This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Foy, R. A. Articles by Paul, R. J. Search for Related Content PubMed PubMed Citation Articles by Foy, R. A. Articles by Paul, R. J. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB HYDROGEN OXYGEN

(Circulation Research. 1997;80:21-27.)
© 1997 American Heart Association, Inc.

Articles

The Effects of Hypoxia on pH_i in Porcine Coronary Artery Endothelium and Smooth Muscle

A Novel Method for Measurements in Endothelial Cells In Situ

Robyn A. Foy, Shunichi Shimizu, Richard J. Paul

the Department of Physiology and Biophysics, University of Cincinnati (Ohio) College of Medicine.

Correspondence to Richard J. Paul, Department of Physiology and Biophysics, University of Cincinnati College of Medicine, 231 Bethesda Ave, Cincinnati, OH 45267-0576. E-mail Richard.Paul@UC.EDU

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Endothelium-dependent relaxation of porcine coronary arteries is attenuated under hypoxic conditions. Recent evidence also indicates that pH_i may modulate the release of the endothelium-derived relaxing factor. We tested the hypothesis that hypoxia-induced attenuation of endothelium-dependent relaxation is mediated by alterations in pH_i. We developed a novel method for loading surface cells, whereby endothelial cell pH_i could be measured in situ on the intact porcine coronary artery. Endothelial cells of arterial ring segments were selectively loaded with the fluorescent indicator BCECF-AM. Differential loading of the endothelial cell layer was verified by confocal microscopy. pH_i of the endothelial cells in situ and of endothelium-denuded arteries was measured with a Photon Technology International spectrofluorimeter. The functional integrity of the endothelium was assessed by the endothelium-dependent relaxation to substance P in a paired adjacent ring. In the experimental protocol for pH_i measurements, preparations were perfused with a physiological bicarbonate buffer (pH 7.4), stimulated with KCl (29 mmol/L), and then subjected to hypoxia and reoxygenation. The mean basal pH_i in endothelial cells on the intact six arteries was 6.92±0.07. Addition of KCl to the perfusion medium decreased (P=.025) pH_i to 6.79±0.07. Subsequent bubbling with N₂ increased (P=.009) pH_i to 7.00±0.06, which was reversed by reoxygenation. In contrast to the in situ endothelium, pH_i of the smooth muscle was not significantly altered from its basal value of 7.24±0.06 (n=5) by either KCl or hypoxia. This differential behavior corroborated the confocal data indicating differential dye loading. These data thus suggest that oxygen-sensitive alterations in pH_i may be an important mechanism of signal transduction in endothelial cells.

Key Words: hypoxia • pH • endothelium • smooth muscle • coronary artery

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Hypoxia is known to significantly alter endothelial cell–mediated arterial smooth muscle responses,^{1 2 3 4 5} which may be of major significance in coronary and carotid artery vasospasm. We have previously shown that endothelium-dependent relaxation to A23187, thrombin, substance P, and acetylcholine is severely attenuated in porcine coronary arteries under hypoxic conditions.⁴ Hypoxia-induced attenuation of endothelium-dependent relaxation is believed to occur at the level of the vascular endothelium, since relaxation induced by sodium nitroprusside (SNP), an endothelium-independent vasodilator, was unimpaired by hypoxia.^{4 5} The mechanism(s) underlying the sensitivity of endothelial function to oxygen is not known.

Recently, investigators have demonstrated that ischemia and hypoxia are associated with an influx of Na⁺ into red blood cells, cardiac myocytes, and capillary endothelial cells.^{6 7 8} This influx of Na⁺ is postulated to occur through activation of the Na⁺-H⁺ exchanger, which is stimulated by the switch from aerobic to anaerobic metabolism that occurs during hypoxia.^{6 8} Activation of the Na⁺-H⁺ exchanger results in intracellular alkalinization.⁹ pH_i has been implicated in the flow-induced release of endothelium-derived relaxing factor (EDRF)¹⁰ and in the sustained release of nitric oxide elicited by bradykinin in endothelium.¹¹ Thus, it is plausible to suggest that alterations in pH_i may be an important mechanism of signal transduction in endothelial cells.

The present study was designed to test the hypothesis that endothelium-dependent relaxation is modulated by hypoxia-induced alterations in pH_i in endothelial and/or smooth muscle cells. We developed novel methodology to measure pH_i in in situ endothelial cells of porcine coronary arterial ring segments. Using confocal microscopy, we have demonstrated that the pH-sensitive dye BCECF-AM can be selectively loaded into the surface cell layer and that ratiometric dye technology can be used to measure hypoxia-induced alterations in pH_i. This is the first study to report hypoxia-induced alterations in pH_i in endothelial and smooth muscle cells of an intact coronary artery.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Tissue Preparation
Porcine hearts were obtained daily from a local slaughterhouse and transported to the laboratory immersed in ice-cold lactated Ringer's solution. Proximal left anterior descending coronary arterial rings were obtained by cutting 4-mm lengths of the artery perpendicular to the axis of blood flow. Vessel rings were suspended in 15-mL tissue baths and immersed in bicarbonate-buffered physiological saline solution (PSS) with the following composition (mmol/L): NaCl 118.3, NaHCO₃ 15.0, dextrose 11.1, KCl 4.7, MgSO₄ 1.2, KH₂PO₄ 1.2, EDTA 0.026, and CaCl₂ 2.5. The pH of the bath solution was 7.4 when gassed with 95% O₂/5% CO₂ at 37°C. Arterial rings were allowed to equilibrate for at least 1 hour with an initial resting tension of 40 mN before any experiment was begun. A resting tension of 40 mN has been shown previously to set a length in the range for optimal force development in this tissue.¹² At the end of the equilibration period, the viability of each preparation was evaluated by KCl-induced contraction. The maximal contractile response to KCl (80 mmol/L) was allowed to plateau, and the tissue was subsequently rinsed with PSS. This procedure was repeated until reproducible contractions were achieved. Endothelial function was assessed by the endothelium-dependent relaxation to substance P (0.01 µmol/L) in preparations submaximally contracted (0.9 maximal contractile response^{4 13} ) by 29 mmol/L KCl.

Measurement of pH_i in Endothelium of Intact Porcine Coronary Ring Segments
One coronary ring was used in isometric tension studies to assess the viability and pharmacological responsiveness of the endothelium; the adjacent ring was used to measure pH_i. The ring used in pH_i experiments was cut open and sutured (endothelial side out) onto a U-shaped hook, which was connected to a polytetrafluoroethylene holder. The preparation was mounted in a 2.4-mL cuvette, which was placed in the sample compartment of a Photon Technology International spectrofluorimeter configured for front face measurement. Fluorescence intensity was measured at excitation wavelengths of 505 and 439 nm and an emission wavelength of 523 nm. The background or autofluorescence of the preparation was measured, and the endothelial cells of the intact coronary ring preparation were selectively loaded with BCECF-AM. This was accomplished by using a Hamilton syringe to selectively position a drop of 5 µmol/L BCECF-AM in PSS onto the endothelial cell surface of the mounted tissue. The endothelium-containing side of the artery was exposed for a 2-minute interval and rinsed with PSS. This was repeated six times. As will be documented below, this procedure loads surface cells, whether endothelial or smooth muscle, and will be referred to as the "surface cell loading protocol."

pH_i was measured in endothelial cells of intact coronary ring segments during hypoxia and subsequent reoxygenation. The tissue was continuously perfused with PSS maintained at 37°C and gassed with 95% O₂/5% CO₂ during normoxic conditions. Preparations were sequentially exposed to KCl (29 mmol/L), hypoxia, and reoxygenation. In the presence of 29 mmol/L KCl, hypoxia was induced by perfusing the tissue, for a period of at least 20 minutes, with PSS that was gassed with 95% N₂/5% CO₂. Measurements of the cuvette PO₂ with an oxygen electrode yielded values in the range of 1% to 2% (7 to 14 mm Hg). It is worth noting that to achieve this level of hypoxia considerable care needs be taken with the perfusion to ensure no leakage of O₂, particularly in terms of the perfusion tubing. These values are in the range of PO₂ values achieved by bubbling of an open organ bath with N₂ of 1%, as previously reported.⁴ Reoxygenation consisted of perfusing the tissue with the normoxic PSS in the presence of 29 mmol/L KCl.

Measurement of pH_i in Coronary Artery Smooth Muscle
The coronary ring was everted, and the endothelium was removed by gently rubbing the ring on a piece of filter paper. The preparation was placed on a U-shaped hook and incubated in PSS for 1 hour. At the end of the equilibration period, the U-shaped hook was connected to a polytetrafluoroethylene holder and placed in a 2.4-mL cuvette. The tissue was oriented perpendicular to the beam of light for front face measurements in the spectrofluorimeter. The autofluorescence of the preparation was then measured. For some experiments (Figs 1 and 2, Table), only the surface smooth muscle layer was loaded using the surface cell loading protocol described for endothelial cell loading in the previous section. When loading of all the smooth muscle layers of the denuded artery was desired, perfusion of the tissue in the cuvette was discontinued, and BCECF-AM (5 µmol/L) was added to the cuvette (incubation protocol; see Reference 14). Incubation with the dye was discontinued (generally <1 hour) when the 505 signal was at least 10 times the baseline, at which time the 439 signal was at least three times baseline. Perfusion of the tissue was reinitiated in order to wash out the unhydrolyzed BCECF. After a stable baseline was achieved, pH_i was measured in tissues sequentially exposed to 29 mmol/L KCl, hypoxia, and reoxygenation.

View larger version (59K): [in this window] [in a new window]   Figure 1. Confocal microscopic images of porcine coronary artery. The left images in all cases show a transverse section, and the right images show a luminal section, as depicted by the schematic diagram above. An intact endothelium-containing artery (panel I, upper four images) and, as a control, an endothelium-denuded artery (panel II, lower four images) were loaded with the pH-sensitive dye BCECF using the same protocol. Our dye-loading protocol is designed to selectively load only the endothelium of the intact artery. The artery was then placed in PSS and set in the microscope. Images were made using 488-nm laser excitation, with low levels of white light to visualize the artery. The fluorescence intensity is localized to a luminal plane in both the intact (panel I, upper two images) and endothelium-denuded (panel II, upper two images) preparations. Rubbing the intact preparation to remove the endothelium eliminates the fluorescence (as shown in the lower two images of panel I), suggesting that the endothelium is the primary locus of the fluorescence. Note that in white light, as shown, the internal elastica is seen as the curly light band just inside the luminal surface. When viewed with only 488-nm laser excitation, the view is uniformly dark (data not shown). In contrast, when the smooth muscle of an endothelium-denuded artery is loaded with BCECF using the same protocol, rubbing the preparation does not eliminate the fluorescence (lower two images of panel II). The absence of any significant residual fluorescence after the intact artery is rubbed (lower two images of panel I) indicates that the smooth muscle layer was not loaded, because smooth muscle fluorescence is not eliminated by rubbing the tissue (lower images of panel II). Thus, the loading protocol is specific for endothelium.

View larger version (29K): [in this window] [in a new window]   Figure 2. Measurement of pHi using BCECF-AM in in situ endothelial cells of an intact porcine coronary artery. Top, Emission intensities (cps) measured at 523 nm for excitation at 439 nm (a pH-insensitive wavelength) and 505 nm (the optimum pH-sensitive wavelength). Bottom, Ratio of 505/439 intensities, calibrated in pHi units (see "Materials and Methods"). For both top (paired lines) and bottom (single line) panels, selectively loaded endothelial cells were exposed to the cell-alkalizing agent NH4Cl (30 mmol/L) to alter pHi (middle records). Bottom records of both panels show the response of this same tissue to NH4Cl after removal of the endothelial cell layer by gently rubbing the tissue with a piece of filter paper. The response to NH4Cl is abolished, thereby showing selective loading of the vascular endothelium. Top records of both panels illustrate return of the NH4Cl-induced alkalinization response after reexposure of the smooth muscle to BCECF-AM, using the surface cell loading protocol. This shows that the absence of response (bottom records of both panels) is not simply due to damage of the smooth muscle potentially associated with the denudation.

pH_i Calibration
All measurements in endothelial or smooth muscle cells were expressed as the following ratio: (fluorescence intensity at 505 nm minus background autofluorescence)/(fluorescence intensity at 439 nm minus background autofluorescence). Absolute values of pH_i were calibrated using the high-K⁺ nigericin technique previously described.^{14 15} At the end of the experiment, nigericin (8 µmol/L) was added to the tissue, and a calibration curve was constructed using Na⁺-free, high-K⁺, MOPS-buffered solutions of known pH values (6.8 to 7.8). The ratio of fluorescence intensity at 505 and 439 nm minus the tissue autofluorescence was linearly related to pH_i. A linear regression line for each individual tissue preparation was constructed and used to convert fluorescence intensity ratios to pH_i values.

Confocal Microscopy
The endothelium of an intact artery was loaded with BCECF-AM as described. The artery was then cut into transverse strips 0.5 mm in width and placed in PSS for observation. Images were made with a Bio-Rad MRC-600 confocal imaging system mounted on a Nikon Diaphot inverted microscope equipped for epifluorescence illumination (College of Medicine, confocal microscope facility). Laser excitation (488 nm) was used to excite the fluorescence with sufficient white light to distinguish the tissue from background. These preparations were not fixed or mounted, but conditions were similar to those used in our spectrofluorimeter. Both luminal and transverse images were obtained by rotating the specimen 90°. After one set of images was obtained, the preparation was denuded of endothelium, and further images were taken. As a control, in a paired artery segment, the endothelium was first denuded, and the dye was loaded into the smooth muscle layer using an identical loading procedure. After a set of images was obtained, this preparation was subjected to the same gentle rubbing procedures and a sham denudation, and further images were obtained.

Chemicals
The following drugs were used: substance P and nigericin (Sigma Chemical Co), BCECF-AM (Molecular Probes), and ammonium chloride (Matheson Coleman & Bell Manufacturing Chemists). Nigericin was dissolved in ethanol and BCECF-AM in dimethyl sulfoxide. All other drugs were dissolved in distilled water. All concentrations represent final bath concentrations.

Data Analysis and Statistics
pH_i responses were expressed as mean±SEM, and the difference between the means was analyzed by two-way ANOVA with the Bonferroni method used to test pairwise comparisons. In each protocol, experiments were performed on coronary arteries from at least five hearts. Changes were deemed significant at a value of P<.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Confocal Microscopic Assessment of the Differential Dye Loading Protocol
We used confocal microscopy to validate our protocol for loading surface cells. Fig 1 shows transverse (left images) and luminal (right images) views of an artery subjected to our endothelial cell loading protocol. Panel I (the upper four images) presents confocal images of an intact artery in which the dye was loaded into the endothelium. The upper two images of panel I show a strong fluorescence limited to a narrow luminal band. The luminal surface of this preparation was then rubbed to remove the endothelium. The lower two images of panel I show little remaining fluorescence in the artery denuded of endothelium, despite the higher gain of the instrument. Note that in white light as shown, the internal elastica is seen as the curly light band just inside the luminal surface. When viewed only with 488-nm laser excitation, the view is uniformly dark (data not shown). The loss of fluorescence after removal of the endothelium suggests that the dye was primarily loaded into the endothelial cell layer.

As a control, panel II of Fig 1 (lower four images) shows confocal images from a paired artery segment that was first denuded of endothelium by gentle rubbing and then subjected to the same dye loading protocol used to load the endothelium (panel I). The fluorescence again was seen to be limited to a luminal band (panel II). In contrast to the intact artery (compare lower two images of panels I and II), rubbing these smooth muscle preparations did not affect the strong fluorescence in the luminal region (lower two images of panel II). Dye loaded into the smooth muscle layer could not be removed by the rubbing procedure used in the standard endothelium denudation. Thus, if there was significant loading of the dye into the subendothelial layer of smooth muscle in the intact artery, fluorescence would not be eliminated by rubbing. Since rubbing abolished the fluorescence of the endothelium-loaded artery (see lower two images of panel I), the confocal microscopic evidence indicates that our dye loading protocol is specific for endothelium. This is supported by the following functional data. These images were typical of four such experiments.

pH_i Measurement in Endothelial Cells of Intact Porcine Coronary Ring Segments
We performed the following control experiments to estimate the percentage of the fluorescent signal attributed to dye loaded into the endothelium compared with dye potentially loaded into the underlying smooth muscle. The average fluorescence intensities of the in situ endothelial cells under various conditions are presented in the Table. Endothelial cells of an intact coronary artery were selectively loaded with BCECF-AM (Fig 2) middle records (a) as described in "Materials and Methods." After we loaded the endothelial cells, the pH_i-sensitive 505-nm excitation signal increased by 77%, and the near pH_i-insensitive 439-nm signal increased by 19%. The signal-to-noise ratio of the fluorescence intensities, measured at an emission of 523 nm, was more than sufficient for measurement and calibration of pH_i (see Figs 2 and 3).

View this table: [in this window] [in a new window]   Table 1. Fluorescence Intensities in Surface Cell Loading Protocols

View larger version (23K): [in this window] [in a new window]   Figure 3. Effect of hypoxia and reoxygenation on pHi in endothelial cells of an intact porcine coronary ring segment. Preparations were perfused with bicarbonate-buffered PSS (pH 7.4), stimulated with KCl (29 mmol/L), and subjected to hypoxia (95% N2/5% CO2) and reoxygenation (95% O2/5% CO2). Calibration was achieved using a Na+-free high-K+ MOPS buffer at pH values (6.8 to 7.8) in the presence of nigericin (8 µmol/L).

To functionally assess our surface cell loading protocol for selective loading of BCECF-AM (see "Materials and Methods") into the endothelial cell layer of an intact artery, we used the intracellular alkalinizing agent NH₄Cl (30 mmol/L) as a tool to perturb pH_i. Fig 2, top panel, shows the individual emission (523-nm) intensities for excitations at 439 nm (a pH-insensitive wavelength) and 505 nm. Fig 2, bottom panel, shows the ratio of these intensities, calibrated in pH_i units. In order to verify that the observed alkalinization was attributed to alterations in pH_i in the endothelium, the tissue was denuded, the remaining fluorescence at the appropriate excitation wavelengths (505 and 439 nm) was measured, and the underlying smooth muscle was reexposed to NH₄Cl. It can be noted that a considerable decrease in the fluorescence intensities was observed upon removal of the endothelium (bottom record of Fig 2, top panel) and that little response to NH₄Cl can be seen in the denuded preparation (bottom record of Fig 2, bottom panel). After denudation of the endothelium, the fluorescence intensities at 439 nm and 505 nm represented 104.0±2.5% and 120.4±5.7% (n=5), respectively, of the background or autofluorescence measured at the beginning of the experiment before dye loading. It was determined that the response to NH₄Cl after denudation (n=5) represented 22.0±4.9% of the response observed in endothelium-intact rings. The subsequent lack (in some cases) or attenuation of response to NH₄Cl, in the same tissue once it was denuded, was taken as functional evidence that the endothelial cells had been selectively loaded. In order to demonstrate that the diminished response to NH₄Cl in denuded tissues was not the result of damage to the smooth muscle cells incurred during removal of the endothelium, the underlying smooth muscle was reloaded with BCECF-AM using our surface cell loading protocol, and the response to NH₄Cl was measured (Fig 2, upper records). The surface smooth muscle layer retained its capacity for dye loading, and intracellular alkalinization was observed upon exposure to NH₄Cl.

On the basis of these control experiments, we estimate that at least 78±4.9% of the measured fluorescence signal is attributed to dye loaded into the endothelium. The remaining signal as detected by exposure to NH₄Cl in denuded tissues may represent (1) small changes in orientation or position of the tissue in the light path, (2) dye penetration into the underlying smooth muscle layer, (3) incomplete endothelial denudation, and/or (4) changes in autofluorescence of the smooth muscle layer. During the present study, we developed protocols that permitted the removal and replacement of the preparation in the fluorimeter with minimal alterations in fluorescence signal; any differences were less than the noise in the intensity ratio (see Figs 2 and 3). This implied that changes in orientation were minimal. However, because of the care taken to keep the orientation of the tissue constant, the endothelium may not have been completely removed during the denudation process. We believe that this is the most likely explanation for the small residual response to NH₄Cl. In subsequent experiments (below), the basal pH_i and magnitudes of responses differ in endothelial and smooth muscle cells, further validating the selectivity of our differential loading protocols.

Effects of Hypoxia and Reoxygenation on pH_i in In Situ Endothelial Cells
The purpose of the present study was to determine if alterations in pH_i may have contributed to the reported inhibition by hypoxia of endothelium-dependent relaxation in porcine coronary artery.^{4 13} We showed that 29 mmol/L KCl produced a stable near-maximal contraction in this preparation. Hypoxia caused a 50% decrease in isometric force, which was reversible upon reoxygenation^{4 13} ; see also Fig 5, top. This decrease in force was similar in intact and denuded arteries.

View larger version (22K): [in this window] [in a new window]   Figure 5. Effect of hypoxia and reoxygenation on isometric force (top) and pHi (bottom) in coronary artery smooth muscle loaded with BCECF by the incubation protocol. Open bar indicates the presence of 29 mmol/L KCl; filled bar, aeration with 95% N2/5% CO2 (hypoxia) in place of 95% O2/5% CO2. Calibration (time, >3400 seconds) was achieved using a Na+-free high-K+ MOPS buffer at pH values (6.8 to 7.8) in the presence of nigericin (8 µmol/L). Inset displays the experimental period on an expanded scale. KCl depolarization and hypoxia slightly increased steady state pHi (0.01) in this case, but a tendency toward alkalinization was not statistically significant in the five arteries studied (Fig 4B).

In order to parallel these previously reported functional studies, pH_i was measured in arteries stimulated with KCl (29 mmol/L) and then subjected to hypoxia and reoxygenation. A typical record illustrating the effects of hypoxia and reoxygenation on pH_i is shown in Fig 3. The average data for pH_i in in situ endothelial cells from six arteries is summarized in Fig 4. The mean basal pH_i was 6.92±0.07. Addition of KCl to the perfusion medium significantly decreased pH_i. Under hypoxic conditions over a period of 20 minutes, pH_i increased to a steady state value of 7.00 (P=.009). Upon reoxygenation, pH_i decreased to 6.91 (P=.011). These are the first known measurements of oxygen-sensitive alterations in pH_i in endothelial cells of an intact coronary ring segment.

View larger version (17K): [in this window] [in a new window]   Figure 4. The average values of pHi for the experiment presented in Fig 3 for in situ endothelial cells (n=6, A) and in Fig 5 for smooth muscle (n=5, B) of porcine coronary artery. Measurements were made in the steady state (ss) unless otherwise noted. Bars indicate mean±SEM. *Statistically significant difference with respect to the previous condition, graphed to the left. (P values are .0254, .009, and .011, from left to right, respectively.) Hypoxia significantly affects the endothelium but not the smooth muscle.

Effects of Hypoxia and Reoxygenation on pH_i in Porcine Coronary Smooth Muscle
In a parallel experimental protocol, pH_i was measured in the smooth muscle of endothelium-denuded coronary arteries, loaded by incubation with BCECF. A typical experimental record is shown in Fig 5, and the average steady state data for five arteries are summarized in Fig 4B. In contrast to the endothelium, neither KCl nor hypoxia had any significant effect on smooth muscle pH_i (P<.05, two-way ANOVA). The mean basal pH_i was 7.24±0.06. With KCl stimulation, the steady state pH_i was unchanged. In contrast to the endothelial cell response, there was little change in smooth muscle pH_i in the steady state under hypoxic conditions. In total, the effects of hypoxia on coronary artery smooth muscle pH_i were small and not significant. Thus, pH_i is unlikely to be a factor in the relaxation of this smooth muscle to hypoxia.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
This is the first study to measure alterations in pH_i in endothelial cells of an intact coronary ring segment. We have demonstrated selective loading of the endothelium of an intact tissue with BCECF-AM. Using confocal microscopy, we demonstrated that an intense fluorescence was limited to a narrow luminal band after subjecting an endothelium-intact artery to our loading procedure. Removal of the endothelium by gentle rubbing abolished this fluorescence. Moreover, when the smooth muscle of an endothelium-denuded artery was subjected to the same dye-loading procedure, only surface cells were loaded. Moreover, rubbing the artery in a similar fashion did not eliminate the fluorescence. If substantial loading into a luminal smooth muscle layer occurred during our loading procedure in an endothelium-intact preparation, one would expect significant residual fluorescence after rubbing. This evidence for selective loading was further supported by functional evidence indicating differential pH_i responses of endothelium and smooth muscle measured in the spectrofluorimeter.

The estimated fluorescence signal attributed to dye contained in the in situ endothelium was at least 78% of the total measured signal. We investigated several potential causes that may have accounted for the remaining 22% of the fluorescence signal observed during control experiments: (1) small changes in orientation or position of the tissue in the light path, (2) dye penetration into the underlying smooth muscle layer, (3) incomplete endothelial denudation, and (4) contributions of changes in autofluorescence of the smooth muscle layer. The most likely cause is that not all of the loaded endothelial cells were removed during the denudation process, since extreme care had to be taken in order to ensure that the position and orientation of the sutured ring preparation were unaltered.

Since pH_i was measured in intact and denuded ring preparations, it was possible to distinguish between alterations in pH_i observed in endothelial versus smooth muscle cells. The basal pH_i values and responses to KCl, hypoxia, and reoxygenation were quantitatively different in intact versus denuded preparations. Other investigators have reported basal pH_i values ranging from 6.78 to 7.26 in cultured endothelial cells in the absence and presence of bicarbonate.^{10 16 17} In the present study, the mean basal pH_i in endothelial cells (n=6 arteries) in situ was 6.92±0.07; the mean pH_i in smooth muscle cells (n=18 arteries) was 7.21±0.02. As with all studies using fluorescent dyes, the accuracy of absolute calibration is dependent on several assumptions. It has been reported¹⁸ that BCECF has a heterogeneous distribution in sea urchin eggs, and this type of compartmentation could lead to an underestimate of pH_i. Less is known about compartmentation in mammalian cells. For these reasons, changes in pH_i rather than the absolute values would appear to carry more weight.

Addition of KCl to the perfusion medium tended to decrease pH_i in endothelial cells but did not alter pH_i in denuded preparations. Although it is not possible to completely rule out a contribution of the underlying smooth muscle to the remaining 22% of the fluorescent signal, the differences in response of endothelial cells and smooth muscle support our confocal microscopy data (Fig 1), indicating that the measurements made after our endothelium-selective loading protocol primarily reflect changes in pH_i in the endothelium.

This new technique allowed us to assess the effects of hypoxia on pH_i in in situ endothelial cells. We have shown that in the presence of KCl, hypoxia caused significant increases in pH_i. This was somewhat unexpected, since hypoxia is associated with increased lactate production due to the utilization of anaerobic glycolysis. The basis for this alkalinization is unknown and likely many faceted. Recently, investigators have demonstrated that ischemia and hypoxia are associated with an influx of Na⁺ into red blood cells, cardiac myocytes, and capillary endothelial cells.^{6 7 8} This influx of Na⁺ is postulated to occur through activation of the Na⁺-H⁺ exchanger, which is stimulated by the switch from aerobic to anaerobic metabolism that occurs during hypoxia.^{6 8} Activation of the Na⁺-H⁺ exchanger could result in intracellular alkalinization.⁹ In the case of the coronary artery smooth muscle, we have reported that hypoxia decreased the phosphocreatine content,¹⁹ which could also account for an increase in pH_i. Clearly pH_i regulatory mechanisms must also be involved to compensate for the known increase in lactate production in both endothelial and smooth muscle cells under hypoxic conditions.^{19 20 21 22} It is worth noting that cyanide treatment (which, similar to hypoxia, increases lactate production) has also been reported to be associated with an increase in pH_i in the smooth muscle of rat portal vein as determined by nuclear magnetic resonance techniques.²³

The intracellular alkalinization measured in endothelial cells in this study corresponds temporally with attenuation of endothelium-dependent relaxation previously reported in this same preparation under hypoxic conditions.⁴ It is not possible, however, to make an exact comparison of time courses. This is due to the fact that the relaxation of an intact artery reflects an average response over all smooth muscle layers, whereas pH_i measurements for both endothelium and smooth muscle reflect only those in the outer layers. Our evidence does suggest that intracellular alkalinization may be one possible factor contributing to our previously reported attenuation of endothelium-dependent relaxation by hypoxia.⁴

The mechanism by which hypoxia and/or intracellular alkalinization could modulate endothelial function is unclear. Other investigators have examined the role of cyclooxygenase products, interruption of the cGMP pathway, and inhibition of oxidative metabolism as possible mechanisms for hypoxia-induced attenuation of endothelium-dependent relaxation.^{4 5 24 25} Since it is known that endothelium-dependent vasodilators, such as thrombin, release prostacyclin as well as EDRF and that this release is markedly attenuated during conditions when oxygen concentrations are limited, it is possible that attenuation of endothelium-dependent relaxation during hypoxia is the result of decreased release and/or synthesis of cyclooxygenase products.²⁴ However, Hashimoto et al⁴ have demonstrated that the inhibition of endothelium-mediated relaxation, in the presence of hypoxia, was not due to attenuation of the action of cyclooxygenase products, such as prostacyclin, because relaxation to A23187, thrombin, and substance P was unaffected by exposure to indomethacin, a cyclooxygenase inhibitor. This finding suggests that hypoxia is affecting the synthesis, release, and/or pathway of action of nitric oxide.

EDRF-mediated smooth muscle relaxation is known to occur through the activation of a soluble guanylate cyclase, which causes an increase in smooth muscle cGMP.²⁶ Johns et al⁵ have shown, in rabbit pulmonary artery, that hypoxia decreased basal levels of cGMP and prevented endothelium-mediated increases but did not affect increases in cGMP induced by SNP. These data suggest that hypoxia exerts its effect proximal to the activation of guanylate cyclase, since SNP-induced increases in cGMP were unaffected by hypoxia. This is consistent with the lack of effects of hypoxia on SNP relaxation of coronary artery smooth muscle⁴ and with the relatively minor effects of hypoxia on smooth muscle pH_i seen in the present study.

It is also likely that hypoxia may inhibit the basal release of EDRF, since basal levels of cGMP were reduced.⁵ Alternatively, the effect of hypoxia may be related to the inhibition of oxidative phosphorylation that occurs during limited oxygen supply. Griffith et al²⁵ have shown that endothelium-dependent relaxation, in rabbit aortic strip preparations, is attenuated by inhibitors of oxidative phosphorylation. In order to determine whether hypoxia-induced inhibition of oxidative phosphorylation was responsible for attenuation of endothelium-mediated responses in porcine coronary arteries, Hashimoto et al⁴ treated preparations with the metabolic inhibitor cyanide (5 mmol/L), and the relaxation response to A23187, thrombin, and substance P was measured during in vitro isometric tension studies. In the presence of cyanide, the relaxation response to these endothelium-dependent vasodilators was unaltered, suggesting that the hypoxia-induced switch to anaerobic metabolism was not the cause of attenuation of endothelium-dependent relaxation in porcine coronary arteries. These studies illustrate that hypoxia-induced attenuation of endothelium-dependent relaxation is not the result of inhibition of the action of cyclooxygenase products, the activation of the cGMP pathway, or oxidative metabolism per se. However, we have demonstrated that hypoxia is associated with increases in pH_i that correspond temporally with attenuation of endothelium-dependent relaxation of porcine coronary arteries. Thus, alterations in pH_i may be an important component of signal transduction in endothelial cells.

The present study focused on the cellular level effects of hypoxia, defined here by bubbling with N₂. The resultant PO₂ of 1 to 2 mm Hg would be severe hypoxia but is comparable to that used in previous studies of the effects of hypoxia on endothelial-vascular interactions. In this range, endothelial cell mitochondria would not be expected to be oxygen-limited. Oxygen consumption by the deeper smooth muscle layers on the other hand would likely render these layers anoxic and inhibit their mitochondria. Under severe hypoxia, when other mechanisms, such as those mediated by adenosine are maximized at the resistance vessel level, it is likely that the epicardial arteries may play a role in the regulation of flow. Thus, the mechanisms under these conditions may be of physiological significance to cardiac function.

We have demonstrated the feasibility of measurement of pH_i in endothelial cells in situ in intact coronary ring segments by using ratiometric fluorescent dye technology. This approach offers the potential for measurement of other intracellular ions in endothelial cells in situ. Moreover, it provides a methodology for assessing the effects of in situ contact with vascular smooth muscle on endothelial cell response, which is not possible using conventional measurements on cultured endothelial cells. Using this approach, we have shown that hypoxia elicits an alkalinization in endothelial cells in situ. This alkalinization may play an important modulatory role in the attenuation of endothelium-dependent relaxation observed under hypoxic conditions as well as other conditions, such as metabolic alkalosis.

	Acknowledgments

 
This study was supported by National Institutes of Health grants HL-23240 and TG32 HL-07571 and American Heart Association Grant-in-Aid 92007130. We thank Peggy Sue Bowman and Jianyi Li for their skillful technical assistance.

Received February 23, 1996; accepted October 11, 1996.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Vanhoutte PM, Rimele TJ. Role of the endothelium in the control of vascular smooth muscle function. J Physiol (Paris).. 1983;78:681-686.

2. Vanhoutte PM, Shimokawa H. Endothelium-derived relaxing factor and coronary vasospasm. Circulation. 1989;80:1-9.[Abstract/Free Full Text]

3. Furchgott RF, Vanhoutte PM. Endothelium-derived relaxing and contracting factors. FASEB J. 1989;3:2007-2018.[Abstract]

4. Hashimoto M, Close LA, Ishida Y, Paul RJ. Dependence of endothelium-mediated relaxation on oxygen and metabolism in porcine coronary arteries. Am J Physiol. 1993;265:H299-H306.[Abstract/Free Full Text]

5. Johns RA, Linden JM, Peach MJ. Endothelium-dependent relaxation and cyclic GMP accumulation in rabbit pulmonary artery are selectively impaired by moderate hypoxia. Circ Res. 1989;65:1508-1515.[Abstract/Free Full Text]

6. Cala PM, Anderson SE, Cragoe EJ. Na⁺/H⁺ exchanger exchange-dependent cell volume and pH regulation and disturbances. Comp Biochem Physiol. 1988;90A:551-555.

7. Haigney MCP, Miyata H, Lakatta EG, Stern MD, Silverman HS. Dependence of hypoxic cellular calcium loading on Na⁺-Ca²⁺ exchange. Circ Res. 1992;71:547-557.[Abstract/Free Full Text]

8. Mazzoni MC, Intaglietta M, Cragoe EJ, Arfors KE. Amiloride-sensitive Na⁺ pathways in capillary endothelial cell swelling during hemorrhagic shock. J Appl Physiol. 1992;73:1467-1473.[Abstract/Free Full Text]

9. Frelin C, Vigne P, Ladoux A, Lazdunski M. The regulation of the intracellular pH in cells from vertebrates. Eur J Biochem. 1988;174:3-14.[Medline] [Order article via Infotrieve]

10. Ziegelstein RC, Cheng L, Capogrossi MC. Flow-dependent cytosolic acidification of vascular endothelial cells. Science. 1992;258:656-659.[Abstract/Free Full Text]

11. Fleming I, Hecker M, Busse R. Intracellular alkalinization induced by bradykinin sustains activation of the constitutive nitric oxide synthase in endothelial cells. Circ Res. 1994;74:1220-1226.[Abstract/Free Full Text]

12. Ngai JN, Roth PS, Paul RJ. Effects of endothelium on basal tone and agonist and O₂ sensitivity in porcine coronary artery. J Pharmacol Exp Ther. 1990;252:1053-1059.[Abstract/Free Full Text]

13. Close LA, Bowman PS, Paul RJ. Reoxygenation-induced relaxation of coronary arteries: a novel endothelium-dependent mechanism. Circ Res. 1994;74:870-881.[Abstract/Free Full Text]

14. Nagesetty R, Paul RJ. The effects of pH_i on isometric force and [Ca²⁺]_i in porcine coronary artery smooth muscle. Circ Res. 1994;75:990-998.[Abstract/Free Full Text]

15. Krampetz IK, Rhoades RA. Intracellular pH: effect on pulmonary arterial smooth muscle. Am J Physiol. 1991;260:L1-L6.[Abstract/Free Full Text]

16. Raju N, Kim D, Brock TA. Intracellular alkalinization leads to Ca²⁺ mobilization from agonist-sensitive pools in bovine aortic endothelial cells. J Biol Chem. 1990;265:19071-19076.[Abstract/Free Full Text]

17. Escobales N, Longo E, Cragoe EJ, Danthuluri NR, Brock TA. Osmotic activation of Na⁺/H⁺ exchange in human endothelial cells. Am J Physiol. 1990;259:C640-C646.[Abstract/Free Full Text]

18. Mozingo NM, Chandler DE. The fluorescent probe BCECF has a heterogeneous distribution in sea urchin eggs. Cell Biol Int Reports. 1990;14:689-699.[Medline] [Order article via Infotrieve]

19. Ishida Y, Hashimoto M, Paul RJ. Does a limitation of energy supply to the contractile apparatus underlie the relaxation induced by hypoxia in smooth muscle? In: Siegman MJ, Somlyo AP, Stephens NL, eds. Regulation And Contraction In Smooth Muscle. 1987:463-464.

20. Paul RJ. The chemical energetics of vascular smooth muscle: intermediary metabolism and its relation to contractility. In: Bohr DF, Somlyo AP, Sparks HV, eds. Handbook of Physiology, Section II: The Cardiovascular System. Baltimore, Md: American Physiological Society; 1980:201-235.

21. Dobrina A, Rossi F. Metabolic properties of freshly isolated bovine endothelial cells. Biochim Biophys Acta. 1983;762:295-301.[Medline] [Order article via Infotrieve]

22. Mertens S, Noll T, Spahr R, Krutzfeldt A, Piper HM. Energetic response of coronary endothelial cells to hypoxia. Am J Physiol. 1990;258:H689-H694.[Abstract/Free Full Text]

23. Hellstrand P, Vogel H. Phosphagens and intracellular pH in intact rabbit smooth muscle studied by ³¹P-NMR. Am J Physiol. 1981;248:C320-C329.

24. Hempel SL, Haycraft DL, Hoak JC, Spector AA. Reduced prostacyclin formation after reoxygenation of anoxic endothelium. Am J Physiol. 1990;259:C738-C745.[Abstract/Free Full Text]

25. Griffith TM, Edwards DH, Newby AC, Lewis MJ, Henderson AH. Production of endothelium derived relaxant factor is dependent on oxidative phosphorylation and extracellular calcium. Cardiovasc Res. 1986;20:7-12.[Medline] [Order article via Infotrieve]

26. Fiscus RR. Molecular mechanisms of endothelium-mediated vasodilation. Semin Thromb Hemost.. 1988;14:12-22.

This article has been cited by other articles:

				E. Boedtkjer and C. Aalkjaer Insulin inhibits Na+/H+ exchange in vascular smooth muscle and endothelial cells in situ: involvement of H2O2 and tyrosine phosphatase SHP-2 Am J Physiol Heart Circ Physiol, February 1, 2009; 296(2): H247 - H255. [Abstract] [Full Text] [PDF]

				L. A. Shimoda, T. Luke, J. T. Sylvester, H.-W. Shih, A. Jain, and E. R. Swenson Inhibition of hypoxia-induced calcium responses in pulmonary arterial smooth muscle by acetazolamide is independent of carbonic anhydrase inhibition Am J Physiol Lung Cell Mol Physiol, April 1, 2007; 292(4): L1002 - L1012. [Abstract] [Full Text] [PDF]

				O. Frobert, E. O Mikkelsen, J. P Bagger, and C. H Gravholt Measurement of interstitial lactate during hypoxia-induced dilatation in isolated pressurised porcine coronary arteries J. Physiol., February 15, 2002; 539(1): 277 - 284. [Abstract] [Full Text] [PDF]

				M. Toporsian, K. Govindaraju, M. Nagi, D. Eidelman, G. Thibault, and M. E. Ward Downregulation of Endothelial Nitric Oxide Synthase in Rat Aorta After Prolonged Hypoxia In Vivo Circ. Res., March 31, 2000; 86(6): 671 - 675. [Abstract] [Full Text] [PDF]

				S. Shimizu and R. J. Paul Hypoxia and Alkalinization Inhibit Endothelium-Derived Nitric Oxide But Not Endothelium-Derived Hyperpolarizing Factor Responses in Porcine Coronary Artery J. Pharmacol. Exp. Ther., October 1, 1999; 291(1): 335 - 344. [Abstract] [Full Text]

				S. Fukuda, T. Toriumi, H. Xu, H. Kinoshita, H. Nishimaki, S. Kokubun, N. Fujiwara, H. Fujihara, and K. Shimoji Enhanced beta -receptor-mediated vasorelaxation in hypoxic porcine coronary artery Am J Physiol Heart Circ Physiol, October 1, 1999; 277(4): H1447 - H1452. [Abstract] [Full Text] [PDF]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Foy, R. A. Articles by Paul, R. J. Search for Related Content PubMed PubMed Citation Articles by Foy, R. A. Articles by Paul, R. J. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB HYDROGEN OXYGEN