Abstract
Abstract. In immortalized human kidney epithelial (IHKE-1) cells derived from proximal tubules, two natriuretic peptide receptors (NPR) were identified. In addition to NPR-A, which is bound by atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and urodilatin (URO), a novel form of NPR-B that might be bound by C-type natriuretic peptide (CNP) was identified using PCR. This novel splice variant of NPR-B (NPR-Bi) was also found in human kidney. Whereas ANP, BNP, and URO increased intracellular cGMP levels in IHKE-1 cells in a concentration-dependent manner, CNP had no effect on cGMP levels. To determine the physiologic responses to these agonists in IHKE-1 cells, the membrane voltage (Vm) was monitored using the slow whole-cell patch-clamp technique. ANP (10 nM), BNP (10 nM), and URO (16 nM) depolarized these cells by 3 to 4 mV (n = 47, 7, and 16, respectively), an effect that could be mimicked by 0.1 mM 8-Br-cGMP (n = 15). The effects of ANP and 8-Br-cGMP were not additive (n = 4). CNP (10 nM) also depolarized these cells, by 3 ± 1 mV (n = 28), despite the absence of an increase in cellular cGMP levels, indicating a cGMP-independent mechanism. In the presence of CNP, 8-Br-cGMP further depolarized Vm significantly, by 1.6 ± 0.3 mV (n = 5). The depolarizations by ANP were completely abolished in the presence of Ba2+ (1 mM, n = 4) and thus can be related to inhibition of a K+ conductance in the luminal membrane of IHKE-1 cells. The depolarizations attributable to CNP were completely blocked when genistein (10 μM, n = 6), an inhibitor of tyrosine kinases, was present. These findings indicate that natriuretic peptides regulate electrogenic transport processes via cGMP-dependent and -independent pathways that influence the Vm of IHKE-1 cells.
Natriuretic peptides are a family of structurally related proteins that have a wide spectrum of biologic activities (1). To date, four major natriuretic peptide subgroups have been identified: atrial natriuretic peptide (ANP), urodilatin (URO), brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP). Three types of receptors for these peptides have been described (2, 3). Two of these receptors, i.e., natriuretic peptide receptor A (NPR-A) and NPR-B, contain an intracellular guanylyl cyclase domain that is activated upon binding of these peptides to the extracellular receptor domain. CNP has a high affinity for NPR-B, whereas NPR-A is activated mainly by ANP, URO, and BNP. A third receptor, NPR-C or clearance receptor, serves to remove natriuretic peptides from the plasma. It only has a short cytoplasmic domain and is not coupled to a guanylate cyclase. All known natriuretic peptides bind to this receptor (4).
It is widely accepted that ANP and URO are responsible for the reduction of NaCl reabsorption in the nephron, leading to natriuresis and concomitant diuresis (5, 6). The receptor (NPR-A) has been localized in various sections of the kidney, especially in the glomeruli, thin limb of Henle's loop, cortical and inner medullary collecting ducts, and renal vasculature, using either immunohistochemical or reverse transcription (RT)-PCR techniques (7, 8). The major target sites for ANP in the kidney are considered to be the glomeruli and the inner medullary collecting duct (9). However, in the proximal tubule it has been reported that ANP inhibits Na+ and water reabsorption by an interaction with protein kinases (10, 11) and Na+-coupled phosphate contransport as well as Na+/H+ antiport (12). To date, no such actions have been shown for CNP. Interestingly, specific mRNA for CNP could be demonstrated in proximal convoluted and straight tubules using PCR analysis, indicating a possible autocrine or paracrine action (13).
We demonstrate that in immortalized human kidney epithelial (IHKE-1) cells derived from the proximal tubule, NPR-A is expressed and the natriuretic peptides ANP, URO, and BNP depolarize the membrane voltage (Vm) by inhibition of a K+ conductance via a cGMP-dependent mechanism. Furthermore, we show that a splice variant of NPR-B (NPR-Bi) is present in these cells, as well as in the human kidney, and that CNP depolarizes the Vm through a cGMP-independent mechanism.
Materials and Methods
Cell Culture
IHKE-1 cells (derived from embryonic kidneys) were cultured as described previously (14). In short, IHKE-1 cells were maintained in Dulbecco's modified Eagle medium/F-12 medium (1:1), with 15 mM Hepes, pH 7.3, 1.6 nM epidermal growth factor, 100 nM hydrocortisone, 83 μM transferrin/insulin, 29 nM Na2SeO3, 10 mM NaHCO3, 20 mM L-glutamine, 1000 U/L penicillin/streptomycin, and 1% fetal calf serum, in an atmosphere of 5% CO2/95% air at 37°C. Subculturing was performed using 0.05% trypsin/0.02% ethylenediaminetetra-acetic acid in Mg2+ - and Ca2+-free phosphate buffer. Culture medium was exchanged twice each week. Cells were used from passages 162 to 188. Cells grew polarized on glass coverslips, with the apical surface facing upward, and developed apical microvilli (Figure 1). After the glass coverslips with the cells were transferred from the culture dishes to the perfusion chamber, they were rinsed for at least 20 min before electrophysiologic measurements were started with a standard solution (see below), at 37°C in a constantly running bath. All media, buffers, and growth factors were purchased, at the highest available purity, from Life Technologies (Eggenstein, Germany), Biochrom/Seromed (Berlin, Germany), Calbiochem (Bad Soden, Germany), Sigma (Deisenhofen, Germany), or Merck (Darmstadt, Germany).
Apical cell surface of immortalized human kidney epithelial (IHKE-1) cells derived from human proximal tubules, imaged by scanning electron microscopy. Cells from microvilli, which is similar to the cellular morphologic features of proximal tubule cells in mammalian kidneys. Magnification, ×6000.
Scanning Electron Microscopy
Cultured cells were fixed with 4% paraformaldehyde in phosphate-buffered saline for 1 h and were stored in 8% paraformaldehyde in Hepes buffer until dehydration in ethanol solutions with increasing ethanol content. After the critical point, dried cells were coated with a 10-nm-thick layer of platinum and examined using a Hitachi S-800 scanning electron microscope (Nissan Sanyo, Ratingen, Germany).
cGMP and cAMP Assays
Cells were grown to confluence in 24-multiwell plates (10 mm), washed twice with 2 ml of serum-free phosphate-buffered saline, and then preincubated at 37°C for 5 min in 1 ml of medium containing 1 mM 3-isobutyl-1-methylxanthine. The natriuretic peptides were added for 10 min. The reaction was stopped by removal of the medium, addition of ice-cold ethanol, and storage at -20°C. The ethanol was evaporated, the sediment was resuspended in 150 μl of 50 mM sodium acetate buffer (pH 6.0) and acetylated, and cGMP or cAMP content was measured with a specific RIA (15).
Primers Used
The following PCR primers were used (listed in the 5′ to 3′ direction): NPRA-541f, CAAGCGCTCATGCTCTACGCCTAC; NPRA-1140r, GATGTTCTCCCCATCAGTAACAGTTC; NPRB-97f, GAACACAACCTGAGCTATGCC; NPRB-549r, GGTGAAGTAGTGAGGCCGGTCA; NPRB-2S, CTGGCCTCCCAGGCCGAAATGGTC; NPRB-3A, TTCAGCGCTTGACCATTAGACTCC; NPRBi-1S, GACTCTCACTCCAGCCCTAGTCTC; NPRB-1f, ATGGCGCTGCCATCACTTCTGCTGTTGG; NPRB-2850r, GGGTCGGTGGCGGATGCGAAAGGAAG; NPRB-2760f, CCGAAATGGTCAACGCCATGCACC; NPRB-3337r, CAAGCCAGAGAGGGACAGGTATATGTA; NPRC-428f, GTGGCCCGGCTTGCATCGCACTGG; NPRC-805r, TCCGGATGGTGTCACTGCTCG; UNIP-5, oligo(dT). Primers were obtained from Perkin Elmer (Weiterstadt, Germany) or Biometra (Göttingen, Germany).
RT-PCR
Cells from cultures were lysed in a 4 M guanidinium chloride buffer, and total RNA was isolated using the RNeasy-kit (Qiagen, Hilden, Germany). RNA from human kidney tissue samples was purified as described previously (16). cDNA first-strand synthesis was performed in a total reaction volume of 30 μl, containing 5 μg of total RNA, 200 μM nucleotide triphosphate mixture, and 200 U of RNaseH- SuperScript Plus reverse transcriptase (Life Technologies). One-three hundredth of each cDNA first-strand reaction mixture was then subjected to a 50-μl PCR in a Perkin Elmer model 9600 thermocycler (Perkin Elmer), using 20 pmol of each primer and 1 U of Taq DNA polymerase (Biomol, Hamburg, Germany). Reaction conditions were as follows: 4 min at 94°C, 1 cycle; 30 s at 98°C, 1 min at 58°C, and 1 min 72°C, 35 cycles. PCR products were analyzed by agarose gel electrophoresis using Sau3A-cleaved pUC18 as a size marker.
For receptor-specific PCR, the following primer pairs were used: NPR-A, NPRA-541f/NPRA-1140r; NPR-B, NPRB-97f/NPRB-549r, NPRBi-1S/NPRB-3A, and NPRB-2S/NPRB-3A; NPR-C, NPRC-428f/NPRC-805r. Negative controls included amplification reactions with non-reverse-transcribed RNA, amplifications without template, and amplifications with a single primer. Amplification of the entire coding region of the IHKE-1 NPR-B was accomplished using the primer paris NPRB-1f/NPRB-2850r and NPRB-2760f/NPRB-3337r, which generated overlapping fragments of 2850 and 578 bp, respectively.
Genomic PCR
High-molecular weight genomic DNA from blood samples from three different male individuals was isolated as described (17). Twenty-five nanograms of each DNA was subjected to PCR using the primer pair NPRB-2S/NPRB-3A. Reaction conditions were as described above, with the exception of elongation steps of 2 min at 72°C.
DNA Sequencing
PCR products were separated by standard agarose gel electrophoresis. Gel pieces containing the desired DNA fragments were cut out with a clean scalpel, and the DNA was isolated using the QIAquick™ gel extraction kit (Qiagen, Hilden, Germany). Twenty to 40 ng of the purified fragments or 500 to 800 ng of the plasmid-cloned fragments were sequenced with each of the amplification primers (separately) or vector-derived standard sequencing primers and the DyeDeoxy Terminator cycle sequencing kit (Perkin Elmer). The reaction mixtures were subsequently analyzed using an ABI PRISM 310 DNA fluorescence sequencer (Perkin Elmer).
Patch-Clamp Studies
Coverslips with cultured cells were fixed at the bottom of a perfusion chamber mounted on an inverted microscope (IM 35; Zeiss, Oberkochen, Germany). The perfusion chamber was continuously perfused at a rate of 10 to 30 ml/min with a standard solution containing 145 mM NaCl, 0.4 mM KH2PO4, 1.6 mM K2HPO4, 5 mM D-glucose, 1 mM MgCl2, and 1.3 mM calcium gluconate; the pH was adjusted to 7.4. All agonists were dissolved in this standard solution immediately before use. All experiments were performed at 37°C. Recordings of Vm were made in the cell-attached configuration, using the slow whole-cell method (18). For this method, pipettes were filled with a solution containing 95 mM potassium gluconate, 30 mM KCl, 1.2 mM NaH2PO4, 4.8 mM Na2HPO4, 5 mM D-glucose, 0.73 mM calcium gluconate, 1 mM ethylene glycol bis(β-aminoethyl ether)-N,N,N′,N′-tetra-acetic acid (EGTA), 1.03 mM MgCl2, and 1 mM ATP; the pH was adjusted to 7.2 before 21 μM nystatin (Sigma) was added. The input resistance of these pipettes was 8 to 15 MΩ. Vm and currents across the membrane were recorded continuously with a patch-clamp amplifier (U. Fröbe, Physiologisches Institut, Universität Freiburg, Freiburg, Germany) and plotted with a pen recorder (WeKaGraph WK-250R; WKK, Kaltbrunn, Switzerland). Voltages always refer to the cytosolic face of the membrane.
Measurements of Intracellular Ca2+ Concentrations
The intracellular Ca2+ concentration ([Ca2+]i) of IHKE-1 cells was measured with the Ca2+-sensitive dye Fura-2, as described previously (19, 20). IHKE-1 cells were incubated for 60 min with Fura-2/acetoxymethyl ester (5 μM) dissolved with 0.1 g/L pluronic F-127 in standard solution, at 37°C in the dark. Cells were excited at 340, 360, and 380 nm using a filter wheel (Physiologisches Institut, Universität Freiburg) rotating at 10 Hz, with a Xenon quartz lamp (XBO 75 W; Zeiss) as the light source. Fura-2 emission was recorded at 500 to 530 nm from approximately five cells of each monolayer, using an adjustable aperture. The ratio of the emissions after excitation at 340 and 380 nm at 10 Hz was calculated, and 10 consecutive values were averaged, yielding a time resolution of 1 Hz. Signal noise and autofluorescence were measured before the cells were loaded with Fura-2/acetoxymethyl ester and amounted to <5% of the signal after the cells were loaded. These background values were subtracted from the measured signals for each experiment. Fluorescence at 360 nm was used to detect volume changes, leakage or bleaching of Fura-2, loss of cells, or air bubbles in the area of measurement during the experiment (19). Calibration of [Ca2+]i was attempted at the end of each experiment by incubation of the cells with the Ca2+ ionophore ionomycin (1 μM; Sigma) in the presence (1.3 mM) and nominal absence of Ca2+ (with 5 mM EGTA), according to standard methods (21), assuming a dissociation constant of Fura-2 for Ca2+ of 224 nM.
Biochemical Reagents
All standard chemicals and nucleotides were obtained, at the highest available purity, from Merck, Calbiochem, or Sigma. The natriuretic peptides ANP, URO, BNP, and CNP were synthesized in the Niedersächsisches Institut für Peptid-Forschung (Hannover, Germany).
Statistical Analyses
Data are presented as original recordings from individual experiments or as mean values ± SEM, with n referring to the number of observations. Slow whole-cell experiments were performed in a paired manner, with control periods before and after each experimental maneuver. Pre- and postexperiment control values were averaged and compared with the corresponding experimental value. Therefore, a two-sided paired t test was used to test for statistical significance. P < 0.05 was set as the significance level.
Results
Stimulation of cGMP Production
As a functional index of NPR-A and NPR-B presence, we compared the ability of four natriuretic peptides (ANP, URO, and BNP for NPR-A and CNP for NPR-B) to stimulate cGMP generation in IHKE-1 cells. Figure 2 shows the relationship between the concentration of each peptide and the cellular cGMP response. ANP and URO were equally potent and 1000 times more potent than BNP in stimulating cGMP generation. ANP and URO both showed maximal stimulatory effects at a concentration of 1 nM, with significant effects at 0.1 pM. CNP failed to stimulate cGMP production at concentrations up to 100 nM (Figure 2). All natriuretic peptides failed to increase intracellular cAMP levels (data not shown).
Concentration-dependent increases in cellular cGMP levels via stimulation of particulate guanylate cyclase activity by the natriuretic peptides atrial natriuretic peptide (ANP), urodilatin (URO), brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP), in the presence of 3-isobutyl-1-methylxanthine (1 mM). Confluent cultures of IHKE-1 cells were processed as described in Materials and Methods and were incubated with increasing concentrations of each peptide. The amount of cGMP was measured by RIA. Each point represents the mean of triplicate measurements from three separate samples. The fits were performed freehand or were fitted linearly.
Detection of NPR-A and NPR-B by RT-PCR
Total RNA from IHKE-1 cells was transcribed into cDNA by RT. A subsequent PCR with cDNA from IHKE-1 cells and NPR-A- or NPR-B-specific primers gave rise to DNA products of the expected sizes (600 and 453 bp, respectively). Comparable levels of β-tubulin-specific PCR products were used, to enable semiquantitative interpretation of receptor gene expression (Figure 3). No product was obtained using the NPR-C-specific primers (data not shown). Sequencing of the PCR products verified the amplification of the authentic receptor DNA sequences of NPR-A and NPR-B.
Ethidium bromide-stained agarose gel showing the products of reverse transcription-PCR analysis of IHKE-1 cells. PCR was performed using 25, 30, 35, or 40 cycles (from left to right). Fifteen microliters of each reaction mixture was applied to the lanes. Homogeneous natriuretic peptide receptor A (NPR-A)- and NPR-B-specific PCR products were obtained with the expected sizes of 600 and 453 bp, respectively. The comparable levels of β-tubulin-specific PCR products enabled reliable semiquantitative interpretation of receptor gene expression. No PCR products were obtained from negative controls (C).
Cloning and Detection of an NPR-B Subtype
Because CNP failed to stimulate cGMP generation, we decided to investigate the NPR-B primary transcript in a more precise way. For this purpose, we first amplified two IHKE-1 cell-derived fragments representing the entire coding region of the NPR-B cDNA, as described above. Cloning and sequence analysis of the fragments revealed the existence of a variant of the NPR-B primary transcript, which was probably generated by alternative splicing. This variant contains an additional 71-bp insertion, between the nucleotides in positions 3537 and 3538 of the NPR-B cDNA, and was designated NPR-Bi (Figure 4). The 5′ - and 3′-terminal nucleotides of the insertion perfectly match the consensus sequence for intron boundaries (exon| GTRAG—intron—AG| exon) (22). The deduced amino acid sequence of the NPR-Bi variant corresponds to the initial 941 amino acid residues of the wild-type receptor. However, because of the additional insertion and a frame shift that results in premature termination of translation, the carboxy-terminal 84 amino acids are replaced by 31 deviating amino acids. Because the carboxy terminus of NPR-B contains the guanylyl cyclase catalytic domain, it is most likely that NPR-Bi does not exhibit guanylyl cyclase activity.
Comparison of the nucleotide and amino acid sequences of the published NPR-B and its splice variant (NPR-Bi) with an insertion from IHKE-1 cells. Nucleotides at the beginning and end of the insertion are splice consensus sequences (GT/AG); TGA describes a stop codon (TER), which arises from the frameshift resulting from the insert in the splice variant. The nucleotides between the splice consensus sequences are part of NPR-Bi. The numbers below the sequence indicate the amino acid positions in NPR-B, showing that the insert in the splice variant interrupts Gly963, which happens to be regenerated by the inserted sequence. Black bars, nontranslated regions; light gray bars, extracellular domain; dark gray bars, kinase-like domain; white bars, guanylyl cyclase domain; striped bar, insertion of NPR-Bi.
To verify the existence of the detected NPR-Bi form, we analyzed different cDNA and genomic DNA using NPR-Band NPR-Bi-specific primers (Figure 5). First, we performed PCR with human kidney cDNA and cDNA and cDNA from IHKE-1 cells, using the primer pair NPRBi-1S/NPRB-3A (see above). NPRBi-1S is a sense primer derived from the insertion of the NPR-Bi variant, whereas NPRB-3A is an antisense primer corresponding to the region flanking the downstream terminus of the insertion. In both cases, we obtained homogeneous PCR fragments of the expected size of 169 bp, clearly verifying the existence of the NPR-Bi splice variant. Furthermore, we performed PCR analysis with cDNA from human kidney, cDNA from two independent IHKE-1 samples, and genomic DNA from three different male individuals, using the primer pair NPRB-2S/NPRB-3A (see above). Primer NPRB-2S is a sense primer hybridizing with the cDNA region flanking the NPR-Bi insertion in the upstream direction. In the case of the human kidney sample, as well as the two IHKE-1 cDNA samples, we obtained two different PCR fragments (248 and 319 bp). The size of the smaller fragment corresponds to that expected for NPR-B lacking the 71-bp insertion, whereas the larger fragment corresponds to the NPR-Bi form. Therefore, both forms do exist in kidney and IHKE-1 cells. The same primer pair was used to analyze the genomic DNA samples. In this case we exclusively obtained the larger fragment of 319 bp, indicating that the 71-bp insertion represents an intron normally occurring within the human NPR-B gene. Next, a PCR was performed using NPRBi-1S as the forward primer and UNIP-5, an oligo(dT) primer, as the reverse primer, confirming that mRNA was amplified. In addition, all fragments obtained were sequenced, directly confirming the correctness of the interpretations of the data given above.
Agarose gel electrophoresis of the different PCR fragments obtained using NPR-B- and NPR-Bi-specific primers and different samples of cDNA and genomic DNA (inverse presentation). Amplification of homogeneous cDNA fragments of the expected size of 169 bp from human kidney and IHKE-1 (IHKE A) cDNA using the NPR-Bi-specific primer pair NPRBi-1S/NPRB-3A indicates the occurrence of this receptor form in both samples. The use of the primer pair NPRB-2S/NPRB-3A, flanking the 71-bp insertion of the NPR-Bi-specific cDNA, led to the amplification of small (248 bp) and large (319 bp) fragments from cDNA from human kidney and two independent IHKE-1 samples (IHKE A and IHKE B). These sizes correspond to those expected for the NPR-B and NPR-Bi forms, respectively, indicating the existence of both mRNA variants in the samples analyzed. Using the same primer pair, only the larger fragment was obtained from three different samples of human genomic DNA (Gen. A, Gen. B, and Gen. C; see text), indicating that the 71-bp NPR-Bi insertion represents a normally occurring intron within the human NPR-B gene.
Electrophysiologic Characteristics and [Ca2+]i Measurements
IHKE-1 cells were investigated 5 to 15 d (10 ± 1 d, n = 297) after passaging and had a mean Vm of -54 ± 1 mV (n = 221). There was no detectable specific Na+ conductance, because amiloride (10 μM) failed to hyperpolarize Vm (▵Vm = 1 ± 1 mV, n = 6). After complete removal of extracellular Na+ (replaced by N-methyl-D-glucosamine), cells were initially hyperpolarized by 21 ± 2 mV (n = 36), indicating that these cells do posses Na+-dependent electrogenic transport systems, such as the Na+/glucose and Na+/phosphate transporters or nonselective cation channels.
Reduction of the extracellular Cl- concentration from 145 to 32 mM (replaced by gluconate) revealed a small Cl- conductance, because cells were initially depolarized by 5 ± 1 mV (n = 23). Neither the Cl- channel blocker 5-nitro-2-(3-phenylpropylamino)benzoate (10 μM) nor 4,4′-diisothiocyanatostilben-2,2′-disulfonic acid (0.5 mM) had significant effects on Vm (n = 4 each).
Increasing the extracellular K+ concentration by 15 mM (with a corresponding decrease in the Na+ concentration) resulted in a depolarization of 12 ± 1 mV (n = 66). A comparable response was seen with the K+ channel blocker Ba2+ at a concentration of 1 mM (▵Vm = 14 ± 2 mV, n = 20).
After demonstration of the receptors for natriuretic peptides and observation of increases in intracellular cGMP levels produced by these peptides, we were interested in the physiologic responses to these agonists in IHKE-1 cells. Therefore, we measured the Vm of IHKE-1 cells with the patch-clamp technique. In 47 recordings, ANP (1 nM) depolarized IHKE-1 cells by 4.0 ± 0.4 mV. At a concentration of 0.1 nM, ANP still significantly depolarized Vm, by 3.0 ± 0.6 mV (n = 6). This effect was mimicked by 8-Br-cGMP (0.1 mM), a membranepermeable analog of cGMP (3.8 ± 0.6 mV, n = 15). Figure 6 shows an original recording of Vm, demonstrating the depolarization induced by ANP (10 nM) and the lack of effect in the presence of Ba2+ (1 mM). In four paired experiments in which the effect of ANP was tested in the absence and presence of Ba2+, ANP depolarized Vm by 4 ± 1 mV without Ba2+ and failed to depolarize Vm (-0.4 ± 0.5 mV) in the presence of Ba2+. BNP (10 nM) also depolarized Vm by 4.6 ± 0.7 mV (n = 16), whereas CNP (10 nM, n = 28) and URO (16 nM, n = 7) significantly depolarized Vm, by 3.0 ± 0.3 and 2.9 ± 0.7 mV, respectively. Figure 7 presents a summary of the effects of natriuretic peptides (ANP, URO, BNP, and CNP) and 8-Br-cGMP on Vm. There was no additivity between ANP and 8-Br-cGMP (▵Vm = 0.3 ± 0.1 mV, n = 4). However, in the presence of CNP, 8-Br-cGMP further depolarized Vm significantly, by 1.6 ± 0.3 mV (n = 5). Because CNP depolarized Vm significantly but did not increase intracellular cGMP levels, we investigated the effects of two different activators of the protein kinase A pathway on Vm. Forskolin (1 μM) and 8-Cl-cAMP (0.1 mM) had no effect on the Vm of IHKE-1 cells (▵Vm = 1 ± 1 mV, n = 16 and 6, respectively). sn-1,2-Dioctanoylglycerol (1 μM), a membrane-permeable activator of protein kinase C, also had no effect on Vm (▵Vm = 1 ± 1 mV, n = 8).
Original recording of the membrane voltage (Vm) of IHKE-1 cells with the slow whole-cell method. ANP (10 nM) depolarized Vm by 4 mV, but in the presence of 1 mM Ba2+ it was without effect. The induced depolarization by ANP could be repeated after Ba2+ was washed out. C, control solution.
Summary of the effects of ANP (10 nM), BNP (10 nM), CNP (10 nM), URO (16 nM), and 8-Br-cGMP (10 mM) on Vm of IHKE-1 cells. The numbers in parentheses refer to the number of experiments. *P < 0.05, statistical significance of the effects.
In seven experiments, CNP (10 nM) also failed to change [Ca2+]i. As a positive control for the reactivity of these cells, we used ATP (10 μM), which reversibly and significantly increased [Ca2+]i in these cells, by approximately 300 nM (n = 13). Removal of extracellular Ca2+ significantly reduced [Ca2+]i by 31 ± 14 nM (n = 16).
Because of the one-transmembrane domain structure of the CNP receptor, we also tested genistein (10 μM), an inhibitor of the tyrosine kinases epidermal growth factor receptor and p60v-src (23). In six paired experiments, CNP (10 nM) depolarized Vm by 3.1 ± 0.3 mV, whereas in the presence of genistein these depolarizations were completely abolished (ΔVm = 0.3 ± 0.5 mV). In five of five excised membrane patch experiments, genistein had no direct effect on the K+ channel itself.
Discussion
To elucidate the physiologic functions and possible clinical implications of the actions of members of the natriuretic peptide family in the renal proximal tubule, the NPR system was investigated. We monitored possible changes in cellular cGMP and cAMP concentrations, electrophysiologic parameters, and [Ca2+]i after addition of cGMP-stimulating agonists to immortalized human kidney epithelial (IHKE-1) cells, as a model for the proximal tubule. These cells provide several indications that they are derived from proximal tubules, e.g., they possess brush border membranes with microvilli, they express proximal tubule-specific enzyme markers (maltase, alkaline phosphatase, and leucine aminopeptidase) in the luminal membrane, and they exhibit several Na+-dependent and -independent amino acid and organic cation transport systems (24,25,26).
In the kidney, natriuretic peptides led to natriuresis and diuresis, with glomeruli and inner medullary collecting ducts as their most important sites of action (5, 6, 9). In proximal tubules, it was shown that ANP inhibits Na+ and water reabsorption (10, 27, 28).
In this study, we show an inhibitory effect of natriuretic peptides on a K+ conductance in the apical membrane of IHKE-1 cells. Natriuretic peptides depolarized the Vm of IHKE-1 cells, indicating either activation of a nonselective cation conductance or a Cl- conductance or inhibition of a K+ conductance. Because the depolarizing effects of ANP on the Vm of these cells were completely abolished by Ba2+, a well known K+ channel blocker, it is clear that natriuretic peptides inhibited a K+ conductance in our study. Direct effects of natriuretic peptides on K+ conductances and K+ channels were reported for mesangial cells, where a K+ conductance was stimulated by ANP, BNP, and URO (29) and where a large Ca2+-dependent K+ channel was activated by ANP (30), and for adrenal glomerulosa cells, where ANP also activated a K+ conductance (31). Furthermore, in rat pituitary tumor (GH4Cl) cells, a K+ channel was activated by natriuretic peptides, via cGMP-dependent dephosphorylation (32), and a new endogenous natriuretic factor (LLU-α) was shown to inhibit a K+ channel in the luminal membrane of the thick ascending limb of rat kidney (33). In rat cortical collecting ducts, we recently demonstrated activation of a K+ channel via a cGMP-dependent protein kinase (34, 35).
Inhibition of the K+ conductance by natriuretic peptides, as described in this study, led to a depolarization of 4 mV. This moderate effect decreases the driving force for Na+ -coupled transport systems and thus might explain the reduced Na+ transport seen with ANP in the proximal tubule (10, 12). This effect might be more pronounced in proximal tubules in vivo; in the cultured cells used in this study, the K+ conductance contributes less to Vm, leading to more depolarized baseline Vm values, compared with in vivo studies (36). The effects of natriuretic peptides on Vm and intracellular cGMP levels are already maximal at a concentration of 100 pM, which is within the physiologic range of ANP concentrations in plasma. A more detailed characterization of this K+ conductance, especially a resolution at the single-channel level, determination of its involvement in Na+ -coupled transport processes, and its responsibility for repolarizing Vm, will be the subject of a future study.
Three different receptors for natriuretic peptides have been characterized: NPR-A, NPR-B, and NPR-C (2, 3). In proximal tubules, NPR-A and NPR-B were detected in rat kidneys by RT-PCR microlocalization (7, 13). The existence of NPR-A was also demonstrated in microdissected proximal tubules of rat and rabbit kidneys by peptide binding studies (37). In this study, we detected the messages for NPR-A and NPR-B in IHKE-1 cells using the RT-PCR technique. As expected, all natriuretic peptides increased intracellular cGMP levels; only CNP had no effect on the intracellular cGMP concentration, even at the high concentration of 10-7 M. Therefore, the observed effect of CNP on Vm is apparently not mediated via cGMP. This conclusion is supported by the additivity of effects attributable to CNP and 8-Br-cGMP. Such additivity was not seen with ANP and 8-Br-cGMP. A lack of increase in cGMP levels with CNP was previously described in proximal convoluted tubules (13), despite the demonstrated presence of NPR-B (2, 3). By cloning the complete NPR-B expressed in IHKE-1 cells, we found that these cells predominantly express a splice variant of NPR-B (NPR-Bi). This receptor exhibits a modified guanylyl cyclase domain in which the last 84 amino acids at the carboxy terminus have been replaced by 31 unrelated amino acids. The NPR-Bi insertion contains the conserved intron/exon splicing sites and a lariat sequence (CTGAC), which is necessary for correct splicing (38). A phosphodiester bond is formed between the 5′ terminus of the intron at the RNA level and the 2′-hydroxy group of the adenosine nucleotide of the lariat sequence. Nevertheless, the position of the lariat sequence directly behind the splice donor site is somewhat unusual, if not sterically hindering (for the sequence, see Figure 4). Therefore, splicing may not be as accurate, and both receptor variants are possible. Such a transcript would give rise to a receptor that might still be able to bind CNP because of the unmodified extracellular domain but would not elicit a cGMP response. We also detected a signal for NPR-Bi in whole human kidney, which clearly indicates that this splice variant is not restricted to this cultured cell line. Because we also found a weaker signal for wild-type NPR-B in IHKE-1 cells, there are at least two possible reasons why we did not detect any increase in intracellular cGMP levels: (1) The mRNA is not translated into a functional NPR-B. (2) The wild-type NPR-B is located on the basolateral side of these cells and, because this membrane is barely accessible in our preparation, CNP could not bind this receptor. The aforementioned lack of an effect of CNP on cellular cGMP levels in rat proximal tubules, despite the presence of NPR-B, indicates that NPR-Bi is possibly present in these rat cells as well. ANP, URO, and BNP induced strong intracellular cGMP responses and depolarized Vm; therefore, we must assume that at least NPR-A is located in the apical membrane. The effect of CNP on Vm in these cells was blocked in the presence of genistein, an inhibitor of tyrosine kinases (23). If CNP binds to this receptor, it could be that this receptor splice variant has the capability of autophosphorylation and may act as a tyrosine kinase when regulating the K+ channel in these cells. Inhibition of a K+ channel in T lymphocytes via tyrosine kinase activation was recently shown (39). Because there was no significant change in [Ca2+]i attributable to CNP or any change in Vm attributable to activators of the cAMP pathway (forskolin or 8-Cl-cAMP) or protein kinase C (sn-1,2-dioctanoylglycerol), we can exclude participation of these candidates.
In conclusion, natriuretic peptides that bind to NPR-A regulate a K+ conductance in the apical membrane of human proximal tubule cells via a cGMP-dependent pathway. CNP uses a different, cGMP-independent pathway, which might involve tyrosine kinase activity. Its receptor (NPR-B) is coexpressed with a splice variant (NPR-Bi) that lacks a functional guanylyl cyclase catalytic domain but may act as a tyrosine kinase. The K+ conductance might be an important regulator of Na+ -dependent transport processes in these proximal tubule cells.
Acknowledgments
Acknowledgments
This study was supported by the Deutsche Forschungsgemeinschaft (Schl 277/5-1 to 5-4), the Alexander von Humboldt Foundation, the Zentrum für Innovative Medizinische Forschung Münster (Hi-1-1-II/96-7), the Danish Center for Respiratory Physiological Adaptation, and the Danish Cancer Society (Grants 95-100-40 and 78-5000). We gratefully acknowledge Dr. K. Adermann for help with the synthesis of natriuretic peptides. The authors thank Sabine Haxelmans, Ingrid Kleta, Melanie Klingenberg, Ulrike Opel, Daniela Rehder, and Heike Stegemann for excellent technical assistance.
Footnotes
American Society of Nephrology
The nucleotide sequence reported in this article has been submitted to the EMBL nucleotide sequence database (accession no. A3005282).
- © 1999 American Society of Nephrology