An inducible cell-cell fusion system with integrated ability to measure the efficiency and specificity of HIV-1 entry inhibitors

Abstract

HIV-1 envelope glycoproteins (Envs) mediate virus entry by fusing the viral and target cell membranes, a multi-step process that represents an attractive target for inhibition. Entry inhibitors with broad-range activity against diverse isolates of HIV-1 may be extremely useful as lead compounds for the development of therapies or prophylactic microbicides. To facilitate the identification of such inhibitors, we have constructed a cell-cell fusion system capable of simultaneously monitoring inhibition efficiency and specificity. In this system, effector cells stably express a tetracycline-controlled transactivator (tTA) that enables tightly inducible expression of both HIV-1 Env and the Renilla luciferase (R-Luc) reporter protein. Target cells express the HIV-1 receptors, CD4 and CCR5, and carry the firefly luciferase (F-Luc) reporter gene under the control of a tTA-responsive promoter. Thus, Env-mediated fusion of these two cell types allows the tTA to diffuse to the target cell and activate the expression of the F-Luc protein. The efficiency with which an inhibitor blocks cell-cell fusion is measured by a decrease in the F-Luc activity, while the specificity of the inhibitor is evaluated by its effect on the R-Luc activity. The system exhibited a high dynamic range and high Z'-factor values. The assay was validated with a reference panel of inhibitors that target different steps in HIV-1 entry, yielding inhibitory concentrations comparable to published virus inhibition data. Our system is suitable for large-scale screening of chemical libraries and can also be used for detailed characterization of inhibitory and cytotoxic properties of known entry inhibitors.

Figure 1. Assay scheme for HIV-1 Env-mediated cell-cell fusion.

An activator (tTA) is used for inducible expression of the HIV-1 Envs in the effector cells, allowing Env expression to be turned on by the removal of Dox from the medium. When effector and target cells are cocultivated, cell-cell fusion enables the diffusion of tTA to the target cell. This activates the transcription of the F-Luc reporter gene, which is also under the control of a tTA responsive promoter. F-Luc activity can be measured and is quantitatively related to the extent of fusion.

Figure 2. Expression of HIV-1 Envs in effector cells and HIV-1 receptors in stable target cells.

A. The H-AD8 and H-JRFL cells were induced for 48 hours and then metabolically labeled with ³⁵S overnight in the absence of Dox. Control cells were treated similarly but 2 µg/ml Dox was present in the medium at all times. Cell lysates and supernatants were subjected to radioimmunoprecipitation and analysis on polyacrylamide gels. B. Levels of expression of CD4 and CCR5 receptors on the two target cells were analyzed by flow cytometry. Cells were stained with secondary antibody alone as a control, or with either OKT4 (anti-CD4) antibody or 2D7 (anti-CCR5) antibody followed by the secondary antibody. Color codes are identical for both target cells. MFI, mean fluorescence intensity.

Figure 3. Direct visualization of cell-cell fusion with stably integrated and inducible GFP gene in target cells.

The H-AD8 and H-JRFL effector cells were incubated separately with Cf2-tetO-GFP cells; forty hours later, cells were visualized under a fluorescence microscope. Images were acquired for target and effector cells alone and after cocultivation. The phase contrast and fluorescence images are shown for each culture, except for the transiently transfected cells. Cf2Th-CD4/CCR5 cells that were transiently cotransfected with the plvx-tetO-GFP and tTA expression plasmids (plvx-TetOff Advanced, Clontech) were visualized after 24 hours and are shown in the right top panel.

Figure 4. Effects of different assay parameters on the fusion readout.

A, B. The influence of changes in the number of Cf2luc4#18 target cells (designated Cf2 cells), and either H-AD8 (A) or H-JRFL (B) effector cells, on the readout of the cell-cell fusion assay are shown. Color codes are identical for both panels. C. Same as (A) but with CEM-R5Luc1#21 target cells. D. The effect of different concentration of DMSO on the cell-cell fusion assay. E. Time course of fusion of H-AD8 and H-JRFL cells with Cf2Luc4#18 target cells and of fusion of H-AD8 cells with CEM-R5Luc1#21 cells. Results are representative from those obtained in three independent experiments.

Figure 5. Two levels of control on the fusion assay.

A. The levels of expression of HIV-1 Envs in the H-AD8 effector cells were controlled by varying the concentration of Dox during induction. Cells were washed prior to fusion, and allowed to fuse with Cf2Luc4#18 cells in Dox-free medium, initiating full induction of HIV-1 Envs. The fusion readout is shown after 7.5 and 21.5 hours. B. The level of expression of HIV-1 Envs in H-AD8 and H-JRFL cells, as well as that of F-Luc in the Cf2Luc4#18 target cells, was suppressed during fusion by adding Dox into the medium. The results shown are representative of those obtained three independent experiments.

Figure 6. Design and testing of a dual reporter assay measuring inhibition efficiency and specificity.

A. A dual reporter system was designed by incorporating the R-Luc gene, under the control of a tTA-responsive promoter, into effector cells. When these cells are fused to target cells, two readouts are measured sequentially: F-Luc activity measures the extent of cell-cell fusion and R-Luc activity measures any off-target effect. Specific fusion inhibitors are expected to give low F-Luc and high R-Luc activities, whereas unrelated inhibition results in low activities of both luciferases. B. Various compounds were tested with the dual reporter system using H-AD8#15Ren effector cells and CEM-R5L1#21 target cells. Both F-Luc and R-Luc activities were measured and normalized to those activities seen in cells without any inhibitor. The final concentration of DMSO during the cell-cell fusion assay was 0.1% unless otherwise indicated. T20, T20 peptide (118 nM); Mar, Maraviroc (1.18 µM); CHX, cycloheximide (100 µg/ml); RPA, RPA-T4 (anti-CD4 antibody, 1 nM); OKT, OKT4 (anti-CD4 antibody, 1 nM); Con.Ab, Control antibody (1 nM); NBD, NBD-556 (9.1 µM); Dox, doxycycline (2 µg/ml); EF, efavirenz (an HIV-1 reverse transcriptase inhibitor, 312 nM).

Figure 7. Statistical analysis of the dual reporter system with two different target cells.

A. Distribution of F-Luc and R-Luc signals obtained in the cell-cell fusion assay conducted with or without 2 µg/ml Dox. Readouts of fusion without Dox are from 35 wells of a single plate and those with Dox are from 35 wells from multiple plates. B. Statistical analysis for the precision of the system. Data from 5 (CEM-R5Luc1 cells) or 4 (Cf2Luc4#18) independent experiments were used to calculate the Z'-factor of the assay for each luciferase. Means ± standard deviations are shown for each cell type.

Figure 8. Dose-response curves for inhibition efficiency and specificity of known HIV-1 entry inhibitors.

A. The gp41-directed T20 peptide, gp120-directed BMS-806, and CCR5-directed Maraviroc were tested for inhibition and off-target effects with the dual reporter system using two different target cells, CEM-R5Luc1#21 (left) and Cf2Luc4#18 (right). Inhibition data were fitted to the logistic equation and the IC₅₀ value for each tested compound and each target cell type was calculated. R²>0.96 for all curves. Red lines display relative R-luc activity and indicate the specificity of a compound; green lines are fitted curves for relative F-luc activity and indicate the inhibition of cell-cell fusion for each compound. The results shown are representative of those obtained in three or four independent experiments, each performed with more than three replicates. B. Reported IC₅₀ values for the three inhibitors in antiviral assays , , , , , (inhibition of the HIV-1_MN isolate by BMS-806 was reported to be 743 nM and this atypical value is excluded from the table).