Optogenetic fMRI (ofMRI) is an emerging technology that enables global mapping of the brain neural responses caused by cell-type and spatiotemporal specific neuromodulation¹. Since its introduction, it has been used to map and probe large-scale neural networks and its underlying dynamics^{1, 2} driven by optogenetic stimulation of excitatory neuronal populations. Taking advantage of this capability, we here study the dynamics of the rodent thalamocortical whisker-related somatosensory circuit. The rodent somatosensory circuit serves as an excellent model in advancing our understanding of thalamocortical information processing due to its well-structured network architecture³. The ascending whisker-related somatosensory circuit that terminates in the somatosensory barrel cortex (S1BF) is divided into two main processing pathways⁴, the lemniscal pathway (via ventral posteromedial thalamus (VPM)) and the paralemniscal pathway (via posteromedial complex of the thalamus (POm)). The subdivisions were identified by anterograde axonal tracing from the thalamus⁵ and revealed through electrophysiological studies⁶. However, many electrophysiological studies are still undertaken to delineate the differing roles of both pathways in processing somatosensory information and its significance in the process of multisensory and sensorimotor interactions due to the sheer scale and level of interaction within and beyond the circuit. Hence, our present study aims to map the large-scale activity of the paralemniscal pathway. This coupled with our previous preliminary ofMRI study⁷ on the lemniscal system presents an added dimension to current efforts in understanding how information is relayed within the thalamocortical circuit.

Animal preparation and optogenetic stimulation: AAV5-CaMKIIa::ChR2(H134R)-mCherry was injected to POm of adult rats (200-250g, male, SD strain, n=6). Four weeks after injection, an optic fiber (diameter 450μm) was implanted at the injection site (Figure 1a) to deliver optical stimulation. Blue (473nm) light was presented to animals expressing ChR2 at two different frequencies, 5 and 10 Hz (pulse width: 30% duty cycle, light irradiance: 40mW/mm²). The two stimulation frequencies were chosen to correspond within the range of rodent whisking frequency^{6, 8, 9}. All optical stimulation was presented in a block-design paradigm (20s on and 60s off) (Figure 1b). Opaque tape was used to ensure no light leakage causing visual stimulus.

fMRI acquisition and analysis: fMRI data was acquired at 7T using GE-EPI (FOV=32×32mm2, matrix=64×64, α=56°, TE/TR=20/1000ms, 10 contiguous slices with 1mm thickness). Data were preprocessed before standard GLM analysis was applied to identify significant BOLD responses (p<0.001, FDR corrected). BOLD time courses were extracted from anatomically defined ROI (Figures 2a and 3a).

Figure 2 shows the BOLD responses for excitation of the POm via ChR2 at 5 and 10Hz. Stimulation at 5 Hz evoked positive BOLD responses in the ipsilateral somatosensory barrel cortex (S1BF), bilateral visual cortex (VC) and bilateral superior colliculus (SC). Stimulation at 10 Hz evoked positive BOLD responses only in ipsilateral S1BF.

Figure 3 shows the BOLD responses for excitation of the VPM via ChR2 at 5 and 10 Hz. Stimulation at 5 and 10 Hz evoked positive BOLD responses in the ipsilateral S1BF, ipsilateral secondary somatosensory cortex (S2), ipsilateral motor cortex (MC), and cingulate cortex.

Previous electrophysiological and behavioral studies^{6, 8, 10} postulated that POm and VPM relay different somatosensory information to S1BF, potentially affecting other cortical and subcortical regions. However, no conclusive evidence of such phenomena has been provided so far. Here, our ofMRI study revealed differences when POm and VPM were stimulated within the natural whisking frequency (5-10 Hz). This clearly demonstrated that somatosensory information is relayed differently within these two pathways. VPM stimulation activated the somatosensory and motor cortical regions (5 and 10 Hz) whereas POm stimulation activated the somatosensory cortical and visual cortical and subcortical regions (mainly at 5 Hz). The contribution of each thalamic nucleus to multisensory integration during natural whisking is unknown as both pathways have thalamocortical axonal projections to motor cortex and also sparse long-range cortico-cortical projections from S1BF to the visual regions^{5, 11}. Our findings indicate that VPM/lemniscal pathway plays a larger role in sensorimotor integration whereas POm/paralemniscal system has a larger responsibility in somatosensory-visual integration. Lastly, the frequency-dependent responses observed during POm stimulation are in line with electrophysiological studies suggesting the sensitivity of POm in detecting whisker movement frequencies^{6, 8}. In summary, our ofMRI study directly demonstrated the differences between paralemniscal and lemniscal thalamocortical pathways. It provided insights into the multi-synaptic activity critical for multisensory interactions (sensorimotor and somatosensory-visual). Future studies will be expanded in several directions, e.g., incorporating electrophysiological recordings and characterizing functional implications of thalamocortical stimulation.