Serotonin (5-HT) containing neuroepithelial cells (NECs) are O2 sensitive chemoreceptors found throughout the skin of larval zebrafish (Danio rerio). Zebrafish larvae are sensitive to changes in ambient PO2 as early as 2 days post fertilisation (dpf) and hyperventilate in response to hypoxia beginning at 3 dpf. Tryptophan hydroxylase (tph) is the rate-limiting enzyme in 5-HT synthesis; three tph paralogs are present in zebrafish (tph1a, tph1b and tph2). Although 5-HT has been implicated as a key neurotransmitter mediating hypoxic hyperventilation, it has not been possible to discern the role of 5-HT specifically contained within the NECs in promoting hypoxic hyperventilation. The purpose of this study was to determine the role of NEC 5-HT in regulating the hypoxic ventilatory response in larval zebrafish. It was hypothesised that 5-HT is a key neurotransmitter released from NECs which contributes to hypoxic hyperventilation. Immunohistochemistry was used to determine the distribution of tph paralogs and their role in 5-HT production in NECs. Tph1a was present in NECs and nerves innervating NECs. Exposure to the non-selective tph inhibitor, para-chlorophenylalanine (pCPA), or translational gene knockdown of tph1a, diminished 5-HT expression within NECs. Exposure to acute hypoxia (PO2 = 30 mmHg) revealed a blunted hypoxic ventilatory response (reduced breathing frequency) in fish exhibiting depleted 5-HT in NECs. The hypoxic hyperventilatory response was rescued with application of 5-HT. The results of these experiments demonstrate that tph1a is responsible for 5-HT production in NECs of larval zebrafish, and that 5-HT released from NECs is involved in establishing their hypoxic hyperventilatory response.

Serotonin (5-HT)-containing neuroepithelial cells (NECs) of the gill filament are believed to be the primary O2 chemosensors in fish. In the mammalian carotid body (CB), 5-HT is one of many neurotransmitters believed to play a role in transduction of hypoxic stimuli, with acetylcholine (ACh) being the primary fast-acting excitatory neurotransmitter. Immunohistochemistry and confocal microscopy was used to observe the presence of the vesicular acetylcholine transporter (VAChT), a marker for the presence of ACh, and its associated innervation in the gills of zebrafish. VAChT-positive cells were observed primarily along the afferent side of the filament, with some cells receiving extrabranchial innervation. No VAChT-positive cells were observed in the gills of goldfish; however, certain key morphological differences in the innervation of goldfish gills was observed, as compared to zebrafish. In addition, in zebrafish NECs, whole-cell current is dominated by an O2-sensitive background K+ current; however, this is just one of several currents observed in the mammalian CB. In zebrafish NECs and the CB, membrane depolarization in response to hypoxia, mediated by inhibition of the background K+ (KB) channels, is believed to lead to activation of voltage-gated Ca2+ (CaV) channels and Ca2+-dependent neurosecretion. Using patch-clamp electrophysiology, I discovered several ion channel types not previously observed in the gill chemosensors, including Ca2+-activated K+ (KCa), voltage-dependent K+ (KV), and voltage-activated Ca2+ (CaV) channels. Under whole-cell patch-clamp conditions, the goldfish NECs did not respond to hypoxia (PO2 ~ 11 mmHg). Employing ratiometric calcium imaging and an activity-dependent fluorescent vital dye, I observed that intact goldfish NECs respond to hypoxia (PO2 ~ 11 mmHg) with an increase in intracellular Ca2+ ([Ca2+]i) and increased synaptic vesicle activity. The results of these experiments demonstrate that (1) ACh appears to play a role in the zebrafish, but not goldfish gill, (2) goldfish NECs likely signal hypoxic stimuli primarily via the central nervous system (CNS), (3) goldfish NECs express a broad range of ion channels as compared to the NECs of zebrafish, and (4) goldfish NECs rely on some cytosolic factor(s) when responding to hypoxia (PO2 ~ 11 mmHg). This thesis represents a further step in the study of neurochemical and physiological adaptations to tolerance of extreme hypoxia.

Endogenously produced carbon monoxide (CO) is an important gaseous signalling molecule which regulates a variety of cardiorespiratory functions. CO is produced in cells by the heme oxygenase (HO) family of proteins by the breakdown of heme into equimolar amounts of CO, bilirubin and Fe2+. My thesis focuses on the hypoxia- and hyperoxia-inducible HO-1/CO system exclusively and aims to provide the first evidence that the HO-1/CO system is involved in cardiorespiratory control in the zebrafish (Danio rerio) and goldfish (Carassius auratus). Overall, I hypothesise that the HO-1/CO system acts as a negative regulator of cardiorespiratory function in fish. Using immunohistochemistry, I was able to characterise the distribution of HO-1 and thus reveal the potential for endogenous CO production (from heme breakdown) in branchial and skin neuroepithelial cells (NECs; putative O2 chemoreceptors) and associated innervation as well as the heart of the developing zebrafish larva. The presence of HO-1 in these structures suggests the likelihood of specific and localized production of CO in fish. To assess the functional significance of the HO-1/CO system in control of cardiorespiratory function, I used pharmacological and gene knock down approaches to diminish HO-1 activity, and presumably endogenous CO production, in adult and larval fish, respectively. The results from these experiments provided evidence that 1) CO has an inhibitory influence on ventilation in goldfish and zebrafish but that its function is temperature- and species-dependent and 2) showed that the HO-1/CO system tonically inhibits cardiac activity in larval zebrafish.

Serotonin (5-HT)-containing neuroepithelial cells (NECs) of the gill filament are believed to be the primary O2 chemosensors in fish. In the mammalian carotid body (CB), 5-HT is one of many neurotransmitters believed to play a role in transduction of hypoxic stimuli, with acetylcholine (ACh) being the primary fast-acting excitatory neurotransmitter. Immunohistochemistry and confocal microscopy was used to observe the presence of the vesicular acetylcholine transporter (VAChT), a marker for the presence of ACh, and its associated innervation in the gills of zebrafish. VAChT-positive cells were observed primarily along the afferent side of the filament, with some cells receiving extrabranchial innervation. No VAChT-positive cells were observed in the gills of goldfish; however, certain key morphological differences in the innervation of goldfish gills was observed, as compared to zebrafish. In addition, in zebrafish NECs, whole-cell current is dominated by an O2-sensitive background K+ current; however, this is just one of several currents observed in the mammalian CB. In zebrafish NECs and the CB, membrane depolarization in response to hypoxia, mediated by inhibition of the background K+ (KB) channels, is believed to lead to activation of voltage-gated Ca2+ (CaV) channels and Ca2+-dependent neurosecretion. Using patch-clamp electrophysiology, I discovered several ion channel types not previously observed in the gill chemosensors, including Ca2+-activated K+ (KCa), voltage-dependent K+ (KV), and voltage-activated Ca2+ (CaV) channels. Under whole-cell patch-clamp conditions, the goldfish NECs did not respond to hypoxia (PO2 ~ 11 mmHg). Employing ratiometric calcium imaging and an activity-dependent fluorescent vital dye, I observed that intact goldfish NECs respond to hypoxia (PO2 ~ 11 mmHg) with an increase in intracellular Ca2+ ([Ca2+]i) and increased synaptic vesicle activity. The results of these experiments demonstrate that (1) ACh appears to play a role in the zebrafish, but not goldfish gill, (2) goldfish NECs likely signal hypoxic stimuli primarily via the central nervous system (CNS), (3) goldfish NECs express a broad range of ion channels as compared to the NECs of zebrafish, and (4) goldfish NECs rely on some cytosolic factor(s) when responding to hypoxia (PO2 ~ 11 mmHg). This thesis represents a further step in the study of neurochemical and physiological adaptations to tolerance of extreme hypoxia.

Early development of O2 chemoreception and hypoxic responses under normoxic (150 mmHg) and chronically hypoxic (110 mmHg) conditions were investigated in Xenopus laevis from hatching to 3 weeks post fertilization. Development, growth, O2 consumption, ventilatory and cardiac performance, and branchial neuroepithelial cells (NEC) density and size were determined. At 3 days post fertilization (dpf), larvae started gill ventilation at a rate of 28 ± 4 beats/min and showed increased frequency to 60 ± 2 beats/min at a PO2 of 30 mmHg. Also at 3 dpf, NECs were identified in the gill filament buds using immunohistochemical methods. Lung ventilation began at 5 dpf and exhibited a 3-fold increase in frequency from normoxia to a PO2 of 30 mmHg. Hypoxic tachycardia developed at 5 dpf, causing an increase of 20 beats/min in heart rate, which led to a 2-fold increase in mass-specific cardiac output at a PO2 of 70 mmHg. At 10 dpf, gill ventilatory sensitivity to hypoxia increased, which was associated with the increase in NEC density, from 15 ± 1 to 29 ± 2 cells/mm of filament at 5 and 10 dpf, respectively. Unlike the elevated rate, cardiac and ventilatory volumes were independent of acute hypoxia. Despite increased cardioventilatory frequency, larvae experienced an average of 80% depression in during acute hypoxia. Chronic hypoxia (PO2 of 110 mmHg) decreased mass-specific cardiac performance before 10 dpf. In older larvae (10 to 21 dpf), chronic hypoxia decreased acute branchial and pulmonary hypoxic hyperventilation and increased NEC size. Collectively, these data suggest that larvae exhibit strong O2-driven acute hypoxic responses post-hatching, yet are still O2 conformers. All acute hypoxic responses developed before 5 dpf, and then the effects of chronic hypoxia started to show between 7 and 21 dpf. Thus, the early formation of acute hypoxic responses is susceptible to the environment and can be shaped by the ambient PO2.