Actual promoter activities (solid lines), and activities predicted (dotted line) based on the profile at one promoter.

Alon/AAAS

The trouble with systems biology is that no one wants to make all those measurements. Rate constants and binding assays are just not that exciting. Now, Michal Ronen, Uri Alon (Weizmann Institute of Science, Rehovot, Israel), and colleagues have come up with a method for rapidly determining the expression profiles of thousands of genes, and extrapolating from those profiles to derive protein concentrations and predict responses to other conditions.

The test case for this system is simple enough. The researchers linked GFP to eight operons in the Escherichia coli SOS DNA repair system. This system is controlled by a single repressor LexA. For each operon, Alon determined β, the unrepressed production rate, and k, the effective affinity of the LexA repressor based on half-maximal repression. This work is speeded along by the use of multiwell plates and automated fluorescence measurements, so converting an idea for an experiment to computerized data can take just two days. “ It's like being in a candy shop,” says Alon.Once the kinetics is determined under one condition, this yields what Alon calls “the hidden variable—the profile of the transcription factor in its active form.” When a new condition is imposed, the kinetics of only one operon need be tested. The kinetics of all the others falls out from this result and the β and k values determined earlier.

“Most previous models tried to fit the effect of one transcription event on another transcription event, and they don't go through the hidden variable of the transcription factor activity,” says Alon. “I think that is the critical step.” The next improvement will be to accommodate multiple varying inputs at a single promoter. Once this is achieved, Alon believes he can derive a cell-wide model of connections and connection strengths, and thus understand the underlying logic of the cell. ▪

Reference:

Ronen, M., et al. 2002. Proc. Natl. Acad. Sci. USA. 10.1073/pnas.152046799.