If the Holy Grail of systems biology — or at least one of the grails — is to computationally simulate the behavior of cells, then scientists are going to have to start their crusade by assembling the little clues. In the March 25 issue of Science, researchers took another measured step forward in the attempt to begin understanding how individual cells behave by reporting on their experiments with gene regulation networks.
Michael Elowitz, an assistant professor in the department of biology and applied physics at CalTech, went after a mechanism for gene regulation inside cells that involves a repressor gene preventing the expression of another gene. Being able to describe the relationship between the repressor and target genes mathematically, he reasoned, would potentially provide a way to describe similar relationships in other cell systems, and help advance researchers’ ability to model living cells.
Doing this in the past was problematic, Elowitz points out, partly because no one could measure the parameters associated with gene regulation — after all, it’s hard to take measurements on a single cell level. But Elowitz, along with his collaborators at the Weizmann Institute of Science in Israel and McGill University in Montreal, set up a system for analyzing fluorescent time-lapse microscopy data that allowed the researchers to simultaneously collect data from multiple instruments. After watching the movies of labeled proteins moving about in colonies of cells — an approach Elowitz refers to as “popcorn biochemistry” — the group then used this data to try to understand the variability in the relationship between a repressor gene and its target.
Elowitz and his colleagues found that they could separate the variability, or noise, inherent to the gene regulatory system under investigation into two components: one governed by the rapid fluctuations of a system with a relatively small number of particles, and one governed by the slower effect of other particles in the cell changing over time. Elowitz says the results should contribute to an understanding of the mechanics of gene regulation, and guide other researchers as they attempt to model these types of systems.
“You want to be able to … model the module in the cell you’re interested in, but the problem with that is that other aspects of the cell are always intruding — your little module is not perfectly isolated from the rest of the cell,” Elowitz says. “A very fundamental question that we’re trying to address here is the coupling between one modular piece of genetic circuitry and everything else going on in the cell. What we’re showing here is that even though we can’t model everything else going on in the cell, we can describe how [those other mechanisms] affect our module statistically speaking — and we can do that very accurately, actually.”
— John S. MacNeil
Michael Elowitz homepage
Abstract in Science