Potential application of simple easy-to-use insertion-deletion (InDel) markers in citrus cultivar identification.

We previously developed insertion-deletion (InDel) markers that distinguish three genotypes (two homozygous and one heterozygous) of diverse citrus cultivars. These InDel markers were codominant and could be clearly detected by using simple agarose gel electrophoresis. We sought to establish a method for cultivar identification using these 28 InDel markers to genotype 31 citrus cultivars. The results revealed that a minimum of 6 markers were required to identify individuals using the three-genotype classification method. Furthermore, we found that a simple method for distinguishing between two genotypes (homozygous and heterozygous) could be used to identify individuals using a minimum of 7 markers. Our findings provide a basis for the development of simple and rapid citrus cultivar identification methods.

Introduction

DNA marker analysis has been widely used as a fundamental tool for classification and identification of diverse citrus cultivars (Shimizu ). Several DNA markers have been developed for citruses, including random amplified polymorphic DNA (RAPD), restriction fragment length polymorphism (RFLP), amplified fragment length polymorphism (AFLP), cleaved amplified polymorphic sequence (CAPS), and simple sequence repeat (SSR) markers (Cai , Chen , Jarrell , Ollitrault , Omura , Shimada ). Dominant markers such as RAPDs and AFLPs, are genotyped with and without polymerase chain reaction (PCR) amplification products, depending on the presence or absence of null allele polymorphisms (Iwata ). Consequently, these makers do not enable the determination of homozygous or heterozygous genotypes. In contrast, codominant markers, such as CAPS, provide PCR amplification products irrespective of the presence or absence of sequence variants and can discriminate differences in the zygosity of loci, thereby facilitating a high degree of genotype discrimination. Unlike SSRs, the results are detectable using simple agarose gel electrophoresis. The use of CAPS markers is a simple and effective method for identifying citrus cultivars (Ninomiya , Nonaka ).

We previously developed insertion-deletion (InDel) markers as a DNA profiling technique to identify hybrid citrus embryos (Noda ). The InDel markers, developed based on Satsuma mandarin whole genome (Shimizu ) re-sequencing analysis, are DNA markers that facilitate genotyping of the zygosity of single-locus regions. Additionally, we demonstrated the transferability of these InDel markers to different citrus species. Similar to CAPS markers, these InDel markers are codominant, thereby facilitating genotyping via simple agarose gel electrophoresis. However, they have a distinct advantage over CAPS markers in that they do not require restriction enzyme treatment.

Our objective in this study was to genotype diverse citrus cultivars, and to verify the applicability of our InDel markers as a basis for methods that could be used to identify citrus cultivars based on the genotyping of these markers.

Materials and Methods

Plant material

We used 31 commercially grown citrus cultivars (17 hybrids and 14 indigenous) (Table 1). These citrus plants were grown on a site at the Kumamoto Prefectural Fruit Tree Research Institute in Japan.

DNA extraction

Total genomic DNA was purified using DNAs-ici.-P DNA extraction kit (RIZO Inc., Ibaraki, Japan). The absorbance of the extracted samples was measured at a wavelength of 260 nm using a Nanodrop 1000 spectrophotometer (Invitrogen, Waltham, Massachusetts, USA).

PCR amplification

PCR amplification was performed in 10 μL reaction volumes consisting of 0.1 μL of Blend Taq® (Toyobo, Osaka, Japan) (2.5 units/μL), 1.0 μL of 10 × buffer for Blend Taq®, 1.0 μL dNTP (2 mM each), 0.2 μL each of forward and reverse primers (10 μM), and 1.0 μL of template DNA (40 ng/μL), under the following conditions: denaturation for 5 min at 95°C; 30 cycles of denaturation at 95°C for 30 s, primer annealing at 55°C for 30 s, and primer extension at 72°C for 30 s.

The resulting PCR products were analyzed using electrophoresis on 4% or 2% agarose gels and visualized by staining with ethidium bromide. Agarose suitable for low molecular weight DNA fragment separation (Agarose KANTO HC; Kanto Chemical, Tokyo, Japan) was used for 4% gels, whereas standard agarose (Agarose-LE; Nacalai Tesque, Tokyo, Japan) was used for 2% gels.

T7 endonuclease assay

We used T7 endonuclease (New England Biolabs) to determine whether the PCR product was homoduplex or heteroduplex DNA. This enzyme specifically cleaves double-stranded DNA containing mismatched structures. PCR fragments were divided into two groups, and 0.25 U of T7 endonuclease was added to one group. The mixture was incubated at 24°C overnight. Undigested and digested samples were analyzed using electrophoresis, using the methods described above.

InDel markers

We selected 28 markers (Supplemental Table 1) from 119 InDel markers (Noda ), based on their transferability to other citrus cultivars and their close relatives.

Genotyping using InDel markers

Genetic zygosity can be divided into three genotypes: one type of heterozygosity, i.e., heteroduplex (LS) and two types of homozygosity, i.e., homoduplex (LL and SS). This is defined as the three-genotype classification method (i.e., LS, LL, and SS) where these genotypes are classified individually, and in the two-genotype classification method (i.e., LS and LS/SS), LL and SS are not classified individually.

The genotypes were defined as indicated below (detailed in Supplemental Figs. 1, 2). (1) If the amplified region was heterozygous, it was expressed as LS when two homozygous DNA bands and heterozygous DNA bands were detected. (2) If the amplified region was homozygous, it was expressed as LL when only the longer of the two homoduplex DNA bands was detected, and SS when only the shorter homoduplex DNA band was detected. (3) All other amplified regions were expressed as X.

Data analysis

To assess the genetic diversity in each InDel marker, the expected heterozygosity, observed heterozygosity, and polymorphic information content (PIC) were calculated using MarkerToolKit ver. 1.0 software (Fujii ). Based on the PIC values, the markers were classified as highly informative (PIC ≥0.5), moderately informative (0.25≤ PIC <0.5), or slightly informative (PIC <0.25) (Botstein ).

Additionally, the probability of the presence of one or more cultivars with matching genotypes was calculated with 28 InDel markers, using the aforementioned software, as described by Ukai (2005).

We used the MinimalMarker program (Fujii ) to detect a set of markers for distinguishing all examined cultivars requiring a minimum number of markers.

Results

The band patterns of the PCR products were analyzed prior to genotyping with the InDel markers. All the InDel markers were detected in the three bands. To clarify the allelic relationships between the detected bands, heteroduplex DNA molecules were removed from the PCR products using T7 endonuclease (Fig. 1). As a result, only the topmost band of the three bands detected by all InDel markers was digested. This indicates that the topmost band was a heteroduplex DNA molecule, and the two bands below it were homoduplex DNA molecules. Of the three bands, the bottom two bands were found to be alleles.

Fig. 1.

Electropherograms for heterozygous loci detected with the use of InDel markers. The separation was confirmed by agarose gel electrophoresis using a 4% gel visualized by staining with ethidium bromide. In the analysis, the left lane of each InDel maker was not treated with T7 endonuclease, whereas the right lane was treated with T7 endonuclease, a cleavage enzyme specific for heteroduplex DNA.

Next, we investigated the genotypes of 31 commercially grown citrus cultivars based on the three-genotype classification method. Fig. 2 shows the electropherograms of genotype discrimination for each InDel marker, which was clearly distinguishable on 4% agarose gels. More than 99% of the genotypes were LS, LL, or SS, as indicated in Table 2, which shows the genotypes for the 31 citrus samples. Importantly, however, there were differences in the genotype combinations of each sample, which allowed us to identify all citrus cultivars using our InDel markers.

Fig. 2.

Electropherograms on 4% agarose gels, and genotyping based on the three-genotype classification method using 28 InDel markers. The genotypes obtained using the InDel markers are expressed as follows: (1) When the amplified region was heterozygous, separation of products by agarose gel electrophoresis yielded three bands containing two homoduplex bands (the bottom two of the three bands) and one heteroduplex band (the top band of the three bands), which migrated at a slower rate than the former. The genotype in this case is expressed as LS. (2) When the amplified region was homozygous, only one of the two homoduplex fragment bands was detected. When a longer band is detected, it is expressed as LL, and when a shorter band is detected, it is expressed as SS. For LG2-3, LG2-4, LG3-2, LG4-10, LG5-25, LG8-8, and LG9-7, only one type of homozygous sample was obtained from the samples used. Therefore, for these markers, only one genotype is shown in the figure. The cultivars used to show the electropherograms of the genotypes at each InDel marker are listed below, with the marker name and the cultivar name in parentheses: LG1-7 (Satsuma mandarin, Harehime, and Ariake), LG1-12 (Satsuma mandarin, Asuki, and Rinoka), LG1-15 (Satsuma mandarin, Rinoka, and Auski), LG2-3 (Satsuma mandarin and Ariake), LG2-4 (Satsuma mandarin and Ariake), LG3-1 (Satsuma mandarin, Ariake, and Ootachibana), LG3-2 (Satsuma mandarin and Ariake), LG3-11 (Satsuma mandarin, Haruka, and Ariake), LG4-1 (Satsuma mandarin, Asuki, and Harehime), LG4-2 (Satsuma mandarin, Asuki, and Benibae), LG4-10 (Satsuma mandarin and Ariake), LG4-12 (Satsuma mandarin, E28, and Ariake), LG5-16 (Satsuma mandarin, Banpeiyu, and Ariake), LG5-25 (Satsuma mandarin and Ariake), LG5-26 (Satsuma mandarin, E28, and Asuki), LG5-28 (Satsuma mandarin, Nankou, and Asuki), LG5-30 (Satsuma mandarin, Haruka, and Ariake), LG6-4 (Satsuma mandarin, Asuki, and Ariake), LG6-23 (Satsuma mandarin, Rinoka, and Asuki), LG7-12 (Satsuma mandarin, Banpeiyu, and Hinoyutaka), LG7-14 (Satsuma mandarin, Banpeiyu, and Reiko), LG8-5 (Satsuma mandarin, Harehime, and Benibae), LG8-6 (Satsuma mandarin, Setoka, and Iyo), LG8-8 (Satsuma mandarin and Ariake), LG8-10 (Satsuma mandarin, Ariake, and Rough lemon), LG8-11 (Satsuma mandarin, Ariake, and Tsunokagayaki), LG9-1 (Satsuma mandarin, Benibae, and Asuki), and LG9-7 (Satsuma mandarin and Asuki).

Table 2.

Genotyping of 31 citrus cultivars based on the three-genotype classification method

Code	InDel markers   Cultivar		LG1-7	LG1-12	LG1-15	LG2-3	LG2-4	LG3-1	LG3-2	LG3-11	LG4-1	LG4-2	LG4-10	LG4-12	LG5-16	LG5-25	LG5-26	LG5-28	LG5-30	LG6-4	LG6-23	LG7-12	LG7-14	LG8-5	LG8-6	LG8-8	LG8-10	LG8-11	LG9-1	LG9-7
		Heb	0.487	0.491	0.487	0.248	0.200	0.487	0.121	0.495	0.495	0.481	0.062	0.398	0.367	0.148	0.498	0.473	0.471	0.433	0.499	0.487	0.492	0.495	0.498	0.312	0.248	0.383	0.467	0.175
		Hoc	0.645	0.600	0.645	0.290	0.226	0.710	0.129	0.516	0.581	0.613	0.065	0.226	0.419	0.161	0.419	0.567	0.276	0.433	0.645	0.645	0.677	0.774	0.806	0.387	0.226	0.387	0.548	0.194
		PICd	0.368	0.371	0.368	0.217	0.180	0.368	0.113	0.373	0.373	0.365	0.060	0.319	0.300	0.137	0.374	0.361	0.360	0.339	0.375	0.368	0.371	0.373	0.374	0.263	0.217	0.310	0.358	0.160
1	Ariake		SS	LS	LS	LL	SS	LL	LL	SS	LS	LS	LL	SS	SS	LL	LS	LS	SS	SS	LS	LS	LS	LS	LS	LL	LL	LL	LS	LS
2	Asuki		LS	LL	SS	LL	SS	LS	LL	SS	LL	LL	LL	SS	SS	LL	SS	SS	SS	LL	SS	LS	LS	LS	LS	LL	LL	LL	SS	LL
3	Asumi		LS	LL	SS	LL	SS	LS	LL	SS	LS	LS	LL	SS	SS	LL	SS	LS	SS	LL	LS	LS	LS	LS	LS	LL	LL	LL	LS	LS
4	Benibae		LS	LS	LS	LS	SS	LL	LL	LS	LL	SS	LL	SS	SS	LL	LS	SS	SS	LL	LS	LS	LS	SS	LS	LL	LL	LL	LL	LS
5	E28		LS	LS	LS	LL	SS	LS	LS	SS	LL	LL	LL	LL	SS	LS	LL	LS	LS	LS	LS	LS	LS	LS	LS	LL	LL	LS	LL	LL
6	Harehime		LL	LS	LS	LS	SS	LL	LS	LS	SS	LL	LL	SS	SS	LL	LS	LS	SS	LS	LS	LS	LS	LL	LS	LL	LL	LS	LL	LL
7	Haruka		LS	LS	LS	LS	LS	LL	LL	LL	LS	LS	LL	SS	SS	LL	LL	LS	LL	LL	SS	LS	LS	LS	LS	LS	LS	LS	LS	LL
8	Hinoyutaka		LL	LL	SS	LL	SS	LS	LL	SS	LL	LL	LL	SS	SS	LL	LS	SS	SS	LL	LS	SS	LS	LS	LS	LS	LL	LS	LS	LL
9	Kumamoto EC10		LS	LL	SS	LL	SS	LS	LL	LS	LS	LS	LL	SS	SS	LL	LS	LS	SS	N.D.	LS	LS	LS	LS	LS	LL	LL	LL	LS	LS
10	Kumamoto EC12		LS	LS	LS	LL	SS	LS	LL	SS	LL	LL	LL	SS	SS	LL	SS	LS	SS	LL	LS	LS	LS	LS	LS	LL	LL	LL	LL	LL
11	Nankou		LS	LL	SS	LL	SS	LS	LS	LS	LS	LS	LL	LS	SS	LS	LL	LL	LS	LS	LS	LS	LS	LS	LS	LL	LL	LS	LS	LL
12	Reiko		LS	LS	LS	LL	SS	LS	LL	LS	LS	LS	LL	SS	SS	LL	SS	SS	SS	LL	LS	SS	SS	LS	LS	LS	LL	LS	LL	LL
13	Rinoka		LL	SS	LL	LL	SS	LS	LL	LL	SS	LL	LL	SS	LS	LL	LL	X	LL	LL	LL	LS	LS	LL	LS	LL	LS	LS	LS	LL
14	Seinannohikari		LL	LL	SS	LS	LS	LS	LL	SS	LS	LS	LL	LS	SS	LL	SS	SS	SS	LS	SS	LS	LS	LS	LS	LL	LL	LL	LL	LL
15	Setoka		LS	LS	LS	LL	SS	LS	LL	SS	SS	LL	LL	SS	SS	LL	SS	SS	SS	LS	LS	SS	SS	LL	LL	LL	LL	LL	LS	LL
16	Tsunokagayaki		LL	LS	LS	LS	SS	LS	LL	SS	LS	LS	LL	LL	LS	LL	LS	SS	LS	LL	SS	SS	SS	LS	LS	LS	LL	SS	LS	LL
17	Tsunonozomi		LS	LS	LS	LS	SS	LS	LL	LL	LS	LS	LL	SS	SS	LL	SS	SS	SS	LL	LS	SS	SS	LS	LS	LL	LL	SS	LS	LL
18	Banpeiyu		LS	SS	LL	LL	SS	SS	LL	LL	SS	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	LL	SS	LL
19	Hyuganatu		LL	LS	LS	LS	LS	LS	LL	LL	LS	LS	LL	LS	LS	LL	LL	LS	LL	LL	LS	LS	LS	LS	LS	LS	LL	LS	LS	LL
20	Iyo		LS	LS	LS	LL	SS	LS	LL	LS	LS	LS	LL	LS	LS	LL	LS	LS	LS	SS	SS	LS	LS	LS	SS	LL	LL	LS	LS	LL
21	Kawachibankan		LS	LS	LS	LL	SS	LS	LL	LS	LS	LS	LL	LL	LS	LL	LL	LS	LL	LS	LL	LL	LL	LS	LL	LS	LL	LL	LS	LL
22	Kunenbo		SS	LS	LS	LS	LS	LS	LL	LS	LS	LS	LL	LS	LS	LS	LS	LS	LS	SS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LL
23	Meyer lemon		SS	LL	LS	LL	SS	LS	LL	LL	SS	LL	LS	SS	LS	LL	LL	LS	X	LS	LL	LS	LS	LL	LS	LL	LL	LS	LL	LL
24	Natsumikan		LS	LS	LS	LL	LS	LS	LL	LS	LS	LS	LL	SS	LS	LL	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LL	LS	LL
25	Ootachibana		LS	SS	LL	LL	LS	SS	LL	LS	LS	LS	LL	LL	LS	LS	LL	LL	LL	LL	LL	LL	LL	LS	LL	LS	LS	LL	SS	LL
26	Ponkan		LS	LL	SS	LL	SS	LL	LL	LS	LL	SS	LL	SS	SS	LL	LS	SS	SS	LS	LS	SS	SS	SS	SS	LS	LL	LL	LL	LL
27	Rough lemon		LL	N.D.	LS	LL	SS	LS	LL	LS	LL	SS	LL	SS	LS	LL	LL	LS	LS	LS	LS	LS	LS	LS	LS	LL	SS	LL	LL	LL
28	Satsuma mandarin		LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS	LS
29	Sudachi		LS	SS	SS	LL	SS	LS	LL	LS	LL	LS	LL	LS	LS	LL	LS	LS	N.D.	LL	LS	SS	SS	LS	LS	LS	LS	LL	LL	LL
30	Sweet orange		LS	LS	LS	LL	SS	LL	LL	LS	LS	LS	LL	SS	LS	LL	LS	LS	SS	LS	LS	LS	LS	LS	LS	LL	LL	LL	LS	LS
31	Tankan		LL	LS	LS	LL	SS	LL	LL	LS	LS	LS	LL	SS	SS	LL	SS	SS	LL	LS	LL	SS	SS	LS	LS	LL	LL	LL	LL	LL

Genotypes obtained using the InDel markers were defined as follows: if the amplified region was heterozygous, it was expressed as LS; if the amplified region was homozygous, it was expressed as LL when the longer of the two homoduplex DNA bands was detected, and SS when the shorter of the two DNA bands was detected; any other amplified region was expressed as X. N.D. stands for “not detected.”, He: expected heterozygosity, Ho: observed heterozygosity, PIC: polymorphic information content.

Genetic diversity assessments based on each of the InDel marker genotypes revealed that the expected heterozygosity for these markers ranged from 0.062 (LG4-10) to 0.499 (LG6-23) and averaged 0.389, whereas the observed heterozygosity ranged from 0.065 (LG4-10) to 0.806 (LG8-6) and averaged 0.458. Furthermore, the PIC ranged from 0.060 (LG4-10) to 0.375 (LG3-23), with an average of 0.304, and the PIC statistics revealed that most of the InDel markers were moderately informative (highly informative: 0/28, moderately informative: 21/28, slightly informative: 7/28) (Table 2).

We calculate the minimum number of markers required to identify each cultivar based on the three genotypes (LS, LL, and SS, and excluding X). The results revealed that a minimum of 6 markers, with 32 possible combinations, were necessary for distinguishing the 31 cultivars (Supplemental Table 2).

When using the three-genotype classification method, the differences in the sizes of the LL and SS products were between 17 and 29 base pairs (bp), thereby rendering it difficult to determine whether the band was LL or SS in a single lane of an agarose gel. Accordingly, to facilitate the detection of slight differences in length, it is best to use high-resolution 4% agarose gels. Additionally, to compare band positions, it is preferable to run a sample of the LS genotype as an internal standard alongside samples of LL or SS genotypes. Thus, we designed a method of cultivar identification that is simpler and less expensive than the three-genotype classification method. The newly developed two-genotype classification method is based on the detection of only two genotype classes, namely, homoduplex (i.e., LL or SS) and heteroduplex (i.e., LS), without distinguishing between LL and SS. This enables simple discrimination, as LS is detected as three bands and LL or SS is detected as single band. Additionally, the genotype can be identified on a 2% gel without comparison with the internal standard (LS).

Supplemental Fig. 3 shows the electropherograms obtained using this method. Although the bands were not as distinct as those on the 4% gel, the information was sufficient to distinguish between the LS and LL/SS genotypes. Supplemental Table 3 shows the results of genotyping, including only two genotypes (LS and LL/SS). Base on the genotype data obtained for LS and LL/SS (excluding X), we calculated the minimum number of markers required to identify each of the assessed cultivars and found that it was possible to identify each cultivar using a minimum of 7 markers (Supplemental Table 4), which is slightly higher than the 6 markers required to distinguishing the three genotypes. These 7 markers yielded 19 possible combinations.

In the case of the three-genotype classification method, we found that the probability of the presence of one or more cultivars with matching genotypes detected using the 28 InDel markers ranged from less than 10–30 (Banpeiyu) to 6.2 × 10–5 (Kumamoto EC10), with a median of 2.5 × 10–8 (Haruka). In the case of the two-genotype classification method, the probability ranged from 1.5 × 10–10 (Satsuma mandarin) to 1.8 × 10–4 (Kumamoto EC12), with a median of 2.4 × 10–6 (Setoka) (Table 3). Simplifying the genotype classification from three to two genotypes would reduce the ability to identify a variety by several magnitudes. This was the case for Ootachibana, for which the probability decreased significantly to a probability of 1.3 × 10–8. Because all Satsuma mandarin loci were heterozygous, there was no change in their probability.

Table 3.

The probability of the presence of one or more cultivars with matching genotypes detected using the 28 InDel markers

Cultivars	Classification methods		B/A
	Three genotypesa (A)	Two genotypesb (B)
Ariake	1.8 × 10–7	1.9 × 10–5	104.2
Asuki	4.5 × 10–8	2.8 × 10–5	635.4
Asumi	2.0 × 10–6	4.3 × 10–5	22.0
Benibae	9.2 × 10–9	1.3 × 10–6	137.5
E28	2.2 × 10–8	9.6 × 10–7	43.9
Harehime	1.2 × 10–8	2.7 × 10–7	22.9
Haruka	2.5 × 10–8	1.5 × 10–6	62.0
Hinoyutaka	1.2 × 10–7	5.4 × 10–6	46.5
Kumamoto EC10	1.2 × 10–5	5.8 × 10–5	4.8
Kumamoto EC12	6.2 × 10–6	1.8 × 10–4	29.4
Nankou	1.3 × 10–8	2.2 × 10–7	17.8
Reiko	1.8 × 10–6	3.4 × 10–5	18.6
Rinoka	6.1 × 10–10	1.7 × 10–6	2786.7
Seinannohikari	1.3 × 10–8	9.1 × 10–7	71.7
Setoka	1.5 × 10–8	2.4 × 10–6	163.1
Tsunokagayaki	2.0 × 10–9	1.5 × 10–6	743.9
Tsunonozomi	1.2 × 10–7	4.0 × 10–4	346.9
Banpeiyu	10–30>	2.1 × 10–7	–
Hyuganatu	1.1 × 10–7	2.7 × 10–6	23.5
Iyo	6.4 × 10–8	2.5 × 10–6	39.0
Kawachibankan	6.2 × 10–9	6.3 × 10–6	1016.0
Kunenbo	2.1 × 10–9	4.4 × 10–8	20.8
Meyer lemon	9.7 × 10–10	3.6 × 10–7	369.4
Natsumikan	3.0 × 10–6	4.3 × 10–6	1.4
Ootachibana	1.0 × 10–14	1.3 × 10–8	1229019.0
Ponkan	1.0 × 10–10	1.4 × 10–7	1376.2
Rough lemon	5.4 × 10–8	3.8 × 10–5	693.7
Satsuma mandarin	1.5 × 10–10	1.5 × 10–10	1.0
Sudachi	5.1 × 10–8	1.2 × 10–6	22.7
Sweet orange	7.2 × 10–6	2.0 × 10–5	2.8
Tankan	7.3 × 10–8	8.1 × 10–6	110.9

Classification into three genotypes: one heterozygous (LS) and two types of homozygous (LL and SS). Classification into two genotypes: heterozygous (LS) and homozygous (LL or SS).

Discussion

Several studies have identified citrus cultivars using CAPS markers in Japan, and the efficacy of this technique has previously been demonstrated (Ninomiya , Nonaka ). Similar to CAPS markers, the method used in this study can identify cultivars using a combination of different DNA marker genotypes as indicators. Ninomiya reported that 33 cultivars could be identified using 8 of the 9 assessed CAPS markers, whereas Nonaka reported that 59 cultivars could be identified using 8 of the 19 CAPS markers. In this study, we identified 31 cultivars using 6 of the 28 InDel markers by the three-genotype classification method and 7 of the 28 markers by the two-genotype classification method. Thus, InDel and CAPS markers can be considered comparable regarding the minimum number of markers required for cultivar identification.

However, InDel markers have two disadvantages compared with CAPS markers when differentiating three genotypes. These disadvantages are the necessity of an internal standard and the use of high-resolution 4% agarose gel to distinguish small PCR-amplified DNA fragments, which is more expensive than the commonly used 2% gel. We compensated for these negatives by developing a method for distinguishing between heterozygous and homozygous genotypes.

InDel markers are codominant markers that can be readily detected by simple agarose gel electrophoresis, similar to CAPS markers. However, compared with CAPS markers, InDel markers enable more rapid processing and cost less, as their use does not necessitate a restriction enzyme treatment step.

It is desirable to have consistent genotype changes for highly advanced discrimination of cultivars because of recombination during crossbreeding. Of the cultivars genotyped in this study, only ‘Haruka’ and its parents ‘Hyuganatsu’ and ‘Natsumikan’ were available for the parent-progeny test. No genotypic inconsistencies for any of the 28 InDel markers were found in this trio (Table 2). In addition, we previously produced hybrid individuals by crossing Satsuma mandarin strains and detected no genotypic inconsistencies in either the parent or their hybrid strains (Noda ).

Although ‘Nishinoka’ and ‘Shiranui’ have been reported to be the female and male parents of ‘Kanpei’ respectively, at the time of breeding (Shigematsu ), based on genotyping using CAPS markers, Ninomiya established that the true male parent was ‘Ponkan’. Using the same cultivars, we conducted a parent-progeny test using InDel markers (Supplemental Table 5). When ‘Nishinoka’ and ‘Ponkan’ were assumed to be the parents of ‘Kanpei’, we detected no genetic inconsistencies in genotyping for any of the 28 InDel markers. However, when ‘Shiranui’ is assumed to be the male parent of ‘Kanpei’, we identified genetic inconsistencies with respect to InDel markers LG5-6 and LG5-30. Furthermore, when ‘Nishinoka’ and ‘Shiranui’ were assumed to be the parents, we detected genetic inconsistencies at InDel markers LG1-7, LG5-6, and LG5-30. Thus, the results obtained for parent-progeny analysis based on the use of InDel markers are clearly consistent with those obtained using CAPS markers. Consequently, these findings indicate that InDel markers can be used effectively in addition to CAPS markers to verify genetic identity in trios and can thus be used to identify citrus cultivars in Japan.

By accumulating further genotypic data on a wide range of citrus cultivars, including registered cultivars and their parents, we anticipate that this method will become more practical for cultivar identification than other methods.

InDels tend to occur less frequently in the genome than SSRs or single nucleotide polymorphisms (SNPs) (Garcia-Lor , Wang ). Moreover, InDels and SNPs have a lower degree of intraspecific differentiation than that observed in SSRs, thereby making it easier to obtain consistent genotypes within species. This advantage enable more effective analysis of interspecific differentiation (Garcia-Lor , Ollitrault ) and indicates that InDels and SNPs provide consistent information on interspecific polymorphisms. These makers would be useful as identification markers for cultivars bred by crossing. However, some citrus cultivars have been bred based on intraspecific mutations, such as bud mutations, and to date, there have been no reports of DNA markers that can be used to differentiate mutant strains from wild types of the same species. Although we have also studied the genotypes of several mutant strains (Aoshima unshu, Kawata unshu, Imamura unshu, and Shirakawa unshu) of Satsuma mandarins using our InDel markers, we were unable to detect inter-strain polymorphisms for any of these markers (Noda ). SSRs have been shown to be the most polymorphic markers at the intraspecific level (Ollitrault ). Thus, in addition to polymorphic markers, such as InDel markers, which are stable between species, the use of potential intraspecific markers, such as SSRs, may enhance the adaptability of marker-based detection to the identification of a wide range of citrus cultivars, including mutant strains and hybrid cultivars.

We examined a relatively broad range of citrus cultivars (31) in this study. However, as of Noda , there are 87 citrus cultivars for which breeder’s rights have been granted in Japan (http://www.hinshu2.maff.go.jp/vips/cmm/apCMM110.aspx?MOSS=1), and breeders will continue to develop more cultivars in the future. Using the 28 InDel markers assessed in this study, we were able to distinguish 31 citrus cultivars using a minimum of 6 markers (32 sets). Additionally, we developed a clear and simple method for distinguishing between two genotypes that required a minimum of 7 markers (19 sets). We also established that the probability of detection of one or more cultivars with matching genotypes using the 28 InDel markers is extremely low (up to 1.8 × 10–4), even with the simple two-genotype classification method. Thus, these estimates indicate that the newly developed method for InDel marker-based cultivar identification can be applied to many more cultivars than those assessed in this study.

Author Contribution Statement

TN, KD, and YN conceived and designed the experiments. TM prepared citrus genetic resources. TN performed the experiments. The manuscript was drafted by TN, edited by YN, and was subsequently reviewed by all authors.

Supplemental Figures

Supplemental Tables

A. Hybrid cultivars
Code	Cultivar name	Female parent	Male parent	Application Numbera	Literature Cited
1	Ariake	Sweet orange	Clementine	N/A	Yamada et al. (1995)
2	Asuki	Okitsu-46	Harumi	32235
3	Asumi	Okitsu-46	Harumi	26542
4	Benibae	HF-9	Encore	18646
5	E28	Nankou	Amakusa	15522
6	Harehime	E-647	Satsuma mandarin	13718
7	Haruka	Hyuganatsu	Natsumikan	6328
8	Hinoyutaka	Kiyomi	Ponkan	12477
9	Kumamoto EC10	Ariake	Harumi	24962
10	Kumamoto EC12	Harehime	Harumi	31959
11	Nankou	Satsuma mandarin	Clementine	2282
12	Reiko	KyEn-5	Murcott	15934
13	Rinoka	Lisbon lemon	Hyuganatsu	28741
14	Seinannohikari	KyOw-21	Youkou	21449
15	Setoka	Kuchinotsu-37	Murcott	10852
16	Tsunokagayaki	KyOw-14	Encore	21450
17	Tsunonozomi	Kiyomi	Encore	25000

Application No. of the plant variety protection system in Japan.

B. Indigenous cultivars
Code	Cultivar name	Scientific name	Strain name
1	Banpeiyu	Citrus maxima Merr.	(Stock strain)
2	Hyuganatu	C. tamurana hort. ex Tanaka	(Stock strain)
3	Iyo	C. iyo hort. ex Tanaka	(Stock strain)
4	Kawachibankan	C. maxima Merr.	(Stock strain)
5	Kunenbo	C. nobillis Lour. Kunep Tanaka	(Stock strain)
6	Meyer lemon	C. meyerii Yu. Tanaka	(Stock strain)
7	Natsumikan	C. natsudaidai Hayata	Kawanonatsudaidai
8	Ootachibana	C. maxima Merr.	(Stock strain)
9	Ponkan	C. reticulata Blanco	(Stock strain)
10	Rough lemon	C. jambhiri Lush	(Stock strain)
11	Satsuma mandarin	C. unshiu Marcov.	Aoshima unshu
12	Sudachi	C. sudachi hort. ex. Shirai	(Stock strain)
13	Sweet orange	C. sinensis (L.) Osbeck	Washington nevel
14	Tankan	C. tankan Hayata	(Stock strain)