Computational method identifies genes that regulate the cell´s machinery

An interdisciplinary team of researchers from the Hebrew University of Jerusalem, Stanford University and Tel Aviv University has developed the first computational method that can identify clusters of genes responsible for controlling processes within a cell, when those clusters become active, and, most importantly, how the clusters are regulated.

In a paper published online May 12 by Nature Genetics, the researchers report that their method revealed several previously unknown control, or regulatory, genes from Saccharomyces cerevisiae, better known as baker’s yeast.

“Revealing the control mechanism for gene clusters is crucial for understanding how cells respond to internal and external signals,” said team member David Botstein, a professor of genetics at Stanford. “Because this new method can predict the functions of regulator genes and their targets, we’re given a lot of insight into the roles of as yet unknown regulatory genes.”

Each cell in our body contains a copy of the genome, a long DNA molecule that stores the genetic information inherited from our parents. The DNA molecule encodes for a large number of genes, each of which is the recipe for building a specific protein. The cell’s machinery is formed from collections of different proteins.

Although all cells in our body contain identical DNA molecules, they dramatically differ in their protein makeup: muscle cells do not contain many of the proteins that are needed by liver cells, and vice-versa. Precise regulation of gene expression (that is, when the genes are activated or deactivated) is crucial to ensure that the right proteins are being made at the right time. Understanding the processes that are responsible for this gene regulation is a central question in genetics and molecular biology that has important implications for understanding how cells function and how diseases that involve breakdown in regulatory processes, such as cancer, can develop.

The common approach to uncovering regulatory processes is through many detailed and elaborate experiments that examine one piece of the puzzle at a time. Recent scientific and technological breakthroughs offer the hope of approaching this problem from a global perspective. The recently completed genome sequencing project enabled the development of microarray assays that measure the expression levels of thousands of genes in one experiment. These essentially give the scientists a snapshot of how much each gene is being used in the cells they examine. A typical experiment would measure the gene expression of thousands of genes in dozens of different experimental conditions to see which genes are expressed in each condition.

Microarray experiments provide a huge amount of valuable information. Scientists can see how different genes change their expression under different conditions and from that deduce clues about their role. Unfortunately, the underlying regulatory mechanisms that are responsible for these changes are far from being transparent in these data, and there is an ongoing debate as to the extent to which it is possible to discover actual regulators from expression data alone.

In the Nature Genetics article, the interuniversity research team provides the first large-scale method to address this question. Dr. Nir Friedman, the Harry and Abe Sherman senior lecturer in computer science at the Hebrew University, and his team are one of the leading groups working on reconstructing regulatory mechanisms from microarray expression measurements.

In collaboration with Stanford computer scientist Daphne Koller, they developed a new computer method for extracting regulatory circuits from large collections of gene expression measurements. Friedman and Koller applied this method to experimental data from Saccharomyces cerevisiae, one of the primary model organisms in genetics. The method identified modules of genes that are co-regulated and identifies the regulatory genes that tell each module of genes to turn on or off — in other words, to start or stop making proteins. The proteins from each module, in turn, are responsible for a different cell process. These processes include converting sugar to energy, responding to stress, folding proteins, and building cellular components such as the nucleus.

The Friedman and Koller team show that the method can not only recognize known regulatory modules, or clusters, but also predict previously unknown clusters and the genes that regulate those clusters. Moreover, the method can make predictions as to the exact conditions under which each regulatory gene is activated. In collaboration with the Botstein group at Stanford, several precise predictions about the role of proteins with unknown function were evaluated in biological experiments that examine the effect of “knocking out” a regulator under the conditions where it is predicted to be active. These experiments confirmed three of the predictions in the lab.

The authors of the Nature Genetics paper also include graduate students Eran Segal (Stanford), Dana Pe’er (the Hebrew University), and two post-doctoral fellows Michael Shapira (Stanford) and Aviv Regev (Tel Aviv University, currently a fellow at Harvard University’s Bauer Center for Genomics Research). The work of the research team was supported in part by a grant from the U.S. National Science Foundation.