Nitric Oxide and Calcium Participate in the Fine Regulation of Mitochondrial Biogenesis in Follicular Thyroid Carcinoma Cells^*

Human Follicular Thyroid Carcinoma Cell Lines and Conditions of Cell Growth

Two human follicular thyroid carcinoma cell lines cultured at 37 °C, 5% CO₂ were used: FTC-133 and RO82 W-1 (Interlab Cell Line Collection, National Institute for Cancer Research, Genoa, Italy). FTC-133 cells were grown in Dulbecco's modified Eagle's medium (DMEM)/F-12 (Jacques Boy, Reims, France), and RO82 W-1 cells were grown in a mixed 60:30 DMEM/F-12 plus endothelial basal medium (PAA, Pasching, Austria). Both media were supplemented with 10% fetal bovine serum (Jacques Boy), 1% l-glutamine (Dominique Dutscher, Brumath, France), and an antibiotic mixture (Thermo Fisher Scientific, Waltham, MA).

To study the effects of NO, the two cell lines were treated once a day for 4 days with the nitric oxide donor S-nitroso-N-acetyl-dl-penicillamine (SNAP) (Merck Calbiochem, Darmstadt, Germany) made up to a final concentration of 50 or 100 μm (corresponding to 50 and 100 nm equivalent NO during several h) in the selected medium or with equivalent volumes of DMSO (Sigma-Aldrich) for the control cells. We verified that the 4.6-h half-life of SNAP was sufficient to increase nitrite production by 2.3-fold (50 μm) and 4.1-fold (100 μm) after 24 h of treatment. To control the mediating effect of NO/cGMP on PRC-related mitochondrial biogenesis, cells were pretreated with a 1 μm concentration of the protein kinase G inhibitor KT5823 (EMD, San Diego, CA) 30 min before 50 μm SNAP treatment during 96 h.

To study the effects of calcium on mitochondrial biogenesis and function, RO82 W-1 cells were treated 5 h/day for 4 days with the calcium ionophore ionomycin (Merck Calbiochem) made up to a final concentration of 2 μm in the culture medium. We further used the cytosolic calcium chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid-acetoxymethyl ester (BAPTA/AM) (Merck Calbiochem) made up to a final concentration of 2–8 μm, with a combination of ionomycin and BAPTA/AM or with equivalent volumes of DMSO for control cells. To avoid cell death due to the toxic effects of intracellular Ca²⁺ ion chelation, the BAPTA/AM concentration was adjusted to cell density in the flask for each day of treatment as follows: 2 μm the first day, 4 μm the second day, 6 μm the third day, and 8 μm the fourth day. Duplicate cell counts were made with a Z1 Coulter particle counter (Beckman Coulter, Fullerton, CA). To control the mediating effect of calcium on PRC-related mitochondrial biogenesis, cells were pretreated with a 10 μm concentration of the calcineurin inhibitor CN585 (EMD) during ionomycin and BAPTA/AM treatments.

Quantitative RT-PCR Analysis

Total RNA was isolated from cultured cells using the RNeasy kit (Qiagen, Hilden, Germany). The integrity of RNA was determined using a Bio-Analyzer 2100 (Agilent Technologies, Waldbronn, Germany). Reverse transcription was performed on 2 μg of RNA with the Advantage RT-for-PCR kit (Clontech), following the manufacturer's recommendations. cDNA real-time quantifications were performed in a 96-well plate on a Chromo4 apparatus (Bio-Rad) using SYBR Green I dye as a fluorescent signal (iQ SYBR Green Supermix, Bio-Rad) according to the manufacturer's instructions. The expression of eight selected genes (CREB1, cytochrome c (CYT C), ERRα, NRF-1, PGC-1α, PRC, SOD2, and UCP2) was measured to determine the impact of calcium and NO modulation on key factors controlling mitochondrial biogenesis and reactive oxygen species production. For RO82W-1 cells treated with ionomycin and BAPTA/AM, some of the most differentially expressed genes on microarray, such as DRP1, TEP1, CRYAB, FIS1, PDE1A, GREB1, TFAM, RINT1, and MFN2, were controlled by quantitative RT-PCR. The amount of cDNA for selected genes was normalized by the quantification of β₂-microglobulin (B2M) or β-globin (for microarray confirmation) cDNA. Each sample was assayed in duplicate using the primers described in Table 1.

Western Blot Analysis

Dry cell pellets were mixed with a buffer containing 10 mmol/liter HEPES (pH 7.9), 1.5 mmol/liter MgCl₂, 10 mmol/liter KCl, 0.5% Nonidet P-40, and a protease inhibitor mixture and then centrifuged at 13,000 × g for 5 min. Supernatants were collected, and the protein concentration was determined by the bicinchoninic assay kit (Uptima, Interchim, Montluçon, France) with bovine serum albumin as standard. Samples containing 20 μg of total protein were applied to 12% SDS-polyacrylamide gel electrophoresis and hybridized with either cytochrome c (Ab13575) or α-tubulin (Ab7291) antibodies at a dilution of 1:5000 (both from Abcam, Paris, France). Cytochrome c was chosen as validated target gene for direct regulation by PRC and PGC-1α coactivators (5). After overnight incubation with primary antibodies, membranes were washed before incubation with the corresponding horseradish peroxidase (HRP)-conjugated secondary antibody, which was detected with a chemiluminescent detection system (ECL Plus, Amersham Biosciences). All of the assays were performed in triplicate.

Enzymatic Activities

The activities of COX, citrate synthase (CS), and LDH were measured on cell lysates, at 37 °C, in a cell buffer (250 mm saccharose, 20 mm Tris, 2 mm EGTA, 1 mg/ml bovine serum albumin, pH 7.2) (50 μl/10⁶ cells) using a Beckman U 640B spectrophotometer (Beckman Coulter). COX activity was measured in a 100 mm KH₂PO₄ buffer containing 1 mg/ml bovine serum albumin using 15 μm reduced cytochrome c and 2.5 mm β-D dodecylmaltoside (25). The CS activity was measured in a reaction medium consisting of 0.15 mm 5,5′-dithiobis(2-nitrobenzoic acid), 0.5 mm oxaloacetic acid, 0.3 mm acetyl-CoA, and Triton X-100 (0.1%). The LDH activity was measured in a reaction medium consisting of 95 mm Tris, 10 mm pyruvate, and Triton X-100 (0.25%) using 300 μm NADH. Specific enzymatic activities were expressed in mIU (i.e. nmol of cytochrome c, 5,5′-dithiobis(2-nitrobenzoic acid), NADH/min/mg protein, respectively). The cellular protein content was determined using the bicinchoninic assay kit with bovine serum albumin as a standard. All of the reagents were obtained from Sigma-Aldrich except for Tris and saccharose (Eurobio, Les Ulis, France).

Measurement of Free Intracellular [Ca²⁺]

FTC-133 and RO82 W-1 cells were seeded in a 96-well plate to a density of 50,000 cells/well 15 h before the experiment and were grown in the selected medium. Cells were then loaded with 5 μm FURA-2/AM (Merck Calbiochem) and 0.1% pluronic acid F-127 (Invitrogen) at 37 °C for 40 min in a PSS buffer (140 mm NaCl, 4 mm KCl, 1 mm MgCl₂, 330 μm NaH₂PO₄, 10 mm Hepes, 11.1 mm glucose, pH 7.4) with or without 2 mm CaCl₂ (Redundant), supplemented with 0.1% bovine serum albumin (PSS/BSA). Cells were washed once with PSS/BSA and incubated for 30 min at 37 °C in 200 μl of PSS/BSA to allow the complete de-esterification of the dye. The buffer was then removed, and cells were loaded with 2 to 8 μm of BAPTA/AM at 37 °C for 40 min in PSS/BSA. Cells were washed once with PSS/BSA and incubated for 30 min at 37 °C in 200 μl of PSS/BSA. The buffer was then removed, and 90 μl of PSS/BSA were added to each well before kinetic measurement of fluorescence intensity at 37 °C using a Mithras LB940 spectrophotometer (Berthold Technologies, Bad Wildbad, Germany). For each well, the cell preparation was alternatively excited for 30 s at two wavelengths, 340 and 380 nm, and the fluorescence emission was detected at 510 nm with a counting time of 150 ms. The ionomycin solution was then injected to attain a final concentration of 2 μm/well, and the plate was shaken before measurement of fluorescence intensity at 510 nm for 160 s. The background was determined by measuring the fluorescence of a well loaded without FURA-2/AM and BAPTA/AM. The fluorescence intensity ratio, F₃₄₀/F₃₈₀, was used to evaluate variations in cytosolic calcium concentration ([Ca²⁺]_c).

Microarray Analysis

cDNA preparations obtained from RO82 W-1 cells after treatments with ionomycin and BAPTA/AM were hybridized in duplicate on human 4 × 44,000 expression chips (Agilent Technologies, Santa Clara, CA) after RNA amplification and cDNA labeling using Cy3/Cy5 dyes according to the manufacturer's recommendations. Slides were analyzed with the feature extraction software after scanning on the DNA microarray scanner (Agilent Technologies). Data are available in the GEO data base (GSE 26237). The hierarchical gene clustering was computed on median gene-centered and log-transformed data using average linkage and uncentered correlation distances. Computations and visualization were performed using Cluster and TreeView software (26). The Expression Analysis Systematic Explorer (EASE) and Gene Set analysis were used to determine the statistical over represented and differentially expressed genes (27). Gene ontology enrichments in gene lists were searched for by means of the GOMiner (28). The most abundant gene ontology terms, representing at least 5% of the genes in the lists, with p values lower than 0.05, were considered for interpretation.

Confocal Microscopy

FTC-133 and RO82 W-1 cells were grown on coverslips and treated with either ionomycin (4 days at 2 μm) or BAPTA/AM (days 1–2, 2 μm; day 3, 4 μm; day 4, 6 μm). Mitochondrial staining was performed by incubating cells with MitoTracker Red CMXRos (100 nm) for 15 min at 37 °C, and endoplasmic reticulum staining was performed by incubating cells with ER-Tracker Green FM (1 μm) for 20 min at 37 °C. Cells were subsequently washed twice in PBS and fixed for 15 min in 3.7% paraformaldehyde, and nuclei were stained with 300 nm DAPI (Invitrogen). Cells were mounted with Vectorshield (Vector Laboratories, Burlingame, CA) and analyzed on a Zeiss LSM700 scanning microscope (Carl Zeiss, Göttingen, Germany) equipped with a ×63 oil immersion objective. The 555-nm laser was used for excitation of MitoTracker Red, the 488-nm laser was used for ER-Tracker Green, and the 405-nm laser was used for DAPI. Image analysis was performed using Zeiss Zen-light 4.8 software.

Statistical Analysis

Data are represented as mean values ± S.D., with n representing the number of experiments. The statistical significance of the variations observed was assessed using the Mann-Whitney test. Differences were considered significant at p < 0.05. Significances of differential expression between microarray and quantitative RT-PCR results were assessed by the t test. All analyses were performed using StatView 5.0 (SAS Institute, Gary, NC).

Effects of NO on Cell Proliferation and Mitochondrial Function in FTC-133 and RO82 W-1 Cells

At 96 h of SNAP treatment, cell proliferation decreased by 41.5% (50 μm SNAP) or 74.0% (100 μm SNAP) in FTC-133 cells and by 56.0% (50 μm SNAP) or 80.9% (100 μm SNAP) in RO82 W-1 cells, compared with time 0 (Fig. 1A). We controlled that the decrease in cell proliferation was due to quiescence rather than to cell death. Thus, the inhibitory action of SNAP was similarly effective on cell proliferation in the oxidative and the glycolytic cell lines.

To evaluate the effects of SNAP on mitochondrial function and glycolytic metabolism, we measured the enzymatic activities of mitochondrial complex IV (COX) and LDH (Fig. 1B). Whereas the inhibition of COX activity was proportional to SNAP concentration at 24 h in both cell lines, differences were observed at 96 h; with 100 μm SNAP, a toxic effect was observed in RO82 W-1 cells with complete inhibition of COX activity and enhancement of LDH activity. For FTC-133 cells, the inhibition of COX activity was proportional to the duration of treatment and the SNAP concentration, but the remaining COX activity was sufficient to maintain an oxidative metabolism with no increase of LDH activity. Conversely, the basal COX activity was very low in RO82 W-1 cells compared with FTC-133 cells (24).

Thus, 50 μm SNAP induced a greater decrease of COX activity at 96 h in FTC-133 cells, which mainly depend on oxidative phosphorylation to produce energy, compared with the glycolytic RO82 W-1 cells. The decrease in FTC-133 cell proliferation was proportional to that of COX activity, whereas the decrease in RO82 W-1 cell proliferation could not be exclusively attributed to changes in mitochondrial metabolism.

Effects of NO on Mitochondrial Biogenesis in FTC-133 and RO82 W-1 Cells

To determine the role of the NO pathway in the induction of mitochondrial biogenesis, we investigated the effects of SNAP on the transcriptional efficiency of several genes involved in mitochondrial function, transcriptional regulation, and coactivation (Fig. 2A). Because the 100 μm SNAP treatment led to a toxic effect at 96 h in RO82 W-1 cells, we explored the NO effect only for the 50 μm SNAP treatment.

At 96 h of SNAP treatment, PRC expression was significantly higher in both cell lines compared with controls. However, this increase was greater in RO82 W-1 cells (108.5%) than in FTC-133 cells (48.0%). PGC-1α mRNA expression was also enhanced at 96 h of SNAP treatment but at a lower level than that of PRC: 72.8% in RO82 W-1 cells and 27.6% in FTC-133 cells. Among the transcription factors, CREB, NRF-1, and ERRα, known to be involved in either PGC-1α- or PRC-dependent mitochondrial biogenesis, NRF-1 and ERRα were overexpressed in RO82 W-1 cells at 96 h, whereas CREB and NRF-1 were overexpressed in FTC-133 cells. Interestingly, SNAP enhanced the mRNA expression of ERRα by 58.7% at 96 h only in RO82 W-1 cells in which the PRC/PGC-1α ratio as well as the induction of PRC were higher.

The expression of cytochrome c, a marker of mitochondrial biogenesis, was exclusively increased in RO82 W-1 cells, by 89.2% for mRNA and by 1.38-fold for protein at 96 h of SNAP treatment (Fig. 2B). No change in cytochrome c expression was found in FTC-133 cells at the same time. Thus, in these thyroid cell lines, NO induced mitochondrial biogenesis more efficiently through the PRC than the PGC-1α coactivator pathway. The overexpression of UCP2 in RO82 W-1 cells reflects the reduction in the efficiency of ATP production while mitochondrial biogenesis is being induced.

Effects of Ionomycin and BAPTA/AM on [Ca²⁺]_c in FTC-133 and RO82 W-1 Cells

We measured the intracellular calcium response to the calcium ionophore ionomycin, in FTC-133 and RO82 W-1 cells (Fig. 3). Ionomycin injection (2 μm) led to a rapid increase in cytosolic calcium concentration. This increase could be due either to the generation of Ca²⁺-permeable pores in the plasma membrane, allowing the influx of Ca²⁺ from the extracellular medium, or to the induction of an efflux of Ca²⁺ from intracellular stores. When cells were incubated in PSS buffer containing 2 mm Ca²⁺, the calcium peak was 1.7-fold higher in FTC-133 cells than in RO82 W-1 cells, whereas in PSS containing 0 mm Ca²⁺, which revealed the part of ionomycin-induced [Ca²⁺]_c increase due to the efflux from intracellular stores, the calcium peak in FTC-133 cells was similar to that in RO82 W-1 cells (Fig. 3A). Thus, in FTC-133 cells, the increase in [Ca²⁺]_c was due in equal parts to the Ca²⁺ efflux from intracellular stores and to the extracellular Ca²⁺ entry. In contrast, in RO82 W-1 cells, ionomycin mainly induced the Ca²⁺ efflux from intracellular stores.

When RO82 W-1 cells were preincubated with the cytosolic calcium chelator BAPTA/AM (Fig. 3B), the response to 2 μm ionomycin was partly reduced with 2 μm BAPTA/AM but strongly diminished with 4, 6, or 8 μm BAPTA/AM. Conversely, in FTC-133 cells, preincubation with BAPTA/AM (2–8 μm) failed to reduce the response to 2 μm ionomycin. This may be explained by the low cytosolic calcium store in RO82 W-1 cells.

Effects of Calcium on Cell Proliferation and Mitochondrial Biogenesis and Function in RO82 W-1 Cells

In RO82 W-1 cells, after 4 days of intermittent treatment, 2 μm ionomycin or 2–8 μm BAPTA/AM decreased cell proliferation by 33.5 and 44.5%, respectively, whereas combined treatment decreased it by 60.6% (Fig. 4A).

Concerning the effects of intracellular calcium levels on mitochondrial function, there was a drastic (73.5%) inhibition of COX activity in ionomycin-treated cells, whereas the chelation of cytosolic Ca²⁺ by BAPTA/AM had no significant effect (Fig. 4B). After combined treatment with ionomycin and BAPTA/AM, COX activity was also strongly decreased, as observed after ionomycin treatment alone. Thus, our results suggest that COX activity depends closely on the mobilization of calcium stores in RO82 W-1 cells.

The effects of the variation of cytosolic Ca²⁺ levels on the transcription of genes involved in mitochondrial biogenesis and function are shown in Fig. 4, C and D. After 4 days of intermittent treatment, ionomycin and BAPTA/AM induced opposite effects on gene transcription in RO82 W-1 cells. Ionomycin reduced the expression of PRC, CREB, NRF-1, CYT C, and UCP2 by 15.0, 13.9, 22.1, 15.7, and 19.0%, respectively, whereas the expression of ERRα remained unchanged. In contrast, BAPTA/AM increased the expression of PRC, CREB, ERRα, CYT C, and UCP2 by 58.5, 25.0, 33.6, 48.3, and 24.2%, respectively, whereas neither ionomycin nor BAPTA/AM modified the expression of PGC-1α mRNA. These results showed that the increase in [Ca²⁺]_c led to the negative regulation of mitochondrial biogenesis at the transcriptional level, independent of a PGC-1α effect. Similar results were obtained after treatment with BAPTA/AM alone or combined treatment with ionomycin and BAPTA/AM (data not shown). In accordance with mRNA results, BAPTA/AM treatment induced a 6.2-fold increase in the cytochrome c protein level compared with controls (Fig. 4C). These results suggest that cytosolic calcium levels play an important role in the regulation of mitochondrial biogenesis.

Effects of Calcium on Mitochondrial and Endoplasmic Reticulum Networks in RO82 W-1 Cells

Regarding pangenomic microarray results and main gene ontologies (supplemental Fig. 1), we observed that ionomycin treatment induced the differential expression of 55 genes, whereas BAPTA/AM treatment induced the differential expression of 169 genes. Using quantitative RT-PCR analysis, we verified that seven (DRP1, TEP1, CRYAB, TFAM, PDE1A, GREB, and RINT1) of the eight most differential genes identified on microarray were confirmed in their expression (Fig. 5A). Among significant ontologies, no pathways involving NO were identified with ionomycin and BAPTA/AM treatments, suggesting that the NO and calcium pathways were not interrelated in the RO82 W-1 cells. Microarray analysis revealed that mitochondrial dynamics was regulated by calcium in RO82 W-1 cells. The expression of three genes involved in mitochondrial fusion (MFN2) and fission (DRP1 and FIS1) was specifically analyzed by quantitative RT-PCR after ionomycin and BAPTA/AM treatments (Fig. 5B). The fission/fusion gene expression ratio (DRP1/MFN2) was in favor of fission activation in the case of BAPTA/AM treatment (ratio 3.6) compared with ionomycin (ratio 1.2) and controls (ratio 1).

To better understand the role of calcium distribution in cellular compartments in mitochondrial biogenesis and function in RO82 W-1 cells, we examined the mitochondrial and endoplasmic reticulum networks after ionomycin or BAPTA/AM treatments (Fig. 5C). The BAPTA/AM treatment led to fragmentation of the mitochondrial network (Fig. 5C, a) independent of the modification of the endoplasmic reticulum (Fig. 5C, b and c). This fragmentation affecting specifically the mitochondrial network could be due to an imbalance between the fission and fusion processes. Thus, the remodeling of the mitochondrial network and the availability of mitochondrial calcium following BAPTA/AM treatment may lead to the induction of mitochondrial biogenesis.

Molecular Pathways Involved in NO- and Ca²⁺-mediated Mitochondrial Biogenesis in RO82 W-1 Cells

To explore whether variations of intracellular NO and calcium concentrations were causative or correlative of mitochondrial biogenesis regulation, we examined the effects of kinase or phosphatase inhibitors on the expression of key genes for mitochondrial biogenesis. When the NO/cGMP pathway was inhibited by the protein kinase G inhibitor KT5823 at 1 μm, the expression of PRC, PGC-1α, NRF-1, and CYTC was significantly down-regulated (Fig. 6A). This showed the involvement of the NO/cGMP pathway in the regulation of PRC and PGC-1α expression impacting mitochondrial biogenesis.

Because activation of the calcineurin pathway was recently shown to be involved in the translocation of DRP1 to mitochondria and in nuclear translocation of a number of transcription factors in response to mitochondrial respiratory stress (29, 30), we have chosen to explore the effects of the calcineurin inhibitor CN585 on the regulation of mitochondrial biogenesis by calcium. The expression of PGC-1α and NRF-1 was significantly increased when CN585 at 10 μm was added during ionomycin treatment (Fig. 6B). However, this was not associated with significant variation of CYT C expression. On the contrary, the addition of CN585 calcineurin inhibitor during BAPTA/AM treatment led to significant down-regulation of PRC, NRF-1, and CYT C expression without variation of PGC-1α expression. This showed the involvement of calcineurin in the regulation of PRC-related mitochondrial biogenesis induced by BAPTA/AM treatment.