Identification and characterization of stable QTLs for vascular bundle number at the panicle neck in rice (Oryza sativa L.)

Abstract

A large vascular bundle number (VBN) in the panicle neck in rice (Oryza sativa L.) is related to the ability to transport assimilates from stem and leaf to reproductive organs during seed maturation. Several quantitative trait loci (QTLs) for VBN have been identified by using segregating populations derived from a cross between indica and japonica rice cultivars. However, the detailed location, effect, and interaction of QTLs for VBN were not understood well. Here, to elucidate the genetic basis of VBN, we identified three stable QTLs for VBN—qVBN5, qVBN6 and qVBN11—by using 71 recombinant inbred lines derived from a cross between indica ‘IR24’ and japonica ‘Asominori’. We confirmed their positions and characterized their effects by using chromosome segment substitution lines (CSSLs) with an ‘IR24’ genetic background. qVBN6 had the most substantial effect on VBN, followed by qVBN11 and qVBN5. We developed pyramided lines carrying two QTLs for VBN to estimate their interaction. The combination of qVBN6 and qVBN11 accumulated VBN negatively in the pyramided lines owing to the independent actions of each QTL. The QTLs detected for VBN will enhance our understanding of genetic mechanisms of VBN and can be used in rice breeding.

Introduction

Rice (Oryza sativa L.), an essential staple crop for more than half of the world’s population, is grown mainly in East, Southeast, and South Asia. Owing to the continuously increasing population size and loss of arable land resulting from land degradation and salinization of irrigated areas, increasing rice production per unit area is a major goal of breeders (Khush 1997). Grain yield is determined by four yield components: spikelet number per panicle, panicle number per plant, grain weight, and grain fertility. To increase yield potential, increasing the source size, sink size, and translocation capability are crucial (Donald 1968, Lee and Chae 2000). The vascular system plays an important role in transporting photosynthesis products from source to sink, and its transport capacity influences grain yield (Cui et al. 2003, Fukuyama et al. 1999). In the development of hybrid rice, a small vascular bundle number (VBN) in the panicle neck restricts grain filling by limiting the transport of assimilates to the sink (Peng et al. 1999, Xu et al. 2005).

Vascular bundles interconnect all parts of a plant and are responsible for transporting photosynthesis products, minerals, and water around (Lucas et al. 2013). Rice has two types of vascular bundles: small ones in the rudimentary glumes of spikelets, and large ones that play a major role in nutrient uptake and ripening rate, which in turn affect grain weight (Chaudhry and Nagato 1970). Wide genetic diversity among rice cultivars underlies the number of large vascular bundles in the panicle neck; the VBN of indica rice tends to be greater than that of japonica rice (Fukuyama and Takayama 1995, Fukuyama et al. 1999, Liu et al. 2016, Zhai et al. 2018); and the ratio of VBN to primary branch number differs between them (Fukuyama and Takayama 1995, Fukuyama et al. 1999).

VBN is influenced mainly by inherited factors but also by environmental factors such as nitrogen availability and plant density. Through the use of segregating populations such as recombinant inbred lines (RILs) and double haploid lines (DHs), genetic factors that control VBN in the panicle neck of O. sativa have been detected as quantitative trait loci (QTLs) (Bai et al. 2012, Cui et al. 2003, Sasahara et al. 1999, Zhang et al. 2002); and genome-wide association studies (GWAS) have identified QTLs for VBN (Liao et al. 2021, Zhai et al. 2018). Some genes underlying QTLs for VBN have been isolated to their chromosomal regions and cloned, and include ABERRANT PANICLE ORGANIZATION 1 (APO1) on chromosome (Chr.) 6 (Terao et al. 2010), NARROW LEAF 1 (NAL1) on Chr. 4 (Fujita et al. 2013, Qi et al. 2008), and LVB9/DENSE AND ERECT PANICLE 1 (DEP1) on Chr. 9 (Fei et al. 2019).

In this study, RILs derived from a cross between indica rice ‘IR24’ and japonica rice ‘Asominori’ were used to elucidate the genetic basis of VBN. Chromosome segment substitution lines (CSSLs) can be used for developing materials for confirming QTL and for evaluating QTL epistasis through pyramiding (Yamamoto et al. 2009). Each detected QTL was confirmed in CSSLs carrying ‘Asominori’ chromosomal segments in the ‘IR24’ genetic background. QTL analysis of VBN was conducted using RILs, and each QTL’s phenotypic effects were verified and validated using CSSLs. To understand interaction of QTLs, the effects of pyramiding pair of the QTLs were evaluated.

Materials and Methods

Plant materials

For QTL analysis in 2016, 2017, and 2018, we used 71 RILs derived from a cross between ‘Asominori’ and ‘IR24’ followed by the single-seed-descent method (Tsunematsu et al. 1996). For evaluating the effect of QTLs in 2018 and 2019, we used three CSSLs with the ‘IR24’ genetic background—IAS30, IAS39, and IAS14—carrying target substitution chromosomal segments of ‘Asominori’ (Kubo et al. 2002). For QTL confirmation in 2019, we used three F₂ populations derived from a cross between each of those CSSLs and ‘IR24’ (168 individuals each). To characterize the gene interactions among the detected QTLs, we developed and characterized pyramided lines (PYLs) each carrying two QTLs in the ‘IR24’ genetic background: we developed three F₁ plants from crosses between IAS30 (carrying qVBN5) and IAS39 (carrying qVBN6), IAS30 and IAS14 (carrying qVBN11), and IAS39 and IAS14. These F₁ plants were self-pollinated to develop F₂ populations of 96 plants each. PYLs were selected from each F₂ population by marker-assisted selection: PYLs 1–4 from IAS30/IAS14, PYLs 5–7 from IAS39/IAS14, and PYLs 8–10 from IAS30/IAS39. These 10 PYLs were self-pollinated, and the effects of pyramiding were evaluated in 24 plants of each F₃ line in 2020. For evaluating the effect of QTLs in 2017, 2018 and 2019, we used three CSSLs with the ‘Asominori’ genetic background—AIS38, AIS49, and AIS76—carrying target substitution chromosomal segments of ‘IR24’ (Kubo et al. 2002).

Plant growth condition and evaluation of VBN

Plants were grown in a paddy field of Saga University (33°14ʹ32ʺN 130°17ʹ24ʺE). At 28 days after sowing, seedlings were transplanted at one plant per hill with 20 cm between hills and 25 cm between rows. Inorganic fertilizer was applied at 40 kg/ha N, 17.5 kg/ha P, and 33 kg/ha K. At 2 to 3 weeks after heading, the tallest panicles among the individual plants were collected for counting of VBN in the panicle neck. Fresh peduncles were sliced at about 1 cm below the panicle base node, and the large vascular bundles were counted under a microscope, in 5 stems each of RILs and PYLs, 10 of CSSLs, and every plant in the F₂ populations.

Extracting DNA and genotyping

Approximately 2–4 cm of leaves was collected directly and freeze-dried for 48 h. Total DNA was extracted by the potassium acetate method (Dellaporta et al. 1983). Polymorphic SSR markers (McCouch et al. 2002) and indel markers (Yonemaru et al. 2015) were used for genotyping of segregating populations; these co-dominant markers distinguish between the two parental lines. Polymerase chain reaction (PCR) was performed using GoTaq master mix (Promega) at 96°C for 5 min; 35 cycles of 30 s at 96°C, 30 s at 55°C, and 30 s at 72°C; and a final at 25°C for 1 min. The amplified PCR products were run in 4% agarose gel at 200 V with 0.5 μg/mL ethidium bromide in 0.5 × TBE buffer for 60–120 min.

DNA markers

QTLs in RILs were found using the genotyping data from RFLP markers each year (Tsunematsu et al. 1996). QTL analysis for VBN in each population used SSR markers located around the substituted chromosome locations in the CSSLs. We selected 15 SSR/indel markers on Chr. 5 around RFLP marker interval C128–R2117 (22.6–23.4 Mbp) for genotyping the IR24/IAS30 F₂ population; 7 markers on Chr. 6 around C962–Ky11 (~28.6 Mbp) for genotyping the IR24/IAS39 F₂; and 12 markers on Chr. 11 around R2918–C3029A (~2.2–2.4 Mbp) for genotyping the IR24/IAS14 F₂ (Supplemental Table 1).

QTL analysis

We performed composite interval mapping in Windows QTL Cartographer v. 2.5 software (Wang et al. 2012) to identify QTLs for VBN. The critical threshold values of logarithm of odds (LOD) that were calculated by 1000 permutation tests with significant level at P < 0.05 on each population were 3.15 in 2016, 3.23 in 2017 and 3.10 in 2018 for RILs and 2.07 for IAS30/IR24 F₂ population, 1.67 for IAS14/IR24 F₂ population and 1.89 for IAS39/IR24 F₂ population.

Statistical analysis

One-way ANOVA assessed the phenotypic differences among parents, CSSLs carrying single, and PYLs carrying paired QTLs. Tukey Kramer’s test was performed for multiple comparison of VBN among parents, CSSLs, and PYLs and Dunnett’s test was used to compare VBN between ‘Asominori’ and CSSLs using the R software, version 3.5.2.

Results

Identification of QTLs for VBN in RILs

The VBN of ‘Asominori’ was 10.3–11.8 in all 3 years, and that of ‘IR24’ was 20.4–24.8 (Fig. 1). Those of the RILs were 10–23 in 2016, 11–23 in 2017, and 10–21 in 2018. The continuous frequency distributions of VBN in the RILs imply that multiple genetic factors control VBN.

Fig. 1.

Frequency distributions of vascular bundle number in panicle neck in RILs in (A) 2016, (B) 2017, (C) 2018. Bars indicate means in parents with standard deviation.

Four QTLs for VBN were detected: qVBN5 and qVBN6 in 2016, 2017, and 2018; qVBN11 in 2016 and 2017; and qVBN4 in 2018 (Table 1). qVBN5 had PVE of 10.3%–28.6%, qVBN6 of 13.6%–18.3%, qVBN11 of 11.8%–24.0%, and qVBN4 of 10.4%. The ‘Asominori’ alleles at qVBN5, qVBN6, and qVBN11 decreased VBN, whereas that at qVBN4 increased VBN.

Table 1. Detection of QTLs for vascular bundle number in RILs derived from IR24/Asominori

Year	QTL	Chr.	Interval marker	LOD	Additive effecta	R2 (%)
2016	qVBN5	5	C128–R2117	4.6	–1.27	16.3
	qVBN6	6	C962–Ky11	4.2	–1.31	16.3
	qVBN11	11	R2918–C3029A	3.4	–1.14	11.8
2017	qVBN5	5	C128–R2117	6.9	–1.67	28.6
	qVBN6	6	Xnpb135–C962	3.8	–1.21	13.6
	qVBN11	11	C794A–C83B	6.3	–2.19	24.0
2018	qVBN4	4	C335–C621B	3.5	0.84	10.4
	qVBN5	5	C128–R2117	3.4	–0.86	10.3
	qVBN6	6	C962–Ky11	5.2	–1.15	18.3

^a Additive effects indicate ‘Asominori’ allele.

Validation of single QTL effects on VBN in ‘IR24’ genetic background

In 2018, VBN of IAS39 (carrying qVBN6) was 13.1, and that of IAS14 (carrying qVBN11) was 16.3 (Fig. 2), significantly lower than that of ‘IR24’ (21.2). On the other hand, VBN of IAS30 (carrying qVBN5) was 20.4, similar to that of ‘IR24’. The ‘Asominori’ alleles at qVBN6 and qVBN11 decreased VBN by 8.1 in IAS39 and by 4.9 in IAS14. In 2019, VBN of IAS39 was 15.5 and that of IAS14 was 18.8, significantly lower than that of ‘IR24’ (22.1). On the other hand, VBN of IAS30 was 22.1, close to that of ‘IR24’. The ‘Asominori’ alleles at qVBN6 and qVBN11 decreased VBN by 6.6 in IAS39 and by 3.3 in IAS14. The effects of the three QTLs were consistent between years, implying minor environmental influence on VBN. These results indicate that the ‘Asominori’ alleles at qVBN6 and qVBN11 stably decreased VBN and validate their effects in the ‘IR24’ genetic background.

Fig. 2.

Effect of each QTL for vascular bundle number in panicle neck in parents and CSSLs. Bars indicate standard deviation. Bars with the same letter are not significantly different between genotypes by Tukey–Kramer multiple comparison test (P < 0.05).

Along with evaluating of ‘Asominori’ allele in ‘IR24’ genetic background, we used three CSSLs with the ‘Asominori’ genetic background—AIS38 (carrying qVBN5), AIS49 (carrying qVBN6), and AIS76 (carrying qVBN11)—carrying target substitution chromosomal segments of ‘IR24’. The VBN of these CSSLs showed slightly incerease in VBN as compare to that of ‘Asominori’ (Supplemental Table 2). For confirmation of these QTLs, the detections of QTLs for VBN under ‘Asominori’ genetic background were difficult because of small difference of VBN between AIS and ‘Asominori’. Therefore, we performed following experiments under ‘IR24’ genetic background to confirm these detected QTL for VBN in RILs.

Confirmation of QTLs for VBN using CSSLs

VBN of the IR24/IAS30 F₂ population was 18–31, when that of ‘IR24’ was 23.2 and that of IAS30 was 22.6 (Fig. 3A). Two QTLs were detected on Chr. 5: qVBN5.1, with PVE of 12.6%, and qVBN5.2, with PVE of 13.7%. The ‘Asominori’ alleles decreased VBN by 1.09 and 1.18, respectively (Table 2). VBN of the IR24/IAS14 F₂ population was 16–28, when that of ‘IR24’ was 22.2 and that of IAS14 was 17.6 (Fig. 3B). One QTL, qVBN11, was detected on Chr. 11, with PVE of 17.9%. The ‘Asominori’ allele decreased VBN by 1.28 (Table 2). VBN of the IR24/IAS39 F₂ population was 14–29, when that of ‘IR24’ was 24.6 and that of IAS39 was 16.8 (Fig. 3C). One QTL, qVBN6, was detected on Chr. 6, with PVE of 61.9%. The ‘Asominori’ allele at qVBN6 decreased VBN by 3.35 (Table 2). These results confirm the presence of QTLs for VBN, and that the ‘Asominori’ alleles at qVBN6, qVBN5, and qVBN11 had negative effects on VBN.

Table 2. Detection of QTLs on Chrs. 5, 6, and 11 in CSSLs with ‘IR24’ background

QTL	Chr.	Interval marker	Interval marker (Mbp)a	LOD	Additive effectb	Dominant effect	PVE (%)
qVBN5.1	5	RM18751–RM18821	21.11–22.52	4.6	–1.09	0.54	12.6
qVBN5.2	5	RM7081–RM7446	24.44–24.82	5.0	–1.18	0.75	13.7
qVBN6	6	RM20546–Indel493	27.02–27.65	34.5	–3.35	–1.45	61.9
qVBN11	11	Indel8807–Indel8810	1.18–1.28	7.0	–1.28	0.02	17.9

^a On ‘Nipponbare’ genome sequence.

^b Negative sign indicates negative-effect ‘Asominori’ allele.

Fig. 3.

Frequency distributions of vascular bundle number in panicle neck in F₂ populations derived from (A) IAS30/IR24, (B) IAS14/IR24, (C) IAS39/IR24. Bars indicate means in parents with standard deviation.

Evaluation of pyramid effects of two QTLs for VBN in ‘IR24’ genetic background

From the IAS30/IAS14 F₂, PYLs 1–4 (qVBN5 + qVBN11) were selected by markers RM3351, RM6841, C5 Indel8795, and C5 Indel8837. From the IAS39/IAS14 F₂, PYLs 5–7 (qVBN6 + qVBN11) were selected by markers RM400, RM6395, C5 Indel8795, and C5 Indel8837. And from the IAS30/IAS39 F₂, PYLs 8–10 (qVBN5 + qVBN6) were selected by markers RM3351, RM6841, RM400, and RM6395. VBNs of PYLs 1–4 (qVBN5 + qVBN11) were 17.0–17.6, similar to that of IAS14 (qVBN11) but significantly lower than those of IAS30 (qVBN5) and ‘IR24’ (Figs. 4A, 5A–5C, 5E, 5F). VBNs of PYLs 5–7 (qVBN6 + qVBN11) were 11.6–14.0. Those of PYLs 6 and 7 were significantly lower than those of both parental lines, and that of PYL 5 was marginally lower than that of IAS39 (Figs. 4B, 5D, 5E, 5G). VBNs of PYLs 5–7 were not significantly different from that of ‘Asominori’ but were significantly lower than that of ‘IR24’. VBNs of PYLs 8–10 (qVBN5 + qVBN6) were 14.8–17.2, not significantly different from that of IAS39, but significantly lower than those of IAS30 and ‘IR24’ (Figs. 4C, 5C, 5D, 5H).

Fig. 4.

Effects of pyramiding QTLs for vascular bundle number in ‘IR24’ genetic background. (A) qVBN5 + 11. (B) qVBN6 + 11. (C) qVBN5 + 6. Bars with the same letter are not significantly different between genotypes by Tukey–Kramer multiple comparison test (P < 0.05).

Fig. 5.

Images of cross-section in the vascular bundle at the panicle neck of parental varieties: (A) ‘Asominori’, (B) ‘IR24’, CSSLs with ‘IR24’ genetic background: (C) IAS30, (D) IAS39, (E) IAS14 and PYLs: (F) PYL1, (G) PYL6, (H) PYL8.

Discussion

VBN is a quantitative trait controlled by multiple genes. Many QTLs for VBN in rice have been identified in segregating populations and by GWAS (Bai et al. 2012, Cui et al. 2003, Liao et al. 2021, Sasahara et al. 1999, Zhai et al. 2018, Zhang et al. 2002). RILs and DHs are useful for identification of QTLs with effects across different environments (Collard et al. 2005, Kearsey and Farquhar 1998), because environmental influence can be minimized by using multiple replicates over multiple years. Sasahara et al. (1999) detected five QTLs for VBN in 3 years’ evaluation by using Asominori × IR24 RILs. They detected one QTL for VBN at 28.64 Mbp on Chr. 6 in all 3 years, one at 26.79 Mbp on Chr. 4 in 2 years, and one at 3.81 Mbp on Chr. 11 and two at 22.56 and 23.29 Mbp on Chr. 5 in 1 year each. Our QTLs correspond to the locations of previously reported QTLs in Sasahara et al. (1999) (Table 1) but the locations of QTLs were not exactly overlapped. There were several possibilities for this difference caused by different growing environments and methods of sampling for observing VBN. In Sasahara et al. (1999), RILs were grown in Joetsu (37°6ʹ N, 138°15ʹ E) and Niigata (37°55ʹ N, 139°3ʹ E), whereas RILs in this study was grown in Saga (33°14ʹ N, 130°17ʹ E). The temperature during rice growing season were different between these areas and might affect to VBN. Also, 3 panicles were collected from longer tiller in Sasahara et al. (1999), while 5 panicles were collected from the tallest tiller among the individual plants in this study. Our number of replications for observed VBN on RILs was higher than that of previous study and it might be affected to different accuracy for QTL detections. In addition, we detected stable QTLs with strong effects and high PVEs: qVBN5 and qVBN6 (detected in 3 years) and qVBN11 (detected in 2 years) had PVE values of 18.3%–28.6%. Sasahara et al. (1999) detected two QTLs for VBN on Chrs. 4 and 6 with PVEs of 20%–23%. A significant proportion of QTLs affecting a trait are active across multiple environments, and highly heritable traits are more repeatable and stable across environments (Paterson et al. 1991, Tanksley 1993). Thus, qVBN5, qVBN6, and qVBN11 were stable owing to their large effects (PVE 18.3%–28.6%) and frequent appearance (2–3 years). Moreover, QTLs for VBN on Chrs. 6 and 11 in the RILs were confirmed as qVBN6 (27.02–27.65 Mbp) and qVBN11 (1.18–1.28 Mbp) by using F₂ populations from IAS39 or IAS14, respectably. Meanwhile two QTLs for qVBN5, qVBN5.1 and qVBN5.2, around 21.11–22.52 and 24.44–24.82 Mbp, respectively (Table 2) were close to the regions of two QTLs detected by Sasahara et al. (1999) around 22.56 and 23.29 Mbp. These results suggest that CSSLs have higher sensitivity for QTL detection and can accurately evaluate effects of a single QTL. These results will contribute to knowledge of the complex genetics of VBN in the panicle neck in rice and set the foundation for potential fine mapping and cloning of these loci.

The phenotypic variation of each QTL detected in the RILs differed by year. qVBN5 and qVBN6 appeared in all 3 years, with PVEs of 10.3%–28.6% and 13.6%–18.3%, respectively; qVBN11 appeared in 2 years, with PVEs of 11.8%–24.0%; and qVBN4 appeared only in 2018, with PVE of 10.4% (Table 1). The total PVE of the three QTLs detected in each year (44.4% in 2016, 66.2% in 2017, and 39.0% in 2018) cannot explain all of the difference in VBN between the parents. Other genetic factors might also control VBN. In addition, the differences in results between our study and Sasahara et al. (1999) might be due to testing of RILs derived from the same parents in different environments, implying the influence of environmental factors on VBN and the error of collecting the panicle. In this study, ‘IR24’ had around 15 panicles and each panicle in one plant had slightly different VBN. Therefore, to prevent the error of collecting panicle, we selected thicker panicle neck and tallest panicle in one plant because the VBN was correlated to panicle thickness (Lee et al. 1992). Thus, we could identify three QTLs for VBN using RILs in multiple years and three QTLs were confirmed in F₂ populations.

The QTLs for VBN on Chrs. 5 and 11 only detected in segregating populations derived from a cross between ‘Asominori’ and ‘IR24’. In contrast, QTLs on Chrs. 4 and 6 were located near QTLs for VBN reported previously: on Chr. 4 at 31.07–31.26 Mbp (Zhai et al. 2018) and at 22.0–26.8 Mbp (Cui et al. 2003). Here, qVBN4 was located around 24.7 Mbp, near NAL1 (Os04g0615000) at around 31.2 Mbp on Chr. 4. NAL1 controls the leaf vein pattern and increases panicle neck diameter and VBN (Fujita et al. 2013, Qi et al. 2008). qVBN6 was located around 28.6 Mbp, near APO1 (Os06g0665400) at around 27.46 Mbp on Chr. 6. APO1 enhances the development of vascular bundles, promoting carbohydrate translocation to the panicle (Terao et al. 2010) and therefore they might house the same gene.

The phenotypic effects of QTLs cannot be precisely confirmed through the use of RILs (Yamamoto et al. 2009), but can be determined through the use of CSSLs. As the genotypes of CSSLs are uniform and phenotypic variation is due mainly to the substitution segments carrying target QTLs, each QTL detected in the RILs could be validated after the evaluation of CSSLs in two years. VBNs showed similar tendencies between years, and only IAS14 and IAS39 not IAS30 were lower than that of ‘IR24’ (Fig. 2), implying a stable effect of these QTLs. The VBN of IAS39 carrying qVBN6 decreased from 6.6 to 8.1 compared to that of ‘IR24’. The VBN of IAS14 carrying qVBN11 decreased from 3.3 to 4.9 compared to that of ‘IR24’. The genetic background of each CSSL is ~93% similar to ‘IR24’ (Kubo et al. 2002). Thus, the reductions of VBN in each CSSL were caused by the presence of the ‘Asominori’ chromosomal segment in the target region of each detected QTL. Additionally, the substituted ‘Asominori’ segment still present in the CSSL (i.e. IAS30 carrying ‘Asominori’ segment on Chrs. 1 and 11, IAS14 carrying ‘Asominori’ segment on Chr. 3) may also relate to a decrease in VBN in each CSSL. On the other hand, the effect of qVBN5 on IAS30 (carrying qVBN5) was similar to that of ‘IR24’. IAS30 was taller and had a thicker culm than ‘IR24’, to which we attribute the higher VBN. Lee et al. (1992) found a significant correlation between VBN, panicle length, and internode thickness. Therefore, peduncles on taller plants are thicker, and the plants have a higher VBN.

To understand the interactions of QTLs related to VBN, we developed pyramided lines carrying pairs of QTLs. The VBNs of PYLs 1–4 (qVBN5 + qVBN11) and of PYLs 8–10 (qVBN5 + qVBN6) were the same as those of the parental lines IAS14 and IAS39, respectively. Based on our observations, plants of PYLs 1–4 and PYLs 8–10 were taller than ‘IR24’ owing to inheritance of the tall plant type from IAS30 (carrying qVBN5). Therefore, tall plant type might influence peduncle thickness and VBN. The VBNs of PYLs 5–7 (qVBN6 + qVBN11) were significantly less than those of both parental lines, so the effects of the two QTLs on VBN might be independent in the ‘IR24’ genetic background. In addition, the VBN of PYL 6 was 11.6, close to that of ‘Asominori’ (11.2 VBN), whereas PYL 7 differed from those of both IAS39 and IAS14. Thus, the interaction between qVBN6 and qVBN11 is additive effect. qVBN6 and qVBN11 decreased respectively by 6.6–8.1 and 3.3–4.9 VBN in IAS39 and IAS14, and both by 10 VBN in PYL6. In several previous study, the interactions between QTLs for morphological traits such as panicle architecture, grain length, grain width has been identified as additive effect (Ando et al. 2008, Liang et al. 2021). The pyramiding effects of ‘Habataki’ alleles on qSBN1, which controlled secondary branch number, and qPBN6, which controlled primary branch number showed an additive increase in spikelet number per panicle (Ando et al. 2008). The pyramiding effects of Z563 alleles on qGL3-1, qGL3-2, and qGL7, which controlled grain length, resulted in an additive decrease in grain length, whereas pyramiding effects of qGW3-1 and qGW3-2 which controlled grain width were additive increase in the grain width (Liang et al. 2021). There were no previous studies for revealing pyramiding effects on QTLs for VBN. In our study, the interaction of qVBN6 and qVBN11 was found to be additive to decrease in VBN. The additive interaction between QTLs for VBN was similar to previous studies for interaction of QTLs for morphological traits.

In this study, we confirmed and validated the detected QTLs in ‘Asominori’ and ‘IR24’ genetic background. In CSSLs with ‘IR24’ genetic background: IAS14, IAS30 and IAS39, the effect of ‘Asominori’ allele on QTLs decreased by 1–7 VBN. However, the effect of ‘IR24’ allele on QTLs increased by only 1–2 VBN in CSSLs with ‘Asominori’ genetic background: AIS38, AIS49 and AIS76 (Supplemental Table 2). The effect of QTLs on VBN in the ‘IR24’ genetic background was larger than that in the ‘Asominori’ genetic background. In the F₂ IAS30/IR24 for qVBN5 and IAS14/IR24 for qVBN11, QTLs for VBN were detected and showed phenotypic variation with 12.6%–13.7% and 17.9%, respectively (Table 2). In the F₂ AIS38/Asominori for qVBN5 and AIS76/Asominori for qVBN11, QTLs for VBN were detected but showed less phenotypic variation with 8.1% and 7.5%, respectively (Supplemental Table 3). Also, the phenotypic variance of QTLs on VBN in the ‘IR24’ genetic background was larger than that in the ‘Asominori’ genetic background. The effects of QTLs for VBN were different between ‘Asominori’ and ‘IR24’ genetic background.

Previously, there are several studies found that there was a significant correlation between the VBN on the panicle neck and panicle structure in segregating populations of the indica and japonica rice cross (Fukuyama and Takayama 1995, Liao et al. 2021, Sasahara et al. 1999, Zhai et al. 2018). In our study, CSSLs carrying a single QTL for VBN with ‘Asominori’ genetic background increased VBN comparing with ‘Asominori’ and also influenced panicle architecture: the primary branch number (PBN), secondary branch number (SBN), and total spikelet numbers (TSN) (Supplemental Table 4). The PBN, SBN and TSN on AIS49 carrying qVBN6 had significantly higher than on ‘Asominori’ in two years. PBN and TSN on AIS38 carrying qVBN5 and SBN and TSN on AIS76 carrying qVBN11 were significantly higher than that on ‘Asominori’ in 1 or 2 years. With increasing VBN, these QTLs increased branch and spikelet number on panicles. Terao et al. (2010) reported that ‘Habataki’ allele on APO1 increased PBN and VBN that enhanced carbohydrate translocation to panicles, resulting in higher grain yield per plant. The effects of ‘IR24’ allele on qVBN5, qVBN6, and qVBN11 tended to increase TSN and VBN, as well as ‘Habataki’ allele on APO1. Based on these results, there is possibility that ‘IR24’ allele on qVBN5, qVBN6, and qVBN11 would increase grain yield in japonica varieties by increasing branch and spikelet number and VBN. However, to confirm the effects of these QTLs in detailed, it is necessary to evaluate yield and yield components in future study. Once the location of these QTLs is narrowed down by fine mapping, the QTLs for VBN could be introduced into other japonica varieties and in developing indica-japonica cross cultivars as well as higher yield varieties to improve the yield trait through MAS. In addition, one or two QTLs might not be sufficient to increase VBN in the ‘Asominori’ genetic background. Therefore, finding several QTLs for VBN with minor effects will be necessary to develop japonica varieties with high VBN.

Author Contribution Statement

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by HTLN, SS, YN and ZD. The manuscript was written by HTLN, SHZ and DF.

Acknowledgments

We thank Prof. Hideshi Yasui and Assoc. Prof. Yoshiyuki Yamagata, belong to Kyushu University, for providing experimental materials through the National Bio-resource project. RILs, IAS, and AIS were provided by the rice seed stock center of Kyushu University with support in part by National Bio-Resource Project of the MEXT, Japan. Ha Thi Le Nguyen was funded by the Vietnamese government for “the critical program of biotechnology development and application in agriculture and rural development”. This research was part of the dissertation submitted by the first author in partial fulfilment of the Ph.D. degree. All authors have provided consent.

Literature Cited

•  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.
•  Bai,  X.,  B.  Wu and  Y.  Xing (2012) Yield-related QTLs and their applications in rice genetic improvement. J Integr Plant Biol 54: 300–311.
•  Chaudhry,  F.M. and  K.  Nagato (1970) Role of vascular bundles in ripening of rice kernel in relation to the locations on panicle. Jpn J Crop Sci 39: 301–309.
•  Collard,  B.C.Y.,  M.Z.Z.  Jahufer,  J.B.  Brouwer and  E.C.K.  Pang (2005) An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142: 169–196.
•  Cui,  K.,  S.  Peng,  Y.  Xing,  S.  Yu,  C.  Xu and  Q.  Zhang (2003) Molecular dissection of the genetic relationships of source, sink and transport tissue with yield traits in rice. Theor Appl Genet 106: 649–658.
•  Dellaporta,  S.L.,  J.  Wood and  J.B.  Hicks (1983) A plant DNA minipreparation: Version II. Plant Mol Biol Report 1: 19–21.
•  Donald,  C.M. (1968) The breeding of crop ideotypes. Euphytica 17: 385–403.
•  Fei,  C.,  X.  Geng,  Z.  Xu and  Q.  Xu (2019) Multiple areas investigation reveals the genes related to vascular bundles in rice. Rice (N Y) 12: 17.
•  Fujita,  D.,  K.R.  Trijatmiko,  A.G.  Tagle,  M.V.  Sapasap,  Y.  Koide,  K.  Sasaki,  N.  Tsakirpaloglou,  R.B.  Gannaban,  T.  Nishimura,  S.  Yanagihara et al. (2013) NAL1 allele from a rice landrace greatly increases yield in modern indica cultivars. Proc Natl Acad Sci USA 110: 20431–20436.
•  Fukuyama,  T. and  T.  Takayama (1995) Variations of the vascular bundle system in Asian rice cultivars. Euphytica 86: 227–231.
•  Fukuyama,  T.,  H.  Sasahara and  Y.  Fukuta (1999) Variation of vascular bundle system corresponds to indica, tropical- and temperate-japonica differentiation of Asian rice (Oryza sativa L.). Breed Sci 49: 15–19.
•  Kearsey,  M.J. and  A.G.L.  Farquhar (1998) QTL analysis in plants; where are we now? Heredity 80: 137–142.
•  Khush,  G.S. (1997) Origin, dispersal, cultivation and variation of rice. Plant Mol Biol 35: 25–34.
•  Kubo,  T.,  Y.  Aida,  K.  Nakamura,  H.  Tsunematsu,  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.
•  Lee,  D.J.,  B.S.  Vergara,  O.B.  Zamora,  B.K.  Kim and  J.C.  Chae (1992) Development of vascular bundles in the peduncle of different tillers and its relationship to panicle characteristics in rice. J Crop Sci Biotechnol 37: 155–165.
•  Lee,  D.J. and  J.C.  Chae (2000) Effect of nitrogen split application methods on development of vascular bundle and yield components of rice cultivars. J Crop Sci Biotechnol 45: 237–240.
•  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.
•  Liao,  S.,  J.  Yan,  H.  Xing,  Y.  Tu,  H.  Zhao and  G.  Wang (2021) Genetic basis of vascular bundle variations in rice revealed by genome-wide association study. Plant Sci 302: 110715.
•  Liu,  T.,  W.  Bi,  J.  Zhang,  Y.  Cui,  Z.  Yan,  Y.  Wang,  C.  Fei,  H.  Xu,  L.  Tang,  W.  Chen et al. (2016) Characterization of the relationship between vascular bundles features and indica-allelic frequency using a seventh filial generations of indica × japonica rice crosses. Euphytica 209: 739–748.
•  Lucas,  W.J.,  A.  Groover,  R.  Lichtenberger,  K.  Furuta,  S.R.  Yadav,  Y.  Helariutta,  X.Q.  He,  H.  Fukuda,  J.  Kang,  S.M.  Brady et al. (2013) The plant vascular system: evolution, development and functions. J Integr Plant Biol 55: 294–388.
•  McCouch,  S.R.,  L.  Teytelman,  Y.  Xu,  K.B.  Lobos,  K.  Clare,  M.  Walton,  B.  Fu,  R.  Maghirang,  Z.  Li,  Y.  Xing et al. (2002) Development and mapping of 2240 new SSR markers for rice (Oryza sativa L.). DNA Res 9: 199–207.
•  Paterson,  A.H.,  S.  Damon,  J.D.  Hewitt,  D.  Zamir,  H.D.  Rabinowitch,  S.E.  Lincoln,  E.S.  Lander and  S.D.  Tanksley (1991) Mendelian factors underlying quantitative traits in tomato: Comparison across species, generations, and environments. Genetics 127: 181–197.
•  Peng,  S.,  K.G.  Cassman,  S.S.  Virmani,  J.  Sheehy and  G.S.  Khush (1999) Yield potential trends of tropical rice since the release of IR8 and the challenge of increasing rice yield potential. Crop Sci 39: 1552–1559.
•  Qi,  J.,  Q.  Qian,  Q.  Bu,  S.  Li,  Q.  Chen,  J.  Sun,  W.  Liang,  Y.  Zhou,  C.  Chu,  X.  Li et al. (2008) Mutation of the rice narrow leaf1 gene, which encodes a novel protein, affects vein patterning and polar auxin transport. Plant Physiol 147: 1947–1959.
•  Sasahara,  H.,  Y.  Fukuta and  T.  Fukuyama (1999) Mapping of QTLs for vascular bundle system and spike morphology in rice (Oryza sativa L.). Breed Sci 49: 75–81.
•  Tanksley,  S.D. (1993) Mapping polygenes. Annu Rev Genet 27: 205–233.
•  Terao,  T.,  K.  Nagata,  K.  Morino and  T.  Hirose (2010) A gene controlling the number of primary rachis branches also controls the vascular bundle formation and hence is responsible to increase the harvest index and grain yield in rice. Theor Appl Genet 120: 875–893.
•  Tsunematsu,  H.,  A.  Yoshimura,  Y.  Harushima,  Y.  Nagamura,  N.  Kurata,  M.  Yano,  T.  Sasaki and  N.  Iwata (1996) RFLP framework map using recombinant inbred lines in rice. Japan J Breed 46: 279–284.
• Wang, S., C.J. Basten, P. Gaffney and Z.B. Zeng (2012) Windows QTL Cartographer software version 2.5. (https://statgen.ncsu.edu/qtlcart/WQTLCart.htm) Department of Statistics, North Carolina State University, Raleigh, NC.
•  Xu,  Z.,  W.  Chen,  L.  Zhang and  S.  Yang (2005) Design principles and parameters of rice ideal panicle type. Chin Sci Bull 50: 2253–2256.
•  Yamamoto,  T.,  J.  Yonemaru and  M.  Yano (2009) Towards the understanding of complex traits in rice: Substantially or superficially? DNA Res 16: 141–154.
•  Yonemaru,  J.I.,  S.H.  Choi,  H.  Sakai,  T.  Ando,  A.  Shomura,  M.  Yano,  J.  Wu and  S.  Fukuoka (2015) Genome-wide indel markers shared by diverse Asian rice cultivars compared to Japanese rice cultivar ‘Koshihikari’. Breed Sci 65: 249–256.
•  Zhai,  L.,  T.  Zhen,  X.  Wang,  Y.  Wang,  K.  Chen,  S.  Wang,  Y.  Wang,  J.  Xu and  Z.  Li (2018) QTL mapping and candidate gene analysis of peduncle vascular bundle related traits in rice by genome-wide association study. Rice (N Y) 11: 13.
•  Zhang,  Z.H.,  P.  Li,  L.X.  Wang,  C.J.  Tan,  Z.L.  Hu,  Y.G.  Zhu and  L.H.  Zhu (2002) Identification of quantitative trait loci (QTLs) for the characters of vascular bundles in peduncle related to indica-japonica differentiation in rice (Oryza sativa L.). Euphytica 128: 279–284.