Knowledge of the molecular structure of trimeric Env on intact viruses and delineating the mechanisms of transmission are central to the design of effective immunogens and therapeutic agents to combat HIV/AIDS. In addition, related enveloped viruses such as influenza and Ebola may share similar mechanisms for viral entry, and as such structural studies of these viruses may offer insight towards vaccine design for all three of these viruses. We have continued to make significant progress towards these goals over the last year. HIV-1 infection begins with the binding of trimeric viral envelope glycoproteins (Env) to CD4 and a co-receptor on target T-cells. Understanding how these ligands influence the structure of Env is of fundamental interest for HIV vaccine development. Using cryo-electron microscopy, we have determined a large number of structures for native HIV-1 Env in its unbound state, bound to soluble CD4, or bound to neutralizing antibodies. These studies described a series of conformations prior to the pre-hairpin state of the complex. Using a cleaved, solubilized version of the HIV-1 Env, we were able to achieve higher resolution structures of the trimer in its pre-fusion (6 Angstrom resolution) and open, activated (9 Angstrom resolution) conformations, revealing that the outer gp120 subunits rotate around three internal helices during spike opening. We are continuing to explore the mechanism for neutralization by a variety of different neutralizing antibodies, and to determine what aspects of binding lead to neutralizing potential. In the past year, we have applied these same techniques to Ebola envelope glycoprotein and to influenza HA trimers, as these enveloped viruses have similarities in structure, and may share a similar mechanism for entry. The Ebola virus is an emerging pathogen that has become a critical target for vaccine and therapeutic development. Ebola displays many copies of a single complex, the envelope glycoprotein, on the surface of mature virions. Broadly neutralizing antibodies directed at the envelope glycoprotein have proven effective in preventing viral fusion; however, certain structural features of the envelope glycoprotein, such as the mucin domains, which appear to enhance Ebola infectivity, have remained poorly understood. Understanding the placement of the mucin domain is critical, especially considering that many neutralizing antibodies bind in close proximity to this region. Using cryo-electron tomography, we determined the first structure of the Ebola glycoprotein showing the mucin domain, a large and highly heterogeneous region that can impede antibody binding. Building upon this work, we are now working to determine the structure of Ebola envelope glycoprotein bound to antibodies currently being used as therapeutics for the virus. These structures may offer insight into the most effective mechanism to treating and/or preventing infection by this virus. Two years ago we reported the first determination of the structure of the native influenza HA trimer on the 2009 pandemic strain bound to the neutralizing antibody C179 using cryo-electron tomography. In continuation of these studies, we are exploring differences in structure between different variants of HA trimers with the goal of discovering the structural elements that will best elicit neutralizing antibodies in a universal influenza vaccine. To further understand the HIV-1 infection process, we have also undertaken a structural study of the HIV-1 core biogenesis. The HIV-1 core, which carries the viral genetic material, forms at the same time as the HIV-1 virion. In the initial stages of core formation, the HIV-1 gag polyprotein coalesces on the internal surface of a budding virion, forming a lattice structure. A series of cleavage events, which are required for proper core formation, release the capsid subunits, allowing the mature core to form. Classical models for core formation have suggested that these subunits release as monomers into the lumen of the forming virion, before spontaneously nucleating, growing into an ordered structure from the narrow end. We used cryo-electron tomography to visualize cores within virions as well as multi-core containing compartments derived from HIV-1 cell culture supernatants. These tomograms revealed a number of key features: first, native cores do not display the entirely ordered structure suggested by the classical models. Instead, the cores have numerous defects, as well as geometries unlikely to form by nucleation and growth models. More importantly, our imaging also revealed structures that appear to be partially formed, rolling cores, tethered to the inside of the membrane. These findings led us to suggest a new model for core formation: we propose that during cleavage of the capsid protein from the gag lattice, the release of stresses allows the capsid subunits to undergo a non-diffusional phase transition to the mature lattice structure. This transition in turn causes the capsid sheet to curl away from the membrane, rolling into the classical conical core shape, while leaving numerous cracks and other defects due to mismatches from the phase transition.
Showing the most recent 10 out of 33 publications
