Phylogenetic studies on three Helicotylenchus species based on 28S rDNA and mtCOI sequence data

Phylogenetic analyses

Multiple sequence alignments and subsequent comparison of the newly acquired (Tables 1 and 2) and GenBank-derived Helicotylenchus spp. sequences were performed using ClustalW algorithm as implemented in the BioEdit program v. 7.2.5 (Hall, 1999). The final multiple-sequence alignments were 595 and 588 nucleotide sites long in the case of 28S rDNA and mtCOI data, respectively. For the 28S rDNA data set, two Rotylenchus and two Hoplolaimus species were selected as outgroup: R. uniformis (MG653537, acquired in this study), R. magnus (EU280789), H. seinhorsti (DQ328752) and H. galeatus (EU626787). Heterodera glycines (HM462017) and R. uniformis (MG663121-2; this study) sequences were used as outgroups in mtCOI data set. Substitution models were tested using the “FindModel,” an online implementation of MODELTEST (Posada and Crandall, 1998). A GTR + I + G substitution model was chosen for both 28S rDNA and mtCOI DNA data sets. The Bayesian phylogenetic trees were constructed with the program MrBayes v. 3.1 (Ronquist and Huelsenbeck, 2003). In both cases, two independent runs were performed with four Markov chains per run. In case of the 28S rDNA, the program was run for 3,000,000 generations and in case of mtCOI for 1,000,000 generations, with a sampling frequency of 100 generations. Burn-in trees of 50,000 and 25,000 generations, respectively, were discarded. For the mtCOI data, the partition by codon position was set. The sampled trees from each run were combined in a single 50% majority-rule tree. The stabilization of the likelihood and parameters was checked with the program Tracer v. 1.6 (Rambaut et al., 2014).

Results

Helicotylenchus canadensis (Waseem, 1961).

(Fig. 1A and D; Table 3).

Sequence analysis

All sequences reported in this study have been deposited in GenBank and their accession numbers are listed in Table 2. By the use of the newly designed primers, we obtained mitochondrial sequences from 4 representatives of H. canadensis, 11 of H. pseudorobustus and 8 of H. varicaudatus (Table 2). The particular amplicon and a derived size of the newly obtained sequences varied depending on the primer combination used (Table 1). The designed primers, in most combinations, targeted the region relatively closer to the 5′ end of the mtCOI gene sequence. The longest PCR amplicons, which were approximately 700 bp, were obtained using the M3.5 F and M8aR primer combination, which also amplified a part of the mitochondrial sequence closer to the 3′ end of the mtCOI gene. However, the most successful primer combination was the M2F and RH1R one, which allowed for the amplification of PCR product of approximately 600 bp from all analyzed Helicotylenchus species.

There was no intra-specific nucleotide variation concerning the four mtCOI sequences derived from H. canadensis. The potential minor nucleotide differences concerned few degenerate nucleotides, namely, K (G or T), Y (C or T), and R (A or G). The highest amplification success was obtained for the H. pseudorobustus and H. varicaudatus. By the use of three of the four presented primer combinations, we were able to obtain amplicons and subsequent mitochondrial sequences from 11 and 8 nematode individuals, respectively. The variability of the sequence amplified from H. pseudorobustus involved only some minor ambiguous nucleotides, possibly due to sequencing error. The only significant nucleotide substitution (A/T) was localized in the middle of the sequence at site 418 (MG663108), which had no influence on the potential amino acid sequence translation. Importantly, the more insightful analysis of the H. pseudorobustus mitochondrial nucleotide sequences revealed that all of them contain multiple TAA and TAG codons, which in the regular invertebrate mitochondrial genetic code would be translated as termination codons. A comparison of these sequences with the publicly available sequences derived from other Hoplolaimidae species (Fig. 3) suggests that Tyrosine (Y) and Phenylalanine (F) codons are the most vulnerable to such substitutions in this family, and that H. oleae-derived sequences also contain TAA codons (Cantalapiedra-Navarrete et al., 2013; van den Berg et al., 2016; Palomares-Rius et al., 2018). Both TAA and TAG triplets were also detected in sequences derived from two Scutellonema: S. brachyurus and S. truncatum, reported from study of van den Berg et al. (2013, 2017). No amino acid sequence ambiguities were predicted for the two other Helicotylenchus species analyzed here. Furthermore, no intra-specific variation on the nucleotide level was observed in the case of mtCOI sequences obtained from H. varicaudatus individuals.

Figure 3:

28S rDNA fragments were successfully acquired from four H. canadensis, nine H. pseudorobustus, and four H. varicaudatus individuals (Table 2). All fragments were approximately 550 nucleotides long and they were characterized by a lack of intra-specific sequence variation.

The highest values of inter-specific variations were observed between H. pseudorobustus and two other species: 32 to 39% in case of mitochondrial sequences and 12 to 13% in case of the 28S rDNA. The lowest inter-specific variation was between H. canadensis and H. varicaudatus, which was on the level of 16 to 20% in the case of mitochondrial sequences and 5% in the case of 28S rDNA.

28S rDNA and mtCOI-based phylogenetic relationships of the investigated Hoplolaimidae species

Bayesian phylogenetic analysis of the 28S rDNA sequences derived from various Helicotylenchus species resulted in a tree topology (Fig. 4) that agrees with the phylogenetic data presented by Subbotin et al. (2015), Palomares-Rius et al. (2018), and Shokoohi et al. (2018). In total, 12 Helicotylenchus clades (I to XI, as defined in Subbotin et al. 2015; number XII was given in this study to the H. oleae lineage) were recovered, and the Helicotylenchus sequences acquired here were inferred to belong to two of them. According to the expectations, H. pseudorobustus sequences were placed in Clade I and clustered with the other publicly available H. pseudorobustus sequences. The rest of the new sequences derived from the two other analyzed Helicotylenchus species were inferred to be in Clade IX. However, the representatives of each of the two species were redistributed as lineages separated by closer relationships to other species. Sequences of H. canadensis clustered within Clade IX, together with representatives of H. vulgaris Type A, whereas the H. varicaudatus-derived sequence was inferred to be a sister lineage of all other Clade IX taxa. Designating Hoplolaimus species as outgroup, Rotylenchus uniformis (MG653537, obtained in this study), together with its congener R. magnus (EU280789), formed the sister group to Helicotylenchus.

Figure 4:

Congruent with results of 28S rRNA analyses above, the topology of the phylogenetic tree inferred from mtCOI showed that H. canadensis and H. varicaudatus share the closest pairwise relationship of the three Helicotylenchus species analyzed (Fig. 5) and they were inferred to be in Clade IX of the 28S rDNA-based tree. The species H. pseudorobustus (representative of Clade I) represented a separate lineage, whose most recent common ancestor with the former two species was that of all Helicotylenchus species. However, the relationships between the three distant lineages (Clades I, IX and XII) have not been resolved on the basis of the available mitochondrial data, most probably due to insufficient number of analyzed taxa.

Figure 5:

Discussion

Morphological and molecular data in this study included three Helicotylenchus species: H. canadensis, H. pseudorobustus and H. varicaudatus, extracted from various locations and vegetation types in Poland (Table 2). The morphological evaluation of the H. pseudorobustus nematode individuals and their positioning in the Clade I of the Helicotylenchus 28S rDNA tree agreed with the type characterization of this species, as described in Subbotin et al. (2015). Newly generated 28S rDNA sequences from H. canadensis and H. varicaudatus that are positioned in Clade IX are reported for the first time for those species. Morphological and morphometrical studies on these species were congruent with previous descriptions. The inferred, close phylogenetic relationship of H. canadensis to Helicotylenchus vulgaris (Yuen, 1964) supports morphological comparisons of these two species as being very similar. The most clearly observable differences between these two species were the length of the tail and the width of the annuli at the tail end: the tail of H. vulgaris is shorter than the tail of H. canadensis, and the middle annuli at the end of the H. vulgaris tail are narrower than other annules, whereas the annuli of the H. canadensis tail are of equal width. The mtCOI-based phylogenetic analysis included four newly sequenced Helicotylenchus spp. (Fig. 5) and it also revealed that H. canadensis and H. varicaudatus are more closely related to each other than either is to H. pseudorobustus.

By providing the 28S rDNA and mtCOI data acquired in this study, the molecular characteristics of H. canadensis, H. pseudorobustus and H. varicaudatus were substantially enriched. So far, H. canadensis and H. varicaudatus had been characterized only by 18S rDNA fragments (van Megen et al., 2009; Schreck Reis et al., 2010). The 28S rDNA region is often more variable than the small subunit ribosomal DNA (Hillis and Dixon, 1991; Holterman, 2008) and it has been useful for resolving phylogenetic relationships within Hoplolaiminae (Bae et al., 2009; Subbotin et al., 2011, 2015; Palomares-Rius et al., 2018). Accordingly, the majority of the Helicotylenchus sequences deposited in GenBank comprise 28S rDNA data. Although GenBank already included 18S rDNA and 28S rDNA sequences for H. pseudorobustus, this study provides the first mtCOI fragments for this species. Characterized by a very high sequence variability, the mitochondrial data of Helicotylenchus species as well as other nematode genera are slowly increasing, being particularly useful in supporting the process of delineation of the most dubiously designated species.

The mitochondrial DNA region encoding for mtCOI gene has been widely explored for barcoding application purposes in many animal organisms (Rach et al., 2017). It was also used in Hoplolaimidae and other nematode groups (Kanzaki and Futai, 2002; Derycke et al., 2010; Riepsamen et al., 2011; Gutiérrez-Gutiérrez et al., 2012; Kumari and Subbotin, 2012; van den Berg et al., 2013; Cantalapiedra-Navarrete et al., 2013; Holguin et al., 2015; Vovlas et al., 2015; van den Berg et al., 2017; Sanchez-Monge et al., 2017; Skwiercz et al., 2017). In this study, to enhance the process of delineation and species identification within genus Helicotylenchus, which was underrepresented in terms of available mtCOI sequences, a new set of mtCOI primers was developed and tested. These newly designed primers offer an alternative to the most frequently used mtCOI primers combinations, especially in a situation when the original ones fail in the amplification of a desired PCR product. Most of the primer combinations presented in this publication allowed for the amplification of the fragments closer to the 5′ end of the assumed cytochrome oxidase c nucleotide sequence. The primer M2F in combination with the RH1R primer was the most successful in our study and allowed for amplification of a mtCOI sequence (600 bp) from all Helicotylenchus species analyzed here. The longest mtCOI sequence, including more nucleotides at its 3′ end, was acquired using the M3.5F and M8aR primer combination on H. pseudorbustus (and R. uniformis that was used as an outgroup species).

The inspection of mitochondrial sequences obtained here shows that all 11 mtCOI sequences from the nematode representatives of H. pseudorobustus were almost identical and included TAA and TAG codons at the same implied amino acid positions. Strikingly, such sequences were obtained each time for the H. pseudorobustus representatives, using three different mtCOI-targeted primer combinations (Table 1), whereas none of the sequences acquired from the other two Helicotylenchus species analyzed here revealed such nucleotide substitutions. In contrast, the homologous triplets in the mtCOI sequences of H. candadensis and H. varicaudatus instead contain triplets translated into tyrosine (TAT) and the phenylalanine (TTT) according to the standard invertebrate mitochondrial genetic code. This finding suggests alterations to the mitochondrial genetic code in H. pseudorobustus, if sequences are indeed translated through putative stop codons in this species.

Genetic code alterations in both nuclear and mitochondrial genomes have been reported in various organisms (Bessho et al., 1992; Telford et al. 2000; Lavrov et al., 2012; Lavrov, 2014; Bezerra et al., 2015; Pánek et al., 2017). In their studies, Jacob et al. (2009) showed that such alterations can also be found in nematodes. In the mitochondrial genome of nematode Radopholus similis (Pratylenchidae), the UAA (TAA) codon is reassigned from translation termination to tyrosine. Consequently, the usage of the TAA codon can be observed in mtCOI sequences acquired from various Hoplolaimidae species, including Rotylenchulus spp. (van den Berg et al., 2016) and the recently investigated H. oleae (Palomares-Rius et al., 2018). Although the simultaneous presence of TAA and TAG codons can be found in the sequences of Scutellonema brachyurus and S. truncatum, derived from the study of van den Berg et al. (2013, 2017), the authors did not elaborate on this matter.

On the contrary, the occurrence of stop codons in the mitochondrial sequences acquired from nematodes could also be an indication of nuclear copies of mitochondrial-derived genes (numts) or COI pseudogenes, as were highlighted by Holguin et al. (2015) in their studies on genetic diversity of Hoplolaimus spp. Although the researchers did not detect such changes in the mtCOI sequences in the species of their interest, they raised an important issue concerning the use of mtCOI sequences in the phylogenetic studies of nematodes. As long as the occurrence of TAG codons in the reported sequences is properly investigated, phylogenetic analyses based on mtCOI sequences should be treated with suitable consideration.