Physiol Rev 83: 1183–1221, 2003; 10.1152/physrev.00010.2003. Regulation and Modulation of pH in the Brain MITCHELL CHESLER Department of Physiology and Neuroscience, Department of Neurosurgery, New York University School of Medicine, New York, New York 1184 1184 1185 1185 1187 1188 1188 1188 1189 1189 1189 1189 1190 1190 1190 1193 1193 1193 1194 1194 1194 1195 1196 1196 1198 1199 1199 1199 1210 1210 1211 1212 Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 I. Introduction II. Buffering Power III. Regulation of Intracellular pH in Vertebrate Neurons A. Neurons from hippocampus B. Neurons from cerebral cortex C. Neurons from cerebellum D. Neurons from spinal cord E. Medullary neurons F. Retinal neurons and photoreceptors G. Sympathetic neurons H. Neoplastic neural cells I. Synaptosomes IV. Regulation of Intracellular pH in Glial Cells A. pH regulation in nonmammalian glial cells B. pH regulation in mammalian astrocytes C. pH regulation in oligodendroglia D. pH regulation in Schwann cells E. pH regulation in glial tumor cells F. pH regulation and volume control in glia V. Molecular Identification of Brain Acid Transporters A. Na⫹/H⫹ exchangers in brain B. HCO⫺ 3 transporters in brain VI. Modulation of Intracellular pH by Neuronal Activity A. Modulation of pHi in neurons B. Modulation of pHi in glial cells VII. Regulation and Modulation of Brain Extracellular pH A. The pH of interstitial fluid B. Changes in interstitial pH caused by neural activity VIII. Effects of Activity-Dependent pH Shifts on Neural Function A. Modulation of function by extracellular alkaline shifts B. Modulation of function by activity-dependent acid shifts IX. Conclusion Chesler, Mitchell. Regulation and Modulation of pH in the Brain. Physiol Rev 83: 1183–1221, 2003; 10.1152/ physrev.00010.2003.—The regulation of pH is a vital homeostatic function shared by all tissues. Mechanisms that govern H⫹ in the intracellular and extracellular fluid are especially important in the brain, because electrical activity can elicit rapid pH changes in both compartments. These acid-base transients may in turn influence neural activity by affecting a variety of ion channels. The mechanisms responsible for the regulation of intracellular pH in brain are similar to those of ⫺ ⫹ other tissues and are comprised principally of forms of Na⫹/H⫹ exchange, Na⫹-driven Cl⫺/HCO⫺ 3 exchange, Na -HCO3 ⫺ ⫺ cotransport, and passive Cl /HCO3 exchange. Differences in the expression or efficacy of these mechanisms have been noted among the functionally and morphologically diverse neurons and glial cells that have been studied. Molecular identification of transporter isoforms has revealed heterogeneity among brain regions and cell types. Neural activity gives rise to an assortment of extracellular and intracellular pH shifts that originate from a variety of mechanisms. Intracellular pH shifts in neurons and glia have been linked to Ca2⫹ transport, activation of acid extrusion systems, and the accumulation of metabolic products. Extracellular pH shifts can occur within milliseconds of neural activity, arise from an assortment of mechanisms, and are governed by the activity of extracellular carbonic anhydrase. The functional significance of these compartmental, activity-dependent pH shifts is discussed. www.prv.org 0031-9333/03 $15.00 Copyright © 2003 the American Physiological Society 1183 1184 MITCHELL CHESLER I. INTRODUCTION II. BUFFERING POWER The study of pH regulation requires the establishment of fundamental concepts of pH buffering and a Physiol Rev • VOL ⫽⫺ dA dpH (1) or ⫽ dB dpH (2) where A and B are the concentrations of strong acid or base, respectively. Accordingly, the buffering power (or buffering capacity) has units of concentration. If one could analyze cytosol in the absence of organellar or metabolic influences, the buffering capacity would be determined by the summed contributions of the separate buffer species. In physiological studies, it is convenient to divide the total intracellular buffering capacity into the bicarbonate-dependent component and the nonbicarbonate or “intrinsic” buffering capacity, given the symbols b and i, respectively. The total intracellular buffering capacity (T) is simply the sum of b and i. The contribution to buffering provided by the CO2HCO⫺ 3 equilibria (b) can be approximated as b ⫽ 2.3 ⫻ [HCO3⫺]i (3) ⫺ where [HCO⫺ 3 ]i is the intracellular HCO3 concentration. The derivation of this formula assumes that cytosolic CO2 is constant, due to its equilibration with extracellular CO2, which serves as a fixed, infinite source. Such a system is said to be “open” with respect to CO2. If one knows the system PCO2 and the intracellular pH (pHi), then [HCO⫺ 3 ]i and b can be calculated from rearrangement of the Henderson-Hasselbalch equation as b ⫽ 2.3[HCO3⫺] ⫽ 83 • OCTOBER 2003 • www.prv.org 2.3 ⫻ S ⫻ PCO2 ⫻ 10pH K 10P a (4) Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 The importance of pH regulation to the viability of cells is widely recognized. The production of intracellular acid is a normal consequence of respiration and, if unchecked, can lead to a marked fall in intracellular pH, compromising vital functions. In this broad respect, the cells of the central nervous system (CNS) do not differ from those found in other tissues. Indeed, many of the mechanisms responsible for long-term “housekeeping” of hydrogen ions are similar. The cellular diversity of the CNS is unique, however, and significant differences can be found in the complement of acid transport mechanisms expressed among subtypes of neurons and glial cells. This has been made apparent by physiological experiments on particular cells and by molecular studies that have revealed regional diversity in the expression of transporter subtypes. The study of pH in the CNS is also distinguished by the occurrence of rapid increases or decreases in H⫹ that arise from electrical activity. These changes take place in time frames from milliseconds to minutes, involve neurons as well as glia, and occur in both the intracellular and extracellular compartments. Given numerous enzymes and ion channels that can be influenced by pH, a possible modulatory role of these pH transients has often been considered, particularly in relation to membrane potential and excitability. In this respect, the mechanisms that generate and regulate these pH changes are of considerable neurobiological interest. The regulation of brain pH, the generation of activitydependent pH changes, and their functional consequences have been the subject of earlier reviews (76, 80, 94). A comprehensive treatment of pH and brain function has also appeared in an excellent volume edited by Kaila and Ransom (152). The present contribution aims to provide an accessible, updated summary of this field and is consequently more circumscribed. To maintain reasonable limits, certain subjects have not received full coverage. Historical aspects and methodological issues have been described elsewhere (76, 149). The effect of pH on numerous enzymes and channels is an extensive subject that is also beyond the scope of this summary. With the exception of particularly notable studies, the reader is directed to various reviews of this topic (22, 211, 288, 298, 307, 319). Important pioneering work on pH regulation in the nervous system of invertebrates has been covered by several earlier contributions (76, 94, 259, 303). This review emphasizes pH studies in the mammalian CNS and ventures into invertebrate studies only when they are particularly relevant to the topic at hand. consistent nomenclature. The definition and measurement of pH, the mechanisms of extracellular and intracellular buffering, and the chemical equilibria involved have been the subject of previous comprehensive reviews (55, 76, 259, 303). The purpose of this section is to highlight the basic concepts of buffering power relevant to physiological studies of pH, before addressing the means of intracellular and extracellular pH regulation in the nervous system. The buffering power of aqueous solutions can be defined in a number of ways (157, 259). The most extensively used definition comes from the early work of Koppel and Spiro (172) and Van Slyke (323), where buffering power () is conceptualized as the concentration of added strong acid (or base) divided by the resulting change in pH. Thus pH REGULATION IN THE BRAIN J ⫽ T ⫻ dpHi dt (5) where J may be expressed as mM H⫹ extruded per unit time. A common mistake in studies of pH regulation is to compare values of dpHi/dt instead of calculating J from Physiol Rev • VOL Equation 5. For example, if the rate of pHi recovery from an acid load were the same in the presence and absence ⫺ of CO2-HCO⫺ 3 , one could not assert that HCO3 -dependent acid extrusion played no role. On the contrary, since T is greater in the CO2-HCO⫺ 3 -buffered media, one should conclude that the acid efflux rate (J) had increased and that HCO⫺ 3 -dependent acid extrusion was indeed present. In such studies, comparisons of J are best performed at the same baseline pHi, since b rises, and i typically falls with elevation of pHi. III. REGULATION OF INTRACELLULAR pH IN VERTEBRATE NEURONS The plasma membrane transport systems responsible for neuronal pHi regulation were first investigated in large invertebrate cells. The modern period of this work was initiated by Roos, Boron, and Thomas, who have reviewed their classic studies (259, 303). The essential features of acid transport systems described in invertebrate neurons can also be found in vertebrate nerve cells. In the first investigation of a vertebrate neuron, the giant reticulospinal cells of the lamprey were found to eliminate acid loads by an amiloride-sensitive Na⫹/H⫹ exchanger and an Na⫹-, ⫺ HCO⫺ 3 -, and Cl -dependent acid extruder (75, 83). In subsequent studies of mammalian neurons, Na⫹/H⫹ exchange and Na⫹-dependent bicarbonate transport were frequently noted. In addition, some nerve cells were found to express a passive Cl⫺/HCO⫺ 3 exchanger, which served as an acid-loading mechanism during recovery from cytosolic alkalization. Significant differences in the function, expression, and pharmacological sensitivity of these transporters have been noted in the investigation of mammalian neurons. Details of these studies are provided below and are summarized in Tables 1 and 2. A. Neurons From Hippocampus Neurons derived from the hippocampus have been the subject of the most extensive studies of intracellular pH (pHi) regulation in the mammalian nervous system. Cultured, fetal, rat hippocampal neurons were investigated by Raley-Susman et al. (250). A key finding was the presence of an Na⫹/H⫹ exchanger insensitive to amiloride and its derivatives. This transporter was the principal mechanism for maintenance of resting pHi and for recovery from acid loads. Evidence for Na⫹/H⫹ exchange in these cells was based on the ion dependence of pHi regulation in HCO⫺ 3 -free media. Thus an acidification was noted upon removal of external Na⫹, and acid extrusion was blocked in the absence of Na⫹, but could be supported by Li⫹. In HCO⫺ 3 -containing solutions, pHi recovery from acid loads was slightly inhibited by the stilbene derivative DIDS and was Na⫹ dependent, suggesting that 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 where PCO2 is expressed in Torr and S is the solubility coefficient of CO2 in mM/torr. It should be noted that the value of b increases steeply as pHi rises, being 2.5-fold greater at pHi 7.4 versus 7.0. In one study of cultured mammalian neurons, intracellular HCO⫺ 3 , while present in significant concentration, was found to contribute little to T (7). The basis of this observation was not clear. Faced with a rapid acid-base challenge, the rate of CO2-HCO⫺ 3 buffering would depend on the presence of carbonic anhydrase (see sect. VIIB, 1–3). It is plausible that insufficient activity of this enzyme limited the contribution of b under these circumstances. The major components of i are associated with the buffering contribution of imidazole residues and phosphate (55). Determination of i for a cell is not always straightforward. To measure i, one might apply a known concentration of acid or base (in the absence of CO2 and HCO⫺ 3 ) and measure the resulting change in pHi. This procedure, however, would rarely give an accurate value, since the imposed pHi change would be countered by plasma membrane acid transport in addition to cytosolic chemical buffering. Acid or base challenges are therefore typically made under conditions in which these transporters are inactive (e.g., under pharmacological blockade or ion substitution) or are minimally active (e.g., using a series of small base challenges). Under these conditions, the ratio of the acid (or base) load to the ⌬pHi (Eq. 1 or 2) is then considered to be a measure of i. Although values of i obtained with such methods are useful in analyzing pH regulatory processes, they should be considered as no better than convenient operational measures, since they ignore the probable contributions of metabolism and organellar transport to the cytosolic H⫹ balance. i has been operationally determined in this way for a number of neurons and glial cells and tends to falls in a range from 5 to 30 mM (76, 159), but may be as high as 40 –50 mM in cells of neonatal brain (256). The value of i can vary with pHi and typically increases as pHi falls (for examples, see Refs. 26, 35, 159). Determination of the intracellular buffering capacity becomes critical when analyzing acid-base transport across the plasma membrane. The rate at which pHi is changed by a transporter has quantitative relevance only in view of the buffering capacity that opposes the pH change. The molar rate of acid efflux from a cell (J) would therefore be given by 1185 1186 TABLE MITCHELL CHESLER 1. Acid regulatory mechanisms identified in neurons of brain and spinal cord Hippocampal Rat cultured fetal Acutely isolated Rat neonatal Rat neonatal/mature Mouse neonatal Cerebral cortical (fetal) Rat cultured fetal Na⫹-Driven HCO⫺ 3 Cl⫺/HCO⫺ 3 Exchange ⫹* ⫹* ⫹ ⫹ ⫺ ⫹ ⫹* ⫹* ⫹ ⫹ H⫹ Pump 250, 251 20 ⫹ ⫹ ⫹ Reference Nos. ⫹ 273 24 251 358 ⫹ ⫹ ⫹ ⫺? ⫹ ⫹ ⫹ 236 8 245 ⫹ ⫹ ⫺ ⫹ ⫹ ⫹ 113 248 ⫹ ⫹ ⫹ 46 ⫹ ⫹ ⫺ ⫺ ⫹ ⫺ 256, 257 256 ⫹ ⫺? ⫹ 146, 216, 265, 268 ⫺ Rows designate evidence for (⫹) or against (⫺) the presence of Na⫹/H⫹ exchange, Na⫹-driven HCO⫺ 3 transport (both Cl -dependent and ⫹ -independent forms), Cl⫺/HCO⫺ exchange, and an ATP-driven H pump. VLM, vasolateral medulla; IO, inferior olive; HYP, hypoglossal nucleus; NTS, 3 nucleus tractus solitarius. *Insensitivity to amiloride or its derivatives. a form of Na⫹-linked HCO⫺ 3 influx also contributed to net acid extrusion. Whether this mechanism was Cl⫺ dependent was not addressed. Baseline pHi of the fetal hippocampal neurons was not affected by removal of Cl⫺ or by application of DIDS, indicating that the fetal cells did not utilize a passive Cl⫺/HCO⫺ exchanger as an acid-loading mechanism 3 (250). This feature was confirmed in a later study of acutely isolated, fetal hippocampal neurons (251). In contrast, acutely isolated adult hippocampal neurons exhibited a pronounced alkalization upon removal of external Cl⫺. In situ hybridization studies targeting the brain isoform of the Cl⫺/HCO⫺ 3 exchanger (see sect. VB3) were in agreement with these findings. Although the message could be identified in fetal brain (171, 251), there was a greater predominance in mature tissue (251). TABLE Later studies of the acid extrusion mechanisms in fetal hippocampal neurons were undertaken by Baxter and Church (20). An Na⫹/H⫹ exchanger insensitive to ethylisopropylamiloride (EIPA) was noted, consistent with earlier data (250). However, in contrast to the previous studies, removal of Cl⫺ produced a DIDS-sensitive, Na⫹-independent alkalization, suggesting the presence of a passive Cl⫺/HCO⫺ 3 exchanger in the fetal-derived neuronal culture. Transition from HCO⫺ 3 -free (i.e., HEPESbuffered) to HCO⫺ 3 -containing solution induced a robust alkalization that was Na⫹ dependent and was blocked by DIDS, or by depletion of internal Cl⫺. These observations suggested that fetal hippocampal neurons can extrude acid using both Na⫹/H⫹ exchange and an Na⫹-driven, Cl⫺/HCO⫺ 3 exchanger. The latter process appeared to play a greater role in pHi regulation at room temperature. Thus 2. Acid regulatory mechanisms identified in neurons outside brain and spinal cord Retina Toad photoreceptors Frog, salamander rods Skate horizontal cells Sympathetic neurons Rat cultured Neoplastic neurons Neuroblastoma Pheochromocytoma Na⫹/H⫹ Exchange Na⫹-Driven HCO⫺ 3 Cl⫺/HCO⫺ 3 Exchange ⫹ ⫹ ⫹ ⫹ ⫹ ⫹ ⫺ ⫹ ⫹? ⫹ ⫹ ⫹ H⫹ Pump Reference Nos. 160, 231 155, 264 135 306 ⫹ ⫹ 21, 212 100 100 ⫺ Rows designate evidence for (⫹) or against (⫺) the presence of Na⫹/H⫹ exchange, Na⫹-driven HCO⫺ 3 transport (both Cl -dependent and ⫹ -independent forms), Cl⫺/HCO⫺ exchange, and an ATP-driven H pump. 3 Physiol Rev • VOL 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 Mouse cultured fetal Cerebellar (fetal) Rat cultured Purkinje Rat cultured granule Spinal cord (embryonic) Rat cultured ventral Medullary (in slices) VLM, IO, HYP NTS Synaptosomes Na⫹/H⫹ Exchange pH REGULATION IN THE BRAIN Physiol Rev • VOL with low sensitivity to amiloride distinguishes hippocampal neurons from the great majority of cells. The nature of this amiloride resistance and its relationship to transporter isoforms remains unclear (see sect. VA). It is notable that acutely isolated CA1 neurons from mice displayed an Na⫹/H⫹ exchanger inhibited by the agent HOE 694 (358). However, the Na⫹/H⫹ exchanger of cultured fetal hippocampal neurons was insensitive to this drug (20). B. Neurons From Cerebral Cortex The regulation of pHi in cortical neurons has been studied in cultured fetal cells obtained from rat (236) and mouse (245). In cultured fetal rat neurons, recovery from intracellular acid loads was partially inhibited by amiloride (or 5-N-methyl-N-isobutylamiloride) and was dependent on external Na⫹, indicating the operation of an Na⫹/H⫹ exchanger (236) [inhibition of acid extrusion by amiloride was also noted by Amos and Richards (8) in cultures of rat neocortical neurons]. Ou-Yang et al. (236) suggested the absence of HCO⫺ 3 -dependent acid extrusion in fetal neurons, based on the following: 1) the similarity ⫺ of acid extrusion rates in HCO⫺ 3 -free and HCO3 -contain⫺ ing solutions, 2) the failure of DIDS or Cl removal to inhibit recovery from acid loading, and 3) similar acid extrusion rates in the presence of amiloride versus amiloride plus DIDS. In contrast, recovery from alkaline loads was inhibited by DIDS and by the absence of Cl⫺, suggesting the presence of a passive Cl⫺/HCO⫺ 3 exchanger. The putative absence of HCO⫺ -dependent acid extru3 sion in the cortical neurons studied by Ou-yang et al. (236) warrants comment. Although calculated acid efflux rates were similar in the presence and absence of HCO⫺ 3 , analysis was highly dependent on accurate determination of the i. It should also be noted that the method of acid loading in these studies was unorthodox, consisting of a transition from 5 to 15% CO2 with compensatory increases ⫺ in solution HCO⫺ 3 (and decrease in Cl ) to maintain a constant bath pH. Given the presence of a Cl⫺/HCO⫺ 3 ⫺ exchanger in these cells, the changes in HCO⫺ 3 and Cl could have confounded interpretation of the pHi recoveries. This study also reported that addition of DIDS to amiloride-containing solution did not further reduce acid extrusion. How this experiment was carried out is uncertain, as a precipitate commonly forms when DIDS is combined with amiloride or one of its derivatives (unpublished observation). It is worth noting that like hippocampal neurons (273), the steady-state pHi of the cultured cortical cells was higher in HCO⫺ 3 -containing media, which would be consistent with the presence of a HCO⫺ 3dependent acid extrusion mechanism. In a later study of cultured fetal neurons from mouse cortex, it was suggested that cells utilized both Na⫹/H⫹ exchange and an HCO⫺ 3 -dependent acid extruder (245). 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 transition from HEPES to HCO⫺ 3 buffer induced an alkalization, or speeded recovery from acid loading, only at 18 –22°C, and application of DIDS caused a fall in pHi at this low temperature, but not at 37°C. At the warmer, physiological temperature, acid extrusion appeared to be dominated by Na⫹/H⫹ exchange. Noncultured hippocampal neurons were investigated by Schwiening and Boron (273). These studies utilized a specific population of pyramidal neurons, acutely isolated from the hippocampal CA1 region of 4- to 14-day-old rats. In the absence of HCO⫺ 3 , the cells exhibited a low baseline pHi (a mean of 6.81) and sluggish recovery from acid loads. In agreement with previous studies of fetal, hippocampal neurons (250), this pHi recovery was inhibited by removal of Na⫹ (suggesting the presence of Na⫹/H⫹ exchange) and showed little sensitivity to amiloride. Unlike fetal hippocampal cultures, where Na⫹/H⫹ exchange was predominant (20), this mechanism played a minor role in acutely dissociated pyramidal neurons (in both the establishment of baseline pHi and the extrusion of acid at 37°C). With addition of external HCO⫺ 3 , a marked increase in baseline pHi occurred. This HCO⫺ 3 -dependent acid extrusion required external Na⫹ as well as internal Cl⫺ and was sensitive to DIDS, indicating the operation of an Na⫹-driven Cl⫺/HCO⫺ 3 exchanger. It was notable that the demonstration of Cl⫺ dependence could not be achieved by incubation in Cl⫺-free media for up to 3 h. To adequately deplete intracellular Cl⫺, repeated activation of the transporter in Cl⫺-free solution was required. These observations urge caution in the interpretation of experiments in which exposure to Cl⫺-free solution fails to alter pHi regulation. A study of pHi regulation in both postnatal immature and mature CA1 neurons from rat was undertaken by Bevensee et al. (24). The basic properties of pHi regulation were similar in the two age groups and featured an EIPA-insensitive Na⫹/H⫹ exchanger and a robust stimu⫹ lation of acid extrusion by HCO⫺ 3 , attributed to Na ⫺ ⫺ driven Cl /HCO3 exchange. A fraction of neurons exhibited a slow recovery from acid loads in the absence of external Na⫹, suggesting the presence of an H⫹ pump. Similar Na⫹-independent pHi recovery was later noted in CA1 neurons acutely isolated from mouse (358). The principal difference between immature and mature CA1 neurons was related to the distribution of baseline pHi. Acutely dissociated neurons derived from both age groups were found to have a wide range of initial pHi values (6.3–7.7) in HCO⫺ 3 -free, HEPES-buffered media. Mature neurons displayed a bimodal distribution, falling into low pHi (mean of 6.68) and high pHi (mean of 7.32) categories. For high pHi neurons, the relationship between pHi and the acid extrusion rate was shifted in the alkaline direction and had a greater slope, both in the absence and presence of HCO⫺ 3. The presence of a predominant Na⫹/H⫹ exchanger 1187 1188 MITCHELL CHESLER C. Neurons From Cerebellum Two studies of nerve cells cultured from rat cerebellum have been performed. Gaillard and Dupont (113) investigated Purkinje cells from neonatal (P0 –P1) rats after 6 –7 days in culture. In HCO⫺ 3 -free media, recovery from acid loading was abolished by withdrawal of external Na⫹ or by exposure to amiloride, indicating the operation of an Na⫹/H⫹ exchanger. In media containing HCO⫺ 3 , recovery from acid loading was also completely blocked by amiloride, which suggested the absence of HCO⫺ 3 -dependent acid extrusion. The presence of passive Cl⫺/HCO⫺ 3 exchange was indicated by the presence of a DIDS-sensitive, Na⫹-independent alkalization upon removal of external Cl⫺. The steady-state pHi of these cells was consistent with this complement of transporters. Unlike hippocampal (273) and cortical neurons (236), the steady-state pHi of the cultured Purkinje cells was lower ⫺ in HCO⫺ 3 -containing (7.06) versus HCO3 -free media (7.40). This finding is consistent with steady-state acid loading via Cl⫺/HCO⫺ 3 exchange and the absence of HCO⫺ -dependent acid extrusion. 3 Studies of rat cerebellar granule cells (cultured from 7-day-old pups) revealed a similar behavior of steadystate pHi (248). In HCO⫺ 3 -containing solution, pHi was lower (7.27) compared with HCO⫺ 3 -free media (7.49). Although completely dependent on external Na⫹, the recovery of pHi from acid loads was only partially blocked by amiloride. In roughly half the cells, this recovery exhibited a dependence on HCO⫺ 3 (but was not blocked by stilbenes). Recovery from acid loading persisted after prolonged exposure to Cl⫺-free media, suggesting the possible operation of Na⫹-HCO⫺ 3 cotransport in these cells. However, as noted by Schwiening and Boron (273), Physiol Rev • VOL internal Cl⫺ is not readily depleted from neurons and may require repetitive acid loading in Cl⫺-free solution to be sufficiently eliminated from the cytoplasm. The presence of a passive Cl⫺/HCO⫺ 3 exchanger in a fraction of the granule cells was suggested by the observation of a reversible alkalization upon withdrawal of external Cl⫺ (however, this effect was not blocked by stilbenes). The lower steady-state pHi noted in HCO⫺ 3 media was consistent with the presence of a Cl⫺/HCO⫺ 3 exchanger that contributed to steady-state acid loading. D. Neurons From Spinal Cord Recently, pHi regulation was studied in neurons cultured from ventral spinal cord of 14- to 15-day rat embryos (46). In HCO⫺ 3 -free media, acid extrusion was mediated by amiloride-sensitive Na⫹/H⫹ exchange. In the presence of HCO⫺ 3 , recovery from acid loading was also significantly slowed by 1 mM amiloride (but was not blocked by 40 M EIPA). Steady-state pHi was reduced by DIDS, suggesting that a HCO⫺ 3 -dependent acid extruder contributed to the maintenance of baseline pHi. Recovery from acid loads in HCO⫺ 3 -containing media was slowed by DIDS and abolished in Na⫹-free solutions. It was suggested that this recovery was mediated by an electrogenic Na⫹-HCO⫺ 3 cotransporter, since a reversible alkalization was observed when external K⫹ was raised to 30 mM. Whether this alkalization was dependent on Na⫹ or HCO⫺ 3 or could be evoked by other means of depolarization was not addressed, however. Thus experiments to more fully distinguish between Na⫹-driven Cl⫺/HCO⫺ 3 exchange and electrogenic Na⫹-HCO⫺ cotransport appear warranted. 3 Removal of external Cl⫺ was found to elicit a reversible alkaline shift in the spinal cord neurons, suggesting the operation of a passive Cl⫺/HCO⫺ 3 exchanger. Baseline pHi was lower in HCO⫺ -containing media compared with 3 HCO⫺ -free, HEPES-buffered solution, consistent with 3 steady-state acid loading via this mechanism. E. Medullary Neurons The neurons of the medulla oblongata are heterogeneous in form and function. Of particular interest in this region are the cells of the chemosensitive areas, which depolarize and fire in response to elevation of CO2 or H⫹. The degree to which this response is triggered by a fall in pHi, a fall in extracellular pH (pHo), or elevation of CO2 per se has been the subject of debate (218). Several recent studies have suggested that a fall in pHi is the principal trigger for the chemosensitive response (110a, 342, 348). For such cells, therefore, the control of pHi is a subject of great interest. The regulation of pHi in chemosensitive neurons has recently been reviewed (249). 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 After acid loading with the NH⫹ 4 prepulse method (38) in HCO⫺ -containing media, the pH 3 i recovery was markedly reduced by EIPA, indicating the operation of an Na⫹/H⫹ exchanger. The rates of pHi recovery in the presence and absence of HCO⫺ 3 (and during exposure to DIDS) were consistent with a HCO⫺ 3 -dependent component of acid efflux. However, the dpHi/dt values were not converted to acid extrusion rates (Eq. 5), and therefore the relative contribution of Na⫹/H⫹ exchange versus HCO⫺ 3 -dependent processes remained unclear. The authors suggested that recovery from alkaline loads occurred by a DIDSinsensitive Cl⫺/HCO⫺ 3 exchanger. In addition, a passive H⫹ influx through voltage-gated H⫹ channels was proposed, since exposure to 1 mM Zn2⫹ slowed the recovery from alkaline loads. Such channels, however, would not be open near the resting membrane potential, as they typically activate near 0 mV. Moreover, there is no compelling evidence for their presence in vertebrate neurons (92). pH REGULATION IN THE BRAIN F. Retinal Neurons and Photoreceptors Early evidence for acid extrusion via Na⫹/H⫹ exchange was provided by Oakley and colleagues (160, 231) in studies of rod photoreceptors from toad. A characterization of pHi regulatory mechanisms was later carried out in rods from frog (155) and salamander retina (264). Both amphibian rods utilized an amiloride-sensitive Na⫹/H⫹ exchanger. In addition, Na⫹-dependent acid extrusion was accelerated by HCO⫺ 3 and inhibited by DIDS, suggesting the presence of an Na⫹-driven HCO⫺ 3 transporter. Whether this mechanism was linked to Cl⫺ was not clear. Evidence for passive Cl⫺/HCO⫺ 3 exchange was found in both frog and salamander rods, as the withdrawal of Cl⫺ produced a rise in pHi, which persisted in the absence of external Na⫹ in the case of frog rods (155). A study of retinal horizontal neurons from the skate also demonstrated the presence of amiloride-sensitive Na⫹/H⫹ exchange, and apparent HCO⫺ 3 -dependent acid extrusion, based on the inhibition of acid extrusion by DIDS (135). Physiol Rev • VOL Removal of external Cl⫺ had no effect on pHi in the horizontal cells, suggesting that these neurons did not possess a passive Cl⫺/HCO⫺ 3 exchanger. G. Sympathetic Neurons One of the first studies of pHi regulation in mammalian nerve cells utilized cultured rat sympathetic neurons (306). Acid extrusion was dependent on external Na⫹ and was inhibited by amiloride, providing clear evidence for Na⫹/H⫹ exchange. The role of HCO⫺ 3 -dependent acid extrusion in these cultured neurons was less certain. While addition of stilbenes did not affect recovery from acid loads, in some cases, pHi recovery was stimulated by addition of 5–10 mM HCO⫺ 3. H. Neoplastic Neural Cells Neuroblastoma cells were among the first neural preparations in which Na⫹/H⫹ exchange was identified (21, 212). Acid extrusion via Na⫹/H⫹ exchange was subsequently reported in other studies of neural tumor cells (111, 204). Dickens et al. (100), studying neuroblastoma and pheochromocytoma (PC12) cells, noted the presence of Na⫹/H⫹ as well as Cl⫺/HCO⫺ 3 exchange. This paper also reported regional heterogeneity of pHi, as the tips of extending neurites were more alkaline (by 0.2– 0.3 pH units) than the cell body. I. Synaptosomes There is wide consensus that acid extrusion from brain synaptosomes is mediated by amiloride-sensitive Na⫹/H⫹ exchange (146, 216, 265, 268). It has been reported that there is no HCO⫺ 3 -dependent acid extrusion in synaptosomes (216, 265). This finding was based on similar rates of recovery from acidosis in the presence and absence of HCO⫺ 3 and lack of sensitivity to stilbene derivatives. However, an identical dpH/dt in the presence of HCO⫺ 3 would imply a greater rate of acid efflux, since intracellular buffering capacity should be elevated in the presence of CO2 and HCO⫺ 3 . Therefore, the possibility of HCO⫺ -dependent acid extrusion in these preparations 3 cannot be excluded. A rise in pHi was noted in synaptosomes exposed to elevated external K⫹ (265) [although in an earlier study (254), elevated external K⫹ did not produce a similar rise in pHi]. This apparent depolarizationinduced alkalization (DIA) persisted in the absence of ⫹ HCO⫺ 3 and Na , suggesting that it was not attributable to electrogenic Na⫹-HCO⫺ 3 cotransport. The cause of the DIA was not elucidated. Synaptosomal acid transport linked to a Cl⫺ flux was suggested by one study in which withdrawal of external 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 Putnam and colleagues (256 –258) imaged pHi from neurons of rat neonatal, medullary brain slices which had been incubated in BCECF. Neurons from two chemosensitive areas, nucleus of the solitary tract (NTS) and ventrolateral medulla (VLM), were compared with cells from the nonchemosensitive inferior olive (IO) and hypoglossal nucleus (HYP). In all cells, recovery from acid loading was completely abolished by amiloride but was insensitive to DIDS, suggesting that brain stem neurons utilize Na⫹/H⫹ exchange but not HCO⫺ 3 -dependent acid extrusion (257). In chemosensitive neurons, elevation of CO2 induced a fall in pHi that recovered only when bath pH was maintained at 7.48 (by simultaneous elevation of external HCO⫺ 3 to 52 mM). In contrast, neurons from nonchemosensitive areas recovered when CO2 was elevated, despite a fall in bath pH to 7.15 (at constant bath HCO⫺ 3 of 26 mM). This difference was attributed to an Na⫹/H⫹ exchange mechanism in the chemosensitive neurons that was more readily inhibited by external H⫹ (256). Immunocytochemical and pharmacological data from chemosensitive neurons in organotypic culture have implicated the type 3 Na⫹/H⫹ exchanger isoform in these cells (348) (i.e., NHE-3; see sect. VA). The presence of Cl⫺/HCO⫺ 3 exchange was suggested in VLM, HYP, and IO neurons, as a DIDS-sensitive alkalization was noted upon withdrawal of external Cl⫺. This exchanger could mediate recovery from alkaline loads in neurons from all three of these regions; however, the Cl⫺/HCO⫺ 3 exchanger in VLM cells was inhibited by high bath pH (7.9), unlike the nonchemosensitive cells. In contrast, there was no apparent Cl⫺/HCO⫺ 3 exchanger in the chemosensitive NTS neurons, which accordingly, could not recover from alkaline loading (256). 1189 1190 MITCHELL CHESLER Cl⫺ caused a DIDS-sensitive rise in pHi. This alkalization, however, was not dependent on HCO⫺ 3 , suggesting the presence of a Cl⫺-H⫹ cotransport mechanism (195). A similar putative Cl⫺-H⫹ cotransporter (or equivalently, a Cl⫺/OH⫺ exchanger) has been noted in cardiac myocytes (183). these cells (294). Evidence for Na⫹/H⫹ exchange (11) and electrogenic Na⫹-HCO⫺ 3 cotransport (12) was also reported for glial cells of the mudpuppy optic nerve. This work is well covered by the review of Deitmer and Rose (94). B. pH Regulation in Mammalian Astrocytes IV. REGULATION OF INTRACELLULAR pH IN GLIAL CELLS 1. Principal acid extrusion mechanisms in mammalian astrocytes A. pH Regulation in Nonmammalian Glial Cells The first detailed investigations of glial pHi were conducted on invertebrate cells which could withstand long-term penetration with pH-sensitive microelectrodes. Pioneering work on the giant glial cells of the leech demonstrated the presence of at least three acid extrusion mechanisms: an Na⫹/H⫹ exchanger, an Na⫹-driven Cl⫺/ ⫹ ⫺ HCO⫺ 3 exchanger, and an electrogenic Na -HCO3 cotransporter (95, 96, 98, 293). Later work indicated the additional presence of a passive Cl⫺/HCO⫺ 3 exchanger in TABLE 3. Acid regulatory mechanisms identified in glia Glia Amphibian glia Na⫹/H⫹ exchange Na⫹-HCO⫺ 3 cotransport Astrocytes ⫹ ⫹ Na /H exchange Na⫹-HCO⫺ 3 cotransport Na⫹-driven Cl⫺/HCO⫺ 3 exchange H⫹ pump Cl⫺/HCO⫺ 3 exchange Oligodendroglia Na⫹/H⫹ exchange Na⫹-HCO⫺ 3 cotransport Cl⫺/HCO⫺ 3 exchange Schwann cells Na⫹/H⫹ exchange Cl⫺/HCO⫺ 3 exchange Glioma cells Na⫹/H⫹ exchange Na⫹-driven Cl⫺/HCO⫺ 3 exchange H⫹ pump Reference Nos. 11 12, 221, 222, 224 26, 53, 87, 163, 166, 205, 247, 277 23, 26, 50, 53, 87, 128, 129, 167, 221–224, 232, 240, 277 203, 167, 277 239, 246, 353 53, 163, 203 41, 42, 162 41, 42, 162 41, 138 217, 267 217 111, 199, 277 277 246, 331 Evidence for particular transport mechanisms for a given cell type is listed. See text for differences in expression that have been noted within a cell category. Physiol Rev • VOL The presence of Na⫹/H⫹ exchange in mammalian astrocytes is well established and uniformly supported. Kimelberg and colleagues (163, 166) provided early evidence for Na⫹/H⫹ exchange in primary astrocyte cultures based on detection of Na⫹-dependent extracellular pH shifts. The use of amiloride or its derivatives to acidify baseline pHi or inhibit pHi recovery from acid loads served as the principal evidence for Na⫹/H⫹ exchange in later astrocyte studies (26, 53, 87, 205, 247, 277). The presence of an Na⫹/H⫹ exchanger in mammalian astrocytes has also been confirmed by immunocytochemical studies which revealed the presence of the amiloridesensitive NHE-1 isoform of this transporter (247) (see sect. VA). In some studies, the amiloride analog EIPA appeared ineffective against the astrocyte Na⫹/H⫹ exchanger (44) or caused a paradoxical increase in pHi (45, 247). Bevensee et al. (26) reported that while amiloride and EIPA could inhibit acid extrusion with similar efficacy at low pHi, EIPA appeared less effective near baseline pHi. Moreover, a fall in pHi was consistently produced by amiloride but not by EIPA. The basis of these observations remains unclear. Inhibition of Na⫹/H⫹ exchange by low external pH was described by Aronson et al. (10) in their studies of the renal transporter. A similar property was noted by Mellergard and Siesjo (205) who found that cultured astrocytes failed to extrude acid in bicarbonate-free media when the external pH was ⬍6.9. This feature has significant implications for the role of astrocyte Na⫹/H⫹ exchange in normal versus pathological settings. While culture studies show a significant contribution of this transporter when astrocyte pHi approaches 6.0 (26), such a low cytosolic pH would probably only arise in vivo under 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 The pH regulatory mechanisms of glial cells are comprised of transport systems very similar to those found in neurons. Physiological studies suggest that glia differ from neurons and from one another in their expression of Na⫹-linked bicarbonate transporters. Electrogenic Na⫹HCO⫺ 3 cotransporters appear to be especially prevalent among glial cells. Details elaborated below are summarized for glial cells according to cell type in Table 3. Studies of intracellular pH in mammalian glia have focused mainly on astrocytes. With a few exceptions, these studies have made use of pH-sensitive fluorescent dyes rather than pH microelectrodes. Almost all investigations have been performed on cultured astrocytes. Generally, the mechanisms of acid transport appear similar to those found in the invertebrate glia; however, some notable differences exist among mammalian astrocyte studies. 1191 pH REGULATION IN THE BRAIN exception of one study (203), its expression in mammalian astrocytes is agreed upon. Among glia, the presence of this transporter was first reported in invertebrates, where Na⫹-HCO⫺ 3 -dependent, electrogenic acid extrusion (13, 96), and a membrane potential-driven shifts in pHi were described (98). The first clue to the presence of this transporter in mammalian astrocytes came from the observation of a depolarization-induced alkalization (DIA) in rat cortical astrocytes studied in vivo. During stimulation of surrounding cortex, a rapid rise in astrocyte pHi was correlated with a depolarization of the astrocyte membrane, attributed to elevation of extracellular K⫹ concentration ([K⫹]o) (81). Addition of Ba2⫹ prevented both the activityevoked depolarization (17) and the cytosolic alkalization, despite a similar evoked rise in [K⫹]o (82). Data consistent with electrogenic Na⫹-HCO⫺ 3 cotransport was subsequently reported in cultured astrocytes (23, 50, 53, 232, 277), reactive astrocytes in gliotic brain slices (128, 129), and acutely isolated astrocytes from retina (221, 223, 224). This body of evidence is formed by two related observations. The first is the occurrence of a hyperpolarization (or an outward current) upon transition from bicarbonate-free to bicarbonate-containing media, which may be inhibited by DIDS or related stilbene derivatives (53, 221, 224, 232). A potential difficulty with this criterion is that the presence of a bicarbonate conductance could give rise to a similar shift in membrane potential (or membrane current under voltage clamp). In a novel, alternative approach, Bevensee et al. (23) withdrew external sodium to reverse the transporter and demonstrated a stilbene-inhibited depolarization. The second observation favoring electrogenic Na⫹HCO⫺ 3 cotransport has been the presence of a DIA in astrocytes. This rise in pHi is typically Na⫹ and HCO⫺ 3 dependent and is often blocked by stilbenes (50, 129, 222, 223, 240). In gliotic brain slices, however, an Na⫹-independent component of the DIA and an insensitivity to stilbenes were noted in reactive astrocytes (129). The dependence of the transport direction upon voltage is governed by the difference between membrane potential and the equilibrium potential for a Na⫹-HCO⫺ 3 cotransporter (ENBC). The general relation for ENBC is given by E NBC ⫽ 2. The astrocyte Na⫹-HCO3⫺ cotransporter The Na⫹-HCO⫺ 3 cotransporter of astrocytes has received considerable attention due to its electrogenic character and the resulting dependence on membrane potential. A form of this transport mechanism was first described in renal proximal tubule (37), and with the Physiol Rev • VOL 关Na⫹兴o关HCO3⫺兴on ⫺ RT ln F共n ⫺ 1兲 关Na⫹兴i关HCO3⫺兴ni (6) ⫹ where n is the net HCO⫺ 3 -Na transport ratio (37, 96). The preponderance of evidence suggests that the astrocyte mechanism has a stoichiometry of 2:1 (23, 50, 232) and that the net change in transport elicited by membrane depolarization is an influx of Na⫹ and HCO⫺ 3. 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 ischemic conditions, where it would be accompanied by a prominent fall in pHo, in the range of 6.2– 6.9 (174 –176, 215, 220, 332). Given the inhibition of this transporter by external H⫹, it seems that the astrocyte Na⫹/H⫹ exchanger would never operate within the pHi range that caused its maximum activation in tissue culture. In fact, under conditions which simulated the external ion shifts, acidosis and hypoxia of ischemic brain, the pHi of cultured astrocytes fell to 6.7– 6.8 and exhibited little or no recovery of pHi over a 20- to 30-min period (28). Virtually all studies agree on the presence of Na⫹and HCO⫺ 3 -linked acid transport in astrocytes. Differences exist regarding the role of Na⫹-driven Cl⫺/HCO⫺ 3 exchange and electrogenic Na⫹-HCO⫺ cotransport. Using 3 mouse cortical astrocytes, Chow et al. (87) reported that recovery from acid loads was not Cl⫺ dependent, and therefore suggested that a form of Na⫹-HCO⫺ 3 cotransport was involved. In contrast, Mellergard et al. (203) noted diminished acid extrusion after withdrawal of Cl⫺ and suggested that their cultured rat cortical astrocytes uti⫹ ⫺ lized Na⫹-driven Cl⫺/HCO⫺ 3 exchange, but not Na -HCO3 cotransport. Curiously, they reported a DIA, a phenomenon typically associated with an electrogenic Na⫹-HCO⫺ 3 cotransporter (see below). Shrode and Putnam (277) provided strong support for both Na⫹-driven Cl⫺/HCO⫺ 3 exchange and electrogenic Na⫹-HCO⫺ cotransport in their 3 cultured rat cortical astrocytes. A later study of cultured cerebellar astrocytes also found evidence for both exchangers (167). A detailed examination of pHi regulation in cultured rat hippocampal astrocytes was conducted by Bevensee et al. (23, 26). Using a fluorescent Cl⫺-sensitive dye to confirm the depletion of cytosolic Cl⫺, these authors were able to convincingly demonstrate that HCO⫺ 3linked acid extrusion did not depend upon Cl⫺ and, accordingly, was attributed to Na⫹-HCO⫺ 3 cotransport. Differences in the identification of acid extrusion mechanisms among astrocytes could be attributed to a number of factors. Discrepancies may arise due to use of low Cl⫺ media to test for Cl⫺-dependent transport. Unless depletion of internal Cl⫺ can be assured, failure to inhibit acid transport after Cl⫺ removal cannot be considered convincing (whereas inhibition of transport after Cl⫺ withdrawal clearly supports Cl⫺ dependence) (277). Differences in data may also originate from altered expression of transporters, arising from varied culture conditions, species of origin (e.g., mouse vs. rat) or brain region (e.g., cortex, vs. cerebellum vs. hippocampus). 1192 MITCHELL CHESLER 3. Na⫹-independent acid extrusion in astrocytes While the ability of astrocytes to extrude acid loads appears to be mainly due to a combination of Na⫹/H⫹ exchange and Na⫹-linked HCO⫺ 3 transport, reports of Physiol Rev • VOL Na⫹-independent transport have appeared. Recording with pH microelectrodes from cultured mouse astrocytes in HCO⫺ 3 -free media, Walz (353) noted a recovery from acute acid loads that was unaffected by either amiloride or withdrawal of external Na⫹. This result stands in marked contrast to a large body of evidence indicating that amiloride-sensitive Na⫹/H⫹ exchange is the principal acid extrusion mechanism of astrocytes in the absence of HCO⫺ 3 (see above). However, it may be noted that penetration of the mouse astrocytes with microelectrodes was made possible after rounding the cells by several hours of exposure to 1 mM dibutyryl cAMP (339). In response to this treatment, it is plausible that changes in transporter expression occurred. Downregulation of the Na⫹/H⫹ exchanger and the expression of a plasmalemmal H⫹ATPase could account for these results. Pappas and Ransom (239) were able to uncover a plasmalemmal vacuolar H⫹-ATPase in cultured rat hippocampal astrocytes that made a small contribution to acid extrusion. This transporter was identified in HCO⫺ 3free media by a small, bafilomycin-sensitive component of acid extrusion that remained in the absence of external Na⫹ or the presence of amiloride. In agreement with these findings, V-type ATPase mRNA has been isolated from cultured astrocytes (246). However, in view of the small component of acid extrusion attributed to this mechanism, its normal function is unclear. In the hippocampal astrocyte study, addition of bafilomycin caused a small intracellular acidification, suggesting a role in the maintenance of steady-state pHi (239). In cultured cortical astrocytes, however, the H⫹-ATPase inhibitors bafilomycin and concanamycin did not affect baseline pHi (5). 4. Cl⫺/HCO3⫺ exchange in astrocytes The presence of a Cl⫺/HCO⫺ 3 exchanger in astrocytes was suggested in early studies of astrocyte Cl⫺ fluxes by Kimelberg et al. (163). In most subsequent reports, the expression of this transporter was noted by the observation of a cytosolic alkalization upon removal of external Cl⫺. This alkalosis was reported in cerebellar (53) and cortical astrocytes (203), where it was inhibited by DIDS. In contrast, in hippocampal astrocytes, withdrawal of Cl⫺ did not alter pHi notably, and it was suggested that this transporter was not prominent in these cells (23). Shrode and Putnam (277) cautioned that alkalization upon withdrawal of external Cl⫺ could be mediated by reversal of a Na⫹-coupled Cl⫺/HCO⫺ 3 exchanger. To convincingly demonstrate the presence of passive Cl⫺/HCO⫺ 3 exchange in cortical astrocytes, these investigators showed that removal of external Cl⫺ produced a DIDS-sensitive intracellular alkalization that persisted in the absence of Na⫹. This process appeared to be active only at an elevated baseline pHi, suggesting a role in the recovery from alkaline loads. 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 For most astrocytes, values of ENBC would be on the order of ⫺80 mV (76), and given a membrane potential of similar magnitude, the transporter would normally be close to equilibrium. Some astrocytes have membrane potentials of ⫺50 mV or less, however (32, 287). For these cells, the energetics of transport may favor a large, steadystate influx of Na⫹ and HCO⫺ 3 (provided ENBC was maintained at a more negative value). However, little is known about how ENBC or transporter stoichiometry varies among astrocyte subtypes. A notable exception comes from the work of Newman (221, 222, 224) who investigated Na⫹-HCO⫺ 3 transport in acutely dissociated Muller cells of the salamander retina. These specialized glia are sometimes considered to be astrocyte variants due to their expression of glial fibrillary acidic protein (27). Muller cells were found to have a transporter with a reversal potential near 0 mV, consistent with a stoichiometry of one Na⫹ per three HCO⫺ 3 . As such, the energetics would favor the efflux of these ions at normal negative membrane potentials. Physiological studies found the transporter predominantly localized to the distal end foot process which normally abuts the vitreous humor. This led to the suggestion that the transporter is specialized to shuttle metabolically produced CO2/HCO⫺ 3 out of the retina (221). In later studies, evidence for Na⫹-HCO⫺ 3 cotransport was also found in Muller cells and astrocytes of rat retina (223). Numerous questions about localization, function, and role of astrocytic Na⫹-HCO⫺ 3 cotransporters remain unsettled. One issue is whether a DIA can always be completely attributed to this mechanism, since this alkalizing response, albeit diminished, can sometimes occur in HCO⫺ 3 -free media (42, 44, 50, 98, 204, 239). These observations might be explained by a mechanism with a high affinity for HCO⫺ 3 , capable of utilizing the small concentrations of this ion generated from metabolically derived CO2 (97). Whether this carrier actually utilizes HCO⫺ 3 or CO2⫺ in the transport mechanism has not been settled, 3 however (see sect. VIIB6). From the functional standpoint, Na⫹-HCO⫺ 3 cotransport appears capable of net acid extrusion in response to cytosolic acidification, as well as the modulation of transmembrane pH, in response to changes in membrane potential. The latter property may be a normal and unique feature of astrocytes, since neural activity rapidly depolarizes glial cells as a result of [K⫹]o elevation (132, 161, 234, 253, 284). The astrocyte Na⫹-HCO⫺ 3 cotransporter may therefore play a role in the modulation of both intracellular and interstitial pH (see sects. VIB and VIIB6). pH REGULATION IN THE BRAIN C. pH Regulation in Oligodendroglia D. pH Regulation in Schwann Cells Primary cultures of Schwann cells from rat sciatic nerve were studied by Nakhoul et al. (217). In addition to a classic, amiloride-sensitive Na⫹/H⫹ exchanger, these Physiol Rev • VOL cells exhibited a passive Cl⫺/HCO⫺ 3 exchanger, revealed by the presence of an Na⫹-independent alkalization upon removal of external Cl⫺. The Cl⫺/HCO⫺ 3 exchanger of the Schwann cells was unaffected by DIDS. These experiments did not address whether an Na⫹-linked HCO⫺ 3dependent acid extrusion mechanism was present. A role for pHi in the proliferation of Schwann cells was suggested in an early study in which mitogenic stimulation resulted in an apparent rise in pHi (267). This alkaline shift was associated with the Na⫹/H⫹ exchanger. Inhibition of Na⫹/H⫹ exchange after addition of a mitogen significantly reduced the degree of subsequent mitosis. E. pH Regulation in Glial Tumor Cells The regulation of pHi was studied in C6 glioma cells by Shrode and Putnam (277) in conjunction with their investigation of rat cortical astrocytes. The glioma cells shared two acid extrusion mechanisms with the astrocytes: a conventional, amiloride-sensitive Na⫹/H⫹ exchanger and a DIDS-sensitive, Na⫹-driven Cl⫺/HCO⫺ 3 exchanger. Unlike the astrocytes, the glioma cells did not have an electrogenic Na⫹-HCO⫺ 3 cotransporter, as they displayed no depolarization-induced alkalization upon elevation of external K⫹. A second study of C6 glioma cells (utilizing NMR spectroscopy) also presented evidence for Na⫹/H⫹ exchange and HCO⫺ 3 transport (111). A comparison of pHi regulation in glioma cell lines (C6 and human U-251, U-118, and U-87) versus primary cortical astrocyte cultures (199) reported a higher steady-state pHi in the gliomas (also noted by Shrode and Putnam for C6 cells) (277) and a higher rate of recovery from acid loads. These effects were attributed to enhanced activity of the Na⫹/H⫹ exchanger in the glioma cell lines, which was not attributed to genetic alterations. Disparity in the efficacy or expression of the HCO⫺ 3 -dependent acid transporters was also noted, as the astrocytes rapidly alkalinized upon introduction of CO2-HCO⫺ 3 , whereas the glioma cells underwent a marked acidification. A study by Volk et al. (331) provided evidence for a vacuolar H⫹-ATPase in C6 glioma cells, based on a fall in pHi observed after application of H⫹-ATPase inhibitors. A voltage-clamp study of bafilomycin-sensitive currents from C6 glioma cells provided additional data supporting the presence of an electrogenic plasmalemmal H⫹ATPase in these cells (246). In a comparison of vacuolar ATPase mRNA in C6 cells, primary astrocyte cultures, and immortalized astrocytes, no differences in levels were noted (246). However, an equivalence in H⫹-ATPase message may not imply similar activity of the transporter in glioma cells and astrocytes. Indeed, significant differences between these cells have been noted in this regard. For example, in the study of Volk et al. (331), the pronounced fall in the steady-state pHi during application of 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 A detailed study of pHi regulation in oligodendrocytes from mouse spinal cord was conducted by Kettenmann and Schlue (162). These cells displayed a high baseline pHi, ranging from 7.2–7.9 in HCO⫺ 3 -containing media of pH 7.4. In the absence of HCO⫺ 3 , pHi was lower, and recovery from acid loading occurred via a typical amiloride-sensitive Na⫹/H⫹ exchange mechanism. In HCO⫺ 3 -containing media, an additional acid extruder was identified that was Na⫹ dependent and apparently not linked to Cl⫺. The observation of a DIA (upon elevation of [K⫹]o) suggested that this mechanism could be an electrogenic Na⫹-HCO⫺ 3 cotransporter; however, this process was unaffected by stilbenes or furosemide. The presence of a Cl⫺/HCO⫺ 3 exchanger in these oligodendrocytes was uncertain. Removal of external Cl⫺ resulted in a higher baseline pHi, consistent with passive Cl⫺/HCO⫺ 3 exchange; however, the removal of HCO⫺ 3 had no complementary effect on Cl⫺ transport (138). Further studies of oligodendroglia were carried out by Gaillard and colleagues (41, 42) using cultured cerebellar cells. In mature oligodendrocytes, baseline pHi was 7.04 in HCO⫺ 3 -containing media (a value considerably lower than that of cultured spinal oligodendrocytes) and was unaffected by stilbenes or Cl⫺ removal, suggesting the absence of a Cl⫺/HCO⫺ 3 exchanger. In oligodendrocyte precursors, however, a DIDS-sensitive, passive Cl⫺/ HCO⫺ 3 exchanger was identified. Steady-state acid loading by Cl⫺/HCO⫺ 3 exchange accounted for a lower baseline pHi of the precursor cells, which averaged 6.88 (41). The differences in pHi noted during oligodendrocyte development may be related to the events leading to termination of cell division, as has been suggested for rat astrocytes (241). In HCO⫺ 3 -free media, both mature cerebellar oligodendrocytes and precursor cells extruded acid via an amiloride-sensitive Na⫹/H⫹ exchanger (41, 42). In the ⫹ ⫺ presence of HCO⫺ 3 , an electrogenic Na -HCO3 cotransporter was also identified. This transporter was unaffected by DIDS as was noted earlier for spinal cord oligodendrocytes (162). Analysis of the rate of alkalization and the calculated HCO⫺ 3 influx induced by graded elevations of external K⫹ suggested a reversal potential close to ⫺60 mV, consistent with an Na⫹-HCO⫺ 3 transport stoichiometry of 1:3 (42). Since the resting potential of these cells was also near ⫺60 mV, it was suggested that this transporter was normally close to equilibrium. 1193 1194 MITCHELL CHESLER H⫹-ATPase inhibitors suggested a particularly prominent role for this mechanism in C6 cells. In contrast, H⫹ATPase inhibitors had little or no effect on baseline pHi in cultured astrocytes (5, 239). F. pH Regulation and Volume Control in Glia V. MOLECULAR IDENTIFICATION OF BRAIN ACID TRANSPORTERS Molecular correlates of several acid transport mechanisms have been identified in the mammalian CNS. Functional characterization in expression systems has served to confirm the physiological identity of these transporters. Localization of particular mRNAs and immunostaining of specific proteins have provided insights into possible regional functions of these molecules. Details provided below are summarized in Table 4. A. Naⴙ/Hⴙ Exchangers in Brain The gene family of Na⫹/H⫹ exchangers identified in vertebrate tissues consists of at least seven members, TABLE 4. Localization of acid transport isoforms in brain Exchangers/Cotransporters Na⫹/H⫹ exchangers NHE1 NHE2 NHE3 NHE4 NHE5 Cl⫺/HCO⫺ 3 exchangers AE3 Na⫹-HCO⫺ 3 cotransporters “NBC” Na⫹-driven Cl⫺-HCO⫺ 3 exchange NDCBE1? Localization Reference Nos. High in cortex, hippocampus, cerebellum Cerebral cortex, brain stem Cerebellum (Purkinje cells) Hippocampus, cortex, brain stem-diencephalon High in hippocampal dentate gyrus, low in CA1 fields and cerebral cortex 191, 235 105, 191 191 30 14 Widespread in CNS 171, 168, 251 Widespread in CNS with glial and neuronal subtypes (see text) 25, 105, 120, 270 Widespread in CNS 340 For differences in results and methodologies, and discussion of NHE-6, see text. CNS, central nervous system. Physiol Rev • VOL 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 The activation of acid transport mechanisms can lead to a net transmembrane flux of osmoles, resulting in changes in cell volume. Linkages between volume regulation of glial cells and the regulation of cytosolic pH can therefore be anticipated. For example, activation of Na⫹/H⫹ exchange can lead to glial swelling, owing to the accumulation of Na⫹ (143a). In addition, the regulatory response to changes in osmolarity involving taurine transport can have immediate, direct effects on pHi (234a). Due to the role of glial swelling in clinical brain edema, this subject has received considerable attention in literature beyond the scope of the present review. For an overview of this topic, the contribution of Kempski et al. (161a) may be consulted. commonly denoted NHE-1 through NHE-7. Deduced amino acid sequences have indicated similarity in membrane-spanning domains and distinct putative regulatory domains in the cytosolic regions (54, 226, 336). The first member cloned was the human growth factor-activated NHE-1 (266), an abundant isoform thought to play a housekeeping role in maintenance of steady-state pHi. NHE-1 mRNA is widespread in the mammalian CNS (191, 235). Immunocytochemical analysis has also identified the protein in assays of whole tissue (105, 247) and astrocyte cultures (247). Transcripts, localized by Northern analysis and in situ hybridization, revealed a marked presence in hippocampus, periamygdaloid cortex, and cerebellum (191). A general increase in message (191) and protein (105) was noted in postnatal cortex. This observation may account for the lower resting pHi and the slower recovery from acid loading reported in fetal versus adult hippocampal neurons (24). However, NHE-1 is typically sensitive to amiloride (or its analogs) (226), a feature at odds with its predominance in hippocampus, where neuronal Na⫹/H⫹ exchange has little or no sensitivity to these agents (250, 273). It is undetermined whether this characteristic of CA1 pyramidal neurons is due to cell-specific alterations of NHE-1 or the presence of an alternative NHE isoform. mRNA for other NHE isoforms has been reported in the mammalian CNS. NHE-2 message was noted in whole brain (343), and later localized to cerebral cortex and brain stem (105, 191), where the mRNA level was apparently one-tenth that of NHE-1 (191). Message for NHE-3, however, was limited to Purkinje cells of the cerebellum (191) (but has been noted in medullary chemosensitive neurons in organotypic cultures, Ref. 348). An in situ hybridization study of NHE-4 reported a predominance of this isoform in hippocampus (30). Compared with NHE-1, -2 and -3, the sensitivity of NHE-4 to amiloride was low. NHE-4 may therefore appear to be a candidate for the neuronal transporter in hippocampus. However, concentrations of amiloride that had no effect pH REGULATION IN THE BRAIN Physiol Rev • VOL pears likely in the case of NHE-6 (47, 209, 230) and NHE-7 (229a). B. HCO3ⴚ Transporters in Brain The transport processes that regulate pHi through the transmembrane flux of bicarbonate fall within the bicarbonate transporter superfamily. Identified and cloned species correspond to the physiologically well-characterized ⫹ ⫺ categories of Cl⫺/HCO⫺ 3 exchange, Na -HCO3 cotrans⫹ ⫺ ⫺ port, and Na -driven Cl /HCO3 exchange. Subspecies of all three transporter classes have been identified in the CNS. 1. The Cl⫺/HCO3⫺ exchanger in the CNS This transporter falls within the anion exchanger (AE) gene family, which consists of at least four members (170). AE3 is expressed in both brain and heart (171), with polypeptide subtypes that differ in their NH2-terminal sequences (185). The AE3 transporter has been localized to CNS neurons (171, 251), and two polypeptide isoforms of the AE3 gene have been separately localized to Muller cells and horizontal neurons within retina (168). While AE3 mRNA was found in both fetal and adult hippocampal neurons, physiological Cl⫺/HCO⫺ 3 exchange appeared nearly absent in the younger cells (251). This suggests that inferences regarding AE3 function made on the basis of message localization should be regarded with caution. In one study, AE3 expression was altered in response to environmental changes. Experiments on cultured hippocampal neurons demonstrated increased AE3 immunoreactivity in response to exposure to ammonia (142). This effect may be related to the depolarizing shift in the GABAA receptor equilibrium potential that has been associated with long-term exposure to ammonium ions (188, 190). 2. Na⫹-HCO3⫺ cotransporters in the CNS Studies of the Na⫹-HCO⫺ 3 cotransporter (NBC) in mammalian brain have indicated a wide distribution in the CNS and the presence of more than one isoform. Using nucleotide probes derived from the electrogenic Na⫹-HCO⫺ 3 cotransporter of rat kidney, in situ hybridization revealed NBC mRNA in olfactory bulb, hippocampal dentate gyrus, and cerebellum, with localization to both glial cells and neurons (270). Immunostaining with polyclonal antibodies indicated that NBC protein was distributed diffusely in cell bodies and processes, as well as in epithelial cells of the choroid plexus, ependyma, and meninges. Colocalization with glial fibrillary acidic protein, microtubule-associated protein, or 2⬘,3⬘-cyclic mononucleotide 3⬘-phosphodiesterase demonstrated expression in particular subpopulations of astrocytes, neurons, and 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 on acid extrusion in hippocampal neurons (250, 273) could inhibit NHE-4 activity by 50 – 60% when expressed in fibroblasts (30). Moreover, NHE-4 could not be activated by acid loading under isosmolar conditions. An increase in acid extrusion required hyperosmolar media, suggesting its activation was linked to changes in cytoskeletal elements brought about by cell shrinkage (31). If NHE-4 displayed similar properties in hippocampal neurons, it could not be responsible for the normal extrusion of H⫹ by these cells. Later studies raise additional questions about the role of NHE-4 in hippocampus, as it was found to be principally located in cerebral cortex and brain stem-diencephalon (105, 191). Cloning and expression of the NHE-5 isoform in PS120 cells revealed a functional Na⫹/H⫹ exchanger with a sensitivity to EIPA. In situ hybridization demonstrated strong localization to the hippocampal dentate gyrus, with lower levels noted in the CA1 fields and cerebral cortex (14). In separate studies, cloning of NHE-5 and expression in fibroblasts (15), or Chinese hamster ovary cells (292), also yielded a functional Na⫹/H⫹ exchanger. Less specific localization to hippocampus and other regions was noted by Northern analysis (15). The NHE-6 isoform has a high degree of sequence identity to a mitochondrial Na⫹/H⫹ exchanger of yeast, termed NHA2. Its abundant expression in mitochondriarich tissues, including brain, increased suspicion that it also was a mitochondrial Na⫹/H⫹ exchanger (230). Recent studies have not confirmed a mitochondrial localization, however, and have provided evidence that the transporter is targeted to endoplasmic reticulum (209), or endosomes (47). NHE-7 is also in organelles, localized to the trans-Golgi network (229a). The function of individual NHE isoforms in brain is unclear, with the possible exception of NHE-1. It is likely that the principal role of NHE-1 in the CNS is the maintenance of steady-state pHi and the recovery from cytosolic acid loads. This is supported by recent studies of a spontaneous slow-wave epilepsy mutant mouse (89, 358). The autosomal recessive defect of these animals arises from a null allele of NHE-1 localized to chromosome four and is associated with ataxic gait, absence and tonicclonic seizures, and selective neuronal death in brain stem and cerebellum (89). In a study of pHi regulation in HEPES-buffered media, acutely dissociated CA1 pyramidal neurons from mutant mice displayed a more acidic state pHi and a lower rate of acid extrusion compared with cells from wild-type animals. In some instances, recovery from acid loading was virtually absent in the neurons from mutant animals (358). The roles of NHE-2 through NHE-7 are more subject to conjecture. Given the well-established utilization of Na⫹/H⫹ exchange in cell volume regulation (336), this function appears plausible, particularly for NHE-4 (30). Regulation of pH within organellar compartments ap- 1195 1196 MITCHELL CHESLER 3. Na⫹-driven Cl⫺/HCO3⫺ exchange in the CNS The Na⫹-driven Cl⫺/HCO⫺ 3 exchanger (NDCBE) was the first acid extrusion mechanism identified in early studies of snail neurons and squid axon (259, 303). In hippocampal neurons, where mammalian neuronal pHi regulation was extensively dissected, NDCBE played a major role in the extrusion of acid (273). An NDCBE cloned from the insulin-secreting cell line MIN6 cDNA library was found to have high mRNA levels in brain, and when expressed in oocytes, it mediated an Na⫹-HCO⫺ 3 influx in exchange for internal Cl⫺ (340). This isoform is likely related to the NDCBE1 cloned and extensively characterized by Grichtchenko et al. (131). Northern blots revealed robust presence of the isoform in all major regions of human brain. When expressed in oocytes, it mediated an Na⫹-dependent, DIDS-sensitive increase in pHi associated with a rise in [Na⫹]i with a 1:2 stoichiometry of Na⫹ to ⫺ HCO⫺ 3 . In addition, isotopic Cl efflux from oocytes was ⫹ ⫺ Na and HCO3 dependent, blocked by DIDS, and required external Cl⫺ to run in reverse mode. VI. MODULATION OF INTRACELLULAR pH BY NEURONAL ACTIVITY Both neurons and glia cells can undergo substantial shifts in pHi in response to electrical stimulation or application of specific membrane ligands. For both classes of cells, these pH changes can occur within seconds, in either the acidifying or alkalizing direction, and can attain magnitudes of several tenths of a pH unit. Physiol Rev • VOL A. Modulation of pHi in Neurons 1. pHi shifts associated with depolarizing stimuli A number of investigators have demonstrated a fall in neuronal pHi associated with membrane depolarization or repetitive spike activity. Many of these studies were first performed on large invertebrate cells. In barnacle photoreceptors, a fall in pHi was correlated with the depolarizing receptor potential (51). Repetitive firing was associated with a cytosolic acidification in molluscan (1) and leech (261) neurons. In the former study it was found that the acidification induced by repetitive spike activity was dependent on the entry of Ca2⫹. Depolarization of snail neurons also elicited a Ca2⫹-dependent acidification (202). Meech and Thomas (200, 201) had reported that injection of Ca2⫹ into snail neurons caused cytosolic acidification and found that this response was partly mediated by mitochondrial Ca2⫹/H⫹ exchange, but was largely unexplained. It was later demonstrated that the Ca2⫹-linked acidification of snail neurons was sensitive to vanadate. This finding implicated the Ca2⫹/H⫹ exchange property of the plasmalemmal Ca2⫹-H⫹-ATPase (reviewed in Refs. 61, 62) in the origin of the intracellular acid shift (274). It should be mentioned that snail neurons can also generate transmembrane H⫹ fluxes though a specific voltage-gated H⫹ conductance (57, 305). Because this conductive pathway activates at potentials positive to the H⫹ equilibrium potential, and exhibits pronounced outward rectification, its main role appears to be the extrusion of H⫹ (reviewed in Ref. 92). Depolarizing stimuli have also been associated with a fall in pHi in vertebrate neurons. In studies of lamprey reticulospinal cells (74) and frog motoneurons (107), intracellular acidosis was associated with depolarization, due to elevation of [K⫹]o or exposure to excitatory amino acids, respectively. In catfish horizontal cells, application of glutamate elicited a fall in pHi that was responsible for suppression of voltage-gated Ca2⫹ channels (102). Cultured mammalian neurons have also been found to acidify in response to excitatory amino acids, and these intracellular acid shifts were linked to entry of Ca2⫹ (60, 134, 143, 341, 346). In a brain stem slice preparation, acid shifts in vagal neurons were noted in response to spike activity or depolarization under voltage clamp. These acid shifts appeared to have a dependence on Ca2⫹ entry as they were blocked by Cd2⫹ and Ni2⫹ but persisted in the presence of tetrodotoxin or HCO⫺ 3 -free saline (316). In addition, activity-dependent, Ni2⫹-sensitive intracellular acid shifts have been reported in rat thalamic neurons (207). In contrast, in hippocampal CA1 neurons studied in acute slices, depolarizing agents [including N-methyl-Daspartate (NMDA)] evoked intracellular acidifications that displayed a minor dependence on Ca2⫹ and were largely attributed to production of metabolic acid (361). 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 oligodendrocytes, respectively. Screening of cDNA libraries resulted in the cloning of two rat brain NBC isoforms, rb1NBC and rb2NBC, which were identical except for a unique 61-amino acid COOH terminus of the latter (25). Generation of polyclonal antibodies to distinguish between the carboxy ends of the isoforms revealed predominant rb1NBC protein in astrocytes compared with neurons, whereas rb2NBC antibody revealed the opposite pattern. When expressed in oocytes, rb2NBC was found to mediate Cl⫺-independent, DIDS-sensitive, electrogenic Na⫹-HCO⫺ 3 cotransport. An NBC isoform cloned by Giffard et al. (120) was similarly found to generate DIDSsensitive, HCO⫺ 3 -dependent currents when expressed in oocytes, and in situ hybridization showed widespread presence in rat CNS, and colocalization with glial acidic fibrillary protein. NBC mRNA was detected in brain at P0 with an increase at P15 and persistence into adulthood. In spinal cord, in contrast, expression was noted by embryonic day 17. Western blots from microsomal preparations of rat brain indicated NBC protein expression as early as embryonic day 16 in cortex and brain stem-diencephalon (105). pH REGULATION IN THE BRAIN Physiol Rev • VOL surface pH from a single cell, or as interstitial pH, when studying responses from a tissue. Although depolarizing stimuli are most often associated with a fall in pHi, alkaline responses have occasionally been noted. Gaillard et al. (114) found that elevation of [K⫹]o elicited a Cd2⫹-sensitive alkalization of cultured cerebellar Purkinje cells, consistent with Ca2⫹ dependence (however, depolarization evoked only acid responses in Purkinje cells studied in acute slices) (349). Recording pHi from the CA1 pyramidal cell layer of hippocampal slices, Zhan et al. (362) found that application of NMDA caused an initial intracellular alkalization followed by an acid shift. The alkaline response was inhibited in low Ca2⫹ media, while the late acid shift appeared to have a metabolic origin. An alkalization of spinal cord neurons was reported in response to elevation of external K⫹ and was speculated to arise via an electogenic Na⫹HCO⫺ 3 cotransporter (46). With prolonged neural activity, cytosolic acidification may occur due to an increase in the production of metabolic acids such as CO2 and lactate. These products could be generated from the energetic demands associated with the influx of Ca2⫹ (341), but could also arise from a host of metabolic needs including the ATP requirements of the Na⫹-K⫹ pump (186). Metabolic processes are likely to underlie the considerable acidosis associated with seizure and spreading depression. Because a metabolically derived acidosis is expected to lower both intracellular and interstitial pH, the acid changes associated with prolonged stimulation, seizure, and spreading depression are considered further below, in context with extracellular acid shifts (see sects. VIIB6 and VIII). 2. pHi shifts associated with GABAA and glycine receptors In characterizing the properties of GABAA receptors, Bormann et al. (34) noted a significant permeability to bicarbonate ions. The equilibrium potential for HCO⫺ 3 (which is the same as that of H⫹) typically lies between 0 and ⫺20 mV. Thus, at more negative membrane potentials, the opening of a GABAA anion channel will result in an HCO⫺ 3 efflux, leading to intracellular acidification and extracellular alkalization. This was first reported by Kaila and Voipio (154), who demonstrated that the opening of GABAA receptors on crayfish muscle produced a bicarbonate-dependent, intracellular acidification accompanied by a surface alkalization. Kaila and colleagues also demonstrated a GABAA receptor-mediated acidosis in crayfish stretch receptors (330), cultured astrocytes (151), and acutely isolated hippocampal CA1 pyramidal neurons (244). In situ evidence for similar intracellular acid shifts was provided by Ballanyi and colleagues (316), who demonstrated an HCO⫺ 3 -dependent, bicuculline-sensitive fall in pHi in vagal neurons that was evoked by 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 An interesting, rapidly spreading acidification of cerebellar cortex (in vivo) was reported by Ebner and colleagues (66, 67), using neutral red as an epifluorescent pHi indicator. These responses were elicited by synaptic activation of the cerebellar Purkinje cells through stimulation of the parallel fibers (66, 67). The basis of these acid responses is unclear. They may be related to a recently reported, rapid, activity-linked, Ca2⫹-related acidification of Purkinje cell dendrites studied in acute cerebellar slices (349). The causes of Ca2⫹-dependent cytosolic acidosis may be multiple. Intracellular acidosis arising from elevated [K⫹]o or glutamate has been linked to the generation of metabolic acid (341) as well as mitochondrial Ca2⫹/H⫹ exchange (346). In a voltage-clamp study of hippocampal CA1 pyramidal cells (317), Ca2⫹ entry through voltage-gated channels caused a cytosolic acid shift that appeared unrelated to mitochondrial Ca2⫹ transport, as it was not significantly reduced when ruthenium red was included in the patch pipettes. The acid shifts were significantly smaller when the electrodes contained 1–5 mM vanadate, or 100 M eosin. These data were consistent with cytosolic acidification mediated by the Ca2⫹/H⫹ exchange property of the plasmalemmal Ca2⫹ATPase, as was noted for snail neurons (274). In evaluating data from such studies it should be borne in mind that the available tools are still rather poor. The use of BCECF to measure acid shifts may confuse the quantitative interpretation of results, as this agent is itself an inhibitor of the plasma membrane Ca2⫹-ATPase (115). For the human erythrocyte pump, the inhibition constant (Ki) of BCECF is 100 M (115), and only 30 M BCECF was able to cause half-inhibition of the plasmalemmal Ca2⫹-ATPase in snail neurons (275). Qualitative difficulties in interpretation can also arise because of the poor specificity of the transport inhibitors. In addition to blocking the plasmalemmal Ca2⫹-ATPase, vanadate has been reported to inhibit mitochondrial Ca2⫹/H⫹ exchange (with a Ki of 0.6 mM) (173) and is well-established as a blocker of sarcoplasmic/endoplasmic reticulum Ca2⫹ (SERCA) pumps (233). SERCA pumps are also inhibited by low concentrations of eosin (with a reported Ki of 0.8 M) (173). Moreover, like the plasmalemmal Ca2⫹-ATPase, the SERCA pumps exchange Ca2⫹ for H⫹ (85, 359). Thus the effects of eosin and vanadate on pHi may not be solely attributable to their inhibition of the plasmalemmal transporter. These considerations underscore the challenges faced when seeking to establish the basis of an intracellular acid shift. Cytosolic acidosis may arise from a wide variety of organellar and metabolic sources in addition to the movement of acid equivalents across the plasma membrane. To distinguish transmembrane movements of acid, it is therefore advantageous if associated pH transients can be recorded in the extracellular domain, either as 1197 1198 MITCHELL CHESLER GABA. Brain stem medullary neurons also exhibited a GABAergic acidosis. In addition, these cells could be acidified by activation of strychnine-sensitive glycine receptors (189), which have an HCO⫺ 3 permeability comparable to that of GABAA receptors (34). 3. Additional modulators of neuronal pHi B. Modulation of pHi in Glial Cells Neuronal activity has long been associated with depolarizing responses of astrocytes, which have been mainly attributed to the transient elevation of [K⫹]o (132, 234, 253, 284). A consequence of astrocyte depolarization is the activation of electrogenic Na⫹-HCO⫺ 3 cotransport, which causes a rapid cytosolic alkaline shift. In the great majority of studies, these alkaline responses were evoked in vitro by the elevation of [K⫹]o (44, 50, 129). The recording of similar alkalizations which followed the elevation and fall of [K⫹]o was reported for cortical astrocytes in vivo by Chesler and Kraig (81, 82) using double-barreled pH microelectrodes. Stimulation of the cortical surface elicited typical slow depolarizing responses that were accompanied by a rapid rise in pHi of several tenths of a unit pH. After tetanic stimulation, the pHi recovered and was sometimes followed by a prolonged intracellular acidosis. During passage of a cortical spreading depression, the alkaline-acid sequence was repeated but was exaggerated in amplitude, with some alkaline transients apPhysiol Rev • VOL 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 A variety of endogenous agents have been found capable of altering cytosolic pH. Connor and Hockberger (88) injected molluscan neurons with cAMP or cGMP and noted either a slow acidification or a biphasic alkalineacid response. In cultured mammalian neurons, an acid shift was elicited by activation of type I metabotropic glutamate receptors (6). The mechanisms underlying these pHi changes have not been determined. An interesting alkalization of acutely isolated hippocampal CA1 neurons was reported by Church and colleagues (280). Norepinephrine, acting through -adrenergic receptors, elicited a concentration-dependent, sustained rise in steady-state pHi and increased the rate of pHi recovery from imposed acid loads. This alkalosis was Na⫹ linked, but was not dependent on HCO⫺ 3 or the elevation of [Ca2⫹]i, and appeared to be mediated by increased activity of the amiloride-insensitive Na⫹/H⫹ exchanger. The effect was linked to G protein-coupled stimulation of adenylate cyclase with subsequent activation of the cAMP-dependent protein kinase. These observations suggest a means by which pHi could vary as a function of the metabolic and electrophysiological history of a set of neurons. In this regard it may be noted that CA1 pyramidal cells have been found to fall into distinct populations with relatively low or high baseline pHi values (24). proaching 0.8 pH units. In recordings from medullary glial cells in vivo, Ballanyi et al. (18) reported similar intracellular alkaline shifts evoked by stimulation of spinal pathways. Intracellular alkaline shifts in response to neural activity have also been reported in the giant glial cells of the leech (an annelid worm) (94). It therefore appears that this glial property has been conserved through evolution. An activity-dependent alkalization of glial cells may have functional significance in several respects. A rise in cytosolic pH can elevate glycolytic rate (321), increase gap junctional conductance (289), and favor the uptake of glutamate (43), or the formation of glutamine (48). On the other hand, when viewed from an extracellular perspective, the Na⫹-HCO⫺ 3 cotransporter would appear to function as an acid secretory mechanism, as discussed in section VIIB6. Another form of astrocyte alkalization has been associated with activation of metabotropic glutamate receptors. Amos and colleagues (5, 6) described a group I mGluR-mediated rise in pHi of cultured cortical and cerebellar astrocytes. The response was independent of Na⫹ but required HCO⫺ 3 (5). The alkaline shift was apparently Ca2⫹ dependent, as it could be triggered by caffeine and ionomycin (5), and was mitigated by intracellular BAPTA (6). Elevation of [Ca2⫹]i appeared to be dependent on its release from intracellular stores, as the response persisted in the absence of external Ca2⫹. Acutely dissociated astrocytes also exhibited an mGluR-mediated alkalization, suggesting that similar responses might occur in vivo (5). As discussed for neurons, glial cell acidosis may be expected to arise from a variety of sources, which include transmembrane acid transport, organellar acid fluxes, and metabolic production of CO2 and lactic acid. Late acidifications of astroctyes studied in vivo were noted by Chesler and Kraig following repetitive neural activity (81, 82). Application of excitatory amino acids was also found to acidify astrocytes in a number of studies (49, 52, 263). Acidosis of glial cells induced by glutamate has been largely attributed to its inward transport (52, 263), as the uptake of this amino acid is coupled to the influx of H⫹ (43, 344, 360). Accordingly, superfusion of hippocampal slices with transportable glutamate analogs was found to cause a fall in tissue pHi (3). Acidification of astrocytes has also been noted in response to the excitatory amino acid kainate, which is not taken up by the glutamate carrier (165). In CO2/HCO⫺ 3containing media, Rose and Ransom (263) found that the kainate-evoked acidification of hippocampal astrocytes was preceded by an alkaline shift. It was proposed that these pHi shifts were governed by the Na⫹-HCO⫺ 3 cotransporter responding to sequential depolarization and hyperpolarization of the membrane. In support of this contention, they found little or no kainate-induced acidification pH REGULATION IN THE BRAIN in CO2/HCO⫺ 3 -free media. In contrast, in cerebellar astrocytes, kainate-evoked acid shifts were prominent in the absence of CO2 and HCO⫺ 3 (52), suggesting an alternate mechanism. The kainate responses of the cerebellar astrocytes did not require external Ca2⫹, unlike acidifications evoked by excitatory amino acid receptors in neurons (60, 134, 143, 341, 346). VII. REGULATION AND MODULATION OF BRAIN EXTRACELLULAR pH A. The pH of Interstitial Fluid Physiol Rev • VOL shifts in blood flow remain plausible, however, and evidence for slow extracellular alkaline shifts arising from postactivity hyperemia has been published recently by Venton et al. (325a). B. Changes in Interstitial pH Caused by Neural Activity Synchronous activation of nerve cells has been shown to cause changes of interstitial pH in a variety of preparations. Reports of activity-dependent pHo shifts go back over 60 years. Many of the earlier measurements are suspect, as extracellular potential shifts were not subtracted from the pH recordings, and the spatial and temporal resolutions were poor (76). The first reliable recordings of pHo transients following synchronous cortical activity were performed in vivo, using pH microelectrodes in combination with an extracellular reference microelectrode to enable subtraction of interstitial potential changes. Stimulated activity, seizure, or spreading depression were associated with an initial interstitial alkaline shift followed by an acidosis that could persist for minutes. This pattern was first noted in cortex (65, 322) and cerebellum (174, 253a) and has been commonly found throughout the CNS. In some instances, however, neural activity has been associated with predominant early acid shifts (see sect. VIIB6). The processes responsible for the generation of activity-dependent pHo shifts have not been fully elucidated. In principle, a large number of transport mechanisms could contribute to the dynamic H⫹ balance of the brain interstitial space. To understand whether a flux of given magnitude can account for an observed pHo transient, however, one must have a quantitative appreciation of the interstitial buffering capacity. Therefore, before addressing the origin of the activity-dependent pHo transients, the nature of pH buffering in the brain interstitial space will be considered. 1. Buffering of brain interstitial fluid The acid-base status of extracellular fluid is governed by the chemical buffering properties of the solution and by the transmembrane flux of acid equivalents. In the adoption of nomenclature, it is formally correct as well as heuristically useful to distinguish between these components. Buffering is a term best reserved for the chemical equilibria within the interstitial fluid which act to mitigate the change in pH arising from an imposed acid or base load. Transmembrane fluxes, in contrast, constitute the source of acid-base loads imposed upon the interstitial fluid. In some instances, a pHo shift caused by one transmembrane flux may be masked by a second flux of opposite direction. The term pH muffling has sometimes been 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 pH microelectrode studies have indicated that brain interstitial fluid is slightly more acidic than blood, with a pH of ⬃7.3 (90, 145, 174). In brain slices, however, the interstitial pH is usually lower, despite the flow of saline buffered to pH 7.4. At the center of a brain slice, baseline pHo values of 7.1–7.2 have been reported (16, 71, 144, 180, 338). Low pHo has also been noted in other isolated CNS preparations including lamprey brain stem (75), frog spinal cord (107), and turtle cerebellum (139). The cause of baseline acidosis in hippocampal slices was studied by Voipio and Kaila (328) using a unique, dual CO2-pH microelectrode. It was demonstrated that the low pH of the interstitial fluid was entirely due to elevation of tissue PCO2. In their interface-style recording chamber, the slice PCO2 ranged from 40 to 50 Torr. In a later study, PCO2 was calculated in submerged hippocampal slices using pH and CO2⫺ 3 -sensitive microelectrodes (78). The submerged slices had a considerably higher baseline CO2 (⬃75 Torr), which again accounted entirely for the difference in pH between bath and interstitial fluid. The high concentration of CO2 in brain slices is likely due to its generation by aerobic metabolism, and poor clearance, owing to lack of blood flow. In a slice, the egress of CO2 would be reliant on diffusion across the tissue and its unstirred layer of surface fluid. CO2 clearance can therefore be expected to be greater (and thus the PCO2 lower) in an interface chamber, where CO2 is able to diffuse across both the upper and lower surfaces (225). In vivo, CO2 is cleared by blood flow, and accordingly, large steady-state accumulation of CO2 is not expected. With a physiological rise in perfusion, due to elevated neural activity, increased clearance of CO2 and local increases in brain pH could arise, in principle. Indeed, observations of rapid, activity-dependent increases in brain interstitial pH were at first attributed to augmented blood flow (322). The identification of similar pH shifts in numerous in vitro studies (where blood flow was absent) indicated that alternate mechanisms must account for these rapid pH changes. Candidate processes are the subject of detailed discussion in the sections which follow. Alterations in brain pH due to physiological 1199 1200 MITCHELL CHESLER CA CO 2 ⫹ H 2 O | 0 H 2 CO 3 | 0 HCO 3⫺ ⫹ H ⫹ (7) Within the physiological time frame, formation of ⫹ HCO⫺ 3 and H from H2CO3 occurs virtually instantly. The hydration of CO2 to form carbonic acid (H2CO3) and its dehydration back to CO2 and water are extremely slow, however, with time constants on the order of 20 s and 1 s, respectively (192). Accordingly, the attainment of equilibrium following an acid or base challenge can require minutes. Biological systems typically overcome these slow kinetics using various forms of the enzyme carbonic anhydrase (CA), which catalyzes both the hydration and dehydration steps. In the brain interstitial space, CA activity is the principal regulatory factor governing the manifestation of pH transients. Indeed, the characterization of Physiol Rev • VOL activity-dependent pHo shifts and the function of interstitial CA are inseparable issues and therefore will be considered together. 2. Interstitial pH transients and CA The first evidence that CA played a role in the regulation of interstitial pH came from the work of Kraig et al. (174). These investigators reported alkaline-acid pHo transients in rat cerebellum in response to stimulation of the parallel fibers, or the passage of a spreading depression. When the cerebellar cortex was superfused with 5 mM acetazolamide, the amplitude of the alkaline and acid shifts increased two- to threefold. Enhancement of evoked alkaline shifts by millimolar concentrations of acetazolamide was also noted in rat cortex (215) and in brain slices (63, 338). Comparable effects, however, could be obtained in hippocampal slices using as little as 1 M acetazolamide (71). At the time of the earliest studies of pHo transients, it was known that brain CA resided in glia (119, 164, 228), within both oligodendrocytes (58) and astrocytes (59). In view of the permeability of cell membranes to acetazolamide (192), and the high concentrations used in the initial experiments, it appeared likely that glial CA would have been inhibited. However, the means by which inhibition of glial intracellular CA could so profoundly affect rapid pHo transients was not obvious, and Kraig et al. (174) were careful to stress that the site of action of acetazolamide was uncertain. Clarification came from a series of experiments using CA inhibitors that do not readily cross cell membranes. Activity-dependent pHo shifts were profoundly affected by these agents. It was therefore reasoned that the site of enzyme activity had to reside in the extracellular space (71, 72, 150). These CA inhibitors included benzolamide, which is relatively impermeant due to its low pKa (318), as well as a set of similarly potent sulfonamides conjugated to high-molecular-weight dextrans (117, 156). In one series of experiments alkaline shifts were evoked in area CA1 of hippocampal slices by either stimulation of the Schaffer collaterals, or by local microejection of glutamate (71, 72). Superfusion of benzolamide (71) or a dextran-conjugated sulfonamide (72) increased the amplitude of these transients severalfold. This increase in the amplitude of an alkaline transient is expected to occur when the pHo shift is generated by the loss of protons from (or addition of OH⫺ to) the interstitial compartment (Fig. 1A). Since CO2-HCO⫺ 3 is the principal extracellular buffer, the ability to replenish extracellular H⫹ hinges on the CA-dependent rate of CO2 hydration. With inhibition of interstitial CA, the generation of H⫹ is reliant on the slow, uncatalyzed hydration of CO2, which leads to a poorly buffered, or “amplified” alkalosis. CA inhibitors will similarly amplify alkaline shifts elicited 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 used to distinguish such a process from the ever-present chemical buffering (304). The buffering capacity of a fluid at equilibrium is governed by the concentrations of the individual buffers and their respective pKa values. In brain interstitial fluid, the principal buffer is CO2/HCO⫺ 3 with a total combined concentration of ⬃27 mM. In comparison, the contributions of other buffers appear negligible (109). Quantification of the CO2/HCO⫺ 3 buffering capacity in the physiological setting typically relies on two assumptions. First, the buffer components are considered to be at equilibrium, and second, the system is treated as “open” with respect to CO2 (i.e., the PCO2 is constant). When these conditions are met, then the extracellular buffering capacity (e) is reliant solely on bicarbonate concentration, and is given by Equation 3. Thus, for an [HCO⫺ 3 ]o of 26 mM, e would be 60 mM. However, as discussed below, neither fixed PCO2 nor its equilibrium with HCO⫺ 3 appear to occur under all conditions in which activity-dependent pHo shifts are generated. Accordingly, the effective buffering capacity can be considerably less than 60 mM. In rat hippocampal slices it has been shown that synchronous neuronal activity can cause interstitial PCO2 to rise by 20 –30 Torr, accounting entirely for the prolonged extracellular acidosis (328). Since the PCO2 is not fixed, interstitial buffering would be less than maximal during these pHo transients. Acid shifts also occur after synchronous activity in vivo (174, 278, 285), but unlike the slice preparation, blood flow is intact and the contribution of CO2 to these phenomena is therefore likely to be less. Over periods of minutes, PCO2 in vivo is governed by the blood flow and pulmonary ventilation, and long-term buffering can be considered to occur at a roughly fixed PCO2. A second important constraint on CO2/HCO⫺ 3 -dependent buffering arises from the kinetics of the carbonic acid reactions pH REGULATION IN THE BRAIN 1201 by any extracellular “proton sink” including alkalizations imposed by local microejection of NaOH (72). Properties of this activity-evoked, HCO⫺ 3 -independent alkalosis are detailed in section VIIB4. In another set of experiments, inhibitors of extracellular CA were used to study extracellular alkaline shifts evoked by activation of GABAA receptors (71, 150, 153, 154). Work on crayfish muscle demonstrated that the efflux of HCO⫺ 3 across GABAA anion channels gives rise to a surface alkalosis (153, 154). In this circumstance, rather than reduce an alkalosis, CA was required to generate the alkaline shift, since the fall in surface H⫹ required the rapid dehydration of H2CO3 (Fig. 1B). Thus, when CA on the surface of the muscle was inhibited with Physiol Rev • VOL benzolamide, the dehydration of H2CO3 was slowed, and the GABA-evoked surface alkalosis was prevented (153). Similar effects were later noted when benzolamide and dextran-conjugated inhibitors of CA were used to study GABAA receptor-mediated extracellular alkalosis in the CA1 region of hippocampal slices (71, 150). Studies of this HCO⫺ 3 -dependent alkalosis are detailed in section VIIB5. The contrasting effects of CA inhibitors on alkaline transients demonstrated that there are two classes of interstitial alkalization that may accompany normal sequences of excitation and inhibition. Excitation appears to generate a rapid proton sink. Because glial cells secrete acid into the interstitial space, and undergo an intracellular alkalization during neural activity, it has been specu- 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 FIG. 1. Effect of extracellular carbonic anhydrase (CA) inhibition on activity-dependent alkaline shifts. A, left: extracellular alkalosis generated by an H⫹ sink (or equivalently an efflux of OH⫺) would be buffered via the CA-catalyzed hydration of CO2. Right: inhibition of extracellular CA by benzolamide would diminish the rate of hydration, permitting the development of a larger alkalosis. B, left: extracellular alkalosis generated by an HCO⫺ 3 source is generated by the CA-catalyzed dehydration of H2CO3. Right: inhibition of extracellular CA by benzolamide slows the dehydration step, thereby limiting development of the alkalosis. 1202 MITCHELL CHESLER 3. Nature of interstitial CA activity in brain Extracellular CA activity has been found in tissues outside the CNS associated with the membrane-tethered isoform CA IV (64). Examples include the luminal brush border of renal tubules (350), the surface of muscle fibers (335), and the luminal endothelial surface in lung (334). In the nervous system, in contrast, there has been little morphological evidence of extracellular CA. Histochemical studies suggested the presence of interstitial CA in dorsal root ganglia (106); however, such methods did not allow definitive extracellular localization. The physiological demonstration of extracellular CA in hippocampal slices (71, 72, 150) nearly coincided with immunocytochemical studies of brain using antibodies against CA type IV. In the latter study, CA IV was found on the luminal membrane of brain endothelial cells, but CA IV staining was not observed in brain tissue proper (118). More recently, immunostaining for CA IV was described on scattered glial cells and neurons in mammalian brain (342). The consistency of the physiological evidence for extracellular CA is at apparent odds with the variability of the morphological studies. This might be explained if interstitial CA activity detected by physiological means in brain slices was an artifact of cellular injury. Indeed, because CA is abundant in glial cells (58, 59, 119, 164, 221a, 228), and there is physiological (207, 244) and morphological evidence (2, 158, 227, 229, 351, 352) for its presence in neurons, it is reasonable to assume that release of intracellular CA would occur when tissue is cut in the preparation of brain slices. Yet, amplification of alkaline transients by acetazolamide (174) and benzolamide (140) was also noted in vivo, suggesting that damage due to slicing was not a primary factor. Indeed, because exogenous CA was found to readily diffuse to and from the interstitial space of brain slices (140), the enzyme would Physiol Rev • VOL be expected to rapidly wash out of the tissue when released from injured cells. Additional evidence that simple spillage of the enzyme could not account for interstitial CA came from a study of hippocampal slices prepared from mutant mice lacking CA II. In a comparison of mutant versus wild-type animals, activity-dependent interstitial alkaline shifts were comparable in amplitude and were amplified in a similar manner by benzolamide (311). Moreover, histochemical studies in brains of CA II-deficient mice detected the presence of membrane-associated CA (255) near neuronal processes. The inability to detect interstitial CA with antisera against CA IV could have several explanations. Extracellular isoforms other than CA IV may account for brain interstitial CA activity. Alternatively, the concentration of CA IV in the interstitial space could be below the detection limit of the immunocytochemical assay. This would not be unprecedented, as early immunocytochemical studies did not detect the CA II isoform in astrocytes (58). These possibilities are not mutually exclusive, and evidence now exists for both a low concentration of interstitial CA IV and the presence of an additional extracellular CA isoform. A suggestion that interstitial CA was present in low concentration came from experiments in which pHo, [HCO⫺ 3 ]o, and PCO2 were monitored during repetitive stimulation of rat hippocampal slices. The rise in [HCO⫺ 3 ]o associated with the extracellular alkaline transient was sometimes less than the value predicted by the Henderson-Hasselbalch equation, suggesting lack of equilibrium ⫹ between CO2, HCO⫺ (79). This result is consis3 , and H tent with an insufficient rate of CO2 hydration owing to low enzyme activity (Fig. 1A). In a later study, addition of exogenous bovine CA to physiological saline decreased evoked alkaline shifts in rat cortical slices by two- to threefold (140). A similar increase in interstitial buffering capacity was also noted when exogenous CA was superfused over rat cortex in vivo (140). The augmentation of buffering occurred when the exogenous CA concentration was ⬃300 nM but was not observed at 30 nM. After addition of 300 nM enzyme, increasing its level to 3,000 nM had no additional effect on the pH transients. Assuming a molar activity comparable to the added CA, these data were consistent with an overall interstitial CA concentration on the order of 30 –300 nM. Tong et al. (310) provided evidence that interstitial CA activity in brain is attributable to CA IV. The type IV CA isoform is tethered to the extracellular face of membranes by a phosphatidylinositol-glycan linkage. Accordingly, treatment with phosphatidlyinositol-specific phospholipase C (PIPLC) can liberate the enzyme and enable its washout (335, 364). When hippocampal slices were incubated with PIPLC, interstitial alkaline shifts were increased markedly compared with controls, and the am- 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 lated that this proton sink is not a glial response, but rather, represents the movement of H⫹ equivalents into a neuronal compartment (76, 94, 252). Possible mechanisms for the proton sink are considered in section VIIB4. GABAA receptor-mediated inhibition, on the other hand, appears to generate a local, extracellular “bicarbonate source.” A second implication of these studies was that a form of CA was present in the interstitial space of the brain slice preparation. Given the similar effects of acetazolamide in vivo (174, 215), it appeared likely that this extracellular CA activity was also present in the intact brain. This was confirmed when benzolamide was found to amplify AMPA-evoked alkaline shifts in vivo (140). Similar amplification of interstitial alkaline transients was later reported in nonmammalian preparations exposed to CA inhibitors, including isolated turtle cerebellum (121) and leech segmental ganglia (262). pH REGULATION IN THE BRAIN 4. Bicarbonate-independent alkaline shifts Although its features have been studied in some detail, the mechanism underlying the activity-evoked, bicarbonate-independent, “net proton sink” (Fig. 1A) has not been established. This alkalosis has been extensively characterized in rat hippocampal slices (63, 71, 72, 124, 130, 144, 177, 237, 238, 282, 283, 329, 338). In this preparation, optical recordings of interstitial pH performed with phenol red (177) or dextran-conjugated fluorescein (124) revealed an onset within 10 ms. Significant characterization of this alkalosis has also been performed in turtle cerebellum (69, 77, 84) and rat lateral geniculate Physiol Rev • VOL nucleus (312), where the basic features appeared similar to the hippocampal response. Differentiation from the GABAA receptor-mediated alkaline shift became apparent in early experiments, since the response persisted in HEPES-buffered Ringer solution (77, 84, 122) and in the presence of picrotoxin (69, 71). These results also indicated that an increase of interstitial HCO⫺ 3 concentration, presumed to arise from activitydependent shrinkage of the extracellular space (101), could not be responsible for the alkalosis (174). Amplification of the responses by extracellular CA inhibitors (Fig. 2A) provided a further means of differentiation from the GABAergic response (71, 72) (discussed in sect. VIIB2). A) IS THE ACTIVITY-DEPENDENT PROTON SINK DUE TO GLUTAMATE UPTAKE? In principle, a number of transport mechanisms could provide a pathway for an HCO⫺ 3 -independent, net proton sink, but might not generate an interstitial base load of sufficient magnitude to account for the observed extracellular pH change. One example is glutamate transport, where the electrogenic, Na⫹-coupled uptake of this amino acid is coupled to an HCO⫺ 3 -independent, influx of H⫹ (43, 344, 360). Under conditions of low extracellular buffering capacity, this mechanism can generate a surface alkalosis on retinal glial cells (43). The carrier can also cause a prolonged intracellular acidosis of brain slices when transported amino acids are superfused over the tissue (3). Yet, glutamate uptake does not appear to be responsible for the Schaffer collateral-evoked interstitial alkaline shift in hippocampal slices. This contention is supported by the fact that the synaptically evoked alkalosis can be blocked by 6-cyano-7-nitroquinoxaline-2,3dione (CNQX) and DL-2-amino-5-phosphonovalerate (APV). In the presence of these ionotropic glutamate receptor antagonists, postsynaptic currents are abolished; however, the presynaptic release of glutamate and its subsequent uptake should be unaffected (70). It is also notable that a similar interstitial alkalosis can be evoked by antidromic activation of the CA1 pyramidal neurons in the presence of CNQX, APV, and picrotoxin (70, 130, 237). These data suggest that in the hippocampal CA1 region, postsynaptic electrical activity is responsible for generation of the extracellular proton sink (see below). Given the influx of an acid equivalent linked to glutamate uptake, this process should contribute in some measure to the interstitial alkalosis during excitatory synaptic transmission. Yet, when transportable glutamate analogs (e.g., D-aspartate) were injected into hippocampal slices, extracellular alkaline shifts could not be detected unless the buffering capacity was reduced with benzolamide (281). Thus, under normal buffering conditions, alkaline shifts generated by glutamate transport do not have a measurable impact on the broad interstitial space in which the carriers reside. It is plausible, however, that pH shifts associated with glutamate uptake attain a sig- 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 plifying effect of benzolamide was largely occluded. The incubation fluid from PIPLC-treated slices contained a low concentration of CA that was insensitive to SDS, a property consistent with CA IV (197, 347). Western blots confirmed the presence of CA IV in this fluid. The immunocytochemical detection of CA IV required the pooled samples from several slices, consistent with a low interstitial CA IV activity. Based on the liberated SDS-insensitive CA IV activity, the interstitial concentration was estimated to be on the order of 100 nM, a figure consistent with the earlier estimate of 30 –300 nM (140). It is notable that treatment with PIPLC never eliminated all interstitial CA activity in hippocampal slices, since addition of benzolamide always caused some enhancement of alkaline transients (310). Incomplete cleavage of phosphatidylinositol-glycan linkages may explain this result. Alternatively, an additional membrane-associated CA could account for the persistence of CA activity. One candidate would be the CA XIV isoform, a membrane-spanning protein with an apparent extracellular catalytic domain. Transcripts of CA XIV have been detected in brain Northern blots (112, 213), and immunocytochemical studies have indicated its presence in medulla, pons, hippocampus, cerebellar granular layer, and major white matter tracts (242). The extent to which CA XIV catalyzes interstitial buffering is not known. Interstitial CA in brain is likely due to surface enzyme located on both glial cells and neurons. An association with glia was noted in two early reports. Physiological studies demonstrated interstitial CA activity in gliotic hippocampal slices, where neuronal elements were virtually absent (128) (see Fig. 5A and sect. VIIB6). Surface CA activity on individual, acutely isolated retinal Muller cells (astrocyte variants) was also detected by Newman (222) (see sect. VIIB6). In a recent study, the reliance of lactate influx on surface CA was used to demonstrate the presence of this enzyme on both astrocytes and neurons (289a). The type of surface CA on the astrocytes and neurons cells was not addressed; however, the immunocytochemical study of CA XIV has suggested that neurons specifically express this isoform (242). 1203 1204 MITCHELL CHESLER nificant magnitude in microdomains close to the sites of transport. The ability of ionotropic glutamate antagonists to prevent the synaptic generation of the proton sink in cerebellum (77) and hippocampus (70) suggested that the response was triggered by excitation of the postsynaptic elements. Indeed, in turtle cerebellum, direct activation of glutamate receptors induced an alkaline shift that persisted after synaptic transmission was blocked by Mn2⫹ or Cd2⫹ (84). In hippocampus, confirmation of the postsynaptic origin came from experiments in which picrotoxin-resistant alkaline shifts were elicited by repetitive antidromic activation of the CA1 pyramidal neurons (70, 237). The antidromically induced alkalization was far smaller than that evoked by orthodromic trains of identical frequency and duration (70). Large alkaline responses could be elicited only by high-frequency antidromic stimulus trains of 50 –200 Hz (130, 237). Curiously, in experiments using phenol red as an extracellular pH indicator, antidromic stimulation of hippocampal CA1 or dentate neurons did not produce any detectable pHo change Physiol Rev • VOL (177). An insufficient rate or duration of antidromic stimulation may account for this result. B) IS THE ACTIVITY-DEPENDENT PROTON SINK DUE TO THE 2⫹ PLASMALEMMAL CA -ATPASE? A mechanism consistent with a postsynaptic origin was suggested by Schwiening et al. (274) in their studies of surface pH on snail neurons. Calcium entry (elicited by voltage-clamp depolarization) triggered a rise in surface pH that was blocked by vanadate, an inhibitor of the plasmalemmal Ca2⫹-ATPase. This transporter is activated by elevation of cytosolic Ca2⫹ and extrudes this ion in exchange for extracellular H⫹ (61, 62). Accordingly, vanadate also slowed the recovery of the cytosolic Ca2⫹ transients. These experiments demonstrated the ability of the plasmalemmal Ca2⫹-ATPase to elicit a surface alkalosis, but it remained unclear whether this transporter was responsible for the activity-dependent proton sink in the vertebrate CNS. Preliminary support for this hypothesis came from experiments in which alkaline shifts were found to be dependent on external Ca2⫹. In the CA1 region of hippocampal slices, alkalizations evoked by 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 FIG. 2. Extracellular alkalosis generated by proton sinks and bicarbonate sources in rat hippocampal slices. A: alternate microejection of GABA or glutamate elicited interstitial alkaline shifts of comparable amplitude and time course. Superfusion of increasing concentrations of benzolamide caused inhibition of the GABA response (revealing an acid shift) and amplification of the glutamate response, consistent with generation of the alkalosis via an HCO⫺ 3 efflux and a proton sink, respectively. [From Chen and Chesler (71).] B: in a rat hippocampal slice, block of excitatory synaptic transmission (with CNQX and APV) permitted selective, high-frequency stimulation of GABAergic inhibitory interneurons. The resulting rapid alkaline transient was largely blocked in the presence of benzolamide. pHo, extracellular pH. [From Kaila et al. (150).] pH REGULATION IN THE BRAIN Physiol Rev • VOL cytosolic Ca2⫹ load is handled by the plasmalemmal Ca2⫹ATPase, a far lower extracellular buffering capacity would be required in order for Ca2⫹/H⫹ exchange to plausibly account for the extracellular proton sink. C) IS THE ACTIVITY-DEPENDENT PROTON SINK DUE TO A CHAN⫹ NEL-MEDIATED H FLUX? In principle, an extracellular proton sink could be generated by a conductive pathway for H⫹ or OH⫺, since the electrochemical gradients for these ions are inward and outward, respectively (76, 259). Voltagegated proton conductances were first characterized in snail neurons (57, 305). While they have also been described in vertebrate cells, there is no evidence for such a conductance in vertebrate neurons. Moreover, voltagegated H⫹ channels typically activate at membrane potentials close to zero, where the electrochemical gradient for H⫹ is outward (reviewed in Ref. 92). This feature is consistent with their presumed role as a rapid H⫹ extrusion mechanism. Channels that are permeant to other ions may conduct H⫹ (or OH⫺) to a limited degree (92). However, several observations make it doubtful that alkaline shifts arise from nonspecific H⫹ conductances of this sort. A variety of channel blockers do not inhibit alkaline shifts (evoked by either glutamate agonists or electrical stimulation), including tetrodotoxin (73), amiloride (77), Ba2⫹ (130), and stilbene derivatives (215, 338). Ca2⫹ channel antagonists have sometimes inhibited alkaline shifts (130, 237), but in many instances the alkalosis was unaffected or only partially reduced by these agents (84, 282). It has been argued that the low activity of H⫹ and OH⫺ (on the order of 10⫺7 M) is inconsistent with a large, channelmediated flux of these ions, unless the channels involved are present in exceptionally high density, as has been suggested for cells that exhibit specific proton conductances (92). D) BICARBONATE-INDEPENDENT ALKALINE SHIFTS IN RETINA. The forms of alkalosis discussed above are considered to arise from mechanisms that remove acid or add base to the interstitial fluid. In the retina, a different form of interstitial alkalization can occur in response to light. In this tissue, the high metabolic rate in the dark leads to a sustained interstitial acidosis (33, 357), which varies in a circadian rhythm (103, 104). When the photoreceptor dark current is transiently turned off during light stimulation, a slow, transient increase in pHo occurs, owing to a decrease in metabolic rate. These slow alkaline shifts have been noted in frog (33) and cat (357) and are most prominent in the proximal retina adjacent to the rods. 5. Bicarbonate-dependent alkaline shifts The means by which the efflux of HCO⫺ 3 generates an external alkalosis is well established. The ability of CO2 to rapidly equilibrate across cell membranes results in an identical equilibrium potential for H⫹ and HCO⫺ 3 which is 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 ionotropic glutamate agonists (238, 283), or by antidromic stimulation (130, 237), were blocked in media containing 0 Ca2⫹ and EGTA. However, subsequent studies revealed that sizable interstitial alkaline shifts could often be obtained in the nominal absence of external Ca2⫹ following interstitial AMPA ejection (282), or during the onset of spreading depression (206). In addition, when alkaline shifts were reduced or abolished in 0 Ca2⫹, addition of Ba2⫹ restored the responses (130, 282, 312). While Ba2⫹ can substitute for Ca2⫹ in carrying charge across voltagegated Ca2⫹ channels (137), the Ca2⫹-ATPase is not known to utilize Ba2⫹ effectively (181, 363). Additional arguments against the Ca2⫹-ATPase mechanism have come from the use of inhibitors. Cd2⫹ is a potent blocker of the plasmalemmal Ca2⫹-ATPase with a Ki near 2 nM (326, 327). At an external concentration of 100 M, Cd2⫹ did not affect the glutamate-evoked alkaline shift in turtle cerebellum (84), and caused just a partial block of the AMPA-induced alkalosis in rat hippocampal slices (282). Carboxyeosin, with a Ki of 20 nM (115), is a more widely used inhibitor of the plasmalemmal Ca2⫹ATPase (86, 110, 126, 276). Instead of blocking this alkalosis, carboxyeosin augmented the interstitial alkaline shifts elicited by antidromic stimulation in hippocampal slices (133). To date, the reduction of the alkaline shift in zero Ca2⫹ solutions is the best evidence favoring the Ca2⫹ATPase hypothesis. In view of the myriad processes influenced by a Ca2⫹ influx, however, a specific link to this Ca2⫹ transporter cannot yet be claimed. Nevertheless, because the plasmalemmal Ca2⫹-ATPse is an obligatory Ca2⫹/H⫹ exchanger, it must contribute to interstitial alkalosis to some degree. Such a ubiquitous mechanism might account for the small extracellular alkaline transients associated with electrical activity in muscle (116) and isolated nerve (91, 108). In the CNS, however, it remains unclear whether the H⫹ flux contributed by this process is sufficient to account for large, rapid, stimulusevoked alkaline shifts. In this regard, it is notable that in cortical pyramidal cells (193), and cerebellar Purkinje neurons (110), cytosolic Ca2⫹ loads are handled mainly by uptake into endoplasmic reticulum and mitochondria, and the fraction of Ca2⫹ entry extruded by the plasmalemmal Ca2⫹-ATPase is accordingly small. These observations, coupled with quantitative consideration of the extracellular alkalosis, place significant restrictions on the Ca2⫹-ATPase hypothesis. For example, if an activity-dependent fall in [Ca2⫹]o was on the order of 200 M (248a) and the entire Ca2⫹ load was rapidly extruded by the plasmalemmal Ca2⫹-ATPase (with a Ca2⫹/H⫹ exchange ratio of 1:2), then the interstitial space would be subjected to a base load of 400 M. To produce an alkaline shift of 0.05– 0.1 pH units, the extracellular buffering capacity would have to be very small, on the order of 4 – 8 mM. Because only a small fraction of the 1205 1206 MITCHELL CHESLER Physiol Rev • VOL ulus train to the Schaffer collaterals. The speed of this early GABAergic response remains unclear, however. Although the ligand-gated activation of the HCO⫺ 3 conductance would suggest rapid generation of the alkalosis, the rate-limiting step in the generation of this response is the catalyzed dehydration of carbonic acid (Fig. 1B) (71, 150), rather than channel gating. Since the GABAA receptor and the strychnine-sensitive glycine receptor have a similar permeability to HCO⫺ 3 (34), it is plausible that glycinergic inhibitory transmission is also associated with HCO⫺ 3 -dependent alkaline pHo shifts. Accordingly, the glycinergic fall in pHi noted in brain stem medullary neurons (189) might be accompanied by a concomitant external alkalosis. This topic remains largely unexplored, however. In the dorsal horn of neonatal rat spinal cord, Sykova and colleagues (147) noted a small glycine-evoked alkaline shift; however, this response persisted in the HEPES-buffered media, suggesting it was not mediated by an HCO⫺ 3 efflux. 6. Mechanisms of activity-dependent acid shifts A) LATE ACID SHIFTS: EVIDENCE FOR A METABOLIC ORIGIN. Slow, activity-dependent acid shifts were noted in the earliest pH microelectrode studies (65, 322). Kraig et al. (174) made the first attempt to elucidate the cause of these phenomena. In the in vivo rat cerebellum, stimulation of the parallel fibers, or induction of spreading depression, elicited a small, brief alkaline transient, followed by a pronounced acidification lasting minutes. In the presence of acetazolamide, the acid shift was increased in size. This observation suggests that the acidosis was due to extrusion of H⫹ rather than the generation of CO2. A persistent acidification was also induced by elevation of [K⫹]o, which could be abolished by fluoride (a glycolytic inhibitor). A similar acidosis was noted in vivo following spreading depression (174, 215) and could be correlated with an increase in brain lactate (215, 269). Taken together, these observations argue that the metabolic production of acid and its subsequent extrusion were responsible for these interstitial acid shifts. Although brain cells express a lactic acid transporter (219), there is no evidence that this mechanism directly contributes to interstitial acidosis. Indeed, in hippocampal slices, lactate transport inhibitors did not affect the extracellular acidosis induced by hypoxia (180), or by elevated K⫹ (128). Given the ability of lactate to serve as an energy source in brain (272), it has been suggested that lactic acid efflux from glial cells is followed by uptake and utilization by neurons (245a, 271, 345). As such, there may be no net effect on interstitial pH. Assuming that prolonged stimulation, spreading depression, and hypoxia cause a fall in pHi, it seems likely that classic acid extrusion mechanisms would be activated, (e.g., Na⫹/H⫹ exchange, Na⫹-driven Cl⫺/HCO⫺ 3 exchange, or electrogenic 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 typically close to ⫺20 mV in vertebrate neurons and glia (76). Thus, at normal negative membrane potentials, there exists an outward electrochemical gradient for HCO⫺ 3 that can drive the efflux of this base to generate an external alkaline shift. A likely pathway for such an efflux became evident when Bormann et al. (34) demonstrated that GABAA and glycine receptor channels had a significant permeability to HCO⫺ 3 . Subsequently, Kaila and Voipio (154) showed that agonist gating of GABAA receptors on crayfish muscle led to an HCO⫺ 3 -dependent alkalization of the surface membrane. This external alkalosis had the features expected of a pH shift generated by an HCO⫺ 3 efflux as it was blocked when surface CA was inhibited by benzolamide (153) and could be mimicked by formate (196), an anion that can also pass the GABAA receptor channel (34). In the vertebrate CNS, extracellular alkaline shifts generated by GABAA receptors were first noted in turtle cerebellum. Superfusion of GABA or GABAA agonists produced a rapid increase in pHo. This alkalosis was HCO⫺ 3 dependent (68, 69), mimicked by formate (69), and inhibited by benzolamide (125). A similar interstitial alkalosis mediated by GABAA receptor agonists was noted in the CA1 stratum pyramidale of rat hippocampal slices (71, 150). Pressure ejection of GABA resulted in a picrotoxinsensitive, bicarbonate-dependent alkalosis that was blocked by extracellular CA inhibitors (Fig. 2A). A more physiological demonstration was provided by Kaila et al. (150) who showed that selective activation of inhibitory interneurons produced an alkaline shift in stratum pyramidale (Fig. 2B) that was augmented by pentobarbital, blocked by picrotoxin, and inhibited by an extracellular, dextran-conjugated CA inhibitor. These studies indicated that the activation of interneurons can give rise to extracellular alkaline shifts localized around GABAA receptors. The significance of these pHo shifts remains to be determined, and the size and speed of these responses during typical sequences of synaptic excitation and inhibition are still unclear. Kaila and colleagues (295) proposed that with prolonged, repetitive stimulation of the Schaffer collaterals, early fatigue of the GABAergic synaptic transmission will limit the contribution of the HCO⫺ 3 efflux to the net alkalosis. Accordingly, alkaline shifts elicited in this manner were amplified by benzolamide (295), signifying prevalence of the excitation-associated proton sink through most of the stimulus train. However, it was noted that high-frequency, short-duration stimulus trains elicited alkaline shifts that were amplified modestly by benzolamide (when compared with the amplification of alkaline transients evoked by low-frequency trains containing the same number of stimuli). Moreover, the alkalosis produced by sufficiently short, high-frequency trains could be inhibited by benzolamide (295). These data suggested that the GABAergic, HCO⫺ 3 -dependent alkalosis is predominant early in a stim- pH REGULATION IN THE BRAIN anisms. In contrast, in isolated neonatal spinal cord, a prominent early alkalosis occurred, which could be elicited with as little as a single shock to the dorsal root (148). Over the first two postnatal weeks, this response converted to a rapid acid shift that increased in prominence over the next three months (Fig. 3A). A similar developmental shift from stimulus-evoked alkaline to acid shift was noted in rat optic nerve (91) (Fig. 3B) and in the CA3 region of hippocampal slices (356). It was suggested that the alkaline to acid transformation had a glial origin, as the time course of the change roughly paralleled glial cell development (91, 148). The idea that glial cells actively secrete acid in response to neural activity has been proposed on many occasions. The observation of a rapid, activity-dependent rise in the cytosolic pH of astrocytes suggested that these cells extrude acid in response to membrane depolarization (81, 82). This hypothesis was tested using Ba2⫹ to inhibit the astrocyte K⫹ conductance and thereby prevent the depolarization that normally results from elevation of [K⫹]o (17). When rat cortex was superfused with Ba2⫹, the stimulus-evoked depolarization and the cytosolic alkalization of astrocytes were both blocked (82). In the cortical interstitial space, Ba2⫹ unmasked or amplified an early stimulus-evoked alkalosis, consistent with the loss of a rapid acid source (82). In neonatal rat spinal cord, Ba2⫹ had a similar effect. After the developmental shift from stimulus-evoked alkalosis to acidosis (148), addition of Ba2⫹ appeared to inhibit the acidification and unmask the underlying alkaline response (290). Moreover, in the spinal cord of irradiated animals, where astrocyte devel- FIG. 3. Developmental onset of rapid, stimulus-evoked acid shifts. A: repetitive stimulation of the dorsal roots elicited an alkaline-acid sequence of pHo transients in the dorsal horn of 4-day-old rats. In older animals, identical dorsal root stimulation elicited a predominant acid shift. [From Sykova et al. (290), copyright 1992, with permission from Elsevier.] B: repetitive stimulation of optic nerve from a 1-day-old rat elicited an initial, small alkalinization, followed by a modest, slow acid shift. In older animals, pure acid shifts were observed. [From Davis et al. (91), copyright 1987, with permission from Elsevier.] Physiol Rev • VOL 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 Na⫹-HCO⫺ 3 cotransport), which could be responsible for the ensuing extracellular acidification. Inhibition of activity-dependent acidosis by amiloride has implicated Na⫹/H⫹ exchange in this regard (338, 262); however, effects of this drug on excitability (301) could not be ruled out. In hippocampal slices there is strong evidence that CO2 generation, rather than acid extrusion, is responsible for slow activity-dependent acid shifts. Voipio and Kaila (328) demonstrated that repetitive activation of the Schaffer collaterals was associated with a slow rise of tissue PCO2 that fully accounted for a concomitant delayed interstitial acidosis. The late acid shift of spreading depression was similarly attributed to CO2 generation in hippocampal slices (297). The hippocampal acid shifts were reduced by benzolamide (328), consistent with a CO2generated acidosis. These results seem to differ from the observations in vivo, where cerebellar (174) and cortical (215) acid shifts were increased by acetazolamide. The ability of blood flow to rapidly clear CO2 in vivo may partly explain these differences. B) EARLY ACID SHIFTS: EVIDENCE FOR A GLIAL ORIGIN. In some instances, neural activity induces a rapid interstitial acidification that can precede or preclude early alkaline transients. For example, initial acid shifts have sometimes been noted at the onset of spreading depression (215, 314). Sykova and Svoboda (291) reported robust, early acid shifts in the dorsal horn of the adult rat spinal cord following peripheral nerve stimulation. These acidifications were inhibited by stilbenes and amiloride, consistent with the involvement of several acid extrusion mech- 1207 1208 MITCHELL CHESLER evoked alkalosis in this preparation was normally masked by the action of the glial electrogenic Na⫹-HCO⫺ 3 cotransporter. The study of extracellular acid shifts evoked by glial cells has led to further insights into the nature of their Na⫹-HCO⫺ 3 cotransporter. In gliotic slices, and in retinal Muller cells, depolarization-induced external acidification was increased after inhibition of extracellular CA (128, 222; Fig. 5A). This suggested that the mechanism of acidification was not solely dependent on the uptake of HCO⫺ 3 , since an acid shift involving loss of interstitial HCO⫺ 3 alone should have been diminished when external CA was blocked (128) (Fig. 5B). It has been suggested that ⫺ CO2⫺ 3 is the ion species utilized to accomplish net HCO3 ⫹ ⫺ flux by the renal Na -HCO3 cotransporter as well as other HCO⫺ 3 carriers (36, 39, 40, 283a). Such a mechanism could explain the effects of CA inhibitors on the extracellular acid shifts produced by glial cell depolarization. Grichtchenko and Chesler (128) proposed that the uptake of Na⫹ and CO2⫺ by astrocytes (thermodynamically 3 equivalent to 1:2 Na⫹-HCO⫺ 3 cotransport) would acidify the extracellular compartment by driving the dissociation ⫹ of HCO⫺ and replenishing CO2⫺ 3 , thereby producing H 3 (Fig. 5C). The fractional increase in [H⫹]o would be greater than the fractional fall in [HCO⫺ 3 ]o, driving the dehydration of H2CO3, and thereby buffering the increase in H⫹. If external CA were blocked, the rate of buffering would be reduced, leading to a greater acid shift (128). Additional evidence consistent with this idea was provided in a study of the surface pH of oocytes expressing both an Na⫹-HCO⫺ 3 cotransporter and extracellular CA IV. Voltage-clamp depolarization produced a surface acidification that was amplified following inhibition of the CA, consistent with the uptake of carbonate ions (127). It should be noted, however, that the same results FIG. 4. Extracellular pH transients in leech ganglion regulated by glial cell membrane potential (Em). Left traces: afferent stimulation evoked a slow glial depolarization and an extracellular acid transient under current (Im) clamp. Middle traces: afferent stimulation during voltage clamp of glial cell unmasked a rapid extracellular alkaline shift. Right traces: an acid shift was again elicited upon return to afferent stimulation under current clamp. [From Rose and Deitmer (260).] Physiol Rev • VOL 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 opment was prevented, the early stimulus-evoked acid shift did not develop, and addition of Ba2⫹ had no effect on the prevailing early alkalosis (290). Taken together, these observations suggest that the effect of Ba2⫹ on activity-dependent pHo transients is due to the arrest of glial acid secretion. Nonspecific actions of Ba2⫹ cannot be discounted, however, particularly when this ion is superfused over a complex preparation. Use of simpler preparations to study glial acid secretion has reduced some of these concerns. Direct evidence for depolarization-induced acid secretion of astrocytes came from studies of gliotic hippocampal slices in which the CA3 neuronal population was replaced by a dense proliferation of reactive astrocytes (128, 129). K⫹-evoked depolarization of the astrocyte population elicited an intracellular alkaline shift (129), and a corresponding interstitial acidosis (128), that were both blocked in the presence of Ba2⫹. The acidosis was dependent on external Na⫹ and ⫹ HCO⫺ 3 , consistent with the operation of electrogenic Na ⫺ ⫹ HCO3 cotransport. Similar characteristics of K -evoked intracellular and surface pH shifts were noted in studies of retinal Muller cells from mudpuppy retina (222). External acidification was recorded with a novel method in which the fluorescent pH indicator BCECF was fixed to a coverslip and imaged below the cell. The external acidification was localized mainly to the end foot process and was reduced in HCO⫺ 3 -free media or by addition of DIDS. The best evidence that glial cells actively secrete acid (via inward electrogenic Na⫹-HCO⫺ 3 cotransport) has come from studies of the leech, where the extracellular fluid of a ganglion is controlled by a single giant glial cell (Fig. 4). Preventing activity-dependent depolarization of the cell under voltage clamp turned the stimulus-induced pHo shift from an acid to an alkaline response (260). This result convincingly demonstrated that the stimulus- pH REGULATION IN THE BRAIN 1209 would be anticipated in the above studies, if a Na⫹-HCO⫺ 3 pair was taken up in exchange for H⫹, as this would result in similar fractional changes in the interstitial ion concentrations. Thus, while the data of the above studies indicate that HCO⫺ 3 is not the sole acid equivalent carried by the Na⫹-HCO⫺ 3 cotransporter, the results can be explained with or without the transport of CO2⫺ 3 . It has been suggested that glial acid secretion is a means of regulating interstitial pH at the onset of neural activity, and serves to “muffle” activity-dependent alkaline shifts (76, 94, 252). Because activity-evoked alkalosis can arise within 10 ms (124, 177), glial acid secretion might be expected to occur with similar speed. The rate of this process is not clear, however. In spinal cord, where early acid shifts are presumed to have a glial origin, pH microPhysiol Rev • VOL electrodes demonstrated an onset within seconds of stimulation (148, 290, 291). Faster changes could not have been detected, however, as the best response time of pH microelectrodes is on the order of 1 s. Studies of glial acid secretion would accordingly benefit from the use of optical recording methods. C) DO SYNAPTIC VESICLE CONTENTS CONTRIBUTE TO EXTRACELLULAR ACIDIFICATION? Extremely rapid, apparent acid transients were reported in hippocampal slices after stimulation of the Schaffer collaterals (177). These changes, detected by phenol red absorbance, peaked ⬃10 ms after an orthodromic stimulus and were followed by slower transients that corresponded to typical alkaline-acid shifts. It was suggested that such an early acidification could be due to release of the acidic contents of synaptic vesicles 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 FIG. 5. Acid shifts evoked by glial acid secretion modified by inhibition of extracellular carbonic anhydrase. A: extracellular acid shifts elicited by depolarization (elevation of [K⫹]o) in a gliotic hippocampal slice were reversibly amplified by benzolamide. [From Grichtchenko and Chesler (127), copyright 1994, with permission from Elsevier.] B: ⫹ uptake of Na⫹ and two HCO⫺ via the hydration of CO2 (left). Addition of benzolamide would 3 would elicit a rise in H ⫹ be expected to reduce the acid transient (right). C: uptake of Na⫹ and CO2⫺ 3 would elicit a rise in H via the dissociation ⫺ 2⫺ ⫹ ⫹ of HCO3 into CO3 and H . The elevation of H would be buffered by the dehydration of H2CO3, and addition of benzolamide would therefore be expected to increase the acid transient. 1210 MITCHELL CHESLER VIII. EFFECTS OF ACTIVITY-DEPENDENT pH SHIFTS ON NEURAL FUNCTION Processes that could be modulated by activity-dependent pH shifts would be expected to have a steep sensitivity to H⫹ near its physiological range. In this category fall a number of membrane conductances including those associated with NMDA receptors (300, 320, 333), GABAA receptors (243, 279, 299), voltage-gated Ca2⫹ channels (19, 141, 308, 309) and proton-gated channels (169, 178; reviewed in Refs. 22, 307, 319, 337). Although channels may have appropriate sensitivities to changes in pH, in most experiments, pH shifts have been elicited by application of neurotransmitters or by artificial, synchronous activation of thousands of neurons. One may therefore question whether these responses are relevant to the in vivo setting. Synchronous or repetitive neural activity, however, is a normal feature of brain function. In hippocampus, for example, large populations of neurons fire together during the theta rhythm at frequencies close to 10 Hz (reviewed in Ref. 56). Stimulation of hippocampal afferents at comparable rates can elicit extracellular alkaline transients on the order of 0.1– 0.2 pH units (63, 71, 144, 295, 355). While pHo shifts have not been recorded in association with normal hippocampal or cortical rhythms, they are known to accompany aberrant neural activity associated with seizure. In vivo, seizure activity has been associated with pronounced interstitial acid shifts (278, 285). In addition, sizable alkaline shifts have been noted at the onset of Physiol Rev • VOL spontaneous epileptiform bursts in isolated guinea pig brain (93) and hippocampal slices (355). Interstitial pH transients have also been noted in response to physiological stimuli. In the dorsal horn of the rat spinal cord, Sykova and Svoboda (291) reported a prolonged fall in pHo (of 0.05– 0.10 pH units) in response to noxious stimulation of the skin. A. Modulation of Function by Extracellular Alkaline Shifts The speculation that channel function is modulated by activity-dependent alkaline pHo shifts has received some experimental support. Enhanced excitability has long been associated with a rise in pHo, while acidosis has been observed to diminish neural activity (9, 16, 144, 182, 194, 324) (for review, see Ref. 286). NMDA receptors were implicated in these effects when the increased excitability of hippocampal slices induced by alkaline saline was prevented by specific antagonists (9). A likely biophysical basis for this observation was provided when it was discovered that NMDA receptors are blocked by external protons, with a steep sensitivity in the physiological range of extracellular pH (300, 320, 333). To implicate a pHo shift in the modulation of a channel, one can alter the buffering capacity and note whether shifts in the amplitude of the pH change elicit expected changes in channel behavior. This may be done by adding or removing major buffers such as sodium bicarbonate or HEPES; however, this approach necessarily entails large changes in the concentration of these substances, which could have numerous unanticipated effects. Inhibition of the interstitial CA with benzolamide is an attractive alternate means of reducing extracellular buffering power, since only micromolar amounts can markedly amplify HCO⫺ 3 -independent, interstitial alkaline shifts. Similar concentrations of benzolamide can also be used to reduce alkaline transients that are generated by a GABAA receptor-mediated HCO⫺ 3 efflux (153). Because this agent can have this dual effect, it is essential to understand the nature of the pH change one is attempting to modulate. In addition, possible nonspecific effects of CA inhibitors should be considered when conducting such experiments. For example, in neonatal hippocampal CA1 neurons, benzolamide partially inhibited low-threshold Ca2⫹ channels (123), which could diminish excitability. Therefore, to correctly interpret data, it is best that both pHo transients and excitability be monitored in such experiments. Taira et al. (296) first attempted to modify excitability using benzolamide in hippocampal slices. An increase in the amplitude (but no change in time course) of Schaffer collateral-evoked excitatory postsynaptic field potentials was noted, which began 5 min after exposure to benzolamide and progressed over the next 35 min. This enhance- 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 (187, 198, 208). These putative fast acid responses were not altered by acetazolamide, however, and were not otherwise confirmed to be pH changes by manipulation of the buffering power. In a study using a fluorescein-dextran probe to record Schaffer collateral evoked pHo shifts, only an early alkaline shift was detected within this time frame (124), and these fluorescent responses were amplified by addition of the CA inhibitor benzolamide (315). The release of vesicular contents into the synaptic cleft should nonetheless alter the pH in this microdomain. DeVries (99) has recently provided evidence that acid vesicular contents released from mammalian cone photoreceptors produce a significant local pH change that has a feedback inhibitory action on the cone calcium channels. The onset of inward cone calcium currents was interrupted by a transient outward current that decayed with a 1- to 2-ms time constant. This brief suppression of the Ca2⫹ current was correlated with the onset of transmitter release, and the effect was diminished when the extracellular buffering power was augmented with HEPES. These data are the best evidence to date that a physiologically relevant acid transient may occur in the synaptic cleft due to release of acidic vesicular contents. pH REGULATION IN THE BRAIN Physiol Rev • VOL fied by benzolamide, but the propagation of spreading depression was no longer enhanced (313, 314). Although these data suggest that endogenous alkaline shifts can have a neuromodulatory role, several caveats remain. Assuming the effects of benzolamide on synaptic transmission and spreading depression were due to the pH sensitivity of NMDA receptors, it still remains unclear whether smaller alkaline shifts, elicited under normal buffering conditions (i.e., in the absence of benzolamide), would have a significant modulatory effect. In addition, while benzolamide did not appear to directly affect NMDA receptors, alternative actions of this drug on pre- or postsynaptic processes remain possible. Similar putative modulation of other channels by endogenous alkaline shifts has not been demonstrated. Enhancement of voltage-gated Ca2⫹ conductances by an alkaline transient appears plausible, particularly for highthreshold channels (307). The GABAA receptor-evoked alkalosis, being mediated by a channel-gated HCO⫺ 3 efflux, may be particularly suited to influence GABAA receptors in a feedback manner (243). The nature of this modulation is unlikely to be uniform, however, since the pHo dependence of GABAA receptors varies with subunit composition (319). Moreover, because generation of this alkalosis relies on interstitial CA, the localization of this enzyme with respect to GABAA receptors may be a critical factor. B. Modulation of Function by Activity-Dependent Acid Shifts Given the decrease in excitability associated with an imposed acidosis, endogenous interstitial acid shifts could be expected to reduce neural activity. However, unlike an extrinsically imposed acidosis, which is unlocalized and slow, endogenous acid shifts show regional differences and can display a range of onset and decay times. Intracellular acid shifts generated by transmembrane H⫹ fluxes may be rapid and localized to a given cell, whereas metabolic acidosis is likely to affect intracellular and extracellular pH across a wide region. Potential effects of these pH changes are therefore considered according to their time course and location. The fastest extracellular acid shifts may be those speculated to occur following the release of the acidic contents of synaptic vesicles (177). The strongest evidence for such modulation comes from a study of retinal cones in which depolarization-evoked Ca2⫹ currents underwent a brief inhibition, coincident with the expected release of synaptic vesicles from these cells (99). Detection of pH transients of appropriate time course, localized to the synaptic cleft, is still awaited. Acid secretion by the astrocyte Na⫹-HCO⫺ 3 cotransporter would be expected to operate on a somewhat 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 ment was not seen in the presence of APV, indicating the involvement of NMDA receptors. In the presence of CNQX and the absence of external Mg2⫹, benzolamide caused a 20 –30% increase in the amplitude of the NMDA receptor-mediated field potential, which did not occur in the presence of APV and Mg2⫹ (without CNQX). Although the pHo transients were not recorded, and GABAA receptors were not blocked, these observations implicated NMDA receptors in a benzolamide-mediated enhancement of excitability and suggested a buffering-mediated effect of the drug. If it were assumed that benzolamide amplified the stimulus-evoked alkaline transients, the results would be consistent with an associated augmentation of NMDA receptor currents, leading perhaps to an enhancement of AMPA receptor-mediated synaptic field potentials through long-term potentiation. In a subsequent study, the effect of benzolamide on NMDA receptor-mediated synaptic transmission was resolved in conjunction with pHo recordings (122). Excitatory postsynaptic currents (EPSCs) were recorded from CA1 pyramidal neurons under whole cell voltage clamp during single shocks to the Schaffer collaterals, while corresponding interstitial alkaline shifts were monitored with a nearby pH microelectrode. Application of benzolamide (in the presence of picrotoxin) increased the alkaline transients, as expected for a net proton sink (72). Benzolamide caused no significant change in the mean amplitude of the EPSCs. However, the duration of the EPSCs was markedly prolonged, and the onset of this effect coincided with the first observed amplification of the alkaline transient. This prolongation was not seen when benzolamide was added in the presence of APV (although the alkaline transients were still amplified), indicating that the effect was mediated via NMDA receptors. Because neither the prolongation of the synaptic response nor the increase in the alkaline transient occurred in HEPES-buffered media, the synaptic enhancement was not due to a direct action of benzolamide on these receptors. It is notable that when optical recordings of Schaffer collateral-evoked alkaline shifts were made under comparable conditions (with picrotoxin and benzolamide), the alkalosis peaked in 20 –200 ms (124). These synaptically evoked alkaline shifts therefore appear welltimed to modulate the decay of the NMDA receptor-mediated synaptic responses, which occur within a similar period (136). Augmentation of NMDA receptors by benzolamide has also been demonstrated in a pathological setting. Spreading depression is associated with a large alkaline transient that is increased by CA inhibitors (174, 215). In hippocampal slices, the latency and propagation velocity of spreading depression were increased by benzolamide, in conjunction with an amplification of the initial alkalosis. When APV was present, the alkalosis was still ampli- 1211 1212 MITCHELL CHESLER Physiol Rev • VOL for the antiepileptic action of this drug (184). However, acetazolamide may have complex and diverse effects on baseline pH in the brain as it has been reported to alkalinize the interstitial space in vivo (174), acidify the interstitial space in brain slices (71, 338), and alkalinize the cytoplasm of thalamic neurons (214). Although acidosis has typically been linked to a fall in excitability, under particular circumstances, a fall in pH may have the opposite effect. Velisek et al. (325) noted that microinfusions with fluid of pH 6.7 had proconvulsant effects in the posterior substantia nigra pars reticulata of rats, but no such effect in the anterior part of the nucleus. Differentiation between extracellular and intracellular sites of action could be difficult in such cases, since application of an acidic fluid is likely to reduce both pHo and pHi. An intracellular acidosis could increase excitability if membrane resistance were increased. This was the case in crayfish muscle where Moody (210) was able to attribute acid-induced hyperexcitability to block of an inward rectifying K⫹ conductance. Augmentation of inward currents by acidosis can also occur in the case of proton-gated channels. As these responses inactivate with sustained acidosis, extracellular acid shifts of sufficient size and speed would be required to increase excitability through this mechanism (22, 179, 337). IX. CONCLUSION A great deal has been learned about the regulation of pHi in cells derived from the vertebrate CNS. The majority of studies have been performed in tissue culture. From the experimental perspective, culture systems are practical, and doubtless will continue to provide important clues to transporter function, particularly as studies focus on molecular sites responsible for the modulation of transporter activity. Issues concerning the biological relevance of culture systems should not be overlooked, however. The degree to which a cultured neuron or glial cell will express its in situ complement of transporters is not always known. Studies using brain slices or acutely dissociated cells, while more technically challenging, have appeal in this regard. Many of the mechanisms responsible for activitydependent modulation of brain pH remain unclear. The study of intracellular pH shifts can benefit from in vitro studies in tissue culture where individual cells are amenable to manipulation. The development of better pharmacological tools to more selectively target H⫹ and Ca2⫹ transport mechanisms would be beneficial for such studies. The adequate stimulus for the generation of a pHi shift in culture may not be biologically relevant, however. In situ studies may be particularly useful in determining how changes in neural activity modulate acid transport systems and how normal shifts in the composition of the 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 slower time scale, commensurate with the time course of membrane depolarization. The onset of external acidification may also rely on the activity of interstitial CA (128). It has often been speculated that this process serves to mask or supercede endogenous, activity-dependent alkaline shifts. Because extracellular alkalization may be expected to increase neural activity, muffling of the alkalosis by glial acid secretion may serve to stabilize neuronal excitability (76, 94, 252) (see sect. VIIB6). Rapid intracellular acidifications, beginning within a second of repetitive spike activity, have recently been reported in somata and dendrites of cerebellar Purkinje cells (349). In addition to their speed, these phenomenon are notable because they can be generated by the repetitive activity of a single neuron and are therefore likely to occur in vivo. The basis of these phenomena remains to be established. Given the inhibitory effect of internal acid on Ca2⫹ channel activity (307), and the preponderance of dendritic Ca2⫹ spikes in Purkinje neurons (302), one may speculate that these cytosolic acid transients act to diminish dendritic electroresponsiveness. Delayed interstitial acid shifts can last for minutes following persistent, synchronous neural activity. These phenomena are the likely result of respiratory by-products such as CO2 and lactic acid (see sect. VIIB6). The association of such responses with in vivo seizure activity (278), and the fall in excitability often associated with external acidosis (9, 16, 144, 182, 194), leads to the obvious speculation that epileptiform events become inhibited, and are possibly terminated, as a result of accruing interstitial acid. Thus, in hippocampal slices, low Mg2⫹induced seizures in hippocampal slices were suppressed by superfusion with Ringer solution of low pH (324). NMDA receptors are an obvious focus of interest in this context, given their inhibition by external H⫹ (300, 320, 333). However, it should be noted that if these forms of acidosis are metabolic in origin, they would be associated with a fall in intracellular as well as interstitial pH. In addition, most experimental manipulations that sought to reduce pHo would be likely to acidify the intracellular compartment as well. Concomitant intracellular acidosis may be expected to limit excitability through actions on a variety of channels (211, 288, 298, 307). It can therefore be difficult to pinpoint the manner in which an observed or imposed extracellular acidosis affects excitability. A number of studies have provided evidence that a fall in intracellular, rather than extracellular, pH is important in arresting seizure activity. Use of weak acids to lower pHi of hippocampal slices was able to suppress epileptiform events (29, 354), and the spontaneous termination of seizure was correlated with intracellular acidosis (354). The use of acetazolamide as an antiepileptic agent may be similarly related to its effects on brain pH. A fall in pHi in hippocampal neurons was noted after application of acetazolamide and was proposed as a basis pH REGULATION IN THE BRAIN Address for reprint requests and other correspondence: M. Chesler, Dept. of Physiology & Neuroscience, NYU School of Medicine, 550 First Ave., New York, NY 10016 (E-mail: [email protected]). REFERENCES 1. Ahmed Z and Connor JA. Intracellular pH changes induced by calcium influx during electrical activity in Molluscan neurons. J Gen Physiol 75: 403– 426, 1980. 2. Aldskogius H, Arvidsson J, and Hansson P. Carbonic anhydrase enzyme histochemistry of cranial nerve primary sensory afferent neurons in the rat. Histochemistry 88: 151–154, 1988. 3. Amato A, Ballerini L, and Attwell D. Intracellular pH changes produced by glutamate uptake in rat hippocampal slices. J Neurophysiol 72: 1686 –1696, 1994. 4. Ammann D, Lanter F, Steiner RA, Schulthess P, Shijo Y, and Simon W. Neutral carrier based hydrogen ion selective microelectrode for extra- and intracellular studies. Anal Chem 53: 2267–2269, 1981. 5. Amos BJ and Chesler M. Characterization of an intracellular alkaline shift in rat astrocytes triggered by metabotropic glutamate receptors. J Neurophysiol 79: 695–703, 1998. 6. Amos BJ, Mathie A, and Richards CD. Activation of group I metabotropic glutamate receptors elicits pH changes in cultured rat cortical glia and neurons. Neuroscience 86: 1109 –1120, 1998. 7. Amos BJ, Pocock G, and Richards CD. On the role of bicarbonate as a hydrogen ion buffer in rat CNS neurones. Exp Physiol 81: 623– 632, 1996. 8. Amos BJ and Richards CD. Intrinsic hydrogen ion buffering in rat CNS neurones maintained in culture. Exp Physiol 81: 261–271, 1996. Physiol Rev • VOL 9. Aram JA and Lodge D. Epileptiform activity induced by alkalosis in rat neocortical slices: block by antagonists of N-methyl-D-aspartate. Neurosci Lett 83: 345–350, 1987. 10. Aronson PS, Suhm MA, and Nee J. Interaction of external H⫹ with the Na⫹/H⫹ exchanger in renal microvillus membrane vesicles. J Biol Chem 258: 6767– 6771, 1983. 11. Astion ML, Chvatal A, and Orkand RK. Na⫹/H⫹ exchange in glial cells of Necturus optic nerve. Neurosci Lett 107: 167–172, 1989. 12. Astion ML, Chvatal A, and Orkand RK. Further studies of electrogenic Na⫹/HCO⫺ 3 cotransport in glial cells of Necturus optic nerve: regulation of pHi. Glia 4: 461– 468, 1991. 13. Astion ML, Obaid AL, and Orkand RK. Effects of barium and bicarbonate on glial cells of Necturus optic nerve. Studies with microelectrodes and voltage-sensitive dyes. J Gen Physiol 93: 731– 744, 1989. 14. Attaphitaya S, Park K, and Melvin JE. Molecular cloning and functional expression of a rat Na⫹/H⫹ exchanger (NHE5) highly expressed in brain. J Biol Chem 274: 4383– 4388, 1999. 15. Baird NR, Orlowski J, Szabo EZ, Zaun HC, Schultheis PJ, Menon AG, and Shull GE. Molecular cloning, genomic organization, and functional expression of Na⫹/H⫹ exchanger isoform 5 (NHE5) from human brain. J Biol Chem 274: 4377– 4382, 1999. 16. Balestrino M and Somjen GG. Concentration of carbon dioxide, interstitial pH and synaptic transmission in hippocampal formation of the rat. J Physiol 396: 247–266, 1988. 17. Ballanyi K, Grafe P, and ten Bruggencate G. Ion activities and potassium uptake mechanisms of glial cells in guinea-pig olfactory cortex slices. J Physiol 382: 159 –174, 1987. 18. Ballanyi K, Muckenhoff K, Bellingham MC, Okada Y, Scheid P, and Richter DW. Activity-related pH changes in respiratory neurones and glial cells of cats. Neuroreport 6: 33–36, 1994. 19. Barnes S and Bui Q. Modulation of calcium-activated chloride current via pH-induced changes of calcium channel properties in cone photoreceptors. J Neurosci 11: 4015– 4023, 1991. 20. Baxter KA and Church J. Characterization of acid extrusion mechanisms in cultured fetal rat hippocampal neurones. J Physiol 493: 457– 470, 1996. 21. Benos DJ and Sapirstein VS. Characteristics of an amiloridesensitive sodium entry pathway in cultured rodent glial and neuroblastoma cells. J Cell Physiol 116: 213–220, 1983. 22. Bevan S. Proton-gated ion channels in neurons. In: pH and Brain Function, edited by Kaila K and Ransom BR. New York: Wiley-Liss, 1998. 23. Bevensee MO, Apkon M, and Boron WF. Intracellular pH regulation in cultured astrocytes from rat hippocampus. II. Electrogenic Na/HCO3 cotransport. J Gen Physiol 110: 467– 483, 1997. 24. Bevensee MO, Cummins TR, Haddad GG, Boron WF, and Boyarsky G. pH regulation in single CA1 neurons acutely isolated from the hippocampi of immature and mature rats. J Physiol 494: 315–328, 1996. 25. Bevensee MO, Schmitt BM, Choi I, Romero MF, and Boron WF. An electrogenic Na⫹-HCO⫺ 3 cotransporter (NBC) with a novel COOH-terminus, cloned from rat brain. Am J Physiol Cell Physiol 278: C1200 –C1211, 2000. 26. Bevensee MO, Weed RA, and Boron WF. Intracellular pH regulation in cultured astrocytes from rat hippocampus. I. Role of HCO3. J Gen Physiol 110: 453– 465, 1997. 27. Bjorklund H, Bignami A, and Dahl D. Immunohistochemical demonstration of glial fibrillary acidic protein in normal rat Muller glia and retinal astrocytes. Neurosci Lett 54: 363–368, 1985. 28. Bondarenko A and Chesler M. Rapid astrocyte death induced by transient hypoxia, acidosis, and extracellular ion shifts. Glia 34: 134 –142, 2001. 29. Bonnet U, Bingmann D, and Wiemann M. Intracellular pH modulates spontaneous and epileptiform bioelectric activity of hippocampal CA3-neurones. Eur Neuropsychopharmacology 10: 97– 103, 2000. 30. Bookstein C, Musch MW, DePaoli A, Xie Y, Rabenau K, Villereal M, Rao MC, and Chang EB. Characterization of the rat Na⫹/H⫹ exchanger isoform NHE4 and localization in rat hippocampus. Am J Physiol Cell Physiol 271: C1629 –C1638, 1996. 31. Bookstein C, Musch MW, DePaoli A, Xie Y, Villereal M, Rao 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 interstitial fluid impact on pH regulation. In this respect, both acute brain slices and organotypic culture systems have appeal, as these preparations maintain an intact interstitial space. In addition, activity-dependent changes in pH may be most relevant when studied in a preparation in which the activation and monitoring of discrete afferent pathways is feasible. Interstitial pH shifts, by definition, require the use of an intact tissue. In some instances, surface pH transients may be studied on individual cells. However, the geometric constraints and particular buffering properties of the intact interstitium are needed to understand how these phenomena arise and are regulated in situ. Studies of interstitial pH shifts have benefited greatly from the advent of liquid membrane pH microelectrodes (4). These devices provide an outstanding signal-to-noise ratio with a response time as fast as 1–2 s. Nonetheless, it is increasingly clear that faster recording methods are required to fully understand the processes that govern interstitial pH. A few studies using optical probes have indicated that activity-dependent pHo shifts can occur within milliseconds. Since these pH transients might be generated in discrete locales, improvements in both spatial and temporal resolution would be beneficial. Further identification and localization of interstitial CA would also add important insights into the microphysiology of brain pH, as the proximity and concentration of this enzyme could be the principal determinant of whether or how a pHo transient is generated. 1213 1214 32. 33. 34. 35. 36. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. MC, and Chang EB. A unique sodium-hydrogen exchange isoform (NHE-4) of the inner medulla of the rat kidney is induced by hyperosmolarity. J Biol Chem 269: 29704 –29709, 1994. Bordey A and Sontheimer H. Postnatal development of ionic currents in rat hippocampal astrocytes in situ. J Neurophysiol 78: 461– 477, 1997. Borgula GA, Karwoski CJ, and Steinberg RH. Light-evoked changes in extracellular pH in frog retina. Vis Res 29: 1069 –1077, 1989. Bormann J, Hamill OP, and Sakmann B. Mechanism of anion permeation through channels gated by glycine and gamma-aminobutyric acid in mouse cultured spinal neurones. J Physiol 385: 243–286, 1987. Boron WF. Intracellular pH transients in giant barnacle muscle fibers. Am J Physiol Cell Physiol 233: C61–C73, 1977. Boron WF. Intracellular pH-regulating mechanism of the squid axon: relation between the external Na⫹ and HCO⫺ 3 dependences. J Gen Physiol 85: 325–345, 1985. Boron WF and Boulpaep EL. Intracellular pH regulation in the renal proximal tubule of the salamander: basolateral HCO⫺ 3 transport. J Gen Physiol 81: 53–94, 1983. Boron WF and De Weer P. Intracellular pH transients in squid giant axons caused by CO2, NH3, and metabolic inhibitors. J Gen Physiol 67: 91–112, 1976. Boron WF and Knakal RC. Intracellular pH-regulating mechanism of the squid axon. Interaction between DNDS and extracellular Na⫹ and HCO⫺ 3 . J Gen Physiol 93: 123–150, 1989. Boron WF and Knakal RC. Na⫹-dependent Cl-HCO3 exchange in the squid axon. Dependence on extracellular pH. J Gen Physiol 99: 817– 837, 1992. Boussouf A and Gaillard S. Intracellular pH changes during oligodendrocyte differentiation in primary culture. J Neurosci Res 59: 731–739, 2000. Boussouf A, Lambert R, and Gaillard S. Voltage-dependent ⫹ ⫹ Na⫹-HCO⫺ exchanger are involved in 3 cotransporter and Na /H intracellular pH regulation of cultured mature rat cerebellar oligodendrocytes. Glia 19: 74 – 84, 1997. Bouvier M, Szatkowski M, Amato A, and Attwell D. The glial cell glutamate uptake carrier countertransports pH-changing anions. Nature 360: 471– 474, 1992. Boyarsky G, Ransom B, Schlue WR, Davis MB, and Boron WF. Intracellular pH regulation in single cultured astrocytes from rat forebrain. Glia 8: 241–248, 1993. Boyarsky G, Ransom BR, Schlue WR, Davis M, and Boron WF. Anomalous stimulation of pH regulation by ethylisopropyl amiloride in cultured mammalian astrocytes. Soc Neurosci Abstr 15: 15, 1989. Brechenmacher C and Rodeau JL. Intracellular pH regulation in ventral horn neurones cultured from embryonic rat spinal cord. Mol Membr Biol 17: 101–108, 2000. Brett CL, Wei Y, Donowitz M, and Rao R. Human Na(⫹)/H(⫹) exchanger isoform 6 is found in recycling endosomes of cells, not in mitochondria. Am J Physiol Cell Physiol 282: C1031–C1041, 2002. Brookes N. Intracellular pH as a regulatory signal in astrocyte metabolism. Glia 21: 64 –73, 1997. Brookes N and Turner RJ. Extracellular potassium regulates the glutamine content of astrocytes: mediation by intracellular pH. Neurosci Lett 160: 73–76, 1993. Brookes N and Turner RJ. K⫹-induced alkalinization in mouse cerebral astrocytes mediated by reversal of electrogenic Na⫹HCO⫺ 3 cotransport. Am J Physiol 267: C1633–C1640, 1994. Brown HM and Meech RW. Light induced changes of internal pH in a barnacle photoreceptor and the effect of internal pH on the receptor potential. J Physiol 297: 73–93, 1979. Brune T and Deitmer JW. Intracellular acidification and Ca2⫹ transients in cultured rat cerebellar astrocytes evoked by glutamate agonists and noradrenaline. Glia 14: 153–161, 1995. Brune T, Fetzer S, Backus KH, and Deitmer JW. Evidence for electrogenic sodium-bicarbonate cotransport in cultured rat cerebellar astrocytes. Pflügers Arch 429: 64 –71, 1994. Burckhardt G, Di Sole F, and Helmle-Kolb C. The Na⫹/H⫹ exchanger gene family. J Nephrol 15 Suppl 5: S3–S21, 2002. Physiol Rev • VOL 55. Burton RF. Intracellular buffering. Respir Physiol 33: 51–58, 1978. 56. Buzsaki G. Theta oscillations in the hippocampus. Neuron 33: 325–340, 2002. 57. Byerly L, Meech RW, and Moody W Jr. Rapidly activating hydrogen ion currents in perfused neurones of the snail, Lymnaea stagnalis. J Physiol 351: 199 –216, 1984. 58. Cammer W. Carbonic anhydrase in oligodendrocytes and myelin in the central nervous system. Ann NY Acad Sci 429: 494 – 497, 1984. 59. Cammer W and Tansey FA. Carbonic anhydrase immunostaining in astrocytes in the rat cerebral cortex. J Neurochem 50: 319 –322, 1988. 60. Canzoniero LM, Sensi SL, and Choi DW. Recovery from NMDAinduced intracellular acidification is delayed and dependent on extracellular bicarbonate. Am J Physiol Cell Physiol 270: C593– C599, 1996. 61. Carafoli E. The calcium pumping ATPase of the plasma membrane. Annu Rev Physiol 53: 531–547, 1991. 62. Carafoli E and Stauffer T. The plasma membrane calcium pump: functional domains, regulation of the activity, and tissue specificity of isoform expression. J Neurobiol 25: 312–324, 1994. 63. Carlini WG and Ransom BR. Regional variation in stimulated extracellular pH transients in the mammalian CNS. Soc Neurosci Abstr 12: 452, 1986. 64. Carter ND, Fryer A, Grant AG, Hume R, Strange RG, and Wistrand PJ. Membrane specific carbonic anhydrase (CAIV) expression in human tissues. Biochim Biophys Acta 1026: 113–116, 1990. 65. Caspers H and Speckmann EJ. Cerebral PO2, PCO2 and pH: changes during convulsive activity and their significance for spontaneous arrest of seizures. Epilepsia 13: 699 –725, 1972. 66. Chen G, Dunbar RL, Gao W, and Ebner TJ. Role of calcium, glutamate neurotransmission, and nitric oxide in spreading acidification and depression in the cerebellar cortex. J Neurosci 21: 9877–9887, 2001. 67. Chen G, Hanson CL, Dunbar RL, and Ebner TJ. Novel form of spreading acidification and depression in the cerebellar cortex demonstrated by neutral red optical imaging. J Neurophysiol 81: 1992–1998, 1999. 68. Chen JC and Chesler M. A bicarbonate-dependent increase in extracellular pH mediated by GABAA receptors in turtle cerebellum. Neurosci Lett 116: 130 –135, 1990. 69. Chen JC and Chesler M. Extracellular alkalinization evoked by GABA and its relationship to activity-dependent pH shifts in turtle cerebellum. J Physiol 442: 431– 446, 1991. 70. Chen JC and Chesler M. Extracellular alkaline shifts in rat hippocampal slice are mediated by NMDA and non-NMDA receptors. J Neurophysiol 68: 342–344, 1992. 71. Chen JC and Chesler M. Modulation of extracellular pH by glutamate and GABA in rat hippocampal slices. J Neurophysiol 67: 29 –36, 1992. 72. Chen JC and Chesler M. pH transients evoked by excitatory synaptic transmission are increased by inhibition of extracellular carbonic anhydrase. Proc Natl Acad Sci USA 89: 7786 –7790, 1992. 73. Chen JCT and Chesler M. NMDA and non-NMDA receptors mediate extracellular alkalinizations in the rat hippocampal slice. Soc Neurosci Abstr 18: 1343, 1992. 74. Chesler M. pH Regulation in Vertebrate Neurons (PhD thesis). New York: New York University, 1985. 75. Chesler M. Regulation of intracellular pH in reticulospinal neurones of the lamprey, Petromyzon marinus. J Physiol 381: 241– 261, 1986. 76. Chesler M. The regulation and modulation of pH in the nervous system. Prog Neurobiol 34: 401– 427, 1990. 77. Chesler M and Chan CY. Stimulus-induced extracellular pH transients in the in vitro turtle cerebellum. Neuroscience 27: 941–948, 1988. 78. Chesler M, Chen JC, and Kraig RP. Determination of extracellular bicarbonate and carbon dioxide concentrations in brain slices using carbonate and pH-selective microelectrodes. J Neurosci Methods 53: 129 –136, 1994. 79. Chesler M and Chen JCT. Evoked shifts in extracellular pH, 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 37. MITCHELL CHESLER pH REGULATION IN THE BRAIN 80. 81. 82. 83. 84. 85. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. Physiol Rev • VOL D, Boron WF, and Haddad GG. Sodium-hydrogen exchangers and sodium-bicarbonate co-transporters: ontogeny of protein expression in the rat brain. Neuroscience 102: 217–228, 2001. 106. Droz B and Kazimierczak J. Carbonic anhydrase in primary sensory neurons of dorsal root ganglia. Comp Biochem Physiol B Biochem 88: 713–717, 1987. 107. Endres W, Ballanyi K, Serve G, and Grafe P. Excitatory amino acids and intracellular pH in motoneurons of the isolated frog spinal cord. Neurosci Lett 72: 54 –58, 1986. 108. Endres W, Grafe P, Bostock H, and ten Bruggencate G. Changes in extracellular pH during electrical stimulation of isolated rat vagus nerve. Neurosci Lett 64: 201–205, 1986. 109. Fencl V. Acid-base balance in cerebral fluids. In: Handbook of Physiology. The Respiratory System. Control of Breathing. Bethesda, MD: Am Physiol Soc, 1986, sect. 3, vol. II, pt. 1, chapt. 4, p. 115–162. 110. Fierro L, DiPolo R, and Llano II. Intracellular calcium clearance in Purkinje cell somata from rat cerebellar slices. J Physiol 510: 499 –512, 1998. 110a.Filosa JA, Dean JB, and Putnam RW. Role of intracellular and extracellular pH in the chemosensitive response of rat locus coeruleus neurones. J Physiol 541: 493–509, 2002. 111. Flogel U, Willker W, and Leibfritz D. Regulation of intracellular pH in neuronal and glial tumour cells, studied by multinuclear NMR spectroscopy. NMR Biomed 7: 157–166, 1994. 112. Fujikawa-Adachi K, Nishimori I, Taguchi T, and Onishi S. Human carbonic anhydrase XIV (CA14): cDNA cloning, mRNA expression, and mapping to chromosome 1. Genomics 61: 74 – 81, 1999. 113. Gaillard S and Dupont JL. Ionic control of intracellular pH in rat cerebellar purkinje cells maintained in culture. J Physiol 425: 71– 83, 1990. 114. Gaillard S, Dupont JL, Bossu JL, and Feltz A. Characteristics of calcium currents and of pH regulation in Purkinje cells maintained in primary culture. Acta Physiol Scand Suppl 582: 38, 1989. 115. Gatto C and Milanick MA. Inhibition of the red blood cell calcium pump by eosin and other fluorescein analogues. Am J Physiol Cell Physiol 264: C1577–C1586, 1993. 116. Gebert G and Friedman SM. An implantable glass electrode used for pH measurement in working skeletal muscle. J Appl Physiol 34: 122–124, 1973. 117. Geers C, Gros G, and Gartner A. Extracellular carbonic anhydrase of skeletal muscle associated with the sarcolemma. J Appl Physiol 59: 548 –558, 1985. 118. Ghandour MS, Langley OK, Zhu XL, Waheed A, and Sly WS. Carbonic anhydrase IV on brain capillary endothelial cells: a marker associated with the blood brain barrier. Proc Natl Acad Sci USA 89: 6823– 6827, 1992. 119. Giacobini E. A cytochemical study of the localization of carbonic anhydrase in the nervous system. J Neurochem 9: 169 –177, 1962. 120. Giffard RG, Papadopoulos MC, van Hooft JA, Xu L, Giuffrida R, and Monyer H. The electrogenic sodium bicarbonate cotransporter: developmental expression in rat brain and possible role in acid vulnerability. J Neurosci 20: 1001–1008, 2000. 121. Gottfried JA and Chesler M. Extracellular carbonic anhydrase activity in turtle cerebellum. Soc Neurosci Abstr 19: 267, 1993. 122. Gottfried JA and Chesler M. Endogenous H⫹ modulation of NMDA receptor-mediated EPSCs revealed by carbonic anhydrase inhibition in rat hippocampus. J Physiol 478: 373–378, 1994. 123. Gottfried JA and Chesler M. Benzolamide inhibits low-threshold calcium currents in hippocampal pyramidal neurons. J Neurophysiol 74: 2774 –2777, 1995. 124. Gottfried JA and Chesler M. Temporal resolution of activitydependent pH shifts in rat hippocampal slices. J Neurophysiol 76: 2804 –2807, 1996. 125. Gottried J and Chesler M. Extracellular carbonic anhydrase in turtle cerebellum. Soc Neurosci Abstr 19: 267, 1993. 126. Green AK, Cobbold PH, and Dixon CJ. Effects on the hepatocyte [Ca2⫹]i oscillator of inhibition of the plasma membrane Ca2⫹ pump by carboxyeosin or glucagon-(19 –29). Cell Calcium 22: 99 – 109, 1997. 127. Grichtchenko II and Boron WF. Surface-pH measurements in 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 86. bicarbonate and CO2 in rat hippocampal slices. Soc Neurosci Abstr 19: 1522, 1993. Chesler M and Kaila K. Modulation of pH by neuronal activity. Trends Neurosci 15: 396 – 402, 1992. Chesler M and Kraig RP. Intracellular pH of astrocytes increases rapidly with cortical stimulation. Am J Physiol 253: R666 –R670, 1987. Chesler M and Kraig RP. Intracellular pH transients of mammalian astrocytes. J Neurosci 9: 2011–2019, 1989. Chesler M and Nicholson C. Regulation of intracellular pH in vertebrate central neurons. Brain Res 325: 313–316, 1985. Chesler M and Rice ME. Extracellular alkaline-acid pH shifts evoked by iontophoresis of glutamate and aspartate in turtle cerebellum. Neuroscience 41: 257–267, 1991. Chiesi M and Inesi G. Adenosine 5⬘-triphosphate dependent fluxes of manganese and hydrogen ions in sarcoplasmic reticulum vesicles. Biochemistry 19: 2912–2918, 1980. Choi HS and Eisner DA. The role of sarcolemmal Ca2⫹-ATPase in the regulation of resting calcium concentration in rat ventricular myocytes. J Physiol 515: 109 –118, 1999. Chow SY, Yen-Chow YC, White HS, and Woodbury DM. pH regulation after acid load in primary cultures of mouse astrocytes. Brain Res 60: 69 –78, 1991. Connor JA and Hockberger P. Intracellular pH changes induced by injection of cyclic nucleotides into gastropod neurones. J Physiol 354: 163–172, 1984. Cox GA, Lutz CM, Yang CL, Biemesderfer D, Bronson RT, Fu A, Aronson PS, Noebels JL, and Frankel WN. Sodium/hydrogen exchanger gene defect in slow-wave epilepsy mutant mice. Cell 91: 139 –148, 1997. Cragg P, Patterson L, and Purves MJ. The pH of brain extracellular fluid in the cat. J Physiol 272: 137–166, 1977. Davis PK, Carlini WG, Ransom BR, Black JA, and Waxman SG. Carbonic anhydrase activity develops postnatally in the rat optic nerve. Brain Res 428: 291–298, 1987. DeCoursey TE and Cherny VV. Voltage-activated hydrogen ion currents. J Membr Biol 141: 203–223, 1994. De Curtis M, Manfridi A, and Biella G. Activity-dependent pH shifts and periodic recurrence of spontaneous interictal spikes in a model of focal epileptogenesis. J Neurosci 18: 7543–7551, 1998. Deitmer JW and Rose CR. pH regulation and proton signalling by glial cells. Prog Neurobiol 48: 73–103, 1996. Deitmer JW and Schlue WR. The regulation of intracellular pH by identified glial cells and neurones in the central nervous system of the leech. J Physiol 388: 261–283, 1987. Deitmer JW and Schlue WR. An inwardly directed electrogenic sodium-bicarbonate co-transport in leech glial cells. J Physiol 411: 179 –194, 1989. Deitmer JW and Schneider HP. Acid/base transport across the leech giant glial cell membrane at low external bicarbonate concentration. J Physiol 512: 459 – 469, 1998. Deitmer JW and Szatkowski M. Membrane potential dependence of intracellular pH regulation by identified glial cells in the leech central nervous system. J Physiol 421: 617– 631, 1990. DeVries SH. Exocytosed protons feedback to suppress the Ca2⫹ current in mammalian cone photoreceptors. Neuron 32: 1107–1117, 2001. Dickens CJ, Gillespie JI, and Greenwell JR. Interactions between intracellular pH and calcium in single mouse neuroblastoma (N2A) and rat pheochromocytoma cells (PC12). Q J Exp Physiol 74: 671– 679, 1989. Dietzel I, Heinemann U, Hofmeier G, and Lux HD. Transient changes in the size of the extracellular space in the sensorimotor cortex of cats in relation to stimulus-induced changes in potassium concentration. Exp Brain Res 40: 432– 439, 1980. Dixon DB, Takahashi K, and Copenhagen DR. L-Glutamate suppresses HVA calcium current in catfish horizontal cells by raising intracellular proton concentration. Neuron 11: 267–277, 1993. Dmitriev AV and Mangel SC. A circadian clock regulates the pH of the fish retina. J Physiol 522: 77– 82, 2000. Dmitriev AV and Mangel SC. Circadian clock regulation of pH in the rabbit retina. J Neurosci 21: 2897–2902, 2001. Douglas RM, Schmitt BM, Xia Y, Bevensee MO, Biemesderfer 1215 1216 MITCHELL CHESLER Physiol Rev • VOL 149. Kaila K and Chesler M. Activity-evoked changes in extracellular pH. In: pH and Brain Function, edited by Kaila K and Ransom B. New York: Wiley, 1998. 150. Kaila K, Paalasmaa P, Taira T, and Voipio J. pH transients due to monosynaptic activation of GABA A receptors in rat hippocampal slices. Neuroreport 3: 105–108, 1992. 151. Kaila K, Panula P, Karhunen T, and Heinonen E. Fall in intracellular pH mediated by GABAA receptors in cultured rat astrocytes. Neurosci Lett 126: 9 –12, 1991. 152. Kaila K and Ransom BR. pH and Brain Function. New York: Wiley-Liss, 1998. 153. Kaila K, Saarikoski J, and Voipio J. Mechanism of action of GABA on intracellular pH and on surface pH in crayfish muscle fibres. J Physiol 427: 241–260, 1990. 154. Kaila K and Voipio J. Postsynaptic fall in intracellular pH induced by GABA-activated bicarbonate conductance. Nature 330: 163–165, 1987. 155. Kalamkarov G, Pogozheva I, Shevchenko T, Koskelainen A, Hemila S, and Donner K. pH changes in frog rods upon manipulation of putative pH-regulating transport mechanisms. Vision Res 36: 3029 –3036, 1996. 156. Karlmark B, Agerup B, and Wistrand PJ. Renal proximal tubular acidification. Role of brush-border and cytoplasmic carbonic anhydrase. Acta Physiol Scand 106: 145–150, 1979. 157. Karmann U and Held DR. Equations treating the pH and HCO⫺ 3 of buffered media as functions of PCO2. Respir Physiol 15: 343–349, 1972. 158. Kato K. Sequence of a novel carbonic anhydrase-related polypeptide and its exclusive presence in Purkinje cells. FEBS Lett 271: 137–140, 1990. 159. Katsura K, Mellergard P, Theander S, Ouyang YB, and Siesjo BK. Buffer capacity of rat cortical tissue as well as of cultured neurons and astrocytes. Brain Res 618: 283–294, 1993. 160. Katz BJ and Oakley B II. Evidence for Na⫹/H⫹ exchange in vertebrate rod photoreceptors. Exp Eye Res 51: 199 –207, 1990. 161. Kelly JP and Van Essen DC. Cell structure and function in the visual cortex of the cat. J Physiol 238: 515–547, 1974. 161a.Kempski O, von Rosen S, Weigt H, Staub F, Peters J, and Baethmann A. Glial ion transport and volume control. Ann NY Acad Sci 633: 306 –317, 1991. 162. Kettenmann H and Schlue WR. Intracellular pH regulation in cultured mouse oligodendrocytes. J Physiol 406: 147–162, 1988. 163. Kimelberg HK, Biddlecome S, and Bourke RS. SITS-inhibitable Cl⫺ transport and Na⫹-dependent H⫹ production in primary astroglial cultures. Brain Res 173: 111–124, 1979. 164. Kimelberg HK, Narumi S, and Bourke RS. Enzymatic and morphological properties of primary rat brain astrocyte cultures, and enzyme development in vivo. Brain Res 153: 55–77, 1978. 165. Kimelberg HK, Pang S, and Treble DH. Excitatory amino acidstimulated uptake of 22Na⫹ in primary astrocyte cultures. J Neurosci 9: 1141–1149, 1989. 166. Kimelberg HK and Ricard C. Control of intracellular pH in primary astrocyte cultures by external Na⫹. Trans Am Soc Neurochem 13: 112, 1982. 167. Ko YP, Lang HJ, Loh SH, Chu KC, and Wu ML. Cl⫺-dependent and Cl⫺-independent Na⫹/HCO⫺ 3 acid extrusion in cultured rat cerebellar astrocytes. Chin J Physiol 42: 237–248, 1999. 168. Kobayashi S, Morgans CW, Casey JR, and Kopito RR. AE3 anion exchanger isoforms in the vertebrate retina: developmental regulation and differential expression in neurons and glia. J Neurosci 14: 6266 – 6279, 1994. 169. Konnerth A, Lux HD, and Morad M. Proton-induced transformation of calcium channel in chick dorsal root ganglion cells. J Physiol 386: 603– 633, 1987. 170. Kopito RR. Molecular biology of the anion exchanger gene family. Int Rev Cytol 123: 177–199, 1990. 171. Kopito RR, Lee BS, Simmons DM, Lindsey AE, Morgans CW, and Schneider K. Regulation of intracellular pH by a neuronal homolog of the erythrocyte anion exchanger. Cell 59: 927–937, 1989. 172. Koppel M and Spiro K. Uber die Wirkung von Moderatoren (Puffern) bei der Werschiebung des Saurebasengleichge wichtes in biologischen Flussigkeiten. Biochem Z 65: 409 – 439, 1914. 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 voltage-clamped Xenopus oocytes co-expressing NBCe1 and CAIV: evidence for CO⫽3 transport. FASEB J 16: 795, 2002. 128. Grichtchenko II and Chesler M. Depolarization-induced acid secretion in gliotic hippocampal slices. Neuroscience 62: 1057– 1070, 1994. 129. Grichtchenko II and Chesler M. Depolarization-induced alkalinization of astrocytes in gliotic hippocampal slices. Neuroscience 62: 1071–1078, 1994. 130. Grichtchenko II and Chesler M. Calcium- and barium-dependent extracellular alkaline shifts evoked by electrical activity in rat hippocampal slices. Neuroscience 75: 1117–1126, 1996. 131. Grichtchenko II, Choi I, Zhong X, Bray-Ward P, Russell JM, and Boron WF. Cloning, characterization, and chromosomal mapping of a human electroneutral Na(⫹)-driven Cl-HCO3 exchanger. J Biol Chem 276: 8358 – 8363, 2001. 132. Grossman RG, Whiteside L, and Hampton TL. The time course of evoked depolarization of cortical glial cells. Brain Res 14: 401– 415, 1969. 133. Guo H, Svichar N, Esquenazi S, Lin A, and Chesler M. Activitydependent extracellular alkaline shifts and [Ca2⫹]o transients in hippocampal slices are increased by carboxyeosin. Soc Neurosci Abstr Program No. 437.7.2002. Abstract Viewer/Itinerary Planner. Washington, DC: Soc. for Neurosci, 2002. Online. 134. Hartley Z and Dubinsky JM. Changes in intracellular pH associated with glutamate excitotoxicity. J Neurosci 13: 4690 – 4699, 1993. 135. Haugh-Scheidt L and Ripps H. pH regulation in horizontal cells of the skate retina. Exp Eye Res 66: 449 – 463, 1998. 136. Hestrin S, Sah P, and Nicoll R. Mechanisms generating the time course of dual component excitatory synaptic currents recorded in hippocampal slices. Neuron 5: 247–253, 1990. 137. Hille B. Ion Channels of Excitable Membranes. Sunderland, MA: Sinauer, 2001. 138. Hoppe D and Kettenmann H. Carrier-mediated Cl⫺ transport in cultured mouse oligodendrocytes. J Neurosci Res 23: 467– 475, 1989. 139. Hounsgaard J and Nicholson C. The isolated turtle brain and the physiology of neuronal circuits. In: Preparations of Vertebrate Central Nervous System In Vitro, edited by Jahnsen H. New York: Wiley, 1990. 140. Huang W, Smith SE, and Chesler M. Addition of carbonic anhydrase augments extracellular pH buffering in rat cerebral cortex. J Neurophysiol 74: 1806 –1809, 1995. 141. Iijima T, Ciani S, and Hagiwara S. Effects of the external pH on Ca channels: experimental studies and theoretical considerations using a two-site, two-ion model. Proc Natl Acad Sci USA 83: 654 – 658, 1986. 142. Irie T, Hara M, Yasukura T, Minamino M, Omori K, Matsuda H, Inoue K, and Inagaki C. Chloride concentration in cultured hippocampal neurons increases during long-term exposure to ammonia through enhanced expression of an anion exchanger. Brain Res 806: 246 –256, 1998. 143. Irwin RP, Lin SZ, Long RT, and Paul SM. N-methyl-D-aspartate induces a rapid, reversible, and calcium-dependent intracellular acidosis in cultured fetal rat hippocampal neurons. J Neurosci 14: 1352–1357, 1994. 143a.Jakubovicz DE and Klip A. Lactic acid-induced swelling in C6 glial cells via Na⫹/H⫹ exchange. Brain Res 485: 215–224, 1989. 144. Jarolimek W, Misgeld U, and Lux HD. Activity dependent alkaline and acid transients in guinea pig hippocampal slices. Brain Res 505: 225–232, 1989. 145. Javaheri S, De Hemptinne A, Vanheel B, and Leusen I. Changes in brain ECF pH during metabolic acidosis and alkalosis: a microelectrode study. J Appl Physiol 55: 1849 –1853, 1983. 146. Jean T, Frelin C, Vigne P, Barbry P, and Lazdunski M. Biochemical properties of the Na⫹/H⫹ exchange system in rat brain synaptosomes. Interdependence of internal and external pH control of the exchange activity. J Biol Chem 260: 9678 –9684, 1985. 147. Jendelova P, Chvatal A, Simonova Z, and Sykova E. Changes in extracellular pH evoked by excitatory and inhibitory amino acids in the isolated rat spinal cord. ENA Satellite Symp EUSEB Basic Clin Neurosci 17: 63– 64, 1994. 148. Jendelova P and Sykova E. Role of glia in K⫹ and pH homeostasis in the neonatal rat spinal cord. Glia 4: 56 – 63, 1991. pH REGULATION IN THE BRAIN Physiol Rev • VOL 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. caused by efflux of formate and acetate anions through GABAgated channels in crayfish muscle fibres. Neuroscience 34: 359 –368, 1990. Matsumura M, Signor G, and Matthews BW. Substantial increase of protein stability by multiple disulphide bonds. Nature 342: 291–293, 1989. Maycox PR, Deckwerth T, Hell JW, and Jahn R. Glutamate uptake by brain synaptic vesicles. Energy dependence of transport and functional reconstitution in proteoliposomes. J Biol Chem 263: 15423–15428, 1988. McLean LA, Roscoe J, Jorgensen NK, Gorin FA, and Cala PM. Malignant gliomas display altered pH regulation by NHE1 compared with nontransformed astrocytes. Am J Physiol Cell Physiol 278: C676 –C688, 2000. Meech RW and Thomas RC. The effect of calcium injection on the intracellular sodium and pH of snail neurones. J Physiol 265: 867– 879, 1977. Meech RW and Thomas RC. Effect of measured calcium chloride injections on the membrane potential and internal pH of snail neurones. J Physiol 298: 111–129, 1980. Meech RW and Thomas RC. Voltage-dependent intracellular pH in Helix aspersa neurones. J Physiol 390: 433– 452, 1987. Mellergard P, Ouyang YB, and Siesjo BK. Intracellular pH regulation in cultured rat astrocytes in CO2/HCO⫺ 3 -containing media. Exp Brain Res 95: 371–380, 1993. Mellergard PE, Ouyang YB, and Siesjo BK. The regulation of intracellular pH in cultured astrocytes and neuroblastoma cells, and its dependence on extracellular pH in a HCO⫺ 3 free solution. Can J Physiol Pharmacol 70 Suppl: S293–S300, 1992. Mellergard PE and Siesjo BK. Astrocytes fail to regulate intracellular pH at moderately reduced extracellular pH. Neuroreport 2: 695– 698, 1991. Menna G, Tong CK, and Chesler M. Extracellular pH changes and accompanying cation shifts during ouabain-induced spreading depression. J Neurophysiol 83: 1338 –1345, 2000. Meyer TM, Munsch T, and Pape HC. Activity-related changes in intracellular pH in rat thalamic relay neurons. Neuroreport 11: 33–37, 2000. Miesenbock G, De Angelis DA, and Rothman JE. Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394: 192–195, 1998. Miyazaki E, Sakaguchi M, Wakabayashi S, Shigekawa M, and Mihara K. NHE6 protein possesses a signal peptide destined for endoplasmic reticulum membrane and localizes in secretory organelles of the cell. J Biol Chem 276: 49221– 49227, 2001. Moody W Jr. Appearance of calcium action potentials in crayfish slow muscle fibres under conditions of low intracellular pH. J Physiol 302: 335–346, 1980. Moody W Jr. Effects of intracellular H⫹ on the electrical properties of excitable cells. Annu Rev Neurosci 7: 257–278, 1984. Moolenaar WH, Boonstra J, van der Saag PT, and de Laat SW. Sodium/proton exchange in mouse neuroblastoma cells. J Biol Chem 256: 12883–12887, 1981. Mori K, Ogawa Y, Ebihara K, Tamura N, Tashiro K, Kuwahara T, Mukoyama M, Sugawara A, Ozaki S, Tanaka I, and Nakao K. Isolation and characterization of CA XIV, a novel membranebound carbonic anhydrase from mouse kidney. J Biol Chem 274: 15701–15705, 1999. Munsch T and Pape HC. Upregulation of the hyperpolarizationactivated cation current in rat thalamic relay neurones by acetazolamide. J Physiol 519: 505–514, 1999. Mutch WA and Hansen AJ. Extracellular pH changes during spreading depression and cerebral ischemia: mechanisms of brain pH regulation. J Cereb Blood Flow Metab 4: 17–27, 1984. Nachshen DA and Drapeau P. The regulation of cytosolic pH in isolated presynaptic nerve terminals from rat brain. J Gen Physiol 91: 289 –303, 1988. Nakhoul NL, Abdulnour-Nakhoul S, Khuri RN, Lieberman EM, and Hargittai PT. Intracellular pH regulation in rat Schwann cells. Glia 10: 155–164, 1994. Nattie EE. Central chemoreception. In: Regulation of Breathing, edited by Dempsey JA and Pack AI. New York: Dekker, 1995. Nedergaard M and Goldman SA. Carrier-mediated transport of 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 173. Kosterin SA, Bratkova NF, Babich LG, Shinlova OP, Slinchenko NN, Shlykov SG, Zimina BP, Rovenets NA, and Velkich TA. Effect of inhibitors of energy-dependent Ca2⫹-transporting systems on calcium pumps of a smooth muscle cell. Ukr Biokhim Zh 68: 50 – 61, 1996. 174. Kraig RP, Ferreira-Filho CR, and Nicholson C. Alkaline and acid transients in cerebellar microenvironment. J Neurophysiol 49: 831– 850, 1983. 175. Kraig RP, Pulsinelli WA, and Plum F. Hydrogen ion buffering during complete brain ischemia. Brain Res 342: 281–290, 1985. 176. Kraig RP, Pulsinelli WA, and Plum F. Carbonic acid buffer changes during complete brain ischemia. Am J Physiol Regul Integr Comp Physiol 250: R348 –R357, 1986. 177. Krishtal OA, Osipchuk YV, Shelest TN, and Smirnoff SV. Rapid extracellular pH transients related to synaptic transmission in rat hippocampal slices. Brain Res 436: 352–356, 1987. 178. Krishtal OA and Pidoplichko VI. A receptor for protons in the nerve cell membrane. Neuroscience 5: 2325–2327, 1980. 179. Krishtal OA and Pidoplichko VI. A receptor for protons in the membrane of sensory neurons may participate in nociception. Neuroscience 6: 2599 –2601, 1981. 180. Krnjevic K and Walz W. Acidosis and blockade of orthodromic responses caused by anoxia in rat hippocampal slices, at different temperatures. J Physiol 422: 127–144, 1990. 181. Kwan CY, Putney J, and James W. Uptake and intracellular sequestration of divalent cations in resting and methacholine-stimulated mouse lacrimal acinar cells. J Biol Chem 265: 678 – 684, 1990. 182. Lee J, Taira T, Pihlaja P, Ransom BR, and Kaila K. Effects of CO2 on excitatory transmission apparently caused by changes in intracellular pH in the rat hippocampal slice. Brain Res 706: 210 – 216, 1996. 183. Leem CH and Vaughan-Jones RD. Sarcolemmal mechanisms for pHi recovery from alkalosis in the guinea-pig ventricular myocyte. J Physiol 509: 487– 496, 1998. 184. Leniger T, Wiemann M, Bingmann D, Widman G, Hufnagel A, and Bonnet U. Carbonic anhydrase inhibitor sulthiame reduces intracellular pH and epileptiform activity of hippocampal CA3 neurons. Epilepsia 43: 469 – 474, 2002. 185. Linn SC, Kudrycki KE, and Shull GE. The predicted translation product of a cardiac AE3 mRNA contains an N terminus distinct from that of the brain AE3 Cl⫺/HCO⫺ 3 exchanger. Cloning of a cardiac AE3 cDNA, organization of the AE3 gene, and identification of an alternative transcription initiation site. J Biol Chem 267: 7927–7935, 1992. 186. Lipton P and Robacker K. Glycolysis and brain function: [K⫹]o stimulation of protein synthesis and K⫹ uptake require glycolysis. Federation Proc 42: 2875–2880, 1983. 187. Liu Y and Edwards RH. The role of vesicular transport proteins in synaptic transmission and neural degeneration. Annu Rev Neurosci 20: 125–156, 1997. 188. Llinas R, Baker R, and Precht W. Blockage of inhibition by ammonium acetate action on chloride pump in cat trochlear motoneurons. J Neurophysiol 37: 522–532, 1974. 189. Luckermann M, Trapp S, and Ballanyi K. GABA- and glycinemediated fall of intracellular pH in rat medullary neurons in situ. J Neurophysiol 77: 1844 –1852, 1997. 190. Lux HD. Ammonium and chloride extrusion: hyperpolarizing synaptic inhibition in spinal motoneurons. Science 173: 555–557, 1971. 191. Ma E and Haddad GG. Expression and localization of Na⫹/H⫹ exchangers in rat central nervous system. Neuroscience 79: 591– 603, 1997. 192. Maren TH. Carbonic anhydrase: chemistry, physiology, and inhibition. Physiol Rev 47: 595–781, 1967. 193. Markram H, Helm PJ, and Sakmann B. Dendritic calcium transients evoked by single back-propagating action potentials in rat neocortical pyramidal neurons. J Physiol 485: 1–20, 1995. 194. Marshall KC and Engberg I. The effects of hydrogen ion on spinal neurons. Can J Physiol Pharmacol 58: 650 – 655, 1980. 195. Martinez-Zaguilan R, Gillies RJ, and Sanchez-Armass S. Regulation of pH in rat synaptosomes. II. Role of Cl⫺. J Neurophysiol 71: 2249 –2257, 1994. 196. Mason MJ, Mattsson K, Pasternack M, Voipio J, and Kaila K. Postsynaptic fall in intracellular pH and increase in surface pH 1217 1218 MITCHELL CHESLER Physiol Rev • VOL tion (DIA) in rat hippocampal astrocytes. J Neurophysiol 72: 2816 – 2826, 1994. 241. Pappas CA, Ullrich N, and Sontheimer H. Reduction of glial proliferation by the K⫹ channel blockers is mediated by changes in pHi. Neuroreport 6: 193–196, 1994. 242. Parkkila S, Parkkila AK, Rajaniemi H, Shah GN, Grubb JH, Waheed A, and Sly WS. Expression of membrane-associated carbonic anhydrase XIV on neurons and axons in mouse and human brain. Proc Natl Acad Sci USA 98: 1918 –1923, 2001. 243. Pasternack M, Smirnov S, and Kaila K. Proton modulation of functionally distinct GABAA receptors in acutely isolated pyramidal neurons of rat hippocampus. Neuropharmacology 35: 1279 – 1288, 1996. 244. Pasternack M, Voipio J, and Kaila K. Intracellular carbonic anhydrase activity and its role in GABA-induced acidosis in isolated rat hippocampal pyramidal neurones. Acta Physiol Scand 148: 229 –231, 1993. 245. Pedersen SF, Jorgensen NK, Damgaard I, Schousboe A, and Hoffmann EK. Mechanisms of pHi regulation studied in individual neurons cultured from mouse cerebral cortex. J Neurosci Res 51: 431– 441, 1998. 245a.Pellerin L, Pellegri G, Bittar PG, Charnay Y, Bouras C, Martin JL, Stella N, and Magistretti PJ. Evidence supporting the existence of an activity-dependent astrocyte-neuron lactate shuttle. Dev Neurosci 20: 291–299, 1998. 246. Philippe JM, Dubois JM, Rouzaire-Dubois B, Cartron PF, Vallette F, and Morel N. Functional expression of V-ATPases in the plasma membrane of glial cells. Glia 37: 365–373, 2002. 247. Pizzonia JH, Ransom BR, and Pappas CA. Characterization of Na⫹/H⫹ exchange activity in cultured rat hippocampal astrocytes. J Neurosci Res 44: 191–198, 1996. 248. Pocock G and Richards CD. Hydrogen ion regulation in rat cerebellar granule cells studied by single-cell fluorescence microscopy. Eur J Neurosci 4: 136 –143, 1992. 248a.Pumain R, Kurcewicz I, and Louvel J. Fast extracellular calcium transients: involvement in epileptic processes. Science 222: 177– 179, 1983. 249. Putnam RW. Intracellular pH regulation of neurons in chemosensitive and nonchemosensitive areas of brain slices. Respir Physiol 129: 37–56, 2001. 250. Raley-Susman KL, Sapolsky RM, and Kopito RR. Regulation of intracellular pH in cultured hippocampal neurons by an amilorideinsensitive Na⫹/H⫹ exchanger. J Biol Chem 266: 2739 –2745, 1991. 251. Raley-Susman KM, Sapolsky RM, and Kopito RR. Cl⫺/HCO⫺ 3 exchange functions differs in adult and fetal rat hippocampal neurons. Brain Res 614: 308 –314, 1993. 252. Ransom BR. Glial modulation of neural excitability mediated by extracellular pH: a hypothesis. Prog Brain Res 94: 37– 46, 1992. 253. Ransom BR and Goldring S. Slow depolarization in cells presumed to be glia in cerebral cortex of cat. J Neurophysiol 36: 869 – 878, 1973. 253a.Rice ME and Nicholson C. Behavior of extracellular K⫹ and pH in skate (Raja erinacea) cerebellum. Brain Res 461: 328 –334, 1988. 254. Richards CD, Metcalfe JC, Smith GA, and Hesketh TR. Changes in free-calcium levels and pH in synaptosomes during transmitter release. Biochim Biophys Acta 803: 215–220, 1984. 255. Ridderstrale Y and Wistrand PJ. Membrane-associated carbonic anhydrase activity in the brain of CA II-deficient mice. J Neurocytol 29: 263–269, 2000. 256. Ritucci NA, Chambers-Kersh L, Dean JB, and Putnam RW. Intracellular pH regulation in neurons from chemosensitive and nonchemosensitive areas of the medulla. Am J Physiol Regul Integr Comp Physiol 275: R1152–R1163, 1998. 257. Ritucci NA, Dean JB, and Putnam RW. Intracellular pH response to hypercapnia in neurons from chemosensitive areas of the medulla. Am J Physiol Regul Integr Comp Physiol 273: R433–R441, 1997. 258. Ritucci NA, Erlichman JS, Dean JB, and Putnam RW. A fluorescence technique to measure intracellular pH of single neurons in brainstem slices. J Neurosci Methods 68: 149 –163, 1996. 259. Roos A and Boron WF. Intracellular pH. Physiol Rev 61: 296 – 434, 1981. 260. Rose CR and Deitmer JW. Evidence that glial cells modulate 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 lactic acid in cultured neurons and astrocytes. Am J Physiol Regul Integr Comp Physiol 265: R282–R289, 1993. 220. Nedergaard M, Kraig RP, Tanabe J, and Pulsinelli WA. Dynamics of interstitial and intracellular pH in evolving brain infarct. Am J Physiol Regul Integr Comp Physiol 260: R581–R588, 1991. 221. Newman EA. Sodium-bicarbonate cotransport in retinal Muller (glial) cells of the salamander. J Neurosci 11: 3972–3983, 1991. 221a.Newman EA. A physiological measure of carbonic anhydrase in Muller cells. Glia 11: 291–299, 1994. 222. Newman EA. Acid efflux from retinal glial cells generated by sodium bicarbonate cotransport. J Neurosci 16: 159 –168, 1996. 223. Newman EA. Sodium-bicarbonate cotransport in retinal astrocytes and Muller cells of the rat. Glia 26: 302–308, 1999. 224. Newman EA and Astion ML. Localization and stoichiometry of electrogenic sodium bicarbonate cotransport in retinal glial cells. Glia 4: 424 – 428, 1991. 225. Nicholson C and Hounsgaard J. Diffusion in the slice microenvironment and implications for physiological studies. Federation Proc 42: 2865–2868, 1983. 226. Noel J and Pouyssegur J. Hormonal regulation, pharmacology, and membrane sorting of vertebrate Na⫹/H⫹ exchanger isoforms. Am J Physiol Cell Physiol 268: C283–C296, 1995. 227. Nogradi A, Kiraly E, and Mihaly A. Neuronal carbonic anhydrase activity in the central nervous system of the rat: light- and electron histochemical investigations of the islands of Calleja. Acta Histochem 85: 187–193, 1989. 228. Nogradi A and Mihaly A. Distribution of carbonic anhydrase activity in the rat central nervous system, as revealed by a new semipermeable technique. Acta Histochem 84: 153–162, 1988. 229. Nogradi A and Mihaly A. Expression and quantitative changes of carbonic anhydrase in developing neurones of rat central nervous system. Int J Dev Neurosci 9: 555–561, 1991. 229a.Numata M and Orlowski J. Molecular cloning and characterization of a novel (Na⫹,K⫹)/H⫹ exchanger localized to the trans-Golgi network. J Biol Chem 276: 17387–17394, 2001. 230. Numata M, Petrecca K, Lake N, and Orlowski J. Identification of a mitochondrial Na⫹/H⫹ exchanger. J Biol Chem 273: 6951– 6959, 1998. 231. Oakley B and Wen R. Extracellular pH in the isolated retina of the toad in darkness and during illumination. J Physiol 419: 353–378, 1989. 232. O’Connor ER, Sontheimer H, and Ransom BR. Rat hippocampal astrocytes exhibit electrogenic sodium-bicarbonate co-transport. J Neurophysiol 72: 2580 –2589, 1994. 233. O’Neal SG, Rhoads DB, and Racker E. Vanadate inhibition of sarcoplasmic reticulum Ca2⫹-ATPase and other ATPases. Biochem Biophys Res Commun 89: 845– 850, 1979. 234. Orkand RK, Nicholls JG, and Kuffler SW. Effect of nerve impulses on the membrane potential of glial cells in the central nervous system of amphibia. J Neurophysiol 29: 788 – 806, 1966. 234a.Olson JE, Putnam RW, Evers JA, and Munoz N. Taurine efflux and intracellular pH during astrocyte volume regulation. Adv Exp Med Biol 442: 229 –235, 1998. 235. Orlowski J, Kandasamy RA, and Shull GE. Molecular cloning of putative members of the Na/H exchanger gene family. cDNA cloning, deduced amino acid sequence, and mRNA tissue expression of the rat Na/H exchanger NHE-1 and two structurally related proteins. J Biol Chem 267: 9331–9339, 1992. 236. Ou-yang Y, Mellergard P, and Siesjo BK. Regulation of intracellular pH in single rat cortical neurons in vitro: a microspectrofluorometric study. J Cereb Blood Flow Metab 13: 827– 840, 1993. 237. Paalasmaa P and Kaila K. Role of voltage-gated calcium channels in the generation of activity-induced extracellular pH transients in the rat hippocampal slice. J Neurophysiol 75: 2354 –2360, 1996. 238. Paalasmaa P, Taira T, Voipio J, and Kaila K. Extracellular alkaline transients mediated by glutamate receptors in the rat hippocampal slice are not due to a proton conductance. J Neurophysiol 72: 2031–2033, 1994. 239. Pappas CA and Ransom BR. A depolarization-stimulated, bafilomycin-inhibitable H⫹ pump in hippocampal astrocytes. Glia 9: 280 –291, 1993. 240. Pappas CA and Ransom BR. Depolarization-induced alkaliniza- pH REGULATION IN THE BRAIN 261. 262. 263. 264. 265. 267. 268. 269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. Physiol Rev • VOL dependence of glutamate receptor-evoked alkaline shifts in hippocampus. Neuroreport 5: 2441–2445, 1994. 283a.Soleimani M and Aronson PS. Ionic mechanism of Na⫹-HCO⫺ 3 cotransport in rabbit renal basolateral membrane vesicles. J Biol Chem 264: 18302–18308, 1989. 284. Somjen GG. Electrophysiology of neuroglia. Annu Rev Physiol 37: 163–190, 1975. 285. Somjen GG. Acidification of interstitial fluid in hippocampal formation caused by seizures and by spreading depression. Brain Res 311: 186 –188, 1984. 286. Somjen GG and Tombaugh GC. pH modulation of neuronal excitability and central nervous system functions. In: pH and Brain Function, edited by Kaila K and Ransom BR. New York: Wiley-Liss, 1998. 287. Sontheimer H, Black JA, Ransom BR, and Waxman SG. Ion channels in spinal cord astrocytes in vitro. I. Transient expression of high levels of Na⫹ and K⫹ channels. J Neurophysiol 68: 985– 1000, 1992. 288. Spray D and Scemes E. Effects of intracellular pH (and Ca2⫹) on gap junction channels. In: pH and Brain Function, edited by Kaila K and Ransom BR. New York: Wiley-Liss, 1998. 289. Spray DC, Harris AL, and Bennett MVL. Gap junctional conductance is a simple and sensitive function of intracellular pH. Science 211: 712–715, 1981. 289a.Svichar N and Chesler M. Surface carbonic anhydrase activity on astrocytes and neurons facilitates lactate transport. Glia 41: 415– 419, 2003. 290. Sykova E, Jendelova P, Simonova Z, and Chvatal A. K⫹-homeostasis and pH-homeostasis in the developing rat spinal cord is impaired by early postnatal X-irradiation. Brain Res 594: 19 –30, 1992. 291. Sykova E and Svoboda J. Extracellular alkaline-acid-alkaline transients in the rat spinal cord evoked by peripheral stimulation. Brain Res 512: 181–189, 1990. 292. Szabo EZ, Numata M, Shull GE, and Orlowski J. Kinetic and pharmacological properties of human brain Na(⫹)/H(⫹) exchanger isoform 5 stably expressed in Chinese hamster ovary cells. J Biol Chem 275: 6302– 6307, 2000. 293. Szatkowski M and Schlue WR. Mechanisms of pH recovery from intracellular acid loads in the leech connective glial cell. Glia 5: 193–200, 1992. 294. Szatkowski MS and Schlue WR. Chloride-dependent pH regulation in connective glial cells of the leech nervous system. Brain Res 665: 1– 4, 1994. 295. Taira T, Paalasmaa P, Voipio J, and Kaila K. Relative contributions of excitatory and inhibitory neuronal activity in alkaline transients evoked by stimulation of Schaffer Collaterals in the rat hippocampal slice. J Neurophysiol 74: 643– 649, 1995. 296. Taira T, Smirnov S, Voipio J, and Kaila K. Intrinsic proton modulation of excitatory transmission in rat hippocampal slices. Neuroreport 4: 93–96, 1993. 297. Taira T, Voipio J, Paalasmaa P, and Kaila K. Interstitial CO2 and pH transients linked to spreading depression in rat hippocampal slices. Soc Neurosci Abstr 18: 175, 1992. 298. Takahashi KI and Copenhagen DR. Modulation of neuronal function by intracellular pH. Neurosci Res 24: 109 –116, 1996. 299. Takeuchi A and Takeuchi N. Anion permeability of the inhibitory post-synaptic membrane of the crayfish neuromuscular junction. J Physiol 191: 575–590, 1967. 300. Tang CM, Dichter M, and Morad M. Modulation of the N-methyl⫹ D-aspartate channel by extracellular H . Proc Natl Acad Sci USA 87: 6445– 6449, 1990. 301. Tang CM, Presser F, and Morad M. Amiloride selectively blocks the low threshold (T) calcium channel. Science 240: 213–215, 1988. 302. Tank DW, Sugimori M, Connor JA, and Llinas RR. Spatially resolved calcium dynamics of mammalian Purkinje cells in cerebellar slice. Science 242: 773–777, 1988. 303. Thomas RC. Experimental displacement of intracellular pH and the mechanism of its subsequent recovery. J Physiol 354: 3P–22P, 1984. 304. Thomas RC, Coles JA, and Deitmer JW. Homeostatic muffling. Nature 350: 564, 1991. 305. Thomas RC and Meech RW. Hydrogen ion currents and intracel- 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 266. extracellular pH transients induced by neuronal activity in the leech central nervous system. J Physiol 481: 1–5, 1994. Rose CR and Deitmer JW. Stimulus-evoked changes of extra- and intracellular pH in the leech central nervous system. I. Bicarbonate dependence. J Neurophysiol 73: 125–131, 1995. Rose CR and Deitmer JW. Stimulus-evoked changes of extra- and intracellular pH in the leech central nervous system. II. Mechanisms and maintenance of pH homeostasis. J Neurophysiol 73: 132–140, 1995. Rose CR and Ransom BR. Mechanisms of H⫹ and Na⫹ changes induced by glutamate, kainate, and D-aspartate in rat hippocampal astrocytes. J Neurosci 16: 5393–5404, 1996. Saarikoski J, Ruusuvuori E, Koskelainen A, and Donner K. Regulation of intracellular pH in salamander retinal rods. J Physiol 498: 61–72, 1997. Sanchez-Armass S, Martinez-Zaguilani R, Martinez GM, and Gillies RJ. Regulation of pH in rat brain synaptosomes. I. Role of sodium, bicarbonate and potassium. J Neurophysiol 71: 2236 –2248, 1994. Sardet C, Franchi A, and Pouyssegur J. Molecular cloning, primary structure, and expression of the human growth factoractivatable Na⫹/H⫹ antiporter. Cell 56: 271–280, 1989. Saunders RD, Brandon YW, and DeVries GH. Role of intracellular pH in the axolemma- and myelin-induced proliferation of Schwann cells. J Neurochem 52: 1576 –1581, 1989. Sauvaigo S, Vigne P, Frelin C, and Lazdunski M. Identification of an amiloride sensitive Na⫹/H⫹ exchange system in brain synaptosomes. Brain Res 301: 371–374, 1984. Scheller D, Kolb J, and Tegtmeier F. Lactate and pH change in close correlation in the extracellular space of the rat brain during cortical spreading depression. Neurosci Lett 135: 83– 86, 1992. Schmitt BM, Berger UV, Douglas RM, Bevensee MO, Hediger MA, Haddad GG, and Boron WF. Na/HCO3 cotransporters in rat brain: expression in glia, neurons, and choroid plexus. J Neurosci 20: 6839 – 6848, 2000. Schurr A, Miller JJ, Payne RS, and Rigor BM. An increase in lactate output by brain tissue serves to meet the energy needs of glutamate-activated neurons. J Neurosci 19: 34 –39, 1999. Schurr A, West CA, and Rigor BM. Lactate-supported synaptic function in the rat hippocampal slice preparation. Science 240: 1326 –1328, 1988. Schwiening CJ and Boron WF. Regulation of intracellular pH in pyramidal neurones from the rat hippocampus by Na(⫹)-dependent Cl(⫺)-HCO⫺ 3 exchange. J Physiol 475: 59 – 67, 1994. Schwiening CJ, Kennedy HJ, and Thomas RC. Calcium-hydrogen exchange by the plasma membrane Ca-ATPase of voltageclamped snail neurons. Proc R Soc Lond B Biol Sci 253: 285–289, 1993. Schwiening CJ and Thomas RC. pH consequences of calcium regulation. In: pH and Brain Function, edited by Kaila K and Ransom BR. New York: Wiley-Liss, 1998. Shmigol A, Eisner DA, and Wray S. Carboxyeosin decreases the rate of decay of the [Ca2⫹]i transient in uterine smooth muscle cells isolated from pregnant rats. Pflügers Arch 437: 158 –160, 1998. Shrode LD and Putnam RW. Intracellular pH regulation in primary rat astrocytes and C6 glioma cells. Glia 12: 196 –210, 1994. Siesjo BK, von Hanwehr R, Nergelius G, Nevander G, and Ingvar M. Extra- and intracellular pH in the brain during seizures and in the recovery period following the arrest of seizure activity. J Cereb Blood Flow Metab 5: 47–57, 1985. Smart TG and Constanti A. A novel effect of zinc on the lobster muscle GABA receptor. Proc R Soc Lond B Biol Sci 215: 327–341, 1982. Smith GA, Brett CL, and Church J. Effects of noradrenaline on intracellular pH in acutely dissociated adult rat hippocampal CA1 neurones. J Physiol 512: 487–505, 1998. Smith SE and Chesler M. Calcium-independent extracellular alkaline shifts evoked by glutamate. Soc Neurosci Abstr 21: 1860, 1995. Smith SE and Chesler M. Effect of divalent cations on AMPAevoked extracellular alkaline shifts in rat hippocampal slices. J Neurophysiol 82: 1902–1908, 1999. Smith SE, Gottfried JA, Chen JC, and Chesler M. Calcium 1219 1220 MITCHELL CHESLER Physiol Rev • VOL 328. 329. 330. 331. 332. 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343. 344. 345. 346. 347. 348. 349. SE. Cadmium inhibition of the erythrocyte Ca2⫹ pump. A molecular interpretation. J Biol Chem 264: 5613–5615, 1989. Voipio J and Kaila K. Interstitial PCO2 and pH in rat hippocampal slices measured by means of a novel fast CO2/H(⫹)-sensitive microelectrode based on a PVC-gelled membrane. Pflügers Arch 423: 193–201, 1993. Voipio J, Paalasmaa P, Taira T, and Kaila K. Pharmacological characterization of extracellular pH transients evoked by selective synaptic and exogenous activation of AMPA, NMDA, and GABAA receptors in the rat hippocampal slice. J Neurophysiol 74: 633– 642, 1995. Voipio J, Pasternack M, Rydqvist B, and Kaila K. Effect of gamma-aminobutyric acid on intracellular pH in the crayfish stretch-receptor neurone. J Exp Biol 156: 349 –360, 1991. Volk C, Albert T, and Kempski OS. A proton-translocating H⫹ATPase is involved in C6 glial pH regulation. Biochim Biophys Acta 1372: 28 –36, 1998. Von Hanwehr R, Smith ML, and Siesjo BK. Extra- and intracellular pH during near-complete forebrain ischemia in the rat. J Neurochem 46: 331–339, 1986. Vyklicky L, Vlachova V, and Krusek J. The effect of external pH changes on responses to excitatory amino acids in mouse hippocampal neurones. J Physiol 430: 497–517, 1990. Waheed A, Zhu XL, and Sly WS. Membrane-associated carbonic anhydrase from rat lung. Purification, characterization, tissue distribution, and comparison with carbonic anhydrase IVs of other mammals. J Biol Chem 267: 3308 –3311, 1992. Waheed A, Zhu XL, Sly WS, Wetzel P, and Gros G. Rat skeletal muscle membrane associated carbonic anhydrase is 39-kDa glycosylated GPI-anchored CA IV. Arch Biochem Biophys 294: 550 –556, 1992. Wakabayashi S, Shigekawa M, and Pouyssegur J. Molecular physiology of vertebrate Na⫹/H⫹ exchangers. Physiol Rev 77: 51– 74, 1997. Waldmann R. Proton-gated cation channels: neuronal acid sensors in the central and peripheral nervous system. Adv Exp Med Biol 502: 293–304, 2001. Walz W. pH shifts evoked by neuronal stimulation in slices of rat hippocampus. Can J Physiol Pharmacol 67: 577–581, 1989. Walz W and Wuttke W. Resistance of astrocyte electrical membrane properties to acidosis changes in the presence of lactate. Brain Res 504: 82– 86, 1989. Wang CZ, Yano H, Nagashima K, and Seino S. The Na⫹-driven Cl⫺/HCO⫺ 3 exchanger. Cloning, tissue distribution, and functional characterization. J Biol Chem 275: 35486 –35490, 2000. Wang GJ, Randall RD, and Thayer SA. Glutamate-induced intracellular acidification of cultured hippocampal neurons demonstrates altered energy metabolism resulting from Ca2⫹ loads. J Neurophysiol 72: 2563–2569, 1994. Wang W, Bradley SR, and Richerson GB. Quantification of the response of rat medullary raphe neurones to independent changes in pH(o) and P(CO2). J Physiol 540: 951–970, 2002. Wang Z, Orlowski J, and Shull GE. Primary structure and functional expression of a novel gastrointestinal isoform of the rat Na/H exchanger. J Biol Chem 268: 11925–11928, 1993. Watzke N, Rauen T, Bamberg E, and Grewer C. On the mechanism of proton transport by the neuronal excitatory amino acid carrier 1. J Gen Physiol 116: 609 – 622, 2000. Wender R, Brown AM, Fern R, Swanson RA, Farrell K, and Ransom BR. Astrocytic glycogen influences axon function and survival during glucose deprivation in central white matter. J Neurosci 20: 6804 – 6810, 2000. Werth JL and Thayer SA. Mitochondria buffer physiological calcium loads in cultured rat dorsal root ganglion neurons. J Neurosci 14: 348 –356, 1994. Whitney PL and Briggle TV. Membrane-associated carbonic anhydrase purified from bovine lung. J Biol Chem 257: 12056 –12059, 1982. Wiemann M and Bingmann D. Ventrolateral neurons of medullary organotypic cultures: intracellular pH regulation and bioelectric activity. Respir Physiol 129: 57–70, 2001. Willoughby D and Schwiening CJ. Electrically evoked dendritic 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 lular pH in depolarized voltage-clamped snail neurones. Nature 299: 826 – 828, 1982. 306. Tolkovsky AM and Richards CD. Na⫹/H⫹ exchange is the major mechanism of pH regulation in cultured sympathetic neurons: measurements in single cell bodies and neurites using a fluorescent pH indicator. Neuroscience 22: 1093–1102, 1987. 307. Tombaugh GC and Somjen GG. pH modulation of voltage-gated ion channels. In: pH and Brain Function, edited by Kaila K and Ransom BR. New York: Wiley-Liss, 1998. 308. Tombaugh GC. Intracellular pH buffering shapes activity-dependent Ca2⫹ dynamics in dendrites of CA1 interneurons. J Neurophysiol 80: 1702–1712, 1998. 309. Tombaugh GC and Somjen GG. Differential sensitivity to intracellular pH among high- and low-threshold Ca2⫹ currents in isolated rat CA1 neurons. J Neurophysiol 77: 639 – 653, 1997. 310. Tong CK, Brion LP, Suarez C, and Chesler M. Interstitial carbonic anhydrase (CA) activity in brain is attributable to membranebound CA type IV. J Neurosci 20: 8247– 8253, 2000. 311. Tong CK, Cammer W, and Chesler M. Activity-dependent pH shifts in hippocampal slices from normal and carbonic anhydrase II-deficient mice. Glia 31: 125–130, 2000. 312. Tong CK and Chesler M. Activity-evoked extracellular pH shifts in slices of rat dorsal lateral geniculate nucleus. Brain Res 815: 373–381, 1999. 313. Tong CK and Chesler M. Endogenous pH shifts facilitate spreading depression by effect on NMDA receptors. J Neurophysiol 81: 1988 –1991, 1999. 314. Tong CK and Chesler M. Modulation of spreading depression by changes in extracellular pH. J Neurophysiol 84: 2449 –2457, 2000. 315. Tong CK and Chesler M. Role of extracellular carbonic anhydrase in the spread and onset of activity-evoked pH shifts. Soc Neurosci Abstr 26: 1624, 2000. 316. Trapp S, Luckermann M, Brooks PA, and Ballanyi K. Acidosis of rat dorsal vagal neurons in situ during spontaneous and evoked activity. J Physiol 496: 695–710, 1996. 317. Trapp S, Luckermann M, Kaila K, and Ballanyi K. Acidosis of hippocampal neurones mediated by a plasmalemmal Ca2⫹/H⫹ pump. Neuroreport 7: 2000 –2004, 1996. 318. Travis DM, Wiley C, Bohdan RN, and Maren TH. Selective renal carbonic anhydrase inhibition without respiratory effect: pharmacology of 2-benzenesulfonamido-1,3,4-thiadiazole-5-sulfonamide (CL 11,366). JPET 143: 383–394, 1964. 319. Traynelis SF. pH modulation of ligand-gated ion channels. In: pH and Brain Function, edited by Kaila K and Ransom BR. New York: Wiley-Liss, 1998. 320. Traynelis SF and Cull-Candy SG. Proton inhibition of N-methylD-aspartate receptors in cerebellar neurons. Nature 345: 347–350, 1990. 321. Trivedi B and Danforth WH. Effect of pH on the kinetics of frog muscle phosphofructokinase. J Biol Chem 241: 4110 – 4112, 1966. 322. Urbanics R, Leniger-Follert E, and Lubbers DW. Time course of changes of extracellular H⫹ and K⫹ activities during and after direct electrical stimulation of the brain cortex. Pflügers Arch 378: 47–53, 1978. 323. Van Slyke D. On the measurement of buffer values and on the relationship of buffer value to the dissociation constant of the buffer and the concentration and the reaction of the buffer solution. J Biol Chem 52: 525–570, 1922. 324. Velisek L, Dreier JP, Stanton PK, Heinemann U, and Moshe SL. Lowering of extracellular pH suppresses low-Mg(2⫹)-induced seizures in combined entorhinal cortex-hippocampal slices. Exp Brain Res 101: 44 –52, 1994. 325. Velisek L, Veliskova J, and Moshe SL. Site-specific effects of local pH changes in the substantia nigra pars reticulata on flurothyl-induced seizures. Brain Res 782: 310 –313, 1998. 325a.Venton BJ, Michael DJ, and Wightman RM. Correlation of local changes in extracellular oxygen and pH that accompany dopaminergic terminal activity in the rat caudate-putamen. J Neurochem 84: 373–381, 2003. 326. Verbost PM, Flik G, Lock RA, and Wendelaar Bonga SE. Cadmium inhibits plasma membrane calcium transport. J Membr Biol 102: 97–104, 1988. 327. Verbost PM, Flik G, Pang PK, Lock RA, and Wendelaar Bonga pH REGULATION IN THE BRAIN 350. 351. 352. 353. 354. 356. 357. Physiol Rev • VOL 358. 359. 360. 361. 362. 363. 364. and darkness on pH outside rod photoreceptors in the cat retina. Exp Eye Res 54: 685– 697, 1992. Yao H, Ma E, Gu XQ, and Haddad GG. Intracellular pH regulation of CA1 neurons in Na(⫹)/H(⫹) isoform 1 mutant mice. J Clin Invest 104: 637– 645, 1999. Yu X, Carroll S, Rigaud JL, and Inesi G. H⫹ countertransport and electrogenicity of the sarcoplasmic reticulum Ca2⫹ pump in reconstituted proteoliposomes. Biophys J 64: 1232–1242, 1993. Zerangue N and Kavanaugh MP. Flux coupling in a neuronal glutamate transporter. Nature 383: 634 – 637, 1996. Zhan R, Fujiwara N, Tanaka E, Shimoji K, Zhan RZFN, Tanaka E, and Shimoji K. Intracellular acidification induced by membrane depolarization in rat hippocampal slices: roles of intracellular Ca2⫹ and glycolysis. Brain Res 780: 86 –94, 1998. Zhan RZ, Fujiwara N, Yamakura T, Taga K, Fukuda S, Endoh H, and Shimoji K. NMDA induces a biphasic change in intracellular pH in rat hippocampal slices. Brain Res 760: 179 –186, 1997. Zhao D and Dhalla NS. Characterization of rat heart plasma membrane Ca2⫹/Mg2⫹ ATPase. Arch Biochem Biophys 263: 281– 292, 1988. Zhu XL and Sly WS. Carbonic anhydrase IV from human lung. Purification, characterization, and comparison with membrane carbonic anhydrase from human kidney. J Biol Chem 265: 8795– 8801, 1990. 83 • OCTOBER 2003 • www.prv.org Downloaded from http://physrev.physiology.org/ by 10.220.32.247 on July 4, 2017 355. pH transients in rat cerebellar Purkinje cells. J Physiol 544: 487– 499, 2002. Wistrand PJ and Kinne R. Carbonic anhydrase activity of isolated brush border and basal-lateral membranes of renal tubular cells. Pflügers Arch 370: 121–126, 1977. Wong V, Barrett CP, Donati EJ, Eng LF, and Guth L. Carbonic anhydrase activity in first-order sensory neurons of the rat. J Histochem Cytochem 31: 293–300, 1983. Wong V, Barrett CP, Donati EJ, and Guth L. Distribution of carbonic anhydrase activity in neurons of the rat. J Comp Neurol 257: 122–129, 1987. Wuttke WA and Walz W. Sodium- and bicarbonate-independent regulation of intracellular pH in culture mouse astrocytes. Neurosci Lett 117: 105–110, 1990. Xiong ZQ, Saggau P, and Stringer JL. Activity-dependent intracellular acidification correlates with the duration of seizure activity. J Neurosci 20: 1290 –1296, 2000. Xiong ZQ and Stringer JL. Extracellular pH responses in CA1 and the dentate gyrus during electrical stimulation, seizure discharges, and spreading depression. J Neurophysiol 83: 3519 –3524, 2000. Xiong ZQ and Stringer JL. Regulation of extracellular pH in the developing hippocampus. Brain Res Dev Brain Res 122: 113–117, 2000. Yamamoto F, Borgula GA, and Steinberg RH. Effects of light 1221
© Copyright 2026 Paperzz