Involvement of clathrin-mediated endocytosis in human immunodeficiency virus type 1 entry

Abstract

Productive entry of human immunodeficiency virus (HIV) is believed to occur by direct fusion at the plasma membrane. Endocytic uptake of HIV particles has been observed in several studies but is considered to be nonproductive, leading to virus degradation in the lysosome. We show here that endocytosis contributes significantly to productive HIV entry in HeLa cells by using trans dominant-negative mutants of dynamin and Eps15. Inducible expression of a dominant-negative mutant of dynamin in a CD4-positive HeLa cell line reduced HIV infection by 40 to 80%. This effect was independent of the infectious dose and was observed for three different isolates. Analysis of reverse transcription products by real-time PCR and of virus entry by delivery of a virion-associated Vpr-beta-lactamase fusion protein revealed a similar reduction, indicating that the block occurred at the entry stage. A strong reduction of HIV entry was also observed upon transient transfection of a different trans dominant-negative variant of dynamin, and this reduction correlated with the relative inhibition of transferrin endocytosis. Expression of a dominant-negative variant of Eps15, which is specific for clathrin-dependent endocytosis, reduced HIV entry in HeLa cells by ca 95%, confirming the role of endocytosis for productive infection. In contrast, no effect was observed for a dominant-negative variant of caveolin. We conclude that dynamin-dependent, clathrin-mediated endocytosis can lead to productive entry of HIV in HeLa cells, suggesting this pathway as an alternative route of virus entry.

FIG.1.

Characterization of the CD4-positive cell line HeLa4D9 stably expressing dyn^ts under the control of a tet-responsive promoter. (A) Fluorescence microscopy analysis of dyn^ts expression (left panel) and endocytosis of transferrin (right panel). HeLa4D9 cells were cultivated for 3 days at 32°C in the presence (c and c′) or absence (a, a′, b, and b′) of tetracycline. Subsequently, cells were shifted to the nonpermissive temperature of 37°C for 1.5 h (a, a′, c, and c′) or were shifted to the permissive temperature of 30°C for 1 h after the 37°C treatment (b and b′). Twenty minutes before fixation, fluorescently labeled transferrin was added to determine inhibition of clathrin-dependent endocytosis. Cells were fixed and stained for the HA epitope tag on dyn^ts. Note that transferrin localization was slightly different when endocytosis assays were performed at 30°C, which is likely to be due to slower uptake kinetics at the lower temperature. (B) FACS analysis of surface expression of CD4 (left panels) and CXCR4 (right panels) on HeLa4D9 cells. Cells were cultivated for 3 days in the presence (heavy gray line) or absence (black line) of tetracycline and were subsequently shifted to the nonpermissive temperature of 37°C for 1.5 h (a and a′) or were kept at the permissive temperature of 30°C (b and b′). Cells were fixed and stained with FITC-labeled antibody against CD4 (a and b) or PE-labeled antibody against CXCR4 (a′ and b′) and analyzed by flow cytometry. Negative controls (gray shading) are unstained cells or cells stained with a PE-labeled isotype control antiserum, respectively. The numbers indicate the percentage of positive cells.

FIG. 2.

Single round entry assay. (A) Schematic depiction of the entry assay. HeLa4D9 cells were cultivated for 3 days at 32°C in the absence (to induce dyn^ts expression) or presence of tetracycline. Subsequently, cells were shifted to the nonpermissive temperature of 37°C for 1.5 h, and 10 mM ammonium chloride (NH₄Cl) was added in some cases to study the influence of endosomal acidification. Cells were infected with HIV-1 particles or viral pseudotypes at the indicated MOIs and were further cultivated at 37°C for additional 4 to 7 h to allow HIV-1 entry. Subsequently, the medium was changed, azidothymidine was added at a concentration of 10 μM to prevent infection at later time points, and tetracycline was added to shut off dyn^ts expression. HIV-1 infection was evaluated after 2 days by staining with an antibody to the viral CA protein and detection with a secondary antibody conjugated to β-galactosidase (indicated as dark cells). The total number of CA-positive cells per well was counted. Punctate patterns on cells indicate normal endocytosis, gray areas indicate inhibition of endocytosis. (B) Representative single-round entry assays. HeLa4D9 cells were infected with HIV-1_NL4-3 at an MOI of 0.06 with HIV-1_NL4-3 pseudotyped with VSV-G or with the HIV-1 O-type isolate MVP8161. Ammonium chloride (NH₄Cl) was added 1.5 h prior to infection as indicated. The total numbers of HIV-1-infected cells are shown for HeLa4D9 cells cultivated in the absence (░⃞) or presence (▪) of tetracycline. Mean values of triplicate infections are shown, with error bars representing one standard deviation.

FIG. 3.

Single round entry assays at different MOI. (A) Single-round entry assay on HeLa4D9 cells infected with HIV-1_NL4-3 at an MOI of 0.08 as described in Fig. 2A. At 2 days after infection, cells were stained for intracellular CA antigen with a PE-conjugated monoclonal antibody and were analyzed by flow cytometry. Dot plots depict side scatter and fluorescence intensities for cells cultivated in the absence (left panels) or presence (right panel) of tetracycline. The percentage of infected cells is indicated. (B) HeLa4D9 cells were infected with HIV-1_NL4-3 at MOIs of 0.08, 0.24, and 0.72 and analyzed by FACS as described above. Percentages of HIV-1-infected cells are shown for HeLa4D9 cells cultivated in the absence (░⃞) or presence (▪) of tetracycline as described in Fig. 2A. The mean values of triplicate infections are shown, with error bars representing one standard deviation.

FIG. 4.

Analysis of RT products in HIV-1 entry. (A) Schematic depiction of HIV-1 RT. Primer tRNA annealed to the primer binding site of the genomic RNA is extended to the 5′ end, and strong-stop DNA is subsequently transferred to the 3′ end of the genome, where it is elongated. RNA is shown as a bold line; DNA is shown as a dotted line. The U3 region amplified by the PCR primers is shaded gray. (B) Kinetic analysis of HIV-1 reverse transcription. HeLaP4 cells (▪) or HeLa cells lacking the CD4 receptor (□) were infected with HIV-1_NL4-3 at an MOI of 0.5. Total cellular DNA was extracted at the indicated time points and analyzed by real-time PCR. The copy number was determined by parallel analysis of a plasmid standard, and the results were normalized for copies of the gapdh gene. Means of duplicate infections are shown. (C) Quantitative analysis of HIV-1 reverse transcription products in cells expressing or lacking dyn^ts. HeLa4D9 cells were infected with DNase-treated HIV-1_NL4-3 at an MOI of 0.29 as described in Fig. 2A. Total cellular DNA was extracted immediately after infection or after 7 h at 37°C and analyzed by real-time PCR as described above. U3 copy numbers (normalized for total DNA) are shown for HeLa4D9 cells cultivated in the absence (░⃞) or presence (▪) of tetracycline. Mean values of three independent PCR analyses for triplicate (0 h) or quadruplicate (7 h) infections are shown, with error bars representing one standard deviation. p.i., postinfection.

FIG. 5.

Analysis of HIV-1 entry using the β-lactamase (Blam) assay. (A) HeLa4D9 cells cultivated in the absence (a and a′) or presence (b and b′) of tetracycline were infected with HIV-1 particles carrying the Blam-Vpr fusion protein at an MOI of 0.185 as described in Fig. 2A. After 5 h at 37°C, cells were loaded with the green fluorescent Blam substrate CCF2/AM. Cells were analyzed for the blue fluorescent cleavage product by fluorescence microscopy at a wavelength of 447 nm (a and b) and for the green fluorescent substrate at a wavelength of 520 nm (a′ and b′) (9, 59). Infected cells were strongly fluorescent in the blue channel (arrows). Note that uninfected cells exhibited a weak blue fluorescence in the blue channel as a result of false coloration. This was due to detection of the green fluorescent substrate in the blue channel but could be clearly distinguished from the blue signal of the cleavage product in the microscope. (B) HeLa4D9 cells cultivated in the absence (░⃞) and presence (▪) of tetracycline were infected with HIV-1 particles carrying the Blam-Vpr fusion protein at MOIs of 0.074 and 0.185. Blam activity was determined by fluorescence microscopy as described above. Mean values of triplicate infections (400 to 800 cells per infection analyzed) are shown, with error bars representing one standard deviation. Background levels determined as relative numbers of blue fluorescent cells in uninfected HeLa4D9 cells were subtracted (6.8 ± 0.8% for cells cultivated in the absence and 0.5% ± 0.17% for cells cultivated in the presence of tetracycline).

FIG. 6.

Analysis of HIV-1 entry in transiently transfected HeLaP4 cells. HeLaP4 cells were transfected with the indicated expression constructs. At 42 h after transfection, cells were infected with HIV-1 particles carrying Blam-Vpr at an MOI of 0.2. Cells were incubated for 5 h at 37°C, and fluorescently labeled transferrin was added 20 min before cells were loaded with the Blam substrate CCF2/AM. Cells were analyzed for endocytosis of labeled transferrin (left panels, a to f), for HIV-1 infection as indicated by the blue fluorescent Blam cleavage product (middle panels, a′ to f′), and for productive transfection as indicated by GFP detection (right panels, a" to f"). Note that uncleaved CCF2/AM was also detected in the GFP channel and obscured weak GFP signals. Accordingly, only strongly GFP-positive cells could be scored and dnEps15GFP, which yielded a very weak signal, could not be detected by fluorescence analysis. Block of transferrin endocytosis was used to identify productively transfected cells in this case. Arrows identify cells with inhibited transferrin uptake (see left panel), arrowheads identify GFP-positive cells infected with HIV-1, and asterisks identify GFP-positive cells not infected with HIV-1. A representative experiment is shown. All signals are shown as false colors.

FIG. 7.

Quantitative analysis of HIV-1 cytosolic entry. HeLaP4 cells were transiently transfected with constructs expressing GFP fusion proteins with wild-type or mutant dynamin (A), wild-type or mutant Eps15 (B), or wild-type or mutant caveolin (C). At 42 h after transfection, cells were infected with HIV-1 particles carrying the Blam-Vpr fusion protein at an MOI of 0.2. Cells were incubated and treated as described in the legend to Fig. 6. HIV-1 infection (as indicated by the blue fluorescent Blam product) of cells was scored in three groups according to expression of the GFP fusion protein and to inhibition of endocytosis. Nontransfected cells (GFP negative) (▪), productively transfected cells exhibiting normal transferrin endocytosis (GFP positive, punctate pattern of fluorescently labeled transferrin) (□), and productively transfected cells with blocked endocytosis (GFP positive, transferrin negative) (░⃞) are indicated. For dnEps15 in panel B, only transferrin internalizing (▪) and transferrin-negative () cells were scored due to the weak GFP fluorescence (compare with Fig. 6). The mean values of three independent infections (five independent infections in case of dnEps15GFP, Eps15GFP, and dyn^K44AGFP) are shown with >400 cells analyzed in total per population. The error bars represent one standard deviation. Background blue fluorescence in uninfected cells was observed in <1% of cells.