Abstract
Neuronal respiration is controlled by ATP demand and Ca2+ but the roles played by each are unknown, as any Ca2+ signal also impacts on ATP demand. Ca2+ can control mitochondrial function through Ca2+-regulated mitochondrial carriers, the aspartate-glutamate and ATP-Mg/Pi carriers, ARALAR/AGC1 and SCaMC-3, respectively, or in the matrix after Ca2+ transport through the Ca2+ uniporter. We have studied the role of Ca2+ signaling in the regulation of mitochondrial respiration in intact mouse cortical neurons in basal conditions and in response to increased workload caused by increases in [Na+]cyt (veratridine, high-K+ depolarization) and/or [Ca2+]cyt (carbachol). Respiration in nonstimulated neurons on 2.5–5 mm glucose depends on ARALAR-malate aspartate shuttle (MAS), with a 46% drop in aralar KO neurons. All stimulation conditions induced increased OCR (oxygen consumption rate) in the presence of Ca2+, which was prevented by BAPTA-AM loading (to preserve the workload), or in Ca2+-free medium (which also lowers cell workload). SCaMC-3 limits respiration only in response to high workloads and robust Ca2+ signals. In every condition tested Ca2+ activation of ARALAR-MAS was required to fully stimulate coupled respiration by promoting pyruvate entry into mitochondria. In aralar KO neurons, respiration was stimulated by veratridine, but not by KCl or carbachol, indicating that the Ca2+ uniporter pathway played a role in the first, but not in the second condition, even though KCl caused an increase in [Ca2+]mit. The results suggest a requirement for ARALAR-MAS in priming pyruvate entry in mitochondria as a step needed to activate respiration by Ca2+ in response to moderate workloads.
Introduction
Oxygen consumption is controlled by the mitochondrial proton electrochemical gradient (ΔμH+; Mitchell and Moyle, 1969). In most cell types, ΔμH+ is mainly used in ATP synthesis. Increases in cell workload will consume ATP and lead to increased ATP production in mitochondria, through the utilization of ΔμH+. This in turn, will stimulate respiration. However, even at high workloads, a rapid formation of ATP through phosphocreatine and the creatine kinase reaction maintains cell ATP at almost constant levels (Cerdan et al., 1990).
In excitable cells, Ca2+ regulates cell function both by activation of ATP consumption (contraction, ion transport) and by activating ATP production through stimulation of oxidative phosphorylation. Ca2+ regulation of oxidative phosphorylation is thought to involve Ca2+ entry in mitochondria through the Ca2+ uniporter (MCU) and activation of three matrix dehydrogenases and complex V (McCormack et al., 1990; Glancy and Balaban, 2012). It may also occur through Ca2+ activation of mitochondrial metabolite transporters, by Ca2+ acting on the external side of the inner mitochondrial membrane. There are two types of such transporters, the aspartate-glutamate carriers (AGCs) and the ATP-Mg/Pi transporters (del Arco and Satrústegui, 1998; Palmieri et al., 2001; Satrústegui et al., 2007; Traba et al., 2012; Amigo et al., 2013).
ARALAR/AGC1 is present in brain and it is a component of the malate aspartate NADH shuttle (MAS). Activation by extramitochondrial Ca2+ of ARALAR-MAS results in an increase in NADH production in neuronal mitochondria (Pardo et al., 2006) and Gellerich et al. (2009, 2012, 2013) have proposed that Ca2+-activation of ARALAR functions as a “gas pedal” to increase pyruvate formation.
SCaMC-3 is the main mitochondrial ATP-Mg/Pi carrier present in brain and liver (del Arco and Satrústegui, 2004; Fiermonte et al., 2004; Amigo et al., 2013). Although not involved in oxidative phosphorylation, which requires adenine nucleotide exchange by the ADP/ATP carriers, adenine nucleotide accumulation in rat liver mitochondria through the ATP-Mg/Pi carrier results in a progressive increase in state 3 respiration (ADP-stimulated respiration; Asimakis and Aprille, 1980; Aprille et al., 1987; Amigo et al., 2013), perhaps by increasing the driving force of the ATP-synthase (Balaban et al., 2009; Glancy and Balaban, 2012).
The control of respiration by Ca2+ in intact neurons is still largely unknown. Hayakawa et al. (2005) described rapid Ca2+-dependent changes in oxygen consumption in response to high KCl in cultured Purkinje neurons, but Mathiesen et al. (2011) found no evidence for a role of cytosolic Ca2+ in activity-dependent rises in cerebral metabolic rate of oxygen (CMRO2) in cerebellar Purkinje neurons in the intact brain. Bak et al. (2012) described a Ca2+-induced increased in flux through the tricarboxylic acid cycle in cerebellar neurons using glucose as substrate. A confounding variable in these and other studies relates to the coincidence of the Ca2+-mechanism with the classical mechanism activating mitochondrial respiration, i.e., ATP demand. Indeed, any Ca2+ signal involves ATP consumption to recover prestimulation values, and the role of Ca2+ versus ADP-stimulation of respiration needs to be established.
The purpose of this work was to study the role of Ca2+ in the control of respiration under basal conditions and in response to an increase in workload in neuronal cultures. Particularly, to evaluate the contribution of the ARALAR-MAS and SCaMC-3 pathways of Ca2+ signaling to mitochondria with respect to that of MCU.
Materials and Methods
Animals.
Male SVJ129 × C57BL/6 mice carrying a deficiency for aralar expression (aralar−/−, aralar+/−, and aralar+/+) obtained from Lexicon Pharmaceuticals were used (Jalil et al., 2005). Mice deficient in SCaMC-3 were generated by Lexicon Pharmaceuticals with a mixed C57BL6/Sv129 genetic background (Amigo et al., 2013). Animals are born in Mendelian proportions and show no evident phenotypic traits. The mice were housed in a humidity- and temperature-controlled room on a 12 h light/dark cycle, receiving water and food ad libitum. All the experimental protocols performed in this study were performed in accordance with procedures approved in the Directive 86/609/EEC of the European Union and with approval of the local Ethics Committee of the Universidad Autónoma de Madrid. All efforts were made to minimize animal suffering.
Genotypes.
Genotypes were determined, as previously described for aralar (Pardo et al., 2006) and SCaMC-3 (Amigo et al., 2013), by PCR using genomic DNA obtained from tail or embryonic tissue samples (Nucleospin tissue kit, Macherey-Nagel). PCR mixtures were preincubated at 94°C for 5 min, followed by 35 cycles of DNA amplification at 94°C for 60 s, 58°C for 60 s, and 72°C for 60 s; the process was finished with incubation at 72°C for 5 min. DNA fragments were separated by electrophoresis on a 1.5% agarose gel.
Neuronal culture.
Cortical neuronal cultures were prepared from E15–E16 mouse embryos as described earlier (Ramos et al., 2003; Pardo et al., 2006). Embryos were obtained from crosses between C57BL6/SV129 aralar+/− mice, C57BL6/SV129 SCaMC-3+/+, or C57BL6/SV129 SCaMC-3−/− mice and nonbrain tissue was used for determination of DNA genotype. Neurons represented >80% of the total cell population (Ramos et al., 2003; Pardo et al., 2006) and included both glutamatergic and GABAergic neurons (results not shown).
Determination of glucose, lactate, and pyruvate.
Glucose, lactate, and pyruvate determinations in media from neuronal cell cultures were performed as follows. Cortical neurons were cultured in 12-well plates for 10 DIV and the experiment was performed in free-serum B27-supplemented neurobasal-A medium with 5 mm glucose. Aliquots of the medium were collected at various times (up to 2 h). Lactate content was determined enzymatically (Cuezva et al., 1982; Sánchez-Cenizo et al., 2010). Glucose and pyruvate concentrations in the media were determined by using kits from Boehringer (glucose) and Instruchemie BV (pyruvate) following the manufacturer's instructions in 48-well microplates, with a FLUOstar OPTIMA reader in the absorbance mode. The consumption of glucose, pyruvate, and net formation of lactate by the cells was calculated on the basis of cellular protein. Results are mean ± SEM (n = 7–15) of two to four independent experiments. Data were statistically evaluated by one-way ANOVA followed by Student's t test. Comparisons between control and aralar-deficient cultures were significant where indicated ***p ≤ 0.001.
Cytosolic Na+ and Ca2+ imaging in primary neuronal cultures.
Neurons growing on poly-lysine-coated coverslips were loaded with 5 μm Fura-2 AM and 50 μm pluronic acid F.127 (Invitrogen) for 30 min at 37°C in Ca2+-free HCSS (120 mm NaCl, 0.8 MgCl2, 25 mm HEPES, 5.4 mm KCl, pH 7.4), 2.5 mm glucose, and washed for 30 min in HCSS (2 mm CaCl2, 2.5 mm glucose). Then coverslips were mounted on the microscope stage equipped with a 40× objective as described previously (Ruiz et al., 1998) and Fura-2 fluorescence was imaged ratiometrically using alternate excitation at 340 and 380 nm, and a 510 nm emission filter with a Neofluar 40×/0.75 objective in an Axiovert 75M microscope (Zeiss). Additions were made as a bolus or, for isosmotic high K+, changing to a “isosmotic” HCSS in which 30 mm NaCl was replaced by 30 mm KCl (90 mm NaCl, 0.8 MgCl2, 25 mm HEPES, 35.4 mm KCl, pH 7.4) during continuous superfusion (≈1.5 ml/min). [Ca2+]i and [Na+]i imaging was performed as described with Fura-2 as Ca2+ indicator (Ruiz et al., 1998; Pardo et al., 2006) and sodium-binding benzofuran isophthalate (SBFI) as a sodium indicator (Rose and Ransom, 1997) at 37°C. For single-cell analysis of [Ca2+]i and [Na+]i the ratio of fluorescence intensity at 340 nm (F(340)) and 380 nm (F(380)) (F(340)/F(380)) was obtained. Signal calibration for Fura-2 [Ca2+]i imaging Ca2+ was performed with 1 μm Br-A23 Ca2+ ionophore (Sigma-Aldrich), with (Rmin) or without (Rmax) preincubation with 2 mm EGTA. Neuronal fluorescence at both wavelengths was corrected for autofluorescence after digitonin-permeabilization and quenching of Fura-2 fluorescence with 4 mm MnCl2. Ratio measurements were converted to Ca2+ concentrations as described previously (Grynkiewicz et al., 1985). Na+ imaging was performed as previously described (Rose and Ransom, 1997), briefly, 20 μm SBFI-AM was loaded during 90 min in the presence of 50 μm pluronic acid F.127 in HCSS 2 mm CaCl2 then washed and equilibrated for another 30 min. Monensin (10 μm) and ouabain (0.1 mm) were added for equilibration of extra- and intracellular [Na+] at the end of the experiments. Image acquisition was performed with the Aquacosmos 2.5 software (Hamamatsu) and data analysis was done with Origin software (OriginLab). When required, BAPTA loading was performed in Ca2+-free HCSS, 50 μm BAPTA-AM coloaded with Fura-2 AM during 30 min, then washed and equilibrated in HCSS 2 mm CaCl2 medium for another 30 min.
Mitochondrial Ca+2 and cytosolic ATP imaging.
To image mitochondrial Ca2+ and cytosolic ATP levels cells were plated onto 4-well Lab-Tek chamber slides and transfected using Effectene (Qiagen) 24 h prior the experiments either with the plasmid coding for mitochondrially targeted ratiometric GEM-GECO-1 (Addgene plasmid 32461), or with the plasmid coding for Cyt GO-ATeam 1 (kindly provided by Dr. H. Imamura, Kyoto University, Japan), respectively, and processed as previously described (Zhao et al., 2011; Nakano et al., 2011). Experiments were performed in 2.5 mm glucose HCSS with either 2 mm CaCl2 or 100 μm EGTA. Additions were made as a bolus or by changes in the medium composition during continuous superfusion. Cells were excited for 100 ms at 436/20 nm for Mit-GEM-GECO1 and at 485/27 nm for GO-ATeam 1, and the emitted fluorescence was collected through a dual pass dichroic CFP-YFP (440–500 and 510–600 nm) alternatively at 480/40 nm (CFP) and 535/30 nm (YFP) for Mit-GEM-GECO-1, and through a FF495-Di03 dichroic at 520/35 nm (GFP) and 567/15 nm (OFP) for GO-ATeam probe. Images were collected every 5 s using a filter wheel (Lambda 10-2, Sutter Instruments; all filters purchased from Chroma) and recorded by a Hamamatsu C9100-02 camera mounted on an Axiovert 200M inverted microscope equipped with a 40×/1.3 Plan-Neofluar objective. Mit-GEM-GECO1 emission ratio was CFP/YFP, whereas GO-ATeam emission ratio was OFP/GFP reflecting mitochondrial Ca2+ and cyt ATP levels, respectively. For Mit-GEM GECO1 imaging ROIs were selected on mitochondrial-containing areas (identified based on their morphology). Single-cell fluorescence recordings were analyzed using ImageJ (NIH) or MetaMorph (Universal Imaging). When required, 50 μm BAPTA-AM (Sigma-Aldrich) loading was performed in Ca2+-free HCSS during 30 min, and then washed and equilibrated in HCSS 2 mm CaCl2 for another 30 min.
Measurement of cellular oxygen consumption.
Cellular oxygen consumption rate (OCR) was measured using a Seahorse XF24 Extracellular Flux Analyzer (Seahorse Bioscience; Qian and Van Houten, 2010). Cortical primary neuronal were plated in XF24 V7 cell culture at 1.0 × 105 cells/well and incubated for 10 d in a 37°C, 5% CO2 incubator in serum-free B27-supplemented neurobasal medium with high levels of glucose. To study OCR at lower glucose concentrations, cultures were preconditioned for 24 h in serum-free B27-supplemented neurobasal-A media with 10 mm glucose, with medium changes every 12 h. Cells were equilibrated with bicarbonate-free low-buffered DMEM medium (without pyruvate, lactate, glucose, glutamine, and Ca2+) supplemented with 15, 5, or 2.5 mm glucose and 2 mm CaCl2 or 100 μm EGTA in conditions of ±Ca2+, for 1 h immediately before extracellular flux assay. In experiments with isosmotic high K+, cultures were first preconditioned for 20 min in 5 mm glucose HCSS in the presence or absence of 2 mm CaCl2. Then, neurons were either maintained in the same medium or stimulated with 30 mm KCl in 5 mm glucose in Ca2+-containing or Ca2+-free isosmotic HCSS medium in which 30 mm NaCl was replaced by 30 mm KCl for 25 min before starting respirometry experiments. Calibration of the respiration took place after the vehicle injection in port A. BAPTA-AM was used at 50 μm in experiments with veratridine and carbachol in DMEM medium or 10 μm BAPTA-AM in the case of isosmotic KCl stimulation in HCSS medium. Loading was performed in Ca2+-free in bicarbonate-free media or HCSS media for 20–30 min, and then washed and equilibrated in 2 mm CaCl2 medium for another 25–30 min. Substrates were prepared in the same medium in which the experiment was conducted and were injected from the reagent ports automatically to the wells at the times indicated. Mitochondrial function in neurons was determined through sequential addition of 6 μm oligomycin, 0.5 mm 2,4-dinitrophenol, and 1 μm antimycin/1 μm rotenone. This allowed determination of basal oxygen consumption, oxygen consumption linked to ATP synthesis (ATP), non-ATP linked oxygen consumption (leak), mitochondrial uncoupled respiration (MUR), and nonmitochondrial oxygen consumption (NM; Qian and Van Houten, 2010; for review, see Brand and Nicholls, 2011).
Results
Glucose utilization in WT and aralar KO and SCaMC-3 KO cultured neurons
Glucose consumption and net lactate formation were measured in culture media from neuronal cultures at 9 DIV using 5 mm glucose over 30–120 min incubation period (Table 1). ARALAR deficiency decreased glucose consumption (by 1.35-fold, p = 0.4, Student's t test, n = 7–15) but did not change lactate formation, unlike previous results obtained using higher glucose concentrations (Pardo et al., 2011). As a result, the percentage of glucose converted into lactate (%Lac/Gluc) tended to increase in aralar KO neurons. In addition, pyruvate consumption was significantly increased, by 1.39-fold (p = 0.00047, Student's t test), in aralar KO neurons (Table 1). These results show that neurons depend on ARALAR for malate/aspartate shuttle activity and glucose oxidation as previously described (Pardo et al., 2011). Glucose utilization and lactate production were the same in SCaMC-3-WT and SCaMC-3-KO neurons (1.10 ± 0.11 and 1.20 ± 0.13 micromoles glucose × mg prot−1 × h−1, and 1.6 ± 0.2 and 2.2 ± 0.4 μmoles lactate × milligrams prot−1 × h−1 in SCaMC-3-WT and SCaMC-3-KO, respectively).
Glucose utilization in WT and aralar KO cultured neurons
Bioenergetic characterization of control and aralar KO and SCaMC-3-KO cultured neurons
We next analyzed the respiratory activity in intact primary neuronal cultures by using a Seahorse XF24 extracellular flux analyzer (Qian and Van Houten, 2010). The basic setup of these experiments and the information obtained is shown in a graphic form in Figure 1A. In control experiments 2,4-dinitrophenol was titrated to obtain the maximum activity of the electron transport chain (data not shown). In other set of controls oligomycin was omitted to correct for its possible effects on the estimation of the maximal uncoupled coupled respiration (Brand and Nicholls, 2011). Under the assay conditions used, no effects of oligomycin on maximal uncoupled respiration were observed (data not shown).
Bioenergetic characterization of aralar WT, aralar KO, SCaMC-3-WT, and SCaMC-3-KO cultured neurons. A, Representative pharmacological profile of oxygen consumption rate in aralar WT neurons showing the sequential injection of agonist/vehicle (Veh) and metabolic inhibitors: oligomycin (Olig, 6 μm), 2,4-dinitrophenol (DNP, 0.5 mm) and antimycin A/rotenone (Ant/Rot, 1.0 μm/1.0 μm) at different time point indicated by dashed lines which allows the determination of basal oxygen consumption (BR), oxygen consumption linked to ATP synthesis (ATP), non-ATP linked oxygen consumption (Leak), maximal uncoupled respiration (MUR) and nonmitochondrial oxygen consumption (NM) as represented graphically. B, C, Effect of glucose and Ca2+ on maximal uncoupled respiration (MUR) in aralar WT vs aralar KO neurons and SCaMC-3-WT versus SCaMC-3-KO neurons. D, Oxygen consumption rate in SCaMC-3-WT versus KO neurons in 2.5 mm glucose, 2 mm Ca2+. E, F, Effect of glucose and Ca2+ on basal respiration capacity in aralar WT versus aralar KO and SCaMC-3-WT versus SCaMC-3-KO neurons. Results are means ± SEM of 3–11 experiments. The effect of Ca2+ on MUR was significant for aralar WT (p < 0.05, two-way ANOVA) but not aralar KO neurons, and the lack of ARALAR caused a significant decrease in MUR at all glucose concentrations in the absence (p < 0.05, two-way ANOVA) or presence of Ca2+ (p < 0.001, two-way ANOVA; ##p < 0.01, post hoc Bonferroni test). Glucose, but not Ca2+, had a significant effect on basal respiration both in aralar WT (p < 0.01, two-way ANOVA) and aralar KO neurons (p < 0.001, two-way ANOVA) and the lack of ARALAR caused a significant decrease in basal respiration both in the presence (p < 0.001, two-way ANOVA) or absence (p < 0.01, two-way ANOVA) of Ca2+. G, H, Oxygen consumption profile in aralar WT and aralar KO cultures in 2.5 mm glucose medium supplemented with 2 mm pyruvate (Pyr) in the presence and absence of 2 mm Ca2+. I, MUR in the presence or absence of 2 mm pyruvate, expressed as percentage of basal OCR in aralar WT and aralar KO neurons (n = 3–11 experiments, Student's t test; *p ≤ 0.05, ***p ≤ 0.001). Data are expressed as mean ± SEM. Statistical analysis was performed using STATISTICA, version7, StatSoft.
The influence of Ca2+ on respiration was addressed by conducting the experiments in the presence and absence of 2 mm CaCl2 in the incubation medium (Fig. 1B,E). The relative roles of ARALAR-MAS and SCaMC-3 in the stimulation of OCR were studied by using primary neuronal cultures derived from aralar (Fig. 1A–E) or SCaMC-3-KO mice (Fig. 1C,D,F). To control for the variations among the parental mouse strains, the wild-type condition in experiments with ARALAR- or SCaMC-3-deficient neurons was that of the specific parental strain (aralar WT or SCaMC-3-WT, respectively). Figure 1G,H illustrates the profiles obtained from WT and aralar KO cultures with 2.5 mm glucose in the presence or absence of Ca2+ and in the presence or absence of 2 mm pyruvate. The behavior of SCaMC-3-KO neuronal cultures was similar to that of WT neurons (Fig. 1D).
MUR
MUR in cultured neurons reflects the maximal respiratory capacity of neuronal mitochondria in intact neurons. It is not affected by the control exerted by proton reentry either through ATP synthase or the proton leak. It is mainly controlled by substrate supply, by the intrinsic respiratory capacity of mitochondria, and by nonmitochondrial respiration (Brand and Nicholls, 2011). Nonmitochondrial respiration was subtracted to calculate MUR and other mitochondrial respiratory parameters.
MUR was found to increase with the glucose concentration (Fig. 1B,C) and was higher in the presence than absence of Ca2+. The lack of ARALAR caused a drastic decrease in MUR, both in Ca2+-free and Ca2+-containing media, at all glucose concentrations, dropping to 20% of WT values at 2.5 mm glucose (Fig. 1B; in the presence of Ca2+ MUR was 667.94 ± 114.36 in WT vs 154.63 ± 31.39 in aralar KO neurons, Student's t test, p = 0.009), which is consistent with the limitation in glucose-derived pyruvate supply to mitochondria when the major NADH shuttle system fails. Indeed, exogenous pyruvate (but not lactate, results not shown) supply, which bypasses the limitation imposed by the lack of MAS, did not change MUR in the presence of Ca2+ (a nonsignificant 1.23-fold increase; Fig. 1G,I) but increased it significantly, by ∼2.17-fold both in the presence and 1.93-fold in the absence of Ca2+ (p = 0.0009 and p = 0.00013 respectively, Student's t test) in aralar KO neurons (Fig. 1G–I). We did not find any effect of 10 mm lactate on MUR in WT or aralar KO neurons (results not shown). This may be due to the strong cytosolic acidification induced by the uptake of lactic acid (Álvarez et al., 2003) which may inhibit glycolysis. Therefore, the effect of lactate has not been studied any further.
As indicated above (Fig. 1B), MUR tended to be higher in the presence than in the absence of external Ca2+ at every glucose concentration. This could reflect an effect of cytosolic Ca2+ on substrate supply, perhaps by activating the malate aspartate shuttle. Indeed, the effect of Ca2+ on MUR was no longer present in ARALAR-deficient neurons especially at the physiological glucose concentrations of 2.5 and 5 mm. On the other hand pyruvate increases MUR in aralar KO neurons regardless of the presence of Ca2+ while in control neurons this effect is less prominent and only in Ca2+-free media (Fig. 1G–I). Indeed, as the presence of the uncoupler prevents Ca2+ uptake in mitochondria and Ca2+-activation of pyruvate dehydrogenase, it is expected to result in a Ca2+-independent respiration on pyruvate. These results are consistent with recent reports showing that removal of the components of the mitochondrial Ca2+ uniporter complex (MCU, MICU1, MCUR1) does not change MUR (Baughman et al., 2011; Mallilankaraman et al., 2012a,b) whereas preventing IP3-mediated Ca2+ signals does reduce MUR (Cárdenas et al., 2010).
Basal respiration
Basal respiration rates increased with glucose concentration in WT, aralar KO, and SCaMC-3-KO neurons, and this was independent of the presence of Ca2+ (Fig. 1E). This is unexpected because the Km for hexokinase is very low, 0.05 mm (Grossbard and Schimke, 1966), but could be related to mild osmotic changes with higher glucose level. The presence or absence of 2 mm pyruvate did not change basal respiration in 2.5 mm glucose in WT, aralar KO, or SCaMC-3-KO neurons (Fig. 1G,H, and results not shown).
Remarkably with 2.5 mm glucose, a concentration close to that normally present in cerebral extracellular fluid in vivo (Lewis et al., 1974), absolute OCR in aralar KO neuronal cultures, was reduced by 46.13 ± 7.46% (p = 0.0067, Student's t test) with respect to WT or SCaMC-3-KO neuronal cultures (Fig. 1E,F). The decrease in OCR in aralar KO neurons persisted at all glucose concentrations indicating that the lack of MAS prevents adequate glucose-derived substrate supply to mitochondria also under basal conditions. However, the fact that basal respiration still proceeds even if halved, suggests the existence of other shuttle systems in these neurons. The participation of mitochondrial ATP synthesis in basal respiration is estimated from the decrease in respiration after the addition of oligomycin. Respiration in the presence of oligomycin reflects the proton leak but as ATP synthase inhibition results in a slight mitochondrial hyperpolarization and the proton leak is voltage-dependent, this approach underestimates ATP synthesis and exaggerates the real proton leak (Brand and Nicholls, 2011). With these caveats, the percentage of OCR involved in ATP synthesis in WT and aralar KO neurons in 2.5 mm glucose was the same, 73.92 ± 1.20%, and the corresponding proton leak is 26.08 ± 1.20%, with an “apparent” mitochondrial respiratory control ratio (ATP synthesis/proton leak) of 3.63 ± 0.56. The presence of pyruvate or the absence of SCaMC-3 (results not shown) did not change the apparent respiratory control ratios (RCR) under basal conditions, but it should be noted that both SCaMC-3-WT and SCaMC-3-KO neurons had higher apparent RCR values with respect to aralar WT or aralar KO neurons, of 5.1 ± 0.48 (SCaMC-3-WT) and 6 ± 0.3 (SCaMC-3-KO) in 2 mm Ca2+ medium, and of 6.08 ± 0.28 (SCaMC-3-WT) and 5.85 ± 1.01 (SCaMC-3-KO) in Ca2+-free medium, these differences should be consequence of the different parental mouse strains.
Effect of agonists on regulation of OCR
Having shown that in neurons using 2.5–5 mm glucose, basal respiration is not limited by substrate supply, we next studied the control of respiration by agents able to increase neuronal workload. Any increase in workload is followed by ATP breakdown and ADP production which is expected to increase OCR. We used veratridine and high K+ which increase workload after plasma membrane depolarization through increases in cytosolic Na+ and Ca2+ and stimulation of the Na+-K+ ATPase, Na+/Ca2+ exchange and the sarcoendoplasmic reticulum Ca2+-ATPase (SERCA) and plasma membrane Ca2+-ATPase (PMCA) pumps. It is expected that in the presence, but not absence, of external Ca2+, Ca2+-regulation of respiration would contribute to the increase in OCR in response to these workloads. It has been proposed that even Ca2+ alone, not ATP demand, stimulates mitochondrial bioenergetics in some neuronal types (Chouhan et al., 2012).
Another way to increase workload is through agents causing Ca2+-mobilization from intracellular stores. The increase in cytosolic Ca2+ will cause ATP breakdown by SERCA and PMCA pumps and would be expected to increase OCR. ATP utilization to recover the resting state is much smaller than that caused by Na+ entry (Attwell and Laughlin, 2001) and consequently, the OCR response should be also smaller. Moreover, removal of external Ca2+ in this condition will reduce Ca2+ signals and workload at the same time.
In either workload condition, Ca2+ may activate OCR by activating ARALAR-MAS and/or SCaMC-3, by way of increasing substrate or adenine nucleotide supply to mitochondria, and Ca2+ may enter mitochondria through the recently identified MCU (Baughman et al., 2011; De Stefani et al., 2011), increasing NADH production in mitochondria and oxidative phosphorylation (Glancy and Balaban, 2012). These mechanisms differ in their S0.5 values for Ca2+-activation, ∼300 nm for ARALAR/AGC1 (Pardo et al., 2006) and in the micromolar range for both SCaMC-3 (Amigo et al., 2013) and MCU (Drago et al., 2011). The relative roles of these pathways in the stimulation of OCR were studied by using primary neuronal cultures derived from aralar or SCaMC-3-KO mice. To control for the variations among the parental mouse strains, the wild-type condition in experiments with ARALAR- or SCaMC-3-deficient neurons was that of the specific parental strain (aralar WT or SCaMC-3-WT, respectively).
Veratridine-induced increase in respiration is activated by Ca2+ and results in a drop in ATP levels
The lipophilic alkaloid neurotoxin, veratridine, binds buried sites in the voltage-dependent Na+ channels in the matrix of the lipid bilayer (Strichartz et al., 1987), causing Na+ channels to remain open, and a pronounced increase in intracellular Na+ concentration (Rose and Ransom, 1997). The plasma membrane depolarization produced by veratridine opens voltage-dependent Ca2+ channels allowing Ca2+ inflow (Rego et al., 2001). Figure 2 shows the effects of veratridine in cortical neurons. Veratridine (50 μm) produced a large and sustained increase in [Ca2+]i which was prevented in Ca2+-free media (Fig. 2A,B). It also induced a sustained rise in mitochondrial Ca2+ levels, reported by mit-GEM-GECO-1, a genetically coded Ca2+ sensor expressed in the mitochondrial matrix (Zhao et al., 2011) which was completely absent in Ca2+-free media (Fig. 2D,E). Cytosolic ATP levels, measured with the low affinity FRET cytosolic ATP sensor GO-ATeam1 (Nakano et al., 2011), were found to drop after veratridine exposure, particularly in Ca2+-free media (Fig. 2G–I). Veratridine induced a sharp increase in OCR of mouse cortical neurons, as previously observed in synaptosomes (Brand and Nicholls, 2011) which resulted in 180.21 ± 4.76% (aralar WT) to 255.02 ± 11.01% (SCaMC-3-WT) increase over basal levels (Figs. 2J, 3E–G and A–C, respectively). The increase was due to coupled respiration and proton leak, as apparent respiratory control values (Figs. 2L, 3D,H) did not change significantly after veratridine addition in Ca2+ media.
Changes in cytosolic and mitochondrial Ca2+, cytosolic ATP, oxygen consumption and cytosolic Na+ in primary neuronal cultures in response to veratridine. A–C, Changes in [Ca2+]cyt, in Fura-2 loaded neurons obtained by stimulation with 50 μm Veratridine (Ver or Verat) in 2 mm Ca2+, Ca2+-free medium, or 50 μm BAPTA preloaded neurons in 2 mm Ca2+ medium. D–F, Corresponding data in neurons transfected with Mit-GEM-GECO1 probe to determine changes in [Ca2+]mit. Recordings from at least 60 cells per condition and two independent experiments were used for [Ca2+]i and a minimum of 15 cells and eight independent experiments for [Ca2+]mit imaging. Individual cell recordings (gray) and average (thick black trace) were shown. G–I, Cytosolic ATP in neurons transfected with cyt-GO-ATeam1 stimulated with veratridine in 2 mm Ca2+ medium, Ca2+-free medium plus 100 μm EGTA and comparison of the two conditions. Recordings from individual cells (gray) and average (black) are shown. The drop of ATP values with respect to basal levels 200 s after veratridine addition were 23.9 ± 1.70% in the presence and 30 ± 1.69% in the absence of Ca2+.(*p = 0.017 two-tailed unpaired Student's t test). J, Veratridine-induced stimulation of OCR in aralar WT neurons under the mentioned Ca2+ and BAPTA conditions. K, L, Stimulation of respiration (as percentage of basal values) and RCR. RCR in nonstimulated state are represented with horizontal lines for each experimental condition (n = 9–11 experiments one-way ANOVA, *p ≤ 0.05, ***p ≤ 0.001, post hoc Bonferroni test). M, N, Changes in [Na+]i, in SBFI-loaded neurons by stimulation with 50 μm veratridine in 2 mm Ca2+ medium or Ca2+-free medium (∼90 neurons per condition). O, Comparison between response in Ca2+ medium (black trace) and Ca2+-free (gray trace) is shown. Increases in normalized SBFI ratio 200 s after veratridine were 21.1 ± 1.02% and 20.3 ± 0.81in the presence or absence of Ca2+, respectively.
OCR responses to veratridine in ARALAR- and SCaMC-3-deficient neurons. A–D, Respiratory profiles in the presence or absence of 2 mm Ca2+, stimulation of mitochondrial respiration and RCRs of SCaMC-3-WT and SCaMC-3-KO neurons stimulated with 50 μm Veratridne (Ver) in 2.5 mm glucose. E–H, Corresponding data from aralar WT and aralar KO neurons. I–L, Corresponding data for the effect of 2 mm pyruvate (Pyr) on veratridine-induced mitochondrial respiration in aralar WT and aralar KO neurons. RCR in nonstimulated state are represented with horizontal lines for each experimental condition. Data are expressed as mean ± SEM from n = 5–6 experiments in SCaMC-3-WT and SCaMC-3-KO cultures, and from n = 4–12 experiments in aralar WT and aralar KO cultures. Two-way ANOVA, *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001, post hoc Bonferroni test.
In the absence of Ca2+, the increase in OCR induced by veratridine was greatly reduced (Figs. 2J, 3B, F). Proton leak values after veratridine exposure were higher in the presence than in the absence of Ca2+ in the external medium, as reflected in the higher RCR in Ca2+-free than Ca2+-containing medium (Figs. 2L, 3D, H). This is likely due to the existence of Ca2+ cycling across the inner mitochondrial membrane through MCU, H+/Ca2+, Na+/Ca2+, and H+/Na+ exchangers (Nicholls, 2005).
The veratridine-induced increase in OCR in Ca2+-free media is driven by the increase in cytosolic Na+ coupled to ATP demand. In the presence of Ca2+, the increase in OCR induced by veratridine is even higher, whereas ATP levels dropped significantly less (Fig. 2I). This may be caused by (1) the increases in cytosolic Na+ and Ca2+ which could result in a further increase in ATP demand in the presence of Ca2+, or (2) the absence of Ca2+-dependent regulation of respiration in Ca2+-free media. To tell apart these possibilities, neurons were preincubated with the rapid intracellular Ca2+-chelator BAPTA-AM (Abramov and Duchen, 2008) and challenged with veratridine in the presence of Ca2+. BAPTA loading blunts cytosolic Ca2+ signals but does not change Ca2+ inflow through voltage-dependent Ca2+ channels (Adler et al., 1991), conditions under which the Ca2+-dependent fraction of the global workload induced by veratridine is maintained. In BAPTA-AM (50 μm) loaded neurons Ca2+ signals in the cytosol, and specially in mitochondria, only appeared gradually and after a substantial delay (Fig. 2A,C) and veratridine-stimulation of OCR was severely decreased (Fig. 2J), indicating that Ca2+-signaling is clearly required for veratridine-induced stimulation of respiration in Ca2+-medium.
We also tested whether the Na+-dependent fraction of workload induced by veratridine was maintained in the absence or presence of Ca2+. Intracellular Na+ was measured in neurons loaded with SBFI-AM, the Na+ indicator (Rose and Ransom, 1997; Fig. 2M–O). As initial SBFI ratios are slightly higher in Ca2+-free than in Ca2+-containing medium, they were normalized to the initial SBFI ratio. Figure 2O clearly shows that the absence of Ca2+ did not modify veratridine-induced Na+ influx. As veratridine-induced drop in cytosolic ATP was still more pronounced in Ca2+-free media, together the results suggest that the lack of Ca2+-regulation of OCR prevents an adequate regeneration of ATP to meet.
Veratridine- and Ca2+-dependent stimulation of respiration is blunted in the absence of SCaMC-3 and ARALAR
The stimulatory effect of veratridine was different in WT and SCaMC-3-KO neurons (Fig. 3A–D). In the presence of Ca2+, the increase in OCR was significantly smaller in SCaMC-3-KO neurons than in controls, but these differences disappeared in Ca2+-free media (Fig. 3C). The requirement for Ca2+ in the effect of SCaMC-3 points to a potential effect of Ca2+-activation of the transporter in the acquisition of a full increase in respiration. This is related with an increase in coupled respiration, as proton leak values are unchanged in SCaMC-3-KO neurons, resulting in significantly higher RCR values in WT than in SCaMC-3-KO neurons (Fig. 3D).
The stimulatory effect of veratridine was strikingly diminished in aralar KO neurons both in the presence and in the absence of Ca2+ (Fig. 3E–H). It should be noted that as the basal rates in the aralar KO cells are about half those of the WT (with or without calcium), the decrease in %OCR stimulation caused by veratridine shown in Figure 3E,F, actually represents a very large difference in absolute OCR. aralar KO neurons respond to veratridine in a Ca2+-dependent way (Fig. 3G), thus Ca2+ entry in mitochondria and activation of dehydrogenases together with activation of SCaMC-3 provide the only pathways of Ca2+ regulation of mitochondrial respiration in these neurons.
To analyze whether the lower response to veratridine was due to limited substrate supply to mitochondria, the effect of pyruvate was analyzed in this experimental condition. We found that pyruvate (2 mm) did not increase veratridine-stimulated respiration in WT, but caused an impressive threefold increase in stimulated respiration in aralar KO neurons (Fig. 3, compare G and K) reaching stimulation values identical to those of WT neurons both in the absence or presence of Ca2+ (Fig. 3I–L). These results clearly show that the failure to respond to veratridine in aralar KO neurons is due to substrate limitation imposed by the lack of MAS.
High K+ induced increase in mitochondrial respiration is highly Ca2+-dependent
High extracellular KCl concentration (30 mm) depolarizes cell membrane, activating the voltage-sensitive Na+ and Ca2+ channels leading to elevation of their cytosolic concentrations in cultured neurons (Courtney et al., 1990; Rose and Ransom, 1997). Channel opening is transient due to the voltage-dependent inactivation of Na+ channels but some of the Ca2+ channels remain open under continuous depolarization (Courtney et al., 1990). The increase in cytosolic Na+ and Ca2+ levels will activate Na+ and Ca2+ extrusion mechanisms (Na+/K+ ATPase pump, SERCA, PMCA, and Na+/Ca2+ exchangers), which will activate ATP turnover.
Figure 4A–P shows the responses to the addition of 30 mm KCl in cortical neurons. KCl addition produces a pronounced elevation of [Ca2+]i in the presence, but not absence, of external Ca2+, with average [Ca2+]i returning to ∼500 nm in the continuous presence of KCl (Fig. 4A,B). Mitochondrial Ca2+ levels increase in parallel to those of cytosolic Ca2+ levels, and this increase is absent in Ca2+-free media (Fig. 4 D,E). The results were essentially the same when KCl addition was performed by isosmotic replacing Na+ for K+ during continuous superfusion (results not shown).
Changes of cytosolic and mitochondrial Ca2+, cytosolic ATP, oxygen consumption, and cytosolic Na+ in primary neuronal in response to KCl. A–C, Changes in [Ca2+]cyt, in Fura-2-loaded neurons obtained by stimulation with 30 mm KCl in 2 mm Ca2+, Ca2+-free medium, or 50 μm BAPTA preloaded neurons in 2 mm Ca2+ medium. D–F, Corresponding data in neurons transfected with Mit-GEM-GECO1 probe to determine changes in [Ca2+]mit. Recordings from at least 60 cells per condition and two independent experiments were used for cytosolic Ca2+ imaging and a minimum of 15 cells and eight independent experiments for mitochondrial Ca2+ imaging. Individual cell recordings (gray) and average (black) were shown. G, H, Cytosolic ATP levels after a switch from HCSS medium to isosmotic KCl medium in which 30 mm NaCl was replaced by 30 mm KCl either in 2 mm Ca2+ medium or 100 μm EGTA medium. I, Comparison between the two conditions. The drop in ATP values 200 s after isosmotic KCl stimulation was 18.6 ± 0.3% and 9 ± 0.1% with respect to basal levels in the presence and absence of Ca2+, respectively (**p = 0.009 two-tailed unpaired Student's t test). J–L, Respiratory profiles, stimulation of mitochondrial respiration, and RCRs of neurons stimulated with hyperosmotic KCl in 2 mm Ca2+, Ca2+-free medium or 50 μm BAPTA preloaded neurons in 2 mm Ca2+. M–O, Corresponding data from neurons stimulated with isosmotic KCl using 10 μm BAPTA. RCR in nonstimulated state are represented with horizontal lines for each experimental condition (n = 8–30 experiments in aralar WT neuronal cultures; one-way ANOVA, *p ≤ 0.05, ***p ≤ 0.001, post hoc Bonferroni test). P, Changes in [Na+]i, in SBFI loaded neurons exposed to 5.4 or 30 mm isosmotic KCl in 2 mm Ca2+ or 100 μm EGTA medium as indicated. Monensin (Monen; 10 μm) and Ouabain (Oua; 0.1 mm) were added for equilibration of extra- and intracellular [Na+] at the end of the experiments. Individual cell recordings (gray) and average (black) were shown (n = 29).
The changes in cytosolic ATP levels, after isosmotic replacement of 30 mm Na+ for 30 mm K+ are shown in Figure 4G–I. A gradual drop in cytosolic ATP levels is observed in the presence of Ca2+ but greater than that in the absence of Ca2+. When 30 mm KCl was added as a bolus, this drop was preceded by a rapid transient increase which was the same in the absence or presence of Ca2+, most likely caused by a hyperosmotic effect, as it is also observed after 30 mm NaCl or 60 mm sucrose addition (results not shown).
The experimental setup of the Seahorse equipment does not allow to add KCl in an isotonic medium. Therefore, we first evaluated the effect of 30 mm KCl addition on OCR in cortical neurons. High potassium produces an immediate increase in OCR of ∼183% (aralar WT neurons; Fig. 4J,K) or 200% (SCaMC-3-WT neurons, data not shown) over basal values in the presence of 2 mm Ca2+ but a much smaller increase of ∼117% (aralar WT; Fig. 4J,K) and 150% (SCaMC-3-WT, data not shown) respectively, in the absence of Ca2+. This stimulation was stable >6–9 min and was not increased in the presence of pyruvate (results not shown). In PC12 cells, KCl effects on OCR are due to an hyperosmotic response (Ashton and Ushkaryov, 2005). However, in cortical neurons the effects of hyperosmotic NaCl or sucrose on OCR were much smaller than those of KCl (results not shown), suggesting that most of the OCR response to KCl addition is not due to the hyperosmotic effect.
We next evaluated the response to 30 mm KCl under isosmotic conditions. Neurons were incubated with isosmotic 30 mm KCl or control (5.4 mm KCl) conditions before OCR determinations were started. As observed, OCR was also increased in isosmotic 30 mm KCl, but slightly less than in hyperosmotic conditions. KCl stimulation was also calcium-dependent, 137.55 ± 8.09% and 113.97 ± 7.96% in the presence and absence of Ca2+, respectively (Fig. 4N). Isosmotic KCl stimulation significantly increased oligomycin-sensitive respiration in the presence of 2 mm Ca2+ (77.24 ± 1.64 and 80.18 ± 0.65 in Ca2+-containing basal and KCl stimulated conditions, respectively, p = 0.025, Student's t test).
To investigate the role of Ca2+-regulation of OCR in the response to isosmotic KCl, neurons were preincubated with 10 μm BAPTA-AM, and challenged with KCl in the presence of Ca2+ to maintain the global workload of K+-depolarization in a Ca2+-medium, while blunting cytosolic Ca2+ signals. Indeed, Figure 4C,F shows that the increases in [Ca2+]i and specially [Ca2+]mit were drastically reduced and delayed under this condition. Interestingly, KCl-stimulation of OCR dropped to the values obtained in Ca2+-free medium (Fig. 4M,N). By using neurons loaded with SBFI-AM we have found that isosmotic KCl causes only subtle changes in cytosolic Na+ both in the presence or absence of Ca2+ (Fig. 4P), in agreement with previous reports (Rose and Ransom, 1997). This suggests that KCl-induced workload is probably related to the increase in cytosolic Ca2+, whereas increased cytosolic Na+ plays a modest role. This is consistent with the findings that (1) KCl-induced fall in cytosolic ATP is hardly detectable in Ca2+-free medium (Fig. 4 G–I) and(2) KCl-stimulation of OCR was much lower in the absence than in the presence of Ca2+ (Fig. 4N). These results clearly indicate that Ca2+-regulation of OCR is absolutely required to reach a full stimulation of coupled respiration by KCl-depolarization.
High K+-stimulation of OCR requires ARALAR-MAS but not SCaMC-3
Figure 5A–J shows the effects of the lack of SCaMC-3 or ARALAR on K+-stimulated OCR in cortical neurons under isosmotic conditions. The lack of SCaMC-3 did not modify the response to high K+ (Fig. 5A–C). However, the lack of ARALAR blocked K+-stimulation of OCR, both in the presence or absence of Ca2+, suggesting that the Ca2+-mediated stimulation strictly requires ARALAR-MAS. Indeed, although in WT neurons K+-stimulated OCR was not increased any further in the presence of pyruvate (Fig. 5, compare F and I), the addition of 2 mm pyruvate, did not change basal OCR but led to a pronounced potentiation of the effects of K+ on aralar KO neuron. In fact, stimulation values were now the same as in WT neurons. These results further indicate that the Ca2+-dependent stimulation of OCR caused by KCl relies on ARALAR as signaling pathway to mitochondria.
OCR-responses to 30 mm KCl in ARALAR- or SCaMC-3-deficient cortical neurons. A–C, Respiratory profiles, stimulation of mitochondrial respiration, and RCRs of SCaMC-3-WT and SCaMC-3-KO neurons stimulated with isosmotic 30 mm KCl in the presence of 2 mm Ca2+. D–G, Corresponding data for aralar WT and aralar KO cultures in the presence or absence of 2 mm Ca2+. H–J, Corresponding data for the effect of 2 mm pyruvate (Pyr) on isosmotic 30 mm KCl-induced respiratory stimulation in aralar WT and aralar KO neurons in the presence of 2 mm Ca2+. RCR in nonstimulated state are represented with horizontal lines for each experimental condition. Data correspond to 5–6 experiments in SCaMC-3-WT and SCaMC-3-KO cultures and 4–24 experiments in aralar WT and aralar KO respectively (two-way ANOVA, *p ≤ 0.05; **p ≤ 0.01; post hoc Bonferroni test).
Carbachol induces a Ca2+-dependent increase in respiration levels which requires the ARALAR-MAS pathway
In cortical neurons, carbachol acts on muscarinic cholinergic, G-protein-coupled receptors, which activate Gq/11 and signal through activation of phospholipase C, generation of inositol 3-phosphate (IP3) and Ca2+-mobilization from intracellular stores, giving rise to a transient [Ca2+]i peak (Kelly et al., 1996; Lucas-Meunier et al., 2003; Kipanyula et al., 2012). The emptying of the endoplasmic reticulum (ER) Ca2+ stores triggered by IP3 activates Ca2+ entry through low-conductance plasmalemmal channels that are regulated by the level of Ca2+ stored in the ER, a process known as store-operated Ca2+ entry (SOCE; Putney, 2009).
Figure 6A,B shows that 250 μm carbachol induced an immediate increase in [Ca2+]i also observed in Ca2+-free medium, followed by a new steady-state [Ca2+]i value above resting levels in the presence of Ca2+ or to a recovery of the resting state in Ca2+-free media in which SOCE-induced Ca2+ entry does not occur. Carbachol-induced [Ca2+]cyt signals hardly reached mitochondria, as no rise in [Ca2+]mit was detected either in the presence or absence of Ca2+ (Fig. 6D,E). The workload induced by carbachol is much smaller than that involving plasma membrane depolarization and does not lead to any changes in cytosolic ATP levels either in the presence or absence of Ca2+ (Fig. 6F). As ATP is used to restore Ca2+ levels to resting conditions, ATP demand is expected to be drastically reduced in a Ca2+-free medium. Accordingly, carbachol induced an increase in OCR in the presence of Ca2+ but not in its absence (Fig. 6G–I) but much lower than that caused by high K+ or veratridine. Carbachol specifically increased oligomycin sensitive respiration from 74.48 ± 1.97% to 84.38 ± 2.21 in basal and carcachol-induced OCR, respectively.
Changes in cytosolic and mitochondrial Ca2+, cytosolic ATP and oxygen consumption in response to carbachol. A–C, Changes in [Ca2+]cyt, in Fura-2-loaded neurons obtained by stimulation with 250 μm Carbachol (Cch) in 2 mm Ca2+, Ca2+-free medium, or 50 μm BAPTA preloaded neurons in 2 mm Ca2+ medium. D, E, Corresponding data in neurons transfected with Mit-GEM-GECO1 to determine changes in [Ca2+]mit. Recordings from at least 60 cells per condition and two independent experiments were used for cytosolic Ca2+ and a minimum of 15 cells and eight independent experiments for mitochondrial Ca2+. Individual cell recordings (gray) and average (black) were shown. F, Corresponding data for cytosolic ATP in neurons transfected with cyt-GO-ATeam1. Drop in ATP values 200 s after carbachol addition were 4.9 ± 0.8% and 2.1 ± 0.4% with respect to basal levels in the presence or absence of Ca2+ (p = 0.92, two-tailed unpaired Student's t test). G–I, Respiratory profiles, stimulation of mitochondrial respiration, and RCRs of neurons stimulated with 250 μm carbachol in 2 mm Ca2+, Ca2+-free medium, or 50 μm BAPTA preloaded neurons in 2 mm Ca2+. RCR in nonstimulated state were represented with horizontal lines for each experimental condition. Data from 18 to 22 experiments in aralar WT cultures (one-way ANOVA, ***p ≤ 0.001, post hoc Bonferroni test).
The role of Ca2+-regulation of OCR with respect to ATP demand in the response to carbachol was evaluated by studying carbachol-stimulation of OCR in the presence of Ca2+ in BAPTA-AM loaded neurons, in which carbachol-induced Ca2+ transients were abolished (Fig. 6C), whereas store-operated Ca2+ entry is maintained or increased by preventing its Ca2+-dependent inactivation (Litjens et al., 2004), a situation entailing conservation or even increase of workload in the absence of Ca2+ signaling. After BAPTA loading, carbachol-stimulation of OCR was halved (Fig. 6G–I), indicating that Ca2+ signaling to mitochondria is required to induce the full response to carbachol. Moreover, as the workload and ATP demand may be larger in BAPTA-loaded than control neurons, the real effect of Ca2+ on carbachol stimulation of OCR is probably underestimated.
Figure 7A–C shows that the lack of SCaMC-3 had no effect on carbachol-induced stimulation of OCR. However, ARALAR deficiency resulted in a smaller stimulation in the presence of Ca2+ (Fig. 7D–F) suggesting that under this condition activation of the ARALAR-MAS pathway is required. The lack of ARALAR reflects a limitation in substrate supply to mitochondria, as pyruvate addition abolished the differences in carbachol-stimulation of OCR between ARALAR-deficient and WT neurons (Fig. 7G–I).
OCR response to Carbachol (Cch) in ARALAR- and SCaMC-3-deficient neuronal cultures. A–D, Respiratory profiles in the presence or absence of 2 mm Ca2+, stimulation of mitochondrial respiration, and RCRs of SCaMC-3-WT and SCaMC-3-KO neurons stimulated with 250 μm carbachol in 2.5 mm glucose. E–H, Corresponding data from aralar WT and aralar KO neurons. G–I, Corresponding data of the effect of 2 mm pyruvate (Pyr) on carbachol-induced mitochondrial respiration in aralar WT and aralar KO neurons. RCR in nonstimulated state are represented with horizontal lines for each experimental condition. Data from 5 to 6 experiments in SCaMC-3-WT and SCaMC-3-KO cultures, and from 4 to 12 experiments in aralar WT and aralar KO cultures (two-way ANOVA, *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001, post hoc Bonferroni test).
The requirement of ARALAR-MAS for the full respiratory response to carbachol in the presence of Ca2+ is possibly mediated through Ca2+-activation of ARALAR in the intermembrane space rather than through changes in matrix Ca2+, as no changes in mitochondrial Ca2+ were observed in response to carbachol, whereas high K+ and veratridine caused rapid increases in mitochondrial Ca2+ in the presence, but not absence, of extracellular Ca2+. The effect of BAPTA on carbachol stimulation of OCR is probably related to its effects in blunting the increase in cytosolic Ca2+ required to activate ARALAR-MAS at the mitochondrial intermembrane space.
Discussion
The role of Ca2+ in tuning ATP production to ATP demand in excitable cells has been known for a long time (Jouaville et al., 1999; Hayakawa et al., 2005; Glancy and Balaban, 2012; Rizzuto et al., 2012). However, in intact neurons the contribution of Ca2+ in adjusting coupled respiration to ATP demand is still controversial (Hayakawa et al., 2005; Mathiesen et al., 2011). A confounding variable in these studies is the fact that any increase in cytosolic Ca2+, either by uptake from the extracellular space or release from intracellular Ca2+ stores, is necessarily coupled to the use of ATP to restore resting Ca2+ levels.
We have now addressed the following fundamental issues: (1) Are any of the brain Ca2+-regulated mitochondrial carriers ARALAR or SCaMC-3 required in basal or maximal respiration of intact neurons using physiological glucose concentrations? (2) Is Ca2+ signaling (independently of Ca2+-dependent increase in ATP demand) required to upregulate respiration in response to an increase in workload in intact neurons? (3) Are ARALAR or SCaMC-3 limiting such a Ca2+-dependent upregulation of respiration?
We have first shown that mitochondrial respiration in nonstimulated intact cortical neurons using physiological glucose concentrations 2.5–5 mm (Lewis et al., 1974) depends on ARALAR-MAS, but not on SCaMC-3, and is reduced by 46% in neurons lacking ARALAR. This clearly indicates a most important role of ARALAR-MAS in shuttling NADH to mitochondria and providing pyruvate supply to the organelle. The lack of effect of external Ca2+ and exogenous pyruvate in basal respiration confirms that the main factor governing basal OCR is workload, i.e., ATP utilization (Brand and Nicholls, 2011).
We next analyzed the effects of different workloads on OCR under conditions in which Ca2+ signaling was allowed or prevented. These were provided by veratridine, high K+-depolarization, and carbachol. In the presence of Ca2+, veratridine causes an increase in cytosolic Ca2+ and Na+ concentrations of 2–3 μm (Fig. 2) and 25 mm (Rose and Ransom, 1997), respectively. As estimated by Attwell and Laughlin (2001), any increase in [Ca2+]cyt from 100 nm to 2–3 μm, in which 39 of every 40 Ca2+ atoms are buffered (Helmchen et al., 1997), would result in a net increase in ATP consumption of ∼0.1 nm ATP in 1 μl cell volume to extrude Ca2+ at a cost of 1 ATP/Ca2+ by plasma membrane or ER ATPases, and a demand of ∼8.3 nm ATP to extrude Na+ at a cost of 1 ATP/3 Na+ through the Na+, K+-ATPase, i.e., a total demand of ∼8.4 nm ATP, most of which are used for Na+ extrusion. In other words, even if the [Ca2+]bound/[Ca2+]free ratio may be higher (Martinez-Serrano et al., 1992), most of the ATP demand is due to Na+ entry. Thirty millimoles of K+ causes a nondetectable increase in [Na+]i and an increase in [Ca2+]i, of 1–2 μm (Fig. 4), resulting in total ATP demand of ∼0.06–0.08 nm per microliter of cell volume. For carbachol, with mean increases in [Ca2+]i of ∼0.1–0.2 μm, ATP demand is only due to Ca2+ extrusion and corresponds to ∼0.01 nm ATP in 1 μl cell volume. In the presence of Ca2+, the response to all three conditions was an immediate increase in oligomycin-sensitive OCR which persisted (KCl and carbachol) or declined (veratridine) during the following minutes.
In all cases the OCR response was severely reduced in the absence of Ca2+, a condition that substantially lowers ATP demand in response to KCl and carbachol, but has a much smaller impact on workload in the case of veratridine, suggesting that lack of Ca2+-regulation rather than lower ATP demand is responsible for the reduced veratridine-stimulation of OCR. Indeed, the decrease in cytosolic ATP levels caused by veratridine is more severe in Ca2+-free media (Fig. 2) indicating that decreased mitochondrial production of ATP, not lower ATP demand, explains the neuronal response to this workload in Ca2+-free medium. All stimuli caused a substantially lower increase in OCR in BAPTA-loaded neurons, in which workload is preserved, clearly indicating that Ca2+-regulation of respiration is required to meet workload demands with an increase in OCR.
We next analyzed the mechanisms responsible for Ca2+-regulation of respiration in the different workloads. For veratridine, the mechanism involved is clearly dependent on MCU as respiration in aralar KO neurons still responds to veratridine in a Ca2+-dependent way (Fig. 3G). In these neurons, Ca2+ entry in mitochondria and activation of dehydrogenases together with activation of SCaMC-3 provide the only pathways of Ca2+-regulation of mitochondrial respiration. As the blunted response to veratridine in aralar KO neurons is reversed by pyruvate supply both in the absence and presence of Ca2+ (Fig. 3), the results suggest that the increase in mitochondrial Ca2+ is insufficient to fully increase respiration upon veratridine challenge and MAS activity to push pyruvate in mitochondria (Gellerich et al., 2009) is clearly required.
On the other hand, the requirement of SCaMC-3 is different, as it is strictly Ca2+-dependent. Isolated brain mitochondria from SCaMC-3 mice exchange [ATP-Mg]2− or [ADPH]2− against Pi2− across the inner mitochondrial membrane with an S0.5 for Ca2+ activation of 3–4 μm within the same range of the MCU (Amigo et al., 2013). The uptake of adenine nucleotides stimulated by a large rise in cytosolic Ca2+ may affect mitochondrial function in two ways: by mass action ratio effects on the ATP synthase, causing an increase in coupled respiration as is thought to occur in hepatocytes (Aprille, 1988), or by increasing the Ca2+ retention capacity of mitochondria exposed to high Ca2+ loads (Amigo et al., 2013). Further work is required to clarify the mechanism involved.
Ca2+-regulation of OCR in response to KCl does not require SCaMC-3 but is absolutely dependent on ARALAR-MAS, as no increase in OCR was obtained in aralar KO neurons, whereas exogenous pyruvate allowed restoration of stimulated respiration to control values. As [Ca2+]mit increases in this condition (Fig. 4D) regardless of ARALAR deficiency (Pardo et al., 2006), these results indicate that [Ca2+]mit by itself is unable to increase respiration in response to this workload. In these conditions, [Ca2+]cyt activation of ARALAR-MAS appears as the only pathway for Ca2+-regulation of respiration in cortical neurons by “pushing” pyruvate into mitochondria, in other words, through an increase in pyruvate transport into mitochondria caused by an increase in cytosolic pyruvate. It is surprising that the increase in [Ca2+]mit is not sufficient to activate pyruvate dehydrogenase (PDH) and other mitochondrial dehydrogenases and thereby “pull” pyruvate into mitochondria as occurs in aralar KO neurons stimulated by veratridine (pyruvate is expected to be pulled into mitochondria through the consumption of pyruvate by PDH and mass action ratio effects on the pyruvate carrier). Activation of PDH may require not only Ca2+-activation of PDH phosphatase but inhibition of PDH kinases (PDKs) which inactivate PDH, through increases in mitochondrial pyruvate/acetyl-CoA and ADP/ATP ratios (Stacpoole, 2012). The high workload induced by veratridine, but not by KCl, associated with a more pronounced increase in mitochondrial ADP/ATP ratio may limit inhibition of PDKs and explain the lack of KCl-stimulated respiration in aralar KO neurons. On the other hand, whether ARALAR-MAS activity is also required to prime mitochondria for Ca2+ stimulation of metabolism is an open question.
Neurons respond to the small workload imposed by carbachol with a modest but sustained increase in OCR. Interestingly, the lack of ARALAR caused a significant decrease in this response in the presence, but not absence, of Ca2+, suggesting that Ca2+-activation ARALAR-MAS is the push mechanism to drive pyruvate in mitochondria upon carbachol exposure. The prominent role of ARALAR-MAS in regulating neuronal respiration in this condition agrees with the small size of the Ca2+ transient and lack of [Ca2+]mit changes caused by carbachol (Fig. 6) and higher affinity of the ARALAR-MAS pathway than MCU for Ca2+ (300 nm vs a few μm), and with the finding of an essential role of AGC1-MAS in increasing mitochondrial NAD(P)H in response to small Ca2+ signals (Pardo et al., 2006).
Together, these results underscore the roles of MCU-[Ca2+]mit, ARALAR-MAS, and SCaMC-3 in upregulating oligomycin-sensitive respiration in cerebral cortex neurons in response to workloads produced by increases in Na+ and/or Ca2+ and robust-to-small Ca2+ signals (Fig. 8). These roles may vary in neurons from the adult brain due to changes in enzyme and transporter composition.
Schematic representation of the effects of the different workloads on Ca2+ regulation of neuronal oxygen consumption. Pathways indicated in green, purple, and red are activated by Veratridine (Ver), KCl, and Carbachol (Cch), respectively. All three stimuli activate ARALAR-MAS, increasing pyruvate supply in mitochondria, whereas only veratridine activates adenine nucleotide uptake through SCaMC-3. Veratridine also stimulates respiration after mitochondrial Ca2+ uptake through the MCU and activation of mitochondrial dehydrogenases. SOCE, Store operated Ca2+ entry; ETC, electron transport chain; AAT,aspartate aminotranferase; AcCoA, acetyl coenzyme A; Asp, aspartate; Cch, carbachol; ER, endoplasmic reticulum; ETC, electron transport chain; Glu, glutamate; IP3, inositol trisphosphate; IP3R, inositol 3-phosphate receptor; α-KG, α-ketoglutarate; mAChR, muscarinic cholinergic G-protein-coupled receptor; MCU, mitochondrial calcium uniporter; MDH, malate dehydrogenase; NCX, sodium calcium exchanger; OAA, oxaloacetate; OGC, oxoglutarate carrier; PDH, pyruvate dehydrogenase; PDKase, pyruvate dehydrogenase kinase; PDPase, pyruvate dehydrogenase phosphatase; PIP2, phosphatidylinositol 4,5-bisphosphate; PLC-β, phospholipase C-β; PMCA, plasma membrane calcium ATPase; PyrC, pyruvate carrier; SERCA, sarco/endoplasmic reticulum Ca2+-ATPase; SOCE, store-operated calcium entry; TCA, tricarboxylic acid cycle; Ver, veratridine; VOCC, voltage operated calcium channel; VONC, voltage operated sodium channel.
MCU-[Ca2+]mit play a prominent role in the response to the large workload produced by veratridine. The ARALAR-MAS pathway also contributes in the regulation of pyruvate supply to mitochondria in this condition, greatly amplifying the OCR response. The ARALAR-MAS pathway plays an outstanding role in the response to smaller workloads, being the only Ca2+-regulation mechanism responsible for upregulation of respiration in response to the small Ca2+ signals produced by carbachol. Surprisingly, it appears to be also the main mechanism responsible for Ca2+-regulation of respiration in response to KCl, which induces large Ca2+ signals in mitochondria. In all cases, ARALAR-MAS through the Ca2+-activation of ARALAR in the intermembrane space provides pyruvate to mitochondria. The mitochondrial ATP-Mg/Pi carrier SCaMC-3 limits respiration only in response to high workloads and robust Ca2+ signals which are able to activate carrier activity, as produced by veratridine, suggesting a requirement of adenine nucleotide uptake in mitochondria to generate a full respiratory response under these conditions.
Footnotes
This work was supported by the Ministerio de Economía Grant BFU2011-30456, by CIBERER (an initiative of the ISCIII), by the Comunidad de Madrid Grant S2010/BMD-2402 MITOLAB-CM (to J.S.), by ISCIII Grant PI080610 (to A.d.A.), and by an institutional grant from the Fundación Ramon Areces to the Centro de Biología Molecular Severo Ochoa. C.B.R is a recipient of a FPU fellowship from the Ministerio de Educación y Ciencia. We thank Dr. Hiromi Imamura, Kyoto University for kindly providing the GO-ATeam plasmids. We thank Isabel Manso, Alejandro Arandilla, and Bárbara Sesé for technical support, José Belio for his help in preparing the figures, and María Angeles Muñoz, from the Unit of Optical and Confocal Microscopy, for her inestimable support.
The authors declare no competing financial interests.
- Correspondence should be addressed to Dr. Jorgina Satrústegui, Universidad Autonoma de Madrid, Centro de Biologia Molecular Severo, Nicolas Cabrera 1, Madrid, 28049, Spain. jsatrustegui{at}cbm.uam.es
