Relationship between Ca²⁺ sparklets and sarcoplasmic reticulum Ca²⁺ load and release in rat cerebral arterial smooth muscle

Abstract

Ca⁺ sparklets are subcellular Ca²⁺ signals produced by the opening of sarcolemmal L-type Ca²⁺ channels. Ca²⁺ sparklet activity varies within the sarcolemma of arterial myocytes. In this study, we examined the relationship between Ca²⁺ sparklet activity and sarcoplasmic reticulum (SR) Ca²⁺ accumulation and release in cerebral arterial myocytes. Our data indicate that the SR is a vast organelle with multiple regions near the sarcolemma of these cells. Ca²⁺ sparklet sites were located at or <0.2 μm from SR-sarcolemmal junctions. We found that while Ca²⁺ sparklets increase the rate of SR Ca²⁺ refilling in arterial myocytes, their activity did not induce regional variations in SR Ca²⁺ content or Ca²⁺ spark activity. In arterial myocytes, L-type Ca²⁺ channel activity was independent of SR Ca²⁺ load. This ruled out a potential feedback mechanism whereby SR Ca²⁺ load regulates the activity of these channels. Together, our data suggest a model in which Ca²⁺ sparklets contribute Ca²⁺ influx into a cytosolic Ca²⁺ pool from which sarco(endo)plasmic reticulum Ca²⁺-ATPase pumps Ca²⁺ into the SR, indirectly regulating SR function.

the opening of sarcolemmal L-type Ca²⁺ channels produces subcellular Ca²⁺ signals called Ca²⁺ sparklets. They play a critical role in the regulation of intracellular Ca²⁺ concentration ([Ca²⁺]_i) in smooth muscle at physiological membrane potentials (i.e., −50 to −30 mV) and extracellular Ca²⁺ (i.e., 2 mM) (2, 22, 36). By controlling [Ca²⁺]_i, Ca²⁺ sparklets could modulate multiple physiological and pathological processes in arterial myocytes including excitability, Ca²⁺ release from intracellular stores, contraction, and gene expression (24, 26).

The release of Ca²⁺ from the SR via small clusters of Ca²⁺-activated ryanodine receptors (RyRs) in smooth muscle cells results in the generation of a local [Ca²⁺]_i signal referred to as a “Ca²⁺ spark” (25). L-type Ca²⁺ channels are known to modulate Ca²⁺ spark activity in arterial myocytes (6, 7, 10). However, unlike ventricular myocytes, where L-type Ca²⁺ channel-mediated Ca²⁺ influx (i.e., a Ca²⁺ sparklet) can activate nearby RyRs to produce a Ca²⁺ spark via a tight, local Ca²⁺-induced Ca²⁺-release mechanism (36), in smooth muscle, functional coupling of L-type Ca²⁺ channels and Ca²⁺ sparks is deemed to be “loose” or indirect (7, 10). In this model, the probability of Ca²⁺ spark activation during an L-type Ca²⁺ channel opening is low. Ca²⁺ influx via L-type Ca²⁺ channels modulates Ca²⁺ spark activity by controlling SR Ca²⁺ load. Increasing SR Ca²⁺ load increases the opening probability and Ca²⁺ flux via RyRs, resulting in higher Ca²⁺ spark amplitude and frequency. Consistent with this model, increasing the opening probability of L-type Ca²⁺ channels by the application of Bay K-8644 or membrane depolarization increases SR Ca²⁺ load and thus Ca²⁺ spark frequency and amplitude in these cells (10, 16, 17, 25).

Optical recordings of Ca²⁺ sparks and Ca²⁺ sparklets have revealed an interesting feature of the channels underlying these events: that their activity varies within arterial myocytes (14, 21–23). These regional variations in Ca²⁺ entry and the interdependency between L-type Ca²⁺ channel and SR function raise a series of important questions. What is the relationship between Ca²⁺ sparks and Ca²⁺ sparklet sites in smooth muscle? Does SR Ca²⁺ load vary locally depending on nearby Ca²⁺ sparklet activity? Do Ca²⁺ sparklets modulate the rate of refilling of SR with Ca²⁺? And, finally, does Ca²⁺ sparklet activity depend on SR Ca²⁺ load?

To address these specific questions, we developed an experimental approach to image Ca²⁺ sparklets and their relative impact on SR Ca²⁺ load and release (e.g., Ca²⁺ sparks) in cerebral arterial myocytes with high temporal and spatial resolution. We found that the SR occupies a vast area with multiple SR-sarcolemmal junctions. Our data indicate that SR Ca²⁺ content and Ca²⁺ spark activity are not modulated locally by nearby Ca²⁺ sparklets, although their activity was necessary for faster SR Ca²⁺ refilling. We also found that L-type Ca²⁺ channel activity was independent of SR Ca²⁺ load in arterial myocytes. This observation ruled out a potential feedback mechanism whereby SR Ca²⁺ load regulates L-type Ca²⁺ channel function. Our data suggest a model whereby L-type Ca²⁺ channel activity induces Ca²⁺ influx into a cytosolic Ca²⁺ pool from which sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) pumps Ca²⁺ into an SR network, in which Ca²⁺ is homogeneously distributed (20).

MATERIALS AND METHODS

Isolation of arterial myocytes.

Sprague-Dawley rats were used in this study in strict accordance with protocols approved by the Animal Care and Use Committee of the University of Washington. Rats were euthanized by an intraperitoneal injection of pentobarbital sodium (250 mg/kg ip). Myocytes were dissociated from cerebral (middle, posterior, and basilar) arteries using standard enzymatic techniques described elsewhere (3). After dissociation, cells were maintained in a nominally Ca²⁺-free Ringer solution until used. Thapsigargin (TG; 1 μM) was included in some of the solutions used to record Ca²⁺ sparklets and whole cell Ca²⁺ currents (I_Ca) to eliminate Ca²⁺ release from intracellular stores and depletes Ca²⁺ from the SR during experimentation.

Plasmid construction and preparation of adenovirus expressing ER-RFP.

Tagged red fluorescent protein (RFP) was modified to contain a calreticulin signal sequence in the NH₂-terminal and an ER retention signal, KDEL (11), in the COOH-terminal. This construct is referred to as ER-RFP. The adenovirus expressing ER-RFP was generated with the ViraPower Adenoviral Expression System using Gateway Technology following the manufacturer's guidelines (Invitrogen). The adenovirus entry clone was created by subcloning ER-RFP into the entry vector pENTR 1A. The adenovirus ER-RFP expression clone was generated via LR recombination to the destination vector pAD/CMV/DEST. Virus particles were generated by transfecting 293A cells with PacI linearized ER-RFP pAD/CMV/DEST DNA with Jet Pei. Virus particles were harvested from 293A cells exhibiting a cytopathic effect using a freeze/thaw method. The virus was tittered to determine the multiplicity of infection.

Cerebral arteries were infected with 1 ml of ER-RFP pAD diluted in 1 ml of DMEM supplemented with 10% FBS and 1% penicillin-streptomycin. Expression of ER-RFP pAD in freshly isolated arterial myocytes was evaluated 48 h postinfection (Fig. 1). To label the sarcolemma of these cells, we used Alexa 488-conjugated wheat germ agglutinin (WGA), as previously described (24). Cells were imaged (512 × 512 pixel images) using our confocal system coupled with a Nikon ×60 oil immersion lens [numerical aperture (NA): 1.4] and a zoom of 3.5 (pixel size = 0.1 μm). The point spread function of our confocal system showed that it has lateral and axial resolutions of 0.25 and 0.80 μm, respectively.

Fig. 1.

The sarcoplasmic reticulum (SR) occupies a vast area with multiple SR-sarcolemmal junctions in cerebral arterial myocytes. Shown are two-dimensional confocal images of ER-red fluoresent protein (RFP)- and wheat germ agglutinin (WGA)-associated fluorescence in representative arterial myocytes. Arrows show the proximity of the SR to the nuclear area. Shown below the ER-RFP image is a two-dimensional image produced by merging the ER-RFP and WGA images. The inset to the right of the merged image shows an enlarged image of the region in the white box. n = 6 cells.

Patch-clamp electrophysiology.

We used the conventional whole cell patch-clamp technique to control membrane voltage and to record I_Ca with an Axopatch 200B amplifier. During experiments, cells were continuously superfused with a solution containing (in mM) 120 N-methyl-d-glutamate, 5 CsCl, 1 MgCl₂, 10 glucose, 10 HEPES, and 20 CaCl₂ and adjusted to pH 7.4. Pipettes were filled with a solution composed of (in mM) 87 Cs-aspartate, 20 CsCl, 1 MgCl₂, 5 MgATP, 10 HEPES, and 10 EGTA and adjusted to pH 7.2 with CsOH. For Ca²⁺ sparklet experiments, 0.2 mM fluo-5F (or rhod-2) was added to the pipette solution. I_Ca was recorded using a step depolarization of 200 ms from a holding potential of −70 to +20 mV. I_Ca was measured as the difference between the peak and sustained currents at the end of the pulse. Under our experimental conditions, these currents have been shown to be produced by nifedipine-sensitive L-type Ca²⁺ channels. A voltage error of 10 mV, attributable to the liquid junction potential of these solutions, was corrected offline. I_Ca was sampled at 20 kHz and low-pass filtered at 2 kHz.

Total internal reflection fluorescence and confocal microscopy and signal mass analysis.

Ca²⁺ sparklets were recorded in patch-clamped (whole cell configuration) cerebral arterial myocytes as previously described (2). Cells were held at a hyperpolarized potential of −70 mV to increase the driving force for Ca²⁺ entry and thus allow the recording of quantal Ca²⁺ sparklets. We used a through-the-lens total internal reflection fluorescence (TIRF) microscope built around an inverted Olympus IX-70 microscope equipped with an Olympus PlanApo (×60, NA: 1.45) oil-immersion lens and an Andor iXON charge-coupled device camera (South Windsor, CT). To monitor [Ca²⁺]_i, cells were loaded with the Ca²⁺ indicator fluo-5F or rhod-2 (200 μM). As in previous studies (2, 22, 23), we used the relatively slow Ca²⁺ buffer EGTA (10 mM) (on rate ≈ 100-fold slower than fluo-5F or rhod-2) to restrict the fluorescence of the indicator to near the site of Ca²⁺ entry (≤1 μm) (40). A similar strategy has been used to record Ca²⁺-release sites in ventricular myocytes using confocal microscopy (1, 35). For further details about this protocol, readers are referred to Amberg et al. (2).

Images were acquired at 100–300 Hz. As previously reported (22, 23), we determined the activity of Ca²⁺ sparklets by calculating the nP_s of each Ca²⁺ sparklet site, where n is the number of quantal levels and P_s is the probability that a quantal Ca²⁺ sparklet event is active. Because Ca²⁺ sparklet activity has a bimodal distribution, Ca²⁺ sparklet sites were grouped into three categories: silent (by default having an nP_s of 0), low (nP_s between 0 and 0.2), and high (nP_s higher than 0.2). A detailed description of this analysis can be found in Navedo et al. (23).

Background fluorescence was subtracted from the total fluorescence signal of fluo-5F- or rhod-2-loaded arterial myocytes. The fluo-5F or rhod-2 fluorescence signal was then converted to Ca²⁺ concentration units using the following “F_max” equation (19):

$$\left[{\mathrm{Ca}}^{2+}\right]=K_{\mathrm{d}}\frac{\mathrm{F}/{\mathrm{F}}_{\max }-\mathrm{1/}{\mathrm{R}}_{\mathrm{f}}}{1-\mathrm{F}/{\mathrm{F}}_{\max }}$$

as previously described before (2, 23). In brief, F is fluorescence, F_max is the fluorescence intensity of fluo-5F or rhod-2 in the presence of saturating free Ca²⁺, K_d is the dissociation constant of the indicator (fluo-5F: 1,280 nM and rhod-2: 600 nM), and R_f is the F_max-to-F_min ratio of the indicators (fluo-5F: 286 and rhod-2: 150). F_min is the fluorescence intensity of the indicator in a solution where the Ca²⁺ concentration is 0. K_d and R_f values were determined in vitro using standard methods and were similar to those reported by others (38). F_max was determined at the end of each experiment by exposing cells to the Ca²⁺ ionophore ionomycin (10 μM) and 20 mM external Ca²⁺.

To label RyRs or SERCA in the SR of arterial myocytes, freshly dispersed cells were incubated with either fluorescent BODIPY FL ryanodine (10 μM) or fluorescent BODIPY FL TG (10 μM), respectively, for 20–30 min. Myocytes were then patch clamped and dialyzed with a pipette solution containing rhod-2 to monitor Ca²⁺ sparklets with our TIRF microscope. TIRF microscopy was used for these experiments as it allows the recording of Ca²⁺ sparklets in relation to subsarcolemmal RyRs and SERCA with high resolution. During analysis, RyR or SERCA images were overlaid onto the Ca²⁺ sparklet stack, and the distance between the center of the sparklet site and the RyR or Ca²⁺-ATPase pump islets was measured. To exclude nonspecific labeling of FL ryanodine (10 μM) or FL TG (10 μM), freshly isolated arterial myocytes were incubated with 10-fold higher concentrations of nonlabeled ryanodine or nonlabeled TG for 10 min before the addition of the fluorescent compounds and imaging of these cells as described above. Under these conditions, labeling of RyRs and SERCA with FL ryanodine and FL TG was significantly decreased to 4 ± 1% (n = 5 cells) and 10 ± 4% (n = 6 cells), respectively (P < 0.05; Fig. 2C).

Fig. 2.

Ca²⁺ sparklet sites coincide with junctional SR expressing sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) and ryanodine (Ry) receptors (RyRs). Shown are total internal reflection fluorescence (TIRF) images of representative arterial myocytes exposed to FL Ry (top right corner in A) or FL thapsigargin (TG; top right corner in B) to label RyRs and SERCA, respectively, in the SR. The white dotted lines represent the sarcolemmal footprint of the cells. Ca²⁺ sparklets were imaged in these cells after they were loaded with rhod-2 via a patch pipette with TIRF microscopy (top left corners in A and B). The traces below the sparklet images are time courses of intracellular Ca²⁺ concentration ([Ca²⁺]_i) in the sites indicated by the green circles. The images in the bottom right corners in A and B were produced by superimposing the FL Ry or FL TG images with the respective Ca²⁺ sparklet images. C: bar graph of the percent area (in μm²) occupied by FL Ry (n = 12 cells) or FL TG (n = 8 cells) labeling in the presence (+) or absence (−) of 10-fold higher concentrations of nonfluorescence Ry or TG in TIRF images of arterial myocytes. D: scatterplot of the distance between FL Ry or FL TG labeling and Ca²⁺ sparklet sites in arterial myocytes. The red lines indicate the median values in each group. *P < 0.05.

To assess the intraluminal SR Ca²⁺ concentration ([Ca²⁺]_SR) and Ca²⁺ sparklets simultaneously using TIRF, we loaded cells with the low-affinity Ca²⁺ indicator mag-fluo-4 (K_d= 22 μm) and the high-affinity dye rhod-2 to differentially load the SR and cytosol, respectively. Because of the thin excitation field created by TIRF (i.e., z-axis resolution of ∼100 nm), we cannot image SR Ca²⁺ content in central SR regions. To overcome this limitation, in some experiments, intraluminal [Ca²⁺]_SR was assessed using a Nikon swept-field confocal microscope. Cells were incubated with 10 μM mag-fluo-4 at 37°C for 45 min, washed with indicator-free solution, and stored at 4°C to unload the cytosolically located mag-fluo-4. Cells were then voltage clamped using the whole cell patch-clamp technique with an internal solution containing rhod-2 to monitor cytosolic Ca²⁺. Cells were illuminated with 488 nm to excite mag-fluo-4 and 568 nm to excite rhod-2. A dichroic mirror and a long-pass interference filter separated the fluorescence from both indicators. Ca²⁺ stores were identified on the basis of the response obtained from SR-loaded cells with mag-fluo-4 to a solution containing 0 mM Ca²⁺ and 10 mM caffeine to release Ca²⁺ from the SR. Under our experimental conditions, using 20 mM Ca²⁺ in the extracellular solution could lead to a larger [Ca²⁺]_SR than in the presence of 2 mM Ca²⁺. Note, however, that while the absolute levels of [Ca²⁺]_SR and rate of loading are likely to be higher in 20 mM external Ca²⁺ than in 2 mM external Ca²⁺, the experimental conditions of our experiments are not likely to alter the spatial distribution of Ca²⁺ in the SR lumen and its relationship to nearby Ca²⁺ sparklet sites.

Simultaneous imaging of Ca²⁺ sparks and sparklets was performed on isolated arterial myocytes loaded with the fluorescent Ca²⁺ indicator fluo-4 AM (50 μM) (30) using a Nikon Swept Field confocal system coupled to a Nikon TE300 inverted microscope equipped with a Nikon ×60 water-immersion lens (NA: 1.4). Images were analyzed with custom software written in IDL language (Research Systems, Boulder, CO). Ca²⁺ sparks were identified with a computer algorithm similar to the one described by Cheng et al. (5). Images were normalized by dividing the fluorescence intensity of each pixel (F) by the average resting fluorescence intensity (F₀) of a confocal image to generate an F/F₀ image. For analysis, we determined the amplitude and frequency of Ca²⁺ signals before and after the application of the RyR blocker tetracaine, the L-type Ca²⁺ channel inhibitor nifedipine, or both. Tetracaine was used for these experiments instead of ryanodine to identify Ca²⁺ sparks because it is membrane permeable and reversibly blocks RyRs in cardiac and arterial myocytes (6, 8, 28, 34).

We used the signal mass approach developed by Zou et al. (41) to determine the amount of Ca²⁺ flux associated with a Ca²⁺ spark or a Ca²⁺ sparklet before and after pharmacological manipulation. For this analysis, the total fluorescence intensity (F_total) associated with a Ca²⁺ signal is determined from raw images by summing the fluo-4 fluorescence from all the pixels within the same area of the image larger than the entire fluorescence signal produced before and after the indicated pharmacological manipulation. The change in F_total (ΔF_total) was then determined by subtracting the total fluorescence before a Ca²⁺ signal from the total fluorescence at each point during that Ca²⁺ signal progression. Signal mass was calculated by determining the peak of the integral of ΔF_total trace over time for each Ca²⁺ signal before and after pharmacological manipulation. For a more detailed description of this analysis, see Amberg et al. (2).

Chemicals and statistics.

All chemicals were purchased from Sigma-Aldrich (St. Louis, MO) unless stated otherwise. Normally distributed data are presented as means ± SE. Two-sample comparisons were made using a Student's t-test. Nonparametric statistical analyses (Mann-Whitney test) were used for nonnormally distributed data. P values of <0.05 were considered statistically significant. In the figures, asterisks indicate a significant difference between groups.

RESULTS

Arterial smooth muscle cells have an extensive SR network.

We determined the spatial distribution of the SR in freshly isolated myocytes from cerebral arteries. To do this, we imaged arterial myocytes infected with a modified tagged RFP fused with a calreticulin entrance sequence in its NH₂-terminal and a four amino acid sequence ER/SR retention signal (KDEL) in its COOH-terminal. We refer to this protein as ER-RFP. Figure 1 shows a two-dimensional confocal image of a representative arterial myocyte expressing ER-RFP. The SR appears in the images as a vast, dense, cell-wide structure that seems to extend from areas adjacent to the nucleus (arrows) to subsarcolemmal regions (n = 6 cells). WGA tagged to Alexa-488 binds sialic acid residues in glycoproteins and, thus, was used to more clearly determine the relationship between the SR and sarcolemma of arterial myocytes in cells infected with ER-RFP (Fig. 1). As previously reported for gastrointestinal and portal vein myocytes (14, 15), ER-RFP fluorescence was observed throughout the cell and in regions of the cell near the WGA fluorescence (Fig. 1, inset).

Ca²⁺ sparklet sites are located near the junctional SR expressing SERCA and RyRs.

The data described above suggest that there are multiple sites within a cell where functional interactions via Ca²⁺ signals between and among Ca²⁺ sensitive proteins in the SR and sarcolemma could take place. Thus, we examined the relationship between Ca²⁺ sparklet sites and regions of SR likely involved in Ca²⁺ uptake (i.e., expressing SERCA) and release (i.e., expressing RyRs) using a TIRF microscope. We recorded Ca²⁺ sparklets in cells that had been exposed to fluorescent FL ryanodine (10 μM) or FL TG (10 μM) to label RyRs and SERCA, respectively (Fig. 2, A and B). Consistent with other studies (12, 15, 27, 32, 33) and the confocal data above, we found in TIRF images that RyRs and SERCA were expressed broadly in regions near the sarcolemma of arterial myocytes. Indeed, quantitative analysis of FL ryanodine and FL TG images suggested that 37 ± 0.5% (n = 12 cells) and 65 ± 0.7% (n = 8 cells) of the sarcolemmal footprint was within close proximity to SR regions expressing RyRs and SERCA, respectively (Fig. 2C). In addition, we found that the majority of Ca²⁺ sparklet sites overlapped with regions of the SR expressing RyR or SERCA. The median distance between Ca²⁺ sparklet sites and RyR- or SERCA-associated fluorescence was <0.2 μm (Fig. 2D). These findings suggest that most Ca²⁺ sparklet sites are located near or at regions of the SR involved in Ca²⁺ uptake and release.

Localization of Ca²⁺ sparklets and Ca²⁺ sparks in arterial myocytes.

We investigated whether Ca²⁺ sparks occur near or at Ca²⁺ sparklet sites in cerebral arterial myocytes. Figure 3A shows a local, transient Ca²⁺ signal that was detected in close proximity to the sarcolemma of an arterial myocyte. To determine the specific contribution of Ca²⁺ sparks and Ca²⁺ sparklets to these Ca²⁺ signals, we imaged [Ca²⁺]_i before and after application of the RyR blocker tetracaine (50 μM) (8) or the L-type Ca²⁺ channel inhibitor nifedipine (10 μM) (2, 22, 23). Note that 50 μM tetracaine had no effect on whole cell L-type Ca²⁺ current density (control: −1.5 ± 0.2 pA/pF vs. tetracaine: −1.2 ± 0.2 pA/pF, P > 0.05, n = 10 cells; Fig. 3A) (8).

Fig. 3.

Ca²⁺ sparklets occur near Ca²⁺ spark sites in arterial myocytes. A, top: representative Ca²⁺ current (I_Ca) recordings under control conditions and after the application of 50 μM tetracaine. Bottom, plot of means ± SE of the peak I_Ca density before (−1.5 ± 0.2 pA/pF) and after treatment with tetracaine (−1.2 ± 0.2 pA/pF, P > 0.05, n = 10 cells). I_Ca was elicited by a step depolarization of 200 ms to +20 mV. B: representative swept-field confocal image of an arterial myocyte showing a transient [Ca²⁺]_i signal. The traces below the image show the time course of [Ca²⁺]_i in the area highlighted by the green circle during control conditions and after the application of 50 μM tetracaine and tetracaine + 10 μM nifedipine. F/F₀, fluorescence intensity/resting fluorescence intensity. C: amplitude histogram of the signal mass amplitude [change in total fluorescent intensity (ΔF_total), in arbitrary units (AU)] during control conditions (top; n = 10 cells). Bottom, graph plotting means ± SE of the signal mass amplitude of Ca²⁺ signals in control conditions and after the application of 50 μM tetracaine. D: means ± SE of the signal mass amplitude of Ca²⁺ signals before and after the application of 10 μM nifedipine (n = 11 cells).

Transient Ca²⁺ signals under control conditions had mean amplitude of 0.7 ± 0.07 ΔF/F₀ and a mean frequency of 0.8 ± 0.08 Hz (Fig. 3B, bottom). In the presence of tetracaine, these Ca²⁺ signals were significantly smaller in amplitude (0.2 ± 0.01 ΔF/F₀) than under control conditions (P < 0.05; Fig. 3B, bottom). Tetracaine-insensitive events were subsequently eliminated by the simultaneous application of 10 μM nifedipine + tetracaine in the solution (Fig. 3B, bottom). We constructed an amplitude histogram of the signal mass (ΔF_total) of these Ca²⁺ signals. Under control conditions, the histogram could be fit with the sum of three Gaussian functions with centers (i.e., mean) at 0.8 × 10⁵, 1.6 × 10⁵, and 3.0 × 10⁵ ΔF_total (χ² = 2.4, n = 10 cells; Fig. 3C, top). In the presence of tetracaine, only events with a mean signal mass amplitude of 0.8 × 10⁵ ΔF_total were observed (χ² = 0.2; Fig. 3C, bottom). These events were eliminated with the simultaneous application of tetracaine + nifedipine. These results suggest that tetracaine-insensitive events are produced by Ca²⁺ sparklets, whereas events with a significantly larger signal mass, which were tetracaine sensitive, are produced by Ca²⁺ sparks in the same microdomain (Fig. 3B).

A testable prediction of this hypothesis is that treatment of arterial myocytes with nifedipine should preserve tetracaine-sensitive events. Consistent with this, the application of 10 μM nifedipine eliminated events with a mean signal mass amplitude of 0.9 × 10⁵ΔF_total and slightly decreased the mean amplitude of remaining events from 1.4 × 10⁵ and 2.9 × 10⁵ ΔF_total to 1.0 × 10⁵ and 1.9 × 10⁵ ΔF_total, respectively (n = 11 cells; Fig. 3D). The decrease in the mean signal mass amplitude of nifedipine-insensitive events (i.e., Ca²⁺ sparks), which were abolished by tetracaine, is likely due to a decrease in SR Ca²⁺ load (6, 7, 10). Taken together, these data indicate that although Ca²⁺ sparklets occur near Ca²⁺ spark sites, they are not directly required for Ca²⁺ spark ignition in cerebral arterial myocytes.

SR Ca²⁺ load does not vary regionally with nearby Ca²⁺ sparklet activity.

A mechanism by which Ca²⁺ sparklets could potentially modulate Ca²⁺ spark activity is by controlling SR Ca²⁺ load, especially in areas near sparklet sites. To investigate this possibility, we simultaneously monitored SR Ca²⁺ and Ca²⁺ sparklet activity using TIRF microscopy. TIRF allowed us to detect Ca²⁺ sparklets and measure potential relative changes in near-membrane SR Ca²⁺ content with a high signal-to-noise ratio. The SR of arterial myocytes was selectively loaded with the low-affinity Ca²⁺ indicator mag-fluo-4 (see materials and methods). Cells were then patch clamped in the whole cell configuration to allow for the dialysis of mag-fluo-4, limiting the fluorescence of this fluorophore to the SR lumen.

To ensure SR loading, cells were patch clamped with a pipette solution containing 100 nM free [Ca²⁺]_i. Rhod-2 was also included in this solution to image cytosolic Ca²⁺ signals. To prevent Ca²⁺ sparks (i.e., Ca²⁺-release events via RyRs located in the SR), cells were continually perfused with the RyR blocker tetracaine (50 μM). As shown in Fig. 4A, cells subjected to this protocol had well-defined zones of mag-fluo-4 fluorescence. The application of a solution containing 10 mM caffeine and no tetracaine resulted in a decrease in mag-fluo-4 fluorescence, thus confirming that these signals arose from Ca²⁺-bound mag-fluo-4 in the lumen of the SR (n = 14 cells; Fig. 4A, bottom).

Fig. 4.

Ca²⁺ sparklet activity does not contribute to regional variations in SR Ca²⁺ concentration ([Ca²⁺]_SR). A, top: representative TIRF image of an arterial myocyte in which the SR was differentially loaded with the slow fluorescence indicator mag-fluo-4 to monitor [Ca²⁺]_SR. Bottom, time course of [Ca²⁺]_SR in the region indicated by the green circle in the TIRF image above and the time course of [Ca²⁺]_SR obtained from a different arterial myocyte using a confocal microscope. The TIRF [Ca²⁺]_SR image was obtained at the time point indicated by the arrow. The dotted line indicates the sarcolemmal footprint obtained with TIRF. B: amplitude histogram of [Ca²⁺]_SR [in F/fluorescent intensity of caffeine (F_caff) units] in multiple SR regions obtained from cells imaged with TIRF (left; n = 14 cells) or confocal (right; n = 7 cells) microscopy. Each histogram could be fit with a single Gaussian function (solid line) with a center at 2.2 ± 0.09 F/F_caff (TIRF) and 2.6 ± 0.2 F/F_caff (confocal). C: bar graph of [Ca²⁺]_SR in areas with and without nearby (≤1 μm) Ca²⁺ sparklet activity (P > 0.05).

The amplitude distribution of Ca²⁺ pools within the SR was assessed by fitting an amplitude histogram of [Ca²⁺]_SR [in F/caffeine fluorescent intensity (F_caff) units] in multiple SR regions, as shown in Fig. 4B. The histogram could be fit (χ² = 2.1) with a Gaussian function with a center at a [Ca²⁺]_SR of 2.2 ± 0.09 F/F_caff. This analysis suggests that the [Ca²⁺]_SR distribution within arterial myocytes has a single mode, indicating that [Ca²⁺]_SR is similar throughout the SR in these cells. Similar results were obtained in mag-fluo-4-loaded cells imaged with a confocal microscope (2.6 ± 0.17 F/F_caff, χ² = 2.8, n = 7 cells; Fig. 4B, right). Consistent with the data described above, we found that [Ca²⁺]_SR was similar in SR regions independently of their association or not with Ca²⁺ sparklet sites (P > 0.05; Fig. 4C).

Ca²⁺ sparklets increase the rate of SR Ca²⁺ refilling in arterial myocytes.

Having established that Ca²⁺ sparklets are not associated with significant variations in local SR Ca²⁺ load in areas close to these Ca²⁺ signals, we investigated whether they influence the rate of loading of the SR with Ca²⁺. In these experiments, we imaged [Ca²⁺]_SR before, during, and after the application of 10 mM caffeine using TIRF (Fig. 5A). [Ca²⁺]_SR decreased rapidly after the application of 10 mM caffeine but recovered back to control levels after the removal of caffeine from the extracellular solution (Fig. 5). The recovery phase of [Ca²⁺]_SR records could be fit with a single exponential function with a time constant of recovery of 38 ± 2 s (n = 11 cells; Fig. 5B). The rate of recovery of [Ca²⁺]_SR after caffeine at multiple SR sites was similar whether or not they were within 1 μm from an active Ca²⁺ sparklet site (P > 0.05; Fig. 5C). In cells where L-type Ca²⁺ channel activity was inhibited with nifedipine, and consequently low activity and persistent Ca²⁺ sparklet events were blocked (2), the rate of recovery of [Ca²⁺]_SR was nearly 12-fold slower (638 ± 170 s, n = 5 cells; Fig. 5D) than under control conditions. This is consistent with the view that Ca²⁺ sparklets contribute to faster SR Ca²⁺ refilling. Together, these findings suggest that while Ca²⁺ sparklets do not contribute to the creation of a localized, heterogeneous SR Ca²⁺ gradient in areas near these events, they are required for faster Ca²⁺ refilling of the SR after it has been depleted.

Fig. 5.

Ca²⁺ sparklets increase the rate of SR Ca²⁺ load in arterial myocytes. A: representative time course of [Ca²⁺]_SR before, during, and after the application of a solution containing 0 mM external Ca²⁺ + 10 mM caffeine to deplete the Ca²⁺ from the SR. The red line indicates the best fit of the recovery phase of [Ca²⁺]_SR with a single exponential function. B: histogram of the rate of [Ca²⁺]_SR recovery after the application of caffeine. This histogram was fitted with a single exponential function with a time constant of recovery (τ_recovery) of 38 ± 2 s (n = 11 cells). C: bar graph of the mean ± SE rate of [Ca²⁺]_SR recovery in areas with and without Ca²⁺ sparklet activity (P > 0.05). D: nifedipine decreased the rate of [Ca²⁺]_SR recovery after the application of caffeine (n = 5 cells).

Ca²⁺ sparklet activity is independent of SR Ca²⁺ content.

Recent studies (29, 37) have reported significant inhibition of L-type Ca²⁺ channel activity after SR Ca²⁺ depletion. Thus, using the approach described above, we determined whether Ca²⁺ sparklet activity depends on [Ca²⁺]_SR in freshly isolated cerebral arterial myocytes. A testable prediction of the aforementioned studies is that depletion of SR Ca²⁺ load should inhibit Ca²⁺ sparklet activity in arterial myocytes. To test this hypothesis, we recorded Ca²⁺ sparklet activity before, during, and after depletion of SR Ca²⁺ load (Fig. 6). We found that Ca²⁺ sparklet activity did not change (P > 0.05) with varied [Ca²⁺]_SR (n = 9 cells; Fig. 6A). Indeed, our data suggest that there is poor correlation (r² = 0.08) between Ca²⁺ sparklet activity (nP_s) and [Ca²⁺]_SR in arterial myocytes.

Fig. 6.

Ca²⁺ sparklet activity is independent of SR Ca²⁺ load in arterial myocytes. A: scatterplot of Ca²⁺ sparklet activity at different [Ca²⁺]_SR (n = 9 cells). Note that Ca²⁺ sparklet activity was poorly correlated with SR Ca²⁺ content (r² = 0.08). n, Number of quantal levels; P_s, probability that a quantal Ca²⁺ sparklet is active. B: amplitude histogram of [Ca²⁺]_SR (in ΔF/F_caff units) in multiple SR regions obtained from cells treated with TG to deplete the SR of Ca²⁺ (n = 7 cells). The histogram could be fit with a single Gaussian function (solid line) with a center at 0.4 ± 0.2 ΔF/F_caff (χ² = 3.1). The dotted line represents the [Ca²⁺]_SR distribution obtained from cells that were not exposed to TG, as in Fig. 4B, right. C, left: representative I_Ca recordings under control conditions and after SR depletion with 1 μM TG. Right, plot of the means ± SE of the peak I_Ca density (in pA/pF, at +20 mV) before and after SR depletion with 1 μM TG (n = 4 cells, P > 0.05).

To determine the implications of these results on cell-wide Ca²⁺ influx, we recorded whole cell I_Ca from freshly isolated arterial myocytes before and after depletion of SR Ca²⁺ with 1 μM TG. For analysis, we only used arterial myocytes that maintained their characteristic spindle-shaped morphology after SR depletion. The amplitude histogram of [Ca²⁺]_SR (in ΔF/F_caff units) shown in Fig. 6B demonstrates that [Ca²⁺]_SR was significantly decreased in cells treated with TG compared with untreated cells (Fig. 4A), thus confirming that this treatment depletes SR Ca²⁺ stores (n = 7 cells). Treatment of cells with TG did not induce a change in peak I_Ca density (−0.31 ± 0.06 pA/pF) compared with control peak I_Ca (−0.29 ± 0.05 pA/pF, P > 0.05, n = 4 cells; Fig. 6C). Collectively, these findings strongly suggest that L-type Ca²⁺ channel activity does not depend on [Ca²⁺]_SR in arterial myocytes.

DISCUSSION

In this study, we used confocal and TIRF microscopy to investigate with high temporal and spatial resolution the relationship between Ca²⁺ sparklets and SR function in arterial myocytes. We report four fundamental observations. First, Ca²⁺ sparklets are located close to SR regions expressing RyRs and SERCA. Second, consistent with these data, Ca²⁺ sparklets occur near Ca²⁺ spark sites in arterial myocytes. Third, local SR Ca²⁺ load and recovery rate do not vary regionally due to the activation of nearby Ca²⁺ sparklets. Yet, sparklets are necessary for faster overall SR Ca²⁺ replenishment. To our knowledge, this is the first time that [Ca²⁺]_SR dynamics and Ca²⁺ sparklet activity have been directly and simultaneously visualized in freshly isolated cerebral arterial myocytes. Finally, our data suggest that L-type Ca²⁺ channel activity is independent of [Ca²⁺]_SR content. This observation is important because it rules out any potential feedback mechanism whereby SR Ca²⁺ load regulates the activity of these channels in arterial myocytes.

Our data suggest that the SR of living cerebral artery smooth muscle is a vast, convoluted, nonuniform organelle that extends throughout the cell and approaches the sarcolemma in multiple regions of the cell. Similar findings have been reported in cultured and freshly isolated arterial myocytes from the gastrointestinal tract, portal vein, and aorta using electron (9) and confocal (12, 13, 15, 32, 39) microscopy. Together with these studies, our data suggest that this SR organization is a conserved feature in smooth muscle from different tissues that creates areas of specialized communication between and among Ca²⁺-sensitive proteins (i.e., L-type Ca²⁺ channels, RyRs, and SERCA) located in the sarcolemma and SR.

The issue of whether L-type Ca²⁺ channels can activate RyRs in the SR that comes into close (10–30 nm) apposition to the surface membrane (i.e., junctional SR) of arterial myocytes has been extensively examined (7, 10, 18). Using a combination of patch-clamp electrophysiology and line-scan confocal microscopy, Kotlikoff's laboratory (7) found that membrane depolarization evoked L-type Ca²⁺ currents and increased the probability of Ca²⁺ spark occurrence. Interestingly, they observed that in most cases, the activation of Ca²⁺ sparks occurred long (>10 ms) after the onset of depolarization and resultant Ca²⁺ currents. On the basis of these data, a model was proposed in which RyRs and L-type Ca²⁺ channels were “loosely” coupled in arterial myocytes. This model is supported by recent data from Moosmang's laboratory (10) demonstrating that Ca²⁺ spark frequency and amplitude are significantly reduced, but not eliminated, in arterial myocytes from a mouse with the Ca_v1.2 gene inactivated. These findings suggest that local, tight coupling between L-type Ca²⁺ channels and RyRs is not necessary for the ignition of Ca²⁺ sparks. However, these results did not address the issue of whether Ca²⁺ sparks and Ca²⁺ sparklets can coexist in the same microenvironment in arterial myocytes.

The data in this study provide compelling evidence in support of the hypothesis that Ca²⁺ sparklets could occur near Ca²⁺ spark sites in cerebral arterial myocytes. First, Ca²⁺ sparklets can occur in regions expressing RyRs in the SR of arterial mycoytes (Fig. 2). Second, the signal mass of Ca²⁺ signals revealed three prominent peaks, two of which were completely eliminated after the application of the RyR blocker tetracaine. The remaining peak seems to correspond to the activity of Ca²⁺ sparklets, since it was eliminated by application of the L-type Ca²⁺ channel blocker nifedipine in the presence of tetracaine. Third, consistent with the “loose coupling” model above, the signal masses of Ca²⁺ signals that seem to correspond to Ca²⁺ sparks were not abolished by nifedipine. Indeed, nifedipine abolished events with a smaller signal mass amplitude but only slightly decreased the amplitude of the remaining larger events. Together, these results suggest that Ca²⁺ sparks and Ca²⁺ sparklets can coexist in the same microenvironment, where Ca²⁺ sparklets can indirectly modulate RyRs activity via the regulation of [Ca²⁺]_SR by cytosolic [Ca²⁺]_i (7, 10).

Our [Ca²⁺]_SR data are consistent with recent studies from the Bers (31) and McCarron (20) laboratories in cardiac and portal vein myocytes that suggest that the SR is an interconnected organelle in which Ca²⁺ can freely diffuse. We found that the [Ca²⁺]_SR distribution was unimodal (Fig. 4, B and C) with a relatively low coefficient of variation (defined as SD/mean) of 0.68 for TIRF data and 0.65 for confocal data. Interestingly, [Ca²⁺]_SR was similar in regions of the SR near high-activity Ca²⁺ sparklet sites and regions of the cells without visible Ca²⁺ sparklet activity. Based on these data, we propose a model for the contribution of Ca²⁺ sparklets to [Ca²⁺]_SR in cerebral arterial myocytes. In this model, Ca²⁺ sparklets contribute to a general cytosolic pool of Ca²⁺ from which the SR can draw to accelerate store refilling. [Ca²⁺]_SR could be transiently higher in areas associated with storage (e.g., calsequestrin), uptake (e.g., SERCA pumps), or release (e.g., RyRs or inositol 1,4,5-trisphosphate receptors), but normally will exist in a free diffusional equilibrium throughout the SR of cerebral arterial myocytes, which is not influenced by the activity of nearby Ca²⁺ sparklets. Future studies should examine this model.

An important observation in the present study is that the activity of Ca²⁺ sparklets is not dependent on [Ca²⁺]_SR. Recent data suggest that stromal interaction molecule 1 (STIM1), which is the main activator of stored-operated channels, directly binds and modulates L-type Ca²⁺ channel activity in rat cortical neurons and cultured A7r5 smooth muscle cells, respectively (29, 37). These two studies have reported significant inhibition of L-type Ca²⁺ channel activity after SR Ca²⁺ depletion, in stark contrast with our results. While the reasons for these opposing findings are unclear, it is possible that variation in the expression of STIM1 in cultured versus freshly isolated cells (used in this study) may contribute to differences in the modulation of L-type Ca²⁺ channels after SR Ca²⁺ depletion. Indeed, a recent report (4) has suggested that STIM1 is abundantly expressed in cultured cells, but very low levels were detected in freshly isolated arterial myocytes. Alternatively, it is possible that L-type Ca²⁺ channels are not regulated by STIM1 after SR Ca²⁺ depletion in cerebral arterial myocytes or that differences in experimental conditions (e.g., different cell types, recording conditions, or depletion methods) may account for the apparent disparities. Regardless, our results rule out a potential feedback mechanism whereby SR Ca²⁺ load regulates sparklet activity in cerebral arterial myocytes. Thus, the physiological relevance of STIM1 regulation of L-type Ca²⁺ channels during SR Ca²⁺ depletion in cerebral arterial myocytes is unclear.

Limitations of the study.

Some of the experimental conditions used in this study could induce changes in [Ca²⁺]_SR that are different to those in arterial myocytes in vivo. For example, as Jaggar's laboratory (6) has shown, incubation of cerebral arteries with tetracaine increases [Ca²⁺]_SR. In addition, the relatively high extracellular [Ca²⁺] (>2 mM) used in some of the experiments in this study may also contribute to elevated [Ca²⁺]_SR. Thus, we cannot rule out the possibility that the absolute levels of [Ca²⁺]_SR and rates of loading could be different in vivo than under our experimental conditions. Despite these limitations, we contend that novel insights into the bases of SR Ca²⁺ accumulation and release and its relationship to Ca²⁺ influx via L-type Ca²⁺ channels have been obtained in the present study. These insights should form the bases of new hypotheses that can be best tested in appropriate animal models.

Conclusions.

In summary, we have provided the first examination of the relationship between Ca²⁺ sparklets and [Ca²⁺]_SR dynamics in freshly dissociated cerebral arterial myocytes. Our findings indicate that the SR in cerebral arterial myocytes extends throughout the cell and lines the internal side of the sarcolemma in specialized areas where it can promote functional interactions between Ca²⁺-sensitive proteins in these two organelles. Ca²⁺ sparklet activity was not associated with the creation of a localized, heterogeneous SR Ca²⁺ gradient, but their activity was necessary for faster SR Ca²⁺ refilling, perhaps due to their influence on steady-state Ca²⁺ entry and global [Ca²⁺]_i (2). Our data provide important evidence indicating that L-type Ca²⁺ channel activity is not modulated by SR Ca²⁺ load in arterial myocytes. Finally, the methods and approaches used here may help determine the mechanisms of regulation of SR function by L-type Ca²⁺ channels during pathological conditions.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute (NHLBI) Grants HL-098200 (to M. F. Navedo) and HL-085686 and HL-085870 (to L. F. Santana) and by American Heart Association Grant 0735251N (to M. F. Navedo). M. A. Nystoriak was supported by the National Institutes of Health-University of Washington Cardiovascular Research Training Program (NHLBI Grant T32-HL-07828). L. F. Santana is an Established Investigator of the American Heart Association.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: Y.T., M.A.N., M.N.-C., and M.F.N. performed experiments; Y.T., M.A.N., and M.F.N. analyzed data; Y.T., M.A.N., M.N.-C., L.F.S., and M.F.N. edited and revised manuscript; Y.T., M.A.N., M.N.-C., L.F.S., and M.F.N. approved final version of manuscript; M.A.N., M.N.-C., L.F.S., and M.F.N. interpreted results of experiments; M.A.N. and M.F.N. prepared figures; L.F.S. and M.F.N. conception and design of research; M.F.N. drafted manuscript.