Learning-dependent functional increase in connectivity among the posterior striatum, lateral geniculate nucleus, and visual cortex in the visual discrimination task

Animals can find important objects automatically and choose proper actions quickly as they repeatedly experience the same situations. Recent studies have suggested that the posterior part of the striatum (pStr) which is innervated from various sensory areas is involved in finding familiar objects automatically and quickly. The next question to be answered is how pStr neurons change their activities and contribute to finding objects quickly through learning in interaction with other regions. In this study, we focused on visual processing and approached the above question by examining the interaction changes among pStr, the visual thalamus (dorsal lateral geniculate nucleus; dLGN) and cortex (visual cortex; VC) through learning. We hypothesized that the functional connectivity from dLGN and VC to pStr were strengthened after learning so that the shortcut pathway via pStr could convey task-related information quickly to the downstream. To test this hypothesis, we recorded the neuronal activities extracellularly using silicon probes from pStr, dLGN and VC simultaneously while rats learned a visual discrimination task in T-maze. We found that pStr neurons fired more after learning, especially in the beginning of the image presentation zone, suggesting pStr is involved in the quick recognition of task-related stimuli. In terms of pStr orchestration with dLGN and VC, 4 Hz coherence between regions got stronger after learning, supporting out hypothesis. Our results indicate that this dLGN/VC-pStr functional connectivity increase plays a key role on automatic recognition of familiar objects.