The negative factor (Nef) is an accessory HIV-1 protein with an extensive cellular interactome and a broad functionality within the HIV replication cycle [1Geyer M, Fackler OT, Peterlin BM. Structure--function relationships in hiv-1 nef EMBO Rep 2001; 2(7 ): 580-., 2Arold ST, Baur AS. Dynamic nef and nef dynamics: How structure could explain the complex activities of this small hiv protein Trends Biochem Sci 2001; 26(6 ): 356-63.]. HIV-1 long-term survivors have been found to harbor viruses with a deficient nef gene, indicating a critical role for Nef in the viral life cycle and in the progression towards AIDS [3Kirchhoff F, Greenough TC, Brettler DB, et al. Brief report.Absence of intact nef sequences in a long-term survivor with nonprogressive hiv-1 infection N Engl J Med 1995; 332(4 ): 228-32., 4Deacon NJ, Tsykin A, Solomon A, et al. Genomic structure of an attenuated quasi species of hiv-1 from a blood transfusion donor and recipients Science Source journal 1995; 270(5238 ): 988-1.]. Nef orchestrates the down-regulation of important surface receptors involved in immune surveillance, such as CD4, MHCI, and MHCII, and activates host cells by triggering signaling pathways involving involving Src family kinases, such as the macrophage-specific hematopoietic cell kinase (Hck) [1Geyer M, Fackler OT, Peterlin BM. Structure--function relationships in hiv-1 nef EMBO Rep 2001; 2(7 ): 580-., 2Arold ST, Baur AS. Dynamic nef and nef dynamics: How structure could explain the complex activities of this small hiv protein Trends Biochem Sci 2001; 26(6 ): 356-63.]. Nef’s functionality is based on numerous protein-protein interactions (PPIs) [2Arold ST, Baur AS. Dynamic nef and nef dynamics: How structure could explain the complex activities of this small hiv protein Trends Biochem Sci 2001; 26(6 ): 356-63., 5Lulf S, Horenkamp FA, Breuer S, Geyer M. Nef surfaces Where to interfere with function. Curr HIV Res 2011; 9(7 ): 543-1.]. The SH3 interaction site on Nef has been identified as a ‘hot spot’ for potential therapeutic intervention due to its highly conserved character [2Arold ST, Baur AS. Dynamic nef and nef dynamics: How structure could explain the complex activities of this small hiv protein Trends Biochem Sci 2001; 26(6 ): 356-63., 5Lulf S, Horenkamp FA, Breuer S, Geyer M. Nef surfaces Where to interfere with function. Curr HIV Res 2011; 9(7 ): 543-1.]. SH3 interacts with a poly-proline type II helix comprising the consensus motif P₇₂xxPxR (HIV-1 NL4-3 nomenclature) and the RT-loop recognition site, a hydrophobic cleft on Nef. The compounds D14, DLC27, and the optimized compound DLC27-14 displayed inhibition of the Nef/SH3 complex formation through targeting of the hydrophobic cleft [6Betzi S, Restouin A, Opi S, et al. Protein protein interaction inhibition (2p2i) combining high throughput and virtual screening: Application to the hiv-1 nef protein Proc Natl Acad Sci USA 2007; 104(49 ): 19256-61., 7Lugari A, Breuer S, Coursindel T, et al. A specific protein disorder catalyzer of hiv-1 nef Bioorg Med Chem 2011; 19(24 ): 7401-6.]. However, all the compounds failed to function in antiviral cell-based assays [6Betzi S, Restouin A, Opi S, et al. Protein protein interaction inhibition (2p2i) combining high throughput and virtual screening: Application to the hiv-1 nef protein Proc Natl Acad Sci USA 2007; 104(49 ): 19256-61., 7Lugari A, Breuer S, Coursindel T, et al. A specific protein disorder catalyzer of hiv-1 nef Bioorg Med Chem 2011; 19(24 ): 7401-6.], thus revealing the need for alternative chemical starting points. To our knowledge, the Nef/SH3 PPI has not been targeted by a biochemical high-throughput screening approach that supports the identification of Nef-specific antivirals.

Here we report on the development and validation of a biochemical time-resolved fluorescence resonance energy transfer (TR-FRET)-based assay for the identification of inhibitors of the Nef/SH3 PPI. The TR-FRET-based assay was paired with an orthogonal biophysical label-free resonant waveguide grating (RWG) assay allowing not only the identification of primary hits that are TR-FRET label-dependent artifacts, but also the confirmation of specifically target-binding compounds.

Fig. (1) shows a schematic of the assay cascade comprising a TR-FRET-based primary screening assay and an orthogonal label-free RWG-based assay. The GST-tagged Hck SH3 domain (GST-SH3) forms a stable complex with the His-tagged Nef protein (His-Nef) in solution (Fig.1A). The affinity of this interaction was quantified previously and found to have a K_d of 0.25 - 1.5 µM [9Breuer S, Schievink SI, Schulte A, et al. Molecular design. functional characterization and structural basis of a protein inhibitor against the hiv-1 pathogenicity factor nef PLoS ONE 2011; 6(5 ): e20033-0., 10Arold S, O'Brien R, Franken P, et al. Rt loop flexibility enhances the specificity of src family sh3 domains for hiv-1 nef Biochemistry 1998; 37(42 ): 14683-91.]. The interaction can be detected homogeneously and therefore automation friendly by TR-FRET between a europium(Eu(III))-conjugated anti-GST monoclonal antibody (α-GST mAb-Eu) and an anti-His XL665-labeled mAb (α-His mAb-XL) in a quaternary complex (Fig.1A). Upon the Eu (III)-cryptate donor excitation at 340 nm, part of the emission at 620 nm activates the acceptor, allophycocyanin XL665, whose emission is detectable at 665 nm. The longevity of the 620 nm emission allows the time-resolved assessment of the 665/620 nm intensity ratio (as a measure of the binding of SH3 to Nef)100 µs after UV excitation of the quaternary complex. Both the large Stokes shift of the Eu(III)-fluorescence and the time resolution of the measurement eliminate most sources of fluorescence interference, making Eu(III) especially well-suited for high-throughput screening of large and diverse compound libraries [11Mathis G. Htrf(r) technology J Biomol Screen 1999; 4(6 ): 309-14., 12Engels IH, Daguia C, Huynh T, et al. A time-resolved fluorescence resonance energy transfer-based assay for den1 peptidase activity Anal Biochem 2009; 390(1 ): 85-7.]. In some cases compounds may have inherent fluorescence that confounds the use of screening using a TR-FRET-based assay. The use of the label-free resonant waveguide grating (RWG) technology for evaluation of fluorescent compounds negates this concern (Fig.1B). The RWG technology utilizes a photonic crystal biosensor surface that reflects broadband light in a narrow range of wavelength (peak wavelength value, or PWV). Any alteration of the mass of an immobilized biomolecule on the biosensor surface, such as through the binding of a compound, alters the dielectric permittivity of the biosensor material, which causes a shift of the reflected wavelength (ΔPWV) proportional to the change in mass [13Cunningham BT, Li P, Schulz S, et al. Label-free assays on the bind system J Biomol Screen 2004; 9(6 ): 481-90.]. Thus, immobilization of the biomolecule on the biosensor surface, followed by the addition of putative chemical or biological binding partners can be monitored kinetically so that one can discriminate binding events on the basis of multiple criteria such as stoichiometry and association kinetics [14Chan LL, Lidstone EA, Finch KE, et al. A method for identifying small-molecule aggregators using photonic crystal biosensor microplates JALA Charlottesv Va 2009; 14(6 ): 348-59.].

Fig. (1A) Principle of a TR-FRET-based Nef/SH3 protein-protein interaction assay. The quaternary complex of His-Nef, GST-SH3, α-GST Eu-labeled mAb, and α-His XL665-labeled mAb generates the proximity for a FRET pair between the donor lanthanide fluorophor europium and acceptor fluorophor XL665. The Eu chelates emit light at 620 nm after excitation at 340 nm and thereby excite neighboring XL665 FRET acceptors. Excited XL665 decays under the emission of photons with a wavelength of 665 nm. The 665 nm emission or the 665/620 nm ratio is proportional to the number of quaternary complexes formed. Excited Eu chelates have a long half-life. The time-resolved detection minimizes interference with extrinsic fluorescence. (B) Schematic of the resonant waveguide grating (RWG)-based assay for monitoring proteinligand or protein-protein interactions in real time. Here, the target protein His-Nef (or off-target GST-SH3) is immobilized on the preactivated aldehyde surface of an optical biosensor whose photonic crystal composition allows the reflection of broadband light only in a narrow range of wavelengths, the peak wavelength value (PWV). Binding of a chemical (compound) or biological (protein) ligand to His-Nef results in a change of the refractory index of the biosensor, measured as a shift of the peak wavelength value (ΔPWV) that is proportional to the change in mass on the surface.

The optimal TR-FRET-based assay conditions were defined in a multi-dimensional approach by varying the protein concentrations of both His-Nef and GST-SH3 and varying the ratios of α-GST mAb-Eu to α-His mAb-XL. Fig. (2A) shows an example of signal development in response to increasing His-Nef and GST-SH3 concentrations at fixed mAb concentrations. As can be seen, the TR-FRET-based assay signal increased in a His-Nef and GST-SH3 concentration-dependent manner, until at high concentrations free His-Nef competed for the binding of detection antibody and blocked the formation of the quaternary complex, resulting in a decrease of the 665/620 nm ratio. Concentrations of 500 nM His-Nef, 10 nM GST-SH3, 0.5 nM α-GST mAb-Eu and 5 nM α-His mAb-XL in a reaction volume of 5 µl in PBS displayed the best signal to background (S/B = 7.5) values with an adequate signal robustness (coefficient of variations < 10%). These assay parameters tolerated dimethyl sulfoxide (DMSO) concentrations of up to 2 % (v/v), and the signal was stable in a time frame of t = 60 – 180 min after antibody addition (data not shown).

Fig. (2A) Optimization of the TR-FRET signal of the quaternary complex. Here, a concentration matrix of His-Nef vs. GST-SH3 was incubated with 0.5 nM Eu-conjugated anti-GST antibody and 5 nM XL-conjugated anti-His6 antibody in a total volume of 5 µl for 2 h at room temperature. (B) TR-FRET-based IC50 analysis of the PPII DLC27. At 500 nM His-Nef and 10 nM GST-SH3 an IC50 of 23.51 ± 1.31 (mean ± s.d.) µM was calculated using a four parameter logistic modelfrom GraphPad Prism 5.0. (C) TR-FRET-based PPI analysis of the Nef mutant P72xxPxR/AxxAxA.10 nM GST-SH3 was incubated with 500 nM wild type Nef, buffer (control) or Nef P72xxPxR/AxxAxA and binding detected using the antibody-based TR-FRET. (D) Scatterplot of a 1536-well LOPAC (Library of Pharmacologically Active Compounds) assay plate. The controls, DMSO (open circles) and DLC27 (open squares), were used to determine the Z’-factor of 0.6 The TR-FRET signal was normalized to the median of the controls.

Based on the optimized assay conditions the potency of a known inhibitor of the Nef/SH3 interaction, DLC27, was evaluated (Fig.2B) [6Betzi S, Restouin A, Opi S, et al. Protein protein interaction inhibition (2p2i) combining high throughput and virtual screening: Application to the hiv-1 nef protein Proc Natl Acad Sci USA 2007; 104(49 ): 19256-61.]. A DLC27 concentration series was evaluated and the IC₅₀ calculated with a four-parameter logistic model from GraphPad Prism 5.0. The IC₅₀ of 23.51 ± 1.31 (mean ± s.d.) µM correlated with previously published data [6Betzi S, Restouin A, Opi S, et al. Protein protein interaction inhibition (2p2i) combining high throughput and virtual screening: Application to the hiv-1 nef protein Proc Natl Acad Sci USA 2007; 104(49 ): 19256-61., 7Lugari A, Breuer S, Coursindel T, et al. A specific protein disorder catalyzer of hiv-1 nef Bioorg Med Chem 2011; 19(24 ): 7401-6.]. As a second proof of concept, we tested a His-Nef mutant lacking the canonical PxxPxR motif: it did not show a TR-FRET-based assay signal (Fig.2C). The feasibility of miniaturization and automation of the TR-FRET-based assay for HTS purposes was examined by testing a 1536-well compound plate from the Library of Pharmacologically Active Compounds (LOPAC, Sigma-Aldrich, St. Louis, MO, USA) (Fig.2D). Individual 50-nl compound aliquots from 1 mM DMSO stocks were transferred using an acoustic dispenser (Labcyte, Sunnyvale, CA, USA) to a well containing the quaternary complex in a 5 µl volume, which was then incubated for 2 h before the TR-FRET-based assay was measured. The Z’ factor [15Zhang JH, Chung TD, Oldenburg KR. A simple statistical parameter for use in evaluation and validation of high throughput screening assays J Biomol Screen 1999; 4(2 ): 67-73.] of this representative test compound plate was calculated to be 0.6. Among the LOPAC compounds, we also placed our positive control DLC27 and D14 as test samples. DLC27 and D14 had been identified as Nef-binding small molecules and inhibitors of the Nef/SH3 protein-protein interaction [6Betzi S, Restouin A, Opi S, et al. Protein protein interaction inhibition (2p2i) combining high throughput and virtual screening: Application to the hiv-1 nef protein Proc Natl Acad Sci USA 2007; 104(49 ): 19256-61.]. Research on DLC27 has continued in a hit-to-lead study, while D14 was not followed up for further antiviral testing [7Lugari A, Breuer S, Coursindel T, et al. A specific protein disorder catalyzer of hiv-1 nef Bioorg Med Chem 2011; 19(24 ): 7401-6.]. Both compounds could be identified as hits and be confirmed in IC50 experiments (Table 1). The robustness and dynamic range of the TR-FRET-based assay allows for both an automated ultra-HTS campaign and ranking of primary hits according to their IC₅₀ values.

Table 1.

Summary of Biochemical and Biophysical Screening Data

To exclude compound artifacts due to fluorescence interference or inhibition of the Nef/SH3 interaction by the virtue of colloidal compound aggregate formation [16McGovern SL, Caselli E, Grigorieff N, Shoichet BK. A common mechanism underlying promiscuous inhibitors from virtual and high-throughput screening J Med Chem 2002; 45(8 ): 1712-22.], we paired the primary TR-FRET-based assay, in an orthogonal approach, with a label-free RWG-based assay format using a 384-well plate setup (Fig.1B). Using the tool compounds DLC27 and D14 for follow up studies, we employed RWG-based competition assays as well as RWG-based direct binding assays to evaluate hits regarding their protein-protein interaction inhibition and direct target binding properties [17Geschwindner S, Carlsson JF, Knecht W. Application of optical biosensors in small-molecule screening activities Sensors (Basel) 2012; 12(4 ): 4311-23.]. Fig. (3A) shows the binding kinetics of His-Nef to GST-SH3 immobilized on an aldehyde-coated biosensor plate in the presence of 50 µM DLC27 or a control (DMSO). Only the control sample shows the expected shift of the reflected wavelength (ΔPWV), confirming that DLC27 disrupts the interaction of Nef and SH3. To further investigate the binding specificity of compounds to the protein target, we immobilized the viral Nef protein or the human SH3 domain on the biosensor surface and then treated each with DLC27 or D14 (Fig.3Band C). The real-time binding kinetics of DLC27 or D14 to immobilized His-Nef or GST-SH3 demonstrated that DLC27 selectively bound to Nef, whereas D14 produced a positive PWV shift with both His-Nef and GST-SH3 (Fig.3B, Cand Table 1). These findings could indicate that D14 shares a binding site between His-Nef and GST-SH3. However, given that the magnitude of the PWV shifts caused by D14 exceeded what can be expected from either the molecular mass or the stoichiometry of compounds, it appears that D14 is a promiscuous compound that disrupts Nef/SH3 interaction.

Fig. (3A) RWG-based competition assay: Inhibition of His- Nef/GST-SH3 complex formation by DLC27. The binding of 150 nM His-Nef to 5 µM GST-SH3 immobilized to the biosensor surface was monitored by RWG in the presence of 1 % (v/v) DMSO or 50 µM DLC27 in PBS in real-time. (B) RWG-based direct binding assay: Binding kinetics of D14 and DLC27 to His-Nef. The peak wavelength shift (ΔPWV) was assessed after the addition of 50µM DLC27 or D14 in 1 % (v/v) DMSO in PBS to 5 µM immobilized His-Nef in real-time. (C) Binding kinetics of D14 and DLC27 to GST-SH3. The !PWV was assessed after the addition of 50 µM DLC27 or D14 in 1% (v/v) DMSO in PBS to 5 µM immobilized GST-SH3 in real-time.

In summary, we have developed a robust and HTS-compatible assay comprising a primary TR-FRET-based assay for identifying inhibitors of Nef/SH3 complex formation. Taking advantage of the emergence of HTS-amenable biophysical methodologies, we combined the TR-FRET-based assay with an orthogonal biophysical label-free RWG-based assay, the HTS resolving power is extended, by allowing screening of fluorescent compounds, as well as the identification of Nef/SH3 promiscuous binders. In combining these two distinct methodologies into a high-throughput assay cascade, novel Nef/SH3 inhibitors can be rapidly selected and ranked from large chemical libraries by their inhibitory activity and specific binding early in a drug discovery campaign (Table1).

The authors confirm that this article content has no conflicts of interest.