Zhu X-M, Yin L, Hood L, Ao P
GenMath, Corp. 5525 27th Ave.N.E., Seattle, WA 98105, USA.
J Bioinform Comput Biol. 2004 Dec;2(4):785-817. doi: 10.1142/s0219720004000946.
Based on the dynamical structure theory for complex networks recently developed by one of us and on the physical-chemical models for gene regulation, developed by Shea and Ackers in the 1980's, we formulate a direct and concise mathematical framework for the genetic switch controlling phage lambda life cycles, which naturally includes the stochastic effect. The dynamical structure theory states that the dynamics of a complex network is determined by its four elementary components: The dissipation (analogous to degradation), the stochastic force, the driving force determined by a potential, and the transverse force. The potential may be interpreted as a landscape for the phage development in terms of attractive basins, saddle points, peaks and valleys. The dissipation gives rise to the adaptivity of the phage in the landscape defined by the potential: The phage always has the tendency to approach the bottom of the nearby attractive basin. The transverse force tends to keep the network on the equal-potential contour of the landscape. The stochastic fluctuation gives the phage the ability to search around the potential landscape by passing through saddle points. With molecular parameters in our model fixed primarily by the experimental data on wild-type phage and supplemented by data on one mutant, our calculated results on mutants agree quantitatively with the available experimental observations on other mutants for protein number, lysogenization frequency, and a lysis frequency in lysogen culture. The calculation reproduces the observed robustness of the phage lambda genetic switch. This is the first mathematical description that successfully represents such a wide variety of major experimental phenomena. Specifically, we find: (1) The explanation for both the stability and the efficiency of phage lambda genetic switch is the exponential dependence of saddle point crossing rate on potential barrier height, a result of the stochastic motion in a landscape; and (2) The positive feedback of cI repressor gene transcription, enhanced by the CI dimer cooperative binding, is the key to the robustness of the phage lambda genetic switch against mutations and fluctuations in kinetic parameter values.
基于我们其中一人最近发展的复杂网络动力学结构理论,以及谢伊和阿克斯在20世纪80年代提出的基因调控物理化学模型,我们为控制噬菌体λ生命周期的遗传开关构建了一个直接且简洁的数学框架,该框架自然地包含了随机效应。动力学结构理论指出,复杂网络的动力学由其四个基本要素决定:耗散(类似于降解)、随机力、由势决定的驱动力以及横向力。从吸引盆、鞍点、峰值和谷值的角度来看,势可被解释为噬菌体发育的景观。耗散导致噬菌体在由势定义的景观中具有适应性:噬菌体总是倾向于接近附近吸引盆的底部。横向力倾向于使网络保持在景观的等势轮廓上。随机涨落使噬菌体能够通过鞍点在势景观周围进行搜索。在我们的模型中,分子参数主要由野生型噬菌体的实验数据确定,并辅以一个突变体的数据,我们对突变体计算得到的结果在蛋白质数量、溶原化频率和溶原培养中的裂解频率方面与其他突变体的现有实验观察结果在数量上相符。该计算重现了观察到的噬菌体λ遗传开关的稳健性。这是首次成功表示如此广泛多样的主要实验现象的数学描述。具体而言,我们发现:(1)噬菌体λ遗传开关稳定性和效率的解释在于鞍点穿越率对势垒高度的指数依赖性,这是景观中随机运动的结果;(2)由CI二聚体协同结合增强的cI阻遏蛋白基因转录的正反馈,是噬菌体λ遗传开关对突变和动力学参数值波动具有稳健性的关键。