Optic nerve head astrocytes become abnormal in eyes that have elevated intraocular pressure, and cultured astrocytes display altered protein expression after being subjected for ≥1 days to elevated hydrostatic pressure. Here we show that 2-h elevated hydrostatic pressure (15 or 30 mmHg) causes phosphorylation of ERK1/2, ribosomal S6 protein kinase (p90RSK), and Na/H exchanger (NHE)1 in cultured rat optic nerve head astrocytes as judged by Western blot analysis. The MEK/ERK inhibitor U0126 abolished phosphorylation of NHE1 and p90RSK as well as ERK1/2. To examine NHE1 activity, cytoplasmic pH (pHi) was measured with BCECF and, in some experiments, cells were acidified by 5-min exposure to 20 mM ammonium chloride. Although baseline pHi was unaltered, the rate of pHi recovery from acidification was fourfold higher in pressure-treated astrocytes. In the presence of either U0126 or dimethylamiloride (DMA), an NHE inhibitor, hydrostatic pressure did not change the rate of pHi recovery. The findings are consistent with NHE1 activation due to phosphorylation of ERK1/2, p90RSK, and NHE1 that occurs in response to hydrostatic pressure. These responses may precede long-term changes of protein expression known to occur in pressure-stressed astrocytes.
- extracellular signal-regulated kinase 1/2
- pH recovery
glaucoma is characterized by progressive, irreversible vision loss that results from retinal ganglion cell death. Disease advancement is marked by gradual deformation of the optic nerve head, the site where ganglion cell axons exit the eye. Glaucoma is commonly, but not exclusively, associated with an increase in intraocular pressure (IOP) (11), and optic nerve head remodeling may be a response to chronically elevated IOP and mechanical deformation. Elevated IOP causes changes in gene expression at the optic nerve head in rat model of glaucoma (21). In primates, alterations in the expression of matrix metalloproteinases (MMPs) and their inhibitors (tissue inhibitors of metalloproteinases, TIMPs) occur in the optic nerve head of experimental glaucoma (1). These changes, which are likely to contribute to remodeling of the glaucomatous optic nerve head, are particularly prominent in optic nerve head astrocytes (2). Altered protein expression was also reported in human glaucomatous optic nerve head astrocytes (3).
Astrocytes are the principal glial cells that regulate the extracellular environment around retinal ganglion cell axons (8). Throughout the central nervous system (CNS), important roles played by astrocytes include regulation of extracellular potassium and pH (38). In the CNS, injury or stress causes normally quiescent astrocytes to become reactive, displaying altered morphology and protein expression, most notably increased glial fibrillary acidic protein (GFAP) (13). Astrocytes are known to respond to a number of different stresses including injury (27), endothelin-1 (19), and oxygen-glucose deprivation (14), and it is possible that such stress factors cause astrocytes to become reactive in the glaucomatous eye. It is interesting that altered elastin expression by optic nerve astrocytes occurs in response to IOP elevation but not axon injury (40), suggesting that at least some astrocyte responses may be specific consequences of a hydrostatic pressure challenge. There have been several studies in which isolated optic nerve astrocytes have been shown to respond to an increase in hydrostatic pressure. After ≥1 days, hydrostatic pressure treatment was found to cause increased expression of nitric oxide synthase-2 in human optic nerve head astrocytes (29). After 24 h, pressure-treated astrocytes display increased expression of heat shock proteins (42), common in cells responding to a stress. DNA microarray analysis points to multiple genes that display either increased or decreased expression in response to elevated hydrostatic pressure (59).
ERK/MAPK is activated chronically in reactive CNS astrocytes (33). Increased ERK/MAPK activity also was observed due to acute focal injury to cultured astroglial monolayer, an effect detectable within 2–10 min and persisting for 4–8 h (34). In the eye, ERK abundance appears abnormally high in optic nerve head astrocytes of monkeys with experimental glaucoma (17). Here we examined the effect of elevated hydrostatic pressure on ERK1/2 phosphorylation in cultured rat optic nerve astrocytes. We also examined phosphorylation of Na/H exchanger (NHE)1 activity because in some cells it is known that ERK1/2 activation can lead to changes in Na/H exchange activity (22). NHE1 is one of the principal ion transport mechanisms in astrocytes. Additionally, NHE1 functions as a scaffold that coordinates interaction between signaling molecules within some cells (4, 39, 41, 46, 47). NHE activation is associated with a variety of cell responses including changes in cell migration, adhesion, proliferation, and apoptosis (24).
MATERIALS AND METHODS
DMEM-F-12 and 2′,7′-bis(2-carboxyethyl)-5(6)-carboxyfluorescein-acetoxymethyl ester (BCECF-AM) were purchased from Invitrogen (Carlsbad, CA). Eagle's minimum essential medium (MEM) was purchased from Mediatech (Manassas, VA). All other chemicals were purchased from Sigma (St. Louis, MO). Polyclonal rabbit anti-GFAP was obtained from Dako Cytomation (Carpinteria, CA). Mouse anti-NHE1 monoclonal antibody was obtained from BD Transduction Laboratories (Lexington, KY). p44/42 MAPK rabbit monoclonal antibody, phospho-p44/42 MAPK (Thr202/Tyr204) mouse monoclonal antibody, phospho (Ser)-14-3-3 binding motif mouse monoclonal antibody, phospho-90-kDa ribosomal S6 protein kinase (p90RSK) (Thr359/Ser363) rabbit polyclonal antibody, and rabbit polyclonal phospho-Src-family (Tyr416) antibody were obtained from Cell Signaling Technology (Danvers, MA). Mouse anti-RSK monoclonal antibody was obtained from GenScript (Piscataway, NJ). The opossum kidney (OK) cell line, which was derived from Virginia opossum, is used widely as a model for mammalian renal proximal tubule (26). The cell line was obtained from Dr. Syed Jalal Khundmiri of the University of Louisville.
Astrocytes were isolated and cultured according to a modification of a previous method (16). Eyes from 1- to 5-day-old rat pups (Hilltop Lab Animals, Scottsdale, PA) were washed in DMEM-F-12 containing penicillin (100 U/ml) and streptomycin (0.1 mg/ml). Optic nerves were dissected, transferred to 35-mm culture dishes, and maintained at 37°C in a humidified atmosphere of 95% air-5% CO2 in DMEM-F-12 medium containing fetal bovine serum (10%), penicillin (100 U/ml), streptomycin (0.1 mg/ml), and epidermal growth factor (EGF, 5 ng/ml) (complete medium). The medium was changed on alternate days. After 7–8 days, when enough cells had grown out onto the petri dish, the remnant of the optic nerve was removed and the cells were stained for expression of GFAP to confirm astrocyte identity. Cells that did not stain positively for GFAP were discarded. GFAP-positive cells were then trypsinized and passaged. Confluence was attained in ∼2 wk. Third- to sixth-passage cells were used in this study. OK cells were cultured in conditions similar to those for astrocytes except that MEM was used without EGF.
Cells were grown to 70–80% confluence on 35-mm culture dishes for pH measurement studies and in 60-mm dishes for Western blot analysis. Pressure was applied by placing the culture dish at the bottom of a glass column filled with 500 ml of DMEM-F-12 complete medium at 37°C equilibrated with 95% air-5% CO2. The column was placed in a humidified atmosphere of 95% air-5% CO2 at 37°C, and the medium was bubbled with 95% air-5% CO2 supplied via a gas frit. The culture dish lid provided the cells some protection against turbulence caused by the gas bubbles. The height of the fluid column was 20 cm; thus pressure at the base of the column was 15 mmHg. Control cells were kept at ambient pressure. As a control for the change in volume of the culture medium, some control groups were transferred to a horizontal tray containing 500 ml of DMEM-F-12 but with a height of fluid of 1.25 cm, about the same as in a normal culture dish. The medium was bubbled with 95% air-5% CO2 supplied via a gas frit positioned at a similar distance from the culture dish as the gas frit in the vertical column and using the same flow rate. Test agents were dissolved either in dimethyl sulfoxide (0.06% final concentration) or in water according to solubility. Control cells received only the vehicle.
Measurement of cytoplasmic pH.
Cells cultured to semiconfluence on 35-mm plastic dishes were loaded with BCECF-AM (5 μM) for 15 min at 37°C in Krebs solution containing (in mM) 119 NaCl, 4.7 KCl, 1.2 KH2PO4, 25 NaHCO3, 2.5 CaCl2, 1.0 MgCl2, and 5.5 d-glucose, pH 7.4. The cells were then washed five times with the Krebs solution and incubated for another 10 min in the Krebs solution to allow deesterification of the dye. Deesterification transforms BCECF-AM (ester form) to membrane-impermeant BCECF (acid form), which is trapped in the cytoplasm. The cells were then washed again to remove any traces of external dye, and the dish was placed in a temperature-controlled perfusion microincubator (PDMI-2, Harvard Biosciences, Holliston, MA) on the stage of an inverted microscope (Nikon Eclipse 80i). The cells were superfused at 37°C with the Krebs solution continuously bubbled with 5% CO2-95% air. Solution inflow was adjusted to 3.0 ml/min with a gravity feed, and the outflow was matched with a peristaltic pump (Watson-Marlow, 520S, Falmouth, UK). The microscope was fitted with a high-resolution video camera (DVC 340M-00-CL) to continuously monitor and record the BCECF fluorescence intensity in the cells. Fluorescence intensity was measured at an emission wavelength of 535 nm and alternating excitation wavelengths of 488 nm and 460 nm programmed by an InCyt Im2 imaging system (Intracellular Imaging, Cincinnati OH). The fluorescence intensity ratio I488/I460 was calibrated by titrating BCECF free acid with a range of buffers with defined pH values (5.49–8.5). Calibration was done in two ways: in the cell perfusion dish and also in a calibration chamber (Intracellular Imaging). There was no difference between the calibrations. The camera and the microscope settings for the calibration and for conducting experiments were the same.
The experimental protocol was as follows. Cells were superfused with Krebs solution for 5 min to obtain a stable baseline cytoplasmic pH (pHi). The superfusate was then switched for 5 min to an ammonium chloride-containing buffer comprising (in mM) 99 NaCl, 4.7 KCl, 1.2 KH2PO4, 25 NaHCO3, 2.5 CaCl2, 1.0 MgCl2, 5.5 d-glucose, and 20 NH4Cl (pH 7.4 and continuously bubbled with 5% CO2-95% air). Ammonium chloride addition causes a rapid rise in pHi. Subsequent replacement of the ammonium chloride with Krebs solution causes a rapid fall in pHi followed by a gradual recovery toward baseline. The pHi recovery rate was measured as the slope of the straight line obtained by plotting pHi values against time (s). To focus on the initial linear phase of pHi recovery, the pHi recovery slope for just the initial 60% of recovery was analyzed. The slope was calculated from the regression equation of the straight line with Excel.
Real-time measurement of pHi under hydrostatic pressure.
A closed chamber (model RC-30, Warner Instruments, Hamden, CT) mounted on the stage of the epifluorescence microscope (Nikon Eclipse 80i) was used to measure pHi in cells subjected to an abrupt 15 mmHg pressure increase. The cells, grown on rectangular glass coverslips, were loaded with BCECF-AM and then washed as described above. The coverslip was mounted to the bottom window of the imaging chamber (RC-30, Warner Instruments). The chamber was then closed tightly, with a second glass coverslip forming the upper window. Chamber volume was 0.5 ml. A peristaltic pump passed Krebs solution through the chamber at 1 ml/min. Hydrostatic pressure was applied to the cells by raising the height of the exit end of the outflow tube. The pressure inside the chamber was measured continuously with a digital pressure transducer (Harvard Apparatus, Holliston, MA). pHi was determined as described above.
Cells were lysed in RIPA buffer containing (in mM) 50 HEPES, 150 NaCl, 1.0 EDTA, 10 sodium pyrophosphate, 2.0 sodium orthovanadate, 10 sodium fluoride, and 1 phenylmethylsulfonyl fluoride (PMSF) with 10% glycerol, 1.0% Triton X-100, 1.0% sodium deoxycholate, and Complete Mini Protease Inhibitor Cocktail tablets (Roche Diagnostics, Indianapolis, IN; 3 tabs/20 ml). The cell lysate was centrifuged at 14,000 g for 60 min, the supernatant was added to Laemmli buffer, and the proteins were separated by electrophoresis on a 7.5% (for NHE1) or 10% [for total (t) and pERK1/2, t- and phospho-p90RSK, and pSrc] SDS-polyacrylamide minigel. Proteins were then transferred by electrophoresis to nitrocellulose membrane, and the membrane was blocked for 1 h with Odyssey blocking buffer (LI-COR, Lincoln, NE). The membranes were incubated overnight at 4°C with the following primary antibodies: anti-NHE1 (1:1,000) mouse monoclonal, anti-p44/42 MAPK rabbit monoclonal (1:1,000), anti-phospho-p44/42 MAPK (Thr202/Tyr204) mouse monoclonal (1:1,000), anti-RSK mouse monoclonal (1:1,000), anti-phospho-p90RSK (Thr359/Ser363) rabbit polyclonal (1:1,000), and rabbit polyclonal phospho-Src family (Tyr416) antibody. All antibodies were diluted in the blocking buffer (LI-COR Odyssey). After three washes in 30 mM Tris, 150 mM NaCl, 0.5% (vol/vol) Tween 20 (TTBS) at pH 7.4, each membrane was incubated for 1 h with an appropriate secondary antibody conjugated with either IR Dye 680 goat anti-mouse secondary antibody (1:40,000) (Alexa Fluor, Invitrogen) or IR Dye 800 goat anti-rabbit secondary antibody (Rockland, Gilbertsville, PA). Protein bands were visualized by infrared LASER scan detection (LI-COR Odyssey). Protein concentration was measured with a bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL) based on a published method (48) using bovine serum albumin (BSA) as the standard.
NHE1 phosphorylation was examined by an immunoprecipitation approach described previously (49). Cells were washed twice with cold PBS (pH 7.4) and then lysed in ice-cold buffer containing (in mM) 20 Tris·HCl, 150 NaCl, 1 EDTA, 1 EGTA, 2.5 sodium pyrophosphate, 1 β-glycerophosphate, 1 Na3VO4, and 100 NaF, with 1% Triton X-100, 0.1% SDS, and protease inhibitors at pH 7.5. The mixture was centrifuged at 14,000 g for 60 min at 4°C, and the supernatant (250 μg protein) was incubated overnight at 4°C with mouse monoclonal antibody against phospho-Ser 14-3-3 protein binding motif (1:20). The mixture, which contained NHE1 immunocomplexes, was added to protein G (Sigma), incubated for 2 h at 4°C, and then washed by centrifugation (14,000 g, 1 min) with the lysis buffer without SDS. The NHE1 immunocomplex was dissociated by adding Laemmli buffer and heating to 70°C for 5 min. The immunoprecipitated sample was then subjected to Western blot analysis for NHE1 as described above. It was not possible to run a loading control because there is no reference protein that we can say with assurance remains unchanged. An equal amount of protein from control and treated cells was used, and the assumption was made that immunoprecipitation and Western blot efficiency was similar in the samples obtained from pressure-treated and control cells.
One-way analysis of variance was used to compare multiple groups of data, and Student's t-test was used to compare two groups of data. A P value of <0.05 was considered significant. Results are shown as means ± SE.
Studies were conducted to examine ERK1/2 in cells subjected to 15 mmHg hydrostatic pressure for 2 h. As judged by Western blot band density ERK1/2 phosphorylation increased significantly (Fig. 1), while total ERK1/2 was unaltered. Some cells were exposed to U0126 (10 μM), an MEK inhibitor, added before challenge with hydrostatic pressure. U0126 abolished the increase in band density of phosphorylated ERK1/2 in cells subjected to 15 mmHg for 2 h (Fig. 1). Added alone, U0126 significantly reduced band density of phosphorylated ERK1/2 in cells kept under normal pressure.
It was reported previously that p90RSK phosphorylation and activation is ERK dependent (30, 60). For this reason, p90RSK phosphorylation status was examined in cells subjected to 15 mmHg. As judged by Western blot band density, p90RSK phosphorylation increased significantly (Fig. 2) while total p90RSK remained unchanged. Cells that were exposed to U0126 (10 μM), added before challenge with hydrostatic pressure, showed no increase in band density of phosphorylated p90RSK (Fig. 2). Added alone, U0126 significantly reduced band density of phosphorylated p90RSK in cells kept under normal pressure.
Both ERK and p90RSK have been shown to phosphorylate and activate NHE1. Previously this laboratory has shown (32) by Western blot analysis that cultured rat optic nerve head astrocytes express NHE1 protein (∼110 kDa), while NHE2, NHE3, and NHE4 are undetectable. As judged by Western blot band density, the amount of NHE1 protein was similar in control cells and cells subjected to 15 mmHg for 2 h (data not shown). However, there was evidence of an increase in NHE1 phosphorylation. To examine NHE1 phosphorylation, cells were exposed to 15 mmHg for 2 h and then lysed and the lysate was subjected to immunoprecipitation using an antibody against phosphoserine. The precipitate was then subjected to Western blot analysis with an antibody directed against NHE1. Material isolated from cells exposed to 15 mmHg stress showed an increase in NHE1 band intensity consistent with increased serine phosphorylation of NHE1 (Fig. 3). When the cells were subjected to 15 mmHg in the presence of 10 μM U0126, the increase in band intensity was almost completely inhibited (Fig. 3A, lane d). It was difficult to run a loading control for the immunoprecipitation study because there was no reference protein that we could say with assurance remained unchanged. While serine-phosphorylated proteins were captured selectively, it is likely that there were different degrees of phosphorylation of various proteins in the pressure-treated cells. Thus we started with equal amounts of protein and assumed that the efficiency of the immunoprecipitation and the subsequent NHE1 Western blot was similar in the treated and control samples.
Figure 4A represents a typical pH recovery trace for a control (atmospheric pressure) experiment in astrocytes after an acid load. After 2-h incubation at elevated hydrostatic pressures, cultured astrocytes showed marked increase in the rate of pHi recovery from an acid load imposed by 5-min exposure to 20 mM ammonium chloride (Fig. 4B). After removal of ammonium chloride, 15 mmHg hydrostatic pressure-treated cells showed more than fourfold increase in pHi recovery rate: 5.03 ± 0.53 × 10−3 pH units/s in control cells and 23.92 ± 1.53 × 10−3 pH units/s in pressure-treated cells. Since the pressure challenge procedure involved incubation of the cells in a large volume of medium (500 ml) in a 20-cm-high vertical column, we conducted parallel control experiments with cells incubated in 500 ml of medium in a horizontal tray such that the submersion depth was not increased. In both cases the medium was bubbled with 5% CO2-95% air from a gas frit at a similar distance from the culture dish. When control cells treated in this manner were subjected to ammonium chloride acidification, they displayed a pHi recovery rate of 4.92 ± 0.39 × 10−3 pH units/s (n = 7), which is similar to the value obtained under normal control conditions [5.03 ± 0.53 × 10−3 pH units/s (n = 9)]. The findings suggest that gas bubbling and switching the cells to a large volume of medium do not lead to a change in pH recovery rate.
To test the effect of a higher pressure, astrocytes were incubated at 30 mmHg pressure for 2 h by submerging them under 40.8 cm of medium. Phosphorylation of ERK, p90RSK, and NHE1 increased in cells subjected to 30 mmHg hydrostatic pressure (Fig. 5), but the magnitude of the change was no different from the magnitude of increase elicited by 15 mmHg. The rate of pH recovery after ammonium chloride acidification in cells subjected to 30 mmHg was 19.07 ± 2.33 × 10−3 pH units/s (n = 6), which is significantly higher than the rate of 5.03 ± 0.53 × 10−3 pH units/s (n = 9) but similar to the rate of 23.92 ± 1.53 × 10−3 pH units/s (n = 10) observed in cells subjected to 15 mmHg.
In a comparative study, cultured OK cells were subjected to a similar challenge with 15 mmHg hydrostatic pressure for 2 h. Lysates from control and pressure-treated OK cells were then probed for ERK and Src phosphorylation. OK cells subjected to 15 mmHg showed a significant increase in ERK phosphorylation (Fig. 6) but not Src phosphorylation (Fig. 7). However, there was no detectable difference between the rate of pHi recovery measured in control OK cells (atmospheric pressure) and OK cells subjected to 15 mmHg hydrostatic pressure [3.08 ± 0.28 × 10−3 (n = 12) vs. 2.81 ± 0.42 × 10−3 (n = 16) pH units/s, respectively].
To examine the contribution of NHE1 to the pressure response, some cultured astrocytes were exposed to the NHE inhibitor DMA (100 μM) for the duration of the experiment. In the presence of DMA, subjecting the cells to 15 mmHg for 2 h failed to increase the pHi recovery rate. Added alone, DMA reduced the pHi recovery rate significantly in cells kept under normal pressure (Fig. 8 and Fig. 9).
Since there was evidence of ERK1/2 phosphorylation in cells subjected to 15 mmHg for 2 h, some cells were exposed to the ERK inhibitor U0126 (10 μM) for the duration of the experiment. U0126 almost abolished the increase in pHi recovery rate in cells subjected to 15 mmHg for 2 h. Added alone, U0126 did not significantly alter the pHi recovery rate in cells kept under normal pressure (Fig. 8 and Fig. 9).
Although hydrostatic pressure appeared to phosphorylate NHE1 and stimulate NHE-mediated pHi recovery, baseline pH was not significantly altered in cells subjected to 15 mmHg for 2 h either in the presence or in the absence of DMA (Table 1). To test whether a transient change of baseline pHi occurs, cultured astrocytes grown on a coverslip were loaded with BCECF and placed inside the pressure chamber mounted on the stage of a fluorescence microscope. Resting pHi was 7.23 ± 0.08 (n = 5) and did not change significantly when pressure was raised abruptly to 15 mmHg (Fig. 10).
In humans, IOP is commonly ∼15 mmHg but can increase by an additional 10–15 mmHg in persons affected with open angle glaucoma (15), the most prevalent form of hypertensive glaucoma. Previously, an increase of 15 mmHg (20 cmH2O) of hydrostatic pressure above the reference point of atmospheric pressure was shown to cause changes in protein expression detected after 1, 3, and 5 days in cultured optic nerve astrocytes (20). At higher pressure (60 mmHg), gene expression changes are observed in as little as 6 h (59). In the present study, increased ERK1/2 phosphorylation was observed in cultured rat optic nerve astrocytes subjected to a hydrostatic pressure of 15 or 30 mmHg for 2 h. The observed increase in ERK1/2 phosphorylation points to activation of the MAPK pathway, a signaling event known to precede activation of a wide array of downstream protein kinases including p90RSK, the 90-kDa ribosomal S6 protein kinase (18). Indeed, Western blot analysis of pressure-treated astrocytes revealed convincing evidence for p90RSK phosphorylation. Importantly, p90RSK phosphorylation was suppressed when cells were subjected to the hydrostatic pressure challenge in the presence of the MEK inhibitor U0126. The findings are consistent with p90RSK activation downstream of ERK1/2 activation.
ERK pathway activation has previously been found to play an important role in the chain of events leading to cultured mouse cortical astrocytes developing a reactive phenotype after exposure to endothelin-1 (13). In cultured astrocytes, ERK1/2 phosphorylation has been observed in response to oxygen-glucose deprivation (22) and mechanical stretch (37). Hydrostatic pressure was reported to activate the ERK pathway and influence proliferation in vascular smooth muscle cells, but pressure >100 mmHg was applied (52). In macrophages, on the other hand, a hydrostatic pressure of 20 mmHg reduces ERK1/2 phosphorylation (44). Thus it is apparent that the same stimulus can cause opposite effects in different cell types. The opposite effect could be explained by the involvement of a complex upstream intracellular cascade converging on MEK to affect ERK. Two such upstream entities shown responsive to pressure include Src (56, 57) and the multifunctional tyrosine kinase focal adhesion kinase (FAK) (9). Pressure has been shown to activate Src in colon cancer cells (51), cardiomyocytes (57), and smooth muscle (56). In contrast, Src-dependent ERK activation does not occur in pressure-treated macrophages, but instead FAK inhibits ERK (44).
Activation of the ERK pathway in pressure-treated astrocytes could lead to many different functional responses. Here we report an effect on the sodium/hydrogen exchanger NHE1. ERK-dependent activation of p90RSK in pressure-treated astrocytes could be responsible in part for the observed phosphorylation of NHE1. The finding that U0126 suppresses both p90RSK and NHE1 phosphorylation is consistent with this notion. Moreover, the ability of p90RSK to phosphorylate NHE1 is well established, and it has been termed the NHE1 kinase (30). In response to growth factors, p90RSK reportedly phosphorylates NHE1 at Ser703 in the regulatory COOH-terminal domain, and this activates the exchanger activity (50). In ischemic cardiomyocytes phosphorylation of Ser703 on NHE1 by p90RSK creates a binding motif for protein 14-3-3 and stabilizes NHE1 in an active state (28). Furthermore, inhibition of p90RSK prevents NHE-mediated ischemia-reperfusion injury (31). However, it also is possible that ERK contributed directly to NHE1 phosphorylation since the distal COOH terminus of NHE1 contains multiple serine and threonine residues that potentially can be phosphorylated by ERK1/2 as well as p90RSK (30). The COOH-terminal cytoplasmic regulatory domain of NHE1 may modulate transport activity by altering the affinity of H+ transport sites in the transmembrane domain (41).
Hydrostatic pressure had a significant effect on the ability of the cells to respond to an acid load: the rate of pHi recovery in pressure-treated astrocytes was increased fourfold. Importantly, the ability of the NHE inhibitor DMA to prevent the response provides strong evidence that NHE activity is stimulated in cells subjected to 15 mmHg hydrostatic pressure. Although multiple NHE isoforms are known to exist, this laboratory has only detected NHE1 in the cultured astrocytes used for the present study (32). Thus the pH recovery findings are consistent with activation of NHE1-mediated proton export in pressure-treated astrocytes, and this functional change in NHE1 is consistent with the observed increase in NHE1 phosphorylation. Moreover, the involvement of the ERK pathway in NHE1 phosphorylation is reflected in the ability of U0126 to abolish both NHE1 phosphorylation and the increase in the rate of pHi recovery in cells subjected to pressure treatment for 2 h.
Although the findings point to NHE activation, hydrostatic pressure did not alter baseline pHi. This may be because astrocytes, like several other cells, likely turn to bicarbonate transporters to regulate modest pHi changes and only call on sodium/hydrogen exchange to reverse large pHi swings. Rat astrocytes have been shown to possess robust bicarbonate-mediated pH-regulating mechanisms (23, 45). In ocular nonpigmented epithelium, which expresses bicarbonate transporters as well as NHE1 and NHE4, recent studies showed that close to the pHi baseline, pHi regulation is bicarbonate dependent and sodium/hydrogen exchange contributes little cytoplasmic pH regulation (43). On this basis, subtle pHi baseline changes due to activation of NHE1 in astrocytes subjected to 15 or 30 mmHg hydrostatic pressure may have been masked because of bicarbonate transporters. Interestingly, in the present study cells treated with DMA or U0126 show some pH recovery. The driving force for this recovery perhaps comes from the bicarbonate transporter or other proton-exporting mechanisms present in this cell.
The rate of pHi recovery after an acid load is a complex parameter that potentially can be influenced by cell buffering capacity and proton-loading mechanisms. In the present study, the contribution of proton-loading mechanisms to the observed rate increase in pressure-treated cells appeared negligible because DMA, which selectively inhibits only NHE-mediated proton transport, completely inhibited the pressure-induced increase in pHi recovery. When we subtracted the rate of pHi recovery in control and pressure-treated cells in the presence of DMA, the difference was negligible. While cellular buffering capacity can vary with pHi (6, 7), the degree of acid loading in ammonium chloride-pretreated cells was similar in all experimental groups: control (6.01 ± 0.06), 15 mmHg pressure treated (6.06 ± 0.11), DMA control (6.07 ± 0.06), DMA + 15 mmHg pressure treated (6.11 ± 0.14), U0126 control (6.03 ± 0.13), and U0126 + 15 mmHg pressure treated (6.10 ± 0.08). This, together with the observed ability of DMA to nearly abolish the increase in pHi recovery rate in pressure-treated cells, suggests that the response is the result of NHE stimulation rather than altered buffering capacity.
It is interesting to note that the magnitude of increases in the rate of pH recovery and phosphorylation of ERK, p90RSK, and NHE1 was similar in cells subjected to 15 mmHg and 30 mmHg hydrostatic pressures. The pressure response appears to have all-or-none characteristics. Interestingly, the ERK inhibitor U0126 did not change basal level of NHE phosphorylation and pHi recovery rate in control cells even though it caused significant reduction in the basal level of ERK phosphorylation. Perhaps there is a threshold level of phosphorylated p90RSK that can phosphorylate NHE1 as well as a threshold level of phosphorylated NHE1 that can influence pHi recovery.
It is understood that 15 mmHg is the approximate normal IOP in rats. Thus the baseline atmospheric pressure condition in the cultured astrocytes used in this study is different from the in vivo reference point. The ability of optic nerve astrocytes to respond to an increase of hydrostatic pressure above atmospheric pressure has been shown on multiple occasions, but other cells including macrophages and chondrocytes also respond to challenge with relatively low (<50 mmHg) hydrostatic pressure (5, 44). Obviously, it is difficult to compare the effect of the hydrostatic pressure applied to cultured astrocytes on a plastic surface to the effect on the optic nerve head in ocular hypertension. In the intact eye the optic nerve head has a tendency to bulge outward because of increased IOP, and this mechanical deformation (12, 58) is likely to have an effect on the cells. In the cultured cells studied here, the hydrostatic pressure was applied equally to both sides of the culture plate and should produce little or no mechanical deformation in the plastic surface on which the cells are growing. Thus the primary pressure-sensing mechanism that triggers pressure responses remains unknown. Through in vitro experiments it is known that pressure can either accelerate or decelerate enzymatic reactions depending on the activation volume of the reactants and the solvents. For example, a pressure of 1,000 bar (1 bar = 750 mmHg) increases the activity of xanthine oxidase and catalase (35, 36) but decreases the activity of superoxide dismutase at pressures between 500 and 1,000 bar (36). Trypsin and chymotrypsin-mediated hydrolysis has also been shown to be accelerated by hydrostatic pressure of 100–700 MPa (1 Pa = 7.5 × 10−3 mmHg) (55). These pressures are much above the physiological range and the pressure used in the present study. However, enzyme reactions are known to behave differently when pressure is applied in a nanoenvironment (54) and perhaps could be different in nanocompartments within a cell. It appears that not all cells respond similarly to hydrostatic pressure. ERK in cultured vascular smooth muscle is reportedly unaffected by hydrostatic pressure <100 mmHg (52). In the present study cultured astrocytes and OK cells responded differently. When OK cells were subjected to the 15 mmHg pressure challenge that stimulated NHE1-mediated proton export in cultured astrocytes, the ability of OK cells to recover from an acid load was not detectably altered. However, pressure-treated OK cells showed significant ERK phosphorylation, albeit to a smaller extent than that shown by pressure-treated astrocytes. This pattern of responses fits with a previous report that indicated that simultaneous activation of both ERK and Src is necessary to activate NHE in OK cells (53). In the present study there was no evidence of Src activation in OK cells subjected to elevated hydrostatic pressure.
NHE1 stimulation is known to occur in cells subjected to a range of different stresses including shrinkage, hypoxia, mechanical deformation, and acidification (39). The cell response to stress may have the effect of stimulating metabolism, and this in turn might require additional proton export capacity. However, it is difficult to tease apart the role played by NHE1 in stressed cells. While there is evidence that NHE1 activation is influenced by the MAPK pathway, there also is evidence to suggest that the MAPK pathway itself is altered as a result of NHE1 activation (25, 39). To make the situation even more complicated, NHE1 has more than one function. While it is an ion transporter, exchanging protons for sodium ions moved across the plasma membrane in the opposite direction, it also plays a key role as an anchoring protein that influences cytoskeletal organization (10) and serves as a point of assembly for a number of signaling proteins.
In summary, a modest hydrostatic pressure increase of 15 mmHg above atmospheric pressure for 2 h caused phosphorylation of ERK1/2, p90RSK, and NHE1 and increased NHE1-mediated pH recovery in astrocytes subjected to an acid load. These responses may precede long-term changes of protein expression known to occur in pressure-stressed astrocytes (20).
This research was supported by National Eye Institute Grant EY-014069.
The authors thank Dr. Syed Jalal Khundmiri for donating OK cells and Dr. Ryan Pelis for useful discussions and suggestions during the experimental stage of this project. The cooperation of Prof. Patricia B. Hoyer, in the Department of Physiology, University of Arizona is gratefully acknowledged.
↵* A. Mandal and M. A. Terán contributed equally to this work.
- Copyright © 2009 the American Physiological Society