Cellular host factors involved in HIV-1 budding

In my dissertation research, I studied the involvement of different cellularpathways in the release of HIV-1 from the host cell, with an emphasis on the ESCRT pathway. The ESCRT pathway comprises a series of protein complexes, including ESCRT-I and ESCRT-III, both of which are required for HIV-1 budding.Chapter 2 focuses on the structural and functional characterization of the ESCRTI complex. To enable biochemical and structural studies, I developed a method for producing pure recombinant full-length human ESCRT-I complex. SAXS studies showed that the full-length human ESCRT-I has an elongated shape. I discovered a potential direct interaction between recombinant ESCRT-I and a fragment of CHMP2A (ESCRTIII). The adaptor protein ALG-2 enhances the ESCRT-I interaction with the ESCRT associated protein ALIX in a calcium-dependent manner. ALG-2 depletion moderately decreased HIV-1 release. However, the effect was not strong, implying that ALG-2 may play a regulatory role in virus release. Chapter 3 describes the interaction of an ESCRT-III related protein (CHMP5) with LIP5 and its importance for HIV-1 budding. LIP5 activates VPS4 oligomerization and ATPase activity, and this enzymatic activity provides the driving force for ESCRTIIIdisassembly and for multiple cycles of virus release. The LIP5 C-terminal VSLdomain binds to VPS4 and the N-terminal (MIT)2 domain binds to many ESCRT-IIIproteins. A flexible linker connects the two LIP5 domains. CHMP5 binds to the LIP5 (MIT)2 domain with unusually high affinity using a novel MIT-MIM mode ofinteraction. Initial evidence for autoinhibition was obtained from fluorescence anisotropy binding experiments using LIP5 and CHMP5. The region of LIP5 autoinhibition was mapped to the upstream region in the LIP5 linker. I propose a model for LIP5 autoinhibition in which the flexible linker of LIP5 binds to its own (MIT)2 domain to compete for CHMP5 binding. Appendix A describes work that suggests that no single ubiquitin-conjugating enzyme is critical for HIV-1 budding. Experiments in Appendix B support the idea that the interaction between the viral protein Gag and the host protein Annexin 2 is indirect.Finally, Appendix C provides a summary of my contributions to the study of the ESCRTIII requirements in HIV-1 budding.