APOBEC3 multimerization correlates with HIV-1 packaging and restriction activity in living cells

Abstract

APOBEC3G belongs to a family of DNA cytosine deaminases that are involved in the restriction of a broad number of retroviruses including human immunodeficiency virus type 1 (HIV-1). Prior studies have identified two distinct mechanistic steps in Vif-deficient HIV-1 restriction: packaging into virions and deaminating viral cDNA. APOBEC3A, for example, although highly active, is not packaged and is therefore not restrictive. APOBEC3G, on the other hand, although having weaker enzymatic activity, is packaged into virions and is strongly restrictive. Although a number of studies have described the propensity for APOBEC3 oligomerization, its relevance to HIV-1 restriction remains unclear. Here, we address this problem by examining APOBEC3 oligomerization in living cells using molecular brightness analysis. We find that APOBEC3G forms high-order multimers as a function of protein concentration. In contrast, APOBEC3A, APOBEC3C and APOBEC2 are monomers at all tested concentrations. Among other members of the APOBEC3 family, we show that the multimerization propensities of APOBEC3B, APOBEC3D, APOBEC3F and APOBEC3H (haplotype II) bear more resemblance to APOBEC3G than to APOBEC3A/3C/2. Prior studies have shown that all of these multimerizing APOBEC3 proteins, but not the monomeric family members, have the capacity to package into HIV-1 particles and restrict viral infectivity. This correlation between oligomerization and restriction is further evidenced by two different APOBEC3G mutants, which are each compromised for multimerization, packaging and HIV-1 restriction. Overall, our results imply that multimerization of APOBEC3 proteins may be related to the packaging mechanism and ultimately to virus restriction.

Keywords: APOBEC3G; brightness; fluorescence fluctuation spectroscopy; mobility; molecular mass complex.

Fig. 1

Brightness and Stoichiometry. (A) A fluorescent protein with a brightness λ is attached to the protein of interest. The brightness of the labeled monomeric protein is identical to that of the fluorescent protein alone. Dimerization of the monomers leads to a doubling of the brightness, because the dimer carries two fluorescent proteins. (B) Brightness of monomeric EGFP (squares) and the dimeric EGFP₂ (circles) in the cytoplasm of U2OS cells. The control experiment demonstrates that the brightness of the label EGFP is concentration independent and that brightness doubling is observed for EGFP₂. Each data point represents the brightness measured in a different cell expressing either EGFP or EGFP₂. The average brightness of EGFP is 1 ± 0.06, while the average normalized brightness EGFP₂ is 1.90 ± 0.10. (C) Brightness titration experiment. Schematic representation of the brightness for a monomer-dimer transition of a labeled protein as a function of its concentration. At low concentrations the equilibrium is shifted towards monomers, which results in a brightness close to 1. High concentrations favor the dimer, which is characterized by a brightness of 2. At intermediate concentrations the brightness is between 1 and 2, which indicates the presence of a mixture of monomers and dimers.

Fig. 2

Multimerization of A3G-EGFP (A) Brightness titration experiments of A3G-EGFP in U2OS cells. Five independent experiments are displayed to illustrate the repeatability. The solid line is the average brightness curve for A3G determined by interpolation of the experimental data. (B) Brightness titration for A3G-EGFP in three different cell lines. (C) Brightness titration experiments with A3G-NTD-EGFP (●) and A3G-CTD-EGFP (▲). The solid line depicts the average brightness of full-length A3G-EGFP for comparison.

Fig. 3

Brightness titration experiments in U2OS cells of (A) A2-EGFP (●), (B) A3A-EGFP (●) and A3C-EGFP (▲), (C) A3B-EGFP and (D) A3D-EGFP (●), A3F-EGFP (◆) and A3H-EGFP (▲). The solid line is the averaged A3G brightness and the dash line indicates the brightness of a monomer.

Fig. 4

A3G-EGFP mutants C97A and E259Q (A) Brightness titration of A3G-E259Q-EGFP (▲) and A3G-C97A-EGFP (●) in U2OS cells. The solid line is the average A3G-EGFP brightness and the dashed line marks the monomeric brightness. (B) The restrictive activity of A3G and mutants as assessed by a single cycle infectivity assay. Viral infectivity was measured as the percentage of fluorescent-positive CEM-GFP reporter cells normalized to results for 0 ng of A3G construct. Error bars indicate the standard deviation from duplicate experiments. (C) Representative immunoblots from one of five experiments. A3G-3xHA expression in producer 293T cells and packaging in viral particles were detected by anti-HA antibody. Anti-tubulin and anti-p24 served as loading controls in cell lysates and viral particles respectively.