The challenge
A central question in neuroscience is how neurons in the primary visual cortex (V1) become selective for specific stimulus orientations. Classical models, proposed by Hubel and Wiesel, suggest that this selectivity emerges from the spatial alignment of thalamic inputs. However, experimental evidence has remained conflicting, particularly in mice, where some studies suggested that orientation selectivity might already be present in thalamic inputs themselves. Resolving this controversy requires direct observation of synaptic activity at the level of individual inputs onto cortical neurons during sensory processing. Yet, technical limitations have historically prevented distinguishing thalamocortical from corticocortical inputs with sufficient spatial and functional precision in vivo.
Our approach
We combined in vivo two-photon glutamate and calcium imaging with optogenetic cortical silencing to identify and functionally characterize individual synapses onto layer 4 neurons in mouse V1. This allowed us to directly distinguish thalamocortical from intracortical inputs and map their functional properties at single-spine resolution.
Our findings
We discovered that thalamocortical synapses provide largely non–orientation-selective input and, strikingly, do not evoke postsynaptic calcium signals. In contrast, corticocortical synapses showed strong orientation tuning and robust calcium responses. These results demonstrate that orientation selectivity is not inherited from thalamic inputs but emerges within cortical circuits. Importantly, the spatial arrangement of thalamic inputs matched the predictions of the classical feedforward model, providing direct experimental validation at the synaptic level.
The implications
These findings redefine how sensory information is transformed in the cortex and highlight distinct synaptic mechanisms underlying cortical computation and plasticity.
Creating SyNergies
By integrating advanced imaging, optogenetics, and circuit-mapping approaches, this study bridges systems neuroscience and synaptic physiology. It provides a mechanistic link between classical theoretical models and modern experimental evidence. The work of Arthur Konnerth and team exemplifies how interdisciplinary strategies can resolve long-standing debates and advance our understanding of cortical computation.
