Kaneko K, Furusawa C
University of Tokyo, Department of Basic Science, Tokyo, Japan.
IET Syst Biol. 2008 Sep;2(5):234-46. doi: 10.1049/iet-syb:20070078.
Biological processes are inherently noisy, as highlighted in recent measurements of stochasticity in gene expression. Here, the authors show that such phenotypic noise is essential to the adaptation of organisms to a variety of environments and also to the evolution of robustness against mutations. First, the authors show that for any growing cell showing stochastic gene expression, the adaptive cellular state is inevitably selected by noise, without the use of a specific signal transduction network. In general, changes in any protein concentration in a cell are products of its synthesis minus dilution and degradation, both of which are proportional to the rate of cell growth. In an adaptive state, both the synthesis and dilution terms of proteins are large, and so the adaptive state is less affected by stochasticity in gene expression, whereas for a non-adaptive state, both terms are smaller, and so cells are easily knocked out of their original state by noise. This leads to a novel, generic mechanism for the selection of adaptive states. The authors have confirmed this selection by model simulations. Secondly, the authors consider the evolution of gene networks to acquire robustness of the phenotype against noise and mutation. Through simulations using a simple stochastic gene expression network that undergoes mutation and selection, the authors show that a threshold level of noise in gene expression is required for the network to acquire both types of robustness. The results reveal how the noise that cells encounter during growth and development shapes any network's robustness, not only to noise but also to mutations. The authors also establish a relationship between developmental and mutational robustness.
生物过程本质上是有噪声的，正如最近对基因表达随机性的测量所强调的那样。在此，作者表明这种表型噪声对于生物体适应各种环境以及对突变的稳健性进化至关重要。首先，作者表明对于任何表现出随机基因表达的生长细胞，适应性细胞状态不可避免地由噪声选择，而无需使用特定的信号转导网络。一般来说，细胞中任何蛋白质浓度的变化都是其合成减去稀释和降解的产物，这两者都与细胞生长速率成正比。在适应状态下，蛋白质的合成和稀释项都很大，因此适应状态受基因表达随机性的影响较小，而对于非适应状态，这两个项都较小，因此细胞很容易因噪声而脱离其原始状态。这导致了一种选择适应状态的新颖通用机制。作者已通过模型模拟证实了这种选择。其次，作者考虑基因网络的进化，以获得表型对噪声和突变的稳健性。通过使用经历突变和选择的简单随机基因表达网络进行模拟，作者表明基因表达中的噪声阈值水平是网络获得这两种稳健性所必需的。结果揭示了细胞在生长和发育过程中遇到的噪声如何塑造任何网络的稳健性，不仅是对噪声的稳健性，还有对突变的稳健性。作者还建立了发育稳健性和突变稳健性之间的关系。
