UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of anoxia-induced neuronal death in hippocampal neurons Fernandes, Herman Brian 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-0376.pdf [ 4.96MB ]
Metadata
JSON: 831-1.0089928.json
JSON-LD: 831-1.0089928-ld.json
RDF/XML (Pretty): 831-1.0089928-rdf.xml
RDF/JSON: 831-1.0089928-rdf.json
Turtle: 831-1.0089928-turtle.txt
N-Triples: 831-1.0089928-rdf-ntriples.txt
Original Record: 831-1.0089928-source.json
Full Text
831-1.0089928-fulltext.txt
Citation
831-1.0089928.ris

Full Text

C H A R A C T E R I Z A T I O N OF ANOXIA-INDUCED N E U R O N A L D E A T H IN H IPPOCAMPAL NEURONS by H E R M A N B R I A N F E R N A N D E S B.Sc. (Pharmacology), McMaster University, 1997 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF S C I E N C E in T H E F A C U L T Y OF G R A D U A T E STUDIES (Neuroscience Graduate Program) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y ^ B R l t i & H C O L U M B I A July 2001 © H e r m a n Brian Fernandes, 2001 UBC Special Collections - Thesis Authorisation Form Page 1 of 1 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t the U n i v e r s i t y o f B r i t i s h Columbia, I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e head o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f The U n i v e r s i t y o f B r i t i s h Columbia Vancouver, Canada http://www.library.ubc.ca/spcoll/thesauth.html 8/22/01 11 A B S T R A C T Cerebral ischemia occurs when the blood supply to the brain, in whole or part, is reduced, resulting in a decrease in the supply of oxygen and glucose to the brain. This type of insult typically results in metabolic dysfunction and widespread neuronal death in affected areas of the brain. A number of in vitro models have been previously described which utilize variations in oxygen and glucose supply, or metabolic inhibition, to simulate ischemia or anoxia. In the present study, a new model of anoxia-induced neuronal death has been developed and characterized. In this model, postnatal hippocampal neuronal cultures are exposed to 5 minutes of anoxia produced by the addition of 2 m M sodium dithionite, an oxygen scavenger. The viability of neuronal cultures was determined using trypan blue exclusion in order to assess the potential of various maneuvers to provide neuroprotection. Using this model, we found that 5 min of anoxia induced by dithionite produces severe neurotoxicity in postnatal hippocampal neurons. This neuronal death is produced by the activation of N M D A and A M P A receptors by glutamate, and can be attenuated by the presence of the non-competitive N M D A receptor antagonist MK-801 during anoxia, or the competitive N M D A and A M P A receptor antagonists D-APV and C N Q X , respectively, when present post-insult. We observed two phases of neuronal death: 1) an acute phase which occurs within 4h and morphologically resembles necrosis, and 2) a delayed phase occurring over a period of 24h which is unmasked by perfusing the cultures post-anoxia for 5 minutes in normoxic buffer. Substitution of external N a + and Cl" with impermeant ions can mitigate a portion of acute neuronal death, however removal of external C a 2 + increases neuronal susceptibility to anoxia. Free radical-mediated damage does not appear to play a role in the acute phase of cell death. Induction of neuronal death by this method is rapid, reliable and consistently produces at least 60% neuronal death by 4 hours post-insult. This model has the advantages of short insult duration, rapid development of neuronal death, and the capacity to produce a delayed neuronal death which may occur via an alternate pathway. T A B L E O F C O N T E N T S A B S T R A C T T A B L E O F C O N T E N T S LIST O F F I G U R E S A C K N O W L E D G E M E N T S I N T R O D U C T I O N I In vitro Models of Anoxia and Ischemia 1) Anoxia/Hypoxia 2) Oxygen-Glucose Deprivation 3) Chemical Anoxia II Apoptosis and Necrosis in Anoxia/Ischemia 4 III Possible Mediators of Anoxia/Ischemia-Induced Neuronal Death 6 1) Glutamate 6 2) Ca 2 + j 12 3) Free Radicals 13 4) P L A 2 and Arachidonic Acid 14 5) Nitric Oxide 16 6) N a + a n d p H 17 IV Specific Aims of This Study 19 M A T E R I A L S A N D M E T H O D S 20 I Hippocampal Cultures 20 II Static Anoxia Model 21 III Perfusion Anoxia Model 23 IV Ca 2 +-Imaging with Fura-2 23 R E S U L T S 27 I Morphological Appearance of Hippocampal Neurons in Culture 27 II The Effect of Anoxia Duration on Neuronal Viability 29 III Toxicity of Sodium Sulfite in Hippocampal Cultures 31 Page ii iv vi viii 1 1 1 2 3 IV Glutamate or Oxygen Free Radical as Possible Toxic Agents of Anoxia-Induced Neuronal Death 33 V Pharmacology of Anoxia-Induced Neuronal Death 36 VI The Effect of Buffer Perfusion on Anoxia-Induced Neuronal Death 42 VII Involvement of External Ions in Anoxia-Induced Neuronal Death 45 VIII Neurotoxicity Produced by Reducing Agents 52 IX Summary of Neuronal Viabilities 54 X Correlation Between Anoxia-Induced Changes in [Ca 2 +]j and Neuronal Death 56 D I S C U S S I O N 60 Morphological Progression of Anoxia-Induced Neuronal Death 60 Glutamate Mediates Anoxia-Induced Neuronal Death 61 Perfusion Anoxia 65 External Ion Substitution - Low Cl" and Low N a + 66 C a 2 + and Anoxia-Induced Neuronal Death 68 Toxicity of N M D A Receptor Redox Modulators 71 Comparison of Dithionite-Induced Model of Anoxia to Other Current In vitro Models 72 Conclusions 78 Limitations of This Study 79 Future Directions 80 A P P E N D I X I Composition of Buffers and Media Used 83 R E F E R E N C E S 84 L I S T O F F I G U R E S Figure 1. Morphology of Hippocampal Neurons In vitro. Figure 2. Varying the Duration of Anoxia Affects the Degree of Neuronal Death Produced Figure 3. Sodium Sulfite is not Acutely Toxic in Hippocampal Neurons in vitro Figure 4. G P T or Pyruvate, but not P B N , Protects Neurons from Anoxia-Induced Neuronal Death When Present During and After Anoxia Figure 5. MK-801 During Anoxia Prevents Acute Anoxia-Induced Neuronal Death Figure 6. Post-Anoxic C N Q X or A P V Prevents Acute Anoxia-Induced Neuronal Death Figure 7. T T X (1 uM) Slightly Improves Neuronal Survival When Present During & After Anoxia Figure 8. MK-801, C N Q X , A P V , T T X , or P B N Do Not Produce Acute Neuronal Death Figure 9. Protocols Used in Perfusion Anoxia Experiments Figure 10. Effect of Flow Environment on Anoxia-Induced Cell Death 4h or 24h Post-Anoxia Figure 11. Post-Anoxic Recovery in Buffer Instead of Conditioned Media Does Not Increase Acute Neuronal Death Figure 12. Effect of External Ion Substitution on Acute Anoxia-Induced Neuronal Death Figure 13. Incubation in Ca 2 +-free Buffer Instead of Media Does Not Increase Acute Neuronal Death Figure 14. Buffer Replacement, not Removal of External C a 2 + for 10 or 30 Minutes, Reduces Acute Anoxia-Induced Neuronal Death Figure 15. 2 m M D T T Has an Acute Toxic Effect on Hippocampal Neurons Which Can Be Blocked by MK-801 Figure 16. Frequency Histogram of Neuronal Viabilities in Hippocampal Cultures Exposed to Normoxia, Glutamate or Anoxia for 5 min Figure 17. Typical Intracellular C a 2 + Responses in Fields of Neurons to 5 Minutes of Normoxia or Anoxia Under Perfusion Conditions Figure 18. Typical Intracellular C a 2 + Responses in Fields of Neurons to 5 Minutes of Normoxia or Anoxia Under Static Conditions V l l l A C K N O W L E D G E M E N T S First and foremost I would like to thank Dr. Ken Baimbridge for his input and advice over the last two years. His approach to supervision allowed me a wide degree of latitude to pursue my research as I saw fit, while directing me during the times I was headed off-course. I could not have finished this project in the desired time without his total support, and all the while he continually encouraged both my academic and extracurricular pursuits. I would also like to thank Dr. John Church, not only for his efforts as a member of my supervisory and examining committees, but particularly for his technical expertise and suggestions on the directions of my study. His input was crucial to the imaging studies performed and his knowledge of relevant literature proved invaluable. I would also like to thank Stella Atmadja for her expertise in cell culture techniques and general lab procedures - without her help, this project would not have been possible. I would also like to thank Drs. E d Moore and Steve Vincent for their roles as members of my supervisory and examining committees, and Dr. Charles Krieger for his contribution as a member of the examining committee. I wish to thank Gord Rintoul, Claudia Krebs, and Claire Sheldon for their assistance at various times during this project; and lastly, the graduate students and staff in the Department of Physiology for a most enjoyable two years. 1 I N T R O D U C T I O N There are a number of different experimental models used to study cerebral ischemia or anoxia, ranging from in vivo animal preparations to acutely dissociated cells and synaptosomes. As the focus of this study is on in vitro models of anoxia/ischemia, this thesis will concentrate on events observed in vitro. I. In vitro Models of Anoxia and Ischemia 1) Anoxia/Hypoxia The simplest way to simulate anoxia in vitro is to expose cells or cell cultures to an oxygen-free environment for a prolonged duration (on the order of hours), while maintaining glucose in the extracellular solution. The exposure buffer is usually pre-gassed with an anaerobic gas mixture to remove dissolved oxygen prior to hypoxic exposure. This method has been used successfully to produce injury in cultured cerebellar granule cells (Dessi et al, 1992); (Dessi et al., 1993a), cortical cells (Rogers and Hunter, 1997), murine neocortical cells (Goldberg and Choi, 1993; Goldberg et al, 1987; Goldberg et al, 1988), and has been used to study hypoxia-induced intracellular ion changes in brain slices (Krai et al, 1993). Dithionite as an Oxygen Scavenger Alternatively, oxygen can be chemically removed from the extracellular solution to mimic anoxic conditions. The oxygen scavenger sodium dithionite has been used to induce anoxic conditions in acutely dissociated hippocampal neurons (Cummins etal, 1991; Friedman and Haddad, 1993; Friedman & Haddad, 1994a; Friedman & Haddad, 1994b; Smith et al, 1998) and cultured cortical and hippocampal neurons (Chidekel et al, 1997; Friedman & Haddad, 1994b; Diarra et al, 1999). Dithionite dissociates in an equilibrium fashion to form the radical 2 SOY, which is thought to be the reducing species in solution (Lambeth & Palmer, 1973). Dithionite consumes molecular O2 in solution via chemical reduction, thereby reducing PO2 while simultaneously generating the superoxide radical in oxygenated solutions (Archer et ai, 1995). Hence, it is recommended that dithionite is used under anaerobic conditions (Lambeth & Palmer, 1973). This limitation of dithionite has been noted in studies of the responses of alveolar lung tissue (Archer et al., 1995) or type I cells of the carotid body (Carpenter et al, 2000) to hypoxia produced by dithionite. Additionally, Carpenter et al. (2000) reported that type I cells in the rat carotid body gated an artefactual C a 2 + current in response to dithionite, and dithionite has been reported to affect the holding currents of current-clamped hippocampal neurons (Gebhardt & Heinemann, 1999). Despite these limitations, the effects obtained by the addition of dithionite have been reported to be due solely to O2 deprivation, in studies using acutely dissociated hippocampal neurons (Friedman & Haddad, 1993; Friedman & Haddad, 1994a; Chidekel et al., 1997). Anoxic insults have been reported to produce glutamate-mediated neuronal damage in both dissociated cell cultures (Goldberg etal., 1987; Madden etal., 1990) and brain (hippocampal) slices (Clark & Rothman, 1987), although evidence that glutamate receptor antagonists are not effective at reducing neuronal damage following oxygen deprivation alone (Friedman and Haddad, 1993; Haddad & Jiang, 1993) exists, suggesting that the nature and duration of insult determine whether both glutamate- and non-glutamate-mediated mechanisms are activated, leading to neuronal damage. 2) Oxygen-Glucose Deprivation Oxygen-glucose deprivation more closely approximates an ischemic, rather than anoxic, insult since an immediate limitation on usable energy is imposed by the absence of glucose. Although the duration of insult required to produce neurotoxicity is not as long as in hypoxic models, cultured neurons need to be incubated for at least 30-40 minutes in a glucose-free buffer in an oxygen-free atmosphere to produce any significant neuronal death during the subsequent 24h period (Goldberg and Choi, 1993; Grabb and Choi, 1999; Gwag etal, 1995) unless argon is used to reduce pC>2, in which case a 20 min insult will reduce neuronal viability to -50 % (Kusumoto etal, 1996). 3) Chemical Anoxia Other common techniques to simulate anoxia/ischemia, referred to as 'chemical anoxia', use chemical agents to interfere with A T P generation. This is normally performed under normoxic conditions. Agents used in this manner include 3-nitropropionic acid (NPA) (Pang and Geddes, 1997; Weih etal, 1999), cyanide (Dubinsky and Rothman, 1991; Patel etal, 1993), azide (Varming et al, 1996; Jorgensen et al, 1999; Regan and Guo, 1999), and iodoacetate (Uto et al, 1995; Dux et al, 1996). N P A , cyanide and azide all inhibit oxidative phosphorylation and A T P production; N P A inhibits succinate dehydrogenase (Riepe etal, 1996), whereas cyanide and azide both block the transfer of electrons between the cytochrome oxidase complex and O2. Iodoacetate inactivates glyceraldehyde dehydrogenase, thereby blocking the glycolytic pathway and dose-dependently reducing A T P production. There are several drawbacks to the use of these existing models. Some (NPA, C N ) require prolonged exposure times to elicit significant neuronal death, and C N has its own technical issues that present an obstacle to its use (Varming et al, 1996). Azide has several side effects that may exacerbate post-anoxic/ischemic injury and confound results (Smith et al, 1991; Varming et al, 1996). In addition, iodoacetate inactivates thiol-containing enzymes that help scavenge free radicals (Uto et al, 1995), further compromising cellular function. 4 II. Apoptosis and Necrosis in Anoxia/Ischemia Neuronal death produced by anoxic or ischemic insults is generally classified into two categories, apoptosis or necrosis. It has been debated as to whether or not these are two distinct processes or simply a continuum. Although there can be certain similarities between the two, most authors agree that apoptosis, or programmed cell death, is highly regulated, whereas necrosis is a result of accidental cellular injury and proceeds in an unregulated manner (Darzynkiewicz et al, 1997). Kerr et al. (1972) described a number of characteristic features of apoptosis, including a progressive cell shrinking, chromatin condensation and the production of membrane-bound cell fragments called apoptotic bodies. Other hallmarks of apoptosis include a controlled internucleosomal cleavage of D N A , resulting in D N A fragments of characteristic sizes, which occurs relatively early in the death process ( D N A laddering), and the maintenance of plasma membrane integrity during early stages (Choi, 1996). Several groups have also reported that apoptosis in neuronal cultures is sensitive to inhibitors of protein synthesis in vitro (Gwag et al., 1995; Nardi et al, 1997; Gwag et al., 1999). In vivo, apoptosis appears to be a characteristic of the delayed neuronal death that follows transient global ischemia. The use of protein synthesis inhibitors reduced the neuronal death (Goto etal., 1990; Shigeno et al., 1990) that was also characterized by D N A laddering and morphological signs of apoptosis (MacManus et al., 1993; Nitatori etal, 1995). A n example of the specific regulation of apoptotis can be found in the activation of caspases required to mediate the later stages of apoptosis. Caspases are proteases which only cleave at cysteine residues within a defined sequence of 4 amino acids at the N-terminus of the target protein (Thornberry and Lazebnik, 1998). They can only be activated through specific pro-apoptotic pathways, requiring the presence of specific cofactors which, in the living cell, 5 may be sequestered in different compartments (Thornberry & Lazebnik, 1998). Inhibition of caspases has been demonstrated to reduce apoptotic cell death following ischemic damage in vivo (Hara et al, 1997; Chen et al., 1998). Hence there are several key features of apoptosis which may be useful for its identification in vivo and in vitro, with certain caveats (see below). Neuronal cell death resulting from excitotoxicity (see below) has been suggested to result primarily from necrosis. The morphological progression of cell death in models of excitotoxicity, with marked cell swelling and fragmentation of cell processes prior to death (Dessi et al., 1993b; Gwag et al, 1995; Matthews et al, 2000), is similar to that described in in vitro models of anoxia/ischemia. For example, in cell cultures exposed to different methods of in vitro anoxia/ischemia, a mode of cell death was reported which lacked the characteristic features of apoptosis (Uto etal, 1995; Kusumoto etal, 1996; Pang and Geddes, 1997). There are a number of features which are common to both apoptosis and necrosis. Late events in both cases resemble each other, for example mitochondrial dysfunction, free radical production, and a loss of A T P (Choi, 1996). For this reason, the time at which cell death is assessed may be important in determining what processes are occurring. Internucleosomal D N A cleavage can also occur in necrotic cell death, and so can cause confusion if used as the sole criteria for determination of apoptosis (Collins et al, 1992). Determination of the mechanism of cell death can become more confusing considering that necrosis and apoptosis can be produced by the same insult in the same neuronal cell culture (Gwag et al, 1995). Hence, for the determination of apoptosis a multiple criteria assessment is ideal, using both descriptive (e.g. morphological appearance) and interventional (e.g. caspase inhibition) methods if possible (Choi, 1996). 6 III. Possible Mediators of Anoxia/Ischemia-Induced Neuronal Death 1) Glutamate Experimental work directed at understanding ischemic brain injury has, over the years, highlighted many events that may mediate neuronal damage. Recent interest has centered on the role of excitatory amino acids (EAAs) in the pathogenesis of ischemic neuronal injury. This well-known process, often referred to as 'excitotoxicity' (Olney et al, 1971; Choi, 1988; Lee et al, 1999), involves the extracellular overflow of endogenous E A A s and the co-ordinate activation of multiple E A A receptor subtypes (in particular, the N-methyl-D-aspartate ( N M D A ) receptor subtype). Neuronal death then ensues, either immediately via an early osmolysis of cells that is dependent upon external Cl" (Rothman, 1985; Olney et al., 1986; Choi, 1987), or after a delay. In the latter case the accumulation of intracellular free C a 2 + ([Ca2+]i) appears to be the final common pathway (Choi, 1987; Rothman etal, 1987; Randall and Thayer, 1992). The release of glutamate during brain anoxia or ischemia results in extracellular concentrations more than sufficient to trigger cell death (Hagberg et al, 1985; Benveniste et al, 1989). However, the source of glutamate has been controversial. Possibilities include the release from synaptic vesicles in nerve terminals, a non-synaptic release from anion channels in astrocytes activated by cell swelling, a Ca2 +-dependent release from astrocytes, or the reversed operation of glutamate transporters. Using specific approaches to systematically block each of these processes in turn, Rossi et al. (2000) concluded that the major source of glutamate was via reversed operation of the high concentration of glutamate transporters found in glutamatergic presynaptic terminals. The release of glutamate from the terminals was also speculated to result from the loss of A T P during a period of anoxia/ischemia leading to the loss of glutamate from synaptic vesicles. The nature of the receptors with which glutamate interacts has also received much attention. Recent work suggests that both synaptic and extra-synaptic N M D A receptors contribute to cell death when exogenous glutamate is applied to cultured neurons, whereas synaptically located N M D A receptors play a greater role when the source of glutamate is predominantly from synaptic release, as occurs during oxygen/glucose deprivation (Sattler et al, 2000). Despite the accumulated evidence in support for a role for N M D A receptors in anoxia/ischemia induced neuron death, clinical trials with a variety of agents aimed at blocking glutamate receptor immediately following a stroke have proved disappointing (Lee etal, 1999). Glutamate Receptors Involved in Anoxic/Ischemic Cell Death There are three families of E A A receptors: N M D A receptors, AMPA/kainate receptors, and metabotropic receptors (Collingridge & Lester 1989; Monaghan et al. 1989). Within each family of receptors exists a number of subtypes, which allows them to play a variety of physiological roles in the nervous system. Each receptor family has specific properties or defining characteristics with respect to subunit composition, pharmacology, electrophysiological properties, and signal transduction / modulatory mechanisms. a) NMDA Receptors The N M D A receptor is thought to be a hetero-oligomer, similar to other ligand gated ion channels. N M D A receptor subunits are encoded by several genes, with 8 splice variants known for the NR1 subunit (Hollmann et al, 1993). There are 4 different isoforms of the NR2 subunit, named N R 2 - A , -B , - C , and -D in the rat (Monyer et al, 1992; Kutsuwada et al, 1992; Ishii et al, 1993), with a possible splice variant existing for N R 2 D (Ishii et al, 1993). The ion channel complex in vivo is composed of at least 1 NR1 subunit which is associated with at least 1 subtype of NR2 subunit, although the exact subunit stoichiometry and number of subunits per receptor is unknown (Dingledine et al, 1999). The NR1 subunit can form functional homomeric channels in Xenopus oocytes (Durand et al, 1992; Sugihara et al, 1992) but not in mammalian cells 8 (Boeckmann and Aizenman, 1994; Ishmael etal, 1996). Receptors containing different NR1 splice variants display subtle differences in pharmacological properties (Zukin & Bennet, 1995). The NR1 subunit contains not only the glutamate binding site (Moriyoshi et al, 1991) but also the binding site for the co-agonist glycine (Chazot et al, 1992). Co-expression of NR1 with at least one subtype of NR2 in both Xenopus oocytes and mammalian cells yields channels with greatly increased conductance and functional responsiveness (Kutsuwada et al, 1992). The NR2 subunit composition affects both the biophysical and pharmacological characteristics of N M D A receptors (Buller et al, 1994; Monyer et al, 1994), and thus provides for greater physiological diversity and function. The N M D A receptor is a ligand gated cation channel that permits entry of N a + and C a 2 + as well as efflux of K + . At resting membrane potentials, M g 2 + blocks the channel pore, but this block is relieved upon membrane depolarization, via other mechanisms (Nowak et al, 1984), including glutamate activation of AMPA/kainate receptors (see below). A number of different factors can modulate the activity of the N M D A receptor. Decreases in extracellular p H (as would occur during ischemia in vivo) inhibit functioning of the receptor by decreasing open channel frequency (Traynelis and Cull-Candy, 1990). MacDonald et al (1989) proposed that phosphorylation of various sites on the N M D A receptor had to occur in order to maintain full channel activity. For example, phosphorylation of intracellular serine/threonine or tyrosine residues on NR2 subunits by protein kinase C leads to an increase in N M D A receptor responses, possibly by reducing the sensitivity of the receptor to inhibition of receptor function by M g 2 + (Chen and Huang, 1992). N M D A receptors can be phosphorylated on serine/threonine residues by protein kinases A & C (as well as Ca2 +/calmodulin-dependent kinase II), and on tyrosine by Src kinase (MacDonald et al, 1998). There are several mechanisms for N M D A receptor desensitization. First, N M D A receptor responses decrease in the continued presence of extracellular glutamate due to a 9 decrease in the affinity of the glycine binding site for the co-agonist glycine. This form of desensitization can be overcome by increasing the concentration of glycine in vitro (McBain and Mayer, 1994). A second mechanism is referred to as Ca2 +-dependent desensitization or inactivation, and can be initiated via the movement of calcium from extracellular to intracellular compartments via a number of different routes (MacDonald et al, 1998). C a 2 + entering the cell via routes other than the N M D A receptor can activate second messenger systems to reduce N M D A receptor activity. For example, synaptic activity can activate post-synaptic calcineurin (via an increase in [Ca 2 +]i) to dephosphorylate the N M D A receptor and reduce its activity (Raman et al, 1996). C a 2 + entering the cell via the N M D A receptor can result in actin depolymerization, uncoupling the N M D A receptor from intracellular scaffolding proteins and thereby reducing receptor activity via decreased open channel probability (Rosenmund and Westbrook, 1993a; Rosenmund and Westbrook, 1993b; Vyklicky, Jr., 1993). N M D A receptor desensitization can also occur via dephosphorylation of tyrosine residues, regulated by tyrosine phosphatase activity in a mechanism unrelated to C a 2 + (Wang et al, 1993). Redox Modulation of the NMDA Receptor The concentration of reducing equivalents in the brain is increased markedly under conditions of mild to moderate stroke (Tanaka et al, 1986). The N M D A receptor complex contains extracellular thiol groups on cysteine residues in the NR1 subunit (Sullivan et al, 1994), which can be modified by either reduction or oxidation, increasing or decreasing (respectively) the open channel frequency of the N M D A receptor (Aizenman et al, 1989; Tang & Aizenman, 1993). Several endogenous redox agents such as oxygen free radicals, oxidized glutathione, pyrroloquinoline quinone, and lipoic acid have been shown to act on the redox site(s) of the N M D A receptor (Choi and Lipton, 2000). The reducing agent dithiothreitol (DTT) has been shown to increase the open channel frequency of N M D A receptor-mediated responses 10 in both in situ and in recombinant systems upon agonist binding, whereas the oxidizing agent 5,5'-dithio-bis-nitrobenzoic acid (DTNB) has been shown to decrease channel open frequency and thus depress N M D A receptor function (Tang & Aizenman, 1993). Previous studies have shown that addition of the sulfhydryl reducing agent mercaptoethanol exacerbates M K - 8 0 1 -sensitive injury caused by chemical anoxia, as did reduced extracellular glutathione (Regan and Guo, 1999). These results suggest that altering the extracellular redox state of the N M D A receptor to the reduced form can enhance neuronal death due to N M D A receptor-mediated injury brought about by A T P depletion. b) AMPA/Kainate Receptors AMPA/kainate receptors are glutamatergic n o n - N M D A receptors which gate a cationic current. To date, 9 basic subunits have been identified, and they can exist in a number of isoforms as a result of alternative splicing (Dingledine et al., 1999). Here we will focus on the A M P A receptor. A M P A receptors are likely hetero-oligomeric, formed from G l u R l , GluR2, GluR3 and GluR4 subunits with 5 subunits per receptor-ion channel (Wenthold etal., 1992). Each GluRl -4 subunit can exist in either 'flip' or 'flop' splice variants, which can affect the biophysical properties (e.g. desensitization) of A M P A receptors (Sommer et al., 1990). Most A M P A receptors in the C N S contain 1 GluR2 subunit, which dictates the permeability of the receptor to calcium; A M P A receptors lacking the GluR2 subunit are Ca 2 +-permeable (Verdoorn et al., 1991). There are a number of studies using different techniques which indicate that calcium-permeable A M P A receptors lacking the GluR2 subunit are present in a number of C N S regions, including hippocampus (Doble, 1999). It was recently shown that following global ischemia, GluR2 m R N A is downregulated in the CA1 region of the gerbil hippocampus prior to neurodegeration, resulting in a functional increase in A M P A receptor-gated C a 2 + currents, 11 providing evidence of A M P A receptor involvement in delayed neuronal death following ischemia (Goiter et al, 1997). A M P A receptors can be activated by the agonists A M P A , kainate, quisqualic acid, and domoic acid, with glutamate acting as the endogenous ligand. Commonly used competitive antagonists include C N Q X and N B Q X (Honore et al, 1988). A M P A receptors likely mediate rapid excitatory synaptic transmission at the majority of glutamatergic synapses, and they are co-localized with N M D A receptors at many central synapses (Bekkers and Stevens, 1989). A M P A receptors are rapidly activated by the low affinity binding of two agonist molecules and the channel gates a cationic current that is carried predominantly by N a + , with some C a 2 + contribution depending on the presence or absence of GluR2. The gating kinetics are much faster than that of N M D A receptors, and A M P A receptors desensitize rapidly in the continued presence of glutamate, with quick inactivation upon removal of glutamate. These qualities facilitate the rapid depolarization of postsynaptic membranes via glutamate binding to A M P A receptors, which results in the relief of Mg 2 + -block of N M D A receptors, allowing the slower onset N M D A current to then contribute to the overall synaptic current (Dingledine et al, 1999). A M P A receptor function can be potentiated following phosphorylation by a number of different kinases, and antagonised by the binding of certain spider and wasp toxins or polyamines such as spermine (Fletcher and Lodge, 1996). Like N M D A receptors, A M P A receptors can also be inhibited by extracellular pH, although this occurs at approximately p H 6.0, limiting its physiological relevance (Dingledine et al, 1999). c) Metabotropic Glutamate Receptors Metabotropic glutamate receptors are g-protein-coupled receptors which are linked to a number of second messenger systems. Eight subtypes have been classified into three families on the basis of the second messenger activated or inhibited. Activation of group I metabotropic 12 glutamate receptors stimulates phospholipase C (PLC) whereas activation of group II or l U receptors produces an inhibition of adenylate cyclase (Bruno et al, 1998) and voltage-sensitive calcium channels (Nicoletti et al, 1996). These receptors are not thought to mediate fast excitatory synaptic transmission, although the extent of their physiological roles has not yet been elucidated. Generally speaking, activation of group II or III metabotropic glutamate receptors may be neuroprotective as this results in a decrease of synaptic excitability (Buisson and Choi, 1995; Nicoletti et al, 1996). However, activation of Group I metabotropic receptors may contribute to cell death by enhancing N M D A receptor-mediated toxicity via phosphorylation of the N M D A receptor, through a protein kinase C-dependent mechanism (Bruno et al, 1995). 2) Ca 2 + i Elevations in intracellular C a 2 + have been implicated in neuronal injury following anoxia/ischemia, and restoration of blood flow in vivo following transient ischemia has been shown to help reduce intracellular calcium load (Silver & Erecinska, 1992; Choi, 1995; Erecinska & Silver, 1996). However, the non-competitive N M D A antagonist, MK-801, was ineffective at accelerating the recovery of [Ca 2 +]i, and at attenuating a secondary intracellular calcium rise thought to lead to cell death (Silver & Erecinska, 1990; Silver & Erecinska, 1992). The subsequent effects of anoxia/ischemia-induced changes in C a 2 + leading to cell death have been widely investigated. Evidence has been obtained for the activation of a number of Ca 2 +-dependent enzymes including nitric oxide synthase (Dawson etal, 1991) and calpain (Siman et al, 1989). More recently emphasis has been placed on the role of mitochondria, and particularly the result of excessive Ca 2 +-uptake into mitochondria (Ankarcrona et al, 1995; Schinder et al, 1996; Vergun et al, 1999) leading to a depolarization of their inner membrane, a loss of A T P synthesis, the production of free radicals, and the release of apoptotic factors such as cytochrome C, procaspases and apoptosis-inducing factor (Shimizu et al, 1999; Susin et al, 13 1999). Indeed, blockade of mitochondrial Ca 2 +-uptake proved to be highly protective in blocking cell death in cultured rat forebrain neurons whereas an elevation in [Ca 2 +]i alone, even when enhanced by the inhibition of mitochondrial Ca 2 +-uptake, was in fact well tolerated by neurons (Stout et al, 1998). Evidence has also been provided suggesting that C a 2 + and nitric oxide act together to depolarize mitochondria (Keelan et al., 1999). 3) Free Radicals Oxygen free radicals are molecules that contain at least one orbital with a single unpaired electron, of which two common examples are superoxide (O2O and the hydroxyl radical (OH-). Peroxide (H2O2) is not a free radical per se, but spontaneously decomposes to form O H - , which is the most reactive of the two (Nakao and Brundin, 1998). Several in vivo studies, including models of ischemia, show that rates of free radical formation increase during in vivo ischemia (Phillis and Sen, 1993; Piantodosi & Zhang, 1996; Sakamoto et al, 1991), but particularly during the early reperfusion period once blood flow (and hence O2 supply) is restored (Siesjo etal, 1989; Oliver etal, 1990; Sakamoto etal, 1991; Phillis and Sen, 1993; Piantodosi & Zhang, 1996). There are several possible mechanisms responsible for the generation of free radicals under these conditions. Free radicals are normally generated by mitochondria at a rapid rate, but under normal conditions this is not toxic to cells (Turrens, 1997). However, under conditions in which the respiratory chain is interrupted (e.g. ischemia), the production of 0 2 " radicals increases (Turrens et al, 1991). It has also been shown that exposure of cultured neurons to N M D A can increase the generation of oxygen free radicals, mediated by an accumulation of C a 2 + i through N M D A receptors (Dugan et al, 1995). As an extension of this idea, under conditions of elevated C a 2 + i (2.5 uM) and A D P , O2" production was greatly increased in mitochondria isolated from C N S tissue (Dykens, 1994). These conditions partially reproduce the intracellular conditions following ischemia, and indicate that increased 14 free radical generation by mitochondria can occur as a consequence of ischemia or anoxia. (Gunasekar etal., 1996) provided additional evidence for this, as metabolic inhibition of granule cell cultures produced N M D A receptor activation and simultaneous generation of oxygen free radicals. Alternatively, increased oxygen free radical generation can occur as a consequence of calcium-dependent activation of enzymes such as phospholipase A 2 ( P L A 2 ) (Chan and Fishman, 1980; Turrens et al., 1991) or xanthine oxidase (Atlante et al., 1997). Anoxia/ischemia-induced increases in [Ca 2 + i ] are thought to increase nitric oxide levels (see Nitric Oxide below), which can lead to excessive free radical production (Dawson et al., 1993). Intracellular organelle damage by free radicals can be a major contributor to ischemic cell death. Peroxidation of membrane lipids results from chemical interactions between membrane lipid constituents and free radicals, compromising plasma membrane integrity and leading to cell death (Braughler and Hall, 1992; Dawson et al., 1993; Sakamoto et al., 1991; Phillis and Sen, 1993). These actions, when combined with the actions of calcium-dependent lipases (see below), may lead to irreversible cell damage. Nuclear D N A can also be a target of free radical-mediated cell damage, triggering repair processes which consume cellular energy reserves (Szabo et al., 1996). Measures to reduce free radical mediated cell damage such as treatment with free radical scavengers (Phillis & Sen, 1993; Newell et al., 1995; Roos & Ericsson, 1999), antioxidants (Yoshida et al., 1984; Uto et al., 1995), or overexpression of superoxide dismutase (Kinouchi et al., 1991) have proven neuroprotective. 4) P L A 2 and Arachidonic Acid Phospholipases A 2 (PLA 2 ) are a group of enzymes which catalyze the hydrolysis of free fatty acids from phospholipids of cell membranes. The products of P L A 2 activity, including arachidonic acid (AA) and lysophospholipids, play a role in signal transduction. There are 15 cytoplasmic as well as secretory forms of PLA2, and some are Ca2 +-dependent, being activated at submicromolar [Ca 2 +]j (Kudo etal., 1993). Glutamate activation of the N M D A receptor can produce a Ca2 +-dependent release of arachidonic acid from striatal neurons and cerebellar granule cells (Dumuis et al., 1988; Lazarewicz et al., 1988), as can the simultaneous activation of A M P A and metabotropic glutamate receptors in striatal neurons (Dumuis et al., 1990). Increased levels of free arachidonic acid have been shown to inhibit Na+-dependent uptake of glutamate in both astrocytes (Barbour et al, 1989) and synaptosomes (Chan et al, 1983), implicating a role for arachidonic acid in exacerbating excitotoxic ischemic injury associated with anoxia/ischemia. In a preparation of mitochondria isolated from rat hearts which had been subjected to 90 min anoxia followed by reoxygenation, an inhibition of the electron transport chain was seen that resulted in increased H2O2 production that prevented A T P generation (Turrens et al, 1991). This Ca 2 +-mediated effect was due to P L A 2 activation in the mitochondria, producing mitochondrial swelling and inhibition of the electron transport chain, thus promoting free radical generation through the production of H2O2 (Turrens et al, 1991). These effects were mimicked by the addition of either A A or high C a 2 + to mitochondria isolated from hearts not subjected to anoxia (Turrens et al, 1991). Hence, it was proposed that during anoxia the activation of mitochondrial P L A 2 by high C a 2 + could lead to the production of A A , leading to or exacerbating mitochondrial impairment and increasing the production of free radicals. Metabolic inhibition of cerebellar granule cell cultures for 15 min with CN" resulted in elevations in both [Ca 2 + ] i and [Na + ]i through a mechanism which could be blocked by the PLA2 inhibitor mepacrine, and replicated using the PLA2 activator melittin, A A , or lysophosphatidylcholine (Chen et al, 1999). After investigating several possible routes of N a + and C a 2 + entry, the authors concluded that these ionic influxes occurred via an A A / lysophosphatidylcholine-activated non-selective cation channel. This provides a possible 16 mechanism for further impairment of cell function by P L A 2 activation under conditions of energy depletion (as in anoxia/ischemia), producing large influxes of C a 2 + and N a + which could result in or exacerbate neuronal injury (Chen et al, 1999). 5) Nitric Oxide Nitric oxide (NO-) fulfills several roles in the body, notably as a mediator of macrophage toxicity (Hibbs et al, 1987) and that of endothelial-derived relaxing factor (Ignarro, 1990; Moncada et al, 1991). Additionally, NO- can act as a messenger in neuronal tissue (Snyder and Bredt, 1991). The synthesis of NO- is mediated by N O synthase (NOS), which has 3 isoforms corresponding to its effects above: inducible N O S (iNOS), endothelial N O S (eNOS), and neuronal N O S (nNOS). N O S uses L-arginine and 0 2 as substrates to produce NO- and its byproduct L-citrulline. Activation of nNOS has been shown to be Ca2 +/calmodulin-dependent, and N O - can subsequently act to stimulate cyclic G M P formation (Bredt and Snyder, 1990), or act itself to produce downstream effects such as activation of p21 (Yun et al, 1998). In addition to its physiological roles, N O has been implicated both in vivo and in vitro as a mediator of neurotoxicity. The N O S inhibitor nitroarginine showed greater efficacy at preventing neurotoxicity in mice after middle cerebral artery occlusion than MK-801 (Nowicki et al, 1991), and was also neuroprotective in a model of focal ischemia in rats (Trifiletti, 1992). Primary cortical cultures from transgenic mice lacking nNOS were found to be much more resistant to N M D A excitotoxicity than similarly treated cultures from wild-type mice, and inhibitors of N O S were found to be neuroprotective in wild-type cultures but not nNOS" cultures (Dawson et al, 1996). Sodium nitroprusside, a N O generator, has been shown to elicit neurotoxicity in a concentration-dependent manner, and inhibitors of N O S have been found to have the same rank order of potency in neuroprotection as they did in inhibiting N O S (Dawson et 17 al, 1993). Inhibition of calmodulin has been found to also prevent N M D A neurotoxicity, consistent with the Ca2 +-dependent activation of nNOS (Dawson et al, 1993). Sattler et al. (1999) demonstrated in cultured cortical neurons that inhibiting the expression of postsynaptic density-95 protein, an N M D A receptor scaffolding protein, selectively blocked Ca 2 +-activated N O production, independent of affecting N M D A receptor or nNOS expression or function. Lastly, nNOS has been shown to be activated in response to middle cerebral artery occlusion in mice, with subsequent production of peroxynitrite (Eliasson et al, 1999). Taken together, these observations suggest that excessive C a 2 + entering through the N M D A receptor during an ischemic episode or excitotoxic challenge can activate N O S , stimulating the production of N O , and produce subsequent neurotoxicity. The primary mechanism of NO-mediated damage appears not to be through N O itself, but through the products of its reactions with free radicals generated during an ischemic insult. For example, the reaction of N O with superoxide produces the peroxynitrite free radical, the marker for which (3-nitrotyrosine) has been localized to infarcted tissue in a mouse model of cerebral ischemia (Eliasson et al, 1999). Interestingly, the same authors also showed that the marker for N O production, L-citrulline, was found in peri-ischemic tissue (containing live neurons), indicating that neurotoxicity could occur in cells other than those that produced N O . This agrees with earlier findings that N O is released by NO-producing neurons to damage adjacent neurons (Dawson et al, 1993) and that those neurons which produce N O appear to be somewhat resistant to NO-mediated toxicity. This may be the result of the presence of superoxide dismutase, which presumably prevents the production of peroxynitrite by preventing the interaction between N O and the superoxide radical (Gonzalez-Zulueta et al, 1998). 6) N a + and p H 18 The loss of ion homeostasis during anoxia/ischemia is not confined to the regulation of C a 2 + . Studies in dissociated hippocampal neurons have indicated that when an increase in [Ca ]i is prevented in neurons following a dithionite-induced anoxic insult, neuronal swelling and injury still occurred, and this injury was attenuated by the replacement of [Na + ] 0 by impermeant ions during anoxia (Friedman and Haddad, 1993). Neuronal injury could also be attenuated if [Na + ] 0 was replaced after anoxia but prior to the rupture of the plasma membrane (Friedman and Haddad, 1994b). Using the same model, it was also shown that a 4 min anoxic exposure resulted in an increase in [Na+]; from -25 m M to ~ 50 m M , which peaked following the anoxic period (Friedman and Haddad, 1994a). Removal of [Na + ] 0 during anoxia also resulted in an decrease in [Ca 2 +]i for the duration of anoxia, which recovered upon return to normoxia (Friedman and Haddad, 1993; Friedman and Haddad, 1994b). Studies in perfused neocortical cultures suggested that anoxia-induced N a + entry into the cell occurred via reverse operation of the N a + / C a 2 + exchanger, exchanging C a 2 + ; for N a + 0 , as this increase in [Na+]i was not sensitive to T T X , glutamate receptor antagonists or inhibition of the C17HCCV exchanger (Chidekel et al., 1997). However, recent studies have shown that glutamate or NMDA-induced increases in [Ca 2 +]i are in part mediated via reverse operation of the N a + / C a 2 + exchanger (Hoyt et al., 1998; Kiedrowski, 1999), as is anoxia-induced injury in the rat optic nerve (Stys et al., 1992). Hence N a + may act as a mediator of anoxia/ischemia-induced neuronal injury, but the roles of N a + and C a 2 + in mediating neuronal injury should not always be considered in isolation. Both extracellular pH (pHo) and intracellular pH (pH) drop to ~ 6.2 during cerebral ischemia in vivo (Silver and Erecinska, 1992). Although originally considered a contributor to neuronal injury, it has more recently been suggested that mild extracellular acidosis can be protective. In in vitro models of oxygen glucose deprivation, maintenance of a decrease in pHo (mimicking in vivo conditions) resulted in a decrease in neurotoxicity (Giffard et al., 1990; Kaku et al., 1993), compared to cultures exposed to the same conditions at normal pHo, most likely due 19 to reduced activation of the N M D A receptor under acidic conditions (Traynelis and Cull-Candy, 1990). In a model of metabolic inhibition using CN" and 2-deoxy-D-glucose, delaying return of pH; to normal values in the reperfusion period through the use of N a + / H + exchange inhibitors in rat neocortical cultures was neuroprotective (Vornov et ah, 1996). This protection could have been due to inhibition of injury mechanisms by intracellular acidosis; alternatively, inhibition of N a + / H + exchange activity could have prevented the exchange of H + ; for N a + Q , thus decreasing the entry of N a + into the cells (Vornov et al., 1996). Subsequent to an accumulation of N a + , reverse operation of the N a + / C a 2 + exchanger (see above) could lead to an increase in [Ca 2 +];, which has been speculated as an injury mechanism in ischemia/reperfusion in the heart (Pierce and Meng, 1992). Given that anoxia can produce changes in pH; in vitro (Diarra et al., 1999), changes in pH; could have an effect on the viability of neurons after anoxia/ischemia in vitro. IV. S P E C I F I C A I M S O F T H I S S T U D Y There are three specific aims of this study: 1) To establish a rapidly acting, reliable model of anoxia-induced neuronal death using cultured hippocampal neurons. 2) To use the model as a basis for investigation to determine the mechanisms of anoxia-induced neuronal death. 3) To investigate whether there is a correlation between anoxia-evoked changes in intracellular calcium and the degree of neurotoxicity produced by an identical anoxic insult. 20 M A T E R I A L S & M E T H O D S I. Hippocampal cultures A l l experiments were performed on primary cultures of postnatal hippocampal neurons derived from 2- to 4-day old male Wistar rat pups using a method modified from that described previously (Abdel-Hamid and Baimbridge, 1997). Al l media and B27 supplement were obtained from Gibco (Grand Island, N Y ) . Several days prior to harvesting hippocampal neurons, 18 mm diameter coverslips (Fisher Scientific) were heated at 70°C for 30 minutes in nitric acid and washed lOx in distilled water prior to autoclaving. From this point, all manipulations were performed under sterile conditions, and all solutions used were purchased sterile or sterilized by filtration or autoclaving. After sterilization, coverslips were soaked overnight in 0.15 M sodium borate, p H 8.4 containing 10 Lig/ml poly-D-lysine (Sigma). After three rinses in sterile distilled water, coverslips were individually coated with 0.15 ml L15 media containing 15.2 u.g/ml laminin (Sigma), made by combining 1 ml of 0.15 M N a C l , 0.05 M Tris (filter-sterilized) containing 167.7 ug laminin with 10 ml of L15 media. After laminin treatment, coverslips were stored at 4°C until the day of culturing. When referring to culture age, the plating day was counted as day zero in vitro (0 DIV). Cultures were used after 7-12 D I V for imaging experiments, and 9-12 D I V for experiments utilizing the anoxia models. On the day of culturing, rat pups between 2-4 days old were anesthetized with CO2 and decapitated. The whole brains were removed and placed immediately into ice-cold LI5 medium containing glucose (34 mM) and gassed with O2. Hippocampi were dissected out, chopped with a scalpel blade and placed into ice-cold L15/glucose gassed with O2 prior to dissociation. 21 To dissociate hippocampal cells, excess L15/glucose was discarded and hippocampal tissue was incubated for 15 minutes in L15/glucose (pH 7.2), containing papain (1 mg/ml, Sigma) and D N A s e (25 (ig/ml, Sigma), gassed with O2 and maintained at 37°C. Enzymatic digestion was terminated by carefully discarding the enzyme-containing L15/glucose supernatant and adding to the hippocampal tissue, 7 ml of 5% CCVequilibrated DME/F12 (hereafter referred to as D M E F ; see Appendix 1 for composition). Using a 5 ml serological pipette, the hippocampal tissue was first pipetted up and down a total of ten times. This procedure was then repeated twice more using a series of fire-polished Pasteur pipettes of decreasing diameters. Dissociated cells were counted using a hemocytometer, and an appropriate volume of D M E F was added to yield a final density of 3 x 105 cells/cm2 when 0.2 ml of cell suspension was added to each coverslip. Coverslips were then transferred to a 37°C incubator for 2 hours before being transferred to 12-well culture plates containing 1 ml/well of D M E F . After 24 hours, D M E F was replaced with 5% CCVequilibrated, B27-supplemented Neurobasal medium ( N B M ; see Appendix 1 for composition). Glial proliferation was inhibited by a single addition of cytosine arabinoside (final concentration: 10 uM) to the culture medium 48 hours after plating. Cultures were replenished by exchanging 0.5 ml of N B M per well with fresh N B M at 4, 7, and 11 DIV. 77. Static Anoxia Model Cultures (grown on coverslips as described) were placed in sterile 12-well plates where each well contained 1 ml of an anoxic or normoxic bicarbonate-buffered salt solution (BBSS, see Appendix 1 for composition). Anoxic B B S S was prepared by pre-gassing with 95% N2 / 5% C 0 2 and adding 10 u.1 of freshly prepared 200 m M sodium dithionite to each well to produce a 2 22 m M final concentration. The anoxic exposure took place immediately after addition of dithionite in a shallow 37°C water bath which was flooded with 95% N 2 / 5% CO2. Cultures were kept in anoxic BSS for the desired duration prior to being rinsed by dipping into a beaker of normoxic B B S S for 3 s and returned to normoxic conditioned media. Control manipulations in normoxic B B S S were always performed in parallel on sister cultures in a 37°C incubator with an atmosphere of 5% CO2 / balance air. Drugs were added as concentrated stock solutions to normoxic or anoxic B B S S , and to conditioned media pre- or post-treatment where indicated, to produce the desired final concentrations. Most drugs (glutamate, sodium dithionite, pyruvate, D T T , MK-801, A P V , T T X , P B N , GPT) were made up as concentrated stock solutions in B B S S ; C N Q X was dissolved in D M S O . For experiments with low external sodium, N a C l in B B S S was replaced by equimolar choline chloride. For experiments with low external chloride, N a C l in B B S S was replaced by equimolar sodium gluconate. For experiments performed in zero external C a 2 + , C a C l 2 was omitted from BBSS. Neuronal viability was determined by assessing the ability of neurons to exclude trypan blue (which stains non-viable cells). At 4 hours post-treatment, cultures were incubated for 5 min in 0.4% trypan blue in HEPES-buffered salt solution (HBSS, see Appendix 1 for composition) at 37°C. Following incubation, cultures were rinsed twice by dipping into a beaker of BSS and transferred to 12 well culture plates containing 1 ml/well H B S S and examined using phase-contrast microscopy. At least 100 neurons were assessed in randomly selected fields from each culture, and the percentage of viable neurons in each culture was calculated. From these percentages, a mean percentage of neuronal viability was determined for each treatment group. Data for each test condition represents a minimum of 3 independent experiments with at least 3 cultures per treatment group in each experiment. Statistical analyses performed included 23 unpaired Student's T-test, or A N O V A followed by Tukey's multiple comparisons test, where appropriate. III. Perfusion Anoxia Model A special anodized aluminum chamber was constructed in the Department of Physiology workshop which allowed for superfusion of up to 4 cultures at once. The chamber was placed in a shallow water bath maintained at 37°C for at least 30 minutes prior to each experiment. For anoxia experiments, the water bath was flooded with 95% N2 / 5% CO2. A single perfusion pump provided both buffer superfusion and suction, maintaining the chamber buffer at a constant level. Neuronal cultures (grown on coverslips as described) were superfused at a rate of 2.0 ml / min for 5 minutes with anoxic B B S S warmed to 3 7 ° C , rinsed by dipping into a beaker of normoxic B B S S for 3 s and returned to normoxic conditioned media. Anoxic B B S S was pre-gassed with 95% N 2 / 5% CO2 prior to use and contained 2 m M sodium dithionite. Control perfusions in normoxic B B S S were always performed in parallel on sister cultures under normal atmospheric conditions at 37°C. In some experiments a 5 min anoxic perfusion was immediately followed by a 5 min normoxic perfusion (as described in Figure 9, Panel B) prior to returning cultures to normoxic conditioned media. When switching from anoxic to normoxic perfusion, the pump rate was increased to 10 ml / min for the first 30 s of normoxia. Neuronal viability was assessed at 4 or 24 hours post-treatment by trypan blue exclusion as described above. Data for each test condition represents a minimum of 3 independent experiments with at least 3 cultures per treatment group in each experiment. Statistical analyses were performed using A N O V A followed by Tukey's multiple comparisons test. 24 IV. Ca2+-imaging with fura-2 i. Loading Fura-2 was purchased as its acetoxymethyl (AM) ester from Molecular Probes Inc. (Eugene, Oregon). Fura-2-AM (1 mg) was dissolved in 1 ml chloroform and then divided into 30 ul aliquots prior to evaporation of the chloroform under vacuum to produce vials containing 30 ug of Fura-2-AM each. These aliquots were stored at -80 °C until use. To load cultures with Fura-2, 25 1^ of anhydrous D M S O and 12.5 ul of pluronic F127 were added, using a fine pipette tip, to a vial of fura-2-AM and then vortex mixed. This mixture was added, while mixing vigorously with a vortex stirrer, to 4 ml of a balanced salt solution (BSS, see Appendix 1 for composition). The mixture was then vigorously vortex mixed for 10 s in order to disperse the fura-2-AM solution. Individual cultures were placed in a well of a 6-well culture dish and incubated for 1 hour at 37°C in 2 ml of this loading media, followed by at least 20 minutes in BSS without Fura-2 at 37°C to allow for complete de-esterification. Neurons were then transferred to fresh BSS prior to mounting on a supervision chamber. For the duration of the imaging experiments, cells were maintained at 37°C. Whenever possible, control and experimental data were collected from paired sister cultures loaded with identical solutions. ii. Imaging Fura-2 fluorescence was measured using a Zeiss Attofluor™ digital fluorescence imaging system controlled by Attofluor imaging software. The microscope used was a Zeiss Axiovert-10 fluorescent microscope equipped with a long distance 40x objective (Zeiss L D UV-Achroplan 0.6, Ph2) and a 100 W mercury arc lamp as a light source. The mercury lamp was used at either full or half power for all of the experiments and no neutral density filters were in place. Fura-2 measurements were made using a dual excitation ratiometric method as described by Diarra et al. 25 (1999). Briefly, cells loaded with Fura-2 were viewed and the dye was excited alternately at either 334 or 380 nm (10 nm band pass filters whose position was determined by a computer-controlled solenoid filter changer) and the resulting fluorescence above 510 nm at each excitation wavelength was recorded. The imaging software operated on a dual-monitor I B M -compatible 80486-33 M H z which controlled the filter changer, camera gain, image processor, data processor, data display and data storage. One of the monitors displayed a pseudocoloured image of the field of view while the second monitor gave the raw emission intensity plotted against time. The photosensitive detector was a high sensitivity, intensified charge-coupled device (CCD) camera and images were digitized to 8 bit resolution with a 512 x 480 pixel frame size on a Matrox-AT image processor. Hippocampal cultures were prepared as described above. Cultures were mounted face-up in a superfusion chamber made in the Department of Physiology workshop. The inflow channel was connected to a perfusion pump while the outflow channel was connected to a suction line which removed all of the superfusate present above a level of 3 mm. For perfusion experiments, neurons were superfused at a rate of 2.0 ml/ min with BSS warmed to 3 7 ° C , thus maintaining the cultures at that temperature. For static experiments the same superfusion chamber was used but the buffer temperature in the superfusion chamber was maintained at 37°C through the use of a foil heater (under the chamber) linked to a microcontroller (Cell MicroControls) which also monitored the temperature of the superfusion chamber via a thermistor within the chamber. In order to select an appropriate field of view, a 12 V 100W halogen lamp was used to visualize the cells under phase illumination. Regions of interest (ROIs) were set at 10 x 10 pixels and placed over the soma of up to 99 neurons in the field of view. The camera gain was set at a level in order to minimize camera saturation while at the same time maximizing image intensity. The frequency of image acquisition was varied between 1 image every second to 1 image every 10 seconds depending on the timepoint within each experiment. Less frequent 26 acquisition rates were used whenever possible to eliminate photobleaching of the fura-2. Calibration of the Fura-2 signal was not attempted. The effects of experimental manipulations on [Ca 2 + ] i are presented as the raw ratio of emission intensities evoked by excitation at 334 nm to those evoked by excitation at 380 nm (I334/I380), without background fluorescence correction. Anoxia was achieved by adding sodium dithionite (final concentration 2 mM) to BSS pre-gassed with 100% N2 or argon. For perfusion experiments anoxia was induced for a 5 minute period by switching the perfusion line from a beaker containing normoxic BSS to a beaker containing anoxic BSS, and was terminated by switching the perfusion line back to normoxic BSS. During anoxic periods, the perfusion chamber was gassed with 100% N2 or argon to maintain anoxia. Baseline measurements were recorded for the first 5 min in normoxic BSS, followed by a 5 min exposure to anoxic BSS. Recording continued following anoxia, as cultures were perfused with normoxic BSS for at least an additional 10 min until the experiment was terminated. For experiments under static conditions baseline measurements were recorded for the first 5 min under normoxic conditions, followed by buffer change to allow exposure to 5 minutes of either anoxia or continued normoxia (for controls). A second buffer change was performed to return to normoxic conditions and measurements were recorded for at least an additional 10 minutes, after which the experiment was terminated. MK-801 (2 u.M) was present in the anoxic BSS for some experiments. To switch between normoxic and anoxic conditions, the perfusion line was transferred to a beaker containing the appropriate buffer (normoxic or anoxic) and the pump was operated at maximum capacity (10 ml / min) for 20 s to flush the previous buffer from the chamber. The pump was then turned off for the desired period of time. 27 R E S U L T S Please note that unless otherwise specified, all results described were obtained under static conditions. /. Morphological Appearance of Hippocampal Neurons in Culture Neuronal cell bodies in hippocampal cultures were easily distinguishable from the underlying layer of glial cells using phase contrast microscopy. Neuronal soma appeared phase-bright with a distinctive 'halo', and neurons formed networks of neuronal processes (Fig. 1A). Although neurons appeared normal immediately following exposure to 5 min of anoxia, neuronal morphology was altered by 1 hour post-exposure with swelling of both cell soma and shrinking of the nucleus. By 4h post-anoxia, neurons were markedly swollen and were unable to exclude trypan blue (Fig. IB), with noticeable fragmentation and degeneration of processes. The nuclei of these neurons became distinct and appeared phase-dark, with the soma lacking the characteristic halo of healthy neurons. The timecourse and morphological progression of this cell death was very similar to that produced by a 5 min exposure to 100 u M glutamate (Fig. IC). The inclusion of 2 u M MK-801 during the anoxia period resulted in a dramatic increase in neuronal viability when assessed at 4h post-anoxia (Fig. ID). At 24h post-anoxia, cellular debris was evident where neurons had been, with nuclear remnants stained by trypan blue; the few neurons that did survive to this point appeared grossly swollen with discontinuous processes, and were unable to exclude trypan blue. Again, the development of anoxia-induced neuronal death was paralleled in cultures exposed to 100 u M glutamate. Based on these observations, it appeared that anoxia produced a neurotoxic insult similar in effect to excitotoxicity. 28 Figure 1. Morphology of hippocampal neurons in vitro. Neurons were exposed to A ) control buffer, B) anoxia, C) 100 u M glutamate, or D) anoxia with 2 u M M K - 8 0 1 , for 5 minutes at 37 °C and returned to a 37 °C incubator for 4 hours. Cultures were then stained with 0.4% trypan blue in H B S S for 5 minutes at 37°C prior to being rinsed in fresh H B S S . Pictures taken under phase contrast optics, size bar = 50 urn. 29 II. The Effect of Anoxia Duration on Neuronal Viability A necessary feature of the anoxia model to be established was the duration of the anoxic insult required to produce neuronal injury. The degree of neuronal injury or death produced by anoxia had to be sufficient so as to be distinguishable from the level of injury or death seen in control cultures, yet not complete, allowing for the determination as to whether or not a given treatment prevented or even exacerbated neuronal death. Hence the relationship between the duration of the anoxic insult and the degree of neuronal death produced was investigated. Hippocampal cultures were exposed to durations of anoxia varying from 2 to 10 minutes, and the amount of cell death, assessed after 4 hours, was compared to cultures exposed for 5 minutes to either normoxic buffer alone or containing 100 LiM glutamate. The results are summarized in Figure 2. Cultures exposed to normoxic buffer for 5 minutes had a mean neuronal viability of 81 ± 1.2 %, whereas cultures exposed to 100 u M glutamate under the same conditions had a viability of 5.4 ± 0.51 % after 4 hours (Fig. 2). Exposing cultures to anoxic periods of varying duration produced intermediate levels of neuronal death. Following 2 minutes of anoxia mean neuronal viability was 40 ± 3.4 %. Exposure to anoxia for either 5 or 10 minutes resulted in significantly lower neuronal viability (22 ± 2.6 % and 27 ± 3.1 % respectively) than 2 minutes, although the decrease in viability was not as profound as that resulting from glutamate exposure. The percentage of viable neurons in cultures exposed to 5 or 10 minutes of anoxia did not differ significantly from each other (Fig. 2). Hence a 5 minute duration of anoxia was established as a suitable exposure time upon which to base our model of anoxia-induced neuronal death. 30 100 -i 90 -80 -• Control • Glutamate (100 uM) S2 min Anoxia 05 min Anoxia HU110 min Anoxia Figure 2. Varying the Duration of Anoxia Affects the Degree of Neuronal Death Produced. Control cultures were incubated for 5 min in normoxic buffer alone. Parallel exposures of experimental cultures to 100 u M glutamate for 5 min, or for 2, 5 or 10 min to anoxia were performed. Neuronal viability, assessed by trypan blue exclusion, was measured 4h post-treatment. Means with different letters differ significantly from one another (PO.05, A N O V A , n=13-21). 31 /ZT. Toxicity of Sodium Sulfite in Hippocampal Cultures The major oxidation product of sodium dithionite in solution is sodium sulfite (Na2S03). To verify that the acute neurotoxicity produced by a 5 minute exposure to anoxia induced by sodium dithionite was not due to a toxic effect of sodium sulfite, hippocampal cultures were incubated in normoxic buffer for 5 minutes with 2 m M sodium sulfite. Neuronal viability was then assessed 4 hours post-treatment, and compared to sham-treated control cultures (Fig. 3). Viability of neurons 4 hours after exposure to 2 m M sodium sulfite in normoxic buffer (80 ± 1.8 %) was not significantly different from those incubated for 5 minutes in normoxic buffer alone (83 ± 2.8 %), indicating that sodium sulfite was not acutely toxic to hippocampal cultures and could not account for the neurotoxicity produced by sodium dithionite. 32 100 90 80 g 70 o D 60 o % 50 JQ .5 40 ^ 30 20 10 0 • Control Sodium Sulfite Figure 3. Sodium Sulfite is not Acutely Toxic in Hippocampal Neurons in vitro. Cultures were exposed to normoxic buffer alone or normoxic buffer containing 2 m M sodium sulfite for 5 minutes. Neuronal viability, assessed by trypan blue exclusion, was measured 4 h post-treatment. Means are not significantly different (P>0.05, Student's T-test, n=10). 33 IV. Glutamate or Oxygen Free Radicals as Possible Toxic Agents of Anoxia-Induced Neuronal Death Glutamate is thought to be one of the main agents responsible for the neurotoxicity of anoxic or ischemic insults in vivo, and is a common factor in in vitro models of ischemic neuronal death. We hypothesized that anoxia was causing the release of glutamate from a number of potential sources and that glutamate was the agent responsible for the majority of neuronal pathology seen following exposure to anoxia. In order to test this hypothesis, neurons were incubated with 10 U/ml of alanine aminotransferase, also known as glutamate-pyruvate transaminase (GPT), during and for 4h after exposure to 5 min of anoxia. As an enzyme co-substrate, pyruvate was also added to a final concentration of 50 m M in order to drive enzyme activity in the direction of glutamate catabolism and consequent production of alanine. G P T was shown to be significantly neuroprotective when present during and after anoxia, with a mean neuronal viability of 71 + 1.7 % compared to 31 ± 2 . 1 % seen in neurons exposed to anoxia without G P T (Fig. 4). For comparison purposes, control viability in these experiments was 84 ± 0.9 % and neuronal viability in glutamate-treated cultures was 6.6 ± 2.5 % after 4 hours (Fig. 4). In order to determine whether pyruvate itself could be neuroprotective, several cultures were treated with 50 m M pyruvate during and following anoxia, without addition of G P T . Pyruvate was also significantly neuroprotective at this concentration, increasing the mean viability of neurons to 49 ± 1.8 % 4 hours after anoxia (Fig. 4). Another putative mediator of neuronal death following anoxia is oxygen free radicals, produced as a consequence of restoration of normoxic conditions following anoxia/ischemia. In order to determine whether free radical-mediated damage could be partially responsible for the neuronal death in our model, the free radical scavenger t-phenylbutyl nitrone (PBN) was added at final concentrations of 100 or 500 L I M to neuronal cultures during, and for 4h after, exposure to anoxia. As shown in Fig. 4, treatment with either 100 or 500 u M P B N (neuronal viabilities of 36 ± 2.8 % or 27 + 2.6 % respectively) had no significant effect upon the degree of cell death induced by anoxia. 35 co c o 1_ 3 o z O n > 100 90 80 70 60 50 40 30 20 10 0 cd ce • Control SPBN (500 uM) IGlutamate ElAnoxia EES PBN (100 uM) IGPT (10 U)/ml) DDI 50 mM Pyruvate Figure 4. GPT or Pyruvate, but not PBN, Protect Neurons from Anoxia-Induced Neuronal Death When Present During and After Anoxia. Control cultures were incubated for 5 min in normoxic buffer alone. Parallel exposures of experimental cultures to 100 u M glutamate for 5 min, or for 5 min to anoxia alone, in the presence of 100 or 500 u M P B N , in the presence of 10 U/ml G P T plus 50 m M pyruvate, or in the presence of 50 m M pyruvate were also performed. Test compounds P B N , G P T and pyruvate were also present in the conditioned media of their respective treatment groups during the 4h post-anoxic period, prior to neuronal viability assessment by trypan blue exclusion. Means with different letters differ significantly from one another (PO.05, A N O V A , n=10-21). 36 V. Pharmacology of Anoxia-induced Neuronal Death Given that glutamate was implicated as the main agent of toxicity, the next series of experiments was designed to elucidate which ionotropic glutamate receptors were activated in the processes leading to neuronal injury. Since both the N M D A and A M P A receptors have both been implicated in excitotoxicity, the non-competitive N M D A receptor antagonist MK-801 (2 LiM) and the competitive N M D A receptor antagonist D - A P V (250 uiM), as well as the competitive A M P A receptor antagonist C N Q X (40 LiM), were tested for potential neuroprotective effects when present during anoxia (Fig. 5). Normoxic control cultures were found to have a neuronal viability of 83 ± 3.4 % after 4 h whereas in cultures exposed to 100 LIM glutamate for 5 min, neuronal viability after 4 h was only 5.4 ± 0.89 %. Cultures exposed to anoxia for 5 minutes had a mean neuronal viability of 26 ± 1.3 %. While 2 u M MK-801 improved neuronal viability (74 ± 2 . 3 %) to levels not significantly different from control cultures, neither 250 LIM A P V (27 ± 1.8 %) nor 40 u M C N Q X (22 ± 2.6 %) were able to improve neuronal viability above that seen in cultures exposed to anoxia without glutamate receptor antagonists (Fig. 5). This neuroprotective profile changed when either C N Q X or A P V were present both during and after anoxia. The inclusion of C N Q X (40 oM) or A P V (250 LIM) in the conditioned media following exposure to 5 min anoxia, in addition to having the antagonists present during the anoxic period, resulted in a dramatic enhancement of neuronal viability (Fig. 6). With C N Q X present during and for 4h after anoxia, a significant improvement in neuronal viability (53 ± 3 . 2 %) relative to anoxia-treated cultures (with or without C N Q X ) was seen. When A P V was present during and for 4h after anoxia, post-anoxic neuronal viability (76 ± 1 . 8 %) improved to levels not significantly different from control cultures (Fig. 6). 37 100 90 80 8 70 o 3 60 CD © 50 . Q .2 40 > £ 30 20 10 0 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • + > • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • + • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Control • Glutamate E3 Anoxia B Anoxia & MK-801 B Anoxia & C N Q X S Anoxia & A P V Figure 5. MK-801 During Anoxia Prevents Acute Anoxia-Induced Neuronal Death. Control cultures were incubated for 5 min in normoxic buffer alone. Parallel exposures of experimental cultures to 100 u M glutamate for 5 min, 5 min to anoxia alone, or in the presence of 2 u M M K -801, 40 u M C N Q X , or 250 u M A P V were also performed. Neuronal viability, assessed by trypan blue exclusion, was measured 4h post-treatment. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=10-l 1). 38 C o 3 CD Z 0 .Q AS • M i > 100 90 80 70 60 50 40 30 20 10 0 a _ L • Control El Co- & Post-anoxia C N Q X S Co- & Post-anoxia A P V Figure 6. Post-Anoxia CNQX or APV Prevents Acute Anoxia-Induced Neuronal Death. This data represents a continuation of the experiment described in Figure 5; Control viability is that from Figure 5, shown here for comparison purposes. Parallel exposures of experimental cultures to anoxia in the presence of 40 u M C N Q X or 250 u M A P V were performed, with C N Q X (40 LIM) or A P V (250 uM) also present during the 4h post-anoxic period. Neuronal viability was assessed by trypan blue exclusion 4 h post-anoxia. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=l 1). 39 The voltage-sensitive sodium channel blocker tetrodotoxin ( T T X ) was also tested for potential neuroprotective effects when present during anoxia only or when present during and for 4h after anoxia. The results of this experiment are summarized in Figure 7. Normoxic control cultures were found to have a neuronal viability of 81 ± 1.2 % after 4 h whereas, in cultures exposed to 100 u M glutamate for 5 min, neuronal viability after 4 h was 5.1 ± 0.50 %. In comparison, the mean neuronal viability for cultures exposed to 5 minutes of anoxia was 31 ± 3.2 %. The mean neuronal viability of cultures exposed to anoxia in the presence of 1 u M T T X (30 ± 4.5 %) was not found to be significantly different from that of cultures exposed to anoxia. However, when 1 u M T T X was present during and post-anoxia, there was a small but significant increase in mean neuronal viability (49 ± 7.4 %) (Fig. 7). The pharmacological agents used were tested to ensure that they were not themselves neurotoxic. Cultures were incubated with 2 u M MK-801, 40 u M C N Q X , 250 u M A P V , 1 u M T T X , or 500 u M P B N for 4 hours in normoxic BBSS; the results are summarized in Figure 8. Control cultures (not incubated with any drug) had a mean neuronal viability of 83 ± 0.82 %. None of the compounds tested produced any significant amount of cell death (Fig. 8). 40 100 n • Control • Glutamate E3 Anoxia STTX Co-Anox a TTX Co & Post Anoxia Figure 7. T T X (1 U.M) Slightly Improves Neuronal Survival When Present During & After Anoxia. Control cultures were incubated for 5 min in normoxic buffer alone. Parallel exposures of experimental cultures to 100 u M glutamate for 5 min, 5 min to anoxia alone, or in the presence of 1 u M T T X (with or without 1 u M T T X post-anoxia), were also performed. Neuronal viability was assessed by trypan blue exclusion 4 h post-anoxia. Means with different letters differ significantly from one another (P<0.05, ANOVA, n=10). 41 c o 1_ 3 d) Q) -Q ro > 100 -, 90 80 70 60 50 H 40 30 20 10 0 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • , • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • ••••••••4 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • c • • • • • • • • • • • • • • • • • • _ • • • • • • • • • ••••••••• ••••+••*••• »•••••+++• •••••••+• • • • • • • • • • • • • • • • • • • • • • • • • • • • • Control MK-801 CNQX SI APV I TTX SPBN Figure 8. M K - 8 0 1 , C N Q X , A P V , T T X , or P B N Do Not Produce Acute Neuronal Death. Control cultures were incubated for 5 min in normoxic buffer alone. Parallel exposures of experimental cultures to 2 u M MK-801, 40 u M C N Q X , 250 u M A P V , 1 u M T T X , or 500 u M P B N in normoxic B B S S for 5 min, were also performed. Neuronal viability was assessed by trypan blue exclusion 4 h post-anoxia. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=10). 42 VI. The Effect of Buffer Perfusion on Anoxia-Induced Neuronal Death One of the goals of this study was to develop an assay to measure neuronal death based upon conditions utilized for previously conducted ion imaging experiments (see Diarra et al., 1999). This would allow a correlation between anoxia-induced intraneuronal ionic changes and neuronal death. However, the anoxia model developed used static conditions, whereas the imaging studies have been conducted on cultures subjected to constant buffer perfusion. Hence, to simulate accurately the conditions used for imaging experiments, a chamber was built to allow perfusion of 4 cultures grown on coverslips at once. This facilitated the study of neuronal viability following exposure to anoxia under perfusion conditions. The protocols followed for perfusion anoxia experiments are shown in Figure 9, and the results are summarized in Figure 10. Neuronal viability 4h after a 5 min normoxic perfusion (84 ± 0.87 %) was similar to that of cultures incubated in normoxic buffer under static conditions, and was not significantly different from the level of neuronal viability 24h after perfusion (Fig. 10). Perfusion anoxia resulted in a viability of 17 ± 2.3 % 4h post-anoxia, similar to that seen under static conditions, and the level of neuronal viability decreased slightly, but not significantly, by 24h. If 5 min of perfusion anoxia was followed by a further 5 min of normoxic perfusion (Fig 9, Panel B), a dramatic increase in acute neuronal viability was observed at the 4h assessment point (60 ± 5.9 %, Fig. 10). However, this acute increase was not maintained by 24h post-anoxia, as neuronal viability (27 ± 5 . 7 %) was decreased to levels comparable to those seen 4h after perfusion anoxia alone (Fig. 10). 43 Anoxia/ Normoxia Recovery A Anoxia Normoxia Recovery B Time Scale 5 min 5 min 4h/24 Figure 9. Protocols Used in Perfusion Anoxia Experiments. The time scales for the experiments performed are 4 and 24 h Perfusion Control or Perfusion Anoxia as indicated above. Panel A outlines the protocol used for Perfusion Control or Anoxia experiments, where perfusion anoxia or normoxia was followed by a 4h recovery period. Panel B shows the protocol followed for 4 and 24 h Perfusion Anoxia + Wash experiments, where 5 min of perfusion anoxia was followed by 5 minutes of normoxic perfusion prior to recovery. Neuronal viability was assessed by trypan blue exclusion after a 4h recovery period for 4h experiments, or after 24h for 24h experiments. 44 EB4h Perfusion Control EH4h Perfusion Anoxia E3 4h Perfusion Anoxia + Wash 10 24h Perfusion Control S 24h Perfusion Anoxia B 24h Perfusion Anoxia + Wash Figure 10. Effect of Flow Environment on Anoxia-Induced Cell Death 4h or 24h Post-Anoxia. Refer to Figure 9 for a description of the protocols and treatment groups. Means with different letters differ significantly from one another (P<0.05, ANOVA, n=9-21). 45 VII. Involvement of External Ions in Anoxia-Induced Neuronal Death A number of extracellular ions have been implicated as mediators of excitotoxic or anoxia-induced neuronal death, including N a + , C f and C a 2 + . Having established the involvement of the N M D A receptor channel (which gates a current of which 2 major components are N a + and C a 2 + ) in this model we next investigated the potential roles that these external ions may play in mediating acute anoxia-induced neuronal death. i. Additional Controls Particular ions were replaced or removed during anoxia only, or during anoxia as well as the post-anoxic recovery period. Since substitution or removal of ions during the post-anoxic period required the post-anoxic recovery to occur in BBSS , rather than conditioned media, a series of control experiments was first performed to examine the effect on neuronal viability of a 4h post-anoxic incubation in BBSS alone. As shown in Figure 11, the viability of control cultures (82 ± .80 %) was not significantly different from cultures which were handled in the same way but allowed to recover from a normoxic exposure in B B S S rather than conditioned media (4h buffer controls, 78 ± 0.93 % viability). Similar results were found with cultures exposed to anoxia; the viability of cultures allowed to recover in conditioned media (Anoxia, 32 ± 3 . 2 % viability) after anoxia was not significantly different from those which recovered in normoxic B B S S (Post-Anoxia Buffer, 36 ± 2.1 %) following anoxia (Fig. 11). Additional control experiments examined the degree to which simply handling the cultures (e.g. rinsing or transfer between buffers) affected neuronal viability compared to cultures that were not handled at all (incubator controls). Incubator controls had a mean neuronal viability of 84 ± 0.81 % (Fig. 11), which was not significantly different from cultures that were handled under normoxic conditions (mean neuronal viability 82 ± 0.80 %). 46 100 i • Control S3 Incubator Control H4h Buffer Control • Glutamate C3 Anoxia M Post-Anoxia Buffer Figure 11. Post-Anoxic Recovery in Buffer Instead of Conditioned Media Does Not Increase Acute Neuronal Death. Incubator Control cultures were not manipulated in any fashion. Control cultures were incubated for 5 min in normoxic B B S S alone and allowed to recover in conditioned media or normoxic B B S S (4h Buffer Control). Parallel exposures of experimental cultures to 100 u M glutamate for 5 min, or 5 min to anoxia with recovery in conditioned media (Anoxia) or BBSS (Post-Anoxia Buffer) were also performed. Neuronal viability was assessed by trypan blue exclusion 4h post-anoxia. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=9). 47 ii. Anoxia Under Low [Na+J0 or Low [Ct]0 Conditions Choline chloride or sodium gluconate were used to substitute for sodium chloride, simulating either low extracellular sodium (low [Na +] 0) or chloride (low [Cl"]0) conditions. Figure 12 summarizes the effects on neuronal viability of low [Na + ] 0 or [Cl"]0 (as well as 0 [Ca 2 + ] 0 ; see below), either during, or during and for 4h after anoxia, and is a continuation of the experiments presented in Figure 11. Substitution for either N a + (36 ± 5.2 %) or Cl" (39 ± 3.1 %) during the anoxic period neither improved nor reduced neuronal viability following anoxia. However, the maintenance of low [Na + ] 0 (54 ± 2.5 % viability) or low [Cl"]0 (55 ± 0.70 % viability) conditions during the 4h post-anoxic recovery period significantly enhanced neuronal viability when compared to cultures exposed to and recovering in normal B B S S (Post-Anoxia Buffer, 36 + 2.1 %). iii. Anoxia Under 0 [Ca ]0 Conditions To reduce extracellular calcium, CaCl2 was omitted from B B S S to simulate nominally Ca 2 +-free conditions. Although omitting C a 2 + during the anoxic period neither improved nor reduced neuronal viability (23 ± 4.8 %) compared to anoxia in normal B B S S , maintenance of 0 [Ca 2 + ] 0 conditions for 4h following anoxia produced a significant decrease in post-anoxic neuronal viability (7.0 ± 2.0 %, Fig. 12), resulting in a level of cell death which was not significantly different from that produced by exposure to 100 u M glutamate (5.4 ± 0.47 %, Fig. ID-Given the finding that 0 [Ca 2 + ] 0 post-anoxia exacerbated anoxia-induced neuronal death, the question arose as to whether the 0 [Ca 2 + ] 0 environment was acutely toxic to the neurons independent of anoxia-induced injury. To examine this possibility, cultures were incubated under normoxic conditions in 0 [Ca 2 + ] 0 buffer for 5 minutes (0 Ca Control) prior to being rinsed 48 100 - i 90 -El Post-Anoxia Buffer E 3 Low Na A 11 Low Na PA • 0 Ca A I 10 Ca PA B Low Cl A m Low Cl PA Figure 12. Effect of External Ion Substitution on Acute Anoxia-Induced Neuronal Death. These data represents a continuation of the experiment described in Figure 11; Post-Anoxia Buffer viability is that from Figure 11, shown here for comparison purposes. Exposures of experimental cultures to 5 min of anoxia in buffer with low [Na + ] 0 , [Cl"]0 or 0 [Ca 2 + ] 0 with recovery in conditioned media (Low Na A , Low C l A or 0 Ca A respectively) or recovery in buffer with low [Na + ] 0 , [Cl~]0 or 0 [Ca 2 + ] 0 (Low Na P A , Low C l P A or 0 Ca P A respectively), were performed. Neuronal viability was assessed by trypan blue exclusion 4h post-anoxia. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=9). 49 and returned to BBSS. This was repeated with cultures being placed in 0 [ C a 2 + ] 0 buffer for an extra 5 (0 Ca5 Control) or 30 (0 Ca30 Control) min prior to being returned to BBSS , or for 4h (0 Ca4h Control) without a return to BBSS prior to viability assessment. As shown in Figure 13, incubating cultures under 0 [Ca ] 0 conditions for all intervals tested did not produce neuronal death beyond that seen in cultures incubated in normoxic BBSS. This would seem to indicate that an enhancement of neurotoxicity occurred only when cultures were exposed to anoxic conditions in a 0 [Ca ] 0 environment. We attempted to determine if 0 [Ca 2 + ] 0 immediately following anoxia (presumably overlapping the anoxia-induced release of glutamate) but limited to a short duration, would result in an enhancement of neuronal viability following anoxia. The results are presented in Figure 14. The data presented are a continuation of that presented in Figure 13, and as such the experimental manipulations occurred at the same time to facilitate comparisons. As shown, neuronal viability was improved in cultures placed into Ca 2 +-free BBSS for 5 (62 ± 3.0 %) or 30 (57 ± 2.9 %) minutes following anoxia in Ca 2 +-free B B S S (prior to being returned to conditioned media) compared to cultures which were exposed to anoxia in normal BBSS and then allowed to recover in conditioned media (Anoxia, 19 ± 2.1 % viability). However, cultures placed into normal B B S S for 5 or 30 minutes following anoxia (in normal BBSS) showed the same significant increase in neuronal viability (65 ± 2.0 % and 68 ± 1.7 %, respectively). This indicated that the replacement of old buffer with fresh buffer rather than the absence of external calcium was responsible for the increases in 9+ neuronal viability seen. Maintenance of cultures in Ca -free B B S S for 30 min after anoxia resulted in slightly, but significantly, lower neuronal viability than maintenance in normal BBSS for 30 minutes after anoxia (Fig. 14). 50 C/> c o 3 0) z J Q (0 100 90 H 80 70 60 50 40 30 H 20 10 0 • C o n t r o l • G l u t a m a t e H 0 C a C o n t r o l m 0 C a 5 C o n t r o l 1 0 C a 3 0 C o n t r o l m 0 C a 4 h C o n t r o l Figure 13. Incubation in Ca 2 +-free Buffer Instead of Media Does Not Increase Acute Neuronal Death. Control cultures were incubated for 5 min in normoxic B B S S and allowed to recover in BBSS. Parallel normoxic exposures of experimental cultures to 100 u M glutamate for 5 min or 5 (0 Ca Control), 10 (0 CalO Control), or 30 (0 Ca30 Control) min to 0 [Csr+]0 buffer prior to recovery in BBSS , or for (0 Ca4h Control) 4h in 0 [ C a 2 + ] 0 buffer without a return to BBSS , were also performed. Neuronal viability was assessed by trypan blue exclusion 4h post-anoxia. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=9-12). 51 100 i 90 -80 -0 Anoxia m 0 Ca5 Anoxia S 0 Ca30 Anoxia H Normal Ca5 Anoxia B Normal Ca30 Anoxia Figure 14. Buffer Replacement, not Removal of External C a 2 + for 10 or 30 Minutes, Reduces Acute Anoxia-Induced Neuronal Death. These data represents a continuation of the experiment described in Figure 13. Exposures of experimental cultures to 5 min of anoxia in B B S S followed by recovery in conditioned media (Anoxia) or B B S S for 5 (Normal C A 5 Anoxia) or 30 minutes (Normal CA30 Anoxia), or anoxia in 0 [Ca 2 + ] 0 with recovery in [Ca 2 + ] 0 buffer for 5 (0 CalO Anoxia) or 30 (0 Ca30 Anoxia) minutes, were performed. Neuronal viability was assessed by trypan blue exclusion 4h post-anoxia. Means with different letters differ significantly from one another (P<0.05, A N O V A , n=9-22). 52 VIII. Neurotoxicity Produced by Reducing Agents The reduction-oxidation state of the N M D A R can affect the functional properties of the associated ion channel, which can directly affect the severity of N M D A receptor-mediated neuronal damage. The effect of shifting the reduction-oxidation state of the N M D A receptor to the fully reduced form on anoxia-induced neuronal death was assessed by using the reducing agent dithiothreitol (DTT). The results of this experiment are summarized in Figure 15. Normoxic control cultures were found to have a neuronal viability of 86 ± 0.68 % after 4 h whereas in cultures exposed to 100 u M glutamate for 5 min, neuronal viability after 4 h was 6.5 ± 0.37 %. In comparison, the mean neuronal viability for cultures exposed to 5 minutes of anoxia was significantly greater (23 + 2.3 %), increasing to 81 ± 1.4 % when 2 |^M-MK801 was present during the anoxic exposure (Figure 15). A 5 min incubation with 2 m M D T T in normoxic B B S S was very toxic, resulting in a mean neuronal viability of 10 ± 0.69 %, not significantly different from that seen in glutamate-treated cultures. When MK-801 (2 uM) was present during the incubation with D T T , neuronal viability increased to 85 ± 1.8 %, and was not significantly different from that seen in control cultures or cultures exposed to anoxia in the presence of MK-801 (Fig. 15). 53 100 i • Control •Glutamate E3 Anoxia Bl Anoxia & MK-801 11DTT B DTT & MK-801 Figure 15. 2 m M D T T Has an Acute Toxic Effect on Hippocampal Neurons Which C a n Be Blocked by M K - 8 0 1 . Control cultures were incubated for 5 min in normoxic B B S S and allowed to recover in conditioned media. Parallel exposures of experimental cultures to 100 L I M glutamate for 5 min, or 2 m M D T T with or without 2 u M MK-801 for 5 min, or 5 min of anoxia with or without 2 u M MK-801 were also performed, with post-treatment recovery in conditioned media. Neuronal viability was assessed by trypan blue exclusion 4h post-anoxia. Means with different letters differ significantly from one another (PO.05, A N O V A , n=9-16) 54 IX. Summary of Neuronal Viabilities One essential characteristic for a model of anoxia-induced neuronal death is its reproducibility, or consistency in producing similar rates of neuronal survival or death in repeated independent experiments over time. We compiled the neuronal viabilities of control, glutamate, and anoxia treatment groups from experiments presented in this study which had all three treatment groups (a total of 20 independent experiments), and plotted a frequency histogram (Figure 16) to establish the ranges for neuronal viability resulting from these treatments in our model. Control cultures had mean neuronal viabilities ranging between 76 and 93 %, with the majority (14/20 experiments) falling between 80 and 90%. Glutamate treatment of cultures produced a very consistent degree of cell death, with neuronal viability never above 10%. Exposure to 5 minutes of anoxia typically reduced cell viabilities to between 10 and 40 % (19/20 experiments) and, on only one occasion, to below 10 %. 55 co •*-> c CD E 'LZ O CL X UJ «4-o 22 20 18 16 14 12 10 8 6 4 2 0 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 Mean Neuronal Viability (%) • Control Cultures •Glutamate Cultures EU Anoxia Cultures Figure 16. Frequency Histogram of Neuronal Viabilities in Hippocampal Cultures Exposed to Normoxia, Glutamate or Anoxia for 5 min. Data shown represent the number of times that the means of neuronal viabilities following a 5 minute exposure to normoxic B B S S (Control Cultures), normoxic B B S S containing 100 L I M glutamate (Glutamate Cultures) or anoxic B B S S (Anoxia Cultures) fall into a given bin defined by the percent mean neuronal viability assessed at 4h post-exposure. Total number of experiments represented equals 20. 56 X. Correlation Between Anoxia-Induced Changes in [Ca2+]i and Neuronal Death To determine whether cell death produced by exposure to 5 min of anoxia could be correlated with acute post-anoxic increases in [Ca 2 +]j, anoxia-induced changes in [Ca 2 +]j were recorded in hippocampal cultures exposed to anoxia under both perfusion and static conditions. Representative results of these experiments are shown in Figures 17 and 18. Under perfusion conditions (Figure 17), mean resting levels of C a 2 * remained constant under normoxic conditions. In response to 5 min anoxia, two types of responses were seen. In the first type of response, a relatively small rise in [Ca 2 +]j starting approximately 2 to 3 minutes after the initiation of anoxia was seen, which peaked 2-3 minutes after return to normoxia and returned to near resting levels post-anoxia. The second type of response was a much larger increase in [Ca 2 +]i, starting within 1 minute of initiation of anoxia. Like the small response, the larger response peaked after the return to normoxia, and the ratio somewhat (but never fully) recovered towards resting levels (Figure 17). Under static conditions (Figure 18), mean resting levels of C a 2 + i remained constant during normoxia, as was seen in the perfusion situation. Also similar to the perfusion situation, there were 2 types of C a 2 + responses to 5 min anoxia under static conditions; a larger response which did not recover following anoxia, and a much smaller response which recovered to resting levels post-anoxia (Figure 18). Since 2 u M MK-801 provided acute neuroprotection in our anoxia model, its effect on 2_|_ anoxia-evoked changes in [Ca ]i when present during the anoxic period was investigated. Preliminary experiments where 2 u M MK-801 was present during the anoxic period were performed; however, since the majority of anoxia-induced increases in Ca 2 + j were very small in amplitude, any effect that MK-801 had on these responses was difficult to interpret (data not shown). 57 One interesting phenomenon noted in our static imaging experiments was a clear correlation between the resting ratio and the magnitude of the ratio increase in response to anoxia. As shown in Figure 18, large response to static anoxia had an initial resting ratio > 0.5, whereas the small response to static anoxia had a resting ratio below 0.5. Out of a total of 12 experiments with static anoxia, ratios from 114 cells were recorded. 90 cells had resting ratios < 0.5, of which 15 had large ratio responses to anoxia. Out of the 24 cells which had resting ratios > 0.5, 20 had large ratio responses to static anoxia. 58 o 00 co C O 0 Anoxia/Normoxia 10 15 Time (min) 20 25 30 Perfusion Anoxia -«- Perfusion Normoxia Perfusion Anoxia (Small Response) Figure 17. Typical Intracellular C a 2 + Responses in Fields of Neurons to 5 Minutes of Normoxia or Anoxia Under Perfusion Conditions. Mean responses of fields of neurons (n=l 1-20 neurons/field) from single coverslips loaded with Fura-2 and exposed to transient anoxic or normoxic periods. A l l coverslips were exposed to, in order: 5 minutes of normoxia, a 5 minute test period of anoxia or normoxia, followed by at least 10 minutes of normoxia. A n increase in [Ca 2 +]j is indicated by an increase in the ratio of emission intensities (I334/I380), calculated as described in Materials & Methods. 59 0 B B i J S ^ J ^ l l l l l i 3 B ! c H 8 f i S 0 10 15 20 Time (min) 25 30 Static Anoxia -•- Static Normoxia - ° - Static Anoxia (Small Response) Figure 18. Typical Intracellular C a 2 + Responses in Fields of Neurons to 5 Minutes of Normoxia or Anoxia Under Static Conditions. Mean responses of fields of neurons (n=7-10 neurons/field) from single coverslips loaded with Fura-2 and exposed to transient anoxic or normoxic periods. A l l coverslips were exposed to, in order: 5 minutes of normoxia, a 5 minute test period of anoxia or normoxia, followed by at least 10 minutes of normoxia. A n increase in [Ca 2 + ] i is indicated by an increase in the ratio of emission intensities (I334/I380), calculated as described in Materials & Methods. 60 DISCUSSION Morphological Progression of Anoxia-Induced Neuronal Death Hippocampal cell cultures have been used by several investigators to investigate neurotoxicity induced by anoxic or ischemic insults using a variety of methods (Rothman, 1983; Dubinsky and Rothman, 1991; Patel et al, 1993; Uto et al, 1995; Dux et al, 1996; Kusumoto et al, 1996; Pang and Geddes, 1997). The morphological appearance of the hippocampal neurons used in our studies matches the descriptions of those used by others. Typically neurons sat atop a glial monolayer and had smooth cell bodies, a phase bright and well differentiated appearance, and intact neurites, with few if any varicosities (Dubinsky and Rothman, 1991; Uto et al, 1995; Kusumoto et al, 1996; Pang and Geddes, 1997). The percentage of viable neurons was relatively consistent in our control cultures, and was similar to that reported by others (Dubinsky and Rothman, 1991; Uto etal, 1995; Kusumoto etal, 1996; Pang and Geddes, 1997). Exposure to high concentrations of glutamate produced a characteristic, acutely developing change in morphology that preceded extensive neuronal death as previously documented in other studies using hippocampal cultures (Patel et al, 1993; Pang and Geddes, 1997). In our studies, anoxia produced similar changes to that of glutamate, with neurons typically appearing swollen by lh post-anoxia with progression to what appeared to be necrotic cell death by 4h post-anoxia. These observations are consistent with the findings of previous studies (Uto etal, 1995; Kusumoto, 1996; Pang and Geddes, 1997). The appearance of our cultures at 4h post-anoxia is also similar to previous descriptions at various times post-insult (depending on the model used), with cell swelling, fragmentation of processes and appearance of cell debris (Rothman, 1983; Kusumoto et al, 1996), and concurrent loss of phase bright appearance accompanied by nuclear shrinkage (Pang and Geddes, 1997). Our model of anoxia consistently reduced neuronal viability to 40% or below, and most often to less than 30%, 61 providing a convenient baseline to study improvements in viability achieved by various manipulations. Glutamate Mediates Anoxia-Induced Neuronal Death Our results indicate that glutamate is the mediator of anoxia-induced neuronal death. There are two direct lines of evidence to suggest this. Firstly, the glutamate-metabolizing enzyme G P T improves neuronal viability when present during and after the anoxic insult. Several glutamate-metabolizing enzymes have been shown to provide neuroprotection in models of excitotoxicity, including glutamine synthetase (Gorovits et al, 1997; Matthews et al, 2000), glutamate dehydrogenase (Matthews et al, 2000) and G P T (Dubinsky and Rothman, 1991; Blitzblau et al, 1996; Matthews et al, 2000). O f these three enzymes, G P T was found to be the most effective at reducing neurotoxicity in response to exogenously applied glutamate (Matthews et al, 2000). G P T catalyzes the conversion of glutamate and pyruvate to alanine and a-ketoglutarate (or vice versa) in an equilibrium reaction. The concentrations of both G P T and pyruvate used in our experiments were based on values established by Matthews et al. (2000) to provide maximal neuroprotection. In order to shift the equilibrium in favour of glutamate catabolism, excess pyruvate was co-applied with GPT. The second line of evidence that suggests glutamate is a major mediator of anoxia-induced neuronal death is that glutamate receptor antagonists can provide partial or complete acute neuroprotection. MK-801 provided complete neuroprotection when present during anoxia, whereas A P V maintained neuronal viability at control levels when present after anoxia, but not when present only during anoxia. C N Q X had a pattern of neuroprotection similar to A P V , though it did not provide complete neuroprotection. Hence both N M D A and A M P A receptors are involved in mediating anoxia-induced neuronal death. In the case of the conditional neuroprotection provided by A P V or C N Q X ; it has been reported that under reducing conditions 62 (established by using D T T ) , competitive N M D A receptor antagonists were ineffective at preventing neuronal death in rat organotypic hippocampal slice cultures following exogenously applied N M D A or 45 min (Vglucose deprivation. Non-competitive antagonists at the glutamate binding site (MK-801, 7-CK) provided the same level of neuroprotection regardless of the reducing environment (Pringle et al, 2000). However, this does not explain the pattern of neuroprotection seen with C N Q X . A second possible explanation is that the majority of glutamate release does not occur until after the anoxic period, that is until reoxygenation. During anoxia, a decrease and subsequent increase in intracellular p H (pHj) has been shown to occur, followed by an overshoot of pH; to alkaline values (Diarra et al., 1999). This recovery and subsequent overshoot of p H following anoxia may stimulate synaptic glutamate release, as an activation of Ca 2 +-dependent neurotransmitter release has been shown to occur in hippocampal nerve terminals (Trudeau et al, 1999) during recovery from intracellular acidification. The non-competitive antagonist M K -801 binds with high affinity (Wong et al, 1986) to its binding site inside the channel pore, and its inhibitory effects are sustained. In contrast A P V and C N Q X are both low affinity competitive antagonists and their inhibitory effects are readily reversed upon washing. Application of either of these compounds after the anoxic period will therefore be required to block any effects of glutamate released during the immediate return to normoxic conditions. Pyruvate as a Neuroprotective Agent Interestingly, pyruvate without G P T also had a moderate neuroprotective effect, though it was not as effective as when present with GPT. This effect may be explained in a number of ways. Firstly, pyruvate may cross the plasma membrane, possibly via a monocarboxylate transporter, and supplement neuronal energy production. In support of this possibility, Brorson et al. (1999) demonstrated that the application of extracellular pyruvate to cultured hippocampal 63 neurons could partially relieve iodoacetate-induced depletion of intracellular A T P , indicating that a mechanism exists in hippocampal neurons for pyruvate to cross the plasma membrane and contribute to the generation of A T P . This confirmed earlier findings by Williams et al. (1996) that extracellular lactate or pyruvate could supplement A T P production in cultured murine striatal neurons in the absence of extracellular glucose, and by (Desagher et al., 1997), who showed that pyruvate uptake occurred in rat striatal neurons. Pyruvate-dependent A T P production could be inhibited by the addition of a-cyano-4-hyroxycinnamic acid (4-CIN), an inhibitor of transmembrane monocarboxylic acid transport (Brorson et al., 1999), and 4-CIN also reversed the inhibition by extracellular pyruvate of glutamate-evoked arachidonic acid release (Williams et al., 1996). Hence pyruvate may aid neuronal survival post-anoxia by supplementing A T P production via oxidative phosphorylation in neurons which have functional mitochondria. Secondly, pyruvate has antioxidant properties and protects against free radical-mediated neuronal damage. For example, pyruvate can protect neurons from peroxide-induced toxicity when present at relatively low (~ 1 mM) extracellular concentrations by undergoing a non-enyzmatic decarboxylation reaction in the presence of H 2 O 2 leading to the formation of carbon dioxide, water and acetate (Desagher et al, 1997). This form of neuroprotection was independent of pyruvate's ability to act as a substrate for Acetyl-CoA generation, as lactate was ineffective at preventing ^(LVinduced neurotoxicity (Desagher et al, 1997). However, this mechanism of pyruvate-mediated neuroprotection is not likely to occur in our model as the free radical scavenger P B N was ineffective at protecting neurons from anoxia-induced death, even at concentrations that effectively protected organotypic hippocampal cultures from (LVglucose deprivation-mediated neurotoxicity (Newell etal, 1995). There is yet a third possible mechanism for pyruvate-mediated neuroprotection. At high concentrations (> 10 mM) pyruvate uptake results in cytosolic acidification due to co-transport of protons into neurons (Desagher et al, 1997), and in our studies a concentration of 50 m M pyruvate was used. Prolonged maintenance of an acidic intracellular p H following chemical anoxia has been postulated to be neuroprotective by inhibiting pH-sensitive injury processes or preventing alkalinization-induced (mediated by C a 2 + ) injury (Bond et al, 1993; Vornov et al, 1996; Jorgensen et al, 1999) and since extracellular pyruvate was present not only during anoxia but also during the 4h post-anoxic recovery, maintenance of an acidic pHj is a likely possibility . However, further studies will be required to determine the exact mechanism(s) of pyruvate mediated neuroprotection in this model. TTX as a Neuroprotective Agent T T X did not provide neuroprotection when present during anoxia. Several other groups have reported a lack of neuroprotection by T T X in different anoxia/ischemia protocols, including combined (LVglucose deprivation (1 L I M T T X ) in striatal slices (Calabresi et al, 1999) and (10 LiM T T X ) neocortical cultures (Goldberg and Choi, 1993), dithionite-induced anoxia (3 u.M T T X ) with acutely dissociated adult hippocampal neurons (Friedman and Haddad, 1993), and dithionite-induced anoxia (1 L I M T T X ) with cultured cortical neurons (Chidekel et al, 1997). Varming et al, (1996) found that T T X (30 or 100 nM) protected cerebellar granule neurons from azide-induced chemical anoxia when exposure was performed in non-depolarizing (i.e. normal [K + ] 0 ) conditions, but was ineffective under depolarized conditions (25 m M [K + ] Q ) where sodium channels would be expected to be largely inactivated and hence synaptically released glutamate would be minimal. This latter finding is compatible with our results since, during anoxia, it would be expected that neurons are depolarized and T T X would not be effective under these conditions. 65 Despite the lack of neuroprotection by T T X during anoxia, neuronal viability improved significantly when T T X was present during the post-anoxic period, possibly by inhibiting spontaneous synaptic release of glutamate from presynaptic nerve terminals. In contrast, Calabresi et al. (1999) concluded that lack of neuroprotection by T T X indicated that synaptic release of glutamate was unlikely in response to ischemia. Although the likely explanation of the protective effect of T T X is the blockade of N a + channels in neurons that repolarize during the immediate post-anoxic period, T T X has also been reported to block spontaneous C a 2 + transients (Chidekel et al., 1997). Such an action could provide an alternate mechanism to prevent spontaneous presynaptic glutamate release. Perfusion Anoxia Anoxia under perfusion conditions appeared to have a neurotoxic effect similar to that seen under static conditions. Neurons were swollen with condensed nuclei approximately l h post-anoxia, and by 4h most of the neurons were unable to exclude trypan blue. This acute toxicity could be mitigated in large part by continuing to perfuse for 5 minutes post-anoxia with normoxic buffer. This additional step may provide protection by washing away extracellular glutamate and metabolic byproducts (Chidekel et al., 1997). Our results suggest that glutamate release occurs within minutes following the end of anoxia and results in substantial excitotoxic cell death unless the glutamate is continuously removed by maintaining perfusion. This contention is indirectly supported by the fact that N M D A receptor antagonists administered 15-30 minutes after the end of 02/glucose deprivation are not neuroprotective (Goldberg and Choi, 1993). The concentration of extracellular glutamate increases following a duration of 02/glucose deprivation which is sufficient to produce neurotoxicity, but this increase does not occur until after a duration of insult sufficient to produce neuronal death has been attained (Goldberg and Choi, 1993). We have shown that 5 minutes of anoxia, even under perfusion conditions, is highly neurotoxic, and so it is reasonable to assume that release of endogenous glutamate can occur during or shortly after anoxia in our model. Our static anoxia experiments, showing that an extra buffer change 5 or 30 minutes after the conclusion of anoxia is also neuroprotective, further support the hypothesis that glutamate release occurs both during anoxia and the acute period following. Although we observed an acute neuronal sparing effect of a normoxic wash after perfusion anoxia, we also observed a delayed component of neuronal death that occurred between 4 and 24h post-anoxia. The question arises then as to whether this delayed component of neuronal death is necrotic or apoptotic in nature. This question is particularly relevant since recent studies suggest that apoptotic neuronal death occurs following cerebral ischemia in vivo (Nitatori et al, 1995; Chopp and L i , 1996; D u et al, 1996). In cell cultures, several studies have shown both acute and delayed neuronal death in response to simulated anoxia or ischemia (Goldberg and Choi, 1993; Gwag et al, 1995; Uto et al, 1995; Kusumoto et al, 1996; Pang and Geddes, 1997). We have not characterized the latter component of neuronal death (but see Future Directions below), but it may represent a distinct death process from that which mediates acute necrotic neuronal death. Other studies have indicated that severe ischemic insults produce an acute necrotic cell death which may mask a parallel, slower acting apoptotic pathway (Gwag et al, 1995). The results of our perfusion anoxia experiments suggest that similar dual processes mediating cell death may be occurring in our static model. External Ion Substitution - Low Cl" and Low Na+ Maintenance of cultures in B B S S following anoxic exposure facilitated an examination of the possible roles that N a + and Cl" played in mediating anoxia-induced neuronal death. We found that low [Na + ] 0 or [Cl"]0 conditions during the anoxic period did not influence neuronal viability following anoxia. Goldberg and Choi (1993) found that 0 2/glucose deprivation in low 67 [Na + ] 0 (27 mM) conditions reduced acute neuronal swelling immediately following exposure, but did not offer additional protection when cultures were returned to their original medium. A similar result was observed under low [Cl"]0 (4 mM) conditions. In our experiment, a significant improvement in post-anoxic neuronal viability was seen when low [Na + ] 0 or [Cl"]0 conditions were maintained during the 4h post-anoxic period. The substitution of choline for N a + produced an increase in neuronal viability equal to that produced by the substitution of gluconate for Cl". The acute neurotoxicity of E A A s has been postulated to be produced by passive Cl" influx, followed by an osmotic water entry and subsequent lysis (Rothman, 1985). Gluconate substitution for Cl" has been shown to inhibit neuronal swelling in a number of injury paradigms including stimulation-evoked depolarization in hippocampal slices (MacVicar and Hochman, 1991), as well as anoxia and oxygen-glucose deprivation in organotypic hippocampal cultures when combined with low [Ca 2 + ] 0 (Newell et al, 1995). Similar neuronal sparing has been demonstrated by removal or replacement of N a + 0 with impermeant ions in models of dithionite-induced anoxia using dissociated hippocampal (Friedman & Haddad, 1993; Friedman and Haddad, 1994b) or cultured cortical neurons (Chidekel etal., 1997), and iodoacetate exposure in cerebellar cultures (Verity etal, 1991). More recently, Cl"-mediated swelling has been dissociated from excitotoxic neuronal death in a model of NMDA-mediated excitotoxicity (Sakaguchi et al, 1999). However, this model does not take into account the metabolic disruptions that may occur as a result of anoxia or ischemia, which would likely exacerbate subsequent excitotoxic stimuli such as glutamate release. Previous research indicates that the glutamate concentration required to produce excitotoxicity in energy-compromised neurons is likely lower than that required to kill neurons in normal cultures (Dubinsky and Rothman, 1991; Novelli etal, 1988, Pang and Geddes, 1997). It should also be noted that post-anoxic low [Na + ] 0 or [Ci"]0 conditions, although significantly improving neuronal viability, still did not prevent a large portion of the neuronal death seen. 68 Although there was still some residual N a + (21.5 mM) and Cl" (7 mM) in the low [Na + ] 0 or [Cl"]0 buffers respectively, the portion of neuronal death which was not prevented by these altered ionic conditions was possibly due to other factors not investigated here. For example, activation of postsynaptic Group I mGluR could amplify the post-synaptic response to glutamate in a protein kinase C-dependent manner by phosphorylation of post-synaptic N M D A receptors, increasing their responsiveness to glutamate and thus enhancing toxicity (Bruno et al, 1995; Buisson & Choi, 1995). In these experiments, [Ca 2 + ] 0 was normal. But the severe neurotoxicity produced by omitting extracellular C a 2 + prevented the investigation of other Ca2 +-dependent processes which could be involved under these conditions. Ca 2 + and Anoxia-induced Neuronal Death Removal of External Ca2+ C a 2 + itself has been shown to be neurotoxic under conditions without excitotoxicity, and capable of inducing either apoptosis or necrosis in cultured cortical neurons (Gwag et al., 1999). Although our original hypothesis was that omission of C a 2 + 0 would improve neuronal viability following anoxia, we found that anoxia in nominally Ca 2 +-free B B S S did not provide neuroprotection, and that the maintenance of 0 [Ca 2 + ] 0 during the post anoxic period exacerbated anoxia-induced neuronal death. This phenomenon has been observed by other groups following simulated anoxia or ischemia. In a model of (Vglucose deprivation-mediated neuronal death, removal of C a 2 + from the exposure buffer increased neuronal death and enhanced acute neuronal swelling, although the latter could be mitigated by adding extracellular sucrose (Goldberg and Choi, 1993). This result is very similar to what was observed in our model. A similar result has been observed in a model of iodoacetate-mediated chemical anoxia (Verity et al., 1991), and a model ofNMDA-mediated excitotoxicity in hippocampal slice cultures (Sakaguchi etal, 1999), 69 where the removal of C a 2 + from the medium had no effect on control cultures or trypan blue exclusion, but exacerbated the onset and magnitude of neuronal injury. Interestingly, in both of these cases low Ca2+-potentiated neuronal swelling and death was sensitive to N M D A receptor antagonists (Verity et al, 1991; Goldberg and Choi, 1993). There are a number of possible explanations for this potentiation of neurotoxicity. Removal of extracellular C a 2 + may enhance N a + influx through N M D A receptors by reducing ionic interference or Ca2 +-dependent N M D A receptor desensitization, which would increase the sodium conductance (Goldberg and Choi, 1993). This hypothesis takes into account the continued effectiveness of N M D A receptor antagonists under low C a 2 + conditions. Swelling under low C a 2 + 0 conditions was also found to be dependent on [Na + ] 0 and [Cl"]0; increased g N a through N M D A receptors would produce a passive transfer of Cl" and water to produce acute swelling (Goldberg and Choi, 1993). The lack of external C a 2 + would also prevent reverse operation of the N a + / C a 2 + exchanger, which under conditions of high Na + ; can operate in reverse mode to extrude N a + (Carini et al, 1997; Hoyt et al, 1998). Low Ca 2 +-mediated swelling may occur by reduction of membrane mechanical stability or enhancement of neuronal excitability under combined conditions of anoxia and 0 [Ca 2 + ] 0 leading to an intracellular N a + load (Konnerth etal, 1986). Imaging of Ca2+i During Static Anoxia Our attempts to correlate anoxia-evoked changes in [Ca 2 + ]; produced puzzling results. Previous studies of anoxia-evoked changes in [Ca 2 +]i using dithionite demonstrated that 5 minutes of anoxia under perfusion conditions produced a transient increase in [Ca 2 +]j which, upon return to normoxia, returned to near baseline levels (Friedman and Haddad, 1993; Diarra et al, 1999). In our hands both perfusion and static anoxia elicited two distinct types of transient increases in intracellular calcium, large and small.,Unfortunately, since the majority of [Ca ]; 70 responses under static conditions were of the small type, it was difficult to establish whether MK-801 , which was effective at preventing acute neurotoxicity in our anoxia model, would have an effect on anoxia-evoked changes in [Ca 2 +]j. It should be noted that despite the inconsistency of Ca 2 +-responses to anoxia, our anoxia model was very consistent in terms of the degree of cell death produced. This suggests that anoxia-evoked increases in [Ca 2 + ] i may not be predictive of neuronal death, unlike a model of 02/glucose deprivation (Kusumoto et al., 1996). Despite this discrepancy, several previous studies have concluded that observed rises in [Ca 2 + ] i after in vitro anoxia/ischemia or excitotoxicity do not necessarily indicate or correlate with cellular death under conditions where oxidative metabolism is compromised (Dubinsky and Rothman, 1991; Friedman and Haddad, 1993; Stout etal., 1998). For example, increases in [Ca 2 + ]; were not always required to produce anoxia-induced injury in hippocampal neurons, since injury still occurred in the presence of Ca 2 +-channel blockers or 0 [Ca 2 + ] 0 , conditions where no anoxia-evoked increase in [Ca 2 + ] i were seen (Friedman and Haddad, 1993). Incubation of hippocampal neurons with CN" produced large elevations in [Ca 2 +]i, but parallel toxicity studies indicated that C N was not neurotoxic (Dubinsky and Rothman, 1991). Hence we do no believe that a lack of anoxia-evoked increase in [Ca 2 +]i under static anoxia conditions indicates a lack of neurotoxic effect. A n interesting observation made in our Ca 2 +-imaging experiments was that cultures 2+ which had an initial I334/I380 > 0.5 were much more likely to respond to anoxia with a large Ca transient than those cultures with an initial I334/I380 < 0.5. We speculate that the different types of C a 2 + responses could be a function of the membrane potential, with neurons that have a slightly depolarized membrane potential producing large C a 2 + transients in response to anoxia while other neurons which have a normal or hyperpolarized membrane potential show little or no response. Consistent with this hypothesis, CA1 pyramidal hippocampal neurons at a holding potential of-50 m V have a significantly higher [Ca 2 + ] i than neurons held at -80 mV, and 71 prolonged membrane hyperpolarization produced sustained decreases in somatic [Ca 2 + ]; in a voltage-dependent fashion (Magee et al, 1996). Similar findings were demonstrated using embryonic hippocampal neurons, with a strong correlation found between resting [Ca 2 +]i and the maximal changes in [Ca 2 +]i evoked by either N M D A or high [K + ] Q application; the higher the resting [Ca 2 +]i, the higher the depolarization-evoked change in [Ca 2 +]; (Nakajima et al, 1993). Toxicity of N M D A Receptor Redox Modulators Since dithionite is a reducing agent and the gating kinetic properties of the N M D A receptor can be modified by agents either oxidizing or reducing the redox site on the receptor (Choi and Lipton, 2000), we decided to compare the effects of dithionite-induced anoxia to reduction of the N M D A receptor using the reducing agent D T T , which has been shown to increase the frequency of N M D A receptor openings in response to agonist binding (Aizenman et al, 1989; Reynolds etal, 1990; Tang and Aizenman, 1993). We hypothesized that D T T would not have an effect on neuronal viability by itself, and intended it for use as a control for redox modulation independent of toxic stimuli. D T T incubation proved very toxic to our cultures, significantly more toxic than dithionite-induced anoxia. Previous research on D T T enhancement of N M D A receptor-mediated neurotoxicity has indicated that it potentiates toxicity of N M D A receptor agonists in neuronal cultures (Aizenman et al, 1989; Levy et al, 1990; Pringle et al, 2000). The concentration of D T T used in our experiments, 2 m M , was used by both Levy et al, (1990) and Pringle et al, (2000) who found it to be non-toxic in their preparations (cultured rat retinal ganglion cells and organotypic hippocampal slice cultures, respectively) when present for the same duration as in the experiments presented in this study. Preliminary experiments in our lab comparing neuronal death produced by (Vglucose deprivation to anoxia-induced neuronal 72 death showed that both methods could produce similar levels of neurotoxicity; hence redox modulation is not required to achieve the levels of neuronal death seen in our model. Although D T T did not produce an observable rapid decrease in p02 measured with an oxygen sensor (data not shown), it has been reported to act as an oxygen scavenger (Hatch, 1993). It should be noted that our results do not indicate the mechanism of D T T toxicity, and so non-specific effects cannot be ruled out. However, our results also do not provide an indication that dithionite alters the redox state of the N M D A receptor, and so further experiments are necessary to examine this possibility. Comparison of Dithionite-induced Model of Anoxia to Other Current In vitro Models O2 deprivation Newell et al. (1995) reported that O2 deprivation (produced by incubation in a buffer made anoxic by gassing with 95 % N2/ 5% CO2 within an anoxic chamber) in organotypic hippocampal cultures resulted in toxicity after 48h that was not prevented by N M D A or A M P A receptor antagonists, but attenuated by P B N . However, others have found that exposing neuronal cultures to prolonged periods of anoxia produced a neurotoxic effect which could be attenuated by the use of N M D A receptor antagonists (Goldberg et al., 1987; Dessi et al., 1992; Rogers and Hunter, 1997). Rogers and Hunter (1997) also noted that C N Q X (100 LLM) was as effective as MK-801 at reducing neuronal death, and the putative glutamate release inhibitor riluzole was also found to prevent anoxic injury in cerebellar granule cell cultures (Dessi et al., 1993a), further implicating glutamate in anoxia-mediated cell injury. A particular drawback to existing in vitro anoxia/hypoxia models is the duration of insult required to produce neuronal injury (Lipton, 1999). When cultures are maintained in their growth medium, anoxic durations of 14-24h in an atmosphere of 95% N2 / 5% CO2 are required to produce widespread neuronal death (Rothman, 1983; Dessi et al, 1992; Dessi et al, 1993a; Rogers and Hunter, 1997). This duration of insult is not only less convenient to use in the laboratory, but is also physiologically inapplicable. The long insult duration required also prevents the separation of acute and delayed cell death processes, limiting the utility of these models. 02/Glucose Deprivation: The omission of glucose from the extracellular environment during anoxia, intended to simulate ischemia, greatly reduces the duration of exposure required to produce neurotoxicity in vitro. This has been demonstrated by Goldberg and Choi (1993), who showed that the duration of anoxia required to produce half-maximal neuronal death in vitro decreases as the concentration of glucose in the extracellular buffer decreases. The neurotoxicity produced by (Vglucose deprivation shares some similarities with our model. For example, the severity of neuronal injury increases with the duration of insult (Goldberg and Choi, 1993; Kusumoto etal, 1996) and, as in our model, the N M D A receptor appears to mediate the majority of neuronal death. For example, (Vglucose deprivation in organotypic hippocampal cultures produced neuronal injury that was sensitive to blockade by N M D A receptor antagonists (Newell et al, 1995; Pringle et al, 2000). Evidence for both N M D A and A M P A receptor involvement was provided by Gwag et al. (1995), who showed that 50 min of (Vglucose deprivation produced necrotic injury in murine cortical cultures which could be attenuated using either MK-801 or C N Q X . Goldberg and Choi (1993) had earlier found that (Vglucose deprivation produced injury featuring acute neuronal swelling and later neuronal degeneration (which had features of necrotic cell death) which could be blocked by G P T , MK-801, dextrorphan, A P V , or 7-CK, while C N Q X (30 u.M) or T T X (10 L I M ) were not protective when present during the insult period. However, AMPA-receptor mediated-toxicity was not ruled out by these results as A M P A receptors have been reported to mediate a slower excitotoxicity than N M D A receptors, and hence A M P A receptor-mediated toxicity could be masked (Goldberg and Choi, 1993). Once again, these details are consistent with our findings using dithionite to induce anoxia. Indeed, our results indicate that A M P A receptors can contribute to anoxia-induced neurotoxicity since C N Q X administered post-anoxia was significantly neuroprotective. Despite these similarities, 02/glucose deprivation still has the disadvantage of requiring prolonged durations of insult to produce injury. Cultured neurons need to be incubated for at least 30-40 minutes in a glucose-free buffer in an oxygen-free atmosphere to produce any significant neuronal death during the subsequent 24h period (Goldberg and Choi, 1993; Grabb and Choi, 1999; Gwag et al, 1995). The one exception to this longer duration is 02/glucose deprivation mediated by argon, but even then a minimum 20 min insult is required to produce significant neuronal death (Kusumoto et al, 1996). Hence our model offers the same or greater capabilities of producing either an acute or delayed neuronal death, mediated by the glutamate binding to the same receptors, while still retaining the advantage of relatively short insult duration and time to development of death. Chemical Anoxia Iodoacetate compares favourably to other in vitro models in that comparatively short incubation times (5 min) are sufficient to produce significant neuronal death (-40%) in hippocampal cultures by 1 h after insult, with a delayed necrotic component which reduced neuronal viability to under 10 % after 24 h (Uto et al, 1995). Toxic effects of iodoacetate include a depletion of A T P levels (due to inhibition of glyceraldehyde dehydrogenase) and an inhibition of protein synthesis preceding neuronal death (Uto etal, 1995) caused by disaggregation of ribosomes (Dux et al, 1996). The mode of neuronal death caused by iodoacetate was concluded to be necrotic (Uto et al, 1995), which was confirmed in another 75 study of iodoacetate-induced neurotoxicity as there were no signs of apoptosis determined by electron microscopy (Dux etal, 1996). Removal of [Ca 2 + ] 0 or addition of MK-801 did not prevent iodoacetate-induced neurotoxicity, an observation which agrees with a previous study in which iodoacetate-mediated neuronal injury could not be prevented by A P V or kynurenate (Verity et al, 1991) . Iodoacetate inactivates thiol-containing enzymes that play an important role in scavenging free radicals. In fact, antioxidant treatment provided significant neuroprotection against iodoacetate toxicity, suggesting the involvement of free radicals as the cause of neurotoxicity (Uto etal, 1995). These findings indicate clear differences in the pharmacology and mechanisms of iodoacetate- and anoxia-induced neurotoxicity. Incubation of cultures with CN" has also been used as an assay for chemical anoxia. It has been shown that N M D A receptors mediate cell injury in hippocampal neurons exposed to CN"-induced chemical anoxia (Patel et al, 1993), reinforcing earlier findings that CN" induced the release of endogenous glutamate in mouse brain slices (Patel et al, 1991). Additionally, Dubinsky and Rothman (1991) showed that both MK-801 and enzymatic glutamate degradation were neuroprotective in cultured hippocampal neurons after exposure to CN", indicating that C N -induced toxicity is mediated by N M D A receptors, a finding confirmed by others (Patel et al, 1993). CN" has also been shown to potentiate N M D A receptor function by modifying the receptor redox site on N R 1 / N R 2 A recombinant receptors (Arden et al, 1998). However, the use of CN" in models of chemical anoxia also has its drawbacks. The main shortcoming of this method of chemical anoxia is the duration of insult required to produce appreciable neuronal death. Patel et al, (1993) found that even after 18 hours of constant exposure to 2 m M CN", the level of neuronal injury achieved was only approximately half-maximal. Also using a concentration of 10 m M C N (for 10 min), Uto et al, (1995) were only able to produce only partial neuronal injury when measured after 24h. Similarly, Dubinsky and Rothman (1991) reported that cultures exposed to 3 m M CN" for 30 min still had mean neuronal 76 viabilities > 75 % after 18 hours. Although an earlier study concluded that CN" produced significant neurotoxicity in hippocampal neurons (Rothman, 1983), the level of neuronal death was later attributed to relatively high glutamate and aspartate concentrations used in the culture media (Dubinsky & Rothman, 1991). Another problem with using CN" is that at physiological pH, CN" is almost exclusively in its free acid form which evaporates quickly, thus leading to difficulties controlling its concentration (Varming et al, 1996); a problem exacerbated by the long duration of exposure required. Azide is another inhibitor of mitochondrial function commonly used to model anoxia or ischemia-induced neuronal death. Although the main toxic effect of azide is thought to be via inhibition of cytochrome a3 and thus A T P synthesis (Vasilyeva et al., 1982; Noumi et al., 1987), it has been reported to affect D N A synthesis (Ciesla et al, 1974) as well as superoxide dismutase function (Misra and Fridovich, 1978). Smith et al, (1991) also suggested that azide, through a conversion to nitric oxide, may enhance glutamatergic synaptic transmission and thereby produce or enhance neurotoxicity. Incubation with azide produced a reversible increase in [ C a 2 + ] i in mouse neocortical neurons, which was sensitive to both N M D A and n o n - N M D A receptor antagonists (Jorgensen et al, 1999). MK-801 provided a significant degree of protection from azide-mediated neurotoxicity, as did T T X under non-depolarizing conditions (Varming et al, 1996). These findings are similar to our model, as 1 u M T T X was found to be protective only when present during the post-anoxic period. Azide toxicity is similar to that induced by CN" exposure in that it targets A T P generation, and incubation of neurons with azide produces an increase in [Ca 2 +]i similar to that seen during incubations with CN" (Dubinsky and Rothman, 1991). However, azide has one particular advantage over CN" in that it is not volatile in solution and thus a steady concentration can be maintained (Regan and Guo, 1999). Unfortunately, azide is also similar to CN" in that a long duration of exposure is required to produce toxicity in neurons. Regan & Guo (1999) 77 reported that a 30 min incubation with azide in glucose-free buffer caused a release of only 10% of total L D H in neuronal cultures after 21-25 hours. Varming et al. (1996) found that 60 min incubations of cerebellar cultures with 100 m M azide reduced neuronal viability to 30 % of control assessed by M T T reduction. Although this represents a significant neurotoxic effect, the duration of insult is still not translatable to the in vivo situation. N P A exposure produced two distinct types of cell death (Pang and Geddes, 1997; Weih et al., 1999). The first resembled an acute necrosis with cell swelling and nuclear shrinkage, and was completely prevented by MK-801 (10 L I M ) . Notably, cell swelling was visible within 1 hour of insult. The second mechanism of cell death appeared to be a delayed apoptosis, as it featured nuclear fragmentation and could be attenuated by including the protein synthesis inhibitor cycloheximide (Pang and Geddes, 1997). Additionally, Pang and Geddes (1997) reported that MK-801 could not attenuate the delayed component of neuronal death following insult, a finding we have observed in preliminary experiments in our model (data not shown). Hence the N P A model of chemical anoxia seems to produce, in hippocampal cultures, very similar effects to the model presented here. As with the other inhibitors of mitochondrial function, impractically long durations of exposure are also the main shortcoming of using N P A as an inducer of chemical anoxia. Fukuda et al. (1998) reported that a 40 minute exposure of cortical or striatal cultures to 1.7 m M N P A was toxic to only 4% of neurons. A 48h exposure of hippocampal neurons to N P A produced neuronal death in a dose-dependent fashion, but at the highest dose tested (15 mM) neuronal viability was still over 20 % (Pang and Geddes, 1997). At lower concentrations, viability was approximately 40 % (5 mM) and 55 % (2 mM). There are several similarities between our model of dithionite-mediated anoxia-induced neuronal death and other current methods of chemical anoxia. Like C N - , azide-, and N P A -mediated chemical anoxia, neuronal death in our model is mediated via glutamate activating 78 N M D A receptors. Predictably, the morphology of the subsequent neuronal death appears similar. In addition, NPA-mediated insult can produce both acute and delayed neuronal death, a characteristic of the anoxia model presented here. However, the short insult duration required to produce neuronal death and rapid development of neurotoxicity in our model represent distinct advantages over each of the aforementioned methods of chemical anoxia. Additionally, the model described here produces a high degree of neurotoxicity, a quality which helps differentiate successful neuroprotective treatments with regard to their effectiveness. Given the differences in iodoacetate-induced neuronal death and the features of cell death produced by our model, it would appear that different toxic processes are occurring and thus the models may not be directly comparable aside from their relatively short duration of insult and rapid development of neurotoxicity. The use of mitochondrial inhibitors to model anoxia may be problematic in that generation of oxygen free radicals probably occurs at a much higher rate using chemical anoxia than under anoxic conditions (Lipton, 1999). In fact, cellular toxicity produced by mitochondrial inhibitors is decreased under anoxic conditions, implicating oxygen free radical generation as a toxic consequence of mitochondrial inhibition, particularly when using iodoacetate or CN" (Borle and Barsic, 1995; Dawson et al., 1993). As such, the use of mitochondrial inhibitors may be appropriate for the study of free radical-mediated cellular damage, but not be suitable as a complete model for anoxia or ischemia-induced neuronal death. Conclusions From this study we conclude that 5 min of anoxia induced by dithionite produces severe neurotoxicity in postnatal hippocampal neurons. Induction of neuronal death by this method is rapid, reliable and consistently produces at least 60% neuronal death by 4h. This neuronal death is produced by the activation of N M D A and A M P A receptors by glutamate, and can be acutely 79 prevented through the use of competitive and non-competitive glutamate receptor antagonists, extracellular glutamate metabolism, and post-anoxic washing or buffer change. There are two phases of neuronal death. The first is an acute phase which occurs within 4h and morphologically resembles necrosis. Substitution of external N a + and Cl" with impermeant ions can mitigate a portion of this acute neuronal death, however removal of external C a 2 + increases neuronal susceptibility to anoxia. Free radical-mediated damage does not appear to play a role in this acute phase. The second phase is a delayed neuronal death, unmasked by measures which prevent acute neuronal death, occurring over a period of 24h. Limitations of This Study There are several general limitations on the use of in vitro techniques to study in vivo phenomena, which can be applied to the field of anoxia/ischemia. Most in vitro methods differ from the in vivo situation in that the duration of insult to induce cell death is generally much longer in vitro; this property is seen not only in primary cultures (Dessi et al, 1992; Dessi et al., 1993a; Goldberg and Choi, 1993; Gwag et al, 1995; Pang and Geddes, 1997; Regan and Guo, 1999), but also in organotypic cultures (Strasser & Fischer, 1995). The method described here does not suffer from this particular shortcoming. However, the lack of a central immune response and inflammation, altered anatomical relationships between cells compared to the in vivo situation (for cell cultures and acutely dissociated cells), and the absence of blood flow and coordinated extracellular p H responses suggest that extrapolation of in vitro results, including those presented here, to an in vivo situation should be done with caution. In addition the differences in relative vulnerabilities of different types of neurons in vivo is not always reproduced in the in vitro situation. While in vivo hippocampal neurons are selectively vulnerable to ischemic insult (Kirino and Sano, 1984; Kirino et al, 1985), the differential 80 sensitivity (notably between hippocampal and cortical neurons) to these insults may not be seen in some cases in vitro (Dux et al, 1996; Kusumoto et al, 1996), while it is maintained in others (Uto et al, 1995). We have not addressed this issue with respect to the dithionite-anoxia model of cell death in the postnatal hippocampal neuronal preparations used in the present study. Future Directions - Proposed Experiments There are several experiments which could be performed to address some of the questions raised by the results presented in this study. 1) Source of Anoxia-induced Glutamate Release Having established that anoxia-induced neuronal death is mediated by glutamate, it would be useful to determine the source of glutamate. Ischemia produces an elevation in extracellular glutamate via both calcium-dependent and -independent mechanisms (Szatkowski and Attwell, 1994), a phenomenon confirmed by others (Patel etal, 1991). In the hippocampus, glutamate release in the hippocampal slice occurred predominantly via reversed uptake of glutamate by neurons under conditions of simulated ischemia (Roettger and Lipton, 1996; Rossi et al, 2000). Hence, inhibitors of glutamate uptake transporters may have utility in determining the source of anoxia-induced glutamate release. Use of these compounds is complicated by the fact that they will exacerbate glutamate-mediated excitotoxicity if release occurs via another mechanism, since glutamate uptake mechanisms will be compromised. A n alternate approach to determining the source of glutamate release would be to use inhibitors of the presynaptic Ca 2 +-channels in hippocampal neurons to determine if Ca 2 +-dependent exocytotic release can occur for a fraction of glutamate release. Several different voltage-gated calcium channel subtypes (N-, P- & Q-type channels) appear to be involved in the presynaptic 81 release of glutamate in hippocampal neurons (Piser et al, 1995b). One snail toxin, co-conotoxin M V I I C , blocks N - , P- & Q-type channels (Hillyard et al., 1992; Randall and Tsien, 1995), and has been shown to block glutamatergic synaptic transmission in the hippocampus (Wheeler et al., 1994). Alternatively, a different toxin derived from the venom of the tarantula spider Grammostola spatulata, co-grammatoxin SLA, has been identified and shown to inhibit N - , P- & Q-type channels (Lampe etal., 1993), completely inhibiting glutamatergic synaptic transmission in cultured rat hippocampal neurons in a reversible manner (Piser et al., 1995a). L-type calcium channels do not appear to be involved in glutamatergic synaptic transmission as the dihydropyridine antagonist nitrendipine has no effect on glutamatergic EPSCs recorded in neurons from hippocampal cultures (Piser et al., 1995b). If exocytotic synaptic release is responsible for all or part of the anoxia-evoked glutamate release, we then expect that treatment with these toxins will improve post-anoxic neuronal viability. 2) Mechanism of Delayed Neuronal Death Following Perfusion/Wash Anoxia To characterize the delayed cell death which can be produced under perfusion conditions, the use of DNA-staining dyes can be used to help determine whether the type of cell death occurring is apoptotic or necrotic in nature. For example, the dye Hoechst 33258 can be used to stain neurons as outlined by (Saudou etal, 1998); alternatively, D A P I staining can be used in the same fashion (Nardi et al, 1997). Neurons can be presumed apoptotic based on the appearance of the stained nuclei under fluorescence; punctate staining and/or fragmentation of the nucleus, as well as degenerated or absent processes indicate apoptotic cells (Sadou et al, 1998). The frequency of apoptotic nuclei could be compared between cultures which undergo acute cell death and those which experience a delay prior to death to determine whether apoptosis makes up an increased component of delayed cell death. 82 If the experiment above indicates that there may be a portion of delayed cell death mediated by apoptosis then another measure to confirm apoptosis, preferably an interventional one such as the use of caspase inhibitors, will be used to assess whether apoptosis can be blocked or reduced. 3) Determination of [glu] by H P L C The measurement of glutamate in extracellular media following simulated anoxia or ischemia can be used to determine when glutamate release is occurring during and following anoxia, and the relative amounts being released. Although the final concentration of extracellular glutamate in neuronal culture medium after either 02/glucose deprivation (Goldberg and Choi, 1993) or N P A exposure (Pang and Geddes, 1997) is not considered high enough to be toxic, the concentration of glutamate in the extracellular medium is probably not a true reflection of the levels activating local postsynaptic glutamate receptors (Goldberg and Choi, 1993). Additionally, previous research indicates that the glutamate concentration required to produce excitotoxicity in metabolically-compromised neurons is likely lower than that required to kill neurons in normal cultures (Dubinsky and Rothman, 1991; Novelli et al., 1988; Pang and Geddes, 1997). A small increase in extracellular glutamate could be a significant finding given that a) the dilution of glutamate in extracellular buffer, and b) neurons in our model are presumably energetically compromised. APPENDIX 1 - Composition of Buffers & Media Used 83 *Note: A l l buffers and solutions were made using sterile distilled water. BSS - contains (mM): NaCl 139, KC1 3.5, N a 2 H P 0 4 3, N a H C 0 3 2, H E P E S acid 6.7, H E P E S - N a 3.3, D-Glucose 5, CaCl 2 1.8, M g C l 2 0.8; together with 2 u M glycine and 0.05% bovine serum albumin, p H 7.35. BBSS - contains (mM): NaCl 127, KC1 3, N a H C 0 3 19.5, N a H 2 P 0 4 H 2 0 1.5, M g S G y 7 H 2 0 1.5, C a C l 2 2, glycine 0.01, glucose 5. DMEF - contains: 29 m M N a H C 0 3 , 12 mg / ml D M E / F i 2 powder (Gibco), 30 Lig/ml penicillin, 25 Lig/ml streptomycin, 10% fetal bovine serum. HBSS - contains (mM): NaCl 137, KC1 5.4, K H 2 P O 4 0 . 4 , N a 2 H P 0 4 0.4, H E P E S 5, glucose 5.6, pyruvate 1. NBM - liquid lx Neurobasal medium (Gibco) supplemented with 30 Lig/ml penicillin, 25 Lig/ml streptomycin and lx B27 supplement. 84 REFERENCES Abdel-Hamid K M , Baimbridge K G (1997) The effects of artificial calcium buffers on calcium responses and glutamate-mediated excitotoxicity in cultured hippocampal neurons. Neuroscience 81: 673-687. Aizenman E , Lipton SA, Loring R H (1989) Selective modulation of N M D A responses by reduction and oxidation. Neuron 2: 1257-1263. Ankarcrona M , Dypbukt J M , Bonfoco E , Zhivotovsky B, Orrenius S, Lipton SA, Nicotera P (1995) Glutamate-induced neuronal death: a succession of necrosis or apoptosis depending on mitochondrial function. Neuron 15: 961-973. Archer SL, Hampl V , Nelson DP, Sidney E , Peterson D A , Weir E K (1995) Dithionite increases radical formation and decreases vasoconstriction in the lung. Evidence that dithionite does not mimic alveolar hypoxia. Circ Res 77: 174-181. Arden SR, Sinor JD, Potthoff W K , Aizenman E (1998) Subunit-specific interactions of cyanide with the N-methyl-D-aspartate receptor. J Biol Chem 273: 21505-21511. Atlante A , Gagliardi S, Minervini G M , Ciotti M T , Marra E , Calissano P (1997) Glutamate neurotoxicity in rat cerebellar granule cells: a major role for xanthine oxidase in oxygen radical formation. JNeurochem 68: 2038-2045. Barbour B , Szatkowski M , Ingledew N , Attwell D (1989) Arachidonic acid induces a prolonged inhibition of glutamate uptake into glial cells. Nature 342: 918-920. Bekkers J M , Stevens C F (1989) N M D A and n o n - N M D A receptors are co-localized at individual excitatory synapses in cultured rat hippocampus. Nature 341: 230-233. Benveniste H , Jorgensen M B , Sandberg M , Christensen T, Hagberg H , Diemer N H (1989) Ischemic damage in hippocampal CA1 is dependent on glutamate release and intact innervation from C A 3 . J Cereb Blood Flow Metab 9: 629-639. Blitzblau R, Gupta S, Djali S, Robinson M B , Rosenberg P A (1996) The glutamate transport inhibitor L-trans-pyrrolidine-2,4-dicarboxylate indirectly evokes N M D A receptor mediated neurotoxicity in rat cortical cultures. Eur JNeurosci 8: 1840-1852. Boeckman F A , Aizenman E (1994) Stable transfection of the NR1 subunit in Chinese hamster ovary cells fails to produce a functional N-methyl-D-aspartate receptor. Neurosci Lett 173: 189-192. 85 Bond J M , Chacon E , Herman B, Lemasters JJ (1993) Intracellular p H and Ca homeostasis in the p H paradox of reperfusion injury to neonatal rat cardiac myocytes. A m J Physiol 265: C129-C137. Borle A B , Barsic M (1995) Chemical hypoxia increases cytosolic Ca and oxygen free radical formation. Cell Calcium 17: 307-315. Braughler J M , Hall E D (1992) Involvement of lipid peroxidation in C N S injury. J Neurotrauma 9 Suppl 1: S1-S7. Bredt DS, Snyder S H (1990) Isolation of nitric oxide synthetase, a calmodulin-requiring enzyme. Proc Natl Acad Sci U S A 87: 682-685. Brorson JR, Schumacker PT, Zhang H (1999) Nitric oxide acutely inhibits neuronal energy production. The Committees on Neurobiology and Cell Physiology. J Neurosci 19: 147-158. Bruno V , Battaglia G , Copani A , Casabona G, Storto M D i G G , V , Ngomba R, Nicoletti F (1998) Metabotropic glutamate receptors and neurodegeneration. Prog Brain Res 116: 209-221. Bruno V , Copani A , Knopfel T, Kuhn R, Casabona G, Dell'Albani P, Condorelli D F , Nicoletti F (1995) Activation of metabotropic glutamate receptors coupled to inositol phospholipid hydrolysis amplifies NMDA-induced neuronal degeneration in cultured cortical cells. Neuropharmacology 34: 1089-1098. Buisson A , Choi D W (1995) The inhibitory mGluR agonist, S-4-carboxy-3-hydroxy-phenylglycine selectively attenuates N M D A neurotoxicity and oxygen-glucose deprivation-induced neuronal death. Neuropharmacology 34: 1081-1087. Buller A L , Larson H C , Schneider B E , Beaton JA, Morrisett R A , Monaghan D T (1994) The molecular basis of N M D A receptor subtypes: native receptor diversity is predicted by subunit composition. J Neurosci 14: 5471-5484. Calabresi P, Marfia G A , Centonze D, Pisani A , Bernardi G (1999) Sodium influx plays a major role in the membrane depolarization induced by oxygen and glucose deprivation in rat striatal spiny neurons. Stroke 30: 171-179. Carini R, de Cesaris M G , Bellomo G, Albano E (1997) Role of N a + / C a 2 + exchanger in preventing N a + overload and hepatocyte injury: opposite effects of extracellular and intracellular C a 2 + chelation. Biochem Biophys Res Commun 232: 107-110. 86 Carpenter E , Hatton CJ, Peers C (2000) Effects of hypoxia and dithionite on catecholamine release from isolated type I cells of the rat carotid body. J Physiol 523 Pt 3: 719-729. Chan P H , Fishman R A (1980) Transient formation of superoxide radicals in polyunsaturated fatty acid-induced brain swelling. JNeurochem 35: 1004-1007. Chan P H , Kerlan R, Fishman R A (1983) Reductions of gamma-aminobutyric acid and glutamate uptake and (Na + + K + ) -ATPase activity in brain slices and synaptosomes by arachidonic acid. J Neurochem 40: 309-316. Chazot P L , Cik M , Stephenson F A (1992) Immunological detection of the N M D A R 1 glutamate receptor subunit expressed in embryonic kidney 293 cells and in rat brain. JNeurochem 59: 1176-1178. Chen J, Nagayama T, Jin K , Stetler R A , Zhu R L , Graham SH, Simon RP (1998) Induction of caspase-3-like protease may mediate delayed neuronal death in the hippocampus after transient cerebral ischemia. J Neurosci 18: 4914-4928. Chen L , Huang L Y (1992) Protein kinase C reduces M g 2 + block of NMDA-receptor channels as a mechanism of modulation. Nature 356: 521-523. Chen W H , Chu K C , W u SJ, W u JC, Shui H A , Wu M L (1999) Early metabolic inhibition-induced intracellular sodium and calcium increase in rat cerebellar granule cells. J Physiol (Lond) 515 (Pt 1): 133-146. Chidekel AS , Friedman JE, Haddad G G (1997) Anoxia-induced neuronal injury: role of N a + entry and Na+-dependent transport. Exp Neurol 146: 403-413. Choi D W (1987) Ionic dependence of glutamate neurotoxicity. J Neurosci 7: 369-379. Choi D W (1988) Calcium-mediated neurotoxicity: relationship to specific channel types and role in ischemic damage. Trends Neurosci 11: 465-469. Choi D W (1995) Calcium: still center-stage in hypoxic-ischemic neuronal death. Trends Neurosci 18: 58-60. Choi D W (1996) Ischemia-induced neuronal apoptosis. Curr Opin Neurobiol 6: 667-672. Choi Y B , Lipton S A (2000) Redox modulation of the N M D A receptor. Cell M o l Life Sci 57: 1535-1541. 87 Chopp M , L i Y (1996) Apoptosis in focal cerebral ischemia. Acta Neurochir Suppl (Wien ) 66: 21-26. Ciesla Z , Mardarowicz K , Klopotowski T (1974) Inhibition of D N A synthesis and cell division in Salmonella typhimurium by azide. M o l Gen Genet 135: 339-348. Clark G D , Rothman S M (1987) Blockade of excitatory amino acid receptors protects anoxic hippocampal slices. Neuroscience 21: 665-671. Collingridge G L , Lester R A (1989) Excitatory amino acid receptors in the vertebrate central nervous system. Pharmacol Rev 41: 143-210. Collins RJ, Harmon B V , Gobe G C , Kerr JF (1992) Internucleosomal D N A cleavage should not be the sole criterion for identifying apoptosis. Int J Radiat Biol 61: 451-453. Cummins TR, Donnelly D F , Haddad G G (1991) Effect of metabolic inhibition on the excitability of isolated hippocampal CA1 neurons: developmental aspects. J Neurophysiol 66: 1471-1482. Darzynkiewicz Z , Juan G, L i X , Gorczyca W, Murakami T, Traganos F (1997) Cytometry in cell necrobiology: analysis of apoptosis and accidental cell death (necrosis). Cytometry 27: 1-20. Dawson V L , Dawson T M , Bartley D A , Uhl GR, Snyder S H (1993) Mechanisms of nitric oxide-mediated neurotoxicity in primary brain cultures. JNeurosci 13: 2651-2661. Dawson V L , Dawson T M , London E D , Bredt DS, Snyder S H (1991) Nitric oxide mediates glutamate neurotoxicity in primary cortical cultures. Proc Natl Acad Sci U S A 88: 6368-6371. Dawson V L , Kizushi V M , Huang P L , Snyder SH, Dawson T M (1996) Resistance to neurotoxicity in cortical cultures from neuronal nitric oxide synthase-deficient mice. J Neurosci 16: 2479-2487. Desagher S, Glowinski J, Premont J (1997) Pyruvate protects neurons against hydrogen peroxide-induced toxicity. JNeurosci 17: 9060-9067. Dessi F, Ben Ari Y , Charriaut-Marlangue C (1993a) Riluzole prevents anoxic injury in cultured cerebellar granule neurons. Eur J Pharmacol 250: 325-328. Dessi F, Charriaut-Marlangue C, Ben Ari Y (1992) Anisomycin and cycloheximide protect cerebellar neurons in culture from anoxia. Brain Res 581: 323-326. 88 Dessi F, Charriaut-Marlangue C, Khrestchatisky M , Ben Ari Y (1993b) Glutamate-induced neuronal death is not a programmed cell death in cerebellar culture. J Neurochem 60: 1953-1955. Diarra A , Sheldon C, Brett C L , Baimbridge K G , Church J (1999) Anoxia-evoked intracellular p H and C a 2 + concentration changes in cultured postnatal rat hippocampal neurons. Neuroscience 93: 1003-1016. Dingledine R, Borges K , Bowie D , Traynelis SF (1999) The glutamate receptor ion channels. Pharmacol Rev 51: 7-61. Doble A (1999) The role of excitotoxicity in neurodegenerative disease: implications for therapy. Pharmacol Ther 81: 163-221. D u C, H u R, Csernansky C A , Hsu C Y , Choi D W (1996) Very delayed infarction after mild focal cerebral ischemia: a role for apoptosis? J Cereb Blood Flow Metab 16: 195-201. Dubinsky J M , Rothman S M (1991) Intracellular calcium concentrations during "chemical hypoxia" and excitotoxic neuronal injury. J Neurosci 11: 2545-2551. Dugan L L , Sensi SL, Canzoniero L M , Handran SD, Rothman S M , L in TS, Goldberg M P , Choi D W (1995) Mitochondrial production of reactive oxygen species in cortical neurons following exposure to N-methyl-D-aspartate. J Neurosci 15: 6377-6388. Dumuis A , Pin JP, Oomagari K , Sebben M , Bockaert J (1990) Arachidonic acid released from striatal neurons by joint stimulation of ionotropic and metabotropic quisqualate receptors. Nature 347: 182-184. Dumuis A , Sebben M , Haynes L , Pin JP, Bockaert J (1988) N M D A receptors activate the arachidonic acid cascade system in striatal neurons. Nature 336: 68-70. Durand G M , Gregor P, Zheng X , Bennett M V , Uhl GR, Zukin RS (1992) Cloning of an apparent splice variant of the rat N-methyl-D-aspartate receptor N M D A R I with altered sensitivity to polyamines and activators of protein kinase C. Proc Natl Acad Sci U S A 89: 9359-9363. Dux E , Oschlies U , Uto A , Kusumoto M , Hossmann K A (1996) Early ultrastructural changes after brief histotoxic hypoxia in cultured cortical and hippocampal CA1 neurons. Acta Neuropathol (Berl) 92: 541-544. Dykens J A (1994) Isolated cerebral and cerebellar mitochondria produce free radicals when exposed to elevated C a 2 + and Na + : implications for neurodegeneration. J Neurochem 63: 584-591. 89 Eliasson M J , Huang Z , Ferrante RJ, Sasamata M , Molliver M E , Snyder SH, Moskowitz M A (1999) Neuronal nitric oxide synthase activation and peroxynitrite formation in ischemic stroke linked to neural damage. JNeurosci 19: 5910-5918. Erecinska M , Silver IA (1996) Calcium handling by hippocampal neurons under physiologic and pathologic conditions. Adv Neurol 71: 119-136. Fletcher E J , Lodge D (1996) New developments in the molecular pharmacology of alpha-amino-3-hydroxy- 5-methyl-4-isoxazole propionate and kainate receptors. Pharmacol Ther 70: 65-89. Friedman JE, Haddad G G (1993) Major differences in C a 2 + ; response to anoxia between neonatal and adult rat CA1 neurons: role of C a 2 + 0 and N a + 0 . J Neurosci 13: 63-72. Friedman JE, Haddad G G (1994a) Anoxia induces an increase in intracellular sodium in rat central neurons in vitro. Brain Res 663: 329-334. Friedman JE, Haddad G G (1994b) Removal of extracellular sodium prevents anoxia-induced injury in freshly dissociated rat CA1 hippocampal neurons. Brain Res 641: 57-64. Fukuda A , Deshpande SB, Shimano Y , Nishino H (1998) Astrocytes are more vulnerable than neurons to cellular Ca2+ overload induced by a mitochondrial toxin, 3-nitropropionic acid. Neuroscience 87: 497-507. Gebhardt C , Heinemann U (1999) Anoxic decrease in potassium outward currents of hippocampal cultured neurons in absence and presence of dithionite. Brain Res 837: 270-276. Giffard R G , Monyer H , Christine C W , Choi D W (1990) Acidosis reduces N M D A receptor activation, glutamate neurotoxicity, and oxygen-glucose deprivation neuronal injury in cortical cultures. Brain Res 506: 339-342. Goldberg M P , Choi D W (1993) Combined oxygen and glucose deprivation in cortical cell culture: calcium-dependent and calcium-independent mechanisms of neuronal injury. J Neurosci 13: 3510-3524. Goldberg M P , Monyer H , Choi D W (1988) Hypoxic neuronal injury in vitro depends on extracellular glutamine. Neurosci Lett 94: 52-57. Goldberg M P , Weiss JH, Pham P C , Choi D W (1987) N-methyl-D-aspartate receptors mediate hypoxic neuronal injury in cortical culture. J Pharmacol Exp Ther 243: 784-791. 90 Gonzalez-Zulueta M , Ensz L M , Mukhina G, Lebovitz R M , Zwacka R M , Engelhardt JF, Oberley L W , Dawson V L , Dawson T M (1998) Manganese superoxide dismutase protects nNOS neurons from N M D A and nitric oxide-mediated neurotoxicity. J Neurosci 18: 2040-2055. Gorovits R, Avidan N , Avisar N , Shaked I, Vardimon L (1997) Glutamine synthetase protects against neuronal degeneration in injured retinal tissue. Proc Natl Acad Sci U S A 94: 7024-7029. Gorter JA, Petrozzino JJ, Aronica E M , Rosenbaum D M , Opitz T, Bennett M V , Connor JA, Zukin RS (1997) Global ischemia induces downregulation of Glur2 m R N A and increases A M P A receptor-mediated C a 2 + influx in hippocampal CA1 neurons of gerbil. JNeurosci 17: 6179-6188. Goto K , Ishige A , Sekiguchi K , Iizuka S, Sugimoto A , Yuzurihara M , Aburada M , Hosoya E , Kogure K (1990) Effects of cycloheximide on delayed neuronal death in rat hippocampus. Brain Res 534: 299-302. Grabb M C , Choi D W (1999) Ischemic tolerance in murine cortical cell culture: critical role for N M D A receptors. J Neurosci 19: 1657-1662. Gunasekar P G , Sun PW, Kanthasamy A G , Borowitz JL, Isom G E (1996) Cyanide-induced neurotoxicity involves nitric oxide and reactive oxygen species generation after N-methyl-D-aspartate receptor activation. J Pharmacol Exp Ther 277: 150-155. Gwag BJ, Canzoniero L M , Sensi SL, Demaro JA, Koh JY, Goldberg M P , Jacquin M , Choi D W (1999) Calcium ionophores can induce either apoptosis or necrosis in cultured cortical neurons. Neuroscience 90: 1339-1348. Gwag BJ, Lobner D , Koh JY, Wie M B , Choi D W (1995) Blockade of glutamate receptors unmasks neuronal apoptosis after oxygen- glucose deprivation in vitro. Neuroscience 68: 615-619. Haddad G G , Jiang C (1993) 02 deprivation in the central nervous system: on mechanisms of neuronal response, differential sensitivity and injury. Prog Neurobiol 40: 277-318. Hagberg H , Lehmann A , Sandberg M , Nystrom B, Jacobson I, Hamberger A (1985) Ischemia-induced shift of inhibitory and excitatory amino acids from intra- to extracellular compartments. J Cereb Blood Flow Metab 5:413-419. Hara H , Friedlander R M , Gagliardini V , Ayata C, Fink K , Huang Z , Shimizu-Sasamata M , Yuan J, Moskowitz M A (1997) Inhibition of interleukin lbeta converting enzyme family proteases reduces ischemic and excitotoxic neuronal damage. Proc Natl Acad Sci U S A 94: 2007-2012. 91 Hatch A L (1993) A modified reagent for the confirmation of blood. J Forensic Sci 38: 1502-1506. Hibbs JB, Jr., Vavrin Z , Taintor R R (1987) L-arginine is required for expression of the activated macrophage effector mechanism causing selective metabolic inhibition in target cells. J Immunol 138: 550-565. Hillyard DR, Monje V D , Mintz IM, Bean BP, Nadasdi L , Ramachandran J, Miljanich G , Azimi-Zoonooz A , Mcintosh J M , Cruz L J , . (1992) A new Conus peptide ligand for mammalian presynaptic C a 2 + channels. Neuron 9: 69-77. Hollmann M , Boulter J, Maron C, Beasley L , Sullivan J, Pecht G , Heinemann S (1993) Zinc potentiates agonist-induced currents at certain splice variants of the N M D A receptor. Neuron 10: 943-954. Honore T, Davies SN, Drejer J, Fletcher EJ , Jacobsen P, Lodge D, Nielsen F E (1988) Quinoxalinediones: potent competitive n o n - N M D A glutamate receptor antagonists. Science 241: 701-703. Hoyt K R , Arden SR, Aizenman E , Reynolds IJ (1998) Reverse N a + / C a 2 + exchange contributes to glutamate-induced intracellular C a 2 + concentration increases in cultured rat forebrain neurons. M o l Pharmacol 53: 742-749. Ignarro L J (1990) Biosynthesis and metabolism of endothelium-derived nitric oxide. Annu Rev Pharmacol Toxicol 30: 535-560. Ishii T, Moriyoshi K , Sugihara H , Sakurada K , Kadotani H , Yokoi M , Akazawa C, Shigemoto R, Mizuno N , Masu M , . (1993) Molecular characterization of the family of the N-methyl-D-aspartate receptor subunits. J Biol Chem 268: 2836-2843. Ishmael JE, Franklin P H , Murray T F , Leid M (1996) High level expression of the N M D A R 1 glutamate receptor subunit in electroporated C O S cells. JNeurochem 67: 1500-1510. Jorgensen N K , Petersen SF, Damgaard I, Schousboe A , Hoffmann E K (1999) Increases in [Ca 2 +]j and changes in intracellular p H during chemical anoxia in mouse neocortical neurons in primary culture. J Neurosci Res 56: 358-370. Kaku D A , Giffard R G , Choi D W (1993) Neuroprotective effects of glutamate antagonists and extracellular acidity. Science 260: 1516-1518. 92 Keelan J, Vergun O, Duchen M R (1999) Excitotoxic mitochondrial depolarisation requires both calcium and nitric oxide in rat hippocampal neurons. J Physiol (Lond) 520 Pt 3: 797-813. Kerr JF, Wyllie A H , Currie A R (1972) Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26: 239-257. Kiedrowski L (1999) N-methyl-D-aspartate excitotoxicity: relationships among plasma membrane potential, N a + / C a 2 + exchange, mitochondrial C a 2 + overload, and cytoplasmic concentrations of C a 2 + , H + , and K + . M o l Pharmacol 56: 619-632. Kinouchi H , Epstein CJ, Mizui T, Carlson E , Chen SF, Chan P H (1991) Attenuation of focal cerebral ischemic injury in transgenic mice overexpressing CuZn superoxide dismutase. Proc Natl Acad Sci U S A 88: 11158-11162. Kirino T, Sano K (1984) Selective vulnerability in the gerbil hippocampus following transient ischemia. Acta Neuropathol (Berl) 62: 201-208. Kirino T, Tamura A , Sano K (1985) Selective vulnerability of the hippocampus to ischemia-reversible and irreversible types of ischemic cell damage. Prog Brain Res 63: 39-58. Konnerth A , Heinemann U , Yaari Y (1986) Nonsynaptic epileptogenesis in the mammalian hippocampus in vitro. I. Development of seizurelike activity in low extracellular calcium. J Neurophysiol 56: 409-423. Krai T, Luhmann HJ, Mittmann T, Heinemann U (1993) Role of N M D A receptors and voltage-activated calcium channels in an in vitro model of cerebral ischemia. Brain Res 612: 278-288. Kudo I, Murakami M , Hara S, Inoue K (1993) Mammalian non-pancreatic phospholipases A2. Biochim Biophys Acta 1170: 217-231. Kusumoto M , Dux E , Paschen W, Hossmann K A (1996) Susceptibility of hippocampal and cortical neurons to argon-mediated in vitro ischemia. J Neurochem 67: 1613-1621. Kutsuwada T, Kashiwabuchi N , Mori H , Sakimura K , Kushiya E , Araki K , Meguro H , Masaki H , Kumanishi T, Arakawa M , . (1992) Molecular diversity of the N M D A receptor channel. Nature 358: 36-41. Lambeth D O , Palmer G (1973) The kinetics and mechanism of reduction of electron transfer proteins and other compounds of biological interest by dithionite. J Biol Chem 248: 6095-6103. 93 Lampe R A , Defeo P A , Davison M D , Young J, Herman JL, Spreen R C , Horn M B , Mangano TJ, Keith R A (1993) Isolation and pharmacological characterization of omega-grammotoxin SLA, a novel peptide inhibitor of neuronal voltage-sensitive calcium channel responses. M o l Pharmacol 44: 451-460. Lazarewicz JW, Wroblewski JT, Palmer M E , Costa E (1988) Activation of N-methyl-D-aspartate-sensitive glutamate receptors stimulates arachidonic acid release in primary cultures of cerebellar granule cells. Neuropharmacology 27: 765-769. Lee J M , Zipfel GJ , Choi D W (1999) The changing landscape of ischaemic brain injury mechanisms. Nature 399: A7-14. Levy DI, Sucher NJ, Lipton S A (1990) Redox modulation of N M D A receptor-mediated toxicity in mammalian central neurons. Neurosci Lett 110: 291-296. Lipton P (1999) Ischemic cell death in brain neurons. Physiol Rev 79: 1431-1568. MacDonald JF, Mody I, Salter M W (1989) Regulation of N-methyl-D-aspartate receptors revealed by intracellular dialysis of murine neurones in culture. J Physiol 414: 17-34. MacDonald JF, Xiong X G , L u W Y , Raouf R, Orser B A (1998) Modulation of N M D A receptors. Prog Brain Res 116: 191-208. MacManus JP, Buchan A M , Hil l IE, Rasquinha I, Preston E (1993) Global ischemia can cause D N A fragmentation indicative of apoptosis in rat brain. Neurosci Lett 164: 89-92. Mac Vicar B A , Hochman D (1991) Imaging of synaptically evoked intrinsic optical signals in hippocampal slices. JNeurosci 11: 1458-1469. Madden K P , Clark W M , Marcoux F W , Probert A W , Jr., Weber M L , Rivier J, Zivin J A (1990) Treatment with conotoxin, an 'N-type' calcium channel blocker, in neuronal hypoxic-ischemic injury. Brain Res 537: 256-262. Magee JC, Avery R B , Christie BR, Johnston D (1996) Dihydropyridine-sensitive, voltage-gated C a 2 + channels contribute to the resting intracellular C a 2 + concentration of hippocampal C A 1 pyramidal neurons. J Neurophysiol 76: 3460-3470. Matthews C C , Zielke HR, Wollack JB, Fishman PS (2000) Enzymatic degradation protects neurons from glutamate excitotoxicity. JNeurochem 75: 1045-1052. 94 McBain CJ, Mayer M L (1994) N-methyl-D-aspartic acid receptor structure and function. Physiol Rev 74: 723-760. Misra HP, Fridovich I (1978) Inhibition of superoxide dismutases by azide. Arch Biochem Biophys 189: 317-322. Monaghan D T , Bridges RJ, Cotman C W (1989) The excitatory amino acid receptors: their classes, pharmacology, and distinct properties in the function of the central nervous system. Annu Rev Pharmacol Toxicol 29: 365-402. Moncada S, Palmer R M , Higgs E A (1991) Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol Rev 43: 109-142. Monyer H , Burnashev N , Laurie DJ, Sakmann B , Seeburg P H (1994) Developmental and regional expression in the rat brain and functional properties of four N M D A receptors. Neuron 12: 529-540. Monyer H , Sprengel R, Schoepfer R, Herb A , Higuchi M , Lomeli H , Burnashev N , Sakmann B , Seeburg P H (1992) Heteromeric N M D A receptors: molecular and functional distinction of subtypes. Science 256: 1217-1221. Moriyoshi K , Masu M , Ishii T, Shigemoto R, Mizuno N , Nakanishi S (1991) Molecular cloning and characterization of the rat N M D A receptor. Nature 354: 31-37. Nakajima K , Harada K , Ebina Y , Yoshimura T, Ito H , Ban T, Shingai R (1993) Relationship between resting cytosolic C a 2 + and responses induced by N - methyl-D-aspartate in hippocampal neurons. Brain Res 603: 321-323. Nakao N , Brundin P (1998) Neurodegeneration and glutamate induced oxidative stress. Prog Brain Res 116: 245-263. Nardi N , Avidan G , Daily D , Zilkha-Falb R, Barzilai A (1997) Biochemical and temporal analysis of events associated with apoptosis induced by lowering the extracellular potassium concentration in mouse cerebellar granule neurons. J Neurochem 68: 750-759. Newell D W , Barth A , Papermaster V , Malouf A T (1995) Glutamate and non-glutamate receptor mediated toxicity caused by oxygen and glucose deprivation in organotypic hippocampal cultures. JNeurosci 15: 7702-7711. 95 Nicoletti F, Bruno V , Copani A , Casabona G , Knopfel T (1996) Metabotropic glutamate receptors: a new target for the therapy of neurodegenerative disorders? Trends Neurosci 19: 267-271. Nitatori T, Sato N , Waguri S, Karasawa Y , Araki H , Shibanai K , Kominami E , Uchiyama Y (1995) Delayed neuronal death in the CA1 pyramidal cell layer of the gerbil hippocampus following transient ischemia is apoptosis. J Neurosci 15: 1001-1011. Noumi T, Maeda M , Futai M (1987) Mode of inhibition of sodium azide on H + -ATPase of Escherichia coli. F E B S Lett 213: 381-384. Novelli A , Reilly JA, Lysko P G , Henneberry R C (1988) Glutamate becomes neurotoxic via the N-methyl-D-aspartate receptor when intracellular energy levels are reduced. Brain Res 451: 205-212. Nowak L , Bregestovski P, Ascher P, Herbet A, Prochiantz A (1984) Magnesium gates glutamate-activated channels in mouse central neurones. Nature 307: 462-465. Nowicki JP, Duval D , Poignet H , Scatton B (1991) Nitric oxide mediates neuronal death after focal cerebral ischemia in the mouse. Eur J Pharmacol 204: 339-340. Oliver C N , Starke-Reed PE, Stadtman E R , Liu GJ, Carney J M , Floyd R A (1990) Oxidative damage to brain proteins, loss of glutamine synthetase activity, and production of free radicals during ischemia/reperfusion- induced injury to gerbil brain. Proc Natl Acad Sci U S A 87: 5144-5147. Olney JW, Ho O L , Rhee V (1971) Cytotoxic effects of acidic and sulphur containing amino acids on the infant mouse central nervous system. Exp Brain Res 14: 61-76. Olney JW, Price M T , Samson L , Labruyere J (1986) The role of specific ions in glutamate neurotoxicity. Neurosci Lett 65: 65-71. Pang Z , Geddes JW (1997) Mechanisms of cell death induced by the mitochondrial toxin 3-nitropropionic acid: acute excitotoxic necrosis and delayed apoptosis. J Neurosci 17: 3064-3073. Patel M N , Ardelt B K , Y i m G K , Isom G E (1991) Cyanide induces Ca 2 +-dependent and -independent release of glutamate from mouse brain slices. Neurosci Lett 131: 42-44. Patel M N , Y i m G K , Isom G E (1993) N-methyl-D-aspartate receptors mediate cyanide-induced cytotoxicity in hippocampal cultures. Neurotoxicology 14: 35-40. 96 Phillis JW, Sen S (1993) Oxypurinol attenuates hydroxyl radical production during ischemia/reperfusion injury of the rat cerebral cortex: an E S R study. Brain Res 628: 309-312. Piantadosi C A , Zhang J (1996) Mitochondrial generation of reactive oxygen species after brain ischemia in the rat. Stroke 27: 327-331. Pierce G N , Meng H (1992) The role of sodium-proton exchange in ischemic/reperfusion injury in the heart. Na^tT exchange and ischemic heart disease. A m J Cardiovasc Pathol 4: 91-102. Piser T M , Lampe R A , Keith R A , Thayer S A (1995a) Complete and reversible block by omega-grammotoxin SIA of glutamatergic synaptic transmission between cultured rat hippocampal neurons. Neurosci Lett 201: 135-138. Piser T M , Lampe R A , Keith R A , Thayer S A (1995b) Omega-grammotoxin SIA blocks multiple, voltage-gated, C a 2 + channel subtypes in cultured rat hippocampal neurons. M o l Pharmacol 48: 131-139. Pringle A K , Self J, Eshak M , Iannotti F (2000) Reducing conditions significantly attenuate the neuroprotective efficacy of competitive, but not other N M D A receptor antagonists in vitro. Eur J Neurosci 12: 3833-3842. Raman LM, Tong G , Jahr C E (1996) Beta-adrenergic regulation of synaptic N M D A receptors by cAMP-dependent protein kinase. Neuron 16: 415-421. Randall A , Tsien R W (1995) Pharmacological dissection of multiple types of C a 2 + channel currents in rat cerebellar granule neurons. J Neurosci 15: 2995-3012. Randall R D , Thayer S A (1992) Glutamate-induced calcium transient triggers delayed calcium overload and neurotoxicity in rat hippocampal neurons. Journal of Neuroscience 12: 1882-1895. Regan R F , Guo Y (1999) Extracellular reduced glutathione increases neuronal vulnerability to combined chemical hypoxia and glucose deprivation. Brain Res 817: 145-150. Reynolds IJ, Rush E A , Aizenman E (1990) Reduction of N M D A receptors with dithiothreitol increases [3HJ-MK-801 binding and NMDA-induced C a 2 + fluxes. Br J Pharmacol 101: 178-182. Riepe M W , Schmalzigaug K , Fink F, Oexle K , Ludolph A C (1996) N A D H in the pyramidal cell layer of hippocampal regions CA1 and CA3 upon selective inhibition and uncoupling of oxidative phosphorylation. Brain Res 710: 21-27. 97 Roettger V , Lipton P (1996) Mechanism of glutamate release from rat hippocampal slices during in vitro ischemia. Neuroscience 75: 677-685. Rogers D C , Hunter A J (1997) Dissociation of effects of glutamate receptor antagonists on excitotoxic and hypoxic neuronal cell death in a novel rat cortical culture system. Brain Res Bull 44: 131-139. Roos M W , Ericsson A (1999) N-tert-butyl-alpha-phenylnitrone reduces the number of microinfarctions in the rabbit brain cortex. Exp Brain Res 124: 271-272. Rosenmund C, Westbrook G L (1993a) Calcium-induced actin depolymerization reduces N M D A channel activity. Neuron 10: 805-814. Rosenmund C, Westbrook G L (1993b) Rundown of N-methyl-D-aspartate channels during whole-cell recording in rat hippocampal neurons: role of Ca + and A T P . J Physiol 470: 705-729. Rossi DJ, Oshima T, Attwell D (2000) Glutamate release in severe brain ischaemia is mainly by reversed uptake. Nature 403: 316-321. Rothman S M (1983) Synaptic activity mediates death of hypoxic neurons. Science 220: 536-537. Rothman S M (1985) The neurotoxicity of excitatory amino acids is produced by passive chloride influx. Journal of Neuroscience 5: 1483-1489. Rothman S M . Thurston JH. Hauhart R E . (1987). Delayed neurotoxicity of excitatory amino acids in vitro. Neuroscience 22: 471-480. Sakaguchi T, Kuno M , Kawasaki K (1999) Disparity of cell swelling and rapid neuronal death by excitotoxic insults in rat hippocampal slice cultures. Neurosci Lett 274: 135-138. Sakamoto A , Ohnishi ST, Ohnishi T, Ogawa R (1991) Relationship between free radical production and lipid peroxidation during ischemia-reperfusion injury in the rat brain. Brain Res 554: 186-192. Sattler R, Xiong Z , L u W Y , Hafner M , MacDonald JF, Tymianski M (1999) Specific coupling of N M D A receptor activation to nitric oxide neurotoxicity by PSD-95 protein. Science 284: 1845-1848. 98 Sattler R, Xiong Z , L u W Y , MacDonald JF, Tymianski M (2000) Distinct roles of synaptic and extrasynaptic N M D A receptors in excitotoxicity. Journal of Neuroscience 20: 22-33. Saudou F, Finkbeiner S, Devys D , Greenberg M E (1998) Huntingtin acts in the nucleus to induce apoptosis but death does not correlate with the formation of intranuclear inclusions. Cell 95: 55-66. Schinder A F , Olson E C , Spitzer N C , Montal M (1996) Mitochondrial dysfunction is a primary event in glutamate neurotoxicity. JNeurosci 16: 6125-6133. Shigeno T, Yamasaki Y , Kato G , Kusaka K , Mima T, Takakura K , Graham DI, Furukawa S (1990) Reduction of delayed neuronal death by inhibition of protein synthesis. Neurosci Lett 120: 117-119. Shimizu S, Narita M , Tsujimoto Y (1999) Bcl-2 family proteins regulate the release of apoptogenic cytochrome c by the mitochondrial channel V D A C . Nature 399: 483-487. Siesjo B K , Bengtsson F, Grampp W, Theander S (1989) Calcium, excitotoxins, and neuronal death in the brain. Ann N Y Acad Sci 568: 234-251. Silver I A , Erecinska M (1990) Intracellular and extracellular changes of [Ca 2 + ] in hypoxia and ischemia in rat brain in vivo. J Gen Physiol 95: 837-866. Silver IA, Erecinska M (1992) Ion homeostasis in rat brain in vivo: intra- and extracellular [Ca ] and [H +] in the hippocampus during recovery from short-term, transient ischemia. J Cereb Blood Flow Metab 12: 759-772. Siman R, Noszek JC, Kegerise C (1989) Calpain I activation is specifically related to excitatory amino acid induction of hippocampal damage. JNeurosci 9: 1579-1590. Smith G A , Brett C L , Church J (1998) Effects of noradrenaline on intracellular p H in acutely dissociated adult rat hippocampal CA1 neurones. J Physiol 512 (Pt 2): 487-505. Smith RP, Louis C A , Kruszyna R, Kruszyna H (1991) Acute neurotoxicity of sodium azide and nitric oxide. Fundam Appl Toxicol 17: 120-127. Snyder SH, Bredt DS (1991) Nitric oxide as a neuronal messenger. Trends Pharmacol Sci 12: 125-128. 99 Sommer B , Keinanen K , Verdoorn T A , Wisden W, Burnashev N , Herb A , Kohler M , Takagi T, Sakmann B , Seeburg P H (1990) Flip and flop: a cell-specific functional switch in glutamate-operated channels of the CNS. Science 249: 1580-1585. Stout A K , Raphael H M , Kanterewicz B l , Klann E , Reynolds IJ (1998) Glutamate-induced neuron death requires mitochondrial calcium uptake. Nat Neurosci 1: 366-373. Strasser TJ, Fischer G (1995) Quantitative measurement of neuronal degeneration in organotypic hippocampal cultures after combined oxygen/glucose deprivation. JNeurosci Methods 57: 177-186. Stys P K , Waxman SG, Ransom B R (1992) Ionic mechanisms of anoxic injury in mammalian C N S white matter: role of N a + channels and N a + - C a 2 + exchanger. J Neurosci 12: 430-439. Sugihara H , Moriyoshi K , Ishii T, Masu M , Nakanishi S (1992) Structures and properties of seven isoforms of the N M D A receptor generated by alternative splicing. Biochem Biophys Res Commun 185: 826-832. Sullivan J M , Traynelis SF, Chen HS, Escobar W, Heinemann SF, Lipton S A (1994) Identification of two cysteine residues that are required for redox modulation of the N M D A subtype of glutamate receptor. Neuron 13: 929-936. Susin S A , Lorenzo H K , Zamzami N , Marzo I, Snow B E , Brothers G M , Mangion J, Jacotot E , Costantini P, Loeffler M , Larochette N , Goodlett DR, Aebersold R, Siderovski DP, Penninger J M , Kroemer G (1999) Molecular characterization of mitochondrial apoptosis-inducing factor. Nature 397: 441-446. Szabo C, Zingarelli B , O'Connor M , Salzman A L (1996) D N A strand breakage, activation of poly (ADP-ribose) synthetase, and cellular energy depletion are involved in the cytotoxicity of macrophages and smooth muscle cells exposed to peroxynitrite. Proc Natl Acad Sci U S A 93: 1753-1758. Szatkowski M , Attwell D (1994) Triggering and execution of neuronal death in brain ischaemia: two phases of glutamate release by different mechanisms. Trends Neurosci 17: 359-365. Tanaka K , Dora E , Greenberg JH, Reivich M (1986) Cerebral glucose metabolism during the recovery period after ischemia— its relationship to NADH-fluorescence, blood flow, E c o G and histology. Stroke 17: 994-1004. Tang L H , Aizenman E (1993) The modulation of N-methyl-D-aspartate receptors by redox and alkylating reagents in rat cortical neurones in vitro. J Physiol 465: 303-323. 100 Thornberry N A , Lazebnik Y (1998) Caspases: enemies within. Science 281: 1312-1316. Traynelis SF, Cull-Candy S G (1990) Proton inhibition of N-methyl-D-aspartate receptors in cerebellar neurons. Nature 345: 347-350. Trifiletti R R (1992) Neuroprotective effects of NG-nitro-L-arginine in focal stroke in the 7- day old rat. Eur J Pharmacol 218: 197-198. Trudeau L E , Parpura V , Haydon P G (1999) Activation of neurotransmitter release in hippocampal nerve terminals during recovery from intracellular acidification. J Neurophysiol 81: 2627-2635. Turrens JF (1997) Superoxide production by the mitochondrial respiratory chain. Biosci Rep 17: 3-8. Turrens JF, Beconi M , Barilla J, Chavez U B , McCord J M (1991) Mitochondrial generation of oxygen radicals during reoxygenation of ischemic tissues. Free Radic Res Commun 12-13 Pt 2: 681-689. Uto A , Dux E , Kusumoto M , Hossmann K A (1995) Delayed neuronal death after brief histotoxic hypoxia in vitro. J Neurochem 64: 2185-2192. Varming T, Drejer J, Frandsen A , Schousboe A (1996) Characterization of a chemical anoxia model in cerebellar granule neurons using sodium azide: protection by nifedipine and MK-801. J Neurosci Res 44: 40-46. Vasilyeva E A , Minkov IB, Fitin A F , Vinogradov A D (1982) Kinetic mechanism of mitochondrial adenosine triphosphatase. Inhibition by azide and activation by sulphite. Biochem J202: 15-23. Verdoorn T A , Burnashev N , Monyer H , Seeburg P H , Sakmann B (1991) Structural determinants of ion flow through recombinant glutamate receptor channels. Science 252: 1715-1718. Vergun O, Keelan J, Khodorov BI, Duchen M R (1999) Glutamate-induced mitochondrial depolarisation and perturbation of calcium homeostasis in cultured rat hippocampal neurones. J Physiol (Lond) 519 Pt 2: 451-466. Verity M A , Torres M , Sarafian T (1991) Paradoxical potentiation by low extracellular C a 2 + of acute chemical anoxic neuronal injury in cerebellar granule cell culture. M o l Chem Neuropathol 15: 217-233. 101 Vornov JJ, Thomas A G , Jo D (1996) Protective effects of extracellular acidosis and blockade of sodium/hydrogen ion exchange during recovery from metabolic inhibition in neuronal tissue culture. JNeurochem 67: 2379-2389. Vyklicky L , Jr. (1993) Calcium-mediated modulation of N-methyl-D-aspartate ( N M D A ) responses in cultured rat hippocampal neurones. J Physiol 470: 575-600. Wang Y T , Pak Y S , Salter M W (1993) Rundown of NMDA-receptor mediated currents is resistant to lowering intracellular [Ca +] and is prevented by A T P in rat spinal dorsal horn neurons. Neurosci Lett 157: 183-186. Weih M , Bergk A , Isaev N K , Ruscher K , Megow D, Riepe M , Meisel A , Victorov IV, Dirnagl U , Dirnagi U (1999) Induction of ischemic tolerance in rat cortical neurons by 3- nitropropionic acid: chemical preconditioning [published erratum appears in Neurosci Lett 2000 Jan 14;278(3):194]. Neurosci Lett 272: 207-210. Wenthold RJ, Yokotani N , Doi K , Wada K (1992) Immunochemical characterization of the non-N M D A glutamate receptor using subunit-specific antibodies. Evidence for a hetero-oligomeric structure in rat brain. J Biol Chem 267: 501-507. Wheeler D B , Randall A , Tsien R W (1994) Roles of N-type and Q-type C a 2 + channels in supporting hippocampal synaptic transmission. Science 264: 107-111. Williams RJ, Maus M , Stella N , Glowinski J, Premont J (1996) Reduced glucose metabolism enhances the glutamate-evoked release of arachidonic acid from striatal neurons. Neuroscience 74: 461-468. Wong E H , Kemp JA, Priestley T, Knight A R , Woodruff G N , Iversen L L (1986) The anticonvulsant MK-801 is a potent N-methyl-D-aspartate antagonist. Proc Natl Acad Sci U S A 83: 7104-7108. Yoshida S, Busto R, Santiso M , Ginsberg M D (1984) Brain lipid peroxidation induced by postischemic reoxygenation in vitro: effect of vitamin E . J Cereb Blood Flow Metab 4: 466-469. Yun H Y , Gonzalez-Zulueta M , Dawson V L , Dawson T M (1998) Nitric oxide mediates N -methyl-D-aspartate receptor-induced activation of p21ras. Proc Natl Acad Sci U S A 95: 5773-5778. Zukin RS, Bennett M V (1995) Alternatively spliced isoforms of the N M D A R I receptor subunit. Trends Neurosci 18: 306-313. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089928/manifest

Comment

Related Items