Summary
Human myxovirus resistance 2 (MX2) can potently restrict HIV-1 and herpesviruses at a post-entry step by a process that requires MX2 interaction with the capsids of these viruses. The involvement of other host cell factors in this process, however, remains poorly understood. Here, we mapped the proximity interactome of MX2 revealing strong enrichment of phenylalanine-glycine (FG)-rich proteins related to the nuclear pore complex as well as proteins that are part of cytoplasmic ribonucleoprotein granules. MX2 interacted with these proteins to form multiprotein cytoplasmic biomolecular condensates that were essential for its anti-HIV-1 and -herpes simplex virus-1 (HSV-1) activity. MX2 condensate formation required the disordered N-terminal region of MX2 and its dimerization. Incoming HIV-1 and HSV-1 capsids associated with MX2 at these dynamic cytoplasmic biomolecular condensates. Our results demonstrate that MX2 forms cytoplasmic condensates that act as nuclear pore decoys, which trap capsids and induce premature viral genome release, and thereby interfere with nuclear targeting of HIV-1 and HSV-1.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
↵* shared senior authorship
The previous preprint has now been peer reviewed and we provide an accordingly revised updated version. In summary the following changes were made: 1. We demonstrate that endogenous, type I Interferon induced MX2 forms biomolecular condensates that contain endogenous FG-Nups. 2. We show that knock down of SAMD4A, TNPO1, and NUP35 significantly reduces HIV-1 and HSV-1 restriction by endogenous, type I Interferon induced, MX2. 3. We show that HIV-1 P90A capsid mutant virus, which is resistant to MX2-mediated restriction, was significantly less associated with the MX2 condensates. 4. We demonstrate the presence of HSV-1 genomes at the surface of and inside MX2 condensates. In addition, we confirmed and quantified trapping of HIV-1 capsids and pre-integration complexes inside MX2 condensates. 5. We now only refer to the dScope tool to predict regions in MX2 NTD with phase separation properties and avoided the use of prion-like domain. In the supplementary materials, there is now a link to Figshare, where life cell imaging videos can be accessed.