Divalent ion block of inward rectifier current in human capillary endothelial cells and effects on resting membrane potential

Abstract

1. Cultured human capillary endothelial cells (HCEC) contain a large inward rectifier current, I_K(IR), that can be abolished by removing external K⁺ or by adding 50 μm Ba²⁺.

2. We show that I_K(IR) is responsible for maintaining the hyperpolarized potential (−60.6 ± 0.5 mV, n = 83) of HCEC. Blocking I_K(IR) with 50 μm Ba²⁺ shifts the zero current level and depolarizes HCEC by 36.5 ± 1.3 mV (n = 4).

3. Increasing external Ca²⁺ concentration ([Ca²⁺]_o) from 0.5 to 7 mm reduces the magnitude of I_K(IR) by 36.5 ± 2.3 % (n = 5) and depolarizes the cells by 10.33 ± 2.4 mV (n = 3), whereas decreasing [Ca²⁺]_o from 1.8 to 0.5 mm increases the amplitude of I_K(IR) by 6.9 ± 1.9 % (n = 4). The relationship between [Ca²⁺]_o and the percentage block of I_K(IR) gives a K_d value of 5.4 ± 0.6 mm at −120 mV.

4. I_K(IR) is also blocked by other divalent ions, with Ba²⁺ >> Sr²⁺ > Mg²⁺ > Mn²⁺= Ca²⁺, and the block of peak current at −120 mV being 85.3 ± 3.2 % (n = 5) for 50 μm Ba²⁺, 62.9 ± 2.2 % (n = 5) for 5 mm Sr²⁺, 40.7 ± 2.5 % (n = 9) for 5 mm Mg²⁺, 33.4 ± 2.1 % (n = 5) for 5 mm Mn²⁺ and 32.9 ± 2.1 % (n = 5) for 5 mm Ca²⁺.

5. The voltage dependence of Sr²⁺ block of peak I_K(IR) occurred with a K_d value of 1.0 ± 0.09 mm for −140 mV, 1.9 ± 0.16 mm for −130 mV, 3.1 ± 0.28 mm for −120 mV, 4.6 ± 0.34 mm for −110 mV and 6.4 ± 0.5 mm for −100 mV (n = 5), with a calculated electrical distance (δ) of 0.44 from the outside.

Endothelial cells are not electrically excitable yet their ionic channels have been postulated to play an important role in the modulation of intracellular calcium and subsequent release of endothelial factors such as nitric oxide (NO), endothelin and prostacyclin (Adams et al. 1989). To date there have been several studies characterizing a variety of ion channels such as receptor-operated non-selective cation channels (Johns et al. 1987; Nilius et al. 1993), K⁺ and cation channels activated by shear stress (Olesen et al. 1988; Schwarz et al. 1992), Ca²⁺-dependent potassium channels (Sauve et al. 1988; Demirel et al. 1994), non-selective cation channels (Fichtner et al. 1987; Inazu et al. 1994) and inward rectifier K⁺ channels in bovine and porcine aortic as well as human endothelial cells (Colden-Stanfield et al. 1987; Johns et al. 1987; Takeda et al. 1987; Silver & DeCoursey, 1990; Jow et al. 1996; for review, see Nilius et al. 1997). However, the physiological roles of these channels are only beginning to be understood.

The inward rectifier channel is the dominant ion channel in whole-cell studies of endothelial cells (Himmel et al. 1994) and has been postulated to play an important role in determining the resting membrane potential in endothelial cells (Campbell et al. 1991; Voets et al. 1996). Resting membrane potential is important in determining the driving force for Ca²⁺ entry and in establishing the gradient for other ion channels (Voets et al. 1996). Wellman & Bevan (1995) suggest that shear stress hyperpolarizes endothelial cells by opening the inward rectifier K⁺ channel, K_IR, thereby increasing the driving force for Ca²⁺ entry and stimulating NO release and relaxation of rabbit cerebral arteries. Membrane depolarization initiates the formation of H₂O₂ and possibly other oxidants in bovine pulmonary arterial endothelial cells (al-Mehdi et al. 1996) and inhibits endothelial relaxing factor (EDRF)/NO synthesis by inhibiting the Ca²⁺ influx pathway (Luckhoff & Busse, 1990; Busse et al. 1991).

One of the aims of the present study was to quantify the contribution of the inward rectifier current to the resting membrane potential of human capillary endothelial cells. The second goal was to explore the effects of extracellular Ca²⁺ on the inward rectifier current and to compare the blocking action of other divalent ions such as Ba²⁺, Sr²⁺, Mg²⁺ and Mn²⁺. We have chosen human capillary endothelial cells to examine these issues as they are clearly relevant to human physiology and we have a good understanding of the ion channels present in them (Jow et al. 1996).

METHODS

Cell culture

Human capillary endothelial cells (HCEC) were obtained from Cell Systems Corporation (WA, USA), and were cultured in growth medium for human capillary endothelial cells (4M1-500, Cell Systems Corporation). HCEC were fed every two days and passaged between 7 and 15 days or at 80–90 % confluence. The HCEC used for this study were from subcultures two to six.

Electrophysiological measurements

HCEC were recorded in whole-cell voltage clamp with the patch-clamp technique (Hamill et al. 1981). Patch electrodes were pulled from borosilicate glass (outer diameter, 1.5 mm; inner diameter, 1 mm; Sutter Instrument Co., CA, USA) and fire polished using an MF-83 Microforge (Narishigi Inc., Tokyo, Japan), and had resistances of 3–4 MΩ. We found that fire polishing was necessary in order to obtain the very high > 30 GΩ seal resistances required to observe the negative resting membrane potential of these cells. The input resistance of cells at −60 mV measured near 10 GΩ and seal resistance in cell-attached modes was greater than 30 GΩ. Membrane potentials were corrected for liquid junction potentials (< 5 mV). The resting membrane potential of single endothelial cells averaged −60.6 ± 0.5 mV (n = 83). The mean cell capacitance was 51.4 ± 6.9 pF (n = 34). All recordings were done at room temperature (25 ± 3°C). An EPC9 amplifier with the acquisition program Pulse-PulseFit (HEKA, Lambrecht, Germany) was used for recording and data acquisition, with Igor Pro (WaveMetrics, Inc., Lake Oswego, OR, USA) and Origin (Microcal Software, Inc., Northampton, MA, USA) used for analysis. Currents were filtered at 3 kHz. All results are expressed as mean values ±s.e.m. The standard pipette solution for whole-cell recordings was (in mmol l⁻¹): 120 potassium aspartate, 20 KCl, 0.5 EGTA, 5 Hepes, 5 phosphocreatine (disodium salt) and 1 MgCl₂ (pH 7.2 with KOH). In some experiments 5 mm MgATP and 10 mm EGTA were included in the pipette solution. We saw no difference in the current with or without MgATP and with high or low EGTA. The standard bath recording solution consisted of (in mmol l⁻¹): 140 NaCl, 5.4 KCl, 5 Hepes, 15 dextrose, 1.8 CaCl₂, 0.4 KH₂PO₄ and 0.3 Na₂HPO₄ (pH 7.2 with NaOH). Additional Ca²⁺, Mg²⁺, Mn²⁺ and Sr²⁺ were added to the standard bath solution by equimolar replacement of NaCl.

RESULTS

Ba²⁺-sensitive inward rectifier K⁺ current

In whole-cell patch recordings, the mean resting membrane potentials of human capillary endothelial cells were tightly grouped around −60.6 ± 0.5 mV, although more depolarized potentials were seen when the seal resistance of the patch electrode was less than 30 GΩ. A large inward current elicited by hyperpolarizing voltage steps was exhibited in 92 % (n = 118) of HCEC. During hyperpolarizing pulses negative to −100 mV, the inward current activated rapidly and decayed with time (Fig. 1A). The current-voltage relationship in Fig. 1E shows a strong inward rectification. The mean peak current amplitude of the inward current at −120 mV was −360.8 ± 49.2 pA or 8.4 ± 0.87 pA pF⁻¹ (n = 38). Application of Ba²⁺ (50 μm) to the outside of the cell inhibited both the peak and the steady-state inward current at −120 mV, by 85.3 ± 3.1 and 87.5 ± 2.3 %, respectively (n =6, Fig. 1A and B). Removing external K⁺ completely eliminated the inward current, which confirms that this inward current is carried by external K⁺ (Fig. 1D and E). From these data, we conclude that the inward rectifier K⁺ current (I_K(IR)) is the dominant inward current in human capillary endothelial cells. Figure 1 also demonstrates the depolarizing shift in the zero current level seen upon block of I_K(IR).

Figure 1. Ba²⁺ block and dependence on external K⁺ of inward rectifier K⁺ current in HCEC.

A, current traces from a HCEC recorded in response to voltage step pulses from −120 to +10 mV in 20 mV intervals at a holding potential of −60 mV. B, current traces from the same voltage pulse protocol after application of 50 μm Ba²⁺. C, current-voltage plot of peak (•) and steady-state (▪) current in standard bath solution, and peak (○) and steady-state (□) current after application of Ba²⁺ for the same cell. D, current traces from a HCEC response to voltage steps of −120 mV to +10 mV from a holding potential of −60 mV with 5.4 mm K⁺ in the standard bath solution. E, current traces from the same voltage protocol in the absence of external K⁺. F, peak current-voltage plot from the currents in D and E before (•) and after (○) removing all external K⁺. Elimination of external K⁺ completely blocks the inward current.

I_K(IR) is blocked by external Ca²⁺

Figure 2A shows typical I_K(IR) traces obtained from HCEC bathed in standard external bath solution containing 1 mm Ca²⁺. Application of the standard external solution with elevated [Ca²⁺] (9 mm) decreased I_K(IR) by 42 % in this cell (Fig. 2B). The residual I_K(IR) was further blocked by 50 μm Ba²⁺ indicating that Ca²⁺ block is less potent than Ba²⁺ block (Fig. 2C). Figure 2D shows the current traces after washout of external Ca²⁺ and Ba²⁺ from the bath with standard external solution containing 1 mm Ca²⁺. Figure 2E shows the current-voltage plot of the currents in 1 mm Ca²⁺, 9 mm Ca²⁺, 50 μm Ba²⁺ and after washout (< 5 min) of external Ca²⁺ and Ba²⁺. This result indicates that the reduction in I_K(IR) caused by external Ca²⁺ and Ba²⁺ is completely reversible.

Figure 2. Ca²⁺ block of I_K(IR) is reversible.

A, inward rectifier current traces of a HCEC bathed in standard bath solution. The cell was held at −60 mV and step potentials were applied from −120 to +20 mV, in 20 mV steps. B, the same cell after superfusion with standard bath solution containing 9 mm Ca²⁺. C, addition of 50 μm Ba²⁺ blocked the remaining current. D, current traces taken after washout of high external Ca²⁺ and Ba²⁺ with standard bath solution. E, current-voltage plots for control, 9 mm Ca²⁺, 50 μm Ba²⁺ and after washout.

Voltage dependence of I_K(IR) and effect of external Ca²⁺

The voltage dependence of I_K(IR) was examined by applying hyperpolarization voltage pulses to potentials more negative than −100 mV. The current shows a biphasic time course with very fast activation that then relaxes to a steady-state level (Fig. 3A). The peak inward current (Fig. 3B, ▪) increased with more negative test potentials, whereas the steady-state current (□) reached a maximum at −120 mV and then declined at more hyperpolarizing potentials. Elevation of [Ca²⁺]_o from the normal 1 mm to 10 mm reduced the magnitude of I_K(IR). Figure 3A shows paired current traces for the same hyperpolarizing voltage pulses before and after increasing [Ca²⁺]_o from 1 to 10 mm. The mean reduction of I_K(IR) peak current at −120 mV was 46.9 ± 1.9 % (n = 5). The reduction in inward rectifier current by [Ca²⁺]_o did not appear to alter the kinetic properties of the current or the rectification of the sustained current at hyperpolarizing potentials.

Figure 3. Block of I_K(IR) by elevated external Ca²⁺.

A, superimposed current traces of the cell bathed in standard bath solution with 1 mm Ca²⁺ and after application of bath solution containing 10 mm Ca²⁺. The current traces were elicited from −150 to −100 mV with a holding potential of −60 mV and pulses of 500 ms duration. B, current-voltage plot of peak current (▪) and steady-state current (□) for control, and peak current (•) and steady-state current (○) after application of bath solution containing 10 mm Ca²⁺.

The external Ca²⁺ concentration-dependent block of I_K(IR) was assessed by measuring the decrease in peak current at −120 mV with various external Ca²⁺ concentrations (Fig. 4). Each cell was exposed to external solutions with various [Ca²⁺]_o from 0.5 to 30 mm. The percentage decrease in current was then plotted against [Ca²⁺]_o (Fig. 4A). The mean data are fitted by the logistic equation:

where I_max is the maximum current block, K_d is the concentration at half-block, and p is the slope of the dose-response curve or the Hill coefficient. This gives a K_d of 5.4 ± 0.6 mm and a p value of 1.07 ± 0.08 (n = 5, −120 mV) for Ca²⁺ block of peak I_K(IR). Figure 4B shows the current traces of a HCEC at −120 mV with various external Ca²⁺ concentrations.

Figure 4. Inhibition of I_K(IR) by a range of external Ca²⁺ concentrations.

A, effect of [Ca²⁺]_o from 0.5 to 30 mm on the mean percentage block of peak inward rectifier current at −120 mV (n = 5). Inhibition of the current at different concentrations of external Ca²⁺ is well fitted by a logistic equation with an EC₅₀ value of 5.4 ± 0.6 mm. B, representative current traces of a HCEC at −120 mV with [Ca²⁺]_o from 0 to 9 mm. C, shifts in zero current level of a HCEC exposed to divalent cations. The cell was held at −60 mV and voltage ramps were applied from −120 to +20 mV. The zero current level occurred at −62 mV with 0.5 mm Ca²⁺, −47 mV with 7 mm Ca²⁺, and −18 mV with 50 μm Ba²⁺ in the bath solution.

The membrane potential is modulated by external Ca²⁺

Elevations of external Ca²⁺ concentration decreased I_K(IR) in all the endothelial cells examined (n = 34). I_K(IR) has been shown by Voets et al. (1996) to be a primary component in establishing the resting membrane potential of bovine pulmonary artery endothelial cells. Therefore, block of I_K(IR) in HCEC should shift the resting membrane potential. Indeed, we found that HCEC membrane potential depolarized as [Ca²⁺]_o was increased from 0.5 to 7 mm. The reversal potential was determined in voltage clamp by noting the zero current level from averages of voltage ramps (−120 to +100 mV). When external Ca²⁺ was increased from 0.5 to 7 mm, the reversal potential shifted 10.33 ± 2.4 mV (n = 3) positive from the initial reversal potential of −59.7 ± 1.8 mV (n = 3). Ba²⁺ at 50 μm is a more potent blocker of the inward rectifier current and, not surprisingly, depolarized the endothelial cells to a much greater extent. Ba²⁺ depolarized the cells by 36.5 ± 1.3 mV from a resting potential of −59 ± 1.1 mV (n = 4). Figure 4C shows data from a single HCEC with the shifts in zero current level after application of 7 mm Ca²⁺ or 50 μm Ba²⁺.

Figure 5 shows the effects of 5 mm Mn²⁺ or Mg²⁺ on I_K(IR). Mn²⁺ (5 mm) in the presence of 1.8 mm external Ca²⁺ blocked 33.4 ± 2.1 % of I_K(IR) in five cells tested. Mg²⁺ (5 mm) with 1.8 mm external Ca²⁺ blocked I_K(IR) by 40.7 ± 2.5 % in nine cells tested. This block did not change the kinetics of the I_K(IR) current.

Figure 5. External Mn²⁺ and Mg²⁺ block of HCEC I_K(IR).

A, current traces from a HCEC in Mn²⁺-free standard bath solution. The cell was held at −60 mV and pulsed from −120 mV to +20 mV in 20 mV steps. B, current traces after switching to a bath solution containing 5 mm Mn²⁺. C, the current-voltage plot of the same cell in standard bath solution and in bath solution containing 5 mm Mn²⁺. D, current traces (same voltage protocol as A) of another HCEC in standard bath solution without Mg²⁺. E, current traces after application of 5 mm Mg²⁺ in the standard bath solution to the same HCEC. F, plot of the current-voltage relationship of the HCEC before and after application of Mg²⁺.

Sr²⁺ produces a voltage-dependent block of I_K(IR)

We observed a profound block of I_K(IR) by Sr²⁺ relative to Ca²⁺, Mg²⁺ and Mn²⁺. The degree of block by 5 mm Sr²⁺ was of roughly the same magnitude as that seen with 50 μm Ba²⁺. Sr²⁺ at 5 mm (−120 mV) blocked 63 ± 2.2 % of peak I_K(IR) in five cells compared with 85.3 ± 3.2 % for 50 μm Ba²⁺ in six cells (−120 mV). Figure 6 shows the effect of 5 mm Sr²⁺ on both peak and steady-state I_K(IR) in HCEC. Sr²⁺ blocked the two components of I_K(IR) to comparable levels (63 ± 2.2 % of peak current and 68 ± 2.1 % of sustained current at −120 mV), with increased block occurring at more negative potentials. Figure 7 further investigates the voltage dependence of this Sr²⁺ block. Figure 7A shows dose-response curves for Sr²⁺ block of fractional peak inward rectifier current for voltages ranging from −140 to −110 mV. The EC₅₀ values were: 1.0 ± 0.09 mm for −140 mV, 1.9 ± 0.16 mm for −130 mV, 3.1 ± 0.28 mm for −120 mV, and 4.6 ± 0.34 mm for −110 mV (n = 5).

Figure 6. Sr²⁺ blocks I_K(IR) in a voltage-dependent manner.

A, I_K(IR) traces during strong hyperpolarizing pulses (−180 to +20 mV) from a holding potential of −60 mV from a HCEC perfused with standard external solution. B, current traces from the same cell in the presence of 5 mm Sr²⁺. Both the transient and steady-state components of the inward current were blocked. C, current-voltage plot for the peak current (▪) and the current at the end of the pulse (•) in standard bath solution, and the peak current (□) and the steady-state current (○) in 5 mm Sr²⁺.

Figure 7. EC₅₀ values for block of I_K(IR) by Sr²⁺ over a range of test potentials (−140 to −100 mV).

A, fractional inhibition of peak inward current vs. external Sr²⁺ concentration at −140 mV (▪), −130 mV (•), −120 mV (▴), −110 mV (▾) and −100 mV (♦). I_Con is the peak current in standard bath solution and I_Sr is the current in the presence of 0.05–10 mm Sr²⁺. The voltage dependence of the Sr²⁺ block was well fitted by the logistic equation. The inset shows a plot of EC₅₀ values vs. test potential for the same cells. The data points were well fitted with a linear regression yielding a slope such that an e-fold change in EC₅₀ occurs over a 27 mV range. B, percentage block of peak I_K(IR)vs. the test potentials for 10 mm Ca²⁺ (▪), 5 mm Mg²⁺ (▴), 5 mm Mn²⁺ (▾) and 5 mm Sr²⁺ (•). C, percentage block of steady-state I_K(IR) by 10 mm Ca²⁺ (▪), 5 mm Mg²⁺ (▴), 5 mm Mn²⁺ (▾) and 5 mm Sr²⁺ (•) for the various test potentials shown.

The Sr²⁺ dissociation constant at a given test potential is described as:

where δ is the fractional electrical distance between the external mouth of the channel and the Sr²⁺ binding site, z is the valency of the blocking ion (Sr²⁺), and V, F, R and T are the potential across the membrane, the Faraday constant, the gas constant and the absolute temperature, respectively. The δ value was calculated using the above equation (Woodhull, 1973) and the slope value of the K_dvs. voltage is shown in the inset of Fig. 7A. The value of δ is 0.44, as measured from the outside, such that Sr²⁺ senses 44 % of the membrane field in reaching its blocking site in physiologically normal internal and external solutions. Alternatively, one can describe this as an e-fold change in EC₅₀ over a 27 mV range. The voltage dependences for the other divalent cations examined are quantified in Fig. 7B and C. Sr²⁺ block of I_K(IR) was highly voltage dependent for both peak and steady-state I_K(IR), suggesting that the ion blocks within the voltage field. Ca²⁺, Mn²⁺ and Mg²⁺ were all largely voltage independent in their block of I_K(IR), although Ca²⁺ showed some weak voltage dependence at the more positive potentials of −80 and −90 mV. These ions are therefore presumably blocking from the outside of the channel.

DISCUSSION

In the present study we have shown that human capillary endothelial cells contain a large inward rectifier current whose block by divalent cations produces significant depolarizations in the HCEC resting membrane potential. This depolarization by divalent cations would be expected to reduce the driving force for agonist-induced Ca²⁺ entry, and as such may play a physiological role in pathological situations in which the concentrations of divalent ions such as Ca²⁺ and Mg²⁺ may be varying, as in hypo- or hyperparathyroidism (Elin et al. 1990), kidney failure (Henning et al. 1968), or treatment with diuretics (Dyckner & Wester, 1984). Modulation of agonist-induced Ca²⁺ entry would alter the endothelial cell sensitivity to the agonists with respect to the Ca²⁺-dependent signals they mediate, such as NO release and intracellular Ca²⁺ release (Adams et al. 1989). Divalent ion modulation of the level of inward rectifier current would also be expected to have effects upon the amount of Ca²⁺ influx elicited by growth factors such as vascular endothelial growth factor (VEGF) that trigger sustained Ca²⁺ entry (Bates & Curry, 1997).

The peak inward rectifier current (I_K(IR)) in these HCEC increases with hyperpolarizing potentials; however, the steady-state current reaches a maximum at about −120 to −130 mV and decreases with more hyperpolarizing potentials (Fig. 3B and 6C). This voltage dependence of I_K(IR) has also been demonstrated in guinea-pig ventricular myocytes (Biermans et al. 1987) and rabbit osteoclasts (Hammerland et al. 1994). The reduction of the steady-state current with extreme hyperpolarizations has been shown to be largely due to block by external Na⁺ (Standen & Stanfield, 1979; Biermans et al. 1987; Voets et al. 1996). In agreement with this we found that in experiments in which external Na⁺ was replaced with NMDG⁺, the voltage-dependent block of I_K(IR) at negative potentials was largely removed (not shown). The present studies were conducted with normal extracellular Na⁺ and K⁺ as we are interested in the physiological role of divalent ion block of I_K(IR). The Na⁺ block of I_K(IR) only occurs at potentials negative to −120 mV and would therefore not be expected to have a large physiological effect on I_K(IR).

The inward rectifier current of these human capillary endothelial cells is blocked by low concentrations (50 μm) of Ba²⁺. Ba²⁺ is known to block inward rectifier K⁺ current at low concentrations in a variety of preparations (Hagiwara et al. 1978; Standen & Standfield, 1978; Kubo et al. 1993). Voets et al. (1996) showed that block of the inward rectifier current by 100 μm Ba²⁺ in bovine pulmonary artery endothelial cells greatly depolarized these cells. We also show that external Ca²⁺ as well as Ba²⁺ depolarize human endothelial cells by blocking I_K(IR). Several groups have reported external Ca²⁺ block of I_K(IR) in guinea-pig cardiac myocytes (Biermans et al. 1987; Shioya et al. 1993), in rabbit osteoclasts (Hammerland et al. 1994), and in rat coronary artery smooth muscle cells (Robertson et al. 1996). However, depolarization by external Ca²⁺ has not been studied in any of these reports.

External divalent cations such as Mg²⁺ and Sr²⁺ reduce the unitary current amplitude of inward rectifier channels in guinea-pig cardiac myocytes, as reported by Shioya et al. (1993). An external Mg²⁺ concentration of 5 mm blocked inward rectifier current by 41.2 % at −80 mV in rat coronary artery smooth muscle cells (Robertson et al. 1996), and 4 mm Mg²⁺ reduced the steady-state I_K(IR) by 85 % at −120 mV in guinea-pig ventricular myocytes (Biermans et al. 1987). In HCEC, both 5 mm Mg²⁺ and 5 mm Mn²⁺ reduced I_K(IR), by 40.7 ± 2.5 % (n = 9) and 33.4 ± 2.1 % (n = 5) at −120 mV, respectively (Fig. 5).

Standen & Stanfield (1978) have previously reported on the block of inward rectifier current in frog skeletal muscle fibres by 10 mm Sr²⁺. They describe the voltage-dependent Sr²⁺ block of the steady-state inward rectifier current as being similar to that of Ba²⁺; in contrast to us they do not see any voltage dependence to the block of the peak inward current in the presence of Sr²⁺. This difference is unlikely to be due to the Sr²⁺ concentration used as we see our effect over a range of Sr²⁺ concentrations (50 μm to 10 mm). We do, however, use physiologically normal 5.4 mm extracellular K⁺ in our bath solution, whereas they use 115 mm K⁺. It is possible that the elevated K⁺ slows the blockade of the channel by Sr²⁺ such that the voltage-dependent block of the peak current is not observed but the block is seen at the end of the pulse. However, they also do not observe a block by extracellular Ca²⁺ in their studies, whereas we do. It therefore seems more likely that there are different inward rectifier channels in these two preparations. Sequence alignment of cloned inward rectifier channels overall shows a high degree of homology in the putative pore-forming region; however, differences do occur and some of these may account for the varying results seen in the literature for this class of channel.

We demonstrate in this study the block of human capillary endothelial inward rectifier current by Ba²⁺, Sr²⁺, Ca²⁺, Mg²⁺ and Mn²⁺. We also quantitatively show the role of I_K(IR) in establishing the resting membrane potential of these human capillary endothelial cells. The block by Sr²⁺ was highly voltage dependent for both the peak and steady-state current. Fits of the data in Fig. 7 show an e-fold change in current blocked over a 27 mV range. This translates into a δ value of 0.44 for Sr²⁺. The strong voltage dependence of Sr²⁺ block relative to Ca²⁺, Mn²⁺ and Mg²⁺ suggests that Sr²⁺ may be binding to a site within the voltage field whereas Ca²⁺, Mg²⁺ and Mn²⁺ appear to block outside the pore. This has been suggested by others for Mg²⁺ and Ca²⁺ block relative to Cs⁺ and Sr²⁺ (Shioya et al. 1993).