Lineage-Specific Viral Hijacking of Non-canonical E3 Ubiquitin Ligase Cofactors in the Evolution of Vif Anti-APOBEC3 Activity

Abstract

HIV-1 encodes the accessory protein Vif, which hijacks a host Cullin-RING ubiquitin ligase (CRL) complex as well as the non-canonical cofactor CBFβ, to antagonize APOBEC3 antiviral proteins. Non-canonical cofactor recruitment to CRL complexes by viral factors, to date, has only been attributed to HIV-1 Vif. To further study this phenomenon, we employed a comparative approach combining proteomic, biochemical, structural, and virological techniques to investigate Vif complexes across the lentivirus genus, including primate (HIV-1 and simian immunodeficiency virus macaque [SIVmac]) and non-primate (FIV, BIV, and MVV) viruses. We find that CBFβ is completely dispensable for the activity of non-primate lentiviral Vif proteins. Furthermore, we find that BIV Vif requires no cofactor and that MVV Vif requires a novel cofactor, cyclophilin A (CYPA), for stable CRL complex formation and anti-APOBEC3 activity. We propose modular conservation of Vif complexes allows for potential exaptation of functions through the acquisition of non-CRL-associated host cofactors while preserving anti-APOBEC3 activity.

FIGURE 1. Proteomic analysis of lentiviral Vif proteins

(A) Percent identity and percent similarity matrix of Vif proteins used in this study. Distance tree generated from Gagpol protein sequence of viruses, with bootstrap support values. (B) Flow-chart of affinity purification – mass spectrometry (AP-MS) pipeline used to identify Vif-interacting host proteins. (C) Representative silver stained SDS-PAGE of eluates from a Vif-Strep purification in transiently transfected HEK293T cells. Asterisks indicate Vif proteins. (D) Cullin ubiquitin ring ligase (CRL) complex proteins identified in AP-MS experiments, colored by SAINT score and clustered hierarchically by correlation. (E) Immunoblot of a Strep affinity purification of Strep-tagged Vif proteins transiently expressed in HEK293T cells, probing for CRL proteins highlighted in panel (D). See also Figures S1, S2; Tables S1, S2.

FIGURE 2. Vif dependence on CBFβ is primate lentivirus-specific

(A) Schema of single-round infection assay. (B–F) Infectivity assays described in panel (A), using Vif proteins and cognate A3 proteins. Bars represent mean ± S.E. of GFP expression from virus reporter lines. Viral, A3, and CBFβ proteins are detected by immunoblot. VLPs: virus-like particles. (G) UV absorbance curves are shown for gel filtration of CUL5-RBX2 (C5R2) alone, or mixed with an excess of indicated Vif complexes (HIV VCBC = Vif-ELOB-ELOC-CBFβ, BIV VCB = Vif-ELOB-ELOC. Peaks are observed at earlier elution volumes when C5R2 is mixed with Vif complexes, indicating E3 complex formation. (H) Coomassie-blue stained SDS-PAGE of peak fractions collected from gel filtration runs shown in panel (G). Vif bands are indicated by an asterisk, and cofactor CBFβ is indicated by a circle. (I) Immunoblot of ubiquitylation reactions with either HIV-1 Vif or BIV Vif E3, using myc-tagged C-terminal domain of human A3F (myc-HsA3F CTD) or bovine A3Z3 (myc-BtA3Z3) as substrate, respectively. Ub: ubiquitin; Me-Ub: methylated ubiquitin.

FIGURE 3. CYPA is Tightly and Uniquely Associated with MVV Vif

(A) Cartoon of double affinity purification experiment. (B) LC-MS/MS mass spectrometry results from the double purification of 2xStrep-tagged Vif proteins and 3xFlag-tagged CUL5. Bars indicate peptide percent coverage of proteins identified in eluates after second purification step. (C) Immunoblot of input lysates, first and second purification eluates used for MS analysis in panel (B). (D) Heatmap of AP-MS data for CYPA in HEK293T and Jurkat T-cell lines. Color indicates SAINT score. (E) Re-probing of immunoblot of Strep affinity purification shown in Figure 1F; ELOC is shown as a control for CRL complex interaction. (F) Strep purification of MVV Vif from transient transfection of ovine FLK cells. CYPA, ELOC, and Vif are detected by immunoblot.

FIGURE 4. CYPA is a Component of the MVV Vif-Hijacked CRL Complex

(A) Co-purification testing in vivo interaction between CYPA and either wild-type or mutant MVV Vif. CYPA-Flag and various MVV Vif-Strep constructs are co-transfected, followed by a Flag immunoprecipation. Co-purification of MVV Vif constructs is assayed by immunoblot. (B) Multiple sequence alignment of MVV, CAEV, and the other Vif proteins used in this study (BIV, FIV, SIVmac, HIV-1) referenced to the first 30 amino acids of MVV Vif. Residues P21 and P24 are highlighted, as well as the region used for CYPA-binding assay in panel (D). Residues are colored by percent identity. (C) Two-dimensional ¹⁵N-¹H chemical shift mapping of CYPA in presence of MVV Vif^17–26 peptide. The resonances of Y48, R55, I56, I57, F60, M61, C62, Q63, G65, G72, L98, S99. A101, Q111, F112, E120, W121 and K125 (orange sticks) shift, then disappear upon addition of the Vif peptide. CsA is labeled in blue. PDB: 1CWA. (D) Column 1: Representative example of a CYPA residue (G135) that is not affected by the presence of the MVV Vif peptide. Column 2: By comparison, R55 and S99 undergo significant chemical shift and intensity reduction. The bars are scaled to the intensity of the HSQC peak at the corresponding Vif concentration. (E) Affinity purification of MVV Vif in presence of a titration of CsA. Co-purification of endogenous Vif interactors is assayed by immunoblot. BC-box and proline mutants are used as controls for ELOC and CYPA binding, respectively. E: Ethanol. (F) UV absorbance curves are shown for gel filtration of CUL5-RBX2 (C5R2) alone, or mixed with an excess of indicated Vif complexes (MVV VCBC = Vif-ELOB- ELOC-CYPA). Peaks are observed at earlier elution volumes when C5R2 is mixed with Vif complexes, indicating E3 complex formation. (G) Coomassie-blue stained SDS-PAGE of peak fractions collected from gel filtration runs shown in panel (F). (H) Immunoblot of methyl-ubiquitylation reactions with MVV E3 and either myc-tagged human A3H (myc-HsA3H) or ovine A3-Z3 (myc-OaA3-Z3). (I) Pair distance distribution function, P(r), calculated from SAXS intensity data. (J) Molecular envelopes of HIV-1 Vif^1–174-CBFβ-CRL5 (left) and MVV Vif-CYPA-CRL5 (right) calculated from P(r). An HIV-1 E3 model was superimposed into both envelopes. See also Figures S3, S5; Table S3.

FIGURE 5. MVV Vif Mutants Deficient in CYPA-Binding are Deficient in A3 Antagonism and Cannot Promote MVV Infectivity in situ

(A) Co-transfection of HA-tagged ovine A3Z2Z3 (OaA3Z2Z3-HA) and either wild-type or proline mutant MVV Vif. A3 stability in the presence of Vif is assayed by immunoblot. (B) Co-affinity purification between OaA3Z2Z3 and MVV Vif constructs that were deficient in OaA3Z2Z3 destabilization in panel (A). Interaction between A3 and Vif proteins is assayed by immunoblotting. (C) MVV spreading assay in ovine primary macrophage cells. Lysates were harvested at various time points post-infection, and virus genome copies were quantified using TaqMan-based real-time PCR assay, mean ± S.E. (n=3). (D) Hypermutation assay of MVV strain KV1772. MVV with either wild-type or mutant vif were subjected to a single-cycle infection assay in primary sheep choroid plexus (SCP) cells, and produced viruses were then used to infect cells, pro-viruses cloned, and assayed for A3-mediated G-to-A mutations. Wild-type, P24A: n=20; P21A: n=16; P21A/P24A: n=17; SLQ::AAA: n=19; Δvif: n=10. Significance values were determined by a one-sided Wilcoxon Rank-Sum Test compared to wild-type; no annotated p-value indicates p-value > 0.05. (E) Tri-nucleotide context of G-to-A mutations measured in panel (C). Other refers to any GNN tri-nucleotide other than GGA or GAA. See also Figure S6.

FIGURE 6. CYPA is Required for MVV Vif A3 Degradation Activity

(A) Comparison of Vif A3 degradation activity in monoclonal CYPA knockdown versus control cells in the presence or absence of CsA. Cells were transiently transfected with HA-tagged ovine A3Z2Z3 (OaA3-Z2Z3) and either wild-type or BC-box mutant (SLQ::AAA) Strep-tagged MVV Vif, then treated 6 hours later with either ethanol (E) or 2 μM CsA overnight. Bars represent HA immunoreactivity normalized first by GAPDH loading control, then to no Vif control for each cell line; mean ± S.E. (n=3). Proteins are detected by immunoblotting. (B) CYPA immunoblot in Jurkat E6-1 CYPA^+/+ “parental” line and derived E6-1 CYPA^−/− knockout (KO) line. (C) Top: Jurkat CYPA^−/− KO cells are transiently transfected with HA-tagged OaA3Z2Z3, Strep-tagged MVV Vif, and Flag-tagged CYPA. eGFP is used as transfection control. Bottom: identical experiment performed in E6-1 CYPA^+/+ control line. (D) Top: Jurkat CYPA^−/− KO cells are transiently transfected with HA-tagged human A3G (HsA3G), Strep-tagged HIV-1 Vif, and Flag-tagged CYPA. eGFP is used as transfection control. Bottom: identical experiment performed in E6-1 CYPA^+/+ control line. (E) MVV Vif activity rescue assay using mutants of CYPA. Strep-tagged MVV Vif, HA-tagged OaA3Z2Z3, and various Flag-tagged CYPA constructs are transfected into Jurkat CYPA^−/− KO line, and Vif activity assessed through A3 stability. See also Figures S4, S7.

FIGURE 7. Modular Conservation of CRL Hijacking and Non-Canonical Cofactor Recruitment by Vif

The host CRL complex hijacked by Vif represents a conserved host-pathogen interaction module. Vif proteins recruit non-canonical host cofactors in a lineage-specific manner within the lentivirus genus.