Structural Evolution of the Dimer interface in Hepadnaviridae

肝病毒科是一个古老的包膜病毒家族，出现于4亿多年前。肝病毒科感染大多数脊椎动物，包括鱼类（中肝病毒和副肝病毒）、两栖动物、爬行动物（疱疹病毒）、鸟类（无肝病毒）和哺乳动物（正肝病毒）。最主要的家族成员是人类乙型肝炎病毒，全世界约有2.5亿慢性携带者。肝病毒科的所有病毒粒子都有由单一类型核心蛋白（HBc）的多个拷贝形成的衣壳。衣壳被多形性脂质包膜包围，该包膜被表面蛋白（HBs）密集包裹。在衣壳内，HBc形成二聚体，在内二聚体界面处形成突出的尖刺。这些尖峰与包膜中的HBs相互作用。系统发育分析和现有的结构知识表明，只有衣壳中HBc二聚体之间的接触是保守的，而内部二聚体界面和突出的尖刺则不是。这表明衣壳的形成是进化维持的，而包膜和分泌的位点适应或与各自的宿主共同进化。虽然人类HBc衣壳的许多结构是已知的，但其他属衣壳的结构是罕见的，并且对于准肝病毒，中肝病毒和疱疹病毒来说是未知的。人乙肝病毒（正肝病毒）和鸭乙肝病毒（阿维肝病毒）衣壳的比较表明，阿维肝病毒在刺突两侧有额外的折叠结构域。这些域很可能与包膜相互作用。相反，原肝病毒的相互作用位点在两种单体的作用下映射到内部二聚体界面。这种与HBs结合的改变伴随着二聚体界面的重组。这就提出了二聚化和HBs结合是如何共同进化的问题。为了解决这个问题，我们将通过电子冷冻显微镜和图像处理确定来自不同属肝病毒科的衣壳的代表性结构。分析将包括古病毒的衣壳，这些衣壳是由鳄鱼、斑胸草雀和斑胸草蜥基因组中的内源性片段重建的。古衣壳的结构将显示出尖刺结构是如何随着地质时代而变化的。最后，利用肽点阵列文库绘制不同属HBs和HBc之间的结合位点，并通过突变确定衣壳上的相互作用位点进行验证。提出的工作解决了肝病毒科结构进化的一个基本方面，目前由于缺乏参考结构和尖峰区域的序列保护而无法建模。

Hepadnaviridae are an ancient family of enveloped viruses that has emerged more than 400 million years ago. Hepadnaviridae infect most vertebrates including fish (meta- and parahepdanavirusses), amphibians, reptiles (herpetohepadnaviruses), birds (avihepadnaviruses) and mammals (orthohepadnaviruses). The most prominent family member is the human Hepatitis B virus with some 250 million chronic carriers world-wide. All virions of hepadnaviridae have capsids that are formed by multiple copies of a single type of core proteins (HBc). The capsids are surrounded by a pleiomorphic lipidic envelope that is densely packed with surface proteins (HBs). Within capsids, HBc forms dimers with protruding spikes at the inner-dimer interface. These spikes interact with HBs in the envelope. Phylogenetic analysis and current structural knowledge show that only the contact between HBc dimers in capsids is conserved while the inner-dimer interface and thus the protruding spike is not. This suggests that capsid formation is evolutionary maintained, while sites for envelopment and secretion adapt to or co-evolve with their respective hosts. While many structures of human HBc capsids are known, structures from capsids of other genera are rare and are unknown for para-, meta- and herpetohepadnaviruses. The comparison of capsids from human HBc (orthohepadnavirus) and duck HBc (avihepadnavirus) show that avihepadnaviruses have additional domains that fold at both sides of the spike. It is likely that these domains interact with the envelope. In contrast, the interaction site in orthohepadnaviruses maps to the inner dimer interface with contributions from both monomers. Such altered binding to HBs is concomitant with a reorganization of the dimeric interface. This raises the question how the dimerization and HBs binding have co-evolved. To address this question, we will determine representative structures of capsids from the different genera of Hepadnaviridae by electron cryo microscopy and image processing. The analysis will include capsids of paleo-viruses, which have been reconstructed from endogenous fragments in the genomes of crocodile, zebra finch and Melopsittacus. The structures of the paleo-capsids will show how the spike architecture has changed over geological ages. Finally, the binding sites between HBs and HBc in different genera will be mapped with peptide spot array libraries and validated by mutations that pinpoint the interaction sites on the capsids. The proposed work addresses a fundamental aspect of the structural evolution in hepadnaviridae that is currently inaccessible to modelling due to a lack of reference structures and sequence conservation in the spike region.