HT-22 cells (Salk Institute, San Diego, CA, USA) were routinely grown in 75 cm² culture flasks with Dulbecco's Modified Eagle's Medium (DMEM) supplemented with fetal bovine serum (10%), penicillin G (150 IU/mL) and streptomycin (50 µg/mL). When reaching confluency, cells were trypsinized using 0.05% trypsin/0.53 mM EDTA solution and reseeded at one-fifth the initial density. For the experiments cells were seeded into the appropriate culture dishes, such as 24 to 96-well plates or on collagen-coated cover glasses in 6-well plates depending on the respective experiments.

Glycer-AGE was prepared as previously described9). Endotoxin-free bovine serum albumin (BSA, 10 mg/mL) was incubated with 100 mM D-glyceraldehyde in phosphate buffer (0.2 M, pH 7.4). The reaction mixture was filtered through microporous membrane filter (0.22 µm pore size) and then incubated for 7 days in the dark at 37℃. Low molecular reactants and unincorporated D-glyceraldehyde were removed by dialysis against 0.1 M phosphate buffer overnight at 4℃. BSA incubated at the same conditions in phosphate buffer without D-glyceraldehyde was used as BSA control. Every AGE preparation was tested for endotoxin by the limulus amebocytelysate assay, and stored at 4℃.

After exposure to experimental protocols, the cells were incubated with 50 µL of 5 mg/mL 3-(4,5-dimethyl-2-thiazyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) at 37℃ for 4 hr. The absorbance value proportional to the formation of formazan from MTT was determined at 540 nm using automatic multiwell spectrophotometer26). MTT reduction values were expressed as percentage of the BSA-control cells.

ATP levels were measured in cells with a luciferin-luciferase assay23). After exposure to experimental procedures, cells were solubilized with 500 µL of 0.5% Triton X-100 and acidified with 100 µL of 0.6 M perchloric acid and placed on ice. Cell suspension was then diluted with 10 mM potassium glutamate buffer containing 4 mM MgSO₄ (pH 7.4), and 100 µL of 20 mg/mL luciferin-luciferase was added to 10 µL of diluted sample. Light emission was recorded at 20 sec with a luminometer (MicroLumat LB96P, Berthold, Germany). Protein content was determined on a portion of the cell sample using Biorad protein assay kit.

Intracellular production of reactive oxygen species (ROS) was determined using DCFH-DA. Cells were preincubated with 10 µM DCFH-DA for 1 hr at 37℃, and then exposed to experimental procedures for various durations. DCF fluorescence was analysed using FACSort Flow Cytometer (Becton-Dickinson Bioscience, San Jose, CA, USA).

Mitochondrial transmembrane potential was measured with DiOC₆(3). Reduction in DiOC₆(3) staining indicates reduction of the mitochondrial transmembrane potential. Cells were loaded with DiOC₆(3) at a final concentration of 50 nM for 20 min at 37℃ in the dark. Cells were washed and resuspended in PBS. The fluorescence intensity was analyzed with a FACsort Becton Dickinson Flow Cytometer (Beckton-Dickinson Bioscience, San Jose, CA, USA).

To determine the formation of mitochondrial permeability transition (MPT) pore, a double staining method with fluorescent dyes, calcein/AM and TMRM was used as described by Lemasters et al.21).

The amount of cytochrome C released from the mitochondrial intermembrane space into the cytosol was determined by a western blot analysis. After washing with ice-cold phosphate-buffered saline, cells were resuspended in permeabilization buffer containing 75 mM NaCl, 1 mM NaH₂PO₄, 8 mM Na₂PO₄, 250 mM sucrose, 1 mM phenylmethylsulfonyl fluoride, additional protease inhibitors, and 0.05% digitonin. Following centrifugation at 15000 rpm at 4℃ for 10 min, the supernatant was separated from the pellet consisting of mitochondria and cellular debris. Equal amounts of protein were loaded on a 15% acrylamide gel and separated by SDS-PAGE. Nitrocellulose paper was then blocked with 0.5% nonfat dry milk, probed with 0.5 µg/mL polyclonal anti-cytochrome C antibodies (1 : 200, Santa Cruz Biotechnology, Dalas, TX, USA), then with secondary antibodies (rabbit) diluted 1 : 5000. Immunoblots were developed with ECL-plus reagents from Amersham according to the manufacturer's instructions.

A quantitative enzymatic activity assay was carried out according to instructions of the R&D (Minneapolis, MN, USA) colorimetric assay kit manufacturer.

Apoptosis was analyzed by staining with Hoechst-33258. For fluorescence microscopy, cells were grown on cover glasses. After exposure to experimental procedures, cells were fixed by 4% paraformaldehyde and loaded for 30 min at 37℃ with 5 µM Hoechst-33258 in HBSS. After washout, cover glasses were mounted on the slide glasses with PBS : Glycerol=1 : 1 solution. Apoptotic cells with condensed or fragmented nuclei were examined using a fluorescence microscopy.

Total RNA was prepared using Trizol reagent (Invitrogen, Carlsbad, CA, USA). RNA concentration was determined by spectrometry. Quantitative real-time PCR was performed with SYBR Green Mastermix (Invitrogen, Carlsbad, CA, USA) and 15 pmol primers for mouse HO-1, and β-actin. Reactions were performed in triplicate and specificity was monitored using melting curve analysis after cycling. Primers used were 5'-GGT GATGGCTTCCTTGTACC-3' (forward) and 5'-AGTGAG GCCCATACCAGAAG-3' (reverse) for HO-1, and 5'-AGAG GGAAATCGTGCGTGAC-3' (forward) and 5'-CAATAGT GATGACCTGGCCGT-3' (reverse) for β-actin.

Western blot analysis was performed to determine the HO-1 protein. Total intracellular protein was isolated by 5 times repeated freeze-thaw lysis in a buffer containing KCl (600 mM), Tris-Cl (20 mM), glycerol (20%), Pefabloc (0.4 mg/mL), leupeptin (10 µg/mL), pepstatin (10 µg/mL), aprotinin (5 µg/mL), and pH 7.8. Protein content was measured by Bradford assay (Bio-Rad Laboratory, Hercules, CA, USA).

Protein samples (40 µg) were electrophoresed on sodium dodecyl sulfate-polyacrylamide (10% gel) and transferred to immobile polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA). The membranes were then blocked with 5% skim milk in Tris-buffered saline containing 0.5% Tween-20 (TBST) for 1 hr at room temperature. Then the membranes were incubated with primary antibodies for HO-1 (StressGen Biotechnology, CA, USA) and β-actin (Gentest). The membranes were washed three times with TBST, and incubated with goat anti-mouse IgG-horseradish peroxidase with a 1 : 5000 dilution in TBST for 2 hr at room temperature. The membranes were washed three times for 10 min with TBS-T again and immunolabeling was visualized using enhanced chemiluminescent HRP substrate (ECL) detection method.

Transient transfection with siRNA and scrambled siRNA against human HO-1 (Santa Cruz Biotechnology, Dalas, TX, USA) were performed using the Superfect® transfection reagent (Qiagen, Hilden, Germany).

The data were presented as the means±standard error. Statistical analyses between two groups were performed by unpaired Student's t-test. Differences among groups were tested by one-way analysis of variance (ANOVA) followed by the Tukey's test. The p value less than 0.05 was accepted to be statistically significant.

In order to determine AGE-induced changes of mitochondrial functional integrity and the effect of sildenafil, MTT reduction ability was examined. There was a time-dependent decrease in the MTT reduction ability in AGE-treated cells. The effect of AGE was dose dependent in the range from 25 to 400 µg/mL (data not shown). Unless otherwise indicated in this study, we treated the cells with 200 µg/mL AGE for 48 hr.

The effect of anti-RAGE antibody was examined to evaluate whether the effect of AGE is ascribed to the interaction with its receptor, RAGE. In cells pretreated with anti-RAGE antibody (1 µg/mL) before exposure to AGE, AGE failed to inhibit MTT reduction ability suggesting that interaction with RAGE is crucial for the cellular actions of AGE (Fig. 1A).

In the presence of sildenafil (as sildenafil citrate, 20 µM), the AGE-induced decrease of MTT reduction ability was significantly attenuated. In the presence of an inactive cell-permeable cGMP analogue Rp-8-Br-cGMP the protective effect of sildenafil was significantly diminished, suggesting that sildenafil exerts its protective effect via a mechanism dependent on cGMP and/or cGMP-dependent protein kinases (PKG) (Fig. 1B).

As ATP is generated primarily by mitochondria through oxidative phosphorylation, intracellular ATP level is another indicator of intact mitochondrial function. There was a time-dependent decrease in intracellular ATP content in AGE-treated cells. In 3 days, intracellular ATP concentration in AGE-treated cells decreased to lower than 25% of its control value. However, in cells pretreated with of anti-RAGE antibody AGE failed to inhibit ATP production (Fig. 2A). Sildenafil significantly reversed the AGE-induced inhibition of ATP generation. In the presence of Rp-8-Br-cGMP the protective effect of sildenafil was significantly diminished (Fig. 2B).

Oxidative stress has been implicated in the AGE-induced neurotoxicity.

Thus to examine whether the protective effect of sildenafil against AGE-induced mitochondrial damage is associated with its effect on oxidative stress, we assessed the changes in ROS generation in AGE-treated cells in the presence and absence of sildenafil. After 30 min incubation with AGE, there was a remarkable increase in ROS formation, which was significantly suppressed in the presence of Rp-8-Br-cGMP (Fig. 3).

Disruption of the inner mitochondrial membrane potential associated with formation of permeability transition pore is a major event to initiate mitochondria-dependent apoptotic signaling21). Therefore, it was examined whether AGE-induced mitochondrial dysfunction is associated with the disruption of mitochondrial membrane potential. Fig. 4A, B represents flow cytometric analysis of DiOC₆(3)-stained cells. In these graphs left shift of cell population from strong (designated as M2) to weak (designated to M1) DiOC₆(3) fluorescence region indicates depolarization of inner mitochondrial membrane potential. The results showed that AGE depolarized the inner mitochondrial membrane potential. Sildenafil significantly reversed the AGE-induced disruption of the mitochondrial membrane potential. In the presence of Rp-8-Br-cGMP the effect of sildenafil was significantly diminished (Fig. 4C).

In confocal micrographs represented in Fig. 5A, B, intact mitochondria accumulated TMR and were discriminated as punctuated bright red spots, whereas injured mitochondria undergoing MPT got deprived of TMR, became permeable to calcein, and as a consequence, loose their visible contour. After 48 hr incubation with AGE, there was a remarkable increase in cells with MPT, and sildenafil significantly suppressed it. In the presence of Rp-8-Br-cGMP the effect of sildenafil was significantly attenuated (Fig. 5C).

Cytochrome C was detected by western blot analysis of the cytosolic fraction from cells treated with AGE for various time periods. In AGE-treated cells cytosolic release of cytochrome C increased significantly in a time-dependent manner, and sildenafil suppressed it (Fig. 6).

Activation of caspase cascades is critical for the execution phase of apoptosis. Caspase-9 is activated by cytochrome C released from mitochondria and thus crucial for the execution of the mitochondria-dependent apoptosis, which in turn activates caspase-3. In AGE-treated cells, 4.1 fold increase in caspase-3 activity was observed (Fig. 7A). In micrographs of cells stained with Hoechst-33258 represented in Fig. 7B, C, apoptotic cells are characterized by nuclear condensation and fragmentation. In parallel with caspase-3 activation there was a significant increase in apoptotic cell number (6.1±1.2% to 37.2±7.1% of total cell population). Sildenafil significantly suppressed both the AGE-induced caspasse-3 activation and apoptosis (Fig. 7D).

Several investigators reported that sildenafil increased heme oxygenase-1 (HO-1) expression and it might play a role in the action of sildenafil1231822). Hence, we examined whether sildenafil has any influence the HO-1 expression. qRT-PCR and Western blot analyses revealed that sildenafil significantly increased expression of both the HO-1 mRNA (Fig. 8A) and protein (Fig. 8B, C).

In order to evaluate whether the effect of sildenafil resulted from the promoted expression of HO-1, effects of HO-1 related molecules were examined. Cobalt protophorphyrin (CoPP), an inducer of HO-1 inducer24), and bilirubin, the metabolic product of HO-124), restored the AGE-induced decrease in MTT reduction ability. On the contrary zinc protoporphyrin IX (ZnPP IX), a HO-1 inhibitor24) significantly diminished the effect of sildenafil (Fig. 9A). Similar results were observed in the experiments to examine the effects on the AGE-induced MPT (Fig. 9B). Overall, these results strongly indicate that sildenafil-induced protective effects are closely related with the increased expression of the HO-1.

In order to confirm further the role of HO-1, we used cells transfected with siRNA. In cells transfected with the HO-1 siRNA the basal expression of the HO-1 protein was significantly diminished. In these cells, CoPP and sildenafil had no influence on the HO-1 expression (Fig. 10A). Likewise, in siRNA-transfected cells, sildenafil failed to restore the AGE-induced changes in MTT reduction ability and MPT induction (Fig. 10B, C). These results further support our hypothesis that promoted expression of HO-1 is closely related with the sildenafil-induced protection mechanism.