We have developed an advanced cell-cell fusion system with a unique configuration in which the tTA transactivator was exploited to regulate expression of all components including HIV-1 Envs, F-Luc and R-Luc. Cocultivation of Env-expressing effector cells and receptor-expressing target cells resulted in robust and reproducible cell-cell fusion. The dual system effectively distinguished specific fusion inhibitors from non-related ones, and reliably measured inhibitory concentrations as well as cytotoxicity for known fusion inhibitors.
The tetracycline–inducible expression system was selected because it is well characterized and offers very tight regulation, with low background levels of transcription and high expression levels 
. Interestingly, Env expression from the non-lentiviral vector pGL4.2-JRFL was consistently higher than Env expression from the lentiviral vector plvx-JRFL in both transiently and stably transfected cells. Although the tTA promoter and Env sequences are identical in both vectors, the flanking sequences in the vector may suppress or enhance tTA-mediated transcription; moreover, distinct integration sites in the stable cell lines may exert different effects on transcription.
The specific design of our cell-cell fusion system offers several advantages: 1) the use of inducible stable cell lines avoids Env-related toxicity and provides consistent expression of HIV-1 Envs relative to repeated transient transfections, which, in addition to being more variable, are time- and resource-consuming; 2) HIV-1 Env expression is tightly regulated and derived from a codon-optimized gene, alleviating the requirement for the HIV-1 Rev protein in the system; 3) tight control of tTA allows a complete shutoff of transcription of all components of the system, as well as the ability to control Env expression level prior to the cocultivation of effector and target cells; 4) R-Luc activity can be used as a surrogate for the level of Env expression on the effector cells because both Envs and R-Luc are regulated by the same tTA protein in these cells; 5) The system can distinguish tTA inhibitors, which will result in off-target effects, from specific fusion inhibitors because R-Luc activity is regulated by tTA; and 6) dual measurements of luciferase activity allow the evaluation of both efficiency and specificity of inhibition in the same cells under the identical experimental conditions, avoiding potential errors associated with measuring these two properties separately. The only limitation of our system is the inability to distinguish fusion inhibitors from specific F-Luc inhibitors. F-Luc inhibitors will lower F-Luc activity without a significant decrease in R-Luc activity, and therefore will be scored as specific fusion inhibitors (false positives). Nevertheless, such compounds are presumably very rare, are easily identifiable in secondary assays, and thus should not significantly compromise the utility of the system to measure specificity.
Some of the above features of our system have been previously used in transient or stable fusion assays. However, these systems have not incorporated all of these features into a single complete system; moreover, the manner in which these elements were utilized differs from ours. One assay monitored the fusion of transiently transfected 293T cells 
; in this case, both transfection efficiency and cytotoxicity were measured with a single R-Luc marker, and thus the contribution of each to the total effect was unknown. A tTA-based system has been used for transient expression of Envs and for detection of fusion inhibition 
. Another fusion system with inducible expression of HIV-1 Envs has also been reported 
. In this system, Dox was used for induction and was required throughout the fusion assay. Coexpression of HIV-1 Rev was required for Env expression, and HIV-1 Tat was used as a transactivator to trigger F-Luc expression in the target cells. Identifying fusion inhibitors in this setting requires an additional step to confirm that the compounds do not inhibit either HIV-1 Rev or Tat. In contrast to our system, this earlier assay did not contain a built-in ability to measure off-target effects.
Our cell-cell fusion system, like other fusion systems, is focused on a single step in the HIV-1 life cycle. Such a strategy will not identify inhibitors of other steps in the virus life cycle, but has several advantages. Compared with more comprehensive antiviral assays, the more focused cell-cell fusion assay requires a shorter time of incubation, may be less sensitive to non-specific effects, and allows the target of any inhibitor to be easily identified. In addition, the cell-cell fusion system does not require the use of special biosafety containment facilities, as no infectious virus is used.
Our dual reporter system was successfully tested against a wide range of inhibitors and control molecules. The system detected known fusion inhibitors and underscored their specificity, whereas control agents did not significantly affect cell-cell fusion. Specific inhibition was documented for inhibitors that target different Env or receptor components of the HIV-1 entry process; for different types of inhibitory molecules such as chemical compounds, peptides and antibodies; and for inhibitors with diverse levels of efficiency [ranging from an IC50 of ~10 µM (NBD-556) to ~3nM (Maraviroc)]. Non-related effects of cycloheximide and tTA-inhibitors were clearly categorized as off-target, as they completely abolished the activity of both luciferases. Measurements of inhibitory concentrations and cytotoxicity of known inhibitors validated the utility of the fusion system. The system was efficiently inhibited in a dose-response manner by all known inhibitors, with different efficacy for the two target cells types tested. CCR5-directed Maraviroc was the most potent inhibitor of fusion to CEM-R5Luc1#21 cells and a less efficient inhibitor for fusion to Cf2Luc4#18 cells. Such differences in potency are expected, since the lower number of target CCR5 molecules on CEM cells, relative to those on Cf2Th cells, enables effective inhibition by even low concentrations of Maraviroc. Our results suggest that the distinct properties of the two target cells available for this cell-cell fusion system may exhibit different levels of sensitivity for the detection of particular types of inhibitors. CEM-R5Luc1#21 cells grow in suspension and are derived from human T-cells that are closely related to authentic target cells during HIV-1 infection; fusion with CEM-R5Luc1#21 cells is inhibited very efficiently by known fusion inhibitors. Cf2Luc4#18 are adherent cells, and are less sensitive to fusion inhibitors, but exhibited a slightly higher Z'-factor with less variation.
Adaptation of the dual reporter system to high-throughput screening in a 384-well format may require additional optimization of cell numbers to adjust for the lower volumes. In addition, commercial luciferase substrate solutions, such as Steady-Glo (Promega), which are designed to have a long half-life for batch processing, have lower luminescence intensities in comparison to other formulations that are designed for high sensitivities (Steady-Glo Technical Manual, Promega). Using such a substrate resulted in lower readouts than the luciferase substrate that was used in our study (data not shown). Moreover, because some decrease of signal readout was observed after long-term propagation of the effector cells, the use of low passage cells is recommended. Preliminary data from scaled-up pilot studies indicate that the cell-cell fusion assay is robust and underscore the importance of using a control system for excluding off-target compounds.
In summary, our assay is simple and efficient. The system has high dynamic range, tolerates DMSO, is precise (high Z'-factor scores) and has been validated with a panel of reference inhibitors directed against different steps in the fusion process, yielding inhibitory concentrations consistent with published data. The dual reporter system can be used for detailed characterization of inhibition versus cytotoxic properties of entry inhibitors. In addition, the system is suitable for a high-throughput screen of chemical compounds and adds a unique layer to the already existing cell-cell fusion systems. We are currently developing additional cell-cell fusion assays that utilize Env derived from different primary HIV-1 isolates. Combining the screening with secondary assays that utilize cells expressing Env from different HIV-1 isolates may identify broad-range fusion inhibitors that can generate novel lead compounds for further development as therapeutics or prophylactic microbicides.