Michael Hemann seeks better ways to deploy chemotherapy drugs and overcome tumor resistance.
Researchers have found they can shut down individual genes by manipulating RNA, potentially opening the way for medicines to treat disorders from high cholesterol to cancer, as well as viral diseases such as AIDS. Now, scientists in the Department of Biology and the Whitehead Institute for Biomedical Research have taken a step toward identifying which human genes can be altered this way.
RNA-mediated cellular processes include a method called RNA interference (RNAi). RNAi has been used to control viral infections in plants, as a tool to probe cell biology by observing the effects of silencing specific genes, and as a method of seeing if potential drugs are having their desired effect.
MicroRNAs are small RNA molecules that also seem to silence genes by interfering with the chemical translation of DNA into protein. Some microRNAs affect genes that serve as master regulators--genes that play a pivotal role in the regulation of other genes. The question is, which mammalian genes code for microRNAs?
In 2003, Associate Professor David Bartel and Assistant Professor Chris Burge (who are also Whitehead researchers) developed a computational method that detects the microRNA genes in different animals. Using this method, they estimated that microRNAs constitute nearly 1 percent of genes in the human genome, making microRNA genes one of the more abundant types of regulatory molecules.
"MicroRNAs are one of the many types of regulatory molecules important in determining which genes are on or off in a particular cell," said Bartel. "Understanding what they do may provide the answers to some unsolved mysteries of gene regulation and help us better understand human biology and disease."
Bartel and Burge then set out to apply a similar approach to defining the relationship between microRNAs and the genes they target. Last month in the journal Cell, their labs reported that they have created a new computational method, called TargetScan, which does just that.
For each microRNA, TargetScan searches a database of messenger RNAs (mRNAs)--chemical messages that transcribe DNA into protein--for regions that pair to portions of the microRNA, and assigns a score to the overall degree of pairing that could occur between the microRNA and each mRNA. Those mRNAs that have high scores conserved in three or more organisms are predicted as targets of the microRNA.
Using this method, the team identified more than 400 genes in the human, mouse and rat genomes likely to be regulated by microRNAs. In addition, TargetScan predicted an additional 100 microRNA targets that are conserved in humans, mice, rats and the pufferfish.
According to Burge, 70 percent of targets predicted by TargetScan are likely to be authentic microRNA targets and the experimental data in the paper supports that a majority of their predictions are correct.
"A detailed understanding of this mechanism will aid in the engineering of new small RNAs that regulate particular target genes while avoiding undesirable side effects," he said. "MicroRNAs or related molecules could potentially be used to therapeutically manipulate gene expression in cases where malfunctioning genes contribute to disease."
A next step is additional work to improve the team's computational methods. These improvements will let the researchers generate more accurate and comprehensive predictions for which genes are regulated by microRNAs, and also will shed more light on the biological roles of the microRNA-mediated gene regulation.
A version of this article appeared in MIT Tech Talk on February 4, 2004.