Impaired pain in mice lacking first-order posterior medial thalamic neurons

Abstract

The thalamus plays an important role in sensory and motor information processing by mediating communication between the periphery and the cerebral cortex. Alterations in thalamic development have profound consequences on sensory and motor function. In this study, we investigated a mouse model in which thalamic nuclei formation is disrupted because of the absence of Sonic hedgehog ( Shh ) expression from 2 key signaling centers that are required for embryonic forebrain development. The resulting defects observed in distinct thalamic sensory nuclei in Shh mutant embryos persisted into adulthood prompting us to examine their effect on behavioral responses to somatosensory stimulation. Our findings reveal a role for first-order posterior medial thalamic neurons and their projections to layer 4 of the secondary somatosensory cortex in the transmission of nociceptive information. Together, these results establish a connection between a neurodevelopmental lesion in the thalamus and a modality-specific disruption in pain perception.

Fig. 1.. Disruption of somatosensory thalamic nuclei in adult SBE1/5^−/− mice.

(A, B) Schematic representation of Parvalbumin (PV) expression in the caudal thalamus of control (according to Allen Brain Atlas) and SBE1/5^−/− mice (please also see schematics of relevant axial level in Sup.Fig.1). The area between the midline nuclei and the VPM/L, which lacks PV staining, represents the POm-HO nucleus. (C-E) Quantification of the VPM/VPL (top) and POm-HO (bottom) relative to the size of the thalamus (n=3). (F, G) Schematic representation of Calbindin2 (CALB2) expression at caudal levels of the thalamus in control (according to Allen Brain Atlas) and SBE1/5^−/− mice (please also see schematics of relevant axial level in Sup.Fig.1). CALB2 displays robust expression in the LP and weaker expression in part of the POm. (H-J) Quantification of the CALB2 stained area within the POm (yellow dashed line) relative to the size of the thalamus in control and SBE1/5^−/− mice (n=3). (K, L) Schematic representation of PV expression in more rostral regions of the thalamus in control (according to Allen Brain Atlas) and SBE1/5^−/− mice. PV shows robust expression in RT and less intense expression in VPM/VPL. (M-O) Quantification of the PV-stained area in VPM/VPL relative to the size of the thalamus in control and SBE1/5^−/− mice (n=4, 2 sections/brain). (P, Q) Schematic representation of CALB2 expression at caudal levels of the thalamus in control (according to Allen Brain Atlas) and SBE1/5^−/− mice. At this level, CALB2 is expressed in the LP, LD and CL nuclei. (R-T) Quantification of the CALB2 stained area in the CL relative to the size of the thalamus in control and SBE1/5^−/− mice (n=3). Scale bar: 300μm in C, H, M and R (also applies for D, I, N, S). Data are represented as mean ± SEM. Statistical analysis was performed using a paired, two-tailed, t-test (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001). Abbreviations of thalamic nuclei: ventral posteriomedial (VPM), ventral posteriolateral (VPL), posterior (PO), posterior medial-higher order (POm-HO), lateral posterior (LP), laterodorsal (LD) and central leteral (CL), vental anterior (VA), ventral lateral (VL), dorsal lateral geniculate (DLG), paracentral (PCN), reticular (RT)

Fig. 2.. Behavioral responses to certain noxious stimuli are altered in SBE1/5^−/− mice.

(A) Schematic representation of the mechanical stimulation assay. A high-speed camera was placed 1–2 feet away from mice contained within a rectangular Plexiglas chamber where they could freely perform. Various innocuous and noxious mechanical stimuli were applied to the hind paw of each mouse through a mesh floor. (B-D) quantification of velocity (B), height (C), and Pain Score (composite of orbital tightening, paw shake, paw guard and jumping) (D), for each stimulus (innocuous: CS=cotton swab, DB=dynamic brush and noxious: LP=light pinprick, HP=heavy pinprick) n=12 for SBE1/5^+/+, n=16 for SBE1/5^+/−, n=17 for SBE1/5^−/−. (E) Schematic representation of the thermal assay. Mice were placed onto a thermal plate that increased in temperature from 25°C (room temperature) to 52°C (noxious temperature) at a rate of 1°C per 6 seconds. The temperature at which mice first responded to thermal pain by jumping from the surface was recorded. (F) Quantification of the threshold temperature for control and SBE1/5^−/− mice (n=6 for SBE1/5^+/+, n=16 for SBE1/5^+/−, n=14 for SBE1/5^−/−). (G) Schematic representation of the formalin assay. Mice were injected subcutaneously with a formalin solution into the left hind paw and placed inside a rectangular Plexiglas chamber where they were recorded with a webcam for 60 minutes. (H-I) Quantification of three different parameters (paw licking, paw shaking, paw guarding) for the first phase of the pain response (0–10min) (H), followed by a second phase of the pain response (10–60min) (I), n=5. Statistical analysis was performed using a 2-way ANOVA test for the mechanical stimuli experiments and the non-paired, two-tailed, t-test for the formalin and thermal plate experiments. Data are represented as mean ± SEM (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).

Fig. 3.. Whisker-related nociception is impaired in SBE1/5^−/− mice

(A-B) GAD67 immunostaining in control and SBE1/5^−/− mice (C-D) Barreloids in the VPM marked by GAD67 immunostaining are compromised in SBE1/5^−/− mice. (E-H) Barrel organization within the primary somatosensory cortex marked by VGLUT2 immunostaining is disrupted on coronal sections (E, F) and flattened cortices (G, H) in SBE1/5^−/− mice (n=3). (I) Schematic representation of the capsaicin assay. Mice were injected with a capsaicin solution (1mg/ml) into the right whisker pad under gentle restraint and immediately placed into a beaker where their behavior was recorded for 10 minutes with a high-speed video camera placed 1–2 feet away. Similar recordings were also performed pre-injection. (J-K) Quantification of the number of facial wipes (J) and time spent wiping (K) in control (n=7) and SBE1/5^−/− (n=7) mice prior to (and after capsaicin injection. Statistical analysis was performed using a non-paired, two-tailed, t-test. Data are represented as mean ± SEM (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001) Scale bars: 300μm in A, C and E (also applies for B, D and F). Abbreviations of thalamic nuclei: ventral posteriomedial (VPM), ventral posteriolateral (VPL), reticular (RT), ventromedial (VM)

Fig. 4.. SBE1/5^−/− mice display reduced activation of VPM and POm-FO neurons in response to trigeminal nociception.

(A-E) c-FOS immunostaining in the VPM, co-labeled with Parvalbumin (PV), is increased in control compared to SBE1/5^−/− mice injected with capsaicin (n=3, 2 sections/brain). (F-J) c-FOS immunostaining in the caudal portion of the POm (POm-HO), delineated by the absence of PV staining, is similar between control and SBE1/5^−/− mice (n=3, 2 sections/brain). (K-O) The number of c-FOS⁺ puncta in the region of the POm-FO (defined by anatomical and molecular landmarks) are significantly decreased in SBE1/5^−/− compared to control mice (n=4). Data are represented as mean ± SEM. Statistical analysis was performed using a paired, two-tailed, t-test: * = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001 Scale bars: 300μm in A and B (also applies for C-N). Abbreviations of thalamic nuclei: ventral posteriomedial (VPM), ventral posteriolateral (VPL), posterior (PO), posterior medial-higher order (POm-HO), posterior medial-first order (POm-FO), lateral posterior (LP), laterodorsal (LD) and central leteral (CL), paracentral (PCN), ventral lateral (VL), centromedial (CM), lateral habenula (LHb), medial habenula (MHb)

Fig. 5.. SBE1/5^−/− mice display reduced neuronal activity in specific somatosensory cortical layers in response to trigeminal nociception.

(A-B) c-FOS immunostaining in different cortical layers of S1 from control (A) and SBE1/5^−/− (B) mice treated with capsaicin. (C) Significant differences in the number of c-FOS+ puncta were observed between genotypes in layers 4, 5a and 6 (n=3, 2 sections/brain). (D-E) c-FOS immunostaining in different cortical layers of S2 from control (D) and SBE1/5^−/− (E) mice treated with capsaicin. (F) Significant differences in the number of c-FOS+ puncta were observed between genotypes in layers 4 and 5a (n=3, 1–2 sections/brain). (G-H) c-FOS immunostaining in different cortical layers of the primary motor cortex (M1) from control (G) and SBE1/5^−/− (H) mice treated with capsaicin. (I) No significant differences in the number of c-FOS+ puncta were observed between genotypes (n=3, 1–2 sections/brain). Data are represented as mean ± SEM. Statistical analysis was performed using a paired, two-tailed, t-test (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001).Scale bars: 200μm in A, D and G (also applies for B, E and H).

Fig. 6.. Thalamocortical projections from POm-FO, but not POm-HO, are absent in SBE1/5^−/− mice.

(A-C) Retrograde labeling with the CTB-488 viral vector injected into S1 at Bregma – 1.5mm. The back filled area (white dashed line) in the region of the VPM (co-labeled with PV) was reduced in SBE1/5^−/− compared to control mice after normalizing to thalamic area (n=3). (D-F) Retrograde labeling with the CTB-488 viral vector injected into S1 at Bregma – 2.1mm. A smaller percentage of back fill was detected in the VPM (co-labeled with PV) of SBE1/5^−/− compared to control mice, (n=3, 2 sections/brain). (G-K) Anterograde labeling with AAV1/Syn:GFP injected into POm-HO. Thalamocortical projections into layers 5a and 1 of both S1 and S2 were equivalent between control and SBE1/5^−/− mice (n=3). Co-immunostaining was performed with PV (red) and DAPI (blue) to identify thalamic nuclei and cortical layers, respectively. (L-P) Anterograde labeling with AAV1/Syn:RFP injected into POm-FO. Thalamocortical projections into layer 4 of S2 were present in control (N) but absent in SBE1/5^−/− mice (P) (n=3). Co-immunostaining was performed with CALB2 (green) and DAPI (blue) to identify thalamic nuclei and cortical layers, respectively. Data are represented as mean ± SEM. Statistical analysis was performed using a paired, two-tailed, t-test (** = p ≤ 0.01). Scale bars: 300μm in A, H and M (also applies for C-F, I-K and N-P). Abbreviations of thalamic nuclei: posterior medial-higher order (POm-HO), posterior medial-first order (POm-FO)

Fig. 7.. Schematic model depicting alterations in somatosensory thalamocortical circuits between control and SBE1/5^−/− mice that contribute to differences in their response to acute pain.

Formation of the VPM (gray oblong) in caudal (Bregma: −1.4mm) and rostral (Bregma: −2mm) regions of the thalamus is defective in SBE1/5^−/− mice, as are thalamocortical projections to S1L4 (gray lines), which compromise barrel formation in S1 (gray ovals). The rostral POm, corresponding to POm-FO (red oblong), extends thalamocortical axons to S2L4 (red lines), which fail to form in SBE1/5^−/− mice. The caudal POm, corresponding to POm-HO (green), extends thalamocortical axons to layers 5a and 1 of both S1 and S2, which appear unaffected in SBE1/5^−/− mice. Abbreviations of thalamic nuclei: ventral posteriomedial (VPM), posterior medial-higher order (POm-HO), posterior medial-first order (POm-FO)