Identification of ‘Haryejosaeng’ mandarin by multiplex SNP genotyping

Background Most the are difficult to improve through hybridization because of polyembryony and male sterility. Therefore, their improvement has mostly been based on the selection of nucellar embryo and bud mutation. These cultivars are supplied to breeders and farms at the seedling stage, which renders their identification based only on morphological traits. In addition, because these seedlings originate from nucellar embryo and bud mutation selection, they are genetically very similar. Therefore, the present study was carried out to develop markers that can specifically and rapidly distinguish ‘Haryejosaeng,’ which is generally supplied to breeders, from other satsuma mandarin cultivars that are planted on farms. Polymerase chain reaction (PCR) was performed to distinguish ‘Haryejosaeng’ from other 8 cultivars (‘Haryejosaeng’- breeder’s stock, ‘Miyagawa wase,’ ‘Okitsu wase,’ ‘Yura wase,’ ‘Miyamoto wase,’ ‘Ueno wase,’ ‘Yonezawa wase,’ and ‘Nichinan 1 gou’) using 6 single nucleotide polymorphism (SNP) markers specific for ‘Haryejosaeng’ and one SNP primer pair, which was used as the negative control. Using a multiplex PCR, SNP markers P1 (HL-SNP-SCAF_2-23997586-F and HL-SNP-SCAF_2-23997586-R), P2 (HL-SNP-SCAF_2-36059523-F and HL-SNP-SCAF_2-36059523-R), and P5 (HL-SNP-SCAF_9-30793978-F and HL-SNP-SCAF_9-30793978-R) simultaneously yielded 165, 150, and 526 bp amplicons, respectively, for Haryejosaeng only. The SNP markers were further validated by high-resolution melting analysis. The multiplex PCR based on P1/P5 and P2/P5

wase' and 'Okitsu wase' [1]. Nucellar seedlings (plants developed from nucellar embryos) of the highquality cultivar 'Haryejosaeng' [2], selected by crossing 'Tachima wase' with C. × natsudaidai 'Hayata' in the Citrus Research Institute, Jeju, Korea, are currently distributed to farmhouses. The fruits of this cultivar are bright orange-yellow and mature in early November; they weigh approximately 80-90 g and have a more compressed-oblate globose shape than 'Miyagawa wase' fruits (the main cultivar grown in Jeju). Furthermore, their rind has the characteristic thickness of mandarin fruits that allows easy peeling. The tree vigor of 'Haryejosaeng' is stronger, total soluble solids in its fruits are approximately 1 Brix higher, and acid content is approximately 0.1% lower than those of 'Miyagawa wase' [2].
It is generally difficult to improve satsuma mandarin cultivars through hybridization because of polyembryony and male sterility [3]. Accordingly, most of these cultivars have been improved based on nucellar embryos [one of the main features of seed reproduction in citrus cultivars; nucellar embryos are developed from nucellar cell tissues surrounding the embryo sac in citrus [4]] and selection of bud mutants with low citric acid and high sugar content (i.e., bud mutation selection). As such, these plants have the same genetic traits as the maternal line [5,6], and variants are difficult to identify based on morphological traits only. For instance, 'Okitsu wase' is an improved cultivar originating from the selection of nucellar seedlings of 'Miyagawa wase' [7], whereas 'Morita Unshiu' was selected from a bud mutation of 'Miyagawa wase' and 'Nichinan 1 gou' was selected from a bud mutation of 'Okitsu wase' [2,8,9]. As per the guidelines of the International Union for the Protection of New Varieties of Plants (UPOV) and Korea Seed and Variety Service (KSVS), it is necessary to develop technologies that can specifically distinguish 'Haryejosaeng' from other mandarin cultivars being cultivated domestically and abroad [10] to protect the intellectual property rights of breeders.
However, there are no reports on the use of these markers in cultivars that are difficult to distinguish morphologically and genetically, such as those of C. unshiu mandarin.
In this context, the present study was conducted to identify SNP markers that can discriminate between 'Haryejosaeng' and other cultivars, and to rapidly verify those markers using the HRM analysis. In addition, we developed a method based on multiplex PCR for rapid, cost-effective, accurate, and reproducible selection of 'Haryejosaeng.'

Selection of SNP markers
After trimming the raw data obtained from sequencing, the genome coverage for each of the cultivars was contrasted with the standard genome of C. × clementina 'Clemenules' (approximately 367 Mbp); it was 20.70 times for 'Haryejosaeng,' 17.92 times for 'Miyagawa wase,' 18.14 times for 'Miyamoto wase,' 16.14 times for 'Okitsu wase' and 17.66 times for 'Nichinan 1 gou' ( Table 1). The mapping ratio of the reads to the C. × clementina 'Clemenules' genome ranged from 90.45% to 91.44% (Table 1).
Among the 49,111 SNPs from the interspecific integrated SNP matrix based on the consensus sequence obtained by mapping the reads of 'Haryejosaeng' and 'Miyagawa wase' to the genome sequence of C. × clementina 'Clemenules,' 3,639 SNPs were selected for designing the primers (data not shown). In addition, 77 candidate SNPs were selected by comparing the sequences of 'Haryejosaeng' with those of 'Miyagawa wase,' 'Miyamoto wase,' 'Okitsu wase,' and 'Nichinan 1 gou' using Integrative Genomics Viewer (IGV) software (Additional file 1: Table S1). Forty-one candidate markers with GC content suitable for primer design were selected. Based on the results mentioned above, 6 positive markers and one negative marker were selected using a PCR assay of 'Haryejosaeng' and control cultivars ( Table 1 and Additional file 2: Table S2).

Multiplex PCR using the SNP markers
A multiplex PCR was employed to obtain amplified products in a single PCR using the selected discrimination markers (P1/P5 and P2/P5) for 'Haryejosaeng.' The P1 and P5 primer pairs simultaneously produced amplicons of approximately 160 and 550 bp, respectively, in the multiplex PCR. In contrast, none of the reactions performed on other cultivars yielded amplicons (Fig. 3a). In addition, the exact size of the amplified product and the non-specific amplicons confirmed that the size of the amplified PCR products was 165 and 526 bp for P1 and P5, respectively; no non-specific bands appeared in other cultivars (Fig. 3b). The P2 and P5 primer pairs simultaneously produced amplicons of approximately 150 and 550 bp, respectively, in the multiplex PCR. In contrast, none of the reactions performed on other cultivars generated amplicons (Fig. 3c). In addition, the exact size of the amplified product and the non-specific amplicons confirmed that the size of the amplified PCR products was 150 and 526 bp for P2 and P5, respectively, and no non-specific bands appeared in other cultivars (Fig. 3d).

Verification of 'Haryejosaeng' in farmhouses
A multiplex PCR was performed using the SNP markers P1/P5 and P2/P5 to specifically identify  (Table 2).

Discussion
Most citrus cultivars reproduce by sexual hybridization, frequent somatic mutation, and nucellar polyembryony; the latter is a feature of citrus in which the nucellar tissue that surrounds the embryonic sac containing the zygotic embryo develops one or more embryos that have the same genetic constitution as the maternal tissue cells [32, [51][52][53]. Satsuma mandarin cultivars have been particularly difficult to improve through hybridization because of polyembryony and male sterility [3,54]. Accordingly, satsuma mandarin cultivars are bred by asexual methods, such as nucellar embryos [4,53] and bud mutation [3,6], and therefore have the same genetic and morphological traits as the mother plant [4]. 'Haryejosaeng' was developed from nucellar seedlings [4] of satsuma mandarin ('Tachima wase'), and it shows higher total soluble solids, and lower acidity than the most common satsuma mandarin ('Miyagawa wase') cultivated on farms [2] and has longer juvenile stages [55]. It is, therefore, difficult to morphologically distinguish the cultivars supplied to farms at the seedling stage [8].
Recently, the genotypes of satsuma mandarin were analyzed using SSR markers, but they could not be distinguished [56], suggesting both morphological and genetic similarities among the different cultivars. Although only 3 of the 7 sets of SNP markers used in the present study could differentiate 'Haryejosaeng' from other satsuma mandarin cultivars, this SNP-based method can be utilized to identify molecular markers capable of distinguishing very close varieties by relying on the differences among nucleotide sequences [18,32]. Therefore, we identified SNP markers that were heterozygous in 'Haryejosaeng' and homozygous in 4 control cultivars, as well as SNP markers that were homozygous in 'Haryejosaeng' and heterozygous in 4 control cultivars (Additional file 1: Table S1).
Using these markers, we could specifically distinguish 'Haryejosaeng' from other cultivars.
We identified 77 candidate SNP markers and selected 6 PCR-positive and one PCR-negative SNP markers among them. The efficacy of these markers in discriminating between cultivars was evaluated ( Fig. 1a and e). Because DNA amplification and sequencing are not required for confirming the results of PCR-based HRM analysis, this methodology can be used to verify the efficiency of SNP markers and detect SNP mutations [41,57,58]. It is a simple, fast, and inexpensive method for detecting polymorphisms [41,42,59]. However, because of the high cost of the equipment required for the HRM analysis, PCR-based SNP marker analysis is considered more efficient than PCR-based HRM analysis [40,42]. Among the 6 positive SNP markers (P1-P6), we were able to select 3 (P1, P2, and P5) that could distinguish 'Haryejosaeng' from other cultivars at the early seedling stage.
Overall, the application of the results obtained in the present study will reduce the time and cost involved in the identification of 'Haryejosaeng' mandarin at the seedling stage, a cultivar that is 8 commercially supplied to farmers.

Conclusion
A method that can distinguish between C. unshiu mandarin cultivars based on their genetic characteristics has not yet been developed. It is, therefore, important to develop markers to identify breeds suitable for cultivation. The multiplex PCR method using the P1, P2 and P5 SNP markers selected in the present study is expected to reduce the time and cost associated with the supply of 'Haryejosaeng' to farms. 'Tachima wase' was produced by crossing the satsuma mandarin 'Tachima wase' with pollens of C. natsudaidai 'Hayata'; the total soluble solids to acidity rate is higher in the improved cultivar than in the common farm cultivar 'Miyagawa wase' and in the mother plant) ( Table 3). Total genomic DNA was extracted from the samples by automatic nuclear extraction (MX 16; Promega, Madison, WI, USA) and stored at -20 °C until use.

Primer design for the SNP/HRM analysis
For designing the primers, we first performed standard genome sequencing of 4 control cultivars ('Miyagawa wase,' 'Miyamoto wase,' 'Okitsu wase,' and 'Nichinan 1 gou') on the Illumina HiSeq 2500 platform (Illumina, San Diego, CA, USA) using the genome of Citrus × clementina 'Clemenules' as reference. The sequence information (approximately 6 Gb) was obtained according to a previously established method [46]. A paired-end DNA library was constructed for genome sequencing using the TruSeq DNA Library Prep Kit (Illumina). The generated raw reads were trimmed using DynamicTrim and LengthSort in the SolexaQA package (v.1.13) [47]. The clean reads were mapped to the standard genome of C. × clementina 'Clemenules' [46], and the consensus sequence was obtained using the Burrows-Wheeler Aligner (BWA) program [48].
Using the selected SNPs, SNP-specific primers were synthesized for the identification of 'Haryejosaeng.' The SNPs were located at the 3′-end of each sequence, and, for specificity, an artificial SNP was introduced in the third base at the 3′-end. In addition, for the re-verification of the SNP marker, primer sets for the HRM analysis were designed on both sides of the SNP locus based on the results of genome sequencing (Table 4).

Cultivar selection using the PCR-based SNP markers
For the selection of 'Haryejosaeng,' multiple PCRs were carried out using the different primer sets (6 pairs) and a negative control primer set (Table 4) nuclease-free water [49]. The PCRs were performed as follows: initial denaturation at 94 °C for 5 min, followed by 30 cycles of 95 °C for 1 min, 58 °C (50 °C for P7 primer) for 30 s, and 72 °C for 30 s, and a final extension at 72 °C for 5 min. The PCR products were electrophoresed on a 1.2%-1.5% agarose gel at 100 V for 30 min or were reconfirmed using a QiAxcel Advanced System electrophoresis device (Qiagen, Hilden, Germany).
For the selection of 'Haryejosaeng,' multiple PCRs were carried out using the different primer sets (6 pairs) and a negative control primer (one set; Table 4). A multiplex PCR was performed using a combination of 2 selected SNP primers. All the multiplex PCRs were performed under the same conditions. nuclease-free water [49]. The PCRs were performed as follows: initial denaturation at 94 °C for 5 min, followed by 30 cycles at 95 °C for 1 min, 58 °C for 30 s, and 72 °C for 30 s, and a final extension at 72°C for 5 min. The multiplex PCR products were electrophoresed on a 1.2% agarose gel at 100 V for 30 min or were reconfirmed using a QiAxcel Advanced System electrophoresis device (Qiagen).

HRM analysis
The HRM analysis was performed according to the modified method described by Park

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its additional files.

Competing interests
The authors declare that they have no competing interests.  z (total length of trimmed reads/total length of raw reads) × 100 y Region except for base = N (unmapped region) within the consensus sequence (reads-mapping region/total length of the reference genome) × 100.      Table 1