SEAMEO SEARCA Biotechnology Information Center
The Network News, Vol. IX No. 11
30 November - 06 December 2005 issue

Imagine the diversity of rice that the International Rice Research Institute (IRRI) conserves in the International Rice Genebank. The Philippines based repository, responsible for safekeeping all known types of rice, contains more than 100,000 strains and varieties (each is referred to as an "accession"). Many of these comprise a mixture of different genotypes. Each rice genotype - that is genetic makeup that defines each type of rice - has an estimated 50,000 genes. Every genes comes in an unknown number of different versions, known as alleles, and each allele may change the way the rice looks or grows or tastes. Consider the incalculable number of different possible combinations of all the different versions, and you begin to comprehend the diversity of rice.

Try a simple calculation, assuming that only two alleles of each gene actually work: write down the number "! 1" and then write 15,000 zeros after it. Equivalently, say "million" a thousand million times (it'll take you 12 years without sleeping). Give or take a few thousand zeros, that's approximately the number of combinations of alleles that might make a recognizable rice plant. Then consider the enormous complexity of interacting biochemical reactions that drive the life of any organism - each allele may have a different effect on any one of the thousands and thousands of biochemical steps. Changing one step produces a series of cumulative effects, altering each subsequent step and, ultimately, the overall biochemical process. The point is that a seemingly genetic difference can produce significant differences in the end product. Each gene affects many traits and each trait is controlled by many genes.

Rice agriculture depends on this diversity. If a new rice disease appears, researchers can search the genebank for resistant varieties. The knowledge required to make rice more t! olerant of drought, for example, exists within the alleles in the collection. The genebank contains the diversity of alleles we need to respond to changes in climate, consumer expectations, agricultural technologies and government priorities.

The entire genebank collection may contain samples of most working versions of each rice gene. The full value of the collection is being, and will be, realized through plant breeding - combining the best alleles from different accessions to create superior new combinations of the traits needed by farmers and consumers. In this way, researchers can breed nutritious, high-quality, high-yielding rice varieties that are resistant to pests and diseases and tolerate stresses such as drought, flooding, low or high temperatures and poor soils.

This seems simple enough in principle, but leaves us with some burning questions. How can we identify the "best" allele of each gene? When a new disease appears, how can we know which alleles offer resistance to that disease? And once we know which alleles, how can we find which of the genebank's more than 100,000 accessions contains them? The challenge is formidable. We are yet to discover the function of most rice genes, or which alleles are possible for most of the genes.

Compounding the difficulty, much of the genetic variation is "hidden" in two ways. First, the effect of an allele depends on the genetic background - the genetic composition of the rest of the genome - and may not be expressed in the accessions that contain it. (The rice genome is the complete set of genetic material contained in, and responsible for, a rice plant.) Second, even where an allele is expressed, it takes a lot of research to tease out its effect from the effects of all other genes in the genome. Finding the unknown valuable alleles in the collection is called allele mining. Discovering all there is to know about the genetic diversity of rice is way beyond the capacity of current technologies. The necessary first step to actually mining for new alleles in the genebank collection is to decide which part of the genome we should researchers look at? Discovering the important genes involves an intensive series of genetic analyses of a small, carefully selected set of genotypes. This area of functional genomics, or gene discovery, allows us to decide which parts of the genome determine agronomic traits of interest. The answer depends on which traits we are interested in - grain quality, nutritional value, disease resistance, tolerance of poor soils and so on. The output of this research is a set of "candidate genes" - genes that we believe may have a certain functional significance.

Having chosen the candidate genes for exploration, we can start the serious business of allele mining - discovering new alleles at the selected genes. This means working through the collection to find all the alleles of these selected genes. Researchers can't just star! t with the first accession and work through the collection. Such an approach would be inefficient, since the second accession, for example, might be similar to the first at the chosen genes, so analyzing that second accession wouldn't give us much additional information. Instead, we begin by choosing a subset of highly distinctive accessions. This subset i know as a "core collection".

To choose the best core collection, researchers collect a wide range of evidence on diversity, then sample accessions representative of this diversity. One easy generic factor is geographic origin. Traditional varieties from different parts of the world have had an independent history of domestication for thousands of years, and are therefore likely to show differences across the whole genome. This way, researchers can discover at least the majority of new alleles in a relatively small number of accessions.

However, even a good core collection won't allow us to discover all possible all! eles. Plant breeders are familiar with the concept that breeding is a "numbers game". Breeders need to screen large numbers of plants in order to find the rare valuable genotypes. The same applies to allele mining - if a valuable allele is present in only one of the 100,000 plus accessions, we will miss it from a core, collection. Ultimately, we may have to screen the whole collection. With allelemining technologies rapidly becoming cheaper and faster, this will soon be within our grasp.

However, simply discovering the new alleles is not the end of the story. Each time we discover a new allele at a candidate gene, we then have to determine its agronomic significance. Here we go back to a new round of functional genomics research to assess the value of the new allele.

By discovering the full diversity of available alleles and their agronomic significance, we can finally look forward to genebanks achieving their full potential - contributing to sustainable development! by enabling us to deploy the right alleles in the right places at the right time.