Ca2+-activated potassium channels
bovine serum albumin
Ca2+-induced Ca2+ release
full duration at half-maximal amplitude
full width at half-maximal amplitude
inositol trisphosphate receptor
no reverse transcriptase
physiological salt solution
region of interest
smooth endoplasmic reticulum Ca2+ ATPase
smooth muscle cell
- • Feed arteries and arterioles, respectively, control the magnitude and distribution of blood flow to skeletal muscle but regional differences in the regulation of vasomotor tone are poorly understood.
- • To provide this insight, we investigated functional roles and molecular expression of the calcium-release channels, ryanodine receptors (RyRs) and inositol 1,4,5-trisphosphate receptors (IP3Rs) in smooth muscle cells (SMCs) of isolated pressurized vessels of mice.
- • In feed arteries, SMCs displayed localized calcium sparks and more global calcium waves. In arterioles, SMCs exhibited only calcium waves.
- • Calcium signalling and vasomotor tone were governed by both RyRs and IP3Rs in feed arteries, while only IP3Rs were functional in arterioles. Regional differences were also manifest in the expression profile of RyR isoforms.
- • This new perspective offers the potential for developing novel strategies to target therapeutic interventions to selective regions of vascular beds.
Abstract We tested the hypothesis that vasomotor control is differentially regulated between feed arteries and downstream arterioles from the cremaster muscle of C57BL/6 mice. In isolated pressurized arteries, confocal Ca2+ imaging of smooth muscle cells (SMCs) revealed Ca2+ sparks and Ca2+ waves. Ryanodine receptor (RyR) antagonists (ryanodine and tetracaine) inhibited both sparks and waves but increased global Ca2+ and myogenic tone. In arterioles, SMCs exhibited only Ca2+ waves that were insensitive to ryanodine or tetracaine. Pharmacological interventions indicated that RyRs are functionally coupled to large-conductance, Ca2+-activated K+ channels (BKCa) in SMCs of arteries, whereas BKCa appear functionally coupled to voltage-gated Ca2+ channels in SMCs of arterioles. Inositol 1,4,5-trisphosphate receptor (IP3R) antagonists (xestospongin D or 2-aminoethoxydiphenyl borate) or a phospholipase C inhibitor (U73122) attenuated Ca2+ waves, global Ca2+ and myogenic tone in arteries and arterioles but had no effect on arterial sparks. Real-time PCR of isolated SMCs revealed RyR2 as the most abundant isoform transcript; arteries expressed twice the RyR2 but only 65% the RyR3 of arterioles and neither vessel expressed RyR1. Immunofluorescent localisation of RyR protein indicated bright, clustered staining of arterial SMCs in contrast to diffuse staining in arteriolar SMCs. Expression of IP3R transcripts and protein immunofluorescence were similar in SMCs of both vessels with IP3R1>>IP3R2>IP3R3. Despite similar expression of IP3Rs and dependence of Ca2+ waves on IP3Rs, these data illustrate pronounced regional heterogeneity in function and expression of RyRs between SMCs of the same vascular resistance network. We conclude that vasomotor control is differentially regulated in feed arteries vs. downstream arterioles.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The control of tissue blood flow is coordinated between resistance arteries upstream from the microcirculation and arterioles embedded within the tissue (Segal, 2005). Throughout resistance networks, relaxation and contraction of smooth muscle cells (SMCs) increases or decreases blood flow, respectively. The regulation of SMC intracellular Ca2+ ([Ca2+]i) is integral to vasomotor control. In resistance arteries, ryanodine receptors (RyRs) (Nelson et al. 1995; Knot et al. 1998; Gollasch et al. 2000; Westcott & Jackson, 2011) and inositol 1,4,5-trisphosphate receptors (IP3Rs) (Zhao et al. 2008; Mufti et al. 2010; Westcott & Jackson, 2011) contribute importantly to Ca2+ signalling underlying myogenic tone. Calcium released from RyRs as localised ‘sparks’ controls the open-state probability of large-conductance, Ca2+-activated K+ channels (BKCa), providing an important negative feedback signal to contraction (Nelson et al. 1995; Knot et al. 1998; Gollasch et al. 2000; Westcott & Jackson, 2011). Ryanodine receptors may also contribute to more global intracellular Ca2+ signals such as Ca2+ waves (Collier et al. 2000; Gordienko & Bolton, 2002; Tumelty et al. 2011; Westcott & Jackson, 2011) via Ca2+-induced Ca2+ release (CICR).
In rat retinal arterioles, RyRs underlie Ca2+ sparks and control BKCa activity of SMCs (Curtis et al. 2004; Tumelty et al. 2007). Nevertheless, because RyRs may amplify Ca2+ signals and contribute to more global Ca2+ events in these cells, an overall excitatory role was proposed (Curtis et al. 2004; Tumelty et al. 2007, 2011), which effectively enhanced myogenic and agonist-induced tone (Fellner & Arendshorst, 2005; Balasubramanian et al. 2007; Fellner & Arendshorst, 2007). In the hamster cremaster muscle, RyRs in SMCs of second-order arterioles were functionally silent as they did not contribute to Ca2+ signals or to myogenic tone (Westcott & Jackson, 2011). However, in feed arteries supplying the same muscle, RyRs provided negative feedback to myogenic tone (Westcott & Jackson, 2011). These observations imply that regional and species-dependent differences can exist in the function of RyRs between resistance arteries upstream from the microcirculation and the downstream arterioles they supply within the tissue.
The mechanisms responsible for the heterogeneous function of RyRs have not been established. Prior studies suggest that differences in RyR isoform expression contribute to heterogeneous function of RyRs (Coussin et al. 2000; Ji et al. 2004; Dabertrand et al. 2006; Zheng et al. 2008). Calcium sparks depend upon the expression of RyR1 or RyR2 (Coussin et al. 2000; Ji et al. 2004), while RyR3 may inhibit Ca2+ sparks (Löhn et al. 2001; Jiang et al. 2003). In SMCs isolated from small pulmonary arteries, large pulmonary arteries and mesenteric arteries, differences in expression levels of RyR isoforms were proposed to account for the heterogeneity in spatiotemporal properties of Ca2+ sparks (Zheng et al. 2008). With the exception of an immunofluorescence study of rat retinal arterioles (Curtis et al. 2008), the expression of RyR isoforms in SMCs of resistance networks has not been examined, nor has there been a comparison of the expression of RyR isoforms between upstream feed arteries and downstream arterioles.
The goal of the present study was to test the hypothesis that vasomotor control is differentially regulated between feed arteries and downstream arterioles of the cremaster muscle of C57BL/6 mice. We specifically questioned whether substantial differences exist in the function of RyRs between upstream feed arteries and downstream arterioles supplying the mouse cremaster muscle. To investigate underlying mechanisms, we used quantitative real-time PCR (RT-PCR) and immunofluorescence to determine whether differences exist in the expression or localisation of RyR isoforms between SMCs of respective resistance vessels. Further, we investigated the function and expression of IP3Rs between SMCs of the same vessels to determine whether regional differences were apparent in this complementary intracellular Ca2+ signalling pathway. Our findings illustrate that regional heterogeneity is manifest in the function and expression of RyRs between SMCs of feed arteries and arterioles supplying mouse skeletal muscle. In contrast, the expression of IP3Rs and dependence of Ca2+ waves upon the function of IP3Rs appear similar between respective resistance vessels. Thus, regional differences in vasomotor control may be determined more by the functional expression of RyRs than by IP3Rs in SMCs of respective branches of the resistance network.
Animal and vessel preparation
All experiments were approved by, and conducted in accordance with, the guidelines of the Institutional Animal Care and Use Committee at Michigan State University, and were performed in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals (2011). Male C57BL/6 mice (8–12 weeks, 20–30 g, Jackson Laboratories, Bar Harbor, MA, USA) were killed by CO2 asphyxiation followed by cervical dislocation. Feed arteries supplying the cremaster muscle (small branches off the iliac artery) were dissected in situ as described in Westcott & Jackson (2011). To obtain second-order arterioles, cremaster muscles were removed and placed in Ca2+-free physiological salt solution (Ca2+-free PSS) containing (in mm): 140 NaCl, 5 KCl, 1 MgCl2, 10 Hepes, 10 glucose (pH 7.4, 295 mosmol (kg H2O)−1) maintained at 4°C. Individual arterioles were hand dissected while viewing through a stereomicroscope as described previously (Burns et al. 2004; Jackson et al. 2008; Westcott & Jackson, 2011). Isolated vessels were transferred to a cannulation chamber using a 50–100 μl Wiretrol pipette (Drummond Scientific, Broomall, PA, USA).
Intact vessels were cannulated onto glass micropipettes positioned within the chamber and secured using 11-0 ophthalmic suture (Ashaway Line and Twine, Ashaway, RI, USA). The chamber was then secured to the stage of a microscope (Leica DMIL, Wetzlar, Germany) where vessels were visualized, warmed and pressurized to 80 cmH2O and allowed to develop myogenic tone (Burns et al. 2004; Jackson et al. 2008; Westcott & Jackson, 2011). All feed arteries and arterioles studied had at least 20% resting myogenic tone compared with their maximum diameters obtained in Ca2+-free PSS (% myogenic tone = (maximum diameter – resting diameter)/maximum diameter × 100%, where maximum diameter is the diameter in Ca2+-free PSS at 80 cmH2O and resting diameter is the diameter in Ca2+-replete PSS at the same pressure). In some experiments, vessels were pressurized to 20 cmH2O and did not develop tone. For functional studies, vessels were superfused continuously with PSS containing Ca2+ (in mm): 140 NaCl, 5 KCl, 1.8 CaCl2, 1 MgCl2, 10 Hepes, 10 glucose (pH 7.4), either alone or containing a pharmacological agent. In studies using more than one pharmacological agent per vessel, the order of drug application was randomized across experiments.
Cannulated vessels were loaded with the intensiometric Ca2+ indicator dye Fluo-4 AM by bath incubation at room temperature. The dye solution consisted of PSS containing 5 μm Fluo-4 AM dye (Invitrogen, Carlsbad, CA, USA) in 0.5% dimethyl sulfoxide (DMSO) and 0.1% bovine serum albumin (BSA; US Biochemical, Cleveland, OH, USA). This solution was added to the cannulation chamber for 2 h, followed by a 30 min period of superfusion with PSS and gradual warming to 37°C (feed arteries) or 34°C (arterioles) (Westcott & Jackson, 2011). All vessels were imaged using a 40× water-immersion objective (numerical aperture, 0.8; working distance, 3 mm; Leica). Fluo-4 fluorescence at 526 nm was acquired at 30 frames s−1 using a spinning-disc confocal system (CSU-10B, Yokagawa Corporation of America, Sugar Land, TX, USA) with 488 nm laser illumination (Solamere, Salt Lake City, UT, USA) and an intensified CCD camera (XR Mega-10, Stanford Photonics, Palo Alto, CA, USA). Each recording was made at 30 frames s−1 for 16.7 s and consisted of 500 frames each 1024 × 1024 pixels (0.17 μm × 0.17 μm per pixel). The Z-resolution with the confocal head (50 μm pinholes) and 40× objective in PSS was 1.77 μm (Westcott & Jackson, 2011). Images were recorded using Piper software (Stanford Photonics) and analysed using SparkAn (courtesy of M. T. Nelson and A. D. Bonev, University of Vermont) and ImageJ (Abramoff et al., 2004) software. The spatiotemporal properties of Ca2+ sparks and waves were quantified by their occurrence (number of SMCs displaying respective events/total number of SMCs viewed), amplitude (F/Fo), frequency (number of events observed per second in a SMC, Hz), full width at half-maximal amplitude (FWHM, μm) and full duration at half-maximal amplitude (FDHM, s) were determined as described previously (Westcott & Jackson, 2011). The observed cell lengths (length of SMCs within the confocal slice) were also measured in both vessel types as a reference dimension for the FWHM of Ca2+ events. The properties of Ca2+ sparks and waves were measured using SparkAn by placing a 10 × 10 pixel ROI on the cells of interest within each vessel and measuring the amplitude and FWHM of events occurring within the recording period. Relative global fluorescence was evaluated by comparing total fluorescence in the presence and absence of each drug studied using ImageJ software. Briefly, a large ROI was placed over all SMCs within the field of view, and total fluorescence was averaged throughout the recording period under both conditions. Global fluorescence during drug treatment was then normalized to the fluorescence recorded under control conditions.
Vessel diameters were measured asynchronously from Ca2+ measurements by using dim transillumination while focusing at the vessel mid-plane as reported in Westcott & Jackson (2011).
Enzymatic isolation of smooth muscle cells for RT-PCR
Mouse cremaster muscle feed arteries or arterioles were dissected as described above and SMCs dissociated as described (Jackson et al. 1997). Briefly, all vessels of a given branch order obtained from one mouse were pooled into 1 ml of dissociation solution (Ca2+-free PSS to which 10 μm sodium nitroprusside, 10 μm diltiazem, 1% BSA and 100 μm CaCl2 were added). Vessels were then incubated for 35 min in a 37°C solution containing 26 Units ml−1 papain and 1 mg ml−1 dithioerythritol followed by incubation for 19 min in a 37°C solution containing 1.95 Units ml−1 collagenase, 0.15 mg ml−1 elastase and 1 mg ml−1 soybean trypsin inhibitor. This solution was then replaced with 4 ml of ice-cold dissociation solution and incubated for 10 min at room temperature to allow vessel segments and dissociated cells to settle. The supernatant was then replaced with 1 ml Ca2+-free PSS. Vessel segments were then triturated using a 1 ml Eppendorf-style pipette to release the SMCs. This solution containing suspended SMCs was transferred to a 35 mm culture dish mounted on the stage of an inverted microscope for harvest of SMCs for quantitative real-time PCR as reported (Hakim et al. 2008).
Isolation and quantitation of RNA
Control tissues (heart, brain, diaphragm and intestinal smooth muscle) were dissected and immersed immediately in RNAlater (Ambion, Austin, TX, USA). Tissues were removed from RNAlater, processed using RNeasy tissue kits (Qiagen, Germantown, MD, USA) according to the manufacturer's protocol and the resultant RNA was quantitated with a Nanodrop 1000 (Thermo Scientific, Wilmington, DE, USA).
Samples of 50 SMCs or an equivalent volume of bath solution (control) were prepared for RT-PCR using the Cells-to-PreAmp CT kit (Applied Biosystems, Foster City, CA, USA, no. 4387299) per manufacturer's instructions. Control tissue RNA isolates were reverse transcribed (2 μg per reaction) using High Capacity RNA to cDNA kits (Applied Biosystems, no. 4387406), per manufacturer's protocol. Resultant cDNA was quantitated with the Nanodrop 1000. Reactions (20 μl) were prepared with Applied Biosystems TaqMan Gene Expression Master Mix (no. 4369016) and inventoried Gene Expression assays (see Table 1 for a list of primers used). Data for RyR1–3 and for IP3R1–3 were generated using independent samples from different animals. All samples were run in triplicate for 40 cycles on an ABI 7500 Thermocycler according to manufacturer's instructions including no reverse transcriptase (NRT) controls. Relative abundance of target RNA was normalized to α-smooth muscle actin RNA, which showed no significant differences among samples. The PCR efficiency (E) was calculated as 10(−1/slope) with the slope estimated from serial dilutions of cDNA from respective control tissues (RyR1, diaphragm; RyR2, heart; RyR3, brain; α-actin, IP3R1, 2 and 3, intestinal smooth muscle, see Table 1). The use of control tissues was necessitated by the low expression level of transcripts in arterial and arteriolar SMCs (Table 2). Efficiency-normalized relative abundance of target RNA was calculated as (Eα-actinCT)/(EtargetCT) (Pfaffl, 2001) where PCR crossing points (CT) were determined using ABI 7500 software (v2.02). When no PCR products were detected within 40 PCR cycles, samples were assigned CT values of 41 so that statistical comparisons of relative abundance could be performed. The n values for RT-PCR experiments refer to the number of independent replicates, each from a different animal.
|Gene product||Gene assaya||Source tissueb||Standard curve CT rangec||Sloped||r 2e||PCR efficiencyf|
|RyR1||Mm01175172_g1||Diaphragm||17–32||–3.46 ± 0.05||0.999||1.95|
|RyR2||Mm00465877_m1||Heart||19–33||–3.31 ± 0.03||0.999||2.01|
|RyR3||Mm01328421_m1||Brain||22–37||–3.54 ± 0.10||0.998||1.92|
|IP3R1||Mm00439917_m1||Intestinal SMCs||24–38||–3.64 ± 0.10||0.986||1.88|
|IP3R2||Mm00444937_m1||Intestinal SMCs||26–37||–3.70 ± 0.07||0.981||1.86|
|IP3R3||Mm00446540_m1||Intestinal SMCs||22–34||–3.71 ± 0.08||0.984||1.86|
|α-actin||Mm01204962_gH||Intestinal SMCs||21–36||–3.73 ± 0.05||0.999||1.85|
|Gene product||Reference tissuea||CT for reference tissueb||CT for arteriolar SMC||CT for feed artery SMC|
|RyR1||Diaphragm||17.1 ± 0.8 (3)||>40 (8)c||>40 (8)|
|RyR2||Heart||20.0 ± 0.5 (3)||> 29.4 ± 0.5 (7)d||28.5 ± 0.5 (8)|
|RyR3||Brain||22.8 ± 0.1 (3)||> 32.1 ± 0.5 (7)d||> 32.3 ± 0.3 (7)d|
|IP3R1||Intestinal SMCs||24.3 ± .5 (5)||31.7 ± 0.4 (5)||30.7 ± 0.4 (5)|
|IP3R2||Intestinal SMCs||26.1 ± 0.3 (5)||> 36.0 ± 0.5 (3)e||35.4 ± 0.6 (5)|
|IP3R3||Intestinal SMCs||22.8 ± 0.3 (5)||>40 (5)f||> 34.0 ± 0.6 (2)g|
|α-actin||Intestinal SMCs||21.3 ± 0.2 (3)||22.9 ± 0.5 (13)||22.5 ± 0.4 (13)|
Individual feed arteries or cremaster arterioles were isolated as above except that BSA was omitted from the dissection solution. Vessels were cannulated and pressurized as described above (see Vessel cannulation, above), then fixed in 4% paraformaldehyde for 20 min while remaining cannulated and pressurized. After three washes in phosphate-buffered saline (PBS), vessels were removed and pinned onto Sylgard 184 (Dow Corning, Midland, MI, USA) within 35 mm culture dishes. Incubation in all subsequent solutions took place at 4°C with primary and secondary antibodies diluted in PBS + 0.1% Triton X-100 to permeabilise cells. Vessels were incubated for 60 min in 10% normal goat serum, then washed 3 times in PBS and placed overnight in solutions containing the primary antibody against RyR1/2 (mouse monoclonal, 1:500, R-129, Sigma, St Louis, MO, USA), IP3R1 (mouse monoclonal, 1:500, clone L24/18, Neuromab, Davis, CA, USA), IP3R2 (rabbit polyclonal, 1:500, AB9074, Millipore, Billerica, MA, USA) or IP3R3 (rabbit polyclonal, 1:500, AB9076, Millipore). Vessels were then washed with PBS, placed into PBS + 10% normal goat serum for 60 min, washed again in PBS, and incubated for 90 min in Alexa Fluor-488-labelled goat secondary antibodies (1:500, Invitrogen). After a final wash in PBS, slides were mounted using ProLong Gold with DAPI (Invitrogen) to stain cell nuclei to aid in the identification of the position of confocal slices (not shown). Coverslips were positioned and sealed in place using clear nail polish.
In control experiments, mouse hepatocytes were isolated and dissociated as described in Kim et al. (2006) then 400 μl of the cell suspensions were spun onto glass microscope slides (Shandon Cytospin 4 Centrifuge; Thermo Scientific, Ashville, NC, USA) for 3 min at 64 × g. Cells on the slides then were fixed with 4% paraformaldehyde for 20 min followed by three washes in PBS, then stained for immunofluorescence as described above.
All slides were imaged using an Olympus FluoView FV1000 confocal laser-scanning microscope using an Olympus PLAPON 60× oil immersion objective (numerical aperture, 1.42) at the Michigan State University Centre for Advanced Microscopy. Whole vessels were imaged with a 2× optical zoom and 2 μm Z-slices. Hepatocytes were imaged with a 1× optical zoom and 0.5 μm Z-slices.
Expression patterns for RyRs and IP3Rs in SMCs of intact vessels and the fluorescence intensity of IP3R staining within primary hepatocytes were quantified using ImageJ (Abramoff et al., 2004). To analyse the expression pattern in SMCs, a reference line was drawn along a tangential confocal optical slice through a vessel to generate a profile of the intensity of immunofluorescent staining along the reference line. As an index of the (non-)uniformity of immunofluorescence, these profile data were used to determine the coefficient of intensity variation (standard deviation divided by mean). Mean fluorescence intensity values were determined by placing a region of interest (ROI) over the entire vessel or group of hepatocytes and the mean intensity within the ROI measured. For IP3Rs, these values then were normalized to the fluorescence from samples stained for IP3R1.
Ryanodine was obtained from Ascent Scientific (Bristol, UK), xestospongin D and 2-APB from Calbiochem (San Diego, CA, USA), and paxilline from Tocris (Ellisville, MO, USA). Fluo-4 was obtained from Invitrogen. All other reagents were obtained from Sigma-Aldrich. All drugs were dissolved in DMSO and then diluted to their final concentrations in PSS. All working solutions contained <0.1% DMSO, with the exception of U73122 and U73343, which contained 1% DMSO. Vehicle controls confirmed that 1% DMSO alone had no effect on the diameters of cannulated vessels.
Data analysis and statistics
Data are shown as means ± 95% confidence intervals for the occurrence of Ca2+ sparks and waves, or as means ± SEM for all other data. Relative expression data were screened for outliers using Grubb's test (http://www.graphpad.com/quickcalcs/Grubbs1.cfm), and identified outliers were excluded from statistical comparisons. No more than one outlier was identified, and thus excluded from consideration from any data set. Statistical significance was determined using Student's t tests or analysis of variance followed by post hoc Tukey's tests using Microsoft Excel and Graphpad Prism (La Jolla, CA, USA), respectively. All statistical comparisons were performed at the 95% confidence level.
Characterization of Ca2+ signals in feed arteries and arterioles
Smooth muscle cells from feed arteries loaded with Fluo-4 displayed two types of Ca2+ signals under control conditions: local microdomain increases in [Ca2+]i and more global elevations in [Ca2+]i (Fig. 1 and Supplemental movies S1 and S2). These differences in spatial domains within SMCs (along with pharmacological evidence below) defined whether a given Ca2+ event was a spark or a wave, respectively (Westcott & Jackson, 2011). The spatiotemporal characteristics of Ca2+ sparks and waves from feed arteries are shown in Fig. 1 and listed in Table 3 (also see Supplemental movies S1 and S2). In addition to having a smaller spatial spread (FWHM), Ca2+ sparks were of shorter duration (FDHM) than waves. In contrast to feed arteries, SMCs of arterioles loaded with Fluo-4 exhibited only Ca2+ waves with no evidence of Ca2+ sparks (239 SMCs from 13 arterioles; see Supplemental movie S3). These data imply that fundamental differences exist in the nature of Ca2+ signals between SMCs of cremaster feed arteries and arterioles in male C57BL/6 mice at 8–12 weeks of age.
|Observed cell length||FWHM||Frequency||Amplitude|
|Feed arteries, sparks||72.0 ± 2.5||5.7 ± 0.5||0.14 ± 0.02||1.54 ± 0.05||0.60 ± 0.05||15|
|Feed arteries, waves||74.0 ± 1.9||46.6 ± 1.9*||0.22 ± 0.02||1.67 ± 0.05||1.15 ± 0.19*||15|
|Arterioles, waves||23.5 ± 0.6†||17.6 ± 0.6*†||0.52 ± 0.08*||1.82 ± 0.04*||0.87 ± 0.08*||13|
Role of ryanodine receptors in cremaster feed arteries and arterioles
In feed arteries, the occurrence of Ca2+ waves (Fig. 2A) and Ca2+ sparks (Fig. 2B) was nearly abolished when exposed to the RyR antagonist ryanodine (10 μm), suggesting that RyRs contribute to both Ca2+ signals in SMCs of these vessels. In concert with this inhibition of spatiotemporal Ca2+ signals, total Fluo-4 fluorescence increased (Fig. 2C and E), suggesting a rise in global [Ca2+]i levels produced by ryanodine. Consistent with this interpretation, ryanodine also produced constriction (Fig. 2D and F). Similar results were obtained using tetracaine (100 μm), a RyR antagonist that does not deplete intracellular Ca2+ stores (Curtis et al. 2008). In the presence of tetracaine (100 μm) the occurrence of Ca2+ waves decreased from 46 ± 11% of 79 cells to 7 ± 8% of 85 cells, the amplitude of the few remaining waves decreased from 1.78 ± 0.09 F/Fo to 1.31 ± 0.03 F/Fo and their frequency decreased from 0.29 ± 0.04 Hz to 0.05 ± 0.01 Hz (n= 4, P < 0.05 for each comparison). Tetracaine abolished Ca2+ sparks. Prior to tetracaine exposure, Ca2+ sparks were observed in 26 ± 11% of 78 SMCs examined, with mean amplitudes of 1.5 ± 0.08 F/Fo and frequencies of 0.18 ± 0.02 Hz (n= 4). In the presence of tetracaine, no Ca2+ sparks were observed in 85 SMCs of 4 feed arteries. Similar to the actions of ryanodine, the inhibition of Ca2+ waves and sparks with tetracaine was accompanied by both a rise in global [Ca2+]i (1.56 ± 0.03-fold increase in global Fluo-4 intensity, n= 4, P < 0.05) and vasoconstriction (diameter decreased from 167 ± 6 μm to 131 ± 10 μm; n= 4, P < 0.05).
Complementary experiments in arterioles yielded distinctly different results. Ryanodine (10 μm) had no significant effect on Ca2+ wave parameters, global [Ca2+]i or vessel diameter, suggesting that RyRs are silent in the SMCs of these microvessels (Fig. 3). Similar results were obtained using tetracaine (100 μm). Calcium wave occurrence (35 ± 11% of 74 SMCs in PSS vs. 32 ± 11% of 75 SMCs in tetracaine), amplitude (1.78 ± 0.14 F/Fo in PSS vs. 1.66 ± 0.03 F/Fo in tetracaine) and frequency (0.18 ± 0.03 Hz in PSS vs. 0.20 ± 0.02 Hz in tetracaine) were not significantly affected (n= 4, P > 0.05 for each comparison). Global Ca2+ (Fluo-4 intensity in tetracaine relative to PSS = 1.14 ± 10) and arteriolar diameter (26 ± 1 μm in PSS vs. 26 ± 1 μm in tetracaine) likewise were not significantly affected by tetracaine (n= 4, P > 0.05 for both comparisons).
Functional coupling of RyRs to BKCa in feed arteries but not arterioles
The increase in global [Ca2+]i (Fig. 2C and E) and the constriction induced by ryanodine (Fig. 2D and F) or tetracaine (see data above) with loss of Ca2+ sparks (Fig. 2B and data above) in feed arteries are consistent with inhibition of the negative feedback function of RyRs in resistance artery SMCs (Nelson et al. 1995; Jaggar et al. 1998; Knot et al. 1998; Westcott & Jackson, 2011). To test this relationship in murine resistance vessels, we probed for functional coupling between RyRs and BKCa by examining interactions between ryanodine and BKCa blockers. When applied alone, the BKCa antagonist paxilline (100 nm; Fig. 4A) or the ryanodine receptor antagonist ryanodine (10 μm; Fig. 4B) each caused significant and comparable constriction of feed arteries. However, in the presence of paxilline, application of ryanodine did not cause additional constriction (Fig. 4A). Similar results were obtained using tetraethylammonium (TEA, 1 mm) to block BKCa: TEA alone constricted feed arteries from a resting diameter of 53 ± 4 μm to 34 ± 2 μm (n= 5, P < 0.05). Subsequent addition of ryanodine (10 μm) in the presence of TEA had no further effect on diameter (32 ± 2 μm; n= 5, P > 0.05). In the same vessels, ryanodine alone constricted the vessels by 14 ± 2 μm (n= 5). Thus, as with paxilline, pre-treatment with TEA inhibited ryanodine-induced constriction. These data suggest that RyRs contribute to the regulation of myogenic tone through their control of BKCa function.
The inhibition of constriction to ryanodine in feed arteries that were first exposed to paxilline or TEA (Fig. 4A and data above) was not attributable to the initial constriction induced by these BKCa blockers, because a similar degree of constriction induced by the voltage-gated Ca2+ channel agonist Bay K 8644 (5 nm), did not prevent ryanodine from producing an additional decrease in diameter: feed arteries constricted from 142 ± 11 μm to 104 ± 11 μm upon exposure to 5 nm Bay K 8644 (n= 4, P < 0.05). Subsequent addition of ryanodine (10 μm), in the presence of Bay 8644, further reduced feed artery diameter to 75 ± 15 μm (n= 4, P < 0.05).
Reversing the order of paxilline and ryanodine exposure had differential effects: when ryanodine was applied first, subsequent addition of paxilline produced significant additional constriction (Fig. 4B). Thus, ryanodine did not eliminate the response to paxilline. These data suggest that in addition to control exerted by RyR-dependent Ca2+ signals, BKCa are probably controlled by additional factors, such as Ca2+ influx through voltage-gated Ca2+ channels (Guia et al. 1999; Westcott & Jackson, 2011).
Although ryanodine alone had no effect on the diameter of arterioles (Fig. 3), these microvessels responded to BKCa blockade. Paxilline (100 nm) constricted arterioles by 49 ± 4% (n= 4, P < 0.05) with no further effect of ryanodine (Fig. 5A). As shown in Fig. 5B, paxilline was still able to constrict arterioles in the presence of ryanodine (10 μm). Similar results were obtained with TEA (1 mm): this BKCa blocker constricted arterioles from 24 ± 2 μm to 12 ± 2 μm (n= 5, P < 0.05). After washout of TEA, diameter returned to 23 ± 2 μm. Subsequent addition of ryanodine (10 μm) was without effect, yet the addition of TEA in the presence of ryanodine constricted arterioles to 11 ± 2 μm (n= 5, P < 0.05), which was not different from the response to TEA alone (P > 0.05).
As ryanodine did not affect Ca2+ signals, resting diameter or responses to BKCa blockers in arterioles, we tested for the presence of functional RyRs in these SMCs using the RyR agonist caffeine (Westcott & Jackson, 2011). Caffeine (10 mm) constricted arterioles by 11 ± 2 μm (n= 5, P < 0.05) from a resting diameter of 26 ± 2 μm, confirming functional RyRs. Although ryanodine (10 μm) alone had no effect on arteriolar diameter (27 ± 3 μm; n= 5, P > 0.05), it abolished constriction induced by caffeine (diameter change, 0.7 ± 0.4 μm; n= 5, P > 0.05). These data indicate that functional RyRs are present in SMCs of second-order cremaster arterioles of the mouse. Though these RyRs are apparently ‘silent’, ryanodine can effectively block responses to their activation by caffeine.
In the absence of regulation by Ca2+ released from internal stores, BKCa may be regulated by Ca2+ influx through voltage-gated Ca2+ channels (Westcott & Jackson, 2011). To test this hypothesis in the mouse, we studied the effects of the BKCa blocker TEA (1 mm) on the diameter of cremaster muscle feed arteries (Fig. 6A) and second-order arterioles (Fig. 6B) in the absence or presence of the voltage-gated Ca2+ channel agonist Bay K 8644 (5 nm). Vessels were studied at low intravascular pressure (20 cmH2O) to preclude myogenic tone (Westcott & Jackson, 2011) as voltage-gated Ca2+ channels should then be inactive (Brayden & Nelson, 1992). This assumption was confirmed here by removal of extracellular Ca2+: in the presence of PSS-containing Ca2+, diameter was 42 ± 3 μm, and in PSS with 0 Ca2+ diameter was 45 ± 3 μm (n= 5, P > 0.05). Similar results were obtained in arterioles: in the presence of PSS-containing Ca2+, diameter was 20 ± 2 μm, while in PSS with 0 Ca2+ diameter was 21 ± 3 μm (n= 5, P > 0.05). With 20 cmH2O transmural pressure in PSS-containing Ca2+, TEA (1 mm) had no effect on the resting diameter of either feed arteries (Fig. 6A) or arterioles (Fig. 6B) as predicted for vessels that do not have myogenic tone. Nevertheless, exposure to Bay K 8644 constricted feed arteries and arterioles significantly (Fig. 6A and B). Moreover, subsequent application of TEA (1 mm), in the presence of Bay K 8644 produced even greater constriction (Fig. 6A and B), supporting the hypothesis that BKCa activity is linked to that of voltage-gated Ca2+ channels in SMCs of both vessels.
In other types of vascular smooth muscle, activation of voltage-gated Ca2+ channels leads to an increase in RyR activity (Jaggar et al. 1998; Essin et al. 2007). To test whether RyRs influenced the apparent activity of BKCa in the presence of Bay K 8644, these experiments were repeated after pre-treatment with ryanodine (10 μm). In feed arteries (Fig. 6A) and arterioles (Fig. 6B), ryanodine had no significant effect on constriction to either Bay K 8644 or subsequent exposure to TEA. These data indicate that the ability of Bay K 8644 to impart reactivity to TEA does not depend on RyRs. In turn, we propose that BKCa activity can be coupled to the activity of voltage-gated Ca2+ channels in murine cremaster muscle feed arteries and second-order arterioles independent from RyRs.
Role of IP3Rs and phospholipase C in feed arteries and arterioles
Inositol trisphosphate receptors provide an alternate signalling pathway for Ca2+ release from intracellular stores and these Ca2+ channels are governed by IP3 generated by PLC activity (Hill et al. 2001; Potocnik & Hill, 2001). In SMCs of feed arteries, application of the IP3R antagonist xestospongin D (5 μm) significantly decreased the occurrence of Ca2+ waves and reduced the amplitude and frequency of remaining waves (Fig. 7A). In contrast, Ca2+ sparks were unaffected (Fig. 7B). Xestospongin D also decreased global Ca2+ (Fig. 7C) and produced vasodilatation (Fig. 7D). To independently confirm a role for IP3Rs, experiments were repeated with a structurally different IP3R antagonist, 2-aminoethoxydiphenyl borate (2-APB). Treatment with 2-APB (100 μm) also inhibited Ca2+ wave occurrence, amplitude and frequency, reduced global [Ca2+]i and attenuated myogenic tone with no significant effect on Ca2+ sparks (Fig. 7E–H). The PLC antagonist U73122 (10 μm), also inhibited Ca2+ waves, global [Ca2+]i and myogenic tone without altering the occurrence or properties of Ca2+ sparks (Fig. 7I–L).
Consistent with results in feed arteries, xestospongin D, 2-APB and U73122 (Fig. 8) each decreased the occurrence of Ca2+ waves in SMCs of cremaster arterioles and reduced the amplitude and frequency of those waves that remained (Fig. 8A, D and G). These three antagonists also decreased global Ca2+ levels (Fig. 8B, E and H) and dilated arterioles (Fig. 8C, F and I). In contrast, U73343 (10 μm), the inactive analogue of U73122, had no significant effect: Ca2+ wave occurrence (56 ± 10% of 93 SMCs in PSS vs. 60 ± 10% of 88 SMCs in U73343), amplitude (1.8 ± 0.1 F/Fo in PSS vs. 1.8 ± 0.1 in U73343) and frequency (0.79 ± 0.10 Hz in PSS vs. 0.99 ± 0.20 Hz in U73343) were not significantly different from values in PSS (n= 6, P > 0.05 for each comparison). Global [Ca2+]i (Fluo-4 intensity in U73343 = 1 ± 0.1 relative to PSS, n= 6) and vessel diameter (31 ± 2 μm in PSS vs. 29 ± 2 μm in U73343, n= 5) were also unchanged by U73343 (P > 0.05 for both comparisons). These data collectively suggest that IP3Rs and phospholipase C contribute to Ca2+ waves and myogenic tone in both feed arteries and arterioles supplying the mouse cremaster muscle.
Effect of Ca2+ store depletion on Ca2+ signals and myogenic tone in feed arteries and arterioles
Treatment with U73122 can deplete intracellular Ca2+ stores, possibly through inhibition of the smooth endoplasmic reticulum Ca2+ ATPase (SERCA) (Macmillan & McCarron, 2010). Therefore, we examined the effects of a known SERCA inhibitor (thapsigargin) on Ca2+ signals and myogenic tone in feed arteries and arterioles. Thapsigargin (100 nm) decreased the occurrence, amplitude and frequency of Ca2+ waves and abolished Ca2+ sparks in SMCs of feed arteries (Fig. 9A and B). In addition, thapsigargin caused a significant increase in the global Ca2+ signal and constricted feed arteries (Fig. 9C and D). In SMCs of arterioles, thapsigargin (100 nm) had similar effects: inhibition of Ca2+ waves, increased global Ca2+ and vasoconstriction (Fig. 9E–G). The reciprocal effects on global [Ca2+]i and vessel diameters between U73122 (Figs 7 and 8, bottom panels) and thapsigargin suggest that the effects of U73122 on Ca2+ signals and myogenic tone (Figs 7 and 8) are unlikely to be caused simply by depletion of intracellular Ca2+ stores.
Expression and localisation of RyRs in feed arteries and arterioles
As an initial test of the hypothesis that differences in the expression and localisation of RyR isoforms contribute to the heterogeneity of function of RyRs observed between feed arteries (Figs 2 and 4) and arterioles (Figs 3 and 5), we examined the expression of RyR isoform transcripts in SMCs isolated from each vessel type. We found no detectable expression of RyR1 (n= 8 samples each from a different mouse) in pre-amplified samples even after 40 rounds of PCR (Table 2). In contrast, transcripts for both RyR2 and RyR3 were observed in feed artery and arteriolar SMCs (Table 2). In feed artery SMCs, RyR2 was detected in 8 of 8 samples, whereas 7 of 8 samples were positive for RyR2 in cremaster SMC (Table 2). Transcripts for RyR3 were detected in 7 of 8 samples from both feed arteries and arterioles. RyR2 was the most abundant transcript in both vessels, followed by RyR3 (Fig. 10). However, the level of expression of these isoforms displayed significant regional heterogeneity. Feed artery SMCs expressed 2.1 ± 0.3-fold more RyR2 than did SMCs from arterioles (n= 7, P < 0.05) (Fig. 10). Conversely, feed artery SMCs expressed only 0.7 ± 0.1-fold as much RyR3 as did arteriolar SMCs (n= 6, P < 0.05). The ratio of RyR2/RyR3 expression was 3.7 ± 0.4-fold greater in feed artery SMCs than in arteriolar SMCs (n= 6, P < 0.05). Importantly, the lack of RyR1 expression in feed arteries and arterioles was not due to an incorrect primer or technical limitation because positive control RT-PCR experiments using RNA from homogenates of diaphragm muscle readily detected RyR1 transcripts (Tables 1 and 2). Expression of RyR1 also was observed in samples of brain (CT= 22.2 ± 0.5, n= 3) and heart (CT= 28.0 ± 0.3, n= 3), although at much lower levels than diaphragm muscle.
The presence and localisation of RyR protein in intact, pressure-fixed-feed arteries and arterioles was investigated using an antibody that recognizes both RyR1 and 2 (RyR1/2). In confocal image slices from intact fixed feed arteries, RyR1/2 staining appeared as bright clusters within the SMCs (Fig. 11). In contrast, RyR1/2 staining in the SMCs of intact fixed arterioles was less intense and had a diffuse staining pattern (Fig. 11). No fluorescence was observed when primary antibodies were excluded from the staining process in either preparation (n= 3, data not shown). Along the reference line for evaluating fluorescence intensity, feed arteries had a significantly higher coefficient of variation vs. arterioles (Fig. 11, right panel) confirming that the pattern of RyR antibody immunostaining differs between SMCs of respective vessels. Given the lack of expression of RyR1 mRNA in these SMCs (Fig. 10), these data suggest that there is a significant difference in the expression and localisation of RyR2 between cremaster feed arteries and downstream, second-order arterioles in the mouse. The expression of RyR3 was not examined due to lack of a suitable antibody for this RyR isoform.
Expression and localisation of IP3Rs in feed arteries and arterioles
As IP3Rs govern Ca2+ signals and myogenic tone in both feed arteries and second-order arterioles (Figs 7 and 8), we examined the expression of the IP3R isoforms in SMCs isolated from respective vessels. Quantitative RT-PCR showed a similar pattern of expression of the IP3R isoforms in both vessels: expression of IP3R1 was highest, with significantly less IP3R2 and IP3R3 (Table 2 and Fig. 12). Immunofluorescent staining of intact vessels revealed bright, clustered staining for IP3R1 similar to that seen for RyR1/2 in the feed arteries (Fig. 13A, top panel). A low level of staining was detected for IP3R2 (Fig. 13A, middle panel) and a small but detectable amount of staining was observed for IP3R3 (Fig. 13, bottom). No staining was observed when primary antibodies were omitted (n= 3, data not shown).
Analysis of the staining pattern for IP3R1 showed no significant difference in the coefficient of variation for fluorescence intensity along feed arteries (0.22 ± 0.06) compared with arterioles (0.22 ± 0.04; P > 0.05, n= 3 each). When the overall fluorescence intensity of staining for each IP3R isoform was expressed relative to IP3R1 there was ∼70% less staining for IP3R2 and ∼90% less staining for IP3R3 in both vessel types, with no differences in the distribution between arteries and arterioles (Fig. 13B, left panel). As there was little staining for either IP3R2 or IP3R3 in SMCs, we performed a series of positive control experiments using freshly isolated mouse primary hepatocytes, according to the recommendations of the antibody manufacturer. In these positive controls, hepatocytes stained for all three IP3R isoforms (Fig. 13A, right panels). In contrast to the profile observed in SMCs, fluorescence intensity for IP3R2 and IP3R3 staining of hepatocytes was twice that of IP3R1 (Fig. 13B, right panel). These data show that the IP3R2 and IP3R3 antibodies are effective in cells containing the appropriate IP3R isoforms and demonstrate that the IP3R1 antibody does not routinely produce a higher signal than the antibodies for IP3R2 and IP3R3. Together, the RT-PCR and immunofluorescence data suggest that IP3R1 is the predominant IP3R isoform expressed in both cremaster feed arteries and second-order arterioles in the mouse.
Blood flow control in vascular resistance networks involves dilatation and constriction of proximal feed arteries and downstream arterioles. Feed arteries located external to the tissue govern the volume of blood entering a microvascular network while arterioles within the tissue govern flow distribution according to local metabolic demand. Though changes in vasomotor tone are integral to flow control throughout resistance networks, regional heterogeneity in the mechanisms underlying the regulation of SMC contraction and relaxation is poorly understood. The present findings illustrate striking differences in the function of RyRs between feed arteries and second-order arterioles of the mouse cremaster muscle. Our studies illustrate that RyRs participate in Ca2+ signals and the negative feedback regulation of myogenic tone in feed arteries. In contrast, RyRs remain functionally silent in second-order arterioles without contributing either to Ca2+ signals or to myogenic tone. The consistency of the present observations with recent findings in the hamster (Westcott & Jackson, 2011) suggest that the differences in function of RyRs that we have identified between feed arteries and downstream arterioles are not unique to a single species but may have more general application. Further, we show for the first time that the expression and localisation of RyR isoforms differs significantly between SMCs of feed arteries vs. arterioles. In contrast, the expression of IP3R isoforms was similar between SMCs in respective resistance vessels, as was the dependence of Ca2+ waves on these Ca2+ channels.
Smooth muscle Ca2+ signals in feed arteries vs. arterioles
In SMCs of murine cremaster feed arteries, RyRs appear to contribute to both Ca2+ sparks and Ca2+ waves (Fig. 2). This behaviour is consistent with our recent findings in SMCs of hamster feed arteries (Westcott & Jackson, 2011) and earlier studies of rat retinal arterioles (Tumelty et al. 2007, 2011). There are two scenarios by which RyRs may participate in the genesis of Ca2+ waves in SMCs of these resistance vessels. First, Ca2+ released by RyRs may trigger release of Ca2+ through IP3R (i.e. CICR), thereby amplifying the RyR Ca2+ signal and producing the waves observed. Alternatively, RyRs may amplify Ca2+ released through IP3R (also via CICR) to produce Ca2+ waves. Both of these possibilities are consistent with our present findings that either RyR blockers or inhibitors of IP3R/PLC abrogated Ca2+ waves (Figs 2 and 7). This behaviour contrasts with our findings in SMCs of arterioles, where Ca2+ waves depended solely upon IP3R (Figs 3 and 8).
The inhibition of Ca2+ waves by ryanodine and tetracaine observed in feed arteries could result from depletion of Ca2+ stores or from the interactions of RyR blockers with IP3R (McCarron et al. 2003; MacMillan et al. 2005). However, we think that these explanations are unlikely for the following reasons. First, neither ryanodine (Fig. 3) nor tetracaine (Results) altered Ca2+ waves in arteriolar SMCs, thereby excluding non-selective effects of the RyR blockers on IP3R. Given the similarity in IP3R expression observed here (Figs 12 and 13), it is unlikely that these agents would affect IP3R in SMCs of one vessel, but not of the other vessel. Second, the presence of functional (but otherwise silent) RyRs in SMCs of arterioles (Fig. 6B), coupled with the lack of effect of ryanodine and tetracaine on otherwise robust Ca2+ signals in these SMCs precludes effects attributable to depletion of Ca2+ stores. Finally, tetracaine appears to block RyRs without producing store depletion (Curtis et al. 2008). Thus, our data support the hypothesis that RyRs directly participate in Ca2+ waves in SMCs of feed arteries controlling blood flow to the mouse cremaster muscle.
Blockade of BKCa with either paxilline (Fig. 4A) or TEA (Results) prevented ryanodine-induced constriction of feed arteries. These data are consistent with previous findings that RyRs contribute to the negative feedback regulation of myogenic tone in resistance arteries and that this function involves RyR-dependent control of BKCa (Nelson et al. 1995; Jaggar et al. 1998; Gollasch et al. 2000). The lack of effect of ryanodine on myogenic tone in the presence of the BKCa blockers was not due to functional antagonism resulting from constriction produced by the BKCa blocker, because induction of similar constriction with Bay K 8644 did not inhibit further constriction in response to ryanodine (Results). Thus, our data support the concept that BKCas mediate the vasomotor effects of constitutively active RyRs in feed arteries supplying the mouse cremaster muscle. However, we do not think that RyRs are the sole controllers of BKCa in these vessels because effective inhibition of RyRs did not abolish the vasomotor effects of BKCa blockade with paxilline (Fig. 4B). In turn, these data suggest that an additional source of activator Ca2+ is involved in the control of BKCa in SMCs of cremaster feed arteries, e.g. voltage-gated Ca2+ channels.
In SMCs of second-order cremaster arterioles, RyRs were functionally silent, as they did not contribute either to myogenic tone or to Ca2+ signals (Figs 3 and 5). These data indicate that the Ca2+ responsible for control of BKCa activity in these SMCs must arise from a source other than RyR-related Ca2+ signals, which is consistent with the behaviour of SMCs in arterioles of the hamster cremaster muscle (Westcott & Jackson, 2011). Based on prior studies in neurons (Sun et al. 2003; Grunnet & Kaufmann, 2004; Berkefeld et al. 2006) and coronary artery SMCs (Guia et al. 1999), we speculated that Ca2+ influx through voltage-gated Ca2+ channels may control BKCa in SMCs of cremaster arterioles (Westcott & Jackson, 2011). In the present study, we tested this hypothesis by examining the effects of TEA (a BKCa blocker) on the diameter of arterioles at low pressure in the absence and presence of the voltage-gated Ca2+ channel agonist Bay K 8644 (Schramm et al. 1983). In arterioles pressurized to 20 cmH2O, myogenic tone was absent, with (presumably) low activity of voltage-gated Ca2+ channels (Brayden & Nelson, 1992), and TEA alone had no effect on diameter (Fig. 6B). However, in the presence of Bay K 8644, the addition of TEA produced robust vasoconstriction that was unaffected by the blockade of RyRs with ryanodine. These data support the hypothesis that BKCas in arteriolar SMCs are controlled by Ca2+ influx through voltage-gated Ca2+ channels independent of RyR activity. A similar relationship appears to exist in SMCs of feed arteries based upon the similarity of their behaviour to these interventions (Fig. 6A), which may, in turn, account for the ryanodine-insensitive portion of responses to BKCa blockade observed in feed arteries (Fig. 4B). Additional research will be required to critically test the hypothesis that Ca2+ influx through voltage-gated Ca2+ channels activates BKCa in SMCs of arterioles and feed arteries and to determine whether these channels are co-localized with BKCa.
Ryanodine receptor and IP3 receptor expression
Ryanodine receptors We report, for the first time, the expression of RyR isoforms in cremaster feed arteries and downstream arterioles and illustrate regional differences that may contribute to the functional heterogeneity that we describe for respective vessels. The SMCs from both vessels expressed only transcripts for RyR2 and RyR3, with RyR2 more highly expressed in feed artery vs. arteriolar SMCs. In contrast, RyR3 was more highly expressed in arteriolar vs. artery SMCs (Fig. 10). When presented as a ratio, RyR2/RyR3 was nearly 4-fold higher in the SMCs from arteries compared with arterioles. These differences in RyR isoform expression may, in turn, account for the functional differences that we observed between respective vessels in Ca2+ signalling and the regulation of vasomotor tone. First, RyR2 has been shown to contribute to Ca2+ sparks in other SMCs (Coussin et al. 2000; Ji et al. 2004). Second, the RyR3 isoform (or one of its splice variants) has been proposed to inhibit Ca2+ spark formation (Löhn et al. 2001; Jiang et al. 2003; Dabertrand et al. 2006). Thus, the higher ratio of RyR2/RyR3 that we observed in SMCs of arteries vs. arterioles is consistent with our observation of RyR-dependent Ca2+ sparks in arteries and their absence from arterioles.
The predominant expression of RyR2 that we observed in SMCs of both resistance vessels studied here is consistent with recent findings in rat cerebral arteries (Vaithianathan et al. 2010) as well as earlier studies in rat aorta and pulmonary arteries (Yang et al. 2005). However, they contrast with studies of rat basilar and femoral arteries in which SMCs exhibited predominant expression of RyR3 (Salomone et al. 2009), with no expression of RyR1 or RyR2. Studies of rat aortic SMCs also suggested that RyR3 was the most highly expressed isoform (Vallot et al. 2000). These data collectively support the concept of regional differences in the pattern of expression of RyR isoforms in vascular SMCs.
In contrast to studies of vascular SMCs from larger vessels (Ledbetter et al. 1994; Neylon et al. 1995; Coussin et al. 2000; Löhn et al. 2001; Yang et al. 2005; Zheng et al. 2005, 2008; Vaithianathan et al. 2010), we found that neither feed artery nor arteriolar SMCs expressed RyR1 (Fig. 10). We do not think that this is due to a technical limitation in our methods, because we used validated primers, and verified that these primers appropriately identified RyR1 expression in mouse skeletal muscle, brain and heart (see Results and Tables 1 and 2). We conclude that if RyR1 is expressed in SMCs from these vessels, then its expression must be at a level that is below our limits of detection, and far less than levels for RyR2 and RyR3. It is worthy to note that most prior studies have used homogenates of intact vessels. Thus, message from cells other than vascular SMCs may have been amplified and thereby have contributed to message levels reported.
Differences in the localisation of RyR proteins expressed within SMCs are likely to contribute to the functional heterogeneity observed between SMCs of feed arteries compared with arterioles. Using an antibody that reportedly detects RyR1 and RyR2 (Sigma), we found significant differences in the pattern of immunostaining between SMCs of respective vessels (Fig. 11). Given the lack of RyR1 transcripts, the fluorescence observed using this antibody probably represents the RyR2 protein in SMCs of both vessels. In the feed arteries, RyR1/2 staining appeared as clusters within SMCs, whereas staining in the arteriolar SMCs was more uniform. In SMCs of rabbit portal vein, Ca2+ spark-generating sites also exhibited clustering (Gordienko et al. 2001), suggesting that such grouping of RyRs may be required to form nexi for Ca2+ spark generation. Thus, differential localisation of RyR proteins may also contribute to the regional heterogeneity in RyR function observed here. By inference, a similar explanation may account for such regional heterogeneity observed in corresponding vessels of the hamster (Westcott & Jackson, 2011).
Inositol trisphosphate receptors and PLC We report, for the first time, the expression and localisation of IP3R isoforms in cremaster muscle feed arteries and arterioles of the mouse. We found predominant expression of IP3R1, with much lower levels of IP3R2 and IP3R3 (Fig. 12). This pattern was reproduced when IP3R isoform protein expression was examined by immunofluorescence (Fig. 13). Furthermore, the pattern of expression of these isoforms (IP3R1 > > IP3R2 > IP3R3) appeared similar between feed artery and arteriolar SMCs. These data suggest that the differences that we observed in IP3R function between feed arteries and arterioles are unlikely due to gross differences in IP3R isoform expression. Our data support previous findings suggesting that IP3R1 is the major IP3R isoform expressed in vascular smooth muscle (Nixon et al. 1994; Tasker et al. 1999; Grayson et al. 2004) and are consistent with the hypothesis that IP3R1 contributes to Ca2+ waves and myogenic tone in mouse cremaster feed arteries and arterioles. A key difference between vessels is that Ca2+ waves in SMCs of feed arteries were inhibited by antagonists of either RyRs or IP3Rs, which implies that the PLC–IP3R pathway could not sustain Ca2+ waves in feed artery SMCs in the absence of RyRs. In contrast, Ca2+ waves in SMCs of arterioles were independent of RyRs. In turn, these data suggest that there are differences in function of IP3Rs between the feed arteries and the arterioles that warrant further investigation.
Phospholipase C has been proposed to participate in the mechanisms responsible for myogenic tone in arteries (Osol et al. 1993; Coats et al. 2001; Jarajapu & Knot, 2002) and arterioles (Bakker et al. 1999; Westcott & Jackson, 2011). The present data are consistent with this hypothesis. Three structurally different inhibitors of the PLC–IP3R signalling pathway (xestospongin D, 2-APB and U73122) inhibited SMC Ca2+ waves and myogenic tone in both feed arteries and arterioles (Figs 7 and 8). The similar effects produced by these inhibitors suggest that non-specific actions were minimal. We therefore conclude that PLC and IP3R play a major functional role in governing myogenic tone in resistance vessels of the mouse. Although U73122 may cause depletion of intracellular Ca2+ stores by inhibition of SERCA (Macmillan & McCarron, 2010), we do not think that such actions contribute to the inhibition of Ca2+ waves and myogenic tone for two reasons. First, in feed arteries, U73122 selectively inhibited Ca2+ waves without affecting Ca2+ sparks (Fig. 7). In contrast, the SERCA inhibitor, thapsigargin, nearly abolished both Ca2+ waves and Ca2+ sparks (Fig. 9). Second, both feed arteries and arterioles dilated (during significantly lowered global Ca2+) when exposed to U73122, whereas inhibition of SERCA with thapsigargin produced vasoconstriction in conjunction with a significant rise in global Ca2+ (Figs 7–9). Similar reasoning applies to interpreting the actions of 2-APB and xestospongin D (Figs 7 and 8). Although 2-APB may have off-target effects, such as blockade of non-selective cation channels (Bootman et al. 2002), the similarity in response to U73122 and xestospongin D suggests that such off-target effects do not account for the actions of 2-APB in our experiments. Also, recent studies suggest that IP3Rs may be involved in the activation of plasma membrane cation channels in vascular smooth muscle (Xi et al. 2008), casting some doubts on the non-selectivity of 2-APB. Thus, our pharmacological interventions support the conclusion that the PLC–IP3R pathway contributes to Ca2+ signalling and myogenic tone in mouse cremaster feed arteries as well as second-order arterioles.
What remains to be established are the identification of the isoform of phospholipase C that is involved, and the mechanisms responsible for pressure-dependent activation of PLC. Studies in rat cerebral arteries and the renal circulation have suggested that stretch of the plasma membrane activates highly expressed G-protein-coupled receptors, such as AT1 receptors for angiotensin II, to activate PLC-β and contribute to myogenic tone (Mederos y Schnitzler et al. 2008). It also has been proposed that depolarization-induced activation of L-type Ca2+ channels can activate G-proteins and downstream PLC-β signalling in SMCs from arteries (del Valle-Rodriguez et al. 2003; Urena et al. 2007). Membrane depolarization due to inhibition of K+ channels and activation of non-selective cation channels and subsequent activation of L-type Ca2+ channels is central to myogenic tone (Davis & Hill, 1999; Hill et al. 2001; Welsh et al. 2002; Earley et al. 2004). Thus, pressure-induced membrane depolarization might also contribute to activation of PLC-β in feed arteries and arterioles. Recent studies have also suggested that PLC-γ (Garcia & Earley, 2011) or PLC-δ1 (Clarke et al. 2008) may participate in the regulation of myogenic tone in vascular smooth muscle. Additional research will be required to define the isoform(s) of PLC that govern myogenic tone in cremaster feed arteries and arterioles.
Summary and conclusions
Feed arteries and arterioles serve complementary roles in controlling tissue blood flow. Whereas both vessels are integral to resistance networks, feed arteries govern flow magnitude into the microcirculation while arterioles govern flow distribution within the tissue. Consistent with the hypothesis that vasomotor tone in respective sites may be differentially regulated, the present study is the first to illustrate regional heterogeneity in the function and expression of RyRs between SMCs in cremaster feed arteries compared with downstream, second-order arterioles of the mouse. In feed arteries, RyRs underlie Ca2+ sparks and Ca2+ waves and contribute to the negative feedback regulation of myogenic tone. In distinct contrast, RyRs are functionally silent in arteriolar SMCs, without contributing either to Ca2+ signals or to myogenic tone under physiological conditions. Nevertheless, RyR function can be demonstrated in arteriolar SMCs by activating them with caffeine at low transmural pressure. Based on these findings, we propose that differences in the level of expression of RyR2 and RyR3 transcripts together with differences in the localisation of RyR protein contribute to the observed functional heterogeneity between respective vessels, including the regulation of BKCa function. Whereas BKCa function appears to be coupled to the activity of RyRs and voltage-gated Ca2+ channels in SMCs of feed arteries, BKCa in SMCs of arterioles function independent from RyRs. Instead, Ca2+ influx through voltage-gated Ca2+ channels controls BKCa activity in these microvessels.
In striking contrast to the heterogeneity observed in the function and expression of RyRs between vessels, IP3Rs exhibit similar patterns of expression with similar functional roles in SMCs of feed arteries and downstream arterioles: IP3Rs underlie Ca2+ waves and contribute to myogenic tone in both vessels. However, unlike arterioles, IP3Rs in feed arteries were incapable of sustaining Ca2+ waves in the absence of RyR function. This previously unrecognized heterogeneity in the function of IP3Rs and/or upstream PLC between SMCs of cremaster feed arteries and second-order arterioles requires further investigation.
Resolving regional heterogeneity in the mechanisms contributing to SMC Ca2+ signalling and myogenic tone provides critical new insight for explaining how blood flow distribution can be regulated within the microcirculation independent of changes in total flow entering the tissue through feed arteries. Conversely, the magnitude of blood flow can be regulated (along with systemic perfusion pressure) by feed arteries upstream from the microcirculation independent from redistributing flow within the tissue. This new perspective provides both a foundation and a rationale for developing therapeutic interventions to selectively modulate and improve blood flow distribution and tissue oxygenation without untoward side-effects, such as orthostatic hypotension, that accompanies drugs that indiscriminately affect the function of SMCs in both resistance arteries and arterioles.
Study conception and design: E.B.W., W.F.J. and S.S.S; Sample collection and data analysis: E.B.W., E.L.G. and W.F.J.; Manuscript editing: E.B.W., W.F.J. and S.S.S. All authors approved the final version for publication. This work was completed at Michigan State University, East Lansing, MI, USA.
Author's present address
E. B.Westcott: Department of Medical Pharmacology and Physiology, MA415 Medical Sciences-1 Hospital Drive, University of Missouri, Columbia, MO 65212, USA.
The authors thank Aaron Fullerton and Dr Kazuhisa Miyakawa for providing the primary hepatocytes used in these studies. This work was supported by National Institutes of Health grants RO1 HL086483, RO1 HL 32469 and PO1 HL070687 and by AHA Fellowship 0815778G (to E.B.W.).