利用分子動力學與密度泛函方法研究普昂蛋白類澱粉樣纖維之結構與特性

Abstract

因為結構變化異常導致疾病的蛋白質，最廣為人所知，並且被大量、持續地研究的，就是普利子 (Prion Protein；Prion)、或稱普利昂蛋白。一般狀況下，在生物體中作用正常，稱為「cellular PrP」( PrPC )；而產生結構上的變化之後，是一種具有感染力的蛋白質粒子，稱作「scrapie PrP」 ( PrPSc )。包含數個錯誤褶疊產物的寡聚合物 (oligomer)，將「傳染」其他的正常的 PrPC，使結構產生變化、並且參與堆疊 (aggregation)。變性的過程中，蛋白質可以高度堆疊 (highly packed) 的 β-褶板二級結構比例大量增加，並形成所謂類澱粉樣纖維 (amyloid fibril)。這在各種神經退化性 (neurodegeneration) 病變中均會被發現。於是，解開形成大量 β-褶板、並繼續堆疊的原因與機制，將是預防或是治療此類病變的第一步。 然而，利用各種實驗方法，能得到的結構訊息有限，只能得到片面的結構參數，如：纖維長度、直徑、原子與原子核之間距離與骨幹扭轉角等等，或是概括的性質，如：二級結構為螺旋或摺板。仍然無法詳細描述體系由單股 β-褶板延伸、形成原纖維，或是原纖維之間或有、或無層狀結構，並繼續堆疊成為類澱粉狀纖維的動態或靜態資訊。此時，分子模擬藉由描述體系中各種作用力，並在因而得到的位能面上尋找能量最低的最穩定結構與其特性，便是上述探求蛋白質摺疊過成各種穩定狀態：自然狀態、過渡態、與更穩定的錯誤摺疊聚集物等的良好方法。 本實驗室已經對人類普昂蛋白127-147多肽片段 (huPrP127-147) 進行一系列實驗，本論文將所得之結構參數列為分子模型建構之考量。接著以分子動力學方法，探求普昂蛋白的可能型態。當進一步分析模擬得到的通心粉狀結構（penne，其直徑與穿透式電子顯微鏡觀察成熟類澱粉樣纖維成長過程中，相對較短、較細的中間型態相符合，我們相信這是形成原纖維初期 (多肽片段股數大於16股)，主要因為靜電作用力而產生的穩定分子結構。 另外本實驗室一個獨立的研究發現，以同位素標定對齊羰基碳的狀況之下，倉鼠普昂蛋白109-122片段 (HaPrP109-122) 的紅外光譜測量，相對於其他標定樣品所得 Amide I’吸收帶，有額外的 8cm-1 紅位移，為了釐清成因是否主要來自於單層之中對齊羰基碳拉伸運動的躍遷偶極耦合作用，我們利用密度泛函從頭算方法，對不同的結構模型進行分析、以及利用古典偶極耦合模型進行近似計算，得到來自於單層之中的躍遷偶極耦合作用較層與層之間之作用結果強一個數量級的結果，故可證明額外的 Amide I’ 紅外光吸收帶，主要來自於單層結構中的躍遷偶極耦合。

Prion proteins, which lead many mammal diseases from the abnormally structural variations, result in numerous scientific investigations in different aspects. Those proteins usually functionalize well in biological environment, called cellular prion protein (PrPC). Until main transformation on secondary structure to beta-sheet riched form, the infectious protein 'particles' become scrapie prion protein (PrPSc) named after the first observed prion disease in sheep. Oligomer including several mis-folding prion would infects other normal cellular prion proteins and makes them aggregate in further. In the process of denaturation, the ratio of highly packed beta-sheets in prion proteins will increase, gradually and finally form the so-called amyloid fibrils. Amyloid fibrils are discovered in many neurodegeneration diseases. Therefore, figuring out the mechanism(s) of transformation between PrPC and PrPSc, the reason aggregation proceeds will be the first step to prevent and even to remedy this kind of diseases. However, structural information from experimental data is limited and partial. Those parameters are the length and diameter of filaments from TEM, inter-atomic distances and backbone torsional angles from solid state NMR, and the deterministic secondary structure, etc. Those data can not give us the direct evidences in describing the system from single strand to extended beta-sheets with or without layer structure and from protofibers to mature fibers. Thus, molecular simulations with description of interactions in the system, will be an important way to sample and search the structures in potential energy surface. Then we may be able to comprehend how those relatively more stable states in the process of prion protein transformation. We already got several information from investigation of human prion protein fragment 127-147 (huPrP127-147) and using those parameters during modeling the system. After modeling, we use molecular dynamics simulations as the way to find out the most stable structure and how the stripe formed as the TEM image. In our surprise, we get penne-like hyper-structure as the number of strands larger than 16. We think this spectacular model might be related to the initial state of the process forming protofibers. On the other hand, research on the other prion system, hamster prion protein fragment 109-122 (HaPrP109-122), shows that there is an additional 8 cm-1 red-shift of amide I' band of isotopically labeled AA specimen comparing with other samples of different labeling scheme. We believe the major reason of this red shift result from the transition dipole coupling (TDC) interaction of local stretching mode of labeled carbonyl carbon on Ala117 residue in-register to other one. But in this system, there might be contribution from inter-layer TDC. In order to clarify the source of major TDC effect, we use ab initio calculation of density funcational theory and classic model of dipolar coupling mechanism. The results show that intra-layer contribution is larger than inter-layer one in at least one order of magnitude.