Mice detect and discriminate temporal patterns of mesoscale optogenetic cortical stimulation

Optogenetic stimulation is widely used to establish causal links between neural activity, behavior, and sensory perception, yet animals may also perceive the neural stimulation itself, a phenomenon termed "optoception." We hypothesized that cortical optogenetic stimulation would generate an artificial percept that could be distinguished from externally evoked sensations and characterized psychometrically in an operant task. We tested this hypothesis using mesoscale cortical photostimulation delivered by a 3 mm, 459 nm LED implanted over auditory cortex. Four mice expressing Channelrhodopsin-2 (ChR2) in excitatory neurons learned a head-fixed go/no-go task in which 20 Hz cortical photostimulation served as the target. ChR2-positive mice detected cortical photostimulation relative to LED-off catch trials, distinguished it from both an external LED matched in wavelength and pulse rate and from an acoustic noise matched in amplitude-modulation rate, and discriminated between 20- and 10-Hz cortical photostimulation. Reducing implanted-LED optical power initially lowered rate-discrimination sensitivity, which then increased across training sessions, demonstrating perceptual learning. In two mice tested with 11 randomly interleaved rates spanning 10-20 Hz, lick probability followed a saturating rate-dependent function. Two ChR2-negative littermates failed to acquire photostimulation-locked behavioral responding. Together, these findings show that mesoscale cortical optogenetic stimulation generates a graded, learnable artificial percept whose behavioral readout depends on photostimulation rate and optical power. This single-LED preparation provides a cost-efficient and wavelength-extensible platform for psychometric analysis of optoception.