Over the last decade intense efforts

Over the last decade, intense efforts have been devoted toward identifying IN inhibitors as potential drugs against HIV. Raltegravir and elvitegravir are integrase strand transfer inhibitors (INSTIs) that have been approved for therapy and dolutegravir is another INSTI that is currently in advanced clinical trials (Raffi and Wainberg, 2012). However, resistance to these INSTIs has emerged and a significant degree of cross-resistance exists among the compounds (Garrido et al., 2011, Mesplède et al., 2012, Wainberg et al., 2012). Recent efforts have turned to discovery of novel IN inhibitors, that target non-active sites of IN (Voet et al., 2009, Luo and Muesing, 2010, Al-Mawsawi and Neamati, 2011, Malet et al., 2012, Christ and Debyser, 2013). These novel IN inhibitors might avoid the problem of cross-resistance with currently available INSTIs, and might therefore be used in combination with them or with other types of anti-HIV drugs. For example, a small molecule termed BI 224436 specifically inhibits the 3′ processing reaction by targeting the non-catalytic site of IN (Yoakim et al., 2011). Another type of compounds are LEDGINs, that inhibit interactions between IN and lens epithelium derived growth factor/p75 (LEDGF/p75), a cellular co-factor essential for viral replication. LEDGINs bind to the IN dimer interface, a site distinct from the active site of IN. Both of these types of compounds, also referred to as allosteric IN inhibitors, allosterically inhibit IN catalytic activity as well as HIV-1 replication in infected Rosiglitazone HCl receptor (Bardiot et al., 2010, Christ et al., 2010). Since IN must bind to viral DNA prior to the 3′-processing step, the inhibition of IN binding to viral DNA also represents an interesting target (Voet et al., 2009, Al-Mawsawi and Neamati, 2011). The IN binding inhibitors, referred to as INBIs, should prevent the subsequent 3′-processing and strand transfer reactions (Voet et al., 2009, Al-Mawsawi and Neamati, 2011). For example, a potent derivative of styrylquinoline, FZ41, can inhibit the binding of IN to viral DNA at an IC50 of 0.75μM (Carayon et al., 2010). Importantly, FZ41 also inhibits 3′-processing and strand transfer in vitro, and blocks HIV in cell culture at concentrations of 4–10μM (Bonnenfant et al., 2004, Deprez et al., 2004). High-throughput screening (HTS) and rational design of novel chemotypes have enabled the discovery of several HIV-1 IN inhibitors (Savarino, 2006, Christ et al., 2010, Yoakim et al., 2011). Traditionally, biochemical assays have been used to screen compound libraries for identification of novel inhibitors. In the search for more potent INBIs, convenient and high-throughput methods for measuring HIV-1 IN DNA binding activity are desirable. Although, several in vitro biochemical methods that measure HIV-1 IN DNA binding activity have been described, including electrophoretic mobility shift assay, nitrocellulose filter assay, chemical and/or UV light cross-linking, fluorescence correlation spectroscopy, and surface plasmon resonance (Christ et al., 2011), these are all low throughput. More recently, fluorescence anisotropy and fluorometric assays that measure the DNA-binding activities of HIV-1 IN have been reported (Deprez et al., 2004, Anisenko et al., 2012), but these have not been validated for HTS or used for the discovery of INBIs. Here, we have modified a simple fluorescence microplate-based assay for detection of protein–DNA interactions (Zhang et al., 2003), and now report on its use for the discovery of INBIs. This assay is robust, inexpensive, requires no specialized instrumentation, and is readily amenable to HTS.
Materials and methods
Results
Discussion The ability to accurately, rapidly and efficiently identify active compounds from large chemical libraries is a goal of HTS assays. Although several in vitro HIV-1 IN DNA binding assays have been reported, none of them has been validated for HTS or used for the discovery of INBIs. Here, we describe an easy, inexpensive, and robust fluorescence assay for measuring HIV-1 IN DNA binding activity. This assay has been optimized in regard to concentrations of IN, LTR DNA substrate, salt, and time (Fig. 1, Fig. 2, Fig. 3). The signals measured in this assay for controls with either IN or LTR alone were at background levels. The apparent Kd value, the influence of ionic strength on IN DNA binding affinity and the non-specific binding of IN to random DNA, as measured by competition experiments (Figs. 2B, C and 4A), are consistent with previous studies (Carayon et al., 2010, McNeely et al., 2011). Our assay was also validated using a well-characterized INBI, FZ41 (Fig. 4B). Additionally, we used our assay to measure the IN DNA binding activities of a small library of natural products, resulting in the identification of a compound termed nigranoic acid as a new INBI (Fig. 5A). The latter compound also inhibited 3′-processing activity, as measured in a standard gel-based 3′-processing assay (Fig. 5B). Nigranoic acid was previously reported to inhibit HIV-1 RT at an IC50 of 159.4μM (Sun et al., 1996) and was also shown to moderately inhibit HIV-1 protease and IN strand transfer (Peng et al., 2010). However, its activity against HIV-1 IN binding to DNA has not previously been characterized. Nigranoic acid moderately inhibits the cytopathic effects of HIV-1 in C8166 cells at EC50 values from 22 to 35μM (Xiao et al., 2006, Sun et al., 2011, Yang et al., 2012), consistent with its anti-HIV-1 IN activity, which is more potent than its anti-RT activity (Fig. 6). Although the exact target site of nigranoic acid on IN is unknown, studies to more fully characterize and optimize this compound and its analogs are in progress.