Two structural switches in HIV-1 capsid regulate capsid curvature and host factor binding

Abstract The mature HIV-1 capsid protects the viral genome and interacts with host proteins to travel from the cell periphery into the nucleus. To achieve this, the capsid protein, CA, constructs conical capsids from a lattice of hexamers and pentamers, and engages in and then relinquishes multiple interactions with cellular proteins in an orchestrated fashion. Cellular host factors including Nup153, CPSF6 and Sec24C engage the same pocket within CA hexamers. How CA assembles pentamers and hexamers of different curvatures, how CA oligomerization states or curvature might modulate host-protein interactions, and how binding of multiple co-factors to a single site is coordinated, all remain to be elucidated. Here, we have resolved the structure of the mature HIV-1 CA pentamer and hexamer from conical CA-IP 6 polyhedra to high resolution. We have determined structures of hexamers in the context of multiple lattice curvatures and number of pentamer contacts. Comparison of these structures, bound or not to host protein peptides, revealed two structural switches within HIV-1 CA that modulate peptide binding according to CA lattice curvature and whether CA is hexameric or pentameric. These observations suggest that the conical HIV-1 capsid has different host-protein binding properties at different positions on its surface, which may facilitate cell entry and represent an evolutionary advantage of conical morphology. Significance statement HIV-1 particles contain a characteristic, conical capsid that shields the genome from the cellular immune system and recruits cellular proteins to direct the capsid to the nucleus. The cone forms from hexamers of CA protein, and twelve pentamers that accommodate curvature. We obtained detailed 3D models of pentamers and hexamers at positions on capsid surfaces with different curvatures. We find two places in CA that switch conformation according to the local capsid curvature and whether CA is in a pentamer or hexamer. We also obtained models of CA bound to peptides from cellular proteins. The data show how switches in CA help it form a cone shape, and interact differently with cellular proteins at different positions on the cone surface.