| Summary: | Odours are first represented in the brain as spatiotemporal maps of activity in the olfactory bulb (OB). Imaging and electrophysiological studies have shown that these maps are both temporally and spatially complex and unique to each odour. Behavioural tasks that probe perceptual differences between odours suggest that odours that evoke similar spatial activity maps in the OB are perceived as similar. However, combination of lesion and behavioural experiments of either the olfactory epithelium or bulb has suggested that rodents can detect and discriminate between odours using minimal stimulus-related input. This has led to a consensus in the field that sensory inputs to the olfactory system contain significant redundant signal and that spatial activity maps are unnecessary for odour coding. The work presented here used a go/no-go behavioural paradigm to investigate the ability of mice not just to detect or discriminate odours after nasal epithelial lesion but also to recognise odours – which enables odour quality perception to be probed. Intrinsic optical imaging was used in the same animals, to observe changes in odour-evoked signals in the OB before and after lesion. The results revealed that even moderatechanges to intrinsic activity maps caused deficits in both odour discrimination and recognition, suggesting that perception of odour quality was significantly altered. Reduction in odour inputs could be equivalent to reducing the intensity of inputs, so alterations to odour quality perception after changes in odour concentration were also examined. Recognition scores were reduced when mice were presented with a familiar odour at an unfamiliar concentration, suggesting odour perception was also significantly altered by reduction of stimulus intensity. In order to determine whether reductions in recognition score caused by lesioning and change in odour concentration had different perceptual origins, mice were trained to generalise across odour concentrations and tested for recognition after lesion. This revealed that impaired recognition after lesion resulted, not from experiencing an altered odour concentration, but from perception of apparent novel odour qualities. Consistent with this, intrinsic imaging data revealed that relative intensity of glomerular activity following lesions was altered compared with maps recorded in shams or by varying odour concentration. Long-standing theories of sensory coding suggest that sensory systems actively match odours in the environment with stored stimulus templates. Odours familiar before lesioning were re-learnt more rapidly after lesioning than novel odours were learnt either before or after lesioning. This suggests that stored templates of familiar odours were compared to moderately altered incoming inputs and, with reinforcement, were rapidly incorporated into those templates. In all, this work suggests that odour quality perception requires comprehensive matching of input patterns to stored representations, suggesting that spatial activity maps are a crucial component of odour coding.
|
|---|
