AUTHOR COMMENTARIES

Would you please sum up your 2002 Science paper, "A draft sequence of the rice genome (Oryza sativa L. ssp japonica)," for our readers?

Syngenta’s 2002 Science paper on the draft sequence of the rice genome describes the creation and analysis of a DNA sequence dataset covering greater than 99% of the rice genome. It also reviews the assembly of that dataset into mapped sequences representing the majority of protein coding genes. Rice was the first crop species to be sequenced to this degree. The approach the authors used demonstrated that lower cost "draft" sequencing methods are applicable to relatively large genomes.

Another group, the Beijing Genomics & Bioinformatics Institute (BGI) used a similar approach to create the draft sequence of another rice variety which was published in the same issue of Science. The International Rice Genome Sequencing Project (IRGSP, an international effort being executed in the public sector) was projected to be complete by 2008 and was using an approach that resulted in more complete and more accurate coverage of the genome.

What were the major findings and implications from this paper? 

There were three major findings from this publication. The first was about the technical approach chosen. Before this research project was executed and published, it wasn’t clear if a random fragment "draft" sequencing method could be used successfully on such a large genome. Prior to this effort, the so-called "shotgun sequencing method" had only been applied to smaller genomes, like those of single-celled organisms. This publication brought the understanding that the majority of the genome, regardless of size, could be sequenced and mapped using a random-fragment approach at a significant cost savings over map-based approaches. Such random-fragment approaches are now the most commonly used method in sequencing large plant genomes.

The second major finding was the fact that the gene number was considerably lower than initially estimated for rice. Original estimates were based on the limited amount of genome sequence information generated by the IRGSP on rice chromosome 1. The predicted gene numbers reported in preliminary project summaries on the internet were as high as 90,000. Our analysis predicted the gene number to fall between 32,000 and 50,000 genes, depending on informatics criteria and confidence levels used. IRGSP also lowered the estimated number from 90,000+ genes in its early publications to ~50-60,000 genes and lower as data became available and the final publication of the map-based sequence was released. This discrepancy or variation in gene number was later reported by Jeff Bennetzen’s group to be due to counting repetitive elements as genes. Some specific repetitive elements can pick up portions of genes and amplify them in the transposition process.

A third important finding of this research was the similarity in the classification of rice genes relative to Arabidopsis genes. The two plants had a very similar distribution of genes in various functional classes. In addition, the classification of transcription factors encoding proteins that regulate gene expression was very similar between Arabidopsis and rice even though it was distinct from other sequenced species like nematodes, fruit flies, and yeast.

Our analysis suggested very similar genes would be found in all cereals, and further suggested that the extent of colinearity of the cereal genomes would be quite high, as predicted earlier from genomic mapping efforts. The similarity between the genes of various cereal species has since been supported by the fact that many genes, as well as regulatory regions, retain their function when transferred between related crops.

What are the applications for rice genome data?

Corn, wheat, and rice represent approximately 70% of total global crop production with over half a billion tons generated annually for each crop. Needless to say, these are very important crops to mankind. Given a growing human population, the yield of these crops will need to be improved on an ever-decreasing acreage of cultivatable land. Currently the yield gains for corn are around 1-2% annually, and these gains haven’t changed significantly for several decades despite higher levels of investment and manpower.

All cereals, such as corn, wheat, rice, and barley, share a common ancestor and are therefore closely related in genes that control traits of interest to producers and consumers. The genome of rice will help elucidate gene function in all cereals and will facilitate more rapid adoption of molecular breeding technology. Together these advances should help maintain or accelerate the yield gains needed to keep up with the growing population and the need for crop products in renewable energy developments.

How was this paper received by the community?

This publication was received by the community with mixed responses. The rice genome sequence data from the Syngenta project was not immediately released to the public, but instead donated to the IRGSP to allow the public project to be accelerated and released as a single higher-coverage assembly. The sequence was also provided to the BGI in Beijing at a later date to allow them to analyze the assembly, compare the japonica rice variety to the indica variety, and publish their results (along with a deposit of all Syngenta raw data into the GenBank trace file database). Individual academic researchers could gain access to the Syngenta sequence assembly under a specific agreement.

Some academic researchers preferred that the data for the entire sequence would have been submitted immediately to GenBank. Others understood the approach in which a private company provided data and funding which resulted in the acceleration of the availability of both draft and final sequence data. In the end, the IRGSP effort was both altered to produce a draft sequence and accelerated to completion several years ahead of the planned completion date of 2008.

Both the Syngenta draft japonica sequence and the BGI indica draft sequences were reported to be incomplete in the final map-based sequence of the rice genome published by the IRGSP. This claim has since been refuted by a recent study comparing the draft sequence with the map-based sequence (Matsumoto T, et al., "The map-based sequence of the rice genome," Nature 436[7052]: 793-800, 11 August 2005). Both the random fragment and map-based sequencing approaches have specific inherent disadvantages. However the speed and cost-efficiency of the random-fragment sequencing approach has made it the method of choice for all new large genome projects since the completion of the human, Arabidopsis, and rice genomes.

What initially sparked your interest in this line of research?

The main driving force behind Syngenta’s interest in this project was the lack of genome sequence information for basic crops at the time this project was started in 1998. Even though the public rice genome project had been initiated in the early 1990s, only 0.5% of the genome had been released to public databases at the time this project was started, and even less for corn and wheat. The desire to adopt molecular breeding technology and enable reverse genetics approaches in cereals drove this project into a high-priority position. Rice has a much smaller genome than corn, and its genome is also very small compared to wheat, so it was chosen as the model crop species for cereal genomics efforts. The commercial applications, however, were mainly intended for corn.

Where have you taken this work since this paper? Where do you see this research going in the next 10 years?

This work has been used within Syngenta to identify candidate genes and molecular markers useful for commercial crop enhancement. It has also allowed us to gain experience in cereal genomics with a smaller genome model. The rice genome work helped Syngenta scientists develop specific gene expression microarrays that have been used for hundreds of experiments internally and with academic collaborators looking at development and responses to the environment in various cereals of commercial interest.

Over the next 10 years the completed rice genome sequence (both the Syngenta project as well as the public projects) will be used to help identify genes from other cereals and validate how these genes function. Since the genome of corn will be released in February of 2008, the conserved regions of rice and corn will help in identification of the complete set of genes present in these important cereal crops.

The conserved non-coding regions will be identified through comparative genomics approaches. It is likely that the knowledge of these conserved regions will change our understanding of what a plant gene really is, just as studies of the human genome have done for mammals. The genome sequences will also enable efficient whole genome expression technologies, enhanced molecular breeding for improved traits, and more efficient forward and reverse genetics. Genome sequences will serve a foundation role in many new approaches toward elucidation of the role of genes in development and environmental responses. The precise function of any given gene will still remain very challenging and require a large research effort to complete, but the genome sequence will certainly facilitate that effort.

Stephen A. Goff
Senior Syngenta Fellow
Syngenta Biotechnology Inc.
Research Triangle Park, NC, USA