STRUCTURAL BIOLOGY OF VIRUS ASSEMBLY

工作总结。背景。我们的目标是阐明控制病毒组装的分子机制，最终目标是确定抗病毒化合物的靶点，表征病毒的抗原性，并了解大分子复合物的组装。本研究主要关注衣壳成熟过程中的大规模构象变化以及病毒粒子与宿主细胞的相互作用。我们在过去一年的主要进展如下：(1)乙型肝炎病毒衣壳（核心抗原）。尽管有有效的疫苗，乙型肝炎病毒仍然是一个巨大的公共卫生问题，促使进一步努力阐明其复制周期。我们一直在研究它的衣壳蛋白（核心抗原），我们首先发现它形成二聚体，能够自组装成两种不同大小的颗粒——一种是T=3的二十面体对称（90二聚体，直径28nm），另一种是T=4的（120二聚体，直径34nm）。我们还发现HBV的衣壳蛋白与大多数其他病毒不同，主要是α -螺旋状的。1997年，我们用空前高分辨率的0.9nm低温电子显微图计算了密度图，其中大部分二级结构都是直接可见的，包括形成二聚化基序的4-螺旋束。我们随后的研究（FY98）使用高分辨率标记技术确定了末端和内部六肽的精确位置（res. 78-83），从而通过我们的密度图描绘了多肽链的整体路径。隐含的折叠？在过去的几个月里，另一个实验室的x射线晶体学已经证实了这一发现。在1999财年，我们在HBV方面的主要工作是进一步完善我们的重金属簇标记技术的敏感性，以便能够更广泛和更深入地在其他复合物的研究中使用它们。我们的第一个成功是将十一原子金簇（十一金簇）连接到核心抗原的c端。最近，我们已经能够在相同的位置看到一个更小的标签，4原子四钠簇。(2)疱疹病毒衣壳组装与成熟。疱疹病毒构成了一个种类繁多的大型复杂DNA病毒家族。该家族的8个成员会导致人类疾病，包括皮肤病。我们研究了几种疱疹病毒家族成员的复杂衣壳组装过程，包括1型单纯疱疹病毒、类人猿巨细胞病毒和？去年卡波西肉瘤相关疱疹病毒在之前的工作中，我们将电子显微镜与生物化学和遗传学相结合，定义了疱疹病毒的分子解剖结构，并表征了六种主要衣壳蛋白在组装中的单个作用。尽管在序列水平上存在广泛的进化差异，但这些特性在不同的疱疹病毒中被证明是密切保守的，并且与任何其他已知的病毒家族非常不同。不过?令我们惊讶的是，我们已经确定了疱疹病毒和DNA噬菌体各自的组装途径之间越来越多的定性相似性。在1999财年，我们的主要重点是进一步表征HSV-1原衣壳，这是一种在结构、组成和稳定性上与成熟衣壳有显著不同的前体颗粒。在1997年，我们在用表达疱疹病毒衣壳蛋白的重组杆状病毒感染昆虫细胞的粗匀浆进行的体外组装实验中发现了这种颗粒。在1999年，我们通过在体外从纯化的蛋白质组装相同的颗粒，确定了这些蛋白质的形态发生能力，并确认了它们作为强制性前体的地位，并直接从受感染的细胞中分离出这些颗粒。后一项实验是用病毒蛋白酶的突变体进行的，在这种突变体中，原附体在感染细胞中积累：相比之下，原附体在野生型感染中很少见，因为它们在组装后不久就成熟了。我们现在的重点转向原衣壳成熟的动态方面。(3)噬菌体衣壳的结构、组装和成熟。我们对噬菌体衣壳组装的主要兴趣在于伴随前体衣壳成熟的巨大构象变化。这些变化是不可逆的，经常涉及亚基的部分再折叠，并且是严格控制的。因此，它们为深入了解蛋白质的大规模构象变化提供了独特的机会。尽管噬菌体在大小、序列和其他特性上差异很大，但成熟转化是一个普遍的特征，我们研究了几个系统中的这一行为事件，以利用每个系统的有利方面。99财年，前三年启动的T4研究取得了成果。T4的特点是其体积大，结构复杂，并且在其外表面结合了两种辅助蛋白。其中一种蛋白质gpSOC （12 kDa）通过在分子间界面上以钳状方式结合来稳定衣壳。我们现在已经能够可视化这两种蛋白质与T4衣壳表面的结合。噬菌体λ对应的蛋白称为gpD。受FCRF-Frederick的同事最近对gpD晶体结构测定的启发，我们分离了λ衣壳，并通过冷冻显微镜在1.5 nm分辨率下对其结构进行了表征。我们的地图清楚地显示了gpD三聚体结合在三角位点上，建立了晶体结构代表了蛋白质的生物活性状态，并揭示了它的结合模式。进一步有效地利用晶体学数据是我们最近对HK97系统的研究的核心，斯克里普斯研究所的同事已经解决了HK97系统的成熟衣壳（在这种情况下，没有辅助蛋白：而是通过共价交联稳定了该衣壳）。我们已经通过冷冻显微镜确定了前体衣壳的结构到~ 1.2 nm的分辨率，并将这两种信息来源结合起来，目的是在原子细节上描述与成熟相关的构象变化。(4)脊髓灰质炎病毒与宿主细胞的相互作用：受体结合。当小核糖核酸病毒（如人类病原体脊髓灰质炎病毒）遇到易感细胞时，病毒会与细胞受体结合。与包膜病毒（通过与细胞膜融合或内吞作用进入细胞）不同，脊髓灰质炎病毒的进入似乎是通过其衣壳的变化介导的，这种变化发生在其与受体相互作用之后，并导致其RNA释放到宿主细胞中。在过去五年进行的一个项目达到高潮时，我们完成了对与转移过程有关的脊髓灰质炎病毒的两种构象的结构分析，并提交了一篇描述这些结果及其影响的论文供发表。再一次，这项研究结合了晶体学和冷冻显微镜，系统地调整了脊髓灰质炎病毒（自1985年以来已知）的晶体结构，以适应我们的病毒两种细胞进入状态的冷冻显微镜模型。由此产生的变化似乎在微观上模拟了地震中构造板块的运动，因为考虑蛋白的刚性-桶结构域发生了移动，产生了RNA可能被释放的缝隙。我们也收到了可溶性受体的寄售？一种3域类igg分子，详细研究了其与病毒颗粒的结合方式。(5) HIV gp41外结构域三聚体淀粉样沉积与艾滋病相关痴呆有关。痴呆性艾滋病患者的病情与HIV包膜糖蛋白的gp41外域的不溶性聚集体在神经组织中的沉积有关。gp41三聚体的结构是已知的：每个亚基由两个由环路连接的α螺旋组成。一组三个螺旋形成一个盘绕线圈和其他三个，相反的方向，螺旋包裹他们。在这个IATAP项目的支持下，我们通过电子显微镜研究了这些聚集体，结合了传统的阴性染色，未染色标本的暗场扫描透射电子显微镜（STEM）的质量测定，以及非常规的阴性染色，其中钒酸盐包埋的材料通过暗场STEM可视化。综上所述，这些观察结果表明，聚集体由不同数量的gp41三聚体组成（大约7到70个），通过它们在聚集体核心（内部）联系的连接环之间的相互作用联系在一起。填料排列不是很规则，但三聚体以长轴径向取向为主。这些聚集体的发病机制可能与淀粉样蛋白有关。在许多疾病状态下积累的难以消化的蛋白质聚集体，其区别在于，尽管与富含β片的传统淀粉样蛋白不同，gp41聚集体主要是α -螺旋状的。

Summary of Work. Background. We aim to elucidate the molecular mechanisms that control the assembly of viruses with the ultimate goals of defining targets for antiviral compounds, characterizing viral antigenicity, and understanding the assembly of macromolecular complexes in general. This research focuses on the large-scale conformational changes that accompany capsid maturation and on the interaction of virions with host cells. Our major progress over the past year is as follows: (1) Hepatitis B Virus Capsid (Core Antigen). Despite the availability of effective vaccines, HBV remains a public health problem of immense proportions, motivating further efforts to elucidate its replicative cycle. We have been studying its capsid protein (core antigen), which we first found to form dimers that are capable of self-assembly into particles of two different sizes - one with T=3 icosahedral symmetry (90 dimers; 28nm diameter), the other with T=4 (120 dimers; 34nm diameter). We also found that the capsid protein of HBV, unlike most other viruses, is predominantly alpha-helical. In 1997, we calculated a density map from cryo-electron micrographs at the unprecedentedly high resolution of 0.9nm, in which much of the secondary structure was directly visible, including the 4-helix bundle that forms the dimerization motif. Our subsequent research (FY98) used high resolution labelling techniques to establish the precise locations of both termini and an internal hexapeptide (res. 78-83), thus delineating the overall path of the polypeptide chain through our density map. The implied fold ? a novel one for a capsid protein - has been confirmed by X-ray crystallography from another laboratory in the past few months. In FY99, our main effort on HBV has been to further refine the sensitivity of our heavy metal cluster labelling techniques in order to be able to use them more widely and more incisively in studies of other complexes. Our first success was with an 11-atom gold cluster, undecagold, attached to the C-terminus of core antigen. More recently, we have been able to visualize an even smaller label, the 4-atom tetrairidium cluster, at the same site. (2) Capsid Assembly and Maturation of Herpesviruses. Herpesviruses constitute a widely diversified family of large, complex, DNA viruses. Eight members of the family cause diseases in humans, including skin diseases. We have studied the elaborate process of herpesvirus capsid assembly for several family members, including herpes simplex type 1, simian cytomegalovirus, and ? in the last year - Kaposi sarcoma-associated herpesvirus. In previous work, we combined electron microscopy with biochemistry and genetics to define the molecular anatomy of herpesviruses and to characterize the individual roles of the six major capsid proteins in assembly. These properties have turned out to be closely conserved among different herpesviruses, despite wide evolutionary divergence at the sequence level, and quite distinct from those of any other known family of viruses. Nevertheless ? to our surprise - we have identified a growing number of qualitative similarities between the respective assembly pathways of herpesviruses and DNA bacteriophages. In FY99, our main focus has been to further characterize the HSV-1 procapsid, a precursor particle that differs markedly in structure, composition, and stability from the mature capsid. In FY97, we discovered this particle in in vitro assembly experiments performed using crude homogenates of insect cells infected with recombinant baculoviruses expressing herpesvirus capsid proteins. In FY99, we have established the morphogenic competence of these proteins by assembling identical particles in vitro from purified proteins, and confirmed their status as obligatory precursors and by isolating such particles directly from infected cells. The latter experiments were performed with mutants in the viral protease for which, procapsids accumulate in infected cells: in contrast, procapsids are rare in wild-type infections because they mature shortly after they are assembled. Our focus is now turning to dynamic aspects of procapsid maturation. (3) Structure, Assembly, and Maturation of Bacteriophage Capsids. Our primary interest in bacteriophage capsid assembly lies in the monumental conformational changes that accompany maturation of precursor capsids. These changes are irreversible, frequently involve partial refolding of the subunits, and are stringently controlled. As such, they afford unique opportunities for insight into large-scale conformational changes in proteins. Although phages vary widely in size, sequence, and other properties, the maturation transformation is a universal feature and we study this behavioral event in several systems to exploit expedient aspects of each. In FY99, studies on T4 initiated over the preceding three years have come to fruition. T4 is distinguished by its large size, order of complexity, and in the binding of two accessory proteins to its outer surface. One of these proteins, gpSOC (12 kDa), stabilizes the capsid by binding in a clamp-like fashion over intermolecular interfaces. We have now been able to visualize the binding of both of these proteins to the T4 capsid surface. The corresponding protein of bacteriophage lambda is called gpD. Motivated by the very recent crystal structure determination of gpD by colleagues at FCRF-Frederick, we isolated lambda capsids and characterized their structure at 1.5 nm resolution by cryo-microscopy. Our map clearly shows gpD trimers bound over the trigonal sites, establishing that the crystal structure represents the biologically active state of the protein and revealing its mode of binding. Further productive exploitation of crystallographic data is central to our most recent work on the HK97 system, whose mature capsid has been solved by colleagues at Scripps Institute (and which, in this instance, has no accessory proteins: rather this capsid is stabilized by covalent crosslinks). We have determined the precursor capsid structures to ~ 1.2 nm resolution by cryo-microscopy and are combining the two sources of information with the aim of developing an account of the maturation-related conformational changes in atomic detail.(4) Interaction of Poliovirus with Host Cell: Receptor Binding. When a picornavirus such as the human pathogen, poliovirus, encounters a susceptible cell, the virus binds to a cellular receptor. Unlike enveloped viruses, which enter cells by fusing with the cell membrane or endocytosis, poliovirus entry appears to be mediated by changes in its capsid that follow its interaction with the receptor and lead to discharge of its RNA into the host cell. In the culmination of a project pursued over the preceding five years, we have completed our structural analysis of two conformers of poliovirus that are implicated in the transfer process and have submitted for publication a paper describing these results and their implications. Again, this study involved a combination of crystallography and cryo-microscopy, whereby the crystal structure of poliovirus (known since 1985) was systematically adjusted to fit our cryo-microscopic models of the two cell-entry states of the virus. The resulting changes appear to emulate, in miniature, the movements of tectonic plates in an earthquake as the rigid beta-barrel domains of the consid proteins shift, creating gaps through which the RNA may be released. We have also received consignments of a soluble form of the receptor ? a 3-domain IgG-like molcule, and studied in detail its mode of binding to the virus particle.(5) Amyloid-like Deposits of HIV gp41 Ectodomain Trimers Implicated in AIDS-Related Dementia. The condition of demented AIDS patients has been correlated with the deposition in neurological tissue of insoluble aggregates of the gp41 ectodomain of the HIV envelope glycoprotein. The structure of the gp41 trimer is known: each subunit consists of two alpha-helices connected by a loop. One set of three helices forms a coiled-coil and the other three, oppositely oriented, helices pack round them. In this new project, pursued within the aegis of the IATAP program, we have studied these aggregates by electron microscopy, using a combination of conventional negative staining, mass determinations by dark-field scanning transmission electron microscopy (STEM) of unstained specimens, and unconventional negative staining, whereby vanadate-embedded material is visualized by dark-field STEM. Taken together, these observations imply that aggregates consist of variable numbers (approximately 7 to 70) of gp41 trimers, associated by means of interactions among their connecting loops which associate at the core (interior) of the aggregate. The packing arrangement is not very regular but the trimers are predominantly oriented with their long axes oriented radially. The pathogenesis of these aggregates may relate to that of amyloid ? indigestible aggregates of proteins that accumulate in a number of disease states, with the distinction that although unlike conventional amyloid, which is rich in beta sheets, the gp41 aggregates are primarily alpha-helical.