INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE LABILE GAS nitric oxide (NO) is generated in the kidney and exerts direct effects on transport in several nephron segments. In proximal tubule cells, NO has been shown to inhibit both apical Na⁺-H⁺ exchange (32) and Na⁺-K⁺-ATPase activity (22) via pathways involving generation of cGMP. Thick ascending limb apical and basolateral Na⁺-H⁺ exchange is inhibited by NO, as is cortical collecting duct transport of sodium and water (9, 35). The inhibitory actions of NO on sodium and water transport along the nephron suggest that NO generated locally could promote natriuresis and diuresis. Consistent with this view, recent studies in the isolated perfused rat thick ascending limb showed that the NO donor spermine NONOate blocks chloride reabsorption by selectively inhibiting the apical Na⁺-K⁺-2Cl cotransporter, with no effect on Na⁺-K⁺-ATPase activity or apical K⁺ permeability (28). In isolated rat cortical thick ascending limbs, inhibition of endogenous NO production by the NO synthase (NOS) inhibitor N^G-nitro-L-arginine has been observed to stimulate chloride transport, suggesting that thick limb NO generation directly regulates cotransport (30).

Regulation of apical Na⁺-K⁺-2Cl cotransport is of critical importance in cells of the macula densa, which are involved in the tubuloglomerular feedback (TGF) response. Macula densa cells express high levels of the neuronal isoform of NOS (nNOS) (25), and several studies indicate that NO released by macula densa nNOS inhibits TGF (13, 16, 40). In rabbit afferent arterioles with attached macula densae, NO has been shown to act directly on macula densa cells to block TGF responses in a cGMP-dependent fashion (31, 39). Because NO donors significantly inhibit the Na⁺-K⁺-2Cl cotransporter in rat thick ascending limb (28), one possibility for the attenuating effect of NO on TGF is via inhibition of the apical Na⁺-K⁺-2Cl cotransporter (NKCC2) (41) in macula densa.

The purpose of the present studies was to investigate the effects of NO on Na⁺-K⁺-2Cl cotransport and signaling pathways for NO in a renal epithelial cell line (MMDD1) with properties of macula densa cells (42). We determined that these cells express both NKCC1 and NKCC2 isoforms of the Na⁺-K⁺-2Cl cotransporter, and we demonstrated that NO potently inhibits both apical and basolateral cotransport. Furthermore, we uncovered a novel role for cytochrome P-450 (also referred to as P-450) in mediating responses to NO in these cells.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture. A renal epithelial cell line, MMDD1, with properties of macula densa cells, was used for these studies, kindly provided by Dr. J. Schnermann (National Institutes of Health, Bethesda, MD). These cells were derived from SV40 transgenic mice, using fluorescent-activated cell sorting of renal tubular cells labeled with segment-specific lectins (42). The cell line has been shown to express well-known markers of macula densa, such as cyclooxygenase 2 (COX-2), nNOS, ROMK, and NKCC2. In the present studies, MMDD1 cells at passages 5-20 were routinely trypsinized and suspended in DMEM nutrient mixture-Ham's F-12 (DMEM/F-12) supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 µg/ml). The cells were plated onto six-well culture dishes and incubated at 37°C in a humidified atmosphere of 95% room air-5% CO₂. For experiments involving measurement of apical and basolateral Na⁺-K⁺-2Cl cotransport, cells were plated onto six-well cell culture inserts (0.4-µm pore size, Falcon, VWR Canlab, Mississauga, ON, Canada). The media was changed every 2 days, and once the cells reached confluence (typically in 3-4 days), the cells were maintained in serum-free DMEM/F-12 for 24 h before experiments.

RT-PCR for nNOS, NKCC1, NKCC2, and cytochrome P-450 2J5. Total RNA was isolated from MMDD1 cells, using a commercial kit (RNeasy, Qiagen, Chatsworth, CA). RNA (1 µg) was treated with DNase I and then reverse-transcribed using random hexamers (2.5 µM) and murine leukemia reverse transcriptase (RT; 2.5 U/µl) (Gene Amp RNA PCR core kit, Applied Biosystems Roche Molecular System, Branchburg, NJ). To control for possible genomic DNA contamination, all experiments included a reaction in which RT was omitted from the transcription buffer. PCR amplification was carried out in 100 µl of a solution containing 2.5 U ampliTaq DNA polymerase, 2 mM MgCl₂, 1× PCR buffer II (Gene Amp PCR core kit), and 1 µM of the sense and antisense oligonucleotide DNA primers for the cDNA of interest. For nNOS, the sense primer was 5'-ACTGCTCAAAGAGATAGAACC-3' and the antisense primer was 5'-GGTAATTCTCCAAACCGAG-3', corresponding to nucleotides 695-716 and 1201-1219 of the murine nNOS cDNA, respectively (26), and predicting amplification of a 525-bp cDNA product. For NKCC1, the sense primer was 5'-CAGAGACAATTTGAAGACC-3' and the antisense primer was 5'-CCAACGTCAGCATGAGAT-3', corresponding to nucleotides 5305-5323 and 5968-5985 of the murine NKCC1 cDNA, respectively, and predicting a PCR product of 681 bp (6). For NKCC2, the sense primer was 5'-CCCGCTCTCTTGGATATATAAC-3' and the antisense primer was 5'-GCTCCGAACAAATTCTTCCG-3', corresponding to nucleotides 2301-2321 and 2790-2809 of the murine NKCC2 cDNA, respectively (14), and predicting a PCR product of 509 bp. RT-PCR specific for the B isoform of NKCC2 [found in macula densa and thick ascending limb (41)] was performed with sense primer 5'-GTGTGATTATCATCGGCTTAGC-3' and antisense primer 5'-ATCAAGCCTATTGCACCACC-3', corresponding to nucleotides 853-874 and 977-1006 of the B isoform of murine NKCC2 and predicting a PCR product of 154 bp (14). For cytochrome P-450 isoform CYP2J5, the sense primer was 5'-ATCAGAGAAGCGAAAAGAATGTAG-3' and the antisense primer was 5'-TACTCAGTCTTAGTCTCCTTTACC-3', corresponding to nucleotides 1549-1572 and 1810-1833 of the murine cytochrome P-450 2J5 cDNA, respectively, and predicting a PCR product of 285 bp (21). PCR was routinely performed for 35 cycles, with a hot-start at 95°C for 5 min, followed by cycles at 95°C for 60 s, 58°C for 30 s, and 72°C for 60 s, followed by extension at 72°C for 10 min. DNA samples were electrophoresed on 2% agarose gels stained with ethidium bromide. DNA sequencing was performed by the DNA Sequencing Facility, University of Ottawa.

Measurement of Na+-K+-2Cl cotransport activity. The activity of Na⁺-K⁺-2Cl cotransport was measured by the uptake of ⁸⁶Rb⁺, in the presence of ouabain and in the presence or absence of the inhibitor bumetanide, essentially as described (12). Briefly, confluent cells were rinsed twice in Earle's salt solution consisting of (in mM) 140 NaCl, 5 KCl, 1 MgSO₄, 1.8 CaCl₂, 5 glucose, and 25 HEPES (pH 7.40 with Tris), and then incubated in this solution for 30 min at 37°C. Cells were then incubated for an additional 15 min in this solution, supplemented with or without agonists/antagonists, at room temperature. For assay of Na⁺-K⁺-2Cl cotransport, the incubation solution was then changed to one supplemented with 1 µCi/ml ⁸⁶Rb⁺ (Amersham, Oakville, ON, Canada), with 2 mM ouabain, and with or without 6 µM bumetanide (Sigma, St. Louis, MO), at room temperature. After 5 min, the uptake was terminated by removal of the solution, followed by four rapid washings with ice-cold 100 mM MgCl₂, buffered with 10 mM HEPES (pH 7.4 with Tris). Culture dishes were air-dried, and cells were solubilized in 1 N NaOH with 0.1% SDS, followed by measurement of radioactivity by liquid scintillation spectrometry. For cells grown on six-well cell culture inserts, activity of apical or basolateral Na⁺-K⁺-2Cl cotransport was determined by adding ⁸⁶Rb⁺ to the apical or basolateral surfaces, respectively, in the presence or absence of bumetanide (6 µM). Bumetanide (6 µM) was added and maintained in the media bathing the opposing cell membrane. Protein content was measured by the Bradford assay method, using BSA as standard (Bio-Rad, Montreal, QC, Canada). Activity of Na⁺-K⁺-2Cl cotransport is the ouabain-insensitive, bumetanide-sensitive component of ⁸⁶Rb⁺ uptake and is expressed as picomoles of ⁸⁶Rb⁺ per milligram of protein per 5 min, with all experiments performed in duplicate. In preliminary experiments, Na⁺-K⁺-2Cl cotransport activity was linear for 30 min under these assay conditions.

Measurement of cAMP and cGMP. MMDD1 cells were grown to confluence on 24-well plastic dishes and then incubated for 24 h in serum-free DMEM/F-12 medium. For assays of cAMP or cGMP, cells were incubated at 37°C for 15 min in DMEM/F-12, supplemented with 3-isobutyl-1-methylxanthine (5 × 10⁴ M), 0.5% BSA, in the presence or absence of agonists. Medium was then aspirated and replaced with ice-cold 10% trichloroacetic acid (TCA; vol/vol). After a further 30 min, samples were extracted four times with 4 vol of water-saturated ether and brought to pH 7.0 with Tris. Aliquots were assayed for cAMP or cGMP, using radioligand competitive binding assay kits, containing [³H]cAMP (Intermedico, Markham, ON, Canada) or [³H]cGMP (Amersham), as we performed previously (5, 32).

Measurement of cytosolic calcium concentration. Cytosolic calcium concentration ([Ca²⁺]_i) was measured in MMDD1 cells, essentially as we described (5). Cells were grown to confluence on glass coverslips and loaded with 5 µM of fura 2-acetoxymethyl ester (fura 2-AM) in the presence of 0.005% Pluronic-F127 for 45 min at room temperature. Cells were then continuously perfused at 37°C with a solution consisting of (in mM) 105 NaCl, 24 NaHCO₃, 2 Na₂HPO₄, 5 KCl, 1.0 MgSO₄, 1.5 CaCl₂, 4 lactic acid, 5 glucose, 1 alanine, and 10 HEPES (pH 7.3), as well as 0.2% BSA, and NO donors were added at various times. Fluorescence was measured from dual monochromators set at 340 and 380 nm, using a computer-linked analytic system (Photon Technology International, South Brunswick, NJ). Fluorescence emission was measured by photon counting, and the corrected fluorescence emission intensity ratio, from 340- and 380-nm excitation, was monitored continuously in selected cells and used as an indicator of [Ca²⁺]_i.

Western blotting of ERK1 and ERK2. Phosphorylation of ERK1 (p44) and ERK2 (p42) was measured by Western blotting. After incubation with or without donors of NO for 15 min, cells were lysed in an ice-cold buffer consisting of 50 mM Tris · HCl (pH 8.0), 150 mM NaCl, 1% Nonidet P-40, 1 mM EDTA, 1 µg/ml leupeptin, 1 µg/ml pepstatin, 1 µg/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, and 1 mM sodium fluoride. After quantification of proteins, equal amounts of protein lysates (10 µg) were run on 15% SDS-polyacrylamide gels and transferred to nitrocellulose membranes (Bio-Rad). The membranes were incubated in TBS-T buffer [10 mM Tris (pH 7.6), 150 mM NaCl, 0.05% Tween 20, 0.5% skim milk] for 1 h at room temperature. The membranes were then incubated for 16 h at 4°C with 1:500 dilution polyclonal antibody to phosphorylated rat ERK1 and ERK2 (Santa Cruz Biotechnology, Santa Cruz, CA), followed by incubation with 1:1,000 dilution of anti-rat secondary antibody conjugated to horseradish peroxidase (Amersham). After membranes were washed, phosphorylated proteins were detected by enhanced chemiluminescence (Amersham) on Hyperfilm (Amersham). Prestained standards were used as molecular weight markers (Bio-Rad), and to control for protein loading, all membranes were stripped and reprobed with polyclonal antibody to unphosphorylated rat ERK1 and ERK2 (Santa Cruz Biotechnology). Signals on Western blots were quantified by densitometry and corrected for unphosphorylated ERK1 and ERK2 levels, using an image analysis software program (Kodak Densitometer 1S440CF).

Eicosanoids. The epoxyeicosatrienoic acids (EETs) and 20-hydroxyeicosatetraenoic acid (20-HETE) were obtained from Sigma as HPLC-purified compounds.

Data analysis. Results are presented as means ± SE of experiments performed in duplicate. Significance was determined by Student's t-test and by ANOVA for cases with multiple comparisons. Significance is considered as P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Characterization of MMDD1 cells. The cell line used in these studies has previously been shown to express a number of specific markers for macula densa, including nNOS, NKCC2, and COX-2. In our laboratory, we confirmed expression of nNOS by RT-PCR, with restriction endonuclease digestion with BglII generating fragments of the expected sizes (400 and 125 bp) (Fig. 1A). RT-PCR also revealed the presence of NKCC2 (Fig. 1B). DNA sequencing of the NKCC2 PCR product confirmed its identity to the mouse NKCC2 isoform (14). RT-PCR was also performed using primers that generated a 154-bp product specific for the B isoform of murine NKCC2 (not shown) (14). DNA sequencing of this product confirmed the presence of the B isoform in MMDD1 cells. In contrast, RT-PCR using 5'-primers specific for the A and F isoforms of NKCC2 did not generate PCR products, although both products were amplified from mouse kidney RNA (not shown). RT-PCR also demonstrated the presence of NKCC1 mRNA in these cells (Fig. 1C), confirmed by DNA sequencing.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Expression of neuronal nitric oxide synthase (nNOS), NKCC1, and NKCC2 mRNA in MMDD1 cells by RT-PCR. A: representative agarose gel showing DNA size ladder (lane 1), 525-bp nNOS PCR product (lane 2), and 400- and 125-bp nNOS PCR products following digestion with BglII. B: representative agarose gel depicting DNA size ladder (lane 1), 509-bp PCR product for NKCC2 in MMDD1 cells (lane 2), reaction lacking reverse transcriptase as negative control (lane 3), 509-bp PCR product for NKCC2 in mouse kidney cortex as positive control (lane 4), and corresponding reaction lacking reverse transcriptase (lane 5). C: representative gel showing DNA size ladder (lane 1), 681-bp product for NKCC1 in MMDD1 cells (lane 2), corresponding reaction lacking reverse transcriptase (lane 3), 681-bp PCR product for NKCC1 in mouse kidney cortex (positive control) (lane 4), and corresponding reaction lacking reverse transcriptase (lane 5).

Effect of NO on Na+-K+-2Cl cotransport. NO has been shown to inhibit Na⁺-K⁺-2Cl activity in the rat thick ascending limb (28) and to inhibit the Na⁺H⁺ exchanger in proximal tubule and thick ascending limb (9, 32). We first determined the effects of NO donors on Na⁺-K⁺-2Cl activity in MMDD1 cells grown on plastic dishes. Both sodium nitroprusside (SNP) and S-nitroso-N-acetylpenicillamine (SNAP) caused concentration-dependent inhibition of ouabain-insensitive, bumetanide-sensitive ⁸⁶Rb⁺ uptake (Fig. 2). Significant inhibition was observed at concentrations of SNP greater than or equal to 10⁶ M and at concentrations of SNAP greater than or equal to 10⁷ M. SNP or SNAP had no significant effect on the bumetanide-insensitive component of ⁸⁶Rb⁺ uptake (not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Dose-dependent inhibition of Na+-K+-2Cl cotransport by sodium nitroprusside (SNP) and S-nitroso-N-acetylpenicillamine (SNAP) in MMDD1 cells. Cells were grown to confluence on plastic culture dishes and incubated with varying concentrations of SNAP (A) or SNP (B), followed by assay of bumetanide-sensitive 86Rb+ uptake, in the presence of ouabain. Results are means ± SE of experiments performed in duplicate. *P < 0.001 vs. control (C); **P < 0.05 vs. C; n = 6.

5

4

1

1

4

5

1

1

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Effect of scavenging of nitric oxide (NO) with hemoglobin (HB) on SNP- and SNAP-mediated inhibition of cotransport. Bumetanide-sensitive 86Rb+ uptake, in the presence of ouabain, was measured in cells grown on plastic dishes. A: effect of preincubation with HB (5 × 105 M) on SNAP (104 M)-mediated inhibition of cotransport. B: effect of HB (5 × 105 M) on SNP (104 M)-mediated inhibition of cotransport. Results are means ± SE of experiments performed in duplicate. *P < 0.005 vs. C; n = 4.

4

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Inhibition of apical and basolateral Na+-K+-2Cl cotransport by SNAP in MMDD1 cells. Cells were grown to confluence on cell culture inserts, followed by measurement of apical or basolateral bumetanide-sensitive 86Rb+ uptake, in the presence of ouabain and in the presence or absence of SNAP (104 M). A: apical cotransport activity (n = 4). B: basolateral cotransport activity (n = 5). Results are means ± SE of experiments performed in duplicate. *P < 0.03 vs. C.

6

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Effect of the inhibitor of nNOS, 7-nitroindazole (7-NI), on apical and basolateral Na+-K+-2Cl cotransport. MMDD1 cells were grown on cell culture inserts and incubated with the nNOS inhibitor 7-NI (106 M) before measurement of bumetanide-sensitive 86Rb+ uptake. A: apical cotransport activity. B: basolateral cotransport activity. Results are means ± SE of experiments performed in duplicate. *P < 0.03 vs. C; **P < 0.05 vs. C; n = 4.

Signaling pathways for inhibition of Na+-K+-2Cl cotransport by NO. In many cells, NO stimulates soluble guanylate cyclase, leading to elevation of cGMP levels. In MMDD1 cells grown on regular plastic dishes, SNP or SNAP had no significant effect on cGMP levels, at concentrations up to 10³ M. As a positive control, atrial natriuretic peptide (ANP; 10⁶ M) caused an eightfold increase in cGMP release in these cells (Fig. 6A). Similarly, neither NO donor had any significant effect on cellular levels of the second messenger cAMP (Fig. 6B).

View larger version (10K): [in this window] [in a new window]   Fig. 6.   Effect of NO donors on cGMP and cAMP in MMDD1 cells. A: cells were incubated with varying concentrations of SNP or SNAP for 15 min, followed by assay of cGMP levels. As a positive control, cells were incubated with atrial natriuretic peptide (ANP; 106 M). B: cells were incubated with SNP or SNAP (103 and 104 M) for 15 min, followed by assay of cAMP levels. As a positive control, cells were incubated with forskolin, an activator of adenylate cyclase (105 M). Results are means ± SE of experiments performed in duplicate (n = 4). *P < 0.001 vs. C.

3

4

8

View larger version (19K): [in this window] [in a new window]   Fig. 7.   SNAP has no effect on cytosolic calcium concentration in MMDD1 cells. Cells were loaded with fura-2 AM as described in MATERIALS AND METHODS, followed by administration of SNAP (104 M) at the indicated point. As a positive control, bradykinin (108 M) caused a significant increase in cytosolic calcium. Shown is a representative tracing of experiments in which either SNP or SNAP was administered (n = 4).

View larger version (29K): [in this window] [in a new window]   Fig. 8.   SNP and SNAP do not affect phosphorylation of ERK1 and ERK2. Shown is a representative immunoblot (n = 4) for phosphorylated ERK1 (44 kDa) and ERK2 (42 kDa) from cell extracts (105g) derived from cells incubated with SNP or SNAP (104 M). No significant effect on phosphorylation was observed, and there was also no effect on levels of nonphosphorylated ERK1 or ERK2 by immunoblot analysis (not shown).

3

4

1

1

4

3

1

1

3

4

5

View larger version (25K): [in this window] [in a new window]   Fig. 9.   Effect of inhibition of cytochrome P-450 on SNAP-mediated inhibition of cotransport. Cells were preincubated with the P-450 inhibitors 1-aminobenzotriazole (ABT; 103 M; A and B) or ketoconazole (Keto; 1.5 × 105 M; C and D) before determination of the effect of SNAP (104 M) on cotransport. A and C: apical cotransport activity. B and D: basolateral cotransport activity. Results are means ± SE of experiments performed in duplicate. *P < 0.001 vs. C; **P < 0.005 vs. C; n = 5 (ABT) or n = 3-4 (Keto).

Effect of EET and 20-HETE on Na+-K+-2Cl cotransport. To determine possible mediators of the effect of cytochrome P-450 on inhibition of cotransport by NO, cells were incubated for 15 min with metabolites of P-450. The EETs 8,9-EET and 11,12-EET (5 × 10⁶ M) had no significant effect on apical or basolateral cotransport activity (Fig. 10). In contrast, in cells grown on plastic, incubation with 14,15-EET caused a concentration-dependent inhibition of total Na⁺-K⁺-2Cl cotransport (Fig. 11A). In cells grown on cell culture inserts, both apical and basolateral cotransport was also inhibited by preincubation with 14,15-EET (5 × 10⁷ M) (Fig. 11, B and C).

View larger version (12K): [in this window] [in a new window]   Fig. 10.   Effect of 8,9-epoxyeicosatrienoic acid (EET) and 11,12-EET on Na+-K+-2Cl cotransport. MMDD1 cells were grown on cell culture inserts and were incubated for 30 min with 8,9-EET or 11,12-EET (5 × 106 M) before assay of cotransport activity. Cells were incubated with SNAP (104 M) as a positive control. A: apical cotransport activity. B: basolateral cotransport activity. Results are means ± SE of experiments performed in duplicate. *P < 0.005 vs. C; n = 3.

View larger version (12K): [in this window] [in a new window]   Fig. 11.   Inhibition of Na+-K+-2Cl cotransport by 14,15-EET. A: dose-dependent inhibition of cotransport in MMDD1 cells grown on plastic dishes. Cells were incubated with varying concentrations of 14,15-EET for 30 min before assay of cotransport. B: effect of 14,15-EET (5 × 107 M) on apical cotransport in cells grown on culture inserts. C: effect of 14,15-EET (5 × 107 M) on basolateral cotransport in cells grown on culture inserts. Results are means ± SE of experiments performed in duplicate. *P < 0.001 vs. C; **P < 0.05 vs. C; n = 3 (A) and n = 8 (B and C).

5

View larger version (13K): [in this window] [in a new window]   Fig. 12.   Effect of 20-HETE on Na+-K+-2Cl cotransport in MMDD1 cells. Assay of Na+-K+-2Cl cotransport was performed on confluent cells grown on cell culture inserts following preincubation with 20-HETE (5 × 106 M). A: apical cotransport activity. B: basolateral cotransport activity. Results are means ± SE of experiments performed in duplicate. *P < 0.005 vs. C; n = 8.

View larger version (11K): [in this window] [in a new window]   Fig. 13.   Effect of the -hydroxylase inhibitor HET0016 on SNAP-mediated inhibition of apical Na+-K+-2Cl cotransport. Bumetanide-sensitive 86Rb+ uptake was measured in confluent cells grown on cell culture inserts after preincubation with the selective -hydroxylase inhibitor HET0016 (105 M), with or without SNAP (104 M). Results are means ± SE of experiments performed in duplicate. There was no significant difference between the effect of SNAP and SNAP + HET0016; n = 8. *P < 0.001 vs. C; **P < 0.01 vs. C.

View larger version (37K): [in this window] [in a new window]   Fig. 14.   Expression of cytochrome P-450 (CYP) 2J5 in MMDD1 cells by RT-PCR. Representative agarose gel showing DNA 100-bp ladder (lane 1), 285-bp CYP 2J5 PCR product from MMDD1 cells (lane 2), corresponding reaction lacking reverse transcriptase (lane 3), 285-bp PCR product for CYP 2J5 from mouse liver (lane 4), and corresponding reaction lacking reverse transcriptase (lane 5).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Two isoforms of the electroneutral, loop diuretic-sensitive Na⁺-K⁺-2Cl cotransporter have been identified and are encoded by separate genes. The NKCC1 isoform is found on the basolateral membranes of secretory epithelia, and in the kidney it is localized to outer and inner medullary collecting ducts, glomerular and extraglomerular mesangium, and to afferent arteriole (18). The NKCC2 isoform is found exclusively in the kidney, and in the rabbit and mouse it consists of splice variants (A, B, F), with the B variant expressed specifically on the apical surface of cells of the cortical thick ascending limb and macula densa (41). The mouse renal epithelial cell line (MMDD1) used in the present study was derived from SV40 transgenic mice and retains some well-known properties of macula densa, such as expression and activity of NKCC2, COX-2, and nNOS (42). We confirmed these properties, and we demonstrated expression of the B variant of NKCC2 in these cells, and not the A or F variants, suggesting derivation of these cells from macula densa or cortical thick ascending limb (42). We also demonstrated that these cells express NKCC1. When MMDD1 cells were grown on cell supports, bumetanide-sensitive ⁸⁶Rb⁺ transport was present on apical and basolateral membranes, indeed with a relative abundance of basolateral cotransport. Our major finding is that NO inhibits apical and basolateral Na⁺-K⁺-2Cl cotransport activity in these cells. The inhibitory mechanism does not involve generation of cGMP, but our data suggest a novel pathway, with a requirement for cytochrome P-450 epoxygenase activity.

In macula densa, transport through the apical membrane Na⁺-K⁺-2Cl cotransporter is thought to be the initiator of the TGF response, resulting in afferent arteriolar constriction. The nNOS isoform is highly expressed in macula densa, and studies show that NO produced in the macula densa blunts TGF (13, 16, 37, 40). Because our data show that the NKCC2 B isoform is present in MMDD1 cells, the results are consistent with the hypothesis that NO derived from macula densa or adjacent thick limb cells in vivo could act in an autocrine or paracrine fashion to inhibit apical cotransporter activity. In recent studies in rabbit afferent arterioles with attached macula densae, inhibition of macula densa nNOS with 7-NI increased TGF responses, and it was also suggested that NO derived from the thick ascending limb could act as a paracrine factor to inhibit TGF at the macula densa (39). Similarly, our data in MMDD1 cells indicate that inhibition of nNOS with 7-NI caused a small but significant increase in cotransporter activity. Autocrine inhibition of apical cotransport by NO could therefore represent a mechanism for precise modulation of the TGF mechanism, independently of the direct inhibitory effects of NO on afferent arteriolar tone.

Although previous studies examined effects of NO donors and endogenous NO on apical Na⁺-K⁺-2Cl cotransport (NKCC2), there is little information on the effects of NO on basolateral cotransport via NKCC1. In rat aortic smooth muscle, NO inhibits basal NKCC1 activity and reduces its phosphorylation (1). In our studies, MMDD1 cells expressed abundant basolateral cotransport activity. Earlier studies by Lapointe and colleagues (3, 19) did not identify basolateral cotransport activity in rabbit macula densa. Although the presence of NKCC1 on the basolateral membrane of murine macula densa cells is possible, it is surprising that basolateral cotransport activity far exceeded apical cotransport in our study. Accordingly, it is possible that the MMDD1 cells have undergone dedifferentiation, with new expression of NKCC1, or they may contain nonepithelial cells or cells from tubular segments known to express NKCC1.

In many cell types, including proximal tubule, thick ascending limb, and cortical collecting duct, NO stimulates soluble guanylate cyclase, leading to production of cGMP (8, 27, 32, 34). In these segments, inhibition of transport pathways by NO is at least partly dependent on generation of cGMP, and recent in vivo studies in rats indicate that increases in renal interstitial cGMP levels promote natriuresis (17). In thick ascending limb, NO inhibits chloride reabsorption by activation of cGMP-stimulated phosphodiesterase, which decreases cAMP levels (27). In contrast, our data indicate that NO had no effect on cGMP or cAMP levels in MMDD1 cells. ANP significantly stimulated cGMP production. The effects of ANP on juxtaglomerular cell renin release have been studied in animal models (10, 38), although there is presently no information on direct effects of ANP on macula densa function. Our data suggest that MMDD1 cells express ANP receptors and also contain the particulate, but not the soluble, guanylate cyclase. In contrast, studies in an isolated rabbit afferent arteriole preparation by Ren et al. (31, 39) showed that NO produced by the macula densa acts on macula densa cells to inhibit TGF, via a pathway involving activation of guanylate cyclase and generation of cGMP. It is possible that macula densa responses to NO may be species specific or that MMDD1 cells may have reduced expression of guanylate cyclase. However, the potent inhibition of cotransport by NO in the present study does not exclude the possibility that in vivo cGMP generation may also induce inhibition.

There has been recent interest in the effects of NO that are independent of cGMP production, including effects of NO metabolites such as peroxynitrite. In the present studies, NO did not alter cell levels of calcium. Of interest, previous studies suggest that the B₂ isoform of the bradykinin receptor is not expressed in human macula densa (33). However, MMDD1 cells demonstrated reproducible calcium transients in response to bradykinin, suggesting expression of bradykinin B₁ or B₂ receptors in these cells. In mesangial cells in culture, NO inhibits mechanical stretch-induced ERK activity and nuclear translocation (15). Although we did not measure ERK activity directly, we observed no effect of NO on basal phosphorylation of ERK1 or ERK2 in MMDD1 cells.

Our studies uncovered a unique signaling pathway for NO in MMDD1 cells. ABT, an agent that blocks the renal metabolism of arachidonic acid to EETs and 20-HETE by cytochrome P-450 (36), completely reversed the inhibitory effect of exogenous NO on apical and basolateral Na⁺-K⁺-2Cl cotransport, suggesting that cytochrome P-450 is involved in NO signaling. Similarly, the inhibitor of P-450 epoxygenase ketoconazole reversed the inhibitory effect of SNAP. Neither ABT nor ketoconazole alone affected cotransport activity. Because incubation with 7-NI caused small but significant increases in apical and basolateral cotransport, this suggests that either ABT or ketoconazole does not completely inhibit P-450-mediated production of EETs at baseline, induced by endogenous NO, but instead blocks NO-stimulated P-450 activity. In this regard, incubation of rat renal microsomes with ABT at concentrations of 10 mM or greater was required to induce 80% inhibition of P-450 in vitro (24). An alternate possibility is that basal NO production in these cells inhibits cotransport activity via a pathway separate from that involving activation of P-450 by exogenous NO.

Although both apical and basolateral cotransport activities were inhibited by NO and this was reversed by inhibition of cytochrome P-450, and mimicked by 14,15-EET, only apical cotransport was inhibited by the arachidonic acid -hydroxylase product 20-HETE. In rabbit thick ascending limb, 20-HETE has also been shown to inhibit Na⁺-K⁺-2Cl cotransport (7). However, in MMDD1 cells, the inhibitor of -hydroxylase HET0016 did not have a significant effect on SNAP-mediated inhibition of apical Na⁺-K⁺-2Cl cotransport, whereas ABT and ketoconazole completely blocked the inhibitory effect of SNAP. Taken together, these data suggest that the inhibitory effect of NO may be mediated by selective activation of cytochrome P-450 epoxygenase activity in these cells, with no effect on the -hydroxylase activity. The data also suggest that NKCC1 and NKCC2 have distinct regulatory mechanisms with regards to these arachidonic acid metabolites.

Along the nephron, P-450-derived eicosanoids, including EETs and 20-HETE, have been shown to influence sodium and water transport. NO has been shown to inhibit cytochrome P-450 isoenzymes of the 1A, 2B1, 3C, and 4A classes by forming iron-nitrosyl complexes at the heme binding sites (2, 29). Recent in vivo studies in rats suggest that the renal vascular effects of NO are mediated via inhibition of P-450 and activation of guanylate cyclase, whereas the natriuretic effects are independent of P-450 (20). Our data suggest that in MMDD1 cells, NO may indeed activate a cytochrome P-450 isozyme with selective inhibition of the cotransporter by 14,15-EET.

Although kidney microsomal content of P-450 activity is significant, knowledge of kidney P-450 isoforms is incomplete. In mouse kidney, the CYP2J5 isoform is expressed in abundance in proximal tubules and collecting ducts (21). This isoform is associated with production of EETs, of which the 14,15-EET is predominant (21). Our studies demonstrate that mRNA for this isoform is present in MMDD1 cells. Further studies are required to determine whether the CYP2J5 isoform is expressed in macula densa in vivo and whether NO activates this enzyme, leading to production of EETs.

In summary, we showed that NO inhibits apical and basolateral Na⁺-K⁺-2Cl cotransport activity in MMDD1 cells, independently of guanylate cyclase activation. The inhibitory mechanism may involve a cytochrome P-450-dependent pathway, possibly mediated by production of 14,15-EET.