Novel synthetic and poly-substance misuse pose major public health concerns due to their rapid

synthesis, widespread availability, unknown toxicity, and high lethality. Rapid and reliable

hazard assessment methods are therefore urgently needed to determine the toxicity of these

substances as soon as they are identified, however current approaches are limited by time and

ethical constraints. This thesis evaluates the 4 days post-fertilisation (dpf) larval zebrafish

(Danio rerio) model as a rapid, high-throughput, and ethical tool bridging in vitro and in vivo

approaches. The light/dark behavioural assay, which is the most used for investigating drug-

induced effects, lacked experimental consistency and was initially validated through

experimental optimisation and development of a robust normalisation approach to improve

reproducibility. Model predictive validity was then confirmed with fentanyl, diacetylmorphine

(heroin), and remifentanil. Fentanyl bioavailability was assessed using liquid chromatography

with tandem mass spectrometry (LC-MS/MS). Behavioural effects were reversed with

naloxone, confirming opioid receptor activation. In vivo whole-brain Ca2+

-transient imaging

identified significantly affected brain regions with fentanyl (e.g., anterior commissure) and low

concentration hyperactivity was reversed with the dopamine D2 receptor antagonist, sulpiride,

implicating dopaminergic mechanisms. Functional connectivity brain analyses revealed

disrupted sensory-motor pathways between the torus semicircularis and eminentia granularis.

Furthermore, the model demonstrated suitability for assessing the behaviour of g-aminobutyric

acid (GABA)A positive allosteric modulator’s (PAM) and N-methyl-D-aspartate (NMDA)

receptor antagonists. Finally, co-administration of fentanyl and ethanol produced a novel

hyperactive phenotype, replicable in adult zebrafish, involving multiple neurotransmitter

systems with significant neuronal dysregulation identified. Taken together, these findings

demonstrate the utility of the model as a first-tier rapid hazard assessment tool of synthetic

substances and poly-substance interactions, supporting their integration into toxicological

research. However, caution for translational extrapolation to human risk evaluations remain.