HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplemental data All Versions of this Article: 578/2/481    most recent jphysiol.2006.119008v1 Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gover, T. D. Articles by Weinreich, D. Search for Related Content PubMed PubMed Citation Articles by Gover, T. D. Articles by Weinreich, D. Related Collections Neuroscience

¹ The Neuroscience Program, University of Maryland, Baltimore, MD 21201-1559, USA
² Medical Biotechnology Center, University of Maryland Biotechnology Institute, and Departments of
³ Physiology
⁴ Pharmacology and Experimental Therapeutics, University of Maryland, School of Medicine, Baltimore, MD 21201-1559, USA
⁵ Department of Biochemistry and Immunology, Institute of Biological Sciences, Federal University of Minas Gerais, Belo, Brazil 31270-901

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Ca²⁺ is vital for release of neurotransmitters and trophic factors from peripheral sensory nerve terminals (PSNTs), yet Ca²⁺ regulation in PSNTs remains unexplored. To elucidate the Ca²⁺ regulatory mechanisms in PSNTs, we determined the effects of a panel of pharmacological agents on electrically evoked Ca²⁺ transients in rat corneal nerve terminals (CNTs) in vitro that had been loaded with the fluorescent Ca²⁺ indicator, Oregon Green 488 BAPTA-1 dextran or fura-2 dextran in vivo. Inhibition of the sarco(endo)plasmic reticulum Ca²⁺-ATPase, disruption of mitochondrial Ca²⁺ uptake, or inhibition of the Na⁺–Ca²⁺ exchanger did not measurably alter the amplitude or decay kinetics of the electrically evoked Ca²⁺ transients in CNTs. By contrast, inhibition of the plasma membrane Ca²⁺-ATPase (PMCA) by increasing the pH slowed the decay of the Ca²⁺ transient by 2-fold. Surprisingly, the energy for ion transport across the plasma membrane of CNTs is predominantly from glycolysis rather than mitochondrial respiration, as evidenced by the observation that Ca²⁺ transients were suppressed by iodoacetate but unaffected by mitochondrial inhibitors. These observations indicate that, following electrical activity, the PMCA is the predominant mechanism of Ca²⁺ clearance from the cytosol of CNTs and glycolysis is the predominant source of energy.

(Received 10 August 2006; accepted after revision 6 November 2006; first published online 9 November 2006)
Corresponding author D. Weinreich: Department of Pharmacology and Experimental Therapeutics, University of Maryland, School of Medicine, Bressler Research Building, Room 4-002, 655 West Baltimore Street, Baltimore, MD 21201-1559, USA. Email: dweinrei{at}umaryland.edu

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Primary sensory neurons play a vital role in a wide range of physiological and pathophysiological processes, including detecting noxious and non-noxious stimuli, releasing trophic factors into peripheral tissues, and initiating neurogenic inflammation. Regulation of these and other processes by sensory neurons is, in many ways, dependent on intracellular free Ca²⁺ concentration ([Ca²⁺]_i) and therefore on the mechanisms that underlie Ca²⁺ homeostasis.

Ca²⁺ has many diverse and important functions in cells and consequently its regulation, both temporally and spatially, must be tightly controlled for proper signalling to occur. In neurons, this regulation includes voltage-dependent Ca²⁺ channels (VDCCs) on the plasma membrane (PM), so that electrical activity may be ‘sensed’ by the internal signalling mechanisms of the cell. Peripheral sensory afferents express multiple types of VDCCs, including L-, N-, R- and T-type (Fox et al. 1987; Schild et al. 1994; Kim & Chung, 1999). In addition to Ca²⁺ influx across the PM, sensory neurons have effective mechanisms for amplifying Ca²⁺ signals, including Ca²⁺-induced Ca²⁺ release (CICR) from intracellular stores through ryanodine receptors. Ca²⁺ can also be released from intracellular stores through inositol 1,4,5-trisphosphate receptors (Thayer et al. 1988). To terminate Ca²⁺ signals, efficient Ca²⁺ clearance mechanisms must also exist. These include transport of Ca²⁺ across the PM by the PM Ca²⁺-ATPase (PMCA), and by the Na⁺–Ca²⁺ exchanger (NCX) (Verdru et al. 1997; Usachev et al. 2002). Ca²⁺ may also be transported from the cytosol into Ca²⁺-sequestering intracellular organelles, the most prominent of which are the endoplasmic reticulum (ER) and mitochondria. Both organelles have been shown to take up Ca²⁺ from the cytoplasm of sensory neurons (Shishkin et al. 2002).

To date, investigations of Ca²⁺ homeostasis in primary sensory afferents have been limited to cell body and nerve trunk preparations (Thayer & Miller, 1990; Mayer et al. 1999). Although these studies have greatly increased our knowledge of sensory neuron physiology and Ca²⁺ homeostasis, caution is necessary when extrapolating such studies to peripheral sensory nerve terminals. Until recently, the small size (0.15–0.25 µm in diameter; Whitear, 1960) and physical inaccessibility of these terminals have precluded direct measurements of Ca²⁺ currents and Ca²⁺ signalling. We have developed a novel preparation for measuring Ca²⁺ transients evoked by electrical or chemical stimulation in the sensory nerve terminals of the rat cornea (Gover et al. 2003). The cornea has unique properties that make it an ideal preparation for studying Ca²⁺ dynamics in nociceptive sensory nerve terminals. In addition to being transparent, the cornea has the greatest density of sensory nerve innervation of any tissue (Lele & Weddell, 1956). The nerve terminals of the cornea are all free nerve endings which reside in the superficial epithelial cell layers no more than 50 µm from the surface of the cornea (Fig. 1; Zander & Weddell, 1951; MacIver & Tanelian, 1993). Ultrastructural studies have demonstrated that corneal nerve terminals are truly ‘free’, without Schwann cell ensheathment or fine specializations (Muller et al. 1996). The combination of a high density of nociceptive innervation, simple tissue structure, proximity of nerve terminals to the tissue surface and the transparency of the cornea makes the cornea an exceptional preparation for functional neuronal imaging. In the current work, we have used this preparation to examine [Ca²⁺]_i regulation during electrically evoked Ca²⁺ transients.

View larger version (106K): [in this window] [in a new window]   Figure 1.  Anatomy of sensory nerve terminals in the rat cornea A, flattened Z-stack image of cornea loaded with FM1-43. On the basis of their distinctive morphology, nerve fibres and epithelial cells are artificially coloured in yellow and green, respectively. Arrow indicates nerve terminal residing in the superficial epithelial cell layer. Scale bar, 50 µm. B, same Z-stack as in A but flattened with a 90 deg orientation with respect to A. In this orientation, the thickness of the epithelial cell layer (Epi Layer) is evident. Double arrows indicate surface of the epithelial cell layer. Asterisk indicates subepithelial nerve plexus residing in the stromal layer (collagen layer) of the cornea. Scale bar, 100 µm.

	Methods

Top Abstract Introduction Methods Results Discussion References

Experiments were performed on isolated corneas from male Sprague-Dawley rats (140–300 g). All animal procedures were approved by the Institutional Animal Care and Use Committee of the University of Maryland, Baltimore. For cornea dye loading, 0.25–1.0 µl of a solution containing 0.9% w/v NaCl, 20% w/v Oregon Green 488 BAPTA-1 dextran (OGB-1 dextran, 10 kDa; Molecular Probes, Eugene, OR, USA) or 10% w/v tetramethylrhodamine dextran (10 kDa; Molecular Probes) and 1–2% v/v Triton X-100 (Sigma, St Louis, MO, USA) was deposited on each cornea of a ketamine-anaesthetized animal for 1 min. After dye exposure, the eyes were rinsed with Ca²⁺- and Mg²⁺-free phosphate buffered saline (PBS). Rats were killed 16–60 h later by pentobarbital (100 mg kg^–1, I.P.). Corneas were dissected directly from the animal immediately after death. Isolated corneas were maintained in an oxygenated Locke solution containing (mM): glucose 10, NaCl 136, KCl 5.6, NaH₂PO₄ 1.2, NaHCO₃ 14.3, MgCl₂ 1.2 and CaCl₂ 2.2; pH 7.4. For experiments determining the effects of alkaline pH on Ca²⁺ regulation, we maintained the cornea in a solution containing (mM): glucose 10, NaCl 150, KCl 5.6, NaHCO₃ 5.0, tris(hydroxymethyl) aminomethane (Tris base) 10, MgCl₂ 1.2 and CaCl₂ 2.2.

Electrical stimulation

Corneal nerve terminals (CNTs) were activated by electric field stimulation with platinum electrodes placed on either side of the recording chamber. Electrical stimulation (50 pulses 10 Hz, 2 ms duration) were generated by a stimulator and delivered through a stimulus isolation unit (both from Astro-Medical, West Warwick, RI, USA). The stimulation frequency was chosen to be representative of physiological firing frequencies in CNTs (MacIver & Tanelian, 1993). The stimulation voltage was greater than threshold to ensure consistent neuronal activation (120–150 V), as detected by transient rises in [Ca²⁺]_i. Corneas were continuously superfused with oxygenated Locke solution at room temperature (22–24°C) except where noted.

Calcium imaging

Ca²⁺ imaging of isolated corneas was performed on an upright epifluorescence microscope (Laborlux 12; Leitz, Wetzlar, Germany) equipped with a water-immersion objective (40 x; NA, 0.8; Zeiss, Jena, Germany). Tissues were excited by the output of a 100 W mercury arc lamp that passed through a 480 nm bandpass filter (30 nm bandwidth). Fluorescence emission was passed through a 535 nm bandpass filter (40 nm bandwidth) before capture by a cooled CCD camera (Retiga Exi, Q-Imaging, Burnaby, Canada). An example image of a typical CNT loaded with OGB-1 dextran is shown in Fig. 1A of the Supplemental Material. The CNTs lie in the epithelial layer of the cornea, as seen through brightfield microscopy (Fig. 1B of the Supplemental Material). Excitation exposure was controlled by an electromechanical shutter (Uniblitz; Vincent Associates, Rochester, NY, USA) gated by transistor–transistor logic (TTL) pulses. Images were acquired at 0.5 Hz with 4 x 4 binning. MetaMorph software (Universal Imaging, Dowingtown, PA, USA) was used for instrument control, image acquisition and analysis. Ca²⁺ imaging of isolated corneas loaded with the Ca²⁺ indicator fura-2 dextran (10 kDa; Molecular Probes) was performed on an inverted microscope (TE200, Nikon) equipped with a UV-transmitting objective (40 x, NA, 1.4; SuperFluor; Nikon). Fura-2 dextran fluorescence was alternately excited by 340 and 380 nm light from a monochromator (PolyChrome II; TILL Photonics, Gräfelfing, Germany), and fluorescence emission was passed through a 510 nm bandpass filter (full width at half maximum, 40 nm) before being captured with a cooled CCD camera (CoolSnap HQ; Roper Scientific, Tucson, AZ, USA). MetaFluor software was used for instrument control and data analysis.

Confocal imaging

Imaging of CNTs loaded with FM1-43 was performed on a Zeiss LSM 510 NLO microscope. Whole corneas were incubated in 5 µM FM1-43 (Molecular Probes) in Locke solution for 1 h at room temperature and then washed for 1 h in normal Locke solution. FM1-43 was excited at 488 nm (argon ion laser), and its fluorescence was filtered (500–550 nm bandpass) before quantification. Scans were at 512 pixel x 512 pixel resolution. Pixel dwell time was 1.6 µs, and intensity was digitized at 12 bit resolution. Z-stacks were analysed and compressed with MetaMorph software. Confocal microscopical imaging of nerve fibres labelled with tetramethylrhodamine dextran was performed on an inverted microscope with a 40 x water-immersion objective (Axiovert, LSM 410; Zeiss). Corneal sections were excited at 568 nm, and fluorescence emission was passed through a 590 nm long-pass filter before photometric detection.

Verification of drug access to the corneal epithelium

In studying sensory nerve terminals in intact corneas, we used the pharmacological agents cyclopiazonic acid (CPA), carbonyl cyanide 3-chloro-phenylhydrazone (CCCP) and KB-R7943, which, respectively, inhibit the sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA), mitochondria and the NCX. The corneal epithelium, within which CNTs reside and whose cells surround the CNTs intimately, poses a barrier to the penetration of some drugs (Maurice & Mishima, 1984). Therefore it was important to verify that the drugs we used could gain access to the corneal epithelium. Corneal epithelial cells were loaded with OGB-1 dextran by ‘biolistic’ delivery. Briefly, 25 mg tungsten particles (1 µm diameter) were washed three times in 100% ethanol and spread out on a glass microscope slide to dry. OGB-1 dextran (5 mg) dissolved in 250 µl distilled water was placed on the tungsten particles and allowed to dry. Dye-coated particles were scraped off the glass slide, suspended in 1 ml 100% ethanol, and sonicated for 1 min. The particles were separated from ethanol by centrifugation and resuspended in 1 ml 100% ethanol. This suspension was stored at –20°C until use. Tungsten particles were delivered to the cornea with a custom-fabricated gene gun modified from the particle inflow gun (PIG) designed by Finer et al. (1992). The stock tungsten particle suspension was diluted 1 : 10 with 100% ethanol. The diluted suspension (1 µl) was shot from the gun onto the whole cornea, on the intact eyeball. After loading, the corneas were allowed to recover for 30 min prior to recording. Corneal epithelial cells loaded with OGB-1 dextran were monitored by Ca²⁺ imaging as 5 µM CPA, 2 µM CCCP or 5 µM KB-R7943 were bath applied to the cornea. In all cases, drug application evoked a Ca²⁺ response in 3 min (Fig. 2A–C in the Supplemental Material). These tests confirm that the drugs used in this study readily permeate the corneal epithelium – well within the time frame of drug incubation used in the experiments presented below.

Data analysis

Ca²⁺ indicator measurements for isolated corneas are reported as the fractional change in fluorescence intensity relative to baseline fluorescence (F/F₀), which was determined as follows. Within a temporal sequence of fluorescence images, a region of interest (ROI) was drawn around a selected portion of each CNT to be analysed. The fluorescence signal from each terminal was calculated as the pixel-averaged intensity within each ROI. The precise boundary of each ROI enclosing a terminal was replicated and placed over an adjacent region of the image that was devoid of nerve processes. The pixel-averaged intensity from each ROI was taken as background and was subtracted from the fluorescence signal obtained for the corresponding nerve terminal to yield the background-corrected fluorescence signal (F) from that terminal. In electrical stimulation experiments, sequences of images were acquired over extended time periods. In these experiments, we typically observed a slight downward drift in baseline, which was principally attributable to photo-bleaching of the indicator. In such cases, the drift was always well fitted by a low-amplitude single-exponential decay. The fitted baseline value (F₀) at every time point was then used to calculate F/F₀. The decay of Ca²⁺ transients were determined from F/F₀ traces that were smoothed by three-point adjacent averaging. F/F₀ values are reported as mean ± S.E.M. For CNTs loaded with fura-2 dextran, [Ca²⁺]_i was derived using the ratio method of Grynkiewicz et al. (1985). All fura-2 fluorescence records were corrected for background fluorescence by subtracting the light intensity measured in an adjacent region as described above. [Ca²⁺]_i was calculated using the equation of Grynkiewicz et al. (1985):

(1)

where R = F₃₄₀/F₃₈₀, with F₃₄₀ and F₃₈₀ being fura-2 emission intensities measured with 340 and 380 nm excitation, respectively; R_min and R_max are the values of R at zero and saturating Ca²⁺ concentrations, respectively; and S_f2/S_b2 is the ratio of emission intensities for the Ca²⁺-free and Ca²⁺-bound indicator when excited at 380 nm. The K_d used for fura-2 dextran was 240 nM; R_min, R_max and S_f2/S_b2 were determined in vitro. Ca²⁺ transients were smoothed by three-point adjacent averaging prior to analysis. [Ca²⁺]_i values are reported as mean ± S.E.M. In all cases, n refers to the number of CNTs recorded; one or two CNTs were recorded from each corneal section. Statistical significance was determined by paired t test; P < 0.05 was considered significant. For comparison of groups of CNTs, statistical significance was determined by unpaired t test; P < 0.05 was considered significant.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Role of ER in Ca²⁺ homeostasis

Previously, we have reported that action potential (AP)-evoked Ca²⁺ influx into CNTs occurs primarily through L-type VDCCs (Gover et al. 2003). We performed a series of experiments to determine whether intracellular Ca²⁺ stores have a functional role in either releasing Ca²⁺ into, or removing Ca²⁺ from, the cytoplasm subsequent to Ca²⁺ influx through the PM. Electrical field stimulation (50 stimuli at 10 Hz) applied to the isolated cornea reliably evoked Ca²⁺ transients in individual CNTs. In any particular CNT, the Ca²⁺ transients were reproducible in both amplitude and time course of decay, as demonstrated by the representative experiment shown in Fig. 2A, where two Ca²⁺ transients evoked by identical stimuli 10 min apart were essentially identical. Similar results were obtained for a total of seven CNTs studied: F/F₀ of the first electrically evoked Ca²⁺ was 0.36 ± 0.18, while the second transient evoked 10–20 min later showed peak F/F₀ of 0.36 ± 0.15 (P = 0.973). The time for the Ca²⁺ transient to decay from 95% to 10% of peak amplitude (t_95–10) was 43.9 ± 5.7 and 31.7 ± 3.8 s for the first and second stimulus, respectively (n = 7, P = 0.136). To determine the absolute change in [Ca²⁺]_i in CNTs during AP activity, we evoked Ca²⁺ transients electrically in CNTs loaded with the ratiometric Ca²⁺ indicator, fura-2 dextran. We observed that electrically evoked Ca²⁺ transients had a peak Ca²⁺ rise of 442 ± 114 nM from a resting [Ca²⁺]_i of 92 ± 12 nM (n = 7; 50 field stimuli at 10 Hz; Fig. 2A inset). The high reproducibility of the evoked Ca²⁺ transients in individual CNTs enabled us to use the same nerve terminals as their own controls in subsequent experiments where Ca²⁺ transporter antagonists were used.

View larger version (15K): [in this window] [in a new window]   Figure 2.  Reproducibility of Ca2+ transients evoked in corneal nerve terminals (CNTs) by electrical field stimulation and the unimportance of intracellular Ca2+ stores in Ca2+ regulation A, two Ca2+ transients evoked 10 min apart by electric field stimulation (50 stimuli at 10 Hz) in an individual sensory nerve terminal in an isolated cornea. The first (blue) and second (red) Ca2+ transients were essentially super-imposable. In seven CNTs studied, the two transients did not differ in amplitude (P = 0.973) or decay time (P = 0.136). Arrow marks time of stimulation. Inset, Ca2+ transients evoked by electric field stimulation (50 stimuli at 10 Hz) in individual CNTs loaded with the Ca2+ indicator fura-2 dextran (average data from n = 7 CNTs). Scale bars represent 100 nM [Ca2+] (vertical) and 10 s (horizontal). B, Ca2+ transient evoked in CNTs by electric field stimulation (50 stimuli at 10 Hz; blue trace, average data from nine CNTs). After incubation with 10 µM CPA for 10 min, identical electrical stimulation evoked a Ca2+ transient that was similar in peak amplitude (P = 0.571) and 95%–10% decay time (P = 0.294) (red trace, average data from n = 9 CNTs).

The activity of enzymes controlling intracellular Ca²⁺ stores is expected to be highly temperature-dependent (Dode et al. 2001). Therefore, conducting our experiments at room temperature may have minimized the contribution of SERCA to Ca²⁺ regulation and, consequently, may have contributed to negative results with SERCA blockers. To determine whether the ineffectiveness of SERCA blockers to modify CNT Ca²⁺ transients was due to lower-than-physiological temperatures, we repeated the SERCA experiments with CPA at a more physiological temperature. We observed that Ca²⁺ transient decay times were significantly faster when CNTs were excited at 35–37°C rather than at room temperature (P < 0.001, n = 40 and 12 for 22–24°C and 35–37°C, respectively; Fig. 4). Ca²⁺ transient peak amplitudes were not significantly different at room temperature (22–24°C) or physiological temperature (35–37°C) (P = 0.183). Although the Ca²⁺ transients evoked at 35–37°C were significantly faster, our CPA experiments performed at 35°C mirrored those done at room temperature: electrically evoked Ca²⁺ transients peaked at F/F₀ of 0.53 ± 0.14 and 0.46 ± 0.11 before and after incubation with 5 µM CPA for 15 min, respectively (n = 3, P = 0.124). The corresponding t_95–10 values were 14.6 ± 5.5 and 11.9 ± 2.5 s (n = 3, P = 0.614). The above results demonstrate that in CNTs, ER Ca²⁺ stores contribute neither to the rise nor to the fall of the Ca²⁺ transient.

View larger version (14K): [in this window] [in a new window]   Figure 4.  Ca2+ transient decay times but not peak Ca2+ transient amplitudes are dependent on temperature and glycolysis Peak amplitudes and 95%–10% decay times of Ca2+ transients were plotted for individual CNTs excited by electrical field stimulation (50 stimuli at 10 Hz). Each point represents data from an individual CNT excited either at room temperature (22–24°C, blue) or at physiological temperature (35–37°C, red and green). There was no correlation between peak amplitudes and 95%–10% decay times (R2 = 0.00027). Average 95%–10% decay times were significantly different between experiments performed at 22–24°C and 35–37°C (P < 0.001, n = 40 and 12, for 22–24°C and 35–37°C, respectively). For experiments with iodoacetate, the temperature was 36°C; pre- and post-drug data are shown by green and black points, respectively, and pre- and post-drug data from the same CNT are connected by a thin line. Solid black line is the result of linear regression performed on pre-drug data points acquired at room temperature.

Another organelle that could contribute to Ca²⁺ homeostasis is the mitochondrion (Werth & Thayer, 1994). To determine the contribution of mitochondria to Ca²⁺ regulation, we used the protonophore CCCP. The ability of mitochondria to sequester Ca²⁺ depends on the normally high electrical potential across the mitochondrial membrane (inside negative). Disruption of this large membrane potential with protonophores leads to release of Ca²⁺ from the mitochondrial stores and disables Ca²⁺ uptake from the cytosol by the mitochondria (Werth & Thayer, 1994). In CNTs at room temperature, before and after a 20 min incubation with 2 µM CCCP, the amplitudes of electrically evoked Ca²⁺ transients were not significantly different at F/F₀ of 0.32 ± 0.12 and 0.31 ± 0.089, respectively (n = 5, P = 0.985; Fig. 3A). Similarly, the Ca²⁺ transient decay times were also not affected: t_95–10 of 50.8 ± 4.6 and 48.7 ± 0.7 s before and after CCCP treatment, respectively (n = 5, P = 0.75).

View larger version (12K): [in this window] [in a new window]   Figure 3.  Glycolysis not mitochondrial respiration is critical for Ca2+ clearance in corneal nerve terminals (CNTs) A, Ca2+ transient evoked in CNTs by electric field stimulation (50 stimuli at 10 Hz; blue trace, average data from n = 5 CNTs). After incubation with 2 µM CCCP for 20 min, electrical stimulation evoked a Ca2+ transient that was similar to the first transient in amplitude (P = 0.985) and 95%–10% decay time (P = 0.75) (red trace, average data from n = 5 CNTs). B, Ca2+ transients evoked in CNTs by electric field stimulation (50 stimuli at 10 Hz) before (blue) and after (red) 20 min incubation with 1 mM ouabain (traces are average data from n = 5 CNTs). Inhibition of the Na+–K+-ATPase by ouabain reduced the Ca2+ transient amplitude by 82 ± 1.2% (P = 0.042, n = 5). C, Ca2+ transients evoked in CNTs by electric field stimulation (50 stimuli at 10 Hz) before (blue) and after (red) 20 min incubation with 2 mM iodoacetate (IAA) (traces are average data from n = 5 CNTs). Inhibition of glycolysis by IAA reduced the Ca2+ transient amplitude by 82 ± 4.3% (P = 0.003, n = 5) and slowed its decay 3.5-fold (P = 0.035). All experiments were done at 35–37°C.

Dissipation of the mitochondrial membrane potential by protonophores leads to rapid depletion of the ATP reserve in neurons (Budd & Nicholls, 1996). Vital transporters such as the Na⁺–K⁺-ATPase and the PMCA should be severely blocked by ATP depletion, and yet elimination of mitochondria as a source of ATP had no effect on the Ca²⁺ transients in CNTs. Direct blockade of these transporters, however, did severely affect the normal functioning of CNTs. Incubation with ouabain, a Na⁺–K⁺-ATPase inhibitor, strongly attenuated the electrical-evoked Ca²⁺ transient in five out of six CNTs tested. The amplitudes of the electrically evoked Ca²⁺ transients before and after a 20 min incubation with 1 mM ouabain were significantly different at F/F₀ of 0.67 ± 0.25 and 0.075 ± 0.043, respectively (n = 5, P = 0.042; Fig. 3B). In one CNT, 1 mM ouabain appeared to have no affect. Next, we investigated whether glycolysis could be a major source of ATP in CNTs. It is remarkable that a 20 min incubation at 36°C with 2 mM iodoacetate, an inhibitor of glycolysis, reduced the electrically evoked Ca²⁺ transient in CNTs by 82 ± 4% (n = 5; Fig. 3C). Indeed, in one CNT, electrical stimulation evoked no Ca²⁺ response at all. The Ca²⁺ transient amplitudes (F/F₀) before and after iodoacetate incubation were 0.775 ± 0.098 and = 0.134 ± 0.027, respectively (n = 5, P = 0.003). The control Ca²⁺ transients decayed with t_95–10 = 9.7 ± 2.1 s (n = 4), whereas the small Ca²⁺ transients evoked in the presence of iodoacetate decayed much more slowly (t_95–10, 32.5 ± 6.7 s; n = 4, P = 0.035). The above observations suggest that inhibition of glycolysis had at least two related manifestations. First, Na⁺–K⁺-ATPase activity was attenuated, leading to reduced Na⁺ and K⁺ electrochemical gradients, which could not support AP generation under electric field stimulation. This, in turn, limited activation of L-type VDCCs, which are the principal mediators of electrically evoked Ca²⁺ transients in CNTs (Gover et al. 2003). Second, the PM Ca²⁺- ATPase activity was also attenuated, so that basal [Ca²⁺]_i could not be restored efficiently even after a small [Ca²⁺]_i rise. Below, we further demonstrate the importance of the PMCA in CNTs. It could be the case that the differences in the time course of the Ca²⁺ transient decay before and after incubation with iodoacetate merely reflect a dependence of Ca²⁺ clearance kinetics on [Ca²⁺]_i (i.e. larger Ca²⁺ transients are cleared faster). This proved not to be the case: we observed no correlation between the peak amplitude and decay time of the Ca²⁺ transients (R² = 0.00027; Fig. 4).

Role of the NCX in Ca²⁺ efflux

We next performed a series of experiments to determine how Ca²⁺ is extruded from the cytosol across the PM. Two mechanisms for transporting Ca²⁺ across the PM were explored: the NCX (Verdru et al. 1997) and the PMCA (Benham et al. 1992). To determine the role of the NCX in Ca²⁺ clearance during an electrically evoked Ca²⁺ transient, we used the isothiourea derivative KB-R7943, which selectively inhibits NCX with an IC₅₀ of 1 µM under physiological conditions (Kimura et al. 1999). Treatment with KB-R7943 did not significantly affect electrically evoked Ca²⁺ transients in CNTs: Ca²⁺ transient amplitudes (F/F₀) 0.516 ± 0.098 and 0.512 ± 0.09, respectively, before and after a 10 min incubation with 10 µM KB-R7943 (n = 6, P = 0.969; Fig. 5A). The corresponding t_95–10 values were also similar, being 24.8 ± 8.3 and 24.4 ± 5.0 s, respectively (n = 6, P = 0.943). These observations suggest that in CNTs, the NCX is not a significant Ca²⁺ efflux mechanism during an electrically evoked Ca²⁺ transient.

View larger version (15K): [in this window] [in a new window]   Figure 5.  The plasma membrane Ca2+-ATPase (PMCA), but not the Na+–Ca2+ exchanger, plays a significant role in Ca2+ homeostasis in corneal nerve terminals (CNTs) A, Ca2+ transients evoked in CNTs by electrical field stimulation (50 stimuli at 10 Hz) before (blue) and after (red) 10 min incubation with 10 µM KB-R7943 (traces are average data from n = 6 CNTs). The Ca2+ transients had similar amplitudes (P = 969, n = 6) and 95%–10% decay times (P = 0.943). B, Ca2+ transients evoked in a CNT by electric field stimulation (50 stimuli at 10 Hz) before (blue) and after (red) 25 min exposure to alkaline extracellular pH (pHo 9.0) to block the PMCA. The Ca2+ transient amplitudes were similar (P = 0.688, n = 6); however, decay of the transients was much slower at pH 9.0 than at pH 7.3 (95%–10% decay times were 2-fold longer; P = 0.006).

Finally, we tested the possibility that the PMCA is the predominant mechanism for Ca²⁺ extrusion across the PM of CNTs. Although there are no selective PMCA antagonists at present, alkaline extracellular pH can inhibit the PMCA; the ATP-dependent Ca²⁺ pump removes Ca²⁺ from the cytoplasmic side of the PM and, in exchange, protons are counter-transported across the PM into the cytoplasm (Carafoli, 1992). Therefore, reducing the extracellular proton concentration attenuates PMCA activity. At physiological extracellular pH (pH_o 7.3), the amplitude of electrically evoked Ca²⁺ transients in CNTs was 0.82 ± 0.29 (n = 6), which was unchanged after exposure to pH_o = 9.0 for 25 min (0.84 ± 0.32; n = 6, P = 0.688). In remarkable contrast, the shift in pH_o from 7.3 to 9.0 increased t_95–10 2-fold, from 27.5 ± 4.8 to 56.0 ± 9.6 s (n = 6, P = 0.006; Fig. 5B). The effect of alkaline pH_o was reversed upon return to physiological pH_o (data not shown). These results indicate that the PMCA is the principal mechanism for Ca²⁺ removal from the cytoplasm of CNTs during a Ca²⁺ transient.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
The goal of this study was to understand Ca²⁺ homeostasis in peripheral nerve terminals of primary sensory neurons. We used a newly developed in vitro preparation, CNTs of the rat labelled with fluorescent Ca²⁺ indicators (Gover et al. 2003). In cell bodies and axons of primary sensory neurons, many diverse processes have been shown to control the clearance of Ca²⁺. These include distinct ATP-dependent Ca²⁺ pumps in the plasma, endoplasmic reticulum and mitochondrial membranes, NCX and Ca²⁺ buffering proteins (Thayer & Miller, 1990; Verkhratsky & Shmigol, 1996; Mayer et al. 1999). In order to determine which of these processes may support Ca²⁺ homeostasis in CNTs, we studied the effects of a variety of pharmacological reagents that interfere with different Ca²⁺ clearance mechanisms. The actions of these reagents were assessed by measuring the magnitude and time course of the Ca²⁺ transients evoked by electrical field stimulation of CNTs in vitro.

CICR in CNTs

We have demonstrated that the L-type VDCC is the major ion channel responsible for Ca²⁺ influx across the PM of CNTs during electrical activity (Gover et al. 2003). Application of SERCA inhibitors (CPA or DBHQ) did not significantly affect the amplitude or the kinetics of the Ca²⁺ transients in CNTs, suggesting that ER Ca²⁺ stores do not play a significant role in electrically evoked Ca²⁺ signalling in the peripheral nerve terminals of primary sensory neurons.

CICR from intracellular Ca²⁺ stores has been characterized in all primary afferent cell bodies studied to date, yet its physiological functions in these neurons are incompletely defined. One known function of CICR is to amplify AP-evoked Ca²⁺ transients to support a slowly developing and long-lasting (many seconds) after-spike hyperpolarization recorded in vagal afferent cell bodies (Moore et al. 1998). This slow afterpotential controls spike frequency accommodation and excitability of vagal afferent somata (Weinreich & Wonderlin, 1987). In contrast to primary sensory somata, Ca²⁺ transients recorded in CNTs were not modified by SERCA inhibitors such as DBHQ or CPA. Non-sensory peripheral nerve terminals such as motor nerve terminals, sympathetic nerve terminals and cholinergic nerve terminals have been reported to utilize CICR (Onodera, 1973; Smith & Cunnane, 1996; Narita et al. 2000). The difference in Ca²⁺ release mechanisms between afferent and efferent nerve terminals may be due to the different requirements for different rates of neurotransmitter release. Owing to their large surface to volume ratio, CNTs may not require CICR to substantially elevate [Ca²⁺]_i. We estimate that the surface to volume ratio of a 1 µm diameter CNT is 29 times greater than that of the average trigeminal ganglion cell body. This greater surface to volume ratio could enable CNTs to support robust Ca²⁺ transients following activation of VDCCs even without Ca²⁺ release from intracellular stores.

The role of mitochondria in Ca²⁺ homeostasis

In addition to SERCA, mitochondria can also have a significant effect on Ca²⁺ buffering in nerves by attenuating the amplitude and prolonging the duration of Ca²⁺ transients (Werth & Thayer, 1994; Shishkin et al. 2002). Ultrastructural studies have revealed prominent clusters of mitochondria throughout CNTs (Hoyes & Barber, 1976; Beckers et al. 1992). It was therefore surprising that manipulation of mitochondrial Ca²⁺ buffering had no effect on the Ca²⁺ transients in CNTs recorded at either room temperature (22–24°C) or physiological temperature (35–37°C). Lack of an observable effect of mitochondria on CNT Ca²⁺ transients may be attributable to several causes. The first is technical: mitochondrial Ca²⁺ uptake generally functions when [Ca²⁺]_i is severely elevated, such as following excessive nerve stimulation or pathological stimuli (Thayer et al. 2002). Thus, it was possible that the electrical stimulation protocols employed in the present work did not produce as large a Ca²⁺ load in CNTs. Direct measurements showed that electrically evoked Ca²⁺ transients had a peak Ca²⁺ concentration of 442 ± 114 nM from a resting [Ca²⁺]_i of 92 ± 12.0 nM. Werth & Thayer (1994) showed that mitochondria buffered Ca²⁺ in dorsal root ganglion somata when electrically evoked Ca²⁺ loads reached > 450 nM. Therefore, we conclude that the observed lack of effect of mitochondria on Ca²⁺ transients is unlikely to be attributed to our electrically evoked Ca²⁺ transients having been too small. The reasons for the apparently minor role of mitochondria observed in CNTs may be biological. First, mitochondria in CNTs are not homogeneously distributed; there is preferential localization of mitochondria in CNTs at the periphery of the cornea, whereas CNTs closer to the centre of the cornea appear to have fewer mitochondria (Hoyes & Barber, 1976). Second, it is possible that the CNT mitochondrial Ca²⁺ stores are already nearly full prior to electrical stimulation and therefore are unable to take up more Ca²⁺ during nerve stimulation. Finally, because the surface to volume ratio in CNTs is so large, PM-associated Ca²⁺ efflux mechanisms may predominate over intracellular Ca²⁺-sequestering mechanisms.

The role of glycolysis in Ca²⁺ homeostasis

The likely depletion of ATP after dissipation of the mitochondrial membrane potential by protonophores (Budd & Nicholls, 1996) appeared to have little effect on electrically evoked Ca²⁺ transients in CNTs. We have demonstrated that multiple ATP-driven pumps are active and necessary for normal CNT functioning. These pumps include the ouabain-sensitive Na⁺–K⁺-ATPase and the PMCA. We have also demonstrated that glycolysis is required for normal CNT function. After incubation with the glycolysis inhibitor, iodoacetate, the electrically evoked Ca²⁺ transients were significantly smaller and decayed back to baseline significantly more slowly. There was a modest increase in resting [Ca²⁺]_i in CNTs upon application of iodoacetate (F/F₀ = 0.39 ± 0.019, n = 5). The increase in resting [Ca²⁺]_i indicates that glycolysis is necessary for Ca²⁺ homeostasis in CNTs even under resting conditions. Because the resting [Ca²⁺]_i only increased 40%, the observed attenuation of the electrically evoked Ca²⁺ transients is physiological and not the result of indicator saturation. For reasonable resting [Ca²⁺]_i values of 50, 75 and 100 nM, the maximum values of F/F₀ for OGB-1 dextran are 3.6, 2.6 and 2.1, respectively (Gover et al. 2003). The rise in F/F₀ baseline in response to iodoacetate was always much smaller than these predicted maxima, suggesting that our measurements were unlikely to have been distorted by indicator saturation. Our findings suggest that inhibition of glycolysis leads to reduced Na⁺ and K⁺ electrochemical gradients (as a result of reduced Na⁺–K⁺-ATPase activity) and reduced PMCA activity, so that basal [Ca²⁺]_i could not be restored efficiently even after a small [Ca²⁺]_i rise. The small rise in [Ca²⁺]_i observed in the presence of iodoacetate may result from direct activation of VDCCs. We have observed that after incubation with the Na⁺ channel blocker, lidocaine, field stimulation can produce small Ca²⁺ transients in some nerve terminals (data not shown). We therefore conclude that our field stimulation protocol may depolarize some CNTs sufficiently to activate VDCCs directly. Given that CNTs derive their energy predominantly from glycolysis, the question arises as to the role of mitochondria in CNT function. One possibility is that mitochondria in CNTs serve an important role following nerve terminal injury.

Role of plasmalemmal NCX

Most of the Ca²⁺ in the cytoplasm of CNTs during a Ca²⁺ transient must leave through the PM. In sensory somata or peripheral axons, blocking the NCX had little, if any, effect on Ca²⁺ transients (Thayer & Miller, 1990; Wächtler et al. 1998; see however, Verdru et al. 1997). In this study, selective blockade of NCX with KB-R7943 did not alter Ca²⁺ transients in CNTs. Lack of NCX activity in the CNTs is surprising, because the NCX seems well suited for Ca²⁺ removal in a structure with a very large surface to volume ratio and therefore presumably high Ca²⁺ turnover. Indeed, immunostaining has revealed NCX localization at neuromuscular junctions, and at synapses in spinal and brain tissue (Luther et al. 1992). Functionally, the NCX does not contribute to the decay of Ca²⁺ transients in vagal axons, which also presumably have large surface to volume ratio (Wächtler et al. 1998).

PM Ca²⁺ pumps

Removal of Ca²⁺ across the PM of sensory somata appears to be dominated by the PMCA (Usachev et al. 2002). Unfortunately, there are currently no selective pharmacological inhibitors of the PMCA. In an ATP-dependent process, for each Ca²⁺ ion extruded, the PMCA counter-transports two protons (H⁺) into the cytosol. Thus, removal of extracellular H⁺ effectively inhibits PMCA-mediated Ca²⁺ efflux (Carafoli, 1992). In CNTs, electrically evoked Ca²⁺ transients decayed much more slowly in the presence of alkaline extracellular saline. This effect was reversed upon return to physiological pH (data not shown).

The apparent simplicity of CNTs with respect to Ca²⁺ homeostasis may be understandable in view of the large surface-to-volume ratio of the nerve terminals. The large surface-to-volume ratio implies that AP-evoked Ca²⁺ influx can elevate [Ca²⁺]_i much more effectively in CNTs than in sensory somata. Therefore, Ca²⁺ amplification through a mechanism such as CICR would not be necessary in CNTs. Similarly, the large surface to volume ratio also means that efflux mechanisms would remove Ca²⁺ from the cytosol more efficiently in CNTs than in sensory cell bodies. For these reasons, it is perhaps not surprising that the geometrical differences between the cell bodies and the nerve terminals translate into mechanistic differences in ion transport.

Collectively, our studies on CNTs reveal that the major pathway for Ca²⁺ influx following an AP is via L-type VDCCs (Gover et al. 2003), and that the PMCA is the predominant Ca²⁺ clearance mechanism in these nerve terminals. It is well established that the cornea is innervated by both nociceptive C- and A-fibre types and most of these fibres are sensitive to capsaicin (MacIver & Tanelian, 1993; Gover et al. 2003). It is therefore noteworthy, that our studies demonstrate homogenous Ca²⁺ regulatory mechanisms across fibre types. It will be interesting to determine how these Ca²⁺ regulatory systems support known functions of CNTs; namely, nociception and secretion of neurotrophic factors.

	References

Top Abstract Introduction Methods Results Discussion References

 
Beckers HJ, Klooster J, Vrensen GF & Lamers WP (1992). Ultrastructural identification of trigeminal nerve endings in the rat cornea and iris. Invest Ophthalmol Vis Sci 33, 1979–1986.[Abstract/Free Full Text]

Benham CD, Evans ML & McBain CJ (1992). Ca²⁺ efflux mechanisms following depolarization evoked calcium transients in cultured rat sensory neurones. J Physiol 455, 567–583.[Abstract/Free Full Text]

Budd SL & Nicholls DG (1996). A reevaluation of the role of mitochondria in neuronal Ca²⁺ homeostasis. J Neurochem 66, 403–411.[Medline]

Carafoli E (1992). The Ca²⁺ pump of the plasma membrane. J Biol Chem 267, 2115–2118.[Free Full Text]

David G & Barrett EF (2000). Stimulation-evoked increases in cytosolic [Ca²⁺] in mouse motor nerve terminals are limited by mitochondrial uptake and are temperature-dependent. J Neurosci 20, 7290–7296.[Abstract/Free Full Text]

Dode L, Van Baelen K, Wuytack F & Dean WL (2001). Low temperature molecular adaptation of the skeletal muscle sarco (endo) plasmic reticulum Ca²⁺-ATPase 1 (SERCA 1) in the wood frog (Rana sylvatica). J Biol Chem 276, 3911–3919.[Abstract/Free Full Text]

Finer JJ, Vain P, Jones MW & McMullen MD (1992). Development of the particle inflow gun to plant cells. Plant Cell Rep 11, 232–238.

Fox AP, Nowycky MC & Tsien RW (1987). Single-channel recordings of three types of calcium channels in chick sensory neurones. J Physiol 394, 173–200.[Abstract/Free Full Text]

Gover TD, Kao JPY & Weinreich D (2003). Calcium signaling in single peripheral sensory nerve terminals. J Neurosci 23, 4793–4797.[Abstract/Free Full Text]

Grynkiewicz G, Poenie M & Tsien RY (1985). A new generation of Ca²⁺ indicators with greatly improved fluorescence properties. J Biol Chem 260, 3440–3450.[Abstract/Free Full Text]

Hoyes AD & Barber P (1976). Ultrastructure of the corneal nerves in the rat. Cell Tissue Res 172, 133–144.[CrossRef][Medline]

Inesi G & Sagara Y (1994). Specific inhibitors of intracellular Ca²⁺ transport ATPases. J Membr Biol 141, 1–6.[Medline]

Kim HC & Chung MK (1999). Voltage-dependent sodium and calcium currents in acutely isolated adult rat trigeminal root ganglion neurons. J Neurophysiol 81, 1123–1134.[Abstract/Free Full Text]

Kimura J, Watano T, Kawahara M, Sakai E & Yatabe J (1999). Direction-independent block of bi-directional Na⁺/Ca²⁺ exchange current by KB-R7943 in guinea-pig cardiac myocytes. Br J Pharmacol 128, 969–974.[CrossRef][Medline]

Lele PP & Weddell G (1956). The relationship between neurohistology and corneal sensibility. Brain 79, 119–154.[Free Full Text]

Luther PW, Yip RK, Bloch RJ, Ambesi A, Lindenmayer GE & Blaustein MP (1992). Presynaptic localization of sodium/calcium exchangers in neuromuscular preparations. J Neurosci 12, 4898–4904.[Abstract]

Lytton J, Westlin M, Burk SE, Shull GE & MacLennan DH (1992). Functional comparisons between isoforms of the sarcoplasmic or endoplasmic reticulum family of calcium pumps. J Biol Chem 267, 14483–14489.[Abstract/Free Full Text]

MacIver MB & Tanelian DL (1993). Structural and functional specialization of A and C fiber free nerve endings innervating rabbit corneal epithelium. J Neurosci 13, 4511–4524.[Abstract]

Maurice DM & Mishima S (1984). Ocular pharmacokinetics. In Pharmacology of the Eye. Handbook of Experimental Pharmacology, ed. Sears ML, pp. 19–116. Springer, New York.

Mayer C, Quasthoff S & Grafe P (1999). Confocal imaging reveals activity-dependent intracellular Ca²⁺ transients in nociceptive human C fibres. Pain 81, 317–322.[CrossRef][Medline]

Moore KA, Cohen AS, Kao JP & Weinreich D (1998). Ca²⁺-induced Ca²⁺ release mediates a slow post-spike hyperpolarization in rabbit vagal afferent neurons. J Neurophysiol 79, 688–694.[Abstract/Free Full Text]

Muller LJ, Pels L & Vrensen GF (1996). Ultrastructural organisation of human corneal nerves. Invest Opthalmol Vis Sci 37, 476–488.[Abstract/Free Full Text]

Narita K, Akita T, Hachisuka J, Huang S, Ochi K & Kuba K (2000). Functional coupling of Ca²⁺ channels to ryanodine receptors at presynaptic terminals. Amplification of exocytosis and plasticity. J Gen Physiol 115, 519–532.[Abstract/Free Full Text]

Onodera K (1973). Effect of caffeine on the neuromuscular junction of the frog, and its relation to external calcium concentration. Jpn J Physiol 23, 587–597.[Medline]

Schild JH, Clark JW, Hay M, Mendelowitz D, Andresen MC & Kunze DL (1994). A- and C-type rat nodose sensory neurons: model interpretations of dynamic discharge characteristics. J Neurophysiol 71, 2338–2358.[Abstract/Free Full Text]

Shishkin V, Potapenko E, Kostyuk E, Girnyk O, Voitenko N & Kostyuk P (2002). Role of mitochondria in intracellular calcium signaling in primary and secondary sensory neurons of rats. Cell Calcium 32, 121–130.[CrossRef][Medline]

Smith AB & Cunnane TC (1996). Ryanodine-sensitive calcium stores involved in neurotransmitter release from sympathetic nerve terminals of the guinea-pig. J Physiol 497, 657–664.[Medline]

Thayer SA & Miller RJ (1990). Regulation of the intracellular free calcium concentration in single rat dorsal root ganglion neurons in vitro. J Physiol 425, 85–115.[Abstract/Free Full Text]

Thayer SA, Perney TM & Miller RJ (1988). Regulation of calcium homeostasis in sensory neurons by bradykinin. J Neurosci 11, 4089–4097.

Thayer SA, Usachev YM & Pottorf WJ (2002). Modulating Ca²⁺ clearance from neurons. Front Biosci 7, d1255–1279.[Medline]

Usachev YM, DeMarco SJ, Campbell C, Strehler EE & Thayer SA (2002). Bradykinin and ATP accelerates Ca²⁺ efflux from rat sensory neurons via protein kinase C and the plasma membrane Ca²⁺ pump isoform 4. Neuron 33, 113–122.[CrossRef][Medline]

Verdru P, De Greef C, Mertens L, Carmeliet E & Callewaert G (1997). Na⁺-Ca²⁺ exchange in rat dorsal root ganglion neurons. J Neurophysiol 77, 484–490.[Abstract/Free Full Text]

Verkhratsky A & Shmigol A (1996). Calcium-induced calcium release in neurones. Cell Calcium 19, 1–14.[CrossRef][Medline]

Wächtler J, Mayer C & Grafe P (1998). Activity-dependent intracellular Ca²⁺ transients in unmyelinated nerve fibres of the isolated adult rat vagus nerve. Pflugers Arch 435, 678–686.[CrossRef][Medline]

Weinreich D & Wonderlin WF (1987). Inhibition of calcium-dependent spike after-hyperpolarization increases excitability of rabbit visceral sensory neurones. J Physiol 394, 415–427.[Abstract/Free Full Text]

Werth JL & Thayer SA (1994). Mitochondria buffer physiological calcium loads in cultured rat dorsal root ganglion neurons. J Neurosci 14, 348–356.[Abstract]

Whitear EC (1960). An electron microscope study of the cornea in mice, with special reference to the innervation. J Anat 94, 387–409.[Medline]

Zander E & Weddell G (1951). Observations of the innervation of the cornea. J Anat 85, 68–69.[Medline]

	Acknowledgements

 
We thank Dr Michael Gold for helpful discussions and for reviewing a previous version of this manuscript. This work was supported by grants from NIH (NS-22069 to D.W. and GM-56481 to J.P.Y.K.) and Conselho Nacional para o Progresso da Ciência (to T.H.V.M.).

This article has been cited by other articles:

				Y. V. Medvedeva, M.-S. Kim, and Y. M. Usachev Mechanisms of Prolonged Presynaptic Ca2+ Signaling and Glutamate Release Induced by TRPV1 Activation in Rat Sensory Neurons J. Neurosci., May 14, 2008; 28(20): 5295 - 5311. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental data All Versions of this Article: 578/2/481    most recent jphysiol.2006.119008v1 Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gover, T. D. Articles by Weinreich, D. Search for Related Content PubMed PubMed Citation Articles by Gover, T. D. Articles by Weinreich, D. Related Collections Neuroscience