The causal gene of a novel small and round seed mutant phenotype (srs3) in rice was identified by map-based cloning and named the SRS3 gene. The SRS3 gene was grouped as a member of the kinesin 13 subfamily. The SRS3 gene codes for a protein of 819 amino acids that contains a kinesin motor domain and a coiled-coil structure. Using scanning electron microscopy, we determined that the cell length of seeds in the longitudinal direction in srs3 is shorter than that in the wild type. The number of cells of seeds in the longitudinal direction in srs3 was not very different from that in the wild type. The result suggests that the small and round seed phenotype of srs3 is due to a reduction in cell length of seeds in the longitudinal direction. The SRS3 protein, which is found in the crude microsomal fraction, is highly expressed in developing organs.

The alpha subunit of plant heterotrimeric G proteins (Galpha) plays pivotal roles in multiple aspects of development and responses to plant hormones. Recently, several lines of evidence have shown that Galpha participates in brassinosteroid (BR) responses in Arabidopsis and rice plants. In this study, we conducted a comprehensive analysis of the roles of the rice Galpha in the responses to BR using a defective mutant of the Galpha gene, T65d1. Decreased sensitivity to 24-epi-brassinolide (24-epiBL) in the T65d1 mutant was observed in many processes examined, e.g. in the inhibition of root growth and the promotion of coleoptile elongation. The T65d1 mutant also showed similar phenotypes to those of BR-deficient mutants, such as the specifically shortened second internode and the constitutive photomorphogenic growth phenotype under dark conditions. However, a negative feedback effect by 24-epiBL on the expression of BR biosynthetic genes was observed in the T65d1 mutant, and the levels of BR intermediates did not fluctuate in this mutant. To determine the epistatic relationship between the T65d1 mutant and d61-7, a weak allele of a rice BR receptor mutant, the two mutants were crossed. The T65d1/d61-7 double mutant showed no epistasis in the elongation inhibition of the internodes, the internode elongation pattern, the leaf angle and the morphological abnormality of leaf, except for the vertical length of seed and the seed weight. Our results suggest that the rice Galpha affects the BR signaling cascade but the Galpha may not be a signaling molecule in BRI1-meditated perception/transduction.

We used site-directed mutagenesis to engineer two constitutively active forms of the alpha subunit of a rice heterotrimeric G protein. The recombinant proteins produced from these novel cDNAs had GTP-binding activity but no GTPase activity. A chimeric gene for a constitutively active form of the alpha subunit was introduced into the rice mutant d1, which is defective for the alpha-subunit gene. All the transformants essentially showed a wild-type phenotype compared with normal cultivars, although seed sizes were substantially increased and internode lengths also showed some increase.

It has been shown that the disruption of the α-subunit gene of heterotorimeric G-proteins (Gα) results in dwarf traits, the erection of leaves and the setting of small seeds in rice. These mutants are called d1. We have studied the expression profiles of the transcripts and translation products of rice Gα in ten alleles of d1 including five additional alleles newly identified. By RT-PCR, the transcripts of the Gα gene were detected in the all d1 alleles. By western blot, the Gα proteins were not detected in the plasma membrane fractions of the d1 alleles with the exception of d1-4. In d1-4, one amino acid change in the GTP-binding box A of the Gα protein was occurred and even in this case the Gα protein was only just detectable in the plasma membrane fraction. Given that the Gα protein did not accumulate in the plasma membrane fraction in d1-8 which has a deletion of just a single amino acid in the Gα protein, it is likely that a proper conformation of the Gα is necessary for accumulation of Gα protein in the plasma membrane. Nine alleles of d1 showed a severer phenotype whilst d1-4 exhibited a mild phenotype with respect to seed size and elongation pattern of internodes. As brassinosteroid signaling was known to be partially impaired in d1s, the sensitivity to 24-epibrassinolide (24-epiBL) was compared among d1 alleles in a T65 genetic background. Only d1-4 showed responses similar to wild type rice. The results show that the d1-4 mutant is a mild allele in terms of the phenotype and mild hyposensitivity to the exogenously applied 24-epiBL.

The alpha subunit of heterotrimeric G-proteins (G alpha) is involved in a broad range of aspects of the brassinosteroid (BR) response, such as the enhancement of lamina bending. However, it has been suggested from epistatic analysis of d1 and d61, which are mutants deficient for G alpha and the BR receptor BRI1, that G alpha and BRI1 may function via distinct pathways in many cases. In this study, we investigated further the genetic interaction between G alpha and BRI1. We report the analysis of transformants of T65d1 and T65d1/d61-7 into which were introduced a constitutively active form of G alpha, Q223L. The application of 24-epi-brassinolide (24-epiBL) to T65d1 expressing Q223L still resulted in elongation of the coleoptile and, in fact, it was enhanced over the wild-type plant (WT) level in a concentration dependent manner. In T65d1/d61-7 expressing Q223L, the seed size was enlarged over that of d61-7 due to activation of G alpha. These results suggest that Q223L is able to augment the BR response in response to 24-epiBL and also that Q223L functions independently of BRI1 in the process of determining seed morphology, given that Q223L was functional in the BRI1-deficient mutant, d61-7.