Article pubs.acs.org/ac
Screening of Small-Molecule Inhibitors of Protein−Protein Interaction with Capillary Electrophoresis Frontal Analysis Mei Xu,† Chao Liu,† Mi Zhou,† Qing Li,† Renxiao Wang,*,† and Jingwu Kang*,†,‡ †
State Key Laboratory of Bioorganic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Lingling Road 345, Shanghai 200032, China ‡ ShanghaiTech University, Yueyang Road 319, Shanghai 200031, China S Supporting Information *
ABSTRACT: A simple and effective method for identifying inhibitors of protein−protein interactions (PPIs) was developed by using capillary electrophoresis frontal analysis (CE-FA). Antiapoptotic B-cell-2 (Bcl-2) family member Bcl-XL protein, a 5-carboxyfluorescein labeled peptide truncated from the BH3 domain of Bid (F-Bid) as the ligand, and a known Bcl-XL-Bid interaction inhibitor ABT-263 were employed as an experimental model for the proof of concept. In CE-FA, the free ligand is separated from the protein and protein−ligand complex to permit the measurement of the equilibrium concentration of the ligand, hence the dissociation constant of the protein−ligand complex. In the presence of inhibitors, formation of the protein−ligand complex is hindered, thereby the inhibition can be easily identified by the raised plateau height of the ligand and the decayed plateau of the complex. Further, we proposed an equation used to convert the IC50 value into the inhibition constant Ki value, which is more useful than the former for comparison. In addition, the sample pooling strategy was employed to improve the screening throughput more than 10 times. A small chemical library composed of synthetic compounds and natural extracts were screened with the method, two natural products, namely, demethylzeylasteral and celastrol, were identified as new inhibitors to block the Bcl-XL-Bid interaction. Cell-based assay was performed to validate the activity of the identified compounds. The result demonstrated that CE-FA represents a straightforward and robust technique for screening of PPI inhibitors.
P
high-throughput screening (HTS) faces an enormous challenge. Most currently available chemical libraries are not amenable to screening of PPI inhibitor, because the contact surfaces of PPIs is large and devoid of well-defined cavities for accommodating a small molecule.14,15 Alternatively, fragment-based drug discovery (FBDD) is proved to be more amenable to this purpose.15 Moreover, natural products and natural-product-like molecules are considered as a promising class of chemicals for screening of PPI inhibitors, since biology has evolved these molecules for binding at protein interfaces.7 Thus far, various techniques, such as fluorescence polarization (FP), fluorescence resonance energy transfer (FRET), and nuclear magnetic resonance (NMR), as well as surface plasmon resonance (SPR), are commonly used to identify and characterize small molecule inhibitors of PPIs.3,15 Each technology has its own strength and weakness. One weakness of these techniques is that they suffer from high false positive rates.16 Capillary electrophoresis (CE) is not only a powerful separation technique but also a versatile platform for studying the molecular recognition interactions, including protein−small
rotein−protein interactions (PPIs) play a critical role in most biological processes, such as metabolism, structural organization, signal transduction, immune recognition, and gene transcription.1,2 Dysregulated PPIs are often associated with various diseases, such as cancer and Alzheimer’s disease.3,4 Therefore, pathologically relevant PPIs are often considered as a vast class of therapeutic targets.5−7 For example, the Bcl-2 family of proteins are a major group of apoptosis regulators, which includes both antiapoptotic members (such as Bcl-2, Bcl-XL, Mcl-1, and Bcl-w), pro-apoptotic members (such as Bak, Bax, and Bid).8 The major gateway to cell apoptosis is guarded by Bcl2 and its antiapoptotic members. Overexpression of antiapoptotic Bcl-2 family proteins enables cancer cells to evade apoptosis leading to the resistance to chemotherapy and radio therapy.9,10 Therefore, Bcl-2 family proteins have been considered as attractive molecular targets for developing new cancer therapies since the beginning of this century.11,12 To date, more than 40 PPIs have been targeted; however, only a few compounds targeting PPIs have entered clinical trials owing to the challenge associated with screening inhibitors of PPIs.7,13,14 Therefore, more effective strategies for screening PPI inhibitors are badly needed. Unlike the traditional targets, i.e., enzymes and small-molecule receptors, discovering small-molecule inhibitors of PPIs with © 2016 American Chemical Society
Received: April 12, 2016 Accepted: July 17, 2016 Published: July 18, 2016 8050
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Article
Analytical Chemistry
fine powder and was ultrasonically extracted with 70% (v/v) ethanol solution for three times. After filtration and removal of the solvent by rotary evaporation, the natural extracts were obtained. The sample of ABT-263 was purchased from Selleckchem (Houston, DE). All solutions were prepared with ultrapure water purified by a Milli-Q water purification system (Millipore, Milford, MA). Instrumentation. All CE experiments were performed on a P/ACE MDQ CE system equipped with a laser-induced fluorescence (LIF) detector (Beckman Coulter, CA). A semiconductor laser was used as an excitation source (488 nm), and the electropherograms were recorded by monitoring the emission of fluorescence at 520 nm. The fused silica capillaries with a dimension of 50 μm i.d. (370 μm o.d.) × 40 cm (29.5 cm to the detection window) were purchased from Sumtech (Hebei, China). Capillary Electrophoresis Frontal Analysis. A new capillary was conditioned with 0.1 M NaOH solution for 30 min, followed by flushing with deionized water for 5 min. The running buffer for CE-FA experiments was composed of 30 mM sodium phosphate buffer (pH 7.5) containing 50 mM NaCl. Stock solutions of Bcl-XL (10 μM) and F-Bid (500 μM) were prepared with 30 mM phosphate buffer (pH 7.5) containing 50 mM NaCl and 0.015% Brij-35(v/v). The stock solutions of ABT263 in 10 mM were prepared by dissolving a certain amount of ABT-263 in DMSO, then diluting with 30 mM phosphate buffer (pH 7.5) containing 50 mM NaCl and 0.015% Brij-35(v/v). A series of ABT-263 solutions were obtained by diluting the stock solution with the above-mentioned buffer. In order to determine the dissociation constant Kd of the complex formed by Bcl-XL and F-Bid, a series of mixtures of FBid and Bcl-XL were prepared by varying the concentrations of the former from 50 to 350 nM and holding the Bcl-XL concentration constant at 200 nM. For inhibition assays, a series of mixtures of F-Bid, Bcl-XL, and the test compounds were prepared by keeping the concentrations of Bcl-XL and F-Bid constant. These mixtures were incubated for at least 15 min in the sample tray at the ambient temperature. Samples were injected by pressure at 55 mbar for 40 s unless noted otherwise. Separations were performed by applying a voltage of 15 kV unless noted otherwise. The capillary temperature was set at 25 °C. Between runs, the capillary was flushed sequentially with 0.1 M NaOH, deionized water, and running buffer for 1 min to improve the repeatability of the measurement. Improvement of the Screening Throughput by Sample Pooling. For improving the screening throughput, the sample pooling strategy was adopted. The stock solutions of all 60 synthesized compounds were prepared by dissolving a certain amount of compound in DMSO to give a concentration of ∼4 mM. These samples were randomly divided into six groups. In the same group, the solutions of each compound were pooled together into a sample vial and then diluted with 30 mM phosphate buffer (pH 7.5) containing 50 mM NaCl and 0.015% Brij 35 (v/v). For the inhibition assay, the pooled samples were mixed with F-Bid and Bcl-XL yielding a final concentration of 10 μM for each compound. The final DMSO concentration must be less than 3% (v/v) in the assay solution. Natural extract represents a complex mixture containing various natural products; therefore, natural extract is inherently a sample pooling. Determination of Inhibition Percentage. In CE-FA, the inhibition percentage is defined as
molecule drug,17,18 protein−peptide,19 protein−protein,20 protein−DNA,21−23 protein−aptamer,24−27 protein aggregation,28 and antibody−antigen.29 However, the works concerning the screening of PPI inhibitors by using CE are very few. Most recently, Rauch et al. demonstrated that capillary zone electrophoresis (CZE) based affinity method can be an effective platform for screening of PPI inhibitors.30 Compared to other techniques, the CZE method displayed several advantages, such as having better specificity to distinguish the false positive results owing to the CE separation.30 Capillary electrophoresis frontal analysis (CE-FA) is often used to measure the dissociation constant of protein−ligand interactions.31,32 In CE-FA, a large volume of pre-equilibrated sample is injected into the capillary to maintain the steady-state equilibrium conditions in the process of separation because most of the ligand and protein zones are overlapped. The elution profile of the free ligand is a plateau peak which allows the determination of free ligand concentration from the height of the plateau.31,32 Compared to the CZE based affinity method,30 CEFA is more robust because the plateau height and the corresponding concentration is not affected by changes in migration times, EOF, length of capillary, or applied voltage.31 Moreover, CE-FA is able to measure the interactions having rapid dissociation rate and low affinity.33 Therefore, CE-FA is a more versatile technique for identifying PPI inhibitors. The aim of this study is to develop a CE-FA method for screening of PPI inhibitors and apply the method to screen PPI inhibitors in natural products. A recombinant Bcl-XL protein and a peptide truncated from the BH3 domain of the Bid protein (26mer) were employed as a proof of concept. The established method was validated by using a known Bcl-XL inhibitor ABT263. In addition, an equation was proposed so that the measured IC50 values can be converted into the inhibition constant Ki values. Finally, a small chemical library consisting of 60 synthesized compounds and 45 natural extracts from the Chinese traditional herbs were screened by the method. Demethylzeylasteral and celastrol were identified to be the new inhibitors of Bcl-XL-Bid interaction.
■
EXPERIMENTAL SECTION Chemicals and Reagents. Recombinant human Bcl-XL protein was expressed and purified according to the protocols described in the literature.34,35 The protein has 209 amino acid residues plus a His tag at the N terminal. The purified protein was dialyzed against 30 mM phosphate buffer (pH 7.5) containing 1 mM EDTA and 50 mM NaCl. The protein concentration was determined as ∼0.65 mM by using the protein assay kits (BioRad, Hercules, CA). The protein solution was frozen at −80 °C before use. The fluorescently labeled Bid-BH3 peptide (FQEDIIRNIARHLAQVGDSMDRSIPPG) was ordered from GL Biochem (Shanghai, China). Because the peptide was fluorescently labeled and its amino acid sequence was derived from the BH3 domain of Bid, it was denoted as F-Bid. Brij 35 was purchased from Sigma (St. Louis, MO). Dimethyl sulfoxide (DMSO), sodium chloride (NaCl), and sodium hydroxide (NaOH) were obtained from Shanghai Chemical Reagent (Shanghai, China). Sodium phosphate monobasic (anhydrous) was obtained from Sinopharm (Shanghai, China). All the synthesized compounds were purchased from Specs (Zoetermeer, The Netherlands). A total of 45 herbs were purchased from Shifeng Biology (Shanghai, China) and Shanghai Pureone Biotechnology (Shanghai, China) (see Table S1 in the Supporting Information). Each herb sample was ground into a 8051
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Article
Analytical Chemistry
Figure 1. Chemical structures of F-Bid, ABT-263, demethylzeylasteral, and celastrol.
inhibition% =
H − H0 × 100 H100 − H0
(1)
where H represents the plateau height of the free ligand at any inhibitor concentrations; H0 is the plateau height at 0% inhibition; H100 is the plateau height obtained at 100% inhibition, i.e., the maximum plateau height obtained in the absence of protein. Determination of the Dissociation and the Inhibition Constants. When ligand L binds to protein P to form a complex PL with a 1:1 stoichiometric ratio, the binding event can be described by eq 2: L + P ⇋ PL
1 [PL] = Kd [L][P]
[PL] [PL] [L] = = [P]t [P] + [PL] Kd + [L]
[P][I] [PI]
(9)
[PL]0 [PL]0 [P]0 [L]0 /Kd = = [P]t [PL]0 + [P]0 [P]0 [L]0 /Kd + [P]0 =
(3)
[L]0 [L]0 + Kd
(10)
By rearrangement: [PL]0 =
[L]0 [P]t [L]0 + Kd
(11)
where [L]0, [P]0, and [PL]0 represent the equilibrium concentrations of the ligand, protein, and protein−ligand complex, respectively, in the absence of inhibitor in the system. When 50% inhibition is achieved, one has
(4)
[PL]50 [PL]50 = [P]t [PL]50 + [P]50 + [PI]50 =
[L]50
[L]50 + Kd + Kd[I]50 /K i
(12)
By rearrangement, [PL]50 =
(5)
[L]50
[L]50 [P]t + Kd + Kd[I]50 /K i
(13)
where [L]50, [P]50, [I]50, [PL]50, and [PI]50 represent the equilibrium concentrations of the ligand, protein, inhibitor, protein−ligand complex, and protein−inhibitor complex, respectively, when 50% inhibition is achieved.
Conservation of mass requires that [L]t = [L] + [PL]
(8)
When inhibitor is absent in the system, one has
(2)
where [P]t represents the total concentration of the protein. The r value is calculated by the ratio between [PL] and [P]t; while, [PL] is obtained by subtracting [L] from [L]t. Plotting r as a function of [L] yields the binding curve. The dissociation constant Kd can be computed by the nonlinear curve fitting using the Sigmaplot 12.0 (Systat Software, CA). When a competitive inhibitor I is present in the system, a binary equilibrium system can be formed and the system is described by eq 5: P + L + I ⇋ PL + PI
[I]t = [I] + [PI]
Ki =
where [P], [L], and [PL] represent the equilibrium concentrations of the protein, ligand, and complex, respectively. Because only [L] can be directly measured in the CE-FA experiment, the ligand−protein binding ratio r is used to construct the binding curve based on eq 4 r=
(7)
where [L]t, [I]t represent the total concentrations of the ligand and inhibitor, [I] and [PI] represent the equilibrium concentrations of the inhibitor and protein−inhibitor complex, respectively. The inhibition constant Ki of the inhibitor are expressed as
The association constant Ka or the dissociation constant Kd can be described by eq 3: Ka =
[P]t = [P] + [PL] + [PI]
(6) 8052
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Article
Analytical Chemistry At 50% inhibition, 50% of protein−ligand complex will dissociate, so one has [PL]0 = 2[PL]50
(14)
Substitute eq 11 and eq 13 into eq 14, one has 2[L]50 [L]0 = [L]0 + Kd [L]50 + Kd + Kd[I]50 /K i
(15)
By rearrangement, Ki =
[I]50 [L]50 /Kd + 2[L]50 /[L]0 − 1
(16)
[I]50 can be computed according to eq 17. The detailed process for deriving eq 17 and the equation for computation of [L]50 is shown in the Supporting Information. ⎛ [L]t ⎞ − 1⎟ + [L]t − [L]50 [I]50 = IC50 − [P]t + Kd⎜ ⎝ [L]50 ⎠ (17)
where IC50 is the total concentration of the inhibitor when 50% inhibition is achieved. Finally, the IC50 value can be converted into the Ki value by using eq 16. Cell-Based Assay. Cell apoptosis assays were performed with HT-29 cells by using an annexin V-FITC apoptosis detection kit (KeyGEN Biotech, China). After incubation in culture plates for 12 h, cells were exposed to 0.1% DMSO (control), 1 μM demethylzeylasteral, or 1 μM celastrol for 12 h at 37 °C. After washing twice with PBS, cells were resuspended in the annexinV-binding buffer and incubated with annexinVFITC/PI in the dark for 15 min according to the manufacturer’s instructions. Subsequently, the cells were analyzed by flow cytometry. The data were processed by using FloMax software.
Figure 2. (A) Electropherograms illustrating the formation of the complex between F-Bid and Bcl-XL, (B) binding curve of F-Bid and BclXL. Conditions: capillary, 50 μm i.d. (370 μm o.d.) × 40 cm (effective length, 29.5 cm); background electrolyte, 30 mM phosphate buffer (pH 7.5) containing 50 mM NaCl; applied voltage, +15 kV; column temperature, 25 °C. Samples were injected at 55 mbar for 40 s. Peaks: 1 = free F-Bid; 2 = F-Bid-Bcl-XL complex. In the binding curve, the concentration of Bcl-XL was fixed at 200 nM and the concentration of FBid varied from 50 to 350 nM.
■
RESULTS AND DISCUSSION Method Development and Validation. The recombiant Bcl-XL and fluorescently labeled peptide F-Bid were employed to develop the method. The structure of F-Bid was shown in Figure 1. In our initial experiment, the detection of F-Bid suffered from an erratic fluorescent intensity. We found that the fluorescent intensity was stabilized and enhanced significantly by addition of 0.015% (v/v) Brij35 in the solutions of protein and F-Bid. Because Brij35 is a nonionic surfactant, it may prevent the fluorescence molecules from quenching by the dissolved oxygen in the solution. Moreover, addition of 50 mM NaCl in the BGE is necessary to minimize the adsorption of the protein and peptide onto the capillary wall.36 Under the LIF detection, besides ligand F-Bid, the ligand− protein complex can also be detectable. As shown in Figure 2A, the LIF traces of the CE-FA experiments display two adjacent plateau peaks corresponding to free F-Bid and the ligand− protein complex, respectively. Also in Figure 2A, the plateau height of the free F-Bid decreased, but that of the complex increased with the increase of the protein−ligand ratio indicating that the protein−ligand complex formed dose-dependently. In CE-FA, injection of a large volume of sample solution is necessary to keep the equilibrium in the process of CE-FA separation.31 That is the reason why the plateau peak shape is always obtained in CE-FA. We investigated the effect of injection time on the formation of the plateau peak over range from 10 to 80 s. As shown in Figure S1 (Supporting Information), an ideal
plateau was obtained when the injection time lasted for 40 s. The effect of the applied voltage on the separation was also investigated over the range from 10 to 20 kV. Ideally, the zone of the free ligand should be partially separated from that of the complex. As shown in Figure S2, applying 15 kV voltage allowed a suitable separation between two zones. Subsequently, the effect of the incubation temperature and the time scale on the formation of the complex was investigated. We found that the binding of F-Bid to Bcl-XL was not sensitive to the incubation temperature ranging from 20 to 37 °C (Figure S3); therefore, the incubation can be carried out at room temperature (about 25 °C). It was observed that the plateau peak height of free F-Bid kept constant when the incubation time was over 15 min until 200 min (Figure S4). That means the equilibrium had been attained after incubation for 15 min. Therefore, all samples were incubated at least for 15 min before injection. Finally, the repeatability of the method was evaluated in terms of the plateau height. The RSD for the intraday (n = 6) repeatability was measured as 2.7%; while the RSD for interday 8053
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Article
Analytical Chemistry
Figure 5. Dose−response inhibition curves of demethylzeylasteral (A) and celastrol (B). The concentrations of Bcl-XL and F-Bid were fixed at 0.1 μM. Other conditions as in Figure 2.
Figure 3. (A) Electropherograms illustrating the inhibition of binding of F-Bid to Bcl-XL by ABT-263, (B) the dose−response inhibition curve of ABT-263.The concentrations of Bcl-XL and F-Bid were fixed at 0.15 μM and 0.1 μM, respectively. Other conditions as in Figure 2.
significant difference, the approach for measuring the plateau height at the beginning part of the plateau is more straightforward. Figure 2B shows the binding curve which was constructed by plotting the ligand−protein binding ratio r against the free concentrations of F-Bid. Each concentration point was measured in triplicate, and the average value was used to construct the curve. Fitting the data with nonlinear regression gives the dissociation constant Kd as 33 nM. This result is in good agreement with the value (Kd = 41 nM) measured by our fluorescence polarization-based assay using the same recombination Bcl-XL protein and F-Bid.37 Determination of the Competitive Inhibition Constant. The chemical structure of the known inhibitor ABT-263 is shown in Figure 1. The dose−response inhibition curve of ABT-263 was constructed by varying the inhibitor concentrations from 1 × 10−9 to 1 × 10−4 M, meanwhile fixing the concentrations of F-Bid and Bcl-XL at 1 × 10−7 M and 1.5 × 10−7 M, respectively (Figure 3B). Each data was measured in triplicate, and the average values were used to construct the curve. With the established inhibition curve, the half maximal inhibitory concentration IC50 of ABT263 was measured as 220 nM. Although IC50 is commonly used as a measurement of potency of antagonist drug in pharmacological research, it is difficult to compare the IC50 values measured under different experimental conditions and from different laboratories, because it depends on the experimental conditions. For example, the IC50 value increases with increase of the concentrations of protein or ligand. On the contrary, Ki belongs to the equilibrium constant which is theoretically independent of the experimental conditions except the temperature.38 Therefore, the inhibition constants Ki is recommended to measure the potency of the inhibitors. In enzyme inhibition studies, the Cheng−Prusoff equation has been commonly used to convert the IC50 value into the Ki
Figure 4. Electropherograms illustrating the identification of inhibitor in natural extract. The concentrations of Bcl-XL and F-Bid were fixed at 0.1 μM. Other conditions as in Figure 2.
(n = 3) repeatability was measured as 4.8%. That means the method has a satisfactory repeatability. Determination of the Dissociation Constant of the Protein−Ligand Complex. In order to determine the concentration of free F-Bid, a calibration curve was constructed over the concentration range from 2 × 10−9 to 1 × 10−6 M getting a linear equation: y = 4.13 × 107x − 0.095 (R2 = 0.999). In our experiment, the plateau height at the beginning part of the plateau was used for quantitative analysis rather than using the average height over the plateau. We compared the difference in terms of the repeatability. In the same experiment, the RSD for the average height over the plateau was measured as 1.2%, whereas the RSD for the height at the beginning part of the plateau was measured as 1.4% (n = 3). Although there is no 8054
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Article
Analytical Chemistry
protein−ligand affinity is based on the degree of polarization of the fluorescence, which is inversely proportional to the rotation rate of the labeled ligand. Limited by the principle of FP, the concentration of the free ligand cannot be directly measured. Therefore, a complex quadratic equation should be solved to calculate the free concentration of protein at 0% inhibition. However, in CE-FA the concentration of free ligand can be measured accurately, which will simplify the approach for computation of the inhibition constant Ki value. To this end, we proposed a new equation, which is more simple and straightforward. Among the commonly used techniques for studying molecule−molecule interactions, very few methods are able to directly measure the concentration of the free ligand. In this aspect, CE-FA has its unique advantage in PPI studies and screening of PPI inhibitors. With eq 16 and eq 17, the Ki value of ABT-263 was measured as 22 nM, which is comparable to 8 nM that we measured with the FP-based method. The different data should be due to the different measurement principle. Screening Data Evaluation. The data quality and robustness of the method were evaluated by the Z′ factor, which is measured by using the following equation.40 Z′ = 1 −
3σpositive + 3σnegative μpositive − μnegative
where σpositive and σnegative are the standard deviations of the positive and negative signals, respectively; μpositive and μnegative are the averages of the positive and negative signals, respectively. The data statistics were shown in Figure S5. The Z′ factor was determined as 0.86, implying a satisfactory quality. Inhibitor Screening. Before screening the chemical library, the effect of the DMSO in sample solutions on the complex formation had to be evaluated to avoid the artificial interference. As shown in Figure S6, the influence of DMSO can be neglected when the DMSO content in sample solutions is less than 3% (v/ v). Further, the stability of plateau height of free peptide in the process of screening was also evaluated. In the course of 100 successive injections, the plateau heights maintained a reasonable stability (Figure S7). A sample pooling strategy was applied to improve the screening throughput. As described in the Experimental Section, with the sample pooling strategy, a mixed sample containing 10 compounds was assayed in one run. That means the screening throughput is improved more than 10 times. In the chemical library, only the Tripterygium extract was identified to be active. As shown in Figure 4, compared to the negative control (no inhibition), increased plateau height of free F-Bid and decayed plateau of the ligand−protein complex were observed when the extract of Tripterygium was assayed. Subsequently, eight commercially available ingredients of Tripterygium were assayed individually at the concentration level of 10 μM. However, only demethylzeylasteral and celastrol were figured out to be the inhibitors for blocking the interaction between Bcl-XL and F-Bid (also in Figure 4). The structure of demethylzeylasteral and celastrol were shown in Figure 1. The dose-dependent inhibition curve for demethylzeylasteral and celastrol were constructed to verify the results (Figure 5). For demethylzeylasteral and celastrol, the IC 50 values were determined as 7.2 μM and 18.2 μM, respectively. Further, the Ki values for demethylzeylasteral and celastrol were computed as 1.5 μM and 3.8 μM, respectively. Cell-based assay was performed to validate the bioactivity of demethylzeylasteral and celastrol. The effect of demethylzeylas-
Figure 6. Cell-based assays showing promotion of cell apoptosis by treating the cells with demethylzeylasteral or celastrol. HT-29 cells were treated with 0.1% DMSO (A), 1 μM demethylzeylasteral (B), and 1 μM celastrol (C) for 12 h. Cell apoptosis was determined by annexin VFITC/PI staining assay.
value.39 However, the Cheng−Prusoff equation is specially designed for enzymatic inhibition but not suitable for PPI inhibition study. This is because, in the enzyme inhibition study, the substrate concentration is much higher than the concentration of enzyme, so that the change of the substrate concentration can be negligible. While this is not the truth in the PPI inhibition study. Therefore, Wang et al. has derived a new mathematical equation used for computation of the Ki value of PPI inhibitors.39 In their FP method, the measurement of the 8055
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Analytical Chemistry
■
ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundations of China (Grants 21375140 and 21175146) and the National Key Laboratory of Organic Biochemistry Opening Foundations.
teral and celastrolwere on promoting cell apoptosis was assayed by using annexin V-FITC/propidium iodide (PI) double staining assay. Human colorectal cancer cell line, HT29 cells, was used as the cell model. As shown in Figure 6, after treating the cells with 1 μM demethylzeylasteral or celastrol for 12 h, the percentages of apoptotic cells were determined as 8.2% and 14.3%, respectively, by flow cytometry. The effect of demethylzeylasteral and celastrol on cell apoptosis is significant compared to 2.5% apoptotic cells in the control assay. The data indicate that demethylzeylasteral and celastrol have the activity to promote cell apoptosis. Demethylzeylasteral and celastrol are the ingredients of the Traditional Chinese Medicine Leigong teng (Tripterygium glycosides), which is used to the treatment of a broad spectrum of autoimmune and inflammatory diseases in China. Celastrol has been reported to induce apoptosis through activation of numbers of pathways.41,42 However, its mechanism of action, to the best of our knowledge, has never been reported before.
■
CONCLUSIONS The CE-FA-based method used for screening of small-molecule inhibitors of PPIs has been established and validated. Utilizing the high sensitive LIF detection greatly facilitates the application of CE-FA for identifying PPI inhibitors. Meanwhile, sample pooling strategy is the necessary consideration to overcome the inherent shortcoming of CE concerning the low throughput. By using the sample pooling strategy, the screening throughput can be increased at least 10 times. Compared with the alternative methods used for screening of PPI inhibitors, the great advantage of our method is that the CE-FA technique allows the equilibrium concentration of the free ligand to be measured straightforwardly. That greatly facilitates the determination of dissociation constant Kd of the complex and the inhibition constant Ki. Moreover, the method displays advantage, such as minimum consumption of sample and reagent. The merit of the method is demonstrated by the discovery of two natural products, namely, demethylzeylasteral and celastrol as the new inhibitors of Bcl-XL protein. The method may be used as a general method for screening of PPI inhibitors. ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b01430. Effect of the injection time on the formation of the plateau; effect of applied voltage on the separation of F-Bid and FBid-Bcl-XL complex; effect of incubation temperature on F-Bid-Bcl-XL binding; dependence of F-Bid-Bcl-XL binding on incubation time; effect of DMSO concentrations on blocking F-Bid-Bcl-XL binding; statistical evaluation of the screening data; stability data for 100 successive injections; and detailed calculation of Ki and compound library for screening (PDF)
■
REFERENCES
(1) Songyang, Z.; Cantley, L. C. Trends Biochem. Sci. 1995, 20, 470. (2) Pawson, T.; Gish, G. D.; Nash, P. Trends Cell Biol. 2001, 11, 504. (3) Milroy, L.-G.; Grossmann, T. N.; Hennig, S.; Brunsveld, L.; Ottmann, C. Chem. Rev. 2014, 114, 4695−4748. (4) Vidal, M.; Cusick, M. E.; Barabasi, A. L. Cell 2011, 144, 986−998. (5) Fry, D. C.; Vassilev, L. T. J. Mol. Med. (Heidelberg, Ger.) 2005, 83, 955−963. (6) White, A. W.; Westwell, A. D.; Brahemi, G. Expert Rev. Mol. Med. 2008, 10, e8. (7) Arkin, M. R.; Tang, Y.; Wells, J. A. Chem. Biol. 2014, 21, 1102− 1114. (8) Purtilo, D. T.; Grierson, H. L. Cancer Genet. Cytogenet. 1991, 51, 143. (9) Youle, R. J.; Strasser, A. Nat. Rev. Mol. Cell Biol. 2008, 9, 47−59. (10) Lessene, G.; Czabotar, P. E.; Sleebs, B. E.; Zobel, K.; Lowes, K. N.; Adams, J. M.; Baell, J. B.; Colman, P. M.; Deshayes, K.; Fairbrother, W. J.; Flygare, J. A.; Gibbons, P.; Kersten, W. J.; Kulasegaram, S.; Moss, R. M.; Parisot, J. P.; Smith, B. J.; Street, I. P.; Yang, H.; Huang, D. C.; Watson, K. G. Nat. Chem. Biol. 2013, 9, 390−397. (11) Soriano, P.; Montgomery, C.; Geske, R.; Bradley, A. Cell 1991, 64, 693. (12) Sawyer, T. K. Biopolymers 1998, 47, 243. (13) Mullard, A. Nat. Rev. Drug Discovery 2012, 11, 173−175. (14) Wells, J. A.; McClendon, C. L. Nature 2007, 450, 1001−1009. (15) Meireles, L. M.; Mustata, G. Curr. Top. Med. Chem. 2011, 11, 248−257. (16) Curpan, R. F.; Simons, P. C.; Zhai, D.; Young, S. M.; Carter, M. B.; Bologa, C. G.; Oprea, T. I.; Satterthwait, A. C.; Reed, J. C.; Edwards, B. S.; Sklar, L. A. Assay Drug Dev. Technol. 2011, 9, 465−474. (17) Riesová, M.; Svobodová, J.; Ušelová, K.; Tošner, Z.; Zusková, I.; Gaš, B. J. Chromatogr. A 2014, 1364, 276−288. (18) Zinellu, A.; Sotgia, S.; Scanu, B.; Pisanu, E.; Giordo, R.; Cossu, A.; Posadino, A. M.; Carru, C.; Pintus, G. J. Chromatogr. A 2014, 1367, 167−171. (19) Sun, Y.; Cressman, S.; Fang, N.; Cullis, P. R.; Chen, D. D. Y. Anal. Chem. 2008, 80, 3105−3111. (20) Shimura, K.; Waki, T.; Okada, M.; Toda, T.; Kimoto, I.; Kasai, K.i. Electrophoresis 2006, 27, 1886−1894. (21) Berezovski, M.; Krylov, S. N. Anal. Chem. 2005, 77, 1526−1529. (22) Stebbins, M. A.; Hoyt, A. M.; Sepaniak, M. J.; Hurlburt, B. K. J. Chromatogr., Biomed. Appl. 1996, 683, 77−84. (23) Berezovski, M.; Krylov, S. N. J. Am. Chem. Soc. 2003, 125, 13451− 13454. (24) Berezovski, M.; Nutiu, R.; Li, Y.; Krylov, S. N. Anal. Chem. 2003, 75, 1382−1386. (25) German, I.; Buchanan, D. D.; Kennedy, R. T. Anal. Chem. 1998, 70, 4540−4545. (26) Buchanan, D. D.; Jameson, E. E.; Perlette, J.; Malik, A.; Kennedy, R. T. Electrophoresis 2003, 24, 1375−1382. (27) Zhang, H.; Li, X.-F.; Le, X. C. J. Am. Chem. Soc. 2008, 130, 34−35. (28) Picou, R. A.; Schrum, D. P.; Ku, G.; Cerqua, R. A.; Kheterpal, I.; Gilman, S. D. Anal. Biochem. 2012, 425, 104−112. (29) Heegaard, N. H. H. J. Chromatogr. A 1994, 680, 405−412. (30) Rauch, J. N.; Nie, J.; Buchholz, T. J.; Gestwicki, J. E.; Kennedy, R. T. Anal. Chem. 2013, 85, 9824−9831. (31) Østergaard, J.; Heegaard, N. H. H. Electrophoresis 2003, 24, 2903−2913. (32) Busch, M. H. A.; Carels, L. B.; Boelens, H. F. M.; Kraak, J. C.; Poppe, H. J. Chromatogr. A 1997, 777, 311−328. (33) Heegaard, N. H.; Nilsson, S.; Guzman, N. A. J. Chromatogr., Biomed. Appl. 1998, 715, 29−54.
■
■
Article
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. Phone: 86-21-54925128. *E-mail:
[email protected]. Phone: 0086-21-54925385. Fax: 0086-21-54925481. Notes
The authors declare no competing financial interest. 8056
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
Article
Analytical Chemistry (34) Zhou, B.; Li, X.; Li, Y.; Xu, Y.; Zhang, Z.; Zhou, M.; Zhang, X.; Liu, Z.; Zhou, J.; Cao, C.; Yu, B.; Wang, R. ChemMedChem 2011, 6, 904− 921. (35) Rampler, E.; Stranzl, T.; Orban-Nemeth, Z.; Hollenstein, D. M.; Hudecz, O.; Schloegehofer, P.; Mechtler, K. J. Proteome Res. 2015, 14, 5048−5062. (36) Liu, C.; Kang, J. J. Chromatogr. A 2012, 1238, 146−151. (37) Yang, C. W.; Chen, S.; Zhou, M.; Li, Y.; Li, Y. F.; Zhang, Z. X.; Liu, Z.; Ba, Q.; Li, J. Q.; Wang, H.; Yang, X. M.; Ma, D. W.; Wang, R. X. ChemMedChem 2014, 9 (7), 1436−1452. (38) Nikolovska-Coleska, Z.; Wang, R.; Fang, X.; Pan, H.; Tomita, Y.; Li, P.; Roller, P. P.; Krajewski, K.; Saito, N. G.; Stuckey, J. A.; Wang, S. Anal. Biochem. 2004, 332, 261−273. (39) Yung-Chi, C.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099−3108. (40) Zhang, J.-H.; Chung, T. D. Y.; Oldenburg, K. R. J. Biomol. Screening 1999, 4, 67−73. (41) Wong, K. F.; Yuan, Y.; Luk, J. M. Clin. Exp. Pharmacol. Physiol. 2012, 39, 311−320. (42) Felder, S.; Zhou, M.; Hu, P.; Urena, J.; Ullrich, A.; Chaudhuri, M.; White, M.; Shoelson, S. E.; Schlessinger, J. Mol. Cell. Biol. 1993, 13, 1449.
8057
DOI: 10.1021/acs.analchem.6b01430 Anal. Chem. 2016, 88, 8050−8057
