Material Information

Endocannabinoids Linking GABA and Glutamate in the Hilus
Nahir, Benjamin
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (145 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences
Neuroscience (IDP)
Committee Chair:
Frazier, Charles J.
Committee Members:
Roper, Steven N.
Foster, Tom
Gu, Jianguo
Demarse, Thomas
Graduation Date:


Subjects / Keywords:
Calcium ( jstor )
Dentate gyrus ( jstor )
Depolarization ( jstor )
Endocannabinoids ( jstor )
Hippocampus ( jstor )
Interneurons ( jstor )
Neurons ( jstor )
Rats ( jstor )
Receptors ( jstor )
Synapses ( jstor )
Neuroscience (IDP) -- Dissertations, Academic -- UF
dsi, endocannabinoid, epilepsy, gaba, glutamate, hilus, hippocampus, inhibition, ltd, memory, presynaptic
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Medical Sciences thesis, Ph.D.


The dentate hilus represents a significantly understudied and poorly understood region of the hippocampus. In particular, basic knowledge is lacking regarding hilar mossy cells and the regulation of their afferent activity. This deficit is striking given the apparent central role of mossy cells in the development and onset of temporal lobe epilepsy (TLE). The studies presented in this dissertation were designed to examine the regulation of excitability in mossy cells through presynaptic control of both their excitatory (glutamatergic) and inhibitory (GABAergic) afferent axons. Two systems in particular retrograde endocannabinoid signaling and ambient GABA were studied that, although researched in detail elsewhere in the brain, have not previously been explored in the hilus. Our initial experiments demonstrated the capability of hilar mossy cells to produce and release endocannabinoids (eCBs) in response to postsynaptic depolarization. Upon release, these eCBs activated the presynaptic CB1 receptor and induced the short term plasticity known as depolarization induced suppression of inhibition (DSI), observed as a transient reduction in GABA release from a subset of hilar interneurons. Further experiments revealed that hilar DSI is mechanistically similar to DSI observed in other brain areas such as the cerebellum and area CA1 of the hippocampus. In a separate study, we investigated the presence and role of presynaptic muscarinic acetylcholine and GABAB receptors on glutamatergic terminals impinging onto mossy cells. Similar to reports in area CA3, we found that both mossy fiber (MF) synapses (from granule cells of the dentate gyrus) and non-MFs to hilar mossy cells expressed presynaptic GABAB receptors (GABABRs) while only non-mossy fiber terminals expressed muscarinic receptors capable of directly inhibiting neurotransmission. MFs, on the other hand, were indirectly inhibited by muscarinic receptor activation. Surprisingly, we observed a preferential tonic activation of MF GABABRs that produced a steady state inhibition of glutamate release from these terminals. We did not find a similar inhibitory tone at non-MF terminals, even with pharmacologically elevated levels of ambient GABA, suggesting a functional spatial gradient over which ambient GABA can exert its effects. Our studies also indicated that the source of this GABAergic tone was spillover of GABA released during action potential dependent exocytosis, enhanced by muscarinic depolarization of a subset of interneurons. Careful examination of our prior results suggested that the neurons responsible for mediating extrasynaptic GABA concentrations may be the same ones that are inhibited by eCB signaling (DSI). As a first step toward testing whether these two signaling systems are linked, we considered the possibility that eCBs might also produce a long term depression of these same interneurons (iLTD). In fact, we did observe iLTD in the hilus that relied on eCB signaling as well as activation of postsynaptic metabotropic glutamate receptors (mGluRs). Our initial results suggested that hilar iLTD was functionally similar to iLTD observed in CA1 and elsewhere in the brain. However, surprisingly we found that low concentrations of tetrodotoxin enhanced iLTD and further that spontaneous glutamatergic transmission seemed to occlude our ability to generate iLTD. These results implied that spontaneous excitatory signaling may produce a constant eCB tone similar to the ambient GABA tone observed earlier and that this mGluR mediated eCB tone could endogenously inhibit GABA release at a low level. The presence of two potent presynaptic regulatory mechanisms in the hilus and a possible functional link between them (i.e. eCBs modulating the level of ambient GABA, thereby preferentially altering glutamatergic transmission from mossy fibers) will likely have a profound impact on the study of epilepsy and could lead to the development of new therapeutic interventions for patients suffering TLE. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2009.
Adviser: Frazier, Charles J.
Electronic Access:
Statement of Responsibility:
by Benjamin Nahir.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Nahir, Benjamin. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2009 ( lcc )


This item has the following downloads:

Full Text




2 2009 Ben Nahir


3 To all my ladies: Mom, my fiance Shelby, Spunky Chaka and Racegirl


4 ACKNOWLEDGMENTS I would first like thank my Mom, my fiance Shelby, Andres (Andy) del Campo, and Warren Katzenstein, four people whose love and support have been instrumental to my success throughout my life. I also wish to thank my friends from the Frazi er lab Chinki Bhatia, Mackenzie Hofmann, and Casie Lindsly including Matt and Nicole Parker, Mark Samols, April Dao, Steve Ziegler and Lindsay Levkoff Cort and Erin Bouldin, W. Michael (Mike) Dismuke, Ty Hesser, and David Tamayo. I wish to extend a special thanks to my mentor, Dr. Charles J. Frazier, for allowing me the opportunity and support to pursue my graduate work in his laboratory. His guidance and confidence in me have been invaluable for my development as a research scien tist, and for that I am extremely grateful I would also like to thank my committee members, Drs. Jianguo Gu, Thomas Foster, Thomas DeMarse, and Steven Roper, for their insig htful comments and suggestions. Additionally, I wish to thank Dr. Sue Semple Rowland for ensuring the smooth running of the Neuroscience branch of the Interdisciplinary Program and for tirelessly supporting myself and all the Neuroscience graduate student s. I would like to thank M s. Betty J. Streetman, Valerie Cloud Driver, and Susan Gardner, for their guidance patience, and assistance with any and all necessary paperwork. Finally, I must thank my high school biology teacher, Dr. Steven J. Robinson, who se knowledge, instruction, and obvious love of science provided the initial spark for my pursuit of a career in research.


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF FIGURES ................................ ................................ ................................ ......................... 8 ABSTRACT ................................ ................................ ................................ ................................ ... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 13 The Hippocampus ................................ ................................ ................................ ................... 13 Anatomy ................................ ................................ ................................ ................................ 14 The Hilus and Mossy Cell s ................................ ................................ ................................ ..... 17 Synaptic Transmission ................................ ................................ ................................ ............ 20 Presynaptic Modulation of Neurotransmission ................................ ................................ ...... 24 Endocannabinoids ................................ ................................ ................................ ............ 24 Endocannabinoid Mediated Plasticities ................................ ................................ ........... 28 GA BA ................................ ................................ ................................ .............................. 30 Summary ................................ ................................ ................................ ................................ 34 2 MATERIALS AND METHODS ................................ ................................ ........................... 36 Drugs an d Suppliers ................................ ................................ ................................ ................ 36 Animal Procedures and Tissue Preparation ................................ ................................ ............ 37 Extracellular Solutions ................................ ................................ ................................ ............ 38 Intracellular Solutions ................................ ................................ ................................ ............. 38 Whole Ce ll Recording ................................ ................................ ................................ ............ 39 Mossy Cell Identification ................................ ................................ ................................ ....... 39 Stimulation ................................ ................................ ................................ .............................. 40 Focal Drug Application ................................ ................................ ................................ .......... 41 Data Analy sis ................................ ................................ ................................ .......................... 42 Spontaneous and Miniature Synaptic Events ................................ ................................ .. 42 Evoked Synaptic Events ................................ ................................ ................................ .. 42 DSI ................................ ................................ ................................ ................................ ... 42 Tonic and Phasic Current Mea surements ................................ ................................ ........ 43 Statistics ................................ ................................ ................................ ........................... 45 3 ENDOCANNABINOID MEDIATED DEPOLARIZATION INDUCED SUPPRESSION OF INHIBITION IN HILAR MOSSY CELLS OF THE RAT DENTATE GYRUS ................................ ................................ ................................ ............... 49 Introduction ................................ ................................ ................................ ............................. 49 Results ................................ ................................ ................................ ................................ ..... 50 DSI of Spontaneous IPSCs in Hilar Mossy Cells ................................ ............................ 50


6 DSI of Evoked IPSCs in Hilar Mossy Cells ................................ ................................ .... 52 DSI in Hilar Mossy Cells is Mediated by Calcium Dependent Release of eCBs that Act on Presynaptic CB1 Receptors ................................ ................................ .............. 53 Activation of Presynaptic CB1 Receptors Preferentially Inhibits Calcium De pendent Exocytosis ................................ ................................ ................................ .. 54 Endocannabinoid Mediated Signaling in the Dentate Gyrus is Subject to Tight Spatial Constraints ................................ ................................ ................................ ....... 55 Discussion ................................ ................................ ................................ ............................... 56 The Role of Muscarinic Acetylcholine Receptors in DSI ................................ ............... 57 Presynaptic Effects of Endocannabinoids ................................ ................................ ....... 59 Spatial Constraints on Endocannabinoid Dependent Signaling ................................ ...... 60 4 PRESYNAPTIC INHIBITION OF EXCITATORY AFFERENTS TO HILAR MOSSY CELLS ................................ ................................ ................................ ................................ .... 7 0 Introduction ................................ ................................ ................................ ............................. 70 Results ................................ ................................ ................................ ................................ ..... 71 Presynaptic GABA B Receptors are Expressed on Mossy Fiber Inputs to Mossy Cells ................................ ................................ ................................ ............................. 71 Presynaptic GABA B Receptors on Mossy Fiber Inputs to Mossy Cells are Both Sensitive to Ambient GABA and Tonically Active Under Control Conditions .......... 73 Bath Application of Muscarinic Agonists Produces GABA B Receptor Mediated Inhibition of Mossy Fiber Inputs to Mossy Cells ................................ ........................ 73 Bath Application of Muscarinic Agonists Produces Increases in Ambient GABA that are Strongly Correlated with Increases in Spontaneous IPSCs. ........................... 74 Non Mossy Fiber Mediated Excitatory Inputs to Mossy Cells are Directly Inhibited by Both GABA B Receptors and Muscarinic Acetylcholine Receptors ....................... 76 Origin of Non MF Inputs ................................ ................................ ................................ 78 Discussion ................................ ................................ ................................ ............................... 79 GABA B Mediated Inhibition of Miniature EPSCs ................................ .......................... 80 Direct and Indirect Inhibition of Minimally Evoked EPSCs. ................................ .......... 81 5 METABOTROPIC GLUTAMATE RECEPTOR MEDIATED ILTD OF GABAERGIC AFFERENTS TO HILAR MOSSY CELLS ................................ ................................ .......... 92 Introduction ................................ ................................ ................................ ............................. 92 Results ................................ ................................ ................................ ................................ ..... 94 Metabotropic Glutamate Receptor Activation Induces iLTD in the Hilus ..................... 94 Hilar iLTD Requires Endocannabinoid Signaling ................................ .......................... 95 Spontaneous Glutamate Release Leads to Endogenous Occlusion of Hilar iLTD ......... 96 Discussion ................................ ................................ ................................ ............................... 98 Endocannabinoid Production via mGluR Activation ................................ ...................... 98 Role of Hilar iLTD ................................ ................................ ................................ .......... 99 Endogenous Occ lusion of Hilar iLTD ................................ ................................ ........... 100 6 CONCLUSION ................................ ................................ ................................ ..................... 111


7 Summary and Discussion ................................ ................................ ................................ ..... 111 Perspectives and Future Studies ................................ ................................ ........................... 121 LIST OF REFERENCES ................................ ................................ ................................ ............. 124 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 145


8 LIST OF FIGURES Figure page 1 1 Drawing of hippocampal formation, as seen by Ramn y Cajal. ................................ ...... 35 2 1 Mossy cell anatomy. ................................ ................................ ................................ .......... 46 2 2 Isolation of minimally evoked mossy fiber mediated evoked excitatory postsynaptic currents (EPSCs) ................................ ................................ ................................ ............... 47 2 3 Simultaneous measurement of phasic and tonic currents. ................................ ................. 48 3 1 Depolarization of hilar mossy cells transiently reduces both frequency and amplitude of carbac hol ( CCh ) induced theta band spontaneous inhibitory postsynaptic currents ( sIPSCs ) ................................ ................................ ................................ ............................ 62 3 2 Depolarization of hilar mossy cells reduces frequency but not amplitude of non theta ................................ ................................ 63 3 3 Depolarization induced suppres sion of inhibition ( DSI ) of evoked IPSCs in hilar mossy cells is enhanced by bath applied CCh. ................................ ................................ .. 64 3 4 DSI in hilar mossy cells depen ds on postsynaptic calcium influx and is blocked by a type 1 cannabinoid ( CB1 ) receptor antagonist. ................................ ................................ .. 65 3 5 Bath application of an e xogenous CB1 agonist reduces the amplitude of evoked IPSCs in hilar mossy cells and occludes DSI. ................................ ................................ ... 66 3 6 Depolarization induced release of endogenous cannabinoids (eCBs) preferentially inhibits calcium dependent exocytosis. ................................ ................................ ............. 67 3 7 Bath application of WIN55,21 2 2 preferentially inhibits calcium dependent exocytosis. ................................ ................................ ................................ .......................... 68 3 8 Spatial constraints on endocannabinoid mediated signaling in the d entate gyrus. ............ 69 4 1 Baclofen reduces miniature EPSC ( mEPSC ) frequency without affecting amplitude. ..... 85 4 2 Baclofen blocks minimally evoked EPSCs ( meEPSCs ) from isolated mossy fiber (MF) terminals via presynaptic GABA B receptors. ................................ ........................... 86 4 3 Presynaptic subtype B GABA receptors ( GABA B Rs ) on MF terminals respond to increases in ambient GABA. ................................ ................................ .............................. 87 4 4 Mossy fiber inputs to mossy cells are subject to tonic inhibition by ambient GABA, mediated through GABA B receptors. ................................ ................................ ................. 88 4 5 Muscarinic inhibition of mossy fibers is reversed by a selective GABA B R antagonist. ... 89


9 4 6 CCh induced increases in sIPSCs produce significant increases in subtype A GABA receptor ( GABA A ) mediated tonic inhibition of hilar mossy cells. ................................ ... 90 4 7 Non MF inputs to mossy cells express both GABA B Rs and mAChRs. ............................ 91 5 1 Activation of group I metabotropic glutamate receptors ( mGluRs ) inhibits eIPSCs to hilar mossy cells. ................................ ................................ ................................ .............. 103 5 2 DHPG selectively inhibits CB1R+ hilar interneurons and induces mGluR mediated, eCB dependent long term depression ( eCB mGluR LTD ) at these synapses to mossy cells. ................................ ................................ ................................ ................................ 104 5 3 Blockade of CB1R by AM251 prevents eCB mGluR mediated depression of meIPSCs. ... 105 5 4 CB1R knockout mice are unaffected by DHPG while wild type cousins maintain their DHPG sensitivity. ................................ ................................ ................................ .... 106 5 5 Metabotropic glutamate receptor type 1 (mGluR1) mediates eCB mGluR in hilar mossy cells in response to treatment with DHPG. ................................ ................................ ...... 107 5 6 10 nM TTX enhances LTD of inhibitory terminals ( iLTD ) without affecting sIPSCs. .. 108 5 7 Spontaneous glutamate release occludes DHPG mediated inhibition. ............................ 109 5 8 10 nM TTX preferentially inhibits sEPSCs. ................................ ................................ .... 110


10 ABSTRACT OF DISSERTATION PRESENT ED TO THE GRADUATE S CHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLME NT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY ENDOCANNABINOIDS: LI NKING GABA AND GLUTA MATE IN THE HILUS By Ben Nahir August 2009 Chair: Charles J Frazier Major: Medical Sciences Neuroscience The dentate hilus represents a significantly understudied and poorly understood region of the hippocampus. In particular, basic knowledge is lacking regarding hilar mossy cells and the regulation of their afferent activity. This deficit is striking given the apparent central role of mossy cells in the development and onset of temporal lobe epilepsy (TLE). The studies presented in this dissertation were designed to examine the regulation of excitability in mossy cells through presynaptic control of both t heir excitatory (glutamatergic) and inhibitory (GABAergic) afferent axons. Two systems in particular retrograde endocannabinoid signaling and ambient GABA were studied that, although researched in detail elsewhere in the brain, have not previously bee n explored in the hilus. Our initial experiments demonstrated the capability of hilar mossy cells to produce and release endocannabinoids (eCBs) in response to postsynaptic depolarization. Upon release, these eCBs activated the presynaptic CB1 receptor a nd induced the short term plasticity known as depolarization induced suppression of inhibition (DSI), observed as a transient reduction in GABA release from a subset of hilar interneurons. Further experiments revealed that hilar DSI is mechanistically sim ilar to DSI observed in other brain areas such as the cerebellum and area CA1 of the hippocampus.


11 In a separate study, we investigated the presence and role of presynaptic muscarinic acetylcholine and GABA B receptors on glutamatergic terminals impinging onto mossy cells. Similar to reports in area CA3, we found that both mossy fiber (MF) synapses (from granule cells of the dentate gyrus) and non MFs to hilar mossy cells expressed presynaptic GABA B receptors (GABA B Rs) while only non mossy fiber terminals expressed muscarinic receptors capable of directly inhibiting neurotransmission. MFs, on the other hand, were indirectly inhibited by muscarinic receptor activation. Surprisingly, we observed a preferential tonic activation of MF GABA B Rs that produced a steady state inhibition of glutamate release from these terminals. We did not find a similar inhibitory tone at non MF terminals, even with pharmacologically elevated levels of ambient GABA, suggesting a functional spatial gradient over which ambient GAB A can exert its effects. Our studies also indicated that the source of this GABAergic tone was spillover of GABA released during action potential dependent exocytosis, enhanced by muscarinic depolarization of a subset of interneurons. Careful examinati on of our prior results suggested that the neurons responsible for mediating extrasynaptic GABA concentrations may be the same ones that are inhibited by eCB signaling (DSI). As a first step toward testing whether these two signaling systems are linked, w e considered the possibility that eCBs might also produce a long term depression of these same interneurons (iLTD). In fact, we did observe iLTD in the hilus that relied on eCB signaling as well as activation of postsynaptic metabotropic glutamate recepto rs (mGluRs). Our initial results suggested that hilar iLTD was functionally similar to iLTD observed in CA1 and elsewhere in the brain. However, surprisingly we found that low concentrations of tetrodotoxin enhanced iLTD and further that spontaneous glut amatergic transmission seemed to occlude our ability to generate iLTD. These results implied that spontaneous excitatory signaling may produce a


12 constant eCB tone similar to the ambient GABA tone observed earlier and that this mGluR mediated eCB tone coul d endogenously inhibit GABA release at a low level. The presence of two potent presynaptic regulatory mechanisms in the hilus and a possible functional link between them (i.e. eCBs modulating the level of ambient GABA, thereby preferentially altering gl utamatergic transmission from mossy fibers) will likely have a profound impact on the study of epilepsy and could lead to the development of new therapeutic interventions for patients suffering TLE.


13 CHAPTER 1 INTRODUCTION The Hippocampus One of the hallmarks of the mammalian brain is its capacity for encoding past events and experiences and recalling these memories later. Everyday interactions, such as driving to and from work or greeting people on the stree t, depend on our ability to create and store new memories as well as call upon old ones. While the exact nature of memory storage is still unclear with theories ranging from single synapse single memory to network activity and pattern completion certain r egions within the brain have been identified as indispensible for memory formation, perhaps the most important of which is the hippocampus. As one of the more electrically unstable regions in the brain (relatively susceptible to runaway excitation, pres umably due to its high level of computational activity), the hippocampus has been pinpointed as one of the primary foci for a very debilitating form of epilepsy : temporal lobe epilepsy (TLE). In the mid to drug treatments received uni or bilateral medial temporal lobe resections (including removal of the hippocampus) as a last ditch effort to reign in their seizures. While such extreme surgical techniques generally proved effective, they had devastatin g side effects; specifically, the patients experienced enduring anterograde amnesia (Scoville and Milner, 1957). Later clinical studies identified specific portions of the medial temporal lobe and more specifically of the hippocampus that were involved in certain aspects of memory formation (Zola Morgan et al., 1986). Since then, countless experimental reports, in animal models ranging from rats to monkeys, have demonstrated the necessity of an intact hippocampus for proper memory storage. And while curr ent opinions differ as to its exact role in memory formation, storage, and recall, the overarching message is clear.


14 Anatomy The beginning of the 20 th century saw neuroscientists exploring in finer and finer detail the anatomical features of specific brain regions. Using newly invented histological techniques such as Golgi and Nissl staining, scientists were able to go beyond gross anatomical descripti ons and begin identifying individual cell layers and distinct types of neurons. The Spanish physician Santiago Ramn y Cajal provided some of the most complete histological studies of the time, including that of the hippocampal formation (Fig. 1 1). As first described by Cajal and later expounded upon by his student, Rafael Lorente de N, the hippocampal formation is comprised of a few different cellular regions with two well defined cell layers. The hippocampal formation is most easily r ecognized by i ts distinct shape two interlocking C shapes described by the principal cell layers o f the dentate and hippocampus but the structure as a whole includes the entorhinal cortex, dentate, hippocampus, subiculum, presubiculum, and parasubiculum. The cells of t he entorhinal cortex, presubiculum, and parasubiculum are somewhat scattered in their arrangement; by contrast, the hippocampus, dentate, and to a lesser extent the subiculum each contain a distinct, identifiable cell layer surrounded by more diffuse cellu lar organization. The cells residing in these well defined strata are mainly glutamatergic and are termed principle cells, as they are primarily responsible for facilitating information flow through the hippocampal formation. The principal cells of the hippocampus are called pyramidal cells, owing to their somatic shape. Lorente de N subdivided the pyramidal cell layer into three primary regions based on their synaptic connectivity: CA1, CA2, and CA3 (the name CA4 was given to the region of the pyramid al blade extending into the hilus but these cells are often grouped together with CA3). Pyramidal cells are bipolar in nature; each cell projects a primary apical dendrite into stratum radiatum between the blades of the pyramidal cell layers. The


15 proxim al portion of this dendrite from CA3 pyramidal cells is linear through s. lucidum but branches multiple times upon entering s. radiatum In general, the apical dendrites of all pyramidal cells run parallel into s. radiatum prior to branching. Pyramidal c ells also extend basal dendrites below the cell layers into stratum oriens The axon of each pyramidal cell arises from the basal portion of the cell as well and projects initially into s. oriens For CA3 cells, the main axonal trunk eventually crosses t he pyramidal blade and projects into s. radiatum of CA1, forming the Schaffer collateral pathway. CA3 axon collaterals also give rise to a back projection into the hilus along with a subset that leave the hippocampus entirely, extending into the septal nu cleus as well as crossing the anterior commissure to form connections with the contralateral hippocampus. Like CA3, axons from CA1 pyramidal cells initiate in the basal part of the cell, projecting into s. oriens These axons primarily project to layers III/IV of the entorhinal cortex but a significant portion of CA1 pyramidal cells also form connections with other cortical structures. The dentate region houses the second principal cell layer, the dentate gyrus which is made up of granule cells. Granule cells are also bipolar with their dendrites branching extensively into stratum lacunosum moleculare S. lacunosum moleculare is described in thirds; granule cell dendrites receive the majority of their excitat ory innervation in the outer two thirds of slm from cells in layers II/III of the entorhinal cortex. Within the inner third of slm granule cells also receive some excitatory input from hilar mossy cells (see below). Granule cell axons, called mossy fibe rs, extend directly into the hilus; the primary mossy fiber projection from every granule cell coalesces to form the mossy fiber tract and project into CA3 in the narrow band called stratum lucidum just above CA3. Mossy fibers in s. lucidum synapse onto p roximal portions of the apical dendrites of CA3 pyramidal cells. These synapses are characterized by


16 large presynaptic boutons that engulf postsynaptic spines. Axon collaterals branch off of mossy fibers in the hilus, forming synaptic contacts with hilar neurons as well. Surrounding these two principal cell layers is a vast network of inhibitory interneurons (for an extensive review, see Freund and Buzsaki, 1996). The main role of these interneurons is to modulate synaptic activity between principal c ells, thus altering the information transmitted through the hippocampus. Interneuron architecture is extremely varied, from simple small unipolar cells to large extensively branched basket cells. Unlike principal cell synaptic contacts onto other excitat ory neurons, glutamatergic synapses at interneuron dendrites are small, commonly formed as en passant synapses or by small filopodial extensions. GABAergic axons from hippocampal interneurons form small, symmetric synapses on their target dendrites. Hipp ocampal neurons also receive non GABAergic or glutamatergic extrinsic innervations as well, including cholinergics from the septal nucleus, noradrenergics from the locus coeruleus, and serotonergics from the raphe nuclei (Johnston et al., 1998). The synap tic organization of the hippocampus immediately affords the structure different discrete routes of information processing. The primary route by which information travels through the hippocampus is known as the trisynaptic pathway (Amaral and Witter, 1989) Neurons from layer II of the entorhinal cortex project primarily to the outer two thirds of the molecular layer. Known as the perforant path, these axons synapse on the dendritic trees of the granule cells, comprising the first of the three synapses in the pathway. Granule cells then contact CA3 pyramidal cells via the mossy fiber tract in s. lucidum as described, forming the second synapse. Schaffer collateral axons from CA3 contact CA1 pyramidal cells in s. radiatum forming the third and final syn apse in the pathway. A secondary monosynaptic loop exists between the entorhinal cortex and CA1 as well. Neurons from entorhinal layers II and III


17 also synapse directly onto distal dendrites of CA1 pyramidal cells that in turn project back to the entorhi nal cortex. Interneurons modulate information flow within these synaptic pathways through feedforward and feedback inhibitory circuits. The Hilus and Mossy Cells The hilar region just below the dentate gyrus is uniquely situated to modify neurotransmis sion near the beginning of the trisynaptic pathway. Many different interneurons exist within the hilus, including axo axonic, basket, and hilar commissural associational pathway (HICAP) related interneurons (Freund and Buzsaki, 1996). The vast majority o f these interneurons receive excitatory drive from at least some of the perforant path, associational commissural projections, mossy fibers, or recurrent CA3 axons (Zipp et al., 1989; Scharfman, 1991; Kneisler and Dingledine, 1995 a, b; Freund and Buzsaki, 1996; Johnston et al., 1998). With the granule cells as one of their primary synaptic targets, these interneurons can affect how active the granule cell layer is in response to perforant path activation, thereby effectively controlling overall activity al ong the trisynaptic pathway. Among hilar interneurons, spiny mossy cells exhibit many unique characteristics; notably, unlike all other hilar interneurons, they are excitatory, releasing glutamate from their synaptic terminals rather than GABA (Soriano and Frotscher, 1994; Scharfman, 1995). Even more unusual, mossy cells contribute only minimally to the local laminar circuitry of the hilus; after a short distance, their axons project along the septotemporal axis of the hippocampus, with targets in both the ipsilateral (associational) and contralateral (commissural) hippocampi (Laurberg and Sorensen, 1981; Amaral and Witter, 1989; Buckmaster et al., 1992; Johnston et al., 1998). Mossy cell axon contacts include hilar interneurons but their primary targe ts are thought to be proximal dendrites of distant granule cells (Leranth et al., 1990; Scharfman et al., 1990; Buckmaster et al., 1996; Jackson and Scharfman, 1996). Such distant axonal arborization


18 is thought to allow mossy cells to coordinate interlami nar activity, a stark contrast to the intralaminar local circuitry associated with most other hippocampal neurons (Henze and Buzsaki, 2007). The two most common types of synaptic input to mossy cells are glutamatergic and GABAergic. The primary source of glutamatergic innervation is the mossy fiber axon collateral from the granule cells (Scharfman et al., 1990). Similar to those observed at the mossy fiber CA3 synapse in s. lucidum mossy fiber axon collaterals form large, asymmetric boutons that engul f thorny excrescences (spines) on the proximal dendrites of the mossy cells (Amaral, 1978; Chicurel and Harris, 1992; Henze et al., 2000). The large size of these synapses has led to speculation and study regarding their signaling properties; in particula r, mossy fiber boutons commonly contain upwards of 30 active zones (Claiborne et al., 1986). Other intrinsic sources of excitation include recurrent axons from the CA3 pyramidal cells and invasions into the hilus by the perforant path (Scharfman, 1991; Li et al., 1994; Scharfman, 1994a). These afferents generally synapse on much smaller spines in the distal portions of the mossy cell dendritic tree, although mossy fibers can also contact mossy cells at these distal sites (Frotscher et al., 1991). Other h ilar mossy cells represent a fourth possible source of excitation, but this suggestion is more contentious especially considering a definitive mossy cell mossy cell synaptic contact has yet to be demonstrated (Scharfman et al., 1990; Larimer and Strowbri dge, 2008). Along with their glutamatergic afferents, mossy cells also receive significant GABAergic input (Larimer and Strowbridge, 2008). Hilar interneurons that release cholecystokinin (CCK) or parvalbumin preferentially target mossy cells over other inhibitory hilar cells (Acsady et al., 2000). Mossy cells also likely receive significant cholinergic input as projections from the medial septum heavily innervate the hilus (Johnston et al., 1998).


19 As evidenced by many electrophysiological recordings, synaptic inputs to mossy cells are often quite spontaneously active (much of which is thought to originate from large boutons formed by the mossy fibers) (Strowbridge et al., 1992; Scharfman, 1993; Henze and Buzsaki, 2007). Mossy cells themselves show onl y weak to moderate spike accommodation during prolonged depolarizing events and are often very excitable, a dubious characteristic considering the low seizure threshold reported for the hippocampus as a whole (but see, Soltesz et al., 1993; Henze and Buzsa ki, 2007). In fact, mossy cells are extremely susceptible to excitotoxicity, a feature often associated with hippocampal insult or traumatic brain injury. Of particular note are the significant mossy cell dysfunction and death that occur in parallel with TLE onset (Nadler et al., 1978; Lowenstein et al., 1992). Consequently, the three main theories describing the etiology of TLE have all given mossy cells considerable weight. One theory, the Irritable Mossy Cell hypothesis, suggests that while most moss y cells die following the initiation of status epilepticus, those that remain are exquisitely sensitive to their excitatory drive and consequently become hyperactive (Santhakumar et al., 2000; Ratzliff et al., 2002). Their hyperactivity leads to overdrive of the granule cells and subsequent seizure onset. A second theory, the Dormant Basket Cell hypothesis, purports that loss of excitatory output following mossy cell death during epilepsy leads to a silencing of hilar basket cells, the main inhibitors of granule cell activity (Sloviter, 1991, 1994; Sloviter et al., 2003; Zappone and Sloviter, 2004). This loss of inhibition then allows excess electrical activity to enter the hippocampus via the dentate. The third theory, Mossy Fiber Sprouting, suggests th at loss of hilar neurons in general with special emphasis on mossy cells causes mossy fibers from the granule cells to seek out new synaptic targets (Sutula et al., 1989; Babb et al., 1991; Golarai et al., 1992; Leite et al., 1996). These new targets are often other granule cells, leading to the creation of an excitatory feedback circuit within the granule


20 cell layer. Although each of these theories differs on the exact etiology of TLE, mossy cells are clearly very important for maintaining the electrical homeostasis of the hippocampus. Synaptic Transmission Neurons, as all cells, are physically isolated from one another; yet communication between neurons is critical for our survival. Two main structures are responsible for allowing the flow of informa tion between neurons in the brain: electrical and chemical synapses. Electrical synapses (gap junctions) permit current to move directly from one cell to another via ionic flow. Chemical synapses, on the other hand, require a signaling molecule to transm it information. A wide variety of chemical synapses, including axo axonic, dendro dendritic, axo somatic, and so on, exist in the mammalian CNS. Most commonly, however, neurons communicate through axo dendritic connections, with the signaling neuron givi ng rise to a presynaptic terminal from its axon and the receiving neuron creating a postsynaptic density in its dendritic tree. Differences in synapse size and shape exist as well, with excitatory terminals generally forming large, asymmetrical synapses a nd inhibitory terminals forming smaller, symmetrical synapses (Ribak and Seress, 1983; Schlander et al., 1987; Frotscher et al., 1992; Torrealba and Muller, 1999). Large synaptic boutons (i.e. from hippocampal mossy fibers) are generally found at the end of axon branches while en passant synapses arise as varicosities along the axon itself (Hamlyn, 1962; Amaral and Dent, 1981; Acsady et al., 1998). Communication at chemical synapses requires vesicular release of neurotransmitter from the presynaptic termi nal and binding and activation of postsynaptic receptors. While transmitter type can vary dramatically between cells, the rules governing synaptic transmission are essentially uniform. Synaptic transmission was initially studied in detail by Paul Fatt an d Bernard Katz in the early 1950s (Fatt and Katz, 1950; Del Castillo and Katz, 1954; Boyd and Martin, 1956). Using the frog neuromuscular junction (NMJ) preparation, their research led


21 them to propose the Quantal Theory of synaptic signaling. While centr al synapses differ from the NMJ, some of the underlying principles proposed by Fatt and Katz still apply; specifically, three variables govern the observed postsynaptic response to presynaptic activity: the number of available release sites (active zones, n), the probability of release of a single vesicle (p), and the size of the postsynaptic response to a single release event (quantal size, q). The final response observed in the postsynaptic cell is the product of these three variables (n*p*q). In cases where the number of release sites is unknown, the mean quantal content (m) represents the product of n and p. Since neurotransmitter packaging into vesicles is generally uniform (although mechanisms exist to alter vesicular loading), if mean quantal conte nt remains stable (i.e. in the absence of presynaptic changes such as short term facilitation) then the average synaptic response is determined primarily by receptor type and density in the postsynaptic membrane. Thus, differences in receptor expression s ignificantly impact neuronal communication at various synapses. Unlike in the periphery at central synapses, characteristics such as multiple release sites, non uniform and non independent exocytosis, and variable vesicle pool sizes often make them ill suited to traditional quantal analysis. However, under appropriately controlled conditions, quantal analysis can be and has been used effectively to illuminate signaling properties at different central synapses (Jonas et al., 1993; Bekkers and Stevens, 19 95; Isaacson and Walmsley, 1995; Rosenmund and Stevens, 1996). In particular, quantal analysis is especially useful when determining whether the effect of a pharmacological agent is presynaptic or postsynaptic. However this method is only practical when m is small (usually achieved by reducing or eliminating action potential generation and propagation); the presence of voltage gated ion channels in the postsynaptic membrane leads to a breakdown in the linear relation


22 between quantal release and quantal s ize (although development of the voltage clamp as well as intracellular blockade of voltage gated ion channels has allowed researchers much more flexibility in measuring multiquantal release events). Following synaptic release of neurotransmitter be it quantal or multiquantal a wide variety of receptors exist to translate intercellular communication into an intracellular signal. Ionotropic receptors respond to ligand binding by opening a pore that allows one or more ion types to traverse the membrane. The ubiquitous excitatory transmitter glutamate, for example, causes membrane depolarization by binding to ionotropic glutamate receptors, which include AMPA, NMDA, and kainate (Hollmann and Heinemann, 1994). Activation of AMPARs by glutamate opens an ion channel permeable to sodium, potassium, and sometimes calcium. The aminobutyric acid (GABA), on the other hand, activates ionotropic GABA A receptors, allowing chloride ions to enter the cell and hyperpolarize the membrane. Pore forming subunits generally combine with other subunits to build the final receptor structure. These receptors can be homomeric (containing identical subunits) or heteromeric, with variations in receptor subunit composition significantly altering th e Vastly outnumbering ionotropic receptors are the metabotropic G protein coupled receptors (GPCRs). These receptors, which often have ionotropic counterparts, affect cellular function and membrane properties through recruitment and activation of specific G proteins. In general, GPCRs are single proteins, consisting of seven membrane spanning domains that form an oblong oval shape. Diversity of related GPCRs is most often ach ieved through amino acid modifications of the inter membrane spanning regions. Related metabotropic receptors (i.e. types 1 5 muscarinic acetylcholine receptors and groups I IV metabotropic glutamate receptors)


23 achieve differential function by recruiting different classes of G proteins. While ionotropic receptors are limited to changing current flow and the conductance of various ion s metabotropic receptors retain the ability to alter a vast array of intracellular and membrane properties. GPCRs can dire ctly couple to ion channels, changing ionic conductance in response to activation; alternatively, they can activate second messenger systems, leading to a decrease in vesicular docking or an increase in protein phosphorylation, for example. The array of d ifferent neurotransmitters, ionotropic and metabotropic receptors along with the sheer number of possible receptor expression patterns allow for an incredible variety in synaptic signaling (Strader et al., 1994; Brown and Sihra, 2008). Since the discove ry of classical synaptic signaling unidirectional, from the presynaptic ter minal to postsynaptic receptors other forms of neurotransmission have been identified that complement and modify such signaling. Receptors for synaptically released neurotransmitte rs, for example, exist outside the postsynaptic density (site of synaptic action) (Eccles et al., 1963; Schoepp, 2001; Engelman and MacDermott, 2004), suggesting that neurotransmitter may spill out of the synaptic cleft (Dzubay and Jahr, 1999; Glykys and M ody, 2007; Nahir et al., 2007) or that its release can occur ectopically (i.e. from sites other than the specialized presynaptic terminal), or both (Matsui and Jahr, 2003; Duguid et al., 2007). The presence of receptors on the axon and presynaptic termina l itself also suggest different means of neuronal communication. Again, spill over or ectopic release of transmitter plays a role in activating these receptors but their spatial orientation suggests that synapses may often be bidirectional sites of commun ication, employing retrograde as well as anterograde signaling (Duguid and Smart, 2004; Feil and Kleppisch, 2008; Lovinger, 2008).


24 Presynaptic Modulation of Neurotransmission Endocannabinoids Endocannabinoids are endogenously synthesized molecules that act at the same receptor 9 tetrahydrocannabinol (THC), the active ingredient in marijuana. While the psychotropic effects of marijuana have been known and exploited for centuries, only within the past 50 years has research begun to elucidate the molecular mechanisms behind these effects (Gaoni and Mechoulam, 1964). In the mammalian brain, endocannabinoids (eCBs) are formed as metabolites of lipid breakdown. Although a number of different molecules fall into the endocannabinoid family (including noladin ether, virodhamine, and N arachidonyldopamine), arachidonylethanolamine (anandamide, AEA) and 2 arachidonyl glycerol (2 AG) appear to be the primary compounds responsible for mediating endogenous cannabinoid signaling (Piomelli, 2003). AEA and 2 AG production occur via similar but independent pathways. The primary precursor to anandamide is phosphatidyl ethanolamine (PE). PE is converted to N arachidonoyl PE by N acyl transferase (NAT); phospholipase D then cleaves the fatty acid head group via hydrolysis of the phosphodiester bond, converting N arachidonoyl PE into anandamide (Di Marzo et al., 1994; Cadas et al., 1997). The initiation of anandamide production, however, is highly calcium dependent. NAT, the enzyme responsible for catalyzing th e passage of the arachidonic acid head group from phosphatidylcholine to PE, is dependent on both Ca 2+ and cAMP: Ca 2+ engages and activates NAT while cAMP enhances NAT activity via protein kinase A (PKA) dependent phosphorylation (Cadas et al., 1996). Like AEA, 2 AG requires the removal of a lipid head group; phospholipase C (PLC) converts the abundant precursor phosphatidylinositol (PI) to 1,2 diacylglycerol (DAG) which is then broken down into 2 AG by DAG lipase (DGL) (Stella et al., 1997). For 2 AG production, PLC acts as the Ca 2+ sensor,


25 similar to NAT in the AEA production pathway. In particular, the specific subtype of PLC determines its Ca 2+ sensitivity and can significantly affect the amount of 2 AG produced for a given calcium concentration (F ukudome et al., 2004; Hashimotodani et al., 2005). A putative Ca 2+ independent pathway has also been proposed for 2 AG production, although the individual steps in this process are unclear (various studies have reported this pathway includes all, part, or none of the PLC/DGL machinery). Ca 2+ independent production of 2 AG occurs following activation of somatic group I metabotropic glutamate receptors (mGluRs) (Maejima et al., 2001b; Varma et al., 2001). An alternative, theoretical Ca 2+ dependent pathway exists for 2 AG production as well. In this case, PI is converted to a lysophospholipid by phospholipase A1 (PLA1); the lysophospholipid is then broken down by lyso PLC to 2 AG (Farooqui et al., 1989; Higgs and Glomset, 1994; Piomelli, 2003). Since 2 AG is often used as a precursor or end product for numerous different intracellular pathways, it is found at a much higher basal concentration than is AEA (Stella et al., 1997). As a result of eCBs being produced through lipid metabolism, they are generally thought to neurotransmitters). However, recent research has demonstrated the presence of active transport of these molecules as well (Ronesi et al., 2004). In the extracellular space, diffusion of these molecules is tightly controlled, primarily because these compounds can diffuse passively across the membrane back into cells. But their uptake is aided and accelerated by facilitated diffusion via neuronal and astr ocytic transporters (Beltramo et al., 1997; Hillard et al., 1997). Once back in the cell, AEA and 2 AG are broken down by hydrolytic enzymes: fatty acid amide hydrolase (FAAH) cleaves anandamide, forming arachidonic acid and ethanolamine (Hillard et al., 1995;


26 Ueda et al., 1995) while 2 AG hydrolysis occurs via monoglycerol lipase (MGL) (Goparaju et al., 1999; Dinh et al., 2002). two receptors discovered and their amin o acid sequences identified (Matsuda et al., 1990; Munro et al., 1993). Of the two receptors, CB1 is most commonly found in the central nervous system while CB2 is generally associated with the periphery and immune system. Other uncloned CB receptors hav e been proposed but although electrophysiological evidence supports their existence (Hajos et al., 2001; Rouach and Nicoll, 2003; Hoffman et al., 2005), such a receptor has yet to be isolated and sequenced. The CB1 receptor is considered one of the most u biquitously expressed GPCRs in the entire mammalian brain, ranging from the neocortex all the way to the cerebellum (Piomelli, 2003). The first immunohistochemical markers proved inadequate at identifying the truly vast expression profile of this receptor Initial reports suggested CB1 was segregated almost exclusively to GABAergic terminals; in the hippocampus in particular, its expression was most closely linked with CCK 8 positive interneurons (Tsou et al., 1998; Katona et al., 1999; Marsicano and Lutz 1999; Tsou et al., 1999; Hajos et al., 2000; Katona et al., 2001). Yet, behavioral data suggest that exogenously delivered cannabinoids can exert either a pro or anti epileptic influence (Lutz, 2004). Pro epileptic effects would be expected if cannabi noids exclusively inhibited GABAergic terminals, leading to a general loss of inhibition in the hippocampus and presumably an increase in excitability. An anti epileptic outcome is more indicative of a reduction in general excitability. In fact, recently developed antibodies clearly demonstrate the presence of CB1 on glutamatergic terminals (Katona et al., 2006; Monory et al., 2006), particularly in the hippocampus, a finding confirmed through electrophysiological studies (Hofmann et al., 2008).


27 CB1 is coupled to the inhibitory G i/o proteins (a few studies suggest an association with G s and G q/11 both excitatory G proteins but excitatory effects of CB1 activation are rarely observed (Lauckner et al., 2005). Activation of presynaptic CB1 decreases adeny lyl cyclase (AC) activity (leading to a subsequent drop in cAMP levels) (Childers and Deadwyler, 1996), inhibits VGCCs (Mackie and Hille, 1992; Twitchell et al., 1997), and activates K + channels (Mu et al., 1999; Kreitzer et al., 2002). Decreased AC activ ity leads directly to a reduction in PKA. This interaction with its effector molecule Rab3A and inhibiting vesicle priming (Wang et al., 1997; Chevaleyre et al., 20 capable of activating non CB receptors; AEA in particular has been shown to bind the transient receptor potential vanilloid type I (TRPV1) receptor (Pertwee, 2005). Due to their lipophilicity, eCBs can also affect a variety of receptor independent changes. These include direct interactions with voltage gated ion channels (including Ca 2+ K + and Na + ), ligand gated ion channels (including serotonin, nicotinic acetylcholine, and TR P family channels), membrane ion transporters, and intracellular proteins and pathways (Oz, 2006). However all eCB effects studied in this dissertation occurred through activation of the CB1 receptor. In studying eCB transmission, many labs including our own have benefited greatly from the development and use of genetically modified mice lacking the CB1 receptor (CB1R / ). Three separate strains of CB1R / mice were independently generated, with the mice used in the work for this dissertation derived fro m the Zimmer strain (Zimmer et al., 1999). With any genetic mutation, the absence of a naturally occurring receptor can lead to developmental and behavioral changes as well as adaptive or compensatory cellular and molecular changes. Extensive studies of CB1R / mice suggest that no significant behavioral deficits exist and that


28 the most notable difference is an increased mortality of knockout mice (~30% compared to ~5% in wild type) (Zimmer et al., 1999). Some behavioral changes do occur, however, includ ing reduced locomotion and increased anxiety. With respect to the work presented in this dissertation, the most relevant effect of eCB signaling is its role in memory. Human and rodent studies together demonstrated the ability of cannabinoids (THC in hum ans and endogenous in rodents) to impair memory formation and LTP. Reports differ greatly, however, regarding the effects of genetic deletion of the CB1 receptor: some have demonstrated increased LTP and enhanced performance on memory tasks while other s r eported no significant difference between CB1R / and wild type littermates in their performance in the Morris water maze (a common rodent memory task) (Valverde et al., 2005). Although the ultimate effects of CB1R deletion are not yet clear, these geneti cally modified mice have proven invaluable for demonstrating the role of CB1 receptors, especially when used in conjunction with other physiological and pharmacological tools. Endocannabinoid Mediated P lasticities Since CB1 was initially observed at GABAergic terminals, the majority of studies dealing with retrograde eCB signaling report its effects at these inhibitory synapses. Consequently, the most commonly observed form of eCB mediated presynaptic inhibition is depolarization induced suppression of inhibition (DSI) (Kreitzer and Regehr, 2001a; Maejima et al., 2001a; Wilson and Nicoll, 2001; Hofmann et al., 2006). Following a depolarizing event in the postsynaptic cell (generally experimentally induced depolarization but physiologically relevant p resynaptic activity has been shown to elicit synapse specific DSI within the dendritic tree (Brown et al., 2003), Ca 2+ influx leads to a brief rise in calcium dependent eCB production, usually 2 AG (eCB depol ). The eCBs are then released into the extracell ular space and diffuse in a retrograde fashion to the presynaptic terminal, activating the CB1 receptors and reducing exocytosis. In this


29 case, eCB production and signaling is transient since the calcium driving this process relies on membrane depolarizat ion. Accordingly, CB1 receptor activation too is transient and thus intraterminal changes are kept to a minimum. Yet synaptic inhibition still occurs, in this case primarily through closure of VGCCs neighboring the CB1 receptors, leading to a smaller cal cium rise in response to an action potential invasion of the terminal. But once eCB signaling stops, CB1 receptors go quiet and synaptic transmission returns to normal. An identical phenomenon has been observed at excitatory synapses as well (depolarizat ion induced suppression of excitation; DSE) (Kreitzer and Regehr, 2001a; Chiu and Castillo, 2008; Hofmann et al., 2008). While DSI has been extensively studied elsewhere in the brain, knowledge regarding the scope of eCB signaling in the hilus was signifi cantly lacking. In fact, in the whole of the hippocampal formation, DSI had yet to be demonstrated at any local circuit synapse (i.e. not targeting principal cells). Consequently, the project presented in Chapter 3 was designed to explicitly test the cap ability of mossy cells to produce eCBs and determine to what extent this signaling could affect GABAergic hilar communication. Under appropriate conditions, CB1 receptors can also affect a long term reduction in neurotransmission. Unlike DSI and DSE, eCB LTD is less prevalent at CB1R + synapses. Surprisingly, the presence of both postsynaptic group I mGluRs and presynaptic CB1Rs is not always sufficient to induce eCB mGluR LTD (Chiu and Castillo, 2008). However, at synapses susceptible to eCB LTD, endocann abinoid signaling can have significant functional implications not only for the target synapses but for neighboring synapses as well. For example, high frequency stimulation (HFS) applied to the Schaffer collateral pathway leads to LTD of inhibitory s. ra diatum interneurons (iLTD) (Chevaleyre and Castillo, 2003). Specifically, excessive drive of the SC axons causes glutamate to reach postsynaptic group I mGluRs leading


30 to dendritic release of eCBs as described; the eCBs in turn act to reduce GABAergic tr ansmission from s. radiatum interneuron terminals neighboring the excitatory synapses formed by the SC axons. This heterosynaptic depression of sr interneurons reduces the threshold for LTP induction at nearby SC synapses, presumably allowing for SCs to p ositively control their synaptic strength (Chevaleyre and Castillo, 2004). Endocannabinoid mediated LTD need not occur only at GABAergic synapses; in fact, LTD of CB1 + glutamatergic terminals (eLTD) is more prevalent and generally more reliable than iLTD (Robbe et al., 2002; Safo and Regehr, 2005; Sergeeva et al., 2007; Adermark et al., 2009; Penzo and Pena, 2009). GABA GABA, formed when glutamic acid decarboxylase (GAD) breaks down glutamate, is one of two principally used inhibitory neurotransmitters in the CNS and both ionotropic GABA A and metabotropic GABA B receptors mediate its effects. GABA A Rs are selective chloride channels that also passage HCO 3 (bicarbonate) across the membrane. Early in development, the intracellular Cl concentration is high in many brain regions, causing the equilibrium potential of the ion to be slightly positive of the resting membrane potential of many neurons. Consequently, activation of GABA A Rs and opening of the Cl channel leads to an efflux of chloride and depolarization of the cell. The functional purpose of this initial excitatory affect is unclear but proposed hypotheses suggest that GABA is used as a neurotransmitter prior to glutamate and therefore GABAergic signaling must necessarily be e xcitatory to promote synaptic growth and enhance the strength of nascent synapses (for review, see Cherubini et al., 1991). Eventually, following increased expression of the K + /Cl cotransporter (KCC2) during development, the chloride gradient shifts, pus hing the Cl equilibrium potential negative of the resting membrane potential and causing subsequent GABA signaling through GABA A Rs to be inhibitory. In extreme cases, activity of GABAergic interneurons on small cellular structures such as dendritic


31 spine s can cause a breakdown of the chloride gradient. With the loss of chloride movement, bicarbonate, whose ionic gradient causes it to oppose chloride movement, can depolarize the membrane (Kaila and Voipio, 1987; Grover et al., 1993). More often, however, if Cl no longer moves across the membrane, active GABA A receptors still mediate inhibition through shunting. GABA B receptors also mediate inhibitory actions of GABAergic signaling, indirectly through G protein activation. GABA B Rs are coupled to the inh ibitory G i/o proteins and their activation dissociates and activates the G and G subunits. These subunits can affect a wide range of intracellular pathways, although they primarily mediate inhibition through activation of inwardly rectifying potassium channels (GIRKs) and closure of VGCCs (Thalmann and Ayala, 1982; Gahwiler and Brown, 1985; Marshall, 2008; Raiteri, 2008). Activation of GIRKs leads to postsynaptic hyperpolarization and inhibition of neuronal excitability. Closure of VGCCs and reduction of the membrane Ca 2+ conductance prevents the postsynaptic cell from responding as efficiently to depolarizing events. Presynaptic and extrasynaptic GABA receptors, both GABA A Rs and GABA B Rs, mediate inhibitory actions using similar mechanisms as those e mployed at the postsynaptic density (Lanthorn and Cotman, 1981; Kullmann et al., 2005). However, the types of receptors present at these different sites vary dramatically in their composition and sensitivity for GABA. Located at the postsynaptic density are phasic GABA A receptors that express a low affinity for GABA and rapidly desensitize (Dutar and Nicoll, 1988). While all GABA A Rs contain five subunits, usually consisting of three distinct subtypes receptors, on the other hand, generally have a higher affinity for GABA than their phasic


32 (Stell and Mody, 2002; Yeung et al., 2003; Mtchedlishvili and Kapur, 2006). These receptors have been reported to inhibit activity via both hyperpolarization and depolarization. As expected, hyperpolarization by presy naptic GABA A Rs moves the membrane potential further away from the threshold necessary for VGCC activation and Ca 2+ influx (Ruiz et al., 2003). The exact mechanism by which GABA A R mediated depolarization inhibits neurotransmission is unclear but theories s uggest that inactivation of the voltage sensitive sodium channels necessary for action potential generation as well as voltage dependent inactivation of VGCC may reduce action potential amplitude or prevent AP propagation entirely. At presynaptic terminal s, depolarization may decrease the driving force on Ca 2+ effectively reducing transmitter release probability. Shunting may also play a significant role in extrasynaptic receptor mediated inhibition (for review, see Kullmann et al., 2005). Similar to G ABA A Rs, metabotropic GABA B Rs are heterodimers, containing a GABA B1 and GABA B2 subunit each, with two isoforms identified for the GABA B1 subunit: GABA B1a and GABA B1b (Kaupmann et al., 1997). Genetic and receptor autoradiographic studies suggest that presyn aptic receptors preferentially include the GABA B1a isoform while postsynaptic receptors use the GABA B1b isoform (Vigot et al., 2006). Presynaptic GABA B Rs inhibit synaptic transmission primarily by reducing calcium conductance via modulation of P/Q type, a nd to a lesser extent N type, calcium channels (Wu and Saggau, 1995; Chen and Regehr, 2003; Lei and McBain, 2003). They may also enhance a background potassium conductance, further hyperpolarizing the presynaptic membrane (Fearon et al., 2003). While GA BA transporters located both on neurons and neighboring glial cells are generally effective at removing the majority of GABA from the extracellular space (Borden, 1996; Conti et al., 2004), some neurotransmitter is clearly present at a basal concentration


33 outside the synaptic cleft. The high affinity afforded to extrasynaptic GABA receptors allows them to respond to such low levels of ambient GABA. This tonic receptor activation leads to a steady state offset of the membrane potential (Kaneda et al., 1995 ) that, although metabolically expensive to maintain a stable Cl gradient in the presence of open channels, might allow for bi directional modulation of membrane and network excitability. Regulation by tonic activation of extrasynaptic GABA receptors occ urs frequently throughout the brain but often in a cell type specific manner (Nusser and Mody, 2002; Porcello et al., 2003; Semyanov et al., 2003; Cavelier et al., 2005). The source of ambient GABA remains controversial. One theory suggests that the sam e transporters responsible for taking up GABA from the extracellular space may reverse their action and extrude GABA back out of the cell (Richerson and Wu, 2003). An alternative theory purports that action potential dependent synaptic release of GABA can overwhelm uptake mechanisms and receptors, causing GABA to spill out of the synaptic cleft and build up around the terminal (Alle and Geiger, 2007; Glykys and Mody, 2007). Evidence for this theory includes the observation that an action potential depende nt increase in GABA release can preferentially inhibit one terminal type without affecting a second, while both impinge on the same cell and express the necessary presynaptic receptors (Nahir et al., 2007). At the onset of my graduate work, mossy fiber syn apses in CA3 were known to express presynaptic GABA A Rs (Ruiz et al., 2003) and GABA B Rs (Vogt and Regehr, 2001), both of which were subject to tonic activation by ambient GABA. The question of direct modulation of MFs in CA3 by mAChRs, however, was less clear; an early report intimated the presence of muscarinics on MF terminals (Williams and Johnston, 1990) but a more recent study suggested the observed muscarinic inhibition was due to indirect activation of GABA receptors (Vogt and


34 Regehr, 2001). In contrast, recurrent CA3 connections have been shown to express both presynaptic GABA B Rs and mAChRs, facilitating direct inhibition by both GABA and acetylcholine (Vogt and Regehr, 2001). Knowing this, the project presented in Chapter 4 was designed to compare GABAergic and cholinergic regulatory mechanisms at mossy fiber and non mossy fiber glutamatergic synaptic contacts in the hilus (specifically at hilar mossy cells) with those found in area CA3. The projects from Chapters 3 and 4 were designed independently and each of the mechanisms described therein stand alone as an effective way to regulate neurotransmission in the hilus of the dentate gyrus. However, we discovered a common theme potentially linking the two mechanisms together that led us to hypothesize that the level of hilar ambient GABA could be regulated by endocannabinoid media ted synaptic plasticity. The project in Chapter 5 is the first step taken toward testing this hypothesis; more specifically, we proposed that eCB mediated plasticity in the hilus may not be limited to DSI but might also include iLTD. Thus, we sought to d iscover whether mossy cells could produce eCBs in an mGluR dependent manner and whether they could subsequently induce iLTD at their CB1R + afferent terminals. Summary The work reviewed here demonstrates the importance of the hippocampus to the brain an d, more specifically, the hilar mossy cell to the hippocampus and reveals a general lack of understanding of mossy cell physiology and function as compared to hippocampal principal cells. Unique among hilar interneurons, mossy cells project to both the ip silateral and contralateral hippocampi and over a significant distance along the septotemporal axis, presumably coordinating granular activity in distinct lamina. Due to their susceptibility to excitotoxicity and central role in competing TLE theories, th is dissertation was designed to elucidate mechanisms by which mossy cell afferent activity is controlled. Our results


35 demonstrate the independent importance of both retrograde eCB signaling and ambient GABA in regulating presynaptic innervations of mossy cells and also suggest a possible link between the two, indicating multiple time scales and multiple levels of presynaptic regulation. Figure 1 1. Drawing of hippocampal format ion, as seen by Ramn y Cajal. Abbreviations: DG dentate gyrus; CA3 co rnus ammoni, area 3; CA1 cornus ammoni, area 1; EC entorhinal cortex. Modified from Ramn y Cajal, 1911.


36 CHAPTER 2 MATERIALS AND METHOD S Drugs and Suppliers The following chemicals were obtained from Tocris Cookson (Ellisville, MO): AM 251 (N (piperidin 1 yl) 5 (4 iodophenyl) 1 (2,4 dichlorophenyl) 4 methyl 1Hpyrazole 3 carboxamide); CB1R antagonist WIN55,212 2 ((R) (+) [2,3 dihydro 5 methyl 3 (4 morpholinylmethyl)pyrrolo[1,2,3 de] 1,4 benzoxazin 6 yl) 1 napthalenylmethanone); CB1R agonist DHPG ((RS) 3,5 dihydroxyphenylglycine); group I mGluR agonist MPEP (2 methyl 6 (phenylethynyl)pyridine hydrochloride); selective mGluR5 antagonist LY367385 ((S) (+) a amino 4 carboxy 2 methylbenzeneacetic acid); selective mGluR1 antagonist aminobut yric acid) CCh ( carbachol ; carbamoylcholine chloride); mAChR antagonist CGP 52432 ([3 [[(3,4 dichlorophenyl)methyl ]amino]propyl](diethoxymethyl)phosphinic acid); GABABR antagonist DCG 2 dicarboxycyclopropyl)glycine); group II mGluR agonist TTX (tetrodotoxin); voltage gated Na+ channel blocker The following chemicals were obtained from Sigma Aldrich (St. Louis, MO): CCh ( Carbachol ; carbamoylcholine chloride) NBQX (1,2,3,4 tetrahydro 6 nitro 2,3 dioxo benzo[f]quinoxaline 7 sulfonamid e); competitive AMPA/Kainate receptor antagonist DNQX (6,7 Dinitroquinoxaline 2,3(1H,4H) dione); competitive AMPA/Kainate receptor antagonist APV (DL 2 amino 5 phosphonovaleric acid); NMDA receptor antagonist TTX (tetrodotoxin) SR101 (Sulforhodamine 101); fluorescent dye BAPTA (Ethylenedioxybis(o (Ethylenedioxy)dianiline tetraacetic acid); membrane impermeant fast Ca2+ chelator


37 Baclofen (() (Aminomethyl) 4 chlorobenzenepropanoic acid); GABA B R agonist NO 7 11 (1 [2 [[(Diphenylmethylene)imino]oxy]ethyl] 1,2,5,6 tetrahydro 3 pyridinecarboxylic acid); inhibitor of GAT1 (GABA transporter) Muscarine (Tetrahydro 4 hydroxy N,N,N,5 tetramethyl 2 furanmethanammonium); mAChR agonist PTX (picrotoxin); GABA A R antagonist Atropine (endo () (Hydroxymethyl)benzeneacetic acid 8 methyl 8 azabicyclo[3.2.1]oct 3 yl ester) ; mAChR antagonist QX 314 Cl (Lidocaine N ethyl chloride); intracellular voltage gated Na+ channel blocker DMSO (dimethyl sulfoxide); dissolving agent Commo n compounds (i.e. CaCl 2 sucrose, KCl) used for intracellular and extracellular solutions were obtained from either Sigma Aldrich or ThermoFisher Scientific (Pittsburgh, PA). The fluorescent dye Alexa Fluor 594 was obtained from Molecular Probes (Invitrog en, Eugene, OR). Ketamine hydrochloride was purchased from Webster Veterinary (Sterling, MA). Animal P rocedures and T issue P reparation Hippocampal tissue from male Sprague Dawley rats or C57BL6/J mice between ages 18 and 25 days postnatal were used for this dissertation. Individual animals were given an intraperitoneal (IP) injection of ketamine hydrochloride (80 100 mg/kg) and decapitat ed using a small animal guillotine. The brain was immediately removed from the skull and placed in ice cold artificial cerebral spinal fluid (ACSF) used for dissection (see Extracellular Solutions). Pelco 100 Series 3000 Vibratome (Pelco, Redding, CA) and incubated between 30 and 35C for 30 min in dis sec ting ACSF then equilibrated to room temperature prior to use. Tissue slices were generally used within 8 h following dissection. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Florida and fit with the general animal welfare guidelines issued by the National Institutes of Health.


38 CB1R / a nd wt C57BL6/J mice were cousins, obtained by crossing homozygous knockout or wild type littermates, respectively, descended from heterozygous parents. Seminal heterozygous breeders were descended from the line developed by A. Zimmer at NIMH and were a ge nerous gift from Dr. Carl Lupica (NIDA). Extracellular S olutions Dissecting ACSF contained (in mM): 124 sodium chloride (NaCl), 2.5 potassium chloride (KCl), 1.2 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 ). Experimental ACSF contained (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 For the experiments presented in figures 3 3 to 3 6, 3 8, and 5 2 to 5 8 CCh (3 as present in the ACSF. When recording inhibitory postsynaptic currents (IPSCs), added to the ACSF. For all experiments, ACSF was saturated with an oxygen/carbon dioxide mixture (95% O 2 / 5% CO 2 ). Experiments were generally performed at 30 1C and heated ACSF was perfused acros s the tissue at a rate of 2.0 mL /min. Cannabinoid rela ted compounds (WIN and AM 251) were dissolved in DMSO and added to the ACSF for a final concentration of Intracellular S olutions To record EPSCs (i.e. Fig. 2 1, 5 8, and Chapter 4, excluding Fig. 4 6), an internal solution (low chloride) was used that contained (in mM): 140 cesium methanesulfonate ( CsCH 3 SO 3 ) 1 MgCl 2 3 NaCl, 0.2 Cs EGTA, 10 HEPES, 4 Na 2 adenosine triphosphate ( ATP ) 0.3 Na guanosine triphosphate ( GTP ) and 5 QX 314 Cl. When recording IPSCs, the internal solu tion (high chloride) contained (in mM): 95 CsCH 3 SO 3 55 CsCl, 1 MgCl 2 0.2 Cs EGTA, 10


39 HEPES, 2 Na 2 ATP, 0.3 Na GTP, and 5 QX 314 Cl. All internal solutions were adjusted using cesium hydroxide (CsOH) to a final pH of ~7.3. 10 mM BAPTA was added to the high chloride internal to block calcium in the postsynaptic mossy cell (Fig. 3 4A). On the day of u se, SR101 visualization via epifluorescence. Whole C ell R ecording Slice tissue was visualized using infrared differential interference contrast microscopy (IR DIC) on a n Olympus BX51WI microscope. Whole cell patch clamp experiments were conducted using borosilicate glass pipettes pulled by a Flaming/Brown P 97 electrode puller (both from Sutter Instruments, Novato, CA). Path electrode tip resistance typically ranged be membrane rupture, access resistance (R a uncompensated. R a was monitored throughout all experiments with any cell exceedi rejected from analysis. Also, any cell whose holding current (I hold ) became more negative than 500 pA when voltage clamped at 70 mV was discontinued and not used for analysis. Voltage clamp experiments were performed using an Axon Multiclamp C ommander 700A or 700B amplifier (Molecular Devices, Sunnyvale, CA). Data were sampled at 20 kHz, filtered at 2 kHz, and digitally recorded by a Digidata 1322A or 1400A analog to digital (A/D) converted using Clampex v. 9 or 10 (Molecular Devices). Mossy C ell I dentification Mossy cells were identified both pre and post hoc based on both physiological and anatomical features. In general, mossy cells exhibit the following attributes: 1) larger than average soma (as measured by membrane capacitance, gener 100 pA), spontaneous EPSCs when voltage clamped at 70 mV, 3) multipolar (multiple dendrites


40 projecting in different directions), and 4) spiny (large thorny excrescences on proximal dendrites). Membrane capacitance wa s measured immediately upon whole cell break in and synaptic activity was qualitatively determined over the first few minutes of experimental recording. Anatomical features were identified at the end of every experiment using epifluorescence. Since hilar interneurons are extremely varied in their physiology, including large and spiny neurons, only those neurons possessing all four criteria were considered mossy cells and used for analysis. A sample fluorescent image is included in figure 2 1. Stimulati on Evoked responses were obtained using an electrode connected to a constant stimulus isolator (World Precision Instruments, Sarasota, FL). For experiments presented in figures 3 3 to 3 6, 3 8, and 5 1, bulk evoked IPSCs (eIPSCs) were elicited with a co ncentric bipolar electrode (FHC, Bowdoin, ME) at a rate of 0.33 Hz. Minimally evoked EPSCs and IPSCs (meEPSCs and meIPSCs, respectively; see Chapters 4 and 5) were elicited using a monopolar electrode placed in a standard recording pipette filled with ACS F. Current intensity generally ranged from 20 to The goal of minimal stimulation is to isolate one or just a few (i.e. < 4) presynaptic axons. To ensure such isolation, meEPSC s and meIPSCs were subject to the following criterion: 15 2A). This criterion was sufficient for the meIPSC experiments presented in Chapter 5. Since the minimal stimulation experiments in Chapter 4 required the isolation of mossy fibers from non mossy fibers, four additional criteria were used along with the sharp threshold to distinguish these different glutamate inputs: rise time, latency of event onset, frequency facilitation, and DCG IV sensitivity. The first two event rise time and latency from stimulus


41 were much shorter for MFs than non MFs. Third, mossy fibers express robust short term potentiation (frequency facilitation). Briefly, a five pulse train of stimuli (16.67 Hz) was delivered to the isolated axon, resulting in a significant potentiation of MF meEPSCs (Fig. 2 2B) while non MF meEPSCs exh ibited only a small increase. Fourth, MF terminals express presynaptic group II metabotropic glutamate receptors and activation of these mGluRs significantly inhibits MFs vs. non MFs (some perforant path axons also express the same receptors but are less well inhibited by their activation). Therefore, isolated meEPSCs were considered MF in origin if they fit the first 3 criteria and were inhibited by more than 80% by DCG IV (Fig. 2 2C). An exception to the 80% rule was made for those experiments presente d in Fig. 4 2, when some cells were treated with DCG IV prior to full recovery from baclofen. In these cases, a 50% block was considered acceptable. Focal D rug A pplication In some cases, direct local drug application was necessary. To deliver drugs di rectly to their target, glass pipettes were pulled with an inner diameter slightly larger (~1 used for whole cell patch clamp and then filled with drugs diluted to the appropriate concentration in ACSF. Spritz pipettes were connected to a Picospritzer III (General Valve, Fairfield, NJ). For the experiment in figure 3 sucrose and placed at the top of the tissue, above the patched mossy cell. Sucrose was delivered for 2 min at ~20 psi. For isola tion of mossy fibers as presented in figure 4 IV was loaded into the spritz pipette. The pipette was placed such that a small depression was observed in the tissue just above the cell during pressure ejection of the drug (10 s, ~20 psi). For the glutamate spritz also shown in figure 4 2, pipettes were loaded with 50 10 ms spritz at 20 psi was delivered at 0.05 Hz. The glutamate spritz was intended to replicate a


42 synaptic response; therefore, a small pressure wave was ob served on the soma of the patched cell for each application. Data A nalysis Spontaneous and M iniature S ynaptic E vents Spontaneous and miniature IPSCs (sIPSCs and mIPSCs, respectively) were detected, with appropriate parameters, using either MiniAnalysis v. 6.03 (Synaptosoft, Decatur, GA) or custom written event detection software in OriginPro v. 7.5 or 8 (OriginLab, Northhampton, MA). When using MiniAnalysis, event tables were imported to OriginPro for further analysis. Power analysis for data presented in figures 3 1 and 3 where absolute power is the summation of all spectral bins in the stated range. Evoked S ynaptic E vents Minimally evoked EPSCs and IPSCs were manually analyzed for contamination by spontaneous synaptic activity and multiple peaks (i.e. responses from multiple presynaptic axons) and rejected if such problems were observed. Cells with > 50% sweeps re moved for such problems were not analyzed further. Errors are reported as SE. DSI s baseline period, 5 s depolarization from 70 to 0 mV, 90 s recovery period, 60 s interval period (to separate sets) For DSI of evoked responses, eIPSCs were elicited at 0.33 Hz in both the baseline and recovery periods. The measurement periods from the relevant periods were: baseline 24 s (8 sweeps) immediately prior to depolarization; DSI 6 s (2 sweeps) immedia tely following depolarization; recovery 24 s (8 sweeps) at end of recovery period (66 90 s after depolarization). Average response amplitude or sIPSC frequency was used to determine the level of DSI.


43 The following statistical test was developed to obj ectively assess the presence of DSI: the DSI period was compared to the baseline and recovery periods. If the DSI period was significantly smaller (see Statistics section) than both the baseline and recovery periods (2.37 standard error (SE) of the rele Allowing for natural fluctuations in cell signaling, the average of 3 to 5 sets was required to show DSI in order to conclude the presence of retrograde endocannabinoid signaling. Tonic and P hasic C urrent M easurements Changes in spontaneous IPSCs (phasic current) are most often measured using traditional event detection. Changes in tonic current, on the other hand, are most often measured as differences in the holding current ( I hold ) However, such measurements must be carefully obtained to prevent contamination by sIPSCs (Scimemi et al. 2005; Prenosil et al. 2006). Alternatively tonic currents have been characterized using noise analysis (Brickley et al. 1996; Mtchedlishvili and Ka pur 2006) or by fitting a Gaussian curve to an all points histogram of sampled data (Wall and Usowicz 1997; Petrini et al. 2004; Glykys and Mody 2006; Glykys et al. 2006). Unfortunately, these techniques generally fail to describe simultaneously changes i n both tonic and phasic currents. For the current manuscript, we developed an analytical method directly based on that described in a recent report from Glykys and Mody (SFN abstract, 327.4, 2006). In brief, continuous time series data was recorded at 20 kHz, down samp led to 5 kHz, and broken into 5 s intervals. For each consecutive interval, an amplitude histogram was constructed from all 25,000 data points. A nonlinear least squares fit (NLSF) was then performed on a portion of each amplitude histo gram using the Gaussian function given in Eq uation 2 1


44 (2 1) where A is the area under the normal curve, w is the width at peak amplitude, and x c is the x value at center of the curve. Importantly, the portion of the amplitude histogram used for NLSF fitting excluded all amplitude bins left of center that had less than the number of counts in the largest amplitude bin. Since physiological nois e in the absence of spontaneous IPSCs is expected to be normally distributed, x c from the best fit curve was accepted as the mean tonic current for each interval. In order to calculate phasic current we relied on the fact for a purely Gaussian distributio n, x c of the best fit curve obtained as above will be equal to the mean value of all the data in the bin. Thus, in intervals with little or no spontaneous activity, the difference in area between the predicted Gaussian curve and experimental data should b e very close to 0, as described by Equ ation 2 2. (2 2) By contrast, in intervals with appreciable spontaneous activity (due to increased sISPC frequency and/or amplitude), spontaneous inward currents create a leftward skew to the a mplitude histogram that will not affect x c from the best fit curve, but that will cause a significant shift in the sum of (x x c ) away from 0. Therefore, this sum was considered to represent total phasic current per interval. Glykys and Mody have tested this technique against real and simulated sIPSCs and reported it to be sensitive to minor (~15%) changes in both phasic and tonic currents. We verified our own code in two ways. First we demonstrated that adding a standing current to all points selective ly modulates tonic current, while doubling the amplitude of all points > 2


45 standard deviations from the mean selectively modulates phasic current (data not shown). Second, we plotted phasic and tonic currents analyzed in 1 s intervals on top of raw sIPSCs as in Fig. 2 3. These plots emphasize the ability of the tonic current to accurately identify I hold regardless of sIPSC frequency, and also demonstrate that phasic current clearly increases with event frequency, showing sensitivity to as little as 1 event Statistics Statistical hypothesis testing was done using OriginPro and Excel 2003 or 2007 (Microsoft, Seattle, WA). In general, two tailed one or two sample (where appropriate) tests were used to determine significance of bath applied drugs to evoked synaptic events (including eIPSCs, meIPSCs, and meEPSCs). Null hypotheses: mean = 1 for data normalized to baseline mean (one sample); mean1 mean2 = 0 (two sample). Kolmogorov Smirnoff (KS) tests were used to determine changes in frequency a nd amplitude of mIPSCs. Pearson correlati on and linear regression analyses were used to ascertain correlations between drug effects. Unless otherwise stated, all drug effects are listed as percent of baseline (100) standard error (SE). Unless otherwise stated, a single asterisk (*) indicates p < 0.05 while a double asterisk (**) indicates p < 0.01.


46 Fi gure 2 1. Mossy cell anatomy. An example of a mossy cell with a large, multipolar dendritic tree. Large thorny excrescences on the proximal dendrites are indicative of mossy fiber (MF) synapses. Smaller spines are located on more distal dendrites and include both MF and non MF glutamatergic contacts. GABAergic synapses are dispersed over both the soma dendritic tree. The axon is clearly vi sible as a much thinner aspiny process with many small filopodial extensions.


47 Figure 2 2. Isolation of minimally evok ed mossy fiber mediated EPSCs. A) An example of a Stim ulation duration was 0.1 ms. Insets represent individual and average responses at every stimulus intensity (gray), with the average response overlaid in black. Scatter plot indicates average amplitude and standard error of the sweeps shown. B) The respo nse shown in A also demonstrated strong frequency facilitation when stimulated average of 5 trials while the scatter plot indicates the peak amplitude of each pulse normalize d to p1 and measured from the average sweep. C) This response also IV. Specifically, p 1 was inhibited by 83% while p2 was blocked by 80%. The paired pulse ratio increased from 1.9 to 2.26. All panels in this figure represent data from a single representative experiment. The latter two features (as indicated in panels B C) identified this response as being of MF origin.


48 Figure 2 3. Simultaneous measurement of phasic and tonic currents. A) In the absence of spontaneous activity, tonic current (dashed lines) accurately reflects I hold while phasic current (dotted lines) calculated as described in the methods is near 0. For simplicity, phasic current is plotted in all pane ls as an offset from tonic current. B) Phasic current shows excellent sensitivity to spontaneous activity. In this panel, a single event produces a phasic current of 7.3 pA when calculated over a 1 s interval. C) Phasic current clearly increases with th e level of spontaneous activity, while tonic current continues to accurately report I hold even during periods of very high event frequency. Legend and scales apply to all panels. Data are sIPSCs taken from baseline (A B) and early CCh period (C) of cell shown in Fig. 4 8A.


49 CHAPTER 3 ENDOCANNABINOID MEDIATED DEPOLARIZAT ION INDUCED SUPPRESSION OF INHIBITION IN HILAR MOSSY CELLS OF THE R AT DENTATE GYRUS 1 Introduction Depolarization induced suppression of inhibition (DSI) is a form of short term synaptic p lasticity that is dependent upon retrograde transmission mediated by endogenous cannabinoids (eCBs) DSI has been observed in a number of brain areas where presynaptic type 1 cannabinoid ( CB1 ) receptors are expressed, including the amygdala, substantia ni gra, basal ganglia, neocortex, cerebellum, and hippocampus (Llano et al., 1991; Pitler and Alger, 1992b; Kreitzer and Regehr, 2001b; Wilson and Nicoll, 2001; Trettel and Levine, 2003; Yanovsky et al., 2003; Zhu and Lovinger, 2005; Engler et al., 2006). In fact, across many areas of the central nervous system variations in factors such as the mechanism of induction (and source) of eCB release, the duration of eCB dependent signaling, and the location of cannabinoid receptor expression have implicated eCBs either directly or indirectly in multiple forms of synaptic plasticity (Kreitzer and Regehr, 2001a; Gerdeman et al., 2002; Robbe et al., 2002; Chevaleyre and Castillo, 2003, 2004; Hashimotodani et al., 2005; Safo and Regehr, 2005). Several recent lines of evidence have begun to suggest that eCBs might play a similarly important role in modulating synaptic activity in the dentate gyrus, an area that serves as the entry point for the primary afferent projections to the hippocampus from both the medial septum and the entorhinal cortex (Johnston et al., 1998). For example, it is clear that both CB1 receptors and the enzymes necessary for degradation of eCBs are expressed in the dentate gyrus (Tsou et al., 1998; Katona et al., 1999; Katon a et al., 2000; Romero et al., 2002; Gulyas et al., 2004). 1 mediated depolarization induced suppression of inhibition in hilar mossy cells of the rat dentate gyrus. Nov;96(5):2501 12


50 Further, previous reports have demonstrated that exogenous CB1 agonists can modulate the activity of inhibitory inputs to dentate granule cells (Nakatsuka et al., 2003) and that eCB dependent sign aling is enhanced in the dentate following febrile seizures (Chen et al., 2003). Finally, a recent study has provided the first clear evidence that DSI may be induced by experimental excitation of granule cells under normal conditions (Isokawa and Alger, 2005). In the present study we test the hypothesis that endocannabinoid dependent retrograde signaling can also be initiated by depolarization of mossy cells found in the hilar region of the dentate gyrus. These cells are unique among local circuit neur ons in the hippocampus and dentate gyrus in that they are glutamatergic rather than GABAergic (Scharfman, 1995). They are also extremely sensitive to both ischemia and excitotoxicity (Freund and Magloczky, 1993; Hsu and Buzsaki, 1993; Magloczky and Freund 1993), have a strong longitudinal projection (Amaral, 1978; Amaral and Witter, 1989; Buckmaster et al., 1992; Buckmaster and Schwartzkroin, 1994), and have been consistently implicated (through either their loss or dysfunction) in several competing theor ies on the etiology of temporal lobe epilepsy (Sloviter, 1991; Lothman et al., 1996; Houser, 1999; Santhakumar et al., 2000; Ratzliff et al., 2002; Ratzliff et al., 2004). Our results indicate that excitation of hilar mossy cells produces a robust inhibit ion of local GABAergic transmission and suggest a prominent role for eCB dependent retrograde signaling in hilar neurophysiology. Results DSI of Spontaneous IPSCs in Hilar Mossy Cells Identified hilar mossy cells (see Chapter 2) were voltage clamped at 70 mV in whole cell mode using a CsCH 3 SO 3 based internal solution that contained ~60 mM CsCl. Approximately 3 min after obtaining a whole cell recording, the ionotropic glutamate receptor antagonists NBQX (10 M), and APV (40 ) were bath applied in order to isolate spontaneous IPSCs and 3


51 CCh was applied in an attempt to increase spontaneous inhibitory postsynaptic current ( sIPSC ) frequency. Mossy cells could be readily divided into three groups based on the nature of the s pontaneous activity that emerged under these experimental conditions. In one group of cells, application of CCh produced a large, sudden, and stable increase in sIPSC frequency and amplitude (Fig. 3 1A). In four cells in which this effect was observed, th e average sIPSC frequency after application of CCh was 14.5 3.42 Hz while the average amplitude was 71.5 6.44 pA. A power analysis revealed that spontaneous activity in these cells uniformly showed 1 or more strong peaks in the theta band (e.g. 4 14 H z, Fig. 3 1B, 3.38%). In this population of cells, depolarization from 70 mV to 0 mV for 5 s produced robust DSI that could be observed as a transient decrease in both frequency and amplitude of sIPSCs (Fi g. 3 1C E, 46.4 13.3% and 49.9 amplitude was observed, but sIPSCs still occurred in the pres ence of 3 CCh at a frequency amplitude was significantly smaller (8.1 1.3 Hz, and 39.8 3.80 pA, respectively), and a power analysis uniformly failed to revea l any significant peaks between 2 and 50 Hz (e.g. Fig. 3 2B). Interestingly, we found that depolarization of hilar mossy cells could still produce a robust decrease in the frequency of these sIPSCs (Fig. 3 2C, E, 34.3 in this population there was no longer any detectable effect of depolarizatio n on sIPSC amplitude (Fig. 3 2D, E, 102 2.10% of baseline, n=3, p = 0.36). Finally, in a third group, cells either failed to show clear increases in sIPSC frequency (beyond 3 Hz) a fter bath application of 3 CCh or showed transient bursts in activity that


52 were not sustained enough to allow for reliable detection of DSI. This group represented the strong majority (~90%) of all mossy cells examined. Consequently, we also examined the ability to detect DSI in hilar mossy cells as a transient reduction in the amplitude of evoked IPSCs. DSI of Evoked IPSCs in Hilar Mossy Cells Bicuculline sensitive (e.g. GABA A receptor mediated) evoked IPSCs were generated in the presence of NBQX and APV at a rate of 0.33 Hz using a bipolar stimulator placed in the hilus. In the absence of bath applied CCh, depolarization of hilar mossy cells from 70 mV to 0 mV for 5 s transiently reduced evoked IPSC ( eIPSC ) amplitude by 18.9 5.00% (Fig. 3 3A, n=6 p<0.01). Interestingly, when DSI was examined in a separate population of mossy cells exposed to 3 CCh, the effect was significantly more robust (Fig. 3 3B, 30.69 3.8% reduction in baseline, n=13, p<.001). Conversely, we also determined that withi n a single population of mossy cells, DSI observed in the presence of 3 CCh was significantly reduced by bath application of 5 atropine (from 34.0 1.4% in CCh to 19.2 4.2% following wash in of test for means, data not shown). In order to demonstrate that this apparent inhibition of DSI by atropine was not due to rundown, we tested DSI at ~3 min intervals for at least 45 min in three cells that exhibited robust DSI in the presence of CCh. Our results indicated that DSI was generally quite stable over this time frame (e.g. 48 5% reduction, first three sets, 45 5% reduction, last three sets, data not shown). Because the effect of CCh on DSI of eIPSCs was much more reliable than its effect on sIPSC frequency and amplitude, further experiments into the mechanism of DSI in hilar mossy cells relied on evoked responses.


53 DSI in Hilar Mossy Cells is Mediated by Calcium Dependent Release of eCBs that Act on Presynaptic CB1 Receptors We observed th at the magnitude of DSI in hilar mossy cells depends directly on both depolarization duration and on postsynaptic calcium influx. Specifically, in separate groups of cells, we measured DSI of 10.3 1.84% following a 0.1 s depolarization, 21.9 2.91% fol lowing a 1 s depolarization, and 30.7 3.8% following a 5 s depolarization (n=8, 9, and 13 respectively, data not shown). Further, we demonstrated that DSI was largely eliminated in cells that were filled with 10 mM BAPTA via the recording pipette (eIPSC amplitude following depolarization was reduced by only 2.53 2.08%, n=18, Fig. 3 4A). We also found that there was no effect of depolarization on the response to rapid local application of exogenous GABA (100 via a picospritzer) in four cells that co llectively showed DSI of eIPSCs of 20 3.0% (data not shown). Cumulatively, these results suggest that calcium dependent release of a retrograde messenger is involved in DSI as observed in hilar mossy cells, and consistent with other systems, several lin es of experimental evidence implicate eCBs in that role. Specifically, we found that in a group of mossy cells that showed 30 5.0% reduction in eIPSC amplitude following a 5 s depolarization to 0 mV, wash in of AM minutes produced a statistically significant block of DSI (to 9.0 3%, n=6, p<0.01, Fig. 3 4B). Similarly, incubation of slices in 5 AM 251 prior to whole cell recording completely eliminated DSI (eIPSC amplitude following depolarization was 100.76 2.08% of control, n=14, Fig. 3 4C). We also demonstrated that bath application of WIN55,212 eIPSC amplitude (by 36.4 7.17%, n=4, a value comparable to that transiently produced by a 5 s depolarization) and completely occludes DSI (Fig. 3 5A).


54 Activation of Presynaptic CB1 Receptors Preferentially Inhibits Calcium Dependent Exocytosis We next sought to determine whether calcium dependence of exocytosis from GABAergic afferents to hilar mossy cells predicts sensitivity to CB1 mediated inhibition. As a first approach we examined the effect of depolarization induced release of eCBs on action potential independent miniature IPSCs recorded in the presence o f 1 TTX. Because voltage gated calcium channels rely upon the depolarization initiated by TTX sensitive Na + channels for their activation, these conditions are expected to dramatically reduce calcium dependent exocytosis. In order to eliminate false negatives due to failures in retrograde transmission, experiments on miniature IPSCs were only completed in cells that had previously demonstrated robust DSI of eIPSCs under normal conditions. In 6 of 8 cases we found that depolarization of hilar mossy c ells (from 70 mV to 0 mV for 5 s) had no significant effect on frequency or amplitude of miniature IPSCs ( mIPSCs ) recorded in normal ACSF (frequency: 111.5 12.6% of control, amplitude: 98.7 2.44% of control, p>0.4 in both cases, Fig. 3 6D). In sharp contrast, in 8 of 11 cases where mIPSCs were tested in ACSF that contained 15 mM KCl and 5 mM CaCl 2 (conditions designed to rescue calcium dependent exocytosis) depolarization reduced mIPSC frequency by 32.9 4.6% without affecting mIPSC amplitude (101 1.59% of control, p<0.01, p>0.5 respectively, Fig. 3 6D). In fact, in three cells where DSI was examined in all three conditions, we sequentially observed robust DSI of eIPSCs, followed by no DSI of mIPSCs, followed by robust DSI of mIPSCs in high KCl a nd CaCl 2 One such series of ex periments is shown in Fig. 3 6A C. Although we did observe DSI in 2 of 8 cases in normal media, these cells had notably high mIPSC frequency than others for these conditions (13.7 3.1 Hz vs. 6.2 2.2 Hz respectively). Therefore we assessed mIPSC frequency in three cells after a 25 minute wash in of an external


55 solution that contained 0 mM Ca 2+ 3.9 mM Mg 2+ and 100 BAPTA AM and found it significantly reduced (e.g. from 7.0 1.4 Hz to 4.1 0.7 Hz, p=0.03 (one taile d), 0.07 (two tailed), data not shown). We also tested the ability of bath applied WIN55,212 2 (5 ) to modulate the frequency of calcium independent mIPSCs induced by local application of a hypertonic solution (100 mM sucrose, for previous evidence of calcium independence, see Fatt and Katz, 1952; Rosenmund and Stevens, 1996; Khvotchev et al., 2000). Specifically, we found that local application of 100 mM sucrose (2 min X ~20 pounds per square inch ( psi ) ) to the surface of the slice just above a patche d hilar mossy cell reliably increased mIPSC frequency to ~500% of control (e.g. to ~20 Hz), and further that this effect was unaltered in a separate population of cells that had been pretreated with 5 WIN55,212 2 for a minimum of 12 min (Fig 3 7A C). F inally, we demonstrated that although WIN55,212 2 failed to reduce the frequency of calcium independent mIPSCs induced by sucrose, it was able to reverse presumably calcium dependent increases in mIPSC frequency produced by bath applying 10 mM KCl (Fig. 3 7D E). Endocannabinoid Mediated Signaling in the Dentate Gyrus is Subject to Tight Spatial Constraints A final question of interest was whether eCBs released due to depolarization of one mossy cell could lead to reduction in transmitter release from GA BAergic afferents to nearby mossy cells. In order to address this question we performed paired whole cell recordings from 14 mossy cells (7 pairs) in which both cells showed robust DSI upon depolarization and in which the cell somas were located within 10 three dimensional space). In that population, we failed to observe any detectable DSI in the non depolarized cell in 11 of 14 cases (DSI in depolarized cells was 27.4 4.65%, n=14, while baseline in non depolarized cells was reduced by only 4.22 2.75% from baseline, n=11, Fig. 3


56 8B C). However, in 3 of 14 cases, some DSI was detectable in the non depolarized cells. In those 3 pairs, both the distance between cell somas and the extent of DSI ob served in the non from 3% to 58%, average 25.6 16.7%, respectively). The most prominent example of this phenomenon is shown in Fig. 3 8D E. Nevertheless, across the entire dataset there was no correlation between the amount of DSI observed in the non depolarized cell and the somatic Further, this correlation remained i nsignificant when the somatic distance was scaled by the amount of DSI in the depolarized cell (p=0.27). Discussion In the present study we used whole cell recording techniques in acute preparations of the rat dentate gyrus to examine retrograde transmis sion initiated by depolarization of hilar mossy cells. Our results indicate that depolarization of hilar mossy cells produces calcium dependent release of eCBs that is facilitated by activation of muscarinic acetylcholine receptors. We further provide sp ecific experiments to indicate endocannabinoids liberated via depolarization of mossy cells activate presynaptic CB1 receptors on traditional GABAergic afferents, and that this activation (whether by endogenous or exogenous agonists) preferentially inhibit s calcium dependent exocytosis. We further provide evidence from paired whole cell recordings that indicates eCB mediated retrograde transmission in this system is subject to tight spatial constraints. Overall, our results represent just the second examp le of eCB mediated retrograde signaling occurring under normal conditions in the dentate gyrus (Isokawa and Alger 2005), the first formal report of DSI at this synapse, and indeed, the first report of endocannabinoids being liberated by depolarization of a local circuit neuron in the hippocampus or dentate gyrus


57 (although see Howard et al., SFN Abstract 808.11, 2003). There are several areas in which our results deserve careful comparison to those previously reported in other systems. The Role of Muscarini c Acetylcholine Receptors in DSI In early experiments on DSI in area CA1 of the hippocampus, CCh was often employed because of its ability to increase sIPSC frequency and amplitude (Pitler and Alger, 1992a b ; Behrends and ten Bruggencate, 1993). DSI of s pontaneous IPSCs was almost always absent in cells that lacked this CCh induced effect and yet DSI of evoked IPSCs was readily apparent, even in the absence of CCh and presence of atropine (Martin and Alger, 1999). Cumulatively, those observations led to the conclusion that activation of muscarinic acetylcholine receptors ( mAChRs ) was not necessary for DSI per se but rather allowed easy detection of DSI by increasing the activity of DSI sensitive (e.g. CB1 positive) interneurons. Implicit in that conclusion, and strongly supported by immunohistochemical analysis (e.g. Katona et al., 1999; Tsou et al., 1999), was the suggestion that there exists a clear population of GABAergic afferents to CA1 pyramidal cells that are CB1 negative. In the present study, we report robust DSI of large amplitude, high frequency, CCh induced sIPSCs. These sIPSCs show strong peaks a t theta frequencies upon spectral analysis, and DSI is apparent as a transient reduction in frequency, amplitude, and theta power. However, while Martin and Alger (1999) report sustained large amplitude CCh induced sIPSCs in > 50% of CA1 pyramidal cells e xamined, we see similar activity in < 10% of mossy cells tested. The mechanism by which CCh increases the amplitude of sIPSCs in these cases is not clear. One possibility is that CCh causes a dramatic increase in quantal content, while a competing explan ation is that CCh promotes synchronous release from somatic and perisomatic GABAergic inputs. While we did not perform experiments aimed directly at this question, both early


58 speculation and more recent experimental work in area CA1 favors the later hypot hesis (Alger et al., 1996; Reich et al., 2005). In the present study, we also report DSI of smaller amplitude sIPSCs that lack clear peaks in the power spectrum at theta frequencies. In these cases DSI is apparent as a transient reduction in sIPSC frequ ency, with no detectable change in sIPSC amplitude. While it might be argued that this represents a new and non traditional form of DSI expression, it is in fact exactly what would be predicted from a purely presynaptic effect of eCBs on asynchronous rele ase events of low quantal content. In that light, it becomes interesting to note that, in contrast to our results, CCh insensitive presumably non theta sIPSCs recorded in CA1 pyramidal cells have generally been reported to be DSI insensitive (Martin and A lger, 1999; Martin et al., 2001). While the reasons for this apparent difference are not yet entirely clear, immunohistochemical studies have demonstrated that innervation by parvalbumin immunoreactive terminals, although present, is much weaker in mossy cells than in traditional hippocampal principal cells, while innervation by cholecystokinin immunoreactive terminals remains robust (Acsady et al., 2000). It should be interesting for future studies to begin to examine the resultant differences in CB1 med iated modulation of network activity. We have further demonstrated that although DSI of evoked IPSCs is apparent in the absence of cholinergic stimulation, the magnitude of this effect is enhanced by bath application of 3 CCh in an atropine sensitive ma nner. Based on current data we cannot definitively determine whether this effect of CCh on DSI of eIPSCs is mediated presynaptically or postsynaptically. A potential presynaptic mechanism might involve a CCh mediated increase in the number of DSI sensiti ve afferents that are recruited by the stimulus; while a postsynaptic mechanism might involve direct mAChR mediated facilitation of depolarization induced eCB


59 release. While significant further work will be necessary to distinguish between these possibi lities in the current system, significant precedence for the postsynaptic hypothesis is developing based on recent work in CA1 (Kim et al., 2002; Ohno Shosaku et al., 2003; Hashimotodani et al., 2005). Presynaptic Effects of Endocannabinoids Although it is clear that activation of CB1 receptors is negatively coupled to exocytosis in a number of CNS synapses, the precise mechanism by which this occurs has been a matter of some controversy. Activation of CB1 receptors in cultured hippocampal neurons has been directly linked to G i /G o mediated inhibition of N and P/Q type calcium channels (Twitchell et al., 1997; Sullivan, 1999). This result has been used to suggest that presynaptic CB1 receptors in acute CNS preparations may inhibit transmitter release prima rily by reducing action potential mediated calcium influx. Indeed, consistent with that hypothesis, miniature IPSCs (recorded in the presence of 1 TTX) in hippocampal slices have generally been shown to be insensitive to CB1 activation in normal media, but sensitive to both CB1 agonists and Cd 2+ (indicating their calcium dependence) following bath application of high KCl (Pitler and Alger, 1994; Hajos et al., 2000; Hoffman and Lupica, 2000; Wilson and Nicoll, 2001). By contrast, endogenous and/or exog enous CB1 agonists have been shown to inhibit presumably calcium independent miniature synaptic events in several other brain areas including the cerebellum, nucleus accumbens, and spinal cord (Llano et al., 1991; Hoffman and Lupica, 2001; Kreitzer and Reg ehr, 2001b; Morisset and Urban, 2001; Robbe et al., 2001; Diana et al., 2002). Cumulatively, these data have implied that, contingent on the specific synapse involved, activation of CB1 receptors may inhibit exocytosis either by reducing calcium influx in to the presynaptic terminal and/or via direct modulation of the release machinery. Our own results in the dentate gyrus indicate that depolarization induced release of eCBs generally fails to reduce mIPSC frequency in normal


60 ACSF, and yet reliably does so in external solution that contains 15 mM K + and 5 mM Ca 2+ Similarly, we have further demonstrated that WIN55,212 2 fails to reduce robust calcium independent exocytosis produced by focal application of a hypertonic solution, and yet reduces the frequenc y of mIPSCs recorded in the presence of high extracellular K + These results are consistent with the conclusion that CB1 receptors on GABAergic afferents to hilar mossy cells preferentially inhibit calcium dependent exocytosis, but do not directly speak t o whether the site of action is at voltage gated calcium channels and/or downstream of calcium influx. An important footnote here is that we did observe DSI of mIPSCs in normal ACSF in 2 of 8 cases. One possible explanation for these outliers is that the re is an additional mechanism for CB1 mediated inhibition of calcium independent exocytosis. However, we also noted that mIPSC frequency could sometimes be reduced in hilar mossy cells by switching to a Ca 2+ free external containing 100 BAPTA AM, sugge sting that TTX may not always completely eliminate calcium dependent exocytosis of GABA. Considered together, these results suggest that the calcium dependence of mIPSCs in hilar mossy cells most likely predicts their sensitivity to inhibition by CB1 acti vation. In that respect, our conclusions parallel those of Yamasaki et al. (2006), who recently and convincingly demonstrated that the differential sensitivity of miniature excitatory postsynaptic currents ( mEPSCs ) vs. mIPSCs in cerebellar Purkinje cells to WIN55,212 2 in normal (2 mM) Ca2+ can be explained by a previously unnoted differential Ca 2+ dependence of the miniature events. Spatial Constraints on Endocannabinoid Dependent Signaling Despite the rapid nature of recent progress, relatively few studies have directly examined spatial constraints on eCB dependent retrograde signaling. In the cerebellum, strong depolarization of Purkinje cells has produced clear spread of eCB dependent ret rograde signaling to both inhibitory and parallel inputs to non depolarized cells (Vincent and Marty, 1993; Kreitzer


61 et al., 2002), and yet more selective induction of eCB release has shown exquisite specificity in retrograde transmission, even within the dendritic tree of a single Purkinje cell (Brown et al., 2003). In area CA1 of the hippocampus, Wilson and Nicoll (2001) have reported that eCB inhibition detectable in non depolarized cells is strongly correlated with somatic distance from a depolarized cell. Our results in the dentate gyrus indicate that it is generally quite difficult to detect spread of eCB dependent signaling between mossy cells with somatic distan ces between depolarization eliminates the possibility of CB1 negative afferents masking the spread of eCB dependent signaling. Although we observed exceptional cases where spread of eCB dependent signaling clearly affected afferent inputs to a non depolarized cell, we did not find any clear correlation between this effect and somatic distance and/or magnitude of DSI. In our view this suggests that eCB dependent retrog rade signaling initiated by depolarization of mossy cells is important, experimental detection of spread of retrograde transmission to non depolarized cells likely depe nds on one or more uncontrolled factors such as degree of dendritic overlap.


62 Figure 3 1. Depolarization of hilar mossy cells transiently reduces both frequency and amplitude of C Ch induced theta band sIPSCs. Data in panels A D are all from a single representative experiment. A) Left: Bath application of 3 M CCh caused a rapid increase in sIPSC frequency and amplitude. Right: several minutes later, depolarization of the mossy cell from 70 mV to 0 mV for 5 s (bar) caused a transient reduction in b oth frequency and amplitude of CCh induced sIPSCs. Insets are averages of 3 6 sIPSCs recorded during the baseline, DSI, and recovery period s, respectively (see Chapter 2, DSI for details) B) A power analysis of the 1 min of data prior to depolarization reveals several strong peaks in the theta band. These peaks were completely absent during the DSI period (data not shown). Average theta power for this cell during the 1 min preceding depolarization was 50% (see Chapter 2). C D) Cumulative probability histograms reveal that depolarization of this cell produced a statistically significant decrease in both frequency and amplitude of sIPSCs. Events were collected from the baseline and DSI periods. Thick lines in each histogram represent the baseline peri od, thin lines represent the DSI period (p<0.001 in both cases, K S Test). E) Summary plot indicating the average effect of depolarization on sIPSC frequency and amplitude in 4 cells that displayed an effect of CCh as shown in panel A. At least 2 3 sets of DSI were averaged for each cell to get a representative effect. Every cell tested that exhibited high theta power sIPSCs after application of CCh showed robust DSI of both frequency and amplitude. A single asterisk indicates p < 0.05, while a double asterisk indicates p < 0.01 on a two tailed Test, n=4.


63 Figure 3 2. Depolarization of hilar mossy cells reduces frequency but not amplitude of non theta sIPSCs recorded in the presence of 3 CCh. Data in panels A D are all from a s ingle representative experiment. A) Sample sIPSCs recorded in the presence of 3 CCh in a cell where CCh failed to produce large amplitude high frequency theta band sIPSCs as in Fig. 3 1. Insets highlight the effect of depolariz ation from 70 mV to 0 mV for 5 s (bar). B) Power analysis indicating that this cell, as every cell that lacked a clear CCh effect on sIPSCs, failed to show clear peaks in the theta band. C D) Cumulative probability histograms reveal that depolarization of this cell produced a statistically significant decrease in sIPSC frequency (p<0.001, K S Test), with no change in amplitude (p= 0 .87, K S Test). E) Summary plot indicating the average effect of depolarization on sIPSC frequency and amplitude in 3 cells that lacked an effect of CCh as shown in panel A, but still had baseline sIPSC freq uency Three sets of DSI were averaged for each cell to get a representative effect. Asterisk p < 0.05, on a two Test, n=3.


64 Figure 3 3. DSI of evoked IPSCs in hilar mossy cells is enhanced by bath applied CCh. Evoked IPSCs were generated at a rate of 0.33 Hz using a bipolar stimulator placed in the hilus. NBQX and APV were present for all experiments. A) In the absence of bath applied CCh, depolariza tion of the mossy cell from 70 mV to 0 mV for 5 s (bar) caused a mild but significant reduction of the mean eIPSC amplitude. B) An identical depolarization produces significantly greater DSI in mossy cells exposed to 3 CCh. Insets in A B are averages of 3 6 traces collected during the baseline and recovery periods, or of the two traces immediately following depolarization for the DSI period. C) Summary data indicating amount of DSI produced in mossy cells in the absence and presence of CCh. Third ba r shows, for comparison, the amount of DSI observed in CA1 pyramidal cells in the absence of CCh using identical techniques. Numbers on the bars are n values. Double asterisks indicate statistical te st (null hypothesis, mean=0). Single asterisk indicates two tailed p value = 0.07, one tailed p value = 0.04 on an unpaired two test assuming equal variances (F=0.36).


65 Figure 3 4. DSI in hilar mossy cells depends on postsynaptic ca lcium influx and is blocked by a CB1 receptor antagonist. A) When mossy cells were filled with 10 mM BAPTA (delivered via the internal solution), eIPSC amplitude following depolarization was 97.5 2.08% of baseline, n=18. B) DSI was significantly reduc ed by bath application of 5 AM 251. Filled circles indicate magnitude of DSI in a group of 6 mossy cells in control conditions (30 5.0%). Open circles indicate amount of DSI observed in the same 6 mossy cells following wash in of 5 AM 251, a CB1 3%). This represents a statistically Test). C) When slices were pre incubated in 5 M AM 251 prior to whole cell recording, DSI was comp letely eliminated (average eIPSC amplitude following depolarization was 99.2 1.31% of control, n=14). D) Summary plot indicating the amount of DSI observed in mossy cells filled with 10 mM BAPTA, mossy cells not exposed to AM 251, and mossy cells pretr eated with AM 251 (5 M) prior to whole cell recording. Insets in A C are averages of 2 6 individual traces recorded during the times indicated. In panel B, sample traces for all three periods are overlaid both before and after wash in of AM 251.


66 Fi gure 3 5. Bath application of an exogenous CB1 agonist reduces the amplitude of evoked IPSCs in hilar mossy cells and occludes DSI. A) DSI of evoked IPSCs was generated in hilar mossy cells once every 3.6 min using techniques and experimental conditions identical to those in earlier figures. Filled circles represent the normalized average amplitude for the baseline period; open circles represent the normalized average amplitude for test period (see Chapter 2). All data are normalized to the baseline amp one representative cell where black traces represent average eIPSC amplitude during the baseline period, and gray traces indicate average eIPSC amplitude during the test period. A p air of traces is shown from data collected both before and after treatment with 5 M WIN 55,212 2. B) Summary data for this experiment are presented as in prior figures, n =4.


67 Figure 3 6. Depolarization induced release of endogenous cannabinoids prefer entially inhibits calcium dependent exocytosis. Data presented in panels A C were all collected se quentially from a single cell. A) DSI of evoked IPSCs is present, generated as in earlier figures. Filled circles represent normalized averages from 3 set s of DSI. Insets are averages of 2 6 individual traces from the indicated period. B) Traces at the top show representative miniature IPSCs recorded following bath application of 1 M TTX as observed both before (Base) and immediately after (No DSI) a 5 s depolarization from 70 mV to 0 mV. A cumulative probability histogram is presented below the traces to indicate that depolarization did not decrease the frequency of mIPSCs in this cell. In fact, in this case, a slight, but not significant, increase w as observed (p=0.12, K S Test). C) Traces at the top show representative mIPSCs recorded in the same cell after increasing external KCl to 15 mM and external CaCl2 to 5 mM. Under these conditions, depolarization produced a significant reduction in mIPSC frequency (p=0.001, K S Test). D) Summary plot indicating average affect of depolarization in the conditions described in panels A C. Numbers on the bar are n values. Single asterisk indicates p=0.03 on unpaired Test. Double asterisks indi Test (Null hypothesis, mean=0).


68 Figure 3 7. Bath application of WIN55,212 2 preferentially inhibit s calcium dependent exocytosis. A) Application of hypertonic solution (100 mM sucrose, applied via a picospritzer to the surface of the slice, ~20 psi X 2 min, bars) caused a significant and reproducible increase in miniature IPSC frequency (App1: 506 104%, App2: 441 68.7%, n=5). B) An identical experiment was performed on a group of mossy cells pre treated wit h 5 M WIN55,212 2 for a minimum of 12 min (App. 1: 572 113%, App. 2: 493 106%, n=6). C) Summary plot indicating that there was no significant difference in the effect of sucrose on mIPSC frequency in the WIN55,212 2 treated group (dark bars) vs. the control group (white bars) on either the first or second sucrose application. The effect of sucrose was calculated as the average mIPSC frequency from 1 to 2 min after starting the sucrose application divided by the average mIPSC frequency in the 1 min i mmediately preceding sucrose application. D) Summary plot indicating that bath applied 5 WIN55,212 2 is able to effectively reverse KCl mediated increases in mIPSC frequency (n=3 out of ~6 tested. Cells were not included in analysis if 10 mM KCl fail ed to significantly increase mIPSC frequency). E) Raw data from a representative experiment recorded in the presence of 10 mM KCl (but before application of 5 M WIN55,212 2 (left traces), and ~25 min after application of WIN55,212 2 (right traces).


69 F igure 3 8. Spatial constraints on endocannabinoid mediated signal ing in the dentate gyrus. A) Dual whole cell patch clamp recording was used to monitor DSI of eIPSCs in both a depolarized cell and a non depolarized cell simultaneously. Pairs were rejec ted from analysis unless both cells showed robust DSI upon (their own) depolarization. B) Across all pairs examined, average DSI in the depolarized cell was 27.4 4.65%, n= 14 cells, 7 pairs. B) In 11 of 14 cases no DSI was detectable in the non depol arized cell (average eIPSC amplitude following depolarization was 95.8 2.75% of control). C) Summary plot indicating that in most cases, endocannabinoids released by depolarization of one mossy cell failed to induce DSI of afferent inputs to a nearby m ossy cell. Numbers on the bars are n values. D) Raw data from a single paired recording indicating an exceptional case where robust DSI in the depolarized cell clearly produced robust inhibition of afferent inputs to the non depolarized cell. E) Summar y plot indicating the average amount of DSI observed in both the depolarized and non depolarized cell in this pair over three trials (73.0 13.5% and 57.8 21.09%, respectively, See Results for additional information).


70 CHAPTER 4 PRESYNAPTIC INHIBITI ON OF EXCITATORY AFFERE NTS TO HILAR MOSSY C ELLS 2 Introduction Hilar mossy cells are the only glutamatergic non principal cells in the hippocampus or dentate gyrus (Soriano and Frotscher, 1994). Unlike conventional hippocampal principal cells, hilar mossy cel ls project largely along the septo temporal axis of the hippocampus where they are thought to innervate both dentate granule cells and GABAergic interneurons (Amaral and Witter, 1989; Schwartzkroin et al., 1990; Buckmaster et al., 1992; Scharfman, 1995; Bu ckmaster et al., 1996; Jackson and Scharfman, 1996). They also have a strong contralateral projection, are extremely sensitive to excitotoxicity, and are strongly implicated in the etiology of temporal lobe epilepsy (Sloviter, 1991; Bekenstein and Lothman 1993; Sloviter, 1994; Buckmaster et al., 1996; Buckmaster and Dudek, 1997; Jefferys and Traub, 1998; Buckmaster and Jongen Relo, 1999; Ratzliff et al., 2002; Coulter, 2004; Santhakumar et al., 2005). Mossy cells are believed to receive direct monosyn aptic excitatory inputs from both dentate granule cells and CA3 pyramidal cells (Ribak et al., 1985; Ishizuka et al., 1990; Frotscher et al., 1991; Penttonen et al., 1997; Coulter, 2004). They are thus largely driven by axon collaterals of some of the mos t heavily studied synaptic pathways in the CNS. For example, the mossy fiber ( MF ) pathway from dentate granule cells to CA3 pyramidal cells is well known for its tight spatial localization, strong frequency facilitation, clear sensitivity to group II meta botropic glutamate receptor ( mGluR ) and presynaptic form of long term potentiation (for review, see Henze et al., 2000; Urban et al., 2 47


71 2001; Nicoll and Schmitz, 2005). Collateral projections of mossy fibers to interneuron targets in CA3 have also been extensively studied, and importantly, have shown both anatomical and functional differences when compared with mossy fiber inputs to CA3 pyramidal cells (Toth et al., 2000; Lei and McBain, 2002; Lawrence et al., 2004; Pelkey et al., 2006). Nevertheless, at present, very little is known about either the physiology or pharmacology of collateral inputs from either mossy fibers or CA3 p yramidal cells to hilar mossy cells. This represents a notable gap in the literature as the effect of these inputs on mossy cell function is likely to have profound consequences for both normal and abnormal hippocampal physiology. One of the first studie s to directly address these questions has recently demonstrated presynaptic plasticity of isolated mossy fiber inputs to mossy cells (Lysetskiy et al., 2005). The current study attempts to further efforts in this area by specifically examining both GABAer gic and cholinergic mechanisms for presynaptic modulation of excitatory inputs to hilar mossy cells. Our results indicate a prominent role for GABA B receptor mediated inhibition of both MF and non MF inputs to hilar mossy cells and also demonstrate both d irect (non MF) and indirect (MF) mechanisms for cholinergic inhibition. We also provide the first clear demonstration that hilar mossy cells express high affinity GABA A receptors capable of responding to changes in ambient GABA and present evidence that i n this system, cholinergic changes in ambient GABA are tightly linked to action potential dependent inhibitory neurotransmission. Results Presynaptic GABA B Receptors are Expressed on Mossy Fiber Inputs to Mossy Cells Miniature excitatory postsynaptic currents (mEPSCS) were recorded in the presence of 1 TTX from hilar mossy cells voltage clamped at 70 mV with a CsCH 3 SO 3 based internal solution. Under these conditions, bath application of baclofen caused a signific ant reduction in mEPSC frequency (to 65.9 4.64% of baseline, n=9, p<0.0001, Fig. 4 1C) without affecting


72 mEPSC amplitude (99.6 4.26% of baseline, n=9, p>0.05, Fig. 4 1D). This effect of baclofen was completely abolished by pretreatment with CGP 52432 (frequency: 99.2 12.07% p>0.05, Fig. 4 1C; amplitude 90.1 6.11%, p>0.05, Fig. 4 1D; n=8). Interestingly, baclofen induced inhibition of mEPSC frequency was absent when the experiment was repeated in ACSF that contained 2.5 mM Mg 2+ 1 mM Ca 2+ and 200 2+ conditions designed to further reduce the probability of observing calcium dependent mEPSCs (frequency: 99.3 12.37%, p>0.05; amplitude: 100.3 4.95%, p>0.05; n=6, data not shown). These findings clearly suggested that activation of presynaptic GABA B receptors can reduce transmission from excitatory afferents to hilar mossy cells, but did little to distinguish which specific glutamatergic inputs were sensitive. In order to address this question more directly we used minimal stimulation techniqu es, coupled with bath or local application of DCG IV to isolate mossy fiber inputs to hilar mossy cells (see Chapter 2). We found that bath application of baclofen reduced the amplitude of EPSCs evoked by minimal stimulation of identified mossy fiber inpu ts to 9.9 1.37% of baseline (n=14, p < 0.0001, Fig. 4 2B). Consistent with a presynaptic site of action, baclofen also increased the failure rate of minimally evoked EPSCs (meEPSCs, see Chapter 2) from 11.3 3.02% to 75.9 4.97%, and increased the coe fficient of variation from 0.79 0.09 to 1.36 0.11 (n=14, p < 0.001 in both cases, data not shown). In order to further reinforce the conclusion that presynaptic GABA B receptors were responsible for baclofen mediated inhibition of mossy fiber dependen t meEPSCS, we also demonstrated that identical application of baclofen had no significant effect on mossy cell responses to local application of exogenous glutamate (50 200 96.4 9.7% of baseline, n=10, summarized in Fig. 4 2B).


73 Presynaptic GABA B R eceptors on Mossy Fiber Inputs to Mossy Cells are Both Sensitive to Ambient GABA and Tonically Active Under Control Conditions We found that the amplitude of mossy fiber mediated meEPSCs recorded from hilar mossy cells in the presence of 50 PTX was redu ced by bath application of the GABA uptake inhibitor NO 711 to 40.85 11.49% of baseline levels (n=6, p<0.03; Fig. 4 3B). Further, in 7 out of 9 separate experiments, we found that bath application of 10 CGP 52432, also in the presence of 50 PTX, i ncreased the amplitude of mossy fiber mediated meEPSCs to 254.7 33.5% of baseline (n=7, p<0.005, Fig. 4 4B). In both cases, apparent GABA B mediated modulation of meEPSC amplitude appeared to be presynaptic in nature. Specifically, concurrent with NO 71 1 mediated inhibition of meEPSCs we observed a clear increase in failure rate (from 8.5 4.4% to 22.9 6.43%, n=6, p=0.03, Fig. 4 3C) and a similar increase in coefficient of variation (from 0.74 0.08 to 1.04 0.11, n=6, p<0.01, Fig. 4 3D). Conversel y, CGP 52432 mediated increases in meEPSC amplitude were accompanied by a clear decrease in both failure rate (from 33.3 8.39% to 7.0 3.1%, n=7, p<0.01, Fig. 4 4C) and CV (from 0.99 0.09 to 0.64 0.06, n=7, p< 0.01, Fig. 4 4D). Note that in 2 of 9 cases bath application of CGP produced much milder increases in amplitude (to 110 2.90% of baseline) without consistent changes in failure rate and CV. Bath Application of Muscarinic Agonists Produces GABA B Receptor Mediated Inhibition of Mossy Fiber Inputs to Mossy Cells We next tested the hypothesis that putative presynaptic muscarinic receptors on mossy fiber inputs to mossy cells could provide an additional mechanism for presynaptic inhibition of mossy fiber tra nsmission. Consistent with that hypothesis, we initially found that in 11 of 12 cells tested, bath application of muscarinic agonists (e.g. 3 CCh or 5 muscarine) reduced mossy fiber mediated meEPSCs recorded from hilar mossy cells to 52.0 5.69% of baseline (n=11, p<0.01, data not shown, 1 cell showed no significant change). This effect was


74 accompanied by a significant increase in the coefficient of variation from 0.52 0.06 to 0.63 0.07 (n=11, p=0.05, data not shown). To provide further eviden ce that the observed muscarinic responses to locally applied exogenous glutamate (100.1 6.74% of baseline, n=9, p>0.5, data not shown). However, for 6 cells in w hich application of a muscarinic agonist reduced the amplitude of mossy fiber mediated meEPSCs to 55.9 6.81%, subsequent application of CGP 52432 completely reversed the effect (e.g. to 108 11.6% of baseline, n=6, p<0.01, Fig. 4 5B). This result is mo re consistent with an indirect mechanism of muscarinic inhibition (i.e. one that does not involve activation of muscarinic acetylcholine receptors ( mAChRs ) directly on the terminal); however because previous experiments had indicated that mossy fiber input s to hilar mossy cells are subject to tonic inhibition by ambient GABA, we also tested the ability of muscarinic agonists to inhibit mossy fiber transmission to mossy cells in slices pretreated with CGP 52432. Under those conditions, no muscarinic inhibit ion of mossy fiber transmission was observed and, in fact, a slight but statistically insignificant increase in meEPSC amplitudes was noted (e.g. to 119.9 21.9%, n=8, p=0.39, data not shown). Bath Application of Muscarinic Agonists Produces Increases in Ambient GABA that are Strongly Correlated with Increases in Spontaneous IPSCs. The data described above clearly implicated presynaptic GABA B receptors on mossy fiber terminals to mossy cells in muscarinic inhibition of glutamatergic transmission. These ob servations raised two interesting questions. First, can hypothetical changes in ambient GABA produced by bath application of muscarinic agonists be detected by additional means that are not dependent on mossy fibers? Second, if bath application of muscar inic agonists does indeed increase levels of ambient GABA, what is the mechanism? We were able to address both of these questions simultaneously using a recently developed analytical technique that allows


75 independent quantification of both tonic and phasi c GABA A receptor mediated currents during a period of high spontaneous activity (see Chapter 2). In brief, mossy cells were voltage clamped at 70 mV using an internal solution that contained ~60 mM chloride, ionotropic glutamate receptors were blocked by bath application of NBQX (10 ) and APV (40 ) and spontaneous IPSCs (sIPSCs) mediated by synaptic activation of postsynaptic GABA A receptors were observed as inward currents. After a 10 minute baseline period, 3 CCh was bath applied, exactly as in previous experiments with meEPSCs (see Fig. 4 6A for sample recording). In 6 of 10 cells tested, conventional event analysis showed clear and sustained increases in sIPSC frequency, and in some cases, clear increases in sIPSC amplitude as well (data not s hown). 1.68 pA / 5 s interval (n=6, p=0.05, Fig. 4 6C). This represented an average increase in phasic current to 767.4 280.9% of baseline. Importantly, the CCh ind uced changes in phasic current were also average increase in tonic current of 23.2 4.61 pA (n=6, p<0.005, Fig. 4 6C). Several lines of evidence suggest that t he changes in tonic current observed in this manner are directly caused by CCh mediated action potential dependent increases in ambient GABA, and subsequent activation of high affinity GABA A receptors expressed by hilar mossy cells. For example, in these same six cells we found that subsequent application of 50 PTX restored tonic current to 97.6 3.14% of baseline levels (Fig. 4 6C), and often reduced phasic current to below baseline levels (e.g. to 53.7 23.5% of baseline, n=6, p=0.11, Fig. 4 6C). F urther, an analysis of mean change in both phasic and tonic currents (expressed in pA / interval) across all three experimental conditions revealed an extremely strong correlation between these currents (R 2 =0.996, slope=5.48, p=0.002, Fig. 4 6D). Consiste nt with that result, we also found that in


76 the 4 of 10 cells that failed to show sustained increases is sIPSC frequency or phasic current following bath application of CCh, no significant effect on tonic current was observed (103 5.31% of baseline, p=0.5 3). In order to explicitly eliminate possible contributions of postsynaptic mAChRs to observed CCh mediated increases in tonic current, we also demonstrated that bath application of CCh failed to change holding current measured by conventional means unde r experimental conditions designed to eliminate most forms of glutamatergic and GABAergic transmission (e.g. in the ACSF that contained low calcium and high Mg 2+ as described i n Chapter 2). Specifically, I hold following bath application of CCh under those conditions was 100 0.75% of baseline (n=3, p=0.71, data not shown). Finally, repeating this experiment in the absence of PTX also failed to restore CCh mediated effects on holding current in four cells tested suggesting action potential independent mechanisms likely do not contribute substantially to CCh mediated changes in ambient GABA (data not shown). Non M ossy F iber M ediated E xcitatory I nputs to M ossy C ells are D irectly I nhibited by B oth GABA B R eceptors and M uscarinic A cetylcholine R eceptors The nature of the experiments presented in Figs. 4 2 to 4 5 is such that many runs were completed on isolated inputs that subsequently did not meet the criteria described in Chapter 2 to be classified as mossy fibers. It is thus potentially informative to compare the results of these putative non MF inputs to those described above. Such analysis of these parallel experiments suggested several similarities and one prominent difference Specifically, like MF inputs to mossy cells, putative non MF inputs were strongly inhibited by baclofen (to 18.2 3.4% of baseline, n=7, p<0.01, data not shown), and in 2 of 4 cases, showed clear increases in amplitude following bath application of CGP 52432 (e.g. to 379 72.3% of baseline, data not shown, 2


77 cells failed to show robust increases). Further, in 23 of 26 cells, bath application of muscarinic agonists reduced meEPSC amplitude to 47.0 4.68% of baseline. However, strikingly unlike MF inp uts, putative non MF inputs to mossy cells failed to demonstrate CGP 52432 mediated reversal of muscarinic inhibition. Specifically, we tested the ability of CGP 52432 to reverse inhibition produced by muscarinic agonists in 9 cells where meEPSCs failed t o meet criteria to be categorized as mossy fibers. In those 9 cells, meEPSCs amplitude following application of muscarinic agonists was 40.6 7.55% of baseline, and this value remained virtually unchanged (at 40.9 10.8% of baseline, data not shown) fol lowing bath application of CGP 52432. One potential criticism of this analysis is that putative non MF inputs in these cases often still showed significant sensitivity to DCG IV. For example, meEPSCs in the 9 cells described above that lacked CGP 52432 mediated reversal of muscarinic inhibition were sill reduced to 47.8 5.9% of baseline following bath application DCG IV. Perforant path inputs to the dentate gyrus have been reported to have intermediate sensitivity to DCG IV (Kilbride et al., 1998; Die trich et al., 2002; Kew et al., 2002), however this could also simply indicate that some of our minimally evoked responses had mixed MF and non MF origin. Therefore, we also tested the ability of 3 10 CCh to inhibit meEPSC amplitudes recorded from moss y cells in slices pre treated with both DCG IV and CGP 52432. Again, in sharp contrast to the results reported above for MF inputs, we found that robust inhibition of non MF inputs was still apparent under those conditions. Specifically, bath application of 3 10 CCh reduced the amplitude of non MF mediated meEPSCs to 55.4 8.49% of baseline (n=12, p<0.01, Fig. 4 7B). This effect was also accompanied by a significant increase in failure rate (from 18.8 4.25% to 38.8 7.32%, n=12, p<0.01) and an ave rage increase in CV (from 0.88 0.06 to 1.07 0.10, n=12, p=0.12). Finally, we also noted that bath application of baclofen reduced the amplitude of non MF


78 mediated EPSCs in slices pretreated with DCG IV to 30.8 4.1% of baseline (Fig. 4 7B), also conc urrent with a significant increase in both failure rate (from 19.0 7.26% to 67.0 5.71%) and CV (from 0.7 to 1.1, n=7, p<0.01 in both cases). Collectively these results strongly suggest that the majority of all glutamatergic inputs to mossy cells expre ss presynaptic GABA B receptors that are subject to tonic inhibition by ambient GABA, but that only non MF inputs also express presynaptic mAChRs capable of directly inhibiting glutamatergic transmission at the terminal. Origin of N on MF I nputs In addition to MF inputs from dentate granule cells, existing data suggest that mossy cells may receive glutamatergic input from both the perforant path and from CA3 pyramidal cells (Kohler, 1985; Ishizuka et al., 1990; Muller and Misgeld, 1991; Scharfman, 1991, 1994a ). Our experiments with minimally evoked EPSCs were designed, first and foremost, to isolate MF mediated inputs to mossy cells. As such, they are not particularly useful for distinguishing between various potential sources of non MF responses. In order to address this question more directly, we performed a final series of experiments designed to selectively activate excitatory inputs to hilar mossy cells originating either from the CA3 pyramidal cell layer or from the perforant path. Putative inputs fr om CA3 pyramidal cells to hilar mossy cells were isolated by stimulating the CA3 pyramidal cell layer with a bipolar stimulator (0.1 ms duration, generally < 100 A) in the presence of both DCG IV and CGP 52432. In 8 of 10 cases CCh produced an atropine s ensitive reduction in EPSC amplitude (specifically to 56.7 7.11 % of baseline in the presence of CCh, p < 0.001, Fig. 4 7C, recovered to 71.8 9.48% of baseline, p=0.01, within 15 min of bath application of atropine, 2 cells exhibited atropine insensiti ve changes in EPSC amplitude). These data suggest that CA3 inputs to hilar mossy cells, like our non MF mediated minimally evoked EPSCs, are directly inhibited by activation of mAChRs. Nevertheless, it is important to


79 note that unlike the minimally evoke d experiments, these experiments were done in the absence of PTX in order to reduce bursting in the CA3 pyramidal cell layer. Therefore we also demonstrated that bath application of NO 711 had no effect on EPSCs generated in hilar mossy cells under identi cal conditions (106 16.8% of baseline, n=6, p=0.72). That experiment largely eliminated the possibility that the CCh mediated inhibition of CA3 evoked EPSCs was produced by increases in ambient GABA and subsequent activation of presynaptic GABA A recepto rs. In sharp contrast, we found that stimulation of the middle to outer perforant path (also in the presence of DCG IV and CGP 52432) failed to produce detectable EPSCs in hilar mossy cells in ~75% of attempts, even when stimulus intensity surpassed 200 A This is not likely to indicate that perforant path inputs to mossy cells have high sensitivity to DCG IV, first because previous work has suggested perforant path inputs have intermediate sensitivity to DCG IV, and second because identical stimulation f requently produced detectable EPSCs in dentate granule cells (data not shown). Further, even when small EPSCs were detected in hilar mossy cells subsequent to stimulation of the perforant path, only minimal sensitivity to CCh was observed (10.4 0.57%, n =3, p=0.003, data not shown). Cumulatively, these experiments suggest that the minimally evoked EPSCs of non MF origin described earlier are more likely to represent afferent input from CA3 pyramidal cells than from the perforant path. Discussion This stu dy represents the first detailed effort to examine the ability of both GABAergic and cholinergic compounds to modulate glutamatergic transmission to hilar mossy cells. Our results demonstrate that both mossy fiber and non MF glutamatergic inputs to mossy cells express presynaptic GABA B receptors at their terminals that are capable of directly inhibiting glutamate release. These receptors are tonically activated by ambient GABA in in vitro preparations, and can be further activated by blockade of GABA tran sporters. We also report that bath application


80 of muscarinic agonists directly inhibits release of glutamate from non MF inputs while indirectly inhibiting mossy fiber inputs by driving action potential dependent increases in ambient GABA. Finally, we de monstrate for the first time that hilar mossy cells express high affinity GABA A receptors that produce tonic GABAergic currents that are sensitive to changes in the concentration of ambient GABA. GABA B M ediated I nhibition of M iniature EPSCs We found that bath application of baclofen produces a CGP 52432 sensitive reduction in the frequency of mEPSCs recorded from hilar mossy cells without changing their amplitude, strongly suggesting a presynaptic site of action. While this represents the first report of GABA B mediated inhibition of excitatory inputs to hilar mossy cells, presynaptic inhibition mediated by GABA B receptors is quite common in the CNS (for review, see Bowery, 1993). At various glutamatergic terminals, both calcium dependent (for review, see Misgeld et al., 1995; Dittman and Regehr, 1997; Takahashi et al., 1998; Chen and Regehr, 2003; Sakaba and Neher, 2003; Ishikawa et al., 2005), and calcium independent (Jarolimek and Misgeld, 1992; Scanziani et al., 1992; Capogna et al., 1996; Dittman and Regehr, 1996; Sakaba and Neher, 2003; Kolaj et al., 2004) mechanisms of GABA B mediated inhibition have been described. In the present study, GABA B mediated inhibition of mEPSCs was largely eliminated in ACSF containing low Ca 2+ high Mg 2+ and 200 Cd 2+ While this does not prove GABA B receptors at these terminals are directly coupled to calcium channels, it does suggest that GABA B R activation preferentially inhibits a calcium dependent release process at these synapses. In that regard, our conclusions are similar to those reported by Lei and McBain (2003) describing GABA B mediated inhibition of glutamatergic inputs to stratum radiatum interneurons in area CA3, although in their study, additional KCl was needed to evoke calcium dependent and baclofen se nsitive mEPSCs. By contrast, Scanziani et al. has reported baclofen mediated inhibition of Cd 2+ insensitive mEPSCs


81 in CA3 pyramidal cells (Scanziani et al., 1992). Overall, these findings reinforce the conclusion that the mechanism of GABA B mediated inhi bition can vary substantially from synapse to synapse, and may even differ, based on postsynaptic target, in collateral synapses from the same axon. Another interesting aspect of our results in this area is the implication that mEPSCs recorded from hilar mossy cells in the presence of TTX in normal ACSF retain significant calcium dependence. While just the opposite is often assumed to be the case, there is growing precedent for the idea of calcium dependent exocytosis in the presence of TTX at other syna pses. For example, perfusion of calcium free external containing BAPTA AM has recently been shown to eliminate CB1 mediated inhibition of mIPSCs in cerebellar Purkinje cells (Yamasaki et al., 2006), and to directly reduce mIPSC frequency in a subset of hi lar mossy cells (Hofmann et al., 2006). Consistent with these findings, others have noted a role for presynaptic calcium stores in promoting spontaneous calcium transients associated with large amplitude mIPSCs (Llano et al., 2000). Direct and I ndirect I nhibition of M inimally E voked EPSCs. The minimal stimulation techniques we used to isolate mossy fiber vs. non mossy fiber inputs to hilar mossy cells are very similar to those employed by other labs, and are based on well recognized features of mossy fib er transmission (Henze et al., 2000; Toth et al., 2000; Walker et al., 2001; Cossart et al., 2002; Lysetskiy et al., 2005). Our experiments with minimally evoked EPSCs clearly indicated that MF inputs to mossy cells express presynaptic GABA B receptors, ar e subject to tonic inhibition by ambient GABA, lack presynaptic muscarinic receptors, and yet can be indirectly inhibited by CCh induced and action potential dependent increases in ambient GABA. This result complements a very recent report of CCh induced and yet GABA mediated effects on mossy cell gain (Kerr and Capogna, 2007), and it is also


82 generally consistent with previous work on MF inputs to CA3 pyramidal cells. For example MF inputs to CA3 pyramidal cells also express presynaptic GABA B receptors, a nd are subject to tonic inhibition by ambient GABA that is mediated by both GABA B and GABA A receptors (Hirata et al., 1992; Vogt and Nicoll, 1999; Vogt and Regehr, 2001; Ruiz et al., 2003). Similarly, although an early study of cholinergic modulation of M F inputs to CA3 pyramidal cells concluded that they were subject to direct inhibition by presynaptic mAChRs (Williams and Johnston, 1990), a more recent study revealed an indirect mechanism dependent on GABA B receptors that is consistent with our findings in the hilus (Vogt and Regehr, 2001). One aspect of our study not generally available in work on other MF terminal zones is the implication that levels of ambient GABA around MF terminals in the hilus are tightly coupled to action potential dependent rele ase of GABA. This conclusion is consistent with studies in developing and mature cerebellum that showed a clear link between sIPSCs and tonic currents mediated by ambient GABA (Kaneda et al., 1995; Brickley et al., 1996; Wall and Usowicz, 1997; Carta et a l., 2004). It is also consistent with cell culture studies of both cerebellar and hippocampal neurons that show TTX mediated reductions in tonic GABAergic currents (Leao et al., 2000; Petrini et al., 2004), and with in vitro work in hippocampus directly i mplicating kainic acid ( KA ) induced increases in sIPSCs with GABA A mediated changes in holding current (Frerking et al., 1999). Nevertheless, it is clear that action potential independent mechanisms for regulating ambient GABA have also been implicated in some studies (Wall and Usowicz, 1997; Rossi et al., 2003). With respect to non MF inputs to hilar mossy cells, our results mirror previous studies on the terminal zones for CA3 pyramidal cell axons in indicating direct presynaptic inhibition mediated by bo th presynaptic GABA B receptors and mAChRs (Hounsgaard, 1978; Vogt and Regehr, 2001; De Sevilla et al., 2002; Fernandez de Sevilla and Buno, 2003; Lei and McBain,


83 2003). This raises the interesting question of whether the majority of our non MF mediated in puts in fact represent collateral projections from CA3 pyramidal cells to hilar mossy cells. On the one hand, event latency analysis was consistent with that hypothesis in indicating that the non MF inputs we observed, like previously identified collatera l inputs from CA3 (Amaral, 1978; Schwartzkroin et al., 1990; Frotscher et al., 1991; Scharfman, 1994b), are likely to synapse on more distal dendrites of hilar mossy cells than MF inputs. On the other hand, our experiments with meEPSCs were not directly d esigned to distinguish putative CA3 inputs from other types of non MF mediated responses. Therefore, in order to address that question more directly, we examined responses in mossy cells that were evoked by direct stimulation (with a bipolar stimulator) o f the CA3 pyramidal cell layer and the perforant path. Our results indicated that, like our minimally evoked non MF mediated population, EPSCs evoked in mossy cells by direct stimulation of CA3 are inhibited by bath application of CCh, even in the presenc e of DCG IV and CGP 52432. Although these experiments were done in the absence of PTX, increases in ambient GABA mediated by NO 711 failed to produce comparable inhibition of these CA3 evoked EPSCs. In sharp contrast, most mossy cells failed to show dete ctable responses to stimulation of the perforant path under identical conditions, although granule cells were still responsive. Further, those mossy cells that did exhibit small EPSCs in response to perforant path stimulation lacked robust inhibition by C Ch. Cumulatively these data suggest that our non MF mediated and minimally evoked EPSCs are indeed more likely mediated by collateral inputs from CA3 than from the perforant path. Nevertheless it must be noted that we cannot as yet explicitly rule out a contribution of mossy cell inputs to other mossy cells, although previous work has suggested that granule cells and molecular layer interneurons are likely the primary targets of mossy cell axons (Wenzel et al., 1997).


84 A final interesting question with reg ard to our non MF mediated minimally evoked EPSCs is why they lacked any evidence of indirect inhibition mediated by GABA B receptor activation. Given our results with mossy fibers, and the clear expression of GABA B receptors on non MF terminals, at least a partial reversal of mAChR mediated inhibition of non MFs might have been expected. Since changes in ambient GABA responsible for indirect inhibition of MF inputs are highly action potential dependent, we expect that spatial factors most likely account f or this apparent discrepancy. If that were the case, we would predict that synapses from mAChR expressing interneurons are likely to terminate on more proximal dendrites of hilar mossy cells, where they would be closer to MF than to non MF inputs. Nevert heless, based on present data, we cannot explicitly rule out the possibility that presynaptic GABA B receptors on non MFs have lower sensitivity to changes in ambient GABA than those on MFs, or the possibility that GABA B and mAChR dependent inhibition of no n MFs share a common presynaptic signaling pathway and thus can fully or partially occlude one another. In summary, our results have examined several specific mechanisms that modulate glutamatergic transmission to hilar mossy cells, and have identified c lear differences between two primary types of excitatory afferents. We believe that further studies into specific mechanisms that regulate mossy cell excitability will be necessary to develop a more complete understanding of hippocampal function and dysfu nction ranging all the way from memory consolidation to epileptogenesis.


85 Figure 4 1. Baclofen reduces mEPSC frequenc y without affecting amplitude. A) Cumulative increased the interevent interval of miniature EPSCs (baseline IEI: 179.5 4.31 ms; baclofen IEI: 328.2 10.92 ms; p<0.001, K S test) with only a negligible change in amplitude (baseline amplitude: 43.3 0.89 pA; baclofen amplitude: 41.4 1.22 pA, data not shown). Baclofen had no significant effect on mEPSC interevent interval (p=0.06, K S test) or amplitude (p=0.05, K S test, data not shown). Insets: consecutive 2 s intervals during either the baseline or after 6 min of baclofen wash in. C) Summary graph of baclofen effect on frequency. Baclofen significantly reduced mEPSC frequency (34.1 blocked this effect (0.8 12.07%, n=8, p>0. 05). D) Summary graph of baclofen effect on amplitude. Baclofen did not significantly affect the amplitude in either control conditions (0.4 4.26% reduction from baseline, n=9, p>0.05) or eline, n=8, p>0.05).


86 Figure 4 2. Baclofen blocks meEPSCs from isolated mossy fiber terminals via presynaptic GABA B receptors. s local IV completely blocked meEPSCs ( 1.8 0.7% of baseline). Following recovery from local application of DCG baclofen reduced meEPSC amplitude to 10.4 1.0% of baseline. During the baclofen washout, meEPSC amplitude showed considerable recovery (to 68.3 4. 9% of baseline). Insets: Individual traces from the relevant periods are shown in grey, with the average trace overlaid in black. B) Summary graph comparing baclofen effect on MF meEPSCs and exogenous glutamate spritz. Cells treated with DCG IV either b efore or after baclofen were pooled and showed significant inhibition by the GABA B agonist (90.1 1.37%, n=14, p<0.0001). By contrast, baclofen failed to inhibit exogenously applied glutamate (50


87 Figure 4 3. Presy naptic GABA B Rs on MF terminals respond to increases in ambient GABA. A) Single cell recording shows that bath application of 1 M NO 711 reversibly depressed average meEPSC amplitude, strongly suggesting that ambient GABA directly inhibits mossy fiber inp uts. Insets: Individual traces from the relevant periods are shown in grey, with the average trace overlaid in black. B) Summary graph of effects of NO 711 and DCG IV on meEPSCs. In 6 cells, NO 711 significantly inhibited meEPSC amplitude (to 40.9 11. 49% of baseline, p<0.03) while subsequent DCG IV treatment further reduced the amplitude (to 13.2 2.01% of baseline, p<0.01). C D) Summary plots for the effect of NO 711 on failure rate and CV. Connected points represent data from a single cell. In 6 cells, NO 711 significantly increased the average failure rate (from 8.5 4.4% to 22.9 6.43%, n=6, p=0.03) and coefficient of variation (from 0.74 0.08 to 1.04 0.11, n=6, p<0.01).


88 Figure 4 4. Mossy fiber inputs to mossy cells are subject to t onic inhibition by ambient GABA, mediated through GABA B re ceptors. DCG IV in a representative cell. CGP 52432 significantly increased the average meEPSC amplitude (to 211.4 14.29% of baseline) and decreased the fa ilure rate (0 failures in CGP 52432). Subsequent bath application of DCG IV reduced meEPSC amplitude by 95.6 1.9%, and significantly increased the failure rate (to 75.5%), suggesting the responses resulted from mossy fiber stimulation. Insets: Individu al traces from the relevant periods are shown in grey, with the average trace overlaid in black. B) Summary graph of CGP 52432 effect on meEPSCs. In 7 of 9 cells, CGP 52432 significantly increased meEPSC amplitude (to 254.7 33.5% of baseline, p<0.005) while DCG almost completely blocked all meEPSCs (10.8 2.0% of baseline, p<0.005). Error bars indicate SE. C D) Summary plots for the effect of CGP 52432 on failure rate and CV. Connected points represent data from a single cell. CGP 52432 significant ly reduced the average meEPSC failure rate (from 33.3 8.39% to 7.0 3.1%, n=7, p<0.01) and coefficient of variation (from 0.89 0.09 to 0.58 0.06, n=7, p<0.01).


89 Figure 4 5. Muscarinic inhibition of mossy fibers is reversed by a selective GABA B R antagonist. A) Single cell recording shows that bath application of 5 M muscarine decreased meEPSC amplitude. This effect was reversed by 10 M CGP 52432, implying that muscarinic receptors indirectly inhibit neurotransmitter release from mossy fibers. Insets: Individual traces from the relevant periods are shown in grey, baseline) while 10 CGP 52432 M reversed this effect (to 108 11.6% of baseline, p< 0.01). Bath application of 1 M DCG IV significantly reduced meEPSC amplitude (to 14.7 1.86% of baseline, p<0.05).


90 Figure 4 6. CCh induced increases in sIPSCs produce significant increas es in GABA A mediated tonic inhibition of hilar mossy cells A) Sample current trace from a mossy cell application, a sudden, dramatic increase in both sIPSC frequency and amp litude was observed. The effect on sIPSCs was accompanied by a negative shift in the holding clipped to allow for visualization of changes in both phasic and tonic currents. B) Effects of CCh and PTX on tonic (black dots) and phasic (white dots) currents in the significantly increased both tonic and phasic currents (by 15.81 pA and 3.3 pA, res effects (p<0.001 in both cases). C) Summary graph of effects of CCh and PTX on phasic (white bars) and tonic (black bars) currents. In 6 of 10 cells, CCh significantly increas ed the phasic current (by 4.29 1.68 pA from baseline, p=0.05) while PTX reversed this effect ( 0.65 0.27 pA from baseline, p>0.05). CCh also significantly increased the tonic current (by 23.2 4.61 pA from baseline, p<0.005) while PTX reversed this i ncrease ( 4.91 4.76 pA from baseline, p>0.05). Error bars indicate SE. D) Average changes in tonic current were plotted against changes in phasic current for each condition and fit by linear regression. The calculated slope is 5.48 with R 2 =0.996, p=0. 002, indicating a strong correlation between phasic and tonic currents across all three experimental conditions.


91 Figure 4 7. Non MF inputs to mossy cells express both GABA B Rs and mAChRs. A) In a representative cell pretreated with PTX (50 M), CGP 524 32 (10 M), and DCG IV bath application of atropine (5 M) rescued meEPSCs to near baseline levels. Insets: Individual traces from the relevant periods are shown in grey, with the average trace overlaid in black. B) Summary plot of independent effects of muscarinic and GABAB receptor activation observed in 12 cells. White bars: Bath application of CCh (3 10 M) reduced the average meEPSC amplitude (55.4 8.49% of baseline, n 79.4 13.4%, n=12, p=0.02). Grey bar: In a separate group of 7 cells pretreated with PTX (50 M) and DCG IV (1 M), bath application of baclofen (10 M) significantly reduced the average meEPSC amplitude (30.8 4.1% of baseline, p<0.0001). C) Summary plot indicating that CA3 evoked EPSCs recorded from hilar mossy cells in the presence of DCG IV and CGP 52432 were significantly reduced by bath application of 3 M CCh, yet no similar effect was produced by bath application of NO 711 in 6 separate experiments.


92 CHAPTER 5 METABOTROPIC GLUTAMA TE RECEPTOR MEDIATED ILTD OF GABAERGIC AFFERENTS TO HILAR M OSSY CELLS Introduction c strength is fundamental to its development as well as its capacity for creating and storing new memories. Presynaptic receptors are ideally placed to ex pressed at various synapses throughout the brain, either augmenting or diminishing transmitter exocytosis. Presynaptic GABA receptors, both ionotropic GABA A Rs and metabotropic GABA B Rs, are present at terminals throughout the brain and their activation generally reduces neurotransmission (Eccles et al., 1963; Axmacher et al., 2004; Vigot et al., 2006). Extracellular (ambient) GABA acting at these receptors can establish an inhibitory tone, affording those terminals a bi directional response to fluctuati ons in the ambient GABA concentration. In the hippocampus, mossy fiber axons of the granule cells express presynaptic GABA A Rs (Ruiz et al., 2003). We have previously shown that, in the hilus of the dentate, mossy fibers as well as non MF glutamatergic ax ons express presynaptic GABA B Rs but that only those receptors on the mossy fiber collaterals contacting mossy cells are responsive to endogenous ambient GABA. Further, the source of this ambient GABA appears to be a class of hilar interneurons that expres s somatic muscarinic acetylcholine receptors (Nahir et al., 2007). Activation of these receptors creates synchronous membrane oscillations and firing patters in these neurons (Reich et al., 2005; Hofmann et al., 2006). This hyperactivity leads to excess GABA spilling out of the synaptic cleft and an overall increase in the extracellular GABA concentration (Nahir et al., 2007). Endocannabinoids (eCBs) regulate neurotransmission in the CNS primarily via retrograde activation of the presynaptic type 1 canna binoid ( CB1 ) receptor (Katona et al., 1999;


93 Wilson et al., 2001; Kawamura et al., 2006). Depending on the duration of CB1R recruitment, eCBs can affect either a short term or long term synaptic depression. The most commonly observed short term presynapti c plasticity depolarization induced suppression of inhibition generally occurs due to a transient rise in postsynaptic intracellular calcium (eCB depol ) and subsequent eCB production and signaling (Wilson and Nicoll, 2001; Brenowitz and Regehr, 2003; Fo ldy et al., 2006; Hofmann et al., 2006). By contrast, CB1R mediated long term synaptic depression arises when eCBs are produced via activation of postsynaptic group I metabotropic glutamate receptors ( mGluRs ; eCB mGluR ) (Ohno Shosaku et al., 2002; Chevaley re and Castillo, 2003; Piomelli, 2003; Maejima et al., 2005). We have previously demonstrated that hilar mossy cells can produce eCB depol that results in depolarization induced suppression of inhibition ( DSI ) at a subset of their GABAergic afferents. In particular, eCBs preferentially inhibit those same hilar interneurons whose activity synchronizes (in the theta band) following activation of somatic mAChRs (Hofmann et al., 2006). Since hilar eCB signaling can modulate activity in those interneurons resp onsible for maintaining extracellular GABA, we questioned whether mossy cells could affect an enduring change in the ambient GABA concentration. Transient inhibition of CB1R + interneurons due to DSI, however, is clearly inadequate at altering the temporal concentration of extracellular GABA. Alternatively, eCB mGluR which has been shown to induce LTD at inhibitory terminals (iLTD) in areas such as the amygdala, cerebellum, striatum, and area CA1 of the hippocampus (Azad et al., 2004; Chevaleyre and Castil lo, 2004; Chevaleyre et al., 2006; Safo et al., 2006; Volk et al., 2006; Adermark and Lovinger, 2009), could modify ambient GABA levels by reducing recurrent GABA exocytosis. Therefore, we tested the ability of mossy cells to produce eCB mGluR and whether pharmacological activation of this pathway could lead to hilar iLTD. We


94 found that mossy cells could produce eCB mGluR primarily through activation of mGluR1, and prolonged eCB signaling led to selective iLTD of CB1R + hilar interneurons. In contrast with iLTD from other areas, we also uncovered an apparent endogenous activation of this iLTD system. Results Metabotropic Glutamate Receptor Activation Induces iLTD in the Hilus Identified hilar mossy cells were voltage clamped at 70 mV with a CsCH 3 SO 3 based internal containing ~60 mM chloride. Our lab has previously demonstrated that mossy cells are capable of producing eCBs that transiently inhibit a subset of GABAergic afferents (Hofmann et al., 2006). We questioned whether inhibitory inputs were a lso susceptible to other forms of retrograde, eCB mediated plasticity. The group I mGluR agonist DHPG has been shown to induce mGluR dependent LTD of inhibitory synapses (iLTD) via retrograde eCB signaling in area CA1 (Chevaleyre and Castillo, 2003). To test for a DHPG mediated eCB dependent LTD in the hilus, we used a concentric bipolar electrode to elicit evoked inhibitory postsynaptic currents ( eIPSCs ; 0.33 Hz) from hilar interneurons (see Methods). In 6 cells tested, treatment with DHPG induced a rob ust acute inhibition followed by a prolonged depression (DHPG: 64.6 5.57%, p<0.01; LTD: 73.9 3.33%, p<0.01; Fig. 5 1A, D). A recent report identified a dependence of eCB mGluR eCB production pathway (Edwards et a l., 2008). To prevent any inadvertent priming, we tested for the presence of DSI at the end of each experiment. Allowing at least 20 minutes for DHPG facilitate testing for DSI, after which we observed robust DSI in 5 cells tested (25.6 7.09% inhibition, p<0.05, Fig. 5 1B, D; we were unable to test for DSI in one cell).


95 Hilar iLTD Requires Endocannabinoid Signaling To determine the extent of eCB signaling in this mGluR dependent LTD, we used minimal stimulation to isolate individual DSI + and DSI GABAergic afferents. Since hilar mGluR LTD did not appear to require priming, we tested for DSI prior to treatment with DHPG in all subsequent experiments. In six cells, DSI minimally evoked IPSCs ( meIPSCs ) showed no response, either acute or long term, to DHPG treatment (DSI: 2.0 3.47% inhibition; DHPG: 114.8 9.42%; washout: 102.8 13.72%; p>0.1; Fig. 5 2A, C). Conversely, DSI + meIPSCs (39.3 5.84%, n=10, p <0.01; Fig. 5 2B, D) exhibited both significant acute and long term reductions following DHPG application (DHPG: 66.7 4.02%, p<0.01; washout: 81.5 7.09%, p<0.05; n=10; Fig 2B, D). Interestingly, in three of ten DSI + afferents, DHPG failed to produce a long term inhibition of meIPSCs (109.9 3.69%, p>0.1) while the remaining seven afferents showed LTD comparable to that observed with bulk stimulation (69.3 4.82%, p<0.01). The eCB system was further tested, first using the selective CB1R antagonist A M251. For the experiment presented in Fig. 5 20 minutes prior to DHPG application. Depolarization failed to induce any DSI in 6 cells tested in the presence of AM251 ( 5.1 3.90%, p<0.1; Fig. 5 3A, C). In line with this result, treatment with DHPG failed to produce any inhibition meIPSCs (DHPG: 105.8 11.18%, n=6; washout: 95.5 7.30%, n=4; p>0.5; Fig. 5 3B, C). Since we could not rule out the possibility of recruiting only CB1R afferents i n the presence of AM251, CB1R knockout mice and their wild type cousins were used to confirm the role of CB1R activation during DHPG mediated inhibition. Similar to treatment with AM251, CB1 / mice exhibited no DSI ( 2.9 4.08% inhibition, n=13, p=0.5; Fig. 5 4A, C) and only very slight DHPG mediated inhibition (DHPG: 90.1 3.44%, p<0.05, n=13; washout: 111.0 16.39%, p>0.5, n=7; Fig. 5 4B, D). CB1 +/+ mice, on the other hand, exhibited simi lar DSI of


96 meIPSCs (58.9 6.01% inhibition, p<0.01, n=5; Fig. 5 4A, C) compared to DSI + afferents isolated in rats and showed an acute DHPG inhibition followed by a slight but statistically insignificant long term depression (DHPG: 68.3 3.59%, p<0.01; w ashout: 91.5 6.10 %, p>0.1; n=5; Fig. 5 4B, D). Cumulatively, these results suggest that DHPG mediated iLTD in the hilus relies on retrograde eCB signaling acting at presynaptic CB1 receptors. Spontaneous Glutamate Release Leads to Endogenous Occlusion o f Hilar iLTD Both mGluR1 and mGluR5 have been shown to elicit eCB LTD in different brain regions; therefore, we sought to determine whether one or both of these group I metabotropic glutamate receptors were involved in our hilar iLTD. DHPG was applied in the presence of the selective significantly affected baseline meIPSC amplitudes (LY: 98.8 5.83%, n=5; MPEP: 102.5 5.47%, n=8; p>0.05; Fig. 5 5A, C), only LY prevented DHPG from inhibiting meIPSCs (DHPG+LY: 92.5 10.80%, p>0.05; DHPG+MPEP: 58.9 3.30%, p<0.01; Fig. 5 5). The levels of DSI, tested before application of either antagonist, did not differ significantly from control (pre LY: 40.2 8.10%; pre MPEP: 32.6 3.77%, p>0.05 for both compared to control; Fig. 5 5A, C). Recently, Heifets, et al. (2008) demonstrated that presynaptic activity regulates iLTD of CA1 interneurons in response to mGluR activation. Initially, we sought to determine whether hilar iLTD sh owed the same dependence on presynaptic calcium. A low concentration of TTX (10 nM) has been shown previously to significantly reduce presynaptic action potential propagation, thereby preventing sponta neous rises in terminal calcium while still allowing f or stimulation induced neurotransmission (Liu et al., 2004; Heifets et al., 2008). In our hands, however, 10 nM TTX failed to change the average sIPSC frequency (control: 19.6 2.11 Hz, n=9; TTX: 25.3 2.32 Hz, n=8; p>0.05) or amplitude (control: 68.5 13.67 pA, n=9; TTX: 68.3


97 7.34 pA, n=8; p>0.05) observed prior to isolation of meIPSCs (Fig. 5 6A). While the level of DSI only slightly increased (48.6 5.29% inhibition, n=8, p>0.05 compared to control; Fig. 5 6B, C), surprisingly DHPG caused a much more robust acute inhibition followed by a significantly enhanced iLTD (DHPG: 48.4 6.07%; LTD: 59.1 7.07%; n=8, p<0.05 in both cases compared to control; Fig. 5 6B, C). A slightly higher concentration of TTX (30 nM) significantly reduced both the freq uency and amplitude of spontaneous IPSCs ( sIPSCs ; freq: 9.6 2.27 Hz, p<0.01; amp: 28.4 2.95 pA; p<0.05; n=7; Fig. 5 6A) but also prohibited sustained activation of meIPSCs. To explain the striking difference of iLTD between control and 10 nM TTX withou t an obvious effect on sIPSCs, we hypothesized that basal spontaneous glutamate release might occlude the effect of DHPG via low level activation of postsynaptic mGluR1, thus reducing both acute and long term reductions in meIPSCs. If true, this would sug gest that low TTX might preferentially reduce spontaneous excitatory postsynaptic potentials ( sEPSCs ) We tested this hypothesis in two ways. First, we looked for a relationship between the level of sEPSCs observed at the beginning of either the control or 10 nM TTX experiments and the corresponding level of acute DHPG inhibition. In control conditions, we observed a significant inverse correlation between the average sEPSC amplitude and the level of DHPG mediated inhibition 0.779, adjuste d R 2 : 0.551, p=0.013; n=9; right panel, Fig. 5 7A). A similar link r: 0.673, adjusted R 2 : 0.375, p<0.05; n=9; left panel, Fig. 5 7A). Together, these result s indicated that the spontaneous activity observed at the beginning of an experiment should directly predict the efficacy of DHPG. Two sample cells are shown in Fig. 5 7B, with sample current traces showing the amount of spontaneous activity and the inset s showing the ability of


98 DHPG to inhibit the corresponding meIPSCs. As is clearly evident, DHPG is much less efficacious in the cell with higher sEPSC activity. By contrast, both of these correlations were disrupted in experiments with 10 nM TTX (amp: Pe 2 : 0.244, p>0.1; 2 : 0.221, p>0.1; n=8, Fig. 5 7C). The sample cells in Fig. 5 7D demonstrate that spontaneous activity no longer predicts the inhibitory effects of DHPG. Second, we examined directly the effect of 10 nM TTX on isolated sEPSCs. In 6 cells tested in the presence of the GABA A conditions), 10 nM TTX produced a significant reduction in sEPSC frequency (79.3 0.07%, p<0.05; Fig. 5 8B) without affecting the amplitude (98.5 0.44%; Fig. 5 8B). Discussion In the current study, we have demonstrated a form of synaptic plasticity that can significantly alter how hilar mossy cells integrate excitatory and inhibitory signals. Spe cifically, pharmacological activation of group I mGluRs induces both an acute and long lasting depression of GABAergic signaling to mossy cells in a CB1R dependent manner. Unlike iLTD observed in other brain areas, the hilar iLTD system appears to be endo genously activated at a low level by the large amount of spontaneous glutamate released from excitatory terminals contacting mossy cells. Endocannabinoid Production via mGluR Activation Endogenous cannabinoids are known to modulate synaptic transmission in response to either increases in postsynaptic calcium or activation of postsynaptic group I metabotropic glutamate receptors (Kreitzer and Regehr, 2001a; Wilson and Nicoll, 2001; Hofmann et al., 2006, 2008). While calcium sensitive eCB production occur s via the diacyl glycerol activation is less clear (Varma et al., 2001; for review, see Chevaleyre and Castillo, 2003; Freund


99 et al., 2003; Maejima et al., 2005; Edwards et a l., 2006). We have demonstrated that pharmacological activation of mGluRs on hilar mossy cells produces a CB1R dependent reduction in GABAergic transmission, similar to that observed in other brain areas, including the amygdala (Azad et al., 2004), cerebe llum (Galante and Diana, 2004), striatum (Maccarrone et al., 2008; Adermark and Lovinger, 2009), and CA1 region of the hippocampus (Chevaleyre and Castillo, 2003; Edwards et al., 2006). Although treating slices with LY367385 prevented DHPG from producing a significant reduction in meIPSC amplitude, a small inhibition was still observed, suggesting that while the mGluR1 receptor subtype likely predominates on mossy cells, mGluR5 may be present in small quantities or play a minor role in producing eCBs. R of the Ca 2+ dependent eCB production pathway, even though eCB mGluR appeared to bypass both the diacyl glycerol lipase and steps. In contrast to CA1, we o bserved no dependence of DHPG mediated eCB production on prior somatic depolarization, suggesting mossy cells do not require priming of the eCB depol pathway for eCB mGluR However, the high degree of spontaneous activity commonly observed in mossy cells in vitro may cause them to be intrinsically primed for eCB mGluR This represents an intriguing possibility particularly in light of the apparent occlusion observed under normal experimental conditions. If the level of spontaneous activity is indicative of the amount of basal glutamate released in vivo then eCBs may play an especially important role in sculpting hilar network activity. Role of Hilar iLTD LTD at inhibitory synapses due to eCB mGluR (iLTD) was initially described in the basolateral amygdala i n response to low frequency stimulation ( LFS ) (Marsicano et al., 2002) and later, using high frequency stimulation ( HFS ) in stratum radiatum of area CA1 in the hippocampus (Chevaleyre and Castillo, 2003). Such stimulation causes bolus glutamate release

PAGE 100

100 t hat activates dendritic mGluRs, mobilizes eCB production, and leads to prolonged retrograde activation of CB1Rs and alterations in presynaptic release machinery (Chevaleyre et al., 2007; Heifets et al., 2008). We report that hilar mossy cells are capable of inducing a similar eCB mGluR LTD of inhibitory, CB1R + synapses. The interneuron terminals susceptible to iLTD are the same that 1) undergo DSI and 2) modulate the level of hilar ambient GABA (Nahir et al., 2007). The second point is particularly notewo rthy since MF inputs to mossy cells are selectively sensitive to changes in ambient GABA while other, non MF glutamatergic terminals are not. Since eCB mGluR leads to a significant reduction in evoked GABA release, it is not unlikely that a parallel reduct ion in spontaneous GABA release occurs. If this is true, iLTD may very well be capable of modulating ambient GABA and thus MF transmission to mossy cells. Further studies will be necessary to determine if and to what extent iLTD impacts hilar glutamaterg ic signaling. Endogenous Occlusion of Hilar iLTD In their study of CA1 iLTD, Heifets et al. (2008) showed that stratum radiatum interneurons require a basal level of terminal Ca 2+ in order to induce iLTD. Although we initially used 10 nM TTX to address this same question in the hilus, the failure of TTX to reduce spontaneous GABA release prevented us from determining to what extent presynaptic activity affects hilar iLTD. However one important difference between the experimental conditions used in this study and those used in CA1 was the presence of the muscarinic agonist carbachol. Carbachol has often been used to enhance both interneuron firing and eCB depol (Kim et al., 2002; Ohno Shosaku et al., 2003; Reich et al., 2005); in the hilus, the large mAChR mediated increase in sIPSC frequency likely overwhelmed the ability of such a low concentration of TTX to prevent action potential propagation. While unable to reduce CCh enhanc ed sIPSCs, 10 nM TTX significantly increased both the acute DHPG inhibition and subsequent LTD. One possible explanation is that such a low concentration of TTX increased the terminal membrane resistance;

PAGE 101

101 such an increase then might cause a greater calciu m influx in response to an action potential invasion, leading to better transduction of the presynaptic CB1R signal and ultimately better iLTD. Although we cannot rule out this possibility, a more likely scenario (as supported by our current data) is tha t 10 nM TTX is more efficacious at reducing sEPSCs than sIPSCs and that, by reducing spontaneous glutamate release, treatment with TTX removed an endogenous system occluding our ability to induce DHPG mediated iLTD. In support of this hypothesis, we found that 10 nM TTX produced a slight but significant reduction in isolated sEPSCs, even in the presence of CCh. Furthermore, we found an inverse correlation between the level of sEPSC activity and subsequent DHPG mediated acute inhibition. When treated with low TTX, this correlation broke down, suggesting that TTX was hindering the ability of spontaneous glutamate to affect eCB mGluR A possible source for this occlusion is glutamate spillover from synaptic terminals that activates extrasynaptic mGluR1. In line with this, we found that the frequency of large (> 35 pA) sEPSCs was also inversely correlated with the level of acute DHPG inhibition. Such large amplitude events may indicate 1) that the release site is very close to the soma and 2) that the synapt ic event is likely multiquantal and includes multiple active zones releasing transmitter simultaneously. Both of these features are characteristic of mossy fiber transmission to mossy cells, suggesting that mossy fiber neurotransmission may be responsible for endogenous activation of mGluR1 and subsequent occlusion of eCB mGluR However, more detailed work will be necessary to determine the exact source of glutamate responsible for occluding iLTD. Strikingly, in most brain areas where eCB mGluR has been rep orted, LTD of excitatory synapses (eLTD) appears to be much more prevalent than iLTD (Watabe et al., 2002; Chevaleyre

PAGE 102

102 and Castillo, 2003; Azad et al., 2004; Safo and Regehr, 2005; Safo et al., 2006; Volk et al., 2006; Centonze et al., 2007; Sergeeva et al. 2007). The paucity of iLTD from other parts of the brain is particularly surprising given the ubiquitous expression of CB1R at GABAergic terminals throughout the brain (Tsou et al., 1998). The underlying reasons for these differences remain unclear. W hile some areas in which iLTD has been observed, such as area CA1 of the hippocampus, exhibit organized, overlapping glutamatergic and GABAergic innervation (Amaral and Witter, 1989), others like the amygdala do not (Sah et al., 2003). However, in CA1, eC B mGluR suggesting a relatively close proximity of excitatory and inhibitory terminals is necessary to overcome the limited diffusion distance of eCB and allow for endogenous, het erosynaptic iLTD. Initial attempts to induce eCB mGluR in mossy cells using exogenous stimulation rather than pharmacological manipulation were unsuccessful, though given the relatively disorganized nature of hilar innervation to these cells, a lack of stim ulator induced eCB mGluR is perhaps not surprising (Scharfman and Schwartzkroin, 1988; Frotscher et al., 1991; Scharfman, 1994a). In typical hippocampal slice preparations, mossy fiber axons extend only two to three hundred microns into the hilus before th ey are severed (Ruiz et al., 2003). Given the focal nature of minimal stimulation (and even bulk stimulation with a bipolar electrode) in conjunction with the lack of intact axons, it is likely very difficult to activate many glutamatergic afferents formi ng clustered synapses on proximal portions of the mossy cell dendritic tree. This does not mean, however, that coordinated activation of granule cells in vivo, for example, could not mobilize eCB production through concentrated glutamatergic transmission. Moreover the presence of an endogenous mechanism occluding iLTD seems to indicate that in fact this system exists and is in use naturally within the hilus. A better understanding of the spatial distribution of granule cells

PAGE 103

103 in the dentate blade whose a xon collaterals impinge on specific mossy cells will be necessary to determine whether concentrated granule cell excitation can induce iLTD. Figure 5 1. Activation of group I mGluRs inhibit s eIPSCs to hilar mossy cells. A) Application of the group I mG (0.2 Hz) to hilar mossy cells followed by a long term depression (iLTD). IPSCs were evoked using a bipolar stimulator placed between the mossy cell and the dentate gyrus. B) Average DSI time c ourse (n=5) of eIPSCs (0.33 Hz) to mossy cells. A brief (5 s) depolarization of hilar mossy cells produced robust DSI of eIPSCs that lasted ~30 s. C) Averaged traces from a sample experiment in which DHPG produced robust acute and long term depression an d subsequent depolarization led to transient DSI. D) Summary graph illustrating the inhibitory effects of DHPG on eIPSCs (left panel). The right panel shows individual cell responses to treatment with the mGluR agonist. In general, eIPSCs exhibited robu st inhibition with only a moderate washout. Numbers on bars are n values.

PAGE 104

104 Figure 5 2. DHPG selectively inhibits CB1R+ hilar interneurons and induces eCBmGluR LTD at these synapses to mossy cells. IPSCs were elicited using a monopolar electrode (meIPSCs) at 0.33 Hz for DSI, 0.5 Hz for DHPG application. A) Sample CB1R interneuron showing no DSI following a 5 s depolarization of the hilar mossy cell. + interneuron demonstrating r obust DSI and significant acute DHPG inhibition followed by iLTD. Insets for both A and B are averaged sweeps from following time periods: DSI sets (Base = 8 sweeps immediately preceding depol; DSI = 2 sweeps immediately succeeding depol; Rec = final 8 sw eeps); DHPG test (Base = 3 minutes preceding DHPG application; DHPG = final 3 minutes of DHPG application; Washout = final 3 minutes (17 20) at end of DHPG washout). C) Summary graph showing lack of DSI and subsequent inability of DHPG to inhibit meIPSCs (left panel). Right panel shows individual cell responses to treatment with DHPG. With one exception, cells showed no inhibition (and slight potentiation) in response to DHPG treatment. D) Summary graph showing co expression of DSI, acute DHPG inhibitio n, and iLTD (left panel). Right panel shows individual cell responses. All DSI + cells responded with acute DHPG inhibition while 7 of 10 showed robust iLTD as well.

PAGE 105

105 Figure 5 3. Blockade of CB1R by AM251 prevents eCB mGluR m ediated depression of meIPS Cs. Insets (A and B): average traces from relevant time periods for sample cell. C) Summary graph showing lack of DSI and DHPG induced inhibition (both acute and long term) in the presence of AM251.

PAGE 106

106 Figure 5 4. CB1R knockout mice are unaffected by DHPG while wild type cousins ma intain their DHPG sensitivity. A) Sample DSI sets from a CB1R / mou se (top) and CB1R +/+ mouse. Insets: average traces from corresponding sample cells for both DSI and DHPG treatments. B) From the same cells shown in panel A, DHPG has no effect on CB1R / meIPSCs while mGluR activation induces an acute inhibition of CB1R +/+ meIPSCs. C) As expected, CB1R / mice do not express DSI while meIPSCs from CB1R +/+ mice show significant inhibition following mossy cell depolarization. D) Summary graphs showing relative DHPG sensitivity. Top: DHPG produced a small but significant reduction in meIPSCs from CB1R / mice. Bottom: meIPSCs from CB1R +/+ mice exhibited strong short term and a small, though statistically insignificant sensitivity to mGluR activation.

PAGE 107

107 Figure 5 5. Metabotropic glutamate receptor type 1 (mGluR1) mediates eCB mGluR in hilar mossy cells in re sponse to treatment with DHPG. A) Treatment with the mGluR1 antagonist DSI but no response to DHPG in the presence of LY367385. B) Summary graph showing DSI but lack of DHPG effect. C) Unlike LY367385, the mGluR5 antagonist DSI as well as acute and long term inhibition following treatment with DHPG. D) Summary graph showing DSI and maintenance of inhibition with DHPG. iLTD in this case was present but not significant (p=0.12). Insets (A and C): average traces from relevant time periods including baseline prior to drug treatments.

PAGE 108

108 Figure 5 6. 10 n M TTX enhances iLTD without affecting sIPSCs. CCh, 10 nM TTX failed to change the sIPSC frequency (a slight increase was observed) or amplitude while 30 nM TTX produced a significant reduction in both (bottom summary graphs). T op 3 panels are sample current recordings from three separate voltage clamped ( 70 mV) mossy cells in the three different conditions. As is evident, 10 nM TTX is unable to reduce CCh enhanced sIPSCs. Although 30 nM did reduce sIPSC activity, it also prev ented the acquisition of stable, sustained meIPSCs. B) Sample cell showing robust DSI and enhanced acute DHPG inhibition and iLTD in the presence of 10 nM TTX. C) Top panels are average meIPSCs from sample cell in B taken from relevant time periods. Bot tom left summary graph shows slight increase of DSI efficacy in the presence of TTX. Bottom right graph shows statistically significant enhancement of both short and long term DHPG mediated inhibitions compared to control conditions.

PAGE 109

109 Figure 5 7. Spo ntaneous glutamate release occl udes DHPG mediated inhibition. A) Correlation plots demonstrating inverse relation between initial sEPSC frequency and amplitude (from beginning of each experiment) and corresponding acute inhibition of meIPSCs by DHPG. Ave rage sEPSC amplitude showed a very strong inverse correlation suggesting that more endogenous glutamate predicts less effective DHPG. B) Two sample cells showing relation between sEPSC activity and DHPG efficacy. DHPG is much more effective at inhibiting meIPSCs in quiet cell than in cell with much more afferent activity. C) Correlation plots (as in A) for experiments with 10 nM TTX. Spontaneous EPSC data were gathered prior to treatment with 10 nM TTX. In this case, low TTX disrupted the correlation b etween sEPSC activity (for both frequency and amplitude) and DHPG efficacy. D) Sample cells from TTX data set showing that differences in activity do not predict ability of DHPG to inhibit meIPSCs. Note: for the frequency DHPG correlation plots in panels A and C, sEPSCs were high pass filtered at 35 pA.

PAGE 110

110 Figure 5 8 10 nM TTX p referentially inhibits sEPSCs. A) Sample current traces from cell demonstrating reduction in sEPSC frequency by 10 nM TTX. TTX was applied in fort to mimic as closely as possible conditions under which TTX failed to inhibit sIPSCs. Right two panels are cumulative probability histograms for interevent interval and amplitude of the same sample cell. B) In 6 cells, 10 nM TTX significantly reduced (by 20%) sEPSC frequency without affecting average amplitude. Right two panels show effect of TTX on total number of events (per measurement period) and average amplitude for each cell.

PAGE 111

111 CHAPTER 6 CONCLUSION Summary and Discussion The hippoca mpus is an incredibly complex, important structure in the mammalian brain. Its unique anatomical architecture and laminar coordination of activity affords the hippocampus significant computational power, a tool necessary for its central role in memory pro cessing. Such a high level of activity, however, can easily lead to electrical instability, a feature that can result in epileptic seizures. In fact, the hippocampus has been determined as one of the primary foci for seizure generation in temporal lobe e pilepsy, one of the most common forms of the disease. In particular, mossy cells of the hilar region below the dentate appear to play a central role in the etiology of temporal lobe epilepsy ( TLE ) their loss and/or dysfunction may lead directly to exces sive electrical activity escaping the dentate region and spreading to the hippocampus and beyond. The dearth of research focused on general mossy cell physiology is surprising given their role in TLE and general susceptibility to excitotoxic death in conj unction with their ability to coordinate granule cell activity in distant regions of the dentate. Understanding the mechanisms that control mossy cell excitability may provide valuable new insights into seizure generation and offer new therapeutic targets for epilepsy treatment. Since presynaptic changes in neurotransmission can profoundly affect activity in the postsynaptic cell, the primary goal of this dissertation was to enhance the current state of knowledge regarding mechanisms regulating synaptic s ignaling from neurons innervating mossy cells. While many different systems exist to modulate neurotransmission from the presynaptic terminal, we were primarily interested in the role of presynaptic GABA receptors and retrograde endocannabinoid ( eCB ) sig naling. The initial study, presented in Chapter 3, demonstrated the ability of mossy cells themselves to regulate their afferent activity via the endocannabinoid

PAGE 112

112 signaling system. Retrograde eCB signaling has recently become a very popular area of study and although depolarization induced suppression of inhibition ( DSI ) has been described elsewhere in the hippocampus and brain as a whole, this research represents the first demonstration of eCB signaling in the hilar region of the dentate. Pioneering stu d ies involving eCB signaling and DSI often used muscarinic acetylcholine receptor ( mAChR ) agonists or endogenous acetylcholine release to enhance spontaneous activity in interneurons (Pitler and Alger, 1994; Wilson and Nicoll, 2001). Depolarization of int erneurons by muscarinic agents generally leads to a very robust and reproducible increase in spontaneous inhibitory postsynaptic current (s IPSC ) frequency, making DSI more readily observable. Unlike other areas of the hippocampus, activation of hilar musc arinic receptors only occasionally led to increased spontaneous activity. Of those instances when carbachol successfully enhanced interneuron firing, we observed two distinct effects: an increase in sIPSC frequency or an increase in both frequency and amp litude. We attributed a rise in frequency alone to augmented asynchronous release from interneurons while enhanced frequency and amplitude was likely due to synchronous GABAergic signaling. As expected, the asynchronous sIPSCs were generally smaller than synchronous events (although both were larger than sIPSCs in the absence of CCh). The reasons for a lack of synchrony in some cases (or no effect of muscarinic signaling at all) are unclear but one possibility is that the lack of effect is an artifact in herent in the experimental design. The nature of slice physiology discrete, disconnected segments of brain tissue probably imparts certain limitations on the phenomena observable by the experimenter. If true, such an attribute may explain the unrelia ble effects of muscarinic agonists on hilar circuitry. For instance, some GABAergic neurons contain the innate ability to, when driven by muscarinic receptors, synchronize oscillations in their membrane potentials in

PAGE 113

113 the absence of ionotropic glutamate re ceptor signaling (Reich et al., 2005). However, this synchronicity likely relies on a threshold level of network connectivity that may be compromised in the slice preparation. For those exceptions when robust synchronous release was observed, spontaneou s activity was coupled to the theta frequency band, one of a few important frequency ranges observed in all animals. In particular, theta is generally associated with two distinct behavioral states: free moving exploration and random eye movement (REM) sl eep (Henze and Buzsaki, 2007). Theta oscillations are common to the hippocampus as a whole but the fact that muscarinic receptor activation caused these oscillations to appear is especially interesting in light of the fact muscarinics are also capable of enhancing calcium dependent eCB production (and hence DSI). Though the mechanisms for this enhancement have yet to be worked out in the hilus, they are likely very similar (if not identical) to the well described cascade that occurs following mAChR activa tion in other brain areas, such as CA1 and the cerebellum (Kim et al., 2002; Ohno Shosaku et al., 2003). Regardless of how muscarinics enhance DSI, the fact that the same signaling molecule may play a dual role in the hilus could have profound implication s for both hilar and general hippocampal network activity. Hilar interneurons can be grouped based on a number of different features, in this case whether they express the type 1 cannabinoid ( CB1 ) receptor at their axon terminals. Immunohistochemical evidence from later studies directly supports our conclusion that only a subpopulation of hilar interneurons is CB1R + (Katona et al., 1999; Tsou et al., 1999). Along with this target specificity, endocannabinoid signaling is subject to tight spatial constraints (~20 imply that hi lar eCBs should generally affect only afferents to a single cell. While this was

PAGE 114

114 usually the case, a few instances clearly demonstrated the ability of eCBs released from one mossy cell to significantly inhibit GABAergic afferents impinging on a neighborin g mossy cell. While Chapter 3 deals exclusively with regulation of inhibitory GABAergic transmission via retrograde eCB signaling Chapter 4 focuses on control of glutamatergic signaling, in this case through modulation of ambient GABA. Although the ma jority of GABAergic signaling is associated with fast phasic inhibition, ambient GABA and tonic inhibition are now recognized as important tools for adjusting the average release rate of terminals over a long time. Accordingly, we searched for presynaptic GABA receptors at glutamatergic terminals impinging on hilar mossy cells. We found that, in the hilus, both mossy fiber axons and non mossy fibers (composed primarily of recurrent axons from CA3 pyramidal cells) express presynaptic metabotropic GABA B rec eptors whose activation efficiently reduces vesicular exocytosis from these terminals. Pharmacological blockade of GABA transporters artificially elevated the concentration of extracellular GABA, selectively reducing transmitter release from MF terminals. Interestingly, blockade of GABA B receptors in normal conditions (i.e. without artificially elevated ambient GABA) produced a robust increase in MF signaling to mossy cells. When considered together, these data suggest that 1) mossy fibers are selectivel y sensitive to increased extracellular GABA and 2) an ambient GABA tone is normally present in the hippocampus at a sufficient level to create a tonic inhibition of hilar mossy fibers (likely even greater in the intact brain than observed in our tissue sli ce preparation). The fact that the basal ambient GABA concentration is enough to partially inhibit mossy fiber exocytosis suggests that these terminals are capable of and primed for a bi directional change in neurotransmission mediated by changes in ambie nt GABA.

PAGE 115

115 The body of work presented in Chapter 4 is in fact the result of two initially independent projects coalescing through a common mechanism. While searching for presynaptic GABA B receptors, we also sought to confirm the presence of putative mAChR s on hilar glutamate terminals (Williams and Johnston, 1990; but see, Vogt and Regehr, 2001). Even as we verified the presence of muscarinics on non MF synapses to mossy cells, blockade of the GABA B receptor system demonstrated that inhibition of MFs by p utative mAChRs was actually the result of indirect inhibition via alteration of ambient GABA. These results imply that muscarinic signaling is capable of affecting glutamatergic neurotransmission in the hilus through very different mechanisms and likely d isparate time scales. Presynaptic mAChRs on non MF terminals may allow faster, more transient changes in glutamate release from these sites; indirect inhibition of MF synapses by way of ambient GABA, although slower, may be more enduring than direct presy naptic inhibition. Since muscarinic signaling clearly affected the ambient GABA concentration, how was this change achieved? And further, what is the primary source of extracellular GABA? Although other studies have addressed the second question, the a nswer is still a matter of debate. Since somatic mAChRs are capable of depolarizing hilar interneurons, we posited that overdrive of these interneurons led to a spillover of synaptically released GABA that bound GABA B receptors on nearby MFs. Using the m ossy cells as sensors for ambient GABA, we clearly saw increased synaptic release driven by muscarinic activation and observed a concomitant tonic current carried by extrasynaptic GABA A receptors. Further analysis of these data revealed a strong correlati on between the changes in phasic and tonic currents, indicating that GABA spillover was responsible for changing the level of ambient GABA and likely plays a significant role in setting the overall extracellular GABA concentration.

PAGE 116

116 the hilus is still unclear; one possibility is that the inhibitory tone provides a constitutively active regulatory mechanism for preventing over excitation of the mossy cells. Specifically, many of the spontaneous excitatory events recorded in mossy cel ls are thought to originate from the large mossy fiber boutons. If true, in the absence of an inhibitory tone mossy fiber terminals might release substantially more glutamate than normal, creating a standing depolarization in the mossy cells. With mossy cells already susceptible to excitotoxicity, such a depolarized state could put the hilar network in a precarious position and ultimately lower the seizure threshold for the hippocampus as a whole. An underlying theme in both Chapters 3 and 4 is the occurr ence of Ca 2+ dependent neurotransmission in the presence of tetrodotoxin (TTX), the voltage gated Na + channel blocker. TTX has long been employed as a way to isolate action potential independent release events (miniature IPSCs, mIPSCs). Miniature IPSCs h ave usually been assumed Ca 2+ independent; however, our results that both CB1Rs and GABA B Rs selectively reduced Ca 2+ dependent release clearly add to a growing body of evidence that suggests spontaneous Ca 2+ dependent exocytosis can occur without the neces sity of an action potential from the soma. The need for both calcium dependent and calcium independent spontaneous neurotransmission in the absence of somatic drive (i.e. action potentials) is unclear. One possibility is that constant contact in the form of AP independent release maintains synaptic stability without significant metabolic cost, required for AP generation and propagation, to the presynaptic cell. The importance of the work presented in Chapters 3 and 4 should not be undervalued. Both amb ient GABA and DSI represent two potent, independent mechanisms for regulating neurotransmission to mossy cells. Data and ideas in Chapter 5, built on the following hypothesis and observations, embody a first attempt to link these two independent systems t ogether with an

PAGE 117

117 eye towards demonstrating how mossy cells might regulate their own glutamatergic afferents. In designing this project, previous data from our lab exposed a biphasic effect of eCB signaling at glutamate terminals to mossy cells an initial depression of ne urotransmission (depolarization induced suppression of excitation, DSE; qualitatively very similar to DSI) followed by a more robust potentiation. Given that presynaptic CB1Rs inhibit neurotransmission, the initial decrease in glutamate s ignaling seemed likely to mimic DSI; the mechanism responsible for increasing glutamate release, on the other hand, was elusive at best. Another project from our lab, along with recent histological studies (Katona et al., 2006; Monory et al., 2006), demon strated the presence of CB1 receptors on glutamate terminals in CA3; the distribution of these receptors together with our own pharmacology and electrophysiology studies indicated that mossy fibers in stratum lucidum were devoid of CB1R expression while re current CA3 axons both expressed CB1R and were sensitive to short term eCB plasticity (DSE) (Hofmann et al., 2008). With these data in mind and because MFs and recurrent CA3 axon collaterals make up the bulk of glutamatergic inputs to hilar mossy cells, we set out to explain the potentiation of sEPSCs. One possible explanation is that an eCB mediated reduction of GABA release leads to a disinhibition of glutamate transmission, specifically from mossy fibers, played out as a potentiation of sEPSCs. Caref ul examination of the data presented in Chapters 3 and 4 provided motivation for this hypothesis and work presented in Chapter 5 represents a first step in testing it. As mentioned, recurrent axon collaterals from CA3 pyramidal cells express the CB1 recep tor at their terminals while mossy fibers do not. Thus, mossy fibers are more likely than CA3 axons to be responsible for the increased glutamate release in the presence of endocannabinoid signaling. Since MFs are selectively regulated by ambient GABA an d mossy cell based eCB signaling can

PAGE 118

118 reduce GABA transmission, endocannabinoids might be capable of affecting the level of ambient GABA, thereby altering mossy fiber neurotransmission. Clearly, the transient nature of DSI precludes it from significantly impacting the overall extracellular GABA concentration. Rather, a long term reduction of GABA release is probably required. Nevertheless, with an appropriate induction mechanism, endocannabinoids are able to generate long term depression ( LTD ) of exocyto sis in other brain regions (Chevaleyre and Castillo, 2003; Centonze et al., 2007; Sergeeva et al., 2007). Specifically, postsynaptic group I metabotropic glutamate receptors (mGluRs) recruit excitatory G proteins that stimulate a calcium independent (and thus prolonged) eCB production, leading to longer activation of CB1 receptors. This extended CB1R activity results in intraterminal changes that reduce vesicular release probability; at inhibitory terminals, this phenomenon is known as iLTD. While we hav e yet to show that an eCB mediated reduction in GABA release will cause a decrease in the ambient GABA concentration, this point is moot without the presence of iLTD at these terminals. Therefore, the experiments presented in Chapter 5 were motivated by t he question of whether CB1R + interneurons impinging onto mossy cells can express iLTD. High frequency stimulation of glutamate axons has been shown to induce iLTD elsewhere in both the hippocampus and brain (Chevaleyre and Castillo, 2003; Azad et al., 20 04; Chevaleyre and Castillo, 2004; Volk et al., 2006). However, the jumbled nature of hilar architecture, diffuse mossy cell dendritic tree, and sheer number of granule cells (each with a low probability of contacting a specific mossy cell) make stimulate d focal glutamate release a difficult prospect indeed. Yet pharmacological treatment with both agonist and antagonists revealed the presence of a postsynaptic group I mGluR (specifically mGluR1) that, upon activation, was capable of producing endocannabin oids. Further, only CB1R + interneurons responded to this activation by

PAGE 119

119 expressing both an acute and long term depression, as demonstrated both by pharmacological blockade of the eCB signaling system as well as genetic manipulation to remove the CB1 recept or from the brain entirely. Recent studies have delved deeper into the nature of iLTD and found, among other things, that the CB1 receptor produces iLTD through a calcium dependent signaling cascade (Chevaleyre et al., 2007; Heifets et al., 2008). A num ber of different methods have been employed to reduce presynaptic calcium, including lowering extracellular calcium, treating the tissue with a membrane permeant fast calcium chelator (to prevent calcium from activating intracellular pathways), and using l ow concentrations of TTX to partially block voltage gated sodium channels (Liu et al., 2004; Heifets et al., 2008). This last technique hinders action potential propagation leading to less calcium change in the presynaptic terminal. Our efforts to dampe n presynaptic calcium increases by using a low concentration of TTX provided some surprising results. In fact, rather than preventing iLTD induction (as reported elsewhere), low TTX enhanced hilar iLTD. That low TTX failed to inhibit iLTD at all is perha ps not entirely surprising; throughout the experiments presented in Chapter 5, the tissue was constantly perfused with carbachol, the same muscarinic agonist responsible for depolarizing and driving interneurons. Consequently, low TTX did not reduce the i nitial sIPSC frequency and only slightly reduced the overall level of presynaptic activity present during these experiments, but not to a level expected to significantly alter action potential propagation and subsequent presynaptic calcium influx. However the mechanism by which iLTD increased is unclear. One possibility is that by partially blocking voltage gated sodium channels, low TTX effectively increased the membrane resistance (R m ) of the presynaptic axon and terminal. This increase in R m might pr event some basal shunting, thus causing those action potentials that do

PAGE 120

120 propagate all the way from soma to synapse to be more effective at raising intraterminal calcium via voltage sensitive calcium channels. Alternatively, low TTX might produce enhanced iLTD by removing endogenous occlusion of this mechanism. Since these experiments were performed in the presence of ionotropic glutamate receptor antagonists, the effect of low TTX on glutamate exocytosis is difficult to assess. However, separate experim ents demonstrated a slight but significant reduction in sEPSC frequency. Furthermore, both the initial sEPSC average amplitude and the average frequency for large sEPSCs (i.e. < 35 pA) measured at the beginning of control experiments were inversely correl ated with the ability of DHPG to produce eCB mediated inhibition; treatment with low TTX, on the other hand, disrupted these correlations. Under normal conditions (i.e. without TTX), blockade of the ionotropic glutamate receptors might lead to a higher ba sal binding of the mGluRs. Treatment with low TTX may then free up more mGluR binding sites by reducing basal glutamate release, thus facilitating DHPG mediated eCB production and CB1R activation. In line with this hypothesis, DHPG produced significantly more inhibition, both acute and long term, in the presence of low TTX. Since G protein signaling cascades are amplified many times over, this change in acute inhibition could be sufficient to induce a much more robust iLTD. While we found it difficult t o generate endogenous iLTD via exogenous stimulation (rather than pharmacological manipulation), the fact that endogenously released glutamate appeared to occlude iLTD could very likely indicate that this mechanism is already employed by the hippocampus at a low background level. In area CA1, iLTD of stratum radiatum interneurons reduces postsynaptic inhibition, lowering the threshold for long term potentiation ( LTP ) induction at the Schaffer Collateral CA1 synapse (Chevaleyre and Castillo, 2004). Unli ke the traditional LTP expressed at that

PAGE 121

121 synapse, mossy fiber LTP is expressed presynaptically (although postsynaptic LTP at MF synapses has been reported in CA3). If hilar iLTD does play a role in reducing ambient GABA and thereby preferentially enhancin g neurotransmission from mossy fibers onto mossy cells, these disinhibited synapses may be more likely to undergo LTP. Since granule cells act as a relay for the majority of information into the hilus from the perforant path, strengthening of the mossy fi ber mossy cell synapses could alter how that information is integrated and eventually propagated throughout the hilar and dentate networks. Alternatively, too much excitatory activity might drive the endogenous production of eCB mGluR that, on top of a background eCB mGluR tone, could act as a seizure generator by increasing excitation of mossy cells. Perspectives and Future Studies Hilar circuitry results in sparse connectivity between granule cells and mossy cells. In general, while mossy cells receive many excitatory inputs, each mossy fiber only contacts a very few mossy cells during its traversal of the hilus. This may be particularly important since granule cells themselves very likely receive convergent innervation from many different entorhinal i nputs. While neighboring granule cells do not generally share postsynaptic mossy cell targets, the extensive axonal branching by perforant path projections in the molecular layer of the dentate may result in granule cells with shared postsynaptic targets receiving common presynaptic input. Consequently, novel information (presumably representing a new environmental experience) from the entorhinal cortex may diverge across the entire dentate gyrus but could ultimately converge again onto specific subpopula tions of mossy cells. With hilar interneurons oscillating in the theta band, a sudden convergence of excitatory input to certain mossy cells could lead to DSI or iLTD, resulting in a temporary or more permanent disruption of interneuron activity and there by allowing those mossy cells to coordinate general dentate activity based on the novel entorhinal signal. The ability of mossy cells to direct this harmonization

PAGE 122

122 might dramatically impact how well a novel experience is propagated through the hippocampus and ultimately encoded as a new memory. We currently lack the detailed knowledge of the dentate to credibly identify which granule cells target specific mossy cells. This deficit is particularly glaring considering we have yet to identify an endogenous in duction mechanism for iLTD. With the development of new electrophysiological tools, such as light activated ion channels (i.e. channelrhodopsins) (Zhang et al., 2006; Berndt et al., 2009) and voltage sensitive dyes (Kee et al., 2008), researchers may soon be able to visualize and control the activity of discrete neuronal populations. The exquisite control these techniques provide, in conjunction with the knowledge gained from studies like those presented in this dissertation, could eventually allow resear chers to ask and answer questions such as what effect directed, biologically relevant excitatory input has on overall network activity. Since phenomena such as DSI and iLTD, although often synapse specific, are viable systems for modulating synaptic activ ity utilized by mossy cells throughout the hilus, a particularly interesting question is what types of input signals to the dentate (from the entorhinal cortex) are responsible for activating these systems. Are those signals the same ones that represent n ew memories or are they distinct, separate control signals used solely for the regulation of hippocampal information throughput? An interesting complementary path of study is how hilar interneurons interact with each other. More specifically, do CB1R + i nterneurons contact other hilar interneurons and if so, are DSI and iLTD available at those terminals as well? From our studies with muscarinic agents, the CB1R + interneurons appear to be responsible for most of the theta oscillations we observed in vitro But very likely, both CB1R + and CB1R interneurons receive excitatory drive from the same sources (mossy fibers, pyramidal cells, perforant path, etc.). Could these interneurons

PAGE 123

123 compete with and inhibit each other, thereby altering hilar network activity along with the mossy cells? While DSI and iLTD appear more likely to work in a focused manner, modulating activity at a select group of synapses, the ambient GABA system seems better suited for more global control of network activity. However, the nature of ambient GABA in vivo is still unclear. Is inhibitory tone truly responsible for maintaining selective control over hilar mossy fibers or is the apparent functional spatial gradient more of an evolutionary novelty? Since epilepsy induces a hyperexcitable state in the hippocampus and both mossy fibers and mossy cells have been implicated as vital components to this change, comparison of the ambient GABA system and inhibitory tone in normal and epileptic tissue could provide valuable insights into both onset and treatment of the disease. If inhibitory tone really does dampen mossy fiber activity, we might expect that ambient GABA is reduced in epileptic tissue, that mossy fibe rs themselves respond differently to tonic activation of their presynaptic receptors, or that they no longer express the same presynaptic flora as compared to healthy tissue. Identifying these differences could ultimately help explain why, for instance, m Basic in vitro electrophysiology can only go so far towards answering these and other questions regarding the functional implications of synaptic signaling systems like endocannabinoids and ambient GABA. To more fully und erstand the brain, these same types of questions how signaling systems interact with each other must be applied to larger, more intact preparations. However, the foundation provided by such mechanistic studies as those presented here is pivotal to all owing us to ask those next questions.

PAGE 124

124 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. Journal of Neuroscience 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. Journal of Neuroscience 20:6907 6919. Adermark L, Lovinger DM (2009) Frequency dependent inversion of net striatal output by endocannabinoid dependent plasticity at different synaptic inputs. J Neurosci 29:1375 1380. Adermark L, Talani G, Lovinger DM (2009) Endocannabinoid dependent plasticity at GABAergic and glutamatergic synapses in the striatum is regulated by synaptic activity. Eur J Neurosci 29:32 41. Alger BE, Pitler TA, Wagner JJ, Martin LA, Morishita W, Kirov SA, Lenz RA (1996) Retrograde signalling in depolarization induced suppression of inhibition in rat hippocam pal CA1 cells. Journal of Physiology 496:197 209. Alle H, Geiger JR (2007) GABAergic spill over transmission onto hippocampal mossy fiber boutons. J Neurosci 27:942 950. Amaral DG (1978) A Golgi study of cell types in the hilar region of the hippocampus in the rat. J Comp Neurol 182:851 914. Amaral DG, Dent JA (1981) Development of the mossy fibers of the dentate gyrus: I. A light and electron microscopic study of the mossy fibers and their expansions. J Comp Neurol 195:51 86. Amaral DG, Witter MP (1989) Th e three dimensional organization of the hippocampal formation: a review of anatomical data. Neuroscience 31:571 591. Axmacher N, Winterer J, Stanton PK, Draguhn A, Muller W (2004) Two photon imaging of spontaneous vesicular release in acute brain slices an d its modulation by presynaptic GABA A receptors. Neuroimage 22:1014 1021. Azad SC, Monory K, Marsicano G, Cravatt BF, Lutz B, Zieglgansberger W, Rammes G (2004) Circuitry for associative plasticity in the amygdala involves endocannabinoid signaling. J Neur osci 24:9953 9961. Babb TL, Kupfer WR, Pretorius JK, Crandall PH, Levesque MF (1991) Synaptic reorganization by mossy fibers in human epileptic fascia dentata. Neuroscience 42:351 363.

PAGE 125

125 Behrends JC, ten Bruggencate G (1993) Cholinergic modulation of synapti c inhibition in the guinea pig hippocampus in vitro: excitation of GABAergic interneurons and inhibition of GABA release. J Neurophysiol 69:626 629. Bekenstein JW, Lothman EW (1993) Dormancy of inhibitory interneurons in a model of temporal lobe epilepsy. Science 259:97 100. Bekkers JM, Stevens CF (1995) Quantal analysis of EPSCs recorded from small numbers of synapses in hippocampal cultures. J Neurophysiol 73:1145 1156. Beltramo M, Stella N, Calignano A, Lin SY, Makriyannis A, Piomelli D (1997) Functional role of high affinity anandamide transport, as revealed by selective inhibition. Science 277:1094 1097. Berndt A, Yizhar O, Gunaydin LA, Hegemann P, Deisseroth K (2009) Bi stable neural state switches. Nat Neurosci 12:229 234. Borden LA (1996) GABA transp orter heterogeneity: pharmacology and cellular localization. Neurochem Int 29:335 356. Bowery NG (1993) GABAB receptor pharmacology. Annual Review of Pharmacology and Toxicology 33:109 147. Boyd IA, Martin AR (1956) The End Plate Potential in Mammalian Mus cle. Journal of Physiology London 132:74 91. Brenowitz SD, Regehr WG (2003) Calcium dependence of retrograde inhibition by endocannabinoids at synapses onto Purkinje cells. J Neurosci 23:6373 6384. Brickley SG, Cull Candy SG, Farrant M (1996) Development o f a tonic form of synaptic inhibition in rat cerebellar granule cells resulting from persistent activation of GABA A receptors. J Physiol 497 ( Pt 3):753 759. Brown DA, Sihra TS (2008) Presynaptic signaling by heterotrimeric G proteins. Handb Exp Pharmacol: 207 260. Brown SP, Brenowitz SD, Regehr WG (2003) Brief presynaptic bursts evoke synapse specific retrograde inhibition mediated by endogenous cannabinoids. Nature Neuroscience 6:1048 1057. Buckmaster PS, Schwartzkroin PA (1994) Hippocampal mossy cell func tion: a speculative view. [Review]. Hippocampus 4:393 402. Buckmaster PS, Dudek FE (1997) Network properties of the dentate gyrus in epileptic rats with hilar neuron loss and granule cell axon reorganization. J Neurophysiol 77:2685 2696.

PAGE 126

126 Buckmaster PS, Jon gen Relo AL (1999) Highly specific neuron loss preserves lateral inhibitory circuits in the dentate gyrus of kainate induced epileptic rats. Journal of Neuroscience 19:9519 9529. Buckmaster PS, Wenzel HJ, Kunkel DD, Schwartzkroin PA (1996) Axon arbors and synaptic connections of hippocampal mossy cells in the rat in vivo. J Comp Neurol 366:271 292. Buckmaster PS, Strowbridge BW, Kunkel DD, Schmiege DL, Schwartzkroin, Pa (1992) Mossy cell axonal projections to the dentate gyrus molecular layer in the rat hip pocampal slice. Hippocampus 2:349 362. Cadas H, di TE, Piomelli D (1997) Occurrence and biosynthesis of endogenous cannabinoid precursor, N arachidonoyl phosphatidylethanolamine, in rat brain. Journal of Neuroscience 17:1226 1242. Cadas H, Gaillet S, Beltramo M, Venance L, Piomelli D (1996) Biosynthesis of an endogenous cannabinoid precursor in neurons and its control by calcium and cAMP. Journal of Neuroscience 16:3934 3942. Capogna M, Ga hwiler BH, Thompson SM (1996) Calcium l atrotoxin on spontaneous and evoked synaptic transmission in the hippocampus. J Neurophysiol 76:3149 3158. Carta M, Mameli M, Valenzuela CF (2004) Alcohol enhances GABAergic transmission to cerebellar granule cells via an increase in Golgi cell excitabilit y. Journal of Neuroscience 24:3746 3751. Cavelier P, Hamann M, Rossi D, Mobbs P, Attwell D (2005) Tonic excitation and inhibition of neurons: ambient transmitter sources and computational consequences. Prog Biophys Mol Biol 87:3 16. Centonze D, Rossi S, Pr osperetti C, Gasperi V, De CV, Bari M, Tscherter A, Febbraro F, Bernardi G, Maccarrone M (2007) Endocannabinoids limit metabotropic glutamate 5 receptor mediated synaptic inhibition of striatal principal neurons. Mol Cell Neurosci 35:302 310. Chen C, Regeh r WG (2003) Presynaptic modulation of the retinogeniculate synapse. J Neurosci 23:3130 3135. Chen K, Ratzliff A, Hilgenberg L, Gulyas A, Freund TF, Smith M, Dinh TP, Piomelli D, Mackie K, Soltesz I (2003) Long term plasticity of endocannabinoid signaling i nduced by developmental febrile seizures. Neuron 39:599 611. Cherubini E, Gaiarsa JL, Ben Ari Y (1991) GABA: an excitatory transmitter in early postnatal life. Trends in Neurosciences 14:515 519. Chevaleyre V, Castillo PE (2003) Heterosynaptic LTD of hippo campal GABAergic synapses: a novel role of endocannabinoids in regulating excitability. Neuron 38:461 472.

PAGE 127

127 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 CNS. Annual Review of Neuroscience 29:37 76. Chevaleyre V, Heifets BD, Kaeser PS, Sudhof TC, Castillo PE (2007) Endocannabinoid mediated long term plasticity requires cAMP/PKA signaling and RIM1 alpha. Neuron 54:801 812. Chicurel ME, Harris KM (1992) Three dimensional analysis of the structure and composition of CA3 branched dendritic spines and their synaptic relationships with mossy fiber boutons in the rat hippocampus. J Comp Neurol 325:169 182 Childers SR, Deadwyler SA (1996) Role of cyclic AMP in the actions of cannabinoid receptors. Biochem Pharmacol 52:819 827. Chiu CQ, Castillo PE (2008) Input specific plasticity at excitatory synapses mediated by endocannabinoids in the dentate gyrus. Neu ropharmacology 54:68 78. Claiborne BJ, Amaral DG, Cowan WM (1986) A light and electron microscopic analysis of the mossy fibers of the rat dentate gyrus. J Comp Neurol 246:435 458. Conti F, Minelli A, Melone M (2004) GABA transporters in the mammalian cere bral cortex: localization, development and pathological implications. Brain Res Brain Res Rev 45:196 212. Cossart R, Epsztein J, Tyzio R, Becq H, Hirsch J, Ben Ari Y, Crepel V (2002) Quantal release of glutamate generates pure kainate and mixed AMPA/kainat e EPSCs in hippocampal neurons. Neuron 35:147 159. Coulter DA (2004) Functional consequences of hilar mossy cell loss in temporal lobe epilepsy: proepileptic or antiepileptic? Epilepsy Curr 4:250 253. De Sevilla DF, Cabezas C, De Prada AN, Sanchez Jimenez A, Buno W (2002) Selective muscarinic regulation of functional glutamatergic Schaffer collateral synapses in rat CA1 pyramidal neurons. J Physiol 545:51 63. Del Castillo J, Katz B (1954) Quantal components of the end plate potential. Journal of Physiology 124:560 573. Di Marzo V, Fontana A, Cadas H, Schinelli S, Cimino G, Schwartz JC, Piomelli D (1994) Formation and inactivation of endogenous cannabinoid anandamide in central neurons. Nature 372:686 691. Diana MA, Levenes C, Mackie K, Marty A (2002) Short t erm retrograde inhibition of GABAergic synaptic currents in rat Purkinje cells is mediated by endogenous cannabinoids. J Neurosci 22:200 208.

PAGE 128

128 Dietrich D, Kral T, Clusmann H, Friedl M, Schramm J (2002) Presynaptic group II metabotropic glutamate receptors r educe stimulated and spontaneous transmitter release in human dentate gyrus. Neuropharmacology 42:297 305. Dinh TP, Carpenter D, Leslie FM, Freund TF, Katona I, Sensi SL, Kathuria S, Piomelli D (2002) Brain monoglyceride lipase participating in endocannabi noid inactivation. Proc Natl Acad Sci USA 99:10819 10824. Dittman JS, Regehr WG (1996) Contributions of calcium dependent and calcium independent mechanisms to presynaptic inhibition at a cerebellar synapse. J Neurosci 16:1623 1633. Dittman JS, Regehr WG ( 1997) Mechanism and kinetics of heterosynaptic depression at a cerebellar synapse. J Neurosci 17:9048 9059. Duguid IC, Smart TG (2004) Retrograde activation of presynaptic NMDA receptors enhances GABA release at cerebellar interneuron Purkinje cell synapse s. Nature Neuroscience 7:525 533. Duguid IC, Pankratov Y, Moss GW, Smart TG (2007) Somatodendritic release of glutamate regulates synaptic inhibition in cerebellar Purkinje cells via autocrine mGluR1 activation. J Neurosci 27:12464 12474. Dutar P, Nicoll RA (1988) A physiological role for GABA B receptors in the central nervous system. Nature 332:156 158. Dzubay JA, Jahr CE (1999) The concentration of synaptically released glutamate outside of the climbing fiber Purkinje cell synaptic cleft. J Neurosci 19:5 265 5274. Eccles JC, Schmidt R, Willis WD (1963) Pharmacological Studies on Presynaptic Inhibition. J Physiol 168:500 530. 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 USA 105:8142 8147. Engelman HS, MacDermott AB (2004) Presynaptic ionotropic receptors and control of transmit ter release. Nat Rev Neurosci 5:135 145. Engler B, Freiman I, Urbanski M, Szabo B (2006) Effects of Exogenous and Endogenous Cannabinoids on GABAergic Neurotransmission between the Caudate Putamen and the Globus Pallidus in the Mouse. J Pharmacol Exp Ther 316:608 617. Farooqui AA, Rammohan KW, Horrocks LA (1989) Isolation, characterization, and regulation of diacylglycerol lipases from the bovine brain. Ann N Y Acad Sci 559:25 36. Fatt P, Katz B (1950) Membrane potentials at the motor end plate. J Physiol 3 :46p 47p.

PAGE 129

129 Fatt P, Katz B (1952) Spontaneous subthreshold activity at motor nerve endings. J Physiol 117:109 128. Fearon IM, Zhang M, Vollmer C, Nurse CA (2003) GABA mediates autoreceptor feedback inhibition in the rat carotid body via presynaptic GABA B rec eptors and TASK 1. J Physiol 553:83 94. Feil R, Kleppisch T (2008) NO/cGMP dependent modulation of synaptic transmission. Handb Exp Pharmacol:529 560. Fernandez de Sevilla D, Buno W (2003) Presynaptic inhibition of Schaffer collateral synapses by stimulati on of hippocampal cholinergic afferent fibres. Eur J Neurosci 17:555 558. Foldy C, Neu A, Jones MV, Soltesz I (2006) Presynaptic, activity dependent modulation of cannabinoid type 1 receptor mediated inhibition of GABA release. J Neurosci 26:1465 1469. Fre rking M, Petersen CC, Nicoll RA (1999) Mechanisms underlying kainate receptor mediated disinhibition in the hippocampus. Proc Natl Acad Sci USA 96:12917 12922. Freund TF, Magloczky Z (1993) Early degeneration of calretinin containing neurons in the rat hip pocampus after ischemia. Neuroscience 56:581 596. Freund TF, Buzsaki G (1996) 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. Fr otscher M, Soriano E, Leranth C (1992) Cholinergic and GABAergic neurotransmission in the fascia dentata: electron microscopic immunocytochemical studies in rodents and primates. Epilepsy Res Suppl 7:65 78. Frotscher M, Seress L, Schwerdtfeger WK, Buhl E ( 1991) The mossy cells of the fascia dentata: a comparative study of their fine structure and synaptic connections in rodents and primates. J Comp Neurol 312:145 163. Fukudome Y, Ohno Shosaku T, Matsui M, Omori Y, Fukaya M, Tsubokawa H, Taketo MM, Watanabe 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. Gahwiler BH, Brown DA (1985) GABA B receptor activated K + current in voltage clamped CA3 pyramidal cells in hippocampal cultures. Proc Natl Acad Sci USA 82:1558 1562. Galante M, Diana MA (2004) Group I metabotropic glutamate receptors inhibit GABA release at interneuron Purkinje cell synapses through endocannabinoid production. J Neurosci 24:4865 4874.

PAGE 130

130 Gaoni Y, Mechoulam R (1964) Isolation, structure and partial synthesis of an active constituent of hashish. JAmChemSoc 86:1646 1647. Gerdeman GL, Ronesi J, Lovinger DM (20 02) Postsynaptic endocannabinoid release is critical to long term depression in the striatum. Nature Neuroscience 5:446 451. Glykys J, Mody I (2006) Hippocampal network hyperactivity after selective reduct ion of tonic inhibition in GABA A receptor alpha5 su bunit deficient mice. J Neurophysiol 95:2796 2807. Glykys J, Mody I (2007) The main source of ambient GABA responsible for tonic inhibition in the mouse hippocampus. J Physiol 582:1163 1178. Glykys J, Peng Z, Chandra D, Homanics GE, Houser CR, Mody I (2006 ) A new naturally occurring GABA(A) receptor subunit partnership with high sensitivity to ethanol. Nature Neuroscience. Golarai G, Cavazos JE, Sutula TP (1992) Activation of the dentate gyrus by pentylenetetrazol evoked seizures induces mossy fiber synaptic reorganization. Brain Research 593:257 264. Goparaju SK, Ueda N, Taniguchi K, Yamamoto S (1999) Enzymes of porcine brain hydrolyzing 2 arachidonoylglycerol, an endogenous ligand of cannabinoid receptors. Biochem Pharmacol 57:417 423. Grover LM, La mbert NA, Schwartzkroin PA, Teyler TJ (1993) Role of HCO 3 ions in depolarizing GABA A receptor mediated responses in pyramidal cells of rat hippocampus. J Neurophysiol 69:1541 1555. Gulyas AI, Cravatt BF, Bracey MH, Dinh TP, Piomelli D, Boscia F, Freund 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. Hajos N, Ledent C, Freund TF (2001) Novel cannabinoid sensitive receptor medi ates 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. Hamlyn LH (1962) Fine Structure of Mossy Fibre Endings in Hippocampus of Rabbit. Journal of Anatomy 96:112 &. Hashimotodani Y, Ohno Shosaku T, Tsubokawa H, Ogata H, Emoto K, Maejima T, Araishi K, Shin HS, Kano M (2005) Phospholipa se Cbeta serves as a coincidence detector through its Ca 2+ dependency for triggering retrograde endocannabinoid signal. Neuron 45:257 268.

PAGE 131

131 Heifets BD, Chevaleyre V, Castillo PE (2008) Interneuron activity controls endocannabinoid mediated presynaptic plast icity through calcineurin. Proc Natl Acad Sci USA 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 hippocamp al mossy fiber pathway: a review. Neuroscience 98:407 427. Higgs HN, Glomset JA (1994) Identification of a phosphatidic acid preferring phospholipase A1 from bovine brain and testis. Proc Natl Acad Sci U S A 91:9574 9578. Hillard CJ, Wilkison DM, Edgemond WS, Campbell WB (1995) Characterization of the kinetics and distribution of N arachidonylethanolamine (anandamide) hydrolysis by rat brain. Biochim Biophys Acta 1257:249 256. Hillard CJ, Edgemond WS, Jarrahian A, Campbell WB (1997) Accumulation of N arachi donoylethanolamine (anandamide) into cerebellar granule cells occurs via facilitated diffusion. J Neurochem 69:631 638. Hirata K, Sawada S, Yamamoto C (1992) Quantal analysis of suppressing action of baclofen on mossy fiber synapses in guinea pig hippocamp us. Brain Research 578:33 40. Hoffman AF, Lupica CR (2000) Mechanisms of cannabinoid inhibition of GABA(A) synaptic transmission in the hippocampus. J Neurosci 20:2470 2479. Hoffman AF, Lupica CR (2001) Direct actions of cannabinoids on synaptic transmissi on in the nucleus accumbens: a comparison with opioids. J Neurophysiol 85:72 83. 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 hip pocampus. European Journal of Neuroscience 22:2387 2391. 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) Excitatory afferents to CA3 pyramidal cells display differential sensitivity to CB1 dependent inhibition of synaptic transmission. Neuropharmacology. Hollmann M, Heinemann S (1994) Cloned Glutamate Receptors. Annual Review of N euroscience 17:31 108. Hounsgaard J (1978) Presynaptic inhibitory action of acetylcholine in area CA1 of the hippocampus. Experimental Neurology 62:787 797.

PAGE 132

132 Houser CR (1999) Neuronal loss and synaptic reorganization in temporal lobe epilepsy. AdvNeurol 79: 743 761. Hsu M, Buzsaki G (1993) Vulnerability of mossy fiber targets in the rat hippocampus to forebrain ischemia. Journal of Neuroscience 13:3964 3979. Isaacson JS, Walmsley B (1995) Counting quanta: direct measurements of transmitter release at a centra l synapse. Neuron 15:875 884. Ishikawa T, Kaneko M, Shin HS, Takahashi T (2005) Presynaptic N type and P/Q type Ca 2+ channels mediating synaptic transmission at the calyx of Held of mice. J Physiol 568:199 209. Ishizuka N, Weber J, Amaral DG (1990) Organiz ation of intrahippocampal projections originating from CA3 pyramidal cells in the rat. J Comp Neurol 295:580 623. 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. Jarolimek W, Misgeld U (1992) On the inhibitory actions of baclofen and gamma aminobutyric acid in rat ventral midbrain culture. J Physiol 451:419 443. Jefferys JGR, Traub RD (1998) 'Dormant' inhibitory neurons: do they exist and what is their functional impact? Epilepsy Research 32:104 113. Johnston D, Amaral DG, Shepherd GM (1998) Hippocampus. In: The synaptic organization of the brain, pp 417 458. New York Oxford: Oxford University Press. Jonas P, Major G, Sakmann B (1993) Quantal components of unitary EPSCs at the mossy fibre synapse on CA 3 pyramidal cells of rat hippocampus. J Physiol 472:615 663. Kaila K, Voipio J (1987) Postsynaptic fall in intracellular pH induced by GABA activated bicarbonate conductance. Nature 330:163 165. Kaneda M, Farrant M, Cull Candy SG (1995) Whole cell and sing le channel currents activated by GABA and glycine in granule cells of the rat cerebellum. J Physiol 485 ( Pt 2):419 435. Katona I, Sperlagh 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, Rancz EA, Acsady L, Ledent C, Mackie K, Hajos N, Freund TF (2001) Distribution of CB1 cannabinoid receptors in the amygdala and their role in the cont rol of GABAergic transmission. Journal of Neuroscience 21:9506 9518.

PAGE 133

133 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. Journal of Neurosci ence 26:5628 5637. Katona I, Sperlagh B, Magloczky Z, Santha E, Kofalvi A, Czirjak S, Mackie K, Vizi ES, Freund TF (2000) GABAergic interneurons are the targets of cannabinoid actions in the human hippocampus. Neuroscience 100:797 804. Kaupmann K, Huggel K Heid J, Flor PJ, Bischoff S, Mickel SJ, McMaster G, Angst C, Bittiger H, Froestl W, Bettler B (1997) Expression cloning of GABA(B) receptors uncovers similarity to metabotropic glutamate receptors. Nature 386:239 246. Kawamura Y, Fukaya M, Maejima T, Yos hida 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. Journal of Neuroscience 26:2991 3001. Kee MZ, Wuskell JP, Loew LM, Augustine GJ, Sekino Y (2008) Imaging activity of neuronal populations with new long wavelength voltage sensitive dyes. Brain Cell Biol 36:157 172. Kerr AM, Capogna M (2007) Unitary IPSPs enhance hilar mossy cell gain in the rat hippocampus. J Physiol 578: 451 470. Kew JN, Pflimlin MC, Kemp JA, Mutel V (2002) Differential regulation of synaptic transmission by mGlu2 and mGlu3 at the perforant path inputs to the dentate gyrus and CA1 revealed in mGlu2 / mice. Neuropharmacology 43:215 221. Khvotchev M, Lonart G, Sudhof TC (2000) Role of calcium in neurotransmitter release evoked by alpha latrotoxin or hypertonic sucrose. Neuroscience 101:793 802. Kilbride J, Huang LQ, Rowan MJ, Anwyl R (1998) Presynaptic inhibitory action of the group II metabotropic glutamate receptor agonists, LY354740 and DCG IV. Eur J Pharmacol 356:149 157. 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 Kneisler TB, Dingledine R (1995a) Spontaneous and synaptic input from granule cells and the perforant path to dentate basket cells in the rat hippocampus. Hippocampus 5:151 164. Kneisler TB, Dingledine R (1995b) Synaptic input from CA3 pyramidal cells to dentate basket cells in rat hippocampus. Journal of Physiology 487:125 146. Kohler C (1985) A projection from the deep layers of the entorhinal area to the hippocampal formation in the rat brain. Neuroscience Letters 56:13 19.

PAGE 134

134 Kolaj M, Bai D, Renaud LP (2 004) GABA B receptor modulation of rapid inhibitory and excitatory neurotransmission from subfornical organ and other afferents to median preoptic nucleus neurons. J Neurophysiol 92:111 122. Kreitzer AC, Regehr WG (2001a) Retrograde inhibition of presynapti c 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 endogenous cannabinoids. J Neurosci 21:RC17 4. Kreitzer AC, Carter AG, Regehr WG (2002) Inhibition of interneuron firing extends the spread of endocannabinoid signaling in the cerebellum. Neuron 34:787 796. Kullmann DM, Ruiz A, Rusakov DM, Scott R, Semyanov A, Walker MC (2005) Presynaptic, extrasyna ptic and axonal GABA A receptors in the CNS: where and why? Prog Biophys Mol Biol 87:33 46. Lanthorn TH, Cotman CW (1981) Baclofen selectively inhibits excitatory synaptic transmission in the hippocampus. Brain Research 225:171 178. Larimer P, Strowbridge B W (2008) Nonrandom local circuits in the dentate gyrus. J Neurosci 28:12212 12223. Lauckner JE, Hille B, Mackie K (2005) The cannabinoid agonist WIN55,212 2 increases intracellular calcium via CB1 receptor coupling to G q/11 G proteins. Proc Natl Acad Sci USA 102:19144 19149. Laurberg S, Sorensen KE (1981) Associational and commissural collaterals of neurons in the hippocampal formation (hilus fasciae dentatae and subfield CA3). Brain Res 212:287 300. Lawrence JJ, Grinspan ZM, McBain CJ (2004) Quantal transmission at mossy fibre targets in the CA3 region of the rat hippocampus. J Physiol 554:175 193. Leao RM, Mellor JR, Randall AD (2000) Tonic benzodiazepine sensitive GABAergic inhibition in cultured rodent cerebellar granule cells. Neuropharmacology 39 :990 1003. Lei S, McBain CJ (2002) Distinct NMDA receptors provide differential modes of transmission at mossy fiber interneuron synapses. Neuron 33:921 933. Lei S, McBain CJ (2003) GABA B receptor modulation of excitatory and inhibitory synaptic transmissi on onto rat CA3 hippocampal interneurons. JPhysiol 546:439 453. Leite JP, Babb TL, Pretorius JK, Kuhlman PA, Yeoman KM, Mathern GW (1996) Neuron loss, mossy fiber sprouting, and interictal spikes after intrahippocampal kainate in developing rats. Epilepsy Research 26:219 231.

PAGE 135

135 Leranth C, Malcolm AJ, Frotscher M (1990) Afferent and efferent synaptic connections of somatostatin immunoreactive neurons in the rat fascia dentata. J Comp Neurol 295:111 122. Li XG, Somogyi P, Ylinen A, Buzsaki G (1994) The hippocam pal CA3 network: an in vivo intracellular labeling study. J Comp Neurol 339:181 208. Liu HN, Kurotani T, Ren M, Yamada K, Yoshimura Y, Komatsu Y (2004) Presynaptic activity and Ca 2+ entry are required for the maintenance of NMDA receptor independent LTP at visual cortical excitatory synapses. J Neurophysiol 92:1077 1087. Llano I, Leresche N, Marty A (1991) Calcium entry increases the sensitivity of cerebellar Purkinje cells to applied GABA and decreases inhibitory synaptic currents. Neuron 6:565 574. 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. Nature Neuroscience 3:1256 1265. Lothman EW, Bertram EH, III, Kapur J, Bekenstein JW (1996) Temporal lobe epilepsy: studies in a rat model showing dormancy of GABAergic inhibitory interneurons. Epilepsy Res Suppl 12:145 156. Lovinger DM (2008) Presynaptic modulation by endocannabinoids. Handb Exp Pharmacol:435 477. Lowenstein DH, Thomas M J, Smith DH, McIntosh TK (1992) Selective vulnerability of dentate hilar neurons following traumatic brain injury: a potential mechanistic link between head trauma and disorders of the hippocampus. Journal of Neuroscience 12:4846 4853. Lutz B (2004) On dem and activation of the endocannabinoid system in the control of neuronal excitability and epileptiform seizures. Biochem Pharmacol 68:1691 1698. Lysetskiy M, Foldy C, Soltesz I (2005) Long and short term plasticity at mossy fiber synapses on mossy cells in the rat dentate gyrus. Hippocampus 15:691 696. Maccarrone M, Rossi S, Bari M, De CV, Fezza F, Musella A, Gasperi V, Prosperetti C, Bernardi G, Finazzi Agro A, Cravatt BF, Centonze D (2008) Anandamide inhibits metabolism and physiological actions of 2 arac hidonoylglycerol in the striatum. Nature Neuroscience 11:152 159. Mackie K, Hille B (1992) Cannabinoids inhibit N type calcium channels in neuroblastoma glioma cells. Proc Natl Acad Sci U S A 89:3825 3829. Maejima T, Ohno Shosaku T, Kano M (2001a) Endogeno us cannabinoid as a retrograde messenger from depolarized postsynaptic neurons to presynaptic terminals. Neurosci Res 40:205 210.

PAGE 136

136 Maejima T, Hashimoto K, Yoshida T, Aiba A, Kano M (2001b) Presynaptic inhibition caused by retrograde signal from metabotropic glutamate to cannabinoid receptors. Neuron 31:463 475. Maejima T, Oka S, Hashimotodani Y, Ohno Shosaku T, Aiba A, Wu D, Waku K, Sugiura T, Kano M (2005) Synaptically Driven Endocannabinoid Release Requires Ca 2+ Assisted Metabotropic Glutamate Receptor Sub type 1 to Phospholipase C {beta}4 Signaling Cascade in the Cerebellum. Journal of Neuroscience 25:6826 6835. Magloczky Z, Freund TF (1993) Selective neuronal death in the contralateral hippocampus following unilateral kainate injections into the CA3 subfie ld. Neuroscience 56:317 335. Marsha ll FH (2008) The role of GABA B receptors in the regulation of excitatory neurotransmission. Results Probl Cell Differ 44:87 98. Marsicano G, Lutz B (1999) Expression of the cannabinoid receptor CB1 in distinct neuronal su bpopulations in the adult mouse forebrain. European Journal of Neuroscience 11:4213 4225. Marsicano G, Wotjak CT, Azad SC, Bisogno T, Rammes G, Cascio MG, Hermann H, Tang J, Hofmann C, Zieglgansberger W, Di Marzo V, Lutz B (2002) The endogenous cannabinoid system controls extinction of aversive memories. Nature 418:530 534. Martin LA, Alger BE (1999) Muscarinic facilitation of the occurrence of depolarization induced suppression of inhibition in rat hippocampus. Neuroscience 92:61 71. Martin LA, Wei DS, Alg er BE (2001) Heter ogeneous susceptibility of GABA A receptor mediated IPSCs to depolarization induced suppression of inhibition in rat hippocampus. J Physiol 532:685 700. Matsuda LA, Lolait SJ, Brownstein MJ, Young AC, Bonner TI (1990) Structure of a cannab inoid receptor and functional expression of the cloned cDNA. Nature 346:561 564. Matsui K, Jahr CE (2003) Ectopic release of synaptic vesicles. Neuron 40:1173 1183. Misgeld U, Bijak M, Jarolimek W (1995) A physiological role for GABA B receptors and the effects of baclofen in the mammalian central nervous system. Prog Neurobiol 46:423 462. Monory K et al. (2006) The endocannabinoid system controls key epileptogenic circuits in the hippocampus. Neuron 51:455 466. Morisset V, Urban L (200 1) Cannabinoid induced presynaptic inhibition of glutamatergic EPSCs in substantia gelatinosa neurons of the rat spinal cord. J Neurophysiol 86:40 48. Mtchedlishvili Z, Kapur J (2006) High Affinity, Slowly Desensitizing GABA A Receptors Mediate Tonic Inhibi tion in Hippocampal Dentate Granule Cells. Molecular Pharmacology 69:564 575.

PAGE 137

137 Mu J, Zhuang SY, Kirby MT, Hampson RE, Deadwyler SA (1999) Cannabinoid receptors differentially modulate potassium A and D currents in hippocampal neurons in culture. J Pharmacol Exp Ther 291:893 902. Muller W, Misgeld U (1991) Picrotoxin and 4 aminopyridine induced activity in hilar neurons in the guinea pig hippocampal slice. J Neurophysiol 65:141 147. Munro S, Thomas KL, bu Shaar M (1993) Molecular characterization of a periph eral receptor for cannabinoids. Nature 365:61 65. Nadler JV, Perry BW, Cotman CW (1978) Intraventricular kainic acid preferentially destroys hippocampal pyramidal cells. Nature 271:676 677. Nahir B, Bhatia C, Frazier CJ (2007) Presynaptic inhibition of exc itatory afferents to hilar mossy cells. J Neurophysiol 97:4036 4047. Nakatsuka T, Chen HX, Roper SN, Gu JG (2003) Cannabinoid receptor 1 activation suppresses inhibitory synaptic activity in human dentate gyrus. Neuropharmacology 45:116 121. Nicoll RA, Sch mitz D (2005) Synaptic plasticity at hippocampal mossy fibre synapses. Nat Rev Neurosci 6:863 876. Nusser Z, Mody I (2002) Selective modulation of tonic and phasic inhibitions in dentate gyrus granule cells. J Neurophysiol 87:2624 2628. Ohno Shosaku T, Sho saku J, Tsubokawa H, Kano M (2002) Cooperative endocannabinoid production by neuronal depolarization and group I metabotropic glutamate receptor activation. Eur J Neurosci 15:953 961. 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 (2006) Receptor independent actions of cannabinoids on cel l membranes: focus on endocannabinoids. Pharmacol Ther 111:114 144. Pelkey KA, Topolnik L, Lacaille JC, McBain CJ (2006) Compartmentalized Ca 2+ channel regulation at divergent mossy fiber release sites underlies target cell dependent plasticity. Neuron 52: 497 510. Penttonen M, Kamondi A, Sik A, Acsady L, Buzsaki G (1997) Feed forward and feed back activation of the dentate gyrus in vivo during dentate spikes and sharp wave bursts. Hippocampus 7:437 450. Penzo MA, Pena JL (2009) Endocannabinoid mediated long term depression in the avian midbrain expressed presynaptically and postsynaptically. J Neurosci 29:4131 4139.

PAGE 138

138 Pertwee RG (2005) Pharmacological actions of cannabinoids. Handb Exp Pharmacol:1 51. Petrini EM, Marchionni I, Zacchi P, Sieghart W, Cherubini E (2004) C lustering of extrasynaptic GABA A receptors modulates tonic inhibition in cultured hippocampal neurons. Journal of Biological Chemistry 279:45833 45843. Piomelli D (2003) The molecular logic of endocannabinoid signalling. Nat Rev Neurosci 4:873 884 Pitler TA, Alger BE (1992a) Cholinergic excitation of GABAergic interneurons in the rat hippocampal slice. J Physiol 450:127 142. Pitler TA, Alger BE (1992b) Postsynaptic spike firing reduces synaptic GABA A responses in hippocampal pyramidal cells. J Neu rosci 12:4122 4132. Pitler TA, Alger BE (1994) Depolarization induced suppression of GABAergic inhibition in rat hippocampal pyramidal cells: G protein involvement in a presynaptic mechanism. Neuron 13:1447 1455. Porcello DM, Huntsman MM, Mihalek RM, Homan ics GE, Huguenard JR (2003) Intact synaptic GABAergic inhibition and altered neurosteroid modulation of thalamic relay neurons in mice lacking delta subunit. J Neurophysiol 89:1378 1386. Prenosil GA, Schneider Gasser EM, Rudolph U, Keist R, Fritschy JM, Vo gt KE (2006) Specific subtypes of GABA A receptors mediate phasic and tonic forms of inhibition in hippocampal pyramidal neurons. J Neurophysiol. Raiteri M (2008) Presynaptic metabotropic glutamate and GABA B receptors. Handb Exp Pharmacol:373 407. Ramn y C ajal S (19 11 ) His and 2. Translated by A Maloine Paris. 1911 Ratzliff AH, Santhakumar V, Howard A, Soltesz I (2002) Mossy cells in epilepsy: rigor mortis or vigor mortis? Trends Neurosci 25: 140 144. Ratzliff AH, Howard AL, Santhakumar V, Osapay I, Soltesz I (2004) Rapid deletion of mossy cells does not result in a hyperexcitable dentate gyrus: implications for epileptogenesis. J Neurosci 24:2259 2269. Reich CG, Karson MA, Karnup SV, Jones LM, Alger BE (2005) Regulation of IPSP theta rhythm by muscarinic receptors and endocannabinoids in hippocampus. J Neurophysiol. Ribak CE, Seress L (1983) 5 Types of Basket Cell in the Hippocampal Dentate Gyrus a Combined Golgi and Electron Microscopic Stud y. Journal of Neurocytology 12:577 597. Ribak CE, Seress L, Amaral DG (1985) The development, ultrastructure and synaptic connections of the mossy cells of the dentate gyrus. J Neurocytol 14:835 857.

PAGE 139

139 Richerson GB, Wu Y (2003) Dynamic equilibrium of neurotr ansmitter transporters: not just for reuptake anymore. J Neurophysiol 90:1363 1374. Robbe D, Alonso G, Duchamp F, Bockaert J, Manzoni OJ (2001) Localization and mechanisms of action of cannabinoid receptors at the glutamatergic synapses of the mouse nucleu s accumbens. J Neurosci 21:109 116. Robbe D, Kopf M, Remaury A, Bockaert J, Manzoni OJ (2002) Endogenous cannabinoids mediate long term synaptic depression in the nucleus accumbens. Proc Natl Acad Sci USA 99:8384 8388. Romero J, Hillard CJ, Calero M, Rabano A (2002) Fatty acid amide hydrolase localization in the human central nervous system: an immunohistochemical study. Brain Res Mol Brain Res 100:85 93. Ronesi J, Gerdeman GL, Lovinger DM (2004) Disruption of endocannabinoid release and striatal long term depression by postsynaptic blockade of endocannabinoid membrane transport. J Neurosci 24:1673 1679. Rosenmund C, Stevens CF (1996) Definition of the readily releasable pool of vesicles at hippocampal synapses. Neuron 16:1197 1207. Rossi DJ, Hamann M, Attwell D (2003) Multiple modes of GABAergic inhibition of rat cerebellar granule cells. J Physiol 548:97 110. Rouach N, Nicoll RA (2003) Endocannabinoids contribute to short term but not long term mGluR induced depression in the hippocampus. EurJ Neurosci 18:1017 1020. Ruiz A, Fabian Fine R, Scott R, Walker MC, Rusakov DA, Kullmann DM (2003) GABA A receptors at hippocampal mossy fibers. Neuron 39:961 973. Safo PK, Regehr WG (2005) Endocannabinoids control the induction of cerebellar LTD. Neuron 48:647 659. Safo PK, Cravatt BF, Regehr WG (2006) Retrograde endocannabinoid signaling in the cerebellar cortex. Cerebellum 5:134 145. Sah P, Faber ES, Lopez De AM, Power J (2003) The amygdaloid complex: anatomy and physiology. Physiol Rev 83:803 834. Sakaba T, Neher E (2003) Direct modulation of synaptic vesicle priming by GABA B receptor activation at a glutamatergic synapse. Nature 424:775 778. Santhakumar V, Aradi I, Soltesz I (2005) Role of mossy fiber sprouting and mossy cell loss in hyperexcitability: a network m odel of the dentate gyrus incorporating cell types and axonal topography. J Neurophysiol 93:437 453.

PAGE 140

140 Santhakumar V, Bender R, Frotscher M, Ross ST, Hollrigel GS, Toth Z, Soltesz I (2000) Granule cell hyperexcitability in the early post traumatic rat dentat e gyrus: the 'irritable mossy cell' hypothesis. J Physiol 524 Pt 1:117 134. Scanziani M, Capogna M, Gahwiler BH, Thompson SM (1992) Presynaptic inhibition of miniature excitatory synaptic currents by baclofen and adenosine in the hippocampus. Neuron 9:919 927. Scharfman HE (1991) Dentate hilar cells with dendrites in the molecular layer have lower thresholds for synaptic activation by perforant path than granule cells. Journal of Neuroscience 11:1660 1673. Scharfman HE (1993) Characteristics of spontaneous and evoked EPSPs recorded from dentate spiny hilar cells in rat hippocampal slices. J Neurophysiol 70:742 757. Scharfman HE (1994a) Evidence from simultaneous intracellular recordings in rat hippocampal slices that area CA3 pyramidal cells innervate dentat e hilar mossy cells. J Neurophysiol 72:2167 2180. Scharfman HE (1994b) EPSPs of dentate gyrus granule cells during epileptiform bursts of dentate hilar "mossy" cells and area CA3 pyramidal cells in disinhibited rat hippocampal slices. Journal of Neuroscien ce 14:6041 6057. 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, Schwartzkroin PA (1988) Further studies of the effe cts of somatostatin and related peptides in area CA1 of rabbit hippocampus. Cell Mol Neurobiol 8:411 429. Scharfman HE, Kunkel DD, Schwartzkroin PA (1990) Synaptic connections of dentate granule cells and hilar neurons: results of paired intracellular reco rdings and intracellular horseradish peroxidase injections. Neuroscience 37:693 707. Schlander M, Thomalske G, Frotscher M (1987) Fine Structure of Gabaergic Neurons and Synapses in the Human Dentate Gyrus. Brain Research 401:185 189. Schoepp DD (2001) Unv eiling the functions of presynaptic metabotropic glutamate receptors in the central nervous system. J Pharmacol Exp Ther 299:12 20. Schwartzkroin PA, Scharfman HE, Sloviter RS (1990) Similarities in circuitry between Ammon's horn and dentate gyrus: local interactions and parallel processing. Progress in Brain Research 83:269 286. Scimemi A, Semyanov A, Sperk G, Kullmann DM, Walker MC (2005) Multiple and plastic receptors mediate tonic GABAA receptor currents in the hippocampus. Journal of Neuroscience 25:1 0016 10024.

PAGE 141

141 Scoville WB, Milner B (1957) Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry 20:11 21. Semyanov A, Walker MC, Kullmann DM (2003) GABA uptake regulates cortical excitability via cell type specific tonic i nhibition. Nature Neuroscience 6:484 490. Sergeeva OA, Doreulee N, Chepkova AN, Kazmierczak T, Haas HL (2007) Long term depression of cortico striatal synaptic transmission by DHPG depends on endocannabinoid release and nitric oxide synthesis. European Jou rnal of Neuroscience 26:1889 1894. Sloviter RS (1991) Permanently altered hippocampal structure, excitability, and inhibition after experimental status epilepticus in the rat: the "dormant basket cell" hypothesis and its possible relevance to temporal lobe epilepsy. Hippocampus 1:41 66. Sloviter RS (1994) The functional organization of the hippocampal dentate gyrus and its relevance to the pathogenesis of temporal lobe epilepsy. Annals of Neurology 35:640 654. Sloviter RS, Zappone CA, Harvey BD, Bumanglag A V, 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. Soltesz I, Bourassa J, Deschenes M (1993) The Behavior of Mossy Cells of the Rat Dentate Gyrus during Theta Oscillations in Vivo. Neuroscience 57:555 564. Soriano E, Frotscher M (1994) Mossy cells of the rat fascia dentata are glutamate immunoreactive. Hippocampus 4:65 69. Stell BM, Mody I (2002) Receptors with different affinities mediate phasic and tonic GABA A conductances in hippocampal neurons. Journal of Neuroscience 22:RC223. Stella N, Schweitzer P, Piomelli D (1997) A second endogenous cannabinoid that modulates long term potentiation. Nature 388:773 778. Strader CD, Fong TM, Tota MR, Underwood D, Dixon RA (1994) Structure and function of G protein coupled receptors. Annu Rev Biochem 63:101 132. Strowbridge BW, Buckmaster PS, Schwartzkroin PA (1992) Potentiation of spontane ous synaptic activity in rat mossy cells. Neuroscience Letters 142:205 210. Sullivan JM (1999) Mechanisms of cannabinoid receptor mediated inhibition of synaptic transmission in cultured hippocampal pyramidal neurons. J Neurophysiol 82:1286 1294. Sutula T, Cascino G, Cavazos J, Parada I, Ramirez L (1989) Mossy fiber synaptic reorganization in the epileptic human temporal lobe. Annals of Neurology 26:321 330. Takahashi T, Kajikawa Y, Tsujimoto T (1998) G Protein coupled modulation of presynaptic calcium curr ents and transmitter release by a GABA B receptor. J Neurosci 18:3138 3146.

PAGE 142

142 Thalmann RH, Ayala GF (1982) A late increase in potassium conductance follows synaptic stimulation of granule neurons of the dentate gyrus. Neuroscience Letters 29:243 248. Torrealb a F, Muller C (1999) Ultrastructure of glutamate and GABA immunoreactive axon terminals of the rat nucleus tractus solitarius, with a note on infralimbic cortex afferents. Brain Research 820:20 30. Toth K, Suares G, Lawrence JJ, Philips Tansey E, McBain CJ (2000) Differential mechanisms of transmission at three types of mossy fiber synapse. Journal of Neuroscience 20:8279 8289. Trettel J, Levine ES (2003) Endocannabinoids mediate rapid retrograde signaling at interneuron right arrow pyramidal neuron synapse s of the neocortex. J Neurophysiol 89:2334 2338. 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. Twitchell W, Brown S, Mackie K (1997) Cannabinoids inhibit N and P/ Q type calcium channels in cultured rat hippocampal neurons. Journal of Neurophysiology 78:43 50. Ueda N, Kurahashi Y, Yamamoto S, Tokunaga T (1995) Partial purification and characterization of the porcine brain enzyme hydrolyzing and synthesizing anandami de. Journal of Biological Chemistry 270:23823 23827. Urban NN, Henze DA, Barrionuevo G (2001) Revisiting the role of the hippocampal mossy fiber synapse. Hippocampus 11:408 417. Valverde O, Karsak M, Zimmer A (2005) Analysis of the endocannabinoid system b y using CB1 cannabinoid receptor knockout mice. Handb Exp Pharmacol:117 145. Varma N, Carlson GC, Ledent C, Alger BE (2001) Metabotropic glutamate receptors drive the endocannabinoid system in hippocampus. J Neurosci 21:RC188. Vigot R, Barbieri S, Brauner Osborne H, Turecek R, Shigemoto R, Zhang YP, Lujan R, Jacobson LH, Biermann B, Fritschy JM, Vacher CM, Muller M, Sansig G, Guetg N, Cryan JF, Kaupmann K, Gassmann M, Oertner TG, Bettler B (2006) Differential compartmentalization and distinct functions of G ABA B receptor variants. Neuron 50:589 601. Vincent P, Marty A (1993) Neighboring cerebellar Purkinje cells communicate via retrograde inhibition of common presynaptic interneurons. Neuron 11:885 893.

PAGE 143

143 Vogt KE, Nicoll RA (1999) Glutamate and gamma aminobutyric acid mediate a heterosynaptic depression at mossy fiber synapses in the hippocampus. Proc Natl Acad Sci USA 96:1118 1122. Vogt KE, Regehr WG (2001) Cholinergic Modulation of Excitatory Synaptic Transmission in the CA3 Area of the Hippoca mpus. Journal of Neuroscience 21:75 83. Volk LJ, Daly CA, Huber KM (2006) Differential roles for group 1 mGluR subtypes in induction and expression of chemically induced hippocampal long term depression. J Neurophysiol 95:2427 2438. Walker MC, Ruiz A, Kull mann DM (2001) Monosynaptic GABAergic signaling from dentate to CA3 with a pharmacological and physiological profile typical of mossy fiber synapses. Neuron 29:703 715. Wall MJ, Usowicz MM (1997) Development of action potential dependent and independent sp ontaneous GABA A receptor mediated currents in granule cells of postnatal rat cerebellum. European Journal of Neuroscience 9:533 548. Wang Y, Okamoto M, Schmitz F, Hofmann K, Sudhof TC (1997) Rim is a putative Rab3 effector in regulating synaptic vesicle fu sion. Nature 388:593 598. Watabe AM, Carlisle HJ, O'Dell TJ (2002) Postsynaptic induction and presynaptic expression of group 1 mGluR dependent LTD in the hippocampal CA1 region. J Neurophysiol 87:1395 1403. Wenzel HJ, Buckmaster PS, Anderson NL, Wenzel ME Schwartzkroin PA (1997) Ultrastructural localization of neurotransmitter immunoreactivity in mossy cell axons and their synaptic targets in the rat dentate gyrus. Hippocampus 7:559 570. Williams S, Johnston D (1990) Muscarinic depression of synaptic tran smission at the hippocampal mossy fiber synapse. J Neurophysiol 64:1089 1097. 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 sp ecificity of endocannabinoid signaling in the hippocampus. Neuron 31:453 462. Wu LG, Saggau P (1995) GABA B receptor mediated presynaptic inhibition in guinea pig hippocampus is caused by reduction of presynaptic Ca 2+ influx. Journal of Physiology 485:649 657. Yamasaki M, Hashimoto K, Kano M (2006) Miniature Synaptic Events Elicited by Presynaptic Ca 2+ Rise Are Selectively Suppressed by Cannabinoid Receptor Activation in Cerebellar Purkinje Cells. Journal of Neuroscience 26:86 95.

PAGE 144

14 4 Yanovsky Y, Mades S, Misg eld U (2003) Retrograde signaling changes the venue of postsynaptic inhibition in rat substantia nigra. Neuroscience 122:317 328. Yeung JY, Canning KJ, Zhu G, Pennefather P, MacDonald JF, Orser BA (2003) Tonically activated GABA A receptors in hippocampal n eurons are high affinity, low conductance sensors for extracellular GABA. Molecular Pharmacology 63:2 8. Zappone CA, Sloviter RS (2004) Translamellar disinhibition in the rat hippocampal dentate gyrus after seizure induced degeneration of vulnerable hilar neurons. J Neurosci 24:853 864. Zhang F, Wang LP, Boyden ES, Deisseroth K (2006) Channelrhodopsin 2 and optical control of excitable cells. Nat Methods 3:785 792. Zhu PJ, Lovinger DM (2005) Retrograde endocannabinoid signaling in a postsynaptic neuron/syna ptic bouton preparation from basolateral amygdala. J Neurosci 25:6199 6207. Zimmer A, Zimmer AM, Hohmann AG, Herkenham M, Bonner TI (1999) Increased mortality, hypoactivity, and hypoalgesia in cannabinoid CB1 receptor knockout mice. Proc Natl Acad Sci USA 96:5780 5785. Zipp F, Nitsch R, Soriano E, Frotscher M (1989) Entorhinal fibers form synaptic contacts on parvalbumin immunoreactive neurons in the rat fascia dentata. Brain Res 495:161 166. Zola Morgan S, Squire LR, Amaral DG (1986) Human amnesia and the medial temporal region: enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus. Journal of Neuroscience 6:2950 2967.

PAGE 145

145 BIOGRAPHICAL SKETCH Though born in Jerusalem, Israel, Eugene, O regon Originally interested in mathematics, his high school biology teacher piqued his curiosity of neuroscience with only a single discussion on neurons. Undergraduate life took Ben to Harvey Mudd College in southern California where he recei Thanks to his experience as an undergraduate intern at the Whitney Marine Biology Laboratory in St. Augustine, F lorida Ben enrolled in the Interdisciplinary Program through the College of Medicine at th e University of Florida in the f all of 2004. In April 2005, he joined the laboratory of Dr. Charles J. Frazier studying the regulation of synaptic signaling. Over the course of his graduate career, he has published three peer reviewed academic researc h papers and is currently in the process of submitting a fourth. Following graduation, Ben plans to marry, move to Portland, Oregon, and begin work as a postdoctoral fellow in the laboratory of Dr. Craig Jahr at the Vollum Institute, a part of Oregon Heal th and Science University.