| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institute for Medical Neurobiology, University of Magdeburg, Germany;
Department of Physiology, University of Alberta, Canada; and
* Neuroscience Research Group, University of Calgary, Canada
1Correspondence: Institute for Medical Neurobiology, University of Magdeburg, Leipziger Str. 44, 39120 Magdeburg, Germany. E-mail: thomas.horn{at}medizin.uni-magdeburg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-nitro-L-arginine. Calcium transients were prolonged by conjoint application of NMDA and NO. We conclude that NMDA-evoked [Ca2+]i transients are modulated by endogenous NO production, which leads to release of calcium from the mitochondrial pool. An NO-activated mitochondrial permeability transition pore may lead to cell death after overstimulation of NMDA receptors.—Horn, T. F. W., Wolf, G., Duffy, S., Weiss, S., Keilhoff, G., MacVicar, B. A. Nitric oxide promotes intracellular calcium release from mitochondria in striatal neurons.
Key Words: nitric oxide synthase • NOS isoform • NMDA • NO production
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
and during hypoxia. There is accumulating evidence that mitochondria comprise a target for NO and that NO cytotoxicity may be mediated by inhibition of the respiratory chain (3
4
5)
and depleting ATP (6
, 7)
. Inhibition of respiration could contribute to the proapoptotic effects of NO by membrane potential reduction, ensuing activation of a mitochondrial permeability transition pore (PTP) (8
, 9)
, and release of the caspase activator cytochrome c (10)
. Aside from NO production from cytosolic NOS, mitochondria themselves express a NOS isoform (10
, 11)
and possess nitrite reductase activity (12)
, additional sources of intramitochondrial NO. Excitotoxicity can evoke an NO-dependent mitochondrial depolarization and [Ca2+]i deregulation (13
, 14)
. A mechanism that results in increased [Ca2+]i after NO exposure would allow NO to influence other [Ca2+]i-sensitive (but not directly NO-sensitive) targets and exert feedback regulation on nNOS activity. Second, positive feedback between [Ca2+]i elevation and NO production may result in deregulation of [Ca2+]i and cell death.
We imaged fluorescent calcium indicator dyes in cultured striatal neurons to test whether exogenous NO could cause mitochondrial calcium release and if NMDA-induced NO production augments NMDA-induced [Ca2+]i signals.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Striatal primordia from E14 mouse embryos were mechanically dissociated in growth media and plated onto 12 or 22 mm glass coverslips precoated with poly-ornithine (10 mg/mL). Neurons were grown in serum-free culture medium within a humidified atmosphere aerated with 5% CO2 in air at 37°C and maintained for 7–17 days in vitro before experiments. Sample cell cultures underwent routine checks for neuron content by immunohistochemistry for the neuronal marker protein MAP-2.
Fura-2 imaging
Coverslips were incubated at 37°C for 40 min in a solution of minimum essential medium with 16.6 µM Fura-2 (Molecular Probes, Eugene, OR) predissolved in dimethyl sulfoxide (DMSO; Sigma, St. Louis, MO) plus 10% pluronic acid detergent (Molecular Probes). After dye loading, cultures were rinsed with medium and transferred to the stage of an inverted epifluorescence microscope (Zeiss Axiovert 10 with a plan-neofluar 40x water/oil immersion objective). Cells were superfused with control saline containing (in mM): 120 NaCl, 5 KCl, 10 HEPES, 1.3 MgCl2, 10 glucose, 2 CaCl2, at pH 7.35. For studies involving NMDA receptor stimulation, MgCl2 was removed. For calcium-free saline, CaCl2 was omitted and 2 mM EGTA (Sigma) added. High extracellular [K+]o (50 mM) saline was made by equimolar substitution of KCl for NaCl.
For imaging of Fura-2 fluorescence, excitation light was provided by a 75 W Xenon lamp (Zeiss) gated by an electronic shutter (Uniblitz T132). A motor-driven filter changer situated between the light source and microscope controlled Fura-2 excitation wavelengths (340 and 380 nm, Omega Optical). Fura-2 fluorescence emission, filtered at 510 nm long-pass, was recorded using a CCD camera (Cohu 6500) coupled to an image intensifier (Video Scope KS 1380). At each wavelength, 32 frames were averaged by an 8 bit imaging board (DT 2867 or DT 3155, Data Translation). Shutter duration, filter changes and image acquisition were controlled by Axon Imaging Workbench software (Axon Instruments, Union City, CA). Acquisition frequency was 1 averaged image per 30 or 60 s. Analysis was performed online or after image acquisition by averaging ratio values within boxes overlying images of cell somata.
Confocal imaging of fluo-3, JC-1, and TMRM
Stock solutions of 5 mM fluo-3 AM in DMSO + 20% pluronic acid were prepared and stored in 1 µL aliquots at -20°C. Neurons grown on 22 mm glass coverslips were incubated in fluo-3 diluted to a final concentration of 5 µM in medium for 40 min. After dye loading, cultures were rinsed with HEPES buffer and transferred into a stainless steel chamber (Attofluor, 2 mL volume) that was mounted on a thermostatically controlled (37°C) stage of a laser scanning microscope (Zeiss LSM 410) mounted on an inverted microscope (Axiovert 135). Cells were observed using a Zeiss 64x oil immersion lens. For imaging of fluo-3 fluorescence, excitation light was provided by an argon laser at 488 nm. Fluo-3 fluorescence emission, filtered at 505 nm long-pass, was recorded using the photomultiplier of the LSM410. Image acquisition frequency was dependent on the experimental protocol and ranged from 1 image per 5 to 30 s. Cells were superfused with HEPES buffer containing (in mM): 120 NaCl, 5 KCl, 10 HEPES, 10 glucose, 2 CaCl2, at pH 7.35, gassed with oxygen at 37°C, at a rate of 2 mL/min using a peristaltic pump (Gilson) and allowed to equilibrate for 
15 min. After start of imaging, the cultures were superfused for a further 10 min with HEPES buffer to wash out excess dye. Cultures were then subjected to the treatment protocols as described. Analysis of fluorescence intensity was performed after image acquisition by averaging intensity values within boxes overlying images of cell somata using the imaging software of the Zeiss LSM. Data were normalized and averages of image intensities calculated.
For imaging mitochondrial polarization status, cells were incubated with either 2 µg/mL JC-1 (5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolylcarbocyanine iodide) or 3 µM TMRM (perchlorate tetramethylrhodamine methyl ester, both from Molecular Probes) in medium at 37°C for 20–30 min. The fluorescence of the JC-1 monomer and J aggregate, representing cytosolic and mitochondrial JC-1, respectively, was excited at 488 nm (Argon laser) using a dichroic filter at 505 nm and the fluorescence was divided to two channels: band-pass 515–565 nm and long-pass 570. For imaging TMRM fluorescence excitation was set at 543 and emission filtered at 570 nm long-pass. The filtered fluorescence light was recorded by photomultipliers. Within one set of experiments, gain and offset of the imaging program were kept constant.
Images were analyzed for changes in fluorescence intensities within regions of interests (boxes drawn over mitochondria or cytosol areas) using the image analysis program of the Zeiss confocal microscope and values were plotted as changes from baseline.
Electrophysiology
Whole-cell voltage clamp recordings were obtained from striatal neurons in cell culture to examine the potential modulation of NMDA-evoked currents by agents that altered NMDA-evoked Ca2+ signals. Patch pipettes were filled with (in mM): 120 KCl, 5 MgCl2, 40 HEPES, 10 EGTA, 2 Mg ATP (pH=7.35); recordings were obtained and digitized using P clamp software (all from Axon Instruments). External bath solution contained (in mM): 140 NaCl, 5 KCl, 10 HEPES, 10 glucose, 2 CaCl2 (pH=7.35). For drug application, drugs were added to the external bath solution and perfused onto cells via an 8 channel microperfusion system with solenoid-driven switching. Solution completely bathed all parts of the cells in the microscopic field and switching times were <30 ms.
Drug application
To study the effect of NOS activity on the NMDA-evoked [Ca2+]i increases, striatal neurons were incubated for 40 min with NOS substrates and/or NOS competitive antagonists and subsequently given three 2-min applications of 200 µM NMDA at 7-min intervals. This paradigm was repeated under four conditions: [1] pretreatment with 100 µM L-arginine (L-Arg), [2] 1 mM N-nitro-L-arginine (N-Arg), [3] L-Arg (100 µM) + N-Arg (1 mM), [4] N-Arg (1 mM) + D-arginine (D-Arg, 100 µM). All these agents were purchased from Sigma.
For application of authentic NO, control HEPES buffer was deoxygenated by 20 min ventilation with Argon gas in a sealed flask. NO stock solution was prepared by saturating the content of one of these flasks with pure NO gas (Union Carbide, Chicago, IL). Stock solution concentration was 1–3 mM as measured by a chemiluminescence NO detector (see below). Dilutions (1:100) were made by transferring aliquots in gas-tight glass syringes from NO stocks to other sealed flasks or gas-tight Hamilton glass syringes containing deoxygenated HEPES buffer. Cell cultures were exposed to single or repeated pulses of NO-containing medium by switching superfusion medium from control saline to saline containing diluted NO. Control saline was superfused using a peristaltic pump (Gilson, Middleton, WI) and NO-containing saline was superfused using either a peristaltic pump or a syringe pump (KD Scientific, New Hope, PA).
Nitrite/nitric oxide analysis
Since NO levels in fluids can vary due to changes in ambient pressure and temperature during preparation of stock solution, changes in NO levels in the cell culture holding chamber were measured continuously during the experiments by selective amperometric oxidation using an NO-sensitive electrode (ISO-NO-METER, World Precision Instruments, Sarasota, FL). For estimating true NO concentrations, these NO-sensitive electrode measurements were calibrated by generating stoichiometric standards from the reaction:
 |
The concentration of NO in the calibration solutions as well as the direct measurements of the NO electrode were confirmed in samples drawn from the holding chamber of the electrodes by a chemiluminescent assay using a SIEVERS NO-ANALYSER in conjunction with the computerized data analysis program NOAWIN as described before (16)
. Nitrite in the biological samples was reduced to NO by potassium iodide in the presence of acetic acid as described above. The generated NO was carried together with the free NO of the sample from the reaction vessel to the analysis chamber by a steady flow of N2. Chemiluminescence that resulted from the reaction of ozone with NO was measured via a photomultiplier. The instrument was calibrated by injection of different NaNO2 concentrations with a fixed sample volume.
NO levels were adjusted to levels ranging from 1 to 8 µM by adjusting the amount of NO stock solution injected into the HEPES buffer container to keep experimental conditions constant.
For controls, other cultures were treated with only deoxygenated saline or deoxygenated saline with sodium nitrite (NaNO2, Sigma). The cGMP analog 8-bromo-cGMP (Molecular Probes) was superfused in control saline with ambient O2 or in saline deoxygenated by continuous gassing with pure N2. Other subsets of cultures were treated with different substances [thapsigargine, ruthenium red, ryanodine, cyclosporine A (CSA), carbonyl cyanide m-chlorophenylhydrazone (CCCP)] that are known to interfere with distinct elements of intracellular calcium regulation or mitochondrial function. Experimental treatment protocols are outlined in the graph and figure legends. All chemicals were purchased from Sigma except where noted. Stock solutions were prepared in either water (ruthenium red, ryanodine), DMSO (thapsigargine), or ethanol (CCCP, CSA). Vehicle-alone treatments were used as controls.
Statistical analysis
If not stated otherwise, all experiments were performed at least in duplicate. Each individual statistical group contained data from at least 20 individually measured cells.
For the continuous fluorescence measurement experiments, data were presented for each test condition as a change in normalized fluorescence (mean±SE). Each condition was tested on at least two coverslips from no fewer than two different culture dates typically yielding a total cell of 20–30 cells per condition. Statistical significance was determined by ANOVA single-factor analysis and values with P < 0.05 were considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A and Fig. 2
). Analysis of the spatial distribution of these signals revealed [Ca2+]i increases in the soma and dendrites (Fig. 1A
, C). The kinetics of these responses was highly variable, even between neighboring neurons (Fig. 2)
, indicating that these differences were not due to bath NO gradients. Responses could be slowly developing (first and second trace) or rapidly activating after variable delays (third and fifth trace) whereas 24.4% of the cells examined showed no response or very small increases (fourth trace).
|
|
Averaged responses from many cells were characterized by a more gradually developing increase over 3–10 min, which reached a stable plateau in the continued presence of NO and reversed to baseline after washout (Fig. 3
A). In this and subsequent figures, cells that showed no detectable NO-evoked [Ca2+]i increase (25.6% of cells examined; e.g., Fig. 2
, fourth panel) were not included in the averages. Although an increase in the extracellular concentration of authentic NO was correlated with an increase in neuronal [Ca2+]i, it is possible these increases were actually mediated by products of NO oxidation. Even under low oxygen tension, the reaction of NO with molecular oxygen to form nitrogen dioxide (2NO+O2
2NO2) could proceed rapidly in the highly concentrated NO stocks and superfusates. Oxygenated NO-treated saline showed no detectable levels of NO and superfusion of this oxidized NO solution did not evoke a Ca2+ increase in neurons that subsequently responded to NO in deoxygenated saline (Fig. 3B
, n=4, total of 16 cells). In an aqueous environment, the primary oxidation product of NO is nitrite, NO2- (17)
, so we tested whether nitrite (in the form of NaNO2) could elicit [Ca2+]i responses. Similar to oxygenated NO, NaNO2 (1–50 mM) did not elicit a [Ca2+]i response in neurons that did respond to subsequent NO application (Fig. 3C
, n=3, total of 31 cells). NO exerts many physiological effects through activation of guanylate cyclase and production of cGMP (18)
. Application of 8-bromo-cGMP (0.5 or 1 mM), a cell-permeating cGMP analog, for 20 min in deoxygenated saline did not increase [Ca2+]i, indicating that cGMP alone is not the mediator of the NO-induced [Ca2+]i response (Fig. 3D
, n=2 cultures, total of 25 cells). Since NO reacts with O2, one could argue that the observed effects are due to the lack of O2 in the superfusion medium. Therefore, we tested the effect of oxygen-free buffer on [Ca2+]i. The omission of oxygen by ventilating the perfusion buffer with argon before applying it to the cells did not alter the baseline of [Ca2+]i (79 cells from 2 coverslips, data not shown), indicating that the rise in [Ca2+]i due to NO is not caused by the lack of oxygen in the perfusate.
|
We also tested whether the [Ca2+]i responses observed during NO exposure are reversible and repeatable. Two consecutive NO exposures separated by 15 min revealed that the NO-mediated [Ca2+]i increase was indeed repeatable (Fig. 4
A, average of 21 cells from 3 coverslips). The peak magnitude of the second response was 150% larger than the first, suggesting that either the initial NO application sensitized some Ca2+-mobilizing mechanism(s) to NO or suppressed neuronal [Ca2+]i buffering. Inhibition of Ca2+ buffering was unlikely, however, because calcium responses evoked by 50 mM [K+]o depolarization after repeated NO exposure (456±46% increase in F340/380; Fig. 4
) were not significantly different from control [K+]o responses (425±29% increase in F340/380, not shown).
|
We next examined the source of this [Ca2+]i increase by analyzing the effects of extracellular [Ca2+]o removal on NO responses. Removal of [Ca2+]o during the plateau phase of the NO response did not reduce the magnitude of this component, indicating that Ca2+ influx is not required for maintenance of the plateau (Fig. 4B
, average of 24 cells from 3 coverslips). Subsequent removal and readdition of NO in the continued absence of [Ca2+]o elicited a [Ca2+]i response, indicating that response initiation does not depend on Ca2+ influx (average 49 cells from 6 coverslips). Although a contribution from Ca2+ influx cannot be discounted, these results indicate that NO promoted release of Ca2+ from internal stores and that this release alone was sufficient to evoke significant increases in [Ca2+]i. We used several pharmacological antagonists to attempt to define the internal Ca2+ store that NO acts on (Fig. 5
A, 22 cells from 2 coverslips). Thapsigargin, which irreversibly blocks endoplasmic reticulum Ca-ATPase and depletes IP3-sensitive Ca2+ stores, evoked a large transient increase in [Ca2+]i but did not block the increase elicited by subsequent NO application (Fig. 5A
). In contrast, numerous cells that first appeared unresponsive to NO exhibited such Ca2+ increases after thapsigargin application, suggesting that internal sequestration by the Ca-ATPase normally helps to limit the NO-induced Ca2+ increase (data not shown). Repeated superfusion of cell cultures with NO in the presence of thapsigargin did not alter the response pattern of [Ca2+]i during exposure to NO (Fig. 5A
). Similarly, ryanodine, which blocks Ca2+-induced Ca2+ release, and ruthenium red, which blocks Ca2+ release mediated by Ca2+/Mg2+ATPase, did not block NO-induced Ca2+ increases (Fig. 5B
).
|
Effect of CCCP, CSA, and glucose deficiency on NO-evoked [Ca2+]i signals
We examined the possibility that NO caused the release of Ca2+ from mitochondria. We first tested the effect of the mitochondrial uncoupler CCCP on the NO-induced [Ca2+]i signal. Cell cultures were treated with 2 µM CCCP for 1 min before administration of an NO pulse (Fig. 6
, total of n=78 examined cells). The increase in [Ca2+]i in response to an NO pulse was reduced when the mitochondrial uncoupler CCCP (1 µM in HEPES buffer throughout the experiment) was present in the perfusion medium vs. NO application alone. Two fluorescent probes (JC-1 and TMRM) were used to study the response of the neuronal mitochondrial membrane potential to the NO exposure. Figure 7
shows representative images of neurons before (control), during NO exposure (NO 5 µM), and after NO exposure (washout, 2 coverslips each condition, n=20 cells each). The green fluorescence of the JC-1-loaded cells and the bright cytosolic fluorescence of the TMRM-loaded cells indicate a breakdown of the mitochondrial membrane potential. We observed that the membrane potential decreased within seconds of NO exposure in contrast to the [Ca2+]i response, which occurred within minutes of the NO stimulus. The mitochondria appear to recover their membrane potential (Fig. 7
, washout), pointing to a reversible effect of NO.
|
|
Another set of cell cultures (Fig. 8
) was subjected to two consecutive exposures to NO-containing perfusion buffer (control). One subset of these cultures was also treated with CSA (2 µM) after the first of two NO pulses (CSA-treated). The control cultures displayed an increase in [Ca2+]i during each NO pulse. In cultures treated with CSA, the increase in [Ca2+]i during the second NO exposure was reduced by 53% (Fig. 8)
.
|
We investigated the effect of glucose deprivation on the response pattern of the CSA-sensitive [Ca2+]i transients during NO exposure. When cell cultures were superfused with glucose-free medium 10 min before the NO application, the amplitude of the NO-induced increase was potentiated (Fig. 9
). This potentiated effect was depressed to almost control levels when 2 µM CSA was added to the medium. Glucose deprivation alone did not have an effect on [Ca2+]i. CSA was used only at this concentration, since it is known that this substance has other effects at higher concentrations.
|
Effect of NO synthase activity on NMDA-evoked [Ca2+]i signals
We next studied NMDA-induced [Ca2+]i signals to determine whether NO-dependent mitochondrial Ca2+ release can modulate these signals. Three 2 min applications of 200 µM NMDA at 7 min intervals in the presence of the NOS substrate L-arginine (100 µM) resulted in three distinct [Ca2+]i signal types: reversible (r) responses (>50% [Ca2+]i recovery between applications and complete recovery within 11 min of the third NMDA application); sustained (s) responses (<50% recovery between applications and incomplete recovery 11 min after the third); and uncontrolled (u) [Ca2+]i responses (maintained irreversible [Ca2+]i overshoot after the second or third NMDA application) (for representative traces, see Fig. 10
A). These response types were observed in 43 ± 12%, 40 ± 7%, and 17% ± 8% of cells, respectively (totals from 4 coverslips, mean±SE; Fig. 10C
). Repetition of this stimulus paradigm in the presence of the competitive NOS inhibitor N-nitro-L-arginine (N-Arg, 10 µM) greatly reduced the propensity for the later two response types (sustained: 15±4%; uncontrolled: 0.3±0.25%, Fig. 10C
; for representative traces, see Fig. 10B
). The action of N-Arg was not the result of depressing NMDA-activated currents because the coapplication of N-Arg had no significant effect on NMDA-activated currents (98.5±1.7% of control, n=4, data not shown). The suppression of sustained and uncontrolled [Ca2+]i responses was reversed by readdition of L-arginine (100 µM), but not by the inactive enantiomer D-arginine (D-Arg, 100 µM; Fig. 10C
). These effects are consistent with the substrate specificity of neuronal NOS and indicate that sustained and uncontrolled [Ca2+]i responses are mediated by NMDA-activated NO release.
|
To prove that NO can indeed directly modulate NMDA-induced [Ca2+]i transients, we applied two consecutive NMDA pulses (200 µM, 1 min long) separated by a period of 15 min, which caused two distinct [Ca2+]i transients (Fig. 11
A). In one subset of cultures we applied NO (1–3 µM) together with the second pulse of NMDA, resulting in an increase in the duration of the NMDA-induced [Ca2+]i transient (Fig. 11B
). When the same experiment was performed in the presence of CSA (2 µM), this effect was abolished (Fig. 11C
), whereas CSA alone did not have any effect on the NMDA-induced [Ca2+]i transient (Fig. 11D
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
induced a slow rise in basal [Ca2+]i that was independent of extracellular [Ca2+]o. Thus, NO acting directly or through an unidentified intermediate caused the release of Ca2+ from intracellular stores in neurons. Furthermore, CSA, a blocker of the PTP channel, inhibited NO-evoked calcium elevations, indicating that one source of this calcium is the mitochondria. We reported previously that repeated application of NMDA, an activator of NO synthase and NO release, evokes three distinct [Ca2+]i signal pattern (termed reversible, sustained, and uncontrolled) in striatal neurons (20)
. We show here that the frequency of prolonged [Ca2+]i increases (sustained and uncontrolled) was significantly suppressed by inhibition of NO synthase activity, implying that NO potentiates NMDA-induced [Ca2+]i signals by activating additional [Ca2+]i-mobilizing pathways. From these data, we propose that NMDA-activated [Ca2+]i signals can be shaped by initial Ca2+ influx through NMDA receptor channels and by delayed NO-mediated Ca2+ release from internal stores, one of which is the mitochondria. By modulating the kinetics of NMDA-evoked [Ca2+]i signals, NO may modulate the physiological and pathological effects of NMDA receptor stimulation.
In this study we used authentic (gaseous) NO, as it allows for more precise control of extracellular NO concentration and because some classic NO donors exert contrary effects mediated partly by donor breakdown products (i.e., ferrocyanide ions; ref 21
). For instance, whereas iron-cyanide NO donors decreased NMDA-stimulated [3H]norepinephrine release from hippocampal slices, such release was enhanced by authentic NO (22)
. NO donors can be cytotoxic at concentrations below those at which gaseous NO is deleterious (23)
, underscoring the nonspecificity of these donors and the importance of using authentic NO. Elevations of [Ca2+]i were not observed in response to NO oxidation products in the present experiments, suggesting that NO itself is the primary response mediator. We applied NO after the control experiments to ensure that cultures were viable and still had the potential to respond to the stimuli with an increase in intracellular calcium. In both control experiments (Fig. 3B, C
), such calcium increases were induced during our standard NO application. Although it appears that the NO-induced [Ca2+]i responses after either oxygenated NO (Fig. 3B
) solution or nitrite application (Fig. 3C
) were blunted compared to NO alone (Fig. 3A
), this finding does not jeopardize the interpretation of our results because repeated NO application (see Fig. 4
) does not result in blunting of the second calcium transient. The lack of calcium responses during the NO+O2 or NO2- treatment and the absence of such a blunting effect when repeated NO applications were given exclude a contribution of metabolites or pH changes due to oxidation. Our control experiments used a range of nitrite concentrations from low to excessive levels to demonstrate that under no circumstances can one observe a rise in intracellular calcium. We cannot exclude that the oxygenation of the NO solution produced metabolites or pH changes that affected the subsequent calcium response during NO. In any case, the conclusion we draw from these results is that such metabolites do not play a role in the effects we observed.
NO-induced [Ca2+]i signals were reversible and NO did not alter [Ca2+]i signals induced by another form of stimulation (high K+ induced depolarization), indicating these responses did not result from nonspecific disruption of neuronal [Ca2+]i homeostasis.
NO has been shown to suppress evoked [Ca2+]i signals in neurons through either inhibition of NMDA currents (24
, 25)
or voltage-gated Ca channels (26)
, but modulation of basal [Ca2+]i in striatal neurons has not been reported previously. In non-neuronal cells, NO-induced Ca2+ release has been reported in pancreatic ß cells (27)
and sea urchin eggs (28)
. NO is the major activator of soluble guanylate cyclase (18)
resulting in cGMP production. In pancreatic ß cells, NO-evoked Ca2+ release can be mediated by cGMP kinase-dependent activation of adenosine-diphosphate-ribosyl cyclase with subsequent production of cyclic adenosine-diphosphate ribose, a putative endogenous activator of Ca2+-sensitive Ca2+ channels (28
, 29)
. 8-Bromo-cGMP, a cell-permeant analog of cGMP, did not increase [Ca2+] (as has been shown in other cells), indicating that a contribution from this pathway is unlikely in these neurons. In the present study, a second NO application resulted in a [Ca2+]i increase 1.5-fold larger than the first, suggesting an action of NO or an intermediate (e.g., cGMP) on internal Ca2+ store capacity or NO sensitivity.
Our observations that prolonged (sustained and irreversible) NMDA-evoked [Ca2+]i signals were attenuated by NOS inhibition and that NO elicited Ca2+ release from mitochondria are highly suggestive that prolonged NMDA-evoked [Ca2+]i responses were similarly maintained by NO-mediated mitochondrial Ca2+ release In addition to direct NO-mediated Ca2+ release, other mechanisms may contribute to this augmentation. For example, NO-mediated Ca2+ mobilization may activate protein kinase C and we have shown previously that prolonged NMDA-evoked [Ca2+]i signals could be reduced by blockers of protein kinase C (PKC) (20)
. PKC can also initiate prolonged NO production by reducing the Ca2+ sensitivity of NOS (30)
whereas NO can sustain Ca2+ influx through NMDA receptor channels (31)
. Blockade of NOS activity exerted a proportionately greater reduction of uncontrolled responses (98% reduction in response frequency) than sustained (62.5% reduction), suggesting that NO-induced intracellular Ca2+ release makes a greater contribution to the total [Ca2+]i increase in cells exhibiting this response type.
There is accumulating evidence that mitochondria play a crucial role in excitotoxicity-induced apoptosis (32
, 33)
. Here we show that energy depletion potentiates the NO-induced effect on [Ca2+]i. Moreover, our data indicate that NO leads to a decrease of the mitochondrial membrane potential of intact striatal neurons. This finding is consistent with the earlier report that NO disrupts mitochondrial respiration in isolated mitochondria (34)
and hippocampal neurons, resulting in a reversible decrease of the mitochondrial membrane potential (6)
. Mitochondrial compromise and neurodegeneration induced by perinatal asphyxia are prevented by NOS inhibition (35)
. Using direct NO application, Brorson et al. (6)
showed a progressive concentration-dependent depletion of cellular ATP. The permeability transition and associated release of cytochrome c are thought to be part of the apoptotic process. The PTP may be involved in neuronal injury due to excitotoxicity (32
, 36
, 37)
, ischemia (38
39
40)
, or the parkinsonian neurotoxin 1-methyl-4-phenylpyridinium (41)
. The relationship between cytochrome c release and PTP opening, and the effects of NO that may be produced under such pathological conditions (42)
, are not clearly established. NO, S-nitrosothiols, and peroxynitrite are reported to variously inhibit or promote PTP opening (43
, 44)
. Our data show that the NO-induced Ca2+ release is sensitive to CSA. This is consistent with reports that show neuroprotective effects of CSA in a variety of neurodegeneration models (32
, 45)
reported to involve the action of NO (46
, 47)
. Inhibition of respiration could lead to PTP opening, which in turn could cause the release of calcium from mitochondria. NO donors trigger apoptosis in PC12 cells that is partially mediated by opening the mitochondrial PTP, release of cytochrome c, and subsequent caspase activation. NO-induced apoptosis is blocked completely in the absence of glucose, probably due to the lack of ATP (46)
, suggesting that mitochondria may be involved in both types of cell death induced by NO donors: necrosis by respiratory inhibition and apoptosis by opening the PTP.
Another possibility is that NO-induced cytosolic [Ca2+] increases are due to the gradual accumulation of Ca2+ as a result of the loss of mitochondrial calcium uptake mechanisms. However, the CSA sensitivity that we report in this study points to a PTP-mediated process.
Considering that prolonged [Ca2+]i elevations can lead to cell death, those cells that showed sustained or irreversible [Ca2+]i responses may represent a subpopulation of striatal neurons with a high sensitivity to NMDA-mediated neurotoxicity. NMDA-evoked NO production has been implicated in excitotoxicity, and we show that NO can increase neuronal [Ca2+]i directly, suggesting a possible role of NO production in potentiating ([Ca2+]i-dependent) NMDA-mediated neurotoxicity. NMDA receptor activation is a more potent stimulation of NO than activation of non-nMDA glutamate receptors (48)
, and several reports demonstrate that blockade of NOS can ameliorate NMDA toxicity (1
, 2)
; but see (49)
. Moreover, NMDA-induced [Ca2+]i signals are more deleterious than equivalent signals produced by other glutamate receptors or by depolarization (50)
, implying that NMDA neurotoxicity is mediated by factors downstream of [Ca2+]i elevations. For example, NMDA-induced Ca2+ influx leads to NO production (51)
and arachidonic acid-dependent superoxide production (52)
, the combination of which could exert a powerful cytotoxic oxidative stress. Further clarification of the actions of NO on neuronal [Ca2+]i metabolism and the link between permeability transition and mitochondrial calcium release may yield a better understanding of the processes that lead to NMDA/NO-mediated neuronal death and, by extension, point to methods of neuroprotection.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 7, 2002.
Revision received May 30, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Wang, A. G. Teschemacher, J. F. R. Paton, and S. Kasparov Mechanism of nitric oxide action on inhibitory GABAergic signaling within the nucleus tractus solitarii FASEB J, July 1, 2006; 20(9): 1537 - 1539. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. T. A. Meij, C. L. Haselton, K. L. Hillman, D. Muralikrishnan, M. Ebadi, and L. Yu Differential Mechanisms of Nitric Oxide- and Peroxynitrite-Induced Cell Death Mol. Pharmacol., October 1, 2004; 66(4): 1043 - 1053. [Abstract] [Full Text] [PDF]
|
|
|
|
