Teleosts, or ray-finned fish, are the largest and most diverse class of vertebrates in the world. Among this diversity includes many differences in fin structure. Zebrafish typically have only soft rays in their fins, while acanthomorph fish have both soft and spiny rays in select fins. How this difference between fin elements arose remains an unanswered question in evolutionary biology, although hox gene expression patterns could play a prominent role in this evolutionary event. Hox genes encode transcription factors that are important for patterning during development. Specifically, the hox13 genes have been shown to be essential for proper limb and fin patterning and are expressed only in soft rays and not spiny rays during acanthomorph fin development. Hoxa13 -/-, Hoxd13 -/- mice completely lack the autopod in their developing limbs. The function of the fish orthologs hoxa13a, hoxa13b, and hoxd13a are not as well defined in fin ray development, although clear structural changes to the fin rays such as truncations and loss of joints, bifurcations and actinotrichia can be seen in the absence of these genes. In this project, various zebrafish compound mutants for hoxa13a, hoxa13b, and hoxd13a are compared to gain insight into the individual roles of each hox13 gene during fin development. From these observations, hoxa13a and hoxa13b appear to have a more prominent role in fin ray patterning, and only require one copy to produce joints, bifurcations and actinotrichia in the rays of every fin. Hoxd13a however requires two copies to perform a similar function. After generating and observing triple hox13 mutant zebrafish, a comparison between these mutant rays and acanthomorph spines was performed using micro-CT scanning and in situ hybridization. Triple hox13 mutant fin rays were found to have highly ossified fin rays as seen in spines, as well as proportionally increased expression of the spine marker alx4a during early development. All in all, the triple hox13 mutant rays appear to be forming an intermediate structure between soft rays and spiny rays, highlighting the potential impact of hox13 downregulation in the evolution of acanthomorph spines.

The most frequent mutation identified in both benign nevi and malignant melanoma is the constitutively activating V600E substitution of BRAF. However, how additional mutations co-operate with BRAFV600E to promote subtypes of melanoma in animals is only beginning to be understood. In this thesis, I generate and analyze zebrafish BRAFV600E melanoma models and also develop the first animal model for BRAFV600E nevus recurrence. In my first data chapter, I develop a unique animal model of nevus recurrence. In people it is not uncommon for a nevus to recur following removal, even when no pigmented nevus cells remain. The biology of how and why this happens is unknown. By partial amputation of the nevus in the zebrafish tail fin, we described both nevus regrowth, as well as nevi that do not regrow. Utilising melanin as a lineage tracer, I was able to show that recurrent nevi are repopulated from an unpigmented precursor population. This suggested that BRAFV600E nevi are supported by an undifferentiated stem cell population that is recruited to regenerate and pigment the nevus after removal. In my second data chapter, I use genetics to develop BRAFV600E zebrafish models of melanoma. In collaboration with Dr. James Lister and Professor Jeroen den Hertog, I describe three differing models of zebrafish melanoma. All three models show progression to melanoma, and in collaboration with Dr. Marie Mathers I establish that while BRAFV600E is present in all three models, co-operating mutations affect melanoma pathology. In my third data chapter, I develop tools to study the molecular differences in the BRAFV600E melanoma models. I described the optimisation of a broad range of antibodies, raised against human peptides due to the lack of reliable antibodies in the zebrafish field. I use punch core biopsies of both zebrafish and human tumours, and whole sagittal sections of juvenile zebrafish, to show specific staining throughout many organs of the developing fish. I then use some of these antibodies to analyse molecular pathways in the melanoma models.

Zebrafish swimming behavior is used in a new, automated drug assay system as a biomarker to measure drug efficiency to prevent or restore hearing loss. This system records video of zebrafish larvae under infrared lighting using Raspberry Pi cameras and measures fish swimming behavior. This automated system significantly reduces the operator time required to process experiments in parallel. Multiple tanks, each consisting of sixteen experiments are operated in parallel. Once a set of experiments starts, all data transfer and processing operations are automatic. A web interface allows the operator to configure, monitor and control the experiments and review reports. Ethernet connects the various hardware components, allowing loose coupling of the distributed software used to schedule and run the experiments. The operator can configure the data processing to be done on the local computer or offloaded to a high-performance computer cluster to achieve even higher throughput. Computationally efficient image processing algorithms provided automated zebrafish detection and motion analysis. Quantitative assessment of error in the position and orientation of the detected fish uses manual data analysis by human observers as the reference. The system error in orientation and position is comparable to human inter-operator error.

Freshwater (FW) fish inhabit hypotonic environments that can vary markedly both spatially and temporally with respect to ambient salt levels and pH. Despite the chemical variability of FW, fish maintain ionic homeostasis (ionic regulation) and pH homeostasis (acid-base regulation) by manipulating ion transport mechanisms within ion-transporting cells (ionocytes) localised to the gills of adults and the skin of larvae. Ionocytes are mitochondrion-rich (MR) cells that, depending on subtype, express specific ion transporters that facilitate the movement of salts and acid-base equivalents across the gills or skin. In zebrafish (Danio rerio), one of the most well-studied ionocytes is termed the H+-ATPase-rich (HR) cell, which is presumed to be a significant site of transepithelial Na+ uptake/acid-secretion. Proteins that have been found in fish zebrafish HR cells include the Na+/H+ exchanger (NHE3), carbonic anhydrases (CA17a and CA15a), proton ATPase (H+-ATPase) and the ammonia channel, ammonia conducting rhesus C glycoprotein b (Rhcgb), which are all thought to function in Na+ uptake acid–base regulation. Ionic and acid-base regulation are achieved both by adjustments to the activity level of these ion transport proteins, but also by regulating the numbers of specific ionocyte subtypes (e.g. HR cells) during acclimation to environments differing in ionic composition or pH. In previous studies, the quantitative assessment of mRNA levels for genes involved in ionic and acid-base regulation relied on measurements using homogenates derived from whole body (larvae) or gill (adult). Such studies cannot distinguish whether any differences in gene expression arise from adjustments of ionocyte subtype numbers, or transcriptional regulation within individual ionocytes. Surprisingly, there are no data on ionocyte-specific gene expression responses in zebrafish exposed to varying environments including acidic or Na+-deficient water. To rectify this gap in the current knowledge, this thesis utilized the florescence activated cell sorting (FACS) approach to separate the HR cells from other cellular sub-populations. The technique was used to measure the gene expression of several HR cell specific transporters and enzymes in isolated HR cells from zebrafish larvae exposed to low pH (pH 4.0) or low Na+ (5 μM) conditions. The data suggest that treatment of larvae at 4 days post fertilization with acidic water caused an increase in h+atpase, ca17a, ca15a, nhe3b and rhcgb mRNA levels in sorted HR cells. These observations suggest the existence of multiple mechanisms of acid secretion in HR cells of larval zebrafish in acidic water; one in which acid secretion via NHE3b is linked to ammonia excretion via Rhcgb, and another facilitated by H+-ATPase. Furthermore, these results provide molecular evidence to support roles for the CA isoforms in acid-base regulation in HR cells. On the other hand, the low Na+ treatment data suggest that nhe3b and rhcgb are the dominant genes maintaining Na+homeostasis. In summary, the results of this thesis demonstrate that acclimation to low pH or low low Na+ environmental conditions is facilitated by HR cell proliferation and HR cell-specific transcriptional control.

Oculocerebrorenal syndrome of Lowe is a rare X-linked disorder caused by mutation of the inositol 5-phosphatase OCRL1. Lowe Syndrome manifests as renal tubular dysfunction, neurological and ocular defects. OCRL1 uses its catalytic domain to hydrolyze two phosphoinositide species, PI(4,5)P2 and PI3,4,5)P3. It is involved in regulation of membrane trafficking, actin dynamics, cytokinesis and ciliogenesis. OCRL1 interacts with IPIP27A and B, which have been shown to be key players in endocytic trafficking in mammalian cells, specifically in the recycling of proteins from early and recycling endosomes to both the plasma membrane and trans-Golgi network. It has been proposed that defective endocytic trafficking may be responsible for the renal tubulopathy seen in Lowe Syndrome patients, characterized by low molecular weight proteinuria and aminoaciduria, but this hypothesis has yet to be tested. Using zebrafish as a model for Lowe syndrome, we show that depletion of OCRL1 can indeed cause defects in endocytosis in the renal tubule. This coincides with a reduction in levels of the multi-ligand receptor megalin, reduced abundance of the endocytic apparatus and increased numbers of enlarged lysosomes in the kidney tubular cells. We also show that knocking-down Pip5K in the OCRL1 mutants to rebalance PI(4,5)P2 levels can rescue the endocytic defect. This indicates that tight control of PI(4,5)P2 level is essential for efficient endocytic trafficking in vivo. Importantly, this finding suggests that Pip5K may be a valuable therapeutic target for patients with Lowe Syndrome. To further characterize the molecular mechanisms by which OCRL1 promotes endocytosis, we have focused on the recently identified OCRL1 interaction partners IPIP27A and B, which are known to function in endocytosis and receptor recycling. Here we report identification and characterization of the zebrafish Ipip27s, including analysis of conservation and expression profiles. To assess Ipip27s function in vivo, KO zebrafish lines were generated using TALENs. This was successful for Ipip27A, but so far not for Ipip27B. Functional analysis using the Ipip27A KO line and KD with morpholinos revealed that both Ipip27s contribute to neural development and may participate in ciliogenesis. Moreover, preliminary analysis indicates an important role for Ipip27A within the endocytic pathway in the kidney tubule, where its loss phenocopies many aspects of the OCRL1 mutant phenotype.

The zebrafish has an amazing capacity for regeneration which includes regeneration of neurons within the central nervous system (CNS) both during development and into adulthood. This attribute makes the zebrafish a valuable tool in the study of regeneration. In this thesis, the research focussed on the regeneration of a specific type of cell in the CNS, dopaminergic (DA) neurons. The DA system of the zebrafish is believed to be evolutionarily conserved with comparable DA populations found in the brain of mammals. Dissimilar to mammals, however, the zebrafish is capable of regenerating various types of neurons and their axons. Thus, the zebrafish DA system provides an excellent model to study replacement of this specific and important cell type in the adult CNS. We have developed a novel toxin ablation paradigm to specifically ablate select groups of DA neurons in the adult zebrafish diencephalon, leaving other DA populations unaffected. To do this a selective DA toxin, 6-hydroxydopamine, was used. One of the ablated DA diencephalic populations is the only source of dopaminergic spinal innervation in the zebrafish. Their ablation leads to a loss of DA spinal axons following our toxin ablation. The ascending projection of the diencephalic population ablated by the toxin has been suggested as the most likely candidate for a zebrafish equivalent of the mammalian nigro-striatal pathway. The loss of cells is very specific and reproducible, indicating that these cells are particularly vulnerable to the toxin. Quantification of affected populations at various time-points post ablation was carried out to determine the capacity for regeneration of DA neurons in the CNS of zebrafish. This revealed that in some populations neuron numbers returned to those seen in controls. However, in other populations neuron numbers only partially recovered even at late time points. We have shown that this recovery is due to neurogenesis; furthermore, by inducing inflammation after the toxin treatment the recovery of DA cell numbers was accelerated by 50%. Regenerated cells originated from Olig2 positive ependymo-radial glial cells found bordering the diencephalic ventricle. We aimed to investigate the function of this group of ablated neurons through a battery of behavioural tests. These tests revealed deficits in the toxin treated animals’ fine movement, such as is necessary for maintaining shoal cohesion and breeding behaviours, whereas general movement behaviours were not found to be impaired. Zebrafish embryos also present as a great resource in the screening of drugs. Their fast and well characterised early development makes them an ideal tool for investigating previously untested neuroprotectants. A reproducible ablation paradigm similar to that established in the adults was also established in the zebrafish embryo. This was then used as a tool to investigate potential novel neuroprotectants. This screen revealed two new flavonoid compounds which had the ability to induce full protection of the affected dopaminergic cells in the zebrafish embryonic brain. The embryonic ablation model therefore represents a vertebrate in vivo model system for future high throughput screening of neuroprotective compounds against toxin induced DA cell loss. Ultimately, understanding how zebrafish functionally regenerate dopaminergic neurons using this ablation model will likely provide a useful tool into the research of neurodegenerative diseases, such as PD.

Myelin is a protective layer wrapped around axons which helps them conduct electrical signals rapidly, and provides them with metabolic support. In the central nervous system (CNS), myelin is produced by specialised glial cells called oligodendrocytes. Loss of myelin (demyelination) is associated with degeneration of axons and many neurodegenerative disorders, including multiple sclerosis (MS). The restoration of myelin sheaths by remyelination may protect axons and help functional recovery of patients, but achieving this requires better understanding of how the process unfolds at the cellular level. To investigate the processes of de- and remyelination in vivo, I have characterised a transgenic zebrafish line in which expression of the bacterial enzyme nitroreductase (NTR) is driven under the myelin basic protein promoter, thus in myelinating glia. I treat larvae with the NTR substrate metronidazole (Mtz). The reaction between NTR and Mtz results in a toxic metabolite which selectively kills NTR-expressing cells. The treatment with Mtz consistently ablates two-thirds of oligodendrocytes while not harming the animals otherwise. Myelin sheaths continue to deteriorate after the end of the treatment, such that seven days later, extensive demyelination is observed by electron microscopy. By 16 days after Mtz-treatment, robust recovery has occurred, with no discernible axon loss and myelin thickness restored to control levels. At this time point, oligodendrocyte numbers have also returned to control levels. During the demyelinated phase, I observe a striking increase in microglia and macrophages in the spinal cord. In order to study the role of the innate immune system in recovery, I used a mutant line, irf8-/- which lacks a transcription factor essential for development of microglia and macrophages. I am in the process of determining the ability of these mutants to regenerate their oligodendrocytes and myelin; preliminary results suggest that they are able to restore their myelin sheaths fully, but seem to have a delay in regenerating their oligodendrocytes compared to wild-types. The model I have established can be used in the future to better understand the consequences of demyelination to axon health, as well as chemical screening to identify compounds that could accelerate the remyelination process or enhance the thickness of myelin generated during remyelination. Insights arising from such studies will be useful in designing strategies to reduce axon loss and improve myelin regeneration in demyelinating diseases.

Myelin is critical for the operation of a functional vertebrate nervous system, allowing for rapid saltatory conduction and providing trophic support to axons. In multiple sclerosis (MS), the immune system attacks myelin sheaths, leading to de-myelination of axons. De-myelinated axons not only lose their ability to conduct rapid nerve impulses, but are themselves susceptible to damage and loss. Long term demyelination leads to neuronal loss and the devastating symptoms of secondary stages of MS. One therapeutic approach which has been suggested is to improve the ability of oligodendrocyte precursor cells (OPCs) to differentiate into mature re-myelinating oligodendrocytes. This process is known to occur in vivo, however, the myelin produced appears reduced and the efficiency with which OPCs differentiate into myelinating oligodendrocytes (OLs) varies greatly. For example, the ability of OPCs from older mice to differentiate is reduced compared to those from young mice. This fact taken alongside the presence of many OPCs in some MS lesions which have failed to re-myelinate makes identifying compounds which can increase OPC differentiation into OLs a key goal for MS drug development. In this work, I use OPC to OL differentiation during zebrafish development as a model for differentiation of OLs more generally. Zebrafish are widely used for chemical screening, with recent developments in genetic manipulation, such as CRISPR/Cas9 gene editing and Tol2 transgenesis, allowing for production of targeted mutations and fluorescent reporter lines respectively. Retinoid X receptor-γ (RXRγ) has previously been identified as being transcriptionally upregulated during re-myelination. Moreover, treatment with 9-cis-retinoic acid, an agonist for the receptor, has been shown to improve remyelination in vivo in rats with toxin induced focal demyelination. I first present a manual chemical screen of a library of compounds designed to target RXRγ, from which I identify several compounds which reproducibly increase OLs in zebrafish. In order to assess whether any of the hit compounds could be acting as agonists for RXRγ, I have created a double knockout zebrafish lacking both genes coding for the zebrafish RXRγ homologues (rxrga and rxrgb). This line has been used to test the activity of hit compounds in a RXRγ loss of function background. Following this first chemical screen, it was clear that great improvements could be made to both throughput and robustness if the screen was automated. Using a commercially available fish handling robot which automates the imaging of plates of zebrafish embryos, known as the VAST BioImager, the throughput of our assay was increased from ~40 fish per day to up to around 300 to 400. We combined the VAST BioImager with a state of the art spinning disk confocal microscope, giving us (to the best of our knowledge) the world's fastest in vivo vertebrate screening system capable of orienting fish and imaging at sub-cellular resolution. This significant increase in rate of fish processing led to the need for an increase in the rate of image analysis. Much of the gains in throughput would be lost to time counting cells, and so I developed software to automate the image processing and analysis. The software developed is shown to closely match the abilities of a human to identify compounds which give significant increases in OLs, with very little human intervention required. In the final section of this work I present an example screen performed using the VAST BioImager in combination with the automated cell counting software, which I developed. The hits from this screen highlight our ability to automatically identify compounds that increase the number of OLs in the developing zebrafish. This method is broadly applicable to other central nervous system cell types and other methods of analysis can be integrated into the presented screening software.

A wide array of species, from ants to humans, live or forage in groups. Shoaling – the formation of groups by fish – confers protection from predation and enhances foraging. However, little is known about the detailed characteristics or the dynamics of shoaling. Shoaling is a complex social interaction and a better understanding of its mechanisms and limitations would permit the study of natural and induced changes on social behavior generally in fish. Here, I present data on the shoaling characteristics of zebrafish (Danio rerio). Novel tracking techniques are used to extract detailed trajectories of all members of a free-swimming shoal of zebrafish. Multiple measures of shoaling – such as distributions of nearest neighbor distances, shoal polarizations, and speeds – are calculated, to better describe the subtleties of the behavior including, for the first time, the high resolution spatio-temporal dynamics of shoaling. In addition, a novel criterion is introduced to determine when and how individual fish or sub-groups leave the shoal. Comparisons are presented between the shoaling characteristics of three populations of zebrafish (LFWT, SFWT, AB) and between days and hours of repeated exposure to the same testing environment, demonstrating the gradual effects of habituation on shoaling. In addition, the effects of manipulating the number of fish in the shoal, hunger levels, and predation threat are also examined, lending empirical support to ecological theories on the adaptive functions of the behavior. Finally, the data are compared to two leading theoretical models of shoaling and a novel simulation approach is suggested. The data strongly suggest that various aspects of shoaling in zebrafish are constantly changing, complex, and flexible, representing a dynamic form of social cognition. The study of these characteristics sheds much-needed light on complex social interactions in this popular genetic model organism, which may eventually lead to a better understanding of social behaviors in other species, including our own.

A wide array of species, from ants to humans, live or forage in groups. Shoaling – the formation of groups by fish – confers protection from predation and enhances foraging. However, little is known about the detailed characteristics or the dynamics of shoaling. Shoaling is a complex social interaction and a better understanding of its mechanisms and limitations would permit the study of natural and induced changes on social behavior generally in fish. Here, I present data on the shoaling characteristics of zebrafish (Danio rerio). Novel tracking techniques are used to extract detailed trajectories of all members of a free-swimming shoal of zebrafish. Multiple measures of shoaling – such as distributions of nearest neighbor distances, shoal polarizations, and speeds – are calculated, to better describe the subtleties of the behavior including, for the first time, the high resolution spatio-temporal dynamics of shoaling. In addition, a novel criterion is introduced to determine when and how individual fish or sub-groups leave the shoal. Comparisons are presented between the shoaling characteristics of three populations of zebrafish (LFWT, SFWT, AB) and between days and hours of repeated exposure to the same testing environment, demonstrating the gradual effects of habituation on shoaling. In addition, the effects of manipulating the number of fish in the shoal, hunger levels, and predation threat are also examined, lending empirical support to ecological theories on the adaptive functions of the behavior. Finally, the data are compared to two leading theoretical models of shoaling and a novel simulation approach is suggested. The data strongly suggest that various aspects of shoaling in zebrafish are constantly changing, complex, and flexible, representing a dynamic form of social cognition. The study of these characteristics sheds much-needed light on complex social interactions in this popular genetic model organism, which may eventually lead to a better understanding of social behaviors in other species, including our own.