Update your browser to view this website correctly. Update my browser now
The route of HIV-1 entry for productive infection in CD4+ host cells is a fundamental question for the molecular understanding of HIV-1 infection and transmission. Although direct fusion has long been thought to be the mode of entry, recent studies have suggested that productive entry of HIV-1 may actually occur through dynamin-dependent endocytosis. In several of these studies, dynasore, a noncompetitive inhibitor of the GTPase activity of dynamin, has been used to support this conclusion. Here we show that dynasore does produce inhibitory effects on the productive infection of HIV-1 in several commonly used cell lines. This effect is present regardless of the methods used to facilitate the infection of HIV-1. However, transferrin uptake remains fully functional in these cell lines upon dynasore treatment. Therefore, the inhibition on HIV-1 infection by dynasore in these cell lines is due to an effect that is independent of transferrin endocytosis. The use of dynasore in probing the role of endocytosis in HIV-1 infection should be corroborated by other methods.
Dynamin is a large GTPase whose membrane fission activity is required for clathrin-mediated endocytosis and also some clathrin-independent endocytic processes. Dynasore, originally discovered as a noncompetitive inhibitor of the GTPase of dynamin, has been widely used as a cell-permeable and fast-acting inhibitor to study the function of dynamin in various cellular events. Recently in several studies, dynasore has been used to probe the role of endocytosis in productive infection of HIV-1 in several different host cells. The inhibition on HIV-1 infection by dynasore has been interpreted as the involvement of dynamin-dependent endocytosis that leads to productive HIV-1 infection. Although dynasore was shown to be relatively specific and did not appear to affect dynamin-independent cellular functions, recent studies using dynamin triple knockout cells have revealed the robust off-target effects of dynasore on membrane ruffling and fluid-phase endocytosis. These studies indicate that dynasore can inhibit dynamin-independent cellular functions and highlight the presence of unknown targets of dynasore that remain to be identified. It also raises the caution on the use of this inhibitor in probing dynamin-dependent cellular functions.
The cellular route of HIV-1 entry for productive infection in CD4+ host cells is a fundamental question for the molecular understanding of HIV-1 infection and transmission. To elucidate the role of dynamin-dependent endocytosis on productive infection of HIV-1, we have thus evaluated the effect of dynasore on HIV-1 infection in several well-established cell lines.
To investigate the entry of HIV-1 into host cells, we have chosen to work with three different cell lines that are well established in the literature for study of HIV-1 infection in cell culture: (1) TZM-bl cells, a HeLa-derived indicator cell line that was engineered to overexpress CD4 and CCR5 receptors, which is widely used for assaying the infectivity of HIV-1 virions in cell culture; (2) Rev-CEM cells, an engineered indicator T cell line that has been demonstrated to report authentic HIV-1 infection; and (3) SUP-T1 cells, a non-engineered T cell line derived from an 8 year old boy who had a relapse of T-cell non-Hodgkin’s lymphoma. To determine the effects of dynasore on HIV-1 infection in the above cell lines, we have treated these cells with various concentrations of dynasore for 30 min at 37°C before HIV-1 infection. These cells were then infected by HIV-1 for 2 hours at 37°C (Materials and Methods), followed by continued incubation at 37°C until the quantitation of HIV-1 infection. To control for the potential toxicity of dynasore on these cultured cells, we also prepared mock infection in the absence of virus but in the presence of various concentrations of dynasore, and measured the number of viable cells at each concentration by trypan blue staining. As shown in figure 1A (a, c and e), the fractions of remaining HIV-1 infectivity relative to those at zero concentration of dynasore are plotted in solid curves as a function of dynasore concentration in the culture ([Dynasore]) for these cell lines. The fractions of viable cells relative to those at zero concentration of dynasore are also plotted in white bars at each [Dynasore]. From these plots, it is clear that dynasore exerts inhibitions on productive HIV-1 infection at nontoxic concentrations, although the degree of inhibition varies from cell line to cell line. At 200 µM dynasore, both TZM-bl cells (Fig. 1Aa) and Rev-CEM cells (Fig. 1Ab) display inhibition of productive HIV-1 infection by 50%. However, this inhibition is subtle for SUP-T1 cells within the range of [dynasore] that we can test. Significant cell death occurs at 100 µM dynasore or above for SUP-T1 cells.
For a side-by-side comparison with other inhibitors, we also carried out studies in the above cell lines using another well-established HIV-1 entry inhibitor, T20. The action of T20 is to bind the pre-hairpin intermediates of envelope glycoproteins during membrane fusion, which prevents six-helical bundle formation and thus blocks membrane fusion and HIV-1 infection. As shown in figure 1A (b, d and f), the fractions of remaining HIV-1 infectivity relative to those at zero concentration of T20 are plotted in solid curves as a function of T20 concentration in the culture ([T20]) for these cell lines. The fractions of viable cells relative to those at zero concentration of T20 are also plotted in white bars at each [T20]. For all 3 cell lines, T20 can achieve close to 100% inhibition of HIV-1 infection under nontoxic concentrations. This comparison between dynasore and T20 also suggests that dynasore is not as potent as T20, and the pathways of HIV-1 infection blocked by dynasore may thus represent only a fraction of the pathways available for HIV-1 infection in these host cells.
The observed inhibition in TZM-bl cells by dynasore is qualitatively consistent with Miyauchi et al., who first reported the inhibitory effect of dynasore on productive HIV-1 infection in TZM-bl cell line. Indeed, the same isolate of HIV-1 but pseudotyped with the envelope glycoprotein of vesicular stomatitis virus (VSV-G) produced quantitatively very similar patterns of inhibition in TZM-bl cells (open squares in Fig. 1Ba), suggesting that it is impairment of an endocytic process that partially blocks the infection of both viruses. Furthermore, the trend of inhibition remains quantitatively the same when we conducted infection in the absence of DEAE-dextran (Fig. 1Bb), suggesting that it is not the presence of DEAE-dextran that sensitized the virus to an endocytic process. Moreover, this inhibitory effect remained when we changed the viral envelope to that of HXB2, one of the prototype HIV-1 clones from infected human T cell lines (Fig. 1Bc), although the degree of inhibition for HXB2 infection that we have observed at 80 µM dynasore was less than that by Miyauchi et al.. Furthermore, for both NL4-3 and HXB2 enveloped viruses, spinoculation onto TZM-bl cells pretreated with dynasore produced comparable inhibition as infection in the presence of DEAE-dextran (Fig. 1Bc). Thus, this inhibition by dynasore on productive HIV-1 infection in TZM-bl cell line is real, regardless of the methods used to facilitate HIV-1 infection.
To probe the mechanism of dynasore inhibition of HIV-1 infection, we then assayed the transferrin uptake in the above 3 cell lines. Cellular uptake of transferrin has been characterized as a classical example of clathrin-dependent endocytosis, in which dynamin is required for clathrin-coated vesicles to pinch off from the plasma membrane. Transferrin binds at the cell surface to its corresponding receptor and is internalized by receptor-mediated endocytosis, a process that requires the formation of clathrin-coated pit. We incubated Alexa-488 labeled transferrin with various host cells that were pretreated with dynasore at 37°C for 30 min, and then used acid wash to remove surface-bound transferrin molecules before subjecting cells to flow cytometry for analysis. Across the range of dynasore concentrations that we have investigated for viral infection in each cell line, no apparent inhibition of transferrin uptake was observed at 37°C (white bars in Fig. 1C), the condition under which we conducted virus infection. In contrast, transferrin uptake can be almost completely blocked by shifting cells to 4°C for transferrin binding and uptake (black bars in Fig. 1C). These results indicate that transferrin uptake is fully functional in these cell lines despite the treatment with dynasore at concentrations that HIV-1 infection has been compromised.
As a further support to the above flow cytometry results, we have conducted confocal imaging for TZM-bl cells after incubation with Alexa-488 conjugated transferrin. As shown in figure 1D, Alexa-488 containing endosomal vesicles can be clearly observed both in the absence and presence of dynasore, indicated by the green punctate signals around the inner edge of the plasma membranes. Moreover, live cell imaging experiments using Alexa-594 conjugated transferrin revealed no apparent inhibition of transferrin uptake by dynasore at any time during the experiments. As shown in supplementary Movies 1 and 2, fluorescent endosomes were clearly visible both in the absence and presence of dynasore. The dynamic movements of these endosomes inside the cells are quite evident in both cases.
In this study, the effect of dynasore on productive HIV-1 infection in 3 commonly used cell lines has been tested. Inhibition on HIV-1 infection was observed in all 3 cell lines, although the degree of inhibition varied. However, the observed inhibition is not correlated with any reduction in transferrin uptake. Instead, transferrin uptake remains fully functional in all these cell lines despite the pretreatment of these cells with dynasore at concentrations that partially block HIV-1 infection. These observations revealed that HIV-1 entry does not share the same pathway as transferrin uptake in these cell lines, and suggest that the inhibitory effect of dynasore on HIV-1 infection is not due to the impairment of clathrin-dependent endocytosis. Rather, dynasore displays an apparent complex effect in its use as an inhibitor for HIV-1 infection in these cell lines. Recent studies using dynamin triple knockout cells have revealed the inhibitory effects of dynasore on fluid-phase endocytosis and peripheral membrane ruffling, even though all 3 isoforms of dynamin are absent. Therefore, the molecular targets of dynasore during HIV-1 infection in these cell lines remain to be identified.
Because the above studies were all conducted in cell culture, the relevance of these results for HIV-1 infection in vivo remains to be tested.
The inhibitory effect of dynasore observed for HIV-1 pseudotyped with VSV-G (Fig. 1Ba) suggests that an endocytic process was affected by dynasore. However, the complex effect of dynasore renders it difficult to precisely assess the role of endocytosis for productive infection of HIV-1 virions in these cell lines. Independent methods or more specific inhibitors are required in order to draw definitive conclusions. Among these, the K44A dominant-negative mutant of dynamin is likely to be a good candidate, which can be used for more specific inhibition of dynamin-dependent endocytosis. The role of dynamin-dependent endocytosis on HIV-1 infection can then be determined.
Cell culture and reagents
HEK 293T/17 cells (ATCC, Manassas, VA) and TZM-bl cells were cultured at 37°C with 5% of CO2 in complete media containing 90% of DMEM and 10% of FBS (HyClone, Laboratories, Logan, UT). Rev-CEM cells and SUP-T1 cells were maintained in RPMI 1640 supplemented with 10% FBS. Throughout, all cell lines were discarded after 10 passages and new aliquots of frozen cells were thawed to improve reproducibility of virion production and infection experiments. Dynasore (Sigma-Aldrich) stock solution was prepared in DMSO. Control experiments showed that DMSO has no effect on HIV-1 infectivity. T20 (Roche) stock solution was prepared in distilled and deionized milliQ water. Inhibitor stock solutions were aliquoted, flash-frozen, and stored in -80°C freezer.
Production of single-cycle HIV-1 virions
To produce single cycle HIV-1 virions, 106 293T cells in a 2 ml culture volume were seeded overnight in a 35 mm dish before transfection using the TransIT LT-1 transfection reagent (Mirus Bio, Madison, WI). For each dish, 1 mg of the provirus-containing plasmid pNL4-3E- was used to make the transfection reagent mixture, together with 1 mg envelope expression plasmid for NL4-3 or HXB2 as indicated in the text. The transfection reagent mixture was incubated at room temperature for 15 min before drop wise addition to the culture media. At 6 h post transfection, the culture media together with the transfection reagents was replaced with fresh complete media and the incubation was continued at 37°C with 5% CO2. At 24 h post transfection, the entire culture media containing single-cycle HIV-1 viruses was collected and filtered through a 0.45 µm syringe filter (Millex-HV PVDF, Millipore) in less than 10 min on average. The filtrate was then aliquoted on ice, flash-frozen in liquid nitrogen and stored in a -80°C freezer. Control experiments showed that virion infectivity remains constant during incubation on ice for up to 8 h and flash-freezing in complete media leads to no loss of virion infectivity. The single cycle virions pseudotyped with either VSV-G or HXB2 envelope were produced using the exact procedures above except that the envelope plasmid for NL4-3 was replaced with that of VSV-G or HXB2. To produce single-cycle virions tagged with mCherry, the same procedures were used as described above except the input DNA used during transfection. In detail, HIV virions carrying free mCherry were generated by transfection of 293T cells with 0.5 mg pNL4-3E- plasmid, 0.5 mg pNL4-3E-MA-mCherry-CA plasmid and 0.1 mg NL4-3 envelope plasmid in 2 ml volume in a 35 mm dish. The pNL4-3E-MA-mCherry-CA plasmid encodes a provirus in which the mCherry protein was inserted between the matrix and capsid domains of the Gag protein, and the mCherry protein sequence was flanked by 2 HIV protease cleavage sites.
Infection assays in each cell line
For infection assay in TZM-bl cell line, 8×104 TZM-bl cells in a 1 ml culture volume were seeded in each well of a 12 well plate 1 day prior to infection. On the next day, virus stocks taken out of -80°C freezer were placed in a room temperature water bath until just thawed, and serially diluted in complete media containing 20 mg/ml DEAE-dextran. 100 ml of viruses at each dilution were layered on top of the cell and the infection was continued for 2 h at 37°C with gentle rocking every 30 min. For spinoculation in TZM-bl cells, upon mixing the virus with cells in the absence of DEAE-dextran, the mixture in a 12 well plate was centrifuged at 1200 g for 2 h at 25°C. At the end of 2 h, 1 ml of complete media was added to each well and the incubation was continued at 37°C for 48 h with 5% CO2. At the end of 48 h, cells were fixed in 2% gluteraldehyde at room temperature for 5 min. After fixation, the cells were washed 3 times with PBS, and stained for 50 min at 37°C using cell staining solution provided in the beta-galactosidase staining kit (Mirus Bio, Madison, WI). After incubation, the cells were washed 3 times with milliQ water and the number of blue cells in each well was counted with a Nikon TS100-F inverted microscope. For infection assay in Rev-CEM or SUP-T1 cell lines, 200 µl of cells were incubated with 100 µl of 50% free mCherry-labeled virus in 1.5 ml polypropylene test tube for 2 h at 37°C with continuous gentle rocking on a nutator in the presence of DEAE-dextran at designated concentrations. After 2 h, free virions were removed by washing with complete media. The cells were further incubated for 5 days at 37°C in 2 ml of complete media in a 12 well plate. At the end of 5 days, cells were fixed with 2% paraformaldehyde for 5 min at room temperature and washed with PBS. For both Rev-CEM cells and SUP-T1 cells, infected cells express mCherry protein and are quantitated with a flow cytometer using 561 nm laser (iCyt Synergy, Sony). For each cell line, the concentration of DEAE-dextran was optimized to obtain the highest infectious virus concentration for a given batch of virus in each cell line, which resulted in 20, 5 and 10 mg/ml, for TZM-bl, Rev-CEM and SUP-T1 cells, respectively. At these concentrations of DEAE-dextran for each cell line, the fractions of live cells as quantitated by trypan blue assay 16, 17 were all above 95%.
Infection assays in the presence of inhibitors
First, to determine the toxic concentrations of each inhibitor on cell growth, TZM-bl, Rev-CEM, or SUP-T1 cells were pretreated with various concentrations of dynasore or T20 for 30 min at 37°C, and further incubated for 2 or 5 days, following the infection assay conditions but without virus infection. Cells were then collected and incubated with 0.4% w/v trypan blue dye for 5 min at room temperature and the fraction of live cells were counted under a bright-field microscope. T20 didn’t show any apparent toxicity within the range of concentrations used for all cell lines. However, high concentrations of dynasore did show toxicity. For virus infection in the presence of dynasore, all cells were pretreated with a certain concentration of dynasore in the absence of serum for 30 min at 37°C, and then incubated with the virus. The concentrations of dynasore are reflective of the level reported in literature. The virus incubation also included dynasore at desired concentrations, although there was no difference in the inhibition on viral infection between the absence and presence of dynasore during virus incubation for all three cell lines (data not shown).
Transferrin uptake in each cell line
For transferrin uptake assay in TZM-bl cells, 3×104 TZM-bl cells in 1 ml complete media were seeded in each well in a 12 well plate and incubated overnight. On the second morning, the cells were pretreated with dynasore at various concentrations for 30 min at 37°C and then incubated with 20 µg/ml of Alexa-488 conjugated transferrin prepared in DMEM without FBS (Invitrogen) for 5 min at 37°C followed by at 4°C for 20 min. Controls were incubated with the same concentration of transferrin conjugates for 25 min at 4°C throughout. All the subsequent procedures were carried out at 4°C. First, cells were washed with pre-chilled PBS 6 times and incubated with pre-chilled pH 2.0 buffer for 5 min at 4°C followed by pre-chilled PBS washing twice to remove transferrin bound on cell surface. The pH 2.0 buffer was made using 500 mM NaCl and 0.2 N glacial acetic acid in milliQ water. Next, TZM-bl Cells were trypsinized, fixed with 4% paraformaldehyde for 10 min at room temperature, washed with PBS 3 times, and analyzed with a flow cytometer (FACS Canto II) using a 488 nm laser 3. For transferrin uptake in Rev-CEM or SUP-T1 cells, the exact procedures as above were followed except the following modifications: cells and transferrin were incubated in 1.5 ml test tube instead of 12 well plate; no trypsinization was performed for these suspension cells, and washing of the cells were done by mixing with each solution and centrifuge at 1000 g for 5 min, followed by resuspension of the cells in designated buffers, all carried out at 4°C.
Confocal imaging of TZM-bl cells
For confocal imaging of transferrin uptake in TZM-bl cells, cells were seeded on Poly-L-Lysine coated coverslip in a 12 well plate and incubated overnight at 37°C with 5% CO2. On the second morning, the cells were pretreated with 0 or 200 µM dynasore for 30 min at 37°C and then incubated with 20 µg/ml of Alexa-488 conjugated transferrin prepared in DMEM without FBS (Invitrogen) for 5 min at 37°C followed by at 4°C for 20 min. The cells were then washed with pre-chilled PBS 6 times and incubated with pre-chilled pH 2.0 buffer for 5 min at 4°C followed by pre-chilled PBS washing twice to remove transferrin bound on cell surface. The cells on coverslip were then fixed with 4% paraformaldehyde for 10 min at room temperature, and washed with Tris buffered saline. The cells on coverslip were then stained with Cholera Toxin Subunit B conjugated with Alexa-555 (Invitrogen) at a concentration of 1 µg/ml in PBS for visualization of plasma membranes. The coverslip was then washed with PBS and mounted onto a glass cover slide with 3 µl mounting media, sealed with nail polish, and imaged using an Olympus FluoView 500 Laser Scanning Confocal Microscope. Confocal images were collected using 100× oil immersion objective. The fluorophores were excited using 488 nm laser for Alexa-488 conjugated transferrin and 543 nm laser for Alexa-555 conjugated Cholera Toxin Subunit B.
Live cell imaging of transferrin uptake in TZM-bl cells
Exponentially growing TZM-bl cells were trypsinized and seeded onto chambered coverglass for at least 6 h before experiments. The cells were pretreated with 0 or 108 µM dynasore for 30 min at 37°C. At time 0, 5 µg/ml of Alexa-594 conjugated transferrin diluted in DMEM was added to cells, incubated for 5 min, and then replaced immediately with warm complete media without transferrin. For live cell imaging, the chambered coverglass was mounted onto a Tokai Hit stage incubator installed on an Olympus IX71 inverted fluorescence microscope (Olympus, Center Valley, PA). The cells were maintained at 37°C with a supply of 5% CO2 directly to the stage incubator. Live cell fluorescence images were collected with a 100× oil immersion objective (Olympus, N.A. 1.4) and imaged onto an iXon3 897 back-illuminated EMCCD camera (Andor Technology, Belfast, Northern Ireland). The Alexa-594 conjugated transferrin was excited under epifluorescence illumination by a linearly polarized 592 nm laser line (VFL-P-1000; MPB Communications Inc., Montreal, Canada) with an irradiance of 20 W/cm2. The excitation filter was 588/30 nm and the emission filter was 629/40 nm. Fluorescence images were captured at an exposure time of 100 ms and the movies were taken at a rate of 1 frame per second. Stage drift was negligible during this time frame as revealed from images of fiducial markers before and after the movies. The movies shown were taken at 20 min after the addition of Alexa-594 conjugated transferrin to both dynasore treated and untreated TZM-bl cells.
This work was supported by NIH Director’s New Innovator Award 1DP2OD008693-01 to WC, and also in part by Research Grant No. 5-FY10-490 to WC from the March of Dimes Foundation. M.C.D. is supported by a NIH postdoctoral fellowship awarded under F32-GM109771.
We thank Cheng Lab members for helpful discussions. The following reagents were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH): pNL4-3 from Dr. Malcolm Martin; TZM-bl cells from Dr. John C. Kappes, Dr. Xiaoyun Wu and Tranzyme Inc, Rev-CEM cells from Dr. Yuntao Wu and Dr. Jon W. Marsh; SUP-T1 cells from Dr. James Hoxie; pHXB2-env from Dr. Kathleen Page and Dr. Dan Littman; and T20 fusion inhibitor from Roche. The plasmid encoding VSV-G was through Addgene from Dr. Bob Weinberg.
All studies conducted in this work were approved by the University of Michigan IBC (IBC00000560).