INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ENDOTHELIUM IS AN ESSENTIAL component of the vascular wall and plays a fundamental role in hemostasis. Particularly, endothelial cells play an important role in maintaining vascular tone by releasing vasodilators such as endothelium-derived relaxing factor or nitric oxide (8, 19) and prostacyclin (18), as well as releasing vasoconstrictors such as endothelin (30). Changes of intracellular pH (pH_i) affect the enzyme activity of nitric oxide synthase, endothelin-converting enzyme in endothelial cells, and further affect their release (1, 17). pH was also found to mediate channel activities in endothelial plasma membrane, including the opening of calcium channels and calcium-activated potassium channels (24, 28), as well as affect calcium homeostasis inside the cells (32). Because of direct contact with flowing blood, endothelial cells are exposed to a wide diversity of physical and chemical stimuli, and pH_i is constantly affected.

It has been shown that hemodynamic shear stress can lead to a decrease in pH_i in rat aortic endothelial cells (33). Reducing the pH of perfusion solution can also cause a decrease in pH_i in cultured human umbilical vein endothelial cells (HUVEC; unpublished observation). It is therefore important to understand the pH_i regulation mechanisms in endothelial cells. Previous studies have demonstrated the existence of Na⁺/H⁺ exchange as an acid extruder in HUVEC (7, 9) and Na⁺-independent Cl/HCO₃ exchange as an acid loader in rat aortic endothelial cells (33). HCO₃-dependent acid extrusion pathways were found to play an important role in pH_i regulation in many other cell types (3, 12, 27). Recently, Na⁺-dependent Cl/HCO₃ exchange as an acid extruder has been shown to function in piglet cerebral microvascular endothelial cells (10) and rat aortic endothelial cells (31). However, little is known about HCO₃-dependent acid extrusion mechanisms in the HUVEC. We therefore studied acid efflux of cultured HUVEC in HCO₃ as well as in nominally HCO₃-free HEPES buffers in the present study.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolation and culture of HUVEC. HUVEC were obtained by collagenase (Sigma Chemical, Poole, UK) digestion of human umbilical vein using the method described previously (26). The cells were cultured in medium 199 with Earle's salts (ICN Biochemicals), containing 20% FCS, 100 IU/ml penicillin, 100 µg/ml streptomycin, and 0.01% heparin. HUVEC were subcultured using trypsin-EGTA, and seeded on gelatin-coated coverslips (6-mm diam, Chance Propper). Fourth to eighth passaged HUVEC were used for pH_i measurements. Twenty-four hours before the measurement, culture medium containing 20% FCS was removed and replaced by 1% FCS to minimize growth factor-mediated Na⁺/H⁺ exchange activity (20).

Measurement of pH_i. The pH_i of a single HUVEC was measured using carboxyseminaphthorhodafluor-1 (SNARF-1), a dual-emission fluoroprobe (4). One coverslip with HUVEC was placed at the bottom of the perfusion chamber, and cells were loaded by incubation in a 5 µM solution of SNARF-1-acetoxymethyl ester at room temperature for 15 min. Perfusion was then started at a constant flow rate of ~2.3 ml/min. The chamber temperature was maintained at 37°C. Individual cells were excited at 540 ± 12 nm, and fluorescence was measured simultaneously at 590 ± 5 and 640 ± 5 nm using an inverted microscope (Nikon Diaphot) converted for epifluorescence. The signals were then digitized at 0.5 kHz (CED1401). The emission ratio of 590/640 was calculated and converted to a linear pH scale using in situ calibration data from the nigericin technique (4). Finally, the pH_i signal was averaged over 0.5-s intervals.

Solutions. HEPES-buffered Tyrode contained (in mM) 140 NaCl, 4.5 KCl, 2.5 CaCl₂, 1 MgCl₂, 11 glucose, and 20 HEPES, pH adjusted to 7.4 at 37°C with NaOH. In HCO₃-buffered Tyrode, NaCl concentration was reduced to 117 mM and HEPES was replaced with 23 mM NaHCO₃. After initial review, more experiments were carried out in the HCO₃-buffered Tyrode plus 12 mM NaCl to maintain constant osmolarity between the HEPES-buffered and HCO₃-buffered Tyrode. The data obtained in this osmolarity- compensated HCO₃ buffer [e.g., resting pH_i 7.27 ± 0.02, n = 7; H⁺ efflux (J_H) 2.80 ± 0.48, n = 4] showed no significant difference from those in original HCO₃ buffer (see RESULTS). In Na⁺-free Tyrode, Na⁺ was replaced with N-methyl-D-glucamine. In Cl-free buffer, Na-gluconate, K-gluconate, Ca-gluconate (12 mM), and MgSO₄ were used. All HCO₃-buffered solutions were equilibrated with 5% CO₂-95% air and had a pH of 7.4 at 37°C. NH₄Cl, amiloride (or dimethyl amiloride, DMA), and DIDS (Sigma) were added to the solutions without osmotic compensation.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In nominally HCO₃-free HEPES-buffered Tyrode (extracellular pH 7.4), steady-state pH_i of a single HUVEC used in the study was 7.21 ± 0.01 (n = 57) at 37°C. After intracellular acid loads, induced by the NH₄Cl-removal method (21), pH_i in the cell recovered to resting level in a pH_i-dependent manner. The recovery was inhibited by 1.5 mM amiloride (n = 8; Fig. 1A), a classical Na⁺/H⁺ exchanger inhibitor (6), or in Na⁺-free Tyrode (n = 15; Fig. 2A), consistent with Na⁺/H⁺ exchange activity in these cells as reported previously (7, 9).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   A: intracellular pH (pHi) recovery in HEPES- and HCO3-buffered Tyrode. Recovery of pHi in HEPES buffer after an acid load is shown on left. After second acid load 1.5 mM amiloride was added, and pHi recovery was inhibited. Perfusion solution was then changed to HCO3 buffer in continued presence of amiloride. pHi started to recover after transient decrease of pHi. B: comparison of H+ efflux (JH) through Na+/H+ exchange and HCO3-dependent acid extrusion pathways. pHi recoveries in HEPES () and HCO3 buffer () shown in A were converted into JH according to equation JH = total · dpHi/dt, where dpHi/dt is change in pHi over time, total = i + CO2, CO2 = 2.3 [HCO3]i, and [HCO3]i is intracellular [HCO3]. From Henderson-Hasselbalch equation, [HCO3]i = [HCO3]o · 10(pHi pHo), where [HCO3]o is extracellular [HCO3]. Calculated JH values were plotted against pHi, and lines of best fit were derived by linear least-square regression.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   A: comparison of pHi recovery in HEPES- and HCO3-buffered Tyrode in absence of extracellular Na+. Recovery of pHi after acid load is inhibited in HEPES (left) but not in HCO3 buffer (right). B: pHi recovery in absence of extracellular Na+ in HCO3- buffered Tyrode. After acid load, pHi recovery still occurred in Na+-free solution but not fully, compared with that in Na+-containing solution. Addition of Na+ back to normal level caused further pHi recovery, even in presence of 1.5 mM amiloride.

In 5% CO₂, 23 mM HCO₃-buffered Tyrode, the steady-state pH_i in individual HUVEC was 7.26 ± 0.02 (n = 25), slightly more alkaline than that in HEPES buffer (7.21 ± 0.01). These values are similar to those of rat aortic endothelial cells reported in a previous study (7.27 ± 0.02 and 7.22 ± 0.03, respectively) (33). This suggests that HCO₃-dependent pH regulation mechanisms, like in many other tissues, may play a major role in maintaining resting pH_i level in endothelial cells under physiological conditions. Incubating the cells in Na⁺-free HCO₃ buffer decreased the resting pH_i by 0.09 ± 0.02 (restabilized at 7.16 ± 0.02, n = 5; data not shown). Figure 1A shows one of eight experiments where pH_i recovery after an acid load was inhibited by 1.5 mM amiloride in HEPES-buffered Tyrode. However, after the perfusion solution was changed to HCO₃-buffered Tyrode, pH_i started to recover in the presence of 1.5 mM amiloride following a transient drop [due to hydration of diffused CO₂ into the cell. This process equilibrated quickly, as shown in Fig. 1, because of the presence of carbonic anhydrase in the HUVEC (15, 16)]. This suggests that there are also HCO₃-dependent acid extruders in the HUVEC, as found in other types of cells (3, 12, 27). pH_i recovery through the HCO₃-dependent pathway was able to bring pH_i back to resting level.

To assess and compare J_H by different pathways, the intracellular intrinsic buffering power (_i) of individual cultured HUVEC was estimated using a method of stepwise reduction of external NH⁺₄ (from 20 mM, n = 7) (12). _i (in mM) equals [NH⁺₄]_i/pH_i, where [NH⁺₄]_i equals ([H⁺]_i × [NH⁺₄]_o)/[H⁺]_o (i means intracellular and o means extracellular for these concentrations.). The _i in HUVEC was plotted as a function of pH_i and fitted by linear equation (least squares; r = 0.95). _i equals 3.65 pH_i plus 41.10.

The pH_i recovery in HEPES buffer (through Na⁺/H⁺ exchange) and in HCO₃ buffer in the presence of amiloride (through HCO₃-dependent acid extruders) in the experiment shown in Fig. 1A was further plotted with J_H vs. pH_i. J_H was calculated according to the equation used in a previous study (J_H = _total · dpH_i/dt, where dpH_i/dt is change in pH_i over time; see details in legend) (12). From the linear-fitted data shown in Fig. 1B, the J_H through HCO₃-dependent pathways was faster than that of Na⁺/H⁺ exchange over the entire pH_i range. Pooled data from a total of six similar experiments showed that J_H through HCO₃-dependent pathways was on average nearly 1.5-fold greater than Na⁺/H⁺ exchange (2.73 ± 0.36 vs. 1.85 ± 0.36 meq/min, respectively) at mean pH_i of 7.04. These data suggested that under physiological conditions HCO₃-dependent acid extruders may contribute more than one-half in pH_i recovery from intracellular acidosis in the HUVEC.

In the absence of extracellular Na⁺, HCO₃dependent pH_i recovery after an acid load was not fully inhibited (Fig. 2); pH_i recovered but only back to 7.09 ± 0.03 (n = 20). For the experiments as shown in Fig. 2B, pH_i was allowed to restabilize at resting level after switching to HCO₃ buffer before acid loading was applied, so that the experiments were carried out in a steady state for HCO₃ equilibrium. These results suggest that there are probably two HCO₃-dependent acid extrusion pathways in HUVEC. One is dependent on extracellular Na⁺ and can mediate the pH_i recovery fully back to resting level (7.26 ± 0.02), and the other is independent of extracellular Na⁺ but can only mediate partial pH_i recovery. The acid efflux (J_H) through the latter was calculated to be 1.94 ± 0.47 meq/min at pH_i of 7.0, compared with 2.92 ± 0.67 meq/min for total HCO₃-dependent acid extrusion pathways (n = 4).

DIDS, a stilbene-derived anion exchanger inhibitor, has been shown to inhibit to a varying extent Cl/HCO₃ exchange (12), Na⁺-dependent Cl/HCO₃ exchange (23), and Na⁺-HCO₃ symport (12). We studied the effect of DIDS on the HCO₃-dependent acid extruders in HUVEC. DIDS (0.5 mM) completely inhibited pH_i recovery, at pH_i of 7.08 in HCO₃-buffered Tyrode, in the presence of amiloride or DMA [a more potent and relatively selective amiloride analog for the inhibition of Na⁺/H⁺ exchange (14)], suggesting inhibition of the Na⁺-dependent pathway (n = 3; Fig. 3). The same concentration of DIDS, however, inhibited pH_i recovery after an acid load by only 24.61 ± 0.42% at mean pH_i of 6.97 in Na⁺-free HCO₃-buffered Tyrode (n = 5, data not shown). This suggests that the Na⁺-independent pathway in HUVEC is less sensitive to DIDS.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Effect of DIDS on pHi recovery in HCO3-buffered Tyrode. As indicated, 0.5 mM DIDS slowed down pHi recovery after acid load. After DIDS was quickly washed off, pHi started to recover again.

We next tested the Cl dependency of the HCO₃dependent acid extrusion pathways in the HUVEC. As shown in Fig. 4A, intracellular alkalinization occurred after the perfusion solution was changed to Cl-free Tyrode (probably due to activation of HCO₃ influx in exchange of Cl efflux via Cl/HCO₃ exchange). pH_i recovery mediated by the HCO₃-dependent acid extruders in HUVEC was gradually slowed during consecutive acid loads, as intracellular Cl was depleted; the pH_i after each recovery was gradually elevated. The same results were obtained in nine other experiments. After extracellular Cl was returned to its normal concentration, pH_i recovery speeded up as intracellular Cl was replenished. These results suggest that both HCO₃-dependent acid extrusion pathways in HUVEC depend on intracellular Cl. Furthermore, after a long incubation in both Na⁺- and Cl-free HCO₃ buffer, in which only the Na⁺-independent pathway works, pH_i recovery after an acid load was fully inhibited (Fig. 4B, n = 5). This indicates that the Na⁺-independent HCO₃-dependent acid extrusion pathway is also Cl dependent. Finally, it was found that this novel Na⁺-independent but Cl- and HCO₃-dependent pathway was inhibited by 88.04 ± 4.76% in 145 mM K⁺ Tyrode solution (n = 7).

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Effect of removal of extracellular Cl on pHi recovery in normal or extracellular Na+-free HCO3-buffered Tyrode. A: series of intracellular acid loads were induced in absence of extracellular Cl and after extracellular Cl returned back to normal level. Dimethyl amiloride (DMA, 100 µM) was added to inhibit Na+/H+ exchange activity. Recovery of pHi was slowed down gradually after each acid load in extracellular Cl-free solution. B: pHi recovery after acid load in both extracellular Na+ and extracellular Cl-free HCO3 buffer.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study we demonstrated that in addition to Na⁺/H⁺ exchange there are two HCO₃dependent acid extrusion pathways in the HUVEC. Both Na⁺/H⁺ exchange and HCO₃-dependent acid extrusion pathways may play a major role in pH_i recovery in intracellular acidosis and in maintaining basal level pH_i under physiological conditions. The Na⁺- and Cl-dependent pathway, which can bring pH_i back to resting levels after an acid load, resembles the Na⁺-dependent Cl/HCO₃ exchange found in squid giant axons (3), snail neurons (27), and the chicken heart (13). The other HCO₃-dependent acid extruder in the HUVEC, which is Cl dependent but Na⁺ independent, has not been documented.

It has been reported in ureter smooth muscle cells that under extreme intracellular acidosis, HCO₃ conductance could contribute to pH_i recovery (2). However, it is unlikely that the Cl- and HCO₃-dependent but Na⁺-independent pH_i recovery we observed in the HUVEC could be due to HCO₃ influx through channels. The resting membrane potential of HUVEC has been estimated at 55.3 mV using perforated patch-clamp technique (22). At this membrane potential, the calculated pH_i equilibrium for HCO₃ is pH 6.50, which means that influx of HCO₃ can only occur at pH_i below this value. On the contrary, the Na⁺-independent acid extrusion pathway in HUVEC can actually bring pH_i back to 7.09 ± 0.03. Furthermore, if HCO₃ conductance could function as a base loader in the HUVEC, it should be enhanced by depolarization of membrane potential in a high-K⁺ solution. Instead, it was found that raising extracellular K⁺ inhibited the pH_i recovery. Therefore, the possibility of HCO₃ influx through a channel to work as an alkaline loader in HUVEC can be ruled out.

As previously shown in sheep cardiac Purkinje strands (29) and ureter smooth muscle cells (2), Cl/HCO₃ exchange was not reversed to contribute to acid extrusion. Taking typical values of 150 mM for extracellular Cl, 30 mM for intracellular Cl [the intracellular Cl concentration in HUVEC was calculated to be ~35 mM according to Cl reversal potential (22), although no data are available from direct measurement], and 7.40 for extracellular pH, the predicted equilibrium pH_i is 6.70 for Cl/HCO₃ exchange. Thus Cl/HCO₃ exchange would not be at equilibrium in a resting HUVEC with normal extracellular pH and pH_i >6.70. It would be energized to mediate HCO₃ efflux in exchange for Cl influx to act as an acid loader as reported previously in rat aortic endothelial cells (33). Actually, what we have found was that pH_i recovered back to 7.09 ± 0.03 after an acid load through the Na⁺-independent and HCO₃-dependent acid extrusion pathway. Only in the absence of extracellular Cl, as shown in Fig. 4, did the reversal of Cl/HCO₃ exchange occur, promoting Cl efflux and HCO₃ influx and causing alkalization. Therefore, it is less likely that Cl/HCO₃ exchange works in a reversed mode as a possible mechanism for the Na⁺-independent but Cl- and HCO₃-dependent pH_i recovery in HUVEC.

Recently, we have observed the involvement of carbonic anhydrase in rapid local pH_i regulation in cardiac myocyte, using a fast flow exchange perfusion chamber (unpublished observation). Vascular endothelial cells including HUVEC have been known to express several types of carbonic anhydrase in and/or on the surface of the cells (15, 16). Kurtz (11) demonstrated a plasma membrane H⁺-ATPase in rabbit S₃ proximal tubule, which can regulate pH_i in a Na⁺-independent manner but requires intracellular Cl. Chen and Boron (5) showed that CO₂/HCO₃ can stimulate a H⁺ pump in the same tissue, which explained the pH_i recovery mediated by a Na⁺-independent pathway in HCO₃ buffer. We have recently reported that lowering extracellular pH can mediate a fall of pH_i in guinea pig ventricular myocyte through a novel Cl-dependent but Na⁺-independent and DIDS-insensitive pathway (25). Whether the same or a similar mechanism as mentioned previously operates in the HUVEC to account for the Cl-dependent but Na⁺-independent pH_i recovery remains unclear. Further studies are needed to confirm the properties of the HCO₃- and Cl-dependent but Na⁺-independent acid extrusion pathway found in HUVEC in this study.