Improvement in group identification of dojo loach, Misgurnus anguillicaudatus, using PCR-restriction fragment length polymorphism

In most Japanese populations of dojo loach (Misgurnus anguillicaudatus), gonochoristic diploids of genetically diversified groups (A and B, further subdivided into B1 and B2) are present, whereas unisexual clonal lineages inhabit certain localities in the Hokkaido and Ishikawa Prefectures in Japan. Through a series of genetic studies including DNA markers, the clonal loaches were deemed to originate from a hybridization event(s) between the A and B1 groups. However, combined analyses with other DNA markers are needed to identify each genetic group. In this study, we improved the PCR-restriction fragment length polymorphism (RFLP) analysis of the recombination activating gene 1 (RAG1) gene using digestion with two restriction enzymes, PvuII and StuI. The improved RAG1-RFLP analysis showed different fragment patterns for each group: two fragments (245 and 198 bp) for group A, three fragments (198, 147, and 98 bp) for group B1, and a single fragment (443 bp) for group B2. The clonal loaches exhibited four fragments (245, 198, 147, and 98 bp) derived from both groups A and B1. Moreover, the DNA markers were able to detect two different hybrid genotypes (A × B2 and B1 × B2). Thus, the improved RAG1-RFLP markers allowed for quick and accurate group identification of the dojo loaches.


Introduction
In most Japanese populations of dojo loach, Misgurnus anguillicaudatus (Cobitidae; Teleostei), bisexually reproducing gonochoristic diploids (2n = 50) are present, whereas unisexual clonal lineages inhabit certain localities in the Hokkaido and Ishikawa Prefectures in Japan (Morishima et al. 2002(Morishima et al. , 2008Arai and Fujimoto 2013). The clonal diploids generate unreduced diploid eggs that develop by gynogenesis without any genetic contribution of sperm from sympatric bisexual wild types (Itono et al. 2006(Itono et al. , 2007Arai and Fujimoto 2013). Previous population genetic studies using allozymes (Khan and Arai 2000), microsatellites (Arias-Rodriguez et al. 2007), and sequences of the control region in mitochondrial DNA (mtDNA-CR) (Morishima et al. 2008) clarified that there are two highly diversified groups, A and B (the latter further subdivided into B1 and B2) in the Japanese wild populations. In Nakaikemi Wetland, Fukui Prefecture, although groups A and B loaches are sympatric, reproductive isolates between the two groups were confirmed and suggested Japanese dojo loach should be recognized as two distinct biological species (Okada et al. 2017). Sequence analyses of recombination activating gene 1 (RAG1) and interphotoreceptor retinoid-binding protein 2 (IRBP2) genes also supported the presence of diverse groups (Yamada et al. 2015). A hybrid origin between groups A and B1 was strongly suggested in clonal loaches because of the heterozygosity of RAG1 and IRBP2 sequences (Yamada et al. 2015). Restriction fragment length polymorphism (RFLP) analyses of RAG1 sequences with the restriction enzyme PvuII provided different fragment patterns among the groups. Specifically, groups A and B2 showed a single 1 3 fragment (443 bp), while group B1 showed two fragments (296 and 147 bp) (Fujimoto et al. 2017). Clonal loaches had three fragments (443, 296, and 147 bp) derived from both groups A and B1 (Fujimoto et al. 2017). Although the RAG1-RFLP marker is a useful tool for identifying genetic groups in dojo loaches, discrimination of groups A and B2, and the hybrids between groups A and B2 is impossible because all of the individuals show a single fragment (443 bp). Similarly, it is impossible to distinguish clonal loaches from hybrids between groups B1 and B2 because the three fragments that are detected are the same sizes (443, 296, and 147 bp).
Different nuclear DNA markers, ManDra (hereafter designated as ManDra-B in this paper), ManDra-A, and Man-Bgl, were developed from repetitive sequences isolated by digestion of genomic DNA with the restriction enzymes DraI and BglII (Fujimoto et al. 2017;Kuroda et al. 2021). The DNA markers ManDra-B and ManDra-A were designed to amplify isolated repetitive sequences by PCR and were used for grouping based on the electrophoretic patterns of the PCR products. Specifically, ManDra-B yields ladder-like electrophoretic patterns in group A, but smear-like patterns in groups B1 and B2 (Fujimoto et al. 2017). In contrast, ManDra-A shows smear-like patterns in group A, but ladder-like patterns in group B1 (Kuroda et al. 2021). Thus, both ManDra-B and ManDra-A show smear-like patterns in clonal loaches (Fujimoto et al. 2017;Kuroda et al. 2021). Similarly, for the ManBgl marker, a 400 bp fragment has been amplified by PCR in group A, while a 460 bp fragment without the 400 bp fragment has been shown in groups B1 and B2 ( Therefore, combined genetic analyses using the abovementioned DNA markers (RAG1-RFLP, ManDra-B, Man-Dra-A, and ManBgl) are needed to completely distinguish the genetic groups of dojo loaches. Here, we have improved the RAG1-RFLP marker using two restriction enzymes, PvuII and StuI. The improved marker allowed quick and accurate identification of each group (A, B1, and B2), clonal lineage, and even hybrid genotypes (A × B2 or B1 × B2).

Experimental animals
In total, 105 dojo loach (M. anguillicaudatus) individuals were collected from 12 localities in Japan (Table 1; Fig. S1). Although most individuals had been grouped (except those from Nanae and Abashiri) by genetic analyses of mtDNA-CR RFLP haplotypes, ManDra-B, ManBgl, and RAG1-RFLP genotypes in previous studies (Morishima et al. 2008;Yamada et al. 2015;Fujimoto et al. 2017), mtDNA-CR RFLP haplotypes, ManDra-A, ManDra-B, and the improved RAG1-RFLP marker were analyzed for all samples in this study (Table 2). Table 2 shows the previous DNA datasets ( Morishima et al. 2008;Fujimoto et al. 2017), as well as the results of the new DNA analyses in this study.

Group identification by mtDNA-CR RFLP haplotypes
Genomic DNA was extracted from tissue samples using a standard phenol/chloroform protocol (Asahida et al. 1996). The mtDNA-CR was amplified by PCR using a previously published primer set (0F 5′-CTG ACA TTC CGA CCA ATC   MtDNA genome was determined using control region RFLP haplotypes (CR-RFLP). Nuclear genome was comprehensively determined from results of ManDra-A genotype, ManDra-B genotype, and improved RAG1-RFLP Dashed boxes indicate previous results obtained from the same individuals including the RAG1-RFLP markers (standard methodology, not improved) described in Morishima et al. (2008) andFujimoto et al. (2017) initial denaturation for 3 min at 93 °C, followed by 30 cycles of denaturation for 1 min at 93 °C, annealing for 1 min at 58 °C, and extension for 1 min at 72 °C. The PCR products were digested using the restriction enzymes HaeIII and HinfI (Takara Bio, Shiga, Japan) (Morishima et al. 2008). Five microliters of each PCR product was mixed with 1.0 μL restriction enzyme, 1.0 μL 10 × M Buffer (for HaeIII) or 10 × H Buffer (for HinfI) (Takara Bio), and 3.0 μL double distilled water in a 0.2 mL microcentrifuge tube. After incubation at 37 °C for 9 h, 5.0 μL of the digested sample was electrophoresed on a 1.5% agarose gel for 40 min at 100 V and visualized with ethidium bromide. According to the method described by Morishima et al. (2008), the genetic group of each sample was determined by the RFLP haplotype of the mtDNA-CR.

Group identification by nuclear DNA markers ManDra-A and ManDra-B
The repetitive sequences, ManDra-A and ManDra-B, were amplified by PCR using previously reported primer sets  RAG1-RFLP marker (Fujimoto et al. 2017) was developed using RAG1 gene sequences (527 bp) of groups A (AB698051-AB698056), B1 (AB698049-AB698050, AB698057-AB698060), and B2 (AB698061-AB698064) determined by Yamada et al. (2015) (Fig. S2). To clarify improved RAG1-RFLP marker can identify each group accurately compared to RAG1-RFLP marker, same RAG1 gene sequence datasets (Yamada et al. 2015) were used as representative sequence of each group in this study. Optimal restriction enzymes that allow identification of each group (A, B1, B2, and clonal loaches) from the sizes and numbers of the digested fragments were selected using CLC Genomics Workbench (ver. 9.5.3) (QIAGEN, Venlo, Netherlands).

Improvement of RAG1-RFLP marker analysis
Primer set RAG1-M.aF (5′-GTT TGA ATG GCA GCC AGC TCTG-3′) and RAG1-M.aR (5′-CCA CAA ACA TGA GAC ACA GAG GTC -3′) was designed to amplify 443 bp of the RAG1 gene region (Fig. S2) (Fujimoto et al. 2017). PCR analyses were performed with 1.0 μL of genomic template DNA (100 ng/μL), 3.6 μL double distilled water, 5.0 μL 2 × Quick Taq HS DyeMix, and 0.2 μL of each primer (10 μM). The PCR cycling conditions were as follows: initial denaturation for 2 min at 94 °C, followed by 35 cycles of denaturation for 1 min at 94 °C, annealing for 1 min at 60 °C, extension for 1 min at 68 °C, and a final extension for 7 min at 68 °C. Five microliters of each PCR product was mixed with 1.0 μL each of restriction enzymes PvuII and StuI (New England Biolabs, Massachusetts, USA), 1.0 μL CutSmart Buffer (New England Biolabs), and 2.0 μL double distilled water in a 0.2 mL microcentrifuge tube. After incubation at 37 °C for 9 h, 5.0 μL of the digested sample was electrophoresed on a 1.5% agarose gel for 40 min at 100 V and visualized with ethidium bromide.

Results and discussion
The mtDNA haplotypes and nuclear genotypes from the 12 localities are shown in Table 2. The combined use of two restriction enzymes (PvuII and StuI) allowed the identification of each group (A, B1, and B2) from the sizes and numbers of the digested fragments of the amplified RAG1 gene region (443 bp) (Fig. S2). The sequences of group A (AB698051-AB698056) contained a restriction site for StuI (AB698051 was shown in Fig. S2). The sequences of group B1 (AB698049-AB698050 and AB698057-AB698060) contained a restriction site for StuI and a restriction site for PvuII (AB698049 was shown in Fig. S2). There were no restriction sites for StuI and PvuII in the group B2 sequences (AB698061-AB698064) (AB698061 was shown in Fig. S2). Thus, three diversified groups (A, B1, and B2) and the clonal loaches showed different electrophoretic fragment patterns using the improved RAG1-RFLP (Fig. 1). Specifically, two fragments (245 and 198 bp), three fragments (198, 147, and 98 bp), and a single fragment (443 bp) were detected in groups A, B1, and B2, respectively (Fig. 1). The clonal loaches exhibited four fragments (245,198,147, and 98 bp) derived from both groups A and B1 (Fig. 1). Thus, the improved RAG1-RFLP markers clearly distinguished the genetic groups in the dojo loaches. Group B2 is known as non-native loaches artificially introduced from Chinese continent (Koizumi et al. 2009;Shimizu and Takagi 2010;Matsui and Nakajima 2020). Thus, the improved RAG1-RFLP marker should be a strong tool for discriminating Japanese native populations and invasive populations.

3
Moreover, the DNA markers allowed the detection of various natural hybrid genotypes. For example, A × B1 hybrid genotype was found sympatrically with group B1 loaches in Obama in Fukui Prefecture (Table 2). Two different hybrid genotypes (A × B2 and B1 × B2) were found sympatrically, as well as groups A, B1, and B2 loaches in Futtsu in Chiba Prefecture ( Table 2). The clonal loaches had specific mtDNA-CR haplotype III, which has been classified into four lineages by random amplified polymorphic DNA (RAPD)-PCR and DNA fingerprints (Morishima et al. 2008). Individuals with hybrid genotypes in Obama and Futtsu may carry out clonal reproduction, although their mtDNA-CR haplotypes V and VII differs from that of the clonal loaches. This occurs because clonal reproduction is closely associated with hybridization in many species and is observed in clonal loaches that are supposed to be of hybrid origin between groups A and B1 (Dawley, 1989;Vrijenhoek, 1994;Beukeboom and Vrijenhoek, 1998;Lamatsch and Stöck, 2009;Arai and Fujimoto, 2013). Thus, experiments using artificial fertilization should be performed in the future to confirm whether unreduced diploid gametes are produced.