Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Anoxia and Na⁺/H⁺ exchange activity in rat hippocampal neurons Sheldon, Claire Alexis 2004

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2005-995488.pdf [ 15.85MB ]
Metadata
JSON: 1.0092383.json
JSON-LD: 1.0092383+ld.json
RDF/XML (Pretty): 1.0092383.xml
RDF/JSON: 1.0092383+rdf.json
Turtle: 1.0092383+rdf-turtle.txt
N-Triples: 1.0092383+rdf-ntriples.txt
Original Record: 1.0092383 +original-record.json
Full Text
1.0092383.txt
Citation
1.0092383.ris

Full Text

\ ANOXIA AND Na+/H+ EXCHANGE ACTIVITY IN RAT HIPPOCAMPAL NEURONS by CLAIRE ALEXIS SHELDON B.Sc. (Physiology), University of Alberta, 1996 M.Sc. (Physiology), University of British Columbia, 1999 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF COMBINED DOCTOR OF MEDICINE AND DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Anatomy, Cell Biology & Physiology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October, 2004 © Claire Alexis Sheldon, 2004 11 ABSTRACT In the present study, the effects of anoxia on intracellular pH (pHj) and intracellular free sodium concentration ([Na+]i) were examined in isolated rat hippocampal neurons loaded with H+- and/or Na+-sensitive fluorophores, and the contribution of changes in Na+/H+ exchange activity to the changes in pHj and [Na+], observed during and after anoxia were assessed. This assessment was aided by the development of a microspectrofluorimetric technique which permitted concurrent measurements of pHj and [Na+]j in the same neuron. I found that, in hippocampal neurons, NaVH"^ exchange activity was reduced shortly following the onset of anoxia, possibly as a result of declining internal ATP levels, and did not contribute to the increases in pHj or [Na+]j observed at this time. In contrast, Na4/!!4' exchange activity was stimulated immediately after anoxia and contributed to acid extrusion and Na+ influx during this particularly vulnerable period. As a result, the reported neuroprotective actions of Na+/H+ exchange inhibitors are likely mediated in the immediate post-anoxic period, consequent upon reductions in acid extrusion and/or internal Na+ loading. A Zn2+-sensitive H+ efflux pathway, possibly a voltage-activated H4" conductance activated by membrane depolarization, also contributed to acid extrusion during and immediately after anoxia and may act to limit the potentially detrimental activation of Na+/H+ exchange activity observed after anoxia. The final series of experiments identified additional mechanisms that contribute to the changes in [Na+]i evoked by anoxia in cultured postnatal rat hippocampal neurons. Na+ influx occurred through multiple pathways, the relative contributions of which differed not only during and after anoxia but also in neurons maintained in culture for different durations of time. Understanding the fundamental cellular mechanisms that contribute to anoxia-evoked changes in pHj and [Na+]i in mammalian central neurons may uncover novel therapeutic strategies for the treatment of stroke. Ill TABLE OF CONTENTS Page Abstract ii Table of Contents iiList of Tables viii List of Figures ix Acknowledgements xii CHAPTER ONE - Introduction 1.0. General introduction 1 1.1. Clinical background1.2. Neuropathology 2 1.3. Pathophysiology 5 1.3.1. [Ca2+] j and excitotoxicity 7 1.3.2. Role of changes in pH 9 1.3.2.1. Historical background 9 1.3.2.2. pH: neurotoxic or neuroprotective? 10 1.3.2.3. pH: extracellular or intracellular? 2 1.3.2.4. pHj: relevance to anoxia and the timing of its actions 14 1.3.3. Role of changes in [Na+]i 15 1.3.4. Pathophysiology: Summary 7 1.4. Maintenance of intracellular pH1.4.1. Na+/H+exchange 18 1.4.1.1. General structure and expression patterns in non- 18 neuronal tissues 1.4.1.2. Expression patterns in nervous tissue 20 1.4.1.3. Na+/H+ exchange activity in rat hippocampal neurons 22 1.4.1.4. Na /H+ exchange: relevance to anoxia 24 1.4.2. HC03"-dependent pH; regulating mechanisms 6 1.4.2.1. Neuronal HC03"-dependent pHj regulation 26 1.4.2.2. HC03-dependent pHj regulation: relevance to anoxia 28 1.4.3. Additional pHj regulating mechanisms 30 1.5. Synthesis and objectives 32 IV CHAPTER TWO - General Methods 2.0. Cell preparation 44 2.0.1. Choice of experimental preparations 42.0.2. Acutely isolated adult rat hippocampal CA1 pyramidal neurons 45 2.0.3. Postnatal rat hippocampal neuronal cultures 46 2.1. Solutions and test compounds 47 2.2. Induction of anoxia 9 2.3. Micro spectrofluorimetry 50 2.3.1. Dye loading 1 2.3.2. Imaging equipment 52 2.3.3. Calculation of pHis [Na+]j and [Ca2+]i 53 2.3.3.1. BCECF2.3.3.2. HPTS 55 2.3.3.3. SBFI 6 2.3.3.4. Fura-2 8 2.4. Experimental procedures and data analysis 5CHAPTER THREE - Intracellular pH Response to Anoxia in Acutely Isolated Adult Rat Hippocampal CA1 Pyramidal Neurons: Reduced Na+/H+ exchange activity during anoxia 3.0. Introduction 73 3.1. Materials and methods 74 3.1.1. Experimental preparation 73.1.2. Recording techniques 5 3.1.3. Experimental maneuvers  5 3.1.4. ATP determination 76 3.1.5. Statistical analysis 7 3.2. Results 73.2.1. Steady-state pH, under normoxic conditions 77 3.2.2. Steady-state pHj response to anoxia 78 3.2.3. Contribution of changes in [Ca2+]i to the changes in pH* observed 80 during anoxia 3.2.4. Na+/H* exchange activity during anoxia 82 3.2.4.1. Steady-state pHj measurements3.2.4.2. Recovery of pHi from imposed internal acid loads 84 3.2.4.3. Role of internal ATP depletion 88 3.3. Discussion 91 3.3.1. Characterization of the changes in pHj observed during and 91 following anoxia in adult rat hippocampal CA1 pyramidal neurons 3.3.2. Reduced contribution from Na+/H+ exchange to acid extrusion 93 during anoxia V CHAPTER FOUR - Intracellular pH Response to Anoxia in Acutely Isolated Adult Rat Hippocampal CA1 Pyramidal Neurons: Increased Na+/H+ exchange activity after anoxia 4.0. Introduction 119 4.1. Materials and methods 120 4.1.1. Experimental preparation 124.1.2. Experimental maneuvers4.2. Results 121 4.2.1. Na+/H+ exchange activity in the immediate post-anoxic period 121 4.2.2. Contribution of a Na 0- and HC03~-independent mechanism to 125 acid extrusion during and following anoxia 4.2.3. Effects of changes in pH0 128 4.3. Discussion 130 4.3.1. Na+/H+ exchange activity after anoxia 13 0 4.3.2. Potential contribution of a gu+ to the increases in pHj during and 133 after anoxia 4.3.3. Synthesis of Chapters 3 and 4 135 CHAPTER FIVE - Changes in [Na+]j Induced By Anoxia in Isolated Rat Hippocampal Neurons: Role of Na+/H+ exchange activity 5.0. Introduction 157 5.1. Materials and methods 15 8 5.1.1. Experimental preparation 158 5.1.2. Recording techniques 15 8 5.1.3. Internal ATP determination 159 5.1.4. Experimental procedures and data analysis 160 5.2. Results 161 5.2.1. Anoxia-induced increases in [Na+]j in acutely isolated adult rat 161 hippocampal CA1 pyramidal neurons 5.2.2. Anoxia-induced increases in [Na+]j in cultured postnatal rat 162 hippocampal neurons 5.2.3. Role of Na+/H+exchange activity 165 5.2.4. Role of HC03-dependent mechanisms 167 5.3. Discussion 169 5.3.1. Resting [Na+] j under normoxic conditions 165.3.2. Anoxia-evoked increases in [Na+]i 169 5.3.3. Contribution of Na+/H+exchange activity 171 5.3.4. Contribution of HC03~-dependent mechanisms 173 5.3.5. Summary 175 VI CHAPTER SIX - Concurrent Measurement of pHj and [Na+]j with Fluorescent Indicators: A further evaluation of the role of Na+/H+ exchange in anoxia-evoked changes in pHj and [Na+]j 6.0. Introduction 190 6.1. Materials and methods 192 6.1.1. Experimental preparation 196.1.2. Dye-loading and recording techniques 192 6.1.3. Calculation of [Na+]i and pHj 194 6.1.4. Data analysis 195 6.2. Results 197 6.2.1. Separation of SBFI and SNARF fluorescence emissions in situ 197 6.2.2. Full calibrations of SBFI, carboxy SNARF-1 and SNARF-5F 199 ratio values in situ 6.2.3. Effects of changes in [Na+] on pHj measurements with carboxy 201 SNARF-1 and SNARF-5F in situ 6.2.4. Concurrent measurements of pH; and [Na+]i in rat hippocampal 201 neurons 6.2.5. Contribution of Na+/H+exchange activity to anoxia-evoked 203 changes in pHj and [Na^ji 6.3. Discussion 205 6.3.1. Part 1: The development of microspectrofluorimetric methods for 205 the concurrent measurement of pHj and [Na+]i 6.3.1.1. Part 1: Technical considerations 206 6.3.1.2. Part 1: Summary 208 6.3.2. Part 2: Anoxia-evoked changes in pHj and [Na+]i 209 CHAPTER SEVEN - Additional Mechanisms Contributing to Anoxia-Evoked Increases in [Na+]i in Cultured Postnatal Rat Hippocampal Neurons 7.0. Introduction 239 7.1. Materials and methods 240 7.1.1. Experimental preparation and solutions 247.1.2. Recording techniques 241 7.1.3. Experimental procedures and data analysis 247.2. Results 242 7.2.1. Increases in [Na+]j during anoxia 247.2.1.1. Role of ionotropic glutamate receptor-operated channels 242 7.2.1.2. Role of voltage-activated Na+ channels 247.2.1.3. Role of plasmalemmal Na/Ca exchange and 243 Na+/K+/2C1" cotransport 7.2.1.4. Role of non-selective cation channels 244 7.2.2. Increases in [Na+]i after anoxia 246 Vll 7.2.2.1. Role of ionotropic glutamate receptor-operated channels, voltage-activated Na+ channels and Na+/K+/2C1" cotransport 246 7.2.2.2. Role of plasmalemmal Na+/Ca2+ exchange 247 7.2.2.3. Role of non-selective cation channels 248 7.3. Discussion 247.3.1. The role of ionotropic glutamate receptor-operated channels 249 7.3.2. The role of voltage-activated Na+channels 250 7.3.3. The role of Na+/Ca2+ exchange 251 7.3.4. The role of Na+/K+/2C1" cotransport 254 7.3.5. The role of a Gd3+-sensitive mechanism 255 7.3.6. Age-dependence of the increases in [Na+]i observed during and 257 following anoxia 7.3.7. Synthesis of Chapters 5, 6 and 7 258 CHAPTER EIGHT - Summary and Conclusions 8.1. Experimental protocols and preparations 273 8.2. Changes in pHj and [Na+]j during and after anoxia 276 8.3. Contribution of pHj regulating mechanisms to the changes in pHj and [Na+]j 278 evoked by anoxia 8.3.1. HC03"-dependent pHj regulating mechanisms 278 8.3.2. Na+/H+exchange activity 280 8.3.2.1. Potential implications of anoxia-evoked changes in 282 Na+/H+ exchange activity 8.3.2.2. Na+/H+exchange inhibitors: neuroprotective actions and 285 therapeutic potential 8.3.3. Role of a putative voltage-activated proton conductance 286 8.4. On the mechanisms contributing to anoxia-evoked changes in [Na+]j 288 BIBLIOGRAPHY 29viii LIST OF TABLES Page Table 1.1. Clinical trials of selected agents in acute stroke 3 5 Table 2.1. Composition of commonly used experimental solutions 61 Table 2.2. List of pharmacological agents 62 Table 2.3. Composition of solutions used for in situ calibrations of pH and Na+- 64 sensitive fluorophores Table 3.1. Anoxia-evoked changes in steady-state pHj 97 Table 5.1. Contribution of pHj regulating mechanisms to the increase in [Na+]i observed 178 during anoxia Table 5.2. Contribution of pHj regulating mechanisms to the increase in [Na+]j observed 179 following anoxia under 0 [K+]0 conditions Table 6.1. Calibration parameters for SBFI, carboxy SNARF-1 and SNARF-5F in 214 single dye- and dual-dye loaded hippocampal neurons Table 6.2. NH44" prepulse- and anoxia-evoked changes in pHj and [Na+]i in hippocampal 215 neurons loaded with a SNARF-based fluorophore and/or SBFI Table 7.1. Potential mechanisms contributing to the increase in [Na+]i observed during 259 anoxia Table 7.2. Potential mechanisms contributing to the increase in [Na+]j observed after 260 anoxia under 0 [K ]0 conditions ix LIST OF FIGURES Page Fig. 1.1. Pathways of ischemic cell death 36 Fig. 1.2. A schematic illustration of the pattern of ionic and electrical changes induced 38 by anoxia or ischemia in mammalian central neurons Fig. 1.3. An illustration of the pHj regulating mechanisms present in rat hippocampal 40 neurons Fig. 1.4. The contribution of Na+/H+ exchange to myocardial injury induced by 42 ischemia Fig. 2.1. A schematic representation of the optical equipment used in neurons single- 65 loaded with a dual-excitation fluorophore (i.e. BCECF, HPTS, SBFI or fura-2) Fig. 2.2. Sample in situ calibration plot for BCECF 67 Fig. 2.3. In situ calibration of SBFI at 37°C, pH0 7.35 9 Fig. 2.4. Consistency of rates of pHj recovery from internal acid loads imposed under 71 control conditions Fig. 3.1. Steady-state pHj changes evoked by transient periods of anoxia 98 Fig. 3.2. Relationship between pre-anoxic resting pHi values and anoxia-evoked 100 changes in steady-state pHi Fig. 3.3. Effects of anoxia on steady-state pHj and fura-2-derived BIj^/BI^o ratio 102 values in the presence and absence of external Ca2+ Fig. 3.4. Anoxia-evoked changes in pET, measured with HPTS 104 Fig. 3.5. The effects of external Na+ substitutions on the magnitude of the fall and rise 106 in pHj observed during anoxia under HC03"-free, Hepes-buffered conditions (pH0 7.35, 37°C) Fig. 3.6. Rates of pHj recovery from internal acid loads are reduced during anoxia 108 Fig. 3.7. Rates of pHj recovery from internal acid loads prior to and during anoxia in 110 neurons with "low" resting pHi values Fig. 3.8. pH; recovery from acid loads imposed prior to and during anoxia under 112 reduced pH0 conditions Fig. 3.9. Treatment with 2-DG and antimycin A slows rates of pH, recovery from 115 internal acid loads Fig. 3.10. pHj recovery from internal acid loads in creatine pretreated neurons 117 Fig. 4.1. Effects of external Na+ substitutions on the increase in pHj observed 139 following 5 min anoxia Fig. 4.2. Recovery of pHj from internal acid loads imposed immediately after anoxia 141 X Fig. 4.3. Recovery of pHj from internal acid loads imposed immediately after anoxia 143 in "low" pHj neurons Fig. 4.4. Effects of modulating the activity of the cAMP/PKA pathway on the pH; 145 response to anoxia Fig. 4.5. Effects of changes in perfusate composition on the increases in pHj observed 147 during and following 5 min anoxia 9+ Fig. 4.6. Influence of Zn on the recovery of pHj from internal acid loads imposed 149 immediately after anoxia Fig. 4.7. Effect of high [K+]0 on pHj recovery from intracellular acid loads imposed 151 under normoxic Na+0-free, nominally HC03 -free, Hepes-buffered conditions (pH0 7.35) Fig. 4.8. Representative traces of the effects of inhibitors of P-type H+,K+-ATPase and 153 V-type H+-ATPase activity on pHi recovery from intracellular acid loads imposed under high [K+]0 conditions (pH0 7.35) Fig. 4.9. Effects of reduced pH0 on rates of pHj recovery from acid loads imposed in 155 the immediate post-anoxic period Fig. 5.1. Anoxia-evoked changes in [Na+]i in rat hippocampal neurons 180 Fig. 5.2. Contribution of reduced Na+,K+-ATPase activity to the increase in [Na+] j 182 observed during anoxia Fig. 5.3. Changes in [Na+]i observed after anoxia during inhibition of Na+,K+-ATPase 184 activity Fig. 5.4. Effects of modulating Na /H exchange activity on the increase in [Na+]i 186 observed after anoxia (Na,K-ATPase inhibited) Fig. 5.5. Contribution of HC03'-dependent mechanisms to the increase in [Na+]i 188 observed after anoxia (Na+,K - ATPase inhibited) Fig. 6.1. A schematic representation of the optical equipment used to measure [Na+]i 216 and pHi in hippocampal neurons loaded with SBFI and/or carboxy SNARF-1 or SNARF-5F Fig. 6.2. Fluorescence emissions from single- and dual-dye loaded hippocampal 218 neurons Fig. 6.3. Quenching effects between SBFI and SNARF-based fluorophores 221 Fig. 6.4. In situ calibration of SBFI at 37°C, pH0 7.35 223 Fig. 6.5. In situ calibration of carboxy SNARF-1 225 Fig. 6.6. Sodium sensitivity of carboxy SNARF-1 in situ 227 Fig. 6.7. Changes in pHj and [Na+]i observed in rat hippocampal neurons in response 229 to intracellular acid loads imposed by the NH/ prepulse technique Fig. 6.8. Changes in pHj and [Na+]i induced by anoxia in rat hippocampal neurons 231 Fig. 6.9. Reducing [Na+]0 limits the increases in pHj and [Na+]j observed immediately 233 after anoxia XI Fig. 6.10. Relationships between changes in pHi and [Na+]i observed in the period 235 immediately after anoxia (Na,K-ATPase inhibited) Fig. 6.11. The influence of maneuvers which inhibit Na+/H+ exchange activity on the 237 anoxia-evoked changes in pHj and [Na+]j measured concurrently in individual cells co-loaded with either carboxy SNARF-1 or SNARF-5F and SBFI Fig. 7.1. Ionotropic glutamate receptor-operated channels do not contribute to the 261 increase in [Na+]j observed during anoxia under the present experimental conditions Fig. 7.2. Role of voltage-activated Na+ channels, Na+/Ca2+ exchange and Na+/K+/2CT 263 cotransport in the increase in [Na+]j observed during anoxia Fig. 7.3. Effect of Gd on the increase in [Na ], observed during anoxia 265 Fig. 7.4. Role of reactive oxygen species in the increase in [Na+]j observed during 267 anoxia + 2~f" Fig. 7.5. Effects of maneuvers which modulate Na /Ca exchange activity on the 269 increase in [Na ]j observed after anoxia (Na ,K -ATPase inhibited) Fig. 7.6. Effects of Gd3+ on the increase in [Na+]j observed after anoxia (Na+,K+- 271 ATPase inhibited) Fig. 8.1. A schematic representation of the mechanisms found in the present studies to 292 contribute to the changes in pHj observed during and after anoxia in rat hippocampal neurons. Fig. 8.2. A schematic representation of the mechanisms found in the present studies to 294 contribute to the changes in [Na+]i observed during and after anoxia in rat hippocampal neurons. Xll ACKNOWLEDGEMENTS I would first like to thank John Church for his support and guidance and for the seemingly endless time and energy he has dedicated towards my training. My sincerest thanks. I would also like to express my gratitude to the members of my supervisory committee, Drs. Kenneth Baimbridge, Steven Kehl and Lynn Raymond, for sharing their time, scientific knowledge and words of encouragement with me. To the members of the entire Departments of Physiology and Anatomy and Cell Biology (students and faculty alike), and especially to Sally Osborne, Pauline Dan, Claudia Krebs, Herman Fernandes, Tony Kelly, May Cheng - thank you for your camaraderie and support. To Val Smith, Candace Hofmann, Mareika Grant, Gord Rintoul, Mike Rauh and Paul Yong, thank you for always being such wonderful friends! Finally, to the people that I hold dearest to my heart, Craig, Mary, Bill, Tess, Mia, Jay, Signy, Toby and Tori, I could not have done this without your never-ending love, humor and, at times, ridicule. I CHAPTER ONE INTRODUCTION 1.0. General Introduction 1.1. Clinical background Stroke is the third leading cause of death in the Western world. In Canada, 16 000 people die as a result of strokes every year and, in addition to those, 300 000 people are living with the disabling effects of stroke (Heart and Stroke Foundation of Canada website, www.heartandstroke.ca'). While the number of stroke patients across Canada is expected to increase in the next two decades, therapies to date have had limited success. Thrombolytic and neuroprotective therapies represent two fundamental approaches to the treatment of strokes. Approximately 80% of strokes occur when the blood supply to the brain has been interrupted (ischemic strokes) and, in the majority of these cases, are consequent upon a physical blockage of the arteries supplying the central nervous system by thrombi or emboli. Since the energy required to maintain neuronal function and integrity is derived principally from oxidative phosphorylation, thrombolytic therapies are aimed at restoring the supply of oxygen and glucose necessary for normal cerebral function (Erecihska & Silver, 1989; Silver et al. 1997). The use of the intravenous thrombolytic agent, tissue plasminogen activator (tPA), administered within 3 h after stroke onset, is the only clinically approved treatment for acute ischemic strokes (see NINDS group study, 1995). However, it is estimated that only 1% of all ischemic stroke patients are treated with tPA (Schellinger et al. 2004), the major reason for this failure of treatment being the arrival and identification of eligible patients more than 3 h after stroke onset. There are also inherent risks associated with the resumption of blood flow which demand careful patient monitoring during and after thrombolytic therapy: reperfusion may be associated with an 2 increased risk of haemorrhage (see NINDS group study, 1995), abnormal neuronal activity (e.g. Xu & Pulsinelli, 1996; Reese et al. 2002), vascular endothelial damage and reactive oxygen species generation (e.g. Kent et al. 2001), all of which may promote the development of subsequent tissue damage. Finally, the efficacy of thrombolytic therapies is also limited by variations in clot composition and size (Broderick & Hacke, 2002a). The second approach to the treatment of acute ischemic strokes is neuroprotective agents. These therapies are designed to limit the cellular mechanisms leading to neuronal death and are, in large part, based on basic science research examining the neuronal response to periods of ischemia. More than 40 potential neuroprotective agents, directed towards a number of different fundamental mechanisms, have been examined in human trials; however, the vast majority of these have had limited clinical efficacy (see Table 1.1: Dietrich, 1998; Lee et al. 1999; Lo et al. 2003), an observation which is perhaps not surprising considering the large number of factors involved in initiating and maintaining the mechanisms thought to be responsible for ischemic neuronal death (Fig. 1.1). It is notable that, by influencing several of these mechanisms, mild hypothermia represents a successful neuroprotective strategy, although its utility is also limited by the lack of feasible techniques to induce hypothermia in a large number of patients (Broderick & Hacke, 2002b). Despite the potentially tragic consequences for patients with stroke and their families, as well as the growing socio-economic impact, effective treatments for ischemic strokes are clearly lacking. 1.2. Neuropathology In this section, I will outline some of the features of cell damage observed in response to periods of cerebral ischemia. It is recognized that "the immediate effect of ischemia is similar to that of anoxia" (Somjen, 2004) and, as a result, the initial events that occur in response to ischemia (i.e. oxygen and glucose deprivation) can be effectively represented by oxygen deprivation alone 3 (anoxia). Indeed, as described below in Section 1.3, the pattern of ionic and electrical events initiated during anoxia and ischemia are similar, although the period of time over which they occur is shorter with more severe insults (e.g. Silver et al. 1997; Centonze et al. 2001). As ischemia is experienced most often clinically, in the section that follows I will use the term ischemia, although in later sections, in light of the similar pathophysiological mechanisms initiated during anoxia and ischemia, both terms will be employed. A number of key factors determine the extent and pattern of damage observed in response to periods of cerebral ischemia. First, cellular populations within the brain have differing intrinsic sensitivities to periods of ischemia: neurons are most sensitive while glia and endothelial cells are more tolerant to periods of ischemia (Pulsinelli et al. 1982; Bramlett & Dietrich, 2004). Second, the magnitude of ischemic cell damage is proportional to the severity and duration of the ischemic insult. Periods of complete ischemia will produce more pronounced cell damage than equivalent periods of incomplete ischemia and, similarly, longer durations of ischemia will produce increasing amounts of damage (see Silver & Erecihska, 1990; Dietrich, 1998). Third, the type of ischemic insult experienced will determine the pattern of cell damage. In response to periods of global ischemia, which occur following the interruption of the supply of oxygen and glucose to the entire brain, as in cardiac arrest, selectively vulnerable cells (for example, CA1 pyramidal neurons of the hippocampus and GABAergic spiny neurons of the striatum) are specifically damaged. Focal ischemia, which occurs following the occlusion of blood vessel(s) supplying a particular region of the brain, produces a 'core' region of severe neuronal damage and a surrounding penumbra where blood flow is less markedly reduced and neurons remain potentially salvageable (reviewed by Back, 1998; Lipton, 1999). Fourth, although events occurring during ischemia initiate cellular dysfunction, further cell damage can 4 be influenced significantly by events that occur during early recirculation (e.g. Gao et al. 1998; Taylor et al. 1999). Despite these various considerations, ischemic neuronal death can be classified in two general ways: acute vs. delayed and necrotic vs. apoptotic. Acute neuronal death occurs as neurons die rapidly following the onset of ischemia and, in these cases, ischemia tends to be severe and/or the neurons are selectively vulnerable to the ischemic episode. Morphologically, acute neuronal death is, in large part, necrotic, characterized by the appearance of darkened nuclei, swollen organelles and a loss of plasma membrane integrity. There are 3 categories of necrotic cell death: edematous cell change, ischemic cell change, and homogenizing cell change (Fig. 1.1; Lipton, 1999). While each category has its own distinquishing features, in most cases, the morphological changes reflect excessive Na+ and Ca2+ entry: accompanying anion (e.g. CI" and HCO3") and water entry promote cell swelling while increases in intracellular free Ca2+ concentrations ([Ca ]j) activate proteolytic and lipolytic enzymes (Goldberg & Choi, 1993; Lee et al. 1999; Lipton, 1999; Small et al. 1999; Yuan et al. 2003). In contrast to acute neuronal death, delayed neuronal death occurs from hours to days, or more, after an ischemic insult. In contrast to the necrotic changes associated with acute neuronal death, delayed neuronal death is, in many cases, apoptotic in nature, characterized by chromatin condensation, cell shrinkage and the generation of apoptotic cell bodies (Lipton, 1999; Small et al. 1999). Changes in [Ca2+]j and the internal concentration of K+ ions ([K+]j) appear to be early events that trigger downstream apoptotic cascades; however, the precise contribution of these (and other) mechanisms to the initiation of apoptosis, and the identities of the downstream effector mechanisms, remain an area of active investigation (Lee et al. 1999; Snider et al. 1999; also see Banasiak et al. 2004 for an illustration of the relationship between [Na+]i and hypoxia-induced apoptosis). Recent evidence suggests that both necrosis and apoptosis may be activated in parallel in the ischemic brain, possibly 5 through common intracellular cascades, and that the resulting observed morphologies lie along a continuum between necrosis and apoptosis (see Snider et al. 1999). A final distinct form of cell death, characterized by a condensed cytoplasm containing many large lysosomes, is autophagocytotic cell death; however, there is limited information regarding both its underlying mechanisms and its relative contribution to ischemic neuronal death (Lipton, 1999; Yuan et al. 2003; also see Florez-McClure et al: 2004). Taken together, these observations suggest that, while many factors influence the extent and pattern of ischemic cell damage, common pathophysiological events may contribute to the initiation and regulation pf both necrotic and apoptotic forms of neuronal death. In the following section, I will discuss the contribution of early changes in the concentrations of intracellular ions to the pathophysiology of ischemic cell damage. 1.3. Pathophysiology Ischemia is an extremely complex metabolic insult during which diverse cellular events are initiated that, in turn, are critical determinants of subsequent functional and structural changes and eventual cell death. It is generally acknowledged that declining intracellular ATP levels and the accompanying changes in the concentrations of internal ions are critically important in initiating ischemic cell damage (Fig. 1.1; also see Somjen, 2002). The extra- and intracellular ionic changes that occur during and following anoxia or ischemia have been studied extensively (for reviews see Hansen, 1985; Erecihska & Silver, 1994; Martin et al. 1994; Lipton, 1999). As illustrated in Fig. 1.2, anoxia- or ischemia-induced changes in ion concentrations (both extra- and intracellular) can be described in 3 consecutive phases: during phase 1, there are slow changes in extra- and intracellular ion concentrations, which presumably reflect the declining capacity of neurons to maintain internal ATP levels (and the consequent inhibition of the activities of ATP-dependent ion pumps, notably the Na ,K -ATPase); the onset of phase 2 is associated with a precipitous and striking depolarization (referred to as 'ischemic' or 'anoxic' depolarization) and is accompanied by marked ionic dysregulation and an increased likelihood of irreversible neuronal injury; finally, slow changes in extra- and intracellular ion concentrations occur during phase 3. Thus, in response to ischemia, there are reductions in the external concentrations of sodium, calcium and chloride ions ([Na+]0, [Ca2+]0 and [CT]0, respectively) and increases in the external concentration of potassium ions ([K+]0). Corresponding increases in [Ca2+]i and the internal concentrations of Na+ and CI" ions ([Na+]j and [Cl"]i, respectively), and decreases in [K+]j, are also observed. Changes in extra- and intracellular pH (pH0 and pHj, respectively) also occur: falls in pH0 and pH, are seen soon after the onset of anoxia or ischemia and may be interrupted by external alkaline transients upon anoxic depolarization before giving way to further acidic shifts. The period of time over which this sequence of events occurs depends on the nature and severity of the insult. For example, during focal ischemia, neurons located within the penumbra exhibit less marked falls in pHj, increases in [Ca ]j and [Na ]\ and are less likely to undergo 'anoxic depolarization', compared to the neurons located within the adjacent core of the ischemic tissue (Back, 1998; Lipton, 1999). Finally, if the anoxic or ischemic insult is transient in nature, the restoration of internal ATP levels, transmembrane ion gradients and membrane potential may be possible; however, this recovery does not occur consistently and the restoration of oxygen and glucose does not necessarily result in the return of ion homeostasis and membrane repolarization (Ekholm ef a/. 1993; O'Reilly et al. 1995; Tanaka et al. 1997). In light of these marked ionic shifts, in the following sections (Section 1.3.1 - 1.3.4), I will outline the potential contributions of changes in the extra- and intracellular concentrations of Ca2+, H+ and Na+ ions to the pathophysiology of ischemic cell death. 7 1.3.1. [Ca I, and excitotoxicity The contribution of Ca2+ ions to anoxia- or ischemia-induced neuronal damage has received particular attention, notably within the framework of the excitoxic model of cell injury (Olney et al. 1971). Both anoxia and ischemia are associated with the release of excitatory neurotransmitters (most notably glutamate) into the synaptic cleft (e.g. Benveniste et al. 1984; Santos et al. 1996). The resulting prolonged activation of Af-methyl-D-aspartate (NMDA) and other subtypes of ionotropic glutamate receptors (e.g. a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid; AMP A) elicits rises in [Na+]i and [Ca2+]j which, acting together, initiate neuronal damage by aggravating the decline in internal ATP levels, further promoting ion dysregulation and activating a variety of intracellular degradative enzymes (Choi, 1990; Herman et al. 1990; Mitani et al. 1994; Kristian & Siesjb, 1997). The central role of glutamate and Ca2+j to anoxia-induced neuronal damage is supported by a number of key findings: i) extracellular glutamate levels increase during ischemia in vivo (e.g. Benveniste et al. 1984; Choi, 1990) or metabolic inhibition in vitro (e.g. Goldberg & Choi, 1993; Kimura et al. 1998); ii) glutamate receptor antagonists, under certain circumstances, may reduce neuronal death in vivo (e.g. Simon et al. 1984; Church et al. 1988; but see Nurse & Corbett, 1996) and reduce Ca2+ influx and neuronal death in neuronal cultures exposed to oxygen-glucose deprivation (e.g. Goldberg et al. 1987; Kaku et al. 1993; but see Newell et al. 1995); and iii) the susceptibility to ischemia in vivo (e.g. Jensen, 2002) and excitotoxicity in vitro (e.g. Cheng et al. 1999) parallel developmental increases in ionotropic glutamate receptor expression (but see Marks et al. 2000). Finally, postischemic enhancements of NMD A- and non-NMDA receptor-mediated Ca2+ influx have been observed 1 - 12 h following transient ischemia 8 (e.g. Pellegrini-Giampietro et al. 1997; Mitani et al. 1998; Xu et al. 1999) and may contribute to further neuronal dysfunction and death at this time (Gao et al. 1998; Lipton, 1999). In contrast, glutamate receptor antagonists fail to influence neuronal survival following periods of prolonged (>90 min) oxygen and glucose deprivation in vitro (Aarts et al. 2003; also see Obeidat et al. 2000), and their proposed neuroprotective effects in in vivo animal models (see above) have been attributed to drug-induced reductions in cerebral temperature (e.g. Nurse & Corbett, 1996). Most importantly, however, glutamate receptor antagonists have proven ineffective in improving clinical outcome following a stroke (Table 1.1; e.g. Lee et al. 1999; Lees et al. 2000). These negative findings have supported suggestions that glutamate-mediated excitotoxicity cannot completely account for the detrimental effects of anoxia or ischemia on neuronal viability. Indeed, some studies have illustrated differences in the time course of glutamate-induced neuronal dysfunction compared with that induced by anoxia or ischemia (see Friedman & Haddad, 1993; Chow & Haddad, 1998). In addition, while increases in [Ca2+]i, specifically Ca2+ entry through NMDA receptor-operated channels (Tymianski et al. 1993; Sattler et al. 1998), contribute to glutamate-induced neuronal death in vitro, anoxia- and ischemia-induced neuronal death is not necessarily dependent on an increase in [Ca ]\ (e.g. Friedman & Haddad, 1993; Li et al. 1996) and the increases in [Ca2+]i that occur in selectively vulnerable hippocampal CA1 pyramidal neurons in vivo during and upon recovery from transient periods of global ischemia are only partially influenced by pharmacological inhibitors of NMDA receptors (Silver & Erecihska, 1990 and 1992). This may, at least in part, reflect the facts that NMDA receptors are inhibited by reductions in pH0 (e.g. Vyklicky et al. 1990; Traynelis & Cull-Candy, 1991) and that marked falls in pH0 occur during anoxia and ischemia in vivo (see above and Section 1.3.2). 9 These apparently contradictory findings have highlighted the complex regulation of anoxic/ischemic neuronal damage and have prompted renewed interest into the identification of additional glutamate receptor-independent events that might contribute to the pathogenesis of anoxic/ischemic neuronal damage. 1.3.2. Role of changes in pH 1.3.2.1. Historical background The importance of pH to the effects of ischemia on neuronal function and death has been recognized for over 25 years (reviewed by Siesjo et al. 1996). Early studies demonstrated key relationships between pre-ischemic plasma glucose levels, the magnitudes of the interstitial acidosis during ischemia and the extent of subsequent neuronal damage observed in response to global ischemia (Ljunggren et al. 191 A; Diemer & Siemkowicz, 1981; Siemkowicz & Hansen, 1981). Initially demonstrated by Myers and Yamaguchi (1977), and confirmed by many others, pre-ischemic hyperglycemia enhances infarct size and limits functional recovery following ischemia compared to normoglycemic controls, a relationship which appears to hold true clinically as patients hyperglycemic prior to stroke show significantly worse outcomes (see Kent et al. 2001; Parsons et al. 2002). After infusing animals with varying amounts of glucose, Li et al. (1995) observed a close correlation between intra-ischemic pH0 and extent of cell necrosis measured 7 days following 10 min global ischemia and further illustrated that the effect of hyperglycemia to aggravate neuronal damage occurs over a narrow range of glucose concentrations and pH0 values (with the toxic effects of pH0 observed at pH0 values < 5.8 - 6.4; Li et al. 1995; also see Hum et al. 1991). In additional studies, normoglycemic animals were subjected to transient ischemia under hypercapnic conditions to lower intra-ischemic pH0 to values similar to those observed in hyperglycemic animals (pH0 -5.8-6.3; Katsura et al. 1994; 10 also see Hurn et al. 1991) and, in these animals, ischemia-induced neuronal damage was similarly aggravated despite the absence of hyperglycemia, providing further support for a critical role of pH0 in the development of ischemia-induced neuronal damage in vivo. However, further studies, employing global and focal models of ischemia, have produced conflicting results; studies have reported reductions, increases and no changes in infarct size under hyperglycemic conditions (e.g. LeBlanc et al. 1993; Li et al. 1996; Schurr et al. 1999). In an attempt to reconcile these differences, Sapolsky et al. (1996) argued that the influence of acidity is not likely to be simply toxic or beneficial to an ischemic neuron but rather "represents the summation and counterbalancing of the salutary and deleterious effects of acidosis". In the following sections, the apparently dichotomous role that changes in pH0/pHj may have in the pathogenesis of ischemic neuronal damage will be discussed further. 1.3.2.2. pH: neurotoxic or neuroprotective? Changes in pH0, of varying magnitudes and directions, are commonly observed within the central nervous system and may function as signalling events. Normal synaptic transmission and postsynaptic receptor activation both elicit changes in pH0 (see Krishtal et al. 1987; Jarolimek et al. 1989; Chesler & Kaila, 1992; Rose & Deitmer, 1995). In hippocampal slices, orthodromic electrical stimulation (2.5 to 20 Hz for 20 - 60 s) evokes external alkaline transients of up to 0.2 pH units, often followed by prolonged (e.g. min) extracellular acidifications of up to 0.1 pH units (Jarolimek et al. 1989). These changes in pH0 may have important regulatory functions; for example, synaptic vesicle fusion releases protons into the synaptic cleft which may, in turn, feedback and inhibit the voltage-activated Ca channels initially responsible for their release (DeVries, 2001; also see Balestrino & Somjen, 1988; Church & McLennan, 1989; Tombaugh, 1994; Tombaugh & Somjen, 1996; Church et al. 1998). 11 During pathological events, such as cerebral ischemia, epileptic seizures and spreading depression, dramatic alterations in cerebral pH0, ranging in magnitude from 0.1 to greater than 1 pH units, can occur (e.g. Mutch & Hansen, 1984; Hansen, 1985; Menna et al. 2000; Tong & Chesler, 2000) and, as under normoxic conditions, may have important downstream consequences. Indeed, seizure activity induces decreases in pH0 that may act to limit or even terminate the seizure activity itself (e.g. Aram & Lodge, 1987; Tong & Chesler, 2000). Mild falls in pH0, such as those observed during the early stages of anoxia or ischemia, suppress neuronal excitability and reduce neuronal energy demand by decreasing NMDA receptor-mediated currents (see above) and voltage-activated Na+ and Ca2+ currents and augmenting GABAA receptor-mediated CI" currents (Krnjevic & Walz, 1990; Cummins et al. 1991; Tombaugh & Somjen, 1996; Krishek et al. 1996). These, and other, effects may underlie the neuroprotective actions of mild extracellular acidosis beyond that produced by glutamate receptor antagonists alone (Tombaugh & Sapolsky, 1990; Kaku et al. 1993). In contrast to the neuroprotective effects of mild reductions in pH0, marked falls in pH0 may kill neurons (reviewed by Tombaugh & Sapolsky, 1993; Siesjo et al. 1996; see also Kraig et al. 1987; Nedergaard et al. 1991). Excessive and/or prolonged extracellular acidosis rapidly depletes tissue ATP levels and, following an anoxic insult, stimulates lipid peroxidation, protein denaturation and the accumulation of free radicals (Nedergaard & Goldman, 1993; Tombaugh & Sapolsky, 1993; Siesjo et al. 1996; Trafton et al. 1996; Raley-Susman & Barnes, 1998), effects which may contribute to the observed ability of marked external acidosis (pH0 <~ 6.5) to induce necrotic and apoptotic neuronal death in hippocampal slices (Ding et al. 2000). Finally, acting at least in part by enhancing the activities of NMDA receptors and high-voltage activated Ca2+ channels, increases in pH0 promote neuronal excitability and the development of seizure activity, enhance the propagation of spreading depression and worsen functional recovery from global 12 ischemia (Aram & Lodge, 1987; Church & McLennan, 1989; Tombaugh & Somjen, 1996; Hum et al. 1997; Tong & Chesler, 2000). All together, pH0 is a key determinant of neuronal function and viability during and following an anoxic/ischemic insult, although the basis for its effects remains incompletely understood. 1.3.2.3. pH: extracellular or intracellular? The great majority of studies which have examined the effects of changes in pH0 on neuronal function, whether under physiological or pathophysiological conditions, have not assessed the possible involvement of concomitant changes in pHj. In contrast to many peripheral cell types (e.g. myocytes, neutrophils and cardiac Purkinje fibers), wherein pH0 has only a minor influence on pHj (the slope of linear regression line relating pHj to pH0, ApHj:ApH0, = 0.35 - 0.40; see Aickin, 1984; Wilding et al. 1992), in many types of mammalian central neurons, pH0 is a critical determinant of pHj (ApHi:ApH0 = 0.75; Ou-Yang et al. 1993; Sanchez-Armass et al. 1994; Church et al. 1998; also see Ritucci et al. 1998 for data comparing the pH0-dependency of pHi in chemosensitive vs. nonchemosensitive medullary neurons). Thus, the influence of pH0 on neuronal function under normoxic and anoxic conditions may, at least in part, be secondary to changes in pHj. Changes in neuronal pHj can also occur in the absence of marked changes in pH0. Membrane depolarization can elicit falls in pHj in invertebrate (e.g. Ahmed & Conner, 1980) and mammalian central neurons (Trapp et al. 1996a; Zhan et al. 1998; Meyer et al. 2000; Willoughby & Schwiening, 2002; reviewed by Ballanyi & Kaila, 1998). Through mechanisms dependent on Ca2+ influx and HCO3" efflux, respectively, glutamatergic and GABAergic neurotransmission can 13 similarly elicit falls in pHj (Kaila & Voipio, 1987; Irwin et al. 1994; Wang et al. 1994; Trapp et al. 1996a; Wu et al. 1999). By influencing a range of intracellular processes, from the activities of internal metabolic pathways and intracellular second messenger systems to the activities of voltage- and ligand-gated ion channels, intrinsic changes in pHj can modulate not only neuronal activity under normoxic conditions but may also be important determinants of neuronal viability in response to anoxic or ischemic insults. For example, the activity of phosphofructokinase is markedly pHj dependent with a change as small as 0.1 pH unit being able to completely activate or inactivate the glycolytic pathway (Busa & Nuccitelli, 1984). Similarly, adenylate cyclase and cyclic nucleotide phosphodiesterase, those enzymes responsible for cAMP synthesis and hydrolysis, respectively, are regulated by pHj such that an increase in pH, causes an elevation in intracellular cAMP levels (Busa & Nuccitelli, 1984). Additional intracellular second messenger cascades, including Ca2+/calmodulin (e.g. Busa & Nuccitelli, 1984), nitric oxide synthase (e.g. Anderson & Meyer, 2000; Conte, 2003) and phospholipase Ai/arachidonic acid (e.g. Stella et al. 1995), are also sensitive to changes in pPF,. The activities of these, and other, intracellular pathways are affected by anoxia and ischemia and appear to be important determinants of subsequent neuronal function and viability by regulating, for example, neurotransmitter release, the generation of reactive oxygen species and membrane integrity (reviewed by Meldrum, 1996; Wieloch et al. 1996; Sapirstein & Bonventre, 2000; Prast & Philippu, 2001); thus, the possibility exists that changes in pHj may influence the neuronal response to anoxia or ischemia by regulating the activities of diverse intracellular signalling cascades. Analogous to the protective effects of mild falls in pH0 (see above), mild reductions in 2"1" pHj may exert a neuroprotective effect by inhibiting, for example, voltage-activated Ca currents (e.g. Takahashi & Copenhagen, 1996; Tombaugh & Somjen, 1997), neurotransmitter release (e.g. Drapeau & Nachshen, 1988; Chen et al. 1998b), and gap junctional conductances (thereby 14 reducing neuronal synchrony and, possibly, anoxia-induced epileptiform activity; Spray & Bennett, 1985; Church & Baimbridge, 1991; Perez-Velazquez et al. 1994; Xiong et al. 2000). On the other hand, the neurotoxicity associated with exposure to highly acidic media (pH0 < 6.5) appears to be a function of the degree and duration of the intracellular acidification produced consequent upon the fall in pH0 (Hum et al. 1991; Nedergaard et al. 1991). Marked decreases in pHj are capable of inducing cellular dysfunction by initiating DNA damage and the production of free radicals (Siesjo et al. 1996; Vincent et al. 1999), as well as inhibiting Ca2+- and voltage-activated K+ channels (see review by Tombaugh & Somjen, 1998; also see Church, 1992; Church et al. 1998; Liu et al. 1999; Kelly & Church, 2004) and promoting cellular swelling (Jakubovicz & Klip, 1989). Recent studies have also illustrated that falls in pHi are early events in mitochondria-dependent apoptosis; considering that optimal caspase activity is observed between pH 6.3 to 6.8, marked falls in pHj are capable of regulating downstream caspase activity (Matsuyama et al. 2000; Takahashi et al. 2004). Finally, although changes in pHj are determined by changes in [rf]\ and [HC03~]j, HCO3" ions themselves can have important effects independent from changes in pHj: HCO3" ions can control cAMP signalling (Chen et al. 2000) and the production of reactive oxygen species (e.g. Konorev et al. 2000; Han et al. 2003) and can directly impact neuronal excitability (Bruehl et al. 2000; Gu et al. 2000; Bruehl & Witte, 2003), especially in the post-anoxic period (e.g. Roberts et al. 2000). 1.3.2.4. pH,: relevance to anoxia and the timing of its actions The studies cited above indicate that changes in pH; per se can modulate neuronal activity under normoxic conditions as well as neuronal viability in response to anoxic or ischemic insults. Similar findings have been made in peripheral cell types, most notably in cardiac myocytes and 15 hepatocytes. In these cell types, it appears that the recovery of pHj to physiological values during reperfusion, rather than the fall in pHj observed during anoxia or ischemia, initiates an intracellular cascade of events, including attendant changes in [Na+]i and [Ca2+],, that finally results in cellular damage (Lazdunski et al. 1985; Currin et al. 1991; Bond et al. 1993). The relevance of this series of events has not been established in mammalian central neurons. On the one hand, decreases in pHj during anoxia have been suggested to modulate the susceptibility of neurons to injury (e.g. LeBlanc et al. 1993; Tyson et al. 1993; Roberts & Chih, 1997). On the other hand, neuronal damage in cultured neocortical neurons evoked by metabolic inhibition (a combination of 2-deoxy-D-glucose and cyanide) can be reduced by inhibiting the rate of restoration of pHj to normal values in the period following the metabolic inhibition (Vornov et al. 1996). An understanding of the basis for these apparently disparate findings requires the characterization of the intrinsic changes in pHj which occur in mammalian central neurons during and following periods of anoxia or ischemia. 1.3.3. Role of changes in [Na+"h Early increases in Na+j occur in response to anoxia or ischemia and appear to contribute to the pathophysiology of subsequent neuronal death (see Urenjak & Obrenovitch, 1996; Lipton, 1999). Thus, the removal of extracellular Na+ reduces anoxia- and ischemia-evoked changes in neuronal morphology (e.g. Friedman & Haddad, 1994b; Chidekel et al. 1997; Raley-Susman et al. 2001) and promotes subsequent functional recovery (e.g. Fried et al. 1995; Raley-Susman et al. 2001). The beneficial effects of reduced Na+ entry may reflect a number of factors, including: i) a reduced demand for cellular ATP to maintain the Na+ gradient (e.g. Erecinska et al. 1991; Fowler & Li, 1998; Chinopoulos et al. 2000); ii) decreased neuronal swelling (e.g. Friedman & Haddad, 1994b; Chidekel et al. 1997); iii) a reduction in the magnitude and duration of 'anoxic' 16 membrane depolarization (e.g. Haddad & Jiang, 1993; Tanaka et al. 1997; Calabresi et al. 1999b); iv) limiting reverse-mode glutamate reuptake and/or Na+/Ca2+ exchange (e.g. Taylor et al. 1995; Kimura et al. 1998; Breder et al. 2000); and v) preventing NaVdependent increases in NMDA receptor-mediated currents (Yu & Salter, 1998). In direct contrast to studies in non-neuronal cell types (e.g. cardiac myocytes; Carmeliet, 1999) and myelinated central nervous system axons (Stys, 1998), the mechanisms which mediate Na+ influx in response to anoxia or ischemia in mammalian central neurons remain relatively poorly defined. Although Na+ influx through glutamate receptor-operated channels has received some attention (Miiller & Somjen, 2000, LoPachin et al. 2001), glutamate-mediated excitotoxicity may not be a completely valid model for the direct actions of anoxia or ischemia on neurons (see Section 1.3.1) and few studies (e.g. Chen et al. 1999) have systematically addressed the potential contributions of other mechanisms integral to the cell (e.g. voltage-activated Na+ channels (e.g. Fung & Haddad, 1997; Banasiak et al. 2004), NaVH4" exchange (e.g. Kintner et al. 2004), forward-mode Na /Ca exchange (e.g. Chidekel et al. 1997) and Na+/K+/2C1" cotransport (e.g. Beck et al. 2003)) to the increases in neuronal [Na+]i observed during anoxia or ischemia (see Pisani et al. 1998a; Guatteo et al. 1998; Calabresi et al. 1998 for studies in slice preparations). In addition, despite indications that continued Na+ entry upon reperfusion may be more damaging than Na+ entry during anoxia or ischemia (see Lipton, 1999), the pathways that mediate Na+ entry in the period immediately after anoxia/ischemia have not been characterized and it remains unknown whether these pathways might differ from those active during an insult, as reported for Ca2+ (see Silver & Erecihska, 1990 and 1992). Further studies are necessary, first, to characterize the changes in [Na+]i observed in mammalian central neurons during and following periods of anoxia or ischemia and, second, to examine the mechanism(s) contributing to the changes in [Na+]j observed at these times. 17 1.3.4. Pathophysiology: Summary Internal shifts in [Ca ], pH and [Na ] occur in neurons in response to anoxia or ischemia. While the potential contribution of changes in Ca2+j has received particular attention, the early changes in pHi and Na+i that occur during and/or following periods of anoxia or ischemia also contribute to the pathophysiology of anoxic and ischemic neuronal death. Nevertheless, the mechanisms that contribute to the production of anoxia-induced changes in pHj and [Na+]j in mammalian central neurons remain relatively poorly defined. Given the close inter-relationships and, in fact, linked regulation of the internal concentrations of these ions via the activities of plasma membrane pHj regulating mechanisms, in the following Section, I will discuss the pHj regulating mechanisms present in rat hippocampal neurons and examine their potential abilities to influence not only pHj but also [Na+]j. 1.4. Maintenance of intracellular pH Assuming a resting membrane potential of -60 mV and pH0 7.3, the passive distribution of protons across the neuronal plasma membrane would drive pHi to ~ 6.3 (Roos & Boron, 1981; Chesler, 2003). Given that the cytosolic compartment is significantly more alkaline than this value, mechanisms must be in place to extrude intracellular protons or other acid equivalents. However, compared to non-neuronal and invertebrate neuronal cell types, relatively few studies have investigated pH; regulating mechanisms in vertebrate and, in particular, mammalian central neurons, even under normoxic conditions (reviewed by Chesler, 2003). Given the importance of pH, (as well as [Na+]0 in regulating neuronal excitability and viability, the characteristics of the mechanisms involved in the regulation of pHj (as well as [Na+]j) are of critical importance. To date, two major classes of pHj regulating mechanisms have been found to be present in neurons of the mammalian central nervous system: i) HCCV-independent Na+/H+ exchangers; 18 and ii) HCOy-dependent exchangers and co-transporters (Fig. 1.3). However, given the variety of mechanisms which have been found to participate in pHj regulation in invertebrate neurons and vertebrate non-neuronal cells, it is likely that additional pHj regulating mechanisms will be identified in mammalian central neurons. 1.4.1. Na+/H+ exchange 1.4.1.1. General structure and expression patterns in non-neuronal tissues In many cell types, pHj is regulated primarily by a family of Na+/H+ exchangers, transmembrane proteins that mediate the electroneutral exchange of an intracellular proton for an extracellular sodium ion. The first mammalian Na+/H+ exchanger isoform was cloned and sequenced in 1989 (NHE isoform 1, NHE1; Sardet et al. 1989) and since that time, seven additional mammalian isoforms have been cloned (NHE2 - 8), each demonstrating varying tissue and/or intracellular distributions (for reviews see Wakabayashi et al. 1997; Putney et al. 2002; Orlowski & Grinstein, 2004; also see Numata et al. 1998; Brett et al. 2002b). NHE1 is expressed on the plasma membrane of virtually all cell types and fulfills the role of a "house-keeping" acid extrusion mechanism, whereas other isoforms exhibit a more restricted distribution. NHE2 and NHE3, for example, are predominantly expressed in intestinal and renal epithelial cells, whereas NHE6 and NHE7, which share only -20% amino acid homology with other isoforms, are expressed in membranes of intracellular organelles (see Brett et al. 2002b). Despite their varied tissue distributions, Na+/H+ exchangers share a common structure. Hydropathy profiles predict that Na+/H+ exchangers exist as integral membrane proteins with 12 transmembrane domains (see Fliegel, 2001). Located on the cytoplasmic face of the exchange mechanism is a 'H+ sensor' which allosterically controls the activity of the exchange mechanism in response to an intracellular acidosis (see Aronson, 1985). Additional sites located on the large intracellular C-19 terminus mediate isoform-specific regulation of transport activity by a variety of intracellular regulatory proteins and second-messenger pathways (see Fliegel, 2001). Thus, the activities of distinct NaVrf" exchanger isoforms can be modulated by multiple mechanisms, ranging from direct interactions with regulatory proteins (e.g. calmodulin, Na^H*" exchanger regulatory factors, NHERFs; Hall et al. 1998) and lipids (e.g. phosphatidylinositol-4,5-bisphosphate, PIP2; Aharonovitz et al. 2000) to direct phosphorylation by protein kinases (e.g. cAMP-dependent protein kinase (PKA), mitogen-activated protein kinase) as well as phosphorylation-independent mechanisms (e.g. in some cases, the regulation of NHE1 by protein kinase C (PKC); Fliegel, 2001). The effects of these, and other, mechanisms vary between NHE isoforms: for example, while NHE1 activity, in some cells, does not appear to be regulated by the activity of the cAMP/PKA pathway (Borgese et al. 1992), the same signalling cascade inhibits NHE3 (and, possibly, NHE5) indirectly by influencing its interaction with NHERF (Hall et al. 1998; Attaphitaya et al. 2001; also see Szaszi et al. 2001) and stimulates PNHE (the NHE isoform found in trout red blood cells) by direct phosphorylation (Borgese et al. 1992; also see Pedersen et al. 2003 for a similar stimulatory effect on the NHE isoform expressed in winter flounder red blood cells). Moreover, the regulation of the activity of a given Na+/H+ exchanger isoform by mitogens and hormones may occur through diverse signal transduction pathways. In enteric endocrine and astrocytoma cells, for example, adrenaline (acting via P2 adrenoceptors) and somatostatin stimulate and inhibit, respectively, Na+/H+ exchange activity; despite the fact that these receptors are coupled to adenylate cyclase, the regulation of Na+/H+ exchange, in these cases, is independent of cAMP accumulation (Barber et al. 1989; also see Isom et al. 1987). In addition to their established role in pHj regulation, Na+/H+ exchangers also act to regulate [Na+]i. For example, basal permeability of cardiac myocytes to Na+ is, in part, 20 determined by Na+/H+ exchange activity (e.g. Frelin et al. 1984; Despa et al. 2002). NaVFT1"-exchanger-induced increases in [Na+]j have also been observed following imposed internal acid loads (e.g. Deitmer & Ellis, 1980, Vaughan-Jones, 1988) and in response to various hormones (e.g. Hou & Delamere, 2002; Cingolani et al. 2003) and these increases in [Na+]i can have several important consequences, from activating intracellular signalling cascades to regulating the activity of the Na+/Ca2+ exchanger (e.g. Hayasaki-Kajiwara et al. 1999; Trudeau et al. 1999; Mukhin et al. 2004). Na+/H+ exchangers also act as plasma membrane anchors for the actin-based cytoskeleton and, thereby, are involved in the control of cellular adhesion and migration (Putney et al. 2002). Considering these diverse properties, it is not surprising that Na+/H+ exchangers are involved in many physiological functions, including not only the regulation of pHj and [Na+]j but also the control of cellular volume and growth. 1.4.1.2. Expression patterns in nervous tissue Few studies have examined the expression of NHE isoforms in nervous tissue. Ma and Haddad (1997) and Douglas et al. (2001) used in situ hybridization and western blot analysis, respectively, to localize the distribution of NHE isoforms within various regions of rat brain. NHE1 was present in all regions of rat brain examined and, in a similar manner, NHE4 was also widely expressed, albeit at lower levels. The expressions of NHE2 and NHE3 were predominantly restricted to the cerebellum, although NHE3 expression was observed in some chemosensitive neurons of the brainstem (also see Wiemann et al. 1999). NHE5 is almost exclusively expressed in the brain (Baird et al. 1999; Szabo et al. 2000) and appears to accumulate in the dendrites and axons of cultured hippocampal neurons transiently transfected with NHE5 (Szaszi et al. 2002). Finally, the intracellular NHE isoforms, NHE6 and NHE7, have 21 also been found in brain tissue (Numata et al. 1998; Numata & Orlowski, 2001; also see Brett et al. 2002b). In the midst of this growing family of NaVH4" exchange proteins, the specific NHE isoform(s) present in rat hippocampal neurons remains unclear despite the fact that this cell type has been the subject of the most extensive studies of neuronal intracellular pH regulation within the mammalian central nervous system (see Chelser, 2003). In contrast to NHE 1-8 and functional Na+/H+ exchange activity found in other mammalian central neurons and glial cells (and, indeed, in mouse hippocampal neurons; e.g. Vornov et al. 1996; Bevensee et al. 1997; Pedersen et al. 1998; Yao et al. 1999), all of which can, to a greater or lesser extent, be pharmacologically inhibited by amiloride, amiloride analogues and benzoylguanidinium compounds (Tse et al. 1993; Yun et al. 1995), Na+/H+ exchange activity in rat hippocampal neurons is essentially insensitive to these compounds, with concentrations as large as 1 mM having no effect on antiport activity (Raley-Susman et al. 1991; Schwiening & Boron, 1994; Baxter & Church, 1996)3. Harmaline appears to be the only pharmacological inhibitor of Na+/H+ exchange activity in rat hippocampal neurons, although its poor selectivity for Na+/H+ exchange and its autofluorescent properties have restricted its use (see Raley-Susman et al. 1991; Schwiening & Boron, 1994; Baxter & Church, 1996). Although NHE4 mRNA has been detected in rat hippocampus and NHE4 exchange activity is relatively resistant to known pharmacological inhibitors, antiport activity apparently only contributes to acid extrusion under hyperosmotic conditions (-490 mOsm) and, as such, its contribution to cytosolic pHj regulation under more physiological conditions remains unclear (Bookstein et al. 1996; Chambrey et al. 1997a and b). 3 It is notable that, in rat cortical neurons, Ou-Yang et al. (1993) observed variable degrees of sensitivity to amiloride inhibition. 22 NHE5 mRNA has also been found in rat hippocampal neurons but, while this relatively amiloride-resistant isoform shares some functional similarities with Na+/H+ exchange activity in rat hippocampal neurons (Attaphitaya et al. 1999. and 2001; Szabo et al. 2000), distinct differences exist between the regulation of NHE5 activity (expressed in cell lines) and NaVH4" exchange activity in rat hippocampal neurons. For example, NHE5 activity is inhibited by activation of the cAMP/PKA pathway (e.g. Attaphitaya et al. 2001) whereas functional Na+/H+ exchange activity in rat hippocampal neurons increases upon activation of this signalling pathway (see below), an effect that is shared by the PNHE in trout red blood cells (Borgese et al. 1992) and a novel sperm-specific NHE (Wang et al. 2003a). It is apparent that further studies are required to elucidate the protein(s) contributing to functional NaVH4" exchange activity in rat hippocampal neurons; however, successful identification of the specific NHE isoform present in rat hippocampal neurons has been limited by the relative lack of specificity of available antibodies (see Hill et al. 2002). 1.4.1.3. Na+/H*" exchange activity in rat hippocampal neurons Under normoxic conditions at 37°C, the resting pHj of rat hippocampal neurons is maintained, at least in part, by NaVH* exchange activity (reviewed by Chesler 2003; also see Raley-Susman et al. 1991; Baxter & Church, 1996; Bevensee et al. 1996; Smith et al. 1998; and see Putnam 2001; Ou-Yang et al. 1993; Sanchez-Armass et al. 1994; Yao et al. 1999 for illustrations of similar findings in rat brainstem and neocortical neurons, rat brain synaptosomes and mouse hippocampal neurons, respectively). Although, as noted above, the identity of the NHE isoform(s) present in rat hippocampal neurons remains unclear, NaVlrE exchange activity in these cells shares a number of key characteristics with Na+/H+ exchange in other cell types. First, acid 23 extrusion via NaVH4" exchange is dependent on the presence of external Na+, with a Km value for external Na+ ions similar to that found for other NaVH*" exchangers (see Raley-Susman et al. 1991). Studies have suggested that Na+/H+ exchange activity in rat hippocampal neurons, as in other cell types, is electroneutral (Raley-Susman et al. 1991), although the reversal of NHE 1 activity has been associated with the development of a proton conductance (Demaurex et al. 1995; see Section 1.4.3). Second, extracellular Li+, but not A^-methyl-D-glucamine (NMDG+), is an effective external Na+ substitute (Kinsella & Aronson, 1981; Raley-Susman et al. 1991; Baxter & Church, 1996). Third, in common with NaVH1" exchangers in other cell types, NaVH4" exchange activity in rat hippocampal neurons exhibits an exquisite sensitivity to changes in pHj, presumably reflecting the presence of an allosteric intracellular IT/ modifier site (the 'H+ sensor'; see Section 1.4.1.1). Fourth, NaVH* exchange activity in rat hippocampal neurons is inhibited by reductions in pH0 and ambient temperature, consistent with observations in other cell types (Vaughan-Jones & Wu, 1990; Baxter & Church, 1996). Finally, as in non-neuronal cell types (reviewed by Fliegel, 2001; Hayashi et al. 2002), rat hippocampal neuronal Na+/H+ exchange activity can be regulated by a number of intracellular signalling cascades, including the cAMP/PKA (Smith et al. 1998) and Ca2+/calmodulin pathways (Church et al. 2001). It is important to note that the regulation of Na+/H+ exchange activity may differ between neurons in different areas of the brain (e.g. cerebellum vs. hippocampus) and even between different types of neurons in a given brain region (e.g. medulla). For example, in rat cerebellar granule cells, PKC is primarily involved in the regulation of Na+/H+ exchange (cf rat hippocampal neurons; Gaillard & Dupont, 1990) and Na+/H+ exchange activity in neurons of the ventrolateral medulla is more sensitive to inhibition by falls in pH0 than exchange activity in neurons located in the neighbouring inferior olive (Ritucci et al. 1997 and 1998). These data re-emphasize two key considerations about Na+/H+ exchange activity in rat hippocampal neurons: i) that NaVH* 24 exchange activity in rat hippocampal neurons can respond to changes in both the external (e.g. changes in pH0 and [Na+]0) and internal (e.g. changes in pHi and the activities of intracellular second messenger systems) microenvironments; and ii) that the control of NaVH4" exchange activity varies between isoforms and/or the cell type in which the specific isoform is expressed; thus, the control of Na^H4" exchange activity in rat hippocampal neurons cannot be predicted on the basis of findings in other cell types. 1.4.1.4. NaVH4" exchange: relevance to anoxia The contribution of Na+/H+ exchange to the regulation of pHj and [Na+]i during and following periods of ischemia has been extensively studied in cardiac myocytes (see reviews by Karmazyn, 1999; Avkiran, 2001). Key experiments performed by Karmazyn (1988) provided initial evidence that Na+/H+ exchange activity in the ischemic/reperfused heart contributes to cellular injury. Whether NaVH* exchange activity in cardiac myocytes remains functional during ischemia remains somewhat controversial (Hurtado & Pierce, 2001), but it is generally accepted that, in response to a marked intracellular acidosis and the accumulation of regulatory factors, such as catecholamines and lysophosphatidlycholine, Na+/H+ exchange activity in cardiac myocytes is markedly activated at the time of reperfusion (Avkiran & Haworth, 1999; Karmazyn et al. 1999). Although this acts to restore pHj, concomitant Na+ influx leads to reversal of the Na+/Ca2+ exchanger and an elevation of [Ca24-], (Fig. 1.4); however, it remains unknown whether the cardioprotective effects of NaVH4" exchange inhibitors are a result of limiting the recovery of pH; and/or Na+ entry (mediated by NaVH4" exchange) or subsequent Ca2+ entry (mediated by reverse Na+/Ca2+ exchange; Avkiran, 2001). Nevertheless, the relative resistance of NHE 1 null mutant mice to cardiac ischemia-reperfusion injury underscores the importance of NaVH4" exchange activity to the pathophysiology of cell death following cardiac ischemia (Wang et al. 2003b). 25 That NaVH4" exchange may play an analogous role in cerebral ischemia was initially suggested in studies by Vornov et al. (1996) in cultured rat neocortical neurons, where Na+/H+ exchange inhibitors were found to inhibit pHj recovery following periods of metabolic inhibition and improve neuronal survival. More recently, in cell types in which NaVH4" exchange activity is sensitive to pharmacological inhibitors, NaVH4" exchange inhibitors have been shown to reduce the extent of neuronal damage following periods of cerebral ischemia in vivo (Kuribayashi et al. 1999; Phillis et al. 1999) and following periods of oxygen-glucose deprivation or glutamate application in vitro (Horikawa et al. 2001a; Matsumoto et al. 2003). Thus, it is becoming increasingly apparent that changes in neuronal NaVH4" exchange activity may, at least in part, determine the extent of neuronal cell damage observed following periods of anoxia or ischemia. Studies performed in vivo or in slice preparations in vitro have suggested that changes in Na+/H+ exchange activity contribute to anoxia/ischemia-evoked changes in pH0 and pHj (Ohno et al. 1989; Obrenovitch et al. 1990; Pirttila & Kauppinen, 1992). Few studies, however, have examined the effects of anoxia or ischemia on NaVH4" exchange activity, either during or following the insult, under conditions in which the observed changes in NaVH4" exchange activity can be attributed to an intrinsic neuronal response to the period of anoxia or ischemia. It also remains unclear whether changes in NaVH4" exchange activity contribute to the potentially detrimental alterations in pHj and [Na+]j observed during and/or following anoxia/ischemia. In support, studies employing cultured postnatal rat hippocampal, cultured fetal mouse neocortical and acutely isolated mouse hippocampal neurons have illustrated that changes in Na+/H+ exchange activity occur in response to anoxia and contribute to the anoxia-evoked changes in pHj (Diarra et al. 1999, J0rgensen et al. 1999; Yao et al. 2001). 26 1.4.2. HCOT-dependent pH, regulating mechanisms The HCO3'-transporter family includes 10 related proteins with wide tissue distributions (Romero et al. 2004). According to their respective functions, these transporters can be classified into three groups: a family of electroneutral acid-loading anion exchangers (which, in forward-mode, exchange external CI" for internal HCO3"); electroneutral acid-extruding Na+-coupled CI7HCO3" transporters (which, in forward-mode, exchange internal CI" for external HCO3"); and electrogenic Na+/HC03" cotransporters which, depending on their transport stoichiometry and the prevailing membrane potential, can act as acid-loading or acid-extruding mechanisms. These diverse transport mechanisms share between 20 and 70% amino acid identity, exist as integral membrane proteins with 10-14 transmembrane domains, and share sensitivities to inhibition by disulfonic stilbene derivatives (Romero et al. 2004). 1.4.2.1. Neuronal HCO-f-dependent pH; regulation That pHj regulation in a given cell type may be dependent on the activities of more than one plasmalemmal transport mechanism was initially described in skeletal muscle (Aickin & Thomas, 1977) and, in the following years, in invertebrate (Moody, 1981; Schlue & Thomas, 1985) and vertebrate (Chelser, 1986) central neurons. It is now generally accepted that HCO3-dependent pHj regulating mechanisms act in concert with Na+/H+ exchange in the maintenance of pHj in many types of mammalian central neurons (reviewed by Chesler, 2003; also see Raley-Susman et al. 1991; Ou-Yang et al. 1993; Raley-Susman et al. 1993; Schwiening & Boron, 1994; Baxter & Church, 1996; Smith et al. 1998; Brett et al. 2002a). In rat hippocampal neurons, two HCCV-dependent pHj regulating mechanisms have been identified to date: i) a Na+-independent C1"/HC03" antiporter which, acting in forward-mode, transports HCO3" out of the cell, thereby acting as an acid loader (alkali extruder); and ii) a Na+-dependent C17HC03" 27 antiporter, which transports HCO3" into the cell in exchange for internal CI", thereby functioning as an acid extruder. The former mechanism can reverse under conditions of extreme intracellular acidosis to participate in acid extrusion (Baxter & Church, 1996). Both C17HCCV exchange proteins (Na+-independent and Na+-dependent) have been found in brain homogenates (Kobayashi et al. 1994; Wang et al. 2000; Grichtchenko et al. 2001) and distinct neuronal populations of the brain (e.g. hippocampus, cerebellum and cortex; Douglas et al. 2003; Giffard et al. 2003). In addition, both are sensitive to inhibition by 4,4'-diisothiocyanatostilbene-2,2'-disulfonate (DIDS; Schwiening & Boron, 1994; Baxter & Church, 1996). There is less functional evidence for the contribution of NaVHCCV cotransport to pHj regulation in mammalian central neurons (see Pocock & Richards, 1992; Schwiening & Boron, 1994; Baxter & Church, 1996; Bevensee et al. 2000; Schmitt et al. 2000); however, recent studies have demonstrated electrogenic Na+/HC0y cotransporter expression in rat brain (possibly located on neuronal processes; see Giffard et al. 2000; Schmitt et al. 2000) and, in mouse hippocampal neurons, Na+/HCOy cotransport may be activated in response to anoxia (Yao et al. 2003). Studies examining the regulation of the activities of C17HCOy exchangers in hippocampal neurons are limited (see Brett et al. 2002a). However, in a manner similar to Na+/H+ exchange, C17HCOy exchangers in non-neuronal cell types respond to a variety of extracellular and intracellular factors, such as changes in pH0 and the activities of various intracellular second messenger systems (e.g. Boron et al. 1979; Vigne et al. 1988; Ludt et al. 1991). Importantly, the regulation of the activity of a given OVHCOy exchanger differs from cell type to cell type and, furthermore, Na+/H+ exchange and ClVHCOyexchange activities in a given cell type are often independently regulated. In osteoblasts, for example, a rise in [Ca2+]j stimulates Na+-independent C17HCOy exchange but has no effect on Na+/H+ exchange activity whereas a rise in intracellular cAMP inhibits both exchangers (Green & Kleeman, 1992). In 28 acutely isolated adult rat hippocampal CA1 pyramidal neurons, NaVrT*" exchange activity is increased while Na+-dependent CT/HCO3" exchange activity is decreased upon activation of the cAMP/PKA pathway (Smith et al. 1998; Brett et al. 2002a). To complicate matters further, the control of CT/HCO3" exchange activities (Na+-dependent and Na+-independent) by the cAMP/PKA pathway in rat hippocampal neurons is dependent on resting pHj values; for example, in neurons with low resting pH/values (pHi < -7.20), PKA activation increases Na+-independent CI/HCO3" exchange activity and decreases Na+-dependent CT/HCO3" exchange activity, whereas opposite effects are seen in neurons with high resting pHj values (pHj > -7.20; Brett et al. 2002a). 1.4.2.2. HCOy-dependent pHj regulation: relevance to anoxia Acting alongside Na+/H+ exchange, HCO3"-dependent transport mechanisms may contribute to the neuronal pHj and [Na+]i (and [Cl"]j) responses to anoxia or ischemia and may, in turn, be important determinants of the extent of cell damage observed. In cardiac myocytes, electrogenic NaVHCOy cotransport may augment the increases pH; and [Na+]i observed in response to anoxia or ischemia (see Lemars, 2001). With the use of a neutralizing antibody, Khandoudi et al. (2001) demonstrated that, in isolated rat hearts, inhibition of electrogenic NaVHCOy cotransport reduced cellular damage and improved functional recovery following an ischemic insult. On the basis of NMR-based pHj measurements in hippocampal slices, Pirttila and Kauppinen (1994) similarly suggested that anoxia caused changes in the activities of HC03"-dependent pHj regulating mechanisms, although this study was unable to differentiate between the possible contributions of neuronal vs. glial elements to the results obtained. In isolated rat hippocampal neurons, Na+-dependent CI7HCO3" exchange activity leads to increases in [Na+]i and pH; under normoxic condititions (Rose & Ransom, 1997; Brett et al. 2002a) and, if electrogenic Na+/HC03" 29 cotransporters are expressed by neurons, membrane depolarizations observed during (and possibly following) anoxia may enhance inward Na+/HCOy cotransport activity that, in turn, may serve to increase [Na+]j, pHj and hyperpolarize the membrane potential. In contrast, however, NaVHCOy cotransport (with a 1:3 stoichiometry) in mouse hippocampal neurons is reportedly activated during anoxia and appears to contribute to anoxia-induced falls in pHj and membrane depolarizations (Yao et al. 2003). It is similarly unclear whether CT/HCO3" exchangers contribute to the production of the anoxia-, ischemia- and excitotoxin-induced increases in [Cl"]i that have been measured in rat hippocampal and neocortical slices (Rothman, 1985; Jiang et al. 1992; Inglefield & Swartz-Bloom, 1998a and b). This uncertainty reflects that fact that CT /HCO3" exchangers act in concert with a diverse family of channels (e.g. volume-sensitive anion channels), plasmalemmal pumps (e.g. Cl"-ATPases) and other transporters (e.g. Na+/K+/2C1" cotransport) to maintain neuronal CI" homeostasis (see Vaughan-Jones, 1979; Aickin & Brading, 1984; Kaila, 1994; Irie et al. 1998). In support of the contribution of CIVHCCV exchange to the neuronal response to anoxia or ischemia, DIDS reduces the extent of neuronal damage in cultured cortical neurons following periods of oxygen-glucose deprivation (Tauskela et al. 2003), protects against excitotoxic damage in chick retinal cells (Zeevalk et al. 1989), reduces apoptotic cell death in cultured cerebellar granule cells (Franco-Cea et al. 2004), prevents ouabain-induced release of glutamate (Estevez et al. 2000) and delays the onset of hypoxic depolarization (Muller, 2000). However, few studies have carefully examined the mechanisms underlying these effects (see Tauskela et al. 2003) and, as such, DIDS may be acting to inhibit not only Na+-dependent and Na+-independent C17HCOy exchangers but also a range of HCCV-independent processes that may be of importance in modulating anoxic or ischemic cell death (e.g. chloride channels, K+/C1" transport, 30 glutamate uptake and mitochondrial release of free radicals; see Han et al. 2003; Malek et al. 2003; Tauskela et al. 2003). Although Na+-dependent and Na+-independent CT/HCO3" exchange activities are capable of influencing pHj, [Na+]i and [CT]i in neurons under normoxic conditions, further experiments are required to examine the potential contributions of these transport mechanisms to the changes in pHj and [Na+]j observed in hippocampal neurons in response to anoxia. 1.4.3. Additional pH, regulating mechanisms Although Na+/H+ exchange and CI7HCO3" exchangers (Na+-dependent and Na+-independenf) have been formally identified as important pHj regulating mechanisms in rat hippocampal neurons, other mechanisms may also contribute (see Fig. 1.3). In the following paragraphs, I will briefly outline some of these potential mechanisms and discuss their possible relevance to the neuronal response to anoxia. First, there is a close relationship between pHj and [Ca ]j, an appreciation of which has led to the development of techniques for the concurrent measurement of pHj and [Ca2+]j using microspectrofluorimetry or ion-selective microelectrodes (ISMs; e.g. Martinez-Zaguilan et al. 1991 and 1996; Wiegmann et al. 1993; Austin et al. 1996). In addition to H+ and Ca2+ ions competing for common intracellular binding sites, Ca , both directly and via intracellular second messenger cascades, can regulate the activities of pHj regulating mechanisms and, thus, pH, (Vaughan-Jones & Wu, 1990; Sanchez-Armass et al. 1994; see also Gordienko et al. 1996). It is also perhaps not surprising that one or more mechanisms exist to regulate in concert the internal concentrations of both ions. One such mechanism, initially described in snail neurons, is the Ca2+,H+-ATPase which extrudes intracellular Ca2+ ions in exchange for extracellular protons 31 (Schwiening et al. 1993). This mechanism also exists in rat hippocampal CA1 (Trapp et al. 1996b) and rat cerebellar granule (Wu et al. 1999) neurons and, in the latter, contributes to glutamate receptor-mediated intracellular acidosis (also see Irwin et al. 1994; Wang et al. 1994). Ischemia or anoxia may affect the activity of this ATPase, given the facts that these insults lead to rises in [Ca ]\ and an external acidosis, both of which would act to enhance to its activity (see Schwiening et al. 1993; Ou-Yang et al. 1994a). In contrast, other studies have found that Ca2+,H+-ATPase activity is inhibited during metabolic insults (e.g. Kass & Lipton 1989; Pereira et al. 1996; Castilho et al. 1998; Wu et al. 1999; Zaidi & Michaelis 1999; Chinopoulos et al. 2000). Second, in a variety of cell types, including snail neurons, a voltage-activated YC conductance (gH+) contributes to the recovery of pHj following intracellular acid loads imposed during membrane depolarization (Meech & Thomas, 1987; Byerly & Suen, 1989; Kapus et al. 1993; Gordienko et al. 1996). gH+s demonstrate an extremely high selectivity for protons and, upon activation, are capable of producing dramatic shifts in pH, without demonstrating rapid current inactivation or desensitization (Meech & Thomas, 1987; Kapus et al. 1993; Lukacs et al. 1993; DeCoursey & Cherny, 1994a). Although the existence of a gH+ in mammalian central neurons has not been investigated, it is possible that it could contribute to acid efflux during anoxic or ischemic insults which are associated with prolonged membrane depolarizations. In addition, as noted above, evidence indicates that gH+s can couple to Na+/H+ exchange (Demaurex etal. 1995). Third, FT^K*- ATPases and LL-ATPases represent members of a diverse family of P- and V-type ATP-driven cation transporters: the former are primarily found in the stomach, colon and kidney (van Driel & Callaghan, 1995) while the latter have been primarily characterized in 32 cardiac myocytes, renal epithelia and osteoclasts (Kurtz, 1987; Nelson & Klionsky, 1996). Both types of ATPases contribute to pHj regulation and, in the case of H4"-ATPases, may act to limit Na+/H+ exchange-induced Ca2+ overload observed during metabolic inhibition in cardiomyocytes (Karwatowska-Prokopcauk et al. 1998). Although Bevensee et al. (1996) speculated that H+-ATPases may contribute to NaVindependent acid extrusion from rat hippocampal neurons, this possibility has not been formally examined (see Yoshinaka et al. 2004 for an illustration of V-type ATPases localized to rat brain synaptic vesicles). Finally, although neurons are primarily metabolically aerobic, during periods of intense neuronal activity or periods of ATP depletion, anaerobic metabolism becomes the key, albeit less efficient, energy-producing pathway and, as a result, lactic acid accumulates intracellularly (Ljunggren et al. 1974; Hope et al. 1988; Jarolimek et al. 1989). Given a pKa of ~ 3.9, at physiological pHi values lactic acid will exist primarily in its anionic form, limiting its ability to cross the lipid membrane (but see Dringen et al. 1995). Characterized in peripheral cell types and limited populations of mammalian central neurons, a lactate/H+ cotransport mechanism removes intracellular lactate and, in this way, may contribute to pH, regulation (Assaf et al. 1990; Nedergaard & Goldman, 1993; Juel, 1997). Although studies in non-neuronal cell types support a role for this mechanism in contributing to the recovery of pHj following transient periods of anoxia (Vandenberg et al. 1993), studies in mammalian central neurons to date have failed to document a similar role (see Fujiwara et al. 1992; Diarra et al. 1999). 1.5. Synthesis and objectives The contribution of changes in pHj and [Na+]j to the pathophysiology of ischemic cell damage has been best investigated in non-neuronal cell types, notably cardiac myocytes and hepatocytes. 33 In these cell types, Na+/H+ exchange activity is activated at the time of reperfusion and, although this acts to restore pHi, concomitant Na+ influx leads to reversal of the Na+/Ca2+ exchanger and an elevation of [Ca2+]-,. Nevertheless, it remains unknown whether the cardioprotective actions of NaVLE1" exchange inhibitors result from limiting the rate of recovery of pHi, reducing internal Na+ loading and/or decreasing the subsequent entry of Ca2+. Despite the fact that changes in pHj and [Na+]j play important roles in the pathophysiology of anoxic and ischemic neuronal death, much less is known about the changes in pHj and [Na+]j that occur in mammalian central neurons during or following periods of anoxia or ischemia, and few studies have examined the role of transport mechanism(s) in the production of the changes in pHj and [Na+]j observed. Thus, the overall aims of the present study were to characterize the changes in pHj and [Na+]i which occur in response to transient periods of anoxia in isolated rat hippocampal neurons and to assess the role of a variety of mechanisms, especially Na+/H+ exchange, in the production of the ionic changes observed. Experiments were performed using isolated rat hippocampal neurons in order to isolate the intrinsic changes in pHj and [Na+]j that occur in response to anoxia (the choice of experimental preparations employed in these studies is discussed further in Section 2.0.1). The principal objectives of the present study are: 1) To characterize the changes in pHj and [Na+]j which occur in isolated rat hippocampal neurons during and following transient anoxic insults and to assess the role of changes in NaVrE exchange activity to the changes in pHj and [Na+]j observed at these times (Chapters 3, 4 and 5). 34 2) To examine further the contribution of NaVH4" exchange to anoxia-evoked changes in pHj and [Na+]i in isolated hippocampal neurons by developing a microspectrofluorimetric technique for the concurrent measurement of both ions (Chapter 6). 3) To explore systematically the potential contributions of other mechanisms integral to the cell to the changes in [Na+]j observed during and following anoxia observed in isolated rat hippocampal neurons (Chapter 7). These studies were driven by the contention that an understanding of the fundamental cellular mechanisms that contribute to anoxia-evoked changes in pH, and [Na+]j in mammalian central neurons may provide novel insights into the pathogenesis of anoxic/ischemic cell death. 35 Table 1.1: Clinical trials of selected agents in acute stroke Proposed principle mechanism of action Drug name Trial Status Glutamate receptor antagonist Voltage-activated Ca2+ channel antagonist Voltage-dependent K+ channel agonist Na+ channel antagonist GABAA receptor agonist YM872 ZK-200775 CGS 19755 Aptiganel Dextrorphan Dextromethorphan Magnesium Remacemide ACEA 1021 GV 150526 Eliprodil Nimodipine Flunarizine BMS-204352 Fosphenytoin Clomethiazole Phase II: ongoing Phase II: abandoned Phase III: no efficacy Phase III: no efficacy Phase II: abandoned Abandoned Phase III: no efficacy Phase III: borderline efficacy Phase I: abandoned Phase III: ongoing Phase III: abandoned Phase III: no efficacy Phase III: no efficacy Phase III: no efficacy Phase III: no efficacy Phase III: no efficacy Table adapted from Lee et al. 1999. 36 Fig. 1.1. Pathways of ischemic cell death. This figure, taken from Lipton (1999), illustrates some of the major events that are hypothesized to contribute to ischemic cell death and also illustrates the extremely complex interactions between these events. Column 5 lists five principal morphological forms taken by dying or dead cells after an ischemic insult. Determining how these end stages are reached is the ultimate goal of research on ischemic cell death. Column 4 lists six critical functional or structural changes, all of which appear to occur as a result of ischemia. Column 3 lists actions that are likely to cause the long-term functional changes described in column 4. These are termed "perpetrators" because they are considered to be key damaging events in ischemic cell death. No direct effects of the perpetrators listed on the critical functional changes, shown in column 4, are implied, because none has been completely established. Columns 2 and 1 show changes in some of the many variables initiated by anoxia or ischemia, the most important end result of which is considered to be the activation of perpetrators, but which may also have more direct effects on cell viability. Events located within the same toned horizontal bands are linked by direct causal interactions. Causal interactions are also indicated by including changes in column 1 within a box whose outline color is same as that of the variable they are changing (shown in column 2). Abbreviations: Depol, depolarization; pHj, intracellular pH; Nai, intracellular Na+; Cai, intracellular Ca2+; P, permeability; FFA, free fatty acids; PAF, platelet-activating factor; e Transport, electron transport; A carriers, changes in the activities of transport mechanisms. 37 Pathways of Ischemic Cell Death Induction of Long-Term Functional Damage 1 Initiators and Activators -ur Perpetrators t Glutamate • Ca, Map Kinase • •"Transport t Glycolysis A Carriers 11-. 1 for • Ca, tGlutamate 4 Nitric Oxide "i 1 Gene Activation J 1 tCa, ^ 1 Free Radicals ] • tpH I J - fFfA — •feTransport Protease Activity (Calpain) Proteolysis Critical Functional & Structural Changes • Ca, f Free Radicals -0* Nitric Oxide Peroxynitrite Free Radical Action & Peroxynitrite A Proteins & Phospholipids & DNA Phospholipase Activity Phospholipid Changes PolyADPribose Polymerase Permeability ilsla/K ATPase Mitochondrial Dysfunction »Protein Synthesis Cytoskeletai Prolonged Changes in Kinases or Phosphatases Cell Death End Stages I H 1 Edematous Cell Change Ischemic Cell Change Homogenizing Cell Change Apoptotic Cell Change Autophago-cytotic Cell Change • • 38 Fig. 1.2. A schematic illustration of the pattern of ionic and electrical changes induced by anoxia or ischemia in mammalian central neurons (adapted from Martin et al. 1994). Phase P. changes in membrane potential (Vm) usually begin shortly following the onset of anoxia or ischemia and neurons may respond with a slight membrane depolarization (upper record) or hyperpolarization (lower record), depending largely on the type of neuron from which recordings are made. Decreases in both pH0 and pHj precede changes in Vm and the external and internal concentrations of other ions. As internal ATP levels decline, Na+,K+-ATPase activity becomes compromised: [Na+]i and [K+]0 increase and further promote the gradual membrane depolarization seen after the initial transient depolarization or hyperpolarization. It is notable that the onset and magnitude of the electrical and ionic changes observed in Phase 1 vary between brain regions and even between different neurons within a given brain region (e.g. Leblond & Krnjevic, 1989; Cowan & Martin, 1992). Phase 2: rapid depolarization (i.e. 'anoxic depolarization') with accompanying large changes in external and internal ion concentrations. Phase 3: without reperfusion, membrane repolarization will not occur and the internal and external concentrations of ions will not be restored. The period of time over which this sequence of events occurs depends on the nature and severity of the insult. 39 PHASE 1 Initial change in membrane potential (either small depolarization or hyperpolarization) Gradual changes in external and internal ion concentrations: e.g. slow rises in [K+L. and [Na+J Membrane repolarization and restoration of ion homeostasis dependent on the return of adequate supplies of oxygen and glucose Initiation of anoxia or ischemia 1 PHASE 2 Rapid depolarization Large changes in the external concentrations of ions: falls in [Ca2*],,, [Na+]0, and [CtL and rises in[K*L Large changes in the internal concentrations of ions: rises in [Ca2+]i, [Na+]|, and [CH and falls in W 40 Fig. 1.3. An illustration of the pHj regulating mechanisms present in rat hippocampal neurons (adapted from Schwiening, 2002). Two electroneutral mechanisms contribute to acid extrusion. First, a Na+/H+ exchanger which, in rat CA1 neurons, is insensitive to pharmacological inhibition with amiloride, amiloride analogues or benzoylguanidinium compounds (e.g. HOE 694). Second, a Na+-dependent C17HC03~ exchanger which is sensitive to stilbenes such as DIDS. As in snail neurons, a putative gH+ may also contribute to acid extrusion under depolarizing conditions. Two acid-loading mechanisms have also been identified. First, a DIDS-sensitive Na+-independent C17HC03" exchanger which is the primary means by which CA1 neurons recover from internal alkaline loads. Second, a plasmalemmal Ca2+,H+-ATPase functions to extrude internal Ca ions and, in doing so, generates a fall in pH;. Although distinct isoforms of mammalian Na+/H+ and Na+-independent C17HC03" exchangers have been identified and the first mammalian members of the electroneutral Na+-driven C17HC03" exchanger family have been cloned, the precise molecular identities of the exchangers present in rat hippocampal CA1 neurons remains unknown. 41 42 Fig. 1.4. The contribution of NaVH4" exchange to myocardial injury induced by ischemia (adapted from Sheldon & Church, 2002b). (1) Before ischemia, NaVH4" exchange and Na+/Ca2+ exchange are operating to extrude H* and Ca2+ ions, respectively. Na+,K+-ATPase activity is functional and acts to maintain an inwardly directed Na+ gradient. NHE, NaVH4" exchange, NCX, Na+/Ca2+ exchange operating in forward-mode. (2) In response to periods of ischemia/reperfusion, there is an inhibition of Na+,K+-ATPase activity and an activation of NaVH4" exchange activity; this combination allows intracellular Na+ ions to accumulate. As noted in the text, several, although not all, studies have suggested that Na+/H+ exchange activity in cardiac myocytes is inactive or minimally active during ischemia. It is acknowledged, however, that NaVH4" exchange is activated immediately upon reperfusion. (3) The resulting increase in intracellular Na+ ions helps to drive Na+/Ca2+ exchange activity into reverse, importing Ca2+ ions. rNCX, Na+/Ca2+ exchanger operating in reverse-mode. (4) Through a variety of intracellular cascades, the increase in intracellular Ca2+ initiates cellular damage and death. These cascades include Ca2+-dependent proteases, phospholipases and the generation of free radicals. The thin arrow extending from NaVH4" exchange to cell death indicates that, although the mechanisms are poorly understood, increased NaVH4" exchange activity can contribute to cellular damage independent of changes in intracellular Ca2+. This may include the damaging effects of increases in intracellular Na+ and intracellular pH. 44 CHAPTER TWO GENERAL METHODS 2.0. CELL PREPARATION 2.0.1. Choice of experimental preparations It has long been known that changes in neuronal pHj occur during and following anoxia or ischemia in vivo and in slice preparations in vitro (for reviews see Erecinska & Silver, 1994; Siesjo et al. 1996; Lipton, 1999); however, it is difficult under these experimental conditions to separate the contribution of various cell types, including glia, to the changes observed from volume-averaged measurements and additional confounds, such as concurrent changes in pH0, [K+]0 and neurotransmitter release (each of which can affect steady-state pHj (and [Na+]i) and the activities of pHj regulating mechanisms), complicate the characterization of underlying mechanisms (see Erecinska & Silver, 1994; Pirttila & Kauppinen, 1994). In this regard, isolated neurons offer a distinct advantage and not only are hippocampal neurons, in particular, vulnerable to the effects of anoxia or ischemia but also they have been the subject of the most extensive studies of pHj regulation in any type of mammalian central neuron (see Chesler, 2003). Thus, experiments were performed using isolated rat hippocampal neuronal preparations. Because the sensitivity to anoxic/ischemic cell damage (e.g. Rothman, 1983; Di Lorteo & Balestrino, 1997) and the mechanisms that regulate pHj (e.g. Bevensee et al. 1996) are developmentally regulated, where possible, experiments were performed employing acutely isolated adult rat hippocampal CA1 pyramidal neurons in preference to cultured rat hippocampal neurons. Differences exist between the ischemic situation in vivo compared to isolated neuronal preparations (e.g. changes in extracellular fluid volume and composition; Lipton, 1999); however, similar to measurements made in vivo and in slice preparations in vitro (see Section 1.3), isolated neurons exhibit elevations in [Ca ]\ and [Na ], and membrane depolarizations in 45 response to periods of anoxia or oxygen and glucose deprivation, and these insults lead to subsequent cell death (e.g. Goldberg & Choi, 1993; Friedman & Hadddad, 1994a; Chen et al. 1999; Diarra et al. 1999; Mazza et al. 2000; Fernandes, 2001; Aarts et al. 2003). 2.0.2. Acutely isolated adult rat hippocampal CA1 pyramidal neurons Where possible, experiments were performed using acutely isolated adult rat hippocampal CA1 pyramidal neurons. Methods used to isolate these neurons were modified from techniques developed by Kay and Wong (1986) and Mody et al. (1989). Male Wistar rats (200 - 260 g) were obtained from The Animal Care Center (University of British Columbia) and housed under conditions of controlled temperature (20 - 22°C) and lighting (lights on 0600 - 1800). Food (Lab Diet, PMI Feeds Inc., St. Louis, MO) and water were available ad libtium. All procedures conformed to guidelines established by the Canadian Council on Animal Care and were approved by The University of British Columbia Animal Care Committee. Animals were anesthetized with 3% halothane in air and decapitated. Brains were removed rapidly and placed in ice-cold (4 - 8°C) HCO3"-containing medium previously equilibrated with 5%C02/95%02 (Solution 1; Table 2.1). One of the hippocampi was separated from the surrounding tissue, transverse hippocampal slices (450 um) were obtained with a Mcllwain tissue chopper and collected in ice-cold HC03"/C02-buffered medium. The slices were transferred to an incubation chamber containing HC03"/C02-buffered medium (at 32°C) and were allowed to recover for at least 1 h. Three hippocampal slices were then enzymatically digested at 32°C in 2 ml of HC03"/C02-buffered medium containing 1.5 mg ml"1 pronase (protease type XIV bacterial from Streptomyces griseus; Sigma-Aldrich Canada Ltd., Oakville, ON). After 30 min, the CA1 regions were removed under a dissecting microscope and triturated with fire-polished Pasteur 46 pipettes of diminishing tip diameters (0.7, 0.5, 0.3 and 0.2 mm) in 0.5 ml of standard loading medium, pH 7.35 at room temperature (Solution 3; Table 2.1). The triturated suspension was deposited onto a cleaned glass coverslip mounted in a temperature-controlled perfusion chamber so as to form the floor of the chamber. Neurons were allowed to adhere to the substrate (i.e. coverslip) for 15 min at room temperature prior to loading with a fluorophore. Freshly isolated hippocampal CA1 pyramidal neurons were chosen for study based on morphological criteria established by Schwiening and Boron (1994), i.e. a smooth, non-granular appearance; a single major process (presumably an apical dendrite) projecting from one pole of the soma which was at least three times (typically >5 times) the length of the diameter of the cell body; and the presence of two or more smaller processes (basal dendrites) at the opposite pole. 2.0.3. Postnatal rat hippocampal neuronal cultures Due to the prolonged time required to load the acetoxymethyl ester (AM) form of the Na+-sensitive dye sodium-binding benzofuran isophthalate (SBFI; see Section 2.3.1), cultured postnatal rat hippocampal neurons were employed in the majority of experiments examining the changes in [Na+]i which occur in response to transient periods of anoxia. Primary cultures of hippocampal neurons were prepared from 2-4 day old postnatal Wistar rats. Rat pups were anesthetized and decapitated. Brains were removed rapidly and collected in ice-cold Leibovitz L-15 medium (Invitrogen Canada Inc., Burlington, ON) supplemented with 34 mM glucose (L-15/G). Hippocampi were removed, collected in ice-cold L-15/G and then incubated for 15 min at 37°C in L-15/G medium containing 1 mg ml"1 papain (from Papaya Latex; Sigma Chemical Co.) and 25 pg ml"1 DNAse (type II from Bovine pancreas; Sigma Chemical Co.). Afterwards, the L-15/G medium was discarded and replaced with Dulbecco's Modified Eagle Medium F-12 47 (Invitrogen Canada Inc.) supplemented with 29 mM NaHC03 and 10% fetal bovine serum (pH 7.4 at 37°C after equilibration with 5% CO2; Sigma Chemical Co.). Hippocampi were then mechanically dissociated using fire-polished Pasteur pipettes of decreasing tip diameters. A hemocytometer chamber was used to count the number of cells within a sample of the cell suspension and a dilution factor was calculated in order to plate neurons at a density of 3 - 8 x 10 neurons cm" onto 18 mm glass coverslips. Coverslips were coated with poly-D-lysine (100 pg ml"1; Sigma Chemical Co.) and laminin (16.7 pg ml"1; Sigma Chemical Co.). Neurons were allowed to adhere to substrate for 2 h before coverslips were transferred into 12 well plates. After 24 h, the growth medium was fully changed to Neurobasal Medium A (Invitrogen Canada Inc.) supplemented with B-27 Supplement (Invitrogen Canada Inc.), 0.5 mM glutamine (Invitrogen Canada Inc.), 100 U ml"1 penicillin (Sigma Chemical Co.) and 100 pg ml"1 streptomycin (Sigma Chemical Co.). The cultures were fed every 4-5 days by half-changing the existing medium with fresh Neurobasal Medium A. Glial proliferation was inhibited 48 h after initial plating by adding 10 pM cytosine-(3-D-arabinofuranoside hydrochloride (Sigma Chemical Co.). Each coverslip consisted primarily of hippocampal neurons with a maximum of 15% cells being glial. Neuronal cultures were used 6-14 days after plating. 2.1. SOLUTIONS AND TEST COMPOUNDS The compositions of the HC03"/C02-buffered media and nominaliy-HC03"/C02-free, Hepes-buffered media commonly used in experiments are detailed in Table 2.1. Normoxic HCO37CO2-buffered solutions were equilibrated with 5% C02/95% air, giving a final pH value of 7.35 (at 37°C); during perfusion with these media, the atmosphere in the recording chamber contained 5% C02/95% air. Hepes-buffered saline was titrated to pH 7.48 (at room temperature; 22°C) in 48 order to achieve a final pH of 7.35 - 7.36 at 37°C. Experiments were performed at 37°C, unless otherwise noted. When external Na+ was reduced to 2 - 4 mM, A^-methyl-D-glucamine (NMDG+) or Li+ were employed as substitutes in Hepes-buffered media and solutions were titrated to pH 7.35 with 10 M HC1 or 2 M LiOH, respectively (Solutions 4 and 5; Table 2.1). Given the use of sodium dithionite to induce anoxia (see below) and the need to maintain [Na+]0 constant during an experiment, external Na+-free media could not be employed. Nevertheless, 2-4 mM Na+0 is considerably less than the apparent Km of Na+/H+ exchange in rat hippocampal neurons for external Na+ (Km = 23 - 26 mM; Raley-Susman et al. 1991) and rates of acid extrusion from rat hippocampal neurons in the complete absence of external Na+ are not influenced by the addition of 2 - 4 mM Na+ (C. Brett, C. Sheldon and J. Church, unpublished observations). In experiments in which Na+-free, Hepes-buffered media were employed, NaCl and NaHaPO^ were omitted and NMDG and/or KC1 were employed as substitutes; solutions were titrated to pH 7.35 with 10 M HC1 or KOH, respectively. For Ca2+-free media, CaC^ was omitted, [Mg2+] was increased to 3.5 mM and 200 uM ethylene glycol-bis(f3-aminoethyl ether) N, N, JV, /V-tetraacetic acid (EGTA) was added. Solutions containing 20 - 40 mM NH4CI were prepared by equimolar substitution for NaCl. In HC03"-free solutions containing Ni2+, Zn2+ or Gd3+, MgS04 was replaced with MgCb and NaH2P04 was omitted (see Caldwell et al. 1998). Corning 240 and 440 pH meters (Corning Inc., Corning, NY), calibrated daily, were utilized to measure the pH of all solutions. A list of pharmacological agents used in the studies is presented in Table 2.2. Unless otherwise noted, test compounds were obtained from Sigma-Aldrich Canada Inc. 2',5'-dideoxyadenosine (DDA) was obtained from Biomol Research Laboratories Inc. (Plymouth Meeting, PA). The Rp- isomer of adenosine-3',5'-cyclic monophosphorothioate (i?p-cAMPS, 49 Na+ salt) was obtained from Biolog Life Science Institute (La Jolla, CA). Arachidonyltrifluoromethyl ketone (AACOCF3) was obtained from Calbiochem (San Diego, CA). (55',10/?)-(+)-5-methyl-10,l l-dihydro-5//-dibenzo[a,d]cyclohepten-5,10-imine maleate (MK-801), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, disodium salt), 2-[2-[4-(4-nitrobenzyloxy)phenyl] ethyl] isothiourea mesylate (KB-R7943) and 7-chloro-5-(2-chlorophenyl)-l,5-dihydro-4,l-benzothiazepin-2(37/)-one (CGP-37157) were obtained from Tocris Cookson Inc. (Ellisville, MO). Bafdomycin A\, omeprazole and 2-methyl-8-(phenylmethoxy)imidazo[l,2-a]pyridine-3-acetonitrile (SCH-28080) were generous gifts from Dr. V. Palaty, AstraZeneca and Schering Canada Inc., respectively. 2.2. INDUCTION OF ANOXIA In the great majority of experiments, anoxia was induced by the addition of 1 - 2 mM sodium dithionite (Na2S204), an O2 scavenger, to the superfusing medium (see Friedman & Haddad 1993; Nowicky & Duchen, 1998; Diarra et al. 1999; Yao et al. 2001 and 2003; Paquet-Durand & Bicker, 2004). Dithionite-containing media (50 ml) were prepared fresh prior to every experiment in which neurons were exposed to anoxia and solutions were equilibrated with 100% argon (Ar; in Hepes-buffered media) or 95% Ar/5% C02 (HC037C02-buffered media) for 10 -15 min immediately prior to use. During anoxia, the atmosphere in the recording chamber was switched from room air to 100% Ar (Hepes-buffered media) or from 95% air/5% CO2 to 95% Ar/5%) CO2 (HC037C02-buffered media). The P02 in media containing 1- 2 mM sodium dithionite was measured with a Radiometer ABL 500 blood gas analyzer calibrated for low P02 values; in samples obtained anaerobically from the recording chamber, P02 was <1 mm Hg (n = 6). Similar P02 values were measured during experiments in which an oxygen electrode (ISO2; 50 World Precision Instruments Inc., Sarasota, FL) was placed in the recording chamber. Dithionite anions (S2O4 "), or the associated SO2" monomers, act to reduce soluble 02 and in doing so produce S042", S032" and H20 (see Lambeth & Palmer, 1973; Camacho et al. 1995). By products of these reactions (see Camacho et al. 1995) may account for the ability of sodium dithionite to influence pulmonary vasoconstriction (Archer et al. 1995) and catecholamine release (Carpenter et al. 2000) in a manner independent of its ability to reduce P02. Therefore, control experiments were performed to verify that the observed changes in pHj and [Na+]i evoked by exposure to solutions containing sodium dithionite were related to its O2 scavenging property (see Chapters 3 and 5). 2.3. MICROSPECTROFLUORIMETRY All ion-sensitive fluorescent probes were obtained from Molecular Probes Inc. (Eugene, OR). In the majority of experiments, pHj measurements were obtained with the dual-excitation ratiometric fluorophore, 2',7'-bis-(2-carboxyethyl)-5(6)-carboxyfluorescein (BCECF), or either of the dual emission seminaphthorhodafluor ratiometric indicators, carboxy SNARE-1 or SNARF-5F carboxylic acid. In some experiments, 8-hydroxypyrene-l,3,6-trisulfonic acid (HPTS) was used to measure pHj. Changes in [Na ]\ and [Ca ]\ were measured using the dual-excitation fluorophores SBFI and fura-2, respectively. Details of the techniques used for dye loading, dye calibration and the conversion of ratio values to ion concentrations (i.e. pHj, [Na ]\ and [Ca ]j) for BCECF, HPTS, SBFI and fura-2 are presented in the following Sections 2.3.1 - 2.3.3. Details regarding the use of carboxy SNARF-1 and SNARF-5F for measurements of pHj are provided in Chapter 6. 51 2.3.1. Dye loading Fluorophores, except HPTS (see below), were loaded into neurons in their AM esters form. AM esters are hydrophobic and uncharged, allowing passage across plasma membranes; upon entry into cells, AM esters are hydrolysed by intracellular esterases to produce the hydrophilic, polyanionic free acid forms of the fluorophores which become trapped intracellularly. BCECF-AM and SBFI-AM were prepared as 1 and 5 mM stock solutions, respectively, in DMSO and were stored at -60°C. Fura-2-AM was dissolved in chloroform and divided into 30 pi aliquots (1 pg pi"1) after which the chloroform was removed by vacuum evaporation. Fura-2-AM-containing vials were stored at -60°C and, on the day of use, fura-2-AM was prepared as a 1 mM stock in anhydrous DMSO. Acutely isolated neurons were loaded with BCECF or fura-2 by incubation with 2 pM BCECF-AM for 15 min at room temperature or 7 uM fura-2-AM for 30 min at 36°C. To load acutely isolated neurons with HPTS, a membrane-impermeant dye, neurons were exposed to 40 mM HPTS during the enzymatic treatment of hippocampal slices with protease and the mechanical trituration of microdissected hippocampal CA1 regions. To load acutely isolated neurons with SBFI, cells were triturated in the presence of 25 uM SBFI-AM and 5 mg ml"1 bovine serum albumin; following trituration, neurons were incubated with 25 uM SBFI-AM in the presence of 0.05% Pluronic acid F-127 for 30 - 45 min (see Chapter 5 for details of the SBFI-loading procedure used for postnatal hippocampal neuronal cultures). Following loading, neurons were superfused (2 ml min"1) with the initial experimental solution at 37°C for 15 min prior to the acquisition of data. 52 2.3.2. Imaging equipment Microspectrofluorimetric measurements were made employing a fluorescence ratio-imaging system (Atto Instruments Inc., Rockville, MD; Carl Zeiss Canada Ltd., Don Mills, ON) equipped with two intensified charge-coupled device cameras (Atto Instruments Inc.). The dual-excitation ratio method was used to determine pHj, [Na+]i and [Ca2+]i with BCECF (or HPTS), SBFI and fura-2, respectively; Fig. 2.1 provides a schematic illustration of the equipment used and details of the excitation and emission filters employed with each fluorophore. Following excitation at the appropriate wavelengths, BCECF, HPTS, SBFI or fura-2 fluorescence emissions were measured by one of two intensified charge-coupled device cameras (Camera 1 in Fig. 2.1). The camera gains at each excitation wavelength for a given dye were set to maximize image intensity, minimize the possibility of camera saturation, and were held constant throughout an experiment. Images were digitized at 8 bit resolution. Data were obtained from multiple neuronal cell bodies simultaneously, with each cell body delineated as a region of interest (ROI). As an indication of background fluorescence, data were also obtained from a cell-free zone throughout the course of an experiment. In order to minimize photobleaching of the dye and UV light-induced damage to the neurons, a computer-controlled high-speed shutter restricted the exposure of neurons to light to periods of data acquisition. Where possible, a variable intensity lamp control (Attoarc, Carl Zeiss Canada Ltd.) was employed to reduce the intensity of the mercury arc lamp and neutral density filters were placed in the light path to reduce the intensity of the incident light at each excitation wavelength. Ratio pairs were acquired at 1 - 15 s intervals throughout the course of an experiment. 53 2.3.3. Calculation of pHu fNali and rCa2+li 2.3.3.1. BCECF Fluorescence emissions at >520 nm were obtained from ROIs placed on individual neuronal somata and raw intensity data at each excitation wavelength (488 and 452 nm) were corrected for background fluorescence prior to calculation of the background-corrected BCECF emission intensity ratio {BI^IBUsi)- Analysis was restricted to those neurons able to retain BCECF throughout the course of an experiment (see Bevensee et al. 1995). The one-point high-[K+]/nigericin technique was employed to convert BUulBUsi ratio values into pHj values. At the end of an experiment, neurons loaded with BCECF were exposed to a pH 7.00, high-[K+] solution containing 10 pM nigericin (Solution 7; Table 2.3; see Baxter & Church, 1996). Nigericin, a carboxylic ionophore, equilibrates cytoplasmic and extracellular [K+] and, in doing so, equilibrates pH0 to pHj (see Thomas et al. 1979). This method provided, for every cell from which experimentally-derived ratio values were analyzed, a BU%%IBUs2 ratio value corresponding to pH 7.00. It has been suggested that the high-[K+]/nigericin technique may introduce errors when employed to measure absolute pHj values (Boyarsky et al. 1996a and 1996b). It is important to note, however, that the interpretation of BCECF-derived measurements of changes in pH, and rates of pH; recovery from imposed internal acid loads (see Section 2.4) are minimally affected by the application of the correction factors determined in the studies of Boyarsky and colleagues (1996a and 1996b). In addition, activity-induced changes in pHj observed in invertebrate glia were not different when comparing measurements made simultaneously using ISMs and BCECF (the latter being calibrated by the high-[K+]/nigericin technique; Nett & Deitmer, 1996). BI^IBUsi ratio values obtained during the calibration period (at pH 7.00) were used as normalization factors for 54 experimentally-derived BI^IBUsi ratio values and the resulting normalized ratio values were converted to pHj using the equation pH = pKa + log [(Rn - Rn(min)) /(Rn(max) - Rn)] (Equation 2.1) where Rn is the experimentally-derived BU^IBhn ratio value normalized to pH 7.00, Rn(min) and Rn(max) are the minimum and maximum obtainable values for the normalized ratio (i.e. at low and high pH values, respectively) and pKa represents the -log of the dissociation constant for BCECF. Rn(min), Rn(max) and pKa were derived from non-linear least-squares regression fits to normalized background-subtracted ratio values vs. pH data obtained in full calibration experiments (Fig. 22A). Full in situ calibrations of BCECF were performed by exposing neurons to 10 uM nigericin-containing, high-[K+] media (Solution 7, Table 2.3) titrated to a range of pH values (pH -5.5 to -8.5 in 0.5 pH unit increments). BI^IBI^ ratio values obtained during the course of a full calibration were normalized to the BUnlBUsi ratio value obtained at pH 7.00 and the resulting normalized BUulBhn ratio values (Rn) were plotted as a function of pH (Fig. 2.25). For the fourteen full calibration experiments utilized in analyzing all BCECF-derived experimental data, the mean values for Rn(max), Rn(min) and pKa were (mean + s.E.M.) 1.98 ± 0.04, 0.52 ± 0.01 and 7.31 ± 0.02, respectively. These values were not dependent on the temperature at which the full calibration was conducted nor the age of the hippocampal neurons used (data not shown). Full calibrations were performed whenever the mercury arc lamp was replaced or the optical set-up of the imaging system was altered. Nigericin can adhere to perfusion tubing and/or perfusion chambers and, by acting as an acid-loading K+/H+ exchanger, can alter pHj (Richmond & Vaughan-Jones, 1997; Bevensee et al. 55 1999a). Thus, after every one-point calibration or full in situ calibration in which nigericin was employed, perfusion lines were replaced and the imaging chamber was decontaminated by soaking first in ethanol, then in 20% Decon 75 (BDH Inc., Toronto, ON) and rinsed vigorously with water (see Richmond & Vaughan-Jones, 1997; Bevensee et al. 1999a). Selected experiments, in which BCECF was used as the pHj indicator, were also repeated using an experimental chamber that had never been exposed to nigericin; although the data from these experiments were not calibrated (and, therefore, are not presented in Chapters 3 or 4), the BCECF-derived BI^/BI^ ratio values obtained were not different from those recorded during the course of equivalent experiments conducted in nigericin-decontaminated chambers, suggesting that nigericin contamination does not contribute to the results obtained in the present studies. 2.3.3.2. HPTS Fluorescence emissions were measured at >520 nm and raw intensity data at each excitation wavelength (452 and 380 nm) were corrected for background fluorescence prior to calculation of the background-corrected HPTS emission intensity ratio (BI^BI^). The one-point high-[K+]/nigericin technique was employed to convert 5/452AB/380 ratio values into pH; values using the equation pH = [pAa + l0g(l/p)] + l0g[(Rn - Rn(min))/(Rn(max) - Rn)] (Equation 2.2) where Rn is the BIASJIBI^Q ratio normalized to unity at pH 7.00 and 1/(3 = fn2a/fn2b, where fn2a and f„2b are the normalized background-subtracted fluorescence intensities at the acidic and basic 56 extremes while exciting the dye at 380 nm (A,ex(2) in Fig. 2.1). The parameters of Equation 2.2 were derived from full calibration experiments, as described for BCECF. 2.3.3.3. SBFI SBFI-derived fluorescence emissions were obtained from ROIs placed on individual neuronal somata and raw intensity data at each excitation wavelength (334 and 380 nm) were corrected for background fluorescence prior to calculation of the background-corrected SBFI emission intensity ratio (5/334/5/330; see Fig. 2.1). A one-point calibration technique was developed to convert 5/334/5/330 ratio values into [Na+]; values (see Diarra et al. 2001 for full details). In brief, at the end of an experiment in which changes in [Na^j were measured, SBFI-loaded neurons were exposed to a pH 7.35 medium containing 10 mM Na+ and 4 pM gramicidin D, a channel-forming ionophore which acts to rapidly equilibrate external and internal monovalent cations (Solution 11, Table 2.3). 5/334/5/3go ratio values obtained during the calibration period ([Na+]j = 10 mM) were used as normalization factors for experimentally-derived 5/334/5/330 ratio values and the resulting normalized ratio values were converted to [Na+]j using the equation [Na+] = p/Td[(Rn - Rn(min))/(Rn(max) - Rn)] (Equation 2.3) where Rn is the experimentally-derived 5/334/5/330 ratio value normalized to [Na+] = 10 mM, Rn(min) and Rn(max) are the minimum and maximum obtainable values for the normalized ratio (i.e. at low and high [Na+]i values, respectively) and p/^d is the product of K~d, the dissociation constant of SBFI for Na+, and p (see below). Parameters required for the conversion of experimentally-derived 5/334/5/330 ratio values to [Na+]j (i.e. p, Ka, Rn(min), and Rn(max)) were 57 determined from full in situ calibrations in which neurons were exposed to pH 7.35 media containing eight different [Na+] values (range, 0-130 mM; Solutions 8-15, Table 2.3) in the presence of 4 uM gramicidin D (Fig. 2.3^4). BI334/BI3&0 ratio values measured at each [Na+] were normalized to BI^IBI^Q ratio values obtained at [Na+] = 10 mM; the resulting normalized 5/334/73/380 ratio values were plotted as a function of [Na+] and fit to the equation where A is a constant from which Rn(max) can be calculated (Rn(max) = A + Rn(min)) and B represents the product of $KA (Fig. 2.35). Measured following excitation at 380 nm (A.ex(2) in Fig. 2.1), P represents the ratio of the normalized fluorescence intensity of SBFI in its 'free' and 'bound' forms (i.e. normalized background-subtracted emission intensities in absence of Na+ and in the presence of saturating concentrations of Na+; 5/n(380f) and 5/n(380b), respectively). To determine p (and, thus, Kd), BIJSQ values from full calibrations were normalized to BI^o values measured at [Na+] = 10 mM and the resulting 5/n(380) values were plotted as a function of [Na+]. 5/n(380f) and #fn(380b) represent 5/n(380) values at the maximum and minimum extremes of the graphed function, respectively (Fig. 2.3Q. Illustrated in Fig. 2.3C, and data points were fit to the equation where c,D, and E are constants from which p can be derived; BInQ^o) = C + E at [Na ] = 0 mM, #fn(380) = C at saturating concentrations of [Na+] and D is [Na+]j at (C + E)/2 (this parameter is not required to calculate P). Rn = Rn(min) + [A([Na+])/(B + [Na+])] (Equation 2.4) £/n(380) = c + [E (D)/(D + [Na+])] (Equation 2.5) 58 For the eighteen full calibration experiments utilized in analyzing all SBFI-derived data, mean values of Rn(min), Rn(max), and p£d were 0.79 ± 0.06, 2.37 ± 0.16 and 55.81 ± 4.32 mM, respectively. Representative mean values for P and were 2.64 ± 0.45 and 21.32 ± 2.32 mM, respectively (n = 6). These values were not dependent on the temperature at which calibrations were conducted nor the age of the hippocampal neurons used (data not shown). Full calibrations were performed whenever the mercury arc lamp was replaced or the optical set-up of the imaging system was altered. To prevent contamination of the perfusion chamber with gramicidin D, perfusion lines were replaced and the perfusion chamber was decontaminated after each experiment (see Section 2.3.3.1). 2.3.3.4. Fura-2 Calibration of the fura-2 signal was not attempted and the effects of experimental maneuvers on [Ca2+]i are presented as changes in fura-2-derived BIj^/BI^o ratio values. Nevertheless, under conditions identical to those employed in the present experiments, this laboratory has found that a 5/334/5/380 ratio value of ~ 0.5 (as was observed in quiescent neurons in the present study) represents an [Ca2+]i ~ 80 nM (see Church et al. 1998). 2.4. EXPERIMENTAL PROCEDURES AND DATA ANALYSIS The effects of anoxia and other experimental maneuvers were examined on steady-state pHj and rates of pHj recovery from internal acid loads imposed by the NH4+-prepulse technique (as established by Boron & DeWeer, 1976). In experiments in which rates of pHj recovery were examined, 2-3 consecutive intracellular acid loads were imposed, the first one (or two) being 59 employed to calculate control rates of pHj recovery for a given neuron and the second (or third) being performed under a test condition. Rates of pHj recovery from imposed acid loads were determined by fitting the recovery portions of the pH record to a single exponential function of the form where a represents the pHj at the point of maximum acidification, b is the pHj range of recovery and c is the time constant. The first derivative of this function was then used to determine rates of pHi change as a function of time (see Wu & Vaughan-Jones, 1994; Baxter & Church, 1996; Smith etal. 1998) Instantaneous rates of pHj recovery (dpHj/df) under control and test conditions were calculated at 0.05 unit intervals of pHj from the point of maximum acidification: formal statistical comparisons were performed at the same absolute pHj values. There were no significant differences between the rates in pHj recovery observed when two (or more) consecutive internal acid loads were imposed under control conditions (Fig. 2.4). Data are reported as mean ± s.E.M. In experiments employing acutely isolated neurons, the accompanying n value refers to the number of neurons from which data were obtained. In experiments in which neuronal cultures were used, the accompanying n value refers to the pHi = a + b(l -e"ct) (Equation 2.6) dpHi/dr = bc(0 (Equation 2.7) 60 number of neuron populations (i.e. coverslips) from which data were obtained (measurements made from 2-5 different batches of neuronal cultures). 61 Table 2.1: Composition of commonly used experimental solutions Standard Standard Standard Low Na+ Low Na+ HC037C02- Hepes- loading (NMDG+) (Li+) buffered buffered (1) (2) (3) (4) (5) NaCl 127.0 136.5 133.5 2.0-4.0 2.0-4.0 NaHC03 19.5 - 3.0 - -KC1 3.0 3.0 3.0 3.0 3.0 CaCl2 2.0* 2.0 2.0 2.0 2.0 NaH2P04 1.5 1.5 1.5 - -MgS04 1.5 1.5 1.5 1.5 1.5 D-glucose 17.5 17.5 17.5 17.5 17.5 NMDG+ - - - 134-136 -LiCl - - - - 134- 136 Hepes - 10.0 10.0 10.0 10.0 Titrated - 10M 10M 10M 2M with: NaOH NaOH HC1 LiOH All concentrations are presented in mM. The standard HC03"-containing solution (Solution 1) was equilibrated with 5% C02 in balance air (normoxia) or balance argon (anoxia). Solutions for use with postnatal hippocampal neuronal cultures contained 10, not 17.5 mM, D-glucose. *HC03_ /C02-buffered medium used during the preparation of hippocampal slices and acutely isolated rat hippocampal CA1 pyramidal neurons contained 1 mM CaC^. Abbreviations: NMDG+, N-methyl-D-glucamine+ 62 Table 2.2: List of pharmacological agents Compound - proposed mechanism of action Solvent [Stock] mM Storage [Test] uM AACOCF3 - inhibitor of cytosolic PLA2 DMSO 10 -60°C 15-30 Bafilomycin A] - inhibitor of H4-ATPase DMSO 2 -20°C 1 -2 Bepridil - inhibitor of Na7Ca2+ exchange Hepes-buffered media - - 50 Bumetanide - inhibitor of Na+/K+/2C1" cotransport DMSO 50 - 50- 100 CGP-37157 - inhibitor of plasmalemmal and mitochondrial Na+/Ca2+ exchange DMSO 25 -3°C 25 CNQX - inhibitor of non-NMDA ionotropic glumate receptor-operated channels Ultra-pure H20 20 -20°C 20 DDA - inhibitor of adenylate cyclase DMSO 100 -60 °C 100 DIDS - inhibitor of HC03'-dependent pH( regulating mechanisms DMSO 100 - 200 Digitonin - selective permeabilization of the plasma membrane Ultra-pure H20 20 - 20 Gramicidin D - pore-forming ionophore; equilibrates [Na+]0 and [Na+]j 50:50 ethanol/methano 1 (v/v) 50 -60 °C 4 KB-R7943 - inhibitor of reverse-mode Na+/Ca2+ exchange Ultra-pure H20 5 -60°C 1- 10 Lidocaine - inhibitor of voltage-gated Na+ channels Ultra-pure H20 500 -3°C 250-500 L-NAME - inhibitor of nitric oxide synthase Hepes-buffered media - - 500 MK-801 - inhibitor of NMDA ionotropic glumate receptor-operated channels Ultra-pure H20 20 -20°C 2 Nifedipine - inhibitor of L-type voltage-gated Ca2+ channels DMSO 10 - 10 Nigericin - carboxylic carrier ionophore; equilibrates [K+]j and [K+]0 Ethanol 10 -60°C 10 Noradrenaline - full P-adrenoceptor agonist Ultra-pure H20 10 -60°C 10 Omeprazole - inhibitor of H+,K+ ATPase DMSO 25 -60°C 50 Ouabain - inhibitor of Na+,K+ ATPase Hepes-buffered media - - 500 Propranolol - full P-adrenoceptor antagonist Ultra-pure H20 50 -60°C 20 i?p-cAMPS - inhibitor of PKA Ultra-pure H20 50 -20°C 50 63 SCH-28080 - inhibitor of Ff,K+ ATPase Tetrodotoxin - inhibitor of voltage-gated Na+ channels Ultra-pure H20 DMSO 250 0.1 -60°C -3°C 500 Verapamil - non-selective inhibitor of voltage-gated Ca2+ channels Trolox - antioxidant Ultra-pure H20 Hepes-buffered media 500 1000 300 In the absence of an indicated stock concentration, test solutions were prepared directly in Hepes-buffered saline. In situations in which stock solutions were prepared fresh daily, no storage temperature is indicated. Ultra-pure H2O was obtained with a Milli-Q UF Plus Reagent Grade Water Purification System (Millipore, Mississauga, ON). Abbreviations: AACOCF3, arachidonyltrifluoromethyl ketone; CGP-37157, 7-chloro-5-(2-chlorophenyl)-l,5-dihydro-4,l-benzothiazepin-2(3#)-one; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione (disodium salt); DDA, 2',5'-dideoxyadenosine; DIDS, 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid; DMSO, dimethylsulphoxide; KB-R7943, 2-[2-[4-(4-nitrobenzyloxy)phenyl]ethyl]isothiourea mesylate; L-NAME, A^-nitro-L-arginine methyl ester; MK-801, (55,10i?)-(+)-5-methyl-10,1 l-dihydro-5#-dibenzo[a,d]cyclohepten-5,10-imine maleate; i?p-cAMPS, Rp- isomer of adenosine-3',5'-cyclic monophosphorothioate; SCH-28080, 2-methyl-8-(phenylmethoxy)imidazo[l,2-a]pyridine-3-acetonitrile; Trolox, 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylate. 64 Table 2.3: Composition of solutions used for in situ calibrations of pH and Na+-sensitive fluorophores High 0Na+ 3Na+ 6Na+ 10 20 40 80 130 [K+] Na+ Na+ Na+ Na+ Na+ (7) (8) (9) (10) (11) (12) (13) (14) (15) NaCl - - - - - - - 10.0 30.0 KC1 - 30.0 30.0 30.0 30.0 30.0 30.0 20.0 -CaCl2 1.0 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 NaH2P04 1.5 - - - - - - - -MgS04 1.5 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 Na Glu 10.0 - 3.0 6.0 10.0 20.0 40.0 70.0 100.0 KGlu 130.5 100.0 97.0 94.0 90.0 80.0 60.0 30.0 -D-glucose 17.5 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 Hepes 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 All concentrations are presented in mM. High-[K+] solutions (Solution 7) were titrated to a range of pH values (pH ~ 5.5 to -8.5 in 0.5 pH unit increments) with 10 M KOH and contained 10 pM nigericin. High-[K+] solutions for use with postnatal hippocampal neuronal cultures contained 10 mM D-glucose. Solutions with different Na+ concentrations (Solutions 8-15) were titrated to pH 7.35 (temperature-corrected) with 10 M KOH and contained 4 pM gramicidin D. Abbreviations: Na Glu, sodium gluconate; K Glu, potassium gluconate. 65 Fig. 2.1. A schematic representation of the optical equipment used in neurons single-loaded with a dual-excitation fluorophore (i.e. BCECF, HPTS, SBFI or fura-2). Neurons were excited with light provided by a 100 W Hg lamp (solid and dotted lines) and band-pass filtered through alternating interference excitation wavelength filters (X,eX(i) and A,ex(2))- Filtered excitation light then reflected off a dichroic mirror (Dichroic 1), passed through the objective (Zeiss LD Achroplan, n.a. 0.60, 40x) and illuminated the fluorophore loaded into neurons. At each excitation wavelength, fluorescence emissions (thick dotted line) passed through Dichroic 1 and a subsequent emission filter (Xem) before being detected by Camera 1. The table presented below the diagram details the filters employed to measure fluorescence from neurons loaded with BCECF, HPTS, SBFI or fiira-2 (i.e. Xex(i), A-ex(2), Dichroic 1 and ^em)- Ratio values (measured as the ratio of the background-subtracted emission intensity detected following excitation at the first excitation wavelength (5/ex(i)) to the background-subtracted emission intensity detected following excitation at the second excitation wavelength (5/ex(2))) were measured for each of the indicated fluorophores. * different filters were employed when SBFI was employed simultaneously with carboxy SNARF-1 or SNARF-5F (see Chapter 6). LP indicates that the filter is a long-pass filter. Compare with Fig. 6.1, a schematic diagram of the same optical equipment with slight modifications to allow the concurrent measurement of pHi and [Na+]i in the same cell. 66 Objective Dichroic 1 Dichroic 2 (not in place) Emission Filter Camera 1 0 Excitation Filters A-ex(1) ex(2) Lamp Camera 2 Excitation Filters ^ex(1) ^-ex(2) BCECF 488 + 5 452 ± 5 HPTS 452 ±5 380 + 5 SBFI* 334 + 5 380 ± 5 fura-2 334 + 5 380 ± 5 Dichroic 1 Emission Measured filter ratio ^em B/a<l/B/a0 510 520LP BWfiCe 510 520LP BUBIm 395 420LP eWB/a» 395 420LP flfe/Bbo 67 Fig. 2.2. Sample in situ calibration plot for BCECF. A, cells were exposed to nominally HCO3" /CCVfree, Hepes-buffered, high-[K+] solutions containing 10 pM nigericin at 37°C and the pH0 (and therefore pHj) values indicated above the record, which is the mean of data obtained from 17 cultured postnatal rat hippocampal neurons recorded on a single coverslip. B, plot of pHj against the resulting background-subtracted normalized ratio value (Rn). Rn was calculated as the quotient of the average background-corrected ratios for all neurons on a given coverslip at each pH value and the average background subtracted ratio value determined in the same neurons at pH 7.0. The curve is a result of a non-linear least squares regression fit to Equation 2.1. For this particular calibration, the values of Rn(min), Rn(max), and pKa were 0.44, 2.01, and 7.25, respectively. Error bars are S.E.M. (n = 3); where absent, error bars lie within the symbol area. A B 7.00 6.04 7.51 6.51 7.98 7.00 8.48 5.54 cm 2.5 2.0 1.5 1.0 0.5 —i— 20 —i— 40 —i— 60 Time (min) 80 100 0.0 69 Fig. 2.3. In situ calibration of SBFI at 37°C, pH0 7.35. A, a full calibration experiment in which 7 SBFI-loaded cultured hippocampal neurons were exposed to 4 uM gramicidin D-containing solutions at the [Na+] values (in mM) indicated above the records. Shown are the mean changes in normalized BI334/BI3W ratio values (Rn), which increased as [Na+] increased. B, plots of [Na+] vs. Rn obtained from experiments of the type shown in A (n = 4). The solid line represents the result of a three-parameter hyperbolic fit of the data points to Equation 2.4 and was used to determine the values of the SBFI calibration parameters (i.e. P^d, Rn(min) and Rn(max))- For this calibration, the values of Rn(min), Rn(max) and were 0.77, 2.52 and 66.02, respectively. C, plots of [Na+] vs. 5/n(38o) obtained from experiments of the type shown in A and from the same experiments used to determine the plot shown in B. The curve is the result of a three-parameter hyperbolic decay fit to Equation 2.5 and was used to determine the values of 5/n(380f) and 5/n(380b), #/n(380) values in the absence of Na+0 and in the presence of saturating concentrations of [Na+]0, respectively. For this calibration, 5/n(380f) and 5/n(380b) were 1.26 and 0.37, respectively. Thus, the calculated P and Ka values were 3.41 and 19.36 mM, respectively. In B and C, error bars are S.E.M; where absent, error bars lie within the symbol area. 70 71 Fig. 2.4. Consistency of rates of pHj recovery from internal acid loads imposed under control conditions. A, a representative record of the changes in pHj observed in an acutely isolated adult rat hippocampal CA1 pyramidal neuron in response to two consecutive internal acid loads, imposed under control conditions using the NH4+ pre-pulse technique. This trace was obtained under nominally-HC03"free, Hepes-buffered conditions, pH0 7.35, 37°C. B, the pHj dependencies of rates of pH, recovery following an initial (open circles) and a second (filled circles) internal acid load imposed under Hepes-buffered control conditions. Rates of pHj recovery were evaluated at 0.05 pH unit intervals of pHj and error bars represent S.E.M (n = 18). Continuous lines represent the weighted non-linear regression fits to the data points indicated for the first and second acid loads (see Motulsky & Ransnas, 1987). 72 73 CHAPTER THREE INTRACELLULAR pH RESPONSE TO ANOXIA IN ACUTELY ISOLATED ADULT RAT HIPPOCAMPAL CA1 PYRAMIDAL NEURONS: REDUCED Na+/H+ EXCHANGE ACTIVITY DURING ANOXIA4 3.0. INTRODUCTION As noted in Chapter 1, while the contribution of Ca ions to neuronal injury has received particular attention, there is renewed interest in the role of changes in pHj in neurodegenerative phenomena. As outlined in Section 2.0.1, while it has long been known that changes in neuronal pHi occur during and following anoxia or ischemia in vivo and in slice preparations in vitro (for reviews see Erecinska & Silver, 1994; Siesjo et al. 1996; Lipton, 1999), it is difficult under these experimental conditions to assess either the intrinsic changes in pHj which occur in neurons in response to anoxia or ischemia or the contribution of intrinsic alterations in the activities of neuronal pHj regulating mechanisms to the pH* changes observed. Recent studies, largely employing cultured fetal or postnatal neurons, point to the involvement of changes in Na+/H+ exchange activity in the neuronal pHj response to anoxia (Diarra et al. 1999; J0rgensen et al. 1999; Messier et al. 2004; also see Yao et al. 2001 for studies in acutely isolated mouse hippocampal neurons). Moreover, examined in vivo and in vitro, selective pharmacological inhibitors of Na+/H+ exchange exert protective effects on neurons in which the transport mechanism is sensitive to such compounds (e.g. Vornov et al. 1996; Horikawa et al. 2001a and b). However, it remains unclear whether potentially detrimental changes in neuronal 4 A version of this chapter has been published. Sheldon, C. and Church, J. (2004) Reduced contribution from Na+/H+ exchange to acid extrusion during anoxia in adult rat hippocampal CA1 neurons. J. Neurochem. 88: 594-603. 74 Na+/H+ exchange activity occur during and/or following anoxia (see Mutch & Hansen, 1984; Obrenovitch et al. 1990; Taylor et al. 1996). In addition, the sensitivity of mammalian central neurons to the damaging effects of anoxia (Kass & Lipton, 1989; Friedman & Haddad, 1993; Roberts & Chih, 1997; Isagai et al. 1999) and the mechanisms that serve to regulate neuronal pHj (Raley-Susman et al. 1993; Bevensee et al. 1996; Roberts & Chih, 1997; Douglas et al. 2001) are developmentally regulated, and it remains unclear whether findings made in phenotypically relatively immature cells in culture can be applied to more mature neurons, especially rat hippocampal CA1 pyramidal neurons that are particularly vulnerable to the damaging effects of anoxia. Thus, the aims of this first study were: i) to characterize the steady-state pH; changes that occur during and following transient periods of anoxia in hippocampal CA1 pyramidal neurons acutely isolated from adult rats; and ii) to examine whether NaVH1" exchange activity in adult rat hippocampal CA1 neurons remains functional during anoxia. Experiments examining changes in Na+/H+ exchange activity in the period immediately following anoxia are presented in Chapter 4. 3.1. MATERIALS AND METHODS 3.1.1. Experimental preparation Acutely isolated adult rat hippocampal CA1 pyramidal neurons, loaded with either BCECF or fura-2, were used in the majority of experiments presented in this Chapter. Solutions containing 17.5 mM 2-deoxyglucose (2-DG) were prepared by equimolar substitution for D-glucose. 75 3.1.2. Recording techniques Details of the techniques used for dye loading, dye calibration and the conversion of BCECF-derived BI^/BUsi ratio values to pHj values are presented in Chapter 2. It has been reported that BCECF inhibits the plasmalemmal Ca^H*-ATPase in erythrocytes (IC50 ~ 100 uM; Gatto & Milanick, 1993). Because the Ca^H1-ATPase in rat hippocampal neurons is an acid-loading Ca2+/H+ exchanger (Trapp et al. 1996b; see Fig. 1.3), anoxia-evoked changes in pHj measured with BCECF may be influenced by a reduction in background acid loading consequent upon inhibition of the ATPase. Therefore, HPTS, a fluorescent ratiometric H+-sensitive indicator that is reported not to inhibit activity-dependent pHj changes in snail neurons (Willoughby et al. 1998), was employed in a limited number of experiments to measure anoxia-evoked changes in pH,; details of the techniques used for dye loading, dye calibration and the conversion of HPTS-derived BI^IBh&o ratio values to pHi values are detailed in Chapter 2. 3.1.3. Experimental maneuvers The effects of anoxia were examined on steady-state pHi and rates of pHi recovery from internal acid loads imposed by the NH4+-prepulse technique. To compare the steady-state pH; changes evoked by anoxia under the various experimental conditions, three parameters were measured. The magnitude of the fall in pHi observed during anoxia was measured as the difference between the pre-anoxic resting pHj value and the minimum pHj value observed during anoxia. The magnitudes of the increases in pHj observed during and following anoxia were measured as the difference between the pHi value observed immediately prior to the return to normoxia and the minimum pHj value observed during anoxia, and the difference between the highest pHi value observed after anoxia and the pre-anoxic steady-state pHi value, respectively. In experiments in 76 which rates of pHi recovery during anoxia were examined, acid loads were imposed such that the peak of the internal acidification occurred at approximately the same time at which steady-state pHj during anoxia reached its minimum value (-2.5 min following the start of anoxia). Instantaneous rates of pHj recovery were then determined at -30 s after the peak acidification (i.e. at -3 min after the start of anoxia). 3.1.4. ATP determination Cellular ATP content was measured using the Molecular Probes ATP determination kit. The luciferin-luciferase assay is based on luciferase's requirement for ATP in the production of light. Experimental samples contained 5-6 CA1 principal cell layers microdissected from hippocampal slices, and were either exposed to anoxia or incubated under normoxic conditions with 5 pg ml"1 antimycin A and 17.5 mM 2-DG (to block oxidative phosphorylation and glycolysis, respectively; see Aharonovitz et al. 2000; Szabo et al. 2000) for the durations indicated in the Results. At the same time as experimental samples were exposed to anoxia or metabolic inhibition, paired samples were maintained in Hepes-buffered saline for an equivalent period of time. In all subsequent steps, samples were kept on ice. Following control or test treatments, samples were lysed by the addition of 0.1 M NaOH/1 mM EDTA and, after centrifugation, the supernatant was neutralized with 0.5 M perchloric acid (see Sheline et al. 2000) and the pellet was used to determine protein content (see below). Ten microlitre aliquots of the supernatant were removed and mixed with 200 pi Reaction Solution which contained (in mM): 25 Tricine buffer (pH 7.8), 5 MgS04, 0.1 EDTA, 0.1 sodium azide, 1 dithiothreitol, 0.5 d-luciferin and 1.25 pg ml"1 firefly luciferase. Sample bioluminescence was detected with a Berthold LB9507 Lumat luminometer (Fisher Scientific Ltd., Ottawa, ON). In all cases, 77 measurements were made in triplicate and data are presented as percentage declines from paired control measurements. Low-concentration ATP standard solutions were prepared by diluting a 5 mM ATP-containing solution in ultra-pure distilled and autoclaved H2O and were used to generate a standard curve relating measured luminescence to moles of ATP. Similarly, using the Bio-Rad DC Protein Assay kit (Bio-Rad Laboratories Inc., Mississauga, ON), standard curves relating absorbance measured at 595 nm (A595) and concentration of protein were generated and the protein content of the pellet was determined. Thus, the total content of ATP (moles) and protein (mg) of the lysates prepared from CA1 regions could be determined and, assuming a cytosolic volume of 2.4 pi mg"1 protein (Chinopoulos et al. 2000), concentrations of ATP could be estimated. 3.1.5. Statistical analysis Data are reported as mean ± s.E.M. and the accompanying n value refers to the number of acutely isolated CA1 neurons from which data were obtained. In experiments in which internal ATP content was measured, n refers to the number of samples examined under a given experimental condition. Statistical analysis was performed with Student's two-tailed t test, paired or unpaired as appropriate, with significance assumed at the 5% level. 3.2. RESULTS 3.2.1. Steady-state pHqinder normoxic conditions Under HC03"/C02-buffered conditions at pH0 7.35 and 37°C, resting pHj was distributed in a Gaussian manner around a mean of 7.34 ± 0.05 (range pH 7.07 - 7.78; n = 18). In nominally 78 HC037C02-free, Hepes-buffered medium at pH 7.35 and 37°C, steady-state pHj was 7.19 ± 0.01 (range pH 6.34 - 7.74; n = 330) and the distribution of resting pHj values was fit with the sum of two Gaussian distributions with means at pHj 6.90 ± 0.02 and pHj 7.35 ± 0.01. The mean resting pHi values and their distributions under both HCO37CO2- and Hepes-buffered conditions were similar to those reported previously by our laboratory (Smith et al. 1998; Brett et al. 2002a) and others (Bevensee et al. 1996) for acutely isolated mature rat hippocampal CA1 pyramidal neurons at 37°C. 3.2.2. Steady-state pHj response to anoxia The steady-state pHj changes evoked by 5 min periods of anoxia, induced by sodium dithionite, were first examined under HCCV/CCVbuffered conditions at pH0 7.35. The results are presented in Table 3.1 and a representative response is illustrated in Fig. 3.1^4. Anoxia elicited a triphasic pattern of steady-state pHi changes which consisted of an initial acidic shift following the induction of anoxia, a subsequent rise in pHj in the continued absence of 02 and, finally, a further internal alkalinization upon the return to normoxia which recovered slowly towards resting pHi values. A clear change in the rate of increase of pHj was observed during the transition from anoxia to normoxia in 10/14 neurons subjected to 5 min anoxia under HCCV/CCVbuffered conditions (corresponding changes were observed in 29/38 "high" pHj neurons under nominally HCCV/CCVfree conditions; see below). It was necessary to assess the possibility that sodium dithionite might induce changes in pHj via mechanisms unrelated to its 02 scavenging property. To do so, HCCV/CCVbuffered medium was bubbled vigorously with 95% ultrahigh purity Ar/5% C02 for periods of 1 to >18 h. In samples obtained anaerobically from the recording chamber, the P02 in medium bubbled with 79 Ar for 1 h was 25.3 ± 0.8 rnrn Hg (n — 4) whereas, measured in 8 different samples, the PQ2 in medium bubbled with Ar for >18 h was <1 mm Hg, and was not significantly different from that measured in medium containing 1 or 2 mM sodium dithionite. When a 5 min period of anoxia was imposed by exposing neurons to HC037C02-buffered medium that had been equilibrated with 5% C02/95% Ar for >18 h, the resultant steady-state pH, changes were not significantly different to those observed when the P0L was reduced to <1 mm Hg by the addition of sodium dithionite under identical buffering conditions (Table 3.1; Fig. 3.15). Thus, the steady-state pHj changes evoked by exposure to media containing sodium dithionite reflect a reduction in P02 and are not secondary to any additional properties of the 02 scavenger. Next, to assess the potential contribution of HCO3" ions and HCOV-dependent pHj regulating mechanisms to anoxia-evoked changes in steady-state pHi, experiments were repeated under nominally HC037C02-free, Hepes-buffered conditions. Neither the decrease in pHj nor the subsequent rise in pHj observed during anoxia were significantly different in the absence or presence of HCO3" (Table 3.1; also see Pirttila & Kauppinen, 1994). The increase in pHj observed following the return to normoxia was larger under Hepes- than under HCO37CO2-buffered conditions; however, this effect failed to reach statistical significance (Table 3.1; also see Pirttila & Kauppinen, 1994; Bevensee & Boron, 2000). In a portion of neurons (13/51) examined under Hepes-buffered conditions, 5 min anoxia elicited a different pattern of pHj changes to that described above. In these neurons, the fall in pHj during anoxia was significantly smaller (0.03 ± 0.01 pH units) than the response observed in the majority of cells examined under identical buffering conditions, and the small acidification gave way to a marked internal alkalinization (0.50 ± 0.04 pH units) that started during anoxia and continued into the post-anoxic period (Fig. 3.2A). This 'atypical' pattern of pHj changes was also 80 observed under HC037C02-buffered conditions in 3/14 neurons and, interestingly, is reported to be the usual response of mouse hippocampal CA1 neurons to 02 deprivation under Hepes-buffered conditions (Yao et al. 2001). It is noteworthy that neurons which exhibited this 'atypical' response had low resting pHj values (under Hepes-buffered conditions, the average pre-anoxic resting pHj value observed in neurons which exhibited the 'atypical' response was 6.95 ± 0.06 compared with 7.33 ± 0.04 in neurons which exhibited the more 'typical' response; n = 13 and 38, respectively; P < 0.05) and that the magnitudes of the fall in pHj observed during anoxia and the increase in pH, observed following anoxia appeared related to the pre-anoxic resting pH; value of a given neuron (Fig. 3.25, C). This finding is consistent with the possibility (Bevensee et al. 1996; Smith et al. 1998; Brett et al. 2002a) that a "low" pHj population of mature rat hippocampal CA1 pyramidal neurons exists that exhibits a distinct pattern of pHj regulation. 2+ 3.2.3. Contribution of changes in [Ca 1, to the changes in pH, observed during anoxia In adult CA1 neurons, anoxia leads to a disruption of internal ion homeostasis that is associated with energy failure and an abrupt depolarization of the plasma membrane (reviewed by Hansen, 1985; Lipton, 1999; also see Rader & Lanthorn, 1989; Silver & Erecihska, 1990; Tanaka et al. 1997). In the present study, anoxia evoked a 2.0 ± 0.4 (n = 8) unit increase in the fura-2 5/334/5/380 ratio value, which commenced at approximately 2 min after the induction of anoxia (as did the rise in pHj observed during anoxia; Fig. 3.3^4) and which remained elevated after the return to normoxia for as long as stable recordings could be maintained (up to 25 min following the end of an anoxic insult; also see Friedman & Haddad, 1993; Kubo et al. 2001). The potential contribution of changes in [Ca2+]i to the changes in steady-state pHj evoked by anoxia was assessed by imposing anoxia under external Ca -free conditions. As shown in 81 9+ Fig. 3.35, exposure to Ca -free medium caused a 0.30 ± 0.02 (n = 6) ratio unit decrease in resting BI^/B^o values and anoxia failed to induce the rapid and marked rise in BI334/BI3W 9-4-values that was observed in the presence of Ca (the increase in the 5/334/5/380 value observed under external Ca2+-free conditions was 0.02 ± 0.01 ratio units)5. In parallel experiments in 2_|_ BCECF-loaded neurons, exposure to Ca -free medium evoked an increase in steady-state pHi of 0.11 ± 0.04 pH units (n = 6), as previously reported (Smith et al. 1998). Once a new steady-state pHj value had been reached, a 5 min period of anoxia induced a triphasic pHj response, the individual components of which were not significantly different to those observed in the presence of 2 mM Ca2+0 (Table 3.1; Fig. 3.35). Following the initial fall in pHj, the increase in pHj that occurred during anoxia in the present study has been observed only relatively infrequently in neurons in vivo or in slice preparations in vitro (see Mabe et al. 1983; Fujiwara et al. 1992; Silver & Erecinska, 1992; Pirttila & Kauppinen, 1992; Melzian et al. 1996). Considering that BCECF inhibits the plasmalemmal acid-extruding Ca2+,FT-ATPase in erythrocytes (see Section 3.1.2), the possibility 9+ existed that the rises in pHj measured with BCECF under the high [Ca ]j conditions that pertain during anoxia may be artifacts consequent upon reduced background acid loading. Measured in HPTS-loaded neurons, however, the increases in pHj observed during and following 5 min anoxia were not significantly different to those observed in BCECF-loaded cells (Table 3.1; Fig. 3.4). Thus, in agreement with reports in which activity-dependent changes in pHj have been recorded in BCECF-loaded neurons (e.g. Trapp et al. 1996a; Wu et al. 1999), these data led me 5 Interestingly, in 4/6 neurons examined under Ca2+-free conditions, a small 0.08 ± 0.02 ratio unit increase in BI^JBI3m values was observed in the immediate post-anoxic period (Fig. 3.3B); although the basis for this transient increase was not investigated, it may reflect Ca2+ release from intracellular stores consequent upon anoxia-evoked changes in pH| (see Ou-Yang et al. 1994b). 82 to conclude that BCECF is an appropriate pH indicator for use in the present experiments. Other factors that may, in part, contribute to the absence of observable rises in pHj during periods of anoxia or ischemia in vivo and in slice preparations in vitro are the concomitant falls in pH0 observed in these multicellular preparations (cf the present experiments in which pH0 was maintained at 7.35 prior to, during and following anoxia; see Chapter 4). These results indicate that the typical steady-state pHj response of acutely isolated adult rat hippocampal CA1 pyramidal neurons to 5 min anoxia consists of an initial fall in pHj upon the induction of anoxia, a subsequent rise in pHj in the continued absence of 02, and a further internal alkalinization upon the return to normoxia. Because all three phases were not significantly influenced by the presence of HCO3", subsequent experiments were conducted in the nominal absence of HCO37CO2 to examine the effects of anoxia on Na+/Ff" exchange activity. 3.2.4. NaVH4" exchange activity during anoxia 3.2.4.1. Steady-state pPF, measurements As noted in Chapter 1, Na+/H+ exchange is the dominant acid extrusion mechanism in rat hippocampal neurons under HC037C02-free conditions but, unusually, this transport mechanism is insensitive to amiloride, amiloride derivatives and benzoylguanidinium compounds (Raley-Susman et al. 1991; Schwiening & Boron, 1994; Baxter & Church, 1996; Bevensee et al. 1996). To inhibit Na+/H+ exchange, therefore, neurons were perfused with reduced-Na+, NMDG+-substituted medium. In agreement with previous reports in rat hippocampal neurons (Baxter & Church, 1996; Bevensee et al. 1996; Smith et al. 1998), prolonged exposure to reduced-Na+, NMDG+-substituted medium was marked by an initial intracellular acidification which then slowly recovered over the following 20 - 30 min to a new steady-state pHj value of 7.35 ± 0.04 in = 20); this pHj value was not significantly different either to the pH, value observed prior to the 83 reduction in Na 0 (7.34 ± 0.03; P = 0.85) or to the mean pHj value typically observed prior to the induction of anoxia in the presence of normal [Na+]0 (7.33 ± 0.04; n = 38; P = 0.77). Under these reduced-Na+0, NMDG+-substituted conditions, 5 min anoxia evoked an internal acidification, the magnitude of which was significantly reduced compared to the fall in pHj observed in the presence of normal Na+0 (Fig. 3.5,4). Illustrated in Fig. 3.55, the magnitude of the increase in pH; observed during anoxia was not significantly influenced under NMDG+-substituted conditions. In contrast to NMDG+, Li+ can act as a substrate for Na+/H+ exchange (see Aronson, 1985; Jean et al. 1985). Consistent with previous reports in rat hippocampal neurons (Raley-Susman et al. 1991; Baxter & Church, 1996; Smith et al. 1998), exposure to reduced-Na+0, Li+-substituted medium caused a transient fall in steady-state pHj which recovered within 5-10 min to a pH; value (7.40 ± 0.02; n = 13) which was not significantly different either to the pHj value observed prior to the reduction of Na+0 (7.38 ± 0.02) or to the pH; value typically observed prior to the induction of anoxia in the presence of normal [Na+]0 (see above; P = 0.36 and 0.29, respectively). Under Li+0-substituted conditions, the magnitude of the fall in pHj induced by 5 min anoxia was not significantly different from that observed in the presence of normal Na+0 but was significantly greater than that observed after prolonged exposure to NMDG+-substituted medium (Fig. 3.5A). The magnitude of the increase in pHj observed during anoxia was not influenced under Li+0-substituted conditions (Fig. 3.55). Similar results, under NMDG+ and Li+-substituted conditions, were obtained when anoxia was imposed in the presence of 5 mM, rather than 17.5 mM, glucose (not shown; see Sheldon & Church, 2004). Taken together, these results suggest the possibility that Na+/H+ exchange activity in rat CA1 neurons becomes inhibited soon after the induction of anoxia. Under normoxic conditions, Na+/H+ exchange in rat hippocampal neurons is active at resting pHj (see Raley-Susman et al. 84 1991; Schwiening & Boron, 1994; Baxter & Church, 1996; Bevensee et al. 1996) and, in the presence of normal Na+0 or under Li+0-substituted conditions, reduced Na+/H+ exchange activity during anoxia would be expected to augment the internal acidosis produced during anoxia (also see Maduh et al. 1990; Chambers-Kersh et al. 2000). In contrast, under conditions where Na+/H+ exchange was blocked prior to the induction of anoxia by prolonged exposure to NMDG+-substituted medium, inhibition of Na+/H+ exchange activity by anoxia is precluded and will not contribute to the fall in pHj during anoxia; thus, the observed reduction in the magnitude of the anoxia-induced acidification when NMDG+, as opposed to Li+, was employed as an external Na4" substitute. However, because blockade of Na+/H+ exchange failed to affect the magnitude of the rise in pHj during anoxia, additional mechanism(s) must contribute to this phase of the anoxic pHj response (see Chapter 4). 3.2.4.2. Recovery of pH, from imposed internal acid loads The steady-state pHj measurements detailed above suggest that NaVH4" exchange activity may decline soon after the onset of anoxia. To further investigate this possibility, I compared rates of pHj recovery from intracellular acid loads imposed under HCCV/COvfree, Hepes-buffered conditions prior to and during anoxia. As illustrated in Fig. 3.6A, under control conditions pHj recovery from an acid load imposed during anoxia was slowed, compared to that observed prior to anoxia in the same cell. Examined in a total of 9 neurons with a mean steady-state pHi of 7.38 ± 0.04, instantaneous rates of pHj recovery were reduced significantly during anoxia at all absolute values of pHj (Fig. 3.65); at a common test pHj of 6.80, for example, there was a 47% decrease in the rate of pHj recovery during anoxia. The increases in pHj evoked by NH44" (quantified by taking the difference 85 between the steady-state pHj immediately prior to the application of NH4"1" and the maximum pHj observed during its application; see Smith et al. 1998) were similar prior to and during anoxia (0.25 ± 0.02 and 0.21 ± 0.04 pH units, respectively; n = 9 in each case; P = 0.36), suggesting that marked alterations in intracellular buffering power are unlikely to contribute to the reduction in rates of pHj recovery observed during anoxia. Next, internal acid loads were imposed prior to and during anoxia under reduced-[Na+]0, NMDG+-substituted conditions (Na+/H+ exchange blocked). Consistent with previous reports in rat hippocampal neurons (Schwiening & Boron, 1994; Baxter & Church, 1996; Bevensee et al. 1996), rates of pHj recovery prior to anoxia were significantly reduced, compared to rates of pHj recovery established in the presence of normal Na+0 (Fig. 3.6C, £>). In contrast, rates of pHj recovery during anoxia were not significantly different from those established during anoxia in the presence of normal Na+0 (Fig. 3.673). Also consistent with the possibility that functional NaVH1" exchange activity is reduced during anoxia, plots of the differences between rates of pHj recovery under normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions both prior to and during anoxia (Fig. 3.6E) revealed a reduced contribution from Na+0-dependent mechanism(s) to pHi recovery from acid loads during anoxia. Interestingly, under NMDG+-substituted conditions, rates of pHi recovery from internal acid loads imposed during anoxia were increased compared with rates observed prior to anoxia under NMDG+-substituted conditions (Fig. 3.6Q, suggesting that a Na+0- and HCCV-independent acid extrusion pathway is activated by anoxia. The potential mechanism(s) underlying the Na+0-independent recovery of pHj observed during anoxia in rat hippocampal neurons are examined in Chapter 4. Rates of pHj recovery were also measured prior to and during anoxia in 3 "low" pHj neurons (resting pHj 7.05 ± 0.08) and these data are illustrated in Fig. 3.7. Consistent with 86 observations made by Bevensee and colleagues (1996), rates of pHj recovery observed in "low" pHj cells prior to anoxia were slower at given absolute values of pHj than rates observed in "high" pHi neurons prior to anoxia (compare Figs. 3.7'A, 5 and 3.65, D), a difference that reflects reduced Na+0-dependent acid extrusion at each pHj value in "low" pHj cells (and, therefore, is not apparent under NMDG+-substituted conditions; see Bevensee et al. 1996). In "low" pHj cells, rates of pHj recovery from internal acid loads imposed prior to anoxia were not different from rates of pHj recovery established during anoxia under Na+0-containing (Fig. 3.7,4) or reduced Na+0, NMDG+-substituted conditions (Fig. 3.75). Illustrated in Fig. 3.7C, in "low" pHi cells, the contribution of Na+0-dependent mechanism(s) to pH, recovery from acid loads was nevertheless reduced during anoxia. However, this effect was less pronounced than observed in "high" pHj cells, a difference which likely reflects the reduced activity of Na+0-dependent pHj regulating mechanism(s) prior to anoxia in "low" vs. "high" pHj cells (Bevensee et al. 1996). Thus, at a common test pHj of 6.80, there were -6 and 4-fold reductions in Na+0-dependent rates of pHj recovery in "high" and "low" pHj cells, respectively (compare Figs. 3.7C and 3.675). Due to the limited number of "low" pHj cells isolated, no attempt was made to characterize further the influence of anoxia on Na+0-dependent pHi recovery in "low" pHj neurons. In light of the fact that Na+/H+ exchangers possess internal H+ modifier site(s) that modulate transport activity, the observed functional reduction in the contribution of Na+/H+ exchange to pHj recovery from acid loads imposed during anoxia may simply reflect the frequency of relatively "high" pHj cells that were found at pH0 7.35. Therefore, internal acid loads were imposed prior to anoxia at pH0 7.35 and then during anoxia at pH0 6.60, conditions that mimic the changes in pH0 that occur in response to anoxia in vivo {n = 9; Fig. 3.8,4). Although the minimum pHj values imposed by NH4+ prepulses during anoxia at pH0 6.60 were lower than those observed at pH0 7.35 (pHj -6.10 and -6.80, respectively; also see Vornov et al. 87 1996), rates of pHj recovery during anoxia at pH0 6.60 were further reduced (P < 0.05), rather than increased, from those observed during anoxia at pH0 7.35 (Fig. 3.8Q. This result is consistent with the possibility that Na+/H+ exchange continues to be inhibited during anoxia at low pHj values; however, the fact that rates of pHi recovery during anoxia at pH0 6.60 were slower than those observed during anoxia at pH0 7.35 (Fig. 3.85, Q suggests the possibility that low pH conditions might be affecting the activity of an additional mechanism that participates in acid extrusion during anoxia in rat CA1 neurons (e.g. the Na+0-independent, HCO3"-independent H+-efilux pathway referred to above; see Chapter 4). To more rigorously assess the effects of anoxia on Na+/H+ exchange activity at low pH0/pHj, the Na+0-dependent component of pHj recovery from internal acid loads was assessed by imposing acid loads prior to and during anoxia at pH0 6.60, under both normal Na+0-containing (n = 9) and reduced-Na+0, NMDG+-substituted (n = 10) conditions. As illustrated in Fig. 3.8C, rates of pHj recovery prior to anoxia at pH0 6.60 were significantly reduced, compared to those observed prior to anoxia at pH0 7.35, consistent with the known effect of falls in pH0 to inhibit the activities of Na+/H+ exchangers (e.g. Jean et al. 1985; Wu & Vaughan-Jones, 1997). However, rates of pH, recovery during anoxia at pH0 6.60 were not significantly different to rates observed prior to anoxia at pH0 6.60 (Fig. 3.8C; P = 0.68). In addition, although rates of pHj recovery during anoxia at pH0 6.60 were not significantly different under normal Na+0-containing vs. reduced-Na+0, NMDG+-substituted conditions (Fig. 3.8Q, plots of the differences between rates of pHj recovery under Na+0-containing and NMDG+-substituted conditions both prior to and during anoxia revealed a reduced contribution from Na+0-ctependent mechanism(s) to pHj recovery from acid loads during anoxia at pH0 6.60 (Fig. 3.8D). The reduced rate of Na+-dependent pHi recovery during anoxia at pH0 6.60 compared to pH0 7.35 (compare Figs. 3.6E 88 and 3.8/J) is consistent with a low pH0-induced inhibition of residual Na+/H+ exchange activity during anoxia. 3.2.4.3. Role of internal ATP depletion In all cell types studied to date, optimal Na+/H+ exchange activity requires the presence of normal physiological levels of intracellular ATP (Demaurex & Grinstein, 1994; Wu & Vaughan-Jones, 1994; Demaurex et al. 1997; Wakabayashi et al. 1997; Szabo et al. 2000). This raises the possibility that an anoxia-induced fall in internal ATP levels (see Erecinska & Silver, 1994) might contribute to the anoxia-evoked decline in Na+/H+ exchange activity. This was examined using a number of different approaches. First, to assess whether rates of pHj recovery from acid loads imposed in rat CA1 neurons in the nominal absence of HCO3" are sensitive to internal ATP depletion, microdissected CA1 regions were incubated with 2-DG and antimycin A under normoxic conditions. Consistent with previous reports (e.g. Kass & Lipton, 1982; Obrenovitch et al. 1990; Carter et al. 1995), resting ATP levels were 10.6 ± 3.5 pmol g"1 protein (n = 6), equivalent to -4.4 mM assuming a cytosolic volume of 2.4 pi mg"1 protein (see Chinopoulos et al. 2000). After 10 min treatment with 2-DG and antimycin A, there was an 80 ± 13% fall in internal ATP levels to a value below the KD of Na+/H+ exchange for ATP (see Discussion). At the time that ATP levels were reduced by 2-DG and antimycin A, rates of pHj recovery from imposed acid loads were slowed, compared with rates measured prior to ATP depletion in the same neurons (Fig. 3.9,4). At a common test pHi of 7.00, for example, there was a 53% decrease in the rate of pHi recovery (P < 0.05), which was not further slowed when the experiments were repeated under reduced-Na+0, NMDG+-substituted conditions (Fig. 3.95). However, plotting the difference between rates of pHj recovery under 89 normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions prior to and following treatment with 2-DG and antimycin A revealed a reduced contribution from Na+0-dependent mechanism(s) to pHi recovery from imposed acid loads in the presence of 2-DG and antimycin A (Fig.3.9Q. , Next, I examined whether anoxia imposed under my experimental conditions results in intracellular ATP depletion. Consistent with previous reports (e.g. Obrenovitch et al. 1990; Erecinska & Silver, 1994; Fowler & Li, 1998; Lipton, 1999), after 3 min anoxia there was a 65 + 4% (n = 3) fall in internal ATP levels, which declined further to 76 ± 4% (n = 2) after 5 min anoxia. Thus, pHj recovery from acid loads imposed during anoxia was slowed at a time when cellular ATP was depleted. Pretreatment of hippocampal slices with 10 mM creatine for >2 h has been shown to increase intracellular phosphocreatine levels in hippocampal neurons and delay the depletion of internal ATP during 02 deprivation (e.g. Kass & Lipton, 1982; Lipton & Whittingham, 1982; Carter et al. 1995; Balestrino et al. 1999). Therefore, in the third series of experiments, I examined whether this maneuver could preserve internal ATP levels and concomitantly attenuate the anoxia-induced decline in rates of pIL recovery from acid loads observed in untreated neurons. In creatine-treated slices, 3 min anoxia caused a 38 ± 6% (n = 4) fall in ATP levels, a reduction significantly less (P < 0.05) than that observed in untreated slices. In neurons isolated from creatine-treated slices, rates of pHj recovery from acid loads imposed during anoxia were not significantly different from rates of pHj recovery established in the same neurons prior to anoxia (Fig. 3.10,4, B). Intracellular acid loads were then imposed in neurons isolated from creatine-treated slices both prior to and during anoxia under reduced-Na+0, NMDG+-substituted conditions. At a common test pHj of 7.00, rates of pHj recovery from acid loads imposed under 90 NMDG+-substituted conditions both prior to and during anoxia were slowed by -60%, compared with rates of pHi recovery observed in the presence of normal Na+0 (Fig. 3.105). Thus, Na+-dependent acid extrusion mechanism(s) remain functional during anoxia in neurons isolated from creatine-treated slices. Indeed, plotting the difference between rates of pH* recovery measured under normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions prior to and during anoxia revealed that, in contrast to slices that had not been treated with creatine (see Fig. 3.6E), the contribution of Na+0-dependent mechanism(s) to pHj recovery from acid loads during anoxia is preserved in neurons isolated from creatine-treated slices (Fig. 3.10Q. Recently, it has been suggested that the inhibitory effect of ATP depletion on Na+/H+ exchange activity is attributable, at least in part, to a decreased availability of plasmalemmal PIP2 (Aharonovitz et al. 2000). Furthermore, reductions in plasmalemmal PIP2 levels have been observed following chemical ATP depletion and periods of ischemia (Sun & Hsu, 1996; Aharonovitz et al. 2000). To start to examine the possibility that NaVH* exchange activity in rat hippocampal neurons is influenced by plasmalemmal PIP2, cultured postnatal rat hippocampal neurons were incubated overnight with 5 mM neomycin and the effect of this treatment on rates of pHj recovery from acid loads imposed under HCOy/CCh-free, normoxic conditions was examined. Neomycin interferes with the ability of PIP2 to interact with membrane and cytoskeletal proteins, such as phospholipases C and D, actin and the Na+/H+ exchanger (see Abdul-Ghani et al. 1996; Castillo & Babson, 1998; Aharonovitz et al. 2000). Paired intracellular acid loads were not possible and rates of pHi recovery in neuronal cultures that had or had not been treated with neomycin were compared in parallel experiments. At a common test pHi of 6.90, rates of pHj recovery from imposed intracellular acid loads were 7.63 ± 2.6 x 10"J and 14.2 ± 2.0 x 10"3 pH units s"1 in the presence and absence of neomycin pretreatment, a 46% decrease 91 in the instantaneous rate of pHj recovery following treatment with neomycin (n = 6 in both cases; P<0.05). 3.3. DISCUSSION 3.3.1. Characterization of the changes in pH, observed during and following anoxia in adult rat hippocampal CA1 pyramidal neurons The typical steady-state pHj response to anoxia in acutely isolated adult rat hippocampal CA1 pyramidal neurons consisted of an initial fall in pHj, a subsequent rise in pHj in the continued absence of O2, and a further internal alkalinization upon the return to normoxia. In -25% of neurons examined, and almost exclusively in neurons with resting pHj values < 7.20 ("low" pHi neurons), a small acidification was observed during anoxia that gave way to a marked internal alkalinization beginning during anoxia and continuing into the post-anoxic period. In both "low" and "high" pHj neurons, the pHj changes were observed under constant external conditions and, as such, represent the intrinsic pHj response of the neurons to anoxia. A further discussion of the observations made in "high" vs. "low" pHj neurons will be presented in Chapter 4. Although studies in hippocampal slices have suggested that there are developmental changes in the neuronal pHi response to anoxia (see Roberts & Chih, 1997), the typical pattern of pHj changes observed in the present study was similar in most respects to that found in cultured postnatal rat hippocampal neurons (Diarra et al. 1999) and cultured fetal mouse neocortical neurons (J0rgensen et al. 1999). The major differences between the previous work in cultured postnatal hippocampal neurons (Diarra et al. 1999) and the present work in acutely isolated adult CA1 neurons are the relatively persistant increases in [Ca2+]j and pHj observed following 5 min 92 anoxia that, in cultured postnatal neurons, occur only after > 10 min anoxia. These differences may reflect, at least in part, the more marked and more persistent membrane depolarization that occurs in adult, compared with fetal or postnatal, hippocampal neurons on withdrawal of metabolic substrates (Bickler et al. 1993; Isagai et al. 1999; Nabetani et al. 1997; Tanaka et al. 1997 and 1999). In contrast to excitotoxin-evoked reductions in pHj which, under normoxic conditions, are largely consequent upon increases in [Ca ]j and the subsequent activation of a plasmalemmal Ca^Ff-ATPase (e.g. Hartley & Dubinsky, 1993; Irwin et al. 1994; Trapp et al. 1996b; Wu et al. 1999), changes in [Ca2+]j do not appear to be major determinants of anoxia-evoked changes in pH, in hippocampal CA1 neurons. Thus, as previously reported in cultured postnatal rat hippocampal (Diarra et al. 1999) and fetal mouse neocortical (Jergensen et al. 1999) neurons, despite the marked reduction in the anoxia-evoked increase in [Ca2+]i observed in the absence of Ca2+0, the pHj response to anoxia was not significantly affected. Experiments in which HPTS was employed as the pH; indicator did not support the possibility that inhibition of the plasmalemmal Ca2+,H+-ATPase by BCECF might account for the increases in pHj observed 9+ during (or after) anoxia in the presence of external Ca . Rather, my findings are consistent with the possibility that Ca2+,H+-ATPase activity may be inhibited by metabolic insults (Kass & Lipton, 1989; Pereira et al. 1996; Castilho et al. 1998; Wu et al. 1999; Zaidi & Michaelis, 1999; Chinopoulos et al. 2000) and therefore contribute little to background acid loading despite the marked increase in [Ca2+]i evoked by anoxia in adult CA1 neurons. 93 3.3.2. Reduced contribution from NaVH* exchange to acid extrusion during anoxia Anoxia induces a marked decline in HCO3"-independent, Na+0-dependent acid extrusion from adult rat hippocampal CA1 neurons. The only established HCO3"-independent, Na+0-dependent acid extrusion mechanism that supports Na+ and Li+, but not NMDG+, transport in this cell type is Na+/H+ exchange. As such, the results of the present study are consistent with the possibility that Na+/H+ exchange activity in adult rat CA1 pyramidal neurons declines soon after the onset of anoxia. The present finding is consistent with extensive studies in cardiac myocytes (e.g. Bond et al. 1993; Park et al. 1999; Satoh et al. 2001) and supports previous suggestions, made largely on the basis of pH0 measurements in vivo and in slice preparations in vitro, that Na+/H+ exchange activity in brain tissue is compromised during anoxia (Pirttila & Kauppinen, 1992; Taylor et al. 1996; Chambers-Kersh et al. 2000; but see Yao et al. 2001). However, it contrasts with the fact that Na+/H+ exchange is the primary mechanism whereby pH; recovers from the internal acidosis imposed by the application of excitotoxins under normoxic conditions (e.g. Hartley & Dubinsky, 1993; Koch & Barish, 1994), highlighting a further difference between the effects of excitotoxins and anoxia on central neuronal function (see Chow & Haddad, 1998). Given that the majority of the experiments in the present study were performed under constant extracellular conditions, the reduction in observable Na+/H+ exchange activity during anoxia is not secondary to anoxia-evoked changes in the composition of the microenvironment. In particular, although reductions in pH0 (as occur during anoxia in vivo and in slices in vitro; Mutch & Hansen, 1984; Obrenovich et al. 1990; Erecinska & Silver, 1994) are known to inhibit Na+/H+ exchange activity (e.g. Jean et al. 1985; Vaughan-Jones & Wu, 1990; Wakabayashi et al. 1997; Wu & Vaughan-Jones, 1997), the present results indicate that a fall in pH0 is not an absolute requirement for reduced antiport activity during anoxia in rat CA1 neurons. In addition, although anoxia-evoked increases in [Na+]j (Chapters 5 and 6; also see Chen et al. 1999) would 94 act to reduce the thermodynamic driving force for Na+/H+ exchange (see Vaughan-Jones & Wu, 1990; Wu & Vaughan-Jones, 1997), calculations indicate that the quotient [Na+]0/[Na+]j remains greater than [H+]0/[H+]i during anoxia at either pH0 7.35 or pH0 6.60, thereby favoring net FT efflux (a further discussion of the thermodynamics of NaVH4" exchange activity is presented in Chapter 6). This is in agreement with studies in guinea pig neocortical slices (Pirttila & Kauppinen, 1992) as well as cardiac myocytes (e.g. Park et al. 1999; Moor et al. 2001) and indicates that factor(s) other than changes in transmembrane H+ and/or Na+ gradients must contribute to the lack of observable Na+/H+ exchange activity in rat CA1 neurons during anoxia. Rather, the present results are consistent with the possibility that the decline in Na+/H+ exchange activity during anoxia might, at least in part, be consequent upon the fall in internal ATP levels which occurs rapidly after the induction of anoxia in adult rat CA1 neurons (also see Obrenovitch etal. 1990; Lipton, 1999). In all cases studied to date, physiological levels of internal ATP are required for optimal NaVH4" exchange activity (Demaurex & Grinstein, 1994; Wakabayashi et al. 1997; Szabo et al. 2000). In AP-1 cells transfected with Na+/H+ exchanger isoform 1 (NHE1), for example, half-maximal activation of the antiporter occurs at -5 mM ATP (Demaurex et al. 1997). In the present study, rates of pHj recovery from acid loads imposed during anoxia were slowed at a time when internal ATP levels were reduced from -4.4 mM under resting conditions to -1.5 mM, consistent with the established ATP dependence of not only NHE1 but also NHE5 (a candidate for the relatively amiloride-resistant Na+/H+ exchanger found in rat CA1 neurons; Szabo et al. 2000). Both the reduced slope of the rate of Na+0-dependent pHj recovery vs. pHj relationship and the acidic shift in the pHj dependence of the rate of Na+0-dependent pHj recovery from acid loads observed during anoxia (Figs. 3.6E and 3.7C) are also consistent with previous findings that internal ATP depletion decreases the affinities of Na+/H+ exchangers for internal protons and 95 lowers their maximum transport velocities (Demaurex & Grinstein, 1994; Wakabayashi et al. 1997; Szabo et al. 2000). The involvement of internal ATP depletion in the decline in NaVH4 exchange activity during anoxia is also suggested by the present findings that: a) incubation with 2-DG and antimycin A under normoxic conditions produced not only a similar fall in internal ATP levels to that observed during anoxia but also reduced rates of Na+0-dependent pHj recovery from internal acid loads to a similar extent; and b) creatine pretreatment not only limited anoxia-evoked reductions in ATP levels but also attenuated anoxia-induced reductions in rates of Na+-dependent pHj recovery from imposed acid loads. Although depletion of cellular ATP reduces the activities of all known Na+/H+ exchanger isoforms, it is also apparent that Na+/H+ exchange transport activity is not necessarily dependent on the direct hydrolysis of ATP (reviewed by Demaurex & Grinstein, 1994; Fliegel, 2001). In this regard, recent evidence indicates that the effect of acute ATP depletion to decrease NHE1 transport activity is in large part consequent upon the depletion of plasmalemmal PIP2, rapid reductions in which occur not only following chemical ATP depletion (Aharonovitz et al. 2000) but also in response to short (e.g. 3 min) periods of cerebral ischemia (Sun & Hsu, 1996). Indeed, in initial experiments, sequestration of PIP2 by pretreatment with neomycin (see Aharonovitz et al. 2000) was associated with slowed rates of pHj recovery from internal acid loads imposed under normoxic conditions, raising the possibility that NaVH4 exchange activity in rat CA1 neurons might also be regulated by the availability of PIP2. Although additional experiments are required to substantiate or refute this possibility, these experiments would provide novel insights into the second-messenger control of NaVH4 exchange activity in rat hippocampal neurons. In addition, the apparent relationship between NaVH4 exchange activity and internal ATP (and/or PIP2) levels would act to link the activity of the exchanger with the metabolic state of the cell. A reduction in antiport activity during a period of metabolic stress 96 may, for example, limit its contribution to potentially detrimental elevations in [Na+]j and (via reverse Na+/Ca2+ exchange) [Ca2+]i, albeit at the expense of a reduced rate of acid extrusion. The observation that the increase in pHj observed during anoxia was not inhibited under reduced Na+0 conditions indicates that Na+/H+ exchange does not make a major contribution to this phase of the pHi response, as expected if exchange activity is reduced shortly following the onset of anoxia. Similar findings have been made in rat central neurons in slice preparations (Pirttila & Kauppinen, 1992) and in primary culture (Diarra et al. 1999), although it appears contrary to a recent report in mouse hippocampal neurons (Yao et al. 2001). In the present study, it was observed that, under NMDG+-substituted conditions, rates of pHj recovery from internal acid loads imposed during anoxia were increased compared with rates observed prior to anoxia under NMDG+-substituted conditions (Fig. 3.6Q, suggesting that a Na+0- and HCO3"-independent acid extrusion pathway is activated by anoxia. In the following Chapter, I will examine further the contribution of this alternate acid extrusion pathway to the neuronal pHj response to anoxia. In conclusion, the present study suggests that Na+/H+ exchange activity in adult rat hippocampal CA1 pyramidal neurons is reduced during anoxia. These findings suggest that, as in cardiac myocytes (Bond et al. 1993; Park et al. 1999), the neuroprotective effects of selective Na+/H+ exchange inhibitors (e.g. Vomov et al. 1996; Phillis et al. 1999; Horikawa et al. 2001a and b) are unlikely to be exerted during anoxia. 97 Table 3.1: Anoxia-evoked changes in steady-state pHj Buffering condition n Magnitude (pH units) of: pHj decrease pHj increase pHj increase during anoxia during anoxia after anoxia HC037C02 14 0.17 ±0.02 0.07 + 0.01 0.15±0.05NS HC037C02 (Ar)* 4 0.14 ±0.01 0.08 ±0.01 0.18 ±0.07 Hepes 38 0.15 ±0.01 0.05 ±0.01 0.21 ±0.02 Hepes, Ca2+0-free 6 0.18 ±0.03 0.06 ±0.01 0.18 ±0.04 Hepes (HPTS)1" 4 0.08 ± 0.02 0.06 ± 0.03 0.23 ± 0.07 Experiments were performed at 37°C, pH0 7.35. Unless otherwise noted, 5 min anoxia was induced with sodium dithionite and BCECF was employed as the pH indicator. * 5 min anoxia t was induced by exposure to medium equilibrated with 5% C02/95% Ar for >18 h. HPTS was the pH indicator. N.S. indicates no statistically significant difference between the corresponding parameter obtained in response to 5 min anoxia under HCOy/COrfree, Hepes-buffered conditions (unpaired, two-tailed Student's f-test; P = 0.19). The pHj decrease during anoxia is the difference between the pre-anoxic steady-state pHi value and the lowest pHi value observed during anoxia. The pHi increases during and after anoxia are, respectively, the difference between the pH, value observed immediately prior to the return to normoxia and the minimum pHj value observed during anoxia, and the difference between the highest pH* value observed after anoxia and the pre-anoxic steady-state pHi value. 98 Fig. 3.1. Steady-state pHi changes evoked by transient periods of anoxia. A, shown are the steady-state pHj changes evoked by 5 min anoxia in a single acutely isolated adult rat hippocampal CA1 pyramidal neuron. Anoxia was imposed under HCCV/CCVbuffered conditions by exposure to medium containing sodium dithionite. Beneath the pHi trace is shown the 5/452 values (filled circles) employed in the measurement of pHj. The stability of the 5/452 values indicates that the relatively persistent nature of the increase in pHj observed after anoxia is not an artifact produced by a decline in /i/452 values consequent upon a deterioration of membrane integrity (see Chapter 2; Bevensee et al. 1995). B, a 5 min period of anoxia was imposed by exposure to HCCV/CCVbuffered medium that had been bubbled vigorously with 5% C(V/95% ultrahigh purity Ar for 20 hours. In A and B, records were obtained at 37°C and pH0 was 7.35 throughout. Time (min) Time (min) 100 Fig. 3.2. Relationship between pre-anoxic resting pHj values and anoxia-evoked changes in steady-state pHj. A, comparison of the 'typical' (solid line) and 'atypical' (open circles) pHj responses to 5 min anoxia in two different BCECF-loaded neurons exposed to sodium dithionite-containing, Hepes-buffered medium. Note the low resting pHj value in the neuron which responded to anoxia with a small reduction in pH, that gave way to a large internal alkalinization that started during anoxia and continued into the post-anoxic period. B, scatter plot relating magnitude of the fall in pHj observed during anoxia to pre-anoxic resting pHj values under HC037C02-(open circles) and Hepes-(filled circles) buffered conditions (n - 14 and 51, respectively). Solid and dashed lines represent linear regression fits to data obtained under HCO37CO2- and Hepes-buffered conditions, respectively (correlation coefficient = 0.80 and 0.67, respectively; P < 0.0005 in each case). C, scatter plot relating magnitude of the rise in pHi observed following anoxia to pre-anoxic resting pHj values under HC037C02.(open circles) and Hepes-(filled circles) buffered conditions (n = 14 and 51, respectively). Solid and dashed lines represent linear regression fits to data obtained under HCO37CO2- and Hepes-buffered conditions, respectively (in each case, correlation coefficient = 0.63; P < 0.02 in each case). resting pHt resting pH, 102 Fig. 3.3. Effects of anoxia on steady-state pHj and fura-2-derived 5/334/5/380 ratio values in the presence and absence of external Ca2+. A, in the presence of 2 mM external Ca2+, 5 min anoxia imposed under Hepes-buffered conditions induced a typical pattern of pHj changes (filled circles). Compare with Fig. 3AA, the same experiment conducted under HCCV/CCVbuffered conditions in a neuron with a similar resting pHj prior to anoxia. The open circles illustrate the changes in fura-2-derived 5/334/5/380 ratio values (representing changes in [Ca2+]i) evoked by 5 min anoxia in a parallel experiment under identical conditions employing a sister neuron. 5, upon exposure to Ca -free medium, pH; (filled circles) increased to a new steady-state value. The break in the pHj trace indicates a 2 min gap in the recording. When a new steady-state pH, value had been reached, 5 min anoxia induced a triphasic pattern of pHj changes, none of the components of which were significantly different from those observed in the presence of 2 mM Ca o- In contrast, measured in a sister neuron in a parallel experiment under identical conditions, the rise in 5/334/5/330 ratio values (open circles) was significantly attenuated (note the change of scale for the 5/334/5/330 axis between A and 5). There was also a small, reversible rise in 5/334/5/330 ratio values in the immediate post-anoxic period (see text). All records were obtained at 37°C under Hepes-buffered conditions at pH0 7.35. 103 104 Fig. 3.4. Anoxia-evoked changes in pHj measured with HPTS. A representative record of the pH, changes evoked by 5 min anoxia (sodium dithionite) in a neuron in which HPTS was employed as the pHj indicator. The experiment was performed under HCCV-free, Hepes-buffered conditions at 37°C and pH0 7.35 (see Table 3.1). 105 7.4 n 6.9 4 1 1 1 0 5 10 15 Time (min) 106 Fig. 3.5. The effects of external Na+ substitutions on the magnitudes of the fall (A) and rise (B) in pHj observed during anoxia under HCOV-free, Hepes-buffered conditions (pH0 7.35, 37°C). Data were obtained under control conditions (normal Na+0-containing; open bars); under reduced-Na+0, NMDG+-substituted conditions (hatched bars); and under reduced-Na+0, Li+-substituted conditions (cross-hatched bars). * indicates P < 0.05 compared to control or Li+-substituted conditions. N.S. indicates no significant difference (P = 0.20) between the fall in pHi evoked by anoxia under normal Na+0-containing compared to Li+0-substituted conditions. A B N.S 0.10 f | 0.08 O CD -o 'E o> as & 0.06 CO X o c CO D) c •a 0.04 4 0.02 A 0.00 i i Control conditions (n = 38) NMDG+-substituted (n = 20) «™ Li+-substituted (n = 13) 108 Fig. 3.6. Rates of pHj recovery from internal acid loads are reduced during anoxia. All experiments were performed under Hepes-buffered conditions at pH0 7.35, 37°C. A, following the first NH4+-induced intracellular acid load, pHj was allowed to recover. A second acid load was then imposed after the start of anoxia. Inset, superimposed records of the recoveries of pHi from acid loads imposed prior to (filled circles) and during (open circles) anoxia; the rate of recovery of pHj was reduced during anoxia. B, the pH, dependencies of rates of pHj recovery prior to (filled circles) and during (open circles) anoxia under control conditions (normal [Na+]0). Continuous lines represent the weighted nonlinear regression fits to the data points indicated for each experimental condition (n = 9 in each case). C, the pHj dependencies of rates of pHj recovery prior to (filled squares) and during (open squares) anoxia under reduced Na+0, NMDG+-substituted conditions. Continuous lines represent the weighted nonlinear regression fits to the data points indicated for each experimental condition (n = 5 in each case). D, rates of pHj recovery from internal acid loads imposed prior to anoxia under normal Na+0-containing conditions (black bar) were faster than those observed prior to anoxia under reduced-Na+0, NMDG+-substituted conditions (hatched bar) and during anoxia, under both normal Na+0-containing (open bar) and reduced-Na+0, NMDG+-substituted conditions (cross-hatched bar) (P < 0.05 in each case). There was no significant difference (N.S., P = 0.50) between rates of pHj recovery from acid loads imposed during anoxia under normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions. Rates of pHj recovery shown were determined at a common test pHj of 6.80. E, the Na+0-dependent component of pHj recovery prior to (filled circles) and during (open circles) anoxia, revealed by plotting the differences between the regression fits of pHj vs. dpHj/dr plots obtained under normal Na+0-containing conditions and reduced-Na+0, NMDG+-substituted conditions. B Anoxia 0.016 •— Prior to anoxia During anoxia NMDG+-substituted • Prior to anoxia I anoxia Prior to anoxia During anoxia 0.000 i 1 Y 1 1 r 6.8 6.9 7.0 7.1 7.2 7.3 7.4 D 0.000 0.016 T 1 1 1 ^T^ 1 6.8 6.9 7.0 7.1 7.2 7.3 7.4 PHi 0.000 ••• Prior to anoxia (n = 9) FEES?) Prior to anoxia, NMDG+-substituted (n i i During anoxia (n = 9) During anoxia, NMDG+-substituted (n pHi 110 Fig. 3.7. Rates of pHj recovery from internal acid loads prior to and during anoxia in neurons with "low" resting pHj values. All experiments were performed under Hepes-buffered conditions at pH0 7.35, 37°C. A, the pHi dependencies of rates of pHj recovery prior to (filled circles) and during (open circles) anoxia under control conditions (normal [Na+]0) were determined in 3 neurons in experiments of the type illustrated in Fig. 3.6,4. Continuous lines represent the weighted nonlinear regression fits to the data points indicated for each experimental condition. B, rates of pHj recovery from internal acid loads imposed prior to anoxia under normal Na+0-containing conditions (black bar) were faster than those observed prior to anoxia under reduced-Na+, NMDG+-substituted conditions (hatched bar; P < 0.05). There was no significant difference (N.S., P = 0.56) between rates of pHj recovery from acid loads imposed during anoxia under normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions (compare with Fig. 3.6Z), the same finding in "high" pHj neurons). Rates of pHj recovery are illustrated at a common test pHj of 6.80. D, the Na+0-dependent component of pHj recovery prior to (filled circles) and during (open circles) anoxia, revealed by plotting the differences between the regression fits of pHj vs. dpHj/d/^ plots obtained under normal Na+0-containing conditions and reduced-Na+0, NMDG+-substituted conditions (compare with Fig. 3.6E, the similar finding in "high" pHj neurons). Ill B 0.020 Prior to anoxia During anoxia 0.000 "o .t; c c CD Z> & i. I Q. "a 0.015 n i2 0.010 0.005 0.000 Prior to anoxia During anoxia 0.010 to x 0.005 D-X Q. 0.000 N.S. Prior to anoxia (n = 3) sssssa Prior to anoxia, NMDG+-substituted (n = 5) ' i During anoxia (n = 3) During anoxia, NMDG+-substituted (n =5) pH. 112 Fig. 3.8. pHj recovery from acid loads imposed prior to and during anoxia under reduced pH0 conditions. Experiments were performed under Hepes-buffered conditions at 37°C. A, an initial acid load was imposed prior to anoxia at pH0 7.35. After the recovery of pHi, pH0 was reduced to 6.60 and, when pHj had stabilized at a new resting level (pH 6.80 ± 0.05, n = 9), a second acid load was imposed during anoxia. B, the pHj dependencies of rates of pHi recovery from internal acid loads imposed during anoxia under normal Na+0-containing conditions at pH0 6.60 (open squares). Also illustrated are the pHi dependencies of rates of pHj recovery from internal acid loads imposed prior to (filled circles) and during (open circles) anoxia under normal Na+0-containing conditions at pH0 7.35 (see Fig. 3.65). A comparison for overall coincidence of the regression fits representing the pH* dependencies of rates of pHj recovery under pH0 7.35 and pH0 6.60 conditions indicated that the rate of pHj recovery from acid loads imposed during anoxia at pH0 6.60 was significantly slower (P < 0.05) than the rate established during anoxia at pH0 7.35. C, rates of pHj recovery from internal acid loads imposed prior to and during anoxia under the conditions shown on the figure, measured at a common test pH, of 6.40. Rates of pHj recovery from acid loads imposed prior to (black bar) and during (open bar) anoxia under normal Na+0-containing conditions at pH0 7.35 were estimated by extrapolating the weighted nonlinear regression fits relating absolute pHi values to the rates of pHj recovery obtained under each experimental condition (see B). At pH0 6.60, there was no significant difference (N.S., P = 0.35) between rates of pHj recovery from internal acid loads imposed during anoxia under Na+0-containing or reduced-Na+0, NMDG+-substituted conditions. D, the Na+0-dependent component of pHj recovery prior to and during anoxia at pH0 6.60, revealed by plotting the difference between the regression fits of pHj vs. dpHj/dr plots obtained under normal Na+0-containing 113 conditions and reduced-Na+0, NMDG+-substituted conditions (note the change in scale of the y-axis from Q. Rates were measured at a common test pHj of 6.40. 114 B pH0 6.60 x CL 0.018 '(/) 0.015 CO 'E 0.012 I 0.009 a. T3 0.006 ±~ D_ 0.003 0.000 l r 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 Time (min) PH| D 0.025 -| 'co CO 0.020 -X 0.015 -0.010 -±~ CI TS 0.005 -0.000 -N.S. i i T llllllll Prior to anoxia, pH0 7.35 i i During anoxia, pH0 7.35 Prior to anoxia, pH0 6.60 (n = 9) n Prior to anoxia, NMDG+-substituted, pH0 6.60 (n = 10) i i During anoxia, pH0 6.60 (n = 9) nun During anoxia, NMDG+-substituted, pH0 6.60 (n = 10) Prior to anoxia, pH0 6.60 During anoxia, pH0 6.60 115 Fig. 3.9. Treatment with 2-DG and antimycin A under normoxic conditions slows the rates of pHj recovery from internal acid loads. A, superimposed records of the recoveries of pHi from acid loads imposed in a CA1 neuron prior to and following 10 min incubation with 2-DG and antimycin A (2-DG + A). The rate of recovery of pHj was reduced following ATP depletion. B, rates of pHj recovery following 10 min incubation with 2-DG + A (open bar) were significantly slower than those observed in the same neurons prior to ATP depletion (black bar). No significant difference (N.S., P = 0.49) was observed between rates of pHj recovery from acid loads imposed following exposure to 2-DG + A under normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions. C, the Na+0-dependent component of pH, recovery prior to (black bar) and following (open bar) 10 min exposure to 2-DG + A, revealed by plotting the difference between the regression fits of pHj vs. dpHj/dr plots obtained under normal Na+0-containing and reduced-Na+0, NMDG+-substituted conditions. In B and C, rates of pHj recovery were determined at a common test pHi of 7.00. 116 B Prior to 2-DG + A During 2-DG + A 1 min c I Q. X Q. 0.008 '</, 0.006 oT X 0.004 0.002 •»»™ Prior to 2-DG + A (n = 5) ezzzza Prior to 2-DG + A, NMDG+-substituted (n = 6) 1=1 During 2-DG + A (n = 5) During 2-DG + A, NMDG+-substituted (n = 6) 0.000 Prior to 2 During 2 DG + A DG +A 117 Fig. 3.10. pHi recovery from internal acid loads in creatine-pretreated neurons. A, superimposed records of the recoveries of pHj from acid loads imposed prior to and during anoxia in a CA1 pyramidal neuron isolated from a hippocampal slice pretreated for 2 h with 10 mM creatine. In contrast to untreated neurons (see Fig. 3.6A), the rate of pHj recovery from the acid load imposed during anoxia was not slowed. B, in neurons pretreated with 10 mM creatine, rates of pHj recovery measured at ~3 min after the start of anoxia under normal Na+0-containing conditions (open bar) were not significantly different to those observed in the same neurons prior to anoxia (black bar; N.S., P = 0.84). Reducing Na+0 (NMDG+-substitution) slowed rates of pHj recovery from acid loads imposed both prior to and during anoxia (hatched and cross-hatched bars, respectively; *, P < 0.05 in each case). C, the Na+0-dependent component of pHj recovery prior to (black bar) and during (open bar) anoxia in neurons isolated from creatine-treated slices, revealed by plotting the difference between the regression fits to pHj vs. dpHj/d^ plots obtained under normal Na+0-containing conditions and reduced-Na+0, NMDG+-substituted conditions. In B and C, rates of pH, recovery were determined at a common test pHj of 7.00. 118 B Prior to anoxia During anoxia 1 min 0.012 0.000 N.S. c <D "O c CD Q. 01 "O i o + 05 X CL T3 0.008 0.006 A 0.004 0.002 0.000 Prior to anoxia (n = 5) Prior to anoxia, NMDG+-substituted (n = 7) During anoxia (n = 5) During anoxia, NMDG+-substituted (n = 7) Prior to anoxia During anoxia 119 CHAPTER FOUR INTRACELLULAR pH RESPONSE TO ANOXIA IN ACUTELY ISOLATED ADULT RAT HIPPOCAMPAL CA1 PYRAMIDAL NEURONS: INCREASED NaVH* EXCHANGE ACTIVITY AFTER ANOXIA6 4.0. INTRODUCTION The potential importance of Na+/H+ exchange activity to ischemic neuropathology is suggested by findings that pharmacological blockers of Na+/H+ exchange exert a protective effect in neurons in which the antiport is sensitive to such compounds (e.g. Vornov et al. 1996; Kuribayashi et al. 1999; Phillis et al. 1999); however, the timing of these neuroprotective actions is uncertain. The results presented in Chapter 3 suggested that NaVH4" exchange activity in acutely isolated adult rat hippocampal CA1 pyramidal neurons is reduced shortly following the onset of anoxia. In the present Chapter, the possibility that NaVH4" exchange activity may become activated following anoxia was investigated. NaVH4" exchange inhibitors have been shown to limit post-ischemic amino acid release (Phillis et al. 1998), free fatty acid efflux (Pilitsis et al. 2001) and cerebral Na+ and water content (Kuribayashi et al. 1999; also see Chapters 5 and 6), suggesting that NaVH4" exchange may be active during reperfusion. Thus, the first aim of this study was to examine NaVH4" exchange activity in acutely isolated adult rat hippocampal CA1 pyramidal neurons in the immediate post-anoxic period. In addition to NaVH4" exchange, studies in cultured postnatal rat hippocampal (Diarra et al. 1999) and fetal mouse neocortical (J0rgensen et al. 1999) neurons point to the involvement of 6 A version of this chapter has been published. Sheldon C. and Church J. (2002) Intracellular pH response to anoxia in acutely dissociated adult rat hippocampal CA1 neurons. J. Neurophysiol 87: 2209-2224. 120 9+ a Zn -sensitive acid extrusion mechanism in the neuronal pHj response to anoxia. Indeed, results presented in Chapter 3 illustrated that a Na+0- and HCdV-independent mechanism(s) contributes to the increase in pHj observed during anoxia. Therefore, the second aim of this study was to further examine the activity of this additional acid extrusion mechanism during and following periods of anoxia in acutely isolated adult rat hippocampal CA1 pyramidal neurons. 4.1. MATERIALS AND METHODS 4.1.1. Experimental preparation In all experiments presented in this Chapter, BCECF-loaded acutely isolated adult rat hippocampal CA1 pyramidal neurons were used. Nominally HC037C02-free, Hepes-buffered media were employed in all experiments, which were conducted at 37°C and pH0 7.35 (unless otherwise noted). 4.1.2. Experimental maneuvers The effects of transient periods of anoxia were examined on both steady-state pHj and on rates of pHj recovery from internal acid loads imposed by the NH4+ prepulse technique. To compare the steady-state pH, changes evoked by anoxia under the various experimental conditions, the magnitudes of the increases in pH; observed during and after anoxia were measured (as described in Chapter 3, Section 3.1.3). In experiments in which rates of pHj recovery were examined, internal acid loads were imposed immediately following 5 min anoxia. In experiments where the composition of the external medium was altered during the recovery of pH, from an intracellular 121 acid load (see Figs. 4.6 and 4.7), individual portions of the recovery were fit to a linear equation, as described by Raley-Susman et al. (1991). Data are reported as mean ± S.E.M. with the accompanying n value referring to the number of neurons from which data were obtained. Statistical analyses were performed with Student's two-tailed paired or unpaired t tests, as appropriate. Significance was assumed at the 5% level. 4.2. RESULTS 4.2.1. Na+/Fft" exchange activity in the immediate post-anoxic period The following series of experiments were designed to examine NaVFf* exchange activity in the immediate post-anoxic period. I first examined whether inhibiting NaVFT1" exchange activity influenced the magnitude of the increase in steady-state pHi observed following anoxia (described in Chapter 3, Section 3.2.2; Fig. 4.1,4). To inhibit Na+/H+ exchange activity, neurons were perfused with reduced-Na+, NMDG+-substituted medium prior to, during and following anoxia. Under these conditions, the increase in pHi observed following anoxia was significantly reduced compared to the increase in pH-, observed following anoxia in the presence of normal Na+0 (Fig. 4.1,4, E). Under reduced-Na+0, Li+-substituted conditions, conditions which support Na+/Fi+ exchange activity (see Raley-Susman et al. 1991; Baxter & Church, 1996), the increase in pHj observed after anoxia was restored to control levels (Fig. 4.1,4, B). The results are consistent with the possibility that Na+/H+ exchange becomes active after anoxia and contributes to the internal alkalinization observed at this time. 122 To further examine Na+/FT exchange activity in the immediate post-anoxic period, I compared rates of pH; recovery from internal acid loads imposed prior to and immediately following anoxia. Examined in 17 neurons with a mean resting pHj of 7.34 ± 0.02, instantaneous rates of pHj recovery were increased significantly after anoxia at all absolute values of pHj (Fig. 4.2A, B). The increases in pHi evoked by NH/ (quantified by taking the difference between the steady-state pH, immediately prior to the application of NH/ and the maximum pHj observed during its application; see Smith et al. 1998) were similar prior to and after anoxia (0.24 ± 0.02 and 0.21 ± 0.02 pH unit increases, respectively; n - 17 in each case; P = 0.16), suggesting that marked alterations in intracellular buffering power are unlikely to underlie the changes in the rates of pHj recovery observed after anoxia. Next, internal acid loads were imposed under reduced-Na+0, NMDG+-substituted conditions in = 5); rates of pHj recovery after anoxia were significantly slower than the corresponding rates observed in the presence of Na+0 (Fig. 4.2/3). Consistent with the possibility that Na+/H+ exchange activation was occurring in the immediate post-anoxic period, plots of the differences between rates of pHi recovery under Na+0-containing and reduced-Na+0 (NMDG+-substituted) conditions both prior to and after anoxia (Fig. 4.2Q revealed an increased contribution from a Na+0-dependent mechanism to pHj recovery from acid loads imposed in the immediate post-anoxic period. Rates of pHj recovery prior to and following anoxia were also determined in 5 "low" pHi neurons (average resting pHj 6.89 ± 0.03) and these data are illustrated in Fig. 4.3. As detailed in Chapter 3 (Section 3.2.4.2), rates of pHi recovery observed in "low" pHj cells prior to anoxia were slower than rates observed in "high" pHj cells prior to anoxia (compare Fig. 4.3^4 with Fig. 4.2/3). Nevertheless, consistent with observations made in "high" pHj cells, in "low" pHj cells, rates of pHj recovery from internal acid loads imposed immediately following anoxia were faster 123 than rates of pHj recovery observed prior to anoxia (Fig. 4.3,4 and B). As illustrated in Fig. 4.35, compared to rates of pHi recovery observed in "low" pHj cells in the presence of normal Na+0, reducing external [Na+]0 (NMDG+-substitution) had only a minor effect on rates of pHi recovery observed prior to anoxia, but significantly slowed rates of pHj recovery observed following anoxia. Thus, analogous with findings made in "high" pHj neurons, the contribution of Na+0-dependent mechanism(s) to pHi recovery from acid loads imposed immediately following anoxia was enhanced in "low" pH; neurons (Fig. 4.3Q. According to Bevensee et al. (1996) and Smith et al. (1998), "low" and "high" pHi adult rat CA1 neurons appear to represent neurons with "low" and "high" levels of NaVFT" exchange activity, respectively. As outlined in Chapter 3, NaVH4" exchange activity was reduced during anoxia and, accordingly, the magnitude of this reduction was larger in "high" vs. "low" pHj neurons. The present results suggest that NaVFT" exchange activity is increased following anoxia in both "high" and "low" pHj neurons; however, this effect is more marked in "low" pHj neurons, presumably reflecting the lower level of NaVFT1" exchange activity prior to anoxia in this group of neurons (at a common test pHj of 6.80, rates of Na+0-dependent pHj recovery immediately following anoxia were increased by -8.5 and 1.3-fold in "low" and "high" pHj neurons, respectively; compare Figs. 4.2C and 4.3Q. Of particular interest is the finding that the apparent activation of Na+/H+ exchange in the immediate post-anoxic period occurred even though external pH was held at a constant value (i.e. pH0 7.35). Thus, a return to normal pH0 values from an external acidification, as would occur after anoxia in vivo, is not an absolute requirement for the post-anoxic activation of Na+/H+ exchange activity in isolated rat hippocampal neurons. One mechanism that could contribute to the activation of Na+/H+ exchange after anoxia is an anoxia-induced change in the activity of intracellular second messenger system(s) which, in turn, act to regulate Na+/H+ exchange activity. In hippocampal neurons, the intracellular concentration of cAMP rises rapidly in the immediate 124 post-anoxic period (e.g. Whittingham et al. 1984; Domanska-Janik, 1996; Small et al. 1996), and our laboratory has shown previously that increases in [cAMP]i, acting via PKA, activate Na+/H+ exchange in acutely isolated rat CA1 neurons under normoxic conditions (Smith et al. 1998). Therefore, I investigated the effect of modulating the activity of the cAMP/PKA system on the rise in steady-state pH; observed after anoxia. As previously reported (Smith et al. 1998), the selective PICA inhibitor i?p-cAMPS (50 uM) failed to affect steady-state pHj under Hepes-buffered, normoxic conditions. However, as illustrated in Fig. 4.4,4, the magnitude of the internal alkalinization observed after anoxia was significantly reduced in the presence of i?p-cAMPS. In contrast, 50 uM i?p-cAMPS failed to significantly affect the increase in pHj observed after anoxia under reduced-Na+0 (NMDG+-substituted) conditions (Fig. 4.45; also see Fig. 4.55 for the effects of NMDG+-substitution on the increase in pHj observed after anoxia in the absence of i?p-cAMPS). Similar results were obtained following pretreatment with the adenylate cyclase inhibitor DDA (100 uM), which reduced the magnitude of the increase in pHj seen after anoxia under normal Na+0-containing conditions to 0.11 ± 0.03 pH units (n = 9; P < 0.05 for the difference to the increase in pHj observed in the absence of DDA). The action of P adrenergic agonists to increase [cAMPJi is potentiated after ischemia (Lin et al. 1983; Domanska-Janik, 1996). Therefore, to further assess the role of the cAMP/PKA pathway in the regulation of Na+/H+ exchange activity in the immediate post-anoxic period, the effects of p adrenergic agonists on anoxia-evoked changes in pHj were examined. Stimulation of the cAMP/PKA pathway with the p-adrenoceptor agonist isoproterenol (10 pM) significantly increased the magnitude of the post-anoxic alkalinization, an effect that was attenuated by pretreatment with the full P adrenoceptor antagonist propranolol (20 uM) or 7?p-cAMPS (50 uM; Fig. 4.45). Furthermore, the 125 isoproterenol-evoked increase in the magnitude of the post-anoxic alkalinization was attenuated significantly under reduced-Na+0 (NMDG+-substituted) conditions and was restored to control values when Li+ was employed as the Na+-substitute (Fig. 4.45). Taken together, the results are consistent with the possibility that anoxia-induced changes in the activity of the cAMP/PKA second messenger system may contribute to the activation of Na+/Ff+ exchange in adult rat hippocampal CA1 pyramidal neurons immediately after anoxia. 4.2.2. Contribution of a NaV and HCCV-independent mechanism to acid extrusion during and following anoxia A number of lines of evidence presented in this and the preceding Chapter suggests that mechanisms other than NaVH4" exchange must contribute to acid extrusion during and following anoxia in adult rat CA1 neurons. Despite marked reductions in NaVFT" exchange activity shortly following the onset of anoxia, increases in pHj, in the presence and absence of Na+0, were often observed during anoxia (Fig. 3.5). In addition, under conditions that inhibit NaVH4" exchange activity (NMDG+-substitution), the increase in pHj observed immediately following anoxia was not fully eliminated (Fig. 4.15). In fact, under NMDG+-substituted, Hepes-buffered conditions, rates of pHi recovery from internal acid loads were increased during and following anoxia, compared to rates established prior to anoxia under the same conditions (see Fig 3.6C and 4.25), suggesting that a Na+0- and HCCV-independent acid extrusion pathway is activated by anoxia. The Na+0-independent internal alkalinizations that occurred during and following anoxia appeared to be associated temporally with marked and persistent increases in [Ca2+]j (see Fig. 3.3A), raising the possibility that they may reflect H+ efflux through a H+-conductive pathway activated by membrane depolarization. In all cell types studied to date, voltage-activated H+ 126 conductances (gH+s) are blocked by micromolar concentrations of Zn2+ (for reviews see DeCoursey & Cherny, 1994a and 2000; Eder & DeCoursey, 2001; also see Cherny & DeCoursey, 1999). Therefore, the following series of experiments examined the potential contribution of a Zn2+-sensitive, voltage-activated EE efflux pathway to the increases in pHj observed during and following anoxia in isolated rat hippocampal CA1 neurons. First, I examined the effects of 100 - 500 uM Zn2+ on the rises in pHj observed during and following anoxia. While the application of Zn2+ did not change resting pHj prior to anoxia, there was a significant reduction in the magnitudes of the rises in pHj observed during and following anoxia (Fig. 4.5A, B). When Zn2+ was applied under reduced-[Na+]0, NMDG+-substituted conditions, the magnitudes of the rises in pHj observed during and following anoxia were reduced to values significantly less than those observed under reduced-[Na+]0, NMDG+-substituted, conditions alone (Fig. 4.5,4, B). Next, internal acid loads were applied during or immediately after anoxia and pHj recovery was allowed to proceed in the presence of 100 - 500 pM Zn2+ and/or under reduced-Na+0, NMDG+-substituted conditions. The recovery of pHj from internal acid loads imposed during anoxia was markedly inhibited under reduced-Na+0 conditions in the presence of Zn2+, and there was an increase in the rate of pHj recovery when Zn2+ was removed from the low-Na+ medium. Rates of pHj recovery were estimated by linear regression fits to pHj data obtained under reduced Na+0, NMDG+-substituted conditions in the presence of Zn2+ and under reduced Na+0, NMDG+-substituted conditions upon the removal of Zn2+ (for -120 s after its removal): the slopes of the fitted lines approximated the rates of pHj recovery (pH units s"1) and, in order to provide estimates of the pH, values at which rates of pHj recovery were measured, the pHj values at the mid-point of the lines (pHo.s) were also determined. Rates of pHi recovery were 1.08 ±1.0 127 x 10"3 and 2.84 ± 1.20 x 10"3 pH units s"1 under reduced Na+0, NMDG+-substituted conditions in the presence and absence of 250 pM Zn2+, respectively (n = 4 in both cases). Thus, there was an increase in the rate of pHj recovery when Zn2+ was removed from the low-Na+ medium, although this did not reach statistical significance (P = 0.46), likely reflecting both the limited number of neurons that could tolerate this experimental series and the different pHi values at which rates were measured (the pHo.5 values at which rates were estimated under reduced Na+0, NMDG+- substituted conditions in the presence and absence of Zn2+ were ~6.4 and 6.7, respectively). The recovery of pHj from an internal acid load imposed immediately following anoxia was also markedly inhibited under reduced-Na+0 conditions in the presence of Zn2+. As illustrated in Fig. 4.6^4, there was a significant increase in the rate of pHj recovery when Zn2+ was removed from the low-Na+ medium, and the rate of recovery increased further upon the reintroduction of normal external Na+. Similar results were obtained in experiments in which pHj recovery from an acid load imposed following anoxia was allowed to proceed initially under control conditions (i.e. in the presence of normal [Na+]0 and absence of Zn2+); in these experiments, reducing external Na+ and/or adding Zn2+ also slowed the rate at which pHj recovered. The pooled results from these series of experiments are presented in Fig. 4.65. Taken together, the results are entirely consistent with the possibilities, raised in light of the steady-state pHj data, that a Na+0- and HCO^-independent, Zn2+-sensitive mechanism contributes to acid extrusion during and after anoxia in acutely isolated adult rat CA1 pyramidal neurons. To assess the possibility that the Zn2+-sensitive component of the recovery of pHi from acid loads imposed during or following anoxia might be activated by membrane depolarization, internal acid loads were applied during normoxia under Na+0-free conditions (i.e. Na+/H+ exchange blocked). As illustrated in Fig. 4.7, pHj recovery in the absence of external Na+ (see 128 Bevensee et al. 1996; Smith et al. 1998) was >2-fold faster under depolarizing (139.5 mM K+0) than under control (3 mM K+0) conditions (n > 5 in each case). Furthermore, whereas 100 pM 9+ Zn failed to affect pHj recovery under control conditions, the rate of pH, recovery under high-[K ]0 conditions was reduced upon the application of Zn . In contrast, as illustrated in Fig. 4.8, the effect of high-[K+]0 to increase rates of pHj recovery from acid loads imposed under normoxic conditions (in the absence of Na+0) was not affected by the P-type H+,K+-ATPase inhibitors omeprazole (50 pM; Wu & Delamere, 1997) or SCH-28080 (500 uM; Petrovic et al. 2002), or the V-type H+-ATPase inhibitor bafilomycin Aj (2 pM; Wu & Delamere, 1997). 4.2.3 Effects of changes in pH„ Anoxia and ischemia in vivo and in slice preparations in vitro lead to reductions in pH0 (e.g. Obrenovitch et al. 1990; Silver & Erecinska, 1990 and 1992; Roberts & Chih, 1997 and 1998). In addition, the activities of Na+/H+ exchangers and gH+s are reduced by falls in pH0 (Green et al. 1988; Vaughan-Jones & Wu, 1990; Wu & Vaughan-Jones, 1997; Ritucci et al. 1998; DeCoursey & Cherny, 2000). Therefore, I examined the effects of lowering pH0 on the magnitudes of the increases in pH; observed during and after anoxia, and compared the anoxia-evoked changes in pHj observed at pH0 6.60 with those changes observed under conditions that inhibit NaVH1" exchange activity and/or gH+s. Lowering pH0 from 7.35 to 6.60 caused a 0.49 ± 0.03 pH unit fall in pH, (n = 19; see Church et al. 1998) and, once pHj had stabilized at a new resting level, anoxia evoked an internal acidification followed by increases in pH; during and after anoxia that were significantly smaller than those observed at pH0 7.35 (Fig. 4.5,4, E). Because Na+/FE exchange activity is reduced during anoxia (Chapter 3, Section 3.2.4.2), the attenuation of the rise in pHj observed during anoxia at pH0 6.60 is consistent with the suggestion that a putative gH+ 129 contributes to the rise in pHj observed during anoxia (see Section 4.2.2). Indeed, there was no difference between the increase in pHi observed during anoxia in the presence of Zn2+ vs. at pH0 6.60 (P = 0.51; Fig. 4.5^4). The effect of pH0 to reduce the increase in pHj observed after anoxia may reflect an inhibitory effect on Na+/ff~ exchange activity and/or a gu+ active in the post-anoxic period. In support, there was no difference between the rise in pHi observed after anoxia at pH0 6.60 compared with the rise in pHj observed after anoxia under reduced Na+0, NMDG-substituted, conditions in the presence of Zn" (P = 0.43; Fig. 4.55). Next, intracellular acid loads were imposed prior to and following anoxia at pH0 6.60 (Fig. 4.9,4; see Chapter 3, Section 3.2.4.2 for the effect of pH0 6.60 on rates of pHj recovery from internal acid loads imposed prior to and during anoxia). Consistent with observations made in Chapter 3, rates of pHj recovery prior to anoxia were decreased at pH0 6.60, compared to rates of pHj recovery observed at the same absolute values of pHj under control (pH0 7.35) conditions (Fig. 4.95). When acid loads were imposed immediately after anoxia, rates of pHj recovery increased, compared to rates of recovery established prior to anoxia also at pH0 6.60 (n = 9; Fig. 4.9,4, 5; P < 0.05 at each absolute value of pHj). Nevertheless, plots of the pHj dependence of the rates of pHj recovery obtained at pH0 6.60 (Fig. 4.95) indicated that rates of pHi recovery after anoxia were reduced at pH0 6.60, compared to rates established after anoxia at pH0 7.35. Qualitatively opposite results were obtained under pH0 7.60 conditions (not shown; see Sheldon & Church, 2002a). Taken together, the results are consistent with contributions from Na+/H+ exchange activity and a putative gH+ to the rises in pHj observed in rat hippocampal neurons immediately following anoxia. The data also indicate that, even at pH0 6.60, an increase in pHj still occur after anoxia, albeit slowly. 130 4.3. DISCUSSION 4.3.1. NaVH4" exchange activity after anoxia In contrast to the decline in observable Na+/H+ exchange activity that occurs in adult rat hippocampal CA1 pyramidal neurons during anoxia (see Chapter 3), the present results suggest that activation of Na+/H+ exchange occurs in this cell type immediately upon reoxygenation. Thus, pHj 'overshoots' following anoxia were reduced either when NMDG+ (but not Li4") was employed as a Na+0 substitute or when pH0 was lowered; in contrast to NMDG4", Li4" can act as a substrate for Na+/H+ exchange, and it is established that Na+/H+ exchange activity can be reduced by falls in pH0 (Green et al. 1988; Vaughan-Jones & Wu, 1990; Baxter & Church, 1996; Wu & Vaughan-Jones, 1997; Ritucci et al. 1998). It is important to note that while pHj 'overshoots' immediately after anoxia (as well as increases in pHj during anoxia) have occasionally been observed in hippocampal slice preparations (see Mabe et al. 1983; Fujiwara et al. 1992; Pirttila & Kauppinen, 1992; Melzian et al. 1996), the more usual response in these preparations comprises a fall in pHj during anoxia and a gradual restoration of pHj towards normal resting levels in the period following the return to normoxia (e.g. Silver & Erecinska, 1992; Roberts & Chih, 1997). In the present study, when anoxia was imposed under pH0 6.60 conditions, the increases in pH; during and after anoxia were greatly reduced and, similar to the changes in pHi observed in response to anoxia in vivo and in slice preparations in vitro, pR, fell during anoxia and gradually recovered upon the return to normoxia. Thus, the apparent differences in the steady-state pHj changes observed in response to anoxia in isolated neurons compared to more complex multicellular preparations are likely, in part, consequent upon the lower pH0 values (along with concurrent changes in [K+]0 and neurotransmitter release) that are associated with the 131 latter preparations and can reduce the activities of the NaVff1" exchanger and other pHj regulating mechanisms (including the Zn -sensitive, putative gH+; see below). Consistent with the steady-state pHj results, rates of pH, recovery from acid loads increased in the period immediately following the return to normoxia, and these increases were attenuated either when NMDG+ was employed as an external Na+ substitute or when pH0 was reduced. The increase in the Na+0-dependent component of pHj recovery from acid loads observed after anoxia (Figs. 4.2C and 4.3Q is also consistent with the activation of Na+/H+ exchange in the immediate post-anoxic period. Although it remains unknown what influence cellular ATP levels and/or post-anoxic changes in [Na+]j may have on NaVFf4" exchange activity following anoxia (see Chapters 5 and 6), the present results are consistent with previous reports not only in cultured postnatal rat hippocampal neurons (Diarra et al. 1999) but in other isolated neuronal preparations in which an involvement of Na+/H+ exchange in the restoration of pHj following anoxia has been demonstrated with selective pharmacological inhibitors (Vornov et al. 1996; J0rgensen et al. 1999; Yao et al. 2001). The results of the present study also support previous suggestions, made on the basis of pH0 measurements, that Na+/H+ exchange activity may contribute to the acidotic [H+]0 shift which occurs in vivo and in slice preparations during early reperfusion (Ohno et al. 1989; Obrenovitch et al. 1990). In cardiac myocytes, it has been proposed that Na+/H+ exchange activity is inhibited during anoxia/ischemia by the extracellular acidosis which occurs at this time, and that the rapid normalization of pH0 immediately upon reperfusion relieves this inhibition, thereby contributing to the activation of Na+/H+ exchange in the immediate post-anoxic period (Lazdunski et al. 1985). In the present study, however, stimulation of Na+/H+ exchange activity occurred after anoxia even when pH0 was maintained at a constant value throughout the anoxic and post-anoxic 132 periods and even when pHj immediately prior to the return to normoxia may not have been markedly decreased from the resting level observed prior to anoxia. Thus, neither a decrease in pHj during anoxia nor a return to normal pH0 values in the immediate post-anoxic period are absolute requirements for the rapid post-anoxic activation of Na+/H+ exchange in adult rat CA1 neurons. In cardiac myocytes, PKC activation also contributes to the rapid activation of Na+/H+ exchange activity during reperfusion (Ikeda et al. 1988; Yasutake & Avkiran, 1995), and the present study points to an analogous contribution from anoxia-evoked changes in the activity of the cAMP/PKA second messenger system in mediating the activation of Na+/H+ exchange in hippocampal neurons in the immediate post-anoxic period. Thus, not only do rapid increases in [cAMPJi occur in hippocampal neurons immediately upon reperfusion but these increases can be maintained for up to 60 min (reviewed by Tanaka, 2001; also see Kobayashi et al. 1977; Whittingham et al. 1984; Blomqvist et al. 1985; Domanska-Janik 1996; Small et al. 1996). In addition, our laboratory has shown previously that, under normoxic conditions, p-adrenoceptor activation, acting via cAMP and PKA, evokes a sustained increase in Na+/H+ exchange activity in acutely isolated adult rat CA1 neurons by producing an alkaline shift in the pHi dependence of the transport mechanism (Smith et al. 1998; also see Connor & Hockberger, 1984 where intracellular injections of cAMP into invertebrate neurons evoked increases in pHj). Consistent with these previous findings, in the present study there was an alkaline shift in the pHi dependence of Na+0-dependent acid extrusion following anoxia (see Figs. 4.2C and 4.3Q. Furthermore, inhibition of adenylate cyclase or PKA reduced the magnitude of the Na+0-dependent component of the pHj 'overshoot' after anoxia (also see Yao et al. 2001) whereas P-adrenoceptor activation augmented the post-anoxic rise in pHj (an effect that was blocked by propranolol, i?p-cAMPS and under conditions where NMDG+, but not Li+, was employed as a 133 Na+0 substitute). The effects of modulating the activity of the cAMP/PKA system on the increase in pHj observed immediately after anoxia are not only consistent with a contribution from Na+/H+ exchange to the post-anoxic increase in pHj but also provide an example of the potential importance of the regulation of neuronal Na+/H+ exchange activity by second messenger systems. 4.3.2. Potential contribution of a gH+ to the increases in pHj during and after anoxia In contrast to the effects of inhibiting Na+/H+ exchange which reduced the rise in pHj observed following anoxia, micromolar concentrations of Zn2+ attenuated the increases in pHj observed both during and following anoxia. Although concurrent pHi imaging and electrophysiological recordings will be required to substantiate or refute the possibility that the effects of Zn2+ may be due to the inhibition of H+ efflux through a H+-conductive pathway activated as a consequence of membrane depolarization, there is precedence for external Na+- and HCdV-independent H+ extrusion from hippocampal neurons under anoxic conditions (Ohno et al. 1989; Pirttila & Kauppinen, 1994; Diarra et al. 1999), and the possible contribution of a gH+ to the rises in pHi observed during and immediately after anoxia in the present experiments is suggested by a number of lines of evidence. First, inhibition by Zn is an identifying characteristic of gH+s (for reviews see DeCoursey & Cherny, 1994a and 2000; Eder & DeCoursey, 2001) and although Zn2+ failed to affect steady-state pHj under normoxic conditions, it attenuated the increases in pHi that occurred during and after anoxia under both Na+0-containing and reduced-Na+0 (NMDG+ substituted) 2_|_ conditions. The effect of Zn to reduce the magnitudes of the alkalinizations observed during anoxia under NMDG+-substituted conditions may reflect the established coupling between Na+/H+ exchange activity and gH+s (Fig. 4.5; DeCoursey & Cherny, 1994b; Demaurex et al. 134 1995). Thus, inhibition of Na+/H+ exchange activity under NMDG+-substituted conditions would potentially act to increase the relative contribution of the Zn2+-sensitive gH+ to acid extrusion under the depolarizing conditions that occur during anoxia. It is important to note that although Zn2+ ions modulate the activities of a variety of ion channels (for reviews see Harrison & Gibbons, 1994; Smart et al. 1994), under the constant perfusion conditions employed in the present experiments, the pH; changes evoked by anoxia are unaffected by NMDA, AMPA or GABAA receptor antagonists, or organic inhibitors of high voltage-activated Ca2+ channels (A. Diarra, C. Sheldon, and J. Church, unpublished observations; also see Chapters 5 and 7). Zn2+ has also been shown to induce falls in pHi in cultured cortical neurons in a manner dependent on Ca2+0 (Dineley et al. 2002); however, in the present study (Chapter 3), the removal of Ca2+0 failed to alter anoxia-evoked changes in pHi. Second, consistent with the steady-state pHj results, Zn2+ decreased rates of pHj recovery from acid loads imposed during and after anoxia, both under control conditions and under conditions where Na+/H+ exchange was inhibited by the substitution of NMDG+ for external Na+. Third, the fact that the Zn2+-sensitive increases in pHi observed during and after anoxia were inhibited by a reduction in pH0 is consistent with the established sensitivity of voltage-activated H+-conducting pathways to the transplasmalemmal pH gradient (DeCoursey & Cherny, 1994a and 2000). Fourth, the Zn -inhibitable internal alkalinizations that occurred during and after anoxia were associated temporally with marked and persistent increases in [Ca ]\ that, in turn, are known to occur in adult CA1 neurons in response to membrane depolarization (Rader & Lanthorn, 1989; Silver & Erecihska, 1990; Tanaka et al. 1997). In this regard, I found not only that the recovery of pHj from internal acid loads imposed during normoxia in the absence of external Na+ was faster under depolarizing (139.5 mM K+0) than under control (3 mM K 0) conditions, but also that Zn only slowed the rate of recovery of 135 pHj in the former case. Arguing against the possibility that the Zn2+-sensitive acid extrusion mechanism might be a H+-conductive pathway is the fact that Zn2+-sensitive increases in pHi after anoxia could occur even when the proton gradient across the plasma membrane was not apparently outwardly directed (i.e. pHj > pH0). However, this observation is tempered by the facts that membrane depolarization occurs during and following anoxia in rat hippocampal neurons (see Chapter 7; also Tanaka et al. 1997) and that the local [H+] in the vicinity of presumed H+-conducting channels may greatly exceed that monitored in bulk cytoplasm. Indeed, as noted by DeCoursey and Cherny (1994a) spatial or temporal pH fluctuations may activate the gu+ in situations not predictable from time-averaged, bulk pH measurements, for example, by fluorescent dyes.' 4.3.3. Synthesis of Chapters 3 and 4 Data presented in Chapters 3 and 4 have examined the contribution of alterations in the activities of pHj regulating mechanisms to the changes in pHj observed during and following transient periods of anoxia in acutely isolated adult rat hippocampal CA1 pyramidal neurons. Thus, NaVH4" exchange, a major acid-extruding mechanism under normoxic conditions in rat hippocampal neurons, becomes inhibited shortly following the onset of anoxia. In contrast, a Zn2+-sensitive alkalinizing mechanism, possibly a gB+, appears to be activated during anoxia as a consequence of membrane depolarization and contributes to acid extrusion at this time. These findings do not preclude contributions from other mechanism(s) to the rises in pHj that sometimes occurred during anoxia, such as a decreased rate of internal acid loading following the onset of anoxic depolarization (see Erecinska et al. 1991; Sanchez-Armass et al. 1994). 136 Following the return to normoxia, the presumed gH+ continues to contribute to acid extrusion. The activity of the putative gH+ may be maintained by the persistant membrane depolarization observed in mature hippocampal neurons in response to anoxia or ischemia. Na+/H+ exchange activity is enhanced following anoxia, an effect which may be mediated, at least in part, by an anoxia-induced activation of the cAMP/PKA second messenger pathway. This finding is consistent with the possibility that the neurotoxic effects associated with post ishemic activation of the cAMP/PKA pathway (e.g. Shibata et al. 1992; Small et al. 1996) may, in part, reflect an activation of Na+/H+ exchange upon reoxygenation. These results do not, however, eliminate the possibility that Na4"/!!4" exchange activity in the post-anoxic period may be regulated concurrently by more than one signaling pathway. Indeed, in mouse hippocampal CA1 neurons, anoxia-induced activation of Na+/FT exchange can be reduced by inhibiting either PKC or PKA (Yao et al. 2001) and, in recent studies, Na4"/!!4" exchange activity in brainstem neurons has been found to be regulated by reactive oxygen species (the production of which is enhanced following periods of anoxia or ischemia; Lipton, 1999; Mulkey et al. 2004; also see Wei et al. 2001 in cardiac myocytes). In addition, it is noteworthy that gH+s can couple to Na+/H+ exchange (DeCoursey & Cherny, 1994b; Demaurex et al. 1995), such that the activation of a gH+ during anoxia would act as an 'acid-relief valve' to limit the potentially detrimental activation of forward Na+/H+ exchange occurs in the immediate post-anoxic period (Vornov et al. 1996). Conversely, the neurotoxic effects associated with micromolar concentrations of Zn24" (e.g. Choi & Koh, 1998; Weiss et al. 2000; Dineley et al. 2003) may, in part, reflect an inhibition of gH+s and augmented Na+/H+ exchange activity upon reoxygenation. Indeed, the activation of Na+/FT exchange activity following anoxia may act to increase the internal Na4" load in the period immediately after anoxia (see Chapter 5) and thereby, for example, worsen cellular energy state 137 (Fried et al. 1995; Chinopoulos et al. 2000), potentiate NMDA receptor-mediated responses (Yu & Salter, 1998; Manzerra et al. 2001), and/or promote the reversal of plasmalemmal Na+/Ca2+ exchange (Kiedrowski et al. 1994). Originally described by Bevensee et al. (1996), CA1 pyramidal neurons can be classified into those exhibiting "high" and "low" levels of NaVFT*" exchange activity under steady-state conditions (also see Smith et al. 1998; Brett et al. 2002a), a finding which supports previous illustrations of intrinsic variations between CA1 pyramidal neurons (e.g. subtle morphological differences and differences in the expression of calcium-binding proteins; Amaral & Witter, 1995; Morris et al. 1995). Notably, however, anoxia-induced changes in NaVFT4" exchange activity were observed in both "high" and "low" pHj cells. It was apparent that neurons expressing "high" levels of NaVFT1" exchange activity prior to anoxia (i.e. "high" pHj neurons; see Bevensee et al. 1996) were more sensitive to the actions of anoxia to reduce NaVFT1" exchange activity than were "low" pHj neurons (which have low levels of NaVFT" exchange activity). Conversely, following anoxia, activation of Na+/H+ exchange activity was more marked in "low" pHj neurons that exhibit relatively "low" levels of NaVFT1" exchange activity prior to anoxia. Although the potential functional significance of these findings is unclear, a number of events that are known to occur in response to anoxia appear dependent on pHj. For example, neurons with pre-anoxic pHj values greater than ~ 7.20 are more likely to undergo hypoxia-induced depolarizations (vs. hyperpolarizations; Cowan & Martin, 1995). There is a growing body of evidence that pharmacological inhibition of Na+/H+ exchange effectively protects against anoxia- and ischemia-induced neuronal injury (e.g. Vornov et al. 1996; Kuribayashi et al. 1999; Phillis et al. 1999). The results presented in Chapters 3 and 4 suggest that any neuroprotective actions of NaVH4" exchange inhibitors would likely be realized in the period immediately following anoxia or ischemia. Whether a similar benefit might be 138 conferred in mature rat hippocampal CA1 pyramidal neurons awaits the identification of pharmacological inhibitors of Na+/H+ exchange in this highly vulnerable cell type. 139 Fig. 4.1. Effects of external Na+ substitutions on the increase in pHj observed following 5 min anoxia. A, the magnitude of the internal alkalinization observed following 5 min anoxia under control conditions (solid line) was reduced under NMDG+0-(open circles) but not Li+0-(filled circles) substituted conditions. Records shown were obtained under Hepes-buffered conditions (pH0 7.35) at 37°C from three different neurons with similar resting pHj values immediately prior to the induction of anoxia. B, effects of changes in perfusate composition on the increase in pH, observed following 5 min anoxia. All experiments were conducted under HCCV/COi-ftee, Hepes-buffered conditions (pH0 7.35, 37°C); error bars are s.E.M. * denotes a statistically significant difference (P < 0.05) compared to control or Li+0-substituted conditions. N.S. indicates no significant difference (P = 0.76) between the increase in pHj observed following anoxia under Na+0-containing compared with Li+0-substituted conditions. 140 e X Q. 0.30 Time (min) Control -o— Reduced Na+0, NMDG+-substituted -•— Reduced Na+„, Li+-substituted 0.25 i lo Q. .t; .£ § .i 1. 0.20 1— CD CO £ x ° £ 11 ro 0 0.15 0.10 0.05 -| 0.00 N.S. Control (n = 38) ezzzza Reduced Na+0, NMDG+-substituted (n = 20) Reduced Na+0, Li+-substituted (n = 13) 141 Fig. 4.2. Recovery of pH; from internal acid loads imposed immediately after anoxia. A, following the first NH4+-induced intracellular acid load, pH, was allowed to recover. A second acid load was then applied after 5 min anoxia. The rate of recovery of pHj was increased in the post-anoxic period, compared to the rate of pHj recovery observed prior to anoxia. B, rates of pHj recovery prior to (filled symbols) and immediately after (open symbols) 5 min anoxia under control (Na+0-containing; circles) and reduced-Na+0, NMDG+-substituted (triangles) conditions. Under both conditions, rates of pHj recovery were increased following anoxia (P < 0.05 at each absolute value of pHj); data points were obtained from 17 and 5 experiments, respectively, of the type shown in A. Continuous lines represent the weighted non-linear regression fits to the data points indicated for each experimental condition. Where missing, standard error bars lie within the symbol areas. C, the Na+0-dependent component of pHj recovery prior to (filled circles) and after (open circles) anoxia revealed by plotting the differences between the regression fits (shown in B) obtained under Na+0-containing and reduced-Na+0 (NMDG+-substituted) conditions. In A -C, data were obtained at 37°C during perfusion with Hepes-buffered media at pH0 7.35. H2 143 Fig. 4.3. Recovery of pHj from internal acid loads imposed immediately after anoxia in "low" pHj neurons. A, the pHj dependencies of rates of pHj recovery prior to (filled circles) and after (open circles) 5 min anoxia under control conditions (normal [Na+]0). Continuous lines represent the weighted nonlinear regression fits to the data points indicated for each experimental condition (n = 5 in each case). B, rates of pHi recovery from internal acid loads imposed following anoxia under normal Na+0-containing conditions (open bar) were faster than those observed prior to anoxia under normal Na+0-containing conditions (filled bar). Also shown are rates of pHj recovery observed prior to (hatched bar) and following (cross-hatched bar) anoxia under reduced-Na+0, NMDG+-substituted conditions. Rates of pHj recovery shown were determined at a common test pHj of 6.80. * denotes a statistically significant difference (P < 0.05). C, the Na+0-dependent component of pHj recovery prior to (filled circles) and after (open circles) anoxia revealed by plotting the differences between the regression fits to pHj vs. dpHj/d/ plots obtained under Na+0-containing and reduced-Na+0 (NMDG+-substituted) conditions. In A -C, data were obtained at 37°C during perfusion with Hepes-buffered media at pH0 7.35. 144 B 0.012 vc/> 0.009 c I 0.006 QL Q. 0.003 Prior to anoxia After anoxia 0.000 6.2 6.4 6.6 6.8 7.0 7.2 7.4 7.6 s.« T3 C C 3 CD CL CD •a 0.008 0.006 H 0.004 4 0.002 4 0.000 Prior to anoxia o— After anoxia 0.010 0.008 CO CO 1 0.006 X - 0.004 X •o 0.002 0.000 Before anoxia (n = 5) GSSSS Before anoxia, NMDG+-substituted (n = 5) ' ' After anoxia (n = 5) After anoxia, NMDG+-substituted (n = 5) i r 6.2 6.4 6.6 6.8 7.0 7.2 7.4 PHi 145 Fig. 4.4. Effects of modulating the activity of the cAMP/PKA pathway on the pHj response to anoxia. A, the magnitude of the internal alkalinization observed following 5 min anoxia under control conditions (filled circles) was increased in the presence of the [3-adrenoceptor agonist isoproterenol (10 pM; open circles) and reduced in the presence of the PKA inhibitor i?p-cAMPS (50 u.M; open triangles). The records shown were obtained under Hepes-buffered conditions (pH0 7.35) at 37°C from three different neurons with similar resting pHj values immediately prior to the induction of anoxia. Pharmacological treatments were applied for >10 min prior to the induction of anoxia and were maintained throughout the records shown. B, effects of the test conditions shown in the figure on the increase in pHj observed after 5 min anoxia. All experiments were performed under nominally HC037C02-free, Hepes-buffered conditions at 37°C, pH0 7.35; error bars are S.E.M. * denotes a statistically significant difference (P < 0.05) compared to control (shown in the first column). j denotes a statistically significant difference (P < 0.05) compared to the value obtained in the presence of 10 pM isoproterenol (shown in the fourth column). 146 A B i—i—i—i Control 50 pM Rp-cAMPS 10 pM isoproterenol Q_ *i .E § <D X •S & OJ ro £ x o c ro •a 3 E 5 ro o 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00 1 nTTTTI Control (n = 38) 50 pM Rp-cAMPS (n= 12) Reduced Na+, NMDG+-substituted + 50 pM Rp-cAMPS (n = 3) 10 pM isoproterenol (n = 9) 10 pM isoproterenol + 20 pM propranolol (n = 4) 10 pM isoproterenol + 50 pM Rp-cAMPS (n = 4) Reduced Na+, NMDG+-substituted + 10 pM isoproterenol (n = 11) Reduced Na+, Lf-substituted + 10 pM isoproterenol (n = 5) 147 Fig. 4.5. Effects of changes in perfusate composition on the increases in pHj observed during (A) and following (B) 5 min anoxia. All experiments were performed under nominally HC037C02-free, Hepes-buffered conditions at 37°C; error bars are s.E.M. * denotes a statistically significant difference (P < 0.05) compared to control, normal Na+0-containing conditions at pH0 7.35 (shown in the first column in both A and B). f denotes a statistically significant difference (P < 0.05) compared to the value obtained under reduced-Na+0, NMDG+-substituted conditions (shown in the second column in both ,4 and B). 148 B x Q. g 0) 'E CO TO 0.10 S 0.08 x a. co o c CO O) c 3 0.06 0.04 0.02 0.00 X 11 * * * 0.30 n i «i „„, Q- •« 0.25 0.20 *= « 0.15 (D •o 3 'E D) CO CO X o c ro g> 0.10 = 0.05 4 0.00 1 t ii • I 1 T Control (n = 38) zzzzzz Reduced Na+0, NMDG+-substituted (n = 20) Essssi 100 - 500 uM Zn2+ (n = 17) Reduced Na+0, NMDG+-substituted + 100-500 uM Zn2+ (n = 9) pH0 6.60 (n= 10) 149 Fig. 4.6. Irrfluence of Zn on the recovery of pHj from internal acid loads imposed immediately after anoxia. A, following a 5 min period of anoxia, an internal acid load was imposed under control (Zn2+-free, NaVcontaining) conditions. At the peak of the acidification, the perfusate was changed to a reduced-Na+, NMDG+-substituted medium containing 100 pM Zn2+ (a to b). From b to c, Zn2+ was removed and pHj recovery was allowed to proceed under reduced-Na+, NMDG+-substituted conditions. At c, the neuron was reperfused with control medium. B, rates of pHj recovery from internal acid loads imposed immediately after anoxia during perfusion with reduced-Na+, NMDG+-substituted medium containing 100 uM Zn2+ (open bar); reduced-Na+, NMDG+-substituted medium (diagonal hatching); Na+-containing medium in the presence of 100 uM Zn2+ (cross hatching); and 2"f" + control (Zn -free, Na -containing) medium (filled bar). The pHo.s values at which rates of pHj recovery were determined were ~ 6.7, 6.9, 7.1 and 7.2, respectively; error bars are S.E.M. | denotes a 2_|_ statistically significant difference (P < 0.05) compared to control (Zn -free, normal Na 0). * denotes a statistically significant difference (P < 0.05) compared to the value obtained under reduced-Na+0, NMDG+-substituted conditions in the presence of 100 uM Zn2+. In A and B, data were obtained at 37°C during perfusion with Hepes-buffered media at pH0 7.35. 150 6 7.9 7.5 Q- 7.1 6.7 H 6.3 NH4+ Anoxia c I CL I CL •o —1 1 1 1 1 1 5 10 15 20 25 30 Time (min) 0.004 -i 0.003 0.002 0.001 0.000 • NMDG+-substituted + 100 u.M Zn2+(r> =11) 177777* NMDG+-substituted (n = 7) 100 p.M Zn2+(n = 7) Na+0-containing, Zn2+ free conditions (n = 16) 151 Fig. 4.7. Effect of high-[K+]0 on pHj recovery from intracellular acid loads imposed under normoxic Na+0-free, nominally HCCV-free, Hepes-buffered conditions (pH0 7.35). A, under control conditions (3 mM KC1; solid line), an internal acid load was applied by the NH4+ prepulse technique and pH; recovered. The rate of pHj recovery was faster under high-[K+]0 conditions (139.5 mM KC1; open circles) compared to control, and the brief application of 100 pM Zn2+ slowed the rate of pHj recovery under high K+0-conditions (filled circles) but had no effect at normal [K+]0 (filled triangles). Records were obtained from four different neurons which exhibited similar minimum pHj values in response to the NH4+ prepulse (with the exception of the control response, NH4+ prepulses have been omitted for clarity). B, rates of pHj recovery (± S.E.M.) from internal acid loads imposed during perfusion with reduced-Na+, NMDG+-substituted medium containing 3 mM K+ either in the absence (open bar) or presence (diagonal hatched bar) of 100 pM Zn2+, and under reduced-Na+0, 139.5 mM K+ substituted conditions in the 9+ absence (filled bar) or presence (cross-hatched bar) of 100 pM Zn . The pHo.5 values at which rates of pHj recovery were estimated were ~ 6.6, 6.8, 6.8 and 6.9, respectively. * denotes a statistically 9+ significant difference (P < 0.05) compared to the value obtained in the presence of 100 pM Zn . 152 B x Q. Time (min) 0.005 i 'in in 0.004 -c 0.003 -X 0.002 -x~ Q. •o 0.001 -0.000 -I 1 J 3 mM K o, 0 Zn (n = 5; —) VTZft 3 mM K+o, 100 uM Zn2+ (n = 5; •) >>H 139.5 mM K+o, 0 Zn2+ (n = 7; O) 139.5 mM K+o, 100 uM Zn2+ (n = 7; •) 153 Fig. 4.8. Representative traces of the effects of inhibitors of P-type H+,K+-ATPase and V-type H+-ATPase activity on pHj recovery from intracellular acid loads imposed under high-[K+]0 conditions (pH0 7.35). A, an internal acid load was imposed using the NH4+ prepulse technique and 50 pM omeprazole was added to the perfusate during the recovery of pHj. B and C, 500 uM SCH-28080 (B) or 2 pM bafilomycin Ai (Q were added to the perfusate immediately upon the removal of NH/. Neither omeprazole, SCH 28080 nor bafilomycin A] had any effect on the recovery of pHi from internal acid loads imposed under the high-[K+]0 (139.5 mM KC1), Na+0-free (NMDG+-substituted) nominally HCCV-free, Hepes-buffered conditions (pH0 7.35) employed throughout all the traces shown. Time (min) Time (min) Time (min) 155 Fig. 4.9. Effects of reduced pH0 on rates of pHj recovery from acid loads imposed in the immediate post-anoxic period. A, an initial acid load was imposed at pH0 6.60 and pET was allowed to recover. The neuron was then exposed to anoxia for 5 min and a second acid load was applied after the return to normoxia. B, rates of pHj recovery from acid loads imposed prior to (filled symbols) and immediately after (open symbols) anoxia at pH0 6.60 (squares) and pH0 7.35 (circles). Continuous lines represent the weighted non-linear regression fits to the data points indicated for each experimental condition. Data collected at pH0 6.60 were obtained from 9 experiments of the type shown in A; where missing, standard error bars lie within the symbol areas. 156 pH0 7.35 A B • Before anoxia o After anoxia 0 5 10 15 20 25 30 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 7.6 Time (min) pHj 157 CHAPTER FIVE CHANGES IN [Na4} INDUCED BY ANOXIA IN ISOLATED RAT HIPPOCAMPAL NEURONS: ROLE OF NaVH" EXCHANGE ACTIVITY 5.0. INTRODUCTION The contribution of NaVH4" exchange to the increases in [Na ]i that occur under conditions of metabolic inhibition has been most extensively established in cardiac myocytes (see Pike et al. 1993; Wu & Vaughan-Jones, 1994). In response to a marked intracellular acidosis and the accumulation regulatory factors, such as catecholamines and lysophosphatidlycholine, NaVH4" exchange activity is activated by periods of ischemia (Avkiran & Haworth, 1999; Karmazyn et al, 1999). Although this acts to restore pHj, there is a concomitant rise in internal Na+ that, in turn, leads to reversal of the Na+/Ca2+ exchanger and a subsequent elevation of [Ca2+]i (e.g. Tani & Neely, 1989; An et al. 2001). While pharmacological inhibition of NaVH4" exchange reduces the extent of ischemic damage in cardiac myocytes, it remains unclear whether the cardioprotective effects result from a reduction in acid extrusion, Na ; accumulation or Ca entry, or through additional so far unidentified mechanism(s) (see Schafer et al. 2000; Avkiran, 2001). Recent studies have also suggested that Na+0 and HC03~-dependent pHj regulating mechanisms, specifically Na+/HC03" cotransport, may act alongside NaVH4" exchange to promote cell damage in cardiac cells in response to ischemia (e.g. Khandoudi et al. 2001; Lemars, 2001). The results presented in Chapters 3 and 4 suggest that NaVFE exchange activity in rat hippocampal neurons, while inhibited during anoxia, is increased in the immediate post-anoxic period, raising the possibility that NaVH4" exchange may contribute to the increases in [Na+]i observed previously in response to anoxia in isolated neurons (Friedman & Haddad, 1994a; Chen et al. 1999; Diarra et al. 2001). In support, changes in NaVFE exchange activity under normoxic conditions (e.g. 158 in response to imposed internal acid loads) are associated with transient increases in [Na+]j in neurons (Moody, 1981; Chesler, 1986; Schwiening & Thomas, 1992; also see Chapter 6). In addition, following transient focal ischemia, Kuribayashi et al. (1999) attributed the neuroprotective effects of Na+/H+ inhibitors to reductions in tissue Na+ and water content. Although less well-characterized, Na+-dependent QTHCCV exchange is similarly capable of eliciting small increases in [Na+]i in rat hippocampal neurons (Rose & Ransom, 1997; also see Thomas, 1977 for similar findings in snail neurons). Thus, the activities of these transport mechanisms may contribute to the potentially detrimental increases in [Na+]i observed in hippocampal neurons in response to anoxia. In the present study, the Na+-sensitive fluorophore, SBFI, was used to: i) characterize the changes in [Na+]j observed during and following anoxia in isolated rat hippocampal neurons; and ii) examine the potential contribution of pHj regulating mechanisms to the observed increases in [Na4], 5.1. MATERIALS AND METHODS 5.1.1. Experimental preparation Unless otherwise noted, primary cultures of hippocampal neurons obtained from 2-4 day postnatal Wistar rats were used. 5.1.2. Recording techniques The methods used to load acutely isolated adult rat hippocampal CA1 neurons with SBFI were presented in Chapter 2. To load cultured postnatal hippocampal neurons with SBFI, coverslips with neurons attached were incubated with 10 uM SBFI-AM (in the presence of 0.10% Pluronic 159 F-127 and 5 mg ml"1 of bovine serum albumin) for 120 - 180 min at 32°C. Following loading, coverslips with neurons attached were mounted into a temperature-controlled perfusion chamber so as to form the base of the chamber and neurons were superfused at a rate of 2 ml min"1 for 15 min with the initial experimental solution (at 37°C) before the start of an experiment. Anoxia-evoked changes in [Na+]i were measured using the dual-excitation ratiometric technique, as detailed in Chapter 2. Neurons loaded with SBFI were alternately excited at 334 and 380 nm and fluorescence emissions were collected from ROIs placed on individual neuronal somata. Raw emission intensity data at each excitation wavelength were corrected for background fluorescence prior to calculation of the ratio (BIHA/BI^O). A one-point calibration technique ([Na+]j = 10 mM) was employed to convert 5/334/5/380 ratio values into [Na+]i values as described (Chapter 2; also see Diarra et al. 2001). It is important to note that, loaded into rat hippocampal neurons, SBFI possesses a negligible sensitivity to K+ (Rose & Ransom, 1997) and a slight sensitivity to changes in pHj, with intracellular acidifications and alkalinizations resulting in apparent decreases and increases in [Na+]j, respectively (see Rose & Ransom, 1997; Nett & Deitmer, 1998). A formal assessment of the pH-sensitivity of SBFI, conducted during the course of this thesis (Diarra et al. 2001) established that the effects of [Na+]i values estimated with SBFI are unlikely to affect the interpretation of results presented here. 5.1.3. Internal ATP determination Cellular ATP content was measured using the Molecular Probes ATP determination kit (see Chapter 3 for details). Transient periods of anoxia were induced in cultured neurons under conditions identical to those used for the microspectrofluorimetric measurements of [Na+]j (i.e. conditions of constant perfusion, 37°C, pH0 7.35). Prior to and at given time intervals during or 160 following anoxia, neurons were lysed by the addition of 40 u.1 of a solution containing 10 mM Tris buffer (pH 7.5), 0.1 M NaCl, 1 mM EDTA and 0.01% Triton X-100 in the presence of a cocktail of protease inhibitors (Roche Diagnostic Canada, Laval, QB) and homogenized with a cell scraper. Ten microlitre aliquots were then removed and used to measure ATP. ATP levels are reported as a percentage decline compared to paired pre-anoxic measurements. 5.1.4. Experimental procedures and data analysis Changes in [Na+]j observed during anoxia (A[Na+]inuring)) were measured as the difference between the pre-anoxic resting [Na+]i value and the peak [Na+]j value observed during a 5 min period of anoxia (see Fig. 5.1). Changes in [Na+]i occurring after anoxia were examined in separate experiments. As illustrated in Fig. 5.3,4, neurons were exposed to 5 min anoxia and, upon the return to normoxia, Na+,K+-ATPase activity was inhibited for 7 min (by perfusion with [K+]-free medium or the application of 500 pM ouabain), revealing continued Na+ entry at this time. The magnitude of the increase in [Na+]j observed after anoxia under these conditions (A[Na+]j(after)) was measured as the difference between the [Na+]j value observed at the end of 5 min anoxia and the [Na+]j value observed at the end of the 7 min exposure to 0 [K+]0 or 500 uM ouabain. Data are reported as mean + s.E.M. with the accompanying n value referring to the number of neuronal populations (i.e. coverslips) from which data were obtained. As detailed in the Results, the magnitudes of the rises in [Na+]j observed during and after anoxia were related to the number of days neurons had been maintained in culture. Therefore, experiments were routinely performed on cultures maintained for 6 - 10 days in vitro (DIV) and, where noted, were repeated using 11-14 DIV neuronal cultures. Measurements of anoxia-evoked increases in 161 [Na+]i observed under a given test condition were normalized to the corresponding [Na+]j measurement made in experiments performed in the absence of a test condition using age-matched sister cultures (yielding Normalized A[Na+]i(during) and A[Na+]i(after) values; see Tables 5.1 and 5.2). Statistical comparisions were performed by comparing absolute [Na+]j measurements (i.e. not normalized A[Na ]i(during) and A[Na ] i(after) values) made under a given test condition to measurements made in age-matched sister cultures under control conditions using Student's two-tailed unpaired Mests. Where appropriate, additional statistical analysis was performed with one-way ANOVA. Significance was assumed at the 5% level. 5.2. RESULTS 5.2.1. Anoxia-induced increases in \Na+\ in acutely isolated adult rat hippocampal CA1 pyramidal neurons Initially, I explored the feasibility of using acutely isolated adult rat hippocampal CA1 pyramidal neurons to examine the contribution of NaVFT1" exchange to the changes in [Na+]i evoked by anoxia. Although 5 min anoxia induced an increase in [Na+]j of 22 ± 16 mM (n = 4), resting [Na+]i in these cells was elevated (27 ± 7 mM) and [Na+]i failed to recover to pre-anoxic values upon the return to normoxia. A similar pattern of changes has been observed by others in acutely isolated hippocampal neurons using the non-ratiometric Na+-sensitive fluorophore, Sodium Green (Friedman & Haddad, 1994a) and may reflect, at least in part, the limited viability of acutely isolated adult neurons and the difficulty with which sufficient SBFI-derived fluorescent signals could be obtained. Therefore all subsequent experiments were performed using postnatal hippocampal neurons in primary culture. Importantly, the pHj response to anoxia in these 162 cultured neurons is similar in nearly all respects to the response in acutely isolated adult rat CA1 neurons described in Chapters 3 and 4: thus, in cultured postnatal hippocampal neurons, NaVFT1" exchange activity is reduced during and activated immediately following anoxia and a Zn2+-sensitive, putative gH+ appears to contribute to acid extrusion during and after anoxia (Diarra et al. 1999). 5.2.2. Anoxia-induced increases in [Na+]i in cultured postnatal rat hippocampal neurons Prior to anoxia, resting [Na+], was 11 ± 1 mM, (n = 444), a value similar to those previously reported by others in either isolated neuronal (Pinelis et al. 1994; Rose & Ransom, 1997; Silver et al. 1997; Chen et al. 1999; Diarra et al. 2001) or brain slice (Guatteo et al. 1998; Calabresi et al. 1998) preparations. As illustrated in Fig. 5AA, a 5 min period of anoxia induced a -15 - 40 mM increase in [Na+]i that began - 90 s into the anoxic insult and recovered to pre-anoxic levels within -6-10 min after the return to normoxia. There was a positive correlation between the magnitude of the anoxia-evoked increase in [Na+]i and the length of time hippocampal neurons had been maintained in culture (Fig. 5.15). When anoxia was imposed under reduced [Na+]0, NMDG+-substituted conditions, the increase in [Na+]i was inhibited (n = 8; Fig. 5.1C), indicating a requirement for Na+ entry. In Chapters 2 and 3, I reported that media containing 1 - 2 mM sodium dithionite have Po2 values <1 mm Hg and that the pHj changes observed during exposure to these media reflect reductions in P0l and are not secondary to any additional properties of the O2 scavenger. In the present series of experiments, the possibility that dithionite-containing solutions may induce changes in [Na+]j via mechanisms unrelated to its O2 scavenging property was assessed in two ways. First, solutions containing 1 or 2 mM sodium dithionite were bubbled vigorously with air 163 for 20 - 30 min, which elevated P02 in these solutions from <1 mm Hg to 152 ± 6 mm Hg (n = 5), as measured with an oxygen electrode (ISO2; World Precision Instruments Inc., Sarasota, FL; also see Carpenter et al. 2000). The increase in [Na+]j observed during exposure to dithionite -containing media equilibrated with air was 3 ± 1 mM (n = 7), significantly (P < 0.05) less than the 23 ± 3 mM (n = 23) increase observed in age-matched cultures exposed to dithionite-containing media equilibrated with 100% Ar (Fig. 5.1,4). As an additional control, standard Hepes-buffered medium was bubbled vigorously with ultra-pure Ar for >18 h, reducing PQ2 in the medium to <1 mM Hg (see Chapter 3; Section 3.2.2). The resultant increase in [Na+]j observed upon exposure to this medium was not different to that observed when P0L was reduced to <1 mM Hg by the addition of sodium dithionite (Fig. 5.15). Thus, the [Na+], changes evoked by exposure to media containing 1 - 2 mM sodium dithionite largely reflect reductions in P0L and are not secondary to any additional properties of the O2 scavenger. As previously described (see Silver et al. 1997), the accumulation of Na+j in neurons during anoxia in part reflects reduced Na+,K+-ATPase activity. In agreement, 5 min applications of standard Hepes-buffered media containing 500 pM ouabain or [K+]0-free medium under normoxic conditions evoked increases in [Na+]i of 27 ± 4 mM (n = 7) and 26 ± 5 mM {n = 7), respectively (Fig. 5.2,4; also see Rose & Ransom, 1997)7. In addition, Na+j accumulation during anoxia occurred at times at which cellular ATP levels were reduced. After 3 min anoxia, internal ATP had fallen to 34 ± 7%, with a further decrease to 24 ± 8% of pre-anoxic values at the end of 7 Five min applications of 1 pM ouabain evoked an increase in [Na+]j that was significantly smaller than that observed in response to 500 pM ouabain (the magnitude of the increase in [Na+]j observed after 5 min 1 pM ouabain was 4 ± 1 mM; n = 5; P < 0.05 compared to the increase in [Na^j observed after 5 min 500 pM ouabain), consistent with the possibility that the maintenance of resting [Na^ in rat hippocampal neurons relies on the activity of a Na+,K+-ATPase isoform which possesses a low-affinity ouabain binding site (Juhaszuva & Blaustein, 1997). 164 5 min anoxia (Fig. 5.25; also see Gleitz et al. 1996). Neuronal cultures were then incubated with 10 mM creatine for >2 h to increase intracellular phosphocreatine levels and delay anoxia-induced falls in ATP (see Chapter 3, Section 3.2.4.3; also see Balestrino et al. 2002). In creatine pretreated neurons, there was a significant attenuation of the fall in ATP observed after 3 min anoxia under control conditions (Fig. 5.25). Furthermore, the magnitude of the increase in [Na+]j observed after 3 min anoxia was reduced by -55% compared to the increase observed in age-matched sister cultures not treated with creatine (Fig. 5.2Q. Following 5 min anoxia, creatine pretreatment failed to limit significantly either the fall in internal ATP or the increase in [Na+]j (Fig. 5.25, Q. For each of the experimental series described above, similar effects were observed in neurons maintained for 11 - 14 DIV (not shown). These results suggest that the maintenance of resting [Na+]j in cultured rat hippocampal neurons is dependent upon the activity of the Na+,K+-ATPase and that the accumulation of [Na+]j during anoxia likely reflects, at least in part, a reduced activity of the pump consequent upon decreases in internal ATP. As illustrated in Figure 5.1,4, [Na+]i recovered to pre-anoxic values within -6-10 min following the return to normoxia. This recovery likely reflects the resumption of Na+,K+-ATPase activity following anoxia (see Ekholm et al. 1993; van Emous et al. 1998). Thus, 5 min after the return to normoxia, internal ATP levels increased by -10% from values measured at the end of anoxia (to 33 ± 7% of pre-anoxic values; n = 4). In addition, inhibition of Na+,K+-ATPase activity ([K+]0-free conditions for a 7 min duration) prevented the recovery of [Na+]i and revealed a secondary post-anoxic increase in [Na+]i that, as in the case of the rise in [Na ]j observed during anoxia, was related to the length of time that neurons had been maintained in culture (Fig. 5.3,4, 5). Similar results were observed if Na+,K+-ATPase activity was blocked after anoxia with 500 uM ouabain (see Fig. 5.35). Once Na+,K+-ATPase activity was re-established by perfusion with 165 standard medium containing 3 mM [K+]0, [Na+]j recovered to pre-anoxic values (Fig. 5.3^4). The secondary post-anoxic increase in [Na+], observed during Na+,K+-ATPase inhibition was blocked when NMDG+ was substituted for external Na+ (n = 4; Fig. 5.3,4), indicating continued Na+ entry in the immediate post-anoxic period. Taken together, these results suggest that: i) the accumulation of Na* during anoxia in cultured postnatal rat hippocampal neurons likely reflects, in part, reduced Na+,K+-ATPase activity, consequent upon decreases in internal ATP; ii) upon the return to normoxia, the resumption of Na+,K+-ATPase activity mediates the recovery of [Na+]i to pre-anoxic levels, and iii) the increases in [Na+]i observed during and after anoxia are strongly dependent on the influx of Na+ from the extracellular space. The pHi measurements described in Chapters 3 and 4 indicated that Na+/Ff" exchange activity in rat hippocampal neurons is inhibited during anoxia and activated in the immediate post-anoxic period. In the following section, I examined the potential contribution of Na+/FT exchange activity to the increases in [Na+]j observed during and following anoxia. As described by van Emous et al. (1998), by inhibiting Na+,K+-ATPase activity, the contribution of Na+/H+ exchange activity to Na+ entry occurring following anoxia could be examined effectively. 5.2.3. Role of Na47Fi+ exchange activity The examination of the contribution of Na+/H+ exchange to anoxia-evoked changes in [Na+]i in rat hippocampal neurons is complicated by the lack of a selective pharmacological inhibitor (see Section 1.4.1; Raley-Susman et al. 1991; Schwiening & Boron, 1994; Baxter & Church, 1996). In Chapters 3 and 4, the effects of anoxia on Na+/H+ exchange activity, and the contribution of these changes in transport activity to the pHj changes observed during and after anoxia, were inferred by determining the Na+0- (and Li+0-) dependency of the anoxia-evoked changes in steady-state pHi and the Na+0-166 dependency of rates of pHj recovery from imposed internal acid loads. Because this approach was not feasible in the present study, I sought to indirectly assess the role of NaVFT exchange activity in the production of anoxia-evoked changes in [Na+]i by testing a number of maneuvers that have previously been found to influence NaVFE exchange activity in rat hippocampal neurons. Harmaline is reported to be a non-selective inhibitor of NaVFE exchange activity in rat hippocampal neurons (Raley-Susman et al. 1991). In agreement, examined under HCCV-free, Hepes-buffered normoxic conditions, harmaline pretreatment (200 uM) reduced rates of pHj recovery from internal acid loads imposed using the NH4+ prepulse technique (see Fig. 5.44, inset). However, consistent with the findings presented in Chapter 3 and 4 which, on the basis of pHj measurements, suggested that NaVFE exchange activity was inhibited during anoxia and stimulated immediately following anoxia, pretreatment with harmaline failed to reduce the rise in [Na+]i observed during 5 min anoxia in either 6 - 10 or 11 - 14 DIV neurons (Table 5.1; Fig. 5.4,4) but reduced the increase in [Na+]j observed following anoxia in both 6-10 and 11-14 DIV neuronal cultures (Table 5.2; Fig. 5.4/4). The pHi measurements presented in Chapter 4 suggested that the activation of NaVFE exchange activity in the immediate post-anoxic period can be inhibited by an extracellular acidosis or inhibition of the cAMP/PKA pathway and, conversely, that exchange activity can be further enhanced by an external alkalosis (see Diarra et al. 1999; Sheldon & Church, 2002a). Consistent with a contribution of NaVFE exchange to Na+ influx in the immediate post-anoxic period, exposure of neurons to pH0 6.60 conditions or 50 uM i?p-cAMPS reduced the magnitude of the increase in [Na+]j observed at this time by 25 - 40% (Table 5.2; Fig. 5.4,4). Conversely, an extracellular alkalosis enhanced the increase in [Na+]j observed following anoxia (Table 5.2; Fig. 5.4^4). 167 The pHj measurements presented in Chapters 3 and 4 also suggested that a Zn2+-sensitive voltage-activated H+ conductance (gH+) may contribute to the dissipation of the internal acid load imposed by 5 min anoxia in rat hippocampal neurons (also see Diarra et al. 1999). It has previously been suggested that inhibition of gH+s (e.g. with Zn2+) may increase the contribution of Na+/H+ exchange to acid extrusion in non-neuronal cell types (Demaurex et al. 1995). In the present study, Zn2+ might therefore be expected to promote Na+ influx at a time when NaVFT" exchange is active (i.e. in the immediate post-anoxic period) but have no effect during anoxia (i.e. at a time when NaVH4" exchange is inhibited). Indeed, exposure of neurons to 100 uM Zn2+ failed to affect significantly the increase in [Na+]j during anoxia (Table 5.1); in contrast, applied immediately after anoxia under K 0-free conditions, Zn significantly enhanced the increase in [Na+]j (Table 5.2; Fig. 5.45). The ability of Zn24" to augment the increase in [Na+]j observed following anoxia was blocked under pH0 6.60 conditions (Table 5.2; Fig. 5.45), consistent with Zn24" indirectly enhancing Na+ influx through NaVH4" exchange. 5.2.4. Role of HCCV-dependent mechanisms The potential contribution of HCO3 -dependent pHj regulating mechanisms to anoxia-evoked increases in [Na+]j was examined by measuring the changes in [Na+]i observed during and after anoxia under HCCV/CCVbuffered conditions. As reported previously (Rose & Ransom, 1997), the transition from a HCCV-free, Hepes-buffered medium to a HCCV/CCVbuffered medium (pH0 constant at 7.35) caused a small (~3 mM) increase in [Na+]j, consistent with the activation of Na+-dependent CT/HCO3" exchange. Under HCCV/CCVbuffered conditions, however, the increase in [Na+]j observed during anoxia was not significantly different to that observed in age-matched sister cultures under HCCV-free, Hepes-buffered conditions (Table 5.1). The addition 168 of 200 uM DIDS under HCOy/CO^-buffered conditions failed to limit the rise in [Na+]j seen during anoxia (Table 5.1), further suggesting that HCOy-dependent pHj regulating mechanisms do not contribute significantly to the increase in [Na+]i observed during anoxia in rat hippocampal neurons. In contrast, the magnitude of the increase in [Na+]i observed after anoxia was consistently greater under HCOV/CCh-buffered conditions than under nominally HCdV-free, Hepes-buffered conditions in 11 - 14, but not 6 - 10, DIV neuronal cultures (Fig. 5.5,4, 73); in neurons 11-14 DIV, the magnitude of the increase in [Na+]i observed following anoxia was 27 + 2 and 50+13 mM under Hepes- and HC037C02-buffered conditions, respectively (n = 5 in each case). The HCOy-dependent increase in [Na+], observed after anoxia in 11 - 14 DIV neurons was blocked by 200 uM DIDS (Fig. 5.5,4, B). Although DIDS is commonly employed as an inhibitor of HCOy-dependent pHi regulating mechanisms, it can also inhibit a variety of cellular events potentially associated with anoxia (e.g. mitochondrial free radical production; see Cabantchik & Greger, 1992; Han et al. 2003; Tauskela et al. 2003). Indeed, 200 uM DIDS reduced slightly, albeit significantly, the increase in [Na+]j observed following anoxia under nominally HCO3"-free, Hepes-buffered conditions; there was no significant difference between the rise in [Na+]j seen after anoxia in the presence of DIDS under HCOy-containing compared to HCOy-free conditions (Fig. 5.573; P = 0.15). Taken together, the results are consistent with the possibility that Na+/FT exchange activity contributes to Na+ influx immediately following, but not during, 5 min anoxia in 6 - 10 and 11 -14 DIV rat hippocampal neuronal cultures. HCCV-dependent mechanisms also appear to contribute to Na+ influx following anoxia, but only in neurons maintained in culture for 11-14 DIV. 169 5.3. DISCUSSION 5.3.1. Resting \Na+\ under normoxic conditions Resting [Na+]j in cultured postnatal rat hippocampal neurons was ~11 mM, a value that is in good agreement with earlier studies in hippocampal neurons (Pinelis et al. 1994; Rose & Ransom, 1997; Diarra et al. 2001). The maintenance of a low resting [Na+]j in the face of a steep inwardly directed electrochemical gradient for Na+ is a common feature of vertebrate and invertebrate neurons (e.g. Thomas, 1972; Deitmer & Schlue, 1983; Chen et al. 1999). In the present study, under normoxic conditions, the application of ouabain or the removal of K+0 caused increases in [Na+]j of ~ 6.0 mM min"1 (estimated from the slope of linear fits to [Na+], measurements obtained during the first 3 min of 0 [K+]0 or ouabain application). As outlined by Rose & Ransom (1997), these rates approximate, in rat hippocampal neurons, a resting molar flux density for Na+ of ~ 16 x 10"12 mol cm" s" (assuming a spherical cell body and cell body diameter of 25 uM). Similar flux values have been estimated in squid axons and snail neurons (Hodgkin & Keynes, 1955; Thomas, 1972; also see Pinelis et al. 1994) and, while they may be influenced by a slowly developing (min) ouabain-induced depolarization sometimes observed in neurons (Thomas, 1972; Fujiwara et al. 1987), they are larger than those observed in non-neuronal cell types (e.g. MacLeod, 1989; Wu & Vaughan-Jones, 1994; Despa et al. 2002). It is apparent that maintenance of resting [Na+]j in rat hippocampal neurons under normoxic conditions reflects a balance between Na+,K+-ATPase activity and ongoing Na+ influx. 5.3.2. Anoxia-evoked increases in rNa+~|j The changes in [Na+]i observed during anoxia similarly reflect a balance between reduced Na+,K+-ATPase activity and ongoing/increased Na+ influx. At the end of 5 min anoxia, I 170 observed an increase in [Na+]j of -15 (6 DIV neurons) to -40 (14 DIV neurons) mM that was dependent on the presence of external Na+ and was reduced when anoxia-induced falls in internal ATP levels were attenuated by creatine pretreatement. The increases in [Na+]j observed in the present study are consistent with those observed previously in a variety of mammalian central neurons in response to anoxia or oxygen-glucose deprivation, not only in culture and slice preparations in vitro (Friedman & Haddad, 1994a; Pisani et al. 1998a; Calabresi et al. 1999b; Diarra et al. 2001) but also in CA1 neurons in vivo in response to 8 min low-flow global ischemia (under which conditions [Na+]j increased by -50 mM; Erecinska & Silver, 2001). In contrast, the changes in [Na+]j observed immediately following anoxia have remained poorly defined (e.g. Taylor et al. 1999; LoPachin et al. 2001). In the present study, the recovery of [Na+]i to resting levels after anoxia reflected a resumption of Na+,K+-ATPase activity in spite of ongoing/increased Na+ influx. Thus, when the Na+/K+-ATPase was inhibited in the immediate post-anoxic period, a further increase in [Na+]j of -30 (6 DIV) to -60 (14 DIV neurons) mM was observed and was blocked in the absence of external Na+. By inhibiting Na+,K+-ATPase activity, the contribution of Na+ influx pathways (e.g. Na+H4" exchange) to Na+ entry occurring following anoxia could be identified; these routes of entry may be of importance under conditions in which the recovery Na+,K+-ATPase activity occurs more slowly and/or by creating local changes in [Na+]i. That the magnitudes of the increases in [Na+]j observed during and after anoxia were related to the number of days that neurons had been maintained in culture may in part account for the previous finding (Jiang et al. 1992) that anoxia-induced falls in [Na+]0 are smaller in brainstem slices taken from neonatal vs. adult rats and may reflect developmental increase in the expression and/or activities of the mechanisms involved in their production (e.g. Bevensee et al. 1996; Sakaue et al. 2000; 171 Douglas et al. 2001; Gibney et al. 2002); this finding is considered further below (see Section 5.3.5 and Chapter 7). 5.3.3. Contribution of Na+/FP exchange activity As noted earlier, the examination of the contribution of Na+/H+ exchange to anoxia-evoked changes in [Na+]i in rat hippocampal neurons is complicated by the lack of a specific pharmacological inhibitor. Therefore, only an indirect assessment of the role of NaVH4" exchange activity to the increases in [Na+]j observed during and following anoxia could be made by testing a number of maneuvers that have previously been shown to influence exchange activity in rat hippocampal neurons. Thus, harmaline (a non-selective inhibitor of NaVFT exchange activity in rat hippocampal neurons; see Raley-Susman et al. 1991), while decreasing rates of pHj recovery from internal acid loads imposed under normoxic, Hepes-buffered conditions, had no influence on Na+ influx occurring during anoxia. In contrast, harmaline limited significantly the increase in [Na+]j observed following anoxia. These observations are consistent with the pHj measurements presented in Chapters 3 and 4 which suggested that, in rat hippocampal neurons, NaVH*" exchange activity is reduced during anoxia and becomes activated in the immediate post-anoxic period. Additional maneuvers which were found, on the basis of pHj measurements, to inhibit Na+/H+ exchange activity in the post-anoxic period (i.e. pH0 6.60 and i?p-cAMPS; see Chapter 4) similarly limited the increase in [Na+]i observed following anoxia. Furthermore, the increase in [Na+]j observed following anoxia was enhanced by maneuvers which stimulate NaVFT exchange activity in the immediate post-anoxic period (i.e. pH0 7.80). In response periods of ischemia, extensive studies have pointed to NaVH4" exchange as an important mechanism that contributes to the increase in [Na+]i seen in cardiac myocytes during reperfusion (reviewed by Karmazyn, 1999; Avkiran, 2001). The present results point to an 172 analogous contribution from Na+H4" exchange activity to the increase in [Na+]j observed immediately following anoxia in rat hippocampal neurons. In a similar manner, following periods of oxygen-glucose deprivation, cortical astrocytes deficient in NHE1 do not demonstrate the internal Na+ loading typically observed in astrocytes expressing functional NHE1 activity (Kintner et al. 2004). The contribution of Na+FT exchange to the increase in [Na+]j observed following anoxia may provide a mechanistic explanation for the neuroprotective effects of Na+Ff" exchange inhibitors in in vivo models of cerebral ischemia. Indeed, in one study, the protective effect of Na+Ff4" exchange inhibitors was attributed to a reduction in cerebral Na+ content (Kuribayashi et al. 1999; also see Matusmoto et al. 2003; Yamamoto et al. 2003). That NaVFT exchange activity can lead to elevations in [Na+]j is not without precedence. Na+/ff~ exchange activity causes increases in [Na+]j under normoxic conditions in many cell types (see Chapter 6; also Moody, 1981; Kaila & Vaughan-Jones, 1987) and it has been suggested that Na+/FT exchanger-induced increases in [Na*]j may play key roles in regulating Na+/Ca2+ exchange activity and/or the activities of intracellular signaling cascades (e.g. Hayasaki-Kajiwara et al. 1999; Trudeau et al. 1999; Mukhin et al. 2004). In these, and possibly other, ways, increased Na+/H+ exchange appears to underlie the increase in presynaptic quantal glutamate release that occurs during the recovery of pHj from imposed internal acidification in hippocampal neurons (Trudeau et al. 1999; also see Nordmann & Stuenkel, 1991; Bouron & Reuter, 1996). The contribution of NaVFT1" exchange activity in the increase in [Na+]j observed following anoxia is considered further in Chapter 6. As detailed in Chapter 4, a Zn -sensitive H efflux pathway (a putative gH+) also contributes to acid extrusion following anoxia in rat hippocampal neurons (also see Diarra et al. 1999), and it has been suggested previously by others that inhibition of gH*s may increase the 173 demand placed on Na+/FT exchange for acid extrusion (see Chapter 4; Demaurex et al. 1995). Consistent with this idea, Zn2+ had no effect on the increase in [Na+]i observed during anoxia but enhanced Na+ influx in the post-anoxic period, an effect that was blocked under pH0 6.60, conditions that are known to inhibit functional NaVFT exchange (Jean et al. 1985; Vaughan-Jones & Wu, 1990; Diarra et al. 1999). Although Zn2+ can modulate the activities of several ion channels and transport mechanisms (see Chapter 4), under the constant perfusion conditions employed in the present study, anoxia-evoked changes in [Na+]i are unaffected by NMD A or AMPA receptor antagonists or blockers of voltage-activated Ca2+ channels (see Chapter 7) and the effects of Zn2+ on the increase in [Na+]i observed following anoxia were observed when the Na+,K+-ATPase was already inhibited. It is of note that Zn2+-induced increases in [Na+]j have been observed previously in cultured cortical neurons and may contribute to a post-ischemic upregulation of NMD A receptor activity (Manzerra et al. 2001). 5.3.4. Contribution of HCO-f-dependent mechanisms In the majority of cells, HCCV-dependent pHj regulating mechanisms act in concert with NaVFT exchange to regulate pHj and some of these HCCV-dependent mechanisms (i.e. Na+-dependent C17HCCV exchange and electrogenic Na+/HCCV cotransport) also transport Na+ ions. In the present study, I found no evidence to suggest that HCCV-dependent mechanisms contribute to Na+ influx during anoxia in rat hippocampal neurons (possibly a result of a decline in internal ATP levels; e.g. Boron et al. 1988). In contrast, HCCV-dependent mechanism(s) appear to contribute to Na+ influx following anoxia. Given the multiple HCCV-dependent processes in rat hippocampal neurons, the identity of those mechanism(s) that contribute to enhanced Na+ influx after anoxia in neurons 11-14 DIV remains unclear. On the one hand, Na+-dependent C17HC03" exchange contributes to acid 174 extrusion in this cell type (Schwiening & Boron, 1994; Baxter & Church, 1996; Brett et al. 2002a) and a post-anoxic activation of exchange activity may account for the HCOy-dependent, DIDS-sensitive Na+ influx observed following anoxia. On the other hand, in non-neuronal cell types, electrogenic Na+/HC03_ cotransport has been found to contribute to Na+-dependent acid extrusion both during and following periods of ischemia (Lamers, 2001; Khandoudi et al. 2001; also see Giffard et al. 2000). Na+/HC03" cotransporters are expressed in discrete populations of central neurons (see Bevensee et al. 2000; Giffard et al. 2000; Schmitt et al. 2000); however, there is little functional evidence for their participation in pHj regulation in rat hippocampal neurons, at least under normoxic conditions (Schwiening & Boron, 1994; Baxter & Church, 1996), although inward (i.e. acid-extruding) Na+/HC03_ cotransport activity could be activated in response to membrane depolarizations observed during anoxia. The possibility that Na+0- and HC03"-dependent acid extrusion (either Na+-dependent CI7HCO3" exchange or Na+/HC03~ cotransport) might contribute to an increase in [Na+]i immediately following anoxia in rat hippocampal neurons is at variance with the previous observation that HC03"-dependent mechanisms appear to limit the magnitude of the internal alkalinization observed in the immediate post-anoxic period (see Chapter 3 for data in isolated adult hippocampal neurons and Diarra et al. 1999 for similar findings in cultured postnatal rat hippocampal neurons). Rather, the observation that the magnitude of the increase in pH, observed following anoxia appears smaller in the presence vs. the absence of HCO3" is consistent to observations made in mouse hippocampal neurons wherein an acid-loading Na+/HC03_ cotransporter is activated during and following transient periods of anoxia (Yao et al. 2003); however, this mechanism would act to extrude Na+ ions. Finally, it is also possible that the complex modulation of voltage-dependent Na+ currents by HC03" ions (e.g. Gu et al. 2000; Bruehl & Witte, 2003) could contribute to the 175 differences in Na+ influx observed after anoxia under HCCV-containing compared with HCCV-free conditions. Although the ability of DIDS to limit the increase in [Na+]j following anoxia under HCO3" /CCVbuffered conditions is consistent with its ability to inhibit HCCV-dependent pHj regulating mechanisms present in rat hippocampal neurons, DIDS also reduced the increase in [Na+]j observed following anoxia under HC03_-free, Hepes-buffered conditions, albeit to a much lesser extent. The latter observation may reflect residual activities of HC03"-dependent, Na+-transporting mechanisms in the nominal absence of HCO3" (see Wu et al. 1994; Deitmer & Schneider, 1998) or the recognized effects of DIDS on HCO3"-independent processes, which include the inhibition of chloride channels, K+/C1" transport and mitochondrial release of free radicals (see Han et al. 2003; Malek et al. 2003; Tauskela et al. 2003). 5.3.5. Summary This present study in rat hippocampal neurons is consistent with previous findings, made in a variety of non-neuronal cell types, that NaVH* exchange and HCCV-dependent mechanism(s) contribute to potentially injurious Na+ influx in the vulnerable period immediately after anoxia. Thus, in isolated rat hippocampal cultures, NaVH4- exchange activity appears to contribute to post-anoxic increases in [Na4];. Although developmental changes in neuronal NaVFT exchange expression and/or activity have been observed (see Bevensee et al. 1996; Ma & Haddad, 1997; Douglas et al. 2001; Nottingham et al. 2001), the contribution of NaVH4" exchange to the increases in [Na+]j observed following anoxia was observed in neurons both 6-10 and 11-14 DIV. In light of the evidence that pharmacological inhibitors of Na+/H+ exchange effectively protect against anoxia- and ischemia-induced neuronal injury, the results presented in Chapters 3 - 5 are consistent with the possibilities that the neuroprotective actions of NaVH4" exchange 176 inhibitors may result from reductions in the rises in pHj and/or [Na+]j that occur during early reperfusion (this is considered further in Chapter 6). HCOy-dependent mechanisms also appear to contribute to the increase in [Na+]j observed following anoxia, although this effect was restricted to neuronal cultures 11 - 14 DIV. Developmental upregulation of the expression of HCOs'-dependent pHj regulating mechanisms has been reported in the central nervous system (Raley-Susman et al. 1993; Kobayashi et al. 1994; Ma & Haddad, 1997; Douglas et al. 2001; Giffard et al. 2003) and, thus, may account for this difference. However, given the presence of multiple HCCV-dependent mechanisms that may contribute to this observed response, together with the complexities of the regulation of such mechanisms in rat hippocampal neurons (see Brett et al. 2002a), further experiments are required to clarify the potential contribution of HCdV-dependent pHi regulating mechanisms to anoxia-evoked increases in [Na+]j in rat hippocampal neurons. Despite the findings summarized above, there are two important limitations of the present study. First, although it is established that NaVH4" exchange activity is a major acid-extruding mechanism in rat hippocampal neurons under nominally HC(V-free conditions, firm conclusions regarding the contribution of NaVH*" exchange activity to anoxia-induced changes in [Na+]i (and pHj) are limited by the non-selective maneuvers that had to be employed to modulate exchange activity. Thus, to more precisely establish the contribution of NaVH4- exchange to anoxia-evoked changes in pHi and [Na+]i, in the following Chapter (Chapter 6), I developed a microspectrofluorimetric technique for the concurrent measurement of both ions in isolated hippocampal neurons. Second, it is clear that the activities of pHi regulating mechanisms can account for neither the increase in [Na+]i observed during anoxia nor all of the Na+ entry that takes place in the immediate post-anoxic period. Thus, in Chapter 7, a study examining additional 177 mechanism(s) that might potentially contribute to the increases in [Na+]i observed during and following anoxia will be presented. 178 Table 5.1: Contribution of pH, regulating mechanisms to the increase in [Na+]j observed during anoxia Treatment Normalized A[Na+ ]i(during) 6-10 DIV 11-14 DIV 200 uM harmaline 0.91 ±0.22 (7) 0.98 ± 0.26 (5) lOOuMZn' 2+ 1.03 ±0.18 (12) 1.17±0.15(15) HCCV/CCVbuffered medium HC037C02-buffered medium + 200 uM DIDS 1.11 ±0.13 (8) 1.28 ±0.28 (3) 0.77 ±0.15 (5) 0.78 + 0.31 (4) N.S. N.S. To generate Normalized A[Na+ ]i(during) values, measurements of A[Na+]i(during) under experimental test conditions were normalized to measurements made in experiments performed on age-matched sister cultures under control conditions. Statistical comparisons were performed by comparing absolute A[Na+] inuring) measurements made under experimental test conditions to measurements made in age-matched sister cultures under control conditions. Numbers in brackets denote the number of neuronal populations (i.e. coverslips) from which the data were generated. 'Neurons were treated with 200 pM harmaline for 120 - 180 min prior to the start of an experiment. DIV, days in vitro. N s indicates no significant difference between the increase in [Na+]i observed during anoxia under HCCV/CCVbuffered conditions, either in the presence or absence of 200 uM DIDS, and the increase in [Na+]i observed during anoxia in age-matched sister cultures under control (HC03"-free, Hepes-buffered) conditions (P = 0.51 and 0.78, respectively). 179 Table 5.2: Contribution of pHi regulating mechanisms to the increase in [Na+]j observed following anoxia under 0 [K+]0 conditions Normalized A[Na+ ]i(after) 6-10 DIV 11-14 DIV Treatment 200 uM harmaline1 0.48 ±0.11 (7)* 0.76 ± 0.08 (7)* pHo6.60 0.5610.10(6)* 0.79 ± 0.05 (7)* 50 uM i?p-cAMPS2 0.69 ± 0.09 (5)* 0.77 ± 0.08 (4)* pHo7.80 1.27 ±0.06 (10)* 1.57 ± 0.41 (4)* 100uMZn2+ 1.87 ±0.27 (7)* 1.38 ± 0.16 (8)* 100 uM Zn2+, pH0 6.60 0.81 ± 0.28 (3) n.d. To generate Normalized A[Na+ Rafter) values, measurements of A[Na+]i(after) under experimental test conditions were normalized to measurements made in experiments performed on age-matched sister cultures under control conditions. Statistical comparisons were performed by comparing absolute A[Na+]i(after) measurements made under experimental test conditions to measurements made in age-matched sister cultures under control conditions. Numbers in brackets denote the number of neuronal populations (i.e. coverslips) from which the data were generated. 'Neurons were treated with 200 uM harmaline for 120 - 180 min prior to the start of an experiment. 2i?p-cAMPS was present in the perfusate during and following anoxia. Alterations in pH0 and exposure to Zn2+ began at the start of perfusion with K+-free media. * indicates statistical significance (P < 0.05) compared with measurements of made in age-matched sister cultures in the absence of treatment. DIV, days in vitro; n.d., not determined. 180 Fig. 5.1. Anoxia-evoked changes in [Na+]j in rat hippocampal neurons. A, 5 min anoxia was imposed under nominally HC03"-free, Hepes-buffered conditions by exposure to medium containing 1-2 mM sodium dithionite and bubbled vigorously with 100% Ar (filled circles). Also shown are the changes in [Na4], evoked by anoxia in a sister culture exposed to medium containing 1 - 2 mM sodium dithionite and bubbled vigorously with air (open circles). B, relationship between the magnitude of the increase in [Na+]j observed during 5 min anoxia (A[Na+]i(during)) and the number of days that neurons were maintained in culture (DIV, days in vitro). Anoxia was imposed under Hepes-buffered conditions either by the addition of 1 - 2 mM sodium dithionite to medium bubbled with 100% Ar (filled circles; n = 21 - 54 for each datum point) or by exposure to medium that had been bubbled vigorously with 100% ultrapure Ar for >18 h (open circles; n > 2 for each datum point). The solid line represents a linear regression fit to the data points obtained when anoxia was imposed by the addition of sodium dithionite (correlation coefficient = 0.96; PO.0001 by one-way ANOVA). Error bars are s.E.M. C, under normal Na+0-containing conditions, anoxia induced an increase in [Na+]i that recovered upon the return to normoxia (filled circles). When anoxia was imposed under reduced Na+0, NMDG+-substituted conditions, the increase in [Na+]j was abolished (open circles). 181 Time (min) DIV Time (min) 182 Fig. 5.2. Contribution of reduced Na+,K+-ATPase activity to the increase in [Na+]j observed during anoxia. A, superimposed records of the changes in [Na+]i observed in response to 5 min exposure to [K+]-free medium (filled symbols) or 500 uM ouabain (open symbols), as indicated by the bar above the traces (compare with Fig. 5.1,4). [Na+]i failed to show complete recovery following ouabain application. B, intracellular ATP levels were determined after 3 or 5 min anoxia induced by exposure to sodium dithionite-containing medium in neurons with (open symbols) or without (filled symbols) pretreatment with 10 mM creatine for >2 h. Measurements were made using neuronal cultures 8-10 DIV and were normalized to values obtained prior to anoxia in age-matched sister cultures in each experimental group. The fall in ATP levels evoked by 3 min anoxia under control conditions was significantly attenuated by creatine pretreatment (*, P < 0.05). Numbers in brackets indicate the number of neuronal populations from which data were obtained. C, in 6 - 10 DIV neurons pretreated with 10 mM creatine for >2 h (open bars), the increase in [Na+]i measured 3 min after the start of anoxia was significantly less than that observed in age-matched sister cultures under control neurons (filled bar; *; P < 0.05). There was no statistical difference between the increase in [Na+]; measured following 5 min anoxia in creatine-treated and untreated cultures (P = 0.57). 183 184 Fig. 5.3. Changes in [Na+]j observed after anoxia during inhibition of Na+,K+-ATPase activity. A, Na+,K+-ATPase activity was inhibited by perfusion with K+-free medium at the end of 5 min anoxia (filled circles), revealing a secondary rise in [Na+]i in the immediate post-anoxic period. Neurons on a different coverslip were exposed to K+- and Na+-free medium immediately after 5 min anoxia (open circles). In the absence of external Na+ (NMDG+-substitution), the increase in [Na+]j following anoxia was abolished. B, the relationship between the magnitude of the increase in [Na+]i observed after anoxia (A[Na+]Rafter)) and the number of days neurons were maintained in culture (filled circles; n = 7 - 31 for each datum point). The solid line represents a linear regression fit to the data points indicated (correlation coefficient = 0.88; .PO.0001 by one-way ANOVA). Also shown are A[Na+]Rafter) values measured during inhibition of the Na+,K+-ATPase with 500 pM ouabain (open circles; n = 3 for each datum point). 185 Time (min) DIV 186 Fig. 5.4. Effects of modulating Na+/FT exchange activity on the increase in [Na+]j observed after anoxia (Na+,K+-ATPase inhibited). A, at the end of 5 min anoxia, neurons were exposed to K+-free medium for 7 min. Compared with changes observed under control conditions (filled circles), the magnitude of the increase in [Na+]j observed after anoxia under K+-free conditions was reduced by pretreatment with harmaline (200 pM for 120 - 180 min; open circles), at pH0 6.60 (open diamonds), or in the presence of the PKA inhibitor, i?p-cAMPS (50 pM, applied at the beginning of anoxia; cross-hairs). In contrast, the magnitude of the increase in [Na+]j observed following anoxia was enhanced at pH0 7.80 (open triangles). Alterations in pH0 began immediately at the start of superfusion with K+-free medium and continued for the duration of the records shown. Inset, examined under normoxic HCCV-free, Hepes-buffered conditions, internal acid loads were imposed using the NH4+ pre-pulse technique in age-matched sister cultures pretreated (open symbols; n = 6) or not pretreated (filled symbols; n = 10) with 200 uM harmaline. Harmaline pretreatment reduced rates of pHj recovery. B, exposure to 100 uM Zn2+ (open squares) immediately following anoxia under under [K+]0-free conditions enhanced the increase in [Na+]j observed at this time, compared with that observed following anoxia in an age-9+ matched sister culture in the absence of Zn (filled circles). In another age-matched sister culture, this effect was abolished when 100 pM Zn was applied at pH0 6.60 (filled squares). Exposure to Zn with or without alterations in pH0 began at the start of perfusion with K -free medium and continued for the duration of the records shown. 187 120 ^ 60 H CO Time (min) Control harmaline (200 uM) pH0 6.60 Rp-cAMPS (50 uM) PH0 7.80 10 15 Time (min) Control Zn2+ (100 u.M) Zn2+(100 nM), pHo6.60 188 Fig. 5.5. Contribution of HCO3"-dependent mechanisms to the increase in [Na+]j observed after anoxia (Na+,K+-ATPase inhibited). A, under HCOV-free, Hepes-buffered (filled circles) or HCO3/CCVbuffered (open circles) conditions, neurons were exposed to 5 min anoxia followed by 7 min perfusion with K+-free medium. 200 uM DIDS reduced the augmented increase in [Na+]j observed after anoxia under HC037C02-buffered conditions (open diamonds; DIDS was added to anoxic media and was present throughout the rest of the record). B, summary of the effects of external buffering conditions and DIDS on the increase in [Na+], observed after anoxia in neurons 6-10 and 11-14 DIV. Neither the addition of HCO3" nor the presence of 200 uM DIDS under HCOy/CCh-buffered or Hepes-buffered conditions, influenced significantly the increase in [Na+]j observed after anoxia in neuronal cultures 6-10 DIV, compared to the increase in [Na+]i observed after anoxia in the absence of HCO3". In contrast, in neuronal cultures 11-14 DIV, the increase in [Na+]i after anoxia was enhanced in the presence of HCO3/CO2 and, under HC037C02-buffered conditions, DIDS reduced the magnitude of the rise in [Na+]j to a value similar to that observed under control conditions (i.e. Hepes-buffered media in the absence of DIDS; P = 0.66). f indicates statistical significance (P < 0.05) compared to measurements made in age-matched sister cultures under HC037C02-buffered conditions in the absence of DIDS. * indicates statistical significance (P < 0.05) compared to measurements made in age-matched sister neurons under nominally HC03~-free, Hepes-buffered conditions in the absence of DIDS. Error bars are S.E.M. 190 CHAPTER SIX CONCURRENT MEASUREMENT OF pHi AND [Na+]i WITH FLUORESCENT INDICATORS: A FURTHER EVALUATION OF THE ROLE OF Na+/H+ EXCHANGE TO ANOXIA-EVOKED CHANGES IN [Na+]i AND pH;8 6.0. INTRODUCTION In the absence of a selective pharmacological agent to inhibit Na+/H+ exchange activity in rat hippocampal neurons, in Chapters 3 - 5, the contribution of NaVH4" exchange activity to the changes in pHj and [Na+]j observed during and following anoxia was inferred, under HC03~-free conditions, by determining the Na+0- (and Li+0-) dependencies of the anoxia-evoked changes in pH; (Chapters 3 and 4) and by examining the effects of non-selective maneuvers that influence NaVff" exchange activity on anoxia-evoked changes in [Na+]j (Chapter 5). The results were consistent with the possibility that Na+/H+ exchange activity is stimulated immediately following anoxia in rat hippocampal neurons and contributes to the regulation of both pHj and [Na+]i at this time. Nevertheless, in the studies presented in Chapters 3-5, fluorescent probe-based measurements of pHj and [Na+]i were conducted separately in experiments performed in parallel, an approach that limits attempts to understand the interrelationships that exist between [H+]j and [Na+]j. It is apparent that concurrent measurements of pHi and [Na+]j would provide further insight into the presumed relationship between anoxia-evoked changes in [H+]i and [Na+]j, particularly in the post-anoxic period, and would strengthen conclusions made regarding the role of Na+/Ff~ exchange activity in the genesis of these ionic changes. 8 A version of this chapter has been published. Sheldon C, Cheng Y.M. and Church J. (2004) Concurrent measurements of the free cytosolic concentrations of H+ and Na+ ions with fluorescent indicators. Pflugers Arch. Epub. 191 Changes in pHj and [Na+]i may be linked directly through mechanisms as diverse as Na+/H+ exchange (e.g. Kaila & Vaughan-Jones, 1987), Na+-dependent OTHCCV exchange (e.g. Rose & Ransom, 1997) and electrogenic NaVHCCV cotransport (e.g. Bers et al. 2003; Deitmer & Schlue, 1989), or indirectly via the coordinated activities of two or more transport mechanisms. In a number of cell types, for example, a rise in [Na+]i promotes reverse-mode Na+/Ca2+ exchange and the subsequent rise in [Ca2+]j can cause a fall in pHj by activating the acid-loading Ca2+,FT-ATPase (e.g. Kiedrowksi, 1999). Furthermore, changes in pHj and [Na+]j can influence the activities of not only pHj regulating transporters (Green et al. 1988; Kaila & Vaughan-Jones, 1987) but also mechanisms that contribute to Na+ flux across biological membranes, including Na+/Ca2+ exchange and Na+/K+/2C1" cotransport (Blaustein & Lederer, 1999; Russell, 2000). This complex relationship between changes in [FT]i and [Na+]j, and the recognized importance of both of these ions as determinants of cell function under a variety of physiological and pathophysiological conditions, further highlights the need for concurrent quantitative measurements of both ions. Thus, in the first part of this study I developed and characterized a novel technique for the near-simultaneous measurement of [Na+]i and pHj in rat hippocampal neurons using, respectively, the dual excitation Na+ indicator SBFI (Minta & Tsien, 1989) and the dual emission seminaphthorhodafluor pH indicators carboxy SNARF-1 (Whitaker et al. 1991) or SNARF-5F (Liu et al. 2001). Although SNARF-5F retains the spectral properties of other SNARF derivatives, it displays a lower pKa value under cell-free in vitro conditions (jpKa ~7.2) that may make it more suitable than carboxy SNARF-1 (pKa -7.5) for measuring changes in pHj below -6.5, as may occur, for example, in mammalian central neurons during anoxia or ischemia. Next, in the second part of this study, I used the technique developed in the first part of the study to perform concurrent measurements of pHi and [Na+]i to further examine the involvement of 192 NaVFT1" exchange activity in the anoxia-evoked changes in pH; and [Na+]j observed in cultured rat hippocampal neurons. 6.1. MATERIALS AND METHODS 6.1.1. Experimental preparation Primary cultures of hippocampal neurons prepared from 2-4 day old postnatal Wistar rats were employed in all experiments presented in this Chapter. 6.1.2. Dye loading and recording techniques In the majority of experiments, changes in pHj were measured with either carboxy SNARF-1 or SNARF-5F carboxylic acid. The acetoxymethyl esters of carboxy SNARF-1 and SNARF-5F (carboxy SNARF-1-AM and SNARF-5F-AM, respectively) were prepared as 10 and 5 mM stock solutions in DMSO, respectively. In a limited number of experiments, BCECF was used to measure anoxia-evoked changes in pHj. Cultured postnatal rat hippocampal neurons were loaded with SBFI in the manner described in Chapter 5. To load SNARF derivatives, either individually or following SBFI-AM incubation, coverslips with neurons attached were placed in standard loading medium containing 0.10% Pluronic acid and either 10 uM carboxy SNARF-1-AM or SNARF-5F-AM for 30 min at 32°C. In experiments in which BCECF was employed as the pHj indicator, coverslips with neurons attached were placed in standard loading medium containing 2 uM BCECF-AM for 30 min at 22°C (Baxter & Church, 1996). Following loading, coverslips were placed in standard loading medium for 20 min to ensure de-esterification of the fluorophore(s) and then mounted in a temperature-controlled perfusion chamber to form the base of the chamber. Neurons were 193 superfused at a rate of 2 ml min"1 for 15 min with the initial experimental solution at 37°C (unless otherwise noted) before the start of an experiment. Details of the recording techniques used to measure pHi in neurons loaded only with BCECF are provided in Chapter 2. In neurons loaded with SBFI and/or a SNARF-based derivative, measurements of [Na+]j and/or pHj were performed using the dual-excitation and dual-emission ratio methods, respectively. The same imaging system employed in studies described in Chapters 3 - 5 to measure pHj or [Na+]j in neurons single-loaded with BCECF or SBFI, respectively, was used to measure pHj and [Na+]i concurrently but with a different filter set and the addition of a second intensified charge coupled device camera. A schematic diagram of the optical equipment used for measurements of [Na+]j and pH; in neurons loaded with SBFI and/or carboxy SNARF-1 or SNARF-5F in the present series of experiments is presented in Fig. 6.1. Filter selection was based upon the published in vitro spectra of SBFI, carboxy SNARF-1 and SNARF-5F (Minta & Tsein, 1989; Martinez-Zaguilan et al. 1991; Liu et al. 2001). In experiments in which SBFI alone was employed, neurons were excited alternately at 334 ± 5 and 380 ± 5 nm and fluorescence emissions at 550 ± 40 nm were detected sequentially by a single camera (Camera 2 in Fig. 6.1). In experiments in which carboxy SNARF-1 or SNARF-5F alone were employed, neurons were excited at 488 ± 5 nm and fluorescence emissions at 550 ± 40 and 640 ± 20 nm were detected simultaneously by two cameras (Cameras 1 and 2, respectively, in Fig. 6.1), the registration of which was confirmed prior to every experiment. In experiments in which neurons were co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F, ratio pairs were collected continuously by alternating between the excitation and emission modes; each automated cycle took -1.5 s to complete, including a -0.5 s delay between collecting SBFI- and SNARF-derived ratio pairs, and was repeated every 2 - 15 s during the course of an experiment. 194 6.1.3. Calculation of \Na+\ and pHj As described in Chapter 2, a one-point calibration technique was employed to convert background-corrected BCECF ratio values to absolute pHj values. Similarly, a one-point calibration technique was employed to convert background-corrected SBFI (5/334/5/380) and SNARF-derived (5/55o/5/64o) ratio values into [Na+]; and pHj values, respectively. At the end of an experiment, neurons loaded with SBFI were exposed to a pH 7.35 medium containing 10 mM Na+ and 4 uM gramicidin D (Table 2.3; Diarra et al. 2001) whereas neurons loaded with carboxy SNARF-1 or SNARF-5F were exposed to a high-[K+], pH 7.00 solution containing 10 uM nigericin (Table 2.3; Baxter & Church, 1996); in neurons loaded with both SBFI and a SNARF derivative, the SBFI and SNARF one-point calibrating media were applied sequentially. The resulting background-corrected ratio values at [Na+]i = 10 mM (for SBFI) and at pHi = 7.00 (for carboxy SNARF-1 or SNARF-5F) were used as normalization factors for experimentally-derived background-subtracted SBFI (5/334/5/380) and SNARF (5/550/5/640) ratio values, respectively. At the end of some experiments conducted at 37°C, I was unable to obtain stable normalizing ratio values for carboxy SNARF-1 or SNARF-5F during the one-point calibration at pH 7.00. In these cases, the one-point calibration was either repeated at pH 7.50 or the experimental data were normalized with ratio values obtained under identical experimental and optical conditions from a fresh sister culture loaded with the appropriate dye(s) and exposed to SNARF calibrating medium at pH 7.00. A similar instability of carboxy SNARF-1 and SNARF-5F ratio values, characterized by anomalous increases and decreases in background-subtracted 550 and 640 nm emission intensities, respectively, was also sometimes experienced during full calibration experiments conducted at 37°C when pH values were <7.00. The reasons for these atypical behaviours, which could occur in neurons loaded with carboxy SNARF-1 or SNARF-5F in the absence or presence of SBFI, remain unclear, although 195 similar difficulties have been experienced by others (e.g. Bassnett et al. 1990; Martinez-Zaguilan et al. 1991; Seksek et al. 1991; Blank et al. 1992; Boyarsky et al. 1996a; Seksek & Bolard, 1996). Nevertheless, reproducible calibration parameters for carboxy SNARF-1 and SNARF-5F at 37°C in situ were obtained and, in the case of carboxy SNARF-1, are consistent with those reported by others (see Section 6.3.1.1). Details regarding the conversion of normalized SBFI ratio values into [Na+]j values are provided in Chapter 2. Normalized carboxy SNARF-1 or SNARF-5F ratio values were converted into pHj values using the equation pH = pKa + logF640rnin/max - log[(Rn - Rn(min)) / (Rn(max) - Rn)] (Equation 6.1) where Rn is the background-subtracted carboxy SNARF-1 or SNARF-5F fluorescence intensity ratio (BI55o/BIe4o) normalized to pH 7.00 (or pH 7.50; see above); pKa is the -log of the dissociation constant of the fluorophore; and F640m\n/max is the ratio of fluorescence measured at 640 nm for low pH (pH 5.5) to that for high pH (pH 8.5; see Buckler & Vaughan-Jones, 1990). The parameters fitting Equation 6.1 were derived from full in situ calibration experiments, as described in Section 6.2.2. 6.1.4. Data Analysis In contrast to the anoxia-evoked changes in pHj observed in acutely isolated adult rat hippocampal CA1 pyramidal neurons, in which pHj rose during anoxic transients and increased further in the post-anoxic period to values above pre-anoxic resting pHj values (see Chapters 3 and 4), in experiments in which SNARF-5F, carboxy SNARF-1 or BCECF were used as pHj indicators, pHj increases during anoxia and pHj 'overshoots' following anoxia were infrequently 196 observed in the cultured postnatal rat hippocampal neurons employed in the present studies. Thus, in experiments in which the changes in pH; observed after periods of anoxia (Na+,K+-ATPase inhibited) were measured, the magnitude of the increase in pHj observed after anoxia was measured as the difference between the pHj value observed at the end of anoxia and the pHj value observed at the end of a 7 min exposure to 0 [K+]0 media (see Table 6.2; Fig. 6.8 D, F). In a similar manner, the magnitude of the increase in [Na+]i observed after anoxia was measured as the difference between the [Na-*]; value observed at the end of anoxia and the [Na+]i value observed at the end of 7 min exposure to 0 [K+]0 media (as described in Chapter 5). The differences between the anoxia-evoked changes in pHj observed in acutely isolated adult vs. cultured rat hippocampal neurons is discussed further in Chapter 8. Results are reported as mean ± s.E.M. In experiments in which internal acid loads were imposed by the NH44" prepulse technique or neurons were exposed to transient periods of anoxia, experiments were performed on at least three coverslips obtained from 2-5 different batches of cultures and the accompanying n value refers to the number of neurons from which data were analyzed. For all other experiments, including full calibration experiments, the accompanying n value refers to the number of cell populations (i.e. number of coverslips) analyzed. Statistical comparisons were carried out using Student's two-tailed t-test, paired or unpaired as appropriate, with a 95% confidence limit. 197 6.2. RESULTS 6.2.1. Separation of SBFI and SNARF fluorescence emissions in situ With the optical filters specified in the Methods, the excitation and emission characteristics of SBFI, carboxy SNARF-1 and SNARF-5F in vitro appear sufficiently distinct to permit the discrimination of NaV and pHj-dependent signals from dual dye-loaded cells. However, the spectral properties of many fluorescent probes, including SBFI (see Negulescu & Machen, 1990; Diarra et ai. 2001) and carboxy SNARF-1 (see Seksek et al. 1991; Martinez-Zaguilan et al. 1996), may differ in situ compared to in vitro. Initially, therefore, I examined whether the behaviours of SBFI, carboxy SNARF-1 and SNARF-5F in hippocampal neurons in situ would allow the isolation of signals coming from the respective dyes under the experimental conditions employed. To do so, the intensities of emitted fluorescence relative to background fluorescence values under a variety of conditions were measured. In neurons single-loaded with SBFI, excitation at 334 and 380 nm (emissions collected at 550 nm) elicited fluorescence signals respectively -20 and -11 times greater than those originating from neurons single-loaded with carboxy SNARF-1, and -16 and -14 times greater than those originating from neurons single-loaded with SNARF-5F (Fig. 6.2A & Q. In neurons single-loaded with carboxy SNARF-1 or SNARF-5F, excitation at 488 nm (emissions collected at 550 nm and 640 nm) elicited fluorescence signals respectively -8 (at 550 nm) and -7 (at 640 nm) times greater than those obtained from neurons single-loaded with SBFI (Fig. 6.25 & Q. When neurons were co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F, fluorescence emissions originating from SBFI and carboxy SNARF-1 or SNARF-5F continued to be adequately resolved (Fig. 6.2/1 - C). Nevertheless, emission intensities measured at 550 nm following excitation at 334 or 380 nm in neurons co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F were significantly reduced, compared to those measured in neurons 198 single-loaded with SBFI (Fig. 6.2A). In addition, following excitation at 488 nm, fluorescence emissions at 550 nm, but not 640 nm, were reduced by -25% in neurons co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F compared to those measured in neurons single-loaded with carboxy SNARF-1 or SNARF-5F (Fig. 6.25). These observations are consistent with the possibility that SNARF derivatives may quench SBFI fluorescence, and vice versa. When neurons were single-loaded with SBFI and subsequently loaded with carboxy SNARF-1 or SNARF-5F (n = 3 in each case), 5/334 and 5/38o values were reduced by 61 ±2 and 63 ± 1% (carboxy SNARF-1; not shown) and by 60 ± 13 and 60 ± 10%) (SNARF-5F; Fig. 6.3,4), respectively. However, because 5/334 and 5/330 values were reduced to a proportionately similar extent, SBFI-derived 5/334/5/380 ratio values were minimally affected by the presence of either carboxy SNARF-1 (not shown) or SNARF-5F (Fig. 6.3,4). Due to the length of time required for SBFI loading, I was unable to further examine the apparent effect of SBFI to quench carboxy SNARF-1 and SNARF-5F 550 nm emissions in a manner similar to that shown in Fig. 6.3,4. Therefore, I measured fluorescence emission intensities at 550 and 640 nm (excitation at 488 nm) in neurons single-loaded with either carboxy SNARF-1 or SNARF-5F at pH 6.0, 7.0 and 8.5, and compared these intensities to those obtained in neurons from sister cultures co-loaded with SBFI and carboxy SNARF-1 or SNARF-5F. Neither carboxy SNARF-1 (not shown) nor SNARF-5F (Fig. 6.35) emission intensities at 550 and 640 nm were significantly influenced by the addition of SBFI, indicating that SBFI does not significantly reduce fluorescence emissions from SNARF-based dyes at any of the pH values examined. Quenching between fluorophores requires intracellular co-localization. Exposure of neurons co-loaded with SBFI and carboxy SNARF-1 or SNARF-5F to 0.005 - 0.01% saponin (n = 3 in each case) reduced emission intensities measured at 550 nm (following excitation at 334, 199 380 or 488 nm) and at 640 nm (following excitation at 488 nm) by >85% (Fig. 6.3A; also see Blank et al. 1992; Rose & Ransom, 1997). Similar findings, which also indicate that SBFI- and SNARF-derived fluorescence emissions originate largely from the cytosolic compartment, were made when co-loaded neurons were exposed to 20 pM digitonin (n = 3 in each case; not shown). 6.2.2. Full calibrations of SBFI. carboxy SNARF-1 and SNARF-5F ratio values in situ Next, full in situ calibrations of SBFI, carboxy SNARF-1 and SNARF-5F in single-loaded neurons and in neurons co-loaded with SBFI and carboxy SNARF-1 or SNARF-5F were performed. As illustrated in Fig. 6.4,4, full SBFI calibrations were performed at 37°C by exposing neurons to pH 7.35 media containing 4 pM gramicidin D at eight different [Na+] values (range, 0 -130 mM). The resulting plot of the data points relating [Na+] to Rn in neurons single-loaded with SBFI is shown in Fig. 6.45 and the resulting fitted SBFI calibration parameters are presented in Table 6.1. Importantly, when neurons single-loaded with SBFI were illuminated at 488 nm (i.e. the excitation wavelength employed for carboxy SNARF-1 and SNARF-5F), fluorescence emissions measured at 550 and 640 nm (i.e. the emission wavelengths of the SNARF dyes) remained small and stable as [Na+] was altered from 0 to 130 mM (Fig. 6AA). Finally, when full SBFI calibrations were performed in neurons co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F, there were no significant differences between the resulting SBFI calibration parameters and those computed from neurons loaded with SBFI alone (Table 6.1; Fig. 6.45). Thus, carboxy SNARF-1 and SNARF-5F do not influence the in situ sensitivity of SBFI to changes in [Na+]i. 200 Full in situ calibrations of carboxy SNARF-1 are presented in Figure 6.5. As described by others (e.g. Buckler & Vaughan-Jones, 1990; Martinez-Zaguilan et al. 1991; Blank et al. 1992), exposing neurons single-loaded with carboxy SNARF-1 to 10 uM nigericin-containing high-K+ solutions at a variety of pH values influenced both 5/550 and 5/640 emission intensities (Fig. 6.5^4). The resulting plots of the data points relating Rn and pH, at both 22°C and 37°C, are shown in Fig. 6.55 and the fitted carboxy SNARF-1 calibration parameters are presented in Table 6.1. Interestingly, the carboxy SNARF-1 /7640min/max value was significantly reduced at 37°C compared to 22°C (Table 6.1) and, in agreement with Ch'en et al. (2003), the dynamic range of the dye's fluorescence ratio increased with increasing temperature (Table 6.1; Fig. 6.55). As noted in the Introduction, SNARF-5F displays a lower pKa value in vitro than carboxy SNARF-1. Therefore, full in situ calibrations (at 37°C) of SNARF-5F single-loaded into hippocampal neurons were performed. The fitted calibration parameters (Table 6.1) confirmed that SNARF-5F also possesses a lower pKa value than carboxy SNARF-1 in situ. Carboxy SNARF-1 and SNARF-5F calibration parameters were then determined at 37°C in neurons also loaded with SBFI (see Fig. 6.55). There were no significant differences between the values of the calibration parameters obtained from neurons loaded with carboxy SNARF-1 or SNARF-5F alone and those obtained from neurons co-loaded with SBFI and carboxy SNARF-1 or SNARF -5F (Table 6.1). Together, the results indicate that SBFI does not influence the in situ sensitivities of carboxy SNARF-1 or SNARF-5F to changes in pHj. 201 6.2.3. Effects of changes in [Na+1 on pH; measurements with carboxy SNARF-1 and SNARF -5F in situ The effects of changes in pHj on [Na+]j estimated with SBFI in situ have been well characterized (see Negulescu & Machen, 1990; Rose & Ransom, 1997; Diarra et al. 2001). To examine the effects of changes in [Na+]i on SNARF-based pHi measurements, neurons were single-loaded with carboxy SNARF-1 or SNARF-5F and [Na ] was varied from 0 to 10 to 130 mM at four different pH values (6.00, 6.50, 7.00 and 7.50) in the presence of 4 pM gramicidin D at 37°C (see Diarra et al. 2001). As illustrated in Fig. 6.6,4, increasing [Na+] from 0 to 10 to 130 mM at a constant pH had minimal effects on carboxy SNARF-1 ratio measurements. Moreover, plots of Rn as a function of pH at each [Na+] (Fig. 6.65) indicated that changes in [Na+] did not alter the computed pKa + logF640mjn/max, Rn(min) or Rn(max) values for carboxy SNARF-1. Thus, neither carboxy SNARF-1-derived Rn nor pHi measurements were influenced significantly by changes in [Na+] in the range 0-130 mM (P > 0.32 at each pH value). Similar findings were made in neurons single-loaded with SNARF-5F (n > 3 at each pH value; not shown). These results indicate that the effects of changes in [Na+]j on SNARF-based pHj measurements are unlikely to affect the interpretation of results under most experimental conditions. 6.2.4. Concurrent measurements of pH; and fNa"1"! in rat hippocampal neurons Consistent with previous measurements in cultured rat hippocampal neurons (Chapters 3-4; also see Baxter & Church, 1996), resting pHj values in cells single-loaded with carboxy SNARF-1 or SNARF-5F were 7.37 ± 0.03 (n = 13) and 7.39 ± 0.02 (n = 57), respectively. Also consistent with previous reports, resting [Na+]j values in cells single-loaded with SBFI were 12 ± 1 mM (n = 44; see Chapter 5 and Rose & Ransom, 1997; Diarra et al. 2001). These values were not 202 different to those obtained in neurons co-loaded with carboxy SNARF-1 and SBFI (resting pHj 7.40 ± 0.02; resting [Na+]j 10 ± 1 mM; n = 51) or neurons co-loaded with SNARF-5F and SBFI (resting pHs 7.36 ± 0.05; resting [Na+]i 10+1 mM; n = 76). To demonstrate the utility of concurrent measurements of [Na+]i and pH, in cells co-loaded with SBFI and a SNARF derivative, rat hippocampal neurons were subjected to internal acid loads imposed by the NH4+ prepulse technique. Under the nominally HC03"-free, Hepes-buffered conditions employed in the present experiments, the recovery of pHi from NH4+-induced internal acid loads in rat hippocampal neurons is mediated in large part by Na+/Ff~ exchange (e.g. Raley-Susman et al. 1991; Baxter & Church, 1996; Bevensee et al. 1996). In light of its lower pKa value, SNARF-5F was employed in these experiments in preference to carboxy SNARF-1. As expected, in neurons loaded with SNARF-5F alone, pHj increased during NH4+ application, fell to below resting levels upon NH44" washout and then recovered (Table 6.2; Fig. 6.7.4). In neurons loaded with SBFI alone, the washout of MTV" was associated with an increase in [Na+]i that subsequently recovered towards resting levels (Table 6.2; Fig. 6.1B). When cells were co-loaded with SNARF-5F and SBFI, the magnitudes of the decrease in pHj and increase in [Na+]j observed upon NH4+ washout were not significantly different from those seen in neurons loaded with either SBFI or SNARF-5F alone (Table 6.2; Fig. 6.7Q and were comparable to the changes measured concurrently with ISMs in crayfish neurons (see Fig. 4 in Moody, 1981). Further, as shown in Fig. 6.1 D, rates of pHj recovery from NH4+-induced internal acid loads were similar in neurons loaded with either SBFI and SNARF-5F or SNARF-5F alone. Finally, measurements in neurons co-loaded with SNARF-5F and SBFI revealed a positive relationship between the magnitude of the recovery of pHj and the increase in [Na+]j seen after NH4+ washout (Fig. 6.1 E). 203 Taken together, these findings support the feasibility of using either carboxy SNARF-1 or SNARF-5F in conjunction with SBFI to concurrently and accurately measure pHj and [Na+]j. Therefore, in the next series of experiments, I used this technique to further examine the contribution of NaVFT1" exchange activity to anoxia-evoked changes in pHj and [Na+]j. 6.2.5. Contribution of Na+/H+ exchange activity to anoxia-evoked changes in pHi and [Na+]i First, I assessed the validity using a SNARF-based fluorophore and SBFI in combination to measure the changes in pHj and [Na+]i evoked by anoxia in cultured rat hippocampal neurons. Similar to the findings presented in Chapters 3-5 (also see Diarra et al. 1999), in cultured neurons loaded only with SNARF-5F (Table 6.2; Fig. 6.84) or SBFI (Table 6.2; Fig. 6.85), 5 min anoxia produced falls in pHi and increases in [Na+]j, respectively, with recoveries toward resting values upon the return to normoxia. Comparable changes in pH; and [Na+]i were observed in neurons co-loaded with SNARF-5F and SBFI (Table 6.2; Fig. 6.8Q or in neurons co-loaded with carboxy SNARF-1 and SBFI (Table 6.2). Next, I used the 0 [K+]0 protocol (Chapter 5) to examine pHj and/or [Na+]j in the immediate post-anoxic period. In rat hippocampal neurons loaded only with SNARF-5F (Fig. 6.8D) or SBFI (Fig. 6.8£), perfusion with [K+]0-free medium for 7 min immediately after the end of 5 min anoxia did not influence the rise in pHj observed following anoxia (compare Fig. 6.SA and 6.8D) and, as discussed in Chapter 5, revealed further Na+ influx occurring at this time (compare Fig. 6.85 and 6.SE). Comparable changes in pHj and [Na+]i, which increased in parallel after anoxia (Na+,K+-ATPase inhibited; Fig. 6.8F), were observed in neurons co-loaded with either SNARF-5F or carboxy SNARF-1 and SBFI (Table 6.2). These results support the feasibility of using either carboxy SNARF-1 or SNARF-5F in conjunction with SBFI to measure concurrently and accurately the changes in pHj and [Na+]i evoked by anoxia in rat hippocampal neurons. 204 If Na+/H+ exchange activity contributes to the increases in pH, and [Na+]j observed in the immediate post-anoxic period (Na+,K+-ATPase blocked), both events should be dependent on external Na+, and the rates at which pHj and [Na+]i increase after anoxia should exhibit an inverse dependency on pHj (see Lazdunski et al. 1985). Indeed, reducing Na+0 (NMDG+-substitution) prevented the rises in pHi and [Na+]j normally observed after 5 min anoxia (compare Figs. 6.9A and B). As illustrated in Fig. 6.95, when Na+0 was returned to normal (in the continued absence of [K+]0), both pH; and [Na+]j rapidly increased. In addition, when rates at which pHj and [Na+]i increased after anoxia were plotted as functions of absolute pHj (Fig. 6.10), both parameters were faster at lower pHj values. Finally, in neurons co-loaded with either carboxy SNARF-1 or SNARF-5F and SBFI, I examined the effects of maneuvers that have previously been found to influence NaVfT exchange activity in rat hippocampal neurons (see Chapters 4 and 5) on anoxia-evoked changes in pHi and [Na+]j. Consistent with previous suggestions that Na+/FT exchange activity is reduced during and activated immediately following 5 min anoxia, pretreatment with 200 u.M harmaline failed to influence the magnitudes of the fall in pHj or increase in [Na+]j observed during anoxia but reduced significantly the magnitudes of the increases in pHj and [Na+]j observed following anoxia (Fig. 6.11). The activation of NaVFT exchange activity in the immediate post-anoxic period can be inhibited by an extracellular acidosis or inhibition of the cAMP/PKA pathway (Chapters 4 and 5). Conversely, as illustrated in Chapter 3, anoxia-evoked changes in pHj are not influenced by the removal of Ca2+0 prior to, during and following anoxia. Lowering pH0 to 6.60 or the application of 50 pM iq>cAMPS, but not perfusion with Ca -free medium, reduced significantly the magnitudes of the increases in pH, and [Na+]j observed after anoxia (Na+,K+-ATPase inhibited; Fig. 6.11). Taken together, the results obtained in neurons co-loaded with a SNARF-based 205 fluorophore and SBFI strengthen previous suggestions that Na+/FT exchange activity in rat hippocampal neurons is increased in the immediate post-anoxic period and contributes to not only the recovery of pHj but also to Na+ influx at this time. 6.3. DISCUSSION There were two distinct objectives of the present study: in the first part of this study, I developed and characterized a technique for the near-simultaneous measurement of [Na+]i and pHj in rat hippocampal neurons, and; in the second part of this study, I employed this technique to further examine the contribution of Na+/FT exchange activity to the anoxia-evoked changes in pHj and [Na+]i in cultured rat hippocampal neurons. 6.3.1. Part 1: The development of microspectrofluorimetric methods for the concurrent measurement of pHj and [Na+1i The dual excitation ratiometric indicator SBFI has been used to measure [Na+]j in a large number of cell types; the ion selectivity of the probe has been assessed, methods for its in situ calibration have been developed and a one-point technique for the conversion of SBFI-derived ratio measurements into [Na+]j has been validated (Diarra et al. 2001). Similarly, the dual emission ratiometric dye, carboxy SNARF-1, has gained wide acceptance as a pH indicator, in part because of its wide dynamic range and good signal-to-noise ratio (see Buckler & Vaughan-Jones, 1990). Nevertheless, its relatively high pKa may limit the accuracy of low pHi measurements, a potential disadvantage that prompted me to assess the utility of SNARF-5F as a pHj indicator in hippocampal neurons. I confirmed that the pKa of SNARF-5F in situ is lower than that of carboxy SNARF-1 but that it behaves in an otherwise similar manner to carboxy SNARF-1. 206 To date, concurrent single cell measurements of [Na+]i (or intracellular Na+ activity) and pHj have been obtained with ISMs, sometimes in conjunction with an ion-sensitive fluorescent probe (e.g. Thomas, 1977; Moody, 1981; Kaila & Vaughan-Jones, 1987; Deitmer & Schlue, 1989; Munsch & Deitmer, 1997; Kilb & Schlue, 1999). Although ISMs offer some advantages over fluorescent probes (see Nett & Deitmer, 1996; Voipio, 1998), they cannot easily be applied to small cells (e.g. mammalian central neurons). In contrast, SBFI and SNARF derivatives offer means for measuring [Na+]j and pHj, respectively, in small cells that are not amenable to stable impalements with ISMs, but heretofore they have only been employed separately9, precluding a clear understanding of the temporal and other relationships between cytosolic Na+ and H+ homeostasis and the role of either ion in the regulation of specific physiological responses. As noted in the Introduction, [H+]j and [Na+]i, like [H+]j and [Ca2+]j (and also [Na+]j and [Ca2+]j), are related by a variety of mechanisms; indeed, an appreciation of the latter relationships has led to the development of techniques for the concurrent measurement of pH; and [Ca2+]i with carboxy SNARF-1 and fura-2 (Martinez-Zaguilan et al. 1991 and 1996) or indo-1 (Wiegmann et al. 1993; Austin et al. 1996), and [Na+]i and [Ca2+]i with SBFI and fluo-3 (Satoh et al. 1994 and 1995) or fluo-4 (Grant et al. 2002). 6.3.1.1. Part 1: Technical considerations A number of conditions must be satisfied for the accurate measurement of the concentrations of two ions simultaneously with fluorescent probes (see Wiegmann et al. 1993; Martinez-Zaguilan et al. 1996): i) a lack of spectral overlap between the probes; ii) minimal interactions between 9 Jung et al. (1995) employed SBFI and carboxy SNARF-1 simultaneously in isolated heart mitochondria but few details were given and the behaviours of the fluorophores when co-loaded were not systematically examined. 207 the probes; iii) a lack of differential compartmentalization between the probes; iv) binding affinities in the physiological range; v) distinct ion selectivities; and vi) lack of toxicity. In the present study, the combination of SBFI and carboxy SNARF-1 or SNARF-5F fulfilled each of these criteria. First, using the optical equipment specified, the fluorescence signals originating from the fluorophores in situ could be adequately resolved (Fig. 6.2). Second, although carboxy SNARF-1 and SNARF-5F quenched the fluorescence of SBFI (Figs. 6.2 and 6.3), SBFI-derived ratio values and the Na+-sensitivity of SBFI were minimally affected (Figs. 6.3 and 6.4). Third, in agreement with previous reports (e.g. Martinez-Zaguilan et al. 1996; Rose & Ransom, 1997), SBFI- and SNARF-derived fluorescence emissions emanated largely from the same (i.e. cytosolic) compartment (Fig. 6.3). Fourth, the calibration parameters obtained in neurons single-loaded with SBFI or carboxy SNARF-1 were consistent with previous in situ estimates (e.g. Buckler & Vaughan-Jones, 1990; Blank et al. 1992; Diarra et al. 2001; Ch'en et al. 2003) and were not significantly influenced by the presence of a second fluorophore (Table 6.1; Figs. 6.4 and 6.5); similarly, the pKa of SNARF-5F was not affected by the presence of SBFI (Table 6.1). Fifth, carboxy SNARF-1 and SNARF-5F in situ are effectively insensitive to changes in [Na+] (Fig. 6.6); previously, SBFI in situ has been found to be weakly sensitive to changes in pH (Negulescu & Machen, 1990; Rose & Ransom, 1997; Diarra et al. 2001). Sixth, although it is well-established that fluorescent probes employed for ratiometric [ionjj determinations may, under some conditions, exert toxic effects (see Nett & Deitmer, 1996; Voipio, 1998 and references therein), resting [Na+]i and pHj values (and rates of pHj recovery from NHV-induced internal acid loads) obtained in neurons co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F were not significantly different from the respective values obtained in neurons single-loaded with either dye (Fig. 6.7) and are consistent with previous measurements in cultured rat hippocampal neurons (see Raley-Susman et al. 1991; Baxter & Church, 1996; Rose 208 & Ransom, 1997; Diarra et al. 2001). In addition, in initial experiments using perforated patch electrophysiological recordings, our laboratory has observed that cultured rat hippocampal neurons co-loaded with SBFI and SNARF-5F have normal resting membrane potentials, exhibit a stable input resistance over time and generate trains of overshooting action potentials in response to membrane depolarization (T. Kelly and C. Sheldon, unpublished observations). These findings suggest that co-loading neurons with SBFI and SNARF-5F does not, at least in the short-term, adversely affect neuronal viability. 6.3.1.2. Part 1: Summary Together, these findings support the feasibility of using SBFI in conjunction with either carboxy SNARF-1 or SNARF-5F to concurrently and accurately measure [Na+]j and pHj. Indeed, no significant differences were observed in the changes in [Na+]i and pHj evoked by NFLi+-induced internal acid loads (Table 6.2; Fig. 6.7) in neurons single-loaded with SBFI or SNARF-5F or co-loaded with both probes. Concurrent measurements of pHj and [Na+]i will help to resolve the interplay between changes in [Na+]i and pHj under various experimental conditions. Thus, the records shown in Fig. 6.7C and E demonstrate that the recovery of pHi from NPLi+-induced internal acid loads in rat hippocampal neurons is accompanied by a transient rise in [Na+]j, a finding that provides further evidence for Na+/H+ exchange (see Thomas, 1977; Moody, 1981; Kaila & Vaughan-Jones, 1987; Munsch & Deitmer, 1997) in a cell type in which the transport mechanism is insensitive to known pharmacological inhibitors (Raley-Susman et al. 1991; Schwiening & Boron, 1994; Baxter & Church, 1996). Concurrent measurements of [Na+], and pHj also provide the data required to correct, on a region-by-region basis, [Na+]j values estimated with SBFI for changes in pH, Using the methods developed during the course of this thesis and detailed in Diarra et al. (2001), the magnitude of the increase in [Na+]j observed in response to 209 NH.4+ washout increased from 12 ± 2 mM (uncorrected) to 16 ± 3 mM (corrected) (P < 0.05 by paired Student's /-tests). The ability to correct apparent [Na+]j values measured with SBFI for the prevailing pHj is another advantage offered by the simultaneous use of SBFI and carboxy SNARF-1 or SNARF-5F. In summary, I have developed a method for the near-simultaneous measurement of [Na+]j and pHj in cells co-loaded with SBFI and carboxy SNARF-1 or SNARF-5F, a technique that offers a useful alternative to the use of ISMs in experiments employing small cells in which information is required on both [Na+]j and pHj and the relationship between these two ions. It is my hope that the method described will serve as a useful reference point for other investigators who wish to study the interrelationships between [Na+]i and pHj regulation in other cell types. 6.3.2. Part 2: Anoxia-evoked changes in pHi and [Na+]j The results obtained from neurons single-loaded with BCECF (Chapters 3 and 4) or SBFI (Chapter 5) suggested that, although NaVFT exchange activity is inhibited shortly following the onset of anoxia, exchange activity is increased in the post-anoxic period and appears to contribute the increases in pHj and [Na+]j observed immediately upon the return to normoxia. These suggestions are strengthened by the results of the present study in which anoxia-evoked changes in pHj and [Na+]j were measured concurrently. First, harmaline pretreatment (which inhibits NaVFT exchange activity in rat hippocampal neurons; see Chapter 5) had no influence on either the fall in pHj or the increase in [Na+]i observed during anoxia10 but reduced significantly 10 The inability of harmaline to reduce the magnitude of the fall in pHj observed during anoxia (as was observed when Na+/H+ exchange activity was inhibited in acutely isolated adult rat hippocampal neurons under reduced Na+0, NMDG+-substituted conditions; see Chapter 3) may reflect either the inability of harmaline pretreatment to fully inhibit Na+/H+ exchange activity or differences in the extent to which Na7rT exchange activity contributes to the maintenance of resting pHj in acutely isolated adult vs. cultured postnatal hippocampal neurons. 210 the increases in both pHi and [Na+]j observed in the immediate post-anoxic period (Na+,K+-ATPase inhibited). Second, consistent with the pHi-dependency of Na+/H+ exchange activity, the rates at which pHj and [Na+]i increased following anoxia (Na+,K+-ATPase inhibited) were both inversely dependent on pHj. Third, maneuvers that have previously been found to inhibit Na+/H+ exchange activity in rat hippocampal neurons reduced significantly the post-anoxic increases in both pH; and [Na+]j. Although previous studies have illustrated that PKA regulates the activity of plasmalemmal Na+/Ca2+ exchange in rat hippocampal neurons under normoxic conditions (e.g. He et al. 1998), in the present study, the prolonged removal of external Ca2+ (i.e. prior to, during and following anoxia, a maneuver which inhibits forward- and reverse-mode Na+/Ca2+ exchange activity) failed to influence the increases in pHj or [Na+]i observed following anoxia and the ability of Rp-cAMPS to reduce post-anoxic Na+ influx was associated with a parallel reduction in the post-anoxic rise in pHj, suggesting that i?p-cAMPS is likely acting to inhibit Na+/H+ exchange activity in the post-anoxic period in rat hippocampal neurons. Finally, concurrent measurements of anoxia-evoked changes in pHj and [Na+], in the same cells confirmed that the interpretation of the anoxia-evoked changes in [Na+]i, measured in neurons single-loaded with SBFI, are unlikely to be influenced by the pH-sensitivity of the dye. In fact, at a time when SBFI-derived [Na+]i values would be expected to be most influenced by changes in pHj (i.e. when pHj and [Na+]j values were reduced and elevated, respectively; see Diarra et al. 2001), the magnitude of the increase in [Na+]j at the end of 5 min anoxia was underestimated by < 4 mM (the magnitude of the increase in [Na+]j observed at the end of 5 min anoxia in 7 - 10 DIV neurons loaded with SBFI and SNARF-5F was 18 + 4 mM and 21+4 mM, before and after correcting the SBFI-derived [Na+]j value for the 0.44 pH units decrease in pHj, respectively; P < 0.05 by paired Student's Mests). 211 In cardiac tissue, post-ischemic increases in pHj and [Na+]j aggravate myocyte damage and the abilities of pharmacological NaVFT exchange inhibitors to slow both the recovery of pHj and Na+ entry upon reperfusion may, together, underlie their cardioprotective actions (reviewed by Avkiran, 2001). Analogous mechanism(s) may underlie the neuroprotective effects of NaVFT exchange inhibitors. Indeed, NaVFT1" exchange inhibitors limit the increases in pHj observed following periods of anoxia or metabolic inhibition in vitro (e.g. Pirtilla & Kauppinen, 1992; Vornov et al. 1996; also see Ohno et al. 1989; Taylor et al. 1996) and reduce Na+ accumulation and water content during reperfusion in vivo (e.g. Kuribayashi et al. 1999; Horikawa et al. 2001a and b). Although the findings of the present study are consistent with an activation of NaVFT exchange activity in the immediate post-anoxic period and its subsequent contribution to the increases in pHj and [Na+]i observed at this time, it remains unclear whether the neuroprotective actions of NaVFf" exchange inhibitors arise from a reduction in FT1" extrusion and/or Na+ entry (see Chapter 8). It is notable that exposure of neurons to K+-free medium immediately following anoxia (in the presence of normal Na+0) had minimal effects on the increases in pHj observed at this time. In non-neuronal cell types, inhibition of Na+,K+-ATPase activity has been reported to reduce (e.g. Kimura & Aviv, 1993) or not influence (e.g. Aickin & Thomas, 1977) the rate of pHj recovery from imposed internal acid loads. When it has been observed, the slowing of NaVFT" exchange has been, in part, attributed to a reduction in the thermodynamic driving force for NaVFT exchange activity (see Wu & Vaughan-Jones, 1997). Vaughan-Jones & Wu (1990) estimated the "thermodynamic driving force" (DF) of NaVFT exchange as follows DF (RT units) = 2.3 RT(log(|^V[Na+]i) + (pH0 - pHO) (Equation 6.2) 212 where, under the conditions of the present study, [Na+]0 and pH0 are 148 mM and 7.35, respectively, and R and T have their usual meanings. In the present experiments in rat hippocampal neurons in which pHj and [Na+], were measured concurrently, there was an ~ 45% reduction in the driving force for forward NaVH4" exchange during exposure to K+-free medium (DF = 3.1 ±0.1 and 1.7 ± 0.1 RT units at the end of 5 min anoxia and the subsequent 7 min K+-free medium, respectively; n = 49 neurons 7-10 DIV). On the other hand, these calculations indicate that even at end of a 7 min exposure to K+-free medium, Na+/FT exchange will continue to mediate acid extrusion (i.e. the values for DF are positive). Moreover, calculated rates of pHj recovery following anoxia under K+0-free conditions were similar to rates observed following anoxia under control conditions (measured at a common test pHi of 6.95, rates of pHj recovery from anoxia-induced •j o 1 falls in pHj were 1.58 ± 0.29 x 10" and 2.38 ± 0.88 x 10" pH units s" in the presence and absence of external K+; n = 6 and 7, respectively; P = 0.35; data obtained from neurons 7-10 DIV). It is apparent that, in the present study, despite reductions in its thermodynamic driving force, Na+/PI+ exchange activity is enhanced in the immediate post-anoxic period, suggesting that the rate of Na+/H+ exchange activity in the period following anoxia is likely regulated by a variety of intracellular events, including anoxia-evoked changes in the activity of the cAMP/PKA pathway, although other mechanisms may be involved (e.g. the restoration of internal ATP levels and/or regulation by free radicals; Yao et al. 2001; Mulkey et al. 2004). In non-neuronal cell types, under conditions of Na+,K+-ATPase inhibition, the ratio between the net flux of FT1" vs. Na+ ions has previously provided reasonable estimates of the coupling ratio of NaVff" exchange (e.g. Grinstein et al. 1984; Aronson, 1985; Kaila & Vaughan-Jones, 1987). However, from the present study, I am unable to provide similar estimates of the stoichiometry of Na+/Ff~ exchange in rat hippocampal neurons: intrinsic H* buffering power cannot be assessed 213 accurately in rat hippocampal neurons (because of the inability to completely inhibit the activities of all pHi regulating mechanisms in rat hippocampal neurons) and it is apparent from the results presented in this study that the increases in pHj and [Na+]j observed following anoxia do not simply reflect only changes in Na+/H+ exchange activity. Rather, as detailed in Chapters 4 and 7, the increases in pHj and [Na+]i observed following anoxia in rat hippocampal neurons reflect co-ordinated changes in the activities of multiple pHj regulating mechanisms and Na+ influx pathways. In summary, the results of the present study, in which anoxia-evoked changes in pHj and [Na+]i were measured concurrently in the same neurons, have strengthened the previous suggestion that NaVFT1" exchange activity in rat hippocampal neurons is inhibited during anoxia and is increased in the immediate post-anoxic period, at which time it contributes to potentially detrimental increases in pHj and [Na+]i-214 Table 6.1: Calibration parameters for SBFI, carboxy SNARF-1 and SNARF-5F in single dye-and dual dye-loaded hippocampal neurons A, SBFI calibration Neurons loaded with Temp Ki p Rn(min) Rn(max) n SBFI 37°C 21.7 ±2.1 4.5 ± 0.7 0.77 ±0.01 3.2 ±0.4 21 SBFI + carboxy SNARF-1 37°C 17.6 ±3.5 5.4 ± 1.3 0.77 ± 0.03 3.2 ± 0.4 5 SBFI + SNARF-5F 37°C 22.7 ±4.8 4.5 ± 1.3 0.78 ± 0.03 3.3 ±0.5 8 B, carboxy SNARF-1 and SNARF-5F calibrations Neurons loaded with Temp pKa F640m\„/max Rn(min) Rn(max) " Carboxy SNARF-1 22°C 7.40 ± 0.06 0.83 ± 0.06 0.34 ± 0.05 1.3 ±0.04 7 Carboxy SNARF-1 37°C 7.59 ±0.13 0.62 ±0.05* 0.18 ±0.04* 1.4 ±0.07 19 Carboxy SNARF-1 + SBFI 37°C 7.62 ± 0.09 0.46 ± 0.07 0.19 ±0.05 1.3 ±0.06 9 SNARF-5F 37°C 7.25 ±0.12 0.45 ± 0.05 0.29 ± 0.06 1.7 ±0.19 12 SNARF-5F + SBFI 37°C 7.30 ± 0.09 0.58 ±0.11 0.27 ± 0.06 1.8 ±0.17 12 Values are means ± s.E.M. *, P < 0.05 compared to the corresponding value at 22°C. 215 Table 6.2: NH4 prepulse- and anoxia-evoked changes in pHj and [Na ]\ in hippocampal neurons loaded with a SNARF-based fluorophore and/or SBFI A, NFlV" prepulse Neurons loaded with Decrease in pIF, on NH4+ Increase in [Na+]j on NH4+ n washout (pH units) washout (mM) SNARF-5F 0.81 ±0.08 39 SBFI 12 ±1 25 SBFI + SNARF-5F 0.87 ±0.12 13 ±2 14 B, During anoxia Neurons loaded with Decrease in pH, (pH units) Increase in [Na+]j (mM) n SNARF-5F 0.40 + 0.02 18 SBFI 18 ±1 19 SBFI + SNARF-5F 0.44 ± 0.03 18 ±4 8 SBFI + carboxy SNARF-1 0.37 ± 0.04 16 + 4 15 C, After anoxia (Na+,K+-ATPase blocked by perfusion with K+-free medium) Neurons loaded with Increase in pH, (pH units) Increase in [Na+]j (mM) n carboxy SNARF-1 0.18 ± 0.02 13 SBFI 20 ± 5 11 SBFI + carboxy SNARF-1 0.25 ± 0.03 22 + 2 7 SBFI + SNARF-5F 0.20 ± 0.02 23 ± 2 2 Values are mean ± S.E.M. In A and B, the changes in pHi and [Na+]j shown are with respect to pre-stimulus resting values (data obtained from neuronal cultures 10-14 and 7-10 DIV, respectively). In C, measurements of the increases in pHj and [Na+]j observed following anoxia (Na ,K -ATPase inhibited) were measured as the differences between the pHj and [Na ]i values observed at the end of 5 min anoxia and the pHj and [Na+]j values observed at the end of 7 min exposure to 0 [K+] medium; data were obtained from neuronal cultures 7-10 DIV. 216 Fig. 6.1. A schematic representation of the optical equipment used to measure [Na+]j and pHj in hippocampal neurons loaded with SBFI and/or carboxy SNARF-1 or SNARF-5F (adapted from Buckler & Vaughan-Jones, 1990). Fluorophores loaded into neurons were excited with light provided by a 100 W Hg lamp and band-pass filtered at 488 ± 5 nm (for SNARF dyes) or alternately at 334 ± 5 and 380 ± 5 nm (for SBFI). Filtered excitation light was reflected off a 505 nm long-pass dichroic mirror (Dichroic 1), passed through a x40 LD Achroplan objective (0.6 N.A.) and illuminated neurons loaded with fluorophore(s). SBFI fluorescence emissions passed through Dichroic 1, reflected off a long-pass dichroic mirror centred at 605 nm (Dichroic 2), passed through a 550 + 40 nm band-pass emission filter and were detected by Camera 2. Fluorescence emissions from carboxy SNARF-1 or SNARF-5F passed through Dichroic 1, were split by Dichroic 2 and passed through 640 ± 20 nm or 550 ± 40 nm band-pass emission filters before being detected by Cameras 1 and 2, respectively. Optical filters, including Dichroic mirrors 1 (505dcxru) and 2 (605drlpxr), were obtained from Chroma Technology Corp. (Rockingham, VT). 217 Excitation wavelengths: 334 nm • 380 nm > 488 nm • Emission wavelengths: 640 nm • 550 nm r Excitation filters I 334 nm Lamp 38p_nm_ 48.8.ntn... Camera 2 Emission Filter 550 nm Objective Dichroic 1 505 nm Dichroic 2 605 nm Emission Filter 640 nm Camera 1 218 Fig. 6.2. Fluorescence emissions from single and dual dye-loaded hippocampal neurons. Neurons were single- or co-loaded with fluorophores, as shown on the Figure, and illuminated at the indicated excitation wavelengths (A,ex). The intensities of emitted fluorescence were measured at either 550 nm or 640 nm (Xem) and, in A and B, are reported as percentages above background fluorescence values (± S.E.M.). A, in contrast to neurons single-loaded with either carboxy SNARF-1 or SNARF-5F, neurons single-loaded with SBFI exhibited large fluorescence emissions (measured at 550 nm) during excitation at 334 or 380 nm. Emission intensities from neurons single-loaded with carboxy SNARF-1 were 5% (excitation at 334 nm) and 9% (excitation at 380 nm) of those observed in neurons single-loaded with SBFI and excited at the same wavelengths. In neurons single-loaded with SNARF-5F and excited at 334 and 380 nm, emission intensities were, respectively, 6% and 7% of those observed in neurons single-loaded with SBFI (compare the appropriate images in the left-hand and middle panels in Q. Also shown are fluorescence emission intensities (measured at 550 nm) during 334 and 380 nm excitation of neurons co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F (see text and the right-hand panel in C). B, in neurons single-loaded with either carboxy SNARF-1 or SNARF-5F, excitation at 488 nm evoked fluorescence emissions at both 550 and 640 nm. In neurons single-loaded with SBFI and also excited at 488 nm, fluorescence emissions measured at 550 and 640 nm were, respectively, 11% and 15% of those observed in neurons single-loaded with carboxy SNARF-1 and 13%o and 13%) of those observed in neurons single-loaded with SNARF-5F (compare the appropriate images in the left-hand and middle panels in C). Also shown are fluorescence emission intensities (measured at 550 and 640 nm) during 488 nm excitation of neurons co-loaded with SBFI and either carboxy SNARF-1 or SNARF-5F (see text and the right-hand panel in C). *, P < 0.05 between the indicated measurements in single- and 219 dual dye-loaded neurons. C, representative pseudocoloured fluorescence emission images from neurons loaded with SBFI (left-hand panel), SNARF-5F (middle panel) or SBFI and SNARF-5F (right-hand panel); each panel consists of four images, captured at the excitation and emission wavelengths indicated on each image in the left-hand panel. Fluorescence intensity i> (percentage above background) (D d> cn co <J1 co o 3 3 3 3 cn co cn oo o o 3 3 3 3 o o o o o Ul o o tO M CO O Ol o o oo o o _1 I I TTTTTTTH I Fluorescence intensity (percentage above background) CD cn ^ cn CD o Co 3 3 3 3 cn o o o o o cn o o ro o o o 2 1 V/////////M CO O CO 3 3 3 3 mrjj ] ESS ] CO W CO O CO ro crj z SJ ro Tl Tl > CT Tl - - 5 o -sa w ^ Tl Z Ul ^ a" S > II Tl CO "* II CO l\J 221 Fig. 6.3. Quenching effects between SBFI and SNARF-based fluorophores. A, in a population of 19 neurons single-loaded with SBFI, background-subtracted emission intensities were measured at 550 nm during excitation at 334 nm (5/334, open circles) and 380 nm (5/380, open diamonds). Perfusion and data collection were then interrupted and 10 uM SNARF-5F-AM was added to the recording chamber. After loading with SNARF-5F, SBFI-derived 5/334 and 5/380 values were reduced to a proportionately similar extent, resulting in little change in SBFI 5/334/5/380 ratio values (filled circles). Data collection was then interrupted again, during which time the gain of camera 2 was adjusted to increase emission intensities measured during excitation at 334 and 380 nm. Finally, the addition of 0.005% saponin (arrow) caused rapid decreases in 5/334 and 5/38o values. Also measured throughout the experiment were background-subtracted emission intensities at 550 nm (5/550, open squares) and 640 nm {BIMO, open triangles) during excitation at 488 nm; 5/550 and 5/640 values increased after neurons were loaded with SNARF-5F and fell rapidly upon the subsequent addition of saponin. Temperature was 37°C throughout. 5, neurons single-loaded with SNARF-5F (filled symbols) or co-loaded with SBFI and SNARF-5F (open symbols) were exposed to SNARF calibration media at pH 6.0, 7.0 or 8.5. The intensities of emitted fluorescence (excitation at 488 nm) were measured at 550 (circles) and 640 (squares) nm and are reported as percentages above background fluorescence values (n > 12 for each data point). Compared with measurements made in neurons single-loaded with SNARF -5F, emission intensities at 550 and 640 nm were not significantly different (P > 0.20 at each pH value) in the presence of SBFI at any of the pH values tested. Error bars are S.E.M. Background-subtracted fluorescence intensity (arbitrary units) 334'D'380 Fluorescence emission intensity (percentage above background) Co ro o o o o o o •o 01 b CT) b b 00 b co b am HOH •3 t-cmirCMi 1—Dm 00 o o o o o • • 0 • 3 3 CO CO cn ^ CO CO cn cn z z 0 z z 0 > > 13 > > 7J 3 7J 73 T1 Tl Tl -Tl cn (In (In n n -n -n + + CO CO CD CD -n n to 223 Fig. 6.4. In situ calibration of SBFI at 37°C, pH0 7.35. A, a full calibration experiment in which 20 neurons single-loaded with SBFI were exposed to 4 uM gramicidin D-containing solutions at the [Na+] values (in mM) indicated above the traces. Shown are the mean changes in emitted background-subtracted fluorescence intensities measured at 550 nm during excitation at 334 nm (5/334, open circles) and 380 nm (5/380, open diamonds) and the mean normalized /J/334AB/380 ratio values (Rn, filled circles), which increased as [Na+] increased. Also shown are the background-subtracted fluorescence emission intensities measured at 550 nm (5/550, open squares) and 640 nm (BI^o, open triangles) during excitation at 488 nm (i.e. the excitation and emission wavelengths used for SNARF measurements); note the change of scale. 5, plots of [Na+] vs. Rn obtained from experiments of the type shown in A. Data points fitted by the continuous line were obtained from four experiments conducted on sister cultures in which neurons were loaded with SBFI alone. Data points fitted by the dashed line were obtained from four experiments conducted on sister cultures in which neurons were co-loaded with SBFI and SNARF-5F. In both cases, error bars are S.E.M. AS detailed in Chapter 2, the curves are the result of at fit to a three-parameter hyperbola (Equation 2.4) to the respective data points indicated and were used to determine the values of the SBFI calibration parameters (i.e. Kd, (3, Rn(min) and Rn(max)) under the different dye loading conditions (Table 6.1; also see Diarra et al. 2001). 225 Fig. 6.5. In situ calibration of carboxy SNARF-1. A, a full calibration experiment performed at 22°C in which 18 neurons single-loaded with carboxy SNARF-1 were exposed to high-[K+], 10 uM nigericin-containing solutions at the pH values shown above the records. Background-subtracted fluorescence emissions measured at 550 nm (BI55Q, open squares) increased upon protonation of the dye while background-subtracted emissions measured at 640 nm (73/640, open triangles) decreased; thus, the resulting background-subtracted BIsso/BIew ratio values normalized to 1.00 at pH 7.00 (Rn, filled circles) increased as pHj fell. Also shown are the background-subtracted fluorescence emission intensities measured at 550 nm during excitation at 334 nm (5/334, open circles) and 380 nm (BI^o, open diamonds) (i.e. the excitation and emission wavelengths used for SBFI measurements); /3/334 and 5/38o emissions from carboxy SNARF-1 (and SNARF-5F; not shown) remained small as pH was altered (note the change of scale). B, plots of pH vs. Rn obtained from experiments of the type shown in A. In neurons single-loaded with carboxy SNARF-1, data points obtained from experiments performed at 22°C and 37°C (n = 3 experiments conducted on sister cultures in each case) were fitted using non-linear least squares regression (dashed and continuous lines, respectively). Also shown are data points obtained from three experiments performed on sister cultures in which neurons were co-loaded with carboxy SNARF-1 and SBFI; these are fitted by the dotted line. In all cases, error bars are S.E.M. The curves were used to determine the values of the carboxy SNARF-1 calibration parameters (i.e. pKa, logF640min/max, Rn(min) and Rn(max)) under the different dye loading conditions (see Table 6.1). 227 Fig. 6.6. Sodium sensitivity of carboxy SNARF-1 in situ. A, neurons loaded with carboxy SNARF-1 were exposed at 37°C to calibration media containing 4 uM gramicidin at pH 6.00 (open squares), 6.50 (open circles), 7.00 (filled circles) or 7.50 (open diamonds). At each pH, [Na+] was changed from 0 to 10 to 130 mM as indicated above the records. BI^IBI^Q ratio values (Rn) were normalized to unity at pH = 7.00 and [Na+] - 10 mM. Inset, to quantify the effects of changes in [Na+] on Rn values measured with carboxy SNARF-1, apparent changes in Rn (ARn) were calculated as Rnr*) - Rn(io) (where x = 0, 10 or 130) for each pH value indicated and plotted as a function of [Na+]. B, Rn values at a given [Na+] were plotted as a function of pH (n = 3 at pH 6.50, 7.00 and 7.50; n = 1 at pH 6.00) and data points were fitted using non-linear least squares regression (continuous, dotted and dashed curves for data obtained at 0, 10 and 130 mM [Na+], respectively). The pKa + log.F640mjn/max, Rn(min) and Rn(max) values for carboxy SNARF-1 did not change significantly as [Na+] was increased from 0 to 10 to 130 mM (pKa + logF640min/max, Rn(min) and Rn(max) values were, respectively, 7.35, 0.44 and 1.25 at 0 mM [Na+]; 7.32, 0.47 and 1.26 at 10 mM [Na4]; and 7.28, 0.46 and 1.30 at 130 mM [Na+]). Inset, Rn values were converted into pHj and apparent changes in pHj (ApHj) were then calculated as pHj(x) -pHj(io) (where x = 0, 10 or 130) for each pH value indicated and plotted as a function of [Na4]. In the insets in A and B, each datum point represents measurements made in 3 separate experiments; error bars are S.E.M. and the continuous lines represent linear regression fits to the data points indicated for each pH value. 228 229 Fig. 6.7. Changes in pHj and [Na+]j observed in rat hippocampal neurons in response to intracellular acid loads imposed by the NH.4+ prepulse technique. A, in a neuron single-loaded with SNARF-5F, washout of NFf4+ evoked a fall in pH, which gradually returned to the resting level. B, in a neuron single-loaded with SBFI, an increase in [Na+]j occurred upon the washout of NH4+. C, in a neuron co-loaded with SNARF-5F and SBFI, the changes in pH; (filled circles) observed on NH44" washout were temporally associated with a transient increase in [Na+]i (open circles). D, rates of pHj recovery from NH4+-induced internal acid loads measured in neurons either single-loaded with SNARF-5F (solid circles; n = 39) or co-loaded with SNARF-5F and SBFI (open circles; n = 14). Error bars are s.E.M. and continuous lines represent weighted nonlinear regression fits to the data points indicated for each experimental condition. Rates of pHj recovery in neurons single-loaded with SNARF-5F were not significantly different from those in neurons co-loaded with SNARF-5F and SBFI (P > 0.05 at each absolute pH; value). E, scatter plot demonstrating the relationship between the magnitude of the recovery of pHj (measured as the difference between the minimum pHj value attained after NH4+ washout and the steady-state pHj value reached after recovery) and the increase in [Na+]i observed after NH4"1" washout in 14 neurons co-loaded with SNARF-5F and SBFI. The continuous line is a linear regression fit to the data points shown (correlation coefficient = 0.68; P < 0.05). 230 Recovery of pHj (pH units) 231 Fig. 6.8. Changes in pH, and [Na+]j induced by anoxia in rat hippocampal neurons. A, in a neuron single-loaded with SNARF-5F, 5 min anoxia evoked a fall in pHj that recovered towards the resting value upon the return to normoxia. B, in a neuron single-loaded with SBFI, 5 min anoxia induced an increase in [Na+]j that recovered to the resting value upon the return to normoxia. C, in a neuron co-loaded with SNARF-5F and SBFI, 5 min anoxia induced a fall in pH; (solid circles) and an increase in [Na+]i (open circles). Upon the return to normoxia, both pHf and [Na+]i recovered toward pre-anoxic values. D, in a neuron single-loaded with SNARF-5F, inhibition of Na+,K+-ATPase activity (by perfusion with K+-free medium) for 7 min following 5 min anoxia did not influence the recovery of pHj (compare with A). E, in a neuron single-loaded with SBFI, inhibition of Na+,K+-ATPase activity following 5 min anoxia revealed a secondary increase in [Na+]j in the post-anoxic period (compare with B). Once Na+,K+-ATPase activity was re-established (i.e. 3 mM [K+]0), [Na+]j recovered to pre-anoxic values. F, in a neuron co-loaded with SNARF-5F and SBFI, both pHj and [Na+]j increased during the period of Na+,K+-ATPase inhibition after anoxia. In A - F, traces were obtained from neurons 7-10 DIV. 232 B Anoxia Anoxia 5 10 15 20 25 30 Time (min) CL o T 1 r 10 15 20 25 Time (min) Time (min) D CL Anoxia 1 1 1 1 1 1 5 10 15 20 25 30 Time (min) 5 10 15 20 25 30 Time (min) x CL O T—i—i—i—r 5 10 15 20 25 30 Time (min) 233 Fig. 6.9. Reducing [Na+]0 limits the increases in pHj and [Na+]-, observed immediately after anoxia. A, during 5 min anoxia, pHj fell and [Na+]j increased. Upon the return to normoxia (Na+,K+-ATPase inhibited), both pHi (filled circles) and [Na+]i (open circles) increased (compare with Fig. 6.8F, an identical experiment performed in a neuron 7 DIV). B, during 5 min anoxia, in the presence of normal Na+0, pHj (filled squares) fell and [Na+]j (open squares) increased. Following anoxia, neurons were perfused with K+- and Na+-free medium. Reducing Na+0 (NMDG+-substitution) prevented the rises in pHj and [Na+]j observed after anoxia in the presence of normal Na+0 (compare with A). Upon the return to normal Na+0 (in the continued absence of [K+]0; arrow), both pHj and [Na+]j increased. This trace is representative of data obtained from 6 additional neurons co-loaded with SBFI and SNARF-5F. 234 Time (min) Time (min) 235 Fig. 6.10. Relationships between changes in pHj and [Na+]i observed in the period immediately after anoxia (Na+,K+-ATPase inhibited). A, measured in 20 neurons 7-10 DIV (co-loaded with SNARF-5F and SBFI), rates at which pHj (open circles) and [Na+]j (filled circles) increase in the immediate post-anoxic period (Na+,K+-ATPase blocked) show an inverse dependence on absolute pHj values. Similar results were observed in neuronal cultures 11-14 DIV (B; n= 10). Rates at which pHi and [Na+]j increased following anoxia under [K+]0-free conditions were determined by fitting the pHi and [Na+]j records obtained under [K+]0-free conditions to single exponential functions. The first derivatives of these functions were used to determine rates of pHj recovery and [Na+]i rise as functions of time. Rates of pHi and [Na+]i rises were determined at 0.05 pH unit and 5 mM intervals, respectively. The pHj values at which rates of [Na+]i rises were measured were determined from obtained curve-fitted parameters. Error bars represent S.E.M. 236 e 0.10 i 0) 0.09 -0.08 -0.07 -+ ro z 0.06 -0.05 -• 0.04 -0.03 -7-10 DIV 4.0 3.5 |- 3.0 2.5 - 2.0 - 1.5 1.0 cr 0.5 co^ o.o " o Q. -a _i . q w "U I c 3 E 0.12 0.11 •a +— 0.10 ro T3 0.09 0.08 11 - 14 DIV mil 6.6 6.8 7.0 7.2 7.4 7.6 I I I I I I I 6.8 6.9 7.0 7.1 7.2 7.3 7.4 7.5 7.6 2.5 2.0 1.5 1.0 0.5 0.0 237 Fig. 6.11. The influence of maneuvers which inhibit NaVFT1" exchange activity on anoxia-evoked changes in pHj and [Na+]i measured concurrently in individual cells co-loaded with either carboxy SNARF-1 or SNARF-5F and SBFI. The magnitudes of the falls in pHj (bars with horizontal lines) and increases in [Na+]i (cross-hatched bars) observed during anoxia were measured as the difference between the minimum pHj and maximum [Na+]j observed at the end of 5 min anoxia and the pre-anoxic resting pHj and [Na+]j values, respectively. The magnitudes of the increase in pHj (filled bars) and [Na+]j (open bars) observed after anoxia (Na+,K+-ATPase inhibited) were measured as the difference between the pH, and [Na+]i values observed at the end of anoxia and the pHj and [Na+]i values observed at the end of 7 min 0 [K+]0, respectively. Measurements under an experimental test condition were normalized to measurements made in experiments performed on age-matched sister cultures under control conditions. Statistical comparisons were performed by comparing the absolute measurements of the anoxia-evoked changes in pH, and [Na+]j (i.e. not normalized) made under experimental test conditions to corresponding measurements made in age-matched sister cultures under control conditions. * indicates statistical significance (P < 0.05) compared with measurements made in sister cultures under control conditions. All data were obtained from neuronal cultures 7-10 DIV. 238 F—| Fall in pH, during anoxia ^ Rise in pH, after anoxia Rise in [Na+]| during anoxia [~~J Rise in [Na+], arlfer anoxia 1.4 n Control 200 nM Control 200 pH0 6.60 50 0 [Ca ] harmaline harmaline (n = 9) Rp-cAMPS in-^Q\ (n = 10) (n=10) n = 9) 239 CHAPTER SEVEN ADDITIONAL MECHANISMS CONTRIBUTING TO ANOXIA-EVOKED INCREASES IN [Na+]i IN CULTURED POSTNATAL RAT HIPPOCAMPAL NEURONS 7.0. INTRODUCTION The detrimental effects of increases in [Na+]j evoked by anoxia or ischemia are well-established; however, in contrast to non-neuronal cell types (e.g. cardiac myocytes; Carmeliet, 1999) and myelinated central nervous system axons (Stys, 1998), the mechanisms which mediate Na+ influx in response to anoxia or ischemia in mammalian central neurons remain relatively poorly defined. Although Na+ influx through glutamate receptor-operated channels has received some attention (Mtiller & Somjen, 2000a, LoPachin et al. 2001), few studies have examined the potential contributions of mechanisms integral to the cell to the increases in [Na+]i observed during anoxia or ischemia (e.g. Chen et al. 1999 for [Na+]j measurements in cerebellar granule cells during metabolic inhibition; also see Guatteo et al. 1998; Pisani et al. 1998a; Calabresi et al. 1999b for studies in slice preparations). In addition, despite indications that continued Na+ influx upon reperfusion may be more damaging than Na+ entry during anoxia or ischemia (Lipton, 1999; also discussed in Chapter 8), the pathways that mediate Na+ entry immediately after anoxia/ischemia have not been characterized and it remains unknown whether these 9+ pathways might differ from those active during an insult, as reported for Ca (e.g. Silver & Erecihska, 1990 and 1992; Stys & LoPachin, 1998). As detailed in Chapters 5 and 6, postnatal rat hippocampal neurons responded to 5 min anoxia with an increase in [Na+]j of -15 - 40 mM that, upon the return to normoxia, recovered to pre-anoxic values despite continued Na+ entry. Under conditions which inhibit the Na+,K+-ATPase following anoxia (perfusion with K+0-free medium or the application of ouabain), a 240 further elevation in [Na+]i of -30 - 60 mM was revealed. The increases in [Na+]j observed during and following anoxia were related to the duration that neurons had been maintained in culture (within the examined range of 6 - 14 DIV) and, in all neurons examined, were dependent on the influx of external Na+ ions. The results presented in Chapters 5 and 6 also indicated that Na+/H+ exchange contributes to Na+ influx immediately after, but not during, anoxia; thus, it is apparent that other mechanisms must contribute to the Na+ influx occurring during and following anoxia in rat hippocampal neurons. In this study, I assessed the potential contribution of intrinsic mechanisms other than Na+/H+ exchange to the rises in [Na+]i that occur during and following 5 min anoxia in postnatal rat hippocampal neurons. 7.1. MATERIALS AND METHODS 7.1.1. Experimental preparation and solutions Primary cultures of hippocampal neurons prepared from 2-4 day old postnatal Wistar rats were employed in all experiments presented in this Chapter. Hyperosmolar solutions were prepared by adding 50 or 100 mM mannitol to standard, Hepes-buffered solutions: the osmolalities of these solutions were measured with an uOsmette osmometer (Precision Systems, Inc., Natick, MA), calibrated before use. The osmolalities of standard Hepes-buffered media in the absence of mannitol and of standard Hepes-buffered media to which 50 and 100 mM mannitol had been added were 296 ± 2 (n = 4), 345 ± 1 (n = 2) and 392 ± 2 mOsm kg H20"' (n = 4), respectively. 241 7.1.2. Recording techniques Details of the techniques used to load cultured postnatal rat hippocampal neurons with SBFI-AM are presented in Chapter 5. Details of the techniques used to measure SBFI-derived emission intensity ratios and the conversion of these ratio values to [Na+]j are described in Chapter 2. 7.1.3. Experimental procedures and data analysis As described in Chapter 5, changes in [Na+]j observed during anoxia (A[Na+]i(during)) were measured as the difference between the pre-anoxic resting [Na+]j value and the [Na+]j value observed at the end of 5 min anoxia. Changes in [Na+Jj observed following anoxia (A[Na+]j(after)) were measured as the difference between the [Na+]j value observed at the end of 5 min anoxia and the [Na+]j value observed at the end of a 7 min exposure to 0 [K+]0 or 500 pM ouabain. Data are reported as mean ± s.E.M. with the accompanying n value referring to the number of neuronal populations (i.e. coverslips) from which data were obtained. In light of the findings presented in Chapter 5 that the increases in [Na+]j observed during and after anoxia were related to the duration of time that neurons had been maintained in culture, measurements of anoxia-evoked increases in [Na+]j observed under a given test condition were normalized to [Na+]i measurements made in control experiments performed on the same day using age-matched sister cultures (see Tables 7.1 and 7.2). Statistical comparisons were performed by comparing absolute [Na+]j measurements (i.e. not normalized A[Na+]i(during) and A[Na+]j(after)) made under a given test condition to the corresponding measurements made in age-matched sister cultures under control conditions using Student's two-tailed unpaired r-tests. Significance was assumed at the 5% level. 242 7.2. RESULTS 7.2.1. Increases in [Na*], during anoxia 7.2.1.1. Role of ionotropic glutamate receptor-operated channels Ionotropic glutamate receptor activation did not contribute to the increase in [Na+]i induced by anoxia under the constant supervision conditions of the present experiments. Thus, in neurons maintained for either 6 - 10 or 11 - 14 DIV, the addition of 2 uM MK-801 and/or 20 uM CNQX failed to influence significantly the magnitude of the increase in [Na+]j observed during anoxia (Table 7.1; Fig. 7.1). In contrast, the same concentrations of MK-801 and CNQX together abolished the increase in [Na+]j evoked by 30 s applications of 20 uM NMDA and 20 uM AMP A under normoxic conditions (Fig. 7.1, inset). 7.2.1.2. Role of voltage-activated Na+ channels In whole-cell recordings obtained using the perforated patch (amphotericin B) configuration (courtesy of T. Kelly), the average resting membrane potential of cultured postnatal rat hippocampal neurons prior to anoxia was -62 ± 1 mV and 5 min anoxia evoked a depolarization of 21 ±2 mV in = 3), similar to the depolarizations observed by others in a variety of isolated mammalian central neurons in reponse to 5 min anoxia (e.g. Haddad & Jiang, 1993) or 30 min metabolic inhibition (e.g. Pisani et al. 1997; Aarts et al. 2003). Despite the anoxia-induced membrane depolarization, 1 uM TTX failed to affect the increase in [Na+]j observed during anoxia in neurons maintained either for 6 - 10 or 11 - 14 DIV (Table 7.1; Fig. 7.2A). In contrast, 1 uM tetrodotoxin (TTX) reduced the increases in [Na+]j evoked by 60 s applications of 50 mM 243 K o under normoxic conditions from 12 ± 3 (n = 5) to 3 ± 1 (n = 6) mM (Fig. 7.2A, inset, P < 0.05). 7.2.1.3. Role of plasmalemmal Na+/Ca2+ exchange and Na+/K+/2C1" cotransport While the increase in [Ca2+]j observed in cultured postnatal rat hippocampal neurons during 5 min anoxia (see Diarra et al. 1999) could activate forward mode Na+/Ca2+ exchange and thereby contribute to the influx of Na+ during anoxia, a rise in [Na+]j could promote reverse-mode operation of the exchange mechanism and, thus, Na+ efflux (Blaustein & Lederer, 1999). Forward- and reverse-mode operation of the plasmalemmal Na+/Ca2+ exchanger can be inhibited by bepridil (50 pM) and KB-R7943 (1 pM), respectively, while elevated concentrations of KB-R7943 (10 uM) have been reported to inhibit both forward- and reverse-mode Na+/Ca2+ exchange activity in hippocampal neurons (Breder et al. 2000). Neither bepridil (50 pM) nor KB-R7943 (1 or 10 pM) influenced the magnitude of the increase in [Na+]j observed during anoxia (Table 7.1; Fig. 7.25). In addition, neither CGP-37157 (25 pM), an inhibitor of plasmalemmal Na /Ca exchange in cerebellar granule cells (Czyz & Kiedrowski, 2003; but see Zhang & Lipton, 1999 for an illustration of CGP-37157 acting to inhibit mitochondrial Na+/Ca2+ exchange in rat hippocampal slices) nor the removal of external Ca2+ prior to and during anoxia, influenced the increase in [Na+]j observed during anoxia (Table 7.1). Inhibition of Na+/K+/2C1" cotransport with bumetanide reduces infarct volume and brain edema following focal cerebral ischemia (Yan et al. 2001 and 2003). Given that Na+/K+/2C1" cotransport protein expression and transport activity increases with time in cultured neocortical neurons (Sun & Murali, 1999; Beck et al. 2003), the effects of bumetanide were examined in 6 -10 and 11 - 14 DIV cultured neurons. Exposure to 50 - 100 pM bumetanide did not affect resting 244 [Na+]i in either 6 - 10 (n = 5) or 11 - 14 (n = 9) DIV neuronal cultures (also see Rose & Ransom, 1997) and failed to influence the magnitude of the increase in [Na+]j observed during anoxia in 6 - 10 DIV cultures (Table 7.1). In contrast, in neurons 11-14 DIV, bumetanide caused a significant reduction in the rise in [Na+], observed during anoxia (Table 7.1; Fig. 7.2Q. 7.2.1.4. Role of non-selective cation channels Non-selective cation channels (NSCCs) can be activated during anoxia or ischemia as a result of membrane stretching, increases in [Ca ]i or free radical production (e.g. Chen et al. 1999; Barros et al. 2001; Aarts et al. 2003) and have been found to participate in the production of a variety of events that occur in response to anoxia or ischemia in isolated neurons (e.g. Chen et al. 1997; Chen et al. 1998a; El-Sherif et al. 2001; Aarts et al. 2003; Limbrick et al. 2003; Smith et al. 2003). To examine the potential contribution of Na+ influx through NSCCs to the rise in [Na+]j observed during anoxia, I applied Gd3+, an established blocker of stretch-activated and other types of NSCCs (see Caldwell et al. 1998). In the presence of 30 or 100 uM Gd3+, the anoxia-induced increase in [Na ]•, was reduced by -40 % (Fig. 7.3A, D). When neurons were exposed to 50 or 100 mM mannitol to limit anoxia-induced cell swelling (see Alojado et al. 1996; Hasbani et al. 1998), the increase in [Na+]i observed during anoxia was enhanced, not reduced, when compared to measurements made under control conditions (Fig. 7.35, D), an effect which may reflect the activation of Na+ influx pathways under these conditions (e.g. Shrode et al. 1997; Bevensee et al. 1999b; Gosmanov et al. 2003). Nevertheless, applied in the presence of 100 mM mannitol, Gd3+ continued to reduce the magnitude of the increase in [Na+]i observed during anoxia (Fig. 7.3C,D), suggesting that the ability of Gd3+ to significantly reduce the increase in [Na+]j observed during anoxia cannot 245 simply reflect its ability to inhibit stretch-activated NSCCs. Similarly, the finding presented above that the increase in [Na+]j during anoxia is not reduced in the absence of external Ca2+ suggests that the ability of Gd3+ to reduce the increase in [Na+]i during anoxia is not likely mediated via an inhibition of NSCCs activated by increases in [Ca2+]j (we were unable to examine the effects of flufenamate, an inhibitor of Ca2+i-activated NSCCs in hippocampal neurons, on the increases in [Na+]j during anoxia because it evoked variable increases in resting [Na+]i; also see Partridge & Valenzuela, 2000). In addition, neither the broad-spectrum Ca2+ channel blockers, Ni2+ and verapamil, nor the L-type Ca2+ channel blocker, nifedipine, influenced significantly the increase in [Na+]i during anoxia (Fig. 7.3D). These findings, together with those presented above that the increase in [Na+]i during anoxia is not influenced significantly by CNQX, indicate that the effect of Gd to limit Na influx during anoxia is likely independent of its ability to block voltage-activated Ca2+ channels (e.g. Boland et al. 1991; Elinder & Arhem, 1994; Caldwell et al. 1998) or AMPA/kainate receptors (e.g. Huettner et al. 1998; Lei & MacDonald, 2001). To examine the potential role of reactive oxygen species (ROS) in activating Gd3+-sensitive Na+ influx through NSCCs, neuronal cultures were pretreated with the antioxidant, trolox (1 mM for 2 - 3 h prior to anoxia; see Papadopoulos et al. 1998; Vergun et al. 2001). Following pretreatment, the magnitude of the increase in [Na+]j during anoxia was reduced by -40% (Fig. 7.4) and 30 pM Gd3+ failed to exert an additional inhibitory effect on the increase in [Na+]i observed during anoxia (Fig. 7.4). Although these results support the possibility that NSCCs activated by ROS may contribute to the increases in [Na+]j observed during anoxia, neither 15-30 uM AACOCF3 nor 500 uM L-NAME had a significant effect on the increases in [Na+]i during anoxia (Fig. 7.45), suggesting a lack of involvement of cytosolic phospholipase A2 246 (PLA2) and nitric oxide synthase (NOS) in the generation of the reactive oxygen species involved (also see Vergun et al. 2001). I was unable to examine the potential effects of manganese (III) tetrakis (4-benzoic acid) porphyrin (MnTBAP, an O2" scavenger; see Patel et al. 1996) or mepacrine (a non-specific PLA2 inhibitor; see Chen et al. 1999), which were highly fluorescent during excitation at both 334 and 380 nm. 7.2.2. Increases in [Na+"|i after anoxia 7.2.2.1. Role of ionotropic glutamate receptor-operated channels, voltage-activated Na+ channels and Na+/K+/2C1" cotransport Analogous with their inability to influence Na+ influx during anoxia, the addition of 2 uM MK-801 and 20 uM CNQX failed to significantly influence the magnitude of the increase in [Na+]i observed after anoxia (Na+,K+-ATPase inhibited; Table 7.2). Similarly, the magnitude of the increase in [Na+]; observed following anoxia was not different in the presence or absence of 1 uM TTX (Table 7.2). One hundred micromolar TTX {n = 3 neuronal cultures at 9 DIV; not shown) and 250 uM lidocaine (Table 7.2) were also without significant effects, suggesting that Na+ influx via TTX-resistant persistent Na+ channels are not major contributors to the continued Na+ influx found to occur immediately after 5 min anoxia. Finally, in contrast to the ability of bumetanide to limit the rise in [Na+]i observed during anoxia in neurons maintained in culture for 11-14 DIV, 100 uM bumetanide did not reduce the magnitude of the increase in [Na+]i observed following anoxia in 6 - 10 or 11 - 14 DIV neuronal cultures (Table 7.2). 247 7.2.2.2. Role of plasmalemmal Na /Ca exchange The plasmalemmal Na /Ca exchanger, operating in forward mode, has been suggested to contribute, at least in part, to Ca2+ efflux following depolarization-induced increases in [Ca2+]i (e.g. Koch & Barish, 1994; Verdru et al. 1997). This mechanism may thereby contribute to the rise in [Na+]j seen after anoxia under conditions where the Na+,K+-ATPase is blocked. Conversely, the present experimental conditions (in which [Na+]i after anoxia was maintained at a relatively high level) could favor reverse-mode operation of the exchange mechanism and, thus, Na+ efflux (e.g. Czyz et al. 2002; also see Blaustein & Lederer, 1999). Applied immediately after anoxia under conditions where Na ,K -ATPase activity was inhibited (0 [K+]0), 50 pM bepridil significantly reduced the magnitude of the increase in [Na+]j, suggesting that forward-mode Na+/Ca2+ exchange contributes to Na+ influx immediately after anoxia (Table 7.2; Fig 7.5). In contrast, KB-R7943 (1 pM) enhanced the magnitude of the increase in [Na+]j observed after anoxia, suggesting that reverse-mode Na+/Ca2+ exchange may contribute to Na+ efflux at this time (Table 7.2; Fig 7.5). The removal of external Ca immediately after anoxia under [K ]0-free conditions, which would also inhibit reverse-mode Na /Ca exchange, appeared to similarly enhance Na+ influx at this time, although this effect did not reach statistical significance (n = 4; P = 0.35). In contrast, when applied at 10 pM (to inhibit both forward- and reverse- mode Na7Ca2+ exchange), KB-R7943 reduced the increase in [Na+]j observed after anoxia compared to those observed in age-matched sister cultures under control conditions (Table 7.2; Fig. 7.5). Together, these results are consistent with those of White & Reynolds (1995), who reported considerable variability in the contribution of forward-mode Na+/Ca2+ exchange to the recovery of [Ca2+]j following glutamate stimulation in cultured rat forebrain neurons (also see Sidky & Baimbridge, 1997; Verdru et al. 1997), and those of Yu & Choi (1997), who found that Na+/Ca2+ exchangers 248 in neocortical neurons can operate concurrently in forward and reverse directions. These issues are considered further in the Discussion. 7.2.2.3. Role of non-selective cation channels As detailed above, 30 uM Gd3+ significantly reduced the increase in [Na+]j observed during anoxia via a mechanism that appeared dependent on the production of ROS. Because ROS production is enhanced upon reoxygenation (see Lipton, 1999), Gd3+ was applied immediately after anoxia under K+0-free conditions. As illustrated in Fig. 7.6, 30 nM Gd3+ significantly reduced the magnitude of the increase in [Na+]i seen after anoxia in neurons 6-10 DIV. Similar effects were observed in neurons maintained in culture 11-14 DIV (Normalized A[Na+]j(after) = 0.73 ±0.11;n = 5;P< 0.05). Analogous to findings made during anoxia (see above), the effect of Gd3+ to reduce the magnitude of the increase in [Na+]i observed after anoxia was not affected by perfusion with 100 mM mannitol but was occluded by pretreatment with 1 mM trolox (Fig. 7.65). Although 500 uM L-NAME was without effect, in contrast to observations made during anoxia, 15-30 uM AACOCF3 significantly reduced the magnitude of the increase in [Na+]i observed after anoxia (Fig. 7.65), suggesting that ROS derived via the PLAi/arachidonic acid pathway may play a role in regulating Na+ influx in the immediate post-anoxic period, possibly by influencing the activity of a Gd -sensitive NSCC. 7.3. DISCUSSION In isolated rat hippocampal neurons, [Na+], increases during 5 min anoxia and, shortly following the return to normoxia, recovers to resting values. As detailed in Chapters 5 and 6, the increase in [Na+]i observed during anoxia reflects both reduced efflux, consequent upon an inhibition of 249 Na+,K+-ATPase activity, and ongoing/increased entry of Na+ ions. Similarly, the change in [Na+]i observed in the period immediately following anoxia reflects a balance between re-established Na+,K+-ATPase activity and continued Na+ influx. Although Na+/FT exchange activity (and, possibly, the activities of HCCV-dependent pH; regulating mechanisms) contributes to Na+ influx immediately following anoxia, it cannot account for all Na+ entry at this time and, in addition, does not appear to contribute to Na+ influx during anoxia. In the present Chapter, I considered other mechanisms that might potentially account for Na+ entry occurring during and/or immediately after anoxia. Given that there is no a priori reason that the mechanism(s) contributing to Na+ entry during and following anoxia are the same, the study provided novel insights into the similarities and differences of [Na+]j regulation between these two vulnerable periods (see Lipton, 1999 and Silver & Erecinska, 1990 and 1992, for similar comments 2"f" concerning the regulation of [Ca ]j during and following transient ischemia in vivo). Moreover, during the course of these studies, it was also found that the increases in [Na+]j observed during and following anoxia were related to the duration of time neurons were maintained in culture (within the examined range 6-14 DIV). 7.3.1. The role of ionotropic glutamate receptor-operated channels Although the application of glutamate receptor agonists under normoxic conditions evoked increases in [Na+]i (see Fig. 7.1; also see Rose, 2002; also see Lasser-Ross & Ross, 1992; Knopfel et al. 2000), the increases in [Na+]j observed during or following (Na+,K+-ATPase blocked) anoxia under the constant perfusion conditions of the present experiments were not dependent on the activation of ionotropic glutamate receptor-operated channels. Glutamate-dependent increases in [Na+], have been observed during oxygen-glucose deprivation (e.g. Miiller & Somjen, 2000a; LoPachin et al. 2001); however, this is not a consistent finding. For example, 250 Guatteo et al. (1998) and Pisani et al. (1998a) observed that a combination of ionotropic and metabotropic glutamate receptor antagonists failed to limit the increases in [Na+]j observed during periods of hypoxia in midbrain neurons and periods of oxygen-glucose deprivation in cortical neurons, respectively. In a similar manner, in a study employing isolated cerebellar granule cells, Chen et al. (1999) reported that ionotropic glutamate receptor-operated channels do not contribute to the increases in [Ca ]\ observed during periods of metabolic inhibition. Thus, the present experimental conditions have allowed me to characterize those components of the [Na+]i response to anoxia that are independent of glutamate receptor-operated channels and are intrinsic to the neuron itself. 7.3.2. The role of voltage-activated Na+ channels TTX-sensitive Na+ channels did not contribute significantly to the increase in [Na+]j observed during anoxia (cf prolonged hypoxia; Banasiak et al. 2004); TTX and lidocaine also failed to attenuate Na+ influx in the immediate post-anoxic period. Other studies, employing isolated neuronal and brain slice preparations, have also suggested voltage-activated Na+ channels are only minor contributors to anoxia-evoked increases in Na+ (e.g. Pisani et al. 1998a; Chen et al. 1999; Muller & Somjen, 2000b; but see Fung et al. 1999; Raley-Susman et al. 2001). Indeed, it has been reported that anoxia inhibits whole-cell Na+ currents in isolated hippocampal and neocortical neurons (Cummins et al. 1993; O'Reilly et al. 1997; Mironov & Richter, 1999; but see Hammarstrom & Gage, 2000 for data illustrating an hypoxia-induced activation of TTX-sensitive persistant sodium currents in hippocampal neurons). In light of the relatively small membrane depolarization observed during anoxia in the present studies, I may be underestimating the contribution of voltage-activated Na+ channels to the anoxia-evoked increases in [Na+]i; however, even in studies in which more profound hypoxia- or ischemia-251 induced membrane depolarizations were observed, TTX similarly had only minor influences on the extent of membrane depolarization (Calabresi et al. 1999b) and magnitude of Na+ entry (Muller & Somjen, 2000b). As a result, our findings support suggestions that the neuroprotective actions of Na+ channel blockers may be mediated at a presynaptic locus (e.g. Taylor et al. 1995; Gleitz et al. 1996; Strijbos et al. 1996; Probert et al. 1997; Kimura et al. 1998; Raley-Susman et al. 2001). These present results do not, however, rule out potential contribution(s) from TTX-and/or lidocaine-insensitive voltage-activated mechanisms to the production of the increases in [Na+]i observed during or following anoxia. Indeed, in preliminary experiments in SBFI-loaded hippocampal neurons voltage-clamped at -60 mV, 5 min anoxia evoked smaller increases in [Na+]i than those recorded simultaneously from unpatched SBFI-loaded neurons present on the same coverslip (data obtained from neurons 13 DIV; C. Sheldon and T. Kelly, unpublished observations). The mechanism(s) underlying this apparent reduction remain to be determined. 7.3.3. The role of Na+/Ca2+ exchange I found no evidence to suggest that Na /Ca exchange (forward- or reverse-mode) contributes to the changes in [Na+]i observed during anoxia in rat hippocampal neurons. In a similar manner, the 2+ "1" removal of Ca 0 did not influence the increase in [Na ], observed during metabolic inhibition in rat cerebellar granule cells (Chen et al. 1999) and bepridil failed to influence the increase in [Na+]j observed during periods of oxygen-glucose deprivation in rat cortical neurons (Pisani et al. 1998a). These findings may reflect the inhibitory effects of reductions in pHj and/or internal ATP levels on Na+/Ca2+ exchange activity (see Blaustein & Lederer, 1999). Driven by the increase in [Na+]i and membrane depolarization that occur during anoxia, reverse-mode Na+/Ca2+ exchange activity may contribute to Na+ efflux (and Ca2+ entry) in the 252 immediate post-anoxic period. Assuming a 3:1 stoichiometry, the reversal potential of Na+/Ca2+ exchange (ENcx) immediately upon the return to normoxia can be estimated by ENCx = 3 E Na+ - 2 ECa2+ (Equation 7.1) where ENa+ and ECa2+ are the equilibrium potentials for Na+ and Ca2+ and values of ENCx more negative than membrane potential indicate reverse-mode Na+/Ca2+ exchange (i.e. Na+ efflux; Blaustein & Lederer, 1999). Employing [Na+]j measurements made in this thesis (see Fig. 7.2Q and [Ca2+]i measurements made in the same cultured postnatal rat hippocampal neurons under identical conditions (e.g. Diarra et al. 1999), [Na+]j and [Ca2+]j values at the end of 5 min anoxia were estimated as -30 mM and -500 nM, respectively. Using these values, ENCX immediately upon the return to normoxia under the present experimental conditions was calculated as —90 mV (also see Yu & Choi, 1997; Blaustein & Lederer, 1999; Czyz et al. 2002), a value significantly more negative than that pertaining prior to anoxia (ENcx was —50 mV assuming 10 mM [Na+]j, 80 nm [Ca2+]i, 148 mM [Na ]0 and 2 mM [Ca ]0) and more negative than measurements of membrane potential made under the present experimental conditions at the end of 5 min anoxia (—40 mV; see Section 7.2.1.2). These calculations suggest that Na+/Ca2+ exchange is likely functioning in reverse-mode to extrude Na+ ions immediately after anoxia. Consistent with this possibility, KB-R7943 (1 uM) enhanced the magnitude of the increase in \Na~\\ observed after anoxia (Na+,K+-ATPase inhibited). In hippocampal slices, KB-R7943 at higher concentrations (10 pM) non-selectively inhibits both forward- and reverse-mode Na /Ca exchange (Breder et al. 2000; see also Iwamoto et al. 1996). In the present study, KB-R7943 (10 pM) no longer increased but, rather, significantly reduced the increase in [Na+]j observed after anoxia and this inhibitory effect was similar to that evoked by the forward-mode Na+/Ca2+ 253 exchange inhibitor, bepridil (50 uM). A previous study (Yu & Choi, 1997) illustrated that forward-and reverse-mode Na+/Ca2+ exchange can operate concurrently in cortical neurons; in the same study, it was found that glutamate exposure, despite causing a negative-shift in ENCX (which would promote reverse-mode Na+/Ca2+ exchange) preferentially enhanced forward-mode Na+/Ca2+ exchange. Similarly, the results of the present study suggest that while reverse-mode Na+/Ca2+ exchange may be acting to extrude Na+ ions immediately after anoxia, forward-mode Na+/Ca2+ exchange may also be active at this time and contribute to concurrent Na+ influx. Nevertheless, it should be noted that bepridil (30 - 50 uM) has also been reported to inhibit reverse-mode Na+/Ca2+ exchange and, in this way, augments ischemia-induced depolarizations in striatal neurons (Calabresi et al. 1999a) and reduces anoxia-induced Ca influx in rat optic nerve preparations (Brown et al. 2001). The role of Na+/Ca2+ exchanger-mediated increases and decreases in [Na+]j (mediated by forward- and reverse-mode Na+/Ca2+ exchange, respectively) to the pathogenesis of ischemic cell death is unclear. Arguing against their importance are the observations that inhibition of forward-mode Na+/Ca2+ exchange (which would limit Na+ entry) typically aggravates neuronal death while inhibition of reverse-mode Na+/Ca2+ exchange (which would increase [Na+]i) typically limits neuronal death, in response to ischemia/reperfusion or hypoxia/hypoglycemia (e.g. Schroder et al. 1999; Breder et al. 2000; Matsuda et al. 2001; see also Hoyt et al. 1998 in which 10 uM KB-R7943 failed to have any effect on the viability of cultured cortical neurons following glutamate excitotoxicity). On the other hand, Calabresi and colleagues (1999a) have suggested that, in striatal neurons, reverse-mode Na /Ca exchange may play a protective role by reducing internal Na+ accumulation and promoting membrane repolarization following transient periods of oxygen-glucose deprivation. 254 In summary, in light of the relatively non-selective nature of the pharmacological agents available to assess forward- and reverse-mode Na+/Ca2+ exchange activity and the fact that the direction of Na+/Ca2+ exchange will be not only influenced by [Na+]j and [Ca2+]i but also membrane potential values attained during and following anoxia, the contribution of Na /Ca exchange to the + 2~t~ regulation of [Na ]j (and [Ca ]j) in rat hippocampal neurons during and immediately after anoxia under the present experimental conditions is difficult to determine precisely. Further investigation into anoxia-evoked changes in Na /Ca exchange activity and the effects of these changes to anoxia-evoked changes in [Na+]i and [Ca2+]i will necessitate simultaneous measurements of not only [Na+]j and [Ca2+]i but also membrane potential (e.g. Ginsburg et al. 2002 in cardiac myocytes). Notably, the results of the present experiments do not exclude the possibility that, as in cardiac myocytes, Na+/FT exchanger-induced increases in [Na+]j in rat hippocampal neurons in the post-anoxic period might contribute to the reversal of Na /Ca exchange activity at this time (see Chapters 5 and 6). 7.3.4. The role of Na+/K+/2C1" cotransport In the present study, the Na+/K+/2C1" cotransport inhibitor, bumetanide, failed to affect resting [Na+]i suggesting that, in agreement with Rose & Ransom (1997), Na+/K+/2C1" cotransport does not mediate Na+ entry in rat hippocampal neurons under resting conditions. In contrast, bumetanide reduced the rise in [Na+]j observed during anoxia in 11 - 14, but not 6-10, neurons suggesting that Na+/K+/2C1" cotransport contributes to Na+ influx during anoxia in neurons expected to express significant levels of functional transporters (i.e. neurons maintained in culture for at least 11 DIV; Sun & Murali, 1999). In a similar manner, Na+/K+/2C1" cotransport also contributes to the accumulation of internal Na+ ions in cortical astrocytes in response to periods of oxygen-glucose deprivation (Lenart et al. 2003), though not in cerebellar granule neurons during metabolic inhibition (Chen et al. 1999). These results, however, contrast with the rapid 255 inactivation of Na+/K+/2CT cotransport reported in neocortical slices upon the initiation of oxygen-glucose deprivation (Yamada et al. 2001). Interestingly, Na+/K+/2C1" cotransport contributes of Na+ entry during ischemia in cardiac myocytes; immediately upon reperfusion, Na+/K+/2C1" cotransport mediates Na+ efflux in this cell type (Anderson et al. 1996), a finding that may account for the ability of bumetanide to enhance (albeit not significantly) the rise in [Na+]i observed in the post-anoxic period in 11 - 14, but not 6 - 10, DIV neuronal cultures. Previous studies have reported that bumetanide limits neuronal death observed in response to periods of oxygen-glucose deprivation in vitro (Beck et al. 2003) and reduces infarct size in vivo when applied during, but not following, ischemic episodes (Yan et al. 2001 and 2003). The present study suggests that these neuroprotective actions may reflect, in part, a reduction in Na+ influx during these insults. 7.3.5. The role of a Gd3+-sensitive mechanism Gd3+, a non-selective blocker of NSCCs, attenuated the increases in [Na+]j observed during and following anoxia. From the other results presented in this study, the inhibitory effects of Gd3+ on the anoxia-evoked changes in [Na+]i are unlikely to reflect the ability of Gd3+ to inhibit AMPA/kainate receptors or voltage-activated Ca2+ channels (Boland et al. 1991; Elinder & Arhem, 1994; Huettner et al. 1998; Lei & MacDonald, 2001). Moreover, hyperosmolar conditions failed to reduce anoxia-evoked increases in [Na+]i and did not occlude the effect of Gd3+ to limit the increases in [Na+]j during and following anoxia, suggesting that mechanogated NSCCs are not major contributors to the increases in [Na+]j seen under the present experimental conditions. In contrast, pretreatment of neuronal cultures with trolox occluded the effect of Gd3+ to limit anoxia-evoked increases in [Na+]i, suggesting that the inhibitory effects of Gd3+ on Na+ influx during and following anoxia may be dependent on the production of reactive oxygen 256 species. Although Gd3+ reduces the increase in [Ca2+]j observed during 5 min anoxia in 7 - 9 DIV cultured postnatal rat hippocampal neurons under conditions identical to those used in the present experiments (n = 7; A. Diarra & J. Church, unpublished observations), the pathway involved in the production of ROS in the present experiments appears to differ from that involved in the activation of the recently described Gd3+-sensitive NSCC that contributes to Ca2+ influx during prolonged (>30 min) oxygen-glucose deprivation in cultured mouse cortical neurons (Aarts et al. 2003) in that it does not appear to depend on free radical production via the NOS/nitric oxide pathway. Although further investigation is required to identify the free radical species interacting with the Gd3+-sensitive Na+ influx pathway described in the present study, PLA2 may be involved in the immediate post-anoxic period. PLA2 activity, acting through a cascade of events that may involve the production of reactive oxygen species, has been shown to activate NSCCs in response to periods of metabolic inhibition in cerebellar granule cells (Chen et al. 1999). Interestingly, PLA2 activity is enhanced with an increase in pHi (such as that which occurs upon the return to normoxia; e.g. Harrison et al. 1991; Stella et al. 1995; Phillis & O' Regan, 2004) and is increased in the hippocampus immediately following oxygen-glucose deprivation (Arai et al. 2001), where it has been shown to maintain post-ischemic membrane depolarization (Tanaka et al. 2003) and contribute to neuronal death (Arai et al. 2001). The results of the present study suggest that the established effects of Gd3+ to reduce neuronal death following periods of oxygen-glucose deprivation in vitro (Aarts et al. 2003) and to reduce brain edema following traumatic brain injury in vivo (Vaz et al. 1998) may reflect the ability of Gd3+ to limit not only Ca2+ but also Na+ influx both during and immediately after these insults. 257 7.3.6. Age-dependence of the increases in fNa*"], observed during and following anoxia The results presented in Chapter 5 illustrated that the increases in [Na+]j observed during and after anoxia displayed a clear dependence on length of time hippocampal neurons were maintained in culture. The larger magnitudes of the increases in [Na+]i observed during and following anoxia in more phenotypically mature neuronal cultures may contribute to the enhanced vulnerability of these neurons to transient anoxic insults, an observation which has previously been ascribed to differences in [Ca2+]i entry and/or the activities of 'Ca2+-dependent lethal processes' (e.g. Rothman, 1983; Di Lorteo & Balestrino, 1997; Sattler et al. 1998; Keelan etal. 1999). In both 6-10 and 11-14 DIV neurons, creatine incubation limited the increase in [Na+]j during anoxia. In addition, a Gd3+-sensitive pathway and Na+/K+/2C1" cotransport contributed to the increase in 6 - 10 and 11-14 DIV neurons, respectively. That Na+/K+/2C1" cotransport fails to contribute to the increase in [Na+]j observed during anoxia in 6 - 10 DIV neurons is consistent with the reported in vivo and in vitro developmental regulation of Na+/K+/2C1" cotransport activity (Plotkin et al. 1997; Sun & Murali, 1999). Immediately after anoxia, continued Na+ influx was mediated, in part, by Na+/H+ exchange activity (Chapters 5 and 6); in addition, a Gd3+-sensitive pathway and Na+/Ca2+ exchange activity were also found to contribute to Na+ influx at this time. Although the potential contribution of Na+/Ca2+ exchange activity to Na+ influx after anoxia in 11 - 14 DIV neurons was not examined, given the developmental upregulation of Na+/Ca2+ exchanger expression and activity (Sakaue et al. 2000; Gibney et al. 2002), anoxia-evoked changes in Na+/Ca2+ exchange activity are likely to play a role in 11 - 14 DIV neurons as well. Indeed, the larger increases in [Na+]i observed following anoxia in 11 - 14, compared to 6 -10, DIV neurons may reflect the developmental regulation of the expression and/or activities of 258 Na+/FT exchangers (e.g. Bevensee et al. 1996; Douglas et al. 2001) and/or Na+/Ca2+ exchangers (e.g. Sakaue et al. 2000; Gibney et al. 2002). 7.3.7. Synthesis of Chapters 5. 6 and 7 Cultured postnatal rat hippocampal neurons respond to 5 min periods of anoxia with an increase in [Na+]i of-15 - 40 mM. The increase in [Na+]j observed during anoxia is, in part, dependent on external Na+ influx through a putative Gd3+-sensitive NSCC in 6 - 10 DIV neurons and Na+/K+/2C1" cotransport in 11 - 14 DIV neurons. In addition, in both 6 - 10 and 11 - 14 DIV cells, reduced Na+,K+-ATPase activity, consequent upon declining internal ATP levels, also contributes to the internal Na+ accumulation during anoxia. It is of note that the activation of NSCCs and Na+/K+/2C1" cotransport as well as Na+,K+-ATPase inhibition, can initiate and promote cell swelling (e.g. Chen & Simard, 2001; Xiao et al. 2002; Beck et al. 2003) and, in this way, their contributions to the increase [Na+]i during anoxia may mediate, at least in part, the acute neurotoxic effects of anoxia. Upon the return to normoxia, Na+,K+-ATPase activity mediates the recovery of [Na+]i in the face of continued Na+ entry. In addition a putative Gd3+-sensitive NSCC, Na+/Ca2+ exchange and NaVFT exchange also contribute to the increase in [Na+]i observed after anoxia. In summary, the present study represents one of the first detailed descriptions of the changes in [Na+]i observed in isolated mammalian central neurons during and after transient periods of anoxia. A number of mechanisms that contribute to the observed anoxia-evoked changes in [Na+]i have also been identified and, as a result, these mechanisms may be important in regulation of anoxic/ischemic cell death. 259 Table 7.1: Potential mechanisms contributing to the increase in [Na+]j observed during anoxia Normalized A[Na+ ]i(during) Treatment 6-10 DIV 11-14 DIV 2uMMK-801 1.14 ±0.14 (8) 0.98 ±0.12 (4) 20uMCNQX 1.01 ±0.03 (5) 1.17 ±0.31 (3) 2pMMK-801 +20pMCNQX 0.98 ±0.18 (4) 0.96 ±0.26 (7) luMTTX 1.04 ±0.11 (4) 0.91 ±0.17(5) 50 uM bepridil 0.94 ±0.17 (7) n.d. 1 pM KB-R7943 1.05 ± 0.28 (5) n.d. 10 pM KB-R7943 1.02 ± 0.17 (6) ad. 25 pM CGP-37157 1.02 ±0.28 (4) n.d. 0 Ca2+0 0.92 ± 0.16 (5) 1.04 ± 0.20 (4) 50- 100 uM bumetanide 1.12 ±0.19 (5) 0.61 ± 0.05 (9)* To generate Normalized A[Na+ ]i(during) values, measurements of A[Na+ ]i(during) under a given experimental test condition were normalized to [Na+]i measurements made in age-matched sister cultures under control conditions. Statistical comparisons were performed by comparing absolute A[Na+ ]i(during) values (i.e. not normalized) made under a given experimental test condition to measurements made in age-matched sister cultures under control conditions. * indicates statistical significance (P < 0.05) compared to measurements made in age-matched sister neurons in the absence of treatment. Numbers in brackets denote number of neuronal populations (i.e. coverslips) from which the data were obtained. DIV, days in vitro; n.d., not determined. 260 Table 7.2: Potential mechanisms contributing to the increase in [Na+]j observed after anoxia under 0 [K+]0 conditions Normalized A[Na+ Rafter) Treatment 6-10 DIV 11-14 DIV 2 uM MK-801 + 20 uM CNQX 1.25 ±0.13 (3) 1.06 ± 0.25 (2) 1 uM TTX 0.98 ± 0.20 (5) 0.92 ± 0.02 (2) 250 uM lidocaine n.d. 0.89 + 0.29 (3) 100 uM bumetanide1 0.91 ± 0.27 (3) 1.21 ± 0.25 (4) 50 uM bepridil 0.43 ± 0.12 (6)* n.d. lpMKB-R7943 1.56 + 0.25(3)* n.d. 10pMKB-R7943 0.67 ± 0.09 (5)* n.d. To generate Normalized A[Na+ ]i(afier) values, measurements of A[Na+ ]i(after) under a given experimental test condition were normalized to [Na+]j measurements made in age-matched sister cultures under control conditions. Statistical comparisons were performed by comparing absolute A[Na+ ] i(after) values (i.e. not normalized) made under a given experimental test condition to measurements made in age-matched sister cultures under control conditions. * indicates statistical significance (P < 0.05) compared to measurements made in sister age-matched neurons in the absence of treatment. Numbers in brackets denote the number of neuronal populations (i.e. coverslips) from which data were obtained. 'Experiments examining the effect of bumetanide on the increase in [Na+]i observed following anoxia were performed using 500 pM ouabain rather than [K+]-free medium to inhibit Na+/K+-ATPase activity following anoxia. DIV, days in vitro; n.d., not determined. 261 Fig. 7.1. Ionotropic glutamate receptor-operated channels do not contribute to the increase in [Na+]i observed during anoxia under the present experimental conditions. A, the rise in [Na+]j evoked by anoxia under control conditions (filled circles) was not significantly affected by the presence of 2 uM MK-801 and 20 uM CNQX (open circles). Inset, in a different neuronal culture, MK-801 (2 uM) and CNQX (20 uM) abolished the increase in [Na+]i evoked by 30 s applications of 20 uM NMDA and 20 uM AMPA in the presence of 2 uM glycine (denoted by short bars above the record); the record is representative of