RNA Molecules: Noise model overturns accepted notions
Modeling the effects of noise on a system helped a Virginia Tech team of biologists and engineers discover an inconsistency in two commonly accepted notions about single cells: how many messenger RNA (mRNA) molecules are in a cell, and how long they live. Their results affect the understanding of how cells process information and were published this spring in the Proceedings of the National Academy of Sciences (PNAS).
In a cell, information is processed through a molecular network of genes and proteins. Messenger RNA is the molecule that carries information from the gene to the cell’s ribosomes, where proteins are made. Because of their small size, cells are sensitive to random fluctuations in the number of molecules being created or destroyed at any given moment. Yet, despite the noise in the system from variations in mRNA molecules, cells typically have reliable communications for essential processes such as DNA replication and cell division.
Because of the randomness of the mRNA numbers, the team investigated the system through a stochastic modeling perspective. For yeast cell-cycle genes, the literature reported, on average, only one mRNA molecule per gene per cell and that each mRNA molecule lives, on average, for 15-20 minutes before it degrades.
ECE’s Bill Baumann worked on the model with John Tyson, University Distinguished Professor of biology, Mark Paul of mechanical engineering, and Sandip Kar, a postdoctoral associate. “When we looked at the effects of noise on our calculations, we found a tradeoff,” Baumann said. “If the mRNA molecules are short-lived, then a small number of them are sufficient for the cell to function. If they are long-lived, however, small numbers are a problem: they create too much noise.” The team concluded that since experimental data revealed long lifetimes, there had to be higher numbers of mRNA molecules than conventionally accepted.
“We came to our conclusion based on our model of a yeast cell, and recent experimental work agrees that the numbers of molecules had been underestimated in the past,” Baumann said.
Electrical engineers typically model the effects of noise for radio receivers and other communications applications.
“This is a different application of the same theory, except that as a biological system, it’s highly nonlinear,” he said.
Tyson, Baumann, and Paul are investigators on a National Institutes of Health (NIH) funded research project to build more elaborate and more accurate models of noise in the control system of yeast cells.