Progress in FY2014 was in the following areas: (1) HIV-1 CAPSID ASSEMBLIES. We published the first full solid state NMR chemical shift assignments for the HIV-1 capsid protein (CA) in assembled form, as well as the results of measurements of site-specific dynamics and quantitative conformational constraints for HIV-1 CA in tubular assemblies. These data indicate that the protein structure in CA assemblies is very similar to the structure in soluble CA dimers and monomers (as expected), but that minor conformational changes are evident in specific segments, primarily the """"""""linker"""""""" between N-terminal and C-terminal domains (NTD and CTD) and short loops that connect helical segments. These data also indicate that intermolecular contacts and interfaces are well ordered, especially at the critical NTD-NTD interfaces in hexamer rings, the critical CTD-CTD dimerization interfaces, and local 3-fold symmetry axes involving CTD-CTD interactions. Previous work by other groups had suggested that structural variations in dimerization interfaces and around 3-fold axes might account for variable capsid lattice curvature, inconsistent with our solid state NMR measurements. From our data, it is apparent that specific regions of CA are dynamically disordered in CA assemblies, especially the linker segment and certain inter-helical loops. Our hypothesis is that these dynamically disordered segments accomodate the lattice curvature. In work not yet published, we have examined the R18L mutant of CA, which forms either spherical assemblies or planar 2D lattices under certain assembly conditions. Surprisingly, solid state NMR spectra of planar and spherical R18L-CA assemblies are nearly identical to those of tubular wild-type CA assemblies. No new signals are detected, indicating that CA is not more rigid or more structurally ordered in a planar lattice (where in principle all CA molecules are structurally equivalent) than in spherical or tubular assemblies. This supports our hypothesis that variable lattice curvature is accomodated by dynamically disordered segments. We have also begun quantitative solid state NMR measurements of interatomic distances in the CTD-CTD dimerization interface in the context of tubular CA assemblies, with the goal of distinguishing among several candidate structures that have been proposed from x-ray crystallography and solution NMR of soluble CTD dimers. Results so far favor the dimerization interface structure seen in solution NMR. Eventually, these measurements will lead to a detailed understanding of CA dimerization in the context of the CA lattice, an issue that has not been resolved in prior studies by electron microscopy and crystallography.
