Unlike germination and seedling stage, QTLs for reproductive cold tolerance in rice were mapped mostly on chromosomes 3, 4, 6, 10 and 12 by different researchers using different mapping populations involving diverse cold tolerant donor germplasm (Table 2, Fig.1). Around 40 QTLs governing cold tolerance at reproductive stage, which largely control seed set rate through molecular and physiological changes in plants, have been reported so far. Saito et al. (1995) mapped two QTLs on chromosome 3 and 4 from Norin PL8. A later study detected two OTLs, Ctb-1 and Ctb-2 governing spikelet fertility under cold stress on chromosome 4 using a set of NILs derived from a cross between Kirara397/Norin PL8 (Saito et al. 2001). Ctb-1 was further fine mapped to a 17 kb region (Saito et al 2004) and the putative gene that controls cold tolerance at booting stage encoding F-box protein as part of E3 ubiquitin ligase complex (Skp1) was detected (Saito et al. 2010). Using spikelet fertility as criteria for cold tolerance at booting stage, many other QTLs were detected by different researchers (Table 2 and 4). Takeuchi et al. (2001) mapped eight QTLs using DH lines of Akihikari/Koshihikari including the ones with higher contributions for general cold tolerance (qCT-7, on chromosome 7), and cold tolerance related to culm length (qCL-1, on chromosome 1) and heading date (qHD-3-2, on chromosome 3 and qHD-6, on chromosome 6), explaining 22.1%, 31.1%, 15.5%, and 50.5% of the respective phenotypic variation.
The QTLs for booting stage cold tolerance, qCTB2a and qCTB3, were mapped by Andaya and Mackill (2003b) using 191 RIL derived from a cross between M-202 and IR50. Eight QTLs based on variation in spikelet fertility, qCTB-1-1, qCTB-4-1, qCTB-4-2, qCTB-5-1, qCTB-5-2, qCTB-10-1, qCTB-10-2, and qCTB-11-1, were mapped using a RIL population derived from a cross between a cold told tolerant japonica cultivar ‘Kunmingxiaobaigu’ (KMXBG) and a cold-sensitive japonica cultivar, ‘Towada (Xu et al. 2008). QTL analysis with composite interval mapping has identified three main-effect QTLs, qPSST-3, qPSST-7, and qPSST-9, respectively, on chromosomes 3, 7, and 9 (Suh et al. 2010) showing an additive effect on the rate of spikelet fertility under cold stress. Another QTL for seed fertility under cold stress, qLTB3, was identified on the long arm of chromosome 3 from the cold-tolerant breeding line ‘Ukei 840’ in F2 and BC1F2 populations from crosses between ‘Ukei840’ and ‘Hitomebore’. The cold tolerance of ‘Ukei 840’ was derived from the Chinese cultivar ‘Lijiangxintuanheigu’ (Shirasawa et al. 2012). Seven SNP markers have been reported in five genes within the qLTB3 region, all of them causing amino acid substitutions. One of those SNPs (in Os03g0790700 gene) caused a mutation in a conserved amino acid and was considered the strongest candidate for conferring cold tolerance. The Os03g0790700 gene encodes a protein similar to the Arabidopsis AAO2 aldehyde oxidase, which is believed to function in ABA biosynthesis (Koiwai et al. 2004; Seo et al. 2004). Zhou et al. (2010) reported a QTL for cold tolerance at the booting stage on chromosome 7, named qCTB7, which explained 9 and 21% of the phenotypic variances in the F2 and F3 generations, respectively. Twelve putative cold tolerance genes from this QTL region were identified by fine mapping and candidate gene cloning. On the basis of genetical and physical mapping, the authors suggested that two other QTLs previously identified in the same location (qRCT7 – Dai et al. 2004; and qCT-7 – Takeuchi et al. 2001) may correspond to the same locus as qCTB7. Although identified in diverse genetic backgrounds and environments, the three QTLs explained similar percentages of the phenotypic variance, ranging from 20.6% to 22.1%. In a 92-kb region of qCTB7, Zhou et al. (2010) identified two auxin response genes (Os07g0576100 and (Os07g0576500), two hydrolase genes (Os07g0575800 and Os07g0577300), and one ubiquitin-conjugating enzyme E2 gene (Os07g0577400). The phytohormone auxin plays a central role in almost every aspect of plant growth and development and several auxin-responsive genes have been implicated in both biotic (e.g. pathogen infection (Ding et al. 2008)) and abiotic stress responses (e.g. desiccation, low temperature, and salinity (Hannah et al. 2005; Jain and Khurana 2009; Song et al. 2009)). Kuroki et al. (2007) identified a QTL for booting stage cold tolerance on chromosome 8 (qCTB8) through the analysis of F2, F3, and F7 populations, using SSR markers. This QTL explains 26.6% of the phenotypic variance. About 30 open reading frames were identified at the 193 kb region of qCTB8. One of them encodes (Os08g05570) monodehydroascorbate reductase (MDAR), which was shown to be upregulated by cold treatment in rice anthers during the young microspore stage. The amount of pollen available for fertilization is directly related to anther length, and cold affects pollen grain maturation, reducing fertility as a consequence. Cold-tolerant varieties belong larger anthers and, consequently, they produce a larger number of pollen grains than susceptible varieties do. Therefore, a strong correlation was suggested to exist between cold tolerance QTLs and anther length QTLs, being the amount of pollen produce an important component of the tolerance mechanism (Saito et al. 2001). In the qCTB8 region, there are at least six genes (Os08g05510 encodes MYB Family transcription factor, Os08g05520 encodes MYB-like DNA binding protein domain, Os08g05580, Os08g05590 and Os08g05600 encode aquaporin protein, Os08g05610 and Os08g05610 encode cytochrome P450) that encode protein related to cold tolerance at vegetative stage. From a set NILs derived from Kirara397/Norin PL8, Saito et al. (2001) identified a marker which is tightly linked with Ctb-2 on chromosome 4, was significantly associated with anther length.
It has been suggested that QTLs related to cold tolerance in the germination stage are independent of QTLs conferring tolerance at the vegetative and reproductive phase (Saito et al. 2001; Andaya and Mackill, 2003a,b; Fujino et al. 2004), indicating that cold tolerance may be developmentally regulated and growth stage-specific. High-quality sequence data of the rice genome has provided a genome-wide SNP resource (IRGSP 2005) that leads to high-quality and reliable markers. Until the present moment, however, only a few studies have been able to link cold tolerance to SNPs in rice (Koseki et al. 2010, Shirasawa et al. 2012), and only one (Kim et al. 2011) could effectively locate an SNP which resulted in amino acid substitution (I99V – Ile to Val) leading to reduced enzyme activity (glutathione transferase isoenzyme OsGSTZ2). According to the authors, this functional difference in the OsGSTZ2 isoform could explain the differential response observed between cold-tolerant and cold-sensitive rice cultivars.