A lysine ring in HIV capsid pores coordinates IP6 to drive mature capsid assembly

Abstract

The HIV capsid self-assembles a protective conical shell that simultaneously prevents host sensing whilst permitting the import of nucleotides to drive DNA synthesis. This is accomplished through the construction of dynamic, highly charged pores at the centre of each capsid multimer. The clustering of charges required for dNTP import is strongly destabilising and it is proposed that HIV uses the metabolite IP6 to coordinate the pore during assembly. Here we have investigated the role of inositol phosphates in coordinating a ring of positively charged lysine residues (K25) that forms at the base of the capsid pore. We show that whilst IP5, which can functionally replace IP6, engages an arginine ring (R18) at the top of the pore, the lysine ring simultaneously binds a second IP5 molecule. Dose dependent removal of K25 from the pore severely inhibits HIV infection and concomitantly prevents DNA synthesis. Cryo-tomography reveals that K25A virions have a severe assembly defect that inhibits the formation of mature capsid cones. Monitoring both the kinetics and morphology of capsids assembled in vitro reveals that while mutation K25A can still form tubes, the ability of IP6 to drive assembly of capsid cones has been lost. Finally, in single molecule TIRF microscopy experiments, capsid lattices in permeabilised K25 mutant virions are rapidly lost and cannot be stabilised by IP6. These results suggest that the coordination of IP6 by a second charged ring in mature hexamers drives the assembly of conical capsids capable of reverse transcription and infection. HIV protects its RNA genome while copying it into DNA by carrying out reverse transcription inside its capsid. This is accomplished by importing nucleotides through highly charged pores at the centre of each capsid multimer. These pores contain two rings of positively charged residues–R18 and K25 –but assembling capsids with these features is challenging because they are intrinsically destabilising. Here we show that the metabolite IP6 coordinates both residues within the pore to drive the assembly of stable capsids capable of nucleotide import. R18 or K25 mutants lose infectivity and the ability to synthesise DNA but have differing assembly phenotypes. Mutant K25A is unable to undergo efficient capsid assembly, while replacing K25 with a neutral polar residue partially restores assembly but not infectivity. We propose that IP6-driven assembly is conserved by HIV not because it is the only way to build a capsid, but because it allows the construction of a capsid with a charged pore that can import nucleotides.