Molecular Characterization and Phylogeny of Ditylenchus weischeri from Cirsium arvense in the Prairie Provinces of Canada

In total, 45 samples of Ditylenchus were examined, of which 35 were from C. arvense plants and six from C. arvense in grain samples. Four samples of Ditylenchus dipsaci were also obtained from garlic bulbs, two from each province of Ontario and Quebec. Code and origin of the samples are given in Table 1. Most of the genomic DNA from samples was obtained from the study of Tenuta et al. (2014). The samples were coded according to their origin: (1) pea harvested grain with C. arvense seed (PG) from six fields across the Prairie Provinces, PG-(58-AB) and PG-(33) from Alberta, PG-(76-SK) and PG-(44) from Saskatchewan, and PG-(81-MB) and PG-(84-MB) from Manitoba; (2) C. arvense from four pea grain fields, CT-(81-MB), CT-(84-MB), CT-(B-MB), CT-(C-MB), all from Manitoba; (3) C. arvense from roadsides of agricultural lands at two locations in Manitoba, CT-(Roadside1-2-MB); (4) C. arvense collected from five agricultural fields in Manitoba CT-(C11F11-MB) and CT-(C11F17-MB), in Saskatchewan CT-(C11F7-SK) CT-(C11F18-SK) and CT-(C11F19-SK); (5) C. arvense collected from 23 locations in agricultural ditches in Manitoba CT-(GL1 to GL23-MB); (6) garlic infested with D. dipsaci, two from each province of Ontario G-(Guelph-ON- and G-(C58-ON), and Quebec G-(PQ-GLG), and G-(PQ-D77); and (7) dead and dried individuals of D. weischeri, CT-(RU) from Russia.

Briefly, nematodes had been recovered from samples using a technique based on soaking. For this, stems, leaves and flowers of C. arvenss were chopped separately and soaked in distilled water in Petri dishes overnight. Similarly, samples of PG (pea harvested grain contaminated with creeping thistle seed), and pieces of stems and corm from the garlic samples G-(Guelph-ON), G-(C58-ON), G-(QP-GLG), and G-(QP-D77) were also soaked overnight in water. After soaking, the samples were checked using a stereo microscope and several individual juvenile, female, and male nematodes were picked using an eyebrow brush. Selected nematodes were stored at 4°C in 1.5-ml tubes containing either in distilled water when the nematodes needed to be used immediately or DESS solution for long-term preservation of the nematodes (Yoder et al., 2006).

Genomic DNA had been recovered from juvenile or female individuals using a protocol described previously by Subbotin et al. (2005). Briefly, one nematode was cut into two or three segments under a stereo microscope in a drop of worm lysis buffer (WLB, 10 Mm Tris-HCl pH 8.0, 50 mM KCl, 1.5 mM MgCl₂, 1 mM DTT, and 0.45% Tween 20) using a sterilized surgical design scalpel blade No 11 (Fisher Scientific, ON, Canada). Approximately 15 to 20 µl of the nematode suspension was then transferred to a PCR reaction tube (Eppendorf) and 2 µl proteinase K was added; the tubes were then incubated at 60°C and 95°C for 60 and 10 min, respectively. The DNA was used immediately or frozen at −20°C before use in PCR. The quality and quantity of the DNA were determined suitable for PCR by a spectrophotometer using a NanoDrop ND-1000 (Thermo Scientific, Mississauga, ON) at 260 nm.

PCR reactions were conducted to a 25 µl final reaction volume. Each reaction contained 10 µl of 2× Master Mix (Qiagen, GmbH, Hilden, Germany) including dNTPs and Taq polymerase, 0.5 µl of each primer, 2 µl of DNA extract, and 12.5 µl of dH₂O. ITS (rDNA), LSU D2D3 28S rDNA, partial segment of SSU 18S rDNA, and Hsp90 genes were targeted for PCR amplification. The whole ITS rDNA including ITS1, 5.8S, and ITS2 was PCR amplified using the forward primer TW-81 (5′-GTTTCCGTAGGTGAACCTGC-3′) and AB-28 reverse (5′-ATATGCTTAAGTTCAGCGGGT-3′) (Joyce et al., 1994). When the ITS rDNA could not be amplified using this primer set, the PCR was run with the forward primer rDNA1 (5′-TTGATTACGTCCC TGCCCTTT-3′) and reverse primer rDNA2 (5′-TTTCACTCGCCG TTACTAAGG-3′) developed by Vrain et al. (1992). The LSU D2D3 28S DNA gene was amplified using the forward primer D2A (5-CAAGTACCGTGAGGG AAAGTTG-3) and the reverse primer D3B (3-TCCTCGGAAG GAACCAGCTACTA-3) as used by De Ley et al. (1999). In some cases, this primer set failed; to overcome this issue an initial amplification using a nested forward primer D2AA (5′-GAAACGGATAGAGCCGACGT AT-3′) and reverse D3BB (5′-CCAAGTCAGACGA TCGATT-3′) with D2A-D3B was used. The D2AA and D3BB primers were designed on the basis of available sequences of Ditylenchus spp. in GenBank. A partial segment of the SSU 18S DNA gene was amplified (Blaxter et al., 1998) using the forward primer G18S4 (5-GCTTGTCTCAAAGATTAAGCC-3) and the reverse primer 18SP (5-TGATCCWMCRGCAGGTCAC-3). The pair of degenerate primer U831 (5′-AAYAA RACMAADCCNTYTGGAC-3′) and L1110 (5′-TCRCARTTVTCCATGA TRAAVAC-3′) (Skantar and Carta, 2004) was used to amplify the partial region of the Hsp90 gene. For the samples which were not amplified using this set of primers, a nested primer of 97F (5′-GCGCAATCCGRMYGAAGT-3′) and 93 R (5′-ACGGCATTCTTGC TCTTCTT-3′) was designed and used in PCR. The PCR conditions were 3 min denaturation at 94°C, followed by 30 s at 94°C, 45 s at 54°C, and  min (for D2A-D3B; D2AA- D3BB) or 2 min (for ITS rDNA) for 34 cycles. As a final step, a 10 min extension at 72°C was applied for all the PCR reactions. Each PCR run included DNA of a positive control of known DNA concentration to ensure performance of the reaction. For all PCR reaction runs, a PCR tube without DNA was used as negative controls. All PCR amplifications were performed using an Eppendorf thermal cycler (Master Cycler Pro, Hamburg, Germany).

To analyze the results of PCR amplification, the PCR products were prepared for gel electrophoresis by mixing 2 µl of the PCR product with loading buffer in water and added to 1% agarose gel with 0.5× TAE buffer to which GelRed fluorescent dye was included (1 µl, 30 ml⁻¹, Biotium, Inc, Hayward, CA). Gel electrophoresis was run for 45 min under constant current (45 to 50 mA) and amplification products viewed using a Gbox gel capture imaging system (SYNGENE, Synoptic, Ltd. Cambridge, UK) equipped with a UV transiiluminator.

PCR products were loaded on 1% agarose gel prepared in 0.5× TAE buffer, excised and purified for further analysis. Purification was performed using a Qiaquick gel extraction kit (Qiagen, GmbH, Hilden, Germany) according to manufacturer protocol. Purified PCR products were sequenced directly in both strand directions using the original primers of each amplified gene. Sequences with background noise were cloned and sequenced. Cloning was performed after ligation of 30 to 50 ng of PCR product to the pGEM-T Easy Vector System (Promega, Madison, WI). The ligation product was then transformed into DH bacteria following the supplier protocol (BioLabs, New England, MA). Plasmids of positive white colonies were isolated using the UltraClean Standard Mini Plasmid Prep Kit (MO BIO Laboratories, Carlsbad, CA); sequencing was performed with the universal vector primers M13F (5′-TGTAAAACGACGGCCAGT-3′) and M13R (5′-GTAATACGACTCACTATAGGGC-3′).

Raw sequence results were visualized using the program Chromas Lite 2.01 (Technelysium Pty Ltd, South Brisbane, Australia), edited as needed and then assembled. Additional editing of sequences was performed using Gendoc 2.7 (Nicholas et al., 1997). All sequencing was performed by the core sequencing facility of Macrogen Corp (Rochville, MD).

Multiple sequence alignments were performed for the sequences generated in this study together with selected 59 published sequences of the ITS rDNA gene from Ditylenchus species deposited in GenBank (Table 2). For each gene, multiple sequence alignments were performed using MegAlign 7.1.0 of Lasergene (Dnastar Inc, Madison,WI) for sequences generated in this study along with those of the same gene for D. dipsaci, D. weischeri, D. gigas, D. destructor, and the unidentified Ditylenchus spp. in the GenBank. The accessions used were mainly from Helder et al. (2004), Subbotin et al. (2005, 2011), Chizhov et al. (2010), and Vovlas et al. (2011) (Table 2). After alignment, consensus parts of each alignment set from the primer ends were trimmed using Gendoc 2.7 (Nicholas et al., 1997) and analysis was performed on a 631 bp region across ITS1, 5.8S, and ITS2 genes. Comparison of the sequence for nucleotide composition and position was then performed among the studied samples along with 24 sequences retrieved from GenBank for D. dipsaci, D. gigas, and selected D. spp. (Table 3).

New sequences generated in this study of the SSU 18S rDNA gene were aligned with six sequences of D. dipsaci and one of D. gigas retrieved from GenBank (Table 2). The sequences were all trimmed to 1390 bp and intra- and inter-sequence variation analyses, similarity matrix, and pairwise genetic distance were estimated using MEGA5 (Tamura et al., 2011).

Alignment of 29 LSU D2D3 28S rDNA sequences from D. dipsaci, D. gigas, D. destructor, D. halictus (Giblin-Davis et al., 2010) and Ditylenchus sp from GenBank (Table 2), and new sequences from this study were done as described above. Analysis was performed on 637 bp of the trimmed sequences of the LSU D2D3 28S rDNA region. Each of the studied gene sequences were analyzed by identifying single-nucleotide polymorphisms (SNPs), base pair composition, and interspecies and intraspecies sequence variation.

The sequences obtained in this study with 14 sequences available from the Hsp90 gene of Ditylenchus species including three of D. gigas, D. weischeri, and Ditylenchus sp, plus five sequences of D. dipsaci were retrieved for GenBank (Table 2), and subjected to multiple sequence alignment and pairwise distance analysis.

Phylogenetic analysis was performed on the ITS rDNA, LSU D2D3 28S rDNA, and Hsp90 genes. The most appropriate nucleotide substitution and parameter values were determined using jModelTest 2.1.7 (Darriba et al., 2012). The gamma distribution shape parameter and substitution rate in the Akaike information criterion were included in the analysis to determine the best fitting evolutionary model. Based on the previously published data on Ditylenchus (Vovlas et al., 2011, Subbotin et al., 2005, 2011), the following outgroup taxa were used, D. myceliophagous Goodey 1958 (DQ151458 and AM232236) for ITS rDNA and D. halictus (AY589364) for LSU D2D3 28S rDNA. Ditylenchus sp. (AY603512) was used as the outgroup for Hsp90 analysis. For each of the studied genes, Bayesian interference analysis was then performed to confirm the tree topology using MrBayes 3.2.6 (Ronquist et al., 2012) to generate a 50% majority rule consensus. Tree View 1.6.6 (Page, 1996) was then used to visualize trees and the posterior probabilities are given on appropriate clades.

Amplification of the ITS rDNA resulted in a single amplicon sized 719 or 964 bp with the primer pair AB-28/ TW-81 or rDNA1/rDNA2, respectively. The LSU D2D3 28S rDNA gene was amplified for 25 samples. The amplicon size obtained was 783 bp using the primer pair D2A- D3B and 660 bp when the nested primer pair D2AA- D3BB was used. No product for the LSU D2D3 28S rDNA was obtained for garlic samples with the primers sets D2A- D3B or when nested with D2AA- D3BB (Table 1).

For SSU 18S rDNA, the size of the PCR product was primarily 1514 bp and in some cases it was 1620 bp. No SSU 18S rDNA PCR product was obtained for samples CT-(C11F11-MB), CT-(C11F17-MB), CT-(C11F19-SK), PG-(44-SK), G-(Guelph-ON), G-(C58-ON), G-(PQ-GLG), and G-(PQ-D77).

The PCR amplified product of the Hsp90 gene was 303 bp or 212 bp when the nested primer 97F/93R was used in PCR. For samples PG-(58-AB), CT-(C11F18-SK), and CT-(C-MB), an additional PCR product of 558 or 484 bp was also amplified when the primer set U831/L110 was used. The Hsp90 gene was amplified from most of the samples, but not from samples CT-(Roadside1-MB, and Roadside2-MB), CT-(C11F7-SK), CT-(C11F17-MB), CT-(81-MB, 84-MB, B-MB), and PG-(44-SK, 76-SK). No amplification was observed for the negative control without DNA.

For some nematode samples more than one sequence was assembled; as a consequence, from the 47 studied samples, a total of 74 sequences were assembled from ITS rDNA (Table 1). Forty-seven sequences were from nematodes retrieved from C. arvense including 25 samples coded CT-(GL-1 to GL-23-MB), seven from non-pea agricultural land CT-(C11F7-SK and C11F19-SK), 2 from CT-(C11F11-SK and C11F17-MB), six from four fields of pea grain in Manitoba CT-(81-,84-,B-,C-MB), four from CT-(Roadside1 and Roadside2-MB), and three from CT-(RU). Nine sequences were obtained from nematodes isolated from pea harvested seed contaminated with thistle seeds, including one from each collection of PG-(33-AB), PG-(76-SK), PG-(81-MB), PG-(84-MB), and five from pea grain field collections of PG-(58-AB). Next to these sequences, 18 sequences originated from D. dipsaci of the garlic collection including six from each G-(Guelph-ON) and G-(PQ-GLG), and 3 from each G-(C58-ON) and G-(PQ-D77). Results of the sequence comparison are presented in Table 3.

Comparison of the Canadian D. weischeri sequences obtained with the Russian haplotype of this species deposited in GenBank showed only one single-nucleotide transition (T/A) at position 174. Intraspecies sequence variation among all D. weischeri analyzed in this study, including the Russian haplotype, was limited to only two sites presenting a transversion (A/T) in position 78 bp located in ITS1 and a transition (C/T) in position 515 bp, located in ITS2.

All the D. weischeri sequences showed a dissimilarity of 5% and 3.8% in the whole ITS rDNA sequence of D. dipsaci and D. gigas, respectively. Including all the sequences for analysis showed an overall dissimilarity of 6.4% between D. weischeri and rest of the samples including D. dipsaci, D. gigas, and D. sp. A comparison of nucleotide composition and position among the selected samples is given in Table 3. There were 11 SNPs specific to all D. weischeri but not in D. dipsaci, D. gigas, and D. sp, with nucleotides and position as follows: (A/T-22), ( -/A-244), (T/A-278), (T/G-460), (GA/AC-529-530), (AG/GA-531-3), (G/A-595), (-/A-601), (G/T-602), (TA/GT-613-14). Nucleotide composition (%) for each of the species was: D. weischeri (T = 29.8%), (C = 20.7%), (A = 25.2%), (G = 24.4%), and a GC content of 45.1%; D. dipsaci (T = 29.9%), (C = 21.3%), (A = 24%), (G = 24.8%), and a GC content of 46.1%; D. destructor (T = 33.3%), (C = 21.2%), (A = 20.1%), (G = 25.2%), and a GC content of 46.4%; D. gigas (T = 29.8%), (C = 20.7%), (A = 25.3%), (G = 24.2%), and a GC content of 49.9%; Ditylenchus. sp: (T = 30.2%), (C = 19.3%), (A = 27.1%), (G = 23.4%), and a GC content of 42.7%.

A total of 27 sequences of LSU D2D3 28S rDNA were obtained (Table 1). Nineteen sequences were from C. arvense and included 11 from samples coded CT-(GL-1 to GL-23-MB), one from each non-pea agricultural land CT-(C11F19-SK) in Saskatchewan, and CT-(C11F11-MB) in Manitoba, three fields of pea grain in Manitoba CT-(84,B,C-MB), two from CT-(Roadside1-MB and Roadside2-MB), and one from CT-(RU).

Also seven sequences were obtained from pea harvested seed contaminated with thistle seeds, including three from PG-(33-AB), and one from each PG-(81-MB), PG-(84-MB), PG-(44-SK), and PG-(58-AB). No sequences were obtained from any of D. dipsaci from garlic, including G-(Guelph-ON), G- (PQ-GLG), G-(C58-ON), and G-(PQ-D77). Intraspecific variation in sequences of D. weischeri was observed for nucleotide transversions (G/T and A/C) at position 565 bp and 561 bp, respectively, and one transition (C/T) at position 564 bp (Table 4). Sequences could not be assembled from samples from garlic (G-ON and G-PQ) due to the presence of background noise in the raw sequence data. An aalysis of the aligned sequences showed 2.6% dissimilarity between D. dipsaci retrieved from GenBank and D. weischeri from this study, which corresponded to 23 SNPs. Similarly, there were 11 SNP differences between D. wiescheri and D. gigas. The nucleotide composition of the 637 bp section of LSU D2D3 28S rDNA were: D. weischeri (T = 24.9%), (C = 20.3%), (A = 24%), G = (30.8%), D. dipsaci (T = 25.2%), (C = 20.2%), (A = 23.9%), (G = 30.8%), D. destructor (T = 24.9%), (C = 21.3%), (A = 22.2%), (G = 31.7%), and D. gigas (T = 25.7%), (C = 20.3%), (A = 23.7), (G = 30.4).

In total, 22 partial sequence of the SSU 18S rDNA gene was obtained of which 11 sequences were from C. arvense which include one from each CT-(GL-5-MB), CT-(GL-6-MB), and CT-(GL-7-MB) from Manitoba (Table 1). Sequences were not obtained from samples PG-(44-SK) and PG-(81-MB). Sequences were obtained for CT-(84,B,C-MB), three from CT-(Roadside1-MB, and Roadside2-MB), and three from CT-(RU), two from each PG-(33-AB) and PG-(58-MB), and one from each PG-(76-SK) and PG-(84-MB). Two sequences were obtained from D. dipsaci from garlic, including one from each G-(Guelph-ON) and G-(C58-ON). The results from sequence comparison showed no intraspecies sequence variation among D. weischeri individuals. Also, Canadian D. weischeri SSU 18S rDNA sequences were 98% identical to those deposited in GenBank for D. dipsaci and D. gigas.

A total of 18 sequences were assembled from the examined individuals (Table 1). Ten of these sequences were from C. arvense and included one from each CT-(GL6-MB, GL7-MB, GL8-MB, GL-10-MB), (C11F11-MB, C11F18-SK, C11F19-SK) and two from CT-(84-MB). Also from pea grain samples with creeping thistle seeds, one sequence from each PG-(33-AB), PG-(58-AB), and PG-(81-MB) was obtained. From the garlic samples two sequences from G-(Guelph-ON) and one from each G-(C58-ON), G-(PQ-GLG) and G-(PQ-D77) was assembled. The samples with positive PCR reaction for amplification of the Hsp90 gene and their sequences are presented in Table 1. In GenBank, 12 sequences were available and compared with three D. weischeri from Russia, four D. dipsaci, two D. gigas, and three Ditylenchus sp. Intraspecific sequence analysis of 230 bp of Hsp90 obtained from D. weischeri in this study showed transitions (A/G and C/T) and two transversions (G/C and G/T) at positions 11, 165, 37, and 205, respectively. Sequence variation between Canadian D. weischeri individuals and the haplotype of this species from Russia was eight nucleotides (3.8%); for the rest of the samples the variation was 6.5%. Hsp90 sequences for D. dipsaci obtained from garlic in this study was identical to those of the same nematode species deposited in GenBank from Allium cepa L., 1753, and Fragaria L., 1753, sp., except for few point mutations. All D. dipsaci had 6.7% dissimilarity to D. weischeri. The result of pairwise sequence distance matrix (data not shown) for these samples showed that the overall mean distance was 0.095 among all analyzed samples, and 0.001 within D. weischeri. The shortest and the longest distance was 0.0% and 2.47%, respectively, among all the analyzed samples. The nucleotide composition for the Ditylenchus species was: (T = 31.3%), C = (22.3%), (A = 25.9%), (G = 20.5%) for D. weischeri, and (T = 31%), (C = 17.7%), (A = 26.5), (G = 24.8) for D. dipsaci.

A total of 142 sequences determined in this study were submitted to GenBank on October 18 and 19, 2017. The GenBank accession numbers for samples sequenced in this study are given in Table 1

Hierarchical clustering analysis based on ITS rDNA sequences yielded clear groupings among the examined samples with D. weischeri present in a distinct clade. The maximum parsimony base tree is presented in Figure 1. The resulting tree clustered all the 132 sequences into two main clades (I and II) with a posterior probability value of 100. Clade I comprised of D. dipsaci, D. gigas, D. weischeri, and an unknown Ditylenchus species (D. sp.); clade II contained only D. destructor. Samples in clade I were divided into two sub-clades (III and IV). Sub-clade III comprised all samples from C. arvense including CT-(GL-MB), CT-(81-MB, B-MB, C-MB and 84-MB), CT-(C11F7-SK, C11F17-MB, C11F18-SK, C11F19-SK), CT-(Roadside1-MB and Roadside2-MB), CT-(RU), PG-(33-AB, 44-SK, 58-AB,76-SK, 81-MB, 84-MB), and the sample of D. weischeri retrieved from the GenBank. Ditylenchus gigas and Ditylenchus sp. composed a separate group in sub-clade III. This sub-clade was branched by a posterior probability value of 80 from the second sub-clade (IV) that included all D. dipsaci samples obtained in this study, G-(Guelph-ON), G- (PQ-GLG), G-(C58-ON), and G-(PQ-D77) and those from GenBank. None of the studied samples from C. arvense or sequences of D. gigas and Ditylenchus sp from GenBank were clustered with sub-clade IV that contained solely D. dipsaci.

Figure 1:

Hierarchical clustering analysis of 48 partial sequences of LSU D2D3 28S rDNA of samples obtained in this study and species of D. dipsaci, D. gigas, and D. destructor retrieved from the GenBank is presented in Figure 2. Two distinct clades were formed that delineated D. destructor from clades that contained D. dipsaci, D. wiescheri, and D. gigas (posterior probability value of 98). All the samples of creeping thistle from this study including CT-(GL-MB), CT-(81, B, C and 84), CT-(C11F7-SK, C11F17-MB, C11F18-SK, C11F19-SK), CT-(Roadside1-MB and Roadside2-MB), CT-(RU), PG-(33-AB, 44-SK, 58-AB, 76-SK, 81-MB, 84-MB) were clustered together and were separated from D. dipsaci and D. gigas with a posterior probability value of 98. Sequences of a Ditylenchus sp. were positioned between the two clades, one contained D. destructor and the second one contained D. dipsaci samples. The two samples of D. gigas placed in a sub-clade with a posterior probability value of 100 and separated from D. weischeri of Canadian origin with a posterior probability value of 98. Interestingly, a sequence of D2-D3 from a sample of Ditylenchus deposited in GenBank under accession number FJ70364 as D. dipsaci was also grouped with D. weischeri. The sample was acquired from field pea from Prairie Canada and provided for the accession deposited by Douda et al. (2013) by the Canadian Food Inspection Agency. Clearly, the accession is for D. weischeri and not D. dipsaci.

Figure 2:

The Hsp90-based parsimonious tree is presented in Figure 3. The sequence of the sample of D. weischeri from the original study of Chizhov et al. (2010), coded as CT-(RU), was grouped in a main clade together with PG-(33-AB, 81-MB, 84-MB, 58-AB), CT-(C11F11-MB, C11F18-SK, C11F19-SK), and CT-(GLG-MB). This clade was separated from two other clades that grouped D. dipsaci or D. gigas. The Ditylenchus sp. from garlic (G-(Guelph-ON), G-(C58-ON), and G-(PQ-D77) was clustered with D. dipsaci retrieved from GenBank. Since the SSU 18S rDNA gene based phylogeny is more applicable for higher taxon, i.e. family and higher, phylogenetic analysis was not performed on SSU 18S rDNA gene sequence data. Moreover, there were limited accessions of SSU 18S rDNA gene sequences in GenBank making the phylogenetic tree analysis not possible.

Figure 3:

Results of analysis of the different genes, ITS rDNA, LSU D2D3 28S rDNA, SSU 18S rDNA, and Hsp90 provide evidence that D. weischeri that is parasitizing C. arvense is genetically distinct from D. dipsaci to be distinguishable from each other. Sequence analysis of the studied genes showed the presence of heterogeneity between individuals of this species. Except for the ITSrDNA gene, there were not many sequences available for other genes in the GenBank database. Even for D. weischeri, there were few ITS rDNA sequences from a limited number of individuals. Among the studied genes, the least sequence divergence was observed for the SSU 18S rDNA gene. The Hsp90 gene showed the highest variability of sequence among species, and is the most informative marker in terms of discrimination between species, and thus likely the reason that enabled us to use it for designing species-specific primers in diagnostic PCR, (Madani et al., 2015).

Comparison of the ITS 1 and ITS 2 among the studied samples of D. weischeri and D. dipsaci showed that a higher nucleotide sequence variation existed in ITS2, thus ITS2 possibly alone would be enough informative for the differentiation of this species. This is in accordance with the published data showing suitability of ITS 2 for barcode identification and discrimination of main taxa in plants and animals (Hui et al., 2010). The GC content for D. weischeri, D. dipsaci, and D. gigas showed to be higher for the D. dipsaci group than the others. The two sequences of D. dipsaci each isolated from pea and garlic (Vovlas et al., 2011) were identical, but they had 22% dissimilarity to a sequence of D. dipsaci from pea (Helder et al., 2004).

While our results showed 98% similarity of Canadian D. weischeri to the SSU 18S rDNA sequence of D. dipsaci, we noticed some discrepancies in the sequences of this gene deposited in the GenBank. For example, the two sequences from two pea isolates of D. dipsaci each from the study of Volvas et al. (2011) and Helder et al. (2004) showed differences at 22%, and two sequence of D. dipsaci each isolated from pea and garlic (Vovlas et al., 2011) showed identical sequences. It was not possible to compare the Canadian samples of D. weischeri for the LSU D2D3 28S rDNA gene sequence due to lack of available sequences for the Russian D. weischeri.

All dendrograms generated in this study provided almost the same pattern for grouping of D. weischeri compared to the other Ditylenchus species examined. Trees were constructed based on the ITS rDNA, LSU D2D3 28S rDNA, and Hsp90 genes. These are the most common molecular markers for phylogenetic analysis; in contrast, we did not use the SSU 18S rDNA gene for analysis as it is less common, had limited sequence accessions of other Ditylenchus in GenBank, and that SSU 18S rDNA gene is more suitable for taxon discrimination at the family or higher level (Holterman et al., 2009; Wu et al., 2015).

Congruent phylogenies using each of ITS rDNA, LSU D2D3 28S rDNA, and Hsp90 genes have been a robust method for integrative diagnostics, particularly for closely related species such as the cyst-forming nematodes (i.e., Skantar et al. 2012; Madani et al., 2017) and Ditylenchus species (Douda et al., 2013). Samples of D. weischeri used in this study were positioned in a separate clade for each of the generated trees using those genes. The three sequences of Hsp90 retrieved from the Genbank for D. weischeri also were positioned next to all D. weischeri in a sub-clade on the Hsp90-based phylogenetic tree. Also in ITS rDNA tree a sample of D. weischeri collected in Saskatchewan (731-CT-C11F7) was positioned in a sub-clade with the sequence of this nematode from Russia. Although sequence similarity search using the Blastn option in NCBI, using sequences of D. weischeri generated in this study showed high similarity to the ITS rDNA of D. gigas, phylogenetic analysis positioned this species in a separated clade.

In conclusion, sequence analysis of the four genes studied provide more insight to the genome of Ditylenchus species and in particular D. weischerii, for the discrimination of this species from the closely related species of D, dipsaci complex, i.e. D. dipsaci. This was confirmed by applying phylogenetic analysis using the most commonly used genes for plant parasitic nematodes, i.e. ITS rDNA, LSU D2D3 28S rDNA, and Hsp90. Results from this study and of past reports of plant host preference and morphology differences provide further evidence that D. weischeri is distinct to be a separate species from D. dipsaci. Some differences in gene sequences and morphology to the Russian haplotype and less severe symptoms of disease on C. arvense in Prairie Canada suggests that the Canadian and Russian populations of D. weischeri may be diverging.