Isoflurane Reduces the Carbachol-evoked Ca²⁺Influx in Neuronal Cells

SH-SY5Y human neuroblastoma cells were cultured in Roswell Park Memorial Institute 1640 medium with L-Glutamine, supplemented with penicillin (50 U/ml), streptomycin (50 μg/ml), and 12% fetal bovine serum at 37°C, in a humidified atmosphere containing 5% CO². All cell culture components were Gibco BRL products purchased from Life Technologies (Rockville, MD). Experiments were performed on monolayer of cells as previously reported.20Cells were plated on glass coverslips (25-mm diameter) at a density of 2–4 × 10⁴cells/ml (2 ml cell suspension/35 mm culture dish) and used when they formed a confluent monolayer (∼10–16 days after plating). During experimentation cells were continuously perfused with a HEPES buffer containing (in mM) 140 NaCl, 5 KCl, 5 NaHCO³, 10 HEPES, 1 MgCl², 1.5 CaCl², 1 adenosine triphosphate, and 10 glucose (pH 7.4). Experiments were performed at 37°C and the temperature was controlled with a Dual Heater controller TC-344A and an inline heater SH-27B (Warner Instruments Inc, Hamden, CT). The exchange of the solution was carried out with a manifold. The solutions containing 1 mm carbachol, with and without 100 nm conotoxin GVIA, 10 μm nitrendipine, or 100 μm La³⁺, were prepared using the HEPES buffer. Saturated isoflurane (Ohmeda Caribe Inc., Guayana, PR) solutions were prepared in HEPES buffer 24 h in advance in gas-tight containers and diluted to the final concentration (1 mm) immediately before use as previously described.19 

SH-SY5Y cells were loaded with the fluorescent Ca²⁺indicator Fura-228by incubating the cells attached on coverslips in the culture medium containing 5 μm of the acetoxymethyl ester of the dye (Fura-2 am; Molecular Probes, Eugene, OR) for 30 min under culture conditions. After loading, cells were washed three times with the HEPES buffer, and the coverslips were placed into the perfusion chamber and perfused (250 μl/min) for 30 min with the HEPES buffer at 37°C before being exposed to the various drugs. The HEPES buffer without or containing the drugs was perfused at a speed of 250 μl/ml.

The perfusion chamber was set on an inverted microscope (DIAPHOT 300; Nikon, Melville, NY), equipped with a 40× oil-immersion objective (N.A.1.30; Nikon). The microscope was connected to a high-speed multiwavelength illuminator (DeltaRAM V; Photon Technology International Inc., Lawrenceville, NJ). The excitation wavelengths for Fura-2 (340 nm and 380 nm) were alternately generated by a monochromator (every 0.02 s). The emitted fluorescence (from the alternated excitation at 340 and 380 nm) from 15 to 20 cells was filtered with the fluorescence barrier filter BA 515 nm, collected with a photomultiplier (PMT01–710; Photon Technology International Inc.), and digitized at 50 Hz.

Data collection and analysis was carried out using the software Felix (version 1.42a, Photon Technology International), Clampfit (Pclamp 8; Axon Instruments, Foster City, CA), and GraphPad Prism (GraphPad Software, Inc., San Diego, CA). For each treatment (corresponding to data in each figure), experiments were done under different conditions on sister cultures (same plating day) and on three to five culture sets (different plating days). The averaged traces shown in the figures were obtained by lining up the peak values for the evoked [Ca²⁺]^cyttransients. Unless otherwise indicated, the areas were obtained over a period of 300 s starting from the onset of the carbachol evoked [Ca²⁺]^cyttransient on each trace. In figures the data represent the delta ratio (Δ ratio) of the emission of Fura-2 at 515 nm generated by excitation at 340 and 380 nm (ratio 340/380).

Comparison between different groups was performed using unpaired two-tailed Student t  test when there was only one treatment (fig. 1) and one-way analysis of variance Newman-Keuls test when there was more than one treatment (figs. 2 and 3) using the GraphPad Prism (GraphPad Software, Inc.) software.

We have reported that isoflurane (1 mm) reduced the carbachol-evoked [Ca²⁺]^cyttransient (fig. 1A-C).20The isoflurane effect includes a reduction in the peak and area under the peak but not in the width at 50% peak height of the carbachol-evoked [Ca²⁺]^cyttransient (fig. 1A, B, C, D). When the concentration of extracellular Ca²⁺was reduced from 1.5 mm to 150 μm, carbachol still evoked a [Ca²⁺]^cyttransient but its magnitude was lower and its decay was speeded up, as indicated by the reduction in the width at 50% peak height (fig. 1E, F, G, H). These results are consistent with previous observations in SH-SY5Y cells and indicate that the carbachol-evoked [Ca²⁺]^cyt-transient results from Ca²⁺release from intracellular Ca²⁺stores and Ca²⁺entry through the plasma membrane.21,24,27Interestingly, in the presence of low extracellular Ca²⁺, the carbachol-evoked [Ca²⁺]^cyttransient became insensitive to isoflurane (fig. 1E–H,versus  1A–D).

Removal of extracellular Ca²⁺, even for short periods, may induce partial depletion of intracellular Ca²⁺stores. Hence, the elimination of the isoflurane-sensitive component of the carbachol-evoked cytoplasmic Ca²⁺response may still involve reduction of Ca²⁺release from intracellular store rather than elimination of Ca²⁺entry through the plasma membrane. To distinguish between these possibilities we blocked Ca²⁺entry through the plasma membrane by using the nonselective cationic channel blocker La³⁺. La³⁺has been shown to block various voltage-dependent Ca²⁺channels,29,30as well as other cationic channels known as capacitative Ca²⁺channels.26La³⁺alone did not significantly affect the peak but decreased the area and reduced the width at 50% peak height of the carbachol-evoked [Ca ²⁺]^cyttransient (fig. 2). These results suggest that Ca²⁺entry through the plasma membrane mostly contributes to the decay phase of the carbachol-evoked [Ca²⁺]^cyttransient, whereas the decrease in the carbachol-evoked [Ca²⁺]^cytpeak in low extracellular Ca²⁺may reflect partial Ca²⁺depletion from intracellular Ca²⁺stores. In the presence of La³⁺, isoflurane did not produce an additional change in the carbachol-evoked [Ca²⁺]^cyttransient (fig. 2). The effect of La³⁺was stronger than the effect of isoflurane (fig. 2A,versus  fig. 1A), and the main difference was that La³⁺, but not isoflurane, strongly decreased the width of the carbachol-evoked [Ca²⁺]^cyttransient (fig. 1D,versus  fig. 2D). The latter suggests that the isoflurane effect appears to be mostly at the plasma membrane, probably by blocking a cationic channel.

Carbachol, through activation of muscarinic receptors, has been shown to affect voltage-dependent Ca²⁺channels31,32and to allow Ca²⁺entry through other nonselective cationic channels.33–35In SH-SY5Y cells, the predominant voltage-dependent Ca²⁺channels are L-type and N-type.36,37We tested whether these voltage-dependent channels contributed to the carbachol-evoked [Ca²⁺]^cyttransient and, if so, whether they were the isoflurane targets underlying the isoflurane reduction in the carbachol response. It was found that the N-type Ca²⁺channel blocker ω-conotoxin GVIA at a supramaximal concentration (100 nm) did not affect the carbachol-evoked [Ca²⁺]^cyttransient (fig. 3A), whereas the L-type Ca²⁺channel blocker nitrendipine at a supramaximal concentration (10 μm) reduced the carbachol-evoked [Ca²⁺]^cyttransient (fig. 3B). This indicated that under these conditions, exposure to carbachol increases Ca²⁺entry through L-type, but not N-type, Ca²⁺channels. Surprisingly, in the presence of nitrendipine, isoflurane further reduced the peak and area of the carbachol-evoked [Ca²⁺]^cyttransient (fig. 3C, D) without affecting the width at 50% peak (fig. 3E). Therefore, the isoflurane effect on the carbachol-evoked [Ca²⁺]^cyttransient is not attributable to the isoflurane effects on the L-type or N- type Ca²⁺channels but to an isoflurane effect on a La³⁺-sensitive plasma membrane cationic channel.

As previously reported,21,24,27it was found that in the human neuroblastoma cell line SH-SY5Y cells, the carbachol-evoked [Ca²⁺]^cyttransient involves both Ca²⁺release from intracellular Ca²⁺stores and Ca²⁺entry through the plasma membrane. Moreover, we found that the blocking effect of isoflurane on the carbachol-evoked [Ca²⁺]^cyttransient appears to be mediated by blocking the carbachol-evoked Ca²⁺entry through the plasma membrane. This isoflurane sensitive Ca²⁺entry involves a cationic channel that is different from the L-type or N-type voltage-dependent Ca²⁺channels. These results together with our previous observations19,20indicate that at the concentrations used, isoflurane blocks only part of the carbachol-evoked [Ca²⁺]^cytresponse, apparently at a site at the plasma membrane that is distal to the muscarinic receptor.

In SH-SY5Y cells the carbachol-evoked [Ca²⁺]^cyttransient is blocked by atropine20,38and is resistant to the N-type channel blocker (ω-conotoxin).38Previously, it was found that the carbachol-evoked [Ca²⁺]^cytincrease was also resistant to a maximal effective concentration (1 μm) of the L-type channel blocker dihydropyridine +PN 200–110.38However, in this study we found that the L-type channel blocker, nitrendipine (10 μm) reduced the carbachol-evoked [Ca²⁺]^cyttransient without eliminating its sensitivity to isoflurane. This was surprising because volatile anesthetics are known to reduce the magnitude of L-type Ca²⁺channel currents.39,40If L-type channels are contributing to the carbachol-evoked [Ca²⁺]^cyttransient, blocking them should reduce the isoflurane sensitivity of the carbachol-evoked [Ca²⁺]^cyttransient, which did not occur. One possible explanation is that at the high concentration of nitrendipine used in this study, nitrendipine may be reducing the carbachol-evoked [Ca²⁺]^cyttransient by interfering with the G-protein-linked muscarinic receptors rather than by blocking L-type Ca²⁺currents.41–43 

Because La³⁺, but not N-type or L-type Ca²⁺channel blockers, eliminated the isoflurane action on the carbachol-evoked [Ca²⁺]^cyttransient, isoflurane may be reducing the carbachol-evoked [Ca²⁺]^cyttransient by blocking a cationic channel at the plasma membrane. As La³⁺, but not isoflurane, reduced the width at 50% peak, it indicates that SH-SY5Y cells may express several nonvoltage dependent cationic channels that mediate Ca²⁺influx upon muscarinic activation and that isoflurane acts only in a subgroup of these channels.

There are at least two possible candidates for the isoflurane-sensitive plasma-membrane cationic channel, an inositol IP3-activated IP3-channel and a capacitative Ca²⁺channel. There is evidence suggesting the presence of plasma membrane-IP3 receptors in mammalian neurons. IP3-activated inward Ba²⁺currents have been recorded in excised inside-out patches of primary cultured Purkinje cells44and in olfactory neurons of rat.45However, halothane has been shown to increase, rather than decrease, Ca²⁺currents through IP3 receptors.46Moreover, it has been argued that there are no IP3 receptors on the plasma membrane but a group of IP3 receptors located very close to the plasma membrane that on activation in turn activate cationic channels on the plasma membrane.47,48Capacitative Ca²⁺influx is mediated by channels that are opened in response to depletion of intracellular Ca²⁺stores.49There appear to be various types of capacitative Ca²⁺channels.49,50Opening of capacitative channels after activation of G-protein-linked receptors, such as muscarinic receptors, involves receptor-mediated activation of phospholipase C and Ca²⁺release by IP3.49Isoflurane has been shown to inhibit the histamine-induced Ca²⁺influx in primary cultures of human endothelial cells.51In rat glioma C6 cells, volatile anesthetics appear to have different inhibitory effects on capacitative Ca²⁺influx such that strong inhibition is observed with halothane but not with enflurane.52It is then possible that isoflurane is inhibiting muscarinic-activated capacitative Ca²⁺influx in the SH-SY5Y cells.

We would like to postulate that this isoflurane-sensitive cationic channel contributes either to the anesthetic potency or to the side effects of isoflurane. The previously reported variable effects of muscarinic blockers (as with other G-protein linked receptors) on the minimal alveolar anesthetic concentration of inhaled anesthetics might in part reflect differences in magnitude of the muscarinic-mediated modulation of various cationic channels. Although the muscarinic-mediated activation of the isoflurane-sensitive cationic channel might reduce the isoflurane potency, the muscarinic-mediated inhibition of voltage-dependent channels31,32might increase the isoflurane potency. The net effect of a muscarinic agent on the isoflurane potency for reducing the muscarinic-evoked increases in [Ca²⁺]^cytwould then depend on the contribution of each of the cationic channels in the different brain and spinal regions where the muscarinic agents are applied.

We previously reported that the isoflurane-action on the carbachol-evoked [Ca²⁺]^cyttransient required that the caffeine-sensitive Ca²⁺stores were not depleted (by either KCl or caffeine pretreatment) and that ryanodine-sensitive Ca²⁺release channels were open.19,20One possible explanation is that there is an open conformation of ryanodine-sensitive Ca²⁺release channels that interacts with the muscarinic-activated cationic channels and prevents their opening. Another explanation is that because of distinct spatial distribution, Ca²⁺release through the ryanodine-sensitive channels blocks the isoflurane-sensitive cationic channels, whereas Ca²⁺release through IP3-sensitive channels opens the isoflurane-sensitive cationic channels.

In summary, we postulate that in SH-SY5Y cells there is an isoflurane-sensitive cationic channel at the plasma membrane that is activated by carbachol and inhibited by La³⁺, isoflurane, and, possibly, through an interaction with the ryanodine-sensitive Ca²⁺release channel or by a Ca²⁺release through these channels (fig. 4). As discussed above, a possible candidate is a isoflurane-sensitive capacitative Ca²⁺channel. This potential target of isoflurane may serve as a site at which isoflurane may affect at least some of the actions of most of the G-protein linked receptors. The magnitude of the isoflurane effect on a given G-protein linked receptor would then be determined in part by the ability of the receptor to activate these isoflurane-sensitive capacitative Ca²⁺channels.

The authors thank June Biedler, Ph.D. for providing the SH-SY5Y human neuroblastoma cells (Sloan-Kettering Institute for Cancer Research, Rye, New York).