INTRODUCTION
Dryophytes japonicus is widely distributed across North- East Asia, including South Korea, whereas Dryophytes suweonensis is known to inhabit only the western parts of the Korean Peninsula based on the Taebaek mountain range (Do et al. 2017). Between the two tree frog species, D. suweonensis is an indigenous species of South Korea, and its population size has decreased owing to a continuous loss of habitats. The species has been designated as a Class I Endangered Wildlife Species in South Korea and a IUCN Red List Endangered (EN) Species (IUCN 2017) due to various causes such as hybridization. Meanwhile, D. suweonensis and its close relative D. japonicus are sympatric species sharing the same habitats (Roh et al. 2014). D. suweonensis has been reported to differ from D. japonicus in terms of its call and facial shape, but it is difficult to visually discriminate the two species based solely on morphological differences. Notably, in the case of female frogs that do not call, accurate species identification is extremely difficult (Kuramoto 1980;Borzée et al. 2013). Therefore, the first step in a series of studies toward the conservation and restoration of D. suweonensis is accurate species identification.
Previous studies on the genetic species identification of land animals including amphibians have, to date, applied a variety of methods including DNA sequence analysis and comparisons, restriction fragment length polymorphisms (RFLP), amplified fragment length polymorphisms (AFLP), random amplified polymorphic DNA (RAPD), single-strand conformation polymorphisms (SSCP), and species-specific PCR analysis (Bertolini et al. 2015;Kanthaswamy 2015). Among these methods, DNA sequence analysis and comparisons were often involved with regions of mitochondrial 12S and 16S ribosomal DNA (rDNA), D-loops, cytochrome b (CYB), and cytochrome c oxidase I (COI), as well as nuclear 28S rDNA, in studies investigating inter- or intra-species variation (Kocher et al. 1989;Bataille et al. 1999;Bellis et al. 2003;Girish et al. 2005;Berry and Sarre 2007;Imaizumi et al. 2007;Tobe and Linacre 2008;Robertson et al. 2009;Fitzcharles 2012). Compared to nuclear DNA, mitochondrial DNA has a higher copy number to allow for easier analysis; this is associated with the advantage of the accumulation of a large number of genetic mutations between species or individuals owing to a relatively high rate of mutation. The species identification of D. suweonensis versus D. japonicus is mostly conducted via the sequencing of mitochondrial DNA (Lee and Park 1992;Lee et al. 1999;Borzée et al. 2017;Lee et al. 2017), which requires considerable time and cost. High resolution melting (HRM), however, allows for the immediate detection and analysis of the differences in dissociation temperatures of the double helix DNA chains, as steadily increasing temperatures are applied to the amplified PCR products. This leads to the rapid and accurate characterization of genotypes of different species with varying sequence data, without the PCR product purification and DNA sequencing steps (Reed et al. 2007).
There are also non-invasive DNA sampling methods using environmental DNA (eDNA), such as feces, urine, or mucus, without direct capture of the target species (Goldberg et al. 2015;Sigsgaard et al. 2015). Recent studies have shown that the eDNA, found as DNA fragments of organisms in various environments, such as soil and water, is highly useful in species identification (Thomsen and Willerslev 2015). This method generally involves the collection of samples from water or soil to detect a rare or endangered species. It has been actively applied to numerous studies worldwide (Jerde et al. 2011;Thomsen et al. 2012;Pilliod et al. 2013;Cardás et al. 2020). HRM analysis using eDNA leads to more rapid, accurate, and cost-effective results than conventional methods (Martinou et al. 2010). In this study, a rapid and accurate method of HRM analysis was developed for species identification of the two tree frog species (D. japonicus and D. suweonensis) that inhabit South Korea, with verification of the potential use of this novel non-invasive method using eDNA for species identification.
MATERIALS AND METHODS
1. Sample collection and genomic DNA extraction
D. suweonensis and D. japonicus were captured from April to June, 2021, in the following regions (the number in parentheses indicates the number of captured animals): D. suweonensis (2) and D. japonicus (5) in Asan, Chungcheongnam- do; D. suweonensis (3) and D. japonicus (3) in Chungju, Chungcheongbuk-do; D. suweonensis (1) in Pyeongtaek, Gyeonggi-do; D. suweonensis (2) in Wanju, Jeollabuk-do. The direct capture of animals was performed at night when D. suweonensis was active. The animals that were either calling or mating were captured while the investigator walked along the rice paddies, the main habitat of the species. To extract the genomic DNA (gDNA), each frog was made to hold a sterile swab in its mouth for approximately 1 min to collect oral epithelial cells (Goldberg et al. 2003). From the sterile swabs with oral epithelial cells, the gDNA was extracted using the DNeasy Blood & Tissue Kit (QIAGEN, Germany) according to the manual. The quantity of the extracted gDNA was analyzed using a spectrophotometer (DeNovix DS-11 FX, DeNovix, USA).
2. Sample preparation for eDNA analysis
For non-invasive species identification using eDNA, three frogs each of D. japonicus captured in Jeollabuk-do and D. suweonensis captured in Chungcheongnam-do were used. First, plastic containers (6.5×6.5×20 cm) were rinsed in bleach to remove any residual DNA and washed clean with sterile tertiary distilled water. Each frog was washed with sterile tertiary distilled water and placed in a plastic container with 200 mL of tertiary distilled water. The control was tertiary distilled water without a frog. To prevent cross-contamination, the investigator wore a mask and lab gloves for replacing each frog. After 30 min of immersion in tertiary distilled water, 200 mL of the immersion water was collected from each container. The eDNA from the immersion water was filter-concentrated using a cellulose nitrate filter (Ф0.45 μm, Whatman, Germany). Next, 20 μL of proteinase K and 280 μL of ATL buffer of the DNeasy Blood & Tissue Kit (QIAGEN, Germany) were added, and the filter was completely destroyed by adding three stainless steel beads of 2.4 mm in diameter (OMNI International, Kennesaw GA, USA) and using the Bead Ruptor Elite (OMNI International, USA). The same volume of AL buffer was added, and the mixture was incubated at 56°C for 12 h. The subsequent steps were carried out as described in the manual, and eDNA was ultimately extracted.
3. PCR primer design and DNA sequencing
To design the PCR primers for mitochondrial 12S rDNA of D. japonicus and D. suweonensis, the data of congeneric species were downloaded from the GenBank database of the National Center for Biotechnology Information (NCBI, USA). The DNA sequence matrix was drawn using BioEdit 7.2.5 (https://www.mbio.ncsu.edu/bioedit/bioedit.html; Biological Sequence Alignment Editor) (Hall 1999), and the PCR primer combinations with high conservation were designed using Sequence Manipulation Suite ver. 2 (http://www.bioinformatics.org/). Here, their length was set to the 19-21 mer range, GC content was set to 36.84- 47.62%, and Tm was set to 56.8-65°C. To perform the PCR amplification of the mitochondrial 12S rDNA, the AccuPower® PCR PreMix (Bioneer, Korea) was used. The PCR reaction solution consisted of 10 ng of gDNA, 5 μM of forward primer (ANU-MT-00018f: 5′-AAAGCRTAG CACTGAAAATG-3′), 5 μM of reverse primer (ANUMT- 01017r: 5′-TCGGTGTAAGCGAGATGCTTT-3′), and sterile distilled water to make up a final volume of 20 μL. Using the ProFlex PCR System (Life Technologies Corporation, Carlsbad, CA, USA), the PCR amplification comprised an initial 1 min of initial denaturation at 95°C and 35 cycles that consisted of 20 s denaturation at 95°C, 20 s binding at 55°C, and 1 min extension at 72°C, with a final extension for 5 min at 72°C. Here, hot-start PCR was performed, wherein the PCR tube that contained all PCR reactants was placed in the PCR device to start the amplification reaction when the temperature (95°C) of the initial denaturation was reached. The resulting PCR products were purified using the AccuPrep® PCR Purification Kit (Bioneer, Korea) according to the manual. The BigDyeTM Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Waltham, MA, USA) and DNA Analyzer 3730xl (Thermo Fisher Scientific) were used for DNA sequence decoding. Here, the primers used in the cycle sequencing were the same as the forward and reverse primers in the PCR reaction. After error correction using BioEdit 7.2.5, the unnecessary parts were adequately cut to complete the contig, and finally, the DNA sequence was decoded.
4. HRM analysis
The mitochondrial 12S rDNA of D. japonicus and D. suweonensis was analyzed to explore the regions that exhibit an adequate level of interspecies genetic variation. A new set of PCR primers was prepared for the flanking regions of such sites (Fig. 2). For the HRM analysis, the Melt- DoctorTM HRM Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) was used according to the manual; 10 μL of MeltDoctorTM HRM Master Mix, 10 ng of gDNA or 2 μL of eDNA, 5 μM of forward primer (HYL-12S-0250f: 5′-GTTACACCACGAGGCTCA-3′), 5 μM of reverse primer (HYL-12S-0343r: 5′-TGAGTTTCTTAAGAA CAAGCG-3′), and 6 μL of sterile distilled water comprised the PCR reactant solution with a 20 μL final volume. Using the QuantStudio 5 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA), the PCR reaction began with an initial 10 min denaturation at 95°C, followed by 40 cycles that consisted of 15 s denaturation at 95°C and 1 min binding/extension at 60°C. The HRM analysis was performed through 10 s denaturation at 95°C and 1 min binding at 60°C for the melt curve and dissociation steps, and through 15 s denaturation at 95°C and 15 s binding at 60°C for the HRM.
RESULTS
Through the novel design of a PCR reaction in this study, to amplify the mitochondrial 12S rDNA region of D. japonicus and D. suweonensis using PCR primers, a product with the predicted size was detected for all 16 frogs. Subsequent DNA sequencing identified 915 bp in D. suweonensis and 916-917 bp in D. japonicus. A comparison of the two sequences allowed for the discrimination of the two species based on an indel (insertion or deletion) mutation of 2 bp, and a species-specific insertion or deletion was found in one of eight D. japonicus frogs. Between the two species, 46 bp of mutations were identified in total; of these, 45 bp comprised a region of interspecific variation between D. suweonensis and D. japonicus and 1 bp was a region of intraspecific variation for D. suweonensis. Based on this, the intraspecific haplotype was analyzed, and D. suweonensis and D. japonicus each showed three haplotypes, of which sequences were registered in the GenBank database (Gen- Bank No. OK156170-OK156175).
The HRM analysis in this study using the novel primer design showed that PCR amplification of the gDNA of all frogs was successful. The melting peak of the D. suweonensis sample was 76.4±0.06°C, whereas that of the D. japonicus sample was 75.0±0.05°C, suggestive of a mean interspecies variation of 1.31°C. In addition, the aligned melt curve and difference plot melt curve led to the easy discrimination of the two species (Fig. 3), in agreement with the DNA sequence result. Thus, the PCR primer design for HRM analysis produced in this study based on the mitochondrial 12S rDNA was shown to be highly useful for the differentiation of D. japonicus and D. suweonensis.
Moreover, HRM analysis using eDNA obtained from the immersion experiment showed a difference of 1.10- 1.31°C in the melting curve analysis for D. japonicus and D. suweonensis. For distilled water, used as a negative control, melting peaks were not observed. In other words, the melting peaks of D. suweonensis were all identical at 76.3°C, whereas those of D. japonicus were 75.0±0.14°C (Fig. 4). This result thus coincided perfectly with the HRM analysis using gDNA.
DISCUSSION
In previous studies, D. suweonensis and D. japonicus were found to have 12.90-15.66% variation in the mitochondrial CYB region (Yang et al. 1997; Lee et al. 1999), whereas the mitochondrial 12S rDNA region showed 7.17% genetic distance based on 68 bp point mutations of 938 bp (Lee et al. 1999). Moreover, the region comprises a highly conserved region (Arif and Khan 2009). In this study, the two species were found to have 4.91% variation, which deviated from the report of Lee et al. (1999), but the resolution was determined to be adequate for differentiating the maternal properties of the two species. In performing HRM analysis, the use of a PCR primer combination that includes a greater area of differential regions could increase the efficiency of species identification (Perini et al. 2020). In this study, the PCR primer design contained 7 bp interspecies point mutation regions from 246 to 359 bp, indicating a 6.19% genetic point mutation rate. The selected region thus exhibited a higher mean interspecies point mutation rate than the 12S rDNA region.
DNA sequencing requires considerable time and cost as it involves a series of steps, including the PCR reaction, PCR product purification, DNA sequence decoding, and sequence alignment (Garritano et al. 2009). In contrast, HRM analysis can be completed within approximately 2 h, allowing for rapid analysis, as the gDNA extraction and PCR amplification are immediately followed by the melting curve analysis (Erali et al. 2008;Garritano et al. 2009). In addition, HRM analysis is associated with a very low risk of cross-contamination to allow for accurate analysis, as the PCR amplification and melting curve analysis can be performed simultaneously within a single container (Erali et al. 2008;Garritano et al. 2009). The cost is also far lower than that of conventional DNA sequencing. Thus, the method of molecular identification developed in this study using the novel PCR primer designed for HRM analysis to discriminate between D. japonicus and D. suweonensis led to the same result as DNA sequencing but was faster, more accurate, and cost-effective.
Recently, various studies have been conducted regarding the use of eDNA to detect amphibian species (Bedwell et al. 2021). In general, the skin of amphibians is moist with granular glands that show the characteristic secretion of intestinal fluid, toxic granules, or mucus mixtures (Mills and Prum 1984;Toledo and Jared 1995;Clarke 1997). As amphibian species have this mucus-secreting skin, it is likely that the immersion of two different species in distilled water for only 30 min will allow for the secretion of eDNA at an adequate quantity for the HRM analysis.
Non-invasive species identification is widely applied in conservation genetics as there is no need to directly capture wildlife species. As it is especially useful to study small and rare species (Rees et al. 2014), related methods have been developed for different species (e.g. Eggert et al. 2001;Palomares et al. 2002). The non-invasive method of sample collection for anuran amphibians has mostly relied on the collection of oral epithelial cells (Pidancier et al. 2003;Broquet et al. 2007). However, this method is also not suitable in terms of minimizing the stress imparted on the animals, as their mouths must be forced open while the body of the animal is restrained in the hand of the investigator. The method in this study, however, involves a simple, 30 min immersion of the animal (D. japonicus and D. suweonensis) in distilled water to obtain the eDNA for subsequent molecular identification, which is thought to sufficiently reduce the stress that might arise during the handling of the animal. This non-invasive method is likely to minimize stress upon species identification, the most basic process in the conservation and management of endangered species.
The results in this study suggested that the HRM analysis using eDNA is a non-invasive method for the rapid and accurate identification of two morphologically similar tree frog species, D. japonicus and D. suweonensis. With the accumulation of melting curve profiles of sympatric species and their availability in databases, this method will be applicable for molecular identification in the future. Further studies should investigate the level of eDNA generation and its detection with respect to various factors such as immersion time and animal size.
ETHICAL APPROVAL
All applicable international, national, and/or institutional (NIEIACUC-2020-012) guidelines for the care and use of animals were followed.