Novel Tests of Complex Recognition Memory in Animals and Humans

Abstract

This thesis sought to address specific methodological issues relating to tasks of recognition memory in animals and humans. Such tasks are very widely used, so the need to reduce variability and improve the translation of animal work to humans is apparent. Study 1 sought to develop a reliable testing method based on the spontaneous recognition paradigm that would reduce the animal numbers required for such tasks. Rats displayed significant performance in multiple recognition tasks carried out in the continual trials apparatus, which allows for multiple trials within a session. Approximately 50% fewer animals were required for statistically meaningful results, compared to studies using the standard one trial a day paradigm. Study 2 sought to further develop the continual trials apparatus for an episodic-like memory task for rodents. This study focussed on the development of an object preference task to investigate the behavioural parameters that would affect recognition in the test phase of the E-maze task. Study 3 aimed to investigate whether the continual trials apparatus could be effectively applied with immediate-early gene imaging during a recognition memory task. Animals tested with novel stimuli showed greater fos expression than animals tested with familiar objects, though not significantly. Finally, Study 4 focussed on the translation of animal models of memory to humans. The analysis of receiver-operating characteristics was used to derive a quantifiable distinction between recollection- and familiarity-based processes of recognition, in a task based on paradigms typically used with rodents.
The key findings of the work in this thesis include evidence of substantial animal reduction using a new behavioural apparatus for assessing recognition memory in rodents, and the successful development of an analogous human task of memory in which processes of recognition can be dissociated and quantified. These two key findings make a significant contribution to the field of recognition memory research as the new rodent behavioural tasks are a clear improvement on standard tasks with the potential to reduce variance and animal numbers, and reducing the reliance on human subjects’ introspective accounts of memory in Study 4 provides a shift away towards better controlled behavioural studies in humans, which more closely reflects the studies carried out with animals, and provides strong validation for particular animal models. Through further validation, the simplicity of the human memory task could make it a useful candidate for assessing different forms of recognition memory with neuropsychological subjects.