Genetic analysis of cold tolerance at the booting stage of rice in Hokkaido

Spikelet fertility of rice decreases if rice plants are exposed to low temperature at the booting stage, due to the failure of microspore development under low temperature conditions. Cold tolerance at the booting stage is one of the most important objectives in breeding programs in Hokkaido. In order to incorporate cold tolerance of foreign rice varieties into Hokkaido, Silewah, which is a tropical japonica variety from Indonesia, was backcrossed to a breeding line, Hokkai241, and a cold-tolerant variety, Norin-PL8, was developed. However, the chromosomal locations of the cold tolerance genes of Norin-PL8 had not been determined. In the present study, I analyzed the quantitative trait loci (QTLs) for cold tolerance of Norin-PL8 using DNA markers. I. Identification of chromosomal segments associated with cold tolerance at the booting stage 1. Graphical genotype analysis of Norin-PL8 Norin-PL8 was developed by the cross Silewah//Hokkai241. Therefore, the cytoplasm of Norin-PL8 is derived from Silewah. I examined the cold tolerance of F_1 plants of reciprocal crosses. There was no difference between the crosses, indicating that the cytoplasm is not associated with cold tolerance. In order to analyze the genotype of Norin-PL8, I examined the restriction fragment length polymorphism (RFLP) patterns of Silewah, Hokkai241 and Norin-PL8. Out of the 240 RFLP markers used, 102 markers gave polymorphic bands. All of the polymorphic patterns could be divided into two groups, those monomorphic between Silewah and Norin-PL8 and those monomorphic between Hokkai241 and Norin-PL8. The former patterns were detected by 11 markers (two markers on chromosome 1, three markers on chromosome 3, five markers on chromosome 4 and one marker on chromosome 8). The results indicate that the segments on chromosomes 1, 3, 4 and 8 were introgressed from Silewah into Norin-PL8. HokkaiPL4 is a sister line of Norin-PL8. Although HokkaiPL4 is also cold-tolerant, it is a little inferior to Norin-PL8. I examined the RFLP patterns of HokkaiPL4 and found that the segments on chromosomes 1 and 4 were introgressed from Silewah into HokkaiPL4 but that the segments on chromosomes 3 and 8 were not. The results suggest that the main cold tolerance gene is located on chromosome 1 or 4. 2. Association between introgression and cold tolerance BC_1F_4 plants of the cross Kirara397/Norin-PL8//Kirara397 were grown in a paddy field in 1993, a year in which there was a very cool summer. I examined the spikelet fertility and genotypes of RFLP markers for 43 BC_1F_4 plants. The results indicated that the introgressions on chromosomes 3 and 4 were associated with cold tolerance. The introgression on chromosome 8 was not associated with cold tolerance. I could not examine the association between the introgression on chromosome 1 and cold tolerance because the RFLP markers on chromosome 1 were monomorphic between Kirara397 and Norin-PL8. In order to verify the results, 92 BC_1F_5 lines of the cross Kirara397/Norin-PL8//Kirara397 were tested for cold tolerance and RFLP genotypes. Cold tolerance was evaluated on the basis of spikelet fertility of rice plants irrigated with deep water controlled at 19℃. For the markers on chromosomes 3 and 4, the mean spikelet fertility of the lines with the Norin-PL8 allele was significantly higher than that of the lines with the Kirara397 allele. On the other hand, for the marker on chromosome 8, the mean spikelet fertility of the lines with the Norin-PL8 allele was equivalent to that of the lines with the Kirara397 allele. The results indicated that the introgressions on chromosomes 3 and 4 were associated with cold tolerance. The association between introgressions and cold tolerance was examined also in F_9 lines between Hokkai241 and Norin-PL8. Out of 503 microsatellite markers used, 6 markers on chromosomes 1, 3 and 4 were polymorphic between Hokkai241 and Norin-PL8. For the markers on chromosomes 1 and 4, the mean spikelet fertilities of the lines with the Norin-PL8 allele were higher than that of the lines with the Hokkai241 allele by 8.5％ and 17.7％, respectively. On the other hand, for the marker on chromosome 3, the mean spikelet fertility of the lines with the Norin-PL8 allele was equivalent to that of the lines with the Hokkai241 allele. Therefore, the introgressions on chromosomes 1 and 4 were associated with cold tolerance in the population between Hokkai241 and Norin-PL8. The results showed that introgressions on chromosomes 1, 3 and 4 were associated with cold tolerance and that the introgression on chromosome 4 had the largest effect on cold tolerance. II. Map position and function of the cold tolerance gene 1. Interval mapping of the cold tolerance gene In order to determine the precise map position of the QTL for cold tolerance on chromosome 4, I developed advanced backcross progenies by using Kirara397 as a recurrent parent and conducted interval mapping. Log-likelihood (LOD) scores were relatively high from the center to the proximal end of the introgression and a maximum LOD score was observed between the RFLP markers R2737 and XNpb102, indicating that the QTL is most likely positioned between R2737 and XNpb102. I developed a set of recombinant inbred lines (RILs) from the recombinants in the segregating population and compared their cold tolerance. BT4-9-7 had the Norin-PL8 allele for R738 and had the Kirara397 allele for the other markers, and its cold tolerance was equivalent to that of Kirara397. Cold tolerance of BT4-76-2, which had the Norin-PL8 allele for R738 and R2737, was higher than that of Kirara397. This result suggests the presence of the QTL between R738 and XNpb102, coinciding with the LOD peak observed in the interval mapping. Cold tolerance of BT4-74-8, which had the Norin-PL8 allele for R740 and the Kirara397 allele for XNpb267, was equivalent to that of Kirara397, indicating that there is no QTL from R740 to the distal end. However, cold tolerance of BT4-70-1 was higher than that of BT4-74-8 and Kirara397. The genotype for the region from R740 to SCAB11 is different between BT4-70-1 and BT4-74-8. The LOD curve is even in this region, suggesting the presence of a hidden LOD peak. These results indicate the presence of another QTL between R740 and SCAB11. I found that there are two QTLs for cold tolerance in the long arm of chromosome 4 and designated the distal one and proximal one as Ctb1 and Ctb2, respectively. Cold tolerance of the RILs that harbor both QTLs was indistinguishable from that of the RILs that harbor either QTL, suggesting that the interaction of Ctb1 and Ctb2 is not additive. Minimum and maximum genetic distances between Ctb1 and Ctb2 were 5.6 cM and 17.2 cM, respectively. 2. Fine physical mapping of the cold tolerance gene For the fine mapping of Ctb1, 1,255 individuals of the segregating population were screened for recombination between the Ctb1 flanking markers and 49 RILs were developed. I selected 2 RILs that had the Kirara397 allele for the Ctb1 region (NS), 10 RILs that had the Norin-PL8 allele for R740 and the Kirara397 allele for XNpb267 (NA), 6 RILs that had the Norin-PL8 allele for XNpb267 and the Kirara397 allele for XNpb264 (NB), 7 RILs that had the Norin-PL8 allele for XNpb264 and the Kirara397 allele for SCAB11 (NC) and 2 RILs that had the Norin- PL8 allele for the Ctb1 region (NT). NS and NT were used as cold-sensitive and cold-tolerant controls, respectively. The spikelet fertility of NC was higher than that of NS, NA and NB and equivalent to that of NT, suggesting that Ctb1 is located around XNpb264. I found that two bacterial artificial chromosome (BAC) clones cover the region between XNpb264 and SCAB11. Based on the DNA sequences of the BAC clones, I developed simple sequence length polymorphism (SSLP) markers : BAC1, BAC22, BAC9, PNK2, BAC29 and BAC28. I found that one of the NC RILs, BT4-12-110-7, was heterozygous for the region between BAC1 and SCAB11. In order to test whether Ctb1 is located between BAC1 and SCAB11, I developed near-isogenic lines, BT4-12-110-7K and BT4-12-110-7P, that had the Kirara397 allele and the Norin-PL8 allele, respectively, for the markers between BAC1 and SCAB11. Cold tolerance of BT4-12-110-7P was significantly higher than that of BT4-12-110-7K, indicating that Ctb1 is located between BAC1 and SCAB11. To narrow down the Ctb1 region, I increased the number of RILs by screening 753 individuals of the segregating population and developed four more RILs. I genotyped the ten RILs using the SSLP markers. Cold tolerance of BT4-27-486-2, which had the Norin-PL8 allele for BAC9 and the Kirara397 allele for PNK2, was significantly higher than that of the cold-sensitive control, indicating that Ctb1 is located between BAC1 and PNK2. Cold tolerance of BT4-18-170-2 and BT4-29-279-6 were equivalent to that of the cold-sensitive control, while the cold tolerance of BT4-27-355-3 was significantly higher than that of the cold-sensitive control. BT4-18-170-2, BT4-29-279-6 and BT4-27-355-3 had the Norin-PL8 allele for BAC1 and the Kirara397 allele for BAC22. The results indicate that Ctb1 is located between BAC1 and BAC22. The physical distance between BAC1 and BAC22 is 56 kb. 3. Function of the cold tolerance gene Anther length is one of the factors involved in cold tolerance because anther length is correlated with amount of pollen. The anther lengths of Norin-PL8 and Hokkai241 were 2.85 mm and 2.13 mm, respectively, and the anther length of F_7 lines between Hokkai241 and Norin-PL8 were continuously distributed between those of the parents. The introgression on chromosome 4 was significantly associated with anther length in the F_7 population. In order to examine the association between QTLs for cold tolerance and anther length, I measured the lengths of 80 anthers from each RIL used for the interval mapping. The RILs harboring Ctb1 and/or Ctb2 produced larger anthers than did the RILs without Ctb1 and Ctb2. This result suggests that both Ctb1 and Ctb2 express cold tolerance by producing a large anther. I found seven open reading frames (ORFs) in the 56-kb Ctb1 region. Two ORFs coded for receptor-like protein kinases that play an important role in plant signal transduction. Two ORFs had F-box motifs. The F-box is involved in the ubiquitin-proteasome pathway. One ORF contained a BAG domain and a ubiquitin domain, suggesting that it plays a role in the physical linkage between the Hsp/Hsc chaperon system and the ubiquitin-proteasome. Functions of the remaining two ORFs are unknown. Of the seven ORFs identified in the Ctb1 region, five ORFs appeared to be associated with a signal transduction pathway or a ubiquitin-proteasome pathway. Since both pathways function in various aspects of growth and development, it is difficult to speculate about the ORF responsible for cold tolerance or anther length. Expression analyses and transformation experiments of the candidate genes are necessary for identification of the Ctb1 gene.

イネは穂ばらみ期に低温に遭遇すると, 小胞子の発育が阻害を受け, 不稔を発生する。冷夏に見舞われやすい北海道では, 穂ばらみ期耐冷性は重要な育種目標の一つである。北海道農業試験場（現北海道農業研究センター）では, 北海道へ外国稲の耐冷性を導入するために，インドネシア原産の熱帯ジャポニカ品種である Silewah へ北海道の育種系統である北海241号を連続戻し交配し, 中間母本農8号を育成した。本研究では DNA マーカーを用いることによって中母農8の穂ばらみ期耐冷性に関する量的形質遺伝子座（QTL）の遺伝解析を試みた。