BioWorld International Correspondent

LONDON - Those involved in drug discovery now have useful new tools at their disposal, including libraries of retroviral vectors, which instruct mammalian cells to switch off specific genes.

The scientists who developed the libraries predict that they will speed up the identification of drug targets. While validating it, they have pinpointed one known and five new genes involved in a cellular pathway that is commonly mutated in cancer.

Rene Bernards, professor of molecular carcinogenesis at the Netherlands Cancer Institute in Amsterdam, told BioWorld International: "We believe that these libraries will have a major impact on drug discovery. Researchers can now choose to inactivate many genes, selectively, one by one in a high-throughput fashion, and ask whether a biological process that they find interesting is being disrupted."

Bernards and his colleagues constructed a set of almost 24,000 retroviral vectors designed to target and suppress almost 8,000 human genes - three vectors for each gene. They describe the development of the library in the March 25, 2004, issue of Nature in a paper titled "A large-scale RNAi screen in human cells identifies new components of the p53 pathway."

In the same issue of Nature, Gregory Hannon and colleagues at Cold Spring Harbor Laboratory in New York, report their creation of two similar libraries, each targeting almost 10,000 human genes and around 5,000 mouse genes. Their paper is titled "A resource for large-scale RNA interference-based screens in mammals."

Bernards' group already has sold copies of its library to several pharmaceutical companies. The library also is available to researchers in academia.

The vectors made by both groups rely on the technique known as RNA interference, or RNAi. Scientists first have to establish the sequence of messenger RNA made by the gene they want to silence and then manufacture double-stranded RNA with the same sequence. In invertebrate cells, adding the double-stranded RNA causes the messenger RNA with the same sequence to be broken down. The gene cannot make its protein and effectively is silenced.

Although that strategy works in Drosophila and Caenorhabditis elegans, mammalian cells do not respond the same. When mammalian cells detect long double-stranded RNA, an antiviral response comes into play and shuts down all protein synthesis.

Shorter stretches of RNA, which are 21 nucleotides long and called small interfering RNAs (siRNAs), can work, but only for short periods. Bernards' group solved the problem two years ago, when they reported a method of making the cells manufacture the siRNAs themselves.

That was done by infecting the cells with retroviruses that had been engineered to include additional genetic material that would direct the synthesis of siRNAs. The extra genetic material produces RNA fragments that fold back on themselves to form short pieces of double-stranded RNA and a loop, and are therefore called short hairpin RNAs. The loop is removed by cellular enzymes, leaving an siRNA molecule. That method, Bernards and his colleagues reported in 2002, can stably suppress gene expression.

After that, producing the library was straightforward, but "a lot of work," Bernards said. The team chose to focus on genes involved in the cell cycle, the regulation of transcription, stress signaling, signal transduction and important biological processes, as well as genes implicated in cancer and other diseases. They made three vectors to target each gene, expecting that at least one would silence the gene in question.

They then tried to use the library to identify genes that play a role in the p53-signaling pathway, which is mutated in more than half of all human cancers.

"In basically one experiment," Bernard said, "we found five new genes that are involved in this pathway, plus we identified p53 itself. When we inactivated these genes, the cells became unresponsive to antiproliferative p53 signaling.

"This is a new tool for answering interesting biological questions and for discovering interesting new drug targets, and in [the] future we plan to use it for more of the same," Bernards said. "We already have a bundle of genes in other cancer-relevant pathways that we plan to look at over the next few months."