QTL ANALYSIS OF DEHYDRATION TOLERANCE AT SEEDLING STAGE IN RICE (ORYZA SATIVA L.)

Dehydration tolerant QTLs at the seedling stage in rice (Oryza sativa L.) was studied using green shoot length as a parameter, which was measured during the recovery period after dehydration stress. According to the results of pilot experiment carried out to evaluate the two parental rice cultivars for the level of dehydration tolerance, at 4 day and 5-day dehydration stress, Japonica parent Hyogokithanishiki (HGKN) showed more tolerance than that of Hokuriku-142 Indica parent (HOK). Quantitative trait loci (QTLs) associated with dehydration tolerance was identified using 163 recombinant inbred lines derived from these parents on a linkage map constructed with 95 simple sequence repeat (SSR) markers. Three dehydration tolerant minor QTLs were identified on chromosome 1, 2 and 3 at composite interval mapping by WinQTLCARTOGRAHER. According to the additive effect plots the QTLs on chromosome 1 and 2 have the allelic effect from the Indica parent Hokuriku-142 and QTL on chromosome 3 has allelic effect from Hyogokithanishiki Japonica parent. However, according to the 1000 time permutation test, these QTLs didn’t exceed the threshold LOD value.


INTRODUCTION
Drought stress is the most common adverse environmental condition that can seriously reduce crop productivity. Increasing crop resistance to drought stress would be the most economical approach to improve agricultural productivity and to reduce agricultural use of fresh water resources. Cultivated rice (Oryza sativa L.) demands tremendous amount of water during growth. Development of new rice cultivars with drought tolerance saves a great amount of water and also helps to increase yield stability. Drought resistance in rice depends on one or more factors of the following components: ability of roots to exploit deep soil water, the capacity for osmotic adjustment and the ability to control and reduce water loss (Ngueyen et al. 2004). Drought tolerant rice cultivars exhibit various adaptations such as earlymaturity, flowering even under little or no rain, stop growing as soil moisture declines, leaf rolling and response of root distribution to soil drying (Azhiri-Sigari et al. 2000). *Corresponding author: lankaranawake@agbio.ruh.ac.lk However, drought tolerance is unmanageable to molecular genetics studies mainly due to limited awareness of specific traits linked to drought tolerance. Furthermore, it is difficult to conduct drought stress treatments in a quantitative and reproducible way. These difficulties have significantly impeded research on plant drought tolerance. Limited awareness of plant phenotypes specifically conferred by drought stress has prevented researchers from using traditional genetics approaches to directly study drought stress tolerance (Tripathy et al. 2000). The biological basis for drought tolerance is still largely unknown and few drought tolerance determinants have been identified (Ludlow and Muchow, 1990;Bohnert et al. 1995;Araus et al. 2002;Bruce et al. 2002). Undisclosed drought tolerance mechanism has stagnated both traditional breeding efforts and use of modern genetics approaches in the improvement of drought tolerance of crop plants (Passioura 2010;Sinclair 2011).
Despite the lack of understanding of drought tolerance mechanisms, physiological and mo-98 lecular biological studies have documented several plant responses to drought stress (Bohnert et al. 1995;Blum, 1996;Ingram and Bartel, 1996;Bray, 1997;Schroeder et al. 2001;Luan, 2002;Sharp et al. 2004;. Molecular approaches to develop drought tolerant rice cultivar have been applied by QTL analysis. Integration of publicly available rice genetic map covered by microsatellite markers with trait data of diverse mapping populations has been successfully identified chromosomal regions associated with some secondary traits of drought response, such as osmotic adjustment (Lilley et al. 1996;Zhang et al. 1999Zhang et al. , 2001Robin et al. 2003), cell membrane stability (Tripathy et al. 2000), abscisic acid (ABA) content (Quarrie et al. 1994(Quarrie et al. , 1997, stomatal regulation (Price et al. 1997), leaf water status, and root morphology (Champoux et al. 1995;Ray et al. 1996;Price and Tomos 1997;Yadav et al. 1997;Ali et al. 2000;Courtois et al. 2000;Zheng et al. 2000;Zhang et al. 2001;Kamoshita et al. 2002;, membrane stability and osmotic adjustment (Li et al. 2005), leaf size and ABA accumulation (Robbin et al. 2003). Wang et al. (2004) showed the reduction in shoot height under dehydration stress in dehydration susceptible rice cultivar, CT9993 while dehydration tolerant IR2266 was not affected. Genomic regions of rice chromosomes with drought-resistant loci have been reported by Adam et al. (2002), Tripathy et al. (2000, Zeng et al. (2000).
In the present study we used F 6 generation of inbred line population for QTL analysis at the seedling stage. The Indica parental cultivar, Hokurikuriku-142 and Japonica parental cultivar, Hyogo-kithanishiki are genetically divergent types, which provides enough segregation of both molecular marker alleles and ergonomically important genes for QTL analysis. The drought tolerant data were collected according to Ranawake and Nakamura (2011). The objective of this study was to investigate the effectiveness of new breeding materials for rice dehydration tolerant through genetic re-combination between Indica and Japonica rice that have no previous evidence on drought tolerant traits.

MATERIALS AND METHODS
F 6 generation of 163 recombinant inbred line (RIL) population derived from Japonica rice cultivar, Hyogo-Kithanishiki (HGKN), and an Indica rice cultivar, Hokuriku-142 (HOK) were used for QTL analysis. Hokuriku was bred from a cross between a Korean cultivar, `Milyang 21 and an IRRI line ` IR-2061-214-31` in the Hokuriku Agricultural Experimental station, Japan (Misawa et al. 2000).

Evaluation of inbred line population for dehydration tolerance:
Two week old rice seedlings were undergone a 5 day drying period. Five-day after complete drying plants were watered and 10 days after recovery period plants were evaluated for drought tolerance (Ranawake and Nakamura 2011). Experiment was carried out according to the complete randomized block design and repeated three times. Average green shoot length was used as the parameter to evaluate the inbred line population on the 10 th day of the recovery period followed by 5-day dehydration stress.
DNA extraction, marker genotyping and construction of linkage map: DNA extraction was done and primer pairs for rice microsatellites (simple sequence repeat: SSR) were surveyed for polymorphisms. The linkage map was constructed using MAPMAKER 3.0, for 100 SSR markers. Composite interval mapping was performed to identify significant QTLs with forward-backward regression. The experiment-wise LOD threshold level was determined by computing 1,000 permutations, as implemented by the WinQTLCARTOGRA-HER (Wang et al. 2003 http://statgen.ncsu.edu/ qtlcart/WinQTLCart). These permutations can account for non-normality in marker distribution and trait values. The level of significant LOD value for QTLs in this study was determined to be 3.01 at p<0.05.

RESULTS AND DISCUSSION
Evaluation of two parental rice cultivars for dehydration tolerance: Considering the fact that two parental cultivars showed the highest significant difference in the green shoot length at 5-day dehydration stress, this stress condition was decided to apply for the evaluation of RILs (Fig.1).

Evaluation of inbred line population for dehydration tolerance:
The frequency distribution of average green shoot length in the RIL population on the 10 th day of the recovery period after 5-day completely drying period showed a transgressive segregation in both directions (Fig. 2). This indicates that the alleles for drought tolerance in the population have donated from both drought tolerant parent and drought susceptible parent.
QTL analysis of dehydration tolerance in inbred line population: QTL analysis was performed using WinQTLCARTOGRAHER version 2.0 (Wang et. al, 2003). To understand the significance of these QTLs, composite interval mapping was performed. Minor dehydration tolerant QTLs with LOD score more than 2 were detected on chromosome 1, 2 and 3 at composite interval mapping of the present study (Fig.3, Fig. 4). A total of 24% According to additive effect plots depicted in the composite interval map plot, the QTLs on chromosome 1 and 2 had the allelic effect from the Indica parent Hokuriku-142 and QTL on chromosome 3 had allelic effect from Hyogokithanishiki Japonica parent (Table 1).
There are many evidences on presence of drought tolerant candidate regions on chromosome 1, 2 and 3. RFLP linkage map for Co39 x Moroberekan recombinant inbred population which was used for QTL analysis for osmotic adjustment and lethal osmotic potential also indicated that chromosomes 1 and 3, possess the QTLs associated with root characters (Champoux el al., 1995).  has been selected rice chromosome 1 for identification of candidate genes for drought stress tolerance in rice.  and Lafitte et al., (2002) also give evidence on presence of drought tolerant QTLs on chromosome 1. In the region of RM472-RM104 on chromosome 1, drought tolerant QTLs for root volume and number of days to leaf rolling were detected by Yue et al., (2006). QTL were also identified for root thickness and root weight (Zheng et al. 2000), as well as for relative water content, leaf rolling, and leafdrying score (Babu et al. 2003) under drought stress. Further, a total of three QTLs on chromosomes 2, 4, and 7 were detected for relative growth rate in rice (Kato et al. 2008).
Chromosome 2 was found to be associated with the grain yield or yield components un- Fig. 3: QTLs controlling dehydration tolerance at the seedling stage identified by the composite interval mapping in the 5-day dehydration stress der drought stress conditions by Yue et al. (2008). Yue et al. (2008) has also reported presence of drought resistant QTLs both on long arm and short arm of the chromosome 3 for flag leaf length, flag leaf width and plant height under drought stress. This chromosomal region on chromosome 3 coincides with the same chromosomal regions where we detected minor QTL on chromosome 3 (Table  1). These all findings align with the findings of our study at least proving that rice chromosomes 1, 2 and 3 are responsible for drought tolerance at the seedling stage. However considering background genetic effect, these QTLs do not exceed threshold LOD value computed by the 1000 time permutation test in WinQTLCARTOGRAHER version 2.5. According to results of QTL analysis there was a narrow variation in the detected QTLs for dehydration tolerance. This might be the reason that two parental rice cultivars had narrow but significant difference at the level of dehydration tolerance which was not wide enough to detect strong dehydration tolerant QTLs.