Expression of the Inwardly Rectifying K⁺ Channel Kir2.1 in Native Bovine Corneal Endothelial Cells

Abstract

purpose. To determine the presence of Kir2.1 channels in native bovine corneal endothelial (BCE) cells and assess their contribution to the resting membrane potential.

methods. RT-PCR and Western blot analysis were used to detect the expression of Kir2.1 mRNA and protein in native BCE cells. Whole-cell patch-clamp recording was used to characterize Kir2.1 currents in freshly isolated, single BCE cells, as well as in BCE cell clusters. The contribution of Kir2.1 channels to the membrane potential (V _m) was assessed by whole-cell recording in the zero-current clamp mode in the absence and presence of Ba²⁺.

results. RT-PCR analysis confirmed that Kir2.1 was expressed in the native BCE cells. Western blot analysis with native BCE cell protein and a polyclonal anti-Kir2.1 antibody revealed a ∼60-kDa band that was blocked by the corresponding synthetic Kir2.1 peptide. Both single BCE cells and BCE cell clusters exhibited an inwardly rectifying K⁺ (Kir) current that was dependent on the extracellular K⁺ concentration. The Kir current was blocked by external Ba²⁺ or Cs⁺ in a voltage- and concentration-dependent manner. In 5 mM K⁺ Ringer’s, the V _m of cell clusters averaged −40.0 ± 4.1 mV (n = 14) and in 140 mM K⁺ Ringer’s it depolarized to −7.4 ± 1.8 mV. Application of Ba²⁺ in 5 mM K⁺ Ringer’s produced a concentration-dependent depolarization of V _m, with 10 mM Ba²⁺ depolarizing V _m from −53.4 ± 4.8 mV to −27.8 ± 6.3 mV (n = 6).

conclusions. Native BCE cells express functional Kir2.1 channels that help determine the membrane potential and probably also play a role in transendothelial transport.

The cornea consists of a stratified epithelium, an underlying stroma, and a nonvascular monolayered endothelium. The corneal endothelium is responsible for pumping salt and water from the corneal stroma to aqueous humor, thus regulating the state of corneal stromal hydration and corneal clarity. ^{1 2} Although K⁺ channels have been implicated in this transport function, ³ they are poorly understood at the functional and molecular levels. Rae and co-workers identified in native corneal endothelial cells from several vertebrate species two types of K⁺ channel: a transient outwardly rectifying K⁺ channel, ⁴ and a Cs⁺-sensitive but Ba²⁺-insensitive inwardly rectifying K⁺ channel whose gating properties result in outwardly rectifying macroscopic currents. ⁵ Recently, we showed that in cell culture, bovine corneal endothelial (BCE) cells express a Ba²⁺-sensitive, inwardly rectifying K⁺ (Kir) channel with electrophysiological properties nearly identical with those of the cloned Kir2.1 channel expressed in Xenopus oocytes. ⁶ About the same time, Rae and Shepard ⁷ reported a full-length coding region of Kir2.1 cloned from native rabbit corneal endothelial cells and a partial coding sequence of Kir2.1 cloned from native BCE cells. However, to date, functional Kir2.1 channels have not been reported in native corneal endothelial cells from any species, nor has expression of Kir2.1 protein been demonstrated. 

In this study, we combined molecular biological, biochemical, and electrophysiological approaches to show that Kir2.1 channels are in fact expressed in native BCE cells, where they play an important role in setting the membrane potential. Some of these data have been reported in abstract form (Yang, et al. IOVS 2001;42:ARVO Abstract 2704). 

Methods

Fresh BCE Cell Preparation

Bovine eyes were obtained from a local abattoir and rapidly transported to the laboratory in ice-cold HEPES-buffered Ringer’s (HR) solution (135 mM NaCl, 5 mM KCl, 10 mM glucose, 1.8 mM CaCl₂, 1.0 mM MgCl₂, and 10 mM HEPES, [pH 7.4]). Adhering extraocular tissue was removed from the globes by dissection. The corneas were excised and placed endothelial side up in a plastic holder. ⁸ The endothelial surface was rinsed and subsequently incubated at 37°C for 30 minutes in cell isolation solution (135 mM N-methyl-d-glucamine [NMDG]-Cl, 5 mM KCl, 10 mM HEPES, 3 mM EDTA-KOH, 10 mM glucose, 3 mM cysteine, 0.06 mg/mL papain, titrated to pH 7.4 with NMDG-free base). Papain was chosen after preliminary experiments indicated that corneal endothelial cells were more viable when dispersed with papain, as opposed to trypsin or dispase II. The endothelial surface was then gently washed with HR solution and the endothelial cells dislodged from Descemet’s membrane by gentle rubbing with a silicone rubber spatula. The dislodged cells suspended in HR solution were collected. The cells from 2 to 30 corneas were pooled and used for electrophysiological recording, total RNA preparation, or whole-cell protein preparation. 

Total RNA Isolation

Total RNA was extracted from freshly isolated native BCE cells (Trizol reagent; Life Technologies, Inc., Rockville, MD), according to the manufacturer’s instructions. RNA pellets were suspended in diethylpyrocarbonate-treated water and quantified by ultraviolet (UV) spectrophotometry. 

RT-PCR Analysis

Total RNA isolated from freshly isolated native BCE cells was reverse transcribed with random decamers with reverse transcriptase (RetroScript; Ambion, Austin, TX) according to procedures outlined in the manufacturer’s instructions. PCR was performed with a primer set specific for Kir2.1 in different species. The forward primer designed from base pairs 848 to 872 of the bovine Kir2.1 coding region (GenBank AY052548; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) ⁶ was 5′-GYA ARC AGG ACA TYG ACA AYG CAG A-3′, and the reverse primer from base pairs 1253-1227 was 5′-GGC TCT AGA GGT ACR CTK GCC TGG TTG -3′. The housekeeping gene, glyceraldehyde-α-phosphate-dehydrogenase (GAPDH), served as a control. The forward primer for GAPDH was 5′-GTG AAG GTC GGA GTC AAC G-3′ from base pairs 113-131 of the human GAPDH sequence (GenBank AF261085); the reverse primer was: 5′-GAG ATG ATG ACC CTT TTG GC-3′ from the region 468-439. The oligonucleotides were synthesized by Integrated DNA Technologies, Inc. (Coralville, IA). The PCR products were generated with DNA polymerase (SuperTaq-Plus; Ambion) and cycled 30 times for GAPDH or 40 times for Kir2.1 (1 minute at 94°C, 1 minute at 50°C, and 1 minute at 72°C), followed by a 7-minute extension at 72°C. The PCR products were separated by 1.5% agarose gel electrophoresis. The sequence of the Kir2.1 RT-PCR product was confirmed by DNA sequencing analysis. Sequencing was performed by the DNA Sequencing Core Facility at the University of Michigan. 

Transfection

The human embryonic kidney cell line HEK 293 expressing simian virus (SV) 40 large T antigen was cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal calf serum (FCS), 100 U/mL penicillin, and 100 μg/mL streptomycin. The expression construct pcDNA3.1/Kir2.1/green fluorescent protein (GFP) was generated by inserting the Kir2.1 coding sequence (GenBank AY052548) ⁶ in-frame into the pcDNA 3.1/GFP, with the GFP tag fused to the 3′ end of the Kir2.1. HEK 293 cells were transfected with 0.8 μg/mL of expression plasmid cDNAs (pcDNA3.1/GFP or pcDNA3.1/Kir2.1/GFP) using a lipophilic transfection agent in serum-free medium (LipofectAmine 2000 and Opti-MEM I, respectively; Life Technologies, Inc., Rockville, MD) according to the manufacturer’s protocol. Whole-cell extracts were prepared for Western blot analysis 72 hours after transfection. Freshly dissociated cells were used for electrophysiological recording 48 to 72 hours after transfection. 

Western Blot Analysis

Whole-cell lysates from freshly isolated native BCE cells (40 μg/lane) or from HEK 293 cells expressing Kir2.1 (5 μg/lane) were subjected to a 4% to 20% linear gradient Tris-HCl gel (Ready Gel; Bio-Rad Laboratories, Hercules, CA). After electrophoresis, proteins were transferred to nitrocellulose membrane and immunoblotted with affinity-purified rabbit polyclonal anti-Kir2.1 antibody (Alomone Laboratories, Jerusalem, Israel) at a dilution of 1:400, followed by development with enhanced chemifluorescent (ECF) substrate (Amersham Pharmacia Biotech, Piscataway, NJ) and chemifluorescence visualization using a phosphorescence imager (Phosphorimager; Molecular Dynamics, Sunnyvale, CA). ⁹ The specificity of anti-Kir2.1 antibody staining was assessed by peptide blocking studies in which the antibody was incubated with the Kir2.1 peptide antigen (Alomone Laboratories) before immunoblot analysis. 

Electrophysiological Recordings

Whole-cell recordings were performed with standard techniques, essentially as described by Hughes and Takahira. ¹⁰ Freshly isolated native BCE cell clusters (5–20 cells) or individual single cells were placed in a continuously perfused Lucite recording chamber. Cells selected for recording had a bright appearance under phase-contrast microscopy. Except where noted, all experiments were conducted at room temperature (23–25°C). Patch pipettes were pulled from glass tubing (7052; Garner Glass, Claremont, CA) with a multistage programmable puller (Sutter Instruments, San Rafael, CA) and heat polished to resistances in the range of 3 to 5 MΩ just before use. Voltage-clamped currents were recorded by whole-cell patch-clamp recording with an amplifier (Axopatch 1D; Axon Instruments, Foster City, CA). The average membrane capacitance (C _m) of single cells was 7 ± 3 pF (±SD, n = 13). C _m of cell clusters was roughly proportional to the number of cells within the cluster. For four clusters containing 18 ± 3 cells (range: 14–20), C _m averaged 107 ± 13 pF. Both voltage-clamp and zero-current clamp data were acquired and analyzed on computer (pCLAMP 8 software; Axon Instruments, Union City, CA). Membrane potential (V _m) was measured in the whole-cell recording mode under zero-current clamp. All voltages were corrected for an offset potential resulting from the liquid junction potentials between the pipette tip and bath solution, which averaged −10 mV. Series resistance averaged 11.8 ± 4.5 MΩ (±SD, n = 13) and was not compensated. 

It is possible that cells within clusters were not voltage clamped to the same value because of space-clamp limitations. This effect would be greatest at voltages farthest removed from the reversal potential and could lead to an underestimation of the degree of inward rectification. However, this problem is not relevant to measurements of open-circuit potential or reversal potential. 

Solutions

The standard bath solution was HR (described earlier). In experiments testing the effects of extracellular K⁺ concentration on current or V _m, a portion of NaCl in HR was replaced with KCl to achieve [K⁺] + [Na⁺] = 140 mM. In blocking experiments, BaCl₂ or CsCl₂ was added to the bath solution to the final concentrations indicated. Bicarbonate-buffered Ringer’s consisted of 120 mM NaCl, 5 mM KCl, 27.5 mM NaHCO₃, 10 mM glucose, 1.8 mM CaCl₂, 1.0 mM MgCl₂ (pH 7.4) and was bubbled with 5% CO₂-95% O₂. The osmolality of all external solutions was 288 ± 5 mmol/kg, except when they contained 10 mM BaCl₂ or CsCl, in which case, solutions had osmolalities of 320 ± 7 and 312 ± 4 mmol/kg, respectively. Control experiments using isotonic solutions in which BaCl₂ or CsCl were substituted for NaCl confirmed that changes in current and membrane voltage were due to the blockage of ion channels as opposed to cell shrinkage. 

The pipette solution consisted of 30 mM KCl, 83 mM potassium gluconate, 5.5 mM EGTA-KOH, 0.5 mM CaCl₂, 4 mM MgCl₂, 10 mM HEPES, and 4 mM K₂ATP and was titrated to pH 7.2 with KOH. The osmolality of the pipette solution was 244 ± 5 mmol/kg. 

Results

RT-PCR Analysis of Kir2.1 Expression

Figure 1 demonstrates the presence of Kir2.1 mRNA transcript in native BCE cells. RT-PCR of native BCE mRNA with a primer pair specific for Kir2.1 resulted in a single band of the expected size (406 bp), identical with the predominant band that resulted from cloned Kir2.1 cDNA. The identity of the product was confirmed to be Kir2.1 by DNA sequencing. No band was observed in a reaction of template and primers without reverse transcriptase (minus RT), indicating that the 406-bp band was generated from mRNA rather than contaminating genomic DNA. RT-PCR with PCR primers for GAPDH also resulted in a band of the expected size, establishing the integrity and viability of the RNA. 

Western Blot Analysis of Kir2.1 Expression

To examine the presence of Kir2.1 in native BCE, we performed Western blot analysis with a commercially available, affinity-purified polyclonal anti-Kir2.1 antibody raised against a synthetic peptide corresponding to amino acid residues 392-412 at the carboxyl terminus of human and bovine Kir2.1. Figure 2 shows that Kir2.1 immunoreactivity was present both in native BCE cells and in GFP-tagged Kir2.1-transfected HEK 293 cells. The identity of the labeled protein band as Kir2.1 was confirmed by the absence of staining when anti-Kir2.1 antibody was preincubated with the synthetic Kir2.1 peptide (Fig. 2A , lane 2; Fig. 2B , lane 3), or when HEK 293 cells were mock transfected (Fig. 2B , lane 1). The blots shown in Figure 2 are representative of three independent experiments. 

In the lane corresponding to native BCE cells, Kir2.1 protein migrated at an apparent molecular mass of ∼60 kDa in 4% to 20% SDS-PAGE (Fig. 2A) . This is greater than the predicted size of unmodified Kir2.1 (48.3 kDa) and probably is due to posttranslational modification. In the lane corresponding to GFP-tagged Kir2.1-transfected HEK 293 cells (Fig. 2B , lane 2), the expressed fusion protein had an apparent MW of ∼80 kDa, which is roughly equal to the algebraic sum of the molecular weights of GFP (27 kDa) and Kir2.1 (48.3 kDa). 

Expression of Inwardly Rectifying K⁺ Current in Native BCE Cells

To evaluate native BCE cells for expression of functional Kir2.1 channels, we recorded whole-cell currents in single, acutely dissociated cells. Figure 3A shows families of currents recorded in a representative cell bathed in 5 mM K⁺ and 140 mM K⁺ Ringer’s. Figure 3B shows the corresponding current–voltage (I-V) relationships averaged from six cells. In 5 mM K⁺ Ringer’s the I-V relationship was essentially linear, but exhibited a somewhat larger slope conductance at voltages negative to about −80 mV. In contrast to previous reports on rabbit corneal cells, ⁴ we did not observe transient or sustained outwardly rectifying currents, even when the membrane potential was held at −60 mV between voltage steps (not shown). When extracellular K⁺ concentration ([K⁺]_o) was increased to 140 mM, inward current increased dramatically and shifted the cell’s zero-current potential in the positive direction. Membrane hyperpolarization from a holding potential of −10 mV rapidly activated an inward current that was sustained except at strong negative potentials, when slow inactivation occurred. In six cells, increasing [K⁺]_o from 5 to 140 mM K⁺ increased the inward slope conductance measured between −140 and −150 mV from an average of 0.26 ± 0.12 to 0.76 ± 0.20 nS and depolarized V _m from −22.4 ± 3.0 to −3.13 ± 2.3 mV. Both of these K⁺-induced changes are consistent with the presence of inwardly rectifying K⁺ channels, but the relatively depolarized membrane potential in 5 mM K⁺ suggested that the membrane had a significant leak conductance. In addition, in most cells, a noisy current appeared after several minutes of recording that made further characterization of the inwardly rectifying K⁺ current difficult. 

We suspected that the leak conductance and noisy current may have been the result of membrane damage due to cell isolation. To test this, we measured currents from small cell clusters consisting of 5 to 20 cells. BCE cell clusters exhibited inwardly rectifying K⁺ currents that were similar to those depicted in Figure 3 , except that they were nearly an order of magnitude larger (Figs. 4A 5A 6A) , indicating that cells in the cluster were electrically coupled. Increasing the [K⁺]_o depolarized the zero-current potential (V ₀) from −39.6 ± 3.0 to −8.2 ± 0.5 mV and increased the inward slope conductance measured between −140 and −150 mV from 7.7 ± 2.9 to 17.1 ± 4.5 nS (Fig. 4) . The more negative membrane potential in BCE cell clusters compared with single BCE cells indicates a higher relative K⁺ conductance, consistent with a lower leak conductance. Because of this and the fact that noisy currents were rarely observed, subsequent experiments were performed on BCE cell clusters. 

Block of Inwardly Rectifying Current by External Ba²⁺ and Cs⁺

It is well established that Kir2.1 channels are blocked by extracellular Ba²⁺ in a time- and voltage-dependent manner. ¹¹ To characterize further the inwardly rectifying K⁺ current in native BCE cell clusters, we examined its sensitivity to blockage by extracellular Ba²⁺ in the presence of 140 mM K⁺. Figure 5 illustrates that external Ba²⁺ produced a time-, voltage-, and concentration-dependent block of inwardly rectifying current in native BCE cell clusters. In the presence of 10 μM Ba²⁺, voltage steps from a holding potential of −10 mV to more negative voltages produced a rapid activation of current followed by a slow, time-dependent decrease (Fig. 5A) . At a given concentration of Ba²⁺, the degree of current inhibition in the steady state increased with membrane hyperpolarization (Fig. 5B) , consistent with the binding of Ba²⁺ to a site within the channel pore that senses the electric field across the membrane. We calculated the dissociation constant (K _d) for the Ba²⁺- induced block according to the first-order equation  

\[I_{\mathrm{B}}/I_{0}\ {=}\ 1/(1\ {+}\ {[}\mathrm{B}{]}/K_{\mathrm{d}}\]

where I _B/I ₀ is the ratio of currents measured in the presence and absence of the blocker and [B] is the blocker concentration. The K _d for the Ba²⁺-induced block at −100 mV averaged 1.2 ± 0.3 μM (mean ± SEM, n = 4). This value is comparable to those obtained from native Kir2.1 channels in cultured BCE cells (2.7 ± 0.4 μM at −100 mV) ⁶ and cloned Kir2.1 channels expressed in Xenopus oocytes (2.6 ± 0.5 μM ⁶ or 0·8 μM ¹² at −100 mV). 

In addition to being blocked by Ba²⁺, Kir2.1 channels are rapidly blocked by external Cs⁺ in a voltage- and concentration-dependent manner. ^{6 11 13 14 15} Figure 6A shows the effect of Cs⁺ on whole-cell currents recorded from a representative BCE cell cluster bathed in 140 mM K⁺ Ringer’s. In the presence of 100 μM Cs⁺, inward currents at large negative potentials exhibited a rapid block, consistent with the fast kinetics of Cs⁺ binding and unbinding in Kir2.1 channels. ^{11 13 16 17} Steady state I-V relationships obtained in the absence and presence of Cs⁺ are shown in Figure 6B . As the membrane was hyperpolarized, the fraction of current blocked by Cs⁺ increased, suggesting that Cs⁺ blocks the channel by binding to a site within the channel pore. Figure 6 also shows that the extent of current inhibition was dependent on Cs⁺ concentration. The K _d for the Cs⁺-induced block at −100 mV averaged 206.6 ± 66.1 μM (n = 4). This value is similar to those obtained previously from native Kir2.1 channels in cultured BCE cells (96.8 ± 19.0 μM at −100 mV) ⁶ and cloned Kir2.1 expressed in Xenopus oocytes (69.9 ± 8.9 μM at −100 mV). ⁶  

Taken together, the sensitivity of the inwardly rectifying K⁺ current to extracellular Ba²⁺ and Cs⁺ suggest that it is mediated by Kir2.1 channels. 

Impact of Kir Channels on the Resting Membrane Potential

To evaluate the contribution of Kir channels to the membrane potential (V _m), we recorded V _m from native BCE cell clusters in whole-cell configuration under the zero-current clamp condition while testing the effects of elevated [K⁺]_o and K⁺ channel blockers. Figure 7A depicts the V _m response of a representative BCE cell cluster to increases in [K⁺]_o. Increasing [K⁺]_o from 5 to 140 mM caused a significant depolarization, indicating that V _m was determined in part by a K⁺ conductance. Figure 7B summarizes the results of similar experiments on 14 native BCE cell clusters. In the presence of 5 mM K⁺, V _m averaged −40.0 ± 4.1 (mean ± SEM; range: −23 to −73 mV) and in the presence of 140 mM K⁺, it averaged −7.4 ± 1.8 (range: −18 to +2.4 mV). To obtain a quantitative estimate of the contribution of Kir conductance to V _m, we applied the chord conductance equation  

\[V_{\mathrm{m}}\ {=}\ (g_{\mathrm{K}}/g_{\mathrm{T}})E_{\mathrm{K}}\ {+}\ (1\ {-}\ g_{\mathrm{K}}/g_{\mathrm{T}})E_{\mathrm{r}}\]

where g _K is the inwardly rectifying K⁺ conductance, g _T is the total conductance, 1 − g _K/g _T is residual conductance normalized to the total conductance, E _K is the K⁺ equilibrium potential, and E _r is the reversal potential of the residual current. Assuming that 10 mM Ba²⁺ blocks g _K specifically and completely, E _r can be taken to be −28 mV (discussed later). This value of E _r is close to the equilibrium potential for Cl⁻ (−32.5 mV), suggesting the residual current is primarily Cl⁻ current. We used this value and equation 2 to calculate that the relative Kir conductance (g _K/g _T) of BCE cells bathed in 5 mM K⁺ averages 0.22 and can be as high as 0.81. The depolarization of V _m to near 0 by 140 mM K⁺ can be accounted for by a K⁺-induced increase in g _K, combined with the depolarization of E _K to 0 mV. 

To obtain further evidence that Kir channels contribute to V _m, we examined the effect of Ba²⁺ on membrane voltage of native BCE cells bathed in 5 mM K⁺ Ringer’s. Figure 8A shows V _m recorded from a representative native BCE cell cluster bathed in 5 mM K⁺ Ringer’s and in 5 mM K⁺ Ringer’s plus the indicated concentrations of Ba²⁺. Application of Ba²⁺ produced a concentration-dependent depolarization of V _m. The results obtained in six cell clusters with initial V _m less than −40 mV are summarized in Figure 8B . In the absence of Ba²⁺, V _m averaged −53.4 ± 4.8 mV, and in the presence of 0.1, 1.0, and 10 mM Ba²⁺, it depolarized to −49.9 ± 4.4 mV (NS), −40.6 ± 4.7 mV, and −27.8 ± 6.3 mV, respectively. In contrast to Ba²⁺, Cs⁺ had only minor effects on V _m (not shown)_. In four cell clusters, 10 mM Cs⁺ depolarized V _m from −58.3 ± 7.5 mV to −52.7 ± 8.3 mV (NS). Similar results were obtained in BCE cell clusters superfused with bicarbonate-buffered Ringer’s at 37°C: in three cell clusters, 10 mM Ba²⁺ depolarized V _m from −41.0 ± 3.5 to −8.4 ± 1.1 mV, whereas in one cell cluster, 10 mM Cs⁺ depolarized V _m from −47.9 to only −46.7 mV. 

On the surface, the relative insensitivity of V _m to Cs⁺ compared with Ba²⁺ might seem to argue against a role for Kir2.1 channels in the generation of the membrane potential. In fact, the block of Kir2.1 channels by Cs⁺ is more strongly voltage dependent, such that inward but not outward currents are inhibited. ⁷ Because Kir2.1 currents would be outward under our experimental conditions (V _m > E _K), one would expect Cs⁺ to have little effect on V _m. To test this prediction, we performed experiments in Kir2.1-transfected HEK293 cells under identical conditions and found that 10 mM Cs⁺ had no significant effect on V _m: in three cells, V _m averaged −79.4 ± 0.4 mV under control conditions and −79.1 ± 1.1 mV in the presence of 10 mM Cs⁺. In contrast, 10 mM Ba²⁺ depolarized V _m in these same cells from an average of −79.7 ± 0.4 to −37.2 ± 7.9 mV. Hence, our findings in BCE cells are consistent with the notion that Kir2.1 channels are an important determinant of V _m. 

Discussion

In this study, we used RT-PCR and Western blot analyses to show that native BCE cells express Kir2.1, an inwardly rectifying K⁺ channel subunit that is widely expressed in excitable and nonexcitable cells. ^{15 18} We also obtained functional evidence for Kir2.1 channels in electrophysiological recordings from single BCE cells and isolated BCE cell clusters, which exhibited an inwardly rectifying K⁺ conductance with biophysical properties and a blocker sensitivity profile consistent with the Kir2.1 channel. This channel is a major component of the BCE K⁺ conductance and appears to be an important determinant of the membrane potential. These results extend our previous observation that functional Kir2.1 channels are expressed in cell cultures of BCE cells. 

Molecular Evidence for Kir2.1 Channel Subunit Expression in Native BCE Cells

In the vertebrate eye, Kir2.1 mRNA has been detected in chick, rabbit, and human lens epithelium, ¹⁹ mouse retina, ²⁰ native rabbit corneal epithelium and endothelium, ⁷ and cultured BCE cells. ⁶ In this study, we detected a 406 bp RT-PCR product in native BCE cells using Kir2.1-specific primers. This band is unlikely to be the result of genomic DNA contamination because no DNA was amplified in absence of reverse transcriptase (Fig. 1) . Hence, we conclude that the product was amplified from Kir2.1 mRNA. 

To examine Kir2.1 expression at the protein level, we performed Western blot analysis with a polyclonal antibody raised against human and bovine Kir2.1. We detected a ∼60 kDa band in native BCE cells and confirmed its identity as Kir2.1 by peptide blocking experiments (Fig. 2) . This molecular weight is larger than that expected for unmodified Kir2.1 (48.3 kDa), probably because of posttranslational modification. Kir2.1 possesses several potential glycosylation sites as well as putative phosphorylation sites for protein kinase A (S425), protein kinase C (S3, T6, S357, and T383), and tyrosine kinase (Y242 and Y 366). ^{6 11 19 21 22}  

Electrophysiological Evidence for Kir2.1 Channel Expression in Native BCE Cells

Whole-cell recording from both isolated single cells and small clusters of native BCE cells revealed an inwardly rectifying current that increased markedly when extracellular [K⁺] was elevated (Figs. 3 4) . Associated with this current increase was a positive shift in zero-current potential, indicating that the change in current amplitude was at least in part the result of an increase in K⁺ conductance. This behavior is qualitatively similar to that of Kir2.1 channels and other members of the Kir channel family, with conductance activated by increases in extracellular K⁺ concentration. ^{6 11}  

In addition to being activated by extracellular K⁺, Kir channels are sensitive to voltage-dependent blockage by extracellular Ba²⁺ and Cs⁺. We found that the inwardly rectifying K⁺ current in native BCE cells was blocked by external Ba²⁺ in a voltage- and concentration-dependent manner with a K _d at −100 mV of 1.2 ± 0.3 μM. This value compares well with the K _d for the Ba²⁺-induced block of cloned Kir2.1 channels ^{6 12 23} and native Kir2.1 channels in different cell types, ^{24 25} including cultured BCE cells, ⁶ but it also agrees with the K _d’s of two other members of the Kir2 channel subfamily, Kir2.2 and Kir2.3. ^{26 27 28 29 30} Members of the Kir2 channel family, however, can be distinguished on the basis of their sensitivity to Cs⁺. Compared with Kir2.2 ^{26 27} and Kir2.3 ^{28 29 30} channels, Kir2.1 channels possess an affinity for Cs⁺ that is about two orders of magnitude lower. ^{11 16 17} The Kir conductance observed in native BCE cells had K _d of 207 ± 66 μM at −100 mV, which is in accordance with the low sensitivity of Kir2.1 channels. Taken together, the pharmacology of the Kir conductance in native BCE cell clusters suggests that they are composed of Kir2.1 channels. 

To our knowledge, the present study is the first to identify Kir2.1 currents in native corneal endothelial cells. Rae and Shepard ⁷ demonstrated the presence of Kir2.1 mRNA in native corneal endothelial cells from a variety of species, but failed to observe Kir2.1 currents. They argued that Kir2.1 protein expression could be as low as several copies per cell and therefore would be difficult to detect. In this study, we were able to resolve inwardly rectifying K⁺ currents in single isolated BCE cells. We can arrive at a rough estimate of the number of channels in these cells by dividing the leak-corrected macroscopic conductance observed in 140 mM K⁺ (∼0.5 nS) by the single-channel conductance (20–25 pS). ^{11 31} Assuming a channel-open probability of 1, this approach yields an estimate of 20 to 25 channels per cell. Because open probability almost certainly is lower, this number is likely an underestimate. 

Previously, Watsky et al. ⁴ and Rae et al. ⁵ described two types of Ba²⁺-insensitive K⁺ channels expressed in rabbit corneal endothelial cells: a transiently activated, outward rectifying K⁺ channel ⁴ and a temperature- and anion-stimulated K⁺ channel. ⁵ The latter exhibits mild inwardly rectifying unitary currents, but outwardly rectifying macroscopic currents due to an increase in open probability with depolarization. ³² In the present study on native BCE cells, we did not observe either current. The reason for these differences between rabbit and BCE cells is unclear, but it may reflect a species-specific difference in gene expression. 

Physiological Roles of Kir Channels in Native BCE Cells

In other cell types, Kir channels play a fundamental role in intracellular homeostasis and in the generation of the resting membrane potential. ^{11 15 16 25 33 34} Our patch-clamp recordings from freshly isolated clusters of native BCE cells indicated a resting V _m of −40 ± 4 mV, consistent with previous measurements in corneal endothelial cells using microelectrodes ^{35 36} and patch pipettes. ^{32 37 38} These values, however, are significantly depolarized from the K⁺ equilibrium potential, indicating that other channels, such as nonselective cation channels ³² and Cl⁻ channels, ^{38 39 40} also contribute to the membrane potential. Nonetheless, we estimate that Ba²⁺-sensitive K⁺ channels contribute approximately 20% of the whole-cell conductance when BCE cells are bathed in 5 mM K⁺ solution. The current–voltage relationships shown in Figures 3 and 4 indicate that Kir2.1 is a dominant K⁺ channel in BCE cells and support the idea that it underlies the Ba²⁺-sensitive K⁺ current. Thus, we conclude that the Ba²⁺-induced depolarization of membrane potential is due to the blockage of Kir2.1 channels. We cannot exclude the possibility, however, that BCE cells may also contain a different type of Ba²⁺-sensitive K⁺ channel that we were unable to detect. 

The corneal endothelium is vital to the maintenance of corneal clarity. The corneal stroma tends to swell due to the imbibition pressure produced by an abundance of proteoglycans, ⁴¹ and this process is counteracted by the endothelial fluid pump, which secretes fluid into the aqueous compartment. A substantial portion of this fluid transport is coupled to active Cl⁻ transport, which is mediated by entry across the basolateral membrane through the Na⁺/K⁺/2Cl⁻ cotransporter and exits across the apical membrane through Cl⁻ channels. ^{3 40 41 42} In addition, active HCO₃ ⁻ secretion generated by influx through a basolateral membrane Na⁺/2HCO₃ ⁻ cotransporter and efflux across the apical membrane via the Cl⁻/HCO₃ ⁻ exchanger and anion channels is also important. Quinidine, a potent blocker of both Kir2.1 ⁴³ and a transient outwardly rectifying K⁺ channel, ⁴ has been shown to elicit a swelling rate in corneas similar to that produced by ouabain, ⁴ suggesting that either or both of these K⁺ channels play a major role in the active transport processes underlying the fluid pump in corneal endothelium. 

The secondary active transport of Cl⁻ and HCO₃ ⁻ is ultimately driven by the transmembrane electrochemical gradient for Na⁺ that is established by the basolateral membrane Na⁺,K⁺-adenosine triphosphatase (ATPase). As in other cells, Na⁺,K⁺-ATPase activity is critically dependent on the presence of an efflux pathway to recycle K⁺, and Kir2.1 channels are likely to function in this regard. Another way K⁺ channels may be linked to the fluid pump is by affecting the electrochemical driving forces on Cl⁻ and HCO₃ ⁻ transport. Because Kir2.1 channels contribute to the membrane potential, their hyperpolarizing influence promotes the efflux of Cl⁻ and HCO₃ ⁻ through anion channels in the apical membrane. ^{3 40 41 42} Ba²⁺-sensitive K⁺ channels have also been implicated in cell volume regulatory responses of BCE cells to hypotonic challenge, where they provide a pathway for the obligatory efflux of cations that accompanies Cl⁻ efflux through swelling-activated anion channels. ³⁶ Our results suggest that the K⁺ channels involved in this process may be Kir2.1. 

Taken together, this work provides the first evidence that native BCE cells express functional Kir2.1 channels. These channels help determine the membrane potential and may also play a role in transendothelial ion transport and cell volume regulation. 

 Supported by National Eye Institute Grants EY11793 (SAE) and EY08850 (BAH) and Core Grant EY07703.

 Submitted for publication December 19, 2002; revised March 17, 2003; accepted March 28, 2003.

 Disclosure: D. Yang, None; D.K. MacCallum, None; S.A. Ernst, None; B.A. Hughes, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

RT-PCR analysis of Kir2.1 expression in native BCE cells. Total RNA isolated from native BCE cells was treated with DNase I, then reverse transcribed with random decamers. PCR was performed using Kir2.1-specific primers (yielding a predicted PCR product of 406 bp) or GAPDH-specific primers (yielding a predicted PCR product of 356 bp). Cloned Kir2.1 cDNA: the positive control PCR reaction.

Figure 2.

 

Western blot analysis of Kir2.1 expression in native BCE cells. (A) Whole-cell lysates (40 μg/lane) from native BCE (A, lanes 1, 2) were subjected to a 4% to 20% SDS-PAGE, transferred to nitrocellulose membranes, and immunoblotted with anti-Kir2.1 antibody (lane 1) and anti-Kir2.1 antibody preabsorbed with synthetic peptide (lane 2). Kir2.1 protein (∼60 kDa; lane 1) was resolved with a polyclonal anti-Kir2.1 antibody at a dilution of 1:400. The ∼60-kDa band was blocked by the synthetic Kir2.1 peptide used to generate the antibody (lane 2). (B) Whole-cell lysates (5 μg/lane) from HEK 293 cells transfected with pcDNA3.1/GFP (lane 1) or pcDNA3.1/Kir2.1/GFP (lanes 2, 3) were subjected to 4% to 20% SDS-PAGE, transferred to nitrocellulose membranes, and immunoblotted with anti-Kir2.1 antibody (lanes 1, 2) and anti-Kir2.1 antibody preabsorbed with its synthetic peptide (lane 3). The anti-Kir2.1 antibody recognized an ∼80 kDa protein (GFP-tagged Kir2.1) in HEK 293 cells transfected with pcDNA3.1/Kir2.1/GFP (lane 2), but not in HEK 293 cells transfected with pcDNA3.1/GFP (lane 1). The ∼80-kDa band was blocked by the synthetic Kir2.1 peptide (lane 3).

Figure 3.

 

Inwardly rectifying K⁺ currents in single isolated BCE cells. (A) Whole-cell current recorded from a representative native BCE cell showing a large increase in inward current produced by elevating the extracellular K⁺ concentration from 5 (top) to 140 (bottom) mM. The horizontal line to the left of the current records indicates the zero-current level. The voltage protocol used to evoke these currents is shown below the panels. (B) Summary of current-voltage relationships obtained in six single BCE cells bathed in 5 and 140 mM K⁺. Data represent the mean ± SEM.

Figure 4.

 

Inwardly rectifying K⁺ currents in BCE cell clusters. (A) Whole-cell current records from a representative native BCE cell cluster showing a large increase in inward current produced by increasing extracellular K⁺ concentration from 5 (top) to 140 (bottom) mM. The horizontal line to the left of the current records indicates the zero-current level. The voltage protocol used to evoke these currents is shown below the panels. (B) Summary of current-voltage relationships obtained in five BCE cell clusters bathed in 5 and 140 mM K⁺. Note the large increase in inward current and positive shift in zero-current potential (V ₀). In 5 mM K⁺, V ₀ ranged from −49 to −31 mV and in 140 mM K⁺, it ranged from −9.3 to −6.7 mV. Currents in each cell were normalized to the current at −160 mV in 140 mM K⁺.

Figure 5.

 

Inwardly rectifying K⁺ currents in native BCE cells are blocked by external Ba²⁺. (A) Whole-cell currents recorded from a representative native BCE cell cluster bathed in 140 mM K⁺ Ringer’s and in 140 mM K⁺ Ringer’s plus 10 μM Ba²⁺. External Ba²⁺ induced a slow time-dependent block of inward K⁺ currents. (B) Current-voltage relationships of native BCE cell clusters in the presence of the indicated concentrations of Ba²⁺. Data represent the mean ± SEM in four BCE cell clusters. Currents in each cell cluster were normalized to the current at −160 mV measured in the absence of Ba²⁺.

Figure 6.

 

Inwardly rectifying K⁺ currents in native BCE cells are blocked by external Cs⁺. (A) Whole-cell currents recorded from a representative native BCE cell cluster bathed in 140 mM K⁺ Ringer’s and in 140 mM K⁺ Ringer’s plus 100 μM Cs⁺. External Cs⁺ induced a rapid block of inward K⁺ currents. (B) Current-voltage relationships of native BCE cell clusters in the presence of the indicated concentrations of Cs⁺. Data represent the mean ± SEM for four BCE cell clusters. Currents in each cell cluster were normalized to the current at −160 mV measured in the absence of Cs⁺.

Figure 7.

 

K⁺ conductance contributes to the membrane potential of native BCE cells. (A) Membrane potential (V _m) recorded from a representative native BCE cell cluster exposed to two concentrations of K⁺ (5 and 140 mM). Elevated [K⁺]_o caused significant depolarization, indicating that K⁺ conductance contributes to V _m. (B) Summary of V _m obtained in native BCE cell clusters bathed in 5 and 140 mM K⁺. In the presence of 5 mM K⁺, V _m ranged from −73 to −23 mV and in the presence of 140 mM K⁺, it ranged from −18 to +2.4 mV. Data represent the mean ± SEM in 14 native BCE cell clusters.

Figure 8.

 

Ba²⁺ depolarizes the membrane potential of native BCE cells in a concentration-dependent manner. (A) V _m recorded from a representative native BCE cell cluster bathed in 5 mM K⁺ Ringer’s and in 5 mM K⁺ Ringer’s plus the indicated concentrations of Ba²⁺. Ba²⁺ produced a concentration-dependent depolarization of V _m. (B) Summary of V _m obtained in six native BCE cell clusters bathed in 5 mM K⁺ Ringer’s and in 5 mM K⁺ Ringer’s plus the indicated concentrations of Ba²⁺. Application of Ba²⁺ in the presence of 5 mM K⁺ produced a concentration-dependent depolarization of V _m, with 10 mM Ba²⁺ depolarizing V _m from −53.4 ± 4.8 mV to −27.8 ± 6.3 mV.