Anatomically and functionally distinct thalamocortical inputs to primary and secondary mouse whisker somatosensory cortices

Abstract

Subdivisions of mouse whisker somatosensory thalamus project to cortex in a region-specific and layer-specific manner. However, a clear anatomical dissection of these pathways and their functional properties during whisker sensation is lacking. Here, we use anterograde trans-synaptic viral vectors to identify three specific thalamic subpopulations based on their connectivity with brainstem. The principal trigeminal nucleus innervates ventral posterior medial thalamus, which conveys whisker-selective tactile information to layer 4 primary somatosensory cortex that is highly sensitive to self-initiated movements. The spinal trigeminal nucleus innervates a rostral part of the posterior medial (POm) thalamus, signaling whisker-selective sensory information, as well as decision-related information during a goal-directed behavior, to layer 4 secondary somatosensory cortex. A caudal part of the POm, which apparently does not receive brainstem input, innervates layer 1 and 5A, responding with little whisker selectivity, but showing decision-related modulation. Our results suggest the existence of complementary segregated information streams to somatosensory cortices.

Fig. 1. Whisker somatosensory thalamic nuclei and their cortical projections revealed through AAV-mediated anterograde trans-synaptic gene expression.

a AAV1 viral vector was injected in Pr5 of the brainstem to express Cre-recombinase in a trans-synaptic anterograde manner. A second AAV injection in the thalamus expressing a Cre-dependent tdTomato fluorescent protein resulted in labeling of VPM neurons receiving direct inputs from Pr5. Left: schematic of the injection protocol. Middle: Example coronal section with VPM neurons expressing tdTomato in comparison to a reference atlas (distance from bregma indicated). Right: Axonal innervation of VPM neurons in wS1. This experiment was repeated in four mice with similar results. b Same as a, but for POm neurons receiving direct inputs from Sp5i. This experiment was repeated in ten mice with similar results. c Same as b, but for POm neurons not expressing Cre-recombinase through trans-synaptic transfection from Sp5i injections. Here, the second viral vector injected in the thalamus only allowed expression of eYFP conditionally on the absence of Cre-recombinase. This experiment was repeated in seven mice with similar results. d Examples of laminar-specific axonal innervation in somatosensory cortices originating from different thalamic nuclei. e Normalized fluorescent expression profile averaged over mice (n = 4 mice for VPM first-order (VPM-FO), n = 10 mice for POm first-order (POm-FO), n = 7 mice for POm higher-order (POm-HO)). Shaded areas: s.e.m. f Schematic of the different somatosensory thalamocortical circuits. The schematic drawings of the brain in panels a–c are reproduced from Paxinos and Franklin (2001) with permission from Elsevier.

Fig. 2. Ex vivo whole-cell recordings in brain slices of POm-FO and POm-HO inputs to excitatory neurons across layers in wS2.

a Schematic showing the strategy used to selectively express ChR2-eYFP in FO or HO subdivisions of POm. Method used to activate thalamocortical axons expressing channelrhodopsin-2 in order to evoke postsynaptic potentials in the somatosensory cortex is illustrated. A 470 nm wavelength light was delivered with a LED light source coupled with a 1 mm optic fiber onto wS2. Inset: Two-photon microscopy image of an in vitro whole-cell patch–clamp recording of two neurons filled with Alexa 594. b Confocal z-projection of wS2 in a parasagittal slice after fixation. ChR2-eYFP was expressed in POm-FO axons, and recorded neurons were filled with biocytin followed by staining with streptavidin conjugated to Alexa 647. This experiment was repeated in three mice with similar results. c Light-evoked excitatory postsynaptic potentials (EPSPs) from recorded neurons labeled in b following 1 ms light pulses. d Top: Peak amplitude of EPSPs evoked by optogenetic stimulation of POm-FO axons recorded in cortical excitatory neurons (N = 3 mice, n = 39 neurons) across different layers in wS2. On each box, central mark indicates the median and edges indicate 25th and 75th percentiles. The whiskers extend from the minimum data point comprised within 1.5× of the interquartile range to the 25th percentile and from the maximum data point comprised within 1.5× of the interquartile range to the 75th percentile. Bottom: same responses normalized to the average peak EPSP recorded in the main input layer (L4) for each experiment. e–g Same as b–d but for POm-HO axon stimulation during whole-cell recording of neurons in wS2 (N = 3 mice with similar results, n = 34 neurons). Here, EPSPs were normalized to the average peak EPSPs from L5A neurons for each experiment.

Fig. 3. Calcium imaging of thalamocortical axons during passive whisker stimulation.

a Schematic of viral vector injection and microprism implantation for in vivo calcium imaging of thalamocortical axons expressing GCaMP6s. b Intrinsic optical signal in wS1 during whisker stimulation. This experiment was repeated in 14 mice with similar results. c Two-photon image of VPM axons in wS1. Similar results were observed in three mice. d Left: Example of a region of interest for axonal segments with high calcium signal correlation in wS1 (red). Right: Corresponding calcium responses to C2 or B2 whisker stimulation in z-score across trials. The average calcium responses are shown below. e Normalized calcium responses averaged over all axons for each population (3 mice, n = 62 axons for VPM-FO; 3 mice, n = 101 axons for POm-FO; 8 mice, n = 86 axons for POm-HO). Inset: Response latency comparison (Kruskal–Wallis two-sided test with Bonferroni correction, **p = 0.008 for VPM-FO vs. POm-FO, **p = 0.004 for VPM-FO vs. POm-HO, p = 1 for POm-FO vs. POm-HO). Boxplot: central mark indicates the median and edges indicate 25th and 75th percentiles. Whiskers extend to the largest or smallest point comprised within 1.5× of the interquartile range from both edges. f Left: Two-photon image of the calcium response (ΔF/F) for C2 (green) and B2 (red) whisker stimulation overlaid on top of VPM axonal innervation (gray) in wS1. Middle: Distribution of whisker selectivity indices for the corresponding axonal population (3 mice, n = 62 axons). Right: Distribution for the absolute value of the whisker selectivity index. Red arrow: average value. g Same as f, but for POm-FO axons in wS2 (3 mice, n = 101 axons). Inset in the right panel: Response latency comparison between tuned (|WSI | ⩾0.75) and untuned (|WSI | < 0.75) axons (Kruskal–Wallis two-sided test, **p = 0.002). Boxplot statistics same as in (e) inset. h Same as f, but for putative POm-HO axons located in layer 1 (8 mice, n = 86 axons). Kruskal–Wallis test with Bonferroni correction comparing |WSI | : p = 0.003 for VPM-FO vs. POm-FO; p = 4 × 10⁻¹¹ for VPM-FO vs. POm-HO; and p = 4 × 10⁻¹¹ for POm-FO vs. POm-HO.

Fig. 4. Distinct sensory information in parallel thalamocortical pathways during goal-directed behavior.

a Schematic of a mouse performing a two-whisker discrimination task, and the average lick rate over all imaging sessions (>4 days of training) for all stimulus conditions across mice (mean lick rates over n = 14 mice, Kruskal–Wallis test with Bonferroni correction: p = 0.02 for C2 vs. B2; p = 2 × 10⁻⁸ for C2 vs. No stim; and p = 0.009 for B2 vs. No stim). Boxplot: central mark indicates the median and edges indicate 25th and 75th percentiles. Whiskers extend to the largest or smallest point comprised within 1.5× of the interquartile range from both edges. b Calcium responses (z-score) for an example VPM-FO axon during lick trials upon C2 or B2 whisker stimulation. Trials are ordered according to lick reaction times, which are shown with a white line on color maps. Average responses are shown below for lick and no-lick conditions. c Same as b, but for a POm-FO axon. d Left: Calcium responses averaged over all VPM-FO axons with significant responses to whisker stimuli during lick and no-lick trials, normalized to the no-lick condition. Dark lines: mean value and shaded areas: s.e.m. Middle: Early phase of the response over the first 0.4 s. Right: Comparison of the response amplitude between lick and no-lick conditions averaged over gray area (0–0.266 s) in middle panel (Wilcoxon paired two-sided test, p = 0.94, N.S. not significant). e, f Same as d, but for POm-FO axons (***p = 3 × 10⁻⁵ for (e)) and POm-HO layer 1 axons (***p = 1 × 10⁻⁵ for (f)), respectively. g Distributions of Pearson correlation coefficient between reaction times and calcium response latencies for all axons with significant responses in lick trials (n = 169 for VPM-FO, n = 264 for POm-FO, n = 281 for POm-HO layer 1 axons). Colored bars: Pearson coefficient with p < 0.05. h Distributions of whisker selectivity index absolute values comparing all axonal populations and lick/no-lick conditions for axons with significant sensory responses in no-lick condition (Kruskal–Wallis two-sided test with Bonferroni correction; VPM-FO: ***p = 1 × 10⁻¹² for lick vs. no-lick, POm-FO: p = 0.32 for lick vs. no-lick N.S. not significant, POm-HO: ***p = 3 × 10⁻¹⁰ for lick vs. no-lick, *p = 0.034 for VPM-FO vs. POm-FO in lick condition, ***p = 5 × 10⁻¹⁵ for POm-FO and POm-HO in lick condition). Boxplot statistics as in (a).