Background During development, endothelial cells acquire mesenchymal-like properties to migrate and facilitate normal vascular formation. This process of transformation is known as endothelial to mesenchymal transition (EndMT) and has also been implicated in diseases like vascular pathologies contributing to endothelial inflammation, atherosclerosis and tumour angiogenesis. The Angiomotin family of scaffold proteins play a role in transducing mechanical force at cell junctions. Of this family, Angiomotin-Like 2 (AmotL2) localises to endothelial cell junctions and was recently found to play a role in regulating endothelial cell mechanosensing and inflammation. Methods/Materials Primary human endothelial cell lines (HUVEC) were cultured and manipulated in vitro to investigate the role of AmotL2 in EndMT. Lentiviral short hairpin RNA interference was employed in AmotL2-loss-of-function studies, (produced using HEK - Human Embryonic Kidney - cells) to generate knockdown(kd) cells. Western blotting (WB) was used to assess AmotL2 depletion and changes in protein expression of key EndMT markers. qPCR was performed to look at the same at a transcriptional level. Immunofluorescent staining and confocal imaging were performed to validate WB and qPCR results as well as to study protein localisation. Results AmotL2 was found to regulate Snail1 and N-cadherin at both protein and mRNA levels. Morphological findings displayed the AmotL2kd cells to be elongated, deviating from the regular cobblestone morphology observed in control cells. An increase in scaffold protein levels was observed in the AmotL2 kd samples. Similar results were seen in qPCR data where increased mRNA expression was observed in the AmotL2 kd samples for the same targets. On analysis of IF image data, more nuclear staining was observed in the kd samples. qPCR analysis done on samples treated with TGF-β, exhibited an increase in mRNA expression of targets involved in the EndMT pathway in the treatment samples against the controls. Conclusion The results suggest that AmotL2 plays a role in EndMT by affecting the transcription factors and proteins involved in the pathway, which leads to changing morphology and behaviour of the cells. Looking into more targets involved in EndMT may give us a better understanding of how this process leads to diseases like atherosclerosis and tumour angiogenesis.

本論文提供了一個句子重組試題編製的環境，協助教師編製句子重組試題，同時學生也能夠在此編製環境中練習句子重組試題。 句子重組試題即是要求學生把試題給的一組詞彙組合成特定詞序的正確語句之題型，該試題類型可以檢驗學生對於句型和文法的知識。然而試題所給的詞彙集合往往除了可以組合成教師想要學生回答的正確語句之外，也可以組合成其它的合法語句。為了能辨識學生的回答，把這些合法語句以人工方式逐一建置為答案卻對出題教師造成了負擔。 我們建構了一個電腦輔助句子重組試題編製的環境來減輕出題教師的負擔。為了讓電腦可以恰當地判斷學生的回答，我們的編製環境限制了試題詞彙集的相對位置，藉此約束學生只能排出教師預設的特定答案。同時在出題教師建置試題答案時，我們的編製環境也試圖提供所有可能的合法詞序之語句，供出題教師參考。但本論文的研究經驗顯示要自動協助出題教師預示所有可能的合法詞序之語句卻是一件艱難的工作，而且這一研究問題與語法學有密切關係。 本論文以基礎詞組為主軸，透過合併詞組和史丹佛剖析器的操作建構出英文句子重組試題編製環境，供教師編輯與學生練習。同時，我們在論文中也提報了中文句子重組試題編製環境的初步探討。 / This thesis presents a computer assisted environment for authoring test items for sentence reconstruction. Not only the teacher can author the test items for sentence reconstruction, but also the student can practice the test items in this environment. A test item for sentence reconstruction asks the student to arrange the shuffled words in a correct order, and this type of tests can examine the knowledge of sentence patterns and grammars of the student. However, the rearranged sentence may match with not only the correct sentence that the teacher wants but also other sentences which are legal. But enumerating all possible legal and acceptable answers for judging the answer of the student manually makes the teacher taking a big load. We construct a computer assisted environment for authoring test items for sentence reconstruction to lighten the load of the teacher. For the purpose of judging the answer of the student by a computer easily, the relative locations of the words are restricted so that we can restrict the sentences that the student arranges. When the teacher provides the correct answers, we try to find and return all of the sentences which may be legal for the teacher's consideration. However, our experience shows that it is difficult to find all of the legal sentences for a given set of words, and this problem associates closely with a certain syntactic research work. This thesis depends on basic word groups to construct an environment of test item authoring for English sentence reconstruction by merging word groups and using the Stanford Parser, and report an initial study of an environment of test item preparation for Chinese sentence reconstruction.