|Home | About | Journals | Submit | Contact Us | Français|
Recent data from mesenteric and cerebral beds have revealed spatially restricted Ca2+ transients occurring along the vascular intima that control effector recruitment and vasodilation. While Ca2+ is pivotal for coronary artery endothelial function, spatial and temporal regulation of functional Ca2+ signals in the coronary endothelium is poorly understood.
We aimed to determine whether a discrete spatial and temporal profile of Ca2+ dynamics underlies endothelium-dependent relaxation of swine coronary arteries (SCA).
Using confocal imaging, custom automated image analysis, and myography we show that the SCA endothelium generates discrete basal Ca2+ dynamics including isolated transients and whole-cell propagating waves. These events are suppressed by depletion of internal stores or inhibition of inositol 1,4,5-trisphosphate receptors (IP3R), but not by inhibition of ryanodine receptors or removal of extracellular Ca2+. In vessel rings, inhibition of specific Ca2+-dependent endothelial effectors, namely small and intermediate conductance K+ channels (KCa3.1, KCa2.3) and endothelial nitric oxide synthase (eNOS), produces additive tone, which is blunted by internal store depletion or IP3R blockade. Stimulation of endothelial IP3-dependent signaling with substance P causes idiosyncratic changes in dynamic Ca2+ signal parameters (active sites, event frequency, amplitude, duration, and spatial spread). Overall, substance P-induced vasorelaxation corresponded poorly with whole-field endothelial Ca2+ measurements, but corresponded precisely with the concentration-dependent change in Ca2+ dynamics (linearly translated composite of dynamic parameters).
Our findings show that endothelium-dependent control of SCA tone is determined by spatial and temporal titration of inherent endothelial Ca2+ dynamics that are not represented by tissue-level averaged Ca2+ changes.
The endothelium is a crucial regulator of coronary blood flow, and endothelial dysfunction is a hallmark of vasospasm and coronary artery disease (CAD).1–5 Endothelial vasoregulation occurs through graded recruitment of various Ca2+-dependent cellular effectors, including endothelial nitric oxide synthase (eNOS) and small/intermediate conductance Ca2+-activated K+ (KCa2.3 and KCa3.1) channels.6–8 Endothelial NOS and KCa3.1/2.3 channels respond to Ca2+-mobilizing stimuli such as laminar shear stress and local mediators (e.g. acetylcholine, bradykinin and substance P),9–13 eliciting nitric oxide (NO) production and endothelium derived hyperpolarization (EDH), respectively. Both NO and EDH promote relaxation of underlying vascular smooth muscle causing vasodilation, and are key controllers of blood pressure and flow14–17 including real-time regulation of coronary perfusion.12,18 Arachidonic acid metabolites19 such as prostacyclin (PGI2) and epoxyeicosatrienoic acids (EETs),20–22 as well as hydrogen peroxide (H2O2),23 have also been implicated in endothelial responses to local stimuli and shear stress. The endothelium plays a particularly important role in the autoregulated coronary circulation by supporting metabolic dilation and moderating myogenic constriction to help meet the blood flow demands of the heart. This crucial role is highlighted during atherosclerosis wherein the progressive inability of the endothelium to override arterial tone exacerbates myocardial ischemia. Multiple Ca2+-dependent pathways control endothelial regulation of coronary artery function at the macro-and microvascular level, but few studies have directly assessed the underlying endothelial Ca2+ signaling in these vessels and none have addressed the spatial and temporal regulation of these signals. Hence, our understanding of the specificity and graded recruitment of endothelial influence remains very limited.
Recent studies suggest that local transient endothelial Ca2+ changes are sufficient to direct functional responses in intact blood vessels.17,24–28 In fact, evidence from mouse mesenteric,24,26 rat cerebral,27 and hamster skeletal muscle29 arteries indicates physiologic endothelial signaling involves spatially and temporally restricted Ca2+ transients that are frequency-tuned by the type and level of stimulation. In mouse mesenteric arteries, basal Ca2+ events emit repetitively from distinct sites along the endothelium, originating from distinct clusters of inositol 1,4,5-trisphosphate receptors (IP3Rs) on the endoplasmic reticulum (ER) membrane.24 These spatially restricted dynamics serve as a constant vasorelaxing impetus by activating nearby KCa3.1 channels and eliciting EDH, particularly at myoendothelial projections (MEPs) where portions of endothelial cell membranes protrude through holes in the internal elastic lamina to form close associations or gap junction communication with smooth muscle cells.30,31 In addition, direct endothelial stimulation can increase local Ca2+ influx events through transient receptor potential channels (e.g. TRPV4 or TRPA1) on the plasma membrane.26,27,32 This Ca2+ entry also targets endothelial KCa channels (KCa3.1 and KCa2.3), and our recent findings suggest this KCa activation may further expand Ca2+ dynamics by promoting additional influx.28 Overall, it is implied that sensitive and adjustable scenarios for vasoregulation depend on highly dynamic endothelial Ca2+ signaling patterns. The role of such dynamic Ca2+ signaling in the coronary artery endothelium remains unknown.
While focus has clearly shifted to elucidation of complex Ca2+ signals within the vasculature, discerning and quantifying complex dynamic Ca2+ signals in intact tissues is challenging. We recently developed and implemented a custom analysis module, LC_Pro33 that allows for definitive tracking of a wide range of Ca2+ dynamics, including local transients and waves, within broad cellular fields while avoiding user error or bias. Here, we employ this discriminating and comprehensive Ca2+ analysis approach to assess functional Ca2+ dynamics in the SCA endothelium.
Domestic male and female juvenile pigs were sedated with ketamine and xylazine and subsequently euthanized with sodium pentobarbital (58 mg/kg, i.v.) followed by thoracotomy as approved by the University of South Alabama Institutional Animal Care and Use Committee. Branches of the left anterior descending artery were dissected from the right ventricle in cold HEPES/bicarbonate-buffered physiological saline solution (PSS).
Coronary artery segments were carefully opened longitudinally, pinned luminal-side-up on small silicone (Sylgard) blocks, loaded with the Ca2+ indicator Fluo-4 AM, and placed in a HEPES PSS-containing chamber as previously described.28 Ca2+-dependent fluorescence was measured with a spinning-disk laser confocal (8 frames/sec). Fluorescence data was processed using the custom ImageJ plug-in, LC_Pro, which is specifically designed to 1) detect and track sites of dynamic Ca2+ change above statistical noise (p<0.01), 2) define regions of interest (ROI; 5 μm diameter) at active site centers and 3) analyze average fluorescence intensities within ROIs. For assessments before and after drug addition, the brief period of bath change (~2 seconds) was omitted from analysis to avoid movement artifact. Fluorescence data is expressed as F/F0, where F0 is determined by a linear regression of base data at each ROI. Some preparations were co-loaded with Fluo-4 AM and DAR-4M AM and imaged simultaneously in order to track both Ca2+ and NO-dependent fluorescence.
Opened arteries pinned to Sylgard blocks were fixed, permeabilized, and stained as previously described.24 Data were obtained using a Nikon A1 confocal microscope (NIS-Elements and ImageJ software).
Artery rings (~500 μm diameter) were mounted in an isometric force myograph (), stretched to optimal length and allowed to equilibrate for 30 minutes before being exposed to indicated protocols Artery viability was tested using 60 mmol/L KCl. For some experiments, 1–10 nmol/L of the thromboxane analog U44619 was included in the bath to establish stable tone (~2 mN). In other experiments assessing acute relaxations, vessels were pre-contracted with 10–100 nmol/L U44619. All drugs were added directly to the bath and data was recorded using Chart software at 3 hz.
Data were expressed as means ± SEM. Group data were subjected to one-way analysis of variance (ANOVA) and individual comparisons were made by Tukey post-test. Parameter data was analyzed using ANOVA with linear mixed effects model. P values less than 0.05 were considered statistically significant. An expanded Materials and Methods section can be found in the Online Supplement.
We performed confocal imaging on opened, branches of swine coronary arteries loaded with Fluo-4 in order to characterize basal endothelial Ca2+ dynamics. Survey of the intima layer under basal conditions (~120 endothelial cells per field assessed at 8.1 frames/sec with no flow), revealed a variety of basal Ca2+ events, ranging from isolated transients (<5 μm2) to broad cellular and multi-cellular waves (>30 μm2) (Figure 1). Events were detected in approximately 35% of the sampled endothelial cells over a 10-minute period (Figure 1A). As depicted in Figure 1B, basal Ca2+ events were considerably variable in size and duration. Localized events tended to originate from 1–3 discrete sites per cell. In most cases, local signals triggered waves that propagated along the long axis of the cell, with the largest Ca2+ increase occurring at or near the nucleus. Cumulative assessment of basal Ca2+ in a single cell over 10 minutes shows discrete regional and focal heterogeneity of aggregate signal. Figure 2A shows a recording of the Ca2+ dynamics occurring within the endothelial field depicted in Figure 1A. Notably, despite the disparate dynamics, the average Ca2+ signal within the whole field remained essentially constant over the full recording (Figure 2A, red line) due to the asynchronous nature of the signals. A summary of Ca2+ parameters obtained from automated analysis (LC_Pro) of eight arteries from eight different animals is shown in Figure 2B. On average, 10.2 ± 1.6 events occurred at 8.9 ± 0.6 distinct sites each minute per field. Individual event parameters (amplitude, duration and spatial spread) all exhibited positively skewed distributions, with mean values of 1.5 ± 0.02 F/F0, 13.0 ± 0.4 sec, and 33.9 ± 2.6 μm2, respectively. Properties of basal Ca2+ dynamics are summarized in Online Table I.
Next, we assessed the intracellular or extracellular source(s) of the basal endothelial Ca2+ events (Figure 3). Replacement of the chamber bath with Ca2+-free PSS for 5 minutes had no significant effect on the occurrence of Ca2+ events, suggesting minimal contribution of extracellular Ca2+ influx (Figure 3A and B). However, depleting internal endoplasmic reticulum Ca2+ stores using the SERCA inhibitor cyclopiazonic acid (CPA; 10 μmol/L for 15 minutes) suppressed basal events by more than 90%. While blockade of ryanodine receptors (RyR) with 10 μmol/L ryanodine had no appreciable effect on the Ca2+ events, inhibition of IP3Rs with either 30 μmol/L xestospongin C or 100 μmol/L 2APB greatly attenuated their occurrence. Notably, similar blunting of Ca2+ events was observed following inhibition of phospholipase C (PLC) with U73122. Immunostaining performed on open vessel preparations revealed a distinctive distribution of IP3Rs around the nucleus and along the axis of endothelial cells (Figure 3C). Figure 3D shows an endothelial cell within an open artery preparation that was probed for IP3R following Ca2+ imaging. Notably, the regional intensity of cytosolic Ca2+ signal associated with a single wave corresponds with regional IP3R density. Together, these data suggest that basal IP3-induced Ca2+ release from the endoplasmic reticulum is the predominant mechanism contributing to basal Ca2+ dynamics in SCA endothelium.
In order to assess whether the basal endothelial Ca2+ signals along the endothelium provide an impetus for persistent modulation of coronary artery tone through activation of Ca2+-dependent endothelial effectors, we performed isometric force measurements in isolated artery rings (Figure 4A and B). While blockade of cyclooxygenase with indomethacin (10 μmol/L) or removal of H2O2 with PEG-catalase (500 U/ml) had no effect on tone (Online Figure I), inhibition of eNOS with NG-nitro-L-arginine (LNNA; 200 μmol/L) caused substantial contraction (16.7 ± 1.7 mN) as did blockade of both KCa2.3 and KCa3.1 channels with apamin (Apa; 0.5 μmol/L) and charybdotoxin (Chtx; 0.1 μmol/L) (3.4 ± 0.6 and 5.1 ± 0.7 mN, respectively). The effects of LNNA, Apa and Chtx were additive, resulting in a net increase in tone equivalent to that elicited by direct depolarization with 60 mmol/L KCl. Changing the order of application did not influence the individual impact of each drug and the cumulative contractions could be effectively relaxed by treatment with the NO donor sodium nitroprusside (SNP, 1 μmol/L), the KATP channel opener pinacidil (1 μmol/L), or the voltage-gated Ca2+ channel blocker nifedipine (1 μmol/L) (Online Figure II), indicating preservation of smooth muscle function, including direct NO- and hyperpolarization-dependent relaxation mechanisms. Importantly, we found that physical disruption of the endothelium caused an increase in resting tone (68 ± 6 % of maximal KCl response; Online Figure III) and prevented responses to LNNA, Apa, and Chtx. Although Chtx can inhibit large conductance Ca2+-activated K+ (BK) channels as well as KCa3.1 channels, we found that the selective BK channel blocker iberiotoxin (0.1 μmol/L) had no significant effect on either arterial tone or Chtx-induced contractions (Online Figure IV). TRAM-34 was not employed as a KCa3.1 channel inhibitor in the current study since it was found to elicit relaxation of SCAs through an endothelium-independent mechanism.34 Immunostaining performed in open arteries revealed expression of eNOS, KCa2.3, and KCa3.1 within the coronary artery endothelium (Figure 4C) with eNOS primarily distributing along the peripheral plasma membrane and near the nucleus in densities corresponding with the Golgi apparatus. KCa3.1 largely concentrated in focal plaques and KCa2.3 was found primarily along the endothelial cell-endothelial cell borders. Tracking Ca2+ and NO-dependent fluorescence (DAR-4M AM) simultaneously revealed spatial and temporal correspondence of local Ca2+ elevations and NO production at both central and peripheral regions of the cell (Figure 4D).
In order to determine whether the persistent modulation of tone by eNOS and/or KCa channels is directly dependent on ongoing basal endothelial Ca2+ signals, we performed additional myography (Figure 5A and B). While inhibition of RyRs (ryanodine, 10 μmol/L) had no significant effect on the contractions induced by Apa and Chtx or LNNA, blockade of IP3Rs (xestospongin C, 30 μmol/L or 2APB, 100 μmol/L) or PLC (U73122 10 μmol/L) significantly reduced contractions elicited by all three inhibitors (p<0.05 vs control, n = 6–7) compared to KCl. Overall, inhibition of IP3-dependent signaling nearly abolished Apa/Chtx-induced contractions and significantly attenuated LNNA-induced contractions. Ca2+-free solution could not be employed for these experiments since it directly undermines smooth muscle contraction; however, inhibition of nonselective cation channels with Gd3+ (100 μmol/L) had little effect on Apa/Chtx or LNNA-induced contractions, suggesting minimal influence of extracellular Ca2+ influx on KCa channel or eNOS modulation of tone under the conditions studied.
Substance P is a potent endothelium-dependent dilator in the coronary circulation. It acts through Gq protein-coupled receptor stimulation, which leads to phospholipase C (PLC) dependent IP3 elevation and Ca2+ mobilization. Figure 6A shows representative concentration-dependent substance P relaxations of U46619 pre-contracted SCAs. These relaxations were greatly impaired or abolished by pretreatment with U73122. Notably, relaxation to high substance P concentration (30 nmol/L) was blocked by the combination of LNNA, Apa and Chtx, but treatment with LNNA alone or the combination of Apa and Chtx had minimal effect. On the other hand, relaxation to low substance P concentration (0.03 nmol/L) could be effectively blocked with LNNA but not with Apa and Chtx (Online Figure V), suggesting preferential eNOS recruitment with low-level stimulation. In order to assess the influence of substance P on existing endothelial Ca2+ dynamics, we performed Ca2+ imaging experiments on open coronary arteries (see Online Supplement movie files). Concentration-dependent effects of substance P on dynamic Ca2+ signals and related parameters are shown in Figure 6B and summarized in Figure 7. Stimulation with substance P (10−13 – 10−6 mol/L) increased the occurrence of discrete endothelial Ca2+ dynamics in a concentration-dependent manner. At low concentrations (≤ 10−10 mol/L), substance P primarily increased the number of Ca2+ events while having little impact on the properties of the events themselves (i.e. amplitude, duration, and spread). This increase in events was mainly due to recruitment of new active sites (i.e. events firing at distinct spots or cells that were not active before substance P exposure; Figure 6B right panels). Importantly, evaluation of the whole endothelial field (Figure 6B; red line) rather than individual ROIs indicated little or no net Ca2+-dependent fluorescence change at 10−10 mol/L substance P although relaxations were measurable. At higher substance P concentrations (>10−10 mol/L), the number of Ca2+ events saturated while the magnitude of individual events increased; this included expansion of amplitude, duration and spread (see Figure 7). At the highest substance P concentration tested (~10−6 mol/L), essentially all active Ca2+ sites generated singular biphasic responses with extended plateaus (in excess of 100 seconds).
Figure 8A (left panel) shows specific substance P-mediated Ca2+ parameter changes (means from Figure 7) plotted as percent change from basal. This provides a comparable linear index of change for all the dynamic parameters (sites, events per site, amplitude, duration and spatial spread). The whole-field Ca2+ change (i.e. average fluorescence change over the entire sampled field rather than within discrete ROIs) is also plotted for the same data set. The right panel of Figure 8A shows the same data wherein the dynamic Ca2+ parameters are combined into a single curve based on the average percent change in parameters at each substance P concentration. Notably, the composite Ca2+ dynamics increased by 265% over the full substance P concentration range while whole-field Ca2+ only increased by 33%.
In order to relate functional responses directly to measured endothelial Ca2+, we plotted concentration-dependent substance P-induced vasorelaxation with the composite change in Ca2+ dynamics and the relative change in whole-field Ca2+ signal, all as percent of maximal response (Figure 8B). Comparison of normalized nonlinear regression curves shows that concentration-dependent vasorelaxation and Ca2+-dynamics superimpose (logEC50 −11.13 ± 0.06 vs. −11.31 ± 0.09, respectively) whereas whole-field Ca2+ is significantly right-shifted compared to both (logEC50 −9.75 ± 0.11, p<0.01 vs. relaxation and Ca2+-dynamics). These data indicate that changes in endothelial Ca2+ dynamics along the coronary artery intima correspond explicitly with concentration-dependent relaxation of coronary arteries to substance P, and that this functional response is poorly predicted by changes in global endothelial Ca2+ levels.
In the current study, we provide the first characterization of intrinsic endothelial Ca2+ signals directing basal and stimulated vasorelaxation of SCAs. Our data reveal a spatially and temporally diverse assortment of Ca2+ dynamics occurring along the coronary artery endothelium that are highly dependent on Ca2+ release from the endoplasmic reticulum through IP3Rs. Functional findings suggest the inherent Ca2+ signals exert a persistent vasorelaxing influence on coronary arteries through activation of eNOS and KCa channels (KCa2.3 and KCa3.1). Moreover, graded GqPCR stimulation elicits expansion of existing endothelial Ca2+ dynamics through frequency modulation and both spatial and temporal signal amplification. Importantly, this idiosyncratic Ca2+ signal expansion corresponds explicitly with coronary artery vasorelaxation whereas global Ca2+ within the same endothelial field does not. Together, our findings expose a previously unappreciated dynamic Ca2+ signaling framework within the coronary artery intima that may underlie definitive profiles of effector recruitment and vascular function.
The endothelium of the coronary circulation is absolutely crucial for control of cardiac blood flow and distribution. Its vital role is underscored by the predictable progression from endothelial dysfunction to ischemic heart disease. Given the accepted functional importance of Ca2+ in the coronary vasculature, the paucity of studies directly addressing Ca2+ signaling in the coronary endothelium is surprising. Recent studies of cultured cells and coronary artery segments have begun to expose key elements of endothelial Ca2+ control through direct Ca2+ measurements.22,35–37 Our goal was to apply new approaches for vascular imaging and analysis to unravel spatial and temporal detail of Ca2+ signaling in the endothelium of coronary arteries. For these studies we assessed arteries from pigs as they provide a highly relevant model of human coronary function and disease.38 Assessing vessel segments rather than dispersed cells provided a comprehensive look at endothelial signaling in the context of the intact tissue. A particularly novel aspect of the study was the use of the analysis module LC_Pro to capture the full spectrum of dynamic Ca2+ activity along the arterial endothelium and provide definitive metrics of event location, frequency, and size. We previously applied similar high-content analysis to elucidate functional Ca2+ signaling in rodent mesenteric and cerebral arteries.27,28 Here, we expose robust basal Ca2+ signals in the SCA endothelium encompassing a broad range of focal transients and cellular waves. Although we focused on moderate sized vessels (~0.5 mm), we found similar event distributions in large (~1 mm) and small (~0.2 mm) arteries, with smaller arteries exhibiting a somewhat higher occurrence of events (Online Figure VI). Overall, our findings suggest the initiation and relative spread of the Ca2+ events is highly dependent on IP3R release. In fact, it appears the ongoing Ca2+ activity is largely if not entirely supported by basal IP3, similar to previously described Ca2+ pulsar activity in mouse mesenteric arteries.24 Despite clear similarities to Ca2+ pulsars, basal SCA endothelial Ca2+ events are generally lower in frequency and more variable in size. Indeed, Ca2+ pulsars mainly focus around isolated basolateral IP3R clusters while SCA endothelial events are more likely to propagate as waves. This may be due to the longitudinal arrangement of ER/IP3Rs in the SCA endothelium that promotes directional Ca2+-induced Ca2+ release39,40 along the cell axis. The breadth of Ca2+ events predicts a distinct profile of cellular targets in the SCA. It is also notable that quite disparate Ca2+ events can occur at a single site within the SCA endothelium (Online Figure VII), suggesting that specific sites do not necessarily produce one type of event. Hence, even basal conditions may be sufficiently flexible to shape local signals (i.e. relative IP3 gradients). Interestingly, the largest and longest-lasting Ca2+ events in the coronary endothelium occur around or within the nucleus. The potential role of these nuclear Ca2+ surges in directing endothelial protein expression patterns or processes such as apoptosis warrants further study.41
Findings over the past two decades indicate that NO and EDH account for the crucial Ca2+-dependent endothelial regulation of coronary flow. NO production is well-recognized as a ubiquitous mechanism of coronary vasodilation, both in epicardial arteries and in the microcirculation,42–44 and its loss portends endothelial dysfunction and heart disease. The pivotal impacts of KCa2.3 and KCa3.1 channels in the coronary circulation have been established more recently9,10, and include acute EDH-mediated vasorelaxation, contribution to pressure-flow autoregulation,18 and compensatory vasodilation under conditions of reduced NO bioavailability, including obesity.6,35 Our current findings indicate that basal Ca2+ dynamics in SCA endothelium exert a persistent relaxing influence on tone through sustained recruitment of both eNOS and KCa channels, with NO production predominating (~70% NO, ~30% KCa). Because these functional endothelial influences occur in the absence of shear stress or endothelial agonist, we submit the coronary circulation is hardwired for constitutive vasoregulation, regardless of overt stimulation. Notably, this implies compensatory EDH signaling is not an alternative to NO, but rather a built-in feature of the coronary endothelium. It is important to note that because basal endothelial Ca2+ dynamics are asynchronous and localized in nature, functional signaling can occur while average Ca2+ over a broad intimal field remains essentially constant. The broad implication is that important background regulation occurs (i.e. modulation of myogenic constriction) even as the endothelium remains globally stable and responsive to acute stimulus.
The basal NO influence in the SCA differs from the predominant KCa3.1 channel-driven EDH in mouse mesenteric artery, implying preferential targeting in different beds and species. Such scenarios are ultimately dependent on both the specific spatiotemporal Ca2+ signaling profile and the existing Ca2+-dependent effector distribution. Post-transcriptional lipid modification45–47 and association with trafficking proteins such as caveolin48–50 and AKAP15026 allow different effectors to localize to discrete microdomains (i.e. eNOS and KCa2.3 to Golgi or caveolae and KCa3.1 to basolateral myoendothelial projections), thereby determining their accessibility to certain Ca2+ signals. The prevalent NO signaling in swine SCA endothelium may reflect targeting of both peripheral and Golgi-associated eNOS50 by local near-membrane events and nuclear-centric Ca2+ waves. Indeed, we show NO production at discrete central and peripheral sites corresponding with basal Ca2+ dynamics in SCA endothelial cells. Still, ~30% of the basal endothelial modulation of tone (mainly through NO) is independent of IP3 signaling (see Figure 5B), possibly reflecting the contribution of a separate Ca2+ source or eNOS activation by phosphorylation, independent of Ca2+.51
Previous studies have shown that direct stimulation of endothelial Gq-coupled receptors increases the frequency and/or spatial coverage of dynamic endothelial Ca2+ signals in mouse mesenteric arteries.24,25,28 Similarly, de novo recruitment of endothelial Ca2+ wavelets has been reported in phenylephrine-stimulated retractor muscle feed arteries through myoendothelial communication of IP3.29 In the current study, exposure of SCAs to the potent IP3-elevating vasodilator substance P consistently promoted distinct bursts of Ca2+ dynamics along the endothelium. While multiple Ca2+ sources may contribute, this behavior is consistent with IP3-sensitized Ca2+ store release, similar to the clustered firing of Ca2+ transients in Xenopus oocytes upon acute elevation of intracellular IP3 concentration.53 Timing of these events is likely dependent on the density, latency and refractory nature of the IP3Rs themselves. Here, we show that the graded expansion of discrete endothelial Ca2+ signals by substance P corresponds explicitly with acute vasorelaxation. Specifically, we found that the net change in dynamic Ca2+ signal parameters, represented as a linearly translated composite of event number (sites and frequency) and size (amplitude, duration and spread), superimposed on concentration-dependent substance P relaxation of arterial tone while the average Ca2+ signal of the endothelial field did not. This functional expansion of Ca2+ signaling involves both spatial and temporal components; most notably recruitment of new Ca2+-liberating sites at lower concentrations and increasing event duration at higher concentrations (see Figure 8). Coincidentally, assessment of Ca2+ event “area under curve” (AUC), which takes into account both event amplitude and duration (F/F0*s; measured from half-max to event peak), showed changes similar to duration alone, suggesting it is also not an independent predictor of low-concentration substance P vasorelaxation. This is likely due to the fact that modest stimulation elicits both large and small events, so average event size does not necessarily change. However, when we consider not only the size but also the number of events, the relationship between Ca2+ dynamics and function becomes highly predictive. The sensitive correspondence of new site recruitment to vasorelaxation is similar to that observed in rat cerebral arteries following direct endothelial stimulation27,52 and may represent a conserved mode of amplifying endothelial responses across multiple circulations.
It should be noted that at low levels of stimulation, subtle changes in local Ca2+ dynamics (i.e. recruitment of new sites) can easily be hidden within the broad sampled field, similar to basal signals. This is apparent in Figure 6B where new Ca2+ events evoked by very modest substance P stimulation contribute minimally to the average field Ca2+ due to their spatial restriction and general asynchrony over the time course. Thus, not only is global endothelial Ca2+ change a poor index of vasorelaxation in the coronary circulation, it would seem to be particularly inadequate at assessing the subtle perturbations associated with physiologic stimuli. Although it is currently difficult to explicitly link individual Ca2+ parameter changes to distinct shifts in effector recruitment profiles, assessment of low and high-concentration substance P stimulation in the current study suggests subtle Ca2+ signal expansion (mainly new sites) at low stimulation levels may preferentially target eNOS, It will be important to determine whether impairment of such sensitive Ca2+-eNOS coupling underlies the early and specific loss of NO signaling with progressive coronary artery disease. It is worth noting that at higher levels of acute endothelial stimulation the overall synchrony of dynamic events tends to increase, and this corresponds with the timing and peak of arterial vasorelaxation. This relative harmony of independent signals may offer another important level of physiologic tuning and also warrants further study. One limitation of the current study was that evaluations of Ca2+ signaling and functional responses had to be performed in separate vascular preparations in order to provide adequate endothelial access for imaging. While this restricts direct inferences between endothelial Ca2+ and arterial tone, studies were performed in parallel using the same animals to preserve continuity of measurements.
In addition to IP3Rs, stimulation of membrane TRP channels can also increase dynamic endothelial Ca2+ signaling.26 These channels respond to various stimuli such as stretch, shear, temperature and second messengers54 to increase focal Ca2+ entry. In particular, activation of TRPV4 channels in mesenteric artery endothelium, either directly or through acetylcholine stimulation, increases the occurrence of Ca2+ transients at the plasma membrane,24 and these signals can be augmented by KCa3.1/2.3 channel positive feedback.28 In the current study, we found no significant influence of Ca2+ entry on basal Ca2+ dynamics, nor did we find any significant effect of TRPV4 blockade on substance P relaxation (Online Figure VIII). While this indicates minimal involvement of TRPV4 channels under the conditions studied, recent studies have implicated these channels in shear stress responses,21,55 including NO-dependent, flow-mediated dilation of human coronary arteries.36 This could have important implications in the context of the current study since stimulated Ca2+ influx would superimpose on existing intrinsic Ca2+ dynamics, not only increasing membrane delimited signals but also likely enhancing Ca2+-induced Ca2+ release from ER IP3Rs. Elucidating the complex effects of differential shear stress profiles on endothelial Ca2+ dynamics is a high priority for future studies. The current work provides a functional signaling framework and definitive analysis approach for discerning these impacts.
Although we found no H2O2 contribution to the basal endothelial modulation of coronary artery tone, such signaling is likely enhanced under pathological conditions. In patients with CAD, Ca2+-dependent flow-mediated dilation switches from a predominant NO-mediated mechanism to one dominated by H2O2.37 The cause of this pathological transition is unknown but likely involves a distinct shift in endothelial Ca2+ patterning to one favoring chronic mitochondrial and/or NADPH oxidase production of reactive oxygen species.56,57 Future studies should identify whether a distinctive pattern switch underlies endothelial dysfunction in developing coronary disease. It should also be noted that the use of juvenile pigs in the current study allowed us to limit variability due to age and gender. Because endothelial dysfunction and coronary artery disease manifest differently over the life-span of men and women,16 future applications of the high-content analysis described here should be particularly useful for discerning sex and age-related disparities.
In summary, the coronary artery endothelium is an active interface, critical to cardiac health. The pro-dilatory, anti-inflammatory, and anti-proliferative influences of the endothelium are all inherently dependent on real-time control of diverse Ca2+ signals. The current study provides evidence that coronary endothelial function is fundamentally encoded by a persistent mosaic of Ca2+ transients. This intrinsic profile of Ca2+ dynamics establishes an essential active framework for directing and tuning endothelial responses in the coronary circulation. Moving forward, it will be imperative that coronary endothelial Ca2+ signaling be resolved with adequate spatial and temporal detail to characterize stimulus-specific responses and define transitions from physiologic to pathologic signaling patterns.
Although endothelial Ca2+ signaling is a pivotal determinant of coronary artery function and the progression of coronary artery disease, the spatial and temporal regulation of discrete Ca2+ signals remains unclear. Here, we aimed to determine the distinct profile of Ca2+ dynamics underlying endothelium-dependent relaxation in SCAs. Using confocal imaging and custom automated image analysis, we found that the SCA endothelium possesses a characteristic pattern of endothelial Ca2+ events encompassing a wide range isolated transients and whole-cell waves. Derived largely from the transient release of internal Ca2+ stores, these events drive constant endothelium-dependent vasorelaxation through persistent recruitment of eNOS and Ca2+-activated K+ channels (KCa3.1 and KCa2.3), effectors known to play an important role in the modulation of coronary artery tone. We also find that acute endothelial stimulation with the vasodilator substance P causes distinct changes in both the spatial and temporal properties of dynamic Ca2+ signals, and these changes correspond precisely with concentration-dependent vasorelaxation whereas whole-field (averaged) endothelial Ca2+ changes do not. We propose that the direct tuning of Ca2+ transients (i.e. spatial recruitment as well as changes in frequency, amplitude, duration and spread) underlies physiologic and pathophysiologic signaling in the coronary circulation, and this signaling may not be adequately represented by tissue-level Ca2+ changes.
We thank Dr. Steve Ballard for his assistance in obtaining and utilizing swine tissues.
Sources of funding: This work was supported by grants from the NIH (HL085887 and S10RR027535) and the American Heart Association (0435051N).
Calcium Cycling/Excitation-Contraction Coupling
Endothelium/Vascular Type/Nitric Oxide
Ion Channels/Membrane Transport
Basic Science Research