Propagated Endothelial Ca2+ Waves and Arteriolar Dilation In Vivo
Measurements in Cx40BAC–GCaMP2 Transgenic Mice
To study endothelial cell (EC)– specific Ca2+ signaling in vivo we engineered transgenic mice in which the Ca2+ sensor GCaMP2 is placed under control of endogenous connexin40 (Cx40) transcription regulatory elements within a bacterial artificial chromosome (BAC), resulting in high sensor expression in arterial ECs, atrial myocytes, and cardiac Purkinje fibers. High signal/noise Ca2+ signals were obtained in Cx40BAC-GCaMP2 mice within the ventricular Purkinje cell network in vitro and in ECs of cremaster muscle arterioles in vivo. Microiontophoresis of acetylcholine (ACh) onto arterioles triggered a transient increase in EC Ca2+ fluorescence that propagated along the arteriole with an initial velocity of ≈116 μm/s (n=28) and decayed over distances up to 974 μm. The local rise in EC Ca2+ was followed (delay, 830±60 ms; n=8) by vasodilation that conducted rapidly (mm/s), bidirectionally, and into branches for distances exceeding 1 mm. At intermediate distances (300 to 600 μm), rapidly-conducted vasodilation occurred without changing EC Ca2+, and additional dilation occurred after arrival of a Ca2+ wave. In contrast, focal delivery of sodium nitroprusside evoked similar local dilations without Ca2+ signaling or conduction. We conclude that in vivo responses to ACh in arterioles consists of 2 phases: (1) a rapidly-conducted vasodilation initiated by a local rise in EC Ca2+ but independent of EC Ca2+ signaling at remote sites; and (2) a slower complementary dilation associated with a Ca2+ wave that propagates along the endothelium.
A remarkable range of endothelial cell (EC) Ca2+ signaling is implicated in regulating the resistance microvasculature.1–5 Conducted electrical signals travel for millimeters along the vessel wall,6 mediate coordinated vasomotor responses to localized stimuli,7–9 and involve distinct Ca2+ signals within smooth muscle (SM) and ECs. For example, arteriolar dilation in response to the endothelium-dependent vasodilator acetylcholine (ACh) involves an increase in ECs2 and decrease in SM10 Ca2+. Although recent intravital studies have demonstrated a crucial role for the endothelium in conducted vasodilation,9 the extent to which Ca2+ signals are transmitted along the arteriolar wall and the mechanisms underlying Ca2+ transmission are controversial. Some studies indicate that EC Ca2+ responses to dilatory stimuli are localized,1,2 whereas others indicate that Ca2+ signals can travel along the endothelium for a millimeter or more.5,11
The majority of information concerning microvascular Ca2+ signaling has been derived from isolated vessels studied in vitro,4,5,12 largely because of the difficulty of selectively loading ECs or SM cells with Ca2+ sensitive dyes in vivo.1,10 A fundamental limitation to isolated arterioles is their disconnection from networks in which they normally reside. The extent to which manipulation of arterioles to obtain dye loading and the loss of physiological parameters (eg, pressure or flow) alter Ca2+ signaling is unknown. Therefore, alternative approaches are necessary to investigate arteriolar Ca2+ signaling during blood flow control in the intact system.
We have recently reported the development of a genetically encoded Ca2+ indicator with optical properties to enable in vivo measurements.13 Previous genetic targeting of the EC lineage has relied on partial promoter constructs with variable expression in adult arterial tissues.14 To address this limitation, we used a bacterial artificial chromosome (BAC) transgenesis approach using the connexin40 (Cx40) gene locus, as Cx40 is expressed in all arterial ECs, as well as in cardiac Purkinje fibers.15 Here we report the development of transgenic mice in which GCaMP2 replaces the initial coding sequence of Cx40 within a locus-spanning BAC. We have used these mice, which have endothelium lineage-specific Ca2+ signaling, to test the hypothesis11 that the conduction of vasodilation along arteriolar networks in vivo involves a wave of Ca2+ traveling along the endothelium.
Materials and Methods
Generation of Tg(RP24–255O4-GCaMP2)1Mik Mice
A BAC clone containing approximately 85 kb and 61 kb of 5′ and 3′ DNA flanking the Cx40 locus was modified by insertion of a GCaMP2-IRES-GCaMP2-pA cassette to replace 6 nucleotides at the initiation codon of Cx40 in exon2 by homologous recombination and used as a random insertion transgene.16 BAC targeting and founder genotyping are described in the supplemental Methods (available online at http://circres.ahajournals.org).
Adult Cx40BAC-GCaMP2 (9 to 16 weeks) mice (n=19; Cornell Core Transgenic Mouse Facility, Cornell University, Ithaca, NY) were anesthetized, the heart removed, or the left cremaster muscle prepared as described.9 Up to 3 arterioles were studied per mouse, with each Ca2+ wave counted as an event; n refers to the number of events studied. For imaging and data details see supplemental Methods.
Efficient EC Targeting in Cx40BAC-GCaMP2 Mice
To circumvent the problem of limited adult EC expression levels and incomplete lineage specificity associated with minimal promoter fragments such as Tie214 or VE-cadherin,17 we chose to replicate the endogenous expression pattern of Cx40 by BAC transgenesis. A bicistronic construct (GCaMP2-IRES-GCaMP2) was targeted to the start codon in a Cx40 -spanning BAC18 (Figure 1A) and the recombineered BAC injected into fertilized oocytes, placing GCaMP2 under Cx40 transcriptional control without overexpression of Cx40 (open reading frame in BAC disrupted) or disruption of an endogenous Cx40 allele (result of knock-in). Random insertion founders were identified by detection of BAC backbone and Cx40-IRES bands (Figure 1B).
Cx40 gap junction proteins are highly expressed in arterial/arteriolar ECs, cardiac Purkinje fibers, and atrial myocytes.9,15 Transgene expression was first tested by coimmunostaining for GCaMP2 and anti-Cx40; Cx40 staining in arterial/arteriolar ECs appeared as punctate labeling at the periphery of GCaMP2-positive cells, and staining was completely coincident in Cx40 positive cells (Figure 1C). Costaining of GCaMP2 and smooth muscle α-actin (Figure 1D) indicated intense GCaMP2 immunoreactivity in arteriolar and capillary endothelium of the lung (distinguished by the presence of SM cells), whereas venous endothelium lacked GCaMP2 expression. Fluorescence was not observed in cremaster veins/venules, whereas ECs of arteries/arterioles were intensely fluorescent. The Cx40BAC construct specified GCaMP2 expression in arteries, arterioles, and capillaries of diverse tissues (Figure 1E).
Cardiac GCaMP2 Expression and Function in Cx40BAC-GCaMP2 Mice
GCaMP2 was expressed in the adult heart, where it was observed in atrial myocytes, the Purkinje cell network, and ECs lining cardiac arterioles (Figure 2A); no expression was detected in ventricular myocardium, consistent with the pattern of Cx40 expression.15 We confirmed the function of GCaMP2 in the Purkinje layer in cut-open, superfused and Langendorf -perfused Cx40BAC-GCaMP2 hearts. Stimulation of the Purkinje network with a bipolar electrode resulted in fluorescent Ca2+ transients at the stimulus frequency over the entire network (Figure 2B), in the absence of equivalent signals in the ventricular myocardium. Confocal measurements in spontaneously beating hearts revealed robust fluorescent transients confined to the Purkinje network (Figure 2C).
EC Ca2+ Signaling and Endothelium-Dependent Vasodilation In Vivo
To investigate the role of Ca2+ signaling during endothelium-dependent vasodilation in vivo, we visualized 35 arterioles in the superfused cremaster muscle of 14 adult Cx40BAC-GCaMP2 mice and delivered a focal ACh stimulus using microiontophoresis (typically 1 μA; 1000 ms; 1 μm micropipette tip). Arterioles displayed a resting diameter of 30±2 μm before stimulation, dilated by 40±4% (n=33) at the site of stimulation, and recovered to resting diameter after stimulation. Acetylcholine triggered a rapid rise in fluorescence at the local site, and the Ca2+ signal propagated bidirectionally along the endothelium (Figure 3). The local rise in EC Ca2+ occurred with a half time of 250±31 ms (n=10), and the mean peak fluorescence increase was 0.39±0.04 (ΔF/F0;n=10). The increase in Ca2+ fluorescence propagated along the endothelium at a velocity of 116±6 μm/s over the first 200 μm (n=28). As the Ca2+ wave propagated away from the stimulus site, the rate of rise and peak ΔF/F0 decreased progressively, decaying to oscillations at the most remote extent of wave propagation (Figure 3B, 3D, and 3F; supplemental Movie I). Although the nonratiometric nature of GCaMP2 obviates quantitative measurements of Ca2+, these oscillations could be clearly observed above background levels (Figure 3B). Ca2+ waves also propagated into and along branches arising from the stimulated arteriole (Figure 3A, 3C, 3D, 3E; supplemental Movies I and II). Varying the ACh stimulus (100 to 1000 ms pulse) revealed a dose-dependent increase in the distance Ca2+ waves propagated with no change in the initial velocity (data not shown). In 14 experiments performed under identical conditions, the Ca2+ wave decayed by 50% in 204±17 μm. However, low level oscillations were often detected for distances up to approximately 1 mm (supplemental Movie I). These EC oscillations at remote sites have not been reported previously and may underlie changes in cell function that extend well beyond areas of EC direct activation.
After the initial local rise in EC Ca2+, vasodilation spread rapidly along the arteriolar network. As shown in Figure 3C and 3F, this rapidly-conducted vasodilation traveled much faster (>2 mm/s) than the Ca2+ wave and occurred at remote sites without a corresponding rise in Ca2+. The time to initiate dilation 400 μm upstream from the stimulus was 878±36 ms (n=9), which was similar to the delay recorded at the site of stimulation (831±63 ms; n=8). Remarkably, at intermediate sites the rapidly-conducted vasodilation was followed by a later phase of dilation that coincided with the arrival of the Ca2+ wave (Figure 3A, compare diameters between 1.25 and 4.5 s), suggesting 2 distinct mechanisms of endothelium-dependent vasodilation (see below). ACh stimuli of shorter duration (eg, 50 or 100 ms) evoked transient, submaximal increases in local EC Ca2+ and diameter; rapidly conducted vasodilation occurred in all experiments in which brief delivery of ACh resulted in a submaximal increase in EC Ca2+ (n=17), suggesting an equivalent threshold.
In contrast to the actions of ACh, focal delivery of the nitric oxide donor SNP, which relaxes SM independent of the endothelium through guanylyl cyclase–dependent mechanisms,19 produced local vasodilation without increasing EC Ca2+ (Figure 4; supplemental Movie III) and no vasomotor response at intermediate or remote sites, further suggesting that the local rise in EC Ca2+ in response to ACh initiates rapidly-conducted vasodilation. Arteriolar fluorescence appeared to decrease during dilation to SNP (Figure 4A), but this was explained by redistribution of GCaMP2 over a larger area within the optical section (Figure 4B). Total fluorescence integrated across the entire vessel (ΣF) varied by <2.5% between rest and SNP dilation (P>0.05, n=5), and a similar redistribution effect was observed for remote dilations to ACh extending beyond the Ca2+ wave (Figure 3F, site 2). Thus simple distension of the arteriolar wall does not increase EC Ca2+.
At the site of ACh delivery the initial rise in Ca2+ preceded the onset of dilation (Figure 5A and 5B; supplemental Movie IV). We further resolved Ca2+ signals in individual ECs by confocal imaging (n=13 cells) during 100-ms ACh stimuli (Figure 5C and 5D); EC Ca2+ rose rapidly, preceding the local dilation, then oscillated asynchronously at a mean frequency of 1.2±0.1 Hz (Figure 5B and 5D; supplemental Movie V). These observations collectively indicate that endothelium-dependent vasodilation is preceded by a rise in EC Ca2+, which triggers both the rapid conduction of vasodilation and an ensuing wave of Ca2+ along the endothelium (see below).
Separation of Ca2+-Dependent Vasodilation and the Vasomotor Response
As discussed above, ACh evoked a rapid dilation at remote sites that preceded arrival of a Ca2+ wave (Figure 3C and 3F), and a slower dilation followed the arrival of the Ca2+ wave. In 8 experiments examined 200 μm upstream from the stimulus site, rapid vasodilation preceded the Ca2+ wave by 681±45 ms. The secondary dilation can be seen most clearly at sites 300 to 400 μm from the local stimulus when the Ca2+ wave arrives (Figure 3A, bottom images). This secondary response was manifested as a maintained dilation during the period of increased endothelial Ca2+ while upstream segments recovered (Figure 5E and 5F; supplemental Movie VI). As the rapid dilation has been linked to the transmission of a hyperpolarizing signal,3,6 we term this process “electrically-conducted vasodilation” and the latter process “Ca2+ wave–dependent vasodilation”.
We examined the degree to which arteriolar dilation was spatially and temporally associated with the arrival of the Ca2+ wave in 9 experiments, by determining the dependence of peak dilation on distance from the stimulus (Figure 5G). Consistent with a Ca2+-dependent component of dilation, the peak diameter change was greater (P<0.05) at 300 μm (56±8%) than at 1000 μm (33±7%). The respective changes (ΔΣF/ΣFo) in Ca2+ fluorescence were +40±13% and −21±15% (P<0.01). The pattern of dilation along the arteriole was also consistent with a Ca2+-dependent component, as peak dilation did not decay monotonically, but dropped off after 300 μm, a point consistent with the decay of the Ca2+ wave (50% decay at 204 μm). Peak dilation at the intermediate (300 μm) site followed the increase in Ca2+, peak Ca2+ occurring at 3.2±0.5 s and peak (second component) dilation occurring at 10.1±1.1 s after ACh stimulation. This was also reflected as sustained dilation at the intermediate arteriolar segments; between 5 and 15 s after stimulation arterioles further dilated 12.4% at 300 μm, whereas at 1000 μm arterioles were recovering tone during this period (−6±6%; P<0.05) (Figure 5E and 5F; supplemental Movie VI). Thus for regions in which the Ca2+ wave propagated, arteriolar dilation was greater, occurred subsequent to the peak rise in Ca2+, and was sustained longer than at remote sites at which only the rapidly-conducted vasodilation occurred.
Figure 6 illustrates the time dependence of Ca2+ wave propagation and dilation analysis in a representative arteriole, comparing responses to ACh at the local site of stimulation (yellow bar) where both processes overlap, a remote site (blue bar) to which the propagated Ca2+ wave does not reach and only electrically-conducted vasodilation occurs, and an intermediate site (red bar), at which electrically conducted vasodilation occurs, followed by a secondary dilation associated with arrival of the Ca2+ wave (supplemental Movie VII). A virtual linescan across the vessel wall at the intermediate site (Figure 6B) reveals an initial rapid dilation that is independent of Ca2+ at that site and achieves a brief plateau (the apparent decrease in wall fluorescence in the linescan is explained in Figure 4B). The Ca2+ wave reaches the intermediate site of the linescan as shown by a rise in fluorescence and a secondary dilation, resulting in a biphasic dilation (Figure 6C), with the Ca2+-dependent component of dilation occurring at a slower rate. Linescans taken near ACh delivery or at the most remote site do not display a biphasic vasodilation (Figure 6D), but for different reasons. At the furthest site, conducted vasodilation occurs without a change in fluorescence because the Ca2+ wave does not propagate this far (or decays to ineffective oscillations) and only the rapid component is observed. By contrast, at the site of ACh exposure, there is a marked rise in Ca2+ that persists through the entire vasomotor response (Figure 6E), obscuring the separation of respective processes.
Endothelial cell signaling underlies an array of vascular responses that coordinates tissue blood supply. It is clear that propagated vasomotor responses in arterioles involve signals between individual ECs, as well as between endothelial and SM cells.6,12 The study of these cell–cell interactions has been hampered by difficulties in applying techniques that monitor intercellular signaling in the intact vascular bed, which prompted us to develop a genetic strategy to direct the expression of the genetically encoded Ca2+ indicator GCaMP2 to the arterial endothelium. When expressed as a transgene in mice, this indicator protein has several important attributes including high brightness and dynamic range along with stable optical properties.13 Here we report the development of mice with high GCaMP2 expression in arteriolar endothelial and Purkinje cells and the use of these mice to examine Ca2+ signaling during endothelium -dependent vasodilation in vivo.
Rationale for the Genetic Strategy
We initially attempted to use the minimal Tie2 promoter to direct GCaMP2 expression, as this promoter has been used to drive expression of EGFP in ECs.14 Unfortunately, Tie2-GCaMP2 transgenic mice displayed little detectable indicator expression within ECs. To achieve EC expression we inserted the GCaMP2 cDNA into the initial codon of Cx40 within a locus-spanning BAC, an approach that avoided limitations of Cx40 promoter constructs.15,16,20 We have found that BAC transgenesis provides an excellent recapitulation of the endogenous gene, enabling both lineage restriction and high levels of transgene expression required for detection of fluorescent protein-based genetic indicators.16
In the mouse, Cx40 mRNAs are transcribed by alternative splicing, resulting in 3 different 5′ UTRs mRNAs.15,18 We inserted GCaMP2 into the common Cx40 start codon in exon2 of the BAC by homologous recombination, placing the sensor cDNA under the control of Cx40 transcriptional elements. Insertion of the GCaMP2 disrupts exon2 of Cx40 within the BAC, preventing Cx40 overexpression from the BAC transgene (Figure 1). Transgene expression was consistent with endogenous Cx40,15 and costaining with Cx40 antibodies revealed excellent expression overlap. Although the expression of Cx40 in capillaries remains controversial,9 GCaMP2 expression in pulmonary capillaries (Figure 1) provides support for the presence of this gap junction protein in some capillary beds.
Ca2+ Signaling in the Cardiac Conduction System
The Purkinje network coordinates ventricular excitation and Purkinje–ventricular junctional coupling is critical in the generation of ventricular arrhythmias.21 Activation of this system is poorly studied in situ because of the difficulty of selectively loading and recording optical signals from this network. Recently, intracellular Ca2+ measurements were reported in Purkinje fibers22 by loading Langendorf -perfused hearts with Fluo3/AM and staining the network by acetylthiocholine iodide. Within the ventricle, Cx40BAC-GCaMP2 mice express GCaMP2 exclusively in the Purkinje network, providing a simple and powerful method to examine Ca2+ signaling in these myocytes and activation of the network. These studies can be combined with myocardial cell labeling, enabling the study of junctional coupling.
The Role of Ca2+ in Endothelium-Dependent Dilation of Arterioles In Vivo
We used Cx40BAC-GCaMP2 mice to examine the relationship between EC Ca2+ signaling and endothelium-dependent vasodilation. Stimulation with ACh elicited an initial local rise in Ca2+ and vasodilation, similar to observations in isolated arterioles2,23 and consistent with findings that ACh-induced vasodilation is inhibited by chelation of intracellular Ca2+ in ECs.23 A local rise in EC Ca2+ stimulates Ca2+-activated K+ channels, resulting in hyperpolarization and rapidly-conducted vasodilation.3,6,24,25 Ca2+ release within ECs is not required for vasodilation since SNP relaxed arterioles with no change in Ca2+ (Figure 4; supplemental Movie III) and rapidly-conducted vasodilations occurred without a change in Ca2+ at distances >1 mm (Figures 3 and 6⇑). Importantly, the propagating wave of Ca2+ along the endothelium evoked a second phase of dilation, indicating that a rise in EC Ca2+ is itself sufficient to produce vasodilation. This was most clearly seen by the separation of the electrically conducted and Ca2+ wave-dependent vasodilation in time (Figure 6). These findings are consistent with studies in isolated hamster feed arteries in which the rapidly-conducted vasodilation was blocked by inhibition of Ca2+-activated K+ channels, revealing a more slowly conducting vasodilation that was preceded by a Ca2+ wave along the endothelium.3 As this slower vasomotor response was blocked by inhibiting nitric oxide synthase and cyclooxygenase, it can be explained by the release of autacoids from ECs subsequent to the rise in Ca2+11.
Our findings indicate that the endothelium-dependent arteriolar response to ACh consists of 2 distinct complementary processes in vivo. The first mechanism involves a rapidly-conducted vasodilation (ie, electromechanical) that is triggered by a local rise in Ca2+ within ECs and extends to encompass sites beyond the propagating EC Ca2+ wave. The second mechanism involves the slower, and more restricted, propagation of a Ca2+ wave along the endothelium, resulting in additional vasodilation in regions to which the Ca2+ wave spreads. As illustrated in Figure 6, these distinct mechanisms overlap, but can be temporally and spatially separated at sites of intermediate distance from the stimulus.
A major question concerns the mechanism of Ca2+ wave propagation along arteriolar ECs. In the present study, Ca2+ waves propagated at an average initial velocity of 116 μm/s, similar to a recent report of Ca2+ wave propagation velocity (111 μm/s) in isolated feed arteries,5 but in contrast to a study in isolated hamster cheek arterioles reporting that EC Ca2+ responses evoked by ACh do not propagate.2 These differences may relate to loading conditions or constraints on imaging after dye loading of isolated vessels. Although Ca2+ waves appeared to slow as they traveled (Figure 3), we suggest that their propagation along the endothelium requires an active mechanism, as the initial velocity is more than 1000-fold faster than expected for free diffusion.26,27 Nevertheless, propagation of Ca2+ waves is markedly slower than that of hyperpolarization and rapidly-conducted vasodilation.3,6,24,25 It is possible that regenerative release of InsP3 triggered by ACh24,25 underlies Ca2+ wave transmission along arterioles. The propagation velocity observed here is markedly faster than the rate of InsP3 Ca2+ waves traveling through cellular cytoplasm (20 μm/s),28,29 suggesting that additional coupling mechanisms between cells may be involved.
In summary, Cx40BAC-GCaMP2 mice represent an important tool in the study of cardiovascular biology and will be useful for future studies of Ca2+ signaling within arterial/arteriolar ECs. The present findings demonstrate the first use of these animals to resolve distinct yet complementary endothelium-dependent signaling pathways for coordinating vasodilation along branches of the arterial network in vivo during blood flow control to skeletal muscle. We show that vasodilation in response to ACh is initiated by a local rise in EC Ca2+, which triggers a rapidly-conducted vasodilation that encompasses multiple branches and occurs independent of EC Ca2+ at remote sites. In turn, the EC Ca2+ wave propagates intercellularly to contribute additional vasodilation. Taken together, respective signaling pathways can coordinate the onset and magnitude of vasodilation throughout resistance networks.
We thank Dr Alcaraz for GCaMP2 expression evaluation and Pat Fisher for immunohistochemistry.
Sources of Funding
This work was supported by NIH grants HL45239 and DK65992 (M.K.); F2HL76999 (Y.T.); Hl69097, HL70722, and HL057929 (G.S.); HL41026 and HL56786 (S.S.) and grant-in-aid from Ministry of Education, Culture, Sports, Science, and Technology of Japan (J.N.).
↵*The first three authors contributed equally to this study.
Duza T, Sarelius IH. Localized transient increases in endothelial cell Ca2+ in arterioles in situ: implications for coordination of vascular function. Am J Physiol Heart Circ Physiol. 2004; 286: H2322–H2331.
Dora KA, Xia J, Duling BR. Endothelial cell signaling during conducted vasomotor responses. Am J Physiol Heart Circ Physiol. 2003; 285: H119–H126.
Yashiro Y, Duling BR. Participation of intracellular Ca2+ stores in arteriolar conducted responses. Am J Physiol Heart Circ Physiol. 2003; 285: H65–H73.
Uhrenholt TR, Domeier TL, Segal SS. Propagation of calcium waves along the endothelium of hamster feed arteries. Am J Physiol Heart Circ Physiol. 2007; 292: H1634–H1640.
Emerson GG, Segal SS. Electrical coupling between endothelial cells and smooth muscle cells in hamster feed arteries: role in vasomotor control. Circ Res. 2000; 87: 474–479.
de Wit C, Roos F, Bolz SS, Pohl U. Lack of vascular connexin 40 is associated with hypertension and irregular arteriolar vasomotion. Physiol Genomics. 2003; 13: 169–177.
Figueroa XF, Paul DL, Simon AM, Goodenough DA, Day KH, Damon DN, Duling BR. Central role of connexin40 in the propagation of electrically activated vasodilation in mouse cremasteric arterioles in vivo. Circ Res. 2003; 92: 793–800.
Looft-Wilson RC, Payne GW, Segal SS. Connexin expression and conducted vasodilation along arteriolar endothelium in mouse skeletal muscle. J Appl Physiol. 2004; 97: 1152–1158.
Brekke JF, Jackson WF, Segal SS. Arteriolar smooth muscle Ca2+ dynamics during blood flow control in hamster cheek pouch. J Appl Physiol. 2006; 101: 307–315.
Budel S, Bartlett IS, Segal SS. Homocellular conduction along endothelium and smooth muscle of arterioles in hamster cheek pouch: unmasking an NO wave. Circ Res. 2003; 93: 61–68.
Dora KA, Doyle MP, Duling BR. Elevation of intracellular calcium in smooth muscle causes endothelial cell generation of NO in arterioles. Proc Natl Acad Sci U S A. 1997; 94: 6529–6534.
Tallini YN, Ohkura M, Choi BR, Ji G, Imoto K, Doran R, Lee J, Plan P, Wilson J, Xin HB, Sanbe A, Gulick J, Mathai J, Robbins J, Salama G, Nakai J, Kotlikoff MI. Imaging cellular signals in the heart in vivo: Cardiac expression of the high-signal Ca2+ indicator GCaMP2. Proc Natl Acad Sci U S A. 2006; 103: 4753–4758.
Tallini YN, Shui B, Greene KS, Deng KY, Doran R, Fisher PJ, Zipfel W, Kotlikoff MI. BAC Transgenic Mice Express Enhanced Green Fluorescent Protein in Central and Peripheral Cholinergic Neurons. Physiol Genomics. 2006; 27: 391–397.
Hisatsune H, Matsumura K, Ogawa M, Uemura A, Kondo N, Yamashita JK, Katsuta H, Nishikawa S, Chiba T. High level of endothelial cell-specific gene expression by a combination of the 5′ flanking region and the 5′ half of the first intron of the VE-cadherin gene. Blood. 2005; 105: 4657–4663.
Koeppen M, Feil R, Siegl D, Feil S, Hofmann F, Pohl U, de Wit C. cGMP-dependent protein kinase mediates NO- but not acetylcholine-induced dilations in resistance vessels in vivo. Hypertension. 2004; 44: 952–955.
Ungvari Z, Csiszar A, Koller A. Increases in endothelial Ca(2+) activate K(Ca) channels and elicit EDHF-type arteriolar dilation via gap junctions. Am J Physiol Heart Circ Physiol. 2002; 282: H1760–H1767.
Jacobs MH. Diffusion Processes. Ergebn Biol. 1935; 12: 1–160.
Allbritton NL, Meyer T, Stryer L. Range of messenger action of calcium ion and inositol 1,4,5-trisphosphate. Science. 1992; 258: 1812–1815.
Newman EA, Zahs KR. Calcium waves in retinal glial cells. Science. 1997; 275: 844–847.