Background

Bagridae is an important family of catfishes which belongs to Order Siluriformes and consists of more than 220 species of 21 genera (Ferraris 2007). It is commonly found in Africa and Asia and has abundant species diversity. In China, the family includes approximately 30 species of 4 genera which inhabit in Yangtze River, Pear River, Heilongjiang River and Yellow River (Qin et al. 2010). In the past decades, their populations have decreased rapidly and almost disappeared in many river systems due to over fishing, environmental pollution and other human disturbances (Luo et al. 2000; Wang et al. 2006; Xiao et al. 2014). In the meantime, an increasing number of Bagridae fishes have been cultured due to a high market demand. And to improve the growth performance and survival rate, some crossbreeding programs have been conducted, such as Pseudobagrus ussuriensis × Pelteobagrus vachelli (Cai et al. 2011; Qin et al. 2012; Wang et al. 2013), Tachysurus fulvidraco × Pelteobagrus vachelli (Wang et al. 2012; Zhang et al. 2016), Tachysurus fulvidraco × Leiocassis crassilabrus, and Pelteobagrus vachelli × Leiocassis crassilabrus (Wang 2013). The taxonomy of Bagridae is confusing and the validity of some catfishes should be further investigated and identified, what is more, taxonomic identification of hybrids is even more difficult. The main intention of crossbreeding is to obtain commercial advantages. Both yellow catfish (Tachysurus fulvidraco) and Chinese longsnout catfish (Leiocassis longirostris) in Bagridae have become important economic freshwater fishes with great market demand and high price in China due to their tender flesh, rich nutrition, few bones, convenient cooking and good taste (Liang et al. 2012; Xiao et al. 2014; Shen et al. 2014). To further improve the growth traits, they have also been used to produce hybrids.

Although crossbreeding could improve the genetic traits of animals, releases and escapes of hybrid individuals from fish farms have potential environment impacts. Hybrids can lead to the habitat changes for some fish species and change species composition in wild populations due to the change of diet or survival competition. Moreover, if the hybrids are fertile, the harmful effects of hybridization may cause pollution of gene pool and the extinction of some populations and species (Alllendorf et al. 2001). Therefore, accurate identification of fish hybrids is critical to the sustainable aquaculture development. Reliable identification methods of hybrids can not only serve for the effective management of crossbreeding practices but also are used to monitor their negative impacts. Hybrid authentication is usually difficult and uncertain by morphological methods. However, molecular identification method is an ideal and alternative method to the traditional morphological discrimination in hybrid identification (Scribner et al. 2000). In the present study, we will try to establish an effective identification method of T. fulvidraco, L. longirostris and their hybrids using the polymerase chain reaction (PCR) based on mitochondrial and nuclear molecular marker.

Methods

Experimental samples

In this study, 30 T. fulvidraco individuals and 30 L. longirostris individuals were used to analyze parental lineages. For the interspecific hybrid, 30 specimens of FTL (crosses using T. fulvidraco as female and L. longirostris as male) and FLT (crosses using L. longirostris as female and T. fulvidraco as male) were genetically analyzed. All samples were obtained from the experimental farm of Yangtze River Fisheries Research Institute, Chinese Academy of Fishery Sciences (YFI). The fish specimens were stored in the fish museum of YFI. All experimental procedures for the target fish were carried out in accordance to the standards of the Animal Care Policy of YFI.

DNA extraction, PCR amplification and sequencing

Total genomic DNA was extracted from muscle tissue using the traditional phenol–chloroform extraction method (Taggart et al. 1992). DNA quality was determined by electrophoresis in a 1% agarose gel. Cytochrome C Oxidase subunit I (COI) sequences were obtained by the described primer pairs in Table 1 which were designed from the COI sequences of T. fulvidraco (GenBank accession no. HM641815) and L. longirostris (GenBank accession no. NC014586). For identifying the hybrid, the nuclear ITS sequences of T. fulvidraco (10 individuals) and L. longirostris (10 individuals) were firstly obtained by the published primer pairs ITSF/ITSR in Table 1 (Yang et al. 2010). Then the differentiation of two catfish species was analyzed and the primer pairs ITSPF/ITSPR were designed to amplify the specific regions (ITSP) of different sequences length based on the obtained ITS sequences.

Table 1 Information of primers pairs and sizes of the PCR products
Full size table

DNA amplifications were carried out by the polymerase chain reaction (PCR) in a total reaction volume of 25 μL. Each 25 μL PCR reaction system for COI and ITS contained 1 μL of 10 mM each primer, 1 U Taq DNA polymerase (TaKaRa, Japan), 2.5 μL 10 × PCR buffer (100 mM Tris–HCl, 500 mM KCl, 15 mM MgCl2; TaKaRa), 2 μL 10 mM dNTP (TaKaRa, Japan) and about 50 ng genomic DNA template. The PCR reaction for ITSP contained 1 μL of 10 mM each primer, 1.25 U LA Taq (TaKaRa, Japan), 12.5 μL 2 × GC bufferI (5 mM MgCl2, TaKaRa), 4 μL 10 mM dNTP (TaKaRa, Japan) and about 50 ng genomic DNA template. All PCR reactions were performed on S1000™ Thermal Cycler (BIO-RAD, USA) based on the different conditions. The procedures for COI: pre-denaturing at 94 °C for 5 min; 35 cycles of denaturing at 94 °C for 30 s, annealing at 52 °C for 30 s, and extending at 72 °C for 2 min; and a final extension at 72 °C for 10 min. The procedures for ITS: pre-denaturing at 94 °C for 5 min; 35 cycles of denaturing at 94 °C for 30 s, annealing at 54 °C for 30 s, and extending at 72 °C for 45 s; and a final extension at 72 °C for 10 min. The procedures for ITSP: pre-denaturing at 94 °C for 1 min; 30 cycles of denaturing at 94 °C for 30 s, annealing at 60 °C for 30 s, and extending at 72 °C for 2 min; and a final extension at 72 °C for 5 min.

The PCR products (COI and ITS) were run on 2.0% agarose gels for 1 h at 80 V and then purified by a DNA Agarose Gel Extraction Kit (Axygen, USA). The purified PCR products were cloned into pMD 18-T vector and sequenced by ABI 3730 automated sequencer (Applied Biosystems, USA).

Data analysis

The obtained COI and ITS sequences were edited and aligned using the Clustal W (Hall 1999). A neighbor-joining (NJ) tree was constructed based on the COI gene sequences of all individuals from each group using MEGA Version 6 (Tamura et al. 2013). Silurus meridionalis was used as outgroup. The uncorrected p-distances model was carried out and node support was assessed based on 1000 bootstrap replicate. The different regions of ITS sequences in length were identified using MEGA Version 6 (Tamura et al. 2013). The primer pairs ITSPF/ITSPR (Table 1) were designed to amplify the specific ITS sites of T. fulvidraco and L. longirostris.

DNA fragment size analysis

DNA fragment sizes of ITSP were determined by electrophoresis on 2% agarose gels with ethidium bromide (1 ng/mL) for 1 h at 80 V. The agarose gels were observed and captured by Gel document system (Clinx, China).

Results

To identify the maternal parentage, all individuals were amplified using the primers COIF/COIR for gene COI. For the COI sequences, the fragment sizes were approximately 1500 bp and there was no obvious difference in length among T. fulvidraco, L. longirostris, FTL and FLT (Table 1). Sequence analysis indicated that all individuals were divided into two groups, with one group including all the individuals of T. fulvidraco and hybrid FTL, and the other containing the rest L. longirostris and hybrid FLT. Similarly, the results showed that all individuals were divided into two clades as the former sequences analysis from phylogeny tree analyses (Fig. 1). The hybrid FTL and T. fulvidraco were grouped into one clade, and the hybrid FLT and L. longirostris into the other clade. These results suggested that all individuals came from two different maternal parentages, T. fulvidraco and L. longirostris, due to the maternal inheritance characteristics of mitochondrial. However, paternal line could not be identified from the sequences of COI gene.

Fig. 1
figure 1

Phylogenetic tree based on COI sequences of Tachysurus fulvidraco, Leiocassis longirostris and their hybrids. The dendrogram was constructed based on the COI gene sequences by the neighbour-joining method. Silurus meridionalis was used as outgroup. FTL T. fulvidraco(♀) × L. longirostris(♂) FLT L. longirostris(♀) × T. fulvidraco(♂) 1–30 represents individual number of each population

Full size image

To detect the paternal line, ITSF/ITSR was firstly used to amplify the ITS sequences and to differentiate between T. fulvidraco and L. longirostris. By comparing the specific sites, some valuable regions were found. Then the primer pairs ITSPF/ITSPR were designed to obtain the amplification of the specific region for ITS (ITSP) based on their different sequences. Furthermore, all individuals were amplified and then detected by electrophoresis in 2% agarose gel. The results showed that there was only one band for T. fulvidraco individuals (about 670 bp) and L. longirostris individuals (about 600 bp) (Fig. 2), while a heterozygous pattern with two bands (600 and 670 bp) was observed for hybrids FTL and hybrids FLT, respectively. Together with the result of maternal line, the paternal line could thus be inferred (Fig. 2).

Fig. 2
figure 2

Electrophoresis patterns of the species-specific ITSP primers for ITS region of hybrids identification. 1–3 T. fulvidraco 4–6 FTL: T. fulvidraco(♀) × L. longirostris (♂) 7–9 L. longirostris 10–12 FLT: L. longirostris(♀) × T. fulvidraco(♂) M molecular marker. Two types of ITS sequence existed in hybrid individuals, one type of ITS is T. fulvidraco, and the other is L. longirostris

Full size image

In conclusion, a method based on sequence variation was established for taxonomic identification of hybrids. Firstly, we used the mitochondrial COI primers to obtain the sequences of the individuals and construct phylogeny tree to determine their maternal parentage by analyzing their homology and phylogenic relation. Then we used the nuclear marker ITSPF/ITSPR to detect the hybrids and their paternal line. The application of this mitochondrial COI and internal transcribed spacer ITS marker has been proved to be feasible to identify the hybrids.

Discussion

Hebert et al. (2003) proposed a molecular technique named DNA barcoding to identify species and COI gene has then been commonly used as a standardized marker. Though COI gene is widely used as a genetic marker for fish species authentication, it is not an appropriate marker for all aspects of species identification (Clark 2015). The COI gene sequences only could not determine a hybrid because mtDNA exhibits maternal inheritance. The COI gene of hybrid F1 showed identical characteristic with its maternal parentage and can not represent different alleles (Rubinoff et al. 2006). Thus, hybrid individuals could be misidentified as their maternal parent species based solely on COI gene. Other genetic markers must be further considered due to the potential problem for accurate identification based on single COI gene (Rubinoff et al. 2006).

Recently, some researches have been carried out to identify hybrids and detect the hybridization events based on the mitochondrial and nuclear molecular markers. Hashimoto et al. (2011) established the molecular appraised technology for Serrasalmid fish and their hybrids based on mitochondrial genes (COI and Cytb) and nuclear genes (RAG2). The molecular diagnostic method was also established for hybrids between Netropical catfish species Pseudoplatystoma corruscans and Pseudoplatystoma reticulatum by mitochondrial 16S and nuclear genes RAG2 (Prado et al. 2011). Dio et al. (2015) identified the hybridization of burrfish between Chilomycterus antillarum and Chilomycterus schoepfii using COI gene sequences and AFLP technology. The interspecies hybridization in the freshwater stingrays Potamotrygon motoro and P. falkneri was revealed by mitochondrial gene (COI and Cytb) and nuclear microsatellite markers (Cruz et al. 2015). The results show that the established identification methods can be rapidly implemented and effectively determine the hybrid individuals. The ITS region exhibited low variations within species and high variations between species (Yu et al. 2006). What is more, it has several advantages due to its rapid evolution, easy isolation and non-coding structure (Chow et al. 2009). It has been widely used in the molecular identification and systematics studies for species discrimination (Chow et al. 2009; Yang et al. 2010; Zhu et al. 2011). As a result of hybridization, the nuclear ITS locus is a heterozygote consisted of heterologous alleles, one of which comes from maternal line and the other from paternal line for diploid individuals. Hence, it could be used to successfully identify hybrids by the heterozygous pattern in this study.

Primer design must be considered as a robust diagnostic technology. First, the conservative and consistent sites were used to design the primer pairs to simultaneously achieve the specific ITSP region sequences of T. fulvidraco individuals, L. longirostris and their hybrids. Second, there were different length PCR products by amplification of designed species-specific primers, and the species-diagnostic bands could be easily distinguished (Pank et al. 2001). In this study, the amplification sequences gained for T. fulvidraco individuals and L. longirostris individuals by primers ITSPF/ITSPR were approximately 670 and 600 bp, respectively, which could be easily and directly observed from the agarose gel. Another consideration factor was to ensure the high efficiency of the amplification reaction (Pank et al. 2001). In our study, we increased the amplification success rates and simultaneously obtained two different-sized products by utilization of LA Taq and GC buffer.

Establishing identification technology for hybrids is vital to monitor hybridization programs and manage crossbreeding progress (Alllendorf et al. 2001). Meanwhile, hybridization programs should be continuously monitored to assess their impacts and safeguard wild populations since it is almost impossible to recover the population characteristics if the wild populations were suffered from genetic contamination. In this study, molecular diagnostic method for the target fish species was successfully established, and it could be applied for identifying not only eggs, larvae and young individuals during the breeding procedure but also fish meat and fish products. It could also become a reference for establishing detection technique for hybrids in other groups of fishes.

Conclusions

The molecular identification methods for T. fulvidraco, L. Longirostris and their hybrids were successfully established. Meanwhile, these molecular diagnostic tools could also be used routinely to assess breeding plans of fish farms, better manage fish hybrids, as well as to monitor and minimize the negative impacts resulting from the implementation of hybridization projects in aquaculture industry.