To ensure that reversible arrest of the infection with NVP and synchronization of reverse transcription would enable specific detection of integrase inhibitors, compounds with known modes of action were tested including the -diketo-acid-based integrase inhibitor L731,988 ,naphthyridine carboxamide inhibitor elvitegravir (in clinical phase III studies) and raltegravir. In addition, RT inhibitors (zidovudine, tenofovir, NVP, efavirenz) and entry inhibitors (enfuvirtide, BMS806) were evaluated. All compounds were assayed for inhibition of wild-type HIV-1 IIIB replication (EC50) and cytotoxicity (CC50) as described previously angiogenesis inhibitors and were confirmed as potent inhibitors (Table 1).

To determine the optimal time point for specific identification of integrase inhibitors in the CIS assay, test compounds should be added when reverse transcription is largely completed to prevent identification of RT inhibitors and where the majority of viral DNA integration into the host chromosome has yet to occur (Fig. 1). Four and a half hours after removal of NVP proved the optimal time point for test compound addition as at that time more than 50% of the viral population had completed reverse transcription and 80% remained in a pre-integration stage (Fig. 3). Consequently, RT inhibitors added at this time point were unable to elicit an EC50 value whereas INIs generated EC50 values that were consistent with the activity in the antiviral assay. This observation was confirmed experimentally by the clear distinction in dose-responses of entry inhibitor BMS806, NRTI tenofovir and integrase inhibitor L870,810 (Fig. 4). As the VSV pseudo-typed Paclitaxel HIV-1 based lacks env, BMS806 was inactive. TNF did not inhibit to greater than 50%, although the highest concentration (100 M) almost reached 50%, but this is probably due to cytotoxicity. In contrast, the integrase inhibitor L870,810 generated an expected dose response curve with an EC50 of 13 nM (Fig. 4). In addition, all integrase inhibitors were identified as hits (SI > 4) in the CIS assay, in contrast to all entry and RT inhibitors, which were inactive in the assay or showed poor selectivity indices (Table 1). The shorter assay duration of the CIS assay compared to the antiviral assay, can be seen as a minor limitation as the selectivity indices might result in a higher false positive rate, as the assay is less prone to potentially toxic aspects of compounds. On the other hand, this enable the identification of novel chemical structures as starting point for drug discovery.

Therefore, based on the obtained dose response curves, with corresponding EC50 and CC50 values, integrase inhibitors can be identified during a high throughput campaign with the CIS assay. The hit rate of a pilot screen of 11,021 compounds in highthroughput format was 0.34% after hit confirmation. The identified hits showed micromolar potency (data not shown). The robustness of the CIS assay was confirmed by determining the Z value during independent screening experiments and resulted in a Z value of 0.68, indicating good robustness (Zhang et al., 1999). Furthermore during these independent screening experiments a fixed set .Library screening to identify novel compounds is a central endeavor of drug discovery in many therapeutic areas. Therefore, screening technologies and assays are under constant development and refinement to enable identification of inhibitors of new therapeutic targets, discovery of inhibitors with novel modes of action, screening larger compound libraries, reducing the number of false positives, and increasing speed of throughput. Typically, biochemical ‘reductionist’ assays and cellular phenotypic or multi target assays represent two strategies for novel inhibitor discovery, where screening efforts in HIV-1 integrase inhibitor programs have centered on biochemical strand transfer assays  .A novel cellular integrase screening (CIS) assay was developed, enabling the identification of integrase inhibitors utilizing a replication incompetent HIV-1 based lentiviral vector. The CIS assay employs a concept also used in time of addition assays, however in the CIS assay the bcl-2 inhibitor viral population is synchronized using a temporary arrest at the level of reverse transcriptase activity. In addition, pseudo-typed retroviral particles with the envelope glycoprotein of Vesicular Stomatitis Virus eliminated sensitivity to HIV entry inhibitors. Furthermore, circumventing the inhibitory effect of RT inhibitors by focusing the CIS assay toward integrase specific inhibitors proved beneficial, where the temporary arrest of the infection during reverse transcription enabled synchronization of reverse transcription complexes. The CIS assay was validated by testing known INIs, entry and RT inhibitors and comparing the activities with the results of a cellular antiviral assay.Therefore, reversible arrest of reverse transcription in combination with addition of compounds at a specific time point enabled identification of integrase inhibitors in a high throughput screening setting. This contrasts with an assay method where compounds are added after synchronization at the entry step, as both RT and integrase compounds would show inhibitory effects and their mechanisms of action would be difficult to resolve. In conclusion, the results demonstrate that the CIS highthroughput screening assay represents a promising tool for the identification of HIV integration-specific hits from compound libraries.