1 EXPANDING THE ROLE OF CANNABINOIDS IN THE SYNAPTIC PHYSIOLOGY OF THE HIPPOCAMPUS By MACKENZIE E. HOFMANN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREME NTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Mackenzie E. Hofmann
3 To my parents
4 ACKNOWLEDGMENTS I would first like to thank my parents for all their love and support over the years. They have always suppor ted and encouraged all of my academic endeavors and I owe my successes in life to them. I also would like to thank my brother Nick and my sister Caitie for their love, support, and always keeping a smile on my face. In addition, I thank the friends I hav e made in graduate school for always making lab a great place to be: Dr. Ben Nahir, Dr. Shelby Nahir, Casie Lindsly, and Dr. Chinki Bhatia. I would also like to thank my best friends Ryan McGhee and Seth Hennes. I especially thank Dr. Jason Frazier for f irst hiring me as a lab technician and then supporting me in my pursuit of a PhD in his lab. My future successes will be a direct result of his ability to prepare his students for a future in academic research. I would also like to thank him for his frie ndship over these years. In addition, I would like to thank my committee members Dr. Carney, Dr. Ache, and Dr. Peris for their assistance in my development as a scientist. I thank Dr. Jeff Tasker for his encouragement and support in pursuing a career in science. Finally, I would also like to extend thanks to Dr. Ron Dimock for providing me the opportunity to participate in scientific research where the first seed was planted for my future pursuits in academic research.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ........................... 10 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 Introduction ................................ ................................ ................................ ............. 15 Literature Review ................................ ................................ ................................ .... 19 Introduction to the Hippocampus ................................ ................................ ...... 19 Hippocampal Anatomy ................................ ................................ ..................... 20 CA3 pyramidal cells ................................ ................................ ................... 22 Mossy cells ................................ ................................ ................................ 24 Synaptic Transmission ................................ ................................ ..................... 26 Action potential dependent exocytosis ................................ ....................... 29 Action potential independent exocytosis ................................ .................... 29 Production and Breakdown of E Cs ................................ ................................ ... 32 GABA centric View of Cannabinoids in the Hippocampus ................................ 33 Functional Relevancy of Cannabinoids in the Hippocampus ............................ 37 Cholinergics in the hippocampus ................................ ............................... 38 Linking mAChRs and ECs for functionally relevant DSI ............................. 39 CB1 Independent Effects of Cannabinoids ................................ ....................... 40 2 MATERIALS AND METHODS ................................ ................................ ................ 44 Hippocampal Slice Preparation ................................ ................................ ............... 44 Electrophysiology ................................ ................................ ................................ .... 44 Identification of Mossy Cells ................................ ................................ ................... 47 Data Analysis ................................ ................................ ................................ .......... 48 Evoked EPSCs/IPSCs ................................ ................................ ...................... 48 DSI/DSE ................................ ................................ ................................ ........... 48 ADP ................................ ................................ ................................ .................. 48 Spontaneous and Miniature EPSC/IPSC Detection ................................ ......... 49 Statistics ................................ ................................ ................................ ........... 49 Drugs and Solutions ................................ ................................ ............................... 49 External Solutions ................................ ................................ ............................ 49 Internal Solutions ................................ ................................ .............................. 50 Drugs ................................ ................................ ................................ ................ 50
6 3 EXCITATORY AFFERENTS TO CA3 PYRAMIDAL CELLS DISPLAY DIFFERENTIAL SENSITIVITY TO CB1 DEPENDENT INHIBITION OF SYNAPTIC TRANSMISSION ................................ ................................ .................. 51 Introduc tion ................................ ................................ ................................ ............. 51 Results ................................ ................................ ................................ .................... 53 DSE is Present in Area CA3 ................................ ................................ ............. 53 WIN55,212 2 Sel ectively Inhibits Glutamate Release from A/C but not MF Inputs to CA3 Pyramidal Cells Via Activation of CB1 Receptors. .................. 54 Endocannabinoid Mediated Retrograde Signaling at Isolated A/C Inputs t o CA3 Pyramidal Cells Depends on the CB1 Receptor, Postsynaptic Calcium Influx, and Activation of mAChRs. ................................ ................... 55 Discussion ................................ ................................ ................................ .............. 56 4 MUSC ARINIC RECEPTOR ACTIVATION MODULATES THE EXCITABILITY OF HILAR MOSSY CELLS THROUGH THE INDUCTION OF AN AFTERDEPOLARIZATION ................................ ................................ ..................... 67 Introduction ................................ ................................ ................................ ............. 67 Results ................................ ................................ ................................ .................... 69 mAChR Activation Produces an Afterdepolarization in Hilar Mossy Cells but not in Other Hilar Neurons ................................ ................................ ............. 69 ADP Depends on a Calcium Activated Non Selective Cation Channel ............ 71 ADP Induction Can Result in the Release of ECs ................................ ............ 74 Discussion ................................ ................................ ................................ .............. 75 5 CANNABINOID RECEPTOR AGONISTS POTENTIATE ACTION POTENTIAL INDEPENDENT RELEASE OF GABA IN THE DENTATE GYRUS THROUGH A CB1 RECEPTOR INDEPENDENT MECHANISM ................................ .................. 86 Introduction ................................ ................................ ................................ ............. 86 Results ................................ ................................ ................................ .................... 87 WIN55,212 2 Produces a CB1 Receptor Independent Increase in mIPSC Freq uency ................................ ................................ ................................ ..... 87 Ruthenium Red Does Not Produce or Enable WIN55,212 2 Mediated and CB1 Receptor Independent Increases in mIPSC Frequency ........................ 88 AEA Also Produces Clear CB1 Receptor Independent Increases in mIPSC Frequency but the TRPV1 Agonist Capsaicin Does Not ............................... 90 Mechanism of WIN55,212 2 Mediated Facilitation of mIPSCs ......................... 91 WIN55,212 2 Has No Effect on Action Potential Dependent Exocytosis in the Presence of the CB1 Receptor Antagonist AM251 ................................ 96 Calcium is Required for WIN55,212 2 Mediated and CB1 Receptor Independent Increases in mIPSC Frequency ................................ ................ 97 Depolarization of Hilar Mossy Cells Sufficient to Produce Robust Activation of CB1 Receptors Has Only Minimal Effects on mIPSC Frequency .............. 99 Discussion ................................ ................................ ................................ ............ 101
7 6 CONCLUSIONS ................................ ................................ ................................ ... 122 Summary and Future Directions ................................ ................................ ........... 122 Differential Expression of CB1 at Glutamatergic Synapses to CA3 Pyramidal Cells ................................ ................................ ........................... 123 EC Involvement in Synaptic Plasticity at CA3 CA3 Synapses ........................ 124 mAChR Activation in Mossy Cells Results in an ADP ................................ .... 125 Induction of an ADP Can Cause DSI ................................ .............................. 127 Novel Effect of Cannabinoids on GABAergic Signaling to Mossy Cells .......... 128 Cannabinoid modulation of actio n potential independent exocytosis ....... 129 GPCR involvement in the novel cannabinoid effect ................................ 130 Potential physiological relevance of AEA in the novel cannabinoid effect 133 Perspectives ................................ ................................ ................................ ......... 135 Cannabinoids and Epilepsy ................................ ................................ ............ 135 Cholinergics and Epilepsy ................................ ................................ .............. 137 Cannabinoid System Following Chronic THC ................................ ................. 138 LIST OF REFERENCES ................................ ................................ ............................. 140 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 161
8 LIST OF FIGURES Figure page 3 1 DSE is observed in CA3 pyramidal cells and requires a ctivation of the CB1 receptor ................................ ................................ ................................ .............. 62 3 2 Functional expression of the CB1 receptor on A/C but not MF aff erents to CA3 pyramidal cells ................................ ................................ ............................ 63 3 3 WIN55,212 2 decreases the amplitude of minimally evoked A/C afferents in wild type but not CB1 / mice. ................................ ................................ .............. 64 3 4 DSE is observed in minimally evoked A/C afferent s and requires both the CB1 receptor and post synaptic calcium influx ................................ ................... 65 3 5 DSE of minimally evoked A/C afferents requires mAChR activation and is unaffected by inhi bition of nitric oxide sy nthase ................................ .................. 66 4 1 Muscarine can induce an ADP in mos sy cells but not interneurons ................... 80 4 2 ADP depends on openin g a voltage dependen t channel ................................ .... 81 4 3 Induction of the ADP does not depend on voltage gated sodium channels but does require sodium ions ................................ ................................ ................... 82 4 4 ADP s in mossy cells depend on calcium and depend on I CAN ............................ 83 4 5 ADP can cause a re lease of ECs resulting in DSI ................................ .............. 84 5 1 Bath appli cation of cannabinoid agonist WIN55,212 2 causes an increase in mIPSC frequen cy without altering amplitude ................................ .................... 110 5 2 Bath application of WIN55,212 2 does not alter intrinsic properties of hilar mossy cells under voltage clamp ................................ ................................ ...... 111 5 3 WIN55,212 2 mediated facilitation of mIPSCs is not dependent on CB1, CB2, or TRPV1 receptors ................................ ................................ .......................... 112 5 4 Ruthenium red does not independently produce, and is not required to observe, WIN55,212 2 mediated increases in mIPSC frequency ..................... 113 5 5 Bath application of AEA also selectivel y increases mIPSC frequency without altering amplitud e, area, rise time, or decay ................................ ..................... 114 5 6 Low doses of WIN55,212 2 effectively increase mIPSC frequency in hilar mossy cells. AEA is a more eff icacious agonist than 2 AG ............................. 115
9 5 7 Suramin effectively blocks WIN55,212 2 mediated facilitation of mIPSCs, but the non selective P2X rec eptor antagonist PPADS does not ........................... 116 5 8 WIN55,212 2 mediated potentiation of mIPSCs is blocked by a selective PKA inhibitor, and is both mimic ked and occluded by forskolin ................................ 117 5 9 Cannabinoid agonists have no effect on action potential dependent exocytosis in the presence of the CB1 receptor antagonist AM251 .................. 118 5 10 WIN55,212 2 mediated increases in mIPSC frequency ar e likely to depend on increases in pr esynaptic calcium concentration ................................ ........... 119 5 11 Depolarization of DSI positive hilar mossy cells produces a small, but comparatively long lasti ng, increase in mI PSC frequency ................................ 121
10 LIST OF ABBREVIATION S 2 AG 2 arachidonoylglycerol A/C associational/commissural ACSF artificial cerebral spinal fluid ADP afterdepolarization AEA anandamide CB1 cannabinoid type 1 CB1 / cannabinoid type 1 knockout CB2 cannabinoid type 2 CNS central nervous system COX cyclooxygenase DSE depolarization induced suppression of excitation DSI depolarization induced suppression of inhibition EC endogenous cannabinoid eEPSC evoked excitatory postsynaptic curren t eIPSC evoked inhibitory postsynaptic current GABA amino butyric acid GPCR G protein coupled receptor HETE hydroxyeicosatetraenoate I CAN calcium activated non selective current iLTD inhibitory long term depression LOX lipoxygenase LPA lysophosphatidic acid LTD long term depression LTP long term pote ntiation
11 mAChR muscarinic acetylcholine receptor meEPSC minimally evoked excitatory postsynaptic current MF mossy fiber mGluR metabotropic glutamate receptor mIPSC miniature inhibitory postsynaptic current NMG N methyl D glucamine PKA protein kinase A S1P sphingosphine 1 phosphate THC ( ) trans 9 tetrahydrocannabinol
12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXPANDING THE ROLE OF CANNABINOIDS IN THE SYNAPTIC PHYSIOLOGY OF THE HIPPOCAMPUS By Mackenzie E. Hofmann August 2010 Chair: Charles J. Frazier Major: Medical Sciences Neuroscience Since the initial discovery that endogenous cannabinoids can act as a retrograde messenger, there has been an e xplosion of research demonstrating the diverse role they can play in modulating neurotransmission. In the hippocampus they have been shown to modulate both short term and long term plasticities and they have been implicated in the etiology and potential t reatment of epilepsy. While much is known about cannabinoid function in the hippocampus, there still remain many unanswered questions. The studies in this dissertation were designed to answer some of these questions about how cannabinoids can modulate sy naptic transmission in this area of the brain. In our initial study, we investigated the role of the cannabinoid type 1 (CB1) receptor at glutamatergic synapses in CA3 pyramidal cells because CB1 at glutamatergic synapses has been understudied compared to CB1 at GABAergic synapses in the hippocampus. We discovered bath application of a cannabinoid agonist inhibited glutamate release from associational/commissural (A/C) synapses but not mossy fiber (MF) synapses demonstrating differential expression of the CB1 receptor. Depolarization of CA3 pyramidal cells caused a calcium dependent release of endogenous cannabinoids (ECs) resulting in a transient decrease in the release of
13 glutamate from A/C synapses. This effect was also sensitive to a CB1 receptor ant agonist and muscarinic acetylcholine receptor (mAChR) antagonist. To our knowledge, this represents the first report of depolarization induced suppression of excitation (DSE) in CA3 pyramidal cells. In an additional study we investigated how the activatio n of mAChRs can affect the intrinsic excitability of mossy cells, and how these changes can play a role in the release of ECs. Similar to reports in CA1, we found that application of mAChR agonists induced an afterdepolarization (ADP) in hilar mossy cells Interestingly, the ADP was absent in other hilar interneurons. In mossy cells the ADP was dependent upon the influx of sodium, the presence of postsynaptic calcium, and the activation of the calcium activated non selective channel (I CAN ). Previous res earch has indicated that depolarization induced suppression of inhibition (DSI) could not be obtained in vivo in the hippocampus, but we now provide evidence for a potential pathway for physiologically relevant DSI. We demonstrated that induction of an AD P can cause a release of ECs leading to DSI. This ADP DSI was similar in magnitude to that elicited with a conventional 5 second depolarization and was also blocked by a CB1 receptor antagonist. Finally, during a previous study on DSI in hilar mossy cells we discovered a novel effect of cannabinoids in the dentate gyrus that we have now further investigated. We found that both endogenous and synthetic cannabinoids could cause an increase in action potential independent exocytosis of GABA through a CB1 rece ptor independent mechanism. This was unlike traditional cannabinoid signaling which causes a decrease in action potential dependent exocytosis of GABA through the CB1 receptor. The
14 excitatory effect of cannabinoids was still present in CB1, CB2, and TRPV 1 antagonists and preserved in CB1 / mice. Our data indicate that uncoupling G proteins from their receptors with suramin prevents the facilitation of action potential independent exocytosis suggesting a role for a G protein coupled receptor (GPCR) In support of these results, we found that activation of adenylyl cyclase and protein kinase A were also necessary for the novel cannabinoid effect suggesting the signaling cascade: S that a depolarization that could induce EC production and DSI could cause a small but significant increase in action potential independent exocytosis. Our results imply cannabinoids can have a novel, CB1 independent effect on synaptic transmission in the dentate hilus. Overall these studies add significant results to the growing literature of cannabinoid mediated modulation of synaptic transmission in the hippocampus. We have demonstrated that CB1 is present at glutamatergic synapses in CA3 pyramidal cell s, ADPs are potential mechanisms for the physiological induction of DSI, and cannabinoids can have a CB1 independent effect on synaptic transmission.
15 CHAPTER 1 INTRODUCTION Introduction For centuries marijuana use has permeated society both as a potential therapy to treat several different symptoms and as a recreational drug for its psychotropic effects. Medicinal marijuana has been shown to reduce pain, alleviate nausea/vomiting, and increase appetite (Carlini, 2004) In addition, there is also substanti al evidence for its use as an anticonvulsant which is of particular interest for the potential treatment of diseases such as epilepsy (Karler and Turkanis, 1981; Lutz, 2004) However, the usefulness of marijuana in treatment regimens has been debated due to the potential negative side effects. Chronic use of marijuana can cause impairments to memory and cognition and can lead to dependence and withdrawal (Carlini, 2004; Ranganathan and D'Souza, 2006; Cooper and Haney, 2009) For many years now scientists have been investigating how marijuana works in the brain in an attempt to find new therapeutic strategies selectively targeting the beneficial effects while avoiding the negative effects of this drug. ( ) trans 9 tetrahydrocannabinol (THC), was first discovered, but it took almost an additional 30 years to find the gene coding for the receptor to which it binds (Gaoni and Mechoulam, 1964) It was in 1990 that the cannabinoid type 1 (CB1) receptor was first clon ed (Matsuda et al., 1990) and soon after the cannabinoid type 2 (CB2) receptor was also discovered (Munro et al., 1993) Both of these receptors are seven transmembrane G protein coupled receptors (GPCRs) that bind to G i/o with vastly different expressio n in the central nervous system (CNS). The CB1 receptor has a wide distribution with
16 particularly strong expression in the cerebellum, cortex, basal ganglia, and the hippocampus (Herkenham et al., 1991; Tsou et al., 1998) In contrast, CB2 is found predo minantly in the immune system but has been reported in the cerebellum and brainstem (Skaper et al., 1996; Van Sickle et al., 2005) While these findings and many others began to shed new light into how marijuana could potentially be mediating its effects on the brain through the CB1 receptor, it took another decade before discovering that cannabinoids could modulate synaptic plasticity by acting as a retrograde messenger. In 2001 Wilson and Nicoll demonstrated that endogenous cannabinoids (ECs) could trave l in a retrograde manner to transiently inhibit GABAergic synapses to CA1 pyramidal cells in the hippocampus. This provided the first evidence that ECs were the responsible retrograde messenger in a form of short term plasticity called depolarization indu ced suppression of inhibition (DSI). The role of CB1 in DSI was confirmed because cannabinoid antagonists blocked it, cannabinoid agonists occluded it, and it was absent in CB1 knockout (CB1 / ) mice (Wilson et al., 2001; Wilson and Nicoll, 2001) Shortl y following this study DSI was also shown in the cerebellum along with depolarization induced suppression of excitation (DSE) (Kreitzer and Regehr, 2001b, a; Maejima et al., 2001) Since this initial discovery extensive progress has been made in understan ding the role of CB1 in synaptic physiology and in linking CB1 function to both the good and bad effects of marijuana. For example, research has shown that the CB1 receptor can play an important role in modulating both long term depression (LTD) and long term potentiation (LTP) (Kano et al., 2009) which are thought by many to be the synaptic level constituent of learning and memory (Lynch, 2004; Massey and Bashir, 2007) In
17 fact, chronic administration of THC to rats results in an inability to obtain LTP possibly accounting for the memory and cognition impairment observed in humans (Hoffman et al., 2007) The anticonvulsant effect of THC and other cannabinoids has been shown to be dependent on activation of CB1 receptors (Wallace et al., 2001; Wallace et al., 2002; Wallace et al., 2003; Blair et al., 2006) In support of these findings recent data also indicates that the CB1 receptor is important for reducing seizure susceptibility (Monory et al., 2006) and is down regulated in human epileptic tissue (Lu danyi et al., 2008) suggesting a potentially significant role for these receptors in epilepsy. These studies and many others have begun to show the diverse role cannabinoids can play in synaptic physiology and to provide insight for new potential therapeu tic strategies. Our lab is particularly interested in studying both learning and memory and epilepsy in the hippocampus. Due to the involvement of cannabinoids in these areas of study, we sought to further investigate the role of cannabinoids in the synap tic physiology of the hippocampus. While significant progress has been made to date there still remain many unanswered questions. Several of these questions form the foundation for the chapters of this dissertation in an effort to expand our knowledge of cannabinoid function in the hippocampus. First, the initial immunohistochemistry performed with a CB1 antibody showed that CB1 was found almost exclusively at GABAergic synapses in the hippocampus (Tsou et al., 1998) These data are at odds with the kno wn anticonvulsant and antiepileptic effects of cannabinoids through the CB1 receptor which suggests CB1 should be present at glutamatergic synapses (Wallace et al., 2001; Wallace et al., 2002; Wallace et al., 2003; Blair et al., 2006) In addition, these results created a very GABA centric focus of cannabinoid function in the
18 hippocampus. Consequently, the first question we wanted to answer in Chapter 3 was whether the CB1 receptor was present at glutamatergic synapses to CA3 pyramidal cells and if so, ar e they involved in cannabinoid dependent short term plasticity. Second, to date it has been difficult to get a form of cannabinoid dependent short term plasticity, DSI, in an in vivo system bringing into question its physiological relevance (Hampson et a l., 2003) Several different studies have demonstrated that activation of muscarinic acetylcholine receptors (mAChRs) increases both the amplitude and duration of DSI (Kim et al., 2002; Ohno Shosaku et al., 2003; Fukudome et al., 2004; Hofmann et al., 200 6; Hashimotodani et al., 2008) and activation of these receptors can also change the intrinsic properties of hippocampal neurons (Fraser and MacVicar, 1996; McQuiston and Madison, 1999; Lawrence et al., 2006a; Lawrence et al., 2006b) The next question w e asked in Chapter 4 was whether mAChR activation changes the intrinsic properties of mossy cells providing a physiological mechanism for the induction of DSI. Third, recent findings from our lab while investigating DSI at hilar mossy cells indicated tha t cannabinoids could have a novel effect on GABAergic signaling. Many labs have demonstrated that cannabinoids can have diverse functions independent of CB1 and possibly through novel cannabinoid receptors (Oz, 2006a; Kreitzer and Stella, 2009) The fina l question we wanted to answer in Chapter 5 was whether cannabinoids can modulate neurotransmission to mossy cells through a CB1 receptor independent pathway. The following literature review seeks to first provide the relevant background on the hippocampu s and synaptic transmission to build the proper foundation for investigating questions about cannabinoids in this area. The second part of the
19 literature review will describe in more detail why these are important questions to study in an effort to expand our knowledge of cannabinoid function in the hippocampus. Literature Review Introduction to the Hippocampus Since the inception of brain investigation, the hippocampus, located in the temporal lobes, has been a focal point of research. The name originat ed from the Bolognese anatomist Giulio Cesare Aranzi for its striking similarity to the seahorse. For many years there was diverse speculation into the function of the hippocampus ranging from olfaction to emotion (Andersen et al., 2007) While the funct ion was debated it was well known that removal of this structure could alleviate seizures in temporal lobe epilepsy and autopsied epileptic patients often presented with hippocampal sclerosis (Walker et al., 2007) In fact it was the attempted treatment o f this disease which led to the discovery of the function of the hippocampus. In 1953 the famous patient H.M. had a bilateral resection of the medial portions of the temporal lobe removing both hippocampi in an attempt to eliminate his seizures which had proven to be resistant to antiepileptic drugs. After the surgery, doctors discovered the removal of his hippocampi resulted in profound memory loss. This memory loss included anterograde amnesia, an inability to remember material after the operation, an d retrograde amnesia, an inability to remember information for a period prior to his operation, and was published in a seminal paper by Scoville and Milner (1957) Additional work with patient H.M. allowed Milner to make several other important observatio ns. She noted the surgery did not create a loss in general intellect and did not impair immediate or working memory. In addition, damage to the hippocampus did not affect all forms of long term memory suggesting the hippocampus
20 is responsible for certain types of memory formation (Stark, 2007) The work by Milner and countless other neuroscientists have provided a vast array of evidence that the hippocampus is the gateway for learning and memory (Squire and Zola Morgan, 1991) As such, there is high int erest in understanding the mechanisms responsible for this process. Insights could lead to potential therapeutics for the memory loss associated with aging, and they could provide additional knowledge about how drug abuse affects learning and memory. In addition, the hippocampus is also of particular interest in hypoglycemia (Walker et al., 2007) Hippocampal Anatomy In order to discuss in detail the questions we want to ask about cannabinoids in the hippocampus it is necessary to first understand the functional roles of CA3 pyramidal cells and hilar mossy cells. The goal of this section is to first place these cell types in the context of hippocampal anatomy as a whole before det ailing the specific characteristics of each. The hippocampus is composed of two primary cell layers that ammonis. The dentate granule cells are elliptical shaped cells w hose dendrites extend throughout the inner, medial, and lateral sections of stratum moleculare. The axonal projections of granule cells are termed mossy fibers (MFs) and they form connections with interneurons throughout the hilar region of the dentate gy rus and with basket cells in the granule cell layer. The MFs also project into area CA3 forming synapses at CA3 pyramidal cells and local area interneurons specifically in stratum lucidum. Granule cells receive intrinsic connections from various interneu rons including basket cells, somatostatin expressing interneurons, and hilar mossy cells. The primary extrinsic
21 connection to granule cells, confined to the medial and lateral layers of stratum moleculare, is called the perforant path which originates fro m the entorhinal cortex. They also receive extrinsic synaptic connections from the medial septal nucleus and the supramammillary area. distinct sections based on anatomy and connectivity: CA1, CA2, and CA3. This section will briefly discuss CA1 and CA2 pyramidal cells with a more detailed discussion of CA3 pyramidal cells below. Both CA1 and CA2 pyramidal cells have basal dendrites that extend into stratum oriens and an apical dendritic arbor that extends into stratum radiatum and stratum lacunosum moleculare. CA2 pyramidal cell axons synapse onto local area interneurons and the apical dendrites of CA1 pyramidal cells, while CA1 pyramidal cell axons project to the subicu lum, deep layers of the entorhinal cortex, and local area interneurons. The primary intrinsic synaptic connections to CA1 pyramidal cells originate from ipsilateral and contralateral CA3 pyramidal cells, termed the Schaeffer collaterals, CA2 pyramidal cel ls, and local area interneurons. Both cell types receive extrinsic innervations from the entorhinal cortex via the perforant path and from the medial septal nucleus (Amaral and Lavenex, 2007) In addition to the primary cell layers, the hippocampus conta ins a plethora of different types of interneurons found in all layers of the hippocampus (for a comprehensive review see Freund and Buzsaki, 1996) Typically interneurons are considered local circuit neurons because their dendritic and axonal arbors stay within the subfield of their cell bodies (e.g. interneurons with cell bodies in stratum radiatum of CA1 typically have axonal and dendritic arborizations within area CA1). The synaptic
22 connections formed by interneurons are found in diverse locations at axons, somas, and dendrites of both interneurons and the primary cell layers each uniquely positioned for modulating synaptic transmission. While the primary cell layers are glutamatergic, that is they release glutamate from their axon terminals, the majo rity of interneurons release the inhibitory neurotransmitter amino butyric acid (GABA); however, there is one notable exception in the dentate gyrus, the hilar mossy cell, which is described in detail below. When considering hippocampal anatomy it is also important to understand how information flows through the hippocampus. The classic unidirectional circuit through the hippocampus is called the trisynaptic pathway. The signal originates in the entorhinal cortex, enters the dentate gyrus through the hippocampal fissure via the perforant path, and is transmitte d to the granule cell dendrites (synapse 1). From the granule cells the signal travels via the MFs to the proximal dendrites of CA3 pyramidal cells (synapse 2). Finally, the CA3 pyramidal cells send the signal to the CA1 pyramidal cells via the Schaeffe r collaterals (synapse 3) and the CA1 pyramidal cells transmit the signal back out of the hippocampus (Amaral and Lavenex, 2007) CA3 pyramidal cells The CA3 pyramidal cell layer begins in the hilar region of the dentate gyrus and extends until the end o f stratum lucidum, a region known as the end bulb. These cells have one dendritic arbor that projects apically into stratum lucidum, radiatum, and laconosum moleculare, while a second dendritic arbor also extends basally into stratum oriens. The axonal p rojections of CA3 pyramidal cells that synapse onto CA1 pyramidal cells are termed the Schaeffer collaterals while the projections that synapse on other CA3 pyramidal cells are termed the associational/commissural (A/C) fibers (discussed
23 below). These axo ns can project to both the ipsilateral and contralateral CA1 and CA3 regions of the hippocampus. The primary intrinsic connections to CA3 pyramidal cells are the MFs from dentate granule cells (discussed below), A/C fibers, and local area interneurons. T hese cells also receive extrinsic innervations from the perforant path and medial septal nucleus (Amaral and Lavenex, 2007) Mossy fiber synapses. MFs are the axonal projections of dentate granule cells and traverse stratum lucidum forming proximal synap ses on the apical dendrites of CA3 pyramidal cells. These synapses form large thorny excrescences, which appear as large spines along the proximal dendrite s under fluorescent microscopy. Each CA3 pyramidal cell receives about 50 different MF synapses and each individual MF forms synapses on 11 18 CA3 pyramidal cells (Henze et al., 2000) MFs form more synapses with stratum lucidum interneurons (40 50 interneurons per MF) which are involved in a feed forward circuit with MF axons. These synapses also hav e a higher potency for spike transmission causing a net inhibitory effect and the activation of a subset of CA3 pyramidal cells (Acsady et al., 1998; Henze et al., 2000; Lawrence et al., 2004) Associational/Commissural synapses. A/C fibers are the recurr ent connections from other CA3 pyramidal cells, which can come from either the ipsilateral or contralateral hippocampus. These synapses are located in the distal dendrites in stratum radiatum and in the basal dendrites of stratum oriens (Amaral and Lavene x, 2007) Each CA3 pyramidal cell receives approximately 12,000 synapses from other CA3 pyramidal cells and subsequently is connected to approximately 2% of the CA3 pyramidal cells between both hippocampi (Amaral et al., 1990; Rolls and Kesner, 2006)
24 MFs and A/C afferents are hypothesized to perform distinct functions when considering the role of CA3 pyramidal cells in memory. The MF CA3 synapse is important for initiating new memories, while the CA3 CA3 synapses form an autoassociative network responsib le for the formation, storage, and recall of episodic memories (Rolls and Kesner, 2006) As an autoassociative network CA3 pyramidal cells can fire in synchrony, increasing the strength of the synaptic connections comprising the network (Bains et al., 199 9) ; however, these recurrent connections also make them vulnerable for the initiation and progression of seizures (Colom and Saggau, 1994; Dzhala and Staley, 2003) In fact, research has shown CA3 can be a source for seizure initiation that spreads to CA1 and the dentate gyrus (Dzhala and Staley, 2003; Wittner and Miles, 2007) Mossy cells Mossy cells are unique hilar interneurons because they are glutamatergic as opposed to GABAergic. Their dendrites are multipolar and extend throughout the hilus but rar ely into the molecular layer of the dentate gyrus. Mossy cell axons project to both the ipsilateral and contralateral hippocampus forming connections with dentate granule cells in the inner molecular layer. These connections are of particular interest be cause mossy cell axons travel in the septo temporal direction and consequently activate granule cells in different subfields from the parent mossy cell. This arrangement also prevents the possibility of a recurrent excitatory loop because MFs form synapti c connections to mossy cells (i.e. a granule cell mossy cell granule cell loop). In fact, the MF synapses are what give mossy cells their name by forming large thorny microsco py. It is believed that mossy cells might receive inputs from as many as 100
25 granule cells while each granule cell only forms synaptic contacts with 1 2 mossy cells (Henze and Buzsaki, 2007) Besides granule cells, mossy cells also form synapses onto hil ar interneurons but it appears they are rarely connected to other mossy cells (Larimer and Strowbridge, 2008) They receive glutamatergic inputs from the perforant path, CA3 pyramidal cells, and semilunar granule cells and GABAergic inputs from other hila r interneurons (Henze and Buzsaki, 2007; Scharfman, 2007; Williams et al., 2007; Larimer and Strowbridge, 2008) Interestingly, mossy cells receive 15 40 times more perisomatic inhibitory inputs from cholecystokinin and parvalbumin expressing interneurons than other hilar interneurons (Acsady et al., 2000) They also potentially receive diverse extrinsic connections including cholinergic afferents from the medial septal nucleus, noradrenergic afferents from the pontine nucleus locus coeruleus, dopaminergi c inputs from the ventral tegmental area, and serotonergic afferents from the raphe nuclei (Amaral and Lavenex, 2007) Of particular interest are the cholinergic inputs from the medial septal nucleus which form synapses on the soma and proximal dendrites of mossy cells implying acetylcholine can play an important role in their synaptic physiology (Deller et al., 1999) While mossy cells are not considered part of the trisynaptic pathway, the septo temporal projection of their axons places them in a unique position to control the output from dentate granule cells. Their longitudinal axons could allow for synchronizing spatially distinct groups of granule cells by providing excitatory feedback. In addition, their connections with interneurons along the same axis can drive local inhibitory interneurons to silence certain groups of granule cells (Henze and Buzsaki, 2007) The connections to and from mossy cells also have profound implications in the etiology of
26 temporal lobe epilepsy. In epileptic models the re is a significant loss of hilar neurons including mossy cells (Buckmaster and Jongen Relo, 1999) and there are several theories describing the implications of losing these cells (Ratzliff et al., 2002) The dormant basket cell theory finds that the dec reased excitatory stimulation of interneurons due to loss of mossy cells results in the disinhibition of granule cells and increased hyperexcitability (Sloviter et al., 2003) while the irritable mossy cell theory claims that the surviving mossy cells ampl ify the hyperexcitable state of granule cells (Santhakumar et al., 2000) Surviving mossy cells from epileptic tissue can also generate spontaneous epileptiform burst discharges (Scharfman et al., 2001) Finally, in epileptic models the decreased connect ions between MFs and mossy cells results in MFs forming new connections with granule cells, termed mossy fiber sprouting, and these synapses contribute to hyperexcitability by creating an excitatory feedback circuit (Wenzel et al., 2000; Gorter et al., 200 1; Buckmaster et al., 2002) Overall, mossy cells are unique hilar neurons that play a diverse role in the synaptic physiology of the dentate gyrus and are implicated in the etiology of epilepsy. Synaptic Transmission This dissertation focuses on the modu lation of synaptic transmission by cannabinoids in the hippocampus. As such this section will provide the relevant background for understanding general synaptic transmission. In addition, initial data acquired for the potential CB1 independent effect of cannabinoids indicate a differential effect on the two pathways for neurotransmitter release: action potential dependent exocytosis and action potential independent exocytosis. Consequently, the second half of this section will detail the differences bet ween these two types of exocytosis
27 There are two primary forms of communication between neurons: electrical synapses and chemical synapses. Electrical synapses can be formed by adjacent cells through gap junctions. Here, proteins known as connexins ar e located in the plasma membranes of both cells and create a pore connecting the cytoplasm of the two cells. These pores are permeable to small ions such as sodium, potassium, and chloride allowing for the passage of current between cells. Through this p athway cells can rapidly transmit electrical signals to each other such as changes in voltage. In contrast chemical synapses are formed by nearby cells whose cytoplasm is not connected. Instead, they communicate through the release of a neurotransmitter which produces a signal in the other cell. This dissertation focuses on cannabinoid modulation of this form of neuronal communication and is described in further detail below. A typical chemical synapse is composed of an axon terminal from a presynaptic c ell and a dendrite, soma, or axon of the postsynaptic cell separated by a small gap termed the synaptic cleft. At the synaptic cleft the presynaptic cell releases a neurotransmitter that subsequently diffuses across the gap to bind to postsynaptic recepto rs resulting in the transmission of a signal between the two cells. In the presynaptic terminal neurotransmitter is stored in globular pieces of the plasma membrane called synaptic vesicles. Under the electron microscope these vesicles are clustered alon g the plasma membrane adjacent to the postsynaptic cell, in an area referred to as the active zone, awaiting release into the synaptic cleft. Release of these vesicles can occur by two different pathways: action potential dependent exocytosis and action potential independent exocytosis (described in detail below) (Fain, 1999) Exocytosis requires the fusion of the synaptic vesicle to the plasma membrane through
28 the interaction of several proteins. The synaptic vesicle protein synaptobrevin must form the 7S complex with two proteins imbedded in the plasma membrane called syntaxin and SNAP 25 in a process called priming. After priming, synaptotagmin binds to the complex and acts as a calcium sensor for fast exocytosis. Following release of neurotransmitt er there is a recycling of the vesicles in a process called endocytosis that are re cycled locally and refilled with neurotransmitter. In contrast, clathrin mediated endocytosis occurs after fusion of the vesicle with the plasma membrane following exocytosis. This pathway also involves a longer process of recycling through endosomes (Fa in, 1999; Sudhof, 2004) Once neurotransmitter is released into the synaptic cleft it can bind to two different classes of postsynaptic receptors: ionotropic receptors and metabotropic receptors. Activation of ionotropic receptors results in the opening of an ion channel that will be permeable to specific ions such as sodium, potassium, and chloride. These receptors are responsible for fast neuronal communication resulting in changes to the membrane potential of the postsynaptic cell. In contrast, metab otropic receptors are GPCRs and activation of these receptors is slower but often longer lasting. These have seven transmembrane regions and when bound by a neurotransmitter they activate a G protein. G proteins are composed of an subunit, a subunit, and a subunit which can activate various ion channels and second messenger cascades resulting in a wide array of possible effects on cellular physiology. There are many different types of neurotransmitters that can be released at chemical synapses to activate these two
29 classes of receptors, and the two most common neurotransmitters in the CNS are the inhibitory neurotransmitter GABA and the excitatory neurotransmitter glutamate (Fain, 1999) Action potential dependent exocytosis A ction potentials are large regenerative depolarizations that neurons use to transmit electrical signals over long distances. They are initiated at the axon hillock, the point of connection between the axon and soma of a neuron. In action potential depend ent exocytosis, an action potential travels down the axon of a cell to the axon terminal where the electrical signal will be converted to a chemical signal through the release of neurotransmitter. The action potential invades the presynaptic terminal and causes a depolarization resulting in the opening of calcium channels. The increase in calcium activates the exocytotic machinery causing the synaptic vesicles to fuse with the plasma membrane and release transmitter into the synaptic cleft. The calcium d ependence of action potential dependent signaling is known as the calcium hypothesis. In fact there is a strong relationship between the calcium influx and the change in the conductance of the postsynaptic cell. It is through action potential dependent e xocytosis that neurons actively communicate to each other. Information is encoded in the frequency of action potentials, converted to a chemical signal, and then translated by the postsynaptic neuron (Fain, 1999) Action potential independent exocytosis A ction potential independent exocytosis is the spontaneous release of neurotransmitter from presynaptic terminals. The original observation of these events from intracellular recordings of the frog neuromuscular junction led to the quantal theory of transm itter release. When the probability of release was reduced by lowering the
30 calcium concentration the endplate potentials, recorded in the postsynaptic cell and produced by presynaptic stimulation, were reduced to an amplitude similar in size to spontaneou s miniature endplate potentials. In addition, it was observed that in low calcium conditions some responses appeared to consist of the sum of two or three miniature endplate potentials. It was reasoned that perhaps there was an irreducible amount of neurot ransmitter that could be released and they referred to this as quanta. This produced the idea that each synaptic vesicle had a specific quantal content. The quantal content, m could be calculated as the product of the number of release sites of neurotra nsmitter, n and the probability of release, p when the probability of release was decreased in low calcium conditions ( m=np ) (Fatt and Katz, 1952; Del Castillo and Katz, 1954; Fain, 1999) For many years quantal transmission, and action potential indep endent exocytosis, was thought to be calcium independent, but research in recent years has brought the validity of this theory into question. For example, in Purkinje cells ryanodine sensitive intracellular calcium transients are responsible for large ampl itude miniature inhibitory postsynaptic currents (mIPSCs) (Llano et al., 2000) In addition, the frequency of mIPSCs can be reduced when the solution is changed to a calcium free media (Li et al., 1998; Hofmann et al., 2006; Yamasaki et al., 2006) To ac count for the different types of release and relative calcium sensitivities, it is believed that there are two calcium sensors regulating exocytosis: a slow sensor with a higher affinity for calcium regulating action potential independent release and a fas t sensor with higher calcium cooperativity regulating action potential dependent release (Sun et al., 2007) Finally, unlike action potential dependent exocytosis which is the active propagation of a signal from one neuron to another, the functional sign ificance of action potential
31 independent exocytosis has been debated due to its spontaneous nature. Recent research has shown this type of exocytosis can regulate protein synthesis in dendrites (Sutton et al., 2004) maintain synaptic connections and rece ptor clustering in the absence of action potential dependent exocytosis (McKinney et al., 1999; Saitoe et al., 2001) and regulate the formation of dendritic spines (McKinney et al., 1999) In addition, spontaneous release also has been implicated in chan ging firing rates of neurons (Carter and Regehr, 2002; Sharma and Vijayaraghavan, 2003) and synchronizing cells responsible for pulsatile hormone release (Popescu et al., 2010) Both action potential dependent and action potential independent release of neurotransmitter can be modulated by receptors located on the presynaptic terminal. Activation of different receptors can either enhance or inhibit both types of exocytosis; however, there is a debate whether inhibition/excitation of the two types of exoc ytosis occurs independently. This stems from the uncertainty about the origin of synaptic vesicles for each pathway, i.e. do they come from the same vesicle pool or separate vesicle pools. Some research has provided evidence that synaptic vesicles releas ed during action potential independent exocytosis do not recycle into the same vesicle pool as those released following stimulation (Sara et al., 2005; Mathew et al., 2008) In contrast, other studies have demonstrated that vesicle recycling following spo ntaneous release populates the same pool as those released following an action potential (Prange and Murphy, 1999; Groemer and Klingauf, 2007) In addition, it is also unresolved whether each type of exocytosis requires the same release machinery (for a g ood review of these issues see Glitsch, 2008; Wasser and Kavalali, 2009) Regardless of the specifics of separate vesicle pools or different release machinery,
32 presynaptic modulation of transmitter release is an important component when studying synaptic physiology because presynaptic receptors can have diverse impacts on how information is transmitted throughout the CNS. Chapters 3 and 4 look at how ECs can presynaptically modulate action potential dependent release to CA3 pyramidal cells and mossy cells respectively, while Chapter 5 focuses on how ECs can alter action potential independent signaling. Production and Breakdown of ECs This final section of the general background provides a brief description of ECs as they will be discussed throughout this dissertation. There are two primary ECs found in the CNS: 2 arachidonoylglycerol (2 AG) and arachidonylethanolamide (anandamide, AEA) (Devane et al., 1992; Stella et al., 1997) Because these are lipid molecules that can pass through the plasma membrane they are thought to be produced in an activity dependent manner through different enzymatic cascades. The primary pathway for the production of 2 AG is from the hydrolysis of inositol phospholipids by phospholipase C into diacylglycerol and a subsequent hydrolysis by diacylglycerol lipase (Stella et al., 1997) AEA is formed when N arachidonoyl phosphatidylethanolamine is hydrolyzed by NAPE PLD (Di Marzo et al., 1994) The breakdown of these two ECs occurs either through hydrolysis or oxidation. The en zymes responsible for the hydrolysis of 2 AG and AEA are monoacylglycerol lipase (MGL) forming arachidonic acid and glycerol and fatty acid amide hydrolase (FAAH) forming arachidonic acid and ethanolamine, respectively (Sugiura et al., 2002; Kano et al., 2 009) In the hippocampus it appears that MGL is located in the axon terminals of granule cells, CA3 pyramidal cells, and a subpopulation of interneurons while FAAH is restricted to the soma and dendrites of the principal neurons (Gulyas et al., 2004) Th e oxidation of these ECs results from either
33 cyclooxygenase (COX) or lipoxygenase (LOX) forming prostaglandins and hydroxyeicosatetraenoates (HETEs), respectively. GABA centric View of Cannabinoids in the Hippocampus Following the initial discovery of cann abinoid involvement in DSI at CA1 pyramidal cells by Wilson and Nicoll (2001) the hippocampus became a hotbed for studying the synaptic physiology of cannabinoids. However, the majority of the new research focused on CB1 at GABAergic synapses with much l ess work devoted to our understanding of CB1 at glutamatergic synapses. This initial focus at GABAergic synapses was due to the lack of staining in pyramidal cell bodies with the CB1 antibody. Instead they were surrounded by immunoreactive fibers suggest ing CB1 was present predominantly at GABAergic terminals (Tsou et al., 1998) In addition, CB1 appeared to localize to cholecystokinin containing interneurons over parvelbumin containing interneurons (Katona et al., 1999; Tsou et al., 1999) The continua tion of studies at GABAergic synapses has led to many new and significant findings about cannabinoids in the synaptic physiology of the hippocampus. In addition to CA1, DSI was found in stratum radiatum interneurons (Ali, 2007) dentate granule cells (Iso kawa and Alger, 2005) and hilar mossy cells (Hofmann et al., 2006; Howard et al., 2007) Cannabinoids have also been shown to be involved in forms of long term plasticity. At CA1 pyramidal cells the induction of DSI allowed normally ineffective trains o f EPSCs to produce LTP (Carlson et al., 2002) Production of ECs following mGluR activation can also produce a form of inhibitory long term depression (iLTD) at CA1 pyramidal cells (Chevaleyre and Castillo, 2003, 2004; Chevaleyre et al., 2007; Edwards et al., 2008; Heifets et al., 2008) and hilar mossy cells (Nahir et al., 2010) In addition to mGluR mediated cannabinoid release, mAChRs were also found to modulate cannabinoid production (Kim et al.,
34 2002; Ohno Shosaku et al., 2003; Fukudome et al., 2004; Hashimotodani et al., 2005; Hofmann et al., 2006) Finally, evidence for cannabinoid tone was demonstrated at mossy fiber associated interneurons in CA3 where these cells were silenced unless a cannabinoid antagonist was applied (Losonczy et al., 2004) While there has been many advances made about our understanding of cannabinoids at GABAergic synapses significantly less is known about how CB1 affects the release of glutamate. Although initial data indicated that CB1 was absent at glutamatergic synapses in the hippocampus many labs still searched for cannabinoid modulation of glutamate (Tsou et al., 1998) The lack of CB1 expression at these synapses was surprising given data showing that cannabinoids could act as anticonvulsants and modulate seizure thr esholds (Wallace et al., 2001; Wallace et al., 2002; Wallace et al., 2003; Blair et al., 2006) Several different labs investigated this discrepancy by studying cannabinoid modulation of neurotransmission from Schaeffer collaterals to CA1 pyramidal cells; however, contradictory data created a controversy over the presence of CB1 at these synapses. Some researchers observed a clear effect of the CB1 agonist WIN55,212 2 that was sensitive to AM 251, a CB1 antagonist, and absent in CB1 / mice (Ohno Shosaku et al., 2002; Slanina and Schweitzer, 2005; Takahashi and Castillo, 2006) while other researchers found a WIN55,212 2 effect that was insensitive to AM 251 and present in CB1 / mice (Hajos et al., 2001; Hajos and Freund, 2002b) Since this initial debat e several discoveries have tipped the scales in favor of the presence of CB1 at CA3 CA1 synapses. The first piece of evidence supporting this conclusion was from the production of new CB1 antibodies with increased detection for CB1 at glutamatergic termin als. These studies found staining for CB1 in stratum radiatum of CA1 suggesting
35 it was present at CA3 CA1 synapses, however it should be noted that there was significantly less intense staining for CB1 at glutamatergic synapses compared to GABAergic synap ses (Katona et al., 2006; Kawamura et al., 2006; Monory et al., 2006) In addition to this finding, several labs produced data showing where the differences in the prior studies might have originated. One possibility for the discrepancies could be due to species and strain differences in the expression of CB1 (Hoffman et al., 2005; Haller et al., 2007) In support of this result it was recently determined that basal levels of endogenous adenosine vary between species, not expression of CB1, and the inter actions between presynaptic adenosine and cannabinoid receptors could account for the controversy over CB1 expression. Tonic activation of adenosine receptors can reduce or eliminate the ability of cannabinoid agonists to inhibit glutamate release through CB1 (Hoffman et al., 2010) Finally, it was also determined that WIN55,212 2 could also be acting in a CB1 independent manner through the inhibition of N type voltage gated calcium channels (Nemeth et al., 2008) This controversy has now been reduced to a simmer but the initial damage was done by discouraging labs from pursuing research on CB1 at glutamatergic synapses. While less contentious than CA1, CB1 has also been found at glutamatergic terminals in the dentate gyrus. Cannabinoid agonists were rep orted to alter the paired pulse facilitation and input/output curve of perforant path stimulation (Kirby et al., 1995) In support of these findings, cannabinoid antagonists blocked the decreases in perforant path stimulation generated by an acetylcholine sterase inhibitor suggesting a cholinergic induced production of cannabinoids to inhibit these inputs (Colgin et al., 2003) A recent study provided a more detailed analysis in dentate granule cells and found that they
36 could produce ECs resulting in DSE. However, this reduction in glutamate release was only found in the inner molecular layer and not the medial and lateral perforant path suggesting CB1 is present at mossy cell axons but not perforant path inputs (Chiu and Castillo, 2008) In comparison to CA1 and the dentate gyrus relatively little is known about cannabinoids at glutamatergic terminals to CA3 pyramidal cells. Early research demonstrated that cannabinoid agonists reduced the power of gamma oscillations, synchronous neuronal activity between 30 100 Hz, from field recordings in the CA3 pyramidal cell layer suggesting cannabinoids could modulate neurotransmission (Hajos et al., 2000) Recent research also indicated that cannabinoid agonists can reduce average firing rates and bursting charact eristics of CA3 pyramidal cells further emphasizing the effect cannabinoids can have in this area (Goonawardena et al., 2010) While these initial studies provide evidence that cannabinoids can modulate glutamatergic signaling in area CA3, they do little to describe these mechanisms at the synaptic level. In support of these electrophysiological findings, recent immunohistochemistry has demonstrated expression of the CB1 receptor at glutamatergic axons in CA3. Using a new C terminal antibody CB1 was foun d at glutamatergic terminals in stratum radiatum but not stratum lucidum in area CA3 suggesting they were present at A/C afferents but not MF inputs (Katona et al., 2006; Monory et al., 2006) These data further emphasize the need for the investigation of CB1 at glutamatergic synapses to CA3 pyramidal cells. Perhaps CB1 at these locations could be responsible for the anticonvulsant effects of cannabinoids because CA3 has been demonstrated as a location for seizure initiation (Dzhala and Staley, 2003; Witt ner
37 and Miles, 2007) In fact, recent research has shown that cannabinoids can decrease synchrony in CA3 pyramidal cells (Goonawardena et al., 2010) and conditional deletion of CB1 from glutamatergic terminals increases the susceptibility to kainic acid induced seizures (Monory et al., 2006) Chapter 3 focuses on providing physiological evidence supporting differential expression of CB1 at glutamatergic terminals in CA3 pyramidal cells while also adding to our minimal understanding of cannabinoids at glu tamatergic synapses in area CA3 and the hippocampus in general. Functional Relevancy of Cannabinoids in the Hippocampus While more is known about CB1 at GABAergic terminals than at glutamatergic terminals in the hippocampus there still remains many unanswe red questions including the functional relevancy of DSI. Traditionally DSI is produced by depolarizing the postsynaptic cell for 3 5 seconds from 70 to 0 mV, a protocol not mimicked by neurons, which brings into question whether DSI occurs in the brain. To address this issue, firing patterns observed in vivo during either spatial exploration or delayed non match to sample were employed but they failed to elicit DSI in CA1 pyramidal cells. Several other types of pulse trains ranging from 5 400 Hz, inclu ding trains at theta frequencies, also were insufficient to induce EC production (Hampson et al., 2003) It was determined through neuronal modeling that to obtain DSI at normal firing patterns in the hippocampus it would require the summation of 2 or mor e convergent synaptic inputs (Zhuang et al., 2005) ; however, this has not yet been shown experimentally. Unlike the hippocampus the functional relevancy of ECs has been well documented in the cerebellum. In Purkinje cells elegant work has demonstrated bri ef stimulation of parallel fibers, within normal firing ranges of these neurons, can produce ECs resulting in DSI with high synapse and spatial specificity (Brown et al., 2003) In
38 addition, pairing of parallel fiber and climbing fiber stimulation greatly enhances the release of ECs, and DSI, indicating their involvement in associative short term plasticity at parallel fibers (Brenowitz and Regehr, 2005) In stellate basket cells, also located in the cerebellum, DSE can be obtained at in vivo firing rates when paired with low levels of activation of mGluR1/5 receptors to lower the calcium requirement for EC production (Myoga et al., 2009) Perhaps a similar pairing of in vivo firing rates with the activation of receptors that enhance cannabinoid productio n could allow for the induction of functionally relevant DSI in the hippocampus. Hilar mossy cells have been shown to release ECs resulting in DSI, and this DSI was enhanced following activation of mAChRs (Hofmann et al., 2006) The interaction between E C production and mAChR activation could provide a viable option for producing functionally relevant DSI in the hilus. Before proposing this possible link in mossy cells we must first consider the general background of the cholinergic system in the hippoca mpus. Cholinergics in the hippocampus The hippocampus receives cholinergic input from the cholinergic basal forebrain, a formation of cholinergic nuclei projecting to areas throughout the cortex. Specifically, the diagonal band of Broca and the medial sep tum innervate neurons within the hippocampus (Frotscher and Leranth, 1985; Niewiadomska et al., 2009) These afferents form contacts at spines, dendritic shafts, and cell bodies at pyramidal cells (CA1,CA2, and CA3), granule cells, and interneurons in str atum radiatum, oriens, lacunosum moleculare and the hilar region of the dentate gyrus (Frotscher and Leranth, 1985) Recent research has shown abundant cholinergic synapses at the soma and proximal dendrites of hilar mossy cells suggesting acetylcholine c ould play a key role in the synaptic physiology of the dentate hilus (Deller et al., 1999) In addition to the
39 cholinergic fibers from the basal forebrain, there is also evidence for a sparse population of cholinergic but non GABAergic interneurons in the hippocampus, however the function of these neurons is still unknown (Frotscher et al., 2000) Release of acetylcholine can activate mAChRs, which are metabotropic receptors that can be located both pre and postsynaptically. There are five subtypes, M 1 M 5 which all have seven transmembrane regions. Typically, M 1 M 3 and M 5 activate phospholipase C, couple to G q/11 and are located postsynaptically, while M 2 and M 4 inhibit adenylyl cyclase, couple to G i/o and are located presynaptically (Volpicelli and Levey, 2004) In addition, mAChRs can also modulate several types of ion channels including I h channels, multiple types of potassium channels, voltage dependent calcium channels (Cobb and Davies, 2005) and their activation can initiate other signaling ca scades including mitogen activated protein kinases and small GTPases (Volpicelli and Levey, 2004) Through these diverse pathways, these receptors have been shown to have a wide array of effects on synaptic physiology in the hippocampus. Linking mAChRs an d ECs for functionally relevant DSI Since cannabinoids are released from the postsynaptic cell we are most interested in changes to the postsynaptic cell caused by mAChR activation when considering a possible link between these two systems for functionally relevant DSI. mAChRs can cause several different changes to the intrinsic properties of neurons, but the most interesting for our studies is the production of an afterdepolarization (ADP). When mAChRs are activated, a brief depolarization can result in a much longer depolarization during which action potentials are often fired. In the hippocampus these ADPs can be found in both CA1 pyramidal cells and local area interneurons (Fraser and MacVicar, 1996; McQuiston and Madison, 1999; Lawrence et al., 2006a ; Lawrence et al., 2006b)
40 Of particular interest, the induction of an ADP requires postsynaptic calcium (Fraser and MacVicar, 1996; Lawrence et al., 2006b) a feature shared with the production of ECs in mossy cells (Hofmann et al., 2006) In addition, brief stimulation to stratum oriens can cause CA1 interneurons to fire action potentials followed by long lasting ADPs (McQuiston and Madison, 1999) This creates a potentially interesting scenario where in vivo firing patterns could induce an ADP resulti ng in a calcium dependent depolarized state that could induce the production of ECs. Mossy cells would be uniquely positioned for such a scenario because they receive somatic and dendritic cholinergic synapses (Deller et al., 1999) however, little is kno wn about potential intrinsic changes following mAChR activation. Chapter 4 investigates whether mAChR activation can result in an ADP in mossy cells and if so, can this ADP act as a physiologically relevant mechanism for the induction of DSI. CB1 Indepen dent Effects of Cannabinoids Previous work in our lab has demonstrated that mossy cells can release ECs resulting in DSI (Hofmann et al., 2006) While investigating whether CB1 at these synapses could modulate action potential independent exocytosis we ca me across some interesting and novel data. Typically activation of CB1 only inhibits action potential dependent exocytosis, and our data indicated an excitatory effect of cannabinoids on action potential independent exocytosis. This data was particularly intriguing because previous research had suggested a possible novel cannabinoid receptor in the hippocampus (Di Marzo et al., 2000; Breivogel et al., 2001) While the most interesting finding would be for a novel cannabinoid receptor there is also a rich and diverse literature on cannabinoid interactions with other types of known receptors and with the structure and fluidity of the plasma membrane (Oz, 2006a) Consequently,
41 Chapter 5 investigates the novel effect of cannabinoids we discovered in an attem pt to determine whether it was through a CB1 independent pathway and if so, the potential mechanism behind this effect. The remainder of this section focuses on providing the relevant background for CB1 independent effects of cannabinoids. Since the ini tial discovery of CB1 and CB2 there has been a continued search for other cannabinoid receptors to account for the effects observed in the presence of antagonists and/or in CB1 / mice. Both AEA and WIN55,212 2 have been shown to stimulate GTP S binding i n CB1 / mice suggesting a possible novel cannabinoid receptor (Di Marzo et al., 2000; Breivogel et al., 2001) In the knockout mice significant activity of cannabinoids was observed in membranes from brain stem, cortex, hippocampus, diencephalon, midbrai n and spinal cord and absent in basal ganglia and cerebellum suggesting a diverse and distinct distribution in the CNS (Breivogel et al., 2001) An additional study discovered that cannabinoids can inhibit the release of [ 3 H]glutamate in the hippocampus o f CB1 / mice (Kofalvi et al., 2003) In fact the controversy over CB1 at Schaeffer collaterals (described above) was briefly attributed to CA3 here is still speculation about novel cannabinoid receptors in the CNS (Kreitzer and Stella, 2009) One potential novel cannabinoid receptor is the peroxisome proliferator activated receptor whose activation leads to the transcription of target genes ofte n associated with lipid metabolism and inflammation. Several endogenous and synthetic cannabinoid ligands such as AEA and WIN55,212 2 can activate this receptor causing an increase in transcriptional activity (O'Sullivan, 2007) Two other families of rec eptors, sphingosp h ine 1 phosphate (S1P) and
42 lysophosphatidic acid (LPA), could potentially be additional novel cannabinoid receptors because they are lipid receptors that share significant homology with CB1 (Kreitzer and Stella, 2009) In support of this hypothesis, S1P was recently discovered to be the first endogenous antagonist for the CB1 receptor (Paugh et al., 2006) Finally, recent research has determined that GPR55, an orphan GPCR, is a novel cannabinoid receptor. Activation of GPR55 by cannabino ids such as AEA induces calcium release from internal stores in an IP 3 and phospholipase C dependent pathway and can inhibit M type potassium channels (Ryberg et al., 2007; Lauckner et al., 2008) There is potential that other orphan GPCRs could also be n ovel cannabinoid receptors, such as GPR3, GPR6, and GPR12, because they share high levels of homology with CB1 (Kreitzer and Stella, 2009) While it is intriguing to consider novel cannabinoid receptors, there is also a large amount of literature demonstra ting cannabinoid interactions with a wide range of known receptors and ion channels. In addition, cannabinoids can also integrate into the plasma membrane due to their lipophilicity causing non specific effects. Cannabinoids can interact with several dif ferent kinds of voltage gated ion channels including sodium, potassium, and calcium channels, and they can also interact with ligand gated ion channels including nicotinic acetylcholine receptors, serotonin receptors, glycine receptors, NMDA receptors, kai nate receptors, and transient receptor potential channels. Cannabinoids can also bind to other GPCRs such as mAChRs and metabotropic serotonin receptors (Oz, 2006a) In the hippocampus, specifically, there are several reports of such interactions includi ng calcium channels (Shen and Thayer, 1998; Nemeth et al., 2008) glycine receptors (Lozovaya et al., 2005) and transient
43 receptor potential channels (Koch et al., 2010) It is important to note that the vast majority of non CB1 effects described above, i n general, require elevated concentrations ( M range) of cannabinoids, and as such could be avoided by using concentrations that would only mediate CB1 dependent effects (for a comprehensive review see Oz, 2006a) An additional source of CB1 independent s ignaling could be receptor independent. Changes to the lipid bilayer have been proven to significantly alter ion channel structure and function. As lipid molecules cannabinoids are endogenous compounds with the potential to integrate into the plasma memb rane resulting in a non specific effect on neuronal signaling (Oz, 2006a) In fact, it was recently demonstrated that changes in the membrane cholesterol content can even influence protein kinase activities (Smith et al., 2010) Finally, it is also worth noting that the breakdown products of ECs can also have pharmacological activity. Prostaglandins, produced by the oxidation of ECs by COX, can abolish DSI, enhance glutamatergic neurotransmission, and augment LTP (Sang et al., 2005; Sang et al., 2006; Yan g et al., 2008; Yang et al., 2009) In addition, production of HETEs by LOX is important for the induction of mGluR LTD in hippocampal slices (Feinmark et al., 2003) Overall there is a wide range of potential CB1 independent effects of cannabinoids whic h will be considered while investigating our novel effect of cannabinoids in Chapter 5.
44 CHAPTER 2 MATERIALS AND METHOD S Hippocampal Slice Preparation Male Sprague Dawley rats between the ages of 17 and 25 were given an intraperitoneal injection of ketamine (80 100 mg/kg) and decapitated using a small animal guillotine. The brain was quickly removed and placed in ice cold artificial cerebral spinal fluid (ACSF, see below for solution) and 300 m thick horizontal slices were cut on a Pelco Series 3000 Vibrat ome (Pelco, Redding, CA). Slices were incubated in a water bath between 30 35C for 30 minutes and then allowed to equilibrate to room temperature for 30 minutes and constantly bubbled with 95% O 2 /5% CO 2 In some experiments either CB1 / or litter match ed wild type C57BL6/J mice aged 17 25 days old were used following the same protocol. These mice were a generous gift from Dr. Carl Lupica of NIDA and are descendents of the line developed by A. Zimmer of NIMH. All protocols were approved by the Univers ity of Florida IACUC and conform to the animal welfare standards issued by the National Institute of Health. Electrophysiology Cells were visualized using an Olympus BX51WI infrared differential interference contrast microscope. Recording pipettes were pu lled for whole cell patch clamp recordings using a Flaming/Brown electrode puller (Sutter P 97, Sutter Instruments, Novato, CA). Typical resistance for the electrode tips was between 3 6 M and were filled with various internal solutions depending on the experiment (see below for ingredients). The access resistance was monitored throughout experiments and typically ranged between 10 and 30 M Recordings were perform ed in either voltage clamp or current clamp using an Axon Multipclamp 700A or
45 700B amplifier (Molecular Devices, Sunnyvale, CA). Data were sampled at 20 kHz, filtered at 2 kHz, and digitally recorded by a Digidata 1322A using Clampex 9/10 (Molecular Devic es, Sunnyvale, CA). In Chapter 3 recordings were performed from CA3 pyramidal cells in voltage clamp at a holding potential of 70 mV. All experiments involving DSE contained 3 M carbachol, a mAChR agonist, and 50 M picrotoxin, a GABA A receptor antagon ist, was used in all experiments except for Figure 3 1. For experiments in Figure 3 1 a concentric bipolar stimulator attached to a stimulus isolation unit (World Precision Instruments, Sarasota, FL) was placed in the dentate granule cell layer to obtain evoked excitatory postsynaptic currents (eEPSCs). For all other experiments involving eEPSCs minimal stimulation techniques were used to isolate either A/C afferents or MF afferents as previously described (Nahir et al., 2007) Using a glass pipette fill ed with ACSF responses were elicited with a 0.1 ms stimulation with a typical intensity <100 A. Responses had a sharp stimulus threshold and were stable with modest additional increases in intensity (see Figure 3 2D). MF responses were isolated in strat um lucidum, showed strong frequency facilitation, and were reduced by at least 60% following bath application of the group II mGluR agonist DCG IV. In contrast, A/C afferents were isolated in stratum radiatum, lacked frequency facilitation, and were insen sitive to DCG IV. In some experiments 2 photon based epifluorescence microscopy (Prairie Technologies, Middletown, WI) was used to place stimulators filled with Alexa 594 for isolation of A/C and MF inputs. This process significantly decreased the time a nd effort in the isolation of MFs. In some cases the glutamate antagonists DNQX (20 M) and APV (40 M) were bath applied to verify that the isolated responses
46 were glutamatergic in origin. Minimally evoked EPSCs (meEPSCs) were evoked at 0.2 Hz except in experiments involving DSE where they were evoked at 0.33 Hz. For DSE experiments, a one minute baseline was obtained followed by a depolarization of the CA3 pyramidal cell from 70 to 0 mV for 5 seconds and a 2 minute recovery period. This constituted o ne set and at least 2 sets were run for all DSE experiments. In Chapter 4 all experiments were performed in the presence of 20 M DNQX and 40 M APV to block glutamatergic transmission. Current clamp was used for ADP experiments and the holding current wa s adjusted so the membrane potential was 60 mV. This adjustment was made before and after application of the mAChR agonist muscarine. To investigate ADPs, a 500 ms depolarizing pulse was applied and adjusted such that a continuous train of action potent ials occurred. The ADP was allowed to last up to 25 seconds before it was terminated by a 1 s, 100 pA hyperpolarizing pulse that was present at the end of each sweep. Two of these depolarizing sweeps were interleaved by sweeps with a 100 pA hyperpolari zing pulse. The intersweep interval was 30 s and this constituted one set of ADP measurements. To investigate changes in input resistance in Figure 4 2 a 500 ms, 40 pA hyperpolarizing pulse was applied before the depolarizing pulse, and immediately foll owing the depolarizing pulse it was applied once every second until the end of the sweep. To prevent action potential contamination of these measurements during the ADP the experiments were also performed in voltage gated sodium channel blocker TTX (1 M) For experiments in Figure 4 5 a concentric bipolar stimulator connected to a stimulus isolation unit was placed in the dentate hilus and responses were evoked in voltage clamp at 0.33 Hz. Current intensity varied between 30 and 100 A and lasted
47 for 0. 1 ms. These experiments used a high chloride internal in order to observe inward currents (see below for ingredients). To investigate the ability of an ADP to induce DSI a mixed mode (voltage clamp/current clamp) protocol was used. First, a 1 min baseli ne of eIPSCs was obtained in voltage clamp followed by a switch to current clamp. Immediately a 500 ms depolarizing pulse was applied in current clamp to induce an ADP and allowed to continue for up to 5 s before the switch back to voltage clamp. At this time eIPSCs were evoked again for 1.5 min and this was one set of DSI. Each cell had 3 5 sets of ADP DSI. Following completion, conventional DSI was tested purely in voltage clamp using a 5 s depolarization from 70 to 0 mV in the same cell. For experi ments in Chapter 5 mossy cells were recorded in voltage clamp at a holding potential of 70 mV. All experiments contained 20 M DNQX and 40 M APV to block glutamatergic transmission. To obtain miniature IPSCs the voltage gated sodium channel blocker TTX (1 M) was used. For experiments in Figure 5 9 evoked IPSCs were generated with either a concentric bipolar stimulator or a glass pipette filled with ACSF (for minimal stimulation) that was placed in the hilus and attached to a stimulus isolation unit. Stimuli were generated at 0.2 Hz with a duration of 0.1 ms. Minimally evoked IPSCs displayed a sharp current threshold and did not change amplitude dramatically with 10 15 A of additional stimulation. For experiments in Figure 5 11, DSI was obtained as described above in Chapter 4 for conventional DSI. Identification of Mossy Cells For experiments in Chapter 4 and Chapter 5, hilar mossy cells were identified using a variety of electrophysiological and anatomical criteria as previously reported (Frazier e t al., 2003; Hofmann et al., 2006; Nahir et al., 2007; Nahir et al., 2010) Briefly, mossy cells typically had a large capacitance (>200 pF) and a high frequency of
48 post synaptic currents often with a large amplitude (> 200 pA) when held at 70 mV. After the experiments, the cells were viewed under fluorescent microscopy. Mossy cells typically were multipolar with multiple large thorny excrescences on the proximal dendrites. Data Analysis Evoked EPSCs/IPSCs For experiments involving evoked or minimally e voked EPSCs/IPSCs, individual sweeps were analyzed for contamination by spontaneous events and were manually eliminated from analysis. We looked for sweeps that showed clean, single peak, short latency responses, and that had a smooth rising phase. To re main in the data set, cells were required to have less than 50% of the sweeps removed. These data were analyzed using ClampFit version 9 (Molecular Devices, Sunnyvale, CA). DSI/DSE To calculate DSI the average amplitude of the 2 sweeps (observed over 6 s) immediately following the depolarization (or ADP) were divided by the average amplitude of the 8 sweeps (observed over 24 s) prior to the ADP or depolarization. For DSE the average amplitude of the 4 sweeps (observed over 12 s) following depolarization w as divided by the average amplitude of the 8 sweeps prior to depolarization. For both DSI and DSE the value was then converted to be expressed as a percent change from baseline. ADP ADPs were measured by taking the area underneath the signal immediately a fter the depolarizing step until the resting membrane potential came within 2 mV of 60 mV. Values reported for ADPs are the average of that measurement taken from two
49 depolarizing sweeps. This analysis was automated using software written in OriginPro 8 (OriginLab, Northampton, MA) by Ben Nahir. Spontaneous and Miniature EPSC/IPSC Detection Spontaneous and miniature EPSCs/IPSCs were detected using appropriate parameters in MiniAnalysis v. 6.03 (Synaptosoft, Decatur, GA), or using similar parameter based event detection software written in OriginC by Dr. Charles J. Frazier. Detected events were then further analyzed using OriginPro 8. Statistics Statistical analysis was performed in either Excel 2003/2007 (Microsoft, Seattle, WA) or OriginPro 8. For anal ysis of spontaneous or miniature EPSC/IPSC frequency and amplitude for individual cells a Kolmogorov Smirnov (K S) test was used. test. Error bars in all figures re present the standard error. Drugs and Solutions External Solutions For our hippocampal slice preparation we incubated our slices in an ACSF containing higher magnesium and lower calcium to prevent excitotoxic death of our cells which consisted of (in mM) : 124 sodium chloride (NaCl), 2.5 potassium chloride (KCl), 1.23 sodium phosphate (NaH 2 PO 4 ), 2.5 magnesium sulfate (MgSO 4 ), 10 D glucose, 1 calcium chloride (CaCl 2 ), and 25.9 sodium bicarbonate (NaHCO 3 ). This solution was also used for experiments in Fig ure 5 10 and is referred to in the text as low calcium solution. For our experiments slices were bathed in an ACSF as follows (in mM): 126 NaCl, 3 KCl, 1.2 NaH 2 PO 4 1.5 MgSO 4 11 D glucose, 2.4 CaCl 2 and 25.9 NaHCO 3 In Figure 4 3, N methyl D glucamin e (NMG) was substituted for NaCl.
50 Internal Solutions For whole cell patch clamp recordings, recording pipettes were filled with various internal solutions depending on the experiment. Below is a list, by chapter, of the ingredients for each internal used (in mM) and all solutions were adjusted to a pH of 7.3: Chapter 3 140 CsMeSO 3 1 MgCl 2 3 NaCl, 0.2 Cs EGTA, 10 HEPES, 4 Na 2 ATP, 0.3 Na GTP, 5 Qx 314 Cl, and 0.063 sulforhodamine 101 or 0.1 Alexa 594 and for experiments in Figure 3 4C 10 BAPTA was added Chapter 4 140 K gluconate, 8 KCl, 0.1 CaCl 2 2 MgCl 2 1 EGTA, 2 Na 2 ATP, 0.3 Na GTP, 10 HEPES, and 0.063 sulforhodamine 101 and for experiments in Figure 4 4A 10 BAPTA was added. In addition a separate internal was used for Figure 4 5 consisting of 90 K gluconate, 55 KCl, 0.1 CaCl 2 2 MgCl 2 1 EGTA, 2 Na 2 ATP, 0.3 Na GTP, 10 HEPES, and 0.063 sulforhodamine 101. Chapter 5 90 or 105 CsMeSO 3 55 CsCl, 1 MgCl 2 0.2 EGTA, 10 HEPES, 2 Na 2 ATP, 0.3 Na GTP, 5 Qx 314 Cl, and 0.063 sulforhodamine 101. Drugs Mo st drugs were made into aliquots and stored at 20 C and diluted prior to experiments. The following lipophilic compounds were mixed in dimethylsulfoxide (DMSO) while all other drugs were mixed in water for storage: WIN55,212 2, WIN55,212 3, AM 251, AM 6 30, forskolin, AEA, JZL 184, capsaicin, capsazepine, and 2 AG. JZL 184 was purchased from Cayman Chemical (Ann Arbor, MI) and EGTA AM was purchased from Calbiochem (San Diego, CA). All other drugs and the salts used for external or internal solutions wer e purchased from Tocris (Ellisvile, MO), Sigma (St. Louis, MO), or Fisher Scientific (Pittsburgh, PA).
51 CHAPTER 3 EXCITATORY AFFERENTS TO CA3 PYRAMIDAL CEL LS DISPLAY DIFFERENT IAL SENSITIVITY TO CB1 D EPENDENT INHIBITION OF SYNAPTIC TRANSMIS SION 1 Introductio n In recent years it has become clear that endogenous cannabinoids (ECs) are released from many neurons in an activity dependent fashion. Subsequent to activity (or depolarization) dependent release, ECs act as retrograde messengers capable of inhibiting transmitter release through activation of presynaptic cannabinoid receptors (Kreitzer and Regehr, 2001a; Ohno Shosaku et al., 2001; Wilson and Nicoll, 2001) When this process occurs at GABAergic terminals it is known as depolarization induced suppression of inhibition (DSI), while a similar effect on excitatory (usually meaning glutamatergic) terminals is referred to as depolarization induced suppression of excitation (DSE). While actions of endogenous cannabinoids have been well documented at both glut amatergic and GABAergic synapses in many areas of the CNS, it is fair to say that, to date, the strong majority of well described EC effects in the hippocampus depend on activation of presynaptic CB1 receptors expressed on GABAergic terminals. These inclu de not only activity dependent short term effects such as DSI (Kreitzer and Regehr, 2001b; Wilson and Nicoll, 2001; Trettel and Levine, 2003; Yanovsky et al., 2003; Isokawa and Alger, 2005; Hofmann et al., 2006) but also longer term inhibition where posts ynaptic metabotropic receptors for glutamate or acetylcholine appear to be intimately involved in EC release (Chevaleyre and Castillo, 2003, 2004; 1 Reprinted with permi ssion from: Hofmann ME, Nahir B, Frazier CJ (2008) Excitatory afferents to CA3 pyramidal cells display differential sensitivity to CB1 dependent inhibition of synaptic transmission. Neuropharmacology 55:1140 1146.
52 Edwards et al., 2006) Cumulatively these results have made a compelling case that ECs play an integral role in regulation of inhibition in the hippocampus. By contrast, it has been somewhat more difficult to determine the role of ECs in modulating excitatory transmission in this area. A lack of immunohistochemical evidence for CB1 on glutamatergic terminals i n the hippocampus has, until very recently, been difficult to reconcile with clear evidence of antiepileptic effects of CB agonists (Wallace et al., 2001; Blair et al., 2006) and numerous (but often conflicting) reports of cannabinoid dependent modulation (or even DSE) of the Schaffer collateral pathway (Hajos et al., 2001; Hajos and Freund, 2002b; Ohno Shosaku et al., 2002; Hoffman et al., 2005; Domenici et al., 2006; Kawamura et al., 2006; Takahashi and Castillo, 2006; Nemeth et al., 2008) Very recentl y, immunohistochemical studies using novel C terminal antibodies have positively identified CB1 on a subset of glutamatergic terminals in the hippocampus, and provided evidence implicating them in modulating the threshold for kainate induced seizures (Kato na et al., 2006; Monory et al., 2006) The current study was motivated by these recent immunohistochemical findings, and designed to directly test the hypothesis that isolated glutamatergic inputs to CA3 pyramidal cells will show differential sensitivity to CB1 agonists. Towards that end, we used both conventional and two photon guided minimal stimulation techniques to examine the effects of CB1 agonists on minimally evoked mossy fiber (MF) and associational/commissural (A/C) projections to CA3 pyramidal cells. Our results indicate that CB1 dependent inhibition of A/C mediated transmission occurs subsequent to both bath application of exogenous agonists and depolarization mediated release of
53 endogenous agonists. By contrast, MF projections to CA3 pyramid al cells appear to lack functional CB1 receptors. Results DSE is Present in Area CA3 In an early set of experiments, we observed that both spontaneous and stimulus evoked EPSCs recorded from CA3 pyramidal cells are subject to DSE. Specifically, in the pre sence of 3 M carbachol (CCh), a muscarinic acetylcholine receptor agonist (mAChR), depolarization of CA3 pyramidal cells from 70 mV to 0 mV reduced both frequency and amplitude of spontaneous EPSCs by 14.5 4.39% and 10.4 1.47% (n=7, p<0.01, Fig. 3 1B ). Similarly, identical depolarization depressed EPSCs generated with a bipolar stimulator placed in the granule cell layer (0.1 msec, 200 900 A) by 23.1 3.88% of baseline (n=15, p<0.01, Fig. 3 1C). DSE of stimulus evoked EPSCs was eliminated by bath application of the CB1 antagonist AM251 and the mAChR antagonist atropine, but was insensitive to simultaneous bath application of the GABA A antagonist PTX and the GABA B antagonist CGP (%change was 0.80 6.19%, n=3, p=.9, Fig. 3 1D; 1.93 3.81%, n=5, p=. 6, Fig. 3 1E; and 16.3 1.62%, n=4, p<.01, Fig. 3 1E; respectively). These results strongly suggest that at least a subset of excitatory afferents to CA3 pyramidal cells are modulated by endocannabinoid dependent retrograde signaling. However, they are insufficient to distinguish reliably between specific types of excitatory inputs, as evidenced by clear activation of recurrent connections via granule cell stimulation in prior studies (Xiang and Brown, 1998; Nicoll and Schmitz, 2005)
54 WIN55,212 2 Selecti vely Inhibits Glutamate Release from A/C but not MF Inputs to CA3 Pyramidal C ell s Via Activation of CB1 Receptors In order to more precisely identify which glutamatergic inputs to CA3 pyramidal cells were susceptible to DSE, we used both conventional and 2 photon guided minimal stimulation techniques to isolate either MF or A/C afferents to CA3 pyramidal cells (see Methods). We found that bath application of the CB1 agonist WIN55,212 2 (5 M ) reduced the amplitude of A/C mediated EPSCs (evoked with a mini mal stimulator placed in stratum radiatum) in 18 of 22 cases (by 34.42 4.67%, p<0.001, Fig. 3 2C). This effect was associated with a clear increase in the coefficient of variation (from 0.31 0.02 to 0.43 0.05, n=18, p=0.01, data not shown) suggestin g that the site of action is likely presynaptic. By sharp contrast, we found that identical application of WIN55,212 2 had no effect on MF mediated EPSCs that were evoked with a minimal stimulator placed in stratum lucidum (%reduction: 8.43 7.57%, n=8, p=0.30, Fig. 3 2C). This differential sensitivity to CB1 agonists was also strikingly apparent in a series of experiments in which two separate minimal stimulators were used to simultaneously isolate both an A/C and a MF input to a single CA3 pyramidal c ell (Fig. 3 2A). Under those conditions, bath application of WIN55,212,2 again clearly inhibited the A/C (stratum radiatum evoked and DCG IV insensitive) but not the MF (stratum lucidum evoked and DCG IV sensitive) EPSCs (% reduction: 50.2 6.09% vs. 18 .1 8.29% respectively, n=6, p = 0.02, Fig. 3 2B). We next determined that WIN55,212 2 mediated inhibition of isolated A/C inputs to CA3 pyramidal cells was blocked by pre incubation with AM 251 (% reduction: 3.37 8.65%, n=6, p=0.71, Fig. 3 2C), sugge sting that the effect is dependent on activation of CB1 receptors. Consistent with that interpretation we noted that WIN55,212 2 had no effect on isolated A/C afferents to CA3 pyramidal
55 cells in CB1 / mice, and yet was still effective on wild type contro ls (%reduction: 11.93 10.63%, n=7 vs. 41.96 6.33%, n=4, respectively, p<.01, Fig. 3 3A). Finally, because a previous study reported an effect of persistently activated cannabinoid receptors on a subpopulation of hippocampal interneurons (Losonczy et al., 2004) we decided to test for an endocannabinoid mediated inhibitory tone at A/C inputs to CA3 pyramidal cells. Our results indicated that minimally evoked A/C mediated EPSCs were insensitive to bath application of 5 M AM 251 (%reduction: 1.10 14. 23%, n=9, p=0.99, data not shown), indicating there is no apparent tonic CB1 receptor activation at these synapses in vitro. Endocannabinoid Mediated Retrograde Signaling at Isolated A/C Inputs to CA3 Pyramidal Cells Depends on the CB1 R eceptor P ost synap tic Calcium I nflux and Activation of mAChRs Next we demonstrated that depolarization of CA3 pyramidal cells in the presence of 3 M CCh reduces A/C mediated meEPSC amplitude by 30.1 2.89% (n=9, p<.01, Fig. 3 4A). This effect is significantly more robust than DSE of spontaneous events (p<0.01) and larger than DSE of EPSCs generated with a bipolar stimulator in the granule cell lay er (p=0.11). Within 90 seconds of depolarization, meEPSC amplitude had recovered to 94.4 3.38%, p=0.12, suggesting there is not a long lasting component to this plasticity. Further characterization of minimally evoked A/C CA3 DSE revealed that it is lar gely blocked by preincubation with AM 251 (%reduction: 8.12 6.37%, n=6, p=0.23, Fig. 3 4B), indicating a dependence on CB1 receptors, and that it is eliminated by application of 10 mM BAPTA via the recording pipette (%reduction: 1.64 5.76%, n=7, p=0.7 8, Fig. 3 4C), indicating a dependence on postsynaptic calcium influx. We further
56 report that DSE of minimally evoked A/C afferents to CA3 pyramidal cells (again in the presence of 3 M CCh) was blocked by bath application of 5 M atropine (control: 31.3 4.60%, atropine: 1.33 6.35%, n=4, p<0.05, Fig. 3 5A). This effect was not due to rundown of the EC system as DSE could be maintained over an identical time period in the absence of atropine (first 4 5 sets: 27.8 2.25%, last 4 5 sets: 26.6 8.08%, n=2, p=0.88, data not shown). Finally, we note that CCh dependent DSE of A/C inputs to CA3 pyramidal cells is not blocked by subsequent bath application of 100 M L NAME, a nitric oxide synthase inhibitor (control: 30.6 4.44%, L NAME: 27.8 3.16%, n=4, p=0.49, Fig. 3 5B). This potentially further distinguishes CCh dependent DSE of A/C inputs to CA3 pyramidal cells from DSI of evoked IPSCs recorded from CA1 p yramidal cells in the presence of CCh, the latter of which has recently been demonstrated to be sensitive to disruption of NO synthesis (Makara et al., 2007) Discussion There are two central conclusions of this study. The first is that recurrent connecti ons between CA3 pyramidal cells express functional CB1 receptors while mossy fiber inputs do not. This conclusion is based in large part on experiments in which meEPSCs were recorded from individual CA3 pyramidal cells. We found, in brief, that mossy fib er mediated meEPSCs were not significantly altered by bath application of a CB receptor agonist, while A/C inputs were strongly inhibited in an AM 251 sensitive manner. Further, the effect of CB1 activation on A/C mediated synaptic transmission was absent in CB1 / animals and yet still present in wild type controls. The second central conclusion of this study is that CB1 positive A/C inputs to CA3 pyramidal cells are subject to DSE that depends on presynaptic CB1 receptors, postsynaptic calcium
57 influx, a nd interestingly, activation of mAChRs. Cumulatively, these results provide clear physiological evidence indicating that glutamatergic inputs to CA3 pyramidal cells have differential sensitivity to EC mediated and CB1 dependent retrograde inhibition. Whil e this is the first study to use physiological techniques to directly examine the EC sensitivity of either A/C or MF inputs to CA3 pyramidal cells, significant effort has been made to characterize EC sensitivity of A/C inputs to CA1 pyramidal cells, often with mixed results. For example, several studies have reported not only that A/C inputs to CA1 pyramidal cells are inhibited by bath application of WIN55,212 2, but also that this inhibition is retained in CB1 / animals suggesting the presence of a novel CB receptor. Further, this effect of WIN55,212 2 in CB1 / animals was noted to be insensitive to AM 251 but antagonized by SR141716A (Hajos et al., 2001; Hajos and Freund, 2002b, a) By contrast, other studies have reported clear sensitivity of A/C inp uts to CA1 pyramidal cells that are both sensitive to AM 251 and absent in CB1 / animals (Kawamura et al., 2006; Takahashi and Castillo, 2006) while others yet have reported species and strain specific differences in EC dependent modulation of Schaffer c ollaterals (Hoffman et al., 2005; Haller et al., 2007) Although the reasons for these discrepancies remain largely unclear, one recent report has provided evidence suggesting that WIN55,212 2 can directly inhibit N type voltage gated calcium channels in a concentration range often used in EC studies (including ours), and that this CB receptor independent effect of WIN55,212 2 may explain some previous reports that were interpreted to indicate presence of a novel CB receptor (Nemeth et al., 2008) In the present study, we noted no consistent evidence of WIN55,212 2 mediated inhibition of MFs, of A/Cs in CB1 / animals, or of A/Cs in slices pre treated with AM 251. Further,
58 it remains unclear whether a direct action of WIN55,212 2 on voltage gated calcium channels could account for the SR141716A sensitive effect of WIN55,212 2 in CB1 knockout animals previously described. On the other hand, in experiments where WIN55,212 2 was effective, we did note some variability in the magnitude of the effect. For exa mple, the average effect of WIN55,212 2 on A/C inputs in the absence of AM 251 varied between 34 and 50% inhibition in our dual stimulation, single stimulation, and wild type mouse experiments; however there was no significant difference between these grou ps when compared on a one way ANOVA. Similarly, the effect of WIN55,212 2 on isolated MF mediated EPSCs in our single stimulation experiments ranged from 39% potentiation to 27% inhibition, but were not significantly different than the null hypothesis on average (8.43 7.57%, n=8, p=0.30, see Results). While such variability is not unexpected in experiments that rely on meEPSCs, we cannot definitively rule out the hypothesis that CB1 receptor independent sites of action for WIN55,212 2 exist in area CA3. It is possible that changes in the cutting solution, incubation procedure, and flow rate such as described by Nmeth et al. (2008) would help to reveal such additional mechanisms of action. However at present our results in CA3 most closely parallel the subset of studies in CA1 that conclude that the CB1 receptor in particular is both necessary and sufficient to mediate the effects of WIN55,212 2 on action potential dependent glutamate release from A/C fibers (Kawamura et al., 2006; Takahashi and Castill o, 2006) Perhaps one of the most unusual features of the A/C CA3 DSE reported here is the apparent high level of dependence on activation of mAChRs. In many original studies of EC mediated retrograde signaling, low concentrations of bath applied CCh
59 we re used primarily with the intention of facilitating EC sensitive spontaneous IPSCs. However, later work in area CA1 of the hippocampus has revealed two specific mechanisms through which postsynaptic mAChRs may facilitate EC signaling. In one mechanism, described by Kano and colleagues, mAChR dependent activation of a calcium sensitive PLC subunit results in greater calcium dependent EC synthesis and release upon robust depolarization (Hashimotodani et al., 2005) However, another mechanism has been sugg ested by a recent study which intriguingly noted that DSI evoked in the presence of CCh has a high dependence on NO mediated signaling, while DSI evoked in the absence of CCh does not (Makara et al., 2007) In the present study we have tested the hypothes is that mAChR dependent DSE might depend on NO by establishing DSE and then bath applying an NO synthase inhibitor. Our results indicated that A/C CA3 DSE, although atropine sensitive, is apparently not NO dependent. Thus further work will be necessary t o fully characterize the mechanism of mAChR action in this system. In closing, it is worth noting that much of the work described above was done during a time when CB1 receptor expression on glutamatergic terminals in the hippocampus was not expected. In fact, until recently, detailed immunohistochemical studies had clearly concluded that CB1 expression in the hippocampus was largely if not exclusively confined to the axon terminals of CCK positive GABAergic neurons (for review see Frazier, 2007) This co mmon viewpoint was dramatically altered very recently when new immunohistochemical studies using novel C terminal antibodies unambiguously revealed CB1 receptor expression on glutamatergic terminals (Katona et al., 2006; Monory et al., 2006) This express ion was indeed lower than on
60 GABAergic terminals, but was eliminated by elegant technology that allowed for creation of CB1 knockouts targeted specifically at glutamatergic neurons (see also Marsicano et al., 2003) Importantly, the CB1 expression on glut amatergic terminals was found to be high in extrapyramidal and extragranular cell layers but was virtually absent in stratum lucidum of area CA3, implying that mossy fiber transmission might be insensitive to CB1 dependent modulation. The potential signif icance of CB1 on glutamatergic terminals was indicated by increased susceptibility of glutamate specific CB1 knockouts to kainic acid induced seizures (Marsicano et al., 2003; Monory et al., 2006) Although the results of these studies were quite striking they still largely lacked direct functional tests of their conclusions at a cellular level, particularly in area CA3. Thus, the significance of the current study is also derived, in part, from its ability to use physiological approaches at a cellular le vel to both confirm and extend conclusions based on recent immunohistochemical advances. We believe that these types of results are rapidly bringing the GABAergic centric view of EC signaling in the hippocampus to an end. In fact, in addition to the resu lts presented here and discussed above there is also new evidence indicating that glutamatergic axon terminals in the inner molecular layer of the dentate gyrus are sensitive to CB agonists, and that hilar mossy cells express CB1 receptors (Monory et al., 2006; Chiu and Castillo, 2008) Further, a synthetic enzyme for anandamide has now been identified in apparently CB1 negative mossy fiber terminals (Nyilas et al., 2008) This finding reinforces the important point that the lack of CB1 mediated effects o n MF inputs apparent in our work does not necessarily rule out other potential roles for MFs in EC dependent signaling. Collectively, these results suggest a strong role for ECs in modulating recurrent
61 excitatory circuits in the hippocampus, a rich future for research on EC dependent modulation of excitability, and real potential to make significant strides towards a better understanding of neurological disorders of both memory and excitability.
62 Figure 3 1. DSE is observed in CA3 pyramidal cells and re quires activation of the CB1 receptor. A : Cumulative probability histograms of the interevent interval (left) and amplitude (right) in a representative cell Depolarization induced a statistically significant decrease in both freque ncy and amplitude of s pontaneous EPSCs [P <= 0.001 in both cases, K S Test]. Insets : Raw traces from the representative cell before (Base) and after (DSE) depolarization. B : DSE of spontaneous EPSCs (sEPSCs) recorded in the presence of 3 M CCh. C: DSE of evoked EPSCs (eEP SCs) stimulated at 0.33 Hz with a bipolar stimulator placed in the granule cell layer. D: AM251 sensitivity of stimulus evoked DSE. Bars in B D indicate a 5 second depolarization from 70 to 0 mV. Insets in C D are averages of 4 8 consecutive sweeps fo r each time period taken from a representative cell. E : Summary plot indicating magnitude of DSE in various experimental conditions. Numbers on/by each bar are n values. *p < 0.01.
63 Figure 3 2. Functional expression of the CB1 receptor on A/C but not MF afferents to CA3 pyramidal cells. A: 2 photon image of a CA3 pyramidal cell indicating typical stimulator placement for dual stimulation experiments. SR, stratum radiatum. SL, stratum lucidum. B : Simultaneous recording of minimally evoked MF (top) a nd A/C (bottom) afferents from another CA3 pyramidal cell clearly demonstrate differential sensitivity to both 5 M WIN55,212 2 and 1 M DCG IV. Insets: Averages of 6 consecutively accepted sweeps during baseline ( black ) and following application of WIN5 5 ,212 2 ( gray ). C : MF input were recorded simultaneously. In other experiments either a MF or an A/C input (but not both) were recorded. D : Representative plot depicting the sha rp threshold found in a minimally evoked isolated A/C afferent. Insets : Overlay of raw data (gray) and average (black) for depicted stimulation intensity. All experiments contained 50 M picrotoxin in the external solution. *p<0.001. **p = 0.02.
64 Figure 3 3. WIN55,212 2 decreases the amplitude of minimally evoked A/C afferents in wild type but not CB1 / mice. A: Representative experiments indicating that 5 M WIN55,212 2 reduced meEPSCs of an isolated A/C afferent in a WT mouse ( left ), and yet f ailed to reduce A/C mediated meEPSCs in a CB1 / mouse ( right ). These responses were isolated in the continual presence of 1 M DCG IV and 50 M picrotoxin. B : Summary plot from experiments with wild type and CB1 / mice. Insets: Averages of 6 consecut ively accepted sweeps during baseline ( black ) and following application of WIN55 ,212 2 ( gray ). *p < 0.01. **p < 0.01.
65 Figure 3 4. DSE is observed in minimally evoked A/C afferents and requires both the CB1 receptor and post synaptic calcium influx. All experiments contained 3 M CCh, 50 M picrotoxin, and 1 M DCG IV in the external solution. A: Normalized average of meEPSCs elicited at 0.33 Hz. Cells were depolarized from 70 to 0 mV for 5 s (bar). DSE was blocked by 5 M AM 251 ( B ) and by 10 mM BAPTA, a calcium chelator, in the internal solution ( C ). D : Summary plot depicting the percent reduction in all experimental conditions. Insets : Average trace for 4 8 consecutive sweeps for each time period from a representative cell. *p < 0.01.
66 Fi gure 3 5. DSE of minimally evoked A/C afferents requires mAChR activation and is unaffected by inhibition of nitric oxide synthase. All experiments contained 3 M CCh, 50 M picrotoxin, and 1 M DCG IV in the external solution. A: DSE is present in cont rol conditions (closed circles) but is blocked after application of 5 M atropine (open circles). B: DSE is present in control conditions (closed circles) and is unaltered after application of 100 M L NAME, a NO synthase inhibitor (open circles). Bars in A B represent cell depolarization from 70 to 0 mV for 5 s. C: Summary plot depicting the effects of atropine and L NAME on DSE. Insets: Average of 4 8 sweeps before (black line) and after (gray line) depolarization from representative cells. *p < 0.05.
67 CHAPTER 4 MUSCARINIC RECEPTOR ACTIVATION MODULATES THE EXCITABILITY OF HILAR MOSSY CELLS TH ROUGH THE INDUCTION OF AN AFTERDEPOLARIZATION 2 Introduction The septo hippocampal cholinergic system has been widely implicated in hippocampal dependent learn ing and memory tasks (Frotscher and Leranth, 1985; Gold, 2003; Hasselmo, 2006) Although extensive effort has been devoted to understanding how cholinergic compounds modulate both cells and circuits in area CA1 and CA3 of the hippocampus (Cobb and Davies, 2005; Lawrence, 2008) comparatively little is known about cholinergics in the hilus. This area of the dentate gyrus is home to an unusual class of local circuit neurons known as hilar mossy cells. Mossy cells are the only hippocampal local circuit neur ons that have a glutamatergic phenotype. They are also unusual in having an axonal projection that runs primarily along the septotemporal axis of the hippocampus, where they likely provide excitatory input to both hilar interneurons and dentate granule ce lls (Amaral and Witter, 1989; Buckmaster et al., 1992; Scharfman, 1995; Jackson and Scharfman, 1996; Larimer and Strowbridge, 2008) These unusual features leave mossy cells well positioned to play a powerful role in modulation of information transfer fro m the dentate to CA3 (Henze and Buzsaki, 2007) and have also promoted significant interest in understanding their specific role in both memory formation and epileptogenesis. However, it is important to note that the hilar region of the dentate gyrus is a lso home to a much more heterogeneous group of more traditional GABAergic local circuit neurons. Attempts have been made to 2 Reprinted with permission from: Hofman n ME, Frazier CJ (2010) Muscarinic receptor activation modulates the excitability of hilar mossy cells through the induction of an afterdepolarization. Brain Res 1318:42 51.
68 categorize GABAergic neurons in the hilus based on general axonal arborization, specific postsynaptic targets, and immunohistochemi cal markers, among other things, yet significant work in this area remains to be done (for recent review see Houser, 2007) Recent evidence has indicated that within the hilar region of the dentate gyrus, monosynaptic connections between GABAergic interne urons and mossy cells are more common than any other pairwise combination (Larimer and Strowbridge, 2008) and further that many GABAergic contacts to mossy cells originate from cholecystokinin r ather than parvalbumin p ositive terminals (Acsady et al., 200 0) Our lab has recently shown that bath application of a muscarinic acetylcholine receptor (mAChR) agonist in the hilus can raise ambient GABA and selectively modulate glutamatergic signaling through activation of presynaptic GABA B receptors (Nahir et a l., 2007) However, we have also noted that identical treatment with cholinergic agonists enhances depolarization induced suppression of inhibition as produced by direct depolarization of mossy cells (Hofmann et al., 2006) These types of findings sugges t multiple sites of action and an overall powerful role for muscarinic systems in the modulation of hilar networks, and yet direct effects of muscarinic agonists on the intrinsic properties of hilar neurons have yet to be carefully examined. In the present study we report that bath application of 5 M muscarine typically depolarizes hilar mossy cells and also activates a muscarinic afterdepolarization (ADP). The ADP observed in mossy cells is likely carried by a calcium sensitive non selective cation channe l (I CAN ), and can lead to endogenous cannabinoid (EC) release and a subsequent inhibition of GABAergic transmission that resembles depolarization induced
69 suppression of inhibition. By contrast, 50% of non mossy cells tested were depolarized by muscarine, but none exhibited a similar muscarinic ADP. Results mAChR Activation Produces an Afterdepolarizati on in Hilar Mossy Cells but not in Other Hilar Neurons In this study we investigated the effect of mAChR activation on the intrinsic excitability of neurons located in the hilar region of the rat dentate gyrus. Mossy cells were patch clamped (see methods) in voltage clamp mode using a K gluconate based internal solution. After briefly recording spontaneous EPSCs, ionotropic glutamate receptor antagonists (2 0 M DNQX and 40 M APV) were bath applied, the recording mode was switched to current clamp, and holding current was adjusted to obtain a resting membrane potential of 60 mV (average I Hold : 4.05 5.57 pA, n=40). Under these conditions we applied a 500 ms depolarizing pulse calibrated to produce a continuous train of action potentials (279 18.8 pA on average, n=40). Following the depolarization, the membrane voltage quickly returned to 60 mV showing no evidence of an ADP (Fig. 4 1A, control). In ord er to examine the effect of mAChR activation, muscarine (5 M ) was bath applied while membrane potential was continuously monitored in current clamp. In 78% of mossy cells tested, bath application of muscarine caused a direct depolarization of majority of those case s (25 of 31) this depolarization was sufficient to induce firing of action potentials. We believe this direct depolarization is likely to depend on muscarinic modulation of a potassium conductance, as comparable bath application of a muscarinic agonist (3 M CCh) had minimal effect on holding current or input resistance in 34 mossy cells voltage clamped with a CsMeSO3 based internal solution (change
70 from baseline: 4.34 3.52% and 2.60 8.43%, respectively, n=34, p>0.05, data not shown). Nevertheless, in virtually all mossy cells tested (regardless of the magnitude of the direct depolarization) bath application of muscarine also resulted in the expression of a robust ADP observed following a 500 ms depolarizing pulse (Fig. 4 1A, muscarine). Over 40 mossy cells tested the average area of the ADP was 101 11.6 mV*s in the presence of muscarine compared to 3.57 0.45 mV*s before muscarine application (p < 0.05, Fig. 4 1C). Although (as noted above) the holding current was adjusted to maintain the membrane potential at 60 mV both before and after application of muscarine, the absolute size of the depolarizing step was the same in both conditions. Further, in a subset of cells tested, we noted that bath application of the general mAChR antagonist atropine ( 5 M ) completely blocked the muscarinic ADP (89.0 27.3 mV*s in muscarine vs. 4.32 1.13 mV*s in muscarine + atropine; n=9, p<0.05). By contrast, across 10 non mossy hilar neurons examined in an identical fashion, 50% were depolarized by 5 mV in respon however, all 10 non mossy hilar neurons examined lacked a muscarinic ADP in response to both moderate (192 28.4 pA) and large (4 29 42.1 pA) depolarizing pulses (Fig. 4 1B, control: 2.23 1.70 mV*s; muscarine: 1.93 0.86 mV*s; n=10, p > 0.05, and higher stim: 1.83 1.02 mV*s, n=7, p > 0.05). There was no difference between non mossy hilar neurons that were depolarized by musca rine and those that were not in terms of whole cell capacitance (109.88 20.64 pF vs. 137.63 22.63 pF), input resistance (213.97 60 mV ( 34.00 5.10 pA vs. 30 25.30 pA). Similarly, while there was some variability
71 among non mossy hilar neurons in firing pattern in response to depolarization, in degree of afterhyperpolariz ation observed after an action potential, and in sag currents observed in response to a hyperpolarizing pulse, none of these features were clearly related to the propensity to be depolarized by bath application of muscarine. Thus significant additional an atomical and immunohistochemical work would be necessary to further sub divide non mossy hilar cells based on susceptibility to mAChR mediated depolarization. Instead, for the remainder of this manuscript, we focused our attention on further characterizin g the robust muscarinic ADP that was uniquely observed in hilar mossy cells. ADP Depends on a Calcium Activated Non Selective Cation Channel As a first step, we asked whether induction of a muscarinic ADP in hilar mossy cells was consistent with opening (a s opposed to closing) of an ionic conductance. Hyperpolarizing steps of 40 pA for 500 ms were applied before, during, and after the ADP in the presence of 1 M TTX (see below for TTX experiments) to prevent action potentials from contaminating the measur ements (Fig. 4 2A). During the ADP there was a significant reduction in the change in voltage produced by the hyperpolarizing step, which returned to baseline levels after the ADP (Fig. 4 2C; pre ADP: 7.72 0.49 mV; ADP: 3.33 0.24 mV; post ADP: 7 .30 1.07 mV, n=4, p<0.05). Although other voltage dependent conductances may contribute to this observation, these data are consistent with the hypothesis that activation of the ADP is associated with a decrease in the input resistance, and thus likely the opening of an ion channel. To further characterize the current involved in the ADP we next investigated its voltage sensitivity in the absence of TTX. After observing an ADP at 60 mV we adjusted the holding current to obtain membrane potentials of 70 mV and 80 mV and compared the area of
72 the ADP at each membrane potential ( current injected for depolarization at each membrane potential was adjusted to evoke a similar number of action potentials ). We found the ADP was present but reduced at both 70 mV and 80 mV when compared to 60 mV (Fig. 4 2C,E; 60 mV: 145 33.2 mV*s; 70 mV: 68.6 12.9 mV*s; 80 mV: 29.4 7.70 mV*s; n=8, p<0.05). To verify this was not due to rundown we again examined the ADP at 60 mV, after having tested the hyperpola rized voltages, and noted no difference from the initial measurements (145 33.2 mV*s vs. 148 31.9 mV*s, n=8, p>0.05). The reduction in the area of the ADP at progressively more hyperpolarized membrane potentials suggests that there is a voltage sensit ive component to the muscarinic ADP observed in hilar mossy cells. Previous research has shown several different currents involved in the production of ADPs throughout the CNS including a TTX sensitive Na + current, an mGluR activated inward current, and a calcium activated non selective current, I CAN (Fraser and MacVicar, 1996; Young et al., 2004; Yue et al., 2005) As was apparent above, we found the muscarinic ADP as observed in hilar mossy cells does not require TTX sensitive Na + channels. For these e xperiments we first obtained an ADP in the presence of muscarine and then examined the effect of 1 M TTX on the ADP. We found that bath application of TTX eliminated the ADP (Fig. 4 3A,C; muscarine: 79.9 20.9 mV*s; muscarine + TTX: 5.5 2.8 mV*s; n= 5; p<0.05), however, this could be due to the reduced depolarization (or possibly reduced calcium influx) produced by the current step in the absence of action potentials. Consequently, we increased the amount of current injected in the presence of TTX (b y an average of 426 32 pA) and were ultimately able to induce an ADP that was not significantly different in area than
73 that observed in the absence of TTX (Fig. 4 3A,C; higher stim: 21.8 12.2 mV*s; n=5; p>0.05). This observation suggests that while T TX sensitive Na + channels may contribute indirectly to the induction of the ADP, they are not required to carry the current. However, as the ADP is clearly depolarizing, we next asked whether the current carried by the ADP is heavily dependent on sodium i ons. To test this, we applied ACSF with NMG substituted for NaCl to create a nominally Na + free solution. We found application of this solution eliminated the ADP (Fig. 4 3B,C; muscarine: 90.9 16.6 mV*s; muscarine + NMG: 1.63 0.253 mV*s; n=5, p<0.0 5). However, in contrast to our observations with TTX, in NMG containing solutions the ADP was not rescued with larger current injections (Fig. 4 3B,C; higher stim: 6.49 3.23 mV*s; n=5, p<0.05). This result implies that a large portion of the inward c urrent observed during an ADP is carried by sodium ions. While closing a K + selective channel may produce robust depolarization, opening of a Na + permeant channel is required to produce an inward current. As such, these data also reinforce the results of the input resistance experiments in suggesting the current that carries the ADP is most likely to depend on opening, rather than closing, of an ion channel Next we sought to examine the calcium dependence of the ADP as observed in hilar mossy cells. Towa rds that end, we placed 10 mM BAPTA, a calcium chelator, into our internal solution. In the presence of BAPTA, we were unable to produce an ADP following muscarine application (Fig. 4 4A,C; control: 0.60 0.16 mV*s; muscarine: 2.12 0.78 mV*s; higher stim: 3.71 1.75 mV*s; n=5, p>0.05). For same day controls we were able to produce an ADP with muscarine absent BAPTA suggesting the agonist was functionally active (Fig. 4 4A,C). Interestingly, although BAPTA internal completely
74 eliminated the muscari nic ADP, it failed to significantly reduce the direct depolarization caused by bath application of 5 M muscarine (change in membrane potential: 15.6 8.61 vs. 11.6 2.59 mV, n=31,5, p>0.05). Collectively, all the results presented above are consistent w ith the hypothesis that the muscarinic ADP as observed in hilar mossy cells is mediated by activation of a calcium activated and non selective cation channel (I CAN ). To test that hypothesis directly, we bath applied the relatively nonselective I CAN antago nist flufenamic acid (FFA, 100 M) (Partridge and Valenzuela, 2000; Ghamari Langroudi and Bourque, 2002; Pressler and Strowbridge, 2006; Pressler et al., 2007) and found that mossy cells are unable to produce an ADP in the presence of both FFA and 5 M muscarine (Fig. 4 4B,C; muscarine: 109 27.3 mV*s; FFA: 5.12 2.03 mV*s; higher stim: 6.86 4.95 mV*s; n=6). ADP Induction Can Result in the Release of ECs After characterizing the ADP, we sought to investigate a possible functional significance for this prolonged depolarization. Previous results have indicated that mossy cells can release ECs in a calcium dependent manner to transiently reduce inhibitory transmission in a phenomenon known as depolarization induced suppression of inhibition (Hof mann et al., 2006; Howard et al., 2007) However, failed attempts to elicit DSI at in vivo firing rates in CA1 have brought into question whether the five second depolarizations typically used to induce EC release are physiologically relevant (Hampson et al., 2003) We reasoned that since the ADP results in a prolonged depolarization, then it could possibly be a physiological mechanism for the release of ECs from mossy cells. To test this hypothesis we switched to a high chloride internal
75 solution so we could observe evoked IPSCs (eIPSCs) at 70 mV in voltage clamp. After observing an ADP in current clamp we switched to voltage clamp, and used a concentric bipolar stimulator to evoke IPSCs. Cells were stimulated at 0.33 Hz for a one minute baseline, fol lowed by a mode switch to current clamp. Once in current clamp we applied a depolarizing pulse (as in previous experiments) to induce an ADP. Three to five seconds after ADP onset the mode was switched back to voltage clamp and IPSCs were evoked for an a dditional 1.5 minutes. Using this approach we found that ADPs terminated after three seconds transiently reduced eIPSC amplitude by 13.5 3.3% (n=3, p=0.05). Further, in a separate group of cells where the ADP was terminated after five seconds evoked IP SC amplitude was reduced by 19.2 1.86% (p<0.01, Fig. 4 5A). By comparison, in the same cells a conventional 5 second depolarization from 70 to 0 mV in voltage clamp reduced eIPSC amplitude by 30.6 3.42% (n=5, p<0.01, Fig. 4 5B). These data suggest that ADP induction may lead to cannabinoid release and CB1 receptor dependent inhibition of evoked GABAergic transmission. In order to further validate that hypothesis, we noted that both ADP induced suppression of inhibition and conventional DSI are abs ent in slices pretreated the CB1 receptor antagonist AM 251 (ADP induced suppression of inhibition: 3.30 1.55%, n=5, p>0.05, Fig. 4 5D; DSI: 7.93 3.87%, n=5, p>0.05, Fig. 4 5E). Importantly, the average area of the ADP observed in the presence of AM 2 51 (66.6 23.1 mV*s) was not significantly different than that which produced transient suppression of evoked IPSCs in the absence of AM 251 (p>0.05). Discussion In the present study we have shown evidence that activation of mAChRs differentially affect s the intrinsic excitability of mossy cells and other hilar interneurons.
76 Application of muscarine often resulted in a direct depolarization of mossy cells, likely through modulation of a Cs + sensitive conductance, and further activated a robust ADP. The ADP observed in hilar mossy cells is carried by a calcium activated and Na + permeant ion channel that has mild voltage sensitivity, and is further blocked by the I CAN antagonist FFA. Further work revealed that activation of an ADP in hilar mossy cells pr oduces transient CB1 receptor dependent suppression of inhibition similar to that produced by a five second depolarization in voltage clamp (DSI). By contrast, only fifty percent of other hilar neurons tested were directly depolarized by muscarine, and on ly mossy cells had a muscarinic ADP. ADPs have been described in many different areas of the CNS including the hippocampus (Fraser and MacVicar, 1996; Jensen et al., 1996; McQuiston and Madison, 1999; Young et al., 2004; Lawrence et al., 2006b) olfactor y bulb (Pressler and Strowbridge, 2006; Pressler et al., 2007) and prefrontal cortex (Haj Dahmane and Andrade, 1996) Research has shown the mechanism responsible for induction of these ADPs can vary. For example, prior studies have implicated both mGlu R activated inward currents (Young et al., 2004) and TTX sensitive Na + channels (Yue et al., 2005) in the induction of an ADP. However, one of the most common mechanisms is the opening of a calcium activated non selective cation current (I CAN ), especially for muscarinic induced ADPs (Fraser and MacVicar, 1996; Haj Dahmane and Andrade, 1998; Lawrence et al., 2006b; Pressler et al., 2007) Similar to other I CAN mediated ADPs, we find that the ADP in hilar mossy cells is sodium and calcium dependent, and is blocked by the I CAN antagonist FFA. In fact, several properties of I CAN could account for the cell to cell variability in size and duration of the ADP we observed. Previous
77 research has shown that t he open probability for I CAN currents depends upon the c ytoplasmic calcium concentration which is regulated in large part by mitochondria (Razani Boroujerdi and Partridge, 1993; Partridge, 1994) Thus, c ell to cell variability in calcium uptake or modulators of mitochondrial function could allow for the variab ility in the area of the ADP (Mattson, 2007) It has also been suggested that phosphorylation by PKA can decrease the amount of time I CAN is in the open state (Partridge, 1994) which could also be responsible for the cell to cell variability. In the p resent study we also implicated the ADP as a possible physiological mechanism for the production of ECs and subsequent inhibition of GABAergic transmission. In our lab we have previously shown that a depolarization in mossy cells can produce ECs in a calc ium sensitive manner resulting in DSI; however, to induce EC production we used a non physiological depolarization ( 70 to 0 mV for 5 s) consistent with previously published reports of DSI (Kreitzer and Regehr, 2001b; Ohno Shosaku et al., 2001; Wilson and Nicoll, 2001; Hofmann et al., 2006) Previous research in the hippocampus was unable to reproduce DSI using in vivo firing rates bringing into question whether DSI could be induced under physiological conditions (Hampson et al., 2003) However, our data implicate an ADP induced depolarization as a possible physiological mechanism capable of producing sustained calcium influx and production of endocannabinoids. Potentially c onsistent with these data, our lab and others have previously demonstrated enhance d DSI following activation of mAChRs (Kim et al., 2002; Ohno Shosaku et al., 2003; Hashimotodani et al., 2005; Hofmann et al., 2006) A recent study in cerebell ar stellate basket cells showed that brief trains of somatic action potentials can result in in creases in dendritic calcium causing the release of E Cs and a
78 transient inhibition of glutam atergic transmission; however, to obtain E C release at in vivo firing rates low level activation of mGluR1/5 receptors was required to lower the calcium requirement of E C release (Myoga et al., 2009) This report provides a possible precedent for the pairing of action potentials with the activation of mAChRs to obtain DSI in mossy cells. In fact, mossy cells are well situated for this scenario because they have cho linergic synaptic connections to their soma and proximal dendrites (Deller et al., 1999) Future studies should investigate the in vivo activation of the ADP to further characterize this possible physiological mechanism for induction of EC dependent signa ling. These experiments could be attempted with either local application or uncaging of acetylcholine to mimic endogenous release. An additional possibility would be to test the hilus for intact cholinergic medial septal inputs using a previously describ ed septo hippocampal parasagittal slicing technique used in both CA1 and CA3 (Toth et al., 1997; Widmer et al., 2006) It will also be interesting to investigate other functional consequences of mAChR activation in hilar mossy cells. Previous research in the olfactory bulb and hippocampus has shown application of muscarine results in decreases in action potential latency and jitter to sinusoidal waves (Lawrence et al., 2006a; Pressler et al., 2007) In addition, mAChR activation increases the firing reli ability at theta frequencies in hippocampal interneurons (Lawrence et al., 2006a) and increases the number of action potentials fired in response to multiple stimuli resulting in increased output of GABAergic transmission in olfactory granule cells (Press ler et al., 2007) Finally, previous research has implicated activation of I CAN in the rising phase of excitotoxicity due to its lack of inactivation (Partridge et al., 1994; Tatsumi and Katayama, 1994; Partridge and
79 Valenzuela, 1999) Because mossy cell s are susceptible to ischemic and excitotoxic death (Freund and Magloczky, 1993; Hsu and Buzsaki, 1993; Magloczky and Freund, 1993) and they have been implicated in theories of epilepsy (Lothman et al., 1996; Houser, 1999; Santhakumar et al., 2000; Ratzlif f et al., 2002; Sloviter et al., 2003) prolonged activation of an ADP, and consequent activation of I CAN could provide a possible mechanism for mossy cell death or the initiation, progression, or maintenance of seizures. Perhaps mAChR activation could pu sh mossy cells past the threshold of excitotoxicity by pairing the activation of the ADP and a reduction in the GABAergic input to mossy cells (due to the production of ECs). Indeed, previous research has implicated I CAN in sustaining cortical seizures ( Schiller, 2004) and antiepileptic drugs have been shown to reduce CCh induced depolarizing plateau potentials in the subiculum (D'Antuono et al., 2007)
80 Figure 4 1. Muscarine can induce an ADP in mossy cells but not interneurons. A, left: A mossy ce ll was held at a membrane potential of 60 mV and a 500 ms depolarizing pulse was applied to evoke a train of action potentials. After the depolarization the cell immediately returned to 60 mV. Middle: Following application of 5 M muscarine an ADP was induced following the depolarizing current which could last for several seconds. Right: The ADP could be blocked by 5 M atropine, a mAChR antagonist. B, left : A hilar interneuron was depolarized resulting in an immediate return to 60 mV. Middle: Af ter application of muscarine, unlike mossy cells, there was no ADP. Right: An ADP could still not be produced with a larger somatic current injection. C: Summary plot showing the average area of ADP following the depolarizing pulse. The three bars left of the hash marks summarize the results of experiments in hilar mossy cells, while the three bars to the right of the hash marks are from hilar interneurons. *p<0.05 compared to baseline. **p<0.05 compared to muscarine. Higher stim was tested in 7 of 10 non mossy hilar neurons.
81 Figure 4 2. ADP depends on opening a voltage dependent channel. A, left: A 500 ms, 40 pA hyperpolarizing pulse was applied before and after the depolarization in the presence of 1 M TTX to investigate changes in input re sistance. Right: An inset of three hyperpolarizing pulses from the representative cell to the left: black represents the pulse before the depolarization, light gray is the first pulse after depolarization, and dark gray is a pulse after the membrane pot ential returned to 60 mV. B: A single cell representation of the effect of voltage on the ADP. The ADP is smaller at more hyperpolarized potentials (middle, 70 mV and right, 80 mV) compared to control (left, 60 mV). C: after an ADP. *p<0.05 compared to pre ADP. D: Summary plot of the area of the ADP at different membrane potentials. *p<0.05 compared to control (before muscarine). **p<0.05 compared to muscarine at 60 mV.
82 Figure 4 3. Induction of the ADP does not depend on voltage gated sodium channels but does require sodium ions. A: A representative cell before (left trace) and after (middle left trace) application of muscarine resul ting in an ADP. When 1 M TTX is applied the ADP is blocked (middle right trace) until the depolarizing current is increased resulting in a smaller ADP (right trace). B: A representative cell before (left trace) and after (middle left trace) application of muscarine resulting in an ADP. When the cell is perfused with a nominally sodium free ACSF containing NMG substituted for NaCl the ADP is blocked (middle right trace) even after increased depolarization (right trace). C: Summary plot of the ADP area under different conditions. *p<0.05 compared to control. **p<0.05 compared to muscarine.
83 Figure 4 4. ADPs in mossy cells depend on calcium and depend on I CAN A: The internal solution was filled with 10 mM BAPTA, a calcium chelator, which blocked the ability to induce an ADP even with a higher stimulation (left, middle left, and middle right trace). A same day control showed an ADP could be produced with the muscarine (right trace). B: A representative cell before (left trace) and after (middle left trace) application of muscarine. The I CAN blocker, FFA (100 M), blocked the ADP (middle right trace) and could not be recovered with higher stimulation (right trace). C: Summary plot of the ADP area under different conditions. For the BAPTA data set either 5 M muscarine or 3 M carbachol was used. *p<0.05 compared to control. **p<0.05 compared to muscarine.
84 Figure 4 5. ADP can cause a release of E Cs resulting in DSI. eIPSCs were evoked with a concentric bipolar stimulator placed in the hilu s at 0.33 Hz in the presence of 5 M muscarine or 3 M carbachol. A: For ADP DSI a mixed mode protocol was used. Following a baseline the cell was switched to current clamp and a 500 ms depolarizing pulse was applied evoking a train of action potentials and resulting in an ADP. Five seconds after the depolarization the cell was switched back to voltage clamp and eIPSCs were evoked again at 0.33 Hz. The ADP resulted in a transient decrease in the eIPSC amplitude (DSI). The inset below the graph is an A DP from a representative cell in current clamp. B: For the DSI protocol, instead of switching to current clamp the cell was depolarized in voltage clamp from 70 mV to 0 mV for 5 seconds resulting in a transient decrease in the eIPSC amplitude (DSI). A B: The insets to the right of the graphs are the average of 8 sweeps for the baseline (black), 2 sweeps for DSI (dark gray), and 8 sweeps for recovery (light gray). C: Summary plot of the amount of DSI for 3 second and 5 second ADPs and conventional DSI (5 second depolarization). *p<0.05. D: ADP DSI was attempted after a 20 minute preincubation in 5 M AM 251, a CB1 receptor antagonist. Induction of a similar ADP no longer resulted in a transient decrease in eIPSCs. Inset below the graph is an ADP
85 f rom a representative cell in current clamp. E: A depolarization from 70 to 0 mV for 5 seconds also did not result in a transient decrease in eIPSCs in the presence of AM 251. D E: The insets to the right of the graphs are the average of 8 sweeps for th e baseline (black), 2 sweeps for DSI (dark gray), and 8 sweeps for recovery (light gray). F: Summary plot of the amount of DSI under each induction protocol in the presence of AM 251.
86 CHAPTER 5 CANNABINOID RECEPTOR AGONISTS POTENTIATE ACTION POTENTI AL INDEPENDENT RELEASE OF GABA IN THE DENTA TE GYRUS THROUGH A C B1 RECEPTOR INDEPENDENT MECHANISM Introduction In 2001, endogenous cannabinoids were identified as retrograde messengers in a form of short term synaptic plasticity known as depolarization indu ced suppression of inhibition (DSI, Ohno Shosaku et al., 2001; Wilson and Nicoll, 2001) In this form of plasticity, endogenous cannabinoids, synthesized postsynaptically, act presynaptically at CB1 receptors to inhibit action potential dependent exocytos is. Since the initial discovery that cannabinoids act as retrograde messengers, CB1 dependent DSI has been discovered at numerous inhibitory synapses in multiple areas of the brain (for example see Trettel and Levine, 2003; Bodor et al., 2005; Zhu and Lov inger, 2005; Hofmann et al., 2006) In recent years, several lines of evidence have also highlighted a previously underappreciated role for cannabinoids in modulation of glutamatergic transmission. Specifically, new antibodies have revealed previously un detectable levels of CB1 receptors in specific glutamatergic terminals (Katona et al., 2006; Monory et al., 2006) while elegant knockout studies have revealed that selective loss of these receptors reduces the threshold for kainic acid induced seizures (M arsicano et al., 2003; Monory et al., 2006) Further, physiological studies from our lab have recently confirmed the selective expression of functional CB1 on glutamatergic terminals in area CA3 predicted by earlier immunohistochemical work (Hofmann et al ., 2008) However, even with this broadened perspective, it is noteworthy that virtually all previously characterized effects of cannabinoids on synaptic transmission in the CNS involve activation of presynaptic CB1 receptors and subsequent reduction in th e
87 probability of action potential evoked transmitter release (for review see Freund et al., 2003; Diana and Marty, 2004; Chevaleyre et al., 2006; Mackie, 2008) In the present study, we describe an effect of cannabinoid ligands that is very different. Sp ecifically, we find that bath application of WIN55,212 2 or anandamide produces a clear increase in the frequency of miniature IPSCs recorded from hilar mossy cells in the rat dentate gyrus, without altering event amplitude, area, rise time, or decay. The effect differs from previously reported effects of cannabinoids on synaptic transmission in many respects. First, it is not mediated by CB1 receptors, CB2 receptors, or vanilloid type I receptors, and is still present in CB1 / animals. Second, it selec tively modulates action potential independent exocytotic events. Third, it promotes exocytosis rather than inhibits it. Fourth, it appears to depend critically on a cannabinoid (CB) mediated increase in calcium concentration rather than an inhibition of calcium influx, and fifth, it is only weakly invoked by postsynaptic stimulation that produces robust DSI. Therefore, we believe this manuscript represents the initial description of a new form of cannabinoid mediated modulation of synaptic transmission. Results WIN55,212 2 Produces a CB1 Receptor Independent I ncr ease in mIPSC F requency We examined the effect of cannabinoid agonists on GABA A mediated miniature inhibitory postsynaptic currents (mIPSCs) recorded from hilar mossy cells voltage clamped at 7 0 mV. GABAergic mIPSCs were isolated by bath application of the ionotropic glutamate receptor antagonists (NBQX, 10 M or DNQX, 20 M ) and APV (40 M ), and the voltage gated Na + channel blocker tetrodotoxin (TTX, 1 M ). In our initial series of experiments, the frequency of mIPSCs (typically < 3 Hz) was enhanced
88 to a mean rate of 15.2 2.65 Hz by bath application of r uthenium red, a secretagogue that facilitates calcium independent exocytosis when applied externally (Trudeau et al., 1996b; Sciancalepore et al., 1998) Under these conditions, we noted that bath application of 5 M WIN55,212 2, a synthetic cannabinoid r eceptor agonist, clearly caused a further increase in mIPSC frequency (to 129 7.31% of baseline, p=0.003), without significantly altering event amplitude, area, rise time, or decay (Fig. 5 1). This data, combined with a lack of effect on postsynaptic me mbrane properties (Fig. 5 2), strongly suggests a presynaptic site of action. Cannabinoid agonists have been reported to produce CB1 mediated inhibitory effects on calcium dependent exocytosis from many different nerve terminals throughout the CNS, includ ing at the GABAergic afferents to hilar mossy cells studied here (Hofmann et al., 2006; Howard et al., 2007) By sharp contrast, the excitatory effect on action potential independent mIPSCs we now report does not appear to be mediated by presynaptic CB1 r eceptors because it is insensitive to 5 M AM251 (a CB1 receptor antagonist), and because it is still present in CB1 / animals (Fig. 5 3A,C). Further we determined that this effect of WIN55,212 2 is not blocked by 1 M AM630, or by 10 M capsazepine, suggesting it is not dependent on CB2 recep tors, or vanilloid type I (TRPV1) receptors, respectively (Fig. 5 3B,C). Cumulatively, these data reveal what appears to be a novel effect of cannabinoids on synaptic transmission in the CNS. However, in order to further substantiate that conclusion, a number of control experiments were necessary. Ruthenium Red Does Not Produce or Enable WIN55,212 2 Mediated and CB1 Receptor I ndepend ent I ncreases in mIPSC Frequency A first series of control experiments was designed to eliminate the hypothesis that the ob served WIN55,212 2 mediated facilitation of mIPSCs had an unexpected
89 dependence on ruthenium red. Towards that end, we demonstrated that ruthenium red mediated increases in mIPSC frequency are stable over a long time period in the absence of WIN55,212 2. Specifically, event frequency measured 10 14 minutes after establishing a stable baseline was 106 5.18% of baseline (n=5, p=0.30) when WIN55,212 2 was not applied (Fig. 5 4A, C). Conversely, we also demonstrated that WIN55,212 2 can produce a similar f acilitation of mIPSCs in the absence of ruthenium red, again with no significant change in amplitude. Specifically, in 7 of 17 cells tested WIN55,212 2 increased mIPSC frequency by 1Hz, achieving an average frequency of 226 28.9% of baseline (p=0.005, Fig. 5 4B C for summary data, Fig. 5 4D E for single cell example). Cumulatively, these data are important because they demonstrate that ruthenium red does not produce, and is not required to observe, the WIN55,212 2 mediated enhancement of mIPSCs appare nt in Figs. 5 1 and 5 3. However, it is worth noting that in the absence of ruthenium red baseline frequency was very low, and success rate (see below) for observing WIN55,212 2 mediated facilitation of mIPSCs of 1Hz was reduced to ~40% from nearly 100%. Specifically, in the remaining cells tested without ruthenium red mIPSC frequency was unaltered (at 103 5.23% of baseline, p=0.58, data not shown) following application of WIN55,212 2. Cumulatively these issu es combine to produce significantly larger WIN55,212 2 mediated increases in normalized frequency with much greater variability in experiments that lack ruthenium red. That is why, in our judgment, use of this compound was advantageous for certain experim ents throughout the manuscript. It is important to note, however, that we were careful to avoid the use of this compound in experiments where its effect on voltage gated calcium channels might have confounded the results (see below).
90 AEA Also Produces C lear CB1 Receptor Independent Increases in mIPSC F requency but the TRPV1 Agonist Capsaicin Does Not We next sought to test the hypothesis that the ability to facilitate mIPSCs is not unique to WIN55,212 2. We reasoned that if the WIN55,212 2 mediated enha ncement of mIPSCs does in fact represent a novel form of cannabinoid dependent signaling, it should also be produced by non synthetic cannabinoid ligands. Indeed, we found that bath application of 0.5 M anandamide (AEA) also produced a robust increase in mIPSC frequency in the presence of ionotropic glutamate receptor antagonists, TTX, and ruthenium red (Fig. 5 5). Like the earlier effect of WIN55,212 2, AEA increased mIPSC frequency without significa ntly altering event amplitude, area, rise time or decay (Fig. 5 5C) Further, we no ted that bath application of 2 M capsaicin a vanilloid type I receptor agonist, failed to produce a similar increase in mIPSC frequency ( 102 16.4% of baseline, n =5 p= 0 .92, data not shown ). This result is consistent with the failure of capsazepine to block the effects of WIN55,212,2 noted earlier and also more broadly with the hypothesis that the effects of WIN 55,212 2 described here are TRPV1 independent In order to confirm that the TRPV1 compounds used in these studies were effective, we used identical solutions to demonstrate that capsazepine effectively blocks capsaicin induced calcium influx as observed in cultured dorsal root ganglion neurons previously loaded w ith Fluo 3 (1 hour incubation in Fluo 3 AM + 20% pluronic acid at 37 C). Specifically we found application of 2 M capsaicin increased baseline fluorescence intensity in cultured DRG neurons to 287 11.3% of baseline (n=7 cells, p=0.03), and that this e ffect was completely eliminated in the same 7 cells by bath application 10 M capsazepine (98.4 8% of baseline, p=0.62, data not shown).
91 Mechanism of WIN55,212 2 Mediated Facilitation of mIPSCs Next we considered three hypotheses regarding the mechanism of WIN55,212 2 and AEA mediated facilitation of mIPSCs. The first hypothesis was that these CB ligands may, due to their highly lipophilic nature, alter membrane dynamics and/or fluidity and thereby produce non specific effects on synaptic transmission. Several lines of evidence argue against this hypothesis. First, although non specific action of CB ligands would be expected to affect all terminals equally, it is apparent that not all GABAergic afferents to hilar mossy cells are responsive to CB ligands in the presence of DNQX, APV, and TTX. Specifically, as reported above, we observed an increase in mIPSC frequency of 1 Hz in 7 of 17 trials in the absence of ruthenium red (Fig. 5 4). Further analysis of this data indicates that there is no differenc e in baseline event frequency (2.8 0.65 Hz vs. 2.7 0.23 Hz, p=0.80), baseline event amplitude (19.7 9.00 pA vs. 19.45 8.29 pA, p=0.86), initial (just after patching) input resistance (183 159 36 pA vs. 136 14.4 pA, p=0.518) among cells that show WIN55,212 2 mediated increases in mIPSC frequency and those that do not (n=7,10 respectively). Thus, some synapses are non responsive. Second, although non specific effects of cannabinoid ligands typically require at least low M concentrations of agonist, we find clear effects on mIPSC frequency at nM doses of both AEA and WIN55,212 2. For example, a robust effect of AE A at 0.5 M is described above and illustrated in Fig. 5 5. Subsequent experiments revealed that bath application of 50 nM WIN55,212 2 increased mIPSC frequency to 131.3 4.1% of baseline (Fig. 5 6A B, n=5, p<0.01), while 250 nM produced a comparable eff ect (Fig.
92 5 6A B, 128.4 6.7% of baseline, n=5, p=0.013). These concentrations are substantially lower than those generally reported to alter membrane dynamics (Lundbaek et al., 2004; Lundbaek et al., 2005; Sogaard et al., 2006; Bruno et al., 2007; Lundb aek, 2008) and yet they produce results that are strikingly similar to those produced by 5 M WIN55,212 2 (Fig. 5 1). Third, although non specific effects would be expected to depend critically on the size and lipophilicity of the agonist, we find very different effects produced by lipid signaling molecules that are structurally very similar. Specifically, we noted that bath application of 2 arachidonoylglycerol (2 AG) at 30 M increased mIPSC frequency to just 114.9 4.31% of baseline (Fig. 5 6C D, n=4, p=0.04), without altering amplitude (96.5 1.53% baseline, p>0.05). Importantly, this ef fect is significantly smaller than that produced by AEA at 0.5 M a 60 fold lower concentration (Fig. 5 6C D, p=0.016), and it is not increased substantially in the presence of a monoacylglycerol lipase inhibitor JZL 184, which slows endogenous breakdown of 2 AG (freq: 120 11.2% of baseline, n=6, p=0.75 vs. 2 AG absent JZL 184, data not shown). Collectively, we believe these observations (a subset of unresponsive synapses, effective doses in the nM range, and apparent preference for AEA over 2 AG) provi de a fairly compelling argument that the effects of CB ligands on mIPSC frequency are not likely to be produced by a non specific interaction with cell membranes. The second hypothesis we considered was that the effects described here are produced by speci fic interactions with cannabinoid ligands and either membrane bound or internal targets that are not G protein coupled receptors (GPCRs). Indeed, there is significant precedent for the interaction of cannabinoid ligands, including AEA, with a
93 wide range o f specific ionophores and other non GPCR targets (for example see recent reviews by Murat Oz, 2006a, b) However, a careful review of this literature reveals that the vast majority of such effects (like those on membrane fluidity) require concentrations o f agonist in excess of 1 M and often in excess of 10 M Thus the experiments with nM concentrations of WIN55,212 2 and AEA described above also preclude most precedents for specific non GPCR targets of cannabinoid ligands. However, we did specifically test the hypothesis that the effects of CB agonists described here were caused by inhibition of the TASK 1 potassium channel. We tested this hypothesis directly because Maingret et al. (2001) reported that TASK 1 has a pharmacological profile similar to t hat described here (although low M doses were used), and because a functional role for TASK channels has recently been demonstrated in a subset of hippocampal neurons (e.g. see Talley et al., 2001; Taverna et al., 2005; Torborg et al., 2006) Our results indicated that WIN55,212 2 still produced robust increases in mIPSC frequency in the presence of the TASK 1 antagonist bupivacaine (20 M 175 23.8% of baseline, n=7, p=0.02, data not shown). Further, we verified the effectiveness of the antagonist by showing it produced a significant increase in input resistance measured in CA1 pyramidal cells voltage clamped at 55 mV (from 172 23.7 to 207 38.4, n=5, 1 tailed p=0.04 ). Thus the effects of WIN55,212 2 on mIPSCs described here do not appear to be TASK 1 dependent, and again, occur at concentrations lower than typically shown to interact with other known non GPCR targets. Based on the above results, we next sought to provide a more direct test of a third hypothesis, that activation of an as yet unidentified GPCR underlies the effect s of WIN55,212 2 and AEA on mIPSCs. Towards that end we examined the effect of bath
94 applied WIN55,212 2 (5 M) on mIPSCs (absent ruthenium red) in slices that had been preincubated for 20 minutes in 20 M suramin, a compound that has previously been shown to effectively uncouple G proteins from their effectors (Beindl et al., 1996; Freissmuth et al., 1996; Kiss, 2005; Wang and Hatton, 2007) Our results indicated that success rate dropped from 41% (7 of 17) in control conditions to 0% (0 of 10) in slices treated with suramin. Similarly, average mIPSC frequency after application of WIN55,212 2, and averaged across all cells, dropped from 154 19.1% of baseline (n=17) in control conditions to 100 20.3% of baseline (n=10) in suramin (Fig 5 7A,C ). Thus, WIN55,212 2 mediated facilitation of mIPSCs is sensitive to suramin. However, in addition to its effects on GPCR dependent signaling, suramin has been reported to be an effective P2X receptor antagonist (McLaren et al., 1995; Bultmann et al., 1996) The refore we also tested the effects of another non selective P2X receptor antagonist, PPADS (20 M), on WIN 55,212 2 mediated facilitation of mIPSCs. Slices were pretreated with PPADS for 20 minutes in a manner identical to the suramin experiments, and WIN55,212 2 was again tested absent ruthenium red. Our results indicated that WIN55,212 2 still ca uses a clear increase in mIPSC frequency in the presence of PPADS. Specifically, across all 12 cells tested, average frequency increased to 169 34% of baseline following application of WIN55,212 2 ( Fig. 5 7B,C, n=12, p=0.07). Further, examined individu ally, WIN increased mIPSC frequency in 5 of 264 60% of baseline (p=0.05). These results indicate a statistically significant effect of WIN55,212 2 on mIPSC frequency in the presence of PPADS that is nevertheless not significantly different from the effect observed in control conditions (p=0.54).
95 In order to further evaluate the GPCR hypothesis, we next asked what type of signaling cascade might plausibly underlie WIN 55,212 2 mediated facilitation of mIPSCs. Previous work in hippocampal cultures has suggested that protein kinase A (PKA) can increase release probability independent of changes in calcium concentration (Trudeau et al., 1996a) Therefore we hypothesized that a GPCR G s adenylyl cyclase (AC) cyclic adenosine monophosphate (cAMP) PKA dependent signaling cascade might be involved. Consistent with that hypothesis, we found that preincubation of slices for 1 hour in 10 M H 89 (a selective PKA inhibit or) eliminated WIN55,212 2 mediated potentiation of mIPSCs (Fig. 5 8A, n=6, p=0.72 for frequency in WIN55,212 2 vs. control). Further, we noted that bath application of forskolin (20 M which raises cAMP via activation of AC) not only produces a similar potentiation of mIPSC frequency as observed in hilar mossy cells (to 176 22% of baseline, n=5, p=0.03), but also occludes further potentiation by WIN55,212 2 (Fig. 5 8B, n=5, p>0.05). These results lend additional strength to the hypothesis that activ ation of an as yet unknown GPCR is responsible for WIN55,212 2 mediated facilitation of mIPSCs, and at this point we believe this is the most likely mechanistic explanation. However it is important to note that in a last series of experiments on this topi c we examined the effects of WIN55,212 3, an enantiomer of WIN55,212 2 that is functionally inactive at CB1 receptors (Kuster et al., 1993; Al Hayani and Davies, 2002; Blair et al., 2006) Somewhat surprisingly, we found bath application of 5 M WIN55,21 2 3 produced clear increases in mIPSC frequency (to 137 13.7% of baseline, n=8, p=0.03) without altering amplitude (100 2.3% of baseline, n=8, p=0.89). This effect is essentially identical to that produced by 5 M WIN55,212 2 in Figure 5 1. At 50 nM, the apparent
96 effect of WIN55,212 3 was more variable and somewhat slower to develop, but was still comparable in overall magnitude (data not shown). While ineffectiveness of WIN55,212 3 in these experiments would have strengthened the argument for a spec ific ligand receptor interaction, lack of stereoselectivity in a functional assay, although quite unusual, does not directly diminish it. Please see the discussion for a more detailed consideration of this unexpected observation. WIN55,212 2 Has No Effe ct on Action Potential Dependent Exocytosis in the Presence of the CB1 Receptor A ntagonist AM251 Although the nature of the experiments described above suggests a selective effect on action potential independent exocytosis, we reasoned that if action poten tial mediated exocytosis was also subject to CB1 receptor independent potentiation by WIN55,212 2, the effect might have been overwhelmed in prior studies by CB1 receptor mediated inhibition. In order to test this possibility, we examined the effect of ba th application of WIN55,212 2 and, in separate experiments, AEA, on evoked IPSCs in slices pretreated with AM251. Our results indicate that WIN55,212 2 had no significant effect on either eIPSC amplitude or paired pulse ratio under these conditions (Fig. 5 9A B). Similar results were also obtained when examining the effect of WIN55,212 2 on minimally evoked IPSCs in the presence of AM251 (Fig. 5 9C). Collectively, t hese data are consistent with the hypothesis that CB agonists selectively facilitate mIPSC frequency via CB1 receptor independent modulation of an exocytotic release process that is mechanistically distinct from conventional action potential dependent exocytosis. Interestingly, a number of recent studies have provided evidence from other termi nals which suggests that miniature and evoked synaptic responses are mediated by release from distinct vesicle pools that may have distinct calcium sensitivity (for review see
97 Glitsch, 2008; Wasser and Kavalali, 2009) Therefore, we next attempted to deter mine whether there is any role for calcium ions in WIN55,212 2 mediated and CB1 receptor independent facilitation of mIPSCs in the present system. Calcium is Required for WIN55,212 2 Mediated and CB1 R e ceptor I ndepende nt Increases in mIPSC Frequency We u sed five different experimental approaches to manipulate the availability of calcium ions, and in each case, we tested the ability of bath application of WIN55,212 2 to increase mIPSC frequency. Two of these manipulations, bath application of low calcium external solution or Cd 2+ ions, were designed to reduce availability of external calcium. Two other manipulations, prior incubation with BAPTA AM or EGTA AM, were designed to chelate calcium intracellularly throughout the slice, although at notably differ ent rates. One manipulation, addition of 10 mM BAPTA to the internal solution, was designed to selectively chelate calcium in the postsynaptic cell without affecting calcium availability elsewhere in the slice. Our results indicate that both reduced ava ilability of external calcium and global chelation of internal calcium dramatically reduce the ability of WIN55,212 2 to facilitate mIPSC frequency, while selective chelation of calcium in the postsynaptic cell does not (Fig. 5 10). However, a detailed co nsideration of these data can be considerably more complicated than that. It is important to note that all of these experiments were conducted in the absence of ruthenium red. This choice eliminated concern about unanticipated effects of ruthenium red on calcium homeostasis in general or voltage gated calcium channels in particular, but also meant that we had to contend with a control dataset (as presented in Figure 5 4B, and Fig. 5 10A B) that had both a low baseline frequency and greater variability in response to WIN55,212 2. Because of
98 these features, a careful analysis of all experiments that altered calcium availability required a consideration of changes in both population effect and success rate. As above, success rate was defined as the percenta ge of experiments (in each experimental group) in which bath application of WIN55,212 2 caused a sustained increase in baseline frequency of 1Hz. The average baseline frequency across all groups was 2.3 0.15 Hz, n=55. For a less stringent measure of success rate, see the inset panel in Fig. 5 10D, and Fig. 5 10 legend. The population effect was defined as the average change in baseline frequency across all runs (including both successes and failures). When these variables are considered independent ly, it is clear that global manipulations of calcium availability reduced both the population effect and the success rate (Fig. 5 10C). For example, there was no significant effect of WIN55,212 2 on mIPSC frequency (measured across all cells) in low calci um external, in 200 M Cd 2+ or following incubation in BAPTA AM. However, significant increases in mIPSC frequency (again measured across all cells) were observed in control conditions, and in cells backfilled with 10 mM BAPTA. A smaller, but statistica lly significant increase in mIPSC frequency was noted in cells preincubated in EGTA AM. Similarly, success rate (measured by the 1 Hz rule) ranged from a high of 56% in cells backfilled with BAPTA (41% in the control group) to a low of 15% in low Ca 2+ e xternal solution. Importantly, success rate was lower than control in every experimental group that involved a global manipulation of calcium availability (Fig. 5 10C). While these results are quite striking, in our view, an independent analysis of eit her population effect or success rate is potentially misleading and runs the risk of
99 underestimating the overall loss of effectiveness of WIN55,212 2. Therefore we next defined the effectiveness of WIN55,212 2 in each experimental group as the product of the success rate and the population effect. Thus, as the success rate approaches 0, so does the effectiveness of WIN55,212 2. By contrast, as success rate approaches 100%, the effectiveness of WIN55,212 2 approaches the population effect. We believe thi s is not only a reasonable approach to the data at hand, but that it in fact provides the cleanest and most accurate view of the overall effect of our various calcium manipulations. The results of this analysis are presented in Fig. 5 10D, where it become s apparent that the overall effectiveness of WIN55,212 2 is reduced to < 20% of control conditions by most global manipulation of calcium (31% of control in EGTA AM), and yet is comparatively unaffected by chelation of calcium exclusively in the postsynapt ic cell. In our view, this data strongly suggests that there is a key role for calcium ions, likely in the presynaptic terminal, in producing WIN55,212 2 mediated, and yet action potential independent, facilitation of mIPSCs. Some more detailed questions regarding interpretation of this dataset are considered in the discussion. Depolarization of Hilar Mossy Cells Sufficient to Produce Robust Activation of CB1 Receptors Has Only Minimal Effects on mIPSC F requency Previous work from our lab, and others, has demonstrated that depolarization of hilar mossy cells produces robust CB1 receptor dependent depolarization induced suppression of inhibition (DSI, Hofmann et al., 2006; Howard et al., 2007) Therefore, in a final series of experiments, we sought to dete rmine whether release of endogenous cannabinoids sufficient to induce DSI would also increase mIPSC frequency. To accomplish that goal, mossy cells were voltage clamped at 70 mV with a high Cl internal solution, and evoked IPSCs were generated using a b ipolar stimulator in the
100 presence of ionotropic glutamate receptor antagonists and carbachol. After recording a stable baseline, mossy cells were depolarized from 70 mV to 0 mV for five seconds (2 5 depolarizations delivered on 3 minute intervals) to tes t for DSI. Cells that demonstrated robust DSI (nearly 100%) were considered to be clearly releasing endogenous cannabinoids in response to depolarization. These same cells were then exposed to TTX (1 M ), and ruthenium red (50 M ) as in Fig. 5 1 of the present manuscript. Under these conditions, DSI positive mossy cells were given an identical series of depolarizations while we monitored mIPSC frequency and amplitude. Our results indicate that depol arization of hilar mossy cells sufficient to produce robust DSI (to 45.2 4.7% of baseline, n=6, p<0.001, Fig. 5 11A), also caused a minimal but statistically significant increase in mIPSC frequency (to 106 2.3% of baseline when measured 0 12 seconds af ter depolarization, n=6, p=0.04), without altering mIPSC amplitude (101 1.45% of baseline, n=6, p=0.48, Fig. 5 11 B C). Overall, these results indicate that the retrograde messenger responsible for DSI is in fact likely able to activate the novel CB lik e receptor described here. However, the fact that depolarization induced increases in mIPSC frequency are much more modest than those observed following bath application of either WIN55,212 2 or AEA suggests that endogenous activation of this system may b e more closely tied to an endogenous ligand different from that which mediates DSI, and that release of such ligand may optimally be derived from a different source (than mossy cells), or depend on a different signal (than postsynaptic calcium influx). In terestingly, we also noticed that mIPSC frequency was still enhanced (to 107 2.26% of baseline, n=6, p=0.04) when measured 1.5 2.0 minutes after depolarization. This finding indicates that although modest in magnitude,
101 depolarization induced increases i n mIPSC frequency are in fact longer lasting than DSI. Discussion In the present study we describe a novel effect of the cannabinoid receptor ligand WIN55,212 2 in the CNS. Specifically, we find this ligand can increase the frequency of miniature IPSCs re corded from hilar mossy cells without altering amplitude, area, rise time, or decay. This effect is not blocked by CB1, CB2, or TRPV1 antagonists, is still present in CB1 / animals, and yet can also be produced by AEA. Further, this effect is blocked by an agent that uncouples GPCRs from their effectors (suramin), is blocked by a selective inhibitor of PKA (H 89), and is mimicked by forskolin induced activation of AC. Importantly, forskolin mediated potentiation of mIPSCs also prevents further increases from being induced by bath application of WIN55,212 2. Interestingly, we also noted that this novel effect of CB agonists is highly restricted to modulation of action potential independent exocytotic mechanisms, and yet still retains a high degree of calc ium dependence. Finally, we report that depolarization of hilar mossy cells sufficient to produce CB1 dependent inhibition of action potential evoked release (DSI) produces only a very small, but apparently long lasting increase in mIPSC frequency. Overa ll, the present results represent identification of a novel effect of cannabinoids on synaptic transmission that is quite distinct from previously described phenomena. One obvious question of extreme interest is the specific mechanism through which canna binoids enhance mIPSC frequency. In the present study we presented several lines of evidence that argue against an essential non specific action of these lipophilic ligands on membrane dynamics and/or fluidity. First, we found that success rate for obser ving cannabinoid mediated facilitation of mIPSCs was < 100% in control
102 conditions, implying that there are some negative synapses, even among inhibitory inputs to hilar mossy cells. Second, we found that nM concentrations of both WIN55,212 2 and AEA produ ce clear increases in mIPSC frequency. By contrast, low M (or higher) concentrations are typically required to produce non specific effects. Third, we found a high degree of agonist specificity, apparent in the bath application experiments with 2 AG and AEA (Fig. 5 6) and in the weak effect of depolarization on mIPSCs observed in cells that expressed robust DSI (Fig. 5 11). Collectively, these data make a strong argument that a specific ligand receptor interaction is involved. In fact, the effectiveness of sub M concentrations of cannabinoids also argues to a large extent against most previously identified specific interactions of cannabinoids with non GPCR targets (for review see Oz, 2006a, b) However, additional work also provided more direct evidence that the effects of CB ligands described here may in fact be GPCR dependent. Specifically, we found that inhibition of G protein dependent signaling by suramin eliminates WIN55,212 2 mediated facilitation of mIPSCs while another non selective P2X receptor antagonist, PPADS, does not. Further, additional wo rk with H 89 and forskolin produced results that are consistent with the more specific hypothesis that a GPCR G s AC cAMP PKA dependent signaling cascade is involved. While we believe the data presented here are sufficient to conclude that such a signaling cascade is likely involved in CB mediated potentiation of mIPSCs, we appreciate that definitive proof of the hypothesis will require identification of the specific gene product for a new (or possibly known) GPCR. In terms of the potential can didates, it is worth reiterating that multiple lines of evidence (insensitivity to AM 251, presence in CB1 / animals, preference for AEA over 2 AG, weak response to depolarization evoked
103 release of endocannabinoids, facilitatory effect on exocytosis, sele ctive modulation of action potential independent release events, and lack of stereoselectivity for WIN55,212) all indicate that the phenomenon described here does not depend on CB1Rs, while straightforward pharmacological experiments have also ruled out CB 2 and TRPV1. Indeed, the last feature in the list above, apparent lack of stereoselectivity for WIN,55,212, was quite unexpected and deserves particular consideration here. At face value, lack of stereoselectivity is inconsistent with the hypothesis that a specific ligand receptor interaction is involved, and thus it may be considered our sole contradictory observation. On the other hand, in our opinion this observation does not currently outweigh the multiple other arguments in favor of a GPCR dependent mechanism. In fact, we believe there are at least two other plausible explanations for the apparent lack of stereoselectivity we observed that are still consistent with the overall GPCR hypothesis. For example, one potential explanation is that a functi onal assay that involves both measurement of mIPSCs and bath application of sub M concentrations of lipophilic compounds is simply not clean or precise enough to reveal stereoselectivity that would otherwise be apparent in an appropriate concentration ran ge. In this event, it may be possible to reveal stereoselectivity in a binding assay once the target protein is known. However, another possibility that should not be dismissed is that the ligand protein interaction underlying this phenomenon truly lack s strong stereoselectivity. While at present, there is no clear reason to favor one of these possibilities over the other, it is important to note that lack of stereoselectivity in a specific ligand receptor interaction is not without precedent. For exam ple, metabotropic lysophosphatidic acid (LPA) receptors are equally activated by stereoisomers of LPA
104 despite recognizing other ligands in a stereoselective manner (for review see Yokoyama et al., 2002; Tigyi and Parrill, 2003) Intriguingly, LPA itself i s a phospholipid derivative with structural similarity to AEA and some metabolic overlap with 2 AG. Thus the lack of stereoselectivity observed here may be an intriguing clue that LPA, s phingosine 1 phosphate and related receptor systems should be consid ered as potential candidates for modulation of action potential independent exocytosis. For relevant overview on these systems and a broader consideration of the potential for novel cannabinoid receptors see Kreitzer et al. (2009) While to our knowledge this is the first study to reveal such clear CB1 receptor independent effects of cannabinoids on action potential independent exocytosis, it is important to note that t wo prior studies have described an ability of CB agonists to stimulate [ 35 S]GTP S binding in brain membranes from CB1 / mice (Di Marzo et al., 2000; Breivogel et al., 2001) Strikingly the only effective agonists, of many tested, were WIN55,212 2 and AEA. Further, the effect of these ligands on [ 35 S]GTP S binding was largely insen sitive to the CB1 receptor antagonist SR141716A, was noted in hippocampal as well as whole brain preparations, and overall, like ECs and fatty acid amide hydrolase, had a distribution distinct from CB1 and CB2. It seems plausible, at least, that the GPCR described in these studies could be the same one involved here. By contrast, we do not believe the recently discovered novel cannabinoid receptor, GPR55, is likely to be involved in our experiments because prior work has shown that this receptor does not bind, and is not activated by, WIN55,212 2 (Ryberg et al., 2007; Lauckner et al., 2008)
105 Another issue that requires particular consideration is the role of calcium ions in WIN55,212 2 mediated increases in mIPSC frequency. Although it has been over 50 ye ars since action potential independent miniature synaptic events were first noted at the neuromuscular junction, until recently such transmission was commonly assumed to be calcium independent. Recent evidence has strongly challenged that hypothesis on se veral levels. For example, although the finding is not universal, an increasing number of studies have now indicated that removal of external calcium and/or chelation of internal calcium can effectively reduce the frequency of miniature postsynaptic curre nts in a variety of preparations (Li et al., 1998; Braga et al., 1999; Hofmann et al., 2006; Yamasaki et al., 2006) Similarly, large miniature events at specific cerebellar synapses have been linked to spontaneous calcium sparks that depend on presynapti c calcium stores (Llano et al., 2000) Further, at the Calyx of Held elegant recent work suggests that distinct calcium sensors underlie calcium dependent modulation of synchronous and asynchronous release (Sun et al., 2007) while other noteworthy studie s have investigated the possibility that action potential independent exocytosis draws largely on a distinct pool of presynaptic vesicles with unique sensitivity to calcium induced exocytosis (Sara et al., 2005; Groemer and Klingauf, 2007; Wasser and Kaval ali, 2009) Indeed, the results presented here are quite consistent with the idea that calcium ions can effectively regulate the rate of action potential independent exocytosis. Specifically, we find that WIN55,212 2 mediated facilitation of mIPSCs is s everely reduced by multiple manipulations that globally reduce availability of calcium ions, but is resistant to chelation of calcium postsynaptically. Thus our primary conclusion in this area is that WIN55,212 2 mediated facilitation of mIPSCs requires c alcium. It is unclear
106 at this point whether an increase in calcium is involved, or merely a minimum basal level is required. Further, an additional caveat here, that also applies to any specific GPCR involved, concerns the site of action. It is clear t hat calcium is not required postsynaptically. In the simplest case any GPCR involved would be expressed presynaptically, and any required calcium would be present in the presynaptic terminal. However, we cannot yet definitively rule out the possibility t hat a GPCR mediated increase in calcium concentration in some cell other than the one patched could ultimately modulate synaptic transmission via an unknown messenger. Indeed, there is precedent available for cannabinoid mediated and calcium dependent rel ease of extracellular messengers from non neuronal cells, although the presence of ionotropic glutamate receptor antagonists in all of our experiments rules out the specific mechanism described in that work (Navarrete and Araque, 2008) Another set of very interesting questions raised by these findings has to do with the source / mechanism of release of an endogenous ligand capable of increasing action potential independent exocytosis. In the present study, we report that WIN55,212 2 mediated facilitation of mIPSCs is mimicked by bath application of anandamide, but that depolarization of DSI positive hilar mossy cells produces only a very modest increase in mIPSC frequency. While it has not been determined specifically at this synapse, the majority of avai lable data suggests that 2 AG is likely to be the retrograde messenger in most depolarization induced (and calcium) dependent forms of DSI or DSE in the CNS. By contrast, AEA is implicated more strongly in other forms of cannabinoid mediated plasticity, i s only a partial agonist for the CB1 receptor, and is much less abundant in the CNS (Hillard, 2000; Di Marzo et al., 2002; Sugiura et
107 al., 2002; Chevaleyre et al., 2006) Further, several lines of evidence have suggested that one or more non CB1 non CB2 A EA specific receptors remain to be identified (for review see Di Marzo et al., 2002; Di Marzo, 2006; Sugiura et al., 2006) Indeed, additional work presented here has demonstrated that exogenous AEA is a more effective agonist (for producing increases in mIPSC frequency) than exogenous 2 AG, even when 2 AG is applied in the presence of a monoacylglycerol lipase inhibitor, JZL 184. Consistent with that observation, we would also expect that endogenous activation of the system described here will ultimately turn out to be more tightly coupled to systems that regulate production of AEA or its metabolites (activity dependent or not), than to postsynaptic mechanisms clearly optimized for activity dependent activation of CB1. While significant future work will be necessary to substantiate these ideas, it is very intriguing to note that a synthetic enzyme responsible for the production of AEA has been recently reported to be present in abundance in the presynaptic terminals of glutamatergic axons, with particular ly high expression in the hilar region of the dentate gyrus (Nyilas et al., 2008) An additional topic worth some discussion is the ultimate physiological relevance of CB mediated potentiation of action potential independent release, i.e. could it plausibl y occur in the intact brain? We believe it could. That conclusion is based on two primary observations. First, we found that an imperfect technique for sampling the activity of incoming afferents (whole cell recording) is still sufficient to observe cann abinoid mediated increases in mIPSC frequency in nearly 50% of trials completed in the absence of ruthenium red (40% success in our primary control dataset, up to 66% success in the presence of postsynaptic BAPTA, Fig. 5 10). Second, we found that a
108 prima ry endogenous cannabinoid (AEA) is clearly effective at sub M concentrations that are well within the physiological range. We believe these observations strongly imply that this mechanism could be relevant to normal function of inhibitory synapses to hilar mossy cells. While we acknowledge that we do not know ex actly why success rate increases substantially in the presence of ruthenium red, this observation does not diminish the fact that the phenomenon can readily be observed in the absence of ruthenium red. In fact, there are several possible explanations for the increased success rate observed in ruthenium, some of which implicate this compound as a tool that may ultimately be useful in future efforts to reveal the specific molecular mechanism responsible for CB mediated increases in action potential independe nt exocytosis. Finally, with respect to the potential functional significance of the phenomenon, it should be noted that previous work in other systems has implicated action potential independent synaptic transmission in a wide range of specific effects including clustering of postsynaptic glutamate receptors (Saitoe et al., 2001) formation of dendritic spines (McKinney et al., 1999) and related maintenance of synaptic connections in the absence of action potentials (McKinney et al., 1999) In addition a number of studies have indicated an ability of miniature synaptic potentials to affect firing rates and inhibitory tone (Lu and Trussell, 2000; Carter and Regehr, 2002; Sharma and Vijayaraghavan, 2003) At a basic level, what virtually all of these spe cific mechanisms (of necessity) have in common is an ability to detect and respond to low levels of exocytosed neurotransmitters. Thus, in our view, the CB mediated increase in action potential independent exocytosis identified here deserves further study as a new
109 physiological mechanism potentially able to produce functional activation of a variety of high affinity extrasynaptic and presynaptic GABA receptors recently identified in this area (e.g. Kerr and Capogna, 2007; Nahir et al., 2007) That said, w e might expect to find both spatially and temporally selective effects, produced by activation of GABAergic receptors present within small microdomains where ambient GABA is transiently increased by spillover from CB mediated and action potential independe nt exocytosis.
110 Figure 5 1. Bath application of cannabinoid agonist WIN55,212 2 causes an increase in mIPSC frequency without altering amplitude. A) Top panels are raw data traces representative of mIPSCs recorded in various experimental conditions (se e labels). Note that mIPSC frequency increases with bath application of ruthenium red (50 M), and then further increases with bath application of WIN55,212 2 (5 M). Lower panels are cumulative probability histograms constructed from the same cell. In this cell WIN55,212 2 caused a decrease in interevent interval (IEI) from 100 2.18 ms to 83.2 1.17 ms (p<0.001, Kolmogorov Smirnov (K S) test) with no change in amplitude (baseline 15.9 0.2 pA; WIN55,212 2 15.6 0.1 pA ; p=0.35, K S test). B) Event frequency (open circles) and average amplitude (closed circles) were calculated in 30 sec ond bins, normalized to the level observed immediately prior to application of WIN55,212, and averaged across five cells. Numbers indicate different conditions observed over the course of the experiment as follows: (1) Mossy cells display large amplitude and high frequency spontaneous EPSCs immediately after patching in control conditions. (2) Application of NBQX, APV, and TTX blocks both ionotropic glutamate receptors and action potentials, leaving much lower frequency, and much smaller, miniature IPSCs (3) Subsequent application of ruthenium red produced an increase in mIPSC frequency that stabilized within ~15 min, and (4) Bath application of WIN55,212 2 in the presence of ruthenium red produces a further increase in event frequency, without alterin g event amplitude. Note again that data were normalized to the stable level observed in NBQX, APV, TTX, and ruthenium red. In some panels in later figures similar experiments are illustrated from this point forward. C) Summary plot for all cells tested by WIN55,212 2 in the presence of ruthenium red (n=10) indicating that other event parameters (area, rise time, and decay) were also unaffected by WIN55,212 2. Inset shows the average of all mIPSCs from a representative cell during the baseline and measu rement periods.
111 Figure 5 2. Bath application of WIN55,212 2 does not alter intrinsic properties of hilar mossy cells under voltage clamp. Throughout the study no consistent effect of WIN55,212 2 on holding current (I hold ), access resistance (R a ), or in put resistance (R m ) was noted. However many experiments were conducted as gap free recordings and thus only occasional measurements of R a and R m were available. In order to more directly test the effects of WIN55,212 2 on intrinsic properties, I hold R m and R a were measured every 5 seconds during bath application of WIN55,212 2 in 8 cells that had been voltage clamped for at least 20 minutes. DN QX, APV, and TTX were present for all experiments. Our results indicated that there was no significant effect of WIN55,212 2 on I hold or R m The specific values for these parameters are apparent on the figure above (n=8, p>0.05 in both cases). The change in R a is statistically significant on a two sample paired t test (p=0.015). However, some increase in R a is typical over time. The change observed here is from 12.0 0.58 MOhm in the baseline period to 12.8 0.65 MOhm in the test period. Based on the time course, this change is unlikely to be rela ted to the bath application of WIN55,212 2, and based on the magnitude (and direction) it is extremely unlikely to cause an apparent increase in mIPSC frequency absent an actual eff ect.
112 Figure 5 3. WIN55,212 2 mediated facilitation of mIPSCs is not dep endent on CB1, CB2, or TRPV1 receptors. A) In the presence of 5 M AM251, bath application 5 M WIN55,212 2 increased the frequency of mIPSCs to 148 16.5% of the level observed in ruthenium red (p=0.03 n=7), without altering amplitude. B) In the prese nce of 10 M capsazepine (CPZ), a TRPV1 antagonist, WIN55,212 2 potentiated the frequency to 140.4 7.85% of the level observed in ruthenium red (p=0.01, n=4). C) Summary plot indicating the WIN55,212 2 mediated increases in mIPSC frequency were observe d not only in the presence of AM251 and CPZ, but also in CB1 / animals (166 5.66% of baseline, n=3, p=0.007) and in the presence of 1 M AM630, a CB2 receptor antagonist (146 14.3% of baseline, n=5, p=0.03). Data were analyzed and presented as descri bed in Fig. 5 1B.
113 Figure 5 4. Ruthenium red does not independently produce, and is not required to observe, WIN55,212 2 mediated increases in mIPSC frequency. A) 50 M ruthenium red produces an increase in miniature IPSC frequency that remains stable in the absence of WIN55,212 2 (106 5.18%, measured 10 14 minutes after normal time for WIN55,212 2 application, p=0.3 ,n=5). Data in this plot are aligned by the time of ruthenium red application. Data is not shown from 3 to 7 minutes because the amou nt of time between application of ionotropic glutamate receptor antagonists and ruthenium red was not constant in all cases (minimum: 6 min, maximum: 10 min). B) WIN55,212 2 mediated increases in mIPSC frequency can still be observed in the absence of rut henium red (mIPSC freq: 226 28.9% of control, p=0.05, n=7 of 17, successes only, open circles, see main text), with no corresponding change in amplitude (filled circles). When averaged across all cells (successes and failures), mIPSC frequency was incre ased by WIN55,212 2 to 154 19.1% baseline (open gray circles, n=17, p=0.01). C) Summary: RR in the absence of WIN55,212 2 had no late effect on event frequency, while WIN55,212 2 still increased event frequency in the absence of RR in 7 of 17 of trials For the analysis in panels B C, successes are runs in which bath application of WIN55,212 14 minutes after application). D) Raw data from a representative successful run in the absence of ruthenium red E) Cumulative probability histograms for the same cell as presented in D, quantified over the normal control and measurement periods. These last two panels are presented to clearly emphasize that ruthenium red is not required to observe clear WIN55,212 2 mediated facilitation of mIPSCs.
114 Figure 5 5. Bath application of AEA also selectively increases mIPSC frequency without altering amplitude, area, rise time, or decay. A) Top panels are raw data traces representative of mIPSCs recorded in NBQX, APV, T TX, and ruthenium red (left traces) and in the same cell following addition of 0.5 M AEA (right traces). Lower panels are cumulative probability histograms constructed from the same cell. In this cell AEA caused a clear decrease in interevent interval ( IEI) without altering event amplitude. B) Across five cells tested, AEA significantly increased mIPSC frequency (to 151.6 9.7% of baseline, p<0.01, n=5), without altering event amplitude. Data in this panel is analyzed and presented as described in th e legend for Fig. 5 1B. C) Summary plot indicating that other event parameters (area, rise time, and decay) were also unaffected by AEA. Inset shows the average of all mIPSCs from a representative cell during the baseline and measurement periods.
115 Fig ure 5 6. Low doses of WIN55,212 2 effectively increase mIPSC frequency in hilar mossy cells. AEA is a more efficacious agonist than 2 AG. A) Left panel: Bath application of WIN55,212 2 (50 nM) increased mIPSC frequency to 131.3 4.1% of baseline (n=5, p<0.01). Right panel: a comparable effect was noted at 250 nM (Frequency: 128.4 6.7% of baseline, n=5, p=0.01, Amplitude: 106 2.7% of baseline, n=5, p=0.09). DNQX, APV, TTX, and RR were present for these experiments, and applied as described in Fig. 5 1. B) Summary panel for data presented in A. C) The effect of AEA (0.5 M, grey symbols) is replotted from Fig. 5 5 to facilitate easy comparison to the effects of 2 AG (30 M, open symbols, Frequency: 114.9 4.31% of baseline, n=4, p=0.04, Amplitu de: 96.5 1.53% baseline, n=4, p>0.05). DNQX, APV, TTX, and RR were also present for these experiments, and applied as described in Fig. 5 1. D ) Summary panel for the data presented in panel C. Note that the effects of AEA (0.5 M) on mIPSC frequency are significantly larger than those produced by 2 AG at 60 fold higher concentration (30 M, p=0.016).
116 Figure 5 7. Suramin effectively blocks WIN55,212 2 mediated facilitation of mIPSCs, but the non selective P2X receptor antagonist PPADS does not. A) Slices were preincubated for 20 min in 20 M suramin (a compound that effectively uncouples GPCRs from their effectors), and then transferred to a patch clamp rig for whole cell recording. WIN55,212 2 (5 M) was bath applied in the presence of DNQX, APV, TTX, and suramin but absent ruthenium red. No change was noted in mIPSC frequency ( 100 20.3% of baseline, n=10, p=0.99 ) or amplitude (101 2.82% of baseline, n=10, p=0.79). The few points with high frequency (and high standard error) at approximately the 17 minute mark are produced by 2 of 10 cells that showed brief flutters of mIPSCs shortly after application of WIN55,212 2. B) In addition to its effects on GPCRs, suramin has been reported to be a non selective P2X receptor antagonist. This panel demonstrates that another non selective P2X receptor antagonist (PPADS, 20 M) fails to reproduce the effects of suramin give n identical preincubation and application protocols. Specifically WIN55,212 2 (5 M) increased mIPSC frequency in the presence of PPADS to 169 34% of baseline (n=12, p=0.066), without altering amplitude (98.2 2.00% of baseline, n=12, p=0.39). C) Sum mary data for experiments presented in panels A B.
117 Figure 5 8. WIN55,212 2 mediated potentiation of mIPSCs is blocked by a selective PKA inhibitor, and is both mimicked and occluded by forskolin. A) Slices were preincubated in 10 M H 89 for at least 1 hour, and H 89 was maintained in the bath during whole cell recording. In the presence of H 89 bath application of 5 M WIN55,212 2 fails to increase mIPSC frequency. B) Bath application of forskolin (20 M ) causes a clear increase in frequency of mIP SCs without altering amplitude. This effect is very similar to the effect of WIN55,212 2 illustrated in Fig. 5 1. Further, subsequent application of WIN55,212 2, after forskolin, fails to produce any further potentiation in mIPSC frequency. C) Summary p lot of experiments in panels A B. The bar graph for WIN55,212 2 applied after forskolin in this panel shows the effect of WIN55,212 2 relative to the level of activity obtained in forskolin. DNQX, APV, TTX, and RR were also present for the experiments i llustrated in this figure, and were applied as described in Fig. 5 1
118 Figure 5 9. Cannabinoid agonists have no effect on action potential dependent exocytosis in the presence of the CB1 receptor antagonist AM251. Evoked IPSCs were generated at a rate of 0.33 Hz using a bipolar (panel A B) or minimal (glass monopolar) stimulator (panel C) placed in the hilus. NBQX and APV were present for all experiments. A) Bath application of 5 M WIN55,212 2 or 0.5 M AEA had no significant effect on evoked IPSC amp litude in slices pretreated with 5 M AM251. B) Similarly, under these conditions, CB agonists had no effect on the paired pulse ratio of responses generated with a bipolar stimulator. C) WIN55,212 2 also had no effect on PPR of minimally evoked responses (meIPSCs). Raw data traces in panels B and C are normalized to the P1 peak.
119 Figure 5 10. WIN55,212 2 mediated increases in mIPSC frequency are likely to depend on increases in presynaptic calcium concentration. NB QX, APV and TTX were present for all experiments. A) Under control conditions (open circles) WIN55,212 2 significantly increased mIPSC frequency (data from Fig 5 4 B, successes only). By contrast, WIN55,212 2 mediated enhancement of mIPSC frequency was severely reduced by bath application of low Ca 2+ external solution (see methods) or 200 M Cd 2+ (n=13 and 10 respectively, data shown is average of all experiments). B) Similarly, preincubation with either 100 M EGTA AM or 100 M BAPTA AM (n=12 and 5 respectively) severely reduced the effect of WIN55,212 2 on mIPSC frequency. C) Top panel: Effect of WIN55,212 2 on mIPSC frequency, averaged across all cells tested (both successes and failures), is shown for every experimental condition. Bottom panel: similar plot showing success rate for bath application of WIN55,212 2 in every experimental condition. Success rate was defined as the percentage of trials where WIN55,212 2 caused an increase in mIPSC frequency of at least 1 Hz, measured from 10 14 minutes after onset of application. Both the population effect and the success rate are reduced by global manipulations of calcium (gray bars) but largely spared by selective
120 chelation of calcium in the postsynaptic cell (10 mM BAPTA internal, open bar on the right). All calcium manipulations were done as pretreatments. D) Effectiveness of WIN55, 212 2 in every experimental condition is expressed as the product of the success rate and the population effect. Values are normalized to control conditions. Effectiveness of WIN55,212 2 was severely reduced by all global manipulations of calcium, but wa s not reduced by chelation of calcium in the postsynaptic cell. Inset shows result of an identical analysis where success rate was defined as the percentage of cells that produced a statistically significant increase in baseline frequency (p<0.05) over the same time period. This was a less restrictive test of success, but the overall results of the analysis were quite similar. Bars are presented in an identical order.
121 Figure 5 11. Depolarization of DSI positive hilar mossy cells produces a small, but c omparatively long lasting, increase in mIPSC frequency. DNQX and APV are present for all experiments. Evoked IPSCs were generated at a rate of 0.33 Hz using a bipolar stimulator. A) In the presence of bath applied CCh (3 M), depolarization of mossy cell s from 70 to 0 mV for 5 s (bar) caused a significant reduction of the mean eIPSC amplitude (DSI, by 45.21 4.49%, n=6, p<0.001, measured over first 6 seconds after depolarization). B) After application of 1 M TTX and 50 M ruthenium red, identical depol arization of these same cells produced a small, but statistically significant increase in mIPSC frequency (106 2.3% of baseline when measured 0 12 seconds after depolarization, n=6, p=0.04), without altering amplitude. Interestingly, this change was als o clearly longer lasting than DSI (107 2.26% of baseline, n=6, p=0.04 when measured 1.5 2.0 minutes after depolarization). C) Summary plot indicating the effect of depolarization on evoked IPSCs vs. mIPSCs in the same population of cells (n=6).
122 CHAPT ER 6 CONCLUSIONS Summary and Future Directions Although there are many potentially beneficial effects of marijuana, it was deemed a drug of abuse and made illegal in 1937. However, recently there has been a growing trend to legalize medicinal marijuana in the United States, but this has not been without its controversy. With the identification of the receptor responsible for many of the effects of THC on the brain, scientists are gaining new insights into how both endogenous and synthetic cannabinoids fun ction in the CNS with a keen focus towards developing new therapeutic strategies to avoid the controversial issues associated with marijuana. Two of the effects of marijuana in the brain, the ability to impair memory and to work as an anticonvulsant, are of particular importance in the hippocampus, a structure intimately tied to memory and epilepsy. Research has demonstrated the endogenous cannabinoid system is a key contributor to both of these areas of hippocampal research. Cannabinoids can modulate LT P and iLTD, processes important in learning and memory (Carlson et al., 2002; Chevaleyre and Castillo, 2003) and the CB1 receptor has been shown to reduce susceptibility to seizures (Monory et al., 2006) While significant progress has been made towards understanding the synaptic physiology of cannabinoids in the hippocampus, many unanswered questions remain. The primary goal of this dissertation was to answer some of these questions in an effort to expand our knowledge of cannabinoid function in the hip pocampus. This section will break down some of the key contributions of our studies towards addressing these issues and discuss their impact on the synaptic physiology of the hippocampus.
123 Differential Expression of CB1 at Glutamatergic Synapses to CA3 Pyr amidal Cells In Chapter 3 we explored the role of cannabinoids at glutamatergic synapses to CA3 pyramidal cells. Compared to GABAergic synapses relatively little is known about CB1 at glutamatergic synapses in the hippocampus as a whole and at CA3 pyramid al cells in particular. While initial immunohistochemistry supported the focused research on CB1 at GABAergic synapses (Tsou et al., 1998) a new CB1 antibody provided evidence for CB1 at glutamatergic terminals. In these studies, CB1 appeared to be pres ent in stratum radiatum but not stratum lucidum suggesting differential expression of CB1 at A/C and MF synapses to CA3 pyramidal cells, respectively (Katona et al., 2006; Monory et al., 2006) Our research provided the first physiological evidence suppor ting the immunohistochemistry for the new CB1 antibody. Following application of cannabinoid agonists, A/C afferents but not MF afferents to the same CA3 pyramidal cell were inhibited. This differential sensitivity has several different implications. F irst, the discovery of CB1 at A/C afferents contributed to the debate on CB1 at CA3 CA1 synapses. A/C afferents are axon collaterals of Schaeffer collaterals, so our data support the findings suggesting CB1 is present at this synapse unless there is diffe rential expression of CB1 depending on the axonal target. Second, any ECs released in stratum lucidum will provide the ability to preferentially inhibit synapses other than MFs. This is of particular importance considering a recent study that found a can nabinoid tone at nearby mossy fiber associated interneurons (Losonczy et al., 2004) Third, instead of acting directing at MF terminals, cannabinoids could modulate MFs indirectly through changes in ambient GABA similar to mAChR activation in this area (V ogt and Regehr, 2001) Reduction in the cannabinoid tone could unsilence the mossy fiber associated interneurons which would be uniquely positioned to release
124 GABA that could bind to the GABA A and GABA B receptors present at MF terminals (Vida and Frotsche r, 2000; Vogt and Regehr, 2001; Ruiz et al., 2003; Losonczy et al., 2004) Fourth, recent research has also shown that MF terminals contain the necessary machinery for the production of AEA (Nyilas et al., 2008) If MFs are capable of releasing AEA ther e will not be a direct autoinhibitory effect due to the lack of the CB1 receptor. It is also interesting to speculate whether MFs are responsible for the EC tone in stratum lucidum through the production and release of AEA from their terminals. This woul d provide a potential mechanism for MFs to control excitability of local produce AEA could have a profound impact on our understanding of how MFs can control synaptic tr ansmission in stratum lucidum, a key area for information flow through the hippocampus. EC Involvement in Synaptic Plasticity at CA3 CA3 Synapses After providing evidence that CB1 was present at A/C synapses we investigated the potential role of ECs in sho rt term plasticity at these synapses. While the role of ECs in modulating synaptic plasticity at glutamatergic terminals has been limited to date, release of ECs has been shown to cause DSE at Schaeffer collaterals and presumed mossy cell axons (Ohno Shos aku et al., 2002; Chiu and Castillo, 2008) Our data demonstrated that CA3 pyramidal cells could produce ECs in a calcium dependent manner resulting in DSE which provided the first report of EC mediated synaptic plasticity in area CA3. Interestingly, thi s DSE was different than DSE/DSI in other areas of the hippocampus. Typically activation of mAChRs enhances the release of ECs and the size of DSI/DSE (Kim et al., 2002; Ohno Shosaku et al., 2003; Fukudome et al., 2004; Hofmann et al., 2006; Chiu and Cast illo, 2008) At this synapse we discovered
125 they must be activated to get release of ECs. While it was recently discovered that mAChR enhanced cannabinoid release can have a high dependence on nitric oxide (Makara et al., 2007) we found this was not the case in CA3 pyramidal cell DSE. The requirement for mAChR activation implies the medial septal nucleus, which does innervate CA3 pyramidal cells (Frotscher and Leranth, 1985) will likely play an intimate role in EC release. The medial septal nucleus is implicated in producing theta rhythms, which are oscillatory patterns seen in the membrane potential or action potential firing of neurons in the 4 14 Hz range (Buzsaki, 2002; Yoder and Pang, 2005) Perhaps the CB1 receptor at A/C afferents is a key compo nent in synchronizing CA3 pyramidal cells in the theta band in a coordinated effort with cholinergic afferentation. Prolonged CB1 receptor activation has already been shown to reduce synchrony in CA3 pyramidal cells (Goonawardena et al., 2010) but it is unknown what effect brief activation of CB1 might have on synchrony. These synapses could also have a profound effect on how we view CB1 as an anticonvulsant especially when considering data demonstrating CA3 as a location for seizure initiation (Dzhala a nd Staley, 2003; Wittner and Miles, 2007) Continued investigation of the relationship between the CB1 receptor at A/C afferents and synchronized activity of CA3 pyramidal cells will be important to increase our understanding of epileptic brains and their potential treatment. mAChR Activation in Mossy Cells Results in an ADP The primary goal of Chapter 4 was to study the potential role of mAChRs in obtaining functionally relevant DSI. However, before investigating this potential link, we first studied the effect of mAChR activation on the intrinsic properties of mossy cells. Mossy cells are well positioned for cholinergic modulation with connections to their soma and dendrites (Deller et al., 1999) While there was previous evidence that
126 mAChR activation contributed to the synaptic physiology of these cells (Hofmann et al., 2006; Nahir et al., 2007) there was no research indicating any intrinsic changes following activation of these receptors. We discovered an ADP in mossy cells following activation of mAChRs providing the first evidence for changes to their intrinsic properties. This prolonged depolarized state often caused the mossy cell to fire action potentials, which could have a profound effect on its axonal targets. In the olfactory bulb, mAChR activation on granule cells resulted in ADPs and persistent action potential firing. This increase in firing caused a potentiation of mitral cell inhibition through increased release of GABA from the granule cells (Pressler et al., 2007) How the change s to action potential firing in mossy cells affects their axonal targets has yet to be determined and should be investigated in future studies. There are at least two distinct potential outcomes: increased excitatory drive of hilar interneurons and incre ased excitatory drive of dentate granule cells. In the first scenario the increased activity of interneurons could silence subsets of granule cells, while in the second scenario the increased firing of mossy cells could increase the firing of specific su bsets of granule cells or potentially synchronize granule cells in different subfields of the dentate. Overall, investigating how ADPs affect the output of mossy cells could provide important information about the role they play in regulating information flow from the dentate to CA3 pyramidal cells. Finally, previous studies have found that activation of mAChRs can result in additional intrinsic changes such as decreases in action potential jitter and latency to sinusoidal waves and increases in firing re liability (Lawrence et al., 2006a; Pressler et al., 2007) Future studies should also investigate whether similar
127 changes can occur in mossy cells to further address the role mAChR activation can have on their intrinsic properties. Induction of an ADP Can Cause DSI Perhaps the most intriguing finding of Chapter 4 is that an ADP is a possible physiological mechanism for DSI. Previous research has questioned whether DSI was functionally relevant in the hippocampus (Hampson et al., 2003) Our data indicates that following a 3 5 second ADP we could get a transient inhibition of evoked IPSCs that was sensitive to the CB1 antagonist AM 251. These data provide the first evidence that DSI might have functional relevancy in the hippocampus; however, future work i s needed because we used bath application of a cholinergic agonist providing continuous activation of mAChRs not likely in the hilus. The next step in these experiments will be to attempt to get ADP induced DSI with a local application of acetylcholine pa ired with brief stimulation. An additional experiment, although ambitious, would be to attempt to get ADP DSI from endogenous acetylcholine by stimulating medial septal inputs using a previously described septo hippocampal parasagittal slicing technique ( Toth et al., 1997; Widmer et al., 2006; Goutagny et al., 2008) If ADP DSI could be produced from the stimulation of endogenous cholinergic afferents, then it is interesting to speculate on how the timing of cholinergic input could affect signaling in th e hilus. As mentioned earlier, the medial septal nucleus is implicated in the production of theta rhythms in the hippocampus (Buzsaki, 2002; Yoder and Pang, 2005) and perhaps the induction of DSI following an ADP could be a mechanism for synchronizing ne urons in the hilus. DSI could silence a subset of interneurons and also potentially disinhibit MF inputs through a reduction in ambient GABA that would decrease activation of presynaptic GABA B receptors (Nahir et al., 2007) Finally, there are also sever al other areas for potential
128 investigation into the functional relevancy of DSI in mossy cells. Recent work in our lab has demonstrated that mGluR activation can produce ECs resulting in iLTD (Nahir et al., 2010) Activation of mGluRs paired with in vivo firing rates could result in DSI similar to that found in a study on cerebellar stellate basket cells (Myoga et al., 2009) In addition, perforant path stimulation can cause plateau potentials in semilunar granule cells resulting in a long lasting barrage of synaptic activity in hilar mossy cells (Larimer and Strowbridge, 2010) Perhaps this barrage of EPSCs could sufficiently depolarize the mossy cell to allow for EC production. Further investigation into these alternative mechanisms could demonstrate a wide range of potential pathways for the production of ECs and induction of physiologically relevant DSI. Novel Effect of Cannabinoids on GABAergic Signaling to Mossy Cells In Chapter 5 we investigated a novel, CB1 independent cannabinoid effect on GABA ergic afferents to hilar mossy cells previously discovered in the lab. This initial finding was particularly interesting due to previous research indicating there is a potential novel, non CB1 cannabinoid receptor in the hippocampus (Di Marzo et al., 2000 ; Breivogel et al., 2001) and due to the speculation of novel cannabinoid receptors in the CNS in general (Kreitzer and Stella, 2009) However, these intriguing propositions of a novel receptor must also be weighed against the known interactions of cannab inoids with various receptors and potential non specific effects of cannabinoids through changes to the plasma membrane (Oz, 2006a) Our data indicate that our novel cannabinoid effect facilitates action potential independent exocytosis through an unknown GPCR. In addition, we found that AEA instead of 2 AG appears to be the primary EC for our novel effect. This section is broken down below to discuss each of these findings in more detail, to describe their impact on the synaptic physiology of the
129 hippoc ampus, and to speculate on potential future investigations to increase our understanding of this novel cannabinoid effect in the hippocampus. Cannabinoid modulation of action potential independent exocytosis Traditional cannabinoid signaling is a CB1 recep tor dependent inhibition of action potential dependent exocytosis. In contrast, the novel cannabinoid effect we discovered is a CB1 receptor independent excitation of action potential independent exocytosis. Even in the presence of a CB1 antagonist we we re unable to unmask an increase in the amplitude of evoked or minimally evoked IPSCs following cannabinoid application. The ability to differentially modulate action potential independent exocytosis through this cannabinoid effect is particularly interest ing. Many presynaptic receptors can alter both action potential dependent and independent exocytosis, while others such as the CB1 receptor can only modulate action potential dependent exocytosis. In this study we have provided specific evidence for a pa thway that can specifically alter action potential independent exocytosis. This could give neurons the ability to regulate processes such as protein synthesis in dendrites and maintenance of synaptic connections, which are dependent on this type of exocyt osis, without affecting action potential dependent neurotransmission (McKinney et al., 1999; Sutton et al., 2004) Our data also suggest that there is either separate vesicle pools for these two types of exocytosis or different release machinery because t hey can be modulated independently. Finally, CB1 dependent modulation of neurotransmission is known to occur at both glutamatergic and GABAergic synapses in both the hippocampus and other areas of the CNS. To date it is unknown whether our novel effect i s also present at glutamatergic synapses or whether it is present in other cell types. Future studies should focus on addressing
130 these issues, however this will be noticeably easier if the responsible receptor and/or gene product were also discovered. GPC R involvement in the novel cannabinoid effect After our initial determination that our novel cannabinoid effect was independent of CB1, we quickly checked two other receptors often activated by cannabinoids: the CB2 receptor and the transient receptor pot ential current. Our data indicate that activation of these receptors was also not responsible for the increase in action potential independent exocytosis. While there is data suggesting a novel cannabinoid receptor is present in the hippocampus (Di Marzo et al., 2000; Breivogel et al., 2001) and an abundance of research has demonstrated that ECs can activate a wide array of different receptors, we needed to confirm that cannabinoids were not acting through a non specific effect before a continued consider ation of potential receptors (Oz, 2006a) There are several lines of evidence suggesting our cannabinoid effect is in fact receptor mediated. First, two characteristics of a non specific effect would be a consistent cannabinoid effect (i.e. few if any fa ilures) and no ligand specificity for cannabinoids with similar structure. Our data indicated that cannabinoids only increased action potential independent exocytosis approximately 50% of the time in the absence of ruthenium red. In addition, AEA and 2 A G, ECs with similar structures, appear to have significantly different effects at vastly different concentrations (AEA worked better at a 60 fold lower concentration). Both of these results would argue against a non specific effect of cannabinoids. Secon d, non specific drug effects typically occur at M or higher concentrations, and we found effects of both WIN55,212 2 (50 nM) and AEA (500 nM) at submicromolar concentrations. Finally, the most direct piece of evidence against a non specific effect was ou r data suggesting the involvement of a GPCR. Suramin,
131 which uncouples G proteins from their receptors, prevented cannabinoids from increasing the frequency of action potential independent exocytosis suggesting a GPCR was responsible for the observed effec t. While this data provided evidence against a non specific effect, it also narrowed down the potential receptors cannabinoids could be interacting with because the GPCR dependence removed all the potential ionotropic receptors from the list; however, can nabinoids have been demonstrated to interact with a variety of non CB1 GPCRs (Oz, 2006a) It did not come as a surprise that a GPCR was involved in the increase of action potential independent exocytosis because previous research suggested the presence of a novel cannabinoid receptor in the hippocampus that was a GPCR (Di Marzo et al., 2000; Breivogel et al., 2001) In addition, similar to the CB1 receptor, the newly identified cannabinoid receptor GPR55 is also a metabotropic receptor (Ryberg et al., 2007 ; Lauckner et al., 2008) These studies imply it is likely that any new cannabinoid receptors that are discovered are likely to be GPCRs. One potential way to assist with the identification of our receptor would be to determine the responsible subunit because most GPCRs are typically associated with one subunit of the G protein. Our data indicate that PKA and adenylyl cyclase are part of the signaling cascade for the potentiation of action potential independent exocytosis. Typically activation of th ese effector molecules are part of the signaling cascade: G S adenylyl cyclase PKA. Additional work will need to be done to verify that our novel cannabinoid effect acts through the stimulation of G S This could be accomplished through the use o f cholera toxin which stimulates G S and should occlude any cannabinoid effect. An additional clue to a potential receptor comes from the lack of stereoselectivity of our
132 cannabinoid effect. Our results indicate that both WIN55,212 2 and its analog WIN55 ,212 3 are equally effective at potentiating action potential independent exocytosis. While most receptors are stereoselective there is evidence for GPCRs that lack stereoselctivity such as LPA (Yokoyama et al., 2002) LPA is particularly interesting as a novel cannabinoid receptor because it shares good homology with CB1 (Kreitzer and Stella, 2009) In fact LPA has been shown to activate cAMP and PKA (Rhee et al., 2006) and can trigger exocytosis in chromaffin cells (Pan et al., 2006) In CA1 activati on of LPA 2 receptors resulted in a potentiation of mEPSC frequency, however there was no change to mIPSC frequency (Trimbuch et al., 2009) While there was no facilitation of mIPSC frequency arguing against LPA as the novel cannabinoid receptor, there cou ld be a different expression pattern of LPA 2 between the hilus and CA1. An additional lysopholipid receptor of interest is S1P because it too shares good homology with CB1 (Stella, 2009) In hippocampal neurons sphingosine kinase 1, a key enzyme for S1P production, can be found and activation of S1P 1 receptors can cause a facilitation of glutamate release (Kajimoto et al., 2007) While there is no evidence yet for endogenous production of S1P from these neurons, cerebellar granule cells and astrocytes ha ve shown the ability to produce S1P following exogenous stimuli (Anelli et al., 2005) In addition, S1P has been shown to be a potential endogenous CB1 antagonist (Paugh et al., 2006) In this scenario S1P could not only increase the release of GABA thro ugh an action potential independent mechanism but also by relieving CB1 mediated inhibition of GABA release through an action potential dependent pathway. LPA receptors and S1P receptors are promising potential receptors for our novel cannabinoid effect o n action potential independent exocytosis
133 and worthy of future investigation. Overall, our data indicate that neurons can potentially modulate action potential independent exocytosis through an unknown GPCR that activates the adenylyl cyclase cAMP PKA signaling cascade. Potential physiological relevance of AEA in the novel cannabinoid effect While finding the receptor would be optimal for evaluation of the physiologically relevant molecule for the novel cannabinoid effect, our data does provide some cl ues towards figuring this out. We demonstrated that AEA appears to be significantly more efficacious at producing our effect than 2 AG even when breakdown of 2 AG is inhibited. This difference was seen even though 2 AG was used at a 60 fold higher concen tration. In additional support of these results we found that the same depolarization that could produce robust DSI only modestly increased the frequency of action potential independent exocytosis. Typically it is believed that 2 AG is the EC produced fo r DSI, although this has not been definitively determined at mossy cells. Regardless, these data further support our observation that AEA is more likely than 2 AG to be the relevant cannabinoid at our receptor, but a complete dose response curve will be n ecessary for full evaluation. If our hypothesis does hold true, then 2 AG might be considered a partial agonist for our receptor. This is particularly interesting because AEA has the characteristics of a partial agonist at CB1 (Hillard, 2000) This prov ides a potential scenario that each EC has a different primary site of action and modulates different pathways of exocytosis. It is also interesting to speculate about the source/mechanism of release if AEA is the primary ligand. Recent research found th e enzymatic machinery responsible for the production of AEA in MF terminals (Nyilas et al., 2008) While it has yet to be proven if they will in fact produce AEA, this could be a possible source of it for our novel cannabinoid effect. It is also importan t to note that we
134 do not know the exact site of action for the cannabinoids. While we do believe the final site of action is at the presynaptic terminal because we observed an increase in the frequency but not amplitude of mIPSCs, there is the possibility that the initial site of action of cannabinoids could be located elsewhere. For example, recent research has shown that ECs can bind to CB1 on astrocytes causing them to release glutamate (Navarrete and Araque, 2008) While this exact scenario is unlike ly because our effect is CB1 independent and we have blocked ionotropic glutamate receptors, perhaps a similar mechanism through the activation of a different receptor could occur. This would imply that cannabinoids cause the release of some unknown retro grade messenger making this a much more complex phenomenon; however, at this time we are continuing to evaluate the more simplistic model that cannabinoids are interacting directly with a receptor located at the presynaptic terminal until data proves other wise. It will also be important to consider additional potential molecules besides AEA for our novel cannabinoid effect. While our data indicate that AEA is an efficacious ligand, we could be in actuality observing an effect of an AEA breakdown product. Both prostaglandins and HETEs have been shown to have pharmacological activity that can modulate neurotransmission. For example, prostaglandins can enhance glutamatergic transmission (Sang et al., 2005) and HETEs are important for the induction of mGlu R LTD (Feinmark et al., 2003) Future studies could investigate whether these endogenous compounds can effectively induce increases to action potential independent exocytosis. Finally, the effectiveness of LPA and S1P could also be assessed because their receptors have been proposed as possible novel cannabinoid receptors (Kreitzer and Stella, 2009) All of these different experiments to determine
135 potential physiologically relevant molecules will also help down the road in the discovery of the receptor b y providing a more detailed pharmacological profile. Overall, Chapter 5 has contributed to our understanding of cannabinoids in the hippocampus by laying the foundation for a CB1 receptor independent effect of cannabinoids in the synaptic physiology of t he dentate hilus. Ultimately, discovery of the gene product for this unknown GPCR will be necessary for continued significant progress because it will allow us to determine its specific distribution in the CNS. In addition, it will enable us to determine the most likely endogenously relevant molecule, investigate the production and breakdown of the molecule, and provide clues for studying the functional significance of the increase in action potential independent exocytosis. Perspectives As mentioned thro ughout this dissertation, one of the reasons scientists have continued to study cannabinoids is to parse out the mechanisms behind both the good effects (role as an anticonvulsant) and bad effects (impairment of memory and cognition) of marijuana. Through this process scientists have gained additional insights about the role CB1 plays in epilepsy, while we have also begun to understand the effects of chronic THC on cannabinoid mediated plasticities. This section will focus on how the findings in this diss ertation fit into what we know about epilepsy and chronic use of THC and the additional questions they raise towards achieving a comprehensive understanding of these important areas of study. Cannabinoids and Epilepsy Previous research has demonstrated tha t specific deletion of CB1 from glutamatergic terminals and not GABAergic terminals increases the susceptibility to
136 seizures (Monory et al., 2006) Also CB1 is downregulated in human tissue specifically at glutamatergic synapses over GABAergic synapses in cluding robust decreases in stratum radiatum of area CA3 (Ludanyi et al., 2008) Our data provided physiological evidence that A/C afferents do express CB1 which can be modulated by ECs. Perhaps this synapse in particular is important for the anticonvuls ant effects of cannabinoids (Wallace et al., 2001; Wallace et al., 2002; Wallace et al., 2003; Blair et al., 2006) and the increased susceptibility to seizures observed in transgenic mice (Monory et al., 2006) There is evidence that cannabinoids can dec rease synchrony in CA3 pyramidal cells which would support this view (Goonawardena et al., 2010) In addition, CA3 can be a location for seizure initiation and maybe decreased expression of CB1 at A/C afferents increases the susceptibility to seizure ini tiation (Dzhala and Staley, 2003; Wittner and Miles, 2007) While it is known that CB1 expression specifically at glutamatergic synapses over GABAergic synapses is important for seizure susceptibility, it is unknown at this time whether CB1 expression at specific glutamatergic synapses, such as A/C synapses, are more important than others. Perhaps targeting an increased expression of CB1 at A/C afferents through gene therapy could reduce the number or duration of seizures for epileptic patients due to its ability to reduce synchrony in CA3 pyramidal cells. Future studies should continue to investigate this synapse because it could potentially be a key area for CB1 to be involved in seizure prevention. While we are still in the initial stages of characte rizing the novel cannabinoid effect it is interesting to speculate about how it could change following epileptogenesis and its potential role as a new target for future therapies. If this system was down regulated
137 following epilepsy, similar to the CB1 re ceptor, there would be a potential decrease in ambient GABA and an increase in the excitability of the dentate. Perhaps this system could be important for small adjustments in the overall excitability of the hippocampus. Through changes in the amount of activation, there could be adjustments to the level of ambient GABA causing a general increase or decrease in inhibition. It is also intriguing to consider activation of this novel cannabinoid system as a potential anticonvulsant. Increasing the inhibito ry tone through selective activation of this system could provide a mechanism for reducing the frequency of seizures by increasing the threshold for seizure initiation. In fact, in CA3 pyramidal cells the synthetic cannabinoid agonist HU 210 can reduce bu rsting in a CB1 independent manner. Potentially, HU 210 could be increasing GABA released from local interneurons through our novel cannabinoid effect to decrease bursting in these cells (Goonawardena et al., 2010) Of course the role of this system in e pilepsy is all purely speculative because we need to know more about it first like, for example, whether the unknown receptor also causes an increase in the action potential independent release of glutamate. However, with a clear understanding of the rece ptor, endogenous ligand, and functional significance we will be able to better assess the potential role of the novel cannabinoid effect in epileptogenesis and in future therapeutic strategies. Cholinergics and Epilepsy Besides the cannabinoid system, th ere are also changes to many different signaling systems following epilepsy including the cholinergic system (Friedman et al., 2007) Epileptogenesis can cause decreases in the activity of acetylcholine esterases, the enzyme responsible for the breakdown of acetylcholine (Freitas et al., 2006; Pernot et al., 2009) Changes to this enzyme in the entorhinal cortex resulted in an increased
138 sensitivity to application of acetylcholine often causing seconds long seizure like events (Zimmerman et al., 2008) A t hilar mossy cells the increased presence of acetylcholine could cause an increase in the duration or frequency of ADPs. This could result in increased cannabinoid production reducing the amount of GABA released from a subset of interneurons potentially increasing excitability in the dentate. In addition, changes in the ADPs could impact the excitability of granule cells by either decreasing their excitability through increased drive of hilar interneurons or increasing their excitability through increase d depolarization from the mossy cell axons. Depending on which of these potential outcomes would prevail, the induction of an ADP could play either an anticonvulsant or proconvulsant role, respectively. In fact, the current underlying the ADP, I CAN has been previously implicated in sustaining cortical seizures (Schiller, 2004) While the effect of ADPs on mossy cell output will need to be studied first, it will be interesting to investigate how the ADP changes following epileptogenesis and how that might affect excitability in the dentate gyrus. There is also a possibility that I CAN could be responsible for mossy cell death in epilepsy because this channel has been implicated in excitotoxicity due to its lack of inactivation (Partridge et al., 1994; Tats umi and Katayama, 1994; Partridge and Valenzuela, 1999) Perhaps trauma or ischemia could cause a prolonged ADP leading to mossy cell death. Overall there are several different possibilities to consider about the role ADPs can play in the epileptic brain Cannabinoid System Following Chronic THC A final point of consideration is how chronic drug abuse can affect the different roles for cannabinoids studied in this dissertation. While the effects of THC on memory and cognition have been well documented (C arlini, 2004; Ranganathan and D'Souza, 2006) it has only been more recently that we are beginning to understand the effect of
139 chronic THC on synaptic physiology. Recent research has shown that exposure to THC can inhibit different forms of synaptic plast icity in the hippocampus such as DSI (Mato et al., 2004) and LTP (Hoffman et al., 2007) A similar effect of THC on DSI or ADP DSI in mossy cells has yet to be determined. THC has also been shown to decrease firing and bursting of CA3 pyramidal cells (Go onawardena et al., 2010) and it will be interesting to determine if this is a result of activating CB1 at A/C afferents. In addition, it should be determined whether THC can block DSE at these synapses. Finally, future studies should investigate whether acute THC can cause an increase in action potential independent exocytosis and if so, whether chronic THC changes the ability to observe this novel cannabinoid effect. Investigating the effect of THC on these different cannabinoid mediated forms of plast icity will provide insight to the potential negative effects of THC on memory and cognition at the level of the synapse. Overall the research presented within this dissertation has added significant data to the rich and diverse field of cannabinoids in the hippocampus. Through a greater understanding of the function of this system we can now ask additional questions about the role of cannabinoids in synaptic physiology and how the cannabinoid system changes in diseased or altered states. While further wor k is obviously necessary these data will hopefully begin to lay the foundation for new potential therapeutic strategies by expanding our knowledge of cannabinoids at glutamatergic synapses, strengthening our understanding of the relationship between cholin ergics and the cannabinoid system, and identifying a novel role for cannabinoids in the synaptic physiology of the hippocampus.
140 LIST OF REFERENCES Acsady L, Kamondi A, Sik A, Freund T, Buzsaki G (1998) GABAergic cells are the major postsynaptic targets of mossy fibers in the rat hippocampus. J Neurosci 18:3386 3403. Acsady L, Katona I, Martinez Guijarro FJ, Buzsaki G, Freund TF (2000) Unusual target selectivity of perisomatic inhibitory cells in the hilar region of the rat hippocampus. J Neurosci 20:6907 6 919. Al Hayani A, Davies SN (2002) Effect of cannabinoids on synaptic transmission in the rat hippocampal slice is temperature dependent. Eur J Pharmacol 442:47 54. Ali AB (2007) Presynaptic Inhibition of GABAA receptor mediated unitary IPSPs by cannabin oid receptors at synapses between CCK positive interneurons in rat hippocampus. J Neurophysiol 98:861 869. Amaral DG, Witter MP (1989) The three dimensional organization of the hippocampal formation: a review of anatomical data. Neuroscience 31:571 591. Amaral DG, Lavenex P (2007) Hippocampal neuroantomy. In: The Hippocampus Book (Andersen P, Morris R, Amaral DG, Bliss T, O'Keefe J, eds), pp 37 114. New York City: Oxford University Press. Amaral DG, Ishizuka N, Claiborne B (1990) Neurons, numbers and the hippocampal network. Prog Brain Res 83:1 11. Andersen P, Morris R, Amaral DG, Bliss T, O'Keefe J (2007) Historical perspective: Proposed functions, biological characteristics, and neurobiological models of the hippocampus. In: The Hippocampus Book (Ande rsen P, Morris R, Amaral DG, Bliss T, O'Keefe J, eds), pp 9 22. New York City: Oxford University Press. Anelli V, Bassi R, Tettamanti G, Viani P, Riboni L (2005) Extracellular release of newly synthesized sphingosine 1 phosphate by cerebellar granule cell s and astrocytes. J Neurochem 92:1204 1215. Bains JS, Longacher JM, Staley KJ (1999) Reciprocal interactions between CA3 network activity and strength of recurrent collateral synapses. Nat Neurosci 2:720 726. Beindl W, Mitterauer T, Hohenegger M, Ijzerma n AP, Nanoff C, Freissmuth M (1996) Inhibition of receptor/G protein coupling by suramin analogues. Mol Pharmacol 50:415 423.
141 Blair RE, Deshpande LS, Sombati S, Falenski KW, Martin BR, DeLorenzo RJ (2006) Activation of the cannabinoid type 1 receptor med iates the anticonvulsant properties of cannabinoids in the hippocampal neuronal culture models of acquired epilepsy and status epilepticus. J Pharmacol Exp Ther 317:1072 1078. Bodor AL, Katona I, Nyiri G, Mackie K, Ledent C, Hajos N, Freund TF (2005) Endo cannabinoid signaling in rat somatosensory cortex: laminar differences and involvement of specific interneuron types. J Neurosci 25:6845 6856. Braga MF, Pereira EF, Marchioro M, Albuquerque EX (1999) Lead increases tetrodotoxin insensitive spontaneous rel ease of glutamate and GABA from hippocampal neurons. Brain Res 826:10 21. Breivogel CS, Griffin G, Di Marzo V, Martin BR (2001) Evidence for a new G protein coupled cannabinoid receptor in mouse brain. Mol Pharmacol 60:155 163. Brenowitz SD, Regehr WG (2 005) Associative short term synaptic plasticity mediated by endocannabinoids. Neuron 45:419 431. Brown SP, Brenowitz SD, Regehr WG (2003) Brief presynaptic bursts evoke synapse specific retrograde inhibition mediated by endogenous cannabinoids. Nat Neuros ci 6:1048 1057. Bruno MJ, Koeppe RE, 2nd, Andersen OS (2007) Docosahexaenoic acid alters bilayer elastic properties. Proc Natl Acad Sci U S A 104:9638 9643. Buckmaster PS, Jongen Relo AL (1999) Highly specific neuron loss preserves lateral inhibitory cir cuits in the dentate gyrus of kainate induced epileptic rats. J Neurosci 19:9519 9529. Buckmaster PS, Zhang GF, Yamawaki R (2002) Axon sprouting in a model of temporal lobe epilepsy creates a predominantly excitatory feedback circuit. J Neurosci 22:6650 6 658. Buckmaster PS, Strowbridge BW, Kunkel DD, Schmiege DL, Schwartzkroin PA (1992) Mossy cell axonal projections to the dentate gyrus molecular layer in the rat hippocampal slice. Hippocampus 2:349 362. Bultmann R, Wittenburg H, Pause B, Kurz G, Nickel P, Starke K (1996) P2 purinoceptor antagonists: III. Blockade of P2 purinoceptor subtypes and ecto nucleotidases by compounds related to suramin. Naunyn Schmiedebergs Arch Pharmacol 354:498 504. Buzsaki G (2002) Theta oscillations in the hippocampus. Neur on 33:325 340.
142 Carlini EA (2004) The good and the bad effects of ( ) trans delta 9 tetrahydrocannabinol (Delta 9 THC) on humans. Toxicon 44:461 467. Carlson G, Wang Y, Alger BE (2002) Endocannabinoids facilitate the induction of LTP in the hippocampus. N at Neurosci 5:723 724. Carter AG, Regehr WG (2002) Quantal events shape cerebellar interneuron firing. Nat Neurosci 5:1309 1318. Chevaleyre V, Castillo PE (2003) Heterosynaptic LTD of hippocampal GABAergic synapses: a novel role of endocannabinoids in re gulating excitability. Neuron 38:461 472. Chevaleyre V, Castillo PE (2004) Endocannabinoid mediated metaplasticity in the hippocampus. Neuron 43:871 881. Chevaleyre V, Takahashi KA, Castillo PE (2006) Endocannabinoid mediated synaptic plasticity in the C NS. Annu Rev Neurosci 29:37 76. Chevaleyre V, Heifets BD, Kaeser PS, Sudhof TC, Castillo PE (2007) Endocannabinoid mediated long term plasticity requires cAMP/PKA signaling and RIM1alpha. Neuron 54:801 812. Chiu CQ, Castillo PE (2008) Input specific plas ticity at excitatory synapses mediated by endocannabinoids in the dentate gyrus. Neuropharmacology 54:68 78. Cobb SR, Davies CH (2005) Cholinergic modulation of hippocampal cells and circuits. J Physiol 562:81 88. Colgin LL, Kramar EA, Gall CM, Lynch G ( 2003) Septal modulation of excitatory transmission in hippocampus. J Neurophysiol 90:2358 2366. Colom LV, Saggau P (1994) Spontaneous interictal like activity originates in multiple areas of the CA2 CA3 region of hippocampal slices. J Neurophysiol 71:1574 1585. Cooper ZD, Haney M (2009) Actions of delta 9 tetrahydrocannabinol in cannabis: relation to use, abuse, dependence. Int Rev Psychiatry 21:104 112. D'Antuono M, Kawasaki H, Palmieri C, Curia G, Biagini G, Avoli M (2007) Antiepileptic drugs and musca rinic receptor dependent excitation in the rat subiculum. Neuropharmacology 52:1291 1302. Del Castillo J, Katz B (1954) Quantal components of the end plate potential. J Physiol 124:560 573.
143 Deller T, Katona I, Cozzari C, Frotscher M, Freund TF (1999) Chol inergic innervation of mossy cells in the rat fascia dentata. Hippocampus 9:314 320. Devane WA, Hanus L, Breuer A, Pertwee RG, Stevenson LA, Griffin G, Gibson D, Mandelbaum A, Etinger A, Mechoulam R (1992) Isolation and structure of a brain constituent th at binds to the cannabinoid receptor. Science 258:1946 1949. Di Marzo V (2006) Non CB1 non CB2 receptors for endocannabinoids. In: Endocannabinoids: The brain and body's marijuana and beyond. (ONaivi ES, Sugiura T, Di Marzo V, eds), pp 151 174. Boca Rato n: Taylor and Francis. Di Marzo V, De Petrocellis L, Fezza F, Ligresti A, Bisogno T (2002) Anandamide receptors. Prostaglandins Leukot Essent Fatty Acids 66:377 391. Di Marzo V, Fontana A, Cadas H, Schinelli S, Cimino G, Schwartz JC, Piomelli D (1994) Fo rmation and inactivation of endogenous cannabinoid anandamide in central neurons. Nature 372:686 691. Di Marzo V, Breivogel CS, Tao Q, Bridgen DT, Razdan RK, Zimmer AM, Zimmer A, Martin BR (2000) Levels, metabolism, and pharmacological activity of anandam ide in CB(1) cannabinoid receptor knockout mice: evidence for non CB(1), non CB(2) receptor mediated actions of anandamide in mouse brain. J Neurochem 75:2434 2444. Diana MA, Marty A (2004) Endocannabinoid mediated short term synaptic plasticity: depolari zation induced suppression of inhibition (DSI) and depolarization induced suppression of excitation (DSE). Br J Pharmacol 142:9 19. Domenici MR, Azad SC, Marsicano G, Schierloh A, Wotjak CT, Dodt HU, Zieglgansberger W, Lutz B, Rammes G (2006) Cannabinoid receptor type 1 located on presynaptic terminals of principal neurons in the forebrain controls glutamatergic synaptic transmission. J Neurosci 26:5794 5799. Dzhala VI, Staley KJ (2003) Transition from interictal to ictal activity in limbic networks in vi tro. J Neurosci 23:7873 7880. Edwards DA, Kim J, Alger BE (2006) Multiple mechanisms of endocannabinoid response initiation in hippocampus. J Neurophysiol 95:67 75. Edwards DA, Zhang L, Alger BE (2008) Metaplastic control of the endocannabinoid system at inhibitory synapses in hippocampus. Proc Natl Acad Sci U S A 105:8142 8147. Fain GL (1999) Molecular and cellular physiology of neurons. Cambridge, Mass.: Harvard University Press.
144 Fatt P, Katz B (1952) Spontaneous subthreshold activity at motor nerve en dings. J Physiol 117:109 128. Feinmark SJ, Begum R, Tsvetkov E, Goussakov I, Funk CD, Siegelbaum SA, Bolshakov VY (2003) 12 lipoxygenase metabolites of arachidonic acid mediate metabotropic glutamate receptor dependent long term depression at hippocampal CA3 CA1 synapses. J Neurosci 23:11427 11435. Fraser DD, MacVicar BA (1996) Cholinergic dependent plateau potential in hippocampal CA1 pyramidal neurons. J Neurosci 16:4113 4128. Frazier CJ (2007) Endocannabinoids in the dentate gyrus. Prog Brain Res 163: 319 815. Frazier CJ, Strowbridge BW, Papke RL (2003) Nicotinic receptors on local circuit neurons in dentate gyrus: a potential role in regulation of granule cell excitability. J Neurophysiol 89:3018 3028. Freissmuth M, Boehm S, Beindl W, Nickel P, Ijzer man AP, Hohenegger M, Nanoff C (1996) Suramin analogues as subtype selective G protein inhibitors. Mol Pharmacol 49:602 611. Freitas RM, Sousa FC, Viana GS, Fonteles MM (2006) Acetylcholinesterase activities in hippocampus, frontal cortex and striatum of Wistar rats after pilocarpine induced status epilepticus. Neurosci Lett 399:76 78. Freund TF, Magloczky Z (1993) Early degeneration of calretinin containing neurons in the rat hippocampus after ischemia. Neuroscience 56:581 596. Freund TF, Buzsaki G (199 6) Interneurons of the hippocampus. Hippocampus 6:347 470. Freund TF, Katona I, Piomelli D (2003) Role of endogenous cannabinoids in synaptic signaling. Physiol Rev 83:1017 1066. Friedman A, Behrens CJ, Heinemann U (2007) Cholinergic dysfunction in tempo ral lobe epilepsy. Epilepsia 48 Suppl 5:126 130. Frotscher M, Leranth C (1985) Cholinergic innervation of the rat hippocampus as revealed by choline acetyltransferase immunocytochemistry: a combined light and electron microscopic study. J Comp Neurol 239: 237 246. Frotscher M, Vida I, Bender R (2000) Evidence for the existence of non GABAergic, cholinergic interneurons in the rodent hippocampus. Neuroscience 96:27 31.
145 Fukudome Y, Ohno Shosaku T, Matsui M, Omori Y, Fukaya M, Tsubokawa H, Taketo MM, Watana be M, Manabe T, Kano M (2004) Two distinct classes of muscarinic action on hippocampal inhibitory synapses: M2 mediated direct suppression and M1/M3 mediated indirect suppression through endocannabinoid signalling. Eur J Neurosci 19:2682 2692. Gaoni Y, Me choulam R (1964) Isolation, Structure, and Partial Synthesis of an Active Constituent of Hashish. Journal of the American Chemical Society 86:1646 1647. Ghamari Langroudi M, Bourque CW (2002) Flufenamic acid blocks depolarizing afterpotentials and phasic firing in rat supraoptic neurones. J Physiol 545:537 542. Glitsch MD (2008) Spontaneous neurotransmitter release and Ca2+ -how spontaneous is spontaneous neurotransmitter release? Cell Calcium 43:9 15. Gold PE (2003) Acetylcholine modulation of neural sy stems involved in learning and memory. Neurobiol Learn Mem 80:194 210. Goonawardena AV, Riedel G, Hampson RE (2010) Cannabinoids alter spontaneous firing, bursting, and cell synchrony of hippocampal principal cells. Hippocampus. Gorter JA, van Vliet EA, Aronica E, Lopes da Silva FH (2001) Progression of spontaneous seizures after status epilepticus is associated with mossy fibre sprouting and extensive bilateral loss of hilar parvalbumin and somatostatin immunoreactive neurons. Eur J Neurosci 13:657 669. Goutagny R, Manseau F, Jackson J, Danik M, Williams S (2008) In vitro activation of the medial septum diagonal band complex generates atropine sensitive and atropine resistant hippocampal theta rhythm: an investigation using a complete septohippocampal pr eparation. Hippocampus 18:531 535. Groemer TW, Klingauf J (2007) Synaptic vesicles recycling spontaneously and during activity belong to the same vesicle pool. Nat Neurosci 10:145 147. Gulyas AI, Cravatt BF, Bracey MH, Dinh TP, Piomelli D, Boscia F, Freu nd TF (2004) Segregation of two endocannabinoid hydrolyzing enzymes into pre and postsynaptic compartments in the rat hippocampus, cerebellum and amygdala. Eur J Neurosci 20:441 458. Haj Dahmane S, Andrade R (1996) Muscarinic activation of a voltage depe ndent cation nonselective current in rat association cortex. J Neurosci 16:3848 3861. Haj Dahmane S, Andrade R (1998) Ionic mechanism of the slow afterdepolarization induced by muscarinic receptor activation in rat prefrontal cortex. J Neurophysiol 80:119 7 1210.
146 Hajos N, Freund TF (2002a) Distinct cannabinoid sensitive receptors regulate hippocampal excitation and inhibition. Chem Phys Lipids 121:73 82. Hajos N, Freund TF (2002b) Pharmacological separation of cannabinoid sensitive receptors on hippocampal excitatory and inhibitory fibers. Neuropharmacology 43:503 510. Hajos N, Ledent C, Freund TF (2001) Novel cannabinoid sensitive receptor mediates inhibition of glutamatergic synaptic transmission in the hippocampus. Neuroscience 106:1 4. Hajos N, Katona I, Naiem SS, MacKie K, Ledent C, Mody I, Freund TF (2000) Cannabinoids inhibit hippocampal GABAergic transmission and network oscillations. Eur J Neurosci 12:3239 3249. Haller J, Matyas F, Soproni K, Varga B, Barsy B, Nemeth B, Mikics E, Freund TF, Hajos N (2007) Correlated species differences in the effects of cannabinoid ligands on anxiety and on GABAergic and glutamatergic synaptic transmission. Eur J Neurosci 25:2445 2456. Hampson RE, Zhuang SY, Weiner JL, Deadwyler SA (2003) Functional significance of cannabinoid mediated, depolarization induced suppression of inhibition (DSI) in the hippocampus. J Neurophysiol 90:55 64. Hashimotodani Y, Ohno Shosaku T, Maejima T, Fukami K, Kano M (2008) Pharmacological evidence for the involvement of diacylglycerol lipase in depolarization induced endocanabinoid release. Neuropharmacology 54:58 67. Hashimotodani Y, Ohno Shosaku T, Tsubokawa H, Ogata H, Emoto K, Maejima T, Araishi K, Shin HS, Kano M (2005) Phospholipase Cbeta serves as a coincidence detector through its Ca2+ dependency for triggering retrograde endocannabinoid signal. Neuron 45:257 268. Hasselmo ME (2006) The role of acetylcholine in learning and memory. Curr Opin Neurobiol 16:710 715. Heifets BD, Chevaleyre V, Castillo PE (2008) Interneuron activi ty controls endocannabinoid mediated presynaptic plasticity through calcineurin. Proc Natl Acad Sci U S A 105:10250 10255. Henze DA, Buzsaki G (2007) Hilar mossy cells: functional identification and activity in vivo. Prog Brain Res 163:199 216. Henze DA, Urban NN, Barrionuevo G (2000) The multifarious hippocampal mossy fiber pathway: a review. Neuroscience 98:407 427.
147 Herkenham M, Lynn AB, Johnson MR, Melvin LS, de Costa BR, Rice KC (1991) Characterization and localization of cannabinoid receptors in rat brain: a quantitative in vitro autoradiographic study. J Neurosci 11:563 583. Hillard CJ (2000) Biochemistry and pharmacology of the endocannabinoids arachidonylethanolamide and 2 arachidonylglycerol. Prostaglandins Other Lipid Mediat 61:3 18. Hoffman AF Macgill AM, Smith D, Oz M, Lupica CR (2005) Species and strain differences in the expression of a novel glutamate modulating cannabinoid receptor in the rodent hippocampus. Eur J Neurosci 22:2387 2391. Hoffman AF, Oz M, Yang R, Lichtman AH, Lupica CR (2 007) Opposing actions of chronic Delta9 tetrahydrocannabinol and cannabinoid antagonists on hippocampal long term potentiation. Learn Mem 14:63 74. Hoffman AF, Laaris N, Kawamura M, Masino SA, Lupica CR (2010) Control of cannabinoid CB1 receptor function on glutamate axon terminals by endogenous adenosine acting at A1 receptors. J Neurosci 30:545 555. Hofmann ME, Frazier CJ (2010) Muscarinic receptor activation modulates the excitability of hilar mossy cells through the induction of an afterdepolarization Brain Res 1318:42 51. Hofmann ME, Nahir B, Frazier CJ (2006) Endocannabinoid mediated depolarization induced suppression of inhibition in hilar mossy cells of the rat dentate gyrus. J Neurophysiol 96:2501 2512. Hofmann ME, Nahir B, Frazier CJ (2008) Ex citatory afferents to CA3 pyramidal cells display differential sensitivity to CB1 dependent inhibition of synaptic transmission. Neuropharmacology 55:1140 1146. Houser CR (1999) Neuronal loss and synaptic reorganization in temporal lobe epilepsy. Adv Neur ol 79:743 761. Houser CR (2007) Interneurons of the dentate gyrus: an overview of cell types, terminal fields and neurochemical identity. Prog Brain Res 163:217 232. Howard AL, Neu A, Morgan RJ, Echegoyen JC, Soltesz I (2007) Opposing modifications in in trinsic currents and synaptic inputs in post traumatic mossy cells: evidence for single cell homeostasis in a hyperexcitable network. J Neurophysiol 97:2394 2409.
148 Hsu M, Buzsaki G (1993) Vulnerability of mossy fiber targets in the rat hippocampus to fore brain ischemia. J Neurosci 13:3964 3979. Isokawa M, Alger BE (2005) Retrograde endocannabinoid regulation of GABAergic inhibition in the rat dentate gyrus granule cell. J Physiol 567:1001 1010. Jackson MB, Scharfman HE (1996) Positive feedback from hilar mossy cells to granule cells in the dentate gyrus revealed by voltage sensitive dye and microelectrode recording. J Neurophysiol 76:601 616. Jensen MS, Azouz R, Yaari Y (1996) Spike after depolarization and burst generation in adult rat hippocampal CA1 p yramidal cells. J Physiol 492 ( Pt 1):199 210. Kajimoto T, Okada T, Yu H, Goparaju SK, Jahangeer S, Nakamura S (2007) Involvement of sphingosine 1 phosphate in glutamate secretion in hippocampal neurons. Mol Cell Biol 27:3429 3440. Kano M, Ohno Shosaku T Hashimotodani Y, Uchigashima M, Watanabe M (2009) Endocannabinoid mediated control of synaptic transmission. Physiol Rev 89:309 380. Karler R, Turkanis SA (1981) The cannabinoids as potential antiepileptics. J Clin Pharmacol 21:437S 448S. Katona I, Spe rlagh B, Sik A, Kafalvi A, Vizi ES, Mackie K, Freund TF (1999) Presynaptically located CB1 cannabinoid receptors regulate GABA release from axon terminals of specific hippocampal interneurons. J Neurosci 19:4544 4558. Katona I, Urban GM, Wallace M, Ledent C, Jung KM, Piomelli D, Mackie K, Freund TF (2006) Molecular composition of the endocannabinoid system at glutamatergic synapses. J Neurosci 26:5628 5637. Kawamura Y, Fukaya M, Maejima T, Yoshida T, Miura E, Watanabe M, Ohno Shosaku T, Kano M (2006) The CB1 cannabinoid receptor is the major cannabinoid receptor at excitatory presynaptic sites in the hippocampus and cerebellum. J Neurosci 26:2991 3001. Kerr AM, Capogna M (2007) Unitary IPSPs enhance hilar mossy cell gain in the rat hippocampus. J Physiol 578:451 470. Kim J, Isokawa M, Ledent C, Alger BE (2002) Activation of muscarinic acetylcholine receptors enhances the release of endogenous cannabinoids in the hippocampus. J Neurosci 22:10182 10191.
149 Kirby MT, Hampson RE, Deadwyler SA (1995) Cannabinoid s selectively decrease paired pulse facilitation of perforant path synaptic potentials in the dentate gyrus in vitro. Brain Res 688:114 120. Kiss T (2005) G protein coupled activation of potassium channels by endogenous neuropeptides in snail neurons. Eur J Neurosci 21:2177 2185. Koch M, Kreutz S, Bottger C, Grabiec U, Ghadban C, Korf HW, Dehghani F (2010) The cannabinoid WIN 55,212 2 mediated protection of dentate gyrus granule cells is driven by CB(1) receptors and modulated by TRPA1 and Ca(v)2.2 channe ls. Hippocampus. Kofalvi A, Vizi ES, Ledent C, Sperlagh B (2003) Cannabinoids inhibit the release of [3H]glutamate from rodent hippocampal synaptosomes via a novel CB1 receptor independent action. Eur J Neurosci 18:1973 1978. Kreitzer AC, Regehr WG (2001 a) Retrograde inhibition of presynaptic calcium influx by endogenous cannabinoids at excitatory synapses onto Purkinje cells. Neuron 29:717 727. Kreitzer AC, Regehr WG (2001b) Cerebellar depolarization induced suppression of inhibition is mediated by endo genous cannabinoids. J Neurosci 21:RC174. Kreitzer FR, Stella N (2009) The therapeutic potential of novel cannabinoid receptors. Pharmacol Ther 122:83 96. Kuster JE, Stevenson JI, Ward SJ, D'Ambra TE, Haycock DA (1993) Aminoalkylindole binding in rat cer ebellum: selective displacement by natural and synthetic cannabinoids. J Pharmacol Exp Ther 264:1352 1363. Larimer P, Strowbridge BW (2008) Nonrandom local circuits in the dentate gyrus. J Neurosci 28:12212 12223. Larimer P, Strowbridge BW (2010) Represe nting information in cell assemblies: persistent activity mediated by semilunar granule cells. Nat Neurosci 13:213 222. Lauckner JE, Jensen JB, Chen HY, Lu HC, Hille B, Mackie K (2008) GPR55 is a cannabinoid receptor that increases intracellular calcium a nd inhibits M current. Proc Natl Acad Sci U S A 105:2699 2704. Lawrence JJ (2008) Cholinergic control of GABA release: emerging parallels between neocortex and hippocampus. Trends Neurosci 31:317 327. Lawrence JJ, Grinspan ZM, McBain CJ (2004) Quantal tr ansmission at mossy fibre targets in the CA3 region of the rat hippocampus. J Physiol 554:175 193.
150 Lawrence JJ, Grinspan ZM, Statland JM, McBain CJ (2006a) Muscarinic receptor activation tunes mouse stratum oriens interneurones to amplify spike reliability J Physiol 571:555 562. Lawrence JJ, Statland JM, Grinspan ZM, McBain CJ (2006b) Cell type specific dependence of muscarinic signalling in mouse hippocampal stratum oriens interneurones. J Physiol 570:595 610. Li YX, Zhang Y, Lester HA, Schuman EM, Davi dson N (1998) Enhancement of neurotransmitter release induced by brain derived neurotrophic factor in cultured hippocampal neurons. J Neurosci 18:10231 10240. Llano I, Gonzalez J, Caputo C, Lai FA, Blayney LM, Tan YP, Marty A (2000) Presynaptic calcium stores underlie large amplitude miniature IPSCs and spontaneous calcium transients. Nat Neurosci 3:1256 1265. Losonczy A, Biro AA, Nusser Z (2004) Persistently a ctive cannabinoid receptors mute a subpopulation of hippocampal interneurons. Proc Natl Acad Sci U S A 101:1362 1367. Lothman EW, Bertram EH, 3rd, Kapur J, Bekenstein JW (1996) Temporal lobe epilepsy: studies in a rat model showing dormancy of GABAergic i nhibitory interneurons. Epilepsy Res Suppl 12:145 156. Lozovaya N, Yatsenko N, Beketov A, Tsintsadze T, Burnashev N (2005) Glycine receptors in CNS neurons as a target for nonretrograde action of cannabinoids. J Neurosci 25:7499 7506. Lu T, Trussell LO ( 2000) Inhibitory transmission mediated by asynchronous transmitter release. Neuron 26:683 694. Ludanyi A, Eross L, Czirjak S, Vajda J, Halasz P, Watanabe M, Palkovits M, Magloczky Z, Freund TF, Katona I (2008) Downregulation of the CB1 cannabinoid recepto r and related molecular elements of the endocannabinoid system in epileptic human hippocampus. J Neurosci 28:2976 2990. Lundbaek JA (2008) Lipid bilayer mediated regulation of ion channel function by amphiphilic drugs. J Gen Physiol 131:421 429. Lundbaek JA, Birn P, Tape SE, Toombes GE, Sogaard R, Koeppe RE, 2nd, Gruner SM, Hansen AJ, Andersen OS (2005) Capsaicin regulates voltage dependent sodium channels by altering lipid bilayer elasticity. Mol Pharmacol 68:680 689.
151 Lundbaek JA, Birn P, Hansen AJ, So gaard R, Nielsen C, Girshman J, Bruno MJ, Tape SE, Egebjerg J, Greathouse DV, Mattice GL, Koeppe RE, 2nd, Andersen OS (2004) Regulation of sodium channel function by bilayer elasticity: the importance of hydrophobic coupling. Effects of Micelle forming amp hiphiles and cholesterol. J Gen Physiol 123:599 621. Lutz B (2004) On demand activation of the endocannabinoid system in the control of neuronal excitability and epileptiform seizures. Biochem Pharmacol 68:1691 1698. Lynch MA (2004) Long term potentiatio n and memory. Physiol Rev 84:87 136. Mackie K (2008) Signaling via CNS cannabinoid receptors. Mol Cell Endocrinol 286:S60 65. Maejima T, Hashimoto K, Yoshida T, Aiba A, Kano M (2001) Presynaptic inhibition caused by retrograde signal from metabotropic gl utamate to cannabinoid receptors. Neuron 31:463 475. Magloczky Z, Freund TF (1993) Selective neuronal death in the contralateral hippocampus following unilateral kainate injections into the CA3 subfield. Neuroscience 56:317 335. Maingret F, Patel AJ, Laz dunski M, Honore E (2001) The endocannabinoid anandamide is a direct and selective blocker of the background K(+) channel TASK 1. EMBO J 20:47 54. Makara JK, Katona I, Nyiri G, Nemeth B, Ledent C, Watanabe M, de Vente J, Freund TF, Hajos N (2007) Involvem ent of nitric oxide in depolarization induced suppression of inhibition in hippocampal pyramidal cells during activation of cholinergic receptors. J Neurosci 27:10211 10222. Marsicano G, Goodenough S, Monory K, Hermann H, Eder M, Cannich A, Azad SC, Casci o MG, Gutierrez SO, van der Stelt M, Lopez Rodriguez ML, Casanova E, Schutz G, Zieglgansberger W, Di Marzo V, Behl C, Lutz B (2003) CB1 cannabinoid receptors and on demand defense against excitotoxicity. Science 302:84 88. Massey PV, Bashir ZI (2007) Long term depression: multiple forms and implications for brain function. Trends Neurosci 30:176 184. Mathew SS, Pozzo Miller L, Hablitz JJ (2008) Kainate modulates presynaptic GABA release from two vesicle pools. J Neurosci 28:725 731.
152 Mato S, Chevaleyre V, Robbe D, Pazos A, Castillo PE, Manzoni OJ (2004) A single in vivo exposure to delta 9THC blocks endocannabinoid mediated synaptic plasticity. Nat Neurosci 7:585 586. Matsuda LA, Lolait SJ, Brownstein MJ, Young AC, Bonner TI (1990) Structure of a cannabin oid receptor and functional expression of the cloned cDNA. Nature 346:561 564. Mattson MP (2007) Mitochondrial regulation of neuronal plasticity. Neurochem Res 32:707 715. McKinney RA, Capogna M, Durr R, Gahwiler BH, Thompson SM (1999) Miniature synaptic events maintain dendritic spines via AMPA receptor activation. Nat Neurosci 2:44 49. McLaren GJ, Kennedy C, Sneddon P (1995) The effects of suramin on purinergic and noradrenergic neurotransmission in the rat isolated tail artery. Eur J Pharmacol 277:57 61. McQuiston AR, Madison DV (1999) Muscarinic receptor activity induces an afterdepolarization in a subpopulation of hippocampal CA1 interneurons. J Neurosci 19:5703 5710. Monory K et al. (2006) The endocannabinoid system controls key epileptogenic circ uits in the hippocampus. Neuron 51:455 466. Munro S, Thomas KL, Abu Shaar M (1993) Molecular characterization of a peripheral receptor for cannabinoids. Nature 365:61 65. Myoga MH, Beierlein M, Regehr WG (2009) Somatic spikes regulate dendritic signaling in small neurons in the absence of backpropagating action potentials. J Neurosci 29:7803 7814. Nahir B, Bhatia C, Frazier CJ (2007) Presynaptic inhibition of excitatory afferents to hilar mossy cells. J Neurophysiol 97:4036 4047. Nahir B, Lindsly C, Fra zier CJ (2010) mGluR mediated and endocannabinoid dependent long term depression in the hilar region of the rat dentate gyrus. Neuropharmacology 58:712 721. Navarrete M, Araque A (2008) Endocannabinoids mediate neuron astrocyte communication. Neuron 57:88 3 893. Nemeth B, Ledent C, Freund TF, Hajos N (2008) CB1 receptor dependent and independent inhibition of excitatory postsynaptic currents in the hippocampus by WIN 55,212 2. Neuropharmacology 54:51 57.
153 Nicoll RA, Schmitz D (2005) Synaptic plasticity at hippocampal mossy fibre synapses. Nat Rev Neurosci 6:863 876. Niewiadomska G, Baksalerska Pazera M, Riedel G (2009) The septo hippocampal system, learning and recovery of function. Prog Neuropsychopharmacol Biol Psychiatry 33:791 805. Nyilas R, Dudok B, Urban GM, Mackie K, Watanabe M, Cravatt BF, Freund TF, Katona I (2008) Enzymatic machinery for endocannabinoid biosynthesis associated with calcium stores in glutamatergic axon terminals. J Neurosci 28:1058 1063. O'Sullivan SE (2007) Cannabinoids go nucle ar: evidence for activation of peroxisome proliferator activated receptors. Br J Pharmacol 152:576 582. Ohno Shosaku T, Maejima T, Kano M (2001) Endogenous cannabinoids mediate retrograde signals from depolarized postsynaptic neurons to presynaptic termin als. Neuron 29:729 738. Ohno Shosaku T, Tsubokawa H, Mizushima I, Yoneda N, Zimmer A, Kano M (2002) Presynaptic cannabinoid sensitivity is a major determinant of depolarization induced retrograde suppression at hippocampal synapses. J Neurosci 22:3864 387 2. Ohno Shosaku T, Matsui M, Fukudome Y, Shosaku J, Tsubokawa H, Taketo MM, Manabe T, Kano M (2003) Postsynaptic M1 and M3 receptors are responsible for the muscarinic enhancement of retrograde endocannabinoid signalling in the hippocampus. Eur J Neurosci 18:109 116. Oz M (2006a) Receptor independent actions of cannabinoids on cell membranes: focus on endocannabinoids. Pharmacol Ther 111:114 144. Oz M (2006b) Receptor independent effects of endocannabinoids on ion channels. Curr Pharm Des 12:227 239. Pa n CY, Lee H, Chen CL (2006) Lysophospholipids elevate [Ca2+]i and trigger exocytosis in bovine chromaffin cells. Neuropharmacology 51:18 26. Partridge LD (1994) Cytoplasmic Ca2+ activity regulation as measured by a calcium activated current. Brain Res 647 :76 82. Partridge LD, Valenzuela CF (1999) Ca2+ store dependent potentiation of Ca2+ activated non selective cation channels in rat hippocampal neurones in vitro. J Physiol 521 Pt 3:617 627. Partridge LD, Valenzuela CF (2000) Block of hippocampal CAN cha nnels by flufenamate. Brain Res 867:143 148.
154 Partridge LD, Muller TH, Swandulla D (1994) Calcium activated non selective channels in the nervous system. Brain Res Brain Res Rev 19:319 325. Paugh SW, Cassidy MP, He H, Milstien S, Sim Selley LJ, Spiegel S, Selley DE (2006) Sphingosine and its analog, the immunosuppressant 2 amino 2 (2 [4 octylphenyl]ethyl) 1,3 propanediol, interact with the CB1 cannabinoid receptor. Mol Pharmacol 70:41 50. Pernot F, Carpentier P, Baille V, Testylier G, Beaup C, Foquin A, Fi lliat P, Liscia P, Coutan M, Pierard C, Beracochea D, Dorandeu F (2009) Intrahippocampal cholinesterase inhibition induces epileptogenesis in mice without evidence of neurodegenerative events. Neuroscience 162:1351 1365. Popescu IR, Morton LA, Franco A, D i S, Ueta Y, Tasker JG (2010) Synchronized bursts of miniature inhibitory postsynaptic currents. J Physiol 588:939 951. Prange O, Murphy TH (1999) Correlation of miniature synaptic activity and evoked release probability in cultures of cortical neurons. J Neurosci 19:6427 6438. Pressler RT, Strowbridge BW (2006) Blanes cells mediate persistent feedforward inhibition onto granule cells in the olfactory bulb. Neuron 49:889 904. Pressler RT, Inoue T, Strowbridge BW (2007) Muscarinic receptor activation modu lates granule cell excitability and potentiates inhibition onto mitral cells in the rat olfactory bulb. J Neurosci 27:10969 10981. Ranganathan M, D'Souza DC (2006) The acute effects of cannabinoids on memory in humans: a review. Psychopharmacology (Berl) 188:425 444. Ratzliff AH, Santhakumar V, Howard A, Soltesz I (2002) Mossy cells in epilepsy: rigor mortis or vigor mortis? Trends Neurosci 25:140 144. Razani Boroujerdi S, Partridge LD (1993) Activation and modulation of calcium activated non selective c ation channels from embryonic chick sensory neurons. Brain Res 623:195 200. Rhee HJ, Nam JS, Sun Y, Kim MJ, Choi HK, Han DH, Kim NH, Huh SO (2006) Lysophosphatidic acid stimulates cAMP accumulation and cAMP response element binding protein phosphorylation in immortalized hippocampal progenitor cells. Neuroreport 17:523 526. Rolls ET, Kesner RP (2006) A computational theory of hippocampal function, and empirical tests of the theory. Prog Neurobiol 79:1 48. Ruiz A, Fabian Fine R, Scott R, Walker MC, Rusako v DA, Kullmann DM (2003) GABAA receptors at hippocampal mossy fibers. Neuron 39:961 973.
155 Ryberg E, Larsson N, Sjogren S, Hjorth S, Hermansson NO, Leonova J, Elebring T, Nilsson K, Drmota T, Greasley PJ (2007) The orphan receptor GPR55 is a novel cannabinoi d receptor. Br J Pharmacol 152:1092 1101. Saitoe M, Schwarz TL, Umbach JA, Gundersen CB, Kidokoro Y (2001) Absence of junctional glutamate receptor clusters in Drosophila mutants lacking spontaneous transmitter release. Science 293:514 517. Sang N, Zhang J, Chen C (2006) PGE2 glycerol ester, a COX 2 oxidative metabolite of 2 arachidonoyl glycerol, modulates inhibitory synaptic transmission in mouse hippocampal neurons. J Physiol 572:735 745. Sang N, Zhang J, Marcheselli V, Bazan NG, Chen C (2005) Postsyn aptically synthesized prostaglandin E2 (PGE2) modulates hippocampal synaptic transmission via a presynaptic PGE2 EP2 receptor. J Neurosci 25:9858 9870. Santhakumar V, Bender R, Frotscher M, Ross ST, Hollrigel GS, Toth Z, Soltesz I (2000) Granule cell hype rexcitability in the early post traumatic rat dentate gyrus: the 'irritable mossy cell' hypothesis. J Physiol 524 Pt 1:117 134. Sara Y, Virmani T, Deak F, Liu X, Kavalali ET (2005) An isolated pool of vesicles recycles at rest and drives spontaneous neuro transmission. Neuron 45:563 573. Scharfman HE (1995) Electrophysiological evidence that dentate hilar mossy cells are excitatory and innervate both granule cells and interneurons. J Neurophysiol 74:179 194. Scharfman HE (2007) The CA3 "backprojection" to the dentate gyrus. Prog Brain Res 163:627 637. Scharfman HE, Smith KL, Goodman JH, Sollas AL (2001) Survival of dentate hilar mossy cells after pilocarpine induced seizures and their synchronized burst discharges with area CA3 pyramidal cells. Neuroscien ce 104:741 759. Schiller Y (2004) Activation of a calcium activated cation current during epileptiform discharges and its possible role in sustaining seizure like events in neocortical slices. J Neurophysiol 92:862 872. Sciancalepore M, Savic N, Gyori J, Cherubini E (1998) Facilitation of miniature GABAergic currents by ruthenium red in neonatal rat hippocampal neurons. J Neurophysiol 80:2316 2322. Scoville WB, Milner B (1957) Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry 20:11 21.
156 Sharma G, Vijayaraghavan S (2003) Modulation of presynaptic store calcium induces release of glutamate and postsynaptic firing. Neuron 38:929 939. Shen M, Thayer SA (1998) The cannabinoid agonist Win55,212 2 inhibits calcium channels by receptor mediated and direct pathways in cultured rat hippocampal neurons. Brain Res 783:77 84. Skaper SD, Buriani A, Dal Toso R, Petrelli L, Romanello S, Facci L, Leon A (1996) The ALIAmide palmitoylethanolamide and cannabinoids, but not anandamide, are protective in a delayed postglutamate paradigm of excitotoxic death in cerebellar granule neurons. Proc Natl Acad Sci U S A 93:3984 3989. Slanina KA, Schweitzer P (2005) Inhibition of cyclooxygenase 2 elicits a CB1 mediated decrease of excitatory tran smission in rat CA1 hippocampus. Neuropharmacology 49:653 659. Sloviter RS, Zappone CA, Harvey BD, Bumanglag AV, Bender RA, Frotscher M (2003) "Dormant basket cell" hypothesis revisited: relative vulnerabilities of dentate gyrus mossy cells and inhibitory interneurons after hippocampal status epilepticus in the rat. J Comp Neurol 459:44 76. Smith AJ, Sugita S, Charlton MP (2010) Cholesterol dependent kinase activity regulates transmitter release from cerebellar synapses. J Neurosci 30:6116 6121. Sogaard R, Werge TM, Bertelsen C, Lundbye C, Madsen KL, Nielsen CH, Lundbaek JA (2006) GABA(A) receptor function is regulated by lipid bilayer elasticity. Biochemistry 45:13118 13129. Squire LR, Zola Morgan S (1991) The medial temporal lobe memory system. Science 253:1380 1386. Stark C (2007) Functional role of the human hippocampus. In: The Hippocampus Book (Andersen P, Morris R, Amaral DG, Bliss T, O'Keefe J, eds), pp 549 580. New York City: Oxford University Press. Stella N (2009) Endocannabinoid signaling in microglial cells. Neuropharmacology 56 Suppl 1:244 253. Stella N, Schweitzer P, Piomelli D (1997) A second endogenous cannabinoid that modulates long term potentiation. Nature 388:773 778. Sudhof TC (2004) The synaptic vesicle cycle. Annu Rev Neurosci 2 7:509 547. Sugiura T, Kobayashi Y, Oka S, Waku K (2002) Biosynthesis and degradation of anandamide and 2 arachidonoylglycerol and their possible physiological significance. Prostaglandins Leukot Essent Fatty Acids 66:173 192.
157 Sugiura T, Kishimoto S, Oka S Gokoh M (2006) Biochemistry, pharmacology and physiology of 2 arachidonoylglycerol, an endogenous cannabinoid receptor ligand. Prog Lipid Res 45:405 446. Sun J, Pang ZP, Qin D, Fahim AT, Adachi R, Sudhof TC (2007) A dual Ca2+ sensor model for neurotrans mitter release in a central synapse. Nature 450:676 682. Sutton MA, Wall NR, Aakalu GN, Schuman EM (2004) Regulation of dendritic protein synthesis by miniature synaptic events. Science 304:1979 1983. Takahashi KA, Castillo PE (2006) The CB1 cannabinoid receptor mediates glutamatergic synaptic suppression in the hippocampus. Neuroscience 139:795 802. Talley EM, Solorzano G, Lei Q, Kim D, Bayliss DA (2001) Cns distribution of members of the two pore domain (KCNK) potassium channel family. J Neurosci 21:74 91 7505. Tatsumi H, Katayama Y (1994) Brief increases in intracellular Ca2+ activate K+ current and non selective cation current in rat nucleus basalis neurons. Neuroscience 58:553 561. Taverna S, Tkatch T, Metz AE, Martina M (2005) Differential expressi on of TASK channels between horizontal interneurons and pyramidal cells of rat hippocampus. J Neurosci 25:9162 9170. Tigyi G, Parrill AL (2003) Molecular mechanisms of lysophosphatidic acid action. Prog Lipid Res 42:498 526. Torborg CL, Berg AP, Jeffries BW, Bayliss DA, McBain CJ (2006) TASK like conductances are present within hippocampal CA1 stratum oriens interneuron subpopulations. J Neurosci 26:7362 7367. Toth K, Freund TF, Miles R (1997) Disinhibition of rat hippocampal pyramidal cells by GABAergic afferents from the septum. J Physiol 500 ( Pt 2):463 474. Trettel J, Levine ES (2003) Endocannabinoids mediate rapid retrograde signaling at interneuron right arrow pyramidal neuron synapses of the neocortex. J Neurophysiol 89:2334 2338. Trimbuch T et a l. (2009) Synaptic PRG 1 modulates excitatory transmission via lipid phosphate mediated signaling. Cell 138:1222 1235. Trudeau LE, Emery DG, Haydon PG (1996a) Direct modulation of the secretory machinery underlies PKA dependent synaptic facilitation in hi ppocampal neurons. Neuron 17:789 797.
158 Trudeau LE, Doyle RT, Emery DG, Haydon PG (1996b) Calcium independent activation of the secretory apparatus by ruthenium red in hippocampal neurons: a new tool to assess modulation of presynaptic function. J Neurosci 1 6:46 54. Tsou K, Mackie K, Sanudo Pena MC, Walker JM (1999) Cannabinoid CB1 receptors are localized primarily on cholecystokinin containing GABAergic interneurons in the rat hippocampal formation. Neuroscience 93:969 975. Tsou K, Brown S, Sanudo Pena MC, Mackie K, Walker JM (1998) Immunohistochemical distribution of cannabinoid CB1 receptors in the rat central nervous system. Neuroscience 83:393 411. Van Sickle MD, Duncan M, Kingsley PJ, Mouihate A, Urbani P, Mackie K, Stella N, Makriyannis A, Piomelli D Davison JS, Marnett LJ, Di Marzo V, Pittman QJ, Patel KD, Sharkey KA (2005) Identification and functional characterization of brainstem cannabinoid CB2 receptors. Science 310:329 332. Vida I, Frotscher M (2000) A hippocampal interneuron associated with the mossy fiber system. Proc Natl Acad Sci U S A 97:1275 1280. Vogt KE, Regehr WG (2001) Cholinergic modulation of excitatory synaptic transmission in the CA3 area of the hippocampus. J Neurosci 21:75 83. Volpicelli LA, Levey AI (2004) Muscarinic acetylc holine receptor subtypes in cerebral cortex and hippocampus. Prog Brain Res 145:59 66. Walker M, Chan D, Thom M (2007) Hippocampus and human disease. In: The Hippocampus Book (Andersen P, Morris R, Amaral DG, Bliss T, O'Keefe J, eds), pp 769 802. New York City: Oxford University Press. Wallace MJ, Martin BR, DeLorenzo RJ (2002) Evidence for a physiological role of endocannabinoids in the modulation of seizure threshold and severity. Eur J Pharmacol 452:295 301. Wallace MJ, Wiley JL, Martin BR, DeLorenzo RJ (2001) Assessment of the role of CB1 receptors in cannabinoid anticonvulsant effects. Eur J Pharmacol 428:51 57. Wallace MJ, Blair RE, Falenski KW, Martin BR, DeLorenzo RJ (2003) The endogenous cannabinoid system regulates seizure frequency and duratio n in a model of temporal lobe epilepsy. J Pharmacol Exp Ther 307:129 137. Wang YF, Hatton GI (2007) Dominant role of betagamma subunits of G proteins in oxytocin evoked burst firing. J Neurosci 27:1902 1912. Wasser CR, Kavalali ET (2009) Leaky synapses: regulation of spontaneous neurotransmission in central synapses. Neuroscience 158:177 188.
159 Wenzel HJ, Woolley CS, Robbins CA, Schwartzkroin PA (2000) Kainic acid induced mossy fiber sprouting and synapse formation in the dentate gyrus of rats. Hippocampus 10:244 260. Widmer H, Ferrigan L, Davies CH, Cobb SR (2006) Evoked slow muscarinic acetylcholinergic synaptic potentials in rat hippocampal interneurons. Hippocampus 16:617 628. Williams PA, Larimer P, Gao Y, Strowbridge BW (2007) Semilunar granule cells : glutamatergic neurons in the rat dentate gyrus with axon collaterals in the inner molecular layer. J Neurosci 27:13756 13761. Wilson RI, Nicoll RA (2001) Endogenous cannabinoids mediate retrograde signalling at hippocampal synapses. Nature 410:588 592. Wilson RI, Kunos G, Nicoll RA (2001) Presynaptic specificity of endocannabinoid signaling in the hippocampus. Neuron 31:453 462. Wittner L, Miles R (2007) Factors defining a pacemaker region for synchrony in the hippocampus. J Physiol 584:867 883. Xiang Z, Brown TH (1998) Complex synaptic current waveforms evoked in hippocampal pyramidal neurons by extracellular stimulation of dentate gyrus. J Neurophysiol 79:2475 2484. Yamasaki M, Hashimoto K, Kano M (2006) Miniature synaptic events elicited by presyna ptic Ca2+ rise are selectively suppressed by cannabinoid receptor activation in cerebellar Purkinje cells. J Neurosci 26:86 95. Yang H, Zhang J, Andreasson K, Chen C (2008) COX 2 oxidative metabolism of endocannabinoids augments hippocampal synaptic plast icity. Mol Cell Neurosci 37:682 695. Yang H, Zhang J, Breyer RM, Chen C (2009) Altered hippocampal long term synaptic plasticity in mice deficient in the PGE2 EP2 receptor. J Neurochem 108:295 304. Yanovsky Y, Mades S, Misgeld U (2003) Retrograde signali ng changes the venue of postsynaptic inhibition in rat substantia nigra. Neuroscience 122:317 328. Yoder RM, Pang KC (2005) Involvement of GABAergic and cholinergic medial septal neurons in hippocampal theta rhythm. Hippocampus 15:381 392. Yokoyama K, Ba ker DL, Virag T, Liliom K, Byun HS, Tigyi G, Bittman R (2002) Stereochemical properties of lysophosphatidic acid receptor activation and metabolism. Biochim Biophys Acta 1582:295 308.
160 Young SR, Chuang SC, Wong RK (2004) Modulation of afterpotentials and fi ring pattern in guinea pig CA3 neurones by group I metabotropic glutamate receptors. J Physiol 554:371 385. Yue C, Remy S, Su H, Beck H, Yaari Y (2005) Proximal persistent Na+ channels drive spike afterdepolarizations and associated bursting in adult CA1 pyramidal cells. J Neurosci 25:9704 9720. Zhu PJ, Lovinger DM (2005) Retrograde endocannabinoid signaling in a postsynaptic neuron/synaptic bouton preparation from basolateral amygdala. J Neurosci 25:6199 6207. Zhuang S, Hampson RE, Deadwyler SA (2005) B ehaviorally relevant endocannabinoid action in hippocampus: dependence on temporal summation of multiple inputs. Behav Pharmacol 16:463 471. Zimmerman G, Njunting M, Ivens S, Tolner EA, Behrens CJ, Gross M, Soreq H, Heinemann U, Friedman A (2008) Acetylch oline induced seizure like activity and modified cholinergic gene expression in chronically epileptic rats. Eur J Neurosci 27:965 975.
161 BIOGRAPHICAL SKETCH Mackenzie was born in La Crosse, Wisconsin but spent the majority of his childhood growi ng up in Lon gwood Florida. Pursuing his life long interest in science, Winston Salem, North Carolina. Following graduation in 2003, Mackenzie took a job as a lab technician working with Dr. Jason Frazier at the University of Florida. During this time he learned the fundamentals of neurophysiology and decided to pursue a PhD in neuroscience. In the fall of 2006, Mackenzie entered the Interdisciplinary Program through the College of Medi cine at the University of Florida and continued his research lab he has published three first author journal articles and has submitted a fourth.