Harvest and primary culture of the murine aldosterone-sensitive distal nephron

Abstract

The aldosterone-sensitive distal nephron (ASDN) exhibits axial heterogeneity in structure and function from the distal convoluted tubule to the medullary collecting duct. Ion and water transport is primarily divided between the cortex and medulla of the ASDN, respectively. Transcellular transport in this segment is highly regulated in health and disease and is integrated across different cell types. We currently lack an inexpensive, high-yield, and tractable technique to harvest and culture cells for the study of gene expression and physiological properties of mouse cortical ASDN. To address this need, we harvested tubules bound to Dolichos biflorus agglutinin lectin-coated magnetic beads from the kidney cortex and characterized these cell preparations. We determined that these cells are enriched for markers of distal convoluted tubule, connecting tubule, and cortical collecting duct, including principal and intercalated cells. In primary culture, these cells develop polarized monolayers with high resistance (1,000-1,500 Ω * cm²) and maintain expression and activity of key channels. These cells demonstrate an amiloride-sensitive short-circuit current that can be enhanced with aldosterone and maintain measurable potassium and anion secretion. Our method can be easily adopted to study the biology of the ASDN and to investigate phenotypic differences between wild-type and transgenic mouse models.

the aldosterone-sensitive distal nephron (ASDN) is defined as the distal convoluted tubule through the connecting tubule to the cortical and medullary collecting duct and is responsible for several important ion transport pathways. The distal convoluted tubule transports sodium (Na⁺), potassium (K⁺), magnesium, and calcium and is divided both functionally and anatomically into the distal convoluted tubule, part 1 and part 2 (DCT1 and DCT2)(64). The connecting tubule (CNT) and cortical collecting duct (CCD) are composed mainly of two interdigitating cell types: 1) principal cells (PCs) and 2) intercalated cells (ICs). Principal cells are responsible for regulating Na⁺, K⁺, and water homeostasis (50), and the main carrier of Na⁺ in PCs is the epithelial sodium channel (ENaC). ENaC is stimulated by several hormones including aldosterone, is regulated in health and disease, and its importance in determining sodium and potassium balance is dependent on the activity of the Na-Cl cotransporter (NCC) in neighboring DCT cells (5). Principal cells also secrete potassium through renal outer medullary potassium (ROMK) and big potassium (BK) channels (50, 70, 71). Intercalated cells have traditionally been shown to mediate acid-base and potassium transport (23), but more recent studies demonstrate an additional role for these cells in mediating direct sodium reabsorption and in regulating sodium transport across PCs through ENaC (26, 58, 69). Thus the ASDN represents a tightly regulated, multifaceted, interdependent nephron segment.

Isolation and primary culture of nephron segments of the ASDN have primarily been restricted to the medullary collecting duct, and thus current methodologies limit better characterization of the roles of the cortical ASDN in physiological and pathophysiological states or from wild-type and transgenic mice (e.g., for gene expression, biochemical, cytological, and electrophysiological studies). Existing methodologies that isolate DCT or CCD include laser capture microdissection (12, 51) and transgenic mice expressing green fluorescent protein with Complex Object Parametric Analyzer and Sorter (COPAS) technology (36, 39). These techniques are not easily accessible to most laboratories and require technical expertise and/or specialized mouse strains. Laser capture microdissection also generates a low yield of cells that cannot be typically used for biochemistry. Moreover, existing technologies do not capture cells that represent the heterogeneity that is present in the ASDN. Thus more accessible methods that concomitantly produce a higher yield of primary cells from the ASDN are needed.

Different lectins bind specifically to different segments of the mammalian nephron (21, 30, 31) and thereby may be useful for isolation of cells from the ASDN. The Dolichos biflorus agglutinin (DBA) lectin has been used to localize and isolate collecting duct cells from the rodent medulla (17, 44, 62, 76, 77), but this reagent has not been previously validated for isolation of cortical cells. Herein, we describe the development and validation of a relatively inexpensive and simple method to harvest and culture epithelial cells from murine ASDN. We have performed biochemical, cytological, and electrophysiological assays to validate this method, and these preparations provide a model for quantitative, segment-specific studies.

MATERIALS AND METHODS

Animals.

Six- to eight-week-old C57BL/6 mice (Jackson Laboratories, Bar Harbor, ME) were provided food and water intake ad libitum and maintained through a 12:12-h light-dark cycle in a climate-controlled environment. The Institutional Animal Care and Use Committee at Stanford University approved the experiments, and mice were euthanized in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Antibodies.

We employed rabbit polyclonal antibodies for detection of endogenous aquaporin-2 (gift of Robert Fenton, University of Aarhus); Tamm-Horsfall protein (Santa Cruz Biotechnology, Santa Cruz, CA); NCC (gift of David Ellison and James McCormick, Oregon Health Sciences University); zona occludin 1 (ZO-1; Invitrogen, Camarillo, CA); acetylated tubulin (Sigma-Aldrich, St. Louis, MO); TRPV5 (Everest Biotech, Ramona, CA); SGK1 (Sigma-Aldrich); and β-actin (EMD Millipore, Billerica, MA). We also generated and validated a mouse monoclonal antibody to detect mouse α-ENaC (Fig. 1).

Fig. 1.

Characterization of α-epithelial Na channel (ENaC) antibody. A: epitope derived from N-terminal intracellular sequence, L(74)IEFHRSYRE(83), (arrow) predicting migration of ∼30- and ∼90-kDa bands for cleaved and uncleaved α-ENaC, respectively. Cleavage occurs on extrafacial loop (scissors). B: detection of transfected α-ENaC in HEK293T cells (mock transfection, left; cotransfection with Nedd4-2, middle; and alone, right). C: detection (arrow) of endogenous uncleaved α-ENaC in immortalized mpkCCD_c14 cells treated as indicated with vehicle or aldosterone (Aldo; 6 h) and vehicle, LY294002 (LY), or insulin for the final 2 h. Cleaved and uncleaved α-ENaC represented by arrows (black and gray, respectively). WCL, whole cell lysate.

Generation of α-ENaC antibody.

An immunogen was generated to the peptide sequence corresponding to amino acids 74–83, which is conserved from human to rat to mouse and doesn't share homology with β- or γ-ENaC. Peptide synthesis, purification, verification of peptide sequence (by mass spectrometry), conjugation to carrier (keyhole limpet hemocyanin), injection of mice (female BALB/c mice), formation of hybridoma cell lines, collection of ascites, and confirmation of specificity by ELISA were all carried out by a commercial vendor (Abmart, Beijing, China).

Immortalized cell culture.

Human embryonic kidney (HEK293T) cells and mouse polarized kidney cortical collecting duct (mpkCCD_c14) cells were maintained and cultured as described previously (8). MpkCCD_c14 cells were pretreated for 6 h with vehicle (100% ethanol) or 1 μM aldosterone (Sigma-Aldrich) with 2 final hours of vehicle (DMSO, Sigma-Aldrich), 50 μM LY294002 (EMD Millipore), or 100 nM insulin (Sigma-Aldrich).

Transient transfection of Nedd4-2 and α-ENaC.

HEK293T cells were transiently transfected as indicated with FLAG-tagged mouse Nedd4-2 with α (HA-tagged)-, β-, and γ-subunits of mouse ENaC as described previously (8). Twenty-four hours after transfection, cells were harvested in 500 μl lysis buffer containing protease and phosphatase inhibitors as described previously (18). Samples were separated by SDS-PAGE and immunoblotted with anti-α-ENaC antibody.

Western blotting: SDS-PAGE.

For immunoblots, samples lysed in lysis buffer {20 mM Tris·HCl, pH 8.5, 1% Nonidet P-40, 1% deoxycholate acid, 62.5% EDTA containing protease inhibitors [1 mM PMSF, 1 mM, benzamidine, and 1× Complete Protease Inhibitor Cocktail (Roche, Indianapolis, IN)] and phosphatase inhibitors [25 mM sodium fluoride, 2 μM microcystin-LR, and 1× Phosphatase-Inhibitor Cocktail Sets I and II (EMD Millipore)]} were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes and incubated with indicated antibodies and ECL (EMD Millipore). As indicated, we blotted endogenous 14-3-3 proteins as a loading control with anti-pan-14-3-3β (K-19) antibody (Santa Cruz Biotechnology).

Immunofluorescence microscopy.

Mice were anesthetized using inhaled isoflurane and perfused with chilled PBS (pH 7.4). Kidneys were removed, decapsulated, bisected sagittally, and incubated in 4% paraformaldehyde for 16 h overnight at 4°C. The tissue was then incubated in 30% sucrose in PBS for 24 h, embedded in Tissue-Tek optimal cutting temperature (OCT) compound (10.24% polyvinyl alcohol, 4.26% polyethylene glycol), frozen on liquid nitrogen-cooled N-methyl butane, sliced into 5-μm sections, mounted on superfrosted glass slides, and saved overnight at −80°C. We permeabilized slides in CSK buffer [50 mM NaCl, 300 mM sucrose, 10 mM PIPES (pH 6.8), 3 mM MgCl₂, 0.5% (vol/vol) Triton X-100] for 15 s then washed twice in PBS for 5 min and incubated in blocking solution (PBS containing 50 mM NH₄Cl, 25 mM l-lysine, 25 mM glycine, 0.2% BSA, 20% donkey serum) for 2 h at room temperature. Sections were then incubated with biotin-conjugated DBA (Vector Laboratories, Burlingame, CA) and/or primary antibody diluted in DAKO antibody diluent (Dako North America, Carpinteria, CA) overnight at 4°C. Sections were rinsed three times with PBS- 0.1% BSA and incubated with the appropriate secondary reagents [Alexa 488-conjugated donkey anti-rabbit antibody or Alexa 568-conjugated streptavidin (Life Technologies, Grand Island, NY)] for 2 h at room temperature. Slides were rinsed in PBS and mounted with Vectashield hardset media with 4,6-diamidino-2-phenylindole (DAPI; Vector Laboratories), and images were obtained using a confocal microscope (Leica SP2).

Isolation and culture of primary cell preparations.

We perfused mice with cold 1× PBS for 3–4 min, decapsulated the kidneys, coarsely dissected cortices, minced in modified Ringer's buffer (mRB; 118 mM NaCl, 16 mM H⁺-HEPES, 17 mM Na⁺-HEPES, 14 mM glucose, 3.2 mM KCl, 2.5 mM CaCl₂, 1.8 mM Mg₂SO₄, and 1.8 mM KH₂PO₄), and pelleted at 175 g for 8 min in a swinging-bucket centrifuge with an A-4-62 rotor (Eppendorf 5810R, Hauppauge, NY) at room temperature. Pellets were resuspended in digestion buffer [0.2% collagenase (Worthington Biochemical, Lakewood, NJ); 0.2% hyaluronidase in mRB (Sigma-Aldrich)] and incubated at 200 rpm for 45 min in a shaker (Thermo Scientific MAKQ420HP, Waltham, MA) at 37°C. To mechanically disrupt tubular fragments, we passed the pellet 10–15 times through flame-tapered 9-in. pasteur pipettes and then reincubated it with DNAse I (Life Technologies) for 25 min at 37°C. Glomeruli and undigested tubules were captured on a 40-μm cell strainer (BD Biosciences, San Jose, CA) and washed. The sieved cells and tubules were then centrifuged three times at 28 g for 3 min, resuspended in mRB, and rocked with biotin-conjugated DBA and streptavidin-linked beads (Dynabeads M-280 Streptavidin, Life Technologies) for 15 min at room temperature (ATR Rotamix, Laurel, MD). Magnetically bound cells/tubules were washed three times and eluted twice with 150 mM α-N-acetyl galactosamine (NAG; Sigma-Aldrich) in mRB for 45 min at room temperature. We separated eluted cells and tubules from magnetic beads and recovered cells by brief centrifugation. We then washed and resuspended eluted cells and tubules in mRB and plated them in DMEM with 10% FBS and penicillin/streptomycin (100 IU/ml/100 μg/ml, Mediatech, Manassas, VA). After 48 h, cells were trypsinized, and we seeded 150,000 cells/cm² onto collagen-coated filter cups in defined media [Renal Epithelial Cell Growth Basal Medium (REBM), Lonza, Allendale, NJ] with supplements, growth factors, and antibiotics (Bullet Kit: hydrocortisone, FBS, gentamicin, hInsulin, hEGF, transferrin, triiodothyronine, epinephrine, Lonza). After 7 days (5 days after replating), cell sheets were incubated for 24 h in basal REBM to deplete them of growth factors and supplements.

RNA extraction and semiquantitative PCR.

Total RNA from kidney, liver, DBA-bound, and unbound cell preparations was isolated using an RNeasy Mini kit per the manufacturer's instructions (Qiagen, Redwood City, CA). Reverse transcription was performed according to the manufacturer's instructions (SuperScript First-Strand Synthesis System for RT-PCR, Life Technologies) from 1 μg of total RNA using random hexamers. To study gene expression, we synthesized primers for nephron segment-specific markers using custom-designed (PrimerBlast tool, NCBI) or published sequences (Table 1). PCR reactions were performed with cDNA, 300 nM for each primer, and 2× Fast SYBR Green PCR master mix (Life Technologies). Fivefold serial dilution standard curves for all primers were made to test the efficiency of each primer pair. We performed reactions, including no-template controls, in triplicate in the StepOnePlus Real-Time PCR System (Life Technologies). To confirm specificity of amplicon products, a dissociation curve was performed for each PCR reaction. Complementary DNA from liver was used as a negative control for nephron segment-specific markers (data not shown). The ΔΔCt method was used to measure relative quantification (48) of each marker to β-actin.

Table 1.

Primers for semiquantitative PCR

Transcript	Forward Primer	Reverse Primer
NPHS1	TGCTGCCTTACCAAGTCCAG	GCTTCTGGGCCGGGTATTTT
SGLT2	TGGCGGTGTCCGTGGCTTGG	CGGACACTGGAGGTGCCAGATAGC
GLUT2 (65)	ATCGCCCTCTGCTTCCAGTAC	GAACACGTAAGGCCCAAGGA
UTA2	GCAATCTCGGGTTGCTTGGGCA	ACACGGCCATCAGGAGCCCCA
ROMK (4)	CCGTGTTCATCACAGCCTTCTT	CCGTAACCTATGGTCACTTGGG
NKCC2 (4)	CCGTGGCCTACATAGGTGTT	GGCTCGTGTTGACATCTTGA
NCC (73)	CAGTGCCTGGTGCTTACAGGGC	CATCATGCAGGACACCAATG
TRPM6 (36)	AAAGCCATGCGAGTTATCAGC	CTTCACAATGAAAACCTGCCC
α-ENaC (59)	TGCTCCTGTCACTTCAGCAC	CCCCTTGCTTAGCCTGTTC
Pendrin (68)	GCTGGCCTCATCTCAGCTG	GCAAGGGTTCCAGAAGCCT
AE1 (6)	AGGACCTGGTGTTGCCAGAG	CGGTTATGCGCCATGGA
AQP2 (4)	TCACTGGGTCTTCTGGATCG	CGTTCCTCCCAGTCAGTGT
β-Actin (59)	GGTCAGAAGGACTCCTATGTGG	TGTCGTCCCAGTTGGTAACA

See the text for definitions.

Immunocytochemistry.

All solutions were made in PBS-CM [PBS (pH 7.4) containing 1 mM CaCl₂ and 0.1 mM MgCl₂]. Cells were fixed in filter cups in 4% paraformaldehyde for 15 or 30 min at room temperature or 4°C, respectively, permeabilized in 0.2% Triton X-100 for 10 min, and blocked in 0.2% gelatin or 3% BSA with 0.1% Tween and 5% normal donkey serum in PBS-CM. Antibody dilutions were made in blocking solution and incubated for 30–60 min at 37°C in a humidified chamber for both surfaces. Filters were cut off from the cups and mounted in DAPI-mounting media. We imaged Z-slices using a Leica-SP2 confocal microscope.

Quantification of cell types in polarized monolayers.

Primary cells from three separate mice were cultured on permeable supports for 7 days as described above and then incubated for 24 h in basal REBM followed by 1 μM aldosterone for 6 h to enhance the sensitivity of detection of α-ENaC. Immunocytochemistry was performed as described above, with primary antibodies against NCC, TRPV5, and α-ENaC. Cells were counted from four representative images at ×400 total magnification per filter (mean cells counted per filter, 1,726 ± 85). Cells were classified as DCT if NCC(+); CNT principal cells if TRPV5(+), α-ENaC(+); CCD principal cells if NCC(−), TRPV5(−), and α-ENaC(+)(32); or other if not positively stained for any of the three markers.

Electrophysiology.

We measured potential difference and transepithelial resistance (R_te) using an Evometer “chopstick voltmeter” (World Precision Instruments, Sarasota, FL). Cell sheets were deemed viable once R_te exceeded 1,000 Ω·cm², which occurred 5–6 days after harvesting. One day before the experiments, cell sheets were incubated in basal REBM for 24 h, then treated with 1 μM aldosterone or an equivalent volume of vehicle (100% ethanol) 6 h before short-circuit current measurement (I_sc). Cell sheets were studied in Ussing chambers (Physiological Instruments, San Diego, CA), as described previously (53). Briefly, cell sheets were bathed (exposed area 1.2 cm²) in Krebs-Henseleit solution (in mM: 118 NaCl, 25 NaHCO₃, 1.2 KH₂PO_4, 4.7 KCl, 10 glucose, 1.2 CaCl₂, and 1.2 MgCl₂) and gassed with 95% oxygen and 5% CO₂ to produce a pH of 7.4 maintained at 37°C. Transepithelial voltage (V_te) was clamped to 0 mV, and the resulting I_sc was displayed continuously on a chart recorder. A set voltage pulse (0.5–2.0 mV) was applied across the cell sheet for 200 milliseconds every 1–20 s, and R_te was determined from the ratio of applied voltage to additional clamp current required. I_sc and R_te stabilized within the first 10 min of mounting of the tissue, at which point we added a series of pharmacological agents (Sigma-Aldrich) to the luminal side. Cell sheets were treated with 10⁻⁵ M amiloride to inhibit ENaC-mediated Na⁺ absorption, 10⁻³ M barium chloride to inhibit potassium channels, and/or 10⁻⁵ M ATP to stimulate calcium-dependent anion secretion. All agents were added as 1:1,000 dilutions of stocks in water. Vehicle controls had no effect on I_sc.

Statistical analysis.

Data analysis was performed using a paired or unpaired t-test using the Analysis Toolpack of Microsoft Excel. Results are provided as means ± SE; n = number of samples; N = number of mice; and we defined statistical significance at a P value <0.05.

RESULTS

DBA-linked biotin selectively binds the connecting tubule and collecting duct from the mouse kidney.

DBA, a lectin that binds terminal NAG, has previously been shown to bind the rat proximal tubule and rat/mouse collecting duct (21, 22, 30, 31), and DBA-biotin can be linked to streptavidin-coated magnetic beads for isolation of lectin-bound cells. To determine the utility of DBA-biotin to harvest cortical tubule cells, we first tested the specificity of DBA-biotin in an adult mouse kidney (Fig. 2). On longitudinal sections through a single tubule, DBA staining primarily colocalized with cortical tubules stained for aquaporin-2, a marker of principal cells in CNT and the cortical and medullary collecting duct (27, 45) (Fig. 2A). Similar to aquaporin-2 immunostaining, DBA staining was more pronounced in medullary than in cortical tubules. DBA-stained cortical tubules shared minimal overlap with markers for the thick ascending limb of the loop of Henle (Tamm-Horsfall protein) (29, 74) or for DCT (NCC) (7, 32) (Fig. 2, B and C, respectively). This expression profile predicts that cell preparations harvested from adult murine kidney cortex using DBA-conjugated beads would be enriched for CNT and CCD, but not thick ascending limb or DCT.

Fig. 2.

Dolichos biflorus agglutinin (DBA) lectin colocalizes with markers of the murine connecting tubule and collecting duct. Immunofluorescence microscopy is shown of representative adult mouse kidney sections with staining for α-N-acetyl galactosamine (NAG) using DBA-biotin and markers for various subtypes of tubular epithelial cells: aquaporin-2 (AQP2; A); principal cells: Tamm-Horsfall protein (THP; thick ascending limb; B); and Na-Cl cotransporter (NCC; distal convoluted tubule; C). Cortical and medullary sections are labeled as indicated. The red signal indicates DBA-bound NAG, and the green signal indicates colocalizing markers. The blue signal indicates nuclear staining (4,6-diamidino-2-phenylindole; DAPI). G, glomeruli in cortical sections. The scale bars are indicated in micrometers.

Gene expression of primary cell preparations.

As shown in Fig. 3A, we used DBA-conjugated magnetic beads to isolate cells/tubules from perfused adult mouse kidney cortex. We grossly separated cortical from medullary tissue and then added collagenase/hyalurondiase to minced cortical tissue. The resultant preparation contained a heterogeneous mixture of single cells, short tubules, longer undigested tubules, and glomeruli. We removed glomeruli and longer undigested tubules by sieving through a 40-μm strainer. Finally, we used DBA-conjugated magnetic beads to select DBA-expressing cortical cells. The protocol from death to elution requires 4–5 h.

Fig. 3.

Gene expression profile of stage-specific isolation of cell preparations using DBA-lectin. A: flow chart of isolation protocol from digestion of tissue to isolation and elution of DBA-bound cells using magnetic beads. B: real-time PCR relative quantification of segment-specific genes from whole kidney (light gray bars); DBA-bound cells (black bars); and unbound cells (white bars). Segment-specific genes are assigned as indicated. G, glomerulus; PT, proximal tubule; tDL, thin descending limb of loop of Henle; TAL, thick ascending limb; DCT, distal convoluted tubule; CNT, connecting tubule; CCD, cortical collecting duct; MCD, medullary collecting duct. *P < 0.05 between selected populations vs. whole kidney; N = 3 mice.

We compared relative gene expression profiles from this preparation to that from whole kidney (Fig. 3, B and C). The volume of cells bound to DBA beads was approximately one-tenth of the volume of unbound cells. Expression of nephrin (NPHS1), a marker for glomeruli (52), was low in DBA-bound and unbound preparations compared with whole kidney. Expression of markers for the medulla, including UTA2 (thin descending limb of the loop of Henle) (13) and aquaporin-2 (medullary >> CCD) (27), was also low in both DBA-bound and unbound preparations. In contrast, expression of markers for the proximal tubule, including SGLT2 and GLUT2 (65), was significantly enriched in unbound preparations. Surprisingly, expression of markers for the distal cortical tubule, including NCC (DCT1 and DCT2), TRPM6 (DCT1 > DCT2) (72), ENaC (DCT2 to principal cells of the CNT and CCD) (43), and pendrin (intercalated cells from CCD) (25, 40), were enriched in DBA-bound preparations. This was unexpected because DBA staining by immunofluorescence microscopy was much stronger in only CNT and CCD (Fig. 2). Markers that are not confined to the distal cortical tubule from DCT to CCD (e.g., ROMK and AE1) (34, 40) were present but not enriched in DBA-bound preparations. Notably, our DBA-bound preparations contained markers for DCT1 and DCT2 (NCC, TRPM6) and intercalated cells (pendrin, AE1), supporting our observation that we were digesting tubular fragments from the distal nephron rather than individual principal cells.

Primary culture of DBA-bound preparations demonstrates a tight, differentiated epithelial monolayer composed predominantly of principal cells from the CCD.

To characterize the electrophysiological properties of DBA-bound cell preparations, we examined whether these cells differentiate and polarize in cell culture. As depicted in Fig. 4A, we incubated DBA-bound preplated cells with trypsin and plated single cells on permeable supports to create a monolayer. As shown in Fig. 4B, transepithelial resistance rose significantly each day in these preparations. At day 7, we fixed and stained cells for markers of a differentiated, tight epithelium. As shown in Fig. 4, C and D, we stained cell monolayers with ZO-1 and α-tubulin to mark tight junctions and primary cilia, respectively (49, 61). Z-stack images show a polarized monolayer with a single cilium at the apical surface. We next evaluated the proportion of these cells that belong to each segment of the ASDN.

Fig. 4.

Development of a high-resistance, polarized monolayer of distal nephron cells in culture. A: flow chart of plating and culture of DBA-bound preparations. B: transepithelial resistance leading up to day of treatment; n = 34 wells; N = 4 mice. *P < 0.05 vs. the resistance on the prior day. Also shown is confocal microscopy of immunocytochemical stain of slides from a representative culture of cells from day 0 for zonula occludin (ZO)-1 (green; C), a marker for tight junctions, and acetylated tubulin (green; D), a marker for primary cilia. The red signal indicates DBA-bound NAG, and the blue signal indicates nuclear staining (DAPI). Images are shown in the xy-plane, and in D the xz-stack is shown below. The scale bars are indicated in micrometers. *P < 0.05 vs. the resistance on the prior day. E: individual (circles) and mean (rectangles) values for cells identified as DCT (light gray), CNT principal cells (dark gray), CCD principal cells (black; PC), and other (white); n = 4 images/mouse; N = 3 mice. *P < 0.05 vs. CCD principal cells.

As shown in Fig. 4E, we estimate that 61.4 ± 1.85% of cultured cells are principal cells from the CCD, and the remainder are DCT (4.8 ± 0.43%), principal cells from CNT (7.4 ± 0.57%), and other cell types (26.4 ± 1.43%). We next evaluated whether these cultured cells retain the function of channels from these segments of the nephron.

Primary culture of DBA-bound preparations retain expression and electrophysiological properties of the aldosterone-sensitive distal nephron.

At day 7, we incubated cells in basal media for 24 h and treated with either vehicle or 1 μM aldosterone. We then harvested these cells and probed for the aldosterone-stimulated genes SGK1 and α-ENaC (67). As shown in Fig. 5A, protein expression of both gene products markedly increased with aldosterone compared with vehicle treatment at 6 and 24 h. In Fig. 5B, we show the staining against α-ENaC in primary cell preparations treated with vehicle or aldosterone for 24 h, as indicated. In Ussing chambers, we detected a significantly increased amiloride-sensitive apical-to-basal sodium current (Fig. 5C) with 1 nM and 1 μM aldosterone- vs. vehicle-treated cells (1 nM, 32.8 ± 2.6; 1 μM, 45.7 ± 3.5 vs. vehicle, 28.2 ± 1.3 μA/cm²; P < 0.05) at 6 h (Fig. 5D).

Fig. 5.

Aldosterone responsiveness of primary distal nephron cultures. A: representative immunoblots from lysates of cells plated for 7 days, then incubated in depleted media for 24 h, followed by treatment with vehicle (V) or 1 μM aldosterone (A) as indicated. Arrows indicate expected bands for SGK1 and α-ENaC. B: confocal microscopy of representative immunocytochemical stain of primary cells treated with vehicle (top) or aldosterone (Aldo; bottom) for 24 h. The green signal indicates ZO-1; the red signal indicates α-ENaC; and the blue signal indicates nuclear staining (DAPI). The scale bars are indicated in micrometers. C: representative tracings of amiloride-sensitive short-circuit current from cells plated for 7 days, depleted for 24 h, treated with vehicle (light gray trace), 1 nM aldosterone (dark gray trace), or 1 μM aldosterone (black trace) for 6 h, and mounted in Ussing chambers. Cells were treated with amiloride as indicated by the black bar. The scale bar is indicated in seconds (sec). D: individual (circles) and mean (rectangles) values for amiloride-sensitive current from vehicle (light gray)-, 1 nM aldosterone (dark gray)-, or 1 μM aldosterone-treated (black) cells; n = 6 filters; N = 6 mice. *P < 0.05 vs. vehicle-treated cells. #P < 0.05 vs. 1 nM aldosterone-treated cells.

Interestingly, the electrical response of DBA cell preparations to aldosterone was dependent on tissue culture conditions. The aldosterone-sensitive induction of SGK1 and α-ENaC was observed in cells cultured in either REBM or mpkCCD_c14 media (3), a defined medium commonly used for immortalized principal cells. However, aldosterone-induced, amiloride-sensitive current was present only with incubation of cells in REBM (Fig. 6).

Fig. 6.

Comparison of aldosterone response using different defined media. DBA-isolated primary cells from mouse cortical collecting tubule were plated on semipermeable, collagen-coated filters, cultured in mpkCCD_c14 vs. Renal Epithelial Cell Growth Basal Medium (REBM) for 7 days, and corresponding depleted media for 24 h before treatment with vehicle (ethanol) vs. 1 μM aldosterone for 6 h. Shown are transepithelial resistance (A) and potential difference (B) for days before incubation in depleted media (day 0); n = 26–36 wells; N = 6 mice/group. C: change in equivalent current from 0 to 6 h with vehicle vs. aldosterone treatment for cells grown in mpkCCD_c14 (light gray bars) vs. REBM (black bars); n = 6–13 wells; N = 6 mice/group. D: representative immunoblots from lysates of cells plated for 7 days, then depleted for 24 h, followed by treatment with vehicle (V) or 1 μM aldosterone (A) as indicated. Arrows indicate expected bands for SGK1, α-ENaC, and 14-3-3. See materials and methods for details. #P < 0.05 vs. the measurement on the prior day for cells incubated in mpkCCD_c14 or REBM, respectively. *P < 0.05 between vehicle- and aldosterone-treated cells.

Potassium and anion currents of primary cell preparations.

To detect the presence of other ion currents from our primary cell preparations, as shown in Fig. 7, A and B, we also measured apical barium-sensitive currents that are likely mediated by ROMK channels on the apical surface (19, 70). Under short-circuit conditions, we detected no significant difference in aldosterone- vs. vehicle-treated cells in the presence of amiloride (1.4 ± 0.3 vs. 2.0 ± 0.5 μA/cm², P = 0.45). We also detected ATP-sensitive anion currents (Fig. 7, C and D), presumably due to chloride secretion (54), and these currents were not influenced by aldosterone under short-circuit conditions (54) (aldosterone, 0.9 ± 0.1; vehicle, 0.5 ± 0.5 μA/cm², P = 0.28). Potassium and anion currents were of comparable magnitude to those measured in other CCD cell lines (14, 54).

Fig. 7.

Potassium and anion currents of primary distal nephron cultures. A: representative tracings of barium-sensitive current from cells plated for 7 days, depleted for 24 h, treated with vehicle (light gray trace) or aldosterone (black trace) for 6 h, and mounted in Ussing chambers. Cells were treated with barium as indicated by the black bar. Tracings are normalized to baseline short-circuit current. B: individual (circles) and mean (rectangles) values for barium-sensitive current from vehicle (light gray)- or aldosterone-treated (black) cells. C: representative tracings of ATP-sensitive transient and sustained current from cells plated for 5 days, depleted for 24 h, treated with vehicle (gray trace) or aldosterone (black trace) for 6 h, and mounted in Ussing chambers. Cells were treated with ATP as indicated by the black bar. B: individual (circles) and mean (rectangles) values for ATP-sensitive sustained current from vehicle (light gray)- or aldosterone-treated (black) cells. The scale bars are indicated in seconds (sec); n = 3–4 chambers; N = 3 mice.

DISCUSSION

The ideal preparation of murine ASDN would be a primary culture of a heterogeneous mixture of cells with sufficient yield for gene expression, biochemical, cytological, and electrophysiological studies. The method for isolating and growing these cells would be efficient, tractable, and available for different types of transgenic lines or a treated or diseased mouse model. The protocol would be standardized and require no specialized technique or expensive equipment. The cells would need to retain their differentiated state as polarized renal tubular epithelia and be electrically active.

The DBA lectin binds NAG and has been described as a marker of the collecting duct in the developing and adult kidneys of several mammalian species, including the mouse, rat, and rabbit (21, 30, 31, 38, 57), and other investigators have used DBA lectin binding to isolate medullary collecting duct cells (17, 42, 62, 76, 77). We have advanced this technology and developed the use of DBA-biotin linked to streptavidin-coated beads for isolation and primary culture of adult mouse cortical distal tubule. To isolate cortical preparations, we crudely separated cortex from medulla, and we also took advantage of the relative resistance of medullary fragments to digestion, which has been attributed to the presence of desmosomes (1, 62, 75). Aquaporin-2 is more abundantly expressed in medullary than in cortical collecting duct (27); hence, when we isolated cortical tissue, we lost yield of the aquaporin-2-expressing cells. However, we retained DBA-bound cortical cells, which developed high-resistance epithelia and retained sensitivity to aldosterone.

Primary culture of these DBA-bound cortical cells recapitulates a polarized tubular epithelial monolayer ex vivo. First, primary cultures of cortical preparations express both ZO-1 and primary cilia, indicating that these cells do not dedifferentiate in cell culture. Moreover, the presence of primary cilia in our primary cells indicates that our preparations exhibit characteristics of renal cortex because inner medullary collecting duct cells have been reported to lack primary cilia (9). Second, these preparations exhibit high transepithelial resistance, comparable with those found in M-1, mpkCCD₁₄, or mCCD₁₁ cells (3, 16, 63). Last, these high-resistance monolayers demonstrate functional segment-specific sodium, potassium, and chloride channels. Taken together, primary cultures of these preparations represent viable, terminally differentiated distal cortical tubular epithelia.

These primary cultures were a heterogeneous collection of cell types, comprised mostly of principal cells of the CCD, but the proportion of each segment represented is consistent across preparations (Fig. 4E). Approximately one-quarter of the cells did not stain positive for markers of DCT, CNT, or CCD, and we presume that the majority of these unlabeled cells are intercalated cells. This would closely approximate the ratio of principal cells to intercalated cells in mouse CNT and CCD (24). We treated cells with aldosterone for 6 h to maximize detection of α-ENaC, but we cannot exclude that changes in the culture media could enhance expression, and therefore alter identification, of other cell types. For example, we did not treat cells with angiotensin II or hypotonic low-chloride buffer to stimulate NCC (36) or calciotropic hormones to stimulate TRPV5 expression (66).

The cortical ASDN is responsive to aldosterone in vivo (46), and an ideal preparation should retain this property ex vivo. Aldosterone treatment of our cell preparation induced protein expression of SGK1 and α-ENaC, comparable to what has been demonstrated in vivo (33). Aldosterone also increased amiloride-sensitive sodium current when measured either by short-circuit current in Ussing chambers (Fig. 5) or by equivalent current using chopstick voltmeters (Fig. 6). When measured in Ussing chambers, increased amiloride-sensitive current was observed with physiological concentrations of aldosterone (1 nM), similar to mCCD₁₁ cells (16). Interestingly, to retain electrical sensitivity to aldosterone, the cells require a defined medium that is slightly different from the common media used for growth of immortalized principal cells, mpkCCD_c14 or mCCD₁₁ (Fig. 6) (3, 16). We speculate that the higher baseline potential difference and hence, equivalent current, and the insensitivity to aldosterone observed with primary cells in mpkCCD_c14 media may have favored a DCT2/CNT phenotype. In contrast, cells incubated with REBM, which exhibited a lower baseline but higher aldosterone-stimulated sodium current, resembled a CCD phenotype (43). REBM has three known differences from mpkCCD_c14 media: 1) a lack of supplemental selenium; 2) substitution of dexamethasone with hydrocortisone; and 3) the presence of epinephrine. Hydrocortisone is less potent and less stable than dexamethasone (37), and norepinephrine has been shown to variably promote apical chloride secretion or sodium reabsorption in CCD cells (10, 35). Perhaps these differences may prime the cells for aldosterone-stimulated sodium uptake. The mechanism by which REBM promotes electrical sensitivity of our cell preparations to aldosterone is not known, but once discovered, may provide insight into the determinants that dictate axial heterogeneity between the DCT2/CNT and CCD.

Potassium secretion, a vital component of ion transport along the ASDN, has been reported in immortalized mCCD_c11 cells (14) and patch-clamp studies of CCD (15, 47), but not in primary cultures isolated with DBA beads or COPAS. Under short-circuit conditions, we obtained a barium-sensitive potassium current that was not directly modified by aldosterone, similar to that reported for immortalized principal cell preparations obtained from microdissected CCD (14). Further studies may determine the magnitude of potassium flux attributed directly or indirectly to aldosterone. The response to apical barium likely reflects secretion through ROMK, as opposed to iberitoxin-sensitive, barium-insensitive BK channels on the apical membrane. Since both apical and basal potassium channels in this CCD are mutated in human disease (14), our preparations may provide a novel culture system that can be used to study the regulation of potassium transport in the ASDN.

More specialized techniques such as microdissection (2, 12, 51, 56) yield a higher purity of specific nephron segments but compromise on time, yield, and feasibility for most investigators. The COPAS technique provides high yield and purity for DCT or CD (20, 36, 39) but requires the time and expense associated with breeding to fluorescent reporter mice (that may not be of similar background to one's target mouse line), specialized equipment, and technical personnel. The hypotonic lysis method for isolation of the medullary collecting duct is fairly inexpensive, but this method cannot be used for sensitive measurements (e.g., oxygen consumption) (62) and does not capture cortical segments of the ASDN. Patch-clamp studies of split-open murine CCD provide for precise electrophysiology but require trained personnel and cannot be used for gene expression, biochemical, or cytological studies (60). Tubular microperfusion is also a superior technique for segment-specific electrophysiology and flux studies (e.g., flow-mediated potassium secretion), but requires young mice, a trained perfusionist, has lower yield, and cannot capture the CNT or medullary CD due to the requirement for a single linear tubule (55).

Our method has several limitations. First, there is inherent variability with partial digestion of cortical tubules; however, we contend that sufficient digestion was achieved as our final preparation demonstrated depletion of medullary cell markers, (e.g., aquaporin-2), and enrichment of cortical markers, (e.g., α-ENaC and pendrin). Second, our technique is limited in cell specificity compared with more precise isolation methods such as microdissection or cell-sorting by COPAS. The presence of DCT cells in our preparation was unexpected based on the observation that DBA does not stain for nephron segments that express NCC. We infer that DCT cells were harvested by DBA-conjugated beads because bound tubules may contain DCT, or NAG that is present in DCT may bind DBA more efficiently under harvest vs. immunohistochemical conditions (41). With our culture conditions, we preserved only a minority of DCT cells compared with more specific techniques such as COPAS, and thus these cells are not a good model for studying NCC-mediated transport.

Despite these limitations, our technique can be used to characterize distal nephron function in wild-type or transgenic mice, and the yield of cells from this technique is sufficient for gene expression, biochemical, cytological, and electrophysiological studies. Moreover, the lack of specificity allows one to capture the entirety of the ASDN, from DCT to CCD, including intercalated cells; thus our method can be utilized to study the relationship between cell types in this highly regulated and interdependent stretch of the nephron. Our primary culture method may be superior to immortalized epithelia such as M-1 (63), mpkCCD_c14 (3), or mCCD_c11 (16) cells because it does not depend on the immortalization process, which may alter cell cycle or differentiation pathways. Immortalized cell lines also provide a high-yield continuous source of cells but are limited to one cell type and may not be useful to study the integrated ASDN from treated or transgenic mice. The addition of markers for DCT, including NCC and TRPM6, provide for an alternative to COPAS (36), mpkDCT (11), or mDCT₁₅ (28) cells for biochemical or immunocytochemical characterization of this segment. Previously available techniques have both strengths and limitations, but our preparation from the mouse kidney cortex using DBA-coated beads provides for a pragmatic approach.

In summary, we have optimized the protocol of Grupp et al. (17) using DBA-conjugated beads for an inexpensive, efficient, and simple technique that provides sufficient yield and purity for phenotypic characterization of ASDN from mice.

GRANTS

M. Labarca received support from the Becas Chile scholarship. J. M. Nizar is supported by the Tashia and John Morgridge Endowed Postdoctoral Fellowship and the Child Health Research Institute at Stanford University. E. M. Walczak is supported by the NIH (T32 DK007357). V. Bhalla has received support from the NIH (5R03 DK083613), and the American Society of Nephrology (Carl W. Gottschalk Research Grant), and currently receives support from the NIH (1R01 DK091565).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: M.L., J.M.N., E.M.W., and W.D. performed experiments; M.L., J.M.N., E.M.W., W.D., and V.B. analyzed data; M.L., J.M.N., E.M.W., A.C.P., and V.B. interpreted results of experiments; M.L., J.M.N., E.M.W., and V.B. prepared figures; M.L., J.M.N., E.M.W., A.C.P., and V.B. edited and revised manuscript; M.L., J.M.N., E.M.W., W.D., A.C.P., and V.B. approved final version of manuscript; V.B. provided conception and design of research; V.B. drafted manuscript.