Investigating the recruitment and synchronization of smooth muscle cells (SMCs) is the key to understanding the physical mechanisms leading to contraction and spontaneous diameter oscillations of arteries, called vasomotion. We improved a method that allows the correlation of calcium oscillations (flashing) of individual SMCs with mean calcium variations and arterial contraction using confocal microscopy. Endothelium-stripped rat mesenteric arteries were cut open, loaded with dual calcium fluorescence probes, and stimulated by increasing concentrations of the vasoconstrictors phenylephrine (PE) and KCl. We found that the number and synchronization of flashing cells depends on vasoconstrictor concentration. At low vasoconstrictor concentration, few cells flash asynchronously and no local contraction is detected. At medium concentration, recruitment of cells is complete and synchronous, leading to strip contraction after KCl stimulation and to vasomotion after PE stimulation. High concentration of PE leads to synchronous calcium oscillations and fully contracted vessels, whereas high concentration of KCl leads to a sustained nonoscillating increase of calcium and to fully contracted vessels. We conclude that the number of simultaneously recruited cells is an important factor in controlling rat mesenteric artery contraction and vasomotion.
- cellular communication
- potassium chloride
by measuring the mean intracellular calcium concentration ([Ca2+]i) of a given tissue region, previous studies have shown that the contraction of arteries and arterioles is due to the simultaneous increase of [Ca2+]i in all smooth muscle cells (SMCs) of the arterial wall (3, 9, 19, 20). Only recently, Peng et al. (14) evaluated the [Ca2+]i at the level of individual cells to investigate in detail the onset of spontaneous diameter oscillations, called vasomotion. The authors developed a new hypothesis to explain how synchronization of different SMCs is achieved to produce vasomotion: diameter oscillations are accompanied by a rhythmic depolarization of the cell membrane, which induces synchronous entry of extracellular Ca2+ in all SMCs through voltage-operated Ca2+ channels; this, in turn, synchronizes Ca2+ release from the sarcoplasmic reticulum through Ca2+-induced Ca2+ release and starts vasomotion (12).
Phenylephrine (PE) is a widely accepted vasoconstrictor and is used to study the mechanism of Ca2+ variations in SMCs (1, 6–8, 18); after binding to the α1-adrenoceptor of SMCs, it activates phospholipase C and releases inositol 1,4,5-trisphosphate [Ins(1,4,5)P3] and Ca2+ from the endoplasmic reticulum (10). Zang et al. (21) have shown that activation of the α1-adrenoceptor with PE strongly depends on the recruitment of SMCs and that the number of SMCs exhibiting an increase of [Ca2+]i grows gradually with PE concentration, finally leading to an “all-or-none” response. The concentration-dependent response of arteries upon PE stimulation suggests that the concentration of the vasoconstrictor is an important parameter. However, a dose-response curve at the cellular level has not yet been well established. To test whether a similar effect can be obtained by using nonselective vasoconstrictors, we also stimulated artery strips with increasing concentrations of KCl. High extracellular [K+] leads to depolarization of the cell membrane and opens voltage-gated Ca2+ channels without directly inducing the intracellular cascades described for PE activation.
Different methods have been used to investigate [Ca2+]i changes at the cellular level: Zang et al. (21) mounted whole intact arteries on a glass cannula, whereas Ruehlmann et al. (15) investigated rabbit vena rings, mounted isometrically; similar approaches have been used by Peng et al. (14) and Miriel et al. (11) to study arteries. Because these studies have been carried out under isometric conditions, they do not allow the correlation of Ca2+ dynamics with vessel contraction. Furthermore, they used a virtual line scan to study the recruitment of SMCs, which does not allow for spatial resolution. We developed a confocal microscopy method to simultaneously study Ca2+ dynamics in various individual cells of an arterial strip that was free to contract in the direction of the SMC axis; potential motion artifacts were eliminated by using a ratiometric Ca2+ fluorescence measurement (2). Our method allows the bidimensional analysis of spatial Ca2+ dynamics in the SMC population on a large number of cells.
The correlation of local contraction with local Ca2+ dynamics enabled us to understand the mechanism of vasomotion at a cellular level and address the following issues: 1) how does the recruitment of SMCs depend on the concentration of the vasoconstrictors PE and KCl? 2) what is the minimum quantity of cells exhibiting synchronous Ca2+ oscillations to provoke a significant local contraction? and 3) is there random calcium activity in individual cells when the tissue presents no significant mean calcium variations and no global contraction in the absence of stimulation or at a low PE and KCl level?
MATERIALS AND METHODS
Preparation of Arteries
Male Wistar rats (250–350 g) were obtained at the animal house of the Swiss Federal Institute of Technology and treated in agreement with the Care of Animals (edited by l'Académie Suisse des Sciences Médicales and la Société Hélvétique des Sciences Naturelles) and were briefly anesthetized with halothane and killed by decapitation. The mesenteric arcade was excised and placed in a modified Krebs-Ringer solution [physiological saline solution (PSS)] at 4°C containing (in mM) 145 NaCl, 5 KCl, 1 CaCl2, 0.5 MgSO4, 1 Na2HPO4, 20 HEPES, 23 Tris base, and 10.1 glucose. First- or second-order arteries were then cleaned from the surrounding tissues, cut open, and gently rubbed with a cotton bud to remove the endothelium. The use of endotheliumdenuded vessels allowed us to directly load SMCs with fluorescent Ca2+ dyes and assured us that the measured fluorescence exclusively originated from Ca2+ variations of SMCs. Furthermore, removal of the endothelium ensured that the observed effect stemmed uniquely from SMC dynamics and excluded the influence of endothelial cells (ECs). All manipulations were carried out at room temperature.
[Ca2+]i was expressed as a fluorescence intensity ratio obtained from fura red-AM and fluo 4-AM fluorescence (2); ratioing (fluo 4 to fura red) was performed on a pixel-by-pixel basis. This method eliminates motion artifacts and out-of-focus movements of the strip, because an increase in [Ca2+]i simultaneously increases the fluorescence of fluo 4 and reduces that of fura red. A concentration of 40 μM fura red and 20 μM fluo 4 was prepared using PSS; Pluronic F-127 (2% wt/vol) was added to improve loading of SMCs with 1 ml of this solution at 37°C for 30 min. After the dye loading, the artery strip was mounted intimal face down in a home-built observation chamber and stretched onto the glass coverslip to avoid long-axial movement. During measurements, the strip was continuously superfused with PSS perpendicular to the vessel axis at a rate of 1.5 ml/min; the tissue was rinsed for 15 min with PSS. To maximally reduce motion artifacts, the superfusion medium was not pulsed but was led into the chamber using gravitation. On the distal side of the observation chamber, PSS was removed using a pump and was not recirculated.
For [Ca2+]i fluorescence imaging, we used an Axiovert 135M inverted microscope linked to a LSM 410 laser-scanning unit (Zeiss; Oberkochen, Germany). The microscope was equipped with a Zeiss Fluar ×10 objective (numerical aperture 0.5). The 488-nm excitation wavelength of an argon ion laser was used to excite the probes, and emitted light was passed through the FT510 and FT560 dichroic beam splitter, respectively. The fluo 4 and fura red fluorescences were simultaneously recorded on two different photomultipliers (PMT). PMT1, equipped with a BP 510-525 filter, recorded the fluo 4 fluorescence (maximal emission at 516 nm), whereas PMT2, equipped with an LP 590 filter, recorded the fura red fluorescence (maximal emission at 650 nm). All optical and digital parameters were controlled with software on an IBM-compatible control computer. Sequences of 300 images were taken at a rate of 2 frames/s for up to 160 s on a tissue surface of 150 × 150 μm (256 × 256 pixels). Given the typical size of a SMC of ∼5 × 50 μm, we were able to simultaneously study ∼80 cells. All acquisitions were performed at room temperature within <2 h after dye loading. Immediately after the acquisition was started, PE and KCl were added to the chamber for the whole time of an acquisition. Arteries were used for a maximum of five stimulations with vasoconstrictors in steps of increasing concentration.
A processing and analyzing software (Metamorph, Visitron) was used to calculate the mean ratio and local strip contraction. Dye-loaded artery strips exhibit horizontal lines perpendicular to the SMCs, which represent collagen fibers (Fig. 1A). They are autofluorescent in both channels and do not appear after ratioing. The distance between two of these elastic fibers has been determined to measure the local contraction. The strip was clamped (constant length) in a way that allowed it to freely contract perpendicularly to the vessel axis. SMC contraction occurred perpendicularly to the vessel axis at approximated constant force given by the stretch in the vessel axis direction. Dynamics of [Ca2+]i of individual cells were measured by following the corresponding region of interest throughout all frames of an acquisition.
All chemicals were obtained from Sigma (Buchs, Switzerland) except for the dyes and pluronic acid, which were obtained from Molecular Probes (Leiden, The Netherlands),
We observed three distinct modes of SMC recruitment corresponding to stimulation with PE in three different concentration ranges, tested on fifteen arterial strips from nine animals. To obtain quantitative data, one segment per strip (Fig. 1A) has been selected, representing a bidimensionnal network of SMCs. After removal of the endothelium, only few ECs remained (bright yellow or red horizontal spots on Fig. 1A). Green regions correspond to internal elastic membranes between SMCs and ECs, which have not been totally removed and which do not appear on the ratio. To understand the mechanism of SMC recruitment after PE stimulation, a representative population of cells was selected and followed over time. Regions of interest were chosen on individual SMCs (yellow, Fig. 1A) in such a way that they could unambiguously be discriminated on each frame. The mean fluorescence ratio (MFR) of the entire segment (Fig. 1B) and of all regions of interest (numbered lines) was then simultaneously recorded over a time period of 160 s; local contraction was measured on every acquired frame.
At low PE concentrations (<0.4 μM), only few cells were asynchronously recruited (flashing) and no local contraction was observed (Fig. 2A). A small number of cells did not flash at any time (cells 4 and 5), and some cells only flashed once (cells 1, 2, 3, 6, and 8), whereas some cells oscillated more frequently (cells 7 and 9). Averaging of flashing over all cells of the region resulted in an almost constant MFR, which means that the recruitment of cells was not strong and/or synchronous enough to have a visible mechanical effect at the vessel level.
At medium PE concentrations (from 0.4 to ∼0.8 μM), the number of flashing cells significantly increased; the amplitude and periodicity of local contraction and Ca2+ dynamics were strongly dependent on the vasoconstrictor concentration, as presented for the responses at 0.6 μM PE (Fig. 1B) and 0.4 μM PE (Fig. 2B), respectively. Stimulation with 0.6 μM PE triggered a first period of asynchronous flashing of SMCs with no variations in the MFR and no local contraction (Fig. 1B). This phase was followed by synchronization of SMC flashing (a), which resulted in a clear peak in the MFR after 22 s, inducing a local contraction with a delay of 2.8 ± 0.4 s (mean ± SE; n = 12). SMC recruitment then became completely synchronized and more rapid, leading to a higher mean Ca2+ peak and stronger local contraction (b). Finally, mean [Ca2+]i decreased, leading to vessel relaxation. At this point, few cells exhibited weak and regular flashing (cells 1, 3, 5, and 8). This sequence of events was followed by multiple synchronized Ca2+ and contraction events: the strip exhibited vasomotion. Interestingly, the amplitude of MFR peaks decreased over time but without changing the level of local contraction, correlating with less important relaxations.
Changing PE concentration in the medium range from 0.6 to 0.4 μM resulted in a similar sequence of events but without stimulating vasomotion (Fig. 2B). It began with partial and asynchronous SMC recruitment leading to a small increase in the MFR (phase a), which was longer compared with 0.6 μM PE, but no local contraction was observed. In the following phase b, SMCs were only incompletely synchronized. This was visible on the MFR by a broad and low peak compared with the following phase c. Here, all cells were simultaneously recruited, leading to maximal contraction and a maximal transient increase of mean Ca2+. After this period, SMCs were only recruited asynchronously, without any indication of another total and synchronous event within 30 s; this is in contrast to vasomotion obtained after 0.6 μM stimulation.
At high PE concentrations (>0.8 μM), SMC recruitment appeared to be complete almost immediately and was followed by a strong local contraction (Fig. 2C); however, vasomotion was never observed. The vessel remained constricted, but the [Ca2+]i was still oscillating with all cells recruited synchronously. Oscillations in MFR exhibited decreasing amplitude and seemed to fade away completely at the end of the acquisition.
Cell recruitment analysis. To determine the level of SMC synchronization, we further evaluated the maximum number of cells recruited simultaneously as a function of PE concentration (Fig. 3). “X”s (Fig. 3A) represent the variations of fluorescence amplitude between the amplitude of first cell flashing (α in Fig. 1B) and the maximum of amplitude of the main peak on the MFR (β in Fig. 1B). These variations were proportional to the synchronization of SMC recruitment and were fitted with a sigmoidal plot. Three phases were distinguished: 1) at low PE concentrations, Ca2+ oscillated in few individual SMCs without reaching a sufficient number of flashing cells and/or synchronization to provoke a local contraction; 2) at high PE concentrations, the vessel remained totally contracted, but some Ca2+ oscillations still appeared; and 3) only at medium PE concentrations (shaded rectangle, from 0.4 to ∼0.8 μM, Fig. 3A) did Ca2+ oscillations induce vasomotion. The right border of the shaded rectangle (∼0.8 μM) represents the threshold between vasomotion and permanent tonic contraction of the arterial strip. Therefore, vasomotion appears only in a small range of PE concentrations, although Ca2+ oscillations existed indeed from 0.2 to 2 μM PE. Because of high variability between experiments, the determined threshold for tonic contraction is not well defined; however, during the establishment of the methodology, we never observed vasomotion at PE concentrations above 0.8 μM. Synchronization of SMC flashing in the intermediate PE concentration range grows with increasing concentration of PE (Fig. 3A). The solid circles represent relative local contraction variations, i.e., the local contraction maximum variation divided by the resting state; again, a sigmoid fits well (dotted line). The solid diamonds represent the relative mean fluorescence variations of three SMCs per experiment (in %) as a function of PE concentration. The horizontal line represents the mean value of all these points: 41 ± 7% (mean ± SE). It indicates that the maximum Ca2+ concentration reached in individual cells is independent of PE concentration. The slope of the mean fluorescence sigmoid was steeper than the slope of the local contraction curve, but both reached a plateau at similar PE concentration (∼0.6 μM), indicating that the arterial strip was totally contracted due to a maximum of synchronous SMC recruitment.
Time interval to synchronization. Another important parameter of synchronization kinetics is the time necessary to reach a first maximum of synchronously recruited SMCs, following flashing of the first SMC (from α to β in Fig. 1B), as presented on Fig. 3B. Synchronization time decreased exponentially with increasing PE concentration. Although the tissue never achieved total SMC recruitment at 0.2 μM PE, a low number of simultaneously flashing cells resulted in a visible peak after 97 ± 32 s on the MFR, which is not negligible. Variability in asynchronous flashing increased with decreasing PE concentration, as indicated by the larger error bars (means ± SE).
Cell population behavior. To compare cells from all experiments performed with PE, we selected cells with at least two transient Ca2+ increases over the time of observation (Fig. 4A). The time to reach a first peak (t1 in Fig. 4A) gave the x-coordinate in Fig. 4B; the y-coordinate was determined by the time lapse between the first and the second peak (T1 in Fig. 4A), thus giving the first point in Fig. 4B (t1, T1). The following points were plotted with coordinates t2 and T2 and so on.
Measurements were carried out at 0.2 (Fig. 4C), 0.4 (Fig. 4D), 0.6 (Fig. 4E), and 1.5 μM PE (Fig. 4F), and different symbols are attributed to each experiment. The grouping of points on the vertical axis indicated matching periods between two Ca2+ increases of cells from different experiments, which, however, were not synchronous. Substantial accumulation on the time axis represented a perfect overlap of Ca2+ oscillations (see dotted circles in Fig. 4). Generally, from 0.2 to 1.5 μM PE, the bandwidth of periods decreased by 66 ± 0.1%, demonstrating increasing synchronization with increasing PE concentration. The number of clusters (dotted circles), relating to the effective proportion of synchronized SMCs, similarly increased. At 0.2 μMPE (Fig. 4C), there were no clusters and the two experiments were distinct; at 0.4 μM PE, periods were more mixed, whereas at 0.6 μM PE (Fig. 4E) asynchronous flashing was particularly visible just after stimulation. It has to be stressed that, although the period may be the same for various cells, the time interval indicated whether the cells were flashing with the same period, albeit with a certain phase or time shift between each other. After 50 s, small clusters indicated increasing synchronization. At 1.5 μM PE (Fig. 4E), cells were obviously synchronized and no individual flashing appeared before the onset of mean Ca2+ oscillations.
To compare the recruitment of SMC after PE stimulation, we repeated our experiments using the non-specific vasoconstrictor KCl in concentrations ranging from 0.5 to 400 mM (nine arteries from five animals). Three different SMC behaviors were observed, corresponding to three KCl concentration ranges.
At low KCl concentrations (from 0.5 to ∼10 mM), cells flashed asynchronously and sporadically, with some completely quiescent cells (Fig. 5A, 0.5 mM KCl). Lines 1–5 (Fig. 5) represent Ca2+ dynamics of individual cells (in arbitrary units), and the lines below indicate the MFR of the whole observed region and the local contraction. Because of the lack of SMC synchronization, the MFR remained constant and no local contraction was observed.
At medium KCl concentrations (from ∼10 to ∼35 mM), arteries first exhibited asynchronous SMC flashing (Fig. 5B, 15 mM KCl), followed by a synchronous Ca2+ increase in all cells of the region of interest, which was slower compared with medium PE concentrations. The slow increase of MFR was followed by a tonic contraction, which was accompanied by asynchronous Ca2+ oscillation in all cells; vasomotion was never observed.
At high KCl concentrations (≥35 mM), cells were completely depolarized (13) and showed a sustained [Ca2+]i increase, apparent from the MFR (Fig. 5C, 50 mM KCl). Concentrations above 50 mM KCl did not change this behavior.
In the present study, we used a ratiometric principle to study [Ca2+]i dynamics in SMCs of the arterial wall with a confocal microscope, using the combination of fluo 4-AM and fura red-AM. This method allowed us to eliminate motion artifacts and out-of-focus movement of the strip (2, 17). The phase shift between the first occurring Ca2+ variation and the following oscillatory contractions (17) also confirmed that the observed fluorescence changes were not due to motion artifacts. On the other hand, the amplitude of the mean Ca2+ peaks decreased with time, as also observed by Savineau and Marthan (16). This was not due to photobleaching, which would result in a linear decrease of the fluorescence intensity. Furthermore, only the amplitude of the peaks was decreasing, not the baseline fluorescence.
Role of the endothelium. Our observation that PE induces vasomotion in the absence of intact endothelium is contradictory to the conclusion of Peng et al. (14). It can be objected that in Fig. 1A, in addition to SMCs (oriented vertically), a few ECs (oriented horizontally) could be observed. However, we previously observed vasomotion in an artery segment perfused in vitro in the absence of ECs (unpublished observations). In addition, these observations are in accordance with the results of Haddock et al. (5), who also observed vasomotion in irideal arterioles when the endothelium had been removed. Therefore, vasomotion can be induced by PE in the absence of an intact endothelium.
Integration of individual cell Ca2+ variation leading to tissue contraction. In our study, the maximal increase of intracellular Ca2+ in individual cells does not depend on PE concentration (Fig. 3A, diamonds and dashed horizontal line). In other words, Ca2+ flashing in an individual SMC is an all-or-none event. Therefore, the amplitude of the observed local contraction is not proportional to the concentration of Ca2+ reached in individual SMCs but rather to the degree of their synchronization. In that way, cells seem to be recruited by their threshold to flash in response to PE stimulation. This would be comparable to the recruitment of motor units in striated muscles to increase the strength of contraction and would allow a fine control of movement of the arterial wall using all-or-none contractile elements.
At low PE concentrations, during the recruitment phase, the vessel does not contract and only local asynchronous flashes occur. At medium PE concentrations, a sufficient number of cells is possibly recruited simultaneously so as to lead to a local contraction. The amplitude of the local contraction is proportional to the number of SMCs flashing synchronously. At high PE concentrations, the frequency of SMC flashing is sufficiently high to induce fusion of the Ca2+ peaks of individual SMCs. This leads to an increase in the interpeak Ca2+ level. We assume that this Ca2+ level is above the maximum Ca2+ level required to cause a supramaximal contraction of the SMCs. Therefore, a tonic contraction developed in that case, despite the fact that mean Ca2+ concentration oscillated in the high concentration range. This interpretation is confirmed by the observation that local contraction is well correlated with the MFR, which reflects in some way the average overall individual SMC contractions. Moreover, we observed that the degree of synchronization of SMC flashing increases with increasing PE concentration. A possible explanation is that the frequency of Ca2+ oscillations in one cell is elevated upon increasing PE concentration; this would increase the probability of a synchronized SMC flashing, as shown by Savineau et al. (16) by stimulating isolated SMCs with acetylcholine. Indeed, the number of clusters increases (dashed-lined circles in Fig. 4), whereas their size decreases at higher PE concentrations, suggesting that the cells' Ca2+ increase gets more synchronized at higher PE concentrations.
When stimulated by increasing KCl concentrations, cytosolic free calcium also oscillated in the SMCs; however, in contrast to the effect of PE, vasomotion was never observed in our experimental conditions. The reason is that synchronization of flashing did not occur between individual SMCs whatever the KCl concentration used to stimulate the local contraction. Three distinct phases have been observed, depending on KCl concentration: 1) at low concentrations, some asynchronous flashing takes place; 2) at medium concentrations, one synchronous Ca2+ increase takes place shortly after addition of KCl and leads to a strong and phasic contraction. This [Ca2+]i increase is preceded and followed by some asynchronous flashing; and finally, 3) at high KCl concentrations, only one synchronous sustained Ca2+ increase was elicited, inducing a tonic contraction, without calcium oscillations.
Therefore, the main important difference observed between the effect of stimulation with PE and high KCl concentration is that synchronization of SMC flashing occurs after PE stimulation but not after high KCl stimulation. PE acts through the activation of phospholipase C, activation of protein kinase C, and its cascade of phosphorylations, production of Ins(1,4,5)P3, and release of Ca2+ from the endoplasmic reticulum to the cytosol (11). The mechanism of resulting oscillations has been described by Gustafsson and Nilsson (4) and is accompanied by membrane potential oscillations (14). The effect of high KCl concentration is to depolarize the membrane potential by changing the K+ gradient across the cell membrane; this opens voltage-gated Ca2+ channels. It is not known whether this is linked to membrane potential oscillation. It could be speculated that the synchronization of flashing SMCs implicates electrochemical signal propagation between SMCs through gap junctions. In that case, the synchronizing signal would be membrane potential oscillations. Thus the absence of an oscillating membrane potential after high KCl stimulation and the observed lack of synchronization would explain the absence of vasomotion. This hypothesis will need further electrophysiological experiments.
In conclusion, our results suggest that the number of simultaneously recruited SMCs is an important factor, controlling rat mesenteric artery contraction and vasomotion rather than simply the SMC [Ca2+]i. Independently from the Ca2+ source (induced by PE or KCl), the cell displays Ca2+ oscillations, but in our experimental conditions vasomotion only occurs after stimulation with PE.
This work was supported by Swiss National Science Foundation Grant FN 31-61716.
The authors thank Drs. M. Capezzali and B. Hinz for careful reading of the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by the American Physiological Society