A visual cortical-lateral posterior thalamic nucleus circuit regulates depressive-like behaviors in male mice

Abstract

Depression, a prevalent psychiatric disorder of ambiguous etiology and high heterogeneity, has been recently linked to the primary visual cortex (V1). However, the precise circuits mediating the impact of V1 on depressive-like behaviors are poorly understood. Here, we demonstrate that the V1, specifically the lateral posterior nucleus of the thalamus (LP)-projecting V1 glutamatergic subpopulation (Glu^V1→LP neurons), shows reduced activity after chronic restraint stress (CRS) in male mice, leading to depressive-like behaviors. Optogenetic or chemogenetic activation of these neurons ameliorated depressive-like behaviors in CRS-depressed mice, whereas reducing activity exacerbated these behaviors. This reduction in Glu^V1→LP neurons activity was predominantly due to a decrease in the guanine nucleotide-binding protein subunit gamma-4 (Gγ4). Overexpression of Gγ4 in the Glu^V1→LP neurons produced antidepressant-like effects, suggesting that Gγ4 is a crucial regulator of mood. Collectively, these results reveal a V1→LP circuit that modulates depressive-like behaviors, suggesting potential targets for therapeutic interventions.

Fig. 1. Reduced neuronal activation in the primary visual cortex (V1) of chronic restraint stress (CRS)-exposed mice is associated with depressive-like behaviors.

a Experimental schematic. C57BL/6 J mice were subjected to 1-day acute restraint stress (ARS) or 21 consecutive days of CRS, followed by an examination of depressive-like behaviors. b, c Sucrose intake percentage and total liquid intake in the sucrose preference test (SPT). Control vs. CRS, P = 0.0001; ARS vs. CRS, P = 0.0003. d Immobile duration in the forced swimming test (FST). Control vs. CRS, P = 0.0004; ARS vs. CRS, P = 0.0147. e Total distance traveled in the open field test (OFT). f, g Representative images and quantification of c-Fos⁺ (green) cells in the different layers of the V1 in control, ARS, and CRS mice. Total, Control vs. CRS, P = 0.0020; ARS vs. CRS, P = 0.0067. L6, Control vs. CRS, P < 0.0001; ARS vs. CRS, P < 0.0001. L5, Control vs. CRS, P = 0.0015; ARS vs. CRS, P = 0.0125. Blue, DAPI staining. Scale bar, 100 µm. Low magnification on the right. Scale bar, 200 µm. Correlation analysis between c-Fos⁺ cells and sucrose preference (h) and immobility duration in the FST (i) of control, ARS, and CRS mice. In (b–e, g), n = 8, 7, 9 mice for Control, ARS, CRS respectively. *p < 0.05, **p < 0.01, ***p < 0.001 by one-way analysis of variance (ANOVA) with Tukey’s post-hoc test (b–e, g); Pearson’s correlation test (h, i); ns, not significant; data represent the mean ± standard error of the mean (SEM). See also Supplementary Fig. 1. Source data are provided as a Source Data file.

Fig. 2. Chemogenetic and optogenetic activation of V1 neurons improves CRS-induced depressive-like behaviors in mice.

a Schematic of injection of the hSyn-hM3D(Gq) AAV viral vector into the V1 region of C57BL/6 J mice to achieve in vivo chemogenetic manipulation. b Representative images showing mCherry expression (red) in the V1 4 weeks after viral injection. Blue, DAPI staining. Scale bar, 200 µm. c Representative images showing c-Fos⁺ (cyan) cells in the V1 after chemogenetic activation. CNO, clozapine N-oxide. Red, mCherry expression. Blue, DAPI staining. Scale bar, 100 µm. Central inset, magnified view of rectangular part; scale bar, 10 μm. d The normalized absolute intensity of c-Fos immunostaining in hM3D(Gq)/mCherry⁺ neurons after the saline or CNO injection in adult C57BL/6 J mice (a.u., arbitrary units). P < 0.0001, n = 5, 4 mice for Saline, CNO respectively. e Behavioral timeline of CRS induction, chemogenetic manipulation, and assessment of depressive-like phenotypes. Data showing the effect of chemogenetic activation of neurons and the CRS paradigm on FST (f) and SPT (g) results. f Control + Pre vs. CRS + Pre, P < 0.0001; CRS + Pre vs. CRS + Post, P = 0.0003; g Control + Pre vs. CRS + Pre, P < 0.0001; CRS + Pre vs. CRS + Post, P = 0.0010 (n = 8 mice/group). h Schematic of hSyn-ChR2 AAV viral vector injection and fiber implantation in the V1. i Representative images showing the locations of fiber tips and ChR2 expression in the V1 4 weeks after viral injection. Scale bar, 200 µm. j Light-evoked action potentials (APs) from a current-clamped neuron infected with ChR2 showing the effect of optogenetic manipulation by 470-nm blue light in the V1 (n = 5 neurons from 3 mice). k Behavioral timeline of CRS induction, optogenetic manipulation, and assessment of depressive-like phenotypes. Data showing the effect of optogenetic activation of neurons and the CRS paradigm on FST (l) and SPT (m) results. l Control + Pre vs. CRS + Pre, P < 0.0001; Control + Pre vs. Control + Optic-Stim, P = 0.0226; CRS + Pre vs. CRS + Optic-Stim, P = 0.0009; m Control + Pre vs. CRS + Pre, P = 0.0062; CRS + Pre vs. CRS + Optic-Stim, P = 0.0036 (n = 8, 9 mice for Control, CRS respectively). ***p < 0.001 by unpaired two-tailed Student’s t-test (d); *p < 0.05, **p < 0.01, ***p < 0.001 by two-way repeated-measures ANOVA with Bonferroni’s post-hoc test (f, g, l, m); data represent the mean ± SEM. See also Supplementary Figs. 2 and 3. Source data are provided as a Source Data file.

Fig. 3. Inhibition of V1 glutamatergic neurons facilitates depressive-like behaviors, whereas activation results in antidepressant-like effects.

a, b c-Fos expression (green) and quantification in immunofluorescence-labeled V1 CaMKIIα⁺ (red) neurons after control or CRS paradigm (P < 0.0001, n = 21 slices/group). a Arrowheads indicate double-labeled cells. Blue, DAPI staining. Scale bar, 25 μm. Enlarged regions on the right. Scale bar, 10 μm. c Fiber photometry setup and GCaMP6s expression in V1 CaMKIIα neurons. Scale bar, 200 μm (left, low magnification), 50 μm (right, high magnification). Representative in vivo calcium fluorescence traces (d) and quantification of calcium peak number (e) demonstrating reduced spontaneous calcium signals (ΔF/F) of CaMKIIα neurons CRS group mice compared with control (Control vs. CRS, P < 0.0001; ARS vs. CRS, P < 0.0001; n = 10 mice/group). Gray circles indicate transient peaks. f, g Schematic of chemogenetic manipulation of CaMKIIα neurons in the V1. Data showing the effect of inhibiting CaMKIIα neurons on FST (h) and SPT (i) results. h P = 0.0016; i P = 0.0019 (n = 10 mice/group). Schematic of hM3D(Gq) AAV viral vector injection into the V1 CaMKIIα neurons (j) and the behavioral timeline of CRS induction, chemogenetic manipulation, and assessment of depressive-like phenotypes (k). Data showing the effect of activating CaMKIIα neurons on FST (l) and SPT (m) results. l Control + Pre vs. CRS + Pre, P = 0.0012; CRS + Pre vs. CRS + Post, P = 0.0023; m Control + Pre vs. CRS + Pre, P = 0.0019; CRS + Pre vs. CRS + Post, P = 0.0001 (n = 10, 9 mice for Control, CRS respectively). **p < 0.01, ***p < 0.001 by unpaired or paired two-tailed Student’s t-test (b, h, i), one-way analysis of variance (ANOVA) with Tukey’s post-hoc test (e), two-way repeated-measures ANOVA with Bonferroni’s post-hoc test (l, m); data represent the mean ± SEM. See also Supplementary Fig. 4–7. Source data are provided as a Source Data file.

Fig. 4. Optogenetic activation of Glu^V1→LP neurons reduces immobility time in the forced swimming test in naïve mice and improves depressive-like behaviors in CRS-depressed mice.

a, b Schematic of anterograde tracing of V1 neurons in adult C57BL/6 J mice (a). Coronal sections showing representative images of virus-infected neurons (b; b1 and b1’: in the V1, b2: in the LP, b3: in the LD, b4: in the SC). LP, lateral posterior thalamic nucleus; LD, laterodorsal thalamic nucleus; SC, superior colliculus. Blue, DAPI staining. Scale bar, 200 μm (b1, b2, b3, b4), 25 μm (b1’). c Schematic of CaMKIIα-ChR2 AAV viral vector injection and fiber implantation in V1. d Representative images showing the locations of fiber tips in the LP and ChR2 expression in the V1 4 weeks after viral injection. Scale bar, 100 µm. e Schematic of optogenetic manipulation and assessment of depressive-like phenotypes. Data showing the effect of activating CaMKIIα neurons on FST (f) and SPT (g) results. f ChR2 + Pre vs. ChR2 + Optic-Stim, P = 0.0044; mCherry + Optic-Stim vs. ChR2 + Optic-Stim, P = 0.0036 (n = 6, 7 mice for mCherry, ChR2 respectively). h Behavioral timeline of CRS induction, optogenetic manipulation and assessment of depressive-like phenotypes. Data showing the effect of activating CaMKIIα neurons on FST (i) and SPT (j) results. i mCherry + Baseline vs. mCherry + Pre, P = 0.0057; ChR2 + Baseline vs. ChR2 + Pre, P = 0.0026; mCherry + Optic-Stim vs. ChR2 + Optic-Stim, P = 0.0033; ChR2 + Pre vs. ChR2 + Optic-Stim, P = 0.0009; j mCherry + Baseline vs. mCherry + Pre, P = 0.0144; ChR2 + Baseline vs. ChR2 + Pre, P = 0.0005; mCherry + Optic-Stim vs. ChR2 + Optic-Stim, P < 0.0001; ChR2 + Pre vs. ChR2 + Optic-Stim, P < 0.0001 (n = 6, 7 mice for mCherry, ChR2 respectively). k, l Schematic of retrograde tracing of V1^LP neurons in adult C57BL/6 J mice (n = 3 mice/group) and representative images of virus-infected V1^LP neurons and terminals in the LP; Blue, DAPI staining. Scale bar, 500 µm (left, low magnification), 200 µm (right, high magnification). m Schematic of V1 injection of AAV-hSyn-ChR2-mCherry in C57BL/6J mice and whole-cell recordings in LP neurons in acute brain slices. n, o Representative traces and quantification of the frequency of light-evoked AP traces in LP neurons before (pre), during (light-on), and after (post) blue-light treatment (Pre vs. Light-ON, P = 0.0139; Light-ON vs. Post, P = 0.0041; n = 12 cells from 4 mice/group). p Representative traces and quantification of the amplitudes of light-evoked postsynaptic currents in LP neurons before and after CNQX treatment (P = 0.0088, n = 10 cells from 4 mice/group). q Representative traces and quantification of the amplitudes of light-evoked postsynaptic currents in LP neurons before and after TTX, TTX + 4-AP treatment (ACSF vs. TTX, P = 0.0050; TTX vs. TTX + 4-AP, P = 0.0016; n = 10 cells from 2 mice/group). *p < 0.05, **p < 0.01 by paired two-tailed Student’s t-test (p), one-way repeated-measures ANOVA followed by Bonferroni’s post-hoc test (o, q), two-way repeated-measures ANOVA with Bonferroni’s post-hoc test (f, g, i, j); data represent the mean ± SEM. See also Supplementary Figs. 8 and 9. Source data are provided as a Source Data file.

Fig. 5. Chemogenetic activation of Glu^V1→LP neurons in the V1 alleviates depressive-like behaviors in mice, while inhibition exacerbates depressive-like behaviors.

a Schematic and representative image of retrograde tracing of Glu^V1→LP neurons in adult C57BL/6 J mice. Scale bar, 100 µm. b, c Representative images and quantification of the percentage of c-Fos⁺ (cyan)/mCherry⁺ (red) cells in the V1 mCherry⁺ population after the control or CRS paradigm (P = 0.0053; n = 6 mice/group). b Arrowheads indicating c-Fos⁺mCherry⁺ double-labeled cells. Gray, DAPI staining. Scale bar, 50 µm. Enlarged regions on the right. Scale bar, 25 μm. d, e Schematic of chemogenetic activation of V1^LP CaMKIIα neurons in adult C57B/L6J mice and behavioral timeline of chemogenetic manipulation and assessment of depressive-like phenotypes. Data showing the effect of activating Glu^V1→LP neurons on FST (f) and SPT (g) results (n = 10, 8 mice for Saline, CNO respectively). f Saline + Post vs. CNO + Post, P = 0.0341. Behavioral timeline of CRS induction, chemogenetic manipulation, and assessment of depressive-like phenotypes (h). Data showing the effect of activating Glu^V1→LP neurons on FST (i) and SPT (j) results after the CRS paradigm. i Saline + Baseline vs. Saline + Pre, P < 0.0001; CNO + Baseline vs. CNO + Pre, P = 0.0005; Saline + Post vs. CNO + Post, P < 0.0001; CNO + Pre vs. CNO + Post, P = 0.0029; j Saline + Baseline vs. Saline + Pre, P = 0.0002; CNO + Baseline vs. CNO + Pre, P = 0.0029; Saline + Post vs. CNO + Post, P = 0.0024; CNO + Pre vs. CNO + Post, P = 0.0017 (n = 10, 8 mice for Saline, CNO respectively). k, l Schematic of chemogenetic inhibition of V1^LP CaMKIIα neurons in adult C57B/L6J mice and behavioral timeline of chemogenetic manipulation and assessment of depressive-like phenotypes. Data showing the effect of inhibiting Glu^V1→LP neurons on FST (m) and SPT (n) results. m Saline + Post vs. CNO + Post, P = 0.0146; n Saline + Post vs. CNO + Post, P = 0.0425; CNO + Pre vs. CNO + Post, P = 0.0040 (n = 8, 10 mice for Saline, CNO respectively). Behavioral timeline of SCRS induction, chemogenetic manipulation, and assessment of depressive-like phenotypes (o). Data showing the effect of inhibiting Glu^V1→LP neurons on FST (p) and SPT (q) results after the SCRS paradigm. p Saline + Post vs. CNO + Post, P = 0.0187; CNO + Baseline vs. CNO + Post, P < 0.0001; q Saline + Post vs. CNO + Post, P = 0.0001; CNO + Baseline vs. CNO + Post, P < 0.0001 (n = 12, 14 mice for Saline, CNO respectively). Behavioral timeline of CRS induction, chemogenetic manipulation, and assessment of depressive-like phenotypes (r). Data showing the effect of inhibiting Glu^V1→LP neurons on FST (s) and SPT (t) results after the CRS paradigm. s Saline + Baseline vs. Saline + Pre, P < 0.0001; CNO + Baseline vs. CNO + Pre, P < 0.0001; Saline + Post vs. CNO + Post, P = 0.0121; CNO + Pre vs. CNO + Post, P = 0.0166; t Saline + Baseline vs. Saline + Pre, P = 0.0005; CNO + Baseline vs. CNO + Pre, P < 0.0001; Saline + Post vs. CNO + Post, P = 0.0006; CNO + Pre vs. CNO + Post, P = 0.0041 (n = 8, 10 mice for Saline, CNO respectively). *p < 0.05, **p < 0.01, ***p < 0.001 by unpaired two-tailed Student’s t-test (c), two-way repeated-measures ANOVA with Bonferroni’s post-hoc test (f, g, i, j, m, n, p, q, s, t); data represent the mean ± SEM. Source data are provided as a Source Data file.

Fig. 6. Knockdown of Gγ4 in V1 excitatory neurons exacerbates depressive-like behaviors in chronic stressed mice, decreases neuronal excitability and enhances inhibitory synaptic transmission.

a, b Representative images and quantification of Gγ4/β-actin in the V1 after the control or CRS paradigms (P = 0.0386, n = 6 mice/group). Correlation between relative Gγ4 protein level and FST immobility time (c) and sucrose intake preference (d) in control and CRS mice (n = 6 mice/group). Representative images (e) of Gγ4 (green) and CaMKIIα (red) and a pie graph of the percentage of Gγ4⁺ population with CaMKIIα⁺ and CaMKIIα^- cells in naïve mice (f) (n = 3 mice/group). Blue, DAPI staining. Scale bar, 50 µm.Representative images (g) and quantification of the fluorescence intensity of Gγ4 (magenta) in the CaMKIIα⁺ (green) population (h) after the 21-day CRS paradigm (P = 0.0005, n = 6 mice/group). Blue, DAPI staining. Scale bar, 50 µm. i Schematic and representative image of injection of the CaMKIIα-shRNA(Gng4) AAV viral vector into the V1 region of C57BL/6 J mice to achieve in vivo knockdown manipulation. Scale bar, 100 µm. Behavioral timeline of CRS induction and assessment of depressive-like phenotypes (j). Data showing the effect of knockdown of Gγ4 in V1 excitatory neurons on SPT (k) and FST (l) results after the CRS paradigm. k shRNA(scramble) + Baseline vs. shRNA(scramble) + CRS, P = 0.0151; shRNA(Gng4) + Baseline vs. shRNA(Gng4) + CRS, P < 0.0001; shRNA(scramble) + CRS vs. shRNA(Gng4) + CRS, P = 0.0028; l shRNA(scramble) + Baseline vs. shRNA(scramble) + CRS, P = 0.0245; shRNA(Gng4) + Baseline vs. shRNA(Gng4) + CRS, P = 0.0201; shRNA(scramble) + Baseline vs. shRNA(Gng4) + Baseline, P = 0.0227; shRNA(scramble) + CRS vs. shRNA(Gng4) + CRS, P = 0.0112 (n = 10, 9 mice for shRNA(scramble), shRNA(Gng4) respectively). m, n Representative images and quantification of the percentage of c-Fos⁺ (green)/mCherry⁺ (red) cells in the V1 mCherry⁺ population after viral injection in control or CRS mice (shRNA(scramble) + Control vs. shRNA(scramble) + CRS, P = 0.0484; shRNA(Gng4) + Control vs. shRNA(Gng4) + CRS, P < 0.0001; shRNA(scramble) + CRS vs. shRNA(Gng4) + CRS, P < 0.0001; n = 5 mice/group). o Comparison of the AP frequency in shRNA(scramble) AAV-infected mice and shRNA(Gng4) AAV-infected mice. Upper-left panel: step depolarizing currents; upper-right panel: representative sample traces induced by a 340-pA depolarization current; lower panel: quantification of AP numbers (shRNA(scramble) vs. shRNA(Gng4), 240 pA to 340 pA, P < 0.05; n = 17, 22 cells for shRNA(scramble), shRNA(Gng4) respectively). p–r Representative trace of mIPSCs in V1 mCherry⁺ neurons and comparison of the mIPSC frequency and amplitude in shRNA(scramble) AAV infected mice and shRNA(Gng4) AAV infected mice (P = 0.0320, n = 16, 15 cells for shRNA(scramble), shRNA(Gng4) respectively). *p < 0.05, **p < 0.01, ***p < 0.001 by unpaired two-tailed Student’s t-test (b, h, r), Pearson’s correlation test (c, d), two-sided Mann Whitney U test (q), two-way repeated-measures ANOVA with Bonferroni’s post-hoc test (k, l, o), two-way ANOVA with Tukey’s post-hoc test (n); data represent the mean ± SEM. a.u., arbitrary units. See also Supplementary Figs. 10–13 and Supplementary Table 1. Source data are provided as a Source Data file.

Fig. 7. Knockdown of Gγ4 in V1 → LP excitatory neurons aggravates depressive-like behaviors, while overexpression alleviates depressive-like phenotypes in CRS mice.

a Schematic of retrograde tracing of Glu^V1→LP neurons in adult C57BL/6 J mice. b, c Representative images and quantification of the fluorescence intensity of Gγ4 (cyan) in mCherry⁺ (red) cells in the V1 after the control or CRS paradigm (P = 0.0016, n = 8 mice/group). Blue, DAPI staining. Green, EGFP expression in the V1. Scale bar, 50 µm. d, e Schematic and representative image of AAV viral vector injection to achieve in vivo knockdown of Gγ4 in V1 → LP excitatory neurons in mice (d). Scale bar, 100 µm. Representative images showing the expression of mCherry (red) and Gγ4 (magenta) in the V1 4 weeks after viral injection (e). Right, quantification of the fluorescence intensity of Gγ4 (magenta) in the mCherry⁺ (red) and mCherry^- cells (P = 0.0006, n = 6 slices/group). Yellow arrowheads indicate mCherry⁺ cells. White arrowheads indicate mCherry^- cells. Blue, DAPI staining. Scale bar, 50 µm. Behavioral timeline of SCRS induction and assessment of depressive-like phenotypes (f). Data showing the effect of knockdown of Gγ4 in V1 → LP excitatory neurons on SPT (g) and FST (h) results after the 7-day SCRS paradigm. g shRNA(scramble) + Baseline vs. shRNA(scramble) + SCRS, P = 0.0226; shRNA(Gng4) + Baseline vs. shRNA(Gng4) + SCRS, P < 0.0001; shRNA(scramble) + SCRS vs. shRNA(Gng4) + SCRS, P = 0.0238; h shRNA(Gng4) + Baseline vs. shRNA(Gng4) + SCRS, P = 0.0123 (n = 8, 9 mice for shRNA(scramble), shRNA(Gng4) respectively). i–k Behavioral timeline of CRS induction and assessment of depressive-like phenotypes (i). Data showing the effect of knockdown of Gγ4 in V1 → LP excitatory neurons on SPT (j) and FST (k) results after the 21-day CRS paradigm. j shRNA(scramble) + Baseline vs. shRNA(scramble) + CRS, P = 0.0306; shRNA(Gng4) + Baseline vs. shRNA(Gng4) + CRS, P = 0.0002; shRNA(scramble) + CRS vs. shRNA(Gng4) + CRS, P = 0.0147; k shRNA(scramble) + Baseline vs. shRNA(scramble) + CRS, P = 0.0493; shRNA(Gng4) + Baseline vs. shRNA(Gng4) + CRS, P = 0.0059; shRNA(scramble) + CRS vs. shRNA(Gng4) + CRS, P = 0.0411 (n = 8, 7 mice for shRNA(scramble), shRNA(Gng4) respectively). Schematic and representative image of AAV viral vector injection to achieve in vivo overexpression of Gγ4 in V1 → LP excitatory neurons in mice (l). Scale bar, 100 µm. Representative images showing the expression of mCherry (red) and Gγ4 (magenta) in the V1 4 weeks after viral injection (m). Right, quantification of the fluorescence intensity of Gγ4 (magenta) in the mCherry⁺ (red) and mCherry^- cells (P = 0.0159, n = 5 slices/group). Yellow arrowheads indicate mCherry⁺ cells. White arrowheads indicate mCherry^- cells. Blue, DAPI staining. Scale bar, 50 µm. n–p Behavioral timeline of CRS induction and assessment of depressive-like phenotypes (n). Data showing the effect of overexpression of Gγ4 in V1 → LP excitatory neurons on SPT (o) and FST (p) results after the 21-day CRS paradigm. o mCherry + Control vs. mCherry + CRS, P < 0.0001; mCherry + CRS vs. OE(Gng4) + CRS, P = 0.0202; p mCherry + Control vs. mCherry + CRS, P = 0.0002; mCherry + CRS vs. OE(Gng4) + CRS, P < 0.0001 (n = 10 mice/group). *p < 0.05, **p < 0.01, ***p < 0.001 by unpaired or paired two-tailed Student’s t-test (c, e, m), two-way repeated-measures ANOVA with Bonferroni’s post-hoc test (g, h, j, k), two-way ANOVA with Tukey’s post-hoc test (o, p); data represent the mean ± SEM. a.u., arbitrary units. Source data are provided as a Source Data file.