Molecular Phylogeny Inferred from 18S rRNA Gene Sequences of Nematodes Associated with Cernuella virgata, a Pest Snail in Australia

Pest snails are economically important pests of the grain industry. Nematode based bioagent appears to be a hope for controlling pest snails in an environment friendly way. Based on the dataset of 18S rRNA gene sequences, we propose a molecular phylogeny of nematodes baited with Cernuella virgata in soils collected from southern states of Australia. A total of 12 species (representing eight genera of nematodes) were identified and the inferred phylogenetic trees (Neighbor-Joining and Minium Evolution) placed them within three (I, IV and VII) out of the seven clades, indicating the possibility of multiple origins of snail parasitism. In Clade I and Clade VII, nematodes associated with Cernuella virgata formed sister group relationships with some slug – parasitic nematodes. We assume that snail – parasitic nematodes and slug parasitic nematodes might share common ancestors in their evolutionary histories.


Introduction
Nematode is one of the most abundant and diverse phylum in the animal kingdom [1]. Due to the lack of objective criteria for assessing homology of morphological characters regarding many nematodes, the systematics of this phylum has been contentious [2]. With the rapid development of molecular phylogeny, the evolutionary history of Nematoda was reassessed and new phylogenetic framework was pointed out [3][4][5]. Nevertheless, little is known about the phylogeny of nematodes associated with terrestrial gastropods, which are economically important invertebrates.
Ross et al. [2] reported the molecular phylogeny of slug-parasitic nematodes based on 18S rRNA gene sequences. A total of eight slugparasitic nematode species (Agfa flexilis, Alloionema appendiculatum, Angiostoma limacis, Angiostoma dentifera, Cosmocercoides dukae, Mermithid sp., Phasmarhabditis Hermaphrodita and Phasmarhabditis neopapillosa) from six families (Agfidae, Alloionematidae, Angiostomatidae, Cosmocercidae, Mermithidae and Rhabditidae) were included in their study. The resulting phylogenetic trees placed eight species within four (I, III, IV and V) out of the five clades of Nematoda, indicating multiple origins of slug parasitism. Five out of the eight nematode species were clustered within Clade V, forming a monophyletic group covering two families (Agfidae, Angiostomatidae) and one genus (Phasmarhabditis). By considering the morphological diversity among these families, they stated that rapid evolution had occurred during the evolutionary history of slug -parasitic nematodes.
While the phylogeny of slug -parasitic nematodes was studied, the phylogeny of snail -parasitic nematodes remains unclear. One of the reasons is that few scientific data are available regarding the snailparasitic nematodes around the world. Our understanding for the snail/nematode associations is mostly based on surveys conducted by Mengert [6] in Germany, Morand [7] in France, Gleich et al. [8] in USA and Charwat and Davies [9] in Australia. Currently the confirmed snail -parasitic nematodes is quite limited (e.g. Angiostoma aspersae (Angiostomatidae), Phasmarhabditis hermaphroditaI (Rhabditidae) and Nemhelix bakeri (Cosmocercidae) [10,11].
Terrestrial snails play a big role in agriculture and other industries. For examples, four introduced species of Mediterranean snails [Cernuella virgata (Da Costa), Theba pisana (Müller), Cochicella acuta (Müller) and Cochicella barbara (Linnaeus)], cause serious damage to the grain industry in Australia each year [12]. To control these pest snails efficiently and environment friendly, nematode -based biological control method was regarded as a priority among other options [9].
Effective use of nematodes requires knowledge of their relationships. Understanding the diversity of nematodes that are parasitic to terrestrial snails (especially for pest snails) and resolving the phylogeny of these nematodes will be useful to the development of nematode -based bioagent against pest snails.
In present study, we conducted a survey in southern Australia to screen nematode species with potentials as parasites of C. virgata. We also aim to solve the phylogenetic relationships of these nematode isolates using data from 18S rRNA gene sequences.

Soil sampling and nematode isolation
Samples were collected from 27 locations in South Australia, Victoria and New South Wales in Australia between August 2007 and September 2008 (Table 1). Sites were chosen based on accessibility and habitats. At each site, five to eight subsamples were taken with at least a two meters distance between them. Each subsample was obtained using a hand trowel from the top soil (10-15 cm deep). Approximately 0.5 kg soil was taken from each spot and was placed in separate polyethylene bag to minimize dehydration. Soil samples were stored in an ice box while being transported to the laboratory. Nematodes were isolated from each sample by baiting with nematode-free snails (C. virgata) as reported by Charwat and Davies [9], then placed in water for 24 hours to release nematodes.

DNA extraction
Nematode DNA was extracted from individual nematodes using a modification of the protocol described by Floyd et al. [13]. In brief, individual nematodes (adults or larvae) were transferred to a 0.2 ml Eppendorf tube containing 20 µl of 0.25 M NaOH, incubated at 25°C for 3-5 hours, then heated at 95°C in a Dri-Block heater (DB-2A: Techne Inc., Duxford UK) for 3 min. The resulting lysate was neutralized with 4µl (1 M) HCl and 10µl 0.5 M Tris-HCl (buffered at pH 8.0), then heated for 3 min at 95°C, followed by addition of 5 µl 2% Triton X-100. The final extract was stored at -20°C for later use.

Choice of DNA markers
Both nuclear and mitochondrial genes (18S rRNA, 28S rRNA, Cytochrome C oxidase I and 16S rRNA) were considered for study. 18S rRNA gene was chosen for three reasons. First, in pilot trials, PCR amplifications were obtained more reliably from 18S rRNA gene than from other candidate genes. Second, a large dataset of sequences was available on GENBANK or NemATOL for many species of nematodes across a range of taxonomic groups [3,[13][14][15]. Third, this gene contains both conserved stem and highly divergent loop regions, making it suitable for taxonomic differentiation [13].
Sequences of purified PCR products were obtained from both directions using the same primer pairs for PCR. Sequencing reactions were performed with the Applied Biosystems BigDyeTM Terminator Ready Reaction Kit (Version 3.1) (Applied Biosystems Ltd). Final capillary separation was carried out at Australian Genome Research Facility Ltd (AGRF), where the samples were analysed using an AB3730xl (Applied Biosystems).

Phylogenetic analysis
Sequence traces were checked for their quality using the Trace Editor of MEGA v 4.0. [16]. A total of 47 DNA sequences were screened for their statistical similarities (positive matrix scores) with 18S rRNA gene sequences of identified nematodes in GENBANK by performing blast search [17]. Among the 12 identified groups, a single DNA sample was selected from each group to align with other 51 nematode 18S rRNA gene sequences that were downloaded from GenBank (Table 2). These additional nematode taxa were chosen based on their taxonomy positions and their relationships with terrestrial molluscs and other invertebrates. The alignments of these DNA sequences were conducted with Clustal X using the default parameters for gap opening and gap extension penalties [18]. A final 543 aligned characters were applied in the phylogenetic analysis. Neighbour-Joining (NJ) and Minimum Evolution (ME) trees were constructed with MEGA v 4.0 [16] using Kimura 2-parameter model. Gaps were treated as missing data in the analysis. The outgroup of Tylenchus arcuatus (Chromadorea, Nematoda) (Accession number: EU306349) was used to root the trees and for character polarization. Bootstrap support was calculated for all analyses using 1000 replicates.

Phylogenetic analysis
Phylogenetic trees (Neighbour Joining and Minimum Evolution) were constructed via phylogenetic analyses of 18S rRNA gene dataset arising from 64 taxa described above. In these phylogenetic trees, seven Clades were revealed and three out of them (clade I, IV and VII) contained nematodes associated with C. virgata (Figure 1, Figure 2). Clade I: Three nematode isolations from the present study, Acrobeloides butschlii (0823), Acrobeloides bodenheimeri (5512) and Cephalobus persegnis (5211) were placed in Clade I (Panagrolaimida) in all topologies (Figure 1, Figure 2). Both NJ and ME trees depicted a sister-group relationship between these taxa and the other two members of Panagrolaimida (Acrobeles complexus and Zeldia punctata). This placement received very strong bootstrap support in both phylogenetic trees (100%).
Clade IV: In both NJ and ME trees, four nematode species from current study, Mononchoides striatus (3912), Pristionchus americanus (4611), Pristionchus lheritieri (3923) and Pristionchus pacificus (3812), were included in this clade IV (Diplogasterida). Among of them, P. americanus (4611), P. lheritieri (3923), and P. pacificus (3812) formed a monophyletic clade with strong bootstrap support (96% in NJ tree and 92% in ME tree). This monophyletic clade is nested within the Clade IV and sister to Mononchoides striatus (3912) and Mononchoides striatus. This result received very strong bootstrap , Oscheius tipulae and Oscheius dolichura) (100% in both NJ and ME trees). Heterorhabditis bacteriophora (0512) was found to cluster with Heterorhabditis bacteriophora under weak bootstrap support. They were sister to Heterorhabditis hepialus and formed a clade with 100% bootstrap support across both phylogenetic trees. Instead of being clustered with other members of Rhabditis, Rhabditis sp. (4411) was found to cluster with Pellioditis typica in all phylogenetic trees with strong bootstrap support (100%). Both NJ and ME trees also depicted a sister -group relationship between these two species and other five slug -parasites (Agfa flexilis, Angiostoma limacis, Angiostoma dentifera, Phasmarhabditis Hermaphrodita and Phasmarhabditis neopapillosa).

Discussion
The present study revealed a total of 12 nematode species that are potentially associated with C. virgata, a pest snail in Australia.
Phylogenetic analyses of 18S rRNA gene sequences placed these nematode species into three large groups: Panagrolaimida, Diplogasterida and Rhabditida (Figure 1, Figure 2), indicating the possible multiple origins of snail parasitism, which is similar to the findings of slug -parasitic nematodes [2].

The relationship of potential snail parasites in relation to other nematodes in clade I, IV and VII
Clade I: In present study, phylogenetic analyses recovered a monophyletic clade I (Panagrolaimida), which includes Acrobeloides bodenheimeri (5512), Acrobeloides butschlii (0823), Cephalobus persegnis (5211), and two other members of Cephalobidae: Zeldia punclata and Acrobeles complexus. This finding is in consistent with the results of Nadler et al. [20], who confirmed the monophyly of cephlobids at superfamily level based on phylogenetic analyses of ribosomal (LSU) sequences data.
Cephalobidae include a diverse array of species ranging from soil dwelling microbivores to parasites of vertebrates and invertebrates [20]. The phylogeny of genera within Cephalobidae (such as Acrobeloides, Cephalobus, Chiloplacus, Eucephalobus and Pseudacrobeles) has been in controversy [20]. Molecular trees did not support traditional genera as natural group [20]. Similarly, morphological characters traditionally applied for distinguishing most genera (e.g. labial variations) were not regarded as diagnostic with the discovery of increasing new species [21]. Such a controversy was reflected in present study. Both NJ and ME trees did not support Acrobeloides and Cephalobus as monophyletic groups.
The NJ and ME trees also depicted a closely related relationship between Clade I and a group including the slug -parasite (Cosmocercoides dukae) and the snail -parasite (Nemhelix bakeri) ( Figure 1, Figure 2), indicating the possibility of a common ancestor between these nematodes.
Clade IV: The monophyly of clade IV (Diplogastropoda), which includes Pristionchus lheritieri (3923), Pristionchus americanus (4611), Pristionchus pacificus (3812), Mononchoides striatus (3912) and Mononchoides striatus, was resolved through both NJ and ME analyses in present study. Strong bootstrap support was observed for this clade in NJ (98%) and ME (94%) trees (Figure 1, Figure 2). This finding is in consistent with the results reported by Fürst von Lieven [22], who constructed a robust cladogram for Diplogastropoda based on morphological data (e.g. the variable structures of the buccal cavity and the function of the stomatal structures).
Traditionally Diplogastropoda was regarded as a sister taxa of Tylenchina [23] because the morphology of pharynx between these two groups is very similar. Data from molecular and ultrastructure, however, strongly object the Diplogasterida/Tylenchida clade [3,24]. The close phylogenetic relationship between Diplogastropoda and Tylenchina is not supported by our results. Neither the NJ analysis nor the ME analysis indicated that Diplogastropoda is a sister taxa of Tylenchina ( Figure 1, Figure 2).
Surveys conducted by Mengert [6], Morand [7], Gleich et al. [8], Charwat and Davies [9] indicated that some species within the Diplogastropoda might associate with terrestrial molluscs parasitically, phoretically or necromenically. The present study supported their findings but was inconsistent with Ross et al. [2], who found no members of the Diplogasteridae were parasitic to slugs.
Such a connection strongly suggests the possibility that snail-parasitic nematodes might share a common ancestor with slug -parasitic nematodes.
The remaining nematode isolate, Mesorhabditis sp. (5112), was separated from other nematode isolates in both NJ and ME analyses ( Figure 1, Figure 2). As a member of Mesorhabditidae, it was expected to cluster with other members of Rhabditida. However, it was actually sister to Clade IV (Diplogasterida) in all phylogenetic trees (84% in NJ and 90% in ME). Additional research is thus required to resolve the phylogenetic position of this taxon.

Other phylogenetic finding in term of nematode phylogeny incurred from this study
While recovering the phylogenetic positions of our 12 nematode isolates, the resulting NJ and ME trees also presented enlightenments on the phylogeny of other nematode taxa.
Mermithida are a group of insect -parasitic nematodes. They are usually associated with arthropods but were also found to be parasites of Molluscs [25]. Our analyses resolved the monophyly of Mermithda (89% in ME tree and 97% in NJ tree). The phylogenetic trees also had moderate to strong support to the sister group relationship between Mermithida and Monochida (70% in ME tree and 96% in NJ tree).
These findings are in consistent with other author's results Megan et al. [1] and Ross et al. [2] but disagree with Stock and Hunt [26], who placed the Mermithidae as a sister group to the plant -parasitic Dorylaimids.
Another clade that was proved to be monophyletic is Sterinernematidae (100% for both NJ and ME). Sterinernematidae is a family of entomopathogenic nematodes (EPN) [27]. It shares similar life history with the other family of entomopathogenic nematodes (Heterorhabditidae) (such as killing insects by realising symbiotically associated bacteria into the hemocoel of insects), but has distantly related phylogenetic relationship with Heterorhabditidae [28]. This situation was reflected in our phylogenetic analyses: the members of Heterorhabditidae (Heterorhabditis bacteriophora and Heterorhabditis hepialus) were placed in clade VII while the member of Sterinernematidae formed a separate clade (clade V) across both NJ and ME trees.
The inferred phylogenetic trees also showed that Sterinernematidae was more closely related to a clade including most Panagrolaimidae (free-living and insect associates). Both NJ and ME trees strongly supported the monophyly of Panagrolaimidae (100% in ME and 99% Are these nematodes really snail parasites? By using C. virgata as baiting material, we found that 12 nematodes species Acrobeloides butschlii (0823), Acrobeloides bodenheimeri (5512), Cephalobus persegnis (5211) pest snail in Australia. Although it is hard to seek testable evidence to confirm this finding, the hypothesis of these nematodes (or some of them) as potential parasites of C. virgata is justified as below.
Reports about bacterivorous nematodes being developed as bioagent against pest slugs (e.g. P. hermaphrodita) have been published [30,31]. From the point of ecological view, all our nematodes isolates fall into the category of free-living bacterivorous nematodes (FLBN). The close relationship between some of our nematodes isolates with some slug parasites were also revealed by the phylogenetic analyses conducted in present study. In this respect, we could not deny the potentiality that bioagents against pest snails such as C. virgata can be developed from these nematode isolates.
All parasitic nematodes were originally evolved from free living nematodes [3]. Parasitism of plants and animals has evolved independently at least nine times in the history of the nematodes [14]. The adoption of parasitism in nematodes probably required either the adaptation of genes present in their free-living ancestors or horizontal gene transfer from bacteria and/or fungus in their environment [32][33][34][35]. Given the fact that our nematode isolates are bacterivirous, and have been isolated from the cadavers of pest snails (C. virgata), it is likely that they could acquire "parasitism genes" from bacteria in their environment, and become parasites of pest snails at some stages of their life cycle. Identification of some "parasitism genes" by examining the expression pattern of their C. elegans orthologs at certain stage of development (e.g. the third larval stage) would be useful in assessing the parasitism of nematodes [32,36]. Further pathogenicity tests are now underway to assess the biocontrol potential of these nematode isolates.

Conclusion
This study presents the molecular phylogeny of nematodes baited from the pest snail of C. virgata in Australia. Both NJ and ME trees constructed based on the dataset of 18S rRNA gene sequences placed 12 nematode isolates into three out of seven Clades (I, IV and VII), suggesting the possibility of multiple origins of snail parasitism. In Clade I and Clade VII, nematodes associated with C. virgata formed sister group relationships with some slug -parasitic nematodes. We assume that snail -parasitic nematodes and slug -parasitic nematodes might share common ancestors in their evolutionary histories.