Previous studies demonstrate that visuo-auditory associative memory is established in the auditory cortex of behaving rats using a classical conditioning protocol. This cross-modal association is blocked by local infusion of a cholecystokinin (CCK) antagonist. Local infusion of a CCK agonist enables auditory cortical neurons to respond to a light stimulus paired with an auditory stimulus in anesthetized rats. In the present study, we explored the mechanisms underlying this visuo-auditory association in the auditory cortex.

First, we observed that even without participation of the auditory thalamus, which was silenced by 6,7-dinitroquinoxaline-2,3-dione (DNQX) application, auditory cortical neurons that previously did not respond to light stimuli exhibited responses after pairing with electrical stimulation in the auditory cortex after CCK infusion into the auditory cortex. Retrograde tracing with cholera toxin subunit B injected into the auditory cortex revealed many retrogradely labeled neurons in both the associative visual cortex and primary visual cortex. By contrast, labeled neurons were rarely observed in the visual thalamus (i.e., lateral geniculate nucleus). Field excitatory post-synaptic potentials evoked by laser stimulation of projection terminals in the auditory cortex from the visual cortex via expression of ChrimsonR (a variant of ChR2) were potentiated after pairing with acoustic stimuli following local infusion of CCK. These results suggest that projections from the visual cortex to the auditory cortex provide a direct anatomical basis for the visuo-auditory association observed in the auditory cortex.

Second, with endogenous CCK release in the auditory cortex from high-frequency optical stimulation of CCK⁺ projections from the entorhinal cortex to the auditory cortex (ENT→AC), inputs from the visual cortex to the auditory cortex (VC→AC) were significantly potentiated after pairing of laser stimulation of projection terminals in the auditory cortex from the visual cortex and auditory stimuli, which induced action potential firing in the auditory cortex. These results demonstrate that cross-modal cortical potentiation can be replicated with endogenous release of CCK.

Third, the potentiation of VC→AC inputs decreased as 1) the frequency of laser stimulation of ENT→AC CCK⁺ projections decreased, 2) the delay between the termination of high-frequency laser stimulation of ENT→AC CCK⁺ projections and pre- and post-synaptic pairing increased, and 3) the delay between activation of VC→AC projections (i.e., pre-synaptic activation) and natural auditory cortex activation (i.e., post-synaptic activation) increased.