Genetic Control of Flowering Time in Rice, a Short-Day Plant

Abstract

Endogenous genetic factors and environmental signals control the time of flowering in plants. One of the environmental signals is photoperiod. Genetic control mechanisms for the photoperiodic response of flowering of long-day plants (LDPs) have been extensively analyzed through the use of Arabidopsis as a model plant (for review, see Coupland, 1998; Levy and Dean, 1998;Samach and Coupland, 2000). In contrast, mechanisms in short-day plants (SDPs) remain unclear, although many physiological studies have been performed on SDPs, such as Pharbitis nil (for review, seeLumsden, 1998). Recent progress in genome analysis has provided a new strategy for analyzing the genetic control of flowering in rice (Oryza sativa; SDP). Several studies have demonstrated that the structure of genes involved in the photoperiodic response of flowering in rice show remarkable similarity to those in Arabidopsis. In rice, genetic analyses of flowering time (often called heading date) have been performed on mutants and natural variants. Several genes involved in the photoperiodic response (photoperiod sensitivity) have been identified (Yokoo et al., 1980; Yamagata et al., 1986; Yokoo and Okuno, 1993; Okumoto and Tanisaka, 1997). A series of nearly isogenic lines (NILs) for several photoperiod sensitivity genes have been developed to facilitate genetic analysis of flowering time in rice (Yamagata et al., 1986). However, the nature of the quantitative inheritance of flowering time has prevented us from performing more detailed analyses, including analysis of epistatic interactions and determination of chromosomal locations of genes. In the last decade, the progress in development of DNA markers made quantitative trait locus (QTL) analysis possible to clarify the number and nature of the genes controlling flowering time in rice (Yano and Sasaki, 1997). Plant materials used for the detection of QTLs and chromosomal locations of QTLs controlling flowering time. A, Mapping populations derived from a cross between Nipponbare and Kasalath. QTL mapping was performed by using F2, BC1F5, BC3F2, and BC4F2 lines. Fine mapping and selection of NILs were conducted by using advanced backcross progeny. B, High-density RFLP linkage map showing chromosomal locations of QTLs (Hd1–Hd14) for flowering time. The development of NILs by marker-assisted selection, in which a small chromosomal segment including the detected QTL of donor variety Kasalath was substituted into the Nipponbare genetic background, has provided many advantages for the genetic analysis of flowering time in rice (for review, see Yano and Sasaki, 1997). For example, the QTL-NILs can be used in the characterization of the photoperiodic response, epistatic interaction analysis, and fine genetic linkage mapping for target QTLs. The QTLs were classified into two groups based on the response of the QTL-NILs to photoperiod. Five QTLs, Hd1,Hd2, Hd3, Hd5, and Hd6, were found to confer the photoperiod sensitivity (Lin et al., 2000;Yamamoto et al., 2000; M. Yano, unpublished data). By the genetic analysis using QTL-NILs, the existence of an epistatic interaction between Hd1 and Hd3 was clarified. It was also suggested that the Kasalath allele of Hd3 itself does not affect photoperiod sensitivity, but that it is involved in enhancement of expression of the Nipponbare alleles of photoperiod sensitivity QTLs, Hd1 and Hd2 (Lin et al., 2000). In addition, epistatic interaction between Hd2 andHd6 was clearly detected in the analysis of the advanced progeny. The effect of the Kasalath allele of Hd6could be observed only in the presence of the Nipponbare allele ofHd2 (Yamamoto et al., 2000). QTL-NILs could also be used for the fine mapping of target QTLs. Five QTLs, Hd1, Hd2, Hd3, Hd6, and Hd9, were mapped precisely on the genetic linkage map as single Mendelian factors (Yamamoto et al., 1998, 2000; Lin et al., 2002; H.X. Lin and M. Yano, unpublished data). Moreover, high-resolution mapping enabled us to dissect two tightly linked loci,Hd3a and Hd3b, in the Hd3 region (Fig.1; Monna et al., 2002). Analysis of the photoperiodic response in NILs of Hd3a and Hd3b revealed that the Kasalath allele of Hd3a promotes flowering under short-day (SD) conditions, and that the Kasalath allele ofHd3b delays late flowering under long-day (LD) and natural field conditions (Monna et al., 2002). Together, it is clearly demonstrated that genetic control mechanisms of flowering in rice could be dissected into each component by a series of genetic analyses of flowering date based on the QTL analysis. A major QTL, Hd1, which controls response to photoperiod, was cloned by means of a map-based cloning strategy (Yano et al., 2000). Hd1 is an orthologue of CO(constans) in Arabidopsis (Putterill et al., 1995) and encodes a protein with the structure of a zinc finger domain and a nuclear localization signal. In addition, structural analysis demonstrated that the major gene controlling the response to photoperiod, photoperiod sensiti