Triple Test Cross (TTC) analyses were carried out to detect epistasis, and to estimate additive and dominance components of genetic variation in two seasons (kharif and spring/summer) for branches per plant, pod clusters per plant, pods per cluster, pod length, biomass, and harvest index in mungbean. The results revealed the absence of epistasis for all the traits in the kharif season but in contrast the presence of epistasis for pod clusters per plant and biomass in the spring/summer season. The partitioning of total epistasis revealed that both i type (additive × additive), and j and l type (additive × dominance, and dominance × dominance) interactions were significant with predominance effect of i type for pod clusters per plant and biomass. The additive and dominance gene effects were highly significant for branches per plant, pod length, and harvest index in both seasons; for pod clusters per plant and biomass in the kharif season; and for pods per cluster in spring/summer season. The magnitude of additive variance was higher for all these traits in both seasons except for pods per cluster and pod length in the spring/summer season, which exhibited a higher value of dominance than additive variance. The direction of dominance was observed towards fewer pod clusters per plant, small pod size, and high harvest index. The predominance of i type interaction and additive gene action for most of the traits in the present investigation suggests the occurrence of selection in the late generations in the segregating population for the improvement of seed yield through yield components in mungbean.
Variations of leaf form, i.e., differences in shape and size, among citrus varieties and the genotype × environment (GE) interactions influencing the variations were examined. Leaf shape was quantitatively measured by the score of the principal components of elliptic Fourier descriptors (EFDs). Leaf size was measured in terms of area and perimeter. The first four principal components of EFDs, which could explain over 90 % of the shape variations, were good measures of the length to width ratio, the position of the center of gravity, the curvature, and the degree of roundness. Nested ANOVA for the leaf form variations in nine citrus varieties indicated that the variations accounted for by the 1st, 2nd and 4th components were inherited. GE interaction of the leaf form was investigated using the data appertaining to seven genotypes from eight locations; the interaction was significant for all the principal components examined except the third. Although the interaction fitted very poorly to a joint linear regression analysis model developed from the data, it was effectively explained by an additive main effect and multiplicative interaction model where the score of the interaction principal component was highly correlated with the stability indices. Our results suggest that the genotype was the main source of variation in leaf shape, but not in size, and that the contribution of GE interaction was minor to both shape and size, although statistically significant at the 1 % probability level.
The suitability of background selection using DNA markers was assessed in a tobacco backcross breeding program for introgression of a potato virus Y (PVY) resistance gene (va/va) derived from a novel resistant source, Kerti No. 1, into the genome background of a commercial variety, Coker319. Ten randomly amplified polymorphic DNA (RAPD) and 76 amplified fragment length polymorphic (AFLP) markers specific to the Kerti No. 1 genome background (background markers) were identified based on the polymorphism analysis with Coker319 and they were used for background selection at loci not linked to va. Two markers flanking va on each side (linkage markers) were used for background selection of the va-carrying chromosome. In each backcross generation, 200 plants were developed and surveyed with only the background markers, while 300 plants were similarly treated with both the background and linkage markers in the BC3S 1 generation. One plant, where all the 86 randomly extracted background marker loci and 2 linkage marker loci were recovered in the recipient genotype, was obtained through only 3 backcrossings and 1 selfing. CKM1 (BC3S2) derived from this BC3S1 plant did not display any significant differences from Coker319 in terms of characters (e.g., yield, number of leaves, length and width of leaves, leaf-color change during the maturation period, yellowing rate in the curing process). In addition, CKM1 was similar to Coker319 in terms of aroma, taste and low smoke quality. These results demonstrated that the genomic background of CKM1 was almost identical with that of Coker319 and that background selection on both a carrier and noncarrier chromosome(s) by DNA markers significantly accelerated the progress of a backcross program in tobacco. CKM1 did not display any adverse effects on the line introgressing the PVY resistance derived from other sources. Therefore, Kerti No. 1 appears to be the most suitable source of PVY resistance for tobacco breeding programs.
Quantitative trait loci (QTLs) that affect sink size and ripening, which are often negatively correlated, were analyzed in two inbred lines of rice (Oryza sativa L.) derived from crosses between the semi-dwarf indica and japonica cultivars. Recombinant inbred lines (RILs) from Milyang23 (indica)/Akihikari (japonica), and back-crossed inbred lines (BCILs) from Sasanishiki (japonica)/Habataki(indica)//Sasanishiki///Sasanishiki were studied over a period of two years. A major QTL that was related to the number of spikelets per panicle was found in the same region of chromosome 1 in both populations. The indica allele increased the number of spikelets and reduced the ripening percentage. The indica allele of the QTL, which was found in both populations but at different locations on chromosome 6, also increased the number of spikelets per panicle, although to a lesser extent than in the case of the QTL on chromosome 1. In this instance, the number of spikelets per panicle did not have a negative effect on the ripening percentage. The increased number of spikelets produced by the QTL on chromosome 6 was due mainly to an increase in the number of primary rachis branches. In contrast, the effect of the QTL on chromosome 1 relied on an increase in the number of secondary rachis branches. In addition, dry matter production tended to increase when the QTL region of chromosome 6 belonged to the indica genotype. Therefore, both panicle structure and source productivity might contribute to the increase of the sink size without reducing the ripening percentage. We also found two loci on chromosomes 11 and 12 of the RILs that were associated with the percentage of incompletely filled spikelets, but which did not affect the sink size. It is considered that these QTLs could contribute significantly to the breeding of cultivars with increased sink size and increased ripening.
Quantitative trait loci (QTLs) that affect the accumulation of nonstructural carbohydrates (NSCs) in the leaf sheaths and culms of rice (Oryza sativa L.), enhancing subsequent grain filling, were analyzed in a population of back-crossed inbred lines (BCILs) from the cross Sasanishiki (japonica)/Habataki (indica)//Sasanishiki///Sasanishiki. QTLs affecting NSC accumulation were detected on chromosomes 1, 4, 5, 7, 11 and 12. Of these, the indica alleles of QTLs on chromosomes 7 and 12 increased NSC accumulation, but were detected for only one year in the two-year study. These loci also showed a strong effect on the number of days to heading, suggesting that their effect on NSC accumulation was linked to environmental conditions before heading. The QTLs on chromosomes 1, 4, 5 and 11 did not affect appreciably the number of days to heading. The QTL on chromosome 1 might be identical with the QTL that increases the number of spikelets per panicle in the indica allele, implying that it has pleiotropic effects. The indica allele of this QTL reduced NSC accumulation, which led to an increased percentage of incompletely filled spikelets. In contrast, the indica allele of the QTLs found on chromosomes 5 and 11 increased NSC accumulation and did not affect appreciably the number of spikelets. These indica allele QTLs also exerted some effect on the reduction of the percentage of incompletely filled spikelets, suggesting the possibility that the expression of these QTLs could improve grain filling in rice.
We have identified two genes, located approximately 2.5 kb apart in the soybean genome, that are closely related to the α subunit of β-conglycinin. One of these genes has been shown to encode the α subunit and the other is very similar to the α subunit gene (referred to as the “ α-related gene”). To determine whether the latter α-related gene expresses the α subunit, the structure of the chromosomal DNA region that contains the two genes was compared among soybean varieties exhibiting different expression levels of the subunit protein. We observed the presence of deletions of the α subunit and α-related genes in some varieties. In the naturally occurring variety Keburi, which accumulates the α subunit protein, the entire α-related gene was deleted in addition to the previously known deletion of the α′ subunit gene. On the contrary, the α subunit gene was deleted in Moshidou Gong 503 and Kari-kei 434. These structural analyses revealed that the α-related gene expresses the α subunit at least in Moshidou Gong 503 and Kari-kei 434. These two varieties do not harbour the α subunit gene, but carry the intact α-related genes and still express the α subunit.
In direct seeding of rice (Oryza sativa L.), rice seedlings are exposed to low oxidation-reduction (redox) potential soil, and to low temperatures in cold areas. In order to trace the genealogy of tolerance to low redox potential among Japanese rice cultivars, we selected 12 Japanese rice cultivars that had been released over the past century and determined the ability of their seedlings to tolerate immersion in low redox culture solution of Eh =−150 to −180 mv. Seedlings with coleoptiles and radicles of about 1 to 2 cm were incubated in low redox potential culture solution for 2 days at 18 and 30°C, and then in distilled water for 3 days at the same temperatures to examine the recovery abilities. The lengths of shoots and roots were measured after pre-germination, after low redox treatment, and after recovery from damage that occurred in the de-oxidized solution. At 30°C, the tolerance of low redox potentials was positively and significantly correlated with recovery from damage that occurred in the de-oxidized solution. The high tolerance of recent cultivars to these low redox potentials at 30°C appeared to originate from Fujisaka 5. However, at 18°C, the differences among the cultivars in their tolerance of low redox potentials and in their ability to recovery became small.
An incompletely dominant gene Ur1 is characterized by undulation of primary and secondary rachis branches. The gene increases spikelet number per panicle owing to increase of secondary branches. The Ur1/ Ur1, Ur1/+ and +/+ genotypes with sd1 (denoted by u, h and d, respectively, viz. three dwarfs) and those without sd1 (U, H and T, respectively, viz. three talls) were produced under the common genetic background of a japonica variety Taichung 65. The six genotypes were grown in 2000. Yield and its related traits were measured. Analysis of variance on yield indicated that the effect of Ur1 was significant whereas the effects of sd1 and the Ur1 × sd1 interaction were nonsignificant. Genotypes u and h had significantly higher yields than d; h showed the highest yield of all genotypes. The three talls had yields in the order of H ³ U ³ T (H > T). Regarding spikelet number per panicle, the genotypes were in the order of u = U > h ³ H > T > d, reflecting the principal effect of Ur1. Regarding panicle number per m2, the effect of Ur1 was not significant. The ripened-grain percentages of the six genotypes were almost in the reverse order of spikelet number per panicle. The 1000 grain weight of the genotypes was in the order of U @ u £ h £ H £ d < T. Sink size-2 (single grain weight × fertilized-spikelet number per m2) showed a high positive correlation with yield. As for LAI, there were no significant differences among the six genotypes. The total weight per m2 at maturity for the three dwarf genotypes were in the order of h ³ u ³ d (h > d). As for harvest index, u and h were higher than d. Similar results were obtained for the three dwarf genotypes under three fertilizer levels in 1998. Thus, Ur1 enlarged sink size through increase of spikelet number per m2 resulting in higher yield. Consequently, Ur1 may be utilized for developing high yielding inbred varieties and F1 varieties at the Ur1/ + genotype.
Rice (Oryza sativa L.) varieties used for brewing sake are commonly characterized by traits such as large grain size with white-core (an opaque structure inside the rice grain). A linkage map was constructed using doubled haploid lines derived from the cross of Reiho (a cooking variety) and Yamada-nishiki (a sake-brewing variety). Random amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP) and simple sequence repeat (SSR) marker systems were employed in QTL analysis. A total of 145 markers were identified and mapped on rice chromosomes. QTLs for plant and grain characters were detected by interval mapping and single point analysis. Several QTLs with a significant contribution were identified for important sake-brewing characters including grain size, grain shape, white-core grain rate and protein content. Several QTLs simultaneously affected the grain weight, width and thickness, while QTLs for the grain length independently affected the grain size. QTLs for the white-core grain rate did not affect the grain size, although one QTL for the white-belly grain rate simultaneously affected the grain weight, width and thickness. Several QTLs were detected for the protein content in both brown and polished rice. One QTL on chromosome 4 that was effective for the decrease of the protein content in polished rice showed a positive relation with the grain length. One QTL with the largest effect on the grain length on chromosome 11 did not contribute to the decrease of the protein content in polished rice. Therefore, it is suggested that the grain length QTL on chromosome 4 might control not only the grain shape but also the internal structure related to the milling efficiency and/or location of the storage protein.
Chromosome substitution series are useful tools for precise mapping of quantitative trait loci (QTLs) and for the evaluation of gene action as a single factor or in the case of gene interactions. In this study, we produced reciprocal sets of chromosome segment substitution lines (CSSLs) between Asominori (Japonica) and IR24 (Indica) in cultivated rice, Oryza sativa L. A set of chromosome segment substitution lines carries a single or a few donor chromosome segments overlapping with neighboring segments in a uniform genetic background. Reciprocal sets were developed by repeated backcrossing to respective recurrent parents and DNA marker-assisted selections. The sets consisted of a series of 70 lines of Asominori CSSLs with IR24 genetic background (designated as ‘IAS’) and a series of 91 lines of IR24 CSSLs with Asominori genetic background (designated as ‘AIS’). These substitution series represented most of the genome of the donor parents on the alien genetic backgrounds. Each of the substitution lines was nearly isogenic to the recurrent parent: the average proportions of recurrent parent’s genome were 93 % and 91 % in the IAS and AIS, respectively. Consistent reciprocal QTL effects for heading date and grain size were observed in both IAS and AIS. The reciprocal substitution set appeared to be an efficient and systematic method for QTL precision analysis.