Effect of nitric oxide donors and noradrenaline on Ca²⁺ release sites and global intracellular Ca²⁺ in myocytes from guinea-pig small mesenteric arteries

Abstract

In smooth muscle the spontaneous Ca²⁺ release from the sarcoplasmic reticulum (SR) occurs at preferred locations called frequent discharge sites (FDSs) giving rise to localized intracellular Ca²⁺ transients (Ca²⁺ sparks). Laser scanning confocal microscopy of fluo-3-loaded single myocytes freshly isolated from small mesenteric arteries of guinea-pig was used to investigate the action of nitric oxide (NO) donors and noradrenaline on the position and activity of FDSs and on global intracellular Ca²⁺ concentration ([Ca²⁺]_i). In 8% of cells ‘microsparks’, Ca²⁺ release events smaller in duration, spread and amplitude than Ca²⁺ sparks were observed. The location of the initiation point of Ca²⁺ sparks observed during line-scan imaging was found to ‘jitter’ by ± 0.41 μm. However, the general position of an FDS within the cell did not change; most FDSs were close (within 1.2 ± 0.1 μm) to the cell membrane and often multiple FDSs occurred in one confocal plane of the cell. In the resting state, NO donors S-nitroso-N-acetylpenicillamine (SNAP; 50 μm) and sodium nitroprusside (SNP; 100 μm) did not change the general position of FDSs and slightly depressed their activity, but did not affect the global [Ca²⁺]_i significantly. Application of noradrenaline (1–10 μm) increased Ca²⁺ spark frequency at existing FDS(s) leading to a Ca²⁺ wave. The increase in FDS activity and in global [Ca²⁺]_i produced by noradrenaline were inhibited by the presence of SNAP or SNP but not by 8-bromoguanosine cyclic 3′,5′-monophosphate (8-Br-cGMP; 100 μm). In the presence of 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ), inhibitor of soluble guanylate cyclase, SNAP and SNP still exerted their effects on the noradrenaline response. These results suggest that SNAP and SNP inhibit the noradrenaline-evoked rise in global [Ca²⁺]_i by a cGMP-independent mechanism and that part of this effect is due to inhibition of the activity of FDSs; moreover, only the activity, but not the position, of FDSs is changed by either stimulant or inhibitory substances.

In smooth muscle cells calcium is released from the sarcoplasmic reticulum (SR) into the cytosol via calcium release events which occur either spontaneously or as a result of cell stimulation. The release occurs through the opening of either ryanodine receptor (RyRs; e.g. Nelson et al. 1995; Gordienko et al. 2001) or inositol trisphosphate receptor calcium channels (IP₃Rs; e.g. Iino, 1987; Boittin et al. 1999). Where several of these calcium release channels open more or less simultaneously, the release of SR calcium (in the presence of a suitable indicator dye) gives rise to a transient localized increase in fluorescence termed a calcium ‘spark’. Such releases occur at one or a few preferred sites in the cell, named frequent discharge sites (FDSs; Gordienko et al. 1998). It was demonstrated that the structural basis of FDSs is portions of the SR enriched in RyRs (Gordienko et al. 2001).

While in cardiac and skeletal muscle the physiological role of Ca²⁺ sparks is to promote contraction, they have been suggested to have a dual role in smooth muscle (Jaggar et al. 2000). On the one hand, Ca²⁺ sparks (usually from the dominant FDS) may precede and then develop into either a spontaneous (in guinea-pig ileal myocytes: Gordienko et al. 1999; in rabbit portal vein myocytes: Gordienko et al. 2001) or a stimulus-evoked Ca²⁺ wave (in guinea-pig mesenteric arteriole myocytes, this study), often giving rise to contraction. It is not yet clear why some Ca²⁺ sparks develop into Ca²⁺ waves and the others do not. On the other hand, it has been suggested that Ca²⁺ sparks have a role in opposing contraction through the activation of Ca²⁺-activated potassium (BK_Ca) channels, causing spontaneous transient outward currents (STOCs) and membrane hyperpolarization (Ganitkevich & Isenberg, 1990; Nelson et al. 1995) or in promoting contraction through opening of calcium-activated chloride channels (Cl_Ca) causing spontaneous transient inward currents (STICs) which depolarize the membrane (ZhuGe et al. 1998; Gordienko et al. 1999).

This study investigates whether vasorelaxant substances elicit Ca²⁺ sparks in different cellular location(s) to those affected by vasoconstrictive substances and thus, by increasing local intracellular Ca²⁺ concentration, might activate a different set of effector molecules within the myocyte. Nitric oxide donors, S-nitroso-N-acetylpenicillamine (SNAP), which is known to hyperpolarize the membrane and decrease [Ca²⁺]_i in rat mesenteric artery (Ghisdal et al. 2000) and sodium nitroprusside (SNP) which relaxes vascular smooth muscle and inhibits the noradrenaline-evoked contraction (Verhaeghe & Shepherd, 1976) were used. Noradrenaline (NA) which leads to an increase in [Ca²⁺]_i, Ca²⁺ waves and contraction of mesenteric artery (Ghisdal et al. 2000; Mauban et al. 2001) was used to stimulate α-adrenoceptors. The effects of SNAP and SNP on the position of FDSs within the cell and on the relative global intracellular Ca²⁺ concentration in vascular myocytes under resting conditions and during stimulation by noradrenaline were examined. It was found that the general positions of FDSs in single myocytes from guinea-pig mesenteric artery were not affected by either NO donors or by noradrenaline and that the number of Ca²⁺ spark discharges at these sites, rather then being increased as has been suggested, was decreased by NO donors and increased by noradrenaline. A preliminary account of this work was presented at the 45th Annual Meeting of the Biophysical Society (Pucovský et al. 2001).

METHODS

Cell preparation

Male Dunkin-Hartley guinea-pigs weighing 250–450 g were killed by cervical dislocation followed by exsanguination, in accordance with UK guidelines for humane killing of experimental animals (Schedule 1). The mesenterium was then taken out and small mesenteric arteries (2nd–4th order branches, diameter ∼100–250 μm) were cleaned of blood and surrounding tissue, cut out and incubated at 37 °C for 35–38 min in a Ca²⁺- and Mg²⁺-free physiological saline solution (PSS; for composition see below) containing collagenase, soybean trypsin inhibitor and bovine serum albumin (all at 1 mg ml⁻¹) and either elastase (0.01% w/v) or protease (0.5 mg ml⁻¹). The pieces of tissue were then taken out, rinsed twice with Ca²⁺- and Mg²⁺-free PSS and gently triturated with a wide-bore glass pipette to yield single cells. Aliquots of the suspension of single cells were then placed in experimental chambers and cells were left at 4 °C for 45–60 min to attach to glass coverslips forming the bottom of the experimental chamber. Fluo-3 AM, the membrane-permeable ester of the Ca²⁺-sensitive fluorescent dye (absorption maximum: 506 nm, emission maximum: 527 nm), was solubilized in dimethyl sulphoxide (DMSO) containing 0.025% (w/v) Pluronic F-127 by ultrasonication for 5 min to yield a stock solution of 4 mm and this solution was then added into PSS so that a final concentration of fluo-3 AM was 5 μm. The final concentration of DMSO was 0.125% (v/v). After the cells had settled on the coverslips, they were incubated in PSS containing fluo-3 AM and left to load the dye for 30 min at room temperature. They were then washed with PSS and left to allow de-esterification of the dye for another 30 min. The cells were stored at 4 °C in PSS (composition (mm): NaCl 120, KCl 6, glucose 12, Hepes 10, MgCl₂ 1.2, CaCl₂ 2.5, pH set to 7.4 with NaOH) and used within 12 h of isolation.

Confocal microscopy

The cells were imaged using a Zeiss LSM 510 laser scanning confocal microscope (Carl Zeiss, Jena, Germany). Experimental chambers containing cells were placed on the stage of an Axiovert 100M inverted microscope. The excitation beam of 488 nm was produced by an argon laser, attenuated to 0.2–0.5% intensity with an acousto-optical tunable filter and delivered to the specimen via a Zeiss Apochromat × 63 oil immersion objective (numerical aperture 1.4). The pinhole size was adjusted to provide a confocal optical section below 1.2 μm. Lateral resolution was established by scanning 0.2 μm fluorescent beads and was found to be 0.4 μm. Emitted fluorescence was captured at wavelengths >510 nm using LSM 510 software (release 2.01, Carl Zeiss, Jena, Germany) running on a Windows NT work station. During a time series protocol, X–Y images (typically 100–200 images per series; image depth of 8 bits) size 256 pixels × 256 pixels, were taken every 494 ms. In some cases, to improve temporal resolution, a single line of 256 pixels, usually positioned over a subplasmalemmal region of the cell parallel to the cell membrane, was scanned every 2.9 ms. Successive images of the scanned line were aligned parallel to each other from left to right to form a line-scan image in which the horizontal dimension reflects time and the vertical dimension shows position along the scan line. X-Y scanning could detect all the FDSs in the confocal plane of the cell but at the expense of temporal resolution. Line scanning could detect every Ca²⁺ spark along the scan line, but only one or few FDSs within the confocal plane could be monitored. Tested substances were applied either focally to the cell by a N₂-driven pressure ejection device using application pipettes (i.d. ∼3 μm) positioned within 80–100 μm of the target cell, or by changing the bathing solution. All experiments were done at room temperature (20–23 °C).

Analysis of data

Raw data were processed and analysed using custom-written routines in IDL 5.4 programming language (Research Systems, Inc., Boulder, CO, USA); statistical evaluation was done using MicroCal Origin software (MicroCal Software, Inc., Northampton, MA, USA) and final images were produced using CorelDraw 7.0 software (Corel Corporation, Ottawa, Ontario, Canada).

To reduce the scatter of points and partly minimize the effect of photomultiplier noise, the raw image data were first smoothed using 3 pixel-wide boxcar averaging and fluorescence values less than 5 intensity units (I.U.) were cut off. The intensity of the fluorescence in the images was then normalized (F/F₀) using a self-ratio approach as described elsewhere (Cheng et al. 1993).

To assess the position(s) of FDSs within the cell, a single image of the cell was constructed from a series of normalized images, in which the number of times the individual pixel values exceeded 1.5 F/F₀ was colour coded. A rise in fluorescence by 50% above the resting level was set as a threshold for detection of calcium ‘sparks’ and ‘microsparks’, as the the vast majority of transient calcium events in these cells had amplitudes above 1.5 F/F₀.

To obtain a single image which quantifies the changes in fluorescence (reflecting the changes in global intracellular Ca²⁺ concentration) the following procedure was used: first, the averages of the control series of images (no solutions added) and of the test series (when either the vehicle solution or test substance were added) were obtained. Then the test series average image was divided by the control series average image to yield an average ratio image. In this image the values of F/F₀ less than 1 (coded in blue) denote pixels where [Ca²⁺]_i has decreased and the values of F/F₀ higher than 1 (coded in brown to yellow) denote pixels where [Ca²⁺]_i has increased. In order to quantify these images, the average change in fluorescence emission per pixel over the entire confocal plane (referred to as ‘average pixel fluorescence’) was calculated, according to the formula:

where F and F′ are fluorescence emission in each pixel in control and test conditions, respectively, and m is the number of pixels in the confocal plane (expressed as%/pixel).

Where appropriate, data are expressed as mean values ± s.e.m. for the number of cells (n) analysed. Statistical significance was calculated using Student's t test for paired observations (except in experiments with 8-Br-cGMP (see below) under resting conditions, where Student's t test for unpaired observations was used) and the differences where P < 0.05 were considered significant.

Chemicals

Collagenase (type IA), trypsin inhibitor (type II-S), elastase (type I), bovine serum albumin, protease (type X), sodium nitroprusside (SNP), (−)-noradrenaline bitartarate salt (NA), N-acetyl-d-penicillamine (APA) and dimethyl sulphoxide (DMSO) were purchased from Sigma-Aldrich Co. Ltd (Poole, Dorset, UK). Fluo-3 acetoxymethyl ester (fluo-3 AM) and Pluronic F-127 were bought from Molecular Probes, Inc. (Leiden, The Netherlands); 8-bromoguanosine cyclic 3′,5′-monophosphate sodium salt (8-Br-cGMP) and 1H-[1,2,4]oxadiazolo [4,3-a]quinoxalin-1-one (ODQ) were from Tocris Cookson Ltd (Bristol, UK) and (±)-S-nitroso-N-acetylpenicillamine (SNAP) was from Calbiochem (Nottingham, UK).

SNAP and APA were dissolved in 10% DMSO. Fluo-3 AM and ODQ were dissolved in 100% DMSO to form stock solutions that were added to PSS. All the other substances were dissolved in deionized water. Solutions of SNAP, SNP and fluo-3 AM were prepared on the day of experiment immediately before use and were used within 7 h. At all times precautions were taken to reduce to a minimum the exposure of these solutions to light.

RESULTS

Spontaneous Ca²⁺ release events

Under resting conditions isolated vascular smooth muscle cells from guinea-pig small mesenteric artery showed spontaneous Ca²⁺ sparks (where fluorescence rose more than 50%) which originated either in one or in several locations in the cell (Fig. 1A; an animated sequence of this (video 1, Multiple FDSs) is presented as supplementary material at The Journal of Physiology online; http://www.jphysiol.org/cgi/content/full/539/1/25). Out of 144 cells, 126 cells (88%) showed spontaneous Ca²⁺ sparks. These occurred at FDSs located close (within 1.2 ± 0.1 μm, calculated from 18 FDSs in 11 cells) to the cellular membrane in 96% of cases (FDSs were apparently located centrally in the confocal plane in 6 cells out of 144). Thirty-five% of cells (51/144) had 1 FDS, 32% (46/144) had 2, 16% (23/144) had 3 and 4% (6/144) had 4 FDSs in the confocal plane. In some myocytes, spontaneous Ca²⁺ waves propagating through part (19/144 cells; Fig. 1B) or whole optical section of the cell (8/144 cells; Fig. 1C) were observed, as reported previously by Gordienko et al. (1998) in guinea-pig ileal myocytes. In most cases, Ca²⁺ waves started in cells with multiple FDSs, originating in one, usually the most active of the FDSs, and lasted up to ∼10 s. Only in 7 cases out of 27 did the Ca²⁺ wave start in a cell with a single FDS. In a few cases (3/144 cells) low amplitude slow [Ca²⁺]_i oscillations (peak F/F₀ less than 1.5 and period of 18.8–24.7 s) were observed (not shown).

Figure 1. Spontaneous Ca²⁺ events in isolated arterial smooth muscle cells.

Sections of longer image series recorded from cells under resting conditions (in PSS) are shown. A, spontaneous Ca²⁺ sparks arise at one or several positions within the confocal plane (a, b, c, d), close to the membrane. B, in 13% of cells spontaneous Ca²⁺ waves not involving the whole cell occurred (starting at ‘f’), originating from FDS (‘e’). C, in 6% of cells spontaneous generalized Ca²⁺ waves occurred (starting at ‘k’), also originating from one (i) of four FDSs (g, h, i, j). It is notable that mostly cells with multiple FDSs (74% of cases) developed spontaneous Ca²⁺ waves. D, in a group of connected cells the calcium wave spreads from the first cell (l) to the next (m, n) with a time lag between maxima of ∼7 s. White numbers in the frames designate the frame number within time series of images. Time difference between frames is 494 ms. Calibration (white horizontal bar in the first frame of each series): 10 μm. Normalized fluorescence intensity is coded as shown on the grey-scale bar at the bottom. The sequences under A and D are also published as supplementary material at The Journal of Physiology online (A), video 1, Multiple FDSs; B, video 2, Asynchronicity, see: http://www.jphysiol.org/cgi/content/full/539/1/25.

In cases when elastase rather than protease was used for dispersion of single cells, the incidence of groups of a few connected cells was higher. This enabled the spread of the Ca²⁺ signal from one cell to another to be observed. Figure 1D shows one such group where [Ca²⁺]_i has risen first in one cell and then, with a time lag of roughly 7 s between the maxima, spread to the next two cells (this sequence (video 2, Asynchronicity) is also available as supplementary material at The Journal of Physiology online; http://www.jphysiol.org/cgi/content/full/539/1/25). The transient nature of Ca²⁺ waves and their asynchronous occurrence in adjacent cells argue in favour of the view that vascular smooth muscle maintains its tension by asynchronous [Ca²⁺]_i oscillations rather than by sustained and simultaneous [Ca²⁺]_i increase (Iino et al. 1994).

Effects of NO donors and cGMP analogue under resting conditions

Application of either SNP (100 μm) or SNAP (50 μm) to the cell suppressed the activity of existing FDSs (Fig. 2Abvs.2Aa and Fig. 2Bbvs.2Ba for SNP and Fig. 2Cbvs.2Ca for SNAP) either completely (e.g. site b in Fig. 2Ba and Bb) or partially (e.g. site d in Fig. 2Ca and Cb), but did not visibly change their position. SNP (100 μm) produced a decrease in the average pixel fluorescence from 4.19 ± 2.18%/pixel (control, only PSS applied; e.g. Fig. 2Bc and Be) to −6.09 ± 3.04%/pixel (n = 6; e.g. Fig. 2Bd and Be) which was on the margin of statistical significance (P = 0.078, which is significant if a one-tailed t test is applied). SNAP (50 μm) only slightly decreased the average pixel fluorescence from 3.87 ± 2.53%/pixel (control, only 0.05% DMSO added; e.g. Fig. 2Cc and Ce) to 0.78 ± 2.71%/pixel (n = 7, P > 0.05; e.g. Fig. 2Cd and e) suggesting that global [Ca²⁺]_i did not change significantly. N-acetyl-d-penicillamine (APA; 50 μm), a product of SNAP decomposition (after SNAP releases NO), did not change [Ca²⁺]_i significantly, fluorescence being 149.1 ± 23.4% of the control in 0.5% DMSO (n = 5, P > 0.05; not shown). Thus there was no evidence for the suggestion (Nelson et al. 1995) that substances relax this smooth muscle by increasing spark discharge.

Figure 2. The effect of NO donors and 8-Br-cGMP on frequent discharge sites and resting intracellular Ca²⁺ in isolated arterial smooth muscle cells.

A, part of normalized series of images under control conditions (Aa) and after addition of 100 μm SNP (Ab). The first thirty images from each series are shown (white numbers in the frames), taken every 494 ms. White horizontal bar in the first frame: 10 μm. From such series as in A the images showing the number of events where fluorescence of individual pixels exceeded 1.5 F/F₀ (to detect FDSs; columns a and b) were constructed. Actions of: 100 μm SNP, B; 50 μm SNAP, C; 100 μm 8-Br-cGMP, D. Column a: control series; column b: after addition of test substance. Images in columns c (vehicle solution present) and d (test substance present) were constructed from raw fluorescence data by dividing the average image of the test series by the average control image before addition of any solution (see Methods). Pixels coded in blue denote decrease and the ones coded in brown to yellow denote increase in fluorescence. Graphs in column e: statistical evaluation of data represented in columns c (vehicle) and d (test substance)(means ±s.e.m.), where the bars to the left show averages of controls and the bars to the right the averages when the substance under test was added. The colour bar at the bottom left applies to columns a and b, and the one at the bottom right to c and d. Except for 8-Br-cGMP, all controls were on the same cell (n = 5 or more); none of the effects is significant (P > 0.05).

The membrane-permeable analogue of cGMP, 8-Br-cGMP, did not have a pronounced effect on FDSs, slightly depressing their activity in some cases (site e in Fig. 2Da and Db) and activating them in the others (e.g. site f in Fig. 2Db). The global [Ca²⁺]_i was not changed significantly by 8-Br-cGMP (from 0.94 ± 1.32%/pixel; n = 7, e.g. Fig. 2Dc, to 3.09 ± 1.10%/pixel in the presence of 8-Br-cGMP; n = 5, e.g. Fig. 2Dd; P > 0.05, evaluated using Student's t test for unpaired observations; Fig. 2De).

Two dimensional (X–Y) scanning of a cell's confocal plane could reveal the positions of FDSs in the plane, but due to low sampling frequency (every 494 ms) could not detect all the Ca²⁺ sparks occuring at the sites and potential fine changes of the position of FDSs. For this reason the most active FDSs within the confocal planes were imaged in line-scanning mode. Line-scan images revealed the actual number of Ca²⁺ sparks discharged from a selected FDS (Fig. 3) and showed that the general position of FDSs was not changed. The position of the initiating point of a Ca²⁺ spark in the FDS was found to ‘jitter’ by 0.41 ± 0.07 μm (n = 7 line-scans, 145 Ca²⁺ sparks analysed; e.g. Fig. 3B and C) which suggests that each such site consists of more than one Ca²⁺ release channel (Parker et al. 1996; Gonzalez et al. 2000; Gordienko et al. 2001). In 6 out of a total of 71 line-scan images (8% of cases) microsparks (Ca²⁺ release events smaller in amplitude and duration than ‘regular’ Ca²⁺ sparks; Gordienko et al. 1998) were observed (e.g. Fig. 3C). A ‘microspark’ typically lasted 74 ± 7 ms and had a width at half-maximum (FWHM) of 1.7 ± 0.1 μm. Calcium ‘sparks’ lasted in average 222 ± 26 ms and had a FWHM of 3.0 ± 0.3 μm (based on analysis of 25 microsparks and 25 sparks in 6 different cells). The average amplitude of microsparks was ∼65% of average spark amplitude.

Figure 3. The effect of NO donors and 8-Br-cGMP on position and activity of single FDSs in isolated arterial smooth muscle cells under resting conditions.

Line-scan images were formed by aligning from left to right the successive fluorescent images of the scan line captured every 2.9 ms. Aa, Ba and Ca are controls; Ab, Bb and Cb were made in the presence of test substances: A, 100 μm SNP; B, 50 μm SNAP; C, 100 μm 8-Br-cGMP. D, statistical evaluation of Ca²⁺ spark frequencies; means ±s.e.m. (n = 5 or more). Note presence of microsparks in C (those in a denoted by *) and the ‘jittering’ of the initiation point of Ca²⁺ spark (-○—○- in B and C).

Line-scan imaging revealed that none of the tested substances changed the number of Ca²⁺ spark discharges from a single FDS significantly, although there was a slight decrease detected (Fig. 3D): 100 μm SNP: 93.62 ± 13.59% of control (n = 5, P > 0.05, Fig. 3A), 50 μm SNAP: 82.49 ± 8.48% (n = 6, P > 0.05, Fig. 3B), 100 μm 8-Br-cGMP: 92.36 ± 13.48% (n = 5, P > 0.05, Fig. 3C). Line scans were made over the most active FDS and this result implies that less frequently discharging FDSs were inhibited (see above and Fig. 2) whereas the most frequently discharging FDS was not significantly affected.

A result that supports the view that ‘microsparks’ are the building blocks of ‘normal’ Ca²⁺ sparks and that their existence might be the cause of ‘jitter’ is shown in Fig. 4. There a portion of the line-scan image from Fig. 3Ca is shown on an expanded time scale containing both ‘microsparks’ and Ca²⁺ sparks. ‘Microsparks’ in this figure arise from several different positions along the scan line and the initiation point of Ca²⁺ sparks shows ‘jitter’. A closer examination of portions of this line-scan image on a further expanded time scale shows that the ‘jitter’ of Ca²⁺ sparks is caused by a different sequence of recruitment of ‘microsparks’ into a Ca²⁺ spark (Fig. 4B–D; position 1 denoted by a continuous, and position 3, denoted by a dashed horizontal line). The example in Fig. 4 shows a Ca²⁺ spark starting as a ‘microspark’ at position 1 and recruiting a ‘microspark’ at positions 2 and 3 (Fig. 4B) thus creating a ‘normal’ Ca²⁺ spark, or vice versa (Fig. 4D; position 2 is included in the Ca²⁺ spark because it lies between 1 and 3), but other sequences of recruitment are also possible (Fig. 4C). The criterion for a single FDS was that it had to discharge Ca²⁺ spark as a single unit, i.e. if Ca²⁺ spark occurred, all of the microsparks' initiation points had to be included in it. The microspark starting at position 4 does not seem to contribute to Ca²⁺ sparks and probably does not belong to the same FDS as microsparks starting at positions 1, 2 and 3.

Figure 4. The relation of ‘microsparks’ to ‘jitter’ of the initiation point of Ca²⁺ sparks.

A is a portion of the line-scan image from Fig. 3Ca shown on a time-expanded scale. B, C and D show portions of the image shown in A, where the time scale was further expanded. Numbers 1–4 under the image in A denote the initiation position of the Ca²⁺ spark or ‘microspark’. Ca²⁺ sparks initiated from position 1 or position 3 were emphasized by horizontal lines (1, continuous line; 3, dashed line). Calibration: horizontal line, 1 s; vertical line, 5 μm (calibration for B also applies to C and D); relative fluorescence intensity was coded as shown on the scale in the bottom right corner. White upward arrows denote Ca²⁺ sparks and ‘microsparks’ starting from position 1, while downward arrows denote those starting from position 3.

Action of noradrenaline

As a weak effect of NO donors on [Ca²⁺]_i might be due to the fact that [Ca²⁺]_i under resting conditions was already low, noradrenaline was used to elevate it and the effects of NO donors were investigated under these conditions.

Administration of noradrenaline, after a lag of between 2 and 20 s, evoked a transient, in most cases generalized, Ca²⁺ wave lasting from about 7 s up to more than 23 s, which originated at a FDS (Fig. 5A). The Ca²⁺ wave was accompanied in some cells by contraction, depending on the concentration of noradrenaline used. Although the rate of onset of [Ca²⁺]_i increase by noradrenaline varied a lot, in general lower noradrenaline concentrations evoked a response more slowly than higher concentrations (not shown). The majority of examined cells both showed Ca²⁺ sparks and responded to noradrenaline. Cells which showed Ca²⁺ sparks but did not respond to noradrenaline were also observed and their lack of response might be explained by digestive enzymes damaging the α-adrenoceptors. Some cells that did not show Ca²⁺ sparks responded to noradrenaline.

Figure 5. The effect of noradrenaline on FDSs and intracellular Ca²⁺ in isolated arterial smooth muscle cells in the presence of NO donors or 8-Br-cGMP.

Parts of normalized series of images taken in the presence of 5 μm noradrenaline (Aa) and in the presence of 5 μm noradrenaline with 50 μm SNAP (Ab). Thirty images from each series are shown (white numbers in the frames), taken every 494 ms. White horizontal bar in the first frame represents 10 μm. Noradrenaline was added 9.9 s before the twenty-first frame. Note that the Ca²⁺ wave f starts from FDS e. B, statistical evaluation of data (means ± s.e.m.; n = 6 or more); the bars to the left show fluorescence change (expressed as%/pixel) of controls and the bars to the right the averages when the test substance was added (a, two applications of 5 μm noradrenaline as a control experiment; b, 50 μm SNAP with 5 μm noradrenaline; c, 100 μm SNP with 1 μm noradrenaline; d, 100 μm 8-Br-cGMP with 3 μm noradrenaline). *Statistically significant (P < 0.05). The sequences under Aa and Ab are also published as supplementary material at The Journal of Physiology online (for video 3, Noradrenaline, and 4, SNAP effect, see: http://www.jphysiol.org/cgi/content/full/539/1/25).

More detailed examination of the action of noradrenaline by line-scan imaging revealed that a low concentration of noradrenaline (e.g. 1 μm) stimulated a FDS to discharge more often (not shown) but, as with NO donors, did not alter the general position of a FDS although some ‘jitter’ was seen. Concentrations of noradrenaline higher than 1 μm usually increased the frequency of Ca²⁺ sparks which then summated into a transient Ca²⁺ wave followed by a decrease in subplasmalemmal Ca²⁺ concentration and cessation of discharges from FDS (Fig. 6D).

Figure 6. The effect of noradrenaline on the position and activity of single FDSs in isolated arterial smooth muscle cells pretreated either with NO donors or 8-Br-cGMP.

Line-scan images were captured every 2.9 ms. Aa, Ba, Ca and Da are controls, panels b were made in the presence of test substance and panels c were made after addition of noradrenaline (in D only noradrenaline was added). A, 100 μm SNP, 1 μm NA; B, 50 μm SNAP, 3 μm NA; C, 100 μm 8-Br-cGMP, 3 μm NA; D, 5 μm NA. Normalized fluorescence intensity is coded as shown on the grey-scale bars under each series of images.

Effects of NO donors and cGMP analogue on noradrenaline-evoked responses

SNAP (50 μm), applied 3–4 min before addition of noradrenaline, inhibited the response of [Ca²⁺]_i to 5 μm noradrenaline (noradrenaline alone: 7.97 ± 2.88%/pixel vs. noradrenaline + SNAP: 2.28 ± 1.62%/pixel, n = 6, P < 0.05; Fig. 5A and Bb). Also, SNAP decreased the activity, but did not affect the general position, of FDSs and did not introduce any new FDSs (not shown). These sequences are available as video 3, Noradrenaline, and video 4, SNAP effect, respectively, published as supplementary material at The Journal of Physiology online; http://www.jphysiol.org/cgi/content/full/539/1/25. APA (50 μm), a product of SNAP decomposition, did not affect the noradrenaline-evoked fluorescence increase (103.1 ± 11.5% of control response; n = 4, P > 0.05; not shown). SNP (100 μm) had similar effects to SNAP: it inhibited the [Ca²⁺]_i increase evoked by 1 μm noradrenaline from 6.58 ± 0.81%/pixel to 3.56 ± 1.30%/pixel, n = 7, P <0.05 (Fig. 5Bc), and suppressed the activity of FDSs by decreasing the number of discharges (not shown). Control experiments showed that a second administration of noradrenaline (5 μm) evoked a [Ca²⁺]_i increase whose magnitude was similar to that evoked by the first administration of noradrenaline (first: 13.22 ± 2.16 vs. second: 17.75 ± 3.67%/pixel, n = 6, P > 0.05; Fig. 5Ba).

In contrast to the NO donors, 8-Br-cGMP (100 μm) was not able to inhibit contraction due to 3 μm noradrenaline and an increase in [Ca²⁺]_i. Control (3 μm noradrenaline): 25.63 ± 6.20%/pixel; 100 μm 8-Br-cGMP + 3 μm noradrenaline: 21.68 ± 3.80%/pixel (n = 8, P > 0.05; Fig. 5Bd).

Line-scan imaging revealed that SNP (Fig. 6A) and SNAP (Fig. 6B) reduced the development of a Ca²⁺ wave evoked by noradrenaline (n = 3 and n = 4, respectively). 8-Br-cGMP did not prevent the development of a Ca²⁺ wave in 3 cases (Fig. 6C) out of 5 and in 2 cases there was an increase in frequency of Ca²⁺ sparks in response to noradrenaline. The positions of FDSs changed neither in the presence of inhibitory substances (SNP, SNAP, 8-Br-cGMP; panels b in Fig. 6), nor in the presence of noradrenaline (panels c in Fig. 6). Also worth noting is the pattern of Ca²⁺ sparks in Fig. 6B where 4 FDSs located next to each other discharge independently and Ca²⁺ sparks do not summate into Ca²⁺ waves.

Role of soluble guanylate cyclase in the action of NO donors

In order to investigate the involvement of soluble guanylate cyclase (sGC) in the action of NO donors, application of NO donor was made in the presence of 10 μm ODQ, an inhibitor of sGC, and noradrenaline. Both SNAP and SNP retained their ability to suppress the activity of FDSs (SNAP: site a in Fig. 7A; SNP: site b in Fig. 7B) and to inhibit the rise in [Ca²⁺]_i in response to noradrenaline (SNAP: to 45.7 ± 12.9% of the control response, n = 6, P < 0.01; Fig. 7Ac; SNP: to 51.6 ± 8.3% of the control response, n = 6, P < 0.01; Fig. 7Bc). The presence of ODQ did not affect the action of 8-Br-cGMP, which acts downstream of sGC. 8-Br-cGMP in the presence of ODQ did not change the rise in [Ca²⁺]_i significantly (to 223.1 ± 133.4% of the control response, n = 5, P > 0.05; Fig. 7Cc).

Figure 7. The role of soluble guanylate cyclase in the effect of NO donors.

Columns a and b, images showing the number of events where fluorescence of individual pixels exceeded 1.5 F/F₀ were constructed from normalized series of images to detect FDSs. Column a, control series (noradrenaline + 10 μm ODQ); column b, noradrenaline + 10 μm ODQ + test substance. A, 5 μm noradrenaline, 50 μm SNAP; B, 1 μm noradrenaline, 100 μm SNP; C, 3 μm noradrenaline, 100 μm 8-Br-cGMP. Graphs in column c, statistical evaluation of fluorescence changes (expressed as% of control increase by noradrenaline; means ±s.e.m.; n = 5 or more), where the bars to the left show controls and the bars to the right the averages when the tested substance was added. *Statistically significant. The colour bar applies to columns a and b. Red letters in columns a and b show FDSs.

DISCUSSION

The results of this study suggest that, in myocytes of guinea-pig small mesenteric arteries, calcium is released from the SR at certain sites, FDSs, and that their activity, i.e. rate of discharge from these sites, is modulated both by the vasoconstrictor noradrenaline, which increases the activity of FDSs, and by vasorelaxants like NO donors, which decrease it in the presence of a vasoconstrictor. There was no evidence in these arterial smooth muscle cells for the suggestion (Nelson et al. 1995) that relaxation was associated with an increase in spark frequency or the occurrence of sparks at new sites. In the present experiments it was found that SNP and SNAP reduced the discharge of sparks in response to noradrenaline and also the frequency of spontaneous sparks contrary to previous observations (Porter et al. 1998). The present experiments, in contrast to those of Porter et al. (1998), compared spark discharge frequency in the same cell in the presence and absence of NO donor. It is also possible that mechanisms are different in rat coronary myocytes which were used by Porter et al. (1998). Although they showed that forskolin relaxation was antagonized by iberiotoxin, a potassium channel blocker, and by ryanodine, a blocker of calcium store release through ryanodine receptors/channels, the actions of iberiotoxin and ryanodine on NO donor relaxations in these small arteries do not seem to have been examined. Not all relaxants may act by increasing spark discharge and several mechanisms are likely to be involved in relaxation by NO donors.

Frequent discharge sites

The fact that FDSs do not change their general position in the presence of excitant or relaxant substances confirms the idea that the sites of Ca²⁺ release correspond to distinct structures in the cell rich in ryanodine receptors and SR membrane, which are located mostly in the subplasmalemmal region (Gordienko et al. 1998, 2001). The majority of cells we studied (87.5%) showed spontaneous Ca²⁺ sparks originating from one, but more often from several FDSs. The arrangement of multiple FDSs in the myocytes of small mesenteric arteries, in contrast to cardiac or skeletal muscle, did not show any discernible pattern and the sites appeared to be scattered in the generally subplasmalemmal cytoplasm.

Out of several FDSs, one was the most active (‘dominant’ or ‘pacemaker’ site) and the spontaneous and noradrenaline-evoked Ca²⁺ waves usually began from this site. It has been shown that Ca²⁺ sparks can be suppressed either with anti-RyR or with anti-IP₃R antibody (Boittin et al. 1999). The fact that noradrenaline increased the activity of the same sites from which spontaneous Ca²⁺ waves originated (presumably caused by activation of RyRs) supports the idea that in smooth muscle there are IP₃ receptors in whose proximity there is relatively large number of ryanodine receptors (Boittin et al. 1999). As a result, regardless of whether the stimulus for Ca²⁺ release is Ca²⁺ itself or IP₃, the Ca²⁺ spark originates from the same site, with slight ‘jittering’ of the exact point of origin which is caused by the fact that Ca²⁺ sparks represent almost simultaneous opening of several Ca²⁺ release channels in a cluster and the channel (or channels) starting the discharge might not be the same one each time (Parker et al. 1996; Gonzalez et al. 2000; Gordienko et al. 2001). Thus calcium-induced calcium release (CICR) plays a crucial role in guinea-pig small mesenteric arteries as in guinea-pig ileal smooth muscle (Gordienko et al. 1998). It seems that CICR is the mechanism by which Ca²⁺ waves, initiated either by IP₃-induced calcium release (IICR) or CICR, propagate through the cell.

Ca²⁺ release events of shorter duration and lower intensity than Ca²⁺ sparks named ‘microsparks’ were described by Gordienko et al. (1998) in the myocytes of guinea-pig ileum. Observation of microsparks in the myocytes of guinea-pig small mesenteric arteries originating from the same general sites as Ca²⁺ sparks supports the view that microsparks are likely to be the building blocks of Ca²⁺ sparks and presumably represent the opening of just one or a few SR Ca²⁺ channels. We show that several microsparks may be recruited into a Ca²⁺ spark in a different sequence, which causes the initiation point of a Ca²⁺ spark to ‘jitter’ by a fraction of a micrometre. In order to define a FDS, it had to discharge a Ca²⁺ spark as a single unit, i.e. the Ca²⁺ spark produced by it would have to include all of the microsparks' initiation points. The mechanism by which microsparks develop into a Ca²⁺ spark is not yet known; sometimes the microsparks trigger regular Ca²⁺ sparks or contribute to them, but on other occasions they appear as isolated events.

The results of Iino et al. (1994), Miriel et al. (1999), Ruehlmann et al. (2000) and Mauban et al. (2001) have suggested that vascular smooth muscle maintains its tension by asynchronous [Ca²⁺]_i oscillations in different cells rather than by sustained and simultaneous [Ca²⁺]_i increases in all cells. Our observation of sequential increases in [Ca²⁺]_i in three myocytes connected in a series is in accordance with this view. Asynchronous Ca²⁺ waves might be a more efficient way for a vascular smooth muscle to maintain its tension: it is enough to produce a Ca²⁺ wave in one cell and the Ca²⁺ thus released could trigger Ca²⁺ waves in neighbouring cells with a time lag which may represent the time necessary for diffusion of activating Ca²⁺ through gap junctions.

Gordienko et al. (1998) postulated that multiple FDSs can be recruited to create a Ca²⁺ wave. However, the presence of two or more FDSs side by side does not automatically mean the creation of a Ca²⁺ wave whenever one of the FDSs discharges Ca²⁺. This process is obviously more complex as there are FDSs that can exist next to each other, yet discharge independently without recruiting one another to create a Ca²⁺ wave.

Ca²⁺ sparks and Ca²⁺ waves have also been observed in intact pressurized vessels. In myocytes within the vessel wall Miriel et al. (1999) have shown that Ca²⁺ sparks originate at preferred locations, which is in accordance with our findings in single myocytes. Ca²⁺ waves observed in pressurized cerebral arteries (Jaggar, 2001) had similar but briefer time courses compared to waves observed in single myocytes in our experiments.

Effects of adrenoceptor activation

We show that a typical response of a guinea-pig small mesenteric artery myocytes to adrenoceptor activation consists of four phases: (1) lag time, presumably during which the concentration of IP₃ increases; (2) increased discharge of Ca²⁺ sparks which summate into (3) generalized Ca²⁺ wave starting from the dominant FDS, and (4) a ‘silent’ period without Ca²⁺ sparks or Ca²⁺ waves.

Inositol trisphosphate, created by PLC activity coupled to α-adrenoceptor stimulation, activates its receptors on the SR and releases Ca²⁺ which in turn stimulates RyRs closely associated or co-operating with IP₃Rs (Boittin et al. 1999). During the onset of action of noradrenaline a small amount of IP₃ is sufficient only to release a small amount of Ca²⁺, capable of stimulating RyRs in the vicinity of IP₃R and thus evoking spatially restricted Ca²⁺ release events (i.e. Ca²⁺ sparks). The frequency of Ca²⁺ sparks gradually increases as the amount of IP₃ builds up, until there is enough Ca²⁺ being released to evoke a chain reaction of CICR manifested as a summation of Ca²⁺ sparks into a Ca²⁺ wave (seen in line-scan images) and recruitment of FDSs over all of the confocal plane to discharge Ca²⁺ (seen in two-dimensional images). However, partial waves were not seen, which is in accordance with the view that individual myocytes respond in ‘all or none’ fashion to α-adrenoceptor activation (Miriel et al. 1999; Mauban et al. 2001; Zang et al. 2001). After a Ca²⁺ wave, [Ca²⁺]_i declines and Ca²⁺ spark discharge is inhibited (Mauban et al. 2001) either through protein kinase C (PKC) activity (Bonev et al. 1997) or because the Ca²⁺ content of stores and thus Ca²⁺ flux through RyRs have been decreased (Komori et al. 1993; Zholos et al. 1994; ZhuGe et al. 1999). As the stores of Ca²⁺ in the SR are limited, the Ca²⁺ wave is transient and after the store content has decreased, [Ca²⁺]_i decreases as well because of the activity of mechanisms removing Ca²⁺ from the cytoplasm (Mogami et al. 1998; Shmigol et al. 2001). Because in the present experiments the action of noradrenaline was monitored only for 25 s and at room temperature (which will slow down biochemical processes in the cell), these results do not exclude the existence of a cyclic process where further Ca²⁺ waves in the continuing presence of adrenoceptor agonist may occur, as observed in intact pressurized vessels at 36–37 °C (Miriel et al. 1999; Mauban et al. 2001; Zang et al. 2001).

Effects of NO donors

Due to the fact that NO is a mixture of three redox forms (for review see Hughes, 1999), the nitric oxide radical (NO•), the nitrosonium cation (NO⁺) and the nitroxyl anion (NO⁻), the effects of NO can be diverse and depend on the prevalence of the NO form taking part in the reaction, its concentration and on the type and proximity of target molecules.

NO donors had a small effect on myocytes of guinea-pig small mesenteric arteries under resting conditions, presumably because [Ca²⁺]_i and the activity of FDSs in such cells were already low. However, in the presence of the NO donors SNAP or SNP the noradrenaline-evoked increase in [Ca²⁺]_i and the activity of FDSs were inhibited in a cGMP-independent way as 8-Br-cGMP, a membrane-permeable analogue of cGMP, did not affect them significantly. This result was further supported by the inability of ODQ, a specific inhibitor of soluble guanylate cyclase, to affect the actions of SNAP and SNP on noradrenaline-evoked responses. ODQ was used at a supramaximal concentration which all but abolishes the increase in cGMP evoked by SNP or SNAP (Garthwaite et al. 1995). Others have found ODQ without effect in coronary resistance arteries (Wiley & Davenport, 2001). MacMillan & Gurney (2001) have reported that 8-Br-cGMP did not mimic the effects of SNP on rabbit aortic myocytes. Both SNAP and SNP had qualitatively the same effects and APA, a product of SNAP decomposition, lacked these effects, which supports the view that both of these NO donors acted through the release of NO.

SNAP or SNP release NO in a reaction catalysed by and taking place in the vicinity of the cellular membrane (Butler & Rhodes, 1997). Because in biological systems the heterolytic cleavage of the S–N bond of S-nitrosothiols predominates over the homolytic one, in the case of SNAP the majority of nitric oxide is released in the form of NO⁺ which might be expected to act independently of guanylate cyclase. SNP is also capable of releasing both NO· and NO⁺ (Feelisch, 1998). NO⁺ is strongly electrophilic and reacts readily with thiol groups of proteins (Gaston, 1999), nitrosylating them. Because the life time of NO⁺ in water is less than a nanosecond (Hughes, 1999), it is reasonable to suppose that it reacts at the place of its generation, i.e. with membrane-bound proteins. SNAP also directly transnitrosylates thiol groups on proteins independently of the formation of nitric oxide (Feelisch, 1998). Nitrosylation of thiol groups of K⁺ channels (e.g. BK_Ca channel: Bolotina et al. 1994; Koh et al. 1995; Abderrahmane et al. 1998; Mistry & Garland, 1998; Lang et al. 2000) may then increase their open probability, hyperpolarize the membrane and lead to decrease in [Ca²⁺]_i and in the frequency of discharge from the FDSs (Jaggar et al. 1998). Ghisdal et al. (2000) have demonstrated in rat mesenteric artery myocytes that SNAP slightly hyperpolarized the membrane and decreased [Ca²⁺]_i in resting myocytes, and in noradrenaline-stimulated myocytes repolarized the membrane and decreased [Ca²⁺]_i, both to resting levels. Simonsen et al. (1999) have shown that 10 μm of SNAP produced 10–15 nm of nitric oxide in tens of seconds after addition to their vessel preparation. Therefore, in our experiments, the concentration of NO released by 50 μm of SNAP could have been in the range 50–75 nm. As the apparent K_m of nitric oxide for guanylate cyclase lies also in the lower nanomolar range (Ignarro et al. 1993), it would be expected that the effect of nitric oxide donors would involve a cGMP-mediated effect (Ghisdal et al. 2000; Pauvert et al. 2000) which inhibits Ca²⁺ release from the SR (Kannan et al. 1997; Ji et al. 1998) or phosphorylates BK_Ca channels via protein kinase G (Robertson et al. 1993; Taniguchi et al. 1993; Zhou et al. 2000). Nevertheless in our experiments no guanylate cyclase effect was detected. Effects of NO independent of cGMP were postulated to operate in aortic myocytes, where the decrease in [Ca²⁺]_i due to sarco/endoplasmic reticulum calcium ATPase (SERCA; Cohen et al. 1999) activity and inhibition of store-operated calcium entry could not be blocked by ODQ (Weisbrod et al. 1998). Another potential target for cGMP-independent effect of NO is the ryanodine receptor, rich in thiol groups (Eu et al. 1999). The nitrosylation of these groups by NO was suggested in cardiac and skeletal types of RyR, where both activation or inactivation of the RyRs is possible, depending on the concentration of NO and on the thiol groups being nitrosylated (Záhradníková et al. 1997; Xu et al. 1998; Suko et al. 1999; Hart & Dulhunty, 2000).

In canine airway smooth muscle NO⁺ is thought to act via a cGMP-independent process to release Ca²⁺ in subplasmalemmal regions, activate the BK_Ca channels and thus relax smooth muscle (Janssen et al. 2000). This should produce an increase in the fluorescence at the edge of the cells and a decrease in central cytosol. However, we did not observe this; the colour-coded images of cells show either uniform increase or decrease all over the confocal plane of the cell, or the regions where [Ca²⁺]_i had increased or decreased are scattered in the confocal plane without obvious pattern. It is possible that [Ca²⁺]_i increases in a very thin layer just under the plasmalemma and in this case the resolution of confocal microscope would be insufficient to detect it. Another possibility, which seems a more plausible explanation in view of our results (nitric oxide-evoked decrease both in [Ca²⁺]_i and in the activity of FDSs) is that Ca²⁺ might be expelled out of the cell rather than taken up into the SR, since uptake of Ca²⁺ into the SR might be expected to increase the rate of discharge from the FDSs (ZhuGe et al. 1999).

Conclusions

Myocytes from guinea-pig small mesenteric arteries were usually found to have multiple sites of Ca²⁺ release which they seem to use both to initiate Ca²⁺ waves leading to contraction and to produce Ca²⁺ sparks. Noradrenaline may recruit new FDSs and increases the activity of ones already discharging; NO donors inhibit the actions of noradrenaline on these sites at least in part through a cGMP-independent pathway, but do not increase spark activity nor recruit new FDSs.

Supplementary material

The online version of this paper (http://www.jphysiol.org/cgi/content/full/539/1/25) contains supplementary material entitled: ‘Multiple FDSs’ (video 1), ‘Asynchronicity’ (video 2), ‘Noradrenaline’ (video 3) and ‘SNAP effect’ (video 4).