TRPV4 channel-dependent negative feedback mechanism regulates G_q protein-coupled receptor-induced vasoconstriction

Abstract

Objective

Several physiological stimuli activate smooth muscle cell (SMC) G_q protein-coupled receptors (G_qPCRs) to cause vasoconstriction. As a protective mechanism against excessive vasoconstriction, SMC G_qPCR stimulation invokes endothelial cell (EC) vasodilatory signalling. Whether Ca²⁺ influx in ECs contributes to the regulation of G_qPCR-induced vasoconstriction remains unknown. Ca²⁺ influx through transient receptor potential vanilloid 4 (TRPV4) channels is a key regulator of endothelium-dependent vasodilation. We hypothesized that SMC G_qPCR stimulation engages endothelial TRPV4 channels to limit vasoconstriction.

Approach and Results

Using high-speed confocal microscopy to record unitary Ca²⁺ influx events through TRPV4 channels (TRPV4 sparklets), we report that activation of SMC alpha₁-adrenergic receptors with phenylephrine (PE) or thromboxane A₂ receptors with U46619 stimulated TRPV4 sparklets in the native endothelium from mesenteric arteries. Activation of endothelial TRPV4 channels did not require an increase in Ca²⁺ as indicated by the lack of effect of L-type Ca²⁺ channel activator or chelator of intracellular Ca²⁺ EGTA-AM. However, gap junction communication between SMCs and ECs was required for PE- or U46619-activation of endothelial TRPV4 channels. Lowering IP₃ levels with phospholipase C inhibitor or lithium chloride suppressed PE-activation of endothelial TRPV4 sparklets. Moreover, uncaging IP₃ profoundly increased TRPV4 sparklet activity. In pressurized arteries, PE-induced vasoconstriction was followed by a slow, TRPV4-dependent vasodilation reflecting activation of negative regulatory mechanism. Consistent with these data, PE induced a significantly higher increase in blood pressure in TRPV4^−/− mice.

Conclusion

These results demonstrate that SMC G_qPCR stimulation triggers IP₃-dependent activation of endothelial TRPV4 channels to limit vasoconstriction.

INTRODUCTION

Activation of G_q protein-coupled receptors (G_qPCRs) in vascular smooth muscle cells (SMCs) has emerged as a central mechanism for vasoconstriction in response to intravascular pressure and neurohumoral mediators^1–6. Sympathetic nerve stimulation predominantly activates G_q-coupled alpha₁-adrenergic receptors (α₁ARs) in SMCs to cause vasoconstriction. Although α₁AR-mediated vasoconstriction is an important component of the flight-or-fight response, excessive vasoconstriction can lower the blood flow to target tissues and elevate blood pressure to abnormal levels. In this regard, endothelial cells (ECs) play an important role in limiting the vasoconstriction induced by SMC α₁ARs. SMC to EC heterocellular communication evokes endothelial vasodilatory mechanisms to regulate vasoconstriction^7–10. However, the individual signaling elements of this negative regulatory mechanism remain unclear.

Recent studies reveal a crucial role for endothelial projections to SMCs, or myoendothelial projections (MEPs), in negative regulation of vasoconstriction in response to α₁AR activation^8–10. MEPs are specialized structures that contain key elements for endothelium-dependent vasodilation including inositol 1,4,5-trisphosphate (IP₃) receptors (IP₃Rs) in the endoplasmic reticulum (ER) and intermediate/small conductance Ca²⁺-sensitive K⁺ (IK/SK) channels and cation channels at the cell membrane^11–14. Moreover, myoendothelial gap junctions (MEGJs) at the end of the MEPs¹⁵ allow for the movement of second messengers and ionic charges between SMCs and ECs^{7–9, 16, 17}. Localized increases in Ca²⁺ at the MEPs have been associated with IK/SK channel activation and endothelium-derived hyperpolarization (EDH) in small, resistance-sized arteries^{11, 12, 18}. Activation of G_qPCRs in SMCs causes vasoconstriction predominantly via two second messengers, IP₃ and Ca²⁺. The diffusion of IP₃ and Ca²⁺ from SMCs to ECs via MEGJs can increase Ca²⁺ inside the MEPs directly or via activation of IP₃Rs on the ER membrane^{7–10, 16, 17}. Whether the diffusion of second messengers from SMCs to ECs following SMC G_qPCR activation also invokes Ca²⁺ influx pathways at the MEPs has not been elucidated.

Previous studies have established that Ca²⁺ influx signals through TRPV4 (transient receptor potential vanilloid 4) channels localized at MEPs regulate EDH-mediated vasodilation in resistance-sized mesenteric, cremaster, and cerebral arteries^{11, 14, 18, 19}. TRPV4 channels are Ca²⁺-selective cation channels that are activated by flow/shear stress, G_q protein signaling, or changes in temperature and pressure^{11, 20, 21}. We recently showed that local, unitary Ca²⁺ influx signals through endothelial TRPV4 channels, termed TRPV4 sparklets, activate IK/SK channels to dilate resistance-sized mesenteric arteries^{14, 18, 22}. Moreover, TRPV4 channels are potentiated by Ca²⁺ and have a binding site for IP₃^{14, 23–27}. Therefore, we hypothesized that SMC G_qPCR activation engages endothelial TRPV4 channels via IP₃ and Ca²⁺ to limit vasoconstriction.

In the current study we utilized differential kinetics and pharmacology to isolate IP₃R-dependent Ca²⁺ signals and TRPV4 sparklets in the native endothelium. Moreover, we provide the first evidence that SMC G_qPCR agonists trigger endothelial TRPV4-IK/SK signaling to negatively regulate vasoconstriction. While SMC G_qPCR-induced activation of TRPV4 channels requires SMC-EC communication via gap junctions, it is independent of the movement of Ca²⁺ from SMCs to ECs or IP₃R-mediated Ca²⁺ release in ECs. Our findings support a novel paradigm that IP₃ is an endogenous activator of TRPV4 channels in the native endothelium and point to IP₃-mediated activation of endothelial TRPV4 channels as a key negative regulatory mechanism for SMC G_qPCR-induced vasoconstriction.

MATERIAL AND METHODS

Materials and Methods are available in the online-only Data Supplement.

RESULTS

Distinct kinetics and pharmacology enable the isolation of TRPV4 Ca²⁺ sparklets and IP₃R-mediated Ca²⁺ signals in the native endothelium from mesenteric arteries (MAs)

SMC α₁AR activation has been previously shown to stimulate IP₃R-dependent Ca²⁺ signals in the native endothelium^{8, 9}. To determine whether SMC α₁AR activation also augments endothelial TRPV4 channel function under basal conditions, it is necessary to separate TRPV4 sparklets from IP₃R-mediated Ca²⁺ signals. We hypothesized that IP₃R Ca²⁺ signals and TRPV4 sparklets can be distinguished based on distinct kinetic and pharmacological properties. Ca²⁺ signals were recorded using high-speed spinning disk confocal microscopy in en face 3^rd-order MAs from transgenic mice that express genetically encoded Ca²⁺ biosensor (GCaMP2) exclusively in ECs. Approximately 15 ECs were seen in a field of view measuring 110X110 μm (Figure 1A, left). Based on the fractional fluorescence (F/F₀) traces, Ca²⁺ signals under basal conditions could be broadly categorized into: short, spike-shaped signals with one distinct peak and longer-lasting signals with square, discrete amplitudes. Consistent with previous reports on IP₃R-mediated Ca²⁺ signals in GCaMP2 mice^{8, 12}, the spike-shaped events had a duration of <300 ms. These events were inhibited by cyclopiazonic acid (CPA, 20 μM), an inhibitor of sarco/endoplasmic reticulum (SR/ER) Ca²⁺-ATPase, but not by the selective TRPV4 channel inhibitor GSK2193874 (GSK219, 100 nM) or 0 mM extracellular Ca²⁺ (Figure 1A, middle & right), indicating Ca²⁺ release from the ER. On the other hand, signals with square amplitudes lasted >300 ms and had a plateau phase with multiple data points at the peak (Figure 1B, left). A multi-Gaussian fit to the all-points histogram of square events yielded a quantal level of 0.185 ΔF/F₀, consistent with the quantal level for TRPV4 sparklets^{14, 18} (Figure 1B, middle). Moreover, the square signals were inhibited by GSK219 and 0 mM extracellular Ca²⁺, but not by CPA, indicative of Ca²⁺ influx through TRPV4 channels at EC membranes (Figure 1B, left & right). These results confirmed the presence of two major Ca²⁺ signals in the native endothelium from MAs under basal conditions, TRPV4 sparklets and IP₃R Ca²⁺ signals. Importantly, the two types of events can be isolated based on their differential kinetics and pharmacology.

Figure 1. SMC α₁AR activation with phenylephrine (PE) increases both TRPV4 Ca²⁺ sparklets and IP₃R-mediated Ca²⁺ signals.

TRPV4 Ca²⁺ sparklets and IP₃R Ca²⁺ signals were isolated using their distinct kinetics and pharmacological properties. (A) Left, a greyscale image of a field of view with ~15 ECs from GCaMP2 mouse. Square boxes represent the regions of interest (ROIs) placed at an event site indicating an IP₃R-mediated Ca²⁺ signal or TRPV4 Ca²⁺ sparklet. Middle, representative F/F₀ traces generated using the ROIs in Figure 1A indicate IP₃R Ca²⁺ signals in the absence or presence of CPA (an inhibitor of sarco/endoplasmic reticulum (SR/ER) Ca²⁺-ATPase, 20 μM). Right, averaged number of IP₃R Ca²⁺ signal sites in the absence or presence of GSK219 (a selective TRPV4 channel inhibitor, 100 nM), 0 mM extracellular Ca²⁺, or CPA (n=5). * p<0.05 vs. CPA. (B) Left, representative F/F₀ traces generated using the ROIs in Figure 1A display TRPV4 Ca²⁺ sparklets under basal conditions or with GSK219 (100 nM). Middle, all-points histogram was constructed from F/F₀ traces obtained from three MAs and was fit with a multi-Gaussian curve. The quantal levels (i.e. step-wise increases in amplitudes) were obtained from the peaks of the multi-Gaussian curve. Right, averaged TRPV4 sparklet activity is presented in the absence or presence of GSK219, CPA, or 0 mM extracellular Ca²⁺, or CPA. TRPV4 sparklet activity is expressed as NP_O per site and number of sparklet sites per field. N represents the number of channels at a site and P_O is the open state probability of the channels. * # p<0.05 vs. Baseline or CPA, respectively. (C) Left, representative F/F₀ traces indicate IP₃R Ca²⁺ signals under basal conditions or in the presence of PE (10 μM). Right, quantified IP₃R Ca²⁺ signals in the absence or presence of PE alone or PE+GSK219. Ca²⁺ signals are presented as event sites per field and total events per field (n=5). * p<0.05 vs. Baseline. (D) Left, representative F/F₀ traces show TRPV4 Ca²⁺ sparklet activity under basal conditions or following PE treatment (10 μM). Right, averaged TRPV4 sparklet activity is presented in the absence or presence of PE alone or PE+GSK219 (n=5). * # p<0.05 vs. Baseline or PE, respectively. Data are presented as mean ± SEM.

Activation of SMC α₁ARs stimulates endothelial TRPV4 sparklets and IP₃R-mediated Ca²⁺ signals

SMC α₁AR activation generates Ca²⁺ and IP₃ as second messengers, which can cross over to the ECs via MEGJs^{7–10, 16, 17}. Both Ca²⁺ and IP₃ have been shown to alter the activity of TRPV4 channels^{14, 23–27}. Therefore, we hypothesized that SMC α₁AR activation stimulates TRPV4 sparklet activity in ECs. We used the following criteria to isolate TRPV4 sparklets from IP₃R-dependent Ca²⁺ signals: 1) durations > 300 ms and 2) inhibition by GSK219. Consistent with previous studies showing α₁AR activation-induced increase in IP₃R-mediated Ca²⁺ signals in ECs^{8, 9}, phenylephrine (PE) stimulated the activity of IP₃R-dependent Ca²⁺ signals in ECs from en face MAs (Figure 1C). Interestingly, SMC α₁AR activation also caused a 3-fold increase in TRPV4 sparklet activity and a 2.5-fold increase in the number of sparklet sites per field (Figure 1D). GSK219 abolished the PE-induced increases in TRPV4 sparklet activity, but not the increase in IP₃R Ca²⁺ signals (Figure 1C and D). These results confirmed that PE increases the activity of both TRPV4 sparklets and IP₃R-mediated Ca²⁺ signals.

SMC G_qPCR activation stimulates endothelial TRPV4 sparklets independent of IP₃R-mediated Ca²⁺ release from the ER

Further mechanistic studies were performed in fluo-4 (a cell-permeable fluorescent Ca²⁺ dye/detector) loaded MAs (Figure 2A, left) to allow visualization of Ca²⁺ in EC and SMC layers from the same artery. The experiments were performed in the presence of CPA (20 μM) to exclusively record and analyze TRPV4 sparklets in absence of IP₃R Ca²⁺ signals. CPA by itself did not alter TRPV4 sparklet activity (Figure 1B, right), suggesting that TRPV4 sparklets do not represent store-operated Ca²⁺ entry. An increase in SMC Ca²⁺ in response to PE was used as a confirmation of arterial health. Approximately 1.5 sparklet sites per field were observed in the presence of CPA alone (Figure 2B, left). PE treatment (10 μM) consistently augmented the activity of TRPV4 sparklets to a level similar to that observed in absence of CPA (Figure 2A, right; Figure 2B). The increased TRPV4 sparklet activity was substantially attenuated by GSK219 treatment (100 nM, Figure 2B, left and middle). The PE-induced increase in TRPV4 sparklet activity was eliminated in the presence of prazosin (an α₁AR antagonist, 500 nM) or GSK219 pre-treatment, and in the MAs from TRPV4^−/− mice (Figure 2B, right). Moreover, addition of xestospongin C (XSC, 3 μM, a potent IP₃R antagonist) in the presence of CPA did not alter PE-induced increase in TRPV4 sparklet activity (Figure 2C), suggesting that TRPV4 sparklets were being activated independent of IP₃R-mediated Ca²⁺ release. Moreover, in the presence of a membrane permeable Ca²⁺ chelator, EGTA-AM (5 μM), PE was able to activate TRPV4 sparklets to a similar extent (Figure 2D), thus ruling out a role for Ca²⁺ in PE-activation of TRPV4 sparklets. Consistent with previous studie^{10, 17}, a strong immunostaining of α_1DAR subtype was observed in SMC layer, but not in EC layer in MAs (Figure 2E). We therefore postulated that SMC α_1DARs activate endothelial TRPV4 channels via SMC-EC heterocellular communication. Another SMC G_qPCR agonist U46619 (a thromboxane A₂ receptor agonist, 1 μM) also caused a 5-fold activation of endothelial TRPV4 sparklets (Figure 2F), supporting the idea that multiple activators of G_q-protein signaling in SMCs stimulate endothelial TRPV4 sparklets.

Figure 2. SMC G_qPCR activation stimulates endothelial TRPV4 sparklets independent of IP₃R-mediated Ca²⁺ release in MAs loaded with fluo-4.

Experiments were performed in the presence of CPA (20 μM). (A) Left, a greyscale image of a field of view with ~15 ECs from C57BL6/J mouse. Square boxes represent the regions of interest (ROIs) placed at TRPV4 Ca²⁺ sparklet sites. Right, representative F/F₀ traces generated using the ROIs on the grayscale image display TRPV4 sparklets before and after the addition of PE (10 μM). (B) Left, averaged basal or PE-induced (10 μM) endothelial TRPV4 sparklet activity before or after GSK219 addition (100 nM, n=9) * # p<0.05 vs. CPA or PE, respectively. Right, summarized effect of PE on TRPV4 sparklet activity in the presence of prazosin (an α₁AR antagonist, 500 nM), GSK219, or in MAs from TRPV4^−/− mice (n=8). (C) Averaged TRPV4 sparklet activity in the absence or presence of XSC, +PE, or +GSK219 (n=4). XSC represents xestospongin C (a potent IP₃R inhibitor, 3 μM). * # p<0.05 vs. CPA+XSC or PE, respectively. (D) Averaged TRPV4 sparklet activity in the presence of GSK101 (a TRPV4 channel agonist; 10 nM) alone, EGTA-AM (a cell permeable Ca²⁺ chelator; 5 μM), and PE (10 μM) (n=5). (E) Representative images for α_1DAR immunostaining (left), nuclear staining with DAPI (middle), and merged staining (right) in the ECs (top) and SMCs (bottom) from an en face MA. The immunostaining image is representative of n=15 separate fields. (F) Left, representative F/F₀ traces display TRPV4 sparklets before and after the addition of U46619 (1 μM). Right, averaged TRPV4 sparklet activity in the presence of CPA, +U46619, or +GSK219 (n=6). U46619 is a thromboxane receptor agonist (1 μM). * # p<0.05 vs. CPA or U46619, respectively. TRPV4 sparklet activity is expressed as NP_O per site and number of sparklet sites per field (Figure 2B–D, F). N represents the number of channels at a site and P_O is the open state probability of the channels. Data are presented as mean ± SEM.

α₁AR activation of endothelial TRPV4 sparklets requires phospholipase C (PLC) signaling and movement of second messengers through MEGJs

IP₃ and Ca²⁺ are the two major second messengers generated by α₁AR signaling via PLC activation in SMCs. Both these second messengers can diffuse to the EC side via MEGJs^{7–10, 16}. Therefore, we determined whether PLC signaling and MEGJ function are necessary for the activation of TRPV4 sparklets by SMC α₁ARs. In the presence of U73122 (a PLC inhibitor, 3 μM), TRPV4 sparklet activity was not altered by PE (Figure 3A). Moreover, gap junction uncoupling agents 18β-glycyrrhetinic acid (18βGA, 30 μM)^{28, 29} or 1-heptanol (heptanol, 1 mM)^30–32 also abolished PE-induced increase in TRPV4 sparklet activity (Figure 3B, Supplemental Figure IA). U73122, 18βGA, or heptanol, by itself, did not alter the activity of TRPV4 sparklets (Supplemental Figures IA, IB). These results suggested that second messengers generated by PLC signaling may diffuse to the ECs through MEGJs and activate TRPV4 sparklets. Lack of effect of PE or U46619 on TRPV4 sparklet activity in the presence of 18βGA or heptanol ruled out a direct effect of these compounds on endothelial cell TRPV4 channel function. PE and U46619 were able to increase SMC Ca²⁺ in the presence of CPA (Supplemental Figures IIA, IIB). Increase in SMC Ca²⁺ by PE or U46619 was not altered by 18βGA (Supplemental Figures IIC, IID), supporting the concept that the effect of PE or U46619 is initiated exclusively in the SMCs.

Figure 3. SMC α₁AR signaling activates phospholipase C and SMC-EC coupling via MEGJs to stimulate endothelial TRPV4 channels in a Ca²⁺-independent manner.

Experiments were performed in the presence of CPA (20 μM). (A) Left, representative F/F₀ traces for the effect of PE (10 μM) on TRPV4 sparklets in the presence of U73122 (a PLC inhibitor, 1 μM). Right, quantified TRPV4 sparklet activity in response to PE in the absence or presence of U73122 (n=7). * p<0.05 vs. CPA. (B) Left, representative F/F₀ traces for the effect of PE (10 μM) on TRPV4 sparklets in the presence of 18βGA (a gap junction uncoupling agent, 30 μM). Right, averaged TRPV4 sparklet activity in response to PE in the absence or presence of 18βGA (n=6). * p<0.05 vs. CPA. (C) Quantified fluorescence intensity indicates whole-cell Ca²⁺ levels (arbitrary fluorescence units or AFU) in SMCs (top) and ECs (bottom) before and after BayK8644 (a voltage-dependent Ca²⁺ channel agonist, 1 μM, n=47–50 cells of each). * p<0.05 vs. Baseline. (D) Representative F/F₀ traces (top) and quantification (bottom) of averaged TRPV4 sparklet activity in the absence or presence of BayK8644 (1 μM) or BayK8644+GSK219 (100 nM, n=6). TRPV4 Ca²⁺ sparklet activity is expressed as NP_O per site and number of sparklet sites per field (Figure 3A, B, and D). N represents the number of channels at a site and P_O is the open state probability of the channels. Data are presented as mean ± SEM.

Increase in SMC Ca²⁺ does not alter endothelial TRPV4 sparklet activity

PE increases SMC Ca²⁺ via IP₃R-mediated Ca²⁺ release from the ER and influx of extracellular Ca²⁺³³. Although endothelial TRPV4 channels can be potentiated by intracellular Ca^{2+14, 18, 25, 27}, it remains unknown whether PE-induced increase in SMC Ca²⁺ is responsible for the activation of endothelial TRPV4 sparklets. Activation of TRPV4 sparklets by PE to a similar extent in presence or absence of CPA and XSC suggests that IP₃R-mediated Ca²⁺ release in SMCs or ECs is not required for the activation of TRPV4 sparklets by PE (Figure 2C). To examine whether an increased Ca²⁺ influx in SMCs can activate TRPV4 sparklets in ECs, BayK8644 (a voltage-dependent L-type Ca²⁺ channel opener, 1 μM) was utilized to selectively increase Ca²⁺ concentration in SMCs. BayK8644 increased global Ca²⁺ levels in SMCs (Figure 3C, top), but did not enhance global Ca²⁺ levels in ECs (Figure 3C, bottom) or endothelial TRPV4 sparklet activity (Figure 3D). These results indicate that PE-induced increase in endothelial TRPV4 sparklet activity is not due to increased SMC Ca²⁺.

IP₃ activates TRPV4 sparklets by increasing burst open times of TRPV4 channels

While PE-induced increase in TRPV4 sparklets was independent of SMC Ca²⁺, the PLC- and MEGJ-dependent nature of TRPV4 sparklet activation by PE suggested a role for SMC-derived IP₃. In this regard, previous studies have stipulated that the movement of IP₃ from SMCs to ECs can increase endothelial Ca^{2+8, 9, 16, 34}. A recent study in expression system unravelled a binding site for IP₃ on TRPV4 channel and showed that binding of IP₃ to this site increases channel activity²⁶. We, therefore, hypothesized that SMC-derived IP₃ activates endothelial TRPV4 channels following α₁AR activation. The presence of lithium chloride (LiCl, inositol monophosphatase inhibitor, 1 mM), which has been commonly used to lower intracellular IP₃ levels^35–38, inhibited PE- and U46619-induced activation of TRPV4 sparklets, suggesting an IP₃-dependent nature of TRPV4 sparklet activation by PE and U46619 (Figure 4A). To determine whether IP₃ can activate TRPV4 sparklets in native endothelium, we uncaged IP₃ using 10 ms pulse of ultraviolet (UV) light. En face mouse MAs were simultaneously incubated with fluo-4 and caged IP₃ (2 μM). CPA (20 μM) was added to prevent IP₃-induced Ca²⁺ release from the ER. UV light alone did not alter the activity of TRPV4 sparklets (Figure 4B). Uncaging IP₃ using UV light increased sparklet activity per site by 5-fold and number of sites per field by 3-fold, respectively (Figure 4C), an effect that was completely suppressed in the arteries pretreated with GSK219 (Figure 4C). Cooperative openings of TRPV4 channels in a cluster has emerged as an important regulatory mechanism for TRPV4 channel activity¹⁴. An analysis of coupling strength among TRPV4 channels in a cluster showed that IP₃ did not alter the cooperative openings (κ values) of TRPV4 channels in a cluster (Figure 4D). Majority of TRPV4 sparklets represent bursts of TRPV4 channel openings^{39, 40}. Uncaging IP₃ increased the burst open times for each quantal level by 2–4-fold (Figure 4E). Taken together, these results suggest that SMC-derived IP₃ activates endothelial TRPV4 channels in response to α₁AR signaling by increasing channel open times.

Figure 4. IP₃ augments endothelial TRPV4 sparklet activity by increasing TRPV4 channel open times, but not cooperativity.

Experiments were performed in the presence of CPA (20 μM). (A) Averaged TRPV4 sparklet activity in response to PE (10 μM) or U46619 (1 μM) in the presence of LiCl (1 mM, n=7). (B) Averaged TRPV4 sparklet activity before and after UV light exposure in the presence of CPA (n=6). (C) Top, representative F/F₀ traces for TRPV4 sparklets before (left) and after (right) IP₃ uncaging in the presence of CPA. Dotted red lines indicate the quantal levels (i.e. stepwise openings of 0–4 TRPV4 channels at a site) derived from all-points histograms (Figure 1B). Bottom, averaged TRPV4 sparklet activity before and after IP₃ uncaging in the absence or presence of GSK219 (100 nM, n=6). * p<0.05 vs. Before IP₃ uncaging. TRPV4 sparklet activity is expressed as NP_O per site or number of sparklet sites per field (Figure 4A and 4B). N represents the number of channels at a site and P_O is the open state probability of the channels. (D) Summarized coupling coefficient (κ) values indicating the effect of IP₃ on cooperative openings of TRPV4 channels at a site (n=14–20 events). Coupling coefficient (κ) values vary from 0 (no coupling or independent gating) to 1 (maximum coupling). (E) Mean burst open times (ms) for each quantal level of TRPV4 sparklets before and after IP₃ uncaging (n=6). * p<0.05 vs. Before IP₃ uncaging. Data are presented as mean ± SEM.

SMC α₁AR signaling activates TRPV4 sparklets exclusively at MEPs

Since MEGJs between endothelial and SMCs form at the ends of MEPs^41–43, we postulated that PE or U46619-induced increase in TRPV4 sparklet activity occurs exclusively at the MEPs. The MEPs are indicated by black holes in the internal elastic lamina (IEL, Figure 5A and 5C). Sparklet sites within 5 μm from the center of a hole in the IEL were considered as MEP sparklet sites¹⁴. PE-induced a 3.5-fold increase in TRPV4 sparklet activity at the MEPs but did not affect the sparklet activity at other sites (Figure 5B). Similarly, U46619 increased sparklet activity at MEPs by 5.5-fold and did not alter sparklet activity at non-MEP sites (Figure 5D). Collectively, these data support the concept that SMC-derived IP₃ locally diffuses through MEGJs and selectively activates TRPV4 channels at the MEPs.

Figure 5. SMC G_qPCR signaling induced an increase in TRPV4 sparklet activity exclusively at the myoendothelial projections (MEPs).

Experiments were performed in the presence of CPA (20 μM). (A) Representative images of an en face MA loaded with fluo-4 and counter-stained with Alexa Fluor 633 hydrazide (10 μM) displays TRPV4 sparklet sites at the holes in internal elastic lamina (IEL), indicative of the MEPs, in the absence (white arrows) or presence (yellow arrows) of PE (10 μM). White and yellow arrows represent TRPV4 sparklet sites that overlapped with MEPs. (B) Summarized data for TRPV4 sparklet activity at the MEP or non-MEP sites in the absence of presence of PE or PE+GSK219 (n=9). * # p<0.05 vs. CPA or PE, respectively. § p<0.05 vs. corresponding value at MEP. (C) Representative Alexa Fluor 633 hydrazide images in the absence (white arrows) or presence (yellow arrows) of U46619 (1 μM). White and yellow arrows represent TRPV4 sparklet sites that overlapped with MEPs. (D) Quantified TRPV4 sparklet activity at the MEPs or non-MEPs in the absence or presence of U46619 or U46619+GSK219 (n=6). * # p<0.05 vs. CPA or PE, respectively. § p<0.05 vs. corresponding value at MEP. TRPV4 sparklet activity is expressed as NP_O per site (Figure 5B and 5D). N represents the number of channels at a site and P_O is the open state probability of the channels. Data are presented as mean ± SEM.

Endothelial TRPV4 channels negatively regulate PE-induced vasoconstriction

Activation of endothelial TRPV4 sparklets promotes endothelium-dependent vasodilation in small MAs^{14, 18}. Therefore, we hypothesized that PE-induced activation of endothelial TRPV4 channels contributes to the feedback regulation of vasoconstriction. Vascular reactivity was assessed in MAs constricted by intravascular pressure of 80 mmHg. PE-induced vasoconstriction peaked at ~2 minutes after addition of PE and was followed by a slow vasodilation (Figure 6A, PE alone, Supplemental Table 1). The vasodilation following PE-induced vasoconstriction has been attributed to negative feedback regulation by endothelium^{7, 44, 45}. The peak vasodilation following PE was significantly attenuated in the presence of TRPV4 inhibitor GSK219, in endothelium-denuded arteries, and in the MAs from TRPV4^−/− mice (Figure 6A–C), indicating that endothelial TRPV4 channels are responsible for the vasodilation following PE-induced vasoconstriction.

Figure 6. TRPV4-IK/SK signaling mediates endothelium-dependent vasodilation following PE-induced vasoconstriction.

(A) Representative diameter traces for a biphasic response to PE (1–10 μM) showing an immediate vasoconstriction followed by a subsequent, slow vasodilation in the absence or presence of GSK219 (100 nM) or Tram-34 (1 μM)+apamin (300 nM), or from EC-denuded arteries. (B) Representative area under curves (AUCs, grey color) or diameter responses (dotted lines) illustrating the secondary vasodilation following PE (1–10 μM)-evoked vasoconstriction in the absence or presence of GSK219 (100 nM) or Tram-34 (1 μM)+apamin (300 nM). (C & D) Averaged AUCs (Figure 6C) or diameter responses (% vasodilation, Figure 6D) to PE (1–10 μM) in the absence or presence of GSK219 (100 nM), Tram-34+apamin, 18βGA (30 μM), Tram-34+apamin+L-NNA (100 μM), in MAs from TRPV4^−/− mice, in EC-denuded MAs, and in the presence of 20 μM CPA (n=4–7). * p<0.05 vs. Control. Data are presented as mean ± SEM. (E) Left, mean arterial pressure in control and TRPV4^−/− mice averaged over three days and three nights. Right, increase in mean arterial pressure after a single bolus injection of PE at noon (0.1 mg/kg, intraperitoneal), n =4. * p<0.05 vs. Control. Data are presented as mean ± SEM.

In systemic arteries, TRPV4 sparklets activate Ca²⁺-sensitive IK/SK channels to cause vasodilation¹⁸. Consistent with these prior findings, vasodilation after PE treatment was significantly reduced in the presence of Tram-34 (1 μM) /apamin (300 nM) (IK/SK channel inhibitors, Figure 6A–D). Addition of nitric oxide synthase (NOS) inhibitor L-NNA (L-N^G-Nitroarginine; N^G-nitro-L-Arginine, 100 μM) in the presence of IK/SK channel inhibitors abolished the remaining vasodilation after PE (Figure 6B–6D), supporting a minor role for NOS in opposing PE-induced constriction. Moreover, vasodilation after PE application was markedly attenuated in the presence of MEGJ uncoupling agent (18βGA, 30 μM, Figure 6C, 6D), indicating that SMC α₁AR initiates SMC-EC gap junction communication that evokes endothelial TRPV4-IK/SK signaling to cause vasodilation following PE-induced vasoconstriction. In the presence of CPA, to eliminate IP₃R-mediated Ca²⁺ release in ECs and SMCs, PE-induced vasoconstriction was followed by a vasodilation similar to that in the absence of CPA (Figure 6C, 6D, Supplemental Figure III). Moreover, the vasodilation was almost entirely inhibited by GSK219 (Supplemental Figure III). Thus, TRPV4 channel-dependent vasodilation following PE addition in the presence of CPA was consistent with that observed in the absence of CPA.

PE induced a larger increase in blood pressure in TRPV4 knockout mice

Radiotelemetric recording of blood pressure showed that the systolic and diastolic blood pressures, and heart rate were not different in global TRPV4 knockout mice when compared to the wild-type control mice (Supplemental Figure IV). However, PE injection (0.1 mg/kg, intraperitoneal) elevated the mean arterial pressure (Figure 6E), as well as both systolic and diastolic blood pressures (Supplemental Figure IV), to a significantly greater extent in TRPV4 knockout mice when compared to wild-type control mice. The heart rate following PE injections was not different between TRPV4 knockout and control mice (Supplemental Figure IV). These results support an important physiological role for TRPV4 channels in negatively regulating α1-adrenergic signaling-induced increase in blood pressure.

DISCUSSION

Heterocellular communication between SMCs and ECs is crucial for the regulation of blood pressure and blood flow on a moment-to-moment basis. One EC is structurally surrounded by several SMCs in the arteriolar wall⁴⁶, thus uniquely positioned to receive signals from SMCs to modulate their contractile state. While the role of endothelium in moderating nerve stimulation-induced vasoconstriction is well-established^{7–10, 17, 34}, the individual signaling elements involved in this negative feedback regulation of vasoconstriction remain unclear. In this study we provide the first evidence that SMC G_q protein signaling activates Ca²⁺ influx through endothelial TRPV4 channels to negatively regulate vasoconstriction. Moreover, activation of endothelial TRPV4 channels by PE appears to be independent of increases in SMC Ca²⁺ and may represent direct activation of TRPV4 channels by IP₃ transfer from SMCs to ECs via MEGJs. SMC G_qPCR activation has emerged as an important mechanism for vasoconstriction in response to intravascular pressure and neurohumoral mediators^1–6. Activation of TRPV4 channels in ECs represents an immediate feedback mechanism for the regulation of vasoconstriction caused by SMC G_qPCR activation. The negative feedback regulation of vasoconstriction mainly involves activation of endothelial TRPV4-IK/SK signaling. Several vascular disorders including hypertension are characterized by sympathetic overactivation and endothelial dysfunction^{47, 48}. Moreover, the function of endothelial TRPV4 channels is impaired in angiotensin II-induced hypertension¹⁴. An impairment of TRPV4 channel-dependent feedback regulation of vasoconstriction may result in higher vasoconstriction and exacerbate the effects of endothelial dysfunction in vascular disorders.

Intracellular Ca²⁺ plays a pivotal role in regulating diverse functions of endothelium^{11, 12, 18, 49, 50}. Previous studies of negative feedback regulation of vasoconstriction have mostly focused on IP₃R-mediated Ca²⁺ release in ECs. While endothelial TRPV4 channels have been shown to regulate endothelium-dependent vasodilation in different vascular beds^{11, 14, 18, 19}, their role in negative feedback regulation of vasoconstriction remains unknown. Our results represent the first evidence for IP₃R and TRPV4 Ca²⁺ signals as two main Ca²⁺ signals in native endothelium under basal conditions. Distinct kinetic and pharmacological properties of these two signals can be utilized to obtain greater mechanistic details on how Ca²⁺ regulates endothelial function under normal and disease conditions. Moreover, an ability to simultaneously record the two signals enables the studies on how they are modulated by physiological stimuli.

MEGJs, an essential component of heterocellular coupling between ECs and SMCs, allow the transfer of small signaling molecules (i.e. Ca²⁺ and IP₃; below 1 KDa of molecular mass)^41–43. Previous studies have postulated that Ca²⁺ and IP₃ produced by SMC G_qPCR signaling diffuse to ECs through MEGJs and alter endothelial Ca²⁺ signaling^{7, 8, 10, 16, 17, 34}. A recent study by Garland et al.¹⁰ suggested that Ca²⁺ influx through SMC voltage-dependent L-type Ca²⁺ channels (VDCCs) increases EC Ca²⁺ signals. While the current study does not exclude the possibility of the Ca²⁺ diffusing across MEGJs, an increase in SMC VDCC activity was unable to stimulate endothelial TRPV4 sparklets (Figure 3D). Moreover, a similar activation of TRPV4 sparklets by PE or U46619 in the presence or absence of VDCC blocker nifedipine (Supplemental Figure V) rules out a role for SMC VDCCs in PE- or U46619-activation of TRPV4 sparklets. Garland et al.¹⁰ demonstrated that nifedipine completely inhibited the effect of 0.3 μM PE, but not 3 μM PE, on endothelial Ca²⁺ signals, suggesting VDCC-independent mechanism for the activation of endothelial Ca²⁺ signals at higher PE concentration (3 μM). Consistent with these data, we find that PE does not activate endothelial TRPV4 sparklets at 0.3 μM (Supplemental Figure VI), and TRPV4 sparklet activation by 1–10 μM PE is not dependent on VDCCs. Thus, it is possible that the activation of endothelial TRPV4 channels by SMC GqPCR signaling is dependent on the concentration of GqPCR agonist, as seen with 1–10 μM PE or 1 μM U46619.

Additionally, lithium chloride (inositol monophosphatase inhibitor, Figure 4A), PLC inhibition (Figure 3A), and uncoupling of gap junctions (Figure 3B, Supplemental Figure I) attenuated the activation of TRPV4 sparklets by PE, suggesting that IP₃ generated by SMC G_qPCR signaling may cross over to the EC side and activate TRPV4 sparklets. This interpretation is consistent with previous studies demonstrating that IP₃ is the key molecule delivered to ECs after α₁AR stimulation and results in localized Ca²⁺ signals-Ca²⁺ pulsars or wavelets-in ECs^{8, 9, 51}. In conjunction with the data that IP₃ activated TRPV4 sparklets, these results support the idea that IP₃ crossing over to the EC side not only activates IP₃R-mediated Ca²⁺ signals but also stimulates TRPV4 channels. In this regard, glycyrrhetinic acids (GAs, 18αGA and 18βGA) have been used to disrupt gap junction coupling between SMCs and ECs in arteries^52–54. In particular, the efficacy and potency of 18βGA for blocking vascular gap junctions were found to be much higher than that of 18αGA²⁸; therefore, 18βGA was employed in this study to examine whether the delivery of SMC-derived second messengers evokes endothelial TRPV4 channel activity. Heptanol has also been used extensively to suppress gap junction communication^30–32. While 18βGA and heptanol (Supplemental Figure I) did not alter basal TRPV4 sparklet activity or SMC Ca²⁺, both the compounds abolished PE-induced activation of TRPV4 sparklets, supporting their effectiveness as MEGJ uncoupling agents.

Our results provide the first evidence of IP₃ as an endogenous activator for TRPV4 channel function in native endothelium. The IP₃-induced increase in TRPV4 sparklet activity was independent of IP₃R-mediated Ca²⁺ release from the ER (Figure 1D, Figure 2). How IP₃ activates TRPV4 channels remains unknown. A recent study delineated that the ankyrin repeat domain (ARD) of TRPV4 channels has an IP₃ binding sites and binding of IP₃ to this site increases TRPV4 channel activity²⁶. In addition, IP₃ application considerably increased whole-cell currents evoked by 4α-PDD (an activator of TRPV4 channels) in HEK293 cells²⁶. TRPV4 sparklets represent bursts of openings of TRPV4 channels in a cluster^{39, 40}. Our results indicate that IP₃ increases the mean burst open times for TRPV4 channels, but does not alter the cooperative openings among the channels in a cluster. We, therefore, postulate that SMC-derived IP₃ stimulates endothelial TRPV4 channels via a direct interaction that increases the channel open times.

Some of the previous studies have proposed a direct movement of free Ca²⁺ from SMCs to ECs as a mechanism for myoendothelial feedback regulation of vasoconstriction^{7, 10}. In the current study an increase in SMC Ca²⁺ did not increase endothelial TRPV4 sparklet activity, a result that is in agreement with previous investigations suggesting that an increase in SMC Ca²⁺ does not alter EC Ca^{2+ 8, 34}. Although there is evidence suggesting that SMC Ca²⁺ diffuses to EC^{7, 10}, any possible diffusion of Ca²⁺ appears to be insufficient to stimulate endothelial TRPV4 channels. Furthermore, a similar activation of endothelial TRPV4 sparklets by PE in the absence or presence of CPA (Figure 1D, Figure 2B), EGTA-AM (Figure 2D), and nifedipine (Supplemental Figure V), and a similar vasodilation to PE in the presence of CPA rule out the role of increased intracellular Ca²⁺ in PE-activation of TRPV4 sparklets. Ca²⁺-independent heterocellular communication between SMCs and ECs could be explained by the following possibilities: 1) The mobility of IP₃ is significantly higher than Ca²⁺ because of vastly different diffusion coefficients of IP₃ and Ca²⁺ (i.e. ~280 vs. 38 μm²/s)⁵⁵. Free Ca²⁺ promptly interacts with Ca²⁺-binding proteins including calmodulin, thereby limiting the mobility of Ca^{2+ 55}. While Ca²⁺-calmodulin complex is too large to be diffused across narrow pores of the MEGJs, SMC-derived IP₃ can freely move to the ECs and participate in myoendothelial feedback regulation. Notably, a recent study showed that the diffusion constant of IP₃ is less than 10 μm²/s in SH-SY5Y neuroblastoma cells⁵⁶. However, further studies and computational modeling are required to confirm the 30-fold slower IP₃ diffusion rate since it could be affected by IP₃R buffering capacity, intracellular IP₃ concentration, photorelease of IP₃, or cell size/subcellular structure⁵⁷. 2) MEGJs are mainly comprised of connexin 37 or 40^{58, 59}. Connexin composition of MEGJs is a key determinant of selective permeability of the MEGJs to IP₃ and Ca^2+60–62. It is plausible that the MEGJs in mouse MAs are selectively permeable to IP₃. Taken together, these results suggest that an increase in SMC Ca²⁺ is not an essential component for TRPV4 channel activation by PE.

PE-induced vasoconstriction is followed by slow vasodilation^{7, 44, 45}, a response that is attributed to myoendothelial feedback regulation. Specifically, PE induces an increase in endothelial Ca²⁺ that activates vasodilatory mechanisms including NO or IK/SK channel-mediated EDH^{8–10, 17}. While PE treatment (1–10 μM) consistently induced the biphasic vasomotor responses in pressurized MAs, the vasodilatory component was abolished in the presence of IK/SK channel inhibitors. It is well-documented that endothelium-dependent vasodilation of resistance-sized arteries is primarily controlled by EDH, whereas NO and/or prostacyclin are the prevailing vasodilators in larger, conduit arteries^{63, 64}. The number of MEPs increases inversely with vessel size¹⁵, supporting the notion that MEP-localized EDH signaling plays a crucial role in resistance-sized arteries. MEPs serve as the cellular platforms that house the biological apparatus including IK/SK channels and MEGJs necessary for EDH-mediated vasodilation^11–14. While TRPV4 sparklets can occur at MEP and non-MEP sites¹⁴, our results that PE activated TRPV4 sparklets selectively at the MEPs support SMC-EC communication at the MEPs and subsequent activation of EDH signaling. It is likely that IP₃ generated by SMC G_qPCR signaling locally diffuses to ECs via MEGJs and initiates EDH signaling by activating TRPV4 channels at the MEPs. Additional inhibition of vasodilation by NOS inhibitor in the presence of IK/SK inhibitors may be an indication for a minor role of NO in myoendothelial feedback regulation in MAs (Figure 6). Compared to the TRPV4 channel inhibitor, IK/SK channel inhibitors in the absence or presence of L-NNA had a larger effect on the vasodilatory component following PE, suggesting that the IP₃R Ca²⁺ signal activated by PE may couple with IK/SK channels and eNOS (Figure 6).

Several limitations of the present investigation need to be acknowledged: 1) There is no tracking of IP₃ as it moves from SMCs to ECs. While this will be an important area for future investigation, the currently available tools do not allow for accurate tracking of IP₃ in intact arteries. This will require development of reliable IP₃ probes or isotope labeling of SMC IP₃. 2) Despite the many advantages of en face preparation (more ECs per field of view, higher signal/noise at high imaging speed), it is likely that SMCs and ECs experience different stretch levels in pressurized and en face preparations, and therefore, have different resting membrane potentials. It was observed that the membrane potential of endothelial cells in en face arteries is ~ −45 mV¹². The SMC membrane potential in en face preparation, and EC membrane potential in pressurized arteries are not known. Although the results in en face preparation are consistent with those from diameter studies in pressurized arteries, minor differences in signaling mechanisms due to differences in membrane potentials cannot be ruled out. 3) The Ca²⁺ images represent snapshots of 1 minute. In this duration, some cells show Ca²⁺ signals, while others are quiescent. Whether this is due to endothelial cell heterogeneity or is a function of probability of finding active sites during the recording duration remains unknown. Definitively addressing this possibility may require data acquisition for several hours. The high imaging rates required for recording TRPV4 sparklets and IP₃ receptor events and the possibility of bleaching the fluorophores or damaging the endothelium limit our ability to record Ca²⁺ signals for several hours. 4) Endothelium-specific TRPV4 knockout mice will provide the most direct evidence for the role of TRPV4 channels in negatively regulating PE-induced vasoconstriction and increase in blood pressure. However, such mice have not been generated. 5) The number of TRPV4 sparklet sites activated by PE and U46619 appear to be different. This could be due to differences in the number of sparklet sites under baseline conditions. Indeed, the number of TRPV4 sparklet sites varies from one field to another, while the sparklet activity/site remains unchanged.

In conclusion, our studies reveal that distinct endothelial Ca²⁺ signals could be isolated and analyzed under physiological conditions using differential kinetics and pharmacology. Moreover, the study also provides first evidence for a novel endothelial mechanism that acts as a feedback regulator for vasoconstriction induced by SMC G_qPCR activation. In agreement with previous studies, this mechanism involves heterocellular communication and transfer of second messengers from SMCs to ECs. This is followed by the activation of EC TRPV4 channels, possibly through the movement of IP₃ from SMCs to ECs and a direct interaction of IP₃ with TRPV4 channels. G_qPCR activation is an important mechanism for SMC contraction. Indeed, numerous physiological vasoconstrictors including intravascular pressure (myogenic tone), α₁AR activation, TXA₂R activation, and angiotensin II receptor activation involve G_q protein signaling in SMCs. Activation of EC TRPV4 channels may be a key mechanism that negatively regulates vasoconstriction to these stimuli, thus preventing excessive vasoconstriction. An abnormal TRPV4 channel-dependent negative feedback regulation of vasoconstriction may result in elevated blood pressure, and therefore is a potential mechanism that could be targeted for treating hypertension.

HIGHLIGHTS.

1. Ca²⁺ influx signals through TRPV4 channels can be distinguished from IP₃R-mediated Ca²⁺ release from the ER in the native endothelium from resistance-sized arteries using distinct kinetics and pharmacology of the two signals.

2. Activation of SMC G_qPCRs initiates SMC-EC communication that stimulates endothelial TRPV4 channel function.

3. SMC-derived IP₃, rather than Ca²⁺, activates endothelial TRPV4 sparklets in response to SMC G_qPCR activation.

4. Endothelial TRPV4 sparklets negatively regulate vasoconstriction caused by the activation of SMC G_qPCRs via TRPV4-IK/SK channel signalling. This signalling pathway represents a novel mechanism for regulation of blood pressure in response to SMC G_qPCR activation.

Abbreviations

α₁ARs

alpha₁-adrenergic receptors

ECs

endothelial cells

EDH

endothelium-derived hyperpolarization

ER

endoplasmic reticulum

G_qPCRs

G_q protein-coupled receptors

IK/SK

intermediate/small conductance Ca²⁺-sensitive K⁺ channels

IP₃

inositol 1,4,5-trisphosphate

IP₃Rs

IP₃ receptors

MEGJs

Myoendothelial gap junctions

MEPs

Myoendothelial projections

NOS

nitric oxide synthase

PE

phenylephrine

PLC

phospholipase C

SMCs

vascular smooth muscle cells

TRPV4

transient receptor potential vanilloid 4

TXA₂R

thromboxane A₂ receptors

VDCCs

voltage-dependent L-type Ca²⁺ channels