The mechanisms by which flow-imposed shear stress elevates intracellular Ca2+ in cultured endothelial cells (ECs) are not fully understood. Here we report finding that endogenously released ATP contributes to shear stress-induced Ca2+ responses. Application of flow of Hanks' balanced solution to human pulmonary artery ECs (HPAECs) elicited shear stress-dependent increases in Ca2+ concentrations. Chelation of extracellular Ca2+ with EGTA completely abolished the Ca2+ responses, whereas the phospholipase C inhibitor U-73122 or the Ca2+-ATPase inhibitor thapsigargin had no effect, which thereby indicates that the response was due to the influx of extracellular Ca2+. The Ca2+ influx was significantly suppressed by apyrase, which degrades ATP, or antisense oligonucleotide targeted to P2X4 purinoceptors. A luciferase luminometric assay showed that shear stress induced dose-dependent release of ATP. When the ATP release was inhibited by the ATP synthase inhibitors angiostatin or oligomycin, the Ca2+ influx was markedly suppressed but was restored by removal of these inhibitors or addition of extracellular ATP. These results suggest that shear stress stimulates HPAECs to release ATP, which activates Ca2+ influx via P2X4 receptors.
- hemodynamic force
- ATP synthase
- and P2X4 receptor
endothelial cells (ECs) cover the inner surfaces of blood vessels and are in direct contact with flowing blood; therefore, they are constantly exposed to fluid shear stress. A number of recent studies reveal that ECs recognize changes in shear stress and transmit signals to the interior of the cell, which leads to cellular responses that involve changes in cell morphology, cell function, and gene expression (2, 8). These EC responses to shear stress are thought to play an important role in blood flow-dependent phenomena such as angiogenesis, vascular remodeling, and atherogenesis.
Shear stress has been demonstrated to activate a variety of molecules known to function in signal transduction including ion channels, G proteins, inositol trisphosphate, and many protein kinases (8), which suggests that multiple pathways are involved in shear stress signal transduction. However, it remains unclear which pathways are primary and which are secondary, and the initial sensor that recognizes shear stress has yet to be identified. Shear stress may activate several pathways simultaneously.
Ca2+ signaling has been demonstrated to be involved in shear stress signal transduction. ECs show an increase in intracellular Ca2+ concentration ([Ca2+]i) in response to shear stress (1, 4), and, at a certain extracellular ATP concentration, shear stress induces a dose-dependent influx of extracellular Ca2+ across the cell membrane (3). A subtype of ATP-gated cation channel, the P2X4 purinoceptor, has even been shown to mediate the shear stress-dependent Ca2+ influx (26, 27). Nevertheless, reports show that Ca2+ responses to shear stress occur even in the absence of extracellular ATP (11, 23), and the role of ATP in shear stress-induced Ca2+ responses is a matter of controversy (9, 18).
ECs are known to synthesize, store, and release a variety of vasoactive substances in response to shear stress. ATP has been shown to be rapidly released by cultured ECs and vascular beds in response to increases in shear stress (6, 13, 17). Thus endogenously released ATP may modify the flow-induced Ca2+ responses, but this issue has not yet been fully investigated. Here we report for the first time that endogenously released ATP mediates shear stress-induced Ca2+ responses in human pulmonary artery ECs (HPAECs).
MATERIALS AND METHODS
Cell culture. HPAECs were obtained from Clonetics (San Diego, CA) and grown on a 1% gelatin-coated tissue culture flask in M199 supplemented with 15% FBS, 2 mmol/l l-glutamine (GIBCO-BRL), 50 μg/ml heparin, and 30 μg/ml EC growth factor (Becton Dickinson). The cells used in the present experiments were from passages 7 and 10.
Flow stimulation and Ca2+ measurement. HPAECs were seeded at a density of 1.5 × 104 cells/cm2 on a coverslip with a surface area of 3.7 cm2 and were cultured for 3 days to form a confluent monolayer that reached a density of ∼4.5 × 104 cells/cm2. The cells were loaded with 5 μmol/l indo 1-acetoxymethyl ester (indo 1-AM; Dojindo), the coverslip was placed in a parallel plate-type flow chamber (FSC2, Bioptechs), and the flow chamber was set on the stage of an inverted microscope (Diaphot 300, Nikon). One end of the chamber was connected to a reservoir filled with Hanks' balanced salt solution (HBSS) via a silicon tube, and the cells were exposed to a stepwise increase in flow rate at 37°C by changing the height of the reservoir and monitoring the flow rate at the outflow tube with a flow sensor (Transonic Systems). The intensity of shear rate (γ, measured in s–1) and shear stress (τ, in dyn/cm2) to the cells were calculated as follows: γ = 6Q̇/ab2 and τ = μγ, where Q̇ is volume flow (in ml/s), a and b are cross-sectional dimensions (in cm) of the flow path, and μ is the viscosity of the perfusate (in Poise). Changes in [Ca2+]i were measured with a confocal laser scanning system (MRC-1000 UV, Bio-Rad) equipped with an ultraviolet argon ion laser as previously described (14). Briefly, light from the laser at a 351-nm wavelength was used to excite the cells through a ×20 objective, and the emitted light was divided into 480- and 405-nm wavelengths (by a beam splitter) and counted with photomultipliers. The time course of the 405-to-480-nm fluorescence ratio (F405/F480) in cells of interest was monitored with the accessory time-course software of the Bio-Rad 1000 UV system.
ATP measurement. ATP released from HPAECs was measured by a luciferin-luciferase assay. The cells were exposed to a stepwise increase in flow rate under the same conditions of Ca2+ measurement at 37°C, and, after the perfusate was collected every 20 s, 100-μl aliquots were applied to the ATP assay system (Toyo Ink; Tokyo, Japan). The ATP assay mixture (luciferase, d-luciferin, and BSA) was injected into each sample, and its relative light intensity was recorded for 10 s in a Lumat LB 9501 luminometer (Berthold; Wildbad, Germany) at room temperature. A calibration curve for ATP concentrations was obtained for each experiment by using the same batch of luciferin-luciferase reagents, and the total moles of ATP released per second per 106 cells were calculated.
Western blot analysis. Western blot analysis was performed as previously described (27). Briefly, cells were dissolved in RIPA buffer (1% Nonidet P-40, 20 mM Tris · HCl, 0.15 M NaCl, 0.5% sodium deoxycholate, 2 mM EDTA, 2 mM EGTA, 0.1% SDS, 0.2 mM Na2MoO4, 10 mM NaF, 1 mM Na3VO4, 1 mM phenylmethylsulfonyl fluoride, 5 μg/ml leupeptin, 5 μg/ml antipain, 5 μg/ml pepstatin A, and 0.2 U/ml aprotinin; pH 7.4) and centrifuged at 26,000 g for 30 min. A 30-μg sample of total cell lysate was resolved in SDS sample buffer (0.2 M Tris · HCl, 18% glycerol, 4% SDS, 0.01% bromophenol blue, and 10% β-mercaptoethanol; pH 8.8) for SDS-PAGE and transferred to an Immobilon membrane (Millipore), where it was incubated for 1 h. The membrane was probed with the anti-P2X4 antiserum (3 μg/ml) and then incubated with anti-rabbit IgG horseradish peroxidase-conjugated antibody. The same membrane was reprobed with monoclonal anti-β-actin antibody (Abcan). The densitometry values of the P2X4 blots were standardized to those of the β-actin blots. An antiserum against human P2X4 receptor protein was generated in rabbits injected with a synthetic peptide (NH2-RLYYREKKYKYVEDYC-COOH) comprising amino acid residues 364–378 of the sequences for the intracellular COOH-terminal domains of the cloned human P2X4 receptor. The peptide was covalently linked to keyhole limpet hemocyanin, and rabbits were immunized by injection with the conjugated peptide every 2 wk for 8 wk. The anti-P2X4 receptor antiserum was then affinity purified with synthetic peptide (P2X4 364–378) immobilized on Sepharose 4B (Asahi Techno Glass).
Competitive PCR. Both sense and antisense primers were synthesized according to the gene-specific sequence (Table 1), and the RNA competitor fragments were obtained as previously described (27). With the use of these primers, DNA competitor fragments were generated and then transcribed into RNA fragments using SP6 RNA polymerase. Competitive PCR was performed by adding 2 μl of the mRNA samples obtained from HPAECs to 2 μl of different 10-fold dilutions of the RNA competitor fragments that ranged from 0.00008 to 800 attomol/μl. After reverse transcription, the PCR reactions were carried out using each target gene-specific primer in a solution that contained ExTaq DNA polymerase (Takara) and [α-32P]dCTP. A second series of competitive PCR assays was then carried out with consecutive 1:2 dilutions of competitors mixed with a constant amount of each target mRNA. The PCR products were separated by electrophoresis in a 5% polyacrylamide gel. The radioactivity of both target mRNA bands and competitor bands (known concentrations) was measured with a GS363 Molecular Imager System. The logarithm of the ratio of target to competitor bands was plotted as a function of the logarithm of the known amount of competitor. The concentration of target mRNA molecules present in cells corresponds to that of competitor at the competition equivalence point [log (target/competitor) = 0]. The β-actin gene, which is a housekeeping gene, was used as a control for variation in the quality and quantity of RNA.
Real-time PCR analysis. Total RNA samples were prepared from HPAECs with Isogen (Nippon Gene; Tokyo). First-strand cDNAs were generated using Moloney murine leukemia virus reverse transcriptase (MMLV-RT; GIBCO-BRL) and RNA primed with oligo(dT) primer. After reverse transcription of RNA into cDNA, real-time PCR was used to monitor gene expression with a Smart Cycler (Cepheid) according to the standard procedure. PCR was performed with a TaKaRa ExTaq R-PCR version (Takara), SYBR green I (BioWhittaker), and the primer pairs (Table 1). The temperature profile included initial denaturation for 30 s at 95°C, followed by 35 cycles of denaturation at 95°C for 15 s, annealing at 60°C for 15 s, elongation at 72°C, and fluorescence monitoring at 85°C. The specificity of the amplification reaction was determined by performing a melting-curve analysis. Standard curves for expression of each gene were generated by serial dilutions of known quantities of the respective cDNA gene template. Relative quantification of the signals was achieved by normalizing the signals of the different genes to GAPDH.
Antisense oligonucleotides. Antisense oligonucleotides (AS-oligos) targeted to the P2X4 receptor and control scrambled oligos (S-oligos) were designed and synthesized by Biognostik (Göttingen, Germany; Ref. 26). The sequences of the phosphorothiorate AS- and S-oligos were 5′-CCTGAAATTGTAGCC-3′ and 5′-TAATCGCTTCAGACG-3′, respectively, and they were FITC labeled at the 5′-end. The AS- or S-oligos were transfected into cells with LipofectAMINE PLUS (GIBCO-BRL).
Statistical analysis. All results are expressed as means ± SD. Statistical significance was evaluated by an ANOVA and a Bonferonni's adjustment applied to the results of a t-test performed with SPSS software. P values <0.01 were regarded as statistically significant.
Flow induces a shear stress-dependent Ca2+ influx into HPAECs. HPAECs were exposed to flow of HBSS, and changes in [Ca2+]i were monitored. [Ca2+]i increased stepwise in tandem with the increases in shear stress, and, when flow was terminated, [Ca2+]i returned to its basal level (Fig. 1A). Chelation of extracellular Ca2+ with EGTA completely abolished the flow-induced Ca2+ responses, whereas the phospholipase C inhibitor U-73122 or the Ca2+-ATPase inhibitor thapsigargin that induces Ca2+-store depletion had no effect (Fig. 1, B–D). When the [Ca2+]i F405/F480 was plotted against shear stress (in dyn/cm2), a linear relationship between the two was observed in the control, but this relationship was absent in the presence of EGTA and was unaffected by U-73122 or thapsigargin. These results indicate that the shear stress-dependent increase in [Ca2+]i was due to the influx of extracellular Ca2+ across the cell membrane.
Shear stress-induced Ca2+ influx is ATP dependent. HPAECs were exposed to apyrase (Enzyme Commission No. 188.8.131.52; Sigma grade III), which degrades ATP and ADP, and their Ca2+ responses to flow were examined. Apyrase markedly inhibited the shear stress-dependent Ca2+ influx in a dose-dependent manner (Fig. 2), which indicates that ATP and ADP are involved in the flow-induced Ca2+ influx in HPAECs. Because no extracellular ATP is present in HBSS, endogenous ATP and ADP appeared to play a role.
HPAECs predominantly express ATP-operated cation channel P2X4. Using a competitive PCR method, we compared the mRNA expression of the P2X subtypes P2X1, P2X2, P2X3, P2X4, P2X5, P2X6, and P2X7. HPAECs had markedly elevated expression of P2X4, whereas expression of P2X1, P2X2, P2X3, P2X5, P2X6, and P2X7 was very low or almost undetectable (Fig. 3). ATP-operated G protein-coupled receptor P2Y1 and P2Y2 expression was also very low compared to P2X4 expression.
P2X4 purinoceptors mediate shear stress-induced Ca2+ influx in HPAECs. AS-oligos targeted to the P2X4 receptor were used to examine whether P2X4 receptors mediate the shear stress-induced Ca2+ influx. The AS-oligos markedly decreased the P2X4 mRNA and protein levels in the HPAECs, whereas control S-oligos had no effect (Fig. 4A). The shear stress-dependent Ca2+ influx was seen in HPAECs exposed to control S-oligos but not in those exposed to AS-oligos (Fig. 4B). These findings suggest that P2X4 receptors play a central role in the shear stress-dependent Ca2+ influx in HPAECs, although the possibility of some involvement by other purinergic receptors besides P2X4 cannot be excluded.
Shear stress stimulates HPAECs to release ATP. To examine whether HPAECs release ATP, the amount of ATP in perfusate was determined by a luciferase luminometric assay. The cells released a very small amount of ATP basally, but when exposed to flow, the amount of ATP released increased in a stepwise manner in response to the stepwise increments in shear stress (Fig. 5).
Flow-induced Ca2+ influx and ATP release are shear stress rather than shear rate dependent. To examine whether the flow-induced Ca2+ influx and ATP release are dependent on shear stress or shear rate, flow-loading experiments were conducted with two perfusates of different viscosities. Stepwise elevation of the shear rate induced increases in [Ca2+]i but to a greater extent with the high-viscosity perfusate (HBSS + 5% dextran; Fig. 6A). This tendency was quantitatively confirmed by plotting the percent increases in [Ca2+]i against shear rate: the different viscosities produced two distinct curves (Fig. 6B, left). By contrast, when plotted against shear stress, they almost formed a single curve (Fig. 6B, right), which indicates that the percent increases in [Ca2+]i are well correlated with shear stress regardless of viscosity. A similar tendency was observed in flow-induced ATP release (Fig. 6C). These findings indicate that the flow-induced Ca2+ influx and ATP release are dependent on shear stress rather than on shear rate.
Endogenously released ATP mediates flow-induced Ca2+ influx. To investigate whether endogenously released ATP is involved in flow-induced Ca2+ influx, ATP release was suppressed with angiostatin or oligomycin, which are inhibitors of ATP synthase (19, 20), and changes in the Ca2+ responses were examined. The ATP release was reduced by angiostatin and oligomycin in a dose-dependent manner (Fig. 7), and the flow-induced Ca2+ responses were also strongly suppressed by angiostatin and oligomycin (Fig. 8). When these inhibitors were removed by flushing, the ATP release and the Ca2+ responses were recovered (Fig. 8). Addition of extracellular ATP also resulted in recovery of the Ca2+ responses in HPAECs treated with angiostatin or oligomycin (Fig. 9). These findings suggest that endogenously released ATP actually mediates the shear stress-induced Ca2+ influx via P2X4 receptors in HPAECs.
The present study demonstrates that a shear stress-dependent increase in [Ca2+]i occurs in HPAECs in response to flow of HBSS and that the increase in [Ca2+]i is due to the influx of extracellular Ca2+ via P2X4 receptors. HPAECs also release ATP in a shear stress-dependent manner in response to flow. When the release of ATP was suppressed with ATP synthase inhibitors, the flow-induced Ca2+ influx became much weaker but was restored by washing away the inhibitors or adding extracellular ATP. Thus it appears that the ATP release occurs in response to flow, and the endogenously released ATP activates P2X4 receptors, thereby facilitating shear stress-dependent Ca2+ influx in HPAECs. An increase in [Ca2+]i is known to enhance the production of a variety of factors involved in the control of vascular tone such as nitric oxide and prostacyclin (21). It was demonstrated (13) in the rat pulmonary circulation perfused with blood-free Krebs solution that an increase in flow induces ATP release from ECs, which leads to pulmonary vasodilation. Increases in pulmonary blood flow caused by rapid constriction of the ductus arteriosus in fetal lambs are also shown (15) to cause pulmonary vasodilation that leads to a decrease in pulmonary vascular resistance. Thus the [Ca2+]i-increasing mechanism to which ATP contributes may play a role in the pulmonary vascular response to the increase in blood flow that occurs, e.g., during acute exercise and in the fetal lung at birth.
Under the present experimental conditions, flow induced Ca2+ influx but not Ca2+ release in HPAECs, which suggests that endogenously released ATP activates P2X4 but not G protein-coupled P2Y receptors. The activation of P2Y receptors increases inositol 1,4,5-trisphosphate [Ins(1,4,5)P3], which binds to Ins(1,4,5)P3 receptors on intracellular Ca2+ stores (such as the endoplasmic reticulum) and releases Ca2+. This may be due to the difference between P2X4 and P2Y in expression level or sensitivity to ATP. Competitive PCR showed that the P2Y1 and P2Y2 mRNA levels were only ∼5% of the P2X4 mRNA levels in HPAECs. P2X4 receptors are most sensitive to ATP, whereas P2Y1 receptors are more sensitive to ADP.
Flow has two effects on ECs. The first involves flow-induced changes in mass transport. If perfusates contain any bioactive substances capable of stimulating ECs, the mass transport of the substance increases as the flow rate or shear rate increases and thereby further stimulates ECs to release ATP or to increase [Ca2+]i. The second effect is shear stress, which mechanically deforms the ECs. To separate these two effects, we conducted flow-loading experiments with two perfusates that have different viscosities, which allowed us to apply different levels of shear stress at the same flow rate. The results clearly demonstrate that the flow-induced ATP release and Ca2+ influx are dependent on shear stress as a mechanical force rather than on flow rate-dependent mass transport.
Apyrase and AS-oligos to P2X4 receptors inhibited the flow-induced Ca2+ responses in HPAECs, which indicates that P2X4 and its ligands ATP and ADP are involved in the responses. Our previous study showed that human umbilical vein ECs (HUVECs) required exogenous ATP to exhibit Ca2+ responses to flow (3, 27). Thus the conditions required for Ca2+ responses seem to differ between HPAECs and HUVECs. This may be explained by the difference in the amount of ATP released in response to shear stress. HPAECs release more ATP in response to shear stress than HUVECs (data not shown). Because of their relatively poor ability to release ATP, HUVECs may require exogenously applied ATP to activate P2X4 receptors. P2X receptors are multisubunit receptors that exist in the form of homomers or heteromers of different subtypes. Our competitive PCR analysis shows that both HPAECs and HUVECs express the P2X subtypes P2X4, P2X5, and P2X7, but P2X4 was by far the most strongly expressed isoform. A recent study demonstrated abundant expression of P2X4 and P2X5 in human EC primary cultures derived from multiple blood vessels (22). Thus it is possible that HPAECs and HUVECs express distinct P2X homomers or heteromers with different affinities for ATP, and this may be related to the different conditions for Ca2+ responses between the two cell lines.
Abundant ecto-enzymes that degrade ATP are present on the EC surfaces, and these ecto-ATPases modify purinoceptor-mediated signaling by exerting an influence on the local concentration of ATP released by ECs. It was recently demonstrated that soluble ATPase is released from ECs in response to shear stress (28). It was also reported that ATP is generated on the surface of ECs by phosphotransfer reactions of ecto-nucleotide kinase (29) and by F0-F1 ATP synthase. Thus the P2X4-mediated Ca2+ response to shear stress seems to be dependent on correlations between two opposing mechanisms: an ATP-increasing mechanism and an ATP-decreasing mechanism. The extent to which each mechanism plays a role in ATP-mediated Ca2+ responses to shear stress, however, remains unclear.
Although the precise mechanism of shear stress-induced ATP release remains uncertain, possible mechanisms include cell lysis, exocytosis of secretary vesicles that contain ATP, ATP release via ATP channels or transporters, or ATP generation on the cell surface. Cell lysis, however, seems unlikely, because none was detected when shear stressed cells were examined microscopically, and cell lysis was quantitated by Trypan blue exclusion tests and lactate dehydrogenase assays (data not shown). Mechanical stimuli can modulate membrane traffic, such as exocytosis and endocytosis (5), and shear stress has been shown to enhance exocytosis of von Willebrand factor (vWF) from the vWF storage organelles (Weibel-Palade bodies), thereby increasing release of vWF from HUVECs (10). Evidence was recently obtained that ATP release by ECs during increased shear stress probably occurs via vesicular exocytosis (7). Inhibitors of vesicular transport such as monensin and N-ethylmaleimide significantly reduced the release of ATP from ECs in response to shear stress. This mechanism appears most possible, but other mechanisms should also be considered. Erythrocytes are reported to release ATP in response to mechanical deformation; this requires the activity of the cystic fibrosis transmembrane conductance regulator, which is a member of the ATP binding cassette (24). However, controversy remains as to whether the cystic fibrosis transmembrane conductance regulator actually mediates ATP release, because there are conflicting reports on its role in the regulation of ATP release (12, 25). Although additional studies are needed to clarify the mechanism of shear stress-dependent ATP release, elucidation of this problem would provide new insight into our understanding of shear stress signal transduction in ECs.
This work was partly supported by Grants-in-Aid for Scientific Research (A); for Scientific Research on Priority Areas; for Young scientists (B); and for Special Coordination Funds for Promoting Science and Technology from the Japanese Ministry of Education, Culture, Sports, Science, and Technology; and a research grant for cardiovascular diseases from the Japanese Ministry of Health and Welfare.
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 © 2003 by the American Physiological Society