Characterization of the Arabidopsis proteome by fluorescent protein tagging

Amitabh Mohanty¹, Vitaly Citovsky², Georgia Drakaki⁴, David Ehrhardt³, David Jackson¹, Shijun Li³, Brigitte Paap², Natasha Raikhel⁴, Sue Rhee³, Narasimha Chary Samboju⁴ and Guo-wei Tian².

¹Cold Spring Harbor Laboratory, Cold Spring Harbor, NY; ²Department of Biochemistry and Cell Biology, State University of New York, Stony Brook, NY; ³ Department of Plant Biology, Carnegie Institution of Washington, Stanford, CA; ⁴Department of Botany and Plant Sciences, University of California, Riverside, CA

Arabidopsis has 25,498 predicted genes, however, only about 70% of the genes have been assigned to any functional category, and less than 10% of the genes have been verified experimentally. There are only very limited reports of systematic analysis of genes of unknown function. Moreover, for the majority of the annotated proteins, the sub-cellular location and/or expression pattern is unknown. Thus, a large gap remains in our understanding of the function of a very significant portion of Arabidopsis gene products, many of which are plant specific and likely play fundamental roles in plant biology. Our consortium aims to fill this gap by systematically analyzing a large number of the functionally unassigned genes by Fluorescent Tagging of their Full-Length Protein products (FTFLP). We plan to characterize the Arabidopsis proteome by seeking three specific aims: 1. Selection and sub-cloning of Arabidopsis genes of unknown function with their potential native regulatory sequences. 2. Fluorescent tagging (using Yellow or Cyan fluorescent proteins) of the genes and insertion into Arabidopsis plants. 3. Analysis of the expression patterns and sub-cellular localization of the YFP/CFP-tagged proteins in planta. Based on the most recent Arabidopsis genome annotations, more than 8,000 genes are annotated as "unknown" or "putative" protein. We have selected a short list of ~800 for initial characterization.

As a pre-requisite we have constructed a reference localization map with ~20 Arabidopsis proteins with known sub-cellular localization patterns. This map will serve as a reference for co-localization studies. For each full-length gene, including upstream and downstream native regulatory sequences are being amplified by PCR from genomic DNA. We use these PCR fragments together with YFP or CFP fragments as templates for a second round of PCR (Triple template overlap PCR) for the insertion of the tag into the gene. Under the control of native promoters, the expression levels of some proteins could be too weak for visualization, so we are also using vectors with multiple copies of the 35S enhancer, to enhance the native expression of target genes. Our recent data from this project will be presented.