The Physiological Role of Serotonergic Transmission in Adult Rat Taste Buds

Home , Lingual papillae, Taste bud, Taste receptor

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of the Ohio State University

Fadi Luc Jaber, BMedSci.

Oral Biology Graduate Program

The Ohio State University

2012

Dissertation Committee

M. Scott Herness

Georgia Bishop

Peter Reiser

Fadi Luc Jaber

2012

ABSTRACT

The role of the neurotransmitter serotonin (5-HT) in the taste bud is not well understood. Traditionally, serotonin was assumed to be the signaling agent between taste receptor cells (TRCs) and afferent nerve fibers. Across mammalian species, serotonin is expressed in a highly conserved manner in the type III cell of the bud, the only cell that forms classic synaptic contacts with sensory afferent nerve fibers. Surprisingly, neurotransmission between TRCs and sensory afferents was discovered to be purinergic, through adenosine triphosphate (ATP) release from type II TRCs, leaving the role of serotonin enigmatic. At present, the consequences of serotonin release from type III cells are still ambiguous. A clue to serotonin’s role came with the discovery of serotonergic receptors within the bud. Of the many serotonin receptors, which form seven major families, the taste bud expresses only two, 5-HT1A and 5-HT3. The 5-HT1A receptor subtype is expressed in TRCs, while the 5-HT3 subtype is assumed to be expressed in the afferent nerve endings. Further, serotonin and the 5-HT1A receptor are expressed in non-overlapping subpopulations of TRCs. This expression pattern led to a new hypothesis for serotonin. In addition to its assumed role of signaling afferent nerve fibers, serotonin may function in a paracrine role among TRCs within the taste bud. At present, signaling mechanisms that define the role of 5-HT1A receptor in the taste bud remain obscure.

The present study seeks to explore this role of 5-HT in the taste bud by using molecular and physiological characterization of the 5-HT1A-expressing TRCs. Three main approaches were used to explore the central hypothesis that 5-HT, released from type III cells, modulates the transduction of multiple taste qualities by directly regulating the release of ATP from type II TRCs following gustatory stimulation. Using the rat as an experimental model, immunocytochemical (ICC) phenotyping of the 5-HT1A-expressing TRC, molecular characterization of this cell type using single cell reverse transcriptase PCR (RT-PCR), and electrophysiological functional study of this receptor using whole nerve chorda tympani recordings were performed.

Immunocytochemical examination of the 5-HT1A-expressing TRC was performed on taste buds located throughout the oral cavity to confirm the prediction that 5-HT1A receptors are expressed in type II cells and are not expressed in type III cells. Paracrine localization of the serotonin and the 5-HT1A-TRC was confirmed in anterior taste buds, supporting the previous paracrine localization observed in posterior taste buds. The expression pattern of 5-HT1A-TRC and a variety of protein markers which have previously been confirmed to be linked to either type II or type III cells in rat taste buds, was explored using double-label immunocytochemistry. 5-HT1A-TRC was consistently found to be co-expressed with all tested markers for type II cells. These markers included the taste-specific G-protein gustducin (GUST), synaptosomal-associated protein 25 (SNAP25), cholecystokinin octapeptide (CCK8), neuropeptide Y (NPY), the neuropeptide Y-1 receptor (NPY1R), and glutamic acid decarboxylase (GAD). On the other hand, 5-

HT1A-TRC expression was completely exclusionary to the type III cell marker, neural cell adhesion molecule (NCAM).

Using the single cell RT-PCR technique, the expression of 5-HT1A receptor was examined with members of the type 1 (T1R) and type 2 (T2R) families of taste receptors,

iii

the candidate sour receptor, polycystic kidney disease 2-like 1 protein (PKD2L1), as well as additional molecular markers. The T1R family of receptors, which respond to sweet and umami stimuli, and the T2R family of receptors, which respond to bitter molecules, are expressed in non-overlapping subpopulations of type II TRCs. It was predicted that

5-HT1A-receptor expression should display significant co-expression with T1R and T2R receptors. Data revealed that 5-HT1A- expressing TRCs were noted to be co-expressed in cells expressing either T1R or T2R receptors. Additionally, 5-HT1A receptors were not co- expressed in cells which expressed the receptor for sour stimuli PKD2L1, expressed in type III cells. Hence the immunocytochemical and single cell RT-PCR data collectively confirmed the expression of the 5-HT1A receptor in type II cells of the taste bud. These findings suggest a modulatory role for serotonin in the bud among the transduction mechanisms of different taste qualities.

To examine the functional role of the 5-HT1A receptors in the taste bud, whole nerve recordings from the chorda tympani (CT) nerve (which innervates the anterior 2/3 of the tongue) were performed before and after jugular injection of serotonin receptor antagonists. It was predicted that inhibition of these receptors would significantly affect neural responses to taste qualities by modulating ATP release from type II TRCs.

Injection of the 5-HT1A receptor antagonist WAY-100635 produced inhibitions in the integrated neural responses to lingually applied sweet, sour, salty, and bitter stimuli. In separate experiments, injections of another 5-HT1A receptor antagonist, NAD-299, produced similar results. On the other hand, injection of ondansetron, a 5-HT3 receptor antagonist, failed to produce any significant change in the chorda tympani responses to these lingually applied stimuli. These data provide solid evidence for a significant modulatory role for serotonin in the modulation of taste information within the taste bud.

Overall, findings from this study suggest that serotonin’s action at the level of the chorda tympani is mediated through its 5-HT1A receptor, found expressed paracrine to 5-HT in type II TRCs, including those that respond to bitter, sweet and umami tastes.

Blocking serotonin’s action on the 5-HT1A receptor caused a generalized inhibition in neural responses to sweet, bitter, salty, and sour tastes, suggesting that serotonin is facilitatory to gustatory signal transmission. At first, this action of serotonin appears to contradict previous reports that showed that, at the cellular level, serotonin release was inhibitory to neurotransmitter (ATP) release from type II cells following gustatory stimulation. However, it is hypothesized that serotonin facilitates gustatory signals by regulating the release of ATP, which would otherwise desensitize post-synaptic purinergic receptors expressed on afferent nerve fibers, resulting in inhibited neural responses. Supporting this hypothesis is the observation that P2X receptors, which are expressed in afferent nerve endings, are known to have a fast rate of desensitization.

Also, preliminary data showed that 5-HT1A activation prevents the resynthesis of the membrane-bound phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) in TRCs, resulting in the inhibition of gap junctions, the putative channels for ATP release from type II TRCs. The study concludes that serotonin is a crucial element in a finely-tuned feedback loop involving the 5-HT1A receptor, ATP, and the P2 purinoceptors. The aim of this feedback loop is to modulate afferent taste signals prior to their transmission across gustatory nerves.

Dedicated to Hannah

ACKNOLEGMENTS

As this journey comes to a close, I look back on the past few years and I realize how fortunate I have been for having Dr. Scott Herness as my mentor and guide. The attributes I found in Dr. Herness truly exemplify excellent mentorship, and I am privileged to have been trained by a scientist of his caliber, and to know him on a personal level. I could not have wished for a better role model. My appreciation also goes to Dr. Fang Li Zhao for always being there when I needed assistance and sound scientific advice. I also thank Tamara Kolli whose help was instrumental in my progress since my starting day, and with whom I shared many enjoyable conversations. I am also grateful for the members of my dissertation committee, Dr. Georgia Bishop and Dr. Peter Reiser for their help and guidance. Throughout my life, my parents have provided me with unconditional love and support, and I am forever indebted to them for encouraging my curiosities and aspirations, and fostering my passion for learning. It is also not without the unwavering patience and support of my wife Mirna that I could have reached this milestone, and I thank her from the bottom of my heart for being beside me, even during the most difficult times. I am incredibly blessed to have my daughter, Hannah, as the ever present source of motivation and happiness in my life. For inspiring me every day, for always giving me reasons to keep calm and carry on, to strive for greatness, and to be a better person, I dedicate to her this work and every future success.

vii

VITA

1996 ………………………..….Collège Notre Dame des Soeurs Antonines – LEBANON

2000 ………….…………….….B.S. Biology, American University of Beirut – LEBANON

2005 ……………………………BMedSci, Medical Science, University of Birmingham – UK

2006 to present ………..Graduate Research Associate, College of Dentistry, Section of Oral Biology, the Ohio State University

PUBLICATIONS

Zhang, Y. Kolli, T., Hivley, R. Jaber, L. Zhao, F-L., Yan, J., and Herness, S. (2010). Characterization of the expression pattern of adrenergic receptors in rat taste buds. Neuroscience 169:1421-1437.

FIELDS OF STUDY

Major Field: Oral Biology

viii

TABLE OF CONTENTS Page Abstract …………………………………………………………………………………………………………………………ii Dedication …………………………………………………………………………………………………………………….vi Acknowledgments ……………………………………………………………………………………………………….vii Vita …………………….……………………………………………………………………………………………………….viii List of Tables ………………………………………………………………………………………………………………..xii List of Figures ……………………………………………………………………………………………………..……xiii

Chapters: 1. Introduction ………………………………………………………………….……………………………..……1 1.1. Taste buds are the end organs of gustation .……………………………………………….3 1.2. Gustatory transduction ..………………………………………………………………….………...4 1.2.1. Signaling agents within the taste bud ..…………………………….…….…...6 1.2.2. Early and late gustatory transduction mechanisms ..…………………...7 1.3. Serotonin in the taste bud …..……………………………………………………………………..8 1.3.1. 5-HT receptors are expressed in the taste bud ….……………………...... 9 1.3.2. Serotonin is released in the taste bud following gustatory stimulation …………………………………..…………………………………………...11 1.4. ATP, purinergic transmission, and the relationship with serotonin....….…....12

2. Methods .………………………………………………………………………………………………….….….17 2.1. Immunocytochemistry .………………………………………………………………….………...17 2.1.1. Subjects ………………………………………………………………………………….….17 2.1.2. Tissue preparation …….……………………………………………………………….18 2.1.3. Conventional immunocytochemistry protocol .………………………....18 2.1.4. Tyramide signal amplification (TSA) immunocytochemistry ..……..19 2.1.5. Experimental controls ..……………………………………………………………...20 2.1.6. Cell counting and data analysis ……………………………………………….….21 2.1.7. Antibodies ..…………………………………………………………………………….….22 2.1.8. Image processing ………………………………………………………………..……..22

2.2. Reverse trascriptase PCR (RT-PCR) ..……………………………………………..…...... 23 2.2.1. Whole papilla RNA extraction ..……………………………….………………….23 2.2.2. cDNA synthesis ….……………………………………………………………………….24 2.2.3. Single-cell RNA extraction and cDNA synthesis .……………………..…..24 2.2.4. Whole taste papilla/single-cell polymerase chain reaction ..……....25 2.2.5. Primers ……………………………………………………………………………………….26 2.2.6. Experimental controls ………………………………………………………..……...27 2.3. Electrophysiological whole nerve chorda tympani (CT) recordings ..………….27 2.3.1. Reasons for using chorda tympani nerve recording to study serotonergic function in the taste bud ……………………….….………….28 2.3.2. Animals and surgery………………………………………………………….……..…29 2.3.3. Nerve recording………………………………………………………………………....31 2.3.4. Gustatory stimulation of the tongue ..………………………………..………31 2.3.5. Drugs ……………………………………………………………………………………...... 33 2.3.6. Experimental controls………………………………………………..………….……33 2.3.7. Data analysis……………………………………………………………………….………34

3. Results .…………………………………………………………………………………………………………...37

3.1. Phenotyping the 5-HT1A-expressing TRCs ………………………………………..………..37 3.1.1. Distribution on 5-HT and 5-HT1A in taste buds of the anterior tongue …………………………………………………………………………….…....….38 3.1.2. Co-expression of 5-HT1A and TRC-specific markers ….……….……..…39 3.1.3. Co-expression of 5-HT1A and taste receptor molecules .…….…..…..42 3.2. Physiological properties of serotonin as measured at the level of the chorda tympani nerve ……………………………………………………………………….…..…46 3.2.1. Validation of experimental preparation ..……………………………..…….47 3.2.2. Chorda Tympani response to ATP jugular injection .…………..…..….47 3.2.3. Chorda Tympani response to saline jugular injection …………....…..49 3.2.4. Effect of blocking 5-HT1A receptors on CT responses to gustatory stimulation …………………………………………………………………………….....50 3.2.5. Effect of blocking 5-HT3 receptors on CT responses to gustatory stimulation ……………………………………………………..………………………...53

4. Discussion……………………………………………………………………………………………………..….55

4.1. 5-HT and 5-HT1A are both expressed in taste buds across the oral cavity …..57 4.2. The serotonin 5-HT1A receptor is expressed in type II TRCs that respond to sweet, bitter, and umami taste stimulation ………………………………………..…….57

4.3. Physiological effects of serotonin at the gustatory cellular and the neural levels ..…………………………………………………………………………………………..….….....63 4.3.1. Jugular injections of gustatory modulators produce immediate effects recordable at the level of the chorda tympani ..……………..65 4.3.2. Blocking the 5-HT1A receptor inhibits chorda tympani responses to lingual taste stimuli …………………………………………………………...... 67 4.4. The physiological mechanisms involved in serotoninergic modulation of gustatory stimuli …………………………………………………………………..……………...... 68 4.5. Conclusion ……………………………………………………………………………….…..………….70

List of references …………………………………..………………………………………………….…..108

LIST OF TABLES

Table Page

1. List of primary antibodies used in the present study………………………………………….….72

2. Primer sequences used in RT-PCR reactions………………………………………………….………73

3. Quantitative immunocytochemical double labeling patterns of 5-HT1A and phenotypic taste receptor cell markers in fungiform papillae of the anterior tongue and in the NID…………………………………………………………………………………………….………..74

4. Quantitative immunocytochemical double labeling patterns of 5-HT1A and phenotypic taste receptor cell markers in circumvallate and foliate papillae of the posterior tongue…………………………………………………………………………………………..………75

5. Quantitative immunocytochemical double labeling patterns of 5-HT1A and phenotypic taste receptor cell markers totaled from all three gustatory sites (posterior and anterior tongue and the NID)………………………………………………………..76

6. Gene expression in single taste cells…………………………………………………………………....77

xii

LIST OF FIGURES

Figure Page

1. Sample integrated whole chorda tympani nerve response to a 0.5 M NaCl solution taste stimulus illustrating response components and measurement of response amplitude ……………………………………………………………………………….……...78

2. Single labeling immunocytochemistry showing TRCs in rat fungiform papillae immunopositive for 5-HT and 5-HT1A ……………………………………...... 79

3. Double labeling ICC showing examples of TRCs in rat CV immunopositive for α- Gustducin, 5-HT1A, and overlaid images of the two…………………………………..….….80

4. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5- HT1A, GAD and overlay……………………………………………………………………….…..…….…81

5. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5- HT1A, SNAP25, and overlay……………………………………………………………………………...82

6. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5- HT1A, CCK-8, and overlay ………………………………………………………………………………..83

7. Double labeling ICC showing examples of TRCs in rat CV immunopositive for NPY, 5-HT1A, and overlay ..………………………………………………………………….….……………....84

8. Double labeling ICC showing examples of TRCs in rat CV immunopositive for NPY1R, 5-HT1A, and overlay ……………………………………………..…………….……….……..85

9. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5- HT1A, NCAM, and overlay ………………………………………………………….……………….…..86

xiii

10. Gel electrophoresis illustrating single cell RT-PCR products from seven taste receptor cells tested with a variety of primer sets corresponding to all seven phenotypic markers, including taste receptor genes. …………………………………..….87

11. Sample integrated whole nerve chorda tympani responses to lingually applied taste solutions …………………………………………………………………………………………………88

12. Sample integrated whole nerve CT responses to quinine and to ATP solutions at different concentrations ………………………………………………………………………………….89

13. Sample integrated CT responses to lingual quinine stimuli (0.3 M) prior to- (PRE), immediately after (11 and 14 min), and 45 min following a 10mg/kg PPADS injection …………………………………………………………………………………………………………..90

14. Effect of P2-purinoceptor antagonist PPADS on CT response to lingual stimuli….91

15. Sample chorda tympani responses to intravenous ATP injections (3 mg/kg) prior to-, immediately following-, and 45 min follow a 10mg/kg PPADS injection …...92

16. Effect of P2-purinoceptor antagonist PPADS on CT response to systemic ATP injection ………………………………………………………………………………………………………....93

17. Normalized summated data averaged from 11 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after a single bolus 300 µL jugular saline injection …………………………………………………………………………………….94

18. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 10 µg/kg …..…………………….95

19. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 25 µg/kg …………………….…..96

xiv

20. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 100 µg/kg …………………...…97

21. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 200 µg/kg ………………….…..98

22. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a NAD-299 solution at 175 µg/kg ……………………….……99

23. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a NAD-299 solution at 350 µg/kg ………………………..…100

24. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 90 minutes after (recovery) a 300 µL single bolus jugular injection of a NAD-299 solution at 700 µg/kg …………………………..101

25. Graphs showing mean percentage differences in CT responses to different taste stimuli taken before (pre-), immediately after (post-), and at 40 to 65 minutes (recovery) after injecting saline or WAY-100635 at various concentrations …..102

26. Graphs showing mean percentage differences in CT responses to different taste stimuli taken before (pre-), immediately after (post-), and at 40 to 65 minutes (recovery) after injecting saline or NAD-299 at various concentrations….……….103

27. Effect of WAY-100635 on mean blood pressure and heart rate …………………….104

28. Normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre),

immediately after- (post), and 30 to 90 minutes after (recovery) a 300 µL single bolus jugular injection of a ondansetron at 1000 µg/kg ………………………………...105

29. Illustration of the proposed mechanism of the relationship between serotonergic and purinergic signaling in the taste bud………………………………………………………..106

xvi

CHAPTER 1

INTRODUCTION

The sense of taste is essential to the survival of all animals. Ubiquitous among most species, the gustatory system has evolved to enable animals to quickly judge the palatability of ingested foodstuffs. In nature, toxins and decayed organic foodstuffs are often characterized by bitter and sour tastes respectively, whereas nutritive, calorie-rich foods are usually associated with sweet taste. In this perspective, the sense of taste provides animals with a rudimentary, innate resource to differentiate at the oral level between harmful and beneficial materials. Gustation originates in the oral cavity upon the ingestion of food and its breakdown into tastants, or small particles that can functionally stimulate receptors in cells known as taste receptor cells (TRCs), contained within the end organ of gustation, the taste bud. Downstream of gustatory stimulation is the generation of action potentials within TRCs, the eventual release of neurotransmitter(s), activation of sensory afferent nerve fibers, and the transmission of gustatory signals to the central nervous system. The physiological function of the taste bud was originally regarded as a detection and relay system for taste signals to the afferent sensory nerves, and the integration of those signals was believed to occur only at the central level in specialized brain loci. However, recent progress in gustatory research and the study of taste transduction mechanisms at the peripheral level has revealed the presence of neuromodulatory mechanisms mediated by an array of neurotransmitters and

neuropeptides that were found indigenously expressed among TRCs. It is now acknowledged that the taste message is modulated at the level of TRCs before the final output, through an elaborate network of autocrine and paracrine intracellular signaling involving excitatory and inhibitory feedback mechanisms. Therefore, the integration of taste information, which determines the intensity and identity of the afferent taste signal, begins at the peripheral level, before the gustatory message is deciphered in the gustatory cortex. In the past few years, focus has been increasingly shifting back towards the exploration of the intrinsic neuromodulatory activity occurring within the taste bud, which is now acknowledged to be indispensable for comprehending the full spectrum of gustatory physiology. In this context, the role serotonin plays in the taste bud warrants special consideration. One of the first neuroactive molecules discovered and characterized in the bud, serotonin was long believed to be the neurotransmitter of taste information between TRCs and afferent nerve fibers1,2. However, new findings eventually prompted a change in this belief, when it was discovered that ATP, rather than serotonin, is the primary gustatory neurotransmitter at the periphery3, leaving the door open to determine the actual role of serotonin in the taste bud. Perhaps, one of the first clues to understanding serotonin’s function in the taste bud came from previous morphological work that demonstrated the expression of serotonin receptors in TRCs4, suggesting the involvement of serotonin in intrinsic modulatory events. This view was bolstered by early behavioral studies in humans that reported altered taste thresholds in patients taking serotonin-specific reuptake inhibitors (SSRI)5,6. In addition, serotonin had been shown to contribute to neuromodulatory circuits and to modulate neurotransmission in the olfactory epithelium and areas of the CNS7,8. More importantly, the amount of released ATP from mouse taste buds was recently found to be affected by serotonergic

stimulation9, hinting to a possible modulatory role for serotonin by influencing the neural output of taste signals. The present work hypothesizes that serotonin’s main role in the taste bud is to modulate taste signals presynaptically by directly influencing the release of the gustatory neurotransmitter ATP from type II TRCs. The study researched the location of the serotonin receptor, 5-HT1A, among TRCs of the bud through immunological and molecular approaches, and examined the net functional effect of serotonin on the flow of gustatory signals by recording neural responses to taste stimulation following the antagonism of serotonergic receptors.

1.1. Taste buds are the end organs of gustation Mammalian species respond to five different taste qualities: sweet, bitter, salty, sour and umami (savory). The functional unit of gustatory chemoreception is the taste bud, containing 50 to 100 closely clustered taste receptor cells (TRCs) with their apex in contact with taste pores on the tongue’s epithelium10. In mammals, taste buds are mostly present on the tongue, but are also found in smaller numbers on the epiglottis, soft palate and upper esophagus10,11. At the surface of the anterior tongue, taste buds are contained in specialized mushroom-like structures known as fungiform papillae12, present mostly at the tongue’s apex. Each fungiform papilla contains a single taste bud in rodents10, and an average of 2 to 4 in humans13. In the posterior tongue, special structures known as circumvallate papillae (CV) form an “O” or a “C” shaped trench that invaginates the tongue’s dorsal epithelium. Each papilla contains 100 to 125 buds, located around the walls of the trench. In humans, 6 to 8 CV papillae form a symmetrical V-shape pointing caudally and terminating near the epiglottis10, whereas in rodents, only one CV papillae is situated at the medial dorsal surface of the tongue 14,15. In both humans and rodents 50 to 125 buds are housed in the clefts of two foliate papillae, each

located on either lateral margins of the posterior tongue16,17. Tactile and thermal stimulations of the tongue are detected by filiform papillae, which lack taste buds and are the most abundant papillae on the surface of the tongue18. Most TRCs in the bud project gustatory microvilli filaments into the lumen of the papilla at the apical side of the bud10. At the basal side, a branch of the afferent gustatory nerve fiber synapses with a group of cells within each taste bud. Taste signals generated in taste buds of the anterior two thirds of the tongue, as well as buds in the palate and naso-incisor duct (NID), are transmitted to the CNS via branches of the chorda tympani (a branch of the facial nerve- CN VII), whereas signals from the posterior third are carried by the glossopharyngeal nerve19,20,21 (CN IX). In the early history of gustatory research, cells in the taste bud were initially classified into four types based on their morphology using electron microscopy 22: Type I cells, also referred to as “dark cells” due to the presence of electron-dense granules in the apex of their cytoplasm23,24. Type II cells, which appear lighter in contrast, with an opaque cytoplasm, and hence are referred to as “light cells”. These cells are characterized by a conspicuously large round nucleus, numerous mitochondria, and prominent smooth endoplasmic reticula. Type III cells, which feature ovoid nuclei and an elongated shape with an intermediary opacity; these are the only cells in the bud that form direct synaptic connections with the afferent nerve fibers. Type IV cells are small, almost circular shaped cells that occupy the basal part of the bud, and are considered the progenitor cells that develop and eventually differentiate into mature TRCs23,24,25,26.

1.2. Gustatory transduction: Before the advent of gene cloning in gustatory research, it was assumed that only type III cells had the ability to detect and process multiple taste stimuli, and then transmit gustatory signals to the sensory afferent nerves via their synaptic connections.

Early electrophysiological experiments revealed that TRCs could respond to a combination of one to four taste qualities27. Signals were thought to be transmitted from TRCs to afferent nerves in a linear fashion. The cloning of taste receptors however prompted a change in this view, when it was discovered that receptors for different taste qualities were expressed in non-overlapping subsets of TRCs28,29. Indeed, different subsets of type II cells, which lack synaptic connections with sensory afferent nerves, were found to express multiple taste receptors30,31, suggesting that all taste modalities are not transduced in type III cells. It also meant that classical synapses with sensory afferent nerves are not essential for the transmission of taste signals. Indeed, distinct subsets of type II cells express different metabotropic receptors for sweet, umami, and bitter tastes in a mutually exclusive pattern. In rodents and humans, the T2R family of metabotropic receptors, coupled to the G-protein gustducin32,33,34,35,36, respond to bitter compounds, whereas sweet and umami tastants are detected by members of the T1R 37,38 family of receptors, coupled mostly to Gα14 rather than gustducin . T2Rs and T1Rs are each expressed in 20 – 30% of type II cells29. Sour stimuli are believed to be transduced in type III cells through the activation of two co-expressed trp channels, PKD2L1And PKD1L3, found exclusively in those cells39,40,41,42. The identity of salt taste receptors in humans remains controversial, and its transduction is not well understood. Humans only perceive sodium and lithium as salty tastes, but other non-sodium salts (e.g. potassium, magnesium, calcium) also have a salt taste component in addition to other tastes. In rodents, transduction of salt is relatively well defined. Salty stimuli are transduced in the taste buds of the anterior tongue by two mutually exclusive subpopulations of TRCs, amiloride-sensitive43 and amiloride- insensitive44 TRCs. The former consist mostly of epithelial sodium channels (ENaC)45,46,47 known for their susceptibility for inhibition by diuretic drugs such as amiloride48,49.

ENaCs are activated by the entry of sodium, lithium and protons, and they are directly gated by sodium concentration. Cell depolarization is triggered by entry of Na+ ions into the cell causing an increase in intracellular calcium levels and neurotransmitter release50. However, those channels are not universally present in the gustatory systems of all species51,52,53, and they are not functional in all their sites of expression across the tongue’s gustatory fields44,54,55. Although the identity of ENaC-expressing TRCs isn’t presently known, previous electrophysiological work has suggested that their expression is exclusionary of type III TRCs56. Salt taste transduction in TRCs of the anterior tongue in rodents is mostly mediated by amiloride sensitive mechanisms48,55, whereas in the posterior tongue, it does not appear to involve amiloride-sensitivity44,54, and is relatively less well understood. At present, the identity of the amiloride-insensitive TRCs and mechanisms is not known, but was shown to involve in part a variant of the transient receptor potential vanilloid receptor 1 (TRPV1) that serves as a nonspecific cation channel57,58.

1.2.1. Signaling agents within the taste bud Despite the identification of taste receptors, the prevailing belief remained that type III cells were the obligatory route for the transmission of taste signals to sensory afferent nerves. But this belief gradually changed as studies began to demonstrate that TRCs respond to endogenous and exogenous modulations. The latter occurs through the activity of circulating hormones including leptin59,60, endocannabinoids61, insulin62, serotonin5 on receptors specifically expressed in TRCs. Exogenous modulators may interfere with the transduction cascades of single or multiple taste modalities and hence alter taste sensitivities and perceptions. Endogenous modulation is mediated by a host of neurotransmitters (serotonin63,64,65, glutamate66, norepinephrine67,68, acetylcholine69, ATP70, and gamma- Aminobutyric acid71), neuropeptides (cholecystokinin72, neuropeptide Y73, vasoactive

intestinal peptide74,75, glucagon-like peptide76, and galanin77). It was discovered that subsets of TRCs express receptors for those agents, and that their activation can occur either in autocrine or paracrine manners, resulting in cellular excitation or inhibition. It then became evident that an extensive network of cell-to-cell interactions was present among cells of the taste bud, which prompted a shift in the traditional view that taste signals proceed in a linear fashion from type III cells to sensory afferent nerve. It is currently accepted that taste signals undergo extensive modulation before their transmission to sensory afferent nerves.

1.2.2. Early and late gustatory transduction mechanisms Sensory taste transduction can be categorized into early and late mechanisms78. The former is the process where a tastant stimulates a TRC through its corresponding receptor causing depolarization and eventually creating an action potential. At the initial stages, tastants interact with receptors to cause changes in the TRC electrical excitability, culminating in the production of an action potential. Events downstream of T1R or T2R receptor stimulation by sweet, bitter, or umami tastants are now well characterized. Ligand (tastant) binding activates the enzyme phospholipase C β2 (PLCβ2), ultimately causing calcium release from intracellular stores and the activation of TRPM5 ion channels, resulting in an influx of Na+ and cell depolarization79,80,81,82. Late transduction mechanisms comprise the series of modulatory mechanisms that shape and refine the taste signal before its transmission to afferent nerves. While events that take place in the early stage of gustatory transduction are relatively well understood, much remains to be learned about the mechanisms underlying modulatory events that take place in the late stage following TRC depolarization, a topic that has recently become the subject of increased research focus. Cell-to-cell communication and taste signal modulation take place during this stage of gustatory transduction, mediated by the actions of the different signaling agents previously mentioned. The

release of serotonin and its consequent effect on TRCs expressing serotonin receptors is one example of late transductory events.

1.3. Serotonin in the taste bud Serotonin, or 5-Hydroxytryptamine (5-HT), is a monoamine neurotransmitter that is primarily found (over 90%) in the enterochromaffin cells in the GI tract and to a lesser degree in the CNS. Outside the CNS, serotonin plays critical roles in physiology of multiple organ systems, including the regulation of food intake and energy balance83, gastrointestinal and endocrine function84, and cardiovascular and pulmonary physiology85. In the brain, serotonin is released mainly from the neurons of the raphe nuclei, all of which are located along the midline of the brainstem and around the reticular formation86. From this point, serotonergic axonal projections radiate to almost every part of the CNS where it modulates neural activity and a wide range of neuropsychological processes87. Serotonin is involved in modulating appetite, metabolism, sexual behavior, mood, body temperature, perception, reward, memory, sexuality, and sleep88,89,90,91. Notably, serotonin plays a modulatory role in sensory systems. In the olfactory bulb of the moth, it modulates the activity within neural circuits of the antennal lobes7. In mice, endogenous serotonergic transmission regulates odor inputs in the olfactory bulb by attenuating odor-evoked synaptic input to glomeruli through activation of the 5- HT2C receptors. Serotonin was also found to inhibit glutamate release from olfactory 8 receptor neuron terminals via GABAB receptors . In the gustatory system, a recent study in humans showed a reduction in sucrose taste threshold and an even bigger reduction in bitter taste threshold when body 5-HT levels are enhanced5. In addition, patients with depression, anxiety, and panic disorders,

as well as individuals taking drugs that change monoamine levels (5-HT and NA), all report altered tastes6. At the cellular level, serotonin was one of the first well characterized agents in the taste bud. PCR and immunolabeling experiments in the posterior tongue and immunolabeling experiments in anterior tongue demonstrated the localization of serotonin in type III cells of mouse, rat, rabbit, and monkey taste buds92,93,94,95,96,97. And since type III cells are the only cells in the bud that appeared to synapse with afferent nerves, and were considered at that time to be the main output from the bud, it was thought that serotonin was the main neurotransmitter of TRCs. Subsequently, serotonergic transmission was widely accepted as the main way gustatory signals are transmitted from type III cells to sensory afferents. However, emerging key findings eventually prompted a change in the view of the role of serotonin in the gustatory periphery.

1.3.1. 5-HT receptors are expressed in the taste bud Although serotonin receptors were expected to be found on gustatory neural afferents, one of the serotonergic receptors, 5-HT1A, was surprisingly found to be expressed in TRCs. Preliminary work in rats, surveying the posterior tongue’s gustatory epithelium for serotonin receptors, using immunolabeling and whole taste bud reverse transcriptase PCR experiments, revealed that of the fourteen serotonergic receptor 4,98 subtypes only 5-HT1A and 5-HT3 were expressed within the taste bud . In those experiments, 5-HT3 receptors were detected only through reverse transcriptase PCR, as immunocytochemistry alone using sections from CV and foliate papillae failed to reveal their location in TRCs. This has led to the conclusion that 5-HT3 may be expressed only on postsynaptic nerve endings, which are undetectable through the relatively low resolution of conventional light microscopy. In support of this notion was the fact that extraction of whole taste buds for RT-PCR experiments inevitably retains post-synaptic

elements, including fragments of the afferent nerve endings. This may explain the expression of mRNA for 5-HT3 in RT-PCR screening experiments and the absence of 5-

HT3-immonopositive cells in ICC experiments. The assumption that 5-HT3 is confined to gustatory nerve endings is also supported by other data from immunolabeling and RT-

PCR work which found 5-HT3 receptors expressed in sensory afferent nerve fibers in the periphery, including the submucous plexus, superior cervical ganglion, nodose ganglion, 99 myenteric plexus, and the dorsal root ganglion . In addition, others have found 5-HT3 receptors expressed in cell bodies of primary sensory neurons of the petrosal ganglion in rats100. These pseudounipolar neurons have peripheral branches that innervate cells in taste buds of foliate and CV papillae, and transportation of 5-HT3 across the neuronal cell bodies to afferent terminals may further explain its likely presence in whole taste bud RT-PCR products.

In contrast to 5-HT3, when the expression of the 5-HT1A receptor subtype was examined using ICC, results showed clearly labeled cells within the bud4. Additionally, when the co-expression of 5-HT and 5-HT1A was examined in the same tissue sections using double label ICC, it was revealed that the two molecules are expressed in non- overlapping subsets of TRCs, suggesting that 5-HT acts on its 5-HT1A receptor in a paracrine manner. Furthermore, 5-HT biosynthesis and reuptake enzymes are also expressed in the taste bud101,102 suggesting that the source for 5-HT in the taste bud is the bud itself, and not from a systemic supply. In response to these findings, alternative theories began to suggest a modulatory role for serotonin in the taste bud, and research began to explore the effect of its release on TRCs. One of the earlier key studies conducted patch-clamping experiments to investigate the possibility that serotonin may influence the electrical excitability of the mammalian TRC. The study found that bath-applying serotonin or serotonin agonists produced inhibition of calcium-activated potassium currents and of voltage-dependent

sodium currents in the TRCs of rat64. In another follow-up study98, these same effects were mimicked by the use of agonists for the 5-HT1A receptor, 1-(1-Naphthyl)piperazine hydrochloride and 8-OH-DPAT ((±)-2-Dipropylamino-8-hydroxy-1,2,3,4- tetrahydronaphthalene hydro-bromide). These data led to the hypothesis that serotonin may affect electrical properties of the taste cell during active stimulation, and suggested a modulatory role for serotonin in late taste transduction events by exerting an inhibitory action on target TRCs.

1.3.2. Serotonin is released in the taste bud following gustatory stimulation A recent study that examined the incidence of serotonin release in mouse taste buds found that it was released upon gustatory stimulation of serotonergic and/or non- serotonergic cells2. By using receptor-bound biosensors, researchers showed that serotonin was released when the buds were depolarized with KCl salt, or stimulated with bitter, sweet, or sour tastants. The mechanism for bitter or sweet-induced release differed from sour-induced release, suggesting that these two groups of qualities activate the 5-HT cell differently. As type III cells are considered to be involved in sour detection it seemed logical that they also release 5-HT following sour stimulation. What was intriguing however, is the fact that they release 5-HT when sweet and/or bitter type II cells are stimulated. In the same study, serotonin release was tested in taste buds isolated from PLCβ2-null mice2 (as previously mentioned, PLCβ2 is an enzyme downstream of T1R/T2R stimulation and is essential for the transduction of sweet, bitter and umami tastes in type II cells). Depolarization by KCl salts still evoked serotonin release from those taste buds, but sweet and bitter taste stimuli did not. Although serotonergic cells do not express sweet or bitter receptors (T1R or T2R), genetic elimination of their downstream cascade in type II cells greatly reduced serotonin release from type III cells in response to bitter and sweet stimulation. This implies that signals originating from type II cells may be necessary for the release of 5-HT from

serotonergic type III cells. Nevertheless, despite the wealth in new findings from the study of serotonergic transmission in the taste bud, they were still too limited to warrant a change in the conventional view on serotonin as the bud’s main neurotransmitter. However, data emerging from the study of purinergic transmission and cell signaling in the taste bud eventually confirmed that ATP, and not serotonin, is indeed the main gustatory afferent neurotransmitter from TRCs.

1.4. ATP, purinergic transmission, and the relationship with serotonin When gustatory tissues were surveyed for purinergic components, it was revealed that gustatory nerves express two ionotropic P2X receptor subunits (P2X2 and 103 P2X3) , indicating that ATP may serve as a neurotransmitter at this level. Furthermore, 70,104,105,106 TRCs were found to express P2X and P2Y purinoceptors , suggesting that ATP might be involved in cell-to-cell communication mechanisms among TRCs in the bud. In a recent key study researching purinergic transmission in the taste bud, recordings from gustatory nerves and two-bottle preference tests in mice lacking

P2X2/P2X3 receptor expression showed that animals became indifferent to sweet, bitter, and umami tastes3. In addition, researchers demonstrated that stimulating stripped gustatory epithelium with bitter tastants induces the release of ATP. In contrast, stimulation of non-gustatory (non-taste bud) epithelium with bitter stimuli caused little or no ATP release. This result, coupled with the behavioral indifference of P2X-KO mice to a vast majority of taste stimuli, suggested a crucial role for ATP as a gustatory neurotransmitter. More recently, another study examined ATP release from isolated mouse TRCs using receptor-bound biosensors, and found that gustatory stimulation of isolated TRCs evokes ATP release from type II cells, and serotonin release from type III cells107. Since

type II cells lack synapses with sensory afferents, taste-evoked ATP has been suggested to occur via pannexin (Px1) or connexin hemichannels108,109 expressed in taste tissue, thereby bypassing the need for vesicular exocytosis through classic synapses. In fact, type II cells are known to exhibit prominent endoplasmic reticula and atypical mitochondria clustered at sites of apposition with nerve fibers17. It has been proposed that these specialized "synaptic ensembles" are part of the high energy machinery for interactions between TRCs and afferent nerve fibers. This machinery provides type II cells, which lack classical synapses with gustatory afferent nerves, the capability to transmit their natively generated taste signals, bypassing type III cells, the presumed obligatory output of the taste bud. Another key study hypothesized that if type III cells –which are both serotonergic and PKD2L1-expressing in their majority– were indeed the bud’s only output, and 5-HT was the main neurotransmitter from TRCs, all tastes would be expected to be abolished when those cells were eliminated39. By genetically engineering mice with targeted genetic ablations of PKD2L1-expressing TRCs, researchers noticed that the animals became completely devoid of taste responses to sour stimuli, yet, sweet, bitter and umami taste perceptions were not affected. These results suggest that at least sweet, salt and umami taste signals are bypassing type III cells in transmitting their signals to target afferent nerves. Collectively, these findings demonstrate that signals originating from type II TRCs, which are the source of ATP release in the bud, are more crucial for the flow of gustatory information than those released by type III TRCs, the source of 5-HT in the bud. Hence, type II, and not type III TRCs, are the primary output of the taste bud, and ATP, and not serotonin, is the obligatory neurotransmitter of gustatory signals from TRCs to sensory afferents. In addition to cementing the view of the role of ATP as the key gustatory neurotransmitter, these findings also propose a neuromodulatory role for

purinergic activity in intracellular signaling among TRCs. ATP’s relationship to serotonin in the taste bud has been recently investigated in a study that examined the impact of released serotonin on ATP release. Researchers stimulated isolated mouse type II TRCs with either serotonin or a serotonin receptor (5-

HT1A) agonist, and found a reduction in ATP release. Also, the same results were produced when a serotonin reuptake inhibitor was applied to the batch medium.

Blocking serotonergic transmission by bath-applying a broadly acting 5-HT1/2 receptor antagonist, or a 5-HT1A receptor antagonist, enhanced taste-evoked ATP release from taste buds. In contrast, bath-applying a potent and selective 5-HT3 receptor agonist did not affect taste-evoked ATP secretion. The study concluded that serotonin may be inhibitory to ATP release from type II cells through negative feedback, possibly mediated by 5-HT1A receptors. The current prevailing opinions support the view that upon gustatory stimulation, ATP (released from type II cells) and serotonin (released from type III cells) both engage in a paracrine and autocrine feedback mechanism that modulates taste signals before their transmission to gustatory nerves. It is now widely acknowledged that taste-evoked ATP plays the main role of gustatory neurotransmitter, mediated by 3,103 P2X receptors on afferent nerve fibers , bypassing the requirement for direct synapses, exclusive to type III cells107,108,109. ATP’s role in cell-to-cell-communication appears to occur through its activation of P2Y and P2X receptors present on TRCs9,70,105,106,110. The study of purinergic and serotonergic transmission strongly points to the involvement of ATP, purinoceptors, serotonin and the 5-HT1A receptor in the putative modulatory feedback loop.

Serotonin is known to co-express in a subset of type III cells with PKD2L1, a candidate sour taste receptor39. As previously mentioned, type III cells do not release ATP, and since blocking purinergic transmission abolishes all tastes, including sour

detection3, it is evident that sour signals originating from type III cells are still being communicated to type II cells, and then transmitted -presumably through ATP- to sensory neural afferents. It is believed that 5-HT is the neurotransmitter that type III cells release upon their stimulation by sour compounds, and therefore, it is a likely assumption that 5-HT receptors are expressed in type II cells. As already stated, only the

5-HT1A receptor subtype is expressed in TRCs, and serotonergic type III cells do not express any serotonin receptors. These observations, coupled with the preliminary ICC data showing the non-overlapping expression of 5-HT and 5-HT1A, demonstrate that 5-

HT targets 5-HT1A receptors in a paracrine manner within the taste bud. One way to understand how 5-HT operates in the bud is to identify the co- expression pattern of its receptor, i.e. to phenotype the 5-HT1A-expressing cell. Since much is known about the location and function of certain TRC-specific markers in the bud, examining the expression of 5-HT1A relative to those markers can help in revealing the identity of TRCs targeted by serotonin release, and hence the type of taste modalities modulated by serotonin. The current study seeks to identify the TRCs and the taste qualities targeted by serotonin in taste buds of the anterior and posterior tongue, and the naso-incisor duct

(NID) in rats. This is achieved by phenotyping the 5-HT1A-expressing cell using immunocytochemistry and single cell reverse-transcriptase-PCR to examine the pattern of expression of 5-HT1A relative to distinct phenotypic markers associated with specific types of cells, and with the transduction of specific tastes. The study also explores the physiological role of serotonin in the bud. At the level of TRCs, serotonin inhibits in a paracrine manner the ubiquitous release of taste- evoked ATP from type II cells following gustatory stimulation. Since gustatory stimulation also evokes 5-HT release from type III cells, it is predicted that 5-HT may modulate the transmission of taste signals via its influence on the amount of

neurotransmitter (ATP) release. The present study explores this hypothesis and examines serotonin’s mode of action in the taste bud, and the scope of its influence on the flow of gustatory signals. The majority of data gathered from the previous studies examining serotonin’s function in the taste bud emerged mostly from in vitro physiological experiments performed using dissociated taste cells, engineered biosensors, and involving the manipulation of the cells’ surrounding media with exogenous reagents. In contrast, the present study explores in real-time at the neural level, the effect of blocking serotonergic transmission on gustatory signals. Using whole nerve chorda tympani recordings, neural responses to taste stimulation are compared before and after intravenous injections of serotonin receptor antagonists, thus allowing study of both the immediate and long-term impacts of serotonin on the transduction of different taste modalities.

CHAPTER 2

METHODS

Three major experimental strategies were employed to investigate the role of serotonin in the taste bud. These were single and double label immunocytochemistry, single cell reverse transcriptase PCR, and electrophysiological whole nerve recordings from the chorda tympani. Adult male Sprague–Dawley rats served as subjects for all experiments. All procedures were approved by The Ohio State University's Institutional Animal Care and Use Committee (IACUC).

2.1. Immunocytochemistry Immunocytochemistry (ICC) double-labeling experiments were performed to phenotype 5-HT1A-expressing TRCs by examining the expression pattern of 5-HT1A relative to specific TRC markers linked to specific taste qualities.

2.1.1. Subjects For tissue harvest, animals were induced into general anesthesia with a 0.09 ml/100 g body weight injection of a mixture of 91 mg/ml Ketamine Hydrochloride (Hospira, Lake Forest, IL, USA) and 0.09 mg/ml Acepromazine Maleate (Phoenix Pharmaceutical, Inc., St. Joseph, MO, USA). Animals were sacrificed by decapitation, and gustatory tissue from anterior tongue (AT) where fungiform papillae are found, naso- incisor duct (NID), and posterior tongue (including the two foliate papillae and single

circumvallate papilla) were quickly excised and fixed by immersion in Bouin's fixative for 5 h at 4 °C. For immunocytochemical experiments investigating 5-HT localization, TRC’s were preloaded with 5-hydroxytryptophan (5-HTP; Sigma-Aldrich Biotechnology, St. Louis, Missouri, USA) at 80 mg/kg by intraperitoneal injection of animals 1 hour before they were sacrificed. 5-hydroxytryptophan (5-HTP) is a naturally occurring amino acid that acts as a chemical precursor for the biosynthesis of 5-HT from tryptophan by decarboxylation. 5-HTP injection before 5-HT immunolabeling is a widely used approach proven to produce stronger 5-HT signals in cells that express 5-HT111,112.

2.1.2. Tissue preparation Tissue blocks were dehydrated, embedded in paraffin for 24 hours, and then sectioned on a rotary microtome at 4 μm thickness. Every series of four sections was serially collected on a Fisherbrand Superfrost Plus glass slide (Fisher Scientific, Pittsburgh, PA, USA). All slides were numbered according to their order in the series. Slides were left to dry at 50 °C for 24 h on a slide warmer. On the day of the experiment, sections were deparaffinized by immersion in xylene (2 x 20 min) and rehydrated in an ethanol gradient (100°, 95°, 70° and 50°; 10 min each) followed by 0.01 M phosphate- buffered saline (PBS; pH 7.4; 3 x 10 min) prior to their processing in either conventional or TSA-amplified immunocytochemistry experiments depending on the nature of the experiment.

2.1.3. Conventional immunocytochemistry protocol Immunocytochemical experiments were performed using either single or double-label protocols. Methods for double-labeling experiments fell into two categories: those in which the two primary antibodies were raised in the same species, and those in which they were not. Conventional double-label ICC protocol was followed

in the former case. After tissue rehydration a barrier was applied around sections on each slide using a hydrophobic barrier pen. A 150 µL blocking solution consisting of 10% normal serum diluted in PBS was then applied on each slide for 1 h at room temperature to minimize non-specific antibody binding during subsequent steps. Donkey or goat whole serum was used to match the species in which the secondary fluorescent antibody was raised. Sections were then incubated in primary antiserum (150 µL per slide, diluted in PBS containing 2% normal serum) at the appropriate dilutions (table 1). Slides were housed in a closed moist chamber overnight at 4 °C to prevent dehydration. The following day, sections were washed in PBS (3 x 10 min) and then incubated with Cy3- or FITC-conjugated secondary antibody (150 µL per slide, at 1:400 in PBS containing 1.5% normal serum) at room temperature for 1 h in the dark. Slides were then washed in PBS (3 x 10 min) and incubated in the second primary antiserum at the appropriate dilution for 24 h at 4 °C. The following day, slides were washed in PBS and incubated with the second secondary antibody for 1 h in the dark and washed again with PBS and finally mounted in Fluoro-Gel and observed under a fluorescent microscope equipped with a digital camera (Nikon EFD-3, Japan).

2.1.4. Tyramide signal amplification (TSA) immunocytochemistry When it was not possible to acquire two primary antibodies raised in different species due to commercial availability, an indirect immunofluorescence double-labeling protocol was employed. This procedure allows detection of two antigens with primary antibodies raised in the same species by altering the dilution and detection method for each antigen. Tyramide Signal Amplification (TSA Biotin System, PerkinElmer, Waltham MA, USA) is an enzyme-mediated detection technique that relies on the catalytic activity of horseradish peroxidase (HRP) to generate high-density labeling of a target protein in situ. One antibody is used at very high dilution with TSA detection whereas the second primary is used with standard dilution and detection methods. This results in a

significant increase in sensitivity without loss of resolution or increase in background. Streptavidin-conjugated IgG Fab fragment (instead of a fluophore-conjugated one) is used to detect the first primary antibody so that the second primary antibody can be used at very low concentrations thus minimizing interference or cross-reaction between the first primary antibody and the second secondary antibody113,. This method also prevents the first secondary antibody from reacting with the second primary antibody. For TSA experiments, tissue sections were first incubated with a solution of 0.5% hydrogen peroxide in methanol for 30 min to eliminate endogenous peroxidase activity. Sections were then incubated for 1 h at room temperature in PBS containing 10% normal serum and 0.3% Triton X-100 to reduce nonspecific antibody binding. Primary antiserum (diluted in PBS containing 2% normal serum) was subsequently applied to the sections and the slides were housed in a closed moist chamber for 36 h at 4 °C. After PBS washing, sections were incubated with biotin-streptavidin-conjugated IgG Fab fragment (1:800 in PBS containing 1.5% normal serum) for 1 h at room temperature and processed according to the kit’s instructions (NEN Life Science Products, Boston, MA, USA). Slides were then mounted in Fluoro-Gel and observed under the fluorescent microscope.

2.1.5. Experimental controls Control slides were included in every experiment to ensure that fluorescent signals did not arise from cross-reactivity of secondary antibodies with the inappropriate primary antibody. To control for false positives and non-specific binding in single-label ICC experiments, two slides were included in each experiment where the first slide lacked the primary antibody, and the second slide lacked the secondary antibody. Absence of a visible signal in those slides upon microscopic observation excluded the occurrence false positive labeling. Similarly, for double-label ICC experiments, including TSA experiments, four slides were included where the first slide lacked the first primary,

the second lacked the second primary, the third lacked the second primary, and the fourth slide lacked the second secondary antibody. In all control slides, cells were expected not to show any positive staining. As a criterion for acceptable labeling, the signal for a target antigen was expected to be visible only on the channel matching the corresponding secondary antibody. Color bleeding to the channel matching the alternate secondary antibody (i.e. when the same cells appear to be labeled on two non- overlapping channels) implied that false positives could be mistakenly interpreted as cells labeled for the other target antigen being investigated. In this case, adjustments were made in the dilutions of the primary antibody until color bleeding ceased to be visible under the channel corresponding to the other secondary antibody.

2.1.6. Cell counting and data analysis Data from anterior and posterior tongue and NID were generated from ICC experiments performed on tissue sections from eight animals using paraffin-embedded anterior tongue sections (containing fungiform papillae), naso-incisor duct (NID) sections, and posterior tongue sections (containing CV and foliate papillae). Double- label ICC experiments were analyzed by counting individual taste buds and taste receptor cells. To avoid counting the same cell more than once in a single 4-section slide, only every fourth section on each slide was chosen for data analysis. This ensured that the sections being considered for cell counting were separated by at least 10 μm, greater than the diameter of a taste cell. Cell counting was repeated twice to confirm the count. Only cells that showed clear apical and/or perinuclear labeling were considered immunopositive. All cell counts were performed under the 40X magnification objective lens. To examine if expression of the antigens in question varied among gustatory fields innervated by different nerves, data obtained from anterior tongue and NID (“anterior field” innervated by the chorda tympani), and those from posterior tongue (“posterior field”, innervated by the glossopharyngeal nerve) were

analyzed separately. A limitation for direct microscopic cell counting was that it only accounted for stained cells within the visible plane of the 4 µm-thick section and it did not fairly represent all labeled cells within the actual tri-dimensional frame of the taste bud. Antigen-expressing cells cannot be presumed to be uniformly distributed throughout the taste bud. Consequently, consecutive tissue sections would most likely show discrepancies in the number of labeled cells. This method therefore serves as an approximate representation of the actual number of immunopositive cells.

2.1.7. Antibodies The antibodies used in ICC experiments included markers for type II and type III taste cells. Primary antibodies for gustducin, NPY, NPY1R, SNAP-25, GAD, and CCK8 were used as type II cell markers, and for NCAM and 5-HT as a type III marker. Antibody descriptions and dilutions are presented in Table 1. The dilution for the antibodies used in conventional and in amplification ICC were empirically determined. Adjustments in dilution were made until no cross reactivity or background staining was seen, in which case, a single labeling dilution series of the primary antibody was carried out in an independent ICC experiment to determine optimum dilution.

2.1.8. Image processing Digital photos of immunofluorescence cells were captured and processed using MetaMorph image analysis software (Molecular Devices LLC, Sunnyvale, CA, USA). Exposure, contrast and brightness were all optimized on a slide by slide basis. Images were captured under the 40X objective lens by a dedicated digital camera (Photometrics CoolSNAP. Tucson, AZ). Images were imported into imaging software (Canvas Illustrator; ACD systems, USA) where they were cropped, adjusted for size, and supplemented with a standardized 20 µm scale bar to produce illustrative figures.

2.2. Reverse transcriptase PCR (RT-PCR) RT-PCR experiments were performed to further investigate the expression of specific markers associated with different types of taste cells in the effort to phenotype the 5-HT1A–expressing cell. These experiments also served to confirm data from ICC experiments and to study the expression of 5-HT1A relative to different taste receptors. Primers used were first tested in control RT-PCR using RNA extracted from individually harvested circumvallate and foliate papillae. For single cell RT-PCR, RNA was extracted from single cells isolated from circumvallate papillae. In both cases, extracted RNA was used to synthesize cDNA, followed by PCR using primers for specific cell markers. RT-PCR products were then analyzed via agarose gel electrophoresis.

2.2.1. Whole papilla RNA extraction Adult male Sprague–Dawley rats (n = 5) were anesthetized and decapitated, their tongues removed and individual papillae (1 x circumvallate + 2 x foliate per animal) excised. Each papilla was trimmed down to approximately 5 mg for each reaction, weighed, and suspended in extraction buffer and then homogenized with a tissue grinder. Total RNA isolation from individually harvested circumvallate or foliate papillae was carried out using the Quick-RNA MiniPrep (Zymo Research Corp. CA, USA). Each tissue suspension was transferred to an RNA spin column supplied in the kit and centrifuged at 14,000 x g for 1 minute. RNA pre-wash buffer was added (400 µl per column) and tubes were centrifuged again (14,000 x g for 1 minute). RNA wash buffer was then added (700 µl per column) and tubes were centrifuged (14,000 x g for 30 seconds). This last step was repeated again with a 400 µl wash buffer. To completely remove the wash buffer, tubes were centrifuged again for 2 minutes at 14,000 x g. RNA was eluted in 40 µl of nuclease-free water in RNase-free tubes by centrifuging at 15,000 x g for 30 seconds.

2.2.2. cDNA synthesis Prior to cDNA synthesis template RNA was desalted and purified from potential genomic DNA contamination. One microliter of RNA was treated with the following components: DNase I, RNase-free (1 µl), 10 x reaction buffer with MgCl2 (1 µl), and 10 µl nuclease-free water. The mixture was incubated for 30 min at 37 °C. A one microliter of 50 mM EDTA was added and the mixture was incubated at 65 °C for 10 minutes. First-strand cDNA was reverse-transcribed from RNA templates using the RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific Life Research Products, Logan, UT, USA): for the annealing step, synthetic oligo (dT)18 primer (1 µl) was added to template RNA, and the mixture was topped-up to a volume to 12 µl with nuclease- free water. For the extension step, the following reagents were added to the mixture in the following order: 5 x reaction buffer (4 µl), RiboLock RNase inhibitor (1 µl), 10mM dNTP mix (2 µl), and finally, RevertAid M-MuLV reverse transcriptase (1 µl). The 20 µl mixture was incubated for 60 min at 42 °C. The cDNA extension reaction was terminated by heating at 70 °C for 5 min.

2.2.3. Single-cell RNA extraction and cDNA synthesis To phenotype the expression pattern of single TRCs, RNA was first amplified in a linear fashion prior to cDNA synthesis using the MessageBOOSTER cDNA Synthesis from Cell Lysates Kit (Cat. No. MBCL90310, EPICENTRE Biotechnologies, Madison, WI, USA). The taste receptor cells used in these experiments were isolated from the taste buds of individual CV papillae that were collected from ten animals under microscopic examination. Isolated cells were dissociated from excised tissue by incubation in cysteine-activated (1 mg/ml) Papain (14 U/ml), divalent-free solution (in mm: 80 NaCl, 5 KCl, 26 NaHCO3, 2.5 NaH2PO4·H2O, 20 d-glucose, and 1 EDTA). Each TRC was carefully suctioned into an ~8 µm-diameter glass pipette in less than 20 nanoliter ECF solution. To verify the presence of only one TRC, each pipette tip was examined under a 200X

magnification lens of a bifocal optical microscope (Nikon, DIAPHOT 200, Japan). The majority of extracted single cells had an appearance morphologically consistent with type II cells, marked by a rather circular shape and short cellular processes. The solution containing the single cell was then expelled into a PCR tube containing 3 μl of the kit’s quick-extract RNA extraction solution. The cell was lysed by vigorous vortex mixing for 1 minute at room temperature. First-strand cDNA was reverse-transcribed from template RNA by adding T7-Oligo(dT) primer and annealing at 65 °C for 5 min and then incubating the reaction with MMLV reverse transcriptase at 37 °C for 60 min. Second-round cDNA synthesis was then performed by adding DNA polymerase and incubating at 65 °C for 10 min. The reaction was terminated by heating at 80°C for 3 min. cRNA was then transcribed from cDNA by incubating the reaction with RNA polymerase at 42 °C for 4 hrs. Finally, RNase-free DNase was added to rid the reaction of any remaining DNA (37 °C for 15 min). The transcribed RNA was purified and desalted using the RNeasy MinElute Cleanup Kit (Qiagen, Inc., Valencia, CA, USA): RNA was washed and centrifuged in RNA spin columns with a succession of buffers and ethanol solutions (RLT, RPE, 100˚ and finally 80˚ ethanol). RNA was eluted in a final volume of 14 µL nuclease-free H2O and concentrated to 3 to 8 µL (the volume recommended by the cDNA synthesis kit) in an Eppendorf Vacufuge, and then used in the second round of 1st-strand cDNA synthesis: random primers, MMLV reverse-transcriptase and buffers were added and the mixture was incubated at 37 °C for 1 hr. Finally, RNase H was incubated with the reaction for 20 min at 37°C to eliminate any remaining RNA, and the reaction was terminated at 95°C for 2 min.

2.2.4 Whole taste papilla/single-cell polymerase chain reaction The iQ™ SYBR® Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA) was used for whole papilla RT-PCR and single-cell RT-PCR to amplify transcribed DNA. To examine the expression of specific markers in taste bud cells or in a single cell, custom-

designed forward and reverse primers were added. Primers sequences are described in Table 2. All primers sequences are written 5’–3′. The standard reaction mixture contained 10 μL iQ SYBR Green Supermix, 1 μL of template cDNA, 1 μL each of 100 μM forward and reverse primers, and 7 μL nuclease- free water. Template cDNA that was produced from total bud RNA was diluted at 1:5 for RT-PCR, but the cDNA produced from a single cell RNA was used undiluted as it was less concentrated than DNA from whole taste tissue. An independent RT-PCR experiment was run for every single cell and for every combination of 2 x CV and 1 x foliate papillae. The PCR profile was 95 °C at 3 min (initialization step, one cycle), 95 °C at 15 sec (denaturation step, 40 cycles) , 58 °C at 30 sec (annealing step, 40 cycles), 72 °C at 30 s (extension step, 40 cycles), and 72 °C at 30 sec (final elongation step, 1 cycle). PCR products were separated by gel electrophoresis on a 1.5% agarose gel containing 0.5 μg/ml Ethidium Bromide. Each lane was loaded with a mixture containing 5 µl of PCR product, 4 µl double-distilled H2O and 1 µl gel loading buffer (Invitrogen 10X Blue Juice, Life Technologies, NY, USA). Bands were measured relative to a 100bp DNA ladder (Lonza SimplyLoad™ DNA ladders, Rockland, ME, USA). The gel was observed under UV light (Benchtop UV transilluminator, UVP systems, Upland, CA, USA), photographed with a Kodak Digital Science DC120 camera (Eastman-Kodak, Rochester, NY, USA), and analyzed with a Kodak Digital Science Electrophoresis Documentation and Analysis System 120 (Eastman-Kodak).

2.2.5. Primers

Primer sets for 5-HT1A were investigated in parallel with the taste transduction receptor molecules T1R3, T2R9, or the G-protein α-gustducin, all markers for type II cells. Other primers included those for the type III cell marker PKD2L1 –the candidate sour taste receptor– and GAD, a type II/III cell marker. To ensure the specificity of the 5-

HT1A bands three different 5-HT1A primer sets were used in whole taste papilla and

single cell RT-PCR. The primers were designed to correspond to three different sequences within the 5-HT1A gene and they were termed: 5-HT1A(A), 5-HT1A(B), and 5-

HT1A(C). These primers were tested in a group of 10 single cells extracted from the CV papillae of five animals (group “A”), along with primers for Gα-gust, T1R3, and T2R9.

2.2.6. Experimental controls In addition to performing whole taste papilla RT-PCR experiments as positive control, primers for cytokeratin 8 (CK8) – a protein ubiquitously expressed across all mature taste receptor cells114 – were included in every single-cell RT-PCR run as a positive whole cell control. Negative control reactions were also included to ensure that the amplified PCR product resulted from cDNA produced from mRNA template and not from either genomic or extraneous sources of DNA: a PCR tube lacking the primers, and another lacking the DNA template, were included to control for carry-over contamination. The latter contained a small aliquot of ECF (in lieu of the DNA template) sampled adjacent to the collected cell, and was processed through RNA amplification and cDNA production to ensure that PCR products did not result from extracellular templates originating from the collected fluid. As another negative control to examine possible contamination in the nuclease-free water used in PCR reactions and gel electrophoresis, a lane in the gel was loaded with only nuclease-free water and loading buffer, omitting the PCR product. Single cells were only included in the data set if they satisfied criteria for both positive and negative controls.

2.3. Electrophysiological whole nerve chorda tympani (CT) recordings Electrophysiological recording from the chorda tympani (CT) or glossopharyngeal nerves has long been used in studying peripheral taste transduction and modulation by signaling agents at the level of the bud115,116,117. Whole nerve recordings offer real-time

visual representation of the collective afferent signals from innervated taste buds. Those signals represent the summation of the various transduction and modulatory mechanisms taking place at the level of the taste buds following lingual gustatory stimulation. Nerve responses to a taste quality can be monitored in wild-type animals and compared to responses from knock-out animals for a specific taste-related protein. Alternatively, responses can be compared before and after the application of a particular agent that has the ability to modulate a specific molecule in the taste transduction cascade. For the purpose of this study, a “pharmacological knock-out” approach was designed whereby CT nerve responses to various lingually applied tastants were compared prior to and following intravascular administrations of antagonists to specific taste signaling molecules. Responses were monitored up to one hour following the injection to measure the recovery from the drug.

2.3.1. Rationale for using chorda tympani nerve recording to study serotonergic function in the taste bud The rationale for using integrated whole nerve chorda tympani recording in combination with systemic drug injections to study the role of 5-HT in taste is multifactorial. First, it allows for the effect of administration of 5-HT receptor antagonists to be analyzed. Second, the CT response to a taste quality represents a real- time representation of integrated responses from a population of taste buds in the papillae being stimulated. Third, if 5-HT does indeed play a role in taste transduction at the level of the taste buds, then interfering with 5-HT’s action is expected to produce noticeable changes in the features of CT responses to taste stimuli. Integrated whole nerve CT recording therefore provides the opportunity to observe, in real time, the net effect of blocking 5-HT receptors on nerve response to lingual taste stimuli. Possible changes in response magnitude, shape, and adaptation to a specific taste quality are all factors that can shed light on 5-HT’s mode of action in the bud.

2.3.2. Animals and surgery Experiments were performed on adult male Sprague-Dawley weighing between 160 to 300 grams. Surgical procedures conformed to the Ohio State University Institutional Animal Care and Use Committee (IACUC) approved protocols. Animals were anesthetized with a 50 mg/Kg intraperitoneal injection of pentobarbital (Nembutal 50mg/mL – Lundbeck Inc., USA) and were maintained at a surgical level of anesthesia – marked by a total absence of pedal withdrawal upon pinching and lack of corneal blink reflex– by administering booster injections of 0.05 ml Nembutal every 30 minutes or as needed. The incision sites were shaved and the animals were laid supine on a heating pad. Body temperature was maintained at 36-37°C and constantly monitored by a rectal probe connected to a dedicated thermometer. The animal’s head was secured through the upper jaw using the clamp of a non-traumatic head holder to provide stability for the surgical procedure. The animal’s tongue was slightly extended out of the mouth to expose the anterior part, and thus facilitate stimulus delivery. Surgical work was performed with the help of a bifocal dissecting microscope (Carl Zeiss 48871, Germany) and the site of surgery was illuminated by an external light source (Leica EL6000, Germany). A tracheotomy was performed to ensure unobstructed breathing: a 1 cm longitudinal midline incision was made in the animal’s neck just above the sternum. The sternohyoid and longus colli muscles were bluntly dissected along the midline and the trachea was exposed. A small transverse incision was made in the cartilage tissue of the trachea for cannulation. The cannula, consisting of 4 cm of polyethylene tubing (1.57 mm internal diameter; Intramedic- NJ, USA) was inserted 1 cm into the trachea and normal respiration was established118. Following the tracheotomy, the left external jugular vein was catheterized for the purpose of drug delivery. The cannula consisted of a 15 cm polyethylene tubing (2 mm internal diameter) attached to a 21-gauge needle (PrecisionGlide 305165, NJ, USA)

and connected to an interchangeable 1 ml syringe (Tuberculin Slip Tip, BD syringe, NJ, USA). Initially, the syringe was filled with a heparinized saline solution (Heparin Lock Flush solution, 10 USP Heparin units/ml, BD PosiFlush, NJ, USA) to prevent clogging of the cannula. The solution was pumped to fill the dead space in the tubing and eliminate air bubbles prior to insertion into the jugular vein. A 5 mm longitudinal incision was made just above the left front leg, lateral to the tracheotomy incision. The underlying adipose tissue and connective tissue were carefully cleared, exposing the jugular vein. Two ligatures (4/0 non-absorbable surgical silk suture, Deknatel, MA, USA) were placed at the rostral and caudal sides of the exposed vein. The rostral ligature was firmly tied to prevent bleeding. A small incision was made in the vein between the two ligatures and the tip of the cannula was inserted in the caudal direction into the vein just before it reached the left atrium (approximately 5 mm). The caudal ligature was then snugly tied, securing the cannula tube to the vein119,120. The chorda tympani nerve was exposed using the lateral approach: a transverse incision was made in the skin overlying the anterior belly of the right digastric muscle, between the rostral lobes of the parotid gland and the mandibular bone. Fibers of the digastric and masseter muscles were bluntly dissected, and the mandible was exposed and freed of connective tissue. The zygomatic arch and ramus of the mandible were cut, and the underlying medial pterygoid muscle was retracted using blunt hooks. The chorda tympani nerve was then visible from the point where it leaves the tympanic bulla caudally, and the point where it separates from the lingual nerve rostrally. The lingual branch of the trigeminal nerve was severed at the point where it exits the foramen ovale, and the nerve was retracted. The CT was carefully freed of the surrounding fascia and connective tissue, and the nerve was severed at the caudal end using fine forceps (Dumont no. 5; Switzerland). The peripheral end of the nerve was then carefully desheathed of its epineurium to facilitate electrical conductivity.

2.3.3. Nerve recording The CT nerve was laid vertically on a 32-gauge platinum/iridium wire electrode, mounted on a micromanipulator apparatus, and connected to an AC pre-amplifier (A-M systems Inc. Model 3000). A second indifferent electrode was inserted in the nearby remnants of the resected digastric muscle. The heating pad, head holder, and stereotactic apparatus were all grounded through a single ground input fed into the amplifier. Neural signals were amplified at a gain of 500X, and filtered between a low and high pass frequency of 3 kHz and 100 Hz respectively. Amplified activity was fed into an oscilloscope (Tektronix, model DPO 2014), and to a dedicated audio monitor (GRASS systems; model AM3). Signals were full-wave rectified and integrated with a time constant of 0.2 sec. on a wide-band A.C to D.C converter (GRASS systems; model 7P3B). Integrated neural activity was displayed by a rectilinear pen on the recording chart of a polygraph D.C driver amplifier (GRASS systems; model 7DAB). Simultaneously, neural activity was monitored on the oscilloscope and through the audio monitor. In the absence of lingual stimulation the integrated spontaneous nerve activity was evident on the recording chart as a stable baseline. Upon taste stimulation, the integrated nerve response appeared as an initial transient upward spike above baseline, representing the summation of action potentials from a group of nerve fibers (phasic response). This portion of the response conveys information on rapid changes in stimulus rate and intensity, and it is quickly followed by a diminished activity that plateaus (tonic response) until the stimulus is terminated 10 seconds following its onset. The tonic portion of nerve response represents nerve adaptation to the stimulus as action potentials continue to be generated in the presence of the stimulus, therefore conveying information on the duration of the stimulus.

2.3.4. Gustatory stimulation of the tongue Gustatory stimulation was achieved by bathing the anterior tongue with 3 mL of

taste solution slowly delivered to the anterior tongue via a sterile Pasteur pipette fitted with a rubber pump. Taste stimuli were selected as prototypical representations of basic taste qualities: Salty, sweet, bitter, and sour, and they consisted of 0.5 M NaCl, 1.0 M sucrose, 0.03 M quinine, or 0.01 N HCl. All solutions were prepared by dissolving reagent grade chemicals in room temperature distilled water. The concentrations of taste solutions produced near maximal CT responses and were chosen based on results from previous CT recording studies121,122,123,124. In preliminary experiments, a different stimulus delivery technique using a gravity feed flow chamber was tested. The flow chamber consisted of a 1 cm diameter glass tube that allowed the anterior tongue to be inserted in the lower opening through a fitted rubber dam. The stimulus was delivered in a 5 ml volume through a funnel attached to the upper opening. The stimulus was terminated by flushing the chamber with 25 ml of distilled water. Drainage occurred through a small opening in the lower half of the tube. The only advantage when compared to the pipette delivery technique was that it allowed for a more spatially inclusive, even, and synchronized coverage of the tongue’s surface by the taste stimulus solution. This produced higher fidelity phasic responses relative to those obtained with pipette delivery where there was higher chance of squirting the bulk of the solution on a smaller area of the tongue at a much higher rate, therefore causing inconsistent magnitudes in the phasic portions of the response. However, this issue was avoided in the pipette method by pumping out the stimulus solution closer to the tongue’s surface at a consistent rate in one slow and controlled release. Repeated receptor stimulation with the same taste quality may cause accumulation of ligand, receptor adaptation and depletion of components of taste transduction cascades (including ATP) downstream of receptor stimulation. Subsequently, integrated nerve responses could decrease in magnitude over time with

high frequency stimulations. Therefore, as a preventative measure, stimuli were delivered in the following sequence: NaCl, sucrose, quinine, and then HCl, with a two- minute interval between consecutive stimuli. Consequently, total time separating two consecutive stimuli of the same taste quality was 8 minutes, allowing ample time for cell recovery. Neural responses had to demonstrate stability prior to drug delivery. Criteria for a stable recording were a steady baseline between stimuli, and consistent magnitude and shape of the first three responses to the same taste stimulus. Once these criteria were satisfied, experimental drug or control vehicle was administered into the jugular vein in the form of a single 300 µL bolus injection delivered over a period of 5 seconds. Nerve responses to taste stimuli were monitored and recorded in the periods before, and at least up to one hour following the injection to examine the immediate, as well as the long term effects of the injected solution.

2.3.5. Drugs

Experimental drugs consisted of two different 5-HT1A receptor antagonists, WAY 100635 (10, 25, 100 or 200 µg/kg BW, Sigma-Aldrich) and NAD-299 (175, 350 or 750

µg/kg BW, Tocris Bioscience, Bristol, UK), and one 5-HT3 receptor antagonist, ondansetron (1000 µg/kg BW – LKT labs. USA). These drugs were chosen due to their selectivity and high potency in blocking their respective 5-HT receptors125,126,127,128,129,130. All drugs were diluted to their appropriate doses in heparinized saline.

2.3.6. Experimental controls To test whether jugular injection of the vehicle solution by itself could influence CT responses to lingual taste stimuli, 300 µl single bolus control jugular injections consisting of heparinized saline were administered in eleven animals while recording CT responses to taste stimuli before and after each injection. As a way of positive control to

examine whether jugular injections were able to quickly reach taste tissue, CT activity was monitored in response to 300 µl single bolus jugular injections of ATP solutions at three different concentrations (0.75, 1.5 or 3 mg ATP/Kg BW) in three different nerve recordings. To examine the effect of blocking ATP receptors on CT responses to jugular ATP injections and to lingual taste stimulation, 300 µl single bolus jugular injections of a non-selective P2-purinoceptor antagonist, PPADS (Pyridoxal-phosphate-6-azophenyl- 2′,4′-disulfonate), were administered at two different concentrations (10 and 20 mg/kg BW)131. Neural activity, response to lingual taste stimuli, as well as nerve response to jugular ATP injections were recorded before and after PPADS injection. Both ATP and PPADS were purchased from the Sigma-Aldrich company. Dosage selection was based on previous studies that used ATP132,133 and PPADS134,135 systemic injections.

2.3.7. Data analysis Integrated CT responses from every experiment were analyzed off-line. Since the magnitude of the phasic portion of nerve response was sensitive to stimulus delivery factors such as the stimulus solution flow rate, and because its duration was very brief, only the tonic portion of the response was considered for data analysis. The amplitude of each response on the recording chart was measured in millimeters as the distance between the baseline and the midpoint on a line drawn through the response plateau (tonic portion of response); (Figure 1). To examine the immediate effect of an injected pharmacological agent on the neural response to a particular taste stimulus, the amplitude of the CT response for that stimulus before the injection was compared with the amplitude of the response taken immediately after the injection. Further, depending on the pharmacokinetics of the injected drug, the CT response- amplitude was predicted to recover from any possible change after a certain period of time following the injection as the drug was metabolized. Therefore, long-term drug effect was factored in the comparison by including responses taken between 30 and 65 minutes after

injection, the presumed recovery phase. To achieve a statistically relevant comparison, for every taste quality stimulus, the amplitudes of the last three responses taken just before drug injection (referred to as “pre-injection“), were averaged together and their mean was compared with that of the amplitude of the first three responses for that stimulus taken immediately after the drug injection (“post-injection”). The two response-amplitude means (pre- and post- injection) were also compared with the amplitude-mean of the last three responses to that same stimulus taken between 30 and 65 minutes after injection (“recovery”). Before corresponding means from separate nerve recording experiments testing a specific injected agent could be averaged together and analyzed, data had to be standardized. In individual experiments, and for every taste quality, the “post-injection” and “recovery” means were normalized to the value of the “pre-injection” mean by dividing each mean by the “pre-injection” mean. Thus, the normalized value of the “pre- injection” mean was always equal to 1.0. Consequently, a normalized “post-injection” or “recovery” mean-value that had a value of 1.0 indicated no change in mean-response amplitude from “pre-injection”. A mean-value above 1.0 signified an increase in mean- response amplitude relative to “pre-injection”, and a mean below 1.0 signified a decrease. At that stage, statistical analysis was carried out. To test the data’s pattern of distribution, the Shapiro-Wilk’s normality test was carried out on each data set from each experiment, with the following null hypothesis

(H0): the population has a normal distribution. The test resulted in p values that were always greater than 0.05, which implied that the alternative hypothesis should be rejected, and that data can be considered normally distributed. Since the goal was to compare the means of two variables and test if the average difference is significantly different from zero, the paired T-test was chosen for statistical analysis with the following null hypothesis (H0): mean response-amplitudes corresponding to two

different stages of recording were not significantly different. A two-tailed significance (2-tailed p value) that was ≥ 0.05 signified that the null hypothesis is valid and that the difference between the two means being compared was not statistically significant. Statistical analysis was performed on the SPSS statistical package (IBM). For every taste quality stimulus corresponding mean-values from all experiments testing the same injected agent were averaged and plotted on an Excel bar chart. Relative data were presented as response means ± standard errors. The following variables were plotted: “mean pre-injection amplitudes”, “mean post-injection amplitudes” and “mean recovery amplitudes”.

CHAPTER 3

RESULTS

The aim of this study was to determine the identity of taste receptor cells targeted by serotonin release, and the functional effects of serotonergic activity in the rat taste bud. Since preliminary data suggest that only the 5-HT1A receptor is specifically expressed in TRCs, phenotyping experiments were performed to localize the expression of this receptor and identify cell types, and hence taste qualities influenced by serotonin release. Two different techniques, double label immunocytochemistry and single cell reverse transcriptase PCR experiments were used to examine the co-expression of 5- HT1A with distinct TRC-specific phenotypic markers. The physiological role of serotonin in the transduction of different taste qualities was studied using electrophysiological whole chorda tympani nerve recordings to examine the effect of blocking serotonergic transmission, by injecting 5-HT1A or 5-HT3 antagonists, on individual gustatory responses. The following sections illustrate combined data from morphological (ICC) and functional (CT recording) experiments that were performed to phenotype the 5-HT1A- expressing TRCs, and determine the role of serotonin in the peripheral gustatory system.

3.1. Phenotyping the 5-HT1A-expressing TRCs Phenotypic markers used in ICC and RT-PCR experiments included TRC type- specific proteins, taste receptor molecules, and elements of the transduction cascades

of specific taste qualities. Their expression in TRCs has been previously characterized in numerous studies, and they can therefore be used as identifiers for specific TRC types and the transduction of specific taste qualities. Consequently, a TRC that co-expresses a certain marker along with 5-HT1A implies the latter is associated with the same taste quality and/or cell type that expresses that marker.

3.1.1. Distribution on 5-HT and 5-HT1A in taste buds of the anterior tongue

All ICC experiments using antibodies directed against either 5-HT or 5-HT1A showed labeled TRCs. All immunopositive cells for either 5-HT or 5-HT1A typically appeared spindle-shaped with evenly dispersed labeling across the cytoplasm, and a dark unlabeled nucleus (Figures 2 – 9). 5-HT-immunopositive TRCs featured tall and slender cell bodies and relatively small nuclei, both characteristics of type III cells. 5-HT1A-immunopositive TRCs on the other hand were relatively shorter and more roundly-shaped, both features of type II cells (Figures 2 – 9).

Previous work has shown that 5-HT and 5-HT1A are expressed in non-overlapping sets of cells in taste buds of the posterior tongue in rats, but no data exists on their expression in taste buds of the anterior tongue or the NID (anterior gustatory field). In the present study, single-label ICC experiments were performed in sections of the anterior tongue to examine the expression of 5-HT and 5-HT1A in taste buds of fungiform papillae. Results showed clearly labeled 5-HT- and 5-HT1A-TRCs in those taste buds (Figure 2). In three separate experiments using three animals, averages of 1.83 ± 0.03

(mean ± standard error/SE) 5-HT-immunopositive TRCs, and 1.97 ± 0.05 5-HT1A- immunopositive TRCs were detected per cross-sectioned taste bud. The expression of 5-

HT1A was also verified in taste buds of the NID where it was studied in double-label ICC experiments. In these taste buds, an average of 1.1 (± 0.1) 5-HT1A-immunopositive cells was detected per cross-sectioned taste bud (Figures 3, 6, 7, and 8).

3.1.2. Co-expression of 5-HT1A and TRC-specific markers

It has been suggested that 5-HT1A receptors are located in type II cells in taste buds4,98, but no previous work has been dedicated to confirm this expression pattern. Moreover, considering the heterogeneity of type II cells, it is not clear whether all or only a subset of those cells express 5-HT1A. The following sections describe results from double label ICC and single cell RT-PCR experiments that were conducted to answer this question by phenotyping the 5-HT1A-expressing TRCs. To obtain a more accurate and comprehensive view of the identity of cells that express 5-HT1A, several protein markers were included in phenotyping experiments. Markers were categorized according to the types of TRCs (type II or III) that express them, and/or to their known association with the transduction of a specific taste quality. Type II cell-specific markers that were used in immunolabeling and molecular phenotyping experiments included the taste-specific G-protein α-gustducin (Gα-gust or GUST), neuropeptide Y (NPY), the neuropeptide Y receptor NPY1R, the neuropeptide cholecystokinin 8 (CCK-8), the enzyme glutamic acid decarboxylase (GAD), the sweet and umami receptor molecule T1R3, and the bitter receptor molecule T2R9. Although expressed predominantly in type II cells, with a minor degree of co-expression in type III cells, the synaptosomal-associated protein 25 (SNAP-25) was also included in ICC double label experiments as a marker mainly associated with type II cells. T1R3 and T2R9 were used only in single cell RT-PCR experiments. Gustducin and GAD were used in both double-label ICC experiments and single cell RT-PCR experiments. Type III cell-specific markers included the neural cell adhesion molecule (NCAM) -used only in ICC experiments- and the candidate sour receptor polycystic kidney disease 2-like 1 protein (PKD2L1), used only in single cell RT-PCR experiments. Table 1 describes all the antibodies used in ICC experiments. To test the specificity of the antibodies selected for ICC experiments, tissue

sections were included in every experiment where the primary or secondary antibody was omitted. As expected, data from this group showed no labeled cells. In the experimental group, results revealed a varied degree of co-expression between 5-HT1A and the different type II cell-specific markers in all the gustatory epithelia that were examined (Tables 3, 4, and 5). Among type II cell-specific markers were the G-protein gustducin, the taste-specific G-protein linked to the T2R family of receptors for bitter taste, and GAD, the rate-limiting enzyme in the decarboxylation of glutamate to GABA. Gustducin has been used in many studies as a specific marker for bitter-processing type II TRCs32,136. GAD has also been demonstrated in previous work to co-localize with type II TRC-specific markers in rat taste buds137. Results showed varying degree of overlap in the expression of 5-HT1A with gustducin (Figure 3) and GAD (Figure 4). On average, in all taste buds examined (posterior and anterior tongue, and NID), out of 135 x 5-HT1A- and 180 x gustducin-expressing cells, about three-quarters (73% ± 2.96; n = 99) of 5-HT1A cells co-expressed gustducin. Conversely, a relatively small percentage of the 34 x 5-HT1A-expressing cells (18.18% ± 2.37; n = 34) co-expressed GAD, which was expressed in 122 cells.

The extent of co-expression between 5-HT1A and SNAP-25 resembled that of 5-

HT1A and gustducin. SNAP-25 is a protein involved in the exocytotic fusion complex of synaptic vesicles taking place during neurotransmitter release. Studies in mice have reported that SNAP-25 is predominantly expressed in type III cells138, whereas in rats, previous work has shown its predominant co-localization with markers associated with type II cells such as gustducin, CCK-8, VIP and GAD137. In posterior tongue taste buds, current double label ICC data showed that most of the 150 x 5-HT1A-immunopositive TRCs (89.33% ± 1.20; n = 134) co-expressed SNAP-25, which was expressed in 162 TRCs (Figure 5). The expression of many neuropeptides including NPY, NPY1R and CCK-8 has

been previously characterized in type II TRCs of taste buds of the posterior tongue in rats. Whereas the expression of NPY and CCK-8 among those cells overlaps to a great degree, NPY and its receptor have been shown to exist in non-overlapping populations of type II TRCs73. Current immunolabeling data showed extensive, but not complete overlap between TRCs that expressed 5-HT1A and those that expressed NPY, NPY1R, and CCK-8 in the gustatory epithelia of the posterior and anterior tongue and the NID (Figures 6, 7, and 8). On average, about three-quarters (73% ± 2.9; n = 95) of the 130 x

5-HT1A-expressing TRCs co-expressed NPY, which was expressed in 190 TRCs. A comparable percentage of 5-HT1A-immunopositive TRCs co-expressed NPY1R or CCK-8.

65.81% (± 4.62; n = 102) of 155 x 5-HT1A- and 160 x NPY1R-immunopositive TRCs co- expressed both markers. Similarly, 72.09% (± 3.00; n = 93) out of 129 x 5-HT1A- and 180 x CCK-8-immunopositive cells expressed both markers. NCAM was the only type III cell-specific marker included in the current double label ICC experiments. NCAM is a glycoprotein involved in cell adhesion and is expressed on the surface of neurons, glia, skeletal muscle and natural killer cells. In the taste bud it is known to be expressed in type III TRCs137. The present data showed no co-expression between 5-HT1A and NCAM (Figure 9). In 36 x taste buds from circumvallate and foliate papillae, 135 cells were NCAM-immunopositive; none overlapped with any of the 180 5-

HT1A-immunopositive cells. This result serves as further confirmation that the expression of 5-HT1A is exclusionary of type III cells of the taste bud. For SNAP25 and NCAM, double-label ICC experiments were performed in anterior and posterior tongue and NID tissue similar to the other markers that were examined. However, it was not possible to achieve reliable SNAP25 and NCAM signals within anterior tongue and NID TRCs. Therefore, for those two markers, data from only the posterior tongue (CV and foliate papillae taste buds) were included in the cell count.

The degree of co-expression of 5-HT1A with all other phenotypic markers,

measured through cell counting, was compared between posterior and anterior gustatory fields (posterior tongue vs. anterior tongue + NID respectively) using the paired samples T-test. Data revealed no significant differences between posterior and anterior fields for the co-expression of 5-HT1A with gustducin (p = 0.062), NPY (p = 0.058), NPY1R (p = 0.053), CCK-8 (p = 0.119), or GAD (p = 0.151). This suggests a similar pattern of co-expression for 5-HT1A in posterior tongue on one hand, and anterior tongue and NID on the other (Tables 3, 4, and 5). Up to this point, current double label ICC data from phenotyping experiments demonstrate the localization of 5-HT1A receptors in a subset of cells associated with the transduction of sweet, umami, and bitter taste stimuli. The fact that the degree of overlap between 5-HT1A and each of the type II-TRC-specific markers was not complete

(less than 100%) suggests that not all type II cells express 5-HT1A.

3.1.3. Co-expression of 5-HT1A and taste receptor molecules Single cell RT-PCR experiments were conducted for two purposes. First, the expression of 5-HT1A relative to taste receptor molecules, which was not possible to investigate in immunolabeling experiments, would be studied through RT-PCR. Secondly, it allowed a limited confirmation of co-expression patterns observed with ICC experiments. Single cell RT-PCR experiments provide a way to survey individually isolated TRCs for the possible DNA expression of any selected molecule(s). Whereas immunolabeling techniques allow the study of the co-expression of only two molecules at a time, the expression of seven different markers can be investigated using the current single cell RT-PCR setting. By extracting total RNA from a single TRC and reverse transcribing the cell’s cDNA, an extensive combination of primers for any proteins of interest can be included in PCR reactions. Therefore, only cDNA sequences that were transcribed from the expressed RNA are amplified when their appropriate forward and reverse primers

were included in the PCR reactions.

In the present study, the expression pattern of 5-HT1A was examined in a total of 51 x single TRCs cells isolated from the CV papillae of ten animals, relative to the expression of a group of phenotypic markers. In all, these markers included gustducin, GAD, the bitter (T2R9), sweet and umami (T1R3), and sour (PKD2L1) receptor molecules. Gustducin and GAD have been included in ICC double label experiments, and their expression relative to 5-HT1A was subsequently confirmed through molecular phenotyping using RT-PCR. Positive control experiments were first conducted by testing all primers, including several primer sets for 5-HT1A, in separate whole taste papilla RT-PCR experiments. In these experiments, CV and foliate papillae were excised from the tongue, homogenized, and total RNA was extracted. Primers were subsequently tested on cDNA reverse transcribed from the extracted RNA. All primers were first tested in whole papillae RT-PCR (Table 2 describes all primers used in the RT-PCR experiments).

Three different primers corresponding to three different regions of the 5-HT1A gene were included to ensure that the RT-PCR product was indeed specific to 5-HT1A. Also included in each single TRC's PCR reactions were primers for cytokeratin 8 (CK8), a protein marker expressed in all mature taste buds. Hence, expression of CK8 in a tested individual cell serves as a positive control and confirms that the cell is a TRC as opposed to a non-gustatory cell. Gel electrophoresis analysis of PCR products showed positive bands for all markers tested, which included CK8, Gus, T1R3, T2R9, PKD2L1, GAD, 5-

HT1A(A), 5-HT1A(B), and 5-HT1A(C), indicating the presence of their DNA, and thus their expression in CV and foliate tissue. Other positive control experiments were performed to verify the specificity of the 5-HT1A primers in individual TRCs. In the first group of experiments (Group “A”), all three 5-HT1A primers (5-HT1A (A), (B), and (C)) were included in PCR reactions performed

on 10 x TRCs, along with primers for Gα-gust, T1R3, and T2R9. Examination of group “A” single cell RT-PCR products revealed that six out of the ten cells were whole mature TRCs marked by the presence of positive bands for the CK8 protein marker in their PCR products (Table 5). Of those six cells, four simultaneously showed specific bands for all three 5-HT1A primers equally, thereby confirming the specificity of the 5-HT1A primers. In subsequent experiments, one of the 5-HT1A primers, 5-HT1A(C), was chosen to be the representative for 5-HT1A, and was included in every PCR run along with the primers for CK8, Gus, T1R3, T2R9, PKD2L1 and GAD. Negative control experiments were performed to ensure that gel bands reflected the presence of DNA from the selected markers, and did not arise from extraneous DNA. In one control, nuclease-free water was substituted for template in the RT-PCR experiment (water control). In another negative control, ECF collected from a region adjacent to a dissociated cell to test for extraneous DNA which might be present in the bathing medium (cell-free control). In other negative controls, primers were omitted. Gel analysis of PCR products from all negative control reactions showed no visible bands. These results validate the specificity of the bands in the experimental group.

In the following set of experiments, the 5-HT1A(C) primer was used in single cell RT-PCR experiments performed in group “B”, which consisted of 35 individually isolated

TRCs. In this group, primers for CK8, Gus, T1R3, T2R9, PKD2L1, GAD, and 5-HT1A were included in the PCR reactions for each tested cell. All those primers represented markers for type II cells, except for the type III-TRC-specific marker, PKD2L1. Gel analysis of group “B” RT-PCR products revealed that 25 out of 35 extracted cells were whole mature TRCs, indicated by the presence of specific bands for CK8 in their PCR products. Consequently, only those 25 cells were included in the data analysis. When data from group “A” and “B” were combined (Table 5), results indicated that among all 51 x single TRCs tested, 31 were CK8-positive, and thus were included in

the data analysis. Of the CK8-positive TRCs, twelve showed specific bands for 5-HT1A.

Interestingly, every 5-HT1A-expressing TRC at least expressed either T1R3 or T2R9. In fact, half of the 12 x 5-HT1A-expressing cells co-expressed T2R9 and the other half co- expressed T1R3, indicating that type II TRCs that process bitter, sweet and umami tastes express 5-HT1A. Nevertheless, out of a total of 12 x T2R9-, and 10 x T1R3-expressing

TRCs, 6 x and 4 x TRCs did not co-express 5-HT1A respectively. This suggests that 5-HT1A is not expressed in all TRCs that process bitter, sweet, and umami.

Further, data also showed that out of 12 x 5-HT1A- and 14-gustducin-expressing TRCS, 7 x TRCs co-expressed both markers. The expression of GAD was examined in 25 x single cells (group “B”) and was found in 6 TRCs. Only two TRCS co-expressed 5-HT1A and GAD. These data were analogous to the current findings from double label ICC experiments, which also revealed a comparable degree of co-expression of 5-HT1A with gustducin or GAD. The data also further suggest the expression of 5-HT1A in type II TRCs. The candidate sour receptor PKD2L1, known to be expressed in most type III cells (which are mostly serotonergic) was found to be expressed in only 2 out of 25 x TRCs. This low yield is consistent with the known anatomy of the taste bud, where type III cells are known to constitute only a minority of cells within the bud39. As predicted, analysis of RT-PCR products revealed that those 2 x TRCs did not co-express any of the type II

TRC-specific markers. Further, they did not express 5-HT1A. As well, the observed expression pattern of multiple type II cell makers within a single TRC was anticipated and served as internal control for the specificity of the technique. Out of the 12 x TRCs that expressed the bitter receptor molecule T2R9, 10 x TRCs co-expressed its linked G-protein α-gustducin. Further, co-expression was observed between gustducin and T1R3 in 5 out of 31 x TRCs, and between gustducin and GAD in 2 of the 25 x TRCs in group “B”. Three of the 6 x GAD-expressing cells also co-expressed T1R3, and only one GAD-positive cell co-expressed T2R9. These patterns of

co-expressions are all consistent with the literature and with the previous immunocytochemical data. Also, as predicted, data did not show any overlap between the type III TRC marker PKD2L1 and any of the markers associated with type II cells, including 5-HT1A. (Table 6 illustrates data from single cell RT-PCR experiments, and Figure 10 illustrates sample gel electrophoresis data from the RT-PCR products of seven different single TRCs). Overall, data from immunolabeling and molecular phenotyping experiments suggest that 5-HT1A is not expressed in type III cells. More crucially, data also revealed that 5-HT1A is expressed to varying degrees with protein markers associated with type II TRCs, including gustducin, NPY, NPY1R, CCK-8, and GAD, as well as molecules for receptors of bitter (T2R9), and sweet and umami tastes (T1R3). These observations are consistent with the notion that serotonin, arising from type III cells, acts in a paracrine manner on type II cells which express the 5-HT1A receptor.

3.2. Physiological properties of serotonin as measured at the level of the chorda tympani nerve

Data from phenotyping experiments are consistent with the notion that 5-HT1A, the only serotonin receptor specific to TRCs, is localized in a subset of type II cells that are involved in processing sweet, bitter and umami tastes. Those TRCs therefore constitute possible targets for serotonergic activity originating from type III cells.

Nevertheless, the outcome of the activation of 5-HT1A receptors, and the net effect serotonin exerts on taste signals was still undetermined. To explore the physiological role of serotonin in the taste bud, chorda tympani responses to various lingual taste stimuli were measured and compared both before and after pharmacologically blocking serotonergic activity (sample integrated CT responses to the four taste stimuli is illustrated in Figure 11). Jugular injections of either

a 5-HT1A or a 5-HT3 antagonist were administered and their effects on sweet, salty, bitter, and sour taste signals were recorded.

3.2.1. Validation of experimental preparation Since the effect of vascular injection of pharmacological agents and chorda tympani neural recordings hadn’t been previously studied, some initial experiments were designed to validate the experimental preparation. Experiments were designed to ensure that agents could reach the taste bud and that intravascular injection itself did not produce changes in CT response to lingual taste stimuli. Determining the time it took for the drug to produce noticeable effects at the level of CT was also necessary. Lastly, since serotonin’s effects are not merely limited to the taste bud, it could be argued that intravascular administration of a 5-HT receptor antagonist might cause non-specific metabolic side effects such as cardiovascular changes that could transiently alter normal physiological blood pressure and tongue perfusion, thereby affecting normal CT responses to lingual taste stimuli. All those factors were addressed in separate recording experiments.

3.2.2. Chorda Tympani response to ATP jugular injection There is now general agreement that ATP is the main gustatory neurotransmitter from TRCs to the afferent nerves. ATP is believed to be released from type II cells as a universal response to their direct stimulation by tastants, or indirectly following the stimulation of type III cells. The released ATP is thought to target ligand-gated P2 purinoceptors on other TRCs and on gustatory afferents. It is predicted therefore that the intravascular injection of ATP would stimulate purinergic receptors, including chorda tympani P2 purinoceptors and produce neural excitations. To test this hypothesis, control experiments were performed in three separate nerve recordings where 300 µl jugular injections of ATP solutions were administered at three different concentrations

(0.75, 1.5 or 3 mg ATP/Kg BW), at 6 minute intervals with the hypothesis that these injections should produce neural excitation. Immediately after ATP injection, there was a transient dose-dependent increase in respiratory rate that lasted between 10 and 20 seconds. This was an expected occurrence resulting from the transient increase in pulmonary artery pressure caused by ATP139,140. In all three recording experiments, ATP injections uniformly caused transient neural excitations integrated as bell-shaped curves of dose-dependent magnitudes (Figure 12). These results confirm that a systemically injected drug can quickly reach taste buds and produce the desired effects as shown by the CT response to the injection. They also confirm that CT response to ATP is excitatory. The effect of pre-administration of a purinergic antagonist was tested on CT responses produced by either intravascular injections of ATP or lingually applied tastants. Separate control nerve recordings where performed where a non-selective P2 purinergic receptor antagonist (PPADS) was systemically injected at two different concentrations (10 and 20 mg/kg BW) while monitoring CT neural activity and nerve response to lingual taste stimulation. The assumption was that blocking P2 receptors should result in neural inhibitions, and consequently smaller CT responses to all four taste stimuli. Furthermore, depending on the pharmacodynamics of the utilized drug, nerve responses were expected to show signs of recovery as the drug was metabolized. Pre-injection and post-injection mean response amplitudes for each taste quality were averaged from three separate experiments and compared. PPADS at 10 mg/kg resulted in significant, reversible decreases in CT responses to all taste qualities (Figures 13 and 14). Data averaged from three experiments showed that, immediately after the injection, there was on average a 53% decrease in mean amplitudes of CT responses to NaCl (p = 0.003), 82% decrease in sucrose responses (p = 0.001), 79% decrease in quinine responses (p = 0.001), and 75% decrease in HCl responses (p = 0.001). PPADS at

20 mg/kg (n = 3) caused an even greater decrease in CT responses: (87% decrease for NaCl response (p = 0.001), 85% for sucrose (p = 0.002), 97% for quinine (p = 0.001), and 97% for HCl (p = 0.001); (figures 13 and 15). In both cases, responses to all taste qualities significantly reversed their decline, and recoveries were observed within 45 minutes after the 10 mg/kg PPADS injection and 60 minutes after the 20 mg/kg injection. The effect of PPADS on CT response to intravenous ATP injection was also significant (Figures 15 and 16). Data averaged from three experiments showed that, immediately following the injection of PPADS at 10 mg/Kg (n = 3), mean CT response amplitude to a 1.5 mg/Kg ATP injection declined by 97% (p < 0.001). On average, responses recovered to half of their pre-injection amplitudes 50 minutes after the injection. Injecting PPADS at 20 mg/Kg (n = 3) virtually abolished CT responses to injected ATP, reducing them by 98.8% (p = 0.001). Responses recovered to almost half (42%) of their pre-injection magnitudes 62 minutes after PPADS injection. Figure 16 illustrates sample CT responses to a 3 mg/kg ATP injection before and after PPADS injection at 10 mg/kg.

3.2.3. Chorda Tympani response to saline jugular injection It was necessary to ensure that the vehicle for the drug solutions used in electrophysiology experiments has no significant effects on CT responses. To examine the net effects of saline jugular injection on CT responses to lingual taste stimuli control recording experiments were performed in ten rats where nerve responses to all four taste qualities were measured in the periods before and after a 300 µL single bolus saline jugular injection was administered. Data averaged from all ten experiments showed that saline injection caused no significant change in nerve response to all taste qualities when post-injection response amplitudes and shapes where compared with pre-injection values (Figure 17). For each taste quality, there was no significant differences between mean CT response amplitudes measured before injection (last 3

responses before injection), immediately after (first three responses after injection), and up to 65 minutes after saline injection (last 3 responses). When pre-saline and post- saline mean response amplitudes were compared for each taste quality, the paired- samples T-test resulted in p-values greater than 0.05 (p = 0.085, 0.063, 0.191And 0.193 for NaCl, sucrose, quinine and HCl respectively). These observations suggest no significant effect for saline injections on CT nerve responses to lingual stimulation by the current concentrations of salt, sucrose, quinine, and HCl solutions.

3.2.4. Effect of blocking 5-HT1A receptors on CT responses to gustatory stimulation

As previously mentioned, both the 5-HT1A and 5-HT3 receptor subtypes are expressed in the taste bud. If the 5-HT1A receptor plays a significant role in mediating the modulatory activity of 5-HT on taste processing, then injecting a 5-HT1A antagonist would be predicted to alter CT responses to the taste modality(s) affected by serotonin.

The same logic applies to the 5-HT3 receptor within the taste bud. To examine the effect of modulating the activity of serotonin receptors on neural gustatory responses, CT responses to lingual taste stimuli were recorded before and after single bolus injections of a 5-HT1A receptor antagonist or a 5-HT3 receptor antagonist.

Two different specific 5-HT1A receptor antagonists were tested, WAY 100635 (10,

25, 100 or 200 µg/kg BW) and NAD-299 (175, 350 or 750 µg/kg BW). The 5-HT3 receptor antagonist was ondansetron (1000 µk/kg BW). Intravascular injections of WAY-100635 significantly reduced CT responses to all four tested taste stimuli. Data averaged from six experiments showed that immediately post-injection WAY-100635 at 10 µg/kg significantly reduced CT response to NaCl by 16% (; p = 0.003), to sucrose by 40% (p = 0.021), to quinine by 21% (p = 0.032), and to HCl by 16% (p = 0.023) (Figure 18). Similar results were observed after injecting WAY- 100635 at the other drug concentrations tested. At 25 µg/kg (n = 6), WAY-100635

significantly reduced the CT response to NaCl by 32 % (p < 0.001), to sucrose by 47 % (p = 0.001), to quinine by 68 % (p < 0.001), and to HCl by 49 % (p < 0.001) (Figure 19). At 100 µg/kg WAY-100635 (n = 8), the inhibition in CT responses was at 31 % for NaCl (p = 0.009), 50 % for sucrose (p = 0.025), 46 % for quinine (p = 0.002), and 36 % for HCl (p = 0.005) (Figure 20). At 200 µg/kg WAY-100635 (n = 6), NaCl responses were reduced by 33 % (p < 0.001), sucrose by 58 % (p < 0.001), quinine by 56 % (p < 0.001), and HCl by 51 % (p = 0.001) (Figure 21). Peak effects of the drug, marked by the degree of response inhibition, were observed in the first ten minutes after each injection, and were followed by a gradual reversal in response decline. On average, recovery phase peaked between 45 and 65 minutes post-injection.

Since the result of blocking 5-HT1A was a consistent inhibition in CT response to all taste qualities, it was deduced that 5-HT1A-mediated serotonergic activity is involved in the processing of sweet, sour, bitter and salt tastes. Furthermore, since blocking serotonergic activity with a serotonin receptor antagonist inhibited gustatory responses, it can be stated that the net effect of serotonin is facilitatory to neural responses.

To confirm the above findings, a second specific 5-HT1A antagonist, NAD-299, was used in CT experiments, with the prediction that similar effects would be observed following its injection. Similar to WAY-100635, jugular injections of NAD-299 at any of the three different concentrations (175, 350, and 700 µg/kg) caused a universal inhibition in neural responses to gustatory stimuli. Significant inhibitions in responses to all taste qualities were observed following drug injection, except for one instance where average NaCl response amplitudes were reduced, but not significantly, post-NAD-299 injection at the lowest dosage used (175 µg/kg). As with WAY-100635, a gradual reversal in response decline also occurred following the injections, except NaCl responses following NAD-299 injection at 175 µg/kg and NaCl, sucrose, and quinine responses following NAD-299 injection at 350 µg/kg. Where it occurred, the recovery phase

peaked between 40 and 65 minutes post-injection, and peak effects of the drug were observed in the first ten minutes after each injection. Injecting NAD-299 at 175 µg/kg (n = 5) significantly reduce the CT response to sucrose by 68 % (p = 0.018), to quinine by 60 % (p = 0.004), and to HCl by 37 % (p = 0.008) (Figure 22). At 350 µg/kg NAD-299 (n = 6), significant reductions were observed in the CT response to NaCl (27 %; p = 0.014), sucrose (35 %; p = 0.025), quinine (34 %; p = 0.023), and to HCl (37 %; p = 0.034) (Figure 23). At 750 µg/kg NAD-299 (n = 6), the inhibition of CT responses was at 56 % for NaCl (p = 0.001), 59 % for sucrose (p = 0.001), 49 % for quinine (p = 0.002), and 33 % for HCl (p = 0.05) (Figure 24). Correlation between drug dose and the degree of inhibition was tested using the Pearson’s correlation test. A stronger correlation is reflected by an r-value (Pearson correlation coefficient) equal or near one. A p-value below 0.05 signifies a statistically significant correlation. For WAY-100635, there was a statistically significant correlation between dosage and the degree of inhibitions in neural responses for sucrose (r = 0.987; p = 0.006), but not for NaCl (r = 0.803 p = 0.099), quinine (r = 0.536; p = 0.232), or HCl (r = 0.738; p = 0.131). The graphs in Figure 25 depict the relative degrees of response inhibitions for each taste quality in response to different concentrations of WAY-100635 injections. Analysis of dose-effect relationship for NAD-299 revealed significant correlation between dosage and the degree of inhibitions in neural responses for NaCl (r = 0.999; p = 0.015), but not for sucrose (r = -264; p = 0.415), quinine (r = -4.21; p = 0.362), or HCl (r = -0.866; p = 0.167); (Figure 26). Although the amount of inhibition caused by either WAY-100635 or NAD-299 did not always statistically correlate to the dosage used, a clear pattern of significant inhibitions, followed by a recovery of neural responses, was observed after each injection.

In summary, the inhibition in CT responses that occurred following the antagonism of 5-HT1A receptors are consistent with the notion that through this particular receptor, 5-HT may be playing a modulatory role in the transduction of salty, sweet, bitter, and sour stimuli by facilitating their neural responses at the level of the chorda tympani. To study any possible metabolic side effects that can alter CT responses, a separate control recording experiment was performed to monitor heart rate and blood pressure both before and after systemic administration of WAY-100635 at the maximum concentration used in this study (200 µg/kg BW). Heart rate and systolic and diastolic blood pressure were recorded at 10 seconds intervals in response to a 300 µL single bolus jugular injection of WAY-100635 at 200 µg/kg. The injection was administered 15 minutes into the recording after a stable baseline was observed and blood pressure and heart rate were monitored up to 35 minutes after the injection. There was a small and brief fluctuation in baseline immediately following the injection but there was no significant change in mean resting blood pressure and heart rate (Figure 27). Due to logistical reasons, this experiment could not be repeated. However, the observations do suggest that intravascular injection of a 5-HT1A receptor antagonist does not produce cardiovascular side-effects significant enough to alter tongue perfusion and influence CT responses to lingual taste stimulation.

3.2.5. Effect of blocking 5-HT3 receptors on CT responses to gustatory stimulation

At present, it is assumed that the 5-HT3 receptor subtype is expressed on postsynaptic nerve endings within the taste bud. To examine any possible role for 5- HT3 in 5-HT’s action in the taste bud, recording experiments were carried out where CT responses to gustatory stimuli were measured before and after a 300 µL single bolus jugular injection of a 1000 µg/kg solution of ondansetron, a 5-HT3 receptor antagonist. Data averaged from 10 separate experiments showed that immediately after injecting

ondansetron, there was no significant change in mean CT response amplitudes to all taste qualities up to 65 minutes post-injection. When pre-injection and post-injection mean response amplitudes were compared for each taste quality, paired-samples T-test resulted in p-values greater than 0.05 (p = 0.295, 0.368, 0.396 and 0.081 for NaCl, sucrose, quinine and HCl respectively) (Figure 28). However, a significant decline in mean CT responses to sucrose was noticed between 55 and 65 minutes post-injection, relative to mean responses taken immediately post injection. Given that this decline occurred significantly after drug injection, it is more likely it arose from other non- specific factors. Since ondansetron caused no significant change in CT responses to all tested taste stimuli, it is evident that 5-HT3 has no major role in the modulatory effect of 5-HT in neural processing of the tested gustatory stimuli. Taken together, findings from these nerve recording experiments demonstrate that: (1) systemically injected reagents are able to reach lingual tissue and target their corresponding receptors. (2) ATP is the main neurotransmitter of the taste bud and blocking its P2 receptors significantly reduces the bud’s neural output to the chorda tympani. (3) Serotonin seems to act as a universal facilitator to all taste qualities. (4) This action of 5-HT appears to be mediated through its 5-HT1A receptor subtype.

In summary, serotonin’s 5-HT1A receptors are expressed in a subset of type II TRCs that include sweet-, bitter-, and umami- processing cells. The physiological action of serotonin is to facilitate chorda tympani responses to sweet, bitter, salty, and sour taste stimuli through activation of 5-HT1A receptors in type II taste receptor cells.

CHAPTER 4

DISCUSSION

This study sought to define the role of serotonin (5-HT) in the mammalian peripheral gustatory end organ, the taste bud, by characterizing cells that are targeted by serotonin in the bud, and by determining the net effect of serotonergic transmission on gustatory signals. The working hypothesis was that serotonergic transmission in the taste bud mainly serves to modulate taste signals at the taste receptor cell level before their synaptic transmission to gustatory afferent nerves. This work is the first to categorize the cells targeted by serotonin in the gustatory papillae of the rat’s tongue (both posterior and anterior) and the naso-incisor duct (NID), and the first to demonstrate the modulatory role of serotonin in the taste bud by interpreting the outcome of its release at the gustatory neural level, downstream of early and late gustatory transduction mechanisms that take place at the cellular level. The study phenotyped the cells that express serotonin’s only TRC-specific receptor in the taste bud, 5-HT1A, and revealed its expression in a subset of ATP-releasing type II cells, comprised mostly of cells that and process sweet, bitter, and umami tastes. ATP is the principle neurotransmitter of gustatory signals from the taste bud3, and previous studies have shown that both ATP and serotonin are released upon gustatory stimulation from type II and III cells respectively2,9. In the present study, physiological experiments demonstrated that serotonin facilitates the transmission of salty, sweet, bitter, and sour signals at the chorda tympani

level through the activation of 5-HT1A receptors. On first sight, this facilitatory action of serotonin may appear to be antithetical to previous reports that serotonin is inhibitory to ATP release at the cellular level9,64,98. This study proposes that serotonin produces its facilitatory action by preventing the too rapid release of ATP, and thus the desensitization of the P2X purinoceptors expressed on gustatory afferent nerve fibers. Findings from this work help to further the knowledge of serotonin and its mode of action at the peripheral gustatory level, and offer new insights into the impact of serotoninergic transmission on the flow process of gustatory information from taste cells to gustatory afferent nerves. Serotonin is a neurotransmitter that has wide-ranging effects and distribution in the mammalian body. Recently, it has been reported to have a modulatory role in the mammalian olfactory bulb8,141. In humans, patients taking serotonin reuptake inhibitors (SSRI) report significantly lower thresholds for sweet and bitter tastes than control subjects5. Serotonin’s function in the mammalian taste bud however is still debated. Its conserved expression in type III cells –the only cells in the taste bud that synapse with gustatory afferent nerve fibers– across many mammalian species had initially made it the most plausible candidate gustatory neurotransmitter between TRCs and gustatory afferent nerves. However, new findings from the study of peripheral taste physiology revealed that eliminating type III cells abolishes only sour taste, leaving other taste modalities intact39. 5-HT-releasing cells are therefore non-essential for tasting bitter, sweet, salty, and umami. Alternatively, studies began to show that the ATP is the main neurotransmitter from TRCs, and that gustatory neurotransmission is contingent on the 3. availability of P2X purinoceptors expressed on sensory afferent fibers ATP is hypothesized to be released in an atypical manner from type II cells, which lack classical synapses, in response to gustatory stimulation of type II and/or type III TRCs107,108109.

The discovery that ATP is the primary peripheral gustatory neurotransmitter has left the door open to uncover the actual role of serotonin in the taste bud, a topic that has regained increased research interest. Recent enquiries into the function of serotonergic transmission have revealed important new findings that seem to suggest a modulatory role for serotonin among TRCs. However, the focus of most studies has been increasingly shifting to purinergic transmission and the role of other signaling agents in the bud, except for few works that have examined the collateral effects of serotonin at the cellular level in relation to other signaling agents. Consequently, the actual role of serotonergic transmission in the taste bud, and the identity of cells and taste qualities modulated by serotonin remain elusive.

4.1. 5-HT and 5-HT1A are both expressed in taste buds across the oral cavity

Previous work had confirmed the expression of both 5-HT and the 5-HT1A receptor subtype in posterior tongue taste buds (circumvallate and foliate papillae taste buds). The current study extended these data to the taste buds of the fungiform papillae and the NID. Confirming the expression of 5-HT and 5-HT1A at those sites was also necessary for the interpretation results from the chorda tympani recording experiments. It was also essential to confirm that the pattern of co-expression of 5-HT1A with the investigated taste associated markers is consistent across taste buds in different sites of the oral cavity. Therefore, ICC experiments were conducted when possible in tissue that included sections from the anterior gustatory field including fungiform papillae and the naso-incisor duct (NID); (Figures 2 – 9).

4.2. The serotonin 5-HT1A receptor is expressed in type II TRCs that respond to sweet, bitter, and umami taste stimulation

Characterization of the 5-HT1A-expressing cell is a fundamental step in the study of serotonergic transmission and its role in the taste bud. The identity of serotonin- releasing cells had been widely confirmed as the type III cells39, but the identity of cells influenced by serotonin is not fully established. It has been suggested that type II cells4, which comprise cells that release ATP, express the 5-HT1A receptor, making them targets for modulation by serotonin. However, it hasn’t been established whether all type II cells express 5-HT1A.

The present study established the expression of serotonin’s 5-HT1A receptor in type II TRCs of the bud. Further phenotyping of the type II cells that expressed 5-HT1A revealed its presence in cells associated with the transduction of sweet, bitter, and umami taste stimuli. A useful strategy to study the function of any signaling agent in the taste bud is to categorize the cells targeted by that agent by examining the expression of its receptor(s). In the present study, determining the identity of cells targeted by serotonin in the taste bud was achieved by examining the co-expression of 5-HT1A receptors with a group of phenotypic markers linked to specific cell types, and the transduction of specific taste qualities. Combined data from double-label ICC from anterior and posterior tongue, and single cell RT-PCR from posterior tongue, showed that 5-HT1A consistently co-expressed with phenotypic markers associated with type II cells (GUST, GAD, NPY, NPY1R, CCK-8, SNAP and GAD), including receptors for sweet and umami (T1R), and bitter (T2R) taste qualities. Further, data showed that 5-HT1A does not co-express with PKD2L1 and NCAM, markers linked to presynaptic (type III) cells, which include serotonergic cells.

The co-expression patterns on 5-HT1A and phenotypic protein markers for type II cells lead to several intriguing signaling possibilities within the taste bud. One particularly interesting type II phenotypic marker used was gustducin, a taste-specific G- protein coupled to the family of metabotropic receptors for bitter tastants, T2Rs35. The

co-expression of gustducin has been documented with other markers associated with type II cells, including neuropeptides NPY and CCK-8, the NPY1 receptor (NPY1R), and the downstream effector protein PLCβ232. In the present study, the co-expression of 5-

HT1A and gustducin was examined using both double label ICC and single cell RT-PCR, and the results revealed extensive co-expression between the two markers in taste buds of the NID, and posterior and anterior tongue. Overall, type II cells expressed gustducin in more cells than 5-HT1A, as revealed by immunolabeling and molecular phenotyping experiments. Further, not all gustducin-positive cells expressed 5-HT1A. It is known that gustducin is expressed in all cells that process bitter tastes32, and all gustducin-positive cells have the ability to release ATP142. Therefore, the incomplete pattern of co- expression between 5-HT1A and gustducin seen in the present study suggests that not all bitter-transducing cells are directly targeted by serotonergic modulation. Nonetheless, it is not impossible for 5-HT1A-negative bitter cells to still be indirectly modulated via intermediary paracrine signals from 5-HT1A-positive cells that are directly inhibited by serotonin.

The expression of 5-HT1A receptor in type II TRCs was further confirmed by data showing its co-expression with other type II cell-associated markers, one of which was for neuropeptide Y (NPY). NPY is a neuropeptide member of the brain-gut peptide family, related to feeding and digestion, and is involved in the regulation of satiety signals in the brain143. Further, NPY has been shown in several independent studies to 144 co-express with serotonin receptors 5-HT1A and 5-HT1C in the rat amygdala . At the periphery, NPY is widely distributed among type II TRCs, and is always co-expressed with CCK-8 and VIP, where it inhibits cellular targets in a paracrine manner via its receptor in the bud73, NPY1 receptor (NPY1R). In the present study, double label ICC experiments for NPY and 5-HT1A, and for NPY1R and 5-HT1A revealed that 73% of the TRCs expressing

NPY and 66 % of those expressing NPY1R co-express 5-HT1A. The pattern of co-

expression was consistent throughout the posterior and anterior tongue, and the NID. These findings imply that cells that release NPY and those that are targeted by NPY can both be targeted by serotonin. Another type II cell-associated marker that was used in double label ICC experiments was CCK-8, an octapeptide hormone of the gastrointestinal system that is widely distributed in the cerebral cortex, striatum and hippocampus145. Like NPY, CCK-8 also plays a role in regulating feeding behavior at the brain level146. In the peripheral gustatory system, it has been located to a subtype of type II cells that sometimes co- express NPY, GUST, and VIP72. CCK’s modulatory function is excitatory and is mediated by the CCK-A receptor expressed in the same cell, thereby producing autocrine positive feedback. As present immunolabeling data revealed, most CCK-8-positive TRCs (~72%) in the NID and posterior and anterior tongue co-expressed 5-HT1A. In addition to the confirmation of 5-HT1A’s expression in type II TRCS, these results suggest that cells targeted by serotonin through the 5-HT1A receptor can be targets for the autocrine excitatory input of CCK-8. CCK-8 has been suggested to be modulatory to bitter taste processing within the taste bud72,147.

The co-expression of SNAP25 and 5-HT1A was also examined in the posterior tongue using double label ICC. SNAP25 is a Q-snare protein involved in the exocytotic fusion complex of synaptic vesicles of a wide variety of cells148. Many studies reported its co-expression in presynaptic (type III cells) in the mouse138,149,150. In those studies, ICC studies indicated that SNAP-25 is expressed in a separate population of TRCs from those expressing the presynaptic membrane protein syntaxin-1149, the sweet and umami receptor molecule T1R3138 or the type II cell-specific protein TRPM5 (transient receptor potential cation channel, subfamily M, member 5)138. In the rat, however, several studies have reported SNAP-25 co-expression with markers associated with type II cells including gustducin151,152, PLA2-IIA153, PLCβ2152, NCAM152, and GAD71,137. GAD, the rate-

limiting enzyme in GABA synthesis, is another marker that has previously been shown to co-express in varying degrees with markers linked to type II cells including GUST, VIP, and SNAP25 in the rat taste bud71,137. In the CNS, it plays the role of an inhibitory 154 neurotransmitter through its GABAA receptor . Further, previous work has established

GABA as an inhibitory gustatory neurotransmitter acting in a paracrine way on its GABAB receptor in the rat taste bud71,155,156, and confirmed its expression with GAD71, NPY73, and gustducin71 in type II TRCs. Based on these data, both SNAP25 and GAD are considered to be markers associated with type II cells in the rat. Present ICC data revealed that 89% of the SNAP25-expressing cells in the posterior tongue taste buds also expressed 5-HT1A. On the other hand, a narrow overlap between GAD and 5-HT1A- positive posterior tongue TRCs was seen in ICC experiments (18% of 5-HT1A-positive cells expressed GAD), and single cell RT-PCR (2 x GAD and 5-HT1A-positive cells out of 6 x

GAD- and 12 x 5-HT1A-positive cells). This pattern of co-expression suggests that a small percentage of type II TRCs that release GABA may be targets for the modulatory action of 5-HT through its 5-HT1A receptor expressed on those cells. The pattern of co- expression between 5-HT1A with SNAP25 and GAD also further confirm the expression of

5-HT1A in type II TRCs. Markers for type III cells included the neural cell adhesion molecule (NCAM), and the candidate sour receptor, polycystic kidney disease 2-like 1 protein (PKD2L1). NCAM is a homophilic binding glycoprotein that is widely distributed in the CNS and the periphery. It is involved in cell-cell adhesion, neurite outgrowth, synaptic plasticity, and learning and memory157. The association of NCAM with markers associated with type III cells has been well documented in rodents’ taste buds23,152,158. As predicted, the present

ICC data did not reveal any co-expression of NCAM and 5-HT1A in TRCs of the posterior tongue. The same pattern of co-expression was seen with the candidate sour receptor

PKD2L1 and 5-HT1A in posterior tongue taste buds, as shown by RT-PCR data. PKD2L1 is

member of the transient receptor potential (TRP) channel family of receptors, and is now believed to be the most likely receptor for sour tastants39. Exclusively expressed in a subset of type III cells, PKD2L1 has been widely used as a reliable marker for those cells. The present study showed using single cell RT-PCR technique that cells expressing

PKD2L1 did not co-express 5-HT1A or any other marker associated with type II cells. This result suggest that 5-HT1A is not expressed in presynaptic (type III) cells, which include serotonergic cells and cells that process sour stimuli.

The expression pattern of 5-HT1A relative to taste receptors, investigated using single cell RT-PCR experiments, suggests a wide distribution of 5-HT1A across cells involved in the transduction of sweet, bitter, and umami tastes. The co-expression of the bitter taste receptor T2R9 in half of the total number of cells that expressed 5-HT1A

(n=6 out of 12 5-HT1A-expressing cells) is further proof that the latter is present in type II TRCs that detect and process bitter tastes. Gustducin, the G-protein coupled to the T2R9 receptor molecule was expected to always co-express with T2R9 in single cell RT-PCR experiments. The present study revealed that of the 14 x gustducin- and 12 x T2R9- expressing cells, 10 x TRCs co-expressed both markers. The other half of the twelve TRCs that expressed 5-HT1A co-expressed T1R3, the receptor molecule for sweet and umami tastants, indicating that 5-HT1A is present in TRCs responsible for the transduction of those two taste qualities. Single cell RT-PCR data also provided internal control for the expression of markers for type II TRCs with each other in the same single cells, and the absence of their expression in cells that expressed PKD2L1. Out of twenty five total TRCs tested, data revealed the co-expression of gustducin and GAD in two TRCs, GAD and T2R9 in one TRC, GAD and T1R3 in three TRCs, and the co-expression of GAD, gustducin, and T1R3 together in another cell. These results agree with findings from previous studies that reported the co-expression of GAD with gustducin and T1R3 TRCs in murine taste

buds71,150,159. Further, out 14 x gustducin- and 10 x T1R3- expressing TRCs, five cells expressed both molecules. This pattern of co-expression is also in line with previous studies that found those two molecules co-localized in mouse TRCs23,160,161,162. Findings from the current experiments therefore serve as internal control for the specificity of the primers used and validate the pattern of co-expression of the selected type II markers. On the other hand, data also showed unexpected co-expression patterns. Among the 32 x CK8-positive TRCs used in RT-PCR, two TRCs co-expressed T1R2 and T2R9 out of the total of the 12 x T2R9- and 10 x T1R3-expressing cells. This was an unexpected outcome, since the current conventional model dictates that a single TRC does not express receptors for more than a single taste modality163. Indeed, the general consensus, based on the studies that examined the interrelationship between the T1R and T2R families of receptors in mammalian taste buds, supports the notion that those two molecules are expressed in non-overlapping populations of TRCs164,165,166,167,168. However, a recent study has reported their co-expression in the solitary chemoreceptor cells in mice169. Therefore, their co-localization in two cells in the current experiments cannot be entirely ruled out as experimental error.

4.3. Physiological effects of serotonin at the gustatory cellular and the neural levels The current phenotyping experiments established that the transduction of sweet, bitter, and umami taste stimuli in type II TRCs may be targets for 5-HT1A- mediated modulation by serotonin released from type III cells. The nature of this modulation has been studied in previous works where it was found that activation of 5- 64 HT1A receptors causes inhibition of target cells in rat taste buds , and inhibition of taste- evoked ATP release from type II TRCs in mice9. These findings are supported by the current data, which show that targets for serotonergic modulation in the taste bud

belong to the type II TRCs, which release ATP in response to taste stimulation.

Furthermore, the current results suggest that 5-HT1A-expressing type II TRCs -which are themselves inhibited by serotonin- can modulate neighboring TRCs, and hence other tastes, by releasing their endogenous neuromodulators, GABA and NPY. Further, based on their expression pattern with the various neuromodulators tested in phenotyping experiments, the 5-HT1A-expressing cell may also be auto-inhibited by the autocrine excitatory input of CCK-8, and be subject to NPY1R-mediated paracrine and/or autocrine inhibitions originating in adjacent NPY-releasing TRCs. Therefore, at the cellular level, serotonin appears to play an inhibitory modulatory action on the transduction of multiple taste modalities in type II cells. This includes the inhibition of taste-evoked ATP release from those cells. Conversely, the effects of serotonin on the flow of taste information have not been investigated at the level of the gustatory nerves. The current study measured those effects by recording whole chorda tympani nerve responses to taste stimuli following the antagonism of serotonin receptors, and demonstrated a crucial role for 5-

HT1A, but not 5-HT3, in mediating serotonin’s effects of in the taste bud. This was an expected outcome considering the results from ICC and RT-PCR experiments, which demonstrated the extensive expression of 5-HT1A in a variety of TRCs that process multiple taste modalities. Indeed, data revealed that blocking the 5-HT1A receptor with two different antagonists (WAY-100635 and NAD-299) at different concentrations caused a universal decrease in CT response to salty, sweet, bitter, and sour lingual taste stimuli. On the other hand, blocking the 5-HT3 receptor with a specific antagonist (ondansetron) caused no significant effects on CT responses. It was concluded therefore that using the present experimental setting, the contribution of 5-HT3 receptors in mediating the effects of serotonin cannot be determined. Ironically, before the discovery that ATP is the primary gustatory neurotransmitter from TRCs, 5-HT3 was

initially hypothesized as the receptor that mediates serotonergic transmission of taste signals to gustatory afferents. To date, several studies have attempted to investigate a possible role for 5-HT3 in mediating the action of serotonin in the taste bud. Although results from molecular and immunolabeling approaches suggest its expression in gustatory nerve ending, electrophysiological and behavioral experiments3,9,98 as well as the current work, do not show a major effect for 5-HT3 in mediating the functional effects of serotonin in the taste bud. It may be possible that 5-HT3 receptors are involved in the modulation of GABAergic, or adrenergic receptors also present on gustatory afferent nerve endings. A different experimental design would be required to test such possibility.

4.3.1. Jugular injections of gustatory modulators produce immediate effects recordable at the level of the chorda tympani Changes in the amplitudes of CT responses to taste stimulation observed after the jugular injection of 5-HT1A receptor antagonists suggest that 5-HT1A is implicated in mediating the modulatory effects of serotonin on gustatory signals. The use of intravascular injections of gustatory modulators in the lingual artery has been previously documented in previous rat whole nerve chorda tympani recording experiments170. In the present study, injections of serotonin receptor antagonists were administered in the jugular vein. Since this new approach has not been previously employed, it was necessary to test the validity and examine the efficacy of the experimental preparation. The study tested whether a drug administered in the jugular vein could reach gustatory tissue and produce immediate detectable effects using chorda tympani recordings. This was achieved by recording CT response to jugular ATP injections, injecting a purinergic blocker (PPADS), and then examining the effects of the blocker on CT responses to gustatory stimuli.

In addition to its presumed presynaptic modulatory role, it is now agreed that that ATP’s most crucial gustatory role is to transmit taste signals from TRCs to gustatory afferent nerves via stimulation of the P2X2/P2X3 purinoceptors expressed on nerve endings9. After performing its physiological roles, taste-evoked ATP is thought to be degraded by ecto-ATPases widely distributed among mammalian TRCs171,172,173. Ecto- ATPases are transmembrane forms of ATPase (adenosinetriphosphatase) enzymes that hydrolyze extracellular 5′-triphosphates including ATP174. Recently, it has been suggested that a member group of the ecto-ATPase family, NTPDase 1, 2, 3, and 8, constitute the most plausible candidates for controlling the levels of extracellular175 with their extracellular active sites and a preferential affinity to ATP176. The presence of these ecto-ATPases explains the rapid decline of the lingually-induced ATP responses observed in the present study. Jugular administration of ATP was therefore predicted to cause transient neural stimulations that are quickly terminated by the activity of ecto-ATPases. As evidenced by the results, jugular injections of ATP at three different concentrations (0.75, 1.5 or 3 mg ATP/Kg BW) produced dose-dependent transient neural responses. These observations serve as further proof of the excitatory action of ATP at the peripheral gustatory level, but more importantly, they confirm that jugular injections of an excitatory molecule can quickly reach the gustatory epithelium and produce recordable responses at the gustatory neural level. To further validate the experimental setup, the study assessed the effects of blocking ATP neurotransmission on CT response to ATP injection and to lingual gustatory stimulations. Dose-dependent inhibitions in responses to taste stimuli and to ATP injections were observed after jugular injections of 10 or 20 mg/kg solutions of the selective P2 purinergic blocker PPADS. These results demonstrate that jugular injection of an inhibitory gustatory agent can also be detected through CT recordings as a decrease in neural responses to lingual taste stimuli.

Further proof of the validity of the experimental setup came from testing the effects of saline injections on CT responses. As demonstrated by the results, neural responses did not significantly fluctuate following saline injections. This outcome ensures that changes in CT responses observed with the subsequent injections of experimental drugs were caused by the functional effects of the drug, and not to any degree by physiological side effects of the vehicle solution.

4.3.2. Blocking the 5-HT1A receptor inhibits chorda tympani responses to lingual taste stimuli

Blocking the 5-HT1A-receptor-mediated serotonergic transmission in the taste bud caused inhibitions in gustatory neural responses that were analogous to the ones caused by blocking P2 purinoceptors using PPADS. Injecting either of the 5-HT1A antagonists (WAY-100635 or NAD-299) in the jugular vein caused immediate declines in the magnitudes of neural response to salty, sweet, bitter, and sour taste stimuli. The inhibitions caused by both drugs were significant at all doses, (with only one exception where NaCl responses did not significantly decrease after injecting NAD-299 at 175 µg/kg). Although in most cases the degree of inhibition did not significantly correlate with the tested doses, there was a clear pattern of decline in neural responses immediately after the injections were administered. A statistically significant correlation between drug dosage and degree of response inhibition may be achieved by increasing the number of experiments. The recovery from the decline in neural responses that was observed with all tested WAY-100635 doses and most tested NAD-299 doses suggest that the effects of either drug on neural responses are not irreversible. It remains to be determined however whether the recovery of responses from their decline -where it occurred- is due to increasing drug metabolism or to compensatory or adaptive mechanisms in response to the impeded 5-HT1A-mediated modulation.

Although umami taste was not included among the stimulus solutions due to the current unavailability of a stimulus that specifically targets umami receptors, it is projected that umami taste would be influenced by serotonin in the same way as the other tastes that were studied. This is based only on the phenotyping experiments in this study which showed the expression of 5-HT1A receptors in umami-, sweet-, and bitter-transducing cells.

4.4. The physiological mechanisms involved in serotoninergic modulation of gustatory stimuli Findings from the current electrophysiological experiments suggest that in its in situ form, serotonin acts to facilitate the transmission of gustatory signals. This implies that serotonin has opposite net actions at the levels of the TRC and the gustatory nerve. Indeed, the first data to establish a physiological effect of serotonin on mammalian taste has demonstrated that activation of 5-HT1A receptors causes inhibition of target TRCs in rat taste buds64. Using patch clamping experiments, researchers demonstrated that activation of the 5-HT1A receptor by exogenously applied 5-HT or a 5-HT1A agonist inhibits calcium-activated potassium current by up to 50%. However, bath-applying a 5-

HT3 receptor agonist did not produce noticeable effects. In another study in mice, it was subsequently demonstrated that taste-evoked ATP release from type II cells was 9 inhibited by 5-HT and enhanced by blocking 5-HT1A with the antagonist WAY-100635 .

Previous findings therefore support the notion of a 5-HT1A-mediated inhibitory effect for serotonin at the cellular level in the mammalian taste bud. The mechanisms through which serotonin inhibits this release may involve a 5-HT1A-mediated modulation of gap junctions, the putative channels for calcium-dependent ATP release downstream of T1R or T2R activation107,108. Activation of either of those two G-protein coupled receptors by bitter, sweet, or umami tastants eventually leads to the activation of PLCβ2, an enzyme

that hydrolyses the membrane-bound phospholipid, phosphatidylinositol 4,5- bisphosphate (PIP2) into diacylglycerol and inositol 1,4,5-trisphosphate. The sequence of 2+ events downstream of PIP2 breakdown includes Ca release from intracellular stores, activation of TRPM5 channels, cellular depolarization, and neurotransmitter (ATP) 80 release . Several independent studies have shown that the cellular levels of PIP2 regulate the signaling activity of ion channels177,178,179,180,181, including connexin and pannexin hemichannels182,183, constituents of gap junctions. Data from preliminary patch-clamping work have also demonstrated that pharmacologically altering PIP2 resynthesis modulates the excitability of TRCs184. The cellular inhibition caused by the 98,185,186,187 activation of 5-HT1A receptors reduces PIP2 resynthesis , and consequentially would inhibit the activity of connexin/pannexin channels, resulting in a reduction in ATP release (Figure 29). In contrast, at the gustatory neural level, the present data appear to offer a conflicting role for serotonin. In the effort to reconcile those findings it is crucial to carefully consider the physiological mechanisms that take place after ATP release following gustatory stimulation (Figure 29). As previously stated, gustatory stimulation ultimately causes type II cells to release ATP, which subsequently assumes the simultaneous roles of a postsynaptic neurotransmitter and a presynaptic neuromodulator. As a neurotransmitter, ATP targets P2X2/P2X3 purinoceptors on afferent nerve endings to transmit the gustatory signals3. On the other hand, ATP’s neuromodulatory role has been found to involve positive autocrine and paracrine 9,106 feedback mechanisms . ATP may activate P2Y1 and P2X2 receptors expressed in type II cells70,110 via autocrine routes, thereby auto-enhancing and triggering its own release respectively. ATP may also activate in a paracrine manner P2Y receptors expressed on type III cells, including serotonergic cells, ultimately triggering serotonin release9. Serotonin can also be directly released from type III cells upon their stimulation with

2 sour stimuli . Regardless of the cause of its release, serotonin activates 5-HT1A receptors expressed in a group of type II cells through a negative paracrine feedback, inhibiting those cells and the release of ATP. Therefore, blocking the 5-HT1A-receptor-mediated serotonergic transmission in the taste bud would be logically assumed to disinhibit ATP release, and thus enhance the stimulus for synaptic P2X purinoceptors. On the surface, this might be expected to result in larger neural responses. The depression in neural responses observed in the present study, although contradictory to initial expectations, can be explained by taking into account the physiological characteristics of P2X purinoceptors on afferent nerve endings. As previously explained, P2X2/P2X3 are ATP- gated receptors that mediate the neurotransmission of the gustatory signals from taste buds. Those receptors are well documented to have a fast rate of activation and desensitization, combined with a slow rate of recovery from desensitization188,189,190. Consequently, in the context of the current study, the magnitude of CT responses is not directly proportional to the levels of taste-evoked ATP release. Without the inhibitory serotoninergic input, which regulates taste-evoked ATP release from type II cells, gustatory neural responses would actually be inhibited thanks to diminished neurotransmission caused by P2X2/P2X3 purinoceptor desensitization. This argument may explain the decrease seen in the amplitude of neural responses to gustatory stimuli following the blockage of serotonergic transmission.

4.5. Conclusion The present study confirmed the hypothesis that serotonin plays a modulatory role on gustatory signals generated in the taste bud. The ultimate goal from the study was to shed more light on the nature of this modulation by identifying its target cells and taste modalities. As interpreted at the neural level through chorda tympani nerve

recording and gustatory stimulation of the anterior tongue, the modulatory effect of serotonin seems to impact sweet, bitter, sour, and salty stimuli. In summary, the present study demonstrated that serotonin is a crucial player in a finely tuned feedback mechanism involving 5-HT1A receptors, ATP, and the P2X2/P2X3 purinoceptors. Serotonin’s primary role in the mammalian taste bud is therefore to modulate taste signals prior to prior to the final neural output by controlling the release of the gustatory neurotransmitter ATP.

Manufacturer, species, Antigen Immunogen Dilution type, catalog number

Santa Cruz Biotech, rabbit A peptide mapping within a highly divergent 1:1000 G polyclonal, affinity purified α gust domain of G of rat origin α gust IgG, sc–395

SNAP- Human crude synaptic immunoprecipitate that Millipore, mouse 1:300 25 recognizes SNAP-25 Protein monoclonal, MAB331

Chemicon, rabbit NCAM Highly purified chicken NCAM polyclonal, immunoaffinity 1:300 purified, AB5032

Immunostar, rabbit Rabbit NPY coupled to methylated BSA with NPY polyclonal, affinity purified, 1:300 glutaraldehyde 22940

Immunostar, rabbit Recognizes amino acids 356-382 of the rat NPY1R polyclonal, affinity purified, 1:300 NPY1 receptor 24506

CCKAr Synthetic peptide rat CCK-A R. Detects CCK- Neuromics, affinity 1:300 A receptor amino acids 256-267 in rat purified, rabbit polyclonal

Chemicon, rabbit CCK-8 Recognizes octapeptide CCK8 1:300 polyclonal, AB1973

Serotonin conjugated to BSA. Recognizes Millipore, rat monoclonal, 5HT 1:150 serotonergic sites is fixed tissue sections MAB352 Synthetic peptide corresponding to a region Chemicon, rabbit 5HT located in the large third intracellular loop of 1:150 1A polyclonal, AB15350 the rat and mouse 5-HT1A receptor protein

Table 1: List of primary antibodies used in the present study

Product Target Accession No. Primers Reference size

(Chen K, et al., CCTCTTCTGCCTCAGTGTCC T1R3 NM_130818 468 bp 2010)191 TAAGCTAGCATGGCGAAGGT

TTTCATGGGCAATCTCCTTC (Zhang Y, et al., T2R9 NM_023999 514 bp 192 CATGTGGCCCTGAGATCTTT 2003)

ATGCAGAACATGAGCATC (Kishi et al., CK8 NM_199370 440 bp 193 ACAGCCACTGAGGCTTTA 2001)

(McLaughlin et GTTGGCTGAAATAATTAAACG Gustducin X65747 251 bp al., 1992)35 ATCTCTGGCCACCTACATC

CAGAGGAAGGTGCTCTTTGG (Uchida S, et al., 5HT (A) NM_012585.1 171 bp 194 1A AAGAAGAGCCTGAACGGACA 2010)

CAGAGGAAGGTGCTCTTTGG (Uchida S, et al., 5HT (B) NM_012585.1 201 bp 1A AGCTTAGGAACTTCGTCGGCA 2010)194

(Borroto-Escuela GGCAGCCAGCAGAGGATGAA 5HT (C) NM_012585.1 336 bp DO et al., 1A CCCCCCAAGAAGAGCCTGAA 2012)195

GAGCTGGTCTTCTTTGTCCG PKD2L1 NM_001106352.1 266 bp (Florea L, et al., CTGCAGTCTCCTTCCAGACC 196 2005)

(Kakinohana O, TCTTTTCTCCTGGTGGTGCC GAD NM_012563.1 373 bp et al., CCCCAAGCAGCATCCACAT 2012)197

Table 2. Primer sequences used in RT-PCR reactions

Tables 3, 4, and 5: Mean 1 represents the mean percentage of 5-HT1A labeled cells which co-expressed the corresponding phenotypic marker per cross-sectioned taste bud. Mean 2 represents the mean percentage of all phenotypic marker labeled cells which co-expressed the corresponding 5-HT1A per cross-sectioned taste bud. Means are expressed ±SE.

Anterior tongue: Mean 1 Total 5-HT Pheno- Double- Mean 2 fungiform 1A (%DL/ taste labeled typic labeled ± SE (%DL/p.m/ ± SE papillae + 5-HT1A buds cells Marker cells TBs/ NID (n = 8 /TBs/ section) animals) section)

5-HT1A/NPY 60 60 48 36 60.00 2.19 75.00 0.98

5-HT / 1A 48 96 80 56 58.33 3.70 70.00 2.03 NPY1R

5-HT1A/GUS 56 44 96 36 81.82 0.63 37.50 6.93

5-HT1A/ 54 48 42 36 75.00 0.98 85.71 0.46 CCK-8

Table 3. Quantitative immunocytochemical double labeling patterns of 5-HT1A and phenotypic taste receptor cell markers in fungiform papillae of the anterior tongue and in the NID

Posterior Mean 1 tongue: CV + Total 5-HT1A Pheno- Double- (%DL/ Mean 2 foliate taste labeled typic labeled 5-HT1A/ ± SE (%DL/p.m/ ± SE papillae (n = buds cells Marker cells TBs/ TBs/section) 8 animals) section) 60 70 142 59 84.29 0.85 41.55 9.11 5-HT1A/NPY

84 59 80 46 77.97 1.04 57.50 3.17 5-HT1A/NPY1R

5-HT /NCAM 180 135 175 0 0.00 0.00 1A

142 150 162 134 89.33 1.20 82.72 2.18 5-HT1A/SNAP

92 91 84 63 69.23 2.38 75.00 1.71 5-HT1A/GUS

5-HT / 1A 84 81 138 57 70.37 0.00 41.30 0.00 CCK-8

Table 4. Quantitative immunocytochemical double labeling patterns of 5-HT1A and phenotypic taste receptor cell markers in circumvallate and foliate papillae of the posterior tongue.

Total: Mean 1 Anterior + Total 5-HT Pheno- Double- Mean 2 1A (%DL/ Posterior taste labeled typic labeled ± SE (%DL/p.m/T ± SE 5-HT / (n = 16 buds cells Marker cells 1A Bs/sec) TBs/sec) animals) 120 130 190 95 73.08 2.90 50.00 9.50 5-HT1A/NPY

5-HT1A/ 132 155 160 102 65.81 4.62 63.75 5.14 NPY1R

5-HT1A/ 180 135 175 0 0.00 0.00 NCAM

5-HT1A/ 142 150 162 134 89.33 1.20 82.72 2.18 SNAP 148 135 180 99 73.00 2.96 55.00 7.72 5-HT1A/GUS

5-HT / 1A 138 129 180 93 72.09 3.00 51.67 8.56 CCK-8 5-HT / 1A 155 135 122 34 25.19 2.37 27.87 4.01 GAD

Table 5. Quantitative immunocytochemical double labeling patterns of 5-HT1A and phenotypic taste receptor cell markers totaled from all three gustatory sites (posterior and anterior tongue and the NID)

Group “A” Cell # CK8 Gus T1R3 T2R9 5HT1A(A) 5HT1A(B) 5HT1A(C) 1 + + - + + + + 2 + - + - - - - 3 + - - + + + + 4 + + - + - - - 5 + - - - + + + 6 + + - + + + +

Group “B”Cell # CK8 Gus T1R3 T2R9 PKD2L1 GAD 5HT1A(C) 7 + + + - - + + 8 + + + - - - - 9 + - + - - - + 10 + + + + - - + 11 + + + + - - + 12 + - + - - + - 13 + - - - - + - 14 + + - + - - + 15 + - + - - - - 16 + - - - + - - 17 + - - + - + - 18 + ------19 + ------20 + + - + - - - 21 + + - - - - - 22 + + - + - - - 23 + + - + - - - 24 + ------25 + ------26 + ------27 + - + - - - + 28 + - - - + - - 29 + - + - - + + 30 + + - + - - - 31 + + - - - + +

Table 6. Gene expression in single taste cells

Figure 1. Sample integrated whole chorda tympani nerve response to a 0.5 M NaCl solution taste stimulus illustrating response components and measurement of response amplitude.

Figure 2. Single labeling immunocytochemistry showing TRCs in rat fungiform papillae immunopositive for 5-HT (left) and 5-HT1A (middle and right) - (scale bar is 20 micron)

Figure 3. (Top) Double labeling ICC showing examples of TRCs in rat CV immunopositive for α-Gustducin (left), 5-HT1A (middle) and overlaid images of the two (right).

(Bottom) TRCs in rat NID taste buds immunopositive for α-Gustducin (left), 5-HT1A (middle) and overlaid images of the two panels (right).

Figure 4. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5-

HT1A (left), GAD (middle) and overlay (right).

Figure 5. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5-

HT1A (left), SNAP25 (middle) and overlay (right).

Figure 6. (Top) Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5-HT1A (left), CCK-8 (middle) and overlay (right).

(Bottom): TRCs in rat NID taste buds immunopositive for 5-HT1A (left), CCK-8 (middle) and overlay (right).

Figure 7. (Top): Double labeling ICC showing examples of TRCs in rat CV immunopositive for NPY (left), 5-HT1A (middle) and overlay (right).

(Bottom): TRCs in rat NID taste buds immunopositive for 5-HT1A (left), NPY (middle) and overlay (right).

Figure 8. (Top): Double labeling ICC showing examples of TRCs in rat CV immunopositive for NPY1R (left), 5-HT1A (middle) and overlay (right).

(Bottom): TRCs in rat NID taste buds immunopositive for 5-HT1A (left), NPY1R (middle) and overlay (right).

Figure 9. Double labeling ICC showing examples of TRCs in rat CV immunopositive for 5-

HT1A (left), NCAM (middle) and overlay (right).

Figure 10. Gel electrophoresis illustrating single cell RT-PCR products from seven taste receptor cells tested with a variety of primer sets corresponding to all seven phenotypic markers, including taste receptor genes. Each row represents PCR results from a single primer set; each column represents results from a single cell.

Figure 11. Sample integrated whole nerve chorda tympani responses to lingually applied taste solutions (0.5 M NaCl, 1.0 M sucrose, 0.03 M quinine, and 1.0 N HCl)

Figure 12. Sample integrated whole nerve CT responses to quinine and to ATP solutions at different concentrations

Figure 13. Sample integrated CT responses to lingual quinine stimuli (0.3 M) prior to- (PRE), immediately after (11 and 14 min), and 45 min following a 10mg/kg PPADS injection

Figures (14 through 26): For every stimulus, blue bars represent averaged (from separate recording experiments) normalized means of amplitudes of the last three consecutive CT responses to that stimulus prior to drug injection (pre-). Red bars represent averaged normalized means of amplitudes of the first three consecutive responses after the injection (post-), and green bars represents averaged normalized means of amplitudes of last three consecutive responses after the injection (between 30 and 65 minutes after injection).

Effect of PPADS @ 10 mg/Kg on CT responses Effect of PPADS @ 20 mg/Kg on CT responses

Figure 14. Effect of P2-purinoceptor antagonist PPADS on the CT response to lingual stimuli: (Left) at 10 mg/Kg (n = 3), PPADS jugular injection caused a 53 % decrease in CT response NaCl, 82 % decrease in CT response to sucrose, 79 % decrease in CT response to quinine, and 75 % decrease in CT response to HCl. (Right) at 20 mg/Kg (n = 3), the decrease in CT response was 87 % for NaCl, 85 % for sucrose, and 97 % for quinine, and 99 % for HCl. Maximum inhibition occurred in the first ten minutes after the injection and responses recovered to 42 % pre-injection values 62 minutes after PPADS injection.

Figure 15. Sample chorda tympani responses to intravenous ATP injections (3 mg/kg) prior to- (left), immediately following- (middle), and 45 min following (right) a 10mg/kg PPADS injection.

Effect of PPADS @ 10 mg/Kg on CT responses Effect of PPADS @ 20 mg/Kg on CT responses to 1.5 mg/Kg ATP injection to 1.5 mg/Kg ATP injection

Figure 16. Effect of P2-purinoceptor antagonist PPADS on the CT response to systemic ATP injection: Summated CT responses to a 300 µL single bolus jugular injection of 1.5 mg/kg ATP solution prior to- (blue bars), immediately following- (red bars), and between 30 and 65 minutes after a 300 µL single bolus jugular injection of PPADS solution at either 10mg/kg (left; n = 3), or a 20mg/kg (right; n = 3). At 10 mg/Kg, PPADS caused a 97% decrease in CT response to ATP injection at 1.5 mg/Kg. Responses recovered to 53 % of pre-injection value 50 minutes after the injection. At 20 mg/Kg, the decrease was 98.8 %. Maximum inhibition occurred in the first ten minutes after the injection, and responses recovered to 42 % pre-injection values 62 minutes after PPADS injection.

Figure 17. Upper left: normalized summated data averaged from eleven experiments showing mean CT response amplitudes for NaCl, sucrose, quinine and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after a single bolus 300 µL jugular saline injection (recovery). Upper right: line plot showing the effect of saline injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording. No significant difference in response magnitudes was observed. Lower: Sample CT responses for all four taste qualities taken prior to, immediately after, and 65 minutes after saline injection.

Figure 18. left: normalized summated data averaged from six experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 10 µg/kg. The injection caused a 16 % decrease in mean CT response to NaCl, 40 % to sucrose, 21 % to quinine and 16 % to HCl. Maximum inhibition occurred in the first ten minutes after the injection, and responses slowly recovered to near pre-injection values 45 minutes following the injection. Right: line plot showing the effect of WAY 100-635 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

Figure 19. left: normalized summated data averaged from six experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 25 µg/kg. The injection caused a 32 % decrease in mean CT response to NaCl, 47 % to sucrose, 68 % to quinine, and 49 % to HCl. Maximum inhibition occurred in the first ten minutes after the injection, and responses slowly recovered to near pre-injection values 50 minutes following the injection. Right: line plot showing the effect of WAY 100-635 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

Figure 20. Upper left: normalized summated data averaged from eight experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 100 µg/kg. The injection caused a 31 % decrease in mean CT response to NaCl, 50 % to sucrose, 46 % to quinine, and 36 % to HCl. Peak inhibition occurred in the first ten minutes post- injection, and responses slowly recovered to near pre-injection values 50 minutes following the injection. Upper right: line plot showing the effect of WAY 100-635 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording. Lower: Sample CT responses for all four taste qualities taken prior to, immediately after, and 65 minutes after saline injection.

Figure 21. left: normalized summated data averaged from six experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a WAY-100635 solution at 200 µg/kg. The injection caused a 34 % decrease in mean CT response to NaCl, 59 % to sucrose, 57 % to quinine, and 51 % to HCl. Maximum inhibition occurred in the first ten minutes after the injection, and responses slowly recovered to near pre-injection values 64 minutes following the injection. Right: line plot showing the effect of WAY 100-635 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

Figure 22. left: normalized summated data averaged from five experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a NAD-299 solution at 175 µg/kg. The injection caused a 1 % non- significant decrease in mean CT response to NaCl, 68 % to sucrose, 60 % to Quinine, and 37 % to HCl. Maximum inhibition occurred in the first ten minutes after the injection (except NaCl responses), and affected responses slowly recovered to near pre-injection values 40 minutes following the injection. Right: line plot showing the effect of NAD-299 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

Figure 23. left: normalized summated data averaged from six experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 65 minutes after (recovery) a 300 µL single bolus jugular injection of a NAD-299 solution at 350 µg/kg. The injection caused a 27 % decrease in mean CT response to NaCl, 35 % to sucrose, 34 % to quinine, and 37 % to HCl. No recovery was noted except for HCl responses, which saw maximum recovery 60 minutes after injection. Right: line plot showing the effect of NAD-299 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

100

Figure 24. left: normalized summated data averaged from six experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 90 minutes after (recovery) a 300 µL single bolus jugular injection of a NAD-299 solution at 700 µg/kg. The injection caused a 56 % decrease in mean CT response to NaCl, 59 % to sucrose, 49 % to quinine, and 33 % to HCl. Maximum inhibition occurred in the first ten minutes after the injection, and affected responses slowly recovered to near pre-injection values 85 minutes following the injection. Right: line plot showing the effect of NAD-299 injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

101

Figure 25. Graphs showing mean percentage differences in CT responses to different taste stimuli taken before (pre-), immediately after (post-), and at 40 to 65 minutes (recovery) after injecting saline or WAY-100635 at various concentrations.

102

Figure 26. Graphs showing mean percentage differences in CT responses to different taste stimuli taken before (pre-), immediately after (post-), and at 40 to 65 minutes (recovery) after injecting saline or NAD-299 at various concentrations.

103

Figure 27. Effect of WAY-100635 on mean blood pressure (left), and heart rate (right). (n=1); WAY-100635 at 200 µg/kg had no significant effects on mean resting blood pressure and heart rate in the period of time following the injection when CT responses would be measured.

104

Figure 28. left: normalized summated data averaged from 6 experiments showing mean CT response amplitudes for NaCl, sucrose, quinine, and HCl prior to- (pre), immediately after- (post), and 30 to 90 minutes after (recovery) a 300 µL single bolus jugular injection of a ondansetron at 1000 µg/kg. The injection caused no significant change in CT response to NaCl, sucrose, quinine, and HCl immediately after injection. Right: line plot showing the effect of ondansetron injection on the differences in response magnitude percentage for each taste quality at the 3 time periods of recording.

105

Figure 29. Illustration of the proposed mechanism of the relationship between serotonergic and purinergic signaling in the taste bud.

106

LIST OF REFERENCES

1. Esakov AI, Golubtsov KV, Solov'eva NA. (1983) Significance of serotonin in the activity of the taste receptor apparatus of the frog Rana temporaria (in Russian). Zh Evol Biokhim Fiziol 19: 62-67.

2. Huang YJ, Maruyama Y, Lu KS, Pereira E, Plonsky I, Baur JE, Wu D, Roper SD. (2005). Mouse taste buds use serotonin as a neurotransmitter. J Neurosci. 25:843-847.

3. Finger, T.E., Danilova, V., Barrows, J., Bartel, D.L., Vigers, A.J., Stone, L., Hellekant, G., Kinnamon, S.C. (2005). ATP signaling is crucial for communication from taste buds to gustatory nerves. Science 310:1495-1499.

4. Kaya, N., Shen, T., Lu, S.G., Zhao, F.L., Herness, S. (2004). Expression and localization of serotonin receptor subtypes in rat taste buds. American Journal of Physiology: Regulatory, Integrative, and Comparative Physiology 286:R649-R658.

5. O’Driscoll S, McRobie E, Ayres C, Mileusnic N, Heath TP, Melichar JK, Donaldson LF. (2008). Serotonin-specific reuptake inhibitors (SSRI) blunt bitter taste acutely (minutes), but enhance it chronically (hours) in normal healthy humans. Proc Physiol Soc. 11:PC102.

6. Heath TP, Melichar JK, Nutt DJ, Donaldson LF. (2006). Human taste threshold are modulated by serotonin and noradrenaline. J Neurosci. 26:12664-12671.

7. Kloppenburg P, Mercer AR. (2008). Serotonin modulation of moth central olfactory neurons. Annu Rev Entomol. 53:179-90.

107

8. Petzold, G.C., Hagiwara, A. and Murthy, V.N. (2009). Serotonergic modulation of odor input to the mammalian olfactory bulb. Nature Neurosci. 12:784-791.

9. Huang YA, Dando R, Roper SD. (2009) Autocrine and paracrine roles for ATP and serotonin in mouse taste buds. J Neurosci 29: 13909-13918.

10. Miller IJ Jr, Spangler K. (1982). Taste bud distribution and innervation on the palate of the rat. Chem Senses 7:99-108.

11. Collings, V.B, (1974). Human Taste Response as a Function of Locus of Stimulation on the Tongue and Soft Palate. Perception & Psychophysics, 16: 169-174.

12. Paran, N., Mattern, C.F.T., Henkin, R.I. (1975). Ultrastructure of the taste bud of the human fungiform papilla. Cell Tissue Res., 161: 1-10.

13. Arvidson K. (1979). Location and variation in number of taste buds in human fungiform papillae. Scand J Dent Res.; 87(6):435-42.

14. Iwasaki S, Yoshizawa H, Kawahara I. (1996). Study by scanning electron microscopy of the morphogenesis of three types of lingual papilla in the mouse. Acta Anat. 157:41-52.

15. Mistretta CM, Baum BJ. (1984). Quantitative study of taste buds in fungiform and circumvallate papillae of young and aged rats. J Anat.;138:323-332.

16. Kobayashi K, Kumakura M, Yoshimura K, Shindo J. (2001). Stereo-structural study of the lingual papillae and their connective tissue cores in relation to ageing changes in the human tongue. Ital J Anat Embryol. 106(2 Suppl 1):305-11.

17. Roper SM, Kinnamon JC.J Comp Neurol. (1988). Ultrastructure of mouse foliate taste buds: synaptic and nonsynaptic interactions between taste cells and nerve fibers. J Comp Neurol. 270(1):11-24, 58-9.

108

18. Kunze K. (1969). The filiform papilla in man as a tactile sense organ. Light and electron microscopic studies. Ergeb Anat Entwicklungsgesch. 41(5):3-64.

19. Finger, T.E., and S.A. Simon. (2002). The cell biology of lingual epithelia. The Neurobiology of Taste and Smell. T.E. Finger, W.L. Silver, and D. Restrepo, editors. Wiley-Liss, New York. 287-314.

20. Travers JB, Grill HJ, Norgren R. (1987). The effects of glossopharyngeal and chorda tympani nerve cuts on the ingestion and rejection of sapid stimuli: an electromyographic analysis in the rat. Behav Brain Res. 25(3):233-46.

21. Frank, M.E. (1991). Taste-responsive neurons of the glossopharyngeal nerve of the rat. J.Neurophysiol., 65:1452-1463.

22. Murray, R.G. (1973). The ultrastructure of taste buds. In: Friedmann, I. (ed.) the Ultrastructure of Sensory Organs. North Holland, Amsterdam. 1–81.

23. Finger, T. (2005). Cell types and lineages in taste buds. Chem. Senses 30:i54 -i55.

24. Delay, R.J., Kinnamon, J.C. and Roper, S.D. (1986). Ultrastructure of mouse vallate taste buds: ii. Cell types and cell lineage. J. Comp. Neurol. 253:242-252.

25. Stone LM, Finger TE, Tam PP et al. (1995). Taste receptor cells arise from local epithelium, not neurogenic ectoderm. Proc Natl Acad Sci USA 92:1916–1920.

26. Farbman AI. (1965). Electron microscope study of the developing taste bud in rat fungiform papilla. Dev Biol. 11:110-135.

27 . Herness, S. (2000). Coding in the taste receptor cells. The early years of intracellular recordings. Physiol. Behav 69: 17-27.

109

28. Chandrashekar, J, M.A. Hoon, N.J.P. Ryba and C.S. Zuker. (2006). The receptors and cells for mammalian taste. Nature 444: 288-294.

29. Gilbertson TA, Damak S, Margolskee RF. (2000). The molecular physiology of taste transduction. Curr Opin Neurobiol. 10: 519-527.

30. Clapp, T.R., Stone, L.M., Margolskee, R.F. and Kinnamon, S.C. (2001). Immunocytochemical evidence for co-expression of Type III IP3 receptor with signaling components of bitter taste transduction. BMC Neurosci. 2, 6.

31. Miyoshi, M.A., Abe, K. and Emori, Y. (2001.) IP3 receptor type III and PLCβ2 Are co- expressed with taste receptors T1R and T2R in rat taste bud cells. Chem. Senses 26:259-265.

32. Hoon M.A., Northup J.K, Margolskee R.F., Ryba N.J. (1995). Functional expression of the taste specific G-protein, alpha-gustducin. Biochem. J. 309 (Pt 2): 629-36.

33. Ruiz-Avila L, Ming D & Margolskee RF. (2000). An in vitro assay useful to determine the potency of several bitter compounds. Chem Senses 25:361-368.

34. Chandrashekar J, Mueller KL, Hoon MA, Adler E, Feng L, Guo W, Zuker CS, Ryba NJ. (2000). T2Rs function as bitter taste receptors. Cell 100(6):703-11.

35. McLaughlin SK, McKinnon PJ, Margolskee RF. (1992). Gustducin is a taste-cell- specific G protein closely related to the transducins. Nature 357:563-569.

36. Boughter, J.D. Jr, Pumplin, D.W., Yu, C., Christy, R.C. and Smith, D.V. (1997). Differential expression of alpha-gustducin in taste bud populations of the rat and hamster. J. Neurosci. 17: 2852-2858.

110

37. Tizzano M, Dvoryanchikov G, Barrows JK, Kim S, Chaudhari N, Finger TE. (2008). Expression of Galpha14 in sweet-transducing taste cells of the posterior tongue. BMC Neuroscience. 9;110.

38. Sainz E, Korley JN, Battey JF, Sullivan SL. (2001). Identification of a novel member of the T1R family of putative taste receptors. J Neurochem.77(3):896-903.

39. Huang AL, Chen X, Hoon MA, Chandrashekar J, Guo W, Tränkner D, Ryba NJP, Zuker CS. (2006). The cells and logic for mammalian sour taste detection. Nature 442:934- 938.

40. Ishimaru Y, Inada H, Kubota M, Zhuang H, Tominaga M, Matsunami H. (2006). Transient receptor potential family members PKD1L3 and PKD2L1 form a candidate sour taste receptor. PNAS 103:12569-12574.

41. Lopez Jimenez ND, Cavenagh MM, Sainz E, Cruz-Ithier MA, Battey JF, Sullivan SL. (2006). Two members of the TRPP family of ion channels, PKD1L3 and PKD2L1 are co-expressed in a subset of taste receptor cells. J Neurochem. 98:68-77.

42. Huang Y-J, Maruyama Y, Stimac R, Roper SD. (2008). Presynaptic (Type III) cells in mouse taste buds sense sour (acid ) taste. J Physiol. 586:2903-2912.

43. Formaker BK, Hill DL. (1988). An analysis of residual NaCl taste response after amiloride. Am J Physiol; 255:R1002-R1007.

44. Formaker BK, Hill DL. (1991). Lack of amiloride sensitivity in SHR and WKY glossopharyngeal taste responses to NaCl. Physiol Behav. 50(4):765-9.

45. Chandrashekar J, Kuhn C, Oka Y, Yarmolinsky DA, Hummler E, Ryba NJ, Zuker CS. (2010). The cells and peripheral representation of sodium taste in mice. Nature 464:297-301.

111

46. Heck GL, Mierson S, DeSimone JA. (1984). Salt taste transduction occurs through an amiloride-sensitive sodium transport pathway. Science. 223(4634):403-5.

47. Mierson S, Olson MM, Tietz AE. (1996). Basolateral amiloride-sensitive Na+ transport pathway in rat tongue epithelium. J Neurophysiol. 76(2):1297-309.

48. Doolin RE, Gilbertson TA. (1996). Distribution and characterization of functional amiloride-sensitive sodium channels in rat tongue. J Gen Physiol. 107:545-554.

49. Eylam S, Spector AC. (2003). Oral amiloride treatment decreases taste sensitivity to sodium salts in C57BL/6J and DBA/2J mice. Chem Senses 28: 447-458.

50. Miyamoto T, Fujiyama R, Okada Y, and Sato T. (2000). Acid and salt responses in mouse taste cells. Prog Neurobiol. 62:135–157.

51. Halpern, B.P. (1998). Amiloride and vertebrate gustatory responses to NaCl. Neurosci. Biobehav. Rev. 23(1):5 -47.

52. McPheeters, M. and Roper, S. D. (1985). Amiloride does not block taste transduction in the mudpuppy, Necturus maculosus, Chem. Senses. 10:341.

53. Hettinger TP, Frank ME. (1990). Specificity of amiloride inhibition of hamster taste responses. Brain Res. 9;513(1):24-34.

54. Kitada, Y. and Mitoh, Y. (1998). Amiloride does not affect the taste responses of the frog glossopharyngeal nerve and submandibular branch of the facial nerve to NaCl. Chem Senses. 23:222.

55. Gilbertson TA and Fontenot DT. (1998). Distribution of amiloride-sensitive sodium channels in the oral cavity of the hamster. Chem. Senses 23: 495-499.

112

56. Bigiani A, Cuoghi V. (2007). Localization of amiloride-sensitive sodium current and voltage-gated calcium currents in rat fungiform taste cells. J Neurophysiol. 98(4):2483-7.

57. Lyall V, Heck GL, Vinnikova AK, Ghosh S, Phan TH, Alam RI, Russell OF, Malik SA, Bigbee JW, DeSimone JA. (2004). The mammalian amiloride-insensitive non-specific salt taste receptor is a vanilloid receptor-1 variant. J Physiol. 558:147-159.

58. Treesukosol Y, Lyall V, Heck GL, DeSimone JA, Spector AC. (2007). A psychophysical and electrophysiological analysis of salt taste in Trpv1 null mice. Am J Physiol Regul Integr Comp Physiol. 292(5):R1799-809.

59. Kawai K, Sugimoto K, Nakashima K, Miura H, Ninomiya Y. (2000). Leptin as a modulator of sweet taste sensitivities in mice. Proc Natl Acad Sci USA. 97: 11044- 11049.

60. Nakamura Y, Sanematsu K, Ohta R, Shirosaki S, Koyano K, Nonaka K, Shigemura N, Ninomiya Y. (2008). Diurnal variation of human sweet taste recognition thresholds is correlated with plasma leptin levels. Diabetes 57: 2661-2665.

61. Yoshida R, Okhuri T, Jyotaki M, Yasuo T, Horio N, Yasumatsu K, Sanematsu K, Shigemura N, Yanamato T, Margolskee R. F, Ninomiya Y. (2010). Endocannabinoids selectively enhance sweet taste. Proc Natl Acad Sci. 107, 935-39.

62. Gilbertson TA, Baquero AF. (2010). Insulin regulates the function of epithelial sodium channels and salt taste preference. In: Proceedings of the Meeting of the Association for Chemoreception Sciences (37th Annual Meeting), Abstr. 17. St. Petersburg, FL,.

63. Delay RJ, Kinnamon SC, Roper SD. (1997). Serotonin modulates voltage-dependent calcium current in Necturus taste cells. J Neurophysiol. 77:2515-2524.

113

64. Herness MS, Chen Y. (1997). Serotonin inhibits calcium-activated potassium and voltage-dependent sodium currents in rat taste receptor cells. J Memdr Biol. 173:127-138.

65. Imendra KG, Miyamoto T, Okada Y, Toda K. (2002). Serotonin differentially modulates the electrical properties of different subsets of taste receptor cells in bullfrog. European Journal of Neuroscience 16:629-640.

66. Caicedo, A., Jafri, M.S., and Roper, S.D. (2000). In situ Ca2+ imaging reveals neurotransmitter receptors for glutamate in taste receptor cells. J. Neurosci. 20:7978-7985.

67. Herness, M.S., and Sun, X.-D. (1999). Characterization of chloride currents and their noradrenergic modulation in rat taste receptor cells. j. Neurophysiol. 82:260-271.

68. Zhang Y, Kolli T, Hivley R, Jaber L, Zhao FI, Yan J, Herness S. (2010). Characterization of the expression pattern of adrenergic receptors in rat taste buds. Neuroscience 169(3):1421-37.

69. Ogura T. (2002). Acetylcholine increases intracellular Ca2+ in taste cells via activation of muscarinic receptors. J Neurophysiol. 87(6):2643-9.

70. Kataoka S, Toyono T, Seta Y, Ogura T & Toyoshima K. (2004). Expression of P2Y1 receptors in rat taste buds. Histochem Cell Biol. 121:419-426.

71. Cao, Y., Zhao F.L., Kolli T., Hivley R., Herness S. (2009). GABA expression in the mammalian taste bud functions as a route of inhibitory cell-to-cell communication. PNAS 106:4006-4011.

72. Herness, S. Zhao, F.L. Lu, S.G., Kaya, N. and Shen, T. (2002). Expression and physiological responses of cholecystokinin in taste receptor cells. J. Neurosci. 22: 10018-10029.

114

73. Zhao, F.L., Shen T., Kaya, N., Lu, S.G., Cao, Y., Herness, S. (2005). Expression, physiological action, and co-expression patterns of NPY in rat taste receptor cells. Proc Natl Acad Sci. 102: 11100-11105.

74. Martin B, Shin YK, White CM, Ji S, Kim W, Carlson OD, Napora JK, Chadwick W, Chapter M, Waschek JA, Mattson MP, Maudsley S, Egan JM. (2010). Vasoactive intestinal peptide-null mice demonstrate enhanced sweet taste preference, dysglycemia, and reduced taste bud leptin receptor expression. Diabetes. 59(5):1143-52.

75. Herness, M.S. (1989). Vasoactive intestinal peptide (VIP)-like immunoreactivity in rodent taste cells. Neuroscience 33: 411-419.

76. Feng XH, Liu XM, Zhou LH, Wang J, Liu GD. (2008). Expression of glucagon-like peptide -1 in the taste buds of rat circumvallate papillae. Acta Histochemic. 110:151-154.

77. Seta Y, Kataoka S, Toyono T, Toyoshima K. (2006). Expression of galanin and the galanin receptor in rat taste buds. Atch histol Cytol 69:273-280.

78. Scott Herness, (2012). The Neurobiology of Gustation: Taste Buds and Transduction Processes. In: Leonard R. Johnson et. al. editors, Physiology of the Gastrointestinal Tract. Oxford: Academic Press, 741-768.

79. Lindemann B (1999). Receptor seeks ligand: on the way to cloning the molecular receptors for sweet and bitter taste. Nat. Med. 5 (4): 381-2.

80. Herness M.S., Gilbertson T.A. (1999). Cellular mechanisms of taste transduction. Annu Rev Physiol. 61:873-900.

81. Margolskee RF (2002). Molecular mechanisms of bitter and sweet taste transduction. J. Biol. Chem. 277 (1): 1-4.

115

82. Ishimaru Y. (2009). Molecular mechanisms of taste transduction in vertebrates. Odotology 97:1-7.

83. Kroeze WK, Hufeisen SJ, Popadak BA, et al. (2003). H1-histamine receptor affinity predicts short-term weight gain for typical and atypical antipsychotic drugs. Neuropsychopharmacology 28:519-26

84. Gershon MD, Tack J.(2007). The serotonin signaling system: from basic understanding to drug development for functional GI disorders. Gastroenterology 132:397-414

85. Rapport MM, Green AA, Page IH. (1948). Serum vasoconstrictor, serotonin; isolation and characterization. J. Biol. Chem. 176:1243-51

86. Frazer, A.; and Hensler, J. G. (1999). Understanding the neuroanatomical organization of serotonergic cells in the brain provides insight into the functions of this neurotransmitter. In Siegel, G. J. Basic Neurochemistry.

87. Serotonin (5-Hydroxytryptamine; 5-HT): CNS Pathways and Neurophysiology. In Encyclopedia of Neuroscience (2009), pp. 715-722, doi:10.1016/B978-008045046- 9.01159-1 Key: citeulike: 9845905

88. Airan RD, Meltzer LA, Roy M, et al. (2007). High-speed imaging reveals neurophysiological links to behavior in an animal model of depression. Science 317:819-2

89. Canli T, Lesch KP. (2007). Long story short: the serotonin transporter in emotion regulation and social cognition. Nat. Neurosci. 10:1103-9

90. Roth BL, Hanizavareh SM, Blum AE. (2004). Serotonin receptors represent highly favorable molecular targets for cognitive enhancement in schizophrenia and other disorders. Psychopharmacology (Berl.) 174:17-24

116

91. Lesch KP. (2005). Serotonergic gene inactivation in mice: models for anxiety and aggression? Novartis Found. Symp. 268:111-40.

92. Fujimoto S, Ueda H, Kagawa H. (1987). Immunocytochemistry on the localization of 5-hyroxytryptophan in monkey and rabbit taste buds. Acta Anat (Basel) 128: 80-83.

93. Kim D-J, Roper SD. (1995). Localization of serotonin in taste buds: a comparative study in four vertebrates. J Comp Neurol. 353:364-370.

94. Nada O, Hirata K. (1975). The occurrence of the cell type containing a specific monoamine in the taste bud of the rabbit’s foliate papilla. Histochemistry 43:237- 240.

95. Ren Y, Shimada K, Shirai Y, Fujimiya M, Saito N. (1999). Immunocytochemical localization of serotonin and serotonin transporter (SET) in taste buds of rat. Molecular Brain Research 74:221-224.

96. Uchida T. (1985). Serotonin-like immunoreactivity in the taste bud of the mouse circumvallate papilla. Jpn. J. Oral Biol. 27:509-521.

97. Yee, C.L., Yang, R., Bottger, B., Finger, T.E., & Kinnamon, J.C. (2001). “Type III” cells of rat taste buds: Immunohistochemical and ultrastructural studies of neuron-specific enolase, protein gene product 9.5, and serotonin. J Comp Neurol 440:97-108.

98. Herness MS, Chen Y. (2000). Serotonergic agonists inhibit calcium-activated potassium and voltage-dependent sodium currents in rat taste receptor cells. J Membr Biol. 173:127-138.

99. Jackson MB and Yakel JL. The 5-HT3 receptor channel. (1995). Annu Rev Physiol 57: 447-468.

117

100. Wang ZY, Keith IM, Olson EB Jr, Vidruk EH, Bisgard GE. (2002). Expression of 5-HT3 receptors in primary sensory neurons of the petrosal ganglion of adult rats. Auton Neuroscience. 10;95(1-2):121-4.

101. Nagai T, Delay RJ, Welton J, Roper SD. (1998). Uptake and release of neurotransmitter candidates, [3H]serotonin, [3H]glutamate, [3H]γ-Aminobutyric acid, in taste cuds of the mudpuppy, Necturus maculosus. J Comp Neurol. 392:199-208.

102. Dvoryanchikov G, Tomchik SM, Chaudhari N. (2007). Biogenic amine synthesis and uptake in rodent taste buds. J Comp Neurol. 505:302-313.

103. Bo X, Alavi A, Xiang Z, et al. (1999). Localization of ATP-gated P2X2 and P2X3 receptor immunoreactive nerves in rat taste buds. Neuroreport. v. 10 p. 1107-1111.

104. Baryshnikov, S.G., Rogachevskaja, O.A., Kolesnikov, S.S. (2003). Calcium signaling mediated by P2Y receptors in mouse taste cells. J. Neurophysiol. 90:3283-3294.

105. Bystrova MF, Yatzenko YE, Fedorov IV, Rogachevskaja OA, Kolesnikov SS. (2006). P2Y isoforms operative in mouse taste cells. Cell Tissue Res. 323:377-382.

106. Fedorov I, Rogachevskaja OA, Kolesnikov SS. (2007). Modeling P2Y receptor-Ca2+ response coupling in taste cells. Biochim Biophys Acta 1768:1727-1740.

107. Huang Y-J, Maruyama Y, Dvoryanchikov G, Pereira E, Chaudhari N, Roper SD. (2007). The role of pannexin 1 hemichannels in ATP release and cell-cell communication in mouse taste buds. PNAS 104: 6436-6441.

108. Romanov RA, Rogachevskaja OA, Khokhlov AA, Kolesnikov SS. (2008). Voltage dependence of ATP secretion in mammalian taste cells. J Gen Physiol. Dec;132(6):731-44.

118

109. Romanov RA, Rogachevskaja OA, Bystrova MF, Jiang P, Margolskee RF, Kolesnikov SS. (2007). Afferent neurotransmission mediated by hemichannels in mammalian taste cells. EMBO Journal 26:657-667.

110. Hayato R, Ohtubo Y, Yoshii K. (2007). Functional expression of ionotropic purinergic receptors on mouse taste bud cells. J Physiol. 584:473-488.

111. Lynn-Bullock CP, Welshhans K, Pallas SL, Katz PS. (2004). The effect of oral 5-HTP administration on 5-HTP and 5-HT immunoreactivity in monoaminergic brain regions of rats. J Chem Neuroanat 27(2):129-38.

112. P.B. Gradwell , B.J. Everitt , J. Herbert . (1975). 5-Hydroxytryptamine in the central nervous system and sexual receptivity of female rhesus monkeys. Brain Research 88 (2):281-293

113. Shindler KS, Roth KA. (1996). Double immunofluorescent staining using two unconjugated primary antisera raised in the same species. J Histochem Cytochem. 44:1331-1335.

114. L. Knapp, A. Lawton, B. Oakley, L. Wong, C. Zhang. (1995). Keratins as markers of differentiated taste cells of the rat. Differentiation. 58(5):341-9.

115. Pfaffmann C. (1941). Gustatory afferent impulses. J Cell Comp Physiol. 17:243-258.

116. Beidler LM. (1953). Properties of chemoreceptors of tongue of rat. J Neurophysiol. 16:595-607.

117. Zotterman Y. (1935). Action potentials in the glossopharyngeal nerve and in the chorda tympani. Skand Arch Physiol. 72:73-75.

119

118. Akihiko Ikeda, Shonosuke Matsushita & Yuzuru Sakakibara. (2010). A simple and reliable method of endotracheal intubation in mice: advantage of exposing the trachea. Scand. J. Lab. Anim. Sci. 36 / 4: 363-268.

119. Harms, P. G. and Ojeda, S. R. (1974) A rapid and simple procedure for chronic cannulation of the rat jugular vein. J. Appl. Physiol. 36: 391-392

120. Thomsen M and Caine SB. (2005). Chronic intravenous drug self-administration in rats and mice. Curr Protoc Neurosci Chapter 9:Unit 9. 20.

121. Hill, D.L., and C.R. Almli. (1980). Ontogeny of chorda tympani nerve responses to gustatory stimuli in the rat. Brain Res. 197:27-38.

122. Beidler LM. (1971). Taste receptor stimulation with salts and acids. In: Beidler LM, editor. Handbook of sensory physiology. New York: Springer-Verlag. pp. 200-220.

123. Shingai T, Beidler LM. (1985). Response characteristics of three taste nerves in mice. Brain Res. 335(2):245-9.

124. Sato T, Beidler LM. (1983). Dependence of gustatory neural response on depolarizing and hyperpolarizing receptor potentials of taste cells in the rat. Comp Biochem Physiol A Comp Physiol. 75(2):131-7.

125. C. A. Fornal, C. W. Metzler, R. A. Gallegos, S. C. Veasey, A. C. McCreary and B. L. Jacobs. (1996). WAY-100635, a potent and selective 5-hydroxytryptamine1A antagonist, increases serotonergic neuronal activity in behaving cats: comparison with (S)-WAY-100135. J Pharm Exp Therapeutics. 278 (2): 752-762.

126. Peter A. Sargent, Karen Husted Kjaer, Christopher J. Bench, Eugenii A. Rabiner, Cristina Messa, Jeff Meyer, Roger N. Gunn, Paul M. Grasby, Philip J. Cowen. (2000). Brain serotonin1A receptor binding measured by positron emission tomography

120

with [11C]WAY-100635: effects of depression and antidepressant treatment. Archieves of General Psychiatry. 57 (2): 174-180.

127. Jerning E, Svantesson GT, Mohell N. (1998). Receptor binding characteristics of [3H]NAD-299, a new selective 5-HT1A receptor antagonist. Eur J Pharmacol. 360 (2- 3): 219–225.

128. Arborelius L, Wallsten C, Ahlenius S, Svensson TH. (1999). The 5-HT(1A) receptor antagonist robalzotan completely reverses citalopram-induced inhibition of serotonergic cell firing. Eur J Pharmacol. 382 (2): 133-138.

129. Worboys PD, Brennan B, Bradbury A, Houston JB. (1996). Metabolite kinetics of ondansetron in rat. Comparison of hepatic microsomes, isolated hepatocytes and liver slices, with in vivo disposition. Xenobiotica. Sep;26(9):897-907.

130. Costall B, Jones BJ, Kelly ME, Naylor RJ, Oakley NR, Onaivi ES, Tyers MB. (1989). The effects of ondansetron (GR38032F) in rats and mice treated subchronically with diazepam. Pharmacol Biochem Behav. Dec;34(4):769-78.

131. Kindig AE, Hayes SG, Hanna RL & Kaufman MP. (2006). P2 antagonist PPADS attenuates responses of thin fiber afferents to static contraction and tendon stretch. Am J Physiol Heart Circ Physiol. 290:H1214-H1219.

132. Seiji Watanabe, Yasuo Kono, Shinya Kaneko, Cephas P. Swamidoss, Paul G. Barash. (1999). Effects of a bolus injection of adenosine triphosphate on atrioventricular conduction and hemodynamics in patients undergoing coronary artery bypass grafting. J of Cardiothoracic and Vascular Anesthesia 13:225-227.

133. Kirkup AJ, Booth CE, Chessell IP, Humphrey PP, Grundy D. (1999). Excitatory effect of P2X receptor activation on mesenteric afferent nerves in the anaesthetised rat. J Physiol. 520:551–563, pmid:10523422.

121

134. Ziganshin, AU. (1993). PPADS selectively antagonizes P2X-purinoceptor-mediated responses in the rabbit urinary bladder. British Journal of Pharmacology 110 (4): 1491–95. PMID 8306091.

135. Lambrecht, G. (1992). PPADS, a novel functionally selective antagonist of P2- purinoreceptor mediated responses. European Journal of Pharmacology 217: 217- 19.

136. Stone LM, Barrows J, Finger TE, Kinnamon SC. (2007). Expression of T1Rs and gustducin in palatal taste buds of mice. Chem Senses. 32(3):255-62.

137. Herness, S., El Dahdah, P., Kolli, T., Cao, Y. (2007). Co-expression patterns of SNAP- 25 with neuropeptides, GAD, and NCAM suggest its expression in multiple cell types. Chemical Senses 32:A94-A95.

138. Clapp TR, Medler KF, S, Margloskee RF, Kinnamon SC. (2006). Mouse taste cells with G protein-coupled taste receptors lack voltage-gated calcium channels and SNAP-25. BMC Biology 4:7.

139. Thomas T, Marshall JM. (1993). The role of adenosine in hypoxic pulmonary vasoconstriction in the anaesthetized rat. Exp Physiol. 78(4):541-3.

140. Antunes VR, Braga VA, Machado BH. (2005). Autonomic and respiratory responses to microinjection of ATP into the intermediate or caudal nucleus tractus solitarius in the working heart-brainstem preparation of the rat. Clin Exp Pharmacol Physiol. 32(5-6):467-72.

141. Ganesh A, Bogdanowicz W, Haupt M, Marimuthu G, Rajan KE. (2010). Role of olfactory bulb serotonin in olfactory learning in the greater short-nosed fruit bat, Cynopterus sphinx (Chiroptera: Pteropodidae). Brain Res. 7:1352:108-17.

122

142. Murata Y, Yasuo T, Yoshida R, Obata K, Yanagawa Y, Margolskee RF, Ninomiya Y. (2010). Action potential-enhanced ATP release from taste cells through hemichannels. J Neurophysiol. 104(2):896-901.

143. Tatemoto K. (2004). Neuropeptide Y: History and Overview. In Michel MC. Handbook of Experimental Pharmacology. 162. Springer. pp. 2-15.

144. Lynch , DR; Walker , MW; Miller , RJ; Snyder , SH. (1989). Neuropeptide Y receptor binding sites in rat brain: differential autoradiographic localizations with 125I- peptide YY and 125I-neuropeptide Y imply receptor heterogeneity. J Neurosci. 9:2607–2619..

145. Beinfeld MC. (2001). An introduction to neuronal cholecystokinin. Peptides 22(8):1197-200

146. Della Fera MA, Baile CA. (1979). CCK-octapeptide injected in CSF causes satiety in sheep. Ann Rech Vet 10(2-3):234-6

147. Lu SG, Zhao FL, Herness S. (2003). Physiological phenotyping of cholecystokinin- responsive rat taste receptor cells. Neurosci Lett. 20;351(3):157-60.

148. Sudhof TC, Rizo J (2002). Snares and Munc18 in synaptic vesicle fusion. Nat Rev Neurosci 3 (8): 641–653.

149. Yang R, Ma H, Thomas SM, Kinnamon JC.J Comp Neurol. (2007). Immunocytochemical analysis of syntaxin-1 in rat circumvallate taste buds. J Comp Neurol. 502(6):883-93.

150. DeFazio RA, Dvoryanchikov G, Maruyama Y, Kim JW, Pereira E, Roper SD, Chaudhari N. (2006). Separate populations of receptor cells and presynaptic cells in mouse taste buds. J Neurosci. 26(15):3971-80.

123

151. Pumplin DW, Getschman E. (2000). Synaptic proteins in rat taste bud cells: appearance in the Golgi apparatus and relationship to alpha-gustducin and the Lewis(b) and A antigens. J Comp Neurol. Nov 13; 427(2):171-84.

152. Ueda K, Ichimori Y, Okada H, Honma S, Wakisaka S. (2006). Immunolocalization of SNARE proteins in both type II and type III cells of rat taste buds. Arch Histol Cytol. 69(4):289-96.

153. Oike H, Matsumoto I, Abe K. (2006). Group IIA phospholipase A2 is co-expressed with SNAP-25 in mature taste receptor cells of rat circumvallate papillae. J Comp Neurol 494:876–886.

154. Baekkeskov S, Aanstoot HJ, Christgau S, Reetz A, Solimena M, Cascalho M, Folli F, Richter-Olesen H, De Camilli P, Camilli PD (1990). Identification of the 64K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase. Nature 347 (6289): 151-6.

155. Cao Y, Zhao FL, Herness MS. (2005). GABA as an inhibitory neurotransmitter in the taste bud. Chemical Senses 30: A105-06

156. Dvoryanchikov G, Huang YA, Barro-Soria R, Chaudhari N, Roper SD. (2011). GABA, its receptors, and GABAergic inhibition in mouse taste buds. J Neurosci. 31(15):5782-91.

157. Suzuki M, Angata K, Nakayama J, Fukuda M. (2003). Polysialic acid and mucin type o-glycans on the neural cell adhesion molecule differentially regulate myoblast fusion. J Biol Chem. 278 (49): 49459-68

158. Smith DV, Akeson RA, Shipley MT. (1993). NCAM expression by subsets of taste cells is dependent upon innervation. J Comp Neurol. 336(4):493-506.

124

159. Obata H, Shimada K, Sakai N, Saito N. (1997). GABAergic neurotransmission in rat taste buds: immunocytochemical study for GABA and GABA transporter subtypes. Brain Res Mol Brain Res. 49(1-2):29-36.

160. Kusakabe Y, Kim MR, Miura H, Shindo Y, Ninomiya Y, Hino A. (2005). Regional expression patterns of T1r family in the mouse tongue. Chem Senses. Suppl 1:i23-4.

161. Margolskee, R. F.; Dyer, J.; Kokrashvili, Z.; Salmon, K. S.; Ilegems, E.; Daly, K.; Maillet, E. L.; Ninomiya, Y.; Mosinger, B. and Shirazi-Beechey, S. P. (2007). T1R3 and gustducin in gut sense sugars to regulate expression of Na+-glucose cotransporter 1. Proc Nat Ac Sc .104.38, 15075-80.

162. Zaza Kokrashvili, Bedrich Mosinger, and Robert F Margolskee. (2009). T1r3 and alpha-gustducin in gut regulate secretion of glucagon-like peptide-1Ann N Y Acad Sci. 1170:91-4.

163. Hoon, M.A., Adler, E., Lindemeier, J., Battey,J.F., Ryba, N.J. and Zuker, C.S. (1999). Putative mammalian taste receptors: a class of taste-specific gpcrs with distinct topographic selectivity. Cell, 96, 541-551.

164. Zhao Gq, Z. Y.; Zhang, Y.; Hoon, M. A.; Chandrashekar, J.; Erlenbach, I.; Ryba, N. J. P.; Zuker, C. S. (2003). The receptors for mammalian sweet and umami taste. Cell 115 (3): 255-266.

165. Nelson g, H. M. (2001). Mammalian sweet taste receptors. Cell 106 (3): 381-390.

166. Adler E, Hoon MA, Mueller KL, Chandrashekar J, Ryba NJ, Zuker CS . (2000). A novel family of mammalian taste receptors. Cell 100: 693-702.

167. Montmayeur J P and Matsunami H . (2002). Receptors for bitter and sweet taste. Curr Opin Neurobiol, 12, 1-8.

125

168. Max M, Shanker Y G, Huang L, Rong M, Liu Z, Campagne F, Weinstein H, Damak S and Margoleskee R F. (2001). TaslRS, encoding a new candidate taste receptor, is allelic to the sweet responsiveness locus Sac. Nat Genet, 28, 58-63.

169. Ohmoto M, Matsumoto I, Yasuoka A, Yoshihara Y, Abe K. (2008). Genetic tracing of the gustatory and trigeminal neural pathways originating from T1R3-expressing taste receptor cells and solitary chemoreceptor cells. Mol Cell Neurosci.38(4):505-17.

170. Simon SA, Liu L, Erickson RP. (2003). Neuropeptides modulate rat chorda tympani responses. Am J Physiol Regul Integr Comp Physiol. 284(6):R1494-505.

171. Iwayama T, Nada O. (1967). Histochemically demonstrable ATPase activity in the taste buds of the rat. Exp Cell Res. Jun;46(3):607-8.

172. Zalewski AA. (1968). Changes in phosphatase enzymes following denervation of the vallate papilla of the rat. Exp Neurol. 22(1):40-51.

173. Akisaka T, Oda M. (1977). The fine structural localization of adenosine triphosphatase activity on the taste bud in the fungiform papillae of the rat. Arch Histol Jpn. 40(1):63-72.

174. Plesner L. (1995). Ecto-ATPases: identities and functions. Int Rev Cytol.158:141-214.

175. Dianna L. Bartel, Susan L. Sullivan, Élise G. Lavoie, Jean Sévigny, and Thomas E. Finger. (2006). Nucleoside Triphosphate Diphosphohydrolase-2 (NTPDase2) is the ecto-ATPase of Type I cells in taste buds. J Comp Neurol. 497(1): 1-12.

176. Kukulski F, Levesque SA, Lavoie EG, Lecka J, Bigonnesse F, Knowles AF, Robson SC, Kirley TL, Sévigny J. (2005). Comparative hydrolysis of P2 receptor agonists by NTPDases 1, 2, 3 and 8. Purinergic Signalling. 1:193-204.

126

177. Masaki Ishikawa, Tsutomu Iwamoto, Takashi Nakamura, Andrew Doyle, Satoshi Fukumoto, and Yoshihiko Yamada. (2011). Pannexin 3 functions as an ER Ca2+ channel, hemichannel, and gap junction to promote osteoblast differentiation. JCB 193:1257-1274

2+ 178. Liu D & Liman ER. (2003). Intracellular Ca and the phospholipid PIP2 regulate the taste transduction ion channel TRPM5. Proc Natl Acad Sci. 100:15160-15165.

179. Suh B-C & Hille B. (2008). PIP2 is a necessary cofactor for ion channel function: How and why? Annu Rev Biophys 37, 175-195.

180. J.C. Saez, V.M. Berthoud, M.C. Branes, A.D. Martinez, E.C. Beyer. (2003), Plasma membrane channels formed by connexins: their regulation and functions. Physiol. Rev. 83:1359-1400

181. Cameron SJ, Malik S, Akaike M, Lerner-Marmarosh N, Yan C, Lee JD, Abe J, Yang J. (2003). Regulation of epidermal growth factor-induced connexin 43 gap junction communication by big mitogen-activated protein kinase1/ERK5 but not ERK1/2 kinase activation. J. Biol. Chem. 278(20):18682-8.

182. Suh B-C & Hille B (2005). Regulation of ion channels by phosphatidylinositol 4,5- bisphosphate. Curr Opin Neurobiol. 15:307-378.

183. Hilgemann DW, Feng S & Nasuhoglu C (2001). The complex and intriguing lives of PIP2 with ion channels and transporters. Science STKE, RE19.

184. Fang-li Zhao and Scott Herness. (2009). Resynthesis of phosphatidylinositol 4,5- bisphosphate mediates adaptation of the caffeine response in rat taste receptor cells J Physiol 587.2:363-377 363.

127

185. Ni YG, Panicker MM, Miledi R. (1997). Efficient coupling of 5-HT1A receptors to the phospholipase C pathway in Xenopus oocytes. Brain Res Mol Brain Res. 51(1-2):115- 22.

186. Raymond JR, Mukhin YV, Gettys TW, Garnovskaya MN. (1999). The recombinant 5- HT1A receptor: G protein coupling and signalling pathways. Br J Pharmacol. Aug;127(8):1751-64.

187. Minisclou C, Benavides J, Claustre Y. (1995). Activation of 5-HT1A receptors inhibits carbachol-stimulated inositol 1,4,5-trisphosphate mass accumulation in the rodent hippocampus. Neurochem Res. 20(9):977-83.

188. Rettinger J, Schmalzing G. (2004). Desensitization Masks Nanomolar Potency of ATP for the P2X1 Receptor J. Biol. Chem. 279: 6426-6433.

189. He ML, Koshimizu TA, Tomić M, Stojilkovic SS. (2002). Purinergic P2X2 receptor desensitization depends on coupling between ectodomain and c-terminal domain. Mol. Pharmacol. 62 (5) 1187-1197

190. Cook SP, McCleskey EW. (1997). Desensitization, recovery and Ca(2+)-dependent modulation of ATP-gated P2X receptors in nociceptors. Neuropharmacology. 36(9):1303-8.

191. Chen K, Yan J, Suo Y, Li J, Wang Q, Lv B. (2010). Nutritional status alters saccharin intake and sweet receptor mRNA expression in rat taste buds. Brain Res. 14;1325:53-62.

192. Zhang Y, Hoon MA, Chandrashekar J, Mueller KL, Cook B, Wu D, Zuker CS, Ryba NJ. (2003). Coding sweet, bitter, and umami tastes: different receptor cells sharing signaling pathways. Cell. 112 (3): 293-301.

128

193. M. Kishi, Y. Emori, Y. Tsukamoto, K. Abe. (2001). Primary culture of rat taste bud cells that retain molecular markers for taste buds and permit functional expression of foreign genes. Neuroscience, 106: 217-225.

194. Uchida S, Hara K, Kobayashi A, Funato H, Hobara T, et al. (2010). Early life stress enhances behavioral vulnerability to stress through the activation of REST4- mediated gene transcription in the medial prefrontal cortex of rodents. J Neurosci 30: 15007-15018.

195. Borroto-Escuela DO et al. (2012). The existence of FGFR1-5-HT1A receptor heterocomplexes in midbrain 5-HT neurons of the rat: relevance for neuroplasticity. J Neurosci 32:6295-303.

196. Florea L, Di Francesco V, Miller J, Turner R, Yao A, Harris M, Walenz B, Mobarry C, Merkulov GV, Charlab R, Dew I, Deng Z, Istrail S, Li P, Sutton G. (2005).Gene and alternative splicing annotation with AIR. Genome Res. 15(1):54-66.

197. Kakinohana O, Hefferan MP, Miyanohara A, Nejime T, Marsala S, Juhas S, Juhasova J, Motlik J, Kucharova K, Strnadel J, Platoshyn O, Lazar P, Galik J, Vinay L, Marsala M. (2012). Combinational spinal GAD65 gene delivery and systemic GABA-mimetic treatment for modulation of spasticity. PLoS One. 7(1):e30561.

129