University of Nevada, Reno
PRESYNAPTIC AND POSTSYNAPTIC
COMPARTMENTS REGULATE NEURONAL CELL
EXCITABILITY AND NEUROPROTECTION
A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of
Philosophy in Cell and Molecular Biology
by
Brendan John Lujan
Dr. Robert Renden/Dissertation Advisor
May, 2016
Copyright by Brendan John Lujan 2016 All Rights Reserved THE GRADUATE SCHOOL
We recommend that the dissertation prepared under our supervision by
BRENDAN JOHN LUJAN entitled Presynaptic and Postsynaptic Compartments Regulate Neuronal Cell Excitability and Neuroprotection
be accepted in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
Robert Renden, Ph.D., Advisor
Christopher von Bartheld, M.D., Committee Member
Ruben Dagda, Ph.D., Committee Member
Thomas Gould, Ph.D., Committee Member
Normand LeBlanc, Ph.D., Committee Member
Minggen Lu, Ph.D., Graduate School Representative
David W. Weh, Ph.D., Dean, Graduate School
May, 2016 i
ABSTRACT
The experiments performed in this dissertation examine the basic synaptic function of neurons in the mammalian central nervous system. In one project, we describe a novel function of NMDA receptors located on the postsynaptic membrane to regulate neuroprotection and synaptic strength. In a second project, we provide evidence that presynaptic neuronal metabolism is crucial for synaptic transmission. Both presynaptic and postsynaptic compartments serve integral roles in maintenance of proper neuronal function.
Although NMDA receptor function has classically been described on the basis of its ionotropic properties, we show a novel function by which ligand binding mediates transmembrane signaling without ion flux. In Chapter 2, we review the role of NMDA receptors in regulating neuronal survival. Next, we describe a novel non-ionotropic signal transduction mechanism for the NMDA receptor in mediating this effect. In Chapter 4, we review the role of NMDA receptors in synaptic plasticity. Next we provide evidence that the non-ionotropic mechanism described above for NMDA receptors in regulation of neuroprotection, also regulates synaptic plasticity. This novel NMDA receptor function was shown to be mediated through both the cell survival promoting Akt-dependent signaling cascade and the ERK pathway.
Neuronal bioenergetics play a crucial role in proper function of information transmission. In Chapters 7 and 8, we investigated the mechanisms of energy homeostasis underlying basal and activity-driven synaptic function. Although the presynaptic compartment clearly places large demand on energy production in maintenance of basic ii
synaptic function, it is currently unclear which mode(s) of energy production exist and
predominate at the presynaptic terminal during ongoing activity. We show source specific
use of cellular energy to regulate the action potential waveform, and downstream
transmission. Further, we suggest presynaptic energy is differentially utilized, and that transmission is dependent on ATP production route during high frequency stimulation in
a vertebrate central synapse. Our study suggests that energy production source is important to maintain functional information processing.
iii
ACKNOWLEDGEMENTS
I would first like to thank Dr. Qi Wan for giving me the opportunity to work in his lab. I entered graduate school with minimal experience and this initial period of my graduate work gave me the confidence to pursue a doctoral degree. I would also like to extend my appreciation to Mingxia Liao, the lab manager in Dr. Qi Wan’s lab, who was able to teach me countless different techniques that will be invaluable tools as I continue to pursue a career in academic research.
To Dr. Robert Renden, your guidance, support and knowledge have welded me into the researcher that I am today. The undying motivation that you have provided has sparked my interest into the field of synaptic physiology. I am grateful that I have discovered an area of science that I will continue to try to understand for many years to come. To all the members of my committee, I appreciate the constructive criticism of my work during my qualification exam and dissertation defense. There was never a shortage of new ideas to pursue to continue to push my project forward. I would also like to extend appreciation toward Dr. Mick Hitchcock, who has provided my stipend support through the last three years. Without this support, my research would not have been possible.
I would like to extend thanks to my immediate family who have continued to encourage me through my graduate school experience. I would like to extend an extra special thanks to my wife Ruby, who has been by my side through the good times and tough times. Without your support, I would not be the man that I am today. iv
TABLE OF CONTENTS
Page
Chapter 1 Introduction...... 1
1.1 Postsynaptic NMDARs regulate neuronal survival and death signaling ...... 2
1.2 Postsynaptic NMDARs regulate bidirectional synaptic plasticity ...... 2
1.3 Presynaptic bioenergetics regulate neurotransmission ...... 3
Chapter 2 Differential Roles of GluN2A- and GluN2B-containing NMDA Receptors in Neuronal Survival and Death ...... 4
2.1 Summary ...... 5
2.2 Introduction ...... 5
2.3 GluN2ARs in neuronal survival ...... 7
2.3.1 GluN2AR-PTEN-TDP43 pathway ...... 8
2.3.2 DJ-1-PTEN-PINK1-GluN2AR pathway ...... 11
2.4 GluN2BR in neuronal death and neurodegeneration ...... 13
2.4.1 DJ-1-PTEN-GluN2BR pathway ...... 14
2.4.2 GluN2BR-PINK1-Akt pathway ...... 15
2.5 Discussion ...... 19
Chapter 3 Glycine Triggers a Non-ionotropic Activity of GluN2A-containing NMDA
Receptors to Confer Neuroprotection ...... 20
3.1 Summary ...... 21
3.2 Introduction ...... 21
3.3 Methods ...... 23 v
3.4 Results ...... 29
3.4.1 Glycine increases Akt phosphorylation independent of Ca2+ influx through
NMDAR channels ...... 29
3.4.2 Elevation of Akt phosphorylation by glycine does not depend on the activation
of glycine receptors ...... 31
3.4.3 Glycine alone enhances Akt activation through a non-ionotropic activation of
GluN2ARs ...... 32
3.4.4 The glycine-GluN1 binding site mediates the non-ionotropic activation of
GluN2ARs ...... 34
3.4.5 Glycine prevents glutamate neurotoxicity-induced neuronal death through non-
ionotropic activation of GluN2ARs ...... 36
3.4.6 The neuroprotective role of non-ionotropic activation of NMDARs by glycine
in ischemic stroke ...... 38
3.5 Discussion ...... 58
Chapter 4 A Synaptic Model of Learning and Memory ...... 64
4.1 Summary ...... 65
4.2 Introduction ...... 65
4.3 AMPARs ...... 68
4.3.1 AMPAR Trafficking ...... 69
4.4 NMDARs ...... 71
4.4.1 NMDAR regulation of ERK 1/2 signaling ...... 72
4.4.2 Differential regulation of NMDARs in bidirectional synaptic plasticity ...... 74 vi
4.5 Discussion ...... 78
Chapter 5 Glycine Potentiates AMPA Receptor Function through Metabotropic
Activation of GluN2A-containing NMDA Receptors ...... 79
5.1 Summary ...... 80
5.2 Introduction ...... 80
5.3 Methods ...... 81
5.4 Results ...... 85
5.4.1 Glycine potentiates AMPA-induced whole-cell currents independent of
NMDAR channel activity...... 85
5.4.2 Glycine enhances AMPAR-mediated synaptic function independent of
NMDAR channel activity...... 87
5.4.3 Potentiation of AMPAR function by glycine requires ERK1/2 activation...... 88
5.4.4 Glycine promotes ERK1/2 activation independent of NMDAR channel pore
activities...... 89
5.4.5 Glycine enhances ERK1/2 activation through a metabotropic activity of
GluN2ARs...... 90
5.4.6 A metabotropic activity of GluN2ARs mediates glycine-induced potentiation
of AMPAR function...... 92
5.5 Discussion ...... 107
Chapter 6 Single Agonist NMDA does not Regulate the Metabotropic Signaling of the NMDA Receptor ...... 109
6.1 Summary ...... 110 vii
6.2 Introduction ...... 110
6.3 Methods ...... 113
6.4 Results ...... 115
6.4.1 Metabotropic NMDAR activation fails to regulate ERK1/2 by the synthetic
agonist NMDA ...... 115
6.4.2 mEPSCs are unresponsive to metabotropic activation of the NMDA receptor116
6.4.3 Single Agonist NMDA does not regulate whole-cell AMPAR-mediated
responses after metabotropic activation of the NMDAR ...... 117
6.5 Discussion ...... 124
Chapter 7 Glycolysis Selectively Shapes the Presynaptic Action Potential Waveform at the Calyx of Held ...... 126
7.1 Summary ...... 127
7.2 Introduction ...... 128
7.3 Methods ...... 130
7.4 Results ...... 135
7.4.1 Presynaptic function at the calyx of Held relies on local ATP production .....135
7.4.2 Basal Ca2+ is not altered by inhibition of glycolysis or OxPhos ...... 137
7.4.3 Local glycolytic ATP is required for maintenance of presynaptic AP waveform138
7.4.4 Ca2+-influx via VGCCs is altered by inhibition of glycolysis ...... 139
7.4.5 Glycolysis inhibition attenuates transmission due to altered AP waveform ...141
7.4.6 Stoichiometric changes in intracellular concentration of Na/K contribute to
altered AP waveform ...... 142 viii
7.5 Discussion ...... 164
7.5.1 Specific isolation of glycolysis versus mitochondrial oxidative phosphorylation165
7.5.2 Glycolysis fuels presynaptic APs ...... 167
7.5.3 Altered AP shape reduces Ca2+ current, fully accounts for smaller EPSC ....168
7.5.4 Presynaptic mitochondrial OxPhos is not required to maintain basal
transmission ...... 169
7.5.5 Physiological Relevance ...... 170
Chapter 8 The Developmental Profile of Presynaptic Energy Utilization during High
Frequency Neurotransmission ...... 172
8.1 Summary ...... 173
8.2 Introduction ...... 174
8.3 Methods ...... 175
8.4 Results ...... 178
8.4.1 Presynaptic depression causes postsynaptic desensitization ...... 178
8.4.2 Presynaptic depression is regulated by ATP derived from both glycolysis and
mitochondrial OxPhos at the developmentally immature calyx ...... 180
8.4.3 OxPhos supports presynaptic Ca2+ buffering during HFS in immature terminal182
8.4.4 Presynaptic energy deficits do not affect recovery after synaptic depression at
the developmentally immature calyx ...... 183
8.4.5 Developmental changes affect presynaptic depression in hearing mice after
inhibition of ATP-production ...... 185 ix
8.4.6 Recovery from synaptic depression is not affected by loss of presynaptic ATP
in post hearing calyx of Held synapses ...... 187
8.4.7 ATP is required for high-frequency activity at the developmentally mature
calyx...... 188
8.4.8 Recovery from synaptic depression at the developmentally mature calyx .....190
8.5 Discussion ...... 210
8.5.1 Energy use in the presynaptic terminal of prehearing mice ...... 210
8.5.2 Developmental shift in the presynaptic metabolic profile ...... 211
8.5.3 ATP is a bottleneck for high frequency transmission...... 212
Chapter 9 Closing Remarks ...... 213
LIST OF FIGURES
Figure 2.1 GluN2ARs and GluN2BRs mediate neuronal survival and death pathways during CNS insult...... 17
Figure 3.1 Enhancement of Akt phosphorylation by glycine in cortical neurons does not
require the channel activities of NMDARs...... 40
Figure 3.2 Enhancement of Akt phosphorylation by glycine in cortical neurons does not
depend on the activation of glycine receptors or the activity of p38-MAPK signaling. .. 42
Figure 3.3 Non-ionotropic activity of GluN2AR mediates glycine-induced enhancement
of Akt phosphorylation ...... 43
Figure 3.4 Glycine-GluN1 binding is required for glycine-induced non-ionotropic activation of GluN2ARs...... 46 x
Figure 3.5 Glycine protects against glutamate neurotoxicity-induced neuronal injury in
cortical neurons through non-ionotropic activation of GluN2ARs...... 49
Figure 3.6 Glycine treatment reduces the infarct area of ischemic brain independent of
glycine receptor activation and the channel activity of NMDARs...... 52
Figure 3.7 Glycine promotes functional recovery of ischemic animals independent of
glycine receptor activation and the channel activity of NMDARs...... 54
Figure 4.1 NMDARs bidirectionally regulate synaptic plasticity ...... 77
Figure 5.1 Glycine enhances AMPAR-mediated whole-cell currents in hippocampal neurons in which the NMDAR channel activity and glycine receptor activation are inhibited...... 94
Figure 5.2 Glycine enhances AMPAR-mediated synaptic currents independent of
NMDAR channel activity...... 96
Figure 5.3 Inhibition of ERK1/2 activation prevents potentiation of AMPAR function by
glycine...... 98
Figure 5.4 Glycine increases ERK1/2 phosphorylation independent of NMDAR channel
activity in hippocampal neurons...... 100
Figure 5.5 Glycine increases ERK1/2 phosphorylation through metabotropic activity of
GluN2ARs in HEK293 cells...... 102
Figure 5.6 Glycine increases ERK1/2 phosphorylation via metabotropic activity of
GluN2ARs in hippocampal neurons...... 104
Figure 5.7 Glycine enhances AMPAR function through metabotropic activity of
GluN2ARs in hippocampal neurons...... 105 xi
Figure 6.1 Metabotropic activation of the NMDAR by single agonist NMDA does not
regulate ERK1/2 signling in hippocampal neurons...... 119
Figure 6.2 Metabotropic activation of the NMDAR by single agonist NMDA does not
regulate mEPSCs ...... 121
Figure 6.3 Metabotropic activation of the NMDAR by single agonist NMDA does not
regulate AMPAR-mediated whole-cell currents ...... 122
Figure 7.1 Cellular ATP from glycolysis, but not mitochondrial OxPhos, is necessary to
maintain basal neural activity at the prehearing calyx of Held...... 145
Figure 7.2 Quantal size and frequency is unaffected by presynaptic inhibition of glycolysis or mitochondrial OxPhos...... 148
Figure 7.3 Resting levels of presynaptic Ca2+ are unaffected by selective loss of
glycolysis or mitochondrial OxPhos...... 150
Figure 7.4 Presynaptic AP waveform is inhibited by loss of glycolytic ATP...... 152
Figure 7.5 Presynaptic Ca2+ currents are altered after loss of glycolysis, but not
mitochondrial OxPhos...... 154
Figure 7.6 Modeling presynaptic Ca2+ current ...... 156
Figure 7.7 Ca2+ currents elicited by replaying recorded APs support IAA inhibition of
AP-evoked ICa...... 158
Figure 7.8 Restoring resting membrane potential only partially rescues AP waveform in
the absence of presynaptic glycolysis...... 160
Figure 8.1 Postsynaptic AMPA receptor desensitization during high-frequency activity in
both prehearing and hearing mice ...... 192 xii
Figure 8.2 High-frequency synaptic neurotransmission is differentially modulated by
ATP source at the developing calyx of Held...... 194
Figure 8.3 OxPhos supports presynaptic Ca2+ buffering at the prehearing Calyx...... 196
Figure 8.4 Mitochondrial OxPhos supports presynaptic Ca2+ buffering high frequency
activity...... 198
Figure 8.5 Recovery of the RRP is not dependent on a specific ATP source in prehearing
terminals...... 200
Figure 8.6 Compensation of ATP production during depression at 100 Hz at the mature
calyx of Held...... 202
Figure 8.7 Recovery from synaptic depression is not dependent on a specific ATP source
in hearing terminals...... 204
Figure 8.8 The role of presynaptic ATP during depression while driving stimulation at
300 Hz at the mature calyx of Held...... 206
Figure 8.9 Recovery from synaptic depression at the developmentally mature calyx
requires optimal ATP production...... 208
LIST OF TABLES
Table 3.1 Modified Neurological Severity Score (mNSS) ...... 56
Table 3.2 The Beam Walk Test Scoring Criteria ...... 57
Table 7.1 Complete descriptive data of sEPSCs...... 162
Table 7.2 Complete descriptive data of APs...... 163
1
Chapter 1 Introduction 2
The synapse is the computational unit of the central nervous system and relies
upon proper function of both presynaptic and postsynaptic compartments. The synapse
regulates important communication events and also provides cues to regulate neuronal survival and death signaling, and synaptic rearrangements that modulate transmission efficacy. The complete mechanisms by which the presynaptic and postsynaptic compartments function are not fully understood.
1.1 Postsynaptic NMDARs regulate neuronal survival and death signaling
The NMDA receptor (NMDAR) is a subtype of ionotropic glutamate receptors located on the postsynaptic membrane that serves important roles in mediating neuroprotection during brain ischemia and traumatic brain injury. NMDARs are heteromeric tetramers that consist of GluN1, GluN2 and GluN3 subunits (Monyer et al.
1992) and are Ca2+-permeable upon activation. Different subunit inclusions confer distinct roles of NMDAR subtypes and link them with different intracellular signaling pathways (Hayashi et al. 2009; Loftis and Janowsky 2003). Although this receptor has been classically defined by its ionotropic signaling capabilities, increasing evidence supports a novel function of this receptor to signal through both ionotropic and metabotropic mechanisms (Stein et al. 2015). However, the molecular mechanisms underlying the differential effects of NMDARs in neuronal survival and death are not fully understood.
1.2 Postsynaptic NMDARs regulate bidirectional synaptic plasticity
Similar to its role in regulation of survival and death signaling, the NMDAR also
serves critical roles in regulation of bidirectional postsynaptic plasticity events. Synaptic 3
plasticity is the ability of a synapse to readily alter its efficiency of connectivity.
NMDARs of specific subunit inclusions also bidirectionally regulate synaptic
potentiation and depression events as well (Kim et al. 2005a). The aforementioned novel
metabotropic signaling ability of the NMDAR has recently been shown to similarly
regulate synaptic plasticity events (Kessels et al. 2013; Nabavi et al. 2013). The complete
mechanism by which the NMDAR can regulate both synaptic potentiation and depression
remains a highly studied topic but the complete molecular mechanism of these plasticity
events remains elusive.
1.3 Presynaptic bioenergetics regulate neurotransmission
Neurons are energetically expensive. While a majority of this energy is expended to regenerate electrical polarization of neurons, the efficient release and recycling of neurotransmitter is also critically important to allow chemical transmission between neuronal presynaptic and postsynaptic compartments. Mitochondria are the major suppliers of cellular energy, generating ATP via oxidative phosphorylation. Recent evidence suggests all ATP may not be equal and thus production route may be important for proper neuronal function (Jang et al. 2016). However, the specific utilization of energy from cytosolic (glycolytic) and/or mitochondrial respiration during synaptic
neurotransmission is unknown.
4
Chapter 2 Differential Roles of GluN2A- and GluN2B-containing NMDA Receptors in Neuronal Survival and Death
Brendan Lujan, Xiaoxuan Liu, Qi Wan
Published in final form: Int J Physiol Pathophysiol Pharmacol. 2012;4(4):211-8. Epub
2012 Dec 2 5
2.1 Summary
Glutamate-induced neurotoxicity is the primary molecular mechanism that
induces neuronal death in a variety of pathologies in the central nervous system (CNS).
Toxicity signals are relayed from extracellular space to the cytoplasm by N-methyl-D- aspartate receptors (NMDARs) and regulate a variety of survival and death signaling.
Differential subunit combinations of NMDARs confer neuroprotection or trigger neuronal death pathways depending on the subunit arrangements of NMDARs and its
localization on the cell membrane. It is well-known that GluN2B-containing NMDARs
(GluN2BR) preferentially link to signaling cascades involved in CNS injury promoting
neuronal death and neurodegeneration. Conversely, less well-known mechanisms of
neuronal survival signaling are associated with GluN2A-containing NMDARs
(GluN2AR)-dependent signal pathways. This review will discuss the most recent signaling cascades associated with GluN2ARs and GluN2BRs.
2.2 Introduction
The NMDARs play functionally diverse roles during physiological and pathophysiological conditions in mammalian organisms. NMDARs are a subtype of ionotropic glutamate receptors permeable to Ca2+ that is responsible for the majority of excitatory neural transmission in the CNS (Dingledine et al. 1999). These ligand-gated receptors have been shown to play crucial roles in regulation of neural development
(Constantinepaton et al. 1990; Kerchner and Nicoll 2008), synaptic plasticity (Barria and
Malinow 2002; Malenka and Nicoll 1999), and glutamate-induced neurotoxicity (Aarts et al. 2002; Choi 1988; Tu et al. 2010). NMDAR dysfunction has been implicated in many 6
CNS pathologies including traumatic brain injury, neurodegenerative diseases and ischemic stroke (Koutsilleri and Riederer 2007; Lee et al. 1999; Liu et al. 2007). The theory of glutamate-induced neurotoxicity is the mechanism of neuronal death thought to underlie many types of CNS injuries. Glutamate-induced neurotoxicity occurs due to an overactivation of glutamate receptors, mainly the NMDARs increase their permeability to
Ca2+ induced by the overly released extracellular glutamate and trigger neuronal death events (Benveniste et al. 1984; Lipton and Rosenberg 1994).
NMDARs are known to be heteromeric tetramers in molecular structure, containing an obligate GluN1 subunit, with GluN2(A-D) and GluN3(A-B) subunits
(Dingledine et al. 1999; Sheng et al. 1994). The receptor is activated by agonist glutamate binding to the GluN2 subunit and co-agonist glycine and/or D-Serine binding to the
GluN1 subunit to regulate channel gating and influx of the pertinent cation second- messenger Ca2+ via the channel pore (Johnson and Ascher 1987; Oliet and Mothet
2009). GluN2ARs and GluN2BRs are the most common NMDAR subtypes found in mammalian CNS. The differences in NMDAR function may be attributed to the specific subunit combinations that are present in each receptor subtype as they show functionally different properties regarding electrophysiology and different sensitivities to regulation by intracellular signals (Buller et al. 1994).
Pharmacological antagonists targeting NMDARs have been unsuccessful in clinical trials in treating various CNS disease states (Ikonomidou and Turski 2002; Kemp and McKernan 2002; Traynor et al. 2006). This could be due simply to the possibility 7
that NMDARs potentiate both cell-survival signaling and cell-death signaling. Thus, nonspecific inhibition of NMDARs will not only block the neuronal death pathways but also inhibit pro-survival signaling. Among many extracellular and intracellular processes that can modulate NMDARs, phosphorylation and dephosphorylation by protein kinases and phosphatases are especially important, because they can critically regulate trafficking, surface expression, and the channel properties of NMDARs (Carroll and
Zukin 2002; Swope et al. 1999). Targeting intracellular pathways that NMDARs govern may provide a stronger mechanistic basis to promote cell survival during CNS injury.
This review will attempt to discuss the roles of GluN2ARs and GluN2BRs in neuronal survival and death and the associated intracellular signaling cascades that are initiated by each receptor subtype and how cross-talk among these pathways can regulate neuronal cell fate in CNS pathologies. Furthermore, the dissection of these intracellular pathways will elucidate prominent therapeutic targets that may play strong roles in regulating neuron cell survival.
2.3 GluN2ARs in neuronal survival
GluN2AR receptor activation has been shown to link preferentially to intracellular signaling cascades during CNS injury that promote cell survival (Anastasio et al. 2009;
Chen et al. 2008; DeRidder et al. 2006), and GluN2ARs localize preferentially to the synaptic zone (Stocca and Vicini 1998; Tovar and Westbrook 1999). One theory explaining the roles of NMDARs during the pathogenesis of CNS injury is built upon the notion that location, either synaptic or extrasynaptic, may determine function. In hippocampal neuronal culture, it has been demonstrated that synaptic activation of 8
GluN2ARs preferentially activate CREB, enhance BDNF gene expression and activate anti-apoptotic signaling pathways; all mechanisms that contribute to neuronal cell
survival during CNS insult (Hardingham et al. 2002). Similarly, GluN2AR activation in
mature rat cortical cultures, located both synaptically and extrasynaptically, confer
neuroprotection against neuronal damage (Liu et al. 2007). In an in-vivo model of rat
focal ischemic stroke, GluN2ARs protect against apoptotic signaling, in part due to the
activation of an Akt-dependent signaling pathway (Liu et al. 2007). In another study that
pharmacologically disabled GluN2ARs, it was shown that neuron cell death increased
after transient global ischemia in rats and that GluN2ARs were responsible for ischemia-
induced activation of the neuroprotective transcription factor CREB, enhancing
expression of gene targets cpg15 and BDNF (Chen et al. 2008). In a traumatic brain
injury model in rat neuron culture, GluN2ARs inhibition were shown to potentiate
caspase-3 activation, a marker of apoptotic signaling (DeRidder et al. 2006). In rat brain
slices and in-vivo rat studies, addition of GluN2AR receptor antagonists to disable
signaling capacity of GluN2ARs resulted in increased cell death, also corresponding to an
upregulation of caspase-3 activity after phencyclidine (PCP) treatment, providing further
evidence that GluN2ARs link to neuroprotective signaling during CNS insult (Anastasio
et al. 2009).
2.3.1 GluN2AR-PTEN-TDP43 pathway
A GluN2AR-PTEN-TDP-43 dependent pathway has been shown to protect
against neuronal injury (Zheng et al. 2012). In an extracellular glutamate accumulation
injury model GluN2AR stimulation, but not GluN2BR stimulation, triggered a reduction 9
in PTEN (phosphatase and tensin homolog) expression (Zheng et al. 2012). PTEN plays a
role in many pathological processes surrounding neuronal injury, such as those associated
with brain ischemia and neurological disease (Gary and Mattson 2002; Omori et al.
2002). Interestingly, PTEN localizes to both the nucleus and cytoplasm of neuronal and
glial cells (Lachyankar et al. 2000; Sano et al. 1999) and plays similar roles in both
locations in promoting apoptosis (Planchon et al. 2008). Recent studies have shown that
specific downregulation of the protein phosphatase activity of PTEN results in decreased
activity of GluN2BRs (Ning et al. 2004). Furthermore, studies by others and us have revealed that suppression of PTEN protects against neuronal death (Cantley and Neel
1999; Chang et al. 2007; Ning et al. 2004). Thus, downregulating protein phosphatase activity of PTEN may represent a novel pharmacological approach for treatment of CNS injury, by which GluN2BR-mediated neuronal death can be prevented without significantly interfering with NMDAR-mediated synaptic function (Hardingham et al.
2002; Vanhoutte and Bading 2003).
The dysfunction of TAR DNA-binding protein-43 (TDP-43) has been recently implicated in neurodegenerative diseases (Arai et al. 2006). However, the physiological and pathophysiological functional profiles of TDP-43 have not yet been fully elucidated.
Knock-down and deletion of nuclear TDP-43 has been shown to be detrimental to neuronal cells, potentiating neurodegenerative signaling while endogenous TDP-43 has been shown to be neuroprotective in the nucleus (Fiesel et al. 2010; Iguchi et al. 2009;
Zheng et al. 2012). A marked increase in TDP-43 expression in the nucleus has been linked to neuron cell viability during in-vitro neurodegenerative injury situations (Zheng 10
et al. 2012). The functional consequence of TDP-43 remains elusive because this protein
shuffles between the nucleus and cytoplasm (Ayala et al. 2008). Although TDP-43 may functionally serve as a neuroprotectant in the nucleus, TDP-43 plays a different role in the cytoplasm where its involvement in protein aggregate formation characterizes the pathogenesis in amyotrophic lateral sclerosis (ALS); (Barmada et al. 2010). It is possible
that increased TDP-43 expression in the nucleus can trigger neuroprotective signaling
pathways whereas its export to the cytoplasm may have more deleterious effects. In response to glutamate accumulation, the endogenous TDP-43, behaving as a pro-survival signaling protein in the nucleus, is only increased in the nucleus and does not translocate to cytoplasm. Interestingly, PTEN has been shown to negatively regulate TDP-43 expression in the nucleus, and the activation of GluN2ARs exerts its neuroprotective effects through suppression of PTEN and subsequent increase in nuclear TDP-43 whereas
GluN2BRs have no effect on PTEN expression (Zheng et al. 2012). This finding demonstrates a GluN2A-activation dependent signaling pathway, namely, the GluN2AR-
PTEN-TDP-43, to trigger neuroprotection through downregulation of PTEN, causing an augmentation in nuclear TDP-43 in a neurodegenerative model. However, evidence linking TDP-43 as being protective in other CNS pathologies such as traumatic brain injury and ischemic stroke has yet to be shown. Thus, delving deeper into the molecular mechanisms that regulate TDP-43 localization and how TDP-43 exerts its differential effects in the cell based on its locale may elucidate strong therapeutic targets to combat
CNS injury. 11
2.3.2 DJ-1-PTEN-PINK1-GluN2AR pathway
In rat cortical cultures, a DJ-1-PTEN-PINK1-GluN2AR signaling cascade has been found to confer neuroprotection (Chang et al. 2010). Study on DJ-1 gene deletion reveals that DJ-1 is an atypical peroxiredoxin-like peroxidase (Andres-Mateos et al.
2007) and loss-of-function mutations in DJ-1 have been identified in patients with early- onset autosomal recessive Parkinsonism (Bonifati et al. 2003), suggesting that dysfunction of DJ-1 may contribute to the dopaminergic neurodegeneration in
Parkinson’s disease (PD). More recent studies have demonstrated that DJ-1 is involved in stroke-induced brain injury (Aleyasin et al. 2007; Yanagisawa et al. 2008), indicating that
DJ-1 dysfunction may play a broad role in the CNS injury situations. DJ-1 suppression led to an increase in PTEN expression in rat cortical culture, and induced a neuroprotective effect by enhancing PINK1-GluN2AR signaling (Chang et al. 2010).
Although increase in PTEN activity has previously been linked to apoptotic signaling pathways, it also has been shown to induce an increase of PINK1 (PTEN- induced kinase 1; also called PARK6) expression and GluN2AR-mediated currents. The newly identified PINK1 gene encodes a serine⁄threonine kinase (Valente et al. 2004).
Similar to DJ-1, loss-of-function mutations in the PINK1 gene have been linked to early onset of PD (Healy et al. 2004; Valente et al. 2004). Inactivation of Drosophila PINK1 results in the progressive loss of dopaminergic neurons (Wang et al. 2006), and functional defects in mitochondria, causing mitochondrial calcium overload, increasing sensitivity to oxidative stress, reducing dopamine release and impairing synaptic plasticity (Gautier et al. 2008; Kitada et al. 2007). PINK1 is neuroprotective in both in 12
vitro and in vivo experimental models (Haque et al. 2008; Wood-Kaczmar et al. 2008).
Whole-cell patch-clamp recordings to measure GluN2AR- and GluN2BR-mediated components of NMDAR currents were performed, with the use of GluN2AR specific antagonist NVP-AAM077 and GluN2BR specific antagonist Ro 25-6981 (Fantin et al.
2008; Liu et al. 2007). In the neurons transfected with PINK1 cDNAs and siRNAs, the overexpression and suppression of PINK1, respectively, increased and inhibited
GluN2AR-mediated currents without effect on GluN2BR-mediated currents (Chang et al.
2010). These data indicate that the PINK1 regulation of NMDAR function resulted from altered GluN2AR activity. As GluN2AR-dependent signaling is believed to be neuroprotective, these results suggest a possibility that PINK1 dysfunction may promote neuronal death through inhibition of GluN2ARs.
In rat cortical cell culture studies, suppressing DJ-1 protein expression in neurons transfected with DJ-1 siRNAs increased GluN2AR mediated whole-cell currents (Chang et al. 2010). Conversely, DJ-1 overexpression in rat cortical neurons inhibited GluN2BR mediated currents (Chang et al. 2010). These results suggest DJ-1 might induce self-
protective signaling through increasing GluN2AR activity and suppressing GluN2BR
activity. Suppression of DJ-1 together with PINK1 knockdown, compared with
suppression of DJ-1 or PINK1 alone, significantly increases NMDAR-mediated neuronal death (Chang et al. 2010). Taken together, these results indicate that NMDAR function is
in part regulated by the DJ-1 -PTEN-PINK1-GluN2AR pathway. As a downstream signal of DJ-1, PINK1 may respond collaboratively to counteract DJ-1 dysfunction-induced 13
neuronal damage, which may delay neuronal death and possibly contributes to the slow
neurodegenerative process in PD.
2.4 GluN2BR in neuronal death and neurodegeneration
As a major player in NMDAR-mediated excitotoxicity, GluN2BR overactivation
during CNS injury couples to cellular death pathways via suppression of CREB-, ERK-,
and PINK1-dependent survival pathways (Martel et al. 2009; Shan et al. 2009; Wang et al. 2004a). A body of evidence provides data suggesting that calcium flux through extrasynaptic GluN2BRs play critical roles in modulating cell death pathways in contrast to GluN2ARs (Hardingham et al. 2002; Vanhoutte and Bading 2003). Dysfunction of the regulatory events governing the endocytosis and insertion of NMDARs to and from the plasma membrane, in both synaptic and extrasynaptic locations, may have deleterious effects on cell survival. Inhibition of GluN2BRs in rats revealed that cell death was decreased after ischemic insult and enhanced preconditioning-induced neuroprotection
(Chen et al. 2008), implicating their involvement in regulation of apoptosis. In cultured mouse cortical neurons, the protein phosphatase activity of PTEN acts as a crucial upstream signal to regulate GluN2BRs (Ning et al. 2004). The protein phosphatase activity of PTEN, through downregulating GluN2BRs, protects against ischemic neuronal death (Ning et al. 2004). GluN2BRs are found more readily in extrasynaptic sites and activation of these receptors inhibits nuclear signaling to CREB, reduces BDNF activity and plays a role in mitochondrial dysfunction ultimately potentiating cellular death
(Stocca and Vicini 1998). How GluN2BR-interacting signaling is involved in neurodegeneration and neuronal death remains to be determined. 14
2.4.1 DJ-1-PTEN-GluN2BR pathway
Recent study indicates that a DJ-1-PTEN-GluN2BR-mediated pathway promotes cell death (Chang et al. 2010). Suppressing DJ-1 protein expression in neurons transfected with DJ-1 siRNAs revealed an increase in not only the GluN2AR-mediated currents but also the GluN2BR-currents (Chang et al. 2010). These results suggest that
DJ-1 dysfunction, while inducing neuronal death through enhancing GluN2BR- dependent cell death signaling, might also promote self-protective signaling through increasing GluN2AR function. The regulation of NMDAR function by DJ-1 could be
attributable to the altered expression of NMDARs on the cell surface (Carroll and Zukin
2002). Neurons with reduced DJ-1 expression through transfection with DJ-1 siRNAs
resulted in an increased surface expression of GluN2B but not GluN2A subunits (Chang
et al. 2010). DJ-1 cDNA transfection led to a decreased surface expression of GluN2BRs
in cortical neurons (Chang et al. 2010). Because the total GluN2B subunit levels were not altered by DJ-1 knockdown or overexpression, the altered surface expression of GluN2B could be the result of an altered delivery and/or internalization of NMDARs. Thus, the post-transcriptional mechanisms may mediate the DJ-1 regulation of GluN2BR surface expression (Chang et al. 2010). As the phosphatase PTEN positively regulates the function of GluN2BRs (Chang et al. 2007; Ning et al. 2004) and DJ-1 is a negative regulator of PTEN (Kim et al. 2005a), PTEN might contribute to DJ-1 knockdown- induced enhancement of GluN2BR function. Indeed, Western blots confirmed that DJ-1 knockdown resulted in an increased protein expression of PTEN in the neurons and that
PTEN inhibitor resulted in a reduction of DJ-1 suppression-induced increase in GluN2BR 15
currents (Chang et al. 2010). Thus, DJ-1 knockdown-induced potentiation of GluN2BR function is mediated in part by increased PTEN expression, which promotes neuronal death and neurodegeneration (Chang et al. 2010).
2.4.2 GluN2BR-PINK1-Akt pathway
GluN2BR can suppress PINK1-Akt pathway to enhance neurodegeneration and neuronal death. To test whether PINK1, in addition to being involved in the pathogenesis of Parkinson’s disease (Valente et al. 2004), plays a role in ischemic neuronal death, the
protein expression of PINK1 in oxygen-glucose deprivation (OGD) conditions in vitro
was observed (Shan et al. 2009). The amount of PINK1 protein was decreased in neurons
exposed to OGD compared with that in control neurons without OGD treatment and the
levels of neuronal injury was increased with the extension of OGD treatment time (Shan et al. 2009). This result suggests that the reduction in protein expression of PINK1 may be involved in ischemic neuron injury. Because inhibition of NMDARs by its antagonist or channel blocker MK-801 significantly reduced OGD induced reduction of PINK1 expression (Shan et al. 2009), the overactivation of NMDARs may be responsible in part
for OGD-induced PINK1 suppression. As overactivation of GluN2BRs plays the major
role in NMDAR excitotoxicity-mediated neuronal death (Chang et al. 2007; Hardingham
et al. 2002; Ning et al. 2004), overactivation of NR2BRs may be involved in OGD- induced PINK1 reduction. By treating neuronal cultures with a GluN2BR antagonist during OGD/reoxygenation, results indicated that OGD-induced reduction of PINK1 expression was inhibited by GluN2BR antagonist (Shan et al. 2009). However, inhibition
of GluN2ARs exhibited no significant effect on OGD-induced decrease of PINK1 16
expression (Shan et al. 2009). Interestingly, the phosphorylation of Akt, a known neuroprotectant, was inhibited while PINK1 protein level was reduced in neurons treated with OGD. The reduction in the levels of Akt phosphorylation was partially recovered by inhibition of GluN2BRs. These results suggest that GluN2BR overactivation triggers a reduction in PINK1 and mediates cell death signaling in part through the suppression of the neuroprotective AKT pathway. 17
Figure 2.1 GluN2ARs and GluN2BRs mediate neuronal survival and death pathways during CNS insult.
Two GluN2AR mediated cell-survival promoting pathways have been
mentioned. GluN2AR-PTEN-TDP-43: Activation of GluN2AR confers
neuroprotection through downregulation of nuclear PTEN and enhancing
function of TDP-43. DJ-1-PTEN-PINK1-GluN2AR: DJ-1 negatively
regulates PTEN function inducing neuroprotection by increasing PINK1
expression in the cytoplasm. PINK1 can positively regulate both GluN2AR
activity and the Akt pathway to promote cell survival. Two GluN2BR 18
mediated pathways regulating cell death have been mentioned. DJ-1-PTEN-
GluN2BR: DJ-1 negatively regulates PTEN function to play a dual role in survival/death by increasing GluN2BR activity. GluN2BR-PTEN-Akt:
GluN2BR activation can negatively regulate PINK1 function and suppress the activation of neuroprotective Akt pathway. 19
2.5 Discussion
The NMDAR plays a major role in excitotoxicity-mediated neuronal death and neurodegeneration in various neurological disorders. However, clinical trials using
NMDAR antagonists have had disappointing outcomes. We now believe that while
NMDAR antagonists reduce GluN2BR-induced neuronal death, the GluN2AR-mediated neuroprotective effect is also suppressed. Thus, uncovering the cellular and molecular mechanisms that specifically link to GluN2AR-mediated neuroprotection or GluN2BR- dependent cell death-promoting signal pathways would provide a molecular basis to develop potent therapeutic strategy. This review discusses the recent progress in understanding how GluN2AR and GluN2BR interact with intracellular signaling to exert their opposing effects.
20
Chapter 3 Glycine Triggers a Non-ionotropic Activity of GluN2A-containing NMDA
Receptors to Confer Neuroprotection
Rong Hu*, Juan Chen*, Brendan Lujan, Tianyuan Cui, Mi Zhang, Zefen Wang, Mingxia
Liao, Zhiqiang Li, Yu Wan, Hongliang Li, Fang Liu, Hua Feng & Qi Wan
*These authors contributed equally to this work
Brendan Lujan contributed to the data presented in Figures 3.1-5.
21
3.1 Summary
Ionotropic activation of NMDA receptors (NMDARs) requires agonist glutamate
and co-agonist glycine. The subtypes of NMDARs play different roles in neuronal
survival and death. Here we show that in the cultured cortical neurons and the HEK293
cells expressing different combinations of NMDARs, glycine alone enhances the
phosphorylation of cell survival-promoting kinase Akt independent of the activation of
glycine receptors or the channel activity of GluN2A subunit-containing NMDARs
(GluN2ARs). The effect of glycine is sensitive to the antagonist of glycine-GluN1 binding site but not that of glutamate-GluN2A binding site. These findings suggest that a non-ionotropic activation of GluN2ARs by glycine results in the increase of Akt phosphorylation. We further show that glycine treatment protects against glutamate neurotoxicity-induced neuronal death through the non-ionotropic activity of GluN2ARs
in cultured cortical neurons. The neuroprotective effect of glycine is blocked by Akt
inhibition. Consistent with the in vitro findings, in an animal model of ischemic stroke,
glycine reduces the infarct area and improves functional recovery independent of glycine
receptor or NMDAR channel activity. Together, our study identifies a non-ionotropic
function of GluN2ARs activated by glycine, which leads to the enhancement of Akt
activation to confer neuroprotection.
3.2 Introduction
The NMDAR is a subtype of ionotropic glutamate receptors that mediate the vast
majority of excitatory neurotransmission in the mammalian central nervous system
(CNS) (Dingledine et al. 1999). NMDARs are ligand-gated Ca2+-permeable channels 22
that consist of GluN1, GluN2 (GluN2A-GluN2D) and GluN3 (GluN3A-GluN3B)
subunits (Monyer et al. 1992). The GluN2A- and GluN2B-containing NMDARs
(GluN2ARs and GluN2BRs) are the major combinations of NMDARs expressed in CNS
(Dingledine et al. 1999). The binding of agonist glutamate to GluN2 subunits and co-
agonist glycine to GluN1 subunits is required to activate GluN2ARs and GluN2BRs
(Johnson and Ascher 1987), which play essential roles in synaptic plasticity (Barria and
Malinow 2002; Malenka and Nicoll 1999), neural development (Constantine-Paton et al.
1990; Kerchner and Nicoll 2008) and glutamate-induced neurotoxicity (Aarts et al. 2002;
Choi 1988; Martel et al. 2009).
Different GluN2 subunits confer distinct roles of NMDAR subtypes and link them
with different intracellular signaling pathways (Hayashi et al. 2009; Kim et al. 2005a;
Loftis and Janowsky 2003). Previous evidence suggests that GluN2BR-mediated
neurotoxicity induces neuronal death (Chen et al. 2008; Liu et al. 2007; Ning et al. 2004),
and that enhancement of GluN2AR activity promotes neuronal survival (Anastasio et al.
2009; DeRidder et al. 2006; Liu et al. 2007). However, the molecular mechanisms
underlying the differential effects of GluN2ARs and GluN2BRs in neuronal survival and
death are not fully understood.
While it is well known for its ionotropic function, increasing evidence indicates
that the NMDAR has non-ionotropic activity (Birnbaum et al. 2015; Kessels et al. 2013;
Nabavi et al. 2013; Stein et al. 2015; Tamburri et al. 2013; Vissel et al. 2001). For example, recent study indicates that ligand binding to NMDARs is sufficient to induce long-term depression (LTD), but does not require ion flow through NMDARs (Nabavi et 23
al. 2013). A non-ionotropic activity is found to be mediated through GluN2BR and is
required for β-amyloid–induced synaptic depression (Kessels et al. 2013; Tamburri et al.
2013). The ion-ionotropic activity of NMDAR signaling is also shown to drive structural shrinkage at spiny synapses (Stein et al. 2015). In the present study, we reveal that glycine alone elicits a non-ionotropic activity of GluN2ARs but not GluN2BRs. We demonstrate that glycine through non-ionotropic activation of GluN2ARs and subsequent enhancement of Akt activation confers neuroprotection.
3.3 Methods
General methods
Randomization was used to assign samples to the experimental groups, and to collect and process data. All animal experiments were approved and carried out in compliance with the guidelines of University of Nevada and Wuhan University School of
Medicine.
Neuronal culture
The cortical neuronal cultures were prepared from female C57BL/6 mice and
Sprague-Dawley rats at gestation day 17 as described (Brewer et al. 1993; Shan et al.
2009). Briefly, dissociated neurons were suspended in plating medium (Neurobasal medium, 2% B-27 supplement, 0.5% FBS, 0.5 μM L-glutamine, and 25 μM glutamic acid) and plated on poly-D-lysine coated Petri dishes. After 1 day in culture, half of the plating medium was removed and replaced with maintenance medium (Neurobasal medium, 2% B-27 supplement, and 0.5 μM L-glutamine). Thereafter, maintenance 24
medium was changed in the same manner every 3 days. The cultured neurons were used for experiments at 12 days after plating.
HEK293 cell culture, plasmids and transfections
HEK293 cells were grown in RPMI 1640 medium (Life Technologies, Grand
Island, NY) supplemented with 10% FBS and Pen/Strep (10 µg/ml). The plasmids of
GFP, GluN1, GluN2A, GluN2B, GluN1(N598Q), and GluN1(N598R) were transfected in cultured HEK293 cells. Transfections were performed using Lipofectamine 2000
(Invitrogen, Carlsbad, CA) as described in our previous studies (Ning et al. 2004; Shan et al. 2009). DNA-Lipofectamine complexes were made in serum-free medium Opti-MEM.
To prevent NMDAR-induced cell death, the transfected HEK293 cells were treated with
1.0 mM DAPV (Anegawa et al. 2000).
GluN2A and GluN2B shRNA lentiviral particles were purchased from Santa Cruz
Biotech (Santa Cruz, CA). These lentiviral particles contain three to five expression constructs each encoding target-specific 19-25 nt (plus hairpin) shRNA designed to knockdown gene expression and are provided as transduction-ready viral particles. The transduction of lentiviral particles was performed in cultured cortical neurons based on the manufacturer’s instructions (Santa Cruz Biotech).
Western blotting
Western blotting assay was performed as described previously (Ning et al. 2004;
Shan et al. 2009). For the detection of phospho-Akt, the samples prepared in the same day were used. The polyvinylidene difluoride membrane (Millipore, Bedford, MA, USA) was incubated with primary antibody against phospho-Akt (Ser473) (Cell Signaling 25
Technology, Beverly, MA), Akt (Cell Signaling), phospho-p38-MAPK (Cell Signaling), p38-MAPK (Cell Signaling), GluN1 (Millipore), GluN2A (Santa Cruz Biotech), or
GluN2B (Santa Cruz Biotech). Primary antibodies were labeled with horseradish
peroxidase-conjugated secondary antibody, and protein bands were imaged using
SuperSignal West Femto Maximum Sensitivity Substrate (Pierce, Rockford, IL, USA).
The EC3 Imaging System (UVP, LLC, Upland, CA) was used to obtained blot images
directly from the polyvinylidene difluoride membrane. For the detection of total Akt, the
same polyvinylidene difluoride membrane was stripped and then reprobed with primary
antibody against total Akt (Cell Signaling Technology). The quantification of Western
blot data was performed using ImageJ software.
Immunocytochemical staining
The assay was performed as described in our previous study (Liu et al. 2006).
Briefly, the cells were fixed with 4% paraformaldehyde in phosphate-buffered saline
(PBS) for 30 min at 25oC, permeabilized (0.25% Triton X-100, 10 min) and blocked in
5% normal goat serum in PBS for 60 min. The cells were incubated with anti-GluN2A
antibody (Santa Cruz Biotech) diluted in 3% normal goat serum in PBS overnight at 4 oC,
and then incubated with fluorochrome-conjugated secondary antibody Alexa Fluor 488
(Invitrogen) diluted in PBS for 1 h at 25 oC. Fluorescent-labeled proteins were imaged
using a 63x objective mounted on a confocal microscope as described previously (Liu et
al. 2006). Images were acquired in the linear range with constant settings and were
analyzed using ImageJ software (Liu et al. 2006). The average fluorescence intensity per 26
unit area was measured. Cells from five separate cultures each were analyzed. The n
value refers to the number of cells analyzed.
Neuronal viability assays
Double staining of propidium iodide (PI) and fluorescein diacetate (FDA) was
performed to detect neuronal viability using a modified procedure (Jones and Senft
1985). Briefly, cultures were rinsed with extracellular solution and incubated with FDA
(5 μM) and PI (2 μM) for 30 min. The cultures were washed with extracellular solution
and then viewed on an Olympus fluorescent microscope (IX51, Olympus). Neuronal
viability was determined by calculating the number of PI-labeled cells over FDA-labeled cells. The investigator for the cell count was blinded to the experimental treatment.
The lactate dehydrogenase (LDH) is a cytoplasmic enzyme retained by viable cells with intact plasma membranes and released from cells with damaged membranes.
The LDH release was measured using CytoTox 96 Cytotoxicity kit based on the manufacturer’s instructions (Promega, Madison, WI) (Shan et al. 2009). The levels of maximal LDH release were measured by treating the cultures with 10× lysis solution
(provided by the manufacturer) to yield complete lysis of the cells. Absorbance data were obtained using a 96-well plate reader (Molecular Devices, Palo Alto, CA) at 490 nm.
According to the manufacturer’s instructions, the LDH release (%) was calculated by
calculating the ratio of experimental LDH release to maximal LDH release.
Focal cerebral ischemia and infarct measurement
Transient focal cerebral ischemia was induced using the suture occlusion
technique (Longa et al. 1989). Male Sprague-Dawley rats weighing 250–300 g were 27
anesthetized with 4% isoflurane in 70% N2O and 30% O2 using a mask. A midline
incision was made in the neck, the right external carotid artery (ECA) was carefully
exposed and dissected, and a 3-0 monofilament nylon suture was inserted from the ECA
into the right internal carotid artery to occlude the origin of the right middle cerebral
artery (MCA) (approximately 22 mm). After 90 minutes of occlusion, the suture was removed to allow reperfusion, the ECA was ligated, and the wound was closed. Sham- operated rats underwent identical surgery and/or intracerebroventricular injections except that the suture was inserted and withdrawn immediately. Rectal temperature was
maintained at 37.0 ± 0.5°C using a heating pad and heating lamp. Rats were killed at
various times after reperfusion after being anesthetized, and the brains were removed for
TTC (2,3,5 -triphenyltetrazolium chloride) staining (Wexler et al. 2002). The brain was
placed in a cooled matrix and 2 mm coronal sections were cut. Individual sections were
placed in 10 cm petri dishes and incubated for 30 min in a solution of 2% TTC in
phosphate buffered saline at 37 °C. The slices were fixed in 4% paraformaldehyde at 4
°C. All image collection, processing and analysis were performed in a blind manner and
under controlled environmental lighting. The scanned images were analyzed using
ImageJ software and the infarct data for all groups were expressed as the ratios of the
infarcted areas to the total brain section areas (Wexler et al. 2002).
Intracerebroventricular administration
For intracerebroventricular injections, the rats were placed on ear bars of a
stereotaxic instrument under anesthesia. Drug infusion to the cerebral ventricle (from the
bregma: anteroposterior, ± 0.8 mm; lateral, 1.5 mm; depth, 3.5 mm) was performed using 28
a 23-gauge needle attached via polyethylene tubing to a Hamilton microsyringe at a rate
of 1.0 μl/ min. Proper needle placement was verified via withdrawing a few microliters of
clear cerebrospinal fluid into the Hamilton microsyringe.
Neurological Severity Scores
The rats were subjected to a modified neurological severity score (mNSS) test as reported previously (Table 1); (Chen et al. 2001). These tests are a battery of motor, sensory, reflex, and balance tests, which are similar to the contralateral neglect tests in humans. Neurological function was graded on a scale of 0 to 18 (normal score, 0; maximal deficit score, 18).
Beam walk test
The beam walk test measures the animals’ complex neuromotor function
(Aronowski et al. 1996; Petullo et al. 1999). The animal was timed as it walked a (90 x 4
x 1.5 cm) beam. A box for the animal to feel safe was placed at one end of the beam. A
loud noise was created to stimulate the animal to walk toward and into the box
(Aronowski et al. 1996; Petullo et al. 1999). Scoring was based upon the time it took the
rat to go into the box (Petullo et al. 1999). The higher the score, the more severe is the neurological deficit (Table 2).
Adhesive-removal test
A modified sticky-tape (MST) test was performed to evaluate forelimb function
(Sughrue et al. 2006). A sleeve was created using a 3 × 1-cm piece of yellow paper tape and was subsequently wrapped around the forepaw so that the tape attached to itself and allowed the digits to protrude slightly from the sleeve. The typical response is for the rat 29
to vigorously attempt to remove the sleeve by either pulling at the tape with its mouth or
brushing the tape with its contralateral paw. The rat was placed in its cage and observed
for 30 s. Two timers were started: the first ran without interruption and the second was
turned on only while the animal attempted to remove the tape sleeve. The ratio of the left
(affected)/right (unaffected) forelimb performance was recorded. The contralateral and
ipsilateral limbs were tested separately. The test was repeated three times per test day,
and the best two scores of the day were averaged. The lower the ratio, the more severe is
the neurological deficit.
Statistics
Student’s T test or ANOVA test was used where appropriate to examine the
statistical significance of the differences between groups of data. Newman–Keuls tests
were used for post-hoc comparisons when appropriate. All results are presented as mean
± SE. Significance was placed at p < 0.05.
3.4 Results
3.4.1 Glycine increases Akt phosphorylation independent of Ca2+ influx through
NMDAR channels
To test the effect of glycine on the activation of cell survival-promoting kinase
Akt (protein kinase B) in cultured mouse cortical neurons in which the channel activity of
NMDARs was completely inactivated, a non-competitive antagonist MK-801 that prevents the flow of ions through the NMDAR channels was used (MacDonald and
Nowak 1990; Rosenmund et al. 1993). To ensure that no Ca2+ passed through NMDAR
channels, MK-801 was added into a extracellular solution (ECS) in which Ca2+ was not 30
included but with the addition of 5 mM Ca2+ chelator EGTA. We named this specific
Ca2+-free ECS as ECS-1 (10 μM MK-801, 5 mM EGTA, 137 mM NaCl, 5.4 mM KCl, 1
mM MgCl2, 25 mM HEPES, 33 mM Glucose, titrated to pH 7.4 with osmolarity of 300-
320 mOsm). Since MK-801 is a use-dependent pharmacological agent, we first treated
the neurons with NMDA (1 μM) and glycine (1 μM) for 1 min that opened the NMDAR
channels and allowed MK-801 in the ECS-1 to fully block NMDARs (Lu et al. 2001;
MacDonald and Nowak 1990; Rosenmund et al. 1993). The cultured neurons were then washed with ECS-1 for three times (10 min wash/each). As shown in Figure 3.1A, this treatment will be referred to as NMDAR channel inactivation procedure.
The activation of Akt was quantified by measuring Akt phosphorylation (p-Akt) on Ser473 in Western blot assay (Luo et al. 2003; Shan et al. 2009). The levels of p-Akt
were quantified by calculating the ratio of p-Akt to total Akt (t-Akt). After the channel activity of NMDARs was suppressed by the NMDAR channel inactivation procedure as shown in Figure 3.1A, the cortical cultures were treated with ECS-1 containing glycine
(100 µM) for 30 min (+Gly group in Figure 3.1B). For the control, the cultures were treated with ECS-1 for 30 min without glycine (-Gly group in Figure 3.1B). We showed that the glycine (100 µM) treatment increased Akt phosphorylation in cortical neurons where the NMDAR channel activity was inhibited (Figure 3.1B). In the same experimental conditions as in Figure 3.1B, the effect of glycine on Akt phosphorylation was dose-dependent (Figure 3.1C). These data indicate that glycine enhances Akt activation in a NMDAR channel activity-independent manner. 31
To provide further evidence that the effect of glycine on Akt phosphorylation was
independent of extracellular Ca2+, we tested the effect of BAPTA, a Ca2+ chelator that
has faster Ca2+-binding kinetics than EGTA (Adler et al. 1991). The experimental
condition was same as that in Figure 3.1B, but BAPTA (0.1, 1 or 5 mM) was included in
the ECS-1 in the +BAP groups (Figure 3.1D). BAPTA was included both during the
NMDAR blockade procedure and during treatment with glycine. Compared with the
group without BAPTA treatment, BAPTA treatment did not interfere with glycine-
induced elevation of Akt phosphorylation in cortical neurons where the NMDAR channel
activity was inhibited (Figure 3.1D).
3.4.2 Elevation of Akt phosphorylation by glycine does not depend on the activation
of glycine receptors
Glycine is the agonist for strychnine-sensitive glycine receptors. Glycine
receptors are expressed in the developing cortex, but not expressed in the mature cortex
(Flint et al. 1998; Lynch 2004). To exclude the possible effect of glycine receptors on the
observed enhancement of Akt activation by glycine, we used the same experimental
design as that in Figure 3.1B, but strychnine was added into the ECS-1 for the treatment
in both –Gly and +Gly groups in Figure 3.2A. Our data showed that strychnine failed to
block the enhancement of Akt phosphorylation by glycine (100 µM) in cortical neurons
where NMDARs are inhibited by the NMDAR channel inactivation procedure (Figure
3.2A). These data indicate that the glycine-induced enhancement of Akt phosphorylation
does not depend on the activation of strychnine-sensitive glycine receptors. 32
To determine whether glycine plays a similar role in rat cortical neurons, we performed the same experiment in cultured rat cortical neurons as that as in Figure 3.2A.
We found that glycine (100 µM) enhanced Akt phosphorylation independent of glycine receptors and the activation of NMDAR channels (Figure 3.2B).
The p38-MAPK is implicated in NMDAR-dependent LTD (Zhu et al. 2002), and was shown recently to be activated by non-ionotropic NMDAR signaling after chemical
LTD induction (Nabavi et al. 2013; Stein et al. 2015). The activation of p38-MAPK is also involved in promoting excitotoxicity (Li et al. 2013). To determine whether glycine altered p38-MAPK signaling independent of glycine receptors and the activation of
NMDAR channels, we tested the effect of glycine on p38-MAPK phosphorylation in cortical neurons following the experimental procedure described in Figure 3.2A. Our results showed that glycine (100 µM) had no significant effect on p38-MAPK phosphorylation in our experimental conditions (Figure 3.2C), suggesting a specific activation of Akt but not p38-MAPK by glycine in a condition where NMDAR channel activities and glycine receptors were suppressed.
3.4.3 Glycine alone enhances Akt activation through a non-ionotropic activation of
GluN2ARs
Our results thus far suggest a non-ionotropic activity of NMDAR to mediate the potentiation of Akt activation by glycine. To provide direct evidence for this possibility, we measured the effects of glycine on Akt phosphorylation in HEK293 cells transiently expressing NMDARs. The cDNAs of GluN1, GluN2A and/or GluN2B subunits were transfected in various combinations into the HEK293 cells (Wan et al. 1997). The 33
channel activities of NMDARs expressed in the transfected cells were inhibited by the
NMDAR channel inactivation procedure as described in Figure 3.1A. Treatment of glycine (100 µM) for 30 min had no effect on Akt phosphorylation in both non- transfected HEK293 cells and the cells transfected with cDNAs of green fluorescence protein (GFP) (Figure 3.3A). However, glycine increased Akt phosphorylation in
HEK293 cells transfected with cDNAs of GluN1+GluN2A following NMDAR channel
inactivation procedure (Figure 3.3B), but not in cells transfected with cDNAs of
GluN1+GluN2B (Figure 3.3C). We also found that glycine did not increase Akt phosphorylation in HEK293 cells transfected with cDNAs of GluN1, GluN2A and
GluN2B, respectively (Figure 3.3D). Together, these results indicate that a non- ionotropic activity of GluN2ARs mediates the elevation of Akt phosphorylation by
glycine.
Amino acid N598 in the GluN1 subunit is a critical residue at the selectivity filter
of the NMDAR channel that determines Ca2+ permeability (Burnashev et al. 1992). The
GluN1(N598Q) mutant and GluN1(N598R) mutant have been shown to cause decreased
Ca2+ permeability of NMDAR channels (Behe et al. 1995; Burnashev et al. 1992; Single
et al. 2000). To test the effect of GluN1(N598Q) and GluN1(N598R) on the enhancement
of Akt activation by glycine, we transfected GluN1(N598Q), GluN2A + GluN1(N598Q),
GluN1(N598R), GluN2A + GluN1(N598R) in HEK293 cells that were treated with
standard ECS. As shown in Figure 3.3E-F, glycine (100 µM) increased Akt
phosphorylation in HEK293 cells transfected with GluN2A + GluN1(N598R) or GluN2A 34
+ GluN1(N598Q). These results provide molecular evidence to support the conclusion
that GluN2AR-mediated Akt activation is independent of Ca2+ influx.
To validate the role of a non-ionotropic activity of GluN2AR in mediating the
enhancement of Akt activation by glycine in cortical neurons, we applied a GluN2A
knockdown approach. The GluN2A protein expression was suppressed in the cultured
cortical neurons transducted with GluN2A shRNA lentiviral particles (Figure 3.3G). The
same experimental design as that in Figure 3.1A was applied to inhibit NMDARs. As shown in Figure 3.3H, glycine (100 µM) increased Akt phosphorylation in neurons transducted with shRNA control, but the effect of glycine was significantly reduced in neurons transducted with GluN2A shRNA. As another control of GluN2A shRNA against GluN2A, the GluN2B shRNA had no influence on the observed effect of glycine
(Figure 3.3I-J). These results lead us to conclude that glycine enhances Akt activation
through a non-ionotropic activity of GluN2ARs in cortical neurons.
3.4.4 The glycine-GluN1 binding site mediates the non-ionotropic activation of
GluN2ARs
To determine how glycine exerts its effect through the non-ionotropic activation
of GluN2ARs, we tested the effects of three NMDAR inhibitors, the glycine-GluN1
binding site antagonist L-689560, the competitive GluN2 antagonist DAPV and the
GluN2B antagonist Ro 25-6981 (Baptista and Varanda 2005; Chang et al. 2010; Fischer
et al. 1997; Vignes and Collingridge 1997), on the glycine-induced Akt activation after the channel activities of NMDARs were blocked. The L-689560, DAPV or Ro 25-6981
was included in the ECS-1 in the wash step of the NMDAR channel inactivation 35
procedure and the step of glycine treatment (Figure 3.1A). The cultures were then treated with ECS-1 containing glycine (100 µM) and one of the three inhibitors for 30 min. We showed that after the channel activities of NMDARs were blocked, L-689560 (50 µM) blocks glycine-induced Akt phosphorylation in the cultured neurons and the HEK293 cells transfected with GluN1+GluN2A (Figure 3.4A-B). But DAPV (50 µM) did not interfere with the enhancement of Akt phosphorylation by glycine (Figure 3.4C-D).
These data suggest that the glycine-GluN1 binding, but not glutamate-GluN2A binding, is required for the non-ionotropic activation of GluN2ARs. As a control experiment, we also tested the effect GluN2B antagonist Ro 25-6981 (5.0 µM). Our data showed that Ro
25-6981 had no significant effect on glycine-induced Akt activation in neurons and
HEK293 cells transfected with GluN1+GluN2A after NMDAR channel activity was inhibited (Figure 3.4E-F).
D-serine is the endogenous agonist of glycine-GluN1 binding site (Oliet and
Mothet 2009). We tested the role of D-serine in Akt phosphorylation in the cortical neurons and the HEK293 cells transfected with GluN1+GluN2A or GluN1+GluN2B following NMDAR channel inactivation procedure. D-serine increased Akt phosphorylation in both cortical neurons and HEK293 cells transfected with cDNAs of
GluN1+GluN2A but not with those of GluN1+GluN2B (Figure 3.4G-I), further implying the role of glycine-GluN1 binding in mediating the effect of non-ionotropic activation of
GluN2ARs. 36
3.4.5 Glycine prevents glutamate neurotoxicity-induced neuronal death through
non-ionotropic activation of GluN2ARs
As Akt is a survival-promoting kinase that plays a crucial role in preventing
neuronal death (Burke 2007; Luo et al. 2003; Manning and Cantley 2007), we measured the effect of non-ionotropic activation of NMDARs by glycine on Akt phosphorylation in
glutamate neurotoxicity-induced neuronal injury. The injury was produced by treating the
cultured cortical neurons with standard ECS containing glutamate (100 μM) and glycine
(1 μM) for 1 h (Figure 3.5A). To block the channel activities of NMDARs, following 1 h
injury and 30 min wash with standard ECS, the cultures were treated with standard ECS
containing 10 µM MK-801 for 23.5 h. For the Control group (Con; Figure 3.5B-F), the
culture was treated with maintenance medium. For the Sham group (Figure 3.5B-F), the
culture was treated with standard ECS for 25 h. For the injury group (Inj; Figure 3.5B-F),
the cultures were treated with standard ECS for 24 h following the injury by glutamate
(100 μM) + glycine (1 μM) for 1.0 h. For the group of glycine, MK-801 or MK-
801+glycine treatment in injured cultures (Inj+Gly, Inj+MK or Inj+MK+Gly; Figure
3.5B-D), following the 1 h injury the cultures were first washed with standard ECS containing MK-801 (10 µM) for three times (10 min wash/each), and then treated with
standard ECS containing glycine (100 μM), MK-801 (10 µM) or MK-801 (10 µM)+ glycine (100 μM) for 23.5 h. For the group of MK-801 treatment in uninjured cultures
(MK; Figure 3.5B-D), the cultures were treated with standard ECS containing MK-801
(10 μM) for 25 h. The Double labeling of propidium iodide (PI) and fluorescein diacetate
(FDA) was also performed to measure neuronal viability (Jones and Senft 1985). The 37
levels of lactate dehydrogenase (LDH) released from injured neurons was measured to
quantify the neuronal damage (Shan et al. 2009). Our data showed that after glutamate neurotoxicity insult, glycine (100 μM) treatment protected against the death of cortical
neurons in which the NMDARs were inactivated (Figure 3.5B-D).
Our results further demonstrated that the neuroprotective effect of glycine (100
μM) was reduced in injured cortical neurons in which the GluN2A expression was
suppressed by GluN2A shRNA (Figure 3.5E-F). The neurons in both shRNA control and
GluN2A shRNA groups were subjected to the same experimental procedures described in
Figure 3.5B-D. We conclude that the neuroprotective effect of glycine is at least in part
mediated through non-ionotropic activation of GluN2ARs in glutamate neurotoxicity- induced neuronal injury.
To determine the roles of Akt activation and glycine-GluN1 binding in glycine- induced neuroprotection, we tested the effect of Akt inhibitor IV and glycine-GluN1 binding antagonist L-689560 in our experimental model. The experimental condition was the same as that described in Figure 3.5A-D. The IV and L-689560 were included in both wash and treatment steps. We found that both IV (1 µM) and L-689560 (50 µM) significantly reduced glycine-induced neuroprotective effect in neurons where the channel activity of NMDARs were inhibited (Figure 3.5G-H; Inj+MK+IV+Gly vs.
Inj+MK+Gly; Inj+MK+L-689560+Gly vs. Inj+MK+Gly). Thus, Akt activation and glycine-GluN1 binding mediate glycine-induced neuroprotection that is mediated through non-ionotropic activation of GluN2ARs. Since IV and L-689560 do not completely block glycine-induced neuroprotection, additional signal pathways may mediate the 38
neuroprotective effect of glycine. An alternative interpretation would be that these drugs
or doses were inadequate to fully block the targets in injury model.
3.4.6 The neuroprotective role of non-ionotropic activation of NMDARs by glycine
in ischemic stroke
Given that glutamate-induced neurotoxicity is a general injury mechanism underlying ischemic/traumatic brain injuries and a variety of neurodegenerative diseases
(Aarts et al. 2002; Dingledine et al. 1999), we tested whether the non-ionotropic activation of NMDARs by glycine conferred neuroprotection in a clinic-relevant rat model of ischemic stroke, the middle cerebral artery occlusion (MCAO); (Sun et al.
2003). At 1.5, 3, 6, 9 or 12 h following ischemia reperfusion, glycine (100 µg/100 g) was administered into the lateral ventricles according to previous reports (De Sarro et al.
2000; Liu et al. 2007; Williams et al. 1995). To suppress the channel activity of
NMDARs and the activation of glycine receptors, at 30 min before glycine injection we
injected MK-801 (8 µg/100 g) and strychnine (1.2 µg/100 g) into the lateral cerebral
ventricles as described previously (Covasa et al. 2004; Williams et al. 1995; Zarrindast et
al. 2006). Treatment of glycine at 1.5, 3 or 6 h after ischemic reperfusion significantly
decreased the infarct area at 24 h after ischemia onset compared with the groups with the
injection of MK-801+strychnine at the same time points (Figure 3.6A-B).
A battery of neurobehavioral tests including modified neurological severity scores
(mNSS) test, beam-walking test and modified sticky-tape (MST) test were performed to
further test the neuroprotective role of glycine (Table 3.1-2); (Chen et al. 2001; Clifton et
al. 1991; Sughrue et al. 2006). The neurological function of stroke animals was evaluated 39
one day before MCAO, and 1, 3, 7 and 14 days after MCAO. Glycine (100 µg/100 g, icv) was injected at 3 h after ischemic reperfusion (De Sarro et al. 2000; Liu et al. 2007;
Williams et al. 1995), and at 30 min prior to glycine injection we injected MK-801 (8.0
µg/100 g) and strychnine (1.2 µg/100 g) into the lateral cerebral ventricles (Covasa et al.
2004; Williams et al. 1995; Zarrindast et al. 2006). These tests were performed by the investigator who was blinded to the experimental groups. Our data showed that compared with rats treated with strychnine and MK-801 treatment (I/R+Stry+MK group), rats treated with glycine (I/R+Stry+MK+Gly group) had significantly lower scores of mNSS test at day 7 and 14 after MCAO (Figure 3.7A), lower scores of beam-walking test at day
3, 7 and 14 after MCAO (Figure 3.7B), and higher ratio of MST test at day 7 and 14 after
MCAO (Figure 3.7C). Together, these results provide functional evidence for the role of non-ionotropic activity of NMDARs in mediating the neuroprotective effect of glycine.
As described above, the injection doses of glycine, MK-801, strychnine, IV and L-
689560 were determined based on both previous reports and our test using multiple doses of these drugs in the MCAO model. 40
Figure 3.1 Enhancement of Akt phosphorylation by glycine in cortical neurons does not require the channel activities of NMDARs.
(A) A schematic diagram showing the NMDAR channel inactivation and
glycine treatment procedure. (B) Glycine (100 µM) increases Akt
phosphorylation (p-Akt) in neurons where NMDAR channel activities are
inhibited (n=9, Student’s T test, *p<0.05 vs. –Gly). (C) Glycine-induced
increase of p-Akt is dose-dependent in neurons where NMDAR channel 41
activities are inhibited (n=6, ANOVA test, *p<0.05 vs. control). (D) The enhancement of p-Akt by glycine (100 µM) is not altered by BAPTA that was included in the ECS-1 (n = 6, ANOVA test, *p<0.05 vs. -BAP). The p-
Akt analyses were normalized to group (1) labeled in the bar graphs unless described elsewhere.
42
Figure 3.2 Enhancement of Akt phosphorylation by glycine in cortical neurons does not depend on the activation of glycine receptors or the activity of p38-MAPK signaling.
(A) Strychnine (10 µM) does not interfere with the enhancement of
p-Akt by glycine (100 µM) in neurons where NMDAR channel
activities are inhibited (n=6, Student’s T test, *p<0.05 vs. –Gly).
(B) In cultured rat cortical neurons glycine enhances Akt
phosphorylation after NMDAR channels and glycine receptors
were inhibited (n=6, Student’s T test, *p<0.05 vs.-Gly). (C)
Glycine has no significant effect on p38-MAPK phosphorylation
(p-p38) in cortical neurons following NMDAR channel
inactivation procedure (n=6; ANOVA test, *p<0.05 vs. Control).
43
Figure 3.3 Non-ionotropic activity of GluN2AR mediates glycine-induced enhancement of Akt phosphorylation
(A) In HEK293 cells without or with GFP transfection, the levels of p-Akt 44
are not altered by glycine (100 µM) treatment after the channel activities of NMDARs are inhibited by the NMDAR channel inactivation procedure
(n=6; ANOVA test). (B) In HEK293 cells transfected with
GluN1+GluN2A cDNAs, glycine (100 µM) increases p-Akt after the channel activities of NMDARs are inhibited (n=9, Student’s T test, *p <
0.05 vs. -Gly). (C) In HEK293 cells transfected with GluN1+GluN2B cDNAs, the levels of p-Akt are not altered by glycine (100 µM) after the channel activities of NMDARs are inhibited (n=6; Student’s T test). (D) In
HEK293 cells transfected with GluN1, GluN2A or GluN2B cDNAs, respectively, the levels of p-Akt are not altered by glycine (100 µM) after the channel activities of NMDARs are inhibited (n=6; ANOVA test). (E)
In HEK293 cells transfected with GluN1(N598Q)+GluN2A, but not
GluN1(N598Q) alone, glycine enhances Akt phosphorylation after the channel activities of NMDARs are inhibited (n=6, ANOVA test, *p<0.05 vs. -Gly). (F) Glycine increases Akt phosphorylation in HEK293 cells transfected with GluN1(N598R)+GluN2A following NMDAR channel inactivation procedure (n=6; ANOVA test, *p<0.05 vs. -Gly). (G) The
GluN2A protein expression in cortical neurons is suppressed by GluN2A shRNA (n=6, Student’s T test, *p<0.05 vs. shRNA control). (H) GluN2A knockdown by GluN2A shRNA attenuates glycine-induced increase of p-
Akt in cortical neurons where NMDAR channels are inhibited (n=6,
ANOVA test, *p<0.05 vs. shRNA control; #p<0.05 vs. GluN2A shRNA; 45
**p<0.05 vs. shRNA control+Gly). (I) The GluN2B protein expression in cortical neurons is suppressed by GluN2B shRNA transduction (n=6,
Student’s T test, *p<0.05 vs. shRNA control). The GluN2B shRNA and shRNA control were purchased from Santa Cruz Biotechnology. (J)
GluN2B knockdown by GluN2B shRNA does not interfere with glycine- induced increase of p-Akt in cortical neurons where NMDAR channels activities are inhibited (n=6, ANOVA test, *p<0.05 vs. shRNA control;
#p<0.05 vs. GluN2B shRNA). Gly: glycine.
46
Figure 3.4 Glycine-GluN1 binding is required for glycine-induced non- ionotropic activation of GluN2ARs.
(A) Glycine-GluN1 binding site antagonist L-689560 (50 µM) blocks
glycine (100 µM)-induced increase of p-Akt in cultured cortical neurons
after the channel activities of NMDARs are inhibited (n=6, ANOVA 47
test, *p<0.05 vs.-Gly). (B) L-689560 (50 µM) blocks glycine (100 µM)-
induced increase of p-Akt in HEK293 cells transfected with GluN1 +
GluN2A after the channel activities of NMDARs are inhibited (n=6,
ANOVA test, *p<0.05 vs.-Gly). (C) Glutamate-GluN2 binding site antagonist DAPV (50 µM) does not interfere with glycine (100 µM)- induced increase of p-Akt in cultured cortical neurons following the
NMDAR channel inactivation procedure (n=6, Student’s T test, *p<0.05 vs.-Gly). (D) DAPV (50 µM) does not interfere with glycine (100 µM)-
induced increase of p-Akt in HEK293 cells transfected with
GluN1+GluN2A following the NMDAR channel inactivation procedure
(n=6, Student’s T test, *p<0.05 vs.-Gly) (E) GluN2BR antagonist Ro 25-
6981 (5 µM) does not interfere with glycine (100 µM)-induced increase
of p-Akt in cultured cortical neurons following the NMDAR channel
inactivation procedure (n=6, ANOVA test, *p<0.05 vs.-Gly). (F) Ro 25-
6981 (5 µM) does not interfere with glycine (100 µM)-induced increase
of p-Akt in HEK293 cells transfected with GluN1+GluN2A following
the NMDAR channel inactivation procedure (n=6, ANOVA test,
*p<0.05 vs.-Gly). (G) D-serine increases the level of p-Akt in cultured
cortical neurons after the channel activities of NMDARs are inhibited
(n=6, ANOVA test, *p<0.05 vs. Control). (H) D-serine increases the
level of p-Akt in HEK293 cells transfected with GluN1+GluN2A after
the channel activities of NMDARs are inhibited (n=5, ANOVA test, 48
*p<0.05 vs. Control). (I) D-serine has no effect on the level of p-Akt in
HEK293 cells transfected with GluN1+GluN2B after the channel activities of NMDARs are inhibited (n=5, ANOVA test). 49
Figure 3.5 Glycine protects against glutamate neurotoxicity-induced neuronal injury in cortical neurons through non-ionotropic activation of
GluN2ARs.
(A) A schematic diagram showing glutamate neurotoxicity injury and 50
glycine treatment procedure. (B) Representative images showing that
glycine (100 μM) reduces glutamate neurotoxicity-induced cell death in neurons where NMDAR channel activity is inactivated. Green: FDA;
Red: PI. Scale bar=25 µm. (C) Summarized data of A (n=5. Total 3136
cells counted for Con group, 2825 cells for Sham group, 3225 cells for
Inj group, 3208 cells for Inj+Gly group, 3003 cells for MK group, 3160
cells for Inj+MK group and 3231 cells for Inj+MK+Gly group.
ANOVA test, *p<0.05 vs. Sham; #p<0.05 vs. Inj; **p<0.05 vs. Inj;
##p<0.05 vs. Inj+MK). (D) In neurons where NMDAR channel
activities are inhibited, glycine (100 μM) prevents glutamate
neurotoxicity-induced increase of LDH release (n=6, ANOVA test,
*p<0.05 vs. Sham; #p<0.05 vs. Inj; **p<0.05 vs. Inj; ##p<0.05 vs.
Inj+MK). (E) Glycine (100 μM) reduces glutamate neurotoxicity-
induced increase of LDH release in neurons where shRNA control is
transfected and NMDAR channel activity is suppressed (n=6, ANOVA
test, *p<0.05 vs. Sham; **p< 0.05 vs. Inj; #p<0.05 vs. Inj+MK). (F)
Glycine (100 μM) does not prevent glutamate neurotoxicity-induced
increase of LDH release in neurons where GluN2A expression is
suppressed by GluN2A shRNA and NMDAR channel activity is
inhibited (n=6, ANOVA test, *p<0.05 vs. Sham; **p<0.05 vs. Inj). (G)
Akt inhibitor IV (1 μM) decreases glycine (100 μM)-induced reduction
of LDH release in neurons where NMDAR channel activity is inhibited 51
(n=6, ANOVA test, *p<0.05 vs. Inj+MK; **p<0.05 vs. Inj+MK+Gly).
(H) Glycine-GluN1 binding antagonist L-689560 (50 µM) decreases glycine (100 μM)-induced reduction of LDH release in neurons where
NMDAR channel activity is inhibited (n=6, ANOVA test, *p<0.05 vs.
Inj+MK; **p<0.05 vs. Inj+MK+Gly). 52
Figure 3.6 Glycine treatment reduces the infarct area of ischemic brain independent of glycine receptor activation and the channel activity of
NMDARs.
(A) Sample images of TTC stained-brain sections collected at 24 h after
ischemia onset. Glycine (100 µg/100 g, icv) was administered at 3 h
following ischemic reperfusion (I/R). At 30 min prior to glycine
injection, MK-801 (8 µg/100 g, icv) and strychnine (1.2 µg/100 g, icv)
were injected. (B) Summarized quantification data indicate that glycine
treatment at 1.5, 3, or 6 h following I/R reduces infarct area after
glycine receptors and NMDARs are inhibited (n=10 animals for each
group; ANOVA test, #p<0.05 vs. I/R+vehicle; *p<0.05 vs.
I/R+Stry+MK). Glycine (100 µg/100 g, icv) was injected at 3 h
following I/R and TTC strained-brain sections were collected at 24 h 53
after ischemia onset.
54
Figure 3.7 Glycine promotes functional recovery of ischemic animals 55
independent of glycine receptor activation and the channel activity of
NMDARs.
For all the experiments, Glycine (100 µg/100 g, icv) was administered
at 3 h following I/R. At 30 min prior to glycine injection, MK-801 (8
µg/100 g, icv) and strychnine (1.2 µg/100 g, icv) were injected. (A)
Animals treated with glycine have lower scores of mNSS test at day 7
and 14 compared with I/R+Stry+MK group (n =10; ANOVA test,
*p<0.05 vs. I/R+Stry+MK). (B) Animals treated with glycine has lower
scores of beam-walking test at day 3, 7 and 14 compared with
I/R+Stry+MK group (n=10; ANOVA test, *p<0.05 vs. I/R+Stry+MK).
(C) Animals treated with glycine have higher ratio in MST test at day 7
and 14 compared with I/R+Stry+MK group (n=10; ANOVA test,
*p<0.05 vs. I/R+Stry+MK).
56
Table 3.1 Modified Neurological Severity Score (mNSS)
57
Table 3.2 The Beam Walk Test Scoring Criteria
58
3.5 Discussion
Using a Ca2+-free ECS-based procedure to inactivate the channel activity of
NMDARs in cultured cortical neurons and HEK293 cells expressing GluN2ARs, we
tested the effect of glycine on Akt phosphorylation, a cellular process playing important
role in neuronal survival. We provided the first evidence that glycine alone induced a
potentiation of Akt phosphorylation independent of the channel activity of NMDARs. We
confirmed that glycine-induced non-ionotropic activation of GluN2ARs, but not
GluN2BRs, mediated the enhancement of Akt activation. Thus, our study identified a
non-ionotropic function of GluN2ARs.
To ensure no channel activities of the NMDARs were contributing to glycine-
induced Akt phosphorylation in our study, we established a NMDAR channel
inactivation procedure to completely inhibit the channel activities of NMDARs (Figure
3.1A). We employed a Ca2+-free ECS containing MK-801 (10 μM) and Ca2+ chelator
EGTA, a specific solution referred as ECS-1. To open the NMDAR channels and allow
MK-801 in the ECS-1 to fully block NMDARs, we first treated the cells with ECS-1
containing NMDA and glycine for 1.0 min. We then washed the cells with ECS-1 for 30
min.
The use of NMDAR inhibitor DAPV leads us to conclude that the effect of
glycine does not require glutamate. It also aids us to exclude the contribution of residual
NMDAR channel activities to the observed effect of glycine. We showed that glycine
increased Akt phosphorylation even after DAPV treatment (Figure 3.4C-D). As DAPV 59
inhibits NMDAR channel activity, this finding further supports the notion that the glycine
effect is independent of NMDAR channel activity.
Increasing evidence supports the non-ionotropic function of NMDARs (Birnbaum et al. 2015; Kessels et al. 2013; Nabavi et al. 2013; Stein et al. 2015; Tamburri et al.
2013; Vissel et al. 2001). It has been recently shown that a non-ionotropic activation of
NMDAR was insensitive to the glycine-GluN1 site antagonist (Kessels et al. 2013;
Nabavi et al. 2013; Stein et al. 2015). However, our study shows that the glycine-GluN1 binding is required to activate the non-ionotropic activity of GluN2ARs (Figure 3.4A-B).
Interestingly, we demonstrate that the competitive GluN2 antagonist DAPV does not
interfere with the enhancement of Akt activation by glycine. Furthermore, we show that the non-ionotropic activity of GluN2ARs is produced with the endogenous glycine-
GluN1 site agonist, D-serine. Together, these findings suggest that glycine triggers a non-
ionotropic activity of GluN2ARs through the glycine-GluN1 binding site.
Glycine is a co-agonist of NMDARs (Johnson and Ascher 1987). The activation
of GluN2ARs and GluN2BRs requires both glutamate and glycine (Johnson and Ascher
1987). It was not clear whether glycine alone had a functional effect on GluN2ARs and
GluN2BRs. We revealed an unexpected role of glycine, independent of glutamate, to
induce a non-ionotropic activity of GluN2ARs. This finding suggests that glycine acts as
a sole agonist to elicit a non-ionotropic activity of GluN2ARs.
By testing the effects of glycine in HEK293 cells transfected with different
combinations of NMDARs, we were able to obtain direct evidence to reveal that a non-
ionotropic activation of GluN2ARs but not GluN2BRs mediates the enhancement of Akt 60
activation (Figure 3.3A-F). Because there were no glutamate was added into ECS-1 in
our experimental conditions, the observations in HEK293 cells also provide further
evidence to support the conclusion that glutamate is not required for non-ionotropic
activation of GluN2ARs by glycine. Thus, the transfected HEK293 cell is a useful system
for our characterization of non-ionotropic activity of GluN2AR.
NMDA-induced Akt phosphorylation has been previously reported in cortical,
striatal and retinal neurons (Mejia-Garcia et al. 2013; Perkinton et al. 2002; Sutton and
Chandler 2002). These studies show that NMDA or glutamate treatment increases Akt phosphorylation. The enhancement of Akt phosphorylation by NMDA or glutamate is blocked by MK-801, suggesting that the ionotropic activity of NMDARs is required for
Akt activation. Since Akt phosphorylation exists after MK-801 treatment or in the Ca2+-
free conditioning, these data indicate that Ca2+-independent signaling pathways
contribute to the Akt activation. Our study provides evidence that even after the cortical
neurons were treated with MK-801 and extracellular Ca2+-free solution, glycine alone
induces a non-ionotropic activity of GluN2ARs to increase the level of Akt
phosphorylation. These results suggest that while Akt activation depends on Ca2+ influx
through NMDARs, a non-ionotropic activity of GluN2ARs also contributes to Akt
activation.
Akt deactivation is believed to be a causal mediator of cell death. Enhancement of
Akt activity exerts pro-survival effect in neuronal injury and neurodegenerative diseases
(Burke 2007; Luo et al. 2003; Manning and Cantley 2007). We focused on Akt activation
by glycine because glycine was shown to have a neuroprotective effect (Liu et al. 2007; 61
Zhao et al. 2005). In this study, we identify Akt as a downstream neuroprotective signal
of glycine that activates non-ionotropic activity of GluN2ARs. We provide evidence that
non-ionotropic activation of GluN2ARs by glycine reduces glutamate neurotoxicity-
induced Akt deactivation and thus prevents cortical neuronal death. Akt is known to
influence neuronal survival through activation or inhibition of substrates (Burke 2007;
Luo et al. 2003; Manning and Cantley 2007). For example, activated Akt promotes
survival through phosphorylation of transcription factors forkhead/FOXO, NF-κB and
mdm2 or through phosphorylation of Bcl-2 family members Bad and Bim. Further study
is needed to determine which Akt-dependent signal pathway mediates the non-ionotropic
activation of GluN2ARs by glycine.
The p38-MAPK signaling mediates excitotoxicity-induced neuronal injury. It is
recently reported that a non-ionotropic activity of GluN2BRs and subsequent p38-MAPK
activation are required for β-amyloid–induced synaptic depression and loss (Kessels et al.
2013; Li et al. 2013). The p38-MAPK is also involved in NMDAR-dependent LTD and is
shown to be activated by non-ionotropic NMDAR signaling after chemical LTD induction (Nabavi et al. 2013; Stein et al. 2015; Zhu et al. 2002). We tested the effect of
glycine on p38-MAPK in cortical neurons following the NMDAR channel inactivation
procedure. In contrast to the effect of glycine on Akt phosphorylation, glycine had no
significant effect on p38-MAPK phosphorylation in our experimental conditions (Figure
3.2C). These data indicate that Akt signaling mediates the effect of non-ionotropic
activity of GluN2ARs but p38-MAPK signaling mediates the effect of non-ionotropic
GluN2BRs. 62
NMDAR-mediated neurotoxicity induces neuronal death and neurodegeneration
in various CNS disorders including ischemic stroke, traumatic brain injury and
neurodegenerative diseases (Koutsilieri and Riederer 2007; Lee et al. 1999; Lipton and
Rosenberg 1994). However, the use of NMDAR antagonists as neuroprotective agents
was disappointing in clinical trials (Ikonomidou and Turski 2002; Kemp and McKernan
2002; Steinberg et al. 1995). A simple possibility is that these antagonists, while suppressing NMDAR-mediated neurotoxicity, block the biological and/or neural
survival-promoting effects of NMDARs (Anastasio et al. 2009; DeRidder et al. 2006;
Dingledine et al. 1999; Liu et al. 2007). Thus, identification of molecular mechanisms by
which specific NMDAR subtype selectively exerts its effect on neuronal survival or death
would provide a critical basis for the development of potent therapy for CNS injuries and
neurodegenerative diseases.
GluN2ARs and GluN2BRs play different role in neuronal survival or death (Chen
et al. 2008; Liu et al. 2007). But the underlying molecular mechanism remains unclear. It
has been recently reported that a non-ionotropic function of NMDARs was required for
β-amyloid–induced synaptic depression and synaptic loss (Birnbaum et al. 2015; Kessels
et al. 2013; Tamburri et al. 2013), providing new evidence for the involvement of
GluN2BRs in neurotoxicity. Our observation for the non-ionotropic activation of
GluN2ARs selectively by glycine explains in part why GluN2AR plays a different role
than GluN2BR in neuronal survival. Unlike the approach blocking the cell death signal,
the neuroprotection mediated by non-ionotropic activity of GluN2ARs is through 63
promoting Akt-dependent neuronal survival signal, which offers no limitation of therapeutic window (Liu et al. 2007). 64
Chapter 4 A Synaptic Model of Learning and Memory
Brendan Lujan 65
4.1 Summary
A notable and fundamental property of central nervous systems (CNS) is the ability to process and retain information. Synapses in the CNS undergo various short- and long-term changes in their strength that regulate the ability to learn new information and store memories (Bliss and Collingridge 1993; Malenka 1994; Zucker and Regehr 2002).
Synaptic plasticity is the ability of a synapse to readily alter its efficiency of connectivity.
The first synapses in the brain identified as undergoing these types of changes were
observed in the hippocampus (Bliss and Gardner-Medwin 1973; Bliss and Lomo 1973).
Brief trains of high frequency stimulation to monosynaptic excitatory pathways in the
hippocampus caused abrupt and sustained increases of synaptic transmission (Bliss and
Gardner-Medwin 1973; Bliss and Lomo 1973). These changes in synaptic strength can be
induced in milliseconds and have been observed to last for hours in-vitro and even days
in-vivo (Bliss and Collingridge 1993; Buzsaki 1980). Thus, these properties have provided the groundwork to model the formation of memories on a synaptic level.
4.2 Introduction
Long-lasting increases in synaptic strength of excitatory neurons is a phenomenon referred to as long-term potentiation (LTP); (Bliss and Lomo 1973), while long-lasting decreases in synaptic strength are referred to as long-term depression (LTD); (Dudek and
Bear 1992). Late forms of LTP, those lasting extended hours and even days, require gene transcription and synthesis of new protein (Nguyen et al. 1994). Substantial evidence
suggests that the hippocampus is an essential component of the CNS that is required for
at least some forms of learning and memory (Squire and Zola-Morgan 1991). 66
Importantly, LTP and LTD are readily inducible specifically in the apical dendrites of
CA1 pyramidal cells of the hippocampus by coordinated afferent stimulation of the
Schaffer collateral and commissural axons by high- and low-frequency stimulation protocols, respectively (Larson et al. 1986; Malenka and Bear 2004; Rose and Dunwiddie
1986). Furthermore, these types of synaptic events have been observed in the
hippocampus during learning (Otto et al. 1991). Because the hippocampus has been
labeled the learning and memory center of the brain, LTP and LTD have been proposed
on a mechanistic level to be functionally relevant in the advanced cognitive processing
activities of the mammalian brain (Doyere and Laroche 1992).
LTP can be described by three basic fundamental properties: cooperativity,
associativity and input-specificity. Cooperativity describes an intensity threshold for
induction, and thus is a function of the intensity and pattern of stimulation (McNaughton
et al. 1978). LTP is an associative property, by which strong activation of synapses in one part of a cell can be coordinated to induce LTP at nearby synapses on the same cell if both are activated in a finite temporal window (Levy and Steward 1979; McNaughton et
al. 1978). In the CA1 region of the hippocampus, LTP is also input-specific, allowing only certain sets of synapses on a cell to be affected, while other nearby cells that do not
receive activity do not share in the potentiation (Andersen et al. 1977; Lynch et al. 1977).
LTP and LTD has been most extensively studied in the hippocampus but has also been
observed in many other brain regions such as the cortex, amygdala and cerebellum
(Bindman et al. 1988; Chapman et al. 1990; Crepel and Jaillard 1991). It is still somewhat
debated whether these long-lasting changes in synaptic efficacy are due to modifications 67
in either presynaptic compartments, postsynaptic compartment or both. However,
literature suggests that the postsynaptic compartment may play a more critical role in
regulation of synaptic plasticity changes, as changes in presynaptic vesicle release probability were not observed during induction of LTP (Manabe and Nicoll 1994).
Furthermore, a retrograde messenger such as nitric oxide, carbon monoxide, arachidonic
acid and platelet-activating factor has yet to be identified and deemed absolutely essential
for induction of these forms of plasticity thus further negating the role of the presynapse
during induction of LTP or LTD (Cummings et al. 1994; Williams et al. 1993).
Both LTP and LTD are triggered in an activity-dependent fashion in which the
presynaptic neuron releases the excitatory neurotransmitter glutamate into the active zone
of a synapse, and binds both AMPA receptors (AMPARs) and NMDA receptors
(NMDARs) on the postsynaptic membrane. It is well-accepted that NMDAR activation is
necessary for many forms of LTP and LTD, and thus requires cell depolarization (Bliss
and Collingridge 1993). Being that AMPARs and NMDARs are often colocalized on the
postsynaptic membrane, changes in long-lasting forms of plasticity are triggered first by
AMPAR activation, which depolarizes the cell via inward flux of Na+ current
(Collingridge et al. 1983). Once cell depolarization has transpired by the activation of
AMPARs, the voltage-dependent block by Mg2+ of the NMDAR is relieved and the
NMDAR becomes permeable to Na+ and K+ as well as the pertinent second-messenger
Ca2+ (Dingledine et al. 1999). The consequential rise in intracellular Ca2+ has, until recently, been deemed essential for induction of LTP and LTD (Regehr and Tank 1990).
Relatively large influx of Ca2+ is thought to lead to LTP while relatively smaller influx 68
of Ca2+ couples to LTD as evidenced by their induction protocols (Sommer et al. 1990;
Yu et al. 1997). Evidence for this model is supported by the fact that specific NMDAR
antagonists have minimal effect on basal synaptic transmission while completely
blocking LTP induction. The detailed mechanism of how these Ca2+-dependent events
are temporally and spatially regulated remains to be elucidated.
4.3 AMPARs
The AMPAR, a class of ligand-gated ion channel involved in most fast glutamatergic signaling transmission, is expressed in most neurons of the mammalian
CNS and is a major effector of plasticity changes (Hollmann and Heinemann 1994;
Sommer et al. 1990). AMPARs are tetramers composed of GluR1-4 subunits and may
exist in either flip or flop isoforms dependent on alternative splicing events (Borges and
Dingledine 1998). Each subunit can form a homomeric functional channel when
expressed in oocytes or transfected cells; however, in neurons, AMPARs are largely
heteromeric and composed of at least two different subunits (Hollmann et al. 1989;
Nakanishi et al. 1990). Further diversity of the AMPAR can be attributed to an RNA-
editing site, denoted the Q/R site in the transmembrane domain of the GluR subunits,
which suffices to act as a switch to control divalent cation permeability (Sommer et al.
1991). Noteworthy is the fact that heteromeric AMPA receptor inclusions with the
GluR2 subunit are impermeable to Ca2+ due to the Q/R editing site (Hume et al. 1991).
Activation of AMPARs is dependent upon agonist glutamate or synthetic agonist α-
amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA; which has higher affinity
than glutamate) binding to the extracellular domain of the receptor which regulates 69
channel gating and permeability of the monovalent cation Na+ and in some cases the
divalent cation Ca2+ (Borges and Dingledine 1998). The number and type of AMPAR subunits expressed in a neuron varies among different neuronal populations, but at least two to four subunits are usually present in each cell (Martin et al. 1993). Evidence for multiple AMPAR subtypes at hippocampal CA1 and CA2 regions has been suggested and the vast majority are those composed of heteromers containing the GluR2 subunit
(Wenthold et al. 1996). Modification in surface expression and changes in the biophysical properties of AMPARs themselves are considered to be the major postsynaptic mechanisms for the regulation of synaptic plasticity (Carroll et al. 2001; Collingridge et al. 2004; Luscher et al. 1999).
4.3.1 AMPAR Trafficking
The first evidence for AMPAR trafficking in neurons was provided when a GFP- tagged GluR1 subunit was expressed in hippocampal neurons and was observed to translocate from the intracellular shaft of dendrites into spines following strong synaptic stimulation and NMDAR activation (Shi et al. 1999). Homomeric GluR1 AMPARs were delivered to synapses upon NMDAR activation concurrently with CaMKII activation and this membrane recruitment was dependent on the C-terminal of GluR1, which associates with the PDZ domain of SAP97 (Hayashi et al. 2000; Leonard et al. 1998). Furthermore,
CamKII has been shown to directly phosphorylate GluR1 subunits providing mechanistic
basis regulating their expression to the neuronal surface (Barria et al. 1997). It has been suggested that GluR4 subunits may serve a fundamental role in developmental plasticity changes while GluR1 subunits serve their primary functional role in plasticity of mature 70
synapses (Zhu et al. 2000). Notably, GluR2 subunit delivery to synaptic sites seems to be
a constitutive process and independent of stimulation (Shi et al. 1999). Similar to GluR1
incorporations into the synapse, GluR2 subunits require the PDZ binding domain on the
C-terminal tail of the receptor and interactions with proteins such as GRIP, ABP or
PICK-1 seem to be important (Daw et al. 2000; Dong et al. 1997; Srivastava et al. 1998;
Xia et al. 1999). The cytoplasmic tail of GluR2 also associates with NSF protein which is
involved in SNARE mediated delivery to the membrane but the exact mechanism is
unclear (Kim and Lisman 2001). Importantly GluR1/GluR2 heteromeric channels can be
inserted to membrane regions by CamKII activation. The GluR3 subunit is often viewed
as functionally synonymous with the GluR2 subunit in plasticity events due primarily to
sequence similarity (Passafaro et al. 2001). Clearly AMPAR insertion to the membrane is
important in plasticity changes but how they arrive to the synapse is not completely clear.
Studies performed using a stargazin-deficient mutant suggest a two-step process by which the AMPARs may first enter extrasynaptic sites and subsequently are directed to synaptic sites (Chen et al. 2000). These observations suggest that this ion channel family may have evolved sophisticated trafficking properties that are amenable to a variety of neuronal plasticity changes.
Counterbalancing surface expression of AMPARs is their internalization via
endocytic pathways and serves to remove AMPARs from the cell membrane (Carroll et
al. 2001). AMPAR internalization is induced during depression protocols, involving
NMDAR activation, and is a dynamin-dependent process (Lin et al. 2000). Preventing
AMPAR internalization blocks LTD induction in both hippocampus and cerebellum 71
(Carroll et al. 2001). After endocytosis, most AMPARs are recycled to the cell
membrane, likely involving the NSF protein, while others are targeted for degradation in
non-recycling endosomes (Nishimune et al. 1998). These studies reveal a strikingly
complex mechanism by which cell excitability can be regulated through AMPAR surface
trafficking.
4.4 NMDARs
NMDARs are ligand-gated ion channels located on the post-synaptic membrane
of neurons that play crucial roles in regulating cell excitability (Dingledine et al. 1999).
NMDARs are tetramers containing an obligate GluN1 subunit with differential
combinations of GluN2 (A-D) and GluN3 (A-B) subunits. Importantly, in hippocampal and cortical brain regions, the receptors primarily expressed are the GluN2A-containing
(GluN2ARs) and GluN2B-containing (GluN2BRs) NMDARs. Their location makes them prime targets as regulators of synaptic plasticity events in these brain regions. The gating of this channel is unique as the receptor requires two agonists to stimulate channel opening. The primary agonist glutamate must associate with the GluN2 subunit while the
co-agonist glycine/D-serine must concurrently bind the GluN1 subunit to initiate
opening. Glutamate is primarily supplied to the postsynaptic neuron via secretion from
the presynaptic bouton whereas glycine and D-serine has been shown to be supplied via
neighboring astrocytes as well as presynaptic neurons (Oliet and Mothet 2009).
Furthermore, this transmembrane channel holds a voltage-dependent property to regulate
gating with the divalent cation Mg2+ sitting in the channel pore under resting membrane
potential (Mayer et al. 1984). Upon presynaptic release of glutamate, membrane 72
depolarization drives a temporally restricted removal of Mg2+ from the channel pore
(Jahr and Stevens 1990). The NMDAR channel gating has thus been described by a
coincidence detection property because both presynaptic release of glutamate and
voltage-dependent removal of Mg2+ must occur simultaneously to mediate Ca2+ influx.
Such high regulation of gating suggests that this protein has essential functional roles.
4.4.1 NMDAR regulation of ERK 1/2 signaling
NMDAR dependent changes in plasticity couple to many intracellular signaling cascades that are upstream of AMPAR trafficking in the postsynaptic membrane. It had
appeared clear until recently that NMDAR-dependent LTP requires CaMKII activation
and Ca2+ influx to the dendritic spine to initiate LTP (Bortolotto and Collingridge 1993).
However, several other protein kinases have also been attributed to playing essential roles
in synaptic plasticity. These kinases include PKA, PKC and cAMP-dependent protein
kinases as well as the more recently discovered role of the mitogen-activated protein
kinases (MAPKs) that activate the extracellular regulated kinase ½ (ERK1/2) in LTP
induction (Boehm et al. 2006; English and Sweatt 1997; Roche et al. 1996).
ERK1/2 are intracellular protein kinases that are activated via their
phosphorylation by extracellular signaling cues, which cause an increase in the active
GTP-bound form of the G protein Ras. Importantly, neuronal Ras and Rap activation are
directly coupled to AMPAR plasticity (Zhu et al. 2002). NMDAR mediated ERK1/2
activation is dependent upon elevated levels of GTP-Ras, which is stimulated by the
activity of guanyl exchange factors (GEFs) and inhibited by the activity of GTPase- activating proteins (GAPs). GTP-Ras then activates Raf kinase which thereby activates 73
MEK whose substrate is ERK1/2 (Nakielny et al. 1992). Neuronal ERK1/2 activation is
mediated by direct membrane depolarization or excitatory glutamatergic signaling (Rosen et al. 1994; Yun et al. 1999). ERK1/2 activation induced by these stimuli are Ras-
dependent and rely upon Ca2+ influx via either NMDAR or voltage-gated Ca2+
channels. However the precise neuronal GEFs and GAPs related to Ras activation have
not been elucidated.
Major advancements elucidating the role of ERK1/2 in neuronal function have
been made using specific antagonists to the MAPK pathway. These inhibitors can be used
in primary neuronal cell culture, organotypic slice culture or perfused directly into living
animals. They are good for studying acute function of the MAPK pathway in neurons as
they act on endogenous proteins and rapidly diffuse to their targets to minimize indirect
effects such as protein synthesis. Both PD 98059 and UO126 target the protein kinase
MEK, the upstream effector of ERK1/2, and have been used in several studies to tease out the role of the MAPK pathway in neuronal function and more specifically synaptic plasticity (English and Sweatt 1997). Importantly neither PD 98059 nor UO126 inhibit
MEK via ATP competition, implying high specificity. Instead, these inhibitors associate with MEK preventing its activation with upstream kinase Raf, and thus preventing MEK
effects on substrate ERK 1/2. The first evidence of ERK1/2 involvement in LTP was
provided when ERK1/2 inhibitor PD 98059 completely blocked LTP induction in the
hippocampus and this conclusion has been widely supported through a number of other
studies (Bolshakov et al. 2000; English and Sweatt 1997; Impey et al. 1998; Patterson et
al. 2001). Furthermore, these inhibitors have been used in behavioral assays of learning 74
and memory with results implicating ERK 1/2 function in cognitive processing capacity.
The two most common behavioral assays to assess long-term memory formation and
function are spatial learning and fear conditioning tests. Importantly, ERK1/2 inhibition
dramatically reduced mammalian performance in these assays via intra-hippocampal
infusion (Schafe et al. 2000; Selcher et al. 1999). Thus, ERK1/2 signaling not only
regulates synaptic plasticity but also learning and memory behavior in a living animal.
4.4.2 Differential regulation of NMDARs in bidirectional synaptic plasticity
A substantial amount of work has been dedicated to understanding the opposing
roles of different NMDAR subtypes on excitatory neural plasticity. Until recently, both
LTP and LTD were thought to require NMDAR receptor activation (Bliss and
Collingridge 1993; Malenka 1994). However, the detailed mechanism by which the same
receptor can be responsible for bidirectional plasticity, namely LTP and LTD, remains a
highly controversial topic. In rat hippocampal slice, field excitatory postsynaptic potential
(fEPSP) recordings were made in the presence of specific pharmacological antagonists to
subtype-specific NMDARs: NVP-AAM077 is a GluN2AR competitive antagonist with
>130 fold affinity for GluN2ARs compared with other NMDAR subtypes (Auberson et
al. 2002). Addition of NVP-AAM077 prevented HFS induced LTP induction and a
saturated form of LTP induced by many HFS protocols, while having minimal effect on
LFS induction of LTD (Liu et al. 2004). This result suggests a preferential activation of
GluN2ARs to achieve LTP induction. This result echoes postulated mechanisms of
previous studies suggesting different NMDAR subtype effects on bidirectional plasticity events in the hippocampus (Hrabetova and Sacktor 1997). 75
Similarly, Ifenprodil and Ro25-6981 are GluN2BR competitive antagonists with
>200-fold affinity for GluN2BRs compared with other NMDAR subtypes (Fischer et al.
1997; Williams 1993). Using a LFS protocol to induce LTD, it was shown that ifenprodil and Ro25-6981 both completely blocked the induction of LTD, suggesting a pivotal role for GluN2BRs for the induction of LTD. The data presented in this study implicating
GluN2BRs in LTD induction has been supported in many other studies performed in many different brain regions (Brigman et al. 2010; Gao et al. 2010; Liu et al. 2004;
Massey et al. 2004). These studies suggest an attractive hypothesis by which GluN2ARs
are responsible for LTP while GluN2BRs are responsible for LTD (Figure 4.1).
On the other hand, several studies have implicated a much more complex
mechanism regulating bidirectional synaptic plasticity in excitatory neurons as a single
NMDAR subtype cannot be responsible for either LTP or LTD, but rather both receptor
subtypes are able to generate LTP and/or LTD (Berberich et al. 2005; Cao et al. 2007;
Miwa et al. 2008). However, not all of these experiments were performed solely in the hippocampus. Data conflicting the hypothesis that LTP is mediated solely by GluN2ARs
was produced using a preparation of the dorsolateral bed nucleus of the stria terminalis in
mice where LTP was still induced in a GluN2AR knockout background (Weitlauf et al.
2005). Similarly, genetically modified mice over-expressing the GluN2B subunit was shown to lead to increased LTP in the hippocampus and increased performance in learning and memory assays (Tang et al. 1999; Wong et al. 2002). In a study performed in the anterior cingulate cortex, it was shown that both GluN2ARs and GluN2BRs 76
contribute to LTD (Toyoda et al. 2005). Thus, NMDAR subtype specificity in regulation of bidirectional synaptic plasticity remains controversial. 77
Figure 4.1 NMDARs bidirectionally regulate synaptic plasticity
Schematic representing synaptic depression (left) and synaptic potentiation
(right). GluN2ARs mediate synaptic potentiation. GluN2AR activation
induces Ca2+ influx and ERK1/2 phosphorylation. Phosphorylated ERK1/2
increases AMPAR insertion to the postsynaptic membrane. GluN2BRs
mediate synaptic depression. GluN2BR activation similarly induces Ca2+
influx but phosphorylates p38. Phosphorylated p38 induces AMPAR
endocytosis causing synaptic depression.
78
4.5 Discussion
Substantial evidence suggests that the NMDAR plays a major role in mediating synaptic plasticity events, and may contribute to learning and memory processes. However, the mechanisms underlying LTP and LTD remain unclear. Although it has been suggested that LTP and LTD are dependent on specific NMDAR subtypes, other studies refute this hypothesis. Thus, further elucidating the mechanism by which bidirectional synaptic plasticity occurs and how these receptor subtypes link to intracellular signaling will provide profound insight into the cognitive processing ability of the CNS. Because the complete mechanism by which LTP and LTD regulate cognitive processing remains unclear, few therapeutic strategies have been developed to treat deficits in cognitive function.
Recent evidence produced in our lab and others shows a novel function of the
NMDAR to regulate synaptic plasticity events (Nabavi et al. 2013). Namely, the
NMDAR may function by a metabotropic mechanism, independent of its channel activities, to regulate neuronal excitability. This recently described novel function of the
NMDAR may help to explain inconsistencies in the literature in regards to information processing mediated by postsynaptic changes in synaptic strength.
79
Chapter 5 Glycine Potentiates AMPA Receptor Function through Metabotropic
Activation of GluN2A-containing NMDA Receptors
Lijun Li*, Rong Hu*, Brendan Lujan, Tianyuan Cui, Juan Chen, Zefen Wang, Yasuko
Nakano, Mingxia Liao, Shuzo Sugita, Liang Zhang, Heng-Ye Man, Hua Feng, Qi Wan
*These authors contributed equally to this work
Brendan Lujan contributed to the data presented in Figures 5.4-5.
80
5.1 Summary
NMDA receptors (NMDARs) are Ca2+-permeable ion channels, whose activation requires agonist glutamate and co-agonist glycine. The channel activity of NMDARs regulates the function of AMPA receptors (AMPARs), a process crucial for synaptic plasticity. Here we report that in mouse hippocampus glycine alone increases AMPAR- mediated synaptic currents independent of both the channel activity of NMDARs and the activation of glycine receptors. The potentiation of AMPAR function by glycine is antagonized by the inhibitor of extracellular regulated kinase ½ (ERK1/2). In cultured hippocampal neurons and HEK293 cells transfected with different combinations of
NMDARs, glycine preferentially acts on GluN2A-containing NMDARs (GluN2ARs) to enhance ERK1/2 phosphorylation. Without depending on the channel activity of
GluN2ARs, glycine increases AMPAR-mediated currents in cultured hippocampal neurons. These results reveal a metabotropic activity of GluN2ARs in mediating glycine- induced potentiation of synaptic AMPAR function through ERK1/2 activation, suggesting a possible role of metabotropic function of GluN2ARs in AMPAR-mediated synaptic plasticity.
5.2 Introduction
While the ionotropic function of NMDARs has been well studied, recent reports suggest that ligand binding to NMDARs is sufficient to induce long-term depression
(LTD); but does not require ion flow through NMDARs (Nabavi et al. 2013; Tamburri et al. 2013). The metabotropic activity of NMDARs mediated through GluN2BR is required for β-amyloid–induced synaptic depression (Kessels et al. 2013; Nabavi et al. 2013; 81
Tamburri et al. 2013). In the present study we reveal a metabotropic activity of GluN2AR
that mediates glycine-induced potentiation of AMPAR function through activation of
ERK1/2, an important intracellular signal involved in synaptic plasticity (Stornetta and
Zhu 2011; Sweatt 2004; Thomas and Huganir 2004).
5.3 Methods
Hippocampal neuronal culture
Hippocampal neuronal cultures were prepared from C57BL/6 mice at gestation day 17 using a modified protocol (Brewer et al. 1993; Shan et al. 2009). C57BL/6 mice were obtained from both Toronto Western Research Institute (TWRI) and University of
Nevada School of Medicine (UNSOM). Briefly, dissociated neurons were suspended in plating medium (Neurobasal medium, 2% B-27 supplement, 10% FBS, 0.5 μM L-
glutamine, and 25 μM glutamic acid) and plated on poly-D-lysine coated Petri dishes.
After 3 days in culture, half of the plating medium was removed and replaced with
maintenance medium (Neurobasal medium, 2% B-27 supplement, and 0.5 μM L-
glutamine). Thereafter, maintenance medium was changed in the same manner every 3
days. The cultured neurons were used for all the experiments at 12-14 days after plating.
All animal work was conducted according to the guidelines set forth by TWRI Canadian
Council on Animal Care Committee (CCACC) and the UNSOM Institutional Animal
Care and Use Committee (IACUC). All procedures were approved by the TWRI CCACC
and the UNSM IACUC.
Electrophysiological recordings 82
For recording NMDAR-mediated whole-cell currents, the recording electrode
resistance was 2-5 MΏ when filled with a standard intracellular solution containing 140
mM CsCl, 2 mM MgCl2, 1 mM CaCl2, 5 mM EGTA, 10 mM HEPES, 4 mM K2ATP,
titrated to pH 7.3 with CsOH and the osmolality was 300–315 mOsm. Bath solution
contained 140 mM NaCl, 5 mM KCl, 2 mM CaCl2, 25 mM HEPES, 33 mM Glucose,
titrated to pH 7.4 with osmolality of 300-320 mOsm. TTX (0.5 μM) was added into the
bath solution to block voltage-gated Na+ channel currents. Neurons were held at +40 mV
under voltage-clamp. NMDAR-mediated whole-cell currents were recorded by pressure
application of 100 μM aspartate and 1 μM glycine (100 kPa, 200 ms) from a micropipette
with its tip located ~20 μm from the recorded cell. Drugs were delivered at intervals of 60
s. Data were acquired with an Axopatch 200B amplifier and pClamp 10 software
interfaced to a Digidata 1322A acquisition board (Molecular Devices, CA), and signals
were filtered at 2 kHz and digitized at 10 kHz.
For recording AMPA-induced whole-cell currents, the cultures were bathed in
standard extracellular solution (ECS; 137 mM NaCl, 2 mM CaCl2, 5.4 mM KCl, 1 mM
MgCl2, 25 mM HEPES, 33 mM Glucose, titrated to pH 7.4 with osmolarity of 300-320
mOsm) or an ECS solution that did not include Ca2+ (ECS-1; 10 μM MK-801, 5 mM
EGTA, 10 μM strychnine, 0.5 μM TTX, 137 mM NaCl, 5.4 mM KCl, 1 mM MgCl2, 25 mM HEPES, 33 mM Glucose, titrated to pH 7.4 with osmolarity of 300-320 mOsm).
Neurons were held at -70 mV under voltage-clamp. AMPAR-mediated whole-cell
currents were evoked by pressure application (100 kPa, 100 msec) of AMPA (100 μM). 83
The other experimental conditions and methods were same as those for recording
NMDAR currents.
Recording of miniature EPSCs (mEPSCs) was performed as described previously
(Liu et al. 2006). The cultures were bathed in the ECS-1 containing 10 μM bicuculline for recording of AMPAR mEPSCs. At least 200 individual AMPAR mEPSCs were collected before and after application of glycine (100 μM). Records were filtered at 2 kHz and analyzed with Clampfit 10.3 (Molecular Devices). The other experimental conditions and methods were same as those of recording for AMPAR-mediated whole-cell currents.
AMPAR-mediated fEPSPs were recorded in hemi-brain slices (400 μm) containing hippocampus prepared by a vibratome (Leica VT 1200s) using C57BL/6 mice
(age of 3~8 weeks). Before decapitation, mice were anaesthetized and underwent trans-
cardiac infusion with a cold choline chloride solution containing (in mM): 50 NaCl, 80
choline chloride, 3.5 KCl, 7 MgCl2, 0.5 CaCl2, 2 NaH2PO4, 5 HEPES and 20 glucose.
Slices were stabilized in oxygenated (95% O2 and 5% CO2) artificial cerebrospinal fluid
(aCSF) containing (in mM): 125 NaCl, 3.5 KCl, 1.25 NaH2PO4, 25 NaHCO3, 2 CaCl2,
1.3 MgCl2, 10 glucose and 2 kynurenic acid (pH 7.4 when aerated with 95% O2 and 5%
CO2) at 35 for 30 min. The slices were then recovered in aCSF containing 10 μM
bicuculline, ℃ 10 μM MK-801 and 5 μM strychnine but no kynurenic acid (aCSF-1) at
room temperature for over 2 h. All recordings were performed by perfusing the slices
(10-15 ml/min) at room temperature with aCSF-1 that was saturated with 95% O2 and
5% CO2 (Wu et al. 2005). Extracellular recording electrodes (1~2 MΩ) filled with aCSF
were used for the recording. The recording electrode was placed in the CA1 apical 84
dendritic layer. Local afferent stimulation was conducted via placing a bipolar tungsten
wire electrode (tip diameter of 50 μm) in CA3 stratum radiatum. Constant current pulses
of 0.1 ms were generated by a Grass stimulator (Grass Technologies, West Warwick, RI)
and delivered through an isolation unit every 30 s. After stable baseline recordings of
fEPSPs over 10 min in slices perfused with aCSF-1, the slices were perfused with aCSF-
1 containing 1 mM glycine for 10 min, and then perfused with the aCSF-1 alone for over
30 min. MultiClamp 700B (Molecular Devices) was used for the recording. Data
acquisition and analysis were performed using DigiData 1322A (Molecular Devices) and
the analysis software pClamp 10 (Molecular Devices). Signals were filtered at 2 kHz and
sampled at 10 kHz.
Transfections
The control siRNAs, GluN2A siRNAs or GluN2B siRNAs were transfected in
cultured hippocampal neurons and the cDNAs of GFP, GluN1, GluN2A, GluN2B and
GluN1 (N598Q) were transfected in cultured HEK293 cells. Transfections were
performed using Lipofectamine 2000 (Invitrogen) as previously described (Ning et al.
2004; Wan et al. 1997).
Western blotting
Western blotting assay was performed as described previously (Liu et al. 2006).
Antibodies against phospho-ERK1/2 (Thr202/Tyr204) (Cell Signaling Technology,
Beverly, MA) and total ERK1/2 (Cell Signaling Technology) were used. For the detection of phospho-ERK1/2, samples prepared in the same day were freshly used for the Western blotting assay for all the experiments. Primary antibodies were labeled with 85
horseradish peroxidase-conjugated secondary antibody. The phospho-ERK1/2 protein
bands were imaged using SuperSignal West Femto Maximum Sensitivity Substrate
(Pierce, Rockford, IL, USA). For the detection of total ERK1/2, the same polyvinylidene
difluoride membrane was stripped and then reprobed with primary antibody against total
ERK1/2 (Cell Signaling Technology). The ERK1/2 protein bands were imaged using
Pierce ECL Western Blotting Substrate (Pierce). The EC3 Imaging System (UVP, LLC,
Upland, CA) was used to obtain Western blot images directly from polyvinylidene
difluoride membranes. The quantification of Western blots was performed using ImageJ
software as previously described (Liu et al. 2006; Ning et al. 2004).
Statistics
All population data were expressed as mean ± s.e. The Student’s T test or the
ANOVA test was used when appropriate to examine the statistical significance of the
differences between groups of data. Significance was placed at p < 0.05.
5.4 Results
5.4.1 Glycine potentiates AMPA-induced whole-cell currents independent of
NMDAR channel activity.
To determine whether AMPAR function is regulated by a metabotropic activity of
NMDARs, we measured the effect of NMDAR co-agonist glycine on AMPAR function
in cultured mouse hippocampal neurons in which NMDARs were inhibited by MK-801, a
non-competitive antagonist preventing the flow of ions through the NMDAR channels
(MacDonald and Nowak 1990; Rosenmund et al. 1993). To ensure that no Ca2+ passed
through NMDAR channels, MK-801 was added into the ECS where Ca2+ was not 86
included but with the addition of 5 mM Ca2+ chelator EGTA and 10 μM strychnine. We
named this specific paradigm of ECS as ECS-1. The glycine receptor antagonist
strychnine was included in the ECS-1 to exclude the possible effects mediated by glycine activation of glycine receptors (Lynch 2004). The cultured neurons were treated with
ECS-1 for 10 min to block the channel activity of NMDARs. This treatment will be referred to as the NMDAR blockade protocol. We verified that the NMDAR blockade protocol inhibited NMDAR-mediated whole-cell currents in our experimental conditions
(Figure 5.1A).
Prior to recording AMPA-induced whole-cell currents, the neuronal cultures were subjected to the NMDAR blockade protocol. AMPAR currents were recorded in ECS-1 with the holding potential at -70 mV. Following a stable recording of AMPAR currents, glycine (100 μM) was continuously puffed onto the recorded neuron for 1 min. We found that the AMPAR peak currents were significantly increased after the treatment of glycine
(100 μM) and the currents were inhibited by specific AMPAR antagonist CNQX (Figure
5.1B).
To determine whether the observed effect of glycine on AMPAR currents occurred at physiologically relevant levels of extracellular Ca2+, we treated the neurons with standard ECS containing 10 μM MK-801, 10 μM strychnine and 0.5 μM TTX for 10 min. We then recorded AMPA-induced whole-cell currents and treated the neurons with glycine (100 μM). As shown in Figure 5.1C, glycine treatment for 1 min increased
AMPAR peak currents in the hippocampal neurons in which NMDAR channels were blocked by MK-801. 87
Endogenous Mg2+ blocks NMDAR channels while AMPAR whole-cell currents
were recorded at the holding potential of -70 mV (Kuner and Schoepfer 1996). To test the
glycine effect in a physiologically relevant condition in which NMDARs are not blocked
by the external application of channel blocker MK-801, we measured AMPAR currents
in neurons treated with standard ECS only containing 10 μM strychnine and 0.5 μM
TTX. We showed that without use of MK-801, glycine treatment (100 μM) for 1 min increased AMPAR peak currents with the holding potential at -70 mV (Figure 5.1D).
Together, these results indicate that glycine potentiates AMPAR function independent of the channel activity of NMDARs.
5.4.2 Glycine enhances AMPAR-mediated synaptic function independent of
NMDAR channel activity.
As the observed enhancement of AMPAR-mediated whole-cell currents might represent an upregulation of synaptic responses of AMPARs to glycine, we tested the effect of glycine on AMPAR-mediated miniature excitatory postsynaptic currents
(mEPSCs). NMDAR channels in the hippocampal neurons were blocked by the NMDAR blockade protocol as described above, and the neurons were bathed in the ECS-1 containing 10 μM GABAA receptor antagonist bicuculline for the entire recording period. Our data showed that AMPAR-mediated mEPSCs were significantly increased in neurons following 1 min treatment of glycine (100 μM), and this enhancement lasted for
~30 minutes (Figure 5.2A). Glycine application increased both amplitude and frequency of AMPAR mEPSCs (Figure 5.2A-C), suggesting an enhancement of postsynaptic
AMPAR function by glycine in a NMDAR channel activity-independent manner. 88
To validate the observed effect of glycine on synaptic AMPAR function in more
physiologically relevant conditions, we recorded field excitatory post-synaptic potentials
(fEPSP) in adult mouse hippocampal slices. AMPAR-mediated fEPSPs were
pharmacologically isolated by treating the slices with GABAA receptor antagonist
bicuculline (10 μM) and NMDAR blocker MK-801 (10 μM). As described in detail in the
Methods, the NMDAR channels in hippocampal slices were blocked by a NMDAR
blockade protocol that was similar to that in hippocampal neuronal cultures. After stable
baseline recordings of AMPAR-mediated fEPSPs for over 10 min, treatment of 1 mM glycine for 10 min led to an increased amplitude of AMPAR-mediated fEPSP in the hippocampus (Figure 5.2D).
5.4.3 Potentiation of AMPAR function by glycine requires ERK1/2 activation.
Because ERK1/2 is critically involved in mediating AMPAR-mediated synaptic plasticity (Kim et al. 2005a), we tested the effect of ERK1/2 inhibition on the upregulation of AMPAR function by glycine (Hotokezaka et al. 2002). The channels of
NMDARs in hippocampal neurons were blocked by the NMDAR blockade protocol, and
the hippocampal neurons were bathed in ECS-1 containing ERK1/2 inhibitor U0126 for
the recordings of both AMPAR-mediated mEPSCs and AMPA-induced whole-cell
currents. We found that U0126 treatment for the entire recording period significantly reduced the upregulation of both AMPAR mEPSC and AMPA-induced whole-cell
currents by glycine (100 μM; 1 min) (Figure 5.3A-D), suggesting that ERK1/2 activation
mediates glycine-induced potentiation of AMPAR function. 89
5.4.4 Glycine promotes ERK1/2 activation independent of NMDAR channel pore activities.
The electrophysiological results (Figures 5.1-3) led us to reason that the NMDAR co-agonist glycine might activate a metabotropic function of NMDARs to enhance
ERK1/2 activation that in turn lead to the enhancement of AMPAR function. To test this possibility, we performed Western blot assay to test the effect of glycine on ERK1/2 activation by measuring the phosphorylation level of ERK1/2 in cultured hippocampal neurons. The levels of ERK1/2 phosphorylation (p-ERK1/2) on Thr202/Tyr204 were quantified by calculating the ratio of p-ERK1/2 to total ERK1/2 (t-ERK1/2). After
NMDARs were blocked by the NMDAR blockade protocol, the cultures were treated with ECS-1 containing glycine (100 μM) for 1 min and then washed with ECS-1 for 30 min. The neurons were then collected for Western blot assay. As shown in Figure 5.4A, glycine increased ERK1/2 phosphorylation in hippocampal neurons where the NMDAR channel activity and glycine receptors were inhibited, and the glycine effect was dose- dependent (Figure 5.4B). In the same experimental conditions, hippocampal neurons were treated with ECS-1 containing 5 mM BAPTA, a Ca2+ chelator that has faster Ca2+- binding kinetics than EGTA (Adler et al. 1991). We found that BAPTA treatment did not interfere with glycine elevation of ERK1/2 phosphorylation (Figure 5.4C). As BAPTA chelates the residual Ca2+ in the ECS-1, this result provides further evidence suggesting that the effect of glycine on ERK1/2 phosphorylation is independent of extracellular
Ca2+. 90
To determine whether the observed effect of glycine on ERK1/2 phosphorylation
occurred at physiologically relevant levels of extracellular Ca2+, we treated the neurons
with standard ECS containing 10 μM MK-801, 10 μM strychnine and 0.5 μM TTX for 10
min. The neurons were treated with standard ECS plus 100 μM glycine for 1 min and
then washed with standard ECS. We found that at 30 min after the treatment of 100 μM
glycine, the levels of ERK1/2 phosphorylation were elevated independent of NMDAR
channel activity and glycine receptor activation (Figure 5.4D).
Consistent with the electrophysiological finding in hippocampal slices (Figure
5.2D), glycine (1 mM) enhanced ERK1/2 phosphorylation independent of NMDAR
channel activity and glycine receptor activation in hippocampal slices (Figure 5.4E). The
treatment procedure for hippocampal slices was same as that of fEPSC recordings
described above.
5.4.5 Glycine enhances ERK1/2 activation through a metabotropic activity of
GluN2ARs.
In order to obtain direct evidence to determine whether a metabotropic NMDAR
mediated glycine potentiation of ERK1/2 activation, we tested the effects of glycine on
ERK1/2 phosphorylation in HEK293 cells transiently expressing NMDARs. The cDNAs
of GluN1, GluN2A and/or GluN2B subunits were transfected in various combinations
into the HEK293 cells (Wan et al. 1997). Prior to the treatment of glycine (100 μM), the
transfected cells were subjected to the NMDAR blockade protocol. ERK1/2
phosphorylation was measured in the transfected cells at 30 min after 1 min treatment of
glycine (100 μM) as described above (Figure 5.4A). We found that glycine had no effect 91
on ERK1/2 phosphorylation in non-transfected HEK293 cells (Figure 5.5A). However, glycine increased ERK1/2 phosphorylation in HEK293 cells transfected with cDNAs of
GluN1+GluN2A (Figure 5.5B) and cDNAs of GluN1+GluN2A+GluN2B (Figure 5.5C),
but not in cells transfected with cDNAs of GluN1+GluN2B (Figure 5.5D). We also
showed that glycine did not increase ERK1/2 phosphorylation in HEK293 cells
transfected with cDNAs of GluN1, GluN2A or GluN2B, respectively (Figure 5.5E).
Thus, glycine preferentially acted on GluN2ARs but not GluN2BRs to enhance ERK1/2
phosphorylation independent of the channel activity of GluN2ARs, indicating that a
metabotropic activity of GluN2ARs mediates glycine elevation of ERK1/2
phosphorylation. As the vast majority of synaptic NMDARs are trimeric NMDARs
containing GluN1, GluN2A and GluN2B subunits, our data indicate that GluN2A but not
GluN2B is required for synaptic metabotropic NMDARs to mediate the glycine effect.
Amino acid N598 is a critical residue at the selectivity filter of NMDAR channel
that determines Ca2+ permeability and GluN1 mutant N598Q has been shown to reduce
Ca2+ permeability (Burnashev et al. 1992). We transfected cDNAs of GluN2A with
GluN1(N598Q) in HEK293 cells, and showed that at 30 min after 1 min treatment of
glycine (100 μM) ERK1/2 phosphorylation in cells co-transfected with GluN2A and
GluN1(N598Q) was increased (Figure 5.5F). Although GluN1(N598Q) only causes a
four-fold decrease of Ca2+ permeability of NMDARs, these results provide molecular
evidence to support the conclusion that GluN2AR-mediated ERK1/2 activation is
independent of Ca2+ influx. 92
We next applied a knockdown approach to validate the role of metabotropic
GluN2ARs in mediating glycine enhancement of ERK1/2 activation in cultured
hippocampal neurons. The GluN2A protein expression was suppressed in neurons transfected with GluN2A siRNA (Figure 5.6A). The NMDAR blockade protocol was used to block NMDAR channels. As expected, ERK1/2 phosphorylation was increased in neurons transfected with the siRNA control at 30 min after 1 min treatment of glycine
(100 μM); (Figure 5.6B), but the effect of glycine was significantly reduced in neurons
transfected with GluN2A siRNA (Figure 5.6B). In contrast, glycine increased ERK1/2
phosphorylation in neurons where GluN2B expression was suppressed by GluN2B
siRNA treatment (Figure 5.6C-D).
5.4.6 A metabotropic activity of GluN2ARs mediates glycine-induced potentiation of
AMPAR function.
We thus far showed that glycine-induced potentiation of AMPAR function
required ERK1/2 activation in a NMDAR channel activity-independent manner, and that
a metabotropic activity of GluN2ARs mediated the elevation of ERK1/2 phosphorylation
by glycine. These results support a possibility that a metabotropic GluN2AR mediates the
potentiation of AMPAR function by glycine. To test this, we first measured the effect of
glycine on AMPA-induced whole-cell currents in neurons transfected with GFP+GluN2A
siRNA. In contrast to neurons transfected with GFP+siRNA control (Figure 5.7A) or
GFP+GluN2B siRNA (Figure 5.7B), neurons transfected with GFP+GluN2A siRNA
exhibited no significant increase of AMPAR currents at 30 min following 1 min
treatment of glycine (100 μM); (Figure 5.7C). The NMDAR blockade protocol was 93
applied to block NMDAR channels in the transfected neurons. These data suggest that glycine activates metabotropic GluN2ARs to enhance AMPAR function. 94
Figure 5.1 Glycine enhances AMPAR-mediated whole-cell currents in hippocampal neurons in which the NMDAR channel activity and glycine receptor activation are inhibited.
(A) Sample currents showing that NMDAR-mediated whole-cell current is
not induced by NMDAR agonist aspartate (100 µM) and co-agonist glycine
(1 µM) in neurons pretreated with ECS-1. (B) AMPA (100 µM)-induced
whole-cell currents are increased by 1 min treatment of 100 µM glycine in
hippocampal neurons after NMDARs and glycine receptors are inhibited
(n=15, *p<0.05). The current is reversibly blocked by AMPAR antagonist 95
CNQX (20 µM). (C) At normal levels of extracellular Ca2+, glycine (100
µM) increases AMPAR currents independent of the channel activity of
NMDARs and the activation of glycine receptors (n=6, *p<0.05). (D)
Without treatment with MK-801, AMPAR peak currents are increased by glycine (100 μM) treatment under the holding potential of -70 mV with which the NMDAR channels are blocked by Mg2+ (n=7, *p<0.05). 96
Figure 5.2 Glycine enhances AMPAR-mediated synaptic currents independent of NMDAR channel activity.
(A) Representative AMPAR-mediated mEPSCs recorded before and 10 min
after treatment of glycine (100 µM) in hippocampal neurons where
NMDARs and glycine receptors are inhibited. (B) Summarized data showing
that mean amplitude of AMPAR mEPSCs is significantly increased after 100 97
µM glycine treatment (n=8, *p<0.05) and that the mean frequency of
AMPAR mEPSCs is also increased after glycine treatment (n=8, *p < 0.05).
(C) Left: sample of averaged AMPAR mEPSCs from the neurons before
(995 events) and after (1407 events) treatment of 100 µM glycine. Right: cumulative probability plots of peak amplitudes of AMPAR mEPSCs (bin size 0.5 pA). The mEPSC amplitude distribution significantly shifts towards greater values after treatment of 100 µM glycine (p<0.05). (D) Left: averaged amplitudes of AMPAR-mediated fEPSP in hippocampal slices recorded before, during and after 1 mM glycine treatment (n=12). NMDARs and glycine receptors in the slices are inhibited before recording. Right: summarized data showing that glycine (1 mM) increases the amplitudes of
AMPAR-mediated fEPSP in hippocampal slices independent of NMDAR channel activity and glycine receptor activation (n =12, *p < 0.05 vs. Before).
98
Figure 5.3 Inhibition of ERK1/2 activation prevents potentiation of AMPAR function by glycine.
(A) Representative AMPAR mEPSCs recorded before and 10 min after 1
min application of 100 µM glycine in cultured hippocampal neurons in which
NMDARs and glycine receptors are inhibited. (B) Summarized data show
that the enhancement of AMPAR mEPSCs by glycine is antagonized by
pretreatment of ERK1/2 inhibitor U0126 (5 µM; n=6). (C) Representative
AMPAR whole-cell currents induced by AMPA (100 µM) recorded before
and 10 min after treatment of 100 µM glycine in hippocampal neurons where
NMDARs and glycine receptors are inhibited. (D) Summarized data show
that the enhancement of AMPAR whole-cell currents by glycine was blocked 99
by U0126 pretreatment (n=8). 100
Figure 5.4 Glycine increases ERK1/2 phosphorylation independent of
NMDAR channel activity in hippocampal neurons.
(A) Glycine (100 μM) increases p-ERK1/2 after NMDARs and glycine
receptors are inhibited (n=8, *p<0.05). (B) Glycine-induced increase of
ERK1/2 phosphorylation is dose-dependent in hippocampal neurons where
NMDARs and glycine receptors are inhibited. (C) BAPTA (5 mM) treatment
does not influence glycine elevation of ERK1/2 phosphorylation in
hippocampal neurons where NMDARs and glycine receptors are inhibited
(n=6, *p<0.05). (D) At normal level of extracellular Ca2+, glycine (100 μM)
increases p-ERK1/2 in hippocampal neurons where NMDARs and glycine
receptors are inhibited (n=6, *p<0.05). (E) Glycine (1 mM) enhances 101
ERK1/2 phosphorylation in hippocampal slices after NMDARs and glycine receptors are inhibited (n=5, *p<0.05). Con: Control; Gly: glycine. 102
Figure 5.5 Glycine increases ERK1/2 phosphorylation through metabotropic activity of GluN2ARs in HEK293 cells.
(A) The levels of ERK1/2 phosphorylation are not altered by glycine (100
μM) treatment in non-transfected HEK293 cells where NMDARs and glycine
receptors are inhibited (n=6). (B) In HEK293 cells transfected with
GluN1+GluN2A cDNAs, the ERK1/2 phosphorylation is increased by
glycine (100 μM) treatment after NMDARs and glycine receptors are
inhibited (n=6, *p<0.05). (C) In HEK293 cells transfected with 103
GluN1+GluN2A+GluN2B cDNAs, the ERK1/2 phosphorylation is increased by glycine (100 μM) treatment after NMDARs and glycine receptors are inhibited (n=6, *p<0.05). (D) In HEK293 cells transfected with
GluN1+GluN2B cDNAs, the levels of ERK1/2 phosphorylation are not altered by glycine (100 μM) treatment after NMDARs and glycine receptors are inhibited (n=6). (E) In HEK293 cells transfected with cDNAs of GluN1,
GluN2A or GluN2B, respectively, the levels of ERK1/2 phosphorylation are not altered by glycine (100 μM) treatment after NMDARs and glycine receptors are inhibited (n=6). (F) In HEK293 cells transfected with
GluN1(N598Q)+GluN2A, glycine enhances ERK1/2 phosphorylation (n = 6,
*p < 0.05 vs. -Gly). Con: Control; Gly: glycine. 104
Figure 5.6 Glycine increases ERK1/2 phosphorylation via metabotropic activity of GluN2ARs in hippocampal neurons.
(A) The GluN2A protein expression in cultured mouse hippocampal neurons
is suppressed by GluN2A siRNA (n=5, *p<0.05). (B) GluN2A knockdown
by GluN2A siRNA blocks glycine-induced increase of p-ERK1/2 in neurons
where NMDARs and glycine receptors are inhibited (n=5, *p<0.05 vs.
siRNA control). (C) The GluN2B protein expression in mouse hippocampal
neurons is suppressed by transfection of GluN2B siRNA (n=5, *p< 0.05). (D)
GluN2B knockdown by GluN2B siRNA does not block glycine increase of
p-ERK1/2 in neurons where NMDARs and glycine receptors are inhibited
(n=5, *p<0.05 vs. siRNA control; #p<0.05 vs. GluN2B siRNA). Gly: glycine.
105
Figure 5.7 Glycine enhances AMPAR function through metabotropic activity of GluN2ARs in hippocampal neurons.
(A) Transfection of siRNA control does not influence glycine-induced
potentiation of AMPAR currents in hippocampal neurons after NMDARs
and glycine receptors are inhibited (n=7, *p<0.05). (1) siRNA control; (2)
siRNA control+Gly. (B) GluN2B knockdown by GluN2B siRNA
transfection does not influence glycine (100 μM) potentiation of AMPA-
induced whole-cell currents in neurons where NMDARs and glycine
receptors are inhibited (n=7, p<0.05). (1) GluN2B siRNA; (2) GluN2B
siRNA + Glycine. (C) Knockdown of GluN2A by GluN2A siRNA 106
transfection blocks glycine-induced potentiation of AMPAR currents in hippocampal neurons after NMDARs and glycine receptors are inhibited
(n=7). (1) GluN2A siRNA; (2) GluN2A siRNA+Gly. Gly: glycine. 107
5.5 Discussion
It is generally believed that the function of NMDAR requires Ca2+ influx through
the NMDAR channel (Dingledine et al. 1999). In this study, with the blockade of
NMDARs, we provide evidence that glycine alone increases AMPAR-mediated synaptic
function without depending on the channel activity of GluN2ARs. In the same
experimental conditions, we demonstrate that ERK1/2 is downstream of GluN2ARs to
mediate glycine-induced potentiation of AMPAR function. Thus, we identify a
metabotropic function of GluN2ARs. Recent evidence indicates that GluN2BR has a
metabotropic activity that is required for β-amyloid–induced synaptic depression and sufficient to induce LTD (Nabavi et al. 2013; Tamburri et al. 2013). Together, these findings suggest that the metabotropic activity of NMDARs has functional significance.
Glycine is a co-agonist of NMDARs (Johnson and Ascher 1987). The channel
activation of GluN2ARs and GluN2BRs requires both glutamate and glycine (Johnson
and Ascher 1987). We reveal an unexpected role of glycine, independent of glutamate, to
trigger the metabotropic activity of GluN2ARs. This finding indicates that, while
function as a co-agonist for the ionotropic activation of NMDARs, glycine alone
activates a metabotropic function of GluN2ARs. Importantly, we show that glycine acts
on the metabotropic GluN2ARs to regulate AMPAR function, implying a functional
significance of glycine-induced activation of metabotropic GluN2ARs.
We demonstrate that GluN2A but not GluN2B subunit is required for the glycine
effect on NMDARs. Although the vast majority of synaptic NMDARs are trimeric
NMDARs containing GluN1, GluN2A and GluN2B subunits, GluN2ARs and GluN2BRs 108
are shown to play different roles in regulating neuronal survival/death and synaptic
plasticity (Chen et al. 2008; Liu et al. 2004; Liu and Zukin 2007). The mechanism
underlying the differential effects of these NMDAR subtypes has been elusive. Our study
reveals that a metabotropic activity of GluN2ARs regulates AMPAR function at synaptic
sites. This observation provides a mechanism that may explain in part why GluN2AR
plays a different role than GluN2BR in synaptic plasticity.
It is not clear how glycine binds to metabotropic GluN2ARs to activate ERK1/2.
One of the possibilities is that a structural rearrangement of GluN1 and/or GluN2A but not GluN2B upon glycine binding may directly or indirectly lead to the activation of
ERK1/2-dependent signaling. The ERK1/2 signaling is known to be activated by
NMDARs and plays an important role in synaptic plasticity (Sweatt 2004; Thomas and
Huganir 2004). NMDAR-dependent ERK1/2 activation involves the small GTPase Ras,
which is stimulated by specific nucleotide exchange factors (GEFs) (Thomas and
Huganir 2004). It has been shown that GluN2ARs and GluN2BRs have antagonistic
actions on Ras-ERK1/2 activation (Kim et al. 2005b). GluN2ARs promote, whereas
GluN2BRs inhibit, Ras-ERK1/2 activation (Kim et al. 2005b). Through Ras-GRF2 (a
Ras-GEF) and ERK1/2 signaling pathway, GluN2AR induces long-term potentiation
(LTP) in CA1 pyramidal neurons of mouse hippocampus (Jin and Feig 2010). It remains
unknown whether the Ras-GRF2 is involved in metabotropic GluN2AR-mediated
enhancement of ERK1/2 activation. 109
Chapter 6 Single Agonist NMDA does not Regulate the Metabotropic Signaling of the NMDA Receptor
Brendan Lujan, Robert Renden, Qi Wan
110
6.1 Summary
NMDA receptors (NMDARs) are Ca2+-permeable ion channels, whose activation requires agonist glutamate and co-agonist glycine. The ionotropic function of the
NMDAR has been well studied and is linked to the function of AMPA receptors
(AMPARs), a process crucial for synaptic plasticity. Here we sought to test the hypothesis that single agonist NMDA can regulate synaptic plasticity independent of the ionotropic function of the NMDAR via an extracellular regulated kinase ½ (ERK1/2) pathway. We targeted the (ERK1/2) signaling cascade due to its well documented involvement in synaptic plasticity events. However, we report that NMDA does not regulate ERK1/2 signaling independent of the NMDAR channel pore by application of the single agonist NMDA. Furthermore, we provide functional evidence that single agonist NMDA does not regulate spontaneous or evoked responses when compared to controls. These data suggest that NMDA does not act as a sole agonist to the NMDAR to regulate metabotropic signaling events and downstream postsynaptic plasticity.
6.2 Introduction
Neurotransmitter receptors are often categorized broadly into two distinct groups: metabotropic receptors that activate intracellular signaling pathways and ionotropic receptors that mediate ion flux. However, it has been suggested that the ionotropic glutamate receptors possess both the capacity to signal through ionotropic and metabotropic mechanisms to affect various aspects of neuronal function (Hayashi et al.
1999; Kessels et al. 2013; Nabavi et al. 2013; Rodriguez-Moreno and Lerma 1998; Vissel et al. 2001; Wang et al. 1997). The first study supporting this idea suggested that AMPA 111
receptors (AMPARs) required ligand binding, but no flux of either Na+ or Ca2+ across
the cell membrane, to sufficiently regulate Gi protein-mediated signaling in primary
cultured cortical neurons (Wang et al. 1997). Furthermore this study provided evidence
that the Gi protein made specific interactions with the GluR1 subunit, assayed by
immunoprecipitation, to mediate these signaling pathways. Another study suggested that
the ionotropic glutamate receptor subfamily member kainate receptor also possessed a
metabotropic signaling mechanism by which a PKC-dependent pathway was described to functionally inhibit GABA input to excitatory neurons in the hippocampus (Rodriguez-
Moreno and Lerma 1998). Subsequently, another study involving the AMPAR provided even more evidence of a metabotropic signaling process as Na+ and Ca2+ were deemed non-essential components in regulation of the Src-family non-receptor protein tyrosine kinase Lyn. The MAPK signaling pathway was activated by Lyn resulting in an increase in mRNA of the brain-derived neurotrophin factor (BDNF), suggesting a role in regulation of synaptic plasticity (Hayashi et al. 1999). These studies suggest that both the
AMPAR and kainate receptor subfamily members possess both ionotropic and metabotropic signaling mechanisms.
The NMDA receptor (NMDAR) was subsequently implicated in possessing both ionotropic and metabotropic signaling capacities. Metabotropic activation of the
NMDAR was suggested to regulate phosphorylation of the C-terminal region of the
GluN2A subunit at Tyrosine 842 by ligand binding, but no ion flux through the NMDAR
(Vissel et al. 2001). Similarly, ligand-binding without ion flux through the NMDAR was shown to be sufficient to induce LTD using a low-frequency LTD induction protocol in 112
acute hippocampal slices in the presence of the NMDAR channel blocker MK-801 and
was P-p38 dependent (Nabavi et al. 2013). The LTD induction was blocked when
NMDA receptor antagonist DAPV, specific to the glutamate binding site on the GluN2
subunit of the NMDAR, was added to the bath. However, these reports were
subsequently refuted in the following year, suggesting LTD induction in the hippocampus
is highly dependent on Ca2+ influx and metabotropic induction of LTD does not exist
(Babiec et al. 2014). Further study suggests that metabotropic activation of the NMDAR also regulates dendritic shrinkage (Stein et al. 2015). Thus, the metabotropic signaling
capacity of the NMDAR remains a highly controversial topic.
We sought to test the hypothesis that metabotropic NMDAR activation can
regulate an ERK1/2-dependent signaling cascade and downstream AMPAR function, which requires ligand-binding but not ion flux by the application of single agonist
NMDA, as opposed to single agonist glycine as shown in Chapter 5. Here we show that
metabotropic activation of the NMDAR after channel block using single agonist
stimulation with chemical NMDA treatment has no effect on ERK1/2 signaling in
cultured hippocampal neurons. Furthermore, we observed no functional changes in the
spontaneous mEPSC frequency nor amplitude. Nor did we observe any change in
AMPA-mediated whole-cell EPSC responses during a 30 minute recording period after
metabotropic activation of the NMDAR compared to control recordings. These data
suggest that single agonist NMDA does not regulate AMPAR function in primary
hippocampal neuronal culture. Furthermore, this study does not support glutamate 113
activation of metabotropic signaling of the NMDAR to regulate AMPAR function, contrary to a previous report in a hippocampal slice preparation (Nabavi et al. 2013).
6.3 Methods
Hippocampal neuronal culture
The hippocampal neuronal cultures were prepared from C57BL/6 mice at gestation day 17 using a modified protocol (Shan et al. 2009). C57BL/6 mice were obtained from the University of Nevada School of Medicine (UNSOM) or Charles River
Labs. Briefly, dissociated neurons were suspended in plating medium (Neurobasal medium, 2% B-27 supplement, 10% FBS, 0.5 μM L-glutamine, and 25 μM glutamic acid) and plated on poly-D-lysine coated Petri dishes. After 3 days in culture, half of the plating medium was removed and replaced with maintenance medium (Neurobasal medium, 2% B-27 supplement, and 0.5 μM L-glutamine). Thereafter, maintenance medium was changed in the same manner every 3 days. The cultured neurons were used for all the experiments at 10-12 days after plating. All animal work was conducted according to the guidelines set forth the UNSOM Institutional Animal Care and Use
Committee (IACUC). All procedures were approved by the UNSOM IACUC.
Western blotting
Western blotting assay was performed as described previously (Liu et al. 2006).
Antibodies against phospho-ERK1/2 (Thr202/Tyr204) (Cell Signaling Technology,
Beverly, MA) and total ERK1/2 (Cell Signaling Technology) were used. For the detection of phospho-ERK1/2, samples prepared in the same day were freshly used for the Western blotting assay for all the experiments. Primary antibodies were labeled with 114
horseradish peroxidase-conjugated secondary antibody. The phospho-ERK1/2 protein bands were imaged using SuperSignal West Femto Maximum Sensitivity Substrate
(Pierce, Rockford, IL, USA). For the detection of total ERK1/2, the same polyvinylidene difluoride membrane was stripped and then reprobed with primary antibody against total
ERK1/2 (Cell Signaling Technology). The ERK1/2 protein bands were imaged using
Pierce ECL Western Blotting Substrate (Pierce). The EC3 Imaging System (UVP, LLC,
Upland, CA) was used to obtain Western blot images directly from polyvinylidene difluoride membranes. The quantification of Western blots was performed using ImageJ software as previously described (Liu et al. 2006; Ning et al. 2004).
Electrophysiological recordings
For the recording of AMPA-induced whole-cell currents, the cultures were bathed in an extracellular solution (ECS; 10 μM MK-801, 5 mM EGTA, 10 μM strychnine, 137 mM NaCl, 5.4 mM KCl, 1 mM MgCl2, 25 mM HEPES, 33 mM Glucose, titrated to pH
7.4 with osmolarity of 300-320 mOsm) with the addition of 0.5 μM TTX. Neurons were held at -70 mV under voltage-clamp. AMPA receptor-mediated whole-cell currents were recorded by pressure application of 100 μM AMPA (100 kPa, 200 ms) from a micropipette with its tip located ~20 μm from the recorded cell. Drugs were delivered at intervals of 3 mins. Data were acquired with an Axopatch 200B amplifier and pClamp 10 software interfaced to a Digidata 1322A acquisition board (Molecular Devices, CA), and signals were filtered at 2 kHz and digitized at 10 kHz.
Recording of miniature EPSCs (mEPSCs) was performed as described previously
(Liu et al. 2006). The cultures were bathed in ECS containing with 10 μM bicuculline to 115
record AMPAR mEPSCs. Five minutes of individual AMPA receptor-mediated mEPSCs were collected before and after application of NMDA (50 μM). Records were filtered at
2.9 kHz and analyzed with a Clampfit 10.3 program (Molecular Devices). The other experimental conditions and methods were same as those of recording for AMPAR- mediated whole-cell currents.
6.4 Results
6.4.1 Metabotropic NMDAR activation fails to regulate ERK1/2 by the synthetic agonist NMDA
In order to test whether the application of the single synthetic agonist NMDA could metabotropically activate NMDAR signaling to regulate ERK1/2, experimental conditions were set to inhibit the NMDAR channel pore by MK-801 in primary neuronal cultures of the hippocampus. MK-801 is an irreversible channel-blocker of NMDARs that resists the channel’s permeability to ionic flow of Na+, K+ and importantly, Ca2+
(MacDonald and Nowak 1990). Since MK-801 is a use-dependent antagonist, primary neuronal cell cultures were treated with an NMDAR blockade protocol: the ECS was applied in the presence of 5 μM glycine and 5 μM NMDA, for 5 min to inhibit
NMDARs. The NMDAR blockade protocol supplies the NMDAR ligands to stimulate channel opening, allowing subsequent blocking of the channel pore by insertion of MK-
801. After the ion pore of the NMDARs had been inhibited with the NMDAR blockade protocol, the neuronal cultures were allowed to assume a steady-state during a 20 minute period before treating with varying concentrations of the synthetic NMDAR agonist
NMDA for 30 min added to the ECS. Subsequently, ERK1/2 activation was assayed via 116
its phosphorylation (p-ERK) on Thr 202 and Tyr 204 residues by Western blot (Figure
6.1A). We observed no significant increases in p-ERK normalized to total whole cell
ERK1/2 (t-ERK) at all concentrations of NMDA tested. This result suggests that
metabotropic activation of the NMDAR does not regulate ERK1/2 signaling. We next
performed a time course experiment to observe whether we could capture the kinetics of
metabotropic NMDAR activation of ERK1/2 signaling. Activation of ERK1/2 signaling
was not effected at all time points tested (Figure 6.1B). These data suggest metabotropic
activation of the NMDAR by high concentrations of NMDA does regulate ERK1/2
signaling independent of Ca2+ ion flux through the NMDAR itself.
To probe the binding location necessary to metabotropically activate the
NMDAR, the NMDAR antagonist DAPV was employed, which competitively inhibits
the receptor at the glutamate binding site on the GluN2 subunit. After NMDAR blockade
but before NMDA treatment, application of 50 µM DAPV had no further effect on the
metabotropic signaling capacity of the NMDAR to regulate ERK1/2 activation (Figure
6.1C). These data suggest that ligand binding to the NMDAR, without ion flux, does not
regulate ERK1/2 signaling via a metabotropic action of the NMDAR through single
agonist application of NMDA.
6.4.2 mEPSCs are unresponsive to metabotropic activation of the NMDA receptor
Spontaneous mEPSCs were recorded from primary hippocampal neuronal
cultures at 10-12 DIV to assess whether metabotropic activation of the NMDAR
regulates spontaneous mEPSC amplitude or frequency. Before cultures were placed into the recording chamber, NMDARs were inhibited with the NMDAR blockade protocol 117
and cells were selected for recording based on pyramidal morphology. A 5 min baseline of spontaneous activity was recorded in the ECS solution in which 10 µM bicuculline,
0.5 µM strychnine and 0.5 µM TTX had been included. After a stable baseline recording was made, cultures were treated with 50 µM NMDA for 3 min and the response spontaneous activity was recorded and compared to baseline (Figure 6.2A). An increase in mEPSC amplitude suggests an increase in postsynaptic surface expression of receptors whereas a decrease in mEPSC amplitude suggests removal of postsynaptic receptors from the neuronal surface. Changes in the frequency of mEPSCs are attributed to variations in vesicle release probability of the presynaptic compartment. No changes were observed to in mEPSC amplitude or frequency (Figure 6.2B-C). From these data we conclude that metabotropic activation of the NMDA receptor by single agonist application of NMDA does not regulate mEPSC amplitude or frequency.
6.4.3 Single Agonist NMDA does not regulate whole-cell AMPAR-mediated responses after metabotropic activation of the NMDAR
To test whether metabotropic activation of the NMDAR by the single agonist
NMDA regulates AMPAR function, we performed whole-cell patch clamp recordings and induced AMPAR activity via puff application of AMPA (100 µM) while recording from the cell body. Before primary hippocampal cultures were placed into the recording chamber, NMDAR channel pores were inhibited with the NMDAR blockade protocol.
Baseline whole-cell AMPAR-mediated EPSCs were compared to response EPSCs that were washed in either bath solution (control) or the bath solution containing 50 µM
NMDA (NMDA treatment) at 30 minutes. The experimental protocol is shown (Figure 118
6.3A). A significant increase was observed in those response EPSCs receiving control treatment (Figure 6.3B-C). Similarly, we observed an increase in EPSCs treated with
NMDA (Figure 6.4D-E). These data suggest that even without metabotropic activation of the NMDA receptor, whole-cell AMPA receptor mediated current amplitude increased simply through the recording protocol. These data suggest that metabotropic activation of the NMDAR by single agonist NMDA does not significantly regulate AMPAR function. 119
Figure 6.1 Metabotropic activation of the NMDAR by single agonist NMDA does not regulate ERK1/2 signaling in hippocampal neurons.
(A) Metabotropic activation of the NMDAR by single agonist NMDA has no
effect on ERK1/2 activation (n≥4 in all conditions tested, p>0.05 vs. control).
Representative blot (left) and summarized data (right) are shown. (B) NMDA
(50 µM) has no effect on ERK1/2 activation up to 30 minutes post treatment
(n=7, p>0.05 vs. control). (C) DAPV (50 µM) has no effect on ERK1/2
activation by NMDA treatment while blocking NMDAR channel pore (n=7, 120
p>0.05). Representative blot (left) and summarized data (right). P>0.05).
121
Figure 6.2 Metabotropic activation of the NMDAR by single agonist NMDA does not regulate mEPSCs
Spontaneous mEPSCs were recorded before and after NMDA treatment (50
µM) while NMDAR channels are inhibited. (A) Representative traces of
baseline (left, black) and response (right, red) spontaneous mEPSCs before
and after metabotropic activation of the NMDAR with NMDA (50 µM).
mEPSC amplitude (B) and frequency (C) were unaffected (n=12, p>0.05).
122
Figure 6.3 Metabotropic activation of the NMDAR by single agonist NMDA does not regulate AMPAR-mediated whole-cell currents
Baseline and response AMPAR-mediated EPSCs were recorded before and 123
after a control or NMDA treatment during NMDAR inhibition. Baseline
EPSC peak amplitudes were compared to those EPSCs recorded 30 minutes after the control or NMDA treatment. (A) Summary of recording protocol.
(B) Baseline (black) and response (gray) sample traces of AMPAR-mediated
EPSCs. (C) Summary data of those cells receiving a control treatment. EPSC peak amplitude was increased (n=7, *p<0.05). (D) Baseline (black) and response (red) EPSCs are shown before and after NMDA (50 µM) treatment while NMDAR channel activities were inhibited. (E) Summary data of those cells receiving NMDA treatment (50 µM). Response EPSCs were potentiated
(n=6, *p<0.05).
124
6.5 Discussion
Here we sought to test if application of the single synthetic agonist NMDA could
metabotropically activate the NMDAR. We show that after the NMDAR channel pore
had been inhibited, that application of NMDA does not regulate ERK1/2 activation
assayed through its phosphorylation on Thr 202 and Tyr 204 residues by Western
blotting. However, recently it has been shown that similar treatment protocol does indeed
regulate p38 kinase signaling (Nabavi et al. 2013). My data strongly refutes the
hypothesis that single agonist application of NMDA regulates ERK1/2 signaling after the
channel pore has been blocked.
We further sought to test if treatment with NMDA had any functional effect on
synaptic transmission after NMDAR channel pores had been inhibited. We observed no
effect of this treatment on spontaneous mEPSC frequency and amplitude. However, we
did observe an increase of AMPAR-mediated whole cell currents within a 30 minute
recording period. Interestingly, we observed potentiated EPSC peak amplitude response
in control conditions as well. These data suggest that the potentiated EPSC responses are
presumably due an artifact present in the primary hippocampal neuronal culture. Because
these neurons were grown in a medium containing both high glucose and glutamate, we
propose that upon switching to our ECS bath recording solution, which contained
substantially lower amounts of glucose and no glutamate, a homeostatic synaptic
plasticity mechanism was occurring. We believe, in response to switching from high to low glucose and eliminating glutamate, a homeostatic mechanism was induced to 125
increase AMPAR surface expression to the membrane in order to compensate for loss of these compounds.
126
Chapter 7 Glycolysis Selectively Shapes the Presynaptic Action Potential Waveform at the Calyx of Held
Brendan Lujan, Christopher Kushmerick, and Robert Renden.
127
7.1 Summary
Mitochondria are major suppliers of cellular energy in neurons; however,
utilization of energy from glycolysis versus mitochondrial oxidative phosphorylation
(OxPhos) in the presynaptic compartment during neurotransmission is largely unknown.
Using presynaptic and postsynaptic recordings from the mouse calyx of Held, we
examined the effect of acute selective pharmacological inhibition of glycolysis or
mitochondrial OxPhos on action potential (AP) generation, Ca2+-influx, and subsequent
synaptic transmission. We show that membrane polarization at the terminal is
preferentially fueled by glycolytic ATP production. Inhibition of glycolysis via glucose
depletion and Iodoacetic Acid (IAA, 1mM) treatment rapidly attenuated synaptic
transmission, due to a smaller and broader presynaptic action potential (AP) waveform.
We show via experimental manipulation and ion channel modeling that the altered AP waveform results in smaller Ca2+-influx, resulting in attenuated excitatory postsynaptic
currents (EPSCs). In contrast, inhibition of mitochondria-derived ATP production via
extracellular pyruvate depletion and bath-applied oligomycin (1 µM) had no significant effect on Ca2+-influx, did not alter AP waveform within the same time frame (up to 30 min), and the resultant EPSC remained unaffected. Thus, non-oxidative glycolysis, but not mitochondrial OxPhos, is requisite for normal synaptic transmission. We propose that glycolytic enzymes are closely apposed to ATP-dependent ion pumps on the presynaptic membrane, as has been shown previously for non-neuronal tissues. This study shows that attenuation of transmission due to acute hypoglycemia results from a 128
single mechanism: a slower, smaller action potential, prior to and independent of any
effect on synaptic vesicle (SV) release or receptor activity.
7.2 Introduction
Central nervous system (CNS) function is acutely dependent on constant energy
supply. Synaptic dysfunction and loss of consciousness due to hypoxia occurs within minutes, and hypoglycemia results in significant cognitive impairment, prior to altered cytosolic ATP (Fleck et al. 1993; Yamane et al. 2000). From studies in hippocampal and cerebellar slice, presynaptic function is estimated to consume ~12% of the neuronal energy budget, with >60% of ATP generated via mitochondrial respiration (Feldman and
Barshi 2007; Hall et al. 2012; Harris et al. 2012). Recent work has pointed to SV recycling as a major consumer of presynaptic ATP, as defects to SV recycling occur after inhibition of either glycolysis or mitochondrial OxPhos (Pathak et al. 2015; Rangaraju et al. 2014). Selective inhibition of one or the other of these pathways has been shown previously to have differential effects on transmission, with various onset time course, depending on the preparation (Ames and Gurian 1963; Okada 1982; Schurr et al. 1989).
However, loss of APs was generally reported to precede loss of synaptic transmission
(Ames and Gurian 1963; Okada 1982).
While presynaptic mitochondria play a significant role in energy production at the
presynaptic terminal following stimulation (Kosterin et al. 2005; Talbot et al. 2007), evidence suggests that glycolysis can sustain marginal cellular energy in the presynaptic terminal. For example, loss of mitochondria from glutamatergic presynaptic terminals impairs sustained transmission, but basal transmission is largely intact (Guo et al. 2005; 129
Sun et al. 2013; Verstreken et al. 2005). Correlative studies indicate that non-oxidative
glycolysis is utilized for neuronal function (Ivanov et al. 2014; Lundgaard et al. 2015). In
vivo, glucose uptake exceeded oxidative consumption in the brain by 2.5-fold (Fox et al.
1988), and stimulation resulted in an increase in non-oxidative glycolysis by-products
(Prichard et al. 1987; Ueki et al. 1988). Under some conditions, glycolysis has even been shown to preferentially support neurotransmission over mitochondrial OxPhos (Bak et al.
2006; Bak et al. 2009; Jang et al. 2016).
Glycolytic enzymes are localized to Na+/K+ ATP-dependent ion pumps at the
plasma membrane of non-neuronal cells (Knull 1978; Paul et al. 1989; Zala et al. 2013),
and also in the brain (Lipton and Robacker 1983; Raffin et al. 1988; Wu et al. 1997);
these pumps are a major consumer of presynaptic energy for generating the
electrochemical gradients that drive the AP, and hypoglycemia quickly results in partial
membrane depolarization, and altered AP waveform in peripheral nerve (Balfour et al.
2006; Stecker and Stevenson 2014). Similarly, glycolytic enzymes have been shown to
preferentially fuel SV refilling (Ikemoto et al. 2003; Ishida et al. 2009; Jang et al. 2016)
and fast axonal transport (Zala et al. 2013). Thus, the routes of ATP regeneration to
support presynaptic function may be segregated; however, the role of glycolysis-
produced ATP on maintenance of synaptic transmission in the intact CNS is largely
unknown.
The aim of this study was to dissect the role of mitochondrial versus glycolytic
production of ATP to support presynaptic function. We used the calyx of Held synapse as
a model for these experiments due to its experimental accessibility and well-described 130
physiology (Borst and van Hoeve 2012; Schneggenburger and Forsythe 2006). We used
pharmacological treatments that selectively inhibit presynaptic glycolysis or
mitochondrial OxPhos. Acute glycolysis inhibition altered the presynaptic AP waveform,
and resulted in attenuated synaptic transmission. Surprisingly, inhibition of mitochondrial
OxPhos had no effect on basal synaptic transmission within 30 min. Our data suggest that
specific routes of ATP production preferentially fuel ATPases to sustain synaptic
neurotransmission. Specifically, we propose that at the presynaptic terminal, glycolysis,
rather than oxidative respiration, provides the energy to maintain the ionic balance that
shapes the AP.
7.3 Methods
Animals
All animals in this study were used in accordance with animal welfare protocols approved by the University of Nevada, Reno. C57BL/6 mice (Charles River Labs)
postnatal day 8-10 of both sexes were used for this study. Data were acquired from 79
animals.
Preparation of acute brain stem slices
Brain stem slices were made as described previously (Renden and von Gersdorff
2007). Briefly, 8-10 day-old mice were euthanized via rapid decapitation, and the brain
removed from the skull and submerged in ice-cold slicing artificial cerebrospinal fluid
(ACSF) solution, containing the following (in mM): 85 NaCl, 2.5 KCl, 25 glucose, 25
NaHCO3, 1.25 NaH2PO4, 75 sucrose, 0.5 CaCl2, 7 MgCl2, 3 myo-inositol, 2 Na-pyruvate,
0.4 ascorbic acid; pH 7.3 when bubbled with carbogen gas (95% O2-5% CO2). Transverse 131
brain stem slices containing the medial nucleus of the trapezoid body (MNTB) were made at a thickness of 200 µm using a vibratome (VT 1200S, Leica Microsystems,
Oberkochen Germany). Slices were transferred to an incubation chamber containing normal ACSF bubbled with carbogen gas for 30-60 minutes at 35°C and maintained thereafter (up to 6 hours) at room temperature (~23°C) until used for recording. Normal
ACSF was composed of the following (in mM): 125 NaCl, 2.5 KCl, 25 glucose, 25
NaHCO3, 1.25 NaH2PO4, 2 CaCl2, 1 MgCl2, 3 myo-inositol, 2 Na-pyruvate, 0.4 ascorbic acid.
Electrophysiology
Slices were transferred to a recording chamber and perfused at ~2 mL/min with a normal ACSF solution bubbled with carbogen gas. All recordings were performed at room temperature. Slices were visualized using infrared gradient contrast (Dodt et al.
2002) with a 60x water-immersion objective (Olympus), and monitored with a CCD
camera (QIClick, QImaging; Surrey, BC Canada). Whole-cell patch clamp or current
clamp recordings were made using a HEKA EPC-10/2 amplifier controlled by
Patchmaster software (HEKA, Ludwigshafen/Rhein Germany). Data were low-pass
filtered at 2.9 kHz and digitized at sampling rates of 10 kHz. Pipettes were pulled from
thick-walled borosilicate capillary glass (1B200F-4; WPI, Sarasota FL) using a P-1000
pipette puller (Sutter Instruments, Novato CA). Electrophysiology data were analyzed
offline using custom-written routines in IGOR Pro (Wavemetrics, Lake Oswego OR).
Postsynaptic voltage-clamp recordings from Principal cells used a pipette solution
containing (in mM): 130 cesium gluconate, 10 CsCl, 5 sodium phosphocreatine, 10 132
HEPES, 5 EGTA, 10 TEA-Cl, 4 Mg-ATP, 0.5 GTP, 5 QX-314 adjusted to pH of 7.2 with
CsOH and 310-315 mOsm. Recording pipettes had 1.5-3 MΩ open tip resistance. Series
resistances for voltage clamped cells ranged from 2-8 MΩ, and were routinely
compensated to <0.5 MΩ. Cells were routinely held at -80 mV command voltage after
correcting for liquid junction potential. All EPSC recordings were further corrected for
residual Rs offline (Schneggenburger et al. 1999; Traynelis 1998). EPSCs were evoked by
placing a bipolar stimulating electrode near the midline and applying a biphasic voltage
waveform (100 µs duration, <5 V). All evoked EPSC recordings were induced by
applying stimulation ≥ 0.5 V over threshold. AMPA-mediated EPSC recordings were
performed in normal ACSF solution, to which we added 50 µM D-AP5 to block
ionotropic NMDA receptors, 0.5 µM strychnine to inhibit glycine receptors, and 10 µM
bicuculline to inhibit GABAA receptors.
Presynaptic AP waveforms were recorded in whole-cell current clamp mode
during bath perfusion of the tissue in normal ACSF with the addition of 0.5 µM
strychnine and 10 µM bicuculline. Presynaptic recording pipettes had open tip resistances
of 4-6 MΩ and were filled with a solution that contained (in mM): 97.5 potassium
gluconate, 32.5 KCl, 10 HEPES, 1 MgCl2, 0.5 EGTA adjusted to pH 7.4 with KOH and
osmolarity of 305-315 mOsm. Presynaptic calyx recordings included 0.25 mg/mL FITC- dextran in the recording pipette to verify the recording was in the presynaptic compartment. Action potentials were elicited via midline afferent fiber stimulation 0.1-
0.5 V above threshold. At least five AP waveforms delivered at ≤0.1 Hz stimulation were
averaged for each recording, and used for analysis. Action potential waveform was 133
isolated from stimulus artifact by subtracting a subthreshold response waveform.
Presynaptic resting potential was normally -75 to -80 mV, prior to liquid junction potential correction. All presynaptic recordings were made within 10 minutes after break-
in, to reduce ATP dialysis from the presynaptic terminal and rundown of activity (Kim et
al. 2007; Renden and von Gersdorff 2007).
Presynaptic Ca2+ currents were isolated by inclusion of 1 µM tetrodotoxin, 10
mM tetraethylammonium (TEA), and 300 µM 4-aminopyridine in the bath solution, to
+ + block Na and K channels, and by using a pipette solution that contained the following
(in mM): 130 cesium gluconate, 15 CsCl, 20 TEA-Cl, 10 HEPES, 0.2 EGTA adjusted to
pH 7.3 with CsOH, and osmolality of 310-315 mOsm. Presynaptic terminals were
clamped at -90 mV and Ca2+ currents were evoked by 1-10 ms voltage steps to varying test potentials. Linear leak and capacitive currents were removed using a P/5 pulse protocol. Recorded presynaptic AP waveforms from control, IAA-, and oligomycin- treated cells were used as voltage-command templates in HEKA Patchmaster to elicit
Ca2+ currents from presynaptic terminals using the same presynaptic pipette internal solution as above, with the addition of 4 mM ATP, 12 mM phosphocreatine, and 0.5 mM
GTP (Renden and von Gersdorff 2007).
Ca2+ imaging
Adeno-associated virus encoding the genetically encoded Ca2+ sensor GCaMP6m
(AV-1-PV2823, Penn Vector Core, University of Pennsylvania; (Chen et al. 2013) was injected into the ventral cochlear nucleus (VCN) at P1. After a 7-day incubation period,
GCaMP6 was selectively expressed at the presynaptic calyceal nerve terminal. Basal 134
Ca2+ was imaged at 0.1 Hz, in the absence of stimulation, using an EM-CCD camera
(Hamamatsu ImagEM X2; Bridgewater NJ). Midline stimulation (100 Hz, 500 ms) was
used to identify healthy infected terminals with low resting [Ca2+]free, and validated
axonal connectivity (see Figure 3). Images were analyzed offline using image analysis
software (Volocity 6.3; Perkin Elmer, Waltham MA). Analysis used the stimulated
response to define the ROI, and measured change in fluorescence normalized to the first
frame in quiescent condition (∆F/F0, %), after background subtraction.
Statistics
Statistical significance was determined by appropriate tests: Paired t-test, One-
way ANOVA, or Two-way ANOVA with post-hoc Holm-Sidak test, as indicated, using
Prism 5.0 software (GraphPad Software, La Jolla CA). Significance is expressed as: * = p
< 0.05, ** = p <0.01, and ***=p<0.001. Data are presented as mean ± SEM.
Modeling
A Hodgkin-Huxley type m2 model of the rat calyx pre-synaptic Ca2+ channel
(Borst and Sakmann 1998; Helmchen et al. 1997) was modified to match our Ca2+ current
recordings from mouse calyx. Specifically, the voltage dependence of the activation variable (m) was shifted by +12 mV, the reversal potential was changed from +43.9 mV to +56 mV, and the maximum conductance was increased from 48.9 nS to 50 nS. The
simulation was programmed in Igor Pro (Wavemetrics, Lake Oswego OR). Model
calculations were compared to the Ca2+ current I-V relationship measured during the last
1 ms of a 10-ms depolarizing voltage step from -70 to +40 in 10 mV steps. For this comparison, the membrane potential during measurements of the I-V relationship was 135
corrected for effects of residual series resistance as Vm′ = Vcmd – Rs·I, where Vm′ is the
corrected membrane potential, Vcmd is the voltage clamp command potential, Rs is the
residual (uncompensated) series resistance and I is the measured current. For model
comparisons with recorded Ca2+ current I-V relationships (Fig 4), measured ICa was
plotted versus corrected membrane potential.
We explored the impact of changes to the Na+ and K+ equilibrium potential on
AP waveform and resting potential using the Neuron simulation environment (Carnevale
and Hines 2006). Our strategy was not to attempt to model as closely as possible the fine
morphological and biophysical details of the calyx of Held. Rather, we looked for a
simple model that would allow us to make general conclusions. For this, we modeled a 1 mm axon of 4 µm diameter and 100 ohm-cm internal resistivity. We included standard
Hodgkin-Huxley Na+, K+, and leak currents, scaled to 24 ˚C using a Q10 of 3 for the
temperature dependence of gating kinetics. Simulated action potentials were generated by
current injection into the distal end of the axon, and action potentials were measured at
the proximal end. The values of ENa and EK were varied as described Figures and Text. In
some simulations, the leak conductance was varied together with EK to maintain the
resting potential at -65 mV.
7.4 Results
7.4.1 Presynaptic function at the calyx of Held relies on local ATP production
In order to dissect the mechanism underlying loss of AP versus synaptic transmission, we constrained ATP production selectively in the calyx of Held presynaptic terminal to either glycolysis or mitochondrial OxPhos, while recording from the 136
innervated postsynaptic principal cell. We relied on bath conditions that have previously
been used to selectively inhibit glycolytic ATP production (no added glucose plus 1 mM
Iodoacetic Acid; IAA), or mitochondrial OxPhos ATP production (no added pyruvate,
myo-inositol, or ascorbic acid; plus 1 µM Oligomycin). In the innervated postsynaptic
cell, ATP was maintained during recordings by including physiological [ATP] in the patch pipette (4 mM, see Methods). Thus, neuronal ATP depletion occurred selectively in the presynaptic terminal in this recording configuration. Stimulating at 0.1 Hz revealed
that inhibition of glycolysis alone strongly attenuated evoked EPSC maximum amplitude
by 36.18 ± 5.18% after 15 minutes (n=6, p=0.0065 versus baseline; Figure 7.1A).
Inhibition of mitochondrial OxPhos had no effect on EPSC amplitude (8.74 ± 9.13%
decrease; n=7, p=0.8284 versus baseline) when treated with oligomycin on a similar
same time scale (Figure 7.1B). Similarly, EPSCs recorded in control conditions were
unchanged on a similar timescale (-5.70 ± 3.67%, n=4, p=0.2008 versus baseline, data
not shown). Notably, inhibition of both glycolysis and mitochondrial OxPhos led to a
biphasic response: a transient increase in EPSC amplitude after approximately 15 minutes
(99.21 ± 39.44% increase; n=5, p=0.0612 versus baseline), followed by a complete loss
of EPSCs (Figure 7.1C). The effect and time course of IAA plus oligomycin on EPSC
size was variable, and not all cells responded with increased EPSC amplitude before loss
of transmission (cf. Figure 7.1Ciii); however, this treatment shows that blocking OxPhos
does impair synaptic function when glycolysis is absent. An increase in evoked
transmission is consistent with membrane depolarization and an increase in intracellular
Ca2+ (Awatramani et al. 2005). 137
One mechanism for attenuated EPSC size in IAA would be decreased SV loading with transmitter. Previous reports indicate that glycolytic enzymes are associated with synaptic transport vesicles (Zala et al. 2013), and changes in intravesicular pH, controlled by v-ATPase function, results in altered quantal size (Rost et al. 2015; Zhou et al. 2000).
Spontaneous excitatory postsynaptic currents (sEPSCs) were recorded to evaluate the effect of glycolysis or OxPhos inhibition on quantal size and kinetics (Figure 7.2).
Selective block of glycolysis or OxPhos did not affect sEPSC frequency or kinetics
(Figure 7.2B-D, and Table 7.1). Simultaneous application of both drugs increased sEPSC frequency significantly (Figure 7.2C), suggesting an increase in presynaptic Ca2+ levels consistent with previous reports (Lee and Kim 2015). However, we did not observe an effect of any treatments on sEPSC amplitudes (Figure 7.2B), indicating that the observed reduction in EPSCs amplitudes was not caused by reduced quantal size.
These data suggest that even during very low levels of synaptic activity, presynaptic ATP production from glycolysis, but not mitochondrial OxPhos, is acutely required to support normal synaptic transmission. Inhibition of ATP production by glycolysis quickly results in decreased transmission, and cannot be compensated for by mitochondrial OxPhos. The lack of effect of IAA on quantal size indicates that attenuation in transmitter release is not due to decreased transmitter loading of SVs.
7.4.2 Basal Ca2+ is not altered by inhibition of glycolysis or OxPhos
Using stereotaxic injection of AAV1-Syn-GCaMP6m into the ventral cochlear nucleus, we transduced globular bushy cells and the calyx of Held terminal with the Ca2+ sensor GCaMP6m (Chen et al. 2013). This viral transduction allowed us to visualize 138
basal free Ca2+ in the intact presynaptic terminal (Figure 7.3A). We observed no significant change in basal Ca2+ signal following IAA or oligomycin exposure, for up to
30 min (Control n=6, -1.25 ± 0.28% baseline; IAA n=7, 4.77 ± 1.58% baseline, p=0.4029 versus control; oligomycin n=9, -0.93 ± 0.67% baseline, p=0.9460 versus control; Figure
7.3B-C). Concurrent exposure to both IAA and oligomycin did result in a dramatic increase in free cytosolic Ca2+, similar to that reported by Lee and Kim (2015) using an
OGD model (IAA + Oligomycin n=7, 26.41 ± 6.49% baseline, p<0.0001 versus control;
Figure 7.3B-C). Taken together, these data suggest that EPSC attenuation due to IAA treatment is independent of basal Ca2+ levels. Additionally, we conclude that the increase in EPSC amplitude and sEPSC frequency due to inhibition of both glycolysis and mitochondrial OxPhos (Figure 7.1C and Figure 7.2C) is a direct result of increased cytosolic free Ca2+ in the presynaptic terminal.
7.4.3 Local glycolytic ATP is required for maintenance of presynaptic AP waveform
The shape of the presynaptic AP directly regulates Ca2+-influx and thus the amount of SV release at the calyx terminal (Ishikawa et al. 2003; Yang and Wang 2006); therefore, we sought to test whether EPSC inhibition by blocking glycolysis or mitochondrial OxPhos was due to altered presynaptic AP shape. We recorded the AP waveform at the calyx evoked by midline stimulation at 0.1 Hz, in control conditions, or in separate recordings after 10-30 min exposure to conditions that inhibited glycolysis or mitochondrial OxPhos, as above (Figure 7.4A). We noted that IAA treatment, but not oligomycin, depolarized the terminal and slowed the AP waveform (control n=8; IAA n=9; glucose-free n=4; oligomycin n=7, Figure 7.4B, Table 7.2). Resting membrane 139
potential was significantly reduced by inhibition of glycolysis (Figure 7.4Bi). AP amplitude, measured from resting potential, was also significantly reduced by glycolysis inhibition (Figure 7.4Bii). Additionally, the temporal aspects of the AP were slowed by glycolysis blockade: increased AP half-width, accompanied by slower rise and fall times, and an increase in AP delay (Figure 7.4Biii-4Biv). To assess specificity of IAA to selectively inhibit glycolytic ATP production, we also recorded APs in ACSF only lacking bath glucose, but without pharmacological inhibitors. Removing glucose from the bath solution resulted in similar changes in membrane potential, AP amplitude, and AP waveform as IAA (Figure 7.4B). Surprisingly, no differences in resting membrane potential, peak amplitude, AP shape, or onset were observed when mitochondrial function was inhibited with oligomycin (Figure 7.4Bi-Biiii). From these data, we conclude that both magnitude and kinetics of the AP waveform are significantly affected by loss of presynaptic ATP derived from glycolysis, but not mitochondrial OxPhos.
These data suggest that glycolysis-derived ATP is being constitutively – and preferentially – used to maintain the presynaptic membrane potential and AP waveform, and altered AP waveform contributes to the reduction in EPSC amplitudes observed under basal stimulation parameters in the absence of presynaptic glycolysis.
7.4.4 Ca2+-influx via VGCCs is altered by inhibition of glycolysis
We sought to test whether presynaptic Ca2+-influx by VGCCs was affected by presynaptic inhibition of glycolysis or mitochondrial OxPhos, as an additional potential mechanism underlying EPSC attenuation. We recorded presynaptic Ca2+ currents in control conditions and following 10-30 min exposure to IAA or oligomycin (Figure 140
7.5A). Presynaptic inhibition of glycolysis or mitochondrial OxPhos did not alter Ca2+
charge (QCa) or peak amplitudes (ICa) due to short AP-like depolarizations (1 ms to 0 mV;
QCa control: n=10, 0.59 ± 0.11 pC; IAA: n=8, 0.50 ± 0.09 pC, p=0.8039 versus control;
oligomycin: n=6, 0.58 ± 0.06 pC, p=0.9538 versus control. ICa control: 0.86 ± 0.18 nA;
IAA: 0.77 ± 0.14 nA, p=0.8950 versus control; oligo: 0.80 ± 0.10 nA, p=0.8950 versus
control; Figure 7.5Aii-Aiii). Although we did not include ATP, GTP or sodium
phosphocreatine in the presynaptic internal solution, we observed minimal rundown of
Ca2+ currents after 5-10 minutes following seal rupture into the whole-cell recording
configuration in all conditions tested at the immature calyx (data not shown; (Kim et al.
2007)). We also stimulated with longer (10 ms) steps, to test activation and current-
dependent inactivation of Ca2+-channels (Figure 7.5B; (Forsythe et al. 1998).
Surprisingly, we observed an increase in ICa at the end of 10 ms step depolarization in the
presence of IAA for depolarizations to potentials more positive than -10 mV (Figure
7.5Bii). When tail currents were normalized to maximum current and fit with a
Boltzmann function (Figure 7.5Biiii), they showed a rightward shift in activation kinetics
in IAA, relative to control (V50 control: -22.63 mV; IAA: -15.87 mV, p=0.0411 versus control; oligomycin: -25.19 mV, p=0.4249). The shift in activation curve, but not amplitude of tail currents, is partially consistent with decreased ATP (Weiler et al. 2014).
In contrast, inhibition of mitochondrial OxPhos by oligomycin had no effect on ICa
voltage dependence. Taken together, these data show that inhibition of glycolysis does
not dramatically affect Ca2+ activation, and does not support a role for altered Ca2+ 141
channel function as the mechanism underlying transmission attenuation due to inhibition
of glycolysis.
7.4.5 Glycolysis inhibition attenuates transmission due to altered AP waveform
It was not immediately clear how the slower and smaller AP waveform induced
by blocking glycolysis would affect subsequent Ca2+ channel activation, Ca2+-influx,
and therefore synaptic transmission. Broader AP results in greater Ca2+-influx at the
calyx, but reduced peak AP amplitude should reduce Ca2+-channel activation (Borst and
Sakmann 1999; Yang and Wang 2006); however, in hippocampus AP broadening results
in increased Ca2+ influx, and transmission potentiation (Geiger and Jonas 2000). We
modeled the expected Ca2+-current generated by the recorded AP waveforms using a
Hodgkin-Huxley type m2 model (see Methods, Figure 7.6). Simulated Ca2+-currents from IAA-treated APs were reduced in amplitude and charge, while oligomycin currents were unaffected (Figure 7.6C). This result indicates that attenuation of transmission due to loss of glycolysis may be due to a single, AP-mediated mechanism.
To experimentally verify whether the altered AP waveform seen in IAA would result in changes in Ca2+ current, we applied the same pre-recorded APs used in the
Ca2+ current simulations as voltage command waveforms in presynaptic calyx recordings, and measured resulting ICa (Figure 7.7). In these experiments, ATP, GTP and phosphocreatine were included in the patch pipette to circumvent any direct effect on
Ca2+-channel activity due to dialysis from the presynaptic terminal. AP command waveforms from control, IAA-treated, and oligomycin-treated terminals were applied to the same cell with 10 sec rest interval between command waveforms. The IAA waveform 142
resulted in significantly decreased peak Ca2+ current relative to control (ICa, control: 1.70
± 0.37 nA, n=7; IAA: 1.42 ± 0.29 nA, n=7; p=0.0195), while Ca2+ current due to
oligomycin-treated waveforms was similar to control (oligomycin: 1.75 ± 0.33 nA, n=7;
p=0.4039 versus control; Figure 7.7Bi). Total charge transfer was also significantly
reduced by IAA-treated AP waveform (QCa, control: 0.75 ± 0.18 pC; IAA: 0.71 ± 0.17
pC, p=0.0099 versus control), while oligomycin increased charge transfer (oligo: 0.90 ±
0.18 pC, p=0.0028 versus control; Figure 7.7Bii). This result provides experimental
validation of Ca2+ current simulations, and indicates that inhibition of glycolysis
decreases presynaptic Ca2+-influx via a slower, smaller AP.
7.4.6 Stoichiometric changes in intracellular concentration of Na/K contribute to altered AP waveform
During ATP deprivation, membrane-anchored ion pumps such as the Na+/K+
ATPase may fail. Under these conditions, we would expect to observe an accumulation of cytosolic [Na+] and loss of [K+] in the presynaptic terminal, which would change the
Nernst equilibrium potentials for these ions and affect their electrochemical driving
forces. Thus, we modeled the presynaptic AP waveform by stoichiometric disruption of
intracellular [Na+] and [K+] using a Hodgkin-Huxley model (see Methods), to see
whether ionic modifications could predict the altered AP waveforms we obtained
experimentally.
In our simulations, ENa was varied from its standard value of +50 mV down to
+18 mV, to mimic loading of the terminal with intracellular Na+, while EK was fixed (-77
mV), and resulted in a loss of peak amplitude accompanied by a delay in spike initiation 143
(Figure 7.8Ai). Conversely, EK was systematically varied from -77 mV up to -64 mV, to
mimic loss of cytosolic K+, with fixed ENa (+50 mV; Figure 7.8Bi). This change resulted in a similar loss of AP peak amplitude, but was accompanied by both membrane depolarization and a delay in spike initiation, similar to the results we observed for IAA treatment. However, if Vm was fixed to control values by current injection, the model
predicted a rescued AP timing and peak amplitude, even in the presence of reduced Ek
(Figure 7.8Bii).
The model described above predicts that when Na+ and K+ intracellular
concentration changes, reduction of AP amplitude can occur for two reasons: directly due
to loss of Na+ driving force, or indirectly due to loss of K+ driving force leading to
membrane depolarization and increased steady-state inactivation of Na+ current (see also
(He and Soderlund 2014). For the latter case (K+), injecting current in terminals affected
by IAA to restore Vm to control values would then rescue the AP waveform. We thus
sought to test whether we could rescue the AP shape in the presence of IAA by returning
resting membrane potential to control resting potential, approximately -87 mV (Figure
7.8C). Injecting current to bring resting membrane potential to control values partially rescued the increase in AP half-width due to IAA (Figure 7.8Cii); however, the changes in AP peak amplitude and AP delay persisted (Figure 7.8Ciii-Civ). Conversely, when terminals in control conditions were slightly depolarized to mimic the effect of IAA on resting membrane potential, AP half-width and AP delay were largely unaffected (Figure
7.8Cii, Civ), though AP peak amplitude was reduced (Figure 7.8Ciii). In all experiments performed in the presence of IAA, we observed exacerbated effects on AP waveform 144
compared to control at all resting membrane potentials. These simulated and
experimental data suggest that the altered presynaptic AP waveform that we obtained due
to IAA treatment, namely decreased membrane potential, and smaller and broader AP
waveform, is due to an additive effect of both reduced Na+ and K+ electrochemical gradients, and Na-channel steady-state inactivation. 145
Figure 7.1 Cellular ATP from glycolysis, but not mitochondrial OxPhos, is necessary to maintain basal neural activity at the prehearing calyx of Held.
EPSCs were recorded from the principal cells of the MNTB, evoked by 146
midline stimulation at 0.1 Hz. The last five EPSCs at times indicated
(control, 1; and drug treatment, 2) were used for analysis, and compared using pairwise t-test. (A) Inhibition of glycolysis with IAA plus glucose-free
ACSF attenuated EPSC size. (Ai) An example recording, where stable baseline was achieved before bath application of glycolysis inhibitor IAA, indicated by red line. Peak EPSC amplitude is shown. (Aii) Representative traces of control (black, time point 1) and OxPhos-only EPSCs (red, time point 2) from the same cell illustrate a loss of EPSC amplitude shortly after inhibition of glycolysis. (Aiii) Pairwise summary data of control EPSC amplitudes and the effect of glycolysis inhibition. (B) Inhibition of mitochondrial OxPhos with oligomycin and pyruvate-free ACSF did not affect EPSC size. Analysis was similar as in panel A. (Bi) An example recording of EPSC peak amplitude including stable baseline before bath application of mitochondrial OxPhos inhibitor oligomycin (blue line). (Bii)
Representative traces of EPSCs in control (black, time point 1) and glycolysis-only condition (blue, time point 2). (Biii) Pairwise summary of
EPSC amplitudes recorded in control, and after inhibition of OxPhos. (C)
Inhibition of both glycolysis and OxPhos resulted in transient increase in
EPSC size, followed by loss of transmission. Analysis was similar as in panel
A. (Ci) An example recording with stable baseline before bath application of
IAA plus oligomycin, indicated by green line. (Cii) Representative traces of the EPSCs in control (black, time point 1) and after inhibition of both 147
glycolysis and OxPhos (green, time point 2) show a transient increase in evoked transmission. Time point 3 (brown) represents complete EPSC failure. (Ciii) Pairwise summary of EPSC amplitudes recorded in control, and after inhibition of both glycolysis and OxPhos. Time point 2 in this case was determined by the maximum % increase in EPSC amplitude after application of both drugs, as the time course was inconsistent from cell to cell.
148
Figure 7.2 Quantal size and frequency is unaffected by presynaptic inhibition of glycolysis or mitochondrial OxPhos.
sEPSCs were recorded from the postsynaptic MNTB neuron before and after
a 30-minute treatment with drug. (A) sEPSCs were unaffected by treatment
with either IAA or oligomycin, however application of both compounds
increased sEPSC frequency. Representative traces of baseline (black) and
drug-treated (red, blue, green) sEPSCs are shown. (B) Pairwise summary
data comparing sEPSC amplitude before and after drug treatment. sEPSC
amplitude was unaffected in all conditions tested. (C) Pairwise summary 149
data of sEPSC frequency. sEPSC frequency was unaffected by treatment of either IAA or oligomycin alone, however application of both drugs in concert led to a significant increase in sEPSC events, by 2-way ANOVA. (D)
Pairwise summary data of sEPSC decay time constant ( ). The decay time constant ( ) was unaffected in all conditions tested.
150
Figure 7.3 Resting levels of presynaptic Ca2+ are unaffected by selective loss of glycolysis or mitochondrial OxPhos.
Genetically-encoded Ca2+ indicator (GCaMP6m) was expressed
selectively in the presynaptic terminal via AAV, and basal Ca2+ levels
were monitored in the selective absence of glycolysis, mitochondrial
OxPhos, or both modes of ATP production. (A) Representative images 151
of a presynaptic calyceal Ca2+ response during a stimulation event
(100 Hz, 500 ms; Ai), used to determine ROI of healthy calyx terminals. (Aii) ROIs used for basal Ca2+ measurement in the same slice, under quiescent (non-stimulation) conditions. (B) Time course of presynaptic Ca2+ measured during bath application of IAA (red), oligomycin (blue) or IAA plus oligomycin (green) over 30 minutes.
Loss of glycolysis or mitochondrial OxPhos independently did not alter presynaptic Ca2+ levels. However, inhibition of both metabolic pathways in concert resulted in a profound increase in presynaptic
Ca2+. (C) Summary data of Ca2+ levels after drug incubation for 15 minutes (Ci), and 30 minutes (Cii), shows an increase in presynaptic
Ca2+ only after inhibition of both glycolysis and mitochondrial
OxPhos.
152
Figure 7.4 Presynaptic AP waveform is inhibited by loss of glycolytic ATP.
AP waveforms were evoked by midline stimulation at 0.1 Hz, recorded in
whole-cell current clamp configuration from the calyx of Held presynaptic
terminal, with Ihold=0 for all cells. (A) Representative APs from separate
terminals in control conditions (black), when glycolysis was blocked with
IAA (red) or zero glucose (gray), and following loss of mitochondrial
OxPhos (blue). (B) Summary data of several AP parameters. (Bi) Presynaptic
resting membrane potential was depolarized in the presence of IAA or zero
glucose. Oligomycin treatment had no effect on Vm. (Bii) AP amplitude,
measured as difference between baseline and AP peak, was reduced in the 153
presence of IAA or zero glucose. Recordings in oligomycin were not different from controls. (Biii) AP width at one-half peak height was increased by IAA and zero glucose, but unaffected by oligomycin. (Biv) Action potential delay measured as time from stimulus to AP peak was also increased by IAA and zero glucose, but unaffected by oligomycin.
154
Figure 7.5 Presynaptic Ca2+ currents are altered after loss of glycolysis, but not mitochondrial OxPhos.
The presynaptic calyx of Held was voltage-clamped in whole-cell
configuration, held at -80 mV, and subjected to short depolarizations to 155
measure Ca2+ channel activation and resulting current (ICa). (Ai) Presynaptic voltage clamp command waveform, and example traces. Command voltage was stepped to 0 mV for 1 ms (top). Representative Ca2+ currents (bottom) in control conditions (black) and in separate cells following treatment with
IAA (red) or oligomycin (blue). (Aii) Peak ICa amplitude was not affected by
IAA or oligomycin, relative to control. (Aiii) Ca2+ current charge (QCa, integral of trace) was also unaffected. (B) Step-depolarizations (10 ms duration) were used to map Ca2+ current-voltage relationship. (Bi) The terminal was stepped in 10 mV increments from -80 mV holding potential
(top). Representative current families (bottom) in control conditions (black), or after pretreatment with IAA (red) or oligomycin (blue). (Bii) Current- voltage relationship using peak Ca2+ current at 10ms for control (black) and after pretreatment with IAA (red) or oligomycin (blue), respectively. IAA resulted in larger currents at steps ≥ 0 mV. (Biii) Tail currents from 10 ms depolarizations showed no difference from control recordings. (Biv)
Normalized tail currents fit by a Boltzmann function showed a rightward shift in activation due to IAA.
156
Figure 7.6 Modeling presynaptic Ca2+ current
Ca2+ currents were modeled using a Hodgkin-Huxley m2 model. (A)
Hodgkin-Huxley parameters for the modeling of presynaptic Ca2+ currents
(see Methods). (B) Recorded Ca2+-current (black symbols) and simulated
Ca2+-current (gray line) at matched command voltages, after series
resistance correction. Inset: example recorded (black) and simulated (gray)
currents due to a 10-ms step depolarization to 10 mV. Simulated currents are
faster than recordings because simulated output was not low-pass filtered.
Scale bars are 1 nA, 5 ms. (C) Representative recorded AP waveforms (top)
and the resulting Ca2+ current simulation (bottom) in control conditions
(black) and after treatment with IAA (red) or oligomycin (blue). Simulated
waveforms were filtered at 2.9 kHz, and aligned on the AP rising phase, for
clarity. Simulated Ca2+ current was reduced in both Qca and Ica in the 157
presence of IAA but unaffected by oligomycin treatment. 158
Figure 7.7 Ca2+ currents elicited by replaying recorded APs support IAA inhibition of AP-evoked ICa.
IAA-mediated AP waveform results in a smaller ICa, validating predictions
from the Ca2+ current simulation. (A) Previously recorded AP waveforms
were used as voltage-command templates (top). Resultant Ca2+ current
recordings are shown (bottom) induced by control (black), IAA (red) or
oligomycin (blue) waveforms, for an example recording. All three currents
were recorded from the same terminal. (B) Summary of effect of AP 159
waveform shape on Ca2+ current peak (ICa) amplitudes (Bi) and current charge (QCa; Bii). Both peak ICa and QCa were significantly reduced by IAA-
AP waveform.
160
Figure 7.8 Restoring resting membrane potential only partially rescues AP waveform in the absence of presynaptic glycolysis. 161
Combined simulation and experimental data suggest that Na and K driving force are altered by glycolysis inhibition. A and B: Simulations from a
Hodgkin-Huxley model used to predict the effect of altering Na+ reversal potential (ENa) and K+ reversal potential (EK) on AP waveform. (Ai) Varying
ENa in the presence of fixed EK (-77 mV) predicted a decrease in AP amplitude and increase in AP delay. (Aii) Fixing Vm at control values (-65 mV) did not further affect the AP waveform shape when ENa was varied. (Bi)
Varying EK in the presence of a fixed ENa (+50 mV) predicts reduced AP peak amplitude, as well as delay in AP spike initiation, and is accompanied by a loss of resting membrane potential as EK decreases. (Bii) Fixing Vm to -
65 mV largely rescued the AP waveform shape, even in the presence of altered EK. (C) Presynaptic calyceal APs were recorded in whole-cell current clamp configuration induced by midline stimulation at 0.1 Hz. Current injection was used to manipulate resting Vm and the resultant APs are shown in the presence and absence of IAA. (Ci) Representative control and IAA treated APs recorded at normal (black, red) and depolarized (gray, pink) resting membrane potentials. (Cii) AP half-width, (Ciii) AP peak amplitude, and (Civ) AP delay, plotted against Vm in the presence and absence of IAA.
Injecting current to restore Vm to normal value in the presence of IAA partially rescued the AP half-width. However, restoring Vm in the presence of
IAA did not rescue the AP peak amplitude nor the delay in spike initiation.
162
Table 7.1 Complete descriptive data of sEPSCs.
Descriptive data for multiple parameters of the sEPSC are shown before and
after treatment with IAA, oligomycin and IAA plus oligomycin. sEPSC
parameters were compared to the internal baseline values via two-way
ANOVA with post-hoc Holm-Sidak test, and the multiplicity-adjusted p-
value is shown. 163
Table 7.2 Complete descriptive data of APs.
Descriptive data for multiple parameters of the AP recorded in control
conditions, in the presence of IAA or zero glucose, and oligomycin. AP
parameters were compared to control using one-way ANOVA with post-hoc
Holm-Sidak test and the multiplicity adjusted P-value is displayed.
164
7.5 Discussion
This study examined the specific contribution of glycolytic- versus mitochondrially-derived ATP to support presynaptic function and neurotransmission. At the calyx of Held, we find that depletion of glucose, or pharmacological inhibition of glyceraldehyde 3-phosphate dehydrogenase with Iodoacetic Acid, significantly attenuated synaptic responses within minutes. Decreased transmission was due to a single mechanism: a slower, broader, and smaller presynaptic AP waveform. Replay of the altered AP command waveform was sufficient to reduce AP-mediated Ca2+-influx, and
inhibit transmission.
Inhibition of mitochondrial OxPhos via bath-applied oligomycin did not affect
synaptic transmission, and showed no effect on AP waveform or presynaptic Ca2+-
influx. These results indicate that glycolysis is acutely required to produce appropriately
shaped APs, and that ATP produced by this route cannot be fully compensated for by
mitochondrial OxPhos. In line with previous reports, we find that substrate depletion plus
pharmacological inhibition of either glycolysis or OxPhos has an effect on transmission
within minutes (Pathak et al. 2015; Rangaraju et al. 2014; Yamane et al. 2000).
Attenuation of the EPSC due to activation of inhibitory presynaptic receptors is not
likely, as Ca2+ currents were not affected (Barnes-Davies and Forsythe 1995; Kimura et
al. 2003). We did not observe hyperpolarization of the presynaptic membrane potential,
or an increase in outward current e.g., due to activation of an ATP-sensitive K+-
conductance, as reported for other neuronal preparations (Duchen 1990; Lee et al. 1995;
Spuler et al. 1988; Sun and Feng 2013; Trussell and Jackson 1987; Zhao et al. 1997). 165
This negative result indicates that these channels may not be robustly expressed at the
calyx presynaptic terminal.
A recent report at the calyx terminal showed that loss of both glycolysis and
mitochondrial OxPhos, mimicking ischemia (Lee and Kim 2015) resulted in intracellular
Na+ and Ca2+ accumulation. Our results differentiate these effects. We propose that
intracellular Na+ accumulation is due primarily to loss of glycolysis, because presynaptic
APs were unaffected by block of mitochondrial ATP production. Either glycolysis or
mitochondrial ATP production appear sufficient to maintain Ca2+ buffering because we
only observe an increase in intracellular Ca2+ when both pathways are inhibited.
7.5.1 Specific isolation of glycolysis versus mitochondrial oxidative phosphorylation
In this study, we suggest we are selectively inhibiting glycolysis or mitochondrial
OxPhos, and isolating these effects to the presynaptic terminal. We isolated the effect of
ATP blockade to the presynaptic terminal, as exogenous ATP and phosphocreatine were supplied to the postsynaptic neuron at physiological levels (4 and 12 mM, respectively) via the patch pipette internal solution, in postsynaptic recordings. We did not observe any effect of our manipulations on quantal size, indicating postsynaptic receptor availability was unaffected (Figure 7.2). Thus, we propose that the effects of inhibition were restricted to the presynaptic terminal. In recordings from the presynaptic terminal, ATP and phosphocreatine were usually absent, and we restricted recordings to <10 min, which has previously been shown to support ATP-dependent processes, at least transiently (Kim et al. 2007). Additionally, stimulation was delivered only at very low frequency, which should limit presynaptic ATP consumption due to SV recycling. 166
Pyruvate, a major substrate for mitochondrial OxPhos, is typically produced as a by-product of glycolysis, and may be depleted during glycolysis blockade, affecting mitochondrial OxPhos. However, several lines of evidence indicate substrate availability is not disrupted when blocking glycolysis with IAA or glucose deprivation. First, neurons express monocarboxylate transporters (MCT2) that allow uptake of pyruvate and lactate from the media, to power mitochondrial OxPhos (Pellerin et al. 2005; Tekkok et al.
2005). Accordingly, in control experiments and those where glycolysis was inhibited, pyruvate (2 mM) was included in the extracellular media. Multiple reports indicate lactate and pyruvate are acceptable substrates for oxidative respiration and can effectively support synaptic function (Bouzier-Sore et al. 2003; Izumi et al. 1994; Schurr et al.
1988). Several other non-glucose substrates can also be metabolized via the TCA cycle to produce energy in the CNS in addition to glucose and lactate/pyruvate (Lee do et al.
2013; Zielke et al. 2009), and are present in our slice preparation. Thus, we propose that exogenous substrates are sufficiently available for mitochondrial OxPhos in the absence of glycolysis, in our experiments. Second, when both IAA and oligomycin are applied together, the effect on transmission and Ca2+-buffering are different and more severe than when applied individually (Figures 7.1-3). The additive effect of glucose/oxygen deprivation has been shown previously as a complete loss of synaptic transmission (Ames and Gurian 1963; Hirsch et al. 1957; Schurr et al. 1988). If mitochondria were dependent on monocarboxylates generated via presynaptic glycolysis, we would expect to see a similar direction and extent of transmission attenuation due to IAA and IAA + oligomycin, which was not observed. This result suggests that mitochondrial OxPhos is 167
not strictly dependent on intact glycolysis in our preparation. Third, intrinsic fluorescence
imaging experiments did not show activity-dependent changes in NADH autofluorescence and flavoprotein signals when glycolysis was inhibited under the same
conditions used here, indicating mitochondrial OxPhos proceeds normally in the absence
of glycolysis, at least for short stimulation trains (Brennan et al. 2006; Duchen 1992).
7.5.2 Glycolysis fuels presynaptic APs
Inhibition of glycolysis resulted in a decrease in basal transmission, as a result of
smaller and slower AP waveform. In addition, resting Vm was significantly reduced.
These defects were not observed for mitochondrial OxPhos inhibition with oligomycin, and were independent of any preceding activity, indicating that glycolysis-mediated ATP production is primarily responsible for maintenance of membrane polarization in the presynaptic terminal. This study is the first attempt to specifically monitor presynaptic function in the absence of glycolysis, but with intact mitochondrial respiration. Our results stand in contrast with the tenet that mitochondrial respiration is the main energy source supporting neuronal function (Harris et al. 2012); however, many of the preceding studies examined CNS function indirectly, and did not have the spatial or temporal level of resolution attained here. Support for glycolysis to power presynaptic function via preferential support of membrane channels comes from fractionation studies in non- neuronal cells, which found glycolytic enzymes associated with plasma and vesicle membranes (Knull 1978; Lim et al. 1983; Mercer and Dunham 1981). Similarly, studies in synaptosomes indicate that loss of glycolysis results in slight (10-20 mV) depolarization of membrane potential (Hrynevich et al. 2015; Kauppinen and Nicholls 168
1986). Two recent reports validate this finding by showing that glycolysis, but not
mitochondrial respiration, are important for maintenance of cytosolic ATP in the
presynaptic terminal (Gazit et al. 2016; Rangaraju et al. 2014). In Drosophila, a temperature-sensitive mutation in phosphoglycerate kinase, another glycolytic enzyme, also exhibited reduced membrane potential in muscle (Wang et al. 2004b). Similarly, we observed depolarization of the presynaptic membrane potential when glycolysis was inhibited (Figure 7.4).
In neuronal brain slice preparations, glucose depletion (or 2-deoxyglucose treatment) has been shown to inhibit synaptic transmission within minutes. In neurons of
the dorsolateral septal nucleus, AP threshold was increased by glucose loss (Shoji 1992).
In a separate study, a period of glucose deprivation resulted in decreased excitatory
synaptic transmission, but did not show a significant decrease in total ATP (Yamane et al.
2000), suggesting localized ATP generation is important for synaptic function. Taken
together, we conclude that hypoglycemia results in loss of synaptic transmission
primarily due to an altered presynaptic AP waveform, and is acutely dependent on
constitutive Na+/K+ ATPase activity.
7.5.3 Altered AP shape reduces Ca2+ current, fully accounts for smaller EPSC
Bath perfusion of IAA resulted in a slower and smaller AP waveform. Neither
IAA nor oligomycin dramatically affected presynaptic ICa due to step depolarization (1-
10 ms). However, IAA-mediated AP waveforms produced a smaller Ca2+-current, shown
by channel simulation and experimental replay of the AP waveform in control terminals.
This reduction in ICa correlates with the observed reduction in EPSC size (Ishikawa et al. 169
2003). Taken together, these data support the hypothesis that IAA/glucose deprivation has a singular effect on presynaptic AP shape, which reduces Ca2-influx and transmitter release, and thus decreases EPSC size.
We propose that inhibition of the presynaptic Na+/K+ ATPase is a likely mechanism for altered AP shape. If Na+ and K+ distribution is disrupted, we expect decreased driving force for both ions (Lipton and Robacker 1983; Raffin et al. 1988; Wu et al. 1997). Simulated changes in Na+ and K+ driving force predict the changes we see in AP waveform due to glycolysis inhibition, and is confirmed by experimental recordings (Figure 7.8). The rapid effect of IAA is similar to the effect of Na+/K+
ATPase inhibition by ouabain (Kim et al. 2007), indicating Na+/K+ ATPases at the presynaptic terminal are constitutively, and perhaps selectively powered by glycolysis.
7.5.4 Presynaptic mitochondrial OxPhos is not required to maintain basal transmission
Mitochondrial OxPhos was suppressed by inhibition of complex V function with oligomycin (1 µM) and concomitant depletion of extracellular pyruvate; a standard method of inhibition, effective within minutes (Rangaraju et al. 2014). In this experimental protocol, glycolysis was left intact. Blocking mitochondrial OxPhos did not reduce EPSC size. We also observed no change in basal free Ca2+, shape of AP waveform, or Ca2+-currents. This result is consistent with those obtained at cerebellar
Purkinje cells, where Ca2+ clearance from soma and dendrites was dependent on glycolysis (Ivannikov et al. 2010). Oligomycin treatment has been shown previously not 170
to alter mitochondrial Ca2+ buffering (Billups and Forsythe 2002; David 1999; Talbot et
al. 2007), and has a specific inhibitory effect on ATP production, with no depolarization of mitochondrial potential (Nicholls and Budd 2000). Accordingly, we observed no
change in cytosolic free Ca2+ due to oligomycin exposure (Figure 7.3). Thus, our results
are in accordance with invertebrate recordings where mitochondrial function was
disrupted, with no effect on normal basal transmission (Guo et al. 2005; Verstreken et al.
2005), and in mammalian terminals in culture that lack mitochondria (Pathak et al. 2015;
Sun et al. 2013). We propose that mitochondrial OxPhos is not necessary to maintain
basal synaptic neurotransmission at the presynaptic terminal.
7.5.5 Physiological Relevance
Hypoglycemia results in rapid loss of cognitive processing in humans (Feldman
and Barshi 2007). We show here that loss of glycolysis results in a specific deficit in AP
shape and timing at the presynaptic terminal, and results in decreased transmission. This
shift in synaptic timing should have profound effects on timing-dependent circuits, e.g. in
the auditory system and cerebellum. Similarly, loss of spike-timing dependent plasticity
in cortical circuits may alter output in cognition, and attention, matching behavioral
observations in humans.
While terminals specialized for high-frequency firing show energy-efficient AP
production and maintenance (Alle et al. 2009), Na+/K+ ATPase activity is constantly
required for appropriate AP timing and shape, and is selectively powered by local
glycolysis. Loss of synaptic timing and decreased synaptic transmission in aglycemic
conditions occurs on the order of minutes, and is independent of synaptic activity. These 171
effects are prior to and independent of changes in excitation-secretion coupling, or receptor activity, and occur in the presence of intact mitochondrial OxPhos. Our identification of the relationship between neuronal glycolysis and synaptic transmission also supports the success of low glucose consumption (e.g. ketogenic diet) in treatment of neuronal hyperexcitability in epilepsy.
172
Chapter 8 The Developmental Profile of Presynaptic Energy Utilization during High
Frequency Neurotransmission
Brendan Lujan, Robert Renden 173
8.1 Summary
The role of energy generation is especially important in neurons, which are
energetically expensive, consuming up to 20% of an organism’s energy at rest. While a
majority of this energy is expended to regenerate electrical polarization of neurons, the
efficient release and recycling of neurotransmitter is also critically important to allow
chemical transmission between neuronal populations, and consumes nearly half of the
presynaptic neuronal energy budget. Mitochondria are the major suppliers of cellular
energy, generating ATP via oxidative phosphorylation. However, the specific utilization of energy from cytosolic (glycolytic) and/or mitochondrial respiration during synaptic neurotransmission is unknown. We use a model synapse with ideal synaptic properties for physiological investigation, the calyx of Held, to test the sources of energy utilized to support high-frequency neurotransmission. We show that inhibition of both glycolysis
and mitochondrial respiration influence the excitatory postsynaptic currents (EPSCs)
during high frequency activity at this mammalian synapse before the onset of hearing, at
P8-10. However, these effects dissipated using the same stimulation parameters (100 Hz,
200ms) in hearing mice, at P16-18. These data suggest a specific metabolic profile exists
to support high-frequency information transmission over the course of development.
Finally, ATP acts as a bottleneck to support high-frequency information transmission
when driving this synapse maximally (300 Hz, 150 ms). This study dissects both the
overlapping and non-overlapping use of presynaptic ATP derived from glycolysis and
mitochondrial respiration in support of high-frequency neurotransmission. 174
8.2 Introduction
Central nervous system (CNS) function is dependent on constant energy supply.
Loss of neuronal energetics has been shown to be a causative factor in a wide variety of cellular dysfunctions, as is the case with brain ischemia (Manzanero et al. 2013).
Additionally, chronic loss of neuronal energy production is widely believed to be a common underlying theme resulting in cell death in a multitude of neurodegenerative diseases, such as Parkinson’s and Alzheimer’s disease, and is a major component in cellular aging (Rodriguez et al. 2015). Interestingly, loss of synaptic function by denervation from postsynaptic targets often precedes the actual cell death in these diseases. Thus, synaptic activity may have a protective role in addition to information transmission. Maintenance of the presynaptic terminal, in particular, places large demand on ATP and recent study suggests that >64% of cellular ATP is used for synaptic transmission in the grey matter (Sengupta et al. 2010). Notably, ATPase activities at the presynaptic terminal are required for maintenance of the resting membrane potential by the Na+/K+ ATPase, which has been shown to directly regulate Na+/Ca+ exchanger
(NCX) function (Lee and Kim 2015). Similarly, ATP is reliably used in maintenance of presynaptic Ca2+ buffering by the plasma membrane calcium ATPases (PMCAs).
Additionally, synaptic vesicle (SV) loading of glutamate by the vesicular glutamate transporter provides another avenue of ATP use at the presynaptic terminal. Localization of mitochondria themselves to the presynaptic compartment relieves cytosolic burden for presynaptic ATP demand, however the complete energetic function of mitochondria at the presynaptic terminal remains unknown in vertebrate systems. Recent work indicates 175
that SV recycling is energetically expensive, and probably is a major source of ATP
consumption at the presynaptic terminal (Pathak et al. 2015; Rangaraju et al. 2014).
Although the presynaptic compartment clearly places large demand on ATP
production in maintenance of basic synaptic function, currently it is unclear which
mode(s) of energy production exist and predominate at the presynaptic terminal during ongoing activity. It has been suggested that mitochondrial OxPhos is the major producer
of ATP in brain (Harris et al. 2012), however increasing evidence supports discrete roles
for glycolytic ATP production in support of neuronal transmission (Bak et al. 2006; Bak
et al. 2009; Jang et al. 2016; Rangaraju et al. 2014).
We test explicitly two modes of ATP production to support presynaptic function.
We selectively inhibit glycolysis or mitochondrial OxPhos specifically at the presynaptic
terminal of the mouse calyx of Held, and examine their contribution to energy production
to support transmission during ongoing activity and following presynaptic depression of
the readily-releasable pool (RRP). We see differential yet overlapping utilization of
glycolysis and mitochondrial OxPhos due to activity. Further, we observe changes in the
reliance of transmission on glycolysis versus OxPhos during postnatal development, after
the onset of hearing. We presume that these changes are part of the synaptic refinement
underlying specialization for high frequency, high fidelity transmission endowed on this
auditory circuit.
8.3 Methods
Animals 176
All animals in this study were used in accordance with animal welfare protocols
approved by the University of Nevada, Reno. C57BL/6 mice (Charles River Labs)
postnatal day 8-10 and 16-18 of both sexes were used for this study.
Preparation of acute brain stem slices
Brain stem slices were made as described previously (Renden and von Gersdorff
2007). Briefly, mice were euthanized via rapid decapitation, and the brain removed from
the skull and submerged in ice-cold slicing artificial cerebrospinal fluid (ACSF) solution,
containing the following (in mM): 85 NaCl, 2.5 KCl, 25 glucose, 25 NaHCO3, 1.25
NaH2PO4, 75 sucrose, 0.5 CaCl2, 7 MgCl2, 3 myo-inositol, 2 Na-pyruvate, 0.4 ascorbic acid; pH 7.3 when bubbled with carbogen gas (95% O2-5% CO2). Transverse brain stem slices containing the medial nucleus of the trapezoid body (MNTB) were made at a thickness of 200 µm using a vibratome (VT 1200S, Leica Microsystems, Oberkochen
Germany). Slices were transferred to an incubation chamber containing normal ACSF bubbled with carbogen gas for 30-60 minutes at 35°C and maintained thereafter (up to 6 hours) at room temperature (~23°C) until used for recording. Normal ACSF was composed of the following (in mM): 125 NaCl, 2.5 KCl, 25 glucose, 25 NaHCO3, 1.25
NaH2PO4, 2 CaCl2, 1 MgCl2, 3 myo-inositol, 2 Na-pyruvate, 0.4 ascorbic acid.
Electrophysiology
Slices were transferred to a recording chamber and perfused at ~2 mL/min with a normal ACSF solution bubbled with carbogen gas. All recordings were performed at room temperature. Slices were visualized using infrared gradient contrast (Dodt et al.
2002) with a 60x water-immersion objective (Olympus), and monitored with a CCD 177
camera (QIClick, QImaging; Surrey, BC Canada). Whole-cell patch clamp recordings
were made using a HEKA EPC-10/2 amplifier controlled by Patchmaster software
(HEKA, Ludwigshafen/Rhein Germany). Data were low-pass filtered at 2.9 kHz and digitized at sampling rates of 10 kHz. Pipettes were pulled from thick-walled borosilicate capillary glass (1B200F-4; WPI, Sarasota FL) using a P-1000 pipette puller (Sutter
Instruments, Novato CA). Electrophysiology data were analyzed offline using custom- written routines in IGOR Pro (Wavemetrics, Lake Oswego OR).
Postsynaptic voltage-clamp recordings from Principal cells used a pipette solution containing (in mM): 130 cesium gluconate, 10 CsCl, 5 sodium phosphocreatine, 10
HEPES, 5 EGTA, 10 TEA-Cl, 4 Mg-ATP, 0.5 GTP, 5 QX-314 adjusted to pH of 7.2 with
CsOH and 310-315 mOsm. Recording pipettes had 1.5-3 MΩ open tip resistance. Series resistances for voltage clamped cells ranged from 2-8 MΩ, and were routinely
compensated to <0.5 MΩ. Cells were routinely held at -70 mV command voltage. All
EPSC recordings were corrected for residual Rs offline (Schneggenburger et al. 1999;
Traynelis 1998). EPSCs were evoked by placing a bipolar stimulating electrode near the midline and applying a biphasic voltage waveform (100 µs duration, <5 V). All evoked
EPSC recordings were induced by applying stimulation ≥ 0.5 V over threshold. AMPA- mediated EPSC recordings were performed in normal ACSF solution, to which we added
50 µM D-AP5 to block ionotropic NMDA receptors, 0.5 µM strychnine to inhibit glycine receptors, and 10 µM bicuculline to inhibit GABAA receptors.
Ca2+ imaging 178
Adeno-associated virus encoding the genetically encoded Ca2+ sensor GCaMP6m
(AV-1-PV2823, Penn Vector Core, University of Pennsylvania; (Chen et al. 2013) was
injected into the ventral cochlear nucleus (VCN) at P1. After a 7-day incubation period,
GCaMP6 was selectively expressed at the presynaptic calyceal nerve terminal. Ca2+ was
imaged at 10Hz, during presynaptic depression trains (100 Hz, 200 ms), using an EM-
CCD camera (Hamamatsu ImagEM X2; Bridgewater NJ). Midline stimulation (100 Hz,
500 ms) was used to identify infected terminals with low resting [Ca2+]free, and validated
axonal connectivity (see Figure 8.4). Images were analyzed offline using image analysis
software (Volocity 6.3; Perkin Elmer, Waltham MA). Analysis used the stimulated
response to define the ROI, and measured change in fluorescence normalized to the first
frame in quiescent condition (∆F/F0, %), after background subtraction.
Statistics
Statistical significance was determined by appropriate tests: Paired t-test, One-
way ANOVA, or Two-way ANOVA with post-hoc Holm-Sidak test, as indicated, using
Prism 5.0 software (GraphPad Software, La Jolla CA). Significance is expressed as: * = p
< 0.05, ** = p <0.01, and ***=p<0.001. Data are presented as mean ± SEM.
8.4 Results
8.4.1 Presynaptic depression causes postsynaptic desensitization
The mature rat calyx of held is able to reliably follow presynaptic spiking activity
at frequencies <600 Hz at physiological temperature (Taschenberger and von Gersdorff
2000). It has been suggested that postsynaptic receptor desensitization occurs in
prehearing rats, however this phenotype is largely abolished after the onset of hearing and 179
developmental maturation (Renden et al. 2005; Taschenberger et al. 2002). Thus, we
sought test if postsynaptic receptor desensitization occurs before and after the onset of
hearing at the mouse calyx of held using high-frequency stimulation (HFS). Presynaptic depression (100 Hz, 200 ms) was induced in P8-10 mice, and EPSCs were recorded from
the innervated postsynaptic principal cell in the presence and absence of kynurenic acid
(2 mM); (Figure 8.1A). Addition of kynurenic acid in the bath solution has previously
been shown to block postsynaptic receptor desensitization at the rat calyx of held (Wong
et al. 2003). In normalized responses, where the first three EPSCs were normalized to the
peak of the initial response, we observed loss of peak amplitude and increased falling
phase of the EPSC in the absence of kynurenic acid (Figure 8.1B). Similarly, normalized
depression trains showed a faster rate of depression when fit with a monoexponential
curve (Figure 8.1C-D). These data suggest that the prehearing mouse calyx of held
undergoes postsynaptic receptor desensitization during ongoing activity as reported
previously in rat before the onset of hearing (Renden et al. 2005; Taschenberger et al.
2002). Likewise, in mature mouse calyceal terminals, presynaptic depression trains (300
Hz, 150 ms) induced EPSCs and these responses were recorded in the presence and
absence of kynurenic acid (1 mM); (Figure 8.1E). Surprisingly, postsynaptic receptor desensitization persisted at developmentally mature terminals (Figure 8.1F-H). These data stand in contrast to previous reports from rat studies where desensitization did not persist in mature terminals (Renden et al. 2005; Taschenberger et al. 2002). We conclude postsynaptic receptor desensitization occurs at the mature mouse calyx both before and after the onset of hearing using our stimulation parameters. We thus included kynurenic 180
acid, 2 mM and 1 mM in P8-10 and P16-18 EPSC recordings, respectively, in all experiments presented hereafter in order to block postsynaptic receptor desensitization.
8.4.2 Presynaptic depression is regulated by ATP derived from both glycolysis and mitochondrial OxPhos at the developmentally immature calyx
Persistent synaptic activity leads to presynaptic depression at the calyx, due to
SV depletion of the readily-releasable pool (RRP); (von Gersdorff and Borst 2002) which stimulates ATP- and Ca2+-dependent pool refilling (Hosoi et al. 2007; Sakaba 2006). We sought to test the reliance of RRP refilling on glycolysis versus OxPhos, during ongoing synaptic activity. We depleted the RRP with a HFS train (100 Hz, 200 ms), a stimulus known to deplete the RRP of SVs at this age (Sakaba and Neher 2001; Schneggenburger et al. 1999). We recorded the resulting EPSCs in conditions that inhibited glycolysis
(IAA), or OxPhos (oligomycin), and these data were compared to control conditions
(control n=11; IAA n=10; oligomycin n=11; Figure 8.2A). Interestingly, when we
compared initial EPSC amplitudes of those cells that were unconditioned and
preconditioned to HFS, we found substantial increases of the initial EPSC in the presence
of oligomycin (% change of unconditioned vs HFS preconditioned initial EPSC
amplitudes: control: 16.39 ± 9.66%, p=0.1723 versus baseline; IAA: 14.95 ± 11.92%,
p=0.3480 versus baseline; oligomycin: 51.17 ± 13.15%, p=0.0115 versus baseline, Figure
8.2B). These data suggest that mitochondrial Ox Phos function regulates EPSC size in an
activity-dependent fashion. We thus analyzed, in all conditions tested, only those cells
that were preconditioned to HFS. Multiple depression trains from the same cell were
averaged and we observed a sharp increase in the maximum amplitude of the 181
preconditioned initial EPSC in the presence of oligomycin (Control: 0.481 ± 0.083 nA;
IAA: 0.309 ± 0.050 nA, p=0.2367 versus control; Oligomycin: 0.814 ± 0.137 nA,
p=0.0450 versus control; Figure 8.2C). In order to evaluate the speed and extent of
presynaptic depression independent of EPSC size, the initial EPSC amplitude of the train
was normalized, and plotted as a function of time (Figure 8.2D). Depression could be
described by a monoexponential curve, where time constant (τ) represents the kinetics of
depression (Taschenberger and von Gersdorff 2000). To our surprise, depression
occurred significantly slower in the absence of glycolysis, while the time constant of
depression was relatively unaffected by blocking OxPhos (Control: 23.46 ± 2.19 ms;
IAA: 35.45 ± 3.24 ms, p=0.0043 versus control; oligomycin: 20.35 ± 2.03 ms, p=0.3788
versus control; Figure 8.2D-E). This analysis also illustrates that the first EPSC in IAA
results in significant facilitation of the second response (Figure 8.2A-D). Consequently,
the paired-pulse ratio (PPR), assumed to be inversely related to release probability, is
significantly increased by IAA, whereas PPR in oligomycin was unaffected (Control:
1.16 ± 0.05; IAA: 1.70 ± 0.12, p=0.0021 versus control; oligomycin: 0.90 ± 0.13,
p=0.0814 versus control; Figure 8.2F). The steady-state response, measured at the end of
the stimulation train, is assumed to represent a balance of SV depletion due to release and
activity-dependent refilling of release sites. Normalized steady-state responses were not affected by either IAA or oligomycin treatment (Measured as % of initial EPSC; control:
21.55 ± 1.95%; IAA: 24.83 ± 3.11%, p=0.7546 versus control; oligomycin: 20.27 ±
4.58%, p=0.7899 versus control; Figure 8.2G). Cumulative EPSC peak amplitude plots
from these stimulation trains can be used to estimate the RRP, given known caveats 182
(Neher 2015). Using this method, we found that apparent RRP size is unaffected by either
IAA or oligomycin (Normalized RRP control: 1.03 ± 0.13 nA; IAA: 1.19 ± 0.12 nA, p=0.4667 versus control; oligomycin: 1.37 ± 0.18 nA, p=0.2112 versus control; Figure
8.2H-I). The initial release probability (Pr) can also be estimated from this analysis, as the fractional contribution of the first response to the RRP estimate. Pr was decreased by blocking glycolysis, and increased by blocking OxPhos. (Pr control: 0.281 ± 0.018; IAA:
0.164 ± 0.016, p=0.0393 versus control; oligomycin; 0.378 ± 0.016, p=0.0340 versus control), consistent with the PPR data (Figure 8.2J).
Taken together and using orthogonal methods, these data indicate that blocking presynaptic glycolysis acutely impairs Pr, slowing SV depletion during a train. This is likely explained as a downstream result of slower and reduced AP waveform (cf. Chapter
6). Conversely, blocking OxPhos increases Pr in an activity-dependent manner, resulting in an increase in EPSC size, without affecting depression kinetics.
8.4.3 OxPhos supports presynaptic Ca2+ buffering during HFS in immature terminal
Although Ca2+ influx for a single AP is small, presynaptic Ca2+ loading during ongoing activity is substantially higher and mitochondria may pay a role in presynaptic
Ca2+ buffering (Billups and Forsythe 2002). We thus sought to test the role of glycolysis and mitochondrial OxPhos in the maintenance of presynaptic Ca2+ after exposure to
HFS. One indirect method to assay intraterminal Ca2+ increases is to measure the frequency of spontaneous excitatory postsynaptic currents (sEPSCs) which are increased during presynaptic Ca2+ loading. We recorded sEPSCs before and after high-frequency 183
activity in the presence and absence of IAA or oligomycin (Figure 8.3A). Basal sEPSC
frequency was unaffected in the presence of either drug alone, however, sEPSC
frequency was selectively increased after HFS in the presence of oligomycin (Figure
8.3B-C). Conversely, sEPSC amplitude was unaffected both before and after HFS in all conditions tested, suggesting no change in SV quantal content nor postsynaptic receptor surface expression. These data suggest presynaptic Ca2+ buffering may be compromised
in the presence of oligomycin after HFS.
To monitor presynaptic Ca2+ levels empirically, the genetically encoded Ca2+ sensor GCaMP6m was transduced into the VCN (Chen et al. 2013). After a 7 day incubation period, the GCaMP6m Ca2+ sensor was selectively expressed at the presynaptic nerve terminal (control: n=3: IAA: n=7; oligomycin: n=4; Figure 8.4A). We observed increased levels of Ca2+ due to HFS at the calyx after exposure to oligomycin while no changes were observed in the presence of IAA compared to control (maximum
% change control: 4.97 ± 1.08%; IAA: 3.32 ± 0.82%, p=0.3505; oligomycin: 10.81 ±
1.59%, p=0.0189; Figure 8.4B-C). From both the sEPSC and GCaMP6m experiments,
we conclude Ca2+ buffering is hindered with the loss of mitochondrial OxPhos after
ongoing synaptic activity at the P8-10 calyx of held.
8.4.4 Presynaptic energy deficits do not affect recovery after synaptic depression at the developmentally immature calyx
We next sought to test the role of glycolytic- versus OxPhos-produced ATP
during presynaptic recovery from depression. Previous work has identified Ca2+- and
ATP-dependent mechanisms for recovery from presynaptic depression at the calyx of 184
Held (Sakaba 2006; Wang and Kaczmarek 1998). We applied a HFS trains (100 Hz, 200
ms), as above to deplete the presynaptic terminal RRP, and subsequently applied test trains of the same frequency and length after various rest periods (20 ms - 13 sec), to measure RRP recovery (control n=11; IAA n=10; oligomycin n=11; Figure 8.5A). The size of the RRP of the conditioning test train was unaffected throughout the duration of the protocol (Figure 8.5B); summary data for RRP recovery from synaptic depression are shown in Figure 8.5C. Recovery curves could be adequately fit with a monoexponential function. Time constants (τ) describing the rate of RRP recovery were not affected by inhibition of either glycolysis or OxPhos (control: 18.84 ± 6.84 sec; IAA: 24.00 ± 9.07 sec, p=0.8527 versus control; oligomycin: 17.69 ± 5.39 sec, p=0.8527 versus control;
Figure 8.5D). Alternatively, we measured the recovery of single EPSC peak amplitudes following the conditioning depression train, to determine functional recovery of synaptic transmission, which should be predominantly affected by changes in Pr. We observed strong potentiation of the initial EPSC peak amplitudes of the conditioning stimulus train after oligomycin treatment (Figure 8.5E), similar to that noted for HFS preconditioned initial EPSC amplitudes in Figure 8.2C. However, initial EPSC recovery from synaptic depression was unaffected by glycolysis or OxPhos inhibitors (Figure 8.5F). Single EPSC recovery curves could also be well fit by a monoexponential time constant (τ), as previously described (control: 7.07 ± 1.35 sec; IAA: 7.38 ± 2.01 sec, p=0.9624 versus
control; oligomycin: 6.56 ± 0.96 sec, p=0.9624 versus control; Figure 8.5F-G) (Wang and
Kaczmarek 1998). From these data, we presume that in the developmentally immature
calyx, ATP derived from either glycolysis or OxPhos is used to support recovery 185
following synaptic depression, but ATP utilized in this process is not dependent on a
specific mode of production.
8.4.5 Developmental changes affect presynaptic depression in hearing mice after
inhibition of ATP-production
The calyx undergoes robust morphological, molecular and functional changes
between the ages of P8-10 and P16-18 to support fast, reliable neurotransmission crucial
for proper sound localization (Schneggenburger and Rosenmund 2015; Taschenberger et
al. 2002; von Gersdorff and Borst 2002). Since many changes occur in the mature
synapse to support high frequency, high fidelity synaptic activity, we queried whether
there is also a shift in ATP dependence in the mature calyx of Held synapse, in order to
support synaptic transmission. We tested for developmental changes in the activity-
dependent role of glycolysis- or OxPhos-derived ATP between prehearing (P8-10,
above), and hearing mice (P16-18). We measured synaptic depression in hearing mice at
P16-18 by driving synaptic transmission at the same frequency as in P8-10 mice (100 Hz,
200 ms; Figure 8.6) and assaying EPSC peak amplitudes from the principal cells of the
MNTB after inhibition of presynaptic ATP-producing pathways (Control: n=9; IAA:
n=4; oligomycin: n=4; Figure 8.6A). Peak EPSC amplitudes were plotted as a function of
time in control, during glycolysis blockade, or blockade of OxPhos. Interestingly, the
initial EPSC peak amplitudes were unaffected in these older animals after preconditioning to HFS (control 0.817 ± 0.229 nA; IAA 0.770 ± 0.215 nA, p=0.9872; oligomycin 0.897 ± 0.221 nA, p=0.9642; Figure 8.6B-C). Notably, we did not observe an increase in the initial EPSC peak amplitude of the depressing train in an activity- 186
dependent fashion after treatment with oligomycin, as observed in P8-10 mice (cf. Figure
8.2C and Figure 8.6C). These data suggest a developmental loss of SV regulation by
OxPhos ATP after the onset of hearing. Peak EPSC amplitudes were normalized to that of the first in the depression train, in an effort to describe the kinetics of depression
(Figure 8.6D). Contrary to depression trains recorded in prehearing mice, we found that synaptic depression was unaffected by loss of glycolysis or mitochondrial OxPhos by the same HFS (time constant (τ); Control: 40.17 ± 3.65 ms; IAA: 52.83 ± 21.63 ms, p=0.5366 versus control; oligomycin: 32.60 ± 1.31 ms, p=0.7893 versus control; cf.
Figure 8.2E and Figure 8.6E). The PPR was also unaffected by either treatment (control:
0.980 ± 0.040; IAA: 0.941 ± 0.053, p=0.8100; oligomycin: 0.984 ± 0.064, p=0.9971;
Figure 8.6F). Steady-state response, measured at the plateau phase of the depression train, was unaffected in all conditions tested (Measured as % initial EPSC; control: 22.23 ±
2.87%; IAA: 18.63 ± 5.64%, p=0.7804 versus control; oligomycin 22.13 ± 6.151%, p=0.9997 versus control; Figure 8.6G). Cumulative EPSC plots reveal that both estimated
RRP size and Pr are unaffected by loss of glycolysis or OxPhos (Normalized RRP control: 1.00 ± 0.15; IAA: 1.18 ± 0.20, p=0.7307; oligomycin: 1.08 ± 0.26, p=0.9259; Pr control: 0.253 ± 0.013; IAA: 0.257 ± 0.029, p=0.3802 versus control; oligomycin: 0.272
± 0.006, p=0.6678 versus control; Figure 8.6H-J).
Taken together, our data support a significant increase in energy efficiency to support high-frequency activity at the post-hearing calyx of Held. These data suggest developmental compensation of cellular energy production to support high frequency neurotransmission at the mature calyx. We conclude that glycolysis and OxPhos are able 187
to fully compensate for selective loss of the other route of ATP production and a possible
role for increased energy efficiency.
8.4.6 Recovery from synaptic depression is not affected by loss of presynaptic ATP
in post hearing calyx of Held synapses
We next looked at recovery from synaptic depression at the mature calyx at P16-
18, using the same recovery from synaptic depression protocol used to assay recovery in
prehearing P8-10 mice (i.e., pairs of 100 Hz, 200ms trains, Figure 8.7A). We found that
the initial RRP size was unaffected in the presence of IAA or oligomycin (control: n=9;
IAA: n=4; oligomycin: n=4; data not shown). Similar to what we observed at P8-10, RRP
recovery was unaffected by treatment with IAA or oligomycin, when the recovery curves
were fit with a monoexponential function (time constant (τ); control: 3.93 ± 1.04 sec;
IAA: 5.53 ± 2.53 sec, p=0.6481 versus control; oligomycin: 1.49 ± 0.25 sec, p=0.3925
versus control; Figure 8.7 B-C). Similarly, functional recovery of synaptic
neurotransmission was assessed by assaying initial EPSC recovery. Notably, we did not
see activity-dependent potentiation of the initial EPSC at the developmentally mature
calyx due to loss of OxPhos (data not shown), however we did observe changes in initial
EPSC recovery due to loss of glycolysis, but not OxPhos (time constant (τ); control: 5.16
± 0.93 sec; IAA: 10.24 ± 2.91 sec, p= 0.0491 versus control; oligomycin: 3.64 ± 0.83 sec,
p=0.4116 versus control; Figure 8.7 D-E). We conclude that cellular energy required for
recovery from synaptic depression has a stronger dependence on ATP derived from
glycolysis, and may not be compensated for by the presence of intact mitochondrial
OxPhos, in contrast to P8-10 animals. 188
8.4.7 ATP is required for high-frequency activity at the developmentally mature calyx
To test whether presynaptic energy utilization is affected at stimulation frequencies at the upper boundary for transmission at the mature calyx at room temperature, we drove brief high-frequency trains of stimulation at 300 Hz (150 ms), a stimulus frequency shown to completely deplete the RRP of SVs at this age, which strongly depresses control synapses to ~10% of initial EPSC (control n=10; IAA n=8; oligomycin n=7; Figure 8.8A); (Mahfooz et al. 2016). We observed an increase in the
HFS preconditioned % change in the presence of oligomycin (% change of unconditioned vs HFS preconditioned initial EPSC amplitudes: control: -38.74 ± 15.65%; IAA: -14.05 ±
5.40%, p=0.0914 versus control; oligomycin: -3.84 ± 4.91%, p=0.0300 versus control;
Figure 8.8B). However, the EPSC peak amplitude was unaffected by either IAA or oligomycin (control 1.312 ± 0.267 nA; IAA 1.593 ± 0.178 nA, p=0.6152; oligomycin
2.033 ± 0.2549 nA, p=0.0889; Figure 8.8C). Normalized EPSC peak amplitudes were plotted as a function of time in control and in the presence of IAA or oligomycin (Figure
8.8D). Normalized depression plots show that depression kinetics were substantially increased by either IAA or oligomycin. This result is contrary to what was observed in
P8-10 and P16-18 animals at 100 Hz (time constant (τ); control: 25.57 ± 3.86 ms; IAA:
14.94 ± 1.61 ms, p=0.0286 versus control; oligomycin: 14.88 ± 1.67, p=0.0343 versus control; Figure 8.8 D-E). The PPR was relatively unaffected in the presence of IAA in contrast to the facilitation observed at P8-10, but significantly reduced by oligomycin
(control: 1.18 ± 0.06; IAA: 1.09 ± 0.05, p=0.6000 versus control; oligomycin: 0.92 ± 189
0.10, p=0.0413 versus control; Figure 8.8F). Steady-state EPSC responses at the end of the 300 Hz train were also suppressed in older terminals by IAA or oligomycin
(measured as % of initial EPSC: control 10.91 ± 1.93%; IAA 4.72 ± 1.25%, p=0.0177
versus control; oligomycin: 5.22 ± 0.98%, p=0.0365 versus control; Figure 8.8G). These
results indicate that, contrary to 100 Hz stimulation, both glycolysis and OxPhos are
required to support high-frequency transmission at 300 Hz, and inhibition of either
pathway likely results in an energy deficit, significantly impacting the ability of the
synapse to faithfully transmit information. The deficit may be caused by decreased RRP
size, or loss of activity-dependent acceleration of RRP refilling.
Cumulative EPSC plots were used to estimate the RRP, and show that apparent
RRP size is unaffected by IAA or oligomycin (Normalized RRP control: 0.9992 ±
0.1636; IAA; 0.9075 ± 0.1510, p=0.8706 versus control; oligomycin: 1.018 ± 0.1177,
p=0.9943 versus control; Figure 8.8H-I). However, Pr was increased after treatment with
oligomycin, consistent with decreased PPR (control: 0.113 ± 0.018; IAA: 0.154 ± 0.016,
p=0.2355 versus control; oligomycin: 0.192 ± 0.023, p=0.0174 versus control; Figure
8.8J). This finding indicates that OxPhos may have a designated role to suppress SV
release during repetitive transmission, early in a high-frequency train.
Taken together, these data support the concept that ATP supply is a critical
bottleneck to support high-frequency transmission at the mature calyx. Further, these data
suggest that OxPhos ATP plays a critical role early in a stimulus train, but that both
OxPhos and glycolysis support maintenance of high-frequency synaptic
neurotransmission, by acceleration of RRP refilling during activity. The exacerbated 190
depression phenotype following inhibition of mitochondrial OxPhos only in P16-18 mice may be due in part to an additional dependence on mitochondrial ATP, via mitochondrial adherence-associated complexes (MACs), a morphological feature present in mature terminals that has been hypothesized to aid in fine-tuning fast neurotransmission.
8.4.8 Recovery from synaptic depression at the developmentally mature calyx
We next asked if recovery from synaptic depression is regulated by ATP derived specifically from cytosolic glycolysis or OxPhos in mitochondria at the developmentally mature calyx. We drove the stimulus trains at 300 Hz (150 ms) to maximally deplete the presynaptic terminal (Figure 8.9A). Fractional RRP recovery was plotted over time, as above (Control n=8; IAA n=5; oligomycin n=7; Figure 8.9B). The recovery curves were best fit with a double-exponential function, consistent with previous reports (Chen et al.
2015; Mahfooz et al. 2016). The initial RRP size of the conditioning train remained unaffected in all conditions tested (data not shown). The slow phase of RRP recovery was not affected by IAA, but was significantly impaired by oligomycin treatment (time constant (τS) control: 3.041 ± 0.466 sec; IAA: 4.258 ± 0.283, p=0.1307 versus control; oligomycin: 4.526 ± 0.428, p=0.0376 versus control; Figure 8.9D). The other components of recovery — fast phase of RRP recovery, and % contribution — remained unaffected
(time constant (τF) control: 0.1486 ± 0.0163 ms; IAA: 0.1905 ± 0.0164 ms, p=0.1705 versus control; oligomycin: 0.1927 ± 0.0154 ms, p=0.1040 versus control; %K fast control: 39.75 ± 1.45; IAA: 44.59 ± 1.35, p=0.6252 versus control; oligomycin: 45.01 ±
6.21, p=5196 versus control; Figure 8.9C, E). These data suggest that OxPhos-derived
ATP at the mature presynaptic terminal may play a more dedicated role in the slow phase 191
of RRP recovery, which cannot be compensated for by glycolytic ATP production. The
fast mode of RRP recovery in the mature calyx terminal may be dependent on ATP, as previously indicated, but shows no specificity for production mode, and loss of one mode can be compensated by the other.
From the same data, initial EPSC recovery was measured when IAA or oligomycin were present (Figure 8.9F). EPSC Recovery was not affected by the loss of
ATP from either glycolysis or OxPhos, similar to the results seen from younger animals
(time constant (τF) control: 0.1651 ± 0.0972 ms; IAA: 0.07123 ± 0.0049 ms, p=0.5110
versus control; oligomycin: 0.1236 ± 0.0129 ms, p=0.8430 versus control; time constant
(τS) control: 7.112 ± 2.520 sec; IAA: 6.585 ± 1.339 sec, p=0.9705 versus control; oligomycin: 4.998 ± 0.567 sec, p=0.5870 versus control; %K fast control: 31.30 ± 8.56%;
IAA: 25.62 ± 2.03%, p=0.7622 versus control; oligomycin: 32.18 ± 4.03%, p=0.9918
versus control; Figure 8.9 G-I). We propose that ATP production is largely compensated
when one pathway is inhibited to support initial EPSC recovery, during rest periods
following a bout of synaptic activity.
192
Figure 8.1 Postsynaptic AMPA receptor desensitization during high- frequency activity in both prehearing and hearing mice 193
Midline stimulation was used to drive high-frequency synaptic activity and the resultant EPSCs were recorded from the principal cells of the MNTB in the presence and absence of kynurenic acid in both prehearing and hearing mice. (A) Representative depression trains in P8-10 mice (100 Hz, 200 ms) in the presence of kynurenic acid (black) and absence (gray). (B) EPSCs normalized to the first response show significant desensitization in the absence of kynurenic acid (control: gray; kynurenic: black). Note loss of peak amplitude and increased falling period in control (gray) trace. (C) EPSC peak amplitudes were normalized to that of the first response and plotted as a function of time to give rise to depression curves. Inclusion of 2 mM kynurenic acid (black trace), alleviated the speed of depression and steady- state responses. (D) Depression curves were well fit with a monoexponential curve. The decay time constant (τ) was faster in control recordings. (E)
Representative depression traces recorded from P16-18 mice (300 Hz, 150 ms) also resulted in significant depression, and also contained desensitization, revealed by inclusion of 1 mM kynurenic acid. (F) The control normalized EPSCs resulted in loss of peak amplitude faster than those recordings in which 1 mM kynurenic acid was included. (G)
Normalized depression curve and (H) decay time constant (τ) further shows desensitization occurs at the hearing calyx. 194
Figure 8.2 High-frequency synaptic neurotransmission is differentially modulated by ATP source at the developing calyx of Held.
Midline stimulation was used to drive high-frequency synaptic transmission
in efforts to empty the RRP at the prehearing calyx of Held (P8-10). (A)
Representative traces of EPSCs recorded from MNTB principal cells during
a 100 Hz, 200 ms stimulus train in control conditions (black), in the presence
of IAA (red) or oligomycin (blue). Kynurenic acid (2 mM) was included in 195
all bath solutions to block postsynaptic AMPA receptor desensitization. (B)
Cells were preconditioned to high frequency activity (20 trains delivered at
100 Hz with varying interstimulus intervals) revealed peak EPSC amplitude is potentiated in the presence of oligomycin. (C) The mean initial EPSC amplitude is shown for control conditions (black), IAA (red), or oligomycin
(blue) after preconditioning. Initial EPSC amplitude was increased in the presence of oligomycin. (D) Peak EPSC amplitudes were normalized to the first response, per cell, and plotted versus time. Data from the 2nd to 20th stimulus were fit by a single exponential decay function in all conditions
(dotted lines). (E) Summary of decay rates from fits to individual cells.
Presynaptic depression is significantly slowed by IAA. (F) The mean paired- pulse ratio (PPR) from individual cells, calculated as amplitude
EPSC2/EPSC1, at 10 ms intervals. PPR is significantly higher in the presence of IAA, but was unaffected by oligomycin. (G) Steady-state of depression was calculated as the average EPSC maximum amplitude at the end of the train, and was unaffected by IAA or oligomycin. (H) Summary cumulative
EPSC amplitude plot, shown as a function of stimulus number. (I) The linear portion of the cumulative EPSC plot (stimuli 10-20) was back-extrapolated to the Y-axis, providing an estimate of the apparent RRP size; normalized RRP was unaffected in all conditions tested. (J) Vesicle release probability (Pr), calculated as the first EPSC amplitude divided by apparent RRP size
(estimated in panel I), was differentially affected by IAA and oligomycin. 196
Figure 8.3 OxPhos supports presynaptic Ca2+ buffering at the prehearing
Calyx.
mEPSCs were recorded before and after a high frequency stimulus train (100
Hz, 200 ms) in P8-10 mice. (A) Representative traces of mEPSCs recorded
before (left) and after (right) a stimulus train are shown in control conditions
(black) or after treatment with IAA (red) or oligomycin (blue). (B) The pre- 197
train mEPSC frequency was unaffected in all conditions tested. (C)
Normalized frequency is plotted as a function of time and binned into 30 second intervals. The stimulus train was delivered at time=0. Normalized frequency was increased following presynaptic depression in the presence of oligomycin. (D) Pre-train amplitude was unaffected in all conditions tested.
(E) Normalized amplitude was unaffected both before and after the stimulus train in all conditions tested. 198
Figure 8.4 Mitochondrial OxPhos supports presynaptic Ca2+ buffering high frequency activity.
Ca2+ levels were monitored using genetically encoded Gcamp6 expressed
selectively in the presynaptic terminal via lentiviral injection into the ventral
cochlear nucleus (VCN). (A) Representative image of a Ca2+ calyceal
response during a high-frequency train. (B) Ca2+ responses presented as %
change are plotted as a function of time in control (black) and after addition 199
of IAA (red) or oligomycin (blue). (C) The maximum % change was increased due to oligomycin treatment. 200
Figure 8.5 Recovery of the RRP is not dependent on a specific ATP source in prehearing terminals.
Conditioning trains were applied to the afferent fiber at 100 Hz, 200 ms to
deplete the RRP, and a subsequent test train was recorded at increasing
interstimulus intervals to measure the time course of RRP recovery. Between 201
sweeps, 30 sec were allowed for full synapse recovery. (A) Representative traces of the recovery from synaptic depression protocol recorded in control conditions (black), or after at least 10 minute treatment with (IAA) red or oligomycin (blue). The representative traces portrayed had a 200 ms rest interval. (B) Conditioning train charge is shown per sweep under control conditions (black) and after treatment with IAA (red) or oligomycin (blue).
Note that response sizes were equivalent across all protocols. (C) Fractional
RRP recovery over time was plotted in control conditions (black), in IAA
(red), or in oligomycin (blue). RRP recovery curves were fit with a single exponential function (dotted lines). (D) RRP recovery time constants, fit from single cells, were unaffected by selective blockade of glycolysis or mitochondrial OxPhos. (E) Initial EPSC amplitudes of the conditioning train are shown per sweep in control (black), or after pretreatment with IAA (red) or oligomycin (blue). Note 1st EPSC peak amplitudes were increased in the presence of oligomycin. (F) Recovery of the initial EPSC, normalized to the depressed state, are plotted for control (black), IAA-treated (red), and oligomycin-treated (blue) cells. Curves were fit with a single exponential function. (G) Initial EPSC recovery time constants are shown for control IAA and oligomycin treatment. No significant differences were observed.
202
Figure 8.6 Compensation of ATP production during depression at 100 Hz at the mature calyx of Held.
Short, high-frequency trains were delivered to P16-18 animals at 100 Hz, 200
ms and EPSCs were recorded from the principal cells of the MNTB.
Kynurenic acid (1 mM) was included in the bath solution in all conditions to
block postsynaptic receptor desensitization. (A) Representative traces of 203
depression trains in control conditions (black) or after treatment with IAA
(red) or oligomycin (blue). (B) Preconditioning to HFS had no effect on
EPSC size. (C) Initial EPSC amplitudes were unaffected in all conditions tested. (D) Normalized synaptic depression during a HFS train. Individual traces were normalized to the first response. Depression curves from the 2nd to 20th stimulus could be adequately fit with a monoexponential decay function (dotted lines). (E) Mean depression time constants are shown for conditions tested. (F) PPR, measured as the peak amplitude of EPSC2/EPSC1, was unaffected by IAA or oligomycin. (G) Steady-state depression, measured at the end of the stimulus train, was unaffected by treatment with IAA or oligomycin. (H) Average cumulative EPSC plots were used to estimate the apparent RRP size and SV Pr. (I) Apparent RRP sizes, measured per cell, were normalized to control. No changes were observed. (J) Pr, measured as
EPSC1/RRP, was not significantly changed after treatment with IAA or oligomycin.
204
Figure 8.7 Recovery from synaptic depression is not dependent on a specific
ATP source in hearing terminals.
High-frequency stimulus trains were applied to the afferent fiber at 100 Hz,
200 ms to deplete the RRP, and EPSCs were recorded at increasing
interstimulus intervals to measure the time course of synaptic recovery.
Between each pair of sweeps, 30 sec were allowed for synaptic recovery.
Kynurenic acid (1 mM) was included in the bath solution in all conditions to 205
block postsynaptic receptor desensitization. (A) Control (black) and IAA-
(red) or oligomycin-treated (blue) conditioning and recovery traces are shown at the 200 ms interstimulus interval. (B) Fractional RRP recovery over time was fit by a monoexponential function (dotted lines). (C) RRP recovery time constants from single cells were unaffected by treatment with IAA (red) or oligomycin (blue). (D) Recovery of the initial EPSC, normalized to the depressed steady-state, are plotted for control (black), IAA-treated (red), and oligomycin-treated (blue) cells. Curves were similarly fit with a monoexponential function. (E) Summary of initial EPSC recovery time constants. Recovery was slowed in the presence of IAA.
206
Figure 8.8 The role of presynaptic ATP during depression while driving stimulation at 300 Hz at the mature calyx of Held.
Short, high-frequency trains (300 Hz, 150 ms) were applied to quickly
deplete the presynaptic RRP, and EPSCs were recorded from the innervated
cell body. Kynurenic acid (1 mM) was included in the bath solution in all 207
conditions to block postsynaptic receptor desensitization. (A) Representative trains recorded in control conditions (black), in IAA (red), or oligomycin
(blue). (B) Responses obtained after HFS preconditioning increases in the presence of oligomycin (C) Initial EPSC amplitudes at the beginning of the stimulus train were similar. (D) EPSC amplitudes were normalized to the first response, and plotted over time. Depression curves from the 2nd to 46th stimulus could be adequately fit with a single exponential decay function
(dotted lines). Both IAA and Oligomycin enhanced presynaptic depression.
(E) Summary plot of mean depression time constants. Oligomycin and IAA resulted in significant speeding of depression. (F) PPR, determined as
EPSC2/EPSC1, was decreased by oligomycin. (G) Steady-state depression, measured at the end of the depressing train, was significantly reduced in both
IAA and oligomycin. (H) Cumulative EPSC plots were used to estimate RRP and release Pr. (I) Summary of RRP size, determined per cell, was unaffected by IAA or oligomycin. (J) Oligomycin significantly enhanced Pr, while IAA had no effect.
208
Figure 8.9 Recovery from synaptic depression at the developmentally mature calyx requires optimal ATP production.
We performed recovery from synaptic depression experiments, similar to
what was described previously, at calyx synapses from P16-18 animals.
However, presynaptic depression was induced by a 300 Hz stimulus train,
150 ms duration, and fractional recovery of the RRP and single EPSCs were
measured following increasing rest periods, from 20 ms-13 sec. Kynurenic 209
acid (1 mM) was included in the bath solution in all conditions to block postsynaptic receptor desensitization. Between pairs of trains, 30 sec rest were allowed for full synaptic recovery. (A) Sample traces for synaptic recovery in control (black) and in the presence of IAA (red) or oligomycin
(blue) at the 200 ms interstimulus interval. (B) Fractional RRP recovery was plotted over time, and could be adequately fit with a double exponential function. Resulting fit parameters, determined per cell, are shown in C-E. (C)
RRP fast time constant, (D) RRP slow time constant, and (E) % fast component. Oligomycin significantly slowed the RRP slow time constant of recovery, while all other components were unchanged relative to control. (F)
Initial EPSC amplitude of the conditioning train is shown, and was unchanged during the experimental protocol. (G) Recovery of the initial
EPSC, normalized against the steady-state depression amplitude, in control and in the presence of IAA or oligomycin. Recovery was similarly fit by a double exponential function, and was not affected by IAA or oligomycin.
Summary of fit constants, per cell, are shown in in G-I. (G) Summary plot of initial EPSC recovery of fast time constant, (H) slow time constant, (I) % fast component, respectively. None of these parameters were affected by either
IAA or oligomycin treatment, relative to control.
210
8.5 Discussion
This study examined the specific contribution of glycolytic- versus mitochondrially-derived ATP to support presynaptic function during high-frequency
neurotransmission. Our data suggest that a specific metabolic profile exists to support
transmission, which changes over the course of synapse maturation. Thus, not only does
our study suggest that ATP production source may be important for reliable transmission,
but that the ATP production source may also change with development in support of
various ATP-dependent processes. Although mitochondrial OxPhos produces relatively
high ATP yield with low production rate, our data suggest that local ATP synthesis from
glycolysis also serves requisite roles in maintenance of presynaptic function during
ongoing activity, consistent with our results from Chapter 6. This hypothesis has been
supported by an increasing number of studies that suggest relatively low yield glycolytic
ATP production is crucial for physiological neuronal function (Jang et al. 2016). The
seemingly mobile ability of glycolysis in ATP production may be important in
maintenance of presynaptic ATPase activity, as evidenced by our study and others (Bak
et al. 2006; Jang et al. 2016).
8.5.1 Energy use in the presynaptic terminal of prehearing mice
We first examined the effects of source specific-inhibition of either glycolytic or
mitochondrial-derived ATP in maintenance of synaptic transmission using short trains of
100 Hz stimulation in young mice, prior to the onset of hearing. We show that inhibition
of glycolysis via IAA treatment and glucose starvation decreases SV Pr during a train of
stimuli, and thus slowed depression in P8-10 animals. This presumably is due to the 211
source-specific effects of IAA directly on the presynaptic AP waveform (cf. Chapter 7).
We also show that inhibition of mitochondrial ATP synthesis, via oligomycin treatment and extracellular pyruvate depletion, inhibits activity-dependent Ca2+ buffering.
Oligomycin treatment also increased SV Pr, and led to EPSC potentiation, without affecting the kinetics of depression. Interestingly, we found no effects of inhibition of either pathway on recovery from synaptic depression, suggesting that in the absence of one ATP production route that the other pathway is able to fully compensate.
8.5.2 Developmental shift in the presynaptic metabolic profile
We propose that a change in the developmental metabolic profile exists at the calyx of Held, as the same experiments performed on P8-10 animals yielded quite different results when performed after the onset of hearing, at P16-18. Namely, inhibition of glycolytic or mitochondrial derived ATP had surprisingly small effects during synaptic depression, using 100 Hz stimulation leading to synaptic reliability at the calyx of Held.
Because the calyx of held has a large (10-fold) safety factor, it may be reasonable to presume each pathway may be more readily compensated for by the other after developmental maturation (Lorteije et al. 2009). Similarly, after preconditioning with
HFS, we did not observe EPSC potentiation, or changes in Ca2+ buffering. This may be explained by tighter coupling of SVs to VGCCs and the formation of Ca2+ nano-domains at the developmentally mature calyx (Chen et al. 2015). Interestingly, we did observe a source-specific effect of IAA on recovery from synaptic depression, consistent with previous reports (Jang et al. 2016). 212
8.5.3 ATP is a bottleneck for high frequency transmission
Limiting presynaptic ATP from either glycolysis or mitochondrial OxPhos
inhibits transmission using short 300 Hz trains in post-hearing animals, at P16-18.
Interestingly, inhibition of either pathway alone resulted in faster presynaptic depression,
suggesting overlapping roles for glycolytic and mitochondrially-derived ATP using HFS
approaching the upper bound for reliable information processing at this age. This result
further suggests that one pathway in unable to compensate for the ATP deficit created by
inhibition of the other pathway alone. These data suggest that ATP is required to
maintain high-frequency transmission and inhibition of either pathway depresses the
amount of ATP available for use. Seemingly, initial priming or RRP refilling is
dependent on ATP derived from both oxidative phosphorylation and glycolysis at this age
(Hosoi et al. 2007; Sakaba 2006). Furthermore, loss of ATP slows the initial rate of SV mobilization from a recycling pool or a premature primed pool to a readily releasable state at the presynaptic terminal. However, the ATP-dependent molecular machinery supporting SV priming remains untested. 213
Chapter 9 Closing Remarks
The studies performed in this dissertation investigated the role of both the presynaptic and postsynaptic neuronal compartments to support proper cellular function.
We have reviewed the postsynaptically located NMDA receptor (NMDAR) and its role in
regulating survival and death signaling as well as its contribution to bidirectional synaptic
plasticity. We show a novel metabotropic function of the NMDAR to regulate both
neuroprotection as well as AMPA receptor (AMPAR) potentiation. The metabotropic
signaling of the NMDAR was shown to couple to the cell-survival protein kinase Akt and
the extracellular regulated kinase ½ (ERK1/2). This previously unknown function of the
NMDAR helps to elucidate the mechanism by which the NMDAR can regulate both
survival and death signaling as well as bidirectional synaptic plasticity.
Future studies should focus on the specific molecular mechanism by which a single ligand binding to the extracellular region of the NMDAR can influence intracellular function without the necessity of ion flux. We hypothesize that single ligand binding to the extracellular binding pocket may be sufficient to induce conformational changes to the intracellular C-terminal domain of specific NMDAR subunits, and that this steric rearrangement is sufficient to initiate cytoplasmic cellular signaling.
Clinical trials for pharmacological antagonists targeting NMDARs have been unsuccessful in treating various central nervous system (CNS) disease states (Ikonomidou and Turski 2002; Kemp and McKernan 2002). This may be due to the simple possibility that inhibition of NMDARs blocks both pro-survival signaling and cell-death signaling.
Targeting the metabotropic activating glycine binding pocket to specifically activate pro- 214
survival pathways, without potentiating cell death signaling, may provide a successful approach by which pro-survival signaling is enhanced without influencing the channel function or other NMDAR subtypes.
In the presynaptic compartment, we have identified a novel role for presynaptic
ATP in support of both basal and high-frequency neurotransmission. Namely, at rest, cellular ATP derived from glycolysis, and not mitochondrial respiration, preferentially supports the presynaptic action potential (AP) waveform. This altered AP waveform was shown to directly reduce presynaptic Ca2+-influx and subsequently inhibit synaptic transmission by decreasing synaptic vesicle release. We have revealed a previously unknown function of ATP derived from glycolysis to support presynaptic function. Using high-frequency stimulation patterns, we uncovered some overlapping functions of presynaptic ATP derived from either glycolysis, or mitochondrial oxidative phosphorylation (OxPhos). Interestingly, we also uncovered non-overlapping functions of cellular ATP in support of high-frequency and high-fidelity information processing. Our data suggest mitochondria preferentially regulate presynaptic Ca2+ homeostasis during repetitive activity in prehearing animals. Further, our data suggest that glycolysis preferentially supports recovery from synaptic depression after the onset of hearing.
Taken together, these data suggest that the specific ATP production route is important for function and that the value of ATP may differ depending on location and route of production.
Since all experiments were performed at room temperature, it would be interesting to see the effects of ATP deprivation at physiologically relevant temperatures. 215
Since glycolysis has a Q10=1.7 (Gray et al. 2006), and the AP waveform becomes even faster at physiological temperature (Taschenberger and von Gersdorff 2000), we may observe a more robust effect on the AP waveform by inhibition of glycolytic ATP under
these conditions. Future studies should also focus on the presynaptic metabolic profile
over the course of development. The calyx of Held continues maturation even after P18,
and it would be interesting to probe the presynaptic ATP requirements to support
synaptic neurotransmission after further refinement of this synapse.
The molecular basis underlying various CNS neurodegenerative disease states
remains unclear, however there is an overarching theme of dysregulation in neuronal
metabolism. Synaptic neuronal terminals are isolated, and located away from the cell body. Thus, synapses must synthesize ATP locally to meet their metabolic demands.
Because synapses are located at relatively distal locations from the cell body, ATP deficits have profound implications in synaptic function. Recently, synaptic vesicle retrieval by endocytic pathways has been suggested to be dependent on specific ATP
sources, but it remains controversial which ATP pathway predominates (Jang et al. 2016;
Pathak et al. 2015). Furthermore, the molecular machinery by which endocytosis occurs
remains unclear, and has been previously thought to be a GTP-dependent process (de
Hoop et al. 1994). Thus, how a decrement in ATP inhibits vesicular endocytosis remains unclear. The calyx of Held may be a good model to test this directly due to its well
described physiology and experimental accessibility. Further understanding of the basic
functions of synaptic transmission may elucidate new therapeutic targets to treat
debilitating CNS disorders. 216
References
Aarts M, Liu YT, Liu LD, Besshoh S, Arundine M, Gurd JW, Wang YT, Salter
MW, and Tymianski M. Treatment of ischemic brain damage by perturbing NMDA
receptor-PSD-95 protein interactions. Science 298: 2002.
Adler EM, Augustine GJ, Duffy SN, and Charlton MP. Alien intracellular calcium
chelators attenuate neurotransmitter release at the squid giant synapse. J Neurosci 11:
1496-1507, 1991.
Aleyasin H, Rousseaux MW, Phillips M, Kim RH, Bland RJ, Callaghan S, Slack RS,
During MJ, Mak TW, and Park DS. The Parkinson's disease gene DJ-1 is also a key
regulator of stroke-induced damage. Proc Natl Acad Sci U S A 104: 18748-18753, 2007.
Alle H, Roth A, and Geiger JRP. Energy-Efficient Action Potentials in Hippocampal
Mossy Fibers. Science 325: 1405-1408, 2009.
Ames A, 3rd, and Gurian BS. Effects of glucose and oxygen deprivation on function of isolated mammalian retina. J Neurophysiol 26: 617-634, 1963.
Anastasio NC, Xia Y, O'Connor ZR, and Johnson KM. Differential role of N-methyl-
D-aspartate receptor subunits 2A and 2B in mediating phencyclidine-induced perinatal neuronal apoptosis and behavioral deficits. Neuroscience 163: 1181-1191, 2009.
Andersen P, Sundberg SH, Sveen O, and Wigstrom H. Specific long-lasting potentiation of synaptic transmission in hippocampal slices. Nature 266: 736-737, 1977.
Andres-Mateos E, Perier C, Zhang L, Blanchard-Fillion B, Greco TM, Thomas B,
Ko HS, Sasaki M, Ischiropoulos H, Przedborski S, Dawson TM, and Dawson VL. 217
DJ-1 gene deletion reveals that DJ-1 is an atypical peroxiredoxin-like peroxidase. Proc
Natl Acad Sci U S A 104: 14807-14812, 2007.
Anegawa NJ, Guttmann RP, Grant ER, Anand R, Lindstrom J, and Lynch DR. N-
Methyl-D-aspartate receptor mediated toxicity in nonneuronal cell lines: characterization using fluorescent measures of cell viability and reactive oxygen species production. Brain
Res Mol Brain Res 77: 163-175, 2000.
Arai T, Hasegawa M, Akiyama H, Ikeda K, Nonaka T, Mori H, Mann D, Tsuchiya
K, Yoshida M, Hashizume Y, and Oda T. TDP-43 is a component of ubiquitin-positive tau-negative inclusions in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Biochem Biophys Res Commun 351: 602-611, 2006.
Aronowski J, Samways E, Strong R, Rhoades HM, and Grotta JC. An alternative method for the quantitation of neuronal damage after experimental middle cerebral artery occlusion in rats: analysis of behavioral deficit. J Cereb Blood Flow Metab 16: 705-713,
1996.
Auberson YP, Allgeier H, Bischoff S, Lingenhoehl K, Moretti R, and Schmutz M. 5-
Phosphonomethylquinoxalinediones as competitive NMDA receptor antagonists with a
preference for the human 1A/2A, rather than 1A/2B receptor composition. Bioorg Med
Chem Lett 12: 1099-1102, 2002.
Awatramani GB, Price GD, and Trussell LO. Modulation of transmitter release by
presynaptic resting potential and background calcium levels. Neuron 48: 109-121, 2005. 218
Ayala YM, Zago P, D'Ambrogio A, Xu YF, Petrucelli L, Buratti E, and Baralle FE.
Structural determinants of the cellular localization and shuttling of TDP-43. J Cell Sci
121: 3778-3785, 2008.
Babiec WE, Guglietta R, Jami SA, Morishita W, Malenka RC, and O'Dell TJ.
Ionotropic NMDA receptor signaling is required for the induction of long-term depression in the mouse hippocampal CA1 region. J Neurosci 34: 5285-5290, 2014.
Bak LK, Schousboe A, Sonnewald U, and Waagepetersen HS. Glucose is necessary to maintain neurotransmitter homeostasis during synaptic activity in cultured glutamatergic neurons. J Cereb Blood Flow Metab 26: 1285-1297, 2006.
Bak LK, Walls AB, Schousboe A, Ring A, Sonnewald U, and Waagepetersen HS.
Neuronal glucose but not lactate utilization is positively correlated with NMDA-induced neurotransmission and fluctuations in cytosolic Ca2+ levels. J Neurochem 109: 87-93,
2009.
Balfour RH, Hansen AM, and Trapp S. Neuronal responses to transient hypoglycaemia in the dorsal vagal complex of the rat brainstem. J Physiol 570: 469-484, 2006.
Baptista V, and Varanda WA. Glycine binding site of the synaptic NMDA receptor in subpostremal NTS neurons. J Neurophysiol 94: 147-152, 2005.
Barmada SJ, Skibinski G, Korb E, Rao EJ, Wu JY, and Finkbeiner S. Cytoplasmic mislocalization of TDP-43 is toxic to neurons and enhanced by a mutation associated with familial amyotrophic lateral sclerosis. J Neurosci 30: 639-649, 2010.
Barnes-Davies M, and Forsythe ID. Pre- and postsynaptic glutamate receptors at a giant excitatory synapse in rat auditory brainstem slices. J Physiol 488 ( Pt 2): 387-406, 1995. 219
Barria A, and Malinow R. Subunit-specific NMDA receptor trafficking to synapses.
Neuron 35: 345-353, 2002.
Barria A, Muller D, Derkach V, Griffith LC, and Soderling TR. Regulatory
phosphorylation of AMPA-type glutamate receptors by CaM-KII during long-term
potentiation. Science 276: 2042-2045, 1997.
Behe P, Stern P, Wyllie DJ, Nassar M, Schoepfer R, and Colquhoun D.
Determination of NMDA NR1 subunit copy number in recombinant NMDA receptors.
Proc Biol Sci 262: 205-213, 1995.
Benveniste H, Drejer J, Schousboe A, and Diemer NH. Elevation of the extracellular
concentrations of glutamate and aspartate in rat hippocampus during transient cerebral
ischemia monitored by intracerebral microdialysis. J Neurochem 43: 1369-1374, 1984.
Berberich S, Punnakkal P, Jensen V, Pawlak V, Seeburg PH, Hvalby O, and Kohr
G. Lack of NMDA receptor subtype selectivity for hippocampal long-term potentiation. J
Neurosci 25: 6907-6910, 2005.
Billups B, and Forsythe ID. Presynaptic mitochondrial calcium sequestration influences
transmission at mammalian central Synapses. J Neurosci 22: 5840-5847, 2002.
Bindman LJ, Murphy KP, and Pockett S. Postsynaptic control of the induction of
long-term changes in efficacy of transmission at neocortical synapses in slices of rat
brain. J Neurophysiol 60: 1053-1065, 1988.
Birnbaum JH, Bali J, Rajendran L, Nitsch RM, and Tackenberg C. Calcium flux-
independent NMDA receptor activity is required for Abeta oligomer-induced synaptic
loss. Cell Death Dis 6: e1791, 2015. 220
Bliss TV, and Collingridge GL. A synaptic model of memory: long-term potentiation in
the hippocampus. Nature 361: 31-39, 1993.
Bliss TV, and Gardner-Medwin AR. Long-lasting potentiation of synaptic transmission in the dentate area of the unanaestetized rabbit following stimulation of the perforant path. J Physiol 232: 357-374, 1973.
Bliss TV, and Lomo T. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J Physiol 232:
331-356, 1973.
Boehm J, Kang MG, Johnson RC, Esteban J, Huganir RL, and Malinow R. Synaptic incorporation of AMPA receptors during LTP is controlled by a PKC phosphorylation site on GluR1. Neuron 51: 213-225, 2006.
Bolshakov VY, Carboni L, Cobb MH, Siegelbaum SA, and Belardetti F. Dual MAP kinase pathways mediate opposing forms of long-term plasticity at CA3-CA1 synapses.
Nat Neurosci 3: 1107-1112, 2000.
Bonifati V, Rizzu P, Squitieri F, Krieger E, Vanacore N, van Swieten JC, Brice A, van Duijn CM, Oostra B, Meco G, and Heutink P. DJ-1( PARK7), a novel gene for autosomal recessive, early onset parkinsonism. Neurol Sci 24: 159-160, 2003.
Borges K, and Dingledine R. AMPA receptors: molecular and functional diversity. Prog
Brain Res 116: 153-170, 1998.
Borst JGG, and Sakmann B. Calcium current during a single action potential in a large presynaptic terminal of the rat brainstem. J Physiol 506: 143-157, 1998. 221
Borst JGG, and Sakmann B. Depletion of calcium in the synaptic cleft of a calyx-type
synapse in the rat brainstem. J Physiol 521: 123-133, 1999.
Borst JGG, and van Hoeve JS. The Calyx of Held Synapse: From Model Synapse to
Auditory Relay. Annu Rev Physiol 74: 199-224, 2012.
Bortolotto ZA, and Collingridge GL. Characterisation of LTP induced by the activation of glutamate metabotropic receptors in area CA1 of the hippocampus.
Neuropharmacology 32: 1-9, 1993.
Bouzier-Sore AK, Voisin P, Canioni P, Magistretti PJ, and Pellerin L. Lactate is a preferential oxidative energy substrate over glucose for neurons in culture. J Cereb Blood
Flow Metab 23: 1298-1306, 2003.
Brennan AM, Connor JA, and Shuttleworth CW. NAD(P)H fluorescence transients after synaptic activity in brain slices: predominant role of mitochondrial function. J
Cereb Blood Flow Metab 26: 1389-1406, 2006.
Brewer GJ, Torricelli JR, Evege EK, and Price PJ. Optimized survival of hippocampal neurons in B27-supplemented Neurobasal, a new serum-free medium combination. J Neurosci Res 35: 567-576, 1993.
Brigman JL, Wright T, Talani G, Prasad-Mulcare S, Jinde S, Seabold GK, Mathur
P, Davis MI, Bock R, Gustin RM, Colbran RJ, Alvarez VA, Nakazawa K, Delpire E,
Lovinger DM, and Holmes A. Loss of GluN2B-containing NMDA receptors in CA1 hippocampus and cortex impairs long-term depression, reduces dendritic spine density, and disrupts learning. J Neurosci 30: 4590-4600, 2010. 222
Buller AL, Larson HC, Schneider BE, Beaton JA, Morrisett RA, and Monaghan
DT. The molecular basis of NMDA receptor subtypes: native receptor diversity is
predicted by subunit composition. J Neurosci 14: 5471-5484, 1994.
Burke RE. Inhibition of mitogen-activated protein kinase and stimulation of Akt kinase
signaling pathways: Two approaches with therapeutic potential in the treatment of
neurodegenerative disease. Pharmacol Ther 114: 261-277, 2007.
Burnashev N, Schoepfer R, Monyer H, Ruppersberg JP, Gunther W, Seeburg PH,
and Sakmann B. Control by asparagine residues of calcium permeability and
magnesium blockade in the NMDA receptor. Science 257: 1415-1419, 1992.
Buzsaki G. Long-term potentiation of the commissural path-CA1 pyramidal cell synapse
in the hippocampus of the freely moving rat. Neurosci Lett 19: 293-296, 1980.
Cantley LC, and Neel BG. New insights into tumor suppression: PTEN suppresses
tumor formation by restraining the phosphoinositide 3-kinase/AKT pathway. Proc Natl
Acad Sci U S A 96: 4240-4245, 1999.
Cao X, Cui Z, Feng R, Tang YP, Qin Z, Mei B, and Tsien JZ. Maintenance of
superior learning and memory function in NR2B transgenic mice during ageing. Eur J
Neurosci 25: 1815-1822, 2007.
Carnevale NT, and Hines ML. The NEURON Book. Cambridge University Press, 2006.
Carroll RC, Beattie EC, von Zastrow M, and Malenka RC. Role of AMPA receptor
endocytosis in synaptic plasticity. Nat Rev Neurosci 2: 315-324, 2001.
Carroll RC, and Zukin RS. NMDA-receptor trafficking and targeting: implications for synaptic transmission and plasticity. Trends Neurosci 25: 571-577, 2002. 223
Chang N, El-Hayek YH, Gomez E, and Wan Q. Phosphatase PTEN in neuronal injury
and brain disorders. Trends Neurosci 30: 581-586, 2007.
Chang N, Li L, Hu R, Shan Y, Liu B, Wang H, Feng H, Wang D, Cheung C, Liao M,
and Wan Q. Differential regulation of NMDA receptor function by DJ-1 and PINK1.
Aging Cell 9: 837-850, 2010.
Chapman PF, Kairiss EW, Keenan CL, and Brown TH. Long-term synaptic potentiation in the amygdala. Synapse 6: 271-278, 1990.
Chen J, Li Y, Wang L, Zhang Z, Lu D, Lu M, and Chopp M. Therapeutic benefit of intravenous administration of bone marrow stromal cells after cerebral ischemia in rats.
Stroke 32: 1005-1011, 2001.
Chen L, Chetkovich DM, Petralia RS, Sweeney NT, Kawasaki Y, Wenthold RJ,
Bredt DS, and Nicoll RA. Stargazin regulates synaptic targeting of AMPA receptors by two distinct mechanisms. Nature 408: 936-943, 2000.
Chen M, Lu TJ, Chen XJ, Zhou Y, Chen Q, Feng XY, Xu L, Duan WH, and Xiong
ZQ. Differential Roles of NMDA Receptor Subtypes in Ischemic Neuronal Cell Death and Ischemic Tolerance. Stroke 39: 3042-3048, 2008.
Chen TW, Wardill TJ, Sun Y, Pulver SR, Renninger SL, Baohan A, Schreiter ER,
Kerr RA, Orger MB, Jayaraman V, Looger LL, Svoboda K, and Kim DS.
Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature 499: 295-300,
2013. 224
Chen ZX, Das B, Nakamura Y, DiGregorio DA, and Young SM. Ca2+ Channel to
Synaptic Vesicle Distance Accounts for the Readily Releasable Pool Kinetics at a
Functionally Mature Auditory Synapse. Journal of Neuroscience 35: 2083-2100, 2015.
Choi DW. Glutamate neurotoxicity and diseases of the nervous system. Neuron 1: 623-
634, 1988.
Clifton GL, Jiang JY, Lyeth BG, Jenkins LW, Hamm RJ, and Hayes RL. Marked
protection by moderate hypothermia after experimental traumatic brain injury. J Cereb
Blood Flow Metab 11: 114-121, 1991.
Collingridge GL, Isaac JT, and Wang YT. Receptor trafficking and synaptic plasticity.
Nat Rev Neurosci 5: 952-962, 2004.
Collingridge GL, Kehl SJ, and McLennan H. Excitatory amino acids in synaptic
transmission in the Schaffer collateral-commissural pathway of the rat hippocampus. J
Physiol 334: 33-46, 1983.
Constantine-Paton M, Cline HT, and Debski E. Patterned activity, synaptic
convergence, and the NMDA receptor in developing visual pathways. Annu Rev Neurosci
13: 129-154, 1990.
Constantinepaton M, Cline HT, and Debski E. Patterned activity, synaptic
convergence and the NMDA receptor in developing visual pathways. Annual Review of
Neuroscience 13: 1990.
Covasa M, Hung CY, Ritter RC, and Burns GA. Intracerebroventricular
administration of MK-801 increases food intake through mechanisms independent of
gastric emptying. Am J Physiol Regul Integr Comp Physiol 287: R1462-1467, 2004. 225
Crepel F, and Jaillard D. Pairing of pre- and postsynaptic activities in cerebellar
Purkinje cells induces long-term changes in synaptic efficacy in vitro. J Physiol 432: 123-
141, 1991.
Cummings JA, Nicola SM, and Malenka RC. Induction in the rat hippocampus of long-term potentiation (LTP) and long-term depression (LTD) in the presence of a nitric oxide synthase inhibitor. Neurosci Lett 176: 110-114, 1994.
David G. Mitochondrial clearance of cytosolic Ca2+ in stimulated lizard motor nerve terminals proceeds without progressive elevation of mitochondrial matrix Ca2+. J
Neurosci 19: 7495-7506, 1999.
Daw MI, Chittajallu R, Bortolotto ZA, Dev KK, Duprat F, Henley JM, Collingridge
GL, and Isaac JT. PDZ proteins interacting with C-terminal GluR2/3 are involved in a
PKC-dependent regulation of AMPA receptors at hippocampal synapses. Neuron 28:
873-886, 2000. de Hoop MJ, Huber LA, Stenmark H, Williamson E, Zerial M, Parton RG, and
Dotti CG. The involvement of the small GTP-binding protein Rab5a in neuronal endocytosis. Neuron 13: 11-22, 1994.
De Sarro G, Siniscalchi A, Ferreri G, Gallelli L, and De Sarro A. NMDA and
AMPA/kainate receptors are involved in the anticonvulsant activity of riluzole in DBA/2 mice. Eur J Pharmacol 408: 25-34, 2000.
DeRidder MN, Simon MJ, Siman R, Auberson YP, Raghupathi R, and Meaney DF.
Traumatic mechanical injury to the hippocampus in vitro causes regional caspase-3 and 226
calpain activation that is influenced by NMDA receptor subunit composition. Neurobiol
Dis 22: 165-176, 2006.
Dingledine R, Borges K, Bowie D, and Traynelis SF. The glutamate receptor ion channels. Pharmacol Rev 51: 7-61, 1999.
Dodt HU, Eder M, Schierloh A, and Zieglgansberger W. Infrared-guided laser stimulation of neurons in brain slices. Sci STKE 2002: pl2, 2002.
Dong H, O'Brien RJ, Fung ET, Lanahan AA, Worley PF, and Huganir RL. GRIP: a synaptic PDZ domain-containing protein that interacts with AMPA receptors. Nature
386: 279-284, 1997.
Doyere V, and Laroche S. Linear relationship between the maintenance of hippocampal long-term potentiation and retention of an associative memory. Hippocampus 2: 39-48,
1992.
Duchen MR. Ca(2+)-dependent changes in the mitochondrial energetics in single dissociated mouse sensory neurons. Biochem J 283 ( Pt 1): 41-50, 1992.
Duchen MR. Effects of metabolic inhibition on the membrane properties of isolated mouse primary sensory neurones. J Physiol 424: 387-409, 1990.
Dudek SM, and Bear MF. Homosynaptic long-term depression in area CA1 of hippocampus and effects of N-methyl-D-aspartate receptor blockade. Proc Natl Acad Sci
U S A 89: 4363-4367, 1992.
English JD, and Sweatt JD. A requirement for the mitogen-activated protein kinase cascade in hippocampal long term potentiation. J Biol Chem 272: 19103-19106, 1997. 227
Fantin M, Auberson YP, and Morari M. Differential effect of NR2A and NR2B subunit selective NMDA receptor antagonists on striato-pallidal neurons: relationship to motor response in the 6-hydroxydopamine model of parkinsonism. J Neurochem 106:
957-968, 2008.
Feldman J, and Barshi I. The effects of blood glucose levels on cognitive performance:
A review of the literature. Moffett Field: NASA Ames Research Center 2007.
Fiesel FC, Voigt A, Weber SS, Van den Haute C, Waldenmaier A, Gorner K, Walter
M, Anderson ML, Kern JV, Rasse TM, Schmidt T, Springer W, Kirchner R, Bonin
M, Neumann M, Baekelandt V, Alunni-Fabbroni M, Schulz JB, and Kahle PJ.
Knockdown of transactive response DNA-binding protein (TDP-43) downregulates histone deacetylase 6. Embo j 29: 209-221, 2010.
Fischer G, Mutel V, Trube G, Malherbe P, Kew JN, Mohacsi E, Heitz MP, and
Kemp JA. Ro 25-6981, a highly potent and selective blocker of N-methyl-D-aspartate receptors containing the NR2B subunit. Characterization in vitro. J Pharmacol Exp Ther
283: 1285-1292, 1997.
Fleck MW, Henze DA, Barrionuevo G, and Palmer AM. Aspartate and glutamate mediate excitatory synaptic transmission in area CA1 of the hippocampus. J Neurosci 13:
3944-3955, 1993.
Flint AC, Liu X, and Kriegstein AR. Nonsynaptic glycine receptor activation during early neocortical development. Neuron 20: 43-53, 1998. 228
Forsythe ID, Tsujimoto T, Barnes-Davies M, Cuttle MF, and Takahashi T.
Inactivation of presynaptic calcium current contributes to synaptic depression at a fast
central synapse. Neuron 20: 797-807, 1998.
Fox PT, Raichle ME, Mintun MA, and Dence C. Nonoxidative glucose consumption
during focal physiologic neural activity. Science 241: 462-464, 1988.
Gao C, Gill MB, Tronson NC, Guedea AL, Guzman YF, Huh KH, Corcoran KA,
Swanson GT, and Radulovic J. Hippocampal NMDA receptor subunits differentially
regulate fear memory formation and neuronal signal propagation. Hippocampus 20:
1072-1082, 2010.
Gary DS, and Mattson MP. PTEN regulates Akt kinase activity in hippocampal neurons and increases their sensitivity to glutamate and apoptosis. Neuromolecular Med 2: 261-
269, 2002.
Gautier CA, Kitada T, and Shen J. Loss of PINK1 causes mitochondrial functional defects and increased sensitivity to oxidative stress. Proc Natl Acad Sci U S A 105:
11364-11369, 2008.
Gazit N, Vertkin I, Shapira I, Helm M, Slomowitz E, Sheiba M, Mor Y, Rizzoli S, and Slutsky I. IGF-1 Receptor Differentially Regulates Spontaneous and Evoked
Transmission via Mitochondria at Hippocampal Synapses. Neuron 89: 583-597, 2016.
Geiger JR, and Jonas P. Dynamic control of presynaptic Ca(2+) inflow by fast-
inactivating K(+) channels in hippocampal mossy fiber boutons. Neuron 28: 927-939,
2000. 229
Gray SR, De Vito G, Nimmo MA, Farina D, and Ferguson RA. Skeletal muscle ATP
turnover and muscle fiber conduction velocity are elevated at higher muscle temperatures
during maximal power output development in humans. Am J Physiol Regul Integr Comp
Physiol 290: R376-382, 2006.
Guo XF, Macleod GT, Wellington A, Hu F, Panchumarthi S, Schoenfield M, Marin
L, Charlton MP, Atwood HL, and Zinsmaier KE. The GTPase dMiro is required for
axonal transport of mitochondria to Drosophila synapses. Neuron 47: 379-393, 2005.
Hall CN, Klein-Flugge MC, Howarth C, and Attwell D. Oxidative Phosphorylation,
Not Glycolysis, Powers Presynaptic and Postsynaptic Mechanisms Underlying Brain
Information Processing. J Neurosci 32: 8940-8951, 2012.
Haque ME, Thomas KJ, D'Souza C, Callaghan S, Kitada T, Slack RS, Fraser P,
Cookson MR, Tandon A, and Park DS. Cytoplasmic Pink1 activity protects neurons
from dopaminergic neurotoxin MPTP. Proc Natl Acad Sci U S A 105: 1716-1721, 2008.
Hardingham GE, Fukunaga Y, and Bading H. Extrasynaptic NMDARs oppose synaptic NMDARs by triggering CREB shut-off and cell death pathways. Nat Neurosci
5: 405-414, 2002.
Harris JJ, Jolivet R, and Attwell D. Synaptic Energy Use and Supply. Neuron 75: 762-
777, 2012.
Hayashi T, Thomas GM, and Huganir RL. Dual palmitoylation of NR2 subunits regulates NMDA receptor trafficking. Neuron 64: 213-226, 2009.
Hayashi T, Umemori H, Mishina M, and Yamamoto T. The AMPA receptor interacts with and signals through the protein tyrosine kinase Lyn. Nature 397: 72-76, 1999. 230
Hayashi Y, Shi SH, Esteban JA, Piccini A, Poncer JC, and Malinow R. Driving
AMPA receptors into synapses by LTP and CaMKII: requirement for GluR1 and PDZ
domain interaction. Science 287: 2262-2267, 2000.
He B, and Soderlund DM. Functional expression of Rat Nav1.6 voltage-gated sodium channels in HEK293 cells: modulation by the auxiliary beta1 subunit. PLoS One 9: e85188, 2014.
Healy DG, Abou-Sleiman PM, Gibson JM, Ross OA, Jain S, Gandhi S, Gosal D,
Muqit MM, Wood NW, and Lynch T. PINK1 (PARK6) associated Parkinson disease in Ireland. Neurology 63: 1486-1488, 2004.
Helmchen F, Borst JGG, and Sakmann B. Calcium dynamics associated with a single action potential in a CNS presynaptic terminal. Biophys J 72: 1458-1471, 1997.
Hirsch H, Euler KH, and Schneider M. Recovery of the brain after complete ischemia in hypothermia. Pflugers Arch 265: 314-327, 1957.
Hollmann M, and Heinemann S. Cloned glutamate receptors. Annu Rev Neurosci 17:
31-108, 1994.
Hollmann M, O'Shea-Greenfield A, Rogers SW, and Heinemann S. Cloning by functional expression of a member of the glutamate receptor family. Nature 342: 643-
648, 1989.
Hosoi N, Sakaba T, and Neher E. Quantitative analysis of calcium-dependent vesicle recruitment and its functional role at the calyx of Held synapse. J Neurosci 27: 14286-
14298, 2007. 231
Hotokezaka H, Sakai E, Kanaoka K, Saito K, Matsuo K, Kitaura H, Yoshida N, and
Nakayama K. U0126 and PD98059, specific inhibitors of MEK, accelerate differentiation of RAW264.7 cells into osteoclast-like cells. J Biol Chem 277: 47366-
47372, 2002.
Hrabetova S, and Sacktor TC. Long-term potentiation and long-term depression are induced through pharmacologically distinct NMDA receptors. Neurosci Lett 226: 107-
110, 1997.
Hrynevich SV, Pekun TG, Waseem TV, and Fedorovich SV. Influence of Glucose
Deprivation on Membrane Potentials of Plasma Membranes, Mitochondria and Synaptic
Vesicles in Rat Brain Synaptosomes. Neurochem Res 40: 1188-1196, 2015.
Hume RI, Dingledine R, and Heinemann SF. Identification of a site in glutamate receptor subunits that controls calcium permeability. Science 253: 1028-1031, 1991.
Iguchi Y, Katsuno M, Niwa J, Yamada S, Sone J, Waza M, Adachi H, Tanaka F,
Nagata K, Arimura N, Watanabe T, Kaibuchi K, and Sobue G. TDP-43 depletion induces neuronal cell damage through dysregulation of Rho family GTPases. J Biol
Chem 284: 22059-22066, 2009.
Ikemoto A, Bole DG, and Ueda T. Glycolysis and glutamate accumlation into synaptic vesicles - Role of glyceraldehyde phosphate dehydrogenase and 3-phosphogylcerate kinase. J Biol Chem 278: 5929-5940, 2003.
Ikonomidou C, and Turski L. Why did NMDA receptor antagonists fail clinical trials for stroke and traumatic brain injury? Lancet Neurol 1: 383-386, 2002. 232
Impey S, Obrietan K, Wong ST, Poser S, Yano S, Wayman G, Deloulme JC, Chan
G, and Storm DR. Cross talk between ERK and PKA is required for Ca2+ stimulation of
CREB-dependent transcription and ERK nuclear translocation. Neuron 21: 869-883,
1998.
Ishida A, Noda Y, and Ueda T. Synaptic Vesicle-bound Pyruvate Kinase can Support
Vesicular Glutamate Uptake. Neurochem Res 34: 807-818, 2009.
Ishikawa T, Nakamura Y, Saitoh N, Li WB, Iwasaki S, and Takahashi T. Distinct roles of Kv1 and Kv3 potassium channels at the calyx of Held presynaptic terminal. J
Neurosci 23: 10445-10453, 2003.
Ivannikov MV, Sugimori M, and Llinas RR. Calcium clearance and its energy requirements in cerebellar neurons. Cell Calcium 47: 507-513, 2010.
Ivanov AI, Malkov AE, Waseem T, Mukhtarov M, Buldakova S, Gubkina O,
Zilberter M, and Zilberter Y. Glycolysis and oxidative phosphorylation in neurons and astrocytes during network activity in hippocampal slices. J Cereb Blood Flow Metab 34:
397-407, 2014.
Izumi Y, Benz AM, Zorumski CF, and Olney JW. Effects of lactate and pyruvate on glucose deprivation in rat hippocampal slices. Neuroreport 5: 617-620, 1994.
Jahr CE, and Stevens CF. A quantitative description of NMDA receptor-channel kinetic behavior. J Neurosci 10: 1830-1837, 1990.
Jang S, Nelson JC, Bend EG, Rodriguez-Laureano L, Tueros FG, Cartagenova L,
Underwood K, Jorgensen EM, and Colon-Ramos DA. Glycolytic Enzymes Localize to
Synapses under Energy Stress to Support Synaptic Function. Neuron 2016. 233
Jin SX, and Feig LA. Long-term potentiation in the CA1 hippocampus induced by
NR2A subunit-containing NMDA glutamate receptors is mediated by Ras-GRF2/Erk
map kinase signaling. PLoS One 5: e11732, 2010.
Johnson JW, and Ascher P. Glycine potentiates the NMDA response in cultured mouse
brain neurons. Nature 325: 529-531, 1987.
Jones KH, and Senft JA. An improved method to determine cell viability by
simultaneous staining with fluorescein diacetate-propidium iodide. J Histochem
Cytochem 33: 77-79, 1985.
Kauppinen RA, and Nicholls DG. Synaptosomal bioenergetics. The role of glycolysis,
pyruvate oxidation and responses to hypoglycaemia. Eur J Biochem 158: 159-165, 1986.
Kemp JA, and McKernan RM. NMDA receptor pathways as drug targets. Nat
Neurosci 5 Suppl: 1039-1042, 2002.
Kerchner GA, and Nicoll RA. Silent synapses and the emergence of a postsynaptic mechanism for LTP. Nat Rev Neurosci 9: 813-825, 2008.
Kessels HW, Nabavi S, and Malinow R. Metabotropic NMDA receptor function is
required for beta-amyloid-induced synaptic depression. Proc Natl Acad Sci U S A 110:
4033-4038, 2013.
Kim CH, and Lisman JE. A labile component of AMPA receptor-mediated synaptic transmission is dependent on microtubule motors, actin, and N-ethylmaleimide-sensitive factor. J Neurosci 21: 4188-4194, 2001. 234
Kim JH, Sizov I, Dobretsov M, and von Gersdorff H. Presynaptic Ca2+ buffers
control the strength of a fast post-tetanic hyperpolarization mediated by the alpha 3
Na+/K+-ATPase. Nat Neurosci 10: 196-205, 2007.
Kim MJ, Dunah AW, Wang YT, and Sheng M. Differential roles of NR2A- and
NR2B-containing NMDA receptors in and AMPA receptor Ras-ERK signaling
trafficking. Neuron 46: 745-760, 2005a.
Kim RH, Peters M, Jang Y, Shi W, Pintilie M, Fletcher GC, DeLuca C, Liepa J,
Zhou L, Snow B, Binari RC, Manoukian AS, Bray MR, Liu FF, Tsao MS, and Mak
TW. DJ-1, a novel regulator of the tumor suppressor PTEN. Cancer Cell 7: 263-273,
2005b.
Kimura M, Saitoh N, and Takahashi T. Adenosine A(1) receptor-mediated presynaptic
inhibition at the calyx of Held of immature rats. J Physiol 553: 415-426, 2003.
Kitada T, Pisani A, Porter DR, Yamaguchi H, Tscherter A, Martella G, Bonsi P,
Zhang C, Pothos EN, and Shen J. Impaired dopamine release and synaptic plasticity in the striatum of PINK1-deficient mice. Proc Natl Acad Sci U S A 104: 11441-11446,
2007.
Knull HR. Association of glycolytic enzymes with particulate fractions from nerve-
endings. Biochim Biophys Acta 522: 1-9, 1978.
Kosterin P, Kim GH, Muschol M, Obaid AL, and Salzberg BM. Changes in FAD and
NADH fluorescence in neurosecretory terminals are triggered by calcium entry and by
ADP production. J Membr Biol 208: 113-124, 2005. 235
Koutsilieri E, and Riederer P. Excitotoxicity and new antiglutamatergic strategies in
Parkinson's disease and Alzheimer's disease. Parkinsonism Relat Disord 13 Suppl 3:
S329-331, 2007.
Koutsilleri E, and Riederer P. Excitotoxicity and new antiglutamatergic strategies in
Parkinson's diseaseand Alzheimer's disease. Parkinsonism Relat Disord 13: S329-S331,
2007.
Kuner T, and Schoepfer R. Multiple structural elements determine subunit specificity
of Mg2+ block in NMDA receptor channels. J Neurosci 16: 3549-3558, 1996.
Lachyankar MB, Sultana N, Schonhoff CM, Mitra P, Poluha W, Lambert S,
Quesenberry PJ, Litofsky NS, Recht LD, Nabi R, Miller SJ, Ohta S, Neel BG, and
Ross AH. A role for nuclear PTEN in neuronal differentiation. J Neurosci 20: 1404-
1413, 2000.
Larson J, Wong D, and Lynch G. Patterned stimulation at the theta frequency is optimal for the induction of hippocampal long-term potentiation. Brain Res 368: 347-
350, 1986.
Lee do Y, Xun Z, Platt V, Budworth H, Canaria CA, and McMurray CT. Distinct
pools of non-glycolytic substrates differentiate brain regions and prime region-specific
responses of mitochondria. PLoS One 8: e68831, 2013.
Lee JM, Zipfel GJ, and Choi DW. The changing landscape of ischaemic brain injury
mechanisms. Nature 399: A7-A14, 1999. 236
Lee K, Dixon AK, Rowe IC, Ashford ML, and Richardson PJ. Direct demonstration
of sulphonylurea-sensitive KATP channels on nerve terminals of the rat motor cortex. Br
J Pharmacol 115: 385-387, 1995.
Lee SY, and Kim JH. Mechanisms underlying presynaptic Ca2+ transient and vesicular
glutamate release at a CNS nerve terminal during in vitro ischaemia. J Physiol 593: 2793-
2806, 2015.
Leonard AS, Davare MA, Horne MC, Garner CC, and Hell JW. SAP97 is associated
with the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor GluR1 subunit. J Biol Chem 273: 19518-19524, 1998.
Levy WB, and Steward O. Synapses as associative memory elements in the hippocampal formation. Brain Res 175: 233-245, 1979.
Li LL, Ginet V, Liu X, Vergun O, Tuittila M, Mathieu M, Bonny C, Puyal J,
Truttmann AC, and Courtney MJ. The nNOS-p38MAPK pathway is mediated by
NOS1AP during neuronal death. J Neurosci 33: 8185-8201, 2013.
Lim L, Hall C, Leung T, Mahadevan L, and Whatley S. Neurone-specific enolase and creatine phosphokinase are protein components of rat brain synaptic plasma membranes.
J Neurochem 41: 1177-1182, 1983.
Lin JW, Ju W, Foster K, Lee SH, Ahmadian G, Wyszynski M, Wang YT, and Sheng
M. Distinct molecular mechanisms and divergent endocytotic pathways of AMPA receptor internalization. Nat Neurosci 3: 1282-1290, 2000.
Lipton P, and Robacker K. Glycolysis and brain-function - K + O stimulation of protein-synthesis and K+ uptake requires glycolysis. Fed Proc 42: 2875-2880, 1983. 237
Lipton SA, and Rosenberg PA. Excitatory amino acids as a final common pathway for neurologic disorders. N Engl J Med 330: 613-622, 1994.
Liu B, Liao M, Mielke JG, Ning K, Chen Y, Li L, El-Hayek YH, Gomez E, Zukin
RS, Fehlings MG, and Wan Q. Ischemic insults direct glutamate receptor subunit 2- lacking AMPA receptors to synaptic sites. J Neurosci 26: 5309-5319, 2006.
Liu LD, Wong TP, Pozza MF, Lingenhoehl K, Wang YS, Sheng M, Auberson YP, and Wang YT. Role of NMDA receptor subtypes in governing the direction of hippocampal synaptic plasticity. Science 304: 1021-1024, 2004.
Liu SJ, and Zukin RS. Ca2+-permeable AMPA receptors in synaptic plasticity and neuronal death. Trends Neurosci 30: 126-134, 2007.
Liu Y, Wong TP, Aarts M, Rooyakkers A, Liu L, Lai TW, Wu DC, Lu J, Tymianski
M, Craig AM, and Wang YT. NMDA receptor subunits have differential roles in mediating excitotoxic neuronal death both in vitro and in vivo. J Neurosci 27: 2846-2857,
2007.
Loftis JM, and Janowsky A. The N-methyl-D-aspartate receptor subunit NR2B: localization, functional properties, regulation, and clinical implications. Pharmacol Ther
97: 55-85, 2003.
Longa EZ, Weinstein PR, Carlson S, and Cummins R. Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke 20: 84-91, 1989.
Lorteije JA, Rusu SI, Kushmerick C, and Borst JG. Reliability and precision of the mouse calyx of Held synapse. J Neurosci 29: 13770-13784, 2009. 238
Lu W, Man H, Ju W, Trimble WS, MacDonald JF, and Wang YT. Activation of
synaptic NMDA receptors induces membrane insertion of new AMPA receptors and LTP
in cultured hippocampal neurons. Neuron 29: 243-254, 2001.
Lundgaard I, Li BM, Xie LL, Kang HY, Sanggaard S, Haswell JDR, Sun W,
Goldman S, Blekot S, Nielsen M, Takano T, Deane R, and Nedergaard M. Direct
neuronal glucose uptake heralds activity-dependent increases in cerebral metabolism. Nat
Commun 6: 2015.
Luo HR, Hattori H, Hossain MA, Hester L, Huang Y, Lee-Kwon W, Donowitz M,
Nagata E, and Snyder SH. Akt as a mediator of cell death. Proc Natl Acad Sci U S A
100: 11712-11717, 2003.
Luscher C, Xia H, Beattie EC, Carroll RC, von Zastrow M, Malenka RC, and Nicoll
RA. Role of AMPA receptor cycling in synaptic transmission and plasticity. Neuron 24:
649-658, 1999.
Lynch GS, Dunwiddie T, and Gribkoff V. Heterosynaptic depression: a postsynaptic
correlate of long-term potentiation. Nature 266: 737-739, 1977.
Lynch JW. Molecular structure and function of the glycine receptor chloride channel.
Physiol Rev 84: 1051-1095, 2004.
MacDonald JF, and Nowak LM. Mechanisms of blockade of excitatory amino acid receptor channels. Trends Pharmacol Sci 11: 167-172, 1990.
Mahfooz K, Singh M, Renden R, and Wesseling JF. A Well-Defined Readily
Releasable Pool with Fixed Capacity for Storing Vesicles at Calyx of Held. PLoS Comput
Biol 12: e1004855, 2016. 239
Malenka RC. Synaptic plasticity in the hippocampus: LTP and LTD. Cell 78: 535-538,
1994.
Malenka RC, and Bear MF. LTP and LTD: an embarrassment of riches. Neuron 44: 5-
21, 2004.
Malenka RC, and Nicoll RA. Neuroscience - Long-term potentiation - A decade of progress? Science 285: 1999.
Manabe T, and Nicoll RA. Long-term potentiation: evidence against an increase in transmitter release probability in the CA1 region of the hippocampus. Science 265: 1888-
1892, 1994.
Manning BD, and Cantley LC. AKT/PKB signaling: navigating downstream. Cell 129:
1261-1274, 2007.
Manzanero S, Santro T, and Arumugam TV. Neuronal oxidative stress in acute
ischemic stroke: sources and contribution to cell injury. Neurochem Int 62: 712-718,
2013.
Martel MA, Wyllie DJ, and Hardingham GE. In developing hippocampal neurons,
NR2B-containing N-methyl-D-aspartate receptors (NMDARs) can mediate signaling to neuronal survival and synaptic potentiation, as well as neuronal death. Neuroscience 158:
334-343, 2009.
Martin LJ, Blackstone CD, Levey AI, Huganir RL, and Price DL. AMPA glutamate
receptor subunits are differentially distributed in rat brain. Neuroscience 53: 327-358,
1993. 240
Massey PV, Johnson BE, Moult PR, Auberson YP, Brown MW, Molnar E,
Collingridge GL, and Bashir ZI. Differential roles of NR2A and NR2B-containing
NMDA receptors in cortical long-term potentiation and long-term depression. J Neurosci
24: 7821-7828, 2004.
Mayer ML, Westbrook GL, and Guthrie PB. Voltage-dependent block by Mg2+ of
NMDA responses in spinal cord neurones. Nature 309: 261-263, 1984.
McNaughton BL, Douglas RM, and Goddard GV. Synaptic enhancement in fascia
dentata: cooperativity among coactive afferents. Brain Res 157: 277-293, 1978.
Mejia-Garcia TA, Portugal CC, Encarnacao TG, Prado MA, and Paes-de-Carvalho
R. Nitric oxide regulates AKT phosphorylation and nuclear translocation in cultured
retinal cells. Cell Signal 25: 2424-2439, 2013.
Mercer RW, and Dunham PB. Membrane-bound ATP fuels the Na/K pump. Studies on membrane-bound glycolytic enzymes on inside-out vesicles from human red cell membranes. J Gen Physiol 78: 547-568, 1981.
Miwa H, Fukaya M, Watabe AM, Watanabe M, and Manabe T. Functional contributions of synaptically localized NR2B subunits of the NMDA receptor to synaptic transmission and long-term potentiation in the adult mouse CNS. J Physiol 586: 2539-
2550, 2008.
Monyer H, Sprengel R, Schoepfer R, Herb A, Higuchi M, Lomeli H, Burnashev N,
Sakmann B, and Seeburg PH. Heteromeric NMDA receptors: molecular and functional distinction of subtypes. Science 256: 1217-1221, 1992. 241
Nabavi S, Kessels HW, Alfonso S, Aow J, Fox R, and Malinow R. Metabotropic
NMDA receptor function is required for NMDA receptor-dependent long-term depression. Proc Natl Acad Sci U S A 110: 4027-4032, 2013.
Nakanishi N, Shneider NA, and Axel R. A family of glutamate receptor genes: evidence for the formation of heteromultimeric receptors with distinct channel properties.
Neuron 5: 569-581, 1990.
Nakielny S, Cohen P, Wu J, and Sturgill T. MAP kinase activator from insulin- stimulated skeletal muscle is a protein threonine/tyrosine kinase. Embo j 11: 2123-2129,
1992.
Neher E. Merits and Limitations of Vesicle Pool Models in View of Heterogeneous
Populations of Synaptic Vesicles. Neuron 87: 1131-1142, 2015.
Nguyen PV, Abel T, and Kandel ER. Requirement of a critical period of transcription for induction of a late phase of LTP. Science 265: 1104-1107, 1994.
Nicholls DG, and Budd SL. Mitochondria and neuronal survival. Physiol Rev 80: 315-
360, 2000.
Ning K, Pei L, Liao M, Liu B, Zhang Y, Jiang W, Mielke JG, Li L, Chen Y, El-
Hayek YH, Fehlings MG, Zhang X, Liu F, Eubanks J, and Wan Q. Dual neuroprotective signaling mediated by downregulating two distinct phosphatase activities of PTEN. J Neurosci 24: 4052-4060, 2004.
Nishimune A, Isaac JT, Molnar E, Noel J, Nash SR, Tagaya M, Collingridge GL,
Nakanishi S, and Henley JM. NSF binding to GluR2 regulates synaptic transmission.
Neuron 21: 87-97, 1998. 242
Okada Y. Reversibility of neuronal function of hippocampal slice during deprivation of
oxygen and/or glucose. Mechanisms of Cerebral Hypoxia and Stroke 35: 191-203, 1982.
Oliet SH, and Mothet JP. Regulation of N-methyl-D-aspartate receptors by astrocytic
D-serine. Neuroscience 158: 275-283, 2009.
Omori N, Jin G, Li F, Zhang WR, Wang SJ, Hamakawa Y, Nagano I, Manabe Y,
Shoji M, and Abe K. Enhanced phosphorylation of PTEN in rat brain after transient middle cerebral artery occlusion. Brain Res 954: 317-322, 2002.
Otto T, Eichenbaum H, Wiener SI, and Wible CG. Learning-related patterns of CA1 spike trains parallel stimulation parameters optimal for inducing hippocampal long-term
potentiation. Hippocampus 1: 181-192, 1991.
Passafaro M, Piech V, and Sheng M. Subunit-specific temporal and spatial patterns of
AMPA receptor exocytosis in hippocampal neurons. Nat Neurosci 4: 917-926, 2001.
Pathak D, Shields LY, Mendelsohn BA, Haddad D, Lin W, Gerencser AA, Kim H,
Brand MD, Edwards RH, and Nakamura K. The Role of Mitochondrially Derived
ATP in Synaptic Vesicle Recycling. J Biol Chem 290: 22325-22336, 2015.
Patterson SL, Pittenger C, Morozov A, Martin KC, Scanlin H, Drake C, and Kandel
ER. Some forms of cAMP-mediated long-lasting potentiation are associated with release of BDNF and nuclear translocation of phospho-MAP kinase. Neuron 32: 123-140, 2001.
Paul RJ, Hardin CD, Raeymaekers L, Wuytack F, and Casteels R. Preferential support of Ca-2+ uptake in smooth-muscle plasma-membrane vesicles by an endogenous glycolytic cascade. Faseb Journal 3: 2298-2301, 1989. 243
Pellerin L, Bergersen LH, Halestrap AP, and Pierre K. Cellular and subcellular distribution of monocarboxylate transporters in cultured brain cells and in the adult brain.
J Neurosci Res 79: 55-64, 2005.
Perkinton MS, Ip JK, Wood GL, Crossthwaite AJ, and Williams RJ.
Phosphatidylinositol 3-kinase is a central mediator of NMDA receptor signalling to MAP kinase (Erk1/2), Akt/PKB and CREB in striatal neurones. J Neurochem 80: 239-254,
2002.
Petullo D, Masonic K, Lincoln C, Wibberley L, Teliska M, and Yao DL. Model development and behavioral assessment of focal cerebral ischemia in rats. Life Sci 64:
1099-1108, 1999.
Planchon SM, Waite KA, and Eng C. The nuclear affairs of PTEN. J Cell Sci 121: 249-
253, 2008.
Prichard JW, Petroff OAC, Ogino T, and Shulman RG. Cerebral lactate elevation by electroshock - A H-1 magnetic-resonance study. Ann N Y Acad Sci 508: 54-63, 1987.
Raffin CN, Sick TJ, and Rosenthal M. Inhibition of glycolysis alters potassium-ion
transport and mitochondrial redox activity in rat-brain. J Cereb Blood Flow Metab 8:
857-865, 1988.
Rangaraju V, Calloway N, and Ryan TA. Activity-Driven Local ATP Synthesis Is
Required for Synaptic Function. Cell 156: 825-835, 2014.
Regehr WG, and Tank DW. Postsynaptic NMDA receptor-mediated calcium
accumulation in hippocampal CA1 pyramidal cell dendrites. Nature 345: 807-810, 1990. 244
Renden R, Taschenberger H, Puente N, Rusakov DA, Duvoisin R, Wang LY, Lehre
KP, and von Gersdorff H. Glutamate transporter studies reveal the pruning of
metabotropic glutamate receptors and absence of AMPA receptor desensitization at
mature calyx of held synapses. Journal of Neuroscience 25: 8482-8497, 2005.
Renden R, and von Gersdorff H. Synaptic vesicle endocytosis at a CNS nerve terminal:
Faster kinetics at physiological temperatures and increased endocytotic capacity during
maturation. J Neurophysiol 98: 3349-3359, 2007.
Roche KW, O'Brien RJ, Mammen AL, Bernhardt J, and Huganir RL.
Characterization of multiple phosphorylation sites on the AMPA receptor GluR1 subunit.
Neuron 16: 1179-1188, 1996.
Rodriguez M, Rodriguez-Sabate C, Morales I, Sanchez A, and Sabate M. Parkinson's disease as a result of aging. Aging Cell 14: 293-308, 2015.
Rodriguez-Moreno A, and Lerma J. Kainate receptor modulation of GABA release involves a metabotropic function. Neuron 20: 1211-1218, 1998.
Rose GM, and Dunwiddie TV. Induction of hippocampal long-term potentiation using physiologically patterned stimulation. Neurosci Lett 69: 244-248, 1986.
Rosen LB, Ginty DD, Weber MJ, and Greenberg ME. Membrane depolarization and calcium influx stimulate MEK and MAP kinase via activation of Ras. Neuron 12: 1207-
1221, 1994.
Rosenmund C, Clements JD, and Westbrook GL. Nonuniform probability of glutamate release at a hippocampal synapse. Science 262: 754-757, 1993. 245
Rost BR, Schneider F, Grauel MK, Wozny C, Bentz CG, Blessing A, Rosenmund T,
Jentsch TJ, Schmitz D, Hegemann P, and Rosenmund C. Optogenetic acidification of
synaptic vesicles and lysosomes. Nat Neurosci 18: 1845-1852, 2015.
Sakaba T. Roles of the fast-releasing and the slowly releasing vesicles in synaptic transmission at the calyx of Held. J Neurosci 26: 5863-5871, 2006.
Sakaba T, and Neher E. Quantitative relationship between transmitter release and calcium current at the calyx of held synapse. J Neurosci 21: 462-476, 2001.
Sano T, Lin H, Chen X, Langford LA, Koul D, Bondy ML, Hess KR, Myers JN,
Hong YK, Yung WK, and Steck PA. Differential expression of MMAC/PTEN in glioblastoma multiforme: relationship to localization and prognosis. Cancer Res 59:
1820-1824, 1999.
Schafe GE, Atkins CM, Swank MW, Bauer EP, Sweatt JD, and LeDoux JE.
Activation of ERK/MAP kinase in the amygdala is required for memory consolidation of pavlovian fear conditioning. J Neurosci 20: 8177-8187, 2000.
Schneggenburger R, and Forsythe ID. The calyx of Held. Cell Tissue Res 326: 311-
337, 2006.
Schneggenburger R, Meyer AC, and Neher E. Released fraction and total size of a pool of immediately available transmitter quanta at a calyx synapse. Neuron 23: 399-409,
1999.
Schneggenburger R, and Rosenmund C. Molecular mechanisms governing Ca(2+) regulation of evoked and spontaneous release. Nat Neurosci 18: 935-941, 2015. 246
Schurr A, West CA, and Rigor BM. Electrophysiology of energy metabolism and
neuronal function in the hippocampal slice preparation. J Neurosci Methods 28: 7-13,
1989.
Schurr A, West CA, and Rigor BM. Lactate-supported synaptic function in the rat
hippocampal slice preparation. Science 240: 1326-1328, 1988.
Selcher JC, Atkins CM, Trzaskos JM, Paylor R, and Sweatt JD. A necessity for MAP
kinase activation in mammalian spatial learning. Learn Mem 6: 478-490, 1999.
Sengupta B, Stemmler M, Laughlin SB, and Niven JE. Action potential energy
efficiency varies among neuron types in vertebrates and invertebrates. PLoS Comput Biol
6: e1000840, 2010.
Shan Y, Liu B, Li L, Chang N, Wang H, Wang D, Feng H, Cheung C, Liao M, Cui
T, Sugita S, and Wan Q. Regulation of PINK1 by NR2B-containing NMDA receptors
in ischemic neuronal injury. J Neurochem 111: 1149-1160, 2009.
Sheng M, Cummings J, Roldan LA, Jan YN, and Jan LY. Changing subunit
composition of heteromeric NMDA receptors during development of rat cortex. Nature
368: 144-147, 1994.
Shi SH, Hayashi Y, Petralia RS, Zaman SH, Wenthold RJ, Svoboda K, and
Malinow R. Rapid spine delivery and redistribution of AMPA receptors after synaptic
NMDA receptor activation. Science 284: 1811-1816, 1999.
Shoji S. Glucose regulation of synaptic transmission in the dorsolateral septal nucleus of
the rat. Synapse 12: 322-332, 1992. 247
Single FN, Rozov A, Burnashev N, Zimmermann F, Hanley DF, Forrest D, Curran
T, Jensen V, Hvalby O, Sprengel R, and Seeburg PH. Dysfunctions in mice by
NMDA receptor point mutations NR1(N598Q) and NR1(N598R). J Neurosci 20: 2558-
2566, 2000.
Sommer B, Keinanen K, Verdoorn TA, Wisden W, Burnashev N, Herb A, Kohler
M, Takagi T, Sakmann B, and Seeburg PH. Flip and flop: a cell-specific functional switch in glutamate-operated channels of the CNS. Science 249: 1580-1585, 1990.
Sommer B, Kohler M, Sprengel R, and Seeburg PH. RNA editing in brain controls a determinant of ion flow in glutamate-gated channels. Cell 67: 11-19, 1991.
Spuler A, Endres W, and Grafe P. Glucose depletion hyperpolarizes guinea pig
hippocampal neurons by an increase in potassium conductance. Exp Neurol 100: 248-
252, 1988.
Squire LR, and Zola-Morgan S. The medial temporal lobe memory system. Science
253: 1380-1386, 1991.
Srivastava S, Osten P, Vilim FS, Khatri L, Inman G, States B, Daly C, DeSouza S,
Abagyan R, Valtschanoff JG, Weinberg RJ, and Ziff EB. Novel anchorage of
GluR2/3 to the postsynaptic density by the AMPA receptor-binding protein ABP. Neuron
21: 581-591, 1998.
Stecker MM, and Stevenson M. Effect of glucose concentration on peripheral nerve and
its response to anoxia. Muscle Nerve 49: 370-377, 2014.
Stein IS, Gray JA, and Zito K. Non-Ionotropic NMDA Receptor Signaling Drives
Activity-Induced Dendritic Spine Shrinkage. J Neurosci 35: 12303-12308, 2015. 248
Steinberg GK, Panahian N, Perez-Pinzon MA, Sun GH, Modi MW, and Sepinwall
J. Narrow temporal therapeutic window for NMDA antagonist protection against focal cerebral ischaemia. Neurobiol Dis 2: 109-118, 1995.
Stocca G, and Vicini S. Increased contribution of NR2A subunit to synaptic NMDA
receptors in developing rat cortical neurons. J Physiol 507 ( Pt 1): 13-24, 1998.
Stornetta RL, and Zhu JJ. Ras and Rap signaling in synaptic plasticity and mental
disorders. Neuroscientist 17: 54-78, 2011.
Sughrue ME, Mocco J, Komotar RJ, Mehra A, D'Ambrosio AL, Grobelny BT, Penn
DL, and Connolly ES, Jr. An improved test of neurological dysfunction following
transient focal cerebral ischemia in rats. J Neurosci Methods 151: 83-89, 2006.
Sun HS, and Feng ZP. Neuroprotective role of ATP-sensitive potassium channels in
cerebral ischemia. Acta Pharmacol Sin 34: 24-32, 2013.
Sun T, Qiao HF, Pan PY, Chen YM, and Sheng ZH. Motile Axonal Mitochondria
Contribute to the Variability of Presynaptic Strength. Cell Reports 4: 413-419, 2013.
Sun Y, Jin K, Xie L, Childs J, Mao XO, Logvinova A, and Greenberg DA. VEGF-
induced neuroprotection, neurogenesis, and angiogenesis after focal cerebral ischemia. J
Clin Invest 111: 1843-1851, 2003.
Sutton G, and Chandler LJ. Activity-dependent NMDA receptor-mediated activation of protein kinase B/Akt in cortical neuronal cultures. J Neurochem 82: 1097-1105, 2002.
Sweatt JD. Mitogen-activated protein kinases in synaptic plasticity and memory. Curr
Opin Neurobiol 14: 311-317, 2004. 249
Swope SL, Moss SJ, Raymond LA, and Huganir RL. Regulation of ligand-gated ion channels by protein phosphorylation. Adv Second Messenger Phosphoprotein Res 33: 49-
78, 1999.
Talbot J, Barrett JN, Barrett EF, and David G. Stimulation-induced changes in
NADH fluorescence and mitochondrial membrane potential in lizard motor nerve terminals. J Physiol 579: 783-798, 2007.
Tamburri A, Dudilot A, Licea S, Bourgeois C, and Boehm J. NMDA-Receptor
Activation but Not Ion Flux Is Required for Amyloid-Beta Induced Synaptic Depression.
Plos One 8: 10, 2013.
Tang YP, Shimizu E, Dube GR, Rampon C, Kerchner GA, Zhuo M, Liu G, and
Tsien JZ. Genetic enhancement of learning and memory in mice. Nature 401: 63-69,
1999.
Taschenberger H, Leao RM, Rowland KC, Spirou GA, and von Gersdorff H.
Optimizing synaptic architecture and efficiency for high-frequency transmission. Neuron
36: 1127-1143, 2002.
Taschenberger H, and von Gersdorff H. Fine-tuning an auditory synapse for speed and fidelity: developmental changes in presynaptic waveform, EPSC kinetics, and synaptic plasticity. J Neurosci 20: 9162-9173, 2000.
Tekkok SB, Brown AM, Westenbroek R, Pellerin L, and Ransom BR. Transfer of glycogen-derived lactate from astrocytes to axons via specific monocarboxylate transporters supports mouse optic nerve activity. J Neurosci Res 81: 644-652, 2005. 250
Thomas GM, and Huganir RL. MAPK cascade signalling and synaptic plasticity. Nat
Rev Neurosci 5: 173-183, 2004.
Tovar KR, and Westbrook GL. The incorporation of NMDA receptors with a distinct subunit composition at nascent hippocampal synapses in vitro. J Neurosci 19: 4180-4188,
1999.
Toyoda H, Zhao MG, and Zhuo M. Roles of NMDA receptor NR2A and NR2B subtypes for long-term depression in the anterior cingulate cortex. Eur J Neurosci 22:
485-494, 2005.
Traynelis SF. Software-based correction of single compartment series resistance errors. J
Neurosci Methods 86: 25-34, 1998.
Traynor BJ, Bruijn L, Conwit R, Beal F, O'Neill G, Fagan SC, and Cudkowicz ME.
Neuroprotective agents for clinical trials in ALS: a systematic assessment. Neurology 67:
20-27, 2006.
Trussell LO, and Jackson MB. Dependence of an adenosine-activated potassium current on a GTP-binding protein in mammalian central neurons. J Neurosci 7: 3306-
3316, 1987.
Tu WH, Xu X, Peng LS, Zhong XF, Zhang WF, Soundarapandian MM, Balel C,
Wang MQ, Jia NL, Zhang W, Lew F, Chan SL, Chen YF, and Lu YM. DAPK1
Interaction with NMDA Receptor NR2B Subunits Mediates Brain Damage in Stroke.
Cell 140: 222-234, 2010. 251
Ueki M, Linn F, and Hossmann KA. Functional activation of cerebral blood-flow and
metabolism before and after global-ischemia of rat-brain. J Cereb Blood Flow Metab 8:
486-494, 1988.
Valente EM, Abou-Sleiman PM, Caputo V, Muqit MMK, Harvey K, Gispert S, Ali
Z, Del Turco D, Bentivoglio AR, Healy DG, Albanese A, Nussbaum R, Gonzalez-
Maldonaldo R, Deller T, Salvi S, Cortelli P, Gilks WP, Latchman DS, Harvey RJ,
Dallapiccola B, Auburger G, and Wood NW. Hereditary early-onset Parkinson's disease caused by mutations in PINK1. Science 304: 1158-1160, 2004.
Vanhoutte P, and Bading H. Opposing roles of synaptic and extrasynaptic NMDA receptors in neuronal calcium signalling and BDNF gene regulation. Curr Opin
Neurobiol 13: 366-371, 2003.
Verstreken P, Ly CV, Venken KJT, Koh TW, Zhou Y, and Bellen HJ. Synaptic mitochondria are critical for mobilization of reserve pool vesicles at Drosophila neuromuscular junctions. Neuron 47: 365-378, 2005.
Vignes M, and Collingridge GL. The synaptic activation of kainate receptors. Nature
388: 179-182, 1997.
Vissel B, Krupp JJ, Heinemann SF, and Westbrook GL. A use-dependent tyrosine dephosphorylation of NMDA receptors is independent of ion flux. Nat Neurosci 4: 587-
596, 2001. von Gersdorff H, and Borst JG. Short-term plasticity at the calyx of Held. Nat Rev
Neurosci 3: 53-64, 2002. 252
Wan Q, Xiong ZG, Man HY, Ackerley CA, Braunton J, Lu WY, Becker LE,
MacDonald JF, and Wang YT. Recruitment of functional GABA(A) receptors to
postsynaptic domains by insulin. Nature 388: 686-690, 1997.
Wang D, Qian L, Xiong H, Liu J, Neckameyer WS, Oldham S, Xia K, Wang J,
Bodmer R, and Zhang Z. Antioxidants protect PINK1-dependent dopaminergic neurons
in Drosophila. Proc Natl Acad Sci U S A 103: 13520-13525, 2006.
Wang JQ, Tang Q, Parelkar NK, Liu Z, Samdani S, Choe ES, Yang L, and Mao L.
Glutamate signaling to Ras-MAPK in striatal neurons: mechanisms for inducible gene
expression and plasticity. Mol Neurobiol 29: 1-14, 2004a.
Wang LY, and Kaczmarek LK. High-frequency firing helps replenish the readily
releasable pool of synaptic vesicles. Nature 394: 384-388, 1998.
Wang P, Saraswati S, Guan Z, Watkins CJ, Wurtman RJ, and Littleton JT. A
Drosophila temperature-sensitive seizure mutant in phosphoglycerate kinase disrupts
ATP generation and alters synaptic function. J Neurosci 24: 4518-4529, 2004b.
Wang YZ, Small DL, Stanimirovic DB, Morley P, and Durkin JP. AMPA receptor-
mediated regulation of a G(i)-protein in cortical neurons. Nature 389: 502-504, 1997.
Weiler S, Krinner S, Wong AB, Moser T, and Pangrsic T. ATP hydrolysis is critically required for function of CaV1.3 channels in cochlear inner hair cells via fueling Ca2+ clearance. J Neurosci 34: 6843-6848, 2014.
Weitlauf C, Honse Y, Auberson YP, Mishina M, Lovinger DM, and Winder DG.
Activation of NR2A-containing NMDA receptors is not obligatory for NMDA receptor- dependent long-term potentiation. J Neurosci 25: 8386-8390, 2005. 253
Wenthold RJ, Petralia RS, Blahos J, II, and Niedzielski AS. Evidence for multiple
AMPA receptor complexes in hippocampal CA1/CA2 neurons. J Neurosci 16: 1982-
1989, 1996.
Wexler EJ, Peters EE, Gonzales A, Gonzales ML, Slee AM, and Kerr JS. An
objective procedure for ischemic area evaluation of the stroke intraluminal thread model
in the mouse and rat. J Neurosci Methods 113: 51-58, 2002.
Williams JH, Li YG, Nayak A, Errington ML, Murphy KP, and Bliss TV. The
suppression of long-term potentiation in rat hippocampus by inhibitors of nitric oxide
synthase is temperature and age dependent. Neuron 11: 877-884, 1993.
Williams K. Ifenprodil discriminates subtypes of the N-methyl-D-aspartate receptor: selectivity and mechanisms at recombinant heteromeric receptors. Mol Pharmacol 44:
851-859, 1993.
Williams KL, Ferko AP, Barbieri EJ, and DiGregorio GJ. Glycine enhances the central depressant properties of ethanol in mice. Pharmacol Biochem Behav 50: 199-205,
1995.
Wong AY, Graham BP, Billups B, and Forsythe ID. Distinguishing between presynaptic and postsynaptic mechanisms of short-term depression during action potential trains. J Neurosci 23: 4868-4877, 2003.
Wong RW, Setou M, Teng J, Takei Y, and Hirokawa N. Overexpression of motor protein KIF17 enhances spatial and working memory in transgenic mice. Proc Natl Acad
Sci U S A 99: 14500-14505, 2002. 254
Wood-Kaczmar A, Gandhi S, Yao Z, Abramov ASY, Miljan EA, Keen G, Stanyer
L, Hargreaves I, Klupsch K, Deas E, Downward J, Mansfield L, Jat P, Taylor J,
Heales S, Duchen MR, Latchman D, Tabrizi SJ, and Wood NW. PINK1 Is Necessary for Long Term Survival and Mitochondrial Function in Human Dopaminergic Neurons.
Plos One 3: 2008.
Wu CP, Cheung G, Rakhshani N, Parvardeh S, Asl MN, Huang HL, and Zhang L.
Ca3 neuronal activities of dorsal and ventral hippocampus are differentially altered in rats after prolonged post-ischemic survival. Neuroscience 130: 527-539, 2005.
Wu K, Aoki C, Elste A, Rogalski-Wilk AA, and Siekevitz P. The synthesis of ATP by glycolytic enzymes in the postsynaptic density and the effect of endogenously generated nitric oxide. Proc Natl Acad Sci U S A 94: 13273-13278, 1997.
Xia J, Zhang X, Staudinger J, and Huganir RL. Clustering of AMPA receptors by the synaptic PDZ domain-containing protein PICK1. Neuron 22: 179-187, 1999.
Yamane K, Yokono K, and Okada Y. Anaerobic glycolysis is crucial for the maintenance of neural activity in guinea pig hippocampal slices. J Neurosci Methods
103: 163-171, 2000.
Yanagisawa D, Kitamura Y, Inden M, Takata K, Taniguchi T, Morikawa S, Morita
M, Inubushi T, Tooyama I, Taira T, Iguchi-Ariga SM, Akaike A, and Ariga H. DJ-1 protects against neurodegeneration caused by focal cerebral ischemia and reperfusion in rats. J Cereb Blood Flow Metab 28: 563-578, 2008. 255
Yang YM, and Wang LY. Amplitude and kinetics of action potential-evoked Ca2+ current and its efficacy in triggering transmitter release at the developing calyx of held synapse. J Neurosci 26: 5698-5708, 2006.
Yu XM, Askalan R, Keil GJ, and Salter MW. NMDA channel regulation by channel- associated protein tyrosine kinase Src. Science 275: 674-678, 1997.
Yun HY, Dawson VL, and Dawson TM. Glutamate-stimulated calcium activation of
Ras/Erk pathway mediated by nitric oxide. Diabetes Res Clin Pract 45: 113-115, 1999.
Zala D, Hinckelmann MV, Yu H, da Cunha MML, Liot G, Cordelieres FP, Marco
S, and Saudou F. Vesicular Glycolysis Provides On-Board Energy for Fast Axonal
Transport. Cell 152: 479-491, 2013.
Zarrindast MR, Jafari-Sabet M, Rezayat M, Djahanguiri B, and Rezayof A.
Involvement of NMDA receptors in morphine state-dependent learning in mice. Int J
Neurosci 116: 731-743, 2006.
Zhao MG, Toyoda H, Lee YS, Wu LJ, Ko SW, Zhang XH, Jia YH, Shum F, Xu H,
Li BM, Kaang BK, and Zhuo M. Roles of NMDA NR2B subtype receptor in prefrontal long-term potentiation and contextual fear memory. Neuron 47: 859-872, 2005.
Zhao YT, Tekkok S, and Krnjevic K. 2-Deoxy-D-glucose-induced changes in membrane potential, input resistance, and excitatory postsynaptic potentials of CA1 hippocampal neurons. Can J Physiol Pharmacol 75: 368-374, 1997.
Zheng M, Liao M, Cui T, Tian H, Fan DS, and Wan Q. Regulation of nuclear TDP-43 by NR2A-containing NMDA receptors and PTEN. J Cell Sci 125: 1556-1567, 2012. 256
Zhou Q, Petersen CC, and Nicoll RA. Effects of reduced vesicular filling on synaptic transmission in rat hippocampal neurones. J Physiol 525 Pt 1: 195-206, 2000.
Zhu JJ, Esteban JA, Hayashi Y, and Malinow R. Postnatal synaptic potentiation:
delivery of GluR4-containing AMPA receptors by spontaneous activity. Nat Neurosci 3:
1098-1106, 2000.
Zhu JJ, Qin Y, Zhao M, Van Aelst L, and Malinow R. Ras and Rap control AMPA receptor trafficking during synaptic plasticity. Cell 110: 443-455, 2002.
Zielke HR, Zielke CL, and Baab PJ. Direct measurement of oxidative metabolism in the living brain by microdialysis: a review. J Neurochem 109 Suppl 1: 24-29, 2009.
Zucker RS, and Regehr WG. Short-term synaptic plasticity. Annu Rev Physiol 64: 355-
405, 2002.