Retrovirus Assembly and Maturation

我们继续分析HIV-1和小鼠白血病病毒（MLV）颗粒的组装和成熟。具体来说，我们已经描述了HIV-1 Gag蛋白的特性，这是HIV-1病毒颗粒的基本组成部分。我们发现Gag蛋白在溶液中处于单体-二聚体平衡。我们确定了导致二聚化的界面；这是一个先前被描述为HIV-1衣壳蛋白（形成Gag的一部分）二聚化的位点。我们还发现肌醇六磷酸（IP6），当加入核酸时，它调节由Gag组装的病毒样颗粒（VLPs）的性质，导致从单体二聚体到单体三聚体平衡的转变。一些证据表明，当核酸和IP6同时存在时，核酸结合到蛋白质的c端核衣壳（NC）结构域，而IP6结合到蛋白质的n端基质（MA）结构域。通过突变Gag中的二聚体界面，我们产生了一种在溶液中保持单体的蛋白质。然后，我们使用各种各样的实验方法，包括凝胶过滤，静态和准弹性光散射，沉降速度分析，小角度中子散射，以及分子模型，来表征这种蛋白质的构象。所有的数据表明，蛋白质在溶液中折叠，其末端在三维空间中彼此靠近。相反，已知该蛋白在组装的病毒颗粒中是一个高度延伸的棒；因此，它在组装时必须经历一个主要的构象变化。我们分析了嵌合蛋白的组装特性，其中HIV-1 Gag的c端结构域，包括NC结构域，被亮氨酸拉链结构域取代。这些蛋白质在体内组装成病毒样颗粒；这些颗粒在形态上与野生型Gag组装的颗粒几乎相同，但似乎不含RNA。这些蛋白可以与野生型Gag结合。当从细菌中纯化时，这些嵌合蛋白是低聚的，但除非添加辅助因子，否则不会组装；辅因子可以是RNA或磷酸肌醇，可能起到中和MA和/或CA结构域正电荷的作用。野生型Gag的正常组装可能也需要这些中和的辅因子，但这种要求在野生型Gag的实验中并不明显，因为在野生型的情况下，RNA也需要NC结合。我们正在对MLV Gag蛋白进行类似的分析。值得注意的是，我们发现该蛋白的特性与HIV-1 Gag有很大的不同。因此，它不会在溶液中寡聚，也不会在溶液中折叠。相反，它看起来是一根延伸的棒，在溶液中与在组装的病毒颗粒中具有大致相同的尺寸。a. SP1结构域在HIV-1组装中的功能分析正常的HIV-1颗粒组装对SP1的变化极为敏感。我们发现，一个代表SP1的18残基肽在高浓度时呈α -螺旋构象，而在低浓度时则不是。我们提出SP1是一个“开关”，当Gag在RNA上寡聚时改变构象。我们将研究这种浓度依赖性变化及其在组装中的可能作用。我们还将对来自MLV Gag的肽进行类似的实验，其中CA-NC边界的序列与HIV-1中的序列完全不同。与Bogdan Dragnea博士合作，我们发现HIV-1 Gag在纳米颗粒“模板”上比在自由溶液中更均匀地组装。我们将利用这一现象来提高粒子的低温电子显微镜（cryo-EM）图像的分辨率。我们还将确定在体内不能正确组装的突变体是否可以在这种新的实验情况下形成球形颗粒。我们发现MLV Gag蛋白的二聚化倾向远弱于HIV-1 Gag。与HIV-1 Gag不同的是，MLV Gag是溶液中的棒状结构，其尺寸与病毒样颗粒（VLP）大致相同。在矩阵（MA）和p12中短的脯氨酸运行，以及CA末端附近的“电线”，有助于其刚性。d.哺乳动物细胞内HIV-1 Gag的构象分析我们已经证明HIV-1 Gag在溶液中是折叠的，但当它组装成全尺寸的VLPs或MA和NC结构域的首选配体存在时，它会延伸。我们将通过荧光共振能量转移（FRET）确定折叠和扩展形式的细胞位置。e. HIV-1 MA和Gag蛋白与人工膜的结合。Hirsh Nanda和Matthias Losche，我们将在控制良好的实验条件下分析HIV-1 MA和Gag与特定组成的人工膜的结合。与项目相关的专利：美国专利#6,001,555：“通过在病毒核衣壳蛋白上附着高度保守的锌指来鉴定和使用灭活HIV-1和其他逆转录病毒的化合物的方法”；一九九九年十二月十四日发布；路易斯·e·亨德森，拉里·o·亚瑟，威廉·g·赖斯和艾伦·雷小山。描述通过与内部病毒成分反应灭活逆转录病毒（包括HIV-1）的化合物。灭活颗粒具有完全天然的外表面，对许多免疫学实验以及可能的疫苗使用都很有用。[对应2011年10月HIV耐药项目实地考察报告中的Rein项目1]

We have continued our analysis of assembly and maturation of HIV-1 and murine leukemia virus (MLV) particles. Specifically, we have characterized the properties of the HIV-1 Gag protein, the fundamental building block of the HIV-1 virus particle. We found that the Gag protein is in monomer-dimer equilibrium in solution. We identified the interface responsible for the dimerization; it is a site previously described as the site at which the HIV-1 capsid protein (which forms a portion of Gag) dimerizes. We also found that inositol hexakisphosphate (IP6), which modulates the nature of the virus-like particles (VLPs) assembled from Gag when nucleic acid is added, causes a shift from monomer-dimer to monomer-trimer equilibrium. Several lines of evidence indicate that when both nucleic acid and IP6 are present, the nucleic acid binds to the C-terminal, nucleocapsid (NC) domain of the protein while the IP6 binds to the N-terminal matrix (MA) domain. By mutating the dimer interface in Gag, we generated a protein which remains monomeric in solution. We then used a wide variety of experimental approaches, including gel filtration, static and quasi-elastic light-scattering, sedimentation velocity analysis, and small-angle neutron scattering, as well as molecular modeling, to characterize the conformation of this protein. All of the data indicate that the protein is folded over in solution, with its ends near each other in 3-dimensional space. In contrast, the protein is known to be a highly extended rod in assembled virus particles; thus, it must undergo a major conformational change when it assembles. We have analyzed the assembly properties of chimeric proteins in which the C-terminal domains of HIV-1 Gag, including the NC domain, is replaced by a leucine-zipper domain. These proteins assemble into virus-like particles in vivo; these particles are morphologically almost identical to those assembled from wild-type Gag, but appear to contain no RNA. These proteins can coassemble with wild-type Gag. When purified from bacteria, these chimeric proteins are oligomeric, but do not assemble unless a cofactor is added; the cofactors can be either RNA or inositol phosphates, and probably function to neutralize positive charges in the MA and/or CA domains. Normal assembly by wild-type Gag presumably also requires these neutralizing cofactors, but this requirement was not apparent in experiments with wild-type Gag because RNA is also required for NC binding in the wild-type case. We are performing a similar analysis of the MLV Gag protein. Remarkably, we find that the properties of this protein are quite different from those of HIV-1 Gag. Thus, it does not oligomerize in solution, and is not folded over in solution. Rather, it appears to be an extended rod, with approximately the same dimensions in solution as in the assembled virus particle. Recent Accomplishments and Current Research: a. Analysis of the function of the SP1 domain in HIV-1 assembly Proper HIV-1 particle assembly is extremely sensitive to changes in SP1. We found that an 18-residue peptide representing SP1 adopts an alpha-helical conformation when it is at high concentration, but not at low concentration. We proposed that SP1 is a "switch" that changes conformation when Gag oligomerizes on RNA. We will study this concentration-dependent change and its possible role in assembly. We will also perform analogous experiments with a peptide from MLV Gag, in which the sequence at the CA-NC border is radically different from that in HIV-1. b. Template-directed in vitro particle assembly In collaboration with Dr. Bogdan Dragnea, we have found that HIV-1 Gag assembles more uniformly on nanoparticle "templates" than out of free solution. We will use this phenomenon to improve the resolution of cryoelectron microscopy (cryo-EM) images of particles. We will also determine whether mutants that fail to assemble correctly in vivo can form spherical particles in this novel experimental situation. c. Characterization of MLV Gag protein We found that MLV Gag protein has a far weaker tendency to dimerize than HIV-1 Gag. Also unlike HIV-1 Gag, MLV Gag is a rod in solution, with approximately the same dimensions as it has in a virus-like particle (VLP). Short proline runs in matrix (MA) and p12, and the "electric wire" near the end of CA, contribute to its rigidity. d. Analysis of conformation of HIV-1 Gag within mammalian cells We have shown that HIV-1 Gag is folded over in solution, but extends when it assembles into full-size VLPs or when preferred ligands for both the MA and NC domains are present. We will determine cellular locations of the folded and extended forms by fluorescence resonance energy transfer (FRET). e. Binding of HIV-1 MA and Gag proteins to artificial membranes In collaboration with Drs. Hirsh Nanda and Matthias Losche, we will analyze the binding of HIV-1 MA and Gag to artificial membranes of defined composition under well-controlled experimental conditions. Patent Linked to Project: U.S. Patent #6,001,555: "Method for Identifying and Using Compounds That Inactivate HIV-1 and Other Retroviruses by Attaching Highly Conserved Zinc Fingers in the Viral Nucleocapsid Protein"; issued December 14, 1999; Louis E. Henderson, Larry O. Arthur, William G. Rice, and Alan Rein. Describes compounds that inactivate retroviruses, including HIV-1, by reaction with internal viral components. The inactivated particles have completely native outer surfaces and are useful for many immunological experiments, as well as for conceivable vaccine use. [Corresponds to Rein Project 1 in the October 2011 site visit report of the HIV Drug Resistance Program]