IC4R005-Microarray-2012-22419825

Project Title

• Efficacy of Microarray Profiling Data Combined with QTL Mapping for the Identification of a QTL Gene Controlling the Initial Growth Rate in Rice

The Background of This Project

• Nowadays, and especially in the tropics, the rice cropping system is shifting from transplanting to direct sowing because it is less labor intensive (Erguiza et al. 1990). For the establishment of such a direct sowing system, strong seedling vigor is one of the most important agronomic traits. The level of seedling vigor is determined by many morphological and physiological traits, among which the germination rate and early seedling growth are the major targets in a breeding program (Williams and Peterson 1973, Sasahara et al. 1986). Moreover, several studies have suggested that phytohormones, such as gibberellin and auxin, are important for their control (Oikawa et al. 2004, Chhun et al. 2007, Lo et al. 2008).
• Microarray-based transcriptome profiling studies have been primarily used to address gene regulations related to morphological and physiological traits of animals and　plants. However, results from these studies tend to identify a large number of differentially expressed genes due to　the complex interactions between various signaling pathways.　Previously, a report　combining expression profiling using　microarrays with genetic mapping significantly contributed to　the isolation of genetic factors in animal research (Aitman　et al. 1999). However, the identification of QTL genes　using this approach has never been reported in plants.
• This study is the first to use this new approach for the isolation of QTL genes in rice; we conducted combined QTL mapping and microarray profiling analyses to identify QTL genes for seedling vigor at the initial growth stage. Our QTL analysis identified two QTLs, early-stage plant develop- ment1/2 (qEPD1 and qEPD2) for increasing plant height and/ or leaf sheath length at the initial growth stage. By combining QTL mapping and microarray profiling, we predicted that qEPD1 or qEPD2 correspond to gibberellin 20 oxidase-1 or -2 genes (OsGA20ox1, 2), respectively, and confirmed this prediction by complementation analysis. Our study demonstrated that the integration of QTL mapping and expression profiling could be a powerful method to narrow down the number of candidate genes for QTL(s) of interest.

Plant Materials & Treatment

• The researchers used a population of 82 BILs (BC 1 F 12 ) derived from a cross between Koshihikari (japonica) and Habataki (indica) to identify QTLs for seedling traits at the initial growth stage and CL at the heading stage. Seeds of all the lines were immersed in water for 2 d and then sown into the cell plug tray (cell count,　14 × 32; tray size, 540 × 280 mm; depth of cell, 30 mm). These　lines were first grown under greenhouse conditions (15 h of　daylight, 25 × C) until 30 d after the germination to evaluate　seedling traits; then, they were transplanted to the paddy　field to measure CL at the heading stage. The phenotypic evaluation in BILs was conducted in three or four replicates. SLs were　selected from the BC4F2　generation.

• We used the Habataki (indica) and Koshihikari (japonica) cultivars as the respective low and high seedling growth cultivars at the initial growth stage (Fig. 1A). After 30 d of germination, the average plant height of Koshihikari was 13.81 cm, whereas that of Habataki was 12.35 cm (Fig. 1A, B). Since the plant height (PH) of rice seedling depends on the total of leaf sheath length (LSL) and leaf blade length (LBL), we measured the LBL and LSL (Fig. 1C, D). The average LSL of Koshihikari and Habataki was 6.91 and 4.88 cm, respectively (Fig. 1C), but there was no significant difference in the LBL (Fig. 1D). These results demonstrate that the initial growth vigor of Koshihikari is mainly due to its longer leaf sheath.
• To identify genes for increasing PH and LSL at the initial growth stage, we conducted QTL analysis by using 82 backcross inbred lines (BILs) derived from these two parents. The PH of 82 BILs varied from 9.73 to 25.48 cm, whereas the LSL varied from 4.08 to 10.88 cm at 30 d after germination (Fig. 2A, B). The frequency of PH and LSL showed an almost normal distribution with a transgressive segregation (Fig. 2A, B). We also conducted QTL analysis of the total culm length (CL) at the heading stage, because we were interested in the difference in PH between the initial growth and the heading stages. CL also showed a continuous frequency distribution with a transgressive segregation (Fig. 2C).

• For PH, one QTL, qPH-3.1 [logarithm of odds (LOD) score = 5.2], was identified within a 25 cM region between markers A-1369Tc and A-362Tc on the bottom part of chromosome 3; this QTL can explain 24.8% of the total phenotypic variation (Fig. 2D, Table 1). Three QTLs for LSL at the initial growth stage were mapped on chromosomes 1 (qLSL-1.1), 3 (qLSL-3.1) and 11 (qLSL-11.1) (Fig. 2D, Table 1). Among these QTLs, qLSL-1.1 (LOD = 4.0), qLSL-3.1 (LOD = 3.9) and qLSL-11.1 (LOD = 3.5) explained 20.3, 19.0 and 16.0% of the total phenotypic variations, respectively (Fig. 2D, Table 1). Consistent with the high PH of Koshihikari being closely linked to its long LSL (Fig. 1), qPH-3.1 for PH and qLSL-3.1 for LSL were co-localized between the same markers on chromosome 3 (Fig. 2D, Table 1). We also performed the QTL analysis of CL using the same BIL population. Two QTLs [qCL-1.1 (LOD = 3.5) and qCL-1.2 (LOD = 4.2)] were detected in the 141.2 and 166.2 cM region on the long arm of chromosome 1 (Fig. 2D, Table 1). qCL-1.1 overlapped with the genomic position of qLSL-1.1 for LSL (Table 1), suggesting that qCL-1.1 might be identical to qLSL-1.1.

• However, we did not detect a significant LOD peak for CL around the qPH-3.1 region on chromosome 3. From previous QTL analyses of rice CL, many QTLs have been identified around the SD1 locus encoding OsGA20ox2 on chromosome 1. Gibberellin is a phytohormone that plays important roles in many aspects of plant growth and development, such as seed germination, stem elongation and flower development (Yamaguchi and Kamiya 2000). Further, by using similar BIL population of Koshihikari and Habataki, Ashikari et al. (2005) identified a QTL for plant height present close to the SD1 locus, and reported that Habataki has a loss-of-functio allele of SD1 with a 383 bp deletion, which shortens the CL via partial inhibition of gibberellin biosynthesis in Habataki. Since the SD1 gene was mapped within the qLSL-1.1 and qCL-1.1 region, it is likely that these two QTLs might be identical to SD1 (see below). In contrast, qCL-1.2, which is located downstream of the SD1 locus, shared a similar interval with qCL1b that was recently identified as a novel QTL for CL.

Labs working on this Project

• Bioscience and Biotechnology Center, Nagoya University, Nagoya, 464-8601 Japan
• Honda Research Institute Japan Co., Ltd., Japan
• RIKEN Plant Science Center, Tsurumi, Yokohama, Kanagawa, 230-0045 Japan

Corresponding Author

Koichiro Aya (Email: koichi.a@agr.nagoya-u.ac.jp )