Abstract
The effect of cyanide on the N-methyl-D-aspartate (NMDA)-stimulated increase in cytosolic free calcium ([Ca++]i) was studied by microfluorescence in fura-2-loaded cerebellar granule cells. The response to NMDA was enhanced by NaCN over a concentration range of 20 to 100 μM. These concentrations of NaCN in the absence of NMDA had no effect on basal [Ca++]i. In comparison, NaCN did not affect K+-depolarization-induced [Ca++]i elevation. The NaCN potentiation of NMDA-evoked [Ca++]i elevation was blocked by addition of Mg++ and by the NMDA receptor antagonists 2-amino-5-phosphono-valeric acid and (+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohept-5,10-imine maleate. Pretreatment of the cells with pregnenolone sulfate or arachidonate, known modulators of the NMDA receptor, enhanced NaCN action. The voltage-sensitive calcium channel blockers nifedepine and diltiazem did not affect the NaCN-induced potentiation. Additionally, the NaCN action was not altered when tetrodotoxin was used to block Na+ channel-mediated glutamate release. In patch-clamp studies, NaCN increased the amplitude and duration of NMDA-stimulated whole-cell currents. NaCN also enhanced the NMDA receptor response in single-channel patch-clamp experiments. In the outside-out patch recording configuration, NaCN increased the NMDA receptor channel opening frequency without affecting single-channel conductance or mean channel open time. These results indicate that cyanide interacts directly with the NMDA receptor channel complex to enhance receptor-mediated responses.
Cyanide is a rapid-acting toxicant in which the CNS is one of the major target organs. Neurological dysfunction induced by cyanide includes respiratory distress, seizures and convulsions (Way, 1984; Borowitz et al., 1992), and in some cases a Parkinson-like condition may develop as a post-toxicity sequela (Uitti et al., 1985;Carella et al., 1988; Valenzuela et al., 1992;Rosenow et al., 1995). In animals, cyanide can produce a Parkinson-like response in which striatal neurodegeneration occurs (Kanthasamy et al., 1994).
Recent studies have shown that an interaction between cyanide and the glutamate neurotransmitter system plays an important part in cyanide neurotoxicity. Cyanide induces the release of glutamate from neuronal stores and alters the brain levels of glutamate (Persson et al., 1985; Patel et al., 1991). The increased extracellular levels of glutamate may result in overstimulation of glutamate receptors, leading to excitotoxic responses (Rothman, 1984). NMDA receptor-mediated Ca++ influx appears to be a key event in the excitotoxic process initiated by cyanide. In neuronal cells, specific NMDA receptor antagonists such as APV and MK-801 block cyanide-induced [Ca++]i elevation (Michaels and Rothman, 1990; Cai and McCaslin, 1992; Patel et al., 1992) and prevent neuronal cytotoxicity (Goldberg et al., 1987; Pauwels et al., 1989; Patel et al., 1993). Neither non-NMDA receptor antagonists nor VSCC blockers alter the cytotoxic response to cyanide.
The purpose of the present study was to characterize further the interaction between cyanide and NMDA receptor-mediated responses using cerebellar granule cells. It was proposed that cyanide exerts a direct effect on the NMDA receptor channel complex to modulate the response to NMDA. It was shown that cyanide enhanced the response to NMDA by increasing the probability of NMDA receptor channel opening, resulting in an increased amplitude and duration of the NMDA response.
Materials and Methods
Cell culture.
Primary cultures of rat cerebellar granule cells were prepared as described previously by Gunasekar et al., (1996). Cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 20 mM glucose, 25 mM KCl and 5000 units/l penicillin/streptomycin at pH 7.4 on poly-l-lysine (MW 30,000–70,000)-coated sterile cover glass in 6-well 35-mm culture dishes. Non-neuronal cell proliferation was prevented by adding cytosine arabinonucleoside (10 μM) 18 hr after plating. Cultures generated by this method contain more than 95% granule cells. Mature cells (9–12 days in vitro) were used in all experiments.
[Ca++]i measurement.
Cytosolic free Ca++ levels were determined in granule cells as previously described (Patel et al., 1994). Briefly, cells grown on glass coverslips were loaded with fura-2 by incubation with 5 μM fura-2 AM for 45 min at room temperature in Locke’s solution containing (in mM) NaCl, 154; KCl, 5.6; MgCl2, 1.0; CaCl2, 2.3; NaHCO3, 3.6; HEPES, 5.0 and D-glucose, 5.6; pH, 7.4. Dye loading was terminated by replacing the loading solution with fresh Locke’s buffer. For microfluorescence measurement of [Ca++]i, the coverslips were mounted in a thermostatically controlled cell chamber maintained at 37°C (Medical System Inc., Greenvale, NY) and placed on an inverted stage Nikon Diaphot-TMD microscope connected to a SLM 8000 C spectrofluorometer (SLM-AMINCO, Inc., Urbana, IL). [Ca++]i measurements were made on a small group (1–3 cells) of fura-2-loaded neurons. After basal [Ca++]i levels were obtained, test compounds were added at the various time points specified via an inlet port connected to syringes for media replacement (the drug solution equilibrated in the chamber within 10 sec of addition), and fluorescence was monitored at 510 nm while excitation wavelength was alternated between 340 nm and 380 nm every 15 sec. Except for studying the effect of extracellular Mg++, we routinely omitted MgCl2 from the solution during the drug treatments. The fluorescence was converted on a real-time basis to Ca++concentrations by use of the SLM 8100 software Intracellular Probe measurement system (SLM-AMINCO, Inc., Urbana, IL) according to the fluorescence ratio method of Grynkiewicz et al., (1985).
Patch-clamp recordings.
Patch-clamp experiments were conducted in primary cerebellar granule cells 10 to 12 days old grown on 25-mm glass coverslips. NMDA receptor-mediated currents were recorded in both whole-cell and single-channel configurations. The same external and internal solutions were used for whole-cell and single-channel recordings. The external solution contained (in mM) NaCl, 150; KCl, 2.8; CaCl2, 1 and Na-HEPES, 10, pH 7.2, and the pipette solution contained CsCl, 150; NaCl, 4; CaCl2, 0.5; K-EGTA, 5 and K-HEPES, 10, pH 7.2. All test compounds were dissolved in the bath solution and diluted to final concentrations as indicated. Drugs were applied to whole cells or excised membrane patches by close distance (< 20 μM) pressure ejection from blunt-tipped (10–20 μm), fire-polished micropipettes. Upon application of pressure, the concentration of drug bathing the cell rises to greater than 90% of the concentration in the micropipette in less than 1 sec (Choi et al., 1977; Dunlap and Fischbach, 1981). For experiments, cultures on glass coverslips were transferred into a chamber made from a 35-mm plastic culture dish containing 1 ml of external solution, which was then placed on the stage of an inverted phase-contrast microscope at room temperature. High-resistance seals were obtained with borosilicate glass patch pipettes (4–7 MΩ). Recordings were performed with an Axo-patch 200A patch-clamp amplifier (Axon Instruments, Foster City, CA). Currents were filtered at 3 kHz with an 8-pole Bessel filter (Frequency Devices, Haverhill, MA). Data acquisition and analysis were performed as previously described (Twitchell and Rane, 1994) with Pulse (Instrutech Corp., Great Neck, NY) and TAC (HEKA Elektronik, Göttingen, Germany) software.
Chemicals.
The following drugs and chemicals were used in this study: Dulbecco’s modified Eagle’s medium, fetal bovine serum, penicillin/streptomycin (Gibco, Grand Island, NY), tissue culture dishes (Costar, Cambridge, MA), glass coverslips (25 mm) (Fisher Scientific, Fair Lawn, NJ), fura-2/AM (Molecular Probes, Eugene, OR), diltiazem hydrochloride (Marion Laboratories, Inc., Kansas City, MO) and MK-801 (Research Biochemicals Inc., Natick, MA). All other chemicals were from Sigma Chemical Co. (St. Louis, MO).
Statistics.
Statistical differences between treatments were determined by Student’s t test or by one-way analysis of variance (ANOVA) with a Newman-Keuls procedure for multiple comparisons. Differences were considered significant at P < .05.
Results
Enhancement of NMDA-stimulated [Ca++]ielevation by cyanide.
NMDA treatment in cultured neurons has been shown to induce [Ca++]i elevation due to activation of the NMDA receptor-coupled ion channels (MacDermottet al., 1986). In cerebellar granule cells, NMDA treatment (50 μM, in the presence of 10 μM glycine) in Mg++-free medium induced a rapid biphasic increase of [Ca++]i followed by a maintained plateau (fig. 1A). Addition of 50 μM NaCN at the NMDA plateau level resulted in a further increase of [Ca++]i. The enhancement of NMDA-induced [Ca++]i elevation by NaCN was not dependent on the order in which NaCN was added.
The magnitude of cyanide-induced enhancement was concentration-dependent with approximately 15%, 25% and 60% increases in the NMDA-stimulated [Ca++]iplateau produced by 20 μM, 50 μM and 100 μM NaCN, respectively (fig. 1B). These doses of NaCN in the absence of NMDA had no effect on the basal [Ca++]i (data not shown).
Effect of cyanide on K+-depolarization-induced [Ca++]i elevation.
The effect of cyanide on K+-depolarization-induced [Ca++]i elevation was also determined. Depolarization-induced Ca++ influx is mediated primarily by activation of VSCCs. In the present study, K+depolarization induced a rapid increase in [Ca++]i that peaked at about 10-fold of basal [Ca++]i. The peak then declined and stabilized at about 4-fold of the basal [Ca++]i, which could be further decreased by the VSCC blocker diltiazem (fig. 2A). When NaCN (50 μM) was added to the K+ response plateau, there was no significant change in [Ca++]i (fig. 2, B and C).
Effects of APV, MK-801 and extracellular Mg++ on cyanide-induced enhancement of NMDA responses.
The effects of extracellular Mg++ and the NMDA receptor antagonists APV and MK-801 on NaCN-induced enhancement of NMDA response were examined. As demonstrated in figure 3A, the addition of 100 μM APV, a competitive NMDA receptor antagonist, induced a rapid decline in [Ca++]i from the NMDA plateau response. Subsequent application of 50 μM NaCN resulted in no further [Ca++]i elevation. MK-801 (1 μM), a noncompetitive NMDA receptor antagonist, showed a similar inhibitory effect on the NaCN action. APV and MK-801 resulted in 59% and 51% attenuation of NMDA-evoked [Ca++]i, respectively, and cyanide did not affect [Ca++]i in the presence of these antagonists (fig. 3B).
The NMDA receptor channel can be blocked by Mg++ in a voltage-dependent fashion (Nowak et al., 1984). In this study, the presence of MgCl2 (1.2 mM) in the extracellular media significantly decreased the NMDA-stimulated [Ca++]i elevation (fig. 3B) and blocked the NaCN-induced enhancement of NMDA response.
Effect of receptor modulators on cyanide-induced enhancement of NMDA responses.
In order to gain insight into the mechanism of cyanide enhancement of the NMDA response, the cells were pretreated with compounds known to modulate the NMDA-induced Ca++influx. PMA did not alter the response to cyanide, whereas DTT and pregnenolone enhanced the response (fig. 4). Arachidonic acid, which is known to potentiate NMDA-induced calcium influx, enhanced the response to NaCN in the presence of NMDA.
Effects of diltiazem and nifedepine on cyanide-induced enhancement of NMDA responses.
NMDA receptor activation promotes the influx of cations such as Na+ and Ca++, which can lead to membrane depolarization. This, in turn, can activate membrane VSCCs, resulting in additional Ca++ influx. To examine the possibility that secondary opening of VSCCs was involved in the NaCN-induced [Ca++]i elevation, we used two VSCC blockers, diltiazem (10 μM) and nifedepine (1 μM). These compounds had no effect on the NMDA response plateau. Additionally, the NaCN response was not affected by treatment with diltiazem or nifedepine (fig. 5, A and B).
Effects of tetrodotoxin on cyanide-induced enhancement of NMDA response.
A possible mechanism for the NaCN enhancement of the NMDA response may involve presynaptic glutamate release, which can then activate NMDA and non-NMDA receptors to promote Ca++influx. NaCN has been shown previously to induce glutamate release in brain slices (Patel et al., 1991), and NaCN-induced neuronal cytotoxicity can be prevented by tetrodotoxin, which blocks Na+ channel-mediated glutamate release (Rothman, 1984). In the present study, tetrodotoxin treatment did not affect the NaCN-induced enhancement of the NMDA response (fig. 6).
Whole-cell recordings.
The microfluorescence studies indicated that NaCN specifically enhanced the NMDA-stimulated elevation of [Ca++]i. To study the underlying mechanism in more detail, we conducted patch-clamp experiments. In patch-clamp studies, the effect of NaCN on NMDA receptor-mediated response was measured directly as changes in receptor-mediated currents.
Figure 7A illustrates representative NMDA-activated whole-cell currents recorded from cerebellar granule cells in the presence or absence of CN. In the presence of glycine (10 μM), NMDA application (50 μM; 8 sec) evoked whole-cell currents with kinetics characteristic of NMDA receptor channels (holding potential, −60 mV). During the 8-sec application of NMDA, there was an initial peak current followed by a gradual decay that reflects NMDA receptor desensitization. When agonist application was terminated, the evoked current showed a slow deactivation time course. The co-application of NaCN (100 μM) increased the amplitude and duration of the NMDA-evoked currents. As shown in figure 7B, the amplitudes of NMDA whole-cell peak currents were doubled in the presence of NaCN. The duration of NMDA response was defined as the time for 80% recovery from peak-level currents. NaCN evoked a 90% increase of the duration of NMDA-stimulated whole-cell currents. NaCN potentiation of whole-cell NMDA currents was observed regardless of the order of NMDA or NMDA/CN application.
Single-channel measurement.
The effect of NaCN on the NMDA receptor response was examined at the single-channel level in excised outside-out membrane patches, a condition in which the involvement of cytoplasmic factors is minimized. Each patch was exposed to both NMDA and NMDA/CN, and changes in channel activity were determined by analysis of 3 to 12 sec of continuous recording for each condition. In the patch shown in figure 8, the presence of NaCN (100 μM) increased the number of NMDA-stimulated channel openings observed for a given recording period. As a result,
Discussion
In cerebellar granule cells, the NMDA-stimulated elevation of [Ca++]i was enhanced 15% to 60% by cyanide at concentrations of 20 to 100 μM, which in the absence of NMDA had no effect on basal [Ca++]i. As demonstrated by patch-clamp studies, cyanide enhanced the NMDA receptor-mediated currents. In the whole-cell configuration, cyanide increased the amplitude and duration of NMDA-stimulated inward currents, and in outside-out patches, it increased the NMDA receptor single-channel opening frequency without affecting the unitary conductance or mean channel open time. Because the effect of cyanide persists in cell-free patches, activation of intracellular second messengers does not appear to be a prerequisite for cyanide’s action. It is concluded that cyanide exerts a direct effect on the NMDA receptor to modulate positively the response to NMDA.
This study confirms and extends the observations of Cai and McCaslin (1992), who concluded that cyanide interacts with the excitatory amino acid receptor. It was observed that cyanide enhanced the influx of Ca++ after activation of the NMDA receptor and that VSCC blockers did not alter the cyanide response. Our previous study showed, in cultured hippocampal neurons, that cyanide in high concentrations (1 mM) enhanced NMDA mediated Ca++ influx and inward current by interacting with the Mg++ block of the receptor (Patelet al., 1994). That effect appeared to be energy-independent and could be explained by a direct interaction of cyanide with an allosteric regulatory site on the receptor.
Activation of NMDA receptor channels and concomitant Ca++influx have been shown to be key events in cyanide-induced excitotoxicity. In a number of neuronal models, cyanide-induced [Ca++]i elevation and cytotoxicity were prevented by selective NMDA receptor antagonists (Michaels and Rothman, 1990; Dubinsky and Rothman, 1991; Cai and McCaslin, 1992; Patelet al., 1992, 1993). It has also been shown that cyanide enhanced MK-801 binding, reflecting an increased NMDA receptor activity and channel opening (Akira et al., 1994; Patel et al., 1994). The present study provides direct evidence that cyanide at biologically relevant concentrations (micromolar range) can potentiate the NMDA receptor-mediated response (Ballantyne, 1983;MacMillan, 1989).
To determine whether the response to cyanide resulted from a direct interaction with the NMDA receptor channel, we determined the effect of cyanide on other processes that regulate Ca++ influx and cytosolic levels. Even though NMDA- and K+-depolarization-induced [Ca++]ielevation reached similar levels, cyanide affected only the NMDA response. This indicates that [Ca++]ielevation per se is not a prerequisite for the enhancement and that it is not due to an altered ability of the cell to buffer changes in cytosolic [Ca++]i levels.
K+-depolarization-induced elevation of [Ca++]i results from opening of VSCCs, followed by their inactivation. The long-lasting plateau results from a small proportion of L-type VSCCs remaining activated over a prolonged period (Murphy et al., 1987). In the present study,l-channels mediated the response to KCl, because diltiazem rapidly lowered the K+ plateau response. The failure of cyanide to affect the K+-induced plateau indicates that cyanide did not influence l-type VSCC-mediated Ca++ influx.
Activation of the NMDA receptor promotes the influx of cations, including Na+ and Ca++, which can lead to membrane depolarization. In turn, depolarization can activate membrane VSCCs, leading to additional Ca++ influx (Mayer and Miller, 1990). Selective pharmacological antagonists were used to differentiate between Ca++ influx through NMDA-gated channels and through VSCCs. The cyanide-induced enhancement of [Ca++]i was blocked by the NMDA receptor antagonists APV and MK-801 and by Mg++. The VSCC blockers nifedipine and diltiazem did not interfere with the cyanide action. It was concluded that the cyanide-induced enhancement of Ca++influx is mediated primarily by NMDA receptor-coupled channels.
Enhancement of whole-cell currents can arise from changes in NMDA receptor single-channel properties, including single-channel conductance, mean channel open time and the frequency of channel openings. In the present study, we found that cyanide increased NMDA receptor channel opening frequency without affecting mean channel open time or single-channel conductance. Like cyanide, a number of other NMDA receptor allosteric modulators also affect the NMDA receptor channel opening probability. The allosteric modulators that increase the opening probability include glycine (Johnson and Ascher, 1987), the polyamine spermine (Rock and MacDonald, 1992), the sulfhydryl reducing agent DTT (Tang and Aizenman, 1993), protein kinase C (Chen and Huang, 1992), arachidonate (Miller et al., 1992) and the neurosteroid pregnenolone sulfate (Bowlby, 1993). Other modulators have been reported to decrease the receptor channel opening probability, including hydrogen ions (Tang et al., 1990), Zn++ (Legendre and Westbrook, 1990), the opioid peptide dynorphin (Chen et al., 1995) and nitric oxide (Fagniet al., 1995). It is interesting to note that in this study, arachidonate and pregnenolone increased the response to cyanide. It is possible that cyanide directly interacts with one or more of the receptor modulatory sites to enhance the NMDA response.
Cyanide neurotoxicity is associated with activation of NMDA receptors and sequent influx of [Ca++] (Dubinsky and Rothman, 1991;Patel et al., 1992, 1993). In neuronal models, cytotoxicity is mediated exclusively by the NMDA receptor, because selective antagonists prevent cell death (Patel et al., 1991). This is caused in part by release of endogenous glutamate by cyanide, leading to receptor activation (Patel et al., 1991). Activation of the receptor by glutamate would be enhanced by a direct interaction of cyanide with the receptor, leading to cytotoxic elevation of [Ca++]. Recently, we have shown in cerebellar granule cells that cyanide-induced activation of the NMDA receptor produces simultaneous generation of nitric oxide and reactive oxygen species, leading to cellular oxidative stress and cytotoxicity (Gunasekaret al., 1996). It is apparent that activation of the receptor is an initiating event in the cytotoxic response to cyanide that is independent of the effect of cyanide on the cell’s energy reserves.
The potentiation of NMDA-induced Ca++ influx by cyanide has important toxicological implications. Many physiological functions of NMDA receptors are mediated by Ca++, including activation of the Ca++-dependent signal transduction cascade involved in nerve cell development and synaptic plasticity (Mayer and Miller, 1990; Gozlan, et al., 1995). In pathological conditions, such as in excitotoxic neuronal injury, excessive NMDA receptor activation and subsequent intracellular Ca++ overload lead to cellular injury (Choi, 1987; Garthwaite and Garthwaite, 1987; Choiet al., 1988). Enhancement of the NMDA response would probably accelerate or potentiate the excitotoxic response and hence may play an important role in cyanide-induced neurotoxicity.
In summary, the interaction of cyanide with NMDA receptor-mediated responses was characterized. Cyanide enhanced NMDA receptor-mediated inward currents as well as [Ca++]i elevation. Because the potentiation of NMDA response was seen in excised membrane patches, the cyanide modulation is the result of a direct action on the NMDA receptor channel complex and does not require an intact cell or activation of intracellular second messengers. It is possible that modulation of NMDA receptor-mediated responses is involved in the acute manifestations of cyanide intoxication due to CNS dysfunction (respiratory distress, seizures and convulsions) and the post-intoxication sequelae associated with selective neurodegeneration.
Footnotes
-
Send reprint requests to: Gary E. Isom, Ph.D, Dept. of Medicinal Chemistry & Molecular Pharmacology, Purdue University, West Lafayette, IN 47907-1334.
-
↵1 This work was supported in part by NIH grant ES04140.
- Abbreviations:
- NMDA
- N-methyl-D-aspartate
- [Ca++]i
- cytosolic free Ca++
- VSCC
- voltage-sensitive calcium channels
- APV
- 2-amino-5-phosphono-valeric acid
- MK801
- (+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohept-5,10-imine maleate
- DTT
- dithiothreitol
- PMA
- phorbol 12-myristate 13-acetate
- Received July 15, 1996.
- Accepted November 8, 1996.
- The American Society for Pharmacology and Experimental Therapeutics