Efforts to identify genes underlying tolerance to drought stress have been mostly focused on model species and major cereal crops such as rice and maize. However, legumes are known to possess high levels of drought tolerance. Legumes are a very large and diverse family of plants that include a number of important crop species. Historically, there has been a focus of research effort on the major legume crop species, such as pea and soybean. However, more recently Medicago truncatula was adopted as a model plant for the study of legume biology due to its small genome (450 Mb). The legume crop chickpea is taxonomically close to Medicago truncatula and displays an ability to endure and grow in relatively low soil water contents. The root system is known to be a primary sensor of drought stress and may play an important role in drought avoidance. Thus, generating and analyzing expressed sequence tags (ESTs) from the roots of crops such as chickpea is highly relevant approach to understanding the biology of drought tolerance and avoidance.

In this study, ESTs were generated by subtracting the expressed genes of a local landrace accession (ICC 4958) from those of a local variety (Annigeri), both which are considered to possess different sources of drought avoidance and tolerance. This subtractive suppressive hybridization approach (SSH) is known to help in the isolation of differentially expressed sequences. The ESTs derived from this process are now available with Genbank (CK148643-CK149150) or can be accessed from a relational database: http://www.icrisat.org/gt1/cpest/home.asp.

A total of 4000 sequence reads were generated from the subtractive suppressive library (SSH). Most of these cDNA are likely to have come from the landrace ICC4958 that served as the tester in the SSH process. Only those sequences that had more than 170bp were processed, while other quality parameters were visually verified through the trace files, leading to 2858 ESTs (with an average length of 492 bp) being chosen for further analysis. Processed ESTs were screened against multiple plant DNA databases (A. thaliana, M. truncatula, G. max, Z. mays and O. sativa in the TIGR-Unique Gene Indices, that represent clustered assemblies of EST sequences and dbEST) using BLASTn and tBlastx. Sequences were clustered into contigs using Sequencher v.4, licensed clustering software. The 2858 ESTs analysed clustered into 210 contigs and 267 singletons. The consensus sequence from contigs was used in searches for conserved domains using RPS-BLAST or reverse-position-specific BLAST that searches the Conserved Domain Database (CDD). The CDD database is a compilation of multiple sequence alignments representing protein domains conserved in molecular evolution and is a useful resource for annotation. Amongst the 210 contigs, 77% (162 of 210) found homology, while 67% (178 of 267) of singletons found homology.

Since it is known that ICC 4958 and Annigeri are very closely related, a subtractive approach was expected to identify only a moderate number of differentially expressed transcripts. A range of bioinformatic analyses were carried out on the ESTs generated from this SSH library. These included putative annotation of the sequences, grouping them into functional categories, clustering, searching for repeats and conserved domains, and primer design. All the information available on the chickpea variety and the sequence data and bioinformatic analysis has been made available through the Chickpea Root EST Database. All of the data is available through web-interfaces; the database allows data mining through a set of query pages and users may also search their sequences against the EST database using the Blast option. The database architecture is based on a SQL-Server (v.7) database server at the backend with Windows 2000 IIS web server and web pages written in ASP. The Blast output analyzer tool, which was developed as part of this work, is also available for download. This tool is a BLAST output-parsing program, which saves the parsed output into a database. The parsed data can be retrieved from the database based on nucleotide/protein scores, description or sequence identity.

This database provides the first insight into the genes that may be associated with root development and stress tolerance in chickpea. ESTs with putative relationships to signal transduction and transcriptional factors, proteases and lipid transfer proteins, transporters and heat-shock proteins, and, an enzyme involved in the synthesis of trehalose which is a known membrane protectant associated with dessication tolerance in resurrection plants were isolated. These annotated ESTs provide a useful resource as candidates for functional genomics and candidate gene mapping of drought tolerance.This database will be an important resource for chickpea genomics scientists and molecular breeders, and provide a platform for studies on legume genomics and drought tolerance by the wider community.