University of Florida Thesis Or Dissertation Formatting Template

OLFACTORY CORTICAL CONTROL DEFINES STRIATAL SENSORY REPRESENTATIONS

KATE A. WHITE

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2018

ACKNOWLEDGMENTS

First and foremost, I would like to thank my family and friends for their unyielding support. To my mother and father: your sacrifices have enabled me to follow my dreams, and it is my sincerest hope that I can one day repay you for all that you’ve done for me. Thank you for all the big and little things you do. To my brother, Dan: thanks for giving me the courage to follow my dreams and push me out of my comfort zone. To my grandfathers: thank you so much for your support and belief in me. To all my wonderful friends who have supported me despite my random absences from their lives, in particular, Sam, Ashli, Rachel, Jessi, Steph, and Sam: thank you so much for your long- distance support that kept me grounded throughout this process.

To my undergraduate mentors: Dr. G Andrew Mickley, who inspired my passion for science with his kind patience and true interest in my ideas and dreams. You gave me the tools and confidence to succeed where I thought I could not. To Dr. Christopher

Turner, who sat with me on weekends, early in the morning, and late in the evening to help me with my undergraduate thesis – you are no longer here, but you left a deep impact on me, even if we only knew one another for a short time. Thank you for teaching me persistence and to expand my scientific horizons.

To my lab mates, past and present: there are too many to name, but special thanks to the following: Sean Copley and Jamie LocPort from the Willis lab, you both made my transition into graduate school a wonderfully warm and welcome one, and I am deeply lucky to call you friends. To Dr. Marie Gadziola, who had the heavy task to teach me just about everything in the Wesson lab: thank you so much for your patience, thoughtfulness, and kindness. To Dr. Luke Stetzik, for being the best lab-roommate

during my time at the University of Florida: you seem to always know what to say, and I sincerely appreciate the guidance you gave me. To Kaitlin Carlson: short acknowledgement words are totally insufficient to thank you, but: I am deeply grateful to have been able to go through graduate school with such a fun, kind, passionate, and hilarious labmate. From struggling together, laughing together, traveling to conferences together, and being moth-lights together: thank you, so much, for being the best labmate, roommate, and friend I could have ever asked for.

To my committees, past and present: Dr. Roy Ritzmann, Dr. Hillel Chiel, Dr.

Brian McDermott, and Dr. Mark Willis from Case Western Reserve University, who gave me indispensable guidance in my early stages. To my current committee: Dr. Daniel

Wesson, Dr. Jeff Martens, Dr. Steve Munger, Dr. Gonzalo Torres, and Dr. Barry Setlow, for being willing to pick up where the other committee left off and giving me fresh and crucial insights into my project. Thank you all so much for shaping my mind as a young scientist. To my collaborators: Dr. Minghong Ma and Dr. Fuqiang Xu, thank you and your labs so much for your support, assistance, and hard work.

Lastly, and importantly, to my husband, Tony White: you’re the only reason I have been able to work so hard for as long as I have, and this PhD is yours just as much as it is mine. You supported me when I felt like I could not achieve my goals, and stepped up when I was swamped with grading and writing and experiments. You somehow managed to accept the fact that I needed to move away for the last year of my graduate training. You drove with me from Cleveland to Gainesville and tried to make this difficult situation somehow fun. Thank you for being by my side through this

process – your support has meant more than you will ever know, and I am so excited to show you that this hard work has been worthwhile.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 3

LIST OF TABLES ...... 8

LIST OF FIGURES ...... 9

LIST OF ABBREVIATIONS ...... 10

ABSTRACT ...... 11

CHAPTER

1 INTRODUCTION ...... 13

Importance of Sensory Information Processing ...... 13 Odor Processing Within Early Olfactory Structures ...... 15 Olfactory Bulb Dynamics During Odor Information Processing ...... 16 The PCX and its Association Fiber System ...... 19 The Role of the OT as an Olfactory and Striatal Structure ...... 28 Focus of Dissertation Research ...... 31

2 A CORTICAL PATHWAY MODULATES SENSORY INPUT INTO THE OLFACTORY STRIATUM1 ...... 35

Preface ...... 35 Materials and Methods...... 37 Animals...... 37 Stereotaxic Surgery and Viral Injections ...... 38 Behavior and Stimulus Presentation ...... 39 Optical Probe Fabrication ...... 40 In Vivo Electrophysiology and Optical Stimulation ...... 41 In Vitro Electrophysiology and Optical Stimulation ...... 41 Histology ...... 43 Data Analysis ...... 45 Results ...... 45 Viral Strategy for the Optogenetic Control of PCX Principal Neurons ...... 45 Activation of PCX Neurons Enhances OT Activity ...... 47 Activation of PCX Association Fibers Within the OT Modulates OT Activity ..... 49 Activation of PCX Association Fibers Bidirectionally Modulates the Representation of Odors in the OT ...... 50 PCX Principal Neurons Synapse with, and Evoke Monosynaptic Responses Within, OT D1- and D2-Type MSNs2 ...... 53 Topographical Organization of PCX Neurons Innervating OT D1- and D2- Type MSNs3 ...... 54

Discussion ...... 56 Contributions ...... 62

3 SUMMARY AND CONCLUSIONS ...... 78

Future Directions for Dissecting Secondary Olfactory System Connectivity ...... 79 Cellular Origins and Terminations of Secondary Structure Connections ...... 80 Neural Consequences of PCX Association Fiber Perturbation ...... 82 Behavioral and Perceptual Roles for the PCX-OT Connection...... 87 Caveats and Considerations ...... 92 PCX and OT as Components of a Cortico-Striatal Loop ...... 95 Role of Recurrent Circuitry in PCX and OT Odor Representations ...... 99

LIST OF REFERENCES ...... 104

BIOGRAPHICAL SKETCH ...... 122

LIST OF TABLES Table page

2-1 Number of mice used for in vivo experiments...... 62

LIST OF FIGURES Figure page

1-1 Simplified map of sensory information processing in the brain...... 33

1-2 Schematic of higher-order olfactory structures in the rodent brain...... 34

2-1 Olfactory fixed-interval task...... 63

2-2 Histological confirmation of electrode/fiber optic implantation sites for in vivo recordings...... 64

2-3 Viral strategy for the optogenetic control of PCX principal neurons...... 65

2-4 VGlut1+ co-labeling with AAV-mCherry PCX neurons...... 66

2-5 Modulation of OT activity by PCX neuron activation...... 67

2-6 Average background firing rate and light-evoked activity during PCX association fiber stimulation...... 68

2-7 PCX association fiber activation biases OT odor representation...... 69

2-8 PCX association fiber activation increases odor-evoked suppression across odors...... 70

2-9 The PCX bidirectionally modulates OT odor-responsive neurons...... 71

2-10 PCX stimulation leads to bidirectional modulation of odor-responsive OT units...... 72

2-11 Whole cell patch clamp recordings of PCX neurons and OT D1 and D2 MSNs reveal unique electrophysiological properties...... 73

2-12 Both OT D1 and D2 MSNs receive monosynaptic and polysynaptic input from the PCX...... 74

2-13 Glutamatergic PCX neurons directly innervate both OT D1 and D2 neurons...... 75

2-14 Innervation of OT MSNs by the PCX is topographically organized...... 76

3-1 Simplified schematic summarizing results...... 103

LIST OF ABBREVIATIONS

MSN Medium Spiny Neuron

OT Olfactory Tubercle

PCX Piriform Cortex

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

OLFACTORY CORTICAL CONTROL DEFINES STRIATAL SENSORY REPRESENTATIONS

Kate A. White

May 2018

Chair: Daniel W. Wesson Co-Chair: Jeffrey R. Martens Major: Medical Sciences - Pharmacology and Therapeutics

The ability for animals to process sensory information arising from the environment is crucial for survival. Sensory information from the environment is extended into the brain in pathways which are largely conserved across both animals and sensory systems. However, the olfactory system utilizes a distributed cortical coding scheme that bypasses obligatory thalamic processing, suggesting that olfactory cortical structures are critical for the encoding of olfactory information. The olfactory system is comprised of numerous cortical structures, including the piriform cortex (PCX) and olfactory tubercle (OT). The PCX, the largest olfactory cortical structure, encodes odor identity, intensity, and is integral for odor learning and memory. A major feature of the PCX is its extensive association fiber system which innervates other olfactory cortices, facilitating dense interconnectivity across the olfactory system. Specifically, the

PCX extends association fibers into the ventral striatum’s OT, which is known to integrate both olfactory and reward information. However, the function of this cortico- striatal pathway is unknown. In this dissertation, I present research which tested the

hypothesis that activation of PCX principal neurons increases OT neural activity in both the presence and absence of odor. I optically stimulated channelrhodopsin-transduced

PCX glutamatergic neurons or their association fibers while recording OT neural activity in mice performing an olfactory task. Activation of PCX neurons or their association fibers within the OT controlled the firing of some OT neurons and bidirectionally modulated odor coding dependent upon the neuron’s intrinsic odor responsivity.

Further, patch clamp recordings and retroviral tracing from D1 and D2 dopamine receptor-expressing OT spiny neurons revealed this input can be monosynaptic and that both cell types receive most of their input from a specific spatial zone localized within the ventro-caudal PCX. These results demonstrate that PCX odor information functionally accesses the direct and indirect pathways of the basal ganglia within the

OT. Broadly, the role for this cortico-striatal pathway yields implications for sensory integration into brain regions responsible for action selection. Furthermore, these results create a foundation on which to understand both higher-order connectivity across sensory systems and inform the way by which these interactions might give rise to sensory perception.

CHAPTER 1 INTRODUCTION

Importance of Sensory Information Processing

The way in which we experience the world is dependent upon our capacity to process and recognize sensory information arising from the environment. Across all animals, the ability to interpret sensory information is integral to both experiencing and surviving the world. Environmental representations are an orchestration of numerous sensory inputs all converging within the brain to form a perceptual whole. Take, for example, an essential survival behavior: the predator-prey interaction. A predator must combine several cues from the environment to catch a prey animal: visual recognition of the prey and the coordination of visual flow throughout the chase; scents and sounds emitted by a prey animal that are used to locate and follow it; tactile input to ultimately attack the prey and end this chase, along with myriad other potential environmental cues. The predator will then consume the prey, integrating gustatory and olfactory cues to determine if the prey is safe to consume. The prey needs to equally integrate these sensory cues from the predator to escape. Thus, the victorious animal is largely determined by its capacity to integrate these sensory cues and derive meaning from them to facilitate appropriate behavioral responses.

Given the integral role sensory information plays in basic survival mechanisms, the importance of understanding the way by which sensory information gives rise to perception cannot be understated. Determining how sensory systems encode and transform inputs creates a broad understanding of basic information processing mechanisms within the brain. Importantly, understanding this signal cascade is not just necessary for non-human animals. Humans require sensory inputs for essential day-to-

day functions, and moreover, numerous disease processes effect the encoding and representation of sensory information across systems. Neurodegenerative diseases often target early brain relays for sensory processing, including Alzheimer’s disease and

Parkinson’s disease (1–3). Sensory input is also aberrantly processed in disorders such as obsessive-compulsive disorder, autism spectrum disorders, and schizophrenia (4–6).

Thus, understanding the basic mechanics underlying sensory processing is not only crucial for determining the way by which the brain processes and interprets sensory information holistically, but will inform the causes of, and treatments for, diseases and disorders of the central nervous system.

Information is processed canonically across sensory systems (Fig 1-1A), and these pathways have been conserved across phyla with minimal modifications between

(7). This signal transduction cascade begins when sensory inputs are transduced by peripheral sensory receptors – for example, in the visual system, rods and cones within the retina absorb photons (8). This information is then extended up into the brain via nerve tracts (e.g. the optic nerve extending visual inputs) (8). From here, this input is then extended into the thalamus, which is known to be a central area for sensory processing that allows for the extraction of specific features of sensory information. It is the responsibility of the thalamus to both encode these features and extend them into downstream cortical structures. For example, visual information from the lateral geniculate nucleus of the thalamus extends into the visual cortex, wherein cortical subregions each encode discrete, defined stimulus features such as edge detection, movement, shape recognition, and depth (8).

This canonical information processing stream is also utilized in audition (9) and somatosensation (10, 11), taking sensory stimuli from the environment and distributing discrete components of this information up into higher-order structures after thalamic processing. However, one sensory system utilizes a different stimulus processing stream: olfaction (Fig 1-1B). The olfactory system both bypasses a mandatory thalamic relay and extends olfactory information directly from the olfactory bulb into numerous cortical structures (12), leading to a distributed coding scheme across the brain. Why did the olfactory system evolve to bypass the thalamus and instead utilize numerous downstream cortical structures to perform stimulus computations? The purpose of this dissertation research was to determine the role of the dense interconnectivity across the olfactory system by focusing on a single higher-order connection: that from the piriform cortex (PCX) into the olfactory tubercle (OT). However, to elucidate to the importance of this connection, it is first necessary to determine the way by which olfactory information is processed prior to entrance into secondary olfactory structures.

Odor Processing Within Early Olfactory Structures

Olfactory information (odor) is a unique sensory stimulus. Olfaction by necessity dictates three-dimensional informational space (13), as incoming odor information is encountered in nature as a plume structure, where odors are released from a source into the air as a volatile stimulus (14). Therefore, olfactory information itself contains both spatial (stimulus location) and temporal (timing of stimulus onset) information (14).

Given the complexity of this stimulus structure, what coding schemes are available to represent such rich information in the brain? Additionally, olfactory information contains more than just stimulus identity or intensity. Odors contain useful information about what the stimulus has meant previously and what it means currently, as well as hedonic

value, or valence (15, 16). Thus, resolving the logic behind the connectivity of the olfactory system, which allows for the optimal representation of such a complex stimulus nature, is a substantial undertaking.

Research over the last century (and beyond – e.g. (17)) has attempted to parse apart the manners by which olfactory information is processed and encoded across the brain, yielding insight into how this informational stream might function. Odors are first brought into the nasal epithelium during inhalation. The sampling of odor information requires active inhalations, or sniffing, as odors cannot enter the epithelium and bind to olfactory receptors without respiration (18). Odors in the environment act as ligands for olfactory receptors on olfactory sensory neurons within the nasal epithelium (7, 19, 20).

Odors and olfactory receptors do not ascribe to a “lock and key” design of ligand- receptor binding, such that one odor binds to a single olfactory receptor. Instead, each of the roughly 1000 olfactory receptor types bind odors with different affinities, or molecular receptive ranges (7, 19, 21–23). Overall, most odors serve as agonists for a number of different olfactory receptors, and those olfactory receptors, in turn, can bind a number of different odors with specific affinities (7, 19). Thus, these factors highlight the complexity of olfactory information input, and might begin to explain why the olfactory system utilizes an intricate coding scheme.

Olfactory Bulb Dynamics During Odor Information Processing

Within the nasal epithelium, there is a spatial distribution of olfactory receptor zones (20, 24, 25). Thus, within this peripheral site, there is an organized topographic representation of olfactory information. Odor information from the epithelium is extended via thousands of olfactory sensory neurons into the first central relay, the olfactory bulb

(12, 26–28). Olfactory sensory neurons expressing a certain receptor type coalesce into

neuropil structures termed glomeruli (26–29). These glomeruli integrate a single olfactory receptor’s information into a discrete space, with each glomerulus being innervated by between 1000-5000 olfactory sensory neurons expressing the same receptor (29, 30). Thus, glomeruli create a spatial representation of odor information across the superficial surface of the olfactory bulb (31–33).

Once odor information is extended into a glomerulus, it is read out by the principal neurons of the olfactory bulb, which are mitral and tufted cells (12). A glomerulus is innervated by the dendrites of anywhere between 20-50 mitral and tufted cells, a smaller proportion of which are mitral cells (34, 35). Importantly, each mitral or tufted cell extends into only one glomerulus (27, 28) and thus represent a single olfactory receptor’s worth of information up into higher-order structures. Considering that mitral and tufted cells extend information directly into these secondary structures, what information is encoded in these neurons, and in what way is this information processed?

Mitral and tufted cells have their own unique connectivity patterns and stimulus response features. Mitral cells are located more superficially than tufted cells within the aptly-named mitral cell layer of the olfactory bulb (29, 36, 37). Mitral cells extend their dendrites into the inner portion of the external plexiform layer (located deep to the mitral cell layer), where they receive more inhibitory inputs from granule cells than tufted cells do and extend their dendrites more extensively (38). Tufted cells are located within the external plexiform layer and extend their dendrites into the deeper outer portion of the external plexiform layer, where they also synapse with granule cells (38). Tufted cells intrinsically have lower thresholds to induce action potentials than mitral cells, and fire

action potentials at a higher frequency than mitral cells (39, 40). Tufted cells both respond to a broader range of odors and lower odor concentrations, yet respond more rapidly than mitral cells (38, 41–43).

Given the greater number of tufted cells extending into glomeruli, tufted cells receive more direct inputs from olfactory sensory neurons than mitral cells (34, 35).

Uniquely, mitral cells receive inputs from dendrites of tufted cells that acquire information from the same glomerulus (sister mitral and tufted cells), indicating that tufted cells are receiving more direct odor input from olfactory sensory neurons, while mitral cells are more likely to receive information previously processed by tufted cells

(40, 44). In terms of density, distribution of tufted cells are sparse within the external plexiform layer, and it is unlikely for a tufted cell to synapse with another; this is not the case with mitral cells, which are often packed more densely together (37). These neurons also project to different higher-order targets within the brain: tufted cells project largely to the OT, anterior olfactory nucleus, olfactory peduncle, and the ventro-rostral portion of the PCX (45–47). Mitral cells project broadly to secondary structures, including the remainder of the PCX, lateral entorhinal cortex, amygdala, and anterior olfactory nucleus, (36, 45, 48). Given these differences, olfactory information from principal neurons likely conveys different facets of olfactory information up into secondary structures.

Mitral and tufted cell outputs are refined by layers of inhibition within the olfactory bulb, specifically among the glomerular and external plexiform layers. Within the glomerular layer, GABAergic juxtaglomerular neurons laterally inhibit other glomeruli, sharpening output at this level (49, 50). This inhibition acts to dampen weaker stimulus

inputs, allowing the strongest input to extend its information up into higher-order structures (51–55). Deeper in the external plexiform layer, granule cells originating from the granule cell layer, the deepest layer of the olfactory bulb, form dendrodendritic synapses with mitral and tufted cells that mediate principal neuron output (53–56). This further serves to sharpen and refine olfactory bulb outputs. These layers of inhibition therefore modify the molecular receptive ranges of mitral and tufted cells, modulating principal cell output and performing complex transformations at this first brain relay (57).

Thus, at this first central structure, odor information has already been spatially segregated dependent upon receptor activation and further refined by inhibitory interneurons. Consequently, information received by olfactory cortical structures is not unprocessed – this odor information undergoes some refinement and sharpening within the olfactory bulb. Once this information is extended into olfactory cortical structures, in what way is this information represented?

The PCX and its Association Fiber System

Considering the projection patterns of olfactory bulb principal neurons, what is the role of each olfactory cortical structure in olfactory system processing? Each one of these higher-order downstream olfactory structures participates in the encoding of olfactory information (e.g. (58–71)), but the role for each structure in olfactory processing is complicated by two features. First, there is a high degree of interconnectivity across cortical structures (29, 72–74)(Fig 1-2), and two, there are feedback projections from a number of these structures onto the olfactory bulb that further modulates afferent outputs (75–77). This alone makes it difficult to determine what any single olfactory structure might do in isolation.

Thus, rather than examining the role of each olfactory structure and what is currently known about their patterns of connectivity, it is more useful to begin with a structure that is rather well-characterized. By beginning with a structure in which there is information on anatomy, circuitry, and knowledge about the way by which this structure encodes olfactory information, the role of connectivity between higher-order structures can be more easily elucidated. Thus, for this dissertation research, I sought to begin with a structure that has been well-studied so that the results revealed by examining a pathway originating from this structure can provide a contextual framework by which to interpret the results. Here, I sought to examine the function of the connectivity from the

PCX into the OT, given what is already know about each structure.

To begin, the PCX is the largest structure receiving direct olfactory bulb input and is considered to be the “primary” olfactory cortex (12, 78). The PCX is a trilaminar paleocortex with both cortical and associational features (12). Based upon anatomy and circuitry, the PCX can be subdivided into the anterior and posterior portions, with further subdivisions of anterior PCX determined by dorsal and ventral segregation via the positioning of the lateral olfactory tract (78). There are a few notable differences between anterior and posterior PCX. The anterior PCX receives denser innervation by olfactory bulb afferents (48, 66, 79, 80). In contrast, the posterior PCX receives fewer afferent inputs from the olfactory bulb but receives associational innervation from anterior PCX and other extrinsic structures, like lateral entorhinal cortex and amygdala

(72, 73, 81). Olfactory bulb inputs rapidly excite anterior PCX neurons, evoking strong postsynaptic potentials, while posterior PCX receives its input from thinner olfactory bulb axons and is activated by this afferent input less frequently and strongly (82, 83).

Thus, it is thought that the anterior PCX represents odor identity, while posterior PCX integrates the broader meaning of an odor due to the segregation of afferent and associational inputs into these subdivisions (15, 16, 78, 84–86).

Similarly, both the anterior and posterior PCX are trilaminar (although the density of lamination changes from anterior to posterior, with thinner layer ii and thicker layer iii throughout the posterior portion)(78). Layer i can be subdivided into layers ia and ib, in which lateral olfactory tract information enters into layer ia and synapses with both principal cells and interneurons of the PCX (12, 78). Layer ib serves the role of integrating associational inputs from PCX principal neurons (87–89). Layers ii and iii are predominantly comprised of glutamatergic principal neurons, which are pyramidal neurons (both superficial and deep) and semilunar cells (12, 87, 90). Pyramidal and semilunar cells share some similarities, with their primary differences arising in their patterns of connectivity both within and outside of the PCX (12, 87, 90). Both pyramidal and semilunar cells receive afferent inputs from the olfactory bulb (87). Interestingly, mitral and tufted cells arising from the same glomerulus do not project their axons to the same neurons in the PCX, both creating a dispersion of olfactory receptor-specific information and immediately disregarding the spatial patterning imposed by glomeruli within the olfactory bulb (91, 92). However, principal neurons receive overlapping patterns of innervation from mitral and tufted cells representing separate glomeruli (91,

93–95), suggesting roles for both distributed and convergent coding of olfactory information within the PCX.

While both semilunar and pyramidal neurons receive inputs from olfactory bulb principal neurons, semilunar cells receive comparably stronger afferent inputs, evoking

larger excitatory post-synaptic potentials, while pyramidal neurons receive stronger associational inputs both from within the PCX itself, and extrinsically from other structures (79, 87, 90, 96, 97). This input scheme is intuitive when considering each cell’s morphology and connectivity patterns. Semilunar cells have a greater number of apical dendrites, with dense distal spines extending towards layer i to receive afferent inputs (78, 87). Pyramidal neurons have basal and apical dendrites that are both spiny in nature, perfectly situating them to receive inputs from both superficial and deep layers of the PCX (78, 87). Input convergence onto pyramidal neurons also favors an associational scheme: a single pyramidal neuron is estimated to receive 200 afferent inputs, but receives 2000 associational inputs, with both inputs converging seemingly randomly onto these principal neurons (98, 99). That is, there appears to be no patterning for which axons from what neuron types synapse onto specific pyramidal neurons.

Beyond these differences, both semilunar and pyramidal neurons display low spontaneous firing rates, at least when compared with the inputs received from mitral and tufted cells from the olfactory bulb, which suggests some level of filtering prior to integration into PCX (41, 93, 100, 101). Semilunar and pyramidal neurons both need input from more than one glomerulus during a discrete time period to evoke a response

(99), and roughly 10% of these PCX principal neurons are responsive to a single olfactory input (66, 101, 102). These similarities suggest that this input likely arises directly from afferent projections onto semilunar cells, while pyramidal neurons are potentially influenced both by afferent and intrinsic associational projections within the

PCX. Thus, the PCX likely comprises two layers of olfactory processing: one that is

responsible for afferent inputs, and the other that is responsible for integrating associational inputs both within and outside of the PCX (87, 90).

What intrinsic features of the PCX allow pyramidal neurons to receive and integrate this associational input? Semilunar cells project largely to other pyramidal neurons in the PCX, while pyramidal neurons synapse primarily with other pyramidal neurons (87, 103, 104). Thus, pyramidal neurons are altogether integrating afferent inputs direct from the olfactory bulb, input from semilunar cells, and additionally, inputs from other pyramidal neurons. This informational integration scheme would likely lead to runaway excitation within the PCX on its own: afferent input onto semilunar cells would lead to excitation of pyramidal neurons, activating larger ensembles of excited pyramidal neurons via glutamatergic connections, and overall leading to global activation of this principal neuron population. Importantly, the PCX mediates its own excitation via principal neuron connectivity with dense and numerous inhibitory interneurons scattered throughout PCX (88, 105).

GABAergic inhibitory interneurons of the PCX play a critical role in shaping olfactory input onto principal neurons, dampening recurrent excitation and allowing for refinement of olfactory inputs intrinsically (88, 105). Inhibitory interneurons (with some disagreement (87)) can be divided into two populations: feedforward interneurons, restricted within layer i, and feedback interneurons, more densely populated within layer iii (88, 89, 106, 107). The two primary feedforward interneuron types are horizontal cells and neurogliaform cells, which act to regulate input extending directly from the olfactory bulb into the PCX (105, 108–110). In contrast, feedback interneurons are confined to the deeper associational layers and regulate feedback onto principal neurons (88, 89,

106, 111). During olfactory bulb stimulation, feedforward inhibitory circuits are depressed across stimulation trains, while feedback inhibition is facilitated, indicating that there is a shift in the inhibited population of neurons which allows for the control of spike timing during sensory input (110). This increase in inhibition is likely due to recruitment of semilunar cells and pyramidal neurons, which then increases the amount of feedback inhibition via recurrent circuitry (105, 110). Inhibitory interneurons also have unique odor tuning properties. As a whole, these neurons are broadly tuned to incoming odor inputs, far more than principal neurons (101, 102). It has been postulated that interneurons receive a higher degree of mitral cell convergence from separate and diverse arrays of glomeruli, allowing for a broad tuning curve (102). This broad tuning has also been demonstrated with odor mixtures, which evoke a strong inhibition across the PCX, allowing changes in odor tuning across ensembles (65, 112–114). These inhibitory interneurons are therefore regulating the activity of principal neurons dependent upon the input stimulus, preventing runaway excitation and shaping the coding of odors in the PCX. Thus, just as there are two layers of glutamatergic olfactory processing within the PCX, there are also two layers of GABAergic mediation of inputs.

In what way do these cellular and circuit-level features converge to modulate olfactory representations within the PCX? While the PCX has cortical components, it also contains features that suggest associational processing (12). Importantly, synaptic features suggest that afferent inputs from olfactory bulb to PCX are not involved in associational processes. Afferent inputs onto principal neurons of the PCX appear to be hardwired, albeit randomly distributed – in contrast, associational connections within the

PCX are highly plastic and undergo significant reconstruction due to long-term

potentiation that is dependent upon stimulus input (115, 116). Indeed, spike timing- dependent plasticity could not be elicited at afferent inputs onto layer ii and iii principal neurons, but could be elicited at their associational synapses, so long as the postsynaptic pyramidal neuron was already active (117). Principal neurons within the

PCX also differ in their short-term synaptic plasticity, which shapes how afferent information is encoded across the PCX (90, 96, 118). Dendrites of both principal neurons types are only weakly activated by incoming inputs, thus neurons likely summate inputs rather than responding strongly and sparsely to a single input (119).

Therefore, principal neurons need to be activated by a number of inputs to be excited, an intrinsic property that depends on the connectivity scheme and number of inhibitory interneurons (87). Overall, both afferent and associational inputs can only evoke a postsynaptic response when a number of inputs converge onto pyramidal neurons, which is an important feature as the connection probabilities of these cells are low, evoked activity of principal neurons is weak, and associational connections are far more numerous (97, 98, 120). This leads to the idea that it is the number of inputs into the

PCX that matters, not necessarily the strength of those inputs, which is reliant upon input from associational networks. Associational input is integral for PCX synaptic plasticity and ultimately, the role of PCX as an associational structure. Thus, both afferent and associational inputs are integral to odor representation within the PCX, which hints at the role the PCX performs in olfactory processing.

The PCX performs numerous roles in the encoding of olfactory information. The

PCX encodes the identity of odors across its structure both in vivo and in vitro (65, 66,

80, 101, 102, 112, 121–123). Importantly, the PCX encodes different odor identities

across distributed, non-topographically-organized neural ensembles, contrasting the ordered inputs from the olfactory bulb (36, 65, 66, 102, 124). PCX neurons tend to be

(although not exclusively) broadly tuned to odors (65, 70, 102, 125–127), although this appears to be a feature mediated by associational connectivity (127, 128). The PCX also rapidly and specifically habituates to odors, decreasing both the firing rate of neurons and the number of neurons that are recruited with sustained odor presentations, both in awake and anesthetized animals (69, 80, 129–133). The PCX responds to odor mixtures, and can distinguish between a novel odor stimulus arising from a background odor or odor mixture (84, 134–136). The PCX also encodes an odor’s intensity (concentration), although the exact way neurons represent this information is subject to debate. There is evidence both for increasing intensity recruiting more PCX neural ensembles (69, 137), or an increase in the firing rate of PCX neurons with greater odor intensity (96, 138, 139). In addition, PCX odor responses are context-dependent, as PCX neurons respond to task-specific stimuli such as anticipation or reward (67, 68, 122, 126). PCX neural ensembles perform pattern completion and separation dependent upon the task (15, 134, 136, 140, 141), leading to odor identification through plastic circuits that can be excited or inhibited based upon previous encounters. This suggests that the PCX forms holistic odor percepts from the convergence of not only olfactory input, but also state and contextual cues that give the odor meaning (12, 15, 16). This gives some indication as to the role of PCX inputs into other structures, leading to the main question of this dissertation research: what is the role of the connectivity from the PCX into other higher-order olfactory structures?

The PCX is an association cortex, and one of its most important features that allows it to be a powerhouse of odor representation arises from its extensive extrinsic association fiber system (12, 72, 74). This association fiber system originates from semilunar and pyramidal neurons within the PCX that extend their axons into outside structures, although further work needs to be done to classify which cell type is responsible for each structure’s input (72, 74, 87, 90, 142, 143). Regardless, these principal neurons extend numerous association fibers into a number of structures, including the OT, orbitofrontal cortex, anterior olfactory nucleus, lateral entorhinal cortex, amygdala, and thalamus (12, 72, 144–146). What is the purpose of this interconnectivity, and what role does this association fiber system ultimately play in olfactory processing?

There is unfortunately little evidence on the role of association fiber connections from PCX onto other structures, and what is currently known on the association fiber system itself is sparse. Anatomical evidence of these associational projections arises from circuit reconstruction largely via intracellular tracing (72, 147, 148), but the function of this connectivity is unknown except for a handful of studies. Lesioning the lateral entorhinal cortex, which is reciprocally connected to the PCX, leads to an increase in

PCX spontaneous activity, impairing an animal’s ability to perform a difficult odor discrimination (61). Additionally, optogenetic stimulation of the basolateral amygdala, also reciprocally connected to PCX, modulates PCX neural activity, decorrelating odor representations within the PCX (149). Beyond this, no other research is available on the role between secondary olfactory structures, and neither of these studies examined the feedforward association fiber connection from the PCX. Despite being the primary

olfactory cortex and having broad influence over numerous brain structures, evidence for the role of this connectivity is scarce. Thus, here I sought to test a single component of this interconnectivity: the connection from the PCX into the OT.

The Role of the OT as an Olfactory and Striatal Structure

The OT is a trilaminar “cortical-like” brain structure, divided into the ventral-most molecular layer (layer i), the dense cell layer (layer ii), and the multiform layer (layer iii)

(150, 151). Layer i contains sparse cell bodies from pial cells, spine-rich cells, and radiate cells (152), and is predominantly comprised of afferent input axons and dendrites from cells within other layers (150). One of the unique features of the OT is that its laminar nature is not smooth, but undulates across the structure (151). This is largely due to two features: cap compartments creating gyri within layer ii, and dense cell clusters (primarily comprised of granule cells) called islands of Calleja within layer iii that push deep into the OT (153–155). The dense cell layer is primarily composed of the principal neuron type within the OT, GABAergic medium spiny neurons (MSNs) that express either dopamine receptor 1 (D1) or 2 (D2) (150, 156). These MSNs also express NMDA and AMPA receptors (157), and thus are able to receive glutamatergic inputs. MSN dendrites extend into layer i to receive afferent inputs and extend their axons deeper into layer iii, forming vertically-oriented continuations with the ventral pallidum and nucleus accumbens (158). Evidence from the nucleus accumbens suggests that these MSNs have the capacity to synapse with and modulate the activity of one another (159). MSNs are not the only cell type within the dense cell layer. Layer ii and iii also contains crescent cells, spine-poor cells, and spindle cells, although further characterization of these cell types will be required to determine their roles within the

OT (150). Layer iii is innervated by axons from other structures, and has sparse cellular

representation, appearing to be comprised mostly of interneurons (150). However, there is a distribution of multiple neuron types across layers ii and iii that are characterized by their spiking activity, creating a heterogenous cell population within the OT (160).

The OT shares similarities with the nucleus accumbens, another ventral striatal structure, in terms of cellular composition and connectivity (150, 161, 162). Considering that the OT contains both D1 and D2 MSNs, it participates in both the direct and indirect pathways of the basal ganglia which facilitate or suppress motivated behaviors (163–

165). The OT both integrates and extends a combination of sensory and reward information into other structures, both within and extrinsic to the basal ganglia. The OT receives inputs from the olfactory bulb, PCX, anterior olfactory nucleus, orbitofrontal cortex, amygdala, hippocampus, thalamus, and brainstem regions, among other areas

(47, 73, 153, 166–171). The OT extends projections into the nucleus accumbens, ventral pallidum, thalamus, hypothalamus, and areas of the brainstem (153, 166, 168–

171). Importantly, as a structure involved in reward processing, the OT has extensive bidirectional connections with the ventral tegmental area (156). Beyond that, the OT is unique because it is a “cortical-like” structure that does not extend association axons to secondary olfactory areas (73). Thus, much in the way that the PCX is highly interconnected with other olfactory structures, the OT is the opposite, receiving olfactory input but projecting predominantly to areas involved in action selection and motivated behaviors (153, 172). Therefore, as much as the OT receives direct projections from the olfactory bulb and dense innervation from the PCX (173), it is far less olfactory and far more striatal, in terms of its cell types, connectivity, and information processing.

The OT processes both olfactory and reward information. Animals will readily self-administer cocaine directly into the OT, even more so than administration into the nucleus accumbens (174). OT stimulation recruits forebrain reward structures including the nucleus accumbens and modulates odor preference behaviors (175). OT neurons encode the magnitude of rewards and are active during motivated behaviors (176). OT lesions also disrupt more complex actions, perturbing sociosexual behaviors, generally decreasing movement, reducing drive to ingest food, and modulating sensory-driven attentive behaviors (177–182). Silencing of OT neurons also decreases preference for chemosignals from oppositely-sexed mice (183), disrupting normal social behaviors.

Amphetamine administration directly into the OT also enhances attention towards an odor (184), suggesting that the OT is a site that integrates and facilitates complex behavioral outputs that are critical to survival, such as the ability to attend to a stimulus, reproduce, and feed. Thus, the OT is of broad interest as it is not only an olfactory sensory structure, but a basal ganglia structure, which has wider implications in terms of behavioral facilitation and output.

The majority of neurons in the OT display low spontaneous firing rates, between

~2-5 Hz dependent upon if the animal is awake or anesthetized (70, 71, 185, 186). OT neurons and PCX neurons are rather similar in terms of olfactory information encoding.

OT neurons can either be excited or suppressed by odor, and a single neuron can encode either a broad or narrow range of odor identities (69–71, 187). OT neurons, like

PCX neurons, also encode both single odors and odor mixtures (71, 185). The number of odor-responsive OT neurons increases as odor intensity increases (69). The OT can habituate to odors, but to a lesser extent than the PCX does (69). Importantly, the OT

encodes the hedonic value of an odor stimulus, or valence. An OT neuron that is responsive to one odor that is coupled with a reward will decrease its responsivity if that reward response is abolished (185). Conversely, a previously unresponsive OT neuron will increase its firing rate to an odor if it goes from being unrewarded to rewarded (185).

Indeed, different subregions of the OT appear to code appetitive and aversive learned odors (188). Overall, this demonstrates the OT’s capacity to encode both olfactory inputs and reward information, highlighting the potential importance of this structure in the integration of inputs necessary for appropriate behavioral facilitation.

Focus of Dissertation Research

Importantly, anatomical evidence demonstrates that the OT receives significantly more inputs from the PCX than from the olfactory bulb, which suggests that the PCX is ideally situated to influence OT odor coding (173). Considering that the PCX and OT perform some overlapping roles in terms of odor processing, what is the role of the connectivity from PCX into the OT? Is the PCX capable of modulating odor-evoked activity within the OT? The function of this connection, to date, is unknown. Further, where this connectivity originates from in the PCX and terminates within the OT has yet to be determined. Thus, the aim of this dissertation research is to establish a role for the connectivity from PCX into OT, both in the presence and absence of olfactory information. Considering how little is known about the topographical features of the

PCX, and the pattern of projections from the association fiber system, in this dissertation research, I sought to map inputs from PCX onto OT neurons, specifically determining if PCX neurons might exert greater influence over D1 or D2 MSNs, and if stimulation of PCX neurons evokes activity in a certain sub-population of OT neurons.

The next chapter will describe the way in which this research was conducted and the results therein, but to summarize: I recorded extracellular, single neuron activity from the OT of mice performing an olfactory task with or without optogenetic PCX principal neuron or association fiber activation. In the first experiment, I defined how

PCX neuron activation modulates background activity within the OT. I next demonstrated how activation of PCX association fibers modulates OT odor coding.

Lastly, as part of a collaborative project, I determined if there was a difference in the projection patterns from PCX neurons onto D1 or D2 MSNs using patch clamp recordings and viral tracing. I found that PCX principal neuron activation led to modulation of OT neuron spontaneous activity, dependent upon if I stimulated the PCX directly, or if I stimulated association fiber axon terminals within the OT itself. By stimulating these neurons and/or their association fibers, I determined that activation of

PCX predominantly suppressed odor responses within the OT at the population level.

However, within a single neuron, there is bidirectional modulation of firing rate to PCX activation that is dependent upon an OT neuron’s intrinsic responsivity to odor. Finally, I determined that PCX principal neurons synapse with both D1 and D2 MSNs without apparent preference, but largely extend from the ventro-caudal portion of the PCX.

Thus, this dissertation work demonstrates that the PCX has the capacity to control sensory representation within the OT, and that the PCX influences both D1 (direct) and

D2 (indirect) pathways of the basal ganglia, ultimately modulating how sensory information might be encoded downstream. This work yields insight into the function of distributed coding schemes in sensory processing, which further informs both sensory and cortico-striatal processing across the brain.

Figure 1-1. Simplified map of sensory information processing in the brain. A) Canonical sensory processing stream utilized by most sensory systems in the mammalian brain. B) Olfactory system processing stream in the mammalian brain.

Figure 1-2. Schematic of higher-order olfactory structures in the rodent brain. Displayed here is a subset of cortical structures within the mouse brain. Arrows indicate the direction of connectivity between structures. OB = olfactory bulb, AON = anterior olfactory nucleus, PCX = piriform cortex, OT = olfactory tubercle, CoA = cortical amygdala, lEC = lateral entorhinal cortex. Simplified and adapted from (29).

CHAPTER 2 A CORTICAL PATHWAY MODULATES SENSORY INPUT INTO THE OLFACTORY STRIATUM1 Preface

How does the brain distribute sensory information to enable the successful encoding of stimuli? In nearly all mammalian sensory systems, information from the environment is transduced by peripheral sensory receptors and relayed into the thalamus, where this input is processed and then distributed into cortical structures that encode stimulus attributes (189). However, this scheme is not utilized in the olfactory system wherein odor information bypasses the thalamus and enters directly into cortical structures, suggesting that it is the interaction of olfactory cortical structures that is integral for the generation of an odor percept (12). In the mammalian olfactory system, odors are transduced by olfactory sensory neurons which extend their axons into olfactory bulb glomeruli (7, 190). From here, mitral and tufted cells (the principal neurons of the olfactory bulb) relay odor information into secondary olfactory structures, including the piriform cortex (PCX), olfactory tubercle (OT), anterior olfactory nucleus, and cortical amygdala, among others (12, 29). In addition to extensive bulbar connections, these cortical structures are also heavily interconnected (for review see

(12, 78, 172)). Thus, there is diffuse connectivity throughout the olfactory system which allows for the potential to transform odor information.

The PCX, often referred to as the ‘primary’ olfactory cortex, extends massive numbers of glutamatergic association fibers throughout the brain which innervate other olfactory structures (12, 78, 146, 148). PCX neural ensembles precisely encode certain odor features (65, 66, 102, 131, 132, 138), suggesting that the association fiber network

1Reprinted with permission from (191)

serves to relay odor information among brain structures. Local association fibers within the PCX shape the encoding of odors by plastic actions and recurrent circuitry (90, 96,

98, 103, 192–194). Thus, do PCX association fibers shape the representation of odors in interconnected olfactory structures? Here we tested this question by interrogating the influence of PCX association fiber input on the OT. The OT receives direct input from the olfactory bulb (45, 47, 173), and like the PCX, encodes odor identity and intensity

(69, 70). However, the OT is the only olfactory structure which is also a component of the ventral striatum (195). This characteristic allows the OT to shape sensory-directed motivated behaviors (153, 156, 176). As is the case with other striatal structures, the principal class of neurons in the OT are medium spiny neurons (MSNs) (196). These

MSNs express either D1- or D2-type dopamine receptors, with striatopallidal MSNs

(those mostly projecting to the ventral pallidum/globus pallidus) expressing D2 receptors and striatonigral MSNs (innervating substantia nigra or ventral tegmental area) expressing D1 receptors (163, 173) (for review (164, 165)). Thus, the OT is preferentially situated, as the olfactory striatum, to send sensory information into downstream basal ganglia structures via the direct (D1) and indirect (D2) pathways, extending projections into both midbrain and other striatal structures (153, 197).

We predicted that the PCX shapes odor processing within the OT by means of the PCX association fiber network. PCX association fibers innervate all layers of the OT

(72, 147, 148), and stimulation of the PCX ex vivo elicits postsynaptic potentials in the

OT (198). In vivo recordings also suggest that network oscillations in the ventral striatum originate in the PCX (199). It remains unknown whether, and if so how, PCX inputs modulate olfactory processing in the OT. Further, the spatial population of

neurons within the PCX that innervate the OT, and the OT neuron types receiving this

PCX input, are both unresolved. To address these voids, we recorded the activity of OT neurons in behaving mice while optically stimulating channelrhodopsin (ChR2)- expressing principal neurons in the PCX and/or their association fibers in the OT while mice performed an olfactory task. We then employed patch-clamp recordings and viral tracing methods to determine the connectivity of PCX neurons upon OT D1 and D2

MSNs. Our work uncovers that the PCX controls both the number of odor responsive

OT neurons and the direction of their response. Moreover, OT D1 and D2 MSNs receive monosynaptic inputs from neurons which are topographically organized to be largely within the ventro-caudal anterior PCX.

Materials and Methods

Animals

For in vivo experiments, 2-5 months old C57BL/6 male mice (n=17 for experiment 1, n=23 for experiment 2) were group-housed on a 12 hour light/dark cycle with food and water available ad libitum except when water was restricted for behavioral training (see Behavior subsection below). Mice were single-housed upon intracranial implantation. For in vitro experiments, D1-tdTomato (200) and D2-EGFP (Tg(Drd2-

EGFP)S118Gsat) BAC (201) transgenic male and female mice (n=9; 1-2 months old) were crossed to obtain mice with dopamine D1 and D2 receptor-expressing MSNs labeled in red and green fluorescence, respectively. All experimental procedures were performed in accordance with the guidelines of the National Institutes of Health and were approved by Institutional Animal Care and Use Committees at all institutions.

Stereotaxic Surgery and Viral Injections

For in vivo and in vitro experiments, mice were anesthetized with isoflurane (2-

4% in oxygen; Patterson Veterinary, Greeley, CO) and mounted in a stereotaxic frame with a water-filled heating pad (38°C) to maintain the mouse’s body temperature.

Anesthetic depth was verified throughout. A local anesthetic (1% bupivacaine, 0.05 ml, s.c.) injection was administered into the wound margin prior to exposing the dorsal skull.

For viral injections, a craniotomy was made above the anterior PCX (A/P: -1.0mm, M/L:

+2.8mm, D/V: +3.5mm), and either a 33Ga Hamilton microsyringe or a glass micropipette was lowered into the PCX. 0.5 µl of

AAV5.CaMKiiα.hChR2.E123T.T159C.p2A.mCherry.WPRE (cell-filling variant) or

AAV5.CaMKiiα.hChR2(H134R).mCherry (non-cell-filling variant), or control vector

AAV5.CaMKiiα.mCherry.WPRE (all undiluted, University of North Carolina Viral Vector

Core, Chapel Hill, NC, USA), was infused by a pump at a rate of 0.05 µl/min. The syringe/pipette was withdrawn, the craniotomy sealed with wax, and the wound margin closed. An additional stereotaxic surgery was performed for in vivo experiments to either implant an optical fiber in the PCX and electrode array in the OT (PCX stimulation) or an optetrode into the OT (PCX association fiber stimulation). For fiber and array implants, a craniotomy was made dorsal to the anterior PCX and a glass optical fiber was lowered into the PCX. Another craniotomy was made dorsal to the OT, and an 8-channel tungsten wire electrode was lowered into the OT. For optetrode implants, a craniotomy was made dorsal to the OT, and an optetrode assembly was lowered ~0.3 mm dorsal to the OT. All craniotomy sites were sealed with wax and the implants cemented in place, along with a head bar for head-fixation during behavioral experiments (see Behavior subsection below). During the recovery period after all

surgeries, mice received a daily injection of carprofen (Pfizer Animal Health) or meloxicam (Patterson Veterinary; 5 mg/kg, s.c. for both) and allowed ad libitum access to food and water.

For viral tracing experiments, viral tools for trans-mono-synaptic labeling were packaged by BrainVTA (BrainVTA Co., Ltd., Wuhan, China). The helper virus AAV9-

EF1a-DIO-histone-BFP-TVA and AAV9-EF1a-DIO-RV-G were titrated at about 3×1012 genome copies per milliliter, and RV-EnvA-ΔG-GFP was titrated at about 2x108 infecting unit per milliliter. The mixture of AAV9-EF1a-DIO-histone-BFP-TVA and AAV9-

EF1a-DIO-RV-G (volume ratio: 1:1, 100 nl in total) was injected into the OT (A/P:

+1.2mm, M/L: +1.1mm, D/V: +5.5mm) in D1R-Cre (n=4) and D2R-Cre (n=4) mice, respectively. Two weeks later, 150nl of RV-EnvA-ΔG-dsRed was injected into the same location of these mice. One week after the RV injection, the mice were perfused transcardially with PBS (Pre-treated with 0.1% diethylpyrocarbonate (DEPC, Sigma)), followed by ice-cold 4% paraformaldehyde (PFA, 158127 MSDS, Sigma)

Behavior and Stimulus Presentation

Following surgical recovery, mice for in vivo experiments were water-restricted on a 23-hour schedule to no less than 85% of their body weight to induce motivation to perform the task. For the fixed-interval olfactory task, mice were head-fixed in a tube with an olfactometer and lick spout directly in front of the mouse’s snout and mouth, respectively. Mice were habituated to head-fixation for 15 minutes/session for 2-3 days prior to stimulus presentation with light isoflurane anesthetization to minimize stress.

Mice then began learning the fixed-interval olfactory task (Fig 2-1), with the following structure for each trial: 1) trial start for 12 seconds, 2) 1 of 9 pseudorandomly presented stimuli for 2 seconds; 3) 8 second post-stimulus rest period; 4) access to fluid reinforcer

(2 mM saccharin) for 10 seconds prior to new trial start to maintain task motivation (see

Fig 2-1). Mice learned to lick for reward to all stimuli presented and worked up to at least 9 trials of each stimulus presentation per session (~1 hour session/day).

Pseudorandomly-presented stimuli were one of the nine following: 4 monomolecular odors (isopentyl acetate, heptanal, ethyl butyrate, 1,7 octadiene; Sigma-Aldrich, St.

Louis, MO), the same 4 odors + 10 Hz (20 ms pulse width) PCX light stimulation, or light stimulation alone. Odors were delivered via an air-dilution olfactometer via independent lines of tubing. All odors were diluted to 1 Torr in mineral oil and delivered to the olfactometer via medical-grade nitrogen at a flow rate of 1L/min. Mice needed to lick for the fluid reinforcer post-stimulus in at least 85% of trials for the session to be included for analysis of neural activity. Trials were divided into correct (licked for fluid reinforcer after stimulus) and incorrect (did not lick for reinforcer after stimulus presentation) responses.

Optical Probe Fabrication

For in vivo experiments, all optical fibers, electrode arrays, and optetrodes were custom-made. For optical fibers, glass multimode fiber (300 µm core, 0.39NA; Thorlabs,

Newton, NJ) was cut to the appropriate length and fastened with optical adhesive

(Norland Products, Cranbury, NJ) in a 2.5 mm ceramic ferrule (Thorlabs). For fixed electrode arrays, tungsten wire (A-M Systems, Sequim, WA) was attached to an

Omnetics connector via silver epoxy, with a stainless steel wire (A-M Systems) serving as the ground wire. Tungsten wires were bundled in two polyimide tubes and cut to the appropriate length. Optetrode assembly was performed in accordance with (202). Light intensity output of the final pre-implanted fiber was 7-10 mW3 and the distance between

tetrode tips and fiber end were 500-800 µm apart to yield a broad light cone for neuron activation.

In Vivo Electrophysiology and Optical Stimulation

The connector of the fixed electrode array or optetrode was connected to a headstage and electrode channels were digitized to acquire multi-unit activity (24 kHz,

200-3kHz band-pass filter) and lick events (300 Hz) for behavioral criterion monitoring.

Light stimulation was provided by 447.5nm LED (Luxeon Rebel ES, Luxeon Stars,

Lethbridge, Alberta) driven by a Thorlabs T-Cube LED driver (Thorlabs) connected to a fiber patch cable (ThorLabs; 300µm core multimode fiber) for temporary connection to the implanted ferrule on the animal. Light stimulus intensity, timing, and frequency were determined after conducting pilot experiments where we stimulated the PCX between 5-

30 Hz at 20 ms pulse width, and intensity was adjusted for every mouse to the lowest intensity possible while still evoking light-induced responses in the OT. For PCX association fiber stimulation experiments, tetrodes on the optetrode were advanced

50µm/session to access new neurons within the OT. For the fixed array, 1 of the 8 channels from the multiunit recordings was used as a local reference channel for signal subtraction.

In Vitro Electrophysiology and Optical Stimulation

For in vitro whole-cell patch clamp recordings, mice were deeply anesthetized with ketamine-xylazine (200 and 20 mg/kg body weight, respectively) and decapitated.

The brains were dissected out and immediately placed in ice-cold cutting solution containing (in mM) 92 N-Methyl D-glucamine, 2.5 KCl, 1.2 NaH2PO4, 30 NaHCO3, 20

HEPES, 25 glucose, 5 Sodium L-ascorbate, 2 Thiourea, 3 Sodium Pyruvate, 10

MgSO4, and 0.5 CaCl2; osmolality ~300 mOsm and pH ~7.3, bubbled with 95% O2-5%

CO2. Coronal sections (250 µm thick) containing the PCX and OT were cut using a

Leica VT 1200S vibratome. Brain slices were incubated in oxygenated artificial cerebrospinal fluid (ACSF in mM: 124 NaCl, 3 KCl, 1.3 MgSO4, 2 CaCl2, 26 NaHCO3,

1.25 NaH2PO4, 5.5 glucose, and 4.47 sucrose; osmolality ~305 mOsm and pH ~7.3, bubbled with 95% O2-5% CO2) for ~30 min at 31ºC and at least 30 minutes at room temperature before use. For recordings, slices were transferred to a recording chamber and continuously perfused with oxygenated ACSF. Fluorescent cells were visualized through a 40X water-immersion objective on an Olympus BX61WI upright microscope equipped with epifluorescence.

Whole-cell patch clamp recordings were made under both current and voltage clamp mode. Recording pipettes were made from borosilicate glass with a Flaming-

Brown puller (P-97, Sutter Instruments; tip resistance 5-8 MΩ). The pipette solution contained (in mM) 120 K-gluconate, 10 NaCl, 1 CaCl2, 10 EGTA, 10 HEPES, 5 Mg-

ATP, 0.5 Na-GTP, and 10 phosphocreatine. Electrophysiological recordings were controlled by an EPC-9 amplifier combined with Pulse Software (HEKA Electronic) and analyzed using Igor Pro. The signals were filtered at 2.9 kHz and acquired at 50 kHz.

Excitatory postsynaptic potentials (EPSPs) were further filtered offline at 20 kHz and excitatory postsynaptic currents (EPSCs) at 0.5 kHz. Junction potential (~9 mV) was corrected offline. Light stimulation was delivered through the same objective via pulses of blue laser (473 nm, FTEC2473-V65YF0, Blue Sky Research, Milpitas, USA) with varying lengths. Viral infection in the PCX was confirmed in brain slices during recordings. No sex differences between male and female mice were evident so data was pooled across mice.

Histology

Histological confirmation of PCX and OT implant sites for in vivo experiments was performed using a DAPI (4’,6-diamidino-2-phenylindole, Invitrogen, Carlsbad, CA) stain of 40 µm coronal brain sections (Fig 2-2). The number of cells transduced in the

PCX was quantified in 4-6 alternate 20 µm sections within a set region of interest in a subset of animals (300µm DV x 500µm ML/section; n=4) and compared to the number of anti-NeuN-labeled neurons to provide an estimate of AAV transduction using Nikon

NIS Elements. The spread of AAV infection was also estimated across all experimental animals using Nikon NIS Elements software (n=12 mice, 0.87mm ± 0.04mm anterior- posterior spread relative to bregma). For VGlut1 confirmation, the number of PCX cells coexpressing viral transduction and VGlut1 was examined in 4-6 alternate 40 µm sections in a subset of mice (n=3; Fig 2-4). For anti-NeuN and VGlut1 immunohistochemistry, free-floating sections were rinsed in tris-buffered saline and diluting buffer, and then blocked in 20% normal donkey serum for 30 minutes. Slices were then incubated for 24 hours at 4°C with the primary antibody anti-VGlut1 antibody

(anti-VGlut1: 1:500 in diluting buffer, Millipore, Burlington, MA; anti-NeuN: 1:1000 in diluting buffer, Abcam, Cambridge, UK). Sections were rinsed with diluting buffer and then incubated in with secondary antibody (anti-VGlut1: donkey anti-guinea pig IgG,

FITC conjugate, 1:375 in diluting buffer, Millipore; anti-NeuN: donkey anti-rabbit IgG,

AlexaFluor 488, 1:500 in diluting buffer, Abcam) for 2 hours at room temperature before being rinsed with tris-buffered saline and then dH20. Tissue was then mounted on slides with DAPI (for anti-VGlut1) or Vectashield mounting medium (for anti-NeuN,

VectorLabs, Burlingame, CA) and imaged. To assess association fiber density in the

OT, the fluorescent intensity of each layer was calculated as a function of background

intensity in NIH ImageJ (203) in 5-6 alternate 40 µm sections in a subset of mice (n=4).

Data from animals with either unrestricted or unsuccessful labeling were excluded from analysis.

To determine the distribution of RV+ cells in the anterior PCX for viral tracing experiments, 40 µm coronal sections were stored at -20ºC floating in 20% glycerine,

30% glycol in PBS. A subset of coronal sections across the anterior PCX (from

~2.34mm – 0.14mm bregma, ~ every 240 µm) were washed in PBS and then mounted onto Superfrost Plus slides with 90% glycerol in PBS and sealed with nail polish.

Images of these sections were captured with the Olympus VS120 virtual microscopy slide scanning system (Olympus, Shanghai, China) using a 10x objective. PCX layers, borders, and rostral-caudal axis were delineated based on a standard brain atlas (204).

The dorsal and ventral subregions of the anterior PCX were divided based on the dorsal edge of the lateral olfactory tract (107).

To confirm that OT neurons are innervated by glutamatergic neurons in the anterior PCX, we performed fluorescent in situ hybridization (FISH) using a VGlut1 probe on the rabies-labeled samples. Anterior PCX coronal sections (15 µm thick) were collected on Superfrost Plus slides and stored at -80ºC. Sections were blocked with 1% hydrogen peroxide in PBS at room temperature, rinsed with PBS, and incubated with

1µg/ml Proteinase K (Sigma, P-6556) at 37ºC. After a 0.2% glycine in PBS rinse, sections were then incubated with 0.25% acetic anhydride in 0.1 M triethanolamine, pH

8.0, and washed again with PBS. VGlut1 probes were pre-diluted to 0.2ng/µl with the hybridization buffer (50% formamide, 300µg/ml tRNA, 10 mM Tris pH 7.5, 10% dextran sulfate, 1x Denhalt’s solution, 600mM NaCl, 0.25% SDS), mixed well, preheated at

80ºC preheated, cooled, and applied to each section. After overnight incubation at 56ºC, the sections were washed, first with 2x SSC, and then 0.2x SSC twice at 65ºC. Once cooled to room temperature, sections were incubated with peroxidase-conjugated anti-

DIG antibody (1:500, Roche Applied Science) at 37ºC for 1 hour, washed with PBST, and lastly trated using TSA-plus Cy3 kit (Perkin Elmer). Since the GFP signal from RV labeled called was quenched during the ISH process, we further performed the immunohistochemical staining for GFP with rabbit anti-GFP (1:1000, Abcam) and treated the sections with Alexa 488-conjugated goat anti-rabbit secondary (1:1000,

Jackson ImmunoResearch), and DAPI and then mounted and sealed as described above. Images were captured as designated above. RV labeled neurons in different anterior PCX subregions and VGlut1 colabeled neurons in the anterior PCX were quantified semi-automatically using FIJI with the cell counter plugin.

Data Analysis

Behavioral and electrophysiological data analyses were performed using custom scripts (Cambridge Electronic Design, MATLAB). Neurons were sorted offline in Spike2 using template matching and principal component analysis, and spike times exported to

MATLAB for analysis. Any putative single unit with >2% of events occurring within a 2 msec inter-spike interval were excluded from analysis (185). Significance level for all statistical tests (paired t-tests, ANOVAs, one-sample z-tests, χ2 with Yates’ correction) was set at p<0.05, and all t-tests performed were paired unless otherwise noted.

Results

Viral Strategy for the Optogenetic Control of PCX Principal Neurons

In order to target PCX neurons and their association fibers innervating the OT for later optogenetic manipulation, we injected an AAV viral vector designed to express

ChR2 and a reporter fluorophore (mCherry) under control of the calcium/calmodulin- dependent protein kinase iiα (CaMKiiα) promoter, focally, in the anterior PCX

(AAV5.CaMKiiα.hChR2(E123T/T159C).p2A.mCherry.WPRE or

AAV5.CaMKiiα.hChR2(H134R).mCherry; Fig 2-3A). This approach led to infection of

PCX neurons (Fig 2-3B-C). Anti-NeuN immunohistochemistry was utilized to label neurons and to aid in quantification of viral infection efficiency. 59.1% of NeuN-labeled neurons within a set PCX region of interest were also mCherry+ (inter-animal range

51.5 – 66.2% ± 5.6% s.d.; Fig 2-3C, see Methods). From all colabeled neurons, the greatest numbers were found in layer ii (78.9%, mean = 101.1 neurons), followed by layers iii and i, respectively (iii: 16.9%, mean = 21.7 neurons; i: 4.2%, mean = 5.5 neurons; F(2,7)=206.47, p<0.0001; Fig 2-3D). Importantly, layer ii contains the densest collection of projection neurons (superficial pyramidal neurons and semilunar cells) extending fibers into the OT (12, 78, 87, 142, 146). While this AAV is, by design, specific for transducing excitatory neurons, we nevertheless confirmed that the neurons infected with AAV and thus expressing ChR2 are indeed glutamatergic by performing anti-VGlut1 (vesicular glutamate transporter 1) immunohistochemistry. Colabeling of mCherry and VGlut1 was observed in these cases which further confirms cellular specificity of the AAV infection (Fig 2-4).

Using this viral approach, we also observed mCherry-expressing PCX association fibers innervating the OT (Fig 2-3E). As anticipated based upon previous tracing studies (9, 37), these fibers were observed in all cell layers of the OT. To quantify the density of these association fibers, we extracted the fluorescence intensity across OT cell layers from 4 mice. Fluorescence intensity was greatest in layer ii,

followed by layers iii and i (F(2,9)=4.63, p=0.04) (Fig 2-3F). Thus, this AAV-based approach allows for the targeting of ChR2 into PCX glutamatergic neurons and their association fibers, including those innervating the OT. We utilized this same viral approach in two in vivo and one in vitro opto-physiological paradigms as described next.

Activation of PCX Neurons Enhances OT Activity

We next sought to address whether the PCX influences the activity of OT neurons in vivo. Three weeks after viral injection as described above, mice were implanted with an optical fiber in the PCX for light stimulation of ChR2-transduced neurons and an electrode array in the OT to record single unit neural activity (Fig 2-5A).

In the same surgery, mice were implanted with a head bar for subsequent head-fixation.

Following several days of recovery, the mice were water deprived and trained to perform a fixed-interval olfactory task (Fig 2-1) in which head-fixed mice received one of four odors for 2 seconds either with or without simultaneous PCX stimulation consisting of a blue light stimulus train. Mice also received pseudorandom trials of light alone.

Odors were presented during the inter-trial interval prior to a window of reward availability, where mice licked to receive a fluid reinforcer. In this design, we ensured the mice were indeed engaged throughout recording sessions by monitoring the occurrence of licking during successive reward windows (Table 2-1). This fixed-interval reinforcement paradigm, which provides reinforcement between all odors in a manner not closely linked in time with odor delivery, ensures all odors are of similar valence to the mice.

Through both post-mortem analyses and behavioral scoring, we focused our analyses on mice that met the following four criteria: 1) AAV expression verified to be restricted within the PCX, 2) electrode arrays confirmed within the OT, 3) fiber tips

localized within the PCX, and 4) criterion-level behavioral performance (Table 2-1).

Among these mice (n=8), 6 had clear multi-unit activity from which single units were sorted. The majority of OT neurons display low background firing rates (69, 185), providing the possibility that even subtle changes in firing may shape network function.

Indeed, the mean background firing rate of the OT units sampled in this experiment was

2.3 ± 1.2 Hz (inter-unit range: 0.0-5.5 Hz; n=58 units; Fig 2-6Ai). We asked whether, and if so how, PCX neuron stimulation may modulate this low OT neuron firing rate. To do this, for every unit we compared 2 seconds of averaged background activity to 2 seconds of averaged light-evoked activity and quantified the magnitude and direction of change upon light stimulation (11-21 trials/unit).

We found that light stimulation of ChR2-transduced CaMKiiα neurons within the

PCX elicits changes in the firing rates of OT units (Fig 2-5B-D). This modulation may occur in the form of brief phasic increases in firing, or those in which the firing is somewhat sustained throughout the light pulses (Fig 2-5Bi-Bii). 27.6% of OT units were significantly modulated by PCX stimulation (16/58 units; observed in 3 out of 6 mice; p<0.05, paired t-tests; Fig 2-5C). The impact of PCX activation was significant at the population level across all of these modulated units (t(15)=-4.79, p=0.00012; Fig 2-5D).

This effect was exclusively due to excitation among these OT units, with all units increasing their firing rate upon light stimulation (Fig 2-5C-D). Some units showed dramatic increases in firing compared to their low background firing rates (e.g., from

~2Hz to near 15Hz; Fig 2-5D). Importantly, the fact that we observed ChR2-dependent modulation of OT units in only 50% of the mice with detectable single units highlights that this effect is not due to non-specific influences of light stimulation on the neurons

(artifact) or the animals (e.g., arousal). Further supporting this, light-evoked responses were not observed in two separate mice injected with AAV5.CaMKiiα.mCherry (n=28 units; p≥0.195, paired t-tests; comparing background and light stimulation periods for each unit; Fig 2-6E). Thus, activation of glutamatergic PCX neurons enhances OT neural activity.

Activation of PCX Association Fibers Within the OT Modulates OT Activity

Is the PCX capable of influencing OT activity directly through its association fiber system? To test this, we adapted a new in vivo preparation wherein we used the same viral approach described above along with an optetrode to directly stimulate PCX association fibers specifically within the OT while simultaneously recording OT unit activity (Fig 2-7A). This approach allowed us to determine the direct contributions of the

PCX upon the OT, while minimizing potential bi- or multi-synaptic influences from other interconnected brain structures which may serve to relay activity from the PCX into the

OT. Through both post-mortem analyses and behavioral scoring, we focused our analyses on 8 mice which met the following three criteria: 1) AAV expression verified to be restricted within the PCX, 2) optetrode arrays confirmed within the OT, and 3) criterion-level behavioral performance (Table 2-1). Among these mice, we extracted 203 single units. Corroborating the previous preparation, the spontaneous firing rates of these OT units were also low, at 2.1 Hz ± 3.2 Hz (inter-unit range: 0-26.6 Hz; Fig 2-

6Aii). Supporting our hypothesis that PCX may influence the OT monosynaptically through its association fiber network, we also observed significant light-evoked responses in this preparation. Overall, the population-level impact of this modulation was less than that we observed upon local PCX activation (i.e. Fig 2-5D; χ2(1)=13.97, p=0.002, comparing modulated units in both experiments). Nevertheless, light-evoked

responses were observed in 5 out of the 8 mice tested (Fig 2-6B-D). Among these mice,

5.1% of all units were significantly modulated (8/156 units; p<0.05, paired t-test; Fig 2-

6B-D). Two of these units displayed significant excitation upon association fiber stimulation (p<0.05, paired t-tests). In 6 units, association fiber stimulation resulted in suppressed background firing rates (p<0.05, paired t-tests) which was significant at the population-level (t(5)=2.96, p=0.032; Fig 2-6C). Further, evoked activity in some units was seemingly well-coupled with each light pulse within the stimulus train (Fig 2-6Dii).

Thus, activation of PCX fibers within the OT itself is capable of modulating the activity of some OT neurons.

Activation of PCX Association Fibers Bidirectionally Modulates the Representation of Odors in the OT

Does the PCX input influence the representation of odors in the OT? As described earlier, mice were presented with pseudorandom trials of four different odors, trials of light alone, or trials of light simultaneously with one of the four odors (odor+light)

(Fig 2-1). We quantified the number of neurons responsive to each of the odors presented with or without activation of PCX association fibers via the optetrode. Across the population of 203 single units, 47.8% (n=97; p<0.05 in t-tests from background to odor-evoked activity) were significantly modulated by at least one of the odors presented. From the activity of these 203 single units, we extracted 812 cell-odor pairs to examine the proportion of units responding to odor compared with simultaneous odor and light. As expected, some cell-odor pairs responded to odors in the form of odor- evoked suppression, whereas others responded by excitation of their firing rates (Fig 2-

7B). Not surprisingly, many OT neurons were not modulated by odor (43). Upon association fiber activation, the proportion of cell-odor pairs significantly excited by odor

decreased 3.7% (82 to 52 cell-odor pairs; χ2(1)=5.77, p=0.016), while the proportion of units significantly suppressed by odor increased 4.7% (85 to 123 cell-odor pairs;

χ2(1)=5.80, p=0.016). This in some instances was due to light-modulation of units not previously modulated by odor in the absence of light. Thus, activating PCX association fibers within the OT changes the proportion of units representing odors.

Does activation of PCX fibers equally modulate odor-evoked OT neuron firing rates across all different odors? To determine this, we examined the number of significantly odor-responsive units for each of the four odors presented either in the presence or absence of PCX stimulation. The number of units responding to two of the four odors presented with concurrent PCX activation were significantly modulated when compared to odor-only activity. There was an increase in the number of units responding both to heptanal (47 to 64) and 1,7 octadiene (27 to 37; Fig 2-8). This shift occurred primarily in suppressed units, increasing their number from 19 to 43 units from heptanal and 20 to 25 for 1,7 octadiene. The number of odor-responsive units to ethyl butyrate remained unchanged (both 32), but there was again an increase in the number of inhibited units (from 26 to 28). Lastly, isopentyl acetate responses decreased from 62 to 40 significant units, but this was largely due to a loss of excitatory units; inhibitory units increased by 5 (from 22 to 27). Thus, at the single unit level, light excitation of

PCX association axons appears to increase the number of odor-responsive units, and more specifically, recruits inhibited units. At a population level, PCX association fiber activation increases the number of units that are odor-suppressed, with no bias within odors.

Our results demonstrating that PCX association fiber activation alters the proportion of odor-modulated OT units next led us to ask if this activation changes a neuron’s odor-evoked firing rate based upon the way that the neuron responds to odors intrinsically (viz., in the absence of light). To address this question, we quantified the differences in odor responses across individual OT cell-odor pairs. Of the 812 cell-odor pairs, 19.6% (n=159) were significantly responsive to at least one of the four odors presented when compared to each cell-odor pair’s average background firing rate

(p<0.05, within cell-odor pair t-tests). 52.2% (n=83; Fig 2-9Ai-iii) of these modulated cell- odor pairs were excited by odor presentation, whereas 47.8% displayed odor-evoked suppression of firing (n=76; Fig 2-9Bi-iii). To determine the direction of change within each population, we compared each cell-odor pair’s odor-evoked response to that elicited by simultaneous odor and light. Dependent upon whether the cell-odor pair was odor-excited or odor-suppressed, OT units displayed a bidirectional response to simultaneous PCX fiber activation. Those units excited by odors were, as a population, less-excited by the same odors in the context of PCX fiber activation (t(82)= 5.13, p<0.0001, Fig 2-9Aii). This effect was most prominent in cell-odor pairs that encoded odors (in absence of light) with low firing rates (Fig 2-9Aiii). Conversely, odor- suppressed units displayed significant excitation during association fiber activation

(t(75)= -5.74, p<0.0001, Fig 2-9Bii). The change in the distribution of cell-odor pairs encoding odor significantly shifted, suggesting that cell-odor pairs with lower odor- evoked firing rates were most greatly modulated by the occurrence of odor and light together (D=0.50, p=0.023, Komolgorov-Smirnoff test; Fig 2-9Biii). Importantly, a similar bidirectional modulation of OT odor coding was also observed in our previous

preparation wherein we stimulated the PCX directly (Fig 2-10). Together, these results demonstrate that PCX association fiber activation biases OT units to encode odors in an opposing manner than they do intrinsically. This indicates that the PCX exerts gain control over OT output.

PCX Principal Neurons Synapse with, and Evoke Monosynaptic Responses Within, OT D1- and D2-Type MSNs2

The predominant cell type in the OT are MSNs which can be divided into those expressing either the D1 or D2 receptor. Given the uniquely important signaling pathways and downstream connectivity of these different neuron types (164, 165), we next addressed whether PCX input is preferentially directed towards D1 or D2 MSNs.

To answer this question, we used an in vitro paradigm in combination with the same viral approach we employed in vivo – however, the injections were performed in D1- tdTomato/D2-EGFP double-transgenic mice to allow for genetically-guided identification of D1 or D2 neurons, respectively (Fig 2-11A). Three to six weeks after AAV injection, brain slices were collected and whole-cell patch clamp recordings performed on OT D1 and D2 MSNs (Fig 2-12A). Consistent with previous electrophysiological characterizations of striatal MSNs (205), D1 and D2 MSNs had low input resistances

(D1 vs D2: 161.7 ± 7.6 vs. 159.1 ± 9.3 MΩ, n = 20 cells in each group; t(38)=0.21, p=0.83). However, D1 MSNs were less excitable compared to D2 MSNs as evidenced by their display of higher firing thresholds and lower firing frequencies upon current injection (206) (Fig 2-11B-C).

As expected, ChR2-expressing principal neurons in the PCX showed instant (<1 msec latency) and largely repeatable responses to blue light stimulation (Fig 2-11D-E).

2Experiments conducted in collaboration with the Ma Lab at University of Pennsylvania

None of the 106 OT neurons recorded from (55 D1 and 51 D2 MSNs) showed instant light-evoked responses, indicating that they were not directly infected by the virus. In a fraction of D1 MSNs in the OT (Fig 2-12B), light pulses evoked excitatory postsynaptic potentials (EPSPs) under current clamp mode. The evoked responses increased with the pulse duration and typically were able to follow 10 Hz stimulations. To differentiate

OT neurons that receive mono- or polysynaptic connections, we recorded light-evoked excitatory postsynaptic currents (EPSCs) under voltage clamp mode and measured the response latency during repeated stimulations (Fig 2-12C). An OT neuron was considered to receive monosynaptic inputs from the PCX if the response latency was <6 msec with the latency jitter <1 msec. Among these, 24% of D1 and 16% of D2 neurons displayed measurable EPSCs upon PCX fiber activation within the OT with 11% of D1 and 14% of D2 neurons receiving monosynaptic excitation (Fig 2-12D). The proportions of D1 and D2 neurons displaying monosynaptic excitation were similar (χ2(1)=0.009, p=0.923), reflecting that PCX inputs to the OT are not biased towards one cell population. Thus, PCX association fiber activation evokes responses, in some cases monosynaptically, in both D1 and D2 OT MSNs. These results reveal that the PCX can directly influence these two dominant and important striatal neuron populations.

Topographical Organization of PCX Neurons Innervating OT D1- and D2- Type MSNs3

Different areas within the anterior PCX may perform unique functions and possess diverse molecular and circuit features (107, 143, 207–212). This led us to ask where projections onto OT D1 and D2 MSNs originate from within the anterior PCX. Do

OT neurons receive input from a spatially-distributed population of PCX neurons? Or,

3Experiments conducted in collaboration with the Xu Lab at the Chinese Academy of Sciences

are the PCX neurons innervating the OT spatially-organized, which would suggest a unique role for PCX cells in that region in the cortico-striatal modulation we uncovered in the previous experiments? To determine the pattern of innervation onto these neuron types, we injected a helper viral vector mixture (AAV9-EF1a-DIO-histone-BFP-TVA and

AAV9-EF1a-DIO-RV-G) into the OT of D1-Cre and D2-Cre mice. Two weeks post- infection, a rabies virus (RV-EnvA-ΔG-GFP; (213)) was injected in the same location.

Tissue was then collected one week following for post-mortem analyses. Using this model, we subsequently determined the number of rabies virus-labeled (RV+) neurons across: 1) PCX layers, 2) the dorsal/ventral axis, and 3) the rostral/caudal axis. First though, we sought to confirm the glutamatergic identity of the cells innervating the OT, as suggested by the results in Fig 2-4. While the PCX association fiber network originates from glutamatergic principal neurons (12, 78, 146), the PCX is comprised of a heterogeneous population of both glutamatergic and GABAergic neurons. Therefore, we performed VGlut1 fluorescent in situ hybridization on a subset of D1- and D2-Cre mouse tissue following RV infection to confirm that indeed the PCX neurons synapsing upon OT MSNs are glutamatergic (Fig 2-13).

This RV approach successfully infected neurons in the anterior PCX following injection into either D1- or D2-Cre mice, demonstrating that PCX neurons send monosynaptic input to both of these cell populations (Fig 2-14A-B), as also supported by our cell-type specific patch clamp recordings (Fig 2-12). We first quantified the difference in innervation patterns of both cell types across PCX layers. Consistent with our in vivo quantification (Fig 2-3D), the majority of neurons innervating both D1 or D2 neurons in the OT originate from within PCX layer ii (D1: F(2,9)=126.0, p<0.0001; D2:

F(2,9)=447.9, p<0.0001 comparing across layers; Fig 2-14C). Next, to determine if different PCX subregions uniquely synapse upon these MSN populations, we quantified the number of RV+ cells in each subregion of PCX across D1-Cre and D2-Cre mice.

This revealed that more neurons in the ventral versus dorsal PCX synapsed upon these neurons (D1: t(3)=8.19, p=0.004; D2: t(3)=8.10, p=0.004; Fig 2-14D). Examining the rostral-caudal distribution of PCX innervation of OT we found that more neurons from the caudal versus rostral PCX synapsed upon both D1 and D2 neurons (D1: t(6)=-

12.57, p<0.0001; D2: t(6)=-1.94, p=0.01; Fig 2-14E). Thus, a ventral-caudal gradient of

PCX neurons innervates, to a similar extent, both OT D1 and D2 MSNs and, therefore,

PCX input into the OT is topographically organized.

Discussion

Extensive anatomical work over the last century has examined the vast and diffuse interconnectivity within the olfactory system (e.g., (12, 34, 47, 78, 167, 214, 215).

While it is assumed that the way in which the olfactory system disseminates odor information across its expansive hierarchical network is critical for olfactory information processing and ultimately odor perception, the function of this inter-regional connectivity is just beginning to be resolved (61, 75, 149, 216, 217). In the present study we contribute to this overarching goal by defining the function and the cells involved in one specific and unique pathway: the cortico-striatal pathway from the ‘primary’ olfactory cortex into the OT. We demonstrated that PCX input onto OT neurons bidirectionally modulates OT neuron activity dependent upon a given neuron’s intrinsic responsivity to odors. PCX fiber activation decreases the firing of neurons excited by odors and increases the firing of those suppressed by odors. Thus, the PCX biases OT odor- evoked activity by transforming the population of OT neurons that are odor responsive.

This shift of responsivity is likely due to convergent effects of PCX inputs as we show here, and the simultaneous integration of olfactory bulb input (218). This is further made possible given the fact that striatal MSNs extend collaterals upon one another, which while shown to modulate neighboring cells (159), are not understood in the context of in vivo physiology nor in the context of behavior.

Regarding perceptual and behavioral outcomes of this bidirectional modulation, what would be the significance of exciting neurons which are otherwise suppressed by odor? Similarly, what would be the function of suppressing neurons which are otherwise excited by odor? While the outcome of this upon odor-guided behavior is unknown, and may be complex, we do know that the OT receives monosynaptic odor input from the olfactory bulb in addition to this PCX input we describe herein. Given this, the OT likely does not ‘need’ this PCX input to receive essential odor information from the olfactory bulb and inform a fundamentally basic odor-guided behavior (‘do I recognize an odor in my environment?’). Instead, it is possible the OT capitalizes upon the extra richness of odor input from the PCX, including that influenced by perceptual learning and experience (122, 134, 136, 219). In this model, the OT is now afforded this ‘refined’ odor input due to the bidirectional change upon PCX fiber activation. We predict this refined odor input facilitates odor-guided behaviors following learning, for instance, in the case of odors with a known reinforced outcome, or valence.

We wish to point out that the amount of OT neuron modulation we observed by

PCX activation is likely an underestimation in both our in vivo and in vitro results. This is especially the case in our paradigms wherein we directly stimulated PCX fibers within the OT itself. First, this may be due to less-than-ideal viral infection efficiency. While a

fair number of PCX neurons were infected (Fig 2-3), this number is by far the complete

PCX glutamatergic cell population. Second, this could be due to the fact that the AAV injections were not targeted specifically to the ventro-caudal portion of the PCX (which we discovered strongly innervates the OT; Fig 2-13). Further, it is likely we severed connections between these structures during slice preparation or surgical implantation of electrodes and optical fibers. Additionally, it is necessary to consider the direct versus indirect effects of the light stimulation paradigms we used. Stimulation of the PCX locally excites numerous glutamatergic neurons, which extend projections both within the PCX and into additional structures via the association fiber network (12, 78, 146).

Thus, changes in OT neuron firing upon local PCX stimulation could be due to a combination of direct PCX inputs and polysynaptic inputs from intermediate structures between the PCX and OT, such as the anterior olfactory nucleus, entorhinal cortex, or amygdala (12, 78). Perhaps this is why our direct activation of PCX fibers within the OT yielded a lower impact upon OT firing rates when compared to local PCX activation.

This would similarly agree with the smaller population-level monosynaptic light-evoked responses in our in vitro preparation. Interestingly, nevertheless, there were significant light-evoked alterations of neuron firing during odor presentation, suggesting that it is a synergistic combination of both odor input and PCX association fiber activation that leads to optimal modulation of OT neuron firing.

Our viral tracing revealed a topographical organization of PCX innervation onto

OT D1 and D2 neurons. This is reminiscent of PCX’s topographically organized output upon the orbitofrontal cortex, which also follows along the major axes of the PCX (144).

Given the distributed and spatially-overlapping representation of odors in the PCX (65,

66, 102) and other olfactory cortices (e.g., (62)), what are the functional implications of this spatially-organized output? The anterior PCX has historically been subdivided into ventral and dorsal regions, with each having differences in access to, and in their responsivity to, sensory input, as well as differences in their expression of molecular markers, and density of cell layers (107, 143, 207–212). Further, there is an established rostral-caudal gradient in the magnitude of inhibition onto PCX pyramidal neurons, with inhibition being greater in the caudal PCX (208, 212). This effect is largely mediated by layer iii inhibitory neurons, and this inhibitory activity pattern is dependent upon sensory experience (212). As there is increasing inhibition caudally, and we demonstrated here that more neurons project into the OT from this caudal region, it is possible that these neurons which innervate the OT are subject to differential types of inhibition dependent upon the quality of incoming odor information. Thus, our work adds to other forms of known anatomical and circuit-level heterogeneity within the PCX and therefore our results inform models for how olfactory information is relayed out of the PCX. Additional

RV injections into other olfactory structures, and perhaps even into subzones within the

OT (the present injections were largely targeted into the antero-medial and antero- lateral OT zones), will be important in defining additional spatial zones within the PCX. It is likely that while the PCX itself represents odors in a distributed, spatially-overlapping manner (65, 66, 102), the topographical organization of PCX efferents as described herein and in (144) allows for the PCX to exert influences which are unique to both the efferent structure, and to odor-evoked activation within PCX spatial zones.

Neurons in the direct pathway, specifically striatonigral MSNs expressing D1 receptors, ultimately disinhibit the thalamus and facilitate motivated behaviors (163).

Neurons of the indirect pathway inhibit the thalamus and diminish motivated behaviors

(220), and are predominantly striatopallidal MSNs expressing the D2 receptor (163).

While we presently do not know specific roles for these cell types within the OT, nor if the strict definition of the direct or indirect pathways applies to these OT cells (165, 221,

222), the OT’s position within the ventral striatum (a component of the basal ganglia) together with our finding that the PCX functionally innervates the OT, adds weight to the possibility that the OT serves as a hub for odor information to enter the basal ganglia.

This is further likely since additional olfactory structures also innervate the OT (153).

What do our results suggest may subserve this ‘hub’ role for the OT? We propose that our finding that both OT D1 and D2 MSNs equally receive PCX input positions them to serve central roles in this hub – affording the OT to exert and extend odor information into downstream structures which comprise both the direct and direct pathways. By this,

PCX input into the OT, or perhaps that arising from another olfactory structure upon the

OT, can functionally be distributed into basal ganglia. This routing of odor information into the direct and indirect pathways likely has major implications for odor approach or aversion behaviors. OT D1 MSNs are considered important in mediating the encoding of both appetitive and aversive learned odors with greater numbers than D2 MSNs

(188). Additionally, dopaminergic input onto OT D1 and D2 MSNs potentiates only D1

MSN responses to olfactory bulb input, suggesting an initial specification of odor information into the OT (223). It is likely that throughout learning, or in states of enhanced motivation, the functional consequence of the observed PCX input may be magnitudes different than that which we report herein. For instance, phasic dopamine release upon OT MSNs may facilitate synaptic input of PCX association fibers and

thereby allow for strengthening of PCX modulation. In these cases, the routing of odor information from the OT will be enhanced and thereby so would the robustness of a possible behavioral response (e.g., odor approach). Our results revealing a source of cortical input to this olfactory striatal structure’s MSN population highlights the OT’s role as a possible hub into basal ganglia. We predict that neuromodulators play an integral role in distributing odor information from the OT into other basal ganglia regions.

In summary, we uncovered a functional role for PCX association fibers in exerting modulatory influences upon the OT. The input upon OT MSN populations by the PCX highlights at least one mechanism whereby the PCX may shape the availability of odor information within an interconnected brain structure. Our finding that this occurred upon both D1 and D2 neuron populations raises the exciting and likely possibility that this pathway informs the display of odor-directed motivated behaviors. As such, one might postulate that the OT, serving as the olfactory striatum and receiving input from numerous olfactory centers (153), participates in concert with other olfactory cortices, beyond the PCX itself, to distribute odor information into critical basal ganglia centers which have direct roles in regulating motor behaviors including stimulus approach and even consummation.

Contributions

I would like to thank all of my coauthors on this work: Yun-Feng Zhang, Zhijian

Zhang, Janardhan P. Bhattarai, Andrew H. Moberly, Estelle in ‘t Zandt, Huijie Mi,

Xianglian Jia, Marc V. Fuccillo, Fuqiang Xu, Minghong Ma, Daniel W. Wesson. This work was supported by NIH NIDCD R01DC014443 and R01DC016519 to D.W.,

R01DC006213 to M.M., F31DC016202 to K.W, and National Natural Science

Foundation of China grant 31771156 to F.X. We thank Marie Gadziola for help with in vivo data handling and Chris Ford for advice on medium spiny neuron recordings. We thank Zhonghua Lu and Liping Wang from the Shenzhen Institutes of Advanced

Technology for providing the VGlut1 probe and platforms for ISH, Yanqiu Li from the

Wuhan Institute of Physics and Mathematics (WIPM) for providing the D1R-Cre/D2R-

Cre mice used in viral tracing, and Lingling Xu from WIPM for managing the microscopy platform.

Table 2-1. Number of mice used for in vivo experiments. PCX neuron PCX association fiber stimulation stimulation Number of mice attempted 17 23 Number of mice with successful AAV 8 9 transduction restricted to the PCX Number of mice with successful 10(7) 22(8) fiber/electrode/optetrode implanta Number of mice shaped to criterion 13(7) 23(8) performance in olfactory taska Average % correct trials across an 97.8% ± 1.2% 93.7% ± 1.1% olfactory task sessionb (93.0% - 100.0%) (85.7%-100.0%) c Number of mice with verified single 6 8 units a() = Number of mice with successful AAV transduction. b() = Range of correct responses across sessions. c Values were averaged across sessions throughout which the optetrode was lowered (2-3 sessions/mouse).

Figure 2-1. Olfactory fixed-interval task. Task structure across three trials of stimulus presentation. Following an inter-trial interval, stimulus presentation begins (2 seconds of: odor, 10 Hz blue intra-cranial light, or odor+light together). 8 seconds following stimulus offset, water-deprived mice were allowed to lick to receive a small fluid reward.

Figure 2-2. Histological confirmation of electrode/fiber optic implantation sites for in vivo recordings. A) Location of implant sites for either i) separate optical fiber into PCX and fixed electrode array in OT preparation (n=6 mice), or ii) optetrode array into the OT preparation (n=8 mice).

Figure 2-3. Viral strategy for the optogenetic control of PCX principal neurons. A) Schematic of the AAV injection procedure into the anterior PCX that was used to express ChR2 in PCX neurons and association fibers. LOT = lateral olfactory tract, PCX = piriform cortex, OT = olfactory tubercle. Following injection of AAV into the PCX (top) 2-3 weeks of time was allowed for viral transduction, following which mice were used for physiological experiments (bottom). B) Representative image of a PCX AAV injection (AAV5.CaMKiiα.hChR2(H134R).mCherry) restricted into specifically the anterior PCX (white dotted outline). Scale bar = 500 µm. C) Representative image of PCX neurons (anti-NeuN, marker for neuronal nuclei) transduced with the viral vector AAV5.CaMKiiα.hChR2.E123T.T159C.p2A.mCherry. WPRE across PCX layers. Scale bar = 100 µm. Anti-NeuN histochemistry performed on tissue from mice not used for recordings. D) Quantification of AAV transduction in the PCX across animals (n=4) as a percentage of mCherry+ cells compared to NeuN+ cells in each PCX cell layer. ***p<0.001. E) Image of mCherry+ association fibers within the OT which originated from the PCX. Scale bar = 100µm. F) The average fluorescence intensity of PCX association fibers in each OT layer across animals (n=4) as a function of background fluorescence.

Figure 2-4. VGlut1+ co-labeling with AAV-mCherry PCX neurons. Representative image of colabeling of mCherry+ (red) and VGlut1+ (green) neurons in the PCX. All scale bars = 100µm.

Figure 2-5. Modulation of OT activity by PCX neuron activation. A) Schematic of the location of the optical fiber in PCX for light stimulation and the fixed electrode array in the OT for extracellular recordings. B) Examples of significant light-evoked activity of two OT single units by stimulation of the PCX. Each raster plot and histogram represent a single unit’s response to light stimulation across 10 trials. Average waveforms for each unit are represented within the histogram. Responses to PCX light stimulation evokes unique temporal excitation, with activity either transient (i) or sustained (ii). C) Distribution of the average light-evoked response across a population of recorded units. A majority displayed excitation to PCX light stimulation (yellow), while some were suppressed (purple) and a small number remained unchanged (gray). Arrows indicate bars where units contributed to the plot in D. Double-line break in x axis denotes change in magnitude binning (from 0.2 to 2) to represent sizable changes in magnitude upon light stimulation. D) Distribution of the change from average background activity to light-evoked activity. Of 16 significantly light- responsive units, all units increased their firing rates, demonstrated in the line graph. As a population, these units were significantly excited from background activity. The mean of this data is denoted by the bold black dashed line in D.

Figure 2-6. Average background firing rate and light-evoked activity during PCX association fiber stimulation. A) Distribution of average background firing rate across both preparations (i) PCX stimulation, ii) PCX association fiber stimulation). Average background firing rate was 2.3 Hz for first preparation and 2.1 Hz for the second. One unit from the first preparation was excluded as its average spontaneous firing rate exceeded three times the standard deviation of the population plus average background (5.9 Hz standard deviation above background, 71.1 Hz outlier unit firing rate). B) The distribution of average evoked response across population of recorded units. Of 203 single units recorded, the majority displayed suppression to activation of PCX association fibers (purple), while some were excited (yellow) and a small number remained unchanged (gray). Arrows indicate bars containing units that contributed to the plot in C. C) Distribution of the change from average background to light-evoked activity. Of 8 significantly light-responsive units, most units decreased their firing rates, demonstrated in the line graph. The mean of this data is demonstrated by the boldened black dashed line. D) Example raster plot and histogram of a significantly light-excited unit (i) v. a non-significant but visually-apparent light-excited unit (ii). Averaged waveform for each unit inset within histogram. E) Average firing rates for background and light stimulation time periods across units (PCX cells without ChR2 expression – AAV5.CaMKiiα.mCherry). Boldened black dashed line = population mean. There is no significant effect of light stimulation on firing rates among individual units (p≥0.195; paired t-tests), or across all units as a population (p=0.28).

Figure 2-7. PCX association fiber activation biases OT odor representation. A) Schematic depicting the location of the optetrode assembly in the OT to record OT extracellular activity and light stimulate PCX association fiber terminals within the OT. B) Population change of odor-responsive OT neurons during PCX light stimulation. Pie charts representing the percent of neurons within the population that were odor-excited (yellow), odor- suppressed (purple), or remain unchanged (gray) with (right) or without (left) PCX light stimulation. *p<0.05.

50 1,7 Ethyl Isopentyl Heptanal Octadiene Butyrate Acetate

40 * * 30

20 Cell Count 10

ExcitedInhibited ExcitedInhibited ExcitedInhibited ExcitedInhibited

Figure 2-8. PCX association fiber activation increases odor-evoked suppression across odors. Population change comparing odor-responsive OT neurons with (blue bars) and without (gray bars) PCX association fiber activation. *p<0.05.

Figure 2-9. The PCX bidirectionally modulates OT odor-responsive neurons. A-B) Modulation of cell-odor pair response from odor alone to odor+light stimulation together. A represents cell-odor pairs significantly excited by odor, while B represents cell-odor pairs significantly suppressed by odor. Panels i-iii use the same conventions for both A and B. i) 2D histograms representing the change in firing rate from 1 second of background activity to 2 seconds of either odor presentation (left) or odor+light presentation (right). Each row is one unique cell-odor pair and is the same for both left and right 2D histograms. Each column is a 100 ms average firing rate bin for that cell-odor pair. End 2D histogram shows the average evoked response (either odor or odor+light) for each cell-odor pair. ii) Boxplot displaying the change in average firing rate across the population for odor alone (gray box) v. odor+light together (blue box). Both plots are significant using a paired t- test comparing odor v. odor+light. p<0.001. Red line designates the mean, the box designates the 50th percentile range, and the whiskers designate the 25th and 75th percentiles. Red dots are outliers. iii) Shows distribution of change from odor alone (gray line) to odor+light together (blue line) from mean firing rate evoked in each case. Pie chart above shows the percent change from odor alone.

Figure 2-10. PCX stimulation leads to bidirectional modulation of odor-responsive OT units. A) Represents the change in cell-odor pair response from odor alone to odor+light together. i) The 2D histogram represents the change in firing rate from 1 second of background to 2 seconds of either odor presentation (left) or odor+light presentation (right). 2D histogram is divided by cell-odor pairs significantly excited (50% excited, 10/20 units, top) or suppressed by odor (50% suppressed,10/20 units, bottom). Each row is one unique cell- odor pair and is the same for both left and right 2D histograms. Each column is a 100 ms average firing rate bin for the cell-odor pair. ii) Boxplot displaying the change in average firing rate across the population for odor alone v. odor+light together. The change in excited population is significant using a paired t-test to compare odor v. odor+light. **t: 3.25(9), p<0.01. iii) Line graph showing the distribution of change from odor alone (gray bars) to odor+light together (blue bars) from mean firing rate evoked for each.

Figure 2-11. Whole cell patch clamp recordings of PCX neurons and OT D1 and D2 MSNs reveal unique electrophysiological properties. A) D1 (red) and D2 (green) MSNs in OT slices visualized under fluorescent illumination. Scale bar = 10 µm. B) Firing patterns of D1 and D2 MSNs in response to current injections (20 pA steps). C) The firing frequencies of D1 and D2 MSNs are plotted against the injected currents. Note that D1 MSNs displayed significantly higher thresholds and fired fewer action potentials than D2 MSNs. **** two-way ANOVA (cell type and current step as two factors), F=31.76(38), p<0.001 for cell type. D) Schematic representing light stimulation of PCX neurons during patch clamp recordings of ChR2 infected PCX neurons. E) An infected pyramidal neuron (red) in the PCX displayed robust responses to blue light stimulation (1ms pulses indicated by blue bars).

Figure 2-12. Both OT D1 and D2 MSNs receive monosynaptic and polysynaptic input from the PCX. A) Schematic of the whole-cell patch clamp recording of OT D1 or D2 MSNs and stimulation of PCX association fibers. B) A D1 MSN in the OT showed light induced EPSPs to different pulse lengths (upper panel) and to a train of 10 Hz stimuli (lower panel) under current clamp mode. C) Light-evoked mono- or poly-synaptic EPSCs in D1 and D2 MSNs under voltage clamp mode. The holding potential was -60 mV. D) Summary of D1 and D2 cells which displayed light-evoked synaptic responses, organized by whether the evoked response was defined as mono- or polysynaptic (See Methods). Values in () = ns.

Figure 2-13. Glutamatergic PCX neurons directly innervate both OT D1 and D2 neurons. A-B) Neurons from D1-Cre (A) and D2-Cre (B) mice. i) Representative images showing the colocalization of VGlut1+ neurons (red) and RV+ neurons (green) in the anterior PCX. Arrows designate colabeled neurons. DAPI (blue) was used as a pan-cellular marker for visualization. Scale bar = 20 µm. ii) Pie charts quantifying the percentage of neurons that are both VGlut1+ and RV+ (green) or RV+ only (gray). The vast majority of RV+ cells were also VGlut1+ in both D1-Cre (z=11.8, p<0.0001; one-sample z-test for proportions) and D2-Cre (D2: z=55.6, p<0.0001; one-sample z-test for proportions) mice.

Figure 2-14. Innervation of OT MSNs by the PCX is topographically organized. A-B) Representative coronal brain sections displaying anterior PCX neurons labeled by the rabies virus (RV) (RV-EnvA-ΔG-GFP – green) from the OT in both D1R-Cre (A) and D2R-Cre (B). Scale bar = 200 µm. C) The percentage of RV+ neurons across all PCX layers. Each dot represents the average for one animal. *p<0.05, **p<0.01 for all. D) Percentage of RV+ neurons localized within the ventral and dorsal PCX subregions. E) Percentage of RV+ neurons localized within the rostral and caudal PCX subregions.

CHAPTER 3 SUMMARY AND CONCLUSIONS The goal of my dissertation research was to understand the role of the connectivity between two olfactory cortical structures: that between the PCX and OT.

My demonstration that the PCX influences the way by which the OT processes and encodes odor information, modulating both D1 and D2 MSNs, fills a void in the field for understanding the way by which cortical structures interact to modify incoming olfactory information. The dense interconnectivity of the olfactory system, due in large part to the expansive association fiber network stemming from PCX principal neurons, is largely unstudied. As evidence accumulates for the role of these connections between structures, it is necessary to consider both the way in which this connectivity modulates sensory perception, and in what way this processing scheme more broadly informs computations across higher-order structures.

Olfactory information is anything but simple. We can conceptualize the burden of processing odor inputs by examining our own experiences. For example, if we caught the scent of our grandmother’s cookies, we would immediately know that: 1) the identity of the smell is cookies (a combination of hundreds of odor ligands); 2) the intensity of the smell, and the smell as being new and different from previous background scents to which we have habituated; 3) the location of the smell, insofar as we are willing to follow it; 4) the emotional connection or hedonic value associated, and; 5) the memory of the smell, likely related to pleasant experiences that have complex and intricate previous associations. These are all deeply complicated problems for the olfactory system to resolve on a rapid timescale. Thus, considering this complexity, it is intuitive that the

olfactory system might require a complicated coding scheme spanning across numerous cortical structures to process this rich information.

Future Directions for Dissecting Secondary Olfactory System Connectivity

Here, I established that the PCX is capable of controlling which OT neurons are responsive to odor information, and the way in which they respond, equally affecting both D1 and D2 MSNs, the principal neurons of the OT. Taken broadly, this implies that one cortical structure has the capacity to control the neural output of another, yielding insight into the way in which this sensory information is both being received and then processed within downstream structures responsible for decision making and behavioral outputs. This cortical influence of an interconnected structure is not occurring in isolation; other interregional computations are happening in parallel across the brain.

The PCX itself is broadly influencing numerous other structures on a similar timescale as it is with the OT. Beyond this, other secondary olfactory structures are influencing the representation of information within one another. Within such a complicated coding scheme, how can one rectify what the role of every connection across the olfactory system might be?

This will be a massive undertaking necessitating a complicated solution. Future work should focus on systematically identifying connections between secondary olfactory structures. For the purposes of planning future experiments stemming from this dissertation research, this topic will be constrained to experiments using in vivo animal preparations that might specifically examine the function of these circuits. To elucidate the role of connectivity between structures, future research should focus on the following questions: 1) identification of cell types involved in this pathway, both upstream and downstream; 2) changes elicited by both activating and silencing this

pathway on neural activity, and; 3) determination of the perceptual and behavioral outcomes to this pathway perturbation. These research directions will determine important components of higher-order olfactory pathways, and the role of these pathways in sensory processing.

Here, I determined the origin of PCX projections into the OT as arising from ventro-caudal anterior PCX principal neurons, and also demonstrated that these neurons terminate equally onto D1 and D2 MSNs within the OT. Further, I elucidated the way in which activation of PCX principal neurons or their association fibers modulates OT neuron activity with and without concurrent olfactory input. For this section, I will describe potential future directions for this research topic, along with any caveats and limitations. I will focus on further research necessary to fully elucidate the role of the projection from the PCX into the OT, which can be extrapolated to other interconnected secondary structures within the olfactory system (and potentially to other inter-cortical connections).

Cellular Origins and Terminations of Secondary Structure Connections

PCX association fiber input into secondary structures can potentially arise from either principal cell type in the PCX: semilunar cells, or pyramidal neurons. Here, I was unable to elucidate specifically which cell type might extend association fibers into the

OT, as both principal neuron types express CaMKiiα, the promoter within the viral vector utilized in my in vivo and in vitro experiments (224). Thus, the ability to modulate OT neuron firing rates, and the direction in which they are modulated, can be controlled by either PCX principal neuron type. Most previous research has either not differentiated between principal neurons types when examining association fiber connections into secondary structures, or they have focused only on pyramidal neuron connections (72,

73, 98, 225). Although it is unknown which cell type might extend its fibers into the OT, connectivity and anatomical evidence suggest that pyramidal neurons extend axons into the OT. Pyramidal neurons integrate information that is direct from the olfactory bulb, semilunar cell inputs, other pyramidal neuron inputs, and inputs from outside structures

(12, 87). PCX pyramidal neuron morphology is also well-suited to both extend associational input and receive this input (78, 87). While there is some limited research to suggest that semilunar cells extend association fibers extrinsically (142, 211, 226), there is no evidence that these cells extend into the OT, suggesting these connections likely arise from pyramidal neurons. If pyramidal neurons are the primary neurons responsible for inputs into the OT, it suggests that pyramidal neurons are extending holistic representations of odor information, as these neurons multiplex information from numerous input sources as given their connectivity (particularly their recurrent circuitry)

(84, 99, 136). Thus, the OT can then encode and represent this holistic olfactory information with only input from pyramidal neurons.

If it is the case that only pyramidal neurons extend input into the OT, how would one go about designing an experiment to determine this connectivity? Previous research utilized a transgenic mouse line that expresses Cre-recombinase in semilunar cells, but not pyramidal neurons (142), to determine connection specificity. This mouse line could be used in conjunction with a viral vector carrying a Cre-dependent viral tracer. Association fibers arising from the PCX into the OT could then be visualized to determine the population of OT neurons that receive input from semilunar cells, if they do at all. If there are no projections, this will confirm that PCX association fiber input is originating only from pyramidal neurons. However, the projection pattern from pyramidal

neurons should be confirmed into the OT utilizing a similar technique, although there are currently no viral vectors or transgenic animals that will specifically label pyramidal neurons. Thus, this line of research is constrained by constructs available to complete this task. In the future, it will be important to determine the cell type that leads to associational connections with the OT, as this will inform the way by which the OT receives olfactory inputs.

I also determined that the downstream targets of PCX principal neurons were both OT D1 and D2 MSNs. This was elucidated by using retroviral tracing in D1-Cre and

D2-Cre mice. Further work will need to resolve if the PCX extends projections onto other neurons within the OT. This is constrained by what is known about neurons in the

OT and what markers these neuron types might express that can be exploited to specifically label or identify these neurons. Thus, in conjunction with the previous experiment, it would be useful to determine OT targets of PCX association fibers through what is already known about these cell types, such as cell morphology. This line of research will elucidate a better understanding of the complete circuit between the

PCX and the OT. Importantly, my dissertation research has determined a necessary starting point, demonstrating a role for both principal neuron types within the PCX and

OT, which creates a framework for future connectivity studies.

Neural Consequences of PCX Association Fiber Perturbation

Here I demonstrated the way in which activation of PCX principal neurons and their association fibers modulate OT neural activity both in the presence and absence of odor inputs. I determined that: 1) the PCX modulates OT spontaneous activity, increasing or decreasing this activity; 2) the PCX globally suppresses OT odor representation, and; 3) the PCX bidirectionally modulates OT odor response dependent

upon the neuron’s intrinsic odor-evoked activity. Thus, this yields insight into the way by which activation of PCX principal neuron inputs modulates olfactory information representation within the OT.

However, there are some important caveats to consider in the interpretation of this activation-based paradigm. In order to understand the way in which one brain structure might impact the neural processing within another, the first approach traditionally seeks to answer the question of what happens in the absence of input from the upstream to downstream structure. Here I chose to activate, first, PCX cell bodies, and next, PCX axon terminals directly within the OT. The choice to activate the PCX, rather inhibit the PCX, was primarily due to technical limitations. Activation of PCX cell bodies or terminals was achieved by channelrhodopsin-based optogenetic manipulations. Channelrhodopsin itself is a light-inducible ion channel with rapid depolarization kinetics. Importantly, the mutation of channelrhodopsin utilized in these experiments has been shown to reliably induce depolarization specifically within pyramidal neurons, one of the principal cells types within the PCX (227). Thus, this manipulation allowed for the ability to rapidly and repeatedly depolarize the principal neurons of the PCX, either at the cell bodies or terminals, due to rapid cation influx at the channel pore.

The limitations arise when considering the inhibition of PCX neurons. Importantly, the optogenetic manipulations currently available widely utilize ion pumps

(halorhodopsin, archaerhodopsin) rather than ion channels, and are generally ineffective and unreliable at axon terminals (228). This means that inhibitory kinetics within these neurons are substantially slower and less reliable, meaning that there was

no available comparable manipulation to the channelrhodopsin-carrying vector utilized here. While there are viral vectors that have been created that carry a mutation of channelrhodopsin that allow for the passage of anions instead of cations, these are currently limited and not available for sale.

There are alternative approaches to consider when attempting to perform the complement of this activation experiment, which formally stated, would be: in what way does silencing of PCX association fiber inputs modulate OT neuron activity? There are two possible alternative approaches that should be attempted in the future to understand how OT neurons might encode odors in the absence of PCX afferents. The first approach would be to utilize a combination of anterograde and retrograde viral vectors to specifically target PCX neurons projecting into the OT. To do this, I would inject a retrograde viral vector carrying Cre into OT, which would back-infect neurons projecting into the OT. Then, I would inject a Cre-dependent inhibitory viral vector

(carrying halorhodopsin or archaerhodopsin under the CaMKiiα promoter) into the PCX, which will infect onto those neurons extending afferents into the OT. Thus, this will circumvent the difficulty of stimulating axon terminals, instead allowing me to stimulate

PCX cells that extend afferents into the OT. This is not without its caveats. This still suffers, although to a lesser degree, the problem of direct or indirect stimulation of OT cell bodies. While I would specifically be targeting the neurons within the PCX that project to the OT, I cannot know if these PCX neurons also project to, and activate, other neurons within extrinsic structures that might potentially feed back into the OT.

An additional alternative approach would be to utilize a viral vector that expresses the inwardly-rectifying potassium channel Kir2.1, which is constitutively

expressed within neurons. This would silence PCX neurons across days. To accomplish this, I would take the same viral approach, which would be to retrogradely infect OT- projecting PCX neurons with Cre, and then inject a Cre-dependent Kir2.1-expressing viral vector into the PCX. Over sessions, we would be able to monitor the way in which

OT neurons respond to odors without PCX input. An important caveat to this approach

(and a similar caveat to those approaches that use DREADDs) would be that we could not look at how neurons respond to odors with and without this PCX silencing. However, in conjunction with our activation paradigm, using one of the proposed silencing designs would provide some insight into general notions for odor coding with and without PCX inputs into the OT.

Using a silencing approach, what outcome would be expected? If excitation of

PCX principal neurons and their association axons led primarily to suppression of odor- evoked activity, but bidirectional control within OT neurons, the opposite is potentially true for silencing of PCX inputs. In the absence of PCX inputs, OT firing rates will be enhanced during odor inputs at the population level. Within individual neurons, odor- excited neurons will further increase their firing rates in the absence of PCX inputs, and odor-suppressed neurons will further decrease their firing rates. If this is the effect, what is the mechanism behind this modulation?

Here I determined that the PCX is capable of controlling the output of OT neurons. Thus, the PCX is likely performing a gain control function within the OT: depending upon the odor input, the PCX can dampen or enhance OT activity. Here, the

PCX served primarily to dampen odor responding in the OT, but this was entirely dependent upon an OT neuron’s intrinsic responsivity. Thus, by performing the opposite

experiment and silencing PCX neurons, the OT can only rely on odor input from the olfactory bulb and other interconnected olfactory structures, whose input is not nearly as extensive into the OT when compared with PCX association fiber input (173). Therefore, within an OT neuron, there is no gain control (i.e. bidirectional modulation) in the absence of PCX inputs, serving to only further enhance odor-excited and further suppress odor-inhibited OT neuron responses. If this is true, it suggests that the PCX exerts considerable control over sensory representation into the ventral striatum. It is important to note that in this framework, the OT is still responding to odor input, but representations are unrefined and largely originate from the olfactory bulb.

Unfortunately, beyond this dissertation research, nothing is known about the way in which the PCX influences odor coding in downstream structures, limiting what can be postulated from previous research. Thus, a priority of future experimentation should be to determine if, indeed, there is an opposing effect of PCX silencing onto OT neurons.

Importantly, here I utilized single neuron extracellular recordings to determine the way in which the PCX modulates the OT. Extracellular recordings are useful to monitor changes in electrical activity originating from many neurons within a downstream structure, thus allowing representation of a heterogenous cell population. The limitation to this, and one that is present in the results here, is that it is difficult to know precisely which neuron types are being recorded from and their connectivity patterns with the

PCX. Thus, to reach this goal, it will be necessary to fluorescently image the neurons within the OT during perturbations of the pathway between the PCX and OT (either activation or inhibition). Fluorescently imaging OT neurons expressing a calcium indicator (e.g. GCaMP for calcium transients (229)) would allow for visualization of

hundreds of OT neurons at a given time, and could be monitored during PCX stimulation (230). I established here that the PCX projects onto, and evokes responses within, D1 and D2 MSNs equally. Future research could utilize Cre-dependent GCaMP- expressing viral vector injections into the OT of D1-Cre and D2-Cre mice, allowing for monitoring of these cell populations with and without PCX activation, and/or olfactory input. What would be the expected outcome of this experiment?

D1 and D2 MSNs project to the direct and indirect pathways, respectively (231).

PCX input is equal onto these neuron types, suggesting that the PCX has the capacity to influence both basal ganglia streams. Thus, I expect that this input would be similar onto both pathways in vivo: both D1 and D2 MSNs can be bidirectionally odor- modulated, either excited or suppressed depending up that neuron’s odor responsivity.

To utilize a different scheme would suggest a biasing of information into the basal ganglia; for example, if a D1 neuron was always excited with PCX inputs, and D2 always suppressed, this would mean that PCX preferentially provides input towards positive, motivating behaviors and has limited influence over aversive behavioral outputs. This informational stream would be illogical for survival behaviors, especially when considering that the OT receives most of its olfactory input from the PCX (173).

The PCX must be able to influence both pathways so that the OT can properly assign a positive or negative value to an odor. Thus, the capacity to influence both input streams into the basal ganglia is important for appropriate motivated behaviors, and future in vivo research will be necessary to determine if this is true.

Behavioral and Perceptual Roles for the PCX-OT Connection

Understanding the cellular origin and termination of connectivity between structures is a necessary starting point, and examining the consequences of this

connectivity on neural activity downstream allows for at least an idea of what this connection might do, both with or without olfactory inputs. However, the way by which the perturbation of this connectivity modulates sensory perception or sensory-driven behaviors is an exciting goal of research on this connectivity. Does excitation or inhibition of a specific pathway modulate the way an animal perceives an olfactory stimulus, and is this modification capable of changing behavioral outputs?

Determining a perceptual or behavioral correlate of a specific connection is a high-risk, potentially low-yield endeavor. As a stimulus, olfactory information carries numerous features, including identity, intensity, valence, and novelty, among others.

The way by which structures might encode facets of this olfactory information, and which cortical structures are responsible for representing this information, has still yet to be fully elucidated. While extensive research has been conducted to understand the role each structure might play in particular facets of olfactory information coding, major discrepancies in understanding the function of each higher-order olfactory structure still remains to be determined. This makes finding the perceptual or behavioral consequences of specific pathways difficult to elucidate, and a major reason as to why this question was not addressed in this dissertation research.

However, there are practical approaches towards elucidating the function of a pathway from a behavioral or perceptual standpoint, as determining what modulation of a pathway means to an awake, behaving animal is an attractive concept. This is a particularly important consideration as it will be necessary to elucidate, in the future, if the change in OT activity upon PCX stimulation is truly perceptible to the animal. To begin, it is necessary to focus on olfactory system structures that have known roles for

encoding specific features of olfactory information. For example, determining the way by which PCX principal neuron excitation into the OT modulates the way an animal perceives and responds to an olfactory stimulus should be a focus of future experiments, and it would be intuitive to begin with features of odor information that each structure is involved in encoding. It is known that both the PCX and OT encode an odor’s identity and intensity (69, 70), and that each are involved in the learning of an olfactory stimulus, in particular, in a reward-based training task (68, 185, 232). Given these possibilities, there are numerous routes one could take to attempt to determine a behavioral or perceptual modification upon perturbation of this pathway.

The first possibility is that the transformation of information from PCX into OT represents odor identity, intensity, or both stimulus features. To determine this, an animal could be trained on an olfactory discrimination task, wherein the animal learns to discriminate between either two different odor identities (of the same intensity), or two different odor intensities (of the same identity). Once the animal can reliably differentiate between these two stimuli, the pathway from the PCX to the OT could be activated or inhibited (ideally both across different sessions) during the odor sampling period. If the animal is no longer capable of discriminating between these two stimuli, the animal’s perception of the odor has changed. This would be a modulation of the way that stimulus features are encoded in the brain.

Considering that the PCX is necessary for odor learning and memory, could it be the case that perturbing the input from PCX into the OT might modulate the ability for an animal to learn a stimulus? Utilizing an olfactory habituation paradigm would be the simplest way to determine if odor learning is impaired with PCX input disturbances into

the OT. To do this, an animal would be presented the same odor stimulus multiple times; however, each odor stimulus would be paired with this PCX activation. It is known that both the PCX and OT adapt (although OT to a lesser extent (69)) to repeated odor presentations). We can measure habituation by measuring the respiration of the animal: if there is a decrease in investigatory sniffing, then the animal is habituating. If the animal does not habituate to this repeated odor presentation, the animal is thus failing to adapt, meaning that PCX axon activation perturbs simple odor learning. If the animal does, however, habituate to the stimulus, it might be the case that the connection from the PCX into the OT is not involved in this type of olfactory learning, and might require a more complicated paradigm.

The final possibility is more complicated, but substantially more likely, given what is known about these two structures. The PCX extends a holistic odor percept into the

OT, allowing the OT to assign a reward value to this odor based upon the stimulus in its entirety. Thus, the PCX extends a percept of odor information that is given a positive or negative valence by the OT, which then is extended downstream into basal ganglia structures responsible for facilitating actions. In this way, PCX input into the OT allows for the facilitation (or suppression) of goal-directed behaviors. The potential processing scheme is complicated and necessitates a more nuanced behavioral approach. A possibility, to this end, would be to perturb this connection during the learning phase of an odor-reward pairing (invoking the idea that the ventral striatum is essential for instrumental learning (233)), and compare this to perturbing this connection after the animal has become an expert at this pairing, which is now represented in the dorsal striatum rather than ventral striatum (234). If perturbation of this pathway can modulate

the ability of the animal to learn this association (which would require activation and suppression not necessarily during the stimulus presentation period, but between the stimulus and the action that allows for a reward) but does nothing to the more robustly- represented odor-reward pairing, it is likely the case that the connection from the PCX into the OT is essential for initially learning odor-outcome pairings.

This interpretation is not without its caveats. The exact timing of information flow entering the OT from the PCX is not known, although work by both Carriero and colleagues (198), and the whole cell patch clamp recordings performed here, suggests that this occurs on a fairly-rapid timescale. Perturbing the connectivity during the stimulus period, or reward receipt period, might be necessary instead. Modifying the output of this connectivity during the stimulus period, however, makes it difficult to separate if the stimulus is being perceived differently, leading to diminished reward responding (as the animal does not recognize the stimulus in which it is supposed to respond to). Alternatively, stimulation during the reward receipt might just be changing the perception of the intensity or magnitude of the reward to that animal. These caveats and nuances will require careful future interpretation so as not to misinterpret or overextend the correlates of perceptual/behavioral outcomes upon modulation of the connection from the PCX into the OT.

One last necessary consideration for the interpretation of any sort of behavioral paradigm is to consider what happens if no effect is seen on olfactory perception or odor-facilitated behaviors at all. This would not necessarily mean that this pathway holds no importance in the olfactory system; instead, suggests the idea of redundancy in the sensory processing streams within the olfactory system. Considering the number

of interconnected structures within the olfactory system, it is not outside of the realm of possibility that there are a few built-in fail-safes; after all, olfactory processing is critical for numerous survival behaviors, so the notion that the brain might have some parallel pathways encoding the same or similar features of odor information is not out of the question. In fact, it has been previously proposed that the olfactory system might indeed contain some of these redundant parallel pathways (235), but future research will be required to determine if this is the case.

Overall, future experiments directly extending from this dissertation research should be the following: 1) determining the connectivity of both semilunar and pyramidal neurons of the PCX onto particular cell types within the OT; 2) inhibiting the connection from the PCX into the OT to examine how this might modulate OT neuron activity; 3) calcium-based imaging D1 and D2 MSNs within the OT during activation or suppression of PCX principal neuron inputs to determine the response properties of each neuron population; 4) attempting to dissect a potential behavioral consequence or perceptual modulation based upon the activation or suppression of the input from PCX principal neurons into the OT. These are all exciting potential future possibilities for this connectivity, and moreover, does not need to be restricted to this connection alone, but can be extrapolated into other interregional connections with slight modifications based upon cell type and known coding principles where necessary.

Caveats and Considerations

While some caveats to the experiments performed here have already been outlined, it is important to thoroughly note the limitations of the experiments that I performed here, and explain what was done to minimize caveats and be considerate of our limitations. The first caveat arises from the in vivo activation paradigm. While

addressed above, to some extent, there are limitations to how we might consider activation of PCX cell bodies and axon terminals on OT activity. First, it is necessary to consider just how biologically relevant the stimulation paradigm was. Here, PCX neurons and axon terminals were stimulated at 10 Hz. In vivo, PCX neurons fire, on average, at 2-5 Hz when animals are anesthetized (70). These neurons can respond to odors with firing rates greater than 5 Hz, but infrequently greater than 30 Hz (70). Thus, the stimulation paradigm utilized here is within the normal range of what would be expected of PCX activation. It should be noted that, as this was in anesthetized mice, we would expect these firing rates to be even lower in awake animals given the increased activation of inhibitory neurons in awake versus anesthetized animals. Even so, this is still within the normal range of PCX activation to odor inputs, which lends credence to the idea that the stimulation paradigm can be interpreted as biologically relevant.

A more philosophical consideration is this: there is (likely) never a natural condition under which so many PCX principal neurons are active all at one time, at the same firing frequency. Previous work has established that the PCX is, at its maximum, responding to odors with roughly 10% of its neurons (98). While the number of active neurons at one time within this paradigm is unknown, the histological quantification suggests that roughly half of all PCX principal neurons are infected. Thus, here I am stimulating somewhere less than half (dependent upon light output power and optical fiber location) of PCX cell bodies or terminals. If this paradigm is 100% effective, such that all infected neurons are being stimulated, PCX input into the OT is enhanced by roughly 5 times its usual rate. Conceptually, this can be interpreted as a mismatch

between the expected PCX input and true PCX input into the OT upon odor presentation. Thus, it is unsurprising that the OT is potentially playing a role in gain control, which is dampening its excitation, and enhancing its inhibition. The PCX is transmitting the message of what odor input it is receiving into the OT, which is much different than what the OT receives from the PCX with odor inputs alone. If the OT is receiving massive glutamatergic input from the PCX into the OT, the OT might simply be suppressing overall odor-evoked activity, as the PCX is disproportionately representing this activity. Considering this in terms of the weight of inputs, there has been a shift in sensory weighting, increasing this representation in the PCX, thus (to balance this input) decreasing this representation in the OT. The only way to truly rectify this role, though, is to perform the inhibitory experiments proposed above.

Another consideration that extends along the lines of biological significance and relevance: these animals are head-fixed, water-restricted, and passively exposed to the odors presented here. It evident that this is not how animals experience odors in day to day life. Why did I choose this specific paradigm to understand the contributions of PCX to OT odor coding? This is primarily due to technical limitations. A main consideration is that the OT is a very ventrally-located brain structure, which means that this structure is subject to picking up electrical movement artifact. When animals move during electrophysiological recordings, large, slow-wave noise is represented on the channels.

While this can be minimized with appropriate referencing and grounding, this is still present and can diminish or hide any true signals that are coming from the OT. In order to minimize this caveat, I chose to head-fix the mice in this paradigm. Mice, when properly acclimated to the head restraint, will only minimally move during the behavioral

session, decreasing the probability of movement noise and increasing the probability of acquiring true multiunit activity from the OT.

An additional consideration to this is the potential effect of stress on the animal.

Head fixation and water restriction are stressful conditions to an animal. To minimize the effect of stress, I introduce water restriction slowly, by tapering animals down from their ad libitum amount to smaller amounts across several days. Animals are also acclimated to head fixation over a number of days before true experimental sessions begin. During the acclimation process, mice are rewarded simply for being head-fixed, which should serve reduce the stress of the context over time. As the animal is slowly introduced to both of these conditions, over days, it is likely that the animal will be minimally stressed during this time period, and indeed, ultimately find this context rewarding over time.

Overall this research served as an important first-pass attempt to understand a potential function of the connection from the PCX into the OT. This research has created a framework for potential important future experiments to be conducted, informing how the PCX might shape odor representation within the OT.

PCX and OT as Components of a Cortico-Striatal Loop

The biasing of input from the PCX into the OT can be interpreted in a broader context beyond sensory system processing. The PCX and OT are unique higher-order structures in that both are “cortical-like” but contain features that are non-cortical. The

PCX functions as an associative structure that is involved in the encoding and representation of holistic sensory information, determining the meaning behind a stimulus (12, 15, 16). However, the PCX is trilaminar and utilizes traditional cortical computations and cell types to process sensory information (12, 78). The OT is a ventral striatal structure in every capacity, from cellular composition, neurotransmitter

release, connectivity, and reward processing, with really the only “cortical” features pertaining to its lamination (236). Despite these nontraditional features, it is possible that the connection from the PCX into the OT comprises a previously unappreciated cortico-striatal loop, meaning that this connectivity potentially has broader implications for the way that sensory information is encoded and extended into action selection centers within the brain.

To begin elucidating if the connection from the PCX into the OT constitutes a cortico-striatal loop, it is first necessary to determine the features of a cortico-striatal loop. Cortico-striatal connectivity on a broad scale allows for the planning and execution of goal-directed behaviors based upon the orchestration of higher-order inputs into the striatum (237). This loop originates from the cortex, which extends a unidirectional input into the striatum as the striatum can only indirectly project back to the cortex (238). The cortex subserves the role of performing top-down modulation onto striatal structures.

Numerous stimulus attributes coalesce to form a perceptual whole in the cortex, which is then extended into the striatum, where this information is processed and assigned a response/outcome (237). There are nuances to this information flow, however, so it is first important to describe canonical features of a cortico-striatal loop and apply these same principles to the connection from PCX into OT, noting exceptions to this rule where applicable.

A cortico-striatal loop is not simply a single-synapse cortex-to-striatum input.

Cortico-striatal loops integrate inputs from thalamic and brainstem areas to facilitate and enact motivated behaviors that require these centers to act in concert: orchestrating motor planning, integrating cognitive assignments of sensory inputs, and including

elements of reward and motivation (237). This circuitry works together so that the cortex extends important stimulus features that require a behavioral output downstream, but also integrate inputs from outside structures that allow cortico-striatal representations to be modified (237). This ability to modify this circuit with new, incoming information leads to the notion that the cortico-striatal loop maintains the task of learning a stimulus, quickly, to execute optimal goal-directed behaviors. Thus, this system is maintained in terms of efficiency, but not hard-wired if contingencies change.

It is useful to think about this concretely in terms of an animal engaged in a learning paradigm. For example, an animal has learned that an auditory cue leads to a sugar reward, so long as the animal performs the appropriate behavior, like lever- pressing or nose-poking. If the animal has been well-trained on this association, the cortico-striatal loop representing this association will have been optimized to rapidly encode this information, determining the required action to receive the reward, and ultimately performing this behavior, reinforcing this stimulus-reward pairing. However, in the event that the contingency changes – for example, a new tone is now rewarded, and the previous tone no longer cues a reward – the animal will need to encode this stimulus information differently to evoke a response via plastic portions of this loop, and it will take time for this response to become rapid and reliable. Thus, this loop first allows for learning of stimulus meaning via circuit plasticity, and then enables stable and rapid representations within this loop over repeated encounters.

Generally, the anatomy of a cortico-striatal loop is as follows: stimulus features are extended into the striatum via the pyramidal neurons of the cortical structure. Most cortico-striatal loops originate from six-layered cortices, which in and of themselves

have the ability to process information across its layers prior to exiting the structure

(237). Connections from pyramidal neurons in these cortices arise from layer iii and v pyramidal neurons after intrinsic modifications of sensory information (237, 239). This information is extended from patchy ensembles of pyramidal cortical neurons into the striatum, projecting into the ventral or dorsal striatum dependent upon the point of origin

(237, 239). This pyramidal neuron input synapses with, and evokes stronger responses within, fast-spiking GABAergic interneurons of the striatum, which in turn project to

MSNs (240). It has been postulated that these interneurons integrate information from numerous cortical targets before evoking post-synaptic potentials within a MSN, meaning that there is at least one striatal relay prior to responsivity by the projection neurons of the striatum. Additionally, dopaminergic innervation arising from the substantia nigra or ventral tegmental area serves to tune a striatal neuron to what stimulus inputs are likely important (241). Indeed, dopaminergic synapses position themselves on MSN dendrites to modulate input from the cortex (241). Thus, there are some processing mechanisms both within the cortex and striatum that might modify or gate incoming information into the cortico-striatal loop.

With these features in mind, is it fair to consider the connection from the PCX into the OT a cortico-striatal loop? Yes, on a few accounts: 1) the PCX integrates sensory information into a holistic odor percept (12, 15, 16), performing the same cortical role proposed in a canonical cortico-striatal loop; 2) the OT encodes the valence of an odor stimulus (185), and is thus involved in encoding outcome contingencies; 3) the OT receives dopaminergic input from the ventral tegmental area (242, 243), which signals salience within this loop; and 4) both the PCX and the OT have connections with

thalamic and brainstem structures (78, 146, 153, 171), thus suggesting the same level of plasticity as proposed above. Altogether, this suggests that the PCX represents numerous facets of olfactory information (that has the potential to be modified via recurrent excitation) and is necessary for the learning of olfactory associations, and since the OT is both perfectly situated to, and has been shown to, encode the value of a stimulus, the connection from the PCX into the OT therefore represents a specialized olfactory cortico-striatal loop.

Role of Recurrent Circuitry in PCX and OT Odor Representations

Here I established that activation of PCX neurons or their association fibers leads to suppression of global OT neuron odor response, yet intrinsically bidirectionally modulates a single neuron’s odor-evoked activity. At first, this appears counterintuitive.

Projections from PCX principal neuron association fibers into the OT are glutamatergic

(73), so one would expect that activation of these fibers leads to an increase in both odor-evoked and spontaneous excitation. Here, while there was excitation of spontaneous firing rates, this did not translate to the same phenomenon during odor presentation. Beyond this, how is it possible that at a population level, PCX activation suppresses an OT odor response, but within neurons, leads to bidirectional excitation/suppression dependent upon the way in which a neuron responds to odor alone? If it is the case that activation of association fibers suppress odor activity, why does this not broadly occur within neurons as well?

First, what are the mechanisms behind global odor-evoked suppression in the

OT from PCX activation? The OT receives olfactory input from numerous structures: directly from the olfactory bulb, heavily from the PCX, and also afferents from the anterior olfactory nucleus and amygdala (36, 147, 148, 161, 162, 168, 173). The latter

three are not direct, which potentially means that the information these structures provide has been processed and filtered to give rise to different qualities of information that are not necessarily redundant with what the olfactory bulb is providing. When all of these structures receive odor information, input originating from these structures appears to converge to elicit some odor-excited responses and some odor-inhibited responses across OT neurons (69–71). It is unknown what OT neuron type is responsible for each type of stimulus-evoked activity, nor is it known how each neuron type is tuned to this input, and further, the degree of axonal convergence onto these OT neurons has not been established. Are certain neurons receiving information from only the olfactory bulb, or only the PCX, or are there neurons that multiplex information from all of these structures? In addition, how might these OT neurons be connected between one another? Considering how little is known, the following framework to interpret the results displayed here is based upon what is known of the cortico-striatal connections and intrinsic dynamics of ventral striatal structures and their connectivity schemes.

With amplified input from the PCX, both by stimulating PCX principal neurons, and stimulating association fibers directly from PCX within the OT, there is widespread suppression to olfactory input. This can be interpreted as the PCX broadly depressing

OT odor response. If the PCX is capable of evoking excitation within OT neurons in the absence of sensory input, it might be the case that this widespread suppression is a function of the synergistic combination of unexpected PCX activation and reception of these other olfactory inputs that are largely glutamatergic. Thus, there is perhaps a compensatory mechanism taking place across all of these inputs to dampen heightened

PCX inputs into the OT.

100

However, the key here is that not all OT neurons are suppressed all the time during concurrent odor presentation and PCX activation. At a single-neuron level, odor- excited neurons become less excited during PCX activation, while odor-suppressed neurons become less suppressed during PCX activation. This means that the population that is responsive (either with excitation or suppression) is switching, but why? The most likely explanation involves the role of recurrent inhibitory dynamics within the OT, as has previously been demonstrated in the nucleus accumbens (159).

Like the OT, the nucleus accumbens is predominantly comprised of these GABAergic

MSNs (150, 161, 162). These MSNs are connected with one another, such that both D1 and D2 MSNs have the capacity to synapse with and inhibit one another (159).

Considering this striatal feature, it is likely the case that odor-evoked responsivity is entirely dependent upon connectivity with both PCX neurons and with other MSNs.

Thus, this connectivity scheme leads to the following circuit-based framework for the role of PCX inputs onto OT neurons during odor presentation (Fig 3-1): OT neurons directly connected to PCX principal neurons are excited by PCX activation, and may be

(although not always) odor-suppressed, leading to excitation during concurrent odor presentation and PCX activation. In contrast, OT neurons not directly connected to PCX neurons are unchanged by/suppressed by PCX stimulation, and may be (although not always) odor-excited, leading to suppression during concurrent odor presentation and

PCX activation. How can this be interpreted at a global scale? OT neurons that are suppressed to odors become less suppressed with PCX activation, leading to either no change from background firing rate, or slight suppression, typically. OT neurons that are not responsive to odor inputs become suppressed with PCX stimulation, likely due to

101

these recurrent GABAergic connections from other MSNs. OT neurons that are excited with odor input either become non-responsive to odors with PCX activation, or become slightly suppressed. This biases the population to represent more suppressed responses globally yet allows for maintenance of within-neuron bidirectional responding.

Thus, this framework proposes that the PCX, while able to evoke excitation within the

OT, serves more broadly during odor presentation to increase the probability of recurrent inhibition within the OT.

102

Figure 3-1. Simplified schematic summarizing results. PCX association fibers release glutamate, exciting the downstream OT MSN. Other interconnected OT MSNs become suppressed by the upstream excited OT MSN, leading to recurrent inhibition within the OT.

103

LIST OF REFERENCES

1. Doty RL (2012) Olfaction in Parkinson’s disease and related disorders. Neurobiol Dis 46(3):527–552.

2. Doty RL (2017) Olfactory dysfunction in neurodegenerative diseases: is there a common pathological substrate? Lancet Neurol 16(6):478–488.

3. Barresi M, et al. (2012) Evaluation of olfactory dysfunction in neurodegenerative diseases. J Neurol Sci 323(1):16–24.

4. Godoy MDCL, Voegels RL, Pinna F de R, Imamura R, Farfel JM (2015) Olfaction in Neurologic and Neurodegenerative Diseases: A Literature Review. Int Arch Otorhinolaryngol 19(2):176–179.

5. Ffytche DH, Blom JD, Catani M (2010) Disorders of visual perception. J Neurol Neurosurg Psychiatry 81(11):1280–1287.

6. Polster MR, Rose SB (1998) Disorders of Auditory Processing: Evidence for Modularity in Audition. Cortex 34(1):47–65.

7. Wilson RI, Mainen ZF (2006) Early events in olfactory processing. Annu Rev Neurosci 29:163–201.

8. Lennie P (1980) Parallel visual pathways: A review. Vision Res 20(7):561–594.

9. Rubel EW, Fritzsch B (2002) Auditory system development: primary auditory neurons and their targets. Annu Rev Neurosci 25:51–101.

10. Abraira VE, Ginty DD (2013) The Sensory Neurons of Touch. Neuron 79(4):10.1016/j.neuron.2013.07.051.

11. Fanselow EE, Nicolelis MA (1999) Behavioral modulation of tactile responses in the rat somatosensory system. J Neurosci 19(17):7603–7616.

12. Haberly LB (2001) Parallel-distributed processing in olfactory cortex: new insights from morphological and physiological analysis of neuronal circuitry. Chem Senses 26(5):551–576.

13. Amoore JE (1970) Molecular basis of odor (Charles C. Thomas, Springfield, Illinois).

14. Vickers NJ (2000) Mechanisms of animal navigation in odor plumes. Biol Bull 198(2):203–212.

15. Gottfried JA (2010) Central mechanisms of odour object perception. Nat Rev Neurosci 11(9):628–641.

16. Wilson DA, Sullivan RM (2011) Cortical processing of odor objects. Neuron

104

72(4):506–519.

17. Ogle W (1870) Anosmia, or Cases illustrating the Physiology and Pathology of the Sense of Smell. Med Chir Trans 53:263–290.

18. Kepecs A, Uchida N, Mainen ZF (2006) The sniff as a unit of olfactory processing. Chem Senses 31(2):167–179.

19. Malnic B, Hirono J, Sato T, Buck LB (1999) Combinatorial receptor codes for odors. Cell 96(5):713–723.

20. Vassar R, Ngal J, Axel R (1993) Spatial segregation of odorant receptor expression in the mammalian olfactory epithelium. Cell 74:309–318.

21. Buck L, Axel R (1991) A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65:175.

22. Araneda RC, Kini AD, Firestein S (2000) The molecular receptive range of an odorant receptor. Nat Neurosci 3(12):1248–1255.

23. Mori K (2006) Maps of Odorant Molecular Features in the Mammalian Olfactory Bulb. Physiol Rev 86(2):409–433.

24. Scott JW, Davis LM, Shannon D, Kaplan C (1996) Relation of chemical structure to spatial distribution of sensory responses in rat olfactory epithelium. J Neurophysiol 75(5):2036–2049.

25. Ma M, Shepherd GM (2000) Functional mosaic organization of mouse olfactory receptor neurons. Proc Natl Acad Sci U S A 97(23):12869–12874.

26. Johnson BA, Woo CC, Hingco EE, Pham KL, Leon M (1999) Multidimensional chemotopic responses to n-aliphatic acid odorants in the rat olfactory bulb. J Comp Neurol 409(4):529–548.

27. Rubin BD, Katz LC (1999) Optical imaging of odorant representations in the mammalian olfactory bulb. Neuron 23(3):499–511.

28. Uchida N, Takahashi YK, Tanifuji M, Mori K (2000) Odor maps in the mammalian olfactory bulb: domain organization and odorant structural features. Nat Neurosci 3(10):1035–1043.

29. Imai T (2014) Construction of functional neuronal circuitry in the olfactory bulb. Semin Cell Dev Biol 35:180–188.

30. Leinders-Zufall T, Greer CA, Shepherd GM, Zufall F (1998) Visualizing odor detection in olfactory cilia by calcium imaging. Ann N Y Acad Sci 855:205–207.

31. Johnson BA, Woo CC, Leon M (1998) Spatial coding of odorant features in the

105

glomerular layer of the rat olfactory bulb. J Comp Neurol 393(4):457–471.

32. Spors H, Grinvald A (2002) Spatio-temporal dynamics of odor representations in the mammalian olfactory bulb. Neuron 34(2):301–315.

33. Thommesen G (1978) The spatial distribution of odour induced potentials in the olfactory bulb of char and trout (Salmonidae). Acta Physiol Scand 102(2):205– 217.

34. Sosulski DL, Lissitsyna Bloom M, Cutforth T, Axel R, Datta SR (2011) Distinct representations of olfactory information in different cortical centres. Nature 472(7342):213–216.

35. Ke M-T, Fujimoto S, Imai T (2013) SeeDB: a simple and morphology-preserving optical clearing agent for neuronal circuit reconstruction. Nat Neurosci 16:1154.

36. Ojima H, Mori K, Kishi K (1984) The trajectory of mitral cell axons in the rabbit olfactory cortex revealed by intracellular HRP injection. J Comp Neurol 230(1):77– 87.

37. Nagayama S, Homma R, Imamura F (2014) Neuronal organization of olfactory bulb circuits. Front Neural Circuits 8:98.

38. Nagayama S, Takahashi YK, Yoshihara Y, Mori K (2004) Mitral and tufted cells differ in the decoding manner of odor maps in the rat olfactory bulb. J Neurophysiol 91(6):2532–2540.

39. Igarashi KM, et al. (2012) Parallel Mitral and Tufted Cell Pathways Route Distinct Odor Information to Different Targets in the Olfactory Cortex. J Neurosci 32(23):7970–7985.

40. Gire DH, et al. (2012) Mitral Cells in the Olfactory Bulb Are Mainly Excited through a Multistep Signaling Path. J Neurosci 32(9):2964 LP-2975.

41. Fukunaga I, Berning M, Kollo M, Schmaltz A, Schaefer AT (2012) Two Distinct Channels of Olfactory Bulb Output. Neuron 75(2):320–329.

42. Ezeh PI, Wellis DP, Scott JW (1993) Organization of inhibition in the rat olfactory bulb external plexiform layer. J Neurophysiol 70(1):263–274.

43. Kikuta S, Fletcher ML, Homma R, Yamasoba T, Nagayama S (2013) Odorant Response Properties of Individual Neurons in an Olfactory Glomerular Module. Neuron 77(6):1122–1135.

44. De Saint Jan D, Hirnet D, Westbrook GL, Charpak S (2009) External Tufted Cells Drive the Output of Olfactory Bulb Glomeruli. J Neurosci 29(7):2043 LP-2052.

45. Haberly LB, Price JL (1977) The axonal projection patterns of the mitral and tufted

106

cells of the olfactory bulb in the rat. Brain Res 129(1):152–157.

46. Skeen LC, Hall WC (1977) Efferent projections of the main and the accessory olfactory bulb in the tree shrew (Tupaia glis). J Comp Neurol 172:1–35.

47. Scott JW, McBride RL, Schneider SP (1980) The organization of projections from the olfactory bulb to the piriform cortex and olfactory tubercle in the rat. J Comp Neurol 194(3):519–534.

48. Nagayama S, et al. (2010) Differential Axonal Projection of Mitral and Tufted Cells in the Mouse Main Olfactory System. Front Neural Circuits 4:1–8.

49. Urban NN, Sakmann B (2002) Reciprocal intraglomerular excitation and intra- and interglomerular lateral inhibition between mouse olfactory bulb mitral cells. J Physiol 542(Pt 2):355–367.

50. Vucinic D (2006) Interglomerular Center-Surround Inhibition Shapes Odorant- Evoked Input to the Mouse Olfactory Bulb In Vivo. J Neurophysiol 95(3):1881– 1887.

51. Kosaka T, Kosaka K (2008) Tyrosine hydroxylase-positive GABAergic juxtaglomerular neurons are the main source of the interglomerular connections in the mouse main olfactory bulb. Neurosci Res 60(3):349–354.

52. Kiyokage E, et al. (2010) Molecular Identity of Periglomerular and Short Axon Cells. J Neurosci 30(3):1185 LP-1196.

53. Isaacson JS, Strowbridge BW (1998) Olfactory reciprocal synapses: dendritic signaling in the CNS. Neuron 20(4):749–761.

54. Urban NN (2002) Lateral inhibition in the olfactory bulb and in olfaction. Physiol Behav 77(4–5):607–612.

55. Linster C, Gervais R (1996) Investigation of the role of interneurons and their modulation by centrifugal fibers in a neural model of the olfactory bulb. J Comput Neurosci 3(3):225–246.

56. Halabisky B, Strowbridge BW (2003) Gamma-frequency excitatory input to granule cells facilitates dendrodendritic inhibition in the rat olfactory Bulb. J Neurophysiol 90(2):644–654.

57. Yokoi M, Mori K, Nakanishi S (1995) Refinement of odor molecule tuning by dendrodenritic synaptic inhibition in the olfactory bulb. Proc Natl Acad Sci 92:3371–3375.

58. Rosenkranz JA, Grace AA (2002) Dopamine-mediated modulation of odour- evoked amygdala potentials during pavlovian conditioning. Nature 417(6886):282–287.

107

59. Schoenbaum G, Chiba AA, Gallagher M (1999) Neural encoding in orbitofrontal cortex and basolateral amygdala during olfactory discrimination learning. J Neurosci 19(5):1876–1884.

60. Xu W, Wilson DA (2012) Odor-evoked activity in the mouse lateral entorhinal cortex. Neuroscience 223(0):12–20.

61. Chapuis J, et al. (2013) Lateral Entorhinal Modulation of Piriform Cortical Activity and Fine Odor Discrimination. J Neurosci 33(33):13449–13459.

62. Kay RB, Meyer EA, Illig KR, Brunjes PC (2011) Spatial distribution of neural activity in the anterior olfactory nucleus evoked by odor and electrical stimulation. J Comp Neurol 519(2):277–289.

63. Kikuta S, et al. (2010) Neurons in the anterior olfactory nucleus pars externa detect right or left localization of odor sources. Proc Natl Acad Sci 107(27):12363–12368.

64. Lei H, Mooney R, Katz LC (2006) Synaptic integration of olfactory information in mouse anterior olfactory nucleus. J Neurosci 26(46):12023–12032.

65. Stettler DD, Axel R (2009) Representations of Odor in the Piriform Cortex. Neuron 63(6):854–864.

66. Rennaker RL, Chen C-FF, Ruyle AM, Sloan AM, Wilson DA (2007) Spatial and Temporal Distribution of Odorant-Evoked Activity in the Piriform Cortex. J Neurosci 27(7):1534–1542.

67. Calu DJ, Roesch MR, Stalnaker TA, Schoenbaum G (2007) Associative Encoding in Posterior Piriform Cortex during Odor Discrimination and Reversal Learning. Cereb Cortex 17(6):1342–1349.

68. Roesch MR, Stalnaker TA, Schoenbaum G (2007) Associative encoding in anterior piriform cortex versus orbitofrontal cortex during odor discrimination and reversal learning. Cereb Cortex 17(3):643–652.

69. Xia CZ, Adjei S, Wesson DW (2015) Coding of odor stimulus features among secondary olfactory structures. J Neurophysiol 114(1):736–45.

70. Payton CA, Wilson DA, Wesson DW (2012) Parallel Odor Processing by Two Anatomically Distinct Olfactory Bulb Target Structures. PLoS One 7(4):e34926.

71. Wesson DW, Wilson DA (2010) Smelling Sounds: Olfactory-Auditory Sensory Convergence in the Olfactory Tubercle. J Neurosci 30(8):3013–3021.

72. Johnson DM, Illig KR, Behan M, Haberly LB (2000) New features of connectivity in piriform cortex visualized by intracellular injection of pyramidal cells suggest that “primary” olfactory cortex functions like “association” cortex in other sensory

108

systems. J Neurosci 20(18):6974–6982.

73. Haberly LB, Price JL (1978) Association and commissural fiber systems of the olfactory cortex of the rat. II. Systems originating in the olfactory peduncle. J Comp Neurol 181(4):781–807.

74. Luskin MB, Price JL (1983) The topographic organization of associational fibers of the olfactory system in the rat, including centrifugal fibers to the olfactory bulb. J Comp Neurol 216:264–291.

75. Otazu GH, Chae H, Davis MB, Albeanu DF (2017) Cortical Feedback Decorrelates Olfactory Bulb Output in Awake Mice. Neuron 86(6):1461–1477.

76. Boyd AM, Sturgill JF, Poo C, Isaacson JS (2012) Cortical Feedback Control of Olfactory Bulb Circuits. Neuron 76(6):1161–1174.

77. Padmanabhan K, et al. (2016) Diverse Representations of Olfactory Information in Centrifugal Feedback Projections. J Neurosci 36(28):7535 LP-7545.

78. Neville KR, Haberly L (2004) Olfactory cortex. The Synaptic Organization of the Brain, ed Shepherd GM (Oxford University Press, New York), pp 415–454. 5th Ed.

79. Hagiwara A, Pal SK, Sato TF, Wienisch M, Murthy VN (2012) Optophysiological analysis of associational circuits in the olfactory cortex. Front Neural Circuits 6:18.

80. Litaudon P, Amat C, Bertrand B, Vigouroux M, Buonviso N (2003) Piriform cortex functional heterogeneity revealed by cellular responses to odours. Eur J Neurosci 17(11):2457–2461.

81. Luskin MB, Price JL (1983) The laminar distribution of intracortical fibers originating in the olfactory cortex of the rat. J Comp Neurol 216(3):292–302.

82. Ketchum KL, Haberly LB (1993) Membrane currents evoked by afferent fiber stimulation in rat piriform cortex. II. analysis with a system model. J Neurophysiol 69:261–281.

83. Ketchum KL, Haberly LB (1993) Synaptic events that generate fast oscillations in piriform cortex. J Neurosci 13(9):3980–3985.

84. Kadohisa M, Wilson DA (2006) Separate encoding of identity and similarity of complex familiar odors in piriform cortex. Proc Natl Acad Sci U S A 103(41):15206–15211.

85. Gottfried JA (2006) Smell: central nervous processing. Adv Otorhinolaryngol 63:44–69.

86. Zelano C, Mohanty A, Gottfried JA (2011) Olfactory Predictive Codes and

109

Stimulus Templates in Piriform Cortex. Neuron 72(1):178–187.

87. Bekkers JM, Suzuki N (2013) Neurons and circuits for odor processing in the piriform cortex. Trends Neurosci 36(7):429–438.

88. Suzuki N, Bekkers JM (2007) Inhibitory interneurons in the piriform cortex. Clin Exp Pharmacol Physiol 34(10):1064–1069.

89. Haberly LB (1983) Structure of the piriform cortex of the opossum. I. Description of neuron types with Golgi methods. J Comp Neurol 213(2):163–187.

90. Suzuki N, Bekkers JM (2011) Two Layers of Synaptic Processing by Principal Neurons in Piriform Cortex. J Neurosci 31(6):2156–2166.

91. Miyamichi K, et al. (2011) Cortical representations of olfactory input by transsynaptic tracing. Nature 472(7342):191–196.

92. Ghosh S, et al. (2011) Sensory maps in the olfactory cortex defined by long-range viral tracing of single neurons. Nature 472(7342):217–220.

93. Haddad R, et al. (2013) Olfactory cortical neurons read out a relative time code in the olfactory bulb. Nat Neurosci 16(7):949–957.

94. Arenkiel BR, et al. (2007) In Vivo Light-Induced Activation of Neural Circuitry in Transgenic Mice Expressing Channelrhodopsin-2. Neuron 54(2):205–218.

95. Apicella CL (2010) No association between oxytocin receptor (OXTR) gene polymorphisms and experimentally elicited social preferences. PLoS One 5:e11153.

96. Suzuki N, Bekkers JM (2006) Neural coding by two classes of principal cells in the mouse piriform cortex. J Neurosci 26(46):11938–11947.

97. Franks KM, Isaacson JS (2006) Strong single-fiber sensory inputs to olfactory cortex: implications for olfactory coding. Neuron 49(3):357–363.

98. Franks KM, et al. (2011) Recurrent Circuitry Dynamically Shapes the Activation of Piriform Cortex. Neuron 72(1):49–56.

99. Davison IG, Ehlers MD (2011) Neural Circuit Mechanisms for Pattern Detection and Feature Combination in Olfactory Cortex. Neuron 70(1):82–94.

100. Gire DH, et al. (2013) Temporal Processing in the Olfactory System: Can We See a Smell? Neuron 78(3):416–432.

101. Zhan C, Luo M (2010) Diverse Patterns of Odor Representation by Neurons in the Anterior Piriform Cortex of Awake Mice. J Neurosci 30(49):16662 LP-16672.

102. Poo C, Isaacson JS (2009) Odor Representations in Olfactory Cortex: Sparse

110

Coding, Global Inhibition, and Oscillations. 62(6):850–861.

103. Large AM, Vogler NW, Mielo S, Oswald A-MM (2016) Balanced feedforward inhibition and dominant recurrent inhibition in olfactory cortex. Proc Natl Acad Sci 113(8):2276–2281.

104. Choy JMC, et al. (2017) Optogenetic Mapping of Intracortical Circuits Originating from Semilunar Cells in the Piriform Cortex. Cereb Cortex 27(1):589–601.

105. Suzuki N, Bekkers JM (2012) Microcircuits Mediating Feedforward and Feedback Synaptic Inhibition in the Piriform Cortex. J Neurosci 32(3):919 LP-931.

106. Kapur A, Pearce RA, Lytton WW, Haberly LB (1997) GABAA-mediated IPSCs in piriform cortex have fast and slow components with different properties and locations on pyramidal cells. J Neurophysiol 78(5):2531–2545.

107. Ekstrand JJ, et al. (2001) A new subdivision of anterior piriform cortex and associated deep nucleus with novel features of interest for olfaction and epilepsy. J Comp Neurol 434(3):289–307.

108. Suzuki N, Bekkers JM (2010) Inhibitory neurons in the anterior piriform cortex of the mouse: Classification using molecular markers. J Comp Neurol 518(10):1670– 1687.

109. Luna VM, Schoppa NE (2008) GABAergic circuits control input-spike coupling in the piriform cortex. J Neurosci 28(35):8851–8859.

110. Stokes CCA, Isaacson JS (2010) From Dendrite to Soma: Dynamic Routing of Inhibition by Complementary Interneuron Microcircuits in Olfactory Cortex. Neuron 67(3):452–465.

111. Ekstrand JJ, Domroese ME, Feig SL, Illig KR, Haberly LB (2001) Immunocytochemical analysis of basket cells in rat piriform cortex. J Comp Neurol 434(3):308–328.

112. Yoshida I, Mori K (2007) Odorant category profile selectivity of olfactory cortex neurons. J Neurosci 27(34):9105–9114.

113. Wilson DA (2003) Rapid, experience-induced enhancement in odorant discrimination by anterior piriform cortex neurons. J Neurophysiol 90(1):65–72.

114. Isaacson JS (2010) Odor representations in mammalian cortical circuits. Curr Opin Neurobiol 20(3):328–331.

115. Franks KM, Isaacson JS (2005) Synapse-specific downregulation of NMDA receptors by early experience: a critical period for plasticity of sensory input to olfactory cortex. Neuron 47(1):101–114.

111

116. Kanter ED, Haberly LB (1990) NMDA-dependent induction of long-term potentiation in afferent and association fiber systems of piriform cortex in vitro. Brain Res 525(1):175–179.

117. Johenning FW, et al. (2009) Dendritic Compartment and Neuronal Output Mode Determine Pathway-Specific Long-Term Potentiation in the Piriform Cortex. J Neurosci 29(43):13649 LP-13661.

118. Oswald A-MM, Urban NN (2012) Interactions between Behaviorally Relevant Rhythms and Synaptic Plasticity Alter Coding in the Piriform Cortex. J Neurosci 32(18):6092–6104.

119. Bathellier B, Buhl DL, Accolla R, Carleton A (2008) Dynamic Ensemble Odor Coding in the Mammalian Olfactory Bulb: Sensory Information at Different Timescales. Neuron 57(4):586–598.

120. Barkai E, Hasselmo ME (1994) Modulation of the input/output function of rat piriform cortex pyramidal cells. J Neurophysiol 72(2):644–658.

121. Howard JD, Plailly J, Grueschow M, Haynes J-D, Gottfried JA (2009) Odor quality coding and categorization in human posterior piriform cortex. Nat Neurosci 12(7):932–938.

122. Gire DH, Whitesell JD, Doucette W, Restrepo D (2013) Information for decision- making and stimulus identification is multiplexed in sensory cortex. Nat Neurosci 16(8):991–993.

123. Miura K, Mainen ZF, Uchida N (2012) Odor Representations in Olfactory Cortex: Distributed Rate Coding and Decorrelated Population Activity. Neuron 74(6):1087–1098.

124. Buonviso N, Revial MF, Jourdan F (1991) The Projections of Mitral Cells from Small Local Regions of the Olfactory Bulb: An Anterograde Tracing Study Using PHA-L (Phaseolus vulgaris Leucoagglutinin). Eur J Neurosci 3(6):493–500.

125. McCollum J, et al. (1991) Short-latency single-unit processing in olfactory cortex. J Cog Neurosci 3:293–299.

126. Schoenbaum G, Eichenbaum H (1995) Information coding in the rodent prefrontal cortex. I. Single-neuron activity in orbitofrontal cortex compared with that in pyriform cortex. J Neurophysiol 74(2):733–750.

127. Poo C, Isaacson JS (2011) A Major Role for Intracortical Circuits in the Strength and Tuning of Odor-Evoked Excitation in Olfactory Cortex. Neuron 72(1):41–48.

128. Tang AC, Hasselmo ME (1994) Selective suppression of intrinsic but not afferent fiber synaptic transmission by baclofen in the piriform (olfactory) cortex. Brain Res 659(1–2):75–81.

112

129. Wilson DA (1998) Habituation of odor responses in the rat anterior piriform cortex. J Neurophysiol 79(3):1425–1440.

130. Wilson DA (1998) Synaptic correlates of odor habituation in the rat anterior piriform cortex. J Neurophysiol 80(2):998–1001.

131. Wilson DA (2000) Odor specificity of habituation in the rat anterior piriform cortex. J Neurophysiol 83(1):139–145.

132. Wilson DA (2000) Comparison of odor receptive field plasticity in the rat olfactory bulb and anterior piriform cortex. J Neurophysiol 84(6):3036–3042.

133. Sobel N, et al. (2000) Time course of odorant-induced activation in the human primary olfactory cortex. J Neurophysiol 83(1):537–551.

134. Chapuis J, Wilson DA (2011) Bidirectional plasticity of cortical pattern recognition and behavioral sensory acuity. Nat Neurosci 15(1):155–161.

135. Kadohisa M, Wilson DA (2006) Olfactory cortical adaptation facilitates detection of odors against background. J Neurophysiol 95(3):1888–1896.

136. Barnes DC, Hofacer RD, Zaman AR, Rennaker RL, Wilson DA (2008) Olfactory perceptual stability and discrimination. Nat Neurosci 11(12):1378–1380.

137. Onoda N, Sugai T, Yoshimura H (2005) Odor-intensity coding in the anterior piriform cortex. Chem Senses 30 Suppl 1:i162-3.

138. Bolding KA, Franks KM (2017) Complementary codes for odor identity and intensity in olfactory cortex. Elife 6:e22630.

139. Roland B, Deneux T, Franks KM, Bathellier B, Fleischmann A (2017) Odor identity coding by distributed ensembles of neurons in the mouse olfactory cortex. Elife 6:e26337.

140. Li W, Howard JD, Parrish TB, Gottfried JA (2008) Aversive learning enhances perceptual and cortical discrimination of indiscriminable odor cues. Science (80- ) 319(5871):1842–1845.

141. Wilson DA, Pham TC, Sullivan RM (1994) Norepinephrine and posttraining memory consolidation in neonatal rats. Behav Neurosci 108(6):1053–1058.

142. Mazo C, Grimaud J, Shima Y, Murthy VN, Lau CG (2017) Distinct projection patterns of different classes of layer 2 principal neurons in the olfactory cortex. Sci Rep 7(1):8282.

143. Haberly LB, Price JL (1978) Association and commissural fiber systems of the olfactory cortex of the rat. I. systems originating in the piriform cortex and adjacent areas. J Comp Neurol 178(4):711–740.

113

144. Chen C-FF, et al. (2014) Nonsensory target-dependent organization of piriform cortex. Proc Natl Acad Sci 111(47):16931–16936.

145. Illig KR (2005) Projections from orbitofrontal cortex to anterior piriform cortex in the rat suggest a role in olfactory information processing. J Comp Neurol 488(2):224–231.

146. Shipley MT, Ennis M (1996) Functional organization of olfactory system. J Neurobiol 30(1):123–176.

147. Schwob JE, Price JL (1984) The development of lamination of afferent fibers to the olfactory cortex in rats, with additional observations in the adult. J Comp Neurol 223(2):203–222.

148. Schwob JE, Price JL (1984) The development of axonal connections in the central olfactory system of rats. J Comp Neurol 223(2):177–202.

149. Sadrian B, Wilson DA (2015) Optogenetic Stimulation of Lateral Amygdala Input to Posterior Piriform Cortex Modulates Single-Unit and Ensemble Odor Processing. Front Neural Circuits 9:81.

150. Millhouse OE, Heimer L (1984) Cell configurations in the olfactory tubercle of the rat. J Comp Neurol 228(4):571–597.

151. Xiong A, Wesson DW (2016) Illustrated Review of the Ventral Striatum’s Olfactory Tubercle. Chem Senses 41(7):549–555.

152. Ribak CE, Fallon JH (1982) The island of Calleja complex of rat basal forebrain. I. Light and electron microscopic observations. J Comp Neurol 205(3):207–218.

153. Wesson DW, Wilson DA (2011) Sniffing out the contributions of the olfactory tubercle to the sense of smell: hedonics, sensory integration, and more? Neurosci Biobehav Rev 35:655–668.

154. Fallon JH, Riley JN, Sipe JC, Moore RY (1978) The islands of Calleja: organization and connections. J Comp Neurol 181:375–395.

155. Millhouse OE (1987) Granule cells of the olfactory tubercle and the question of the islands of calleja. J Comp Neurol 265(1):1–24.

156. Ikemoto S (2007) Dopamine reward circuitry: Two projection systems from the ventral midbrain to the nucleus accumbens-olfactory tubercle complex. Brain Res Rev 56(1):27–78.

157. Surmeier DJ, Day M, Gertler T, Chan S, Shen W (2010) Chapter 6 – D1 and D2 Dopamine Receptor Modulation of Glutamatergic Signaling in Striatal Medium Spiny Neurons. Handb Behav Neurosci 20:113–132.

114

158. Alheid GF, Heimer L (1988) New perspectives in basal forebrain organization of special relevance for neuropsychiatric disorders: the striatopallidal, amygdaloid, and corticopetal components of substantia innominata. Neuroscience 27(1):1–39.

159. Taverna S, Ilijic E, Surmeier DJ (2008) Recurrent Collateral Connections of Striatal Medium Spiny Neurons Are Disrupted in Models of Parkinson’s Disease. J Neurosci 28(21):5504–5512.

160. Chiang E, Strowbridge BW (2007) Diversity of neural signals mediated by multiple, burst-firing mechanisms in rat olfactory tubercle neurons. J Neurophysiol 98(5):2716–2728.

161. Heimer L, Wilson RD (1975) Golgi Centennial Symposium ed Santini M (Raven Press).

162. Heimer L, Zaborszky L, Zahm DS, Alheid GF (1987) The ventral striatopallidothalamic projection: I. The striatopallidal link originating in the striatal parts of the olfactory tubercle. J Comp Neurol 255(4):571–591.

163. Gerfen CR, et al. (1990) D1 and D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal neurons. Science (80- ) 250(4986):1429–1432.

164. Tian X, Kai L, Hockberger PE, Wokosin DL, Surmeier DJ (2010) MEF-2 regulates activity-dependent spine loss in striatopallidal medium spiny neurons. Mol Cell Neurosci 44(1):94–108.

165. Smith RJ, Lobo MK, Spencer S, Kalivas PW (2013) Cocaine-induced adaptations in D1 and D2 accumbens projection neurons (a dichotomy not necessarily synonymous with direct and indirect pathways). Curr Opin Neurobiol 23(4):546– 552.

166. Luskin MB, Price JL (1982) The distribution of axon collaterals from the olfactory bulb and the nucleus of the horizontal limb of the diagonal band to the olfactory cortex, demonstrated by double retrograde labeling techniques. J Comp Neurol 209(3):249–263.

167. White LE (1965) Olfactory bulb projections of the rat. Anat Rec 152:465–479.

168. Ubeda-Bañona I, et al. (2007) Projections from the posterolateral olfactory amygdala to the ventral striatum: neural basis for reinforcing properties of chemical stimuli. BMC Neurosci 8(1):103.

169. Zahm DS, Heimer L (1993) Specificity in the efferent projections of the nucleus accumbens in the rat: comparison of the rostral pole projection patterns with those of the core and shell. J Comp Neurol 327(2):220–232.

170. Fallon JH (1983) The islands of Calleja complex of rat basal forebrain, II.

115

Connections of medium and large sized cells. Brain Res Bull 10:179–199.

171. Vertes RP, Hoover WB (2008) Projections of the paraventricular and paratenial nuclei of the dorsal midline thalamus in the rat. J Comp Neurol 508(2):212–237.

172. Giessel AJ, Datta SR (2014) Olfactory maps, circuits and computations. Curr Opin Neurobiol 24(0):120–132.

173. Zhang Z, et al. (2017) Whole-Brain Mapping of the Inputs and Outputs of the Medial Part of the Olfactory Tubercle. Front Neural Circuits 11:52.

174. Ikemoto S (2003) Involvement of the Olfactory Tubercle in Cocaine Reward: Intracranial Self-Administration Studies. J Neurosci 23(28):9305–9311.

175. FitzGerald BJ, Richardson K, Wesson DW (2014) Olfactory tubercle stimulation alters odor preference behavior and recruits forebrain reward and motivational centers. Front Behav Neurosci 8. doi:10.3389/fnbeh.2014.00081.

176. Gadziola MA, Wesson DW (2016) The Neural Representation of Goal-Directed Actions and Outcomes in the Ventral Striatum’s Olfactory Tubercle. J Neurosci 36(2):548–560.

177. Hagamen TC, Greeley HP, Hagamen WD, Reeves AG (1977) Behavioral Asymmetries Following Olfactory Tubercle Lesions in Cats. Brain Behav Evol 14(4):241–250.

178. Koob GF, Riley SJ, Smith SC, Robbins TW (1978) Effects of 6-hydroxydopamine lesions of the nucleus accumbens septi and olfactory tubercle on feeding, locomotor activity, and amphetamine anorexia in the rat. J Comp Physiol Psychol 92(5):917–927.

179. Hitt JC, Bryon DM, Modianos DT (1973) Effects of rostral medial forebrain bundle and olfactory tubercle lesions upon sexual behavior of male rats. J Comp Physiol Psychol 82(1):30–36.

180. Gervais G (1979) Unilateral lesions of the olfactory tubercle modifying general arousal effects in the rat olfactory bulb. Electroencephalogr Clin Neurophysiol 46(6):665–674.

181. Agustín-Pavón C, Martínez-García F, Lanuza E (2014) Focal lesions within the ventral striato-pallidum abolish attraction for male chemosignals in female mice. Behav Brain Res 259(0):292–296.

182. Menna-Barreto LS, Cipolla-Neto J, Marques MD (1983) Sexual behavior elicited in cage-mates of olfactory tubercle stimulated rats. Physiol Behav 31(4):565–567.

183. DiBenedictis BT, Olugbemi AO, Baum MJ, Cherry JA (2015) DREADD-Induced Silencing of the Medial Olfactory Tubercle Disrupts the Preference of Female

116

Mice for Opposite-Sex Chemosignals. eNeuro 2(5):1–16.

184. Shin R, Cao J, Webb SM, Ikemoto S (2010) Amphetamine Administration into the Ventral Striatum Facilitates Behavioral Interaction with Unconditioned Visual Signals in Rats. PLoS One 5(1):e8741.

185. Gadziola MA, Tylicki KA, Christian DL, Wesson DW (2015) The Olfactory Tubercle Encodes Odor Valence in Behaving Mice. J Neurosci 35(11):4515– 4527.

186. Rampin O, Bellier C, Maurin Y (2012) Electrophysiological responses of rat olfactory tubercle neurons to biologically relevant odours. Eur J Neurosci 35(1):97–105.

187. Murakami M, Kashiwadani H, Kirino Y, Mori K (2005) State-dependent sensory gating in olfactory cortex. Neuron 46:285–296.

188. Murata K, Kanno M, Ieki N, Mori K, Yamaguchi M (2015) Mapping of Learned Odor-Induced Motivated Behaviors in the Mouse Olfactory Tubercle. J Neurosci 35(29):10581–10599.

189. Sherman SM (2012) Thalamocortical interactions. Curr Opin Neurobiol 22(4):575– 579.

190. Shepherd GM, Chen WR, Greer CA (2004) Olfactory bulb. The Synaptic Organization of the Brain, ed Shepherd GM (Oxford Univ Press, Oxford), pp 165– 216. 5th Ed.

191. White KA, et al. (2017) A cortical pathway modulates sensory input into the olfactory striatum. bioRxiv. Available at: http://biorxiv.org/content/early/2017/12/16/235291.abstract.

192. Linster C, Menon A V, Singh CY, Wilson DA (2009) Odor-specific habituation arises from interaction of afferent synaptic adaptation and intrinsic synaptic potentiation in olfactory cortex. Learn Mem 16:452–459.

193. Hasselmo ME, Wilson MA, Anderson BP, Bower JM (1990) Associative memory function in piriform (olfactory) cortex: computational modeling and neuropharmacology. Cold Spring Harb Symp Quant Biol 55:599–610.

194. Barkai E, Bergman RE, Horwitz G, Hasselmo ME (1994) Modulation of associative memory function in a biophysical simulation of rat piriform cortex. J Neurophysiol 72(2):659–677.

195. Heimer L, Switzer RD, Van Hoesen GW (1982) Ventral striatum and ventral pallidum: Components of the motor system? Trends Neurosci 5:83–87.

196. Meredith GE (1999) The synaptic framework for chemical signaling in nucleus

117

accumbens. Ann N Y Acad Sci 877(1 ADVANCING FRO):140–156.

197. Heimer L, Wilson RD (1975) The subcortical projections of the allocortex: similarities in the neural associations of the hippocampus, the piriform cortex, and the neocortex ed Santini M (Raven, New York).

198. Carriero G, Uva L, Gnatkovsky V, de Curtis M (2009) Distribution of the Olfactory Fiber Input Into the Olfactory Tubercle of the In Vitro Isolated Guinea Pig Brain. J Neurophysiol 101(3):1613–1619.

199. Carmichael JE, Gmaz JM, van der Meer MAA (2017) Gamma Oscillations in the Rat Ventral Striatum Originate in the Piriform Cortex. J Neurosci 37(33):7962– 7974.

200. Shuen JA, Chen M, Gloss B, Calakos N (2008) Drd1a-tdTomato BAC Transgenic Mice for Simultaneous Visualization of Medium Spiny Neurons in the Direct and Indirect Pathways of the Basal Ganglia. J Neurosci 28(11):2681–2685.

201. Gong S, et al. (2003) A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature 425(6961):917–925.

202. Anikeeva P, et al. (2011) Optetrode : a multichannel readout for optogenetic control in freely moving mice. Nat Neurosci 15(1):163–170.

203. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Meth 9(7):671–675.

204. Paxinos G, Franklin K (2000) The Mouse Brain in Stereotaxic Coordinates (Academic Press, San Diego). 2nd Ed.

205. Cepeda C, et al. (2008) Differential electrophysiological properties of dopamine D1 and D2 receptor-containing striatal medium-sized spiny neurons. Eur J Neurosci 27(3):671–682.

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

207. Illig KR, Haberly LB (2003) Odor-evoked activity is spatially distributed in piriform cortex. J Comp Neurol 457(4):361–373.

208. Luna VM, Pettit DL (2010) Asymmetric rostro-caudal inhibition in the primary olfactory cortex. Nat Neurosci 13(5):533–535.

209. Haberly LB (1973) Unitary analysis of opossum prepyriform cortex. J Neurophysiol 36(4):762–774.

210. Price JL (1973) An autoradiographic study of complementary laminar patterns of

118

termination of afferent fibers to the olfactory cortex. J Comp Neurol 150:87–108.

211. Diodato A, et al. (2016) Molecular signatures of neural connectivity in the olfactory cortex. 7:12238.

212. Large AM, Vogler NW, Canto MB, Schick P, Oswald A-MM (2017) Opposing spatial gradients of inhibition and neural activity in mouse olfactory cortex. bioRxiv. Available at: http://biorxiv.org/content/early/2017/07/31/152975.abstract.

213. Wickersham IR, et al. (2007) Monosynaptic Restriction of Transsynaptic Tracing from Single, Genetically Targeted Neurons. Neuron 53(5):639–647.

214. Heimer L (1968) Synaptic distribution of centripedal and centrifugal nerve fibers in the olfactory system in the rat. J Anat 103:413–432.

215. Brunjes PC, Illig KR, Meyer EA (2005) A field guide to the anterior olfactory nucleus (cortex). Brain Res Brain Res Rev 50(2):305–335.

216. Howard JD, Kahnt T, Gottfried JA (2016) Converging prefrontal pathways support associative and perceptual features of conditioned stimuli. Nat Commun 11546.

217. Rothermel M, Wachowiak M (2014) Functional imaging of cortical feedback projections to the olfactory bulb. Front Neural Circuits 8. doi:10.3389/fncir.2014.00073.

218. Schneider SP, Scott JW (1983) Orthodromic response properties of rat olfactory bulb mitral and tufted cells correlate with their projection patterns. J Neurophysiol 50(2):358–378.

219. Li W, Luxenberg E, Parrish T, Gottfried JA (2006) Learning to smell the roses: experience-dependent neural plasticity in human piriform and orbitofrontal cortices. Neuron 52(6):1097–1108.

220. Calabresi P, Picconi B, Tozzi A, Ghiglieri V, Di Filippo M (2014) Direct and indirect pathways of basal ganglia: a critical reappraisal. Nat Neurosci 17(8):1022–1030.

221. Ikemoto S, Bonci A (2014) Neurocircuitry of drug reward. Neuropharmacology 76, Part B(0):329–341.

222. Kupchik YM, et al. (2015) Coding the direct/indirect pathways by D1 and D2 receptors is not valid for accumbens projections. Nat Neurosci 18(9):1230–1232.

223. Wieland S, et al. (2015) Phasic Dopamine Modifies Sensory-Driven Output of Striatal Neurons through Synaptic Plasticity. J Neurosci 35(27):9946–9956.

224. Wang X, Zhang C, Szábo G, Sun Q-Q (2013) Distribution of CaMKIIα expression in the brain in vivo, studied by CaMKIIα-GFP mice. Brain Res 1518:9–25.

119

225. Ketchum KL, Haberly LB (1993) Membrane currents evoked by afferent fiber stimulation in rat piriform cortex. I. Current source-density analysis. J Neurophysiol 69(1):248–260.

226. Yang J, et al. (2004) Quantitative analysis of axon collaterals of single neurons in layer IIa of the piriform cortex of the guinea pig. J Comp Neurol 473(1):30–42.

227. Berndt A, et al. (2011) High-efficiency channelrhodopsins for fast neuronal stimulation at low light levels. Proc Natl Acad Sci 108(18):7595–7600.

228. Han X, Boyden ES (2007) Multiple-color optical activation, silencing, and desynchronization of neural activity, with single-spike temporal resolution. PLoS One 2:e299.

229. Pologruto TA, Yasuda R, Svoboda K (2004) Monitoring Neural Activity and [Ca2+] with Genetically Encoded Ca2+ Indicators. J Neurosci 24(43):9572–9579.

230. Helmchen F, Fee MS, Tank DW, Denk W (2001) A miniature head-mounted two- photon microscope. high-resolution brain imaging in freely moving animals. Neuron 31(6):903–912.

231. Kravitz A V, Tye LD, Kreitzer AC (2012) Distinct roles for direct and indirect pathway striatal neurons in reinforcement. Nat Neurosci 15(6):816–818.

232. Gottfried JA, O’Doherty J, Dolan RJ (2002) Appetitive and aversive olfactory learning in humans studied using event-related functional magnetic resonance imaging. J Neurosci 22(24):10829–10837.

233. O’Doherty J (2004) Dissociable roles of ventral and dorsal striatum in instrumental conditioning. Science (80- ) 304:452–454.

234. Vicente AM, Galvão-Ferreira P, Tecuapetla F, Costa RM (2016) Direct and indirect dorsolateral striatum pathways reinforce different action strategies. Curr Biol 26(7):R267–R269.

235. Shepherd GM (1993) Principles of specificity and redundancy underlying the organization of the olfactory system. Microsc Res Tech 24(2):106–112.

236. Meyer G, Wahle P (1986) The olfactory tubercle of the cat. I. Morphological components. Exp Brain Res 62(3):515–527.

237. Haber SN (2016) Corticostriatal circuitry. Dialogues Clin Neurosci 18(1):7–21.

238. Shepherd GMG (2013) Corticostriatal connectivity and its role in disease. Nat Rev Neurosci 14(4):278–291.

239. Wall NR, De La Parra M, Callaway EM, Kreitzer AC (2013) Differential Innervation of Direct- and Indirect-Pathway Striatal Projection Neurons. Neuron 79(2):347–

120

360.

240. Ramanathan S, Hanley JJ, Deniau J-M, Bolam JP (2002) Synaptic Convergence of Motor and Somatosensory Cortical Afferents onto GABAergic Interneurons in the Rat Striatum. J Neurosci 22(18):8158 LP-8169.

241. W. S (1998) Predictive reward signal of DA neurons. J Neurophysiol 80:1.

242. Del-Fava F, Hasue RH, Ferreira JGP, Shammah-Lagnado SJ (2007) Efferent connections of the rostral linear nucleus of the ventral tegmental area in the rat. Neuroscience 145(3):1059–1076.

243. Mooney K, Inokuchi A, Snow JJ, Kimmelman C (1987) Projections from the ventral tegmental area to the olfactory tubercle in the rat. Otolaryngol Head Neck Surg 96(2):151–7.

121

BIOGRAPHICAL SKETCH

Kate A. White (maiden: Tylicki) was born in Cleveland, OH. She graduated high school in 2009 and attended Baldwin Wallace University for her undergraduate studies from 2009-2013. She graduated summa cum laude with her Bachelor of Science degree in neuroscience and psychology in May 2013 under the guidance of Drs. G. Andrew

Mickley and Christopher Turner. Her undergraduate thesis was entitled “The effects of a protein kinase R inhibitor in an animal model of Alzheimer’s beta-amyloidosis,” which was funded through a university grant, and given a regional conference award.

In 2013, she entered the Department of Biology PhD program at Case Western

Reserve University under the neurobiology and neuromechanics concentration. She joined the lab of Dr. Dan Wesson in 2014, where she researched various questions related to olfactory cortical coding. She moved to the University of Florida with the

Wesson lab in 2017 and entered into the PhD program for the Department of

Pharmacology and Therapeutics. During her graduate training, she authored two publications, one in which she was a first author, and the other a second author. She was awarded a Ruth L. Kirschstein National Research Service Award Pre-Doctoral

Fellowship from the National Institute on Deafness and Other Communication Disorders for her dissertation research proposal. She has presented her research both at university symposia and national conferences, including Society for Neuroscience and

Association for Chemoreception Sciences. Kate defended her dissertation in April 2018 and receive her PhD in pharmacology and therapeutics. She will continue to pursue postdoctoral training in systems-level chemosensory neuroscience.

122