OUP user menu

Expression of ryanodine receptor type 3 and TRP channels in endothelial cells: comparison of in situ and cultured human endothelial cells

Ralf Köhler, Susanne Brakemeier, Meike Kühn, Christiane Degenhardt, Heinz Buhr, Axel Pries, Joachim Hoyer
DOI: http://dx.doi.org/10.1016/S0008-6363(01)00281-4 160-168 First published online: 1 July 2001


Objective: Ca2+ mobilization plays an important role in endothelial function by stimulating Ca2+-dependent synthesis of vasodilating factors. In addition to inositol-1,4,5-trisphosphate (InsP3) mediated Ca2+ mobilization, Ca2+ release from ryanodine-sensitive pools and Ca2+-influx through TRP channels have been suggested to be important in endothelial Ca2+-signaling. However, the function and molecular identity of TRP channels and ryanodine receptors in human endothelium in situ are still elusive. We hypothesized that expression of ryanodine-receptors (RyR) and TRP channels differs between human endothelium in situ and in cultured cells. Methods: By combining single-cell RT-PCR and patch-clamp techniques, expression of RyR and TRP channels was determined in situ in endothelial cells of human mesenteric artery (HMAECs) obtained from patients undergoing bowel resection and in the endothelial cell line EA.hy926. Results: At the single cell level, expression of RyR 3 was detected in 25 and 5% of HMAECs and EA.hy926 samples, respectively. Expression of the RyR 1 and 2 was not detected in either HMAECs or EA.hy926. In patch-clamp experiments in HMAECs, applications of caffeine (0.5 mM) induced sustained hyperpolarization mediated by activation of Ca2+-activated K channels. In EA.hy926, caffeine-induced hyperpolarization was not detected. Single HMAECs expressed the TRP genes, TRP1 and TRP3, but not TRP 4 and 6. The TRP1 was the predominantly expressed TRP gene in HMAECs in situ whereas TRP3 expression was rarely detected. EA.hy926 expressed only TRP1. In patch clamp experiments in HMAECs, Ca2+-store depletion activated non-selective cation currents leading to Ca2+ entry. Conclusions: Our findings suggest that, in addition to InsP3 mediated Ca2+ release, Ca2+ release from ryanodine-sensitive stores mediated by RyR3 and Ca2+ entry through TRP1 might represent important components of endothelial Ca2+ signaling in situ and thereby of endothelial function in intact human blood vessels.

  • Calcium (cellular)
  • Endothelial function
  • Endothelial receptors
  • Gene expression
  • Signal transduction

Time for primary review 26 days.

1 Introduction

In endothelial function, modulation of free cytosolic calcium ([Ca2+]i) in response to humoral factors or hemodynamic forces [1,2] is an essential step of intracellular signal transduction. For instance, mobilization of [Ca2+]i stimulates Ca2+ dependent synthesis of vasodilating factors [3] and gene expression [4]. Changes of endothelial [Ca2+]i are mediated by at least five major mechanisms [2]: (1) nonspecific leak, (2) Ca2+ ATPases (3) Ca2+ release from internal stores and Ca2+ entry through (4) mechanosensitive or (5) store-operated cation channels (SOCs).

The rapid increase of [Ca2+]i induced by humoral factors such as bradykinin is mediated by formation of inositol-1,4,5-trisphosphate (InsP3) which binds to InsP3 receptors [5] and thereby initiates Ca2+ release from intracellular stores. In addition to this InsP3-mediated Ca2+-release, also Ca2+ release from ryanodine-sensitive stores has been detected in cultured bovine and human endothelial cells (ECs) [6,7] thus indicating the presence of a ryanodine-sensitive Ca2+-induced Ca2+ release (CIRC) mechanism in the endothelium.

In Ca2+ entry subsequent to receptor/second messenger activation (for review see [8–10]) mammalian homologues of the transient receptor potential (trp) gene originally cloned from Drosophila melanogaster have been implicated. Like trp, all functionally expressed TRP genes code for cation non-selective Ca2+ permeable channels. Six human TRP homologues have been cloned so far [11–15]. TRP1 and TRP4 have been proposed to be store-operated [15], whereas TRP3 and TRP6 appeared to be insensitive to store depletion, but are directly activated by diacylglycerol as reported recently [15]. Expression of some TRP channels has been detected, although inconsistently so, in cultured ECs [16–19].

Despite these existing data obtained from cultured ECs, function and molecular identity of RyR and TRP channels are still elusive in endothelium of intact human blood vessels. Moreover, expression of RyR and TRP genes in endothelium of differentiated vessels might vary essentially from that of proliferating ECs kept under highly artificial cell culture conditions. Therefore, we set out a study to directly identify and characterize RyR and TRP channels in intact endothelium of human mesenteric arteries (MAs). For this purpose, we adapted the single-cell RT-PCR approach in combination with the patch-clamp technique, which allowed a simultaneous molecular biological and functional characterization of RyR and TRP channels in situ by avoiding alterations of channel expression and function due to cell isolation and culturing. To compare channel expression in intact endothelium of MAs with that of cultured ECs, we determined RyR and TRP channel expression in the endothelial cell line (EA.hy926) originated from human umbilical vein ECs.

2 Methods

2.1 Preparation and isolation of MAs

Third order MAs of about 0.5 –1 cm in length×0.4–1.1 mm O.D. were excised out of the arterial arcades from colon specimens of patients subjected to hemicolectomy. Tissue collection was performed following the guidelines of the the local ethics committee. For patch-clamp experiments and cell harvesting procedures, vessel slices of about 2 mm2 were fixed on a holding glass capillary and placed in the experimental chamber mounted on the stage of an inverted microscope (Zeiss Axiovert 135).

EA.hy926 (a kind gift from C.J.S. Edgell) were grown to confluency on uncoated coverslips.

2.2 Harvesting of single endothelial cells

Vessel slices were pre-incubated with 0.05% trypsin and 0.02% EDTA in PBS without Ca2+ and Mg2+ for up to 30 min. The enzymatic reaction was stopped by a 5-min superfusion with PBS containing Ca2+ and Mg2+. Under microscopic control, a slightly rounded HMAEC was then selectively fixed by aspiration at the tip of the patch pipette and mechanically detached from the vessel wall. The enzymatic pretreatment was feasible and efficient to allow mechanical removal of up to 15 single HMAECs in 1 h. Single EA.hy926 were analogously harvested from cell monolayers.

2.3 Reverse transcription

ECs and content (∼6 μl) of the patch pipette were expelled into a test tube containing 1 μl ‘first strand’ buffer (Life Technologies), 0.5 μl dNTPs (10 mM each; Promega), 1 μl ‘random’ hexamer primer (100 μM; Boehringer Mannheim), 1 μl DTT (0.1 M; Life Technologies) and 0.5 μl RNase-inhibitor (40 U/μl; Promega). After one freeze–thaw cycle to induce breakdown of the cell membrane, 0.5 μl SuperScript®RT (200 U/μl; Life Technologies) was added and the final volume (∼10 μl) was allowed to reverse transcribe for 1 h at 37°C. Thereafter, the total RT product was immediately used for the polymerase chain reaction (PCR). In order to exclude contamination by non-endothelial mRNA, samples of the bath solution were aspirated next to the endothelial cells.

2.4 Polymerase chain reaction

A modified single-cell RT-PCR was performed as described previously [20,21]. The most efficient PCR conditions i.e. primer combination with maximal sensitivity, MgCl2 concentration, number of cycles, and annealing temperature were determined by using a serial dilution of human mesenteric artery cDNA. Sequence-specific oligonucleotide-primers (Tib Mol Biol, Berlin) were selected which yielded no PCR products of expected size from human genomic DNA (50 μg/μl) and of cell samples processed without reverse transcription (n=15). The respective primer pairs are listed in Table 1. In a single-cell sample, cDNA of InsP3R, RyR and TRP channels were co-amplified along with the von Willebrand factor (vWF) cDNA as an endothelial cell marker and with myosin heavy-chain (MyHC) cDNA to exclude contamination with smooth muscle cell cDNA. A first ‘multiplex’ PCR was performed in a final volume of 50 μl containing 5 μl PCR-buffer (10×), 2 μl dNTPs (10 mM each), 4 μl MgCl2 (50 mM), 1 μl of each sense-primer (10 pmol), 1 μl of each antisense-primer (10 pmol), ∼10 μl RT-product, and 0.5 μl Taq DNA polymerase (5 U/μl; (all Life Technologies) using a programmable Biozym Maxicycler PTC 9600. Samples were incubated for 5 min at 94°C, followed by 50 cycles of 30 s at 94°C, 1 min at 55°C, 1 min at 72°C, and a final elongation for 10 min at 72°C. In a second multiplex PCR round with ‘nested’ primers, 5 μl of the first PCR product were used for reamplification (45 cycles) with an increased annealing temperature of 60°C to increase the specificity of the PCR. Amplified cDNAs were analyzed on a 2% agarose gel containing ethidium bromide. A 50-bp DNA ladder (Life Technologies) was used as a molecular weight marker. Identities of the amplified cDNA fragments were further confirmed by restriction site analysis and by sequencing using an ABI 377 automatic sequencer (ABI Prism, Weiterstadt, Germany).

View this table:
Table 1

Primer nucleotides

GeneAccession no.Outer primer pairsNested primer pairsProduct sizes (bp)

2.5 Patch-clamp experiments

Membrane currents were recorded with a EPC-9 (HEKA, Lambrecht, Germany) patch-clamp amplifier, low-pass-filtered (−3 dB, 1000 Hz) at a sample time of 0.5 ms [22]. The endothelial membrane potential was recorded in the current clamp mode of the EPC9 patch-clamp amplifier. Patch pipettes were pulled from borosilicate glass capillaries with 0.3 mm wall thickness and had a tip resistance of 2–4 MΩ in symmetrical KCl solution. The seal resistance in cell-attached patches ranged from 4 to 10 GΩ. In whole-cell patch-clamp experiments, the pipette solution contained (in mM): 135 KCl, 4 MgCl2, 1 EGTA, and 5 HEPES (pH 7.2). In a subset of experiments, the pipette solution contained additionally 0.955 mM CaCl2, ([Ca2+]free=3 μM). In single-channel patch-clamp experiments, additional pipette solutions were used: a NaCl pipette solution containing (in mM): 140 NaC1, 4.3 KCl, 1.3 CaC12, 1 MgC12, and 10 HEPES (pH 7.4), and a ‘high’ Ca2+ pipette solution containing: 90 CaCl2, 1 MgC12, and 10 HEPES (pH 7.4). The bath solution contained (in mM): 137 NaCl, 4.5 Na2HPO4, 3 KCl, 1.5 KH2PO4, 0.4 MgCl2, and 0.7 CaCl2 (pH 7.4). The Ca2+ free bath solution contained (in mM): 1 EGTA and 0 CaCl2.

All experiments were performed at 37°C. To destroy RNA-degrading enzymes all solutions were treated with 0.1% diethylpyrocarbonate (Sigma, Deisenhofen, Germany) and autoclaved. Data analysis was performed as described previously [23].

For statistics, the Mann–Whitney U test or χ2 analysis were used as indicated. P values of <0.05 were considered significant. If not otherwise stated, data are given as mean±S.E.

3 Results

3.1 Single-cell RT-PCR in intact endothelium of human MAs and in EA.hy 926

From freshly dissected intact MAs a total of 151 cDNA samples of a single HMAEC were taken and analyzed for gene expression. Expression of vWF was detected in 122 of 151 (81%) samples. The 29 vWF negative samples from MAs yielded no PCR-products of all other genes investigated. These negative samples are most likely explained by a loss of the cell during transfer or mRNA degradation rather than by the harvest of a non-endothelial cell.

For negative controls, every fifth sample was a medium samples (n=30) aspirated near by the endothelial surface. None of these medium sample yielded a positive PCR signal. Expression of MyHC could not be co-detected in any of the cell samples (0/151). From 23 EA.hy926 samples, only 11 samples (48%) gave a positive PCR signal for vWF, which was significantly less when compared to HMAECs (P=0.0015, χ2 analysis). MyHC expression was not detected in EA.hy926 samples.

3.2 Expression of InsP3R1 and RyR in HMAEC from intact MAs and EA.hy926

Expression pattern of InsP3R1, RyR, and TRP channels in single HMAECs and EA.hy926, were analogously investigated using the ‘multiplex’ RT-PCR approach. Fig. 1A and B shows the results of a ‘multiplex’ RT-PCR analysis of single HMAEC and single EA.hy926. Expression of InsP3 R1 was detected in 68 of 122 (56%) vWF positive HMAEC samples. Regarding RyR, expression of RyR3 was detected in 30 (25%) of the samples whereas expression of RyR1 and RyR2 RT-PCR was not observed in any of these cell samples. In EA.hy926, expression of InsP3 R1 was detected in 11 of 16 samples (69%; vs. HMAEC, P=0.52), from which, however, only four samples yielded a vWF RT-PCR signal. In contrast to HMAECs, expression of RyR3 was detected in only 1 of 19 samples (5% vs. HMAEC, P=0.1). Expression of RyR1 and RyR2 were not detected in single EA.hy926.

Fig. 1

Expression pattern of TRP channels, InsP3 R1, and RyR3 in single HMAECs (A) and from human MA and EA.hy926 (B). Ethidium bromide-stained gels of ‘multiplex’ RT-PCR products from vWF positive and MyHC negative HMAEC and from EA.hy926, medium controls, and −RT controls. Molecular weight markers are shown in the lanes of the far right.

3.3 Expression of TRP channels in HMAECs from intact MA and in EA.hy926

In HMAECs, expression of TRP1 was detected in 24 of 146 (16%) vWF-positive cell samples. TRP3 was less frequently detected: in 2 of 69 (3%) cell samples. TRP4 and TRP6 were not detected in these cell samples. In single EA.hy926, the percentage of TRP1-positive cell samples (20 of 26; 77%) was higher than in HMAECs (P<0.001). Expression of TRP3 as well as of TRP4 and TRP6 was not detected in EA.hy926.

3.4 Ca2+ release and Ca2+ entry dependent changes of membrane potential in intact endothelium

Prior to molecular biological studies, we conducted patch-clamp experiments in each vessel preparation to characterize the function of Ca2+-influx and Ca2+-release channels. To monitor changes of free cytosolic calcium, we measured endothelial resting potentials and agonist-induced membrane potential changes induced by Ca2+-release and subsequent stimulation of Ca2+-activated ion channels which served as physiological Ca2+ sensor [20,24,25]. After obtaining stable electrical access high capacitance values of >500 pF indicated an intact electrical coupling of HMAECs. Application of bradykinin (100 nM) induced hyperpolarization of the cell membrane from a resting potential of −28 mV±3 S.E. to −46 mV±4 S.E. (n=14), which lasted for up to 5 min (Fig. 2A). A similar hyperpolarization response from −26 mV±8 S.E. to −51 mV±6 S.E. (n=4) could be evoked by caffeine (0.5 mM, Fig. 2B). In EA.hy926 application of bradykinin, but not of caffeine induced a comparable hyperpolarization (data not shown). Agonist-induced hyperpolarization of HMAECs is completely suppressed by CTX, a blocker of Ca2+-activated potassium channels [16,20]. Function and expression of these Ca2+-activated potassium channels in endothelium of human MAs have recently been described in more detail [20].

Fig. 2

Representative current clamp-recordings of endothelial cell potential of intact human mesenteric artery. Endothelial hyperpolarization was induced by application of 100 nM bradykinin (A) or 0.5 mM caffeine (B). Arrow indicates start of solution exchange. Cell potential changes induced by bradykinin in the absence (C) and after complete store depletion and subsequent re-addition of extracellular Ca2+ without (D) or with 50 μM Gd3+ (E) in the bath solution. Note the more positive cell potential in the absence of free extracellular Ca2+.

For a functional characterization of store-operated Ca2+ entry, we conducted a series of current clamp experiments in the absence or presence of extracellular Ca2+. In the absence extracellular Ca2+, endothelial membrane potential was shifted to more positive values (−4 mV±1 S.E., n=13). Under these conditions, application of bradykinin induced a transient hyperpolarization to −13 mV±4 S.E. mV (n=11), which lasted for 10–20 s. After repetitive agonist stimulation endothelial hyperpolarization was absent indicating complete emptying of intracellular Ca2+ stores (Fig. 2C). Subsequent to emptying of Ca2+ stores, reapplication of extracellular Ca2+ (1 mM, n=5) resulted in Ca2+ influx as deduced from the rapid repolarization to −46 mV±10 S.E. (Fig. 2D). Repolarization was almost completely blocked in the presence of 50 μM Gd3+ (1 mV±2 S.E., P<0.05, Mann–Whitney U test, n=3, Fig. 2E), which is known to inhibit store-operated Ca2+ entry [9].

3.5 Single-channel and whole-cell patch clamp analysis

To further characterize the currents underlying store-depletion-induced membrane potential changes, we conducted single-channel and whole-cell patch clamp experiments in intact endothelium and single uncoupled HMAECs, respectively. In cell-attached patch clamp experiments with a NaCl pipette solution, activation of single-channel currents carried by Na+ were observed within 30 s of the application of 100 nM bradykinin to the bath solution (Fig. 3a). In excised membrane patches with pipette solutions containing either high Na+, K+, or Ca2+ as outlined in detail in Methods, this channel had a mean conductance of 26 pS±5 S.D. for Na+ (n=5), 24 pS±4 S.D. for K+ (n=9), and 6 pS±1 S.D. (n=5) for Ca2+ at negative membrane potentials ranging from −100 to −10 mV (Fig. 3b). Permeability ratios for K+:Na+ and Ca2+:Na+ were 1:1 and 0.1:1, respectively, as calculated from reversal potentials of −1 mV and −37 mV, respectively, by using the Goldmann–Hodgkin–Katz equation [12]. In five experiments application of Gd3+ (20 and 100 μM) to the cytosolic face of the membrane resulted in a dose dependent inhibition of channel activity at a membrane potential of 70 mV (Fig. 3c).

Fig. 3

Representative single channel and whole-cell recordings from electrically uncoupled HMAECs. (a) Activation of a single non-selective channel after stimulation with 100 nM bradykinin. C, denotes closed state of channel. (b) Current–voltage relationship of non-selective cation channel in inside-out patches from HMAECs. Inhibition of single-channel currents by Gd3+ in an inside-out patch (c). Non-selective whole-cell currents induced by cell dialysis with 10 μM InsP3 (d) or with 3 μM Ca2+free (e). Ca2+-dependent whole-cell currents after replacing Na+ by NMDG+ in the external solution (f). Activation of whole-cell currents after stimulation with 100 nM bradykinin (g). Dose-dependent block of non-selective whole-cell currents by Gd3+ (h).

In whole-cell patch clamp experiments in a single electrically uncoupled HMAEC (n=6) lacking Ca2+-activated K+ currents [20], cell dialysis with 10 μM InsP3 via the patch pipette, induced an inwardly directed current of 7.7±3 pA/pF at a holding potential of −50 mV leading to a shift of the interpolated reversal potential from −9±2 mV to −2±1 mV (Fig. 3d). However current responses to IP3 dialysis varied considerably between the cells (IP3-induced inward current at −50 mV: −1.4; −1.9; −4.7; −6.4; −9.1, −22.5 pA/pF). Comparable non-selective currents were activated after agonist-stimulation (Fig. 3e) or when the cells were dialyzed with a pipette solution containing 3 μM Ca2+free (Fig. 3f). The inward current component was eliminated by replacing external Na+ with NMDG+ (Fig. 3g). When chloride was substituted by equal amounts of aspartate in bath and pipette solution, currents responses were not altered, thus, indicating that chloride currents to do not considerably contribute to InsP3 and Ca2+-dependent cell currents. The non-selective cation current was sensitive to Gd3+ with a Kd value of 4±2 μM (n=3, Fig. 3h).

4 Discussion

In the study we applied the single-cell RT-PCR technique to analyze gene expression in single HMAECs from intact endothelium of human mesenteric arteries and in single EA.hy926. Efficiency and specificity of this technique was demonstrated by the high percentage (81%) of cell samples from the endothelium of MAs yielding an endothelial specific vWF RT-PCR product. Due to its high sensitivity, the single-cell RT-PCR is particularly vulnerable to minimal contamination which could lead to false positive PCR signals. Therefore, it was of crucial importance to exclude such cDNA or mRNA contamination by including a high number of negative controls in each PCR analysis. In fact, none of the medium or water controls yielded a positive PCR signal. The harvest of VSMCs or contamination by mRNA from leaking VSMCs appeared to be unlikely, since MyHC gene expression was not co-detected in samples giving vWF RT-PCR signal. This assured us about the endothelial origin of the cDNA.

4.1 InsP3R and RyR in human endothelial cells

InsP3-mediated Ca2+ release from internal stores and InsP3R expression have been extensively characterized in cultured ECs from various species and vessels and is considered the major mediator of agonist-induced Ca2+ release from internal stores [6]. In the present study, the finding of a high percentage of InsP3R1 positive HMAECs from intact MAs supports these previous results. Moreover, the observation of bradykinin-induced Ca2+ mobilization and subsequent activation of Ca2+-dependent K+ channels (KCa) leading to endothelial hyperpolarization [20,24] indicated the presence of functional active InsP3R.

In addition to InsP3-mediated Ca2+ release, a Ca2+ release from ryanodine-sensitive stores has been implicated in endothelial Ca2+ signaling. In previous studies in bovine aortic ECs, RyR proteins have been identified in endoplasmic reticulum membranes by antibody techniques [26]. Moreover, in human ECs and rabbit aortic ECs [27,28], RyR function was demonstrated by monitoring [Ca2+]i and by electrophysiological studies using low concentrations of ryanodine or caffeine as an agonist for RyR stimulation, respectively. More recently, we showed that Ca2+ release from ryanodine-sensitive stores is involved in shear stress-induced Ca2+ mobilization and STOC activation in bovine aortic ECs [20]. However, these studies did not reveal the molecular identity of RyR. In the present in situ study in endothelium of human arteries, we observed expression of RyR3 but not that of the other two subtypes, RyR1 and RyR2. These latter RyR subtypes might thus not be expressed in human mesenteric endothelium.

From our single-cell RT-PCR analysis and patch-clamp data, we conclude that caffeine-induced Ca2+-mobilization is presumably mediated by RyR3 in intact human endothelium. Moreover, this RyR3 mediated Ca2+-signaling is not preserved in cultured ECs as RyR3 expression is greatly reduced and caffeine-induced and Ca2+-dependent hyperpolarization is absent in EA.hy926.

4.2 TRP channels in human endothelial cells

As shown in a previous study [20], stimulation of the endothelium with bradykinin induced a rapid and transient hyperpolarization mediated by KCa activation due to Ca2+ release from InsP3-sensitive stores. This transient hyperpolarization was followed by sustained hyperpolarization, which depended on Ca2+ entry since it was only observed in the presence of extracellular Ca2+. This indicates that a Ca2+ entry mechanism is activated subsequent to store-depletion. This interpretation was supported by our observation that reapplication of extracellular Ca2+ after complete emptying of intracellular stores induced a rapid repolarization which was inhibited by Gd3+, a blocker of Ca2+ entry. In this membranous mechanisms of Ca2+ entry, the TRP channels have been implicated to function as receptor/second messenger-operated Ca2+-entry channels [8–13]. A recent study showed expression of the TRP genes TRP1, TRP3 and TRP4 gene in cultured human umbilical vein ECs [16]. Another study showed expression of TRP1 but not of TRP3 and TRP6 in cultured human pulmonary artery endothelial cells [17]. In the present study on HMAECs, only expression of TRP 1 and TRP3 was observed. TRP1 appeared to be the predominantly expressed member of the TRP genes in the endothelium of MAs. Compared to TRP1, expression of TRP3 seemed to be very low in HMAECs, since expression was detected in only 3% of the cells. Expression of TRP4 and TRP6 genes could not be observed in endothelium of human MAs. The finding of TRP1 expression in HMAECs is consistent with the apparently wide tissue distribution of this gene [29] whereas expression of TRP3, TRP4, and TRP6 have been suggested to be more restricted to brain [30].

The discrepancy between expression of TRP genes in HMAECs and that found in isolated and cultured human ECs [16,17] could be due to alterations of expression of TRP channels as a consequence of cell isolation and cell culture. The differences in channel expression in situ and in cell culture strongly suggests that ion channel expression in cultured ECs does not necessarily reflect the in vivo situation. This suggestion is further supported by our observation that expression of TRP1 in EA.hy926 seems to be considerably higher than in HMAECs and expression of TRP3 is not detectable in this cell line. These findings of high TRP1 expression and missing TRP3 expression suggest that, compared to native ECs, expression of ion channel is considerably altered in EA.hy926 which might be due to artificial culture conditions, such as exposure to FCS containing high levels of a variety of growth factors.

In patch clamp experiments in HMAECs, we detected activation of single cation channels after agonist stimulation and activation of whole-cell cation currents by infusion of InsP3 or Ca2+. The non-selective Ca2+-permeable cation channel observed in the present single channel patch-clamp experiments is characterized by well distinguishable channel openings and had a unitary conductance of 26 pS for monovalent cation. Although there is no direct experimental evidence, this channel matches some electrophysiological properties of a human TRP1 functionally expressed in COS cells [9] and of a non-selective cation channel identified in EA.hy926 [29]. For TRP3, unitary conductances ranging from 17 to 60 pS have been reported [14,30,31]. In contrast to TRP1 and the non-selective channel described in this study, single channel activity of TRP3 is characterized by very short lasting openings [30]. All our efforts failed to record such TRP3 related single-channel currents in HMAECs. This failure might be explained by the low expression of TRP3 in HMAECs or by the forming heteromultimeric channels consisting of TRP1 and TRP3. Interestingly, with respect to the activation mechanism for TRP3, increases of the cytoplasmic Ca2+-concentration have been shown to activate TRP3 channels functionally expressed in COS cells [14]. In whole-cell patch-clamp experiments, we could show that cell dialysis with Ca2+ activated non-selective cation currents. This might indicate that TRP3 channels, presumably as part of heteromultimeric channels, contributes to the Ca2+-activated non-selective cation current in HMAECs.

Overall, at the single cell level, we observed large variations in the whole-cell current response to IP3 dialysis, which might explain by the considerable heterogeneity of TRP channel, IP3R1, and RyR3 expression pattern in HMAECs in situ. It is tempting to speculate that this might indicate the presence of specialized HMAECs for agonist-induced signaling in the endothelial monolayer.

In summary, we established combined single cell RT-PCR and patch-clamp analyses in situ to characterize putative Ca2+-influx channels and Ca2+-release channels of the ER in single HMAECs from very small vessel specimens by avoiding alterations due to cell culture. We demonstrated expression of the putative Ca2+ influx channels TRP1 and TRP3 and endoplasmic Ca2+ release channels InsP3R1 and RyR3 at the single-cell level. In addition to InsP3 mediated Ca2+ store depletion, Ca2+ release from ryanodine-sensitive stores and Ca2+ entry through TRP1/3 channels might represent important components of endothelial Ca2+ signaling and thereby of endothelial function in human blood vessels.


This work was supported by the Deutsche Forschungsgemeinschaft (FOR 341/1/5/7, HO 1103/2–4).


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
View Abstract