Predicting dominant mutations in genetic disorders associated with the misassembly of cytoskeletal proteins

微管蛋白是微管的组成部分，微管在有丝分裂和分子运输过程中具有多种功能，包括染色体排列。因此，微管蛋白突变与许多表型有关，如神经元疾病和对化疗的耐药性。该项目的第一部分包括在细胞水平上从空间和数量上定义不同微管蛋白同型的特性。每个同型可以使用荧光显微镜可视化，提供有关其特定定位以及在细胞周期的特定阶段的持续时间和强度的信息。除此之外，还将开发利用神经细胞作为系统进一步研究微管蛋白同型的技术。为了确定导致致病性的微管蛋白突变的机制，我将重点关注微管蛋白β III和β iv。然后，该项目将继续在体外重建和纯化单个微管蛋白异构体，在那里可以替换不同的异构体，从而分析微管功能的差异。为了实现这一目标，微管蛋白将在真核Sf9细胞中重组表达，并使用亲和纯化进行纯化。这种微管蛋白随后将被用于在体外对单个微管进行重构分析，使用TIRF显微镜来确定不同的同型是否会影响微管动力学。微管蛋白突变体也将同时表达和纯化。为了测试微管蛋白同型是否也调节微管运动的运动性和在微管上停留的时间，将使用实验室中可用的马达进行定量分析以测量分子运动运输，并进行生化分析以检测与其他非运动微管相关蛋白（MAPs）的相互作用。从这项工作中，我期望确定与微管蛋白β III和β IV突变相关的致病性的分子机制。下一阶段涉及使用生物信息学方法来计算分析先前实验工作中突出的相关同种异构体中的突变。广泛的人类变异映射到最近可用的微管结构将用于此，优先考虑以前未被此类方法预测的致病突变。患者基因组数据、多序列比对和分子建模技术等工具也可用于这一目的，通过这些方法指出的最具致病性的突变将被纳入实验工作。这包括直接比较含有优先致病突变的神经细胞与野生型，其中微管功能如前所述进行检查。将生物信息学方法强调的突变应用于这项实验工作将允许确定最致病的突变，因为它们也将在生物学模型中得到验证。最后，在整个项目中使用生物信息学和实验方法收集的数据可以集成到机器学习方法中。这将导致一种算法的设计，能够预测哪些微管蛋白突变最有可能导致神经紊乱或赋予对紫杉醇的抗性。

Tubulin is the building block of microtubules, which carry out a diverse set of functions including chromosomal alignment during mitosis and molecular trafficking. Therefore tubulin mutations are linked to a number of phenotypes such as neuronal disorders and resistance to chemotherapy. The first part of the project involves defining the properties of different tubulin isotypes at the cellular level spatially and quantitatively. Each isotype can be visualised using fluorescence microscopy, giving information about its particular localisation as well as the duration and intensity during specific phases of the cell cycle. Alongside this, techniques will be developed to use neuronal cells as a system to study tubulin isotypes further. To define the mechanism of tubulin mutations that results in pathogenicity, I will focus on tubulin beta III and beta IV.The project will then move on to reconstituting and purifying single tubulin isoforms in vitro, where different isoforms can be substituted allowing differences in microtubule function to be analysed. To achieve this, tubulin will be recombinantly expressed in eukaryotic Sf9 cells and purified using affinity purification. This tubulin will then be used in reconstitution assays to be performed in vitro on individual microtubules using TIRF microscopy to define whether different isotypes influence microtubule dynamics. Tubulin mutants will also be expressed and purified alongside. To test whether tubulin isotype also regulates microtubule motor motility and residency time on microtubule, quantitative analysis will be performed to measure molecular motor transport using available motors in the lab and biochemical assays to detect interactions with other non-motor microtubule-associated proteins (MAPs). From this work, I anticipate to define the molecular mechanism underlying the pathogenicity associated with tubulin beta III and beta IV mutations.The next phase involves the use of bioinformatics approaches to computationally analyse mutations in relevant isoforms highlighted by the previous experimental work. A wide array of human variants mapped onto the recently available microtubule structures will be used for this, prioritising pathogenic mutations that have not been predicted by such methods previously. Tools such as patient genome data, multiple sequence alignments and molecular modelling techniques can also be incorporated for this purpose, where mutations indicated to be the most pathogenic by these methods will be taken into experimental work. This comprises directly comparing neuronal cells containing the prioritised pathogenic mutations to wild type where microtubule function is examined as previously described. Applying the mutations highlighted by the bioinformatics approaches into this experimental work will allow the most pathogenic mutations to be determined as they would also be validated within a biological model. Finally, data gathered throughout the project using both bioinformatics and experimental approaches can be integrated into machine learning methods. This will lead to the design of an algorithm that is able to predict which tubulin mutations are most likely to cause neuronal disorders or confer resistance to taxol.