The biogeography and genetic diversity of A. cerana has extensively been studied for its distributional range in the temperate and tropical regions of Asia (De la Rúa et al., 2000; Deowanish et al., 1996; Sihanuntavong et al., 1999; Smith and Hagen, 1996; Smith et al., 2000; Smith et al., 2004; Takahashi et al., 2007; Tan et al., 2007). These studies exclusively used an internal spacer region located between mitochondrial tRNA^Leu and COII to analyze this species’ genetic diversity (Crozier and Crozier, 1993; Cornuet et al., 1991).

Nuclear loci require four times more the effective population size than mitochondrial loci (Brown, 1983). Thus, several more folds of nucleotides are required from nuclear DNA to obtain equivalent numbers of variable sites to mitochondrial loci in general (Zink and Barrowclough, 2008). Because of this condition, population genetics studies of A. cerana have focused on mitochondrial DNA. In fact, only a limited number of studies have utilized nuclear locioriginating markers, such as the ribosomal internal transcribed spacer 2 (ITS2) (Kim et al., 2010). Nevertheless, development of nuclear loci might be required to achieve a balance of information to expand our understanding of the population genetics aspects of A. cerana. In particular, recent studies have shown that intron sequences have a distinct rate of evolution, making them potentially useful in resolving relationships over a wide range of taxonomic levels (Yu et al., 2011). However, this method of study requires additional experimental efforts, such as isolation of alleles, PCR optimization, and alignment expertise due to insertion/deletion (indel) (Sang, 2002).

In this study, we reviewed the genome database to find introns of cDNA sequences for Apis species, including A. cerana. Initially, six introns were selected and primers were designed, and subsequently two introns, each having proper levels of diversity, were selected. These two introns were used to assess the potential of population genetic markers for A. cerana using 552 clones from 184 individuals of A. cerana collected from 16 localities in four countries including Korea.

A total of 184 A. cerana workers collected from 16 localities in four countries (South Korea, China, Vietnam, and Thailand) were used in this study (Table 1). The field-collected bees were preserved in 95% ethanol for molecular experiments.

[Table 1.] A list of trapping localities, sample numbers and GenBank accession numbers

A list of trapping localities, sample numbers and GenBank accession numbers

Total DNA was extracted from one or two legs using the Wizard Genomic DNA Purification Kit, in accordance with the manufacturer’s instructions (Promega, Madison, WI, USA). While reviewing intron sequences of Apis, six cDNAs, which provide the exon–intron structure, were found in the GenBank: the White gene (Kawakita et al., 2008), MRJP9 gene (Peiren et al., 2008), OR2 gene (Unpublished, GenBank accession number FJ666105), 1 inosital 1,4,5-triphosphate receptor gene (Lo et al., 2010), long-wavelength rhodopsin (LWRh) gene (Kawakita et al., 2008), and mitotic checkpoint control protein (Bub3) gene (Kawakita et al., 2008). These cDNAs included one to three introns and two to four exons. Among each of these, one intron having a relatively longer size and/or a proper site for primer design was selected from each gene. Three or four primers from each intron were designed and tested for proper utility (e.g., amplification success). Detailed primer information and origin are provided in Table 2.

[Table 2.] List of primer sequences used to amplify and sequence the introns of Apis cerana

List of primer sequences used to amplify and sequence the introns of Apis cerana

A 35-cycle amplification (94℃ for 1 min, 50–52℃ for 1 min, and 72℃ for 1 min) process for PCR was conducted after initial denaturation at 94℃ for 5 min and the final extension step continuing for 7 min at 72℃. To confirm successful DNA amplification, electrophoresis was carried out using 0.5× TAE buffer on 0.5% agarose gel. After purification with the PCR purification Kit (Qiagen, Germany), the amplicons were cloned into a pGEM-T Easy vector (Promega, USA). For the cloning process, XL1-Blue competent cells (Stratagene, USA) were transformed with the ligated DNA, and the resultant plasmid DNA was isolated using a Wizard Plus SV Minipreps DNA Purification System (Promega, USA). Two or three clones with limited sampling were sequenced for preliminary experiments and the obtained sequences were analyzed for their variability. Eventually, the White gene and the MRJP9 gene introns were selected and these introns were subjected to full individual screening, analyzing three clones per individual. DNA sequencing was conducted using the ABI PRISM^® BigDye^® Terminator ver. 3.1 Cycle Sequencing Kit with an ABI 3100 Genetic Analyzer (PE Applied Biosystems, USA). All products were sequenced from both strands.

Sequence delimitation was conducted by comparing newly acquired sequences to the GenBank-registered corresponding cDNA using MAFFT ver. 6 (Table 2; Katoh et al., 2002). When one or more nucleotide base or insertion/deletion (indel) position differed to the obtained sequences, the sequences were considered different sequence types.

In order to estimate the variability of the intron sequences, population diversity estimates, such as sequence diversity and nucleotide diversity, both of which are reflective of genetic diversity within a population, were determined using Arlequin ver. 3.5 (Excoffier and Lischer, 2010). The maximum sequence divergence within each locality was estimated via extraction of the within-locality estimates of unrooted pairwise distances from PAUP ver. 4.0b (Swofford, 2002).

As a preliminary experiment, sequencing was performed only for limited individuals: 53 clones from the intron of the White gene, 18 from MRJP9 gene, 24 from OR2 gene, 31 from 1 inosital 1,4,5-triphosphate receptor gene, 27 from LWRh gene, and 29 from Bub3 gene. Six introns provided multiple numbers of sequence types (8–20), but the introns of the Bub3 gene and the LWRh gene provided one and two sequence types, respectively (Table 3). When diversity of sequence types was considered, the intron of one inosital 1,4,5-triphosphate receptor gene showed the lowest as eight (25.8%), excluding the introns of Bub3 gene and LWRh gene. Furthermore, this intron was shorter in original length at 92–94 bp and this may be problematic, considering the potential of variable sites in a given length. The intron of OR2 gene provided the second highest diversity of sequence type (45.83%) and the highest sequence divergence (0.79–7.14%), but the original sequence length of the intron is only the third (118–126 bp), after the introns of the White gene and the MRJP9 gene (158–160 bp and 310–313 bp, respectively). Thus, these were excluded from subsequent full individual screening, and finally, the introns of the White gene and the MRJP9 gene were selected for full investigation to evaluate their potential as nuclear population genetic markers.

[Table 3.] Preliminary information of intron sequenced

Preliminary information of intron sequenced

The sequence analyses of 552 clones from 184 individual bees provided a total of 222 sequence types, ranging in size from 157 to 163 bp for the White gene intron and 141 sequence types, ranging in size from 307 to 357 bp for the MRJP9 intron (Table 4). The G/C content was ~27.8% (72.2% in A/T content) in the White gene intron and 19.5% (80.5% in A/T content) in the MRJP9 intron, indicating that the two introns are highly biased for A/T nucleotides. The sequence divergence of the two introns was 0.6%–7.9% in the White gene intron and 0.26%–17.6% in the MRJP9 intron, indicating moderate to substantial divergence. Previously, Lee et al. (2015) reported the sequence divergence of two mitochondrial non-coding sequences (NC1 and NC2) from the same individual bees as 3.093% (three positions) to 1.031% (one position) and 2.597% to 0.433%, respectively. The NC2 has been proven to be useful for genetic diversity and biogeographic study for worldwide A. cerana populations (Smith and Hagen, 1996; Smith et al., 2000; Smith et al., 2004; Warrit et al., 2006; Takahashi et al., 2007). The NC1 was newly developed in Lee et al. (2015) and also showed near equal variability to NC2. Further, these non-coding sequences each provided ten in NC2 and nine haplotypes in NC1 from the 184 individual bees (Lee et al., 2015). Compared to the above result, in this study, the sequence divergence of the two introns are much larger in terms of number of sequence type and sequence divergence. Thus, these intron sequences might be useful as non-mitochondrial DNA-based molecular markers for future population studies.

[Table 4.] Characteristic of intron sequences

Characteristic of intron sequences

In order to understand the population genetic diversity of A. cerana in populations, several diversity estimations were made (Tables 5 and 6). In the case of the White gene intron, the nucleotide diversity (π) ranged from 0.003117 (locality 3, Hongcheon-gun) to 0.025837 (locality 15, Vin Phuc, Vietnam), showing at least an eight-fold difference between them (Table 6). For the MRJP9 intron, diversity ranged from 0.016541 (locality, 4, Namwon-si) to 0.052468 (locality 15, Vin Phuc, Vietnam), showing at least a three-fold difference (Table 6). Thus, the MRJP9 intron showed two to five fold greater diversity than that of the White gene intron.

[Table 5.] Within-locality diversity estimates of Apis cerana from White gene intron

Within-locality diversity estimates of Apis cerana from White gene intron

[Table 6.] Within-locality diversity estimates of Apis cerana from MRJP9 gene intron

Within-locality diversity estimates of Apis cerana from MRJP9 gene intron

Among the 16 populations, the highest π was found in Vin Phuc, Vietnam (locality 15) in both intron sequences. The second highest estimate were found in Chanthaburi population, Thailand (locality 16) as 0.010435 in the White gene intron and Beijing, China (locality 14) as 0.027619 in the MRJP9 intron. On the other hand, the nine Korean populations ranged from 0.003117 to 0.007170 and 0.015984 to 0.022876 in the White gene intron and MRJP9 gene intron, respectively. This indicates that the Korean populations are relatively low in diversity of the two intron sequences, although their locality of seven (Cheongyanggun) ranked the fifth in the MRJP9 intron and eighth in the White gene intron. In fact, Lee et al. (2015), based on NC2, has also shown that the India, Indonesia, Malaysia, Philippines, and Burma populations were several-fold higher in genetic diversity than the Korean populations, indicating the presence of small effective populations in Korea. Similar to π estimates, haplotype diversity (H) also was the highest in Vin Phuc, Vietnam (locality 15) for both intron sequences. The H of the nine Korean populations ranged from 0.4190 to 0.7059 in the White gene intron and 0.7905 to 0.9673 in the MRJP9 gene intron. These estimates are not high; however, they are better than those of π found in several Korean localities showing more improved ranks in the H than those of π. Nevertheless, it is obvious that the Korean populations have an overall low diversity.

In summary, we selected two intron sequences from the public genome database and sequenced 552 clones from 184 individuals of A. cerana collected from 16 localities in four countries. The high variability and diversity of the intron sequences show their high potential for use as non-mitochondrial DNA-based molecular markers for future population genetics studies.