Intercellular calcium waves are associated with the propagation of vasomotion along arterial strips

Dominique Seppey, Roger Sauser, Michèle Koenigsberger, Jean-Louis Bény, Jean-Jacques Meister

Abstract

Vasomotion consists of cyclic arterial diameter variations induced by synchronous contractions and relaxations of smooth muscle cells. However, the arteries do not contract simultaneously on macroscopic distances, and a propagation of the contraction can be observed. In the present study, our aim was to investigate this propagation. We stimulated endothelium-denuded rat mesenteric arterial strips with phenylephrine (PE) to obtain vasomotion and observed that the contraction waves are linked to intercellular calcium waves. A velocity of ∼100 μm/s was measured for the two kinds of waves. To investigate the calcium wave propagation mechanisms, we used a method allowing a PE stimulation of a small area of the strip. No calcium propagation could be induced by this local stimulation when the strip was in its resting state. However, if a low PE concentration was added on the whole strip, local PE stimulations induced calcium waves, spreading over finite distances. The calcium wave velocity induced by local stimulation was identical to the velocity observed during vasomotion. This suggests that the propagation mechanisms are similar in the two cases. Using inhibitors of gap junctions and of voltage-operated calcium channels, we showed that the locally induced calcium propagation likely depends on the propagation of the smooth muscle cell depolarization. Finally, we proposed a model of the propagation mechanisms underlying these intercellular calcium waves.

  • conducted vasomotor response
  • smooth muscle cell
  • rat mesenteric artery

many kinds of muscular arteries present rhythmic oscillations of their diameter. This phenomenon, called vasomotion, occurs both in vivo and in vitro and is independent of any physiological rhythms, like the heartbeat or the respiratory cycle (1, 10). Vasomotion may have physiological and pathophysiological importance (18). The arterial diameter oscillations are induced by contractions and relaxations of the smooth muscle cells (SMCs) in the arterial wall. In a recent study, we have shown that the endothelium has mainly a modulating role in vasomotion, and that the presence of the endothelial cells (ECs) is not required to obtain vasomotion on rat mesenteric arterial strips (29). We have also demonstrated that a greater vasoconstrictor concentration is needed to induce vasomotion on intact strips than on endothelium-denuded strips. The contraction of the SMCs is due to an increase in their cytosolic calcium concentration ([Ca2+]i) (16, 19). Synchronization between the SMCs of the arterial wall is needed to induce vasomotion (12, 15, 20). However, it has been observed that contractions and dilatations can spread as a wave along the microvasculature during vasomotion (6). As arterial contraction results from an increase in the SMC [Ca2+]i, these contraction waves should be linked to intercellular calcium waves propagating along the artery.

The propagation of arterial diameter variations was observed in the microcirculation in response to a local stimulation (9, 27, 31). Although these conducted vasomotor response propagations were not studied during vasomotion, the propagation mechanisms may be similar. Several studies in arterioles or feed arteries (2, 4, 79, 11, 2528, 31, 33, 34) have investigated the conducted vasodilatation or vasoconstriction in response to a local stimulation with acetylcholine, respectively, phenylephrine (PE) or potassium chloride (KCl). It was concluded that the dilatation propagation is due to a propagation of the hyperpolarization of the membrane potential through gap junctions along the arterioles (7, 25, 33), whereas a propagation of the membrane potential depolarization is responsible for the contraction propagation (4, 8, 33, 34). The cell layer used as a pathway by the electrical signal leading to the conducted vasodilatation, namely vasoconstriction, has been investigated in arterioles (2, 33). The vasoconstriction is conducted by the smooth muscle layer alone, whereas both endothelium and smooth muscle layers transmit the vasodilatation signal. This last result seems to depend on the size of the arteries. Indeed, a study performed on mesenteric arteries showed that, in these larger arteries, the conducted vasodilatation is passing via the endothelium layer only (30). To the best of our knowledge, the conducted vasoconstriction has been investigated only in the microcirculation, where the distance of conducted vasoconstriction or even its existence itself seems to depend on the tissue and the species considered (11, 26, 28).

In the present study, our aim was to investigate the intercellular calcium and contraction waves spreading along the arteries during vasomotion. We have established the velocity of these waves on endothelium-denuded rat mesenteric arterial strips stimulated by PE. Local PE stimulations were then achieved to induce intercellular calcium waves, and the involvement of gap junctions and voltage-operated calcium channels (VOCCs) in the propagation mechanisms was assessed. Finally, a model of the propagation of the intercellular calcium waves based on the experimental data gathered was proposed.

MATERIAL AND METHODS

Preparation of the arteries.

Male Wistar rats weighting 300 ± 50 g were killed by decapitation after anesthesia with 4% of isofluran, in agreement with the care of animals (edited by the “Swiss Academy of Medical Sciences” and the “Helvetic Society of Natural Sciences”). All experiments performed on rat mesenteric arteries were done with the approval of the cantonal veterinary office (authorization number 1799.1) in agreement with the law on animal protection in Switzerland. The mesenteric arcade was excised and immediately placed in a Krebs Ringer solution containing the following (in mM): 145 NaCl, 5 KCl, 1.36 CaCl2, 0.54 MgSO4, 1.14 NaH2PO4, 20 HEPES, 11.56 Tris base, and 11.1 glucose. First- or second-order arteries, which correspond to arteries with a lumen diameter of ∼150–250 μm, were carefully cleaned from their surrounding tissues and opened longitudinally to obtain arterial strips with a length of 8–12 mm. The ECs were removed mechanically with a cotton bud to obtain strips without endothelium. At the end of every experiment, the strips were tonically precontracted with 3 μM PE, and the absence of the endothelium was confirmed by the loss of relaxation to 10 μM acetylcholine (29).

All strips were loaded for 30 min at 37°C, in 500 μl of Krebs Ringer solution containing 40 μM fura red AM, 10 μM fluo-4 AM, and 2% pluronic F-127. The strips were then mounted intima, face up, in a home-built observation chamber. By fixing the two ends of the strips, we avoided the long axial movement, but allowed the strips to freely contract in the direction perpendicular to the vessel axis (12). During the experiments, the strips were continuously superfused with a physiological solution gassed with a mixture of 5% CO2 and 95% air (to ensure a pH of 7.4) and containing the following (in mM): 128.2 NaCl, 4.7 KCl, 1.36 CaCl2, 1.05 MgCl, 0.42 NaH2PO4, 20.2 NaHCO3, and 11.1 glucose. This solution was superfused perpendicularly to the vessel axis at a rate of 2 ml/min. To maximally reduce motion artifacts, the superfusion was not pulsed, but led into the chamber using gravity. Before the first acquisition and between every measurement, the strip was rinsed for 20 min by the superfusion.

Vasoconstrictor stimulation.

The strips were stimulated either globally or locally with PE. In the case of the global stimulations, the whole strip length was continuously stimulated with the same concentration of PE. This was performed by adding PE directly to the superfusion. For the local stimulations, a micropipette was used to restrict the PE delivery to a small region of the strip. The micropipette used had a diameter of 5 μm and was held with a manual micromanipulator (Leica Microsystems, Wetzlar, Germany) 10 μm above the intima part of the strip, perpendicularly to the vessel axis. The injection was made in the direction opposite to the superfusion flow. The PE concentration in the micropipette was 40 μM. An injection time between 0.3 and 0.6 s at a pressure between 40 to 60 hPa was chosen on a microinjector pump, FemtoJet (Vaudaux-Eppendorf AG, Basel, Switzerland), to obtain a stimulated area of the same size in each experiment. Under the conditions used in these experiments, fluorescent dye (rhodamine dextran) ejected from the stimulating pipette could be detected in a region with a diameter of 100–150 μm. Except for the calcium-free experiments, local stimulations were repeated at least two times to ensure that the results were reproducible. All experiments were performed at room temperature.

Chemical and drugs.

All chemicals and drugs were obtained from Sigma (Buchs, Switzerland), except the dyes fluo-4 and fura red, which were obtained from Molecular Probes (Leiden, the Netherlands). The calcium-free solution was prepared like the physiological solution, except that the calcium was omitted and replaced by 2 mM EGTA. There was no incubation delay between control experiments and measurements in the presence of calcium-free solution. Before the experiments with nifedipine, BAY K 8644 or palmitoleic acid (PA), the strips were superfused during 20 min with a physiological solution containing one of the three drugs.

Image acquisition and processing.

Acquisitions were performed with an upright confocal microscope, SP2 (Leica Microsystems, Wetzlar, Germany), equipped with a ×20 water immersion, long working distance, and objective with a numerical aperture of 0.4 (Leica Microsystems, Wetzlar, Germany). A 488-nm wavelength argon laser was used to excite the probes. The emitted light from the two different dyes was collected simultaneously with two different photomultipliers (PMT). The first PMT recorded the fluorescence from fluo-4 by collecting the emitted light with a wavelength between 505 and 560 nm, and the second PMT recorded the fluorescence from the fura red by measuring the light with a wavelength between 625 and 725 nm. Images of 750 × 600 to 750 × 200 μm2 were taken every 675–333 ms, depending on the height of the images. The calcium measurement was made by performing a pixel per pixel ratio of fluo-4/fura red fluorescence intensity. This ratio value was proportional to the calcium concentration in the SMCs and eliminated most of the motion and out-of-focus artifacts (5, 14). To quantify the local strip contraction, two recognizable points forming a segment perpendicular to the vessel axis were tracked during the acquisitions. The distance between these two points was measured, and the local strip contraction was expressed in percentage of the initial segment length (12).

Wave velocity and propagation distance measurements.

To obtain the velocity of the intercellular calcium wave, the mean fluorescence ratio was measured in 10 regions of interest (ROIs) along the strip. Each ROI had a dimension of 30 × 75 μm2, and two consecutive regions were separated by a distance of 75 μm (Fig. 1). A cross-correlation technique was used to compute the delay between the raise in calcium in each ROI along the artery. Simultaneously, the velocity of the contraction wave was measured with an algorithm tracking the strip movements during the whole acquisition (24). For the locally stimulated waves, the distance over which the calcium wave propagated was measured from the border of the stimulated area.

Fig. 1.

Fluorescence image of a rat mesenteric arterial strip loaded with fluo-4 and fura red. The endothelium has been mechanically removed. To measure the velocity of intercellular calcium waves, 10 regions of interest (ROIs; depicted in red on the image) of 30 × 75 μm2 were placed along the strip. The ROIs are separated by a distance of 75 μm. The direction of the superfusion flow is given by the blue arrows.

Statistics.

Data are represented as the mean ± SD. “n” denotes the number of arterial strips used in each experiment, and each strip is coming from a different animal. Student's t-test was used to compare the wave velocity and the propagation distance between the different experiments.

RESULTS

Characterization of the vasomotion propagation: calcium and contraction waves.

When endothelium-denuded rat mesenteric arterial strip was globally stimulated with a PE concentration leading to vasomotion (i.e., a concentration between 0.4 and 0.8 μM), we observed that all parts of the artery did not contract simultaneously, but contraction spread along the strip. This contraction wave was accompanied by an intercellular calcium wave. When PE was applied, the artery first showed a transitory period during which the SMCs flashed asynchronously. Vasomotion appeared only after a delay, ranging from 30 s to 2 min, and sustained calcium and contraction waves spreading along the strips could then be observed. Figure 2A shows a typical recording obtained by measuring the mean fluorescence ratio in 10 ROIs along the vessel axis, as described in material and methods and in Fig. 1. The local contraction at four different places of the strip, corresponding to the ROIs 1, 4, 7, and 10, is also depicted. The calcium waves appeared after a transitory period, and their velocity remained unchanged during the whole experiment. The calcium waves could be observed to propagate from the proximal to distal side of the artery or from the distal to proximal side: there was no privileged direction. As shown on Fig. 2A, each calcium wave is followed by a contraction wave. The contraction waves were followed by a tracking algorithm, and their velocity was determined to be 120 ± 21 μm/s (n = 15). Taking the cross correlation between the calcium evolution in 10 different ROIs along the artery length, the calcium wave velocity has also been measured as 108 ± 22 μm/s (n = 15). This value was the same along the whole observed part of the strip and is not statistically different from the one obtained for the contraction wave (P = 0.14). A video showing intercellular calcium waves spreading during arterial vasomotion is provided as supplemental material with the online version of this article. In some acquisitions, the originating area of the waves was in the microscope field of view, and we observed that it was always the same during a complete acquisition (ROI 5 on Fig. 2B). Stimulating again the strip after 20 min of rinsing with physiological solution, the calcium waves still originated from the same location (Fig. 2C) (n = 7).

Fig. 2.

Typical time courses of the mean fluorescence ratio (MFR) in each of the 10 ROIs placed along strips (see Fig. 1). A: a strip globally stimulated with 0.6 μM phenylephrine (PE) first showed MFR increases, which were not synchronized among the different ROIs (transitory period). Vasomotion was then established, and MFR increases propagated along the arterial strip at a constant velocity. The red curves show the time evolution of the local contraction in four different regions corresponding to ROIs 1, 4, 7, and 10. B: a strip globally stimulated with 0.8 μM PE showed vasomotion and propagations of MFR increase, starting from a ROI localized in the observation field (ROI 5, bold curve). C: after having been rinsed during 20 min with the physiological solution, the strip used in B was again stimulated with 0.8 μM PE. During arterial vasomotion, propagations of MFR increase originated from the same ROI (bold curve) as in B. The broad gray and orange lines were added to better visualize the propagation of the MFR increase and contraction, respectively, along the arterial strips. au, Arbitrary units.

Intercellular calcium waves induced by local PE application.

To define the mechanisms involved during the propagation of vasomotion, a micropipette was used to stimulate a small part of the arterial strip with a high concentration of PE (40 μM). To ensure there was no diffusion of PE to other parts of the artery, the injections were made in the direction opposite to the superfusion flow. Several successive local stimulations were performed to ascertain the observed results.

When a rat mesenteric arterial strip was locally stimulated with 40 μM of PE, a rise in [Ca2+]i was measured in the SMCs localized in the stimulated area, but no change was recorded in the other SMCs of the strip. A propagation of the local calcium rise could only be observed when the strip was globally stimulated by a low PE concentration just below the concentration threshold inducing vasomotion (0.3 μM, Fig. 3A). The propagation was not totally regenerated, and the calcium wave propagated over a finite distance of 385 ± 38 μm (n = 11). The velocity of the intercellular calcium waves induced by a local stimulation was measured to be 100 ± 21 μm/s (n = 11) and remained the same along the whole propagation distance. This velocity is not statistically different from the one measured for the intercellular calcium waves spreading during vasomotion (P = 0.36).

Fig. 3.

Typical time courses of the MFR in each ROI placed along a strip. A: local stimulations of the strip with 40 μM PE showed no MFR increase, except in the stimulated area. There was no propagation when the artery was in its resting state. Combined with a global stimulation with 0.3 μM PE through the superfusion system, each local stimulation induced a propagation of the MFR increase affecting several ROIs. B: the strip used in A was rinsed during 20 min with physiological solution and 50 μM palmitoleic acid (PA). Under a global superfusion with 0.3 μM PE and 50 μM PA, this strip responded only locally to local stimulations with 40 μM PE. The arrows depict the moment when a local injection was performed. The area that is locally stimulated is represented by bold curves, and broad shaded lines have been added to better visualize the propagation of the MFR increase along the arterial strip.

Intercellular communication via gap junctions.

To examine whether the signal mediating the intercellular calcium waves propagated from cell to cell through gap junctions, experiments were performed on strips incubated during 20 min with physiological solution supplemented with PA, a gap junctions uncoupler (13). Strips showing calcium waves under local PE stimulations in the presence of 0.3 μM PE in the superfusion did not show any propagation after treatment with PA (50 μM), although 0.3 μM PE was still present in the superfusion (Fig. 3) (n = 4).

Importance of extracellular calcium.

To analyze the possible contribution of the extracellular calcium and check whether plasmalemmal calcium channels are involved in the propagation of the intercellular calcium waves, the physiological solution was replaced by a calcium-free solution. Strips on which local PE stimulations induced calcium waves in the presence of 0.3 μM PE and the physiological solution in the superfusion (Fig. 4A) did not show any calcium propagation in response to a local stimulation when they were superfused with calcium-free solution and 0.3 μM PE (Fig. 4B). To ensure that the sarcoplasmic reticulum (SR) was not empty at the end of the calcium-free experiments, a high concentration of PE (50 μM) was globally added. An increase of [Ca2+]i was observed in all ROIs (Fig. 4B) (n = 5).

Fig. 4.

Typical time courses of the MFR in each ROI placed along a strip. A: a strip stimulated locally with 40 μM PE showed no propagation of the MFR increase when it was in its resting state and a propagation of the MFR raise around the stimulated area when 0.3 μM PE were added to the superfusion. B: the strip used in A did not show any MFR increase in the neighboring ROIs of the stimulated area when the strip was locally stimulated with 40 μM PE and globally superfused with a calcium-free solution containing 0.3 μM PE. At the end of the recording, a high concentration of PE (50 μM) was added on the whole strip and induced an increase of MFR in all ROIs. The arrows depict the moment when a local injection was performed. The area that is locally stimulated is represented by bold curves, and broad shaded lines have been added to better visualize the propagation of the MFR increase along the arterial strip.

Role of VOCCs.

Nifedipine, a blocker of the L-type VOCCs (32), was added to check the involvement of these channels in the calcium wave propagation mechanism. A typical recording of the effect of nifedipine treatment on the calcium propagation in arterial strips is shown in Fig. 5. As already observed in Figs. 3A and 4A, no induced propagation could be observed in the absence of a global PE stimulation (Fig. 5A). Strips globally stimulated with 0.3 μM of PE showing calcium propagations in response to local stimulations with PE (Fig. 5B) did not show any calcium propagation around the stimulated area when nifedipine (1 μM) and 0.3 μM of PE were present in the superfusion (Fig. 5C, first 150 s) (n = 9). To check if it would still be the case when all of the SMCs of the strip were in a more excited state, we increased the concentration of PE in the superfusion to 1 μM in the presence of nifedipine. This induced an increase in the mean [Ca2+]i in all of the SMCs of the strips with calcium oscillations; however, it was still impossible to induce a calcium propagation around the stimulated area (Fig. 5C, seconds 150–280) (n = 4). The involvement of VOCCs was also tested using BAY K 8644, an L-type VOCC activator (36, 37). Strips that did not show any calcium waves in response to local stimulations when superfused with 0.2 μM PE (Fig. 6A) presented propagations around the stimulated area in response to local stimulation after treatment with BAY K 8644 (300 nM) and the addition of 0.2 μM PE in the superfusion (Fig. 6B) (n = 4). In the presence of BAY K 8644, locally stimulated calcium waves could be induced with a global stimulation of only 0.2 μM PE, but the propagation distance was smaller than the one measured on strips stimulated globally with 0.3 μM PE in the absence of BAY K 8644 (165 ± 21 vs. 385 ± 38 μm). Although the concentration of the global stimulation with PE required to obtain a locally induced calcium wave is decreased in the presence of BAY K 8644, a global stimulation is still needed for a calcium wave to propagate (data not shown).

Fig. 5.

Typical time courses of the MFR in each ROI placed along a strip. A: a strip stimulated locally with 40 μM PE showed no propagation of the MFR increase around the stimulated area when it was in its resting state. B: the strip used in A was globally stimulated with 0.3 μM PE. Each local stimulation with 40 μM PE induced a finite propagation of the MFR raise around the stimulated area. C: after having been superfused 20 min with the physiological solution supplemented by 1 μM nifedipine, the strip used in A and B did not show any MFR increase in the neighboring ROI of the stimulated area, even if the global concentration of PE was increase to 1 μM. The arrows depict the moment when a local injection was performed. The area locally stimulated is represented by bold curves, and broad shaded lines have been added to better visualize the propagation of the MFR increase along the arterial strip.

Fig. 6.

Typical time courses of the MFR in each ROI placed along a strip. A: a strip stimulated locally with 40 μM PE showed no propagation of the MFR increase around the stimulated area when it was globally stimulated through the superfusion system with 0.2 μM PE. B: after having been superfused during 20 min with the physiological solution supplemented by 300 nM BAY K 8644, the strip used in A showed a limited propagation of the MFR raise around the stimulated area. The arrows depict the moment when a local injection was performed. The area locally stimulated is represented by bold curves, and broad shaded lines have been added to better visualize the propagation of the MFR increase along the arterial strip.

Local KCl stimulation.

In a series of experiments, the strips were stimulated locally with KCl instead of PE. The local application of a high KCl concentration (200 mM) with the micropipette induced a [Ca2+]i increase in the SMCs of the stimulated area, but no calcium propagation could be observed in the remainder of the arterial strip, if the strip was in its resting state (Fig. 7A). When the same strip was globally stimulated with a low concentration of PE (0.3 μM) using the superfusion system, calcium propagations could be observed in response to local stimulations with KCl (Fig. 7B). The propagation distance of these calcium waves was measured to be 373 ± 66 μm, and their velocity was 103 ± 8 μm/s (n = 4). These results were not statistically different from the ones obtained when the calcium waves were induced by local stimulations with PE (wave velocity: P = 0.79; propagation distance: P = 0.66).

Fig. 7.

Typical time courses of the MFR in each ROI placed along a strip. A: a strip stimulated locally with 200 mM KCl showed no propagation of the MFR increase around the stimulated area when it was in its resting state. B: when the strip used in A was globally stimulated with 0.3 μM PE, the local stimulation with 200 mM KCl induced a propagation of the MFR raise around the stimulated area. The double arrows depict the moment when a KCl local injection was performed. The area locally stimulated is represented by bold curves, and broad shaded lines have been added to better visualize the propagation of the MFR increase along the arterial strip.

DISCUSSION

Although vasomotion was investigated for years, this study analyzes for the first time its propagation and the associated intercellular calcium waves. The velocity of the calcium and contraction waves has been precisely determined on endothelium-denuded rat mesenteric arterial strips. As shown in a previous publication, vasomotion can be obtained both on intact arteries and on endothelium-denuded strips (29). Unfortunately, the [Ca2+]i dynamics in SMCs can only be observed on endothelium-denuded strips. This is the reason why, in the present study, all of the experiments were performed on strips on which the ECs have been removed. When these strips were stimulated with a PE concentration leading to vasomotion, contraction and calcium waves spreading along the arterial strips were observed. Since the contraction of the artery is induced by an increase in the [Ca2+]i in SMCs (16, 19), it was expected that the velocity of the intercellular calcium waves spreading along the arterial axis during vasomotion was the same as the one of the contraction waves. The experiments proved that it was indeed the case (Fig. 2A). The transitory period during which the SMCs flashed asynchronously before the apparition of vasomotion has already been observed in some other studies (12, 20). It is possible to have an increase in a contraction curve with no calcium increase in the corresponding ROI (Fig. 2A). This can easily be explained by the fact that, if a calcium increase is present in the neighboring ROIs, then the tissue contraction in these ROIs will be transmitted to the surrounding tissue by mechanical displacement. On figures presenting locally induced calcium waves (Figs. 37), only the calcium curves are presented to avoid any misinterpretation that may arise from the presentation of contraction curves.

The presence of calcium waves propagating along the artery brings up the question of their origin and of the possible existence of some regions from which they start. Such starting areas have been observed in some experiments (Fig. 2). Interestingly enough, the starting area remained the same during whole acquisition, even if the strip was stimulated again after having been rinsed with physiological solution during 20 min. This means that the presence of a starting area is not an artifact, but indicates that the tissue is not homogeneous. In the present study, we have not investigated in greater details what singularizes these starting areas from the remainder of the arterial wall, but some hypotheses can be expressed. This singularity could come from the heterogeneity existing between SMCs of the same artery. SMCs show different sensitivity to PE (23). Some SMCs could present more calcium channels, or even the SR density could differ from cell to cell. These possible heterogeneities would induce differences in the SMC [Ca2+]i, causing, for instance, a disparity in SMC oscillation frequency. SMC groups oscillating at a different frequency from the remainder of the arterial wall have already been observed by Rahman et al. (22). The existence of starting areas could also be due to the presence of other cells playing a pacemaker role. Such cells could be similar to the arterial interstitial cells of Cajal-like cells observed in guinea pig mesenteric arteries, which were supposed to play a role in vasomotion (21). All of these different hypotheses are plausible and need to be tested in a forthcoming study.

To define the mechanisms involved in the propagation of the intercellular calcium waves arising during vasomotion, a method allowing the stimulation of only a small region of the arterial strip with a micropipette containing PE was used. The presence of intercellular calcium waves depends on the PE concentration present in the superfusion, i.e., on the global low-PE concentration applied on the arterial strip. When the strips were in their resting state, i.e., when no global stimulation was applied, a local PE stimulation failed to induce a calcium increase in any other cells than the stimulated ones. As we have observed in calcium waves during vasomotion, we tried to approach the vasomotion state by stimulating the strips with a low concentration of PE added to the superfusion. This low-PE global stimulation induced a slight increase in the [Ca2+]i of all of the SMCs, as shown in Fig. 3A. A local PE stimulation was then able to induce a limited intercellular calcium wave propagating from the stimulated area on a distance of ∼400 μm. The velocity of the calcium waves induced by local stimulations with low-PE global stimulation was the same as the one of the waves spreading during vasomotion. This suggests that the calcium waves in these two different configurations are directed by the same propagation mechanisms. The fact that the locally induced calcium waves propagated only on a finite distance, which was not the case for the waves arising during vasomotion, can be explained by the fact that, during vasomotion, all of the SMCs of the strip are in an oscillatory state (12), which is not the case when the strip is globally stimulated with a low concentration of PE.

When the communication through the gap junctions between the SMCs was inhibited with PA, only the stimulated area showed a calcium increase (Fig. 3). This indicates that the wave propagation mechanisms involve the gap junction pathway. This allows us to exclude a propagation via a paracrine mechanism.

As no propagation could be obtained when the physiological solution was replaced with a calcium-free solution (Fig. 4), at least one part of the [Ca2+]i increase during the propagation came from the calcium present in the extracellular medium. This calcium entry through the membrane is then essential for the calcium wave. It is known from the literature that the SR empties when the SMCs stay too long in a calcium-free solution (3). Therefore, only a single local stimulation was performed. At the end of these experiments, a global application of a high concentration of PE induced an important calcium increase in all SMCs. This demonstrates that the intracellular mechanisms inducing calcium increase were still functional at the end of the calcium-free experiments and particularly that the SR was not empty. It also indicates that the impossibility to induce a calcium wave in the absence of extracellular calcium did not come from a deficiency in the calcium release from the SR. The calcium entry through the cell membrane could happen through many different calcium channels of the SMC membrane.

Several studies demonstrated that a local PE stimulation depolarizes the stimulated SMCs (3335), and, in the microcirculation, the conducted vasoconstriction has been shown to propagate in response to the spreading of the depolarization along the arterioles (7, 25, 31, 33). To test whether it was also the case in the larger arteries considered in the present study, experiments with nifedipine were made. We have seen previously that all of the SMCs of the strip have to be excited just below the vasomotion state to locally induce calcium waves. As nifedipine could induce a decrease in the [Ca2+]i in all of the SMCs by closing the VOCCs, we decided to check the ability of a local stimulation to induce a calcium wave, not only with a global stimulation of 0.3 μM PE, but also with a higher global PE stimulation (1 μM). These experiments showed that locally induced calcium waves were inhibited when the VOCCs were blocked by nifedipine, even when the [Ca2+]i in all of the SMCs was increased by a high global stimulation of PE (Fig. 5). From these last results, we conclude that the propagation of a depolarization from the stimulated area to the remainder of the artery leading to an opening of the VOCCs is a reasonable hypothesis for the propagation pathway of the intercellular calcium wave.

It has been shown in mesenteric terminal arterioles that the conducted calcium response is not totally inhibited by nifedipine. A part of the response is due to nifedipine-insensitive VOCCs present in the arterioles: the T-type calcium channels (4, 8). The fraction of the T-type calcium channels in regard to the L-type calcium channels differs largely along the branches of mesenteric arteries. In the proximal branches used in the present study, the fraction of T-type over L-type channels is low. This ratio increases along the arterial tree to reach 90–100% in the terminal branch and arterioles (17). The low density of T-type calcium channels in the arteries used here could explain why the locally induced calcium waves did not propagate without a global excitation of the arterial strips, which was not required in experiments performed in arterioles (2, 4, 8, 9, 26, 28, 33, 34). The importance of the role played by the L-type calcium channel in the propagation of the calcium waves was confirmed using BAY K 8644, an L-type calcium channel activator (Fig. 6). In the presence of this drug, a propagation could be observed on strips globally stimulated with 0.2 μM of PE. This PE concentration was too low to induce any propagation around the locally stimulated area in control experiments. By replacing PE by KCl, no difference in the distance of propagation or in the velocity of the locally induced calcium waves was observed (Fig. 7). This shows that the calcium propagation is not dependent on the vasoconstrictor used for the local stimulation. The key element to induce intercellular calcium waves seems to be the depolarization of the stimulated area.

From the experimental data gathered, it is possible to propose a model of the mechanisms underlying the propagation. The intercellular calcium waves propagating from the stimulated area follow the propagation of a depolarization, which opens the VOCCs present in the SMC membrane. A part of the calcium present in the physiological solution enters the SMCs, but the increase in the [Ca2+]i is not sufficient to reach the threshold required for the calcium-induced calcium release in the SMCs. The [Ca2+]i in the SMCs must be artificially increased in all of the SMCs of the arterial wall by globally stimulating the strip. In our experiments, this was done by adding a low concentration of PE in the superfusion. The effect of BAY K 8644 confirms this hypothesis, since a lower concentration of PE is needed in the superfusion when more calcium enters the cells through VOCCs. The relatively slow calcium wave velocity of almost 100 μm/s measured in regard to the propagation velocity of an electrical signal, which is quasi-instantaneous, could be explained by the time needed for calcium influx through VOCCs and for the calcium-induced calcium release mechanism.

In conclusion, the present study shows for the first time contraction and intercellular calcium waves spreading along arterial strips during vasomotion. Local stimulations of PE performed on endothelium-denuded strips failed to induce calcium propagations when the strips were in their resting state. However, if the [Ca2+]i in all SMCs was slightly increased, intercellular calcium waves propagating from the stimulated area could be observed. A model of the propagation mechanisms of these waves has been suggested. As the calcium waves induced by local stimulations have the same speed as the ones occurring during vasomotion, we propose that the propagation mechanisms are the same in the two different configurations.

GRANTS

This research was supported by the Swiss National Science Foundation grant (FN 310000-114097).

DISCLOSURES

I am not aware of financial conflict(s) with the subject matter or materials discussed in this manuscript with any of the authors, or any of the authors' academic institutions or employers.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
View Abstract