The set and the structure of the polychaet’s Hox clusters

Hox-genes in the ontogenesis of polychaetes

 

Milana A. Kulakova*

Department of Embriology

Saint Petersburg University

Saint Petersburg

Russia

 

 

Abstract

 

The basic plane and evolution of bilateral animals (Bilateria) are closely associated with Hox-genes. In the general case these genes exist in the genome in the form of clusters - complexes of genes with a conservative position of individual genes inside. These genes code transcription factors, containing homeodomain and their expression is ordered along a main body axis in accordance with a colinearity rule. This means that position of gene in cluster determines his spatial and sometimes temporary expression in embryo. In vertebrates and arthropods, which are investigated most detailed, Hox genes define a morphological distinction between different regions of body (tagmas). Boundaries of Hox genes expression are often coincide with borders of tagmas. In addition, these genes also have function in adult organisms, where they are involved in regeneration processes and maintain a tissue homeostasis. In recent years, there was an emergence of interest in the study of Hox genes in animals that do not belong to the traditional models of molecular biology (Drosophila, mouse, Xenopus). The Polychaetes belong to the most intriguing objects for such studies. These are members of the third, poorly studied evolutionary branch of Bilateria - Lophotrochozoa, which itself makes them attractive both for molecular biologists and developmental biologists. Among of the polychaetes, there are groups (families) with morphologically specialized segments, which are grouped in the tagmas (Chaetopteridae) and families with a large number of identical segments (Nereididae). There are species with impressive abilities to regeneration of the head and tail and those that are incapable of regeneration at all. In addition, often a larva exists within the life cycle of polychaetes. The larva may be radically different from an adult worm. How work the Hox genes in such different systems? Can we compare some of their functions with the functions of homologous genes in insects and vertebrates? What is their place in the hierarchy of molecular regulators, which controlling developmental processes in the polychaets? These questions arose not because of idle scientific curiosity. If we can correctly answer them, many pieces of bilaterian animals earliest evolution jigsaw puzzle will fall into place. In this review, I attempted to collect all the modern data on Hox genes in polychaetes. Also there was made some cautious assumptions about the ancestral functions of the Hox cluster.

 

Keywords: Annelida,Hox genes, UrBilateria, Larval development, Regeneration.

Introduction.

This chapter deals with Hox-genes, their structural organization and function features in polychaetes ontogenesis. Within hundreds of evolutionary conservative gene families Hox-genes are particularly important for understanding the relation between individual development and the macroevolution. Edward B. Lewis, the American genetic scientist (Nobel Prize in Medicine 1995) was first who discovered this relation. In 1987 he published in Nature the results of long-term study about segmentation control pattern in Drosophila (Lewis, 1987). Lewis used mutation analysis to consider the structure and the function of specific bithorax locus (a gene complex bithorax or BX-C). As early as 1915 Calvin Bridges (who worked in Thomas Hunt Morgan’s laboratory) isolated and described strange bithorax mutation at this locus in Drosophila. This was a classic homeotic mutation in which third thoracal segment was partially transformed into the second one. In this case the anterior part of haltere was turned into the anterior part of the wing (Bridges and Morgan, 1923).

Lewis knew nothing about nature of factors which are encoded in sequence BX-C, but he made some important conclusions. Most of them proved true brilliantly. There are some of them:

1) BX-C genes form a cluster – the complex of genes, which emerged as a result of tandem duplications of ancestor gene with subsequent mutational divergence of its offsprings;

2) There is direct link of gene’s position at the chromosome with spatial order of gene’s activation (now we call this link by the term “colinearity”);

3) Diptera originated from primitive four-winged ancestors, and insects originated from primitive arthropods, which had legs at all abdominal segments. During the evolution “leg-suppressing” genes emerged, which suppressed legs development at abdominal segments. Also emerged “haltere-promoting” genes which suppressed wing development at the third thoracal segment. The loss-of-function mutation in BX-C may lead to primitive state recapitulation, i.e. to the appearance of four-winged and many-legged flies.

Now we know that in the insect’s evolution there wasn’t an emerging of new Hox-genes, instead “old” genes assumed new development program control functions. However, the logical mainstream used by Lewis, was true.

Just after the publication of this key article, researchers had obtained the tool, which directly linked the molecular control of development and the phylogenetic evolution. There was born the new science – the Evolutionary Developmental biology or EvoDevo.

The search, the cloning and the sequencing of BX-C sequences and of other Drosophila genes (which are causing homeotic transformation) was performed independently by three scientist groups in California, Indiana and Basel (cit. ex Papageorgiou, 2007). As Lewis had predicted, these genes appeared to be serial homologs. It became known that they encode transcription factors with distinctive DNA-binding protein domain – the homeodomain. The nucleotide fragment that encoding this homeodomain, consists of 180 base pairs and is named homeobox. Hox -genes (from words “homeotic” and “homeobox”) in Drosophila are organized in two complexes: ANT-C and BX-C. Later a set of homologous Hox-genes was found in Vertebrates (Ranginib et al., 1989; Godsave et al., 1994; Burke et al., 1995). The mutations of these genes also led to spatial shifts in structures that located along body axis. It suggested that Hox-genes had originated not in insects. Their evolutionary age appeared to be much older. Hox-genes originated in the remote past, deep in Precambrian time. It was about 600 millions years ago (mya), when supposedly lived the last common ancestor of chordates and arthropods.

There was analysis of Hox amino-acid sequences of homeodomains and flanking areas in animals from different evolutionary branches. Its results dramatically changed the phylogeny of bilateral animals (deRosa et al., 1999). It appears that evolutionary tree of bilateral animals is divided near the basement to two stems: Deuterostomia (Chordates, Hemichordates, Echinoderms) and Protostomia. The last ones in turn are divided to Ecdysozoa (Arthropods, Onychophorans, Nematodes, Priapulida and others) and Spiralia (Brachiopods, Nemerteans, Annelids, Mollusca, Platyhelminthes, Rotifers and others) (deRosaetal., 1999). The term “Ecdysozoa” derives from Greek word “ecdysis” for molting, because all Ecdysozoa animals are molting ones, covered by hard cuticle. Their growth is accompanied by periodical molting, when old covers are dropped and the animal is growing quickly (more truly straighten itself up until new cover harden). The Spiralia branch named after common stereotype of early embryonic cleavage. This branch is most heterogeneous and interesting. It include Lophotrochozoa, Platyhelminthes, Gastrotricha, Syndermata and Gnathostomulida (Struck et al., 2014). The Lophotrochozoa group especially intrigues evolutionary biologists, because there is highest divergence of body organization plans. Indeed octopus, murex and ostrea are representatives of same taxon Mollusca, but it is hard to imagine by looking at their appearance. The term Lophotrochozoa itself derives from two other zoological terms: Lopho phorata and Trocho phora. Lophophorata (Brachiopoda, Bryozoa and Phoronida) are animals with lophophore – the especial organ for feed by particles which are suspended in water. Earlier Lophophorata was included in Deuterostomia. The Trochophora is ciliary spherical larva, tipically occurring in Lophotrochozoa (Annelids, Mollusca).

This new “molecular” phylogeny had fitted with previously built phylogeny by 18s rDNA (Field et al., 1988, Aguinaldo et al., 1997) and by mitochondrial DNA (Cohen et al., 1998). The comparison study of Hox-genes sequences suggested that common ancestor of bilateral animals (located at the basement of all three evolutionary branches) already had had at least seven Hox-genes (deRosaetal., 1999). How looked this common ancestor, what processes had led to its appearance and what is the cause of great evolutionary radiation of its descendants – these are the questions which consist a paradigm of modern science about Metazoa evolution.

During last ten years, mechanisms underlying morphological evolution became mostly clear. New outstanding methods emerged, which allow to delicately tune the work of gene of interest, to turn off this gene in suitable time of development, to stimulate its ectopic expression and even to edit a genome in vivo (CRISPR/Cas9 technology). However our knowledge about Hox-gene’s functions and even about patterns of their expression is unequal for different Bilateria branches. While Hox-gene functions within Deuterostomia and Ecdysozoa have been studied in detail at least et the vertebrates and arthropods examples, in contrast these researches of Lophotrochozoa branch are at the beginning stage. Meanwhile the set of molecular tools for control over morphogenesis of these animals is same as in Drosophila or in human. The difference is in strategy of these tools using. If we pay more attention to Lophotrochozoa group, we’ll meet wonderful discoveries in regulatory evolution and epigenetics.

 

Who are Polychaetes?

 

Before we start to discuss polychaete’s Hox-genes, there is necessary to make clear, who are the polychaetes. In early works of classic morphologists (Rouse and Fauchald, 1997) the Annelida phylum conclusively (with wide evidential base) was divided to two monophyletic branches – Polychaeta and Clitellata (Clitellata = Oligochaeta + Hirudinea). But annelid’s systematic was dramatically changed as soon as nucleotide and amino-acid sequences formed the basis for the analysis. At first time there were very few genes and gene families, then mitochondrial genomes, expressed sequence tags (ESTs) libraries and even whole transcriptomes, which was read with the aid of NGS platforms (Illumina) (Zrzavy et al. 2009; Shenet al., 2009; Struck et al., 2011; Weigert et al., 2014, Andrade et al., 2015). Such systematic approach revealed much more complex counter-intuitive links within Annelida (Fig.1). It appears that polychaetes are paraphyletic and the term “polychaete” itself fell out of taxonomic register, becoming the synonym of Annelida as a whole (Struck et al., 2011; Weigert and Bleidorn, 2016). Now, mostly due to traditions (but not according to cladistics) we use the name “Polychaetes” for sea annelids from two different monophyletic branches – Errantia and Sedentaria. This is important, that these two branches had common ancestor in remote period. Fossil record saved petrificated mandibles (scolecodonts), which looks like jaws of modern Eunicida (Errantia) and are dated by the border of Late Cambrian and Early Ordovician, i.e. at least 485 millions years old (Weigert and Bleidorn, 2016). This is possible, that last common ancestor of Errantia and Sedentaria lived even earlier. Following the phylogenetic reconstructions and fossils from Middle-Cambrian sediments of Burgess Shale (Middle Cambrian; ~508 mya) it was successful to restore some traits of this common ancestor (Struck et al., 2011; Parry et al., 2015). This was macroscopic epibenthic worm with paired palps, biramous parapodia, simple (capillary) chaetes and, possible, with aciculae – the chitinous support rods within parapodia (Struck et al., 2011). There are disagreements about aciculae, because annelids from Burgess Shale haven’t them (Parry et al., 2015), but they was found in worm fossil in earlier sediments (Liu et al., 2015).

Errantia and Sedentaria dominate by species quantity and form divergence. They consist the crown group of Annelids. Among them there are segmented and unsegmented worms, there are forms with simple life cycle and with complex one, which includes several larva stages. Besides them, Annelids include six basal branches - Oweniidae, Magelonidae, Chaetopteridae, Sipuncula, Amphinomida, and Lobatocerebrum (Weigert and Bleidorn, 2016). Among them also there are segmented and unsegmented worms. This makes difficult to reconstruct Annelid crown group ancestor.

 

 

The set and the structure of the polychaet’s Hox clusters

 

Annelid’s Hox-genes are studied most widely of all Lophotrochozoa. However if we juxtapose the data about Hox-cluster set and about Hox-genes transcription pattern with phylogenetic map, we’ll see that this data is very sketchy (Fig.1). Clitellata line species are studied better than other lines (Wysocka-Diller et al., 1989; Aisemberg et al., 1993; Snow and Buss, 1994; Cho et al., 2004, Cho et al., 2012; Simakov et al., 2013; Zwarycz et al., 2015; Endoetal., 2016). This is probably attributable to easier maintaining of terrestrial and freshwater annelid cultures. But according to molecular phylogeny and the fossil record the Clitellata appeared in late Paleozoic era and not earlier than in Permian period. They are very different from the ancestor of crown group annelids and basal annelids (Parry et al., 2015).

Almost full sets of Hox-genes were cloned from Urechis unicinetus (Sedentaria, Echiura), Ctenodrilus serratus (Sedentaria, Cirratuliformia) and Myzostoma cirriferum (Errantia; Mizostomida) (Barucca et al., 2016), but the data about their expression still absent yet. The first five Hox-cluster’s genes were cloned from only-explored representative of basal group – Chaetopterus variopedatus (Irvine and Martindale, 2000). Their transcription patterns are well-studied and we discuss about them later.

By now most complete data acquired for two polychaetes from Errantia branch (Alitta virens and Platynereis dumerilii; Family:Nereididae) and for the one polychaete from Sedentaria branch (Capitella capitata; Family: Capitellidae). Of course, two families from near hundred existed ones cannot represent all crown group annelids, but at least there is good that studied species weren’t from same branch.

The full set of Hox-genes that are specific for Lophotrochozoa was first described in Alitta virens (formerly called Nereis virens) (de Rosa et al., 1999). In Alitta genome there are eleven Hox-genes: Hox1 (PG1), Hox2 (PG2), Hox3 (PG3), Hox4 (PG4), Hox5 (PG5), Lox5 (PG6−8), Hox7 (PG6−8), Lox4 (PG6−8), Lox2 (PG6−8), Post2 (PG9+) и Post1 (PG9+). This set is considered as an ancestral complement for Mollusca, Brachiopoda and Annelida (Barucca et al., 2016). In second nereid polychaete (Platynereis dumerilii) at first time had been cloned only nine genes (Kulakova et al.2007; Pfeifer et al., 2012), but later missed orthologs was found (Hox7, Lox4) (Bakalenko N., unpublished data). The expression of these genes begins lately, and there is necessary to consider development stage (when the cDNA is received) to clone them. The size and the structure of nereid polychaete’s Hox-cluster isn’t described in details yet. According to preliminary data that was acquired by pulsed field electrophoresis (PFE) (Andreeva et al., 2001), Alitta ’s Hox-cluster is placed in single locus that isn’t exceeding 2,4 – 2,5 Mb. It suggests that Hox-genes aren’t scattered in genome, i.e. the cluster isn’t atomized like ones in oikopleura or in octopus (Seo et al., 2004; Albertin et al., 2015). But the resolution of PFE doesn’t allow to find out the exact size of cluster and the pattern of its organization (organized, disorganized or split, sensu Denis Duboule, 2007). In Platynereis Hox-genes are also localized within one chromosome, and moreover in one linkage group with some “extended” Hox-genes (eve and mox), and with genes from EHGbox (en, hb9/mnx1, gbx). These genes often flank Deuterostomia Hox-clusters and are considered as a part of ancient synteny between ancestral gene clusters from Mega-cluster of homeobox-containing genes of the Antennapedia class (ANTP-class) (Pollard and Holland, 2000; Hui et al., 2012).

The similar Hox-genes set was found in Capitella capitata (Frobius et al., 2008). The structure of the cluster is studied in details. It is localized in three different scaffolds. In first scaffold at the region of 243 kbp are placed the genes CapI-lab (PG1), CapI-pb (PG2), CapI-Hox3 (PG3), CapI-Dfd (PG4), CapI-Scr (PG5), CapI-lox5 (PG6−8), CapI-Antp (PG6−8), and CapI-lox4 (PG6−8). The remaining genes, CapI-lox2 (PG6−8) and CapI-Post2 (PG9+) occupy 21,6 kbp in second scaffold. The Post1 lies within third scaffold, rounded by non-Hox sequences. It suggests that the Post1-ortholog is not part of the Capitella Hox-cluster. All Capitella Hox-genes have same transcriptional orientation and haven’t ORFs between them. If we summarize Hox-coding areas, then the total cluster’s size will be 345 Kb. The Capitella genome is sequenced (Simakov et al., 2013), but unfortunately the gap between scaffolds is still present. It probably points at the region, that is difficult to read and to assemble, or may be at the great distance between Hox-cluster fragments.

This is interesting that studied Clitellata species (oligochaetes and leeches) often had Hox-cluster reconstructions which led to the loss or to the duplication of certain genes (Simakov et al., 2013; Barucca et al., 2016). For example, leeches haven’t Post1 orthologs, and Hox2 is present only in earthworm Perionyx excavatus. Other earthworm Eisenia fetida has multiply duplications that possibly are whole-genome ones. They increase Hox-genes quantity in genome up to 28 (Zwarycz et al., 2015).

Orthologs of the first five polychaete’s Hox-genes were found in Deuterostomia and Ecdysozoa. It seems, that these genes are inherited from common Bilateria ancestor – UrBilateria (de Rosa et al. 1999). These are most conservative Hox-genes. The annelid’s central Hox-genes (Lox5, Hox7, Lox4 and Lox2) haven’t definite orthologs among vertebrate’s central Hox-genes (Hox6, Hox7, Hox8) and among Arthropods (Antp, Ubx, abd-A) (de Rosa et al. 1999; Hueber et al., 2013; Frobius et al., 2008). To puzzle out in complicated phylogenetic relationships within this sequences group, there was necessary a detailed analysis of signature amino acid motifs (Hueber et al., 2013). It revealed that Protosomia last common ancestor had at least two central Hox-genes – Antp/Hox7-like and abd-A-like. The first sequence in most conservative and was inherited from the Protostomia/Deuterostomia ancestor (Hueber et al., 2013). The Antp/Hox7-like by the independent duplication gave rise to ftz and Antp in Arthropods, and Lox5 and Hox7 in Annelids. The second central gene – the abd-A-like appeared shortly before UrProtostomia division to Ecdysozoa and Spiralia. This ancestral gene gave rise to abd-A in Arthropods and to Lox4 and Lox2 in Annelids. It is interesting, that key gene for Insect development and evolution Ubx appeared relatively late in Arthropods line, probably by ancestral abd-A sequence duplication (Hueber et al., 2013). The affinity between central Hox proteins of Deuterostomia and Protostomia is traced at the level of single sequence type - Antp/Hox7. It suggests that UrBilateria had had one central Hox-gene - Antp/Hox7, which duplicated and diverged by the specific way for each evolution branch.

The Annelid posterior genes are Post1 and Post2. They are representatives of PG9+ and are characteristic for Spiralia. Until quite recently it was believed that Post1 appeared after the duplication of Post1/Post2 ancestral sequence at the level of last common ancestor of Mollusks, Brachiopods and Annelids. It seemed true because in nemertean’s and flatworm’s genomes was found only one posterior gene that is similar to Post2 (Barucca et al., 2016). However in 2016 an article was published, which described single Post-1 gene and 4 Post-2 paralogs, that had been found in Schmidtea mediterranea (Tricladida) (Currie et al., 2016). There is possible, that Post1 had appeared earlier than it was supposed, but it lost repeatedly.

Thus in each evolution branch of Bilateria (Deuterostomia, Ecdysozoa and Spiralia) the orthologs of central and posterior genes duplicated and diverged independently. It has made them useful phylogenetic markers for examination of animals with doubtful systematic place (Chaetognatha, Rotifera etc.). Besides, the structural Hox-cluster evolution is inseparable from the functional evolution. Let’s try to understand, how the Hox-genes expression is connected with polychaete’s morphogenesis features.