1Neuroscience, Physiology and Pharmacology Dept., UCL London, UK
2Max Planck Institute for Biological Cybernetics, Her Institute, Tuebingen, Germany
Understanding how patterns of neuronal activity contribute to odor perception will shed light on how external information is translated by the nervous system. Previously, large-scale lesioning of the nasal epithelium or the main olfactory bulb (OB) has been reported to produce only mild deficits in odor detection and discrimination. This has led to the idea that much of the glomerular activity within the OB is redundant and that odor processing can be achieved with a limited amount of sensory input. Here we have combined behavioural experiments and in vivo imaging to directly examine the relationship between OB activity and odor perception in mice with disrupted olfactory representations.
Water-deprived mice (C57BL/6J, P30-70) were trained to discriminate a monomolecular rewarded odor (S+) from an unrewarded odor (S-) using a go/no-go behavioural paradigm. Subsequently, under pentobarbitone anesthesia (1.1mg/g b.w.), S+ and S- odor - evoked activity on the dorsal surface of the OB was recorded using standard intrinsic-signal imaging. Mice then received a nasal flush of either ZnSO4 (8.4%) to partially ablate sensory input, or NaCl (9%, sham treatment). Three to five days later, mice were assessed on discrimination of an unfamiliar odor pair and then on recognition of the pre-treatment S+ and S- odors. Mice were then re-anesthetised and a second imaging session was carried out to assess the representation of S+ and S- odors.
We find that ZnSO4 - treated mice required significantly more trials to discriminate rewarded unfamiliar odors accurately (≥ 80% correct) than sham treated mice (336 ± 39.9, n = 5 mice vs 166.6 ± 44.4 trials, n = 6 mice; p < 0.05). Both groups however demonstrated equivalent discrimination accuracy scores after 340 trials (90.7 ± 3.7%, vs 91.7 ± 5.9%; p > 0.05) and showed no difference in their mean discrimination times (p > 0.05). For unrewarded familiar odorants, ZnSO4-treated mice showed significantly poorer recognition than sham mice (71 ± 7.4 vs 90 ± 2.8%, p < 0.05) and an initially significant discrimination deficit of rewarded familiar odorants (71.1 ± 7.3 vs 89.8 ± 2.2% p < 0.05; first 20 trials). Comparison of pre- and post-treatment imaging data revealed that the integrity of glomerular-activity patterns observed in the second session correlated with the ability to recognize S+ and S- odors (r = 0.74, p < 0.05, n = 13 mice).
In contrast to previous lesion studies, our data show that perceived odor quality and the ability to distinguish odorants depends critically upon the number and integrity of functional input channels to the OB.