Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The endothelium lining the inner wall of blood vessels is a complex physiologic system playing a crucial role in modulating vessel tone and remodeling, blood fluidity, and cell adhesion (1). The properties of endothelial cells differ from tissue to tissue, being also dependent on the size of the vessel. Small pulmonary arteries with diameters from 30 to 50 µm are the major determinants of pulmonary arterial pressure and vascular resistance (6). For technical reasons, the electrophysiologic investigation of these cells in situ has not been possible so far and therefore a number of questions concerning their basic function remain open. Although the resting potential in endothelial cells was shown to play an important role in Ca²⁺ homeostasis and nitric oxide (NO) synthesis (2, 7, 8), little is known about the mechanisms of its generation and endogeneous regulation. In addition, endothelial cells form a tight syncytium but to our knowledge there is no information about the degree of electrical coupling between intact pulmonary endothelial cells in situ.

To address these questions, we developed a new method for investigating intact cells in fresh slices of the lung. The present results show that the properties of pulmonary endothelial cells substantially differ from those of systemic vessels (9) as well as from properties of pulmonary endothelial cells studied in tissue cultures (22, 23).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Preparation

The experiments were performed on 150- to 200-µm lung slices by means of the patch-clamp technique (24, 25). The procedure for animal decapitation was approved by the local veterinary authority (Regierungspräsidium Giessen). Rats aged 2 to 22 d were quickly decapitated and 1 ml of external solution containing 2% agar (cooled to 39°C) was injected into the trachea using a standard syringe. To prevent the outflow of the agar solution, the trachea was clamped with forceps for several seconds while the rat was immersed in ice-cold saline (3 to 4 min) bubbled with O₂/CO₂ (95%/5%). The lung was dissected and a small tissue block from a peripheral region of the upper lobe was cut out. The block was glued to the glass stage using cyanoacrylate glue and fixed in the chamber of the tissue slicer. The blade of the slicer was slowly moved forward with a speed of 3 mm/min. The slices were subsequently incubated in a standard solution at 37°C for at least 1 h.

Solutions

The standard external solution contained (in mM): NaCl 115, KCl 5.6, CaCl₂ 2, MgCl₂ 1, glucose 11, NaH₂PO₄ 1, and NaHCO₃ 25 (pH 7.4 when bubbled with 95% O₂/5% CO₂). External solutions with different [K⁺]_o were obtained from the standard solution by equimolarly replacing NaCl with KCl. In external low-Na⁺_o solution 115 mM NaCl was substituted by 115 mM tetramethylammonium-Cl. In low-Cl_o solution 115 mM NaCl and 5.6 mM KCl were replaced by 115 mM Na-aspartate and 5.6 mM K-aspartate. In experiments where the pH was varied, the external solutions were buffered with N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (Hepes)-NaOH and contained (in mM): NaCl 141, KCl 5.6, MgCl₂ 1, CaCl₂ 2, glucose 11, and Hepes 10 (pH 6.0, 6.8, 7.4, and 8.0 adjusted with NaOH). Standard internal solution (high-K⁺_i) contained (in mM): NaCl 5, KCl 144.4, MgCl₂ 1, ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid (EGTA) 3, and Hepes 10 (pH was adjusted to 7.3 by 10.6 mM KOH). In internal low-Cl_i solution 144.4 mM KCl was equimolarly substituted with K-aspartate. In high-Ca²_i solution EGTA was omitted and 1 mM CaCl₂ was added. All blockers and substances were directly added to the external or internal solutions. The temperature in all experiments was 21 to 23°C.

Current Recording

Patch pipettes were pulled from borosilicate glass tubes (GC 150; Clark Electromedical Instruments, Pangbourne, UK). The pipettes were fire-polished directly before the experiments and had a resistance of 3 to 7 M. The patch-clamp amplifiers were EPC-7 (List, Darmstadt, Germany) in all voltage- and current-clamp experiments. The effective corner frequency of the low-pass filter was 1 kHz. The frequency of digitization (except recordings in Figures 3B and 3C) was 2 kHz. The data were stored and analyzed with commercially available software (pCLAMP version 5.5.1; Axon Instruments, Foster City, CA). Offset potentials were nulled directly before formation of a seal. Junction potentials and voltage errors due to resistance in series were not corrected.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Ionic permeability of the membrane in SPAECs. (A) ER in SPAECs as a function of [K+]o for 2- (open symbols; five cells) and 22-d-old (filled symbols; seven cells) animals. The data points were fitted using the Goldman-Hodgkin-Katz equation (see MATERIALS AND METHODS) with [K+]i = 155 mM, [Cl]i = 162 mM, and [Cl]o = 126.6 mM giving PCl/PK ratios of 0.018 ± 0.003 and 0.11 ± 0.01 for 2- and 22-d-old rats, respectively. (B) Change in the membrane potential during external perfusion with low-Na+o and low-Clo solutions and with standard solution containing 0.1 mM DIDS in a 22-d-old rat. (C) Change in the ER in a SPAEC of a 22-d-old rat during intracellular perfusion with low-Cli solution. The moment at which the membrane was broken and whole-cell recording configuration was established (beginning of Cl depletion) is indicated by an arrow. The process of diffusion lasted about 5 min, instead of 20 to 30 s reported for the small round cells (35), suggesting ion diffusion through gap-junction channels into the neighboring cells within the endothelial syncytium. In B and C the frequency of digitization was 1 Hz.

The permeability ratio for K⁺ and Cl was calculated using the Goldman-Hodgkin-Katz equation:

The data were fitted using a linear or nonlinear least squares procedure. Numerical values are given as means ± standard error of the mean and fitting parameters as means ± standard error of the fit. In all figures the errors are indicated when exceeding the symbol size.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Anatomy

In the lung slices, the arteries were identified as the blood vessels accompanying a bronchus (Figure 1A; the cross section through the upward branch of the bronchus [b] is clearly seen). On a cross section, the bronchus was easily recognized according to the characteristic flimmering of the bronchial epithelium. The movement of the flimmering epithelium throughout the whole lung slice indicated a good general state of the preparation. The artery was further followed to the periphery until its diameter became 15 to 40 µm. These small vessels were considered as precapillary pulmonary arteries. Their size was about two to five times the diameter of erythrocytes (Figure 1B). On the longitudinal section through the vessel (Figure 1B), one can also recognize the artery wall (white arrows) and two typical enlargements of endothelial nuclei (black arrows). When filled with Lucifer Yellow-containing solution, the endothelial cells appeared fusiform with a length of 80 to 100 µm (Figure 1C). The mean maximum width determined in 40 unstained cells was 2.4 ± 0.08 µm.

View larger version (83K): [in this window] [in a new window]   Figure 1.   Pulmonary endothelial cells in a slice of rat lung. (A) Section of lung with bronchus (b), artery (a), and alveolus (al). The bronchus seen on the surface of the slice represents a dissected upward branch of the main bronchus (indicated by dashed lines) that accompanies the artery. The flimmering epithelium lining the bronchus was used for its identification. (B) Longitudinal section through an artery. White arrowheads indicate the inner artery wall. Black arrowheads show endothelial nuclei. Right, two erythrocytes are shown at the same magnification. (C) Endothelial cell with an attached patch pipette filled with internal solution containing 0.5% Lucifer Yellow. A dye leakage into the intercellular space, presumably through gap-junction channels, led to a remarkable staining of the artery wall. A strong staining of neighboring cells was not observed. Scale bars: A-C, 10 µm.

Experiments were performed on arteries occasionally dissected during slicing (at 0 to 30-degree angles to the vessel axis) and whose inner surface became, therefore, directly accessible for the patch pipette. The endothelial cells could be clearly distinguished from smooth-muscle cells because they did not generate detectable voltage-gated Ca²⁺ and K⁺ currents.

Resting Potential

The membrane resting potential (E_R) measured in small pulmonary artery endothelial cells (SPAECs) changed during postnatal development (Figure 2A). In SPAECs from 2-d-old rats, the E_R was 70.2 ± 1.5 mV (five cells), being close to the equilibrium potential for K⁺ ions (E_K) of 84 mV under our experimental conditions. The cells within a syncytium had an input resistance (R_IN) of 74.8 ± 1.7 M (n = 5; Figure 2B). The E_R increased during postnatal development, reaching the value of 42.6 ± 4.0 mV (five cells) at Day 22. This change in E_R was accompanied by a considerable lowering of R_IN (to 39.2 ± 7.9 M; six cells; Day 22). Although the lowering of R_IN observed for the cells within a syncytium can reflect an additional expression of both ionic and gap-junction conductances, an increase in the E_R is most likely to result from an age- dependent development of ionic conductance selective to ions other than K⁺.

View larger version (12K): [in this window] [in a new window]   Figure 2.   Change in ER and RIN during postnatal development. ER (A) and RIN (B) as a function of age. To determine RIN, the cell was voltage-clamped at 80 mV and 50-ms pulses were applied in depolarizing (to 20 mV; filled symbols) and hyperpolarizing (to 120 mV; open symbols) directions. The data points were connected by eye. Each point in A and B represents the mean of five to 21 measurements.

Under current-clamp conditions, the dependence of the E_R on external K⁺ concentration ([K⁺]_o) for the SPAECs of 2- and 16-d-old rats was investigated (Figure 3A). In cells of newborn rats, the E_R followed the values predicted by the Nernst equation for E_K (Figure 3A, solid line) at [K⁺]_o exceeding 5 mM (Figure 3A, open symbols). For older animals, a stronger deflection from the theoretical curve was observed at [K⁺]_o below 50 mM (Figure 3A, filled symbols). To study which ions besides K⁺ were involved in setting the E_R in endothelial cells of older animals, experiments were performed in which the concentrations of several ions were varied. Under our experimental conditions, the possible candidates were Ca²⁺ (E_Ca = +), Na⁺ (E_Na = +86 mV), and Cl (E_Cl = +5 mV). Removal of Ca²⁺ from external solution (10 cells) or raising internal Ca²⁺ to 1 mM (six cells) changed neither E_R nor R_IN, indicating that Ca²⁺ does not contribute to the resting conductance (not shown). An equimolar substitution of 115 mM external NaCl by tetramethylammonium-Cl had no effect on E_R (Figure 3B; four cells). In contrast, lowering external Cl led to a remarkable membrane depolarization (five cells) whereas external application of the Cl channel blocker 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS) hyperpolarized the membrane to about 80 mV (Figure 3B, eight cells). In the experiment shown in Figure 3C we measured the E_R in the cell using a pipette filled with an internal solution in which 144.4 mM KCl was equimolarly replaced with K-aspartate (low-Cl_i solution). The E_R of 46 mV was measured just after breaking through the membrane (Figure 3C, arrow). As the pipette's low-Cl_i solution diffused into the cell, the E_R hyperpolarized to about 80 mV (10 cells). This suggests that the membrane of SPAECs is permeable for K⁺ and Cl ions. By fitting the data points in Figure 3A with the Goldman-Hodgkin-Katz equation (Figure 3A, dashed lines) the permeability ratios for both ions, P_Cl/P_K, were calculated to be 0.018 ± 0.003 and 0.11 ± 0.01 for 2- and 22-d-old rats, respectively. Thus, the relative membrane permeability to Cl increased more than 6-fold during the first 22 d of postnatal development.

Membrane Conductance

Ionic conductances contributing to the E_R were studied in the voltage-clamp mode. The cells were clamped at a holding potential of 80 mV and 30- or 50-ms voltage steps were applied in both hyperpolarizing and depolarizing directions (Figure 4A). The current amplitudes were measured at the end of the pulse and plotted as a function of voltage (Figure 4B). In spite of the difference in E_R, the current-voltage relationships were linear for the SPAECs of both newborn and older animals, suggesting that the cells did not possess detectable voltage-dependent currents. Varying external Ca²⁺ (0 to 2 mM) did not change the shape of the current-voltage curve, suggesting a negligible contribution of Ca²⁺ and Ca²⁺-activated conductances. Several blockers were tested so as to describe the pharmacologic profile of the conductance. Of all substances tested, only 10 mM Ba²⁺ suppressed the currents (Figure 4C). In the current-clamp mode, Ba²⁺ depolarized the membrane by 10 to 15 mV (five cells). No remarkable changes of the current amplitude were observed in the presence of 20 mM tetraethylammonium (TEA) (n = 6; Figure 4D), 1 mM Cs⁺, 300 µM Zn²⁺, 0 to 5 mM Mg²⁺, 1 mM bupivacaine, 10 mM lidocaine, 5 mM 4-aminopyridine, 1 µM apamin, 100 nM charybdotoxin, 100 nM iberiotoxin, 10 µM glibenclamid, 100 µM quinine, 300 µM quinidine, and 0.1 mg/ml iloprost (a stable analogue of prostacyclin). Internally applied 1 mM Ca²⁺ and 3 mM adenosine triphosphate (ATP) also had no effect on the membrane current.

View larger version (21K): [in this window] [in a new window]   Figure 4.   Membrane conductance in SPAECs. (A) Voltage-clamp recordings of membrane currents in cells of 2- and 16-d-old rats. The cells were held at 80 mV and 30- or 50-ms voltage pulses were applied in depolarizing (to 40, 0, and +40 mV) and hyperpolarizing directions (to 120 mV). The traces show the absolute current. The level of zero current recorded at the resting potential of 69 mV for a 2-d-old rat and 51 mV for a 16-d-old rat is indicated by the dashed line. The current amplitudes were measured at the end of the test pulse. (B) Current-voltage relationship for membrane currents in cells of 2- (filled triangles) and 16-d-old (open circles) animals. The data points were fitted with straight lines. (C) Effect of 10 mM Ba2+ on membrane currents. Suppression of membrane currents was accompanied by membrane depolarization. (D) Lack of blocking effect in the presence of 20 mM TEA.

Effect of pH

The sensitivity of membrane current to changes in external pH was also studied. In seven cells, increasing pH from 7.4 to 8.0 resulted in augmentation of membrane conductance, whereas lowering pH to 6.8 and 6.0 reduced it (Figures 5A and 5B). In current-clamp experiments, the cells responded to acidosis (pH 6.0) with a 15- to 20-mV depolarization and to alkalosis (8.0) with about 5 mV hyperpolarization (Figure 5C).

View larger version (19K): [in this window] [in a new window]   Figure 5.   pH sensitivity of the membrane current and resting potential. (A) Membrane current monitored by 50-ms-voltage pulses from 80 to 40 or 120 mV in external solutions with different pH values. (B) Current-voltage relationships for membrane currents in control (pH 7.4) and in pH 6.0 solutions. (C) pH dependence of the resting membrane potential. Each point represents the mean of five to eight measurements.

Cell-to-Cell Coupling

The present method of recording from intact pulmonary endothelium in tissue slices has made it possible to estimate the degree of electrical coupling between neighboring cells in the direction perpendicular to the longitudinal axis of the artery. These estimations were done using a simultaneous recording with two electrodes (Figure 6A). The first electrode held the membrane potential of one cell at 80 mV in the voltage-clamp mode and applied test voltage pulses of different amplitudes (Figure 6B). With a second electrode a current-clamp recording of the membrane potential in another cell was performed. Because the intact cells had a very small mean width of 2.4 µm and a typical fusiform shape, it was very difficult (for both mechanical and optical reasons) to obtain stable recordings from two neighboring cells. Therefore, we used another method of estimation by recording from two arbitrary cells and plotting the transduction coefficient (k) as a function of the distance between cells (L).

View larger version (28K): [in this window] [in a new window]   Figure 6.   Estimation of cell-to-cell coupling. (A) A scheme of two-electrode measurements. L is the distance between the tips of two pipettes. (B) Voltage recordings in two-electrode experiments. The voltage commands applied to the first electrode are shown on the left (L = 0). The next three families of curves are current-clamp recordings of membrane potentials in three cells, located 2.5, 11, and 28 µm away from the first cell. (C) k as a function of the distance between the cells. Values of k for depolarizing and hyperpolarizing pulses are given by squares and circles, respectively. The data points are fitted using an exponential function given in RESULTS. The upper scale shows the distance in cell widths; 1 unit represents 2.4 µm. The value of k at 1 cell width (2.4 µm), equal to 0.83, corresponds to the transduction coefficient between two neighboring cells.

These experiments were performed using 18- to 22-d-old rats (four animals) with an original E_R of about 42 to 45 mV (Figure 6B, dashed line at 43 mV). The leftmost family of recordings shows the test pulses applied to the first (voltage-clamped) cell (L = 0), held at a potential of 80 mV. The three families of curves on the right are responses of three different current-clamped cells located 2.5, 11, and 28 µm away from the first cell. One can see that the E_Rs in the current-clamped cells are electrotonically influenced by a relatively negative (80 mV) potential of the first cell: the smaller the L, the more negative the membrane potential was measured in the second cell. The amplitude of the membrane response of the current-clamped cells on the voltage pulse applied to the first cell decreased with L. k, determined as a ratio between the amplitudes of cell response [E(L)] and test pulse [E(0)], for negative and positive test pulses is presented in Figure 6C (circles and squares, respectively) as function of L. Assuming that k for L = 0 is 1 (potential at the point of fixation is equal to the test potential) the data were fitted with equation: k = exp[(L/L₀)], giving an electrical length constant L₀ of 13.3 ± 0.8 µm. This equation with L₀ set to 13.3 µm could be further used for estimating the direct transduction between two neighboring cells. Assuming the distance between neighboring cells equal to the cell width of 2.4 µm (see ANATOMY, earlier) the k value was calculated to be 0.83 (Figure 6C; the upper scale shows the distance expressed in cell widths, where 1 U = 2.4 µm).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The membrane of SPAECs was permeable for K⁺ and Cl ions. The resting potential in SPAECs of newborn animals is predominantly determined by a voltage- and TEA- insensitive K⁺ conductance distinct from Ca²⁺-activated, ATP-sensitive, or inward-rectifier conductances. Its voltage-insensitivity and pharmacologic properties, including suppression by external Ba²⁺ and acidosis, are most consistent with those of two pore-domain K⁺ channels (26- 28) highly expressed in the lung tissue (26, 28). Our findings differ from previous reports describing Ca²⁺-activated (11), ATP-sensitive (18), and stretch-activated (21) as well as inward rectifying K⁺ channels (22) in several endothelial preparations. It should be noted, however, that none of these studies was performed on native endothelial cells of small pulmonary arteries. Most of them were carried out on cell lines of umbilical vein endothelium (12, 13, 15, 21), freshly dissociated or cultured systemic artery endothelial cells (11, 14, 18, 19), and coronary capillary fragments (20). There are few studies of cultured pulmonary artery endothelial cells showing nonselective cation channels (23) and inwardly rectifying K⁺ channels sensitive to Ba⁺ and Cs⁺ (22) as well as lack of the endogenous large-conductance Ca²⁺-activated K⁺ channels. Although the last finding is in good agreement with our observations, the principal discrepancy of the membrane properties reported is likely to be due to the difference in either the vessel caliber or the preparation techniques (culture versus fresh tissue slice). We assume that the small pulmonary vascular endothelial cells have their unique electrophysiologic properties.

The membrane permeability to Cl increased during the first 22 d of the postnatal development from P_Cl/P_K = 0.018 to 0.11, leading to an increase in E_R from 70 to about 45 mV. This increase in Cl conductance during the first postnatal weeks associated with growth and maturation could be triggered by the dramatic increase in pulmonary blood flow, which has been shown to affect endothelial chloride conductance (3). A postnatal development of Cl permeability can be important for endothelial cell function inasmuch as the resulting membrane depolarization has influence on the intracellular Ca²⁺ concentration and therefore on Ca²⁺-dependent functions such as release of NO (8).

The acidosis-induced membrane depolarization described here could be relevant for endothelial cell function during disturbances of the acid-base status and may give a possible explanation for the pH sensitivity of the hypoxic pulmonary vasoconstriction (HPV) (29). The main mechanisms underlying HPV are considered to be located in the pulmonary artery smooth-muscle cells (32). The endothelial production of NO, however, modulates HPV (33). As the NO production in endothelial cells is reduced by membrane depolarization (8, 9), the acidosis-induced depolarization of the endothelial membrane may explain the increased hypoxia sensitivity of pulmonary arteries during acidosis.

The endothelial cells in small pulmonary arteries showed tight electrical coupling with each other. The k = 0.83 estimated for two neighboring cells in intact endothelium was close to the value of 0.8 measured in isolated paired guinea-pig heart cells (34). We have also succeeded in estimating the electrotonic length constant in the direction perpendicular to the longitudinal axis of the artery. The length constant of 13.3 µm indicates that most endothelial cells lying in one cross section of a small pulmonary artery with a diameter of 20 to 30 µm are electrically well coupled. There are some reasons to expect that the electrical length constant for the longitudinal direction of the vessel is several times larger. The theoretically calculated electrotonic length constant for one endothelial cell (considered as a capillary of 0.5 µm diameter) is about 1,000 µm (2), indicating that no electrical signal dissipation can be expected within one endothelial cell of 80 to 100 µm length. Therefore, the electrical transduction in the arteries in both longitudinal and perpendicular directions should be limited predominantly by the number of cell-to-cell transduction steps. On this basis, one can assume that the electrical length constants in the longitudinal direction, in which the longer axes of the cells are oriented, will be several times greater than the 13.3 µm measured in our experiments.

In conclusion, the slice preparation allowed the examination of intact endothelial cells in visually identified small pulmonary arteries. The electrophysiologic properties and their maturational changes could have significant impact on the physiologic function of the precapillary pulmonary arteries. The present method can further be used for studying the specific properties of the pulmonary vasculature.