NORTHWESTERN UNIVERSITY Regulation and Function of Striatal

NORTHWESTERN UNIVERSITY

Regulation and Function of Striatal Nitric Oxide-Producing Interneurons

A DISSERTATION

SUBMITTED TO THE GRADUATE SCHOOL

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

for the degree

DOCTOR OF PHILOSOPHY

Field of Neuroscience

Alexandria Ellen Melendez-Zaidi

EVANSTON, ILLINOIS

December 2016

ABSTRACT

Regulation and Function of Striatal Nitric Oxide-Producing Interneurons

Alexandria Ellen Melendez-Zaidi

The striatum is a subcortical nucleus that regulates a number of complex activities ranging from voluntary action selection to the subconscious formation of habit. The coordination of these operations is mediated by the principle cells of the striatum, spiny projection neurons (SPNs). SPN activity is dictated by a confluence of cortical and thalamic glutamatergic inputs, their intrinsic excitability, and striatal microcircuits to achieve a desirable outcome.

Striatal microcircuitry is of particular interest as it has historically, been the most poorly described of these elements due to the inherent difficulty in studying a nucleus without a laminar organization. Consequently, our knowledge of the interneurons that compose these circuits is scant. The advent of transgenic mice and novel constructs allowing for the identification and selective excitation of these cells has provided the opportunity to attempt a comprehensive description of basic striatal physiology. Accordingly, the goal of this thesis work is to determine the function and regulation of one class of striatal interneuron known as the plateau and low- threshold spiking interneurons (PLTSIs). We used a combination of optogenetics, pharmacology, and calcium imaging to test the hypothesis that these cells mediate a form of synaptic plasticity in SPNs. To investigate the regulation of PLTSIs we combined mapping studies using a monosynaptic Rabies virus (RV), optogenetics, and pharmacology.

Glutamatergic inputs onto SPNs were found to undergo long term depression (LTD) when PLTSIs were optogenetically activated. One of many neuromodulators these cells synthesize is nitric oxide (NO). We found that this LTD was dependent on NO and occurred via a postsynaptic mechanism that relied on removal of AMPARs from postsynaptic synapses. 4

What’s more, this form of LTD was distinct from the canonically described endocannabinoid (eCB) LTD.

To move towards an understanding of what regulates NO release, we first asked what regulates PLTSI activity. RV mapping experiments revealed quantitatively similar input from cortical and thalamic nuclei whereas functional studies using optogenetics to selectively excite either region revealed differential responses in PLTSIs. Whereas cortical stimulation induced high-frequency spiking followed by a pause, thalamic activation halted PLTSI firing and was frequently followed by rebound bursting. The thalamic response was due to activation of a disynaptic circuit involving striatal cholinergic interneurons (ChIs) as the thalamic pause was sensitive to antagonists of muscarinic acetylcholine receptors (mAChRs); enhanced by inhibitors of ACh degrading enzymes; and mimicked by direct optogenetic activation of ChIs. The pause and pause-bursting responses were dependent on Gi-couple M4 receptors. Because both the ACh and NO systems are altered in many diseases (e.g. Parkinson’s disease, Levodopa-induced dyskinesia) our data suggests a potential role for PLTSIs and the PLTSI-ChI interaction in initiating or augmenting these states.

ACKNOWLEDGEMENTS

Foremost, I would like to express my gratitude to the funding sources during my time in lab. These were primarily the General Motor Control Mechanisms and Disease Training Program and the National Institute of Neurological Disorders and Stroke Grant NS034696. I am especially grateful for the existence of the former as it has provided me with the unique opportunity to meet and confer with a number of esteemed visiting scientists, to develop an extensive network of colleagues working on related issues and for the chance to present my work in public.

I would like to thank and acknowledge the tremendous amount of work done by the technicians in our lab –Marisha Alicea, Kang Chen, Yu Chen, Bonnie Erjavec, Danielle Schowalter and our lab manager Sasha Ulrich. Their work has made my work and the rest of the lab’s work that much easier to complete. I would like to especially thank Danielle for her friendship and offering a sympathetic ear anytime I needed to complain about experimental failures in particular or life in general. The work presented in chapter 2 was initiated by a former graduate student in the lab, Igor Rafalovich who co-authored and work; I am grateful for his friendship and scientific intelligence. Asami Tanimura, Josh Plotkin, and Shenyu Zhai contributed to data collection, analysis, and scientific discussion. What’s more, they have always made themselves available to me to help troubleshoot experiments or interpret data and for that I am ever grateful.

I have enjoyed being a part of the Surmeier lab for five years and this is in no small part thanks to the wonderful company of people I have been given the opportunity to work with. I am thankful that Wenjie Ren was my rig “neighbor”; she has made even the worst days of data collection enjoyable.

My thesis committee members – Lewis Landsberg, Jack Kessler, Gianmaria Maccaferri, and Richard Miller – have been encouraging and forgiving as I navigated between projects and for that I am so appreciative. My advisor, Jim Surmeier, has been a phenomenal PI and I am 6

honored that he – for whatever reason – allowed me to be a member of the lab. He has allowed me to pursue my own interests, and demonstrated great patience when progress was slow. I am grateful to Jim for allowing me to think independently but helping me when I needed it, and for being so enraptured with science that he understood when – sometimes – you just want to do an experiment because the result will be cool. Working under a first-rate scientist is something that will never leave me and I am truly thankful for that.

I would like to thank all of my family for being so supportive and positive, and for motivating me to keep going when times were hard: my brothers and sister, grandparents – particularly my late grandmother Jo Ellen – my mother- and father-in-law, Appa and Shaad. My mom and dad have always supported me to work hard but never pressured me to be anything other than a “good person”; I hope I have succeeded in the latter respect. I’m sorry for always forgetting to charge my phone.

Finally, my deepest love and gratitude to my greatest sources of happiness/support: my husband, Sehban, who has put up with me through both medical school and graduate school and listened to so many lab stories that he can hold a relatively intelligent conversation about “spiny neurons”; and my daughter “E” for reminding me how wonderful life is.

LIST OF ABBREVIATIONS

2-AG – 2-arachidonylglycerol 5-HT serotonin 5-HT2R – serotonin receptor 6-OHDA – 6-hydroxydopamine, neurotoxin compound that selectively kills dopaminergic and noradrenergic neurons 8-Br-cGMP – 8-Bromoguanosine 3’5’-cyclic monophosphate, cell permeable cGMP analog A2a – adenosine A2a receptor, selectively expressed by iSPNs, mainly in reference to A2a-Cre mice ACh – acetylcholine AChE – acetylcholinesterase ACSF – artificial cerebral spinal fluid AEA – anandamide AFDX-384 – M2/M4 selective antagonist AHP – afterhyperpolarization AM-251 – CB1R antagonist AMPAR – α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor (glutamate) BAC – bacterial artificial chromosome BAY 41-2272 – activator of sGC at an NO-independent regulatory site BK – big conductance Ca2+-activated K+ current Cav – voltage-gated Ca2+ channel Cav1(1.2, 1.3) – L-type Ca2+ channels Cav2/2 – N-type Ca2+ channels CB1R, CB2R – cannabinoid 1 and 2 receptors (Gi/o-coupled) CgCx – cingulate cortex CGP 55845 – selective blocker of GABAB receptors ChAT – choline acetyltransferase ChI – cholinergic interneuron ChR2 – channelrhodopsin CM – centromedian thalamic nucleus CM/Pf – centromedian/parafascicular complex CR – calretinin, in reference to a subpopulation of calretinin expressing striatal interneurons DA – dopamine (DAR, dopamine receptor) D1R – dopamine 1 receptor (dSPN expression) D2R – dopamine 2 receptor (iSPN expression) DIC – differential interference contrast DHPG – selective agonist of group I metabotropic glutamate receptors. DYN – dynorphin, expressed by dSPNs eCB – endocannabinoids; 2-Arachidonoylglycerol and Anandamide 8

eGFP – enhanced green fluorescent protein ENK – encephalin EPSC/P – excitatory post synaptic current/potential EP – entopeduncular nucleus (GPi, primates); projection site of dSPNs FSI – fast-spiking interneuron, express the parvalbumin Ca2+ binding protein GABA – γ-amino butyric acid GABAAR – ionotropic, ligand-gated receptor for GABA; considered inhibitory GABABR – metabotropic receptor for GABA Gabazine – selective blocker of GABAAR, synonymous with SR 95531 GC – guanylyl cyclase; may refer to soluble (sGC) or particulate (pGC); receptor of nitric oxide and atrial natriuretic peptides, respectively GIRK – G protein coupled inwardly-rectifying potassium channel GP – globus pallidus (GPe, primates); projection site of iSPNs GPe – external segment of the globus pallidus; projection site of iSPNs GPi – internal segment of the globus pallidus; projection site of dSPNs GPCR – G protein coupled receptor, class of metabotropic receptors HD – Huntington’s disease IPSC/P – inhibitory post-synaptic current/potential L-DOPA – 3,4-dihydroxy-L-phenylalanina, DA precursor used for symptomatic mgmt. of PD LID – Levodopa-induced dyskinesia L-NAME – NG-Nitro-L-arginine methyl ester hydrochloride, inhibitor of NO synthase LTD – long-term depression LTP – long-term potentiation mAChR – muscarinic acetylcholine receptor, metabotropic M1R – muscarine 1 receptor M4R – muscarine 4 receptor Mecamylamine hydrochloride – non-competitive nicotinic acetylcholine receptor antagonsit mGluR – metabotropic glutamate receptors MT-3 – muscarinic toxin 3, mamba toxin selective for M4 subtype of mAChRs MT-7 – muscarinic toxin 7, mamba toxin selective for M1 subtype of mAChRs NAc – nucleus accumbens; ventral striatum separated into the “core” and “shell” nAChR – nicotinic acetylcholine receptor; ionotropic receptors NADPH diaphorase – dihydronicotinamide adenine dinucleotide phosphate diaphorase; a marker for NOS (see below) NMDAR – N-methyl D-aspartate (NMDA) receptor (glutamate) nNOS – neuronal nitric oxide synthase NO – nitric oxide NO-LTD – nitric oxide induced long term depression nNOS – neuronal nitric oxide synthase, enzyme expressed in PLTSIs responsible for NO production NOS – nitric oxide synthase NPY – neuropeptide Y, peptide expressed by PLTSIs 9

PD – Parkinson’s disease PDE – phosphodiesterase, responsible for cyclic nucleotide degredation PFN – parafascicular thalamic nucleus PKA – protein kinase A (cAMP-dependent protein kinase) PKC – protein kinase C PKG – protein kinase G (cGMP-dependent protein kinase) PLTSI – plateau and low-threshold spiking interneuron PPN – pedunculopontine nucleus PSTH – peri-stimulus time histogram PV – parvalbumin, Ca2+ binding protein expressed by FSIs RV – rabies virus SC – superior colliculus sGC – soluble guanylyl cyclase, receptor of NO SK – small conductance Ca2+ activation K+ channels SKF 82958 – dopamine 1 receptor antagonist SNAP – S-nitroso-N-acetyl-DL-pennicillamine, nitrosothiol derivative that releases NO (NO donor) SNc – substantia nigra pars compacta SNr – substantia nigra pars reticulata; projection site of dSPNs SP – substance P, peptide expressed by dSPNs SPN – striatal projection neuron dSPN – direct pathway SPN, projecting to the iSPN – indirect pathway SPN, projecting to SST – somatostatin, peptide expressed by PLTS class of interneurons STN – subthalamic nucleus TAN – tonically active neuron, often believed to be cholinergic interneurons vGluT – vesicular glutamate transporter VTA – ventral tegmental area VU 0255035 – selective M1 antagonist XE-991 – selective blocker of KCNQ (Kv7) voltage-gated potassium channels, mediates M- current

TABLE OF CONTENTS

ABSTRACT ...... 3

ACKNOWLEDGEMENTS ...... 5

LIST OF ABBREVIATIONS ...... 7

TABLE OF CONTENTS ...... 10

LIST OF FIGURES ...... 11

Chapter 1 Introduction ...... 13

I. Striatal organization and anatomy ...... 16 II. Synaptic plasticity at the striatal synapse ...... 30 III. Pathophysiology in the striatum: a role for dopamine and acetylcholine ...... 39 IV. Dissertation aims and organization ...... 43 Chapter 3 Interneuronal nitric oxide signaling mediates post-synaptic long-term depression of striatal glutamatergic synapses ...... 45

Chapter 3 Cholinergic interneurons modulate PLTSIs through muscarinic signaling ...... 75

Chapter 6 Discussion ...... 114

References ...... 124

LIST OF FIGURES

FIGURE 1.1. THE ORGANIZATION OF BASAL GANGLIA CIRCUITS...... 19 FIGURE 1.2. MICROCIRCUITS WITHIN THE STRIATUM ...... 25 FIGURE 1.3. GLUTAMATERGIC SYNAPSES ONTO SPNS ...... 35 FIGURE 2.1. NO INDUCES LTD AT CORTICOSTRIATAL SYNAPSES THROUGH ACTIVATION OF PKG.51 FIGURE 2.2. NO-LTD IS INDUCED AT CORTICOSTRIATAL SYNAPSES BY SNAP OR 8-BR-CGMP. .. 54 FIGURE 2.3. NO PRODUCING INTERNEURONS CAN INDUCE NO-LTD...... 57 FIGURE 2.4. TRANSFECTED SST-CRE MICE SHOW NO OFF-TARGET LABELING...... 59 FIGURE 2.5. LOCAL INTERNEURONS GENERATE NO-LTD IN A SELECT POPULATION OF SPNS...... 61 FIGURE 2.6. NO-LTD IS INDEPENDENT OF CONVENTIONAL DETERMINANTS OF ECB-LTD...... 64 FIGURE 2.7. NO-LTD IS POST-SYNAPTICALLY EXPRESSED...... 66 FIGURE 2.8. NO-LTD IS AMPLIFIED BY BAY-41...... 70 FIGURE 2.9. NO SIGNALING OCCLUDES ECB-LTD THROUGH L-TYPE CURRENT INHIBITION...... 72 FIGURE 3.1. RETROGRADE LABELING OF MONOSYNAPTIC INPUTS TO STRIATAL SST+ INTERNEURONS...... 90 FIGURE 3.2. SST-TDTOMATO CELLS HAVE CHARACTERISTICS OF PLTSIS...... 92 FIGURE 3.3. SST X AI14 MICE DEMONSTRATE PROMISCUOUS LABELING OF STRIATAL INTERNEURONS...... 94 FIGURE 3.4. CORTICAL AND THALAMIC STIMULATION INDUCE DIFFERENTIAL RESPONSES IN PLTSIS ACTIVITY...... 96 FIGURE 3.5. THALAMIC ACTIVATION INTERRUPTS PLTSI ACTIVITY VIA ACTIVATION OF CHOLINERGIC INTERNEURONS...... 98 FIGURE 3.6. CORTICAL ACTIVATION ALSO ENGAGES CHOLINERGIC INTERNEURONS TO MODULATE PLTSIS...... 100 FIGURE 3.7. SYNAPTIC PROPERTIES OF CGCX AND PFN INPUTS ONTO PLTSIS...... 102 FIGURE 3.8. INTRASTRIATAL RV+ CELLS ARE CHAT NEGATIVE...... 104 FIGURE 3.9. CHI ACTIVATION REPRODUCES THE PFN-INDUCED PAUSE IN PLTSI FIRING...... 106 12

FIGURE 3.10. STRONG MACHR STIMULATION RESULTS IN BURSTING IN PLTSIS...... 108 FIGURE 3.11. PLTSIS IDENTIFIED FROM SST-TDT+ OR NPY-EGFP MICE RESPOND SIMILARLY TO MUSCARINE...... 110 FIGURE 3.12. PLTSI BURSTING IS DEPENDENT ON M4 ACTIVATION...... 112

Chapter 1

Introduction

Thomas Jefferson once said the following:

“Do you want to know who you are? Don’t ask. Act!

Action will delineate and define you.”

For an aspiring physiologist whose studies are focused on the basal ganglia - that part of the brain in which thought and action are clearly and inextricably linked - this quotation speaks profoundly to the intrinsic nature of my work. For it is in the basal ganglia that information is turned into action as both "higher" and "lower" level structures, from cerebral cortex to brainstem nuclei, synapse and have those synaptic messages transformed into an output that can both facilitate movement and influence executive functions (DeLong 1990, Hart, Leung et al. 2014). As such, the basal ganglia are as equally engaged in the brain of an infant performing simple swaying movements to music as it is in the brain of a ballerina executing hundreds of synchronized complex movements, or in the learned associations made by students; one who forgoes studying for an exam for the pleasure of playing video games, and the other who fervently studies in anticipation of good marks on the exam.

And while our actions are significantly dependent on “higher order” structures such as the cerebral cortex in which “thought” is believed to originate, these thoughts are not exempt from the immense processing that must first occur within nuclei of the basal ganglia before that thought or idea is executed. 14

The basal ganglia, and its individual members, have been describe in a variety of ways since they were first observed over 19 centuries ago by Claudius Galenus (Parent 2012). One of the most comprehensive descriptions came from the 17th century English physician Thomas Willis, who was the first to provide a detailed account of the largest member of the basal ganglia, the corpus striatum (Rose 2010, Parent 2012). Although the largest member of the basal ganglia, it went unstudied for nearly two hundred years, until it was flung back into the realm of research by a series of experiments that revealed the onset of robust motor disorders, following lesions of the nucleus (Gerfen and Bolam 2010). Subsequently, a half a century later, striatal dysfunction was linked to the “shaking palsy” first formally described by James Parkinson in the early 1800’s; an observation that significantly contributed to the motivation for researching the function of this nucleus in both models of disease and health (Goetz 2011). Since that time, numerous studies have been conducted which have revealed the complexity of this region of the brain and how integral intact functioning of striatal physiology is in the performance of menial as well as complex tasks (Nelson and Kreitzer 2014, Surmeier, Graves et al. 2014).

Among the brain structures that have been closely studied, the striatum is unique in that it lacks any obvious organization that can be appreciated by the human eye; this is in stark contrast to other brain structures such as the motor cortex, hippocampus, or cerebellum (Gerfen 1989, Gerfen 1992). Therefore, it has been difficult to characterize the subtypes of striatal projection neurons (SPNs), not to mention the numerous populations of interneurons that exist within the nucleus (Tepper, Tecuapetla et al. 2010). This “disorganization” makes it nearly impossible to use traditional patch clamp methods to study the striatum and, consequently, our understanding of the microcircuits which exist in this nucleus is relatively weak (Kravitz, Owen et al. 2013). Fortunately, in only the last two decades have molecular advances and the use of transgenic mice permitted the separation of SPN populations from interneurons and, in this latter category, fast-spiking from tyrosine hydroxylase-expressing from neuropeptide Y-expressing interneurons (Matsushita, Okada et al. 2002, Chattopadhyaya, Di Cristo et al. 2004, Heintz 2004, Gong, Doughty et al. 2007, van den Pol, Yao et al. 2009, Madisen, Zwingman et al. 2010, Taniguchi, He et al. 2011). Additionally, the advent of optogenetics has played a significant role 15

in the resolution of issues associated with selective excitation of discrete cell populations and therefore, has been substantially responsible for a number of recent publications revealing complex interactions between many cell types and their associated transmitters (e.g. GABA, acetylcholine, and dopamine) within the striatum (Boyden, Zhang et al. 2005, Higley, Gittis et al. 2011, Threlfell, Lalic et al. 2012, Nelson, Hammack et al. 2014, Kim, Ganesan et al. 2015, Saunders, Granger et al. 2015).

A more distinct field of study, that has not been as hindered by anatomical constraints is that of striatal plasticity. Both long-term depression (LTD) and long-term potentiation (LTP) have been described to occur in SPNs (Calabresi, Maj et al. 1992, Kreitzer and Malenka 2005, Shen, Flajolet et al. 2008) . While dopamine (DA) and endocannabinoids (eCB) are, without question, the most critical actors in cellular learning, a role for other molecules is clearly apparent (Mathur, Capik et al. 2011, Mathur and Lovinger 2012, Atwood, Kupferschmidt et al. 2014, Pancani, Bolarinwa et al. 2014). One of the most contentious of these molecules is nitric oxide (NO), which is expressed in a small, somewhat mysterious subpopulation of interneurons (Kawaguchi 1993, Kishimoto, Keverne et al. 1993, Gracy and Pickel 1997). NO has been implicated in both LTD and LTP with conflicting results; however, it is speculated that different preparations may have contributed to the discordance in the literature (Calabresi, Gubellini et al. 1999, Centonze, Grande et al. 2003, Sammut, Park et al. 2007). Further complicating these observations are related publications demonstrating that NO and DA can influence the other’s release and that either may alter the intracellular signaling cascade initiated by the other, within SPNs (Shibata, Araki et al. 1996, Hartung, Threlfell et al. 2011). With this in mind, it comes as no surprise that NO signaling is altered in dopaminergic disorders such as Parkinson’s disease (PD) and levodopa-induced dyskinesia (LID); however, it is unknown how these changes are manifested. (Eve, Nisbet et al. 1998, Dehorter, Guigoni et al. 2009, Padovan-Neto, Echeverry et al. 2009, Picconi, Bagetta et al. 2011, Solis, Espadas et al. 2015).

The ensuing studies utilize a combination of contemporary techniques, optogenetic and Rabies viruses, and transgenic mice to address an often referenced but ill-defined topic in striatal physiology: what is the role of the striatal NO-releasing interneurons? How is their activity 16

(and, theoretically, NO release) regulated? And what are the consequences of NO release on SPN activity?

I. Striatal organization and anatomy Efferent inputs modulate striatal output

Although the basal ganglia include the pallidum, substantia nigra, and subthalamic nucleus, it is challenging to discuss these three nuclei without mention of the striatum. This is primarily because the striatum, serving as the input structure to the basal ganglia, acts as a selective colander; heavily processing the glutamatergic afferents impinging on the heavily spiny dendrites of SPNs (Albin, Young et al. 1989, Graybiel 2000, Haber, Kim et al. 2006, Balleine, Delgado et al. 2007). SPNs receive excitatory inputs from nearly every cortical subdivision along with thalamic nuclei; inputs that carry information to direct tasks as distinct as the subdivisions themselves (Kemp and Powell 1970, Jones, Coulter et al. 1977, McGeorge and Faull 1989). Consequently, the striatum has been shown to participate in a plethora of processes which include the execution of voluntary motor movements, sensorimotor learning, the formation of habits, and the assigning of salience to various stimuli (Zink, Pagnoni et al. 2003, Yin and Knowlton 2006, Balleine, Delgado et al. 2007).

What seemingly unites these processes is that they all involve some form of learning. Given that the striatum receives a dense dopaminergic innervation from the substantia nigra pars compacta (SNc) and the ventral tegmental area (VTA) is not surprising: indeed it is theorized that DA from these structures aids in assigning importance to the impinging glutamatergic inputs (Ilango, Kesner et al. 2014, Lammel, Lim et al. 2014, Howe and Dombeck 2016). So vast is this DA innervation, that the axonal arborization of individual DA neurons has been shown to cover as much as 5.7% of the striatal volume (Matsuda, Furuta et al. 2009). Contrast this dense, unrestrained innervation to that of the cortex where it is estimated that, despite the fact that corticostriatal neurons outnumber SPNs by a ratio of 6:1, an individual corticostriatal neuron synapses on only 1% of SPNs in an area in which it’s axon extends (Oorschot 1996). Therefore, 17

each SPN receives inputs from many corticostriatal neurons and the corticostriatal neurons that innervate each SPN must be completely separate and distinct (Cowan and Wilson 1994). These inputs are almost exclusively directed to the spine heads of SPNs – it is estimated that the fraction is well over 90% – and are posited to then be modulated within the striatum by dopaminergic terminals that abut spine necks (Somogyi, Bolam et al. 1981, Bouyer, Park et al. 1984, Moss and Bolam 2008). These observations support theories that the striatum is a structure charged with processing and interpreting glutamatergic inputs, rather than passively relaying their message to basal ganglia output structures (Kincaid, Zheng et al. 1998, Reig and Silberberg 2014).

The subject of striatal glutamatergic inputs cannot be fully addressed without a considerable discussion of the thalamus. While the striatum receives spattered innervation from the intralaminar, ventral anterior, ventral lateral midline nuclear group, and other thalamic nuclei; the centromedian/parafascicular (CM/Pf) complex within the intralaminar subdivision offers the thalamus’ main contribution (McFarland and Haber 2000, Smith, Raju et al. 2004, Wall, De La Parra et al. 2013, Smith, Galvan et al. 2014). In contrast to the cortex, thalamostriatal afferents almost exclusively target dendritic shafts, rather than the spine heads (Sidibe and Smith 1996, Sidibe and Smith 1999). The CM/Pf inputs are unique in that, unlike cortical inputs which are thought to deliver a steady stream of almost excessive information regarding internal and external states, the thalamic inputs are only heavily activated during salient stimuli (Aosaki, Graybiel et al. 1994, Ding, Guzman et al. 2010). Therefore, it is hypothesized that the thalamus is charged with redirecting the “attention” of SPNs so as to optimize action selection and ultimately, survival (Nanda, Galvan et al. 2009, Smith, Raju et al. 2009) This hypothesis is supported by studies that have revealed high release probabilities and profound short term depression at thalamostriatal synapses as compared to corticostriatal synapses (Ding, Peterson et al. 2008). This suggests that these inputs are primed to avert attention but then fade away, allowing for cortical inputs to dominate, as is usual, once the potentially dangerous or novel event subsides. The advent of optogenetics has allowed for even further separation of the CM and Pf nuclei which has revealed that Pf inputs are ‘smaller’ but 18

undergo plasticity whereas CM inputs do not (Ellender, Harwood et al. 2013). This observation is especially interesting when we consider a series of work in which CM/Pf inputs to the striatum were seen to decrease in Parkinson’s disease (PD), concomitant with decreased staining for vGluT2 terminals (Raju, Ahern et al. 2008, Parker, Lalive et al. 2016). Whether a difference exists between CM and Pf in this dopamine depleted stain remains to be seen, as does how the inputs are altered by systemic administration of levodopa (L-Dopa), a common treatment for PD.

Striatal projection neurons

The principle cell of the striatum is the striatal projection neuron (SPN) a GABA-ergic inhibitory cell representing 90-95% of neurons within the human striatum (Oorschot 1996). SPNs can be divided into two classes which are separated by a number of properties to include their axonal projection targets, peptide gene expression profiles and the expression of different subclasses of dopamine receptor (DAR) (Gerfen and Young 1988, Gerfen, Engber et al. 1990, Le Moine and Bloch 1995). Specifically, SPNs whose axons project to the substantia nigra pars reticulata (SNr) and entopeduncular nucleus (EP) also express the peptides substance P (SP) and dynorphin (dyn) and the D1 DAR; whereas SPNs that project to the globus pallidus (GP)

exclusively express encephalin (enk) and D2 DAR (Steiner and Gerfen 1998, Gerfen and Surmeier 2011). These pathways are the so-called “direct” and “indirect” pathways long thought to mediate movement and the cessation of movement, respectively (DeLong 1990, DeLong and Wichmann 2009, Freeze, Kravitz et al. 2013). While this long-held dogma has been upended by recent in vivo studies demonstrating both pathways are concomitantly active during a motor task, it still remains that the two output divisions represent one level of dissimilarity (Cui, Jun et al. 2013).

More recent studies have utilized the differential expression of DARs to generate transgenic mice, which has allowed for the identification of the two populations in silico (Heintz 2004, Gong, Doughty et al. 2007). 19

Figure 1.1. The organization of basal ganglia circuits. (A) Outline of the three, primary cortico-basal ganglia circuits that include the limbic, associative, and motor circuits. The limbic and paralimbic cortices (e.g. cingulate cortex, CgCX; piriform cortex, PRC), hippocampus and amygdala innervates the ventral striatum (which includes the Nucleus Accumbens, NAc) which projections to the ventral pallidum (limbic portions of GPe, ventral GPi). Associative cortices (e.g. dorsolateral prefrontal, orbitofrontal) project to the dorsallateral striatum which comprises the caudate nucleus and the anteromedial portion of the putamen. The motor circuit connects sensorimotor cortex to the dorsolateral striatum which is primarily composed of the posterolateral putamen. (B) A semi-anatomical model of dorsolateral circuits. Direct- and indirect-pathway striatal projection neurons (dSPNs, iSPNs) send GABAergic projections to various basal ganglia output nuclei. dSPNs primarily innervate two inhibitory nuclei: the internal segment of the globus pallidus (GPi) and substantia nigra reticulata (SNr) and give off sparse axon collaterals to the external segment of the globus pallidus (GPe). iSPNs primarily innervate the GPe which send their inhibitory inputs to the subthalamic nucleus (STN). The STN sends excitatory afferents back to GPi and SNr. 20

Inspired by Gremel and Lovinger, 2016; Genes Brain Behav. (B) Adapted from Gerfen and Surmeier, 2011; Annu Rev Neuroscience.

These studies have set the groundwork for foundational papers in the fields of striatal and

basal ganglia research by demonstrating that the D1R-expressing SPNs (dSPNs because of their

“direct” projection) and D2R-expressing SPNs (iSPNs because of their “indirect” projections) are physiologically distinct (Day, Wokosin et al. 2008, Gertler, Chan et al. 2008). Additionally, these studies demonstrated a differential dendritic anatomy between the two subclasses, which resulted in differences in cell excitability. Further, it was demonstrated that dSPNs have a larger dendritic arborization as compared to iSPNs, which was causative of the enhanced excitability in iSPNs as compared to dSPNs. This observation not only furthered our understanding of basic basal ganglia organization and physiology, but also laid the basis for further studies that demonstrated a differential response of the two cell types to disease (Day, Wang et al. 2006, Andre, Cepeda et al. 2011, Fieblinger, Graves et al. 2014).

Interneuron classes

A HISTORY OF STRIATAL INTERNEURONS

While a majority of the cells found in the striatum are SPNs, 3-10% of the estimated ~6 million that make up the nucleus, at least 180,000, are interneurons (Oorschot 1996). Striatal GABAergic interneurons (INs) are similar to cortical interneurons in many respects. Striatal INs are primarily derived from an embryonic source – the medial ganglionic eminence (MGE) – that is distinct from that of projection neurons which are derived from the lateral ganglionic eminence (LGE) (Marin, Andersen et al. 2000). IN subtypes that exist in the cortex largely exist in the striatum as well and serve similar functions in both. The parvalbumin (PV) expressing fast- spiking interneurons (FSIs) are believed to target the soma and proximal dendrites of projection neurons of both nuclei; neuropeptide Y (NPY) expressing neurogliaform cells (NPY-NGF) function similarly as FSIs in both cortex and striatum; somatostatin (SST) expressing, plateau and low-threshold hold spiking cells (PLTSIs) show sparse connectivity with the distal dendrites of pyramidal cells along with SPNs; and calretinin (CR) expressing interneurons are present in 22

both regions but remain largely uncharacterized (Di Cristo, Wu et al. 2004, Wonders and Anderson 2006, Tepper, Tecuapetla et al. 2010, Tremblay, Lee et al. 2016).

One huge difference between cortical and striatal GABAergic interneurons and the circuits in which they participate is that very little is known about the latter when compared to the former. This can be attributed to the fact that the striatum lacks the laminar organization that exists in the cortex which makes it impossible, through traditional electrophysiological methods, to selectively activate subtypes of interneurons known to innervate different regions of projection neurons (Gerfen 1989, Gerfen 1992). Here again, optogenetics and transgenic mice have aided in the study of this field. Participants in the GENSAT project were the first to nearly simultaneously generate a large number of bacterial artificial chromosome (BAC)–eGFP reporter or BAC-Cre driver lines to facilitate the identification of subclasses of neurons (Heintz 2004). Some of the first papers utilizing a reporter to identify striatal interneurons employed either the Lhx6-eGFP mice that selectively expressed eGFP in SOM-, PV-, CR- and NPY-expressing striatal interneurons or the NPY-eGFP mice in which both PLTSI and NPY-NGF cells appear labeled (Partridge, Janssen et al. 2009, Gittis, Nelson et al. 2010). These mice have proved invaluable in the study of intrinsic physiological properties of these cells. Josh Huang’s group is credited with another substantial assist when they generated 20+ Cre and inducible CreER knockin lines to further the study of GABAergic interneuron subclasses (Taniguchi, He et al. 2011). These lines have been paired with viral delivery of Cre-dependent opsins, modified rabies virus, and pharmacogenomics tools to further our understanding of complex circuitry (Boyden, Zhang et al. 2005, Wickersham, Lyon et al. 2007, Dong, Allen et al. 2010, Magnus, Lee et al. 2011).

GIANT ASPINY CHOLINERGIC INTERNEURONS

The difficulty in studying the physiology of striatal GABAergic interneurons was limited not only by the seeming disorganization of the nucleus, but also because of the similarity in appearance between SPNs and these subclasses when viewed through differential interference contrast (DIC) microscopy (Tepper, Tecuapetla et al. 2010). Cholinergic interneurons (ChIs) 23

escaped this problem because their large somatic size of 20-50 µm allows for easy identification despite their sparsity (Ma, Feng et al. 2014). ChI-focused publications dominate studies of all other GABAergic interneurons combined – over 4x over – as can be evidenced by any pubmed search.

Traditionally, ChIs have not been considered “GABAergic interneurons” since they were long considered to exclusively release the neurotransmitter acetylcholine (Ach). However, in recent years ChIs have been shown to have the capacity to co-release glutamate along with ACh and to indirectly release GABA through excitation of presynaptic nicotinic receptors on dopaminergic axons that co-release the inhibitory neurotransmitter along with DA (Higley, Gittis et al. 2011, Threlfell, Lalic et al. 2012). Cholinergic neurons of the basal forebrain have been demonstrated to co-release ACh and GABA, but this seems not to be the case in the striatum (Saunders, Granger et al. 2015). The physiological circumstances in which co-release, or indirect release of GABA, are important are not clear so the focus of this section will be on ACh.

The striatum has some of the highest Ach levels in the brain, which underscores the importance of ChIs in striatal physiology (Mesulam, Mash et al. 1992, Contant, Umbriaco et al. 1996). These cells are autonomously active neurons, firing between 3-10 Hz to produce a tonic level of ACh (Bolam, Wainer et al. 1984, Tepper, Koos et al. 2004). As a result, the phrase “tonically active neuron” (TAN) has become synonymous with ChIs, perhaps mistakenly so (see below section of Plateau and low-threshold spiking interneurons). ChIs have been found both in vivo and in slice to oscillate between three distinct firing states: regular, irregular, and bursting, and are believed to release different amounts of neurotransmitter during these different states of firing (Bennett and Wilson 1999, Bennett, Callaway et al. 2000). It should also be noted that these cells receive an especially robust thalamic input that was thought to preferentially innervate ChIs (as compared to cortical inputs) though recent work has challenged this thought (Ding, Guzman et al. 2010, Kosillo, Zhang et al. 2016). In awake behaving animals, stimulation of this input has been shown to generate a signature response in TANs: a brief period of excitation (burst) followed by a prolonged period of quiescence or diminished activity (pause) and resumption of normal firing (though some cells show the opposite response pattern; pause → 24

rebound burst → normal firing) (Matsumoto, Minamimoto et al. 2001, Doig, Magill et al. 2014). This burst-pause pattern is indistinguishable from the recorded TAN response in awake behaving primates that are presented with a salient stimulus (Aosaki, Graybiel et al. 1994, Aosaki, Kimura et al. 1995). The significance of this response is not wholly recognized but is undoubtedly tied to learning and/or disease since the response dissipates as the stimulus loses its novelty or if animals are treated with 1-methyl-4-phyl-1,2,3,6-tetrahydropyridine (MPTP), causing loss of dopaminergic neurons to mimic a Parkinsonian state (Aosaki, Graybiel et al. 1994).

PLATEAU AND LOW THRESHOLD SPIKING INTERNEURONS

While ChIs are undoubtedly TANs, recent studies demonstrating the existence of a second population of striatal pace making cells inspired a re-interpretation of these classic in vivo studies (Ibanez-Sandoval, Tecuapetla et al. 2011, Beatty, Sullivan et al. 2012, Elghaba, Vautrelle et al. 2016). This second class is comprised of the plateau and low-threshold spiking interneurons (PLTSIs) which express the most neurochemicals of any cell in this nucleus: SST, NPY, neuronal nitric oxide synthase (nNOS), and GABA. PLTSIs show extremely low connectivity – approximately 7% – with SPNs when this connection is interrogated using traditional physiological methods such as dual whole-cell recordings between the two cell types (Gittis, Nelson et al. 2010). This connectivity increased when Alexa 594 was included in the PLTSI targeted patch pipette so as to preferentially record from SPNs within the arborization of PLTSIs but even so the 14% reported was vastly lower than the connectivity percentage between FSIs and SPNs (45%), NPY-NGF interneurons and SPNs (86% connectivity) or tyrosine- hydroxylase interneurons (THINs) and SPNs (40-100%) without the aid of dye in these cases (Koos and Tepper 1999, Czubayko and Plenz 2002, Koos and Tepper 2002, Taverna, Ilijic et al. 2008, Gittis, Nelson et al. 2010).

The paltry GABAergic inputs from PLTSIs onto SPNs have led many to speculate that the primary action of this class of GABAergic interneurons may not be inhibition at all, but neuromodulation (Beatty, Sullivan et al. 2012). 25

Figure 1.2. Microcircuits within the striatum Microcircuits within the striatum are composed of SPNS as well as a variety of striatal interneurons. dSPNs and iSPNs receive innervation from: (A) Cholinergic interneurons (ChIs) that results in modulation of muscarinic acetylcholine receptors (mAChRs) on both cell types. dSPNs express M1 and M4 receptors whereas iSPNs express M1. ChI activation causes a robust change in PLTSI firing that is dependent on mAChRs (Chapter 3). (B) Plateau and low-threshold spiking interneurons (PLTSIs) provide a sparse, GABAergic input onto SPNs and also modulate plasticity via nitric oxide (NO, Chapter 2). (C) ChIs show a robust response to NO signaling as well, suggesting they likely receive innervation from these cells. (D) ChIs also provide strong innervation to various classes of striatal interneurons expressing both nAChRs and mAChRs. The effect of this innervation is to cause action potential (AP) generation via nAChRs and/or release of GABA through activation of nAChRs present on presynaptic terminals. mAChR activation has also been shown to reduce GABA release at the FSISPN synapse. (E) The interneurons contained in the shaded box – Tyrosine-hydroxylase (TH) expressing interneurons (THINs), fast- spiking interneurons (FSIs), and neuropeptide Y (NPY) expressing neurogliaform (NPY-NGF) cells – have been demonstrated to provide strong, inhibitory inputs onto SPNs and are members of the canonical “feed-forward” circuitry. 26

This would not be surprising given the cells only weakly stain for GABA, even in the presence of colchicine (Tepper, Tecuapetla et al. 2010). As PLTSIs are the exclusive source of striatal NO, as well as SST and one of two sources of NPY, their activity may instead reflect global levels of any of these molecules (Duchemin, Boily et al. 2012, Liguz-Lecznar, Urban- Ciecko et al. 2016). It is well appreciated that NO can modulate striatal afferents and – it is debated – intrinsic excitability (see section II). SST and NPY have both been shown to modulate GABAergic synaptic transmission, SPN excitability, and corticostriatal synaptic transmission (Galarraga, Vilchis et al. 2007, Lopez-Huerta, Blanco-Hernandez et al. 2012) (Melendez, unpublished observations). Defining or redefining the role of PLTSIs remains an important question in striatal physiology; as of right now it is a cell without a circuit or function (English, Ibanez-Sandoval et al. 2012, Szydlowski, Pollak Dorocic et al. 2013).

FAST-SPIKING INTERNEURONS

Striatal FSIs are medium-sized, parvalbumin (PV) expressing cells with a high connectivity between themselves and SPNs, and the densest axonal arborization of any other striatal neuron, approaching 1.5-2x it’s dendritic field (Kawaguchi 1993, Tepper and Bolam 2004). FSIs demonstrate some unique electrophysiological signatures which include extremely fast firing rates (upwards of 400 Hz) and a stuttering firing pattern in response to large current injections: though this phenotype is likely species specific as the existence of extensive numbers of FSIs that exhibit stuttering in mice is suggested far less than those stuttering cells found in rats (Koos and Tepper 2002, Bracci, Centonze et al. 2003, Plotkin, Wu et al. 2005, Sciamanna, Ponterio et al. 2015). FSIs, like SPNs, receive innervation from both cortex and thalamus, as well as SNc and ChIs, as suggested by ultrastructural and functional data (Kubota, Inagaki et al. 1987, Wilson, Chang et al. 1990, Ramanathan, Hanley et al. 2002). Pharmacological studies revealed a role for DA and Ach as application of both D1/5 agonists and D2 antagonists have been shown to directly and indirectly modulate FSI firing, whereas their firing is increased in response to carbachol, while simultaneously decreasing their release of GABA via postsynaptic Type-1 nicotinic and muscarinic receptors, respectively (Bracci, Centonze et al. 2002, Koos and Tepper 2002, Wiltschko, Pettibone et al. 2010). 28

NPY-NEUROGLIAFORM, TYROSINE-HYDROXYLASE-EXPRESSING, AND CALRETININ INTERNEURONS

NPY-NGF interneurons were characterized after their NPY-expressing counterpart the PLTSIs, but have proved to have a more robust, or readily seen, role in SPN inhibition. Dual patch clamp recordings from NPY-NGF interneurons and SPNs produces large amplitude inhibitory post-synaptic currents (IPSC) in the latter that is capable of shutting down action potential (AP) generation (Ibanez-Sandoval, Tecuapetla et al. 2011, English, Ibanez-Sandoval et al. 2012). These cells have also been shown to participate in novel intrastriatal circuits between ChIs and SPNs; specifically, NPY-NGF cells offer the non-GABAergic ChIs a means of inhibiting SPNs through excitation of Type I class nicotinic receptors on NPY-NGF interneurons and subsequent inhibition of SPNs (English, Ibanez-Sandoval et al. 2012).

THINs, like NPY-NGF cells, were not originally described in the early, seminal reviews on striatal interneurons, though they were noted almost half a decade before in the monkey caudate nucleus (Dubach, Schmidt et al. 1987, Kawaguchi 1993). Still, the presence and relative abundance of these cells was debated until recently when a transgenic mouse in which eGFP expression is driven by the tyrosine hydroxylase (TH) promoter led to a series of studies demonstrating that THINs were, in fact, comparable in number to other interneuron classes, are GABAergic, do not release DA despite their expression of TH and are highly connected to not only SPNs but other interneuron classes (Matsushita, Okada et al. 2002, Ibanez-Sandoval, Tecuapetla et al. 2010, Xenias, Ibanez-Sandoval et al. 2015). These cells are notable for increasing in number in DA depleted states and housing four subpopulations each with distinct membrane properties, connectivity to SPNs and anatomy (Tashiro, Sugimoto et al. 1989, Unal, Shah et al. 2015).

CR interneurons have been described within the striatum since the earliest studies, but little continues to be known about this cell population as there is not yet a transgenic mouse to identify this population in acute brain slices (Tepper, Koos et al. 2004, Tepper, Tecuapetla et al. 2010). As such, their physiology remains a mystery, as does their connectivity to other SPNs 29

and/or interneurons within the striatum. In both human and non-human primates, these cells far outnumber the PV and SOM/NPY interneurons by as much as 3-4 to 1, and they are the sole cell for which neurogenesis has been reported in response to hypoxia (Wu and Parent 2000). Both of these observations suggests that the cells are critical members of striatal circuitry but how they fit in remains to be determined.

Striatal output pathways

Although interneurons may exert a profound modulation on SPNs, it is the activity of the projections cells that dictates the actions of the basal ganglia (Cui, Jun et al. 2013). As was discussed above, striatal output is separated into one of two pathways which are defined by the SPN subtype that make them up. The indirect pathway, composed of axons from D2R expressing SPNs, extends to the GP (in rodents the GPe) while the direct pathway, made up of D1R expressing axons, primarily innervates the EP (GPi in rodents) and substantia nigra (SN) but also throws off axon collaterals to the GP/GPe (Gerfen, Engber et al. 1990, Kawaguchi, Wilson et al. 1990). Despite the involvement of the GP/GPe in processing striatal output messages from iSPNs, and modulating striatal function through their “ascending” reciprocal connections back onto SPNs, the primary effecter systems activated by the striatum are the SN and EP (Gerfen and Bolam 2010).

The SN and EP both send projections to the pedunculopontine nucleus (PPN), superior colliculus (SC), and thalamus (Gerfen, Staines et al. 1982, Chevalier, Vacher et al. 1985, Kita and Kitai 1987). With respect to the latter nucleus, SN inputs are subdivided into: 1.) those that innervate intralaminar nuclei (e.g. CM/Pf) which synapse back on the striatum and 2.) thalamic nuclei that innervate the cortex (Gerfen and Bolam 2010)It is of interest to note that divisions of the thalamus that are targeted by SN axons vary between rodents and primates owing to more discrete cortical divisions in the latter species (Gerfen and Bolam 2010). For example, the ventromedial and paralaminar medial dorsal nuclei are the primary thalamic targets in rodents whereas the ventral anterior and paralaminar medial dorsal nuclei predominate in primates (Ilinsky, Jouandet et al. 1985). 30

The basal ganglia outputs to the PPN are of particular interest with respect to pathology. Like the basal ganglia, the PPN is implicated in learning and locomotion but is also thought to generate REM sleep (Tsang, Hamani et al. 2010). The appearance of REM behavioral disorder – a condition in which sleep paralysis is impaired – is a common, early symptom that is present in many patients with Parkinson’s disease (Langston 2006). The leap between faulty basal ganglia outputs, due to dopaminergic denervation, and faulty activation of a targeted structure causing pathology is not a difficult leap to make. The job of the superior colliculus rotates between different stages of eye movement. This nucleus is involved in fixed movement, smooth pursuit, convergence and saccades (Gandhi and Katnani 2011). Like the PPN, abnormal eye movements have long been observed in PD patients in the form of an exaggerated glabellar reflex which is hypothesized to occur via abnormal afferents impinging on the SC (Basso, Powers et al. 1996, Bologna, Agostino et al. 2009).

II. Synaptic plasticity at the striatal synapse It is often stated that synaptic plasticity is the cellular correlate of learning; the physical means by which this “learning” occurs is through the strengthening or depression of synapses via mechanisms at both the pre and postsynaptic bouton. However, the phrase “synaptic plasticity” is actually rather unspecific: it may refer to short-term (seconds) or long-term (hours-days) modulation of synaptic connection at both excitatory and inhibitory synapses (Sjostrom, Rancz et al. 2008). Although short-term plasticity (STP) occurs within the striatum, this section will focus on long term changes akin to what Bliss and Lomo described over 50 years ago at the hippocampal performant patch synapse (Bliss and Lomo 1970). Additionally, we will limit the study of plasticity to glutamatergic synapses with a heavy focus on long term depression or long term potentiation. In recent years, GABAergic plasticity in the striatum has gained popularity and a number of papers investigating the role of endocannabinoids, dopamine, and SST in mediating these phenomena have been produced and should be read by any party interested in this matter (Centonze, Saulle et al. 2001, Centonze, Battista et al. 2004, Adermark, Talani et al. 31

2009, Lopez-Huerta, Blanco-Hernandez et al. 2012, Nieto Mendoza and Hernandez Echeagaray 2015). The mechanisms that dictate GABAergic plasticity are similar to glutamatergic plasticity, but diverge just enough that they would require too much of a departure from the focus of this dissertation.

Dopamine

No study of the striatum, whether focused on synaptic plasticity or not, can be undertaken without the mention of the neurotransmitter dopamine (DA). While none of the members of the basal ganglia are immune from the influence of dopaminergic fibers of the SNc, the striatum is without question the favored target and the most affected by this nucleus (Lavoie, Smith et al. 1989). The expression of two classes of DARs on d-and iSPNs are, themselves, responsible for the variance in gene expression pattern between the two cell types (Gerfen, Engber et al. 1990).

That D1 or D2 expression on SPNs should differentially modulate gene expression, or intrinsic cellular physiology, should not be surprising, as the receptors activate completely difference intracellular signaling cascades. DAR as a family are G-protein class receptors (GPCRs) that are

comprised of five members (D1-5) broken into two groups: D1-like receptors (D1, D5) and D2-like

receptors (D2-4) (Neve, Seamans et al. 2004). D1-like receptors are Gs coupled to initiate elevations in intracellular cAMP which activates protein kinase A (PKA) and phosphorylation of

targets sensitive to this kinase; D2-like receptors are Gi/o coupled so that their activation leads to Gα-dependent effects – e.g. inhibition of cAMP – alongside the actions of the Gβγ subunit which will generate diacetylglycerol (DAG), protein kinase C (PKC) and IP3 secondary to activation of phospholipase C (Stoof and Kebabian 1984, Hernandez-Lopez, Tkatch et al. 2000, Svenningsson, Nishi et al. 2004). Though PKA has a multitude of effects, some of the most relevant for plasticity are its ability to promote insertion of glutamate receptor subunits into the membrane and enhance Ca2+ permeability at the GluN2B subunit (Snyder, Allen et al. 2000, Hallett, Spoelgen et al. 2006, Murphy, Stein et al. 2014). D1Rs themselves, independent of phosphorylation, have been recently shown to alter the subcellular localization of NMDA receptors to a more LTP-friendly configuration (Ladepeche, Yang et al. 2013). D2R can also more directly influence synaptic strength via depression of conductance through AMPA 32

receptors and decreasing the phosphorylation at GluA1 Ser845 sites which promotes their removal (Hernandez-Echeagaray, Starling et al. 2004, Hakansson, Galdi et al. 2006).

At first glance the role of dopamine in striatal plasticity seems straightforward: D1R activation will promote long-term potentiation of glutamatergic synapses whereas D1R activation should promote long-term depression by virtue of their effects on cAMP/PKA levels. However, early reports (and more recent ones) have revealed that paradigms utilizing high- frequency stimulation (HFS) to induce LTD work in all SPNs, not the ~50% that would be expected if DARs were exclusively responsible for LTD (Calabresi, Maj et al. 1992, Bagetta, Picconi et al. 2011). Though this observation was not repeatable by other groups who found HTF-LTD only in iSPNs, it raised the question as to whether both cell types were capable of undergoing LTD or LTP, as their differential DAR expression would suggest they were not (Kreitzer and Malenka 2007, Mathur and Lovinger 2012). A sophisticated study utilizing spike time dependent plasticity (STDP) and a recording configuration to maintain the integrity of intracellular signaling cascades found typical, bidirectional Hebbian plasticity in both cell types that was dependent on differential activation or inhibition of DARs on the two cell types (Shen, Flajolet et al. 2008). Additionally, it was found that other GPCRs could influence DARs: when D2R signaling was inhibited and A2a adenosine receptors activated, paradigms normally eliciting LTD were found to cause LTP.

The aforementioned LTD has been critically studied by a number of groups (Kreitzer and Malenka 2008). From these studies we know that it is endocannabinoid (eCB) mediated, induced post-synaptically and expressed pre-synaptically via decreases in Ca2+ at synaptic terminals through N-type Ca2+ channels (Szabo, Lenkey et al. 2014). This mechanism is arguably the most closely studied with respect to striatal plasticity (Mathur and Lovinger 2012). What remains to be completely elucidated is how other receptors might modulate the effects of D1 and D2Rs to push SPNs towards one form of learning or the other. As will be discussed below, other receptors (e.g. muscarinic subtype 4 receptors) whose intracellular pathways also converge on cAMP/PKA can modulate the expression of LTD or LTP in both normal and diseased states (Lerner and Kreitzer 2012, Shen, Plotkin et al. 2015). 33

Endocannabinoids

Endocannabinoids (eCBs) are ligands of the cannabinoid receptors, CB1 and CB2 (Matsuda, Lolait et al. 1990). In the CNS two ligands exist that act on both receptor subtypes; arachidonoyl ethanolamide (anandamide) and 2-arachidonoyl-glycerol (2-AG) (Devane, Hanus et al. 1992, Mechoulam, Ben-Shabat et al. 1995). In the striatum, CB1 receptors are found on glutamatergic, GABAergic, and dopaminergic terminals and similarly reduce the release of their respective neurotransmitters (Gerdeman and Lovinger 2001, Matyas, Yanovsky et al. 2006, Sidlo, Reggio et al. 2008). In the interest of brevity, we will focus on presynaptic eCBRs on glutamatergic terminals. CB1 receptors are activated by eCBs synthesized at the postsynaptic site within the dendritic spines of SPNs (Yoshida, Fukaya et al. 2006). This makes eCBs a retrograde signaling molecule.

eCB synthesis is the result of a strong presynaptic stimulus coupled with postsynaptic depolarizations causing strong activation of group I mGluRs along with elevated intraspine Ca2+ levels, respectively (Mathur and Lovinger 2012). eCB can then activate CB1 receptors which are Gi coupled GPCRs. This activation causes a decrease in neurotransmitter (glutamate) release through a combination of the inhibition of voltage gated calcium channels (VGCCs) and activation of classes of K+ channels – G-protein coupled K+ currents (GIRKs) and A-type K+ channels (Henry and Chavkin 1995, Bacci, Huguenard et al. 2004, Guo and Ikeda 2004). This results in a presynaptic expression of LTD.

Because eCB synthesis is so dependent on the timed convergence of activation of the PLC cascade and elevated intracellular Ca2+ levels, manipulations that interact with either pathway can influence whether LTD occurs. For example, stimulation of D2R or muscarinic subclass 4 receptors (M4R) both decrease AC/cAMP levels (Beaulieu and Gainetdinov 2011, Ockenga, Kuhne et al. 2013). Both manipulations reduce the activation of a regulator of G protein signaling (RGS4), a GTPase that normally functions to inhibit eCB generation (and consequentially, LTD) (Lerner and Kreitzer 2012, Shen, Plotkin et al. 2015). 34

For years, paradigms utilizing intrastriatal stimulation have used the phrase “corticostriatal plasticity”, neglecting the possibility that thalamostriatal synapses, which comprise approximately 40% of glutamatergic synapses in the nucleus, also underwent LTP or LTD (Smith, Raju et al. 2004, Doig, Moss et al. 2010). eCB-LTD has recently been shown to be exclusive to corticostriatal synapses – albeit the presence of low CB1 receptor expression at thalamostriatal synapses – which has been supported by experiments utilizing selective expression of channelrhodopsin (ChR2) in either cortical or thalamic nuclei (Wu, Kim et al. 2015). The existence of differential expression of eCB-LTD is not surprising given other molecules, such as the opioid peptides, that show differential expression of LTD (Atwood, Kupferschmidt et al. 2014).

Two questions remain for closer study: 1.) How might stimulation paradigms (STDP, HFS, intrastriatal stimulation vs. cortical stimulation) bias preparations towards or against eCB- LTD? 2.) How do other forms of LTD interact with eCB-LTD? The second question is particularly important given recent studies demonstrating the occlusive properties that activation of presynaptic serotonergic receptors has on typical corticostriatal eCB-LTD (Kreitzer and Malenka 2007, Picconi, Bagetta et al. 2011, Mathur and Lovinger 2012).

Acetylcholine

Within the striatum, Ach released from ChIs targets two types of receptors: ionotropic nicotinic and metabotropic muscarinic receptors (nAChRs, mAChRs, respectively). Like DA receptors, the latter category can be subdivided into two classes largely known as the M1-like receptors (M1, M3, M5) and M2-like receptors (M2, M4) (Kruse, Kobilka et al. 2014). While M1, M2, and M4 are definitively expressed in the striatum, the presence of M3 varies by study, and M5 appears absent (Weiner, Levey et al. 1990, Hersch, Gutekunst et al. 1994, de la Vega, Nunez et al. 1997). mAChRs, particularly of the M1 class, have been shown to enhance iSPN excitability, by promoting the closure of KCNQ channels responsible for the M-current, as well as the inward rectifiers responsible for setting SPN resting membrane potential (Shen, Hamilton et al. 2005, Shen, Tian et al. 2007). 35

Figure 1.3. Glutamatergic synapses onto SPNs Excitatory, glutamatergic inputs from both cortex and the thalamus impinge on dendritic spine heads of striatal projection neurons (SPNs). Glutamate released from these terminals activates postsynaptic AMPA and NMDA receptors on SPNs to generate excitatory post-synaptic currents/potentials (EPSC/Ps). Glutamatergic activation of type I metabotropic glutamate receptors (mGluRs) activates intracellular signaling cascades that, when temporally aligned with elevated Ca2+ via influx through L-type Ca2+ channels, generates synthesis of endocannabinoids (eCBs) which act as a retrograde messenger to reduce glutamate release at presynaptic, cortical terminals, effectively causing long-term depression (LTD) at this synapse. The expression of DA receptors on the postsynaptic SPN terminals influence how efficacious the generation of eCBs and subsequent plasticity is. The expression of this plasticity is, thus, presynaptic. Other modes of presynaptic plasticity include serotonergic activation of presynaptic 5-HT receptors, muscarinic activation of M2 receptors, and activation of -opioid receptors through pharmacological agents or endogenous peptides (e.g. encephalin). While𝛿𝛿 eCB-dependent LTD is not as robustly present at thalamic synapses, other forms of LTD have been observed which include the activation of µ- opioid receptors, and undetermined mechanisms that generate LTD following both low-frequency and high-frequency stimulation (LFS, HFS) paradigms. Nitric oxide (NO) has been posited to also play a role in striatal LTD, likely through association with a post-synaptically expressed soluble guanylyl cyclase (sGC) it’s cytoplasmic receptor. 36

The M1R class was also shown to enhance currents through NMDAR though other groups have found no enhancement of postsynaptic glutamate receptors in response to ACh (Calabresi, Centonze et al. 1998, Higley, Soler-Llavina et al. 2009). This discordance may be partially attributed to different preparations and experimental design, in so far as the same group that was showed NMDAR conductance increased with M1 activation, also found that HFS-LTP was potentiated when M2 receptors were antagonized, suggesting the presence of a tonic inhibition of presynaptic glutamate (Calabresi, Centonze et al. 1998, Calabresi, Centonze et al. 1998). As such it appears that mAChRs have a dual effect on plasticity that is theorized to be highly dependent on nascent Ach levels (DeBoer and Abercrombie 1996, Martella, Tassone et al. 2009). These findings are consistent with recent studies showing that M4 receptors promote LTD in dSPNs and that this effect is lost in pathological states in which DA and ACh levels become abnormal (Picconi, Bagetta et al. 2011, Shen, Plotkin et al. 2015).

Nitric oxide

Nitric oxide is one of the more contentious molecules associated with striatal synaptic plasticity. Depending on the preparation (in vivo vs. acute brain slice) and the stimulation paradigm utilized (HFS vs. STDP vs. low-frequency stimulation, LFS) groups have seen both LTD and LTP following activation of NO signaling (Centonze, Gubellini et al. 2003, West and Tseng 2011). Nitric oxide can act via activation of its receptor, soluble guanylyl cyclase (sGC, pGC, respectively), or directly through a post-translational modification known as S-nitrosation (Ahern, Klyachko et al. 2002). Though S-nitrosation may play a role in plasticity by targeting many postsynaptic targets – including the α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and NMDA glutamatergic receptors – the NO-sGC pathway is thought to be more important in SPNs (Menniti, Faraci et al. 2006, Garthwaite 2008). This is because SPNs express some of the highest levels of sGC within the brain, as well as various classes of phosphodiesterases (PDEs), which are responsible for degrading cyclic nucleotides such as cyclic guanosine monophosphate (cGMP) that are activated following NO binding of sGC (Menniti, Faraci et al. 2006). 38

Some of the earliest studies of striatal LTD implicated NO in the expression of this form of learning. These studies showed that either HFS or the mixed class PDE inhibitor zaprinast could induce glutamatergic LTD and that either of these methodologies was occluded by application of a pharmacological “donor” of NO, s-nitroso-N-acetylpenicillamine (SNAP) (Calabresi, Maj et al. 1992, Calabresi, Mercuri et al. 1993, Calabresi, Gubellini et al. 2000). Although this work was critical to understanding striatal plasticity it was unclear where the expression of this LTD was occurring and how it was being induced. Furthermore, a distinct line of work in vivo argued for an NO-mediated augmentation of corticostriatal excitability at both loci. Specifically, they showed that concurrent stimulation of frontal cortex in the presence of nNOS inhibitors resulted in short term depression of cortically evoked SPN spiking activity, indicating a tonic potentiation of cortical glutamate release mediated by NO (Sammut, Park et al. 2007). A similar manipulation to decrease NO signaling at the postsynaptic site (SPNs) through the inhibition of its receptor, sGC, resulted in decreases SPN excitability as measured by input resistance, action potential amplitude and spontaneous transitions to the “up-state” (West and Grace 2004, Ondracek, Dec et al. 2008).

Across multiple brain regions, the use of different paradigms and/or preparations (in vivo vs in vitro) to induce plasticity has frequently led to reports of conflicting results; this phenomenon is not limited to the striatum (see above sections on Dopamine and Endocannabinoids). As such the dissent between the two NO camps (those that believe it is an LTD-permissive molecule and those that believe promotes LTP) is not a surprise nor a negation of a role for NO in either form of learning (Charpier and Deniau 1997, Pare, Shink et al. 1998). However, what remains unclear is what exactly NO is doing to cortical synapses within the striatum, whether this manipulation can be extended to the other set of glutamatergic inputs (thalamostriatal inputs), and whether these phenomena is indeed mediated by the nNOS expressing PLTS interneurons.

III. Pathophysiology in the striatum: a role for dopamine and acetylcholine Parkinson’s disease is the second most common neurodegenerative disease and one of the most economically burdensome (www.ninds.nih.gov). The cardinal symptoms of the disease – rigidity, bradykinesia, resting tremor, and postural instability – have helped define the disorder which, historically, has been thought of as a motor disorder (Ward and Gibb 1990, Gelb, Oliver et al. 1999). However, essentially all patients (97%) self-report non-motor symptoms including cognitive dysfunction or dementia, mood disorders, sleep disturbances, autonomic and olfactory dysfunction, and many more (Barone, Antonini et al. 2009). These symptoms are inexorably linked to the death of the dopaminergic neurons in the substantia nigra pars compacta (SNc), but likely, also reflect the dysfunction and death of other vulnerable populations within both the cerebrum and peripheral nervous system (PNS) (ref). In the former population, most of the afflicted neurons lie within the brainstem and also express excess levels of neuromelanin (NM), a pigmented derivative of oxidized catecholamines that is postulated to act as a sink for toxic elements within a cell – attracting metals, iron, etc. (Sulzer, Mosharov et al. 2008). Other similarities among this group of cells are that they are autonomous pacemakers, have massive axonal terminal fields, and eventually exhibit profound Lewy Body pathology (Conway, Rochet et al. 2001, Matsuda, Furuta et al. 2009, Orimo, Uchihara et al. 2011).

These observations have helped form one of the leading theories in the field regarding why SNc (and other) neurons die in PD; this theory centers around a “selective vulnerability” hypothesis (Surmeier, Guzman et al. 2012, Brichta and Greengard 2014). Specifically, dopaminergic SNc neurons engage a unique compilation of channels in their pace making machinery that induces an excessive burden on these cells. That burden arises because moving Ca2+ back outside a cell is energetically expensive – but especially necessary in this cell type given their low expression of calcium binding proteins – and as such, these cells are metabolically taxed throughout their lifetimes (Foehring, Zhang et al. 2009, Surmeier and Schumacker 2013). This vulnerability is underscored by the need to sequester neuromelanin and to also provide mitochondrial and protein support to their massive, mostly unmyelinated axonal tree (Sulzer, Bogulavsky et al. 2000, Zecca, Zucca et al. 2003, Braak and Del Tredici 2004). 40

What results from the death of this neurotransmitter? Traditional models of striatal adaptions to dopaminergic denervation suggest that iSPNs become hyper-excitable whereas dSPNs become hypo-excitability. These two changes subsequently cause cessation of movement due to a dis-inhibition of basal ganglia effectors systems consistent with the bradykinesia in most PD patients (DeLong and Wichmann 2007). On a cellular level, the iSPN hyper-excitability theory is consistent with experimental evidence that has repeatedly demonstrated a pruning of iSPN dendrites and dendritic spines in mice that have undergone unilateral lesioning of their SNc using the toxin 6-hydroxydopamine (6-OHDA) to model PD (Day, Wang et al. 2006, Peterson, Goldberg et al. 2012). Nevertheless, dSPNs were also shown to undergo similar anatomical changes which would suggest they should develop hyper excitability as well. Recent evidence has shown that in spite of the similarity in anatomy in the 6-OHDA state, there is a differential change in excitability that is unexpectedly opposite to what was long predicted: specifically, iSPNs from 6-OHDA treated mice are hypo excitable whereas dSPNs acquire an opposite physiological signature (Fieblinger, Graves et al. 2014). This data has helped shape a new theory of how the striatum responds to dopamine loss. This theory is grounded in homeostatic plasticity. Specifically, SPNs exhibit intrinsic homeostatic adaptations to maintain their relative excitabilities: iSPNs “tune down” their excitability since the molecule (DA) that normally does this is now absent whereas dSPNs become more excitable to compensate for the lack of dopaminergic tone that might normally drive them to this state (Albin, Young et al. 1989, Meurers, Dziewczapolski et al. 2009, Heiman, Heilbut et al. 2014). What fails to occur, however, are the proper homeostatic adaptations to synaptic changes, chief of which is the loss of corticostriatal synapses (Stephens, Mueller et al. 2005, Day, Wang et al. 2006, Villalba, Lee et al. 2009, Fieblinger, Graves et al. 2014). This failure undoubtedly contributes to the disease phenotype though how and why it occurs remains to be determined (Zaja-Milatovic, Milatovic et al. 2005, Smith and Villalba 2008).

As would be expected, synaptic plasticity in Parkinsonian mice is profoundly abnormal. Multiple groups have shown that chronic DA denervation leads to loss of both LTD and LTP within the striatum (Picconi, Centonze et al. 2003, Shen, Plotkin et al. 2015). The mechanisms 41

behind this loss are not as straightforward as simply removing or replacing dopamine (see below section), and it is likely that all striatal microcircuitry is affected and contributes to circuit pathophysiology and symptomatology. For example, it is well documented that PD patients exhibit hyper-cholinergic signaling, a state that likely contributes to the efficacy of anticholinergic drug treatment (Duvoisin 1967, Katzenschlager, Sampaio et al. 2003). However, it should be noted that these effects are mostly limited to lessoning resting tremor (Schwab, Poskanzer et al. 1972, Lang 1984). Faulty cholinergic signaling has been demonstrated in 6- OHDA models in slice, likely resulting from decreased D2R mediated inhibition of ChI firing as well as a reduction in M4 autoreceptor functioning in these cells (Maurice, Mercer et al. 2004, Ding, Guzman et al. 2006, Kharkwal, Brami-Cherrier et al. 2016).

Multiple groups have shown aberrant activity in other interneuron populations. Inhibitory signaling between FSIs and SPNs, PLTSIs and SPNs, and SPNs onto other SPNs via recurrent collaterals are all altered following 6-OHDA treatment (Taverna, Ilijic et al. 2008, Dehorter, Guigoni et al. 2009, Gittis, Hang et al. 2011). The direction of change (enhancement or depression of GABAergic signaling) is highly dependent on the type of presynaptic input. While intrinsic and synaptic changes in and onto cholinergic and fast-spiking interneurons in a DA depleted state have been studied, the same cannot be said for all subclasses of interneurons (Pisani, Bonsi et al. 2003, Damodaran, Cressman et al. 2015, Ztaou, Maurice et al. 2016). These questions must be answered before a comprehensive view of striatal adaptations to DA denervation can be synthesized.

Levodopa induced dyskinesia: a dopamine oscillatory state

Though we are still lacking in disease modifying agents, there are a number of pharmacological and surgical tools directed at treating PD that have proved efficacious in addressing certain symptoms (www.updoate.com). Historically, the most utilized of these is dopamine replacement in the form of the precursor to this transmitter, 3,4-dihydroxylphenyl-L- alanine (L-DOPA) (Rascol, Payoux et al. 2003). With respect to bradykinesia, L-DOPA is the most effective drug, though tremor and rigidity also show some responsiveness (Miyasaki, 42

Martin et al. 2002, Connolly and Lang 2014). Even in younger patients, in which dopamine agonists (e.g. bromocriptine, pramipexole, etc.) are preferred, most will eventually require dopamine replacement as their symptoms become more progressive (Quinn, Critchley et al. 1987, Kumar, Van Gerpen et al. 2005).

L-DOPA is often withheld in younger patients – that will likely live longer and require more years of treatment – for concern of a common sequelae of DA replacement, levodopa- induced dyskinesia (LID) (Quinn, Critchley et al. 1987, Marras and Lang 2008). LID is common, appearing in at least 50% of patients following 5-10 years of L-DOPA use (Olanow, Watts et al. 2001). Though mild dyskinesias are normally tolerated by patients in exchange for the increase in mobility, others experience “on state” choreiform dyskinesia that is so aggressive L-DOPA treatment may need to be temporarily stopped (Quinn 1995). Paralleling these periods is a phenomena known as “wearing off” in which patients stop responding as robustly or respond for shorter time periods, following treatment (Connolly and Lang 2014).

Despite the prevalence and severity of LID, it remains the mainstay treatment of PD (Quinn 1995, Fabbrini, Brotchie et al. 2007). This has led to more resources allocated towards understanding why LIDs occur with the ultimate goal of a complementary therapy to attenuate their development / expression (Cenci and Konradi 2010). Much like PD itself, aberrant DA levels are hypothesized to cause LID; however, in this case it is believed that the influx of DA causes hyper-activation of sensitized DARs (Cenci and Konradi 2010). D1R stimulation forces a mismatch in plastic processes: synapses are pushed towards LTP and unable to undergo LTD or “depotentiation” from LTP (Calabresi, Pisani et al. 2016). Evidence suggests that LTD can be restored by enhancing M4 signaling in dSPNs – which serves to inhibit the inhibition of eCB production – or by inhibiting of striatal PDEs (Picconi, Bagetta et al. 2011, Shen, Plotkin et al. 2015). How, and if the two mechanisms are related remains to be determined, as does how they differ between “on” and “off” states (times of elevated and no striatal DA, respectively). Both elements will be critical to our understanding of how dSPN output becomes hyperactive, causing parallel hyperactivity of movement (Surmeier, Graves et al. 2014). 43

IV. Dissertation aims and organization This dissertation was initiated with the goals of: 1.) determining how inhibitory inputs from PLTSIs shaped SPN dendritic integration and influenced striatal output, and 2.) how these processes were changed in a model of Parkinson’s disease. As data was collected, it became evident that the methodology chosen to investigate these inputs and/or the outcome measures was not efficacious or robust enough to study this synapse. While it has proved possible to execute these types of experiments in both the hippocampus and cortex, both of these structures have a well understood, strictly laminar structure that makes subcellular interrogation less burdensome (Petreanu, Mao et al. 2009, Lovett-Barron, Kaifosh et al. 2014). Furthermore, our results concerning alterations in PLTS interneuron activity in the 6-OHDA model were not aligned with previous publications that had modestly studied this phenomenon and will be briefly discussed in chapter 4 (Dehorter, Guigoni et al. 2009).

Paralleling these questions was a similar project that had begun to reveal the influence of NO on SPN LTD. Given their sparse dendritic and axonal trees and the low probability of PLTSI – SPN synaptic contacts, the initial hypothesis was reevaluated based on the premise that the PLTSIs primarily modulated striatal activity through longer acting neuromodulatory molecules such as NO, SST, and NPY. While pharmacology was utilized as a first means of probing this phenomenon, recent genetic and technical advances allowed for NO to be cemented as a bona fide neurotransmitter. The SST-IRES-Cre and NPY-Cre mice allowed for the selective expression of Cre-dependent viruses in these two lines while the commercial availability of numerous Cre-dependent channelrhodopsin and halorhodopsin variants presented the tools to selectively enhance or inhibit PLTSI activity. Using the former construct, it was revealed that tuning these cells to higher firing frequencies for short periods of time enabled the induction of NO release and, consequently, plasticity at SPN glutamatergic synapses (see Chapter 2).

The NO-LTD work has led to the reiteration of a number of questions, many of which have been previously identified in several publications that have shown a critical role for NO in dopaminergic pathologies (Padovan-Neto, Echeverry et al. 2009, Solis, Espadas et al. 2015). 44

However, it is difficult to hypothesize what a group of cells is doing in “abnormal” conditions when their activity in the unperturbed striatum is so ill defined. Even a question as basic as “what innervates PLTSIs?” was unclear. Fortunately, paralleling the development of Cre- dependent opsins and a myriad of Cre-driver lines, was the creation of a system allowing for Cre-dependent retrograde labeling of select cell populations utilizing Rabies virus. This tool was irrefutably critical to the second half of this dissertation (Chapter 3) which is focused in describing the intrastriatal circuits in which PLTSIs are involved. It was during this course of investigation that a novel role was found for giant cholinergic interneurons and we were able to utilize other transgenic lines (NPY-eGFP, CHAT-Cre) to query the relationship between PLTSIs and ChIs. Finally, the use of pharmacological tools and novel imaging modalities was critical in our ability to tease out the mechanisms underlying the cholinergic modulation of PLTSI activity and the broader implications of this manipulation.

Chapter 3

Interneuronal nitric oxide signaling mediates post-synaptic long-term

depression of striatal glutamatergic synapses

Introduction

The striatum has long been implicated in learning goal-directed behavior and habits (Balleine et al., 2007). This learning is thought to reflect changes in the strength of glutamatergic synapses that dictate the timing and pattern of activity of principal SPNs (Gerfen and Surmeier 2011). In the dorsal striatum, axospinous glutamatergic synapses on SPNs are formed primarily by cortical pyramidal neurons. Long-term depression (LTD) of these synapses was initially described over 20 years ago and has been extensively studied (Calabresi, Maj et al. 1992, Centonze, Picconi et al. 2001, Kreitzer and Malenka 2008, Gerfen and Surmeier 2011, Mathur and Lovinger 2012). The best characterized form of LTD is induced by post-synaptic depolarization and activation of metabotropic glutamate receptors and L-type Ca2+ channels, which trigger the generation of endocannabinoids (eCBs) by SPNs; expression of LTD is mediated by eCB activation of pre-synaptic CB1 receptors, which result in sustained depression of glutamate release (Lovinger, Tyler et al. 1993, Kreitzer and Malenka 2005). Although other neuromodulators can mimic the actions of eCBs at these synapses, no post-synaptically expressed form of LTD has been described in SPNs (Mathur, Capik et al. 2011).

One of the signaling molecules implicated in post-synaptically expressed LTD elsewhere in the brain is NO (Garthwaite 2008). Striatal expression of NO signaling proteins – soluble guanylyl cyclase (sGC) and protein kinase G (PKG) – are among the highest of any brain region, making a role for NO signaling in striatal plasticity plausible (Ariano 1983, Ding, Burette et al. 2004). Indeed, NO has been implicated in striatal synaptic plasticity, but its role is controversial 46

(Calabresi, Gubellini et al. 1999, Sammut, Park et al. 2007), with most of the available data suggesting it is a permissive modulator of eCB-dependent LTD (eCB-LTD) (Centonze, Gubellini et al. 1999, Centonze, Picconi et al. 2001).

Materials and Methods

Slice Preparation: All experiments were performed in accordance with the Northwestern University ACUC and NIH guidelines. Unless otherwise specified, we used 6-12 week old, hemizygous, male C57Bl/6 mice for this study. Drdla-eGFP and Drd2-eGFP BAC transgenic mice were used. In some experiments, BAC transgenic mice were crossed with a transgenic line expressing ChR2 in the cortex (Thyl-ChR2- eGFP; Jackson Laboratory, JAX# 007612). For PLTSI stimulation experiments, male and female SST-IRES-Cre mice were used (Jackson Laboratory). For optogenetic experiments probing thalamostriatal synapses, male Tg(Grp-Cre)KH288 mice developed by the GENSAT Project were utilized, in which ere expression is limited to the central lateral (CL) nucleus of the thalamus. Mice were anesthetized with a ketamine/xylene mixture, perfused with ice-cold, oxygenated artificial cerebrospinal fluid (ACSF) and decapitated. The ACSF contained (in mM): 124 NaCl, 3.5 KCl, 2CaCb, 1 MgCb, 26 NaHC03, 1.2 NaH2P04, and 10 D-glucose. Para-sagittal brain slices, 275-300µM thick, containing striatum were prepared and, after cutting, the slices were transferred to a holding chamber containing ACSF and bubbled continuously with 95% 02 and 5% CO2 at 30-33°C.

Electrophysiology: Individual slices containing dorsal lateral striatum were transferred from the holding chamber to a submersion-style recording chamber. A variable flow mini-pump was used to continuously perfuse ACSF over the slice. All electrophysiological recordings were obtained at elevated temperature (30-33°C) with 10 µM gabazine in the ACSF to block GABAA mediated currents. Whole-cell 47

recordings were performed from D l or D2 SPN s of the dorsolateral striatum, identified under differential contrast microscopy and via their fluorescence. Glass electrodes (3-6 M.O) were filled with a cesium-based solution and cells allowed to fill for 10 minutes before starting experiments. The internal solution contained (in mM): 120 CsMeS03, 5 NaCl, 10 TEA-Cl, 10 Hepes, 5 QX-314, 4 ATP-Mg, .3 GTP- Na, and .25 EGTA. This internal solution was brough to pH 7.2-7.3 and 270-280 mOsm/1. EGTA concentration was increased to SmM for experiments in which PLTSis were optically stimulated. For evoked EPSC recordings, SPNs were voltageclamped at -70 mV and stimuli delivered at O.lHz through a concentric bipolar electrode placed between layer V and VI of the cortex (Frederick haer& Co, ME). For sequential LTD experiments (Fig. 2B, 9BC) an attempt was made to normalize the amplitude of the baseline EPSCs as the pre-incubation with SNAP/BAY41 or 8-Br-cGMP would have been expected to induce NO-LTD and depress cortical EPSCs compared to the control data set. For optogenetic PL TSI LTD experiments, SPN s were first briefly held at +O m V in the absence of gabazine to ascertain the relative connectivity of the SPN to ChR2-infected PLTSis by measuring the postsynaptic inhibitory current (IPSC) that resulted from a full-field (1-Sms) light stimulus from the LED light source. Cells were then brought back down to -70 m V, and gabazine perfused for the duration of the experiment. For perforated-patch recordings, the internal solutions contained (in mM): 126 KMeS04, 14 KCl, 3 MgC12, 0.5 CaC12, 5 EGTA, 10 HEPES; pH was adjusted to 7.25 with NaOH and osmolarity to 275-280 mOsm. Electrical access was obtained using a stock solution of gramicidin (2mg ml- I) in dimethyl sulfoxide (DMSO) which was prepared daily and diluted in the cording solution before use to a final concentration of 2 ug ml-1. Access resistance was monitored throughout recording and cells in which a change >20% baseline was observed were thrown out. In experiments where drugs were transiently applied, application times were 10 minutes following 5 minutes of baseline. In experiments where drugs were perfused throughout the length of the recording, cells 48

were first exposed to drugs at break in for a period of 10-20 minutes before collecting data. The EPSC amplitude was analyzed using custom scripts written in Matlab (Mathworks, Natic, MA).

Viral gene delivery: For PLTSI and thalamostriatal optogenetic experiments we used an adeno-associated virus (AAV) vector carrying a ChR2-eYFP expression construction (AAV2/9-EFla-doublefloxed-hChR2(Hl34R)-eYFPWPRE-HGHpA, supplied by University of Pennsylvania Vector Core; Addgene 20298). We injected 0.3-0.4 µL of virus into SST-IRES-Cre or KH288-Cre mice, respectively, at a rate of -100 nL/minute. Mice were sacrificed 2-3 weeks post infection and brain slices used for electrophyiological recordings. Coordinates for dorsolateral striatum injections into SST-IRES-Cre mice were (relative to bregma): 1.10 mm anterior, -2.00 mm lateral, -3.00 mm ventral; to target CL of KH288-Cre mice, coordinates were (relative to bregma): 1.34 mm posterior, 0.78 mm lateral, 3.1 mm ventral.

2-photon laser scanning microscopy (2PLSM): iSPNs were identified by somatic eGFP expression using 2-photon excited fluorescence induced and detected with a Prairie Ultima laser scanning microscope system (Prairie Technologies). Fluorescent and bright-field images were viewed in register using a Dodt contrast detector system. Cells were patched using video microscopy with a Hitachi CCD camera and an Olympus 60X/ 1.0 NA lens. Green (EGFP) and red (Alexa 568) signals were acquired using 810 nm excitation (Chameleon-XR laser system, Coherent Laser Group, Santa Clara, CA) pulsed at 90 MHz (-250 fs pulse duration). Power attenuation was achieved with two Pockels cells electro-optic modulators (models 350-80 and 350-50, Con Optics, Danbury, CT). The two Pockels cells were aligned in series to provide an enhanced modulation range for fine control of the excitation dose (0.1 % steps over four decades). Following patch rupture the internal solution was allowed to equilibrate for 10-15 minutes before imaging. High magnification z-series of 49

dendrite regions -40-100 µm from the soma were acquired with 0.072 µm2 pixels with a 4 µs dwell time and 0.5µm z-steps. At the end of each experiment whole cell z-series were acquired with 0.36 µm2pixels with a 10 f-LS dwell times and 1 f-lm z- steps.

Ca2+ imaging: SPN s were loaded with 25 µM Alexa 568 and 600 µM Fluo SF via the recording electrode, which also contained (in mM): 120 CsMeS03, 5 NaCl, 10 TEA-Cl, 10 HEPES, 5 QX-314, 4 ATP-Mg, 0.3 GTP-Na. Green fluorescent line scan signals were acquired from dendritic spines at 6 ms per line and 512 pixels per line with 0.08 µm pixels and a 10 µs pixel dwell time. The line scan was started 200 ms before a 300 ms voltage step from -70 to O m V and continued for 1.5 s to obtain the background fluorescence and to record the decay of the optical signal after stimulation. To reduce photo- damage and photo-bleaching, the laser was fully attenuated using the second Pockels cell at all times during the scan except for the period directly flanking the voltage step. Changes in Fluo SF fluorescence were measured as liF/Fo.

2-photon laser uncaging of glutamate (2PLUG): Glutamate uncaging was achieved using a Verdi/Mira laser system. 5 mM MNI-glutamate (Tocris, Cookson, Ellisville, MO) was superfused over the slice at 0.4 ml/hr using a syring pump and mulitbarreled perfusion manifold (Cell MicroControls, Norfolk,VA). Glutamate was uncaged adjacent to individual spines using I ms pulses of725 nm light typically 10- 20 mW in power at the sample plane. Photolysis power was tuned via a third Pockels cell modulator (Con Optics, Danbury, CT) to achieve uncaged-EPSCs (uEPSCs) 5 pA uEPSC amplitudes were measured from averaged (5 repetitions) traces.

Chemical and Reagents: Gabazine (SK95531), 8-Br-cGMP, SNAP, AM251, sulpiride, Bay-41, and CSST were obtained from Tocris. RP-8-Br-PET-cGMP was 50

obtained from Biolog Life Science Institute. L-NAME was obtained from Sigma Aldrich. Custom peptides were synthesized by the Stanford peptide facility and used at a concentration of 1-2 mM.

Data Analysis and Statistics: Data analysis was performed using Clampfit 9.2 (Molecular Devices, Inc., Sunnyvale, CA), Igor Pro 6.0 (WaveMetrics, Lake Oswego, OR), and Matlab. For plasticity experiments, the magnitude of LTD was calculated as a percentage of the baseline where the baseline is defined as the average of minutes 0-5 of recording time. The average of minutes 30-35 of each recording was used to calculate effects post drug application except for Figure IE, 2A, and 6B where minutes 25-30 were used. Statistical analysis was performed using Matlab and Prism (GraphPad Software). For time courses, data are reported as Median, first and third quartiles; for box plots, data are reported as median, first and third quartiles, and minimum and maximum of data set (whiskers). Outliers have been removed in both cases.

Results

NO signaling produced LTD of glutamatergic synapses

To assess the potential role of NO signaling in striatal synaptic plasticity, excitatory postsynaptic currents (EPSCs) evoked in SPNs by electrical stimulation of corticostriatal axons were monitored before, during and after the application of the NO donor (S)-nitroso-N-acetyl-D,L- penicillamine (SNAP, 100 μM) to ex vivo parasagittal brain slices of mouse forebrain (Fig. 1A). In this preparation, cortical axons can be stimulated electrically without directly activating striatal neurons, unlike the coronal brain slice (Kawaguchi, Wilson et al. 1989). Transient application of SNAP led to a persistent depression of corticostriatal EPSCs (Fig. 1B).

Figure 2.1. NO induces LTD at corticostriatal synapses through activation of PKG. (A) Simplified diagram depicting the circuit components being examined. (B) Sample whole cell recording of an SPN before, during and after a 10-minute application of the NO donor SNAP (100 µM). Open circles represent single EPSC events and solid line represents average EPSC size over a one-minute interval. EPSC amplitudes are normalized to baseline responses. Access resistance plotted below. Example EPSCs from this SPN are shown to the right before (pre-i, min 1-5 averaged) and after (post-i, min 30-35 averaged) SNAP application. The stimulus artifact has been suppressed for clarity. (C) PKG inhibition (3 µM, Rp-8-Br-PET-cGMPS, Rp8Br; black trace) blocked SNAP induced LTD (red trace) (control: n = 16 cells; PKG inhibitor: n=6). *p<0.05, signed rank test. (D) Cartoon diagram illustrating the NO signaling cascade. (E) Bath application of 8-Br-cGMP (500 μM) induced LTD in dSPNs (n=7). *p<0.05, signed rank test. Example EPSCs are shown to the right. (F) 8-Br-cGMP induced LTD similarly in iSPNs and dSPNs (dSPNs, from 1E: n=7; iSPN: n=9). *p<0.05, rank sum test. (G) SNAP induced LTD was unaffected by the inclusion of mecamylamine (10 µM) and scopolamine (10 µM) in the bath (n=5). Scale bars (vertical, horizontal): (B) 10 pA, 20 ms; (E) 25 pA, 20 ms. (C,E,G) Data are presented as median, 1st and 3rd quartiles.

Data contributed by Igor Rafalovich and Shenyu Zhai. 52

To verify that our electrical stimulus was not directly exciting SPNs, cortical pyramidal neurons were optogenetically activated to evoke EPSCs; SNAP application also produced a persistent depression of optically evoked corticostriatal EPSCs (Fig. 2A).

NO can have both direct and indirect effects on proteins involved in synaptic transmission (Garthwaite 2008). Because its striatal expression is robust (Ariano 1983), our working hypothesis was that the NO donor effects were mediated by sGC activation. NO stimulates sGC, increasing cytoplasmic cyclic guanosine monophosphate (cGMP) production and the activation of PKG (Fig. 1D). If NO was acting through this signaling cascade, its effects should be mimicked by analogs of cGMP and blocked by inhibitors of PKG. Indeed, antagonism of PKG by including Rp-8-Br-PET-cGMPS (3 µM) in the patch pipette blocked the effects of SNAP (Fig. 1C) and brief bath application of the membrane permeable cGMP analog 8-bromo- cGMP (8Br-cGMP, 500 µM) produced a robust and persistent depression of corticostriatal EPSCs, mimicking SNAP (Fig. 1E). Moreover, the ability of SNAP to decrease EPSC amplitudes was occluded in cells that had been pre-incubated in 8Br-cGMP (Fig. 2B).

The effects of cGMP analogs on corticostriatal EPSCs were similar in direct pathway SPNs (dSPNs) and indirect pathway SPNs (iSPNs) (Fig. 1F), consistent with the broad striatal distribution of signaling molecules in the NO pathway (Bredt, Hwang et al. 1990, Vincent 1994).

Because activation of the NO/cGMP/PKG pathway has been shown to augment the activity of cholinergic interneurons (Centonze, Picconi et al. 2001), the effect of SNAP was examined in the presence of antagonists of cholinergic signaling. Perfusion of the nicotinic receptor antagonist mecamylamine (10 µM) and the muscarinic receptor antagonist scopolamine (10 µM) prior to and throughout SNAP perfusion did not alter the depression of corticostriatal EPSCs (Fig. 1G).

To rule out any effects of cellular dialysis with the patch electrode, experiments using SNAP and 8Br-cGMP were also performed in the perforated patch recording configuration. The responses observed were similar to those seen in whole cell (Fig. 2C). 54

Figure 2.2. NO-LTD is induced at corticostriatal synapses by SNAP or 8-Br-cGMP. (A) An example recording from brain slices taken from a transgenic Thyl-ChR2 mice in which ChR2 is expressed in the cortex (see supplemental methods). Corticostriatal responses were evoked using full field stimulation with an LED light source (0.2 ms stimulus duration). Similar to what was seen with electrical stimulation of cortex, transient perfusion of SNAP (I 00 µM) induced depression of cortical EPSCs. (Control, min 1-5: -38.5 pA; Post SNAP, min 25-30: -24.0 pA, 55.89% of baseline) *p < 0.05, MannWhitney nonparametric test. (B) NO-LTD induced by transient perfusion of SNAP (100 µM) was occluded by pre-incubation with 8-Br-cGMP (500 µM). 8-Br-cGMP was perfused for 10 minutes before, plus during SNAP application (n = 5 cells). p > 0.5, Wilcoxon signed rank test. (C) An example recording utilizing the perforated patch recording configuration demonstrating that NO-LTD induction using 8-BrcGMP was not affected by cellular dialysis associated with the whole-cell recording configuration. Examples traces are shown to the right. Scale bars: (C) 50 pA, 20 ms. (B) Data are reported as median, first and third quartiles.

Data contributed by Shenyu Zhai.

Because the NO-induced reduction in evoked EPSC amplitude persisted for as long as recordings could be maintained, it will be referred to as NO-LTD.

Optogenetic activation of PLTS interneurons induced NO-LTD

Although the effects of SNAP and cGMP analogs were consistent and robust, this does not prove that NO signaling is engaged by the striatal circuitry to control synaptic strength (Feelisch 1998). In the striatum, neuronal nitric oxide synthase (nNOS) is expressed robustly only in interneurons that co-express somatostatin, neuropeptide Y and gamma amino butyric acid (GABA) (Tepper, Tecuapetla et al. 2010). Because of their distinctive physiological properties, these cells also are referred to as persistent and low threshold spiking interneurons (PLTSIs). If NO is a bona fide modulator of plasticity, the activation of PLTSIs and their generation of NO should induce NO-LTD. To test this hypothesis, we used optogenetic methods to selectively express channelrhodopsin2 (ChR2) in striatal PLTSIs (Zhang, Wang et al. 2006, Witten, Steinberg et al. 2011, Holley, Joshi et al. 2015) (Fig. 3A, 4A-C and 5A). Opposed to the case in genetic crosses, viral delivery of a floxed construct resulted in no off-target labeling of other interneuron subtypes (e.g. FSIs, Fig. 4A-C) (Hu, Cavendish et al. 2013). In ex vivo brain slices from these mice, SPNs were sorted into two groups: those in which full-field optogenetic stimulation of PLTSIs evoked GABAergic responses (coupled) and those that showed no response (no coupling) (Fig. 5B). In SPNs coupled to PLTSIs, optical stimulation for 5 minutes at 15 Hz, a firing frequency that spontaneously active PLTSIs do not normally achieve in slice, led to a robust LTD at corticostriatal synapses (Fig. 3B and 5C). Exciting PLTSIs to firing frequencies lower than 15Hz, or stimulating for shorter durations, failed to elicit LTD (Fig. 3D). As expected, when the nNOS inhibitor NG-Nitro-L-arginine methyl ester (L-NAME, 100 μM) or the PKG inhibitor Rp8Br (3 µM) were applied throughout the recording period, LTD induction was blocked (Fig. 3C and 3D). Similarly, pre-incubation of slices with SNAP occluded PLTS induction of LTD (Fig. 5E). In contrast, in SPNs where optical stimulation failed to evoke GABAergic responses (possibly because optical stimulation was ineffective in activating PLTSIs), the same stimulation protocol had no effect on corticostriatal synaptic transmission (Fig. 3D and 5C-D). Taken together, these data argue that NO signaling originating from PLTSIs induces LTD at corticostriatal synapses. 57

Figure 2.3. NO producing interneurons can induce NO-LTD. (A) Simplified diagram depicting the experimental design used to optogenetically probe PLTSI involvement in NO-LTD. (B) Full field LED activation of PLTSIs for 5 minutes at 15 Hz induced NO-LTD at corticostriatal synapses in SPNs synaptically coupled to PLTSIs. Example traces shown to the right. (C) LTD was blocked by continuous bath application of the nNOS inhibitor L- NAME (100 µM) or inclusion of Rp8Br (3 µM) in the patch pipette (control: n=5; L-NAME: n=6; Rp8Br: n=5). Example EPSCs are shown at right. (D) Summary data for NO-LTD induced by PLTSI activation. SPNs not coupled to PLTSIs showed no response to the PLTSI stimulation paradigm (n=4). **p<0.01 Mann-Whitney nonparametric test. Scale bars: (B), (C) 20 pA, 10 ms. (B-C) Data are presented as median, 1st and 3rd quartiles.

Data contributed by Asami Tanimura. 58

Figure 2.4. Transfected SST-Cre mice show no off-target labeling. The striatum of SST-IRES-Cre mice was transfected with a floxed construct expressing ChR2- eGFP. Slices were post-fixed and immunolabeled for PV. (A) A map image showing the extent of spread of the ChR2-eGFP virus and immunolabled PV interneurons (red). The needle track and surrounding viral expression can be seen in the cortex. (B) Zoomed in, 10X, images from the image in panel A corresponding to the yellow boxed regions. Note that there is no apparent co- localization between the PV and ChR2-eGFP expressing SST interneurons. (C) Representative 60X images from the same slice showing no co-localization between the PV and eGFP markers. 60

Figure 2.5. Local interneurons generate NO-LTD in a select population of SPNs. (A) A PLTSI in brain slices prepared from a SST-IRES-Cre mouse in which a ere-dependent ChR2 construct has been stereotaxically delivered (referred to as SST-ChR2 mice) was patched using a Cs+ internal. The cell was subjected to the same 15Hz stimulation protocol that was used to induce NO-LTD in SPN s. The graph demonstrates the recorded PL TSI photocurrent over 1 minute of stimulation, normalized to the first 10% (6 s) of recording time. Circles are averages of responses within a 1 s window. (Avg 1-5 sec= 100.9% baseline; Avg 55-60 sec= 97.9% baseline; Q1=97.48%; Q3=99.6%). p>0.05, MannWhitney test. (B) Two SPNs from another SST-ChR2 mouse were patched. Approximately 5 minutes after break-in, both were briefly stepped from -70 to Om V. PLTSis were stimulated via full field LED illumination for l-5ms and then brought back down to -70 mV, GABAzine applied, and the LTD paradigm described in figure 2 was run. SPNs that exhibited GABAergic inhibitory postsynaptic currents (IPSCs) (top panel) exhibited NO-LTD whereas those that did not show an IPSC (bottom panel) did not exhibit LTD. (C) A summary plot showing average firing rates recorded from PLTSis in SST-ChR2 mice in the cell-attached recording configuration. PLTSI firing rates were consistently lower than the 15 Hz stimulation paradigm (Mean firing frequency and standard deviation= 4.57±3.05 Hz; range: 0.5472 to 10.58 Hz, n = 12). (D) Summary data demonstrating the relationship between the amplitude of the GABAergic IPSCs and interneuron dependent LTD. Stimulating PLTSis at lower frequencies ( 10 Hz) did not cause as robust LTD as higher (15 Hz) frequencies (For cells with IPSC > 10 pA, 10 Hz: depression in 3/5 cells, potentiation in 2/5 cells; 15 Hz: depression in 5/6 cells). (E) PLTSI induction of NO-LTD was occluded by pre-incubation with SNAP (100 µM). (Average min 1-5 = 100.38% baseline; min 30-35 = 93.68% baseline; QI =93.82%; Q3=96.22%. n = 8). p>0.05, Wilcoxon signed rank test. Scale bars: (A) 50 pA, 20 ms. (C) Data are represented as median, first and third quartiles, minimum and maximum of data. (E) Data are represented as median, first and third quartiles.

Data contributed by Asami Tanimura.

NO-LTD was independent of eCb signaling and was expressed post-synaptically

Phenomenologically, the LTD induced by NO signaling (NO-LTD) was very similar to that induced by eCb signaling (eCb-LTD). However, the two forms of LTD had very different determinants. First, NO-LTD induced by cGMP analogs in iSPNs and dSPNs was unaffected by antagonizing D2 dopamine receptors with sulpiride (1 μM) (Fig. 6A). Second, NO-LTD induced by cGMP analogs in iSPNs and dSPNs was unaffected by antagonizing CB1 eCb receptors with AM251 (2 μM) (Fig. 6A). Third, NO-LTD was unaffected by dialyzing neurons with the Ca2+chelator BAPTA (20 mM) (Fig. 6C). Each of these interventions is known to block eCb- LTD (Kreitzer and Malenka 2008, Mathur and Lovinger 2012). Another difference between these forms of plasticity was that NO-LTD was readily induced at thalamostriatal glutamatergic synapses, in contrast to eCb-LTD (Wu, Kim et al. 2015) (Fig. 6C-D).

One of the most important features of eCb-LTD is that while it is induced post- synaptically in SPNs, it is expressed pre-synaptically as a reduction in glutamate release probability (Kreitzer and Malenka 2008, Mathur and Lovinger 2012). Four observations support the proposition that unlike eCb-LTD, NO-LTD was expressed post-synaptically. First, the expression of NO-LTD was not accompanied by a change in paired-pulse facilitation, a hallmark of pre-synaptic LTD (Fig. 6E). Second, the induction of NO-LTD was independent of presynaptic activity; cessation of afferent fiber stimulation during the application of 8Br-cGMP had no effect on the magnitude of NO-LTD (Fig. 6F), unlike eCb-LTD (Adermark, Talani et al. 2009). Third, intracellular dialysis of D15 – a polypeptide previously shown to disrupt the interaction between dynamin and amphiphysin, which is crucial for AMPAR endocytosis (Carroll, Beattie et al. 1999, Morishita, Marie et al. 2005) – blocked NO-LTD expression while dialysis of a scrambled peptide had no effect on NO-LTD (Fig. 7A-B). Moreover, D15 dialysis did not alter eCb-LTD induced by the metabotropic glutamate receptor agonist (S)-3,5- dihydroxyphenylglycine (DHPG, 100 μM) (Fig. 7B). Fourth, transient application of 8Br-cGMP persistently decreased the amplitude of EPSCs evoked by uncaging of glutamate at visualized SPN spines (Fig. 7C-7F).

Figure 2.6. NO-LTD is independent of conventional determinants of eCB-LTD. (A) Application of 8-Br-cGMP (500 f-LM) induced LTD in the presence of the D2R antagonist sulpiride ( 1 f-LM) or in the presence of CB 1 receptor antagonist AM25 l (2 f-LM) in both dSPNs and iSPNs (Sulpiride: dSPNs: 74.72% baseline; Ql=68.53%; Q3=79.93%, n = 7; iSPNs: 77.95% baseline; Ql=69.12%; Q3=82.54%, n = 8; AM251: dSPNs: 63.27% baseline; Ql=49.12%; Q3=78.36%, n = 8; iSPNs: 58.69% baseline; Ql=45.79%; Q3=69.96%, n = 7). * p < 0.05, signed rank test. (B) Activation of PKG by 8-BrcGMP (500 f-LM) induced LTD in the presence of 20 mM BAPTA, loaded via the recording electrode (Post 8Br, min 25-30: 77.0% baseline; QI =72.0%; Q3=79.7%; n = 7). *p < 0.05, signed rank test. (C) Photomicrograph depicting the location of virally delivered ChR2 in the thalamus. (D) Activation of PKG by 8-Br-cGMP (500 µM) induced LTD at optically stimulated thalamic synapses (top panel, 61.78% baseline; QI =51.72%; Q3=65.71 %, n = 6). *p < 0.05, Wilcoxon signed rank test. (E) NO-LTD induced by 8- Br-cGMP did not significantly alter the paired pulse ratio of SPNs (left panel, Pre-LTD: 1.01±0.27, average and standard deviation; Post-LTD, 8Br-cGMP: 1.16±0.33, n = 5). p > 0.05, Wilcoxon signed rank test. Population time courses are shown at right. Traces above show a representative cell in which depression of evoked EPSCs, and no change in PPR were seen. (F) Cessation of cortical stimulation during 8-Br-cGMP application had no effect on LTD induction (75.21 % baseline; Ql=66.45%; Q3=84.61 %, n = 6). *p < 0.05, signed rank test. Scale bars: (C) 1 mm; (D) 10 pA, 10 ms (top), 20 pA, 10 ms (bottom); (E) 50 pA, 20 ms. (A) Data are represented as median, first and third quartiles, minimum and maximum of data. (B, D, E-F) Data are represented as median, first and third quartiles.

Data contributed by Igor Rafalovich and Asami Tanimura.

Figure 2.7. NO-LTD is post-synaptically expressed. (A) 8-Br-cGMP (500 µM) induced NO-LTD which was blocked by intracellular application of the endocytosis-disrupting peptide, D15 (1-2 mM; control: black trace, n=6; D15: red trace, n=6). (B) Quantification of the population data shown in (A). A scrambled peptide (sD15) had no effect on 8-Br-cGMP NO-LTD (1-2 mM; n=6). As expected, sD15 did not disrupt LTD induced by DHPG (100 µM; n=6). * p<0.05, signed rank test. (C) Simplified diagram illustrating 2PLUG experiments. (D) Representative 2PLSM image of iSPN dendritic spine at which 2PLUG was performed (blue circle). (E) 8-Br-cGMP induced a long-lasting decrease in uEPSC amplitude. Example uEPSCs before (pre-inc) and after transient bath application of 8-Br-cGMP (500 µM) are shown to the right. Time matched control traces (no 8-Br-cGMP application; post-control) are shown for comparison. * p<0.05, signed rank test. (F) Summary data for 8Br-cGMP effect in SPN dendritic spines (8-Br-cGMP: n=26 spines, 5 cells; control: n=29 spines, 3 cells). Scale bars: (D) 3 μm; (E) 3 pA (right panel, top) and 5 pA, 50 ms (right panel, bottom). (A, E) Data are presented as median, 1st and 3rd quartiles.

Data contributed by Igor Rafalovich and Joshua Plotkin. 67

Taken together, these results provide strong evidence for a post-synaptic locus of NO- LTD expression, contrasting it with other forms of LTD described in the striatum.

NO signaling blocked eCb-LTD induction

Are NO-LTD and eCb-LTD completely independent? As shown above, NO-LTD does not depend upon signaling events (e.g., CB1R or D2R activation) known to be necessary for induction of eCb-LTD. But could NO signaling blunt eCb-LTD? This would help to explain why the two forms of LTD have not been distinguished to date. The core mechanisms controlling eCb-LTD are well described, making pursuit of this aim tractable (Fig. 9A) (Mathur and Lovinger 2012). As a first step toward characterizing their interaction, NO-LTD was induced by SNAP (100 μM) and a positive allosteric modulator of sGC (BAY-41-2272, 10 μM) (Stasch, Becker et al. 2001); BAY-41-2272 was used because it potentiated the ability of SNAP to induce LTD, was inhibited by Rp8Br, and had no effect on its own (Fig. 8A-B) (Stasch et al., 2001). SPNs were then held at -50 mV, afferent fibers stimulated at low frequency and the mGluR agonist DHPG (100 μM) applied; normally, this is a very robust means of inducing eCb-LTD (Surmeier, Plotkin et al. 2009, Mathur and Lovinger 2012) (Fig. 9B). However, in this situation, the protocol failed to induce a lasting change in EPSC amplitude (Fig. 9B), arguing that NO- LTD blocked eCb-LTD induction. Bath application of 8Br-cGMP had a very similar effect, disrupting eCb-LTD (Fig. 8C). The ability of NO signaling to disrupt eCb-LTD was not presynaptic in origin as direct application of the CB1R agonist WIN55,212-2 mesylate (WIN) induced a robust LTD in the presence of SNAP and BAY-41-2272 (Fig. 9C). The other possibility is that NO signaling blunted eCb generation. The production of eCbs by SPNs requires the opening of L-type Ca2+ channels and an elevation in intracellular Ca2+concentration (Mathur and Lovinger 2012). In endocrine cells, phosphorylation of L-type channels by PKG decreases their opening (Mahapatra, Marcantoni et al. 2012), raising the possibility that NO signaling was suppressing dendritic L-type channels required for sustained eCb synthesis.

Indeed, co-application of SNAP (100 μM) and BAY-41-2272 (10 μM) decreased SPN intraspine Ca2+transients evoked by somatic depolarization (Fig. 9D and 9E). Antagonism of L- type channels with isradipine eliminated the effects of SNAP/BAY on intraspine Ca2+ transients, confirming the L-type specificity of the modulation (Fig. 9D). Thus, NO signaling disrupted the induction of eCb-LTD at least in part by attenuating activity-dependent, post-synaptic Ca2+ entry through L-type channels (Fig. 9F).

Discussion

The data presented make the case that NO release by PLTSIs induces a post-synaptically expressed form of LTD at SPN axospinous glutamatergic synapses. Although common in other regions of the brain (Garthwaite and Boulton 1995, Holscher 1997, Susswein, Katzoff et al. 2004), this is the first demonstration that striatal SPNs manifest this form of plasticity, complementing post-synaptically expressed LTP (Surmeier, Plotkin et al. 2009). The molecular mechanisms underlying NO-LTD remain to be fully elucidated, but our results point clearly to AMPA receptor endocytosis. Elsewhere, PKG, which was necessary for NO-LTD, is known to regulate the activity of DARPP-32 and protein phosphatase 1 – both of which have been implicated in phosphorylation of the GluA2 AMPAR subunit (Walaas, Hemmings et al. 2011).

How does the existence of NO-LTD change our understanding of the striatal network? PLTSIs appear to be part of a corticostriatal feed-forward circuit that is activated in parallel with SPNs (Tepper, Tecuapetla et al. 2010). Thus, NO-LTD could counterpoise dopamine-dependent potentiation of glutamatergic synaptic strength in SPNs during goal-based learning, helping to maintain a stable level of regional responsiveness by diminishing the strength of inactive synapses neighboring those undergoing potentiation. For this mechanism to work, the signaling mechanisms governing LTP induction would have to oppose those of NO-LTD.

Figure 2.8. NO-LTD is amplified by BAY-41. (A) SNAP induced, NO-LTD (grey trace) is enhanced in the presence of the NO-independent activator of sGC, BAY 41-2272 (black trace, control data set from figure 1 C. SNAP Ctr!: 79.24% baseline, Q 1 =70.46%, Q3=89.50%, n = 14 cells. SNAP+BAY: 53.74% baseline, Ql=38.36%, Q3=66.46%, n = 5). *p < 0.05, Mann Whitney U test. Like SNAP induced NO-LTD, this enhancement can be blocked by inclusion of an inhibitor of PKG in the recording pipette (red trace, 144.6% baseline; QI =88%; Q3=184%, n = 5). *p < 0.05 Mann Whitney U test comparing SNAP/BAY to SNAP/BAY+ PKG inhibitor. (B) Summary data demonstrate the effects of SNAP, SNAP+ BAY41-2272, and inhibition with Rp-8-Br-PET-cGMPs. *p < 0.05, Mann Whitney nonparametric test. (C) 8-Br-cGMP inhibited eCB-LTD, similar to SNAP mediated inhibition (88.51 % baseline; Ql=77.29%; Q3=107.44%, n = 8). p > 0.05, Wilcoxon signed rank.

Data contributed by Igor Rafalovich. 71

Figure 2.9. NO signaling occludes eCb-LTD through L-type current inhibition. (A) Simplified diagram outlining the mechanisms behind eCB-LTD. (B) Induction of eCb-LTD by DHPG (100 µM; black trace, n=6) was prevented by SNAP (100 μM) and BAY-41 (10 μM) perfused 10 min before and throughout recording (red trace: n=8). *p<0.05, rank sum test. Example EPSCs are shown to the right. (C) LTD induced by pharmacological activation of presynaptic CB1 receptors with WIN (2 µM) was not blocked by SNAP/BAY-41 (n=7). *p<0.05, signed rank test. (D) 2PLSM images of an iSPN (left) and dendritic spine from which a Ca2+ imaging line scan (middle, blue line) was performed. Bath application of SNAP/BAY-41 reduced peak calcium influx (n=26 spines, 3 cells). This reduction was occluded in the presence of 5 μM isradipine (n=22 spines, 3 cells). (E) Quantification of (D). * p<0.05, signed rank test. (F) Integrated diagram depicting the interaction between eCB and NO LTD. Scale bars represent: (B) 20 pA, 25 ms; (C) 40 pA, 20 ms; (D) 20 μm (left panel), 3 μm (middle panel), 4%∆F/Fo and 200 ms (right panel). (B-C) Data are presented as median, 1st and 3rd quartiles.

Data contributed by Igor Rafalovich and Joshua Plotkin. 73

As Ca2+ entry through NMDA receptors is necessary for LTP induction in SPNs, phosphodiesterase 1 – a striatally enriched, Ca2+/calmodulin dependent phosphodiesterase that preferentially degrades cGMP – might mediate this interaction (Bender and Beavo 2006, Shen, Flajolet et al. 2008). An additional possibility may be that NO released from PLTSIs also simultaneously tunes GABAergic inputs within the striatum, specifically from other interneurons onto SPNs, a possibility that has been described elsewhere (Nugent, Penick et al. 2007, Sagi, Heiman et al. 2014).

In disease states, deficits in NO-LTD could contribute to pathological remodeling of the striatal circuitry. For example, in models of Parkinson’s disease (PD), striatal NO signaling falls (Sancesario, Giorgi et al. 2004). The loss of feed-forward NO-LTD could contribute to hyperactivity of iSPNs and hypokinetic symptoms of the disease (Albin, Young et al. 1989). The sustained elevation in iSPN activity appears to drive profound homeostatic pruning of corticostriatal synapses, disrupting patterns of connectivity shaped by experience. An impairment in NO-LTD might also contribute to levodopa-induced strengthening of dSPN corticostriatal synapses and dyskinesia in PD (Day, Wang et al. 2006, Picconi, Bagetta et al. 2011, Fieblinger, Graves et al. 2014). How pathological changes in NO signaling preferentially effect dSPNs and iSPNs requires further study.

Chapter 3

Cholinergic interneurons modulate PLTSIs through muscarinic signaling

Introduction

The basal ganglia are a group of subcortical structures tasked with coordinating activities that range from executive motor control to the subconscious formation of habit (Shiflett, Brown et al. 2010, Gerfen and Surmeier 2011). A unique member of this group is the striatum, which serves as the primary intercessor of information arriving from nearly all cortical subdivisions, thalamic nuclei, and a spattering of other non-glutamatergic areas; and “output nuclei” of the basal ganglia including the globus pallidus (GP) and substantia nigra (SN) (Smith, Raju et al. 2004, Wall, De La Parra et al. 2013). The striatum is far more than a passive supplier of information; within the striatum afferent inputs undergo “processing” via complex striatal microcircuits in tandem with the cell autonomous mechanisms present at individual synaptic sites (Shiflett, Brown et al. 2010, Gerfen and Surmeier 2011).

In recent years the fundamental actors in these microcircuits have been more closely studied and have become better understood. The principle cells of the nucleus, the striatal projection neurons (SPNs), are undoubtedly important members given their predominance ( 90% of all striatal cells) and extensive axon collaterals (Guzman, Hernandez et al. 2003, Taverna,≈ Ilijic et al. 2008). However, striatal interneurons have also been shown to exert powerful influence over SPN output that exceeds their scarcity in number (Gittis, Leventhal et al. 2011, English, Ibanez- Sandoval et al. 2012). Once thought to be composed of only 4 subpopulations, recent work has revealed at least 4 additional interneuron types (Kawaguchi 1993, Ibanez-Sandoval, Tecuapetla et al. 2010, Ibanez-Sandoval, Tecuapetla et al. 2011, Faust, Assous et al. 2015, Munoz-Manchado, Foldi et al. 2016). Despite this heroic effort to identify and characterize these rare cells, we are still ignorant as to how many of them shape striatal activity. 76

Some of this uncertainty is likely due to the fact that we know even less about what extra and intrastriatal circuits modulate interneuronal activity. One of the most puzzling members of this cohort are the plateau and low-threshold spiking interneurons (PLTSIs). These cells express the most diverse compilation of neurotransmitters and peptides – somatostatin (SST), neuropeptide Y (NPY), nitric oxide (NO), and γ-amino butyric acid (GABA) – all of which have been shown to transform striatal activity (Galarraga, Vilchis et al. 2007, Lopez-Huerta, Blanco-Hernandez et al. 2012, Bari, Dec et al. 2015). However, their importance has been questioned given their presumably sparse connectivity with SPNs (Gittis, Nelson et al. 2010, Ibanez-Sandoval, Tecuapetla et al. 2011).

Recent work from our lab has cemented a series of work implicating PLTSIs and NO in striatal long-term depression (LTD) (Calabresi, Gubellini et al. 1999, Rafalovich, Melendez et al. 2015). Based on the results on this research, we posed the question; what was regulating PLTSI activity and consequentially, NO release? To address this, we mapped inputs to these cells using specific Cre-driver lines and a modified Rabies virus system (Wickersham, Lyon et al. 2007, Wall, De La Parra et al. 2013). We combined this approach with pharmacology, optogenetics, and imaging studies to show that, like other striatal cells, PLTSIs receive robust glutamatergic inputs from limbic cortices as well as intralaminar thalamic nuclei.

Although mapping studies demonstrated robust innervation by both nuclei, it is important to note that the two inputs generated divergent responses in PLTSIs. Specifically, thalamic activation had the overall effect of inhibiting PLTSi firing. While a component of this inhibition was gabazine-sensitive, the most profound effects were mediated by cholinergic interneurons (CHIs) acting on muscarinic acetylcholine receptors (mAChRs) expressed on PLTSIs. This inhibition was occasionally followed by seconds of oscillatory burst firing; both components being atropine- sensitive. Similar effects could be mimicked by optogenetic activation of cholinergic interneurons or bath application of muscarine but were unaffected my muscarinic toxin 7 (MT-7) and blocked

by muscarinic toxin 3 (MT-3); implicating the M2/M4 class of receptors in the effect. These studies present a novel circuit, from PFN-to-ChIs-to-PLTSIs, and a new function for striatal M4 receptors. Finally, because both the acetylcholine (ACh) and NO systems are altered in pathological states – 77

e.g. Parkinson’s disease (PD), or Levodopa-induced dyskinesia (LID) – our data suggests a possible role for PLTSIs and/or the PLTSI-ChI interaction in initiating or augmenting these states.

Materials and Methods

Animals. Adult mice (age 2 – 6 months) of both sexes were used for all experiments; mice hemizygous for the gene(s) of interest were used in all cases unless noted otherwise. Animals used for Rabies virus mapping experiments were SST-IRES-Cre mice and NPY-Cre mice; the latter strain was used as an alternative means of studying PLTSI afferents (Jackson Laboratory). Control rabies virus experiments to compare intrastriatal afferents onto iSPNs with PLTSIs utilized bacterial artificial chromosome (BAC) transgenic mice in which the Adora2a promoter drives Cre expression (A2a-Cre; GENSAT). To identify PLTSIs for pharmacological experiments or experiments in which cortex or thalamus was optogenetically activated, SST-IRES-Cre mice were crossed with Rosa26-STOP-tdTomato (Ai14) mice to generate SST-tdT+ mice (Jackson Laboratory). To study the effect of cholinergic excitation on PLTSI activity we generated bi-transgenics by crossing a BAC transgenic mouse expressing the humanized Renilla green fluorescent protein (NPY-eGFP; Jackson Laboratory) with mice expressing Cre-recombinase under control of the choline acetyltransferase (ChAT) regulatory elements (ChAT-Cre; Jackson Laboratory). Finally, WT C57/B6 mice were used for experiments to detect nitric oxide production. All procedures were performed with the approval of the Northwestern University ACUC and met NIH guidelines.

Slice Preparation. Adult SST-tdT+ mice, NPY-eGFP, NPY-eGFP::ChAT-Cre, and B57BL/6 mice of both sexes were deeply anesthetized with a mixture of ketamine/xylazine and perfused transcardially with ice-cold artificial cerebrospinal

fluid (ACSF) bubbled with 95% O2/5% CO2 and containing (in mM): 125 NaCl, 2.5

KCl, 1.25 NaH2PO4, 2 CaCl2 1 MgCl2, 25 NaHCO3, and 12.5 glucose. The brains 78

were rapidly removed and cut in the coronal (for optogenetic experiments) or sagittal orientation (for pharmacological experiments) at a thickness of 250 µm using a Leica VT1000S slicer. Slices were transferred to a holding chamber where they were incubated at 35°C for 30 minutes and then moved to a room-temperature solution for 30 minutes prior to recording.

Electrophysiology. Slices were transferred from the holding chamber to a recording chamber mounted on an Olympus BX51 upright, fixed-stage microscope where they were continuously perfused with oxygenated ACSF at a temperature between 32- 35°C. A 60X, 0.9 NA water-immersion objective and standard x-cite lamp were used to first identify tdTomato or eGFP fluorescent cells; the morphology of those cells expressing either fluorophore was then examined under standard differential contrast interference microscopy. Cells were patched in either the cell-attached, perforated- patch, or whole-cell recording configurations. The use of cell-attached and perforated-patch techniques was exclusively utilized to record PLTSI firing as the cells exhibited profound rundown and/or alterations in their firing patterns with time, in the whole-cell configuration. Additionally, we could not consistently observe a muscarinic modulation of PLTSI activity if the cell membrane was compromised. Whole-cell recordings were limited to recording synaptic currents evoked by optogenetic stimulation. Patch pipette resistance was between 3.5-5 MΩ. For perforated patch recordings cells were front- and back-filled with a K+-based internal solution composed of (in mM): 126 KMeSO4, 14 KCl, 10 HEPES, 5 EGTA, 0.5

CaCl2, and 3 MgCl2. pH was adjusted to 7.20-7.30 using NaOH, osmolarity was 270- 280 mOsm 1-1. Gramicidin was dissolved in DMSO at a concentration of 2 mg/mL and diluted to a working concentration of 20ug/mL in the filtered K+ internal via sonification. Cell-attached current-clamp recordings were made using ACSF-filled patch pipettes. Whole-cell recording solution contained (in mM): 120 CsMeSO3, 15 CsCl, 8 NaCl, 10 HEPES, 0.20 EGTA, 10 TEA-Cl, 5 QX-314, 4 Mg-ATP, 0.30 Na- 79

GTP. Electrophysiological recordings were obtained using a Multiclamp 700B amplifier (Molecular Devices) interfaced to a PC running pClamp9.2 software (Molecular Devices). Junction potential, 6-8 mV, was not corrected. Drugs were directly applied to the perfusion medium. For incubation experiments, slices were transferred to a separate holding chamber containing an appropriate dilution of drug/toxin before being transferred to the recording chamber and continuing to be perfused with drug. In many cases PLTSI activity was irregular in the first five minutes following sealing on and then became more regular following this time frame. Therefore, for consistency we used an arbitrary time period of 5 minutes after sealing on as the starting point for data collection.

Stereotaxic injections. Mice were anesthetized with isoflurane (1.5-3%; O2 1-1.2 L/min) and placed within a stereotaxic frame. SST-IRES-Cre and A2a-Cre mice underwent unilateral injections of AAV9-EF1a-DIO-HTB (Addgene 44187) at the following coordinates (all coordinates relative to bregma): 0.75 mm anterior, 1.65 mm lateral, 3.30 mm ventral. 90 nL of virus was injected over the course of 2 minutes and the needle left in place for 10 minutes before being slowly removed. 14 days later a EnvA-pseudotyped Rabies virus (RV) in which the native glycoprotein gene has been deleted (EnvA-RVΔG-mCherry) was injected at the same location but using a different angle of approach so as to minimize contamination outside the striatum. Rabies virus was allowed to express for 7 days before mice were used for immunohistochemistry. The cingulate cortex (CgCX) or parafascicular nucleus (PFN) of SST-tdT+ mice were injected with 90-180 nL of AAV2/9-hSyn- hChR2(H134R)-eGFP (University of Pennsylvania Vector Core; Addgene 26973) at a rate of 45 nL/min. CgCx and PFN coordinates were: 0.9 mm anterior, 0.3 mm lateral, 1.6 mm ventral; -2.3 mm anterior, 0.7 mm lateral, -3.35 mm ventral. NPY- eGFP x ChAT-Cre mice were injected with 270 nL of AAV2/9-EF1a-DIO- hChR2(H134R)-eYFP (University of Pennsylvania Vector Core; Addgene 20298) at 80

a rate of 45 mL/min and sacrificed for electrophysiological recordings 14-21 days later.

Preparation of fixed tissue, immunohistochemistry. Mice were deeply anesthetized with isoflurane and a lethal dose of ketamine/xylazine and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA) in PBS. Whole brains were dissected and post-fixed in 4% PFA in PBS at 4°C for 1 day before being sectioned (70-90 µm) in the coronal or sagittal confirmation using a Vibratome. For RV tissue, every third section was mounted on frosted slides and cover slipped using a DAPI- free glycerol mounting medium (Molecular Probes). Select sections from RV-SST- Cre and RV-A2a-Cre mice, along with SST-tdTomato+ and ChR2 injected SST-Cre mice underwent immunohistochemical processing. For ChAT staining: slices were incubated in PBS containing 5% normal donkey serum (NDS) and 0.3% Triton X-100 at room temperature (RT) for 1 h and then transferred to the same solution containing a primary choline acetyltransferase (ChAT) antibody raised in rabbit (Millipore 144P, 1:1000) and incubated at 4°C overnight. For PV staining: slices were incubated in 5% normal goat serum (NGS) and 0.3% Triton X-100 for 1 h at RT then incubated at 4°C overnight in the same solution containing an antibody against parvalbumin raised in mouse (Sigma P3088, 1:1000). Both sets of sections were washed 3x5” in PBS and subsequently incubated in an Alexa Fluor-conjugated secondary antibody diluted in the same blocking solution at 1:500. Slices were imaged with a laser-scanning confocal microscope (Olympus Fluoview FV1000 system).

Drugs. Drugs used included gabazine (10 µM, R&D Systems) to block GABAA

receptors, CGP 55845 (2 µM) to block GABAB receptors, mecamylamine (mec 10 µM, R&D Systems) to block nicotinic acetylcholine receptors. Atropine (1 µM) to block metabotropic acetylcholine receptors and physostigmine (10 µM) to block

acetylcholinesterase. AFDX-384 to indiscriminately block M2/M4 class receptors, 81

muscarinic-toxin 3 and muscarinic-toxin 7 (100 nM, Peptides International) to block the M4 and M1 receptor subtypes, respectively, and VU0255, 035 to block M1 receptors.

Statistical analyses. Data analysis was performed using Clampfit 9.2 (Molecular Devices, Inc., Sunnyvale, CA), Igor Pro 6.0 (WaveMetrics, Lake Oswego, OR), and Matlab. Statistical analysis was performed using Prism (GraphPad Software). Compiled data is reported as mean±s.e.m except for figure 10 in which the median is reported. The following non-parametric statistical tests used were: Mann-Whitney rank sum (for unpaired samples); Wilcoxon rank sum test (paired samples). P > 0.05 was set as the threshold for achieving significance. For box plots, data are reported as median, first and third quartiles, and minimum and maximum of data set (whiskers) with outliers removed.

Results

Mapping monosynaptic inputs to striatal PLTSIs

To determine what innervated PLTSIs of the dorsal striatum, we used SST-IRES-Cre mice in conjunction with a modified Rabies virus (RV) system to selectively map afferent inputs onto these cells (Wickersham, Lyon et al. 2007, Taniguchi, He et al. 2011). We injected 90 nL of AAV9-DIO-HTB into the dorsal striatum of these mice, which allowed for expression of: a nuclear-targeted GFP, the avian receptor, TVA and the Rabies glycoprotein (RG) (Fig. 1A-C). The utility of this virus is three-fold: first, a second injection of an mCherry expressing Rabies virus that has been pseudotyped with the avian sarcoma leucosis virus (EnvA-RVΔG-mCherry) will only infect cells in which the TVA receptor is expressed. Secondly, the presence of the RG in these same cells – and its deletion in the Rabies virus – ensures that retrograde spread of RV will only occur in those cells in which the first virus has been successfully replicated. Third, 82

identification of the number of starter cells, or those cells whose inputs are being mapped by retrograde spread of RV, is facilitated by the co-localization of HTB-eGFP and RV-MCherry A similar strategy has been used previously to map inputs to classes of striatal projection neurons (SPNs) as well as subpopulations of cortical interneurons (Wall, De La Parra et al. 2013, Wall, De La Parra et al. 2016). Two weeks after the initial injection, a second injection of the modified Rabies virus (RV) was made using the same injection coordinates as the primary injection but using an alternate angle of approach so as to minimize the likelihood of creating “starter cells” outside of the dorsal striatum as the cortex also houses SST+ interneurons (Fig. 1A-C). One week following RV injection, mice were perfused with PFA and brain slices cut in the coronal or sagittal conformation; cells that expressed mCherry alone were identified as presynaptic inputs to the starter cell population.

As was expected, these cells showed robust monosynaptic RV+ cells in the cortex (Partridge, Janssen et al. 2009, Gittis, Nelson et al. 2010, Ibanez-Sandoval, Tecuapetla et al. 2011). These inputs were preferentially located in the cingulate and secondary motor cortices (Fig. 1D, n = 3 animals) and projected exclusively from the side ipsilateral to the injection site suggesting there is no innervation from contralaterally projecting intratelencephalic (IT) neurons (Kress, Yamawaki et al. 2013). In 2/3 mice we noted sparse (5-7 cells) in the ipsilateral motor cortex. Additionally, we noted a strong labeling pattern in the thalamus localized to the parafascicular nucleus (PFN, Fig. 1E), a member of the intralaminar nuclei (Smith, Raju et al. 2009). Although the result is previously undescribed, it was not unexpected given the robust inputs that other striatal interneurons – cholinergic (ChIs) and fast-spiking (FSIs) – receive from this region (Ding, Guzman et al. 2010, Sciamanna, Ponterio et al. 2015). Notably, our RV labeling was consistent with earlier reports indicating that these cells receive inputs from both the external segment of the globus pallidus (GPe) and dopaminergic inputs from midbrain dopamine neurons (Fig. 1F-G) (Bevan, Booth et al. 1998, Saunders, Huang et al. 2016).

SST-tdTomato+ cells had characteristics of PLTSIs 83

Although past studies have looked at the effects of intrastriatal activation on PLTSI activity, the results are limited by the recording configuration and inability to specifically activate a population of afferents (Gittis, Nelson et al. 2010, Ibanez-Sandoval, Tecuapetla et al. 2011). To address this issue we crossed SST-IRES-Cre mice with a floxed-tdTomato reporter mouse to generate SST-tdT+ mice (Fig. 2A, left panel) in which we injected a Cre-independent channelrhodopsin (ChR2) construct (Madisen, Zwingman et al. 2010, Taniguchi, He et al. 2011). As was reported in cortex, this cross labeled SST+ PLTSIs but also demonstrated some off target labeling of two other cell populations that had electrophysiological characteristics of NPY- neurogliaform (NPY-NGF) interneurons and FSIs (data not shown) (Hu, Cavendish et al. 2013). Consistent with this observation, we saw sparse co-localization between tdTomato+ cells and cells immunolabeled for PV (Fig. 3). However, we were able to reliably identify PLTSIs based on their fluorescent morphology (Fig. 2A, right panel) and electrophysiological profile in the perforated-patch configuration which revealed spontaneous activity, rebound spiking and TTX- insensitive oscillations (Fig. 2B-C) (Beatty, Sullivan et al. 2012, Song, Beatty et al. 2016).

Furthermore, as was previously reported, these cells showed a robust response to D1/5 receptor stimulation (Fig. 2D-F) that resulted in a significant increase in their firing (Centonze, Bracci et al. 2002). Therefore, SST-tdT+ mice allowed for identification of PLTSIs. Nevertheless, we verified our results (Fig. 4-5, 8) in a different line in which both PLTSIs and NPY-NGF cells are labeled (NPY-eGFP, Jackson) but can be distinguished by spontaneous activity among other characteristics.

Cortical and thalamic stimulation differentially pause PLTSI firing

Given the robust labeling in PFN (Fig. 1E) and in as much as the cells are spontaneous pacemakers that sit close to spike threshold, we hypothesized that both cortical and thalamic inputs would drive PLTSI firing as the cells are. To study this, we targeted ChR2 injections to the CgCx or PFN (Fig. 4A-B) and recorded from identified PLTSIs in the perforated-patch and cell-attached recording configuration 14 days post injection. In cingulate injected mice, optical excitation (20 Hz, 0.5 ms, 400-500 ms duration) evoked action potentials in most PLTSIs that 84

closely followed the optical stimulus (Fig. 4C; responding PLTSIs: n = 24/27 cells). In some cells (n = 15/24 responders) this burst of action potentials was followed by a brief cessation of firing (Fig. 4C-D). The length of the pause appeared to be correlated to the number of spikes in the burst (Fig. 4D; R2=0.27). In other striatal interneurons that show burst-pause responses to afferent stimulation – e.g. ChIs – this pause is dependent on dopamine (Ding, Guzman et al. 2010). However, application of dopaminergic antagonists failed to modulate the pause (n = 5/5 cells, data not shown). Like ChIs, current injection through the recording electrode can also evoke a burst-pause response in PLTSIs (Beatty, Sullivan et al. 2012). Therefore, we concluded that the CgCx evoked pause was partially due to cell intrinsic mechanisms, initiated when cells were driven to spike by excitation of cortical inputs.

Comparatively, the PLTSI response to PFN activation was more variable. In many cells no response (change in firing) was seen (non-responding PLTSIs: n = 8/28). This was not due to a lack of ChR2 expression as neighboring SPNs patched using a Cs+-based internal solution always exhibited robust synaptic currents (data not shown). In 70% of PLTSIs (n = 14/20) that did respond to PFN stimulation, we observed a pause that was similar in duration to that evoked by CgCx stimulation, but without the preceding burst (Fig. 4E-F). Thalamic innervation of ChIs is particularly strong and has been demonstrated to induce a burst-pause response in these cells and subsequent increases in levels of acetylcholine (ACh) and dopamine (DA) as well as subsequent activation of intrastriatal GABAergic interneurons (Sullivan, Chen et al. 2008, Threlfell, Lalic et al. 2012). Therefore, we hypothesized that the PFN-induced pause in PLTSI was mediated through striatal interneurons.

To test this hypothesis, we recorded from PLTSIs in the perforated-patch configuration while stimulating PFN. In 2/10 PLTSIs in which we observed a pause in firing in normal ACSF, the addition of blockers of inhibitory transmission (CGP 2 µm, gabazine 10 µm, mecamylamine 10 µm) eliminated the response. However, the pause in firing remained in 8/10 cells and was present in 3/12 other PLTSIs that were recorded in the blocking solution from the onset of recording (Fig. 4E). The gabazine-sensitivity was suggestive of a role for GABAergic 85

interneurons whereas the gabazine-insensitive component indicated that a parallel process was taking place to modulate PLTSI firing. These cells have been shown to express muscarinic acetylcholine receptors (mAChR) of the M1 and M2/M4 class, and have recently been shown to respond to tonic levels of ACh (Elghaba, Vautrelle et al. 2016). Therefore, we reasoned that this pause might be due to activation of mAChRs and that the variability in cell response was a consequence of variability in cholinergic activation by PFN and/or tonic ACh levels. If this was true, then enhancing acetylcholine levels should increase the number of cells exhibiting a pause response to PFN activation. In the presence of blockers of inhibitory transmission, the addition of physostigmine to the bath resulted in a complete pause or significant slowing of PLTSIs in response to PFN stimulation in 6/10 cells tested (Fig. 5A top and middle panel, 5C). Subsequent application of atropine eliminated this response (Fig. 5A lower panel, 5C). Consistent with what we expected given the RV labeling, and the small percentage of cells demonstrating increases in firing in response to PFN stimulation (Fig. 4E, right panels), when cholinergic and GABAergic signaling were blocked, a brief excitation in response to thalamic stimulation was revealed in most PLTSIs (Fig. 5C). We hypothesized that, though PFN inputs evoked excitatory synaptic currents in PLTSIs, the primary effect was to alter firing through activation of a disynaptic circuit involving both GABAergic interneurons and ChIs (Fig. 5B).

Given this result, we revisited cortical stimulation experiments; augmenting or blocking cholinergic signaling through physostigmine and atropine, respectively. While the addition of physostigmine did not significantly alter the length of the pause response following evoked bursting (data not shown) the addition of atropine completely abolished the pause in 3/6 cells interrogated (Fig. 6A-C). This result suggests that both cortical and thalamic stimulation can excite cholinergic interneurons and evoke ACh release which then modulates PLTSI activity in parallel with excitatory glutamatergic inputs. This result is in line with recent studies that have shown that ChI-evoked DA release can be detected following optogenetic excitation of motor cortex or the PFN (Kosillo, Zhang et al. 2016). We hypothesized that both inputs engaged ChIs and acetylcholinergic signaling to silence PLTSI firing, but cortical inputs were stronger as a consequence of either differential subcellular localization of synapses or differential postsynaptic 86

receptor compositions (Fig. 6D). Accordingly, recording from Cs+-loaded PLTSIs verified this latter hypothesis, revealing significant differences in the amplitudes of evoked AMPA and NMDA currents in PFN and CgCx injected mice, as well as the NMDA/AMPA ratios (Fig. 7A- B).

Intrastriatal RV+ inputs are ChAT-negative

Given the strong cholinergic modulation of PTLSI firing and robust intrastriatal RV+ labeling, we looked to see whether we could identify this latter population as ChIs. We immunolabeled sections from RV injected SST-Cre mice with an antibody against ChAT (Fig. 8B, bottom panels). In sections from 3 mice we saw no co-localization between ChAT labeled and RV+ labeled cells. To verify that this was not a phenomenon specific to PLTSIs, we repeated the RV experiments in Adora2a (adenosine A2a receptor)-Cre mice which allow for targeting of striatal iSPNs (Schiffmann, Jacobs et al. 1991, Gong, Doughty et al. 2007). Similar to what was seen in SST-Cre mice, we failed to detect co-labeled RV+ and ChAT+ striatal cells (Fig. 8A-B, top panel). These results were surprising given the robust innervation of iSPNs by ChIs and previous publications which have used Cre-independent RV to show that a population of intrastriatal afferents onto SPNs were ChAT+ (Descarries and Mechawar 2000, Salin, Lopez et al. 2009). It should be noted that the RV used in this study was significantly different from the modified system used here. Moreover, it seems that the nature of synapses dictate the efficiency of RV spread such that synapses that function through “volume transmission”, as ChIs are hypothesized to do, will be more difficult for the RV to spread across when compared with “typical” tightly opposed terminals (Wall, De La Parra et al. 2013).

Optical activation of ChIs reproduces the PFN-induced pause

As an alternative means of testing the theory that thalamic inputs indirectly modulate PLTSIs via ChIs, bi-transgenic mice were created by crossing NPY-eGFP by ChAT-Cre strains (van den Pol, Yao et al. 2009) (Jackson Labs). Into the dorsal striatum of these mice we injected 270 nL of an eGFP-tagged, Cre-dependent ChR2 (Fig. 9A). Fourteen to twenty-one days 87

following injection we recorded from PLTSIs of these mice. As was expected given our thalamic data, a train of stimulation interrupted firing in PLTSIs by causing a long hyperpolarization that resembled the effects of thalamic activation and, additionally, was sensitive to atropine (Fig. 9B-C). PLTSI-like cells of the cortex show similar pauses in firing which are mediated by activation of GIRK channels (Bacci, Huguenard et al. 2004). Because

PLTSIs have been shown to express M1 and M2/M4 class muscarinic receptors and because GIRKs are potently modulated by this latter class, it was hypothesized that the activation of these receptors were mediating the observed effects in PLTSIs (Ariano and Kenny 1989, Bernard, Normand et al. 1992, Bernard, Laribi et al. 1998). Accordingly, in the presence of antagonists of inhibitory and nicotinic transmission, incubation of slices with the M1 specific toxin, muscarinic toxin-7 (MT-7) demonstrated no effect on the ChI-evoked pause whereas incubation with the M4 specific toxin muscarinic toxin-3 (MT3) eliminated the effect (Fig. 9D). This data reveals that PFN stimulation indirectly produces a pause in PLTSI firing that is mediated by ACh release from ChIs.

Strong mAChR activation results in rebound bursting in PLTSIs

During the course of the thalamic stimulation experiments, we noticed that some PLTSIs had an especially strong response to PFN stimulation when cholinergic signaling was being boosted through physostigmine. Specifically, the strong hyperpolarization previously described was followed by 4.12-12.96 s of burst firing, as defined by a interstimulus interval coefficient of variation (I.S.I., CV) that exceeded 1 (n = 3/6 cells, Fig. 10A) (Beatty, Sullivan et al. 2012). We predicted that this shift in firing pattern was mediated by an especially strong efflux of ACh, or cholinergic innervation of those cells. Accordingly, bath application of muscarine (3-5 µm) consistently caused PLTSIs to shift their firing pattern from regular or irregular pace making to stereotyped bursting for the duration of drug application (responding cells: SST-tdT+ mice, n = 18/29 cells, Fig. 10B). The alteration in pattern evoked by either PFN stimulation or muscarine, and the similarity between the two manipulations, was seen clearly in a joint interspike interval plot and histogram (Fig. 10C-D top and bottom panels) and was recapitulated in PLTSIs 88

identified from NPY-eGFP mice (Fig. 11A-B). As we saw in optogenetic experiments, this

bursting was insensitive to blockade of M1 receptors, as muscarine caused similar charges in

firing in the presence of an antagonist of M1 receptors, VU0255035 (Fig. 12A-B) (Sheffler,

Williams et al. 2009). However, in the presence of a mixed class antagonist of M2/M4 receptors

or the M4 specific toxin MT-3, we observed no change in the patterning of PLTSIs in response to muscarine (Fig. 12C-F). Therefore, we concluded that mAChR stimulation of PLTSIs had a dual effect of halting firing or converting PLTSIs to a bursting mode, depending on acetylcholine levels.

Discussion

The data presented in this work present the first ever study of monosynaptic inputs onto the somatostatinergic class of striatal interneurons. Our results suggest that these cells, like SPNs or ChIs, receive robust innervation from cortical and thalamic regions as well as neuromodulatory structures (e.g. SNc). While the monosynaptic RV mapping is a powerful tool in understanding network connectivity, it is truly a first step: without functional data accompanying mapping studies, it is difficult to ascertain the importance or purpose of one input over another. Our optogenetic studies underscore this point: despite robust labeling in specific thalamic nuclei, excitation of ChR2-labeled cells produced low amplitude, unreliable responses in PLTSIs, particularly when compared with excitation of ChR2-labeled cortical neurons. Though there are numerous explanations for this, the most straightforward hypothesis is that thalamostriatal inputs onto PLTSIs are highly divergent and/or cortical inputs are convergent (e.g. there is a one-to-one ratio between PLTSIs and an individual thalamic neuron whereas cortical neurons synapse on multiple PLTSIs). An additional factor may be the location of synapses within the PLTSI dendritic tree; because these cells have huge arbors that have frequently been reported to approach 1 mm, if thalamic inputs were targeted to the distal dendrites and cortical inputs to proximal sites, those inputs may be drastically attenuated even with Cs+-loaded cells in which dendrites are electrotonically shrunk. 89

Surprisingly, despite the small synaptic events, when thalamic inputs were stimulated while recording from PLTSIs in the cell-attached or perforated patch configurations, the impact of this nuclei had a much more profound effect. Specifically, we found that PLTSI activity could be halted. In extreme cases the cessation in firing was followed by seconds of burst firing before the cells reverted to normal pace making. Though a component of the pause was sensitive to blockers of inhibitory transmission, a majority of the pauses we recorded were eliminated by atropine. What’s more, in cells exhibiting no pause, enhancing cholinergic tone revealed this modulation to be present. Direct optogenetic activation of ChIs confirmed that this effect was

mediated by ChIs and, moreover, was dependent on the activation of M4 class muscarinic acetylcholine receptors. Bath application of muscarine in the presence of various antagonists of

subclasses of mAChRs underscored the dependence of M4 receptor activation on the generation of stereotyped bursting.

The significance of these results is threefold. First, we have verified earlier anatomical studies that suggested that M4 receptors, in addition to being present on dSPNs and ChIs, are also expressed on PLTSIs and have a powerfully modulatory effect on PLTSI patterning (Ariano and Kenny 1989). Secondly, we have described a novel relationship between these subclasses of striatal interneurons. Thirdly, the responses of PLTSIs to thalamic stimulation strongly resembled in vivo recordings from striatal tonically active neurons (TANs) in primates in response to salient stimuli; suggesting that both ChIs and PLTSIs have a role in learning (Aosaki, Graybiel et al. 1994, Aosaki, Kimura et al. 1995). Given the loss of this response in DA depleted states that model Parkinson’s disease (PD) our results are particularly important and suggest that PLTSIs are a contributor to disease pathology and/or the well documented pathological adaptations (Dehorter, Guigoni et al. 2009, Fieblinger, Graves et al. 2014). Though it is likely that this involves alterations in NO-dependent plasticity which PLTSIs are thought to control through synthesis and release of NO, this is speculation (Rafalovich, Melendez et al. 2015). 90

Figure 3.1. Retrograde labeling of monosynaptic inputs to striatal SST+ interneurons. (A) A simplified diagram of the Rabies virus (RV) system. SST-IRES-Cre mice were injected with a Cre-dependent AAV9-EF1a-DIO-HTB virus that allowed for expression of both the TVA receptor and RV glycoprotein (RG) exclusively in Cre-expressing cells. In the striatum, SST- expressing cells are limited to a population of interneurons known as plateau and low-threshold spiking interneurons (PLTSIs) that co-express neuronal nitric oxide synthase (nNOS) and neuropeptide Y (NPY). Two weeks later, a second injection of a modified RV (EnvA-RVΔG- mCherry) was injected at the same location using a different angle of approach to minimize contamination outside of the striatum. The RV can only infect cells in which TVA is expressed, and can only retrogradely spread in those that express the coat protein, RG, as well. (B-C) A representative image of the injection site. Cells that co-express eGFP and RV-mCherry represent the starter cell population for which monosynaptic inputs are mapped. (D) mCherry expressing neurons in M2/Cg cortex. (E) Labeling of thalamic inputs in the parafascicular nucleus (PFN). (F-G) mCherry expressing inputs from the GPe and SNc as previously described for these cells. 91

Figure 3.2. SST-tdTomato cells have characteristics of PLTSIs. (A) Schematic demonstrating the transgenic lines crossed to generate SST-tdTomato+ mice (top left); coronal section of the dorsal striatum demonstrating tdTomato fluorescence (bottom left); a reconstruction of a PLTSI patched during experiments utilizing a Cs+-based internal solution that were simultaneously filled with Alexa 568 (50 µm) to allow for morphological verification of cell type. (B) Fluorescent cells from SST-tdT+ mice had electrophysiological characteristics consistent with PLTSIs including spontaneous firing and rebound bursting with low-threshold spiking (C); TTX-insensitive oscillations. (D) SST-tdT+ cells showed a response to DA agonists that was typical of PLTSIs; SKF82958 (1 µm) resulted in increases in cell firing rate. (E) Joint consecutive I.S.I. plot and histogram of the PLTSI depicted in D. (F) Summary box plots of PLTSI responses to SKF82958 (Hz: Baseline, 8.70±2.061, SKF82958, 12.45±2.632; CV: Baseline, 0.23±0.04, SKF82958, 0.23±0.03; n = 6). * denotes statistical significance as defined by p < 0.05 using the Wilcoxon signed-rank test. Box plots represent 1st and 3rd quartiles, whiskers represent minimum and maximum of data. Scale bars: (B) 20 pA, 500 ms; (C)10 pA, 1 s, dotted line = -55 mV; (D) 10 mV, 1s, dotted line = -55 mV 93

Figure 3.3. SST x Ai14 mice demonstrate promiscuous labeling of striatal interneurons. (A) Coronal section from a SST-tdTomato+ mouse (middle panel) that has been immunolabeled for parvalbumin (PV, left panel) so as to identify fast-spiking interneurons (FSIs). (B) A close- up image of the dorsal striatum from the slice in A. White arrows identify cells that demonstrate co-labeling of parvalbumin and tdTomato, indicating off target expression of the fluorophore in interneuron populations other than PLTSIs. The white arrows indicate PV-expressing FSIs that are labeled with the tdTomato fluorophore. (C) A reconstructed image of cells within the dorsal striatum from A-B. The white arrow indicates a PV-expressing FSI that is labeled with the tdTomato fluorophore. 95

Figure 3.4. Cortical and thalamic stimulation induce differential responses in PLTSIs activity. (A) Cartoon demonstrating brain regions – 1. Cingulate cortex (CgCx), 2. Parafascicular nucleus (PFN) – into which channelrhodopsin (ChR2) was expressed in SST-tdTomato+ or NPY-eGFP mice (SST-tdT+: CgCx, n = 7 mice, PFN, n = 7 mice; NPY-eGFP: CgCx, n = 2 mice; PFN, n = 2 mice). (B) ChR2 expression in CgCx and PFN. (C-D) Example trace: 5 sequential sweeps from a PLTSIs patched in the cell-attached recording configuration in a CgCx injected mouse (left panel). CgCx stimulation consistently resulted in an increase in firing (right panel, responding cells: n = 24/27). (D) In some cells this bursting was followed by a pause in activity (left panel, average duration: 719±127 ms). The pause duration was correlated to the bursting frequency achieved during optical stimulation (right panel, R2=0.27, dotted lines are 99% confidence intervals). The bursting ratio was calculated by dividing the firing frequency during optical stimulation by the basal firing frequency; the pause ratio was calculated by dividing the first I.S.I. after cessation of optical stimulation by the average I.S.I., calculated from baseline firing frequency (right panel, average firing rate at baseline: 5.27±0.51 Hz; average rate frequency during optical stimulus: 13.97±1.55 Hz). (E) Example trace: 5 sequential sweeps from a PLTSI from a PFN injected mouse (left panel). Cells from these mice showed more variable responses with a large number not responding at all (right panel; n = 10/30 cells, 7 mice). The most common response was a pause in PLTSI firing following that was not correlated with a bursting generated by the optical stimuli, as was the case in CgCx mice (n = 14/30 cells, 7 mice). Application of CGP/gabazine failed to eliminate the pause in 8/10 cells tested, as did longer incubation times with the blockers prior to recording (right most panel). (F) The peristimulus time histograms (PSTH) and raster plots (below) for the two cells shown in C and E demonstrate similarities in the degree of the pause following stimulation, but differential response during the period of optical stimulation.

Figure 3.5. Thalamic activation interrupts PLTSI activity via activation of cholinergic interneurons. (A) A representative PLTSI, patched in the perforated-patch recording configuration showed delays in firing following optogenetic stimulation of PFN (data not shown) that were eliminated by application of inhibitors of GABAA, GABAB, and nACh receptors (top panel, Gabazine, 10 µm; CGP 55845, 2 µm; mecamylamine, 10 µm). Subsequent application of physostigmine (10 µm) revealed that PFN stimulation evoked a large, hyperpolarizing potential that halted cell firing for 2-6 sand was blocked by subsequent application of atropine (middle and bottom panels). The PSTH and raster plot for the same cell. PSTH include data presented in below raster plots, 5 sweeps per drug application. (B) Cartoon diagram of intra- and extrastriatal circuits in which PLTSIs are thought involved. Cells receive robust cortical inputs causing consistent excitation of PLTSIs. PFN inputs onto PLTSIs are smaller, generating postsynaptic currents but incapable of generating action potentials. PFN inputs, however, strongly innervate striatal cholinergic interneurons (ChIs) and other classes of GABAergic interneurons so that the overall effect of PFN activation is to indirectly modulate PLTSI firing through activation of disynaptic circuits involving

GABAA and mAChR mediated inhibition in firing. (C) Averaged population responses of PLTSIs showing that PFN stimulation generates a “pause” that is revealed by enhancing cholinergic signaling. (n = 6 cells). Error bars denote S.E.M. Scale bars: (A) 20 mV, 2 s.

100

Figure 3.6. Cortical activation also engages cholinergic interneurons to modulate PLTSIs. (A) A representative PLTSI recorded in the perforated patch configuration demonstrating a burst of action potentials during cortical stimulation followed by a pause in firing; the addition of physostigmine had no significant effect on the length of pause (top). The addition of atropine eliminated the pause while glutamatergic blockers eliminated the burst in firing during optogenetic stimulation (middle and bottom). (B) PSTH for the cell shown in (A) and associated raster plots demonstrating 5 sweeps of the four conditions color coded as indicated in the figure. (C) Population data demonstrates PLTSI firing was significantly different from baseline during the time of optical stimulation (t = 5.25-6 s, two-way ANOVA); no significance was found in time period following stimulation in any of the conditions tested (n = 6 cells). (D) Updated cartoon diagram demonstrating cortical modulation of PLTSIs: optogenetic activation of cortex is hypothesized to activate both ChIs and PLTSIs through postsynaptic glutamatergic receptors. Cortical inputs onto PLTSIs are strong and result in action potential generation followed by a pause: this pause is mediated through both cell autonomous (e.g. sAHP generation) and non-autonomous mechanisms. The latter is thought to be mediated by ChIs, also activated by cortical stimulation that modulate PLTSIs through mAChRs.

(B) Error bars denote S.E.M. Scale bars: (A) 20 mV, 2 s. 101

102

Figure 3.7. Synaptic properties of CgCx and PFN inputs onto PLTSIs. (A) Representative AMPA (black traces) and NMDA (red traces) currents recorded in Cs+-loaded PLTSIs from CgCx and PFN injected mice demonstrate differential ratios and maximum evoked amplitudes. (B) Population data demonstrating the recorded AMPA and NMDA currents in PLTSIs from CgCx and PFN injected mice (Cells: CgCx, n = 9, PFN, n = 10; Mean±S.E.M. AMPA current: CgCx, 139.3±21.70 pA, PFN, 7.72±1.99 pA; NMDA: CgCx, 79.05±13.52 pA, PFN, 14.15±2.73 pA; NMDA/AMPA Ratio: CgCx, 0.57±0.03, PFN, 2.11±0.29). *** denotes significance was found for p < 0.01, Mann-Whitney U test. Box plots represent the median, 1st and 3rd quartiles; whiskers represent the minimum and maximum data points. Scale bars: (A) 10 pA, 20 ms. 103

104

Figure 3.8. Intrastriatal RV+ cells are ChAT negative. (A) Coronal section from an A2a-Cre mouse (Cre expression driven under the Adora2a promoter, A2a adenosine receptor, allowing for targeting of iSPNs) that has been injected with the AAV9- Ef1a-DIO-HTB and Rabies viruses. A2a-Cre mice were subject to the same RV protocol as SST- IRES-Cre mice (Fig. 1). Slices from RV-injected A2a-Cre and SST-Cre mice were immunolabeled with an antibody against choline acetyltransferase (ChAT) to ascertain whether intrastriatal, monosynaptic inputs onto iSPNs or PLTSIs were ChIs. (B) ChAT-stained cells (Alexa 488, long arrows) can be easily differentiated from the DIO-HTB labeled cells (arrow heads) by the distinct morphology of ChIs and their cytoplasmic staining as compared to the nuclear GFP expressed by cells that have taken up the DIO-HTB virus (top left, bottom left panels). We observed no co-localization between ChAT and mCherry+ cells in either A2a-Cre (top panels) or SST-Cre (bottom panels) mice. (n = 3 mice). 105

106

Figure 3.9. ChI activation reproduces the PFN-induced pause in PLTSI firing. (A) Cartoon describing the mice used in these experiments: offspring of BAC NPY-eGFP crossed to ChAT-Cre mice underwent stereotaxic injections of a DIO-ChR2-eGFP targeted to the dorsal striatum (top panel). Specificity of ChR2 expression in CHIs was validated by post-fixing slices that were used for electrophysiological recordings, and immunolabeling for ChAT (bottom panel). (B) A representative PLTSIs; cells were identified by their strong, eGFP fluorescence and patched in the perforated patch or cell-attached recording configuration. In the presence of antagonists of GABAergic and nicotinic transmission (Gabazine, 10 µm; CGP 55845, 2 µm; mecamylamine, 10 µm) optical excitation of ChIs caused a strong hyperpolarization in PLTSIs that was similar to the response evoked by PFN stimulation and, similarly, blocked by nonspecific blockade of mAChRs via atropine application (1 µm). (C) Population PSTHs for PLTSIs recorded from NPY- eGFP::ChAT-Cre mice, compared to the response generated by PFN stimulation (PFN-ChR2: data set from Fig. 5C, ChAT-ChR2: n = 8 cells, 3 mice). (D) Slices were pre-incubated in MT-7; the toxin was included in the recording solution as well. Application of MT-7 had no effect on the ChI-evoked pause (n = 4 cells) whereas incubation with MT-3 significantly decreased the ChI- evoked pause (n = 6 cells). 107

108

Figure 3.10. Strong mAChR stimulation results in bursting in PLTSIs. (A) Example traces (5 sequential sweeps) from a PLTSI patched in a PFN injected animal. This cell exhibited little response to optogenetic excitation both before and after addition of blockers of inhibitory signaling. Addition of physostigmine resulted in a deep hyperpolarization that was followed by burst firing for 4-12 s following cessation of the stimulus. (B) A PLTSI from a SST- tdT+ mouse that exhibited stereotyped burst firing in response to muscarine (5 µM) application. This response was consistent in PLTSIs identified in either SST-tdT+ or NPY-eGFP mice (responding cells: SST-tdT+ mice, n = 18/29 cells; NPY-eGFP mice, n = 4/5 cells). (C-D) Joint interspike interval (I.S.I.) plots for the cells in panels A and B (top panels). Only the first two sweeps from panel A are plotted to so as to keep the number of events similar between stimulation paradigms (optogenetics vs. muscarine; events: n = 260, PFN-ctrl; n = 308, PFN-physo; n = 285, ctrl; n = 247, +musc). Histograms for the above cells (bottom panels). Scale bars: (A) 10 mV, 2 s; (B) 50 pA, 2 s. 109

110

Figure 3.11. PLTSIs identified from SST-tdT+ or NPY-eGFP mice respond similarly to muscarine. (A) Box plots show the change in frequency (left panel) and CV (right panel) evoked by application of muscarine in PLTSIs identified from SST-tdT+ or NPY-eGFP (greyed area) mice. SST-tdT+ (median before and after musc): 4.95 Hz, 5.88 Hz*. CV: 0.36, 0.63*. NPY-eGFP: 6.018 Hz, 14.35 Hz*; CV: 0.4159, 1.581*. (B) Example PLTSI identified in a NPY-eGFP mouse before (above) and during (bottom) application of 5 uM muscarine. * indicates significance at p < 0.05, Wilcoxon signed rank test. Box plots represent the median, and 1st and 3rd quartiles while whiskers indicate the minimum and maximum values from the data set. 111

112

Figure 3.12. PLTSI bursting is dependent on M4 activation. (A) A time course from an individual PLTSI that demonstrates the effect of muscarine on firing regularity as measured by the I.S.I. CV. The CV was calculated every 1 s with a 0.5 s overlap between measures. This cell was incubated in the M1 receptor blocker VU0255035 (5 µM) for at least 10 min prior to recording; the drug was kept on through the duration of the experiment. In the presence of M1 blockade muscarine still induced bursting (CV > 1) in the cell. Example traces from the indicated time points (a. and b.) can be seen to the right of the graph. (B) Box plots show the change in firing frequency and CV induced by muscarine application. The effect of muscarine alone (left plots, un-highlighted region, as reported in Fig. 11) is shown for comparison with muscarine in the presence of M1 blockade (right plots, highlighted region; median firing frequency and CV before and after muscarine: 4.72 Hz, 5.32 Hz; CV: 0.58, 1.20; n = 6). (C) Time course for a PLTSI recorded in the presence of the mixed M2/M4 receptor antagonist AFDX-384 (300 nM). (D) In the presence of AFDX, muscarine caused an increase in discharge rate but did not significantly alter the CV or discharge pattern (median firing frequency and CV before and after muscarine: 5.03 Hz, 9.71 Hz; CV: 0.33, 0.26; n = 9). (E) Time course for a PLTSI incubated and recorded in MT-3. (F) Box plot summary of PLTSI response to muscarine in the presence of M4 receptor blockade (median firing frequency and CV before and after muscarine: 3.69 Hz, 6.33 Hz; CV: 0.42, 0.26; n = 6). * indicates significance at p < 0.05 as compared to the baseline, Wilcoxon signed rank test. Box plots represent the median, and 1st and 3rd quartiles while whiskers indicate the minimum and maximum values from the data set. 113

114

Chapter 6

Discussion

In this thesis, we provide evidence that nitric oxide (NO) is directly responsible for inducing long term depression (LTD) at excitatory synapses impinging on striatal projection neurons (SPNs) of the dorsal striatum. Furthermore, we directly implicated a subpopulation of striatal interneurons - expressing neuronal nitric oxide synthase (nNOS) - in this form of plasticity. We accomplished this by demonstrating that optogenetically driving interneurons to firing rates higher than their normal frequency for short periods of time could, similarly, induce an NO-dependent form of LTD. This finding brings to a close a long series of debates regarding the importance of this molecule in striatal plasticity.

What might push SPNs towards this form of learning in vivo? I address this question in the second portion of my work in which I mapped inputs onto the cells responsible for the generation of NO; the nNOS-expressing plateau and low-threshold spiking interneurons (PLTSIs). I showed that the cells have a pattern of innervation that closely resembles the cholinergic interneuron (ChI), which are the other cells that make up the striatal tonically active neurons (TANs). Furthermore, I describe a novel circuit in which thalamic inputs can indirectly alter PLTSI firing, through ChI-mediated activation of postsynaptic M4 muscarinic acetylcholine receptors (mAChRs).

Earlier studies revealed that NO signaling has a robust excitatory effect on ChIs. Given this information, the results of our research and the abundance of work that has been done to reveal the many roles of mAChRs in the striatum, I predict that the significance of the PLTSI- ChI connection is to tune inputs impinging on the dendritic spines of SPNs and to push cells towards or away from synaptic depression. Furthermore, I hypothesize that PLTSIs, as a newly identified member of the TAN population, share a role in shaping the striatal response to strong thalamic stimuli (e.g. salient stimuli). These newly established roles will require a reexamination 115

of past work that has considered ChI the sole TAN within the nucleus. This retrospection will be crucial to our generating a more complete model of striatal networks, in both health and disease.

Nitric oxide mediates a postsynaptic form of LTD

Within the last ~20 years, considerable evidence has been put forward in support of the idea that NO – whether induced by NO donors or by enhancing signaling via application of precursors of NO – modulates the release of a number of neurotransmitters/modulators (Kiss 2000, Prast and Philippu 2001, West, Galloway et al. 2002). Therefore, though the results presented in Chapter 2 are novel, they are not exactly unexpected. This is particularly true given earlier experiments performed by Calabresi’s group that suggested the involvement of the NO- generating PLTSIs in striatal LTD (Calabresi, Gubellini et al. 1999, Centonze, Gubellini et al. 1999, Centonze, Grande et al. 2003). What makes this work stand out is twofold: first, we have used a combination of optogenetics and pharmacology to show that activation of PLTSIs can induce this form of plasticity and secondly, we have provided evidence that the LTD mediated by NO is distinct from the canonical striatal LTD mediated by eCBs (Mathur and Lovinger 2012).

The experiments in which optogenetic activation of PLTSIs was capable of eliciting LTD in SPNs were not as straightforward as our story indicates. When these experiments were first attempted we found tremendous variability in the SPN response to our stimulus. Only after going back to re-analyze this data did it become apparent that the cells exhibiting the most profound LTD were those same neurons in which the GABAergic innervation from PLTSIs was strong (Fig. 2.5). Aside from its use as a correlational tool, the significance of this observation is not entirely clear. NO synthase (NOS) has been demonstrated to localize to NMDARs via the PSD-95 scaffolding protein, indicating that NO is released in the dendrites (Bredt, Hwang et al. 1990, Garthwaite and Boulton 1995, Brenman, Chao et al. 1996, Christopherson, Hillier et al. 1999, d'Anglemont de Tassigny, Campagne et al. 2007). Contrast this to GABA which is canonically released from presynaptic axonal boutons (Kubota and Kawaguchi 2000). This then weakens the case that the correlation between IPSC amplitude and magnitude of NO-LTD is due 116

to “shared” innervation as the two neurotransmitters are presumably released from different compartments. However, in a separate line of experiments in which reconstructions of PLTSIs were made, we observed substantial overlap between these two structures which could help to explain our results. If this was the case, then the PLTSI generated IPSC was actually a measure of whether there were any interneurons in the vicinity of the SPN we were interrogating. Paralleling this idea is the possibility that the IPSC strength was a measure of how efficiently the ChR2 construct was transfected into interneurons within a given slice.

The second major point that we derived from the work shown in chapter 2 is that NO dependent plasticity is distinct from eCB mediated forms of LTD. This may seem unimpressive given the numerous neuromodulators that induce LTD, independent of CB1R activation (Calabresi, Gubellini et al. 1999, Sergeeva, Doreulee et al. 2007, Mathur, Capik et al. 2011, Atwood, Kupferschmidt et al. 2014). However, NO’s ability to do so was a moot point for the last 15-20 years. This was due, in part, to problematic stimulation paradigms that utilized high- frequency stimulation (HFS) and intrastriatally-placed electrodes; a caveat of the latter is that it induces the release of a multitude of neuromodulators via depolarization of intrastriatal presynaptic boutons which, in turn, influence whether SPNs undergo LTD, LTP, or neither. The experiments describing the NO-mediated inhibition of eCB-LTD are the first to demonstrate that the two forms of plasticity are distinct but intertwined. One of the many questions that remains to be answered is how this inhibition is mediated in vivo, and under what circumstances it is important. Are basal levels of striatal NO sufficient to reliability inhibit eCB generation and subsequent LTD? Or is the opposite case – eCB-LTD predominates until an efflux of NO initiates closure of L-type Ca2+ channels necessary for eCB synthesis – at play?

A solution to this question may be found in pharmacogenetics tools such as the “designer receptors exclusively activated by designer drugs” (DREADDs) or “pharmacologically selective actuator molecules” (PSAM) (Dong, Allen et al. 2010, Magnus, Lee et al. 2011). Pairing these constructs with SST- or NPY-Cre mice and in vivo behavioral assays which evaluate striatal learning (e.g. T-maze) would allow us to test whether either of these hypotheses are at work. An alternative line of research that demands attention is a more thorough examination of the 117

molecular mechanisms in SPNs that influence the NO/sGC/cGMP pathway. Nonselective inhibitors of the cGMP (and cAMP) degrading enzymes, PDEs, have already been shown to induce a form of LTD (Picconi, Bagetta et al. 2011). The advent of more selective compounds will allow us to examine how these enzymes interact with signaling cascades implicated in other forms of LTD (e.g. eCBs) or LTP to push SPN synapses to either state.

PLTSIs are modulated by ionotropic glutamatergic and metabotropic acetylcholine receptors

The results obtained in Chapter 2 beg the question: in a more physiological state (e.g. an intact, freely moving animal) what situations might provoke the induction of NO-LTD? While in vivo studies will be necessary to address this question, the scarcity of information on the intra- and extrastriatal connectivity of PLTSIs forced us to retreat back to a more basic inquiry. This question was straightforward: what regulates PLTSIs? More specifically, where do these cells receive inputs from and how do these afferents impact their tonic activity? Their simple dendritic arborization, paucity of spines, and intrinsic pace making abilities suggested an autonomy that is outside the normal realm of regulation by extrasynaptic inputs. Along this line of thought, it has been demonstrated that basal discharge rate and regularity of cells is unaffected by acute application of blockers of fast synaptic transmission (Beatty, Sullivan et al. 2012). Contradicting this is evidence from ChIs – also unaffected by synaptic blockers – which show alterations in their firing, in response to evoked afferent stimulation (Bennett and Wilson 1998, Beatty, Sullivan et al. 2012).

The initial images of the RV injected SST-Cre mice were somewhat surprising; anyone who has any experience patching PLTSIs is no doubt aware that it is astonishingly difficult to electrically evoke an EPSP/C in these cells that exceeds 20-30 pA, even when cells are loaded with a Cs+-based internal solution to electrotonically shrink dendrites, or the stimulating electrode is placed within the corpus callosum. The images demonstrating dense labeling of neurons within the cingulate and M2 regions of cortex seemed to suggest that this should not be the case. Subsequent optogenetic experiments in which PLTSIs were recorded in the cell- 118

attached configuration using a Cs+-loaded pipette and later broken into so as to record the amplitude of synaptic currents (Chapter 3) revealed that even in cells in which the maximally evoked EPSC was ≤ 30 pA, we could still reliably evoke APs in the cell-attached mode. If PLTSIs were more similar to SPNs – sitting at hyperpolarized potentials around -80 mV in the absence of any excitatory stimuli – than this result would be difficult to accept. However, as autonomous pace making cells, PLTSIs live at depolarized potentials interposed between APs and AHPs rendering them sensitive to even small perturbations in their membrane potential.

An additional surprise stemming from the RV experiments was there was little to no mCherry expression seen in the M1 region of cortex, despite past studies demonstrating that this region delivers a robust innervation to SPNs of the dorsal striatum (Kress, Yamawaki et al. 2013). This may be a result of experimental execution rather than biology. A limitation to the RV technique is that the detected, monosynaptic inputs are dependent on the starter cell population. I biased my striatal injections in the medial orientation – canonically delineated to be more aligned with associative and limbic basal ganglia loops – to decrease the chance of contamination outside the striatum; because of the substantial somatostatin expression in the cortex, this was easy to accomplish (Alexander, DeLong et al. 1986, Ma, Hu et al. 2006). It is well appreciated that corticostriatal synapses onto SPNs are elimination or reorganized in disorders of abnormal dopaminergic signaling such as Parkinson’s disease (PD) or chronic neuroleptic administration (Day, Wang et al. 2006, Sebel, Graves et al. 2016). An obvious question is whether striatal interneurons such as PLTSIs also undergo appreciable changes at their synaptic contacts with cortical nuclei. If this is the case, decreased excitatory drive may contribute to the depressed NO – and somatostatin – signaling that is noted to occur in PD (Eve, Nisbet et al. 1998).

Our electrophysiological data from cortically injected mice made the distinct response of PLTSIs to thalamic stimulation more difficult to reconcile. Despite a qualitatively similar number of RV labeled cells, when recording from PLTSIs in the cell-attached and then whole- cell configurations, I consistently found cells that exhibited a pause or decrease in discharge

when PFN was stimulation, that then had no discernible EPSC upon break in (Vm = -70 mV). 119

After collecting so much “negative” data, I began bringing cells to positive holding potentials; the rationale for this being that many hippocampal interneurons have “silent synapses” (NMDA containing but AMPA-silent) at multiple postnatal time points (Riebe, Gustafsson et al. 2009).

To my surprise, at Vm = +40 mV PFN stimulation now produced outward currents that were sensitive to Gabazine, though in some cases, a minute glutamatergic EPSC remained. We hypothesized that this PFN-generated, inhibitory response could be from either: 1.) a novel population of inhibitory thalamic neurons, or 2.) a disynaptic intrastriatal circuit. Through correspondence with a collaborator, and immunolabeling RV slices for GABA, we found that the second scenario was most likely responsible for the generation of this phenomenon. GABAergic inhibition, however, could not explain the pause response that remained in a majority of PLTSIs,

as the pause persisted in the presence of drugs that block both the nicotinic and GABAA and

GABAB receptors. A role for acetylcholine was hypothesized based on earlier publications from the lab that demonstrated a differential response in ChIs when thalamic versus cortical inputs were stimulated; the former coinciding with a burst of ACh release (Ding, Guzman et al. 2010). Despite the negative immunolabeling data that suggested no or – at most – sparse innervation of PLTSIs by ChIs (Chapter 3), the success of physostigmine to potentiate the PFN-evoked pause in PLTSIs was a clear indication that mAChRs modulate the activity of these cells. Around the same time these experiments were taking place, a study was published that suggested acetylcholine has a net, inhibitory effect over PLTSIs by increasing inhibitory drive onto these cells from GABAergic interneurons, while simultaneously inducing membrane hyperpolarizations (Elghaba, Vautrelle et al. 2016). These results were somewhat at odds with our study: though mAChR activation initially evoked strong hyperpolarizations in PLTS, in a subset of cells these hyperpolarizations were followed by profound burst firing; this high- frequency burst firing was also the predominate response to bath application of muscarine.

Furthermore, blockade of M2/M4 receptors revealed an excitatory effect of muscarine, mediated

by the Gq-coupled M1 receptors these cells have been shown to express (Ariano and Kenny 1989). This is more suggestive of an excitatory effect of acetylcholine. What could explain this discrepancy? One obvious difference between our lines of study is the age of mice used. In our 120

lab, striatal slices younger than p 60 are a rarity. Accordingly, the average age of mice used in these studies was p 63, over twice as old as the adolescent 25 day old animals used in the referenced publication. There is evidence that there is a selective increase in striatal M4 receptors in aged mice compared to a younger cohort (90 vs. 21 days) which could explain the discrepancy between our study and that from Bracci’s group (Tice, Hashemi et al. 1996). However, other groups have found contrasting results (Tayebati, Di Tullio et al. 2004).

Our ability to replicate the pause in PLTSI firing through direct excitation of ChIs affirmed the physiological relevance of this circuit. Although a majority of our data from these mice was generated in the presence of antagonists of nAChRs, in 3 cells where we recorded the response before and after mecamylamine application we did not see any change between the two states. This underscores previous studies that demonstrated that PLTSIs, relative to other GABAergic interneurons, have nominal nAChR expression and implicates mAChRs the primary means by which ChIs alter their activity (Luo, Janssen et al. 2013). Experiments in which we utilized pharmacology to validate the mAChR subtype responsible for the hyperpolarization and

burst firing, affirmed our optogenetic observations which implicated M4 receptors. Although the

evidence for M4 receptor expression on PLTSIs is not as unequivocal as the expression on dSPNs, the effect of MT-3 incubation in eliminating burst firing in the presence of muscarine, along with the optogenetic data from NPY-eGFP x ChAT-Cre mice is a strong indication that these cells do express pertinent amounts of the receptor (Ariano and Kenny 1989). It is

interesting that combining muscarine application with either this toxin or the mixed M2/M4 receptor antagonist AFDX-384 induced a statistically significant increase in cell firing. This

effect was linked to activation of M1 receptors. M1 receptors are well documented to enhance cell excitability in both hippocampal and cortical interneurons, and iSPNs (McCormick and Prince 1985, Shen, Hamilton et al. 2005, Shen, Tian et al. 2007, Yi, Ball et al. 2014). As is the

case in SPNs, the mechanism by which M1 receptors appear to be enhancing PLTSI activity is through closure of the M-current generating KCNQ channels and depolarization of the cell (data

not shown). This is well described. What is puzzling, is why the M1 effect of enhanced cellular discharge is more easily discerned using bath application of drugs as opposed to the direct 121

activation of ChIs or indirectly, via PFN stimulation. One explanation may be that in PLTSIs

there is a differential subcellular localization of mAChR classes; specifically, the M1 receptor is expressed at the soma and more proximally so is quicker to be appreciated by a somatic patch

pipette whereas M2/M4 receptors are in the dendrites. This hypothesis stems from multiple reports which have demonstrated differential localization of classes of mAChRs in both projection neurons and interneurons across many other nuclei (Nathanson 2008).

Before concluding, it is important to note that across all RV mice, we saw profound labeling of monosynaptic inputs within the GPe. This is in line with previous and current studies, that have described the preferential innervation that PV+ neurons of the GP have for striatal interneurons over SPNs (Bevan, Booth et al. 1998, Saunders, Huang et al. 2016).

PLTSI – CHI interactions: An explanation for heterogeneity of TANs in vivo?

The importance of ChIs in regulating normal striatal physiology was cemented and propelled into attention when Ann Graybiel’s group published a line of study that revealed that TANs, as opposed to phasically active neurons (PANs) a.k.a. SPNs, acquire a signature response during sensorimotor learning (Aosaki, Graybiel et al. 1994, Aosaki, Kimura et al. 1995). This response was described as a pause that followed the end of a conditioned stimuli (a “click” sound or LED light flash paired with a juice reward). The pause response was sometimes preceded by a burst of activity and/or followed by a period of rebound bursting. Upon review of the literature, among groups that have performed similar in vivo studies, only two have used immunohistochemical methods to verify that the recorded TANs were, indeed, ChAT+ cells (Inokawa, Yamada et al. 2010, Doig, Magill et al. 2014). Inokawa et. al’s study holds the limitation of not testing any behavioral paradigms to verify that a pause response occurs in these cells, and showed limited success in the juxtacellular neurobiotin labeling needed to stain cells (3/10 cells attempted). In Doig’s study, the group demonstrated impressive care and forethought to the issue of TAN identity and thus, only analyzed cells that were shown to stain positive for ChAT. This care however, was limited to the recordings from rats, and not the primate data in 122

this study. While Doig’s study certainly cements ChIs as TANs, they have not ruled out the possibility of PLTSIs contributing to this population as well. Both our results presented in chapter 2 and 3, and Aosaki’s original papers underscore this idea. In fact, one paper notes that based on the normal distribution of striatal interneurons (SST, PV, calretinin, ChIs) and their electrode placement (within the matrix or border of matrix and striosomes), microelectrode recordings will preferentially interrogate SST+ cells over ChIs by as much as 55 to 12. The validation of PLTSIs as distinct population of TANs suggests that these cells contribute to the learned pause in response to salient stimuli (Tepper, Tecuapetla et al. 2010, Beatty, Sullivan et al. 2012). If this was, in fact, the case, then the cellular plasticity that underlies this form of learning is likely dependent on multiple neuromodulators: dopamine, released by ACh acting on nAChRs on axon terminals; but also NO, SST and/or NPY released from PLTSIs driven to high- frequency bursting following thalamic activation.

A coordinated ACh – NO – DA response is not unheard of. In fact, it has long been appreciated that AChE and dihydronicotinamide adenine dinucleotide phosphate diaphorase (NADPH diaphorase, a marker of NOS) show significant colocalization with respect to striatal compartmentalization ((Sandell, Graybiel et al. 1986). In the nucleus accumbens this relationship is even stronger; NO has been shown to enhance DA release and this enhancement may be modulated by ACh (Hartung, Threlfell et al. 2011). It is tempting to speculate that the function of PFN and/or ACh-mediated bursting in PLTSIs is to release NO and amplify plastic processes coincident to salient stimuli. Alternatively, PLTSI activity could modulate plasticity through release of another neuromodulator, SST. SST release would serve to boost the function of acetylcholine: both decrease presynaptic release of neurotransmitter and both have been shown to enhance SPN excitability; SST doing so through reduction of SK and enhancement of BK channels (Galarraga, Vilchis et al. 2007).

A role for PLTSIs and NO in models of Parkinson’s disease and related pathologies 123

It is well appreciated that PD is a disease of both hypo-dopaminergic and hyper- cholinergic signaling (Duvoisin 1967, Shen, Plotkin et al. 2015). The utility of anticholinergic drugs in symptom management underscores this point (Lang 1984). Could altered ACh levels drive maladaptive changes to striatal architecture by way of NO? Under the premise that the Gi- coupled mAChRs are tied to PLTSI bursting and neuromodulator release, elevated ACh would be predicted to enhance this axis. However, over time it is suggested that mAChRs, in particular

M2s are endocytosed, so an increase in tonic ACh would eventually lead to ablation of the PLTSI response to ACh (Bernard, Laribi et al. 1998). This is one way in which NO-mediated LTD could be lost. The ability to induce this form of plasticity in PD mice through manipulation of intracellular signaling in SPNs is in line with this “presynaptic” model of NO loss by way of decreased synthesizing machinery (Picconi, Bagetta et al. 2011). The study of PLTSIs and NO plasticity in the DA denervated state will be necessary to test this hypothesis.

These studies are critical to complete our understanding of striatal physiology in diseased states. It has been found, repeatedly, that nNOS inhibitors (L-NAME) can decrease the severity of LID (Padovan-Neto, Echeverry et al. 2009, Takuma, Tanaka et al. 2012). Before we can tackle the mechanism behind this observation, and possibly the translation of this intervention, we must fully understand how these cells behave in the unperturbed disease state (PD) and how they are supposed to function, when DA levels are appropriate.

124

References

Adermark, L., G. Talani and D. M. Lovinger (2009). "Endocannabinoid-dependent plasticity at GABAergic and glutamatergic synapses in the striatum is regulated by synaptic activity." Eur J Neurosci 29(1): 32-41.

Ahern, G. P., V. A. Klyachko and M. B. Jackson (2002). "cGMP and S-nitrosylation: two routes for modulation of neuronal excitability by NO." Trends Neurosci 25(10): 510-517.

Albin, R. L., A. B. Young and J. B. Penney (1989). "The functional anatomy of basal ganglia disorders." Trends Neurosci 12(10): 366-375.

Alexander, G. E., M. R. DeLong and P. L. Strick (1986). "Parallel organization of functionally segregated circuits linking basal ganglia and cortex." Annu Rev Neurosci 9: 357-381.

Andre, V. M., C. Cepeda, Y. E. Fisher, M. Huynh, N. Bardakjian, S. Singh, X. W. Yang and M. S. Levine (2011). "Differential electrophysiological changes in striatal output neurons in Huntington's disease." J Neurosci 31(4): 1170-1182.

Aosaki, T., A. M. Graybiel and M. Kimura (1994). "Effect of the nigrostriatal dopamine system on acquired neural responses in the striatum of behaving monkeys." Science 265(5170): 412- 415.

Aosaki, T., M. Kimura, A. M. Graybiel and M. I. o. Technology (1995). "Temporal and Spatial Characteristics of Tonically Active Neurons of the Primate's Striatum." Journal of Neurophysiology 73(3): 19.

Ariano, M. A. (1983). "Distribution of components of the guanosine 3',5'-phosphate system in rat caudate-putamen." Neuroscience 10(3): 707-723.

Ariano, M. A. and S. L. Kenny (1989). "Striatal muscarinic receptors are associated with substance P and somatostatin containing neurons." Brain Res 497: 51-58. 125

Atwood, B. K., D. A. Kupferschmidt and D. M. Lovinger (2014). "Opioids induce dissociable forms of long-term depression of excitatory inputs to the dorsal striatum." Nat Neurosci 17(4): 540-548.

Bacci, A., J. R. Huguenard and D. A. Prince (2004). "Long-lasting self-inhibition of neocortical interneurons mediated by endocannabinoids." Nature 431(7006): 312-316.

Bagetta, V., B. Picconi, S. Marinucci, C. Sgobio, V. Pendolino, V. Ghiglieri, F. R. Fusco, C. Giampa and P. Calabresi (2011). "Dopamine-dependent long-term depression is expressed in striatal spiny neurons of both direct and indirect pathways: implications for Parkinson's disease." J Neurosci 31(35): 12513-12522.

Balleine, B. W., M. R. Delgado and O. Hikosaka (2007). "The role of the dorsal striatum in reward and decision-making." J Neurosci 27(31): 8161-8165.

Bari, A., A. Dec, A. W. Lee, J. Lee, D. Song, E. Dale, J. Peterson, S. Zorn, X. Huang, B. Campbell, T. W. Robbins and A. R. West (2015). "Enhanced inhibitory control by neuropeptide Y Y5 receptor blockade in rats." Psychopharmacology (Berl) 232(5): 959-973.

Barone, P., A. Antonini, C. Colosimo, R. Marconi, L. Morgante, T. P. Avarello, E. Bottacchi, A. Cannas, G. Ceravolo, R. Ceravolo, G. Cicarelli, R. M. Gaglio, R. M. Giglia, F. Iemolo, M. Manfredi, G. Meco, A. Nicoletti, M. Pederzoli, A. Petrone, A. Pisani, F. E. Pontieri, R. Quatrale, S. Ramat, R. Scala, G. Volpe, S. Zappulla, A. R. Bentivoglio, F. Stocchi, G. Trianni and P. D. Dotto (2009). "The PRIAMO study: A multicenter assessment of nonmotor symptoms and their impact on quality of life in Parkinson's disease." Mov Disord 24(11): 1641-1649.

Basso, M. A., A. S. Powers and C. Evinger (1996). "An explanation for reflex blink hyperexcitability in Parkinson's disease. I. Superior colliculus." J Neurosci 16(22): 7308-7317.

Beatty, J. A., M. A. Sullivan, H. Morikawa and C. J. Wilson (2012). "Complex autonomous firing patterns of striatal low-threshold spike interneurons." J Neurophysiol 108(3): 771-781. 126

Beaulieu, J. M. and R. R. Gainetdinov (2011). "The physiology, signaling, and pharmacology of dopamine receptors." Pharmacol Rev 63(1): 182-217.

Bender, A. T. and J. A. Beavo (2006). "Cyclic nucleotide phosphodiesterases: molecular regulation to clinical use." Pharmacol Rev 58(3): 488-520.

Bennett, B. D., J. C. Callaway and C. J. Wilson (2000). "Intrinsic membrane properties underlying spontaneous tonic firing in neostriatal cholinergic interneurons." J Neurosci 20(22): 8493-8503.

Bennett, B. D. and C. J. Wilson (1998). "Synaptic regulation of action potential timing in neostriatal cholinergic interneurons." J Neurosci 18(20): 8539-8549.

Bennett, B. D. and C. J. Wilson (1999). "Spontaneous activity of neostriatal cholinergic interneurons in vitro." J Neurosci 19(13): 5586-5596.

Bernard, V., O. Laribi, A. I. Levey and B. Bloch (1998). "Subcellular Redistribution of m2 Muscarinic Acetylcholine Receptors in Striatal Interneurons In Vivo after Acute Cholinergic Stimulation." J Neurosci 18(23): 10207-10218.

Bernard, V., E. Normand and B. Bloch (1992). "Phenotypical Characterization of the Rat Striatal Neurons Expressing Muscarinic Receptor Genes." J Neurosci 12(9): 3591-3600.

Bevan, M. D., P. A. Booth, S. A. Eaton and J. P. Bolam (1998). "Selective innervation of neostriatal interneurons by a subclass of neuron in the globus pallidus of the rat." J Neurosci 18(22): 9438-9452.

Bliss, T. V. and T. Lomo (1970). "Plasticity in a monosynaptic cortical pathway." J Physiol 207(2): 61p. 127

Bolam, J. P., B. H. Wainer and A. D. Smith (1984). "Characterization of cholinergic neurons in the rat neostriatum. A combination of choline acetyltransferase immunocytochemistry, Golgi- impregnation and electron microscopy." Neuroscience 12(3): 711-718.

Bologna, M., R. Agostino, B. Gregori, D. Belvisi, D. Ottaviani, C. Colosimo, G. Fabbrini and A. Berardelli (2009). "Voluntary, spontaneous and reflex blinking in patients with clinically probable progressive supranuclear palsy." Brain 132(Pt 2): 502-510.

Bouyer, J. J., D. H. Park, T. H. Joh and V. M. Pickel (1984). "Chemical and structural analysis of the relation between cortical inputs and tyrosine hydroxylase-containing terminals in rat neostriatum." Brain Res 302(2): 267-275.

Boyden, E. S., F. Zhang, E. Bamberg, G. Nagel and K. Deisseroth (2005). "Millisecond- timescale, genetically targeted optical control of neural activity." Nat Neurosci 8(9): 1263-1268.

Braak, H. and K. Del Tredici (2004). "Poor and protracted myelination as a contributory factor to neurodegenerative disorders." Neurobiol Aging 25(1): 19-23.

Bracci, E., D. Centonze, G. Bernardi and P. Calabresi (2002). "Dopamine excites fast-spiking interneurons in the striatum." J Neurophysiol 87(4): 2190-2194.

Bracci, E., D. Centonze, G. Bernardi and P. Calabresi (2003). "Voltage-dependent membrane potential oscillations of rat striatal fast-spiking interneurons." J Physiol 549(Pt 1): 121-130.

Bredt, D. S., P. M. Hwang and S. H. Snyder (1990). "Localization of nitric oxide synthase indicating a neural role for nitric oxide." Nature 347(6295): 768-770.

Brenman, J. E., D. S. Chao, S. H. Gee, A. W. McGee, S. E. Craven, D. R. Santillano, Z. Wu, F. Huang, H. Xia, M. F. Peters, S. C. Froehner and D. S. Bredt (1996). "Interaction of nitric oxide synthase with the postsynaptic density protein PSD-95 and alpha1-syntrophin mediated by PDZ domains." Cell 84(5): 757-767. 128

Brichta, L. and P. Greengard (2014). "Molecular determinants of selective dopaminergic vulnerability in Parkinson's disease: an update." Front Neuroanat 8: 152.

Calabresi, P., D. Centonze, P. Gubellini, A. Pisani and G. Bernardi (1998). "Blockade of M2-like muscarinic receptors enhances long-term potentiation at corticostriatal synapses." Eur J Neurosci 10(9): 3020-3023.

Calabresi, P., D. Centonze, P. Gubellini, A. Pisani and G. Bernardi (1998). "Endogenous ACh enhances striatal NMDA-responses via M1-like muscarinic receptors and PKC activation." Eur J Neurosci 10(9): 2887-2895.

Calabresi, P., P. Gubellini, D. Centonze, B. Picconi, G. Bernardi, K. Chergui, P. Svenningsson, A. A. Fienberg and P. Greengard (2000). "Dopamine and cAMP-regulated phosphoprotein 32 kDa controls both striatal long-term depression and long-term potentiation, opposing forms of synaptic plasticity." J Neurosci 20(22): 8443-8451.

Calabresi, P., P. Gubellini, D. Centonze, G. Sancesario, M. Morello, M. Giorgi, A. Pisani and G. Bernardi (1999). "A critical role of the nitric oxide/cGMP pathway in corticostriatal long-term depression." J Neurosci 19(7): 2489-2499.

Calabresi, P., R. Maj, A. Pisani, N. B. Mercuri and G. Bernardi (1992). "Long-term synaptic depression in the striatum: physiological and pharmacological characterization." J Neurosci 12(11): 4224-4233.

Calabresi, P., N. B. Mercuri and G. Bernardi (1993). "Chemical modulation of synaptic transmission in the striatum." Prog Brain Res 99: 299-308.

Calabresi, P., A. Pisani, J. Rothwell, V. Ghiglieri, J. A. Obeso and B. Picconi (2016). "Hyperkinetic disorders and loss of synaptic downscaling." 19(7): 868-875. 129

Carroll, R. C., E. C. Beattie, H. Xia, C. Luscher, Y. Altschuler, R. A. Nicoll, R. C. Malenka and M. von Zastrow (1999). "Dynamin-dependent endocytosis of ionotropic glutamate receptors." Proc Natl Acad Sci U S A 96(24): 14112-14117.

Cenci, M. A. and C. Konradi (2010). "Maladaptive striatal plasticity in L-DOPA-induced dyskinesia." Prog Brain Res 183: 209-233.

Centonze, D., N. Battista, S. Rossi, N. B. Mercuri, A. Finazzi-Agro, G. Bernardi, P. Calabresi and M. Maccarrone (2004). "A critical interaction between dopamine D2 receptors and endocannabinoids mediates the effects of cocaine on striatal gabaergic Transmission." Neuropsychopharmacology 29(8): 1488-1497.

Centonze, D., E. Bracci, A. Pisani, P. Gubellini, G. Bernardi and P. Calabresi (2002). "Activation of dopamine D1-like receptors excites LTS interneurons of the striatum." Eur J Neurosci 15(12): 2049-2052.

Centonze, D., C. Grande, E. Saulle, A. B. Martin, P. Gubellini, N. Pavon, A. Pisani, G. Bernardi, R. Moratalla and P. Calabresi (2003). "Distinct roles of D1 and D5 dopamine receptors in motor activity and striatal synaptic plasticity." J Neurosci 23(24): 8506-8512.

Centonze, D., P. Gubellini, G. Bernardi and P. Calabresi (1999). "Permissive role of interneurons in corticostriatal synaptic plasticity." Brain Res Brain Res Rev 31(1): 1-5.

Centonze, D., P. Gubellini, A. Pisani, G. Bernardi and P. Calabresi (2003). "Dopamine, acetylcholine and nitric oxide systems interact to induce corticostriatal synaptic plasticity." Rev Neurosci 14(3): 207-216.

Centonze, D., B. Picconi, P. Gubellini, G. Bernardi and P. Calabresi (2001). "Dopaminergic control of synaptic plasticity in the dorsal striatum." Eur J Neurosci 13(6): 1071-1077. 130

Centonze, D., E. Saulle, A. Pisani, G. Bernardi and P. Calabresi (2001). "Adenosine-mediated inhibition of striatal GABAergic synaptic transmission during in vitro ischaemia." Brain 124(Pt 9): 1855-1865.

Charpier, S. and J. M. Deniau (1997). "In vivo activity-dependent plasticity at cortico-striatal connections: evidence for physiological long-term potentiation." Proc Natl Acad Sci U S A 94(13): 7036-7040.

Chattopadhyaya, B., G. Di Cristo, H. Higashiyama, G. W. Knott, S. J. Kuhlman, E. Welker and Z. J. Huang (2004). "Experience and activity-dependent maturation of perisomatic GABAergic innervation in primary visual cortex during a postnatal critical period." J Neurosci 24(43): 9598- 9611.

Chevalier, G., S. Vacher, J. M. Deniau and M. Desban (1985). "Disinhibition as a basic process in the expression of striatal functions. I. The striato-nigral influence on tecto-spinal/tecto- diencephalic neurons." Brain Res 334(2): 215-226.

Christopherson, K. S., B. J. Hillier, W. A. Lim and D. S. Bredt (1999). "PSD-95 assembles a ternary complex with the N-methyl-D-aspartic acid receptor and a bivalent neuronal NO synthase PDZ domain." J Biol Chem 274(39): 27467-27473.

Connolly, B. S. and A. E. Lang (2014). "Pharmacological treatment of Parkinson disease: a review." Jama 311(16): 1670-1683.

Contant, C., D. Umbriaco, S. Garcia, K. C. Watkins and L. Descarries (1996). "Ultrastructural characterization of the acetylcholine innervation in adult rat neostriatum." Neuroscience 71(4): 937-947.

Conway, K. A., J. C. Rochet, R. M. Bieganski and P. T. Lansbury, Jr. (2001). "Kinetic stabilization of the alpha-synuclein protofibril by a dopamine-alpha-synuclein adduct." Science 294(5545): 1346-1349. 131

Cowan, R. L. and C. J. Wilson (1994). "Spontaneous firing patterns and axonal projections of single corticostriatal neurons in the rat medial agranular cortex." J Neurophysiol 71(1): 17-32.

Cui, G., S. B. Jun, X. Jin, M. D. Pham, S. S. Vogel, D. M. Lovinger and R. M. Costa (2013). "Concurrent activation of striatal direct and indirect pathways during action initiation." Nature 494(7436): 238-242.

Czubayko, U. and D. Plenz (2002). "Fast synaptic transmission between striatal spiny projection neurons." Proc Natl Acad Sci U S A 99(24): 15764-15769.

d'Anglemont de Tassigny, X., C. Campagne, B. Dehouck, D. Leroy, G. R. Holstein, J. C. Beauvillain, V. Buee-Scherrer and V. Prevot (2007). "Coupling of neuronal nitric oxide synthase to NMDA receptors via postsynaptic density-95 depends on estrogen and contributes to the central control of adult female reproduction." J Neurosci 27(23): 6103-6114.

Damodaran, S., J. R. Cressman, Z. Jedrzejewski-Szmek and K. T. Blackwell (2015). "Desynchronization of fast-spiking interneurons reduces beta-band oscillations and imbalance in firing in the dopamine-depleted striatum." 35(3): 1149-1159.

Day, M., Z. Wang, J. Ding, X. An, C. A. Ingham, A. F. Shering, D. Wokosin, E. Ilijic, Z. Sun, A. R. Sampson, E. Mugnaini, A. Y. Deutch, S. R. Sesack, G. W. Arbuthnott and D. J. Surmeier (2006). "Selective elimination of glutamatergic synapses on striatopallidal neurons in Parkinson disease models." Nat Neurosci 9(2): 251-259.

Day, M., D. Wokosin, J. L. Plotkin, X. Tian and D. J. Surmeier (2008). "Differential excitability and modulation of striatal medium spiny neuron dendrites." J Neurosci 28(45): 11603-11614.

de la Vega, M. T., A. Nunez and J. A. Arias-Montano (1997). "Muscarinic M1 and M3 receptors in rat striatum: a binding study." Arch Med Res 28(4): 493-497. 132

DeBoer, P. and E. D. Abercrombie (1996). "Physiological release of striatal acetylcholine in vivo: modulation by D1 and D2 dopamine receptor subtypes." J Pharmacol Exp Ther 277(2): 775-783.

Dehorter, N., C. Guigoni, C. Lopez, J. Hirsch, A. Eusebio, Y. Ben-Ari and C. Hammond (2009). "Dopamine-deprived striatal GABAergic interneurons burst and generate repetitive gigantic IPSCs in medium spiny neurons." J Neurosci 29(24): 7776-7787.

DeLong, M. and T. Wichmann (2009). "Update on models of basal ganglia function and dysfunction." Parkinsonism Relat Disord 15 Suppl 3: S237-240.

DeLong, M. R. (1990). "Primate models of movement disorders of basal ganglia origin." Trends Neurosci 13(7): 281-285.

DeLong, M. R. and T. Wichmann (2007). "Circuits and circuit disorders of the basal ganglia." Arch Neurol 64(1): 20-24.

Descarries, L. and N. Mechawar (2000). "Ultrastructural evidence for diffuse transmission by monoamine and acetylcholine neurons of the central nervous system." Prog Brain Res 125: 27- 47.

Devane, W. A., L. Hanus, A. Breuer, R. G. Pertwee, L. A. Stevenson, G. Griffin, D. Gibson, A. Mandelbaum, A. Etinger and R. Mechoulam (1992). "Isolation and structure of a brain constituent that binds to the cannabinoid receptor." Science 258(5090): 1946-1949.

Di Cristo, G., C. Wu, B. Chattopadhyaya, F. Ango, G. Knott, E. Welker, K. Svoboda and Z. J. Huang (2004). "Subcellular domain-restricted GABAergic innervation in primary visual cortex in the absence of sensory and thalamic inputs." Nat Neurosci 7(11): 1184-1186.

Ding, J., J. N. Guzman, T. Tkatch, S. Chen, J. A. Goldberg, P. J. Ebert, P. Levitt, C. J. Wilson, H. E. Hamm and D. J. Surmeier (2006). "RGS4-dependent attenuation of M4 autoreceptor 133

function in striatal cholinergic interneurons following dopamine depletion." Nat Neurosci 9(6): 832-842.

Ding, J., J. D. Peterson and D. J. Surmeier (2008). "Corticostriatal and thalamostriatal synapses have distinctive properties." J Neurosci 28(25): 6483-6492.

Ding, J. B., J. N. Guzman, J. D. Peterson, J. A. Goldberg and D. J. Surmeier (2010). "Thalamic gating of corticostriatal signaling by cholinergic interneurons." Neuron 67(2): 294-307.

Ding, J. D., A. Burette, P. I. Nedvetsky, H. H. Schmidt and R. J. Weinberg (2004). "Distribution of soluble guanylyl cyclase in the rat brain." J Comp Neurol 472(4): 437-448.

Doig, N. M., P. J. Magill, P. Apicella, J. P. Bolam and A. Sharott (2014). "Cortical and thalamic excitation mediate the multiphasic responses of striatal cholinergic interneurons to motivationally salient stimuli." J Neurosci 34(8): 3101-3117.

Doig, N. M., J. Moss and J. P. Bolam (2010). "Cortical and thalamic innervation of direct and indirect pathway medium-sized spiny neurons in mouse striatum." J Neurosci 30(44): 14610- 14618.

Dong, S., J. A. Allen, M. Farrell and B. L. Roth (2010). "A chemical-genetic approach for precise spatio-temporal control of cellular signaling." Mol Biosyst 6(8): 1376-1380.

Dubach, M., R. Schmidt, D. Kunkel, D. M. Bowden, R. Martin and D. C. German (1987). "Primate neostriatal neurons containing tyrosine hydroxylase: immunohistochemical evidence." Neurosci Lett 75(2): 205-210.

Duchemin, S., M. Boily, N. Sadekova and H. Girouard (2012). "The complex contribution of NOS interneurons in the physiology of cerebrovascular regulation." Front Neural Circuits 6: 51.

Duvoisin, R. C. (1967). "Cholinergic-anticholinergic antagonism in parkinsonism." Arch Neurol 17(2): 124-136. 134

Elghaba, R., N. Vautrelle and E. Bracci (2016). "Mutual Control of Cholinergic and Low- Threshold Spike Interneurons in the Striatum." Front Cell Neurosci 10: 111.

Ellender, T. J., J. Harwood, P. Kosillo, M. Capogna and J. P. Bolam (2013). "Heterogeneous properties of central lateral and parafascicular thalamic synapses in the striatum." J Physiol 591(1): 257-272.

English, D. F., O. Ibanez-Sandoval, E. Stark, F. Tecuapetla, G. Buzsaki, K. Deisseroth, J. M. Tepper and T. Koos (2012). "GABAergic circuits mediate the reinforcement-related signals of striatal cholinergic interneurons." Nat Neurosci 15(1): 123-130.

Eve, D. J., A. P. Nisbet, A. E. Kingsbury, E. L. Hewson, S. E. Daniel, A. J. Lees, C. D. Marsden and O. J. Foster (1998). "Basal ganglia neuronal nitric oxide synthase mRNA expression in Parkinson's disease." Brain Res Mol Brain Res 63(1): 62-71.

Fabbrini, G., J. M. Brotchie, F. Grandas, M. Nomoto and C. G. Goetz (2007). "Levodopa- induced dyskinesias." Mov Disord 22(10): 1379-1389; quiz 1523.

Faust, T. W., M. Assous, F. Shah, J. M. Tepper and T. Koos (2015). "Novel fast adapting interneurons mediate cholinergic-induced fast GABAA inhibitory postsynaptic currents in striatal spiny neurons." Eur J Neurosci 42(2): 1764-1774.

Feelisch, M. (1998). "The use of nitric oxide donors in pharmacological studies." Naunyn Schmiedebergs Arch Pharmacol 358(1): 113-122.

Fieblinger, T., S. M. Graves, L. E. Sebel, C. Alcacer, J. L. Plotkin, T. S. Gertler, C. S. Chan, M. Heiman, P. Greengard, M. A. Cenci and D. J. Surmeier (2014). "Cell type-specific plasticity of striatal projection neurons in parkinsonism and L-DOPA-induced dyskinesia." Nat Commun 5: 5316.

Foehring, R. C., X. F. Zhang, J. C. Lee and J. C. Callaway (2009). "Endogenous calcium buffering capacity of substantia nigral dopamine neurons." J Neurophysiol 102(4): 2326-2333. 135

Freeze, B. S., A. V. Kravitz, N. Hammack, J. D. Berke and A. C. Kreitzer (2013). "Control of basal ganglia output by direct and indirect pathway projection neurons." J Neurosci 33(47): 18531-18539.

Galarraga, E., C. Vilchis, T. Tkatch, H. Salgado, F. Tecuapetla, T. Perez-Rosello, E. Perez-Garci, E. Hernandez-Echeagaray, D. J. Surmeier and J. Bargas (2007). "Somatostatinergic modulation of firing pattern and calcium-activated potassium currents in medium spiny neostriatal neurons." Neuroscience 146(2): 537-554.

Gandhi, N. J. and H. A. Katnani (2011). "Motor functions of the superior colliculus." Annu Rev Neurosci 34: 205-231.

Garthwaite, J. (2008). "Concepts of neural nitric oxide-mediated transmission." Eur J Neurosci 27(11): 2783-2802.

Garthwaite, J. and C. L. Boulton (1995). "Nitric oxide signaling in the central nervous system." Annu Rev Physiol 57: 683-706.

Gelb, D. J., E. Oliver and S. Gilman (1999). "Diagnostic criteria for Parkinson disease." Arch Neurol 56(1): 33-39.

Gerdeman, G. and D. M. Lovinger (2001). "CB1 cannabinoid receptor inhibits synaptic release of glutamate in rat dorsolateral striatum." J Neurophysiol 85(1): 468-471.

Gerfen, C. R. (1989). "The neostriatal mosaic: striatal patch-matrix organization is related to cortical lamination." Science 246(4928): 385-388.

Gerfen, C. R. (1992). "The neostriatal mosaic: multiple levels of compartmental organization in the basal ganglia." Annu Rev Neurosci 15: 285-320. 136

Gerfen, C. R. and J. P. Bolam (2010). The Neuroanatomical Organization of the Basal Ganglia. Handbook of Basal Ganglia Structure and Function. H. Steiner and K. Y. Tseng. London, Academic Press/Elsevier: 3-28.

Gerfen, C. R., T. M. Engber, L. C. Mahan, Z. Susel, T. N. Chase, F. J. Monsma, Jr. and D. R. Sibley (1990). "D1 and D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal neurons." Science 250(4986): 1429-1432.

Gerfen, C. R., W. A. Staines, G. W. Arbuthnott and H. C. Fibiger (1982). "Crossed connections of the substantia nigra in the rat." J Comp Neurol 207(3): 283-303.

Gerfen, C. R. and D. J. Surmeier (2011). "Modulation of striatal projection systems by dopamine." Annu Rev Neurosci 34: 441-466.

Gerfen, C. R. and W. S. Young, 3rd (1988). "Distribution of striatonigral and striatopallidal peptidergic neurons in both patch and matrix compartments: an in situ hybridization histochemistry and fluorescent retrograde tracing study." Brain Res 460(1): 161-167.

Gertler, T. S., C. S. Chan and D. J. Surmeier (2008). "Dichotomous anatomical properties of adult striatal medium spiny neurons." J Neurosci 28(43): 10814-10824.

Gittis, A. H., G. B. Hang, E. S. LaDow, L. R. Shoenfeld, B. V. Atallah, S. Finkbeiner and A. C. Kreitzer (2011). "Rapid target-specific remodeling of fast-spiking inhibitory circuits after loss of dopamine." Neuron 71(5): 858-868.

Gittis, A. H., D. K. Leventhal, B. A. Fensterheim, J. R. Pettibone, J. D. Berke and A. C. Kreitzer (2011). "Selective inhibition of striatal fast-spiking interneurons causes dyskinesias." J Neurosci 31(44): 15727-15731.

Gittis, A. H., A. B. Nelson, M. T. Thwin, J. J. Palop and A. C. Kreitzer (2010). "Distinct roles of GABAergic interneurons in the regulation of striatal output pathways." J Neurosci 30(6): 2223- 2234. 137

Goetz, C. G. (2011). "The history of Parkinson's disease: early clinical descriptions and neurological therapies." Cold Spring Harb Perspect Med 1(1): a008862.

Gong, S., M. Doughty, C. R. Harbaugh, A. Cummins, M. E. Hatten, N. Heintz and C. R. Gerfen (2007). "Targeting Cre recombinase to specific neuron populations with bacterial artificial chromosome constructs." J Neurosci 27(37): 9817-9823.

Gracy, K. N. and V. M. Pickel (1997). "Ultrastructural localization and comparative distribution of nitric oxide synthase and N-methyl-D-aspartate receptors in the shell of the rat nucleus accumbens." Brain Res 747(2): 259-272.

Graybiel, A. M. (2000). "The basal ganglia." Curr Biol 10(14): R509-511.

Guo, J. and S. R. Ikeda (2004). "Endocannabinoids modulate N-type calcium channels and G- protein-coupled inwardly rectifying potassium channels via CB1 cannabinoid receptors heterologously expressed in mammalian neurons." Mol Pharmacol 65(3): 665-674.

Guzman, J. N., A. Hernandez, E. Galarraga, D. Tapia, A. Laville, R. Vergara, J. Aceves and J. Bargas (2003). "Dopaminergic modulation of axon collaterals interconnecting spiny neurons of the rat striatum." J Neurosci 23(26): 8931-8940.

Haber, S. N., K. S. Kim, P. Mailly and R. Calzavara (2006). "Reward-related cortical inputs define a large striatal region in primates that interface with associative cortical connections, providing a substrate for incentive-based learning." J Neurosci 26(32): 8368-8376.

Hakansson, K., S. Galdi, J. Hendrick, G. Snyder, P. Greengard and G. Fisone (2006). "Regulation of phosphorylation of the GluR1 AMPA receptor by dopamine D2 receptors." J Neurochem 96(2): 482-488.

Hallett, P. J., R. Spoelgen, B. T. Hyman, D. G. Standaert and A. W. Dunah (2006). "Dopamine D1 activation potentiates striatal NMDA receptors by tyrosine phosphorylation-dependent subunit trafficking." J Neurosci 26(17): 4690-4700. 138

Hart, G., B. K. Leung and B. W. Balleine (2014). "Dorsal and ventral streams: the distinct role of striatal subregions in the acquisition and performance of goal-directed actions." Neurobiol Learn Mem 108: 104-118.

Hartung, H., S. Threlfell and S. J. Cragg (2011). "Nitric oxide donors enhance the frequency dependence of dopamine release in nucleus accumbens." Neuropsychopharmacology 36(9): 1811-1822.

Heiman, M., A. Heilbut, V. Francardo, R. Kulicke, R. J. Fenster, E. D. Kolaczyk, J. P. Mesirov, D. J. Surmeier, M. A. Cenci and P. Greengard (2014). "Molecular adaptations of striatal spiny projection neurons during levodopa-induced dyskinesia." Proc Natl Acad Sci U S A 111(12): 4578-4583.

Heintz, N. (2004). "Gene expression nervous system atlas (GENSAT)." Nat Neurosci 7(5): 483.

Henry, D. J. and C. Chavkin (1995). "Activation of inwardly rectifying potassium channels (GIRK1) by co-expressed rat brain cannabinoid receptors in Xenopus oocytes." Neurosci Lett 186(2-3): 91-94.

Hernandez-Echeagaray, E., A. J. Starling, C. Cepeda and M. S. Levine (2004). "Modulation of AMPA currents by D2 dopamine receptors in striatal medium-sized spiny neurons: are dendrites necessary?" Eur J Neurosci 19(9): 2455-2463.

Hernandez-Lopez, S., T. Tkatch, E. Perez-Garci, E. Galarraga, J. Bargas, H. Hamm and D. J. Surmeier (2000). "D2 dopamine receptors in striatal medium spiny neurons reduce L-type Ca2+ currents and excitability via a novel PLC[beta]1-IP3-calcineurin-signaling cascade." J Neurosci 20(24): 8987-8995.

Hersch, S. M., C. A. Gutekunst, H. D. Rees, C. J. Heilman and A. I. Levey (1994). "Distribution of m1-m4 muscarinic receptor proteins in the rat striatum: light and electron microscopic immunocytochemistry using subtype-specific antibodies." J Neurosci 14(5 Pt 2): 3351-3363. 139

Higley, M. J., A. H. Gittis, I. A. Oldenburg, N. Balthasar, R. P. Seal, R. H. Edwards, B. B. Lowell, A. C. Kreitzer and B. L. Sabatini (2011). "Cholinergic interneurons mediate fast VGluT3-dependent glutamatergic transmission in the striatum." PLoS One 6(4): e19155.

Higley, M. J., G. J. Soler-Llavina and B. L. Sabatini (2009). "Cholinergic modulation of multivesicular release regulates striatal synaptic potency and integration." Nat Neurosci 12(9): 1121-1128.

Holley, S. M., P. R. Joshi, A. Parievsky, L. Galvan, J. Y. Chen, Y. E. Fisher, M. N. Huynh, C. Cepeda and M. S. Levine (2015). "Enhanced GABAergic Inputs Contribute to Functional Alterations of Cholinergic Interneurons in the R6/2 Mouse Model of Huntington's Disease." eNeuro 2(1).

Holscher, C. (1997). "Nitric oxide, the enigmatic neuronal messenger: its role in synaptic plasticity." Trends Neurosci 20(7): 298-303.

Howe, M. W. and D. A. Dombeck (2016). "Rapid signalling in distinct dopaminergic axons during locomotion and reward." Nature 535(7613): 505-510.

Hu, H., J. Z. Cavendish and A. Agmon (2013). "Not all that glitters is gold: off-target recombination in the somatostatin-IRES-Cre mouse line labels a subset of fast-spiking interneurons." Front Neural Circuits 7: 195.

Ibanez-Sandoval, O., F. Tecuapetla, B. Unal, F. Shah, T. Koos and J. M. Tepper (2010). "Electrophysiological and morphological characteristics and synaptic connectivity of tyrosine hydroxylase-expressing neurons in adult mouse striatum." J Neurosci 30(20): 6999-7016.

Ibanez-Sandoval, O., F. Tecuapetla, B. Unal, F. Shah, T. Koos and J. M. Tepper (2011). "A novel functionally distinct subtype of striatal neuropeptide Y interneuron." J Neurosci 31(46): 16757-16769. 140

Ilango, A., A. J. Kesner, K. L. Keller, G. D. Stuber, A. Bonci and S. Ikemoto (2014). "Similar roles of substantia nigra and ventral tegmental dopamine neurons in reward and aversion." J Neurosci 34(3): 817-822.

Ilinsky, I. A., M. L. Jouandet and P. S. Goldman-Rakic (1985). "Organization of the nigrothalamocortical system in the rhesus monkey." J Comp Neurol 236(3): 315-330.

Inokawa, H., H. Yamada, N. Matsumoto, M. Muranishi and M. Kimura (2010). "Juxtacellular labeling of tonically active neurons and phasically active neurons in the rat striatum." Neuroscience 168(2): 395-404.

Jones, E. G., J. D. Coulter, H. Burton and R. Porter (1977). "Cells of origin and terminal distribution of corticostriatal fibers arising in the sensory-motor cortex of monkeys." J Comp Neurol 173(1): 53-80.

Katzenschlager, R., C. Sampaio, J. Costa and A. Lees (2003). "Anticholinergics for symptomatic management of Parkinson's disease." Cochrane Database Syst Rev(2): Cd003735.

Kawaguchi, Y. (1993). "Physiological, Morphological, and Histochemical Characterization of Three Classes of Interneurons in Rat Neostriatum." J Neurosci 13(11): 4908-4923.

Kawaguchi, Y., C. J. Wilson and P. C. Emson (1989). "Intracellular Recording of Identified Neostriatal Patch and Matrix Spiny Cells in a Slice Preparation Preserving Cortical Inputs." J Neurophysiol 62(5): 1052-1068.

Kawaguchi, Y., C. J. Wilson and P. C. Emson (1990). "Projection subtypes of rat neostriatal matrix cells revealed by intracellular injection of biocytin." J Neurosci 10(10): 3421-3438.

Kemp, J. M. and T. P. Powell (1970). "The cortico-striate projection in the monkey." Brain 93(3): 525-546. 141

Kharkwal, G., K. Brami-Cherrier, J. E. Lizardi-Ortiz, A. B. Nelson, M. Ramos, D. Del Barrio, D. Sulzer, A. C. Kreitzer and E. Borrelli (2016). "Parkinsonism Driven by Antipsychotics Originates from Dopaminergic Control of Striatal Cholinergic Interneurons." Neuron 91(1): 67- 78.

Kim, J. I., S. Ganesan, S. X. Luo, Y. W. Wu, E. Park, E. J. Huang, L. Chen and J. B. Ding (2015). "Aldehyde dehydrogenase 1a1 mediates a GABA synthesis pathway in midbrain dopaminergic neurons." Science 350(6256): 102-106.

Kincaid, A. E., T. Zheng and C. J. Wilson (1998). "Connectivity and Convergence of Single Corticostriatal Axons." J Neurosci 18(12): 4722-4731.

Kishimoto, J., E. B. Keverne, J. Hardwick and P. C. Emson (1993). "Localization of nitric oxide synthase in the mouse olfactory and vomeronasal system: a histochemical, immunological and in situ hybridization study." Eur J Neurosci 5(12): 1684-1694.

Kiss, J. P. (2000). "Role of nitric oxide in the regulation of monoaminergic neurotransmission." Brain Res Bull 52(6): 459-466.

Kita, H. and S. T. Kitai (1987). "Efferent projections of the subthalamic nucleus in the rat: light and electron microscopic analysis with the PHA-L method." J Comp Neurol 260(3): 435-452.

Koos, T. and J. M. Tepper (1999). "Inhibitory control of neostriatal projection neurons by GABAergic interneurons." Nat Neurosci 2(5): 467-472.

Koos, T. and J. M. Tepper (2002). "Dual cholinergic control of fast-spiking interneurons in the neostriatum." J Neurosci 22(2): 529-535.

Kosillo, P., Y. F. Zhang, S. Threlfell and S. J. Cragg (2016). "Cortical Control of Striatal Dopamine Transmission via Striatal Cholinergic Interneurons." Cereb Cortex. 142

Kravitz, A. V., S. F. Owen and A. C. Kreitzer (2013). "Optogenetic identification of striatal projection neuron subtypes during in vivo recordings." Brain Res 1511: 21-32.

Kreitzer, A. C. and R. C. Malenka (2005). "Dopamine modulation of state-dependent endocannabinoid release and long-term depression in the striatum." J Neurosci 25(45): 10537- 10545.

Kreitzer, A. C. and R. C. Malenka (2007). "Endocannabinoid-mediated rescue of striatal LTD and motor deficits in Parkinson's disease models." Nature 445(7128): 643-647.

Kreitzer, A. C. and R. C. Malenka (2008). "Striatal plasticity and basal ganglia circuit function." Neuron 60(4): 543-554.

Kress, G. J., N. Yamawaki, D. L. Wokosin, I. R. Wickersham, G. M. Shepherd and D. J. Surmeier (2013). "Convergent cortical innervation of striatal projection neurons." Nat Neurosci 16(6): 665-667.

Kruse, A. C., B. K. Kobilka, D. Gautam, P. M. Sexton, A. Christopoulos and J. Wess (2014). "Muscarinic acetylcholine receptors: novel opportunities for drug development." Nat Rev Drug Discov 13(7): 549-560.

Kubota, Y., S. Inagaki, S. Shimada, S. Kito, F. Eckenstein and M. Tohyama (1987). "Neostriatal cholinergic neurons receive direct synaptic inputs from dopaminergic axons." Brain Res 413(1): 179-184.

Kubota, Y. and Y. Kawaguchi (2000). "Dependence of GABAergic synaptic areas on the interneuron type and target size." J Neurosci 20(1): 375-386.

Kumar, N., J. A. Van Gerpen, J. H. Bower and J. E. Ahlskog (2005). "Levodopa-dyskinesia incidence by age of Parkinson's disease onset." Mov Disord 20(3): 342-344. 143

Ladepeche, L., L. Yang, D. Bouchet and L. Groc (2013). "Regulation of dopamine D1 receptor dynamics within the postsynaptic density of hippocampal glutamate synapses." PLoS One 8(9): e74512.

Lammel, S., B. K. Lim and R. C. Malenka (2014). "Reward and aversion in a heterogeneous midbrain dopamine system." Neuropharmacology 76 Pt B: 351-359.

Lang, A. E. (1984). "Treatment of Parkinson's disease with agents other than levodopa and dopamine agonists: controversies and new approaches." Can J Neurol Sci 11(1 Suppl): 210-220.

Langston, J. W. (2006). "The Parkinson's complex: parkinsonism is just the tip of the iceberg." Ann Neurol 59(4): 591-596.

Lavoie, B., Y. Smith and A. Parent (1989). "Dopaminergic innervation of the basal ganglia in the squirrel monkey as revealed by tyrosine hydroxylase immunohistochemistry." J Comp Neurol 289(1): 36-52.

Le Moine, C. and B. Bloch (1995). "D1 and D2 dopamine receptor gene expression in the rat striatum: sensitive cRNA probes demonstrate prominent segregation of D1 and D2 mRNAs in distinct neuronal populations of the dorsal and ventral striatum." J Comp Neurol 355(3): 418- 426.

Lerner, T. N. and A. C. Kreitzer (2012). "RGS4 is required for dopaminergic control of striatal LTD and susceptibility to parkinsonian motor deficits." Neuron 73(2): 347-359.

Liguz-Lecznar, M., J. Urban-Ciecko and M. Kossut (2016). "Somatostatin and Somatostatin- Containing Neurons in Shaping Neuronal Activity and Plasticity." Front Neural Circuits 10: 48.

Lopez-Huerta, V. G., E. Blanco-Hernandez, J. Bargas and E. Galarraga (2012). "Presynaptic modulation by somatostatin in the rat neostriatum is altered in a model of parkinsonism." J Neurophysiol 108(4): 1032-1043. 144

Lovett-Barron, M., P. Kaifosh, M. A. Kheirbek, N. Danielson, J. D. Zaremba, T. R. Reardon, G. F. Turi, R. Hen, B. V. Zemelman and A. Losonczy (2014). "Dendritic inhibition in the hippocampus supports fear learning." Science 343(6173): 857-863.

Lovinger, D. M., E. C. Tyler and A. Merritt (1993). "Short- and long-term synaptic depression in rat neostriatum." J Neurophysiol 70(5): 1937-1949.

Luo, R., M. J. Janssen, J. G. Partridge and S. Vicini (2013). "Direct and GABA-mediated indirect effects of nicotinic ACh receptor agonists on striatal neurones." J Physiol 591(1): 203- 217.

Ma, Y., Q. Feng, L. Ouyang, S. Mu, B. Liu, Y. Li, S. Chen and W. Lei (2014). "Morphological diversity of GABAergic and cholinergic interneurons in the striatal dorsolateral and ventromedial regions of rats." Cell Mol Neurobiol 34(3): 351-359.

Ma, Y., H. Hu, A. S. Berrebi, P. H. Mathers and A. Agmon (2006). "Distinct subtypes of somatostatin-containing neocortical interneurons revealed in transgenic mice." J Neurosci 26(19): 5069-5082.

Madisen, L., T. A. Zwingman, S. M. Sunkin, S. W. Oh, H. A. Zariwala, H. Gu, L. L. Ng, R. D. Palmiter, M. J. Hawrylycz, A. R. Jones, E. S. Lein and H. Zeng (2010). "A robust and high- throughput Cre reporting and characterization system for the whole mouse brain." Nat Neurosci 13(1): 133-140.

Magnus, C. J., P. H. Lee, D. Atasoy, H. H. Su, L. L. Looger and S. M. Sternson (2011). "Chemical and genetic engineering of selective ion channel-ligand interactions." Science 333(6047): 1292-1296.

Mahapatra, S., A. Marcantoni, A. Zuccotti, V. Carabelli and E. Carbone (2012). "Equal sensitivity of Cav1.2 and Cav1.3 channels to the opposing modulations of PKA and PKG in mouse chromaffin cells." J Physiol 590(Pt 20): 5053-5073. 145

Marin, O., S. A. Andersen and J. L. Rubenstein (2000). "Origin and Molecular Specificiation of Striatal Interneurons." J Neurosci 20(16): 6063-6076.

Marras, C. and A. Lang (2008). "Invited article: changing concepts in Parkinson disease: moving beyond the decade of the brain." Neurology 70(21): 1996-2003.

Martella, G., A. Tassone, G. Sciamanna, P. Platania, D. Cuomo, M. T. Viscomi, P. Bonsi, E. Cacci, S. Biagioni, A. Usiello, G. Bernardi, N. Sharma, D. G. Standaert and A. Pisani (2009). "Impairment of bidirectional synaptic plasticity in the striatum of a mouse model of DYT1 dystonia: role of endogenous acetylcholine." Brain 132(Pt 9): 2336-2349.

Mathur, B. N., N. A. Capik, V. A. Alvarez and D. M. Lovinger (2011). "Serotonin induces long- term depression at corticostriatal synapses." J Neurosci 31(20): 7402-7411.

Mathur, B. N. and D. M. Lovinger (2012). "Endocannabinoid-dopamine interactions in striatal synaptic plasticity." Front Pharmacol 3: 66.

Matsuda, L. A., S. J. Lolait, M. J. Brownstein, A. C. Young and T. I. Bonner (1990). "Structure of a cannabinoid receptor and functional expression of the cloned cDNA." Nature 346(6284): 561-564.

Matsuda, W., T. Furuta, K. C. Nakamura, H. Hioki, F. Fujiyama, R. Arai and T. Kaneko (2009). "Single nigrostriatal dopaminergic neurons form widely spread and highly dense axonal arborizations in the neostriatum." J Neurosci 29(2): 444-453.

Matsumoto, N., T. Minamimoto, A. M. Graybiel and M. Kimura (2001). "Neurons in the thalamic CM-Pf complex supply striatal neurons with information about behaviorally significant sensory events." J Neurophysiol 85(2): 960-976.

Matsushita, N., H. Okada, Y. Yasoshima, K. Takahashi, K. Kiuchi and K. Kobayashi (2002). "Dynamics of tyrosine hydroxylase promoter activity during midbrain dopaminergic neuron development." J Neurochem 82(2): 295-304. 146

Matyas, F., Y. Yanovsky, K. Mackie, W. Kelsch, U. Misgeld and T. F. Freund (2006). "Subcellular localization of type 1 cannabinoid receptors in the rat basal ganglia." Neuroscience 137(1): 337-361.

Maurice, N., J. Mercer, C. S. Chan, S. Hernandez-Lopez, J. Held, T. Tkatch and D. J. Surmeier (2004). "D2 dopamine receptor-mediated modulation of voltage-dependent Na+ channels reduces autonomous activity in striatal cholinergic interneurons." J Neurosci 24(46): 10289- 10301.

McCormick, D. A. and D. A. Prince (1985). "Two types of muscarinic response to acetylcholine in mammalian cortical neurons." Proc Natl Acad Sci U S A 82(18): 6344-6348.

McFarland, N. R. and S. N. Haber (2000). "Convergent inputs from thalamic motor nuclei and frontal cortical areas to the dorsal striatum in the primate." J Neurosci 20(10): 3798-3813.

McGeorge, A. J. and R. L. Faull (1989). "The organization of the projection from the cerebral cortex to the striatum in the rat." Neuroscience 29(3): 503-537.

Mechoulam, R., S. Ben-Shabat, L. Hanus, M. Ligumsky, N. E. Kaminski, A. R. Schatz, A. Gopher, S. Almog, B. R. Martin, D. R. Compton and et al. (1995). "Identification of an endogenous 2-monoglyceride, present in canine gut, that binds to cannabinoid receptors." Biochem Pharmacol 50(1): 83-90.

Menniti, F. S., W. S. Faraci and C. J. Schmidt (2006). "Phosphodiesterases in the CNS: targets for drug development." Nat Rev Drug Discov 5(8): 660-670.

Mesulam, M. M., D. Mash, L. Hersh, M. Bothwell and C. Geula (1992). "Cholinergic innervation of the human striatum, globus pallidus, subthalamic nucleus, substantia nigra, and red nucleus." J Comp Neurol 323(2): 252-268.

Meurers, B. H., G. Dziewczapolski, T. Shi, A. Bittner, F. Kamme and C. W. Shults (2009). "Dopamine depletion induces distinct compensatory gene expression changes in DARPP-32 147

signal transduction cascades of striatonigral and striatopallidal neurons." J Neurosci 29(21): 6828-6839.

Miyasaki, J. M., W. Martin, O. Suchowersky, W. J. Weiner and A. E. Lang (2002). "Practice parameter: initiation of treatment for Parkinson's disease: an evidence-based review: report of the Quality Standards Subcommittee of the American Academy of Neurology." Neurology 58(1): 11-17.

Morishita, W., H. Marie and R. C. Malenka (2005). "Distinct triggering and expression mechanisms underlie LTD of AMPA and NMDA synaptic responses." Nat Neurosci 8(8): 1043- 1050.

Moss, J. and J. P. Bolam (2008). "A dopaminergic axon lattice in the striatum and its relationship with cortical and thalamic terminals." J Neurosci 28(44): 11221-11230.

Munoz-Manchado, A. B., C. Foldi, S. Szydlowski, L. Sjulson, M. Farries, C. Wilson, G. Silberberg and J. Hjerling-Leffler (2016). "Novel Striatal GABAergic Interneuron Populations Labeled in the 5HT3a(EGFP) Mouse." Cereb Cortex 26(1): 96-105.

Murphy, J. A., I. S. Stein, C. G. Lau, R. T. Peixoto, T. K. Aman, N. Kaneko, K. Aromolaran, J. L. Saulnier, G. K. Popescu, B. L. Sabatini, J. W. Hell and R. S. Zukin (2014). "Phosphorylation of Ser1166 on GluN2B by PKA is critical to synaptic NMDA receptor function and Ca2+ signaling in spines." J Neurosci 34(3): 869-879.

Nanda, B., A. Galvan, Y. Smith and T. Wichmann (2009). "Effects of stimulation of the centromedian nucleus of the thalamus on the activity of striatal cells in awake rhesus monkeys." Eur J Neurosci 29(3): 588-598.

Nathanson, N. M. (2008). "Synthesis, trafficking, and localization of muscarinic acetylcholine receptors." Pharmacol Ther 119(1): 33-43. 148

Nelson, A. B., N. Hammack, C. F. Yang, N. M. Shah, R. P. Seal and A. C. Kreitzer (2014). "Striatal cholinergic interneurons Drive GABA release from dopamine terminals." Neuron 82(1): 63-70.

Nelson, A. B. and A. C. Kreitzer (2014). "Reassessing models of basal ganglia function and dysfunction." Annu Rev Neurosci 37: 117-135.

Neve, K. A., J. K. Seamans and H. Trantham-Davidson (2004). "Dopamine receptor signaling." J Recept Signal Transduct Res 24(3): 165-205.

Nieto Mendoza, E. and E. Hernandez Echeagaray (2015). "Dopaminergic Modulation of Striatal Inhibitory Transmission and Long-Term Plasticity." Neural Plast 2015: 789502.

Nugent, F. S., E. C. Penick and J. A. Kauer (2007). "Opioids block long-term potentiation of inhibitory synapses." Nature 446(7139): 1086-1090.

Ockenga, W., S. Kuhne, S. Bocksberger, A. Banning and R. Tikkanen (2013). "Non-neuronal functions of the m2 muscarinic acetylcholine receptor." Genes (Basel) 4(2): 171-197.

Olanow, C. W., R. L. Watts and W. C. Koller (2001). "An algorithm (decision tree) for the management of Parkinson's disease (2001): treatment guidelines." Neurology 56(11 Suppl 5): S1-s88.

Ondracek, J. M., A. Dec, K. E. Hoque, S. A. Lim, G. Rasouli, R. P. Indorkar, J. Linardakis, B. Klika, S. J. Mukherji, M. Burnazi, S. Threlfell, S. Sammut and A. R. West (2008). "Feed- forward excitation of striatal neuron activity by frontal cortical activation of nitric oxide signaling in vivo." Eur J Neurosci 27(7): 1739-1754.

Oorschot, D. E. (1996). "Total number of neurons in the neostriatal, pallidal, subthalamic, and substantia nigral nuclei of the rat basal ganglia: a stereological study using the cavalieri and optical disector methods." J Comp Neurol 366(4): 580-599. 149

Orimo, S., T. Uchihara, T. Kanazawa, Y. Itoh, K. Wakabayashi, A. Kakita and H. Takahashi (2011). "Unmyelinated axons are more vulnerable to degeneration than myelinated axons of the cardiac nerve in Parkinson's disease." Neuropathol Appl Neurobiol 37(7): 791-802.

Padovan-Neto, F. E., M. B. Echeverry, V. Tumas and E. A. Del-Bel (2009). "Nitric oxide synthase inhibition attenuates L-DOPA-induced dyskinesias in a rodent model of Parkinson's disease." Neuroscience 159(3): 927-935.

Pancani, T., C. Bolarinwa, Y. Smith, C. W. Lindsley, P. J. Conn and Z. Xiang (2014). "M4 mAChR-mediated modulation of glutamatergic transmission at corticostriatal synapses." ACS Chem Neurosci 5(4): 318-324.

Pare, D., E. Shink, H. Gaudreau, A. Destexhe and E. J. Lang (1998). "Impact of spontaneous synaptic activity on the resting properties of cat neocortical pyramidal neurons In vivo." J Neurophysiol 79(3): 1450-1460.

Parent, A. (2012). "The History of the Basal Ganglia: The Contribution of Karl Friedrich Burdach." Neuroscience & Medicine 03(04): 374-379.

Parker, P. R., A. L. Lalive and A. C. Kreitzer (2016). "Pathway-Specific Remodeling of Thalamostriatal Synapses in Parkinsonian Mice." Neuron 89(4): 734-740.

Partridge, J. G., M. J. Janssen, D. Y. Chou, K. Abe, Z. Zukowska and S. Vicini (2009). "Excitatory and inhibitory synapses in neuropeptide Y-expressing striatal interneurons." J Neurophysiol 102(5): 3038-3045.

Peterson, J. D., J. A. Goldberg and D. J. Surmeier (2012). "Adenosine A2a receptor antagonists attenuate striatal adaptations following dopamine depletion." Neurobiol Dis 45(1): 409-416.

Petreanu, L., T. Mao, S. M. Sternson and K. Svoboda (2009). "The subcellular organization of neocortical excitatory connections." Nature 457(7233): 1142-1145. 150

Picconi, B., V. Bagetta, V. Ghiglieri, V. Paille, M. Di Filippo, V. Pendolino, A. Tozzi, C. Giampa, F. R. Fusco, C. Sgobio and P. Calabresi (2011). "Inhibition of phosphodiesterases rescues striatal long-term depression and reduces levodopa-induced dyskinesia." Brain 134(Pt 2): 375-387.

Picconi, B., D. Centonze, K. Hakansson, G. Bernardi, P. Greengard, G. Fisone, M. A. Cenci and P. Calabresi (2003). "Loss of bidirectional striatal synaptic plasticity in L-DOPA-induced dyskinesia." Nat Neurosci 6(5): 501-506.

Pisani, A., P. Bonsi, D. Centonze, P. Gubellini, G. Bernardi and P. Calabresi (2003). "Targeting striatal cholinergic interneurons in Parkinson's disease: focus on metabotropic glutamate receptors." Neuropharmacology 45(1): 45-56.

Plotkin, J. L., N. Wu, M. F. Chesselet and M. S. Levine (2005). "Functional and molecular development of striatal fast-spiking GABAergic interneurons and their cortical inputs." Eur J Neurosci 22(5): 1097-1108.

Prast, H. and A. Philippu (2001). "Nitric oxide as modulator of neuronal function." Prog Neurobiol 64(1): 51-68.

Quinn, N. (1995). "Drug treatment of Parkinson's disease." Bmj 310(6979): 575-579.

Quinn, N., P. Critchley and C. D. Marsden (1987). "Young onset Parkinson's disease." Mov Disord 2(2): 73-91.

Rafalovich, I. V., A. E. Melendez, J. L. Plotkin, A. Tanimura, S. Zhai and D. J. Surmeier (2015). "Interneuronal Nitric Oxide Signaling Mediates Post-synaptic Long-Term Depression of Striatal Glutamatergic Synapses." Cell Rep 13(7): 1336-1342.

Raju, D. V., T. H. Ahern, D. J. Shah, T. M. Wright, D. G. Standaert, R. A. Hall and Y. Smith (2008). "Differential synaptic plasticity of the corticostriatal and thalamostriatal systems in an MPTP-treated monkey model of parkinsonism." Eur J Neurosci 27(7): 1647-1658. 151

Ramanathan, S., J. J. Hanley, J. M. Deniau and J. P. Bolam (2002). "Synaptic convergence of motor and somatosensory cortical afferents onto GABAergic interneurons in the rat striatum." J Neurosci 22(18): 8158-8169.

Rascol, O., P. Payoux, F. Ory, J. J. Ferreira, C. Brefel-Courbon and J. L. Montastruc (2003). "Limitations of current Parkinson's disease therapy." Ann Neurol 53 Suppl 3: S3-12; discussion S12-15.

Reig, R. and G. Silberberg (2014). "Multisensory integration in the mouse striatum." Neuron 83(5): 1200-1212.

Riebe, I., B. Gustafsson and E. Hanse (2009). "Silent synapses onto interneurons in the rat CA1 stratum radiatum." Eur J Neurosci 29(9): 1870-1882.

Rose, F. C. (2010). "Chapter 39: an historical overview of British neurology." Handb Clin Neurol 95: 613-628.

Sagi, Y., M. Heiman, J. D. Peterson, S. Musatov, M. Scarduzio, S. M. Logan, M. G. Kaplitt, D. J. Surmeier, N. Heintz, P. Greengard and T. R. University (2014). "Nitric oxide regulates synaptic transmission between spiny projection neurons." Proceedings of the National Academy of Sciences of the United States of America 111(49): 6.

Salin, P., I. P. Lopez, P. Kachidian, P. Barroso-Chinea, A. J. Rico, V. Gomez-Bautista, P. Coulon, L. Kerkerian-Le Goff and J. L. Lanciego (2009). "Changes to interneuron-driven striatal microcircuits in a rat model of Parkinson's disease." Neurobiol Dis 34(3): 545-552.

Sammut, S., D. J. Park and A. R. West (2007). "Frontal cortical afferents facilitate striatal nitric oxide transmission in vivo via a NMDA receptor and neuronal NOS-dependent mechanism." J Neurochem 103(3): 1145-1156. 152

Sancesario, G., M. Giorgi, V. D'Angelo, A. Modica, A. Martorana, M. Morello, C. P. Bengtson and G. Bernardi (2004). "Down-regulation of nitrergic transmission in the rat striatum after chronic nigrostriatal deafferentation." Eur J Neurosci 20(4): 989-1000.

Sandell, J. H., A. M. Graybiel and M. F. Chesselet (1986). "A new enzyme marker for striatal compartmentalization: NADPH diaphorase activity in the caudate nucleus and putamen of the cat." J Comp Neurol 243(3): 326-334.

Saunders, A., A. J. Granger and B. L. Sabatini (2015). "Corelease of acetylcholine and GABA from cholinergic forebrain neurons." Elife 4.

Saunders, A., K. W. Huang and B. L. Sabatini (2016). "Globus Pallidus Externus Neurons Expressing parvalbumin Interconnect the Subthalamic Nucleus and Striatal Interneurons." PLoS One 11(2): e0149798.

Schiffmann, S. N., O. Jacobs and J. J. Vanderhaeghen (1991). "Striatal restricted adenosine A2 receptor (RDC8) is expressed by enkephalin but not by substance P neurons: an in situ hybridization histochemistry study." J Neurochem 57(3): 1062-1067.

Schwab, R. S., D. C. Poskanzer, A. C. England, Jr. and R. R. Young (1972). "Amantadine in Parkinson's disease. Review of more than two years' experience." Jama 222(7): 792-795.

Sciamanna, G., G. Ponterio, G. Mandolesi, P. Bonsi and A. Pisani (2015). "Optogenetic stimulation reveals distinct modulatory properties of thalamostriatal vs corticostriatal glutamatergic inputs to fast-spiking interneurons." Sci Rep 5: 16742.

Sebel, L. E., S. M. Graves, C. S. Chan and D. J. Surmeier (2016). "Haloperidol Selectively Remodels Striatal Indirect Pathway Circuits." Neuropsychopharmacology.

Sergeeva, O. A., N. Doreulee, A. N. Chepkova, T. Kazmierczak and H. L. Haas (2007). "Long- term depression of cortico-striatal synaptic transmission by DHPG depends on endocannabinoid release and nitric oxide synthesis." Eur J Neurosci 26(7): 1889-1894. 153

Sheffler, D. J., R. Williams, T. M. Bridges, Z. Xiang, A. S. Kane, N. E. Byun, S. Jadhav, M. M. Mock, F. Zheng, L. M. Lewis, C. K. Jones, C. M. Niswender, C. D. Weaver, C. W. Lindsley and P. J. Conn (2009). "A novel selective muscarinic acetylcholine receptor subtype 1 antagonist reduces seizures without impairing hippocampus-dependent learning." Mol Pharmacol 76(2): 356-368.

Shen, W., M. Flajolet, P. Greengard and D. J. Surmeier (2008). "Dichotomous dopaminergic control of striatal synaptic plasticity." Science 321(5890): 848-851.

Shen, W., S. E. Hamilton, N. M. Nathanson and D. J. Surmeier (2005). "Cholinergic suppression of KCNQ channel currents enhances excitability of striatal medium spiny neurons." J Neurosci 25(32): 7449-7458.

Shen, W., J. L. Plotkin, V. Francardo, W. K. Ko, Z. Xie, Q. Li, T. Fieblinger, J. Wess, R. R. Neubig, C. W. Lindsley, P. J. Conn, P. Greengard, E. Bezard, M. A. Cenci and D. J. Surmeier (2015). "M4 Muscarinic Receptor Signaling Ameliorates Striatal Plasticity Deficits in Models of L-DOPA-Induced Dyskinesia." Neuron 88(4): 762-773.

Shen, W., X. Tian, M. Day, S. Ulrich, T. Tkatch, N. M. Nathanson and D. J. Surmeier (2007). "Cholinergic modulation of Kir2 channels selectively elevates dendritic excitability in striatopallidal neurons." Nat Neurosci 10(11): 1458-1466.

Shibata, M., N. Araki, K. Ohta, J. Hamada, K. Shimazu and Y. Fukuuchi (1996). "Nitric oxide regulates NMDA-induced dopamine release in rat striatum." Neuroreport 7(2): 605-608.

Shiflett, M. W., R. A. Brown and B. W. Balleine (2010). "Acquisition and performance of goal- directed instrumental actions depends on ERK signaling in distinct regions of dorsal striatum in rats." J Neurosci 30(8): 2951-2959. 154

Sidibe, M. and Y. Smith (1996). "Differential synaptic innervation of striatofugal neurones projecting to the internal or external segments of the globus pallidus by thalamic afferents in the squirrel monkey." J Comp Neurol 365(3): 445-465.

Sidibe, M. and Y. Smith (1999). "Thalamic inputs to striatal interneurons in monkeys: synaptic organization and co-localization of calcium binding proteins." Neuroscience 89(4): 1189-1208.

Sidlo, Z., P. H. Reggio and M. E. Rice (2008). "Inhibition of striatal dopamine release by CB1 receptor activation requires nonsynaptic communication involving GABA, H2O2, and KATP channels." Neurochem Int 52(1-2): 80-88.

Sjostrom, P. J., E. A. Rancz, A. Roth and M. Hausser (2008). "Dendritic excitability and synaptic plasticity." Physiol Rev 88(2): 769-840.

Smith, Y., A. Galvan, T. J. Ellender, N. Doig, R. M. Villalba, I. Huerta-Ocampo, T. Wichmann and J. P. Bolam (2014). "The thalamostriatal system in normal and diseased states." Front Syst Neurosci 8: 5.

Smith, Y., D. Raju, B. Nanda, J. F. Pare, A. Galvan and T. Wichmann (2009). "The thalamostriatal systems: anatomical and functional organization in normal and parkinsonian states." Brain Res Bull 78(2-3): 60-68.

Smith, Y., D. V. Raju, J. F. Pare and M. Sidibe (2004). "The thalamostriatal system: a highly specific network of the basal ganglia circuitry." Trends Neurosci 27(9): 520-527.

Smith, Y. and R. Villalba (2008). "Striatal and extrastriatal dopamine in the basal ganglia: an overview of its anatomical organization in normal and Parkinsonian brains." Mov Disord 23 Suppl 3: S534-547.

Snyder, G. L., P. B. Allen, A. A. Fienberg, C. G. Valle, R. L. Huganir, A. C. Nairn and P. Greengard (2000). "Regulation of phosphorylation of the GluR1 AMPA receptor in the neostriatum by dopamine and psychostimulants in vivo." J Neurosci 20(12): 4480-4488. 155

Solis, O., I. Espadas, E. A. Del-Bel and R. Moratalla (2015). "Nitric oxide synthase inhibition decreases l-DOPA-induced dyskinesia and the expression of striatal molecular markers in Pitx3(- /-) aphakia mice." Neurobiol Dis 73: 49-59.

Somogyi, P., J. P. Bolam, S. Totterdell and A. D. Smith (1981). "Monosynaptic input from the nucleus accumbens--ventral striatum region to retrogradely labelled nigrostriatal neurones." Brain Res 217(2): 245-263.

Song, S. C., J. A. Beatty and C. J. Wilson (2016). "The Ionic Mechanism of Membrane Potential Oscillations and Membrane Resonance in Striatal LTS Interneurons." J Neurophysiol: jn.00511.02016.

Stasch, J. P., E. M. Becker, C. Alonso-Alija, H. Apeler, K. Dembowsky, A. Feurer, R. Gerzer, T. Minuth, E. Perzborn, U. Pleiss, H. Schroder, W. Schroeder, E. Stahl, W. Steinke, A. Straub and M. Schramm (2001). "NO-independent regulatory site on soluble guanylate cyclase." Nature 410(6825): 212-215.

Steiner, H. and C. R. Gerfen (1998). "Role of dynorphin and enkephalin in the regulation of striatal output pathways and behavior." Exp Brain Res 123(1-2): 60-76.

Stephens, B., A. J. Mueller, A. F. Shering, S. H. Hood, P. Taggart, G. W. Arbuthnott, J. E. Bell, L. Kilford, A. E. Kingsbury, S. E. Daniel and C. A. Ingham (2005). "Evidence of a breakdown of corticostriatal connections in Parkinson's disease." Neuroscience 132(3): 741-754.

Stoof, J. C. and J. W. Kebabian (1984). "Two dopamine receptors: biochemistry, physiology and pharmacology." Life Sci 35(23): 2281-2296.

Sullivan, M. A., H. Chen and H. Morikawa (2008). "Recurrent inhibitory network among striatal cholinergic interneurons." J Neurosci 28(35): 8682-8690.

Sulzer, D., J. Bogulavsky, K. E. Larsen, G. Behr, E. Karatekin, M. H. Kleinman, N. Turro, D. Krantz, R. H. Edwards, L. A. Greene and L. Zecca (2000). "Neuromelanin biosynthesis is driven 156

by excess cytosolic catecholamines not accumulated by synaptic vesicles." Proc Natl Acad Sci U S A 97(22): 11869-11874.

Sulzer, D., E. Mosharov, Z. Talloczy, F. A. Zucca, J. D. Simon and L. Zecca (2008). "Neuronal pigmented autophagic vacuoles: lipofuscin, neuromelanin, and ceroid as macroautophagic responses during aging and disease." J Neurochem 106(1): 24-36.

Surmeier, D. J., S. M. Graves and W. Shen (2014). "Dopaminergic modulation of striatal networks in health and Parkinson's disease." Curr Opin Neurobiol 29: 109-117.

Surmeier, D. J., J. N. Guzman, J. Sanchez and P. T. Schumacker (2012). "Physiological phenotype and vulnerability in Parkinson's disease." Cold Spring Harb Perspect Med 2(7): a009290.

Surmeier, D. J., J. Plotkin and W. Shen (2009). "Dopamine and synaptic plasticity in dorsal striatal circuits controlling action selection." Curr Opin Neurobiol 19(6): 621-628.

Surmeier, D. J. and P. T. Schumacker (2013). "Calcium, bioenergetics, and neuronal vulnerability in Parkinson's disease." J Biol Chem 288(15): 10736-10741.

Susswein, A. J., A. Katzoff, N. Miller and I. Hurwitz (2004). "Nitric oxide and memory." Neuroscientist 10(2): 153-162.

Svenningsson, P., A. Nishi, G. Fisone, J. A. Girault, A. C. Nairn and P. Greengard (2004). "DARPP-32: an integrator of neurotransmission." Annu Rev Pharmacol Toxicol 44: 269-296.

Szabo, G. G., N. Lenkey, N. Holderith, T. Andrasi, Z. Nusser and N. Hajos (2014). "Presynaptic calcium channel inhibition underlies CB(1) cannabinoid receptor-mediated suppression of GABA release." J Neurosci 34(23): 7958-7963. 157

Szydlowski, S. N., I. Pollak Dorocic, H. Planert, M. Carlen, K. Meletis and G. Silberberg (2013). "Target selectivity of feedforward inhibition by striatal fast-spiking interneurons." J Neurosci 33(4): 1678-1683.

Takuma, K., T. Tanaka, T. Takahashi, N. Hiramatsu, Y. Ota, Y. Ago and T. Matsuda (2012). "Neuronal nitric oxide synthase inhibition attenuates the development of L-DOPA-induced dyskinesia in hemi-Parkinsonian rats." Eur J Pharmacol 683(1-3): 166-173.

Taniguchi, H., M. He, P. Wu, S. Kim, R. Paik, K. Sugino, D. Kvitsani, Y. Fu, J. Lu, Y. Lin, G. Miyoshi, Y. Shima, G. Fishell, Sacha B. Nelson and Z. J. Huang (2011). "A Resource of Cre Driver Lines for Genetic Targeting of GABAergic Neurons in Cerebral Cortex." Neuron 71(6): 995-1013.

Tashiro, Y., T. Sugimoto, T. Hattori, Y. Uemura, I. Nagatsu, H. Kikuchi and N. Mizuno (1989). "Tyrosine hydroxylase-like immunoreactive neurons in the striatum of the rat." Neurosci Lett 97(1-2): 6-10.

Taverna, S., E. Ilijic and D. J. Surmeier (2008). "Recurrent collateral connections of striatal medium spiny neurons are disrupted in models of Parkinson's disease." J Neurosci 28(21): 5504- 5512.

Tayebati, S. K., M. A. Di Tullio and F. Amenta (2004). "Age-related changes of muscarinic cholinergic receptor subtypes in the striatum of Fisher 344 rats." Exp Gerontol 39(2): 217-223.

Tepper, J. M. and J. P. Bolam (2004). "Functional diversity and specificity of neostriatal interneurons." Curr Opin Neurobiol 14(6): 685-692.

Tepper, J. M., T. Koos and C. J. Wilson (2004). "GABAergic microcircuits in the neostriatum." Trends Neurosci 27(11): 662-669.

Tepper, J. M., F. Tecuapetla, T. Koos and O. Ibanez-Sandoval (2010). "Heterogeneity and diversity of striatal GABAergic interneurons." Front Neuroanat 4: 150. 158

Threlfell, S., T. Lalic, N. J. Platt, K. A. Jennings, K. Deisseroth and S. J. Cragg (2012). "Striatal dopamine release is triggered by synchronized activity in cholinergic interneurons." Neuron 75(1): 58-64.

Tice, M. A., T. Hashemi, L. A. Taylor and R. D. McQuade (1996). "Distribution of muscarinic receptor subtypes in rat brain from postnatal to old age." Brain Res Dev Brain Res 92(1): 70-76.

Tremblay, R., S. Lee and B. Rudy (2016). "GABAergic Interneurons in the Neocortex: From Cellular Properties to Circuits." Neuron 91(2): 260-292.

Tsang, E. W., C. Hamani, E. Moro, F. Mazzella, Y. Y. Poon, A. M. Lozano and R. Chen (2010). "Involvement of the human pedunculopontine nucleus region in voluntary movements." Neurology 75(11): 950-959.

Unal, B., F. Shah, J. Kothari and J. M. Tepper (2015). "Anatomical and electrophysiological changes in striatal TH interneurons after loss of the nigrostriatal dopaminergic pathway." Brain Struct Funct 220(1): 331-349.

van den Pol, A. N., Y. Yao, L. Y. Fu, K. Foo, H. Huang, R. Coppari, B. B. Lowell and C. Broberger (2009). "Neuromedin B and gastrin-releasing peptide excite arcuate nucleus neuropeptide Y neurons in a novel transgenic mouse expressing strong Renilla green fluorescent protein in NPY neurons." J Neurosci 29(14): 4622-4639.

Villalba, R. M., H. Lee and Y. Smith (2009). "Dopaminergic denervation and spine loss in the striatum of MPTP-treated monkeys." Exp Neurol 215(2): 220-227.

Vincent, S. R. (1994). "Nitric oxide: a radical neurotransmitter in the central nervous system." Prog Neurobiol 42(1): 129-160.

Walaas, S. I., H. C. Hemmings, Jr., P. Greengard and A. C. Nairn (2011). "Beyond the dopamine receptor: regulation and roles of serine/threonine protein phosphatases." Front Neuroanat 5: 50. 159

Wall, N. R., M. De La Parra, E. M. Callaway and A. C. Kreitzer (2013). "Differential innervation of direct- and indirect-pathway striatal projection neurons." Neuron 79(2): 347-360.

Wall, N. R., M. De La Parra, J. M. Sorokin, H. Taniguchi, Z. J. Huang and E. M. Callaway (2016). "Brain-Wide Maps of Synaptic Input to Cortical Interneurons." J Neurosci 36(14): 4000- 4009.

Ward, C. D. and W. R. Gibb (1990). "Research diagnostic criteria for Parkinson's disease." Adv Neurol 53: 245-249.

Weiner, D. M., A. I. Levey and M. R. Brann (1990). "Expression of muscarinic acetylcholine and dopamine receptor mRNAs in rat basal ganglia." Proc Natl Acad Sci U S A 87(18): 7050- 7054.

West, A. R., M. P. Galloway and A. A. Grace (2002). "Regulation of striatal dopamine neurotransmission by nitric oxide: effector pathways and signaling mechanisms." Synapse 44(4): 227-245.

West, A. R. and A. A. Grace (2004). "The nitric oxide-guanylyl cyclase signaling pathway modulates membrane activity States and electrophysiological properties of striatal medium spiny neurons recorded in vivo." J Neurosci 24(8): 1924-1935.

West, A. R. and K. Y. Tseng (2011). "Nitric Oxide-Soluble Guanylyl Cyclase-Cyclic GMP Signaling in the Striatum: New Targets for the Treatment of Parkinson's Disease?" Front Syst Neurosci 5: 55.

Wickersham, I. R., D. C. Lyon, R. J. Barnard, T. Mori, S. Finke, K. K. Conzelmann, J. A. Young and E. M. Callaway (2007). "Monosynaptic restriction of transsynaptic tracing from single, genetically targeted neurons." Neuron 53(5): 639-647.

Wilson, C. J., H. T. Chang and S. T. Kitai (1990). "Firing patterns and synaptic potentials of identified giant aspiny interneurons in the rat neostriatum." J Neurosci 10(2): 508-519. 160

Wiltschko, A. B., J. R. Pettibone and J. D. Berke (2010). "Opposite effects of stimulant and antipsychotic drugs on striatal fast-spiking interneurons." Neuropsychopharmacology 35(6): 1261-1270.

Witten, I. B., E. E. Steinberg, S. Y. Lee, T. J. Davidson, K. A. Zalocusky, M. Brodsky, O. Yizhar, S. L. Cho, S. Gong, C. Ramakrishnan, G. D. Stuber, K. M. Tye, P. H. Janak and K. Deisseroth (2011). "Recombinase-driver rat lines: tools, techniques, and optogenetic application to dopamine-mediated reinforcement." Neuron 72(5): 721-733.

Wonders, C. P. and S. A. Anderson (2006). "The origin and specification of cortical interneurons." Nat Rev Neurosci 7(9): 687-696.

Wu, Y. and A. Parent (2000). "Striatal interneurons expressing calretinin, parvalbumin or NADPH-diaphorase: a comparative study in the rat, monkey and human." Brain Res 863(1-2): 182-191.

Wu, Y. W., J. I. Kim, V. L. Tawfik, R. R. Lalchandani, G. Scherrer and J. B. Ding (2015). "Input- and cell-type-specific endocannabinoid-dependent LTD in the striatum." Cell Rep 10(1): 75-87.

Xenias, H. S., O. Ibanez-Sandoval, T. Koos and J. M. Tepper (2015). "Are striatal tyrosine hydroxylase interneurons dopaminergic?" 35(16): 6584-6599.

Yi, F., J. Ball, K. E. Stoll, V. C. Satpute, S. M. Mitchell, J. L. Pauli, B. B. Holloway, A. D. Johnston, N. M. Nathanson, K. Deisseroth, D. J. Gerber, S. Tonegawa and J. J. Lawrence (2014). "Direct excitation of parvalbumin-positive interneurons by M1 muscarinic acetylcholine receptors: roles in cellular excitability, inhibitory transmission and cognition." J Physiol 592(16): 3463-3494.

Yin, H. H. and B. J. Knowlton (2006). "The role of the basal ganglia in habit formation." Nat Rev Neurosci 7(6): 464-476. 161

Yoshida, T., M. Fukaya, M. Uchigashima, E. Miura, H. Kamiya, M. Kano and M. Watanabe (2006). "Localization of diacylglycerol lipase-alpha around postsynaptic spine suggests close proximity between production site of an endocannabinoid, 2-arachidonoyl-glycerol, and presynaptic cannabinoid CB1 receptor." J Neurosci 26(18): 4740-4751.

Zaja-Milatovic, S., D. Milatovic, A. M. Schantz, J. Zhang, K. S. Montine, A. Samii, A. Y. Deutch and T. J. Montine (2005). "Dendritic degeneration in neostriatal medium spiny neurons in Parkinson disease." Neurology 64(3): 545-547.

Zecca, L., F. A. Zucca, H. Wilms and D. Sulzer (2003). "Neuromelanin of the substantia nigra: a neuronal black hole with protective and toxic characteristics." Trends Neurosci 26(11): 578-580.

Zhang, F., L. P. Wang, E. S. Boyden and K. Deisseroth (2006). "Channelrhodopsin-2 and optical control of excitable cells." Nat Methods 3(10): 785-792.

Zink, C. F., G. Pagnoni, M. E. Martin, M. Dhamala and G. S. Berns (2003). "Human striatal response to salient nonrewarding stimuli." J Neurosci 23(22): 8092-8097.

Ztaou, S., N. Maurice, J. Camon, G. Guiraudie-Capraz, L. Kerkerian-Le Goff, C. Beurrier, M. Liberge and M. Amalric (2016). "Involvement of Striatal Cholinergic Interneurons and M1 and M4 Muscarinic Receptors in Motor Symptoms of Parkinson's Disease." 36(35): 9161-9172.