Report on Session 6: Expression Profiling

The session was mainly focused on the application of microarrays and proteomics to functional genomic studies in rice. There were 4 speakers Dr Rudy Dolferus from CSIRO Plant Industry in Canberra, Dr Naoki Kishimoto and Dr Junshi Yazaki from the National Institute of Agricultural Biology (Japan), and Dr Nijat Imin from the Australian National University.

The first and last speakers used different approaches in order to understand a major problem facing rice growers in temperate regions; low temperature induced male sterility. When rice undergoing panicle development is subjected to temperatures lower than 15oC male sterility can occur, and there are currently no strongly cold tolerant rice varieties commercially grown in Australia. Little is known about the mechanism and breeding efforts have been largely unsuccessful. It is thought that cold induced sterility is similar to drought induced sterility, and therefore may involve the hormone ABA. The sterility is very stage specific and occurs at the tetrad to young microspore stage of anther development.

Dr Rudy Dolferus utilized a genomic approach in order to identify genes whose expression is altered during cold treatment of rice panicles. Two microarrays were constructed, one containing 18,400 clones from cDNA libraries covering anther development, and another one containing 11,000 consisting of clones taken only from the tetrad stages of anther development. Rice undergoing panicle development were subjected to cold temperatures and anther samples were taken at different developmental stages. It was found that there were many genes that possessed differentially expression when cold treated samples were compared against control samples, and that there were differences in the extent and type of differentially expressed genes observed between the different developmental stages. Due to difficulties in being able to harvest tissue samples at the exact same stages of anther development, many experiments were repeated that narrowed down the total number of clones to 523. Many of these have been sequenced and so far a number of transcription factors, protein kinases and genes encoding metabolic enzymes have been found. Surprisingly a large number of unidentified genes have been uncovered, that is probably a reflection of the limited knowledge currently available about anther development and biochemistry. Dr Dolferus then went on to outline his method of amplify small amounts of mRNA from anther tissue sample for use in microarray hybridisations. The successful 200 to 1000 fold amplification of RNA samples will allow more precise developmental stages to be studied in future.

Dr Nijat Imin used a proteomics approach to the problem of cold induced male sterility. Proteomic is the study of the protein complement of a cell or tissue expressed in the genome at any timepoint and is analyzed using large scale protein separation and identification technologies. The advantages of proteomics is the ability to detect and identify post-transcriptional modifications that can be crucial for protein function, and the ability to study protein-protein interactions. Using two-dimensional gel electrophoresis protein over 3000 protein spots were reproducibly detected when looking at 6 different developmental stages of rice anthers. Over 350 of these proteins were subjected to MALD-TOF MS and putative identities assigned to more than 100. It was found that 40% of these proteins occurred as different isoforms, 25 were metabolic enzymes whereas 20% were stress-related. Protein differential display of at different stages of anther development was then performed. 150 proteins spots were found to be differentially expressed, and 44 were identified. Some of these proteins were found to be tapetum specific expressed at the early stages of development. Proteins associated with sugar metabolism, cell elongation and cell expansion were found to be up-regulated at the late stage. When anthers were subjected to cold treatment 36 proteins were differentially regulated 32 upregulated and 4 downregulated.

There was one question resulting from this presentation.

Q1) Of the 40% of proteins found to have different isoforms what kind of modification did they possess. A1) There were true modifications ie., same gene, and these probably contained altered phosphate and carbohydrate groups. There were some modifications that came from different genes. However we need to do more analysis to determine the exact type of modifications.

Dr Naoki Kishimoto presented work from the rice full-length cDNA project. The role of the project was to isolate full length rice cDNAs from all expressed rice genes. Possessing full length cDNA aids in the correct annotation of the rice genome, provide information on the structure of the transcripts as well as a source of clones for functional genomics projects including microarray and over-under expression of rice genes. It also allows for comparative genomics analyses with Arabidopsis and other Eukaryotes. The project constructed more than 30 cDNA libraries from the rice cultivar Nipponbare, ranging from different tissues such as seeds and calli to rice grown under different environmental stresses. 28,469 unique clones were sequenced. Over 75% had homology to other genes present in the databases (<10^-10 Blast N), and 94% were mapped on the genome. From the 28,000 clones it was found that there were 19,000 transcriptional units (TU), the other 9,000 clones represented different splice variants. Surprisingly only 50% of the TU were present in the Indica genome sequence as of April 2002. The set of 19,000 transcriptional units were then compared to different Eukaryotes, and to Arabidopsis. It was found that 33% of the TU were rice specific. It was revealed that rice did not contain any TIR type resistance genes, although there were 70 NBS-LRR type resistance genes present. The data from this project will soon be released on the KOME web site (Knowledge-based Oryza Molecular-Biology Encylopedia).

There were two questions resulting from this presentation:

Q1) What proportion of the total number of rice genes does your 19 TU represent? A1) It is thought that there are approximately 40,000 unique rice genes, so about 50%.

Q2) Was the discrepancy between the 28,000 unique clones and the only 19,000 TU all due to splice variants. A2) 2,500 TU out of the 19,000 possessed altered splicing. However some of the 2,500 TU had more splice variants than just 2, with 44 being the highest number observed.

Dr Junshi Yazaki presented work on the Rice Expression Database (RED). Microarrays were constructed containing 8987 unique cDNA sequences obtained from the Rice Genome Research Program. Microarrays were printed and distributed to 64 research groups and the results deposited in the database. The database contains over 600 different experiments, and this database will be made available to the public in December (http://cdna01.dna.affrc.go.jp/RMOS/). This database enables people to do blast searches on the genes present on the microarray and to show its genes structure and map position, it also enables expression profiling and gene clustering to be performed using selected experiments. Clusters of genes can be searched common cis regulatory elements. Dr Yazaki then went on to present results from a specific experiment looking at the cross talk between the hormones ABA and GA. Rice callus tissue was treated with either 50uM GA3 or 50uM of ABA were compared to untreated callus and gene expression changes detected using the 8987 rice gene microarray. It was found that 88 changes were detected upon addition of GA and 71 upon application of ABA. There 16 genes in common between the two treatments. 15 of these genes were suppressed in the GA treatment and induced in the ABA treatment. On large group of genes in this group were the Late Embryo Abundant (LEA) proteins. Only one gene was upregulated in both the GA and ABA treatments and this genes is currently uncharacterized.