain and awaits verification by examination in other animal species. In a previous study, we performed a molecular phylogenetic analysis of wide-ranging groups of animals to determine the role of Zic family zinc finger proteins in evolution. We compared both the amino acid (aa) sequence and the exon–intron organization of many interspecies orthologs from major metazoan phyla (Aruga et al. 2006). The study revealed novel evolutionarily conserved domains and gave us a broad understanding of the processes of protein evolution and the traits involved in evolutionary change. We therefore applied the same strategy to Msx family proteins. Some earlier works had pointed out a promising direction for an analysis of the molecular phylogeny of Msx family genes. A pioneering study by Holland (1991) compared vertebrate, ascidian (Ciona), and Drosophila Msx homologs. The relationship between the Msx genes of zebrafish and other vertebrates were investigated in terms of aa sequences and expression patterns (Ekker et al. 1997) and chromosomal synteny (Postlethwait 2006). Perry et al. (2006) focused on phylogenetic comparisons among primate Msx genes. However, these studies are not sufficient for our purpose because they only dealt with the limited evolutionary processes. We report here the molecular phylogeny of Msx family genes, as determined by a comparison of the Msx sequences from 13 animal phyla. We compared the conserved domains of Msx proteins and the exon–intron organization of Msx genes, and we performed a functional analysis of some of the conserved domains. The results indicated that the eumetazoan ancestor already possessed Msx protein with a full set of conserved domains; these domains have diverged strongly in some animal groups. Materials and Methods Database Search The similarity search against the current databases was done with the Blast algorithm (BlastP, TBlastN, Altschul et al. 1990, 1997) against public databases (http://www.ncbi.nlm.nih.gov/blast/, http://www.ensembl.org/index.html). Presence or absence was estimated by a database search with TBlastN algorithm against selected genome sequence databases that used key Mus musculus Msx-1 or Nematostella Msx sequences. A decision on authentic homology was based on reverse Blast analysis in which the sequence that showed the highest score in a certain species was used as a key sequence for the homology search by using the BlastX algorithm against National Center for Biotechnology Information (NCBI) eukaryotic databases. If the sequence that showed the highest Blast score among the identified sequences contained the Msx proteins, this labeled a member of Msx family (table 1). Table 1 Animals Used in This Study Phylum Subphylum Class Species Abbreviation Genome cDNA Msx-Related Genes (Synonym, Accession Number) Chordata     Vertebrata         Mammalia Mus musculus Mm DB DB Msx1 (AAH16426), Msx2 (Q03358), Msx3 (P70354)         Aves Gallus gallus Gg DB DB Msx1 (Hox7, P50223), Msx2 (Hox8, P28362)         Amphibia Xenopus laevis Xl DB Msx1 (AAH81101)         Osteichthyes Danio rerio Dr DB DB Msh-A (Q03357), Msh-B (Q03356), Msh-C (Q01703), Msh-D (Q01704), MsxE (AAB03273)     Cephalochordata         Leptocardia Branchiostoma floridae Bf DB Msx (CAA10201)     Urochordata         Ascidiacea Ciona intestinalis Ci DB DB CAD56691 Molgula oculata Mo DB AAA87223         Appendicularia Oikopleura dioica Od DB AAW24005 Hemichordata         Enteropneusta Saccoglossus kowalevskii Sk DB Msx (ABD97280) Echinodermata         Echinoidea Strongylocentrotus purpuratus Sp DB DB SpMsx (AAB97688) Heliocidaris erythrogramma Her DB Msx (AAY86177) Heliocidaris tuberculata Ht DB Msx (AAY86178)         Asteroidea Asterina pectinifera Ap Fosmid PCR MsxA [Ast-1 (AB302953)]a, MsxB [Ast-3 (AB302954)]a Arthropoda     Chelicerata Limulus polyphemus Lp PCR MsxA (AB302959)a Pandinus imperator Pi PCR MsxA (AB302960)a, MsxB (AB302961)a     Crustacea Artemia franciscana Af PCR Msx (AB302962)a     Insecta Tribolium castaneum Tc DB Muscle segment homeoprotein (AAW21975) Drosophila melanogaster Dm DB DB msh (Q03372) Anopheles gambiae Ag DB DB EAA08817 Apis mellifera Ame DB Msx (AB362784)b Tardigrada Milnesium tardigradum Mt PCR Msx (AB302966)a Nematoda Caenorhabditis elegans Ce DB DB vab-15 (Q09604) Caenorhabditis briggsae Cb DB CAE58718 Mollusca         Cephalopoda Loligo bleekeri Lb PCR Msx (AB302963)a Octopus ocellatus Oo PCR Msx (AB302964)a         Bivalvia Corbicula fluminea Cf Fosmid PCR Msx [Cor-2/Cor-3 (AB302955)]a Brachiopoda         Inarticulata Lingula anatina La PCR Msx (AB302965)a Annelida         Polychaeta Platynereis dumerilii Pd DB Msx (CAJ38810)         Oligochaeta Tubifex tubifex Tt Fosmid PCR MsxA [Tt-1 (AB302956)]a, MsxB [Tt-3 (AB302957)]a         Hirudinida Helobdella sp. Hel DB Msx (AAB37254) Echiura Urechis caupo Uc PCR Msx (AB302967)a Platyhelminthes         Tubellaria Schmidtea mediterranea Sm DB Msx (AB362785)b Thysanozoon sp. Thy PCR Msx (AB302968)a Cnidaria         Hydrozoa Hydra viridis Hvi DB CAA45912 Hydra vulgaris Hvu DB CAB88390 Scolionema suvaense Ss Fosmid PCR Msx [Sco-1/Sco-6 (AB302958)]a Podocoryne carnea Pc DB AAX58756         Anthozoa Nematostella vectensis Nv PCR Msx (AB362783)a Anemonia erythraea Ae PCR Msx (AB302969)a Acropora millepora Ami DB msh3 (ABK41269) Porifera         Demospongiae Ephydatia fluviatilis Ef DB prox3 (AAA20151) Phylum Subphylum Class Species Abbreviation Genome cDNA Msx-Related Genes (Synonym, Accession Number) Chordata     Vertebrata         Mammalia Mus musculus Mm DB DB Msx1 (AAH16426), Msx2 (Q03358), Msx3 (P70354)         Aves Gallus gallus Gg DB DB Msx1 (Hox7, P50223), Msx2 (Hox8, P28362)         Amphibia Xenopus laevis Xl DB Msx1 (AAH81101)         Osteichthyes Danio rerio Dr DB DB Msh-A (Q03357), Msh-B (Q03356), Msh-C (Q01703), Msh-D (Q01704), MsxE (AAB03273)     Cephalochordata         Leptocardia Branchiostoma floridae Bf DB Msx (CAA10201)     Urochordata         Ascidiacea Ciona intestinalis Ci DB DB CAD56691 Molgula oculata Mo DB AAA87223         Appendicularia Oikopleura dioica Od DB AAW24005 Hemichordata         Enteropneusta Saccoglossus kowalevskii Sk DB Msx (ABD97280) Echinodermata         Echinoidea Strongylocentrotus purpuratus Sp DB DB SpMsx (AAB97688) Heliocidaris erythrogramma Her DB Msx (AAY86177) Heliocidaris tuberculata Ht DB Msx (AAY86178)         Asteroidea Asterina pectinifera Ap Fosmid PCR MsxA [Ast-1 (AB302953)]a, MsxB [Ast-3 (AB302954)]a Arthropoda     Chelicerata Limulus polyphemus Lp PCR MsxA (AB302959)a Pandinus imperator Pi PCR MsxA (AB302960)a, MsxB (AB302961)a     Crustacea Artemia franciscana Af PCR Msx (AB302962)a     Insecta Tribolium castaneum Tc DB Muscle segment homeoprotein (AAW21975) Drosophila melanogaster Dm DB DB msh (Q03372) Anopheles gambiae Ag DB DB EAA08817 Apis mellifera Ame DB Msx (AB362784)b Tardigrada Milnesium tardigradum Mt PCR Msx (AB302966)a Nematoda Caenorhabditis elegans Ce DB DB vab-15 (Q09604) Caenorhabditis briggsae Cb DB CAE58718 Mollusca         Cephalopoda Loligo bleekeri Lb PCR Msx (AB302963)a Octopus ocellatus Oo PCR Msx (AB302964)a         Bivalvia Corbicula fluminea Cf Fosmid PCR Msx [Cor-2/Cor-3 (AB302955)]a Brachiopoda         Inarticulata Lingula anatina La PCR Msx (AB302965)a Annelida         Polychaeta Platynereis dumerilii Pd DB Msx (CAJ38810)         Oligochaeta Tubifex tubifex Tt Fosmid PCR MsxA [Tt-1 (AB302956)]a, MsxB [Tt-3 (AB302957)]a         Hirudinida Helobdella sp. Hel DB Msx (AAB37254) Echiura Urechis caupo Uc PCR Msx (AB302967)a Platyhelminthes         Tubellaria Schmidtea mediterranea Sm DB Msx (AB362785)b Thysanozoon sp. Thy PCR Msx (AB302968)a Cnidaria         Hydrozoa Hydra viridis Hvi DB CAA45912 Hydra vulgaris Hvu DB CAB88390 Scolionema suvaense Ss Fosmid PCR Msx [Sco-1/Sco-6 (AB302958)]a Podocoryne carnea Pc DB AAX58756         Anthozoa Nematostella vectensis Nv PCR Msx (AB362783)a Anemonia erythraea Ae PCR Msx (AB302969)a Acropora millepora Ami DB msh3 (ABK41269) Porifera         Demospongiae Ephydatia fluviatilis Ef DB prox3 (AAA20151) NOTE.—DB, sequence collected from NCBI databases; fosmid, isolated as fosmid genomic clones in this study; PCR, cloned by PCR amplification of the cDNA. a Sequence newly determined in this study. b Sequences edited from the publicized draft sequences (http://www.ncbi.nlm.nih.gov/BLAST/tracemb.shtml). Open in new tab Table 1 Animals Used in This Study Phylum Subphylum Class Species Abbreviation Genome cDNA Msx-Related Genes (Synonym, Accession Number) Chordata     Vertebrata         Mammalia Mus musculus Mm DB DB Msx1 (AAH16426), Msx2 (Q03358), Msx3 (P70354)         Aves Gallus gallus Gg DB DB Msx1 (Hox7, P50223), Msx2 (Hox8, P28362)         Amphibia Xenopus laevis Xl DB Msx1 (AAH81101)         Osteichthyes Danio rerio Dr DB DB Msh-A (Q03357), Msh-B (Q03356), Msh-C (Q01703), Msh-D (Q01704), MsxE (AAB03273)     Cephalochordata         Leptocardia Branchiostoma floridae Bf DB Msx (CAA10201)     Urochordata         Ascidiacea Ciona intestinalis Ci DB DB CAD56691 Molgula oculata Mo DB AAA87223         Appendicularia Oikopleura dioica Od DB AAW24005 Hemichordata         Enteropneusta Saccoglossus kowalevskii Sk DB Msx (ABD97280) Echinodermata         Echinoidea Strongylocentrotus purpuratus Sp DB DB SpMsx (AAB97688) Heliocidaris erythrogramma Her DB Msx (AAY86177) Heliocidaris tuberculata Ht DB Msx (AAY86178)         Asteroidea Asterina pectinifera Ap Fosmid PCR MsxA [Ast-1 (AB302953)]a, MsxB [Ast-3 (AB302954)]a Arthropoda     Chelicerata Limulus polyphemus Lp PCR MsxA (AB302959)a Pandinus imperator Pi PCR MsxA (AB302960)a, MsxB (AB302961)a     Crustacea Artemia franciscana Af PCR Msx (AB302962)a     Insecta Tribolium castaneum Tc DB Muscle segment homeoprotein (AAW21975) Drosophila melanogaster Dm DB DB msh (Q03372) Anopheles gambiae Ag DB DB EAA08817 Apis mellifera Ame DB Msx (AB362784)b Tardigrada Milnesium tardigradum Mt PCR Msx (AB302966)a Nematoda Caenorhabditis elegans Ce DB DB vab-15 (Q09604) Caenorhabditis briggsae Cb DB CAE58718 Mollusca         Cephalopoda Loligo bleekeri Lb PCR Msx (AB302963)a Octopus ocellatus Oo PCR Msx (AB302964)a         Bivalvia Corbicula fluminea Cf Fosmid PCR Msx [Cor-2/Cor-3 (AB302955)]a Brachiopoda         Inarticulata Lingula anatina La PCR Msx (AB302965)a Annelida         Polychaeta Platynereis dumerilii Pd DB Msx (CAJ38810)         Oligochaeta Tubifex tubifex Tt Fosmid PCR MsxA [Tt-1 (AB302956)]a, MsxB [Tt-3 (AB302957)]a         Hirudinida Helobdella sp. Hel DB Msx (AAB37254) Echiura Urechis caupo Uc PCR Msx (AB302967)a Platyhelminthes         Tubellaria Schmidtea mediterranea Sm DB Msx (AB362785)b Thysanozoon sp. Thy PCR Msx (AB302968)a Cnidaria         Hydrozoa Hydra viridis Hvi DB CAA45912 Hydra vulgaris Hvu DB CAB88390 Scolionema suvaense Ss Fosmid PCR Msx [Sco-1/Sco-6 (AB302958)]a Podocoryne carnea Pc DB AAX58756         Anthozoa Nematostella vectensis Nv PCR Msx (AB362783)a Anemonia erythraea Ae PCR Msx (AB302969)a Acropora millepora Ami DB msh3 (ABK41269) Porifera         Demospongiae Ephydatia fluviatilis Ef DB prox3 (AAA20151) Phylum Subphylum Class Species Abbreviation Genome cDNA Msx-Related Genes (Synonym, Accession Number) Chordata     Vertebrata         Mammalia Mus musculus Mm DB DB Msx1 (AAH16426), Msx2 (Q03358), Msx3 (P70354)         Aves Gallus gallus Gg DB DB Msx1 (Hox7, P50223), Msx2 (Hox8, P28362)         Amphibia Xenopus laevis Xl DB Msx1 (AAH81101)         Osteichthyes Danio rerio Dr DB DB Msh-A (Q03357), Msh-B (Q03356), Msh-C (Q01703), Msh-D (Q01704), MsxE (AAB03273)     Cephalochordata         Leptocardia Branchiostoma floridae Bf DB Msx (CAA10201)     Urochordata         Ascidiacea Ciona intestinalis Ci DB DB CAD56691 Molgula oculata Mo DB AAA87223         Appendicularia Oikopleura dioica Od DB AAW24005 Hemichordata         Enteropneusta Saccoglossus kowalevskii Sk DB Msx (ABD97280) Echinodermata         Echinoidea Strongylocentrotus purpuratus Sp DB DB SpMsx (AAB97688) Heliocidaris erythrogramma Her DB Msx (AAY86177) Heliocidaris tuberculata Ht DB Msx (AAY86178)         Asteroidea Asterina pectinifera Ap Fosmid PCR MsxA [Ast-1 (AB302953)]a, MsxB [Ast-3 (AB302954)]a Arthropoda     Chelicerata Limulus polyphemus Lp PCR MsxA (AB302959)a Pandinus imperator Pi PCR MsxA (AB302960)a, MsxB (AB302961)a     Crustacea Artemia franciscana Af PCR Msx (AB302962)a     Insecta Tribolium castaneum Tc DB Muscle segment homeoprotein (AAW21975) Drosophila melanogaster Dm DB DB msh (Q03372) Anopheles gambiae Ag DB DB EAA08817 Apis mellifera Ame DB Msx (AB362784)b Tardigrada Milnesium tardigradum Mt PCR Msx (AB302966)a Nematoda Caenorhabditis elegans Ce DB DB vab-15 (Q09604) Caenorhabditis briggsae Cb DB CAE58718 Mollusca         Cephalopoda Loligo bleekeri Lb PCR Msx (AB302963)a Octopus ocellatus Oo PCR Msx (AB302964)a         Bivalvia Corbicula fluminea Cf Fosmid PCR Msx [Cor-2/Cor-3 (AB302955)]a Brachiopoda         Inarticulata Lingula anatina La PCR Msx (AB302965)a Annelida         Polychaeta Platynereis dumerilii Pd DB Msx (CAJ38810)         Oligochaeta Tubifex tubifex Tt Fosmid PCR MsxA [Tt-1 (AB302956)]a, MsxB [Tt-3 (AB302957)]a         Hirudinida Helobdella sp. Hel DB Msx (AAB37254) Echiura Urechis caupo Uc PCR Msx (AB302967)a Platyhelminthes         Tubellaria Schmidtea mediterranea Sm DB Msx (AB362785)b Thysanozoon sp. Thy PCR Msx (AB302968)a Cnidaria         Hydrozoa Hydra viridis Hvi DB CAA45912 Hydra vulgaris Hvu DB CAB88390 Scolionema suvaense Ss Fosmid PCR Msx [Sco-1/Sco-6 (AB302958)]a Podocoryne carnea Pc DB AAX58756         Anthozoa Nematostella vectensis Nv PCR Msx (AB362783)a Anemonia erythraea Ae PCR Msx (AB302969)a Acropora millepora Ami DB msh3 (ABK41269) Porifera         Demospongiae Ephydatia fluviatilis Ef DB prox3 (AAA20151) NOTE.—DB, sequence collected from NCBI databases; fosmid, isolated as fosmid genomic clones in this study; PCR, cloned by PCR amplification of the cDNA. a Sequence newly determined in this study. b Sequences edited from the publicized draft sequences (http://www.ncbi.nlm.nih.gov/BLAST/tracemb.shtml). Open in new tab The 18S ribosomal RNA sequences were collected from the NCBI database (accession number, supplementary text, see Supplementary Material online). These sequences were chosen because they were derived from animal species that were identical, or closely related, to those used in the Msx phylogenetic analysis (table 1). The search for Eh-like sequences was done as follows. A local aa sequence database was constructed by DNASpace (Hitachi Software Engineering, Tokyo, Japan). The database was subjected to a homology search by using the Smith–Waterman algorithm (Smith and Waterman 1981) and the consensus Eh sequences FSV[DE] × [IL][IL] as key sequences. Sequences that gave scores of more than 15 under set parameters (i.e., Matrix, BLOSUM62; initial gap penalty, −5; extension gap penalty, −2) were listed as candidate sequences for the alignment shown in table 2. Table 2 Eh1 Motifs Found in N-terminal Regions of Metazoan Msx Proteins Species Eh1N Eh1C Gg_1 55 FSVEALMA 106 FPSVGALGK Mm_1 61 FSVEALMA 118 FSVGGLL-K Xl_1 67 FSVEALMA 120 YPVGAIM-Qa Dr_E 44 FSVEALMA 68 FSVEVLQLP Bf 46 FSVASLMA 91 FSVEGILSK Ci 76 FSIEFLLS Od 6 FSVDWIIS Sk 73 FSVASLIS 112 FSVEGILSK Sp 61 FRVESLFS 100 HSVENILAK Her 51 FRVESFFS 104 HSVENILAK Ap_Ab 59 FRVESLVG 108 YTVEGILAS Dm 134 FSVASLLA Ame 43 FSVDSLLS Ce 14 FSVESLLT Cf 79 FGVDSIIS 151 FSMDEILGK Pd 132 FSVDSIIS 180 FSVDGILSK Tt_Ab 119 FSVAAIMA 170 FTVDGILGG Ss 53 FSIDYILN Pc 30 FSIDYLLN Nv 29 FSVESLIS 57 FSVESILEK Ami 14 FSVESLIS 43 FSVERLLDK Consensus FSV[D/E]S[L/I][L/I]S FSV[D/E]GILXK Species Eh1N Eh1C Gg_1 55 FSVEALMA 106 FPSVGALGK Mm_1 61 FSVEALMA 118 FSVGGLL-K Xl_1 67 FSVEALMA 120 YPVGAIM-Qa Dr_E 44 FSVEALMA 68 FSVEVLQLP Bf 46 FSVASLMA 91 FSVEGILSK Ci 76 FSIEFLLS Od 6 FSVDWIIS Sk 73 FSVASLIS 112 FSVEGILSK Sp 61 FRVESLFS 100 HSVENILAK Her 51 FRVESFFS 104 HSVENILAK Ap_Ab 59 FRVESLVG 108 YTVEGILAS Dm 134 FSVASLLA Ame 43 FSVDSLLS Ce 14 FSVESLLT Cf 79 FGVDSIIS 151 FSMDEILGK Pd 132 FSVDSIIS 180 FSVDGILSK Tt_Ab 119 FSVAAIMA 170 FTVDGILGG Ss 53 FSIDYILN Pc 30 FSIDYLLN Nv 29 FSVESLIS 57 FSVESILEK Ami 14 FSVESLIS 43 FSVERLLDK Consensus FSV[D/E]S[L/I][L/I]S FSV[D/E]GILXK a The sequence was highly diverged from other Eh1C sequences. In Xenopus tropicalis (AAH62514), Eh1C sequence was FPVGGIMK. b Ap_B and Tt_B Eh1N were identical to those in Ap_A and Tt_A, respectively. Open in new tab Table 2 Eh1 Motifs Found in N-terminal Regions of Metazoan Msx Proteins Species Eh1N Eh1C Gg_1 55 FSVEALMA 106 FPSVGALGK Mm_1 61 FSVEALMA 118 FSVGGLL-K Xl_1 67 FSVEALMA 120 YPVGAIM-Qa Dr_E 44 FSVEALMA 68 FSVEVLQLP Bf 46 FSVASLMA 91 FSVEGILSK Ci 76 FSIEFLLS Od 6 FSVDWIIS Sk 73 FSVASLIS 112 FSVEGILSK Sp 61 FRVESLFS 100 HSVENILAK Her 51 FRVESFFS 104 HSVENILAK Ap_Ab 59 FRVESLVG 108 YTVEGILAS Dm 134 FSVASLLA Ame 43 FSVDSLLS Ce 14 FSVESLLT Cf 79 FGVDSIIS 151 FSMDEILGK Pd 132 FSVDSIIS 180 FSVDGILSK Tt_Ab 119 FSVAAIMA 170 FTVDGILGG Ss 53 FSIDYILN Pc 30 FSIDYLLN Nv 29 FSVESLIS 57 FSVESILEK Ami 14 FSVESLIS 43 FSVERLLDK Consensus FSV[D/E]S[L/I][L/I]S FSV[D/E]GILXK Species Eh1N Eh1C Gg_1 55 FSVEALMA 106 FPSVGALGK Mm_1 61 FSVEALMA 118 FSVGGLL-K Xl_1 67 FSVEALMA 120 YPVGAIM-Qa Dr_E 44 FSVEALMA 68 FSVEVLQLP Bf 46 FSVASLMA 91 FSVEGILSK Ci 76 FSIEFLLS Od 6 FSVDWIIS Sk 73 FSVASLIS 112 FSVEGILSK Sp 61 FRVESLFS 100 HSVENILAK Her 51 FRVESFFS 104 HSVENILAK Ap_Ab 59 FRVESLVG 108 YTVEGILAS Dm 134 FSVASLLA Ame 43 FSVDSLLS Ce 14 FSVESLLT Cf 79 FGVDSIIS 151 FSMDEILGK Pd 132 FSVDSIIS 180 FSVDGILSK Tt_Ab 119 FSVAAIMA 170 FTVDGILGG Ss 53 FSIDYILN Pc 30 FSIDYLLN Nv 29 FSVESLIS 57 FSVESILEK Ami 14 FSVESLIS 43 FSVERLLDK Consensus FSV[D/E]S[L/I][L/I]S FSV[D/E]GILXK a The sequence was highly diverged from other Eh1C sequences. In Xenopus tropicalis (AAH62514), Eh1C sequence was FPVGGIMK. b Ap_B and Tt_B Eh1N were identical to those in Ap_A and Tt_A, respectively. Open in new tab Animals Genomic resources for Scolionema suvaense, Tubifex tubifex, Loligo bleekeri, Octopus ocellatus, Corbicula fluminea, Pandinus imperator, Artemia franciscana, and Asterina pectinifera have been described (Aruga et al. 2006). The maintenance and recovery of Milnesium tardigradum species have been described (Suzuki 2003). Limulus polyphemus, Lingula anatina, Urechis caupo, and Anemonia erythraea were purchased from local vendors. Thysanozoon sp. was collected from the seashore in Kanagawa Prefecture, Japan. Its picture can be seen at http://lcn.brain.riken.jp/photos.html. Species identification was done by sequencing 18S ribosomal RNA genes (data not shown). Polymerase Chain Reaction Cloning of Msx cDNA RNAs were isolated by using TRIzol reagent (Invitrogen, CA) in accordance with the manufacturer's recommendation. cDNAs were generated by using a 3′-Full RACE Core Set (Takara Bio, Shiga, Japan). The homologs were initially identified by nested polymerase chain reaction (PCR) on cDNA or genomic templates. The following primers were used for PCR amplification of the homeodomain (HD) region of Msx family genes. The first PCR was carried out by using MshDF-1, 5′-YTIMGIAARCAYAARACIAAYMG-3,′ and MshDR-1, 5′-TTIGCICKRTTYTGRAACCA-3′, and the second PCR was done by using MshDF-2, 5′-AAYMGIAARCCIMSIACICCITT-3′, and MshDR-2, 5′-CCAIATYTTIAYYTGIGTYTC-3′. Each PCR consisted of 35 cycles of 94 °C for 1 min, 38 °C for 1 min, and 72 °C for 2 min. The PCR was performed with ExTaq DNA polymerase (Takara Bio) in the presence of BD TaqStart anti-Taq antibody (BD Biosciences, San Diego, CA). cDNAs corresponding to HD at their 3′ ends were cloned by using a 3′-Full RACE Core Set (Takara Bio). The aa sequences were deduced from nucleotide sequencing of multiple PCR fragments. Cloning of Genomic DNA Isolation of high molecular weight DNA, fosmid library construction, and library screening were done as described (Aruga et al. 2006). Fosmid genomic libraries were prepared with CopyControl pCC1FOS vector (Epicentre, WI). In this study, we isolated 8 fosmid clones from A. pectinifera [Ast-1 (36,228 bp), Ast-3 (42,842 bp)], C. fluminea [Cor-2 (37,605 bp), Cor-3 (37,272 bp)], T. tubifex [Tub-1 (38,378 bp), Tub-3 (35,661 bp)], and S. suvaense [Sco-1 (42,994 bp), Sco-6 (38,643 bp)]. Cor-2 and Cor-3 were overlapping with a completely matching 34,763-bp sequence. Sco-1 and Sco-6 were overlapping with a completely matching 24,029-bp sequence. Ast-1 and Ast-3 and Tub-1 and Tub-3 were derived from independent genes. Nematostella vectensis genomic DNA was obtained by PCR amplification of N. vectensis genomic DNA using following primers that are based on the draft genome sequences (http://www.ncbi.nlm.nih.gov/BLAST/tracemb.shtml): 5′-CCACCATGGAGGCGGATCGCGATTTGCCT-3′ and 5′-ATAAACTAATGCGGGTGCAGAAAACCTG-3′. DNA Sequencing and Molecular Phylogenic Analysis Sequencing and data assembly were done as described (Toyoda et al. 2002). Genomic sequences of Schmidtea mediterranea, Hydra vulgaris, Caenorhabditis elegans, Strongylocentrotus purpuratus, Ciona intestinalis, and N. vectensis were derived from public databases (http://www.ncbi.nlm.nih.gov/BLAST/tracemb.shtml). Sequence analysis was done with DNASISPro (Hitachi Software Engineering, Tokyo, Japan), Sequencher (Gene Codes, Ann Arbor, MI), and Genetyx (Genetyx, Tokyo, Japan) software. Homology searching was performed against a public database (http://www.ncbi.nlm.nih.gov/BLAST/) by using Blast and discontinuous MEGA Blast. The aa and nucleotide sequences were aligned by ClustalW (Thompson et al. 1994). Some of the aligned sequences were corrected by visual inspection. Ancestral sequences were deduced from the present aa sequence data by using ANCESCON, a distance-based program that gives more accurate ancestral sequence reconstruction than do PAML, PHYLIP, and PAUP* at large evolutionary distances (Cai et al. 2004). Phylogenetic tree analysis was done with MEGA3.1 (Neighbor-Joining [NJ] tree and Maximal Parsimony [MP] tree Kumar et al. 2004) and MrBayes 3.1.2 (Bayesian Inference [BI] tree Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003). NJ tree was based on the distance calculation with point accepted mutation (PAM) matrix (Dayhoff et al. 1978) after removing position containing gaps (complete deletion option). In the NJ and MP trees, the tree reliability was estimated by bootstrap test (Felsenstein 1985) with 1,000 repetitions. In the BI analysis, we used an empirical model (WAG distances, Whelan and Goldman 2001) with gamma, alpha shape parameter, and aa frequencies estimated from the data. We ran 1,000,000 generations with 1 cold and 3 incrementally heated Markov chains, random starting trees for each chain, and trees sampled every 100 generations. We constructed a 50% major rule consensus tree from the last 15,000 trees that were saved (burnin = 2,500). The evolutionary distances in figure 3C were calculated by MEGA3.1. The distances of aa sequences (Msx) and nucleotide sequences (18S RNA) were determined by using the PAM matrix (Dayhoff et al. 1978) and the Tamura–Nei model (Tamura and Nei 1993), respectively. Measurement was done after removal of any alignment gap-containing sites, assuming different evolutionary rates among sites (gamma distribution, α = 0.37 for Msx aa sequences and α = 0.42 for 18S ribosomal RNA nucleotide sequences). The α parameters were estimated by Tree-Puzzle program (Schmidt et al. 2002). For the evolutionary distance analysis, we omitted the paralogs and sequences from identical genera (those in Heliocidaris, Caenorhabditis, and Hydra species), except for one representative sequence. The representative sequences were the most strongly conserved in each group. According to these criteria, Mm_2, Gg_2, Dr_E, Ap_A, Pi_B, Tt_A, Her, Ce, and Hvu were used in the distance analysis, whereas Mm_1, Mm_3, Dr_A, Dr_B, Dr_C, Dr_D, Gg_1, Ap_B, Pi_A, Tt_B, Ht, Cb, and Hvi were not used. Comparison of evolutionary rates among the chordate and cnidarian subgroups (fig. 3D and E) were done by counting different aa residues between the target aa sequences and a reference sequence (Anc_Msx) by using MEGA3.1. Counting was performed for the sequences from the Chordata and Cnidaria. Subsequently, the numbers of residues nonidentical and identical to the reference sequence were placed in 2 × 2 tables to compare the differences between 2 sequences from distinct sister subgroups. Fisher's exact test was performed to evaluate the statistical significance of the differences in the ratios of nonidentical to identical residues. Plasmid Construction and Immunoprecipitation Xl_Msx1 cDNA (a gift from Dr Atsushi Suzuki) was cloned into pCS2 + Myc tag vector (Turner and Weintraub 1994). Nv_Msx genomic DNA and mouse Grg1 cDNA were cloned into pcDNA3.1/His (Invitrogen) that was modified to have an initiation methionine, and either 2 Flag epitope tags, 3 HA tags at their HindIII/KpnI sites. The Grg1 cDNA was obtained from Riken FANTOM clones (http://www.gsc.riken.go.jp/e/FANTOM/) (Carninci et al. 2005). Site-directed mutations were introduced into the protein-coding region according to the method of Ito et al. (1991). The primer sequences will be provided upon request. A truncation (stop) mutant of Xl_Msx1 was generated by inserting a stop linker (5′-TGAATATCA-3′) into a unique SmaI site in the open reading frame (ORF) of Xl_Msx cDNA. Immunoprecipitation was performed essentially as described (Ishiguro et al. 2007). Briefly, COS7 cells were transfected with the epitope-tagged Msx and Grg1 expression vectors with Lipofectamine Plus reagent (Invitrogen). The transfected cells were washed and harvested in PBS(−) containing 1 mM phenylmethylsulfonylfluoride (PMSF), and total cell extracts were prepared with a lysis wash buffer consisting of 20 mM HEPES–KOH, pH 7.8, 10% glycerol, 150 mM NaCl, 0.5 mM dithiothreitol, 0.1 mM ethylenediaminetetraacetic acid, 0.5% NP-40, and 1 mM PMSF. The extracts were incubated with anti-FLAG or anti-HA affinity beads (Sigma, St Louis, MO) at 4 °C for 6 h. The beads were subsequently washed with the buffer. The bound proteins were separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis, transferred onto polyvinylidene difluoride membranes (Nihon Millipore, Tokyo, Japan), and detected by antibodies against the epitope tags using ECL western blotting detection reagent (GE Healthcare, Tokyo, Japan). Results Collection of Msx-Related Sequences from Metazoa A homology search against current sequence databases revealed Msx orthologs could only be detected in metazoans. We did not find definitive Msx orthologs in fungi, algae, plants, and protists. This result is consistent with a previous finding that homeobox genes belonging to the ANTP class, in which Msx is included, are confined to the Metazoa (Holland and Takahashi 2005). We first collected Msx family gene sequences through a database search. The homology search, which covered both complete and incomplete nucleotides and protein sequence databases, revealed the presence of Msx homologs in the Porifera, Cnidaria, Nematoda, Arthropoda, Annelida, Echinodermata, Hemichordata, and Chordata (table 1). Porifera prox3 (Seimiya et al. 1994) was an Msx ortholog, as suggested by the phylogenetic tree analysis in Galle et al. (2005). Genomic information on exon–intron boundaries was available for vertebrates, insects, and a urochordate. To improve the comprehensiveness of our analysis, we newly cloned partial cDNA fragments of Msx homologs from a wide range of animals in the Platyhelminthes, Mollusca, Echiura, Tardigrada, Brachiopoda, Cnidaria, Annelida, Arthropoda, and Echinodermata. In addition, fosmid genomic clones were isolated from starfish (A. pectinifera), sludgeworm (T. tubifex), bivalve (C. fluminea), and jellyfish (S. suvaense), and their entire nucleotide sequence was determined. The number of Msx genes collected in a species varied from 1 to 5. We observed closely related, but significantly diverged, genes (paralogs) in T. tubifex, A. pectinifera, and P. imperator, besides those described earlier in vertebrates (Ekker et al. 1997). In total, we obtained 17 additional Msx sequences from 14 animal species from 9 animal phyla (table 1). The aa Residues Functionally Important in Mammalian Msx HDs are Strongly Conserved in Metazoan Orthologs After sequencing the collected cDNA and genomic clones, the aa sequences were deduced. We first compared the aa sequences of the HD and its C-terminally flanking (CF) region (fig. 1). The alignment revealed that the HDs were strongly conserved. The Msx HD has DNA-binding activity. The residues that are responsible for the molecular interaction with DNA bases (R2, R5, K46, I47, Q50, N51, and R58) or DNA phosphoribosyl backbones (K3, T6, F8, Y25, R31, W48, R53, K55, and K57) (Hovde et al. 2001) were absolutely conserved among the collected sequences. Some residues (F8 and R58) were conserved in Msx, but not in Dlx family proteins that contained HDs most similar to Msx (fig. 1) (Gauchat et al. 2000). The Msx HD is also known to physically interact with proteins that are essential for the molecular function of Msx. K3, R5, and F8 are required for interactions and transcriptional repression by the general transcription factor TFIIF (Zhang et al. 1996). Mutations in human MSX HDs, including R31P (STA, Hu et al. 1998), L13P (corresponding to MSX2 L154P in PFM, Wuyts et al. 2000), RK18-19del (corresponding to MSX2 RK159-160del in PFM, Wuyts et al. 2000), P7H (corresponding to MSX2 P148H in CSO, Jabs et al. 1993) are missense mutations that cause genetic disorders. These sites were conserved among the metazoan Msx proteins, except that RK18-19 had different residues in urochordates, insects, and cnidarian species (fig. 1). These results indicate that functionally important residues in mammalian Msx proteins are strongly conserved in all metazoan Msx HDs. FIG. 1.— Open in new tabDownload slide The aa alignment of HD and a C-terminal flanking region (HD + CF). “Dots” indicate identical aa residues to those in the top sequence (Mm_2), and “hyphens” indicate the spaces artificially inserted in the multiple alignment program. The abbreviations for the animal species names are given in table 1. Anc_Msx indicates the metazoan ancestral Msx sequence deduced by ANCESCON. “Plus” indicates the aa residues that are responsible for molecular interaction with DNA bases, and “asterisk” indicates DNA phosphoribosyl backbones. The locations of residues responsible for human genetic disorders (CSO, PFM, and STA; see text) are indicated at the top of the alignment. S206A indicates an artificial missense mutation introduced into Xl-Msx1 in figure 6. “Shading” indicates the locations of DNA-associated or disease-causative aa residues in the alignment. “Open boxes” and “open circle” indicate the presence of phase-0 and phase-2 introns, respectively. The genes subjected for the exon–intron boundary analysis are shown in figure 4. Bottom horizontal line indicates the region used in the calculation of evolutionary distances. FIG. 1.— Open in new tabDownload slide The aa alignment of HD and a C-terminal flanking region (HD + CF). “Dots” indicate identical aa residues to those in the top sequence (Mm_2), and “hyphens” indicate the spaces artificially inserted in the multiple alignment program. The abbreviations for the animal species names are given in table 1. Anc_Msx indicates the metazoan ancestral Msx sequence deduced by ANCESCON. “Plus” indicates the aa residues that are responsible for molecular interaction with DNA bases, and “asterisk” indicates DNA phosphoribosyl backbones. The locations of residues responsible for human genetic disorders (CSO, PFM, and STA; see text) are indicated at the top of the alignment. S206A indicates an artificial missense mutation introduced into Xl-Msx1 in figure 6. “Shading” indicates the locations of DNA-associated or disease-causative aa residues in the alignment. “Open boxes” and “open circle” indicate the presence of phase-0 and phase-2 introns, respectively. The genes subjected for the exon–intron boundary analysis are shown in figure 4. Bottom horizontal line indicates the region used in the calculation of evolutionary distances. The Degree of HD Sequence Conservation Varies among Taxonomic Groups To determine the structural features of the Msx proteins in each taxon, we first drew a phylogenetic tree, based on the alignment shown in figure 1, by using the NJ (fig. 2), BI (supplementary fig. 1, Supplementary Material online), and MP (supplementary fig. 2, Supplementary Material online) methods. Monophyly of the Hydrozoa (Hvu, Ss, and Pc) was recovered by the NJ/BI/MP trees, and those of the Diptera (Dm and Ag) and Cephalopoda (Oo and Lb) were recovered by the BI/MP trees. However, with the exception of paralogs, there were no other strong groupings supported by multiple trees. Instead, there was a high level of similarity between several evolutionarily distant animal species [e.g., only 2 changes in the 69 aa (9–77) sequences between Mm_2 (Chordata) and La (Brachiopoda) (fig. 1)]. We therefore speculated that the metazoan ancestral Msx sequence was strongly retained in some animal groups but not in others. FIG. 2.— Open in new tabDownload slide NJ tree used for the metazoan Msx HD + CF aa sequences. The tree was rooted with mouse Dlx1 (Mm Dlx1) as an outgroup. Internal labels indicate bootstrap values (1,000 replicates). PAM matrix was used for distance calculation by using the alignment shown in figure 1. Branches with less than 50% bootstrap values were condensed. Abbreviations for animal species names are indicated in table 1. FIG. 2.— Open in new tabDownload slide NJ tree used for the metazoan Msx HD + CF aa sequences. The tree was rooted with mouse Dlx1 (Mm Dlx1) as an outgroup. Internal labels indicate bootstrap values (1,000 replicates). PAM matrix was used for distance calculation by using the alignment shown in figure 1. Branches with less than 50% bootstrap values were condensed. Abbreviations for animal species names are indicated in table 1. To test this idea, we determined the mean distances of the HD + CF sequence between all pairs of 12 taxa (fig. 3A). In this grouping, the Tardigrada were grouped with the Arthropoda in light of their sister-group relationship, as determined by morphological and molecular analyses (Brusca and Brusca 2003); the Echiura were grouped with the Annelida (McHugh 1997; Bleidorn et al. 2003); and the Hemichordata were grouped with the Echinodermata on the basis of recent molecular phylogenetic analyses (Cameron et al. 2000; Peterson 2004). Two classes in the Cnidaria—Hydrozoa and Anthozoa—were tested separately because analysis of the phylogenetic trees suggested that there was a strong difference in tree length between these two taxa. For reference sequence comparison, we deduced the metazoan ancestor Msx sequence (Anc_Msx in fig. 1) by using ANCESCON, a program for the reconstruction of ancestral protein sequences that takes into account the observed variations in evolutionary rates between positions (Cai et al. 2004). We first determined the aa number where substitutions occurred in the Anc_Msx sequence for each taxon (fig. 3B). The sequences from the Vertebrata–Cephalochordata (1 [group no. in fig. 3]), Echinodermata–Hemichordata (3), Mollusca (6), Brachiopoda (7), and Anthozoa (11) had low values (no greater than 6 aa substitutions), whereas those from the Urochordata (2), Hydrozoa (10), and Porifera (12) had large values (22–27 aa substitutions). The mean distances between all pairs of 2 taxa were then determined concerning the Msx and 18S ribosomal RNA sequences (fig. 3C, Supplementary table 1, Supplementary Material online). There were strongly deviated pairs along the distances of the Msx sequences. The lowest group (pairs under the oblique line in fig. 3C) included pairs among the Vertebrata–Cephalochordata–Hemichordata, Echinodermata, Mollusca, Brachiopoda, and Anthozoa groups, whereas the highest one (pairs above the broken line in fig. 3C) contained pairs among mainly the Urochordata, Hydrozoa, and Porifera. From these results, we postulated that the Vertebrata, Cephalochordata, Hemichordata, Echinodermata, Mollusca, Brachiopoda, and Anthozoa Msx sequences shared highly conserved features among all the metazoan Msx sequences. On the other hand, the sequences belonging to the Urochordata, Hydrozoa, and Porifera diverged strongly. FIG. 3.— Open in new tabDownload slide Evolutionary distances computed from metazoan Msx aa sequences. (A) Taxonomic groups used in the analysis. Numbers correspond to those in (B and C). (B) aa substations in comparison with ancestral Msx (Anc_Msx in fig. 1). The numbers of substituted aa residues were counted for each Msx HD + CF sequence. Solid bars indicate mean numbers of substituted aa in each taxon shown in (A). Error bars indicate SDs. Error bars are shown if there is more than one sequence; the number of sequences in each group (n) are indicated in the right column. The within-group distances were as follows: Vertebrata–Cephalochordata–Hemichordata, 0.044; Urochordata, 1.106; Echinodermata, 0.032; Arthropoda–Tardigrada, 0.464; Mollusca, 0.021; Annelida–Echiura, 0.152; Platyhelminthes, 0.657; Hydrozoa, 0.889; and Anthozoa, 0.066. The values for the Vertebrata–Cephalochordata–Hemichordata, Echinodermata, Mollusca, and Anthozoa were clustered below 0.05. (C) Between-group means of evolutionary distances of all pairs of the 12 taxa in (A). Scatter graph indicates distances based on the Msx aa sequences [y axis, normal scale (left graph); log scale (right graph)] and those based on the 18S RNA gene nucleotide sequences (x axis). The ‘oblique lines” are arbitrarily placed for the indication of the pairs with the low (below the continuous line) and high (above the broken line) values along the distance of Msx sequences. The sequences subjected to the analysis in (B) and (C) were selected according to the criteria in Materials and Methods. (D and E) Probabilities of equal molecular evolutionary rate, as determined by Fisher's exact test. The tests were done between pairs of sequence from Cnidaria sister groups (D) or Chordata sister groups (E). The abbreviation of the sequence name can be seen in table 1. In both analyses, the presumptive ancestral Msx sequence (Anc_Msx) was used as a reference sequence. Deuterostomia ancestral sequence was identical to Anc_Msx in the ANCESCON analysis. FIG. 3.— Open in new tabDownload slide Evolutionary distances computed from metazoan Msx aa sequences. (A) Taxonomic groups used in the analysis. Numbers correspond to those in (B and C). (B) aa substations in comparison with ancestral Msx (Anc_Msx in fig. 1). The numbers of substituted aa residues were counted for each Msx HD + CF sequence. Solid bars indicate mean numbers of substituted aa in each taxon shown in (A). Error bars indicate SDs. Error bars are shown if there is more than one sequence; the number of sequences in each group (n) are indicated in the right column. The within-group distances were as follows: Vertebrata–Cephalochordata–Hemichordata, 0.044; Urochordata, 1.106; Echinodermata, 0.032; Arthropoda–Tardigrada, 0.464; Mollusca, 0.021; Annelida–Echiura, 0.152; Platyhelminthes, 0.657; Hydrozoa, 0.889; and Anthozoa, 0.066. The values for the Vertebrata–Cephalochordata–Hemichordata, Echinodermata, Mollusca, and Anthozoa were clustered below 0.05. (C) Between-group means of evolutionary distances of all pairs of the 12 taxa in (A). Scatter graph indicates distances based on the Msx aa sequences [y axis, normal scale (left graph); log scale (right graph)] and those based on the 18S RNA gene nucleotide sequences (x axis). The ‘oblique lines” are arbitrarily placed for the indication of the pairs with the low (below the continuous line) and high (above the broken line) values along the distance of Msx sequences. The sequences subjected to the analysis in (B) and (C) were selected according to the criteria in Materials and Methods. (D and E) Probabilities of equal molecular evolutionary rate, as determined by Fisher's exact test. The tests were done between pairs of sequence from Cnidaria sister groups (D) or Chordata sister groups (E). The abbreviation of the sequence name can be seen in table 1. In both analyses, the presumptive ancestral Msx sequence (Anc_Msx) was used as a reference sequence. Deuterostomia ancestral sequence was identical to Anc_Msx in the ANCESCON analysis. Evolutionary Rates Differ between Two Independent Pairs of Sister Groups The evolutionary rates of 2 representative phyla, the Cnidaria and Chordata, were compared. After comparison between the ancestor sequence and each sequence from the Cnidaria, we compared the diversification rates in any pair of sequences between the Anthozoa and Hydrozoa (fig. 3D, supplementary table 2, Supplementary Material online). In the Chordata, we compared the diversification rates between the Vertebrata, Cephalochordata, and Urochordata (fig. 3E, supplementary table 2, Supplementary Material online). Fisher's exact test of the results rejected the null hypothesis that evolutionary rates were equal between the 2 tested sequences in both the Anthozoa–Hydrozoa and the Vertebrata–Cephalochordata–Urochordata comparisons (all P values were less than 0.01). The analysis indicated clear differences in the evolutionary rates among sister groups in the Cnidaria and Chordata. Comparison of Exon–Intron Organization of Msx Family Proteins We next compared the exon–intron organization of Msx genes (fig. 4). All the Msx genes examined contained at least 1 intron in the protein-coding region. All of the exon–intron boundaries followed the GT/AG rule. Eighteen Msx genes possessed only 1 intron, whereas the other 5 (Ame-Msx, Ci-Msh, Ce-VAB15, Tt-MsxA, and Tt-MsxB) possessed additional introns. The one in the N-terminal region flanking the HD was located in a region with little sequence similarity (fig. 5). However, marked similarities were observed in the distance from the most N-terminal end of the HD (11–28 aa), and the phases of the intron insertion site in the ORF were “1” without exception. Furthermore, Seimiya et al. (1994) putatively assigned a “phase-1” splicing acceptor site in the 17 aa from the N-terminal end of the Ef-Msx (prox3) HD, although the preceding exon was not identified. When we aligned the N-terminal region aa sequence by adjusting the intron position as the cardinal point, there was a weakly conserved sequence near the intron position (fig. 5). The conserved sequence was summarized as FPWMQ, where tryptophan was strongly conserved (hereafter, we call the conserved sequence “PWM”). FIG. 4.— Open in new tabDownload slide Schematic drawing of exon–intron structures of Msx genes from various animals. “Open boxes” and “closed boxes” indicate non–HD and HD areas, respectively, of the Msx protein-coding region. “Horizontal bars” indicate introns, and “asterisks” above the horizontal bars indicate AC introns. FIG. 4.— Open in new tabDownload slide Schematic drawing of exon–intron structures of Msx genes from various animals. “Open boxes” and “closed boxes” indicate non–HD and HD areas, respectively, of the Msx protein-coding region. “Horizontal bars” indicate introns, and “asterisks” above the horizontal bars indicate AC introns. FIG. 5.— Open in new tabDownload slide The aa alignment of the HD N-terminal flanking region. PWM indicates the newly identified conserved domain (shaded, PWM). All the AC introns were located in phase 1 of the “boxed” residues. “Long arrow” indicates the HD region. “Hyphens” and “equals marks” indicate the spaces and deletions, respectively, that were introduced for the alignment. The equals marks in Dm and Ame represent deletions of PPGMFPGAGFGG and HGHLYTPHGGPTSPN, respectively. The “asterisk” in the Ef line indicates the presence of a putative splicing acceptor site in phase-1 of the following HD-containing exon. FIG. 5.— Open in new tabDownload slide The aa alignment of the HD N-terminal flanking region. PWM indicates the newly identified conserved domain (shaded, PWM). All the AC introns were located in phase 1 of the “boxed” residues. “Long arrow” indicates the HD region. “Hyphens” and “equals marks” indicate the spaces and deletions, respectively, that were introduced for the alignment. The equals marks in Dm and Ame represent deletions of PPGMFPGAGFGG and HGHLYTPHGGPTSPN, respectively. The “asterisk” in the Ef line indicates the presence of a putative splicing acceptor site in phase-1 of the following HD-containing exon. On the basis of these observations, we speculated that the intron was acquired in a common ancestor of a eumetazoa, or possibly a metazoan, Msx, and has been retained in the all Msx family genes; we call this intron the AC (absolutely conserved) intron. The other conserved intron position was located near the C-terminal end of the HD at V45 (phase 0) in 4 genes [Tt-MsxA, Tt-MsxB (Annelida), Ame-Msx (Arthropoda), and Ci-Msh (Urochordata)] (fig. 1). Comparison of Non–HD Region aa Sequences The N-terminal region aa sequences have been evolutionarily conserved in a wide range of metazoan animals (table 2). The most conserved ones were located at the N-terminal end, and the consensus sequence can be summarized as FSV[D/E]S[L/I][L/I]S, an Engrailed Homology 1 motif (Eh1), termed Eh1N in this paper. The other one was located between the Eh1N and PWM sequences. The generalized sequence motif of this domain was FSV[D/E]GILXK; it was thus similar to Eh1N and was termed Eh1C. Whereas Eh1N was conserved in all species, Eh1C showed scattered distribution among the invertebrate Msx proteins examined. When we mapped the presence of the Eh1C sequence and PWM sequence in the animal groups, it became clear that the Eh1C sequence was retained in the Vertebrata, Cephalochordata, Hemichordata, Echinodermata, Mollusca, and Anthozoa (Cnidaria) but not in the Urochordata and Hydrozoa (Cnidaria) Msx homologs isolated so far. In the Vertebrata, 2–5 paralogues existed in the examined species, and Eh1C-related sequence could be observed at least in 1 paralogue, but not always in the others. An examination of the current NCBI database revealed that, in the Vertebrata, all the examined Msx1 homologs in Xenopus, Ambyostoma, Notophthalamus, Gallus, Mus, Rattus, Bos, and all primate species kept the same Eh1N sequence, LPFSVEALMAD. All the Msx2 homologs in Gallus, Mus, Rattus, Canis, Bos, and all primates had LPFSVEALMSD (data not shown). In addition to the Eh1-related motifs, short conserved sequence stretches were found in the N-terminal region (supplementary fig. 3A, Supplementary Material online) and in C-terminal flanking of the region shown in fig. 1 (supplementary fig. 3B, Supplementary Material online). The conservation of the 2 regions was limited to the sequences from Vertebrata, Cephalochordata, Echinodermata, Hemichordata, Mollusca, and Annelida. However, their functional significance was not clear at this point. Eh1N and Eh1C can Act as Binding Sites for Groucho-Related Proteins Eh1N and Eh1C were very close to the sequences for binding of the transcriptional corepressor Groucho (Paroush et al. 1994; Fisher et al. 1996; Tolkunova et al. 1998). To evaluate the functional significance of the conserved domains in the N-terminal regions, we generated mutant Msx proteins that had alanine substitutions in the conserved Eh1N, Eh1C, and PWM regions and analyzed their Groucho-related protein-binding abilities. For this purpose, expression constructs for FLAG or HA epitope-tagged mouse Groucho-related-gene1 (Grg1), Myc-tagged Xl-Msx1, and FLAG-tagged Nv-Msx were generated and used for a coimmunoprecipitation assay. When these expression vectors were transfected into COS7 cells, we detected each epitope-tagged protein with the expected molecular weight in an immunoblot assay (fig. 6 and data not shown). FLAG- or HA-Grg1 expression vector was then cotransfected with Myc-Xl-Msx1 or HA-Nv-Msx, respectively, and the Grg1 proteins were immunoprecipitated with anti-epitope tag antibodies. Both Myc-Xl-Msx1 and FLAG-Nv-Msx were coprecipitated with Grg1 (fig. 6B and C). FIG. 6.— Open in new tabDownload slide Functional characterization of the conserved domains. (A) Schematic drawing of conserved domains in the Msx proteins and the aa sequences in the mutant and wild-type Msx proteins. Stop(TGA), a truncation mutant generated by inserting a stop codon just after proline 62 in Xl_Msx1 aa sequence (AAH81101); S206A, a substitution mutation in which serine 206 in the Xl_Msx1 was changed into alanine (position of the targeted serine residue is indicated in fig. 1). Substitution mutations in Eh1N, Eh1C, and PWM sequences of Xl_Msx1 or Nv_Msx are indicated in below the diagram. The highly diverged Eh1C sequence Xl_Msx1 (table 2) was not changed. (B) Immunoprecipitation assay between FLAG-Mm-Grg1– and Myc-Xl-Msx1–derived proteins. (C) Immunoprecipitation assay between HA-Mm-Grg1– and FLAG-Nv-Msx–derived proteins. In (B) and (C), the immunoprecipitation experiments were carried out by using cell extracts from COS7 cells cotransfected with Grg1 and Msx expression vectors. The results of immunoblot using the antibodies indicated on the left side of the panels are shown. The combinations of transfected vectors in each cell lysate are indicated at the tops of the pictures. Immunoblot analyses were performed both on the 6% of cell lysates subjected to the immunoprecipitation experiments (IB, 6% input) and the immunoprecipitated (IP) proteins. FIG. 6.— Open in new tabDownload slide Functional characterization of the conserved domains. (A) Schematic drawing of conserved domains in the Msx proteins and the aa sequences in the mutant and wild-type Msx proteins. Stop(TGA), a truncation mutant generated by inserting a stop codon just after proline 62 in Xl_Msx1 aa sequence (AAH81101); S206A, a substitution mutation in which serine 206 in the Xl_Msx1 was changed into alanine (position of the targeted serine residue is indicated in fig. 1). Substitution mutations in Eh1N, Eh1C, and PWM sequences of Xl_Msx1 or Nv_Msx are indicated in below the diagram. The highly diverged Eh1C sequence Xl_Msx1 (table 2) was not changed. (B) Immunoprecipitation assay between FLAG-Mm-Grg1– and Myc-Xl-Msx1–derived proteins. (C) Immunoprecipitation assay between HA-Mm-Grg1– and FLAG-Nv-Msx–derived proteins. In (B) and (C), the immunoprecipitation experiments were carried out by using cell extracts from COS7 cells cotransfected with Grg1 and Msx expression vectors. The results of immunoblot using the antibodies indicated on the left side of the panels are shown. The combinations of transfected vectors in each cell lysate are indicated at the tops of the pictures. Immunoblot analyses were performed both on the 6% of cell lysates subjected to the immunoprecipitation experiments (IB, 6% input) and the immunoprecipitated (IP) proteins. We also tested the physical interaction between Grg1 and mutant Msx1 proteins that contained substitution mutations in their conserved domains (fig. 6A). We prepared 5 expression vectors with substituted mutant Xl-Msx1 (Stop, Eh1N, PWM, Eh1N&PWM, HD-S206A) (fig. 6A). The mutants lacking Eh1N (Eh1N, Eh1N&PWM) lost binding to Grg1 protein, whereas the PWM and HD substitution mutants retained Grg1-binding abilities comparable to those of wild-type Xl-Msx1 (fig. 6B). Because Nv-Msx contained 2 Eh1-like sequences (Eh1N and Eh1C), we generated expression vectors for 4 Nv-Msx mutant proteins, including either single or combined mutants of the Eh1 domains (fig. 6A). Immunoprecipitation experiments showed that a mutation in Eh1N or Eh1C gave a strong or weak decrement, respectively, in Grg1-binding ability (fig. 6C), whereas the PWM mutants were unaffected. The Eh1N and Eh1C combined mutant (NvMsxEh1N&C) completely lacked the ability to bind to Grg1, suggesting that both domains bind to Grg1. Collectively, these results indicated that the Eh1-like domains can be bound by Grg1 protein in both vertebrate and anthozoan animals. Discussion Functional Significance of Msx Conserved Domains Msx1 HD binds specific DNA target sequences (Hovde et al. 2001). In addition, many proteins bind mammalian Msx1 or Msx2 proteins [Lhx1 to HD (Bendall et al. 1998); Dlx to HD (Zhang et al. 1997); Grg to the N-terminal region (Rave-Harel et al. 2005); PIAS to the C-terminal region (Lee et al. 2006); Pax3 (Bendall et al. 1999), Pax9 (Ogawa et al. 2006), and Histone H1b to the N-terminal region (Lee et al. 2004); and Dlxin-1 (Masuda et al. 2001) and TFIIF (RAP74 and RAP30) to the N-terminal region of the HD (Newberry et al. 1997)]. The physical interaction between these molecules may have acted as a form of evolutionary constraint. The overall conservation of the Msx HD suggests that interaction with these molecules is critical for the molecular functions of Msx. At the aa residue level, all the DNA-interacting and TFIIF-interacting residues in the Msx HD were conserved without exception, suggesting that interaction with these molecules, in particular, is essential for many metazoans to survive against natural selection. Other structure–function information can be obtained through an analysis of the point mutations in human MSX1 and MSX2. Some missense mutations in the HD are associated with STA, PFM, and CSO (fig. 1). Besides these HD mutations, it is noteworthy that a mutation in the methionine of Eh1N (L61K in human MSX1) is associated with oligodontia, a congenital form of tooth agenesis (Lidral and Reising 2002). Our results revealed that these residues are strongly conserved not only in primates, as revealed by Perry et al. (2006), but also in the Msx of many metazoans. The characteristic signs of these congenital anomalies suggest that the disease-associated residues are essential, at least for cranial and tooth development in humans. However, the strong conservation of the residues in invertebrate animals suggests that the same sequences can be used in various developmental and/or survival contexts of invertebrates. This view supports an idea that, in many metazoans, the gene encoding Msx protein is among the most fundamental and versatile. Our results revealed that the Eh1N and Eh1C domains are required for the interaction of Msx with Grg1 in mammalian cells. The Eh1 domain was originally identified in the domain mediating the transcriptional repression of Drosophila engrailed protein through interaction with the transcriptional corepressor, Groucho. Although the Eh1-like domains have been identified in various transcription factors (Copley 2005), a recent study showed that Grg1 can physically interact with Msx1 and can regulate an Msx target gene (Rave-Harel et al. 2005). Our study revealed that Grg1 binding was mediated by an Eh1-like domain in Xl-Msx1 protein and that the 2 Eh1-like sequences in anthozoan Nv-Msx were able to bind to Grg1 in mammalian cells. These results suggest that Groucho-like–protein-binding activity is retained in metazoan Msx proteins. Interestingly, the Nematostella genome contains sequences highly similar to Grg1 (H.T. and J.A., unpublished data). Furthermore, vertebrate, Drosophila, and nematoda Groucho-related genes play critical roles in animal development (Pflugrad et al. 1997; Gasperowicz and Otto 2005). It would be interesting to see how putative Msx-binding domains in Groucho family proteins are conserved in metazoans. Collectively, these results suggest that the physical interaction between Msx and Groucho-like protein was established in the eumetazoan ancestor and has been utilized in various ways in the course of evolution. Variation in the Degree of Protein Structure Conservation in the Msx Family We focused here on the conserved domains in Msx proteins based on the phylogeny of species. Phylogenetic trees of the 4 conserved domains that we characterized (Eh1N, Eh1C, PWM, and HD) showed a correlation between the degree of HD sequence divergence and the absence of Eh1C and PWM sequences (fig. 7). In the Hydrozoa and Urochordata, which showed HD sequence divergence in Msx, conservation of the Eh1C or PWM sequences was not evident. In contrast, the HD sequence of Msx was conserved in the Anthozoa, Echinodermata, Cephalochordata, Vertebrata, and Mollusca. HD in the Annelida, Arthropoda, and Nematoda showed an intermediate degree of sequence diversification; these taxa lacked Eh1C and had incomplete PWM sequences. FIG. 7.— Open in new tabDownload slide Evolutionary process of Msx genes. Hypothetical model that explains the diversity of Msx genes in the Eumetazoa. “Thick horizontal bars” indicate AC (absolutely conserved) introns. Dark, light, and intermediate gray boxes indicate conserved type, diverged type, and intermediate type HD + CF regions, respectively; black thick vertical lines indicate Eh1N; gray thick vertical lines indicate Eh1C; striped boxes indicate PWM; and hatched boxes indicate HD. FIG. 7.— Open in new tabDownload slide Evolutionary process of Msx genes. Hypothetical model that explains the diversity of Msx genes in the Eumetazoa. “Thick horizontal bars” indicate AC (absolutely conserved) introns. Dark, light, and intermediate gray boxes indicate conserved type, diverged type, and intermediate type HD + CF regions, respectively; black thick vertical lines indicate Eh1N; gray thick vertical lines indicate Eh1C; striped boxes indicate PWM; and hatched boxes indicate HD. These results allow us to speculate on the overall Msx evolutionary process (fig. 7). Ancestral Msx appeared in the metazoan ancestor, and this gene probably possessed the AC intron. Because of the absence of the N-terminal region of Ef-Msx, we do not know whether the metazoan ancestor was already equipped with all the conserved domains. However, the eumetazoan ancestor, at the latest, may have already contained the Eh1N, Eh1C, PWM, and HD sequences. The 4 conserved sequences diverged differentially in the course of evolution. The degree of divergence was relatively small in the Anthozoa, Echinodermata, Cephalochordata, Vertebrata, and Mollusca but large in the Porifera, Hydrozoa, and Urochordata. There are several potential pitfalls in this hypothesis. First, we are assuming that the diverged-type Msx genes were generated from conserved-type Msx genes. As an opposing idea, convergence to the one prototype sequence cannot be ruled out at this point. Second, roles of possible phylum-specific or class-specific conserved domains remains unclear due to insufficient numbers of sequences for the within-phyla comparison. In this regard, increasing-complexity-style evolution is still possible in Msx family evolution. Third, we might have missed highly diverged, but functionally equivalent, sequences because conservation of the Eh1N, Eh1C, and PWM sequences relies on short sequence stretches. Fourth, we cannot yet conclude that the variable divergence rates reflect general acceleration (or deceleration) of the molecular evolutionary rate or are limited to particular molecular species. Although we utilized the available 18S ribosomal RNA sequences as references, a comprehensive evaluation should be done with additional references. These points may be readily addressed by an extended analysis of additional molecular species. Our hypothesis could also be evaluated by examining whether or not other genes fundamental to the organization of the animal body [so called “tool-kit” genes (Carroll et al. 2001)] show similar evolutionary tendencies. In a previous study, we found that selective loss of the conserved domain in the Zic family is found in certain animal taxa proteins (Aruga et al. 2006). In the case of Zic family proteins, conserved Zic can be seen in the Arthropoda, Mollusca, Annelida, Echinodermata, and Chordata (vertebrates and cephalochordates), whereas diverged Zic can be seen in the Platyhelminthes, Cnidaria, Nematoda, and Chordata (urochordates). On the basis of the phylogenetic distribution of the conserved Zic proteins, we proposed that the bilaterian ancestors had already acquired the full set of conserved domains that is found in currently living animals. Thus, in the cases of both the Msx and the Zic family proteins, the ancestral genes may already have possessed a set of conserved domains that were selectively lost in the course of evolution. This differential divergence rate among the taxa could be involved in diversification of the organization of the animal body. However, it is premature to conclude this because our understanding of the role of the Msx genes in each animal taxa is very limited. If we consider the Cnidaria sister groups, Anthozoa and Hydrozoa, hydrozoan Msx is expressed in regenerating muscle tissues (Yanze et al. 1999; Galle et al. 2005) and anthozoan Msx in the planula larval ectoderm (de Jong et al. 2006). However, because these reports dealt with expression profiles in only limited stages of the animals' life cycles, we are deterred from considering further the similarities and differences between Msx usage in the two classes of animals. If we consider the chordate Msx homologs, ascidian Msx genes appear to be commonly expressed in muscle, notochord, neural plate precursor cells, and the folding neural plate (Ma et al. 1996; Aniello et al. 1999). Some of these expression sites overlap with those of Msx gene expression in vertebrates (Davidson 1995), suggesting that some of the roles of Msx are shared between ascidians and vertebrates. However, vertebrate Msx genes are expressed in the limb bud, mandibular process, tooth, uterus, and other many organs that are not apparently found in ascidians. It is possible that the negative pressures for structural diversification differ between ascidians and vertebrates. Finally, our results revealed the differential diversification of the Msx protein functional domain that became established in the eumetazoan ancestor in the course of evolution. The similarity of this evolution to that of Zic family proteins raises the possibility that a group of “tool-kit” genes shared the same feature, that is, differential diversification of the conserved domain. We await further extension of the metazoan-wide phylogenetic analysis of developmentally critical genes. Supplementary Material Supplementary text, figures 1–3, and tables 1 and 2 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/). Database deposition: The sequences reported in the paper have been deposited in the GenBank/EMBL/DDBJ database accession numbers. AB302953–AB302969, AB362783–AB362785. We thank Yayoi Nozaki and all the technical staff of the Sequence Technology Team of RIKEN GSC at the Research Resource Center, RIKEN BSI, for their technical assistance. We also thank Atsushi Suzuki (Hiroshima University) for the plasmid. This study was done as a collaborative research project of the Strategic Research Program at RIKEN. It was supported by grants-in-aid from the Japanese Ministry of Education, Science, Sports, and Technology. References Altschul SF , Gish W , Miller W , Myers EW , Lipman DJ . Basic local alignment search tool , J Mol Biol , 1990 , vol. 215 (pg. 403 - 410 ) Google Scholar Crossref Search ADS PubMed WorldCat Altschul SF , Madden TL , Schaffer AA , Zhang J , Zhang Z , Miller W , Lipman DJ . Gapped BLAST and PSI-BLAST: a new generation of protein database search programs , Nucleic Acids Res , 1997 , vol. 25 (pg. 3389 - 3402 ) Google Scholar Crossref Search ADS PubMed WorldCat Aniello F , Locascio A , Villani MG , Di Gregorio A , Fucci L , Branno M . Identification and developmental expression of Ci-msxb: a novel homologue of Drosophila msh gene in Ciona intestinalis , Mech Dev , 1999 , vol. 88 (pg. 123 - 126 ) Google Scholar Crossref Search ADS PubMed WorldCat Arendt D , Nubler-Jung K . Comparison of early nerve cord development in insects and vertebrates , Development , 1999 , vol. 126 (pg. 2309 - 2325 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Aruga J , Kamiya A , Takahashi H , et al. (15 co-authors) A wide-range phylogenetic analysis of Zic proteins: implications for correlations between protein structure conservation and body plan complexity , Genomics , 2006 , vol. 87 (pg. 783 - 792 ) Google Scholar Crossref Search ADS PubMed WorldCat Bendall AJ , Abate-Shen C . Roles for Msx and Dlx homeoproteins in vertebrate development , Gene , 2000 , vol. 247 (pg. 17 - 31 ) Google Scholar Crossref Search ADS PubMed WorldCat Bendall AJ , Ding J , Hu G , Shen MM , Abate-Shen C . Msx1 antagonizes the myogenic activity of Pax3 in migrating limb muscle precursors , Development , 1999 , vol. 126 (pg. 4965 - 4976 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Bendall AJ , Rincon-Limas DE , Botas J , Abate-Shen C . Protein complex formation between Msx1 and Lhx2 homeoproteins is incompatible with DNA binding activity , Differentiation , 1998 , vol. 63 (pg. 151 - 157 ) Google Scholar Crossref Search ADS PubMed WorldCat Bleidorn C , Vogt L , Bartolomaeus T . New insights into polychaete phylogeny (Annelida) inferred from 18S rDNA sequences , Mol Phylogenet Evol , 2003 , vol. 29 (pg. 279 - 288 ) Google Scholar Crossref Search ADS PubMed WorldCat Brusca RC , Brusca GJ . The emergence of the Arthropods, Onychophorans, Tardigrades, Trilobites, and the Arthropod bauplan , Invertebrates , 2003 Sunderland (MA) Sinauer (pg. 461 - 510 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Cai W , Pei J , Grishin NV . Reconstruction of ancestral protein sequences and its applications , BMC Evol Biol , 2004 , vol. 4 pg. 33 Google Scholar Crossref Search ADS PubMed WorldCat Cameron CB , Garey JR , Swalla BJ . Evolution of the chordate body plan: new insights from phylogenetic analyses of deuterostome phyla , Proc Natl Acad Sci USA , 2000 , vol. 97 (pg. 4469 - 4474 ) Google Scholar Crossref Search ADS PubMed WorldCat Carninci P , Kasukawa T , Katayama S , et al. (194 co-authors) The transcriptional landscape of the mammalian genome , Science , 2005 , vol. 309 (pg. 1559 - 1563 ) Google Scholar Crossref Search ADS PubMed WorldCat Carroll SB , Grenier JK , Weatherbee SD . , From DNA to diversity, molecular genetics and the evolution of animal design , 2001 Oxford Blackwell Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Copley RR . The EH1 motif in metazoan transcription factors , BMC Genomics , 2005 , vol. 6 pg. 169 Google Scholar Crossref Search ADS PubMed WorldCat Davidson D . The function and evolution of Msx genes: pointers and paradoxes , Trends Genet , 1995 , vol. 11 (pg. 405 - 411 ) Google Scholar Crossref Search ADS PubMed WorldCat Dayhoff MO , Schwarz RM , Orcutt BC . Dayhoff MO . A model of evolutionary change in proteins , Atlas of protein sequence and structure , 1978 Silver Spring (MD) National Biomedical Research Foundation (pg. 342 - 352 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC de Jong DM , Hislop NR , Hayward DC , Reece-Hoyes JS , Pontynen PC , Ball EE , Miller DJ . Components of both major axial patterning systems of the Bilateria are differentially expressed along the primary axis of a ‘radiate’ animal, the anthozoan cnidarian Acropora millepora , Dev Biol , 2006 , vol. 298 (pg. 632 - 643 ) Google Scholar Crossref Search ADS PubMed WorldCat Du H , Chalfie M . Genes regulating touch cell development in Caenorhabditis elegans , Genetics , 2001 , vol. 158 (pg. 197 - 207 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Ekker M , Akimenko MA , Allende ML , Smith R , Drouin G , Langille RM , Weinberg ES , Westerfield M . Relationships among msx gene structure and function in zebrafish and other vertebrates , Mol Biol Evol , 1997 , vol. 14 (pg. 1008 - 1022 ) Google Scholar Crossref Search ADS PubMed WorldCat Felsenstein J . Confidence limits on phylogenies, an approach using the bootstrap , Evolution , 1985 , vol. 39 (pg. 783 - 791 ) Google Scholar Crossref Search ADS WorldCat Fisher AL , Ohsako S , Caudy M . The WRPW motif of the hairy-related basic helix-loop-helix repressor proteins acts as a 4-amino-acid transcription repression and protein-protein interaction domain , Mol Cell Biol , 1996 , vol. 16 (pg. 2670 - 2677 ) Google Scholar Crossref Search ADS PubMed WorldCat Galle S , Yanze N , Seipel K . The homeobox gene Msx in development and transdifferentiation of jellyfish striated muscle , Int J Dev Biol , 2005 , vol. 49 (pg. 961 - 967 ) Google Scholar Crossref Search ADS PubMed WorldCat Gasperowicz M , Otto F . Mammalian Groucho homologs: redundancy or specificity? , J Cell Biochem , 2005 , vol. 95 (pg. 670 - 687 ) Google Scholar Crossref Search ADS PubMed WorldCat Gauchat D , Mazet F , Berney C , Schummer M , Kreger S , Pawlowski J , Galliot B . Evolution of Antp-class genes and differential expression of Hydra Hox/paraHox genes in anterior patterning , Proc Natl Acad Sci USA , 2000 , vol. 97 (pg. 4493 - 4498 ) Google Scholar Crossref Search ADS PubMed WorldCat Holland PW . Cloning and evolutionary analysis of msh-like homeobox genes from mouse, zebrafish and ascidian , Gene , 1991 , vol. 98 (pg. 253 - 257 ) Google Scholar Crossref Search ADS PubMed WorldCat Holland PW , Takahashi T . The evolution of homeobox genes: implications for the study of brain development , Brain Res Bull , 2005 , vol. 66 (pg. 484 - 490 ) Google Scholar Crossref Search ADS PubMed WorldCat Hovde S , Abate-Shen C , Geiger JH . Crystal structure of the Msx-1 homeodomain/DNA complex , Biochemistry , 2001 , vol. 40 (pg. 12013 - 12021 ) Google Scholar Crossref Search ADS PubMed WorldCat Hu G , Vastardis H , Bendall AJ , et al. (12 co-authors) Haploinsufficiency of MSX1: a mechanism for selective tooth agenesis , Mol Cell Biol , 1998 , vol. 18 (pg. 6044 - 6051 ) Google Scholar Crossref Search ADS PubMed WorldCat Huelsenbeck JP , Ronquist F . MRBAYES: Bayesian inference of phylogenetic trees , Bioinformatics , 2001 , vol. 17 (pg. 754 - 755 ) Google Scholar Crossref Search ADS PubMed WorldCat Ishiguro A , Ideta M , Mikoshiba K , Chen DJ , Aruga J . Zic2-dependent transcriptional regulation is mediated by DNA-dependent protein kinase, poly-ADP ribose polymerase, and RNA helicase A , J Biol Chem , 2007 , vol. 282 (pg. 9983 - 9995 ) Google Scholar Crossref Search ADS PubMed WorldCat Isshiki T , Takeichi M , Nose A . The role of the msh homeobox gene during Drosophila neurogenesis: implication for the dorsoventral specification of the neuroectoderm , Development , 1997 , vol. 124 (pg. 3099 - 3109 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Ito W , Ishiguro H , Kurosawa Y . A general method for introducing a series of mutations into cloned DNA using the polymerase chain reaction , Gene , 1991 , vol. 102 (pg. 67 - 70 ) Google Scholar Crossref Search ADS PubMed WorldCat Jabs EW , Muller U , Li X , et al. (12 co-authors) A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis , Cell , 1993 , vol. 75 (pg. 443 - 450 ) Google Scholar Crossref Search ADS PubMed WorldCat Jumlongras D , Bei M , Stimson JM , Wang WF , DePalma SR , Seidman CE , Felbor U , Maas R , Seidman JG , Olsen BR . A nonsense mutation in MSX1 causes Witkop syndrome , Am J Hum Genet , 2001 , vol. 69 (pg. 67 - 74 ) Google Scholar Crossref Search ADS PubMed WorldCat Khadka D , Luo T , Sargent TD . Msx1 and Msx2 have shared essential functions in neural crest but may be dispensable in epidermis and axis formation in Xenopus , Int J Dev Biol , 2006 , vol. 50 (pg. 499 - 502 ) Google Scholar Crossref Search ADS PubMed WorldCat Kumar S , Tamura K , Nei M . MEGA3: integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment , Brief Bioinform , 2004 , vol. 5 (pg. 150 - 163 ) Google Scholar Crossref Search ADS PubMed WorldCat Lallemand Y , Nicola MA , Ramos C , Bach A , Cloment CS , Robert B . Analysis of Msx1; Msx2 double mutants reveals multiple roles for Msx genes in limb development , Development , 2005 , vol. 132 (pg. 3003 - 3014 ) Google Scholar Crossref Search ADS PubMed WorldCat Lee H , Habas R , Abate-Shen C . MSX1 cooperates with histone H1b for inhibition of transcription and myogenesis , Science , 2004 , vol. 304 (pg. 1675 - 1678 ) Google Scholar Crossref Search ADS PubMed WorldCat Lee H , Quinn JC , Prasanth KV , Swiss VA , Economides KD , Camacho MM , Spector DL , Abate-Shen C . PIAS1 confers DNA-binding specificity on the Msx1 homeoprotein , Genes Dev , 2006 , vol. 20 (pg. 784 - 794 ) Google Scholar Crossref Search ADS PubMed WorldCat Lidral AC , Reising BC . The role of MSX1 in human tooth agenesis , J Dent Res , 2002 , vol. 81 (pg. 274 - 278 ) Google Scholar Crossref Search ADS PubMed WorldCat Ma L , Swalla BJ , Zhou J , Dobias SL , Bell JR , Chen J , Maxson RE , Jeffery WR . Expression of an Msx homeobox gene in ascidians: insights into the archetypal chordate expression pattern , Dev Dyn , 1996 , vol. 205 (pg. 308 - 318 ) Google Scholar Crossref Search ADS PubMed WorldCat Master VA , Kourakis MJ , Martindale MQ . Isolation, characterization, and expression of Le-msx, a maternally expressed member of the msx gene family from the glossiphoniid leech, Helobdella , Dev Dyn , 1996 , vol. 207 (pg. 404 - 419 ) Google Scholar Crossref Search ADS PubMed WorldCat Masuda Y , Sasaki A , Shibuya H , Ueno N , Ikeda K , Watanabe K . Dlxin-1, a novel protein that binds Dlx5 and regulates its transcriptional function , J Biol Chem , 2001 , vol. 276 (pg. 5331 - 5338 ) Google Scholar Crossref Search ADS PubMed WorldCat McHugh D . Molecular evidence that echiurans and pogonophorans are derived annelids , Proc Natl Acad Sci USA , 1997 , vol. 94 (pg. 8006 - 8009 ) Google Scholar Crossref Search ADS PubMed WorldCat Monsoro-Burq AH , Wang E , Harland R . Msx1 and Pax3 cooperate to mediate FGF8 and WNT signals during Xenopus neural crest induction , Dev Cell , 2005 , vol. 8 (pg. 167 - 178 ) Google Scholar Crossref Search ADS PubMed WorldCat Nose A , Isshiki T , Takeichi M . Regional specification of muscle progenitors in Drosophila: the role of the msh homeobox gene , Development , 1998 , vol. 125 (pg. 215 - 223 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Ogawa T , Kapadia H , Feng JQ , Raghow R , Peters H , D'Souza RN . Functional consequences of interactions between Pax9 and Msx1 genes in normal and abnormal tooth development , J Biol Chem , 2006 , vol. 281 (pg. 18363 - 18369 ) Google Scholar Crossref Search ADS PubMed WorldCat Paroush Z , Finley RL Jr , Kidd T , Wainwright SM , Ingham PW , Brent R , Ish-Horowicz D . Groucho is required for Drosophila neurogenesis, segmentation, and sex determination and interacts directly with hairy-related bHLH proteins , Cell , 1994 , vol. 79 (pg. 805 - 815 ) Google Scholar Crossref Search ADS PubMed WorldCat Perry GH , Verrelli BC , Stone AC . Molecular evolution of the primate developmental genes MSX1 and PAX9 , Mol Biol Evol , 2006 , vol. 23 (pg. 644 - 654 ) Google Scholar Crossref Search ADS PubMed WorldCat Peterson KJ . Isolation of Hox and Parahox genes in the hemichordate Ptychodera flava and the evolution of deuterostome Hox genes , Mol Phylogenet Evol , 2004 , vol. 31 (pg. 1208 - 1215 ) Google Scholar Crossref Search ADS PubMed WorldCat Pflugrad A , Meir JY , Barnes TM , Miller DM 3rd . The Groucho-like transcription factor UNC-37 functions with the neural specificity gene unc-4 to govern motor neuron identity in C. elegans , Development , 1997 , vol. 124 (pg. 1699 - 1709 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Postlethwait JH . The zebrafish genome: a review and case study of msx genes , Genome Dyn , 2006 , vol. 2 (pg. 183 - 197 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Ramos C , Robert B . msh/Msx gene family in neural development , Trends Genet , 2005 , vol. 21 (pg. 624 - 632 ) Google Scholar Crossref Search ADS PubMed WorldCat Rave-Harel N , Miller NL , Givens ML , Mellon PL . The Groucho-related gene family regulates the gonadotropin-releasing hormone gene through interaction with the homeodomain proteins MSX1 and OCT1 , J Biol Chem , 2005 , vol. 280 (pg. 30975 - 30983 ) Google Scholar Crossref Search ADS PubMed WorldCat Ronquist F , Huelsenbeck JP . MrBayes 3: Bayesian phylogenetic inference under mixed models , Bioinformatics , 2003 , vol. 19 (pg. 1572 - 1574 ) Google Scholar Crossref Search ADS PubMed WorldCat Satokata I , Ma L , Ohshima H , et al. (14 co-authors) Msx2 deficiency in mice causes pleiotropic defects in bone growth and ectodermal organ formation , Nat Genet , 2000 , vol. 24 (pg. 391 - 395 ) Google Scholar Crossref Search ADS PubMed WorldCat Satokata I , Maas R . Msx1 deficient mice exhibit cleft palate and abnormalities of craniofacial and tooth development , Nat Genet , 1994 , vol. 6 (pg. 348 - 356 ) Google Scholar Crossref Search ADS PubMed WorldCat Schmidt HA , Strimmer K , Vingron M , von Haeseler A . TREE-PUZZLE: maximum likelihood phylogenetic analysis using quartets and parallel computing , Bioinformatics , 2002 , vol. 18 (pg. 502 - 504 ) Google Scholar Crossref Search ADS PubMed WorldCat Seimiya M , Ishiguro H , Miura K , Watanabe Y , Kurosawa Y . Homeobox-containing genes in the most primitive metazoa, the sponges , Eur J Biochem , 1994 , vol. 221 (pg. 219 - 225 ) Google Scholar Crossref Search ADS PubMed WorldCat Smith TF , Waterman MS . Identification of common molecular subsequences , J Mol Biol , 1981 , vol. 147 (pg. 195 - 197 ) Google Scholar Crossref Search ADS PubMed WorldCat Suzuki A , Ueno N , Hemmati-Brivanlou A . Xenopus msx1 mediates epidermal induction and neural inhibition by BMP4 , Development , 1997 , vol. 124 (pg. 3037 - 3044 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Suzuki AC . Life history of Milnesium tardigradum Doyére (tardigrada) under a rearing environment , Zool Sci , 2003 , vol. 20 (pg. 49 - 57 ) Google Scholar Crossref Search ADS PubMed WorldCat Tamura K , Nei M . Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees , Mol Biol Evol , 1993 , vol. 10 (pg. 512 - 526 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Tan H , Ransick A , Wu H , Dobias S , Liu YH , Maxson R . Disruption of primary mesenchyme cell patterning by misregulated ectodermal expression of SpMsx in sea urchin embryos , Dev Biol , 1998 , vol. 201 (pg. 230 - 246 ) Google Scholar Crossref Search ADS PubMed WorldCat Thompson JD , Higgins DG , Gibson TJ . CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice , Nucleic Acids Res , 1994 , vol. 22 (pg. 4673 - 4680 ) Google Scholar Crossref Search ADS PubMed WorldCat Tolkunova EN , Fujioka M , Kobayashi M , Deka D , Jaynes JB . Two distinct types of repression domain in engrailed: one interacts with the groucho corepressor and is preferentially active on integrated target genes , Mol Cell Biol , 1998 , vol. 18 (pg. 2804 - 2814 ) Google Scholar Crossref Search ADS PubMed WorldCat Toyoda A , Noguchi H , Taylor TD , Ito T , Pletcher MT , Sakaki Y , Reeves RH , Hattori M . Comparative genomic sequence analysis of the human chromosome 21 down syndrome critical region , Genome Res , 2002 , vol. 12 (pg. 1323 - 1332 ) Google Scholar Crossref Search ADS PubMed WorldCat Turner DL , Weintraub H . Expression of achaete-scute homolog 3 in Xenopus embryos converts ectodermal cells to a neural fate , Genes Dev , 1994 , vol. 8 (pg. 1434 - 1447 ) Google Scholar Crossref Search ADS PubMed WorldCat van den Boogaard MJ , Dorland M , Beemer FA , van Amstel HK . MSX1 mutation is associated with orofacial clefting and tooth agenesis in humans , Nat Genet , 2000 , vol. 24 (pg. 342 - 343 ) Google Scholar Crossref Search ADS PubMed WorldCat Vastardis H , Karimbux N , Guthua SW , Seidman JG , Seidman CE . A human MSX1 homeodomain missense mutation causes selective tooth agenesis , Nat Genet , 1996 , vol. 13 (pg. 417 - 421 ) Google Scholar Crossref Search ADS PubMed WorldCat Walldorf U , Fleig R , Gehring WJ . Comparison of homeobox-containing genes of the honeybee and Drosophila , Proc Natl Acad Sci USA , 1989 , vol. 86 (pg. 9971 - 9975 ) Google Scholar Crossref Search ADS PubMed WorldCat Wheeler SR , Carrico ML , Wilson BA , Skeath JB . The Tribolium columnar genes reveal conservation and plasticity in neural precursor patterning along the embryonic dorsal-ventral axis , Dev Biol , 2005 , vol. 279 (pg. 491 - 500 ) Google Scholar Crossref Search ADS PubMed WorldCat Whelan S , Goldman N . A general empirical model of protein evolution derived from multiple protein families using a maximum-likelihood approach , Mol Biol Evol , 2001 , vol. 18 (pg. 691 - 699 ) Google Scholar Crossref Search ADS PubMed WorldCat Wilkie AO , Tang Z , Elanko N , Walsh S , Twigg SR , Hurst JA , Wall SA , Chrzanowska KH , Maxson RE Jr . Functional haploinsufficiency of the human homeobox gene MSX2 causes defects in skull ossification , Nat Genet , 2000 , vol. 24 (pg. 387 - 390 ) Google Scholar Crossref Search ADS PubMed WorldCat Wilson KA , Andrews ME , Rudolf Turner F , Raff RA . Major regulatory factors in the evolution of development: the roles of goosecoid and Msx in the evolution of the direct-developing sea urchin Heliocidaris erythrogramma , Evol Dev , 2005 , vol. 7 (pg. 416 - 428 ) Google Scholar Crossref Search ADS PubMed WorldCat Wuyts W , Reardon W , Preis S , Homfray T , Rasore-Quartino A , Christians H , Willems PJ , Van Hul W . Identification of mutations in the MSX2 homeobox gene in families affected with foramina parietalia permagna , Hum Mol Genet , 2000 , vol. 9 (pg. 1251 - 1255 ) Google Scholar Crossref Search ADS PubMed WorldCat Yanze N , Groger H , Muller P , Schmid V . Reversible inactivation of cell-type-specific regulatory and structural genes in migrating isolated striated muscle cells of jellyfish , Dev Biol , 1999 , vol. 213 (pg. 194 - 201 ) Google Scholar Crossref Search ADS PubMed WorldCat Zhang H , Catron KM , Abate-Shen C . A role for the Msx-1 homeodomain in transcriptional regulation: residues in the N-terminal arm mediate TATA binding protein interaction and transcriptional repression , Proc Natl Acad Sci USA , 1996 , vol. 93 (pg. 1764 - 1769 ) Google Scholar Crossref Search ADS PubMed WorldCat Zhang H , Hu G , Wang H , Sciavolino P , Iler N , Shen MM , Abate-Shen C . Heterodimerization of Msx and Dlx homeoproteins results in functional antagonism , Mol Cell Biol , 1997 , vol. 17 (pg. 2920 - 2932 ) Google Scholar Crossref Search ADS PubMed WorldCat Author notes William Martin, Associate Editor © 2007 The Authors. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2007 The Authors.

Basu, Malay Kumar; Rogozin, Igor B.; Deusch, Oliver; Dagan, Tal; Martin, William; Koonin, Eugene V.

doi: 10.1093/molbev/msm234pmid: 17974547

Abstract Some of the principal transitions in the evolution of eukaryotes are characterized by engulfment of prokaryotes by primitive eukaryotic cells. In particular, ∼1.6 billion years ago, engulfment of a cyanobacterium that became the ancestor of chloroplasts and other plastids gave rise to Plantae, the major branch of eukaryotes comprised of glaucophytes, red algae, green algae, and green plants. After endosymbiosis, there was large-scale migration of genes from the endosymbiont to the nuclear genome of the host such that ∼18% of the nuclear genes in Arabidopsis appear to be of chloroplast origin. To gain insights into the process of evolution of gene structure in these, originally, intronless genes, we compared the properties and the evolutionary dynamics of introns in genes of plastid origin and ancestral eukaryotic genes in Arabidopsis, poplar, and rice genomes. We found that intron densities in plastid-derived genes were slightly but significantly lower than those in ancestral eukaryotic genes. Although most of the introns in both categories of genes were conserved between monocots (rice) and dicots (Arabidopsis and poplar), lineage-specific intron gain was more pronounced in plastid-derived genes than in ancestral genes, whereas there was no significant difference in the intron loss rates between the 2 classes of genes. Thus, after the transfer to the nuclear genome, the plastid-derived genes have undergone a massive intron invasion that, by the time of the divergence of dicots and monocots (150–200 MYA), yielded intron densities only slightly lower than those in ancestral genes. Nevertheless, the accumulation of introns in plastid-derived genes appears not to have reached saturation and continues to this time, albeit at a low rate. The overall pattern of intron gain and loss in the plastid-derived genes is shaped by this continuing gain and the more general tendency for loss that is characteristic of the recent evolution of plant genes. intron gain, intron loss, plant evolution, plastid-derived genes Introduction Multiple spliceosomal introns interrupting protein-coding genes and the concurrent splicing machinery are among the defining features of eukaryotes (Doolittle 1978; Gilbert 1978; Mattick 1994; Deutsch and Long 1999). Origin and evolution of introns are often considered within the context of the debate between the introns-early and introns-late concepts, a conundrum that emerged shortly after the discovery of the exon–intron organization of eukaryotic genes (Darnell 1978; Doolittle 1978; Logsdon and Palmer 1994; Logsdon 1998). The introns-early hypothesis (Darnell 1978; Doolittle 1978; Gilbert 1978, 1987; Gilbert and Glynias 1993; Gilbert et al. 1997) (appearing also in the more recent guise of “introns first”[Jeffares et al. 2006]) holds that introns were an intrinsic element of the first protein-coding genes and actually facilitated the origin of proteins via recombination; under this view, the absence of spliceosomal introns in prokaryotes is attributed to complete loss of introns in the course of “genome streamlining.” The introns-late hypothesis counters that prokaryotes never had spliceosomal introns and that the introns and the spliceosome emerged during eukaryotic evolution (Logsdon and Palmer 1994; Stoltzfus 1994; Stoltzfus et al. 1994; Logsdon et al. 1995; Logsdon 1998). At present, despite extensive efforts to identify traces of primordial introns, there seems to be no substantial empirical evidence in support of introns-early, and consequently, spliceosomal introns and the spliceosome are generally considered a key, early eukaryotic innovation (Stoltzfus et al. 1994; Cho and Doolittle 1997; Logsdon et al. 1998; Belshaw and Bensasson 2006; Koonin 2006). Introns and the spliceosome are among those fundamental features of the eukaryotic cell (along with the nucleus, the cytoskeleton, and several others) that, beyond reasonable doubt, were already present in the last eukaryotic common ancestor. The origin of each of these innovations and the earliest steps in their evolution are shrouded in mystery, given that they are present in their (nearly) fully developed form in all extant eukaryotes (Mans et al. 2004; Collins and Penny 2005). The eukaryotic spliceosomal introns as well as the spliceosomal small RNAs that perform critical functions in splicing show significant similarities to self-splicing group II introns (retroelements), found in bacteria and organelles and are thought to have evolved from such elements (Zimmerly et al. 2001; Lambowitz and Zimmerly 2004; Toro et al. 2007). However, the process of the acquisition of spliceosomal introns by eukaryotic genomes is not understood beyond general schemes (Koonin 2006; Martin and Koonin 2006). The dynamics and mechanisms of intron evolution are not well understood either. Generally, the abundance and turnover rate of introns in a genome are thought to be determined by the effective population size and the characteristic mutation rate of the respective species (Lynch and Richardson 2002; Lynch and Conery 2003); however, it has been argued that various selective forces could substantially affect the rates of intron gain and loss (Jeffares et al. 2006). Comparative genomic studies have revealed impressive conservation of intron positions in diverse animals (Raible et al. 2005) and have shown that the positions of many introns are shared by orthologous genes even in distant eukaryotes, such as animals and plants (Fedorov et al. 2002; Rogozin et al. 2003). However, several recent models of the processes of intron gain and loss have yielded widely contradicting scenarios. Some models lead to an evolutionary landscape dominated by intron gain (Qiu et al. 2004), whereas others suggest dominance of intron loss (Roy and Gilbert 2005a, 2005b) or offer intermediate solutions, which imply that intron gain and loss both made important and comparable contribution to the evolution of eukaryotic genes (Rogozin et al. 2003; Csuros 2005; Nguyen et al. 2005). A recent, comprehensive maximum likelihood reconstruction of intron gain and loss in eukaryotic evolution not only supports the latter point of view but also reveals a significant (∼1.8-fold) overall excess of losses and a distinctly nonuniform distribution of gains and losses over the history of eukaryotes (Carmel et al. 2007b). In particular, substantial excess of intron gains over intron losses has been detected only for some of the early intervals of eukaryotic evolution that are associated with major evolutionary innovations, such as the origin of animals. Somewhat unexpectedly, it is extremely hard to detect concrete evidence of intron gain events. It would seem plausible that new introns, at least, in part, are gained via transposition of preexisting introns; however, with a single exception noted below, the attempts to detect homologous introns in different genes within the same genome so far have failed (Fedorov et al. 2003). In complementary studies, comparison of orthologous genes from mammalian genomes failed to reveal any gains at all, suggesting that all introns currently contained in mammalian genes were already present at the time of the radiation of mammalian orders (Roy et al. 2003), and similar results have been reported in a detailed analysis of several plant gene families (Teich et al. 2007). A reconstruction of intron gain and loss in paralogous gene families from animals and plants similarly indicated no significant gain during the last several hundred million years of evolution (Babenko et al. 2004). Recently, a focused attempt was undertaken to identify intron gain in genes of the protist Entamoeba histolytica that appear to have been acquired via horizontal gene transfer from bacteria; any introns present in these genes would be relatively recent gains (Roy et al. 2006). However, Roy et al. detected very few introns in the horizontally transferred genes, indicating a low gain rate. About the only purported showcase of intron gain is the report on an apparent gain of 122 introns in the nematodes of the genus Caenorhabditis since the divergence of the 2 nematodes with sequenced genomes, Caenorhabditis elegans and Caenorhabditis briggsae (Coghlan and Wolfe 2004). However, a recent reanalysis has suggested that most of these purported intron gains might actually represent losses of ancestral introns (Roy and Penny 2006). In addition to the virtual lack of comparative genomic evidence in support of recent intron gain, the mechanisms of new intron insertion into genes are not understood; this contrasts the case of intron loss that is widely believed to be mediated, primarily, by recombination between intron-containing genes and intronless cDNA copies of the corresponding mRNAs (Fink 1987; Mourier and Jeffares 2003). Given the elusiveness of intron gain, evolutionary scenarios are seriously considered under which most introns that are currently present in eukaryotic genes have been inserted at very early stages of the evolution of eukaryotes, whereas subsequent evolution had been dominated by intron loss (Roy and Gilbert 2005a, 2005b; Roy 2006). However, the logic of the aforementioned study of introns in the relatively few horizontally transferred genes of Entamoeba (Roy et al. 2006) applies also to more momentous events in the evolution of eukaryotes. In particular, plants have been shown to carry thousands of genes that have been transferred from the genome of the cyanobacterial endosymbiont to the nuclear genome (Martin et al. 1998, 2002; Timmis et al. 2004). There is no doubt that, at the time of the transfer, each of these genes was intronless, so any introns found in genes of plastid (chloroplast) derivation would result from de novo insertion (gain), even if occurring soon after the symbiosis event. In an attempt to infer the salient features of the intron gain process, we compared the characteristics of introns found in plastid-derived and ancestral eukaryotic genes present in plant genomes. We found that plastid-derived genes have a slightly but significantly lower intron density than ancestral eukaryotic genes in the same plants and that the process of intron acquisition leading to the increase of intron density in plastid-derived genes is still ongoing, albeit at a low rate. Materials and Methods Identification of Orthologous Genes Rice genome annotation and sequences were downloaded from Rice Annotation Project DataBase (http://rapdb.lab.nig.ac.jp/), Arabidopsis genome annotation and sequences were from National Center for Biotechnology Information (NCBI) (ftp://ftp.ncbi.nih.gov/genomes/), and poplar sequences were from the Joint Genome Institute (http://genome.jgi-psf.org/Poptr1_1/Poptr1_1.home.html). Orthologs in rice, poplar, and Arabidopsis orthologs were identified using a modified InParanoid algorithm (O'Brien et al. [2005] and Basu MK, unpublished data). Pairs of orthologous proteins that differed by >25% in length were discarded from the analysis to avoid the potential impact of inaccurate gene annotations. Altogether, ∼2,500 clusters containing exactly 1 orthologous gene from Arabidopsis, poplar, and rice each were identified. Each of the 3 pairwise comparisons between plant species yielded a greater number of orthologs because of lineage-specific duplications, particularly, the whole-genome duplication in poplar (Tuskan et al. 2006). The Arabidopsis–poplar comparison produced ∼7,000 orthologous gene pairs, the Arabidopsis–rice comparison ∼5,500 pairs, and the poplar–rice comparison ∼5,000 pairs. The clusters of orthologous plant genes were partitioned into 3 classes, namely, chloroplast derived, ancestral eukaryotic, and plant specific. The plastid-derived genes were identified by phylogenetic affinity to cyanobacterial genes using a modification of the previously described procedure (Martin et al. 2002). Specifically, each Arabidopsis protein sequence was compared with a data set that included proteins from 200 eubacteria, 24 archaebacteria, and 13 nonphotosynthetic eukaryotes (extracted from the NCBI Genome database, National Institutes of Health [NIH], Bethesda, MD; supplementary table S1, Supplementary Material online) using the BlastP program. The Blast hits with an E value >10−10 and <25% amino acid identity were discarded, and the remaining, highly significant hits were ranked by the percent identities normalized by the fraction of aligned residues. The best hit from each phylum was selected for further analysis. The Arabidopsis protein sequences were aligned with the homologous sequences from other species using ClustalW (Thompson et al. 1994), and sites containing gaps were removed. Phylogenetic trees were built with the Neighbor-Joining using the PROTDIST program of the PHYLIP package (Felsenstein 1996) with the Jones-Thornton-Taylor substitution matrix for the calculation of protein distances and the Neighbor tree reconstruction. The trees in Newick format were parsed using a PERL script, and Arabidopsis genes that had a cyanobacterial homolog as a sister taxon were defined as chloroplast-derived genes. Arabidopsis genes that had homologues only in cyanobacterial genomes were classified as chloroplast-derived genes as well. The plastid-derived genes in poplar and rice were defined as orthologs of the plastid-derived genes of Arabidopsis that were identified in pairwise genome comparisons. The rest of the clusters were classified as ancestral or plant specific using the following criteria. The Arabidopsis proteins from the uncategorized clusters were compared with the nonredundant protein sequence database (NCBI, NIH) using the BlastP program with the composition-based statistical adjustment (Altschul et al. 1997; Schaffer et al. 2001). All clusters containing proteins with highly significant sequence similarity (E value < 10−10) to at least 1 homologous protein in both fungi and metazoa were classified as ancestral. In addition, a subset of more liberally defined ancestral genes was identified using a relaxed criterion (E value < 10−4). The remaining clusters, that is, those for which, under the strict or the relaxed criterion, no significant similarity to nonplant proteins was detected, were classified as plant specific. This procedure partitioned the identified single-gene orthologous clusters from 3 plants into 304 chloroplast-derived, 1,104 (1,279) ancestral, and 545 (650) plant-specific clusters (numbers obtained with the relaxed criterion given in parentheses). All the analyses described below involved comparisons between chloroplast-derived and ancestral genes; the plant-specific genes were disregarded. Conservation of Intron Positions Intron conservation was determined, essentially, as previously described (Rogozin et al. 2003). Amino acid sequences of orthologous plant proteins were aligned using the MUSCLE program (Edgar 2004), and intron positions were then mapped onto the alignments. All the positions that contained 1 or more gaps within 5 amino acid upstream and downstream were discarded from the analysis. Intron Gain and Loss Intron gain and loss in the 2 dicot species, Arabidopsis and poplar, were calculated using Dollo parsimony (Rogozin et al. 2005), with rice as an outgroup. Introns that are shared by 1 of the dicot species and rice are counted as being lost in the second dicot species, whereas introns present in 1 of the dicot species but missing in the other and in rice are counted as gained in the respective dicot species. The same alignment quality criterion was applied as for the analysis of intron position conservation. Information Content of the Splice Junction and Intron Phase Information content at the splice junctions was calculated for the −6 to +4 positions of the donor and acceptor sites of each reliable (no gaps in the alignment within a 5 amino acid window on either side) intron positions in the clusters of orthologous plant genes. The information content was calculated using the formula: where, pik is the nucleotide frequency of ith nucleotide in position k (Stephens and Schneider 1992; Sverdlov et al. 2003). Evolutionary Rate Nonsynonymous (Ka) and synonymous (Ks) substitution rates for each protein-coding gene sequence were calculated as follows. For each protein sequence in every alignment, the corresponding coding sequences (CDSs without introns) were extracted from the database. These sequences were aligned using the protein sequence alignment as guide. The calculation of the Ka, Ks, and Ka/Ks were performed using the method of Nei and Gojobori (1986) as implemented in the program yn00 of the PAML package (Yang 2007). Results and Discussion Plastid-Derived Genes Have Slightly but Significantly Lower Intron Densities than Ancestral Genes and Fewer Conserved Intron Positions In an attempt to characterize the process of intron gain in plastid-derived genes, we compared a variety of features of these genes with the corresponding features of ancestral eukaryotic genes in the same plants. The plastid-derived genes represented an updated set of genes that showed a phylogenetic affinity with cyanobacteria and were identified, essentially, as described previously (Martin et al. [2002] and see Materials and Methods). The ancestral eukaryotic genes were defined as those that are conserved between plants and representatives of other eukaryotic kingdoms and hence appear to antedate the origin of Plantae (disregarding the unlikely possibility of horizontal gene transfer between phylogenetically remote eukaryotes); 2 levels of stringency were employed to delineate a more permissive and a more strictly defined sets of ancestral genes (for details see Materials and Methods). Clearly, the density of introns in plastid-derived genes from all 3 plant genomes, calculated either per gene or per unit length of the CDS and either by comparing the means or the entire density distributions across the plastid-derived and ancestral gene sets, was slightly but significantly lower than the respective densities in the ancestral genes (table 1). Table 1 Intron Density in Plastid-Derived Genes and Ancestral Genes of Plants Plastid Deriveda Ancestral (relaxed ancestral)a P value (P value with relaxed ancestral) Arabidopsis Genes 1,543 6,609 (7,933) Introns 7,904 44,989 (50,073) Amino acids 701,238 3,292,488 (3,823,834) Mean protein length, aa 454 498 Intron content (/gene) 5.12 6.80 (6.31) <2.2 × 10−16 (<2.2 × 10−16)b Mean intron density (/aa) 0.0113 0.0137 (0.0131) <2.2 × 10−16 (<2.2 × 10−16)b Median intron density (/aa) 0.011 0.0139 (0.013) <2.2 × 10−16 (1.895 × 10−09)c Poplar Genes 831 3,059 (3,630) Introns 4,509 21,244 (23,506) Amino acids 349,197 1,503,459 (1,735,468) Mean protein length, aa 420 491 Intron content (/gene) 5.42 6.94 (6.47) 3.126 × 10−13 (1.859 × 10−07) Mean intron density (/aa) 0.0129 0.0141 (0.0135) 3.679 × 10−08 (3.4 × 10−03) Median intron density (/aa) 0.0129 0.0147 (0.0140) 9.73 × 10−03 (0.34) Rice Genes 767 2,716 (3,129) Introns 4,488 20,322 (22,299) Amino acids 332,527 1,361,110 (1,539,890) Mean protein length, aa 434 501 Intron content (/gene) 5.85 7.48 (7.12) 4.412 × 10−16 (5.34 × 10−11) Mean intron density (/aa) 0.0135 0.0149 (0.0145) 8.39 × 10−10 (1.764 × 10−05) Median intron density (/aa) 0.0136 0.0160 (0.0153) 7.46 × 10−06 (7.6 × 10−04) Plastid Deriveda Ancestral (relaxed ancestral)a P value (P value with relaxed ancestral) Arabidopsis Genes 1,543 6,609 (7,933) Introns 7,904 44,989 (50,073) Amino acids 701,238 3,292,488 (3,823,834) Mean protein length, aa 454 498 Intron content (/gene) 5.12 6.80 (6.31) <2.2 × 10−16 (<2.2 × 10−16)b Mean intron density (/aa) 0.0113 0.0137 (0.0131) <2.2 × 10−16 (<2.2 × 10−16)b Median intron density (/aa) 0.011 0.0139 (0.013) <2.2 × 10−16 (1.895 × 10−09)c Poplar Genes 831 3,059 (3,630) Introns 4,509 21,244 (23,506) Amino acids 349,197 1,503,459 (1,735,468) Mean protein length, aa 420 491 Intron content (/gene) 5.42 6.94 (6.47) 3.126 × 10−13 (1.859 × 10−07) Mean intron density (/aa) 0.0129 0.0141 (0.0135) 3.679 × 10−08 (3.4 × 10−03) Median intron density (/aa) 0.0129 0.0147 (0.0140) 9.73 × 10−03 (0.34) Rice Genes 767 2,716 (3,129) Introns 4,488 20,322 (22,299) Amino acids 332,527 1,361,110 (1,539,890) Mean protein length, aa 434 501 Intron content (/gene) 5.85 7.48 (7.12) 4.412 × 10−16 (5.34 × 10−11) Mean intron density (/aa) 0.0135 0.0149 (0.0145) 8.39 × 10−10 (1.764 × 10−05) Median intron density (/aa) 0.0136 0.0160 (0.0153) 7.46 × 10−06 (7.6 × 10−04) NOTE.—aa, amino acid. a The gene sets in poplar and rice were derived by orthology with the respective Arabidopsis genes determined in pairwise genome comparison (see Materials and Methods); the numbers differ because not all Arabidopsis genes had one-to-one orthologs in poplar and rice. b Calculated using Fisher’s exact test. c Calculated by comparing intron density distributions across the respective gene sets using the t-tests. Open in new tab Table 1 Intron Density in Plastid-Derived Genes and Ancestral Genes of Plants Plastid Deriveda Ancestral (relaxed ancestral)a P value (P value with relaxed ancestral) Arabidopsis Genes 1,543 6,609 (7,933) Introns 7,904 44,989 (50,073) Amino acids 701,238 3,292,488 (3,823,834) Mean protein length, aa 454 498 Intron content (/gene) 5.12 6.80 (6.31) <2.2 × 10−16 (<2.2 × 10−16)b Mean intron density (/aa) 0.0113 0.0137 (0.0131) <2.2 × 10−16 (<2.2 × 10−16)b Median intron density (/aa) 0.011 0.0139 (0.013) <2.2 × 10−16 (1.895 × 10−09)c Poplar Genes 831 3,059 (3,630) Introns 4,509 21,244 (23,506) Amino acids 349,197 1,503,459 (1,735,468) Mean protein length, aa 420 491 Intron content (/gene) 5.42 6.94 (6.47) 3.126 × 10−13 (1.859 × 10−07) Mean intron density (/aa) 0.0129 0.0141 (0.0135) 3.679 × 10−08 (3.4 × 10−03) Median intron density (/aa) 0.0129 0.0147 (0.0140) 9.73 × 10−03 (0.34) Rice Genes 767 2,716 (3,129) Introns 4,488 20,322 (22,299) Amino acids 332,527 1,361,110 (1,539,890) Mean protein length, aa 434 501 Intron content (/gene) 5.85 7.48 (7.12) 4.412 × 10−16 (5.34 × 10−11) Mean intron density (/aa) 0.0135 0.0149 (0.0145) 8.39 × 10−10 (1.764 × 10−05) Median intron density (/aa) 0.0136 0.0160 (0.0153) 7.46 × 10−06 (7.6 × 10−04) Plastid Deriveda Ancestral (relaxed ancestral)a P value (P value with relaxed ancestral) Arabidopsis Genes 1,543 6,609 (7,933) Introns 7,904 44,989 (50,073) Amino acids 701,238 3,292,488 (3,823,834) Mean protein length, aa 454 498 Intron content (/gene) 5.12 6.80 (6.31) <2.2 × 10−16 (<2.2 × 10−16)b Mean intron density (/aa) 0.0113 0.0137 (0.0131) <2.2 × 10−16 (<2.2 × 10−16)b Median intron density (/aa) 0.011 0.0139 (0.013) <2.2 × 10−16 (1.895 × 10−09)c Poplar Genes 831 3,059 (3,630) Introns 4,509 21,244 (23,506) Amino acids 349,197 1,503,459 (1,735,468) Mean protein length, aa 420 491 Intron content (/gene) 5.42 6.94 (6.47) 3.126 × 10−13 (1.859 × 10−07) Mean intron density (/aa) 0.0129 0.0141 (0.0135) 3.679 × 10−08 (3.4 × 10−03) Median intron density (/aa) 0.0129 0.0147 (0.0140) 9.73 × 10−03 (0.34) Rice Genes 767 2,716 (3,129) Introns 4,488 20,322 (22,299) Amino acids 332,527 1,361,110 (1,539,890) Mean protein length, aa 434 501 Intron content (/gene) 5.85 7.48 (7.12) 4.412 × 10−16 (5.34 × 10−11) Mean intron density (/aa) 0.0135 0.0149 (0.0145) 8.39 × 10−10 (1.764 × 10−05) Median intron density (/aa) 0.0136 0.0160 (0.0153) 7.46 × 10−06 (7.6 × 10−04) NOTE.—aa, amino acid. a The gene sets in poplar and rice were derived by orthology with the respective Arabidopsis genes determined in pairwise genome comparison (see Materials and Methods); the numbers differ because not all Arabidopsis genes had one-to-one orthologs in poplar and rice. b Calculated using Fisher’s exact test. c Calculated by comparing intron density distributions across the respective gene sets using the t-tests. Open in new tab A potential complication to this conclusion is that the proteins encoded by plastid-derived genes are, on average, shorter than those encoded by ancestral genes (table 1). It has been reported that intron density positively correlates with the length of a gene's CDS (Lynch and Kewalramani 2003); accordingly, it could not be ruled out that the plastid-derived genes had lower intron density than ancestral genes solely because they encoded, on average, shorter proteins. To address this problem, we compared the intron densities in the plastid-derived genes of each of the 3 plant species with those in a randomly generated subset of ancestral genes with CDS length distributions statistically indistinguishable from those in the plastid-derived genes. The difference in intron densities between plastid-derived genes and ancestral genes was detected in these comparisons as well and remained highly significant for Arabidopsis and rice although not for poplar (supplementary table S2, Supplementary Material online). Thus, we conclude that the lower intron density in plastid-derived genes is not, primarily, explained by the differences in the lengths of the encoded proteins. By contrast, a comparison of the distributions of intron lengths in plastid-derived genes and in ancestral genes did not reveal any significant differences, with the median intron lengths being 114 and 113 nt, respectively (fig. 1). Thus, plastid-derived genes and ancestral genes in the same plants have significantly different intron densities but, essentially, the same characteristic intron length. It has been reported that intron length is linked to gene expression level, but although the signs of the correlation appear to be opposite in animals and plants such that highly expressed animal genes have longer introns than lowly expressed ones (Castillo-Davis et al. 2002), whereas, in plants, highly expressed genes have longer introns (Ren et al. 2006). However, the near identity of the distributions of intron lengths in plastid-derived and ancestral genes suggests that the lower intron density in the plastid-derived genes is unrelated to expression but rather is rooted in the history of the 2 classes of genes. FIG. 1.— Open in new tabDownload slide Distributions of intron lengths in chloroplast-derived and ancestral genes of plants. The median values were 114 and 113 nt, respectively. The y axis is shown only partially to emphasize the median values. The box extends to the first quartile above and below the median, and the thin lines extend to the third quartile. The thick black line shows the data beyond the third quartile. FIG. 2.— Open in new tabDownload slide Information content of exonic and intronic positions flanking the splice junctions of chloroplast-derived and ancestral genes of plants. We further compared the level of conservation (the amount of variation) of intron positions in the plastid-derived and ancestral genes. In agreement with the previous observations (Roy and Penny 2007), over 95% of the intron positions were found to be conserved in both classes of genes in all 3 pairwise comparisons of plant genomes. However, the plastid-derived genes consistently showed a lower level of conservation and the difference was highly significant for the comparisons of each of the dicots with rice (table 2), suggesting that lineage-specific loss and/or gain of introns were more prominent in plastid-derived than in ancestral genes. Table 2 Conservation of Intron Positions in Plastid-Derived and Ancestral Genes in Plants Conserved Intron Positions Variable Intron Positions Fraction of Variable Positions Arabidopsis versus Rice     Plastid derived 2,384 106 0.042     Ancestral (relaxed ancestral) 12,216 (13,132) 415 (461) 0.032 (0.033) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.0188     Plastid derived versus relaxed ancestral 0.0333 Arabidopsis versus Poplar     Plastid derived 2,450 61 0.024     Ancestral (relaxed) 12,432 (13,362) 255 (280) 0.020 (0.020) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.192     Plastid derived versus relaxed ancestral 0.226 Rice versus Poplar     Plastid derived 2,420 85 0.033     Ancestral (relaxed) 12,412 (13,360) 328 (357) 0.025 (0.026) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.026     Plastid derived versus relaxed ancestral 0.027 Conserved Intron Positions Variable Intron Positions Fraction of Variable Positions Arabidopsis versus Rice     Plastid derived 2,384 106 0.042     Ancestral (relaxed ancestral) 12,216 (13,132) 415 (461) 0.032 (0.033) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.0188     Plastid derived versus relaxed ancestral 0.0333 Arabidopsis versus Poplar     Plastid derived 2,450 61 0.024     Ancestral (relaxed) 12,432 (13,362) 255 (280) 0.020 (0.020) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.192     Plastid derived versus relaxed ancestral 0.226 Rice versus Poplar     Plastid derived 2,420 85 0.033     Ancestral (relaxed) 12,412 (13,360) 328 (357) 0.025 (0.026) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.026     Plastid derived versus relaxed ancestral 0.027 Open in new tab Table 2 Conservation of Intron Positions in Plastid-Derived and Ancestral Genes in Plants Conserved Intron Positions Variable Intron Positions Fraction of Variable Positions Arabidopsis versus Rice     Plastid derived 2,384 106 0.042     Ancestral (relaxed ancestral) 12,216 (13,132) 415 (461) 0.032 (0.033) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.0188     Plastid derived versus relaxed ancestral 0.0333 Arabidopsis versus Poplar     Plastid derived 2,450 61 0.024     Ancestral (relaxed) 12,432 (13,362) 255 (280) 0.020 (0.020) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.192     Plastid derived versus relaxed ancestral 0.226 Rice versus Poplar     Plastid derived 2,420 85 0.033     Ancestral (relaxed) 12,412 (13,360) 328 (357) 0.025 (0.026) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.026     Plastid derived versus relaxed ancestral 0.027 Conserved Intron Positions Variable Intron Positions Fraction of Variable Positions Arabidopsis versus Rice     Plastid derived 2,384 106 0.042     Ancestral (relaxed ancestral) 12,216 (13,132) 415 (461) 0.032 (0.033) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.0188     Plastid derived versus relaxed ancestral 0.0333 Arabidopsis versus Poplar     Plastid derived 2,450 61 0.024     Ancestral (relaxed) 12,432 (13,362) 255 (280) 0.020 (0.020) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.192     Plastid derived versus relaxed ancestral 0.226 Rice versus Poplar     Plastid derived 2,420 85 0.033     Ancestral (relaxed) 12,412 (13,360) 328 (357) 0.025 (0.026) P values (Fisher’s exact test)     Plastid derived versus ancestral 0.026     Plastid derived versus relaxed ancestral 0.027 Open in new tab A Higher Rate of Intron Gain but Not Intron Loss in Plastid-Derived Genes Compared with Ancestral Genes We further sought to directly compare the rates of intron gain and loss in plastid-derived and ancestral genes of plants. Given the availability of the complete genomes of 3 relatively close plant species with an unequivocally resolved phylogeny, the number of intron gains and losses in each of the 2 dicotyledons (Arabidopsis and poplar) could be determined in a straightforward fashion using the Dollo parsimony method (Rogozin et al. 2003, 2005). Introns that are shared by 1 of the dicot species and rice are counted as being lost in the second dicot species, whereas introns present in 1 of the dicot species but missing in the other and in rice are counted as gained in the respective species. The use of Dollo parsimony as opposed to more complex maximum likelihood models (Carmel et al. 2005, 2007b; Csuros 2005; Nguyen et al. 2005) is justified, in this case, because, at such short evolutionary distances, the contribution of parallel gain of introns in the sites in independent lineages is negligible (Rogozin et al. 2005; Sverdlov et al. 2005); maximum likelihood analysis was not practical because of the small number of potential gains and losses. In a general agreement with a previous report (Roy and Penny 2007), we found that only a few introns were lost or gained after the divergence of the 2 dicots (table 3). In Arabidopsis, there was a substantial excess of losses over gains, again, as previously reported (Roy and Penny 2007); no such excess of intron loss was detected in poplar, which has a much less compact genome than Arabidopsis (Tuskan et al. 2006) (table 3). Although the compared numbers of introns lost and gained were small, a significant excess of gains in plastid-derived genes was detected as compared with ancestral genes (table 3); by contrast, there was no significant difference in the rates of intron loss between the 2 classes of plant genes (table 3). Overall, in Arabidopsis, even in plastid-derived genes, slightly more introns were lost than gained but in poplar, a small net gain of introns was seen (table 3). Although some overestimate of intron gain is possible due to parallel losses (even as not many parallel losses are expected given the small overall number of lost introns), this effect would apply equally to plastid-derived and ancestral genes. Table 3 Intron Gain and Loss in Plastid-Derived and Ancestral Genes of Dicotyledons Plastid-derived (P) Ancestral (A) Ancestral Relaxed (R) Introns 1,310 6,563 7,069 Gain (fraction) 15(0.011) 29(0.004) 32(0.005) Arabidopsis Loss (fraction) 26(0.020) 142(0.022) 160(0.023) P values (Fisher’s exact test) P and A 0.0092 P and R 0.0087 Poplar Gain (fraction) 12(0.009) 40(0.006) 42(0.006) Loss (fraction) 8(0.006) 44(0.007) 46(0.007) P values (Fisher’s exact test) P and A 0.456 P and R 0.458 Total gain 27 69 74 Total loss 34 186 206 P values (Fisher’s exact test) P and A 0.0126 P and R 0.0082 Plastid-derived (P) Ancestral (A) Ancestral Relaxed (R) Introns 1,310 6,563 7,069 Gain (fraction) 15(0.011) 29(0.004) 32(0.005) Arabidopsis Loss (fraction) 26(0.020) 142(0.022) 160(0.023) P values (Fisher’s exact test) P and A 0.0092 P and R 0.0087 Poplar Gain (fraction) 12(0.009) 40(0.006) 42(0.006) Loss (fraction) 8(0.006) 44(0.007) 46(0.007) P values (Fisher’s exact test) P and A 0.456 P and R 0.458 Total gain 27 69 74 Total loss 34 186 206 P values (Fisher’s exact test) P and A 0.0126 P and R 0.0082 Open in new tab Table 3 Intron Gain and Loss in Plastid-Derived and Ancestral Genes of Dicotyledons Plastid-derived (P) Ancestral (A) Ancestral Relaxed (R) Introns 1,310 6,563 7,069 Gain (fraction) 15(0.011) 29(0.004) 32(0.005) Arabidopsis Loss (fraction) 26(0.020) 142(0.022) 160(0.023) P values (Fisher’s exact test) P and A 0.0092 P and R 0.0087 Poplar Gain (fraction) 12(0.009) 40(0.006) 42(0.006) Loss (fraction) 8(0.006) 44(0.007) 46(0.007) P values (Fisher’s exact test) P and A 0.456 P and R 0.458 Total gain 27 69 74 Total loss 34 186 206 P values (Fisher’s exact test) P and A 0.0126 P and R 0.0082 Plastid-derived (P) Ancestral (A) Ancestral Relaxed (R) Introns 1,310 6,563 7,069 Gain (fraction) 15(0.011) 29(0.004) 32(0.005) Arabidopsis Loss (fraction) 26(0.020) 142(0.022) 160(0.023) P values (Fisher’s exact test) P and A 0.0092 P and R 0.0087 Poplar Gain (fraction) 12(0.009) 40(0.006) 42(0.006) Loss (fraction) 8(0.006) 44(0.007) 46(0.007) P values (Fisher’s exact test) P and A 0.456 P and R 0.458 Total gain 27 69 74 Total loss 34 186 206 P values (Fisher’s exact test) P and A 0.0126 P and R 0.0082 Open in new tab The emerging interplay between intron loss and gain in plant genes is fairly complex. In ancestral eukaryotic genes, intron loss decidedly dominates over gain, in agreement with the findings of Roy and Penny (2007). The pattern of intron loss and gain in the plastid-derived genes can be plausibly conceived of as a superposition of 2 opposite trends: 1) the gain of introns that continues, albeit at a low rate, since the time of the original invasion and 2) the general trend of intron loss that is particularly pronounced in the Arabidopsis lineage, probably, due to the evolution toward genome contraction. Differences in Intron Phase Distribution and Splice Sites Information Content between Chloroplast-Derived and Ancestral Genes We further compared additional salient features of introns in chloroplast-derived and ancestral genes, including intron phase distributions and the patterns of information content in and around the donor and acceptor splice sites. It is well known that introns in all species show a substantial excess of phase 0 (i.e., introns located between codons) over phases 1 and 2 (introns located after the first and second bases of a codon, respectively) (Fedorov et al. 1992; de Souza et al. 1998). Subsequently, it has been noticed that relatively new introns have an even greater excess of phase 0 than ancient introns, and this has been explained by the apparent migration of information from the exonic splice signals to the sequence of the intron itself, in the course of an intron's evolution (Sverdlov et al. 2003). Under this hypothesis, introns preferentially fix in phase 0 but, subsequently, some of them slide to phases 1 and 2, thanks to the relaxed selective constraints on flanking nucleotides after the migration of information into the intron. The results of the comparison of the phase distributions of introns in the chloroplast-derived and ancestral genes of plants appeared to be in line with this interpretation of intron evolution. In particular, the introns in chloroplast-derived genes showed a significantly greater excess of phase 0 than ancestral genes (table 4). The comparison of the information content of splice sites also conformed with the previously established pattern. The sites in chloroplast-derived genes showed a higher information content in exonic positions flanking the splice junctions, especially, in the −1 position upstream of the donor site and the +1 position downstream of the acceptor site, and conversely, ancestral genes showed a stronger signal in intronic positions adjacent to the splice junctions (fig. 2). Table 4 Phase Distribution of Introns in Chloroplast-Derived and Ancestral Genes of Plants Phase 0a Phase 1a Phase 2a Total Arabidopsis Chloroplast derived 4,735 (0.599) 1,662 (0.210) 1,507 (0.190) 7,904 Ancestral 26,693 (0.593) 8,639 (0.192) 9,657 (0.214) 44,989 Ancestral relaxed 29,593 (0.590) 9,783 (0.195) 10,697 (0.213) 50,073 P valuesb Chloroplast derived versus Ancestral 2.769 × 10−07 Chloroplast derived versus Ancestral relaxed 2.86 × 10−06 Poplar Chloroplast derived 2,713 (0.601) 889 (0.197) 907 (0.201) 4,509 Ancestral 12,733 (0.599) 4,021 (0.189) 4,490 (0.211) 21,244 Ancestral relaxed 13,980 (0.594) 4,546 (0.193) 4,980 (0.211) 23,506 P valuesb Chloroplast derived versus Ancestral 0.21 Chloroplast derived versus Ancestral relaxed 0.26 Rice Chloroplast derived 2,706 (0.602) 907 (0.202) 875 (0.194) 4,488 Ancestral 12,040 (0.592) 3,979 (0.195) 4,303 (0.211) 20,322 Ancestral relaxed 13,173 (0.590) 4,401 (0.197) 4,725 (0.211) 22,299 P valuesb Chloroplast derived versus ancestral 0.04 Chloroplast derived versus ancestral relaxed 0.03 Phase 0a Phase 1a Phase 2a Total Arabidopsis Chloroplast derived 4,735 (0.599) 1,662 (0.210) 1,507 (0.190) 7,904 Ancestral 26,693 (0.593) 8,639 (0.192) 9,657 (0.214) 44,989 Ancestral relaxed 29,593 (0.590) 9,783 (0.195) 10,697 (0.213) 50,073 P valuesb Chloroplast derived versus Ancestral 2.769 × 10−07 Chloroplast derived versus Ancestral relaxed 2.86 × 10−06 Poplar Chloroplast derived 2,713 (0.601) 889 (0.197) 907 (0.201) 4,509 Ancestral 12,733 (0.599) 4,021 (0.189) 4,490 (0.211) 21,244 Ancestral relaxed 13,980 (0.594) 4,546 (0.193) 4,980 (0.211) 23,506 P valuesb Chloroplast derived versus Ancestral 0.21 Chloroplast derived versus Ancestral relaxed 0.26 Rice Chloroplast derived 2,706 (0.602) 907 (0.202) 875 (0.194) 4,488 Ancestral 12,040 (0.592) 3,979 (0.195) 4,303 (0.211) 20,322 Ancestral relaxed 13,173 (0.590) 4,401 (0.197) 4,725 (0.211) 22,299 P valuesb Chloroplast derived versus ancestral 0.04 Chloroplast derived versus ancestral relaxed 0.03 a Number in the parenthesis indicated fraction of total. b Calculated using chi-square 2 × 3 table with degrees of freedom = 2. Open in new tab Table 4 Phase Distribution of Introns in Chloroplast-Derived and Ancestral Genes of Plants Phase 0a Phase 1a Phase 2a Total Arabidopsis Chloroplast derived 4,735 (0.599) 1,662 (0.210) 1,507 (0.190) 7,904 Ancestral 26,693 (0.593) 8,639 (0.192) 9,657 (0.214) 44,989 Ancestral relaxed 29,593 (0.590) 9,783 (0.195) 10,697 (0.213) 50,073 P valuesb Chloroplast derived versus Ancestral 2.769 × 10−07 Chloroplast derived versus Ancestral relaxed 2.86 × 10−06 Poplar Chloroplast derived 2,713 (0.601) 889 (0.197) 907 (0.201) 4,509 Ancestral 12,733 (0.599) 4,021 (0.189) 4,490 (0.211) 21,244 Ancestral relaxed 13,980 (0.594) 4,546 (0.193) 4,980 (0.211) 23,506 P valuesb Chloroplast derived versus Ancestral 0.21 Chloroplast derived versus Ancestral relaxed 0.26 Rice Chloroplast derived 2,706 (0.602) 907 (0.202) 875 (0.194) 4,488 Ancestral 12,040 (0.592) 3,979 (0.195) 4,303 (0.211) 20,322 Ancestral relaxed 13,173 (0.590) 4,401 (0.197) 4,725 (0.211) 22,299 P valuesb Chloroplast derived versus ancestral 0.04 Chloroplast derived versus ancestral relaxed 0.03 Phase 0a Phase 1a Phase 2a Total Arabidopsis Chloroplast derived 4,735 (0.599) 1,662 (0.210) 1,507 (0.190) 7,904 Ancestral 26,693 (0.593) 8,639 (0.192) 9,657 (0.214) 44,989 Ancestral relaxed 29,593 (0.590) 9,783 (0.195) 10,697 (0.213) 50,073 P valuesb Chloroplast derived versus Ancestral 2.769 × 10−07 Chloroplast derived versus Ancestral relaxed 2.86 × 10−06 Poplar Chloroplast derived 2,713 (0.601) 889 (0.197) 907 (0.201) 4,509 Ancestral 12,733 (0.599) 4,021 (0.189) 4,490 (0.211) 21,244 Ancestral relaxed 13,980 (0.594) 4,546 (0.193) 4,980 (0.211) 23,506 P valuesb Chloroplast derived versus Ancestral 0.21 Chloroplast derived versus Ancestral relaxed 0.26 Rice Chloroplast derived 2,706 (0.602) 907 (0.202) 875 (0.194) 4,488 Ancestral 12,040 (0.592) 3,979 (0.195) 4,303 (0.211) 20,322 Ancestral relaxed 13,173 (0.590) 4,401 (0.197) 4,725 (0.211) 22,299 P valuesb Chloroplast derived versus ancestral 0.04 Chloroplast derived versus ancestral relaxed 0.03 a Number in the parenthesis indicated fraction of total. b Calculated using chi-square 2 × 3 table with degrees of freedom = 2. Open in new tab Conclusions The comparisons of the intron densities and features in plant genes derived from plastids and ancestral eukaryotic genes from the same plants lead to 2 principal, complementary conclusions. First, the chloroplast-derived genes that contained no introns when they resided in the cyanobacterial genome immediately after the symbiosis have accumulated a large number of introns such that the intron density in these genes became close to that in the ancestral genes. The current collection of genomes from the plant lineage is insufficient to reconstruct the dynamics of intron invasion into the plastid-derived genes. However, almost 96% of the intron positions in the plastid-derived genes are conserved between dicotyledons and monocotyledons, indicating that the great majority of introns were acquired by these genes prior to the dicot–monocot divergence 150–200 MYA (Wolfe et al. 1989; Chaw et al. 2004). Previous analyses have suggested that extensive intron gain occurred only during short intervals of eukaryotic evolution, possibly, coinciding with major evolutionary innovations (Babenko et al. 2004; Carmel et al. 2007b). Thus, it seems likely that the bulk of the invasion occurred over a relatively short time span such that, in this respect, the occupation of the plastid-derived genes with introns resembled the original invasion of introns into the genes of the emerging eukaryotic cell, presumably, following the mitochondrial endosymbiosis (Koonin 2006; Martin and Koonin 2006). These results unequivocally show that massive intron gain occurred on at least 1 occasion during eukaryotic evolution other than the original invasion. At present, both demonstrable waves of intron invasion are associated with primary endosymbiosis; it remains yet uncertain whether or not extensive intron gain also accompanies secondary endosymbiotic events. Second, however, we observed that the plastid-derived genes continued to accumulate new introns at a slightly but significantly greater rate than ancestral genes even after the divergence of the 2 dicot species. This suggests that the plastid-derived genes still have not reached the saturation intron density. It is tempting to speculate that this saturation density, which, of course, differs between eukaryotic taxa, has an adaptive value, a notion that is in agreement with the recent demonstration of a positive correlation between intron gain rate and evolutionary conservation of genes (Carmel et al. 2007a). The pattern of intron gain and loss in plastid-derived genes is further complicated by the superposition of the apparent continued gain and the more general trend of intron loss during the recent evolution of plant genes, at least, in the lineages where genome contraction is observed, such as Arabidopsis. The present conclusions give only a rough, first approximation reconstruction of the evolutionary dynamics of introns in the plastid-derived genes because of the small numbers of reliably identifiable intron losses and gains. For a more detailed reconstruction, a diverse sample of genomes from different branches of plants is required. We thank Miklos Csuros for helpful discussions. This work was supported in part by the Intramural Research Program of the National Library of Medicine at National Institutes of Health/Department of Health and Human Services. References Altschul SF , Madden TL , Schaffer AA , Zhang J , Zhang Z , Miller W , Lipman DJ . Gapped BLAST and PSI-BLAST: a new generation of protein database search programs , Nucleic Acids Res , 1997 , vol. 25 (pg. 3389 - 3402 ) Google Scholar Crossref Search ADS PubMed WorldCat Babenko VN , Rogozin IB , Mekhedov SL , Koonin EV . Prevalence of intron gain over intron loss in the evolution of paralogous gene families , Nucleic Acids Res , 2004 , vol. 32 (pg. 3724 - 3733 ) Google Scholar Crossref Search ADS PubMed WorldCat Belshaw R , Bensasson D . The rise and falls of introns , Heredity , 2006 , vol. 96 (pg. 208 - 213 ) Google Scholar Crossref Search ADS PubMed WorldCat Carmel L , Rogozin IB , Wolf YI , Koonin EV . An expectation-maximization algorithm for analysis of evolution of exon-intron structure of eukaryotic genes , Comp Genomics Lect Notes Comput Sci , 2005 , vol. 3678 (pg. 35 - 46 ) Google Scholar OpenURL Placeholder Text WorldCat Carmel L , Rogozin IB , Wolf YI , Koonin EV . Evolutionarily conserved genes preferentially accumulate introns , Genome Res , 2007a , vol. 17 (pg. 1045 - 1050 ) Google Scholar Crossref Search ADS WorldCat Carmel L , Wolf YI , Rogozin IB , Koonin EV . Three distinct modes of intron dynamics in the evolution of eukaryotes , Genome Res , 2007b , vol. 17 (pg. 1034 - 1044 ) Google Scholar Crossref Search ADS WorldCat Castillo-Davis CI , Mekhedov SL , Hartl DL , Koonin EV , Kondrashov FA . Selection for short introns in highly expressed genes , Nat Genet , 2002 , vol. 31 (pg. 415 - 418 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Chaw SM , Chang CC , Chen HL , Li WH . Dating the monocot-dicot divergence and the origin of core eudicots using whole chloroplast genomes , J Mol Evol , 2004 , vol. 58 (pg. 424 - 441 ) Google Scholar Crossref Search ADS PubMed WorldCat Cho G , Doolittle RF . Intron distribution in ancient paralogs supports random insertion and not random loss , J Mol Evol , 1997 , vol. 44 (pg. 573 - 584 ) Google Scholar Crossref Search ADS PubMed WorldCat Coghlan A , Wolfe KH . Origins of recently gained introns in Caenorhabditis , Proc Natl Acad Sci USA , 2004 , vol. 101 (pg. 11362 - 11367 ) Google Scholar Crossref Search ADS PubMed WorldCat Collins L , Penny D . Complex spliceosomal organization ancestral to extant eukaryotes , Mol Biol Evol , 2005 , vol. 22 (pg. 1053 - 1066 ) Google Scholar Crossref Search ADS PubMed WorldCat Csuros M . Likely scenarios of intron evolution , Comp Genomics Lect Notes Comput Sci , 2005 , vol. 3678 (pg. 47 - 60 ) Google Scholar OpenURL Placeholder Text WorldCat Darnell JE Jr . Implications of RNA-RNA splicing in evolution of eukaryotic cells , Science , 1978 , vol. 202 (pg. 1257 - 1260 ) Google Scholar Crossref Search ADS PubMed WorldCat de Souza SJ , Long M , Klein RJ , Roy S , Lin S , Gilbert W . Toward a resolution of the introns early/late debate: only phase zero introns are correlated with the structure of ancient proteins , Proc Natl Acad Sci USA , 1998 , vol. 95 (pg. 5094 - 5099 ) Google Scholar Crossref Search ADS PubMed WorldCat Deutsch M , Long M . Intron-exon structures of eukaryotic model organisms , Nucleic Acids Res , 1999 , vol. 27 (pg. 3219 - 3228 ) Google Scholar Crossref Search ADS PubMed WorldCat Doolittle WF . Genes in pieces: were they ever together? , Nature , 1978 , vol. 272 (pg. 581 - 582 ) Google Scholar Crossref Search ADS WorldCat Edgar RC . MUSCLE: multiple sequence alignment with high accuracy and high throughput , Nucleic Acids Res , 2004 , vol. 32 (pg. 1792 - 1797 ) Google Scholar Crossref Search ADS PubMed WorldCat Fedorov A , Merican AF , Gilbert W . Large-scale comparison of intron positions among animal, plant, and fungal genes , Proc Natl Acad Sci USA , 2002 , vol. 99 (pg. 16128 - 16133 ) Google Scholar Crossref Search ADS PubMed WorldCat Fedorov A , Roy S , Fedorova L , Gilbert W . Mystery of intron gain , Genome Res , 2003 , vol. 13 (pg. 2236 - 2241 ) Google Scholar Crossref Search ADS PubMed WorldCat Fedorov A , Suboch G , Bujakov M , Fedorova L . Analysis of nonuniformity in intron phase distribution , Nucleic Acids Res , 1992 , vol. 20 (pg. 2553 - 2557 ) Google Scholar Crossref Search ADS PubMed WorldCat Felsenstein J . Inferring phylogenies from protein sequences by parsimony, distance, and likelihood methods , Methods Enzymol , 1996 , vol. 266 (pg. 418 - 427 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Fink GR . Pseudogenes in yeast? , Cell , 1987 , vol. 49 (pg. 5 - 6 ) Google Scholar Crossref Search ADS PubMed WorldCat GenBank [Internet] Bethesda (MD) National Center for Biotechnology Information c2002 [cited 2007 Feb 15]. Available from: http://www.ncbi.nlm.nih.gov/Genbank/index.html Gilbert W . Why genes in pieces? , Nature , 1978 , vol. 271 pg. 501 Google Scholar Crossref Search ADS PubMed WorldCat Gilbert W . The exon theory of genes , Cold Spring Harb Symp Quant Biol , 1987 , vol. 52 (pg. 901 - 905 ) Google Scholar Crossref Search ADS PubMed WorldCat Gilbert W , de Souza SJ , Long M . Origin of genes , Proc Natl Acad Sci USA , 1997 , vol. 94 (pg. 7698 - 7703 ) Google Scholar Crossref Search ADS PubMed WorldCat Gilbert W , Glynias M . On the ancient nature of introns , Gene , 1993 , vol. 135 (pg. 137 - 144 ) Google Scholar Crossref Search ADS PubMed WorldCat Jeffares DC , Mourier T , Penny D . The biology of intron gain and loss , Trends Genet , 2006 , vol. 22 (pg. 16 - 22 ) Google Scholar Crossref Search ADS PubMed WorldCat Koonin EV . The origin of introns and their role in eukaryogenesis: a compromise solution to the introns-early versus introns-late debate? , Biol Direct , 2006 , vol. 1 pg. 22 Google Scholar Crossref Search ADS PubMed WorldCat Lambowitz AM , Zimmerly S . Mobile group II introns , Annu Rev Genet , 2004 , vol. 38 (pg. 1 - 35 ) Google Scholar Crossref Search ADS PubMed WorldCat Logsdon JM Jr . The recent origins of spliceosomal introns revisited , Curr Opin Genet Dev , 1998 , vol. 8 (pg. 637 - 648 ) Google Scholar Crossref Search ADS PubMed WorldCat Logsdon JM Jr , Palmer JD . Origin of introns—early or late? , Nature , 1994 , vol. 369 pg. 526 Google Scholar Crossref Search ADS PubMed WorldCat Logsdon JM Jr , Stoltzfus A , Doolittle WF . Molecular evolution: recent cases of spliceosomal intron gain? , Curr Biol , 1998 , vol. 8 (pg. R560 - R563 ) Google Scholar Crossref Search ADS PubMed WorldCat Logsdon JM Jr , Tyshenko MG , Dixon C , D-Jafari J , Walker VK , Palmer JD . Seven newly discovered intron positions in the triose-phosphate isomerase gene: evidence for the introns-late theory , Proc Natl Acad Sci USA , 1995 , vol. 92 (pg. 8507 - 8511 ) Google Scholar Crossref Search ADS PubMed WorldCat Lynch M , Conery JS . The origins of genome complexity , Science , 2003 , vol. 302 (pg. 1401 - 1404 ) Google Scholar Crossref Search ADS PubMed WorldCat Lynch M , Kewalramani A . Messenger RNA surveillance and the evolutionary proliferation of introns , Mol Biol Evol , 2003 , vol. 20 (pg. 563 - 571 ) Google Scholar Crossref Search ADS PubMed WorldCat Lynch M , Richardson AO . The evolution of spliceosomal introns , Curr Opin Genet Dev , 2002 , vol. 12 (pg. 701 - 710 ) Google Scholar Crossref Search ADS PubMed WorldCat Mans BJ , Anantharaman V , Aravind L , Koonin EV . Comparative genomics, evolution and origins of the nuclear envelope and nuclear pore complex , Cell Cycle , 2004 , vol. 3 (pg. 1612 - 1637 ) Google Scholar Crossref Search ADS PubMed WorldCat Martin W , Koonin EV . Introns and the origin of nucleus-cytosol compartmentalization , Nature , 2006 , vol. 440 (pg. 41 - 45 ) Google Scholar Crossref Search ADS PubMed WorldCat Martin W , Rujan T , Richly E , Hansen A , Cornelsen S , Lins T , Leister D , Stoebe B , Hasegawa M , Penny D . Evolutionary analysis of Arabidopsis, cyanobacterial, and chloroplast genomes reveals plastid phylogeny and thousands of cyanobacterial genes in the nucleus , Proc Natl Acad Sci USA , 2002 , vol. 99 (pg. 12246 - 12251 ) Google Scholar Crossref Search ADS PubMed WorldCat Martin W , Stoebe B , Goremykin V , Hapsmann S , Hasegawa M , Kowallik KV . Gene transfer to the nucleus and the evolution of chloroplasts , Nature , 1998 , vol. 393 (pg. 162 - 165 ) Google Scholar Crossref Search ADS PubMed WorldCat Mattick JS . Introns: evolution and function , Curr Opin Genet Dev , 1994 , vol. 4 (pg. 823 - 831 ) Google Scholar Crossref Search ADS PubMed WorldCat Mourier T , Jeffares DC . Eukaryotic intron loss , Science , 2003 , vol. 300 pg. 1393 Google Scholar Crossref Search ADS PubMed WorldCat Nei M , Gojobori T . Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions , Mol Biol Evol , 1986 , vol. 3 (pg. 418 - 426 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Nguyen HD , Yoshihama M , Kenmochi N . New maximum likelihood estimators for eukaryotic intron evolution , PLoS Comput Biol , 2005 , vol. 1 pg. e79 Google Scholar Crossref Search ADS PubMed WorldCat O'Brien KP , Remm M , Sonnhammer EL . Inparanoid: a comprehensive database of eukaryotic orthologs , Nucleic Acids Res , 2005 , vol. 33 (pg. D476 - D480 ) Google Scholar Crossref Search ADS PubMed WorldCat Qiu WG , Schisler N , Stoltzfus A . The evolutionary gain of spliceosomal introns: sequence and phase preferences , Mol Biol Evol , 2004 , vol. 21 (pg. 1252 - 1263 ) Google Scholar Crossref Search ADS PubMed WorldCat Raible F , Tessmar-Raible K , Osoegawa K , et al. (12 co-authors) Vertebrate-type intron-rich genes in the marine annelid Platynereis dumerilii , Science , 2005 , vol. 310 (pg. 1325 - 1326 ) Google Scholar Crossref Search ADS PubMed WorldCat Ren XY , Vorst O , Fiers MW , Stiekema WJ , Nap JP . In plants, highly expressed genes are the least compact , Trends Genet , 2006 , vol. 22 (pg. 528 - 532 ) Google Scholar Crossref Search ADS PubMed WorldCat Rice Annotation Project Database [Internet]. Japan: National Institute of Agrobiological Sciences and National Institute of Genetics; c2005 [cited 2006 Dec 20]. Available from: http://rapdb.lab.nig.ac.jp/ Rogozin IB , Babenko VN , Wolf YI , Koonin EV . Albert VA . Dollo parsimony and reconstruction of genome evolution , Parsimony, phylogeny, and genomics , 2005 Oxford Oxford University Press (pg. 190 - 200 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Rogozin IB , Wolf YI , Sorokin AV , Mirkin BG , Koonin EV . Remarkable interkingdom conservation of intron positions and massive, lineage-specific intron loss and gain in eukaryotic evolution , Curr Biol , 2003 , vol. 13 (pg. 1512 - 1517 ) Google Scholar Crossref Search ADS PubMed WorldCat Roy SW . Intron-rich ancestors , Trends Genet , 2006 , vol. 22 (pg. 468 - 471 ) Google Scholar Crossref Search ADS PubMed WorldCat Roy SW , Fedorov A , Gilbert W . Large-scale comparison of intron positions in mammalian genes shows intron loss but no gain , Proc Natl Acad Sci USA , 2003 , vol. 100 (pg. 7158 - 7162 ) Google Scholar Crossref Search ADS PubMed WorldCat Roy SW , Gilbert W . Complex early genes , Proc Natl Acad Sci USA , 2005a , vol. 102 (pg. 1986 - 1991 ) Google Scholar Crossref Search ADS WorldCat Roy SW , Gilbert W . Rates of intron loss and gain: implications for early eukaryotic evolution , Proc Natl Acad Sci USA , 2005b , vol. 102 (pg. 5773 - 5778 ) Google Scholar Crossref Search ADS WorldCat Roy SW , Irimia M , Penny D . Very little intron gain in Entamoeba histolytica genes laterally transferred from prokaryotes , Mol Biol Evol , 2006 , vol. 23 (pg. 1824 - 1827 ) Google Scholar Crossref Search ADS PubMed WorldCat Roy SW , Penny D . Smoke without fire: most reported cases of intron gain in nematodes instead reflect intron losses , Mol Biol Evol , 2006 , vol. 23 (pg. 2259 - 2262 ) Google Scholar Crossref Search ADS PubMed WorldCat Roy SW , Penny D . Patterns of intron loss and gain in plants: intron loss-dominated evolution and genome-wide comparison of O. sativa and A. thaliana , Mol Biol Evol , 2007 , vol. 24 (pg. 171 - 181 ) Google Scholar Crossref Search ADS PubMed WorldCat Schaffer AA , Aravind L , Madden TL , Shavirin S , Spouge JL , Wolf YI , Koonin EV , Altschul SF . Improving the accuracy of PSI-BLAST protein database searches with composition-based statistics and other refinements , Nucleic Acids Res , 2001 , vol. 29 (pg. 2994 - 3005 ) Google Scholar Crossref Search ADS PubMed WorldCat Stephens RM , Schneider TD . Features of spliceosome evolution and function inferred from an analysis of the information at human splice sites , J Mol Biol , 1992 , vol. 228 (pg. 1124 - 1136 ) Google Scholar Crossref Search ADS PubMed WorldCat Stoltzfus A . Origin of introns—early or late , Nature , 1994 , vol. 369 (pg. 526 - 527 ) Google Scholar Crossref Search ADS PubMed WorldCat Stoltzfus A , Spencer DF , Zuker M , Logsdon JM Jr , Doolittle WF . Testing the exon theory of genes: the evidence from protein structure , Science , 1994 , vol. 265 (pg. 202 - 207 ) Google Scholar Crossref Search ADS PubMed WorldCat Sverdlov AV , Rogozin IB , Babenko VN , Koonin EV . Evidence of splice signal migration from exon to intron during intron evolution , Curr Biol , 2003 , vol. 13 (pg. 2170 - 2174 ) Google Scholar Crossref Search ADS PubMed WorldCat Sverdlov AV , Rogozin IB , Babenko VN , Koonin EV . Conservation versus parallel gains in intron evolution , Nucleic Acids Res , 2005 , vol. 33 (pg. 1741 - 1748 ) Google Scholar Crossref Search ADS PubMed WorldCat Teich R , Grauvogel C , Petersen J . Intron distribution in Plantae: 500 million years of stasis during land plant evolution , Gene , 2007 , vol. 394 (pg. 96 - 104 ) Google Scholar Crossref Search ADS PubMed WorldCat Thompson JD , Higgins DG , Gibson TJ . CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice , Nucleic Acids Res , 1994 , vol. 22 (pg. 4673 - 4680 ) Google Scholar Crossref Search ADS PubMed WorldCat Timmis JN , Ayliffe MA , Huang CY , Martin W . Endosymbiotic gene transfer: organelle genomes forge eukaryotic chromosomes , Nat Rev Genet , 2004 , vol. 5 (pg. 123 - 135 ) Google Scholar Crossref Search ADS PubMed WorldCat Toro N , Jimenez-Zurdo JI , Garcia-Rodriguez FM . Bacterial group II introns: not just splicing , FEMS Microbiol Rev , 2007 , vol. 31 (pg. 342 - 358 ) Google Scholar Crossref Search ADS PubMed WorldCat Tuskan GAS , Difazio S , Jansson J , et al. , (110 co-authors) . The genome of black cottonwood, Populus trichocarpa (Torr. & Gray) , Science , 2006 , vol. 313 (pg. 1596 - 1604 ) Google Scholar Crossref Search ADS PubMed WorldCat Wolfe KH , Gouy M , Yang YW , Sharp PM , Li WH . Date of the monocot-dicot divergence estimated from chloroplast DNA sequence data , Proc Natl Acad Sci USA , 1989 , vol. 86 (pg. 6201 - 6205 ) Google Scholar Crossref Search ADS PubMed WorldCat Yang Z . PAML 4: phylogenetic analysis by maximum likelihood , Mol Biol Evol , 2007 , vol. 24 (pg. 1586 - 1591 ) Google Scholar Crossref Search ADS PubMed WorldCat Zimmerly S , Hausner G , Wu X . Phylogenetic relationships among group II intron ORFs , Nucleic Acids Res , 2001 , vol. 29 (pg. 1238 - 1250 ) Google Scholar Crossref Search ADS PubMed WorldCat Author notes Aoife McLysaght, Associate Editor Published by Oxford University Press 2007. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Published by Oxford University Press 2007.

Tawari, Blessing; Ali, Ibne Karim M.; Scott, Claire; Quail, Michael A.; Berriman, Matthew; Hall, Neil; Clark, C. Graham

doi: 10.1093/molbev/msm238pmid: 17974548

Abstract Genome sequencing of the protistan parasite Entamoeba histolytica HM-1:IMSS revealed that almost all the tRNA genes are organized into tandem arrays that make up over 10% of the genome. The 25 distinct array units contain up to 5 tRNA genes each and some also encode the 5S RNA. Between adjacent genes in array units are complex short tandem repeats (STRs) resembling microsatellites. To investigate the origins and evolution of this unique gene organization, we have undertaken a genome survey to determine the array unit organization in 4 other species of Entamoeba—Entamoeba dispar, Entamoeba moshkovskii, Entamoeba terrapinae, and Entamoeba invadens—and have explored the STR structure in other isolates of E. histolytica. The genome surveys revealed that E. dispar has the same array unit organization as E. histolytica, including the presence and numerical variation of STRs between adjacent genes. However, the individual repeat sequences are completely different to those in E. histolytica. All other species of Entamoeba studied also have tandem arrays of clustered tRNA genes, but the gene composition of the array units often differs from that in E. histolytica/E. dispar. None of the other species’ arrays exhibit the complex STRs between adjacent genes although simple tandem duplications are occasionally seen. The degree of similarity in organization reflects the phylogenetic relationships among the species studied. Within individual isolates of E. histolytica most copies of the array unit are uniform in sequence with only minor variation in the number and organization of the STRs. Between isolates, however, substantial differences in STR number and organization can exist although the individual repeat sequences tend to be conserved. The origin of this unique gene organization in the genus Entamoeba clearly predates the common ancestor of the species investigated to date and their function remains unclear. Entamoeba, tRNA genes, repeated DNA, recombination Introduction In 2005, Entamoeba histolytica HM-1:IMSS became the first Amoebozoan with a published genome sequence (Loftus et al. 2005). Although our understanding of its chromosome organization is still somewhat incomplete, a number of striking findings were made. Among these was the recognition that the tRNA genes were exceptionally abundant, with an estimated 4,500 copies—about 10 times the number in the human genome (Clark, Ali, et al. 2006). Additionally, most are clustered and organized into long tandem arrays that make up over 10% of the genome. Twenty-five distinct arrays have been described and are composed of tandemly repeated units encoding between 1 and 5 tRNA acceptor types. Three arrays also encode the organism's 5S RNA and 1 encodes another RNA suspected to be a small nuclear RNA (snRNA). The intergenic regions between all arrayed tRNA genes in E. histolytica contain complex structures made up of short sequences repeated in tandem. These repeats are of varying size but generally fall in the range of 7–12 bp, although a few are substantially longer (up to 44 bp). In many ways the structures resemble the micro- and minisatellites seen in other eukaryotic genomes. However, micro- and minisatellites are normally randomly dispersed throughout the genome, whereas the tRNA-linked E. histolytica short tandem repeats (STRs) form part of a larger unit that is itself tandemly arrayed. Because E. histolytica appears to lack “traditional” mini/microsatellites elsewhere in its genome, we have chosen to use the term STR for these structures in E. histolytica, to emphasize the apparent distinction between these loci and the mini/microsatellites of other organisms. Although clustering of an organism's tRNA genes has occasionally been observed in other eukaryotes, tandem arrays of tRNA genes have been reported to date only in E. histolytica. In the most closely related organism for which a genome sequence is available, Dictyostelium discoideum, individual tRNA genes are distributed throughout the genome (Eichinger et al. 2005). Observation of a unique gene organization inevitably raises the question of its origins, evolution, and function. To start to address these questions, we have investigated the organization of tRNA genes in other species of Entamoeba and other isolates of E. histolytica. For 4 species, a genome survey was performed, with 20,000 random sequences being obtained for each. The survey sequences have allowed an essentially complete picture of tRNA gene organization to be deduced for each of the 4 species and comparisons have yielded important insights into tRNA array evolution. Sequencing of the intergenic regions from dozens of isolates of E. histolytica has clarified the patterns of tRNA-linked STR evolution in that species. Materials and Methods DNA Isolation Entamoeba dispar SAW760, Entamoeba moshkovskii FIC and Laredo, Entamoeba terrapinae M, Entamoeba invadens IP-1, and Entamoeba ecuadoriensis EC were originally obtained from the American Type Culture Collection (accession numbers 50484, 30041, 30042, 30043, 30994, and 50261, respectively). Organisms were grown in LYI-S-2 medium (Clark and Diamond 2002) with 15% (E. dispar and E. ecuadoriensis) or 10% (others) heat-inactivated adult bovine serum at 36 °C (E. dispar and E. ecuadoriensis) or 22 °C (others). Entamoeba ecuadoriensis cultures were supplemented with antibiotic-inhibited Escherichia coli XL-1 cells; the other species were grown axenically. Isolation of Entamoeba equi Aberystwyth DNA was described previously (Clark, Kaffashian, et al. 2006), whereas E. moshkovskii GAH and WS genomic DNAs were a gift from Jeffrey D. Silberman, University of Arkansas. DNA samples from E. histolytica isolates for STR comparison were obtained during development of a genotyping method (Ali et al. 2005), a majority being from Bangladesh with the others coming from 11 different countries. Genome Survey Sequencing The genomes of E. dispar SAW760, E. moshkovskii FIC, E. terrapinae M, and E. invadens IP-1 were surveyed by sequencing of small insert libraries (ca. 2.5 kb). For genomic library construction, axenic cells were harvested, washed, embedded in low–melting point agarose (CleanCut agarose, Bio-Rad Laboratories Ltd, Hemel Hempstead, UK) and cast into pulsed field gel electrophoresis blocks. Blocks were incubated in lysis buffer with proteinase K at 55 °C for 2 days. After electrophoresis, high molecular weight DNA was purified by agarase digestion, sonicated, and cloned into pUC18 before sequencing. More detailed information on library construction can be found at: http://www.sanger.ac.uk/Teams/Team53/psub/methods.shtml. Based on the average read length (ca. 600 bp) of the 20,000 reads and assuming a similar genome size to E. histolytica (ca. 23 Mb) in all species, the genome coverage of the surveys is approximately 0.5×. Other Sequencing For routine analyses, DNA was isolated, polymerase chain reaction (PCR) amplification performed, and PCR products separated in 1.2–1.5% agarose gels as previously described (Ali et al. 2005). Products were purified using the QiaQuick gel extraction kit (Qiagen Ltd., Crawley, UK), sequenced directly (BigDye v.3.0, Applied Biosystems, Warrington, UK) using the individual amplification primers, and analyzed on an ABI3730 sequencer. For E. ecuadoriensis, E. moshkovskii Laredo, and E. equi, individual array units were obtained by PCR amplification using tRNA primers (Ali et al. 2005), gel purified as above, cloned into pGEM-T Easy (Promega UK, Southampton, UK), and individual plasmids sequenced using M13F and R primers. A few gaps in sequences from the genome surveys were also filled by this method. E. moshkovskii GAH and WS PCR products were sequenced directly using amplification primers. Sequences Deposited The genome survey read data have been deposited in the EMBL database with the following accession numbers: E. invadens AM600998–AM623528, E. dispar AM623529–AM643666, E. moshkovskii AM643667–AM665026, and E. terrapinae AM665027–AM689296. Consensus sequences of tRNA array units from the 4 genome surveys, and other species, and representative sequences from the E. histolytica intraspecific diversity study have been deposited in the databases under the accession numbers EF421258–EF421403 and EF427346–EF427363. tRNA Gene Identification Raw genome survey sequences were scanned using tRNAscan SE (Lowe and Eddy 1997). Reads encoding the same tRNA isoacceptor type were aligned (http://prodes.toulouse.inra.fr/multalin/multalin.html; Corpet 1988), manually edited, and a consensus sequence obtained. STR and Scaffold/Matrix Attachment Regions Identification Tandem Repeat Finder (http://tandem.bu.edu/trf/trf.html; Benson 1999) was used to search for STRs in the consensus sequences. Scaffold/matrix attachment regions (S/MAR) were predicted using MARSCAN (http://bioweb.pasteur.fr/seqanal/interfaces/marscan.html). Nomenclature Array unit names are derived from the encoded tRNAs (and 5S RNA, when present) using the standard single-letter amino acid abbreviation and are depicted in square brackets. The anticodon is included where necessary to distinguish between arrays. Regions of an array are also identified using the single-letter code and anticodons but without the square brackets. For example, A-S and SGCT-D are both intergenic regions in the E. histolytica array [ASD]. The anticodon is necessary for SGCT-D as another array, [SD], contains STGA-D. STR loci are also named after the tRNA genes flanking the intergenic region in which they are found, for example, STR locus SGCT-D. Results The genus Entamoeba consists of species that produce cysts with 1, 4, or 8 nuclei and a few that do not encyst. At present, it appears that cyst nuclear number is a reasonably good indicator of phylogenetic relatedness, with all the 4-nucleated cyst producers forming a robust clade (Clark, Kaffashian, et al. 2006). To date, only species in the 4-nucleated cyst clade have been grown successfully in the absence of bacteria. The species chosen for our genome survey analysis cover the range of diversity among Entamoeba species that can be grown in axenic culture (fig. 1). FIG. 1.— Open in new tabDownload slide Phylogenetic relationships among Entamoeba species. The tree depicted is redrawn from that in Clark, Kaffashian, et al. (2006). Species producing cysts with different nuclear number are indicated. Species referred to in this work are shown in boldface. The scale bar represents the evolutionary distance equivalent to 0.1 changes per site. FIG. 1.— Open in new tabDownload slide Phylogenetic relationships among Entamoeba species. The tree depicted is redrawn from that in Clark, Kaffashian, et al. (2006). Species producing cysts with different nuclear number are indicated. Species referred to in this work are shown in boldface. The scale bar represents the evolutionary distance equivalent to 0.1 changes per site. Entamoeba histolytica/E. dispar comparisons Phylogenetic analyses have consistently shown that E. histolytica and E. dispar are sibling species (Silberman et al. 1999; Clark, Kaffashian, et al. 2006). Indeed, it is only in the past 10–15 years that they have been widely accepted as being distinct (Diamond and Clark 1993). The assembly into arrays of tRNA gene-containing sequences from the E. dispar SAW760 genome survey allowed comparisons between these species to be performed. The gene content of the tRNA array units is identical in the 2 species. The only difference is that although most E. histolytica isolates (including the genome strain HM-1:IMSS) have 2 versions of the array containing AsnGTT and LysCTT genes ([NK1] and [NK2]), which differ completely in their STR regions and other intergenic sequences (Ali et al. 2005), the E. dispar SAW760 genome contains only 1 type of [NK] array. Thus, E. dispar SAW760 has 24 arrays rather than the 25 found in E. histolytica HM-1:IMSS. Note, however, that a few E. histolytica isolates have lost 1 array ([NK1]) and therefore have 24 arrays (Ali et al. 2005). It is possible that the single [NK] array in E. dispar SAW760 may also prove to be an exception. The tRNA/5S RNA gene content and orientation of all arrays is identical in the 2 species. However, the corresponding intergenic sequences generally share no similarity in either their STR or simple sequence regions. This contrasts dramatically with E. histolytica intraspecific variation where, despite substantial STR number variation, the simple sequence regions and the individual repeat sequences that make up the STRs are highly conserved (see below). The ability to design species-specific primers for amplification of intergenic regions from all isolates of each species also confirms that the sequences immediately adjacent to the genes are conserved within each species (Zaki et al. 2002; Ali et al. 2005). Entamoeba moshkovskii Organization Phylogenetic analyses have consistently shown that E. moshkovskii is closely related to E. histolytica and E. dispar (Silberman et al. 1999; Clark, Kaffashian et al. 2006). The organisms are morphologically identical and some E. moshkovskii subtypes have been shown to infect humans, causing a diagnostic problem (Ali et al. 2003). It was not surprising, therefore, to find that most arrays in E. moshkovskii FIC show the same gene content and orientation as in E. histolytica and E. dispar. Indeed, of the 23 arrays in E. moshkovskii 19 have the same organization as arrays in E. histolytica and E. dispar (table 1). There is one major difference, however. In no case is there any evidence for STRs in the intergenic regions. On average the E. moshkovskii array units are significantly smaller than their homologs in E. histolytica and E. dispar (mean size 64% and 61%, respectively). Table 1 Comparative Organization of tRNA Array Units in Entamoeba Species tRNA Isoacceptor type Entamoeba dispar SAW760a Entamoeba moshkovskii FIC Entamoeba terrapinae M Entamoeba invadens IP-1 AlaAGC [AAGC] [AAGC] [AAGC], [ADSS] [AAGC] AlaTGC [ASD] [ADSSD] [ADSD], [ADRML] [ASS] AlaCGC [ALL] [ALL] [ALLSL5] [ALI] ArgACG [R5] [R5] [IR] [VMEDR5E] ArgCCG # # # # ArgCCT [RT] [RT] [PTRL] [PPTRL] ArgTCG [MR] [MR] [MRM5], [ADRML] [MR] ArgTCT [RTCT] [RTCT] [RTCT] [RTCT] AsnGTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] AspGTC [ASD], [SD] [ADSSD] [ADSD], [ADSS], [ADRML] [VMEDR5E], [FVDTX]b, [EIDLLL] CysGCA [SPPCK], [SQCK] [SPPPCK], [SQCK] [CK], [SQCK] [NKQCK] GlnCTG [SQCK] [SQCK] [SQCK], [SQK] [NKQCK] GlnTTG [TQ] [TQ] [VQ51], [VQ52] [TQQ] GluCTC [VME5] [VME5] [VME] [VMEDR5E] [EIDLLL] GluTTC [YE] [YE] [YE] [VMEDR5E], [YE] GlyCCC # # # # GlyGCC [GGCC] [GGCC] [GGCC] [GGCC] GlyTCC [GTCC] [GTCC] [GTCC] #, #SGI HisGTG [HGTG] [HGTG] [HGTG] [HGTG] IleAAT [WI] [WI1], [WI2] [IR], [WI1], [WI2] [WI], [EIDLLL] IleTAT # # # [ALI], #SGI LeuAAG [LT] [LT], [PL] [PTRL], [LT], [PL] [PPTRL] LeuCAG [LS] [LS] [ALLSL5], [LLLSL5] [LS] LeuCAA [ALL] [ALL] [ALLSL5], [LLLSL5] [ALI], [EIDLLL] LeuTAA [ALL], # [ALL], # [ALLSL5], [ADRML], [LLLSL5] [EIDLLL] LeuTAG # # # # LysCTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] LysTTT [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [CK], [SQCK] [NKQCK] eMetCAT [MR] [MR] [MRM5], [ADRML] [MR] iMetCAT [VME5] [VME5], [MV5] [VME] [VMEDR5E] PheGAA [VF] [VF] [VF] [FVV5], [FVDTX]b ProCGG [SPPCK] [SPPPCK] [SPPP] [PPTRL] ProAGG [SPPCK] [SPPPCK] [SPPP] [SPP] ProTGG [PTGG] [SPPPCK] [PTRL], [PL], [SPPP] [SPP], [PPTRL] SerAGA [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [SQCK], [SPPP] [SPP] SerGCT [ASD] [ADSSD] [ADSD], [ADSS] [ASS] SerCGA [LS] [LS] [ALLSL5], [LLLSL5] [LS], #SGI SerTGA [SD] [ADSSD] [ADSS] [ASS] ThrAGT [LT] [LT], [RT] [PTRL], [LT] [PPTRL] ThrCGT [TQ] [TQ] # [TQQ] ThrTGT [TX]b [TX]b [TX]b [FVDTX]b TrpCCA [WI] [WI] [WI1], [WI2] [WI] TyrGTA [YE] [YE] [YE] [YE] ValCAC [VME5] [VME5] [VQ51], [VQ52] [FVV5] ValGAC [VF] [VF] [VF] [FVV5], [FVDTX]b ValTAC [V5] [MV5] [VME] [VMEDR5E] tRNA Isoacceptor type Entamoeba dispar SAW760a Entamoeba moshkovskii FIC Entamoeba terrapinae M Entamoeba invadens IP-1 AlaAGC [AAGC] [AAGC] [AAGC], [ADSS] [AAGC] AlaTGC [ASD] [ADSSD] [ADSD], [ADRML] [ASS] AlaCGC [ALL] [ALL] [ALLSL5] [ALI] ArgACG [R5] [R5] [IR] [VMEDR5E] ArgCCG # # # # ArgCCT [RT] [RT] [PTRL] [PPTRL] ArgTCG [MR] [MR] [MRM5], [ADRML] [MR] ArgTCT [RTCT] [RTCT] [RTCT] [RTCT] AsnGTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] AspGTC [ASD], [SD] [ADSSD] [ADSD], [ADSS], [ADRML] [VMEDR5E], [FVDTX]b, [EIDLLL] CysGCA [SPPCK], [SQCK] [SPPPCK], [SQCK] [CK], [SQCK] [NKQCK] GlnCTG [SQCK] [SQCK] [SQCK], [SQK] [NKQCK] GlnTTG [TQ] [TQ] [VQ51], [VQ52] [TQQ] GluCTC [VME5] [VME5] [VME] [VMEDR5E] [EIDLLL] GluTTC [YE] [YE] [YE] [VMEDR5E], [YE] GlyCCC # # # # GlyGCC [GGCC] [GGCC] [GGCC] [GGCC] GlyTCC [GTCC] [GTCC] [GTCC] #, #SGI HisGTG [HGTG] [HGTG] [HGTG] [HGTG] IleAAT [WI] [WI1], [WI2] [IR], [WI1], [WI2] [WI], [EIDLLL] IleTAT # # # [ALI], #SGI LeuAAG [LT] [LT], [PL] [PTRL], [LT], [PL] [PPTRL] LeuCAG [LS] [LS] [ALLSL5], [LLLSL5] [LS] LeuCAA [ALL] [ALL] [ALLSL5], [LLLSL5] [ALI], [EIDLLL] LeuTAA [ALL], # [ALL], # [ALLSL5], [ADRML], [LLLSL5] [EIDLLL] LeuTAG # # # # LysCTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] LysTTT [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [CK], [SQCK] [NKQCK] eMetCAT [MR] [MR] [MRM5], [ADRML] [MR] iMetCAT [VME5] [VME5], [MV5] [VME] [VMEDR5E] PheGAA [VF] [VF] [VF] [FVV5], [FVDTX]b ProCGG [SPPCK] [SPPPCK] [SPPP] [PPTRL] ProAGG [SPPCK] [SPPPCK] [SPPP] [SPP] ProTGG [PTGG] [SPPPCK] [PTRL], [PL], [SPPP] [SPP], [PPTRL] SerAGA [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [SQCK], [SPPP] [SPP] SerGCT [ASD] [ADSSD] [ADSD], [ADSS] [ASS] SerCGA [LS] [LS] [ALLSL5], [LLLSL5] [LS], #SGI SerTGA [SD] [ADSSD] [ADSS] [ASS] ThrAGT [LT] [LT], [RT] [PTRL], [LT] [PPTRL] ThrCGT [TQ] [TQ] # [TQQ] ThrTGT [TX]b [TX]b [TX]b [FVDTX]b TrpCCA [WI] [WI] [WI1], [WI2] [WI] TyrGTA [YE] [YE] [YE] [YE] ValCAC [VME5] [VME5] [VQ51], [VQ52] [FVV5] ValGAC [VF] [VF] [VF] [FVV5], [FVDTX]b ValTAC [V5] [MV5] [VME] [VMEDR5E] NOTE.—The tRNA gene content of each array is shown using the single-letter amino acid code. The 5S RNA gene is identified by “5” when present. Where 2 distinct variants of an array exist, the numbers 1 and 2 are appended to the gene complement. The E. invadens array containing only 5S RNA genes is not listed. # indicates dispersed tRNA genes. The dispersed cluster of 3 genes in E. invadens is indicated by #SGI. a Unit organization in Entamoeba histolytica HM-1:IMSS (Clark, Ali, et al. 2006) and E. dispar SAW760 is identical except for [NK], which exists as 2 distinct arrays in E. histolytica but only 1 in E. dispar. b X is a gene encoding the same unidentified small RNA in one array in each species (Banerjee and Lohia 2003; Clark, Ali, et al. 2006). Open in new tab Table 1 Comparative Organization of tRNA Array Units in Entamoeba Species tRNA Isoacceptor type Entamoeba dispar SAW760a Entamoeba moshkovskii FIC Entamoeba terrapinae M Entamoeba invadens IP-1 AlaAGC [AAGC] [AAGC] [AAGC], [ADSS] [AAGC] AlaTGC [ASD] [ADSSD] [ADSD], [ADRML] [ASS] AlaCGC [ALL] [ALL] [ALLSL5] [ALI] ArgACG [R5] [R5] [IR] [VMEDR5E] ArgCCG # # # # ArgCCT [RT] [RT] [PTRL] [PPTRL] ArgTCG [MR] [MR] [MRM5], [ADRML] [MR] ArgTCT [RTCT] [RTCT] [RTCT] [RTCT] AsnGTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] AspGTC [ASD], [SD] [ADSSD] [ADSD], [ADSS], [ADRML] [VMEDR5E], [FVDTX]b, [EIDLLL] CysGCA [SPPCK], [SQCK] [SPPPCK], [SQCK] [CK], [SQCK] [NKQCK] GlnCTG [SQCK] [SQCK] [SQCK], [SQK] [NKQCK] GlnTTG [TQ] [TQ] [VQ51], [VQ52] [TQQ] GluCTC [VME5] [VME5] [VME] [VMEDR5E] [EIDLLL] GluTTC [YE] [YE] [YE] [VMEDR5E], [YE] GlyCCC # # # # GlyGCC [GGCC] [GGCC] [GGCC] [GGCC] GlyTCC [GTCC] [GTCC] [GTCC] #, #SGI HisGTG [HGTG] [HGTG] [HGTG] [HGTG] IleAAT [WI] [WI1], [WI2] [IR], [WI1], [WI2] [WI], [EIDLLL] IleTAT # # # [ALI], #SGI LeuAAG [LT] [LT], [PL] [PTRL], [LT], [PL] [PPTRL] LeuCAG [LS] [LS] [ALLSL5], [LLLSL5] [LS] LeuCAA [ALL] [ALL] [ALLSL5], [LLLSL5] [ALI], [EIDLLL] LeuTAA [ALL], # [ALL], # [ALLSL5], [ADRML], [LLLSL5] [EIDLLL] LeuTAG # # # # LysCTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] LysTTT [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [CK], [SQCK] [NKQCK] eMetCAT [MR] [MR] [MRM5], [ADRML] [MR] iMetCAT [VME5] [VME5], [MV5] [VME] [VMEDR5E] PheGAA [VF] [VF] [VF] [FVV5], [FVDTX]b ProCGG [SPPCK] [SPPPCK] [SPPP] [PPTRL] ProAGG [SPPCK] [SPPPCK] [SPPP] [SPP] ProTGG [PTGG] [SPPPCK] [PTRL], [PL], [SPPP] [SPP], [PPTRL] SerAGA [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [SQCK], [SPPP] [SPP] SerGCT [ASD] [ADSSD] [ADSD], [ADSS] [ASS] SerCGA [LS] [LS] [ALLSL5], [LLLSL5] [LS], #SGI SerTGA [SD] [ADSSD] [ADSS] [ASS] ThrAGT [LT] [LT], [RT] [PTRL], [LT] [PPTRL] ThrCGT [TQ] [TQ] # [TQQ] ThrTGT [TX]b [TX]b [TX]b [FVDTX]b TrpCCA [WI] [WI] [WI1], [WI2] [WI] TyrGTA [YE] [YE] [YE] [YE] ValCAC [VME5] [VME5] [VQ51], [VQ52] [FVV5] ValGAC [VF] [VF] [VF] [FVV5], [FVDTX]b ValTAC [V5] [MV5] [VME] [VMEDR5E] tRNA Isoacceptor type Entamoeba dispar SAW760a Entamoeba moshkovskii FIC Entamoeba terrapinae M Entamoeba invadens IP-1 AlaAGC [AAGC] [AAGC] [AAGC], [ADSS] [AAGC] AlaTGC [ASD] [ADSSD] [ADSD], [ADRML] [ASS] AlaCGC [ALL] [ALL] [ALLSL5] [ALI] ArgACG [R5] [R5] [IR] [VMEDR5E] ArgCCG # # # # ArgCCT [RT] [RT] [PTRL] [PPTRL] ArgTCG [MR] [MR] [MRM5], [ADRML] [MR] ArgTCT [RTCT] [RTCT] [RTCT] [RTCT] AsnGTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] AspGTC [ASD], [SD] [ADSSD] [ADSD], [ADSS], [ADRML] [VMEDR5E], [FVDTX]b, [EIDLLL] CysGCA [SPPCK], [SQCK] [SPPPCK], [SQCK] [CK], [SQCK] [NKQCK] GlnCTG [SQCK] [SQCK] [SQCK], [SQK] [NKQCK] GlnTTG [TQ] [TQ] [VQ51], [VQ52] [TQQ] GluCTC [VME5] [VME5] [VME] [VMEDR5E] [EIDLLL] GluTTC [YE] [YE] [YE] [VMEDR5E], [YE] GlyCCC # # # # GlyGCC [GGCC] [GGCC] [GGCC] [GGCC] GlyTCC [GTCC] [GTCC] [GTCC] #, #SGI HisGTG [HGTG] [HGTG] [HGTG] [HGTG] IleAAT [WI] [WI1], [WI2] [IR], [WI1], [WI2] [WI], [EIDLLL] IleTAT # # # [ALI], #SGI LeuAAG [LT] [LT], [PL] [PTRL], [LT], [PL] [PPTRL] LeuCAG [LS] [LS] [ALLSL5], [LLLSL5] [LS] LeuCAA [ALL] [ALL] [ALLSL5], [LLLSL5] [ALI], [EIDLLL] LeuTAA [ALL], # [ALL], # [ALLSL5], [ADRML], [LLLSL5] [EIDLLL] LeuTAG # # # # LysCTT [NK] [NK] [NK1], [NK2] [NKQCK], [NK] LysTTT [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [CK], [SQCK] [NKQCK] eMetCAT [MR] [MR] [MRM5], [ADRML] [MR] iMetCAT [VME5] [VME5], [MV5] [VME] [VMEDR5E] PheGAA [VF] [VF] [VF] [FVV5], [FVDTX]b ProCGG [SPPCK] [SPPPCK] [SPPP] [PPTRL] ProAGG [SPPCK] [SPPPCK] [SPPP] [SPP] ProTGG [PTGG] [SPPPCK] [PTRL], [PL], [SPPP] [SPP], [PPTRL] SerAGA [SPPCK], [SQCK] [SPPPCK], [SQCK] [SQK], [SQCK], [SPPP] [SPP] SerGCT [ASD] [ADSSD] [ADSD], [ADSS] [ASS] SerCGA [LS] [LS] [ALLSL5], [LLLSL5] [LS], #SGI SerTGA [SD] [ADSSD] [ADSS] [ASS] ThrAGT [LT] [LT], [RT] [PTRL], [LT] [PPTRL] ThrCGT [TQ] [TQ] # [TQQ] ThrTGT [TX]b [TX]b [TX]b [FVDTX]b TrpCCA [WI] [WI] [WI1], [WI2] [WI] TyrGTA [YE] [YE] [YE] [YE] ValCAC [VME5] [VME5] [VQ51], [VQ52] [FVV5] ValGAC [VF] [VF] [VF] [FVV5], [FVDTX]b ValTAC [V5] [MV5] [VME] [VMEDR5E] NOTE.—The tRNA gene content of each array is shown using the single-letter amino acid code. The 5S RNA gene is identified by “5” when present. Where 2 distinct variants of an array exist, the numbers 1 and 2 are appended to the gene complement. The E. invadens array containing only 5S RNA genes is not listed. # indicates dispersed tRNA genes. The dispersed cluster of 3 genes in E. invadens is indicated by #SGI. a Unit organization in Entamoeba histolytica HM-1:IMSS (Clark, Ali, et al. 2006) and E. dispar SAW760 is identical except for [NK], which exists as 2 distinct arrays in E. histolytica but only 1 in E. dispar. b X is a gene encoding the same unidentified small RNA in one array in each species (Banerjee and Lohia 2003; Clark, Ali, et al. 2006). Open in new tab Entamoeba moshkovskii Isolate Differences To further investigate the differences in array unit organization between E. moshkovskii and E. histolytica/E. dispar, a few units were sequenced in other E. moshkovskii isolates including some that belong to a different subtype. Again, no STRs were observed; nevertheless, differences in the intergenic region were substantial, both in sequence and length (data not shown). For intergenic region W-I, 2 distinct unit variants were observed in both E. moshkovskii FIC and E. moshkovskii Laredo and the relative similarity suggests that the divergence between the unit variants postdated divergence of the E. moshkovskii strains. The intergenic regions are much more similar within than between strains, although both length and sequence differences are still present. This suggests either that duplication and divergence of arrays occurred independently, or that some “cross-talk” between the 2 arrays has occurred subsequent to strain divergence. There is no evidence of repeats of any type being responsible for the length differences observed in any of the intergenic regions for which sequences from multiple isolates are available. Entamoeba terrapinae and E. invadens These 2 reptilian parasites are not closely related to each other or to the 3 species already mentioned. Not surprisingly, their array organizations are also very different, with only 8 of the 26 distinct arrays in E. terrapinae M and 4 of the 20 arrays in E. invadens IP-1 having the same gene organization as in E. histolytica (table 1). Among the E. invadens arrays is the largest one seen to date with a unit size of 2364 bp, containing 6 tRNA genes and a 5S RNA gene (array [VMEDR5E]). One peculiarity of E. invadens is the existence of an array containing only 5S RNA genes. This is the arrangement of 5S RNA genes found in many other eukaryotes, including humans, but it is not present in the other Entamoeba species studied to date. There is no evidence for non-tRNA/5S RNA genes being part of arrays in the surveyed species, except for homologs of the gene encoding small RNA “X” previously identified in an E. histolytica array (Banerjee and Lohia 2003; Clark, Ali, et al. 2006). This ∼200-nt RNA is thought perhaps to be a small nuclear RNA and is always encoded in arrays adjacent to the gene for tRNA ThrTGT. It is notable that snRNAs have been found occasionally encoded in arrays with the 5S RNA genes of a few organisms, including certain fish (Manchado et al. 2006) and oysters (Cross and Rebordinos 2005). In the latter case there is also a microsatellite present. Cross-Species Comparison of Array Organization Despite the small number of array units with identical organization between the reptilian species and E. histolytica/E. dispar, it is nevertheless easy to see similarities—the context and orientation of adjacent tRNA genes is frequently conserved even if the overall unit organization differs (figs 2 and 3). Indeed, it is possible to produce a network of array structures in which similar tRNA gene arrangements are linked (supplementary figs. 1–5, Supplementary Material online). In several cases (e.g., ValTAC in fig. 2A and SerGCT in fig. 2B), the gene is in a distinct gene/array combination in each species (except E. histolytica/E. dispar) but the similarities in context are nevertheless clear. In most cases, the direction of the fission/fusion events cannot be stated with any certainty, but in 2 cases the order of events can be deduced. FIG. 2.— Open in new tabDownload slide Mapping of array organization onto phylogenetic tree. The relationships among surveyed species from figure 1 (Entamoeba invadens branch shortened for simplicity) are shown adjacent to a depiction of the corresponding array. Arrows indicate the orientation of the tRNA/5S RNA gene and contain the single-letter amino acid code and anticodon for the encoded tRNA. (A) The array unit organization involving the gene encoding tRNA ValTAC. (B) The array unit organization involving the genes encoding tRNAs SerGCT and SerTGA. FIG. 2.— Open in new tabDownload slide Mapping of array organization onto phylogenetic tree. The relationships among surveyed species from figure 1 (Entamoeba invadens branch shortened for simplicity) are shown adjacent to a depiction of the corresponding array. Arrows indicate the orientation of the tRNA/5S RNA gene and contain the single-letter amino acid code and anticodon for the encoded tRNA. (A) The array unit organization involving the gene encoding tRNA ValTAC. (B) The array unit organization involving the genes encoding tRNAs SerGCT and SerTGA. FIG. 3.— Open in new tabDownload slide Array organization of genes encoding ProTGG. The relationships among surveyed species plus Entamoeba ecuadoriensis from figure 1 (Entamoeba invadens branch shortened for simplicity) are shown adjacent to a depiction of the corresponding tRNA-Pro–encoding arrays. Arrows indicate the orientation of the tRNA gene and contain the single-letter amino acid code and anticodon for the encoded tRNA. FIG. 3.— Open in new tabDownload slide Array organization of genes encoding ProTGG. The relationships among surveyed species plus Entamoeba ecuadoriensis from figure 1 (Entamoeba invadens branch shortened for simplicity) are shown adjacent to a depiction of the corresponding tRNA-Pro–encoding arrays. Arrows indicate the orientation of the tRNA gene and contain the single-letter amino acid code and anticodon for the encoded tRNA. In E. histolytica/E. dispar, tRNA ProTGG is encoded on its own in a single-gene array, whereas in the other species it is part of a larger unit and adjacent to a gene encoding tRNA ProAGG (fig. 3). The species most closely related to E. histolytica/E. dispar is the little-studied organism E. ecuadoriensis EC. The equivalent array was cloned and sequenced from this species and its organization shown to be identical to that in E. moshkovskii, namely [SPPPCK]. The most parsimonious explanation for this distribution of organizations is that in the E. histolytica/E. dispar lineage, the tRNA ProTGG gene has “popped out” of the [SPPPCK] array that was present in the common ancestor they shared with E. ecuadoriensis to form a separate array on its own (fig. 3). Similarly, E. ecuadoriensis has an AspGTC gene between AlaTGC and SerGCT genes, as in E. moshkovskii (fig. 2B), again suggesting that fission of the [ADSSD] unit into [ASD] and [SD] occurred in the E. histolytica/E. dispar lineage after divergence from E. ecuadoriensis. Dispersed tRNA Genes In E. histolytica, 5 tRNA isoacceptor types were found to be encoded by nonarrayed genes, dispersed in small numbers throughout the genome. Most of the dispersed tRNA genes found in the other Entamoeba genomes encode the same tRNAs as in E. histolytica. However, there are some exceptions: tRNA ThrCGT is encoded as part of an array with tRNA GlnTTG in all species except E. terrapinae, and all species have a [GlyTCC] array except for E. invadens, where it is dispersed (table 1). Additionally, there are 2 types of tRNA LeuTAA gene that are distinct in sequence, one of which has an intron (LeuTAAi). In E. histolytica/E. dispar and in E. moshkovskii, the gene with the intron is dispersed and the other gene is arrayed. Both types of gene are also present in E. terrapinae and E. invadens, where they are both found in arrays (table 1). Cross-Species Comparison of Array Base Composition In the analysis of E. histolytica arrays (Clark, Ali, et al. 2006), it was noted that the regions of the arrays units between genes were A + T rich (on average 80.0%) and showed a pyrimidine/purine compositional bias (on average 70.8% pyrimidine on one strand). The arrays for the 4 new species were similarly analyzed (table 2). All intergenic regions were A + T rich, although those in E. moshkovskii and E. terrapinae were 10% lower in A + T than the other species. The pyrimidine/purine compositional bias was also strong in E. dispar and E. moshkovskii but was much less pronounced in E. invadens and essentially absent in E. terrapinae arrays. Although this bias is likely to affect DNA structure when present, the lack of conservation of this feature suggests that it is not essential to the present role of the arrays across all species. Table 2 Base Composition of Array Intergenic Regions Organism Mean A + T (%) Range A + T (%) Mean Pyrimidine (%) Range Pyrimidine (%) Entamoeba histolytica HM-1:IMSS 80.0 77.7–83.7 70.8 62.5–76.3 Entamoeba dispar SAW760 81.5 79.3–85.3 67.3 59.9–69.8 Entamoeba moshkovskii FIC 69.3 63.2–71.4 75.8 56.6–79.7 Entamoeba terrapinae M 71.2 67.6–74.3 53.2 50.4–56.8 Entamoeba invadens IP-1 81.8 66.4–84.7 56.9 51.0–74.7 Organism Mean A + T (%) Range A + T (%) Mean Pyrimidine (%) Range Pyrimidine (%) Entamoeba histolytica HM-1:IMSS 80.0 77.7–83.7 70.8 62.5–76.3 Entamoeba dispar SAW760 81.5 79.3–85.3 67.3 59.9–69.8 Entamoeba moshkovskii FIC 69.3 63.2–71.4 75.8 56.6–79.7 Entamoeba terrapinae M 71.2 67.6–74.3 53.2 50.4–56.8 Entamoeba invadens IP-1 81.8 66.4–84.7 56.9 51.0–74.7 Open in new tab Table 2 Base Composition of Array Intergenic Regions Organism Mean A + T (%) Range A + T (%) Mean Pyrimidine (%) Range Pyrimidine (%) Entamoeba histolytica HM-1:IMSS 80.0 77.7–83.7 70.8 62.5–76.3 Entamoeba dispar SAW760 81.5 79.3–85.3 67.3 59.9–69.8 Entamoeba moshkovskii FIC 69.3 63.2–71.4 75.8 56.6–79.7 Entamoeba terrapinae M 71.2 67.6–74.3 53.2 50.4–56.8 Entamoeba invadens IP-1 81.8 66.4–84.7 56.9 51.0–74.7 Organism Mean A + T (%) Range A + T (%) Mean Pyrimidine (%) Range Pyrimidine (%) Entamoeba histolytica HM-1:IMSS 80.0 77.7–83.7 70.8 62.5–76.3 Entamoeba dispar SAW760 81.5 79.3–85.3 67.3 59.9–69.8 Entamoeba moshkovskii FIC 69.3 63.2–71.4 75.8 56.6–79.7 Entamoeba terrapinae M 71.2 67.6–74.3 53.2 50.4–56.8 Entamoeba invadens IP-1 81.8 66.4–84.7 56.9 51.0–74.7 Open in new tab Entamoeba histolytica Intraspecific STR Variation During the development of a genotyping method (Zaki and Clark 2001; Ali et al. 2005), limited sequencing revealed that the basis of PCR product length variation between isolates was variation in the numbers of repeats in the inter-tRNA gene spacer regions. To investigate the patterns of variation, 3 of the STR loci used for genotyping were selected for study in greater depth: R-R from array [RTCT], STGA-D from array [SD], and N-K2 from array [NK2]. In almost every case, an unambiguous sequence was obtained from direct sequencing of the PCR products, indicating that the majority of copies in each array share the same sequence. Nevertheless, it must be remembered that each sequence is a “consensus.” Each of the 3 STR loci studied contains blocks of repeats that show both copy number and sequence variation. Within each block, there are often repeat sequence variants that differ by only 1 base. The start point of the repeat sequence is defined somewhat arbitrarily as “partial repeats” may also be present at the ends of each block. STR locus R-R The intergenic region in array [RTCT] contains 3 blocks with variable repeat copy numbers (Blocks A–C; fig. 4). Block A contains between 3 and 7 tandem copies of a single 8-base sequence. Adjacent to Block A is a tandem duplication of the same sequence that shows no copy number variation between isolates. Block B consists of two 8-base repeat sequence variants, one type occurring only at each end of the block, whereas the central variant can be present in either 3 or 4 copies. Three tandem copies of an 8-base sequence are always found between Blocks B and C. The 8-base repeats present in Blocks A and B and between Blocks B and C are all related in sequence. FIG. 4.— Open in new tabDownload slide Intraspecific differences in STR organization in the Entamoeba histolytica R-R intergenic region. Sequence types identified through sequencing of STR regions from different isolates are shown. Blocks of STRs are indicated, with distinct repeat sequences being assigned different shading. The arrow indicates the position of the tRNA gene and contain the 3-letter amino acid code and anticodon for the encoded tRNA. The complete [RTCT] array unit is depicted as there is only 1 tRNA gene in this array. FIG. 4.— Open in new tabDownload slide Intraspecific differences in STR organization in the Entamoeba histolytica R-R intergenic region. Sequence types identified through sequencing of STR regions from different isolates are shown. Blocks of STRs are indicated, with distinct repeat sequences being assigned different shading. The arrow indicates the position of the tRNA gene and contain the 3-letter amino acid code and anticodon for the encoded tRNA. The complete [RTCT] array unit is depicted as there is only 1 tRNA gene in this array. Block C contains one of the largest repeats found in E. histolytica, a 32-base sequence. This repeat may be present as a single copy plus 2 or 3 tandem copies or the tandem copies may be missing completely, apparently having been deleted in some isolates. The sequence of the 32-base Block C repeat clearly consists of four 8-base sequences related to those in the other blocks. However, the nature of the interstrain variation shows that it evolves as a 32-base sequence. In total, 12 different combinations of sequences for the 3 blocks have been identified in the [RTCT] intergenic region among the 136 sequences obtained to date. Of particular note is the fact that 3 pairs of sequence types have the same overall length (1RR/4RR, 2RR/5RR, and 11RR/12RR), resulting from compensating gains and losses of single repeats of 8 bases in Blocks A and B. This means that some PCR product sizes used in isolate identification can be derived from more than one sequence type. STR Locus STGA-D This intergenic region shows 2 variable blocks (fig. 5). Block B consists simply of an eight-base sequence that can be present in 2 or 3 copies. In contrast, Block A shows extensive variation. Two 9-base repeat sequence variants that differ at one position make up Block A. The first sequence variant may be present in anything from 3 to 19 copies. The second variant is present in either 1 or 2 copies and always at the same end of Block A. FIG. 5.— Open in new tabDownload slide Intraspecific differences in STR organization in the Entamoeba histolytica STGA-D intergenic region. Sequence types identified through sequencing of STR regions from different isolates are shown. Blocks of STRs are indicated, with distinct repeat sequences being assigned different shading. The arrows indicate the position of the tRNA genes and contain the 3-letter amino acid code and anticodon for the encoded tRNA. FIG. 5.— Open in new tabDownload slide Intraspecific differences in STR organization in the Entamoeba histolytica STGA-D intergenic region. Sequence types identified through sequencing of STR regions from different isolates are shown. Blocks of STRs are indicated, with distinct repeat sequences being assigned different shading. The arrows indicate the position of the tRNA genes and contain the 3-letter amino acid code and anticodon for the encoded tRNA. In total, 17 sequence types have been identified among the 128 intergenic regions sequenced to date, and once again certain distinct sequence types have essentially the same lengths as a result of compensating gains and losses of repeats (5SD/6SD, 9SD/11SD, 10SD/12SD/14SD, and 13SD/15SD). STR locus N-K2 This intergenic region shows the greatest degree of diversity of the 3 for which substantial numbers of sequences are available, with 18 sequence types detected among the 53 sequences obtained (fig. 6). Again 2 blocks are present and each consists of two 8-base sequence variants that differ at one position. Although Block B is found in only 3 forms, Block A can contain from 10 to 32 copies of the repeats. Five of the sequence types are of the same length (2NK-5NK and 7NK) as are 2 other pairs (10NK/11NK and 14NK/15NK). In addition to copy number variation, different patterns of sequence-variant interspersion contribute to the diversity. FIG. 6.— Open in new tabDownload slide Intraspecific differences in STR organization in the Entamoeba histolytica N-K2 intergenic region. Sequence types identified through sequencing of STR regions from different isolates are shown. Blocks of STRs are indicated, with distinct repeat sequences being assigned different shading. The arrows indicate the position of the tRNA genes and contain the 3-letter amino acid code and anticodon for the encoded tRNA. FIG. 6.— Open in new tabDownload slide Intraspecific differences in STR organization in the Entamoeba histolytica N-K2 intergenic region. Sequence types identified through sequencing of STR regions from different isolates are shown. Blocks of STRs are indicated, with distinct repeat sequences being assigned different shading. The arrows indicate the position of the tRNA genes and contain the 3-letter amino acid code and anticodon for the encoded tRNA. Other Diversity Most of the sequence diversity detected in the intergenic regions is confined to copy number variation in the repeat blocks. However, a number of point mutations have also been detected, primarily in the nonrepeated regions (data not shown). These are mostly transitions, but occasional transversions and indels have been noted. The mutations are scattered across the intergenic sequences and most have been seen in only 1 isolate. As well as point mutations observed when comparing sequences between isolates, sequencing of individual clones derived from the same PCR product also reveals point mutation differences between individual unit copies from the same array. That these are not amplification artifacts is confirmed by detection of similar features in clones sequenced during the genome project as these were derived from sheared genomic DNA rather than PCR products. Point mutations as well as occasional differences in STR number are observed in the individual reads, although they represent a small minority of the sequences in each case. Geographic Distribution Eleven out of the 12 R-R sequence types were found among the 84 Bangladeshi samples studied, whereas 9 were found among the 54 samples from other countries. This suggests either that most sequence types have a wide geographic distribution or that they have arisen independently in multiple locations. A similar picture emerges from the other loci: 14 out of the 17 S-D sequence types were found in the 80 Bangladeshi samples, whereas 10 were found in 48 samples from elsewhere; 12 out of the 18 N-K2 sequence types were found in 36 Bangladeshi samples with 9 being found in the 18 samples from elsewhere in the world. Discussion Comparative Organization Versus Phylogeny All of the Entamoeba species studied here belong to the 4-nucleated cyst-producing clade, and so at present we cannot say whether tRNA arrays are a characteristic of all species in the genus. Through the E. histolytica genome project and the genome surveys of 4 additional species reported here, we have a complete picture of the tRNA array organization in a cross section of species from this clade. It is possible that a few undetected single-copy dispersed tRNA or 5S RNA genes may exist in the surveyed species as the genome coverage in each case is low (ca. 0.5×). However, it seems unlikely that arrays will have been missed in the surveys and examples of all tRNA isoacceptor types detected in the E. histolytica genome have also been found in the other species. In addition to the surveys, examples of tRNA array units from a few additional species have been sequenced. This was undertaken to clarify the patterns of evolution in the case of E. ecuadoriensis and to confirm the presence of arrays in the case of E. equi, the latter being the earliest branch of the 4-nucleated cyst clade in the most recent phylogenetic tree (Clark, Kaffashian, et al. 2006). Several array units have the same organization in all species examined, which suggests that these may be useful starting points for examining tRNA array organization in species outside this clade. Only in E. histolytica and E. dispar are significant STRs observed in all array units. In other species, small numbers of tandem repeats are seen occasionally in the intergenic regions but there is no consistency to their presence. The one array unit sequenced in E. equi, [HGTG], exists in 2 variants and comes the closest to having a STR as there are 4 tandem copies of a 6-base sequence in each. In E. ecuadoriensis, 3 tandem copies of a 10-base sequence occur in array [SPPPCK], and in E. invadens, 3 tandem copies of an 8-base sequence occur in array [MR]. The other tandem repeats observed are very short (4 bases or less) and few in number or are simple duplications, probably arising by chance through mutation. For 2 species, multiple isolates have been examined at several loci. It was already clear from limited sequencing that the main source of interisolate variation in E. histolytica tRNA arrays was variation in repeat number (Zaki and Clark 2001). The data presented here give an in-depth insight into 3 intergenic regions. However, they also raise new questions. It is unclear why different blocks of repeats in the same intergenic region show dramatically different patterns of diversity. For example, N-K2 has 2 blocks of repeats, 1 of which exists in 11 length variants and the other in only 2, despite the fact that the repeat size is the same and the blocks are separated by only 82 bases. Across all 3 intergenic regions for which significant data are available, there is no obvious pattern. The repeat size, base composition, and relative position in the array unit do not appear to determine whether a block is going to be variable or stable in repeat number. In E. moshkovskii, a smaller number of isolates were compared. The underlying basis for the length variation observed is much less clear as no obvious repeat structure can be discerned in the intergenic regions of the array units studied. There is substantial sequence divergence between isolates. In part, this reflects the fact that E. moshkovskii is actually a species complex (Clark and Diamond 1997), but even between the sequences that can be aligned well the origins of the observed length differences remains unclear. It is notable that substantial differences in tRNA gene number appear to exist between species. The percentage of reads encoding tRNAs varies over 10-fold, from 1.1% in E. terrapinae to 15.7% in E. dispar, indicating a range of between 500 and 7000 tRNA genes, assuming that no differences in “clonability” exist between arrays within or between species. That clonability is a potential confounding factor is illustrated by the observation that no sequences in the E. terrapinae survey contained HisGTG genes; that array sequence was obtained by direct sequencing of PCR products. If correct, the numbers imply significantly fewer tRNA genes and also shorter arrays in E. moshkovskii, as the average unit length differs less than 2-fold between E. moshkovskii and E. dispar. Nevertheless, even in E. moshkovskii, the total number of tRNA genes in the genome is likely to be substantially larger than in most eukaryotes. Origins At present, we do not have data on the tRNA gene organization in subgroups of the genus Entamoeba other than the 4-nucleated cyst-producing clade, nor do we have data on the tRNA gene organization in subgroups of the genus Entamoeba other than the 4-nucleated cyst-producing clade nor do we have any indication of their organization in the most closely related genera identified to date (Endolimax, Pelomyxa, and Mastigamoeba; Silberman et al. 1999; Nikolaev et al. 2006). The origin of this unique gene organization thus cannot be deduced by comparison to related organisms. The origin of the STRs within E. histolytica and E. dispar arrays may be less obscure if only by analogy to what has been observed in other systems. The inter-tRNA regions in these species are extremely A + T rich. As a result, short tandem “duplications” are likely to occur quite frequently by chance mutation. Strand slippage during DNA replication or repair then has the potential to result in repeat propagation (Ellegren 2004), resulting in the STR loci observed. Strand slippage is also likely to be the mechanism generating the interstrain diversity observed. Of interest in this context is the observation that active tRNA genes appear to be “hot spots” for recombination, probably due to RNA polymerase III transcription complexes forming replication fork barriers, leading to replication pauses and resulting in “fragile” DNA (Deshpande and Newlon 1996; Labib and Hodgson 2007). This link of recombination to tRNA genes has been observed in both yeast (where it can result in chromosome rearrangements [Dunham et al. 2002; Admire et al. 2006; Pratt-Hyatt et al. 2006] and retrotransposon integration [Kim et al. 1998]) and Dictyostelium (where tRNA genes are closely linked to retrotransposon integration sites [Szafranski et al. 1999; Siol et al. 2006]). Retrotransposon sequences are frequently found in the regions immediately flanking E. histolytica tRNA arrays (Clark, Ali, et al. 2006) suggesting that a similar link exists here also. Pausing of replication may also increase the likelihood of strand slippage (Ellegren 2004), which is probably essential for tRNA array maintenance. That E. histolytica (and presumably other species of Entamoeba) can undergo DNA repeat expansion was observed as long ago as 1992, when it was reported (Bhattacharya et al. 1992) that repeat numbers were increasing (and decreasing) in a region of the ribosomal DNA episomal circle over quite a short period of time in continuous culture. The repeats in that case were 170 bp in size, intermediate between the STR and tRNA array unit sizes. Whether comparable tRNA array unit copy number changes are occurring in culture is not known at this stage, but based on PCR product sizes, the average number of STRs in intergenic regions does not show rapid variation, the patterns observed in E. histolytica being stable across several years in continuous culture (Ali et al. 2005) and during axenization (Clark, unpublished data). It is not difficult to envision how new tRNA genes can get incorporated into an existing array in a 2-step process. The new genes need to become linked to one unit in an array. This could occur by recombination among distinct array units or by insertion of reverse transcribed tRNAs (t-SINEs; Piskurek et al. 2003). The process responsible for homogenization of repeats in a tandem array is most likely concerted evolution. Concerted evolution has been most extensively studied in the rDNA of eukaryotes (Eickbush and Eickbush 2007) because these genes are usually (although not in Entamoeba) arranged in long tandem arrays as seen for the tRNA arrays described here. This gene conversion-based process will occasionally lead to the spread of that variant unit and the replacement of the existing arrayed copies. The spread of variants is a random process not requiring a selective advantage for the new unit structure. Other types of mechanism could also be involved, such as unequal sister chromatid exchange. The reverse process is more difficult to envision, where one or more genes “leave” an existing array to form a separate array unit. This has occurred in 2 cases described here, the clearest being the ProTGG gene, which has popped out of array [SPPPCK] to form its own array in the E. histolytica/E. dispar lineage (fig. 3). Several ways in which this could have occurred can be envisioned. One or more ProTGG genes may have become “dispersed” from the array, before subsequently being eliminated from [SPPPCK]. The existence of dispersed copies would make elimination of the ProTGG gene from the array selectively neutral. A dispersed gene could then become rearrayed. Alternatively, 2 [SPPPCK] arrays may have coexisted in the genome and have undergone complementary gene eliminations, leading to [ProTGG] in one case and [SPPCK] in the other—no dispersed genes are required in this version. The apparent secondary dispersal of the LeuTAAi gene in the lineage leading to E. histolytica/E. dispar/E. moshkovskii may reflect the first stage of another such event occurring via the first scenario, although it is also possible that incorporation of this gene into arrays has occurred independently in E. invadens and E. terrapinae. Clustering of tRNA genes also occasionally occurs in the absence of the array organization, as 3 dispersed tRNA genes were found within a 0.6-kb region in E. invadens (#SGI in table 1). This may be indicative of the first stage in array formation. Function The function of tRNA arrays in Entamoeba species remains obscure. The existence of low copy-number dispersed genes encoding certain tRNAs suggests that large gene numbers are not needed to provide sufficient tRNA for efficient translation, and there is no link between codon usage and tRNA gene number (Clark, Ali, et al. 2006). It is impossible to know whether all of the arrayed genes are transcribed, but there is no clear evidence from the surveys for any copies being nonfunctional pseudogenes, and it is difficult to envision a 100-fold difference in turnover rate between tRNA types. More likely is the possibility that many copies in an array exist in a transcriptionally inactive form due to differences in chromatin modification. The first potential function to have been proposed is based on the observation of S/MAR consensus sequences in some of the E. histolytica array units (Banerjee and Lohia 2003). This consensus is a degenerate and A + T-rich sequence and was shown to occur frequently by chance in random sequences of the same base composition (Clark, Ali, et al. 2006). To further investigate this possible role, array units from E. dispar were analyzed using MARSCAN. Although S/MAR consensus sequences were found, they were frequently in different array units to those identified in E. histolytica. As the organization of the arrays is fully conserved in the 2 species, the lack of S/MAR conservation casts doubt on there being a matrix-binding role for the arrays. A more recent proposal that the tRNA arrays may be located at chromosome ends and represent functional replacements for more traditional telomere repeats cannot yet be tested (Clark, Ali, et al. 2006). Assembly of the E. histolytica genome is not yet complete (Loftus et al. 2005), although efforts to achieve this are ongoing, and the coverage of the other genomes is too low to permit any testing of this theory. No telomerase or telomere-like sequences were found in the E. histolytica genome nor are they present in the genome surveys. Because E. histolytica chromosomes do not condense, cytogenetic localization of tRNA arrays to chromosome ends will not be possible. If the tRNA arrays are indeed telomeric, they face the same challenge as the structures they replace—how to replicate the ends without constant array length reduction. Conclusions tRNA arrays appear to be a general feature of Entamoeba species’ genomes, but the highly polymorphic STRs are largely confined to E. histolytica and E. dispar. The degree of similarity in organization reflects the phylogenetic relatedness of the species as indicated by small subunit ribosomal RNA analyses. Although most of our initial objectives for this project have been achieved, many questions regarding the origin and function of the tRNA arrays remain. This research was supported by grants from the Biotechnology and Biological Sciences Research Council (UK) (grant #G18391; genome surveys) and from the Wellcome Trust (grant #067314; E. histolytica intraspecific variation) awarded to C.G.C. We thank Natasha Larke, Sarah Sharp, Claire Arrowsmith, Doug Ormond, Zahra Hance, Barbara Harris, David Harris, and Carol Churcher for contributions to the genome surveys. Portions of this work formed part of the PhD thesis of I.K.M.A and the MSc thesis of C.S. References Admire A , Shanks L , Danzl N , Wang M , Weier U , Stevens W , Hunt E , Weinert T . Cycles of chromosome instability are associated with a fragile site and are increased by defects in DNA replication and checkpoint controls in yeast , Genes Dev , 2006 , vol. 20 (pg. 159 - 173 ) Google Scholar Crossref Search ADS PubMed WorldCat Ali IKM , Hossain MB , Roy S , Ayeh-Kumi PF , Petri WA Jr , Haque R , Clark CG . Entamoeba moshkovskii infections in children, Bangladesh , Emerg Infect Dis , 2003 , vol. 9 (pg. 580 - 584 ) Google Scholar Crossref Search ADS PubMed WorldCat Ali IKM , Zaki M , Clark CG . Use of PCR amplification of tRNA gene-linked short tandem repeats for genotyping Entamoeba histolytica , J Clin Microbiol , 2005 , vol. 43 (pg. 5842 - 5847 ) Google Scholar Crossref Search ADS PubMed WorldCat Banerjee S , Lohia A . Molecular analysis of repetitive DNA elements from Entamoeba histolytica, which encode small RNAs and contain matrix/scaffold attachment recognition sequences , Mol Biochem Parasitol , 2003 , vol. 126 (pg. 35 - 42 ) Google Scholar Crossref Search ADS PubMed WorldCat Benson G . Tandem repeats finder: a program to analyze DNA sequences , Nucleic Acids Res , 1999 , vol. 27 (pg. 573 - 580 ) Google Scholar Crossref Search ADS PubMed WorldCat Bhattacharya S , Bhattacharya A , Diamond LS . Entamoeba histolytica extrachromosomal circular ribosomal DNA: analysis of clonal variation in a hypervariable region , Exp Parasitol , 1992 , vol. 74 (pg. 200 - 204 ) Google Scholar Crossref Search ADS PubMed WorldCat Clark CG , Ali IKM , Zaki M , Loftus BJ , Hall N . Unique organisation of tRNA genes in Entamoeba histolytica , Mol Biochem Parasitol , 2006 , vol. 146 (pg. 24 - 29 ) Google Scholar Crossref Search ADS PubMed WorldCat Clark CG , Diamond LS . Intraspecific variation and phylogenetic relationships in the genus Entamoeba as revealed by riboprinting , J Euk Microbiol , 1997 , vol. 44 (pg. 142 - 154 ) Google Scholar Crossref Search ADS PubMed WorldCat Clark CG , Diamond LS . Methods for cultivation of luminal parasitic protists of clinical importance , Clin Microbiol Rev , 2002 , vol. 15 (pg. 329 - 341 ) Google Scholar Crossref Search ADS PubMed WorldCat Clark CG , Kaffashian F , Tawari B , et al. (13 co-authors) New insights into the phylogeny of Entamoeba species provided by analysis of four new small-subunit rRNA genes , Int J Syst Evol Microbiol , 2006 , vol. 56 (pg. 2235 - 2239 ) Google Scholar Crossref Search ADS PubMed WorldCat Corpet F . Multiple sequence alignment with hierarchical clustering , Nucleic Acids Res , 1988 , vol. 16 (pg. 10881 - 10890 ) Google Scholar Crossref Search ADS PubMed WorldCat Cross I , Rebordinos L . 5S rDNA and U2 snRNA are linked in the genome of Crassostrea angulata and Crassostrea gigas oysters: does the (CT)n. (GA)n microsatellite stabilize this novel linkage of large tandem arrays? , Genome , 2005 , vol. 48 (pg. 1116 - 1119 ) Google Scholar Crossref Search ADS PubMed WorldCat Deshpande AM , Newlon CS . DNA replication fork pause sites dependent on transcription , Science , 1996 , vol. 272 (pg. 1030 - 1033 ) Google Scholar Crossref Search ADS PubMed WorldCat Diamond LS , Clark CG . A redescription of Entamoeba histolytica Schaudinn, 1903 (Emended Walker, 1911) separating it from Entamoeba dispar Brumpt, 1925 , J Euk Microbiol , 1993 , vol. 40 (pg. 340 - 344 ) Google Scholar Crossref Search ADS PubMed WorldCat Dunham MJ , Badrane H , Ferea T , Adams J , Brown PO , Rosenzweig F , Botstein D . Characteristic genome rearrangements in experimental evolution of Saccharomyces cerevisiae , Proc Natl Acad Sci USA , 2002 , vol. 99 (pg. 16144 - 16149 ) Google Scholar Crossref Search ADS PubMed WorldCat Eichinger L , Pachebat JA , Glockner G , et al. (97 co-authors) The genome of the social amoeba Dictyostelium discoideum , Nature , 2005 , vol. 435 (pg. 43 - 57 ) Google Scholar Crossref Search ADS PubMed WorldCat Eickbush TH , Eickbush DG . Finely orchestrated movements: evolution of the ribosomal RNA genes , Genetics , 2007 , vol. 175 (pg. 477 - 485 ) Google Scholar Crossref Search ADS PubMed WorldCat Ellegren H . Microsatellites: simple sequences with complex evolution , Nature Rev Genet , 2004 , vol. 5 (pg. 435 - 445 ) Google Scholar Crossref Search ADS PubMed WorldCat Kim JM , Vanguri S , Boeke JD , Gabriel A , Voytas DF . Transposable elements and genome organization: a comprehensive survey of retrotransposons revealed by the complete Saccharomyces cerevisiae genome sequence , Genome Res , 1998 , vol. 8 (pg. 464 - 478 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat Labib K , Hodgson B . Replication fork barriers: pausing for a break or stalling for time? , EMBO Rep , 2007 , vol. 8 (pg. 346 - 353 ) Google Scholar Crossref Search ADS PubMed WorldCat Loftus B , Anderson I , Davies R , et al. (54 co-authors) The genome of the protist parasite Entamoeba histolytica , Nature , 2005 , vol. 433 (pg. 865 - 868 ) Google Scholar Crossref Search ADS PubMed WorldCat Lowe T , Eddy SR . tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence , Nucleic Acids Res , 1997 , vol. 25 (pg. 955 - 964 ) Google Scholar Crossref Search ADS PubMed WorldCat Manchado M , Zuasti E , Cross I , Merlo A , Infante C , Rebordinos L . Molecular characterization and chromosomal mapping of the 5S rRNA gene in Solea senegalensis: a new linkage to the U1, U2, and U5 small nuclear RNA genes , Genome , 2006 , vol. 49 (pg. 79 - 86 ) Google Scholar Crossref Search ADS PubMed WorldCat Nikolaev SI , Berney C , Petrov NB , Mylnikov AP , Fahrni JF , Pawlowski J . Phylogenetic position of Multicilia marina and the evolution of Amoebozoa , Int J Syst Evol Microbiol , 2006 , vol. 56 (pg. 1449 - 1458 ) Google Scholar Crossref Search ADS PubMed WorldCat Piskurek O , Nikaido M , Boeadi , Baba M , Okada N . Unique mammalian tRNA-derived repetitive elements in dermopterans: the t-SINE family and its retrotransposition through multiple sources , Mol Biol Evol , 2003 , vol. 20 (pg. 1659 - 1668 ) Google Scholar Crossref Search ADS PubMed WorldCat Pratt-Hyatt MJ , Kapadia KM , Wilson TE , Engelke DR . Increased recombination between active tRNA genes , DNA Cell Biol , 2006 , vol. 25 (pg. 359 - 364 ) Google Scholar Crossref Search ADS PubMed WorldCat Silberman JD , Clark CG , Diamond LS , Sogin ML . Phylogeny of the genera Entamoeba and Endolimax as deduced from small subunit ribosomal RNA gene sequence analysis , Mol Biol Evol , 1999 , vol. 16 (pg. 1740 - 1751 ) Google Scholar Crossref Search ADS PubMed WorldCat Siol O , Boutliliss M , Chung T , Glockner G , Dingermann T , Winckler T . Role of RNA polymerase III transcription factors in the selection of integration sites by the Dictyostelium non-long terminal repeat retrotransposon TRE5-A , Mol Cell Biol , 2006 , vol. 26 (pg. 8242 - 8251 ) Google Scholar Crossref Search ADS PubMed WorldCat Szafranski K , Glockner G , Dingermann T , Dannat K , Noegel AA , Eichinger L , Rosenthal A , Winckler T . Non-LTR retrotransposons with unique integration preferences downstream of Dictyostelium discoideum tRNA genes , Mol Gen Genet , 1999 , vol. 262 (pg. 772 - 780 ) Google Scholar Crossref Search ADS PubMed WorldCat Zaki M , Clark CG . Isolation and characterization of polymorphic DNA from Entamoeba histolytica , J Clin Microbiol , 2001 , vol. 39 (pg. 897 - 905 ) Google Scholar Crossref Search ADS PubMed WorldCat Zaki M , Meelu P , Sun W , Clark CG . Simultaneous differentiation and typing of Entamoeba histolytica and Entamoeba dispar , J Clin Microbiol , 2002 , vol. 40 (pg. 1271 - 1276 ) Google Scholar Crossref Search ADS PubMed WorldCat Author notes 1 Equal contribution to this work. 2 Present address: Bayer Schering Pharma, Strawberry Hill, RG14 1JA, United Kingdom. 3 Present address: Division of Infectious Diseases and International Health, University of Virginia Health System. 4 Present address: Queen Mary's School of Medicine and Dentistry, Institute of Cell and Molecular Science, Centre for Cutaneous Research, Whitechapel, London, United Kingdom. 5 Present address: School of Biological Sciences, University of Liverpool, Biosciences Building, Liverpool, United Kingdom. Laura Katz, Associate Editor © 2007 The Authors. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. © 2007 The Authors.