Renal Physiology

Renal tubular epithelial cell apoptosis is associated with caspase cleavage of the NHE1 Na+/H+exchanger

Karen L. Wu, Shenaz Khan, Sujata Lakhe-Reddy, Liming Wang, George Jarad, R. Tyler Miller, Martha Konieczkowski, Arthur M. Brown, John R. Sedor, Jeffrey R. Schelling


Renal tubular epithelial cell (RTC) apoptosis causes tubular atrophy, a hallmark of renal disease progression. Apoptosis is generally characterized by reduced cell volume and cytosolic pH, but epithelial cells are relatively resistant to shrinkage due to regulatory volume increase, which is mediated by Na+/H+ exchanger (NHE) 1. We investigated whether RTC apoptosis requires caspase cleavage of NHE1. Staurosporine- and hypertonic NaCl-induced RTC apoptosis was associated with cell shrinkage and diminished cytosolic pH, and apoptosis was potentiated by amiloride analogs, suggesting NHE1 activity opposes apoptosis. NHE1-deficient fibroblasts demonstrated increased susceptibility to apoptosis, which was reversed by NHE1 reconstitution. NHE1 expression was markedly decreased in apoptotic RTC due to degradation, and preincubation with peptide caspase antagonists restored NHE1 expression, indicating that NHE1 is degraded by caspases. Recombinant caspase-3 cleaved the in vitro-translated NHE1 cytoplasmic domain into five distinct peptides, identical in molecular weight to NHE1 degradation products derived from staurosporine-stimulated RTC lysates. In vivo, NHE1 loss-of-function C57BL/6.SJL-swe/swemice with adriamycin-induced nephropathy demonstrated increased RTC apoptosis compared with adriamycin-treated wild-type controls, thereby implicating NHE1 inactivation as a potential mechanism of tubular atrophy. We conclude that NHE1 activity is critical for RTC survival after injury and that caspase cleavage of RTC NHE1 may promote apoptosis and tubular atrophy by preventing compensatory intracellular volume and pH regulation.

  • cell death
  • nephropathy
  • regulatory volume increase
  • renal disease
  • tubular atrophy
  • Na+/H+ exchanger 1

tubular atrophy is a hallmark of chronic renal diseases and is superior to glomerular pathology as a histological predictor of clinical outcomes (41). We have previously shown that renal tubular epithelial cell (RTC) apoptosis is a mechanism of tubular atrophy (20, 43). In the original descriptions of apoptosis morphology, Wyllie et al. (51) termed the process “shrinkage necrosis” due to reductions in cell volume observed during apoptosis. Apoptotic cell shrinkage is achieved by net loss of intracellular osmoles and H2O, as well as by caspase-dependent proteolysis of housekeeping and structural proteins, which mediates cell disassembly. Many studies have also demonstrated that apoptosis is associated with a decrease in cytosolic pH (16, 27, 28, 30, 36, 46), which is required for activation of the caspase cascade (30, 45).

In contrast to neuronal cells and lymphocytes, epithelium-derived cells are relatively resistant to apoptosis following exposure to hypertonic extracellular conditions (6, 34), due to an enhanced capacity to rapidly expand intracellular volume through regulatory volume increase (RVI) pathways (17, 26, 31,33). RVI is achieved by activation of the Na+/H+ exchanger isoform NHE1 and, depending on the cell type, the anion exchanger (AE) 2 isoform of the Cl/HCO 3 exchanger and/or the Na+/K+/2Cl symporter (26,31, 33). The net effect is ion and H2O influx, which leads to intracellular volume re-expansion. If RVI-dependent transporters are robustly activated after initiation of an apoptotic stimulus, restoration of intracellular volume may preempt apoptosis (26, 33). Alternatively, for a cell to undergo apoptosis, RVI must be overcome or inhibited (26,33). Neither AE2 nor the types 1 or 2 bumetanide-sensitive cotransporter (BSC-1 or BSC-2, respectively) isoforms of the Na+/K+/2Cl symporter are expressed in proximal tubule (1, 15, 18), the nephron segment that demonstrates the most abundant apoptosis in animal models of progressive renal disease (20, 43). However, NHE1 is ubiquitously expressed, including within the proximal tubule, suggesting that NHE1 may be critical to RTC survival by promoting resistance to apoptotic cell shrinkage.

In addition to regulating cell volume via RVI, NHE1 mediates other housekeeping functions, such as intracellular pH (pHi) regulation through electroneutral Na+ influx and H+ efflux. NHE1-dependent Na+/H+ exchange has been linked to essential cell functions, such as proliferation (4, 32), whereas diminished NHE1 activity has been associated with lymphocyte apoptosis (27, 40). Moreover, NHE1 has recently been recognized to function as a scaffold for binding with ezrin, radixin, and moesin (ERM) (14), adaptor molecules that link cytoskeleton to plasma membrane proteins, suggesting that NHE1 may facilitate apoptosis resistance by preserving cytoarchitecture and maintaining cell volume independent of Na+/H+ antiporter functions.

Because predicted sequelae of NHE1 inhibition-cell shrinkage, intracellular acidification, and cytoskeleton collapse mimic the apoptotic phenotype, we investigated whether RTC apoptosis is regulated by NHE1 caspase cleavage. We find that RTC apoptosis is associated with caspase-dependent NHE1 degradation. Furthermore, cell culture and whole animal data demonstrate that NHE1 loss-of-function mutations render RTC susceptible to apoptosis. The data are consistent with a mechanism whereby NHE1 degradation causes RTC apoptosis and tubular atrophy, which prevents RVI and promotes intracellular acidosis, an optimum condition for caspase activity.



We used the following: amiloride, 4′,6-diamidino-2-phenylindole (DAPI), ethyl-N-isopropylamiloride (EIPA), hexamethyleneamiloride (HMA), staurosporine (STS), adriamycin hydrochloride (Sigma, St. Louis, MO); z-VAD-fmk, z-DEVD-fmk, Ac-DEVD-CHO (Calbiochem, La Jolla, CA); 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF)-AM (Molecular Probes, Eugene, OR); anti-poly(ADP-ribose)polymerase (PARP) IgG, phycoerythrin (PE)-conjugated annexin V (Pharmingen, San Diego, CA); annexin V, anti-hemagglutinin (HA) IgG (Roche, Indianapolis, IN); horseradish peroxidase (HRP)-conjugated IgG (Santa Cruz Biotechnology, Santa Cruz, CA); green fluorescence protein (GFP) cDNA (Clontech, Palo Alto, CA); Red X-conjugated anti-mouse IgG, FITC-conjugated anti-mouse IgG (Vector Laboratories, Burlingame, CA); [35S]methionine (ICN, Irvine, CA); C57BL/6.SJL +/+, C57BL/6.SJL swe/+, and C57BL/6.SJL swe/swe mice (Jackson Laboratories, Bar Harbor, ME); COOH-terminal, HA epitope-tagged rat NHE1 cDNA (a gift from Dr. J. Orlowski, McGill University); and KR/A and E266I NHE1 mutant cDNAs (gifts from Dr. D. Barber, University of California at San Francisco). Rabbit polyclonal anti-NHE1 IgG was generated against an NHE1 cytoplasmic domain peptide as previously described (23) and affinity purified.

Cell lines.

The human renal proximal tubule (HRPT) cell RTC line (gift from Dr. L. Racusen, Johns Hopkins University) has been extensively characterized (20, 21, 39, 43). HRPT and HEK 293 (ATCC, Manassas, VA) cells were maintained in DMEM-F12 (Gibco-BRL, Rockville, MD) plus 10% fetal bovine serum (Hyclone, Logan, UT) and 1% penicillin-streptomycin-fungizone solution (Sigma). PS120 cells are genetically deficient for NHE1 expression and were derived from control CCL39 fibroblasts (gifts from Drs. D. Grall and J. Pouysségur, University of Nice).

Flow cytometry.

Cells were incubated with STS, lifted with trypsin-EDTA (10 min, 37°C), and washed twice in incubation buffer (10 mM HEPES/NaOH, pH 7.4, 140 mM NaCl, and 5 mM CaCl2) at 4°C. Washed cells were incubated with BCECF-AM (1.6 μM, 30 min, room temperature), PE-conjugated annexin V (15 min, room temperature), and DAPI (2 μg/ml, 15 min, room temperature). Cytosolic pH was measured by BCECF-AM fluorescence, and relative cell volumes were determined from forward vs. side light scatter characteristics with a Becton Dickinson FACS Vantage flow cytometer according to previously described methods (21, 50). Apoptosis was measured by annexin V binding (see below).

Apoptosis assays.

Cells were plated on glass coverslips at 0.25 × 106cells/ml density, grown to 80% confluence, and then maintained in serum-free medium combined with apoptotic stimuli. In some experiments, cells were preincubated with NHE1 inhibitors for 2 h or caspase inhibitors for 1 h before apoptosis induction. Apoptotic cells were identified by simultaneous fluorescent labeling of chromatin with DAPI and externalized phosphatidylserine with annexin V as previously described (20, 21). Random fields were viewed at ×40 magnification with a Nikon epifluorescence microscope (Tokyo, Japan), and the percentage of apoptotic cells was separately determined by two blinded observers from a total of 100–200 cells per experimental condition. Representative fields were photographed with a Spot Digital System camera and analyzed using Image Pro software (Diagnostic Instruments, Sterling Heights, MI).

Plasmid transfections.

Plasmids were transformed into DH-5α-competent bacterial cells according to manufacturer's protocol (Gibco-BRL), extracted using a Maxiprep kit (Qiagen, Valencia, CA), and amplified by culture in Luria-Bertani-ampicillin. GFP, HA-NHE1, KR/A (inhibits ERM binding), and E226I (inhibits Na+/H+ exchange) mutant NHE1 cDNAs were transiently transfected into cells according to previously described methods (21). Briefly, cells were plated in six-well dishes (0.25 × 106 cells/well) and cultured overnight in DMEM-F12 plus 10% fetal bovine serum to achieve 80% confluence. Cells were then washed and incubated with 100 μl of serum-free DMEM (Gibco-BRL) containing 6 μl of Fugene 6 transfection reagent (Roche) and 2.0–3.0 μg of plasmid DNA for 20 min at room temperature. Transfected cells were then cultured in complete media containing DMEM-F12 and 10% fetal bovine serum for an additional 24 h.

Immunoblot analysis.

Methods have previously been described in detail (42). Whole cell lysates were prepared in boiling 2× SDS sample buffer (125 mM Tris, pH 6.8, 2% SDS, 5% glycerol, 1% β-mercaptoethanol, and 0.003% bromphenol blue). Samples were assayed for protein content using protein assay reagents (Bio-Rad, Hercules, CA). Proteins were denatured by boiling for 5 min, and samples (60 μg/lane) were resolved by 8 or 14% SDS-PAGE (Novex, San Diego, CA). Proteins were transferred to polyvinylidene difluoride membranes, blocked with 5% nonfat milk, and incubated with either anti-PARP (1:2,000, 1 h, room temperature) or anti-HA (1:5,000, 1 h, room temperature) antibodies, followed by HRP-conjugated secondary antibody (1:5,000, 1 h, room temperature). Band intensity was detected by enhanced chemiluminescence methods (Amersham Pharmacia Biotech, Arlington Heights, IL) and exposure to Kodak Biomax ML film. In some experiments individual bands were digitized by phosphorimager (Molecular Dynamics, Sunnyvale, CA), quantified with Image Quant 5 software (Molecular Dynamics), and normalized to control values.

Protein degradation by [35S]-labeled pulse chase.

Cells were cultured to subconfluence, washed with PBS, and incubated with [35S]methionine in methionine-free DMEM (0.1 mCi/ml, 2 h, 37°C). Cells were washed with PBS and cultured in complete media (0–6 h, 37°C) with or without STS and peptide caspase inhibitors. Protein lysates (200 μg per sample) were immunoprecipitated with anti-HA (1 μg) or anti-NHE1 IgG (1 μg) and resolved by SDS-PAGE according to previously described methods (42). Autoradiograms were developed from dried gels. In some experiments, individual bands were digitized by phosphorimager (Molecular Dynamics), quantified with Image Quant 5 software (Molecular Dynamics), and normalized to control values.

Immunocytochemistry and fluorescence microscopy.

Methods have previously been described in detail (20, 21,43). Cells were maintained on sterile glass coverslips within six-well plates, fixed in paraformaldehyde (4%, 10 min, room temperature), and blocked with 5% low-IgG BSA and 0.2% Triton X-100 (Sigma) for 30 min at room temperature. Cells were incubated with anti-HA IgG (1:200, 2 h, room temperature), followed by either red X-conjugated or FITC-conjugated anti-mouse IgG (1:200, 2 h, 4°C). Negative controls were cells incubated with isotype-identical IgG, which was immunoreactive with an irrelevant epitope. Coverslips were mounted in antifade, aqueous media containing DAPI (Vectashield; Vector Laboratories) on standard microscope slides. Random fields were viewed by two observers blinded to experimental condition, using a Nikon epifluorescence microscope with appropriate fluorescence filters. Representative fields were photographed with a Spot Digital System camera and analyzed using Image Pro software.

Assay for caspase-3 cleavage of in vitro-translated NHE1.

The DNA template for in vitro translation was created by PCR amplification of the NHE1 cytosolic domain (cNHE1) from rat cDNA with upstream primer 5′-CTACCGCTC- GAGCCACCATGCCCAAGGACCAGTTCATCATTGCC-3′ that contains Xho I restriction endonuclease, Kozak and ATG start sites, and downstream primer 5′-TGCTCTAGACTAGCCCTGCCCTTTGGGGATGAAAGG-3′ containing an XbaI restriction site and stop codon. The cNHE1 construct included the open reading frame encoding amino acids 447–820, which corresponds to the 58 COOH-terminal amino acids within the transmembrane domain and the entire cytosolic domain. The resulting DNA was digested withXho I and Xba I (Gibco-BRL) and 1.1 kb product was cloned into pTNT vector (Promega, Madison, WI). PCR-generated cNHE1 nucleotide sequence was verified by automated sequencing (Cleveland Genomics, Cleveland, OH). [35S]Met-labeled cytoplasmic NHE1 substrate was generated using the reticulocyte lysate-based TNT Quick T7-coupled transcription/translation system (Promega) according to manufacturer's instructions. Briefly, cNHE1 plasmid template (1 μg) was labeled with [35S]Met (50 μCi, 90 min, 37°C). Autoradiograms from dried gels yielded a single 45-kDa band within 2- to 3-h film exposure (not shown). [35S]Met-labeled, in vitro-translated cNHE1 (3 μl) was incubated with or without Ac-DEVD-CHO (100 μM, 2 h, 37°C), followed by 1–2 μl of purified caspase-3 (6 h, 30°C; Pharmingen) in 5 μl of caspase buffer (100 mM HEPES, pH 7.5, 10% sucrose, and 0.1% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate) according to published protocols (10, 22). Peptide products were resolved by 14% SDS-PAGE and examined for cleavage by autoradiography.

Animal models.

C57BL/6.SJL swe/swe mice harbor an NHE1 A1639T point mutation, which introduces a premature stop codon, resulting in truncation between the 11th and 12th NHE1 transmembrane domains and loss of NHE1-dependent Na+/H+ activity (13). Four-week-old swe/swe homozygotes have a brain phenotype that includes ataxia and seizures, but a gross renal phenotype was not observed, as determined by kidney histology and serum Na+, K+, Cl, HCO 3 , urea nitrogen, creatinine, and albumin concentrations. To identify the role of NHE1 in tubulointerstitial disease susceptibility, we induced nephropathy in 8-wk-old wild-type (C57BL/6.SJL +/+), C57BL/6.SJL swe/+, and C57BL/6.SJL swe/swe mice by tail vein injection with adriamycin hydrochloride (10 μg/g) (49). Mice with a C57BL/6 genetic background were specifically chosen because C57BL/6.SJLswe/swe mice survive to adulthood, in contrast to SJL/Jswe/swe, which die at ∼3 wk (3, 13). Furthermore, in contrast to BALB/c mice, C57BL/6 mice do not develop adriamycin nephropathy (49), allowing for more robust comparisons between potentially susceptible (swe/swe,swe/+) and resistant (+/+) strains. Mice were killed 10 days after adriamycin infusion; swe/swe,swe/+, and +/+ animals were distinguished by immunoblotting liver lysates (20 μg protein/lane) with anti-NHE1 antibodies (1:1,000). RTC apoptosis from frozen kidney sections was determined by terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling (TUNEL) assays (Intergen, Purchase, NY) according to previously described methods (21, 43). Three sagittal sections from two mice per genotype were scanned for TUNEL-positive cells by two observers blinded to experimental conditions. Kidney area was determined with a Spot Digital System camera and Image Pro software, and data are expressed as apoptotic RTC/mm2. The animal care protocol was approved by the Institutional Animal Care and Use Committee at Case Western Reserve University School of Medicine.


Data are representative of three to five experiments per condition. Graphical results are expressed as means ± SE unless otherwise indicated. Comparisons between paired samples were made by the Student's t-test. Comparisons between groups containing more than two samples were made by one-way analysis of variance with the Bonferroni, Student-Newman-Keuls, or Kruskal-Wallis tests for multiple comparisons. Statistical significance is defined asP < 0.05.


Apoptotic RTC are shrunken and acidic.

To determine whether apoptotic RTC develop reduced cell volume and cytosolic pH, as has been described in leukocytes, we stimulated cultured RTC to undergo apoptosis with STS. Cell size and pH were determined by flow cytometry. As shown in Fig.1, a greater proportion of RTC incubated with STS displayed smaller cell volumes and lower cytosolic pH, similar to STS- and Fas-induced Jurkat T cell apoptosis (30). These data demonstrate that the apoptotic RTC phenotype includes shrinkage and acidification, consistent with a role for NHE1 inhibition.

Fig. 1.

Apoptotic renal tubular epithelial cells (RTC) are shrunken and acidic. Subconfluent monolayers of cultured RTC were incubated with or without staurosporine (STS, 1 μM, 5 h, 37°C). Apoptosis was determined by annexin V labeling (apoptotic cells are red, nonapoptotic cells are green). Relative cytosolic pH was measured by 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein-AM fluorescence (y-axis) and relative cell volumes by light scatter characteristics (x-axis). Results are representative of 5 separate experiments.

Hypertonicity stimulates RTC apoptosis.

Epithelial cells are relatively resistant to apoptosis induction by anisotonic conditions due to robust RVI (6,34), which is mediated by multiple transporters (26, 31,33). To determine whether cell stress imposed by hypertonicity causes apoptosis, we exposed RTC to increasing extracellular concentrations of the impermeant sugars mannitol and sucrose or NaCl and then simultaneously assayed them for apoptosis by annexin V labeling of externalized phosphatidylserine and chromatin condensation. Figure 2 shows that hypertonicity caused RTC apoptosis in response to all three stimuli in a concentration-dependent fashion, indicating that RVI was surmounted and RTC were susceptible to hypertonic stress-induced apoptosis.

Fig. 2.

Hypertonicity stimulates RTC apoptosis. RTC were incubated in media supplemented with mannitol, sucrose, or NaCl at indicated concentrations for 5 h. Apoptotic cells were identified by fluorescence microscopy analysis of annexin V labeling of externalized phosphatidylserine (A) and 4′,6-diamidino-2-phenylindole (DAPI) staining of condensed chromatin (B). Data represent means ± SE from 3–5 separate experiments.

Hypertonicity stimulates RTC apoptosis by an NHE1-dependent mechanism.

To determine the role of NHE1 in apoptosis induced by hypertonic conditions, we examined the effect of NHE1 inhibitors on RTC apoptosis, assayed by DAPI labeling of condensed chromatin (Fig. 3 A) and annexin V labeling of externalized phosphatidylserine (Fig. 3 B). These studies show that the Na+/H+ inhibitor amiloride caused modest apoptosis but significantly potentiated RTC apoptosis from hypertonic stress. Because amiloride, particularly at high concentrations, may inhibit multiple sodium transporters, including other NHE isoforms expressed in proximal tubule, NaCl-induced apoptosis was assayed following preincubation with NHE1-specific amiloride analogs EIPA and HMA (35, 47). Figure 3, C and D, demonstrates that, like amiloride, EIPA and HMA caused modest apoptosis and both inhibitors enhanced hypertonic NaCl-dependent apoptosis (HMA > EIPA), suggesting that NHE1 inactivation potentiates RTC apoptosis induced by stresses, such as hypertonicity.

Fig. 3.

Hypertonicity stimulates RTC apoptosis by an Na+/H+ exchanger (NHE) 1-dependent mechanism. RTC were preincubated with amiloride for 2 h at indicated concentrations (A and B) or NHE1 inhibitors ethyl-N-isopropyl-amiloride (EIPA, 5 μM, 2 h, 37°C) or hexamethylene amiloride (HMA, 100 μM, 2 h, 37°C;C and D) and then exposed to hypertonic media (DMEM-F12 + 300 mM NaCl, 5 h, 37°C). Apoptosis was determined by DAPI staining of condensed chromatin (A andC) and annexin V labeling of externalized phosphatidylserine (B and D) by immunofluorescence microscopy. Data are presented as means ± SE from 4 separate experiments. *P < 0.05 compared with control (no additions) by ANOVA; ‡P < 0.05 compared with media plus NaCl group by ANOVA.

Hypertonicity-induced RTC apoptosis is mediated by caspase-3 activation and decreased NHE1 expression.

To determine whether NHE1 is linked to caspase activation in apoptosis due to hypertonic stimuli, we incubated RTC with hypertonic NaCl and amiloride and then probed for activation of caspase-3, which is the final downstream executioner caspase in many apoptosis signaling cascades. Figure4 A shows that amiloride accentuated hypertonicity-induced cleavage of the caspase-3 substrate PARP, suggesting that NHE1 activity opposes apoptosis, perhaps by inhibiting decreases in cytosolic pH, which enhance caspase-3 activity (30, 45). Figure 4, B andC, demonstrates that hypertonic NaCl induction of caspase-3 activity was associated with diminished NHE1 expression, indicating that NHE1 may represent a caspase-3 target, which is consistent with the possibility that NHE1 dysfunction could contribute to the acidic and shrunken cell phenotype, as shown in Fig. 1.

Fig. 4.

Hypertonicity-induced RTC apoptosis is mediated by caspase-3 activation and decreased NHE1 expression. A: RTC were preincubated with amiloride at indicated concentrations for 2 h and then maintained in hypertonic media (DMEM-F12 + 300 mM NaCl, 5 h, 37°C). Whole cell lysates were probed for poly(ADP-ribose)polymerase (PARP) expression by immunoblot analysis. Caspase-3 activation was determined by PARP cleavage, which is indicated by the arrow. Human renal proximal tubule (HRPT) cells (B) or hemagglutinin (HA) epitope-tagged NHE1-transfected HEK 293 cells (C) were incubated in serum-free media (−) or in serum-free media plus 300 mM NaCl (+) for 6 h at 37°C. Whole cell lysates were immunoblotted with anti-NHE1 IgG (B,top) or anti-HA IgG (C, top). Lysates were blotted with anti-PARP IgG (B and C,bottom). The 85-kDa PARP cleavage product is marked by the arrow. Results are representative of 3 separate experiments.M r, molecular weight.

RTC apoptosis under isotonic conditions is associated with decreased NHE1 expression.

Because the data indicated that cell shrinkage-induced RTC apoptosis is mediated by NHE1 inhibition and caspase-3 activation, we queried whether NHE1 could be a caspase-3 substrate. To explore this possibility, we transiently cotransfected HEK 293 cells with an NHE1 cDNA construct containing a carboxy-terminal HA tag and GFP cDNA (to mark transfected cells), followed by STS incubation to stimulate caspase-3-dependent apoptosis. Nuclear morphology and NHE1 expression patterns were determined by standard, fluorescence microscopy. Figure 5, Aand B, demonstrates representative, transfected, apoptotic, and nonapoptotic cells. Approximately 20% of all cells underwent apoptosis, but only a small percentage of transfected, apoptotic cells expressed NHE1 on the cell surface (Fig. 5, C and E). Conversely, almost all transfected, nonapoptotic cells expressed NHE1 in a plasma membrane distribution (Fig. 5, C and E). Similar results were observed in RTC (data not shown). The results from these experiments suggest that NHE1 is cleaved during apoptosis and that loss of NHE1 expression renders cells susceptible to apoptosis.

Fig. 5.

RTC apoptosis under isotonic conditions is associated with decreased NHE1 expression. HEK 293 cells were cotransfected with carboxy-terminal, HA-tagged NHE1 and green fluorescence protein (GFP) cDNAs (1 μg/well with each vector) then incubated with STS (1 μM, 5 h, 37°C) to induce apoptosis. Fluorescence (not confocal) micrographs show transfected cells (green, A), apoptosis [fragmented nuclei stained with DAPI (blue)] (B), NHE1 expression by immunocytochemical staining with biotinylated anti-HA IgG and Texas red-conjugated streptavidin (red,C), and merged images (D) fromA–C. Results are representative of 3 experiments.E: 200 GFP-positive cells per experiment were scored for apoptosis, as defined by DAPI labeling of condensed chromatin and plasma membrane NHE1 expression, by immunocytochemical staining, as described in C. Quantitation of NHE1 expression in apoptotic vs. nonapoptotic cells is shown in E. Results represent means ± SE from 3 experiments. *P < 0.05 compared with GFP-positive, apoptosis-negative group by Student'st-test.

NHE1 reconstitution promotes resistance to apoptosis.

To further investigate the role of NHE1 in apoptosis, we compared NHE1-deficient PS120 cells, which were derived from Chinese hamster ovary fibroblasts (37), and NHE1-expressing control fibroblasts (CCL39 cells) for susceptibility to STS-induced apoptosis. As shown by DAPI and annexin V assays in Fig.6, A and B, respectively, apoptosis was observed in a significantly greater percentage of PS120 cells compared with the CCL39 cells, consistent with a recent report in these two cell lines (2). NHE1 function was subsequently addressed by add-back experiments, in which PS120 cells were transiently transfected with increasing concentrations of NHE1 then stimulated with STS and assayed for apoptosis. Figure 6 C demonstrates that NHE1 reconstitution in PS120 cells conferred resistance to apoptosis. At the highest NHE1 expression levels, apoptosis was equivalent to control CCL39 cells (Fig. 6, A and B).

Fig. 6.

NHE1 reconstitution promotes resistance to apoptosis. A and B: NHE1-deficient PS120 fibroblasts and NHE1-expressing CCL39 fibroblasts (control) were treated with STS (1 μM, 5 h, 37°C). Apoptosis was determined by annexin V labeling (A) and DAPI staining of fragmented nuclei (B). Results are means ± SE from 3 experiments. *P < 0.05 compared with PS120 + STS group by Student's t-test.C: PS120 cells were transfected with graded concentrations of HA epitope-tagged NHE1 cDNA (expressed as [NHE1 cDNA] in μg/well) followed by STS stimulation (1 μM, 5 h, 37°C). NHE1 expression is shown by immunoblot (top) and apoptosis of corresponding groups, as determined by annexin V labeling and quantitation by immunofluorescence microscopy (bottom). Histograms are presented as means ± SE. *P < 0.05 compared with untransfected, STS-stimulated cells by ANOVA. D: PS120 cells were transiently cotransfected with GFP (to mark transfected cells) and empty vector (Mock), wild-type NHE1 (WT), an NHE1 mutant that does not bind ezrin, radixin, and moesin (ERM) proteins (KR/A), or a Na+/H+ exchange-defective NHE1 mutant (E266I). Cells were then stimulated with STS (1 μM, 5 h, 37°C) to undergo apoptosis. %Apoptosis in the transfected cell subpopulation was determined by annexin V labeling and immunofluorescence microscopy. Data are presented as means ± SD from 3 experiments. *P < 0.05 compared with mock-transfected, STS-stimulated cells by ANOVA.

To assess which NHE1 domains are required for apoptosis resistance, we transiently transfected PS120 cells with wild-type NHE1, an NHE1 mutant that does not bind ERM proteins due to multiple K/A or R/A substitutions in residues 553–564 (KR/A) or an Na+/H+ exchange-defective NHE1 construct containing a point mutation in the third cytoplasmic loop (E266I). Significant differences in transfection efficiency were not observed between groups (not shown). Cells were then stimulated with STS and assayed for apoptosis. Similar to results in Fig. 6,A and B, PS120 cells were susceptible to apoptosis, which was reversed by transfection of wild-type NHE1 expression (Fig. 6 D). Expression of KR/A mutant NHE1 partially restored cell viability (Fig. 6 D), implying that NHE1-ERM binding is not the sole determinant of apoptosis resistance. Importantly, E266I expression did not rescue cells from apoptosis (Fig. 6 D), which indicates that Na+/H+ exchange is critical for apoptosis resistance.

RTC apoptosis leads to diminished NHE1 expression by protein degradation.

Because NHE1 has a long (∼24 h) half-life (9, 12) and is not significantly regulated by membrane cycling (38), rapid NHE1 disappearance with STS incubation suggests an NHE1 degradation mechanism, rather than suppressed synthesis. To determine more definitively whether STS-induced decreases in cell surface NHE1 were due to protein degradation, we conducted 35S-labeled pulse-chase experiments in HEK 293 cells transfected with HA-NHE1 cDNA and then induced them to undergo apoptosis with STS. Immunoprecipitation with anti-HA IgG revealed diminished35S-labeled NHE1 beginning 2.5 h after STS incubation, which was markedly more pronounced at 5–6 h (Fig.7), indicating that STS stimulates NHE1 degradation. 35S-labeled NHE1 levels did not appreciably change in unstimulated cells from 0 to 4 h, consistent with the long NHE1 half-life (not shown). More importantly,35S-labeled NHE1 levels were significantly greater in unstimulated compared with STS-stimulated cells at 4 h (not shown), further indicating that NHE1 is degraded with apoptosis. Because STS is a known caspase-3 activator, the data also suggest a caspase-3 mechanism of NHE1 degradation.

Fig. 7.

RTC apoptosis leads to diminished NHE1 expression by protein degradation. A: HEK 293 cells were transiently transfected with HA-tagged NHE1 cDNA, metabolically labeled with [35S]methionine, and then treated with 1 μM STS for indicated times. Cell lysates (200 μg/sample) were immunoprecipitated with anti-HA IgG and resolved by SDS-PAGE. Image represents autoradiogram from a dried gel. B: individual bands were digitized by phosphorimager, quantified, and normalized to control (time = 0) values. Results are means ± SE from 3 separate experiments. *P < 0.05 compared with control by ANOVA.

Although HA-tagged proteins have been successfully employed as caspase substrates (8, 11, 48), because the HA epitope sequence contains a putative caspase-3 cleavage site (YPYDPVDYA), pulse chase experiments were also conducted in untransfected RTC, followed by immunoprecipitation of endogenous NHE1 with rabbit polyclonal anti-human NHE1 IgG raised against the membrane-proximal cytoplasmic domain (23). These studies also revealed NHE1 degradation (Fig. 9 A), with a similar kinetic pattern as in Fig. 7, demonstrating that NHE1 is degraded during apoptosis.

NHE1 is degraded by caspase-3.

To determine whether apoptosis-associated NHE1 degradation is due to caspase cleavage, we transiently transfected RTC with carboxy-terminal HA, epitope-tagged NHE1, and stimulated them with STS to undergo apoptosis in the presence or absence of the cell-permeable, broad-spectrum peptide caspase inhibitor z-VAD-fmk or the peptide caspase-3 inhibitor z-DEVD-fmk. Whole cell lysates were probed for NHE1 expression and PARP cleavage by immunoblot analysis. Figure 8 A demonstrates that STS induced concomitant caspase-3 activity and NHE1 degradation in HEK 293 cells, and both caspase-3 activity and NHE1 degradation were partially inhibited by z-DEVD-fmk or z-VAD-fmk preincubation. Similar results were observed in RTC (Fig. 8 B). These data strongly suggest that NHE1 is cleaved by caspase-3 during apoptosis.

Fig. 8.

NHE1 is degraded by caspases. HEK 293 cells (A) or cultured RTC (B) were transiently transfected with carboxy-terminal, HA-tagged NHE1, preincubated with the caspase-3 inhibitor z-DEVD-fmk (100 μM, 1 h, 37°C) or the broad-spectrum caspase inhibitor z-VAD-fmk (100 μm, 1 h, 37°C), and then incubated with STS (1 μM, 5 h, 37°C). Whole cell lysates were probed for NHE1 expression by immunoblotting with anti-HA antibodies. Individual bands were digitized by phosphorimager, quantified, normalized to control = 100, and shown in histograms beneath the corresponding blots. Results are means ± SE from 3 separate experiments. Note thaty-axes are discontinuous. Blots were stripped and re-probed for PARP, with PARP cleavage product designated by the arrows. *P < 0.05 compared with control (no additions) by ANOVA.

To further define whether NHE1 is a caspase substrate, we metabolically labeled RTC, induced them with STS to undergo apoptosis, and analyzed anti-NHE1 immunoprecipitates for potential caspase cleavage products. Figure 9 A reveals simultaneous NHE1 degradation and appearance of several bands ranging in size from M r ∼15 to ∼32 kDa, suggesting that NHE1 is cleaved by caspase-3. To directly assess whether NHE1 is a caspase target, we incubated in vitro-translated cNHE1 with recombinant caspase-3 in the presence and absence of the caspase-blocking peptide Ac-DEVD-CHO and then evaluated it for degradation. As shown in Fig.9 B, in vitro degradation of cNHE1 by caspase-3 resulted in the generation of peptides identical in M rcompared with NHE1 peptides derived from STS-stimulated whole cell lysates (Fig. 9 A) in the absence, but not presence, of the peptide caspase inhibitor. A 37-kDa band was observed in control conditions (Fig. 9 B), which was not inhibited by Ac-DEVD-CHO, suggesting that this product was nonspecifically generated by caspase buffer alone. However, the other DEVD-inhibitable bands (shown by arrows) represent true caspase cleavage products. Together, data from Figs. 8 and 9 provide compelling evidence that NHE1 is a caspase-3 substrate.

Fig. 9.

A: [35S]methionine-labeled RTC, treated with 1 μM STS for indicated times, were immunoprecipitated with affinity-purified rabbit polyclonal anti-NHE1 IgG. B: [35S]methionine-labeled, in vitro-translated cNHE1 protein (3 μl) was preincubated with or without the peptide caspase inhibitor Ac-DEVD-CHO (DEVD, 100 μM, 2 h) and then with recombinant caspase-3 (casp-3, 1–2 μl, 6 h, 30°C) as described in materials and methods. The reaction product was resolved by 14% SDS-PAGE, and the dried gel was exposed to film for 3 h. The most prominent, common bands from A andB are demarcated by arrows. Results are representative of 3 separate experiments. cNHE1, NHE1 cytosolic domain.

NHE1 regulates RTC apoptosis in vivo.

To test the role of NHE1 inactivation as a mechanism of RTC deletion in an in vivo model of progressive renal disease, wild-type (C57BL/6.SJL+/+), NHE1 loss-of-function mutant (C57BL/6.SJLswe/swe), and heterozygote (C57BL/6.SJLswe/+) mice underwent adriamycin infusion to induce nephropathy (49). Kidneys removed 10 days postinfusion revealed little histopathological damage in any of the genotypes (data not shown). RTC apoptosis was rarely observed in wild-type controls (Fig. 10 A), consistent with a failure of adriamycin to cause nephropathy in mice with a C57BL/6 genetic background (49). However, significant increases in RTC apoptosis were observed in C57BL/6swe/swe and C57BL/6 swe/+ compared with C57BL/6+/+ mice (Fig. 10, B and C). Because neither C57BL/6 mice infused with adriamycin nor NHE1 loss-of-function mutant C57BL/6 swe/swe mice demonstrate histological or functional kidney abnormalities (13, 49), the data suggest that NHE1 inhibition unmasked a renal phenotype in adriamycin-treated C57BL/6 mice. In accordance with our previous reports demonstrating that RTC apoptosis precedes and contributes to tubular atrophy (20), these studies indicate that NHE1 confers cytoprotection, whereas loss of RTC NHE1 function is associated with apoptosis and tubular atrophy.

Fig. 10.

NHE1 regulates RTC apoptosis in vivo. Wild-type (C57BL/6 +/+), mutant C57BL/6 swe/swe, and C57BL/6 swe/+ mice were injected by tail vein with adriamycin (10 μg/g) and killed 10 days later. Mice were genotyped by immunoblotting liver lysates (20 μg protein/lane) with anti-NHE1 antibodies (C, top). RTC apoptosis was assayed by immunofluorescence terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling (TUNEL) labeling of frozen kidney sections. Representative fields (×20 magnification) are shown for C57BL/6 +/+ (A) and C57BL/6swe/swe (B) kidneys. Quantitative measures of apoptosis for each genotype are expressed as means ± SE and shown in C, bottom. *P < 0.05 compared with swe/swe and swe/+ groups by ANOVA.


NHE1 is ubiquitously expressed, and activation has been linked to vital housekeeping functions such as cell volume regulation (31,33) and growth factor-dependent proliferation (4, 32,44). Decreased NHE1 activity has been associated with lymphocyte apoptosis (27, 40), but a specific mechanism of NHE1 inhibition in apoptosis has not previously been described. A major finding in our study is that RTC NHE1 is inhibited by caspase cleavage. The results have potentially broad implications to disease pathogenesis, inasmuch as the data were generated in both epithelial and mesenchymal cells and in response to apoptosis induction by hypertonic stress or STS.

Many epithelium-derived cell lines were previously considered to be resistant to apoptotic cell volume reduction due to robust expression and function of RVI pathway components (6). However, our data demonstrate that RTC RVI can be overcome, permitting apoptosis to proceed. Of the transporters that mediate RVI, we focused on NHE1 partly because it is responsible for intracellular volume regulation, and an anticipated consequence of NHE1 inhibition, cell shrinkage, is a characteristic apoptotic feature. Indeed, we found that RTC apoptosis is associated with diminished cell volume, consistent with NHE1 inhibition. Our data do not exclude roles for regulatory volume decrease pathways, which can be activated in apoptosis (33). In addition, other transporters (7, 29, 52) may participate in RTC RVI triggered by apoptotic stimuli. For example, inhibition of NHE3, which is expressed on the apical proximal RTC membrane, could conceivably contribute to RTC shrinkage and acidosis. However, because hypertonic cell shrinkage suppresses NHE3 activity (19), we reasoned that apoptotic cleavage of NHE3 was unlikely to result in significant further suppression of transporter activity. Inhibition of other transporters that have been implicated in RVI, such as the AE2 Cl/HCO 3 exchanger or the BSC-1 and BSC-2 isoforms of the Na+/K+/2Clcotransporter, could also be involved in RTC apoptosis, by impeding intracellular volume expansion. However, prominent functions for these proteins in proximal RTC apoptosis are unlikely, since none are abundantly expressed in the proximal tubule (1,15, 18).

In addition to cell shrinkage, RTC NHE1 degradation was also associated with intracellular acidification (16). The data are in agreement with studies in leukocyte cell lines, which demonstrate that apoptosis is preceded by decreased pHi (16,27, 36). Importantly, NHE1 stimulation prevented intracellular acidification and abrogated apoptosis (16, 36), indicating NHE1 activation opposes apoptosis. Rich et al. (40) demonstrated that human leukemia cells exhibit a significantly higher pHi compared with normal leukocyte lineage cells. Moreover, decreased pHi was associated with increased apoptosis, which was exacerbated by exposure to the NHE1 inhibitor HMA. Barrière et al. (2) recently reported that apoptosis was also associated with intracellular acidification in Chinese hamster lung fibroblasts and that NHE1-induced increases in pHi were sufficient to prevent apoptosis. All of these findings are consistent with observations that the optimum pH for endonuclease and caspase activation is 6.3–6.8 and support the notion that intracellular acidification is critical for apoptosis execution (16,30). Grinstein's group (5, 25) has shown that expression of Δ566 NHE1 mutants (cytoplasmic domain deletion membrane distal to Met566) resulted in impaired osmoregulation, as well as a constitutively decreased resting pHi, suggesting that apoptotic NHE1 cleavage upstream from this domain would yield a similar phenotype. Together, our data in RTC are consistent with NHE1 inhibition contributing to both apoptotic acidification and intracellular volume dysregulation.

Although the role of NHE1 as a Na+/H+ exchanger is applicable to apoptotic cell volume decrease, a number of structural proteins must also be cleaved to achieve a shrunken cell morphology. Denker et al. (14) recently demonstrated that NHE1 is tethered to the plasma membrane and cytoskeleton through direct interaction with actin-binding ERM proteins. On the basis of this discovery, NHE1 would appear to function as a cytoarchitecture scaffold and require disassembly during apoptosis, consistent with ERM protein dissociation from plasma membrane during apoptosis (24). We reasoned that NHE1 may therefore confer apoptosis resistance by anchoring the actin cytoskeleton to the plasma membrane via interactions with ERM proteins. However, expression of NHE1 constructs with point mutations (KR/A) that abolish ERM binding (14) resulted in partial rescue of RTC from apoptosis, whereas expression of E266I mutants with intact ERM binding domains had no effect on apoptosis. Results from these studies suggest that mechanisms in addition to NHE1-ERM interactions must be required for apoptosis resistance. Furthermore, our data do not exclude the possibility that interactions between other NHE1 domains and the cytoskeleton may be important for maintenance of cell volume and resistance to apoptosis. Because the NHE1 KR/A mutant has normal Na+/H+ exchange activity (14), we conclude that Na+ influx and/or H+ efflux are critical NHE1 functions for apoptosis resistance.

Evidence to support the hypothesis that NHE1 is cleaved by caspases includes apoptosis-dependent loss of NHE1 expression due to protein degradation, rescue of NHE1 expression by preincubation with cell-permeable peptide caspase inhibitors, and direct cleavage of in vitro-translated NHE1 by caspase-3. Although consensus caspase cleavage sites are identified in the carboxy-terminal human NHE1 cytosolic tail (e.g., 755-DEED-758), deletion mutation studies predict that cleavage at this site would not result in Na+/H+exchange-dependent osmoregulatory dysfunction (5). Furthermore, cleavage at distal carboxy-terminal site(s) would result in generation of very small fragments and, therefore, would not account for the peptide band pattern observed in Fig. 9, indicating that cleavage at additional membrane-proximal, noncanonical sites is required. Further degradation of NHE1 and/or NHE1 caspase cleavage products by other protease pathways is also possible, although preliminary studies revealed that STS-induced changes in NHE1 expression are not altered by pretreatment with the proteosome inhibitors lactacystin, MG132, and PS-1 (Wu KL and Schelling JR, unpublished observations).

Although neither in vitro stimulus of apoptosis (STS, hypertonicity) is encountered by RTC in vivo, these agents were employed to mimic cell stresses that result in RTC apoptosis in vivo. We have previously shown that RTC apoptosis and tubular atrophy are caused by the in vivo stresses hypoxia and Fas activation in murine models of progressive renal disease (20, 43). In the current studies, the in vivo role of NHE1 was established by demonstration of increased RTC susceptibility to apoptosis in NHE1-deficient mice with adriamycin-induced nephropathy. To maximize the likelihood of detecting apoptotic RTC before obliteration of tubulointerstitial architecture by renal scarring, we killed mice after only 10 days. Because of the short observation interval, animals did not develop tubulointerstitial pathology, although we predict that the natural history of enhanced RTC apoptosis is tubular atrophy and interstitial fibrosis.

NHE1 is commonly referred to as a housekeeping protein, implying that it is pedestrian and unregulated. To the contrary, NHE1 has been shown to mediate vital cell functions, and the current studies establish a new role for NHE1 as a defender against RTC death. We speculate that in initial stages of RTC apoptosis in vivo, e.g., due to inflammation or uremia, NHE1 is likely to be activated in response to cell volume reduction cues. NHE1-dependent RVI may then be sufficient to prevent further cell volume shrinkage and perhaps even promote cell survival, provided the apoptotic stimulus is not too robust. However, once NHE1 is cleaved by caspases, the combined sequelae of NHE1 inhibition, cytosolic acidification, and cell shrinkage promote inexorable RTC apoptosis by optimizing pHi for further caspase activity and by bringing caspases in proximity to substrates via cell shrinkage.


We are grateful to Drs. Barber, Grall, Orlowski, Pouysségur, and Racusen for donation of reagents and to Dr. Eleanor Lederer for thoughtful observations and comments.


  • This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-54178, DK-38558, and DK-57933. J. R. Schelling is an Established Investigator of the American Heart Association.

  • Address for reprint requests and other correspondence: J. R. Schelling, Case Western Reserve Univ., MetroHealth Medical Center Campus, Rammelkamp Center for Education and Research, 2500 MetroHealth Dr., G531, Cleveland, OH 44109-1998 (E-mail: jrs15{at}

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

  • First published November 26, 2002;10.1152/ajprenal.00314.2002


View Abstract