利用兩種非結晶學研究對小型病毒的深入探討

Abstract

現今，結構生物學已經成為一門能有效提供原子層次物理作用力證據的學科。在十年前，X光結晶學繞射曾是最為普遍取得高解析蛋白結構的方法。然而，隨著科技的進步，已經有越來越多無需長晶的結構測定方法被提出，得以打破既有的蛋白晶體製備瓶頸，甚至加速資料的運算處理。在這篇論文研究中，我們將藉由小型病毒顆粒為例，比較兩個現進蛋白結構測定策略: 冷凍電子顯微鏡單顆粒三維重構 (cryo-electron microscopy single particle reconstruction),以及X光自由電子雷射同調繞射成像術 (X-ray free-electron laser coherent diffractive imaging, XFEL-CDI)。此研究不僅是為了兩種光源的評測，也期望能從結果中找到足夠的生物意義。 此研究的第一個部分，我們從蚊子C6/36細胞株中純化出HeDNV病毒，並藉由冷凍電顯單顆粒三維重構技術得到此病毒2.7埃近原子解析度的模型。目前已知蚊子被同時感染densovirus後會顯著的降低登革熱病毒在蚊子細胞的感染程度，故此部分的研究目的是藉由冷凍電顯技術提供densovirus的高解析結構資訊，以利未來對抗蚊子傳播的惡性疾病。HeDNV為蚊子densovirus的一種，在病毒分類學中屬於Brevidensovirus，它包含了一個小型約20-22奈米的20面對稱，無外套膜的病毒外殼，以及長約4千個鹼基的T型髮夾狀單股反向核醣核酸基因組。由於解析度夠高，我們得以分辨組成此病毒顆粒的胺基酸骨幹，也展示了冷凍電顯的強大之處。取出HeDNV外殼的單體組成單元分析後發現，此結構由一個在其他densovirus常見的'jelly roll'核心組成。此外，這個病毒結構展示了其內部的單股核酸基因體有三個區塊能和外殼特定胺基酸形成相互作用力。這是目前已知densovirus中能解析到最多核酸作用的結構模型。甚至，這是第一個陳述詳細的胺基酸-核酸-二價金屬離子共同作用力的densovirus蛋白結構，而這些作用力是維持該病毒單元體相互結合。藉由比較此病毒結構與其他物種的densovirus結構，我們能提出一個不同於傳統基因序列的病毒分類方法，而是以結構觀點出發的演化分析方法。 第二個部分著重在討論另外一種光源：X光自由電子雷射(XFEL)。它具有'飛秒突發'以及'高同調'的特性，藉由與同調光繞射成像術(CDI)的結合，被認定為能實現'無需結晶就能解析結構'的方法。由於我們著眼在接近自然狀況下小型顆粒的結構數據表現，在實驗設計上不同於其他團隊的做法。因此此次研究，選擇了人類B型肝炎的類病毒顆粒做為已知結構標的樣品，欲藉由冷凍電顯的輔助，調查此方法對於30奈米大小不同濃度的的病毒顆粒在低解析度結構因子分布的狀況。其結果發現，在CDI影像中，高濃度的病毒顆粒呈現出有別於低濃度的三個繞射環訊號，經分析後發現該訊號屬於” 低解析度結構因子”。我們藉由調整冷凍電顯的”離焦參數”以及”開啟相位板”增強影像低頻範圍的訊號後，與”CDI”的結果比較後發現，三者的”power-spectra”圖譜分佈吻合，且相位板的開啟能更顯注於觀察低解析度結構因子的資訊。遺憾的是，與冷凍電顯的結果相比，X光自由電子雷射在本次的實驗設計並不能取得更高解析度的資訊。 綜合上述兩部分的實驗，此論文提供了一系列實測比較的數據，期待這個研能為結構生物學研究者在選擇合適的分析光源上有所依據。

Nowadays, structural biology has become a reliable subject for providing biophysical interaction evidence in near atomic scale. Decades ago, X-ray crystallography was the most common way to get high resolution protein structural information. However, with the technologies improve, more and more non-crystallographic methods have been developed to breakthrough sample preparation bottleneck, even more, speed-up data process. In this research, we compared two modern strategies used for measuring small virus capsid protein structural information: cryo-electron microscopy (cryo-EM) single particle reconstruction, and X-ray free-electron laser coherent diffractive imaging (XFEL-CDI), not only for benchmark, but also for dig-out any biological significance among these virus sample targets. In the first part, we purified HeDNV virus from mosquito C6/36 cell line and successfully solved mosquito Haemagogus equinus densovirus (HeDNV) at 2.7 Å near-atomic resolution by cryo-EM single particle reconstruction. Mosquito densovirus infection suggests markedly reduced severity of Dengue virus infection in mosquito cell. The main purpose of this project is to provide high-resolution structural information from a native mosquito densovirus by cryo-EM for fighting mosquito-borne diseases. HeDNV belongs to Brevidensovirus in virus taxonomy, contains small, non-enveloped, T=1 icosahedra (60 subunits), 20-22nm in diameter capsid structure, and a linear, negative-sensed single-stranded DNA of 4–4.2 kb T-shaped hairpin genome. Since the resolution is high enough to interpret amino acid side chain backbone, the results of this research has demonstrate the power of cryo-EM single particle reconstruction. The topology and geometry of HeDNV asymmetric unit shared a classical “jelly roll” core which is highly conserved in all densoviruses. Besides, the electron map distribution of ssDNA genome and capsid protein represents three groups of ordered ssDNA bases interactions with specific amino acids, which is the most abundant of nucleotide density solved so far in densovirus. Furthermore, this is the first report of detailed protein / DNA / metal ion interaction networks contribute to subunit-subunit interaction in densovirus. By comparing of available invertebrate DNV structures, we hope can provide a structural based analysis to classify the sub-classes of densoviruses. Second part, we are interested in another light resource: X-ray free electron laser (XFEL). It provides femto-second pulse and high coherence. With the combination of coherent diffractive imaging (CDI) method, it may suggest to realize “solving structure without crystal.” In order to investigate “low resolution structural order” of XFEL-CDI technique revealing 30nm biological sample and validate by cryo-EM. In this research, we are offered to use a liquid-enclosure chamber for sample loading in order to mimic the native state of biological sample, and we tend to use icosahedral Human Hepatitis B virus (HBV) capsid, a smaller and structure-known virus particle as a benchmark for our XFEL setup, with the help of cryo-EM. The results showed that there are three more diffraction rings appear in high concentrated HBV capsid samples, which represented the “structure factor” of HBV capsid. To validate from cryo-EM by increasing low frequency signal, we used “defocus” and “phase plate” method, compared with CDI results in reciprocal space. The power spectrum of XFEL-CDI, CEM, and CEM with VPP fits well in low frequency represents the structure factor of HBV capsid. The limited power spectrum of was collected in high concentrated HBV capsid solution, brings out the insufficiency of single particle XFEL-CDI in 30nm Virus-like sample so far. Furthermore, phase plate contributes significant contrast intensity in low frequency area, which becomes a useful tool to get macro-molecule shape information. Taken together, I hope this research can provide a clear answer to researchers choosing suitable light source for their structural biology research purpose.