Balancing excitation and inhibition within olfactory bulb glomeruli

BALANCING EXCITATION AND INHIBITION WITHIN OLFACTORY BULB GLOMERULI

by

JOSEPH DONALD ZAK

B.S., University of Michigan, Ann Arbor, 2008

A thesis submitted to the


Faculty of the Graduate School of the University of Colorado in partial fulfillment

of the requirements for the degree of Doctor of Philosophy

Neuroscience Program
 2015

This thesis for the Doctor of Philosophy degree by Joseph Donald Zak

has been approved for the Neuroscience Program

by

Sukumar Vijayaraghavan, Chair 
 Nathan Schoppa, Advisor

Achim Klug Catherine Proenza


Diego Restrepo Angeles Ribera

Zak, Joseph Donald (Ph.D., Neuroscience)


Balancing Excitation and Inhibition Within Olfactory Bulb Glomeruli Thesis directed by Professor Nathan Schoppa.

ABSTRACT

The mammalian main olfactory bulb is organized into discrete functional units known as glomeruli, where first order sensory processing occurs. In addition to its modular structure, the olfactory bulb also exhibits extensive laminar organization. A largely unanswered question in the field is whether olfactory bulb glomeruli have the propensity to facilitate signal discrimination within a single glomerular unit, or if such a mechanism arises through layer-specific interactions between many glomeruli, each corresponding to unique sensory input. Here, I used patch-clamp recordings and calcium imaging techniques to assess the ability of individual glomeruli to serve as sensory gating elements, independent of the activity of neighboring glomeruli. I also considered the ability of the intrinsic circuitry within glomeruli to promote disinhibition of excitatory elements following strong incoming signals.

I first obtained evidence that GABAergic periglomerular (PG) cells that surround olfactory bulb glomeruli are more sensitive to excitatory input than glutamatergic

external tufted (ET) cells that also surround glomeruli. However, I also found that ET cells received substantially more excitatory input from olfactory sensory neurons, potentially offsetting their decreased sensitivity to excitation. To resolve this potential

conflict I undertook a calcium imaging-based population analysis and determined that at the weakest levels of olfactory input, there are substantially more active PG cells than ET cells, thus glomeruli favor inhibition over excitation following weak inputs. Furthermore, I demonstrated that inhibition from PG cells controls excitation of ET cells and

subsequent activity of output mitral cells.

Consistent with a glomerular filtering mechanism I next provided evidence that following strong sensory inputs, endogenous glutamate acts at group II metabotropic glutamate receptors (mGluRs) to down-regulate inhibition within glomeruli. My studies indicate that in well activated glomeruli, glutamate released from ET cells acts

intraglomerularly at group II mGluRs on the dendrites of PG cells, subsequently reducing their cellular excitability and neurotransmitter release. The actions of group II mGluRs on PG cells can contribute to the mechanistic bases of dynamic range expansion in terms of glomerular output by allowing for greater spike output from mitral cells following strong sensory inputs.

The form and content of this abstract are approved. I recommend its publication. Approved: Nathan Schoppa

ACKNOWLEDGMENTS

Numerous individuals have provided all means of intellectual and technical support during the completion of this thesis. Foremost, I want to thank my advisor, Dr. Nathan Schoppa, for his guidance throughout my graduate training. Working with Nathan has been nothing short of a formative experience in my early scientific career. I have been very fortunate to work with him over the past few years, and I feel that a great portion of my development as a scientist has been due to his genuine interest in my success.

Current and past members of the Schoppa lab have also been invaluable to my scientific development. Past members Drs. Jennifer Whitesell and David Sheridan were great sources of technical help and expertise when I joined the lab and were very helpful as I learned patch-clamp electrophysiology and the anatomy of the olfactory system. Dr. Frederic Pouille, Dr. Jennifer Bourne and Britni Sanchez have provided many thoughtful comments on this work and have made the Schoppa lab an enjoyable place to work. I appreciate their insight and suggestions. I would also like to thank Dr. David Gire. Although our time in the Schoppa lab never overlapped, a large portion of my thesis is built upon the data David generated as a graduate student and would not have been possible without his previous studies.

I also acknowledge my thesis committee members, Drs. Sukumar Vijayaraghavan, Achim Klug, Cathy Proenza, Diego Restrepo and Angie Ribera. Their guidance has been instrumental to not only the development of this project, but also the way I think about my work. I thank them for making time for many committee meetings and their insightful

comments. I specifically thank Diego and Sukumar for their many discussions on the function of the olfactory system and guidance outside of this thesis.

The greater Neuroscience Program at the University of Colorado has been an outstanding community to be a part of. We are fortunate to have such wonderful administrative support and strong leadership. The students in the program have been a constant source of camaraderie and support, but perhaps most importantly, intellectual challenge. I have been fortunate to form many great friendships that I look forward to following throughout my scientific career. Working and studying within the Department of Physiology & Biophysics has been an incredibly rewarding experience due to both the collegial atmosphere among students and faculty, but also the breadth of expertise in systems neuroscience.

I owe a great deal of gratitude to Dr. Zhong-wei Zhang at the Jackson Laboratory in Bar Harbor, Maine. During my time in Dr. Zhang’s lab, not only did I develop an interest in electrophysiology, but I also learned rigorous work ethic and persistence. Working with Dr. Zhang instilled in me an enthusiasm and passion for my studies that I hope to carry throughout the rest of my career.

A large portion of my graduate studies were supported by a National Research Service Award from the National Institute on Deafness and Other Communication Disorders (F31 DC013480). I thank the Institute for their support and recognition.

On a more personal note, I would like to thank my family, especially my parents for pushing me to challenge myself in all aspects of life. They have been a tremendous source of support, love and advice. Finally, and most importantly, I thank my wife

Michelle for her constant support. She has been patiently with me throughout this entire process. None of this could have been accomplished without her. Words cannot express my gratitude and admiration for her unwavering support

TABLE OF CONTENTS

CHAPTER

I. INTRODUCTION ... 1

Research in the olfactory system ... 1

Anatomy and structure of the olfactory system ... 3

The peripheral olfactory system ...3

The central olfactory system and beyond ...6

Cellular constituents and neural circuitry of the olfactory system ...8

Intraglomerular circuits ...8

Interglomerular circuits ...13

Cortical and subcortical input to the olfactory bulb ...17

Organization of olfactory bulb glomeruli ...18

Contrast enhancement in the olfactory system ...20

Overview of present work ...28

II. GABAERGIC SIGNAL FILTERING AT OLFACTORY BULB GLOMERULI ...31

Introduction ...31

Materials and methods ...34

Animals ...34

Slice preparation ...34

Electrophysiology ...35

Data analysis ...38

Results ...39

PG cells are more sensative to excitatory iput than ET cells ...39

ET cells receive larger EPSCs for a given level of OSN activity ...45

Fura2, AM fluorescence is correlated with spike output from PG and ET Cells ...47

GABAergic cells in the glomerulus are preferentially activated with low-level sensory input ...55

MC output is correlated with the activity of VGAT- neurons ...60

Inhibition is the dominant synaptic input onto ET cells following weak olfactory input ...65

Group II mGluRs contribute to an inhibitory plateau within glomeruli ...70

Discussion ...75

Activity of GABAergic neurons at a glomerulus selectively filters weak sensory input ...75

Functional implications for odor processing ...77

III. METABOTROPIC GLUTAMATE RECEPTORS PROMOTE DISINHIBITION OF OLFACTORY BULB GLOMERULI THAT SCALES WITH INPUT STRENGTH ...80

Introduction ...80

Materials and methods ...82

Animals ...82

Slice preparation ...83

Data analysis ...88

Results ...90

Group II mGluR activation reduces GABAergic inhibition onto tufted cells and mitral cells ...90

Activation of Group II mGluRs reduces GABA release from PG cells ...95

mGluR-mediated disinhibition is driven by glutamate released from intrinsic bulbar neurons ...98

mGluR activation selectively increases ET cell excitation when OSNs strongly activate glomeruli ...108

MC excitation is enhanced by activation of Group II mGluRs ....113

Discussion ...118

mGluR-mediated disinhibition of olfactory bulb glomeruli ...118

Condition-sepcific enhancement of glomerular output ...121

IV. CONCLUSION AND FUTURE DIRECTIONS ...124

Conclusion ... 124

Future studies ... 128

In vitro based optogenetic approaches ... 128

In vivo based optogenetic approaches ... 130

Contribution of neurogenesis in the olfactory bulb glomerular layer ... 134

Neurologic disease and the olfactory system ... 139

Closing remarks ... 143

LIST OF FIGURES

FIGURE

1.1 Early anatomical description of the olfactory bulb circuity ...2

1.2 Illustration of the olfactory system ...7

1.3 Glomerular circuitry of the olfactory bulb ...10

1.4 Intraglomerular and interglomerular GABAergic neurons in the olfactory bulb ...14

1.5 Interglomerular inhibition and contrast enhancement in the olfactory system ...22

1.6 Periglomerular cell mediated odor evoked inhibition of mitral cells ...24

1.7 Schematic of intraglomerular filtering based on the intrinsic properties of juxtaglomerular neurons ...26

1.8 Computational evidence for intraglomerular signal filtering ...27

2.1 PG cells spike at lower levels of OSN input than ET cells ...40

2.2 Dual PG-ET cell recordings reveal ET cells receive more OSN input at a given level of OSN activity ...47

2.3 Fura2, AM reliably detects single action potentials in PG and ET cells ...52

2.4 GABAergic neurons in the glomerulus are preferentially activated with weak inputs ...56

2.5 MC activity is correlated with the number of active VGAT- neurons at a glomerulus ...61

2.6 Inhibition does not scale with excitation within glomeruli ...67

2.7 Group II mGluRs limit inhibition of ET cells ...72

3.1 Activation of group II mGluRs decreases polysynaptic IPSCs in ET cells ...93

3.3 Activation of Group II mGluRs by endogenous glutamate reduces GABA release from PG cells ...100 3.4 A conditioning stimulus applied to ET cells reduces GABA release from PG

cells ...104 3.5 OSNs are not a major source of glutamate driving activation of group II

mGluRs ...107 3.6 DCG-IV enhances ET cell excitation in response to OSN input ...111 3.7 mGluR-mediated enhancement of ET cell excitation is dependent on the

level of neural activation ...114 3.8 Activation of group II mGluRs enhances the MC response to OSN input ...117 4.1 Model of odor processing ...127

LIST OF ABBREVIATIONS

AON anterior olfactory nucleus

aPC anterior piriform cortex

ATP adenosine triphosphate

cAMP cyclic adenosine monophosphate

BrdU 5-bromo-2’-deoxyuridine CGP55845 (2S)-3-[[(1S)-1-(3,4- Dichlorophenyl-)ethyl]amino-2- hydroxypropyl](phenylmethyl)phosphinic acid hydrochloride ChR2 channelrhodopsin-2 DCG-IV (1R,2R)-3-[(1S)-1-amino-2-hydroxy-2- oxoethyl]cyclopropane-1,2-dicarboxylic acid

DIC differential interference contrast

DHPG (S)-3,5- dihydroxyphenylglycine

DL-AP5 DL-2-amino-5-phosphonopentanoic acid

EPL external plexiform layer

EPSC excitatory postsynaptic current

EPSP excitatory postsynaptic potential

ET external tufted [cell]

Fura2 Acetoxymethyl 2-[5-[bis[(acetoxymethoxy-oxo-

methyl)methyl]amino]-4-[2-[2- [bis[(acetoxymethoxy-

oxo- methyl)methyl]amino]-5-methyl- phenoxy]ethoxy]benzofuran-2-yl]oxazole-5- carboxylate

GABA γ-︎aminobutyric acid

gabazine 4-[6-imino-3-(4-methoxyphenyl)pyridazin-1-yl]

butanoic acid hydrobromide

GAD65 glutamate decarboxylase, 65 kDa isoform

GC granule cell

GCaMP genetically encoded calcium indicator

GnRH gondadotropin-releasing hormone

GTP guanosine triphosphate

sIPSC spontaneous inhibitory postsynaptic current

IPSC inhibitory postsynaptic current

IPSP inhibitory postsynaptic potential

JG juxtaglomerular [cell]

L-AP4 L-(+)-2-amino-4-phosphonobutyric acid

LCA loose cell-attached

LLD long-lasting depolarization

LY341495 2-[(1S,2S)-2-carboxycyclopropyl]-3-(9H- xanthen-9-

yl)-D-alanine

MAP-4 (S)-2-amino-2-methyl-4-phosphonobutanoic acid

MC mitral cell

mGluR metabotropic glutamate receptor

mPG PG cell receiving direct monosynaptic input from

OSN axon terminals

NBQX 2,3-dioxo-6-nitro-1,2,3,4-tetrahydr-

obenzo[f]quinoxaline-7- sulfonamide

NpHR halorhodopsin

OB olfactory bulb

ON olfactory nerve

ONL olfactory nerve layer

OR odorant receptor

OSN olfactory sensory neuron

pPC posterior piriform cortex

PG periglomerular [cell]

pPG PG cell receiving polysynaptic excitation through external tufted cells

ROI region of interest

SA short axon [cell]

TC tufted cell

VGAT vesicular γ-︎aminobutyric acid transporter

VNO vomernasal organ

CHAPTER I INTRODUCTION

Research in the olfactory system

The olfactory system has historically been a target of scientific observation from a neuroanatomical perspective at the cellular level (Cajal, 1891; Reviewed by Gire et al., 2013). Due to the discrete and confined structures of the olfactory bulb, these early studies led to functional inferences on the direction of information flow between neural populations (Figure 1.1). However, only more recently have the circuits comprising the olfactory bulb been more throughly dissected on a neurophysiological level. Strikingly, these studies have revealed a diversity of signaling modalities beyond classic

axodendritic neurotransmitter release, including dendrodendritic signaling (Pinching and Powell, 1971; Schoppa and Urban 2003; Panzanelli et al., 2007; Gire and Schoppa 2009), electrical coupling (Schoppa and Westbrook, 2002; Kosaka and Kosaka, 2004; Christie et al., 2005; Pimentel and Margrie, 2008) and signaling through diffusion-based

mechanisms (Gire et al., 2012). Such modalities would be impossible to infer from neuroanatomical observations alone. Furthermore, the biophysical and biochemical heterogeneity of neuronal populations once assumed to be ubiquitous has revealed several distinct subclasses of both excitatory and inhibitory neurons. The expansion of defined neural subtypes within the olfactory system allows for more complex signal processing. As such, these recent advances have called for a reconsideration of the

Figure 1.1. Early anatomical description of the olfactory bulb circuity.

An example drawing by Santiago Ramon y Cajal of olfactory bulb neurons impregnated using the Golgi Method (Cajal, 1891; Gire et al., 2013). Olfactory sensory neurons (A, bottom) are seen projecting and terminating at individual glomeruli. Cajal correctly determined that each mitral cell (C) projects to only a single glomerulus and possess an axon that, in turn, projects to cortical structures (E). From these observations and

drawings, Cajal was able to correctly discern the bi-directionality of information flow in the olfactory system, both to and from cortex. Cajal also notes the presence of tufted cells and granule cells in his illustration.

mechanisms by which information processing occurs within the olfactory system and in particular, the glomerular layer of the olfactory bulb.

This body of work I present here seeks to further our understanding of the structure-function relationships of the olfactory bulb glomerulus as well as cellular communication and modulation within it. I will first describe the nature of input to the olfactory bulb, followed by an overview of the synaptic connectivity within the bulb and more specifically, the glomerulus. I will then discuss the stimulus-based organization of the glomeruli as it pertains to the phenomenon of contrast enhancement and odor discrimination. At the conclusion of this chapter, I will provide a brief overview of the aims of this thesis as a means of introduction to the work that will follow in subsequent chapters.

Anatomy and structure of the olfactory system

The peripheral olfactory system

The mammalian olfactory system is a unique and well-conserved sensory

modality. Within the rodent genome there are 1,300 olfactory receptor genes that encode for approximately 1,100 unique receptor protein subtypes (Young et al., 2002). Although the human genome contains a similar number of genes, the actual number of receptor subtypes encoded is substantially lower (Glusman et al., 2000; Zozulya et al., 2001). However, despite the reduced receptor subtype diversity observed in humans, the number of discriminable odors to the human nose may exceed one trillion (Bushdid et al., 2014).

At the periphery of the olfactory system, odorant receptors (ORs) located on the cilia of olfactory sensory neurons (OSNs) bind monomolecular features of volatilized odorant molecules (Figure 1.2). Each OSN expresses only a single receptor protein (Ressler et al., 1994) that is most selective to a single monomolecular feature of an odorant molecule, known as a functional group. However, ORs possess the capability to interact with a range of molecules within a functional group across a spectrum of receptor-ligand binding affinities (Gaillard et al., 2002). The exact mechanisms of odor receptor-ligand binding are beyond the scope of this work, but for a review see Fleischer et al., 2009. The range of OR binding affinities across odor functional groups results in graded levels of OSN excitation; thus, OR activity indicates the presence of a functional group on an odorant molecule rather than a perceptible odor stimulus.

OSNs are bipolar neurons with cell bodies found within a specialized epithelial tissue known as the olfactory epithelium. These neurons radially project their cilia into the nasal mucosa where they encounter odorant molecules drawn in to the mucus via interactions with olfactory binding proteins (Breer et al., 1996; Buck et al., 1996; Shepherd et al., 2004).

Upon binding with their odorant ligands, ORs transduce excitatory signals through a G-protein coupled cascade (Ronnett and Moon, 2002). Odorant molecule binding stimulates Golfactory to increase production of 3’,5’-cyclic adenosine

monophosphate (cAMP, Brunet et al., 1996; Belluscio et al., 1998; Wong et al., 2000; Mombaerts, 2004) that in turn leads to an influx of cations through activation of cyclic nucleotide gated ion channels (Firestein and Warblin, 1987; Lowe et al., 1989; Boekhoff

et al., 1990). This conductance results in OSN membrane depolarization, which triggers output in the form of action potentials (i.e., “spikes”). The generation of action potentials in OSNs is generally phase-locked with the peak of inhalation in the respiratory cycle (Cang and Isaacson, 2003; Margrie and Schaefer, 2003; Carey et al., 2009). Interestingly, in the absence of odorant molecules, some OSNs still produce action potentials locked with inhalation due to mechanical activation of cation channels (Grosmaitre et al., 2007; Connelly et al., 2015), although the frequency of OSN spiking is markedly less than when in the presence of an odorant molecule.

Upon exiting the olfactory epithelium, axons of OSNs course through the cribriform plate and form the most distal layer of the olfactory bulb known as the olfactory nerve (ON) layer (Figure 1.2). The axon fibers of OSNs corresponding to common OR types then coalesce and enter as a single fascicle to modular, OR-specific, neuropils called glomeruli (Ressler et al., 1994; Mombaerts et al., 1996; Mori, 1999). Each glomerulus receives convergent input from approximately 25,000 OSNs expressing a single type of OR (Treloar et al., 2002, Doty et al., 2003; Zou et al., 2009) and as such, operates as a functional unit. Within these structures, first-order olfactory information processing occurs. Each OR corresponds to one or two glomeruli on each olfactory bulb, up to four in total. When OSNs target more than one glomeruli, the glomeruli are usually situated within each olfactory bulb such that one is positioned on the contralateral surface from the other (Mombaerts, et al., 1996).

A division of the olfactory system known as the accessory olfactory system detects pheromone cues in the environment through a specialized organ known as the

vomeronasal organ (VNO). The axons of sensory neurons found within the VNO project to the accessory olfactory bulb, located caudally on the dorsal surface of the main olfactory bulb. It is speculated that the accessory olfactory system provides important social cues and contributes to communication in some species (Doving and Trotier, 1998). While there are apparent similarities between the accessory and main olfactory system, the differences between their synaptic organizations are substantial (Jia et al., 1999). The entirety of the research performed herein examined the main olfactory system of the rat (Rattus norvegicus) and as such, the accessory olfactory system will not be discussed further.

The central olfactory system and beyond

The first site of information processing in the olfactory system occurs in unique structures known as glomeruli. Glomeruli are a dense amalgamation of several different cell types, known collectively as juxtaglomerular (JG) cells, as well as centrifugal axonal projections from cortical and subcortical nuclei. Upon initial processing (discussed later in this chapter), output from glomeruli is directed down the axons of two separate classes of projection neurons known respectively as mitral cells (MCs) and tufted cells (TCs). The axons of MCs and TCs travel through the lateral olfactory tract (LOT) and then collateralize to target both cortical and limbic telencephalic structures including, but not limited to, anterior olfactory nucleus (AON), anterior piriform cortex (aPC), posterior piriform cortex (pPC), enthorhinal cortex and cortical amygdala (Haberly and Price, 1977; Nagayama et al., 2010; Gosh et al., 2011; Miyamichi et al., 2011;

Figure 1.2. Illustration of the olfactory system.

A, Olfactory sensory neurons located within the nasal cavity bind with volatilized odorant molecules drawn into the nasal mucosa via olfactory binding proteins. Each olfactory sensory neuron expresses only a single type of olfactory receptor protein and is most selective to a single functional group of an odorant molecule. B, The axons of olfactory sensory neurons course through the cribriform plate where they terminate at glomeruli (C) found at the surface of the olfactory bulb. Each olfactory sensory neuron expressing a common receptor protein converges on the same glomerulus, thus odor specificity is maintained amongst glomeruli. D, Output from glomeruli is directed down the axons of projection neurons known as mitral and tufted cells where they terminate in structures including the piriform cortex (PirCtx) and anterior olfactory nucleus (AON).

B

C D

Sosulski et al., 2011). The olfactory system is unique among sensory systems because it does not incorporate an obligatory thalamic relay; rather signals are sent directly to the cortex from the olfactory bulb.

Interestingly, MCs and TCs differentially target the above-mentioned higher-order neural structures. TCs primarily target the anterior portions of piriform cortex whereas MC projections are dispersed throughout the entire structure. Conversely, TCs densely target the AON; where in contrast, MC projections are quite sparse (Nagayama et al., 2010; Igarashi et al., 2012). In addition to their differential projections, MCs and TCs exhibit distinct temporal dynamics with respect to their output following sensory input. TCs are generally active before MCs and are neatly phase-locked with the peak of inhalation in the respiratory phase, whereas MCs exhibit more temporally diverse

behavior (Cang and Isaacson 2003; Cury and Uchida, 2010; Carey and Wachowiak, 2011; Fukunaga et al., 2012). Their disparate anatomical projections and physiologic properties indicate that MCs and TCs may serve as two distinct channels of olfactory information.

Cellular constituents and neural circuitry of the olfactory system

Intraglomerular circuits

Each glomerulus is a spherical neuropil receiving dense innervation from numerous cell types, known collectively as juxtaglomerular cells that can be broken down into subgroups based on their cellular morphology and biochemical constituents (Figure 1.3). These neurons incorporate and permutate signals from the olfactory

periphery before directing them toward cortical and other telenchephalic structures via the axons of MCs and TCs.

TCs encompass several subgroups of glutamatergic neurons existing outside of the mitral cell layer. These include deep tufted, middle tufted, superficial-middle tufted and external tufted (ET) cells. These cells are classified with respect to the proximity of their soma to the glomerular layer, with deep tufted cells being most proximal to the mitral cell layer and ET cells being found within the glomerular layer. All subgroups of TCs project an apical dendrite to a single glomerulus, and while several groups also possess lateral dendrites, the subtype most relevant to this dissertation, ET cells, only exhibit dendritic processes within the glomerular neuropil. Most TCs receive direct monosynaptic glutamatergic input from the axon terminals of OSNs and feedforward excitation from other TCs affiliated with the same glomerulus. Due to their uniglomerular affiliation, TCs are only capable of receiving excitatory input from OSNs corresponding to a single odorant molecule.

MC cell bodies exist in a neat and well-defined layer (the mitral cell layer) situated internally to the glomerular and external plexiform layer of the olfactory bulb. Each MC possesses a single primary dendrite that extends past the external plexiform layer, where it then terminates and ramifies within a single glomerulus (Figure 1.3B). There are approximately 20 to 25 MCs that project to a given glomerulus, and the network of MCs projecting to a common glomerulus is electrically coupled via gap junctions allowing for non-chemical communication between sister MCs (Schoppa and Westbrook, 2002; Christie et al., 2005). In addition to their primary dendrite, MCs also

Figure 1.3. Glomerular circuitry of the olfactory bulb.

A, abbreviations: PG, periglomerular cell; ET, external tufted cell. + denotes excitation, - denotes inhibition. Olfactory sensory neuron axon terminals provide glutamatergic excitation to ET cells, which, in turn, excite MCs via extrasynaptic glutamate transients (Gire et al., 2012). OSNs can also excite PG cells, although most PG cells are excited in a disynaptic circuit with ET cells as an intermediary (Shao et al., 2009). PG cells provide GABAergic inhibition onto ET cells, thus limiting overall glomerular excitation. B, Cell-fill images of a PG, ET and mitral cell demonstrating distinct cell size and dendritic morphology. PG cell ET cell Mitral Cell 40 µm 40 µm

→

have lateral dendrites that extend into the external plexiform layer, where they form reciprocal dendrodendritic synapses with granule cells. These connections will be discussed in more detail in the subsequent section.

Traditionally, MCs were thought to receive direct excitation from the axon terminals of OSNs; however, recent work has provided evidence that MCs, at best, receive only very small signals at the their soma resulting from OSN input and primarily receive excitation through TCs acting as as relay neurons (De Saint Jan et al., 2009; Najac et al., 2011; Gire et al., 2012). It should be noted that MCs do indeed form direct synapses with OSN axon terminals, but the resulting excitatory signals are shunted via the electrically coupled compliment of MCs projecting to the same glomerulus, as suggested in gap junction knock-out mouse models (Gire et al., 2012). Furthermore, TCs do not signal directly to MCs through traditional dendrodendritic synapses, but rather transmit excitatory input via an accumulation of diffuse glutamate within the glomerular neuropil (Gire et al., 2012).

A recent study identified a subclass of VGLUT3-expressing ET cells that provide little to no input to MCs and are primarily excited by other TCs (Tatti et al., 2014). In keeping with their lack of MC connectivity, the primary function of this subclass of ET cells appears to be driving suppression of glomerular output neurons. Interestingly, the mechanism of MC suppression occurs through direct co-release of γ-aminobutyric acid (GABA) onto TCs from VGLUT3 ET cells. Co-transmission of glutamate and GABA from TCs has not previously been reported in the literature and could add an additional layer of complexity to glomerular processing.

GABAergic interneurons, known as periglomerular (PG) cells, are also central to this research due to both their location in the glomerular layer and their ability to receive direct input from OSNs as well as feedforward excitation through the OSN-ET-PG pathway (Shao et al., 2009; Figure 1.3). PG cells are the most numerous cell type within the glomerular layer that synthesize GABA (Parrish-Aungst et al., 2007; Puopolo and Belluzzi, 1998; Panzanelli et al., 2007). GABA released from PG cells acts through direct synaptic transmission at GABAA receptors throughout the glomerulus, especially at

synapses on ET cells (Murphy et al., 2005, Gire and Schoppa 2009). However, PG cells can also modulate OSN output through activation of metabotropic GABAB and D2

dopamine receptors near their axon terminals (Nickell et al., 1994; Hsia et al., 1999; Berkowicz and Trombley, 2000; Ennis et al., 2001).

In addition to GABAergic transmission to other neurons, PG cells are capable of releasing GABA onto themselves (Smith and Jahr 2002; Murphy et al., 2005). This feature of PG cells is utilized in Chapter III as a method of directly assessing the magnitude of neurotransmitter release from a single neuron.

The primary focus of my research is the interplay between excitation and inhibition that lies downstream of OSN axon terminals but upstream of output MCs. In this circuit, ET cells and PG cells can directly signal to each other through feedforward excitation and inhibition respectively. The bi-directionality of this signaling, coupled with the fact that MCs are excited by ET cells (Gire and Schoppa, 2009; De Saint Jan et al., 2009; Najac et al., 2011; Gire et al., 2012) and inhibited by PG cells (Shao et al., 2009), demonstrates that this circuit plays an important role in shaping the probability of overall

glomerular output. The probabilistic nature of glomerular output may be in large part based on selective, input-intensity-based activation of inhibitory versus excitatory elements within a glomerulus, which will be discussed in greater detail in Chapter II.

Interglomular circuits

Within the olfactory bulb, there are additional cellular elements (e.g., granule cells (GCs) and several subtypes of short axon (SA) cells) that contribute to the inhibitory circuitry (Figure 1.4). A striking feature of these neurons is their ability to allow odorant- specific glomerular units to exert inhibitory influence on neighboring modules within discrete layers of the olfactory bulb. At present there is very little experimental evidence for lateral excitation between modules; however, computational work suggests that ephaptic signaling among OSN axons in the olfactory nerve layer could result in input to glomeruli whose axons have tightly coalesced (Bokil et al., 2001).

SA cells exist in several subtypes, defined by their synaptic partners, biochemical constituents, electrophysiological properties and localization in the olfactory bulb. However, for the purpose of this dissertation, SA cells will refer to neurons whose cell bodies reside in the glomerular layer and radially extend neurites up to several hundred micrometers (Aungst et al., 2003), predominantly within the glomerular layer. Additional SA cell subtypes are primarily found in the granule cell layer and include deep SA cells, Blanes cells and Golgi cells (Price and Powell, 1970; Eyre et al., 2013; Pressler and Stowbridge, 2006).

Figure 1.4. Intraglomerular and interglomerular GABAergic neurons in the olfactory bulb.

A, Periglomerular (PG) cells project a single dendrite into the glomerular neuropil and can provide feedforward inhibition within a single unit. B, Short axon (SA) cells connect many glomeruli via neuritic projections that span up to 1000 µm (Aungst et al., 2003). SA cells primarily inhibit glomeruli through GABAergic synapses onto external tufted (ET) cells (not depicted). C, Granule cells (GC) form reciprocal dendrodendritic synapses at lateral dendrites of mitral cells within the external plexiform layer (EPL). MCs at an excited glomerulus activate GCs who then feedforward inhibition to MCs projecting to surrounding glomeruli.

A

B

All three cell types found in the granule cell layer form reciprocal inhibitory synapses with GCs. Deep SA cells and Blanes cells may exist as overlapping, or partially overlapping, populations due to their similar morphology and connectivity. Golgi cells are identified by a similar, but more highly branched morphology compared with Blanes cells, as well as a characteristic bursting response to excitatory input (Pressler et al., 2013). The studies in the following chapters do not address connectivity in the granule cell layer, and, for this reason, SA cells within the glomerular layer will be the only cell type discussed in further detail.

Due to their GABAergic nature, SA cells mediate inhibition between glomeruli, known as lateral or center-surround inhibition (Aungst et al., 2003; Arevian et al., 2008; Kiyokage et al., 2010, Whitesell et al., 2013). Such a mechanism could, in principle, serve as an alternative or co-mechanism to enhance contrast in odor perception, with respect to feedforward inhibition mediated by PG cells. Recent evidence suggests that SA cells target ET cells through GABAergic synapses and, as such, have been shown to suppress glomerular output (Aungst et al., 2003; Arevian et al., 2008; Whitesell et al., 2013) by limiting ET cell excitation, which, in turn, suppresses feedforward excitation of MCs. SA cells also synthesize dopamine, and when co-released with GABA, can result in temporally bi-phasic synaptic inhibition-excitation responses in ET cells (Kiyokage et al., 2010; Liu et al., 2013).

Granule cells, located deeper in the bulb - within the granule cell layer - form reciprocal dendrodendritic synapses on the lateral dendrites of mitral cells in the external plexiform layer and provide both inter- and intra- glomerular inhibitory drive on to MCs

(Shepherd, 2004). That is, granule cells can provide feedforward inhibition to MCs associated with neighboring glomeruli, as well as provide feedback inhibition to the mitral cell from which it was activated (Jahr and Nicoll, 1982; Schoppa et al., 1998; Isaacson and Strowbridge, 1998). Through these interactions, granule cells are primarily involved in synchronizing the activity of mitral cells within the gamma frequency range.

In the olfactory system, gamma synchrony corresponds to oscillations within the timing of a single sniff, rather than allowing for comparisons between sniffs (Adrian, 1950; Lagier et al., 2004; Galan et al., 2006; Schoppa, 2006). Moreover, recent in vivo studies support the idea of GC mediated gamma synchrony of MCs, but also provide evidence that GCs play only a small role, if at all, on odor evoked inhibition of MCs (Fukunaga et al., 2014).

There are additional cell types whose intrinsic properties and functions have only recently been explored, and these include parvalbumin expressing GABAergic neurons found in the external plexiform layer that form reciprocal synapses at MC lateral

dendrites (Kato et al., 2013; Miyamichi et al., 2013). These neurons are broadly tuned to odor input due to their high connectivity with MCs that project to many unique

glomeruli. Because of this connectivity these cells are uniquely positioned to serve as a broad gain control mechanism as opposed to GCs that exhibit only sparse connectivity between MCs.

Cortical and subcortical input to the olfactory bulb

Although the studies presented in this dissertation do not consider modulatory inputs from cortical and subcortical nuclei, these projections are an important component of olfactory sensory processing. The axons of pyramidal cells in the piriform cortex and AON project back into the olfactory bulb where they form glutamatergic synapses with bulbar cells within several distinct layers (Boyd et al., 2012; Markopoulos et al., 2012; Rothermel and Wachowiak, 2014). So called “top-down” modulation of the olfactory bulb provides a mechanism for the context of the stimulus to impact overall perception. For example, during the behavior of active sniffing, an animal needs to decrease the detection threshold to accommodate for potentially low concentrations of an odor in their environment.

In addition to the above mentioned glutamatergic inputs, the olfactory bulb also receives projections from a host of other modulatory nuclei, including the basal forebrain, locus coerulus and raphe nucleus. In vivo studies have provided evidence that

dopaminergic, adrenergic and GABAergic projections among others, to the olfactory bulb are involved in stimulus discrimination (Escanilla et al., 2009; Doucette et al., 2007; Nunez-Parra et al., 2013); furthermore, several recent studies have examined how

modulatory inputs affect the circuit properties of the olfactory bulb in vitro (Pandipati and Schoppa, 2010; Pandipati et al., 2012; Liu et al., 2012; D’Souza et al., 2013; Schmidt and Strowbridge, 2014). The effects of each modulatory nucleus are entirely too vast to discuss in detail here, but for an excellent review of cholinergic modulation of the olfactory bulb see D’Souza and Vijayaraghavan (2014).

Specifically relevant to this study are centrifugal projections to the olfactory bulb that can excite or suppress the activity of PG cells either by direct or indirect

mechanisms. In addition to direct centrifugal inputs onto PG cells, low-level centrifugal excitation of ET cells could result in PG cell excitation and net glomerular inhibition due to the fact that MCs require strong ET activity to generate diffuse glutamate transients that are necessary to drive output (Gire et al., 2012). The potential for “top-down” modulation to influence processing within glomeruli will be addressed in greater detail in Chapter IV.

Organization of olfactory bulb glomeruli

Glomeruli on the surface of the olfactory bulb are organized with little to no discernible fine-scale chemotopy in terms of their spatial placement (Mori et al., 2006; Soucy et al., 2009, Hammen et al., 2014). As consequence of this topographically sparse arrangement of glomeruli, functional groups of MCs with similar odor tuning are not proximally ordered with respect to neural space (Fantana et al., 2008; Soucy et al., 2009; Ma et al., 2012), and, in fact, MCs projecting to neighboring glomeruli do not share similar odor response profiles (Egaña et al., 2005). Such disorganization is unique with respect to the architecture of most other sensory modalities including the well-studied retinotopy in V1 visual cortex (Tootell et al., 1998; Blasdel and Campbell, 2001),

tonotopy within the auditory cortex and nuclei in the brainstem (Friauf, 1992; Talavage et al., 2003; Kandler et al., 2009), as well as organization of the somatosensory cortex (Kass, 1997). A reoccurring feature within the organization of the central nervous system

is the placement of neural networks corresponding to similar or overlapping receptive fields with regard to the proximity of both stimulus and neural space. An arrangement of overlapping receptive fields allows for the sharpening of the sensory input-output

function through local inhibitory interactions (Kuffler, 1953; Cook and McReynolds, 1998).

In vivo studies have yielded much of our current understanding of how olfactory

bulb glomeruli are ordered with respect to their sensory inputs. Initially, glomeruli were found to display broad organization, with regions of the bulb corresponding to functional groups of related ligands (Meister and Bonhoeffer, 2001). However, more recent studies have found that these broad regions of organization do not extend to a finer scale (Soucy et al., 2009).

In terms of the physical properties of odor molecules, it appears that the olfactory system has abandoned, or alternatively never developed, a systemic organizing principle. It is worth noting that the studies discussed above do not explicitly rule out the possibility of an organizing principle guiding glomerular placement; however, at present we do not understand the rules governing such placement if they do in fact exist. As such, it remains possible that the olfactory system has developed an organization with regard to

perceptual rather than physical qualities of odor molecules. Another possibility is that glomeruli are organized with respect to ethological saliency, whereby odors

corresponding to behavioral outputs can exert influence on exhibition of potentially conflicting behaviors correspond to neighboring glomeruli.

Contrast enhancement in the olfactory system

Most sensory systems, including vision, audition, and somatosensation, are organized around the receptive fields of neural modules in at least one dimension. Therefore, modules corresponding to similar stimuli in physical space are often located near each other in neural space (Hubel and Wiesel 1959; Cook and McReynolds, 1998; Talavage et al., 2000; Kass, 1997). This relationship allows for well-activated neural areas to exert influence over neighboring neural modules with similar or overlapping receptive fields through lateral inhibitory interactions. Often termed “center-surround inhibition”, this mechanism has the ability to increase sensory information transfer by sharpening output from olfactory bulb glomeruli. This, in turn, allows for glomeruli corresponding to closely related inputs to respond more discretely to overlapping inputs.

In the olfactory system, output sharpening could be accomplished through odor-evoked inhibition of output neurons at a subset of glomeruli. In principle, the glomeruli receiving inhibition would correspond to odorant molecules that excited neighboring glomeruli with similar receptive fields.

Such a mechanism in the olfactory system could exist and be mediated by GABAergic SA cells. However, as mentioned above, olfactory bulb glomeruli are only coarsely arranged with respect to the chemical affinity of their afferent OSNs, often with glomeruli corresponding to similar odorant features situated several millimeters away from one another (Soucy et al., 2009). In effect, these long ranges and dispersed

organization leave it unlikely that lateral inhibition, mediated by SA cells, could serve as the mechanism of contrast enhancement in the olfactory system. This is largely due to the

fact that these neurons would have to be highly targeted and span relatively long distances to be effective in this capacity (Figure 1.5).

Recent studies have provided evidence against lateral inhibition as the mechanism of contrast enhancement from a behavioral perspective. Mice are able to identify target odors even against a background of up to 14 molecularly similar odorant molecules (Rokni et al., 2014). If lateral inhibition sharpens the activity among glomeruli corresponding to molecularly or perceptually similar odorant molecules, an expected outcome of the dense background is a suppression of activity at the glomerulus of the target odor. It should be noted that the aforementioned authors did observe a decrease in the discrimination accuracy that was correlated with the number of background odors; whether this is a result of lateral interactions between glomeruli or a simple signal to noise convolution is not easily interpreted from the behavioral assay alone. Furthermore, the results of this same study provide direct evidence that olfaction is not purely a synthetic sense, whereby overlapping odor maps combine to form a new perception, as was once thought.

Similar to SA cells, but on a much finer scale, GCs could also contribute to odor-evoked inhibition of MCs via lateral or center-surround inhibition. Mechanistically, GCs provide feedforward inhibition to MCs corresponding to unique glomerular units. The connections between GCs and MCs are sparse (Kato et al., 2013), and as such the

probability that any individual GC inhibits a neighboring MC is quite low. However, GCs are the most numerous neuron in the olfactory bulb (Shepard et al., 2004) which could compensate for the low probability of GC-MC connectivity.

Figure 1.5. Interglomerular inhibition and contrast enhancement in the olfactory system.

A, If olfactory bulb glomeruli are arranged in a scheme incorporating local chemotopy, lateral inhibition may serve as an effective means of contrast enhancement. In such a scheme, glomeruli corresponding to similar molecular features of odorant molecules are ordered near one another in neural space. Because glomeruli are situated in close

proximity, GABAergic inhibition from one glomerulus can suppress the activity of its neighbors. B, If glomeruli display no clear relationship among neighbors, lateral

inhibition becomes ineffective at suppressing the activity of glomeruli corresponding to similar molecules due to their dispersed placement and the relatively long distances between them. Recent studies have provided evidence that glomeruli are ordered in a course but not fine scale (Meister and Bonhoeffer, 2001; Soucey et al., 2009). Figure adapted from Schoppa, 2009.

Input Output Lateral Inhibition Chemotopic Map Input Output Lateral Inhibition Chemotopic Map A B

Interestingly, recent evidence suggests that GCs do not contribute odor-evoked inhibition of MCs (Fukunaga et al., 2014; Figure 1.6). In vivo experiments, which specifically silenced GCs, revealed that odor-evoked inhibitory responses in MCs persisted in the absence of GC activity (Figure 1.6D). This same work also provides strong evidence that GABAergic PG cells do indeed contribute to the inhibition of MCs (Figure 1.6E). Recent computational work further demonstrates that PG cells may be far more effective at inhibiting the activity of MCs than GCs (Arruda et al., 2013). These studies focus on inhibition of MCs through direct PG cell-MC connectivity; however, a largely ignored, and perhaps more potent mechanism to suppress MC activity, may be operate through PG cell mediated inhibition of ET cells (Gire and Schoppa, 2009) - a crucial driver of MC excitation (Gire et al., 2012). Whether suppression of MC activity is accomplished through direct synaptic connectivity (PG cell to MC synapses) or a

polysynaptic circuit (ET cells to PG cells to MCs) has yet to be investigated.

Central to this body of research is the premise that contrast enhancement in the olfactory system can arise through the intrinsic circuitry of a single glomerulus. Under this hypothesis, weak incoming olfactory signals arising from “off-target” or low-affinity odorant molecules may be able to selectively engage GABAergic PG cells due to their intrinsic biophysical properties, including electrical impedance or input resistance (Figure 1.7A). The selective activity of PG cells would then result in feedforward inhibition onto excitatory cellular elements within the same glomerulus, subsequently preventing output to downstream neural structures. Following stronger input from “on-

Figure 1.6. Periglomerular cell mediated odor evoked inhibition of mitral cells. A, In vivo whole-cell current-clamp recording of MC membrane potential in response to odor presentation in the anesthetized mouse. B, A subset of the MC odor responses resulted in clear hyperpolarizations of the membrane potential. C, Schematic of the potential sources of inhibition onto MCs. GCs synapsing with a MC at a strongly-activated glomerulus could feedforward GABAergic inhibition onto the lateral dendrites of MCs corresponding to neighboring glomeruli. Alternatively, inhibition of MCs could arise though interaction intrinsic to a single glomerulus. In such a hypothesis,

GABAergic PG cells are selectively activated by weak incoming odor signals due to their high input resistance. D, The activity of GCs was silenced by ArchT, a proton pump that results in a decrease in cellular activity. When ArcT was active, the observed MC odor evoked hyperpolarizations persisted. E, When the activity of PG cells was reduced using the same method, MC odor evoked inhibitory responses were reduced in amplitude. Figure from Fukunaga et al., 2014.

A B D E

target”, high-affinity odorant molecules, OSNs can engage ET cells that will subsequently excite mitral cells (Figure 1.7B).

Several additional studies have expounded on intraglomerular contrast

enhancement from a modeling perspective (Figure 1.8; Sethupathy and Cleland, 2006, Linster and Cleland, 2009). These ideas are again based on the high input resistance of PG cells (0.8-1.5 GΩ) compared to MCs (< 160 MΩ; Margrie et al., 2001; Puopolo and Belluzzi, 1998; Smith and Jahr, 2002). However, one caveat to this work is that the glomerular circuitry in the model environment is highly simplified. Absent are ET cells, which we now know drive both excitation of MCs and a large portion of PG cells at a glomerulus (Najac et al., 2011; Gire et al., 2012; Shao et al., 2009). This work proposes that odorant molecules that provide low-level input to glomeruli may be able to

selectively engage PG cells that then provide feedforward inhibition onto MCs, thereby reducing their activity (Figure 1.8C-D). Following input from odorant molecules that provide stronger input, the theoretical excitation received by MCs from OSNs is

sufficient to “out-scale” feedforward inhibition from PG cells and output from glomeruli is generated.

This dissertation seeks to expand on these ideas and provide evidence of

glomerular signal filtering from a physiological perspective. Although largely ignored in the modeling studies, in the work presented below, we will consider the role that ET cells play as both a critical driver of glomerular excitation and inhibition.

Figure 1.7. Schematic of intraglomerular filtering based on the intrinsic properties of juxtaglomerular neurons.

A, Weak incoming olfactory signals, potentially arising from low-affinity “off target” odors, could selectively engage a population of GABAergic PG cells that form direct synapses with the axon terminals of OSNs due to their intrinsic biophysical properties. PG cells have a higher input resistance compared to ET cells or MCs, thus allowing for lower levels of excitatory input from the olfactory periphery to exert larger membrane depolarizations subsequently increasing the probability of output generation from PG cells. PG cells then feed-forward inhibition onto ET cells that further reduces their excitability and ability to feed-forward excitation onto MCs. Following weak inputs sensory signals are

“filtered” from being broadcast via the axons of MCs to higher order structures in the brain. B, High affinity “on target” odor molecules provide greater input to the glomerulus in the form of heightened OSN activity. This greater level of input can engage lower input resistance ET cells, that, in turn, provide feedforward

excitation onto output MCs. These signals are allowed to “pass” through the filter despite the activity of GABAergic PG cells.

→ -+ Periglomerular Cell External Tufted Cell Mitral Cell Weak Input

(”Off Target Odors”)

→

(”On Target Odors”)

Figure 1.8. Computational evidence for intraglomerular signal filtering. A, Tuning curves for MCs and PG cells mapped onto an axis of odor ligand-receptor potency. PG cells are more readily recruited than MCs to low levels of OSN activity and the net effect on MCs is inhibition at low odor ligand-receptor potencies. B, Schematic representation of neuronal responses to a given odorant in the presence of intraglomerular PG-mediated inhibition of mitral cells at two different potencies. At higher ligand-receptor potencies MCs generate output. Adapted from Cleland and Linster, 2009. C, A model MC response to the sequential presentation of a homologous series of nine different odorants. Low affinity odors generate suppressive MC responses. D, Suppressive MC responses depend on GABAA receptors. Odor-evoked inhibition in MCs can be transformed

into excitation when bulbar GABAA receptors are blocked. Figure from Cleland

and Sethupathy, 2006.

Cleland and Linster Decorrelation of odor representations

FIGURE 1 | Comparison of decorrelation models. (A) Schematic

comparison of on-center/inhibitory surround and non-specific decorrelation functions. (i) Two overlapping input representations (α and β) depicted in one dimension. (ii) Canonical on-center/inhibitory-surround decorrelation generates an explicit inhibitory surround in which the shoulders of the input representation are inhibited below baseline, yielding a sharp reduction in overlap among similar representations. This computation is performed by lateral inhibition in the retina and cochlear nucleus, and by the

non-topographical model of olfactory receptive field decorrelation. (iii) A lesser degree of decorrelation can be obtained by broad, non-specific inhibition, including lateral inhibition with an unstructured surround, although this imposes a general reduction in sensitivity across the entire

representation. This is the effect of most lateral inhibitory models studied to date in the olfactory bulb; notably, it does not generate the inhibitory surround observed byYokoi et al. (1995). Whereas both computations can effect a measurable decorrelation in principle, the two transformations differ both qualitatively and in terms of quantitative efficacy. Figure adapted fromCleland (2010). (B) Lateral inhibition. (i) Left panel. Tuning curves for two mitral cells (Mi 1 and Mi 2) with overlapping receptive fields for odorants, prior to the effects of lateral inhibition in a topographical representation scenario. Both neurons are excited by the odorant presented, although Mi 1 is more strongly activated than Mi 2. Right panel. The same two mitral cell tuning curves after the inclusion of lateral inhibition. Now, whereas Mi 1 is still excited by the odorant presented, Mi 2 is inhibited. The abscissa is a hypothetical axis of odor quality. (ii) Schematic representation of neuronal responses to a given odorant in the absence of lateral inhibitory PG axonal projections in a topographical representation scenario. The odorant presented activates the lightly shaded population of OSNs somewhat more strongly than it does the more darkly shaded population of OSNs, evoking a higher spike rate in the

OSN population projecting to the glomerulus on the right. In the absence of inhibition, mitral cells (Mi) are activated in direct proportion to their constituent OSN populations. (iii) Schematic representation of the same two glomeruli and the same odorant presented as in (Bii), with the addition of PG cells that also are activated in direct proportion to their OSN population and deliver lateral inhibition onto mitral cells in the other glomerulus. The mitral cell that is more weakly responsive to the odorant presented [corresponding to the dotted vertical line in (Bi)] is silenced due to this lateral inhibitory input from the PG cell associated with the more strongly activated parent glomerulus.

(C) Intraglomerular inhibition. (i) Tuning curves for mitral and periglomerular

cells mapped onto an abscissa of odor ligand-receptor potency (potency incorporates both ligand-receptor affinity and efficacy terms; for discussion of the effects of odor concentration on this relationship, seeCleland et al., 2007). Both mitral and periglomerular cells are excited by the odorant presented via the activity of their associated OSN populations (Miin, PGin); though PG cells

are more sensitive to this common input (Gire and Schoppa, 2009). Inhibition of mitral cells by PG cells alters the mitral cell tuning curve (Miout), generating

a Mexican-hat inhibitory surround in a metric space defined by odor quality. Figure adapted fromCleland and Sethupathy, 2006). (ii) Schematic representation of neuronal responses to a given odorant in the presence of intraglomerular PG-mediated inhibition of mitral cells. The odorant presented is the same as in (B), exhibiting a stronger potency for the receptors expressed by the OSN population projecting to the glomerulus on the right [the two potencies correspond to those depicted by vertical lines in (Ci)]. Periglomerular cells are activated in direct proportion to their constituent OSN populations, as in the lateral inhibitory case, whereas mitral cells receive both afferent excitation and intraglomerular inhibition, thereby exhibiting sharpened receptive fields with inhibitory surrounds. The mitral cell that is more weakly responsive to the odorant presented is silenced [compare to (Biii)].

BMC Neuroscience 2006, 7:7 http://www.biomedcentral.com/1471-2202/7/7

Page 12 of 18 (page number not for citation purposes) Non-topographical contrast enhancement replicates the canonical observation of olfactory contrast enhancement

Figure 6

Non-topographical contrast enhancement replicates the canonical observation of olfactory contrast enhance-ment. A. A model mitral cell's response to the sequential presentation of a homologous series of nine different odorants.

Nonspecific sinusoidal excitation was added to replicate the respiration-linked activity observed in the cell recorded by these authors (see Methods). The homologous odor series was simulated by altering the odor ligand-receptor affinity of the OSNs that project to the mitral cell depicted from near zero to a maximal value (at odor 4) and back along the trajectory of a normal distribution. The ligand-receptor affinities of the OSNs associated with the other nine glomeruli were sampled randomly from another normal distribution, as if they were exhibiting similar sensitivity profiles to unknown odor series. Odors were applied for 2 seconds (horizontal bar). NTCE fully replicated the Mexican hat contrast enhancement function observed in a mitral cell reported by Yokoi et al. ([10]; their Figure 2A). B. Response dependence on GABAA receptors. Odor-evoked inhibition in

mitral cells can be transformed into excitation when bulbar GABAA receptors are blocked [10]. In the present model, blockade

of GABA-ergic synapses from periglomerular cells effected this reduction of inhibition, replicating the results of Yokoi et al. ([10]; their Figure 5D). A constant background stimulation was applied in all cases to generate tonic spiking so that inhibition could be observed. Control, no odor stimulus was applied. Odor, a 4 second stimulus (using odor 6 from Figure 6A) was applied (horizontal bar), evoking an inhibitory response in the mitral cell. Bicuculline + odor, the same stimulus was applied after block-ing all periglomerular synapses onto mitral cells. NTCE replicated the effects of bicuculline application as shown by Yokoi et al. [10], though the effect was mediated via periglomerular cells rather than granule cells as proposed by those authors.

A odor 1 odor 2 odor 3 odor 4 odor 5 odor 6 odor 7 odor 8 odor 9 control odor bicuculline + odor B 2 sec 2 sec A B C D

Overview of present work

In Chapter II, I describe a mechanism that allows for olfactory bulb glomeruli to serve as a gate or filter to incoming signals. Using a combination of electrophysiological and imaging approaches, this chapter demonstrates that at weak levels of olfactory input, inhibition is the dominant conductance within glomeruli and that this inhibition,

originating from GABAergic PG cells in the glomerular layer,

suppresses overall glomerular excitation and MC output. Previous computational (Sethupathy and Cleland, 2006; Linster and Cleland, 2009) and experimental (Gire and Schoppa, 2009) work has established that such a phenomenon is possible based on the high input resistance of PG cells (Puopolo and Belluzzi, 1998; Smith and Jahr, 2002) and feedforward inhibition from PG cells onto ET cells and MCs. Furthermore, I provide evidence that along the continuum of olfactory input intensities, excitation eventually “out-scales” inhibition, thereby allowing for output to occur.

Electrophysiological measurements determined that PG cells are substantially more sensitive to olfactory input via the axons of OSNs than ET cells; however, ET cells generally receive inputs of larger magnitude per level of OSN activity. Due to these opposing factors, PG and ET cells are often co-active following weak olfactory inputs; nonetheless, the net state of the glomerulus remains inactive in terms of output from MCs. To this end, we find that the magnitude of glomerular output via MCs is correlated with the number of active ET cells at the same glomerulus. Lastly, in this chapter I show that as the magnitude of olfactory input increases, group II metabotropic glutamate receptors (mGluRs) provide a neuromodulatory mechanism that results in a plateau of

inhibition, ensuring that excitation dominates over inhibition and signal passage from the glomerulus occurs. The specific mechanism by which group II mgluRs reduce inhibition within glomeruli will be discussed in the subsequent chapter.

In Chapter III, I discuss a neuromodulatory mechanism that can shift the balance of excitation and inhibition within glomeruli based on the intensity of sensory input via the action of group II mGluRs on GABAergic PG cells. group II mGluRs are inhibitory glutamate receptors that reduce cellular excitability and neurotransmitter release (Anwal, 1999; Hayashi et al., 1993; Tanaguchi et al., 2013). Furthermore, group II mGluRs are generally found pre- or extra-synaptically (Shigemoto et al., 1997; Schoepp, 2001), thus making them an ideal target of extrasynaptic glutamate transients known to exist within glomeruli (Isaacson, 1999; Christie and Westbrook, 2006; Gire et al., 2012). I

demonstrate that activation of these receptors within glomeruli decreases the magnitude of inhibition onto ET cells. Functionally, this reduced inhibitory drive enhances ET cell activity and overall glomerular output. Moreover, my data indicates that the glutamate acting at group II mGluRs is derived from ET cells, acts through diffusion based mechanisms, and results in glomerular disinhibition following strong olfactory inputs.

I conclude this dissertation with a discussion of the data presented in Chapters II and III and its contribution to our understanding of functional odor processing. The results are discussed in terms of both sensory gating and non-topographical contrast enhancement. I discuss future experiments, both in vitro and in vivo, that could broaden our insight as to the mechanistic basis of sensory filtering in the olfactory bulb

neurogenesis in the dynamic remodeling of glomerular networks. Finally, I briefly discuss an application of the olfactory system as a disease model, with a focus on the glomerulus and misregulation of GABAergic circuits within.

CHAPTER II

GABAERGIC SIGNAL FILTERING AT OLFACTORY BULB GLOMERULI

Introduction

A multitude of recent studies have advanced our understanding of both the circuitry and functionality of the olfactory system; however, a largely unanswered question remains: How are molecularly-similar odorant compounds discriminated by the olfactory system? In sensory modalities, including vision (Cook and McReynolds, 1998; Blasdel and Campbell, 2001) and somatosensation (Sur, 1980; Kass, 1997), the

mechanistic basis of contrast enhancement arises through lateral interactions between strongly and less-strongly activated neural modules. Within the circuitry of the mammalian olfactory bulb, lateral interactions between functional units known as

glomeruli have been both well established and extensively studied (Pinching and Powell, 1971; Aungst et al., 2003, Arevian et al., 2008; Kiyokage et al., 2010; Whitesell et al., 2013); however, the ability of these interactions to form the neural basis of contrast enhancement comes into question on two fronts. 1) GABAergic granule cells that form reciprocal dendodendritic synapses with mitral cells (MCs) projecting to discrete

glomeruli, are strongly implicated in synchronizing the output of MCs within the gamma frequency range (Lagier et al., 2004; Galan et al., 2006; Schoppa, 2006) and appear to contribute little to odor-evoked inhibition of mitral cells (Arruda et al., 2013; Fukunaga et al., 2014); and 2) the lack of fine chemotropic organization among glomeruli (Mori et al., 2006; Fantana et al., 2008; Soucy et al., 2009; Ma et al., 2012). Functional evidence

suggests that MCs projecting to neighboring glomeruli do not share overlapping odor response profiles (Egaña et al., 2005). Such disorganization makes it unlikely that lateral interactions mediated by GABAeric short axon (SA) cells are directed in a targeted manner with regard to the molecular structure of odorant molecules and may be better suited to facilitate local gain control (Linster and Cleland, 2009).

The possibility that contrast enhancement in the olfactory system arises in part through local interactions within a single glomeruli (Cleland and Sethupathy, 2006) is particularly attractive and is in large part based on the relative excitability of GABAergic periglomerular (PG) cells and output MCs. PG cells are small neurons with a high input resistance compared to MCs and as such may be more responsive to low-levels of

sensory input to glomeruli (Puopolo and Belluzzi, 1998; Smith and Jahr, 2002). Although past modeling studies have considered the intrinsic excitability of PG cells with respect to MCs, missing from these studies is an analysis of external tufted (ET) cell excitability (Cleland and Sethupathy, 2006; Linster and Cleland, 2009). ET cells, in large part,

control MC excitation (De Saint Jan et al., 2009; Najac et al., 2011; Gire et al., 2012), but are also the primary source of excitation to the majority of PG cells at a glomerulus (Shao et al., 2009). Due to this circuit configuration, with ET cells acting as a “gate-keeper” of both excitation and inhibition in a glomerulus, it is important to understand the

mechanisms by which these opposing pathways are engaged with respect to incoming sensory information.

Our analysis takes into account the intrinsic biophysical properties, including input resistance, of both PG and ET cells. We also consider the relative synaptic

connectivity between olfactory sensory neuron (OSN) axon terminals and PG/ET cells. Our results are in agreement that the circuitry of olfactory bulb glomeruli can support a filtering mechanism; however, in contrast to previous studies, we report that the

glomerular filter operates through ET cells as a critical driver of inhibition following low-level olfactory input.

In this study we used both single and dual patch-clamp recordings as well as calcium imaging techniques to assess the relative excitability of PG cell-mediated GABAergic circuits and ET cell-mediated glutamatergic circuits within a single glomerulus. We obtained evidence that when weak signals arrive at glomeruli, the GABAergic network of PG cells is selectively engaged, which in turn leads to the filtering of these weak incoming signals through synapses on both ET and MCs (Gire and Schoppa, 2009; Shao et al., 2013). However, quite unexpectedly, it appears that the input necessary to activate PG cells arises from a combination of direct OSN input and indirect multi-step excitation through ET cells (Shao et al., 2009). Lastly, we determined that the magnitude of inhibition within glomeruli is many times larger than excitation at the weakest levels of olfactory input and this relationship is inverted following stronger inputs, thereby ensuring that excitation is the dominant conductance within glomeruli following strong inputs.

Materials and methods

Animals

Male and female eight to 20 day-old Sprague Dawley rats obtained from Charles River Laboratories as well as male and female vesicular γ-aminobutyric acid (VGAT)-Venus transgenic rats were used for all experiments. The latter rat model expresses (VGAT)-Venus fluorescent protein under control of the VGAT promoter (Uematsu et al., 2008, strain 2; Wistar background; Whitesell et al., 2013). Animals used for these experiments were both heterozygotes and homozygotes. All experiments were conducted under protocols

approved by the Animal Care and Use Committee of the University of Colorado, Anschutz Medical Campus.

Slice preparation

Acute horizontal olfactory bulb slices (330 ︎µm) were prepared following isoflurane anesthesia and decapitation as described in Schoppa et al. 1998. Olfactory bulbs were rapidly removed and placed in oxygenated (95% O2, 5% CO2) ice cold

solution containing (in mM): 72 sucrose, 83 NaCl, 26 NaHCO3, 10 glucose, 1.25

NaH2PO4, 3.5 KCl, 3 MgCl2, 0.5 CaCl2 adjusted to 295 mOsm. Olfactory bulbs were

separated into hemispheres with a razor blade and attached to a stage using adhesive glue applied to the ventral surface of the tissue. Slices were cut using a vibrating microslicer (Leica VT1000S) and were incubated in a holding chamber for 30 minutes at 32°C. Subsequently, the slices were stored at room temperature. Experiments were carried out under an upright Zeiss Axioskop2 FS Plus microscope (Carl Zeiss MicroImaging) fitted