| Plant development is controlled by spatio-temporal gene expression patterns. Dissection of gene expression patterns might explain the mechanisms under plant growth and development, and also promote genetic improvement of cereal crops. Rice is the staple food crop worldwide, and also the model species for fundamental research in monocot plants. Transcriptional analysis in rice not only helps us to understand the regulation of rice development but also provides hints to study many other crops. Plant architecture is mainly determined by its lateral branches. In rice, tiller branch and panicle branch which are produced at vegetative and reproductive stage respectively, largely determines its plant architecture and yield potential. Therefore, they are regarded as important agronomic traits to be genetically improved in rice breeding. In this study, the gene expression profiles covering the entire life cycle of rice were constructed by dissecting the transcriptomes of the 39 organs, and then the data was compared with the Arabidopsis’ s transcriptomes. A gene regulatory network containing small RNAs and transcription factors which control rice lateral branches development was constructed and the function of the genes within the network had been further investigated. The major results are as follows:1. A gene expression atlas covering 39 organs was constructed by dissecting the transcriptomes of two indica varieties Zhenshan 97 and Minghui 63 using Affymetrix Gene Chip Rice Genome Array. Many approaches were used to evaluate the data quality. The data is an important platform for rice functional genomics.2. The developmental relationship among organs was revealed by the similarity of their transcriptomes. Strikingly, it was found that stamen had a unique pattern which differed from the patterns of other organs and callus had similar gene expression profiles with that to roots.3. A total of 2667 genes were differentially expressed in the developing panicles covering four consecutive stages. Among them, two groups of genes showed the opposite expression trends from early to later stage. The expression levels of the first group genes were gradually decreased from early to later stage, while the secondary ones showed the completely opposite trends. Many known regulators of panicle branch, and three SPLs targeted by mi R156, were involved in the first group, indicating that the genes in this group might be related to branch development. While many MADS-box genes related to flower development were involved in the secondary group. Transcriptional regulation should play a vital role in rice panicle development since transcription factors were both enriched in the two groups.4. 2376 tissue-specifically expression genes and 7276 constitutively expression genes were identified. A total of 19 most stably genes were found. They had constant expression levels regardless of tissue or cell types or even genetic background, environmental clues and exogenous treatments.5. The relatedness of organs between rice and Arabidopsis was revealed by analyzing the similarities of their transcriptomes and 11 reciprocally matching organs were found. The expression levels of orthologs were conserved in the 11 reciprocally matching organs in rice and Arabidopsis, however the expression patterns of more than half of the orthologs were divergent in two species, suggesting the functional divergence of many orthologs.6. Comparison of the tissue-specifically and constitutively expressed genes between rice and Arabidopsis revealed that the tissue-specifically expressed genes were usually species specific as well, while the constitutively ones were usually conserved in both of the two species. It was also found that the orthologs with tissue-specifically expression patterns evolved more rapidly than the ones with constitutively expression patterns.7. The number of tillers was greatly increased by overexpressing mi R156, but the panicle branches were greatly retarded in the overexpressors. The opposite phenotypes were caused by down regulating the mi R156. Thus, mi R156 promoted tiller branch but negatively regulated the activity of inflorescence meristem. As the expression of SPL7, SPL14 and SPL17, all of which were targeted by mi R156, decreased gradually from early to later stages in rice panicle development, we speculated that these three SPL genes might regulate panicle branch development.8. miR529 also regulated SPL genes in rice. The phenotypes which were caused by overexpressing mi R156 could be also observed in the mi R529 overexpressors.9. The rice plant architecture was not changed in spl7 single mutant. However the RNAi lines of SPL14 and SPL17 were reminiscent of the overexpressors of mi R156 and mi R529, of which tiller and panicle branches were increased and reduced respectively. Double RNAi lines of SPL14 and SPL17 showed stronger phenotype than either of the single RNAi lines, indicating the redundant roles of SPL genes in regulating rice development. Both tiller branches and panicle branches were reduced by overexpressing anyone of SPL7, SPL14 and SPL17. The number of secondary branches of each primary branch was also reduced, indicating that SPL genes positively regulated spikelet specification. Moreover, the phenotypes of the mi R156 overexpressors could be partially rescued by overexpressing SPL genes.10. The number of both primary and secondary branches per panicle but not the number of tiller branches was reduced by overexpressing mi R172 and the opposite phenotypes were caused by down regulating mi R172, suggesting that mi R172 negatively regulated inflorescence meristem activity, but positively promoted spikelet specification. The RNAi lines of AP2-like genes, SNB and Os TOE1, which were targeted by mi R172, were reminiscent of the mi R172 overexpressors. While the overexpression lines of SNB and Os TOE1 showed similar phenotype to the plants with lower mi R172 expression, indicating that mi R172 might regulate panicle development through its target genes SNB and Os TOE1.11. AP2-like proteins containing EAR motif, were transcriptional repressors in plants. They interacted with the transcriptional co-repressor ASP1 in vitro and in vivo. This interaction was mediated by the EAR motif of AP2-like protein and the CTLH domain of ASP. The activity of ASP1 was essential for the function of AP2-like genes in regulating panicle branch development.12. Gene expression analysis revealed that mi R156 and mi R172 had complementary expression patterns. The expression of mature mi R172 and its two precursors, pri-mi R172 b and pri-mi R172 d were elevated in the SPL overexpressors. By means of approaches, including Ch IP, yeast-one-hybrid and EMSA, SPL14 was revealed to directly target the promoter of mi R172 to regulate its expression. Genetic analysis further confirmed that SPL genes controlled spikelet specification through mi R172 pathway.13. PAP2/MADS34 also promoted spikelet specification by repressing RCN in rice. Gene expression analysis showed that SPL gene positively regulated PAP2/MADS34, and SPL14 directly bound to the promoter of PAP2/MADS34, revealed by Ch IP and EMSA, suggesting that SPL genes regulated spikelet specification also by affecting PAP2/MADS34-RCN pathway. This was further supported by the fact that RCN1 recovered the abnormal spikelet specification caused by overexpressing SPL genes.14. A large number of genes were differentially expressed in the young panicle of mi R156 overexpressors compared to the wild type. Many known regulators of panicle branches were involved, such as LAX1 and RFL. Moreover, these genes were co-expressed with SPL genes, suggesting that SPL genes might control panicle branch development by interacting with these genes. Genetic analysis showed that the LAX1 and RFL were essential for SPL genes functioning. Yeast-one-hybrid and EMSA revealed that SPL14 bound to the promoter of LAX1.15. Ghd7 regulated the expression of SPL genes. Genetic evidences demonstrated that SPL genes were essential for the roles of Ghd7 in regulating plant height and panicle branches, but not heading date. |
PDF Full Text Request