Protein translocation analysis by automated image processing
Protein translocation analysis by automated image processing
复制标题
通过自动图像处理进行蛋白质易位分析
DOI：
--
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
T. Berlage
中科院分区：
文献类型：
--
作者：
O. Domanova;S. Borbe;Stefanie Mühlfeld;R. Kubitz;D. Häussinger;T. Berlage
许多动态的细胞过程与蛋白质在细胞内和细胞间的转移有关。有几种蛋白质负责跨肝细胞膜的主动运输。例如，胆盐通过嵌入细胞膜的转运蛋白从肝细胞输出到邻近的小管中。在高渗条件下，它们被回收到细胞内的小泡中，运输受到下调。这些转运蛋白的位置可以通过荧光显微镜来探索。对肝转运蛋白的拓扑学分析可能揭示临床相关的功能信息，并可作为其他位置调节过程的范例。一种自动化的方法被开发出来，它可以量化转运蛋白在细胞中的定位。通过自动图像处理，分析激光扫描共聚焦显微镜图像以寻找合适的膜段。获得并量化了蛋白质在膜上的分布情况。比较正常和高渗条件下孵育的大鼠肝组织标本。制定了信息提取和统计评估的全自动化工作流程，并产生了强有力的成果。速度较慢的人工计算可以被快速、完全自动化的方法所取代。此外，新方法只使用了客观和可重复性的标准。在不同的数据集上都表现出了稳健性和稳定的性能。
Many dynamic cellular processes are associated with intraand intercellular translocation of proteins. Several proteins are responsible for the active transport across the membrane of liver cells. For example, bile salts are exported from hepatocytes into adjacent canaliculi by transporter proteins embedded in the membrane. Under hyperosmotic conditions they are retrieved into intracellular vesicles and transport is down regulated. The location of these transporters can be explored by fluorescence microscopy. A toponomic analysis of the hepatic transporters may reveal functional information of clinical relevance and serve as an example of other location-regulated processes. An automated method was developed which quantifies the localisation of transporter proteins in a cell. Laser scanning confocal microscopy images were analysed for suitable membrane segments by automatic image processing. Protein distribution profiles across the membrane were obtained and quantified. Rat liver tissue samples incubated under normal and hyperosmotic conditions were compared. The fully automated workflow for the information extraction and statistical evaluation has been elaborated and produces robust results. Slow manual calculation can be substituted by the faster and completely automated method. Furthermore, the new method only uses objective and reproducible criteria. Robustness and stable performance have been shown on various data sets.