Specificity and Robustness in Transcription Control Networks

Recognition by transcription factors of the regulatory DNA sub-sequences upstream of genes is the fundamental step in controlling gene expression. It is susceptible to mutations of both the sequence and the DNA binding domain of the protein. We define "robustness" of a transcription control networkin terms of the mutation load- the probability of functional failure due to mutation - and analyze the dependence of the mutation load on factor binding specificity and its degree of pleiotropy (i.e., the number of its regulatory targets). The study is based on a novel and realistic model of factor/DNA recognition in the context of relatively simple prokaryotic transcription control architecture. The mutation load computed in this model has a minimum as a function of binding specificity, with the optimal specificity of the factor scaling approximately linearly with its degree of pleiotropy. This prediction is supported by the correlation between the relative specificity of factors and their pleiotropy found in genomic data for the known transcription factors in E.coli. Robustness analysis can be extended to more complex, combinatorial and redundant networks and provides a novel perspective on transcription regulation.