INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION APPENDIX REFERENCES

REGULATION OF CELL VOLUME and regulation of intracellular pH (pH_i) are important for normal cell functioning (5, 19, 21). In many cells, the two regulatory processes involve common membrane ion transporters, specifically the Na⁺/H⁺ exchanger and Cl/HCO₃ exchanger (7, 11). As a result, regulation of cell volume can entail changes in pH_i, whereas regulation of pH_i can entail changes in cell volume (7, 11). This interdependence raises interesting questions about the hierarchy of cell volume regulation vs. pH_i regulation, especially in cells (e.g., macrophages) that exhibit coordinated changes in pH_i and cell volume during cell activation.

The relationship between cell volume and pH_i is complicated further by the potential effects of osmotically induced changes in cell volume on the equilibrium state of intracellular buffers. Osmotically induced dilution or concentration of a closed-system cytosolic buffer (i.e., a buffer such as intracellular protein, which has a cytosolic content that is essentially fixed) should exert similar effects on the concentrations of the buffer, its conjugate species, and H⁺. This would produce little, if any, change in pH_i. However, the situation is different for open-system cytosolic buffers (i.e., buffers that readily permeate the plasma membrane and with a total cytosolic content that can vary rapidly). For example, consider cells that are shrunken by exposure to a hyperosmotic medium in the presence of physiological concentrations of CO₂-HCO₃, an open-system buffer. One expects that osmotic cell shrinkage (reflecting a net loss of cytosolic water) would produce comparable increases in the cytosolic concentrations of HCO₃ and H⁺ with little or no change in the cytosolic Pco₂ (and hence CO₂ concentration). In this instance, cell shrinkage would produce an intracellular CO₂-HCO₃-H⁺ disequilibrium, leading to the net formation of CO₂ from HCO₃ and H⁺ and an increment in pH_i (i.e., an intracellular contraction alkalosis). This change in pH_i is comparable to the effect on plasma pH of reducing extracellular fluid volume in vivo (10). Weak acids (e.g., CO₂) and bases are, by definition, buffers. Thus one might also expect osmotic changes in cell volume to disrupt pH_i in the presence of other membrane-permeant weak acids or bases (e.g., lactic acid or NH₃).

We have shown previously that osmotic shrinkage of resident alveolar macrophages activates the plasmalemmal Na⁺/H⁺ exchanger and, under CO₂-free conditions, causes an increase in the steady-state pH_i (15). In the present study, we determined the effects of osmotic cell shrinkage on the pH_i of single alveolar macrophages in the presence of CO₂. Macrophage exposure to hyperosmotic medium in the presence of CO₂ caused a biphasic rise in pH_i (i.e., an initial rapid increase, followed by a slower increment in pH_i), whereas osmotic shrinkage in the absence of CO₂ caused a slow monotonic rise in pH_i. We then conducted studies with other open-system buffers, specifically a membrane-permeant weak acid (propionic acid) or weak base (NH₃), to test the hypothesis that the initial rapid change in pH_i on osmotic shrinkage reflected a disequilibrium in the total cellular buffer system (intrinsic buffers plus added weak acid-base buffer).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION APPENDIX REFERENCES

Cell preparation. Resident alveolar macrophages were isolated from rabbit lungs, as described previously (1, 2). The cells were suspended in RPMI 1640 supplemented with 25 mM HCO₃, 25,000 U/ml penicillin, and 50 µg/ml gentamicin. The cells then were seeded on 18-mm round glass coverslips at densities of ~10⁶ macrophages per coverslip and maintained overnight in an incubator (37°C, 5% CO₂).

pH_i measurement. The methods for measurement of pH_i have been described previously in detail (6). Briefly, pH_i was measured using a Nikon inverted microscope coupled to a Spex cation measurement system. Cells were exposed to 10 µM 2',7'-bis(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF)-AM (in standard HEPES-buffered solution) at room temperature for 2-10 min. The cells then were continuously superfused with a solution preequilibrated at 37°C. An ~10-µm-diameter spot of light was focused on a single cell on the stage of the microscope. The fluorescence of intracellular BCECF was measured and corrected for background fluorescence unrelated to BCECF. Intracellular BCECF was calibrated to pH_i using the standard high K⁺-nigericin technique (6, 22).

Solutions. The standard HEPES-buffered solution contained (in mM) 135 NaCl, 5 KCl, 1 CaCl₂, 1 MgSO₄, 2 KH₂PO₄, 6 HEPES, and 5 glucose. HEPES-free CO₂-HCO₃ solution was prepared by adding 25 mM NaHCO₃ (replacing NaCl and HEPES) and bubbling with 5% CO₂-balance air. Solutions with propionic acid-propionate were made by replacing 20 mM NaCl with 20 mM sodium propionate. Solutions with NH₃-NH₄⁺ were made by replacing 20 mM NaCl with 20 mM NH₄Cl. Extracellular osmolarity was routinely altered by adding mannitol (Sigma); a few experiments used sucrose (Mallinckrodt) or NaCl (Sigma). For calibration with the high K⁺-nigericin technique, the standard solution was used with Na⁺ replaced by K⁺ and 10 µM nigericin added. All solutions had a pH of 7.4 at 37°C, except for the calibrating solutions, which were titrated to different pH values using N-methyl-D-glucamine or HCl.

Statistics. The data points are presented as means ± SE. Mean differences between populations were compared with the use of the Student's t-test. Nonlinear curve fitting was performed with the use of the Marquardt-Levenberg algorithm (NFIT, Island Products, Galveston, TX). Curve-fitted parameters are presented as means ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION APPENDIX REFERENCES

Effects of increased extracellular osmolarity in the absence of open-system buffers. Figure 1 shows the typical response of macrophage pH_i to osmotic cell shrinkage in the absence of open-system buffers (including CO₂-HCO₃). Exposure to hyperosmotic solution [twice-normal osmolarity (2T) = ~600 mosM] caused the pH_i to increase slowly. The consequent change in pH_i (pH_i) averaged 0.38 ± 0.04 in 18 cells (Table 1). The effect was reversible. When the solution was returned to the original osmolarity, pH_i slowly recovered toward the initial starting value (Fig. 1). In 18 cells, the mean pH_i after 2T removal was 0.34 ± 0.04 (Table 1).

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Representative effects on intracellular pH (pHi) of exposure to twice-normal osmolarity (2T) in CO2-free solution. The external solution was switched between an isotonic solution and a hyperosmotic solution containing 300 mM mannitol, to change the extracellular osmolarity from ~300 to 600 mosM.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   pHi dependence of responses to 2T in CO2-free solution for 18 cells. , Changes in pHi (pHi) during initial exposure to 2T, as a function of the starting pHi. , pHi observed after withdrawal of 2T (Off phase) are plotted with a positive sign.

Effects of increased extracellular osmolarity in the presence of CO₂. In the presence of CO₂, exposure to 2T caused a biphasic rise in pH_i. There was a rapid initial alkalinization from a baseline value of 7.00 ± 0.04 to a shoulder value of 7.10 ± 0.04, followed by a slower rise over the course of >5 min to a new steady-state value of 7.24 ± 0.05 (n = 20 cells) (Fig. 3, Table 1). In 20 cells, the mean pH_i was 0.10 ± 0.01 for the initial rapid portion of the On phase and 0.14 ± 0.04 for the subsequent slower portion. The Off response was monitored for only 1-2 min (Fig. 3). Macrophage pH_i declined rapidly during that period of time to an apparently stable value (pH_i  0.14; Table 1, Fig. 3).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Representative effects on pHi of 2T exposure in the presence of 5% CO2. A: at point a, the external solution was switched to a solution containing 300 mM mannitol to raise the external osmolarity to ~600 mosM. The break in the record between points c and d represents a period of 11 min during which 2T was continuously applied. At point e, the external solution was switched back to the original isotonic solution. B: On phase time course of pHi response is redisplayed (segment a-b of A), along with a single exponential curve fit to the data. Dashed line represents data from Fig. 1 (in absence of CO2) over the same time period of the On response. C: data for the Off phase (segment e-f of A) is redisplayed along with a single exponential curve fit to the data. Dashed line represents data from Fig. 1 (CO2-free) over the same time period of the Off response.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   pHi dependence of responses to 2T in CO2-HCO3 solution for 20 cells. , pHi produced by addition of 300 mM mannitol (On phase), as a function of the starting pHi. The best-fit line to the data (solid line) had slope of 0.21 ± 0.05 and y-intercept of 1.34 ± 0.21 (±SD). Slope was significantly different from zero (P = 0.0014) with R2 = 0.44. , pHi produced by return to normal osmolarity (Off phase), plotted with a positive sign. Best-fit line (dashed line) had slope of 0.25 ± 0.05 and y-intercept of 1.69 ± 0.35 (±SD). Slope was significantly different from zero (P < 0.0001) with R2 = 0.62.

Effects of increased extracellular osmolarity in the presence of propionic acid. If the rapid pH_i response to 2T in CO₂ reflected a disequilibrium in the total cellular buffer system, then other membrane-permeable weak acids would be expected to mimic the effects observed in CO₂. Furthermore, alveolar macrophages have HCO₃-dependent plasmalemma acid-base transporters (3) that could be operating during the osmotic challenge in CO₂. Propionic acid was selected as a membrane-permeant weak acid (4) that is unlikely to support HCO₃-dependent transport.

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Representative effects on pHi of varying osmolarity in the presence of propionic acid. Exposure to 20 mM total propionic acid-propionate caused a rapid fall in pHi, followed by a slower recovery of pHi to near the starting value. The cell was then pulsed with solutions of differing osmolarities: 30 mM mannitol was added for the 1.1T solution (10% increase in osmolarity above normal or 330 mosM); 75 mM mannitol for 1.25T (375 mosM); 150 mM mannitol for 1.5T (450 mosM); 300 mM mannitol for 2T (600 mosM); and 600 mM mannitol for 3T (900 mosM).

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Effect of magnitude of osmotic challenge on pHi in the presence of propionic acid-propionate. The mean pHi values produced during both the On and Off phases (Off plotted with positive sign) for 20 cells are shown as a function of the relative osmolarity of the pulse. The curve is a fit of the data to A + Blog10s, where s represents an ideal shrinkage factor (i.e., extracellular osmolarity relative to 300 mosM). Best-fit values (±SD) were 0.009 ± 0.005 for A and 0.35 ± 0.02 for B in the equation.

Effects of increased extracellular osmolarity in the presence of NH₃. If the rapid pH_i response in the presence of a weak acid was due to changes in the equilibrium characteristics of the total cellular buffer system, then a different response should be observed in the presence of a membrane-permeant weak base. Now the total cellular buffer system consists of a weak base-conjugate acid buffer pair (e.g., NH₃-NH₄⁺) plus the intrinsic buffers. Figure 7A illustrates an experiment in which a cell was exposed to 20 mM NH₃-NH₄⁺ in the nominal absence of CO₂. This solution caused a rapid alkalinization, consistent with the entry of NH₃, a fraction of which combined with intracellular H⁺ to form NH₄⁺. The pH_i subsequently recovered to near the original baseline value, reflecting a compensatory net H⁺ loading of the cell. The cell was then briefly exposed to NH₃ solutions of altered osmolarity. In marked contrast to the results with a weak acid buffer (CO₂ or propionic acid), increases in extracellular osmolarity caused cellular acidifications in the presence of NH₃. In the case with 2T, pH_i was 0.21 for the On response in NH₃ and 0.21 for the Off response (Table 1).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Representative effects on pHi of osmotic challenges in the presence of NH3-NH4+. A: initial exposure to 20 mM total NH3-NH4+ produced a rapid alkalinization, as NH3 entered the cell and combined with H+ to form NH4+. In the continued presence of ammonium, the pHi recovered back toward the starting value over a period of 10-15 min. The cell was then pulsed with differing amounts of added mannitol to alter extracellular osmolarity. B: similar responses to 2T in NH3 were produced when extracellular osmolarity was altered using 300 mM mannitol, 300 mM sucrose, or 150 mM NaCl.

View larger version (13K): [in this window] [in a new window]   Fig. 8.   pHi dependence of responses to 2T in the presence of NH3-NH4+ for 28 cells. , pHi produced by exposure to 2T (On phase) was plotted with a positive sign, as a function of starting pHi. , pHi produced by return to normal osmolarity (Off phase).

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Effect of magnitude of osmotic challenge on pHi in the presence of NH3-NH4+. The mean pHi values produced during both the On and Off phases (On plotted with positive sign) for 50 cells are shown as a function of the relative extracellular osmolarity. The curve is a fit of the data to A + Blog10s. The best-fit values (±SD) were 0.011 ± 0.003 for A and 0.74 ± 0.01 for B.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION APPENDIX REFERENCES

Increases in extracellular osmolarity were accompanied by changes in pH_i under all circumstances examined in this study. Other investigators have determined the effects of osmotically induced changes in cell volume on plasmalemma acid-base transport (7, 11, 13). However, the present results suggest that changes in pH_i also can result from alterations in the equilibrium characteristics of the total cellular buffer system (intrinsic buffers plus added weak acid-base buffers) during changes in extracellular osmolarity and cell volume.

In the presence of weak acid buffers (CO₂ or propionic acid), exposure to 2T caused a rapid, reversible alkalinization, the size of which was dependent on the starting pH_i (see Figs. 3 and 4). Varying the magnitude of the osmotic challenge produced corresponding changes in the pH_i (see Fig. 6). This "intracellular contraction alkalosis" is comparable to the plasma acid-base disturbance described in vivo, wherein contraction of the extracellular fluid volume produces an alkalosis associated with concentrating HCO₃ at a constant CO₂, with intrinsic (i.e., nonbicarbonate) buffers playing an important role in determining the size of the extracellular pH change (10). Conversely, increasing extracellular osmolarity in the presence of a weak base buffer (NH₃) caused the pH_i to decrease rapidly (see Fig. 7), that is, an "intracellular contraction acidosis". Changes in pH_i in the presence of NH₃ were graded with the size of the osmotic challenge (see Fig. 9) but did not correlate with the starting pH_i (see Fig. 8). In absolute terms, the pH_i produced by osmotic cell shrinkage in NH₃ was larger than that observed in CO₂ or propionic acid.

The APPENDIX presents a simple qualitative analysis of the expected pH_i elicited by cell shrinkage in the presence or absence of membrane-permeant weak acid-base buffers (i.e., open-system buffers). A number of assumptions and simplifications are included in the analysis: ideality of the change in cell volume; no cell volume recovery during the pH_i transients; high relative membrane permeabilities to H₂O, CO₂, propionic acid, and NH₃; and low relative membrane permeabilities to H⁺, HCO₃, propionate, and NH₄⁺. The analysis also ignores possible nonideality in changes in intracellular constituents during cell volume changes (e.g., changes in intracellular ionic strength), the relative rates of buffering reactions including catalyzed reactions (e.g., CO₂ hydration/dehydration), the dependence of intrinsic (nonbicarbonate) buffering power (_int) on pH_i, the changes in _int produced by changes in cell volume, and the rates of plasmalemma acid-base transport that could be altered during the pH_i transients or by changes in cell volume directly.

Some of these issues have been addressed previously with alveolar macrophages (15). In suspensions of alveolar macrophages under CO₂-free conditions, addition of 320 mM sucrose to the external solution produced an approximate halving of cell volume, and macrophage volume remained at the smaller level for (at least) 10 min. The shrinkage was associated with an approximate doubling of _int at a given pH_i. Na⁺/H⁺ exchange (amiloride-sensitive and Na⁺-dependent recovery from acid loads) was activated by osmotic cell shrinkage. This is consistent with the slow but large pH_i increase observed in the present study in CO₂-free solution without added weak acid-base (see Fig. 1). In the presence of CO₂, additional acid-base transporters (e.g., Cl/HCO₃ exchangers) could contribute to changes in pH_i after cell volume perturbations. For this reason, we chose propionic acid as a model weak acid; it rapidly permeates the macrophage plasma membrane (4) and is unlikely to support plasmalemma HCO₃-dependent transport. Osmotic cell shrinkage produced similar rapid pH_i transients in the presence of CO₂ or propionic acid. Thus it is unlikely that the responses were mediated via Cl/HCO₃ exchange. The rapid pH_i responses in propionic acid were insensitive to EIPA or bafilomycin A₁. Hence, it also is unlikely that the responses were mediated via Na⁺/H⁺ exchange or the H⁺ pump. Furthermore, osmotic challenges caused pH_i to change in opposite directions in the presence of a weak acid buffer (i.e., contraction alkalosis) vs. a weak base buffer (i.e., contraction acidosis). It is difficult to imagine how shrinkage-activated acid-base transport could produce such divergent pH_i responses. These findings suggest that the rapid pH_i responses were not dependent on the effects of osmotic cell shrinkage on membrane ion transporters. The simplest explanation is that the rapid pH_i responses reflected disequilibria in the total cellular buffer system due to osmotic changes in cell volume.

This interpretation of the data is consistent with all salient features of the pH_i responses. As shown in the APPENDIX, a shrinkage-induced disequilibrium in the total cellular buffer system is predicted to produce a cytosolic alkalinization in the presence of a weak acid buffer (e.g., CO₂ or propionic acid) and a cytosolic acidification in the presence of a weak base buffer (e.g., NH₃). Such paradoxical changes in pH_i were observed experimentally. The analysis in the APPENDIX also predicts that the magnitude of the pH_i response (pH_i) should be dependent on the size of the osmotic challenge (s) and _int, reaching a maximum value of ±log₁₀s when _int is zero (direction of change in pH_i determined by presence of weak acid vs. weak base). The observed pH_i values varied directly with s (see Figs. 6 and 9) and were 35-74% of the predicted maximum, in keeping with the presence of intrinsic buffers in alveolar macrophages (15).

If cell shrinkage induced a disequilibrium in the total cellular buffer system, then reequilibration should be determined by competing reactions between intrinsic buffers and the added weak acid-base buffer (i.e., CO₂-HCO₃, propionic acid-propionate, or NH₃-NH₄⁺). Although the kinetics of these competing reactions are not known, we can compare the relative contribution of each reaction to the final equilibrium state by examining the buffering powers of the individual buffers. Alveolar macrophages in isotonic solution have an _int of ~20 mM/pH at pH_i 7.1 (15). At the same pH_i in the presence of 5% CO₂, the intracellular HCO₃ concentration is calculated to be ~12 mM, yielding an intracellular bicarbonate buffering power (_bicarb) of ~28 mM/pH (assuming the apparent CO₂-HCO₃ equilibrium constant and CO₂ concentration are the same inside and outside the cell and that CO₂-HCO₃ behaves as an open-system buffer such that _bicarb is 2.303 times the concentration of HCO₃). In the presence of 20 mM propionic acid-propionate or NH₃-NH₄⁺, the intracellular propionate buffering power (_prop) is calculated to be ~23 mM/pH and the intracellular ammonium buffering power (_amm) to be ~92 mM/pH (using similar assumptions as above). The absolute pH_i elicited by osmotic cell shrinkage was larger in the presence of NH₃ than weak acids, whereas similar pH_i values were detected in experiments with CO₂ and propionic acid, in keeping with _amm > _{bicarb  prop}.

The approach of comparing buffering powers can also be used to explain the different sensitivities of pH_i to starting pH_i in experiments with weak acids vs. weak bases. _bicarb, _prop, _amm, and _int are all sensitive to pH. In experiments with CO₂ or propionic acid, the conjugate base concentration (and hence, _bicarb or _prop) increased with increments in pH_i, following the law of mass action. In contrast, the _int of alveolar macrophages decreased at higher pH_i values (15). Thus the relationship between the buffering powers of intrinsic buffers vs. the added weak acid buffer (hence, the magnitude of the pH_i responses to osmotic cell shrinkage) should be relatively sensitive to starting pH_i in the presence of CO₂ or propionic acid. On the other hand, in experiments with NH₃, the conjugate acid concentration (hence, _amm) decreased with increments in pH_i, along with the pH-induced decrements in _int. Thus the relationship between the buffering powers of intrinsic buffers vs. the added weak base buffer (hence, pH_i) should be relatively insensitive to starting pH_i in the presence of NH₃. In keeping with this interpretation of the data, pH_i was dependent on the starting pH_i value in weak acid experiments but was not correlated with the starting pH_i in NH₃ experiments.

The present findings are probably not unique to alveolar macrophages. We have previously observed rapid changes in pH_i during changes in extracellular osmolarity in the presence of CO₂-HCO₃ in renal mesangial cells (Boyarsky, unpublished results). Conditions exist in vivo in which cells are exposed to diverse osmotic microenvironments in the presence of multiple open-system buffers (e.g., CO₂ and NH₃), particularly in the kidney. The effects on pH_i of alterations in extracellular osmolarity in a multiple buffer system are unclear. Furthermore, although the present studies focussed on the responses to hyperosmotic cell shrinkage, the pH_i effects were reversed On return to the original tonicity (with consequent cell swelling to recover macrophage volume). This observation suggests that "expansion" acidoses and alkaloses might occur during hyposmotic cell swelling in the presence of weak acids and weak bases, respectively.

The experimental cells were highly sensitive to increases in extracellular osmolarity. Statistically significant changes in pH_i occurred with only 5-10% increases in extracellular osmolarity (see Figs. 6 and 9). This raises the possibility that the responses have physiological/pathophysiological relevance. A large body of evidence indicates that recovery of cell volume following osmotic cell shrinkage involves the activation of plasmalemma acid-base transporters and often is accompanied by transporter-mediated changes in pH_i with a time course of several minutes (8, 9, 12-15, 18, 20). The present results show that cell shrinkage under physiological conditions (5% CO₂) can produce rapid increases in pH_i with a time course of several seconds, most likely due to disequilibria in the total cellular buffering system. The duration of these pH_i shifts was not determined. Nonetheless, such rapid responses will define the initial changes in pH_i that occur after osmotic cell shrinkage under physiological conditions. It is worthwhile to consider, therefore, that rapid buffer-associated pH_i changes are involved in the allosterical activation of the acid-base transporters that play a role in the subsequent recovery of cell volume.

Regardless of the specific mechanisms involved (transporter-mediated or buffer-associated), any change in pH_i that accompanies changes in cell volume will effectively reset the relationship between pH_i and extracellular pH. As such, pH_i responses to cell volume changes will influence the functioning of physiological systems that track extracellular acid-base status. For example, in animal studies, Kasserra and co-workers (16, 17) found that intravenous infusion of a hyperosmotic solution caused a prolonged blood acidosis (dilution acidosis) and hypercarbia without eliciting a compensatory increase in ventilation. The hyperosmotic challenge also produced alkaline shifts in the tissue pH of skeletal muscle (i.e., a tissue contraction alkalosis). Assuming that a similar pH shift occurred in chemoreceptive cells (intracellular alkalosis), the authors suggested that aniosmotic-induced changes in pH_i caused the chemoreceptors to mistrack blood acid-base status, and this impaired respiratory compensation of the acid-base disturbance. The present data suggest that uncoupling of intracellular and extracellular pH can have a rapid onset, before the involvement of volume-activated plasmalemma acid-base transporters.