Materassi Donatello, Roychowdhury Subhrajit, Hays Thomas, Salapaka Murti
Laboratory for Information and Decision Systems, Massachussets Institute of Technology, 77 Massachusetts Avenue Cambridge, MA 02139, USA.
BMC Biophys. 2013 Nov 16;6(1):14. doi: 10.1186/2046-1682-6-14.
Intracellular transport is crucial for many cellular processes where a large fraction of the cargo is transferred by motor-proteins over a network of microtubules. Malfunctions in the transport mechanism underlie a number of medical maladies.Existing methods for studying how motor-proteins coordinate the transfer of a shared cargo over a microtubule are either analytical or are based on Monte-Carlo simulations. Approaches that yield analytical results, while providing unique insights into transport mechanism, make simplifying assumptions, where a detailed characterization of important transport modalities is difficult to reach. On the other hand, Monte-Carlo based simulations can incorporate detailed characteristics of the transport mechanism; however, the quality of the results depend on the number and quality of simulation runs used in arriving at results. Here, for example, it is difficult to simulate and study rare-events that can trigger abnormalities in transport.
In this article, a semi-analytical methodology that determines the probability distribution function of motor-protein behavior in an exact manner is developed. The method utilizes a finite-dimensional projection of the underlying infinite-dimensional Markov model, which retains the Markov property, and enables the detailed and exact determination of motor configurations, from which meaningful inferences on transport characteristics of the original model can be derived.
Under this novel probabilistic approach new insights about the mechanisms of action of these proteins are found, suggesting hypothesis about their behavior and driving the design and realization of new experiments.The advantages provided in accuracy and efficiency make it possible to detect rare events in the motor protein dynamics, that could otherwise pass undetected using standard simulation methods. In this respect, the model has allowed to provide a possible explanation for possible mechanisms under which motor proteins could coordinate their motion.
细胞内运输对于许多细胞过程至关重要，其中很大一部分货物由驱动蛋白通过微管网络进行转运。运输机制的故障是许多医学疾病的基础。现有的研究驱动蛋白如何在微管上协调共享货物转运的方法要么是分析性的，要么基于蒙特卡罗模拟。产生分析结果的方法虽然能对运输机制提供独特见解，但做出了简化假设，难以详细表征重要的运输方式。另一方面，基于蒙特卡罗的模拟可以纳入运输机制的详细特征；然而，结果的质量取决于用于得出结果的模拟运行次数和质量。例如，在这里很难模拟和研究可能引发运输异常的罕见事件。
在本文中，开发了一种半分析方法，该方法能以精确方式确定驱动蛋白行为的概率分布函数。该方法利用了基础无限维马尔可夫模型的有限维投影，保留了马尔可夫性质，并能详细准确地确定驱动蛋白构型，从中可以得出关于原始模型运输特征的有意义推断。
在这种新颖的概率方法下，发现了关于这些蛋白质作用机制的新见解，提出了关于它们行为的假设，并推动了新实验的设计与实现。在准确性和效率方面提供的优势使得能够检测驱动蛋白动力学中的罕见事件，否则使用标准模拟方法可能无法检测到这些事件。在这方面，该模型为驱动蛋白协调其运动的可能机制提供了一种可能的解释。
