Computational Analysis of Microtubule Dynamics for Personalized Cancer Therapy

描述(由申请人提供)：优化的光学和反馈控制显微镜硬件允许高效获取大的、高质量的图像数据集。基于计算机的分析提供了对大量图像数据的快速处理，这是手动无法完成的。然而，计算机视觉系统可以为转化型癌症研究做出的最令人兴奋的贡献是，它可以访问基于图像的信息，而这些信息是肉眼无法访问的。计算机视觉程序可以直接耦合到描述隐藏的、不可见的过程和可测量的图像事件之间关系的数学模型。因此，隐藏进程的行为的变化可以作为图像中的变化来检测。这项研究设想使用这样的算法来获得关于FDA批准的三种紫杉烷对前列腺癌(PC)细胞系中微管细胞骨架的不同影响的统计代表性结果。我之前在基础研究中的工作证明了基于计算机的微管(MT)细胞骨架分析的能力，可以区分肾癌的弱疾病表型，并建立与微管动力学的联系。因此，拟议的翻译研究项目可以通过首次为医生配备计算机辅助工具来影响临床决策，该工具允许为转移性PC患者设计有效的个性化MT靶向化疗。转移性前列腺癌的主要治疗方法是用FDA批准的三种紫杉烷(紫杉醇、多西紫杉醇和卡马西紫杉醇)中的一种进行紫杉烷类化疗。然而，目前还没有办法根据个体患者MT细胞骨架的特定动态行为模式来选择紫杉烷进行化疗。此外，最近的数据表明，AR与MTS结合是为了运输到细胞核，并且在转移性PC患者中有几种临床相关的AR剪接变体。到目前为止，还没有关于野生动物潜在影响的信息。 MT动力学上的AR类型或变异体，因此没有关于作为细胞AR含量的函数的紫杉烷治疗的差异转移PC细胞反应的信息。在初步研究的基础上，我们假设个体PC患者的肿瘤MT动力学存在内在差异，AR变异体的存在影响MT聚合动力学的特定参数。如果是正确的，这一假设对PC的治疗有非常重要的意义。由于不同的微管靶向药物(甚至来自同一类药物，如紫杉烷)影响MT动力学的不同参数，我们可以想象到，我们可以将每种药物与单个肿瘤特异性的MT动力学信号相匹配，以获得最大的治疗效果。

DESCRIPTION (provided by applicant): Optimized optics and feedback-controlled microscope hardware permit efficient acquisition of large, high- quality image datasets. Computer-based analyses deliver fast processing of high volume of image data which manually cannot be accomplished. The most exciting contribution that computer vision systems can make to translational cancer research, however, is to give access to image-based information that is inaccessible by eye. Computer vision programs can be directly coupled to mathematical models that describe the relation between hidden, invisible processes and measurable image events. Changes in the behavior of hidden processes are thus detectable as changes in the image. This study envisages the use of such algorithms to obtain statistically representative results for the differential effects of each of the three FDA-approved taxanes on the microtubule cytoskeleton in prostate cancer (PC) cell lines. My previous work in basic research has demonstrated the ability of computer-based analysis of the microtubule (MT) cytoskeleton to distinguish between weak disease phenotypes and establish links to MT dynamics in renal cell carcinoma. Therefore, the proposed translational research project can impact clinical decision-making by equipping physicians for the first time with a computer-aided tool allowing the design of an effective personalized MT-targeting chemotherapy of metastatic PC patients. Metastatic PC is treated primarily by means of taxane-based chemotherapy with one of the three FDA- approved taxanes (paclitaxel, docetaxel and cabazitaxel). However, currently there is no way of selecting the taxane for chemotherapy based on the particular pattern of dynamic behavior of the MT cytoskeleton in individual patients. In addition, recent data have indicated that AR binds MTs in order to traffic to the nucleus and that there are several clinically relevant AR splice variants i metastatic PC patients. To date, there is no information available on the potential effects of wild type or variant AR on MT dynamics and consequently no information on differential metastatic PC cell response to taxane treatment as a function of cellular AR content. Based on preliminary research, we hypothesize that there are inherent differences in tumor MT dynamics among individual PC patients, and that the presence of AR variants affects specific parameters of MT polymerization dynamics. If correct, this hypothesis has very significant implications for PC treatment. Because different microtubule-targeting drugs (even from within the same class like the taxanes) affect distinct parameters of MT dynamics, it is conceivable that we can match each drug with an individual tumor-specific "MT-dynamics signature" for maximum therapeutic efficacy.