HIV-1 Vif N-terminal motif is required for recruitment of Cul5 to suppress APOBEC3

Abstract

Background: HIV-1 Vif promotes the degradation of host anti-retroviral factor family, APOBEC3 proteins via the recruitment of a multi-subunit E3 ubiquitin ligase complex. The complex is composed of a scaffold protein, Cullin 5 (Cul5), RING-box protein (Rbx), a SOCS box binding protein complex, Elongins B/C (Elo B/C), as well as newly identified host co-factor, core binding factor beta (CBF-β). Cul5 has previously been shown to bind amino acids within an HCCH domain as well as a PPLP motif at the C-terminus of Vif; however, it is unclear whether Cul5 binding requires additional regions of the Vif polypeptide.

Results: Here, we provide evidence that an amino terminal region of full length Vif is necessary for the Vif-Cul5 interaction. Single alanine replacement of select amino acids spanning residues 25-30 (25VXHXMY30) reduced the ability for Vif to bind Cul5, but not CBF-β or Elo B/C in pull-down experiments. In addition, recombinant Vif mutants had a reduced binding affinity for Cul5 compared to wild-type as measured by isothermal titration calorimetry. N-terminal mutants that demonstrated reduced Cul5 binding were also unable to degrade APOBEC3G as well as APOBEC3F and were unable to restore HIV infectivity, in the presence of APOBEC3G. Although the Vif N-terminal amino acids were necessary for Cul5 interaction, the mutation of each residue to alanine induced a change in the secondary structure of the Vif-CBF-β-Elo B/C complex as suggested by results from circular dichroism spectroscopy and size-exclusion chromatography experiments. Surprisingly, the replacement of His108 to alanine also contributed to the Vif structure. Thus, it is unclear whether the amino acids contribute to a direct interaction with Cul5 or whether the amino acids are responsible for the structural organization of the Vif protein that promotes Cul5 binding.

Conclusions: Taken together, we propose a novel Vif N-terminal motif that is responsible for Vif recruitment of Cul5. Motifs in Vif that are absent from cellular proteins represent attractive targets for future HIV pharmaceutical design.

Figure 1

Vif N-terminal amino acids are responsible for Cul5 interaction, in vitro. Vif wild-type and single alanine mutants were co-expressed with N-terminal Cul5, N-terminal 6X-His-CBF-β (residues 1–140), and Elo B/C. Next, the complex was pulled down using nickel affinity purification. A) While Vif wild-type and H28A mutant pull down Cul5 efficiently, H27A, M29A and Y30A mutants are unable to bind Cul5 efficiently. B) Quantitative measurement of the Cul5 band intensity was performed indicating the relative amount of Cul5 bound to Vif wild-type and mutant protein complexes. FL – full length, N-terminal – amino-terminus, 6X-His – 6X histidine tag.

Figure 2

Several Vif N-terminal mutants are unable to degrade APOBEC3G and APOBEC3F. Vif wild-type and mutant proteins along with A3G or A3F were overexpressed in HEK 293 T cells. Two days post-transfection, cells were lysed and proteins were separated by SDS-PAGE and visualized by western blotting. A) APOBEC3F-V5 and B) APOBEC3G-V5 are efficiently degraded by wild-type Vif, but not by several Vif N-terminal mutants in HEK 293 T cells.

Figure 3

Select Vif N-terminal mutants have a reduced ability to bind Cul5 in mammalian cells. HA-tagged Vif wild-type and mutant proteins along with CBF-β were overexpressed in HEK 293 T cells. Two days post-transfection, cells were lysed and cleared lysate was mixed with anti-HA matrix affinity beads for 4-8 hrs. Incubated beads were washed several times followed by elution of bound proteins. Select Vif N-terminal mutants (V25A, H27A, M29A, and Y30A) that do not efficiently degrade A3G and A3F have a reduced ability to co-precipitate Cul5; however, CBF-β and Elo B/C can still bind Vif.

Figure 4

Vif N-terminal mutants localize to cytoplasm, but are inefficient at restoring HIV infectivity. HA-tagged Vif wild-type and mutant proteins along with CBF-β were overexpressed in HEK 293 T cells. Two days post-transfection, cells were lysed and cleared lysate was mixed with anti-HA matrix affinity beads for 4-8 hrs. Incubated beads were washed several times followed by elution of bound proteins. A) Select Vif N-terminal mutants that do not efficiently degrade A3G and A3F have a reduced ability to co-precipitate Cul5; however, CBF-β and Elo B/C can still bind Vif. B) Plasmids (Vif-YFP 2 ug and CBF-β 0.5 ug) were transfected into 293 T cells using Lipofectamine 2000 (Invitrogen), according to the manufacturer’s protocol. Cells were visualized at 25°C using a Zeiss LSM510-Meta confocal imaging system. Imaging demonstrates that the Vif double mutant V25/H27A and single mutant H108A localize to the cytoplasm of the cell similar to wild-type. C) Vif wild-type and mutant containing virus were produced and used to infect MAGI cells. Infected cells were stained using X-gal. The histogram demonstrates that Cul5-binding deficient Vif mutants were inefficient at restoring HIV infectivity in the presence of A3G. Error bars represent the standard error from triplicate experiments. Capsid p24 levels are shown in the western blot.

Figure 5

Recombinant Vif mutants have a lower Cul5 binding affinity compared to wild-type Vif. Isothermal titration calorimetric analyses of the interaction between Cul5 and Vif wild-type and mutant complexes. A) and B) Representative ITC isotherm and table for Vif wild-type and mutants demonstrating a lower affinity between mutant Vif and Cul5 compared with Vif wild-type. C) Vif complex samples run on SDS-PAGE gel and visualized by coomassie stain and western blot. FL – full length.

Figure 6

CD spectroscopy analysis reveals that Vif mutant complexes are structurally different from wild-type. Purified Vif complexes including 6X-His-CBF-β (residues 1–182) and Elo B/C were purified by nickel affinity and size exclusion chromatography. Each complex was analyzed by circular dichroism spectroscopy. A) CD spectra for Vif wild-type and mutant complexes showed a distinction in the minima at 208 and 222 for mutant complexes that do not bind Cul5, suggesting these mutants have more alpha helical structure. B) Spectra analysis reveals differences between wild-type and mutant complex secondary structure and confirms that the mutants that do not bind Cul5 have a higher percentage of alpha helical structures; however, the percentage of beta-sheet structures is reduced.

Figure 7

Model for Vif-E3 ligase complex and schematic diagram of Vif residues required for E3 ligase recruitment. A) Model for Vif-mediated ubiquitination and degradation of A3 molecules via recruitment and assembly of host E3 ligase components. B) Updated schematic diagram of Vif binding motifs, including the Vif N-terminal motif required for efficient Cul5 binding.