We present evidence that Na-K-ATPase in the rat proximal tubule is directly activated by ANG II much faster than previously observed. Specifically, we show that a 2-min exposure to 0.1 and 1 nM ANG II slowed the rate of intracellular sodium accumulation in response to an increase in extracellular sodium added in the presence of gramicidin D. From these data, we show that ANG II directly stimulates Na-K-ATPase activity at rate-limiting concentrations of intracellular sodium. Under these same conditions, exposing proximal tubules to ANG II altered the amount of 32P incorporated into multiple phosphopeptides generated from a tryptic digest of the α-subunit of Na-K-ATPase. Na-K-ATPase was isolated from whole cell lysates by means of a ouabain-affinity column and then separated into its individual subunits by SDS-PAGE. Na-K-ATPase bound to the column in its E2 conformation and was eluted by altering its conformation to E1 using Na+ATP. Na-K-ATPase isolated from cells treated with ANG II eluted more easily from the ouabain-affinity column than Na-K-ATPase isolated from control cells, suggesting that ANG II decreased the affinity of Na-K-ATPase for ouabain. Thus ANG II rapidly stimulated the activity of Na-K-ATPase in 2 min or less by a mechanism that could involve changes in phosphorylation and conformation of Na-K-ATPase. We suggest that the physiological role for rapid direct activation of Na-K-ATPase is greater control of intracellular sodium during sodium reabsorption.
- sodium reabsorption
- sodium entry
angiotensin ii (ANG II) acting via the AT1 receptor stimulates net sodium reabsorption in the proximal tubule of the kidney, which affects blood pressure, contributes to the development of hypertension (27, 37, 47), and is relevant to the treatment of heart failure (9). Sodium moves from the lumen of the tubule across a single layer of epithelial cells and into the peritubular capillaries via three major sodium transport mechanisms. In the apical plasma membrane, the Na/H exchanger moves sodium down its electrochemical gradient into the cell. In the basolateral membrane, the Na/HCO3 cotransporter and Na-K-ATPase export sodium against its gradient into the interstitium. The energy required to accomplish the net transport of sodium across this epithelial barrier is derived from the hydrolysis of ATP by the Na-K-ATPase. This enzyme pump actively transports sodium, which is its rate-limiting substrate, and establishes the electrochemical gradients that drive the movement of sodium via the Na/H exchanger. The exchanger, in turn, removes intracellular protons promoting the formation of bicarbonate, which contributes to the size of the electrochemical gradient that drives sodium out of the cell via the Na-HCO3 cotransporter.
ANG II has been shown to rapidly (≤1 min) stimulate both the Na/H antiporter and the Na/HCO3 cotransporter in rat proximal tubule (22), whereas the minimum reported time for ANG II to directly stimulate Na-K-ATPase activity is 30 to 45 min (2, 7, 21). This apparent delay in activating Na-K-ATPase has significantly contributed to the impression that ANG II primarily regulates sodium reabsorption by controlling the activity of the Na/H exchanger and the Na-HCO3 cotransporter and that the stimulation of sodium reabsorption will be associated with a sustained increase in intracellular sodium concentration ([Na]i) that will secondarily stimulate Na-K-ATPase. This model is supported by evidence that ANG II initially increases [Na]i in proximal tubules (41, 52) and evidence from vascular smooth muscle that stimulation of Na-K-ATPase by ANG II is secondary to sodium entry (8, 43). On the other hand, there is considerable disagreement on the time course of the changes in [Na]i in response to ANG II in the proximal tubule (41, 52), and no one has yet tested how fast the Na-K-ATPase is directly activated by ANG II. Therefore, the extent to which Na-K-ATPase is directly regulated by ANG II during sodium reabsoprtion is unknown. Earlier studies examining the direct effect of ANG II on Na-K-ATPase activity in the proximal tubule were primarily concerned with demonstrating that direct stimulation was distinct from secondary activation and the resulting protocols were not suitable for testing how fast ANG II directly activated Na-K-ATPase activity on a minute time scale (2, 7, 21). We, therefore, tested the hypothesis that Na-K-ATPase activity in the proximal tubule could be directly stimulated with a much more rapid time course.
Using a new method to quickly measure Na-K-ATPase activity in a manner that can clearly separate direct stimulation from the secondary effects of increasing [Na]i, we here report that ANG II directly stimulated the activity of Na-K-ATPase at rate-limiting concentrations of [Na]i in 2 min or less. This time course is consistent with the rate at which the AT1 receptor activates cell signaling pathways (40) and stimulates the Na/H exchanger and the Na-HCO3 cotransporter (22). We also present evidence for a potential mechanism of Na-K-ATPase stimulation by ANG II: altered phosphorylation of the α-subunit of Na-K-ATPase and a decrease in the affinity of the Na-K-ATPase for ouabain, a specific inhibitor.
MATERIALS AND METHODS
Isolation of rat proximal tubules.
Rat renal proximal tubules from male Sprague-Dawley rats (200 to 250 g) were prepared by collagenase dispersion and isolated on Percoll gradients (51). The total yield from two kidneys from one rat was ∼16 mg protein.
Loading tubules with sodium-binding benzofluran isophthalate and the measurement of changes in fluorescence.
Isolated tubules were loaded for 45 min with sodium-binding benzofluran isophthalate (SBFI) at a final concentration of 18 μM in the presence of 0.16% Pluronic F-127, and 0.8% DMSO in 2.5 ml of buffer 1, which contained 125 mM NaCl, 4 mM NaHCO3, 5 mM KCl, 0.74 mM NaH2PO4, 5 mM glucose, 20 mM HEPES (free acid), 1.2 mM MgSO4, 0.56 mM Na2HPO4, 4 mM lactate, 1 mM sodium pyruvate, 1 mM CaCl2·2H2O, 1 mM each glutamine, l-alanine, and sodium butyrate, 0.1% BSA, pH 7.4 (11, 29). Buffer 1 had a total osmotic concentration of ∼310 mosmol/kgH2O by osmomometer (Precision). During loading, tubules were bubbled with 95% O2-5% CO2.
After being loaded, the suspension was divided into four equal aliquots, diluted ∼10-fold in buffer 1, and further bubbled until needed. Immediately before use, the tubules were washed three times with or without 3 mM ouabain in buffer 2, which was the same as buffer 1, except that it contained 75 mM choline chloride (no added NaCl) and CaCl2 was reduced to 0.5 mM. The total sodium in buffer 2 was 8 mM, and the final osmotic concentration was ∼237 mosmol/kgH2O. To wash, tubules were centrifuged at 4°C for 2 min at 36 relative centrifugal force (RCF). After the third wash, the pellet was brought up to a volume of 2 ml with buffer 2 and bubbled for 5 min at room temperature.
Suspended tubules were placed in a stirred cuvette in a spectrofluoremeter (SPEXFLUOROLOG II) at room temperature. The tubules had been in buffer 2 for ∼15 min, and the pH of the solution was between 7.2 and 7.1. The sample was alternately excited at 340 and 380 nm while the fluorescence was measured at 505 nm (26). After ∼1 min, ANG II or buffer was added. Two minutes later, gramicidin D (Gram D) dissolved in DMSO was added to a final concentration of 3 μM followed ∼5 s later by NaCl to give a final concentration of 50 mM and a total osmotic concentration of ∼310 mosmol/kgH2O. After 4 min, the concentration of Gram D was increased to 20 μM and ∼1 min later the NaCl concentration was increased to 125 mM. The final osmotic concentration was ∼450 mosmol/kgH2O. In each experiment, we used four aliquots to measure the rate of sodium accumulation ± ouabain ± ANG II.
Calculation of intracellular sodium.
Based on established procedures (26) and using Eq. 1, we calculated [Na]i from the fluorescence of SBFI (1) where R is the ratio of the fluorescence elicited at 340/380. R0 is the value of R when [Na]i is negligible, and R∞-calc is the calculated value of R∞ at saturating concentrations of sodium, which is appropriate for calculating [Na]i at ∼310 mosmol/kgH2O.
To compensate for slight differences in experimental conditions, the values for R0 and R∞-calc were determined for each individual aliquot. R0 was the value of R before the addition of ANG II (or buffer) when the tubules were suspended in buffer 2 containing 8 mM sodium (Fig. 1). The procedure for calculating R∞-calc depended on the presence and absence of ANG II and ouabain, because they could affect the activity of Na-K-ATPase and hence the value of [Na]i. In the presence of ouabain and the absence of ANG II, we assumed that [Na]i = extracellular sodium concentration ([Na]o) once a steady-state fluorescence was reached in the presence of 20 μM Gram D and 50 mM [Na]o (Fig. 1). Under these conditions, we assumed that PK = PNa and that the value of the membrane potential was zero. R∞-calc was calculated using Eq. 1 from the measured value of R assuming that [Na]i was equal to 50 mM. For tubules in the absence of ouabain ± ANG II and in the presence of ouabain + ANG II, we calculated R∞-calc by multiplying the measured value of R in the presence of 125 mM sodium by a correction factor. The correction factor was equal to the value of R∞-calc for the tubules incubated in the presence of ouabain and the absence of ANG II (as described above) divided by the measured value of R in the presence of 20 μM Gram D and 125 mM sodium. A typical correction factor was ∼1.06. This approach assumes that SBFI will be saturated with sodium when [Na]o is 125 mM and the Gram D concentration is 20 μM. The correction also accounts for the higher osmotic concentration (∼450 mosmol/kgH2O) in the presence of 125 mM sodium. The Kd at ∼310 mosmol/kgH2O was 23 mM as determined by us in proximal tubules (data not shown) following established procedures (26).
Measurement of intracellular free calcium in tubules.
Proximal tubular cells were incubated in buffer 1 with 2 μM fura 2-AM and 0.04% Pluronic F-127 for 45 min at room temperature while being bubbled with 95% O2-5% CO2 (23, 40). Tubules were washed three times in buffer 1 followed by reincubation in buffer 1 with further bubbling for 30 min to allow hydrolysis of internalized fura 2-AM. Approximately 15 min before calcium measurements were begun, the tubules were washed three times with buffer 2 and bubbled with 95% O2-5% CO2 until used. Samples were excited at 340 and 380 nm while the fluorescence was measured at 505 nm in a SPEXFLUOROLOG II at room temperature. Intracellular free calcium ([Ca2+]i) was calculated as previously described (23).
Isolation of the α-subunit of Na-K-ATPase from tubules metabolically labeled with 32P.
Rat proximal tubules isolated from two kidneys were suspended in buffer 3, which contained 50 mM NaCl, 75 mM choline chloride, 0.1 mM NaH2PO4, 5 mM glucose, 20 mM HEPES (free acid), 1.2 mM MgSO4, 4 mM lactate, 1 mM sodium pyruvate, 0.5 mM CaCl2·2H2O, 1 mM each glutamine, l-alanine, and sodium butyrate, and 0.1% BSA, pH 7.4. The tubules were then exposed to 2.5 mCi/ml [32P]orthophosphate for 60 min at room temperature while being bubbled with 95% O2-5% CO2. The tubules were then split into two groups and incubated for 2 min ± ANG II (1 or 0.1 nM). The tubules were collected by centrifugation (∼10 s) at 16,000 RCF and lysed in 1 ml ice-cold buffer containing 0.4 mg SDS/ml, 0.5 mM Chaps, 0.5% DMSO, 25 mM KCl, 5 mM imidazole, 1 mM MgCl2, 1 mM EGTA, phosphatase inhibitors (5 μM phenylarsineoxide, 1 μM microcystin, 1 μM okadaic acid), and protease inhibitors [1.04 mM 4-(2-aminoethyl) benzene sulfonyl fluoride, 15 μM pepstatin A, 14 μM E-64, 36 μM bestatin, 21 μM leupeptin, and 0.8 μM aprotinin] for 5 min on ice. The lysate was centrifuged (2 min) at ∼2,800 RCF to remove nuclear material, and the supernatant was diluted 1:1 with wash buffer containing 30 mM KCl, 5 mM imidazole, 1 mM EGTA, and the same phosphatase/protease inhibitors used in the lysis buffer. This dilution step reduces the concentration of SDS, which would otherwise interfere with binding of Na-K-ATPase to the ouabain-affinity column. The KCl concentration was maintained at 30 mM, because at this concentration the KCl is high enough to maintain binding of Na-K-ATPase to the column (53) and yet is low enough to avoid the formation of insoluble potassium dodecyl sulfate.
Each sample was then added to 1 ml of ouabain-affinity media in a separate small disposable column for isolation and purification of Na-K-ATPase (53). The columns were capped and mixed for 45 min at 4°C to bind Na-K-ATPase to the ouabain-affinity matrix. The columns were set upright, drained, and washed with 3 bed volume of wash buffer, which removed more than 95% of the original cellular proteins. Two bed volumes of wash buffer containing 0.02% SDS were added to the ouabain-affinity matrix and mixed for 20 min at 4°C. The column was then drained. This same procedure was performed a total of four times, because in each successive wash the ratio of SDS to protein is increased, which significantly improves the purification. The columns were next washed with 3 bed volume of a solution containing 156 mM KCl, 5 mM imidazole, 1 mM EGTA, and the same phosphates and protease inhibitors used in the lysing buffer and then with one-half bed volume of elution buffer (150 mM NaCl, 25 mM imidazole, 4 mM EDTA, 3 mM ATPNa2, plus phosphatase, and protease inhibitors). These samples were discarded. Flow through the column was halted and 1 bed volume of elution buffer was added and mixed with the ouabain-affinity media for 30 min at 4°C (53). The column was set upright, and the elution buffer was collected from the bottom of the column and saved. One additional bed volume of elution buffer was added to the top of the upright column, and an equal volume was collected from the bottom. These two fractions, both eluted with Na + ATP, were combined and further processed for the analysis of phosphopeptides. A solution of 2% SDS equal to 1.5 bed volume was added to the top of the column and collected at the bottom. This fraction was used to determine the amount of Na-K-ATPase that remained on the column after eluting with Na + ATP and was not used for the analysis of phosphoproteins.
Separation of phosphopeptides generated by tryptic digestion through two-dimensional thin-layer mapping.
Samples eluted with Na + ATP were concentrated using Centricon-100 units and subjected to SDS-PAGE (32), with 90% of the sample processed for Coomassie blue staining and subsequent processing, while 10% of the sample was transferred to nitrocellulose for Western blotting. The α-subunit was extracted from the stained gel in 50 mM NH4HCO3 (pH 7.5)/0.5 M β-mercaptoethanol/0.1% SDS and then recovered by TCA precipitation, using 20 μg of boiled RNase as carrier. The sample was washed with cold acetone, oxidized with performic acid, and maximally digested overnight with 20 μg of TPCK-treated trypsin (48). The digestion was terminated by repeated cycles in which the sample was diluted with water and subsequently concentrated. The peptides were dissolved in a buffer of 2.2% formic acid, 7.8% acetic acid, 90% water (pH ∼1.9), spotted onto glass-backed, thin-layer cellulose plates and then separated by electrophoresis at 1,500 V for 45 min in a model 7000 Hunter thin-layer electrophoresis apparatus. The plates were dried and the samples were separated overnight in the second dimension of ascending chromatography in 37.5% N-butanol, 25% pyridine, 7.5% acetic acid, and 30% water. Autoradiographs of the dried plates were prepared using Kodak MS film and intensifying screens at −80°, with typical exposure times of 1 wk.
The α-subunit of Na-K-ATPase was visualized by Coomassie blue staining and autoradiography and identified by immunoblotting using 0.5 μg/ml of a polyclonal antibody specific for the rat α-1 isoform of Na-K-ATPase (Upstate Biotechnology) with secondary antibody and enhanced chemiluminescent detection as described (35). The amount of α-subunit on immunoblots was quantified in arbitrary units by densitometry using a Fugi LAS-100 System and Image Gauge v3.3 software.
Statistics and curve fitting.
Statistical tests and curve fitting were done using Origin 6.0, Microcal Software. The Lorentzian equation used to fit the data on sodium accumulation is (2) where A is the total area under the curve from the baseline, Xc is the center, w is the full width of the peak at one-half the height, and Yo is the baseline offset (Origin 6.0, Microcal Software).
Effect of ANG II on Na-K-ATPase activity.
To test for a direct effect of ANG II on Na-K-ATPase at rate-limiting and physiological [Na]i, we measured the rate of sodium accumulation in rat proximal tubules in response to a sudden increase in [Na]o in the presence and absence of ouabain, a specific inhibitor of Na-K-ATPase (Fig. 1). In tubules equilibrated ±3 mM ouabain in a solution (buffer 2) containing 8 mM sodium, the initial level of [Na]i was close to zero and not affected by the addition of a small volume of either buffer (Fig. 1A) or ANG II (Fig. 1B). The addition of 42 mM NaCl and 3 μM Gram D, which increases the membrane permeability to monovalent cations (10), caused an immediate increase in [Na]i (Fig. 1, A and B). Over the next 200 s, [Na]i rose more slowly in the absence of ouabain than in its presence, showing that the Na-K-ATPase in these cells was active (Fig. 1). The rate of sodium accumulation in the absence of ouabain was slowed by the addition of ANG II (Fig. 1B vs. 1A), suggesting that ANG II stimulated Na-K-ATPase activity. Is this stimulation direct or indirect? If direct, then Na-K-ATPase activity at a given [Na]i should be higher in the presence of ANG II. Therefore, we need to determine Na-K-ATPase activity as a function of [Na]i ± ANG II. To carry out this evaluation, we have to further analyze the data in Fig. 1 beginning with an examination of the rates of unidirectional sodium influx and efflux in the presence of ouabain.
The rate of net sodium accumulation in the presence of ouabain was linear up to 35 mM [Na]i and was unaffected by ANG II (Fig. 2). This implies that the rate of unidirectional sodium influx was constant and unaffected by ANG II and that the rate of unidirectional sodium efflux, which increases as a function of [Na]i, remained small enough to be ignored. Thus, as long as [Na]i is ≤35 mM, the size of the ouabain-sensitive difference in [Na]i at any given time is directly proportional to the Na-K-ATPase activity that has occurred up to that moment. Once [Na]i is greater than 35 mM, the ouabain-sensitive difference in [Na]i underestimates Na-K-ATPase activity, because the ouabain-insensitive unidirectional sodium efflux carries a significant fraction of the sodium leaving the cell.
In the experiment shown in Fig. 1, A and B, it took 60 s for [Na]i to reach 35 mM in the presence of ouabain, as determined from the Lorentzian equation fit to the first 70 s of sodium accumulation (Fig. 3, A and B, respectively). The ouabain-sensitive difference in sodium concentration during this 60 s was then calculated from the original data and plotted as a function of the [Na]i in the tubules with a functioning Na-K-ATPase (Fig. 3C). At 60 s, the [Na]i in the tubules incubated in the presence of ANG II and the absence of ouabain was ∼20 mM (Fig. 3B). In the absence of ANG II, it took 44 s for [Na]i to reach 20 mM (Fig. 3A). Thus, to compare the effect of ANG II on Na-K-ATPase activity over the same range of [Na]i, Fig. 3C shows the ouabain-sensitive difference in [Na]i for the first 44 s of Fig. 3A and the first 60 s of Fig. 3B. Comparing the two curves in Fig. 3C, we conclude that ANG II directly stimulated Na-K-ATPase activity, because the ouabain-sensitive difference in sodium concentration at any given [Na]i is clearly greater in the presence of ANG II than in the control. The type of experiment shown in Fig. 3C was repeated five times at 0.1 nM ANG II and six times at 1 nM ANG II with essentially the same results (Table 1). Thus a 2-min exposure to either 0.1 or 1 nM ANG II rapidly activated Na-K-ATPase activity to a similar extent (Table 1).
To test whether measuring ouabain-sensitive sodium concentrations was a reasonable method for estimating Na-K-ATPase activity, we used our available data to estimate absolute Na-K-ATPase activity in more conventional units. Accordingly, Na-K-ATPase activity (nmol Na·mg total cellular protein−1·min−1) was calculated from the ouabain-sensitive change in [Na]i that occurred as [Na]i increased from 15 to 20 mM and the time it took for this change to occur. In the six experiments at 1 mM ANG II presented in Table 1, this interval was 16.3 ± 3.2 s (SE). In the controls, it was 11.6 ± 2.7 s (SE). Knowing the total amount of protein in the sample, we then estimated the intracellular volume using the conversion factor of 0.75 mg protein/mg dry weight (4) and the determination that the dry weight for proximal tubules is 24% of the wet weight (5). Here, we assumed the proximal tubules had a normal volume at ∼310 mosmol/kgH2O, despite not knowing the real volume of the cells during the period of measurement. With the use of this approach, the average pumping rate was ∼35 nmol Na·mg total cellular protein−1·min−1 when [Na]i increased from 15 to 20 mM in the control tubules and ∼70 nmol Na·mg total cellular protein−1·min−1 for the same interval in the presence of 1 nM ANG II. These estimates are well within the range measured by others using ouabain-sensitive 86Rb uptake (1, 3), a widely accepted approach to measuring Na-K-ATPase activity in intact cells. Thus the sizes of the ouabain-sensitive differences in sodium concentrations that we measured are of the correct order of magnitude.
Measurement of free [Ca2+]i.
Because suspending tubules in a hyposmotic solution is known to transiently increase [Ca2+]i and the effects of ANG II are calcium dependent, we measured the resting free [Ca2+]i under the conditions used to test the effect of ANG II on sodium uptake (Fig. 1). The resting free [Ca2+]i was 100 nM (mean of 2 measurements) when the tubules were suspended in buffer 2, which contained 0.5 mM CaCl2. This was the same resting free [Ca2+]i as when the tubules were suspended in a normal isotonic buffer (buffer 1) containing 1 mM CaCl2. Adding 1 nM ANG II to the tubules in buffer 2 triggered a transient increase in free [Ca2+]i, which peaked at 125 nM in 1 min and then declined to baseline during the next minute. For comparison, adding 1 nM ANG II to tubules suspended in an isotonic buffer 1 caused free [Ca2+]i to increase to 150 nM (mean of 2 measurements) in ∼2 min and to fall to baseline over the next several minutes. We also determined that the addition of 3 μM Gram D and 42 mM NaCl did not affect free [Ca2+]i during the 70-s interval used to measure the rate of sodium accumulation. Finally, the extracellular calcium in buffer 2 was set at 0.5 mM based on earlier measurements testing the relationship between extracellular calcium and resting free [Ca2+]i (data not shown).
Effect of ANG II on phosphorylation of Na-K-ATPase.
To investigate the hypothesis that direct and rapid stimulation of sodium pump activity could be due to changes in sodium pump phosphorylation, we radiolabeled oxygenated proximal tubules with [32P]orthophosphate and divided them into two sets (control and exposure to ANG II for 2 min) in buffer 3. The tubules were then separately lysed with detergent in the presence of protease and phosphatase inhibitors, and the sodium pump was purified in parallel on two separate ouabain-affinity columns (materials and methods). The Na-K-ATPase was bound to the column in the presence of 30 mM potassium (53), which puts Na-K-ATPase into a conformation [E2(K)] that binds to ouabain (24). After the column was washed, Na-K-ATPase was eluted with Na + ATP (53), which converts Na-K-ATPase into E1(Na + ATP), a conformation that does not bind ouabain (24). Proteins eluted with Na + ATP were separated by means of SDS-PAGE, stained with Coomassie blue, and the α-subunit was identified by immunoblotting (Fig. 4). The extent of the purification of the α-subunit can be evaluated by comparing the Coomassie blue-stained gel and the associated immunoblot of the whole cell lysate applied to the column (Fig. 4) with those of the purified fraction (Fig. 4). Autoradiographs of the Coomassie-stained gels clearly showed that the α-subunit from rat proximal tubules was a phosphoprotein (as expected), but there was not a consistent effect of 1 or 0.1 nM ANG II on the total incorporation of phosphate (data not shown). Therefore, we excised the α-subunits from the gel and subjected them to maximal cleavage with trypsin. The resulting peptide fragments were then separated in two dimensions (high-voltage electrophoresis and then ascending chromatography) to identify the degree to which individual phosphorylation sites were affected by ANG II treatment. Our data clearly show that ANG II increases the amount of phosphorylation in individual phosphopeptides (B, C, and F in Fig. 5) while having little or no effect on others (D and E) and perhaps decreasing the phosphorylation in another (A) (Fig. 5). Thus the rapid stimulatory effect of ANG II on Na-K-ATPase activity could be mediated by changes in the phosphorylation of the α-subunit. Finally, in these experiments, we observed more than four distinct phosphopeptides, which is more than the number of well-characterized sites of phosphorylation in the rat α-1 isoform.
Effect of ANG II on the affinity for ouabain.
In the process of doing experiments to determine the effect of ANG II on the phosphorylation of Na-K-ATPase, we observed that α-subunit purified from cells treated with ANG II eluted from the column more easily than from control cells (Fig. 6). In all these experiments, we used columns with a bed volume of ∼1 ml and applied ∼100 μg of total α-subunit to each column. The total amount of α-subunit that eluted from each column was equal to the amount eluted with Na + ATP plus the amount subsequently eluted with SDS. On each column, we recovered between 10 and 15 μg of α-subunit, which represents the capacity of the column (53). We did not observe any effect of ANG II on the total amount of Na-K-ATPase that eluted from the columns. This is what one would expect from loading the columns with 6 to 10 times as much Na-K-ATPase as they could bind. On the other hand, Na + ATP eluted ∼70% of the α-subunit that bound to the column from control cells and ∼88% of the α-subunit that was recovered from cells treated with ANG II, a significant difference (P < 0.05; Fig. 6). Because the reversal of ouabain binding is a slow process occurring over many minutes (30) and the binding and elution of rat kidney Na-K-ATPase to the column depend on conformation (53), these results suggest that treatment of the cells with ANG II may facilitate the release of the α-subunit from ouabain.
We showed that direct activation of Na-K-ATPase activity by ANG II occurs with a rapid time course very similar to the rate at which ANG II directly activates Na/H exchanger and the Na-HCO3 transporter (22). Our measurements cannot exclude that the Na/H exchanger and the Na-HCO3 transporter are still activated earlier than the Na-K-ATPase. Nevertheless, the expected physiological consequence of rapid activation of Na-K-ATPase is greater control of intracellular sodium during sodium reabsorption and less disruption of all the cellular functions that depend on the magnitude of the sodium gradient across the plasma membrane. These functions are diverse and significant and include effects on cell volume control, regulation of cell pH, the control of intracellular calcium, and the transport of phosphate, amino acids, and sugars.
To distinguish direct activation of Na-K-ATPase from secondary effects, we developed a unique set of experimental conditions, which include expected changes in cell volume and changes in the size of the sodium and potassium gradients across the plasma membrane. Thus, at the time ANG II was added, the cells were likely undergoing a regulatory volume decrease (RVD) (39). Similarly, adding NaCl to begin the measurement of Na-K-ATPase activity would have caused the cells to shrink and could have triggered a regulatory volume increase (RVI) (39). We do not know how the response of the Na-K-ATPase to ANG II might have been affected by RVD/RVI or by the presence of Gram D, which would have caused a loss of intracellular potassium. Thus it is important that the rapid effect of ANG II on Na-K-ATPase activity also be tested under more standard conditions.
Although we expected 0.1 nM ANG II to stimulate Na-K-ATPase activity, we were initially surprised to observe stimulation at 1 nM ANG II. Others showed that low concentrations of ANG II (10−12 to 10−10 M) stimulate, high concentrations (10−8 to 10−5 M) inhibit, and intermediate concentrations (10−9 M) have little effect on fluid and sodium reabsorption (28, 44, 49). On the other hand, it has recently been shown that 1 nM ANG II, which is the concentration found in the lumen (38), can either stimulate, inhibit, or have no effect on fluid reabsorption in the proximal tubule depending on the modulatory effects of PKC and calcium (13). For instance, 1 nM ANG II had no effect on fluid reabsorption when both [Ca2+]i and PKC were free to respond in a normal way to the addition of ANG II (13). However, if the normal calcium transient produced by ANG II was reduced, 1 nM ANG II stimulated fluid reabsorption (13). Thus we may have observed a stimulatory effect of 1 nM ANG II on Na-K-ATPase activity, because the transient increase in free [Ca2+]i when the tubules were suspended in buffer 2 containing 0.5 mM calcium was suppressed compared with more standard conditions.
Phosphorylation of Na-K-ATPase.
The observation that ANG II altered the phosphorylation of individual phosphopeptides without producing a consistent effect on the net phosphorylation of the α-subunit is likely a consequence of the large number of phosphopeptides generated from the α-subunit and the presence of compensating changes, i.e., the amount of phosphorylation increased in some phosphopeptides (B, C, and F in Fig. 5), decreased in at least one other (A), and did not change in others (D and E). Such a complex pattern of changes in multiple phosphopeptides is common in proteins that are regulated by phosphorylation by more than one mechanism and by more than one signaling pathway (50). Na-K-ATPase is regulated by changes in phosphorylation at multiple sites and by multiple mechanisms (45). To date, there are four well-identified phosphorylation sites on the rat α1-subunit. Three of these are in the NH2-terminal tail: Tyr-10 for an unidentified tyrosine kinase activated by insulin (15), Ser-23 and Ser-16 for PKC (6, 18, 19). 1 The other established site is Ser-943 in the COOH-terminal region, a site phosphorylated by PKA (6, 17, 20). In addition, there are very likely at least two other sites: a serine and a threonine in the COOH-terminal half that are phosphorylated by PKC that have been detected in the α-subunit from the shark (34). Phosphoamino acid analysis of Na-K-ATPase immunoprecipitated from rat proximal tubules (15) and rat soleus muscle (12) also supports the presence of phosphothreonine somewhere in the primary sequence. It is also clear that activation of the AT1 receptor is capable of producing complex activation of signaling pathways that could cause multiple changes in phosphorylation (16). The AT1 receptor, which mediates direct stimulation of Na-K-ATPase activity (7), is coupled to adenylate cyclase via Gi, phospholipase C via Gq (31), and tyrosine-signaling pathways (42). Similarly, in LLC-PK1 cells, a proximal tubule cell line, the AT1 receptor quickly activates phosphatases, including calcineurin, a serine-threonine phosphatase (33). Thus it is quite possible that the observed stimulation of Na-K-ATPase activity is related to changes in the pattern of phosphorylation of the α-subunit without distinct changes in net phosphorylation.
Although the data in Fig. 5 are consistent with the presence of at least six sites of phosphorylation, the number of phosphorylation sites represented by the phosphopeptides is not yet clear for a number of reasons. First, not all of the sites may have incorporated enough 32P to be observed. Second, cleavage by trypsin can sometimes produce two fragments with distinct mobilities from the same phosphorylation site (for instance, due to variable cleavage within dibasic motifs) (25). Third, it is not possible to prove a negative and demonstrate that another protein was not present in the excised piece of gel and did not contribute some of the observed phosphopeptides. This explanation, however, is extremely unlikely, because such a hypothetical protein would have had to copurify with the Na-K-ATPase, which binds and elutes from the ouabain-affinity column in a conformational-dependent manner (53), have subsequently run at the same molecular mass as the α-subunit, and also have been sensitive to ANG II.
Given the number of possible phosphorylation sites that might be involved and the difficulty of comparing our experimental conditions with those of others, it is not yet possible to predict exactly how ANG II-stimulated changes in phosphorylation may have rapidly stimulated Na-K-ATPase activity. An increase in the affinity for sodium would be consistent with the results and conclusions of Aperia et al. (2), who observed direct stimulation at rate-limiting concentrations of sodium after exposure of rat proximal tubules to ANG II and subsequent freezing and thawing to permeabilize the plasma membrane. Similarly, recruitment of pumps to the plasma membrane via exocytosis could explain how ANG II directly stimulated Na-K-ATPase activity in frozen-thawed proximal tubules near saturating concentrations of sodium (7, 21). Evidence that recruitment to the plasma membrane could be at least part of the mechanism by which ANG II directly stimulates Na-K-ATPase has recently been obtained from work in opossum kidney cells expressing the rat α1-subunit (14). In this system, a 10-min treatment with 1 pM ANG II caused an ∼35% increase in ouabain-sensitive 86Rb uptake measured over the subsequent 20 min and an ∼60% increase in the amount of Na-K-ATPase in the plasma membrane. The mechanism of stimulation is apparently related to increased phosphorylation of both Ser-16 and -23, because both these sites are phosphorylated by ANG II, as measured by anti-phosphoserine antibodies, and mutating either of these sites to an alanine blocked the stimulatory effect of ANG II on ouabain-sensitive 86Rb uptake (14).
Change in conformation.
Data in Fig. 6 suggest that treating cells with ANG II could have a lasting effect on the affinity of the Na-K-ATPase for ouabain. We already established that Na-K-ATPase binds to the ouabain-affinity column under conditions that promote the formation of the E2 conformation and elutes from the column under conditions that promote the formation of E1 (53), as expected for the interaction of Na-K-ATPase with ouabain (24). In the present study, Na-K-ATPase was eluted with Na + ATP over a 30-min period at 2°C. We kept the temperature cold to reduce possible changes in the phosphorylation of the α-subunit and chose 30 min because we knew that the rate of release of ouabain from Na-K-ATPase is a slow process (30), especially in the cold (46). Clearly, further experiments are required to determine the reason for the shift in elution observed in Fig. 6 and to determine whether changes in conformation are part of the mechanism by which ANG II rapidly stimulates Na-K-ATPase activity.
In conclusion, we showed that ANG II directly activates Na-K-ATPase at rate-limiting concentrations of [Na]i during the same time frame that it stimulates the Na/H exchanger and the Na-HCO3 cotransporter. Direct stimulation of Na-K-ATPase is correlated with multiple changes in the phosphorylation of the α-subunit, which could be part of the stimulatory mechanism.
This work was supported by National Institutes of Health Grants R01-DK-60752 and R01-CA-81150, the Harper Medical Staff Trust Fund, and the National Kidney Foundation of Michigan.
↵1 The mature form of the rat α1-isoform is missing the first five amino acids of the primary sequence. Hence, Ser-23 in the full-length sequence, as was used here, is the same as Ser-18 in the mature form. Similarly, Ser-943 is the PKA phosphorylation site in the complete sequence.
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.
- Copyright © 2004 the American Physiological Society