UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Peripheral arterial chemoreceptors and their role in cardio-respiratory control Reyes, Catalina 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_fall_reyes_catalina.pdf [ 12.35MB ]
Metadata
JSON: 24-1.0073886.json
JSON-LD: 24-1.0073886-ld.json
RDF/XML (Pretty): 24-1.0073886-rdf.xml
RDF/JSON: 24-1.0073886-rdf.json
Turtle: 24-1.0073886-turtle.txt
N-Triples: 24-1.0073886-rdf-ntriples.txt
Original Record: 24-1.0073886-source.json
Full Text
24-1.0073886-fulltext.txt
Citation
24-1.0073886.ris

Full Text

PERIPHERAL ARTERIAL CHEMORECEPTORS AND THEIR ROLE IN CARDIO-RESPIRATORY CONTROL  by Catalina Reyes  BSc., Universidad de los Andes, 2000 MSc., University of British Columbia, 2006  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (ZOOLOGY)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) June 2013  © Catalina Reyes, 2013  Abstract Peripheral arterial chemoreceptors show anatomical and functional similarities and differences among vertebrate groups. Fishes have widely distributed neuroepithelial cells containing serotonin in all gill arches and extrabranchial sites, while mammals have clustered glomus cells in the carotid bifurcation and aortic arch that contain a number of neurotransmitters. However, we do not know how peripheral arterial chemoreceptors of amphibians and reptiles compare to other groups. My thesis established the location, distribution, neurochemical content, reflex roles and plasticity of peripheral arterial chemoreceptors in representative amphibians and reptiles (snakes (Crotalus durissus); turtles (Trachemys scripta elegans) and frogs (Rana catesbeiana)). I found functional chemosensory areas in the carotid bifurcation, aorta and pulmonary artery of rattlesnakes, the same locations where peripheral chemoreceptors are found in amphibians, turtles and tortoises. I used immunohistochemistry and tract tracing to identify putative O2-sensing cells in snakes, turtles and frogs, and determined their neurochemical content and anatomical relation to branches of the glossopharyngeal and vagus nerves. Although the structure and innervation pattern of these cells is clearly maintained among vertebrate taxa, the types of neurochemicals involved in oxygen chemotransduction seem to have increased in number progressing throughout the vertebrate taxa. While serotonin is found in all vertebrates, the presence of other neurotransmitters varied among species. Catecholamines were found in the chemosensory areas of amphibians, while turtles and snakes contained acetylcholine. While the serotonergic and catecholamine containing cells were organized singly or in small clusters in amphibians and reptiles, cholinergic cells in reptiles were always arranged in large clusters. Unlike mammals, anatomically distinct chemoreceptor groups in snakes did not differ in their reflex response. All chemosensory areas regulated the respiratory and cardiovascular systems, the latter through adjustments in heart rate and the cardiac shunt. Furthermore, changes in the breathing pattern of turtles resulted from daily changes in the sensitivity of chemoreceptors independent of metabolism, indicating that biological rhythms play a role in respiratory control in reptiles. My findings suggest that while O2-sensing structures are essential among vertebrate groups, considerable plasticity exists for the specifics of location and neurochemicals, which is likely related to differing needs to match oxygen supply and demand.  ii  Preface A version of chapter 5 has been published as: “Reyes, C. and W.K. Milsom. 2009. Daily and seasonal rhythms in the respiratory sensitivity of red-eared sliders (Trachemys scripta elegans). Journal of Experimental Biology 212: 3339-3348”. Dr. W.K. Milsom and I conceived of this project. I did all of the laboratory work, performed all analyses and wrote the original manuscript. Dr. W.K. Milsom provided advice on analyses and contributed revisions to this manuscript. In chapter 2 Dr. Brink D. did the tissue preparation and imaging for frog whole mounts (Figures 2.10 and 2.13). I conducted all of the tissue preparation, processing and imaging for tissue sections and performed the analysis and quantification in both tissue sections and whole mounts. In chapter 2 and 3, hematoxylin and eosin stained cross sections were prepared by Wax-it Histology services, Vancouver, BC. All experiments in this dissertation were approved by the UBC animal care committee (A09-0233 and A04-1006).  iii  Table of contents Table of Contents Abstract ......................................................................................................................................... ii Preface .......................................................................................................................................... iii Table of Contents ......................................................................................................................... iv List of Tables ................................................................................................................................ ix List of Figures ................................................................................................................................x List of Abbreviations ................................................................................................................. xvi Glossary ........................................................................................................................................xx Acknowledgements .................................................................................................................... xxi Dedication ................................................................................................................................. xxiii Chapter 1: Introduction .......................................................................................................... 1 1.1 Phylogeny of peripheral arterial chemoreceptors ................................................. 1 1.1.1 Location, innervation and distribution of putative O2 sensing cells ................... 1 1.1.1.1 Fish ................................................................................................................. 2 1.1.1.2 Amphibians ..................................................................................................... 2 1.1.1.3 Turtles ............................................................................................................. 3 1.1.1.4 Lizards ............................................................................................................ 4 1.1.1.5 Snakes ............................................................................................................. 5 1.1.1.6 Crocodilians .................................................................................................... 5 1.1.1.7 Birds ................................................................................................................ 5 1.1.1.8 Mammals: ....................................................................................................... 6 1.2 Characterization of oxygen sensing cells at arterial chemosensitive sites ........... 6 1.2.1 Fish ..................................................................................................................... 7 1.2.2 Amphibians ......................................................................................................... 7 1.2.3 Turtles ................................................................................................................. 9 1.2.4 Lizards ................................................................................................................ 9 1.2.5 Snakes and crocodilians ...................................................................................... 9 1.2.6 Birds .................................................................................................................... 9 1.2.7 Mammals .......................................................................................................... 10 1.3 Oxygen sensing mechanisms .................................................................................. 11 1.4 Overview of peripheral chemoreceptors in vertebrates ...................................... 12 1.5 Significance of locating and characterizing putative oxygen sensing cells in amphibians and reptiles ..................................................................................................... 13 1.6 Cardiovascular anatomy of amphibians and reptiles and presence of cardiac shunts .................................................................................................................................. 14 1.6.1 Amphibians ....................................................................................................... 14 1.6.2 Reptiles ............................................................................................................. 14 1.7 Reflex roles and stimulus specificity of peripheral chemoreceptors in amphibians and reptiles ..................................................................................................... 15 1.7.1 Reflex roles of peripheral arterial chemosensitive areas .................................. 15 1.7.2 Stimulus specificity of peripheral arterial chemoreceptors .............................. 16 1.7.3 Significance of establishing the reflex roles and stimulus modalities of chemosensory areas in amphibians and reptiles ............................................................... 17 1.8 Plasticity of chemoreflexes ..................................................................................... 17 1.9 Model species ........................................................................................................... 18 1.10 Research questions and objectives of the thesis ................................................... 20 iv  Chapter 2: Distribution and innervation of peripheral arterial chemoreceptors in the bullfrog (Rana catesbeiana).................................................................................................... 30 2.1 Summary ................................................................................................................. 30 2.2 Introduction ............................................................................................................ 30 2.3 Materials and methods ........................................................................................... 33 2.3.1 Animals and holding conditions ....................................................................... 33 2.3.2 Vascular and nerve anatomy............................................................................. 33 2.3.3 Tissue sections .................................................................................................. 34 2.3.4 Surgery and tissue preparation.......................................................................... 34 2.3.5 Immunohistochemistry ..................................................................................... 35 2.3.6 Controls ............................................................................................................ 36 2.3.7 Microscopy for cryo-sectioned tissue ............................................................... 36 2.3.8 Whole mount tissue preparation ....................................................................... 37 2.3.8.1 Immunohistochemistry for whole mounts .................................................... 37 2.3.8.2 Microscopy for whole mounts ...................................................................... 38 2.3.9 Quantification ................................................................................................... 38 2.3.9.1 Cell size ........................................................................................................ 38 2.3.9.2 Cell distribution ............................................................................................ 39 2.3.9.2.1 Calculations of cell density in the carotid labyrinth ............................... 39 2.3.9.2.2 Calculations of cell density in the aorta and pulmocutaneous artery...... 39 2.3.10 Haematoxylin and eosin histology.................................................................... 39 2.3.11 Scanning electron microscopy .......................................................................... 40 2.4 Results ...................................................................................................................... 40 2.4.1 Anatomy and innervations of the central vasculature....................................... 40 2.4.2 Controls for immunohistochemistry ................................................................. 41 2.4.3 Potential chemosensory areas ........................................................................... 41 2.4.3.1 Carotid labyrinth ........................................................................................... 42 2.4.3.1.1 Neurochemical content of the carotid labyrinth ...................................... 42 2.4.3.1.2 Cell density in the carotid labyrinth ........................................................ 42 2.4.3.1.3 Innvervation of the carotid labyrinth ...................................................... 43 2.4.3.2 Aorta ............................................................................................................. 43 2.4.3.3 Pulmocutaneous artery.................................................................................. 43 2.5 Discussion ................................................................................................................ 44 2.5.1 Putative neurotransmitters in oxygen sensing cells .......................................... 44 2.5.1.1 Putative oxygen sensing cells in the carotid labyrinth of frogs .................... 46 2.5.1.2 Putative oxygen sensing cells in the aorta and pulmocutaneous artery of frogs ...................................................................................................................... 48 2.5.2 Distribution of putative oxygen sensing cells ................................................... 49 2.6 Conclusion ............................................................................................................... 50 Chapter 3: Distribution and innervation of peripheral arterial chemoreceptors in the red-eared slider (Pseudemys scripta elegans)........................................................................ 66 3.1 Summary ................................................................................................................. 66 3.2 Introduction ............................................................................................................ 66 3.3 Materials and methods ........................................................................................... 69 3.3.1 Animals and holding conditions ....................................................................... 69 v  3.3.2 Vascular and nerve anatomy............................................................................. 70 3.3.3 Tissue sections .................................................................................................. 70 3.3.4 Surgery and tissue preparation.......................................................................... 70 3.3.5 Immunohistochemistry ..................................................................................... 72 3.3.6 Controls ............................................................................................................ 72 3.3.7 Microscopy for cryo-sectioned tissue ............................................................... 73 3.3.8 Quantification ................................................................................................... 73 3.3.8.1 Cell size ........................................................................................................ 73 3.3.9 Haematoxylin and eosin histology.................................................................... 73 3.4 Results ...................................................................................................................... 74 3.4.1 Anatomy and innervations of the central vasculature....................................... 74 3.4.2 Controls for immunohistochemistry ................................................................. 75 3.4.3 Neurochemical content and innervation of chemosensory areas ...................... 75 3.4.3.1 Neurochemical content of the carotid artery ................................................ 76 3.4.3.2 Neurochemical content of the aorta .............................................................. 76 3.4.3.3 Neurochemical content of the pulmonary artery .......................................... 77 3.4.3.4 Neurochemical content of the truncus arteriosus.......................................... 77 3.4.4 Innervation of chemosensory areas and putative oxygen sensing cells ............ 78 3.4.5 Neural crest origin of putative oxygen sensing cells ........................................ 78 3.5 Discussion ................................................................................................................ 78 3.5.1 Innervation of peripheral chemosensory areas ................................................. 79 3.5.2 Putative oxygen sensing cells in the chemosensory areas of turtles ................. 80 3.5.2.1 Distribution of cells containing vesicular acetylcholine transporter in the chemosensory areas of turtles ....................................................................................... 80 3.5.2.1.1 Large polygonal VAChT-IR cells ........................................................... 80 3.5.2.1.2 Populations of AChT-IR cell clusters ..................................................... 81 3.5.2.2 Distribution of serotonergic cells in the chemosensory areas of turtles ....... 82 3.5.3 Absence of catecholamines in the peripheral arterial chemoreceptors of turtles . .......................................................................................................................... 84 3.5.4 Origin of putative oxygen sensing cells............................................................ 84 3.6 Conclusion ............................................................................................................... 85 Chapter 4: Distribution and role of peripheral arterial chemoreceptors in cardiorespiratory control of the South American rattlesnake (Crotalus durissus) ................... 101 4.1 Summary ............................................................................................................... 101 4.2 Introduction .......................................................................................................... 101 4.3 Materials and methods ......................................................................................... 104 4.3.1 Animals and holding conditions ..................................................................... 104 4.3.2 Surgery and instrumentation ........................................................................... 104 4.3.3 Denervation of carotid or pulmonary and aortic branches of the glossopharyngeal/vagus nerve trunk............................................................................... 105 4.3.4 Experimental protocol .................................................................................... 106 4.3.5 Data analysis and statistics ............................................................................. 106 4.3.6 Tissue preparation ........................................................................................... 107 4.3.7 Immunohistochemistry ................................................................................... 108 4.3.8 Controls .......................................................................................................... 108 4.3.9 Microscopy for cryo-sectioned tissue ............................................................. 109 4.3.10 Quantification: cell size .................................................................................. 109 vi  4.4 Results .................................................................................................................... 109 4.4.1 Anatomy and innervation of the central vasculature ...................................... 109 4.4.2 Chemosensory areas in rattlesnakes ............................................................... 110 4.4.2.1 Effects of chemoreceptor stimulation on cardiovascular control ............... 111 4.4.2.1.1 Cardiovascular control by peripheral chemoreceptors in the carotid bifurcation ............................................................................................................... 111 4.4.2.1.2 Cardiovascular control by peripheral chemoreceptors in the aorta....... 111 4.4.2.1.3 Cardiovascular control by chemoreceptors in the pulmonary artery .... 112 4.4.2.2 Effects of chemoreceptor stimulation on respiratory control ..................... 112 4.4.3 Stimulus specificity of peripheral chemoreceptors......................................... 112 4.4.4 Neurochemical content of the chemosensory areas ........................................ 113 4.4.4.1 Controls for immunohistochemistry ........................................................... 113 4.4.4.2 Neurochemical content of the carotid bifurcation ...................................... 113 4.4.4.3 Neurochemical content of the aorta ............................................................ 113 4.4.4.4 Neurochemical content of the pulmonary artery ........................................ 114 4.4.5 Neural crest origin of putative oxygen sensing cells ...................................... 114 4.5 Discussion .............................................................................................................. 114 4.5.1 Limitations of the study .................................................................................. 115 4.5.2 Comparison with other studies ....................................................................... 115 4.5.3 Location of peripheral chemoreceptor in rattlesnakes .................................... 116 4.5.4 Regulatory roles and stimulus specificity of chemoreceptors groups ............ 117 4.5.4.1 Effects of chemoreceptor stimulation on respiratory control ..................... 117 4.5.4.2 Effects of chemoreceptor stimulation on cardiovascular control ............... 118 4.5.4.2.1 Mechanisms of cardiac shunt regulation ............................................... 118 4.5.4.3 Response time of chemoreceptor stimulation ............................................. 119 4.5.4.4 Reflex roles of peripheral chemoreceptors ................................................. 120 4.5.5 Distribution and neurochemical content of putative oxygen sensing cells..... 121 4.6 Conclusion ............................................................................................................. 123 Chapter 5: Daily and seasonal rhythms in the respiratory sensitivity of red-eared sliders (Trachemys scripta elegans) .................................................................................................. 139 5.1 Summary ............................................................................................................... 139 5.2 Introduction .......................................................................................................... 139 5.3 Materials and methods ......................................................................................... 140 5.3.1 Measurement of metabolic rate and ventilation.............................................. 141 5.3.2 Data analyses .................................................................................................. 142 5.3.3 Calculation of the air convection requirement and chemosensitivity............. 142 5.3.4 Statistical analyses .......................................................................................... 143 5.4 Results .................................................................................................................... 144 5.4.1 Daily rhythms in the metabolism and breathing of turtles during exposure to hypoxic-hypercapnia....................................................................................................... 144 5.4.2 Daily rhythms in chemosensitivity ................................................................. 144 5.4.3 Seasonal rhythms in the metabolism and breathing of turtles during exposure to hypoxic-hypercapnia....................................................................................................... 145 5.4.4 Seasonal rhythms in chemosensitivity ............................................................ 146 5.5 Discussion .............................................................................................................. 147 vii  5.5.1 Daily rhythms in ventilatory sensitivity ......................................................... 147 5.5.2 Seasonal rhythms in ventilatory sensitivity .................................................... 149 5.6 Conclusion ............................................................................................................. 151 Chapter 6: General discussion and conclusions................................................................ 163 6.1 Importance of studying peripheral chemoreceptors in amphibians and reptiles . ................................................................................................................................ 163 6.2 Major findings and implications ......................................................................... 164 6.2.1 Location of chemosensory areas in vertebrates .............................................. 165 6.2.2 Anatomical features and neurochemical content of O2-sensing cells ............. 166 6.2.2.1 Anatomical features of putative oxygen sensing cells ................................ 167 6.2.2.2 Neurochemical content of putative oxygen sensing cells ........................... 167 6.2.3 Organization and innervation of putative O2-sensing cells ............................ 170 6.2.3.1 Organization ............................................................................................... 170 6.2.3.2 Innervation .................................................................................................. 171 6.2.4 Distribution of putative O2-sensing cells related to stimulus specificity and reflex roles ...................................................................................................................... 172 6.2.4.1 O2-content sensitive chemoreceptors responsible for cardiovascular control (hypothesis 1).............................................................................................................. 173 6.2.4.2 Cardiovascular regulation by PO2- sensitive chemoreceptors in the pulmocutaneous or pulmonary artery (hypothesis 2) ................................................. 174 6.2.4.3 Presence of two sets of chemoreceptors within the pulmonary artery (hypothesis 3).............................................................................................................. 174 6.3 Caveats ................................................................................................................... 176 6.4 Future research ..................................................................................................... 176 6.5 Concluding remarks ............................................................................................. 177 References ...................................................................................................................................181 Appendices .................................................................................................................................198  viii  List of Tables Table 1.1 Summary of the location, innervation, neurochemical content, distribution and organization of chemosensory areas in vertebrates ...................................................................... 29  Table 2.1 Primary and secondary antibodies used for immunohistochemistry on tissue sections ...................................................................................................................................................... 65  Table 3.1 Primary and secondary antibodies used for immunohistochemistry on tissue sections .................................................................................................................................................... 100  Table 4.1 Primary and secondary antibodies used for immunohistochemistry. ........................ 136  Table 4.2 Effects of NaCN on stroke volume (VStot), mean arterial pressure (MAP) and systemic pulse (Psys).................................................................................................................... 137 Table 4.3 Effects of saline injections (2 ml) on pulmonary blood flow (Q̇ pul), systemic blood flow (Q̇ sys), heart rate (fH), mean arterial pressure (MAP), total ventilation (VTot), breathing frequency (fR) and amplitude (VAMP). ......................................................................................... 138 Table 5.1 Temperature and photoperiod used in the experimental series. ................................. 159  Table 5.2 Mean ± s.e.m. temperature coefficients (Q10). ........................................................... 160 Table 5.3 Day and night values (± s.e.m.) of tidal volume (ml/kg), breathing frequency (breaths/min), frequency of breathing episodes (episodes/hour) and the percent of time spent in apnea for turtles breathing a hypoxic-hypercapnic gas (H-H) and air........................................ 161  Table 5.4 Mean ± s.e.m. tidal volume (ml/kg), breathing frequency (breaths/min), frequency of breathing episodes (episodes/hour), breaths per episode, instantaneous breathing frequency (breaths/min in an episode) and percent time spent in apnea of turtles breathing a hypoxichypercapnic gas (H-H) and air.................................................................................................... 162  ix  List of Figures Figure 1.1. Distribution of O2-sensing chemoreceptors in vertebrates ....................................... 24 Figure 1.2. Schematic explaining the pressure and washout shunt hypotheses for R-L cardiac shunts in reptiles ........................................................................................................................... 25  Figure 1.3. Change in arterial oxygen tension as a result of increasing ventilation, eliminating the R-L shunt or a combination of increased ventilation and elimination of R-L shunt (Wang and Hicks, 1996a). ............................................................................................................................... 26 Figure 1.4. Pulmonary (Q̇ pul) and systemic (Q̇ sys) blood flows in turtles with varying hematocrit in normoxia, during apneas (NVP) and during ventilation (VP) (Wang et al., 1997). ................. 27  Figure 1.5. Mean (± s.e.m.) day (unfilled bars) and night (filled bars) oxygen consumption (A) and total ventilation (B) for outdoor turtles during different seasons measured at mean seasonal temperatures but natural photocycle (summer: 20.8ºC, 16L : 8D; fall: 14.7ºC, 10L : 14D; winter: 9ºC, 9L : 15D; spring: 14.6ºC, 14L : 10D). Comparison of day and night values of oxygen consumption (C) and ventilation (D) for outdoor turtles exposed to the natural photoperiod (photocycle values are the same as the fall values shown in A and B) and to constant darkness (constant dark and no daily thermal cycle) in the fall season ....................................................... 28  Figure 2.1. Schematics showing the orientation of longitudinal tissue sections and whole mounts (A and B). Location in the central vasculature of Rana catesbeiana where cross sections for histological analysis were made (C) ............................................................................................. 52  Figure 2.2. Picture (A) and schematic (B) showing the anatomy and innervation of the central vasculature of Rana catesbeiana .................................................................................................. 53  Figure 2.3. Positive controls for vesicular acetylcholine transporter (VAChT) and the neuronal tracer Cholera toxin B (CTB) in the Jugular ganglia of Rana catesbeiana .................................. 54  Figure 2.4. Carotid labyrinth of Rana catesbeiana ...................................................................... 55 x  Figure 2.5. Cell populations present in the carotid labyrinth or R. catesbeiana .......................... 56  Figure 2.6. Triple immunolabeling for 5HT (red), human natural killer-1 (HNK-1, green) and a nuclear stain (DAPI, blue) in the carotid labyrinth (A), aorta (B) and pulmocutaneous artery (C) of Rana catesbeiana ..................................................................................................................... 57  Figure 2.7. Cell density as a function of distance of the section from the outer edge of the carotid labyrinth (0 µm, insert) for (A) TH-IR cells, (B) 5HT-IR cells, (C) large TH-IR cells and (D) 5HT and TH-IR cells .................................................................................................................... 58  Figure 2.8. Innervation of the carotid labyrinth by the glossopharyngeal (IX) and vagus (X) nerves, shown by the presence of nerve fibers containing the neuronal tracer CTB.................... 59  Figure 2.9. Cell populations and vagal innervation (X) in the aorta of R. catesbeiana.............. 60  Figure 2.10. Distribution of serotonergic cells in the aorta (whole mount) of R. catesbeiana .... 61  Figure 2.11. Distribution of serotonergic cells in the aorta and pulmocutaneous artery of R. catesbeiana ................................................................................................................................... 62  Figure 2.12. Cell populations and vagal innervation (X) in the pulmocutaneous of R. catesbeiana ................................................................................................................................... 63  Figure 2.13. Serotonergic cells in the pulmocutaneous artery constriction (whole mount) of R. catesbeiana ................................................................................................................................... 64  Figure 3.1. (A) Schematic diagram of the major central arteries arising from the heart of the turtle. (B) picture and (C) schematic diagram showing the anatomy and innervation by the vagus nerve (arrows) of the common carotid artery of Trachemys scripta elegans ............................... 86  Figure 3.2. Pictures and schematic diagrams showing the anatomy and innervation by the vagus nerve of the aorta (A, and D) and pulmonary artery (C and F) of Trachemys scripta elegans .... 87 xi  Figure 3.3. Positive control for catecholamines in the adrenal glands of Trachemys scripta elegans .......................................................................................................................................... 88  Figure 3.4. Images to illustrate the level at which longitudinal sections were taken .................. 89  Figure 3.5. Cells containing vesicular acetylcholine transporter in the common carotid artery of Trachemys scripta elegans ........................................................................................................... 90  Figure 3.6. Serotonergic cells in the common carotid artery of Trachemys scripta elegans....... 92  Figure 3.7. Putative oxygen sensing cells in the aorta of Trachemys scripta elegans................. 94  Figure 3.8. Series of polygonal VAChT-IR cells in the aorta of Trachemys scripta elegans ..... 96  Figure 3.9. Pulmonary artery of Trachemys scripta elegans ....................................................... 97  Figure 3.10. Putative oxygen sensing cells in the pulmonary artery of Trachemys scripta elegans ...................................................................................................................................................... 98  Figure 4.1. Schematic (A) and pictures (B-F) showing the anatomy (C) and innervation by the glossopharyngeal/vagus nerve (IX/X nerve, arrows) of the carotid bifurcation (B and D) of Crotalus durissus. (E) the IX/X nerve running caudally by the carotid artery towards the aorta, pulmonary artery and heart (F) ................................................................................................... 124  Figure 4.2. Changes in systemic (A) and pulmonary (B) blood flows and heart rate (C) in Crotalus durissus following injections of NaCN in the carotid bifurcation. .............................. 125 Figure 4.3. Changes in shunt fraction (Q̇ p/Q̇ s; A, D and G), cardiac output (B, E and H) and systemic resistance (C, F and I) following injections of NaCN in the carotid bifurcation, aorta and pulmonary artery of Crotalus durissus ................................................................................ 126  xii  Figure 4.4. Original traces showing the effects of NaCN injections (arrows) in the carotid bifurcation (A and B); aorta (C and D) and the pulmonary artery (E and F) on haemodynamic variables of intact and denervated Crotalus durissus ................................................................. 127  Figure 4.5. Changes in systemic (A) and pulmonary (B) blood flows and heart rate (C) in Crotalus durissus following injections of NaCN in the aorta .................................................... 128  Figure 4.6. Changes in systemic (A) and pulmonary (B) blood flows and heart rate (C) in Crotalus durissus following injections of NaCN in the pulmonary artery................................. 129  Figure 4.7. Relative changes in the total ventilation of Crotalus durissus following injections of NaCN in the carotid bifurcation (A), aorta (B) and pulmonary artery (C) ................................. 130  Figure 4.8. Relative changes in amplitude and breathing frequency in Crotalus durissus following injections of NaCN in the carotid bifurcation (A and B), aorta (C and D) and pulmonary artery (E and F)......................................................................................................... 131  Figure 4.9. Original traces showing the effects of NaCN injections (arrow) in the carotid bifurcation (A); aorta (B) and the pulmonary artery (C) on the ventilation and breathing pattern of intact and denervated Crotalus durissus ................................................................................ 132  Figure 4.10. Putative oxygen sensing cells in a segment of the left common carotid and the carotid bifurcation of Crotalus durissus ..................................................................................... 133  Figure 4.11. Putative oxygen sensing cells in the aorta of Crotalus durissus ........................... 134  Figure 4.12. Putative oxygen sensing cells in the pulmonary artery of Crotalus durissus ....... 135  Figure 5.1. Day and night values of resting ventilation, oxygen consumption and air convection requirement (ACR) ± s.e.m of turtles breathing a hypoxic-hypercapnic gas (A-C) and air (D-F) (modified from Reyes and Milsom, 2010) in different seasons (N=6 for summer and spring, N=8  xiii  for winter and fall) measured at mean seasonal temperatures and natural photocycle (summer: 20.8°C, 16L:8D; fall: 14.7°C, 10L:14D; winter: 9°C, 9L:15D; spring: 14.6°C, 14L:10D) ....... 153  Figure 5.2. Mean daytime (open bars) and nighttime (filled bars) changes in oxygen consumption (A), ventilation (B) and air convection requirement (∆ACR, the hyperventilatory response, C) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas .......... 154  Figure 5.3. Comparison of day (open bars) and night (filled bars) changes in ventilation (A) and ACR (B) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas under the fall photocycle and constant dark...................................................................................................... 155  Figure 5.4. Ventilation oxygen consumption and air convection requirement (ACR) of turtles breathing a hypoxic-hypercapnic gas (H-H) and air (modified from Reyes and Milsom, 2010) in different seasons. Values recorded at seasonal temperatures are given as well as summer and winter values corrected to 14.7°C............................................................................................... 156  Figure 5.5. Changes in oxygen consumption (A), ventilation (B) and air convection requirement (ACR) (C) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas in different seasons ........................................................................................................................................ 157  Figure 5.6. Temperature-corrected (14.7°C) changes in oxygen consumption (A), ventilation (B) and air convection requirement (ACR) (C) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas in different seasons ............................................................................ 158  Figure 6.1. Diagram of the central vasculature of frogs, turtles and snakes showing the location and distribution of putative O2-sensing cells .............................................................................. 180 Figure A1. Positive controls for injections of the neuronal tracer Cholera toxin B (CTB) in the vagus nerve, serotonin (5HT) and catecholamines (tyrosine hydroxylase, TH)……………….198  xiv  Figure B1. Positive controls for vesicular acetylcholine transporter (VAChT), serotonin (5HT) and the neuronal tracer cholera toxin B (CTB) in the jugular ganglia and lungs of Trachemys scripta elegans………………………………………………………………………………….199  xv  List of Abbreviations ∆: delta, change (e.g. ∆V̇ E) ∆V̇ E: change in total ventilation rate ∆V̇ O2: change in oxygen consumption rate ∆ACR: change in air convection requirement (∆V̇ E/ ∆V̇ O2) µm: micrometer 5-HT: 5-hydroxytryptophan (serotonin) or serotonin cellular marker IX: refers to the ninth cranial nerve, glossopharyngeal nerve X: refers to the tenth cranial nerve, vagus nerve ACh: acetylcholine ACR: air convection requirement AMP: adenosine monophosphate ANOVA: analysis of variance Ao: aorta APUD: amine precursor uptake and decarboxylation ATP: adenosine triphosphate BK: background K+ channels ºC: degrees Celsius ºN: degrees North Ca+2: calcium ca: carotid artery CAM: contact adhesion molecules CaO2: arterial content of oxygen cm: centimeter CO2: oxygen content CO2: carbon dioxide CTB: neuronal tracer cholera toxin subunit B CvO2: venous content of oxygen DAPI: 4',6-diamidino-2-phenylindole (cellular marker that labels cell nuclei) fH: heart rate g-s connection: connection between glomus cells and smooth muscle cells H2O2: hydrogen peroxide xvi  H2S: hydrogen sulfide h: hours H-H: hypoxic hypercapnic gas mixture HNK-1: human natural killer-1 (labels a subset of migratory cells derived from the neural crest) HO-2: hemoxygenase-2 HVR: hypoxic ventilatory response Hz: hertz IgG: immunoglobulin G IR: immunoreactive K+: potassium L:D: refers to the light : dark cycle L-R: left to right, refers to a left to right shunt MAP: mean arterial pressure Ml: milliliters mmHg: millimeters of mercury, where 1 mm Hg is 0.1333 kPa or 0.0013 atm MR1: metabolic rate, used for temperature corrections of metabolic rate (Q10 = (MR2/MR1)10/(T2-T1) MR2: metabolic rate, used for temperature corrections of metabolic rate (Q10 = (MR2/MR1)10/(T2-T1) MS-222: tricaine methanesulphonate (anaesthetic) Na+: sodium NaCN: sodium cyanide NaCl: sodium chloride NADPH: nico nicotinamide adenine dinucleotide phosphate Na2HPO4: sodium phosphate dibasic NaH2PO4: sodium dihydrogen phosphate monhydrate NaOH: sodium hydroxide NDS: normal donkey serum NEB’s: neuroepithelial bodies NEC’s: neuroepithelial cells NGS: normal goat serum NTS: nucleus tractus solitarius (or the nucleus of the solitary tract) NVP: non ventilatory period or apnea O2: oxygen xvii  PA: pulmonary artery PBS: phosphate buffer solution PCO2: partial pressure of CO2 PO2: partial pressure of O2 or oxygen tension PaO2: arterial partial pressure of O2 PvO2: venous partial pressure of O2 Ppul: pulmonary blood pressure Psys: systemic blood pressure Pulsesys: systemic blood pressure pulse (Psys-Pdias) Q10: temperature coefficient Q̇ LAo: left aorta blood flow rate Q̇ pul: pulmonary blood flow rate Q̇ sys: systemic blood flow rate Q̇ pul/Q̇ sys: pulmonary to systemic blood flow rate ratio (shunt fraction) Q̇ tot: cardiac output (Q̇ sys + Q̇ pul) R-L: right to left, refers to a right to left shunt RM: Repeated measures ROS: reactive oxygen species Rpul: pulmonary resistance (Ppul/Q̇ pul) Rsys: systemic resistance (Psys/Q̇ sys) s: seconds s.e.m.: standard error of the mean STPD: standard temperature and pressure, dry T1: temperature, used for temperature corrections of metabolic rate (Q10 = (MR2/MR1)10/(T2-T1) T2: temperature, used for temperature corrections of metabolic rate (Q10 = (MR2/MR1)10/(T2-T1) TH: tyrosine hydroxylase VAChT: vesicular acetylcholine transporter (marker for transporter of acetylcholine into storage or synaptic vesicles) VAMP: lung ventilatory amplitude V̇ E: total lung ventilation rate (fR*VT) V̇ O2: oxygen consumption rate VP: ventilatory period xviii  Vstot: total stroke volume (pulmonary + systemic. Q̇ tot/fH) VT: tidal volume VTot; total ventilation (amplitude (VAMP) x breathing frequency (fR)) WGA: neuronal tracer wheat-germ agglutinin  xix  Glossary Anemia: deficiency in the oxygen carrying component of the blood. Anurans: one of the order of amphibians characterized by the absence of a tail (frogs and toads) Non ventilatory period (NVP) or apnea: defined as a respiratory pause that exceeded the duration of two missed breaths Bradycardia: a reduction in heart rate Chelonians: one of the orders of reptiles, which includes turtles and tortoises Circadian rhythms: Endogenous rhythm with a period that is close to the period of the solar day (24 hours). Circannual rhythms: Endogenous rhythm with a period that has a period close to a year. Standard temperature and pressure (dry, STPD): 21 ºC, 1 atmosphere (atm). Q10: Compares the rate of a reaction at two different temperatures. MR1 and MR2 are the rates of oxygen consumption at temperature t1 and t2, respectively. Q10 = (  MR 2  10  MR1  )  ( t 2 − t1 )  Hypercapnia: high levels of carbon dioxide Hyperoxia: higher than normal partial pressure of oxygen Hypoxemia: low oxygen content in the arterial blood Hypoxia: deficiency in the amount of oxygen reaching the tissues Hypoxic ventilatory response: increase in ventilation induced by a deficiency in the amount of oxygen reaching the tissues Oxygen content: (CaO2) total amount of oxygen in the blood Oxygen saturation: amount of oxygen carried by hemoglobin in the blood Oxygen tension: (PO2) oxygen dissolved in plasma P50: PO2 at which blood is 50% oxygenated xx  Acknowledgements I have spent several years at UBC and I have met wonderful people along the way, with whom I have shared this amazing journey and who I owe the guidance and support that have helped me to complete this stage of my life. I first want to thank my supervisor Bill Milsom, for welcoming me in his lab and for his guidance and support throughout my years at UBC. His excitement about science and positivism taught me to always see the best of things and to be open about ideas and possibilities. Thankyou Bill for being a wonderful mentor. I would also like to thank my committee members Dr. Colin Brauner, Dr. Tony Farrell and Dr. Wayne Vogl for their endless support and guidance. I am lucky to have had brilliant scientists guiding me throughout my PhD. Thanks to their dedication and support I have been able to complete this project. I want to give special thanks to Dr. Angelina Fong and Stella Lee without whom I would have not been able to complete my PhD. Angelina mentored me throughout my PhD and advised me in the initial stages of my thesis. She taught me different immunohistochemical and surgical techniques. Stella went with me on my second trip to Brazil, to help with chemical and rattlesnake handling while I was pregnant. She made the long days in the lab more enjoyable and has been an inconditional friend. Angelina and Stella have helped me face the most difficult stages of my PhD and their support and encouragement have made everything possible. I am grateful to Cleo Leite and Angela Scott from whom I learnt the most difficult surgical techniques I used in my thesis. Thankyou for your patience, advice and most importantly all the good times we shared together. I am also thankful to Dr. Matt Ramer who helped me figure out what neuronal tracer to use, after several months of unsuccessful trials. My lab mates have been an important part of my time at UBC. Thanks to Emily Coolidge, Barb Gajda, Cara Lachmuth, Ze Eduardo Carvallo, Charissa Fung, Dr. Cosima Porteus, Dr. Dee Brink, Dr. Graham Scott Dr. Jessica Meir and Yvonne Dzal. You guys have been incredible and I feel lucky to have met and shared wonderful moments with you. I want to thank other people in Zoology who have been my partners in crime throughout my years at UBC. Allison Barnes, Emily Coolidge, Anne Dalziel, Jessica Meir, Katie Kuker, Milica Mandic, Stella Lee, Gigi Lau, Aleeza Gerstein, Jess Hill, Brian Sardella, Sylvia Wood; Dave Toews, Alistair Blachford, Erica Eliason and Jeff Richards. Thanks for your friendship.  xxi  I would also like to thank the Zoology staff; especially Alice Liou for her friendship and for making sure I always had a committee form, and Alistair Blachford for his IT help. Thanks to Bruce Gillespie, Don Brandys, and Vincent Grant for building and fixing my equipment. I am grateful to Garnet Martins and Kevin Hodgson for their advice, training and troubleshooting during my long hours in the bioimaging facility. When I first arrived to UBC I met two amazing women, Janice Meier and Allison Barnes, who have been an important part of my life and are my family away from home. You ladies have filled my life with such joy and sweet memories and I feel really lucky to have friends like you. Thank you for caring and sharing all those special moments with me. I want to thank my family for their unconditional support, encouragement and love. My dad, mom, grandpa, brother, sister and Taquito have gone out of their way to help me complete my degree. They have supported me during the toughest times of my PhD and have celebrated my accomplishments despite being so far away. I love you all. I would also like to thank my lovely husband, Jonathan, my daughter Olivia and Willy who have filled my life with love, joy, laughs and cuddles and who have made this process manageable for me. Jon has also read the whole thesis and given me advice on the structure and organization of this document. He has been incredibly patient and loving. Finally, I would like to thank the Natural Sciences and Engineering Research Council of Canada for funding this research.  xxii  Dedication To my loving family Abey, Alfonso, Camelia, Alejandro, Gabby, Jonathan, Olivia and Willy  xxiii  Chapter 1: Introduction Peripheral chemoreceptors, associated with oxygen sensing are present in all vertebrates studied to date. Oxygen sensing cells have been comprehensively studied in fish, mammals and, to a lesser extent, amphibians. Less is known regarding peripheral chemoreceptors in birds and turtles, and virtually nothing is known about the peripheral arterial chemoreceptors in nonchelonian reptiles (lizards, snakes and crocodilians) (Adams, 1958; Milsom and Burleson, 2007). The aim of my thesis was to test the hypothesis that O2-sensing structures are highly conserved among vertebrates while their locations (there are multiple O2-sensing sites), stimulus modalities, and plasticity (referred as the ability to change within two time frames, daily and seasonally) in their reflex response have evolved to more effectively match ventilation and perfusion, particularly in animals with central cardiac shunts (amphibians and reptiles).  1.1  Phylogeny of peripheral arterial chemoreceptors  1.1.1  Location, innervation and distribution of putative O2 sensing cells  In all species studied to date, peripheral chemoreceptors have always been found associated with the pharyngeal arches or their derivatives (such as aortic arches) and, innervated by the glossopharyngeal and/or vagus nerves (IX and X cranial nerves, respectively) (Adams, 1958; Ishii et al., 1966; Rogers, 1967; Abdel-Magied and King, 1978; Lahiri et al., 1981; Ishii et al., 1985a; Ishii et al., 1985b; Ishii and Ishii, 1986; de Graaf, 1990; Gonzalez et al., 1994; Jonz and Nurse, 2003, 2009) (Fig. 1.1). In embryonic vertebrates generally six pairs of aortic arches arise from the ventral aorta and join the paired dorsal aortae (Kardong, 2006). In teleost fish and larval amphibians the first (mandibular arch) and second (hyoid arch) embryonic aortic arches degenerate. The remaining aortic arches (3-6) perfuse the gills in the adult fish (Jonz and Nurse, 2009). In adult amphibians, reptiles, birds and mammals the aortic arch 5 also disappears and the remaining arches 3, 4 and 6 become the carotid artery, aortic arch and pulmonary artery (pulmocutaneous artery in amphibians), respectively. In amphibians and reptiles the aortic arch remains paired, while in birds and mammals only one aortic arch is retained (right and left systemic arch in birds and mammals, respectively) (Kardong, 2006). In fish, the third pharyngeal arch is innervated by both the IX and X cranial nerves, as is the carotid labyrinth in amphibians while the carotid bifurcations in birds and mammals are either innervated by the X 1  (birds) or IX (mammals) cranial nerves. The fourth and sixth pharyngeal arches in fish are innervated solely by the X cranial nerve as are their derivatives in all other vertebrates (West and Van Vliet, 1992; Gonzalez et al., 1994; Jones and Milsom, 1982; Milsom and Burleson, 2007). 1.1.1.1  Fish  Neuroepithilial cells are currently believed to be the oxygen sensing cells in fish and have been located in all extant gill arches (3-6 embryonic pharyngeal arches), innervated by the IX (first gill arch) and X cranial nerves, as well as in the pseudobranch, innervated by the facial (VII) and IX cranial nerves, and the orobranchial cavity, innervated by the trigeminal (V) and VII cranial nerves (Dunel-Erb et al., 1982; Burleson and Milsom, 1993; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz et al., 2004; Milsom and Burleson, 2007; Bailly, 2009; Jonz and Nurse, 2009; Milsom, 2012). Hypoxia (deficiency in the amount of oxygen reaching the tissues), in most fish, elicits a reduction in heart rate (bradycardia) and an increase in ventilation (through both and increase in frequency and amplitude). The bradycardia and increase in breathing frequency appears to be regulated primarily by chemoreceptors distributed in all gill arches in many fishes (reviewed by Milsom, 2012), while the change in breathing amplitude arises primarily from activation of chemoreceptors in all gill arches, with some contribution from extrabranchial receptors (Sundin et al., 1999; Sundin et al., 2000; Leite et al., 2007). 1.1.1.2  Amphibians  In amphibians the third embryonic pharyngeal arch becomes the carotid artery which contains the carotid labyrinth. This structure is a vascular swelling of the common carotid artery at the site where the internal and external carotid arteries arise. Internally, the carotid labyrinth is a very complex maze-like vascular bed (Toews et al., 1982; Kusakabe, 2002, 2009). It is innervated by a branch of the IX cranial nerve, the carotid nerve, and a fine pharyngeal branch of the X nerve (Adams, 1958; Ishii et al., 1966; Van Vliet and West, 1987b; Kusakabe, 2002, 2009). This structure is considered homologous to a portion of the first gill arch of fishes and to the carotid body in mammals (West and van Vliet, 1992; Milsom and Burlenson, 2007). The hypoxic ventilatory response (increase in ventilation induced by a deficiency in the amount of oxygen reaching the tissues) in amphibians is primarily regulated by the carotid labyrinth. Bilateral denervation of the carotid labyrinth, however, does not abolish the 2  ventilatory response (Van Vliet and West, 1986; West et al., 1987) and two other areas have been suggested to be chemoreceptive; the aortic arch and the pulmocutaneous artery (Lillo, 1980; Hoffmann and de Souza, 1982; Ishii et al., 1985a; Wang et al., 2004). These blood vessels are derived from the fourth and sixth embryonic pharyngeal arches respectively (homologous to the 2nd and 4th gill arches in fishes). The aortic trunk, innervated by a pharyngeal branch of the vagus nerve, is a chemosensory and barosensory area. Perfusion with hypoxic or hypoxichypercapnic (low levels of oxygen and high levels of carbon dioxide) solution or with cyanide (NaCN, inhibits cellular respiration by acting on mitochondrial cytochrome oxidase and blocking electron transport) stimulates nerve discharge and results in elevation of ventilation and blood pressure (Ishii et al. 1985a). The pulmocutaneous artery, however, may also be chemosensory and appears to be the dominant barosensory area in anurans (frogs and toads) (West and Van Vliet, 1983; Van Vliet and West, 1987a). This area is innervated by the recurrent laryngeal nerve, a branch of the X cranial nerve (West and Van Vliet, 1992). Injections of NaCN into the pulmocutaneous artery increase ventilation in anaesthetised frogs and cause bradycardia in submerged frogs, suggesting the presence of chemoreceptors in this area (Lillo, 1980). 1.1.1.3  Turtles  Based on anatomical descriptions, Adams (1962) concluded that the area homologous to the carotid bifurcation in mammals was not located at the extant carotid bifurcation in turtles, but was situated more centrally in the truncal region. During development an anastomosis forms between the developing internal and external carotid arteries high in the neck and the rostral portion of the external carotid atrophies. Thus, the internal carotid artery gives rise to the extant external carotid artery secondarily and the original carotid bifurcation is lost. This region regresses into the dorsal (common) carotid artery (Adams, 1958; Rogers, 1967; Jones and Milsom, 1982). A chemosensory area homologous to the carotid bodies in mammals has been located in the dorsal carotid artery, innervated by the carotid nerve, a branch of the IX cranial nerve, in one species of tortoise (Ishii and Ishii, 1986). However, although carotid nerve discharge from this site significantly increased with NaCN and hypercapnic stimulation, this area did not seem to be very sensitive to hypoxia or hyperoxia (Ishii and Ishii, 1986).  3  A few studies have also found several chemosensory sites in the aortic arch (derivative of the 4th pharyngeal arch), pulmonary artery (derivative of the 6th pharyngeal arch) and truncus arteriosus (Adams 1962; Benchetrit et al. 1977; Ishii et al. 1985b). In the turtle Geoclemmys reevesii (Ishii et al., 1985b) and tortoise (Testudo hermanni) (Ishii and Ishii, 1986) the superior laryngeal nerve (a branch of the X cranial nerve) gives off the superior truncal nerve (homologous to the aortic nerve), which runs along the common carotid artery and innervates the aortic arch. The inferior truncal nerve arises from the glanglion trunci of the vagus and gives off two fine branches near "thymus four", one to the aortic arch and one to the pulmonary arch, and runs further to innervate the common pulmonary artery and bulbus cordis (Ishii et al., 1985b). Electrophysiological studies yield different conclusions about which of these areas are the major chemoreceptive sites in turtles. Benchetrit et al. (1977) suggested that the pulmonary arch was the main chemosensory area in the tortoise based on responses to NaCN injections while Ishii et al. (1985b) implied that the aortic arch was the major reflexogenic area due to its innervation by both the superior and inferior truncal nerves. 1.1.1.4  Lizards  In this group the internal and external carotid arteries arise from the common carotid artery in the neck as in mammals. The internal carotid artery, however, arises by a number of internal openings (Adams, 1958). The superior laryngeal nerve arises from the ganglion nodosum and innervates the origin of the internal carotid artery. Fibers from the glossopharyngeal and sympathetic nerves also innervate this region, although this varies between species (Adams, 1958; Adams, 1962; Rogers, 1967). Furthermore, the truncus arteriosus of lizards which has been proposed to contain baroreceptors may also have a chemosensory role based on the presence of chromaffin cells (Berger et al., 1982). Studies of peripheral arterial chemoreceptors at this and other sites in this group are few. Small groups of epithelioid cells (Adams, 1962) or clusters of 10-12 cells (Rogers, 1967) were found where the internal carotid arises from the common carotid artery. Most studies on lizards have only focused on the carotid arch, while equivalent chemosensory areas reported in turtles have not been examined.  4  1.1.1.5  Snakes  In many snakes, the right common carotid artery is truncated and does not project to the head. The left common carotid artery extends along the neck and bifurcates to form the internal and external carotid arteries just behind the angle of the jaw. The internal and external carotid arteries on the right side are still extant and are connected to those on the left side by an anastomosis just before they enter the head; they receive their blood flow ultimately from the left common carotid artery. The glossopharyngeal nerves appear to innervate the carotid bifurcations bilaterally. However, there has been no evidence to suggest that this area is homologous to the carotid bifurcation in mammals and no specialized cells have been located at this site (Adams, 1958). 1.1.1.6  Crocodilians  In crocodilians, as in turtles the extant carotid bifurcation is secondarily derived. The internal and external carotid arteries arise separately from the brachiocephalic trunk in this group. The right internal carotid artery disappears between its origin from the brachiocephalic trunk to the point where it joins the left internal carotid artery (Reese, 1914). It has been suggested that the chemosensory area homologous to the mammalian carotid body lies at the origin of the left internal carotid artery. Three branches of the X cranial nerve have been found to innervate this area: the vagus gives off a branch above the ganglion nodosum; a branch of the oesophageal nerve that arises from the superior laryngeal nerve, and a branch that arises directly from the superior laryngeal nerve after its anastomosis with the IX cranial nerve. As in snakes no specialized cells have been located in this area (Adams, 1958). 1.1.1.7  Birds  The carotid body in birds is located on the common carotid artery. As in turtles the internal carotid artery divides secondarily and gives rise to the external carotid artery. The site of the embryonic carotid bifurcation, therefore, remains near the heart (Adams, 1958; Jones and Milsom, 1982). The carotid body in birds is innervated by the X cranial nerve (Abdel-Magied and King, 1978) and regulates the ventilatory reflexes to hypoxia and hypercapnia (Jones and Purves, 1970a). In diving ducks, these chemoreceptors are largely responsible for the diving bradycardia (Jones and Purves, 1970b).  5  1.1.1.8  Mammals:  Peripheral chemoreceptors have been comprehensively studied in mammals. These chemoreceptors have been located in the carotid and aortic arteries (Lahiri et al., 1983; Gonzalez et al., 1994) and possibly on the pulmonary artery (Krahl, 1962). Aortic and carotid bodies respond to changes in arterial blood gases and acid-base status by generating reflex responses in the respiratory and cardiovascular systems (Lahiri et al., 1983). The carotid bodies are located near the common carotid bifurcation into the external and internal carotid arteries. These are innervated by the carotid branch (carotid sinus nerve - CSN) of the IX cranial nerve. The sensory cell bodies of the fibres carried in the carotid nerve are in the inferior ganglion (petrosal ganglion). Filaments of the X cranial nerve join the CSN (Gonzalez et al., 1994). The aortic bodies are located in the aortic arch and are innervated by the aortic nerve, a branch of the X cranial nerve. The response of these chemoreceptors to hypoxia (at a constant CO2 tension, PCO2) is reduced compared to that of the carotid chemoreceptor (Lahiri et al., 1981). Aortic chemoreceptor activity correlates with changes in arterial oxygen content (CaO2, total amount of O2 in the blood), while arterial oxygen tension (PaO2, oxygen dissolved in plasma) appears to be the driving stimulus of carotid chemoreceptor activity (Lahiri et al., 1983). It appears that aortic chemoreceptors mainly control cardiovascular reflexes, such as bradycardia (Eyzaguirre et al., 1983), while carotid bodies contribute to the hypoxic ventilatory response (Fitzgerald and Lahiri, 1986).  1.2  Characterization of oxygen sensing cells at arterial chemosensitive sites While most cells respond to hypoxia by reducing their activity, there are a special group of  excitable cells that respond to hypoxia by activating second order sensory afferent neurons (Prabhakar, 2006). These cells include neuroepithelial cells, neuroepithelial bodies and glomus cells (Gonzalez et al., 1994; Cutz and Jackson, 1999; Peers and Kemp, 2001). This section reviews the current state of knowledge of the characteristics (anatomical features, neurochemical content and organization) of O2-sensing cells in vertebrates and the conspicuous gaps in our understanding (Table 1.1). 6  1.2.1  Fish  In fish, oxygen sensing structures in the gills are thought to be neuroepithelial cells (NEC’s) (Dunel-Erb et al., 1982; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz et al., 2004; Saltys et al., 2006; Coolidge et al., 2008; Bailly, 2009; Jonz and Nurse, 2009). NEC’s are also found in the lungs of mammals, where they are grouped into the neuroepithelial bodies (Cutz and Jackson, 1999). NEC’s are structurally similar to glomus cells and it has been suggested that they may be phylogenetic precursors of glomus cells (Jonz and Nurse, 2003; Milsom and Burleson, 2007; Jonz and Nurse, 2009; Jonz and Nurse, 2012; Zachar and Jonz, 2012). Under acute hypoxia, NEC’s release peptide and amine modulators, mainly serotonin (Peers and Kemp, 2001). In fish, the neurotransmitters known to participate in chemotransduction are yet unknown. Serotonin (5HT) is believed to play a key role based on anatomical studies (Jonz et al., 2004; Regan et al., 2011) but acetylcholine (ACh) is believed to be the key transmitter based on behavioral and physiological studies (Jonz and Nurse, 2003; Burleson and Milsom, 1995; Regan et al., 2011). Evidence for the presence of other neurotransmitters in the NEC’s of fish is equivocal. Several studies have failed to find catecholamines in the NEC’s of trout, goldfish (reviwed by Porteus et al., 2012) or catfish (Zaccone et al., 2003), but cultured NEC’s of the channel catfish were shown to be immunoreactive for tyrosine hydroxylase (TH-IR, TH is the rate limiting enzyme in catecholamine production) and to respond to hypoxia (Burleson et al., 2006).  1.2.2  Amphibians In amphibians, glomus cells appear sporadically in the stroma of the carotid labyrinth in  groups of two or three cells (Ishii and Oosaki, 1969). These cells contain a Golgi complex, granulated cytoplasm and dense-cored vesicles (Ishii and Oosaki, 1969; Kobayashi, 1971b; Ishii and Kusakabe, 1982; Kusakabe, 1991). Sustentacular cells surround glomus cells, although only partially, so that glomus cells are in direct contact with endothelial cells (Ishii and Oosaki, 1969; Kobayashi, 1971b). Glomus cells in the carotid labyrinth of anurans possess afferent and efferent synapses (Ishii and Oosaki, 1969). The carotid nerve in toads, for instance, contains fibers from the vagus and sympathetic nerves (Ishii and Ishii, 1973). Slowly conducting fibers of the vagus nerve are chemosensory and rapidly conducting fibers are barosensory (afferent), 7  while sympathetic fibers are vasoconstrictors of the labyrinth (i.e. they are efferent fibers) (Ishii and Ishii, 1973). Stimulation of the carotid labyrinth by hypoxemia (low level of O2 in the blood) results in an increase in ventilation (Ishii et al., 1966). Increased respiration stimulates efferent sympathetic fibers, which depresses chemoreceptor firing. Thus, sympathetic efferents and to some extent efferent vagal fibers seem to modulate chemosensitivity, but only the former regulate vascular tone (West and Van Vliet, 1992). Vascular tone may also be regulated by the close connection that exists between glomus cells and smooth muscle (g-s connection). Efferent stimulation of the IX cranial nerve causes exocytosis of catecholamine-containing vesicles at the g-s connection (Ishii and Kusakabe, 1982). Catecholamines reduce blood flow to the internal and external carotid arteries. ACh and NaCN infusions on the other hand, only reduce blood flow to the internal carotid artery (Kusakabe et al., 1987). Alternatively, neuropeptides in nerve fibers within the carotid labyrinth may also regulate vascular tone (Kusakabe et al., 1991, 1994). As many aspects of the carotid labyrinth of anurans have been well documented, it is surprising that very little is known about their putative neurotransmitters and their role in oxygen sensing. As mentioned above, neuropeptides have been found in nerve terminals innervating the carotid labyrinth, but not in glomus cells. Evidence of the occurrence of catecholamines in the carotid labyrinth has been shown by formaldehyde vapor exposure. While this technique is nonspecific, it has been used to identify biogenic monoamines. Catecholamines fluoresce greenyellow while serotonin fluoresces yellow (Banister et al., 1967; Kobayashi, 1971a). Optimal fluorescence of 5HT, however, requires strong reaction conditions which will cause diffusion of the catecholamine fluorescence. Furthermore, at high concentration, the peak emission of catecholamine fluorescence will be very similar to that of serotonin (Corrodi and Jonsson, 1967), making it difficult to demonstrate the presence of both amines in the same tissue. Glomus cells in the aorta show similar characteristics to those of the carotid labyrinth (Ishii et al., 1985a) and exposure to formaldehyde vapor has been used to show they also possess catecholamines (Banister et al., 1967).  8  1.2.3  Turtles  In turtles, groups of characteristic glomus and sustentacular cells are found in the adventitia of the carotid artery, aortic arch and pulmonary artery. These glomus-like cells contain densecored vesicles (Kusakabe et al., 1988) that may contain catecholamines or 5HT, since monoamine-containing cells have been observed in the wall of the aorta (Ishii et al., 1985b) and in the opening of the main cardiac vessels (Chiba and Yamauchi, 1973) by exposure to formaldehyde vapor. To date, however, no specific neurotransmitters have been identified.  1.2.4  Lizards  Small clusters of cells similar to those of the mammalian carotid body have been located in the adventitia of the carotid arch where the internal carotid artery arises. Exposure to formaldehyde vapor resulted in intense yellow fluorescence in the Japanese lizard, suggesting the presence of 5HT or catecholamines (Kobayashi, 1971a). As in turtles the presence of catecholamines and 5HT has been shown by histochemical techniques, but no specific characterization of neurotransmitters involved in signal chemotransduction has been made.  1.2.5  Snakes and crocodilians  No structures resembling glomus cells of mammals or NEC’s of fish have been described in these groups.  1.2.6  Birds  Glomus tissue responsive to hypoxia has been described in birds (Nye and Powell, 1984), but the neurochemical content of these areas has not been thoroughly studied. The presence of 5HT has been confirmed by formaldehyde vapor (Kobayashi, 1971a) and hypoxia has been shown to elicit the release of 5HT from epithelioid cells in the aorta of chickens (Ito et al., 1997).  9  1.2.7  Mammals  Peripheral arterial chemoreceptors in mammals consist of an association of glomus cells, sustentacular cells and afferent nerve endings (Jones and Milsom, 1982; Lahiri et al., 1983; Gonzalez et al., 1994). Glomus cells are epithelial cells derived from neural ectoderm of the neural crest (Pearse et al., 1973; Kondo et al., 1982) and possess a clear round nucleus, granulated cytoplasm and dense-cored vesicles containing neurotransmitters such as catecholamines (dopamine, adrenaline and noradrenaline), 5HT, substance P, ACh and ATP, which are released in response to hypoxemia (Gonzalez et al., 1994; Peers and Kemp, 2001; Prabhakar, 2006). Glomus cells are surrounded by sustentacular cells characterized by a diskshaped nucleus and non granulated cytoplasm. The afferent nerve endings associated with glomus cells are sensory fibers from the carotid nerve (carotid sinus nerve), a branch of the IX cranial nerve, and the aortic nerve, a branch of the X nerve (Lahiri et al., 1983; Gonzalez et al., 1994). This association of cells may be important for the chemo-transduction of the signal, although the mechanisms by which oxygen is sensed are not entirely understood. The nature of the oxygen sensor has been the focus of recent research and a few hypotheses have been proposed (see Oxygen sensing mechanisms below). In mammals, there is evidence to suggest that ultimately inhibition of K+ channels at low oxygen tension (PO2) causes membrane depolarization and opening of voltage-gated calcium channels. The increase in intracellular calcium initiates the release of neurotransmitters which act on afferent nerve endings increasing sensory discharge (Bunn and Poyton, 1996; Prabhakar, 2000). As mentioned above, glomus cells of the mammalian carotid body contain a plethora of neurotransmitters, some of which play an excitatory, inhibitory or modulatory role (reviewed by Nurse, 2005, 2010). Although the function of different neurotransmitters seems to vary between species (Gonzalez et al., 1994), it has been shown that ACh co-released with ATP are key players in signal chemotransduction, by increasing sensory nerve discharge (Eyzaguirre and Zapata, 1984; Zhang et al., 2000; Nurse, 2005; Shirahata et al., 2007; Nurse, 2010). In most species, 5HT and dopamine seem to be involved in excitatory and inhibitory modulation (Gonzalez et al., 1994; Nurse, 2005, 2010). The significance of having inhibitory neurotransmitters is that it maintains the sensory discharge for the duration of the hypoxic stimulus, by modulating the over-excitation produced by the excitatory neurotransmitters (Prabhakar, 2006).  10  1.3  Oxygen sensing mechanisms Much research has centered on determining the mechanism of oxygen sensing in the carotid  body of different species of mammals. To date, there is no consensus as to which is the oxygen sensor, perhaps due to species differences, but several hypotheses have emerged that fall into two general categories: the redox hypothesis and the heme-protein hypothesis. The redox hypothesis states that changes in PO2 will alter production of intracellular reactive oxygen species (ROS) by a membrane bound NADPH oxidase (nicotinamide adenine dinucleotide phosphate oxidase) or by the mitochondria (the mitochondria-derived ROS hypothesis); and that the change in production of ROS during hypoxia inhibits background K+ (BK) channel activity. NADPH oxidase is a heme-containing enzyme bound to the plasma membrane that produces ROS such as hydrogen peroxide (H2O2) which increases K+ permeability (Gonzalez et al., 2007). The presence of O2 determines the rate of H2O2 production. Thus, the formation of H2O2 is decreased during hypoxia, inhibiting K+ channels (Prabhakar, 2000; Chandel and Schumacker, 2000; Lopez-Barneo et al., 2001; Kemp, 2006). The mitochondrion is another source of ROS. A small amount of the oxygen consumed by the mitochondria is not completely reduced, forming ROS. This hypothesis states that the formation of ROS is increased during hypoxia (Chandel and Schumacker, 2000; Kemp, 2006). The heme-protein hypothesis proposes that a protein capable of reversibly binding O2 at a heme site is the oxygen sensor. Several heme-containing proteins have been proposed as the oxygen sensor: (1) mitochondrial cytochrome oxidase contains heme, and inhibitors of the mitochondrial respiratory chain, such as cyanide, increase sensory discharge similarly to hypoxia (Prabhakar, 2000). (2) Hemoxygenase-2 (HO-2) is found within the complex of a specific type of membrane bound K+ channel. When molecular O2 is available, HO-2 produces carbon monoxide (CO), which increases the activity of these K+ channels. During hypoxia, a decrease in CO leads to a fall in K+ permeability (Prabhakar, 1999; Williams et al., 2004). Evidence also indicates that O2 sensing is linked to metabolic energy production in the mitochondria. It has been proposed that a fall in ATP production during hypoxia increases the AMP/ATP ratio, which in turn activates AMP-activated protein kinase leading to inhibition of K+ channels (Wyatt and Evans, 2007). Recent studies have proposed the role of sulfide (H2S) as an O2 sensor. The levels of intracellular H2S are maintained by the balance between its 11  production and oxidation. Thus, hypoxia leads to accumulation of H2S in the cell (Olson et al., 2006). Support for all the proposed mechanisms of oxygen sensing has been demonstrated in different mammalian species, and it has become clear that no unifying theory can explain how oxygen is sensed in the carotid body of all species. Prabhakar (2006) proposed the chemosome hypothesis which states that multiple oxygen sensors are involved in the response to hypoxia. He argues that this may be advantageous as it allows the glomus cell to respond to a broad range of PO2 and at a faster rate (Prabhakar, 2006). Carotid bodies differ from other oxygen sensing tissue in their rapid response, high sensitivity to small changes in PO2 (from 100 to 80 mmHg) and lack of adaptation to a constant stimulus; i.e. afferent discharge is maintained throughout the hypoxic stimulus (Prabhakar, 2006). As explained above these characteristics could be attributable to the presence of multiple neurotransmitters and oxygen sensors (chemosome hypothesis) within the glomus cells of the carotid body.  1.4  Overview of peripheral chemoreceptors in vertebrates To summarize (Table 1.1), NEC’s occur in all gill arches and in the pseudobranch of fish.  In some species NEC’s are found in the orobranchial cavity as well (reviewed by Smatresk, 1990; Milsom and Burleson, 2007; Milsom, 2012). All gill arches are innervated by the X nerve, while the first gill arch is also innervated by the IX nerve. NEC’s generally do not occur in clusters and evidence suggests that 5HT is the main neurotransmitter present in these cells (reviewed by Jonz and Nurse, 2009, 2012), although other neurotransmitters, such as ACh, catecholamines and nitric oxide may also be involved in oxygen chemotransduction (reviewed by Porteus et al., 2012). Lower tetrapods that have been studied so far (anurans and turtles) appear to have several chemosensory areas and glomus cells within these areas are dispersed Ishii et al., 1966; Ishii and Oosaki, 1969; Kobayashi, 1971b; Ishii and Kusakabe, 1982; Ishii et al., 1985a, b; Ishii and Ishii, 1986; Kusakabe et al., 1988; Kusakabe, 1991). In mammals only two chemosensory sites remain the carotid and aortic bodies. The carotid body is innervated by the IX cranial nerve and the aortic body by the X cranial nerve. Glomus cells occur in clusters and contain a plethora of neurotransmitters (carotid body) (Gonzalez et al., 1994; Nurse, 2005, 2010). NEC’s of fish share a number of features with the glomus cells in mammals, and it has 12  been proposed that NEC’s are the evolutionary precursors of mammalian glomus cells (Jonz and Nurse, 2003; Milsom and Burleson, 2007; Jonz and Nurse, 2009, 2012; Zachar and Jonz, 2012). The phylogenetic trend appears to be a reduction in the number of oxygen sensing areas from fish to mammals (Figure 1) (Milsom, 1998; Milsom and Burleson, 2007), as well as a tendency to form a confined cluster of glomus cells and to increase the number of neurotransmitters involved in signal chemotransduction. However, more information regarding the location, distribution, arrangement and neurochemical content of peripheral chemoreceptors in lower tetrapods is needed to be able to draw conclusions about how peripheral chemoreceptors have evolved.  1.5  Significance of locating and characterizing putative oxygen sensing cells in  amphibians and reptiles In this dissertation I used a comparative approach to determine the location, innervation and neurochemical profiles of putative O2-sensing cells in different species of reptiles (Trachemys scripta elegans and Crotalus durissus) and an amphibian (Rana catesbeiana). Finding oxygen sensing cells associated with aortic arches and innervated by the IX or X cranial nerves, with similar morphology and neurotransmitters as fish NEC’s and mammalian glomus cells would suggest that the mechanisms of O2-sensing are conserved among vertebrates. The location (multiple chemosensory areas), distribution and sensitivities of these cells, however, may have changed so that more effective cardio-respiratory control can be exerted in animals that can control levels of arterial blood gases through changes in ventilation or the degree of cardiac shunt (Wang et al., 1997) (see Reflex roles and stimulus specificity of peripheral chemoreceptors in amphibians and reptiles below). It is important to note that this dissertation examinines peripheral chemoreceptors in just a few species of amphibians and reptiles. The variation in the anatomy and function of the cardiorespiratory system within these groups is large, and has potentially important effects on control systems, such as peripheral chemoreceptors. However, it is crucial to identify and characterize peripheral arterial chemoreceptors in these groups to advance our understanding on how these control systems have evolved in vertebrates. My thesis represents a first step toward understanding the regulation of blood gases in reptiles and amphibians. 13  1.6  Cardiovascular anatomy of amphibians and reptiles and presence of cardiac shunts  1.6.1  Amphibians  The hearts of amphibians have two atrial chambers and a single trabeculate ventricle. The conus arteriosus originates from the ventricle. Internally the conus arteriosus is bisected by a spiral valve that forms two channels within the conus. This provides a mechanism by which mostly oxygenated blood is directed to the systemic arches and deoxygenated blood to the pulmocutaneous arch. The right and left aortae and pulmocutaneous arteries emerge from the truncus arteriosus (Kardong, 2006).  1.6.2  Reptiles  The hearts of turtles and snakes have two completely divided atria and a partially undivided ventricle. A muscular ridge subdivides the ventricle into the cavum pulmonale and the cavum dorsale, which is further divided into the cavum arteriosum and the cavum venosum. Because the dorsolateral end of the muscular ridge is free, there is potential communication between the ventricular chambers during diastole and systole (Farrell et al., 1998; Hicks; 1998, 2002). The pulmonary artery originates from the cavum pulmonale and the right and left aortae arise from the cavum venosum. The cavum arteriosum has no direct output (Burggren, 1987; Wang et al., 1997; Farrell et al., 1998; Hicks, 2002). During diastole deoxygenated blood from the right atrium enters the cavum venosum and subsequently flows to the cavum pulmonale, while oxygenated blood from the left atrium enters the cavum arteriosum. During systole, blood flows from the cavum arteriosum to the cavum venosum and to the left and right aortae. Blood from the cavum pulmonale is ejected directly through the pulmonary arteries (Hicks, 1998). The undivided ventricle of amphibians and reptiles allows for central vascular shunts. During a R-L shunt systemic venous blood bypasses the pulmonary circulation. A L-R shunt consists of recirculation of pulmonary venous blood (oxygenated blood) into the pulmonary circulation. Two hypotheses have been advanced to explain the mechanisms of shunt. The washout hypothesis states that the muscular ridge completely separates the cavum pulmonale and cavum venosum during systole. The R-L shunt results from deoxygenated blood that remained in the cavum venosum at the end of diastole. The L-R shunt results from oxygenated blood left 14  in the cavum venosum after systole that is washed into the cavum pulmonale during diastole (Fig. 1.2) (Wang et al., 1997). The pressure hypothesis states that the muscular ridge does not separate the cavum pulmonale and the cavum venosum during systole so that blood will flow between these chambers according to the differences in the outflow resistances (systemic and pulmonary resistance) (Hicks and Malvin, 1995). Both washout and pressure shunts can occur simultaneously.  1.7  Reflex roles and stimulus specificity of peripheral chemoreceptors in amphibians and  reptiles 1.7.1  Reflex roles of peripheral arterial chemosensitive areas  In order to maintain arterial blood gas homeostasis, adjustments in the cardio-respiratory system must occur to match oxygen supply to oxygen demand. In birds and mammals PaO2 is very similar to the gas tension in the lung. Therefore, efficient regulation of arterial blood gases can be achieved by changes in ventilation alone (Wang and Hicks, 1996a; Wang et al., 1997). In animals with central cardiac shunts, however, adjustments in both the respiratory and cardiovascular systems are necessary for efficient blood gas maintenance (Hicks and Wood, 1989; Wang and Hicks, 1996a; Wang et al., 1997; Wood, 1982, 1984). The relative importance of the respiratory and cardiovascular systems to the regulation of blood gases depends largely on environmental and physiological stressors (hypoxia, hypercapnia and anemia). For instance, in normoxia it is more efficient for turtles to eliminate the R-L shunt than to increase ventilation to increase oxygen delivery to tissues (Wang and Hicks, 1996a), since haemoglobin in blood leaving the lung will already be fully saturated. Consequently increases in ventilation will have little effect on arterial oxygen saturation (amount of O2 carried by hemoglobin in the blood) (Wood, 1984) while reducing the shunt will enhance tissue oxygen delivery. The same is also true in anemia (reduced O2 carrying component of the blood) (Wang et al., 1997). In hypoxia, while changes in ventilation become more effective in increasing PaO2, the combined effect of increasing ventilation and reducing the R-L shunt is larger than adjustments in ventilation alone (Fig. 1.3) (Wang and Hicks, 1996a). It appears that it is most efficient to regulate changes in CaO2 by preferentially regulating the cardiovascular system (heart rate and degree of cardiac shunt), but to regulate changes in PaO2 by adjusting ventilation. To accomplish this, Wang et al. 15  (1994, 1997) have suggested the presence of anatomically separated groups of chemoreceptors that sense the content of oxygen or PO2 (see below) and regulate the cardiovascular and respiratory systems respectively (Wang et al., 1994).  1.7.2  Stimulus specificity of peripheral arterial chemoreceptors  In amphibians and reptiles in the absence of cardiac shunts, CaO2 will depend on the PaO2, hemoglobin concentration and hemoglobin oxygen affinity (often indicated by P50, the PO2 at which blood is 50% oxygenated). During a R-L shunt (blood bypasses the lungs) CaO2 is reduced and PaO2 becomes dependent on the degree of shunt, since this is equivalent to mixing solutions in a closed system. CaO2 is determined by the fraction of shunted blood and the respective CaO2 of the shunted and non-shunted blood, while arterial PaO2 is determined by the CaO2 of the mixed blood and blood oxygen affinity (Wood, 1984). Thus in amphibians and reptiles arterial blood gases can change independent of changes in lung oxygen tension. To date, CaO2 sensing chemoreceptors have only been found in the mammalian aortic bodies. Perfusion of the aortic chemoreceptors is low relative to their oxygen consumption rate and thus their PO2 is affected by changes in CaO2, blood flow and hemoglobin affinity (Lahiri et al., 1981). This occurs because at the tissue level, PO2 is a function of CO2 and cccording to the Fick equation (V̇ O2 = Q̇ x (a-v)CO2), if the oxygen consumption is maintained, then a change in O2 content will lead to a larger change in PO2 when blood flow is low. Other chemoreceptors overtly sensitive to changes in oxygen content have not been found in other vertebrates, including amphibians or reptiles. Ectothermic vertebrates, however, show a stronger correlation between the threshold of the hypoxic ventilatory response and CaO2 than with PaO2 (Glass et al., 1983; Wood, 1984). Furthermore, Wang et al. (1997) have shown that anemic turtles increased pulmonary blood flow during apnea even though no changes in oxygen tension occurred (Fig. 1.4). This led them to suggest two hypotheses: 1) either there was a chemoreceptor group that sensed CaO2 and mainly regulated the cardiovascular system 2) or there was a PO2 sensitive chemoreceptor in the venous circulation since a fall in CaO2 will reduce venous oxygen tension (PvO2) and oxygen content (CvO2) (Wang et al., 1997). To date, attempts to test these hypotheses in toads have not been successful (Wang et al., 1994; Andersen et al., 2003; Wang et al., 2004) and the stimulus specificity and reflex role of distinct chemosensory areas remain unknown. 16  1.7.3  Significance of establishing the reflex roles and stimulus modalities of  chemosensory areas in amphibians and reptiles Studies on the regulation of blood gas homeostasis in vertebrates have often focused on the regulatory role of respiration ignoring the important contribution of the cardiovascular system, particularly in animals that possess central cardiac shunts (Wang and Hicks, 1996b). To fully understand mechanisms involved in cardio-respiratory control it is necessary to identify and characterize the chemosensory areas responsible for sensing the levels of arterial blood gases and determine their relative roles in the various cardio-respiratory reflex responses. In this dissertation I aimed to determine if (1) adjustments in shunt pattern, which affects arterial blood gases, are controlled by peripheral chemoreceptor input, and (2) if the distinct groups of peripheral chemoreceptors described in reptiles have different regulatory roles (cardiovascular or respiratory).  1.8  Plasticity of chemoreflexes Reptiles generally respond to environmental stimuli, such as hypoxia and hypercapnia, by  adjusting cardiorespiratory processes. Generally ventilation and heart rate increase under these conditions (Wang and Hicks, 1996b; West and Van Vliet, 1992). Both the hypoxic ventilatory response (HVR) and hypercapnic ventilatory response are highly variable among species (Milsom, 1990; Milsom, 1998). Jackson (1978) suggested that the lack of uniform responses between studies of various species of reptiles could be derived from differences in the role or relative sensitivities of chemoreceptor groups. The sensitivity of peripheral chemoreceptors to natural stimuli does not only vary between species, but it is also affected by external factors such as temperature (Glass et al., 1983). It has also been shown recently that the sensitivity of chemoreceptors to natural stimuli may vary between day and night (McArthur and Milsom, 1991; Woodin and Stephenson, 1998). It is possible, therefore, that differences in response between different studies may also reflect the time of day or season during which the measurements were made. We have previously shown the existence of endogenous circadian and circannual oscillations in the metabolism and ventilation of red-eared sliders (Trachemys scripta elegans) at rest (Fig. 1.5) (Reyes and Milsom, 2010). Daily and seasonal oscillations in ventilation were 17  accompanied by circadian and circannual changes in the breathing pattern leading to longer apneas at night and during colder seasons, which could reduce the risk of predation, cost of locomotion and potentially cost of breathing during dormancy. Little is known about the mechanisms that generate the arrhythmic breathing pattern of reptiles, but peripheral input is known to be important in determining the breathing pattern of episodic breathers (Shelton et al., 1986). Input from peripheral chemoreceptors conveys information regarding the level of arterial blood gases. Thus, these receptors may play an important role in initiating and terminating nonventilatory periods by adjusting ventilation and perfusion to meet metabolic demands, which will result in changes in the breathing pattern (Shelton et al., 1986). Few studies have addressed the role that biological rhythms play in the control of respiration in reptiles. The purpose of this part of my thesis was to determine whether the daily and seasonal changes in the breathing pattern of turtles either resulted from changes in the sensitivity of chemoreceptors (plasticity) or were modulated solely by daily and seasonal oscillations in metabolism.  1.9  Model species All the species I selected (bullfrogs, red-eared sliders and rattlesnakes) exhibit intermittent  breathing patterns where single breaths or breathing episodes are interspaced by a nonventilatory period (NVP) or apnea (Wood and Lenfant, 1976; Burggren et al., 1977; Shelton and Boutilier, 1982; Shelton et al., 1986; Milsom, 1988; Boutilier, 1990). During the NVP arterial blood oxygen tension falls and CO2 tension rises (PCO2) (reviewed by Shelton and Boutilier, 1982; Burggren and Shelton, 1979). The duration of the NVP is quite variable and the mechanism that controls the NVP length remains unknown (Shelton and Boutilier, 1982; Milsom, 1991). It has been proposed, however, that peripheral chemoreceptor input is important in establishing the breathing pattern characteristic of amphibians and reptiles (Shelton et al., 1986). Furthermore, all of these species have an undivided ventricle that allows for central cardiovascular shunts (Burggren, 1987; Farrell et al., 1998; Hicks, 1998; Hicks, 2002). The pattern and degree of shunt (R-L or L-R shunt and their magnitude) will influence the levels of arterial blood gases (Wang and Hicks, 1996a; Wang et al., 1997). However, the involvement of arterial chemoreceptors in shunt regulation remains largely unknown. Peripheral chemoreceptor input appears to be particularly important for cardio-respiratory control in lower tetrapods. It is 18  therefore surprising that most of the research on peripheral arterial chemoreceptors has focused on fish and mammals, while studies on amphibians and particularly reptiles are limited. Thus, one of the goals of my thesis was to determine the location, innervation and neurochemical content of chemosensory areas in these groups to better understand how peripheral arterial chemoreceptors have evolved in vertebrates. In anurans and chelonians functional chemosensory areas have been identified (Ishii et al., 1966; Ishii et al., 1985a, b; Ishii and Ishii, 1986), but virtually nothing is known regarding putative neurotransmitters that may participate in signal transduction, the type of stimulus that triggers a response and whether the responses to stimulation of different chemoreceptor populations involve respiratory or cardiovascular adjustments. Given that functional chemosensory areas have been identified in anurans and chelonians, members of these groups (Rana catesbeiana and Pseudemys scripta elegans) provide me with a good foundation to study peripheral arterial chemoreceptors and to test techniques that will allow me to locate and characterize these receptors in other reptiles (Crotalus durissus). I chose to work specifically with bullfrogs (Rana catesbeiana) because this species was readily available and the presence of biogenic amines in its carotid labyrinth had been demonstrated by exposure to formaldehyde vapor (Kobayashi, 1971a). It is important to point out that in amphibians most research has been directed towards the study of the carotid labyrinth, while other chemosensory areas have been largely overlooked. Red-eared sliders (Pseudemys scripta elegans) were specifically selected, since this was the species used to determine the effects of circadian and seasonal cycles in the sensitivity of arterial chemoreceptors. This species occupies ponds over most of the Eastern United States, as well as Central America, the Greater Antilles and the northern South American countries (Harles and Morlock, 1978); their distribution ranges up to a latitude of 50 ° N (Ultsch, 1989). Red-eared sliders are primarily diurnal (Harles and Morlock, 1978). Some temperate populations of redeared sliders undergo dormancy from October until March or April, when the water temperature falls below 10ºC (Ernst and Barbour, 1989). They generally spend the winter submerged in a body of water or buried in the mud (Cagle, 1950), where they may become hypoxic or even anoxic (Ultsch, 1989). Given the established daily pattern and seasonality of this species I could  19  investigate the plasticity of the reflex response of peripheral arterial chemoreceptors in two different time frames, daily and seasonally. Given the lack of knowledge about peripheral chemoreceptors in all remaining groups of reptiles (snakes, lizards and crocodilians), studies on any of these groups are novel and interesting. However, I used snakes since peripheral arterial chemoreceptors have not been located in this group and because their anatomy makes them good candidates to test the relative roles of distinct chemosensory areas. The long distance between the carotid and the aortic and pulmonary chemoreceptors allows for better isolation of the chemoreceptor groups while testing for their reflex roles and stimulus sensitivities. I specifically used rattlesnakes (Crotalus durissus) because they have been well described in terms of shunting patterns (and can produce a large R-L shunt (Wang et al., 1998)) and they were readily available in Brazil where these experiments were conducted. The former was particularly important to investigate the effects of peripheral chemoreceptor stimulation on shunt regulation and therefore in the control of arterial blood gases.  1.10 Research questions and objectives of the thesis The overall objective of my thesis was to determine whether O2-sensing structures are highly conserved among vertebrates while their locations (if there are multiple O2-sensing sites), stimulus modalities, and sensitivities in reptiles and amphibians have changed to more effectively control ventilatory and cardiovascular functions. My first objective was to determine if chemosensory areas in lower tetrapods are homologous to those of other vertebrates. Since peripheral arterial chemoreceptors have already been located in anurans and chelonians, but not in other reptiles, my goal was to locate potential chemosensory areas in snakes (Crotalus durissus) based on homology of arterial supply and innervation by the IX and X cranial nerves. I then determined if these areas were functional chemoreceptors by stimulating each site with focal injections of NaCN, which mimics anoxia, and selectively denervating each area to corroborate that the response originated from chemoreceptor stimulation at the specific site (Chapter 4). Furthermore, I characterized putative O2-sensing cells in anurans and two groups of reptiles (turtles and snakes) so that homologies 20  with those of other vertebrates could be derived. I specifically determined whether putative oxygen sensing cells within the chemosensory areas contained neurotransmitters that may participate in O2 signal transduction. I used immunohistochemical markers for ACh, TH and 5HT, which are neurotransmitters present in the carotid bodies of mammals and fish (Chapters 2, 3 and 4). In addition, retrograde tract tracing (neuronal tracer cholera toxin B, CTB) and immunohistochemistry were used together to determine the anatomical relation between oxygen sensing cells and branches from the IX and X cranial nerves (Chapters 2 and 3). I hypothesized that oxygen chemosensory areas occurred at homologous sites across the vertebrates and contained oxygen sensing cells with similar properties (neurotransmitter content, innervation and arrangement), suggesting that oxygen sensing mechanisms have been conserved across vertebrates. The following objectives focused on establishing the physiological characteristics of peripheral arterial chemoreceptors, their role in cardio-respiratory control and the plasticity of the chemoreflex they produce. In particular, my goal was to determine whether distinct populations of chemoreceptors sense oxygen tension, oxygen content, or both (stimulus modality) and establish the role of different chemoreceptor populations in ventilatory regulation and cardiovascular control in rattlesnakes (Chapter 4). I hypothesized that, as in mammals, oxygen chemoreceptors in areas that are well perfused will be stimulated only by changes in oxygen tension, while oxygen chemoreceptors in areas that are poorly perfused or exposed to venous blood will also sense changes in oxygen content. Specifically, those located in areas homologous to the carotid bifurcation of mammals will sense changes in Po2 while those located in the walls of the aorta and pulmonary artery will also sense changes in Cao2. Correlating receptor responses to vascular anatomy will integrate questions 1 and 2 of this thesis (see Summary of research questions below). I predicted that reptiles would have multiple chemosensory sites that are architecturally designed to enhance sensitivity to different stimulus modalities, which should be advantageous for reptiles that regulate blood gases by changes in both ventilation and/or the degree of cardiac shunt. As in mammals, chemoreceptor groups that sense PO2 will primarily regulate ventilation while those that are stimulated by low CaO2 will adjust the cardiovascular system. To determine the stimulus modality of distinct chemosensory areas, I compared the responses of each 21  chemosensory site after stimulation with a bolus of saline (normal PO2 and low CaO2) and a bolus of blood (normal PO2 and CaO2). To establish the presence and reflex roles of specific chemosensory areas, I injected NaCN as close to the site as possible and recorded the time required to initiate a response and the nature of the response (ventilatory or circulatory). Furthermore, to confirm that I had located afferent fibers innervating chemosensory areas, the specific branch of the IXth /Xth cranial nerve to that area was severed followed by a second injection of NaCN. My last objective consisted of determining whether there is plasticity in the reflex response these receptors produced. I tested if the hypoxic-hypercapnic sensitivity of chemoreceptors changed within two time frames (time of the day and time of the year), leading to daily and seasonal changes in the breathing pattern of red-eared sliders independent of changes in metabolism (Chapter 5). To address this objective I exposed turtles to seasonal environmental conditions over a one-year period. I measured oxygen consumption, ventilation and breathing pattern continuously for 24 h in animals breathing air and a hypoxic-hypercapnic gas mixture in each season, after two days of acclimation to the experimental conditions (mean seasonal temperature and natural photoperiod). Oxygen consumption and ventilation were used to calculate the air convection requirement (ACR= V̇ E/V̇ O2, ml air/ml O2) for each hour. Respiratory sensitivity was calculated as the change in total ventilation from breathing air to breathing a hypoxic-hypercapnic gas mixture. In summary I have addressed the following questions and objectives in this dissertation: 1. Are the oxygen sensing structures of amphibians and reptiles homologous to mammalian carotid and aortic bodies? Objective 1: locate peripheral chemoreceptors in frogs, turtles and snakes through their innervation by the glossopharyngeal and vagus nerves (Chapters 2, 3 and 4). Objective 2: Establish whether these putative chemosensory areas are functional in snakes (Chapter 4). Objective 3: Characterize putative oxygen sensing cells and determine whether they contain neurotransmitters that may participate in signal chemotransduction to sensory 22  nerves in frogs, turtles and snakes (immunohistochemistry and neuronal tracer; Chapters 2, 3 and 4). 2. What stimuli are these chemoreceptors sensing? Objective 1: Determine if chemosensory areas are stimulated by changes in oxygen tension or oxygen content (Chapter 4). 3. What is the reflex role of distinct groups of chemoreceptors? Objective 1: Determine the relative reflex role (respiratory or cardiovascular control) of distinct populations of chemoreceptors (Chapter 4). 4. Are chemoreflexes palstic? Objective 1: Determine whether there are circadian and/or circannual rhythms in chemosensitivity (Chapter 5).  23  Figure 1.1. Distribution of O2-sensing chemoreceptors in vertebrates. VII, IX and X refer to facial, glossopharyngeal and vagus nerves. 2-6 refer to arteries supplying the respective embryonic pharyngeal arches in each group. Red, yellow and blue colors represent O2 chemosensory areas homologous to derivatives of the 3rd, 4th and 6th pharyngeal arches, respectively (modified from Van Vliet and West, 1987).  24  Figure 1.2. Schematic explaining the pressure and washout shunt hypotheses for R-L cardiac shunts in reptiles. RA, right atrium; LA, left atrium; CP, cavum pulmonale; CA, cavum arteriosum; CV, cavum venosum; Lao, left aortic arch; RAo, right aortic arch; PA, pulmonary artery (Modified from Hicks and Malvin, 1995).  25  Figure 1.3. Change in arterial oxygen tension as a result of increasing ventilation, eliminating the R-L shunt or a combination of increased ventilation and elimination of R-L shunt (Wang and Hicks, 1996a).  26  Figure 1.4. Pulmonary (Q̇ pul) and systemic (Q̇ sys) blood flows in turtles with varying hematocrit in normoxia, during apneas (NVP) and during ventilation (VP) (Wang et al., 1997).  27  Figure 1.5. Mean (± s.e.m.) day (unfilled bars) and night (filled bars) oxygen consumption (A) and total ventilation (B) for outdoor turtles during different seasons measured at mean seasonal temperatures but natural photocycle (summer: 20.8ºC, 16L : 8D; fall: 14.7ºC, 10L : 14D; winter: 9ºC, 9L : 15D; spring: 14.6ºC, 14L : 10D). Comparison of day and night values of oxygen consumption (C) and ventilation (D) for outdoor turtles exposed to the natural photoperiod (photocycle values are the same as the fall values shown in A and B) and to constant darkness (constant dark and no daily thermal cycle) in the fall season. An asterisk denotes significant differences between day and night; significant differences between seasons are denoted by different letters (Holm Sidak multiple comparison method; for summer, spring, P < 0.05 N=6 and constant dark, N=8 for winter and fall). (Reyes and Milsom, 2010).  28  Table 1.1 Summary of the location, innervation, neurochemical content, distribution and organization of chemosensory areas in vertebrates. Location refers to the site where chemosensory areas are found. Distribution refers to the location of oxygen sensing cells within the chemosensory areas. Organization refers to the arrangement of cells (solitary or in clusters). IX, glosspharyngeal nerve; X, vagus nerve; ph a/aa, pharyngeal arch/aortic arch; Innervat, innervation cb, carotid body; ab, aortic body; Ch, chapter (refers to the chapters of the thesis where some of the gaps were addressed. Group  Fish  Amphibians  Turtles  Lizards  ph a/aa 3rd 4th 5th 6th  chemosensory areas 1st gill arch 2nd gill arch 3rd gill arch 4th gill arch orobranchial cavity  Innervat IX, X X X X V, VII  3rd 4th 6th 3rd 4th 6th 3rd 4th 6th  carotid labyrinth aortic arch pulmocutaneous artery (not confirmed) common carotid artery aortic arch pulmonary artery carotid bifurcation (not confirmed) not known not known  IX, X X X IX X X IX/X  Ch 4  Ch 4  not known  X  common carotid artery aortic arch not chemosensory carotid bifurcation (carotid bodies) aortic arch (aortic bodies) not chemosensory  X X  Snakes Crocodilians Birds  Mammals  3rd 4th 6th 3rd 4th 6th  IX X  Neurochemicals  Distribution  Organization  5HT (mainly) catecholamines?  widely distrib.  mostly solitary  not specific Ch 2  Ch 2  groups 2-3 cells Ch 2  not specific Ch 3  Ch 3  Ch 3  not specific  not known  not known  Ch 4  Ch 4  Ch 4  not known  not known  not known  5HT, catecholamines Ach, ATP, 5HT, catecholamines, etc  widely distrib. clusters conglomeration widely distrib.  clusters (1000+, cb) clusters (5-50, ab)  29  Chapter 2: Distribution and innervation of peripheral arterial chemoreceptors in the bullfrog (Rana catesbeiana) 2.1  Summary Peripheral arterial chemoreceptors have been located in the carotid labyrinth, aortic arch and  the pulmocutaneous artery of frogs. However, the putative neurotransmitters associated with these chemoreceptors have not yet been described. The goal of the present study was to determine the neurochemical content, innervation and distribution of putative oxygen sensing cells in frogs and to derive homologies with peripheral arterial chemoreceptors of other vertebrates. I used retrograde tract tracing together with immunohistochemical markers for acetylcholine (ACh), tyrosine hydroxylase (TH, the rate limiting enzyme in catecholamine synthesis) and serotonin (5HT) to identify putative oxygen sensing cells and to determine their anatomical relation to afferent branches of the glossopharyngeal and vagus nerves (IX and X cranial nerves, respectively). I found potential oxygen sensing cells in all three areas innervated by branches of the X cranial nerve. Only the carotid labyrinth, however, was innervated by the IX cranial nerve. Cells containing either 5HT or TH were found in all three sites, but cells containing both neurotransmitters were only found in the carotid labyrinth. The morphology and size of these cells resemble glomus cells found in other amphibians and mammals, and the 5HTcontaining cells are neural crest derived, providing further evidence for the oxygen sensing function of these cells. The presence of 5HT- and TH-immunoreactive cells in the regions of the aorta, pulmocutaneous artery and the carotid labyrinth appear to reflect a transition between the major neurotransmitters seen in fish (5HT) and mammals (ACh, adenosine and catecholamines).  2.2  Introduction Vertebrates can ensure an adequate supply of oxygen to their tissues under hypoxia through  adjustments in ventilation and perfusion triggered by peripheral arterial chemoreceptors which monitor the levels of oxygen in the blood (Shelton et al., 1986; Gonzalez et al., 1994; Milsom, 1998). These chemoreceptors for the most part consist of glomus cells (mammals) derived from neural crest cells (Pearse et al., 1973; Jones and Milsom, 1982; Lahiri et al., 1983; Gonzalez et al., 1994) or neuroepithilial cells (NEC’s, fish) (Dunel-Erb et al., 1982; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz, et al. 2004). At low oxygen tensions (PO2) these cells 30  release neurotransmitter(s) that stimulate afferent nerve endings (Gonzalez et al., 1994; Bunn and Poyton, 1996; Prabhakar, 2000; Peers and Kemp, 2001; Prabhakar, 2006) (see Chapter 1 for details on oxygen sensing mechanisms). Afferent chemosensory information is relayed by these nerves to the nucleus of the solitary tract (NTS) where it is integrated (Stuesse et al., 1984). Ultimately physiological adjustments are produced by effector organs, so that oxygen supply and demand are matched (Nurse, 2005). This system is critical as a first step to ensure survival during hypoxia. Peripheral chemoreceptors have been found in derivatives of the pharyngeal arches of all vertebrates studied to date and are commonly innervated by the glossopharyngeal (IX) and/or vagus (X) nerves (Adams, 1958; Ishii et al., 1966; Rogers, 1967; Abdel-Magied and King, 1978; Lahiri et al., 1981; Ishii et al., 1985a; Ishii et al., 1985b; Ishii and Ishii, 1986; de Graaf, 1990; Gonzalez et al., 1994; Jonz and Nurse, 2003, 2009). The distribution of these chemosensory areas, however, varies widely among vertebrates and the phylogenetic trend seems to be a reduction in the number of sites from fish to mammals (Milsom and Burlenson, 2007). While oxygen sensing cells in fish are present in all gill arches (Dunel-Erb et al., 1982; Zaccone et al, 1992; Burleson and Milsom, 1993; Jonz and Nurse, 2003; Jonz et al., 2004, Jonz and Nurse, 2009), in mammals only the carotid (derivative of the 3rd pharyngeal arch) and aortic (derivative of the 4th pharyngeal arch) bodies are chemosensory (Lahiri et al., 1983; Gonzalez et al., 1994). Amphibians are intermediate, with oxygen sensing sites located in derivatives of the 3rd, 4th and possibly 6th pharyngeal arches (Adams, 1958; Ishii et al., 1966; Ishii et al., 1985a; Ishii and Ishii, 1986). In amphibians, hypoxia elicits an increase in ventilation, which arises primarily by stimulation of peripheral chemoreceptors in the carotid labyrinth (Boutilier and Toews, 1977; Van Vliet and West, 1992; Wang et al., 1994). The carotid labyrinth is a maze-like vascular bed that arises from the carotid artery (Toews et al., 1982; Kusakabe, 2002, 2009). It is considered homologous to a portion of the first gill arch of fishes and to the carotid body of mammals, based on its embryonic origin and innervation by branches of the IX and X cranial nerves (Adams, 1958; Ishii et al., 1966; Van Vliet and West, 1987b; West and van Vliet, 1992; Kusakabe, 2002; Milsom and Burlenson, 2007; Kusakabe, 2009). Stimulation of the carotid labyrinth fails to produce baroreceptor reflexes, indicating that this area is only chemosensory (Ishii et al., 1966; Hoffmann and de Souza, 1982). Bilateral denervation of the carotid labyrinth, however, does not 31  abolish the chemosensory response (Van Vliet and West, 1986; West et al., 1987) and two other areas, the aortic arch and pulmocutaneous artery have been found to be chemosensory (Lillo, 1980; Hoffmann and de Souza, 1982; Ishii et al., 1985a; Wang et al., 2004), although the evidence for the pulmocutaneous artery is equivocal. These areas originate from the 4th and 6th pharyngeal arches and are innervated by a pharyngeal branch of the X cranial nerve and a branch of the recurrent laryngeal nerve (RLN), respectively. In contrast with the carotid labyrinth, the aorta and pulmocutaneous artery are important barosensory areas (Lillo, 1980; West and Van Vliet, 1983; Ishii et al., 1985a; Van Vliet and West, 1987a; West and Van Vliet, 1992). The ultrastructure of the carotid labyrinth shows similarities with the carotid body of mammals. Glomus cells with dense-cored vesicles, partially surrounded by sustentacular cells and in close proximity to afferent nerve endings have been identified (Ishii and Oosaki, 1969; Kobayashi, 1971b; Ishii and Kusakabe, 1982; Kusakabe, 1991). As many aspects of the carotid labyrinth of anurans have been well documented, it is surprising that so little is known about their putative neurotransmitters and their role in oxygen sensing. Evidence of the occurrence of catecholamines in the carotid labyrinth has been shown by formaldehyde vapor exposure (Banister et al., 1967; Kobayashi, 1971a). While this technique is non-specific, it has been used to identify biogenic monoamines. Glomus cells in the aorta show similar characteristics to those of the carotid labyrinth (Ishii et al., 1985a). However, the innervation, organization and neurochemical content of putative oxygen sensing cells in the aorta and pulmocutaneous artery have not been described in frogs. In mammalian glomus cells, a plethora of neurotransmitters such as catecholamines (dopamine, adrenaline and noradrenaline), acetylcholine (ACh), ATP, Substance P and serotonin (5HT) are believed to participate in signal chemotransduction (Nurse, 2005, 2010), while the Neuroepithelial cells of fish mainly secrete 5HT (Dunel-Erb et al., 1982; Bailly et al., 1992; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz et al., 2004; Bailly, 2009; Jonz and Nurse, 2012). Thus, establishing the neurotransmitter content and expression of enzymes that participate in neurotransmitter synthesis in amphibians may allow us to further explore homologies between chemoreceptor cells in the different vertebrate groups. The objectives of the present study were to characterize the neurochemical profile and innervation of the three chemosensory areas previously identified in frogs (Ishii et al., 1966; Ishii et al., 1985a), and to determine the distribution patterns of putative oxygen sensing cells in these 32  three areas. I hypothesized that within these chemosensory areas I would find sparsely distributed glomus cells containing a similar complex of neurotransmitters to those found in mammalian carotid bodies. To test this hypothesis I used markers for ACh which has been suggested to be the main neurotransmitter to participate in signal transduction in mammalian carotid body; tyrosine hydroxylase (TH), the rate limiting enzyme in catecholamine synthesis (Nurse, 2005); 5HT which is known to be the main neurotransmitter in NEC’s (Dunel-Erb et al., 1982; Bailly et al., 1992; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz et al., 2004; Jonz and Nurse, 2012) and Human natural killer-1 (HNK-1), a marker for proliferative cells that develop from neural crest cells. Furthermore, retrograde tract tracing and immunohistochemistry were used together to determine the anatomical relation between putative oxygen sensing cells and branches from the IX and X cranial nerves. These studies will help in understanding the homologies between neurochemical function and anatomy in amphibians in relation to other, better studied vertebrate groups.  2.3  Materials and methods  2.3.1  Animals and holding conditions  Bullfrogs (Rana catesbeiana, Shaw) (mean mass 0.25± 0.03 kg) were obtained from a commercial supplier (Boreal Northwest, St. Catharines, ON, Canada) and were housed indoors in fiberglass tanks continuously supplied with running water and areas to bask. Frogs were kept at room temperature (~20°C) and under a 12L:12D photoperiod. Animals were fed live king worms three times a week. Frogs were fasted for a week prior to surgery. The holding and experimental procedures followed Canadian Council of Animal Care guidelines and were approved by the University of British Columbia Animal Care Committee (animal care certificate No. A09-0233).  2.3.2  Vascular and nerve anatomy  Dissections were made to identify branches of the IX and X cranial nerves that innervate the pulmocutaneous artery, aorta and carotid labyrinth. The nerve branches and vasculature were visualized using a stereomicroscope (Leica M125, Leica Microsystems, Nussloch, Germany) and digital images were acquired with a camera (Leica DFC295, Leica Microsystems, Nussloch,  33  Germany) attached to the microscope and Leica software Application Suite (LAS V3.6, Leica Microsystems, Nussloch, Germany).  2.3.3  Tissue sections  Cross sections of the carotid labyrinth and longitudinal sections of the pulmocutaneous artery and aorta were used to confirm their innervation by branches of the IX and X cranial nerves using neuronal tract tracing (Cholera Toxin B (CTB, List Biological laboratories, distributed by Cedarlane Laboratories, Hornby, ON, Canada) and Wheat-germ agglutinin (WGA, Vector laboratories, Burlington, ON, Canada)). Afferent and efferent fibres of the IX and X cranial nerves in the central vasculature were visualized using antibodies against CTB and WGA. Antibodies against TH, 5-HT and vesicular acetylcholine transporter (VAChT) were also used. These neurotransmitters are known to participate in signal transduction in the mammalian carotid body (Gonzalez et al., 1994). The HNK-1 antibody was used to identify cells that originate from the neural crest and are proliferative. Cells that are innervated by the IX and/or X cranial nerves, that contain all or some of the neurotransmitters mentioned above and are derived from neural crest cells, are good candidates as oxygen sensing cells.  2.3.4  Surgery and tissue preparation  Frogs were anaesthetised by immersion in Ethyl 3-aminobenzoate methanesulfonate (MS222, 1.5 g/L buffered to pH 7.0; Sigma-Aldrich, USA). Frogs were divided into two experimental groups. An incision was made behind the tympanic membrane and the vagus nerve (Group 1; N=4) or the glossopharyngeal nerve (Group 2; N= 5) were carefully isolated from the jugular vein, carotid artery and surrounding connective tissue. The nerves were injected 2-3 times (1-2 µl total) posterior to the jugular ganglion in the head with CTB (0.75%) and WGA (5%) in distilled water using a 2 µl Hamilton Syringe 7002KH (Reno, NV, USA) with a glass micropipette on the needle. Only one side of the animal was injected with the neuronal tracer, such that the innervations on the other side of the animal served as an internal control. Tissue on the injected and uninjected side of the animal was used to determine the neurochemical content of putative chemosensory areas (N=11). Upon recovery from anesthesia, animals were given an analgesic (intraperitoneal injection of Xylazine (10 mg/kg)). The neuronal tracer was allowed to travel in the injected nerve for 5-7 days after which the animals were deeply anaesthetised with 34  MS-222 and perfused transcardially with heparinised saline (100UI/ml) using a blunt 21 gauge needle connected to a peristaltic pump until the blood vessels appeared clear of blood (400-500 ml). Tissue of interest was then fixed by perfusing the animal with 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer saline (PBS; Na2HPO4, 13.4 g/l; NaH2PO4, 6 g/l; NaCl, 9 g/l; buffered to pH of 7.4 with NaOH). The carotid labyrinths and segments of the pulmocutaneous arteries and aortae were dissected out from the base of the heart to beyond the vessel constriction. To confirm that the neuronal tracer did not spread from the injection site, the IX, X and hypoglossal (XII) nerves from both the injected and non injected sides were collected. The adrenal gland, jugular ganglia and lungs of some of the animals were also collected to use as positive immunohistochemical controls (see Controls section). The collected tissue was post-fixed for 24 h in 4% PFA at room temperature, cryoprotected in 30% sucrose buffer and frozen in Tissue Tek (Sakura, San Marcos, CA, USA) at -80 °C. Twenty micrometer thick longitudinal (aorta and pulmocutaneous artery, (Fig. 2.1A) or cross sections (carotid labyrinth) were made using a cryostat (CM3050, Leica Microsystems, Nussloch, Germany) and serially mounted on Superfrost plus slides (VWR International, West Chester, PA, USA). Slide mounted sections were immediately processed for immunohistochemistry or stored in a -80 ºC freezer until needed.  2.3.5  Immunohistochemistry  Slide-mounted tissue was washed in PBS and blocked in 10% normal goat serum (NGS) or normal donkey serum (NDS) (Jackson Laboratories, distributed by Cedarlane Laboratories, Hornby, ON, Canada), depending on the host-species of the secondary antibody, for 30 minutes. Primary antibodies were diluted (PBS, 0.3% Triton X-100, 1-2% NGS or NDS) according to optimal dilutions previously determined (Table 2.1). Slides were incubated with the primary antibody (individually or in combination) for 48 h at room temperature and then washed in PBS. Following the washes, slides were incubated in the dark with biotinylated or fluorescently labelled secondary antibodies diluted in PBS (with 0.3% Triton X-100 and 1-2% NGS or NDS) (Table 2.1) for 2 h and subsequently washed in PBS. Slides treated with biotinylated secondary antibodies were incubated with a fluorescently conjugated Avidin, Neutravidin for 1 h and washed in PBS. Cell nuclei were labelled with DAPI (Vectashield with DAPI, Vector Laboratories, Burlington, Ontario). Coverslips (#1.5, Fisher Scientific, Ottawa, ON, Canada) 35  were mounted with Vectashield (Vector Laboratories, Burlington, Ontario) to prevent photobleaching and then the coverslips were sealed with nail polish. Processed slide-mounted tissue was stored at 4 ºC in the dark until imaging.  2.3.6  Controls  Controls consisted of excluding the primary antibody to control for non-specific binding of the secondary antibody and of excluding the secondary antibody in cases where a biotinylated secondary antibody was used, to control for non-specific binding of Avidin-Neutravidin reaction. To control for interactions between antibodies single labelled slides were processed in each run. Positive controls for TH and VAChT primary antibodies were performed using the adrenal gland and jugular ganglion respectively. The lungs of frogs in which the vagus nerve had been injected with neuronal tracers were used as positive controls for the primary antibodies against CTB, WGA and 5-HT. The specificity of the primary antibodies was tested by examining the labeling pattern of two different antibodies raised in the same host species. For instance, monoclonal antibodies for HNK-1 and TH raised in mouse and anti goat-CTB, anti goat WGA compared to the anti goat5HT antibody. Furthermore, the specificity of these antibodies had been verified by the suppliers by either using a preabsorption control or Western blots. Specific labeling of cells in the rat brain using the rabbit anti-5HT polyclonal antibody was inhibited by preincubation of diluted antiserum with 500 µM 5HT. Western blots were probed using the mouse anti-TH monoclonal antibody. TH protein of human embryonic kidney cells was identified by a 60 kDa immunoreactive band. See Table 2.1 for antibody and supplier details. Additionally, the specificity of CTB and WGA injections were established by immunohistochemically processing injected nerves (vagus or glossopharyngeal), as well as nearby not-injected nerves (hypoglossal and not injected vagus or glossopharyngeal nerves).  2.3.7  Microscopy for cryo-sectioned tissue  Slides were observed using an epifluorescence light microscope (Axioplan 2, Zeiss, Jena, Germany) equipped with a G365, HQ470 or BP546/12 excitation filter and a LP420, HQ525/50 or LP590 emission filter to detect DAPI, FITC and Cy-3 respectively. Images were captured using a Q-Imaging CCD camera (Model Retiga 1300, Burnaby, BC, Canada) and QCapture 36  image capture software (Version 2.95.0). Immunolabelled slides were also viewed and analysed with an FSX100 (Olympus, Center Valley, PA, USA), equipped with a single-panel color CCD pixel shift type camera. Representative sections were further examined using a spinning disk microscope (Perkin Elmer Ultraview VOX Spinning Disk Confocal, Waltham, MA, USA), equipped with 405 nm, 440 nm, 488 nm and 561 nm lasers and filters 527/55, 445/60 or 615/70, 525/50 or 640/120 for detection of GFP, DAPI and rhodamine. Z-stacks of 80-106 optical sections and 0.19-0.25 µm apart were captured using Leica multi-immersion 20x and 63x glycerol objectives and a Hamamatsu C9100-50 camera. Further image analysis was achieved using ImageJ.  2.3.8  Whole mount tissue preparation  Whole mounts of the carotid labyrinth, pulmocutaneous artery and aorta were used to further demonstrate the distribution of putative oxygen sensing cells along the blood vessel, as a function of distance from the heart to the lung or from the heart to the systemic circulation. As well, I wanted to parse the distribution of putative chemosensors through the depth of the blood vessel (from lumen to basal surface). The vascular tissue was prepared for whole mounts as indicated below. Frogs (N=2) were euthanized using a lethal dose of MS-222 and pithed. The truncus arteriosus, including the spiral valve, left aorta, pulmocutaneous artery and carotid labyrinth were dissected free from each frog all together and transferred into a sylgard bottom dish containing ice cold Rana Ringer solution (NaCl 75 mM, KCl 4.5 mM, MgCl2 1 mM, NaH2PO4 1mM, NaHCO3 40 mM, CaCl2 2.5 mM, glucose 5 mM). Each vessel was isolated, cut open longitudinally and pinned flat with the lumen of the vessel exposed (Fig. 2.1B). The orientation of each vessel with respect to its position relative to the heart was marked. The lumen of each vessel was rinsed with heparinised Rana Ringer solution (0.1 IU/ml) to clear any blood remaining in the blood vessel. The vascular tissue was then fixed with 4% PFA in 0.1 M PBS (see Tissue sections for details) for 12 h at 4ºC.  2.3.8.1  Immunohistochemistry for whole mounts  After fixation whole mounts were washed in PBS containing 5% NDS for 30 minutes. A polyclonal antibody against serotonin (rabbit a-5-HT, 1:100 concentration, Sigma-Aldrich, USA) 37  and a monoclonal antibody against tyrosine hydroylase (mouse a-TH, 1:100 concentration, Immunostar Inc, Hudson, WI, USA) were diluted in PBS containing 0.1% Triton X-100 and 5% NDS. Whole mounts were incubated with the primary antibodies for one day at room temperature while gently agitated using an orbital shaker (Cole Parmer, Vernon Hills, IL, USA) and one day at 4ºC. Subsequently, whole mounts were washed in multiple changes of PBS (5% NDS) for 30 minutes and incubated in secondary antibodies in the dark (1:100 concentration; donkey anti-rabbit Alexa Fluor® 488 (A21206) and donkey anti-mouse Alexa Fluor® 594 (A21204); Molecular Probes, OR, USA) diluted in PBS containing 0.1% Triton X-100 and 5% NDS for one day at room temperature and one day at 4ºC. Whole mounts were washed in multiple changes of PBS (5% NDS, no detergent) for 30 minutes and placed on double bridged slides in a small drop of PBS. Coverslips were placed over the surface of the whole mount and adhered by applying silicon grease on the mounting bridges and sealing the tissue containing compartment with melted paraffin wax. Slight pressure on the coverslip was required to sufficiently flatten the blood vessel.  2.3.8.2  Microscopy for whole mounts  3-D images of the flattened vessels were obtained with an Olympus Multiphoton microscope, using an excitation beam of 800 nm. Emissions were filtered using a wide green filter with a band pass of 495-540 nm and a red filter with a band pass of 570-675 nm, which eliminated any green fluorophore emission bleed through into the red photodetector. Images were captured with a 25x long-working length, high numerical aperture water immersion objective (NA 1.1), enabling adequate image resolution throughout the approximately 100 micron thickness of each whole mounted vessel. The gain and sensitivity for the red and green photodetector channels were the same. Image stacks were collected in sequence starting at the end of each blood vessel closest to the heart and proceeding towards the head or torso.  2.3.9 2.3.9.1  Quantification Cell size  To further describe different cell populations, including candidate chemosensory cells, I measured cell diameters, along the longest axis of the cell in tissue sections from 4-5 animals. Only cells where I could clearly see the nucleus and where I could see the entire labeled 38  cytoplasm surrounding the nucleus were measured. In each animal 10-15 5HT- and THimmunoreactive (IR) cells were measured in different sections of the carotid labyrinth, aorta and pulmocutaneous artery using Volocity imaging software Version (6.1.2) (Perkin Elmer Inc., Waltham, MA, USA) and ImageJ. I corroborated these results by measuring a number of cells in whole mounts using ImageJ. All data are presented as average cell size (µm) ± s.e.m.  2.3.9.2 2.3.9.2.1  Cell distribution Calculations of cell density in the carotid labyrinth  I used serial cross sections of the carotid labyrinth to determine the cell distribution from the outer edge towards the inside of this structure (Fig. 2.7, Insert) Montages of 4-6 sections from each animal (n=5 animals, 25 sections) were made using the stitching function in an FSX100 microscope (Olympus, Center Valley, PA, USA). The imaged sections represented a sample of the whole blood vessel. Cells from different populations in each section were counted and the tissue area calculated using ImageJ. Due to the complex maze-like structure of the carotid labyrinth, the vascular area was calculated by thresholding the image and converting the pixels to a binary image, so that the areas of the lumen sinusoids could be subtracted from the total area. Data are presented as mean cell density ± s.e.m. Linear regressions were used to determine if cell density changed with increased distance from the outer edge (0 µm) of the carotid labyrinth (R version 2.11.1, R development core team, 2010).  2.3.9.2.2  Calculations of cell density in the aorta and pulmocutaneous artery  The occurrence of 5HT and particularly catecholamine containing cells is very irregular throughout the aorta and pulmocutaneous artery making it difficult to quantify the cells in sectioned tissue. Therefore, I used whole mounts to more accurately determine the cell distribution along these blood vessels from the heart to the lung or systemic circulation using the thresholding and cell counter functions in ImageJ.  2.3.10 Haematoxylin and eosin histology Haematoxylin and Eosin (H&E) labeled cross sections were used to visualize the internal structure of blood vessels in the central vasculature and to determine the general location of putative chemoreceptor cells. Longitudinal sections at 20 µm thick or cross sections at 6 µm 39  thick, with two sections skipped in between, were cut through six regions of the central vasculature (Fig. 2.1C) using a Leica rotary microtome (RM2255, Leica Microsystems, Nussloch, Germany). Sections were mounted onto slides. Slides were deparaffinised and rehydrated by running them through changes in xylene (Fisher scientific, Ottawa, ON, Canada), ethanol (100%, 95%, 90%, 80%, 70%; A407, Fisher scientific, Ottawa, ON, Canada) and water. Slides were stained in Gill’s #2 hematoxylin (3801520, Leica Microsystems Inc., Buffalo Grove, IL, USA), then washed with running tap water, rinsed in 1% acid alcohol and washed again. Subsequently the slides were stained in 1% lithium carbonate (Sigma Aldrich, Oakville, ON, Canada), washed with tap water, stained in 2% aqueous eosin (stock solution: Sigma Aldrich, Oakville, ON, Canada) and washed again before dehydrating back to xylene through serial solutions of ethanol. Coverslips (#1, Fisher Scientific, Ottawa, ON, Canada) were mounted with permount (SP15-500, Fisher Scientific, Ottawa, ON, Canada). The H&E stained cross sections were prepared by Wax-it Histology services, Vancouver, B.C.  2.3.11 Scanning electron microscopy To demonstrate the complex structure of the frog carotid labyrinth, resin casts suitable for scanning electron microscopy (SEM) were made. Batson's #17 plastic (Polysciences, Inc., Catalog # 07349, Warrington, PA, USA) was prepared according to standard instructions. In euthanized animals (N=2, as described above), the end of the carotid artery was cut, anterior to the carotid labyrinth and the vessel was flushed with heparinised saline. Batson's resin was then injected into each carotid artery, and the resin was allowed to set. The injected vessel was dissected free and soaked in 1 M NaOH for 48 h which removed most of the tissue. Subsequently the residual tissue was removed with collagenase (Sigma Aldrich, C9891, Oakville, ON, Canada). The carotid labyrinth resin casts were then sputter coated with gold (Cressington Sputter coater 208 HR) and SEM images were collected with a Hitachi S-2600N Variable Pressure Scanning Electron Microscope.  2.4  Results  2.4.1  Anatomy and innervations of the central vasculature  In the bullfrog, the truncus arteriosus leaves the heart and splits on each side into a pulmocutaneous artery and an aortic arch, which immediately gives rise to a common carotid 40  artery. The later gives rise to the carotid labyrinth (Fig. 2.2A). These three areas are derivatives of the pharyngeal arches. The X cranial nerve gives rise to small branches that innervate the aorta and carotid labyrinth (Fig. 2.2B) and to the RLN, which innervates the pulmocutaneous artery before looping underneath it and ascending towards the larynx. The carotid labyrinth is also innervated by a branch of the IX cranial nerve.  2.4.2  Controls for immunohistochemistry  In frogs, chromaffin cells in the adrenal glands labeled for TH (Fig. A1-B), and neuroepithelial bodies and vagal nerve fibers in the lungs labeled for serotonin (rabbit anti-5HT) and CTB, respectively (Fig. A1-A), Furthermore, the labeling patterns of the HNK-1 and TH primary antibodies raised in mouse were very different, as were the reactivity patterns of the anti -CTB and WGA antibodies raised in goat compared to that of the goat anti-5HT antibody, when tested in the carotid labyrinth, aorta and pulmocutaneous artery of frogs. Although, I did not have this type of specificity control for the rabbit anti-5HT antibody in frogs, the pattern of labeling was consistent with that of the goat anti-5HT antibody. In addition I confirmed the specificity of the rabbit anti-5HT antibody in the corresponding tissues of turtles and snakes (see Chapters 3 and 4). In all immunohistochemical runs omitting the primary antibody and the biotinylated secondary antibody (when used) resulted in no detectable labeling, indicating that all antibodies were likely specific in frogs. I did not find any discernible differences in labeling pattern between single and double labeled slides. I confirmed that neuronal tracer microinjections were site specific as I was able to visualize both CTB and WGA in the injected nerves, but not in nearby nerves (data not shown). These controls corroborate that the procedures and antibodies used in this study were effective and likely specific in frogs.  2.4.3  Potential chemosensory areas  I found putative oxygen sensing cells in the carotid labyrinth, aorta and pulmocutaneous artery. These cells were innervated by the IX and\or X cranial nerves and contained 5HT and TH (indicative of catecholamines). I did not find cells containing VAChT in any of the areas. VAChT, however, labeled neurons in the jugular ganglion (Fig. 2.3) indicating that this antibody works well in this species.  41  2.4.3.1  Carotid labyrinth  The carotid labyrinth is a swelling of the carotid artery that gives rise to the external and internal carotid arteries, which deliver oxygenated blood to the head (Fig. 2.4A). Cross sections of the carotid labyrinth show the maze-like internal structure, composed of numerous lumen sinusoids (Fig. 2.4B, C). I found numerous putative oxygen sensing cells arranged singly or in clusters of 2-4 cells in the intravascular stroma. The cells were distributed close to the lumen pockets (Fig. 2.4D).  2.4.3.1.1  Neurochemical content of the carotid labyrinth  I found four distinct populations of cells differentiated based on their neurochemical content, morphology and size. 1) A population of 5HT-IR cells was distributed throughout the stroma of the carotid labyrinth. These cells were oval with large nuclei and a cell diameter of 11.55 ± 0.14 um (Fig. 2.5A). Some of the serotonergic cells colocalized with HNK-1 (Fig 2.6A). 2) There was also a population of small TH-IR cells (Fig. 2.5B), which were pyramidal-like, sometimes with visible projections (Fig. 2.5C). Their size (11.64 ± 0.13 um) and distribution was similar to that of 5HT-IR cells. 3) A population of large (19.52 ± 1.39 um) TH-IR cells with relatively small nuclei (Fig. 2.5D) appeared less frequently than the other two populations and were always located away from the lumen sinusoids (see ∗ labeled structure in Fig. 2.4B). 4) A group of cells that contained both 5HT and TH that resembled the 5HT-IR cells in shape and size were also distributed throughout the stroma of the carotid labyrinth. These cells were less common than singly labeled 5HT and TH cells (2.5E).  2.4.3.1.2  Cell density in the carotid labyrinth  The four cell types varied in density in the carotid labyrinth. Across all animals, 5HT (49.1 cells/mm2) and TH (38.6 cells/mm2) cells were more abundant, while large TH-IR cells (12.4 cells/mm2) were scarce. Only in a few instances 5HT and TH were found in the same cell (colocalized, 4.2 cells/mm2). The density of the cells was unrelated to their position relative to the outer edge of the blood vessel in three out of the four cell types (Fig. 2.7A, C, D). The density of 5HT cells increased near to the center of the carotid labyrinth (density = 26.7 + 0.01*distance, P=0.04, R2=0.24) (Fig. 2.7B).  42  2.4.3.1.3  Innvervation of the carotid labyrinth  The tracer studies confirmed that the carotid labyrinth is innervated by the IX and X cranial nerves. Two of the four cell types described above (5HT and small TH) are innervated by these nerves. Nerve fibers of the IX and X colocalize with 5HT- (Fig. 2.8A, C) and TH-containing cells (Fig. 2.8B, D). Large TH cells did not show any apparent innervation. Since my experiments used up to three labels (DAPI, the neuronal tracer and either 5HT or TH), I could not determine if the population of cells containing both 5HT and TH were innervated. All throughout the carotid labyrinth, I also found cells that were either surrounded by nerve fibers or that had taken up the neuronal tracer (Fig. 2.8B, arrow) but that did not colocalize with the markers for serotonin or catecholamines.  2.4.3.2  Aorta  The internal structure of the aorta lacks the complexity and tortuousity of the carotid labyrinth (Fig 2.9A). However, longitudinal sections of the blood vessel revealed the presence of some lumen compartments (Fig 2.9A, B). I found two populations of cells embedded in the vessel wall with similar size and morphology (Fig. 2.9B-D). Both were oval in shape, TH-IR cells were 12.0 ± 0.4 µm long and 5HT-containing cells measured 12.2 ± 0.23 µm. Some serotonergic cells projected into the lumen compartments as well (Figs. 2.9B and 2.10B). 5HTcontaining cells colocalized with HNK-1 (Fig 2.6B). Both cell types occurred singly or in clusters of 2-3 cells, never colocalized and were distributed throughout the length (from the heart to the systemic circulation) of the blood vessel. Higher densities of serotonergic cells were observed at the wall thickening (constriction) where the artery arches caudally and no cells were found distal to this point (Fig. 2.11, top panel). The distribution of TH-IR cells was irregular and sporadic making it difficult to quantify them. I confirmed the vagal innervation to the aorta using the neuronal tracer CTB. Nerve fibers containing the tracer were observed near the lumen (Fig 2.9C) and innervating the vessel wall. In a few instances, I observed nerve fibers directly innervating putative chemosensory cells (Fig 2.9D, Insert).  2.4.3.3  Pulmocutaneous artery  The internal structure of the pulmocutaneous artery is less complex than that of the aorta (Fig. 2.12A). I found two cell populations in this area. One type of cells labeled for 5HT (12.1 ± 43  0.3 µm) and the other for TH (12.5 ± 0.3 µm). The morphology of the cells was almost identical to those cells in the aorta being embedded in the vessel wall in singles or clusters (Figs. 2.12B and 2.13B). Serotonergic cells colocalizad with HNK-1 (Fig 2.6C). They were distributed throughout the length of the pulmocutaneous artery, but were more abundant in the spongy tissue by the distal pulmocutaneous constriction where the vessel narrows before entering the lung (Figs. 2.11, bottom panel and 2.13). As in the aorta, catecholamine-containing cells appeared sporadically throughout the blood vessel so that I could not quantify TH cell distribution. I confirmed the innervation of the pulmocutaneous artery by the vagus using CTB. Nerve fibers labeled with the neuronal tracer innervated areas along the lumen and within the vessel wall. Both 5HT- and TH-IR cells were also innervated by the X cranial nerve (Fig. 2.12C, D).  2.5  Discussion I identified several likely oxygen sensing cells in bullfrogs distributed throughout the carotid  labyrinth, aortic arch and pulmocutaneous artery. My results show the presence of catecholamines and 5HT, but not ACh, in these putative oxygen sensing cells. These cells are all innervated by the X cranial nerve, while cells in the carotid labyrinth are also innervated by the IX cranial nerve. Serotonergic cells in all three areas seem to be derived from neural crest cells. Putative oxygen sensing cells appear very similar in morphology in all three chemosensory areas, although their distribution and the occurrence of some cell types differ between regions. My study supports the chemosensory function of the pulmocutaneous artery and supports the hypothesis that the carotid and aortic chemosensory areas in frogs are homologous to the carotid and aortic bodies in mammals.  2.5.1  Putative neurotransmitters in oxygen sensing cells  In the present study, I found cells containing either 5HT or catecholamines in the carotid labyrinth, aorta and pulmocutaneous artery of Rana catesbeiana. Cells with both 5HT and catecholamines appear sporadically only in the carotid labyrinth. 5HT and catecholamines (adrenaline, noradrenaline and dopamine) amongst other neurotransmitters have been shown to participate in oxygen chemotransduction in the carotid body of mammals (Gonzalez et al., 1994; Nurse, 2005, 2010). Glomus cells in mammals are derived from APUD (amine-precursor uptake and decarboxylation) cells that have the ability to take up specific amino acids and decarboxylate 44  them to produce the corresponding amine (catecholamines and serotonin, amongst others) and store them in granules (Pearse, 1969). The presence of cells containing either catecholamines (Rana temporaria, Rana catesbeiana, Bufo vulgaris) or 5HT (Rana nigromaculata, Bufo bufo japonicas) has previously been demonstrated in the carotid labyrinth and aorta of amphibians using formaldehyde vapor fluorescence (Banister et al., 1967; Kobayashi, 1971a; Ishii et al., 1985a). This technique is not very specific, however, and does not distinguish well between 5HT and catecholamines. For instance, optimal fluorescence of 5HT requires strong reaction conditions, which then cause diffusion of the catecholamine fluorescence. In addition, at high concentration, the peak emission of catecholamine fluorescence will be very similar to that of 5HT (Corrodi and Jonsson, 1967). This has made it very difficult to demonstrate the presence of both amines in the same tissue and may explain why only 5HT or catecholamines have been so far reported in individual studies of the chemosensory areas of amphibians. Interestingly, although 5HT has been reported in the carotid labyrinth of a couple of amphibian species (Kobayashi, 1971a), the function of this transmitter has not been investigated and the focus has been mainly on the role of catecholamines in afferent and efferent regulation of the carotid labyrinth. The implications of my findings are that not only catecholamines, but also 5HT may be involved in oxygen sensing. The contribution of multiple neurotransmitters in signal transduction also occurs in the glomus cells in the carotid bodies of mammals, with the distinction that while all transmitters are found in the same cells in mammals, different cell types, some containing 5HT and some catecholamines, appear to contribute to this process in amphibians. Despite finding multiple neurotransmitters in the chemosensory areas of bullfrogs, I did not see any evidence of ACh in these regions. I did observe, however, that the marker for ACh labels neuronal bodies in the jugular ganglia, demonstrating that the antibody is effective in this species. Historically 5HT and catecholamines were thought to be the main neurotransmitters involved in hypoxic signal chemotranduction in the NEC’s of fish (Dunel-Erb et al., 1982; Jonz and Nurse, 2003) and in the carotid body of mammals (Gonzalez et al., 1994; Nurse 2005; Prabhakar, 2006). Recent studies, however, have proposed that 5HT alone predominates as the principle neurotransmitter involved in O2-sensing in fish (Jonz and Nurse, 2009, 2012) and that ACh co-released with ATP plays the crucial role in signal chemo-transduction in mammals (Eyzaguirre and Zapata, 1984; Zhang et al., 2000; Nurse, 2005; Shirahata et al., 2007; Nurse, 45  2010). Given the apparent importance of ACh to oxygen sensing in higher vertebrates, its absence from the chemosensory areas of bullfrog is surprising. Other studies have also failed to find any evidence of cholinergic cells in the carotid labyrinth and aorta of amphibians, but injections of ACh in these regions have been shown to cause an increase in chemosensory discharge (Ishii and Ishii, 1967, Ishii et al., 1985a) and a decrease in the internal carotid outflow (Kusakabe et al., 1987). Thus, it appears that ACh may participate in efferent control in amphibians by modulating chemoreceptor sensitivity and vascular tone (West and Van Vliet, 1992). Similar results have been found in fish (Burleson and Milsom, 1993; reviewed by Porteus et al., 2012).  2.5.1.1  Putative oxygen sensing cells in the carotid labyrinth of frogs  I identified four populations of cells in the carotid labyrinth of bullfrogs that differ from each other in their neurochemical content, morphology and/or distribution within the carotid labyrinth. Serotonergic cells, small TH-IR cells and cells where both neurotransmitters colocalize are present, in singles or clusters, throughout the vascular stroma of the carotid labyrinth, not far from lumen sinusoids. The former two cell types are innervated by the IX and X cranial nerves, which implies a role as primary sensors. Cells where both 5HT and TH colocalize are likely to be innervated as well, but I was limited to a maximum of two markers in addition to DAPI in my immunohistochemical procedures. Glomus cells, resembling those in the carotid body of mammals have been observed in the vascular stroma of the carotid labyrinth of Bufo marinus, Bufo vulgaris, Xenopus laevis. (Rogers, 1963; Ishii et al., 1966; Ishii and Oosaki, 1969; Ishii and Kusakabe, 1982). These cells are oval (12 µm diameter) with large nuclei and occasionally processes extend from the cell (Rogers, 1963). The presence of densecored vesicles suggests a secretory function (Ishii et al., 1966; Ishii and Oosaki, 1969; Kobayashi, 1971b). The size and distribution of these cells correspond with that of the 5HT- and small TH-IR cells (11.6 µm and 11.6 µm for 5HT and small TH cells respectively) I identified in bullfrogs. Their morphology, however, more resembles that of 5HT cells, although I never see processes projecting from 5HT cells. In my study processes are characteristic of small catecholamine-containing cells. Several lines of evidence support the chemosensory role of the cells described above. These include their proximity to lumen sinusoids, presence of monoamines, such as 5HT and 46  catecholamines, innervation by the IX and X nerves and the structural similarities to glomus cells of other amphibians and mammals (Ishii et al., 1966; Gonzalez et al., 1994; West and Van Vliet, 1992; Kusakabe, 2002; Nurse, 2005: Kusakabe, 2009). In addition to the chemoreceptor-like anatomical features of TH-IR cells that I describe in my study, others have shown a functional role of catecholamines in chemoreception (Ishii and Ishii, 1967). Injections of catecholamines in the carotid labyrinth of toads affect chemoreceptor discharge and it appears that epinephrine strongly stimulates it, while dopamine inhibits it (Van Vliet and West, 1992), as is the case in the carotid body of mammals (Gonzalez, et al., 1994). Given the anatomical and functional evidence it is likely that small TH-IR cells in this study and that of others (Banister et al., 1967; Kusakabe, 1990) are involved in oxygen chemoreception. The role of 5HT cells in oxygen sensing is more elusive as its effect on afferent nerve discharge has not been investigated in the chemosensory areas of amphibians. Given the lack of functional evidence, I relied on anatomical features to support the notion that these cells are chemosensory (see above). Further evidence is given by the presence of HNK-1 on the surface of 5HT cells. HNK-1 is a marker for proliferative neural crest cells (Hou and Takeuchi, 1994; Clark et al., 2001; Kundrat, 2008; Reyes et al., 2010). Glomus cells in the carotid bodies of mammals and birds have been shown to be derived from neural crest cells (Pearse et al., 1973; Kondo et al., 1982), in fact it has been suggested that the ultrastructural similarities shared by all APUD cells are due to their common origin from neural crest cells (Pearse, 1969). Thus, the origin, morphology and innervation of serotonergic cells also place them as plausible candidates for oxygen sensing. Electrophysiological studies are required to fully substantiate that 5HT- and TH-IR cells are chemoreceptors, but taken together, my data and that of others strongly suggest their involvement in chemosensing. However, I cannot determine if these cells act directly at the reciprocal synapse of the afferent nerve (primary oxygen sensor) or have a paracrine function instead. An alternative scenario that cannot be ruled out is that the primary oxygen sensors are instead the cells where 5HT and catecholamines colocalize. The morphology, distribution and density of these cells correlate with glomus cells described in other amphibians. Ishii et al. (1966) reported the presence of glomus cells in three out of ninety six sections of the carotid labyrinth. Similarly, Ishii and Kusakabe (1982) reported only fifty glomus cells in one carotid labyrinth. Cells containing both, 5HT and TH, also appear sporadically in the carotid labyrinth 47  of bullfrogs (on average 4.2 cells/mm2), whereas cells containing only 5HT or TH are much more numerous (49.1 cells/mm2 and 38.6 cells/mm2, respectively). If this is the case and only cells containing multiple neurotransmitters are the oxygen sensors, then this raises the question of why the other two cell types are so abundant. A possibility is that 5HT- and TH-containing cells have functions other than reflex cardiorespiratory chemoreception. In addition to this reflex role (Ishii et al., 1966), the carotid labyrinth regulates vascular tone and glomus cells are involved in vascular regulation by acting on smooth muscle cells (Kusakabe et al., 1987). The mechanism by which vascular control from glomus cells is achieved involves modulation of chemoreceptor sensitivity by efferent regulation (West and Van Vliet, 1992). The increase in blood flow in the carotid labyrinth caused by excitation of chemoreceptors during hypoxia is regulated by release of acetylcholine from nerve endings onto receptors in the glomus cell. This triggers the release of catecholamines (Banister et al., 1967; Ishii and Kuskabe, 1982) from glomus cells onto smooth muscle cells causing vasoconstriction (Kusakabe et al., 1987). It is possible that small TH-IR cells have a dual function in regulation of vascular tone and chemoreception, or it may be that TH-only containing cells are vasoactive, while cells holding multiple neurotransmitters are chemosensory. In addition to the three populations of cells present in the vascular stroma of bullfrogs, I find a fourth cell type that occurs sporadically in the carotid labyrinth and is always located in clusters away from the lumen sinusoids. These cells are large (19.5 µm) and oval with relatively small nuclei and immuno-reactive for catecholamines (large TH-IR cells). Their location and the fact that they are not innervated, suggest that they are not involved in reflex cardio-respiratory regulation. This population of cells has not been described before.  2.5.1.2  Putative oxygen sensing cells in the aorta and pulmocutaneous artery of frogs  In bullfrogs I find cells containing either 5HT or catecholamines that are embedded in the tunica media of the aorta and are innervated by the X cranial nerve. Their size (12.0 and 12.2 µm in diameter for TH- and 5HT-IR cells, respectively) and morphology are similar to that of oval glomus cells (12 µm in diameter) found in the smooth muscle area of the aorta of toads (Ishii et al., 1985a). The presence of catecholamines, specifically adrenaline, in this area has been demonstrated using formaldehyde vapor exposure and chromatography techniques, although at lower concentrations than in the carotid labyrinth (Banister et al., 1967; Ishii et al., 48  1985a). Contrary to my findings, it was suggested that adrenaline was not found in cells, but present in nerve endings found in the adventitia and media of the aorta (Banister and Mann, 1965). To my knowledge this study is the first to show the presence of 5HT cells in this region. Since the aortic arch is mainly chemosensory, with a minor role in baroreception (Ishii et al., 1985a) it is likely that the 5HT- and TH-IR cells that I find in this area are involved in oxygen sensing. However, the sporadic occurrence of putative oxygen sensing cells in the aorta compared to that of the carotid labyrinth implies that the latter have a more prominent role in chemoreception (West and Van Vliet, 1992). Given that the aorta is chemosensory and I do not find cells were 5HT and catecholamines colocalize, I could argue that the 5HT and small THonly containing cells in the carotid labyrinth are chemosensory and not just cells where both neurotransmitters colocalize. My data also supports the chemosensory function of the pulmocutaneous artery, as has been proposed previously (Hoffmann and de Souza, 1982; Wang et al., 2004). NaCN injections in the pulmocutaneous artery of lightly anesthetized frogs stimulate ventilation when frogs are breathing and cause bradycardia when they are submerged (Lillo, 1980). This study does not provide direct evidence for the presence of peripheral chemoreceptors in the pulmocutaneous artery of amphibians, since no denervations were carried out to ensure that the reflex responses were produced by stimulation of chemoreceptors in that particular area and were not due to recirculation. Thus, the presence of peripheral chemoreceptors in the pulmocutaneous artery has not been proven. In any case, I find cells embedded in the blood vessel wall, which contain either 5HT or catecholamines. These cells are innervated by small branches from the RLN, a branch of the vagus. They are the same size (12.5 and 12.1 µm in diameter for TH- and 5HT-IR cells, respectively) and shape as cells found in the aorta and carotid labyrinth of bullfrogs (present study) and toads (Ishii and Kusakabe, 1982; Ishii et al., 1985a). As discussed above both cell types are likely to be chemosensory. My findings give additional evidence of the chemosensory nature of the pulmocutaneous artery.  2.5.2  Distribution of putative oxygen sensing cells  The stimulus specificity of different chemoreceptor groups has been a matter of discussion for the past two decades. In mammals, the carotid bodies respond to changes in arterial oxygen tension (PaO2), but not in arterial oxygen content (CaO2), while aortic bodies are sensitive to both. 49  Perfusion of the aortic chemoreceptors is low relative to their oxygen consumption and thus their PO2 will be affected by changes in oxygen content, blood flow and hemoglobin affinity (Lahiri et al., 1981). The location of oxygen sensing cells relative to blood supply is presumably an important determinant in their stimulus modality. In this study, 5HT-, small TH-IR cells and cells with both neurotransmitters, in the carotid labyrinth of bullfrogs are in close proximity to the lumen sinusoids. This is also the case for glomus cells in the carotid labyrinths of Xenopus and toads (Ishii and Oosaki, 1969; Ishii and Kusakabe, 1982). I speculate that the location of these cells is optimal to sense changes in oxygen partial pressure (PO2), and in fact, it has been shown that chemoreceptors in the carotid labyrinth of toads are sensitive to PO2, but not to CaO2 (Van Vliet and West, 1992; Wang et al., 1994). The stimulus modality of the other two chemosensory areas in amphibians has not been established but may also include changes in CaO2, as putative oxygen sensing cells are embedded in the vessel wall, further away from the lumen, with the exception of some serotonergic cells in the aorta. However, independent from their location relative to the lumen, the discharge of pulmocutaneous chemoreceptors is likely to correlate with a reduction in oxygen content. These receptors are perfused by mixed venous blood and a fall in CaO2 will reduce venous oxygen tension (PvO2) and oxygen content (CvO2) (Wang et al. 1997; Wang et al., 2004). This chemoreceptor group will be therefore more suitable for monitoring oxygen carrying capacity (Lahiri et al., 1981) than carotid chemoreceptors.  2.6  Conclusion My data allows me to compare the distribution of oxygen sensing cells in amphibians with  those of other, better-studied vertebrate groups. This study is the first to show the presence of both catecholamine- and 5HT-IR cells in the carotid labyrinth, aortic arch and pulmocutaneous artery of bullfrogs and of cells containing both neurotransmitters in the carotid labyrinth. With the exception of the pulmocutaneous artery, these areas have been showed to be functional chemoreceptors. Based on distribution, cell morphology, innervation and neurochemical content, I propose that the three cell types present in these areas are likely to function as oxygen sensing cells. Catecholamine release has been shown to increase sensory nerve discharge, supporting the notion that TH-IR cells are involved in chemotransduction. The role of 5HT in oxygen sensing has not been established in amphibians, but their neural crest origin, anatomical features and similarities to glomus cells in other amphibians imply their role as chemosensory 50  cells. Additionally, this study gives further evidence for the chemosensory function of the pulmocutaneous artery, as I found the same cell types in this area as in the carotid labyrinth and aorta. The presence of putative oxygen sensing cells in derivatives of pharyngeal arches, their innervation by the IX and X cranial nerves, the presence of multiple neurotransmitters in these cells, together with similarities in the ultrastructure of glomus cells in other vertebrates suggest that oxygen sensing structures are highly conserved among vertebrates, but the number of chemosensory areas has changed in amphibians to better regulate arterial blood gases.  51  Figure 2.1. Schematics showing the orientation of longitudinal tissue sections and whole mounts (A and B). Location in the central vasculature of Rana catesbeiana where cross sections for histological analysis were made (C). A: Orientation of longitudinal sections, where the blood vessel was cut sagittally. B: Orientation of whole mounts. The blood vessel was cut open longitudinally and pinned flat with the lumen of the vessel exposed. C: Location in the central vasculature of Rana catesbeiana where cross sections were made for histological analysis. 1) Proximal truncus arteriosus with spiral valve; 2) Distal truncus arteriosus with spiral valve; 3) Junction of the pulmocutaneous artery and aorta; 4) carotid labyrinth midsection; 5) internal carotid artery distal to the carotid labyrinth; 6) aorta midsection 7) Pulmocutaneous artery midsection. Lao, left aorta; LPCa, left pulmocutaneous artery; Lcl, left carotid labyrinth; Leca, left external carotid artery; Lica, left internal carotid artery and tra, truncus arteriosus.  52  Figure 2.2. Picture (A) and schematic (B) showing the anatomy and innervation of the central vasculature of Rana catesbeiana. Scale bar 1 mm. Ao, aorta; Pca, pulmocutaneous artery; cl, carotid labyrinth; eca, external carotid artery; ica, internal carotid artery; IX, glossopharyngeal nerve; X, vagus nerve; RLN, recurrent laryngeal nerve.  53  Figure 2.3. Positive controls for vesicular acetylcholine transporter (VAChT) and the neuronal tracer Cholera toxin B (CTB) in the Jugular ganglia of Rana catesbeiana. Triple immunolabeling for VAChT (green), CTB (red) and a nuclear stain (DAPI, blue) in the jugular ganglia. The images from the green and red channels are shown separately and as a merged image with DAPI (blue). Colocalization appears yellow. B: Double immunolabeling for VAChT (red) and cell nuclei (DAPI, blue) in the Jugular ganglia of R. catesbeiana. Scale bars 10 µm.  54  Figure 2.4.Carotid labyrinth of Rana catesbeiana. A: Scanning electron micrograph of a vascular resin cast of the carotid labyrinth of Rana catesbeiana. View shows the vascular maze (mz) and origin of the internal carotid artery. Scale bar 1 mm. B: Montage of a cross section (20 µm thick) of the carotid labyrinth showing the central chamber (cch) and vascular stroma comprised by the lumen sinusoids (l). Serotonin (5HT, green) and tyrosine hydroxylase (TH, red) immunoreactive cells appear throughout the vascular stroma of the carotid labyrinth. Scale bar 145 µm. C: Hematoxylin and Eosin stained cross section (6 µm thick) of the carotid labyrinth showing the lumen sinusoids (l), endothelial cells (e), smooth muscle cells (sm) and melanocytes (m). Scale bar 25 µm. D: Cross section (20 µm thick) of the vascular stroma of the carotid labyrinth showing triple immunolabeling for 5HT (green), TH (red) and a nuclei stain (DAPI, blue). Inserts show close ups of 5HT- and TH-positive cells. Scale bar 25 µm.  55  Figure 2.5. Cell populations present in the carotid labyrinth or R. catesbeiana. A: Serotonin immunoreactive cell (5HT-IR, green) in the vascular stroma of the carotid labyrinth. B: Cluster of tyrosine hydroxylase immunoreactive cells (TH-IR, red) in the vascular stroma of the carotid labyrinth. C: TH-IR cells with projections. D: A second population of large TH-IR cells in the periphery of the carotid labyrinth. E: Cells that contain both 5HT (green) and TH (red). The images from the green and red channels are shown separately and as a merged image with DAPI (blue). Colocalization appears yellow. Scale bars 10 µm.  56  Figure 2.6. Triple immunolabeling for 5HT (red), human natural killer-1 (HNK-1, green) and a nuclear stain (DAPI, blue) in the carotid labyrinth (A), aorta (B) and pulmocutaneous artery (C) of Rana catesbeiana. The images from the green and red channels are shown separately and as a merged image with DAPI (blue). Colocalization appears yellow. HNK-1 is present in the surface of some serotonin immunoreactive cells, suggesting that these cells are derived from neural crest cells or are proliferating. Scale bars 10 µm.  57  Figure 2.7. Cell density as a function of distance of the section from the outer edge of the carotid labyrinth (0 µm, insert) for (A) TH-IR cells, (B) 5HT-IR cells, (C) large TH-IR cells and (D) 5HT and TH-IR cells. The color of the symbol indicates different animals. The solid regression line in (B) indicates a significant trend (P=0.04, R2=0.24). (cl) carotid labyrinth.  58  Figure 2.8. Innervation of the carotid labyrinth by the glossopharyngeal (IX) and vagus (X) nerves, shown by the presence of nerve fibers containing the neuronal tracer CTB. A: 5HTpositive cell (green) innervated by a branch of the IX nerve (CTB, red). B: TH-positive cell (green) innervated by a branch of the IX nerve (CTB, red). Nerve fibers containing CTB are also found surrounding cells that are not immunoreactive for serotonin or tyrosine hydroxylase (arrow). C: 5HT cells (green) innervated by a branch of the X nerve (CTB, red). D: TH cell (green) innervated by the X nerve (CTB, red). The images from the green and red channels are shown separately and as a merged image with DAPI (blue). Scale bars 11 µm.  59  Figure 2.9. Cell populations and vagal innervation (X) in the aorta of R. catesbeiana. A: Hematoxylin and Eosin stained longitudinal section (20 µm thick) of the aorta showing pockets in the wall of the blood vessel (*). B: Tyrosine hydroxylase (TH, red) and serotonin immunoreactive cells (5HT, green) in the vessel wall of the aorta, close to a lumen pocket (*). C: Vagal fibers containing the neuronal tracer Cholera toxin B (CTB, red) in the vicinity of 5HT-IR cells (green). D: Vagal fibers containing CTB (red) in close proximity to the lumen and innervating a TH-IR cell (green, arrow and insert). (l) lumen. The images from the green and red channels are shown separately and as a merged image with DAPI (blue). Scale bars 10 µm.  60  Figure 2.10. Distribution of serotonergic cells in the aorta (whole mount) of R. catesbeiana. A: Oval 5HT-IR cells (purple) lining the vessel wall project into the lumen (arrows, B, digital resection of stack). 2-D compression of an image stack apprx. 80 µm thick. Serotonin positive cells are sparse and found in small clumps close to the lumen of the vessel. C: Diagram of the position of this region. D and E: Haematoxylin and Eosin staining for the aorta at 4.2 and 25X respectively. Scale bars 50 µm.  61  Figure 2.11. Distribution of serotonergic cells in the aorta and pulmocutaneous artery of R. catesbeiana. Top panel: Montage of an aorta whole mount showing 5HT cell distribution from the heart (left) towards the systemic circulation (right). The table shows the cell density for each panel. Serotonin immunoreactive cells are distributed throughout the blood vessel, but they appear at higher density at the wall thickening. Bottom panel: Montage of a pulmocutaneous artery whole mount showing 5HT cell distribution from the heart (left) towards the lung (right). Serotonin immunoreactive cells are sparsely distributed throughout the blood vessel length. The table shows the cell density for each panel.  62  Figure 2.12. Cell populations and vagal innervation (X) in the pulmocutaneous artery of R. catesbeiana. A: Hematoxylin and Eosin stained longitudinal section (20 µm thick) of the pulmocutaneous showing the blood vessel wall. B: Tyrosine hydroxylase (TH, red) and serotonin immunoreactive cells (5HT, green) embedded in the vessel wall of the pulmocutaneous artery. C: Vagal fibers containing the neuronal tracer Cholera toxin B (CTB, red) in close proximity to a 5HT cell (green). D: Vagal fibers (CTB, red) surrounding a lumen pocket and in close proximity to TH-IR cells (green). E: Higher magnification of panel D, showing TH-IR cells (green). The images from the green and red channels are shown separately and as a merged image with DAPI (blue). Scale bars 10 µm.  63  Figure 2.13. Serotonergic cells in the pulmocutaneous artery constriction (whole mount) of R. catesbeiana. A: Serotonergic "glomus-like" cells (purple) are densely packed in a thickened muscular region of the pulmocutaneous artery, just prior to the constriction of the vessel (2-D compression of an image stack). B: The 5 HT-IR cells are embedded in the vessel wall throughout a depth of 87 microns (digital re-section of the image stack at the position of the arrows). C: Diagram of the position of this region. D: Haematoxylin and Eosin cross-section of this approximate region, 4.2X. Scale bars 50 µm.  64  Table 2.1 Primary and secondary antibodies used for immunohistochemistry on tissue sections Antisera  Antigen  Manufacturer  Host  Dilution  Cat. No.  Secondary antisera1  Primary 5-HT  serotonin  Sigma-Aldrich  rabbit  1:350  S5545  TH  tyrosine hydroxylase  Immuno Star  mouse  1:250 (cl) 1:150 (pc,da)  22141  VAChT  Sigma-Aldrich  rabbit  1:250  V5387  HNK-1  vesicular acetylcholine transporter CD-57  FITC a-rabbita biotinylated a-rabbitg Cy-3 a-mousee Alexa Fluor® 568h biotinylated amousef Alexa Fluor® 488b  BD Pharmigen  mouse  1:200  559048  Alexa Fluor® 488c  anti-CTB  Cholera Toxin B subunit  List Biologicals laboratories  goat  1:1500  703  Alexa Fluor® 568d  anti-WGA  Wheat germ agglutinin  Vector laboratories  goat  1:100  AS-2024  Alexa Fluor® 568d  Secondary1 FITC  rabbit IgG (H+L)a  Cedarlane Labs  donkey  1:200  CLAS101047 A-21206 A-21202 A-11057 20079-521001 BAF018 BA-1000 A-11004  --  Alexa Fluor® 488 rabbit IgG (H+L)b Molecular Probes, Invitrogen donkey 1:200 -c Alexa Fluor® 488 mouse IgG (H+L) Molecular Probes, Invitrogen donkey 1:200 -Alexa Fluor® 568 goat IgG (H+L)d Molecular Probes, Invitrogen donkey 1:200 -Cedarlane labs donkey 1:200 -Cy-3 mouse IgG (H+L)e f* Biotinylated mouse IgG Cedarlane labs donkey 1:200 -Biotinylated rabbit IgGg* Vector laboratories goat 1:500 -h Alexa Fluor® 568 mouse IgG (H+L) Molecular Probes, Invitrogen goat 1:200 -1 Secondary antisera were conjugated with a fluorescent marker a-h Secondary antisera antigen corresponds with primary antibody host *Biotinylated secondary antibodies were visualized with avidn NeutrAvidin®, Oregon green® 488 conjugate (1:200, A6374, Molecular probes)  65  Chapter 3: Distribution and innervation of peripheral arterial chemoreceptors in the red-eared slider (Pseudemys scripta elegans) 3.1  Summary Peripheral arterial chemoreceptors have been located in the common carotid artery, aorta  and pulmonary artery of turtles. However, the putative neurotransmitters associated with these chemoreceptors have not yet been described. The goal of the present study was to determine the neurochemical content, innervation and distribution of putative oxygen sensing cells in the central vasculature of turtles and to derive homologies with peripheral arterial chemoreceptors of other vertebrates. I used tract tracing together with immunohistochemical markers for acetylcholine (ACh), tyrosine hydroxylase (the rate limiting enzyme in catecholamine synthesis) and serotonin (5HT) to identify putative oxygen sensing cells and to determine their anatomical relation to branches of the vagus nerve (X cranial nerve). I found potential oxygen sensing cells in all three chemosensory areas innervated by branches of the X cranial nerve. Cells containing either 5HT or vesicular acetylcholine transporter (VAChT) were found in all three sites. The morphology and size of these cells resemble glomus cells found in amphibians, mammals, tortoises and lizards. Furthermore, I found populations of cholinergic cells located at the base of the aorta and pulmonary artery that are likely involved in efferent regulation of vessel resistance. None of the cholinergic or serotonergic cells colabeled with the Human natural killer-1 marker, indicating that these cells are neither derived from the neural crest nor mature, non-proliferative cells. The presence of 5HT- and VAChT-IR cells in segments of the common carotid artery, aorta and pulmonary artery appear to reflect a transition between cells containing the major neurotransmitters seen in fish (5HT) and mammals (ACh and adenosine).  3.2  Introduction Regulation and maintenance of blood gas homeostasis during periods of hypoxia or  hypoxemia is critical for the survival of vertebrates. Specialized oxygen sensing cells, such as neuroepithilial cells (NEC’s) in the gills of fish (Dunel-Erb et al., 1982; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz, et al. 2004) or glomus cells in the carotid and aortic bodies of mammals (Lahiri et al., 1983; Gonzalez et al., 1994) and in the carotid labyrinth, aorta and pulmocutaneous artery of frogs (Ishii et al., 1966; Lillo, 1980; Hoffmann and de Souza, 66  1982; Ishii and Kusakabe, 1982; Ishii et al., 1985a; Wang et al., 2004) (see Chapter 2) are responsible for the immediate response to changes in oxygen levels. The mechanisms by which chemoreceptor cells sense oxygen remain unknown (Lahiri, 2000; Prabhakar, 2000, 2006). It is well established, however, that at low oxygen tension (PO2), background K+ channels close and the cell membrane of glomus cells depolarizes causing Ca+2 dependent neurosecretion, which stimulates afferent nerve endings (Gonzalez et al., 1994; Bunn and Poyton, 1996; Prabhakar, 2000; Peers and Kemp, 2001; Prabhakar, 2006) (see Chapter 1 for details on oxygen sensing mechanisms). Chemosensory information is relayed by the glossopharyngeal and/or vagus nerves (IX and X cranial nerves, respectively) to the nucleus of the solitary tract (NTS) in the brainstem where it is integrated (Stuesse et al., 1984). Physiological adjustments in ventilation and perfusion are produced to ensure an adequate supply of oxygen to the tissues (Nurse, 2005). In all vertebrates studied so far, peripheral arterial chemoreceptors are associated with derivatives of the pharyngeal arches and are innervated by the IX and/or X cranial nerves. The distribution and number of chemosensory areas, however, varies among vertebrate taxa (Adams, 1958; Ishii et al., 1966; Rogers, 1967; Abdel-Magied and King, 1978; Lahiri et al., 1981; Ishii et al., 1985a; Ishii et al., 1985b; Ishii and Ishii, 1986; de Graaf, 1990; Gonzalez et al., 1994; Jonz and Nurse, 2003; Milsom and Burlenson, 2007; Jonz and Nurse, 2009). Chemoreceptors are found in all gill arches in fish (as well as in the oro-branchial cavity in some species) but only in the carotid artery and aortic arch (derived from the 3rd and 4th pharyngeal arches respectively), in mammals (Dunel-Erb et al., 1982; Lahiri et al., 1983; Smatresk, 1990; Zaccone et al., 1992; Burleson and Milsom, 1993; Gonzalez et al., 1994; Jonz and Nurse, 2003; Jonz et al., 2004; Jonz and Nurse, 2009). Amphibians and reptiles are intermediate, with three chemosensory areas located in derivatives of the 3rd, 4th and 6th pharyngeal arches (Adams, 1958; Ishii et al., 1966; Rogers, 1967; Lillo, 1980; Hoffmann and de Souza, 1982; Ishii et al., 1985a; Ishii et al., 1985b; Ishii and Ishii, 1986; Wang et al., 2004). However, studies in reptiles are incomplete as functional chemosensory areas have so far been identified only in chelonians (Ishii et al., 1985b; Ishii and Ishii, 1986; and now also snakes (see Chapter 4)). In turtles, hypoxia elicits an increase in ventilation (Glass, 1992; Frankel et al., 1969; Lenfant et al., 1970; Jackson, 1973; Benchetrit et al., 1977) due to stimulation of putative peripheral chemoreceptors located in the common carotid artery, aorta, truncus arteriosus and the 67  pulmonary artery (Benchetrit et al., 1977; Ishii et al., 1985b; Ishii and Ishii, 1986). Little is known about these peripheral chemoreceptors in turtles, and the few existing studies come to conflicting conclusions concerning the innervation and the relative importance of the different chemosensory areas in this group. Benchetrit et al. (1977) suggested that the pulmonary arch was the main chemosensory area in the tortoise based on responses to cyanide (NaCN) injections while Ishii et al. (1985b) concluded that the aortic arch was the major reflexogenic area due to its innervation by both the superior and inferior truncal nerves. In their study, the inferior truncal nerve was found to arise from the ganglion trunci of the vagus and innervate the aorta and pulmonary artery, before continuing towards the truncal region where it innervated the pulmonary trunk and the bulbus cordis. The superior truncal nerve ran alongside the carotid artery and innervated the aorta (Ishii et al., 1985b). Although the innervation by the inferior truncal nerve is consistent among studies, that of the superior truncal nerve has been controversial. Adams (1962) found that the superior truncal nerve innervated the truncus and proposed that this was an important chemosensory area based on this innervation and the associatted clusters of epithelioid cells. Evidence for the innervation of the carotid artery is limited. Early studies described innervation by the IX and X cranial nerves to regions below the carotid bifurcation by the head (Ask-Upmark, 1935). This area is not homologous to the carotid bifurcation in mammals (derivative of the 3rd pharyngeal arch). The homologous site is in the common (dorsal) carotid artery at the base of the neck. During development an anastomosis forms between the external and internal carotid arteries and the rostral portion of the external carotid artery disappears. The internal carotid artery gives rise to the extant external carotid artery secondarily and the original bifurcation is lost (see Chapter 1 for details) (Adams 1958; Adams, 1962; Jones and Milsom 1982). In the tortoise, Testudo hermanni, this area is innervated by a branch of the IX cranial nerve (Ishii and Ishii, 1986). Other studies failed to find IX or X cranial innervation to this area (Adams, 1962; Ishii et al., 1985b). In the carotid artery, aortic arch and pulmonary artery of turtles, clusters of characteristic glomus cells surrounded by sustentacular cells are found in the adventitial layer. These cells are in close proximity to efferent and afferent nerve endings and have dense-cored vesicles (Kusakabe et al. 1988) that may contain catecholamines or serotonin (5HT), since monoaminecontaining cells have been observed in the wall of the aorta (Ishii et al. 1985b) and in the 68  opening of the main cardiac vessels (Chiba and Yamauchi, 1973) by exposure to formaldehyde vapor. To date, however, no specific neurotransmitters have been identified in these cells in turtles. The objective of the present study was to characterize the neurochemical content, innervation, organization and distribution of putative oxygen sensing cells in the three chemosensory areas previously identified in turtles (Adams, 1958; Adams, 1962; Benchetrit et al., 1977; Ishii et al., 1985b; Ishii and Ishii, 1986). I hypothesized that glomus-like cells containing a similar complex of neurotransmitters to those found in the mammalian carotid body and in the chemosensory areas of frogs would be distributed in regions of the common carotid artery, aorta and pulmonary artery of turtles. To test this hypothesis I used markers for three neurotransmitters: acetylcholine (ACh), proposed to be the main neurotransmitter involved in signal transduction in mammals; tyrosine hydroxylase (TH), the rate limiting enzyme in catecholamine synthesis (Nurse, 2005, 2010); 5HT, the main neurotransmitter in the NEC’s of fish (Dunel-Erb et al., 1982; Bailly et al., 1992; Zaccone et al., 1992, 1994, 1997; Jonz and Nurse, 2003; Jonz, et al., 2004; Bailly, 2009; Jonz and Nurse, 2012). I also examined these cells for a marker for the human natural killer-1 (HNK-1, a neuronally expressed adhesion molecule), which identifies proliferative cells that develop from neural crest cells. Retrograde tract tracing and immunohistochemistry were used together to determine the anatomical relation between putative oxygen sensing cells and branches from the X cranial nerve. Establishing the neurochemical profiles and innervation of peripheral chemoreceptors in turtles will provide a more comprehensive understanding of the phylogenetic patterns in the neurochemical content and distribution of the chemoreceptor cells of vertebrates.  3.3  Materials and methods  3.3.1  Animals and holding conditions  Ten adult red-eared sliders (Trachemys scripta elegans, Wied) (average mass 0.66 ± 0.06 kg) were obtained from Niles Biological Inc. (Sacramento, California, USA) and housed in indoor tanks (0.9 m x 0.5 m x 0.6 m) containing 0.23 m3 of water with a flow-through system and a constant air temperature of 20 ºC. Turtles were provided with basking platforms and fullspectrum lights with a 12L:12D photoperiod. Animals were fed a mixture of trout chow (Aquamix 400, 5D04), vegetables and fruits as well as vitamin D and calcium supplements three 69  times a week, but were fasted for 7 days before surgery. The holding and experimental procedures followed Canadian Council on Animal Care guidelines and were approved by the University of British Columbia Animal Care Committee (animal care certificate No. A09-0233).  3.3.2  Vascular and nerve anatomy  To identify the innervation by the IX and X cranial nerves to the carotid bifurcation, common carotid artery, aorta and pulmonary artery, I performed dissections of the central vasculature and head using a stereomicroscope (Leica M125, Leica Microsystems, Nussloch, Germany). Digital images were acquired with a camera (Leica DFC295, Leica Microsystems, Nussloch, Germany) attached to the microscope and Leica software Application Suite (LAS V3.6, Leica Microsystems, Nussloch, Germany).  3.3.3  Tissue sections  The neuronal tract tracer Cholera Toxin B (CTB, List Biological laboratories, distributed by Cedarlane Laboratories, Hornby, ON, Canada) was used to further examine the innervation by the X cranial nerve to the carotid bifurcation, common carotid artery, aorta and pulmonary artery. Afferent and efferent nerve fibres of the X cranial nerve in the central vasculature were visualized using an antibody against CTB on longitudinal sections of the four blood vessels. Antibodies against TH, 5HT and vesicular acetylcholine transporter (VAChT) were used to determine the neurochemical content of cells in these areas. These neurotransmitters participate in hypoxia chemotransduction in the mammalian carotid body (Gonzalez et al., 1994). The HNK-1 antibody was used to identify proliferative cells that originate from the neural crest. The presence of all or some of the markers mentioned above, the innervation by afferent fibers from the IX or X cranial nerves and the origin from neural crest are important features of chemosensory cells; I rely on these anatomical features to identify putative oxygen sensing cells in the chemosensory areas previously identified in turtles.  3.3.4  Surgery and tissue preparation  All turtles were given an intramuscular injection of Xylazine (Bayer; 1 mg/kg) for analgesia prior to surgery (N=6). Moderate anesthesia was induced by inhalation of Isoflurane (Baxter Healthcare Corporation, Deerfield, IL, USA) after which turtles were intubated and artificially 70  ventilated using a mechanical ventilator (Harvard Apparatus, Holliston, MA, USA) set at a frequency of 6 breaths per minute and a tidal volume of 20 ml/kg. To obtain a surgical plane of anesthesia, Isoflurane was delivered in a concentration of 4-5% in air. During surgery anesthesia was maintained at 1-2%. An incision was made in the neck and the X cranial nerve was carefully isolated from the carotid artery and surrounding connective tissue. The nerve was injected 2 cm distal from the head one to two times (1-2 µl total) with CTB (1%) in distilled water using a 2 µl Hamilton Syringe (7002KH, Reno, NV, USA) with a glass micropipette attached to the needle. Only one side of the animal was injected with the neuronal tracer such that the innervation on the other side of the animal served as an internal control. Tissue on the uninjected side of the animal (N=6) and from additional animals that were not used for tract tracing (N=4) were used to determine the neurochemical content of putative chemosensory areas (total N=10). Six to seven days after injection, turtles were deeply anesthetised with Isoflurane and perfused transcardially with heparinised saline (100 IU/ml) using a blunt 21 gauge needle connected to a peristaltic pump until the blood vessels appeared clear of blood. Paraformaldehyde (PFA; 4% in 0.1 M phosphate buffer saline (PBS; Na2HPO4, 13.4 g/l; NaH2PO, 6 g/l; NaCl, 9 g/l; buffered to pH of 7.4 with NaOH)) was then perfused to fix the tissue. The carotid bifurcation and segments of the common carotid and pulmonary arteries and aortae were removed from the base of the heart to 1 cm distal to where each vessel noticeably constricts in diameter as it extends peripherally. The vagi from both the injected and uninjected sides were collected to confirm that the neuronal tracer remained in the injected nerve. The adrenal gland, jugular ganglia and lungs of two animals were also collected to use as positive immunohistochemical controls (see Controls section). Following fix perfusion, the collected tissue was pinned to sylgard gel plates and post-fixed for 2-4 h in 4% PFA at room temperature, cryoprotected in 30% sucrose buffer and frozen in Tissue Tek (Sakura, San Marcos, CA, USA) at -80 °C. Longitudinal sections (20 µm, same procedure as explained in Chapter 2, Fig. 2.1A) were made using a cryostat (CM3050, Leica Microsystems, Nussloch, Germany) and serially mounted on Superfrost plus slides (VWR International, West Chester, PA, USA). Slide mounted sections were immediately processed for immunohistochemistry or stored in a -80 ºC freezer until needed.  71  3.3.5  Immunohistochemistry  Techniques for immunolabeling and imaging were performed as described in Chapter 2. Briefly, slide mounted tissue was washed in PBS which was pipetted in and out of the glass holders to prevent the delicate internal structure of the arteries from breaking off from the slides. Sectioned tissue was blocked in 10% normal donkey serum (NDS) or normal goat serum (NGS, Jackson Laboratories, distributed by Cedarlane Laboratories, Hornby, ON, Canada), depending on the host-species of the secondary antibody, for 1 h. Primary antibodies were diluted (PBS, 0.3% Triton X-100, 2% NDS) according to optimal dilutions determined previously (Table 3.1). Slides were incubated with the primary antibody (individually or in combination) for 48 h at room temperature and then washed in PBS. Following the washes, slides were incubated in the dark with fluorescently labelled secondary antibodies diluted in PBS (with 0.3% Triton X-100 and 2% NDS) (Table 3.1) for 2 h and subsequently washed in PBS. DAPI was used to visualize cell nuclei (Vectashield with DAPI, Vector Laboratories, Burlington, Ontario). Coverslips (#1.5, Fisher Scientific, Ottawa, ON, Canada) were mounted with Vectashield (Vector Laboratories, Burlington, Ontario) to reduce photobleaching and then the coverslips were sealed with nail polish. Processed slide-mounted tissue was stored at 4 ºC in the dark until imaging.  3.3.6  Controls  Controls consisted of excluding the primary antibody to control for non-specific binding of the secondary antibody. To control for interactions between antibodies, single labelled slides were processed in each run. Positive controls for TH and VAChT primary antibodies were performed using the adrenal gland and jugular ganglion, respectively. The lungs of turtles in which the X cranial nerve had been injected with the neuronal tracer were used as positive controls for the primary antibodies against 5-HT and CTB. The specificity of the primary antibodies was also tested by examining the labeling pattern of two different antibodies raised in the same host species. For instance, the polyclonal antibodies for VAChT and 5HT raised in rabbit, and the antibodies for CTB and 5HT raised in goat. The specificity of these antibodies had also been verified by the suppliers using a preabsorption control. Specific labeling of the rat brain using the polyclonal antibodies rabbit anti-5HT and goat anti-5HT, was inhibited by preincubation of diluted antiserum with 500 µM and 100 µg/ml of serotonin, respectively. In Western blots VAChT appears as a 70 kDa 72  immunoreactive band and staining of the band is inhibited by pretreatment of the antibody with the immunizing peptide. See Table 3.1 for antibody and supplier details. Additionally, the specificity of CTB injections was established by immunohistochemically processing injected and uninjected X cranial nerves.  3.3.7  Microscopy for cryo-sectioned tissue  Sectioned tissue was observed using an epifluorescence light microscope (Axioplan 2, Zeiss, Jena, Germany) equipped with a G365, HQ470 OR BP546/12 excitation filter and a LP420, HQ525/50 or LP590 emission filter to detect DAPI, Alexa 488 and Alexa 568 or 594 respectively. Images were captured using a Q-Imaging CCD camera (Model Retiga 1300, Burnaby, BC, Canada) and QCapture image software (Version 2.95.0). Representative sections were further examined using a spinning disk microscope (Perkin Elmer Ultraview VOX Spinning Disk Confocal , Waltham, MA, USA), equipped with 405 nm, 488 nm and 561 nm lasers and filters 527/55, 445/60 or 615/70, 525/50 or 640/120 for detection of GFP, DAPI and rhodamine. Z-stacks of 203-266 optical sections and 0.19-0.25 µm apart were captured using Leica multi-immersion 20x and 63x glycerol objectives and a Hamamatsu C9100-50 camera. Images were analyzed using ImageJ.  3.3.8 3.3.8.1  Quantification Cell size  I measured the diameters of different cell types to further characterize candidate chemosensory cells. Only cells where I could clearly identify the nuclei and where I could see the entire labeled cytoplasm surrounding the nuclei were measured. Ten 5HT- and ten VAChTimmunoreactive (IR) cells were measured in different sections of the common carotid arteries, aortae and pulmonary arteries of 4 animals using Volocity imaging software (Version 6.1.2, Perkin Elmer Inc., Waltham, MA, USA) and ImageJ. All data are presented as average cell size (µm) ± s.e.m.  3.3.9  Haematoxylin and eosin histology  Sections from all sites were also stained with Haematoxylin and Eosin (H&E) to observe the internal structure of blood vessels in the central vasculature and to determine the location of 73  putative chemoreceptor cells. Fixed tissue (4% PFA) was placed in specimen cassettes and dehydrated for paraffin embedding by running them through changes in ethanol (70%, 80%, 95%, 100%; A407, Fisher scientific, Ottawa, ON, Canada), xylene (Fisher scientific, Ottawa, ON, Canada) and paraffin at 56 ºC. Longitudinal sections 20 µm thick (N=2) or cross sections (N=1) of 6 µm thick at 7-10 µm intervals, were cut through representative regions of the central vasculature using a Leica rotary microtome (RM2255, Leica Microsystems, Nussloch, Germany). Sections were mounted onto slides. Slides were deparaffinised and rehydrated by running them through changes in xylene, ethanol (100%, 95%, 90%, 80%, and 70%) and water. Slides were stained in Gill’s #2 hematoxylin (3801520, Leica Microsystems Inc., Buffalo Grove, IL, USA), then washed with running tap water, rinsed in 1% acid alcohol and washed again. The slides were then stained in 1% lithium carbonate (Sigma Aldrich, Oakville, ON, Canada), washed with tap water, stained in 2% aqueous eosin (stock solution: Sigma Aldrich, Oakville, ON, Canada) and washed again before dehydrating back to xylene through serial solutions of ethanol. Coverslips (#1, Fisher Scientific, Ottawa, ON, Canada) were mounted with permount (SP15-500, Fisher Scientific, Ottawa, ON, Canada). The H&E stained cross sections were prepared by Wax-it Histology services, Vancouver, B.C.  3.4  Results  3.4.1  Anatomy and innervations of the central vasculature  In the turtle, the pulmonary trunk and the right and left aortae arise from the ventricle. The pulmonary trunk divides into the right and left pulmonary arteries. The right aorta gives rise to the brachiocephalic trunk and continues posteriorly where it joins with the left aorta to form the dorsal aorta. The brachiocephalic trunk bifurcates and branches to give rise to the left and right subclavian and common carotid arteries (Fig. 3.1A). The left and right common carotid arteries each bifurcate (secondarily) by the head to form the internal and external carotid arteries. The aorta, common carotid and pulmonary arteries are innervated by the X cranial nerve, which arises from the jugular ganglion in the head and runs down alongside the common carotid artery. A number of small branches from the X nerve innervate the common carotid (dorsal carotid) artery after it branches from the brachiocephalic trunk (Fig. 3.1B, C, arrows; Figure illustrates arrangement on the left side). The X nerve then continues down and right after going over the aorta it innervates the dorsal side of this blood vessel around the region where it arches 74  (Fig. 3.2A, D; arrows; this figure illustrates arrangement on the left side). Subsequently, a branch of the X, the recurrent laryngeal nerve (RLN), arises, loops under the aorta and ascends to innervate the larynx (Fig. 3.2B, E). The X nerve gives rise to numerous small branches that innervate the ventral side of the pulmonary artery, just before running next to the descending pulmonary artery towards the lungs (Fig. 3.2C, F; arrows). A thin branch continues to innervate the truncus arteriosus.  3.4.2  Controls for immunohistochemistry  In turtles, neurons in the jugular ganglion labeled for VAChT (Fig. B1-A), and neuroepithelial bodies and vagal nerve fibers in the lungs labeled for 5HT and CTB, respectively (Fig. B1-B, C). Furthermore, the labeling patterns of the VAChT and 5HT primary antibodies raised in rabbit were very different, as were the reactivity patterns of the anti -CTB and anti-5HT antibodies raised in goat, when tested in the carotid artery, aorta and pulmonary artery of turtles. Omission of the primary antibody never resulted in detectable labeling in any of the immunohistochemical runs, indicating that all antibodies were likely specific in turtles. Furthermore, no differences in labeling pattern were found between single and double labeled slides. I confirmed the specificity of the neuronal tracer microinjections by visualizing CTB in injected, but not in uninjected X nerves (data not shown). These controls confirm that the procedures and antibodies used in this study are effective and likely specific in turtles.  3.4.3  Neurochemical content and innervation of chemosensory areas  I found three different populations of cells, in the common carotid artery, aortic arch, truncus arteriosus and pulmonary artery of turtles. One population contained 5HT and two others contained VAChT. The two neurotransmitters were never found to colocalize and the morphology and distribution of the three cell types was distinct. One of the populations of VAChT-IR cells and the 5HT-IR cells were innervated by the X cranial nerve. I did not find cells containing catecholamines in any of the chemosensory areas. However, TH (indicative of catecholamines) labeled cells in the adrenal glands of turtles (Fig. 3.3), indicating that this antibody works well in this species. No cells labeled for VAChT, 5HT or TH in the bifurcation of the carotid into internal and external branches near the head. Furthermore, I found no  75  indication that the carotid bifurcation is innervated by the X cranial nerve, as I did not find any labeling by the CTB marker (data not shown).  3.4.3.1  Neurochemical content of the carotid artery  Figure 3.4 describes the orientation of the longitudinal sections described below. The internal structure of the common carotid artery is characterized by the presence of ridges and lumen sinusoids (Fig. 3.5C). In the adventitia layer at the base of the common carotid artery, right after it branches from the brachiocephalic trunk I found clusters of cells that contained VAChT (Fig. 3.5C, E). These cells were oval with large nuclei and a diameter of 10.48 ± 0.3 µm (Fig. 3.5E). Cells containing VAChT were confined to this region and not found elsewhere in the blood vessel. Some 5HT-IR cells were also present in this area, but they never colocalized with VAChT-IR cells. More distally, the internal structure of the common carotid artery becomes denser with numerous lumen sinusoids (Fig. 3.5C). Numerous 5HT-IR cells were found throughout this region of the blood vessel wall arranged singly (Fig. 3.6B, arrowheads; D) or in clusters of 2-5 cells (Fig. 3.6E). These cells were oval in shape, with large nuclei and 11.6 ± 0.4 µm long. They occurred in the vessel wall in close proximity to the lumen sinusoids (Fig. 3.6C). Some 5HT-IR cells were innervated by the vagal fibers (Fig. 3.6D). After about one third of the length of the common carotid artery, from the base of the blood vessel towards the carotid bifurcation, the artery narrows, its internal structure becomes smoother and 5HT-IR cells disappear.  3.4.3.2  Neurochemical content of the aorta  The internal structure of the aorta is more complex than that of the common carotid and pulmonary arteries. Right after the aorta emerges from the ventricle, the internal structure of the blood vessel becomes a complex series of ridges (Fig. 3.7B). In the adventitia layer of this segment of the aorta I found chains of polygonal cells (17.2 ± 0.5 µm long) with large nuclei that contained VAChT (Fig. 3.8B, arrows; C). Subsequently, the internal structure of the aorta becomes wide and spongy (Fig. 3.7B, arrow; D). Numerous circular clusters (58.6 ± 2.8 µm) of oval VAChT-IR cells (diameter: 10.0 ± 0.4 µm) were present in the spongy tissue, towards the heart-end of this segment (Fig. 3.7B, D arrow; F, G). Cholinergic cells surrounded some of the lumen sinusoids in this region (Fig. 3.7F). 5HT-IR cells were present throughout the length 76  (from the heart to the systemic circulation) of the aorta embedded in the vessel wall, but they occurred at higher densities in the spongy tissue (Fig. 3.7C-E, arrowheads). 5HT cells occurred mostly in singles, but were occasionally found in groups of 2-5 cells (Fig. 3.7H, I). These cells were oval with large nuclei and a diameter of 11.0 ± 0.4 µm. The two cell types described above never colocalized, but 5HT-IR cells were found in close proximity to some of the VAChT circular clusters in the spongy region of the artery. Clusters of VAChT-IR cells and 5HT cells were innervated by vagal fibers containing the neuronal tracer CTB (Fig. 3.7F, I).  3.4.3.3  Neurochemical content of the pulmonary artery  The internal structure of the pulmonary artery is smooth compared to the common carotid artery and aorta. The outer wall at the base of the blood vessel is innervated by the X cranial nerve. Very few 5HT-IR cells were found in this region; instead polygonal cells containing VAChT were organized in sequence in the adventitial layer (Fig. 3.9B, C). These cells had similar morphology and size (16.9 ± 0.7 µm) as the VAChT-IR cells found at the base of the aorta. The pulmonary artery narrows significantly at the origin of the arterial ligament (remnant of the ductus arteriosus) (Fig. 3.10A, B). In the vessel wall of this region of the artery I found numerous 5HT-IR cells arranged in singles or pairs that were almost identical to the cells in the aorta in shape and size (11.5 ± 0.2 µm) (Fig. 3.10B, arrowheads, I). Above and below the narrowing of the vessel the artery is composed mainly of spongy tissue with many lumen sinusoids. Within the spongy region I found large clusters of oval VAChT-IR cells (cluster diameter: 51.2 ± 3.0 µm; cell diameter: 9.7 ± 0.3 µm) (Fig. 3.10F, G), as well as singles (11.5 ± 0.2 µm long) and clusters (cluster diameter: 26.8 ± 0.6; cell diameter: 7.6 ± 0.3 µm) of 5HT-IR cells (Fig. 3.10H, I). Although 5HT-IR cells were found throughout the blood vessel, they were more abundant at the narrowing of the vessel and in the spongy tissue surrounding it (Fig. 3.10A, arrow; B, arrowheads). As in the other chemosensory areas described above, 5HT and VAChT did not colocalize in the same cells.  3.4.3.4  Neurochemical content of the truncus arteriosus  In the truncal region close to the heart I found VAChT-IR cells arranged in series. These polygonal cells were located closer to the outer vessel wall and were 18.6 ± 0.5 µm long. Oval  77  5HT-IR cells (12.8 ± 2.5 µm) arranged in singles were observed close to the heart and in the vessel wall shared by the aorta and pulmonary arteries (data not shown).  3.4.4  Innervation of chemosensory areas and putative oxygen sensing cells  The neuronal tracer study confirmed that the carotid artery, aorta, pulmonary arteries and truncus arteriosus are innervated by the X cranial nerve. Nerve fibers containing CTB surrounded some of the lumen sinusoids in all chemosensory regions (Fig. 3.6B, 3.7F) and innervated the region of the aorta and pulmonary artery composed by spongy tissue (Fig. 3.7F, 3.10F, G). Vagal fibers commonly innervate segments closer to the heart and up to where the arteries bend, instead of segments closer to the thoracic-end of the blood vessels. Nerve fibers from the X cranial nerve also innervated 5HT- and oval VAChT-IR cells in all three chemosensory areas (Fig. 3.6D, 3.7F, I, 3.10F-I), except for the common carotid artery where I did not find innervated oval VAChT-containing cells. However, nerve fibers labeled with the tracer surrounded some of the tissue in the region of the blood vessel where these cells were located (Fig. 3.5B). I did not find nerve fibers labeled with the neuronal tracer CTB innervating the series of polygonal VAChT-IR cells at the base of aorta, pulmonary artery and truncus arteriosus, although vagal fibers were observed innervating these three regions. I found cells that were either surrounded by nerve fibers or that had taken up the neuronal tracer throughout the common carotid artery, aorta and the pulmonary artery (Fig. 3.7F, 3.10F). None of these neuronal cells colocalized with the markers for 5HT or acetylcholine.  3.4.5  Neural crest origin of putative oxygen sensing cells  Cells labeled with the HNK-1 marker were observed in the common carotid artery, aorta and pulmonary artery of turtles (data not shown). However, none of these cells were positive for 5HT or VAChT, indicating that these cells are either not derived from the neural crest or that they are mature, non-proliferative cells.  3.5  Discussion This study was designed to investigate the innervation and neurochemical content of  putative oxygen sensing cells in turtles. I examined cells in the common carotid artery, aorta, truncus arteriosus and pulmonary artery, sites previously shown to be chemosensitive in these 78  animals. My definition of a putative arterial oxygen sensing cell was one that had the morphological characteristics displayed by other arterial oxygen sensing cells ranging from NEC’s in fish gills to chromaffin or glomus cells in mammalian carotid bodies; oval cells (~10 µm) with large nuclei, an abundance of vesicles containing neurotransmitters, and innervation by the IX or X cranial nerves (Ishii and Osaki, 1969; Ishii et al., 1985b; Gonzalez et al., 1994; Jonz and Nurse, 2003; Campanucci and Nurse, 2007; Coolidge et al., 2008; Jonz and Nurse, 2009, 2012; Shakarchi et al., 2012; Chapter 2). I identified two different cell types likely to be oxygen sensing cells in all of these regions; cells containing VAChT and cells containing 5HT. Catecholamines were never seen in putative oxygen sensing cells and the VAChT and 5HT were never found in the same cell. The arrangement and distribution of cells containing these chemicals was also distinct. 5HT and ACh have been proposed to participate in hypoxia chemotransduction in mammals and fish (Dunel-Erb et al., 1982; Gonzalez, et al., 1994; Zaccone et al., 1994, 1997; Jonz and Nurse, 2003; Nurse, 2005, 2010; Porteus et al., 2012; Shakarchi et al., 2012; Jonz and Nurse, 2012; Zachar and Jonz, 2012), but their roles in reptiles are unknown. My findings imply that they are involved in oxygen sensing in turtles also.  3.5.1  Innervation of peripheral chemosensory areas  I found that the aorta and common carotid and pulmonary arteries were all innervated by the X cranial nerve, and that this innervation extended to the truncal region. I did not find any innervation of the extant carotid bifurcation, in contrast to observations by Ask-Upmark (1935) who reported that both the IX and X cranial nerves innervated this region in turtles. In this study, the common (dorsal) carotid artery at the base of the neck was well innervated by branches of the X cranial nerve and not by a branch of the IX cranial nerve. Ishii et al. (1985b) also failed to find any innervation from the IX cranial nerve to the common carotid artery of the turtle Geoclemmys reevesii although they did in the tortoise Testudo hermanni (Ishii and Ishii, 1986). As in other studies (Adams, 1958, 1962; Ishii et al., 1985b), I found that the aorta and pulmonary artery were innervated by branches of the X cranial nerve, and that a fine vagal branch continued on to innervate the truncal region. I did not see any innervation by the superior truncal branch of the X cranial nerve to the aorta or truncus, nor did I see a single branch from the inferior truncal nerve to the aorta and pulmonary artery (Adams, 1962; Ishii et al., 1985b). Instead, numerous branches arose from the X nerve to innervate the regions where the blood 79  vessels bend. A similar pattern of innervation with aortic and pulmonary branches originating independently from the ganglion trunci of the vagus was observed in some turtles (Geoclemmys reevesii) by Ishii et al. (1985b).  3.5.2  Putative oxygen sensing cells in the chemosensory areas of turtles  I found glomus or neuroepithelial-like cells containing either VAChT or 5HT in the common carotid artery, aorta and pulmonary artery, but not at the bifurcation of the extant common carotid into the internal and external carotid arteries by the head. Although, Frankel et al. (1969) described a carotid body-like structure below the carotid bifurcation of some turtles; it is now felt that the carotid bifurcation in turtles is secondarily derived, and that the homologous site to the carotid bifurcation in mammals lies at the base of the neck where the common carotid branches from the brachiocephalic trunk. Physiological experiments on the tortoise Testudo horsfieldi (Benchetrit et al., 1977), together with my findings of innervated cells containing either VAChT or 5HT in this region of the common (dorsal) carotid artery supports this notion.  3.5.2.1  Distribution of cells containing vesicular acetylcholine transporter in the  chemosensory areas of turtles 3.5.2.1.1  Large polygonal VAChT-IR cells  Polygonal cells, with large nuclei were arranged in a series of rows in the outer most region of the adventitial layer at the base of the aorta, pulmonary artery and truncus arteriosus. Their size was similar in all three locations (aorta: 17.2 ± 0.5 µm; pulmonary artery: 16.9 ± 0.2 µm and truncus: 18.6 ± 0.5 µm). Given that the morphology (polygonal), size and chain-like arrangement of these cells was not characteristic of the 10-12 µm ovoid mammalian glomus cells (Gonzalez et al., 1994; Campanucci and Nurse, 2007), fish NEC’s (Jonz and Nurse, 2003; Coolidge et al., 2008; Jonz and Nurse, 2009, 2012; Shakarchi et al., 2012) or amphibian putative oxygen sensors (Ishii and Oosaki, 1969; Chapter 2), and that their distribution on the outer most layer of the adventitia would not be favourable for sensing oxygen in the blood, I speculated that these polygonal VAChT-IR cells were not chemosensory. It is likely, however, that these cells serve an important function given their presence at the outflow tract of the ventricle to the aorta and pulmonary artery. A plausible role for the VAChTIR cells located at the base of the great arteries and the truncus arteriosus may be control of 80  outflow resistance by regulation of vasomotor tone at these sites. In animals with an incomplete separation of the systemic and pulmonary circulations, ACh has been shown to regulate the degree of cardiac shunt (Hicks, 1994; Wang et al., 2001b). Infusion of ACh causes the pulmonary artery to vasoconstrict, increasing its resistance relative to that of the aorta. This leads to an increase in systemic blood flow relative to pulmonary blood flow and a net right to left shunt (R-L shunt) (Burggren 1987; Hicks, 1994). Although changes in cardiac shunt are mainly achieved by regulating vascular resistance of the pulmonary artery at the site where the vessel narrows (constriction) (Shelton and Burggre, 1976; Burggren, 1977; Milsom et al., 1977), changes in diameter of the outflow tract of the pulmonary artery (Burggren, 1977) and to a lesser extent that of the aorta also affect vessel resistance (Shelton and Burggren, 1976).  3.5.2.1.2  Populations of AChT-IR cell clusters  Large clusters of oval VAChT-IR cells with large nuclei were observed at the base of the common carotid artery where it branches from the brachiocephalic trunk, and in the spongy tissue surrounding the vessel narrowing of the aorta and pulmonary artery. Occasionally, cells in the common carotid artery were arranged in series surrounding pockets of spongy tissue, and in the aorta surrounding lumen sinusoids. The cells were of similar size in all three blood vessels (aorta: 10.0 µm; pulmonary artery: 9.7 µm and common carotid artery: 10.48 µm). The ovoid morphology and size (~10 µm) of these cells was similar to chemoreceptor cells of other vertebrates (Ishii and Oosaki, 1969; Ishii et al., 1985b; Gonzalez et al., 1994; Jonz and Nurse, 2003; Campanucci and Nurse, 2007; Coolidge et al., 2008; Jonz and Nurse, 2009, 2012; Shakarchi et al., 2012; Chapter 2). The organization of cells in clusters that is characteristic of peripheral chemoreceptors in amphibians and mammals permits chemical and electrical synapses between neighbouring oxygen sensing cells (Kusakabe et al., 1987; Gonzalez et al., 1994; Nurse, 2010). Clusters of granulated cells have also been found in the adventitial layer of the wall of the common carotid artery, aorta and pulmonary artery of the tortoise Testudo hermanni (Kusakabe et al., 1988). Junctional specializations were observed between some of the glomuslike cells, potentially allowing for communication between cells. Furthermore, these cells were shown to contain dense small granules (Kusakabe et al., 1988), which indicate a secretory function. Cell clusters in the tortoise shared a number of anatomical features with the VAChTIR cell clusters I found in the turtles. For instance, clusters in both, were innervated and similar 81  in morphology and size (tortoise: 20-50 µm; turtle: 58.6 µm and 51.2 µm, aorta and pulmonary respectively). It is likely, then, that the granulated cell clusters reported by Kusakabe et al. (1988) correspond to the VAChT-IR cell clusters found in this study and that ACh is stored and released from the dense-cored vesicles. Although, ACh is not found in amphibian chemoreceptor cells (Ishii et al., 1985a; West and Van Vliet, 1992; Chapter 2), it has been shown to be crucial in chemotransduction in the mammalian carotid body, albeit co-released with ATP (Eyzaguirre and Zapata, 1984; Zhang et al., 2000; Nurse, 2005; Shirahata et al., 2007; Nurse, 2010). Infusion of ACh into the gills of fish increased nerve discharge, suggesting that ACh may also be involved in chemoreception in this group (Burleson and Milsom, 1995). Given the anatomical features that characterize VAChT-IR cell clusters in the chemosensory areas of redeared sliders, it is likely that these cells are involved in oxygen sensing. I cannot disregard the possibility that these cells may also have a role in vasomotor regulation. Afferent (sensory) and efferent nerve endings lay close to large clusters of granulated cells in the tortoise Testudo hermanni (Kusakabe et al., 1988). Efferent innervation of glomuslike cells may regulate their sensitivity to hypoxia, as has been shown in the carotid labyrinth of toads (West and Van Vliet, 1992) and in the carotid body of mammals (reviewed by Nurse, 2010). However, it is also possible that ACh release from these cells after efferent stimulation will increase the resistance to blood flow of these vessels, by acting on the smooth muscle at the vessel narrowing. As shown by Milsom et al. (1977) at least in the pulmonary artery, the region where the vessel narrows is highly innervated by the X cranial nerve and is a main area of vasomotor regulation.  3.5.2.2  Distribution of serotonergic cells in the chemosensory areas of turtles  I found numerous 5HT-IR cells spread throughout the truncus arteriosus, aorta, pulmonary artery and the initial portion of the common carotid artery. These cells occurred alone or in groups of 2-5 and were of similar shape and size (11-12 um) in all the regions. These cells were innervated by the X cranial nerve and their morphology and size resembled glomus cells in the carotid body of mammals (Gonzalez et al., 1994) and the vascular stroma of the carotid labyrinth of Bufo marinus, Bufo vulgaris, Xenopus laevis (Rogers, 1963; Ishii et al., 1966; Ishii and Oosaki, 1969; Ishii and Kusakabe, 1982) and Rana catesbeiana (Chater 2). I propose that the  82  5HT-IR cells found in the chemosensory areas of turtles participate in hypoxia chemotransduction. My results provided evidence for possible chemosensing in the truncal region of reptiles. The truncus is an important baroreceptor region in this group (Adams, 1958; Berger, 1987). A chemosensory role for this region was formerly proposed based on the observation of nests of epithelioid cells in close proximity to nerve bundles in the truncal region of lizards and turtles (Adams, 1962; Berger et al., 1982). In this area, the presence of catecholamines was deduced from yellow or yellow-green fluorescent cells arranged singly or in loose groups that were visualized after formaldehyde vapor exposure (Berger et al., 1982; Ishii et al., 1985b). Around the base of the great vessels of the turtle heart, however, clusters of yellow-orange fluorescent cells occurred that were not affected by reserpine treatment (Chiba and Yamauchi, 1973), suggesting that these cells contained 5HT. Although this is consistent with my observations, I found that single cells rather than clusters were more common in the truncal region of red-eared sliders. In any case, the size and shape of 5HT-IR cells in the present study resembled those reported for the granulated cells of the soft-shelled turtle (10-20 um; Chiba and Yamauchi, 1973). The discrepancy in arrangement could be due to species differences. I also found single or small clusters of serotonergic cells in the aorta and pulmonary arteries of turtles. The presence of cells containing biogenic amines has been shown using formaldehyde vapor exposure in the aorta of the turtle Geoclemmys reevesii. Given that these cells gave a yellow-green fluorescence it was concluded that the amine present was a catecholamine. As noted in Chapter 2, this technique is not very specific and can lead to inaccurate interpretations of the results. Thus, it is likely that the neurotransmitter present in these cells was 5HT as I did not find any evidence for cells containing catecholamines in any of the chemosensory areas of turtles. 5HT is the main neurotransmitter involved in oxygen sensing in the NEC’s of fish (Dunel-Erb et al., 1982; Zaccone et al., 1992; Bailly, 2009; Jonz and Nurse, 2003; Jonz, et al. 2004; Porteus et al., 2012; Jonz and Nurse, 2012; Zachar and Jonz, 2012), it also participates in neuromodulation in the carotid body of mammals (Nurse, 2010) and is present in all chemosensory areas of the bullfrog (Chapter 2). Therefore, its presence in putative oxygen sensing cells of turtles is not surprising and it is likely to play a role in oxygen chemotrasnduction.  83  Interestingly, granulated cells were found to be in contact with smooth muscle cells in the pulmonary artery of the tortoise Testudo hermanni, (Kusakabe et al., 1988). This arrangement has been described in the carotid labyrinth of toads and it was proposed that catecholamine release from glomus cells acted on smooth muscle cells to regulate vascular tone (Ishii and Kusakabe, 1982). In the present study, I found large clusters of serotonergic cells only in the pulmonary artery in the spongy tissue surrounding the vessel where it narrows. These cell clusters were innervated, similar in shape and size (tortoise: 20-50 µm; turtle: 26.85 ± 0.6 µm) to the granulated cell clusters found by Kusakabe et al. (1988) in the tortoise. It is possible that these groups of cells are vasoactive acting on smooth muscle cells through the release of 5HT at the site of the vessel narrowing.  3.5.3  Absence of catecholamines in the peripheral arterial chemoreceptors of turtles  In contrast to mammals (Nurse, 2010) and amphibians (Chapter 2), while catecholamines were absent in the chemosensory areas of turtles, TH labeled cells in the adrenal gland, demonstrating that the antibody was effective in turtles. Kusakabe et al. (1988) proposed that catecholamines released from the granulated cells in the chemosensory areas of turtles could be involved in efferent control of vascular tone by the same mechanism present in Xenopus laevis (Ishii and Kusakabe, 1982). This suggestion was based on the observation of yellow-green fluorescent cells in the aorta wall of turtles after exposure to formaldehyde vapor (Ishii et al., 1985b). As mentioned before, it is likely that the biogenic amine responsible for the yellowgreen fluorescence was 5HT and that the clusters of granule-containing cells correspond to the VAChT-IR clusters in the present study. My findings suggest that catecholamines are not involved in chemoreception in turtles, in contrast with mammals and amphibians but comparable to fish.  3.5.4  Origin of putative oxygen sensing cells  Neither serotonergic nor cholinergic cells were labeled by the HNK-1 marker. The HNK-1 monoclonal antibody recognizes an epitope expressed on a number of cell adhesion molecules. This antibody is often expressed in migrating neural crest cells (Bronner-Fraser, 1986) and glomus cells that are neural crest derived. I used this antibody in an attempt to determine the origin of putative oxygen sensing cells in the turtle. However, the fact that putative 84  chemoreceptor cells in the turtle did not co-label with the HNK-1 marker does not prove that these cells are not derived from the neural crest, since cells that are no longer migrating or proliferative do not express this surface marker. Thus, the origin of the 5HT and cholinergic cells in the chemosensory areas of red-eared sliders still remains unknown. A different marker for neural crest derived cells, such as neuronal specific enolase, could be used to address this question.  3.6  Conclusion My data allows me to compare the distribution and neurochemical content of oxygen  sensing cells in turtles with those of other, better-studied vertebrate groups. To my knowledge, this study is the first to show the presence of both ACh- and 5HT-IR cells in the common carotid artery, aorta, pulmonary artery and truncus arteriosus of turtles. With the exception of the truncal region, these areas have been shown to be functional chemoreceptive sites. Based on distribution, cell morphology, innervation and neurochemical content, I propose that the VAChT-IR cell clusters and 5HT-containing cells present in these areas are likely to function as oxygen sensing cells. Additionally, this study suggests that the truncal region may possess putative oxygen sensing cells containing 5HT. However, the involvement of this area in chemoreception maybe small, since no VAChT-IR cell clusters were observed at this location. While multiple neurotransmitters now appear to be involved in O2 sensing in the glomus cells in the carotid bodies of mammals and the chemosensory areas in reptiles (present chapter) and amphibians (Chapter 2) there are some distinctive differences: 1) 5HT and ACh are found in different cells in turtles, while neurotransmitters colocalize in the chemoreceptor cells of mammals (Gonzalez, et al., 1994) and in some cells in the carotid labyrinth of amphibians. 2) Putative oxygen sensing cells in amphibians contain catecholamines and 5HT instead of ACh and 5HT, and 3) in turtles all chemosensory areas are solely innervated by the X cranial nerve, while the carotid body of mammals (Gonzalez, et al., 1994) is innervated by the IX cranial nerve and the carotid labyrinth of frogs is innervated by both, the IX and X cranial nerves. The one thing all groups have in common is the presence of 5HT in O2 chemosensitive regions although its role may shift from primary neurotransmitter to neuromodulator during the phylogenetic progression.  85  Figure 3.1. (A) Schematic diagram of the major central arteries arising from the heart of the turtle. (B) picture and (C) schematic diagram showing the anatomy and innervation by the vagus nerve (arrows) of the common carotid artery of Trachemys scripta elegans. Scale bar in B = 1 mm.  86  Figure 3.2. Pictures and schematic diagrams showing the anatomy and innervation by the vagus nerve of the aorta (A, and D) and pulmonary artery (C and F) of Trachemys scripta elegans. Arrows indicate the nerve branches of the vagus nerve innervating the dorsal side of the aorta (A) and the ventral side of the pulmonary artery C). Note that the RLN do not innervate these areas. Scale bar 1 mm.  87  Figure 3.3. Positive control for catecholamines in the adrenal glands of Trachemys scripta elegans. A: Immunolabeling for tyrosine hydroxylase (TH, red) and a nuclear stain (DAPI, blue) in the adrenal glands. The images from the red and blue channels are shown separately and as a merged image. Scale bar 10 µm.  88  Figure 3.4. Images to illustrate the level at which longitudinal sections were taken. A: Cross section of artery showing the level at which longitudinal sections were taken. B: vessel wall viewed from the luminal side. C: longitudinal section viewed from the luminal side. Note that because of the spongy nature of the vessel wall, “islands”of tissue appear surrounded by either luminal or sinusoidal space.  89  Figure 3.5. Cells containing vesicular acetylcholine transporter in the common carotid artery of Trachemys scripta elegans. A: Drawing of the central vasculature of the turtle. The outlined box shows the segment of the common carotid artery sectioned for immunohistochemical analysis. The arrow shows the region of the common carotid artery where clusters of VAChT-IR cells are found. B: Immunolabeling for the neuronal tracer Cholera toxin B (CTB, red) and a nuclear stain (DAPI, blue) in the common carotid artery of Trachemys scripta elegans. The images from the red and blue channels are shown separately and as a merged image. Fibers from 90  the vagus nerve (CTB, red) are found around some of the lumen sinusoids in the region of the common carotid artery pointed by the arrows in (A) and (C). C: Montage of a longitudinal section (20 µm thick) of the common carotid artery stained with Haematoxylin and Eosin (H&E). The arrow points at the region of the blood vessel where clusters of VAChT-IR cells are found. The left side of the image is closest to the heart. Bc, brachiocephalic trunk. Scale bar 1mm. D: H&E stained region of the common carotid artery were VAChT-IR cell clusters are found (arrows). E: Immunolabeling for vesicular acetylcholine transporter (VAChT, green) and a nuclear stain (DAPI, blue) in the common carotid artery of Trachemys scripta elegans. The images from the green and blue channels are shown separately and as a merged image. Scale bar 25 µm for B, D and E.  91  Figure 3.6. Serotonergic cells in the common carotid artery of Trachemys scripta elegans. A: Drawing of the central vasculature of the turtle. The outlined box shows the segment of the common carotid artery sectioned for immunohistochemical analysis. 5HT-IR cells were found throughout the length of the segment of blood vessel. B: Immunolabeling for the neuronal tracer Cholera toxin B (CTB, red), serotonin (5HT, green) and a nuclear stain (DAPI, blue) in the common carotid artery of Trachemys scripta elegans. The images from the red and green channels are shown separately and as a merged image. Fibers from the vagus nerve (CTB, red) are found surrounding a lumen sinusoid (∗) and innervating some 5HT-IR cells (arrows). 5HTIR cells with no visible innervation are shown by arrowheads. C: Longitudinal section (20 µm thick) of the common carotid artery stained with haematoxylin and eosin. This section corresponds to (B). (∗) shows a lumen sinusoid next to the vessel lumen (to the left). 92  Arrowheads point to the region where 5HT-IR cells are found. D: Immunolabeling for the neuronal tracer CTB (red), 5HT (green) and a nuclear stain (DAPI, blue) showing vagal innervation of a single 5HT-IR cell. The images from the red and green channels are shown separately and as a merged image. E: Immunolabeling for 5HT (red) and a nuclear stain (DAPI, blue) showing a small cluster of 5HT-IR cells in the common carotid artery. The images from the red and blue channels are shown separately and as a merged image. Scale bars 25 µm.  93  Figure 3.7. Putative oxygen sensing cells in the aorta of Trachemys scripta elegans. A: Drawing of the central vasculature of the turtle. Arrow points to the region (spongy tissue) of the aorta where VAChT-IR cell clusters are found and the area of highest 5HT cell density. B: Montage of a longitudinal section (20 µm thick) of the aorta stained with haematoxylin and eosin. The arrow points to the region of the blood vessel where clusters of VAChT-IR cells are found. The left side of the image is the heart-end (∗). The three rectangular outlines correspond to stained sections in panels C (far left), D (middle, spongy tissue) and E (far right, after slight vessel narrowing). Scale bar 1mm. C-E: Similarly stained longitudinal sections (20 µm thick) of the 94  regions of the aorta outlined in (B). Arrowheads show the regions where 5HT-IR cells are found and arrows show the regions where VAChT-IR cell clusters are found. F: Immunolabeling for the neuronal tracer CTB (red), VAChT (green) and a nuclear stain (DAPI, blue). The images from the red and green channels are shown separately and as a merged image. Vagal nerve fibers (red) and VAChT-IR cell clusters (green) occur in the spongy tissue of the aorta (region shown in D). Some VAChT-IR cells are found surrounding a lumen sinusoid. Vagal fibers innervate some VAChT-IR cells (colocalization shown in yellow). G: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue) showing a low magnification image of the region shown in D and F with numerous VAChT-IR cell clusters. The images from the red and green channels are shown separately and as a merged image. H: Immunolabeling for 5HT (green) and a nuclear stain (DAPI, blue) showing a cluster of serotonergic cells in the spongy tissue region of the aorta. The images from the green and blue channels are shown separately and as a merged image. I: Immunolabeling for the neuronal tracer (CTB, red), 5HT (green) and a nuclear stain (DAPI, blue) showing a single serotonergic cell (green) innervated by fibers of the vagus nerve (red) in the aorta of Trachemys scripta elegans. The images from the red and green channels are shown separately and as a merged image. Scale bars 25 µm (C-H). Bc, brachiocephalic trunk; Cca, common carotid artery; Ao, aorta; Pa, pulmonary artery, Scl, subclavian artery; RLN, recurrent laryngeal nerve.  95  Figure 3.8. Series of polygonal VAChT-IR cells in the aorta of Trachemys scripta elegans. A: Drawing of the central vasculature of the turtle. Arrow points to the region of the aorta (outflow tract of the aorta) where a series of polygonal VAChT-IR cell are distributed. B: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue) showing a low magnification image of the outer wall at the base of the aorta where series of polygonal VAChTIR cells are found. The images from the green and blue channels are shown separately and as a merged image. C: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue) show a close up of the region in B. The images from the green and blue channels are shown separately and as a merged image. Scale bar 25 µm.  96  Figure 3.9. Pulmonary artery of Trachemys scripta elegans. A: Drawing of the central vasculature of the turtle. Arrow points to the region of the pulmonary artery (outflow tract of the pulmonary artery) where series of polygonal VAChT-IR cell are located. B: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue) showing a low magnification image of the outer wall at the base of the pulmonary artery where a series of polygonalVAChT-IR cells are found. C: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue) show the region outlined in B. The images from the green and blue channels are shown separately and as a merged image. Scale bar 25 µm.  97  Figure 3.10. Putative oxygen sensing cells in the pulmonary artery of Trachemys scripta elegans. A: Drawing of the central vasculature of the turtle. Arrow points to the region of the pulmonary artery (region of vessel narrowing and surrounding spongy tissue) where large clusters of VAChT-IR, and 5HT-IR cells, as well as small clusters and single 5HT-IR cells are found. B: Montage of the pulmonary artery labeled with a nuclear stain (DAPI, blue). The arrows point at the spongy tissue surrounding the vessel constriction where clusters of VAChTIR cells are found. The arrowheads point at the region where single (region of vessel narrowing and surrounding spongy tissue) and large clusters (spongy tissue) of 5HT-IR cells are found. The left side of the image is the heart-end. The three rectangular outlines correspond to cross sections in panels C (far left, heart-end), D (middle, central region) and E (region of vessel narrowing and surrounding spongy tissue) stained with haematoxylin and eosin. C-E: Cross 98  sections (6 µm thick) of the regions of the pulmonary artery outlined in (B) stained with haematoxylin and eosin. Scale bar 161 µm. F-G: Immunolabeling for the neuronal tracer CTB (red), VAChT (green) and a nuclear stain (DAPI, blue) showing nerve fibers from the vagus in close proximity to clusters of VAChT-IR cells in the spongy tissue surrounding the vessel constriction (F). Some of the VAChT-IR cells in the clusters are innervated by the vagus (G, colocalization shown in yellow). H-I: Immunolabeling for the neuronal tracer CTB (red), 5HT (green) and a nuclear stain (DAPI, blue) showing an innervated large cluster of serotonergic cells (H) and a single innervated serotonergic cell in region of the spongy tissue surrounding the region of narrowing of the pulmonary artery. The images from the red and green channels are shown separately and as a merged image. Scale bars 25 µm for F-I.  99  Table 3.1 Primary and secondary antibodies used for immunohistochemistry on tissue sections Antisera Primary  Antigen  Manufacturer  Host  Dilution  Cat. No.  Secondary antisera  5HT  serotonin  Sigma-Aldrich  rabbit  1:350  S5545  Alexa Fluor® 488a Alexa Fluor® 568b  5HT  serotonin  Immuno Star  goat  1:650 (ao,tr) 1:700 (cc, pa)  20079  Alexa Fluor® 594d  TH  tyrosine hydroxylase  Immuno Star  mouse  1:200 (adrenal)  22141  Alexa Fluor® 594f  VAChT  vesicular acetylcholine Sigma-Aldrich transporter  rabbit  1:350 (ao, pa, tr) 1:150 (cc) 1: 250 (jug gnlg)  V5387  Alexa Fluor® 488a Alexa Fluor® 568b  HNK-1  CD-57  BD Pharmigen  mouse  1:100  559048  Alexa Fluor® 488e  anti-CTB  Cholera Toxin B subunit  List Biologicals laboratories  goat  1:550 1:500 (lung)  703  Alexa Fluor® 568c  Secondary1 Alexa Fluor® 488 rabbit IgG (H+L)a Molecular Probes, Invitrogen donkey 1:200 A-21206 -Alexa Fluor® 568 rabbit IgG (H+L)b Molecular Probes, Invitrogen donkey 1:200 A-10042 -Molecular Probes, Invitrogen donkey 1:200 A-11057 -Alexa Fluor® 568 goat IgG (H+L)c Alexa Fluor® 594 goat IgG (H+L)d Molecular Probes, Invitrogen donkey 1:200 A-11058 -Alexa Fluor® 488 mouse IgG (H+L)e Molecular Probes, Invitrogen donkey 1:200 A-21202 -Alexa Fluor® 594 mouse IgG (H+L)f Molecular Probes, Invitrogen goat 1:200 A-11005 -1 Secondary antisera were conjugated with a fluorescent marker a-h Secondary antisera antigen corresponds with primary antibody host Abbreviations: Aorta (ao), Truncus (tr), Pulmonary artery (pa), Common carotid artery (cc), Adrenal (adrenal gland), Jugular ganglion (jug gngl) 100  Chapter 4: Distribution and role of peripheral arterial chemoreceptors in cardio-respiratory control of the South American rattlesnake (Crotalus durissus) 4.1  Summary Peripheral arterial chemoreceptors monitor the levels of arterial blood gases and adjust  ventilation and perfusion to meet metabolic demands. These chemoreceptors are present in all vertebrates studied to date but have not been described in reptiles other than turtles. The goals of this study were to 1) identify functional chemosensory areas in the South American rattlesnake (Crotalus durissus), 2) determine the neurochemical content of putative chemosensory cells in these areas, and 3) determine the role each area plays in ventilatory and cardiovascular control. To this end, rattlesnakes were instrumented with transonic flow probes, arterial catheters and subcutaneous impedance electrodes to measure shunt fraction, heart rate, blood pressure and ventilation, respectively. The catheters were placed at three putative chemosensory sites, the bases of the aortic arch and pulmonary artery as well as at the carotid bifurcation, for sitespecific activation with sodium cyanide (NaCN, 0.5mg/0.1ml). These same sites were subsequently examined using immunohistochemical markers for acetylcholine, tyrosine hydroxylase (the rate limiting enzyme in catecholamine synthesis) and serotonin (5HT) to identify putative oxygen sensing cells. All three sites were chemosensory and stimulating each led to cardiovascular (shunt fraction and heart rate) and respiratory adjustments although not in an identical fashion. All three chemosensory areas contained cells positive for 5HT, however, cells positive for vesicular acetylcholine transporter were found only in the aorta and pulmonary artery.  4.2  Introduction A primary role of the cardio-respiratory system is to supply oxygen to satisfy the metabolic  demands of tissues. To fulfill this role, peripheral arterial chemoreceptors continuously monitor the levels of oxygen in the blood and adjust ventilation and perfusion to match oxygen supply to demand (Shelton et al., 1986; Gonzalez et al., 1994; Milsom, 1998). This control system is the first line of response when oxygen is limited and is critical to ensure the survival of the animal. Peripheral chemoreceptors have been found in all vertebrates studied so far (Adams, 1958; Ishii 101  et al., 1966; Rogers, 1967; Abdel-Magied and King, 1978; Lahiri et al., 1981; Ishii et al., 1985a, b; Ishii and Ishii, 1986; de Graaf, 1990; Gonzalez et al., 1994; Jonz and Nurse, 2003, 2009). They are always associated with derivatives of pharyngeal arches and innervated by the glossopharyngeal (IX) and/or vagus (X) nerves (Milsom and Burlenson, 2007). In reptiles, however, functional peripheral chemoreceptors have only been identified in turtles, located in the common carotid artery, aortic arch and pulmonary artery (derivatives of the 3rd, 4th and 6th pharyngeal arches respectively) (Ishii et al., 1985b; Ishii and Ishii, 1986; Chapter 3). In lizards, anatomical evidence suggests the presence of chemoreceptor cells in the truncal region (Berger et al., 1982) and at the carotid bifurcation (Adams, 1962; Rogers, 1967), but other areas have not been investigated. Furthermore, nothing is known regarding peripheral chemoreceptors in snakes and crocodilians. Reptiles other than crocodilians have an undivided ventricle (Burggren, 1987; Hicks, 1998; Farrell et al., 1998), which allows them to shunt blood away from the lung (R-L shunt), reducing arterial oxygen or away from the systemic circulation (L-R shunt), increasing arterial oxygen. Thus, in reptiles adjustments in both ventilation and cardiac shunt are necessary to efficiently regulate arterial oxygen (Hicks and Wood 1989; Wang and Hicks 1996a; Wang et al. 1997; Wood 1982, 1984). The relative importance of the respiratory or cardiovascular systems to the regulation of blood gases depends largely on environmental and physiological conditions (normoxia, hypoxia, hypoxemia and anemia). For instance, a reduction in the R-L shunt increases oxygen delivery to the tissues in anemic and normoxic turtles more effectively than an increase in ventilation (Wang and Hicks 1996a; Wang et al., 1997), since haemoglobin in blood leaving the lung is already fully saturated (Wood 1984). During hypoxia, when pulmonary venous blood is not fully saturated, arterial blood gases are more effectively regulated by increasing ventilation in turtles (Wang and Hicks 1996a). Similarly, anemia in toads increases heart rate, while hypoxia elicits changes in ventilation (Wang et al., 1994; Branco and Glass, 1995). Given these findings, Wang et al. (1994, 1997, 2004) proposed that amphibians and reptiles have anatomically distinct groups of chemoreceptors that sense arterial oxygen tension (PaO2) or oxygen content (CaO2) and adjust the respiratory and cardiovascular systems respectively. To date, this hypothesis that the respiratory and cardiovascular systems are independently regulated by different groups of peripheral chemoreceptors has not been successfully tested. 102  Not only is there a lack of information on the location and reflex roles of peripheral arterial oxygen chemoreceptors in reptiles, very little is known about the neurochemical content of their putative oxygen sensing cells. There have been observations of nests of epithelioid cells in the carotid arch of lizards (Rogers, 1967; Berger et al., 1982) and granulated cell clusters in all of the chemosensory areas of the tortoise (Kusakabe et al., 1988). Furthermore, formaldehyde vapor exposure has revealed an incidence of cells containing biogenic amines (serotonin or catecholamines) in the carotid arch of lizards (Kobayashi, 1971a) and the truncus and aorta of chelonians (Chiba and Yamauchi, 1973; Ishii et al., 1985b). More recently, two populations of putative chemoreceptor cells have been described in turtles, containing either vesicular acetylcholine transporter (VAChT) or serotonin (5HT) (Chapter 3). To date, however, no specific neurotransmitters have been identified in other reptiles. Given the foregoing, the objectives of the present study were to (1) locate functional chemosensory areas in rattlesnakes, (2) characterize the cardio-respiratory reflex responses of each area to a hypoxia related stimulus (cyanide, NaCN) and (3) identify putative oxygen sensing cells based on their neurochemical content and anatomical features. I hypothesized that as in frogs and turtles, rattlesnakes would have three distinct chemosensory areas located at the carotid bifurcation and bases of the aortic arch and pulmonary artery, and that each would possess glomus-like cells with similar neurotransmitter profiles to the chemosensory areas described in turtles. Furthermore, I hypothesized that aortic and pulmonary chemoreceptors would primarily regulate the degree of intra-cardiac shunt, while carotid chemoreceptors would primarily regulate ventilation. To test these hypotheses, I located potential chemoreceptor areas through their innervation and association with derivatives of pharyngeal arches and quantified the cardio-respiratory reflex responses to NaCN stimulation at each site. I verified that physiological adjustments were produced by each chemosensory site by selectively denervating each area and reapplying the stimulus. To identify putative chemoreceptor cells I used markers for four neurotransmitters: acetylcholine (ACh), the main neurochemical involved in signal transduction in mammals; tyrosine hydroxylase (TH), the rate limiting enzyme in catecholamine synthesis (Nurse, 2005, 2010); 5HT, the principle neurotransmitter present in the neuroepithelial cells (NEC’s) of fish (Dunel-Erb et al., 1982; Bailly et al., 1992; Zaccone et al., 1992; 1994, 1997; Jonz and Nurse, 2003; Jonz, et al. 2004; Bailly, 2009; Jonz and Nurse, 2012) and Human natural killer 1 (HNK-1), a marker for proliferative cells that develop from neural crest cells. By 103  establishing the location, neurochemical profiles and regulatory roles of peripheral chemoreceptors in snakes I hoped to broaden the understanding of phylogenetic patterns of chemosensing seen in vertebrates. My goal was to determine if cell morphology, size and neurochemical content of O2-sensing structures was highly conserved throughout vertebrates, and whether the location and reflex roles of distinct chemoreceptor groups have changed to more effectively control ventilatory and cardiovascular functions in animals with central vascular shunts.  4.3  Materials and methods  4.3.1  Animals and holding conditions  South American rattlesnakes (Crotalus durissus Linnaeus 1758) (mean mass = 1.4 ± 0.08 kg, N=22) were obtained from the Butantan Institute in São Paulo and transported to Universidade Estadual Paulista (UNESP), Rio Claro, SP, Brazil. The snakes were housed in a vivarium at room temperature (~28°C) under a natural photoperiod. The holding and experimental procedures followed Canadian Council on Animal Care guidelines and were approved by the University of British Columbia Animal Care Committee (animal care certificate No. A09-0233).  4.3.2  Surgery and instrumentation  Snakes were anaesthetised by inhalation of CO2 until all righting reflexes disappeared. The animals were intubated and artificially ventilated using a mechanical ventilator (Harvard Apparatus, Holliston, MA, USA) set at a frequency of 6 breaths per minute and a tidal volume of 25 ml/kg. During surgery anaesthesia was maintained with Isoflurane (0.5-1%; Baxter Healthcare Corporation, Deerfield, IL, USA). A 7-8 cm ventral incision was made cranial to the heart and two 1.5-2R and 2.5R ultrasound, transit-time flow probes (Transonic System, Inc., NY, USA) were placed around the pulmonary artery and left aorta to measure pulmonary (Q̇ pul) and systemic (Q̇ sys) blood flows, respectively. Acoustical gel was infused around the flow probes to enhance the signal. The flow probes were connected to a dual-channel blood flow meter (Transonic T206). Arterial catheters filled with heparinised saline were occlusively placed in a side branch of the pulmonary and caudal arteries for measurements of pulmonary (Ppul) and systemic (Psys) blood pressures, respectively. Both catheters were connected to pressure 104  transducers and the signals were amplified with a Gould D.C. amplifier. Both pressure transducers (systemic blood pressure: Narco Biosystems, Narco Scientific, Houston, TX, model 320-1000E; pulmonary blood pressure: Statham, Hato Rey, Puerto Rico, model P23Db) were placed at the level of the heart and were calibrated daily using a barometer. Heart rate was derived from the blood pressure traces and ventilation was inferred from impedance measured with two sets of needle electrodes placed subcutaneously. Snakes were divided into two experimental groups. In Group 1 (N=10, aorta and N=12, pulmonary artery): for site-specific stimulation of putative aortic and pulmonary chemoreceptors PE50 and PE10, polyethylene catheters filled with heparinised saline were advanced into the right aortic arch through the vertebral artery and into the pulmonary artery at the base of the heart through a small branch of the pulmonary artery, respectively. In Group 2 (N=10), the left carotid was non-occlusively cannulated (percutaneous cannula) and the catheter was advanced towards the carotid bifurcation, for specific stimulation of putative carotid chemoreceptors. The catheters and leads of each probe were secured to the skin with suture and local analgesia (xylocaine) was applied to the incision. Animals were placed in dark plastic boxes and allowed to recover overnight. All experiments were performed at room temperature on fully recovered snakes. At the end of the experiment animals were euthanized and the position of the catheters was verified. The left carotid bifurcation, pulmonary artery, right and left aortae, jugular ganglion and the lungs of five animals were collected for immunohistochemical analysis.  4.3.3  Denervation of carotid or pulmonary and aortic branches of the  glossopharyngeal/vagus nerve trunk Denervation studies were performed to confirm that the responses observed after focal stimulation with NaCN originated from peripheral chemoreceptor activation. In Group 1, branches of the IX/X nerve innervating the pulmonary artery and aortic arch were severed, while the IX/X truncii and the branch that innervates the sphincter of the pulmonary artery were left intact. It was impossible to discriminate between the small branches of the IX/X cranial nerve that innervate the pulmonary artery and aorta, which run together. In Group 2, branches of the IX/X nerve innervating the carotid bifurcation and nearby areas were severed and the main trunk of the IX/X nerve was left intact, while a heavily vascularised and innervated epithelial body sitting at the carotid bifurcation was excised. 105  4.3.4  Experimental protocol  The animals received focal injections of NaCN (1%, volume injected 0.1 ml; 380970, Sigma Aldrich, USA), saline (2 ml; NaCl, S7653, Sigma-Aldrich, USA) and blood (2 ml) through the catheters in the vertebral and side branch of the pulmonary artery (Group 1), and the carotid artery (Group 2). The order of the injections was randomised and at least 20 minutes elapsed between injections so that blood flow, heart rate and blood pressure returned to baseline. Injections of NaCN and blood were followed by 0.8 ml of saline to ensure that no NaCN or blood remained in the cannula. This value was determined by measuring the volume of the catheters. NaCN injections were used to determine the presence of chemoreceptors in the putative chemosensory areas. The volume of NaCN infused in the cannula was small (0.1 ml) to prevent mechanoreceptor stimulation. A dose-response curve was performed to determine the concentration of NaCN to be used and sham saline injections (0.9 ml) confirmed that the responses observed were not an artefact of the injection volume. Blood and saline were used to test whether putative chemosensory sites responded to low CaO2 or low oxygen tension (PO2). Blood was withdrawn from the caudal arterial catheter and rapidly infused into either the aorta, pulmonary or carotid arteries. Withdrawal of blood did not cause any changes in cardio-respiratory variables. The volume and duration of both blood and saline injections were the same (2 ml; 30-45 s).  4.3.5  Data analysis and statistics  Mean blood pressures, blood flows, heart rate, tidal volume and breathing frequency were recorded over a period of 10 minutes before the stimulus was applied and until all physiological variables returned to baseline using WINDAQ acquisition software (version 2.19, Dataq Instruments, Akron, OH, USA) sampling at a rate of 250 Hz per channel. With only one pulmonary artery, measurements of blood flow in the rattlesnake pulmonary artery (Q̇ PA) represent Q̇ pul. Systemic flow (Q̇ sys) was estimated as 3.3 times Q̇ LAo (Galli et al. 2005). Total cardiac output (Q̇ tot) was calculated as Q̇ sys + Q̇ pul. Shunt pattern was calculated as Q̇ pul/Q̇ sys. Total stroke volume (Vstot; pulmonary + systemic) was calculated as Q̇ tot/fH (where fH is heart rate calculated from the blood pressure trace). Systemic (Rsys) and pulmonary (Rpul) resistances were calculated as Psys/Q̇ sys and Ppul/Q̇ pul, respectively, assuming that central venous blood 106  pressures were negligible. Total ventilation (V̇ Tot; amplitude (VAMP) x breathing frequency (fR)) and VAMP are expressed as percentage values relative to the pre-injection values (control). All cardiovascular variables were averaged over the 2-min before NaCN injections and over 1-min intervals after the injection. Respiratory data were averaged over the 4-min period before stimulus application and for 2-min intervals after the injection. All values are presented as mean ± s.e.m. I tested for differences between control values (prior to injection) and values after NaCN, saline or blood injections on blood flows, blood pressures, heart rate and breathing frequency with a one-way repeated measures ANOVA. Holm-Sidak multiple comparisons tests were used to determine pairwise differences. Data that did not meet the assumptions of normal distribution or equal variances were natural log (Ln) transformed or tested for significant differences with a non-parametric Friedman repeated measures analysis of variance. I tested whether the mean of the post-injection values (relative values) for breathing amplitude and total ventilation were higher than a 100% (control) using a one sample t-test. Sigma Stat (version 3.11, Systat Software, IL, USA) or R (R version 2.11.1, R development core team, 2010) were used for all statistical analyses.  4.3.6  Tissue preparation  The right and left carotid bifurcation and segments of the left common carotid artery, pulmonary artery and right and left aortae (N=5) were removed and flushed with heparinised saline (100 UI/ml) using a blunt 21 gauge needle connected to a syringe until the blood vessels appeared clear of blood. The jugular ganglia and lungs of two animals were also collected to use as positive immunohistochemical controls (see Controls section). The tissue was pinned, immersed in paraformaldehyde (PFA; 4%) in 0.1 M phosphate buffer saline (PBS; Na2HPO4, 13.4 g/l; NaH2PO, 6 g/l; NaCl, 9 g/l; buffered to pH of 7.4 with NaOH). The tissue was then washed in PBS and cryoprotected in 30% sucrose buffer and frozen in Tissue Tek (Sakura, San Marcos, CA, USA) at -80 °C. Longitudinal sections (18 µm thick, same procedure as in Chapter 2, Fig. 2.1A) were made using a cryostat and serially mounted on Superfrost plus slides (VWR International, West Chester, PA, USA). Slide-mounted sections were immediately processed for immunohistochemistry or stored in a -80 ºC freezer until needed.  107  4.3.7  Immunohistochemistry  Techniques for immunolabeling, controls and imaging were performed as described in Chapter 2 and Chapter 3. Briefly, slide-mounted tissue was washed in PBS, the buffer was pipetted in and out of the glass holders to prevent the delicate internal structure of the arteries from breaking off from the slides. Following the washes, the sectioned tissue was blocked in 10% normal donkey serum (NDS) (Jackson Laboratories, distributed by Cedarlane Laboratories, Hornby, ON, Canada) for 1 h. Primary antibodies (mouse, anti-TH; goat, anti-5HT, ImmunoStar; rabbit, anti-VAChT, Novus Biological) were diluted (PBS, 0.3% Triton X-100, 2% NDS) according to optimal dilutions determined previously by dilution curves (Table 4.1). Slides were incubated with the primary antibody (individually or in combination) for 48 h at room temperature and then washed in PBS. After the washes, slides were incubated in the dark with fluorescently labelled secondary antibodies diluted in PBS (with 0.3% Triton X-100 and 2% NDS) (Table 4.1) for 2 h and subsequently washed in PBS. DAPI was used to visualize cell nuclei (Vectashield with DAPI, Vector Laboratories, Burlington, Ontario). Coverslips (#1.5, Fisher Scientific, Ottawa, ON, Canada) were mounted with Vectashield (Vector Laboratories, Burlington, Ontario) to reduce photobleaching and the coverslips were sealed with nail polish. Processed slide-mounted tissue was stored at 4 ºC in the dark prior to imaging.  4.3.8  Controls  Controls consisted of excluding the primary antibody to control for non-specific binding of the secondary antibody. To control for interactions between antibodies, single labelled slides were processed in each run. Positive controls for VAChT and 5HT primary antibodies were performed using the jugular ganglion and lungs, respectively. The specificity of the primary antibodies was also tested by examining the labeling pattern of two different antibodies raised in the same host species (i.e. the polyclonal antibodies for VAChT and 5HT raised in rabbit) The specificity of the goat anti-5HT primary antibody had been verified by the supplier on rat hypothalamus and spinal cord using a preabsorption control. Labeling was eliminated by pretreatment of the diluted antibody with 100 µg/ml of serotonin. See Table 4.1 for antibody and supplier details.  108  4.3.9  Microscopy for cryo-sectioned tissue  Sectioned tissue was observed using an epifluorescence light microscope Zeiss AxioObserver Z1 (Carl Zeiss, USA) connected to a halide lamp and equipped with 488049-9901 FL Filter Set 49, 489038-9901 FL Filter Set 38 HE and 489043-9901 FL Filter Set 43 HE filters to detect DAPI, GFP and Cy3 respectively. Images were captured using AxioVision software (Rel. 4.8.3, Carl Zeiss Microscopy, USA). Representative sections were further examined using a spinning disk microscope (Perkin Elmer Ultraview VOX Spinning Disk Confocal, Waltham, MA, USA), equipped with 405 nm, 488 nm and 561 nm lasers and filters 527/55, 445/60 or 615/70, 525/50 or 640/120 for detection of GFP, DAPI and rhodamine. Z-stacks of 117-179 optical sections and 0.2 µm apart were captured using Leica multi-immersion 20x and 63x glycerol objectives and a Hamamatsu C9100-50 camera. Images were analyzed using ImageJ.  4.3.10 Quantification: cell size I measured the diameters of different cell types to further characterize candidate oxygen sensing cells. Only cells where I could identify the nuclei and clearly see the entire labeled cytoplasm were used for quantification. Ten 5HT- and ten vesicular acetylcholine transporter immunoreactive (VAChT-IR) cells were measured in different sections of the carotid bifurcation, right aorta and pulmonary artery of 4 animals using ImageJ software. All data are presented as average cell size (µm) ± s.e.m.  4.4  Results  4.4.1  Anatomy and innervation of the central vasculature  In rattlesnakes, the left and right aortae and the single pulmonary artery all arise directly from the ventricle (Fig. 4.1A, C). The common carotid arteries both arise from the right aorta. The right common carotid artery is truncated and only advances a short distance before terminating. It does not continue to the head region. The left common carotid artery continues rostrally and bifurcates into the internal and external carotid arteries at the base of the head (Fig. 4.1A, B). It also gives rise to an anastomosis that connects the left common carotid artery to the rostral extension of the right common carotid artery at the base of the skull. Thus, even though the right common carotid artery does not extend from the heart to the head, the right internal and external carotid arteries persist and receive blood flow from the left common carotid artery. 109  To identify potential chemosensory areas in the snakes I traced the innervation of the IX/X cranial nerves to the carotid bifurcation, aortae and pulmonary artery, which are derivatives of pharyngeal arches. In rattlesnakes the IX and X cranial nerve trunks unite after leaving the skull. Four or more small branches from the IX/X nerve trunk innervate the carotid bifurcation by the head and a vascularised small glandular body that sits at the carotid bifurcation (Fig. 4.1A, B, D). The left IX/X nerve continues caudally alongside the left common carotid artery (Fig. 4.1E, arrowhead) while the right IX/X nerve supplies the remnant of the right common carotid artery. Three tracheal branches arise from the IX/X nerve trunk to innervate the trachea. Before innervating the heart, numerous small branches arise from the IX/X nerve to innervate the aortae and pulmonary artery in the region proximal to the heart (Fig. 4.1F). Thus, the carotid bifurcation, aortae and the pulmonary artery are extensively innervated by the IX/X nerve trunk.  4.4.2  Chemosensory areas in rattlesnakes  Nerve tracing indicated that the left carotid bifurcation, aortic arch and pulmonary artery could all be functional chemosensory areas in rattlesnakes. To test this I recorded cardiorespiratory variables while injecting each site individually with NaCN, a hypoxic mimetic. I determined that the cardio-respiratory adjustments produced by chemoreceptor stimulation were not due to an injection effect, as saline injections of 0.9 ml (same volume as NaCN injections) did not trigger changes in any of the cardiovascular or respiratory variables. I confirmed that the cardio-respiratory reflexes were produced by stimulation at specific chemoreceptor sites by the absence of a reflex response after NaCN stimulation in animals where the small glossopharyngeal/vagal branches innervating either the carotid bifurcation or the aorta and pulmonary artery were severed (N=4, carotid; N=3, aorta and pulmonary artery). Furthermore, I verified that the ability to shunt was not affected after denervation of a particular chemoreceptor site, by stimulating intact chemoreceptor groups to elicit a reflex response in some of the snakes.  110  4.4.2.1  Effects of chemoreceptor stimulation on cardiovascular control  4.4.2.1.1  Cardiovascular control by peripheral chemoreceptors in the carotid  bifurcation Injections of NaCN into the carotid artery caused Q̇ sys to fall and Q̇ pul to rise (Q̇ sys: P<0.001; Q̇ pul: P=0.015, one-way RM ANOVA) (Fig. 4.2A, B), leading to a reduction in the right to left (R-L) shunt (increase in Q̇ pul/Q̇ sys) (P<0.001, one-way RM ANOVA) (Fig. 4.3A). fH increased significantly (P<0.001, one-way RM ANOVA) (Fig. 4.2C). Vstot (Table 4.2) and Q̇ tot (Fig. 4.3B) remained unchanged after carotid chemoreceptor stimulation (Vstot: P=0.856, Friedman repeated measures analysis of variance; Q̇ tot: P=0.591, one-way RM ANOVA). NaCN injections did not affect systemic mean arterial pressure (MAP) (P=0.177, one-way RM ANOVA), but caused a significant fall in systemic pulse pressure (P<0.001, one-way RM ANOVA) (Table 4.2). Rsys was elevated compared to pre-injection values (P<0.001, one-way RM ANOVA) (Fig. 4.3C). After denervation of the carotid bifurcation, NaCN injections did not elicit any reflex response (Fig. 4.4A, B). This result confirms that the cardiovascular changes I saw were consistent with chemoreceptor activation in the carotid bifurcation.  4.4.2.1.2  Cardiovascular control by peripheral chemoreceptors in the aorta NaCN injections in the aorta led to a significant decrease in Q̇ sys (P<0.001, one-way RM  ANOVA), while Q̇ pul did not change (P=0.645, Friedman repeated measures analysis of variance) (Fig. 4.5A, B). This resulted in a significant increase in Q̇ pul/Q̇ sys and a reduction in the R-L shunt (P<0.001, one-way RM ANOVA) (Fig. 4.3D). Although the immediate response to NaCN was a slight fall in fH and skipped heart beats in some of the snakes; a marked tachycardia and reduced Vstot were observed after three minutes of stimulus application (fH: P=0.013; Vstot: P=0.004, one-way RM ANOVA) (Fig. 4.5C; Table 4.2). Q̇ tot, and MAP did not change (Q̇ tot: P=0.528; MAP: P=0.649, one-way RM ANOVA) (Fig. 4.3E; Table 4.2). NaCN caused a reduction in systemic pulse pressure (Table 4.2) and an increase in Rsys (Pulsesys: P<0.001; Rsys: P=0.002, one-way RM ANOVA) (Fig. 4.3F). After denervation of the aorta all cardiovascular responses to NaCN were abolished (Fig. 4.4C, D), consistent with activation of aortic chemoreceptors.  111  4.4.2.1.3  Cardiovascular control by chemoreceptors in the pulmonary artery  NaCN stimulation of pulmonary chemoreceptors caused a significant fall in both Q̇ sys and Q̇ pul (Q̇ sys: P<0.001; Q̇ pul: P=0.041, one-way RM ANOVA) (Fig. 4.6A, B). Since Q̇ sys was reduced proportionately more than Q̇ pul, shunt fraction increased significantly and the R-L shunt was reduced (P=0.012, one-way RM ANOVA) (Fig. 4.3G). The fall in Q̇ sys was presumably caused by the increase in Rsys (P=0.014, one-way RM ANOVA) (Fig. 4.3I). As in the other two areas, NaCN injections led to a significant increase in fH, although only after a slight fall in the first two minutes following NaCN stimulation (P<0.001, one-way RM ANOVA) (Fig. 4.6C). Stimulation of chemoreceptors in the pulmonary artery led to a decrease in Vstot (Table 4.2), Q̇ tot (Fig. 4.3H), MAP and systemic pulse pressure (P<0.001, one-way RM ANOVA for Vstot, Q̇ tot, MAP and Pulsesys) (Table 4.2). All cardiovascular responses to NaCN were abolished after denervation of the pulmonary artery (Fig. 4.4E, F), consistent with activation of pulmonary chemoreceptors.  4.4.2.2  Effects of chemoreceptor stimulation on respiratory control  NaCN injection into all three areas caused V̇ Tot to increase (carotid: P=0.004 aorta: P=0.002; pulmonary: P=0.007, one sample t-test) (Fig. 4.7). In the carotid and aortic groups changes in V̇ Tot were due to significant increases in fR (carotid: P=0.012, Friedman repeated measures analysis of variance; aorta: P=0.002, one-way RM ANOVA) and VAMP (carotid: P=0.012; aorta: P=0.007, one sample t-test) (Fig. 4.8A, C). The increase in V̇ Tot in response to NaCN injections in the pulmonary artery was caused mainly by a rise in VAMP (P=0.0001, one sample t-test) (Fig. 4.8E, F). After selective denervation of the chemosensory sites V̇ Tot did not change. In fact NaCN injections in the aorta of some denervated rattlesnakes inhibited V̇ Tot, by a mechanism that is unknown (Fig. 4.9).  4.4.3  Stimulus specificity of peripheral chemoreceptors  A bolus injection of saline (2 ml, 30-45 s infusion time), which has a low CaO2, but normal PaO2 did not have any effect on blood flows, shunt pattern, fH, MAP or V̇ Tot (P>0.05, Holm-Sidak pairwise comparisons) (Table 4.3) when injected into any of the three sites. Control injections of blood (2 ml, 30-45 s infusion time) with normal PaO2 and CaO2 did not elicit cardiovascular or ventilatory adjustments either. 112  4.4.4 4.4.4.1  Neurochemical content of the chemosensory areas Controls for immunohistochemistry  In snakes, neurons in the jugular ganglion labeled for VAChT and neuroepithelial bodies in the lungs labeled for serotonin. Furthermore, the labeling patterns of the VAChT and 5HT primary antibodies raised in rabbit were very different, when tested in the carotid bifurcation, aorta and pulmonary artery of snakes. Although, I did not have this type of specificity control for the goat anti-5HT antibody in snakes, the pattern of labeling was consistent with that of the rabbit anti-5HT antibody. In addition I confirmed the specificity of the goat anti-5HT antibody in the corresponding tissues of frogs and turtles (see Chapters 2 and 3). Omission of the primary antibody did not result in detectable labeling in any of the immunohistochemical runs, indicating that all antibodies were likely specific in snakes. Results from single and double labeled slides did not differ in immunolabeling patterns. These controls confirm that the procedures and antibodies used in this study are effective and likely specific in snakes.  4.4.4.2  Neurochemical content of the carotid bifurcation  The internal structure of the common carotid artery at the carotid bifurcation consists of spongy tissue surrounding a series of sinusoids (Fig. 4.10A). I found numerous single 5HT-IR cells embedded in the vessel wall at the carotid bifurcation (Fig. 4.10F; arrowheads B and E) and a lower density of 5HT-IR cells in the common carotid artery just proximal to the bifurcation (Fig. 4.10D). 5HT-IR cells were oval and 11.7 ± 0.3 µm long (Fig. 4.10G). Serotonergic cells were not found in the common carotid artery between the heart and the area just proximal to the bifurcation (data not shown). I saw a VAChT-positive cell in only one animal (N=5) and in this one case both neurotransmitters (ACh and 5HT) were present in the same cell.  4.4.4.3  Neurochemical content of the aorta  I found two cell populations in the aorta of rattlesnakes. A cell type labeled for VAChT and was characterized by clusters of oval cells with large nuclei averaging 10.6 ± 0.2 µm in diameter (Fig. 4.11D, E). They were found in the spongy tissue inside the lumen of the blood vessel (Fig. 4.11C, arrow showing general location). A 5HT-positive cell type was characterized by oval cells arranged singly or in groups of 2 or 3 (12.0 ± 0.3 µm) (Fig. 4.11F). These cells were found embedded in the vessel wall throughout the aorta (Fig. 4.11 B arrowheads; G), but their density 113  was higher in the spongy tissue close to the heart (Fig. 4.11 A, B arrowheads; H). Although both cell types never colocalized, some 5HT-IR cells were found in the same area as VAChT-IR cell clusters.  4.4.4.4  Neurochemical content of the pulmonary artery  I found cells containing VAChT in the pulmonary artery where the artery begins to branch (Fig. 4.12A, arrow; E). These cells were oval with large nuclei (10.7 ± 0.2 µm) and were arranged in large clusters close to the lumen of the vessel (Fig.4.12F). Numerous 5HT-IR cells occurred along the first few centimeters of the artery where it emerges from the heart embedded in the vessel wall (Fig. 4.12A, B, arrowheads). 5HT-IR cells were oval and 11.7 ± 0.3 µm long. They were mostly arranged singly, but occasionally were found in groups of 2 or 3 (Fig. 4.12G, H). As in the other chemosensory areas described above, 5HT and VAChT did not colocalize to the same cells.  4.4.5  Neural crest origin of putative oxygen sensing cells  A few cells labeled with the HNK-1 marker were observed in the adventitia of the aorta, carotid bifurcation and pulmonary arteries of rattlesnakes (data not shown). None of these cells were positive for 5HT or VAChT, indicating that these cells are either not derived from the neural crest or that they are mature, non-proliferative cells.  4.5  Discussion This study identified three functional peripheral chemoreceptive sites in the rattlesnake and,  for the first time, defined the regulatory roles (cardiovascular or respiratory) of the different chemosensory sites. It also identified putative oxygen sensing cells in these regions based on the presence of neurotransmitters involved in oxygen chemotransduction in other vertebrates. The three O2-chemosensory sites were the area surrounding the carotid bifurcation, and the bases of the aortic arch and pulmonary artery. In response to NaCN (mimicking low O2 levels in the blood), rattlesnakes increased ventilation and reduced the R-L intra-cardiac shunt (i.e. they decreased the fraction of blood that bypasses the lung). All chemoreceptors were effective in regulating shunt and ventilation. Putative oxygen sensing cells containing 5HT were present in all chemosensory areas, but those containing VAChT were only found in the aorta and 114  pulmonary artery. The results of this study allowed me to compare the anatomy and function of chemosensory sites of rattlesnakes with other vertebrates.  4.5.1  Limitations of the study  I was unable to separately denervate the aorta and pulmonary artery because the glossopharyngeal/vagal innervation to these areas consists of numerous small branches that run closely together. I therefore cannot rule out the possibility that the reflex response produced by the pulmonary artery was not caused by recirculation of NaCN to the aorta. However, the similar response time and the different cardiovascular and ventilatory effects of pulmonary and aortic injections suggest that this is not the case. Furthermore, given the small concentration and volume of the NaCN injections, the bolus would have become very diluted by the time it reached other chemosensory areas. The ratio between pulmonary and systemic resistance is an important determinant of cardiac shunt (Overgaard et al., 2002). Unfortunately I was unable to record pulmonary blood pressure, due to difficulties with the pressure transducer. In the absence of pulmonary blood pressures and resistance measurements, all inferences about shunt control are deduced from the effects of the stimulant on systemic resistance alone. Tissue from five snakes was collected after termination of the experiments, fixed in PFA and shipped to San Diego, CA where it was processed for immunohistochemistry and histology. The tissue remained in fixative for up to 7 days (depending on the day of collection). I tried to minimize over-fixation of the tissue by collecting it from snakes towards the end of the study and maintaining it at 4 ºC until shipped. Although I obtained positive immunohistochemical results, some antigenicity may have been lost and image sharpness reduced.  4.5.2  Comparison with other studies  Cardiovascular variables under control conditions in the present study were comparable to those obtained in other studies on unanaesthetised and resting rattlesnakes at room temperature (Wang et al., 2001a; Skals et al., 2005; Taylor et al., 2009). The presence of a net R-L or L-R shunt in resting snakes under control condtions was variable. A characteristic net R-L shunt as seen in resting turtles and toads (Shelton and Burggren, 1976; Overgaard et al., 2002; Krosniunas and Hicks, 2003; Andersen et al., 2003) was 115  observed in some rattlesnakes, while in others a L-R shunt prevailed as reported in unanaesthetised resting rattlesnakes (Taylor et al., 2009). The net shunt, regardless of direction, was small. Nevertheless, the direction and magnitude of the cardiac shunt prior to the stimulus could potentially affect the response to NaCN injections. In the presence of central vascular shunts, arterial systemic blood is composed of systemic venous blood (deoxygenated blood) and blood returning from the lungs (oxygenated blood) (Wang et al., 1997). Thus, a R-L shunt will reduce the levels of oxygen in the arterial systemic blood (Wang and Hicks, 1996a), which will presumably stimulate peripheral chemoreceptors and potentially affect their sensitivity to a new stimulus, such as NaCN. This has been reported in turtles where NaCN injections in the pulmonary artery increased ventilation, but NaCN did not have any effect on ventilation after exposure to hyperoxia for one hour (Benchetrit et al., 1977).  4.5.3  Location of peripheral chemoreceptor in rattlesnakes  I identified multiple chemosensory areas in rattlesnakes and confirmed the chemosensory role of these areas by the absence of a reflex after selectively denervating and stimulating each site. Sham injections of saline (same volume and time of injection) further confirmed that the response was not due to stimulation of mechanoreceptors. I also verified that the IX/X nerve trunk remained intact after denervation of the different sites by stimulating a non-denervated area and confirming that changes in heart rate and blood flows still occurred. Peripheral chemoreceptive areas in rattlesnakes were located at the carotid bifurcation, and at the bases of the aorta and pulmonary artery, derivatives of the 3rd, 4th and 6th pharyngeal arches, and were innervated by the IX/X nerve trunk. Peripheral chemoreceptors are found in these same areas in turtles and toads, also innervated by the IX and/or X nerves (Ishii et al., 1966; Benchetrit et al., 1977; Lillo, 1980; Hoffman and de Souza, 1982; Ishii et al., 1985a, b; Ishii and Ishii, 1986). In rattlesnakes, as in toads chemoreceptive cells are present at the carotid bifurcation, a site homologous to the location of the mammalian carotid bodies. I did not find any visible innervation or evidence of putative oxygen sensing cells containing 5HT at other sites along the common carotid artery. In turtles, putative chemoreceptor cells are found in the common carotid artery at the base of the neck just after the carotid artery branches from the subclavian artery. In the case of the turtle, however, this site is homologous to the carotid bifurcation in other species as the extant carotid bifurcation in turtles is secondarily derived 116  during development (Adams, 1958). In all lower tetrapods studied to date (amphibians and reptiles), all the extant derivatives of the pharyngeal arches contain chemosensitive sites. This also appears to be the case in rattlesnakes. It has been proposed that multiple O2 sensing chemoreceptors in distinct anatomical locations may be advantageous to animals with the ability to shunt, as distinct chemoreceptor groups will be exposed to blood with different composition (Wang et al., 1997; Milsom and Burleson, 2007). For instance, carotid and aortic chemoreceptors will be exposed to arterial blood with oxygen levels that will depend on the degree of admixture, while chemoreceptors in the pulmonary artery will be exposed to mixed venous blood and venous oxygen levels will be determined by systemic oxygen delivery (blood flow and CaO2) relative to tissue oxygen uptake (Wang et al., 1997). 4.5.4  Regulatory roles and stimulus specificity of chemoreceptors groups  Injection of NaCN into the regions of the carotid bifurcation and base of the aorta and pulmonary artery of rattlesnakes elicited cardio-respiratory responses. The reflex responses in the present study were abolished after denervation of each putative chemosensory site, indicating that they were not caused by recirculation to other sites.  4.5.4.1  Effects of chemoreceptor stimulation on respiratory control  NaCN injections at each of the three chemosensitive sites in the rattlesnakes increased ventilation. Changes in both breathing frequency and tidal volume were responsible for the increase in ventilation associated with injections into the carotid and aortic chemosensitive areas, while the effect of injections into the pulmonary chemoreceptor area was produced by changes in tidal volume alone. The magnitude of the ventilatory response and the respiratory pattern (tidal volume or breathing frequency) used to produce this response under conditions of low oxygen are highly variable among different species of reptiles (Shelton et al., 1986). For instance, garter snakes exposed to hypoxia increase ventilation mainly by increasing tidal volume accompanied by a small rise in breathing frequency (Bartlett, Jr. and Birchard, 1983). In contrast, the hypoxic ventilatory response (HVR) of the diamondback water snake consists of an increase in breathing frequency and a fall in tidal volume, such that the increase in ventilation is small (Gratz, 1979). My findings indicate that the HVR in the rattlesnake consists of an amplitude and frequency  117  response and that the net effect arises from a balance of changes to the breathing pattern produced by each chemoreceptor group. The ventilatory response of pulmonary and carotid chemoreceptor stimulation was faster and slightly larger than that of the aortic chemoreceptors. The pulmonary artery of turtles has also been found to be an important chemosensory site, as NaCN injections in this area had a major effect on ventilation compared to the effects of stimulating the aorta and truncus arteriosus (Benchetrit et al., 1977). The time course of the contributions by each site also varied with that arising from stimulation of the carotid chemosensitive area being more transient.  4.5.4.2  Effects of chemoreceptor stimulation on cardiovascular control  NaCN injections at all chemosensory sites caused a fall in Q̇ sys, and increases in fH, Rsys and Q̇ pul/Q̇ sys, so that the R-L shunt was reduced. There were, however, some differences in the reflexes produced by the distinct chemoreceptive sites. Stimulation of the carotid and aortic chemosensory areas produced a rise in Q̇ pul, but no change in blood pressure or Q̇ tot. The cardiovascular effects of NaCN injections at pulmonary chemoreceptive sites were large and consisted of a fall in Q̇ pul, Q̇ tot and blood pressure. A reduction of the R-L shunt during hypoxia has been well documented in reptiles (Wang and Hicks, 1996a; Andersen et al., 2003), therefore it is not surprising that NaCN stimulation of all chemosensory sites in rattlesnakes caused a reduction in shunt. Changes in shunt pattern are effective in adjusting arterial blood gases in reptiles and amphibians (Burggren and Shelton, 1979; Wood, 1982, 1984; Burggren, 1988; Burggren et al., 1989; Wang and Hicks, 1996a; Wang et al., 1997).  4.5.4.2.1  Mechanisms of cardiac shunt regulation  The reductions in R-L shunt after chemoreceptor stimulation in this study were consistently achieved by a drop in Q̇ sys accomplished through an increase in systemic resistance. The increase in Rsys and concomitant reduction in the Q̇ sys has been attributed to increased sympathetic tone in anaesthetised rattlesnakes and turtles (Overgaard et al., 2002; Galli et al., 2007). Presumably, the increase in Q̇ pul after stimulation at the carotid and aortic chemoreceptive sites resulted from a fall in Rpul. Unfortunately, I was unable to measure Rpul in this study, but it is well known that regulation of cardiac shunts in reptiles is mainly achieved by cholinergic 118  vagal control of the smooth muscle surrounding the pulmonary artery (Burggren, 1977, Milsom et al., 1977; Hicks and Comeau, 1994; Hicks, 1998; Wang et al., 2001b). Adrenergic stimulation of the systemic circulation and release of vagal tone on the pulmonary circulation together will decrease Q̇ sys and increase Q̇ pul, respectively, reducing the R-L shunt (Hicks, 1994; Comeau and Hicks, 1994; Overgaard et al., 2002). The reduction in the R-L shunt in this study was achieved by a redistribution of blood flows such that cardiac output did not change. Under these conditions the oxygen content of systemic arterial blood should have increased while systemic flow decreased. Activation of pulmonary chemoreceptors, however, significantly reduced both blood flows and Q̇ tot. Since the fall in Q̇ sys was larger than the fall in Q̇ pul the net effect was still a reduction in the R-L shunt. Pulmonary chemoreceptors regulated arterial oxygen by adjustments in total blood flow and the degree of shunt. I do not know what the net effect would be of stimulation at all three sites simultaneously in the rattlesnakes but my findings do indicate that regulation of Q̇ sys plays a key role in control of net cardiac shunt in rattlesnakes.  4.5.4.3  Response time of chemoreceptor stimulation  Changes of cardiovascular and respiratory variables after chemoreceptor stimulation generally started after a short delay (30 and 50-120 s, respectively). However, maximum (significant) effects were not reached until later and this varied between physiological variables and between chemoreceptive sites. The time needed to reach a maximum change in ventilatory and cardiovascular variables in the present study was longer than the recirculation time, but denervation of specific chemosensory areas allowed me to exclude the possibility of stimulation of other areas by recirculation of NaCN. The effects of NaCN infusion into the chemosensory areas of rattlesnakes were also delayed compared to those of NaCN injections into the carotid labyrinth and pulmocutaneous artery of frogs and toads (Lillo, 1980; Wang et al., 2004). Species differences, the concentration of NaCN used, or the fact that the earlier experiments were performed on lightly anesthetized frogs and decerebrated and unidirectionally ventilated toads could explain the differences in time response between studies. However, in anaesthetised turtles, maximum increase in nerve discharge occurred after 1-1.5 minutes post-NaCN injection (Ishii et al., 1985b). The time response in 119  anesthetised turtles is significantly longer than the emergence of a reflex response in anesthetised and decerebrated amphibians, suggesting that in reptiles both signal chemotransduction (turtles) and reflex production (rattlesnakes) take longer than in other vertebrates. 4.5.4.4  Reflex roles of peripheral chemoreceptors  One of the objectives of this study was to determine whether distinct chemoreceptor groups specifically controlled the cardiovascular versus respiratory systems in the rattlesnake. I hypothesized that aortic and pulmonary chemoreceptors would primarily regulate the degree of intra-cardiac shunt, while carotid chemoreceptors would primarily regulate ventilation. Here I demonstrate that all three anatomically separate peripheral chemoreceptive sites in rattlesnakes have similar reflex roles in cardiovascular and respiratory control. Previous studies on toads and turtles showed that reductions in arterial oxygen carrying capacity produce an increase in heart rate and pulmonary blood flow while ventilation remained unchanged, and a decrease in PaO2 increased ventilation (Wang et al., 1994; Branco and Glass, 1995; Wang et al., 1997; Andersen et al., 2003). Several studies found correlations between the hypoxic ventilatory response in reptiles and changes in CaO2 rather than PO2 (Glass et al., 1983). To explain these observations, Wang et al. (1997) postulated two hypotheses: (1) the existence of an CaO2 sensitive chemoreceptor or (2) the presence of a chemoreceptor group in the pulmocutaneous and pulmonary arteries of amphibians and reptiles that specifically affected the cardiovascular system (Wang et al., 1997). The aortic bodies of mammals have been shown to be sensitive to changes in CaO2 and to mainly regulate the cardiovascular system (Jones and Daly, 1997), while the carotid bodies are PO2 sensitive and are predominantly responsible for ventilatory control (Lahiri et al., 1981). Aortic bodies are under-perfused relative to their oxygen consumption, which makes them sensitive to changes in oxygen concentration as well as PO2. The presence of chemoreceptors in the pulmocutaneous or pulmonary arteries of amphibians, turtles and snakes has been demonstrated (Lillo, 1980; Hoffman and de Souza, 1982; Ishii et al., 1985a, b; Wang et al., 2004; present study). Based on these observations both hypotheses are attractive. That both ventilatory and cardiovascular adjustments occured after stimulation of each chemosensory site in rattlesnakes (present study) and toads (Wang et al., 2004), suggests that chemoreceptor groups with distinct reflex roles do not exist in reptiles and amphibians.  120  Furthermore, my data do not support the presence of a group of receptors within the aorta or pulmonary artery that are under-perfused relative to their oxygen consumption making them sensitive to changes in CaO2 as wells as PO2. I find that all populations of putative oxygen sensing cells in each of the chemosensory areas of snakes have similar distributions and are all close to luminal blood flow (see Distribution and neurochemical content of putative oxygen sensing cells below). The same is true for turtles (Chapter 3) and for the aorta and carotid labyrinth of frogs (Chapter 2). This said, I still cannot rule out the possibility that some chemoreceptors at one of these sites have different stimulus modalities. Since stimulation with NaCN does not allow for differentiation between reductions in CaO2 and PO2, I used saline injections (2 ml) to mimic low CaO2, at normal PO2 and blood injections as a control (normal CaO2 and PO2). Saline injections, however, did not affect any of the cardiovascular or respiratory variables in any of the chemosensory areas. Although these findings suggest that peripheral chemoreceptors in rattlesnakes are not sensitive to reductions in CaO2, it is likely that the stimulus was not strong enough to cause a response. All I can conclude is that some receptors at all chemoreceptive sites respond to NaCN by eliciting cardiovascular and ventilatory adjustments.  4.5.5  Distribution and neurochemical content of putative oxygen sensing cells  I found the presence of VAChT and 5HT, but not catecholamines in putative oxygen sensing cells in the rattlesnake. VAChT-IR cells were grouped together in spongy tissue towards the lumen of the aorta and pulmonary artery. These cells were more often found in the distal segments of the chemosensitive areas of this blood vessel. 5HT-IR cells arranged singly or in small clusters of 2-5 were found throughout the chemosensitive areas of the carotid bifurcation, aorta and pulmonary artery. Their density, however, seemed to be higher in spongy-like tissue towards the heart-end of the chemosensitive areas of the aorta and pulmonary artery. Although these neurotransmitters never colocalized, the two cell types occasionally overlapped in the chemosensitive areas of the blood vessels. Putative oxygen sensing cells containing either 5HT or ACh were also present in all the chemosensory areas of red-eared sliders (Chapter 3). The morphology, size and arrangement of putative serotonergic and cholinergic O2 sensing cells in turtles and snakes is remarkably similar and coincides with anatomical features reported for chemoreceptor cells in other vertebrates (Ishii and Oosaki, 1969; Ishii et al., 1985b; Gonzalez et 121  al., 1994; Jonz and Nurse, 2003; Campanucci and Nurse, 2007; Coolidge et al., 2008; Jonz and Nurse, 2009, 2012; Shakarchi et al., 2012; Chapters 2, 3). There are, however, some differences in the distribution and neurochemical content of chemosensory areas between turtles and snakes. First, in rattlesnakes there was no evidence of the polygonal VAChT-IR cells present at the outflow of the aorta and pulmonary arteries in turtles. In Chapter 3, I proposed that these cells were involved in vasomotor regulation in turtles. Rattlesnakes have large R-L shunts (Wang et al., 1998; Galli et al., 2005; Taylor et al., 2009) and, in reptiles, changes in shunt pattern are regulated by cholinergically mediated constriction of the pulmonary artery (Burggren, 1977; Milsom et al., 1977; Burggren, 1987; Hicks, 1994; Hicks, 1998; Wang et al., 2001b), therefore, I would have expected similar ACh-positive cells at the outflow of the aorta and pulmonary artery. As mentioned above clusters of oval VAChT-IR cells are present in the pulmonary artery and aorta of snakes, but their location is by the lumen, not in close contact with smooth muscle, and distal to the heart. These observations taken together with their resemblance to putative O2 sensing cells in turtles (Chapter 3), suggests a chemosensory role. I cannot rule out, however, that VAChT-IR cells in snakes have a dual function in afferent and efferent control. Second, I found no sign of VAChT-IR cells in the carotid bifurcation. Ovoid 5HT-IR cells in the pulmonary vasculature of the snake Acrochordus granulates have been postulated to be chemosensory as well (Donald and Lillywhite, 1989). Their morphology, size (10-16 µm in diameter) and arrangement correspond to serotonergic cells in the carotid bifurcation, aorta and pulmonary artery of rattlesnakes in the present study. Given the anatomical similarities of VAChT- and 5HT-IR cells in rattlesnakes with chemoreceptor cells in other vertebrates and that 5HT and ACh have been proposed to be involved in O2-sensing in mammals and fish (Dunel-Erb et al., 1982; Gonzalez, et al., 1994; Zaccone et al., 1994, 1997; Jonz and Nurse, 2003; Jonz et al., 2004; Nurse, 2005, 2010; Porteus et al., 2012; Shakarchi et al., 2012; Jonz and Nurse, 2009, 2012; Zachar and Jonz, 2012), I postulate that both serotonergic and cholinergic cell types in rattlesnakes are O2 sensors. All data collected to date indicates that putative oxygen sensing cells in reptiles share the same neurochemical content. 5HT is present in peripheral chemoreceptors of all vertebrates studied and the presence of catecholamines and ACh varies between groups. My data on rattlesnakes support the observation of a trend to increase the number of neurotransmitters involved in  122  oxygen signal transduction from fish to amphibians and reptiles to mammal (Milsom and Burleson, 2007).  4.6  Conclusion To my knowledge this is the first study to locate functional peripheral chemoreceptors in a  reptile other than turtles. Their location in derivatives of pharyngeal arches 3, 4 and 6 (carotid bifurcation, aorta and pulmonary artery respectively) is the same as in turtles and amphibians. Having multiple chemoreceptor groups that sense arterial (carotid and aortic chemoreceptors) or mixed venous blood (pulmonary chemoreceptors) may construe an advantage to animals that have intra-cardiac shunts. I also determined the regulatory role (cardiovascular or respiratory) of the different chemoreceptive areas in rattlesnakes and identified putative oxygen sensing cells in these regions. Stimulation of all chemosensory sites with NaCN, which mimics anoxia, elicited adjustments in cardiac shunt and ventilation. That all chemosensory sites produced changes in cardiac shunt fraction suggests that adjustments of the cardiovascular system through changes in blood flows are important in regulating arterial blood gases. Thus, in the presence of low O2 levels in the blood, rattlesnakes increased ventilation, reduced the right to left shunt (i.e. they decreased the fraction of blood that bypasses the lung) and increased heart rate which will maintain blood gas homeostasis. Furthermore, ACh and 5HT may be involved in oxygen chemotransduction in rattlesnakes, as I found populations of putative chemoreceptor cells containing each of these neurotransmitters. These cells resembled those of other vertebrates, particularly those found in turtles, where the same arrangement and neurochemical content occur in putative oxygen sensing cells in the chemosensory areas.  123  Figure 4.1. Schematic (A) and pictures (B-F) showing the anatomy (C) and innervation by the glossopharyngeal/vagus nerve (IX/X nerve, arrows) of the carotid bifurcation (B and D) of Crotalus durissus. (E) the IX/X nerve running caudally by the carotid artery towards the aorta, pulmonary artery and heart (F). ic, internal carotid artery; ec, external carotid artery; Ca, carotid artery; Ao, aorta; PA, pulmonary artery. L and R refer to left and right side of the animal. Scale bar 1 mm (B, D-F) and 5 mm (C).  124  .  Qsys (ml/min/kg)  40  A  30  ∗ ∗  20  ∗  10  .  Qpul (ml/min/kg)  40  ∗ ∗  B  30  20  10 55  C fH (beats/min)  50  ∗ ∗ ∗  45 40 35 30 0  50 100 150 200 250 300 350  Time (s)  Figure 4.2. Changes in systemic (A) and pulmonary (B) blood flows and heart rate (C) in Crotalus durissus following injections of NaCN in the carotid bifurcation. Open symbols indicate pre-injection values (control) and filled symbols indicate post-injection values averaged over a minute. (*) denotes values that are significantly different from the control (Holm-Sidak pairwise comparison) (N=10).  125  Carotid  ∗  ∗ ∗  Carotid  B  C  80  3  2  .  6  .  60  40  Rsys (Psys/Qsys)  A  Carotid  Qtot (Vs x fH)  . .  Qp/Qs (ml/min/kg)  4  80  3  ∗  2  ∗  ∗ ∗  .  Aorta  E  6  .  60  40  Rsys (Psys/Qsys)  D Qtot (Vs x fH)  Qp/Qs (ml/min/kg)  Aorta  F  4  ∗ ∗ 2  1 20  0  Pulmonary  Pulmonary 80  ∗  3  2  .  Pulmonary  H  6  .  60  ∗  ∗  40  Rsys (Psys/Qsys)  G Qtot (Vs x fH)  4  Qp/Qs (ml/min/kg)  2  0  Aorta 4  . .  ∗  4  1 20  . .  ∗ ∗  I  ∗  4  ∗  2  1  20 0  100  200  Time (s)  300  0 0  100  200  Time (s)  300  0  100  200  300  Time (s)  Figure 4.3. Changes in shunt fraction (Q̇ p/Q̇ s; A, D and G), cardiac output (B, E and H) and systemic resistance (C, F and I) following injections of NaCN in the carotid bifurcation, aorta and pulmonary artery of Crotalus durissus. Open symbols indicate pre-injection values (control) and filled symbols indicate post-injection values averaged over a minute. (*) denotes values that are significantly different from the control (Holm-Sidak pairwise comparison) (N=10 for carotid and aortic groups; N=12 for pulmonary group).  126  Figure 4.4. Original traces showing the effects of NaCN injections (arrows) in the carotid bifurcation (A and B); aorta (C and D) and the pulmonary artery (E and F) on haemodynamic variables of intact and denervated Crotalus durissus. Scale bar 10 s.  127  .  Qsys (ml/min/kg)  40  A  ∗ ∗ ∗ ∗  30  20  10  .  Qpul (ml/min/kg)  40  B  30  20  10 55  fH (beats/min)  C 50 45  ∗  40  ∗ ∗  35 30 0  50 100 150 200 250 300 350  Time (s)  Figure 4.5. Changes in systemic (A) and pulmonary (B) blood flows and heart rate (C) in Crotalus durissus following injections of NaCN in the aorta. Open symbols indicate preinjection values (control) and filled symbols indicate post-injection values averaged over a minute. (*) denotes values that are significantly different from the control (Holm-Sidak pairwise comparison) (N=10).  128  .  Qsys (ml/min/kg)  40  A  30  ∗  ∗  20  10  .  Qpul (ml/min/kg)  40  B  30  ∗ ∗  20  10  fH (beats/min)  55  C  50  ∗ ∗  45 40 35 30 0  50 100 150 200 250 300 350  Time (s)  Figure 4.6. Changes in systemic (A) and pulmonary (B) blood flows and heart rate (C) in Crotalus durissus following injections of NaCN in the pulmonary artery. Open symbols indicate pre-injection values (control) and filled symbols indicate post-injection values averaged over a minute. (*) denotes values that are significantly different from the control (Holm-Sidak pairwise comparison) (N=12).  129  .  VTOT (% change)  Carotid 1400  A  1200  P=0.004  ∗  1000 800 600 400 200 0  .  VTOT (% change)  Aorta 1400  B  1200  P=0.002  1000  ∗  800 600  ∗  400 200 0  .  VTOT (% change)  Pulmonary 1400  C  1200  P=0.007  1000 800 600 400 200 0 0  120 240 360 480 600  Time (s)  Figure 4.7. Relative changes in the total ventilation of Crotalus durissus following injections of NaCN in the carotid bifurcation (A), aorta (B) and pulmonary artery (C). Total ventilation is expressed as percentage values relative to the pre-injection values (control). Open symbols indicate pre-injection values (control) and filled symbols indicate post-injection values averaged over two minutes. (*) denotes values that are significantly different from the control (one sample T-test) (N=10 for the carotid and aortic groups; N=12 for the pulmonary group).  130  Carotid A  ∗  6  ∗  4 2 0  Aorta VAMP (% change)  2  200  400 200  800  E  4 2  240  360  Time (s)  480  P=0.007  600  F  ∗  600  ∗  400  ∗  200 0  P>0.05 120  Aorta  Pulmonary  Pulmonary  0  P=0.012  600  0  P=0.002  6  0  ∗  D  4  8  400  ∗  6  0  B  800  VAMP (% change)  fR (breahts/min)  C  Carotid  600  0  P=0.012  8  fR (breahts/min)  800  VAMP (% change)  fR (breahts/min)  8  P=0.0001 0  120  240  360  480  600  Time (s)  Figure 4.8. Relative changes in amplitude and breathing frequency in Crotalus durissus following injections of NaCN in the carotid bifurcation (A and B), aorta (C and D) and pulmonary artery (E and F). Amplitude is expressed as percentage values relative to the preinjection values (control). Open symbols indicate pre-injection values (control) and filled symbols indicate post-injection values averaged over two minutes. (*) denotes values that are significantly different from the control (fR, Holm-Sidak pairwise comparison; VAMP, one sample T-test) (N=10 for the carotid and aortic groups; N=12 for the pulmonary group).  131  Figure 4.9. Original traces showing the effects of NaCN injections (arrow) in the carotid bifurcation (A); aorta (B) and the pulmonary artery (C) on the ventilation and breathing pattern of intact and denervated Crotalus durissus. Scale bar 50 s.  132  Figure 4.10. Putative oxygen sensing cells in a segment of the left common carotid and the carotid bifurcation of Crotalus durissus. A: Montage of a longitudinal section (20 µm thick) of the carotid bifurcation stained with Haematoxylin and Eosin (H&E). The left side of the image is the heart-end (*) (scale bar 1 mm). C-E: H&E stained longitudinal sections (20 µm thick) showing regions of A at higher magnification. B: Higher magnification of C showing the carotid bifurcation. Arrowheads indicate were 5HT-IR cells are found. F-G: Immunolabeling for 5HT (red) and a nuclear stain (DAPI, blue). The images from the red and green channels are shown separately and as a merged image. Scale bars 25 µm (B-G).  133  Figure 4.11. Putative oxygen sensing cells in the aorta of Crotalus durissus. A- C: Haematoxylin and Eosin (H&E) stained longitudinal sections (20 µm thick) of the aorta showing regions proximal to the heart (A, closest to the heart). Arrowheads in A and B indicate the spongy tissue and vessel wall were some 5HT-IR cells are found. Arrow in C indicates the location of VAChT-IR cells. D and E: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue). F-H: Immunolabeling for 5HT (red) and a nuclear stain (DAPI, blue). G shows 5HT-IR cells embedded in the vessel wall (right arrowhead in B) and H shows 5HT-IR cells in the spongy tissue indicated by arrowheads on the left side of panel A and B (F, Insert in H). The images from the red and green channels are shown separately and as a merged image. Scale bars 25 µm.  134  Figure 4.12. Putative oxygen sensing cells in the pulmonary artery of Crotalus durissus. A: Montage of a longitudinal section (20 µm thick) of the aorta stained with Haematoxylin and Eosin (H&E). The left side of the image is the heart-end (*). Arrowheads indicate the location of 5HT-IR cells and the arrow shows the location of VAChT-IR cells (Scale bar 250 µm). B-D: H&E stained longitudinal sections (20 µm thick) showing the region outlined in A at a higher magnification. The arrowhead indicates the location of some 5HT-IR cells. E-F: Immunolabeling for VAChT (green) and a nuclear stain (DAPI, blue). G-H: Immunolabeling for 5HT (red) and a nuclear stain (DAPI, blue). The images from the red and green channels are shown separately and as a merged image. Scale bars 25 µm (B-H). 135  Table 4.1 Primary and secondary antibodies used for immunohistochemistry. Antigen  Manufacturer  Host  Dilution  Cat. No.  Secondary antisera1  5-HT  serotonin  Immuno Star  goat  1:650  20079  Alexa Fluor® 555c  TH  tyrosine hydroxylase  Immuno Star  mouse  dilution curve  22141  Alexa Fluor® 488b  VAChT  vesicular acetylcholine Novus Biologicals  rabbit  1:300  NBP1-46776  Alexa Fluor® 488a  Alexa Fluor® 488b  Antisera Primary  transporter  HNK-1  CD-57  BD Pharmigen  mouse  1:100  559048  Alexa Fluor® 488  rabbit IgG (H+L)a  Molecular Probes, Invitrogen  donkey  1:200  A-21206  --  Alexa Fluor® 488  mouse IgG (H+L)b  Molecular Probes, Invitrogen  donkey  1:200  A-21202  --  Alexa Fluor® 555  goat IgG (H+L)c  Molecular Probes, Invitrogen  donkey  1:200  A-21432  --  Secondary1  1  Secondary antisera were conjugated with a fluorescent marker  a-c  Secondary antisera antigen corresponds with primary antibody host  136  Table 4.2 Effects of NaCN on stroke volume (VStot), mean arterial pressure (MAP) and systemic pulse (Psys). Values are mean ± s.e.m. (N=10, carotid group; N=10, aortic group and N=12, pulmonary group). * indicates a significant effect relative to pre-injection value (control, Holm-Sidak pairwise comparison). Ca: carotid artery, Ao: aorta, PA: pulmonary artery. Time  control 60 s 120 s 180 s 240 s 300 s  Vstot (ml/kg)  MAP (mm Hg)  Pulsesys (mm Hg)  Ca  Ao  PA  Ca  Ao  PA  Ca  Ao  PA  1.5±0.2 1.5±0.2 1.4±0.2 1.4±0.2 1.4±0.2 1.5±0.3  1.7±0.2 1.7±0.2 1.5±0.2 1.4±0.2* 1.4±0.3* 1.4±0.2*  1.5±0.2 1.4±0.1 1.0±0.2* 1.1±0.2* 1.3±0.2 1.3±0.2  41.2±3.7 41.9±3.8 41.9±3.9 43.2±4.2 44.2±4.1 44.4±3.8  31.2±4 30.9±3.9 29.1±3.6 31.2±3.9 33.1±3.5 32.5±3.4  43.2±6.4 38.4±6.5 31.5±4.7* 30.0±3.7* 31.0±3.7* 32.8±4.1*  9.6±1.4 9.5±1.4 7.5±1.3* 7.2±1.2* 7.6±1.4* 8±1.5  8.3±1.9 7.7±1.6 6.4±1.3* 6.5±1.5* 6.1±1.7* 6.2±1.7*  12.3±1.7 13.2±2.4* 9.8±2.4* 9.9±2.5* 10.7±2.2* 11.3±2.2*  137  Table 4.3 Effects of saline injections (2 ml) on pulmonary blood flow (Q̇ pul), systemic blood flow (Q̇ sys), heart rate (fH), mean arterial pressure (MAP), total ventilation (VTot), breathing frequency (fR) and amplitude (VAMP). Values are mean ± s.e.m. (N=9, carotid group; N=8, aortic group and N=9, pulmonary group). VTot and VAMP are expressed as percentage values relative to the pre-injection (control) values. Saline injections had no significant effect relative to pre-injection value (control, Holm-Sidak pairwise comparison). Ca: carotid artery, Ao: aorta, PA: pulmonary artery. Time Saline (2 ml)  control 60 s 120 s 180 s 240 s 300 s Time Saline (2 ml)  control 60 s 120 s 180 s 240 s 300 s  Q̇ pul (ml min-1 kg-1)  Q̇ sys (ml min-1 kg-1)  fH (min-1)  MAP (mmHg)  Ca  Ao  PA  Ca  Ao  PA  Ca  Ao  PA  Ca  Ao  PA  27.3±3.4 25.4±3.7 26.7±3.6 25.5±3.5 24.7±3.2 24.6±3.3  25.7±4.9 25.7±5.0 25.4±4.5 25.4±4.3 23.9±4.5 25.7±4.0  24.8±2.8 24.2±3.0 23.6±3.2 25.1±3.3 24.6±3.2 25.3±3.2  28.1±5.5 28.6±5.7 27.8±5.9 28.3±6.1 27.9±5.9 27.6±5.6  38.0±10.8 37.9±10.2 35.2±7.8 32.9±6.4 32.6±6.5 32.8±6.9  39.7±13.8 41.7±13.1 38.0±13.5 39.6±13.4 40.1±13.4 40.3±13.5  37.1±2.9 37.3±3.0 37.6±2.9 37.6±2.8 37.4±2.8 37.4±3.0  38.6±3.4 37.6±3.4 38.1±3.8 38.0±3.3 38.1±3.5 38.6±3.4  37.3±5.0 37.8±4.7 37.0±4.7 37.9±4.4 38.2±4.6 38.6±4.7  41.9±5.1 42.8±5.3 43.7±5.4 44.3±5.4 44.5±5.2 44.8±5.3  35.0±3.3 35.3±3.4 34.6±3.3 35.2±3.3 36.6±3.7 35.8±3.6  33.3±3.4 33.4±3.1 32.3±2.9 32.3±3.1 32.8±3.2 33.4±3.5  fR (min-1)  VTot (% change)  VAMP (% change)  Ca  Ao  PA  Ca  Ao  PA  Ca  Ao  PA  100 135.5±26.8 146.6±36.2 166.3±31.9 128.6±24.0 165.6±37.9  100 151.4±49.9 65.9±18.5 52.7±21.3 76.4±20.7 68.4±13.2  100 146.1±19.7 104.4±19.1 111.3±22.6 112.4±18.1 72.4±15.5  3.4±0.5 3.3±0.6 3.2±0.6 3.6±0.6 2.9±0.6 3.0±0.6  4.4±0.7 4.9±0.9 3.4±0.9 3.2±1.3 2.6±0.7 3.0±1.0  3.6±0.7 3.6±0.7 3.2±0.7 3.3±0.7 3.6±0.8 2.3±0.5  100 133.4±21.5 181.1±42.9 171.9±36.8 187.8±40.4 201.7±38.4  100 109.9±22.0 81.9±19.2 69.7±25.0 96.3±27.6 95.1±30.3  100 147.5±20.9 111.5±13.6 107.6±14.4 99.1±10.9 95.9±14.4  138  Chapter 5: Daily and seasonal rhythms in the respiratory sensitivity of redeared sliders (Trachemys scripta elegans) 5.1  Summary The purpose of the present study was to determine whether the daily and seasonal changes in  ventilation and breathing pattern previously documented in red-eared sliders resulted solely from daily and seasonal oscillations in metabolism or also from changes in chemoreflex sensitivity. Turtles were exposed to natural environmental conditions over a one-year period. In each season, oxygen consumption rate, ventilation and breathing pattern were measured continuously for 24 h while turtles were breathing air and for 24 h while they were breathing a hypoxichypercapnic gas mixture (H-H). I found that oxygen consumption rate was reduced equally during the day and night under H-H in all seasons except spring. Ventilation was stimulated by H-H but the magnitude of the response was always less at night. On average, it was also less in the winter and greater in the reproductive season. These data indicate that the day to night differences in ventilation and breathing pattern seen previously resulted from daily changes in chemoreflex sensitivity, while the seasonal changes were strictly due to changes in metabolism. Regardless of mechanism, the changes resulted in longer apneas at night and in the winter at any given level of total ventilation facilitating longer submergence at times of the day and year when turtles are most vulnerable.  5.2  Introduction Circadian and circannual rhythms time biological processes so that specific events take place  at appropriate times of the day or year (Underwood, 1992; Tosini et al., 2001; Mortola and Seifert, 2002). Such temporal changes in physiological variables may be particularly crucial for the survival of temperate species, which are subjected to large daily and seasonal variations in their environment (changes in day length and temperature cycles). It has been previously shown that endogenous circadian and circannual oscillations are present in the metabolism and ventilation of red-eared sliders (Trachemys scripta elegans, Wied) (Reyes and Milsom, 2010), and are accompanied by changes in breathing pattern that result in longer apneas at night and during colder seasons. The authors speculated that by reducing trips to the surface to breathe, turtles could reduce the cost of locomotion, risk of predation (Cagle, 1950), and potentially, the cost of breathing (Vitalis and Milsom, 1986) during dormancy. 139  Little is known about the mechanisms that generate the long apneas that are common to arrhythmic breathing in turtles. Endogenous circadian and circannual oscillations in apnea length could reflect oscillations in metabolic rate (the need to breathe) or in chemoreflex sensitivity (the drive to breathe). Chemoreflexes are thought to be important in initiating and terminating periods of apnea (Shelton et al., 1986; Milsom, 1990), but the few studies of circadian and circannual variation in chemoreflex responses, performed in mammals and amphibians, have not always observed rhythms in chemoreflex sensitivity (McArthur and Milsom, 1991; Milsom et al., 1993; Peever and Stephenson, 1997; Rocha and Branco, 1998; Bicego-Nahas and Branco, 1999; Mortola and Seifert, 2000; Stephenson et al., 2000; BicegoNahas et al., 2001; Mortola and Seifert, 2002; Seifert and Mortola, 2002a; Seifert and Mortola, 2002b; Mortola, 2004). In the present study, I sought to determine whether daily and seasonal oscillations in chemosensitivity per se (defined as the increase in ventilation of turtles breathing a hypoxichypercapnic (H-H) gas, after correcting for any changes in metabolism; i.e. the change in air .  .  .  .  convection requirement; ∆ACR = ∆VE/∆VO2, where VE is ventilation and VO2 is oxygen consumption) could be involved in producing the daily and seasonal changes in ventilation and breathing pattern previously reported in red-eared sliders (Reyes and Milsom, 2010). I hypothesized that endogenous rhythms in chemosensitivity contribute to the daily and seasonal oscillations observed in the breathing pattern of red-eared sliders allowing turtles to remain submerged longer at night and in the winter, which may have implications on their survival.  5.3  Materials and methods Red-eared slider turtles (Trachemys scripta elegans, Wied) (average weight=1.09±0.35 kg,  N=8) were obtained from commercial suppliers (Lemberger Company, Wisconsin and Sullivan Company Inc., Tennessee, USA) and were housed in a semi-natural pond (2.9 x 1.9 x 0.6 m, containing 3.3 m3 of water), where they experienced the environmental temperature and natural photoperiod of Vancouver (BC, Canada) throughout the year. Two temperature loggers (DS1921 Thermocron iButton, Dallas Semiconductor Corporation, Sunnyvale, CA, USA) were placed on the basking area (1.25 x 1.9 m) in the shade, and submerged in the pond to record ambient and water temperature respectively, every hour for a year. Prior to each trial, temperatures recorded by the data loggers were averaged to determine the experimental seasonal temperature. Daily photoperiod and rainfall were obtained from the Meteorological Service of Canada website. The 140  holding and experimental procedures followed Canadian Council on Animal Care guidelines and were approved by the University of British Columbia Animal Care Committee (animal care certificate No. A041006). Experiments were carried out over a 12-month period (October to October) on unrestrained turtles that were fasted for 7 days. In each season turtles were removed from the semi-natural pond and placed in an experimental tank (1.22 x 0.49 x 0.64 m) for four days, at the mean seasonal temperature (same day and night water temperatures) and photoperiod prevailing in that season (Table 5.1). The tank was placed inside a wooden box to insulate the turtle from external stimuli (noise and light). Full spectrum lights were set with a timer to control the light and dark cycle. The water entering the tank was filtered and the temperature was controlled by a solenoid system. A flow-through system ensured that the water temperature remained constant within treatments. Turtles were acclimated for two days to adjust to the handling stress, the experimental tank and the lack of a thermal cycle. On the first day following acclimation turtles were given humidified air to breathe for 24 h. On the second day after acclimation turtles were given a H-H gas mixture (8% oxygen-3% CO2) to breathe for 24 h. Water temperature in the experimental tank was recorded every hour for the length of the experimental treatment with a temperature logger (DS1921 Thermocron iButton, Dallas Semiconductor Corporation, Sunnyvale, CA, USA) to ensure that temperature remained steady during the experimental trial. To determine the effects of removing all cues that could give any indication of time of the day, the same protocol was repeated, in the fall except that turtles were maintained under constant darkness for four days. The water temperature used for this experiment was 13.6°C. Day and night differences measured in the autumn under the natural photoperiod were compared with the corresponding day and night values measured under constant darkness.  5.3.1  Measurement of metabolic rate and ventilation .  .  VO2 and VE were measured simultaneously using open-flow respirometry. Air or the H-H gas mixture was delivered through a port in the side of a breathing funnel at a rate of 350 ml/min. The flow rate was regulated with a gas mixing flow meter (Cameron Instrument Company, Port Aransas, TX, USA). The air/gas mixtures entering and leaving the funnel were dried and sampled .  by a gas analyzer (model 222A version 1.02, Raytech Instruments, Vancouver, BC, Canada). VO2 rates were calculated from the flow rate and the difference in fractional concentrations of oxygen 141  between inflow and outflow gas. Rates of oxygen consumption are expressed as ml O2 STPD/min/kg (where STPD is standard temperature and pressure, dry). .  VE was monitored at the outflow of the breathing funnel with a pneumotachograph (Fleisch, Richmond, VA, USA) attached to a differential pressure transducer (Validyne, Northridge, CA, USA) (Funk et al., 1986). The breathing trace was analyzed for breathing frequency (fR, .  breaths/min) and average tidal volume (VT, ml/kg). Total ventilation (VE; fR x VT, ml/min/kg) was calculated from these values. The components of breathing frequency: number and frequency of breathing episodes, number of breaths in each episode, the instantaneous frequency (the frequency of the breaths within an episode) and the percent time spent apneic were also calculated. Apnea was defined as a respiratory pause that exceeded the duration of two breaths.  5.3.2  Data analyses  All data were continuously recorded using WINDAQ acquisition software (version 2.19, Dataq Instruments, Akron, OH, USA). Fall and winter data were analyzed first, for every 1-h segment of the 24-h experiment. Based on this, I determined that analyses of data every other hour was sufficient to accurately describe daily rhythms, thus spring, summer and constant dark runs were analyzed every other hour. Data were averaged over both the daytime and nighttime measurements, and over the full 24 h for all turtles (N=6-8) in each season. Segments of breathing traces were removed from the analyses when turtles were active to ensure that breathing and metabolism values used for the analyses were resting rates. Turtle activity produced water movement that induced pressure changes that were sensed by the pneumotacograph (noise). This method was verified by comparing the noise in the breathing trace with 24-h long video recordings of the turtles during the initial fall trials.  5.3.3  Calculation of the air convection requirement and chemosensitivity  Oxygen consumption and ventilation were used to calculate the air convection requirement .  .  (ACR, ml air/ml O2) for each hour, where ACR = VE/VO2. The hypoxic-hypercapnic ventilatory response was calculated as the absolute change in total ventilation when turtles went from breathing air to breathing the hypoxic-hypercapnic gas .  mixture (∆VE). Since my goal was to determine whether ventilatory sensitivity showed endogenous daily and seasonal rhythms (i.e. independent of temperature and metabolism), I 142  eliminated the effects of temperature variation between seasons on my calculations of chemosensitivity using temperature coefficient values (Q10) to correct all data to 14.7°C (the mean annual water temperature). Q10’s for ventilation, oxygen consumption and ACR were calculated for each turtle under both the air and the hypoxic-hypercapnic treatments. Winter .  .  Q10’s were calculated by inserting the VO2, VE and ACR measured at two different temperatures in this season (19.6ºC and 8.8ºC, indoor and outdoor temperatures respectively) into equation 1 .  .  (Table 5.2). MR2 was the variable to be determined (corrected winter VO2, VE and ACR at T2 = 14.7ºC). MR1 was the winter value measured at 8.8ºC (T1). As the summer outdoor temperature (20.8ºC) was similar to the indoor temperature (19.6ºC), I calculated the temperature coefficients by inserting oxygen consumption, ventilation and ACR values measured at the summer and fall outdoor temperatures (20.8ºC and 14.71ºC, respectively). Fall values were selected because changes between summer and fall for all physiological variables were solely the result of changes in temperature (Reyes and Milsom, 2010). Q10 values and summer values (MR2) measured at 20.8°C (T2) were inserted into equation 1 to calculate oxygen consumption, ventilation and ACR values (MR1) at 14.7°C (T1). Temperature corrected changes in ventilation .  .  (∆VE), oxygen consumption (∆VO2) and air convection requirement (∆ ACR) were obtained .  .  from the temperature corrected values calculated for VE,VO2 and ACR under air and the hypoxichypercapnic gas.  Equation. 1  Q10 = (MR2/MR1)10/(T2-T1)  I eliminated the effect of changes in metabolism (day-night and seasonal) on chemosensitivity by calculating temperature corrected changes in air convection requirement when turtles went from breathing air to breathing the hypoxic-hypercapnic gas (∆ACR). Thus, in .  .  this study I use changes in ACR calculated from temperature corrected values of ∆VE and ∆VO2 as a measure of changes in chemosensitivity.  5.3.4  Statistical analyses  Data are expressed as means ± s.e.m. Differences between day and night, as well as the effects of season on the daily changes were assessed with two-way repeated measures (RM) ANOVA. Differences between seasons (24-hour average) were assessed using one-way repeated 143  measures ANOVA. Holm-Sidak multiple comparisons tests were used to determine pairwise differences. Data that did not meet the assumptions of normal distribution or equal variances were natural log (Ln) transformed. Sigma Stat (version 3.11, Systat Software, IL, USA) was used for all statistical analyses.  5.4  Results  5.4.1  Daily rhythms in the metabolism and breathing of turtles during exposure to  hypoxic-hypercapnia It has been previously shown that there are circadian rhythms in oxygen consumption and ventilation in red-eared sliders breathing air (Reyes and Milsom, 2010) (Fig. 5.1D, E). Here I show for the first time that red-eared sliders also have daily rhythms in oxygen consumption and .  .  ventilation while breathing a hypoxic-hypercapnic gas mixture (VO2: P=0.008; VE: P=0.003, daynight, two-way RM ANOVA) (Fig. 5.1A, B). Daytime breathing frequency was higher than nighttime breathing frequency in all seasons (P<0.001, day-night, two-way RM ANOVA) (Table 5.3), while tidal volume remained relatively constant throughout the 24 h in all seasons (P=0.236, day-night, two-way RM ANOVA) (Table 5.3). Changes in breathing frequency were solely the result of more frequent breathing episodes during the day (P=0.003, day-night, two-way RM ANOVA) (Table 5.3); the number and frequency of breaths in each episode did not vary with time of the day (P=0.373 and P=0.222, number and frequency of breaths respectively, two-way RM ANOVA) (data not shown)). Thus, turtles were apneic longer at night than during the day in all seasons (P=0.002, day-night, two-way RM ANOVA) (Table 5.3). While day-night differences in the breathing pattern of turtles exposed to the hypoxichypercapnic gas were similar to those observed in animals breathing air (Table 5.3) (Reyes and Milsom, 2010), the daytime increases in breathing under hypoxia-hypercapnia were larger than the increases in oxygen consumption, giving rise to daily rhythms in ACR (P<0.001, day-night, twoway RM ANOVA) that were not present in animals breathing air (Fig. 5.1C).  5.4.2  Daily rhythms in chemosensitivity .  Hypoxic-hypercapnia reduced VO2 in the fall (day: 37%, night: 11%), summer (day: 17%, night: 3%) and winter (day: 40%, night: 0.5%). In spring, however, both daytime and nighttime metabolism increased (130% and 56%, respectively) when turtles were breathing the hypoxic144  hypercapnic gas. The magnitude of these changes, in absolute terms, were the same during the day and the night (P=0.165, day-night, two-way RM ANOVA) (Fig. 5.2A). Hypoxic-hypercapnia also caused an increase in ventilation (summer-day: 536%, summernight: 538%; fall-day: 629%, fall-night: 592%; winter-day: 341%, winter-night: 568%; spring-day: 1496%, spring -night: 990%). This increase was reduced at night in fall and spring (P<0.05, Holm-Sidak) (Fig. 5.2B). The ACR increased when animals were breathing the hypoxichypercapnic gas (summer-day: 799%, summer-night: 600%; fall-day: 1627%, fall-night; 902%; winter-day: 749%, winter-night: 953%; spring-day; 1194%, spring-night: 871%) and this increase .  .  in overall air convection requirement (∆ACR: ∆VE/∆VO2) varied between day and night across all seasons (P=0.002, day-night, two-way RM ANOVA) (Fig. 5.2C). Thus, daily changes in ventilation were not simply due to changes in metabolism, but were the result of daily cycles in the sensitivity of chemoreflexes to respiratory stimuli. Exposure to four days of constant darkness blunted daily oscillations in the chemoreflex .  response of red-eared sliders. “Daytime” changes in ventilation (∆VE) under constant darkness were reduced by 63% from day values under the natural photocycle, while the decrease in the night ventilatory response was less (21% decrease under constant darkness). As a result no day-night .  differences in respiratory responses were observed after four days of constant darkness (∆VE: P=0.118; ∆ACR: P=0.480, Holm-Sidak) (Fig. 5.3). 5.4.3  Seasonal rhythms in the metabolism and breathing of turtles during exposure to  hypoxic-hypercapnia As has been shown for red-eared sliders held in normoxia (Reyes and Milsom, 2010), when sliders were given the hypoxic-hypercapnic gas mixture to breathe they also showed a seasonal cycle in metabolism. I measured the highest rates of oxygen consumption in the summer and metabolism decreased seasonally as expected with temperature changes (P<0.001, one-way RM ANOVA) (Fig. 5.4C). Ventilation showed the same general pattern between seasons as metabolism, except that ventilation was not as elevated in summer (P=0.013, one-way RM ANOVA) (Fig. 5.4A). Overall differences in metabolism between seasons seemed to result solely from the effects of temperature on oxygen consumption, since these differences were not apparent after temperature-correction (P=0.266, one-way RM ANOVA, 14.7°C) (Fig. 5.4D). In contrast,  145  ventilation remained elevated in the spring and reduced in the winter after correction for temperature (P<0.001, one-way RM ANOVA) (Fig. 5.4B). Seasonal changes in the oxygen consumption and ventilation of turtles breathing the hypoxichypercapnic gas showed similar trends as those observed in turtles breathing air. Seasonal differences in breathing during hypoxic-hypercapnia were caused by small changes in tidal volume and large changes in breathing frequency (VT: P=0.013; fR: P<0.001, one-way RM ANOVA) (Table 5.4). Breathing frequency in hypoxia-hypercapnia changed due to alterations in the number and frequency of breaths within breathing episodes (P<0.001, one-way RM ANOVA) (Table 5.4), not due to changes in the frequency of breathing episodes (P=0.076, one-way RM ANOVA) (Table 5.4). These changes in the breathing pattern resulted in a considerable increase in the time spent in apnea during winter (P<0.001, one-way RM ANOVA) (Table 5.4). Because ventilation and oxygen consumption changed largely in parallel, ACR changed very little between seasons. However, ACR was reduced in the summer compared to other seasons (P<0.05, Holm-Sidak) (Fig. 5.4E). No seasonal differences were found after correcting for temperature (Fig. 5.4F).  5.4.4  Seasonal rhythms in chemosensitivity  Oxygen consumption fell when turtles were given the hypoxic-hypercapnic gas instead of air in all seasons (13, 26 and 28% reduced in the summer, fall and winter respectively) (P=0.003, oneway RM ANOVA) except spring, when metabolism increased by 95% (Fig. 5.5A). In contrast, ventilation increased in all seasons during hypoxic-hypercapnia (summer: 539%; fall: 615%; winter: 416%; spring: 1302%; P<0.001) (Fig. 5.5B) and this increase was greatest in the spring and lowest in the winter. Hypoxic-hypercapnia increased the ACR in all seasons (summer: 728.6%; fall: 1204%; winter: 604%; spring: 1041%; P=0.001), but the magnitude of this change was similar in all seasons except summer, when it increased the least (Fig. 5.5C). To remove the effects of temperature on the metabolism and ventilation of turtles, I then .  .  temperature-corrected the VE and VO2 values measured in all seasons under both hypoxiahypercapnia and air to 14.7°C using temperature coefficients (Q10 values) and used these values to .  .  calculate the ∆VE, ∆VO2 and ∆ACR. The absolute change in oxygen consumption in the spring was significantly different from that of the fall and summer (P<0.05, Holm-Sidak test). Differences between spring and winter were lost, however, after correcting for temperature (P>0.05, Holm-Sidak test) (Fig. 5.6A). Red-eared sliders still showed elevated ventilatory 146  responses to hypoxic-hypercapnia in the spring and reduced ventilatory responses in the winter when temperature effects were removed (P<0.001, one-way RM ANOVA (Fig. 5.6B). However, after temperature correction, the increase in ventilation relative to metabolism (∆ACR) did not vary between seasons (P=0.428, one-way RM ANOVA) (Fig. 5.6C).  5.5  Discussion The main goal of my study was to determine whether respiratory chemosensitivity varied  daily and/or seasonally in red-eared sliders. Respiratory chemosensitivity is normally defined in mammals as a unit change in ventilation arising from a unit change in arterial gas partial pressure .  (∆VE /∆PaO2/CO2). In the present study I did not measure blood gases due to the difficulty of .  maintaining cannulae open for over a year. As a result I report only the ∆VE between animals breathing air and the hypoxic-hypercapnic gas (i.e. no units of change in stimulus intensity). In .  ectotherms, metabolism and temperature can alter the relationship between VE and PaO2/CO2. I .  .  therefore corrected for these factors by standardizing VE to VO2 (air convection requirement, ACR), and also adjusting these values for changes in temperature between seasons. After taking these precautions, I found that the day to night differences in ventilation resulted from daily changes in chemoreflex sensitivity, but that the seasonal changes were directly related to changes in metabolism.  5.5.1  Daily rhythms in ventilatory sensitivity  Red-eared sliders in this study showed a reduced respiratory response to the hypoxichypercapnic stimulus at night compared to the day. Time of the day has been shown to affect the ventilatory response to hypoxia, hypercapnia, or both stimuli combined in mammals (Stephenson et al., 2000; Jarsky and Stephenson, 2000; Mortola and Seifert, 2002; Mortola, 2004) and birds (Woodin and Stephenson, 1998), and now I have shown this in reptiles. The mechanisms underlying day-night changes in ventilatory responses have not yet been elucidated, although interesting species-specific differences have been found. Rats had a higher ventilatory response to hypoxia and hypercapnia at night, when they were active, than during the day. These oscillations in breathing were correlated to daily changes in metabolism and not to changes in chemosensitivity since the hyperventilatory response did not vary between day and night (i.e. there was no %∆ACR between night and day) (Peever and Stephenson, 1997; Mortola 147  and Seifert, 2000; Seifert and Mortola, 2002a; Seifert and Mortola, 2002b). In humans, on the other hand, the ventilatory response to hypercapnia was higher in the day and was independent of metabolism or sleep (Stephenson et al., 2000; Mortola and Seifert, 2002; Mortola, 2004), indicative of an endogenous daily cycle in chemosensitivity. In the present study I did not attempt to determine to what extent the day night differences seen in chemosensitivity were a result of changes in sleep state. State of arousal is known to affect both the hypoxic and hypercapnic ventilatory responses in mammals. Although, the variability between species is large, the general consensus is that the hypercapnic ventilatory response is blunted during sleep and the hypoxic ventilatory response is reduced or unchanged (Phillipson et al. 1978; Bowes et al., 1981; Hunter et al. 1998; Mortola, 2004). Sleep patterns have been described in turtles and lizards (Flanigan, 1973; Flanigan et al. 1974) and periodic reductions in breathing frequency have been reported in pythons and lizards during “sleep” (Donnelly and Woolcock, 1977; Cragg, 1978). Thus, my data support the notion that turtles have a daily rhythm in respiratory sensitivity that is independent of circadian changes in metabolism, activity and temperature, but I cannot rule out changes in state. Mortola (2004) suggested that the species difference between rats and humans could reflect the fact that circadian changes in ventilation are coupled to changes in metabolism in rats where these changes are rapid and sleep state is fragmented, but endogenous in humans where the changes in metabolism are small and slow and the sleep cycle is prolonged. Metabolism in turtles also changes slowly, and is further constrained by changes in environmental temperature, perhaps increasing selection pressure on alternate mechanisms to reduce ventilation during periods of inactivity. Free-running rhythms (constant temperature and D-D or L-L photocycles) in metabolism and ventilation have been shown in the box turtle, garter snake, european green lizard and redeared slider (Glass et al. 1979; Hicks and Riedesel, 1983; Rismiller and Heldmaier 1991; Reyes and Milsom, 2010). In the present study, I found that 4 days of constant darkness was sufficient to abolish the day-night difference in chemosensitivity. Interestingly, this was caused by a reduction in the “daytime” value, while “nighttime” ventilatory sensitivity did not change from the levels seen under the natural photocycle. The same turtles continued to show day-night differences in metabolism and ventilation after three days in constant darkness while breathing air, although the cycles were blunted and the period was slightly shifted (Reyes and Milsom, 2010). The loss in the day-night difference in chemosensitivity after 4 days of constant darkness does not imply that the daily changes in the physiological variables were not driven by a 148  circadian system. Removal of all environmental cues typically reduces the amplitude of daily rhythms, until the rhythms are lost altogether (Aschoff et al., 1982; Kenagy and Vleck, 1982; Underwood, 1992; Tosini et al., 2001). Tegu lizards have also been shown to loose circadian oscillations in metabolism over time (Milsom et al., 2008). There was large variability in the response of these lizards to constant dark, with some individuals loosing the rhythms after a day in constant dark and others maintaining the rhythms up to 14 days. Unfortunately, it is impossible for me to determine where in the reflex arc that underlies the ventilatory response to hypoxic-hypercapnia, changes occurred, or even whether the changes were in the hypoxic, hypercapnic or both components of the response. While hypoxia is only sensed by peripheral chemoreceptors, hypercapnia stimulates both central and peripheral chemoreceptors and the central brainstem sites involved in processing both stimuli are different (Shelton and Boutilier, 1982; Shelton et al., 1986; Milsom, 1990). I used a hypoxic-hypercapnic gas mixture, however, to mimic the respiratory stimuli that turtles experience during submergence (Burggren and Shelton, 1979; Shelton and Boutilier, 1982). Regardless of the mechanism behind the daily changes in chemosensitivity of turtles, it is clear that nighttime reductions in chemosensitivity should reduce sensitivity to a fall in oxygen stores and an accumulation of CO2 in the blood, and thus allow for longer dives. This appears to be true for other diving species as well. Thus, canvasback ducks showed reduced chemosensitivity to progressive asphyxia during the nighttime that was independent of metabolism. These animals forage underwater at night and increasing the level of blood gases at which chemoreflexes are elicited at this time was suggested to allow them to prolong their dives and foraging times (Woodin and Stephenson, 1998). In red-eared sliders, the endogenous circadian rhythms in metabolism (lower at night) (Reyes and Milsom, 2010) along with daily oscillations in chemoreflex responses (independent of metabolism) allow these turtles to reduce the locomotion costs and predation risks associated with surfacing to breathe at night.  5.5.2  Seasonal rhythms in ventilatory sensitivity  Red-eared sliders showed seasonal differences in their ventilatory response to hypoxichypercapnia: it was reduced in the winter and elevated in the spring. This may not seem surprising since it is well established that sensitivity to hypoxia or hypercapnic-acidosis increase with temperature (Shelton and Boutilier, 1982; Glass et al., 1983; Glass et al., 1985; Funk and Milsom, 1987). In the present study, however, seasonal changes in the hypoxic-hypercapnic 149  response remained after the data were temperature corrected, and appeared to have resulted from the effects of the respiratory stimulus on metabolism and not from changes in chemosensitivity, since the change in air convection requirement (∆ACR) did not vary with season. The reduced ventilatory response during winter coincides with the dormancy period. Temperate reptiles, including red-eared sliders, often undergo active metabolic suppression, when metabolism cannot be sustained under winter environmental conditions (Mayhew, 1965, Bennett and Dawson, 1976; Gregory, 1982; Ultsch, 1989; Zari, 1999). Toads and bullfrogs also show reduced metabolism and a lower ventilatory response to hypoxia (toads; Bicego-Nahas et al., 2001) and hypercapnia (bullfrogs; Bicego-Nahas and Branco, 1999) in the winter. It would appear that seasonal reductions in both, the temperature corrected resting ventilation and in the ventilatory response to hypoxic-hypercapnia of red-eared sliders were due to the orchestrated fall in metabolism. Short photoperiods in the fall trigger behaviours such as fasting and selection of lower temperatures (Cagle 1950; Gregory, 1982; Rismiller and Heldmaier 1982), which will slowly reduce metabolism in preparation for dormancy (Bennett and Dawson, 1976). Turtles in this study underwent anorexia in the late fall and did not eat in the winter. For many reptiles, the highest metabolic rates occur after feeding, with metabolism falling dramatically during fasting. Furthermore, during long periods of fasting the digestive system atrophies removing the daily cost of regeneration (Wang et al. 2001c). Although reduced gut mass and lack of feeding may .  have influenced winter metabolism, the seasonal trends in VO2 cannot be solely explained by changes in feeding regime. Turtles in this study stopped eating voluntarily in mid to late fall and this continued throughout the winter. Thus, if feeding were important, temperature corrected metabolism in the fall (in animals breathing air) should also have been partially reduced and this was not the case (Fig. 5.4C, D). Ventilation increased in all seasons with exposure to hypoxia-hypercapnia, which agrees with other studies on reptiles (Jackson, 1973; Jackson et al., 1974; Glass et al., 1985; Shelton et al., 1986). However, the increase in spring ventilation was larger than that in all other seasons even after temperature correction. Red-eared sliders mate during the spring (Cagle, 1950; Duvall et al., 1982; Ernst and Barbour, 1989; Kuchling, 1999) and hormones involved in reproduction may be the source of the elevated metabolism and the higher ventilatory response during this season (Tatsumi et al., 1997). Higher metabolic rates during oogenesis and mating have also been reported in Lacerta viridis (Rismiller and Heldmaier, 1991). 150  Paradoxically, while the ventilatory response to hypoxic-hypercapnia was greatest in the spring, resting metabolism and ventilation in animals breathing air were very low (Fig. 5.4) (Reyes and Milsom, 2010) remaining at winter levels. In all cases, however, the changes in ventilation and metabolism occurred in parallel, suggesting that the spring time increases in the hypoxic-hypercapnic ventilatory response were a result of spring time increases in the effects of hypoxic-hypercapnia on metabolism and not the result of seasonal changes in chemosensitivity. The reasons for this are not clear. Hypoxia and/or hypercapnia generally suppress metabolism, if anything, and this was the trend that was seen in all other seasons. It is possible that cause and effect are reversed here and that metabolism increases because of the very large increase in ventilation rather than vice versa, but estimates of the oxidative cost of ventilation in turtles remain controversial. Estimates of the relative metabolic cost of breathing based on the elimination of breathing by unidirectional ventilation (Kinney and White, 1977), or based on changes in ventilation and metabolism associated with hypoxic exposure (Wang and Warburton, 1995, Skovgaard and Wang, 2004) suggest that the costs could be as high as 15 to 20% of resting metabolism. Estimates based on changes in ventilation and metabolism associated with hypercapnic exposure, however, suggest the costs are extremely small and often overshadowed by an acidosis induced metabolic suppression leading to a paradoxical decrease in metabolism with an increase in breathing (Wang and Warburton, 1995, Skovgaard and Wang, 2004). This issue remains to be resolved, as does the source of the increase in metabolism described here.  5.6  Conclusion What I have measured in this study is the response to a fixed stimulus applied at the lungs.  The stimuli for the chemoreceptors involved in ventilatory regulation are the arterial blood gases (PO2, PCO2 and pH) at the peripheral receptors and the PCO2/pH of the cerebral spinal fluid (CSF), but these were not measured. As a result, the changes in chemosensitivity that I describe may have been due to changes in the sensitivity of the peripheral arterial chemoreceptors per se, or to changes in the PO2/PCO2/pH at receptor sites. The latter could occur due to daily and/or seasonal changes in the magnitude of the cardiac shunt that is known to be large in turtles. There were seasonal changes in temperature that are also known to give rise to changes in the air convection requirement and hence to resting blood gas tensions. To add to the complexity, similar changes in baseline blood gas tensions initiated from different baseline blood gas levels may give rise to different responses due to the non-linearity of the hypoxic ventilatory response 151  (the same change in blood gases may not amount to the same change in stimulus when starting from different baselines) (Glass, 1992). Finally, even when no changes were seen in chemosensitivity as a function of season, this could have been due to seasonal changes in cardiac shunting acting to maintain constant blood gas tensions (Wang and Hicks, 1996a; Wang et al., 1997). Any such changes, however, are part of daily and seasonal rhythms. In this study I am looking at the whole animal output for a given environmental input and determining the mechanistic basis of these changes will be the next challenge. Mechanism aside, red-eared sliders showed reduced respiratory chemosensitivity at night and in the winter, and enhanced chemosensitivity in the spring. Day and night differences resulted from daily oscillations in the sensitivity of chemoreflexes, while seasonal differences could be explained by the effects of the hypoxic-hypercapnic stimuli on metabolism. Regardless of the different mechanisms, daily and seasonal changes in the ventilatory response together with circadian and circannual rhythms in metabolism (Reyes and Milsom, 2010) benefit turtles by facilitating longer apneas. These physiological changes will reduce time at the surface and may be important for minimizing the cost of locomotion at times when the cost of surfacing or risk of predation may compromise their survival.  152  A P=0.003  VE (ml/min x kg)  150  200  ∗ ∗  D P=0.010  150  VE (ml/min x kg)  200  100  ∗  100  .  .  50  50  ε 0  B P=0.008  0.4  0.3  ∗  0.2  .  ∗  0.1  0.0  E  P=0.012 ε  0.3  0.2  ε 0.1  0.0  C P<0.001  2000  ∗  1500  ACR (VE/VO2)  . .  ACR (VE/VO2)  2000  ε  0  VO2 (ml O2/min x kg)  .  VO2 (ml O2/min x kg)  0.4  Day Night  Air  Hypoxia-hypercapnia  . .  ∗ 1000  500  F P=0.07  1500  1000  500  0  0 Summer Fall  Winter Spring  Summer Fall  Winter Spring  Figure 5.1. Day and night values of resting ventilation, oxygen consumption and air convection requirement (ACR) ± s.e.m of turtles breathing a hypoxic-hypercapnic gas (A-C) and air (D-F) (modified from Reyes and Milsom, 2010) in different seasons (N=6 for summer and spring, N=8 for winter and fall) measured at mean seasonal temperatures and natural photocycle (summer: 20.8°C, 16L:8D; fall: 14.7°C, 10L:14D; winter: 9°C, 9L:15D; spring: 14.6°C, 14L:10D). (*) denotes a difference between day and night within seasons under the hypoxia-hypercapnia mixture (H-H) and (ε) denotes differences between day and night within seasons on air (HolmSidak pairwise comparison). 153  Figure 5.2. Mean daytime (open bars) and nighttime (filled bars) changes in oxygen consumption (A), ventilation (B) and air convection requirement (∆ACR, the hyperventilatory response, C) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas. P values reported correspond to overall differences in mean day and night values (two-way RM ANOVA). (*) denote day and night differences within seasons (Holm-Sidak pairwise comparison) (N=6 for summer and spring, N=8 for winter and fall). 154  160  A  P=0.015  B  P=0.014  .  change in VE (ml/min x kg)  140  ∗  120 100 80 60 40 20 0 1400  change in ACR (VE/VO2)  1200  . .  ∗  1000 800 600 400 200 0  Photocycle  Constant darkness  Figure 5.3. Comparison of day (open bars) and night (filled bars) changes in ventilation (A) and ACR (B) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas under the fall photocycle and constant dark. P values reported correspond to overall differences in mean day and night values (two-way RM ANOVA). (*) denote day and night differences within the photocycle and constant dark treatments (Holm-Sidak pairwise comparison) (N=6 for summer and spring, N=8 for winter and fall).  155  Figure 5.4. Ventilation oxygen consumption and air convection requirement (ACR) of turtles breathing a hypoxic-hypercapnic gas (H-H) and air (modified from Reyes and Milsom, 2010) in different seasons. Values recorded at seasonal temperatures are given as well as summer and winter values corrected to 14.7°C. Seasonal differences between values measured under H-H are indicated by different lowercase letters. Seasonal differences between values measured on air are indicated by different uppercase letters (N=6 for summer and spring, N=8 for winter and fall; one-way RM ANOVA). 156  Figure 5.5. Changes in oxygen consumption (A), ventilation (B) and air convection requirement (ACR) (C) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas in different seasons. Significant differences between seasons (P<0.05, one-way RM ANOVA) are indicated by different letters. (N=6 for summer and spring, N=8 for winter and fall). 157  .  change in VO2 (ml O2/min x kg)  0.2  change in VE (ml/min x kg)  b  14.7 C  P=0.005 0.1  0.0 ab  a  -0.1  160  .  .  A  140  a  B P<0.001  c  . 14.7 C  120 a  100  a  80 b  60 40 20 0  1800  . .  change in ACR (VE/VO2)  1600  C P=0.428  .  14.7 C  1400 1200 1000 800 600 400 200 0 Summer Fall  Winter Spring  Figure 5.6. Temperature-corrected (14.7°C) changes in oxygen consumption (A), ventilation (B) and air convection requirement (ACR) (C) as turtles went from breathing air to breathing a hypoxic-hypercapnic gas in different seasons. Significant differences between seasons (P<0.05, one-way RM ANOVA) are indicated by different letters. (N=6 for summer and spring, N=8 for winter and fall).  158  Table 5.1 Temperature and photoperiod used in the experimental series. SEASONAL CUES Season  Treatment Photoperiod  Temperature  Winter  9L:15D  9°C  Spring  14L:10D  14.6°C  Summer  16L:8D  20.8°C  Fall  10L:14D  14.7°C  Constant darkness (outdoor turtles)  0L:24D  13.6°C  159  Table 5.2 Mean ± s.e.m. temperature coefficients (Q10) used to correct values of oxygen consumption, ventilation and air convection requirement measured in the summer and winter to 14.7°C for comparison with other seasons (see Materials and Methods for details). Condition Air Hypoxia-Hypercapnia  .  .  Season  Q10 (VO2)  Q10 (VE)  Q10 (ACR)  Summer  2.85±0.75  2.13±0.76  0.85±0.18  Winter  2.11±0.44  1.77±0.38  0.96±0.46  Summer  3.03±0.6  1.06±0.2  0.36±0.11  Winter  3.54±0.56  2.8±0.61  0.46±0.11  160  Gas/air H-H H-H H-H H-H Air  Spring  Air  Wint  Air  Fall  Air  Sum  Seaso  Table 5.3 Day and night values (± s.e.m.) of tidal volume (ml/kg), breathing frequency (breaths/min), frequency of breathing episodes (episodes/hour) and the percent of time spent in apnea for turtles breathing a hypoxic-hypercapnic gas (H-H) and air (Reyes and Milsom, 2009) in different seasons. (N=6 for summer and spring, N=8 for winter and fall) and measured at the mean seasonal temperatures and natural photocycle (summer: 20.8°C, 16L:8D; fall: 14.7°C, 10L:14D; winter: 9°C, 9L:15D; spring: 14.6°C, 14L:10D). (*) denotes differences between day and night within seasons under hypoxia-hypercapnia (H-H) and (ε) denotes differences between day and night within seasons on air (Holm-Sidak pairwise comparison). Tidal volume  Breathing  Episodes/hour  % Time in apnea  frequency day  night  day  night  day  night  day  night  11.3±0.6  12.7±0.9  9.6±1.2  6.6±0.6  32.2±3.4  29.9±4.1  74.3±3  80.9±1.6  ∗  ∗  5.5±0.6  6.3±0.7  3.1±0.3  2.1±0.1  22.4±2.9  17.2±1.4  92.5±0.9  94.5±0.4  13.8±0.8  13.1±0.8  10.6±1.3  4.9±0.6  33.5±5.4  16.3±3.6  67.7±4.3  84.4±2  ∗ 6.7±0.8  5.7±0.4  2.9±0.3  ∗ 1.5±0.2  6.8±1.9  3.04±0.7  1.6±0.4  ∗ 5.4±0.9  3.3±0.7  1.5±0.1  19.5±1.3  8.6±1.1  20.1±4.5  0.5±0.1  14±2.3  12.6±4.1  43.1±3.9  5±1.5  7.99±0.6  1.8±0.2 ε  87.6±2.8  93.5±1.9  94.1±0.6  98±0.6  ε 28.7±4.5  ∗ 6.1±0.3  95.9±0.5  ∗  ε 4.9±0.6  92.4±1.1 ε  ∗  ε 20.3±1.4  10.3±1.7  ε  ε 10.2±2.3  15.9±1.9  ∗  72.6±3.5  83.6±2.3  ∗ 1.2±0.2  14.9±1.5 ε  10.7±2.1  94.9±0.4  96.6±0.5  ε  161  Gas/air H-H air H-H H-H air  Spring  air  Winter  air  Fall  H-H  Summer  Season  Table 5.4 Mean ± s.e.m. tidal volume (ml/kg), breathing frequency (breaths/min), frequency of breathing episodes (episodes/hour), breaths per episode, instantaneous breathing frequency (breaths/min in an episode) and percent time spent in apnea of turtles breathing a hypoxichypercapnic gas (H-H) and air (Reyes and Milsom, 2009) and exposed to seasonal conditions (summer: 20.8°C, 16L:8D; fall: 14.7°C, 10L:14D; winter: 9°C, 9L:15D; spring: 14.6°C, 14L:10D). Significant differences between seasons (P<0.05, one-way RM ANOVA) are indicated by different lowercase letters (H-H) and uppercase letters (air) (N=6 for summer and spring, N=8 for winter and fall). Tidal  Breathing  Frequency  Breaths per  Inst.  % time  volume  frequency  of episodes  episode  frequency  apnea  11.8±0.7  8.6±0.9  31.4±3.3  17.8±2.6  0.60±0.03  76.5±2.4  ab  a  ac  a  a  5.8±0.6  2.8±0.3  20.6±2.3  8.5±0.8  0.70±0.02  93.2±0.7  AB  A  A  A  A  A  13.4±0.7  7.2±0.8  23.3±4.1  26.1±5.4  0.50±0.03  77.8±2.9  ab  a  a  a  a  6.1±0.5  2.1±0.2  12.5±1.6  11.3±2.3  0.60±0.03  94.5±0.7  A  B  B  A  A  AC  8.1±1.8  2.1±0.4  15.4±4.1  5.8±1.4  0.30±0.05  91.3±1.9  a  b  b  b  b  4.1±0.6  0.8±0.1  8.4±1.6  3.6±0.4  0.20±0.04  96.6±0.3  B  C  C  B  B  B  19.9±1.2  7.1±0.8  37.1±2.3  13±1.1  0.50±0.03  77.2±2.6  b  a  bc  a  a  6.8±0.3  1.5±0.1  13±1.5  7.5±1.1  0.60±0.03  95.6±0.3  A  B  B  A  A  BC  162  Chapter 6: General discussion and conclusions Vertebrates require a continuous supply of oxygen to metabolizing tissue in order to survive. Oxygen is transported by the cardio-respiratory system in a series of steps: 1) convection of air or water to the gas exchange surface; 2) diffusion into the blood; 3) convective transport by the blood (perfusion) and 4) diffusion from blood to the tissue where oxygen is consumed by the mitochondria to produce ATP (Glass, 1992). At some point in their life history, many vertebrates experience reduced levels of oxygen either by exposure to hypoxia (environmental) or by hypoxemia or anemia (Boutilier, 2001). The abilities and strategies implemented to sustain metabolism when faced with low oxygen vary widely amongst vertebrates (Milsom, 1990). All strategies require the presence of control systems, most of which are dependent on peripheral arterial chemoreceptors that sense the levels of oxygen in the environment (water/air/blood) and trigger cardio-respiratory adjustments so that O2 is supplied to satisfy the oxygen demands of tissues (Shelton et al., 1986). Thus, these chemoreceptors regulate ventilation and perfusion and are the first line of defense against altered conditions for oxygen transport. Indeed, their importance is reflected by the fact that peripheral arterial chemoreceptors are found in all vertebrates studied to date (Adams, 1958; Jones and Milsom, 1982; Milsom 1998; Milsom and Burleson, 2007).  6.1  Importance of studying peripheral chemoreceptors in amphibians and reptiles Recently, much research has focused on the oxygen sensing cells (NEC’s) in fish with the  purpose of establishing homologies with peripheral arterial chemoreceptors of mammals, particularly the carotid bodies (for review see Milsom and Burleson, 2007; Jonz and Nurse, 2009, 2012; Milsom, 2012; Porteus et al., 2012; Zachar and Jonz, 2012). The low solubility of oxygen in water and therefore low oxygen content (CO2), and the common occurrence of aquatic hypoxia (Dejours, 1981) mean that O2-sensing receptors are critical for the survival of fish as they regulate gill ventilation and heart rate to ensure an adequate oxygen supply to the tissues (Taylor et al., 1999; Milsom, 1998; Milsom and Burleson, 2007). For terrestrial vertebrates, even though CO2 in air is not typically limiting, many (particularly amphibians and reptiles) commonly experience low levels of oxygen on a seasonal, daily or a minute by minute basis. For instance, anurans and reptiles often inhabit environments that are severely hypoxic (Ultsch, 1973; Heisler et al., 1982). These include burrows or lakes that freeze over in winter or at high altitudes (Ultsch, 1989; Burggren and Shelton, 1979). Furthermore, 163  arterial blood gases in amphibians and reptiles oscillate due to their intermittent breathing pattern (Shelton et al., 1986; Milsom, 1988; Milsom, 1990; Smatresk, 1990; West and Van Vliet, 1992) and the presence of cardiac shunts (Wood, 1982, 1984; Burggren, 1987; Wang and Hicks, 1996b; Hicks, 1998; Hicks, 2002). Since there is no evidence for central oxygen chemoreceptors in these animals (Johnson et al., 1998; Gargaglioni and Branco, 2004), input from peripheral arterial chemoreceptors alone regulate arterial blood gases during acute hypoxia or hypoxemia (Shelton et al., 1986; Milsom, 1990). Surprisingly, our knowledge of oxygen sensing structures in amphibians and particularly in reptiles is limited. In amphibians, studies of oxygen sensing have focused on the carotid labyrinth and the only reptiles where functional chemosensory areas have been located are chelonians. In order to draw conclusions about the evolution of peripheral arterial chemoreceptors, more research is needed on lower tetrapods that breathe air like mammals, but have unique characteristics, such as the ability to shunt, and low metabolic rates, that may have imposed different evolutionary forces upon the cardio-respiratory control systems. It has been proposed that the widely distributed NEC’s in fish are the precursors of the tightly clustered population of O2-sensing cells seen in the carotid body of mammals (Jonz and Nurse, 2003; Milsom and Burleson, 2007; Jonz and Nurse, 2009; Zachar and Jonz, 2012). I asked where do oxygen sensing cells of amphibians and reptiles fit in the evolution of peripheral chemoreceptors? The main objectives of my thesis were to identify and characterize putative peripheral arterial chemoreceptors in reptiles and amphibians, based on their location, distribution, innervation and neurochemical content, in order to infer if O2-sensing mechanisms have been conserved throughout the vertebrate taxa. Since anurans (Ishii et al., 1985b), chelonians (Ishii et al., 1985a; Ishii and Ishii, 1986) and snakes (present study) have multiple chemosensory sites, I sought to determine if distinct chemoreceptor groups had different stimulus modalities and reflex roles. I also asked if there was plasticity in the chemoreflexes produced by oxygen chemoreceptors that could lead to daily and seasonal changes in the breathing pattern potentially affecting the survival of these animals.  6.2  Major findings and implications My research provides some insights into the evolution of peripheral arterial chemoreceptors.  I located functional peripheral arterial chemoreceptors in rattlesnakes (Chapter 4), and characterized putative oxygen sensing cells in the chemosensory areas of bullfrogs, red-eared 164  sliders and rattlesnakes (Chapters 2, 3 and 4). Using a comparative approach, I propose that some aspects of O2-sensing structures are highly conserved in vertebrates while others show marked differences among taxa. In contrast to the carotid and aortic bodies of mammals, I showed that chemosensory areas in rattlesnakes do not have distinct reflex roles. All sites respond to low oxygen (mimicked by NaCN) by regulating both the respiratory and cardiovascular systems (Chapter 4). I showed that adjustments in the cardiovascular system include changes in cardiac shunt, and that peripheral chemoreceptor stimulation directly affects the pattern and magnitude of shunts in rattlesnakes. Furthermore, my dissertation showed daily but not seasonal plasticity in the sensitivity of the respiratory reflex that is independent of changes in metabolism, and affects the breathing pattern of turtles (Chapter 5). In the following section I review some of my major findings and discuss them in the context of what is known about peripheral chemoreceptors in fish and mammals to provide a more complete understanding of how peripheral chemoreceptors have evolved throughout the vertebrate taxa.  6.2.1  Location of chemosensory areas in vertebrates  I found three functional chemosensory areas in rattlesnakes located at the carotid bifurcation, aortic arch and pulmonary artery, derivatives of the 3rd, 4th and 6th pharyngeal arches respectively (Chapter 4). My findings together with studies on fish, amphibians, turtles and mammals suggest that the location of peripheral arterial chemoreceptors is highly conserved throughout the vertebrate taxa, and that they are always associated with pharyngeal arches or their derivatives (i.e. aortic arches) (reviewed by Milsom and Burleson, 2007). In gill breathers, O2 chemoreceptors are present in all extant derivatives of the pharyngeal arches. For instance, NEC’s have been located in all gill arches (pharyngeal arches 3-6) of fish and larval amphibians (reviewed by Evans et al., 2005; Jonz and Nurse, 2009; Jonz and Nurse, 2012). In adult anurans, chelonians and snakes, chemosensory areas are found in derivatives of the 3rd (carotid labyrinth, common carotid artery or carotid bifurcation in anurans, turtles or snakes; respectively), 4th (aorta) and 6th (pulmocutaneous or pulmonary artery) pharyngeal arches (Ishii et al., 1966; Lillo, 1980; Hoffmann and de Souza, 1982; Ishii et al., 1985a, Ishii et al., 1985b; Ishii and Ishii, 1986; Wang et al., 2004; Chapter 4), which are the only aortic arches present in these groups, as the 5th aortic arch is lost during development (Kardong, 2006). In birds and mammals, chemosensory areas are believed to be confined to two locations in derivatives of the 3rd and 4th pharyngeal arches. In birds, peripheral chemoreceptors have been identified in the aorta (Nye and Powell, 165  1984; Ito et al., 1997) and in the carotid body at the base of the common carotid artery (AbdelMagied and King 1978; Kameda, 2002). As in turtles, the carotid bifurcation in birds is secondarily derived (Adams, 1958; Adams, 1962). In mammals, the carotid bodies and aortic bodies are located at the carotid bifurcation and along the aorta, respectively (Gonzalez et al., 1994). The presence of glomus-like cells in the pulmonary artery of mammals has been previously described (Krahl, 1962), but no functional evidence has been reported (Comroe, 1964) and all evidence points to the lack of O2 chemoreceptors in derivatives of the 6th pharyngeal arch in birds and mammals. Interestingly, in fetal cats and dogs aortic bodies are perfused by the pulmonary artery (Coleridge et al., 1967; Kollmeyer and Kleinman, 1975), but after birth they are solely perfused by the aorta (Comroe, 1964). The function of aortic bodies perfused by pulmonary arterial blood remains unknown, but it is thought to play a role in the initiation of postnatal respiration (Marshall, 1994). All mammals examined so far have carotid bodies (Marshall, 1994), although aortic bodies are present in cats and dogs, they are commonly missing or not well developed in mice, rats and rabbits (Marshall, 1994). The presence of aortic bodies does not seem as consistent as that of the carotid body, and their contribution to the hypoxic response is less important (Fitzgerald and Lahiri, 1986). This is also the case in birds (Smatresk, 1990). Pulmonary receptors appear to be missing in both groups. The reduced number of chemosensory areas in birds and mammals is perhaps related to their completely divided dual circulation and the fact that the gas composition is identical in all systemic arteries. To summarize, chemosensory areas are present in derivatives of all extant pharyngeal arches in fish, amphibians and reptiles. O2-sensing receptors in some fish are also found in the orobranchial cavity (Smatresk, 1990). Chemosensory areas in birds and mammals are limited to two locations (derivatives of the 3rd and 4th pharyngeal arches). My findings in amphibians and reptiles support the hypothesis of a trend to reduce the number of chemosensory areas from fish to amphibians and reptiles to mammals.  6.2.2  Anatomical features and neurochemical content of O2-sensing cells  I found putative oxygen sensing cells in the chemosensory areas of bullfrogs (Rana catesbeiana, Chapter 2), red-eared sliders (Trachemys scripta elegans, Chapter 3) and rattlesnakes (Crotalus durissus, Chapter 4), identified based on their neurochemical content, morphology, innervation and distribution. My dissertation demonstrated the presence of multiple neurotransmitters in the chemosensory areas of bullfrogs, red-eared sliders and rattlesnakes. 166  Previous studies have postulated the presence of biogenic monoamines (catecholamines and serotonin) in the putative chemosensory areas of amphibians and reptiles using a formaldehyde vapor exposure technique that is not very specific. The limitations of this technique were explained in Chapter 1 and 2. In this section I discuss some of the characteristics that make the cells described in Chapters 2, 3 and 4 potential O2 chemoreceptors and compare them with fish NEC’s and avian and mammalian glomus cells to give support to their proposed chemosensory function.  6.2.2.1  Anatomical features of putative oxygen sensing cells  Putative chemoreceptor cells in bullfrogs, red-eared sliders and rattlesnakes were very similar in shape and size. They were oval, with large nuclei and 10.0-12.0 µm in diameter. Similarly, the NEC’s of fish and the glomus cells of the carotid and aortic bodies of mammals are small (10 µm) ovoid cells with large nuclei (Jonz and Nurse, 2009, 2012; Eyzaguirre et al., 1983). As in the glomus cells of birds and mammals (Gonzalez et al., 1994; Kameda, 2002), some catecholamine-containing cells in the carotid labyrinth of bullfrogs had processes, but these cells tended to be more pyramidal like in shape. The morphology of oxygen sensing cells is generally conserved in vertebrates, suggesting that these cells originated from a common precursor cell. Glomus cells in birds and mammals are derived from neural crest cells (Pearse et al., 1973; Eyzaguirre et al., 1983). My findings suggest that this is the case at least for serotonergic (5HT) cells in the chemosensory areas of bullfrogs (Chapter 2). Although I could not confirm the origin of putative oxygen sensing cells in turtles and rattlesnakes, it is possible that these cells are derived from the neural crest, but they are mature, non-migratory cells that no longer express the human natural killer 1 (HNK-1) marker.  6.2.2.2  Neurochemical content of putative oxygen sensing cells  I found multiple neurotransmitters in the chemosensory areas of bullfrogs (Chapter 2), redeared sliders (Chapter 3) and rattlesnakes (Chapter 4) (Fig. 6.1). One of the major findings of this dissertation was the consistent presence of 5HT-containing cells in all the chemosensory areas of the amphibians and reptiles studied. Although their chemosensory role is not proven in these species, their frequent occurrence and resemblance to O2 chemoreceptors in fish and mammals makes them good candidates for O2-sensing cells. In fish, 5HT appears to be the main neurotransmitter involved in oxygen chemotransduction (Dunel-Erb et al., 1982; Bailly et al., 167  1992; Zaccone et al., 1992, 1994; 1997; Jonz and Nurse, 2003; Jonz et al., 2004; Saltys et al., 2006; Bailly, 2009; Jonz and Nurse, 2009, 2012). For instance serotonergic NEC’s in zebrafish (Danio rerio) have been shown to depolarize when stimulated by hypoxia (Jonz et al., 2004) and perfusion of 5HT on isolated gills of trout (Ochorhynchus mykiss) increased afferent nerve discharge (Burleson and Milsom, 1995). In mammals, 5HT seems to only modulate the sensitivity of the carotid body (Nurse, 2005; 2010). Aortic chemoreceptors in birds release 5HT when stimulated by hypoxia (Ito et al., 1999). The involvement of 5HT in oxygen chemotransduction seems to be maintained throughout the vertebrate taxa. Other neurotransmitters are involved in oxygen sensing in birds and mammals. Acetylcholine (ACh) and ATP appear to be the most important neurotransmitters in hypoxia chemotransduction (Eyzaguirre and Zapata, 1984; Zhang et al., 2000; Nurse, 2005; Shirahata et al., 2007; Nurse, 2010) in the mammalian carotid body. They are also present in the aortic bodies of mammals, but their role in signal transduction has not yet been determined (Piskuric et al., 2011; Piskuric and Nurse, 2013). Catecholamines, particularly dopamine are expressed in glomus cells of both the aortic and carotid bodies of mammals (Gonzlaez et al., 1994; Piskuric et al., 2011; Piskuric and Nurse, 2012) and in most species it is an inhibitory neurotransmitter (reviewed by Eyzaguirre et al., 1983; Gonzalez et al., 1994). All these neurochemicals are contained within the same glomus cell (i.e. colocalize). Glomus cells in the carotid body of birds express tyrosine hydroxylase (TH), indicating the presence of catecholamines that colocalize with 5HT (Kameda, 2002). In fish, physiological (trout) and behavioral (mangrove rivulus, Kryptolebias marmoratus) responses to ACh indicate that this neurotransmitter contributes to the hypoxic response (Burleson and Milsom, 1995; Regan et al., 2011), but the mechanisms involved remain unknown. ACh-containing cells have been found in the gills of fish, but never colocalizing with 5HT or with the marker for synaptic vesicles. Thus, it has been proposed that ACh could be released from either afferent or efferent nerve endings and could act on NEC’s to release 5HT (Porteus et al., 2012). The presence of catecholamines in fish NEC’s is equivocal. Several studies have failed to demonstrate the presence of catecholamines in trout, goldfish (reviewed by Porteus et al., 2012) and catfish (Zaccone et al., 2003). However, TH expression has been shown in NEC’s cultures of the channel catfish (Ictalurus punctatus), where two subpopulations were identified, one that depolarized and one that hyperpolarized when exposed to hypoxia (Burleson et al., 2006). 168  In the putative O2 chemoreceptor cells of bullfrogs, red-eared sliders and rattlesnakes I found neurotransmitters other than 5HT, but the neurochemical content differed between the two groups. Catecholamines were present in all the chemosensory areas of bullfrogs (Chapter 2) and injections of epinephrine in the carotid labyrinth of toads (Bufo marinus) stimulate chemoreceptor discharge, while dopamine inhibits it (Van Vliet and West, 1992), as is the case in the carotid body of mammals (Gonzalez et al., 1994). I showed that a small subpopulation of cells in the carotid labyrinth of bullfrogs contained both TH and 5HT, in the aorta and pulmocutaneous artery; however, these neurotransmitters were present in different putative oxygen sensing cells. No evidence for cholinergic cells was found in any of the chemosensory areas of bullfrogs (Chapter 2) or in the carotid labyrinths of other amphibians, but injections of ACh in these regions has been shown to increase chemosensory discharge (Ishii and Ishii, 1967, Ishii et al., 1985a) and decrease the internal carotid outflow (Kusakabe et al., 1987). Thus, as in fish, it appears that ACh may participate in efferent control in amphibians by modulating chemoreceptor sensitivity and vascular tone (West and Van Vliet, 1992). In contrast to amphibians, catecholamines were absent from the chemosensory areas of reptiles (red-eared sliders and rattlesnakes; Chapters 3 and 4, respectively). Instead, AChcontaining cells were found in all chemoreceptive sites of red-eared sliders and rattlesnakes, except for the carotid bifurcation in the snakes. Although the chemosensory role of AChcontaining cells has not been demonstrated in reptiles, the morphology and arrangement (see Organization and innervation of putative O2-sensing cells section below) of these cells is very similar to that of granulated cells in the carotid artery, aorta and pulmonary artery of tortoises. Based on their anatomical features and the presence of dense-cored vesicles, granulated cells in tortoises have been proposed to be functional chemoreceptors (Kusakabe et al., 1988). The anatomical resemblance of ACh cell clusters described in this dissertation to the granulated cells reported by Kusakabe et al. (1988) supports their chemosensory role. To summarize, my findings in amphibians and reptiles support the observation of a trend to increase the neurochemical content in the chemosensory areas of vertebrates from fish to amphibians and reptiles to mammals (Milsom and Burleson, 2007). 5HT is involved in oxygen chemotransduction in all vertebrates, but its role in this process seems to changes throughout the vertebrate taxa. Thus, 5HT is the main neurotransmitter in fish, but has only a neuromodulatory role in mammals, where other neurotransmitters, such as ATP and ACh play a key role (Eyzaguirre and Zapata, 1984; Zhang et al., 2000; Nurse, 2005,; Shirahata et al., 2007; Nurse, 169  2010; Jonz and Nurse, 2009, 2012). The mechanisms of oxygen sensing get more complex, by involving neurotransmitters other than 5HT, first in different putative O2 sensing cells as in bullfrogs, red-eared sliders and rattlesnakes; and then within the same glomus cell in the carotid body of mammals. Having more neurotransmitters and neuromodulators involved in chemoreception may allow for fine tuning of the hypoxic response. For instance, excitatory and inhibitory neurochemicals may act together to ensure that the response is maintained throughout the duration of the stimulus (Prabhakar, 2006). Another possibility is that cell types containing different neurotransmitters could sense different stimulus modalities.  6.2.3  Organization and innervation of putative O2-sensing cells  In addition to a trend of increasing neurochemical content of chemoreceptor cells from fish to mammals, the organization of these cells also appear to become more complex. In this dissertation, I show that the organization of putative chemoreceptor cells in bullfrogs appears to be at an intermediate stage between fish and mammals (Chapter 2). The arrangement of cholinergic putative O2-sensing cells in reptiles (red-eared sliders and rattlesnakes) resembles that of the aortic bodies of mammals, but serotonergic cells retain some of the characteristic arrangement of fish NEC’s. Furthermore, the innervation pattern of the chemosensory areas is conserved throughout the vertebrate taxa.  6.2.3.1  Organization  The NEC’s of fish generally occur as solitary cells spread throughout the gill filament or in some species also in the lamellae (Dunel-Erb et al., 1982; Bailly et al., 1992; Zaccone et al., 1992; Jonz and Nurse, 2003; Coolidge et al, 2008; Saltys et al, 2006; Jonz and Nurse, 2009). Thus, the organization of O2-sensors in fish is not very complex compared to that of the mammalian carotid and aortic bodies. The carotid body consists of a tight conglomeration of thousands of glomus cells surrounded by sustentacular cells (Gonzalez et al., 1994; Jonz and Nurse, 2012). In contrast, aortic bodies are broadly distributed above and below the aorta and (Comroe, 1964) are formed by clusters of 5-50 glomus and sustentacular cells (reviewed by Jonz and Nurse, 2012). Similar to the aortic bodies of mammals, the carotid and aortic bodies of birds consist of small clusters of glomus cells widely distributed throughout the aorta, carotid body and adjacent common carotid artery (Miyoshi et al., 1995; Kameda, 2002). The organization of cells in clusters is significant because it permits communication through chemical and electrical 170  synapses between neighbouring oxygen sensing cells (Gonzalez et al., 1994; Nurse, 2010). Putative oxygen sensing cells in bullfrogs occur throughout the carotid labyrinth, aorta and pulmocutaneous artery, organized in singles or in groups of 2-3 (Chapter 2), suggesting an intermediate stage between fish and mammals. The organization of the two types of putative oxygen sensing cells is very different in reptiles (red-eared sliders and rattlesnakes). Serotonergic cells are mostly found throughout the chemosensory areas of turtles and snakes solitary or in loose groups, while clusters of 2-5 cells are less common. Cholinergic cells consistently occurred in the same region of the blood vessels and in large clusters in both redeared sliders and rattlesnakes. Similarly, large clusters of granulated cells have also been found in the aorta, carotid and pulmonary arteries of tortoises, with junctional specializations between them, which could allow for communication between cells (Kusakabe et al., 1988). The organization of serotonergic cells in red-eared sliders and rattlesnakes resemble that of fish, while cholinergic cells are arranged similarly to the aortic body of mammals.  6.2.3.2  Innervation  Afferent fibers of the glossopharyngeal (IX) and vagus (X) nerves are responsible for relaying messages from chemoreceptor cells to the NTS, where the signal is integrated and cardio-respiratory adjustments triggered to maintain blood gas homeostasis (Gonzalez et al., 1994; Nurse, 2005). The release of neurotransmitters from O2-sensing cells increases sensory nerve discharge. Singles and clusters of putative oxygen sensing cells in bullfrogs (Chapter 2) and red-eared sliders (Chapter 3) are innervated by the vagus nerve. Serotonergic and catecholamine-containing cells in the carotid labyrinth of bullfrogs were also innervated by the glossopharyngeal nerve, suggesting that these are primary sensors. The innervation pattern is similar to that of single NEC’s in fish and clusters of glomus cells in the carotid body of mammals (Kondo, 1976; Jonz and Nurse, 2009, 2012). Unfortunately I cannot distinguish between afferent and efferent innervation with the neuronal tracer I used. Cholera Toxin B is both an anterograde and retrograde tracer (Kobbert et al., 2000). However, afferent and efferent nerve endings have been described in amphibians, tortoises, fish and birds (King et al., 1975; Dunel-erb et al., 1982; Kusakabe et al., 1988; West and Van Vliet, 1992) and presumably the efferent nerves play a role in modulating the sensitivity of chemoreceptors as in mammals (Campanucci and Nurse, 2007; Shirahata et al., 2007). Some putative chemoreceptor cells in bullfrogs and turtles did not co-label with the neuronal tracer. It is possible that the nerve fibers 171  innervating some of the cells were not visible in tissue sections. An alternative is that they are non-innervated cells that have a paracrine function on the surrounding vascular tissue or on innervated glomus-like cells, as has been proposed for non-innervated NEC’s in fish (Coolidge et al., 2008). Similarly, about 5% and 12% of rat and cat glomus cells in the carotid body are not innervated (Kondo, 1976; Eyzaguirre et al., 1983). In conclusion, the innervation by the IX and X cranial nerves to the chemosensory areas and putative oxygen sensing cells seems to be similar in all vertebrates studied to date.  6.2.4  Distribution of putative O2-sensing cells related to stimulus specificity and reflex  roles It has proven difficult to determine the stimulus specificity (oxygen tension, PO2 or content, CaO2) of peripheral chemoreceptors in vertebrates. The ventilatory response to hypoxia is thought to arise from changes in PO2 in most vertebrates. That said, there is a strong correlation between ventilation and CaO2 of arterial blood in many species. For instance, the gill ventilatory response in many fish correlates better to changes in CaO2 than PO2 (Randall, 1982; Smith and Jones, 1982). The hypoxic ventilatory response also correlates better with changes in CaO2 than PO2 in turtles with changing temperature (Glass et al., 1983). Inhalation of carbon monoxide triggers an increase in ventilation in birds (Tschorn and Fedde, 1974). Hypoxia produced increases in ventilation and heart rate in the toad (Bufo paracnemis), while anemia only affected heart rate (Wang et al., 1994). In turtles, hypoxia did not affect blood flow, but anemia reduced the R-L shunt independent of changes in PO2 and ventilation did not change with a fall in CaO2 (Wang et al., 1997). These studies suggest that there are different groups of chemoreceptors with specific stimulus modalities that selectively regulate the respiratory or the cardiovascular systems (Wang et al., 1994; Wang et al., 1997). Wang et al. (1997) proposed three hypotheses that could explain the lack of a hypoxic ventilatory response during anemia: First, the existence of CaO2 sensitive chemoreceptors in the systemic circulation that preferentially regulate the cardiovascular system (Wang et al., 1997), similar to the aortic bodies of mammals (Lahiri et al., 1981). Second, the presence of a chemoreceptor group in the venous circulation or pulmocutaneous and pulmonary arteries of amphibians and reptiles perfused by mixed venous blood and therefore sensitive to both PO2 and O2 content (Wang et al., 1997). Third, the presence of two distinct populations of  172  chemoreceptors within the pulmonary artery: a PO2-sensitive receptor population and an O2 content chemoreceptor group (Wang et al., 2004). One of the goals of this dissertation was to determine the stimulus specificity of the different groups of peripheral arterial chemoreceptors (carotid, aortic and pulmonary O2 receptors) in rattlesnakes. For this I used a bolus injection of saline to simulate low CaO2 at a normal PO2, and compared it to a bolus injection of blood with normal CaO2 and PO2 (same volume and time of injection). None of the chemosensory sites were stimulated by either the bolus of blood (as expected) or the saline injection, suggesting either that peripheral arterial chemoreceptors in rattlesnakes are not stimulated by a fall in the CaO2, or that the stimulus was not strong enough to elicit a response. Therefore, I am unable to conclusively show whether CaO2 or PO2 stimulates breathing and affects shunt. Thus, in this section I discuss the three hypotheses proposed by Wang et al. (1997, 2004) mentioned above within the framework of the anatomical features, distribution and reflex response of peripheral arterial chemoreceptors in bullfrogs, red-eared sliders and rattlesnakes described in this thesis.  6.2.4.1  O2-content sensitive chemoreceptors responsible for cardiovascular control  (hypothesis 1) O2 content chemoreceptors have only been described in mammals whose aortic bodies respond to low CaO2 as well as PO2 and primarily regulate the cardiovascular system (Jones and Daly, 1997), while the carotid bodies are only sensitive to PO2 and are mainly responsible of ventilatory control (Lahiri et al., 1981). Lahiri et al. (1981) proposed that the different sensitivities of aortic and carotid bodies depend on their blood supplies. The blood flow of the carotid body is large compared to that of the aortic body. At the tissue level, PO2 is a function of CO2. According to the Fick equation (V̇ O2 = Q̇ x (a-v)CO2), if the oxygen consumption of both receptors is similar, then a change in O2 content will lead to a larger change in PO2 in regions of low blood flow than regions of high blood flow. The net result is that the aortic bodies will be more sensitive to the CaO2 and hemoglobin oxygen affinity of arterial blood than the carotid bodies. The anatomical data for bullfrogs described in this dissertation supports this scenario. Putative oxygen sensing cells in the carotid labyrinth occur close to lumen sinusoids, while those in the aorta and pulmocutaneous artery are embedded in the blood vessel wall. The difference in distribution of these cells could be significant for their sensitivities to different stimulus 173  modalities. It is possible that the location of cells in the carotid labyrinth is optimal to sense changes in PO2 but not oxygen content, and the location of cells in the pulmocutaneous artery, further away from the lumen (i.e. blood flow) will be affected by changes in PO2 as well as oxygen content. In support to this hypothesis, it has been shown that chemoreceptors in the carotid labyrinth of toads respond to a fall in PO2, but not to changes in oxygen content (Van Vliet and West, 1992; Wang et al., 1994). However, the stimulus modality of the other chemosensory areas in amphibians has not been investigated. My findings in reptiles do not provide evidence for an CaO2-sensitive chemoreceptor, as seen in mammals because the distribution of the putative O2-sensing cells did not differ between the three chemosensory areas in red-eared sliders and rattlesnakes.  6.2.4.2  Cardiovascular regulation by PO2- sensitive chemoreceptors in the pulmocutaneous or  pulmonary artery (hypothesis 2) One hypothesis states that O2-sensing chemoreceptors on the pulmonary artery regulate the cardiovascular system and that carotid and aortic chemoreceptors adjust ventilation (Wang et al., 1997). Peripheral chemoreceptors are found in the pulmocutaneous artery of anurans (Lillo, 1980; Hoffman and de Souza, 1982; Wang et al., 2004) and in the pulmonary arteries of turtles (Geoclemmys reevesii) (Ishii et al., 1985b) and rattlesnakes (Crotalus durissus, Chapter 4). Since these chemoreceptors are perfused by mixed venous blood and both PvO2 and venous oxygen content (CvO2) depend on the systemic flow and arterial oxygen content relative to metabolic rate, a fall in arterial oxygen content will reduce PvO2 and CvO2 assuming that blood flow and metabolic rate remain unchanged (Wang et al., 1997). Although these chemoreceptors are sensitive to PO2, they will be affected by changes in the content of oxygen. The physiological data on rattlesnakes in this dissertation do not support this hypothesis, because both ventilatory and cardiovascular adjustments were produced after stimulation of all chemosensory areas with NaCN. Similarly, injections of cyanide in the carotid labyrinth and pulmocutaneous artery of toads affected both ventilation and the cardiac shunt (Wang et al., 2004).  6.2.4.3  Presence of two sets of chemoreceptors within the pulmonary artery (hypothesis 3)  The lack of evidence to support the hypothesis that PO2 chemoreceptors in the pulmonary circulation could selectively affect the cardiovascular system led Wang et al. (2004) to postulate 174  that two sets of chemoreceptors exist within the pulmonary circulation, one that senses PO2 and regulates ventilation and a second that responds to oxygen content and produces cardiovascular adjustments. This hypothesis has not been tested and my findings may support it if the different types of putative chemoreceptor cells (5HT- and ACh-containing cells in red-eared sliders and rattlesnakes) are each sensitive to a different stimulus modality. Thus, hypoxia and anemia would trigger the release of different neurotransmitters. However, this would not explain how stimulation of the carotid bifurcation of rattlesnakes triggers changes in both ventilation and cardiac shunt, since only serotonergic cells occur in this area. Furthermore, it raises the question of why chemoreceptors in the carotid labyrinth of amphibians are only sensitive to PO2 and not CaO2 (Van Vliet and West, 1992), if the neurochemical content of this area is the same as that of the aorta and pulmocutaneous artery in this group. The same is true for the carotid and aortic bodies in mammals, which have similar neurotransmitters. The NEC’s of fish are either oriented towards the external environment (Jonz and Nurse, 2003) or towards the blood (internal) and respond to changes in water PO2 or blood oxygen tension or content, respectively (Milsom and Burleson, 2007; Jonz and Nurse, 2009). Although variable between species, fish where the bradycardia response to hypoxia is confined to the first gill arch tend to be sensitive to water oxygen tension, while the remaining gill arches respond to external and internal O2 changes and primarily regulate ventilation (Milsom and Burleson, 2007; Milsom, 2012). So it appears that depending on their locations different chemoreceptors are able to sense water or blood O2 levels and either regulate the cardiovascular or respiratory systems. The neurochemical content, however, is the same for externally and internally oriented NEC’s (Porteus et al., 2012), suggesting that is not the release of different neurotransmitters that produces different reflex responses. To conclude, regardless of the stimulus specificity of distinct groups of peripheral arterial chemoreceptors, my findings and that of others (Wang et al., 2004) suggest that all chemosensory areas are capable of producing cardiovascular and ventilatory adjustments in toads and rattlesnakes. Thus, it is possible that the relative reflexes produced under hypoxia or anemia could be the product of a complex integration of the stimuli in the respiratory and cardiovascular centers of the brain (Porteus et al., 2011).  175  6.3  Caveats Although my study identified and characterized putative oxygen sensing cells in the  chemosensory areas of red-eared sliders, rattlesnakes and bullfrogs, chemosensory cells have only been examined in a small fraction of vertebrate diversity. For instance, there are 2700 species of snakes, over 250 species of chelonians and over 5000 species of anurans. There is great variation within these broad taxonomic groups in life histories and cardiovascular anatomy, particularly in reptiles, that could impose different selection pressures upon the cardiorespiratory control systems. Great caution must therefore be exercised when generalizing across the full array of vertebrate diversity, and further studies are needed to determine the generality of the variation in peripheral chemoreceptor features that I identified. Furthermore, eventhough the anatomical features that characterize putative oxygen sensing cells in red-eared sliders, rattlesnakes and bullfrogs suggest a chemosensory role, it is necessary to confirm their chemoreceptive nature before deriving any definitive conclusions about the phylogenetic changes in the peripheral chemoreceptors of vertebrates. Some of these ideas are further discussed in the next section.  6.4  Future research Very few studies have investigated peripheral arterial chemoreceptors in lower tetrapods,  particularly reptiles; this area of study remains wide open for exploration. To date, chemosensory areas have been located in anurans, chelonians and rattlesnakes. I identified putative O2-sensing cells in these areas based on their neurochemical content, innervation and anatomical features. Other studies have identified granulated cells in anurans and chelonians using electron microscopy (Ishii and Oosaki, 1969; Kusakabe et al., 1988). These cells resemble putative oxygen sensing cells described in this dissertation. However, to derive conclusions about the evolution of peripheral arterial chemoreceptors, the chemosensory role of these cells needs to be confirmed using electrophysiological studies. A first step in this direction would be to record from single nerve fibers while stimulating these cells either in situ or in vitro. Single fiber nerve recording, however, may prove difficult as these areas are often innervated by numerous, small nerve branches. An alternative approach would be to use patch clamp electrophysiology on isolated cells and record their response to hypoxia. It has been shown that glomus cells in the carotid body of rats (Wang et al., 2008) undergo hypertrophy and cell proliferation and the NEC’s of zebrafish and mangrove rivulus increase in 176  size (Jonz et al., 2004; Regan et al., 2001) when exposed to sustained hypoxia. Immunohistochemical experiments using the markers for acetylcholine and serotonin before and after exposure to sustained hypoxia may be a useful tool to determine the O2-sensing role of the cells I described in this dissertation. Because different groups of reptiles differ broadly in their cardiovascular anatomy and therefore the magnitude of the cardiac shunt (Hicks, 1998) it would be interesting to determine if the location and regulatory role of peripheral chemoreceptors varies between species with large degrees of cardiac shunt and those with almost complete separation of the systemic and pulmonary circulations. Snakes could be a useful group for this comparison as they all have an undivided ventricle, but the magnitude of the cardiac shunt varies dramatically among species. Since I have identified chemosensory areas in rattlesnakes, this group could be compared to pythons (Python molurus) in which the muscular ridge effectively separates the ventricle during systole (Wang et al., 2003). Furthermore, locating and characterizing peripheral chemoreceptors in varanid lizards and crocodilians may prove fruitful in determining how this control system has changed in vertebrates. Varanid lizards show ventricular pressure separation and high systemic blood pressure, presumably due to their high metabolic rates (Burggren and Johansen, 1982). I discussed the finding that reduction in the chemosensory sites of mammals are probably the result of evolutionary pressures acting on control systems to ensure a precise regulation of arterial gases so that oxygen supply is not compromised. Therefore, locating peripheral chemoreceptors in an extremely active reptile, such as varanid lizards, with an undivided ventricle will help in testing this hypothesis. Crocodilians are the only reptiles with a four chamber heart. Their cardiovascular anatomy does not allow for a R-L shunt, but they can still produce a L-R shunt (Hicks, 1998). Comparing these groups with radically divergent anatomies would allow us to test if changes in the cardiovascular anatomy (3 to 4 chamber heart) correlate with changes in the number, distribution and reflex roles of peripheral chemoreceptors.  6.5  Concluding remarks The main objectives of my dissertation were to determine whether O2-sensing structures are  highly conserved among vertebrates while their locations, reflex roles and degree of plasticity have changed to more effectively regulate the respiratory and cardiovascular systems in animals with central cardiac shunts (amphibians and reptiles). My findings and those of others (Ishii et al., 1966; Ishii et al., 1985a; Ihsii et al., 1985b; Ishii and Ishii, 1986) indicate that amphibians and 177  reptiles have multiple chemosensory areas located in all extant derivatives of pharyngeal arches, as they are in fish. The presence of multiple chemosensory sites in animals with intra-cardiac shunts may be advantageous as the location determines the composition of blood to which these chemoreceptors are exposed. For instance, chemoreceptors in the pulmonary artery provide the central nervous system with information regarding oxygen consumption and oxygen acquisition (Van Vliet and West 1987b, 1987c). The pulmonary arteries of birds and mammals are not chemosensory, indicating that this function has been lost in more recently evolved vertebrates (Milsom and Burleson, 2007). A decrease in the number of chemosensory areas may allow for tighter control of arterial blood gases, which is important for animals with high metabolic rates such as endotherms. My thesis, together with findings in fish and mammals (reviewed by Jonz and Nurse, 2009, 2012) suggest that the innervation pattern, morphology and size of putative O2-sensing cells are maintained throughout the vertebrate taxa. However, there appears to be a trend towards greater complexity in the organization (solitary to clusters) of putative O2-sensing cells and an increase in the neurochemicals involved in O2 chemotransduction in the phylogenetic path from fish to mammals. The change, so far, seems to be from mainly solitary cells containing serotonin in fish, to a few neurotransmitters contained in different cells arranged singly or in clusters in anurans and reptiles (turtles and rattlesnakes), to clusters of glomus cells with multiple neurotransmitters in birds and mammals. It is possible that the formation of clusters and the increase in the neurochemical content allow for more precise regulation of chemoreceptor activity. It appears that all vertebrates share a basic structure in their O2 chemoreceptors (anatomical features and innervation) with modifications to suit the metabolic needs of the different organisms. For instance, the low solubility of O2 in aquatic environments poses a challenge to water-breathers. Thus, multiple oxygen sensing sites with internally and externally oriented NEC’s found in fish may enable them to sense oxygen levels in the blood as well as in their environment (Milsom and Burleson, 2007). The high metabolic rates and low hypoxia tolerance of birds and mammals relative to lower vertebrates require a tightly controlled system that prevents oscillations of arterial blood gases. This may be accomplished by a reduction in the number of chemosensory sites and complex organization of O2-sensing cells functioning as a chemosensory unit. Lastly, chemosensory areas in the carotid bifurcation, aorta and pulmonary artery may allow amphibians and reptiles to regulate arterial blood gases in the face of frequent 178  oscillations in arterial O2 due to an intermittent breathing pattern and the presence of intracardiac shunts (Wood, 1982, 1984; Shelton et al., 1986; Burggren, 1987; Milsom, 1988; Milsom, 1990; Smatresk, 1990; West and Van Vliet, 1992; Wang and Hicks, 1996b; Hicks, 1998; Hicks, 2002). Thus, the changes in O2-sensing structures seem to reflect the needs of each vertebrate class to match oxygen supply and demand.  179  Figure 6.1. Diagram of the central vasculature of frogs, turtles and snakes showing the location and distribution of putative O2-sensing cells. A: central vasculature of the frog. 5HT is shown in red, TH in green and cells where both neurotransmitters colocalized are shown in yellow. B: central vasculature of the turtle showing the location of VAChT cell clusters (putative O2sensing cells) in blue and 5HT-containing cells in red. Polygonal VAChT cells are shown in purple. C: central vasculature of the snake, showing VAChT cells in blue and 5HT-containing cells in red. cl, carotid labyrinth; Lao, left aorta; LPca, left pulmocutaneous artery; LPA, left pulmonary artery; PA, pulmonary artery; Lica, left internal carotid artery and Leca, left external carotid artery.  180  References Abdel-Magied, E.M. and A.S. King. 1978. The topographical anatomy and blood supply of the carotid body region of the domestic fowl. Journal of Anatomy 126: 535-546. Adams, W.E. 1958. The comparative morphology of the carotid body and carotid sinus. Charles C. Thomas, Springfield, Illinois, pp. 184-214. Adams, W.E. 1962. The carotid sinus-carotid body problem in the chelonian (with a note on a foramen of panizza in Dermochelys). Archives Internationales de Pharmacodynamie et de therapie 139: 28-37. Andersen, J.B., M.S. Hedrick and T. Wang. 2003. Cardiovascular responses to hypoxia and anaemia in the toad Bufo marinus. Journal of Experimental Biology 206: 857-865. Aschoff, J., U. Gerecke, Ch. von Goetz, G. A. Groos and F.W. Turek. 1982. Phase responses and characteristics of free-running activity rhythms in the golden hamster: independence of the pineal gland. In: J. Aschoff, S. Dann and G. Groos (Eds.). Vertebrate Circadian Systems, Berlin: Springer-Verlag, pp 129-140.. Ask-Upmark, E. 1935. The carotid sinus and the cerebral circulation. Lund (Ed.). Berlingska Boktryckeriet, pp 374. Bailly, Y., S. Dunel-Erb and P. Laurent. 1992. The neuroepithelial cells of the fish gill filaments: indolamine-immuno-cytochemistry and innervation. Anatomical Record 233: 143-161. Bailly, Y. 2009. Serotonergic neuroepithelial cells in fish gills: cytology and innervation. In: Zaccone, G., E. Cutz, D. Adriaensen, C.A. Nurse, A. Mauceri (Eds.). Airway Chemoreceptors in the Vertebrates: Structure, Evolution and Function. Science Publishers, Enfield, USA, pp 61-97. Banister, R.J. and S.P. Mann. 1965. An investigation of the adrenergic innervation of the heart and major blood vessels of the frog by Falck’s method of fluorescence microscopy. Journal of Physiology 181: 13-15. Banister, R.J., P.J. Portig and M. Vogt. 1967. The content and localization of catecholamines in the carotid labyrinths and aortic arches of Rana temporaria. Journal of Physiology 192: 529-535. Bartlett Jr., D. and G.F. Birchard. 1983. Effects of hypoxia on lung volume in the garter snake. Respiratory Physiology 53: 63-70. Benchetrit, G., J. Armand and P. Dejours. 1977. Ventilatory chemoreflex drive in the tortoise Testudo horsfieldi. Respiration Physiology 31: 183-191. Bennett, A. F. and W. R. Dawson. 1976. Metabolism. In Biology of the Reptilia, vol 5 (ed. C. Gans and W. R. Dawson), pp 127-223. New York: Academic Press. Berger, P.J. 1987. The reptilian baroreceptor and its role in cardiovascular control. American Zoologist 27: 111-120. 181  Berger, P.J., I.L. Gibbins, D.K. Hards and L.J. Crosby. 1982. The distribution and ultrastructure of sensory elements in the baroreceptor region of the truncus arteriosus of the lizard Trachydosaurus rugosus. Cell Tissue Research 226: 389-406. Bicego-Nahas, K. C. and L. G. S. Branco. 1999. Seasonal Changes in the cardiorespiratory responses to hypercarbia and temperature in the bullfrog, Rana catesbeiana. Comparative Biochemistry and Physiology Part A: Molecular and Integrative Physiology 124: 221229. Bicego-Nahas, K. C., L. H. Gargaglioni and L. G. S. Branco. 2001. Seasonal changes in the preferred body temperature, cardiovascular and respiratory responses to hypoxia in the toad, Bufo paracnemis. Journal of Experimental Zoology 289: 359-365. Boutilier, R.G. 1990. Control and co-ordination of gas exchange in bimodal breathers. In: Advances in Comparative and Environmental Physiology 6: Vertebrate Gas Exchange: From Environment to Cell, vol 6. Boutilier, R.G. (Ed.). Springer-Verlag, New York, pp 280-346. Boutilier, R.G. 2001. Mechanisms of cell survival in hypoxia and hypothermia. Journal of Experimental Biology 204: 3171-3181. Boutilier, R.G. and D.P. Toews. 1977. The effect of progressive hypoxia on respiration in the toad Bufo marinus. Journal of Experimental Biology 68: 99-107. Bowes, G., E.R. Townsend, L.F. Kozar, S.M. Bromley and E.A. Phillipson. 1981. Effect of carotid body denervation on arousal response in sleeping dogs. Journal of Applied Physiology 51: 40-45. Branco, L.G.S. and M.L. Glass. 1995. Ventilatory responses to carboxyhaemoglobinaemia and hypoxic hypoxia in Bufo paracnemis. Journal of Experimental Biology 198: 1417-1421. Bronner-Fraser, M. 1986. Analysis of the early stages of trunk neural crest migration in avian embryos using monoclonal antibody HNK-1. Developmental Biology 115: 44-55. Bunn, H.F. and R.O. Poyton. 1996. Oxygen sensing and molecular adaptation to hypoxia. Physiological Reviews 76 (3): 839-885. Burggren, W.W. 1977. The pulmonary circulation of the chelonian reptile: morphology, haemodynamics and pharmacology. Journal of Comparative Physiology B 116: 303-323. Burggren, W.W. 1987. Form and function in reptilian circulations. American Zoologist 27 (1): 5-19. Burggren, W.W. 1988. Cardiovascular responses to diving and their relation to lung and blood oxygen stores in vertebrates. Canadian Journal of Zoology 66: 20-28. Burggren, W.W., M.L. Glass and K. Johansen. 1977. Pulmonary ventilation: perfusion relationships in terrestrial and aquatic chelonian reptiles. Canadian Journal of Zoology 55: 2024-2034. Burggren, W.W. and G. Shelton. 1979. Gas exchange and transport during intermittent breathing in chelonian reptiles. Journal of Experimental Biology 82: 75-92. 182  Burggren, W.W. and K. Johansen. 1982. Ventricular hemodynamics in the monitor lizard Varanus exanthematicus: pulmonary and systemic pressure separation. Journal of Experimental Biology 96: 343-354. Burggren, W.W., A. Smits and B. Evans. 1989. Arterial O2 homeostasis during diving in the turtle Chelodina longicollis. Physiological Zoology 62: 668-686. Burleson, M.L. and W.K. Milsom. 1993. Sensory receptors in the first gill arch of rainbow trout. Respiration Physiology 93: 97-110. Burleson, M.L. and W.K. Milsom. 1995. Cardio-ventilatory control in rainbow trout: I. Pharmacology of branchial, oxygen-sensitive chemoreceptors. Respiration Physiology 100: 231-238. Burleson, M.L., S.E. Mercer and M.A. Wilk-Blaszczak. 2006. Isolation and characterization of putative O2 chemoreceptor cells from the gills of channel catfish (Ictalurus punctatus). Brain Research 1092:100-107. Cagle, F. R. 1950. The life history of the slider turtle Pseudemys scripta troostii (Holbrook). Ecological Monographs 20: 32-54. Campanucci, V.A. and C.A. Nurse. 2007. Autonomic innervation of the carotid body: role in efferent inhibition . Respiratory Physiology and Neurobiology 157: 83-92. Chandel, N.S. and P.T. Schumacker. 2000. Cellular oxygen sensing by mitochondria: old questions, new insight. Journal of Applied Physiology 88: 1880-1889. Chiba, T. and A. Yamauchi. 1973. Fluorescence and electron microscopy of the monoaminecontaining cells in the turtle heart. Cell and tissue research 140: 26-37. Clark, K., G. Vender, B.P. Murray, K. Panfilio, S. Cook, R. Davis, K. Murnen, R.S. Tuan and S.F. Gilbert. 2001. Evidence for the neural crest origin of turtle plastron bones. Genesis 31: 111-117. Coleridge, H.M., J. Coleridge and A. Howe. 1967. A search for pulmonary arterial chemoreceptors in the cat, with a comparison of the blood supply of the aortic bodies in the newborn and adult animal. Journal of Physiology 191: 353-374. Comeau, S.G. and J.W. Hicks. 1994. Regulation of central vascular blood flow in the turtle. American Journal of Physiology 267: 569-578. Comroe, J.H. 1964. The peripheral chemoreceptors. In: Handbook of Physiology, Section 3: Respiration Volume 1, Chapter 23. American Phsysiological Society, Washington, DC. Coolidge, E.H., C.S. Ciuhandu and W.K. Milsom. 2008. A comparative analysis of putative oxygen-sensing cells in the fish gill. Journal of Experimental Biology 211: 1231–1242. Corrodi, H. and G. Jonsson. 1967. The formaldehyde fluorescence method for the histochemical demonstration of biogenic monoamines. The Journal of Histochemistry and Cytochemistry 15 (2): 65-78.  183  Cragg, P.A. 1978. Ventilatory patterns and variables in rest and activity in the lizard, Lacerta. Comparative Biochemistry and Physiology Part A 60: 399-410. Cutz, E. and A. Jackson. 1999. Neuroepithelial bodies as airway oxygen sensors. Respiration Physiology 115: 201-214. de Graaf, P.J.F. 1990. Innervation pattern of the gill arches and gills of the carp (Cyprinus carpio). Journal of Morphology 206: 71-78. Dejours, P. 1981. Principles of Comparative Respiratory Physiology. Elsevier/North-Holland, 2nd ed., Amsterdam. Donald, J.A. and J.B. Lillywhite. 1989. Adrenergic nerves and 5-hydroxytryptaminecontaining cells in the pulmonary vasculature of the aquatic file snake Acrochordus granulatus. Cell and Tissue Research 256: 113-118. Donnelly, P. and A.J. Woolcock. 1977. Ventilation and gas exchange in the carpet python, Morelia spilotes variegate. Journal of Comparative Physiology. 122: 403-418. Dunel-Erb, S., Y. Bailly and P. Laurent. 1982. Neuroepithelial cells in fish gill primary lamellae. Journal of Applied Physiology 53 (6): 1342-1353. Duvall, D., L.J. Jr. Guillette. and R. E. Jones. 1982. Environmental control of reptilian reproductive cycles. In Biology of the Reptilia, vol 13D (ed. C. Gans and F. H. Pough), pp 201-231. London and New York: Academic Press. Ernst. C. H. and R. W. Barbour. 1989. Turtles of the World. Smith Inst. Press. Evans, D.H., P.M. Piermarini and K.P. Choe. 2005. The multifunctional fish gill: dominant site of gas exchange, osmoregulation, acid-base regulation, and excretion of nitrogenous waste. Physiological Reviews 85: 97-177. Eyzaguirre, C., R.S. Fitzgerald, S. Lahiri and P. Zapata. 1983. Arterial Chemoreceptors. In: Shepherd and Abboud, F.M. (Eds.). Handbook of Physiology. The cardiovascular system III (Part 2). American Physiological Society, Bethesda, pp 557-621. Eyzaguirre, C. and P. Zapata. 1984. Perspectives in carotid body research. Journal of Applied Physiology 57: 931-957. Farrell, A.P., A.K. Gamperl and E.T.B. Francis. 1998. Comparative aspects of heart morphology. In: C. Gans and A.S. Gaunt (Eds.). Biology of Reptilia Vol. 19: Morphology G: Visceral organs. SSAR Press, Ithaca, New York, pp 375-424. Fitzgerald, R.S. and S. Lahiri. 1986. Reflex responses to carotid/aortic body simtulation. In: Cherniack, N.S., Widdicombe, J.G. (Eds.). Handbook of Physiology, Section 3: The respiratory system, vol. 2: Control of Breathing, part 1. American Physiological Society, Washington, DC, pp. 313-362. Flanigan, W.F. 1973. Sleep and wakefulness in iguanid lizards, Ctenosaura pectinata and Iguana iguana. Brain Behavior and Evolution 8: 401-436.  184  Flanigan, W.F., C.P. Knight, K.M. Hartse and A. Rechtschaffen. 1974. Sleep and wakefulness in chelonian reptiles. I. The box turtle, Terrapene carolina. Archives Italiennes de Biologie 112 (3): 227-252. Frankel, H.M., A. Spitzer, J. Blaine and E.P. Schoener. 1969. Respiratory response of turtles (Pseudemys scripta) to changes in arterial blood gas composition. Comparative Biochemical and Physiology 31: 535-546. Funk, G.D., C.L. Webb, and W.K. Milsom. 1986. Non-invasive measurement of respiratory tidal volume in aquatic, air-breathing animals. Journal of Experimental Biology 126: 519523. Funk, G. D. and W.K. Milsom. 1987. Changes in ventilation and breathing pattern produced by changing body temperature and inspired CO2 concentration in turtles. Respiratory Physiology 67: 37-51. Galli, G.L.J., N. Skovgaard, A.S. Abe, E.W. Taylor and T. Wang. 2005. The role of nitric oxide in the regulation of the systemic and pulmonary vasculature of the rattlesnake Crotalus durissus terrificuls. Journal of Comparative Physiology B 175: 201-208. Galli, G.L.J., N. Skovgaard, A.S. Abe, E.W. Taylor and T. Wang. 2007. The adrenergic regulation of the cardiovascular system in the South American rattlesnake, Crotalus durissus. Comparative Biochemistry and Physiology A 148: 510-520. Gargaglioni, L.H. and L.G.S. Branco. 2004. Nucleus isthmi and control of breathing in amphibians. Respiratory Physiology and Neurobiology 143: 177-186. Glass, M.L. 1992. Ventilatory responses to hypoxia in ectothermic vertebrates. In: C. Lenfant (Ed.). Physiological adaptations in vertebrates: Respiration, circulation and metabolism. Marcel Dekker, pp 97-118. Glass, M.L., J.W. Hicks, and M.L. Riedesel. 1979. Respiratory responses to long-term temperature exposure in the box turtle, Terrapene ornate. Journal of Comparative Physiology 131: 353-359. Glass, M.L., R.G. Boutilier and N.Heisler. 1983. Ventilatory control of arterial PO2 in the turtle Chrysemys picta bellii: effects of temperature and hypoxia. Journal of Comparative Physiology 151: 145-153. Glass, M. L., R. G. Boutilier and N. Heisler. 1985. Effects of body temperature on respiration blood gases and acid-base status in the turtle Chrysemis picta bellii. Journal of Experimental Biology 114: 37-51. Gratz, R.K. 1979. Ventilatory response of the diamondback water snake, Natrix rhombifera to hypoxia, hypercapnia and increased oxygen demand. Journal of Comparative Physiology 129: 105-110. Gregory, P. T. 1982. Reptilian hibernation. In Biology of the Reptilia, vol 13 (ed. C. Gans and F.H. Pough), pp 53-154. New York: Academic Press. Gonzalez, C., L. Almaraz, A. Obeso and R. Rigual. 1994. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiological Reviews 74 (4): 829-898. 185  Gonzlaez, C., M.T. Agapito, A. Rocher, M.C. Gonzalez-Martin, V. Vega-Agapito, A.Gomez-Nino, R. Rigual, J. Castaneda, A. Obeso. 2007. Chemoreception in the context of the general biology of ROS. Respiratory Physiology and Neurobiology 157: 30-44. Harles, M. and H. Morlock. 1978. Turtles perspective and research. Wiley-Interscience Publication. New York. Heisler, N., G. Forcht, G.R. Ultsch, J.F. Anderson. 1982. Acid–base regulation in response to environmental hypercapnia in two aquatic salamanders, Siren lacertina and Amphiuma means. Respiration Physiology 49: 141–158. Hicks, J.W. 1994. Adrenergic and cholinergic regulation of intracardiac shunting. Physiological Zoology 67 (6): 1325-1346. Hicks, J.W. 1998. Cardiac shunting in reptiles: mechanisms, regulation, and physiological functions. In: C. Gans and A.S. Gaunt (Eds.). Biology of Reptilia, Morphology G, Visceral organs. SSAR Press, Ithaca, NY, pp 425-483. Hicks, J.W. 2002. The physiological and evolutionary significance of cardiovascular shunting patterns in reptiles. News in Physiological Science 17: 241-245. Hicks, J.W. and M.L. Riedesel. 1983. Diurnal ventilatory patterns in the garter snake, Thamnophis elegans. Journal of Comparative Physiology 149: 503-510. Hicks, J.W. and S.C. Wood. 1989. Oxygen homeostasis in lower vertebrates: the impact of external and internal hypoxia. In: S.C. Wood (Ed.). Comparative pulmonary physiology: current concepts. Marcel Dekker, pp 311-341. Hicks, J.W. and S.G. Comeau. 1994. Vagal regulation of intracardiac shunting in the turtle, Pseudemys scripta. Journal of Experimental Biology 186: 109-126. Hicks, J.W. and G.M. Malvin. 1995. Mechanisms of intracardiac shunting in reptiles: Pressure vs washout shunting. In: N. Heisler (Eds.). Comparative and Environmental Physiology. Mechanisms of Systemic Regulation: Respiration and Circulation. Springer Verlag, Berlin, Heidelberg, New York, pp 137-157. Hoffmann, A. and M.B. Cordeiro de Souza. 1982. Cardiovascular reflexes in conscious toads. Journal of the Autonomic Nervous System 5: 345-355. Hou, L. and T. Takeuchi. 1994. Neural crest development in reptilian embryos, studied with monoclonal antibody, HNK-1. Zoological Science 11:423-431. Hunter, J.D., J.Z. McLeod, and W.K. Milsom. 1998. Cortical activation states in sleep and anethesia. II: Respiratory reflexes. Respiratory Physiology 112: 83-94. Ishii, K., K. Honda and K. Ishii. 1966. The function of the carotid labyrinth in the toad. Tohoku Journal of Experimental Medicine 88: 103-116. Ishii, K. and K. Ishii. 1967. Adrenergic transmission of the carotid chemoreceptor impulses in the toad. Tohoku Journal of Experimental Medicine 91: 119-128. 186  Ishii, K. and T. Oosaki. 1969. Fine structure of the chemoreceptor cell in the amphibian carotid labyrinth. Journal of Anatomy 104 (2): 263-280. Ishii, K. and K. Ishii. 1973. Fiber composition and derivation of afferent and efferent nerve fibers in the carotid nerve innervating the carotid labyrinth of the toad. Tohoku Journal of Experimental Medicine 109: 323-337. Ishii, K. and T. Kusakabe. 1982. The glomus cell of the carotid labyrinth of Xenopus laevis. Cell Tissue Research 224: 459-463. Ishii, K., K. Ishii and T. Kusakabe. 1985a. Chemo- and baroreceptor innervation of the aortic trunk of the toad Bufo vulgaris. Respiration Physiology 60: 365-375. Ishii, K. K. Ishii and T. Kusakabe. 1985b. Electrophysiological aspects of reflexogenic area in the chelonian, Geoclemmys reevesii. Respiration Pysiology 59: 45-54. Ishii, K. and K. Ishii. 1986. Glossopharyngeal innervation of chemo- and baroreceptors in the dorsal carotid artery of the tortoise (Testudo hermanni). Respiration Physiology 65: 295302. Ito, S., T. Ohga and Y. Nakazato. 1997. Release of 5-hydroxytryptamine by hypoxia from epithelioid cells of chicken thoracic aorta. British Journal of Pharmacology 122: 799-801. Jackson, D.C. 1973. Respiratory response to hypoxia in turtles at various temperatures. Respiration Physiology 18: 178-187. Jackson, D.C. 1978. Respiratory control in air-breathing ectotherms. In: D.G. Davies and C.D. Barnes (Eds.). Regulation of ventilation and gas exchange. New York: Academic, pp 93130. Jackson, D. C, S. E. Palmer and W. L. Meadow. 1974. The effects of temperature and carbon dioxide breathing on ventilation and acid-base status of turtles. Respiratory Physiology 20: 131-146. Jarsky, T. M. and R. Stephenson. 2000. Effects of hypoxia and hypercapnia on circadian rhythms in the golden hamster (Mesocricetus auratus). Journal of Applied Physiology 89: 2130-2138. Johnson, S.M., R.A. Johnson and G.S. Mitchell. 1998. Hypoxia, temperature and pH/CO2 effects on respiratory discharge from a turtles brain stem preparation. Journal of Applied Physiology 84(2): 649-660. Jones, D.R. and M.J. Purves. 1970a. The effect of carotid body denervation upon the respiratory response to hypoxia and hypercapnia in the duck. Journal of Physiology 211: 295-310. Jones, D.R. and M.J. Purves. 1970b. The carotid body in the duck and the consequence of its denervation upon the cardiac responses to immersion. Journal of Physiology 211: 279294.  187  Jones, D.R. and W.K. Milsom. 1982. Peripheral receptors affecting breathing and cardiovascular function in non-mammalian vertebrates. Journal of Experimental Biology 100: 59-91. Jones, J.F.X. and M. de Burgh Daly. 1997. Reflex cardiac dromotropic responses to stimulation of the carotid and aortic chemoreceptors in the anaesthetized cat. Journal of Physiology 502: 461-467. Jonz, M.G. and C. Nurse. 2003. Neuroepithelial cells and associated innervation of the zebrafish gill: A confocal immunofluorescence study. The Journal of Comparative Neurology 461: 1-17. Jonz, M.G., I.M. Fearon and C. Nurse. 2004. Neuroepithelial oxygen chemoreceptors of the zebrafish gill. Journal of Physiology 560 (3): 737-752. Jonz, M.G. and C.A. Nurse. 2009. Oxygen-sensitive Neuroepithelial cells in the gills of aquatic vertebrates. In: Zaccone, G., E. Cutz, D. Adriaensen, C.A. Nurse, A. Mauceri (Eds.). Airway Chemoreceptors in the Vertebrates: Structure, Evolution and Function. Science Publishers, Enfield, USA, pp 1-30. Jonz, M.G. and C.A. Nurse. 2012. Peripheral chemoreceptors in air- versus water- breathers. Advances in Experimental Medicine and Biology 758: 19-27. Kameda, Y. 2002. Carotid body and glomus cells distributed in the wall of the common carotid artery in the bird. Microscopy Research and Technique 59: 196-206. Kardong, K.V. 2006. Vertebrates: Comparative anatomy, function, evolution. McGraw-Hill Higher Education, 4ht edition, Boston, USA, pp 455-461 Kemp, P.J. 2006. Detecting acute changes in oxygen: will the real sensor please stand up? Experimental Physiology 91 (5): 829-834. Kenagy, G. J. and D. Vleck. 1982. Daily temporal organization of metabolism in small mammals: adaptation and diversity. In Vertebrate Circadian Systems (ed. J. Aschoff, S. Dann and G. Groos), pp 322-338. Berlin: Springer-Verlag. King, A.S., D.Z. King, R.D. Hodges and J. Henry. 1975. Synaptic morphology of the carotid body of the domestic fowl. Cell and Tissue Research 162: 459-473. Kinney, J.L. and F.N. White. 1977. Oxidative cost of ventilation in a turtle, Pseudemys floridana. Respiratory Physiology 31: 327-332. Kobayashi, S. 1971a. Comparative cytological studies of the carotid body 1. Demonstration of monoamine-storing cells by correlated chromaffin reaction and fluorescence histochemistry. Archivum Histologicum Japonicum 33 (4): 319-339. Kobayashi, S. 1971b. Comparative cytological studies of the carotid body 2. Ultrastrucutre of the synapses on the chief cell. Archivum Histologicum Japonicum 33 (5): 397-420. Kobbert, C., R. Apps, I. Bechmann, J.L.Lanciego, J. Mey and S. Thanos. 2000. Current concepts in neuroanatomical tracing. Progress in Neurobioloy 62: 327-351. 188  Kollmeyer, K.R. and L.I. Kleinman. 1975. A respiratory venous chemorecptor in the young puppy. Journal of Applied Physiology 38: 819-826. Kondo, H. 1976. Innervation of the carotid body of the adult rat. Cellular and Tissue Research 173: 1-15. Kondo, H., T. Iwanaga and T. Nikajima. 1982. Immunocytochemical study on the localization of neuron-specific enolase and S-100 protein in the carotid body of rats. Cell and Tissue Research 227: 291-295. Krahl, V.E. 1962. The glomus pulmonale: its location and microscopic anatom. In: de Reuck AVS and M. O’Connor (Eds). Pulmonary structure and function. Ciba Foundation Symposium, London, UK. Krosniunas, E.H. and J.W. Hicks. 2003. Cardiac output and shunt during voluntary activity at different temperatures in the turtle, Trachemys scripta. Physiological and Biochemical Zoology 76 (5): 679-694. Kuchling, G. 1999. Reproductive biology of the chelonian. New York: Springer Berlin. Kundrat, M. 2008. HNK-1 immunoreactivity during early morphogenesis of the head region in a nonmodel vertebrate, crocodile embryo. Naturwissenschaften 95: 1063-1072. Kusakabe, T. 1990. Ultrastructural studies of the carotid labyrinth in the newt Cynops pyrrhogaster. Zoological Science 7: 201-208. Kusakabe, T. 1991. Morphogenesis of the carotid labyrinth in the bullfrog, Rana catesbeiana, during larval development and metamorphosis. Anatomy and Embryology 184: 133-139. Kusakabe, T. 2002. Carotid labyrinth of amphibians. Microscopy Research and Technique 59: 207-226. Kusakabe, T. 2009. The amphibian carotid labyrinth. In: Zaccone, G., E. Cutz, D. Adriaensen, C.A. Nurse, A. Mauceri (Eds.). Airway Chemoreceptors in the Vertebrates: Structure, Evolution and Function. Science Publishers, Enfield, USA, pp 125-139. Kusakabe, T., K. Ishii and K. Ishii. 1987. A possible role of the glomus cell in controlling vascular tone of the carotid labyrinth of Xenopus laevis. Tohoku Journal of Experiemental Medicine 151: 395-408. Kusakabe, T., K. Ishii and K. Ishii. 1988. Dense granule-containing cells in arterial chemoreceptor areas of the tortoise (Testudo hermanni). Journal of Morphology 197: 183-191. Kusakabe, T., P. Anglade and S. Tsuji. 1991. Localization of substance P, CGRP, VIP, neuropeptide Y, and somatostatin immunoreactive nerve fibers in the carotid labyrinths of some amphibian species. Histochemistry 96: 255-260. Kusakabe, T., T. Kawakami and T. Takenaka. 1994. Coexistence of substance P, neuropeptide Y, VIP and CGRP in the nerve fibers of the carotid labyrinth of the bullfrog, Rana catesbeiana: a double-labelling immunofluorescence study in combination with alternate consecutive sections. Cell Tissue Research 276: 91-97. 189  Lahiri, S. 2000. Historical perspectives of cellular oxygen sensing and responses to hypoxia. Journal of Applied Physiology 88: 1467-1473. Lahiri, S., A. Mokashi, E. Mulligan and T. Nishino. 1981. Comparison of aortic and carotid chemoreceptor responses to hypercapnia and hypoxia. Journal of Applied Physiology 51 (1): 55-61. Lahiri, S., N.J. Smatresk and E. Mulligan. 1983. Repsonses of peripheral chemoreceptors to natural stimuli. In: H. Acker and R.G. O’Regan (Eds.). Physiology of the Peripheral Arterial Chemoreceptors. Elsevier, Amsterdam, pp 221-256. Leite, C.A.C., L.H. Florindo, A.L. Kalinin, W.K. Milsom and F.T. Rantin. 2007. Gill chemoreceptors and cardio-respiratory reflexes in the neotropical teleost pacu, Piaractus mesopotamicus. Journal of Comparative Physiology A 193: 1001-1011. Lenfant, C., K. Johansen, J.A. Petersen and K. Schmidt-Nielsen. 1970. Respiration in the fresh water turtle, Chelys fimbriata. Respiration Physiology 8: 261-275. Lillo, R.S. 1980. Localization of chemoreceptors which may cause diving bradycardia in bullfrogs. Canadian Journal of Zoology 58: 931-936. Lopez-Barneo, J., R. Pardal and P. Ortega-Saenz. 2001. Cellular mechanisms of oxygen sensing. Annual Review of Physiology 63: 259-287. Marshall, J.M. 1994. Peripheral chemoreceptors and cardiovascular regulation. Physiological Reviews 74: 543-594. Mayhew, W. W. 1965. Hibernation in the horned lizard, Phrynosoma m’calli. Comparative Biochemistry and Physiology 16: 103-119. McArthur, M.D. and W.K. Milsom. 1991. Changes in ventilation and respiratory sensitivity associated with hibernation in Columbian (Spermophilus columbianus) and Goldenmantled (Spermophilus lateralis) ground squirrels. Physiological Zoology 64 (4): 940959. Milsom, W.K. 1988. Control of arrhythmic breathing in aerial breathers. Canadian Journal of Zoology 66 (1): 99-108. Milsom, W.K. 1990. Control and Co-ordination of gas exchange in air breathers. In: R.G. Boutilier (Ed.). Advances in Comparative and Environmental Physiology, Vol. 6. Springer-Verlag, Berlin, pp 347-400. Milsom, W.K. 1991. Intermittent breathing in vertebrates. Annual Review of Physiology 53: 87105. Milsom, W.K. 1998. Phylogeny of respiratory chemoreceptor function in vertebrates. Zoology 101:316-332. Milsom, W.K. 2012. New insights into gill chemoreception: receptor distribution and roles in water and air breathing fish. Respiratory Physiology and Neurobiology 184: 326-339.  190  Milsom, W.K., B.L. Langille and D.R. Jones. 1977. Vagal lcontrol of pulmonary vascular resitance int he turtle Chrysemys scripta. Canadian Journal of Zoology 55: 359-367. Milsom, W. K., S. Osborne, P.F. Chan, J.D. Hunter and J.Z. MacLeod. 1993. Sleep, hypothermia, and hibernation: metabolic rate and the control of breathing pattern in golden-mantled ground squirrels. In: C. Carey and G.L. Florant (Eds.). Life in the cold. Ecological, Physiological and Molecular mechanisms. Westview Press, Inc., San Francisco: Milsom, W.K. and M.L. Burleson. 2007. Peripheral arterial chemoreceptors and evolution of the carotid body. Respiratory Physiology and Neurobiology 157 (1): 4-11. Milsom, W.K., D.V Andrade, S.P. Brito, L.F. Toledo, T. Wang and A.S. Abe. 2008. Seasonal changes in daily metabolic patterns of tegu lizards (Tupinambis merianae) placed in the cold (17°C) and dark. Physiological and Biochemical Zoology 81(2): 165-175. Miyoshi, M., Y. Hashimoto, Y. Kon and M. Sugimura. 1995. Distribution of epithelioid cells in the wall of the chicken aorta and their functional role as chemoreceptors. Anatomical Record 242: 302-309. Mortola, J. P. 2004. Breathing around the clock: an overview of the circadian pattern of respiration. European Journal of Applied Physiology 91: 119-129. Mortola, J. P. and E. L. Seifert. 2000. Hypoxic depression of circadian rhythms in adult rats. Journal of Applied Physiology 88: 365-368. Mortola, J. P. and E. L. Seifert. 2002. Circadian patterns of breathing. Respiratory Physiology and Neurobioliogy 131: 91-100. Nurse, C.A. 2005. Neurotransmission and neuromoldulation in the chemosensory carotid body. Autonomic Neuroscience: Basic and Clinical 120: 1-9. Nurse, C.A. 2010. Neurotransmitter and neuromodulatory mechanisms at peripheral arterial chemoreceptors. Experimental Physiology 95: 657-667. Nye, P.C.G. and F.L. Powell. 1984. Steady-state discharge and bursting of arterial chemoreceptors in the duck. Respiratory Physiology 56: 369-384. Olson, K.R., R.A. Dombkowski, M.J. Russell, M.M. Doellman, S.K. Head, N.L. Whitfield and J.A. Madden. 2006. Hydrogen sulfide as an oxygen sensor/transducer in vertebrate hypoxic vasoconstriction and hypoxic vasodilation. Journal of Experimental Biology 209: 4011-4023. Overgaard, J., J.A.W. Stecyk, A.P. Farrell and T. Wang. 2002. Adrenergic control of the cardiovascular system in the turtle Trachemys scripta. Journal of Experimental Biology 205: 3335-3345. Pearse, A.G.E. 1969. The cytochemistry and ultrastructure of polypeptide hormone-producing cells of the APUD series and the embryologic, physiologic and pathologic implications of the concept. The Journal of Histochemistry and Cytochemistry 17 (5): 303-313.  191  Pearse, A.G.E., J.M. Polak, F.W. D. Rost, J. Fontaine, C. Le Lievre and N. Le Douarin. 1973. Demosntration of the neural crest origin of type I (APUD) cells in the avian carotid body, using a cytochemical marker system. Histochemistry and Cell Biology 34: 191203. Peers, C. and P.J. Kemp. 2001. Acute oxygen sensing: diverse but convergent mechanisms in airway and arterial chemoreceptors. Respiratory Research 2 (3): 145-149. Peever, J. H. and R. Stephenson. 1997. Day-night differences in the respiratory response to hypercapnia in awake adult rats. Respiratory Physiology 109: 241-248. Phillipson, E.A.,C.E. Sullivan, D.J.C. Read, E. Murphy and L.F. Kozar. 1978. Ventilatory and waking responses to hypoxia in sleeping dogs. Journal of Applied Physiology 44: 512-520. Piskuric, N.A., C. Vollmer and C.A. Nurse. 2011. Confocal immunofluorescence study of rat aortic body chemoreceptors and associated neurons in situ and in vitro. Journal of Comparative Neurology 519: 856-873. Piskuric, N.A. and C.A. Nurse. 2012. Effects of chemostimuli on [Ca2+]I responses of rat aortic body type cells and endogenous local neurons: Comparison with carotid body cells. Journal of Physiology 590: 2121-2135. Piskuric, N.A. and C.A. Nurse. 2013. Expanding role of ATP as a versatile messenger at carotid and aortic body chemoreceptors. Journal of Physiology 591 (2): 415-422. Prabhakar, N.R. 1999. NO and CO as second messengers in oxygen sensing in the carotid body. Respiration Physiology 115: 161-168. Prabhakar, N.R. 2000. Oxygen sensing by the carotid body chemoreceptors. Journal of Applied Physiology 88: 2287-2295. Prabhakar, N.R. 2006. O2 sensing at the mammalian carotid body: why multiple O2 sensors and multiple transmitters? Experimental Physiology 91 (1): 17-23. Porteus, C., M.S. Hendrick, J.W. Hicks, T. Wang and W.K. Milsom. 2011. Time domains of the hypoxic ventilatory response in ectothermic vertebrates. Journal of Comparative Physiology B 181: 311-333. Porteus, C.S., D.L. Brink and W.K. Milsom. 2012. Neurotransmitter profiles in fish gills: Putative gill oxygen chemoreceptors. Respiratory Physiology and Neurobiology 184: 316-325. R Development Core Team. 2010. R: a language and environment for statisctical computing. R Foundation for Statistical Computing, Vienna, Austria. Retrieved Nov 10, 2013 from http://www.R-project.org Randall, D.J. 1982. The control of respiration and circulation in fish during exercise and hypoxia. Journal of Experimental Biology 100: 275-288. Reese, A.M. 1914. The vascular system of the Florida alligator. Proceedings of the Academy of Natural Sciences of Philadelphia 66, 2: 413-425. 192  Regan, K.S., M.G. Jonz, P.A. Wright. 2011. Neuroepithelial cells and the hypoxia emersion response in the amphibious fish Kryptolebias marmoratus. Journal of Experimental Biology 214: 2560-2568. Reyes, C. and W.K. Milsom. 2010. Circadian and circannual rhythms in the metabolism and ventilation of red-eared sliders (Trachemys scripta elegans). Physiological and Biochemical Zoology 83 (2): 293-298. Reyes, M., K. Zandberg, I. Desmawati and M.E. de Bellard. 2010. Emergence and migration of trunk neural crest cells in a snake, the California Kingsnake (Lampropeltis getula californiae). BMC Developmental Biology 10:52. Rismiller, P.D. and G. Heldmaier. 1982. The effect of photoperiod on temperature selection in the European Green Lizard, Lacerta viridis. Oecologia 53: 222-226. Rismiller, P. D. and G. Heldmaier. 1991. Seasonal changes in daily metabolic patterns of Lacerta viridis. Journal of Comparative Physiology B 161: 482-488. Rocha, P. and L. G. S. Branco. 1998. Seasonal changes in the cardiovascular, respiratory and metabolic responses to temperature and hypoxia in the bullfrog Rana catesbeiana. Journal of Experimental Biology 201: 761-768. Rogers, D.C. 1963. Distinctive cell types in the amphibian carotid labyrinth. Nature 200: 492493. Rogers, D.C. 1967. The structure of the carotid bifurcation in the lizards Tiliqua occipitalis and Trachysaurus rugosus. Journal of Morphology 122: 115-130. Saltys, H.A., M.G. Jonz and C.A. Nurse. 2006. Comparative study of gill neuroepithelial cells and their innervation in teleosts and Xenopus tadpoles. Cell and tissue research 323: 1-10. Seifert, E. L. and J. P. Mortola. 2002a. Circadian pattern of ventilation during acute and chronic hypercapnia in conscious adult rats. American Journal of Physiology-Regulatory, Integrative and Comparative 282: R244-R251. Seifert, E. L. and J. P. Mortola. 2002b. Circadian pattern of ventilation during prolonged hypoxia in conscious rats. Respiratory Physiology and Neurobiology 133: 23-34. Shakarchi, K., P.C. Zachar and M.G. Jonz. 2012. Serotonergic and cholinergic elements of the hypoxic ventilatory response in developing zebrafish. Journal of Experimental Biology, in press. Shelton, G.D. and W.W. Burggren. 1976. Cardiovascular dynamics of the chelonian during apnoea and lung ventilation. Journal of Experimental Biology 64: 323-343. Shelton, G.D. and R.G. Boutilier. 1982. Apnoea in amphibians and reptiles. Journal of Experimental Biology 100: 245-273. Shelton, G., D.R. Jones and W.K. Milsom. 1986. Control of breathing in ectothermic vertebrates. In: Fishman AP., N.S. Cherniak, J.G. Widdicombe, S.R. Geiger (Eds.). Handbook of physiology, section 3. The respiratory system, vol II. Control of breathing, part 1. American Physiological Society, Bethesda. 193  Shirahata, M., A. Balbir, T. Otsubo, R.S. Fitzgerald. 2007. Role of acetylcholine in neurotransmissionof the carotid body. Respiratory Physiology and Neurobiology 157: 93–105. Skals, M., N. Skovgaard, A.S. Abe and T. Wang. 2005. Venous tone and cardiac function in the South American rattlesnake Crotalus durissus: mean circulatory filling pressure during adrenergic stimulation in anaesthetised and fully recovered animals. Journal of Experimental Biology 208: 3747-3759. Skovgaard, N. and T. Wang. 2004. Cost of ventilation and effect of digestive state on the ventilatory response of the tegu lizard. Respiratory Physiology and Neurobiology 141: 85-97. Smatresk, N.J. 1990. Chemoreceptor modulation of endogenous respiratory rhythms in vertebrates. American Journal of Physiology 28: R887-R897. Smith, F.M. and D.R. Jones. 1982. The effect of changes in blood oxygen carrying capacity on ventilation volume in the rainbow trout (Salmo gairdneri). Journal of Experimental Biology 97: 325-334. Stephenson, R., R. M. Mohan, J. Duffin. and T.M. Jarsky. 2000. Circadian rhythms in the chemoreflex control of breathing. American Journal of Physiology-Regulatory, Integrative and Comparative 278: R282-286. Stuesse, S.L. W.L.R. Cruce and K.S. Powell. 1984. Organisation within the cranial IX-X complex in ranid frogs: a horseradish peroxidase transport study. Journal of Comparative Neurology 222: 358-365. Sundin, L., S.G. Reid, A.L. Kalinin, F.T. Rantin and W.K. Milsom. 1999. Cardiovascular and respiratory reflexes: the tropical fish, traira (Hoplias malabaricus) O2 chemoresponses. Respiration Physiology 116: 181-199. Sundin, L., S.G. Reid, F.T. Rantin and W.K. Milsom. 2000. Branchial receptors and cardiorespiratory reflexes in a neotropical fish, the tambaqui (Colossoma macropomum). Journal of Experimental Biology 203: 1225-1239. Tatsumi, K., C. K. Pickett, C. R. Jacoby, J.V. Weil and L.G. Moore. 1997. Role of endogenous female hormones in hypoxic chemosensitivity. Journal of Applied Physiology 83 (5): 1706-1710. Taylor, E.W., D. Jordan and J.H. Coote. 1999. Central control of the cardiovascular and respiratory systems and their interactions in vertebrates. Physiological reviews 79: 855916. Taylor, E.W., D.V. Andrade, A.S. Abe, C.A.C. Leite and T. Wang. 2009. The unequal influences of the left and right vagi on the control of the heart and pulmonary artery in the rattlesnake, Crotalus durissus. Journal of Experimental Biology 212: 145-151. Toews, D., G. Shelton and R. Boutilier. 1982. The amphibian carotid labyrinth: some anatomical and physiological relationships. Canadian Journal of Zoology 60:1153-1160. 194  Tosini, G., C. Bertolucci and A. Foa. 2001. The circadian system of reptiles: a multioscillatory and multiphotoreceptive system. Physiology and Behavior 72: 461-471. Tschorn, R.R. and M.R. Fedde. 1974. Effects of carbon monoxide on avian intrapulmonary carbon dioxide sensitive receptors. Respiration Physiology 20: 303-311. Ultsch, G.R. 1973. A theoretical and experimental investigation of the relationships between metabolic rate, body size, and oxygen exchange capacity. Respiration Physiology 18: 143–160. Ultsch, G. R. 1989. Ecology and physiology of hibernation and overwintering among freshwater fishes, turtles and snakes. Biological Reviews 64: 435-516. Underwood, H. 1992. Endogenous rhythms. In: Gans, C. and D. Crews (Eds.). Biology of the Reptilia: hormones, brain and behavior, vol 18. University of Chicago Press. Chicago, pp 229-297. Van Vliet, B.N. and N.H. West. 1986. Cardiovascular responses to electrical stimulation of the recurrent laryngeal nerve in conscious toads (Bufo marinus). Journal of Comparative Physiology B 156: 363-375. Van Vliet, B.N. and N.H. West. 1987a. Response characteristics of pulmocutaneous arterial baroreceptors in the toad, Bufor marinus. Journal of Physiology 388: 55-70. Van Vliet, B.N. and N.H. West. 1987b. Responses to circulatory pressures, and conduction velocity of pulmoncutaneous baroreceptors in Bufo marinus. Journal of Physiology 388: 41-53. Van Vliet, B.N. and N.H. West. 1992. Functional characteristics of arterial chemoreceptors in an amphibian (Bufo marinus). Respiratory Physiology 88: 113-127. Vitalis, T.Z. and W.K. Milsom. 1986. Mechanical analysis of spontaneous breathing in the semi-aquatic turtle, Pseudemys scripta. Journal of Experimental Biology 125: 157-171. Wang, T., L.G.S. Branco and M.L. Glass. 1994. Ventilatory responses to hypoxia in the toad Bufo paracnemis before and after decrease in haemoglobin oxygen-carrying capacity. The Journal of Experimental Biology 186: 1-8. Wang, T. and S.T. Warburton. 1995. Breathing pattern and cost of ventilation in the alligator. American Respiratory Physiology 102: 29-37. Wang, T. and J.W. Hicks. 1996a. The interaction of pulmonary ventilation and the right-left shunt on arterial oxygen levels. The Journal of Experimental Biology 199: 2121-2129. Wang, T. and J.W. Hicks. 1996b. Cardiorespiratory synchrony in turtles. The Journal of Experimental Biology 199: (8): 1791-1800. Wang , T., E.H. Krosniunas and J.W. Hicks. 1997. The role of cardiac shunts in the regulation of arterial blood gases. American Zoologist 37: 12-22.  195  Wang, T., A.S. Abe and M.L. Glass. 1998. Effects of temperature on lung and blood gases in the South American rattlesnake Crotalus durissus terrificus. Comparative Biochemistry and Physiology 121A: 7-11. Wang, T., E.W., Taylor, D. Andrade, A.S. Abe. 2001a. Autonomic control of heart rate during forced activity and digestion in the snake Boa constrictor. Journal of Experimental Biology 204: 3553-3560. Wang, T., S. Warburton, A. Abe and T. Taylor. 2001b. Vagal control of heart rate and cardiac shunts in reptiles: relation to metabolic state. Experimental Physiology 86 (6): 777-784. Wang, T., M. Busk and J. Overgaard. 2001c. The respiratory consequences of feeding in amphibians and reptiles. Comparative Biochemistry and Physiology Part A 128: 535-549. Wang, T., J. Altimiras, W. Klein and M. Axelsson. 2003. Ventricular haemodynamics in Python molurus: separation of pulmonary and systemic pressures. Journal of Experimental Biology 206: 4241-4245. Wang, T., E.W. Taylor, S.G. Reid and W.K. Milsom. 2004. Interactive effects of mechanoand chemo- receptor inputs on cardiorespiratory outputs in the toad. Respiratory Physiology and Neurobiology 140: 63-76. Wang, Z.Y., E.B. Olson Jr., D.E. Bjorling, G.S. Mitchell, G.E, Bisgard. 2008. Sustainedhypoxia induced proliferatin of carotid body type I cells in rats. Nature 104: 803-808. Williams, S.C., P. Wootton, H.S. Mason, J. Bould. D. Iles, D. Riccardi, C. Peers and P.J. Kemp. 2004. Haemoxygenase-2 is an oxygen sensor for a calcium-sensitive potassium channel. Science 306: 2093-2097. West, N.H. and B.N. Van Vliet. 1983. Open-loop analysis of pulmocutaneous barorefles in the toad Bufo marinus. American Journal of Physiology 245: R642-650. West, N.H., Z.L. Topor and B.N. Van Vliet. 1987. Hypoxemic threshold for lung ventilation in the toad. Respiration Physiology 70: 377-390. West, N.H. and B.N. Van Vliet. 1992. Sensory mechanisms regulating the cardiovascular and respiratory systems. In: Environmental Physiology of the Amphibia. ed. M.E. Feder and W.W. Burggren, pp 151-182. Chicago, University of Chicago press. Wood, S.C. 1982. Effect of O2 affinity on arterial PO2 in animals with central vascular shunts. Journal of Applied Physiology 53 (6): 1360-1364. Wood, S.C. 1984. Cardiovascular shunts and oxygen transport in lower vertebrates. American Journal of Physiology 247: R3-R14. Wood, S.C. and C.J.M. Lenfant. 1976. Respiration: mechanics, control and gas exchange. In: Biology of the Reptilia, vol. 5. Gans, C. and W.R. Dawson (Eds.). Academic Press, London, New York, San Francisco, pp 225-274. Woodin, M. and R. Stephenson. 1998. Circadian rhythms in diving behavior and ventilatory response to asphyxia in canvasback ducks. American Journal of Physiology 274: R686R693. 196  Wyatt, C.N. and A.M. Evans. 2007. AMP-activated protein kinase and chemotransduction in the carotid body. Respiratory Physiology and Neurobiology 157: 22-29. Zaccone, G., J.M. Lauweryns, S. Fassulo, G. Tagliafierro, L. Ainis and A. Licata. 1992. Immunocytochemical localization of serotonin and neuropeptides in the neuroendocrine paraneurons of teleost and lungfish gills. Acta Zoologica 73 (3): 177-183. Zaccone, G., S. Fassulo and L. Ainis. 1994. Distribution patterns of the paraneuronal endocrinecells in the skin, gills and the airways of fishes as determined by immunohistochemical and histological methods. Histochemical Journal 26: 609-629. Zaccone, G., S. Fasulo, L. Ainis and A. Licata. 1997. Paraneurons in the gills and airways of fishes. Microscopy Research and Technique 37: 4-12. Zaccone, G., L. Ainis, A. Mauceri, P, Lo Cascio, L.G. Francesco and S. Fasulo. 2003. NANC nerves in the respiratory air sac and branchial vasculature of the Indian catfish, Heteropneustes fossilis. Acta Histochemica 105: 151-163. Zachar, P.C. and M.G. Jonz. 2012. Neuroepithelial cells of the gill and their role in oxygen sensing. Respiratory Physiology and Neurobiology 184: 301-308. Zhang, M., H. Zhong, C. Vollmer, C.A. Nurse. 2000. Co-release of ATP and Ach mediates hypoxic signalling at rat carotid body chemoreceptors. Journal of Physiology 525: 143158. Zari, T. A. 1999. Seasonal acclimatization in metabolic rate of the fan-fingered gecko, Ptyodactylus hasselquistii (Reptilia: Gekkonidae). Journal of Thermal Biology 24: 137142.  197  Appendices Appendix A - Supplementary material for Chapter 2  Figure A1. Positive controls for injections of the neuronal tracer Cholera toxin B (CTB) in the vagus nerve, serotonin (5HT) and catecholamines (tyrosine hydroxylase, TH) in the lung and adrenal gland of R. catesbeiana. A: Triple immunolabeling for CTB (red), 5HT (green) and a nuclear stain (DAPI, blue) in the lung of Rana catesbeiana, showing vagal innervation (red) to the lung and putative Neuroepithelial bodies (green). The images from the green and red channels are shown separately and as a merged image B: Double immunolabeling for TH (red) and cell nuclei (DAPI, blue) in the adrenal glands of R. catesbeiana. Scale bars 10 µm.  198  Appendix B Supplementary material for Chapter 3  Figure B1. Positive controls for vesicular acetylcholine transporter (VAChT), serotonin (5HT) and the neuronal tracer cholera toxin B (CTB) in the jugular ganglia and lungs of Trachemys scripta elegans. A: Immunolabeling for VAChT (red) and a nuclear stain (DAPI, blue) in the jugular ganglia. Scale bar 10 µm. B: Immunolabeling for 5HT (green) and a nuclear stain (DAPI, blue) in the lung. Scale bar 10 µm. The images from the green and blue channels are shown separately and as a merged image. C: Immunolabeling for CTB (red) and a nuclear stain (DAPI, blue) in the lung. The images from the red and blue channels are shown separately and as a merged image. Scale bar 50 µm.  199  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073886/manifest

Comment

Related Items