Male Wistar rats (weighing 300–400 g) were sacrificed using an intraperitoneal injection of pentobarbital (150 mg/kg) according to the animal care and use local committee (Comité d'éthique régional d'Aquitaine). All the experiments were carried out in accordance with the recommendations of the Comité d'éthique régional d'Aquitaine (CEEA 50) and the protocol was approved by the same committee (protocol n° 50110016-A). The left lung was rapidly removed and rinsed in Krebs–HEPES–bicarbonate (KHB) containing (in mM): 118.4 NaCl, 4.7 KCl, 1.2 MgSO₄, 4 NaHCO₃, 1.2 KH2PO₄, 2 CaCl₂, 10 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) and 6 D-glucose, pH 7.4 with NaOH. Intrapulmonary arteries (IPA) of first, second and third order with an external diameter ranging from 300 μm to 2 mm were then dissected free from surrounding connective tissues under binocular control.

EPR recordings were performed as previously described (Billaud et al., 2009). IPA were incubated in the spin trap solution containing 500 μM 1-hydroxy-3-methoxycarbonyl-2,2,5,5-tetramethylpyrrolidin (CMH, Noxygen), 25 μM deferoxamine (Sigma), and 5 mM N,N-diethyldithiocarbamate (DETC, Sigma) in KHB at 37°C for 45 min. 5-HT was added during the spin trap incubation. When indicated, pharmacological inhibitors were also added 30 min prior and during the spin trap incubation. The reaction was stopped by snap freezing the samples in liquid nitrogen. Samples were then analyzed by EPR spectrometry using a tabletop X-band spectrometer miniscope (MS200, Magnettech). Spectra of the oxidized product of CMH (CM^•) were recorded at 77°K using a flask Dewar. Acquisition parameters were as followed: Bo Field: 3341 ± 150 G, microwave power: 10 dB, amplitude modulation: 5 G, sweep time: 60 s, gain: 300 and 3 scans. Signals were quantified by measuring the total amplitude of the signal, after correction of baseline and normalized to the protein quantity of the sample in mg/ml. O2•_ level was expressed as [A/(mg/ml) of proteins], where A corresponds to arbitrary units.

IPA were incubated with or without 5-HT 100 μM in phosphate buffer saline solution during 45 min. Arteries were then homogenized in this incubation medium and centrifuged (10500 g, 20 min at 4°C). Supernatants were further used for the measurement of peroxynitrites. As peroxynitrites are rapidly transformed to their more stable structural isomer, nitrates, supernatants were first incubated with nitrate reductase (Cayman chemicals) to convert nitrates into nitrites. Levels of peroxynitrites were then quantified from nitrites with Griess reagent according to the manufacturer's instructions (Molecular probes). Peroxynitrites levels were normalized to tissue protein content in mg/ml. Results are expressed as [(μM)/(mg/ml) of proteins].

IPA treated or not with 5-HT were homogenized in HEPES buffer containing (in mM): 20 HEPES, 1 EGTA, 210 mannitol, 70 sucrose, pH 7.2 with NaOH. Homogenized samples were centrifuged (1500 g, 5 min at 4°C) and supernatants were used for the determination of superoxide dismutase (SOD) activity. SOD enzyme activity was then determined as described in the procedure of the SOD Assay Kit-WST from Sigma and as carried out in previous studies (Peskin and Winterbourn, 2000). Briefly, the method is based on the inhibition of the SOD activity in presence of the highly water-soluble tetrazolium salt WST-1 [2-(4-Iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt] that produces a colored water-soluble formazan dye (WST1-formazan) upon reduction with O2•_. Levels of SOD activity were determined by measuring the absorbance of the WST1-formazan dye at 450 nm with a microplate reader (EL808, Bio-Tek instruments). SOD standard solutions ranging from 0.001 to 200 Units/ml were used to perform a calibration curve and SOD activity value was read from this curve and expressed as Units/ml for each sample. Tissue protein content of each sample was also quantified in mg/ml to normalize SOD enzyme activity, expressed as Units/mg of proteins.

Briefly, IPA were lysed in RIPA buffer (Sigma) for total protein extraction. The lysed tissues were centrifuged at 15,000 g for 10 min at 4°C. The amount of total protein in the supernatant was measured using a lowry assay (Biorad DC protein assay) following the manufacturer's instructions. The amount of protein in each sample was compared to a standard curve performed with bovine serum albumin (BSA, 0–2 mg/ml).

PASMC were obtained as previously described (Martin et al., 2012). Briefly, IPA from the rat left lung were dissected free from surrounding connective tissues. IPA were cut into small pieces and placed in Hanks Balanced Salt Solution containing 50 μM CaCl₂, 0.5 mg/ml papain, 0.3 mM dithioerythritol, and 0.3 mg/ml collagenase for 10 min at 37°C. IPA were then mechanically and gently agitated using a polished wide-bore Pasteur pipette to release the cells. PASMC were seeded onto 14 mm round glass coverslips, maintained in culture medium (DMEM) supplemented with 1% penicillin– streptomycin, 1% sodium pyruvate, 1% non essential amino acids, and 10% fetal calf serum. Cells were stored at 37°C in a humidified atmosphere gassed with 5% CO₂. When the PASMC reached 80% confluence, they were growth arrested by using serum-free culture medium supplemented with 1% insulin–transferrin–selenium. After 48 h, recordings of cytosolic and mitochondrial calcium, mitochondrial membrane potential, and mitochondrial O2•_ measurements with fluorescent dyes were performed. For mitochondrial respiration assessment, PASMC were seeded in 175-cm² flasks and grown until confluence. Cells were used up to passage 4. PASMC phenotype was confirmed by positive immunostaining for α-smooth muscle actin and calponin.

Endogenous cellular oxygen consumption was monitored on a suspension of PASMC at 37°C in a 1 ml thermostatically controlled chamber (1.0 × 10⁶ cells/ml/run) equipped with a Clark oxygen electrode (Oxygraph system, Hansatech) as previously described (Rossignol et al., 2004). The respiratory buffer was the DMEM growth medium without serum. Cumulative concentrations of 5-HT were added to the chamber when indicated. The basal respiratory rate is expressed as nmol O₂/min/10⁶ cells and the effect of 5-HT was calculated as a percentage of the basal respiratory rate.

Cytosolic and mitochondrial calcium were assessed by simultaneously loading PASMC with the non ratiometric fluo4-AM probe (2 μM) and rhod-2AM probe (1 μM) respectively. PASMC were also loaded simultaneously with rhod-2AM (1 μM) and mitotracker green (400 nM), a mitochondrial marker, for mitochondria labeling. For mitochondrial membrane potential measurements, PASMC were loaded with tetramethylrhodamine methyl ester (TMRM) (100 nM). All the fluorescent dyes used were obtained from Molecular probes (Invitrogen). For all the above mentioned experiments, PASMC were loaded with the dyes for 30 min at 37°C in KHB solution and then washed in KHB solution without the dye for 30 min to allow deesterification of the dye. PASMC were then mounted on the stage of a laser scanning confocal microscope (TE2000, Nikon, Champigny-Sur-Marne, France) with a x 20, 1.40 NA plan apochromat oil-immersion objective. The cells were under continuous bath perfusion with KHB solution at a rate of 2 ml/min and 5-HT 100 μM was bath-applied. Fluo-4AM was excited with an argon laser at 488 nm and emitted light was filtered at 515 ± 30 nm. Rhod-2AM and TMRM were excited with a helium-neon laser at 543 nm and the emitted light was filtered at 605 ± 75 nm. Regions of interest were drawn around each cell and fluorescence was recorded every 2 s by using the EZ-C1 software (Nikon, Champigny-Sur-Marne, France). All experiments were performed at room temperature. The fluorescence values (F) were normalized to the basal fluorescence (fluorescence before application of the agonists or F0) to obtain the fluorescence ratio (F/F0). The area under the curve was calculated with Origin 6.0 software (Microcal). The area under the curve was calculated for a 75 s duration, starting 45 s after the beginning of the recording, which corresponds to the time where the agonist was applied. Results are expressed as (F/F0)s⁻¹.

Mitochondrial O2•_ production was measured with mitoSox red (molecular probes). PASMC were loaded with mitoSox red (4 μM) in KHB for 20 min at 37°C. PASMC were washed twice and mounted on the stage of a laser scanning confocal microscope (TE2000, Nikon, Champigny-Sur-Marne, France) with a x 20, 1.40 NA plan apochromat oil-immersion objective. 5-HT (100 μM) was applied at t = 0 min. MitoSox red was excited with an helium-neon laser at 543 nm and the emitted light was filtered at 605 ± 75 nm. Regions of interest were drawn around each cell and fluorescence was recorded every 15 min for 60 min by using the EZ-C1 software (Nikon, Champigny-Sur-Marne, France). The fluorescence values at the start of the experiment (F0) were subtracted from the fluorescence at t = 60 min (F) to calculate the percentage of increase in mitoSox red fluorescence.

All drugs were diluted in distilled water, except for LOE-908, nicardipine, RHC-80267, and rotenone, which were dissolved in dimethyl sulphoxide (DMSO). Antimycin A was dissolved in ethanol. The maximal concentration of DMSO and ethanol used in experiments was <0.1%.

All results are expressed as mean ± SEM; n indicates the number of rats for EPR, peroxynitrites measurements, SOD activity and mitochondrial respiratory rate, the number of cells for cytosolic, mitochondrial calcium, mitochondrial membrane potential, and mitochondrial superoxide anion recordings. Experiments were performed on a minimum of 3–4 rats. Unpaired t-test was used to compare data obtained from measurements of cytosolic, mitochondrial calcium levels, and mitochondrial membrane potential. Non-parametric one way ANOVA (Kruskal-Wallis) followed by Dunn's multiple comparison tests was used to compare mitochondrial respiration. All other data were analyzed using a non-parametric test for unpaired samples (Mann-Whitney test). All bar graphs and statistics were performed with GraphPad Prism 5. Values of P < 0.05 were considered significant.

We previously demonstrated that 5-HT had a maximal contractile effect at the concentration of 100 μM in rat IPA (Billaud et al., 2009). Moreover, the maximal effect of the anti-oxidant PEG-SOD and PEG-catalase was observed on the contraction to 5-HT 100 μM (Billaud et al., 2009). Therefore, we focused our study on O2•_ produced by 5-HT 100 μM. Indeed, as previously demonstrated (Billaud et al., 2009), a new set of experiments has been performed with electron paramagnetic resonance (EPR) recordings confirming that 5-HT (100 μM) significantly increased O2•_ levels in rat IPA (n = 13 and 12) (Figure 1A). Similarly, 5-HT (100 μM) also significantly increased O2•_ levels in rat aorta (n = 7) (Figure 1B). Since O2•_ can be scavenged by NO to form peroxynitrites (ONOO⁻), we measured ONOO⁻ concentration in rat IPA with Griess reagent as an alternative means to investigate 5-HT-induced O2•_ increase. 5-HT significantly increased ONOO⁻ levels (Figure 1C). Moreover, in rat IPA, 5-HT had no significant effect on the activity of superoxide dismutase, an anti-oxidant enzyme responsible for O2•_ degradation (Figure 1D). Altogether, 5-HT produces O2•_ in IPA and aorta and modification of superoxide dismutase activity cannot explain 5-HT-induced O2•_ increase in IPA.

Pulmonary arterial contraction in response to 5-HT is known to be mediated by O2•_, 5-HT receptors (5-HT_2A and 5-HT_1B/D) and, to a lesser extent, by the 5-HT transporter (Morecroft et al., 2005; Billaud et al., 2009). We have thus used selective pharmacological inhibitors and activators to study the role of the 5-HT transporter, 5-HT₂, and 5-HT_1B/D receptors involved in IPA contraction. EPR recordings showed that ketanserin (1 μM), a 5-HT₂ receptor antagonist significantly decreased the amount of O2•_ produced by 5-HT (n = 9) (Figure 2A) whereas blocking the 5-HT_1B/D receptors or the 5-HT transporter by GR 127935 (1 μM) or citalopram (1 μM), respectively, had no effect (n = 11–12) (Figure 2A). Moreover, α-methyl 5-HT (10 μM), a non-selective 5-HT₂ agonist significantly increased O2•_ levels (n = 6) (Figure 2B).

Given the importance of calcium in the contraction to 5-HT in rat IPA, we have then studied the role of intracellular calcium by depleting the calcium from the sarcoplasmic reticulum with a calcium free KHB solution containing a calcium chelator (0.4 mM EGTA), an activator of ryanodine receptors (caffeine 5 mM) and an inhibitor of the sarcoplasmic reticulum Ca²⁺-ATPase (thapsigargin 1 μM) to prevent calcium re-uptake in the sarcoplasmic reticulum. Such protocol has been shown to fully deplete sarcoplasmic reticulum in calcium in rat pulmonary arteries (Gonzalez De La Fuente et al., 1995; Guibert et al., 1996). Once the sarcoplasmic reticulum was depleted in calcium, extracellular calcium was reintroduced in the incubation medium allowing the activation of store-operated calcium permeable channels in the plasma membrane. The role of extracellular calcium was studied by incubating IPA in calcium free KHB solution containing 0.4 mM EGTA. Unlike the absence of calcium from sarcoplasmic reticulum, absence of extracellular calcium significantly reduced 5-HT-induced O2•_ increase (n = 12–13) (Figure 3A). Moreover, absence of both intra and extracellular calcium did not further affect 5-HT-induced O2•_ production (n = 7) (Figure 3A).

We have then investigated the role of the voltage-dependent and voltage-independent calcium channels on 5-HT-induced O2•_ increase by using their respective blockers in rat IPA. Nicardipine (1 μM), an L- type voltage-dependent calcium channels blocker, did not prevent O2•_ increase in response to 5-HT (Figure 3B). LOE-908 (10 μM), a voltage-independent calcium channel inhibitor and RHC-80267 (50 μM), a DAG lipase inhibitor which will prevent arachidonic acid to stimulate voltage-independent calcium channels, were then used because they were already known to decrease contraction and calcium response to 5-HT in rat IPA (Guibert et al., 2004). LOE 908 and RHC-80267 both significantly decreased 5-HT-induced O2•_ increase (Figure 3C). Altogether, 5-HT stimulation of rat IPA produced O2•_ via the activation of the 5-HT₂ receptors and a calcium influx through voltage-independent calcium channels.

METC is known to be an important source of O2•_ (Perez-Vizcaino et al., 2010). Using EPR recordings in rat IPA, we showed that rotenone (100 nM), an inhibitor of the complex I of the METC significantly decreased the amount of O2•_ produced by 5-HT (100 μM) (Figure 4A). However, antimycin A (10 μM), a complex III cytochrome b_H inhibitor did not exhibit any significant effect on 5-HT-induced O2•_ production in rat IPA (Figure 4A). Rotenone did not modify basal O2•_ level (n = 6—data not shown). To confirm the role of mitochondria, we addressed the effect of 5-HT on the mitochondrial respiratory rate and on the mitochondrial membrane potential. 5-HT significantly decreased the mitochondrial respiratory rate (Figure 4B). The mitochondrial membrane was transiently depolarized in response to 5-HT 100 μM in the presence of extracellular calcium (Figures 5A,B) whereas, in its absence, the depolarization was strongly delayed and sustained (Figures 5B,C).

Since extracellular calcium influx and mitochondria both play a key role in the production of O2•_ by 5-HT (Figures 3A,C, 4A), we hypothesized that calcium influx induced by 5-HT would trigger an increase in mitochondrial calcium. We have thus simultaneously measured cytosolic and mitochondrial calcium levels in PASMC, in response to 5-HT in the presence and in the absence of extracellular calcium. Fluorescent labeling of PASMC with mitotracker green, a mitochondrial marker, and rhod-2, a mitochondrial calcium dye, shows a high level of colocalization demonstrating the specific targeting of rhod-2 to the mitochondrial network (Figures 6A–C). 5-HT (100 μM) increased both cytosolic and mitochondrial calcium levels in the presence of extracellular calcium (Figures 6D,F). In the absence of 5-HT stimulation, both cytosolic and mitochondrial calcium levels did not vary over time (data not shown). In the absence of extracellular calcium, the rise in cytosolic calcium in response to 5-HT was significantly decreased and the rise in mitochondrial calcium almost disappeared (Figures 6E,F). It should be noted that mitochondrial calcium increase is slightly delayed compared to cytosolic calcium increase (Figure 6D) suggesting that calcium influx may trigger the increase in mitochondrial calcium. In addition, we specifically measured mitochondrial O2•_ with the fluorescent probe mitosox. 5-HT (100 μM) significantly increased mitochondrial O2•_ levels (Figures 7Aa,Ab,B) and this O2•_ increase was significantly inhibited in the absence of extracellular calcium (Figures 7Ac,Ad,B) or the presence of rotenone (100 nM) (Figure 7B). Altogether, these results confirm the role of mitochondria, mitochondrial complex I and extracellular calcium in the signaling pathways involved in response to 5-HT in rat IPA.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.