CROI 2020 Abstract eBook

Abstract eBook

Oral Abstracts

Adnectin that binds to CD4 and inhibits downstream actions of gp160. Studies were performed to help elucidate the binding site of the Adnectin on CD4 and understand the mechanism of inhibition. Methods: Hydrogen-deuterium exchange mass spectrometry (HDX) was used to examine comparative deuteration rates of amide backbone protons of CD4, either in the absence or presence of saturating amounts of Adnectin. In addition, crystal structures of CD4 bound to both the Adnectin and a Fab subunit of ibalizumab were solved at a 3.7Å resolution. Cryo-EM studies of Adnectin bound to soluble CD4 were also generated. Finally, mutagenic analyses on CD4 were performed to confirm and extend these findings. Results: Using HDX, CD4 peptides at the N-terminus of D2 and in D3 showed differential rates of deuteration (both enhanced and slowed) in the presence of the Adnectin that mapped predominantly to the D2-D3 interface. The structure of the ibalizumab Fab/CD4 D1-D4/Adnectin complex revealed an extensive interface between the Adnectin and residues on CD4 domains D2-D4 that stabilize a novel T-shaped CD4 conformation. A cryo-EMmap of the gp140/CD4/ combinectin complex clearly shows the bent conformation for CD4 while bound to gp140. Mutagenic analyses on CD4 confirmed that amino acid F202 forms a key interaction with the Adnectin. In addition, amino acid L151 was shown to be a critical determinant of the specificity for binding to human CD4 protein over related primate CD4 molecules. Mutation of L151 to R (the residue present in cynomolgus monkey CD4) abrogated Adnectin binding to human CD4, while the reverse mutation (R151L) restored binding to cynomolgus monkey CD4. Conclusion: The significant conformational change of CD4 upon Adnectin binding brings the D1 domain of CD4 in proximity to the host cell membrane surface and provides a potential explanation for the ability of the CD4-bound Adnectin to inhibit HIV-1 infection. In addition, mutations of D2-D3-interface residues, specifically F202 and L151, dramatically impacted Adnectin binding to human and primate CD4, providing a rationale for the observed species specificity of the Adnectin. 21LB SERINC3/5 PERTURB HIV MEMBRANE FUSION POST-HEMIFUSION AT FUSION-PORE DILATION STEPS Amanda E. Ward 1 , Volker Kiessling 1 , Judith M. White 1 , Owen Pornillos 1 , Barbie K. Ganser-Pornillos 1 , Lukas K. Tamm 1 1 University of Virginia, Charlottesville, VA, USA Background: Serinc3 and Serinc5 are recently described host restriction factors that in the absence of Nef, can block HIV infection by incorporating into budding viral particles and decreasing their ability to infect subsequent cells. Serincs are thought to block the very earliest stages of infection, membrane fusion and cell entry, by an incompletely understood mechanism. Methods: We used giant plasma membrane vesicles (blebs) as model target membranes to study “wildtype” and Serinc-disrupted HIV membrane fusion at a single-particle level with cryoElectron Tomography and Total Internal Reflection Fluorescence (TIRF) microscopy. Results: Using fluorescent reporters of membrane and content mixing, we observed that Serinc3 and Serinc5 do not cause a defect in mixing of the outer lipid leaflets (hemifusion), but a pronounced defect in fusion pore opening. Additionally, cryo-electron tomography of HIV pseudoviruses mixed with blebs showed rearrangements of viral and target membranes and proteins at multiple intermediates steps of HIV membrane fusion. We found that Serinc3 and Serinc5 increased the number of hemifusion and early fusion product events and that many of the fusion products are cinched between former virus and bleb. Conclusion: These results suggest that Serinc3 and Serinc5 create bottlenecks in the process of membrane fusion; a first bottleneck after hemifusion and an additional bottleneck that prevents full fusion pore dilation such that the viral capsid cannot pass into the cytosol. Understanding how Serincs disrupt HIV membrane fusion will clarify the requirements for normal HIV membrane fusion and potentially identify new viral weaknesses that could become drug targets. 22 CRISPR-INDUCED MUTAGENESIS POINTS TOWARD A ROLE OF TRN-SR2 IN HIV NUCLEAR IMPORT Frauke Christ 1 , Julie Janssens 1 , Flore De Wit 1 , Jolien Blokken 1 , Youlia Lampi 1 ,

the importance of TRN-SR2 for HIV nuclear import is generally accepted, the detailed mechanism and role of TRN-SR2 remains under debate. According to one model the direct interaction of TRN-SR2 with HIV integrase drives nuclear import of the pre-integration complex (PIC), alternatively TRN-SR2 may play an indirect role linked to uncoating of the PIC and the protein CPSF6. Methods: We have designed CRISPR-Cas9 guide RNAs targeting exon 2 and 8 of TNPO3 in HeLaP4 cells. After selection of clones with reduced TRN-SR2 expression on both mRNA (QPCR) and protein expression levels (western blotting), a detailed analysis of HIV replication and PIC nuclear import was performed. Results: CRISPR-Cas9 induced DNA breaks in TNPO3 using guide 2 and 8 failed to generate complete knockout clones but instead allowed for selection of 2 HeLaP4 clones with a single allelic KO, resulting in 2-fold reduced TRN-SR2 levels (clone #20 and #25). Nevertheless, HIV single round and multiple round replication was severely hampered in clone #20 and #25. Interestingly genome sequencing of TNPO3 revealed that the remaining allele showed small in-frame deletions resulting in deletion of Aa (V103 and 373LHAL376). We then analyzed the PIC nuclear import in the respective cell lines by QPCR and fluorescent imaging of eGFP-IN labeled PICs. Both techniques evidenced a strong defect in nuclear import. Recombinant TRN-SR2 deletion mutants demonstrated an impairment of the molecular interaction with HIV-integrase. Conclusion: CRISPR-Cas9 targeting two different exons of TNPO3 failed to generate KO cell lines indicating that a full KO of TRN-SR2 might be toxic for HeLaP4. Yet, CRISPR-Cas9 unexpectedly led to mutagenesis. The resulting clones were fully viable but failed to support HIV replication. The block of replication was pinpointed to nuclear import and the corresponding recombinant mutant TRN-SR2 was impaired for interaction with HIV-IN. The presented data support the notion that TRN-SR2 is a genuine co-factor of HIV replication and interacts differently with HIV-IN than with its cellular cargoes. 23 NUCLEAR UNCOATING OF HIV-1 OCCURS NEAR SITES OF INTEGRATION Ryan C. Burdick 1 , Chenglei Li 1 , Mohamed Husen Munshi 1 , Jonathan Rawson 1 , Kunio Nagashima 2 , Wei-Shau Hu 1 , Vinay K. Pathak 1 1 National Cancer Institute, Frederick, MD, USA, 2 Leidos Biomedical Research, Inc, Frederick, MD, USA Background: A critical step in HIV-1 replication is the disassembly (uncoating) of the viral core. Remarkably, the timing and intracellular location of HIV-1 uncoating remain unknown. Studies of HIV-1 uncoating have been hampered by an inability to accurately quantify capsid protein (CA) loss from the viral complexes and by an inability to identify rare infectious viral complexes (~1/50) in infected cells. Methods: We developed methods to label CA with GFP (GFP-CA) in infectious viral complexes and to identify transcriptionally-active proviruses in live-cell imaging assays. We analyzed the dynamics of viral complex association with nuclear envelope and nuclear uncoating, and identified rare viral complexes that integrate to form transcriptionally active proviruses. Results: Using live-cell imaging, we observed >110 GFP-CA-labeled infectious viral complexes that integrated and expressed HIV-1 RNA and the gfp reporter gene. The infectious viral complexes maintained steady GFP-CA fluorescence signals for several hours after nuclear import followed by abrupt (<20 min) GFP-CA loss ~10.5 hours after infection, signifying nuclear uncoating. HIV-1 transcription sites appeared near the sites of nuclear uncoating, indicating that uncoating occurs at or very close to the site of integration. Similar GFP-CA fluorescence intensities of nuclear viral complexes and viral cores in vitro suggest that viral cores in the nucleus retain >90% of the CA and that nuclear uncoating is the major uncoating event. The nuclear GFP-CA-labeled viral complexes rapidly disassembled after treatment of the infected cells with capsid inhibitor PF74 indicating that the nuclear viral complexes retained CA hexamers. Time-of-addition assays with PF74, nevirapine, and raltegravir indicate that nuclear uncoating occurs ~3 hrs after the completion of reverse transcription and ~1 hr before integration. We probed the potential mechanism by which viral cores enter the nucleus and found that cleavage and polyadenylation specificity factor 6 (CPSF6), a host nuclear protein that binds to CA, influences the intracellular location of uncoating and facilitates the nuclear import of intact or nearly intact viral cores. Conclusion: Intact or nearly intact viral cores of infectious viral complexes that retain >90% of their CA enter the nucleus and uncoat near their genomic integration sites just before integration.

Oral Abstracts

Irena Zurnic 1 , Rik Gijsbers 1 , Zeger Debyser 1 1 Katholieke University Leuven, Leuven, Belgium

Background: In order to infect non-dividing cells, HIV needs to cross the nuclear envelop. In 2010 we reported the identification of the importin TRN-SR2 (TNPO3) as the determining host factor for nuclear import. While

8

CROI 2020