Quantitative Behavioural Phenotyping to Investigate Anaesthesia-Induced Neurobehavioural Impairment

Abstract

Background: Vertebrate animal experiments have suggested that anaesthesia exposure to the developing nervous system causes neuroapoptosis and behavioural impairment. Mechanistic understanding is limited, and target-based approaches are challenging. High-throughput methods may be an important parallel approach to drug-discovery and mechanistic research. The nematode worm Caenorhabditis elegans is an ideal candidate as a model for this. A rich subset of its behaviour can be studied, and hundreds of morphological and behavioural features can be quantified, then aggregated to yield a 'signature'. Perturbation of this behavioural signature may provide a tool that can be used to quantify the effects of anaesthetic regimes, and act as an outcome marker for drug screening and molecular target research. Methods. Larval C. elegans were exposed to: isoflurane, ketamine, morphine, dexmedetomidine, and lithium (and combinations). Behaviour was recorded, and videos analysed with automated algorithms to extract behavioural and morphological features. Results: Anaesthetic exposure during early development leads to persisting morphological and behavioural variation (in total, 125 features/exposure combinations). Higher concentrations, and combinations of isoflurane with ketamine, lead to more feature 'hits'. Morphine and dexmedetomidine do not appear to lead to behavioural impairment. Lithium rescues the neurotoxic phenotype produced by isoflurane. Conclusions: Findings correlate well with vertebrate research: impairment is dependent on agent, is concentration-specific, is more likely with combination therapies, and can potentially be rescued by lithium. These results suggest that C. elegans may be an appropriate model with which to pursue phenotypic screens for drugs that mitigate the neurobehavioural impairment. Some possibilities are suggested for how high-throughput platforms might be organised in service of this field.