At #10 is a paper from a team led by Ronald Davis of Stanford University, which uses a knockout approach to investigate the function of individual yeast genes. And at #8 the worm gets a similar treatment, albeit with a very different technique, from a group led by Julie Ahringer at the Wellcome Trust/Cancer Research U.K. Institute and University of Cambridge.

The idea behind gene knockout is simplicity itself: destroy a gene and see what impact that has on the development of the organism. In the past, knockouts were often created at random, adding the additional task of working out which gene had in fact been destroyed. But starting from the sequence, Davis and his group systematically destroyed 5,916 yeast open-reading frames (ORFs, long stretches of DNA that could code for a protein). They replaced the entire gene with an antibiotic-resistance marker that was itself flanked with a unique DNA sequence that could be used as a "molecular bar code" to identify each individual deletion mutant. That enabled the team to grow a mixture of every deletion mutant under various defined conditions—for example, galactose or high salt—and track the performance of the various mutants. The more important a gene is for growth, the more rapidly it declines in the mixture.

The galactose pathway has been extensively studied in yeast, and yet this approach identified 10 genes that had not previously been associated with the use of galactose. Likewise, although the pathways that enable yeast to grow in conditions of high salt were well known, Davis’s team found several new genes that were also necessary under these conditions. These are just two of the fascinating results to emerge from the study, which systematically maps individual genes onto known environmental conditions.

The worm team adopted a similar theoretical approach, but rather than deleting individual genes they used RNA-mediated interference, RNAi. This relies on a cellular defense mechanism triggered by the double-stranded RNA that is characteristic of invasive genetic elements such as certain viruses. The cell defends itself by destroying single-stranded RNA of the same sequence, thus preventing the rogue elements from reproducing themselves and inserting themselves into the cell’s genome. Double-stranded RNA can thus knock out a cellular gene of the same sequence, by calling for the destruction of the mRNA produced by the target sequence. As luck would have it, merely feeding bacteria that contain double-stranded RNA to worms knocks out the corresponding worm gene.

The Cambridge group constructed a library of 16,757 bacterial strains, each expressing double-stranded RNA equivalent to one of the worm’s 19,427 genes. Wild-type worms suffered some sort of disruption from 1,722 of the target genes, roughly 10% of the genes studied. Of these, 1,528 could be assigned to a definite RNAi phenotype, and roughly two-thirds of these had not previously been associated with any biological function in the living worm.

Many of the genes resulted in more than one kind of RNAi phenotype, indicating that the gene in question had many developmental roles. Almost half of the genes were absolutely necessary for the survival of the worm embryo. Of the ones that permitted growth and development, the largest group resulted in some sort of uncoordinated movement. Most fascinating, the group identified 33 genes that were close homologues of human disease genes. Of these, half gave rise to a viable post-embryonic phenotype, while in the whole sample of 1,722 genes only 16% resulted in a viable adult worm. This suggests that the worm may be useful as model of certain human genetic diseases.