PCR-based INDEL markers co-dominant between Oryza sativa, japonica cultivars and closely-related wild Oryza species.

Wild relatives genetically close to cultivars are precious genetic resources for plant breeding. Oryza rufipogon, O. barthii, O. glumaepatula, O. meridionalis and O. longistaminata are such wild species, and are also categorized as AA genome species based on their structural similarities. Chromosome segment substitution lines (CSSLs) are a powerful resource in breeding and genetics, and numerous rice CSSLs have been produced. This study aimed to develop DNA markers for evaluation of CSSLs directly by PCR and subsequent gel electrophoresis. We confirmed that up to 155 of 188 markers developed for detection of japonica-indica INDELs could also detect INDELs between rice cultivars and wild AA-species accessions. Percentages of applicable markers were higher in O. rufipogon accessions (61.7 to 85.6%), and lower in accessions of other four AA species (39.8 to 51.4%). These markers were distributed throughout the rice chromosomes, and will be useful for genotyping of CSSLs and other genetic resources derived from crosses between rice cultivars and closely related wild species.

Introduction

Rice is one of three major crops alongside wheat and maize, and supports a large proportion of the world population’s diet and calories. Therefore, cultivated rice and other crops have been extensively bred for a long time, resulting in development of numerous modern varieties with superior agronomical traits. Local varieties or landraces, widely adapted to various cultivation environments, have been traditionally used as breeding resources. However, recent genome structural analyses revealed that in the breeding lineage of Japanese rice cultivars, a restricted number of superior landraces and their descendants have been repeatedly used as breeding parents (Yamasaki and Ideta 2013). Consequently, the genetic diversity of modern varieties has become smaller, frequently resulting in serious damage to the grain harvest due to genetic breakdown of agronomical traits, such as resistance to plant diseases and insects. Thus, exploration of genetic resources other than landraces is an important and urgent issue in crop breeding. The use of wild plant species is a promising approach to extend the pool of genetic resources.

Wild relatives of the genus Oryza are classified into 21 species (Vaughan and Morishima 2003). Five of them, O. rufipogon, O. barthii, O. glumaepatula, O. meridionalis and O. longistaminata, are genetically close relatives to the cultivated rice species O. sativa and O. glaberrima. Based on the similarity of their genome structures, these seven species are also called AA genome species. F1 hybrids and their descendants can be obtained by the usual crossing and backcrossing in most combinations between AA species, and thus, the wild AA genome species are precious genetic resources for rice breeding. A trial to use these wild resources has already started. Chromosome segment substitution lines (CSSLs), which contain a particular chromosomal segment of a donor parent substituted with the genetic background of a recurrent parent, are a powerful breeding resource and a genetic tool for fine mapping and comprehensive analysis of quantitative trait loci (Yano 2001), and many rice CSSLs have been produced to date (Hirabayashi , Shim , Yoshimura , for example).

This study aimed to develop polymerase chain reaction (PCR)-based insertion/deletion (INDEL) markers to assist in production of rice CSSLs. One merit of co-dominant INDEL markers is easy detection directly by PCR and subsequent gel electrophoresis (Pacurar ). Here, we confirmed that 155 markers out of 188 used for detection of INDELs intraspecific between japonica and indica were also applicable to detect interspecific INDELs between japonica rice cultivars and AA genome wild species.

Materials and Methods

Plant materials

Fourteen accessions of five wild Oryza species were used: O. rufipogon (accession Nos. W0106, W0120, W0137, W1551, W1681), O. barthii (W0049, W1605), O. glumaepatula (W1169, W1171), O. meridionalis (W1297, W2103), and O. longistaminata (W1413, W1508, W1624) (Table 1). Seeds of all accessions were prepared from the wild Oryza collection conserved at the National Institute of Genetics (NIG) (Nonomura ). The japonica cultivars Nipponbare and Taichung 65 (T65), and the indica cv. Kasalath were used as standards of cultivated species (Fig. 1, Supplemental Table 1). All plants were grown in the field of NIG, Mishima, Japan.

Table 1

PCR amplification patterns between japonica cultivars and wild AA genome species using 181 INDEL markers

Accession No.	Species	Origin	No. of INDEL markers*				Total	Application rate (%, Class I/Total)
			Class I	Class II	Class III	Others
W0106	O. rufipogon	India	153	21	6	1	181	84.5
W0120	O. rufipogon	India	129	39	5	7	180	71.7
W0137	O. rufipogon	India	111	51	18	0	180	61.7
W1551	O. rufipogon	Thailand	155	18	8	0	181	85.6
W1681	O. rufipogon	India	147	18	6	10	181	81.2
W0049	O. barthii	unknown	89	54	30	0	173	51.4
W1605	O. barthii	Nigeria	91	58	32	0	181	50.3
W1169	O. glumaepatula	Cuba	89	52	40	0	181	49.2
W1171	O. glumaepatula	Cuba	90	51	40	0	181	49.7
W1297	O. meridionalis	Australia	72	23	85	1	181	39.8
W2103	O. meridionalis	Australia	77	22	81	1	181	42.5
W1413	O. longistaminata	Sierra Leone	73	37	63	8	181	40.3
W1508	O. longistaminata	Madagascar	68	28	61	6	163	41.7
W1624	O. longistaminata	Cameroun	69	31	60	3	163	42.3

See the text about the definition of markers in classes I, II, III and Others.

Fig. 1

Different INDEL patterns among cultivars and AA genome wild rice species. Three independent INDEL markers, chr09.0755, chr10.0007 and chr12.2774, were used for example. C1; class I marker, C2; class II marker, C3; class III marker, CO; class “Others” (See the text and Table 1). Nip (j), Kas (i) and T65 (j) indicate standard rice cultivars, Nipponbare (japonica), Kasalath (indica) and Taichung 65 (japonica), respectively.

Preparation of PCR primers for INDEL detection

The INDEL sites polymorphic between the japonica cv. Nipponbare and the indica cv. 93-11 were first detected by the genome browser of the MSU Rice Genome Annotation Project (http://rice.plantbiology.msu.edu). We selected 785 sites, each of which includes a putative INDEL 15-500 base pairs (bp) in length. PCR primers 20-bp long were designed and confirmed to be usable as sequence-tagged site markers by e-PCR (http://www.ncbi.nlm.nih.gov/tools/epcr/). Subsequent PCR analyses revealed that 595 of the 785 sites could be used as co-dominant INDEL markers between Nipponbare and 93-11. We selected 181 INDEL markers, 160 of which were used in a previous study (Mizuta ), that covered all of the rice chromosomes, and used them in this study.

DNA extraction and PCR analysis

Genomic DNA was extracted from freeze-dried leaf samples according to the CTAB method (Rogers and Bendich 1988). PCR was performed in a 10 μL reaction mixture containing 5 μL GoTaq Green Master Mix (Promega), 5 μM each of a PCR-primer pair, and 2 ng of template DNA. The PCR program used was 98°C for 3 min for initial denaturation, followed by 35 cycles at 94°C for 30 sec, 58°C for 30 sec and 72°C for 1 min, with 72°C for 3 min for a final extension. PCR products were run in 1.5% agarose gels (Certified low range ultra agarose; Bio-Rad) in 0.5x TAE buffer. In Supplemental Tables 1, 2, the length predicted by e-PCR was adopted to determine the size of PCR products; if they were identical between wild species and cultivars on the electrophoretic image, we estimated the approximate size manually by comparison with a size marker, 100 bp DNA Ladder One (Nacalai Tesque).

Results and Discussion

We examined how many of the 181 markers, all of which contain INDELs between cvs. Nipponbare and 93-11 (see Methods), were applicable for detecting polymorphisms between japonica cultivars and each of 14 accessions classified into any of 5 wild species: O. rufipogon, O. barthii, O. glumaepatula, O. meridionalis and O. longistaminata (Table 1).

As a result of PCR analyses, the 181 INDEL markers were classified into three classes: a class I marker that gave co-dominant, polymorphic PCR amplicons between japonica cultivated and wild species (C1 in Fig. 1, for example), class II that gave amplicons, but identical between the two species (C2 in Fig. 1), and class III that gave no amplicon in wild species (C3 in Fig. 1). A class I marker is assumed to contain an INDEL in the internal sequence between a PCR primer pair, with the sequences largely conserved among cultivated and wild species. In case of class II, the internal sequence contains no interspecific INDEL, while sequences corresponding to the primer pair are conserved. A class III marker includes primer sequence polymorphisms deleterious for PCR amplification, and gives no information about the internal sequence polymorphism. Some markers that give multiple or smeared amplicons are inappropriate for genotyping (CO in Fig. 1), and are categorized as “Others” in Table 1. The results of PCR-assisted classification of 181 INDEL markers are summarized in Table 1, and the complete data are shown in Supplemental Table 1. In this study, the class I markers were only defined as “applicable” for genotyping of CSSLs, and the application rate (% of class I out of total markers) was calculated for each wild accession (Table 1).

The applicable marker rates were higher in O. rufipogon accessions. The rates differed intraspecifically among the four accessions we used; the highest was 155 markers (85.6%) in W1551 and the lowest was 111 markers (61.7%) in W0137 (Table 1). The high application rate in this species was largely because it had the fewest class III markers. This is not surprising because O. rufipogon is widely accepted as an ancestral progenitor of O. sativa cultivars (Vaughan and Morishima 2003). These applicable markers were distributed through all rice chromosomes (Fig. 2A), and were sufficient for genotyping of CSSLs and other AA genome-derived genetic resources.

Fig. 2

Distribution of applicable INDEL markers, polymorphic between cv. T65 and closely-related wild rice species; (A) O. rufipogon (acc. No. W0106), (B) O. barthii (W0049), (C) O. glumaepatula (W1169), (D) O. meridionalis (W1297), and (E) O. longistaminata (W1413). The marker names were omitted, but corresponded to those registered in Supplemental Table 2 in the order from the top to the bottom of each chromosome. Centromere positions were indicated by arrowheads.

The application rates declined in the order O. barthii and O. glumaepatula, with the lowest for O. meridionalis and O. longistaminata (Table 1). O. barthii is thought to be an ancestral progenitor of the African cultivated species, O. glaberrima (Vaughan and Morishima 2003), and O. meridionalis and O. longistaminata each have been phylogenetically distinct from ohter AA genome species (Doi , Wang , Xu ). These results were also consistent with our previous report on other INDEL markers (Yamaki ). Thus, we concluded that these results represent generally accepted phylogenetic relationships of AA genome species. In the four wild species other than O. rufipogon, the application rates of INDEL markers were lower, ranging from 39.8 to 51.4%, about half or less than the rates for O. rufipogon accessions (Table 1). However, in each of the four species, at least one or more markers were applicable on each of all chromosomes (Fig. 2B–2D). Thus, these marker sets are also usable for genotyping of CSSLs derived from AA species other than O. rufipogon.

It was noteworthy that a considerable number of wild species-derived amplicons were identical to their corresponding indica-type amplicons in length (Table 2, chr09.0755 in Fig. 1, for example). In particular, more than 85% of the markers applicable for O. rufipogon accessions were of this type. This was consistent with a previous result, in which the level of genetic differentiation of indica cultivars from O. rufipogon species was more modest than that of japonica cultivars (Huang ). Actually, all five O. rufipogon accessions in this study were also used in that report (Huang ). The results of this and previous studies indicate that DNA markers polymorphic for O. rufipogon are obtained more easily in japonica than in indica cultivars. However, we cannot rule out the possibility that the modest differentiation between indica cultivars and O. rufipogon accessions is attributable to mutual introgression at their habitats of origin, such as a paddy, before collection, or introgression during ex situ conservation in germplasm centers. Chromosomal regions tagged with such markers should be handled carefully, particularly when using wild accessions for evolutionary or species differentiation studies. In addition, in CSSL evaluation by INDELs of this study, note that the order of marker alignments not always conserved among the species. Genome rearrangement, such as translocation, inversion and duplication, frequently takes place among Oryza species (Tian ). This study does not take any linkage relation among INDEL markers into consideration.

Table 2

The number of INDEL markers that give indica-type amplicon in wild species used in this study

Accession No.	Species	No. of markers		I/C ratio (%)
		co-dominant with japonica (C)	indica-type amplicon in wild species (I)
W0106	O. rufipogon	153	139	90.8
W0120	O. rufipogon	129	112	86.8
W0137	O. rufipogon	111	96	86.5
W1551	O. rufipogon	155	140	90.3
W1681	O. rufipogon	147	134	91.2
W0049	O. barthii	89	76	85.4
W1605	O. barthii	91	79	86.8
W1169	O. glumaepatula	89	71	79.8
W1171	O. glumaepatula	90	70	77.8
W1297	O. meridionalis	72	43	59.7
W2103	O. meridionalis	77	47	61.0
W1413	O. longistaminata	73	48	65.8
W1508	O. longistaminata	68	46	67.6
W1624	O. longistaminata	69	48	69.6

As a concluding remark, this study showed that up to 155 INDEL markers were co-dominant between japonica rice cultivars and five wild species having AA genome chromosomes (Supplemental Table 1). To make them more convenient for evaluation of CSSLs, the markers were also compiled for each accession (Supplemental Table 2). We are now producing CSSLs of each of the five wild species as donor parents, using cv. T65 as a recurrent parent. All wild accessions used in this study have been confirmed to have strong resistance to bacterial blight diseases (Yoshimura, Nonomura et al., unpublished), and/or to brown planthopper, green rice leafhopper, or both (Yasui, Nonomura et al., unpublished). Thus, it is expected that some of the CSSL descendants bear resistance to diseases or insects inherent in the donor accessions. The INDEL marker sets developed in this study will be a powerful tool for effective selection and evaluation of these CSSLs.