Development of 12 sets of chromosome segment substitution lines that enhance allele mining in Asian cultivated rice

Abstract

Many agronomic traits that are important in rice breeding are controlled by multiple genes. The extensive time and effort devoted so far to identifying and selecting such genes are still not enough to target multiple agronomic traits in practical breeding in Japan because of a lack of suitable plant materials in which to efficiently detect and validate beneficial alleles from diverse genetic resources. To facilitate the comprehensive analysis of genetic variation in agronomic traits among Asian cultivated rice, we developed 12 sets of chromosome segment substitution lines (CSSLs) with the japonica background, 11 of them in the same genetic background, using donors representing the genetic diversity of Asian cultivated rice. Using these materials, we overviewed the chromosomal locations of 1079 putative QTLs for seven agronomic traits and their allelic distribution in Asian cultivated rice through multiple linear regression analysis. The CSSLs will allow the effects of putative QTLs in the highly homogeneous japonica background to be validated.

Introduction

With its wide genetic diversity, Asian cultivated rice (Oryza sativa L.) is grown in diverse agroecological systems (Vaughan et al. 2008a, 2008b, Zong et al. 2007). Such genetic resources are the primary materials for exploring and characterizing beneficial agronomic traits with which to improve rice for future food security. Using a subset of accessions representing the genetic diversity of rice is an efficient approach for this purpose, and genomics-based analyses such as genetic mapping and genome-wide association study have accelerated these efforts (Cu et al. 2021, Ebana et al. 2008, Kojima et al. 2005, Tanaka et al. 2020). However, as agronomic traits are controlled by multiple quantitative trait loci (QTLs) with different magnitudes of effects (Yamamoto et al. 2009, Yano and Sasaki 1997), such analyses do not offer desirable resolution in detecting alleles with minor effects, and have difficulty in accurately assessing the effects of alleles owing to the highly heterogeneous genetic backgrounds and biased allele frequencies among studies (Korte and Farlow 2013, Nagata et al. 2015, Shen and Carlborg 2013). Therefore, establishing an additional workflow that lets us efficiently explore beneficial alleles from diverse genetic resources and validating them in a genetic background of commercial cultivars is necessary to promote the breeding programs.

Advanced backcross populations created by successive backcrossing and marker-assisted selection can be used to overcome these difficulties because their highly uniform genetic background increases the ability to detect QTLs with even minor effects (Yamamoto et al. 2000). For example, genetic analysis using BC₄F₂ lines detected 65 QTLs for grain shape (grain length or width) in a single japonica × indica rice cross, including some newly identified QTLs (Nagata et al. 2015). An analysis using BC₄F₂ lines from 11 diverse donors of Asian cultivated rice detected 255 QTLs for heading date, of which 127 were detected in regions that differed from those of the 650 previously identified QTLs for heading date (Hori et al. 2015). Such evidence confirms that advanced backcross populations are suitable for dissecting agronomic traits and suggests that an increase in the number of donors will contribute to the comprehensive elucidation of the genetic control of agronomic traits in Asian rice cultivars. However, in using BC₄F₂ lines, it is necessary to manage a large number of individuals to estimate QTLs, owing to the segregating nature of the phenotypes of interest; for example, studies have used more than 1000 individuals per cross combination (Hori et al. 2015, Nagata et al. 2015). It is difficult to use such populations for analyzing agronomic traits, such as yield or quantitative disease resistance, that require much labor for trait measurement and care in controlling environmental conditions. Therefore, a procedure that does not require large numbers of plant materials would be desirable for evaluating such traits.

Chromosome segment substitution lines (CSSLs) are a type of advanced backcrossed population wherein usually single chromosome segments from a donor are substituted in the genetic background of a recurrent cultivar, and are usually homozygous. Since respective chromosomal regions are selected to cover all of the donor’s chromosomes in given lines, a set of CSSLs lets us detect QTLs distributed throughout the genome with high sensitivity and by using fewer plants than other genetic mapping populations (reviews: Balakrishnan et al. 2019, Xi et al. 2006, Yamamoto et al. 2009). Through the evaluation of CSSLs, complex agronomic traits such as heading date, yield components, preharvest sprouting resistance, and grain quality have been finely dissected to identify new QTLs in rice (Abe et al. 2013, Ando et al. 2008, Bian et al. 2010, Chen et al. 2007, Ebitani et al. 2005, Hao et al. 2009, Hori et al. 2010, Ishikawa et al. 2005, Kato and Hirayama 2021, Kubo et al. 2002, Li et al. 2022, Liang et al. 2021, Mizuno et al. 2018, Murata et al. 2014, Ookawa et al. 2016, Sun et al. 2022, Takai et al. 2009, Uga et al. 2015, Ujiie et al. 2012, Wang et al. 2021, Wu et al. 2020, Yasui et al. 2010, Yuan et al. 2022, Zeng et al. 2006, Zhang et al. 2019, Zhu et al. 2009). Therefore, increasing the number of CSSLs available for research and evaluation of agronomic traits can contribute to our understanding of the genetic control of agronomic traits and improvement of rice.

Donors for CSSLs are usually selected on the basis of their superiority in certain agronomic traits or from cultivars distantly related to the recurrent parents. Researchers and breeders select donors and recurrent parents to meet their objectives, and therefore those of CSSLs usually differ among projects (Balakrishnan et al. 2019, Fukuoka et al. 2010). More than 50 sets of CSSLs that have been developed used 24 recurrent parents of both indica and japonica groups (Balakrishnan et al. 2019). Since the effects of some QTL alleles may differ in different genetic backgrounds, sets of CSSLs derived from diverse donors in the same recurrent parents are ideal to compare the effects of QTLs for agronomic traits and for systematically choosing preferable alleles in practical breeding programs. Extensive efforts to develop CSSLs in an indica background that meet this requirement have recently been reported (Zhang 2021), whereas efforts to use the genetic variation in japonica genetic backgrounds remain scarce.

To prepare a set of plant materials for a comprehensive analysis of genetic variation in agronomic traits among Asian cultivated rice, we developed 12 sets of CSSLs in the japonica background, 11 with the same genetic background, using donors selected from the World Rice Core Collection (WRC) (Kojima et al. 2005) to represent the genetic diversity of Asian cultivated rice. Using these materials, we found putative QTLs for seven agronomic traits to overview their distribution mode among Asian cultivated rice through multiple linear regression analysis.

Materials and Methods

Plant materials

We used 10 rice cultivars (‘Basilanon’, ‘Bei Khe’, ‘Bleiyo’, ‘Deng Pao Zhai’, ‘Khau Mac Kho’, ‘Khao Nam Jen’, ‘Muha’, ‘Naba’, ‘Qiu Zhao Zong’, ‘Tupa 121-3’; Table 1, Supplemental Fig. 1) selected from the WRC (Kojima et al. 2005) as donors for developing CSSLs in the japonica ‘Koshihikari’ genetic background. We selected them to represent the genetic diversity among Asian cultivated rice on the basis of DNA polymorphisms and geographical origin. We used another two cultivars, ‘Hayamasari’ and ‘Silewah’ (Table 1, Supplemental Fig. 1), that are adapted to high-latitude or cold rice growing areas (Fujino and Sekiguchi 2005, Satake 1979). For assessing cold tolerance, we used ‘Hitomebore’, which is genetically close to ‘Koshihikari’ but heads early, as the recurrent parent in the CSSL for ‘Silewah’. We began with an F₁ population derived from a cross with ‘Koshihikari’ or ‘Hitomebore’ and repeatedly backcrossed the progeny to the original parent to produce BC₄F₁ plants. If BC₄F₁ plants could not be obtained, self-pollinated progeny were obtained from the BC₃F₁. Finally, we selected 36 to 42 CSSLs with a ‘Koshihikari’ background and 35 with a ‘Hitomebore’ background and grew the plants with their parents (Fig. 1). All plants were grown in an experimental field at the National Agriculture and Food Research Organization (NARO), Tsukuba, Japan (36.03°N, 140.11°E). Basal fertilizer was applied at 56 kg/ha N, 56 kg/ha P₂O₅, and 56 kg K₂O/ha. Month-old seedlings were transplanted in mid May at one per hill in plots with a double row for each line, at 18 cm between plants and 36 cm between rows.

Table 1. Overview of 12 sets of chromosome segment substitution lines in Asian cultivated rice

Code	Donor			Recurrent variety name	No. of lines	SL numbers	Averages of donor’s chromosomes per line (Mb)			Range of target chromome per line (Mb)	Recurrent genome (%)b	Un-covered genome (%)c
	Variety namea	Origin	Subspecies				Target	Non-target	Heterozygous
BAK	Basilanon*	Philippines	indica	Koshihikari	46	SL3001–3046	12.08	3.44	2.85	0.52–32.64	90.9	5.4
BKK	Bei Khe*	Cambodia	indica	Koshihikari	43	SL2401–2443	13.69	3.77	2.35	3.07–26.57	92.2	3.6
BLK	Bleiyo*	Thailand	indica	Koshihikari	41	SL3301–3341	11.28	2.65	4.75	0.11–24.96	92.1	4.0
DPK	Deng Pao Zhai*	China	indica	Koshihikari	36	SL3101–3136	15.22	5.60	6.53	2.07–26.19	93.8	0.8
HAK	Hayamasari	Japan	japonica	Koshihikari	40	SL901–940	15.09	2.73	3.12	2.70–35.82	93.3	2.2
KMK	Khau Mac Kho*	Vietnam	japonica	Koshihikari	46	SL2702–2747	12.26	2.25	3.51	1.20–24.98	92.9	3.4
KNK	Khao Nam Jen*	Laos	japonica	Koshihikari	40	SL2801–2840	11.69	1.84	2.23	1.81–23.68	90.9	5.6
MUK	Muha*	India	indica	Koshihikari	42	SL2601–2642	13.15	5.15	2.32	0.22–23.18	91.1	4.5
NAK	Naba*	India	indica	Koshihikari	41	SL3201–3241	14.48	3.13	0.31	1.38–28.81	90.9	0.7
QZK	Qiu Zhao Zong*	China	indica	Koshihikari	43	SL3401–3443	14.16	3.33	3.50	2.49–25.67	94.6	1.1
TUK	Tupa121-3*	Bangladesh	indica	Koshihikari	48	SL2501–2548	13.70	2.56	2.35	2.86–27.17	94.1	1.8
SIH	Silewah	Indonesia	japonica	Hitomebore	35	SL2901–2935	13.78	4.96	3.01	2.10–28.96	89.0	6.2

^a * Selected from world rice core collection (WRC) of the NARO Genebank.

^b Proportion of recurrent genome in each CSSL.

^c Proportion of regions in which none of the lines have donor introgression among CSSLs.

Fig. 1.

Schematic representation of the development of CSSLs. Marker-assisted selection (MAS) was conducted mainly in backcrossed generations (—), but when target chromosomal regions segregated, it was further conducted in fixing generations (- - -). Proportions of the number of lines are indicated below the three types of lines, which differed in the number of backcrosses.

Assessment of agronomic traits

Days to heading of a plant was scored as the number of days from sowing to the appearance of the first panicle. That of parental accessions or individual CSSLs was determined as the day when 50% of individuals in each line headed. Since three donors (‘Khao Nam Jen’, ‘Khau Mac Kho’, and ‘Bleiyo’) did not head under natural daylength conditions, we transferred them to a short-day treatment cabinet (9-h daylength) at 70 days after sowing to promote heading. Culm length was measured as the distance from the ground to the panicle neck of the main stem. Panicle length was measured as the distance from there to the panicle tip. The number of panicles per plant was counted. The seed fertility rate was measured in one representative panicle. Grain length and width were measured as described (Nagata et al. 2015) in 600-dpi images of ~200 seeds per individual in SmartGrain grain shape analysis software (Tanabata et al. 2012). Five plants were evaluated per line for culm length, panicle length, number of panicles, seed fertility rate, grain length, and grain width.

Blast [Pyricularia oryzae (syn. Magnaporthe oryzae)] resistance in the 10 sets of CSSLs whose donors were selected from WRC was evaluated in an experimental field at NARO, Tsukuba, where the predominant fungal race is 037.3, which caused lesions on our susceptible plants. Plants were grown from 50 seeds per line in each of two or three independent experiments. The blast disease severity of 60- to 70-day-old plants was scored from 0 (highly resistant: no symptoms) to 10 (highly susceptible: leaves totally dead) as described (Asaga 1976). ‘Koshihikari’ was grown on either side of each line as a susceptible control.

DNA extraction and marker analysis

Fresh leaves were harvested from field-grown plants, and total DNA was extracted from leaf samples by the CTAB method (Murray and Thompson 1980). We conducted a whole-genome survey using 98 to 132 simple sequence repeat (SSR) markers or indel markers (Supplemental Table 1) in two backcross generations (BC₁F₁ and BC₄F₁) to select target chromosome segments in each CSSL and to minimize non-target chromosome segments from the donor, and surveyed target chromosomes in two other backcross generations (BC₂F₁ and BC₃F₁), and then selected plants homozygous for the target chromosome segments in the self-pollinated progeny (Supplemental Table 2). We determined the genetic backgrounds and sizes of introgressed segments of respective CSSLs, which we characterized using 340–539 SNP markers (average, 450) per population (Supplemental Table 3-1–3-12), using 768-plex SNPs on the Illumina GoldenGate BeadArray technology platform (Illumina, Inc., San Diego, CA, USA) or by using the SNP_TYPE assay in the 96.96 Dynamic Array with integrated fluidic circuit technology in a Fluidigm EP1 system (Standard BioTools Inc., South San Francisco, CA, USA), as described (Ebana et al. 2010). SNP genotyping was based on the manufacturers’ instructions.

Detection of putative QTLs

We determined the position of putative QTLs for respective agronomic traits in respective CSSLs by multiple linear regression analysis in R software (R Core Team 2018). Marker genotypes, values of agronomic traits, and environments (e.g., years and blocks within a field) of individual plants were used as variables to choose a model with the minimum root-mean-squared error (RMSE) by the leave-one-out cross-validation procedure. Variables whose partial regression coefficient differed significantly from 0 at the 8% level were accepted in the model. Coefficients with the highest variance inflation factor (VIF) where VIF >10 were excluded from the model to avoid multicollinearity. The model was refined until VIF of coefficients was <10. The final model was analyzed by multiple linear regression to determine the degree of contribution of coefficients under the conditions of VIF <10, P < 0.05, and the significance of the coefficient among bootstrap samples was P > 0.1. The calculations were conducted in software (“CSSL QTL Detector”) developed for this study. The software has a Japanese language interface specified for QTL detection in CSSLs; it runs under Microsoft Windows 10 and requires. NET Framework 4.6, R, and the DAAG (R) library. It provides an option for a model with the minimum Akaike’s Information Criterion (AIC) by stepwise procedures, and users can determine thresholds of some statistical parameters. The output is a list of basic parameters (AIC value, coefficient of determination, adjusted coefficient of determination, F-value, and P-value in the final model), a list of significant parameters (DNA marker loci, environmental factors such as test year and blocks within the field) with their degree of probability and their estimated effect, a table showing measured and estimated values in respective CSSLs and the recurrent parent, and regression plots obtained by multiple linear regression analysis. A table of putative QTLs, which includes marker genotypes of respective CSSLs and their average of the trait, is output in Microsoft Excel format.

Results

Genomic constitution of CSSLs

We developed 12 sets of CSSLs in japonica rice genetic backgrounds by using marker-assisted selection in >39 000 plants (Supplemental Table 1) and characterized their genomic constitution (Table 1). The number of backcrosses was mostly 4 or 5, but was 3 in 1.4% of the lines (SL2530, SL2531, SL2533, SL2605, SL2606, SL2615, SL3113) (Fig. 1). Each CSSL contained at least one substituted segment of a particular target chromosomal region from the donor and additional small segments in non-target regions in the ‘Koshihikari’ or ‘Hitomebore’ genetic background (Supplemental Table 3-1–3-12). We determined the size of the target donors’ chromosomal segments, non-target regions, and heterozygous regions from the donor in each set of CSSLs (Supplemental Table 3-1–3-12). These values varied among CSSLs (Table 1). The proportion of the recurrent genome in each CSSL, calculated from the size of fragments estimated from the marker genotypes as indicated in Supplemental Table 3-1–3-12, ranged between 89.0% (SIH, see Table 1 for cultivar codes) and 94.6% (QZK), averaging 92.1% (Table 1). The proportion of the chromosomal regions in which none of the CSSLs had the donor’s introgression ranged from 0.7% (NAK) to 6.2% (SIH), averaging 3.3% (Table 1).

Variation of agronomic traits

To clarify the phenotypic variation in CSSLs, we measured eight agronomic traits in the CSSLs, donors, and recurrent parents (Tables 2, 3, Supplemental Table 4-1–4-12). Their averages in a respective set of CSSLs were usually close to the values of the recurrent parents (Tables 2, 3). The range of phenotypic variation of the traits differed among the sets of CSSLs, being generally wider in the populations with indica donors, except in blast disease severity (Table 2). The range of phenotypic variation in CSSLs did not correlate with the degree of difference between parents except in culm length and grain width (Table 2).

Table 2. Ranges and averages of eight agronomic traits in each set of chromosome segment substitution lines

Code	Year	No. of lines evaluateda	Days to heading (day)					Culm length (cm)					Panicle length (cm)					Number of panicle					Fertility (%)					Grain length (mm)					Grain width (mm)				Year	No. of lines evaluated	Blast disease severity score
			Max	Min	Range (%)b	Av.		Max	Min	Range (%)	Av.		Max	Min	Range (%)	Av.		Max	Min	Range (%)	Av.		Max	Min	Range (%)	Av.		Max	Min	Range (%)	Av.		Max	Min	Range (%)	Av.			Max	Min	Range (%)	Av.
BKK	2012	41/43	135	95	40 (38)	109 ± 7.9		101	75	26 (30)	87.3 ± 7.60		23.9	17.1	6.8 (36)	20.4 ± 1.42		14.8	7.8	7.0 (53)	11.9 ± 1.43		98	40	58 (61)	90.7 ± 10.67		7.52	6.70	0.82 (12)	7.18 ± 0.18		3.55	3.03	0.52 (15)	3.39 ± 0.10	2015	43/43	8.8	3.5	5.3 (71)	7.4 ± 0.99
TUK	2012	48/48	128	98	30 (28)	108 ± 4.8		105	73	32 (37)	89.2 ± 6.94		23.6	17.4	6.2 (32)	20.2 ± 1.58		19.0	10.4	8.6 (65)	13.4 ± 1.74		98	81	17 (18)	94.1 ± 3.04		7.80	6.71	1.09 (15)	7.22 ± 0.22		3.58	3.07	0.51 (15)	3.43 ± 0.12	2015	48/48	8.1	3.2	4.8 (65)	6.7 ± 1.30
KNK	2012	39/40	130	105	25 (24)	108 ± 4.8		97	82	15 (17)	87.2 ± 3.61		23.3	18.0	5.3 (28)	20.0 ± 1.28		14.2	8.4	5.8 (44)	12.1 ± 1.23		97	89	8 ( 8)	94.2 ± 1.94		7.51	6.86	0.65 ( 9)	7.14 ± 0.16		3.63	3.30	0.33 (10)	3.46 ± 0.06	2013	40/40	7.9	3.9	4.0 (58)	6.8 ± 0.91
KMK	2012	46/46	130	104	26 (25)	108 ± 4.6		101	77	24 (27)	89.4 ± 4.17		21.9	17.1	4.8 (25)	19.5 ± 1.06		15.8	9.4	6.4 (48)	12.7 ± 1.26		97	83	14 (15)	93.7 ± 2.43		7.75	6.61	1.14 (16)	7.11 ± 0.20		3.64	3.25	0.39 (11)	3.42 ± 0.08	2013	45/46	8.5	3.0	5.4 (78)	7.2 ± 1.24
HAK	2012	38/40	121	92	29 (27)	106 ± 5.0		90	71	19 (21)	85.1 ± 4.25		22.2	17.7	4.5 (24)	19.1 ± 0.92		16.0	11.2	4.8 (36)	13.5 ± 1.28		98	67	30 (32)	93.8 ± 5.38		7.24	6.66	0.58 ( 8)	7.04 ± 0.12		3.46	3.22	0.24 ( 7)	3.39 ± 0.05	–	–	–	–	–	–
DPK	2013	35/36	132	101	31 (29)	111 ± 7.1		106	82	25 (27)	92.6 ± 6.71		22.5	17.9	4.6 (23)	19.9 ± 1.20		15.8	8.8	7.0 (58)	12.7 ± 1.63		99	91	8 ( 8)	96.5 ± 2.16		7.48	6.89	0.59 ( 8)	7.14 ± 0.15		3.57	2.98	0.59 (17)	3.39 ± 0.13	2014	36/36	7.7	4.2	3.4 (48)	6.8 ± 0.63
NAK	2013	40/41	136	91	45 (42)	108 ± 9.1		112	74	38 (42)	91.1 ± 7.68		23.3	17.3	6.0 (30)	20.4 ± 1.21		17.3	9.4	7.9 (65)	12.5 ± 1.73		100	89	11(12)	96.6 ± 2.12		7.61	6.93	0.68 (10)	7.17 ± 0.15		3.64	3.10	0.54 (16)	3.42 ± 0.13	2014	40/41	7.7	4.1	3.6 (50)	6.4 ± 1.03
BLK	2013	38/41	125	105	20 (19)	109 ± 5.1		114	84	31 (34)	94.9 ± 7.25		23.1	17.8	5.3 (26)	20.5 ± 1.30		21.5	10.0	11.5 (95)	13.6 ± 2.15		100	83	17 (18)	96.8 ± 3.36		8.01	6.89	1.12 (16)	7.18 ± 0.22		3.57	3.16	0.41 (12)	3.45 ± 0.10	2014	40/41	8.5	4.4	4.1 (56)	7.2 ± 0.66
QZK	2013	42/43	126	83	43 (41)	104 ± 7.9		113	59	54 (60)	88.9 ± 9.07		22.6	16.0	6.6 (33)	20.2 ± 1.38		16.6	8.8	7.8 (64)	12.6 ± 1.57		100	93	6 ( 6)	97.7 ± 1.50		7.48	6.88	0.60 ( 8)	7.19 ± 0.14		3.66	2.99	0.67 (19)	3.44 ± 0.13	2015	43/43	8.6	2.7	5.9 (79)	6.2 ± 1.81
MUK	2014	41/42	125	86	39 (39)	103 ± 7.5		123	76	47 (53)	95.5 ± 10.00		31.8	16.4	15.4 (73)	21.8 ± 3.06		17.0	9.4	7.6 (60)	12.3 ± 1.66		98	35	63 (67)	92.3 ± 9.55		7.56	6.67	0.89 (13)	7.16 ± 0.21		3.61	3.11	0.50 (14)	3.42 ± 0.09	2015	42/42	8.0	2.9	5.0 (68)	6.1 ± 1.34
SIH	2014	34/35	105	93	12 (13)	96 ± 2.6		103	81	21 (24)	89.9 ± 5.30		24.1	18.5	5.6 (27)	22.1 ± 1.29		16.0	9.8	6.2 (46)	13.3 ± 1.68		98	86	12 (13)	94.9 ± 2.60		7.81	7.14	0.67 ( 9)	7.38 ± 0.18		3.53	3.28	0.25 ( 7)	3.42 ± 0.06	–	–	–	–	–	–
BAK	2014	44/46	122	85	37 (37)	102 ± 5.5		116	79	37 (40)	94.2 ± 6.47		31.0	18.7	12.3 (58)	21.5 ± 2.26		16.2	9.8	6.4 (50)	13.3 ± 1.52		98	84	14 (15)	94.0 ± 2.49		7.60	6.83	0.77 (11)	7.16 ± 0.16		3.55	3.05	0.50 (14)	3.40 ± 0.10	2015	46/46	7.8	2.0	5.8 (77)	6.6 ± 1.36
Average (indica donors)					(34)					(41)					(39)					(94)					(26)					(12)					(15)						(64)
Average (japonica donors)					(22)					(22)					(26)					(44)					(17)					(11)					( 9)						(68)
Correlationc					0.04					–0.58					0.21					–0.32					0.06					0.35					0.52						0.35

^a Data of lines under development in the year of evaluation were excluded.

^b Values in parentheses show the ratio of the range of variation in a set of CSSLs to that of their parents.

^c Correlation between the range of variation in a set of CSSLs and that of their parents.

Table 3. Averages of eight agronomic traits of the recureent and donor parents of the chromosome segment substitution lines

Variety name		Year	Days to heading (day)	Culm length (cm)	Panicle length (cm)	Number of panicle	Fertility (%)	Grain length (mm)	Grain width (mm)	Year	Blast disease severity score
Koshihikari	(r)a	2012	106	86.9 ± 2.74	19.1 ± 0.61	13.3 ± 1.38	94.5 ± 1.03	7.13 ± 0.10	3.42 ± 0.05	2013	7.0 ± 0.46
		2013	106	89.9 ± 2.22	20.2 ± 0.82	12.1 ± 0.91	97.0 ± 1.01	7.07 ± 0.07	3.46 ± 0.03	2014	7.2 ± 0.42
		2014	101	93.0 ± 1.67	21.2 ± 0.93	12.7 ± 1.72	93.4 ± 2.14	7.11 ± 0.05	3.45 ± 0.04	2015	7.5 ± 0.47
Bei Khe	(d)	2012	128	122.7 ± 2.32	26.1 ± 0.55	12.5 ± 0.95	95.0 ± 0.59	8.41 ± 0.15	2.70 ± 0.06	2015	2.0 ± 0.07
Tupa121-3	(d)	2012	105	125.8 ± 3.71	24.5 ± 1.03	12.1 ± 1.40	95.3 ± 1.51	8.10 ± 0.16	2.69 ± 0.06	2015	2.8 ± 1.33
Khao Nam Jen	(d)	2012	165	185.2 ± 11.65	21.2 ± 0.73	5.4 ± 0.55	79.8 ± 7.01	7.48 ± 0.04	3.79 ± 0.06	2013	2.5 ± 1.80
Khau Mac Kho	(d)	2012	124	118.9 ± 3.56	22.6 ± 0.42	5.23 ± 0.61	88.3 ± 3.11	8.08 ± 0.04	3.92 ± 0.05	2013	1.1 ± 0.00
Hayamasari	(d)	2012	79	58.2 ± 1.63	17.7 ± 0.62	15.5 ± 1.81	97.5 ± 0.86	7.19 ± 0.05	3.39 ± 0.02	–	–
Deng Pao Zhai	(d)	2013	120	143.0 ± 5.83	23.4 ± 0.80	18.3 ± 1.89	94.6 ± 1.11	8.12 ± 0.07	3.15 ± 0.03	2014	1.5 ± 0.64
Naba	(d)	2013	130	108.1 ± 4.09	29.0 ± 1.38	10.1 ± 0.89	92.1 ± 1.22	8.49 ± 0.07	2.75 ± 0.01	2014	1.5 ± 0.65
Bleiyo	(d)	2013	118	147.4 ± 6.15	21.3 ± 1.96	10.4 ± 1.95	78.6 ± 5.77	9.27 ± 0.09	2.76 ± 0.06	2014	0.5 ± 0.65
Qiu Zhao Zong	(d)	2013	87	107.0 ± 1.76	21.1 ± 1.04	17.8 ± 5.31	93.2 ± 1.60	8.47 ± 0.07	3.01 ± 0.04	2015	2.7 ± 0.00
Muha	(d)	2014	101	125.3 ± 5.17	26.4 ± 0.92	11.6 ± 0.78	79.0 ± 2.65	8.62 ± 0.10	2.87 ± 0.02	2015	1.0 ± 0.00
Basilanon	(d)	2014	105	133.6 ± 3.26	24.5 ± 0.61	15.4 ± 1.31	72.6 ± 2.14	7.26 ± 0.05	2.53 ± 0.02	2015	2.5 ± 0.62
Hitomebore	(r)	2014	95	89.4 ± 3.64	21.1 ± 0.93	13.4 ± 2.43	95.6 ± 1.77	7.31 ± 0.02	3.37 ± 0.02	–	–
Silewah	(d)	2014	105	159.8 ± 3.55	32.8 ± 0.67	5.8 ± 0.74	88.6 ± 4.14	8.80 ± 0.15	3.54 ± 0.05	–	–

^a r, recurrent; d, donor.

Detection of putative QTLs in the CSSLs

Multiple linear regression analysis was performed in each CSSL to identify putative QTLs for seven traits and their chromosomal locations in the 12 sets of CSSLs detected 1079 putative QTLs, 75 to 217 per trait (Table 4, Supplemental Table 3-1–3-12). Ten or more putative QTLs were found in each set of CSSLs tested for grain length, grain width, and blast disease severity; number of panicles had the fewest. The average numbers of QTLs in the CSSLs of indica donors were larger than those of japonica donors in six of the seven traits, one of which (the difference in culm length) was significant at the 0.1% level (Table 4). The adjusted coefficient of determination obtained for respective trait/CSSLs combinations was highest in grain width, and was >0.80 in 11 of the 12 CSSL sets. In contrast, that in the number of panicles was the smallest in most CSSLs. The adjusted coefficient of determination of the CSSLs of indica donors was larger than that of the CSSLs of japonica donors in all seven traits, one of which (the difference in culm length) was significant at the 0.1% level (Table 4).

Table 4. Summary of multiple linear regression analysis for seven agronomic traits in 12 sets of CSSLs

Code	subspecies	Traita	Number of putative QTLs									Adjusted coefficient of determination
			CL	PL	NP	FR	GL	GW	BS	Total		CL	PL	NP	FR	GL	GW	BS
BAK	indica		22	8	5	13	24	14	16	102		0.86	0.72	0.28	0.43	0.86	0.83	0.77
BKK	indica		19	14	5	11	19	23	14	105		0.86	0.58	0.12	0.77	0.87	0.91	0.63
BLK	indica		16	12	8	10	17	18	17	98		0.85	0.41	0.39	0.96	0.91	0.86	0.87
DPK	indica		18	11	9	8	15	15	11	87		0.82	0.34	0.41	0.65	0.78	0.91	0.67
MUK	indica		19	10	8	5	19	21	18	100		0.86	0.73	0.16	0.08	0.83	0.84	0.90
NAK	indica		20	9	5	15	20	15	10	94		0.87	0.42	0.30	0.53	0.83	0.89	0.59
QZK	indica		14	10	10	10	16	15	14	89		0.87	0.34	0.33	0.57	0.83	0.89	0.88
TUK	indica		21	12	6	10	18	23	18	108		0.79	0.50	0.15	0.58	0.84	0.92	0.85
KMK	japonica		16	12	5	5	18	21	13	90		0.58	0.42	0.12	0.35	0.84	0.82	0.63
KNK	japonica		14	8	2	13	17	18	10	82		0.64	0.42	0.08	0.37	0.79	0.82	0.82
HAK	japonica		7	8	6	5	14	15	n.a.b	55		0.69	0.44	0.16	0.77	0.70	0.71	n.a.
SIH	japonica		11	9	6	7	17	19	n.a.	69		0.81	0.46	0.33	0.38	0.87	0.90	n.a.
Total			197	123	75	112	214	217	141	1079
Averages	12 CSSLs		16.4	10.3	6.3	9.3	17.8	18.1	14.1			0.79	0.48	0.24	0.54	0.83	0.86	0.76
	8 CSSLs (indica donors)		18.6	10.8	7.0	10.3	18.5	18.0	14.8			0.85	0.51	0.27	0.57	0.85	0.88	0.77
	4 CSSLs (japonica donors)		12.0	9.3	4.8	7.5	16.5	18.3	11.5			0.68	0.44	0.17	0.47	0.80	0.81	0.73
T-testc			***	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.			***	n.s.	n.s.	n.s.	n.s.	n.s.	n.s.

^a CL: culm length, PL: panicle length, NP: number of panicle, FR: fertility, GL: grain length, GW: grain width, BS: blast disease severity.

^b Not available.

^c Comparison of avarages in CSSLs of indica and japonica donors: n.s, not significant; ***, P < 0.1%.

To show the chromosomal distribution of putative QTLs for respective traits, Fig. 2 presents the number of QTLs per megabase (Mb) pairs in each chromosome sequence. Putative QTLs for the tested traits were found on all chromosomes; fewer on chromosomes (Chrs.) 4, 10, and 11 and more on Chrs. 6 and 12. Their mode of distribution among chromosomes differed by trait. Fig. 3 presents the chromosomal locations of putative QTLs for respective traits found in two or more donors. In regions where putative QTLs for the same trait were found in multiple donors and their ranges partly overlapped, the consensus region was determined as the position of each putative QTL when the donors’ alleles had the same direction as the recurrent parent’s allele. For example, eight consensus regions for grain length were found on Chr. 2 on the basis of 20 putative QTLs among 12 populations (Fig. 4). Using this criterion, the number of QTLs per 10 Mb was 29.9 across all chromosomes; ≥50 in bins from 0 to 10 Mb on Chr. 5 and 20 to 30 Mb on Chr. 7; and smallest (9.0) in the bin from 10 to 20 Mb on Chr. 8. Note that the bin from 30 to 40 Mb on Chr. 6, where the length of the introduced chromosome was <1.25 Mb, was excluded (Supplemental Table 5).

Fig. 2.

Chromosomal distribution of putative QTLs for seven agronomic traits, with number of putative QTLs per megabase on each chromosome.

Fig. 3.

Chromosomal locations of putative QTLs for seven agronomic traits found in two or more donors. Chromosome numbers are indicated above; scale to show position on chromosomes is indicated on left. Arrows indicate putative QTLs whose donor alleles ↑ increase or ↓ decrease values. Colors distinguish traits. The numbers on the arrows indicate the number of CSSLs with each putative QTL.

Fig. 4.

Locations of putative QTLs for grain length detected on chromosome 2. Codes of CSSLs are indicated above. Gray boxes indicate the most probable regions for putative QTLs based on the final model subjected to multiple linear regression analysis. Open boxes indicate regions that are partly interrupted. Numbers above and below gray boxes show additive effects of donor alleles on grain length in red (positive) or blue (negative). Red dashed rectangles indicate consensus regions (Cons.) among the 12 CSSLs; results are shown as red arrows to the right, together with the numbers of CSSLs that detected each putative QTL. Box on the right shows QTL regions found by Nagata et al. (2015) and the additive effects of the ‘IR64’ allele on grain length. IRK, BC₄F₂ population developed with ‘Koshihikari’ as the recurrent parent. KSI, BC₄F₂ population developed with ‘IR64’ as the recurrent parent.

Discussion

Multiple linear regression analyses using the 12 sets of CSSLs revealed more than 1000 putative QTLs for seven traits on all rice chromosomes. Our results clearly show that a set of CSSLs is a powerful research resource for uncovering diversity in the genetic control of agronomic traits in Asian cultivated rice. The procedures used here screened beneficial QTL alleles for various agronomic traits. As we used a leading commercial japonica cultivar as the genetic background, the QTL alleles are readily transferable into desirable genetic backgrounds in breeding programs targeting the japonica background.

A key point of this study is that QTL alleles from diverse donors were evaluated in a uniform genetic background. Previous studies used various donors, genetic backgrounds, and combinations, so it is difficult to compare their results. Huajingxian 74 is commonly used as a recurrent parent for CSSLs of indica background (Zhang 2021). In contrast, systematic efforts have not been reported for CSSLs of japonica background. We used donors representing the genetic diversity of Asian cultivated rice (Fukuoka et al. 2010, Kojima et al. 2005). Therefore, the number of distinct alleles of a putative QTL within a region would reflect diversity of alleles among donors (Fig. 4, Supplemental Table 3-1–3-12). When the recurrent parents and the donors are distantly related, generally more QTLs are found in CSSLs. To determine the distribution of putative QTLs by chromosome, we obtained differences in mode of distribution and variations among agronomic traits, information rarely obtained from a small number of biparental QTL mapping populations (Fig. 2). Our study assigned QTL clusters to regions that were not reported in a previous curation study (Yonemaru et al. 2010), such as 20–29.7 Mb on Chr. 7 and 0–10 Mb on Chr. 5. Our study gives researchers a comprehensive view of the genetic control of agronomic traits in Asian cultivated rice, more specifically in the japonica genetic background.

Our results show consensus regions for respective QTLs that accord with those reported in previous QTL mapping studies, as represented by grain length on Chr. 2 (Fig. 4). Meta-analysis of QTLs allows us to find stable and reliable QTLs and delimit their confidence intervals (Goffinet and Gerber 2000, Swamy and Sarla 2011). We also found new QTLs that were shared among CSSLs with different donors. For example, a QTL for grain length at 27–29 Mb on Chr. 7 and one for grain length and grain width at 3 Mb on Chr. 10 were located in regions where QTLs for grain shape have not been previously reported (Huang et al. 2013, Jiang et al. 2022, Nagata et al. 2015). QTLs for blast resistance were found in regions where few QTLs have been reported, such as on Chrs. 5 and 7 (Ballini et al. 2008, Fukuoka 2018). Importantly, such evidence cannot be easily obtained from biparental mapping populations owing to background noise of other genes or QTLs for blast resistance. Despite the limitation of mapping resolution in the use of CSSLs, our procedures provide preliminary information on target regions for marker-assisted selection of complex agronomic traits.

Multiple regression analysis allowed us to estimate multiple QTLs in an introduced chromosomal segment from a specific donor (Xu et al. 2010, Zhou et al. 2016). For example, we found two close QTLs for grain length with opposite directions of additive effect around 4 Mb on Chr. 1 (donor: MUK). Similarly, QTLs for grain length and culm length at 38–44 Mb on Chr. 1 (donors: BLK and BAK), for culm length (donor: BAK) or seed fertility (donor: QZK) at 32 Mb on Chr. 4, and for culm length at 2–6 Mb on Chr. 6 (donor: KMK) and at 20 Mb on Chr. 9 (donor: NAK) had additive effects in opposite allelic directions (Supplemental Table 3-1–3-12). Such observations support the idea that the genetic basis of QTL clusters can be explained by tight linkage of genes for various agronomic traits (Luo et al. 2013), as well as by pleiotropy of specific genes (Xue et al. 2008). However, this procedure has low discriminability in selecting the most probable region owing to unbalanced allelic frequency (lower in donor’s allele) and a small number of genotypes, as well as the possibility of overestimating or underestimating the number of putative QTLs. Nevertheless, the analysis highlighted chromosomal regions that should be intensively investigated, and progeny testing using advanced mapping populations will deny or confirm the hypothesis.

We used 539 per population, and an average of 450 markers (Supplemental Table 3-1–3-12), so genotypes of CSSLs were determined more precisely than in previous CSSLs in japonica genetic backgrounds (Abe et al. 2013, Ando et al. 2008, Ebitani et al. 2005, Hori et al. 2010, Ishikawa et al. 2005, Kato and Hirayama 2021, Kubo et al. 2002, Mizuno et al. 2018, Murata et al. 2014, Nagata et al. 2015, Takai et al. 2009). Still, failure to get rid of small chromosomal fragments from the donor in the non-target region(s) or unexpected dropout of the donor’s chromosomal fragment in the target chromosome would decrease the accuracy of prediction of putative QTLs. By using a high-throughput genotyping method such as genotyping-by-sequencing to increase the number of markers (Arbelaez et al. 2015, Fan et al. 2022, Xu et al. 2010) so that the position of the terminal region of the substituted segment can be accurately known, the prediction accuracy can be improved.

In conclusion, we developed 12 sets of CSSLs with japonica rice genetic backgrounds. Their substituted chromosomal segments containing the genetic diversity of Asian cultivated rice were well characterized by SNP markers. Although the CSSLs have a low resolution for QTL mapping owing to a low recombination frequency per chromosome, we estimated many QTLs (with various magnitudes of effect and gene action in each region) by combining comparisons among multiple populations and with multiple regression analyses. In addition, during the development of the CSSLs, we obtained seeds of BC₄F₂ populations in which a particular chromosomal segment was segregating. Progeny testing with segregating BC₄F₂ populations can determine QTL effects and allow further fine mapping, since QTLs of interest may behave as single Mendelian factors. Thus, the CSSLs and advanced mapping populations will be useful for rice researchers and breeders.

Author Contribution Statement

MY and SF conceived and designed the experiments. KN, YN and the other authors developed or determined genotypes. KN, RM and SF phenotyped agronomic traits. TM developed the “CSSL QTL Detector” program. KN analyzed the data. KN, KM and SF wrote the paper. All authors have reviewed drafts of the paper and approved the final draft.

Acknowledgments

We are grateful to the technical staff of the Institute of Crop Science for their technical assistance and management of the rice fields at NARO. We are grateful to Dr. Sunhee Choi for the assistance in field sampling. This work was supported by a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan (NVR0001, IVG2003). We thank two editors from ELSS, Inc. (https://elss.co.jp/en/) for editing our manuscript before submission. The CSSLs can be provided under a material transfer agreement from the Institute of Crop Science, NARO.

Literature Cited

•  Abe,  T.,  Y.  Nonoue,  N.  Ono,  M.  Omoteno,  M.  Kuramata,  S.  Fukuoka,  T.  Yamamoto,  M.  Yano and  S.  Ishikawa (2013) Detection of QTLs to reduce cadmium content in rice grains using LAC23/Koshihikari chromosome segment substitution lines. Breed Sci 63: 284–291.
•  Ando,  T.,  T.  Yamamoto,  T.  Shimizu,  X.F.  Ma,  A.  Shomura,  Y.  Takeuchi,  S.Y.  Lin and  M.  Yano (2008) Genetic dissection and pyramiding of quantitative traits for panicle architecture by using chromosomal segment substitution lines in rice. Theor Appl Genet 116: 881–890.
•  Arbelaez,  J.D.,  L.T.  Moreno,  N.  Singh,  C.W.  Tung,  L.G.  Maron,  Y.  Ospina,  C.P.  Martinez,  C.  Grenier,  M.  Lorieux and  S.  McCouch (2015) Development and GBS-genotyping of introgression lines (ILs) using two wild species of rice, O. meridionalis and O. rufipogon, in a common recurrent parent, O. sativa cv. Curinga. Mol Breed 35: 81.
•  Asaga,  K. (1976) Evaluation standard of rice leaf blast in nursery test. J Agric Sci 31: 156–156 (in Japanese).
•  Balakrishnan,  D.,  M.  Surapaneni,  S.  Mesapogu and  S.  Neelamraju (2019) Development and use of chromosome segment substitution lines as a genetic resource for crop improvement. Theor Appl Genet 132: 1–25.
•  Ballini,  E.,  J.B.  Morel,  G.  Droc,  A.  Price,  B.  Courtois,  J.L.  Notteghem and  D.  Tharreau (2008) A genome-wide meta-analysis of rice blast resistance genes and quantitative trait loci provides new insights into partial and complete resistance. Mol Plant Microbe Interact 21: 859–868.
•  Bian,  J.M.,  L.  Jiang,  L.L.  Liu,  X.J.  Wei,  Y.H.  Xiao,  L.J.  Zhang,  Z.G.  Zhao,  H.Q.  Zhai and  J.M.  Wan (2010) Construction of a new set of rice chromosome segment substitution lines and identification of grain weight and related traits QTLs. Breed Sci 60: 305–313.
•  Chen,  Q.,  J.  Mu,  H.  Zhou and  S.  Yu (2007) Genetic effect of japonica alleles detected in indica candidate introgression lines. Scientia Agricultura Sinica 40: 2379–2387.
•  Cu,  S.T.,  N.I.  Warnock,  J.  Pasuquin,  M.  Dingkuhn and  J.  Stangoulis (2021) A high-resolution genome-wide association study of the grain ionome and agronomic traits in rice Oryza sativa subsp. indica. Sci Rep 11: 19230.
•  Ebana,  K.,  Y.  Kojima,  S.  Fukuoka,  T.  Nagamine and  M.  Kawase (2008) Development of mini core collection of Japanese rice landrace. Breed Sci 58: 281–291.
•  Ebana,  K.,  J.  Yonemaru,  S.  Fukuoka,  H.  Iwata,  H.  Kanamori,  N.  Namiki,  N.  Nagasaki and  M.  Yano (2010) Genetic structure revealed by a whole-genome single-nucleotide polymorphism survey of diverse accessions of cultivated Asian rice (Oryza sativa L.). Breed Sci 60: 390–397.
•  Ebitani,  T.,  Y.  Takeuchi,  Y.  Nonoue,  T.  Yamamoto,  K.  Takeuchi and  M.  Yano (2005) Construction and evaluation of chromosome segment substitution lines carrying overlapping chromosome segments indica rice cultivar ‘Kasalath’ in a genetic background of japonica elite cultivar ‘Koshihikari’. Breed Sci 55: 65–73.
•  Fan,  J.,  H.  Hua,  Z.  Luo,  Q.  Zhang,  M.  Chen,  J.  Gong,  X.  Wei,  Z.  Huang,  X.  Huang and  Q.  Wang (2022) Whole-genome sequencing of 117 chromosome segment substitution lines for genetic analyses of complex traits in rice. Rice (N Y) 15: 5.
•  Fujino,  K. and  H.  Sekiguchi (2005) Identification of QTLs conferring genetic variation for heading date among rice varieties at the northern-limit of rice cultivation. Breed Sci 55: 141–146.
• Fukuoka, S. (2018) Marker-Assisted Gene Pyramiding for Durable Resistance to Blast. In: Sasaki, T. and M. Ashikari (eds.) Rice Genomics, Genetics and Breeding. Springer, Singapore, pp. 393–415.
•  Fukuoka,  S.,  Y.  Nonoue and  M.  Yano (2010) Germplasm enhancement by developing advanced plant materials from diverse rice accessions. Breed Sci 60: 509–517.
•  Goffinet,  B. and  S.  Gerber (2000) Quantitative trait loci: A meta-analysis. Genetics 155: 463–473.
•  Hao,  W.,  M.Z.  Zhu,  J.P.  Gao,  S.Y.  Sun and  H.X.  Lin (2009) Identification of quantitative trait loci for rice quality in a population of chromosome segment substitution lines. J Integr Plant Biol 51: 500–512.
•  Hori,  K.,  K.  Sugimoto,  Y.  Nonoue,  N.  Ono,  K.  Matsubara,  U.  Yamanouchi,  A.  Abe,  Y.  Takeuchi and  M.  Yano (2010) Detection of quantitative trait loci controlling pre-harvest sprouting resistance by using backcrossed populations of japonica rice cultivars. Theor Appl Genet 120: 1547–1557.
•  Hori,  K.,  Y.  Nonoue,  N.  Ono,  T.  Shibaya,  K.  Ebana,  K.  Matsubara,  E.  Ogiso-Tanaka,  T.  Tanabata,  K.  Sugimoto,  F.  Taguchi-Shiobara et al. (2015) Genetic architecture of variation in heading date among Asian rice accessions. BMC Plant Biol 15: 115.
•  Huang,  R.,  L.  Jiang,  J.  Zheng,  T.  Wang,  H.  Wang,  Y.  Huang and  Z.  Hong (2013) Genetic bases of rice grain shape: So many genes, so little known. Trends Plant Sci 18: 218–226.
•  Ishikawa,  S.,  N.  Ae and  M.  Yano (2005) Chromosomal regions with quantitative trait loci controlling cadmium concentration in brown rice (Oryza sativa). New Phytol 168: 345–350.
•  Jiang,  H.,  A.  Zhang,  X.  Liu and  J.  Chen (2022) Grain size associated genes and the molecular regulatory mechanism in rice. Int J Mol Sci 23: 3169.
•  Kato,  K. and  Y.  Hirayama (2021) Development and characterization of chromosome segment substitution lines derived from backcross between japonica donor rice cultivar Yukihikari and japonica recipient cultivar Kirara397. Breed Sci 71: 283–290.
•  Kojima,  Y.,  K.  Ebana,  S.  Fukuoka,  T.  Nagamine and  M.  Kawase (2005) Development of an RFLP-based rice diversity research set of germplasm. Breed Sci 55: 431–440.
•  Korte,  A. and  A.  Farlow (2013) The advantages and limitations of trait analysis with GWAS: A review. Plant Methods 9: 29.
•  Kubo,  T.,  Y.  Aida,  K.  Nakamura,  H.  Tunematsu,  K.  Doi and  A.  Yoshimura (2002) Reciprocal chromosome segment substitution series derived from Japonica and Indica cross of rice (Oryza sativa L.). Breed Sci 52: 319–325.
•  Li,  J.,  H.  Yang,  G.  Xu,  K.  Deng,  J.  Yu,  S.  Xiang,  K.  Zhou,  Q.  Zhang,  R.  Li,  M.  Li et al. (2022) QTL analysis of Z414, a chromosome segment substitution line with short, wide grains, and substitution mapping of qGL11 in rice. Rice (N Y) 15: 25.
•  Liang,  P.,  H.  Wang,  Q.  Zhang,  K.  Zhou,  M.  Li,  R.  Li,  S.  Xiang,  T.  Zhang,  Y.  Ling,  Z.  Yang et al. (2021) Identification and pyramiding of QTLs for rice grain size based on short-wide grain CSSL-Z563 and fine-mapping of qGL3-2. Rice (N Y) 14: 35.
•  Luo,  X.,  S.  Ji,  P.  Yuan,  H.  Lee,  D.  Kim,  S.  Balkunde,  J.  Kang and  S.  Ahn (2013) QTL mapping reveals a tight linkage between QTLs for grain weight and panicle spikelet number in rice. Rice (N Y) 6: 33.
•  Mizuno,  Y.,  U.  Yamanouchi,  T.  Hoshino,  Y.  Nonoue,  K.  Nagata,  S.  Fukuoka,  T.  Ando,  M.  Yano and  K.  Sugimoto (2018) Genetic dissection of pre-harvest sprouting resistance in an upland rice cultivar. Breed Sci 68: 200–209.
•  Murata,  K.,  Y.  Iyama,  T.  Yamaguchi,  H.  Ozaki,  Y.  Kidani and  T.  Ebitani (2014) Identification of a novel gene (Apq1) from the indica rice cultivar ‘Habataki’ that improves the quality of grains produced under high temperature stress. Breed Sci 64: 273–281.
•  Murray,  M.G. and  W.F.  Thompson (1980) Rapid isolation of high molecular weight plant DNA. Nucleic Acids Res 8: 4321–4326.
•  Nagata,  K.,  T.  Ando,  Y.  Nonoue,  T.  Mizubayashi,  N.  Kitazawa,  A.  Shomura,  K.  Matsubara,  N.  Ono,  R.  Mizobuchi,  T.  Shibaya et al. (2015) Advanced backcross QTL analysis reveals complicated genetic control of rice grain shape in a japonica × indica cross. Breed Sci 65: 308–318.
•  Ookawa,  T.,  R.  Aoba,  T.  Yamamoto,  T.  Ueda,  T.  Takai,  S.  Fukuoka,  T.  Ando,  S.  Adachi,  M.  Matsuoka,  T.  Ebitani et al. (2016) Precise estimation of genomic regions controlling lodging resistance using a set of reciprocal chromosome segment substitution lines in rice. Sci Rep 6: 30572.
• R Core Team (2018) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/.
•  Satake,  T. (1979) Two extremely cool-tolerant varieties. Intern Rice Res Newsl 4: 9–10.
•  Shen,  X. and  Ö.  Carlborg (2013) Beware of risk for increased false positive rates in genome-wide association studies for phenotypic variability. Front Genet 4: 93.
•  Sun,  S.,  Z.  Wang,  S.  Xiang,  M.  Lv,  K.  Zhou,  J.  Li,  P.  Liang,  M.  Li,  R.  Li,  Y.  Ling et al. (2022) Identification, pyramid, and candidate gene of QTL for yield-related traits based on rice CSSLs in indica Xihui18 background. Mol Breed 42: 19.
•  Swamy,  B.P.M. and  N.  Sarla (2011) Meta-analysis of yield QTLs derived from inter-specific crosses of rice reveals consensus regions and candidate genes. Plant Mol Biol Rep 29: 663–680.
•  Takai,  T.,  A.  Ohsumi,  Y.  San-oh,  M.R.C.  Laza,  M.  Kondo,  T.  Yamamoto and  M.  Yano (2009) Detection of a quantitative trait locus controlling carbon isotope discrimination and its contribution to stomatal conductance in japonica rice. Theor Appl Genet 118: 1401–1410.
•  Tanabata,  T.,  T.  Shibaya,  K.  Hori,  K.  Ebana and  M.  Yano (2012) SmartGrain: High-throughput phenotyping software for measuring seed shape through image analysis. Plant Physiol 160: 1871–1880.
•  Tanaka,  N.,  M.  Shenton,  Y.  Kawahara,  M.  Kumagai,  H.  Sakai,  H.  Kanamori,  J.  Yonemaru,  S.  Fukuoka,  K.  Sugimoto,  M.  Ishimoto et al. (2020) Whole-genome sequencing of the NARO World Rice Core Collection (WRC) as the basis for diversity and association studies. Plant Cell Physiol 61: 922–932.
•  Uga,  Y.,  Y.  Kitomi,  E.  Yamamoto,  N.  Kanno,  S.  Kawai,  T.  Mizubayashi and  S.  Fukuoka (2015) A QTL for root growth angle on rice chromosome 7 is involved in the genetic pathway of DEEPER ROOTING 1. Rice (N Y) 8: 8.
•  Ujiie,  K.,  T.  Kashiwagi and  K.  Ishimaru (2012) Identification and functional analysis of alleles for productivity in two sets of chromosome segment substitution lines of rice. Euphytica 187: 325–337.
•  Vaughan,  D.A.,  B.R.  Lu and  N.  Tomooka (2008a) The evolving story of rice evolution. Plant Sci 174: 394–408.
•  Vaughan,  D.A.,  B.R.  Lu and  N.  Tomooka (2008b) Was Asian rice (Oryza sativa) domesticated more than once? Rice (N Y) 1: 16–24.
•  Wang,  D.,  K.  Zhou,  S.  Xiang,  Q.  Zhang,  R.  Li,  M.  Li,  P.  Liang,  N.  Farkhanda,  G.  He,  Y.  Ling et al. (2021) Identification, pyramid and candidate genes of QTLs for associated traits based on a dense erect panicle rice CSSL-Z749 and five SSSLs, three DSSLs and one TSSL. Rice (N Y) 14: 55.
•  Wu,  G.,  H.  Deng,  M.  Yu,  Y.  Cai,  D.  Zhou,  J.  Tan,  J.  Yu,  X.  Luo,  S.  Tong,  P.  Wang et al. (2020) Genetic analysis of rice seed recovery under low-temperature conditions using a new CSSL population with a high-density genetic map in rice. Mol Breed 40: 109.
•  Xi,  Z.,  F.  He,  R.  Zeng,  Z.  Zhang,  X.  Ding,  W.  Li and  G.  Zhang (2006) Development of a wide population of chromosome single-segment substitution lines in the genetic background of an elite cultivar of rice (Oryza sativa L.). Genome 49: 476–484.
•  Xu,  J.,  Q.  Zhao,  P.  Du,  C.  Xu,  B.  Wang,  Q.  Feng,  Q.  Liu,  S.  Tang,  M.  Gu,  B.  Han et al. (2010) Developing high throughput genotyped chromosome segment substitution lines based on population whole-genome re-sequencing in rice (Oryza sativa L.). BMC Genomics 11: 656.
•  Xue,  W.,  Y.  Xing,  X.  Weng,  Y.  Zhao,  W.  Tang,  L.  Wang,  H.  Zhou,  S.  Yu,  C.  Xu,  X.  Li et al. (2008) Natural variation in Ghd7 is an important regulator of heading date and yield potential in rice. Nat Genet 40: 761–767.
•  Yamamoto,  T.,  H.  Lin,  T.  Sasaki and  M.  Yano (2000) Identification of heading date quantitative trait locus Hd6 and characterization of its epistatic interactions with Hd2 in rice using advanced backcross progeny. Genetics 154: 885–891.
•  Yamamoto,  T.,  J.  Yonemaru and  M.  Yano (2009) Towards the understanding of complex traits in rice: Substantially or superficially? DNA Res 16: 141–154.
•  Yano,  M. and  T.  Sasaki (1997) Genetic and molecular dissection of quantitative traits in rice. Plant Mol Biol 35: 145–153.
•  Yasui,  H.,  Y.  Yamagata and  A.  Yoshimura (2010) Development of chromosome segment substitution lines derived from indica rice donor cultivars DV85 and ARC10313 in the genetic background of japonica cultivar Taichung 65. Breed Sci 60: 620–628.
•  Yonemaru,  J.,  T.  Yamamoto,  S.  Fukuoka,  Y.  Uga,  K.  Hori and  M.  Yano (2010) Q-TARO: QTL annotation rice online database. Rice (N Y) 3: 194–203.
•  Yuan,  H.,  P.  Gao,  X.  Hu,  M.  Yuan,  Z.  Xu,  M.  Jin,  W.  Song,  S.  Zhan,  X.  Zhu,  B.  Tu et al. (2022) Fine mapping and candidate gene analysis of qGSN5, a novel quantitative trait locus coordinating grain size and grain number in rice. Theor Appl Genet 135: 51–64.
•  Zeng,  R.,  Z.  Zhang,  F.  He,  Z.  Xi,  A.  Talukdar,  J.  Shi,  L.  Qin,  C.  Huang and  G.  Zhang (2006) Identification of multiple alleles at the Wx locus and development of single segment substitution lines for the alleles in rice. Rice Sci (N Y) 13: 9–14.
•  Zhang,  B.,  L.  Shang,  B.  Ruan,  A.  Zhang,  S.  Yang,  H.  Jiang,  C.  Liu,  K.  Hong,  H.  Lin,  Z.  Gao et al. (2019) Development of three sets of high-throughput genotyped rice chromosome segment substitution lines and QTL mapping for eleven traits. Rice (N Y) 12: 33.
•  Zhang,  G. (2021) Target chromosome-segment substitution: A way to breeding by design in rice. Crop J 9: 658–668.
•  Zhou,  Y.,  G.  Dong,  Y.  Tao,  C.  Chen,  B.  Yang,  Y.  Wu,  Z.  Yang,  G.  Liang,  B.  Wang and  Y.  Wang (2016) Mapping quantitative trait loci associated with toot traits using sequencing-based genotyping chromosome segment substitution lines derived from 9311 and Nipponbare in rice (Oryza sativa L.). PLoS One 11: e0151796.
•  Zhu,  W.,  J.  Lin,  D.  Yang,  L.  Zhao,  Y.  Zhang,  Z.  Zhu,  T.  Chen and  C.  Wang (2009) Development of chromosome segment substitution lines derived from backcross between two sequenced rice cultivars, indica recipient 93-11 and japonica donor Nipponbare. Plant Mol Biol Rep 27: 126–131.
•  Zong,  Y.,  Z.  Chen,  J.B.  Innes,  C.  Chen,  Z.  Wang and  H.  Wang (2007) Fire and flood management of coastal swamp enabled first rice paddy cultivation in east China. Nature 449: 459–462.