Mitochondrial demise is a key feature of progressive neuronal death contributing to acute and chronic neurological disorders. Recent studies identified a pivotal role for the BH3-only protein B-cell lymphoma-2 interacting domain death antagonist (Bid) for such mitochondrial damage and delayed neuronal death after oxygen-glucose deprivation, glutamate-induced excitotoxicity, or oxidative stress in vitro and after cerebral ischemia in vivo. Therefore, we developed new N-phenyl–substituted thiazolidine-2,4-dione derivatives as potent inhibitors of Bid-dependent neurotoxicity. The new compounds 6, 7, and 16 were identified as highly protective by extensive screening in a model of glutamate toxicity in immortalized mouse hippocampal neurons (HT-22 cells). These compounds significantly prevent truncated Bid–induced toxicity in the neuronal cell line, providing strong evidence that inhibition of Bid was the underlying mechanism of the observed protective effects. Furthermore, Bid-dependent hallmarks of mitochondrial dysfunction, such as loss of mitochondrial membrane potential, ATP depletion, as well as impairments in mitochondrial respiration, are significantly prevented by compounds 6, 7, and 16. Therefore, the present study identifies a class of N-phenyl thiazolidinediones as novel Bid-inhibiting neuroprotective agents that provide promising therapeutic perspectives for neurodegenerative diseases, in which Bid-mediated mitochondrial damage and associated intrinsic death pathways contribute to the underlying progressive loss of neurons.
Neuronal cell death is a consequence of both acute and chronic neurologic insults and is associated with oxidative stress, breakdown of the mitochondrial membrane potential, and release of mitochondrial death–promoting factors that mediate detrimental DNA damage (Culmsee and Landshamer, 2006). In particular, emerging evidence suggests that disturbed mitochondrial dynamics and integrity are key decision points in the sequence of intrinsic neuronal death signaling (Grohm et al., 2010). Therefore, the regulators of intrinsic pathways of programmed cell death, such as the proapoptotic members of the B-cell lymphoma-2 (Bcl-2) family, may serve as promising therapeutic targets for neuroprotection. In fact, our recent studies demonstrated a pivotal role for the proapoptotic Bcl-2 family protein Bcl-2 interacting domain death antagonist (Bid) in neuronal cell death (Culmsee et al., 2005; Becattini et al., 2006; Landshamer et al., 2008; Grohm et al., 2010) where Bid mediated detrimental effects on mitochondrial integrity and fragmentation, thereby accelerating glutamate toxicity and oxidative stress (Landshamer et al., 2008; Grohm et al., 2010, 2012).
Bid is a widely expressed cytosolic BH3-only protein comprising 195 amino acid residues (22 kDa) in its full-length form; it can be cleaved to a truncated proapoptotic protein of 15 kDa, termed truncated Bid (tBid) (Li et al., 1998). In apoptotic cells, detrimental Bid cleavage is mediated by caspase-8 or calpains, resulting in translocation of tBid to the mitochondrial outer membrane, where it mediates membrane permeabilization and the subsequent release of death-promoting proteins, such as cytochrome c, Smac/DIABLO, or apoptosis-inducing factor (AIF) into the cytosol (Gross et al., 1999; McDonnell et al., 1999; Brustovetsky et al., 2003; Culmsee and Plesnila, 2006). Our previous studies showed that inhibition of Bid by using small interfering RNA provides neuroprotective effects in vitro by inhibiting tBid-induced mitochondrial fragmentation, loss of mitochondrial membrane integrity, mitochondrial AIF release, and cell death (Landshamer, et al., 2008). In vivo, a pivotal role for Bid in mechanisms of delayed neuronal death has been confirmed in models of cerebral ischemia and brain trauma, where Bid knockout mice revealed significantly reduced brain damage compared with wild-type controls (Plesnila, et al., 2001; Yin, et al., 2002; Bermpohl et al., 2006). These results exposed Bid as a promising target for the development of novel therapeutic strategies of neuroprotection with high relevance for the treatment of age-related diseases of the nervous system, where intrinsic death pathways cause impaired mitochondrial integrity and neuronal dysfunction.
So far, only a few small molecules targeting Bid were designed based on multidisciplinary NMR approaches (structure-activity relationships by interligand Nuclear Overhauser Effects) (Becattini et al., 2006). These first Bid inhibitors inhibited tBid-induced second mitochondria-derived activator of caspase/direct IAP binding protein with low pI (Smac/DIABLO) release from isolated mitochondria and attenuated mitochondrial damage, AIF release, and neuronal cell death in models of glutamate toxicity and oxygen-glucose deprivation in neuronal cell lines and primary neurons in vitro (Becattini et al., 2004, 2006; Landshamer et al., 2008; Tobaben et al., 2011).
The aim of the present study was to develop a novel class of small-molecule compounds that target Bid and provide protection against neurotoxic insults. Based on the structures of available Bid inhibitors, a series of N-phenyl–substituted thiazolidine-2,4-dione derivatives were designed and screened for neuroprotection in a model of glutamate toxicity in immortalized mouse hippocampal HT-22 neurons. In these cells, which lack ionotropic glutamate receptors, glutamate-induced cell death is mediated by inhibition of the cellular cystine import, subsequent glutathione depletion, and enhanced formation of reactive oxygen species through increased lipid peroxidation and Bid-dependent mitochondrial damage (Murphy et al., 1989; Burdo et al., 2006). The protective effects of the synthesized compounds were further investigated in a model of glutamate-induced excitotoxicity in cultured cerebrocortical neurons. To investigate the interference with the proapoptotic activities of Bid, we examined the effects of the newly synthesized compounds after tBid expression in HT-22 cells and further analyzed the compounds’ ability to prevent glutamate-induced Bid-mediated impairments in mitochondrial integrity and function.
Materials and Methods
1H-NMR and 13C-NMR spectra were recorded on Jeol ECA-500 and Jeol EX-400 spectrometers (Tokyo, Japan). Mass spectra were recorded on a Varian MAT CH7a and a S.I.S. VG 7070. Reagents and solvents were purchased from abcr (Karlsruhe, Germany), Alfa Aesar (Karlsruhe, Germany), Fluorochem (Hatfield, UK), Matrix Scientific (Columbia, SC), Merck (Darmstadt, Germany), Sigma-Aldrich (Steinheim, Germany), and Thermo Fisher Scientific (Schwerte, Germany) and were purified by distillation or recrystallization, if necessary. Chromatography refers to column flash chromatography using the indicated eluent and Macherey-Nagel (Düren, Germany) silica gel (0.040–0.063 mm).
General Procedure A: Preparation of N-Substituted Thiazolidinediones Derivatives
The respective amine (1.1 equivalent) and methyl thioglycolate (1.0 equivalent) were added to an ice-cooled solution of 1,1′-carbonyldiimidazole (1.1 equivalent) in 30 ml of absolute methylene chloride. The reaction mixture was stirred at room temperature for 3 days and then washed three times with 3 M HCl. The solvent was removed, and the crude solid was purified by recrystallization from toluene or by flash column chromatography.
3-(4-Phenoxyphenyl)Thiazolidine-2,4-Dione (Compound 5).
According to procedure A from 4-phenoxyaniline (1.14 g, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol) and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: recrystallization from toluene to yield a white solid (1.54 g, 5.38 mmol, 96%); 1H-NMR [dimethylsulfoxide (DMSO)-d6] δ 7.41–7.47 (m, 2H), 7.29–7.33 (m, 2H), 7.18–7.22 (m, 1H), 7.07–7.10 (m, 4H), 4.30 (s, 2H); 13C-NMR (DMSO-d6) δ 171.8, 171.3, 157.1, 124.1, 128.0, 129.5, 130.1, 118.2, 34.1. Mass spectrometry (MS) electron ionization high-resolution mass spectrometry (EI-HRMS) m/z calculated for C15H11NO3S: 285.0460; found: 285.0452 [M]+.
3-Phenylthiazolidine-2,4-Dione (Compound 6).
According to procedure A from aniline (0.56 ml, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol) and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: recrystallization from toluene to yield a white solid (856 mg, 4.43 mmol, 79%); 1H-NMR (DMSO-d6) δ 7.46–7.50 (m, 3H), 7.29–7.32 (m, 2H), 4.31 (s, 2H); 13C-NMR (DMSO-d6) δ 171.7, 171.2, 133.3, 129.0, 128.6, 127.7, 34.1. MS (EI-HRMS) m/z calculated for C9H7NO2S: 193.0198; found: 193.0206 [M]+.
3-o-Tolylthiazolidine-2,4-Dione (Compound 7).
According to procedure A from o-toluidine (661 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol) and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: flash column chromatography (EtOAc, isohexane; 4:1) to yield a white solid (360 mg, 1,74 mmol, 31%); 1H-NMR (DMSO-d6) δ 7.37–7.39 (m, 2H), 7.31–7.34 (m, 1H), 7.22–7.24 (m, 1H), 4.33 and 4.47 (AB, JAB = 17.4 Hz, 2H), 2,09 (s, 3H); 13C-NMR (DMSO-d6) δ 171.4, 171.1, 135.8, 132.4, 130.6, 129.4, 128.7, 126.8, 34.3, 16.8. MS (EI-HRMS) m/z calculated for C10H9NO2S: 207.0345; found: 207.0354 [M]+.
3-p-Tolylthiazolidine-2,4-Dione (Compound 8).
According to procedure A from p-toluidine (661 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol) and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol). Purification was as follows: recrystallization from toluene to yield a white solid (801 mg, 3.87 mmol, 69%); 1H-NMR (DMSO-d6) δ 7.29–7.31 (m, 2H), 7.16-7.18 (m, 2H), 4.29 (s, 2H), 2.35 (s, 3H); 13C-NMR (DMSO-d6) δ 171.8, 171.3, 138.4, 130.7, 129.5, 127.5, 34.1, 20.6. MS (EI-HRMS) m/z calculated for C10H9NO2S: 207.0345; found: 207.0349 [M]+.
N-(3-(2,4-Dioxothiazolidin-3-yl)Phenyl)Acetamide (Compound 9).
According to procedure A from N-(3-aminophenyl)acetamide (927 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol). Purificationwas as follows: recrystallization from toluene to yield a white solid (428 mg, 1.61 mmol, 29%); 1H-NMR (DMSO-d6) δ 10.17 (s, 1H), 7.65–7.56 (m, 1H), 7.55–7.57 (m, 1H), 7.38–7.44 (m, 1H), 6.95–6.97 (m, 1H), 4.31 (s, 2H), 2.06 (s, 3H). 13C-NMR (DMSO-d6) δ 171.9, 171.3, 169.0, 140.0, 133.5, 129.2, 122.3, 119.2, 118.2, 34.3, 24.0. MS (EI-HRMS) m/z calculated for C11H10N2O3S: 250.0412; found: 250.0439 [M]+.
5-(2,4-Dioxothiazolidin-3-yl)-2-Methylbenzonitrile (Compound 10).
According to procedure A from 5-amino-2-methylbenzonitrile (815 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol). Purification was as follows: recrystallization from toluene to yield a white solid (588 mg, 2.37 mmol, 42%); 1H-NMR (DMSO-d6) δ 7.78–7.79 (m, 1H), 7.56–7.62 (m, 2H), 4.31 (s, 2H), 2.53 (s, 3H); 13C-NMR (DMSO-d6) δ 171.7, 171.0, 142.5, 132.7, 131.6, 131.5, 131.3, 117.0, 112.3, 34.4, 19.7. MS (EI-HRMS) m/z calculated for C11H8N2O2S: 232.0306; found: 232.0322 [M]+.
3-(4-(Methylsulfonyl)Phenyl)Thiazolidine-2,4-Dione (Compound 11).
According to procedure A from 4-(methylsulfonyl)aniline (1.06 g, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol). Purification was as follows: recrystallization from toluene to yield a white solid (673 mg, 2.48 mmol, 44%); 1H-NMR (DMSO-d6) δ 8.07–8.10 (m, 2H), 7.62–7.64 (m, 2H), 4.33 (s, 2H), 3.29 (s, 3H); 13C-NMR (DMSO-d6) δ 171.7, 171.0, 140.9, 137.6, 128.8, 128.0, 43.3, 34.5; MS (EI-HRMS) m/z calculated for C10H9NO4S: 270.9973; found: 270.9973 [M]+.
3-(3,4-Dimethylphenyl)Thiazolidine-2,4-Dione (Compound 12).
According to procedure A from 3,4-dimethylaniline (748 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: recrystallization from toluene to yield a white solid (533 mg, 2.41 mmol, 43%); 1H-NMR (DMSO-d6) δ 7.25–7.27 (m, 1H), 6.99–7.06 (m, 2H), 4.29 (s, 2H), 2.26 (s, 3H), 2.24 (s, 3H); 13C-NMR (DMSO-d6) δ 171.9, 171.4, 137.3, 137.2, 130.9, 129.9, 128.4, 125.0, 34.1, 19.2, 19.0. MS (EI-HRMS) m/z calculated for C11H11NO2S: 221.0496; found: 221.0515 [M]+.
3-(5,6,7,8-Tetrahydronaphthalen-2-yl)Tiazolidine-2,4-Dione (Compound 13).
According to procedure A from 5,6,7,8-tetrahydronaphthalen-2-amine (908 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: Flash column chromatography (EtOAc, isohexane; 4:1) to yield a white solid (592 mg, 4.43 mmol, 43%). 1H-NMR (DMSO-d6) δ 7.16-7.18 (m, 1H), 6.97–6.99 (m, 2H), 4.28 (s, 2H), 2.73–2.75 (m, 4H), 1.70–1.78 (m, 4H). 13C-NMR (DMSO-d6) δ 172.2, 171.4, 137.6, 137.5, 130.6, 129.5, 127.9, 124.7, 34.1, 28.6, 28.4, 22.4, 22.3. MS (EI-HRMS) m/z calculated for C13H13NO2S: 247.0667; found: 247.0670 [M]+.
3-(4-Methoxyphenyl)Thiazolidine-2,4-Dione (Compound 14).
According to procedure A from 4-methoxyaniline (0.76 g, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: recrystallization from toluene to yield a white solid (1.02 mg, 4.76 mmol, 74%); 1H-NMR (DMSO-d6) δ 7.20–7.22 (m, 2H), 7.03–7.05 (m, 2H), 4.28 (s, 2H), 3.79 (s, 3H); 13C-NMR (DMSO-d6) δ 172.0, 171.5, 159.3, 129.0 125.8, 114.3, 55.3, 34.1. MS (EI-HRMS) m/z calculated for C16H13NO2S: 223.0303; found: 223.0296 [M]+.
3-(4-Benzylphenyl)Thiazolidine-2,4-Dione (Compound 15).
According to procedure A from 4-benzylaniline (1.13 g, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol), and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: recrystallization from toluene to yield a white solid (1.35 g, 4.76 mmol, 85%), 1H-NMR (DMSO-d6) δ 7.26–7.37 (m, 6H), 7.18–7.22 (m, 3H), 4.29 (s, 2H), 3.99 (s, 2H); 13C-NMR (DMSO-d6) δ 171.9, 171.4, 142.2, 140.7, 131.2, 129.2, 128.7, 128.5, 127.8, 126.1, 40.6, 34.2. MS (EI-HRMS) m/z calculated for C16H13NO2S: 283.0667; found: 283.0654 [M]+.
3-(3,4-Dimethoxyphenyl)Thiazolidine-2,4-Dione (Compound 16).
According to procedure A from 3,4-dimethoxyaniline (945 mg, 6.17 mmol), 1,1′-carbonyldiimidazole (1.00 g, 6.17 mmol) and thioglycolic acid methyl ester (0.50 ml, 5.61 mmol), purification was as follows: recrystallization from toluene to yield a white solid (865 mg, 3.42 mmol, 63%); 1H-NMR (DMSO-d6) δ 7.04–7.06 (m, 1H,), 6.92–6.93 (m, 1H), 6.81–6.83 (m, 1H), 4.28 (s, 2H), 3.79 (s, 3H), 3.73 (s, 3H) 13C-NMR (DMSO-d6) δ 166.2, 165.0, 151.0, 149.0, 134.3, 121.9, 115.9, 115.8, 55.7, 34.7. MS (EI-HRMS) m/z calculated for C11H11NO4S: 253.0409; found: 253.023 [M]+.
According to Becattini et al. (2006), docking studies were conducted using the NMR solution structure of mouse BID from the Protein Data Bank entry PDB ID 1DDB (McDonnell et al., 1999). All compounds were converted from SMILES notation into Sybyl mol2 format using CORINA (Molecular Networks GmbH, Erlangen, Germany). Protein preparation and docking were performed using FlexX (Rarey et al., 1996), available within the LeadIT version 2.1.3 from BioSolveIT GmbH (Sankt Augustin, Germany). Pictures for two-dimensional representation of the docking solution were generated with pose view (Stierand and Rarey, 2007), as implemented in LeadIT, and exported from the program in the same way as the three-dimensional representation.
Expression and Purification of Recombinant Bid
Full-length Bid with a hexa-histidine tag at the N terminus was expressed in the pET15b vector (Addgene, Cambridge, MA) in competent Escherichia coli Rosetta 2 (DE3) cells (Novagene, Merck KGaA, Darmstadt, Germany). The protein was recovered in the soluble bacteria fraction and purified by ÄKTAprime plus chromatography on a Ni Sepharose HisTrap FF column, followed by an ion exchange HiTrap Q HP column and a HiLoad 16/600 Superdex 75-pg column (GE Healthcare Bio-Science AB, Uppsala, Sweden). Highly pure protein was eluted as shown by SDS-PAGE and Coomassie staining and stored in 20 mM Tris pH 7.4, 50 mM NaCl at 4°C until use.
Label-Free Biochemical Assay
Biochemical assays were performed using the Corning Epic label-free technology on the EnSpire Multimode Plate Reader (PerkinElmer, Rodgau, Germany) (O’Malley et al., 2007). Bid immobilization on the optical biosensors was accomplished by adding 200 µg/ml protein in 20 mM sodium acetate, pH 4.0, using a 12-channel Thermo Scientific matrix multichannel equalizer pipette followed by overnight incubation at 4°C. The microplate was subsequently washed three times with phosphate-buffered saline (PBS; pH 7.4) buffer with 0.005% Tween by a robotic wash station. After washing, the plate was equilibrated in the assay buffer (PBS, 1% DMSO, pH 7.4) for 2 hours (30 µl). After the incubation, a baseline reading was recorded. After another washing step, 15 µl of the assay buffer, together with 15 µl of various thiazolidinediones, prepared as described previously, was dispensed in the plate. The final readings were taken over a period of 1 hour. The label-free responses were measured as shifts in reflected wavelength and were expressed in picometers. Results were analyzed using the EnSpire label-free user interface software. The difference between the last baseline measurements and the maximum signal was used to determine the KD values. Graphs were generated using GraphPad PrismR V-5.01 (GraphPad Software, La Jolla, CA). Compounds used for the Bid assay were diluted with the assay buffer (PBS, 1% DMSO, pH 7.4) at a working concentration of 200 µM (100 µM final concentration in the plate) and then further diluted in the assay buffer directly in a 384-well polypropylene plate for a total of four different concentrations.
Cell Culture of HT-22 Cells
HT-22 cells, derived from immortalized hippocampal neurons were cultured in Dulbecco’s modified Eagle’s medium (Invitrogen, Karlsruhe, Germany) supplemented with 10% heat-inactivated fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine (PAA Laboratories GmbH, Cölbe, Germany). Cells were cultured at densities of 10,000 to 12,000 cells/well either in standard 96-well plates (Greiner, Frickenhausen, Germany) for MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium-bromide] assays or in E-plates 96 (Roche Diagnostics, Penzberg, Germany) for impedance measurement. For plasmids and gene transfer and measuring of tBid-induced toxicity, cells were cultured at densities of 40,000 cells/wells in standard 24-well plates (Greiner).
Primary Cerebrocortical Neuron Culture
Primary cortical neurons were prepared from embryonic brains (E16-18) of rats. Meninges were removed and the cortical neurons separated by mechanical dissociation and mild trypsinization. Cells were plated at a density of 16,000 cells/well (96-well plates) on polyethyleneimine precoated plates. Neurobasal medium (Invitrogen) supplemented with 5 mM HEPES, 1.2 mM glutamine, 2% (v/v) B27 supplement (Invitrogen) was used as a culture medium. After 7 or 8 days of in vitro culture, cortical neurons were used for further experiments. Compounds were prepared as a stock solution in DMSO. The final concentration of DMSO in all experiments was less than 0.5% v/v, revealing no effect on glutamate-induced cell death. Compounds were applied at a concentration of 25 µM 1 hour before the glutamate challenge and additionally as cotreatment with glutamate. For induction of excitotoxicity, cerebrocortical neurons were exposed to 150 µM glutamate for 24 hours. Experiments were repeated at least three times, and analyses were performed without knowledge of the treatment history.
Induction of Neuronal Cell Death in HT-22 Cells
Neuronal cell death was induced 24 hours after seeding of the cells. Induction of apoptosis was performed by either glutamate-induced toxicity or tBid overexpression.
For measuring glutamate toxicity, cell growth medium was removed and replaced by standard cell culture medium containing glutamate (3 mM) (Sigma-Aldrich, Munich, Germany) and/or novel inhibitors at final concentrations of 0.1 to 100 µM. BI-6C9 (N-[4-[(4-aminophenyl)thio]phenyl]-4-[[(4-methoxyphenyl)sulfonyl]amino]-butanamide; Sigma-Aldrich) was added to the media at a final concentration of 10 µM and used as a positive control for neuroprotection. Between 14 and 16 hours later, cell viability was analyzed.
One hour before transfection with a tBid-expressing plasmid (pIRES-tBid), HT-22 cells were pretreated with BI-6C9 (Sigma-Aldrich) at a final concentration of 10 µM or with novel compounds at final concentrations of 1 to 100 µM. Sixteen to 20 hours after tBid-induced toxicity, cell morphology and cell viability were analyzed.
For analysis of cellular morphology transmission light microscopy of living HT-22 cells was performed using an Axiovert 200 microscope (Carl Zeiss, Jena, Germany) equipped with a Lumenera Infinity 2 digital camera (Lumenera Corporation, Ottawa, ON, Canada). Light was collected through a 10 × 0.25 numerical aperture (NA) objective (Carl Zeiss), and images were captured using phase contrast. The INFINITY ANALYZE software (Lumenera Corporation) was used for digital image recording and image analysis.
Cell Viability Assay and Impedance Measurements
Quantification of cell viability in HT-22 cells was performed either in standard 96-well plates or in standard 24-well plates by MTT (Sigma-Aldrich) reduction at 0.25 mg/ml for 1 hour. After terminating the reaction by removing the media and freezing the plate at −80°C for at least 1 hour, the MTT dye was dissolved in DMSO and absorbance was determined at 590 nm versus 630 nm (FluoStar OPTIMA; BMG Labtech, Offenburg, Germany). Cell viability levels were demonstrated as the percentage of absorption levels in untreated control cells (100% cell viability). DMSO control was used as solvent control. For statistical analysis, experiments were repeated at least three times.
In addition, real-time detection of cellular viability was conducted using the xCELLigence real-time cell analyzer (RTCA)-MP (Roche Diagnostics) by measuring the electrical impedance between microelectrodes integrated into the bottom of custom-made tissue culture plates (E-plates 96). High electrical resistance of cells enables monitoring of changes in cell proliferation and cell viability recorded as the differences in cell impedance, which is displayed as the cell index (CI) as function of time. HT-22 cells were cultured at a density of 10,000 cells/well in 96-well E-plates and treated with glutamate (3 mM) and/or novel compounds (0.1–50 µM) 24 hours after seeding. Recording of CI values and normalization were performed using the RTCA Software 1.2 (Roche Diagnostics). Background impedance caused by the media was recorded before seeding of the cells and subtracted automatically by the RTCA software.
For analysis of total ATP levels, HT-22 neurons were seeded in white 96-well plates (Greiner) for luminescence measurements. Twenty-four hours after seeding, cells were treated with 5 mM glutamate and when indicated additionally with the Bid inhibitor BI-6C9 (10 µM) and novel compounds (20 µM), respectively. Cellular ATP levels were detected 20 hours after glutamate exposure using the ViaLightTM Plus-Kit (Lonza, Verviers, Belgium). After treatment of cells with the nucleotide-releasing reagent and injection of the ATP monitoring reagent into each well, luminescence was detected immediately by the FluoStar plate reader (BMG Labtech). The values were given as relative values in percent of control cells. The experiments were repeated at least three times with an n = 8 per treatment condition.
Measurements of Oxygen Consumption Rate
Oxygen consumption rate (OCR) measurements were assessed using an XF96 Extracellular Flux Analyzer (Seahorse Bioscience, North Billerica, MA), which directly records the OCR in cells that remain attached to the culture plate by using calibrated optical sensors. The OCR recordings were carried out as previously described with minor modifications (Gohil et al., 2010). Briefly, HT-22 cells were seeded in XF 96-well cell culture microplates (Seahorse Bioscience) at a density of 10,000 cells/well in 4.5 g/liter of standard culture medium and incubated at 37°C and 5% CO2 for ∼24 hours and afterward treated with glutamate (5 mM) and/or novel compounds (20 µM) for 20 hours. Before starting the measurements, the growth medium was washed and replaced with ∼180 μl of assay medium (with 4.5 g/liter of glucose as the sugar source, 2 mM glutamine, 1 mM pyruvate, pH 7.35), and cells were incubated at 37°C for at least 60 minutes. Three baseline measurements were recorded before the addition of compounds. To assess mitochondrial dysfunction, respiratory chain poisons were used. The ATP synthase inhibitor oligomycin was injected in port A at a final concentration of 3 µM to measure OCR in the absence of oxidative phosphorylation. The protonophore carbonylcyanide-4-(trifluormethoxy)-phenylhydrazone was subsequently injected in port B at a concentration of 0.4 µM to dissipate the proton gradient across the inner mitochondrial membrane and thereby to assess mitochondrial respiratory capacity (MRC). Coinjection of the complex I/III inhibitors rotenone/antimycin A in port C at concentrations of 1 µM respectively was used to inhibit O2 consumption by the mitochondrial electron transport chain and thus to address nonmitochondrial respiration. Three measurements were performed after the addition of each compound by a 4-minute mix cycle used to oxygenate the medium and a 3-minute measurement cycle to assess respiration. For comparing OCR after compound exposure to OCR in nontreated cells, the absolute cell number is insignificant as a result of the same population of cells being compared. Therefore, results are indicated as normalized OCR (% baseline rate) for each individual cell population to minimize variability resulting from slight differences in plating and viability during culture and treatment time (∼48 hours).
Detection of Mitochondrial Membrane Potential (Δψm)
For the detection of mitochondrial membrane potential (Δψm), cells were treated with 5 mM glutamate and/or the Bid inhibitor BI-6C9 (10 µM) or the novel compounds (20 µM) 24 hours after cell seeding in 24-well plates. Twenty-hours after glutamate exposure, Δψm in whole cells was determined using the MitoPT tetramethylrhodaminethyleste (TMRE) kit (Immunochemistry Technologies, Hamburg, Germany) followed by flow cytometric measurements using the Guava Easy Cyte 6-2 L system (Merck Millipore, Schwalbach, Germany). As a positive control for a complete loss of Δψm, the protonophore carbonyl cyanide m-chlorophenylhydrazone (50 µM) was applied to intact cells. After respective treatment, cells were collected and washed with PBS. After incubation with 0.2 µM TMRE for 30 minutes at 37°C, cells were washed and resuspended in assay buffer. Flow cytometry was performed using fluorescence excitation at 488 nm and TMRE emission at 680 nm. Data were collected from 10,000 cells per condition. High red TMRE fluorescence indicates intact mitochondria, whereas mitochondrial membrane depolarization is evidenced by the drop in red fluorescence. For quantitative analysis, the GuavaSoft Software package was used. Measurements are representative of at least three to five independent experiments, each with n = 3.
Plasmids and Gene Transfer
Plasmid pCDNA 3.1+ was obtained from Invitrogen. The pIRES-tBid vector was generated as previously described (Kazhdan et al., 2006). The ApoAlert pDSRed2-Bid vector, which encodes a biologically active fluorescent fusion protein of Bid and DsRed monomer, was derived from Clontech (Palo Alto, CA).
tBid Overexpression in HT-22 Cells
For plasmid transfections of HT-22 cells, 4 × 104 cells/well were seeded in 24-well plates 24 hours before transfection and incubated under normal growth conditions (37°C, 5% CO2). On the day of transfection, standard growth medium was replaced with 500 µl of fresh standard growth medium. Pretreatment with different substances was performed 1 hour before the following plasmid transfection. pIRES-tBid or empty vector pcDNA 3.1+ was dissolved separately in antibiotic- and serum-free medium, and Attractene transfection reagent (Qiagen, Hilden, Germany) was added to the DNA solution. To allow complex formation, samples were incubated for 10 to 15 minutes at room temperature. Afterward, 60 µl of the transfection mixture was added to the cell culture medium at a final concentration of 2 µg of DNA/well and 4.5 µl of Attractene/well. Control cells were treated with 60 µl antibiotic- and serum-free cell culture medium only, and vehicle controls were additionally treated with 4.5 µl of Attractene/well. Cells were incubated under normal growth conditions until cell morphology and cell viability were analyzed 16 to 20 hours after tBid overexpression.
tBid Overexpression in Human Embryonic Kidney 293 Cells
Human embryonic kidney (HEK293) cells were cultured in Dulbecco’s modified Eagle medium (Invitrogen) supplemented with 10% heat-inactivated fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine (PAA Laboratories GmbH). For tBid overexpression, cells were seeded in 12-well plates at a density of 10,000 cells/well. Twenty-four hours after seeding, cells were transfected with 0.5 µg of pIRES-tBid using effectene (Qiagen) and allowed to express the recombinant proteins for 24 hours. Compound 16 (25 µM) and BI-6C9 (10 µM) were applied on HEK293 cells 1 hour after pIRES-tBid transfection. To analyze the toxicity mediated by tBid overexpression, we performed PI flow cytometric measurements using the Guava Easy Cyte 6-2L system (Merck Millipore). Data were collected from 10,000 cells per treatment condition. For quantification of the data, the GuavaSoft Software package was used. PI staining was used to determine the apoptotic and necrotic cells. Cells were incubated with 2.5 µl PI (PromoKine, Heidelberg, Germany) for 10 minutes at room temperature, and fluorescence of PI was detected with the red filter at 690/50 nm. tBid was detected by the green fluorescent IRES with the green filter at 525/530 nm, and all green cells were gated and analyzed for PI-positive cells.
Immunostaining and Confocal Laser Scanning Microscopy
For analysis of Bid localization and translocalization, HT-22 cells were seeded in ibidi µ-slide eight-well plates at a density of 20,000 cells/well and transfected with the pDSRed2-Bid vector. Mitochondria were visualized by transfection of cells with mito-green fluorescent protein (GFP) as previously described (Landshamer et al., 2008). Plasmid transfection was performed analogous to the tBid transfection as described already herein. Cells were cotreated with glutamate (3 mM) and thiazolidinediones (25 µM) 48 hours after transfection. Immunoanalysis of pDSRed2-Bid and mGFP was performed 18 hours after treatment. After fixation of cells with 4% paraformaldehyde, images were acquired using a confocal laser scanning microscope (Leica SP5; Leica, Wetzlar, Germany). Light was collected through a 63 × 1.4 NA, oil immersion objective. mGFP fluorescence was excited at 488 nm, and emissions were detected between 500 and 535 nm bandwidth. DSred2-Bid fluorescence was detected by excitation at 633 nm and emission between 640 nm and 750 nm. For digital imaging, the software LSM Image Browser 4.2.0 (Carl Zeiss) was used.
Concentration-Response Curves and EC50 Values
To compare the potency of different synthesized compounds, EC50 values were determined based on the concentration-response curves for each of the substances. To determine the concentration required to achieve maximal neuroprotective effects, initially all compounds were screened by their ability to prevent glutamate-induced cell death in concentrations of 10, 20, 30, 40, and 50 µM. Since quantification of cell viability (MTT assay) revealed maximal protective effects of the compounds in a concentration of 20 µM and greater, we could not generate concentration ranges for this (Supplemental Fig. 1). Therefore, compounds with an EC50 < 10 µM were screened in lower concentrations of 0.1 µM up to 50 µM for maximum protection. Cell viability data were normalized between untreated control conditions (0%) and maximum amplitude at 50 µM (100%). Using OriginPro8.5 software (OriginLab Corporation, Northampton, MA), the resulting data were fitted with a sigmoid function following the equation: and EC50 values for all substances were calculated.
All data are given as means ± standard deviation. Statistical multiple comparisons were performed by analysis of variance followed by Scheffé’s post hoc test. Calculations were performed with the Winstat standard statistical software package (R. Fitch Software, Bad Krozingen, Germany).
In an effort to identify structurally novel small-molecule compounds that provide protection against intrinsic mitochondrial death pathways, we focused on the development of potent inhibitors of Bid-mediated neurotoxicity. The previously described Bid inhibitors were obtained by estimating the binding affinity of different fragments through Nuclear Overhauser Effects by Becattini et al. (2006). The binding fragments (i.e., BI-2A7 and BI-2A1) showed binding affinity, but only BI-2A1 reduced the Bid-mediated release of the proapoptotic factor Smac/DIABLO. Combination of two of the fragments yielded BI-6C9, the most active inhibitor described in the series of Becattini et al. (2004) (Fig. 1). We initially started the development of alternative Bid-inhibitors by substituting the p-phenoxyaniline for fragment BI-2A7 (1), thereby omitting the metabolically labile primary aromatic amine and the thioether partial structure. Since the p-phenoxyaniline (4) and its propanoyl derivatives provided slightly neuroprotective activities in a study conducted in parallel, structure 4 was used to further explore structure-activity relationships of this type of compound. Therefore, we incorporated the acylamide partial structure of 4 into a thiazolidine-2,4-dione, concurrently reducing conformational flexibility (Fig. 1).
To establish structure-activity relationships, several thiazolidine-2,4-dione derivatives were prepared by varying the 3-aryl substituent. The desired compounds were obtained following a procedure described by Geffken (1987). Briefly, the appropriately substituted aniline, thioglycolic acid methyl ester, and N,N-carbodiimidazole gave the target thiazolidine-2,4-diones in a one-pot reaction. These compounds were further analyzed for their ability to target Bid and their neuroprotective efficiency (Table 1).
Of note, for some molecules (e.g., compound 7), a rotational barrier is observed as demonstrated by the AB system for the protons at C-5 in the 1H-NMR spectrum. The spatial structures, a two-dimensional illustration concerning the barrier and the corresponding C-5 proton NMR-signals of compound 8 (left hand) and 7 (right hand), are shown in Fig. 2. The reduced flexibility of compound 7 might improve its binding to Bid and thereby increase its neuroprotective property compared with more flexible structures (e.g., compound 8) (Fig. 2).
N-Phenyl–Substituted Thiazolidinediones Provide Neuroprotection against Glutamate-Induced Toxicity.
In a first approach, all synthesized structures were screened for their potential to attenuate glutamate-induced and Bid-dependent neuronal cell death in HT-22 cells. Therefore, HT-22 cells were treated with the synthesized compounds in concentration ranges of 0.1 µM up to 50 µM and exposed to toxic glutamate concentrations of 3 mM. To investigate the protective and toxic properties of all compounds, cell viability was determined by the MTT assay 14 hours after cotreatment.
In the first screening setup, examining substance concentrations between 10 and 50 µM, most of the compounds revealed maximal neuroprotection at a concentration of 20 µM, whereas concentrations higher than 30 µM and up to 50 µM could not further increase the protective properties (Supplemental Fig. 1). To examine the doses required to achieve maximal protection, compounds were further tested in concentration ranges between 0.1 µM and up to 10 µM, and the concentration of 50 µM was set as steady-state dose providing 100% protection. Figure 3 shows the protective effects of three representative 3-aryl substituted thiazolidinedione derivatives, which significantly preserved cell viability of glutamate-treated HT-22 neurons in a concentration-dependent manner between 5 and 50 µM (Fig. 3, A, C, and E). Notably, compounds 6, 7, and 16 achieved full protection against glutamate-induced toxicity at concentrations greater than 20 µM (Fig. 3, A, C, and E; Supplemental Fig. 1, A–E). This pronounced protective effect was comparable to the protective effect of the available Bid inhibitor BI-6C9 that was applied in the screening experiments as a reference for maximal protection in this cell death model. Importantly, treatment of cells with the newly synthesized substances alone revealed no toxic effects in concentrations up to 50 µM (Fig. 3, A, C, and E; Supplemental Fig. 1, A–E). Similar results were obtained for all synthesized compounds mentioned in Table 1.
To further compare the neuroprotective potency of all tested compounds, EC50 values were calculated from the cell viability data at the applied concentration ranges. Since our data revealed full protective effects of the compounds at concentrations of 50 µM (positive control for highest protection), cell viability data were normalized between 0% for glutamate-only treated cells (vehicle) to 100% protection at 50 µM and concentration-response curves were fitted with a sigmoid function (dose-response fit). Concentration-response curves for compounds 6, 7, and 16, which represent the highest therapeutic potency of the tested compounds, are shown in Fig. 2, B, D, and F, and calculated EC50 values of all structures are provided in Table 1.
Real-Time Monitoring of Cell Impedance Confirms Neuroprotection by the N-Phenyl–Substituted Thiazolidinediones.
The neuroprotective properties of the N-phenyl–substituted thiazolidinediones detected by endpoint MTT assays were substantiated by real-time analysis of cell impedance, expressed as normalized cell index (NCI) (Fig. 4). High electrical resistance of cells allows the investigation of cell proliferation and viability, as well as cell morphology and adhesion, since alterations in these cell properties are recorded as changes in cell impedance (Diemert et al., 2012). The RTCA revealed that cellular impedance significantly decreased in cell populations exposed to glutamate (3 mM) 5 to 8 hours after treatment. In contrast, NCI was fully preserved to control levels by cotreating the cells with glutamate (3 mM) and compounds 6, 7, or 16 at a concentration of 50 µM, respectively (Fig. 4, A–C). Notably, once initiated, glutamate exposure triggered the complete loss of cellular impedance within a small time window of 2–4 hours, indicating that neuronal cell death occurred rapidly and in a highly synchronized manner. Nevertheless, this decrease in CI was persistently blocked by our novel inhibitor compounds. Further monitoring of cell impedance confirmed the concentration-dependent protective effects of the tested compounds that were shown by the MTT assay (Fig. 3, A–C). Compounds 6 and 16 revealed transient neuroprotection at concentrations of 5, 10, and 25 µM (6) but showed a sustained protection over time at concentrations of 50 µM (Fig. 4, B and C). Similarly, compound 7 was only transiently protective at concentrations lower than 5 µM but attenuated the complete decline of NCI already at concentrations of 10 µM, although not to levels of control cells. However, persistent neuroprotection against glutamate-induced toxicity was achieved at a concentration of 50 µM (Fig. 3A). These results accentuate that real-time measurements of cell proliferation not only confirmed our findings detected by MTT assay but further allowed for distinction between transient and persistent protective effects. Therefore, this method provided a better estimation of the protective potential of the novel compounds.
To transfer the compound’s neuroprotective effects established in the neuronal HT-22 cell line to primary neuronal cells, the thiazolidinediones were applied to a model of glutamate-induced excitotoxicity in primary cerebrocortical rat neurons. Glutamate challenge mediated significant neuronal toxicity, which was partially rescued by pretreatment with our synthesized thiazolidinediones, particularly with compound 16 (Fig. 4D).
N-Phenyl–Substituted Thiazolidinediones Attenuate tBid-Induced Neuronal Cell Death.
Our previous studies demonstrated a pivotal role for the BH3-only protein Bid in neuronal apoptotic cell death in the present model of glutamate toxicity in HT-22 cells (Landshamer et al., 2008; Grohm et al., 2010). Pharmacological Bid inhibition by the available small-molecule inhibitor BI-6C9 prevented cell death by inhibiting tBid translocation to the mitochondrial membrane, the subsequent release of AIF, cytochrome c, and Smac/DIABLO into the cytosol (Becattini et al., 2004, 2006). To examine the role of our newly synthesized compounds in Bid-mediated neuronal cell death, the novel compounds were applied to HT-22 neurons transfected with a tBid-encoding plasmid, and cell morphology and cell viability were analyzed (Fig. 5). After overexpression of tBid, HT-22 neurons showed excessive alterations in cell morphology: cells clearly appeared shrunken, rounded up, and detached from the bottom of the culture plate (Fig. 5A). In contrast, 1 hour of pretreatment of the cells with the novel compounds (1 µM) rescued HT-22 cells from tBid-induced toxicity and preserved spindle-shaped cell morphology (Fig. 5A). Notably, treatment of the cells with the compounds 6, 7, and 16 at low concentrations of 1 µM induced significant protective effects (Fig. 5). Quantification of cell viability determined by MTT assay 20 hours after tBid transfection confirmed the compound-mediated protection against tBid-induced cell death. Expression of tBid reduced cell viability of HT-22 cells to 20–30% of control levels, whereas cell viability was significantly restored by pretreatment of cells with the compounds 6, 7, or 16 at concentrations of 1 µM, respectively (Fig. 5, B–D). Similar results were observed with pretreatment of the cells with the indicated compounds at higher concentrations of 10 and 50 µM, whereas no toxic effects of the compounds were detected when applied under control conditions (data not shown). To estimate the transfection efficiency of tBid-transfected cells, parallel transfection experiments in HT-22 cells were performed using the current protocols for DNA vector transfection (Attractene; Qiagen). Cells transfected with a GFP of a comparable vector size as the tBid-vector were analyzed by fluorescence microscopy, revealing a transfection efficiency of approximately 70% (Supplemental Fig. 2). These data correlate well to the percentage of tBid-induced cell death and are similar to previous results on the transfection efficiency in HT-22 cells (Landshamer et al., 2008). Thus, it is suggested that treatment of cells with the thiazolidinediones protected all transfected cells. These findings showed that the new N-phenyl–substituted thiazolidinedione compounds were able to rescue HT-22 cells after tBid-overexpression, thereby providing strong evidence for our hypothesis that the newly synthesized compounds provided protection by targeting the proapoptotic protein Bid.
To examine whether the proposed protective potency of the compounds against tBid-induced cell death is specific for neuronal cells or whether they can be transferred to non-neuronal cells, we investigated the compound’s effect on HEK293 cells (Supplemental Fig. 3A). To this end, HEK293 cells were transfected with the tBid-encoding plasmid in the presence and absence of the synthesized thiazolidinediones. Notably, tBid overexpression promoted severe cell damage, which occurred rapidly within 15 hours after transfection, indicating that this cell line is more sensitive to tBid-induced cell death than the current used HT-22 cell line. In HEK293 cells, cotreatment with BI-6C9 (10 µM) provided only a slight protection against tBid toxicity. Interestingly, the protective effect achieved by cotreatment of cells with compound 16 was more pronounced, revealing a small but significant increase in HEK293 cell survival compared with tBid-mediated damage (Supplemental Fig. 3A). These data indicate that the protective potential of the newly synthesized thiazolidinediones is not specific for neurons, although their potency might vary between different cell lines.
For a better understanding of the protective effects provided by the novel compounds and to confirm their ability to inhibit Bid, the compounds were subjected to docking analysis into the three-dimensional structure of Bid. Thereby a possible binding mode of compound 7, which revealed the highest protective potency, was derived by means of docking with the program FlexX (Fig. 6). The thiazolidinedione moiety points into a deeply buried pocket on the surface of Bid, like BI-6C9, and allows interactions with the amino acids Val186 and Ile86 of Bid. Further, a hydrophobic interaction between the aromatic ring of compound 7 with the side chain of Tyr185 of Bid is suggested (Fig. 6, A and B) and substantiate the intended specificity of newly synthesized compounds to target Bid. Finally, we could confirm the compound’s specificity in a cell-free system using purified recombinant Bid protein. The binding of the presented thiazolidinediones to recombinant Bid was measured using the EnSpire label-free platform (Fig. 6, C–E) (O’Malley et al., 2007). Using this biochemical assay, the target Bid was initially immobilized at a concentration of 200 µg/ml onto the amino-coupling surface of the EnSpire label-free biochemical microplate biosensor. After washout of the unbound target and further equilibration of the biosensor, several concentrations of the appropriate compounds were added to the immobilized Bid protein. The background-corrected responses were well fitted by a one-site interaction model, yielding a KD value in the low µM range for each compound tested (KD 17.7 ± 5.2 µM (6), KD 2.2 ± 0.9 µM (7), KD 0.76 ± 0.32 µM (16); Fig. 6, C–E). These findings underscore that the identified thiazolidinediones are able to bind efficiently to Bid.
Novel Compounds Prevent Glutamate-Induced Bid Translocation and Bid-Mediated Mitochondrial Impairments.
Glutamate-induced toxicity involves mainly impairments in mitochondrial integrity and function. Loss of mitochondrial membrane potential (Δψm) and mitochondrial fragmentation trigger the release of death promoting factors into the cytosol, thereby preceding the final execution of caspase-independent cell death (Konig et al., 2007; Landshamer et al., 2008; Grohm et al., 2010). Since these intrinsic cell death features are Bid-dependent and initiated by the translocation of Bid to mitochondria (Plesnila et al., 2001; Culmsee and Plesnila, 2006; Landshamer et al., 2008; Grohm et al., 2010), we further addressed the compound's protective potential against Bid-dependent mitochondrial dysfunction.
To examine the compound’s ability to prevent mitochondrial Bid translocation after the glutamate challenge, HT-22 cells were transfected with a pDsRed2-Bid vector that results in the expression of a red fluorescing Bid residing in the cytosol (Fig. 7A). To visualize mitochondria, cotransfection of cells with mGFP was performed. In line with our previous findings in HT-22 cells, under physiologic conditions, Bid is distributed in the cytosol, whereas glutamate treatment causes the translocation of Bid to mitochondria (Fig. 7A, yellow in merged) (Landshamer et al., 2008). Cotreatment of cells with the compound 16, as a representative of the thiazolidinediones, prevented the accumulation of Bid at the mitochondria and retrained the homogeneous distribution of Bid in the cytosol (Fig. 7A). To further investigate whether hallmarks of Bid-dependent cell death downstream of mitochondrial Bid transactivation are prevented by the presented compounds, we analyzed mitochondrial membrane potential (Δψm) in HT-22 cells. After the translocation of Bid to mitochondria, a significant loss of Δψm was detected, indicated by a drop in red TMRE fluorescence 20 hours after glutamate treatment (Fig. 7, B and C). The uncoupler carbonyl cyanide m-chlorophenylhydrazone was used as a positive control to induce depolarization of the mitochondrial membrane (Fig. 7C). Of note, the application of 20 µM of the novel compounds 7, 16, or 6 did not affect Δψm under control conditions but almost completely attenuated the pronounced breakdown of Δψm after glutamate challenge (Fig. 7, B and C). The protective effects mediated by the novel compounds were comparable to that of the available Bid inhibitor BI-6C9, suggesting that the newly synthesized compounds preserve Bid translocation to mitochondria, thereby maintaining Δψm.
To address whether the N-phenyl thiazolidinediones are also able to prevent the functional integrity of mitochondria, ATP levels of compound-treated HT-22 cells were examined in the presence and absence of glutamate (Fig. 7D). Indeed, the rapid depletion in cytoplasmic ATP caused by glutamate-induced cell injury was completely attenuated by compounds 7, 6, and 16, whereas ATP levels of control cells were not altered (Fig. 7D). Therefore, these data confirmed at a functional level that the novel small molecules preserve mitochondrial integrity and energy production of glutamate-treated HT-22 cells.
Mitochondrial Respiratory Activity Is Maintained by N-Phenyl–Substituted Thiazolidinediones.
Broader insights into the compounds’ ability to rescue metabolic functions were derived from further investigations of mitochondrial bioenergetics (Fig. 8). Therefore, HT-22 cells were challenged with glutamate and/or the novel compounds, and O2 consumption was analyzed using the Extracellular Flux Analyzer (Seahorse Bioscience) designed for the exposure of mitochondria to defined substrates and inhibitors to modulate mitochondrial respiration in adherent cells. In accordance with previous studies reporting that glutamate-induced mitochondrial dysfunction involves changes in mitochondrial acidification and OCR, glutamate exposure of HT-22 cells decreased the OCR, indicating a glutamate-induced reduction in mitochondrial respiration (Fig. 8). After recording the basal mitochondrial respiration, the ATP synthase inhibitor oligomycin was added to uncouple mitochondrial respiration from ATP production. While the OCR of control cells was decreased after oligomycin injection, glutamate-treated cells responded less to oligomycin (Fig. 8). The further application of the protonophor carbonylcyanide-4-(trifluormethoxy)-phenylhydrazone increased OCR to a maximum extent, allowing the examination of the maximum oxygen flux and MRC of cells. After glutamate treatment, both maximum oxygen flux and MRC were significantly decreased but were notably restored by cotreatment of cells with the compounds 7, 6, or 16, respectively (Fig. 8, A–C). Finally, the mitochondrial respiratory chain was inhibited by the application of the complexes I and III inhibitors rotenone and antimycin A. Whereas rotenone/antimycin A–sensitive OCR specifically indicates respiration in mitochondria, rotenone/antimycin A–nonresistant rate identifies nonmitochondrial respiration and allows for calculation of mitochondrial maximum respiration. After onset of glutamate exposure, mitochondrial maximum respiration was markedly reduced in control cells but was preserved by the application of compounds 7, 6, and 16 at concentrations of 20 µM, respectively (Fig. 8, A–C). In summary, the prevention of glutamate-induced reduction in mitochondrial respiratory capacity and maximum respiration provides additional evidence for rescued metabolic functions and maintained energy metabolism in mitochondria of glutamate-treated HT-22 cells that are mediated by the N-phenyl–substituted thiazolidinediones.
In the present study, we developed a series of novel 3-phenyl–substituted thiazolidine-2,4-dione derivatives that display protective effects against glutamate-induced toxicity and, in addition, rescued HT-22 cells against tBid-mediated cell death. Moreover, key features of mitochondrial dysfunction, such as loss of mitochondrial membrane potential described as “point of no return” in the cells commitment to die (Kroemer et al., 2007), as well as impairments in mitochondrial energy metabolism, were effectively prevented by the novel compounds. The observed protective effect in the model of oxidative stress is already promising since the new compounds may provide a protective benefit against pathologic pathways occurring in neurodegenerative diseases where oxidative stress triggers the onset of neuronal cell death. Further, our finding that the indicated structures preserved cell viability in HT-22 cells after tBid overexpression strongly supports the intended specificity of the N-phenyl–substituted thiazolidinediones to provide protection by targeting the proapoptotic protein Bid, which was substantiated by virtual docking analysis. The specific binding of the compounds to the target Bid was finally confirmed by applying the compounds to recombinant Bid protein in biochemical cell-free assay. Of note, we report that glutamate-induced mitochondrial translocation of Bid and the subsequent hallmarks of Bid-dependent intrinsic cell death were significantly prevented by the N-phenyl–substituted thiazolidinediones (Fig. 7) and further demonstrated the compounds’ ability to preserve the oxygen consumption rate of HT-22 cells, and thereby mitochondrial bioenergetics, described to be defective in neurodegenerative disorders (Clerc and Polster, 2012). In addition, we revealed that the proposed protective effects of the presented compounds found in HT-22 cells can be transferred to other cells. Parallel experiments in primary cultured neurons demonstrated the neuroprotective effects of the N-phenyl–substituted thiazolidinediones in a model of glutamate-induced excitotoxicity. Furthermore, the compounds provided slight protection against tBid-induced cell death in HEK293 cells, indicating the compound’s protective potency independent of the cell type. Recent studies demonstrated a crucial role for the BH3-only protein Bid in the model of glutamate-induced oxidative stress in HT-22 cells where small interfering RNA mediated Bid gene silencing and pharmacological Bid inhibition provided neuroprotection in vitro (Landshamer et al., 2008). Further, these previous studies identified depolarization of the mitochondrial outer membrane, mitochondrial fission, and ATP depletion as key events of glutamate toxicity, and these effects were apparently also Bid dependent (Landshamer et al., 2008; Grohm et al., 2010, 2012). A pivotal role of Bid in neuronal cell death was further substantiated by previous findings that reduced Bid expression prevented cell death in a model of oxygen glucose deprivation in vitro and also reduced brain damage in models of cerebral ischemia and brain trauma in vivo (Plesnila et al., 2001; Culmsee et al., 2005). Moreover, recent reports showed that full-length Bid, as well as the truncated form tBid, is sufficient to induce mitochondrial dysfunction and cell death in neurons and that they differ only in kinetics and cell death pathways (Konig et al., 2007). Whereas tBid is suggested to induce rapid mitochondrial damage in a caspase-dependent manner, full-length Bid is thought to be involved in caspase-independent apoptosis with slower kinetics (Ward et al., 2006). In the described model system of glutamate toxicity in HT-22 cells, however, activated full-length Bid as well as tBid may contribute to neuronal cell-death pathways.
Overall, these findings in model systems of neuronal and non-neuronal death in vitro and in vivo strongly suggest that targeting Bid might be a promising tool to develop pharmaceutical drugs for neuroprotection. Indeed, it was recently demonstrated that pharmacological inhibition of Bid with the available inhibitor BI-6C9, which was used as positive control for neuroprotection in the present study, strongly reduced tBid-induced release of Smac/DIABLO from isolated mitochondria in concentrations as low as 20 µM in vitro and prevented effectively tBid-mediated cell death at concentrations of 10–50 µM in cell-based assays of caspase-dependent (Becattini et al., 2004) and caspase-independent cell death (Culmsee et al., 2005; Landshamer et al., 2008; Grohm et al., 2010; Tobaben et al., 2011). In particular, the available Bid-inhibitors reduced caspase-3 activity in tBid-transfected HeLa cells at 50 µM and persistently blocked caspase activity at 100 µM (Becattini et al., 2004). In our model systems in HT-22 neurons, the Bid inhibitor BI-6C9 provided protection against glutamate toxicity as well as tBid-induced death at a concentration of 10 µM. Notably, the newly synthesized compounds provided protective effects against glutamate-induced cell death at concentrations of 5 µM up to 50 µM and attenuated tBid-induced toxicity even at concentrations as low as 1 µM.
These data show great promise for applications in model systems of brain damage in vivo to determined not only their protective properties but also to examine whether the compounds cause side effects in animal models. It is noteworthy that the presented compounds belong to the structural class of thiazolidinediones, previously shown to bind to the peroxisome proliferator activator receptor (PPAR)γ, resulting in antidiabetic effects, but also in severe side effects. In contrast to all known thiazolidinedione-based PPAR ligands, which need to possess a free N-H for PPARγ binding, all of our thiazolidinediones are N-substituted. Therefore, we do not expect them to bind to PPARs. These compounds indicate future therapeutic perspectives for treatment of neurological diseases because of their pronounced potency to block Bid-dependent cell death. Although the precise mechanism by which the thiazolidinediones derivatives bind to and inhibit the activity of Bid remains to be determined, the presented data expose a novel class of potential drug candidates for diseases associated with neuronal injury that involve mechanisms of mitochondrial pathways of programmed cell death.
In conclusion, we report here the design and synthesis of N-phenyl–substituted thiazolidinediones as novel inhibitors against neurotoxicity and indicated inhibition of the BH3- only protein Bid as a key mechanism of the neuroprotective properties of the novel compounds. In particular, the thiazolidinediones derivatives provided significant protection against glutamate toxicity and tBid overexpression in hippocampal HT-22 cells and further preserved detrimental impairments of mitochondrial integrity and function, thereby inhibiting the final execution of intrinsic cell death. Future optimization of the structures should result in compounds with favorable pharmaceutical properties that are suitable for therapeutic strategies in the treatment of neuronal and non-neuronal diseases where Bid has been implicated (Plesnila et al., 2001; Guegan et al., 2002; Friedlander, 2003; Waldmeier, 2003; Culmsee and Plesnila, 2006).
The authors thank Emma Esser for careful editing of the manuscript and Dr. Alexander Seiler (Roche Diagnostics GmbH) for providing support with the xCELLigence system. The authors further thank Dr. Cornelius Krasel for excellent expertise on HEK293 cell transfection and Dr. Hans-Peter Steffens for providing support with the Enspire Multimode Plate Reader (PerkinElmer).
Participated in research design: Oppermann, Culmsee, Schlitzer.
Conducted experiments: Oppermann, Elsässer, Dolga, Doti.
Contributed new reagents or analytic tools: Schrader, Kraus.
Performed data analysis: Oppermann, Wegscheid-Gerlach, Culmsee.
Wrote or contributed to the writing of the manuscript: Oppermann, Culmsee, Schlitzer.
- apoptosis-inducing factor
- B-cell lymphoma-2
- Bcl-2 interacting domain death antagonist
- cell index
- electron ionization high-resolution mass spectrometry
- green fluorescent protein
- human embryonic kidney (cells)
- immortalized mouse hippocampal neurons
- mitochondrial respiratory capacity
- mass spectrometry
- 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
- numerical aperture
- normalized cell index
- oxygen consumption rate
- phosphate-buffered saline
- propidium iodide
- tBid expressing plasmid
- peroxisome proliferator activator receptor
- real-time cell analyzer
- small interfering RNA
- second mitochondria-derived activator of caspase/direct IAP binding protein with low pI
- truncated Bid
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics