研究驅動蛋白Kif11、Kif18A、Kif23 在神經型態發生中扮演之功能

Abstract

作為神經中最主要的細胞骨架，微管在神經型態發生(neuronal morphogenesis)的過程中扮演了一個重要的角色。透過微管動力蛋白(microtubule-based motors)系統，細胞內的物質如蛋白質、胞器、RNA等能夠被運送至細胞各處。而在神經細胞中，為了有效的將物質運送至長距離的樹突與軸突，胞內運輸系統存在的重要性可見一斑。為了研究微管動力蛋白在神經型態發生中的功能，我們開發了一種高含量顯微影像分析技術。此技術是使用P19細胞株作為模型，在P19分化為神經的過程中利用pLenti-shRNA系統性的抑制個別的微管動力蛋白，再透過顯微影像的自動擷取與分析來觀察神經細胞的型態。透過這項技術，我們觀察到Kif11、Kif18A和Kif23個別被抑制後的神經細胞出現了相似的型態：多核變大的細胞本體(soma)以及長度變短的神經突(neurite)。另一方面，利用機器學習的方法(machine learning-based method)來對神經影像作影像特徵的量化及分群也顯示這三個驅動蛋白被抑制後的型態極度相像。因此，我們認為這三個驅動蛋白在神經型態發生中具有類似的功能或參與相同的路徑。在本研究中，我們利用RT-PCR的方式測定出pLenti-shRNA的抑制效率。同時也使用了Kif11的專一性藥物來去除其活性，並得到了與抑制Kif11相似的結果。最後，為了研究在神經型態發生的過程中，Kif11和Kif18A是否具有相同的功能，我們在抑制Kif11的細胞中大量表現Kif18A。然而表現Kif18A並無法恢復(rescue)抑制Kif11所造成的特殊型態。

The microtubule is the major cytoskeleton in neurons and it plays crucial roles during neuronal morphogenesis. Intracellular transport by microtubule-based motors is fundamental for distributing various proteins, organelles, and RNA transcripts. This intracellular transport system is particularly important for neurons in which cellular components need to be transported over long distance into axons and dendrites. To study the function of microtubule-based motors on neuronal morphogenesis, we developed a high-content suppression assay to individually knockdown most microtubule-based motors in the mouse genome. We used mouse P19 embryonal carcinoma differentiated neurons as our model system. High-content screen and analysis were performed to automatically quantify the phenotypic effect of suppressing each individual motor. From this assay, we observed interesting phenotype in Kif11, Kif18A, or Kif23-depleted neurons with multiple nuclei seemingly stuck in one soma and drastically reduced neurite length. Further feature quantification and classification using machine learning-based method demonstrated that the phenotypic effects of depleting Kif11, Kif18A or Kif23 are extremely similar. Therefore, we hypothesized that these motors have similar function or participate in the same pathway during neuronal morphogenesis. RT-PCR was used to verify the knockdown efficiency. In addition, monastrol was used to inactivate Kif11 during neuronal differentiation, and Kif11-inactivated neurons phenocopy shRNA-mediated Kif11-depleted neurons. Finally, we studied the functional overlap between Kif11 and Kif18A by suppressing Kif11 and overexpressing Kif18A in the same cell. However, the results demonstrated that the phenotypes of depleting Kif11 cannot be rescued by overexpressing Kif18A.