Plant parasitic nematodes of the family Trichodoridae cause substantial yield losses in many agricultural crops. Rapid and accurate identification of trichodorids to the species level is critical for selection of appropriate measures for control. This study analysed 99 sequences of the D2–D3 expansion segments of the 28S rRNA gene and 131 sequences of the 18S rRNA gene from the stubby nematodes belonging to the genera Nanidorus, Paratrichodorus and Trichodorus. Species delimiting was based on the integration of morphological identification, which is not provided in the present article, and molecular-based phylogenetic inference and sequence analysis. Twenty-two valid species and several species complexes were identified among nematodes included in the analysis. PCR-RFLPs of the partial 18S rDNA and the D2–D3 expansion segments of the 28S rDNA were tested and proposed for identification of these nematodes. Gel PCR-RFLP profiles and tables with restriction fragment lengths for several diagnostic enzymes are provided for identification. Some problems of taxonomy and phylogeny of nematodes of the family Trichodoridae are also discussed.
Nematodes of the family Trichodoridae are widely distributed in Europe and North America and are also reported from other parts of the world. These nematodes can cause substantial crop losses by acting as plant pathogens and as vectors for plant viruses. Because trichodorid nematode feeding can cause stunting of the roots, they are referred to as stubby root nematodes. One hundred and two species belonging to five genera, Allotrichodorus, Monotrichodorus, Nanidorus, Paratrichodorus and Trichodorus are currently recognized in the Trichodoridae family. Several stubby nematode species are known to vector tobraviruses, which cause economically important diseases in several crops (Decraemer, 1995; Decraemer & Robbins, 2007). Viruses belonging to the genus Tobravirus include Tobacco rattle virus, Pea early-browning virus and Pepper ringspot virus. There is a highly specific relationship between virus and nematode vector, so that particular virus isolates are transmitted only by certain nematode species.
Rapid and accurate identification of trichodorids to the species level is the first critical step for selection of appropriate measures for control of these nematodes. Traditional identification of trichodorids is based on analysis of morphological and morphometrical characters, which often have high intraspecific variability, are complex, difficult and time-consuming. Recently, DNA-based approaches have been successfully adapted for the molecular diagnostics of trichodorids. Blaxter et al. (1998), Boutsika et al. (2004b), Riga et al. (2007), Van Megen et al. (2009) and Duarte et al. (2010) have published sequences of 18S rRNA and ITS rRNA genes for several agriculturally important species of trichodorids. Using sequence data, Boutsika et al. (2004a) and Riga et al. (2007) developed a PCR with specific primers for diagnostics of P.allius, P. macrostylus, P. pachydermus, P. teres, T. primitivus and T. similis. Holeva et al. (2006) designed a real-time PCR assay for detection and quantification of P. pachydermus and T. similis in field samples. Recently, Duarte et al. (2011) developed a PCR-RFLP assay based on the 18S rRNA gene for rapid identification of 12 trichodorid species belonging to the genera Nanidorus, Paratrichodorus and Trichodorus. DNA techniques have been successfully applied to diagnostics of several stubby nematode species. However, many species remain uncharacterized at the molecular level.
The main objectives of this study were to: (i) verify species identification of trichodorid nematodes collected in the Czech Republic, USA and India by using phylogenetic analysis of rRNA gene sequences; (ii) develop a PCR-RFLP assay using D2–D3 expansion fragments of the 28S rRNA gene for diagnostics of stubby nematodes; and (iii) test the PCR-RFLP assay developed by Duarte et al. (2011) with a wider range of trichodorid samples.
Materials and methods
Nematode populations used in this study were obtained from soil samples collected from the Czech Republic, India and USA (California) (Table 1). The nematodes from the USA were extracted from samples using a centrifugal flotation technique (Coolen, 1979; Hooper, 1986a) and the nematodes from the Czech Republic and India were extracted by a sieving and decanting method (Brown & Boag, 1988). Specimens were killed by gentle heat, fixed in 4% formalin or triethanolamine–formalin (TAF, 2% triethanolamine, 7% formaldehyde solution, 91% water) and mounted in anhydrous glycerin for examination (Hooper, 1986b). Morphological identification of specimens was done using keys provided by Decraemer (1995) and Decraemer & Baujard (1998), with corresponding species descriptions. Some of the species from the Czech Republic used here were described morphologically by Kumari (2010) and Kumari & Decraemer (2011). In this study, the species were defined and delimited based on an integrated approach that considered morphological evaluation, molecular based phylogenetic inference (tree based methods) and sequence analysis (genetic distance methods) (Sites & Marshall, 2004).
Table 1. Species and populations of trichodorids and outgroup taxon sequenced in the present study
Identification based on morphology and rRNA sequences
DNA isolation, amplification, cloning and sequencing
Molecular studies of trichodorid samples from the Czech Republic, India and California, USA were conducted using slightly different protocols at the Crop Research Institute (Czech Republic) and at the California Department of Food and Agriculture (USA). Trichodorid nematodes collected from the Czech Republic and India were stored in 1 m NaCl before analysis. Total genomic DNA from individual nematodes was extracted according to the rapid method of Stanton et al. (1998). Four regions, 18S, ITS1, ITS2 and partial 28S of rRNA genes, were amplified using nematode universal primers (Table 2). The 18S rRNA gene was amplified into three fragments. Primer combinations were as follows: first fragment SSU_F_04 + SSU_R_09, second fragment SSU_F_22 + SSU_R_13, and third fragment SSU_F_23 + SSU_R_81. PCR was performed in a 25 μL total volume containing one PCR bead (GE Healthcare), 20·5 μL double distilled sterile water and 2·0 μL each primer (10 pmol μL−1) (synthesized by Generi Biotech). To this, 0·5 μL of DNA was added as a template for PCR. A negative control (sterilized water) was included in all PCR experiments. All PCR reactions were performed on a DNA Engine PTC–1148 thermal cycler (Bio-Rad). The cycling profile for all four markers was as follows: initial denaturation for 3 min at 94°C, followed by 40 cycles of 30 s at 94°C, 30 s at 55°C and 30 s at 72°C, with a final extension at 72°C for 10 min. Amplicons were analysed by electrophoresis and the remaining products were purified using the High Pure Product Purification kit (Roche Diagnostics GmbH) and sequenced in both directions using each primer pair (Macrogen). sequencher 4.8 (Gene codes Corp.) was used to assemble and view each sequence and check for base-calling errors.
Table 2. Primer combinations used in the present study
For stubby nematode species collected in the USA, DNA was extracted using the proteinase K protocol. Several specimens from each sample were put into a drop of water on a glass slide and cut under a binocular microscope. Each nematode specimen was transferred to an Eppendorf tube containing 25 μL double distilled water, 2 μL 10 × PCR buffer and 3 μL proteinase K (600 μg mL−1) (Promega). Tubes were incubated at 65°C for 1 h and then at 95°C for 15 min. Detailed protocols for PCR, cloning and sequencing were as described by Tanha Maafi et al. (2003). The PCR product was purified using the QIAquick Gel Extraction Kit (QIAGEN). The primer sets used for amplification and sequencing of ribosomal RNA gene fragments are given in Table 2. PCR products were purified and run on a DNA multicapillary sequencer at the University of California, Riverside.
Sequences were submitted to GenBank under accession numbers as indicated in Table 1.
The PCR product of the 18S rRNA gene was digested with TaqI, SatI, BseNI or TscAI. The PCR product of D2–D3 expansion fragments of the 28S rRNA gene was digested with BseNI, PstI, PvuII or RsaI. Three to five microlitres of purified PCR products were digested with each of the restriction enzymes. RFLPs were separated by electrophoresis using TAE-buffered gels, stained with ethidium bromide, visualized using a UV transilluminator and photographed. The length of each restriction fragment was obtained by virtual digestion of each sequence using WebCutter 2.0 (http://www.firstmarket.com/cutter/cut2.html).
Sequence and phylogenetic analysis
The new sequences of the 18S rRNA gene and D2–D3 expansion fragments of 28S rRNA gene were aligned using clustalX 1.83 (Thompson et al., 1997) using default parameters of corresponding gene sequences (Blaxter et al., 1998; Boutsika et al., 2004a; Van Megen et al., 2009; Duarte et al., 2010; RC Holeva, MS Phillips, FG Wright, DJ Brown and VC Blok, the James Hutton Institute, Dundee, UK, unpublished data, XQ Li and JW Zheng, Institute of Biotechnology, Zhejiang University, China, unpublished data). Outgroup representatives of the genera Tripyla and Alaimus for the D2–D3 and Tobrilus, Prismatolaimus, Tripyla, Tylolaimophorus and Diphtherophora for 18S data sets were chosen using previous published data (Van Megen et al., 2009). Sequence data sets were analysed with Bayesian inference (BI) using MrBayes 3.1.2 (Huelsenbeck & Ronquist, 2001). BI analysis under the GTR + I + G model for each gene was initiated with a random starting tree and was run with four chains for 1 × 106 generations. The Markov chains were sampled at intervals of 100 generations. Two runs were performed for each analysis. The log-likelihood values of the sample points stabilized after approximately 1000 generations. After discarding burn-in samples and evaluating convergence, the remaining samples were retained for further analysis. The topologies were used to generate a 50% majority rule consensus tree. Posterior probabilities (PP) are given for appropriate clades. Sequence differences between samples were calculated with paup* 4b10 (Swofford, 2003) as an absolute distance matrix and the percentage was adjusted for missing data.
Species identification and delimiting
Ninety-nine and 131 sequences from trichodorids were included in the analyses of the D2–D3 of 28S and 18S rRNA genes, respectively. Forty sequences of the 28S and 12 sequences of the 18S rRNA gene were obtained in the present study. Using traditional morphological characters and molecular criteria (apomorphies and DNA distances), the following 10 species were distinguished within the samples: Nanidorus minor, N. renifer, Paratrichodorus pachydermus, P. porosus, Trichodorus pakistanensis, T. primitivus, T. similis, T. sparsus, T. variopapillatus and T. viruliferus. Two representatives of Trichodorus (Trichodorus sp. C and Trichodorus sp. D) from California were not identified to a species level. Several samples identified as representative of the same morpho-species showed differences in molecular characteristics, and were thus classified here as different species types: Trichodorus sparsus type ‘A’, ‘B’, ‘C’, ‘D’; T. pakistanensis‘A’, ‘B’; P. porosus‘A’, ‘B’, ‘C’; P. pachydermus‘A’, ‘B’; P. teres‘A’, ‘B’, ‘C’; P. hispanus‘A’, ‘B’. The analysis of the D2–D3 of the 28S rDNA sequence data set revealed four unidentified species of Trichodorus, and the partial 18S rDNA sequence data set distinguished four unidentified species of Trichodorus and two unidentified species of Paratrichodorus. Morphological descriptions and identifications of these nematodes are not available. More detailed morphological and molecular analysis is required to further evaluate and identify these samples. A total of 22 valid known species (Trichodorus– 11 species; Nanidorus– three species; Paratrichodorus– eight species) were identified and included in the analyses.
Sequence and phylogenetic analysis
The alignment for the partial 18S rDNA included 136 sequences and was 1137 bases long. Fourteen Trichodorus, three Nanidorus and 10 Paratrichodorus nominal and putative species were included in the analysis.
Intraspecific sequence variations for some species were: P. porosus‘C’, 0–0·5% (0–6 nt); N. minor, 0–0·6% (0–7 nt); N. renifer, 0–0·4% (0–4 nt); T. primitivus, 0–0·3% (0–3 nt); T. pakistanensis, 0–0·7% (0–8 nt); T. sparsus‘A’, 0·2 (2 nt); T. sparsus‘B’, 0–0·3% (0–4 nt); T. similis, 0–0·5% (0–5 nt); and T. nanjingensis, 0–0·9% (0–11 nt). The 18S BI tree included a major weakly supported clade with all Trichodorus species, one highly supported clade with most Paratrichodorus samples, a weakly supported clade with P. porosus‘A’ and ‘B’, and two distinct clades with Nanidorus renifer and N. nanus, respectively, and a group of N. minor sequences (Fig. 1). The genus Trichodorus was monophyletic, whereas Paratrichodorus and Nanidorus were shown to be paraphyletic. Relationships between the major clades were not well resolved. Trichodorus sparsus‘A’, ‘B’, ‘C’, ‘D’, P. teres‘A’, ‘B’, ‘C’ and P. hispanus‘A’, ‘B’ formed corresponding groups of related clades, whereas there were no sister relationships for T. pakistanensis‘A’ with ‘B’ and P. porosus‘C’ with ‘A’+‘B’, respectively.
D2–D3 of 28S rDNA
The alignment for the D2–D3 of the 28S rDNA included 102 sequences and was 819 nucleotides in length. Thirteen Trichodorus, two Nanidorus and three Paratrichodorus nominal and putative species were included in this analysis.
Intraspecific sequence variations for some species were: P. porosus‘C’, 0–0·6% (0–5 nt); P. pachydermus‘B’, 0% (0 nt); T. sparsus‘A’, 0% (0 nt); T. pakistanensis‘A’, 0–0·6% (0–5 nt); T. primitivus, 0·3% (2 nt); N. renifer, 0–1·6% (0–10 nt); N. minor, 0–1·6% (0–11 nt). In the D2–D3 BI tree Trichodorus samples were distributed among six moderate or highly supported major clades, and Nanidorus and Paratrichodorus represented two and three clades, respectively (Fig. 2). The genus Trichodorus was paraphyletic, whereas Paratrichodorus and Nanidorus were monophyletic.
ITS1 of rDNA
The ITS1 sequences were obtained from four species. Comparison of the sequences revealed the following differences: for P. pachydermus [Czech Republic and UK (AJ439513)], 16% (76 nt); for T. similis [Czech Republic and UK (AJ439523)], 1% (13 nt); and for T. sparsus (‘A’ and ‘B’), 20% (220 nt).
ITS2 of rDNA
The ITS2 sequences were obtained from 13 samples. Sequences of Nanidorus minor, P. porosus‘C’ from California, and T. pakistanensis‘A’ from India showed a high level of similarity (>99%) with sequences of corresponding species from China. Sequences of T. similis from the Czech Republic and UK were also similar. Surprisingly, sequences of T. primitivus from the Czech Republic and UK differed by 37%, whereas in the D2–D3 sequences these samples differed only by 0·3%.
The results of RFLPs of the partial 18S rDNA for Nanidorus, Paratrichodorus and Trichodorus using four restriction enzymes are given in Figure 3. Lengths of restriction fragments after digestion of PCR products with six enzymes are presented in Table 3. Restriction of PCR products by BseNI clearly distinguished Nanidorus, Paratrichodorus (except for P. porosus‘C’) and Trichodorus from each other, in numbers and lengths of fragments (Table 3). Restriction of P. porosus‘C’ by this enzyme resulted in two fragments of similar lengths for some Trichodorus species. However, AvaI clearly differentiated this species from all Trichodorus. Three species of Nanidorus are distinguished by SatI, although the differences between N. minor and N. renifer are based on fragments of less than 60 bp, which may not be clearly visible on agarose gels. PCR-RFLP and virtual digestion of 18S rDNA sequences revealed that six enzymes differentiated P.anemones, P. hispanicus‘B’, P. porosus‘C’, P. teres‘A’ and Paratrichodorus sp. ‘C’ from each other and other Paratrichodorus species. Paratrichodorus hispanus‘A’, P. macrostylus and P. pachydermus‘B’ were indistinguishable from each other by any of the enzymes, as well as Paratrichodorus allius from P. teres‘C’. The six enzymes differentiated most Trichodorus species, except for T. beirensis, T.sparsus‘A’ and ‘B’, T. viruliferus and Trichodorus sp. ‘A’, which generated similar RFLP profiles.
Table 3. Length (bp) of restriction fragments after digestion of PCR products obtained from the partial 18S rRNA gene for Nanidorus, Paratrichodorus and Trichodorus
The D2–D3 PCR-RFLP profiles generated by four enzymes for 12 species of trichodorids are given in Figure 4. Lengths of restriction fragments from RFLP for the D2–D3 fragment of the 28S rDNA for Nanidorus, Paratrichodorus and Trichodorus are presented in Table 4. The four restriction enzymes BseNI, PstI, PvuII and RsaI separated all valid and putative species. The results of PCR-RFLP analysis based on all enzymes studied were identical to those expected from in silico analysis.
Table 4. Length (bp) of restriction fragments after digestion of PCR products obtained from the D2–D3 of the 28S rRNA gene for Nanidorus, Paratrichodorus and Trichodorus
aBold number – fragment verified by PCR-RFLP in this study; italics number – addition fragments.
bPresence of several additional bands on a gel not considered after virtual restriction.
405, 214, 181
444, 189, 161, 6
630, 161, 6
627, 161, 6
501, 161, 128, 6
348, 234, 214
583, 110, 103
629, 161, 6
441, 182, 172
581, 110, 104
246, 223, 161, 128, 21, 6
630, 161, 6
472, 161, 158, 6
796, 516, 280
354, 161, 275, 6
472, 161, 158, 6
455, 181, 121, 12
286, 227, 161, 116, 6
402, 227, 161, 6
350, 232, 214
471, 161, 158, 6
471, 161, 158, 6
268, 214, 133, 115, 66
419, 210, 161, 6
480, 181, 133
243, 226, 161, 158, 6
Trichodorus sp. C
455, 161, 115, 66
797, 431, 366
630, 161, 6b
Trichodorus sp. D
616, 115, 65
629, 161, 6b
Trichodorus sp. E
315, 233, 214, 36
260, 222, 161, 149, 6
Diagnostics of trichodorid nematodes is often difficult because of high intra- and interspecific variability of many morphological and morphometrical characters. Results of phylogenetic and sequence analyses may provide additional criteria to help identify and delimit species. In this study, Bayesian inference was used for phylogenetical reconstruction and species delimiting. Although the results show agreement between molecular and morphological identification for many trichodorid species, identification of some samples remains uncertain because of the presence of different sequences under the same specific name in GenBank. In this study a letter code was assigned for samples clustered separately in phylogenetic trees and morphologically identified as representatives of a single trichodorid species.
Phylogenetic and sequence analysis of the partial 18S and 28S rRNA gene sequences revealed a group of related sequences morphologically identified as representatives of the species T. sparsus (types ‘A’, ‘B’, ‘C’, ‘D’). A high level of variation of morphological and morphometric characters has been reported between populations of T. sparsus from different countries (Loof, 1973; Peneva, 1988; Decraemer, 1995). Recently, Decraemer et al. (2008) distinguished three morphotypes of T. sparsus from Serbia based upon a combination of morphological characters and morphometrics, and noticed that separation between different morphotypes was not straightforward. Probably, the T. sparsus group consists of several sibling (or cryptic) species or subspecies. Final identification of T. sparsus isolates and other samples marked in this study as T. pakistanensis‘A’, ‘B’, P. porosus‘A’, ‘B’, ‘C’ and P. pachydermus‘A’, ‘B’ will be possible after a more thorough morphological characterization of populations. Collection and molecular characterization of nematode materials from the type localities may be critical to resolving these identification problems. Species delimitation is controversial and should rely on the consensus of several data sets and criteria. Additional morphological, molecular and biogeographical data should be used to confirm the delimitation of species made in this and other studies.
PCR-RFLP analysis of ribosomal RNA gene sequences is one of the most effective methods of identification of different nematode groups comprising many species. The PCR-RFLP technique is a simple, rapid and cost-effective technique in comparison with other techniques. Because it is assumed that the sequence of the rRNA gene is conserved within a species, but diverse between species, selection of an appropriate gene marker for identification is crucial in developing a diagnostic. Duarte et al. (2011) designed a PCR-RFLP assay for identification of 12 trichodorids using a fragment of the 18S rRNA gene. The present analysis showed that not all species can be identified using this region. The D2–D3 of the 28S rRNA gene has higher interspecific variation and evolves more rapidly than the 18S rRNA gene, and therefore appears to be a more appropriate marker for identification of multiple species when compared with the 18S rRNA gene. Restriction of D2–D3 amplicons by four enzymes produces species-specific restriction patterns for all species analysed. Even though this study demonstrates advantages of PCR-RFLP analysis for the identification of stubby nematodes, ribosomal RNA gene markers may exhibit intraspecific polymorphism, which will cause variation in restriction maps in different geographically distinct populations. Thus, it will be necessary to investigate the variation of sequences from different geographically separated populations for some species before application of this technique.
Phylogenetic relationships within trichodorids were recently studied by Van Megen et al. (2009) and Duarte et al. (2010) using full length 18S rRNA sequences, with only 17 and 15 stubby nematode species included in these analyses, respectively. The present study, which included more species and compared a smaller fragment of the 18S rRNA gene, showed similar relationships between taxa as previous studies. However, this analysis generated low supports for relationships between major clades and also revealed artifactual paraphylies for Paratrichodorus and Nanidorus. Major differences in topologies between trees were observed mainly in positions of poorly supported clades. Observed paraphyly for Trichodorus in D2–D3 of the 28S rRNA gene tree here and the 18S rRNA gene tree by Van Megen et al. (2009), and for Paratrichodorus and Nanidorus in the 18S rRNA gene tree here, might be explained by increasing for absolute total number of species on the branches and nodes with more descendent species, which lead to an unbalanced tree. The increase in imbalance is consistent with a cumulative effect of differences in diversification rates between branches (Holman, 2005). Validity of the genus Nanidorus has been the subject of some debate (Siddiqi, 1980, 2002) and only recently Duarte et al. (2010) proposed accepting Nanidorus as a valid genus based on morphological features and the results of molecular analysis of the 18S rRNA gene. Although the 18S rRNA gene analysis in the present study does not generate enough resolution to understand relationships between Nanidorus, the D2–D3 of the 28S rRNA gene, having a higher phylogenetic signal, gives clear evidence for its monophyly and also confirms closer relationships of this genus with the genus Trichodorus rather than with Paratrichodorus.
Observed clustering of trichodorid species in the tree was generally in agreement with reported phenotypic similarity of male characters (Decraemer, 1995; Decraemer & Baujard, 1998). Sorting of three prime characters (number of ventromedian precloacal supplements, number of ventromedian cervical pappillae and habitus), Decraemer & Baujard (1998) distinguished 13 male groups within Trichodoridae. In the trees here, T. lusitanicus, T. viruliferus, T. beirensis, T. similis, T. primitivus and T. variopapillatus, all belonging to the male Group 12, and T. cylindricus from Group 11, clustered together. Trichodorus sparsus (Group 10) was in a separate clade on D2–D3 of the 28S rRNA and 18S RNA gene trees. Two major subclades of Paratrichodorus on the 18S rRNA gene tree also corresponded to the male groupings: (i) P. allius (Group 2), P. porosus (Group 3), P. teres (Group 4); and (ii) P. anemones, P. pachydermus, P. hispanus, P. macrostylus (Group 6).
Thus, PCR-RFLP and sequencing of ribosomal RNA markers appear to be a useful and appropriate method for characterization and accurate identification of stubby nematodes. However, larger numbers of species and populations from diverse origins and other alternative gene markers should be included and analysed in future studies to confirm the findings made in this work.
The authors acknowledge Professor F. Decraemer for help with identification of some trichodorid samples and Dr C. Blomquist for critical reading of the manuscript draft. The work was partly supported by the Ministry of Agriculture of the Czech Republic, project number MZe – 0002700604, etapa 09 and the Russian Foundation for Basic Research, project number 11-04-01514.