We previously demonstrated that endothelium-derived hyperpolarizing factor (EDHF)-mediated dilations in cerebral arteries are significantly reduced by inhibitors of PLA2. In this study we examined possible mechanisms by which PLA2 regulates endothelium-dependent dilation, specifically whether PLA2 is involved in endothelial Ca2+ regulation through stimulation of TRPV4 channels. Studies were carried out with middle cerebral arteries (MCA) or freshly isolated MCA endothelial cells (EC) of male Long-Evans rats. Nitro-l-arginine methyl ester (l-NAME) and indomethacin were present throughout. In pressurized MCA, luminally delivered UTP produced increased EC intracellular Ca2+ concentration ([Ca2+]i) and MCA dilation. Incubation with PACOCF3, a PLA2 inhibitor, significantly reduced both EC [Ca2+]i and dilation responses to UTP. EC [Ca2+]i was also partially reduced by a transient receptor potential vanilloid (TRPV) channel blocker, ruthenium red. Manganese quenching experiments demonstrated Ca2+ influx across the luminal and abluminal face of the endothelium in response to UTP. Interestingly, PLA2-sensitive Ca2+ influx occurred primarily across the abluminal face. Luminal application of arachidonic acid, the primary product of PLA2 and a demonstrated activator of certain TRPV channels, increased both EC [Ca2+]i and MCA diameter. TRPV4 mRNA and protein was demonstrated in the endothelium by RT-PCR and immunofluorescence, respectively. Finally, application of 4α-phorbol 12,13-didecanoate (4αPDD), a TRPV4 channel activator, produced an increase in EC [Ca2+]i that was significantly reduced in the presence of ruthenium red. We conclude that PLA2 is involved in EC Ca2+ regulation through its regulation of TRPV4 channels. Furthermore, the PLA2-sensitive component of Ca2+ influx may be polarized to the abluminal face of the endothelium.
- fura 2
- manganese quenching
- transient receptor potential
- endothelium-derived hyperpolarizing factor
- phospholipase A2
endothelium-derived hyperpolarizing factor (EDHF)-mediated dilation has been shown to exist in arteries and arterioles of multiple vascular beds. The mechanism of EDHF-mediated dilation is distinct from that of the nitric oxide (NO) and cyclooxygenase (COX) pathways. In rat middle cerebral artery (MCA), EDHF-mediated dilations result from endothelial free Ca2+ concentration ([Ca2+]i) increasing above a threshold of ∼330 nM and subsequent activation of intermediate-conductance Ca2+-activated K channels (IKCa) (15, 17). Activation of the IKCa channels promotes endothelial cell hyperpolarization, which then appears to be directly transferred to the surrounding smooth muscle (25).
In addition, our group recently demonstrated a critical role for cytoplasmic phospholipase A2 (PLA2) in EDHF-mediated dilation in rat MCA (33). In that study, we found that inhibitors of cytoplasmic PLA2 (cPLA2) significantly reduced EDHF-mediated dilations to the purinergic receptor agonist uridine triphosphate (UTP). The specific role of PLA2 in EDHF-mediated dilation, however, was not determined.
The purpose of the present study was to examine the mechanism by which PLA2 contributes to EDHF-mediated dilation. We evaluated two hypotheses regarding endothelial Ca2+ regulation and EDHF-mediated dilation. Our first hypothesis states that PLA2 is critical for normal endothelial Ca2+ regulation following UTP stimulation. In our proposed model, the critical role of PLA2 activation would be to promote the sustained or plateau component of endothelial [Ca2+]i to sufficient levels for IKCa activation (17). To examine the mechanism of Ca2+ regulation further, we examined the role of transient receptor potential (TRP) channels in mediating Ca2+ influx in MCA endothelium. Our second hypothesis states that TRPV4 channels contribute to endothelial Ca2+ influx and are regulated by products generated subsequent to activation of PLA2. For these studies we measured endothelial [Ca2+]i and Ca2+ influx in pressurized MCA, [Ca2+]i in isolated endothelial cells, and message for candidate TRP channels by cell-specific PCR.
Isolated/pressurized artery preparation.
Rat MCA were harvested and mounted in a pressurized vessel chamber as described previously (14, 17). MCA were pressurized to a mean transmural pressure of 85 mmHg, and flow was established through the lumen of the artery (100 μl/min) with PSS of the following composition (in mM): 119 NaCl, 4.7 KCl, 21 NaHCO3, 1.18 KH2PO4, 1.17 MgSO4, 0.026 EDTA, 5.5 glucose, and 1.6 CaCl2. PSS was warmed (37°C) and gassed with 5% CO2-20% O2-75% N2 to maintain a pH of 7.4. Nitro-l-arginine methyl ester (l-NAME) and indomethacin (10 μM each) were included in all experiments to inhibit NO synthase (NOS) and COX, respectively (18). The vessel chamber was custom-manufactured for performing fluorescence measurements in pressurized arteries (ChuelTech, Houston, TX).
In this preparation, compounds could be delivered via the abluminal or luminal perfusate to target them to the smooth muscle side or luminal endothelial side, respectively. Abluminal perfusion delivers compounds to the smooth muscle cells as well as the basolateral face of the endothelium. Luminal perfusion provides for delivery of compounds to the apical face of the endothelium. Because of the tight junctions between the endothelial cells of the MCA, hydrophilic compounds are prevented from passing across the endothelial layer (16).
Endothelial calcium measurements.
[Ca2+]i was selectively measured in the endothelium of pressurized arteries with fura-2 as described previously (14, 17). In brief, fura-2 AM was delivered through the lumen of the artery for a short period of time to selectively load the endothelium. For the isolated endothelial cell studies, cells were allowed to settle on a glass coverslip before incubation in the perfusion chamber with fura-2 AM (0.35 μM) plus Pluronic F-127 (0.01%) for 12–15 min. Cells were then washed with fresh PSS for 15 min to remove excess fura-2 AM and to allow complete deesterification of the dye. The fura-2 emission (520/40 nm) was measured from three excitation wavelengths (340/15, 360/15, and 380/15 nm) at a frequency of 4 Hz. The 360-nm excitation wavelength reflects the Ca2+-insensitive isosbestic point for fura-2. The autofluorescence/background was subtracted before the ratios were calculated. In the isolated cells, we calculated the ratio of 360/380 emissions instead of the 340/380 ratio as in intact arteries because of the lower intensity of the 340-nm emission in the cell preparation. Fura-2 loading and experiments with isolated cells took place at room temperature in PSS of the following composition (in mM): 137 NaCl, 5.6 KCl, 1.6 CaCl2, 1 MgCl2, 10 glucose, and 10 HEPES, with pH adjusted to 7.4 with NaOH.
Manganese quenching measurements.
Manganese (Mn2+) quenching is a technique that is used to evaluate Ca2+ influx into fura-2-loaded cells (9, 23). Mn2+ acts as a Ca2+ surrogate or tracer and is used in the absence or presence of Ca2+. Mn2+ differs from Ca2+ in that its binding to fura-2 results in a quenching of fura-2 fluorescence. Since the intracellular concentration of Mn2+ is negligible, the decay of fura-2 fluorescence in the presence of added extracellular Mn2+ reflects Mn2+ (and Ca2+) entry. The initial rate of quench was used as an index of Ca2+ entry.
We administered MnCl2 either luminally or abluminally to evaluate endothelial Ca2+ influx across the apical and basolateral membrane surface, respectively. The two membrane surfaces are separated by tight junctions that form a barrier between luminal and abluminal compartments (16). Fura-2 loading was restricted to the endothelium (14); therefore, the measured fluorescence quenching is reflective of endothelial Mn2+ influx. For luminal Mn2+ delivery, 300 μM MnCl2 was added to the luminal perfusate and delivered simultaneously with UTP. For abluminal Mn2+ delivery, 150 μM MnCl2 was administered to the abluminal perfusate 2 min before UTP delivery. This time lag was used to allow the Mn2+ sufficient time to reach the basolateral surface of the endothelium before UTP reached the luminal surface. The lower concentration of MnCl2 was used so that quenching due to basal Mn2+/Ca2+ influx would not dominate the measurement.
Cell digestion protocol.
Rat MCA (left and right) were removed from the brain, cleaned of connective tissue, and cut into small rings. Our digestion protocol was modified from a protocol initially established by Jackson et al. (10). The artery segments were incubated for 10 min in a wash buffer solution containing (in mM) 137 NaCl, 5.6 KCl, 0.1 CaCl2, 1 MgCl2, 10 glucose, 10 HEPES, 0.005 verapamil, and 0.1% BSA, with pH adjusted to 7.35 with NaOH. The wash solution was carefully removed and replaced with digestion buffer containing papain (26 U/ml) plus dithioerythritol (1 mg/ml) and incubated for 35 min at 37°C. At the end of the incubation period, the papain solution was carefully removed and the vessel segments were washed with the wash buffer solution. The solution was then replaced with a second digestion solution containing collagenase (Sigma blend; 1.4 mg/ml), elastase (43 U/ml), and soybean trypsin inhibitor (0.9 mg/ml) and incubated for 8 min at 37°C. The artery segments were then washed twice with wash buffer and triturated with a P-200 pipette tip to disperse smooth muscle cells and small endothelial cell clusters. Endothelial cells were found in multi-cell planar sheets among individual smooth muscle cells. The digestion buffer consisted of (in mM) 135 NaCl, 5 KCl, 1.5 MgCl2, 0.42 Na2HPO4, 0.44 NaH2PO4, 4.17 NaHCO3, and 10 HEPES, with pH adjusted to 7.25 with NaOH. The papain, dithioerythritol, and collagenase were purchased from Sigma; elastase was purchased from Calbiochem; and Soybean trypsin inhibitor was purchased from Worthington.
RT-PCR with freshly isolated endothelial and smooth muscle cells was based on the protocol described by Kohler et al. (11) In brief, endothelial or smooth muscle cells were harvested into a new glass pipette (∼50-μm tip opening) that had the tip prefilled with ∼1 μl of RNAlater solution. Smooth muscle cells were identified on the basis of their classic cigar shape morphology. Endothelial cells were harvested as multi-cell planar sheets containing 5–50 cells. Harvested cells were expelled into a PCR tube and frozen in liquid N2. The following components of the first-strand synthesis kit (Invitrogen) were then added to each tube: 5× first-strand buffer (2 μl), 10 mM dNTP (0.5 μl), 50 μM random hexamer (1 μl), 0.1 mM DTT (1 μl), 10 U/μl RNase inhibitor (1 μl), and diethyl pyrocarbonate H2O (1.5 μl). Two rapid freeze-thaw cycles were performed to rupture cells and gain access to the RNA before the addition of 1 μl of Superscript II reverse transcriptase. Samples were incubated at 37°C for 1 h and then stored at −20°C for later use.
Our initial assignment of cell type was based on morphological characteristics. This cell assignment was confirmed by a PCR screening of cell-specific products. The final assignment of endothelial cells required that the cell preparation demonstrate PCR amplicons for endothelial NOS (eNOS) but not for SM22α (and vice versa for smooth muscle cells). Only cell preparations that demonstrated no contamination from the other cell type were used.
PCR was performed for 35–50 cycles using the following temperature protocol: 94°C (30 s), 53–55°C (60 s), and 72°C (30 s), with an initial step at 94°C (2 min) to activate the platinum Taq polymerase (Invitrogen). Gene-specific primers were as follows: ATGGATGAGCCAACTCAAGG (forward) and CCAGCTCTGTCCTCAGAAGG (reverse) for eNOS; GGCAGCTGAGGATTATGGAG (forward) and GCTGGCCTTCCCTTTCTAAC (reverse) for SM22α; TGACTACCGGTGGTGTTTCA (forward) and TGATCCCTGCATAGTGTCCA (reverse) for TRPV1; and ATCAACTCGCCCTTCAGAGA (forward) and GGTGTTCTCTCGGGTGTTGT (reverse) for TRPV4. The predicted size of amplicons were 356, 325, 330, and 339 bp for eNOS, SM22α, TRPV1, and TRPV4, respectively. Primers for TRPV1 and TRPV4 were a gift from Drs. Roger O'Neil and Rachel Brown at the University of Texas Health Science Center at Houston.
A section of brain containing the MCA was removed and quickly frozen in embedding compound. The MCA/brain was sectioned in 10-μm slices and placed on glass slides for immunohistochemistry. In this series of experiments, we evaluated immunoreactivity to an anti-TRPV4 antibody (a gift from Dr. Stefan Heller, Stanford University School of Medicine). Control experiments were performed using either 1) preimmune serum or 2) block solution in place of TRPV4 antibody.
Frozen sections were fixed with 4% paraformaldehyde for 30 min and then washed with PBS (3 × 5 min). Sections were then permeabilized with Triton X-100 (0.2%; 15 min) and blocked with PBS containing 5% goat serum (60 min). Incubation with anti-TRPV4 antibody (or control) was performed overnight at 4°C at 1:300 dilution. The antibody and preimmune solutions were prepared in PBS containing 1% goat serum. Sections were next washed with PBS (3 × 5 min) and incubated with a Cy2-conjugated donkey anti-rabbit IgG secondary (Jackson ImmunoResearch, West Grove, PA). The secondary detection antibody was incubated covered at room temperature (60 min) at 3.8 μg/ml final concentration. The sections were washed with PBS (5 × 10 min) and stained with 4,6-diamidino-2-phenylindole (1:10,000 dilution) before coverslips were applied. Imaging was performed using a Nikon TE-2000 fluorescence microscope with deconvolution capabilities. Sections were acquired at 0.17-μm intervals using a ×60 oil objective (NA 1.40).
Comparisons between concentration-response curves between two groups were performed using two-way repeated-measures ANOVA followed by either a Holm-Sidak or Student-Newman-Keuls test to compare individual points. Unpaired comparisons between two groups were evaluated using a t-test (SigmaStat). Significance was defined as P < 0.05.
Initial experiments were performed with pressurized MCA to determine whether the endothelial [Ca2+]i in response to luminal UTP was altered following PLA2 inhibition. Note that inhibitors of NOS and COX (l-NAME and indomethacin, 10 μM each) were present throughout to focus on the EDHF-mediated component of dilation. Some experiments were performed in the additional presence of PACOCF3, an inhibitor of PLA2. Mean baseline diameter in the presence or absence of PACOCF3 was 192 ± 8 and 183 ± 10 μm, respectively. Mean maximum diameter for the two groups was 291 ± 6 and 281 ± 4 μm, respectively. There was no significant difference between control and PACOCF3 groups for either baseline or maximum diameter (P > 0.05; t-test).
Figure 1 shows representative diameter and endothelial [Ca2+]i measurements in pressurized MCA in response to luminally delivered UTP (10−7–10−5 M). Experiments were performed with vehicle control (1:1,000 EtOH) or with 20 μM PACOCF3, an inhibitor of PLA2. Note that whereas 10−5 M UTP produced sustained maximal dilation in the control arteries, it produced only submaximal oscillations in the PACOCF3-treated arteries. The endothelial [Ca2+]i responses to UTP were likewise attenuated following PLA2 inhibition. The steady-state mean values are summarized in Fig. 2.
In the control group, luminal UTP produced a dose-dependent increase in endothelial [Ca2+]i. Below 10−5 M UTP, no dilation of the MCA resulted, whereas maximal dilation was produced at 10−5 M UTP. This finding is consistent with earlier studies from our laboratory in which a critical [Ca2+]i threshold was found for EDHF-mediated dilations (15). Inhibition of PLA2 with PACOCF3 resulted in significant reduction in both dilation and endothelial [Ca2+]i responses (2-way repeated-measures ANOVA).
It is apparent from the results shown in Figs. 1 and 2 that global endothelial [Ca2+]i responses were reduced following PLA2 inhibition. We next sought to determine whether this reduced [Ca2+]i response was due to reduced Ca2+ influx. We used the Mn2+ quenching technique to evaluate Ca2+ influx from both luminal and abluminal compartments. In this technique, we supplemented the perfusing solution with 150–300 μM MnCl2. Mn2+ acts as a Ca2+ surrogate and enters the cells along the same pathways as Ca2+. However, when Mn2+ binds to fura-2 in the endothelial cell, it causes the quenching of fura-2 fluorescence. Therefore, decay of fura-2 fluorescence measured at the isosbestic wavelength (360 nm) indicates Mn2+ (and thus Ca2+) entry.
In our initial experiments, we administered MnCl2 (300 μM) luminally with UTP (10 μM), based on the assumption that significant Ca2+ influx occurred across the luminal/apical membrane of the endothelium. Surprisingly, the Mn2+ quenching was similar between control and PACOCF3 groups. The quench rate was −21 ± 1 and −24 ± 2%/min, respectively (P > 0.05, n = 5 each). These results suggest that UTP-stimulated Ca2+ influx across the luminal membrane is not altered by PLA2 inhibition.
We next applied MnCl2 (150 μM) to the abluminal compartment to evaluate Ca2+ influx across the basolateral membrane of the endothelium. The addition of MnCl2 alone produced some quenching (Fig. 3A), similar to when we administered MnCl2 alone luminally (n = 3, not shown). The slope of quenching with MnCl2 alone (reflected by the region between MnCl2 addition and UTP addition) was −20.9 ± 2.6 and −19.3 ± 4.7%/min for control and PACOCF3 groups, respectively. The basal Ca2+ influx, reflected by the addition of MnCl2 alone, did not differ between groups (P = 0.78, t-test). The subsequent luminal delivery of UTP produced significantly less quenching in the PACOCF3 group compared with the control group (Fig. 3, A and B). Curves were significantly different with individual differences at 450 and 550 s (2-way repeated-measures ANOVA with Holm-Sidak method for individual comparisons). The initial slope of quenching was −28.1 ± 2.3 and −14.2 ± 1.6%/min for control and PACOCF3 groups, respectively (P < 0.01; t-test). Therefore, it appeared that Ca2+ influx across the basolateral membrane was reduced following PLA2 inhibition.
If PLA2 was indeed involved in endothelial Ca2+ regulation, we would expect that application of compounds normally generated by PLA2 would result in an increase in endothelial [Ca2+]i and promote dilation of the artery. Arachidonic acid (AA) is the primary compound released by cytosolic PLA2. In Fig. 4A, we demonstrate a concentration-dependent increase in endothelial [Ca2+]i by luminally delivered AA. PACOCF3 (20 μM) was present to inhibit endogenous production of AA by PLA2. AA also produced dose-dependent dilation of the artery (Fig. 4B).
TRPV1 and TRPV4 are nonselective cation channels that have been demonstrated to possess significant Ca2+ conductance and are activated by AA or its metabolites (28, 29). We therefore evaluated the expression of these candidate AA-sensitive Ca2+ channels by PCR in whole MCA and freshly isolated endothelial cells (Fig. 5A). Whole MCA demonstrated a clear PCR product for TRPV4 (339 bp) but not TRPV1 (330 bp). Endothelial cells similarly demonstrated PCR product for TRPV4 but not TRPV1. The identity of the endothelial cells was based on morphological characteristics at the time of harvesting as well as expression of “cell-specific” message by PCR (Fig. 5B). Endothelial cell preparations demonstrated PCR product for eNOS (an endothelial marker) but not SM22α (a smooth muscle marker). SM22α was well expressed in whole MCA and isolated smooth muscle cells.
Tissue-specific expression of TRPV4 channel protein was evaluated using immunofluorescence in frozen cross sections of MCA. Deconvolution microscopy was used to obtain sufficient resolution of the very thin endothelial layer. Figure 6A demonstrates TRPV4 channel immunoreactivity in smooth muscle as well as endothelial cells. The TRPV4 channel expression appeared to be considerably more pronounced in the endothelium compared with the smooth muscle. The endothelial cell staining was determined by its location luminal to the internal elastic lamina. Note that endothelial cell nuclei also are present in this region. Figure 6B demonstrates the relatively low level of nonspecific staining in the preimmune serum control. A second control in which the primary antibody was omitted resulted in virtually zero positive staining (not shown).
To further evaluate the role of TRPV channels in UTP-dependent Ca2+ regulation, we used a nonselective TRPV channel inhibitor, ruthenium red (RuR). RuR has been shown to block all known TRPV channels with varying efficacy (2). Administration of RuR (3 μM) resulted in a partial attenuation of the endothelial Ca2+ response to UTP (Fig. 7). The attenuation was statistically significant (2-way repeated-measures ANOVA), although not as complete as with PLA2 inhibition. RuR did not significantly attenuate EDHF-mediated dilations (not shown).
Administration of a selective TRPV4 channel activator, 4α-phorbol 12,13-didecanoate (4α-PDD; 1–3 μM), produced a significant increase in [Ca2+]i within freshly isolated MCA endothelial cells (Fig. 8). The elevated [Ca2+]i was rapidly reduced following the application of RuR (1 μM). The virtually instantaneous response to RuR suggests that the effect was through reduced Ca2+ influx versus an effect on ryanodine receptors (also inhibited by RuR) that would require cellular uptake of the blocker.
In this study, we examined the role of PLA2 in endothelial Ca2+ regulation and EDHF-mediated dilation. We present the following significant findings regarding endothelial Ca2+ regulation in rat MCA: 1) endothelial Ca2+ regulation is PLA2 sensitive, and 2) TRPV4 (but not TRPV1) channels are expressed in cerebral endothelial cells and may contribute to the PLA2-sensitive Ca2+ regulation. Furthermore, we provide evidence suggesting that the PLA2-sensitive Ca2+ influx is polarized to the basolateral membrane.
Endothelium-derived hyperpolarizing factor.
Early studies speculated that EDHF was a factor that originated in the endothelium and then diffused to the smooth muscle. Although there is evidence for a transferable factor in certain coronary and renal beds, other vascular beds appear to lack evidence for transferable factors (1). In rat MCA, it appears that hyperpolarization originates in the endothelium and is then conducted to the surrounding smooth muscle (25). Endothelial cell hyperpolarization results from elevated [Ca2+]i that must reach a critical threshold before activating small- and/or intermediate-conductance KCa channels (15, 17). The transfer of hyperpolarization between endothelium and smooth muscle then likely occurs via intercellular gap junctions between the two tissue types (25). Once the innermost layer of smooth muscle is hyperpolarized, the hyperpolarization can be amplified and propagated to outer layers via the activation of smooth muscle inwardly rectifying K+ channels (4).
Endothelial Ca2+ regulation is PLA2 sensitive.
Activation of PLA2 liberates AA from membrane phospholipids. The resulting AA can then act directly on target proteins or be metabolized to one of several other bioactive metabolites. The evidence for a direct effect of AA on Ca2+ influx in endothelial cells comes from the finding that both AA and nonmetabolized AA metabolites promote Ca2+ influx in intact endothelial preparations (7, 19). In addition, application of AA and analogs to the cytosolic face of inside-out patches stimulates opening of nonselective cation channels. In this latter preparation, the metabolizing enzymes should have been washed away. Evidence also has been presented supporting a role for AA metabolites of the cytochrome P-450 monooxygenases: epoxyeicosatrienoic acids (EETs). Several EET products have been demonstrated to trigger Ca2+ influx in cultured endothelial cells, whereas the inhibition of cytochrome P-450 enzymes results in diminished Ca2+ influx (8, 27, 30).
A previous study from our laboratory demonstrated that activation of endothelial PLA2 was critical for EDHF-mediated dilations (33). Given our current understanding of the EDHF-dependent mechanism in rat cerebral arteries, it seemed unlikely that PLA2 would liberate an endothelial product that must then diffuse to the smooth muscle before exerting a hyperpolarizing effect. Instead, we postulated that PLA2 acts as a critical modulator of some step in the EDHF-dependent mechanism. Given the above-mentioned findings regarding AA and AA metabolites on Ca2+ influx, we evaluated a role for PLA2/AA on MCA endothelial Ca2+ regulation.
As shown in Figs. 1 and 2, inhibition of PLA2 resulted in a profound reduction in endothelial [Ca2+]i responses to UTP. This endothelial [Ca2+]i reflects the sum of Ca2+ release from internal stores and Ca2+ influx from the extracellular milieu. Since Ca2+ influx accounts for the sustained [Ca2+]i response as well as our proposed pathway by which PLA2/AA modulates endothelial [Ca2+]i, we examined Ca2+ influx using the Mn2+ quenching technique. Interestingly, Mn2+/Ca2+ influx across the luminal/apical membrane was unaltered by PLA2 inhibition. In these experiments, we added MnCl2 to the luminal perfusate only. In pressurized cerebral arteries, the endothelial cells are tightly coupled and do not readily allow passage of hydrophilic compounds (16). In contrast to influx across the apical membrane, Mn2+/Ca2+ influx was reduced across the basolateral membrane following PLA2 inhibition. In this later experiment, MnCl2 was delivered to the basolateral surface of the endothelium by being added to the abluminal perfusate. Together, these data suggest that the PLA2-dependent Ca2+ influx occurs primarily across the basolateral membrane of MCA endothelial cells in response to UTP.
Does endothelial Ca2+ influx demonstrate apical-basolateral polarity?
It is well established that endothelial cells in the brain, like many other epithelial cells, exhibit apical/basolateral polarity of expression and function (3, 13, 20, 21, 24). This means that the apical membranes (blood facing) can be quite different from the basolateral membranes (smooth muscle facing) in both protein composition and function. Our study adds to the already extensive list of polarized proteins and functions of cerebral endothelium. In particular, our studies suggest that the extracellular milieu on the adventitial side provides the source of Ca2+ for PLA2-sensitive Ca2+ influx. This is not to say that all Ca2+ influx in response to UTP stimulation occurs across the abluminal membrane. UTP did indeed stimulate Ca2+ influx across the luminal/apical membrane above the basal levels, but not by a PLA2-sensitive mechanism. Thus it appears that more than one Ca2+ influx mechanism exists to promote Ca2+ entry into MCA endothelium, at least one of which is PLA2 dependent. The identity of the channels involved in mediating Ca2+ influx in this tissue remains to be resolved and is discussed further below.
TRPV4 channels are expressed and functional in the endothelium.
From our study, we cannot conclusively state which Ca2+ influx channels are involved in the PLA2-sensitive and -insensitive Ca2+ entry mechanism. However, we do provide evidence indicating a likely involvement of TRPV4 channels. TRPV1 and TRPV4 are logical candidates for Ca2+ influx channels on the basis of a demonstrated presence in other endothelial preparations (6, 12, 30, 31), moderate Ca2+ conductance (2), and known activation by either AA or AA metabolites (2, 28, 30). We found no evidence for TRPV1 channel mRNA in either whole MCA or isolated endothelial cells (Fig. 5). In addition, application of 10 μM capsaicin (a TRPV1 activator) failed to promote an increase in endothelial [Ca2+]i (data not shown). However, using PCR, we did find evidence for TRPV4 message in both whole artery and isolated endothelial cells. Furthermore, TRPV4 protein was identified in both smooth muscle and endothelium by immunohistochemistry (Fig. 6). The presence of TRPV4 mRNA in cerebral smooth muscle cells was recently described (5). Thus our demonstration of TRPV4 protein in MCA smooth muscle was an expected finding. However, we believe the present study to be the first to report TRPV4 mRNA and protein in MCA endothelium. We also found functional evidence for TRPV4 channels in MCA endothelial cells. Application of a selective TRPV4 channel activator, 4α-PDD, promoted a significant increase in [Ca2+]i from freshly isolated MCA endothelial cells. The [Ca2+]i response was rapidly reduced by subsequent delivery of a TRPV channel blocker, RuR, suggesting that the response to 4α-PDD occurred by stimulating Ca2+ influx. In addition, luminal application of AA produced an increase in endothelial [Ca2+]i (Fig. 4A). These latter data demonstrate that AA or an AA metabolite stimulates increased [Ca2+]i in MCA endothelium. The metabolism of AA could conceivably occur via any of the pathways other than COX, which was inhibited in the present studies. If metabolism of AA is required, the most likely pathway would be via cytochrome P-450 epoxygenases to generate EETs. It should be noted that in our laboratory, we have not found a role for the EETs in UTP-mediated dilation (32). Rather, it appears that AA itself acts as the signaling molecule in the rat MCA endothelial cells.
An interesting explanation of our findings is that TRPV4 channels are expressed preferentially on the basolateral membrane. Although we are not aware of any studies that have reported polarity to TRP channel expression, TRP channels have been shown to be rapidly shuttled to the plasma membrane in response to various stimuli (22). Additional studies are required to evaluate the possibility that TRPV4 channel expression is restricted to the basolateral membrane in rat MCA.
In summary, we have demonstrated that endothelial [Ca2+]i and diameter responses to UTP are significantly reduced following PLA2 inhibition. This reduced [Ca2+]i response results from decreased Ca2+ influx across the basolateral membrane but not across the apical membrane. Furthermore, we demonstrate that TRPV4 channels are present and appear to contribute to endothelial Ca2+ influx and EDHF-mediated dilations in rat MCA.
This work was supported by American Heart Association Grants 0230353N (to S. P. Marrelli), 0665100Y (to S. P. Marrelli), and 0270110N (to R. M. Bryan) and National Institute of Neurological Disorders and Stroke Grant R01 NS-46666 (to R. M. Bryan).
We thank Dr. Stefan Heller (Stanford University School of Medicine) for generously providing the TRPV4 channel antibody and preimmune control serum.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by the American Physiological Society