Inhibition of Trpv4 rescues circuit and social deficits unmasked by acute inflammatory response in a Shank3 mouse model of Autism

Abstract

Mutations in the SHANK3 gene have been recognized as a genetic risk factor for Autism Spectrum Disorder (ASD), a neurodevelopmental disease characterized by social deficits and repetitive behaviors. While heterozygous SHANK3 mutations are usually the types of mutations associated with idiopathic autism in patients, heterozygous deletion of Shank3 gene in mice does not commonly induce ASD-related behavioral deficit. Here, we used in-vivo and ex-vivo approaches to demonstrate that region-specific neonatal downregulation of Shank3 in the Nucleus Accumbens promotes D1R-medium spiny neurons (D1R-MSNs) hyperexcitability and upregulates Transient Receptor Potential Vanilloid 4 (Trpv4) to impair social behavior. Interestingly, genetically vulnerable Shank3^+/- mice, when challenged with Lipopolysaccharide to induce an acute inflammatory response, showed similar circuit and behavioral alterations that were rescued by acute Trpv4 inhibition. Altogether our data demonstrate shared molecular and circuit mechanisms between ASD-relevant genetic alterations and environmental insults, which ultimately lead to sociability dysfunctions.

Fig. 1. Downregulation of Shank3 in the NAc during early postnatal development alters social preference.

Schema of injection sites in the NAc with AAV-scrShank3-GFP or AAV-shShank3-luczsGreen in P6 mice (a) or at P90 (d). a′, d′ Representative images of injection sites (scale bar: 500 µm). Left: time spent around the enclosures during the social preference test for mice injected at P6 (b and c) or at P90 (e and f) (paired-samples t-tests for object- vs. social: b t₍₁₂₎ = 6.092, p < 0.001; c t₍₉₎ = 0.409, p = 0.697; e t₍₉₎ = 3.806, p = 0.004; f t₍₆₎ = 6.970, p < 0.001). Right: juvenile preference index for mice injected at P6 (b and c) or at P90 (e and f) (one-sample t-tests against chance level = 0.5: b t₍₁₂₎ = 5.847, p < 0.001; c t₍₉₎ = 0.273, p = 0.791; e t₍₉₎ = 3.928, p = 0.003; f t₍₆₎ = 7.996, p < 0.001). Error bars report SEM.

Fig. 2. Shank3 NAc downregulation alters D1R MSNs excitability. Decreasing the activity of D1R MSNs normalizes sociability deficits.

a Experimental design. Drd1a-dTomato mice were injected neonatally in the NAc with scr or shShank3 virus and whole-cell patch clamp recordings were performed during early adulthood. a′ Representative image of the NAc of a D1R-tom+::shShank3 mouse (scale bar: 50 µm). b Example traces at 300 pA depolarizing current injection in D1R-tom+ MSNs infected with scrShank3 (left) or with shShank3 (right). c Number of action potentials (nAPs) across increasing depolarizing current steps (0-500 pA) for D1R-tom+::scrShank3 and shShank3 MSNs, in presence of picrotoxin and kynurenic acid (repeated measures two-way ANOVA, virus main effect F_{(1, 14)} = 10.88, p = 0.005, current steps main effect F_{(10, 140)} = 7.727, p < 0.001, virus x current step interaction F_{(10, 140)} = 1.626, p = 0.1051, scrShank3 n = 8 cells, 3 mice, shShank3 n = 8 cells, 3 mice). d Example traces at 300 pA depolarizing current injection in D1R-tom- MSNs infected with scrShank3 (left) or with shShank3 (right). e Number of action potentials (nAPs) across increasing depolarizing current steps (0−500 pA) for D1R-tom−::scrShank3 and shShank3 MSNs, in presence of picrotoxin and kynurenic acid (repeated measures (RM) two-way ANOVA, virus main effect F_(1,18) = 0.098, p = 0.758, current steps main effect F_{(10, 180)} = 14.58, p < 0.001, virus x current step interaction F_{(10, 180)} = 0.3254, p = 0. 9736, n = 8 cells, 3 mice (sh), n = 12 cells, 3 mice (scr)). f Experimental design. D1R-Cre positive (D1R:Cre⁺) and negative (D1R:Cre⁻) mice were injected neonatally in the NAc with scr or shShank3 virus and after P30 with AAV-hSyn-DIO-hM4Di-mCherry (DREADD). After 4 weeks, allowing for virus expression, the mice underwent social behavior assessment in the three-chamber task. All mice were intraperitoneally injected with CNO 30 min before starting the test. f′ Representative image of the NAc of a D1R:Cre⁺ mouse infected with shShank3 and DREADD viruses (scale bar: 50 µm). Left: time around the target during the social preference test for D1R:Cre^-::scrShank3 mice: paired-samples t-test for object- vs. social: g t₍₇₎ = 5.453, p = 0.001; D1R:Cre⁺::scrShank3 mice, paired-samples t-test for object- vs. social: h t₍₇₎ = 0.471, p = 0.652; D1R:Cre⁻::shShank3 mice, paired-samples t-test for object- vs. social: i t₍₆₎ = 0.264, p = 0.801 and D1R:Cre⁺::shShank3 mice, paired-samples t-test for object- vs. social: j t₍₈₎ = 3.443, p = 0.009. Right: juvenile preference index (one-sample t-tests against chance level = 0.5: D1R:Cre⁻::scrShank3; t₍₇₎ = 6.395, p < 0.001, D1R:Cre⁺::scrShank3; t₍₇₎ = 0.054, p = 0.958, D1R:Cre⁻::shShank3; t₍₆₎ = 0.334, p = 0.750, D1R:Cre⁺::shShank3; t₍₈₎ = 3.706, p = 0.006). Error bars report SEM.

Fig. 3. Downregulation of Shank3 in D1R MSNs induces alterations in inflammatory mediators and Trpv4 expression.

a Experimental design. Drd1a-dTomato mice were injected neonatally in the NAc with scr or shShank3 virus. At P30 the NAc was dissected and FACsorted in 4 different cell populations (D1R-tom+, D1R-tom-, D1R-tom+::AAV and D1R-tom−::AAV). For each cell population we carried out bulk RNA sequencing. b Worst-case scenario selected altered genes in scr vs sh testing clearly discriminated infected cells, both D1R+ and D1R− in PCA analysis. c While non-infected samples do not share common genes significantly altered in scr vs sh testing, infected D1R+ and D1R- share a core set of 68 altered genes. d Overall GO:Term analysis of infected D1R+ significantly altered genes highlights the relevance of inflammatory mechanisms, as well as cell adhesion-, localization- and movement-related functions. e D1R-tom+ altered genes include genes expressing proteins directly involved in electrophysiological properties, including the Transient receptor potential vanilloid 4 (Trpv4). f Real-time PCR analysis of NAc dissected from P6- or P90-injected mice confirm the upregulation of Trpv4 in P6 sh-infected mice (unpaired t-tests: P6 – Trpv4 t₍₄₎ = 2.980, p = 0.041; P6 – Trpv1 t₍₄₎ = 0.367, p = 0.732; P6 – HCN1 t₍₄₎ = 0.318, p = 0.766; P90 – Trpv4 t₍₅₎ = 0.4203, p = 0.69). g Experimental design. Drd1a-dTomato mice were injected neonatally in the NAc with scr or shShank3 virus and whole-cell patch clamp recordings were performed during early adulthood. h Right: example traces from 300 pA depolarizing current injection in D1R+ MSNs infected with scrShank3 treated with vehicle (left), D1R+ MSNs infected with shShank3 treated with vehicle (middle) and D1R+ MSNs infected with shShank3 treated with HC-067047 (right). Left: number of action potentials (nAPs) across increasing depolarizing current steps (0−500 pA) for D1R-tom+::scrShank3 and shShank3 MSNs in the presence of Trpv4 antagonist (HC-067047) (repeated measures ANOVA, drug main effect F_{(2, 31)} = 5.883, p = 0.007, current steps main effect F_{(10, 310)} = 24.15, p < 0.001, drug by current steps interaction F_{(20, 310)} = 1.685, p = 0.035, n = 12 cells, 4 mice (shShank3-Veh), n = 10 cells, 3 mice (shShank3-Trpv4), n = 12 cells, 4 mice (scrShank3-Veh)). i Experimental design. C57BL6/j mice were injected neonatally in the NAc with scr or shShank3 virus and at P50-60 were bilaterally cannulated above the NAc. After 7 days, mice underwent the three-chamber social interaction assay. ScrShank3 were infused with vehicle (aCSF/DMSO 0.3%). On the other hand, shShank3 mice were infused with either vehicle (aCSF/DMSO 0.3%) or HC-067047 (2 µg in aCSF/DMSO 0.3%) 10 min before to start the test. i′ Representative image of the injection site and cannula placement above the NAc (scale bar: 250 µm). Left: time around the target during the social preference test for mice infected with scrShank3 and infused with vehicle (paired-samples t-test for object- vs. social: j t₍₅₎ = 6.304, p = 0.002), mice infected with shShank3 and infused with vehicle (paired-samples t-test for object- vs. social: k t₍₇₎ = 0.869, p = 0.414) or with HC-067047 (paired-samples t-test for object- vs. social: l t₍₇₎ = 4.324, p = 0.004). Right: juvenile preference index for mice infused either with vehicle or with HC-067047 (one-sample t-tests against chance level = 0.5: j t₍₅₎ = 6.459, p = 0.001; k t₍₇₎ = 1.02, p = 0.342; l t₍₇₎ = 6.078, p = 0.001). m Juvenile preference index comparison between shShank3-vehicle and shShank3-HC-067047 (paired-samples t-test for object- vs. social: t₍₇₎ = 2.6, p = 0.035). Error bars report SEM.

Fig. 4. LPS challenge in Shank3^+/− unmasks social deficits.

a Experimental design. Shank3^+/+ and Shank3^+/− were intraperitoneally injected with LPS or vehicle and 24 h later they were subjected to the three-chamber task. Left: Time spent around the target (paired-samples t-tests for object- vs. social: b t₍₇₎ = 7.686, p < 0.001; c t₍₈₎ = 4.199, p = 0.003; d t₍₈₎ = 3.462, p = 0.008; e t₍₉₎ = 0.935, p = 0.374). Right: juvenile preference index (one-sample t-tests against chance level = 0.5: b t₍₇₎ = 7.2, p < 0.001; c t₍₈₎ = 5.262, p < 0.001; d t₍₈₎ = 3.734, p = 0.006; e t₍₉₎ = 0.9747, p = 0.355). f Experimental design. Shank3^+/+ and Shank3^+/− were intraperitoneally injected with LPS and 7 days later were subjected to a three-chamber task. Left: Time spent around the target (paired-samples t-tests for object- vs. social: g t₍₆₎ = 5.979, p = 0.001; h t₍₉₎ = 2.759, p = 0.022). Right: juvenile preference index (one-sample t-tests against chance level = 0.5: g t₍₆₎ = 6.054, p < 0.001; h t₍₉₎ = 2.463, p = 0.036). i mRNA expression analysis of IL-1β, TNF-α and Trpv4 genes after LPS challenge in Shank3^+/+ (IL-1β one-way ANOVA followed by Sidak’s multiple comparisons test, F_{(2, 9)} = 10.33, p = 0.005; TNF-α Kruskal−Wallis statistic 7.538, p = 0.012; Trpv4 one way ANOVA followed by Sidak’s multiple comparisons test, F_{(2, 9)} = 2.768, p = 0.116). j mRNA expression analysis of IL-1β, TNF-α and Trpv4 genes after LPS challenge in Shank3^+/− (IL-1β Kruskal−Wallis statistic 9.002, p = 0.002; TNF-α one way ANOVA followed by Sidak’s multiple comparisons test, F_{(2, 10)} = 10.27, p = 0.004; Trpv4 one way ANOVA followed by Sidak’s multiple comparisons test: F _{(2, 9)} = 31.26, p < 0.001). k Experimental design. Shank3^+/− were crossed with Drd1a-tdTomato mice labeling specifically D1R-MSNs in a Shank3^+/− background. Ex-vivo patch clamp recordings were made 24 h after the LPS injection. l Whole-cell recording of Trpv4 current 24 h after LPS challenge in Shank3^+/+ and Shank3^+/− mice (repeated measures ANOVA, voltage steps main effect F_{(1.171,25.77)} = 12.11, p = 0.001, genotype main effect F_{(3, 22)} = 0.4152, p = 0.744, genotype by voltage steps interaction F_{(120, 880)} = 1.451, p = 0.002; n = 5 cells, 2 mice (Shank3^+/+), n = 5 cells, 2 mice (Shank3^+/+ + LPS), n = 7 cells, 2 mice (Shank3^−/+) n = 9 cells, 2 mice (Shank3^−/+ + LPS)). m Example traces from 300 pA depolarizing current injection in D1R + MSNs of Shank3^+/− mice 24 h after: vehicle IP injection and treated with vehicle (left), LPS challenge and treated with vehicle (middle), LPS challenge and treated with HC-067047 (right). n Number of action potentials (nAPs) across increasing depolarizing current steps (0−500 pA) for D1R-tom+::Shank3^+/− MSNs after LPS challenge (repeated measures ANOVA, treatment main effect F_{(2, 30)} = 3.034, p = 0.063, current steps main effect F_{(10, 300)} = 28.08, p < 0.001, treatment by current steps interaction F_{(20, 300)} = 2.042, p = 0.006, n = 10 cells, 3−4 mice each group). Error bars report SEM.

Fig. 5. Acute LPS challenge alters the calcium transients of NAc cells in Shank3^+/− mice during sociability test.

a Experimental design. Adult Shank3^+/+ and Shank3^+/− mice were unilaterally injected in the NAc with an AAV-hSyn-chl-jGCaMP7s virus and subsequently an optic fiber was implanted above the ROI. These animals were then tested in the 3- chamber task. a′ Representative image of injection site and optic fiber implantation (scale bar: 500 µm). b Example of recorded ΔF/F during the three-chamber task. The entry in proximity of the stimuli are indicated with bold lines (entry in the juvenile proximity in blue and orange for the object). The time passed in proximity of the stimuli is indicated below the traces. b′ Schema reporting when the experimental animal is entering in proximity to the juvenile stimulus. c, e, h, and j Heatmaps reporting the PETH of normalized (z-score) ΔF/F recorded in the NAc and centered on the entry in proximity of the stimuli (time = 0). d, f, i, and k Left: PETH of normalized (z-score) ΔF/F recorded in the NAc and centered on the entry in proximity of the stimuli (time = 0). Right: Comparison of the juvenile (Jv, in blue) and object (O, in orange) mean z-scores obtained on averaging the normalized ΔF/F 5 s after the entry in proximity of the stimuli (paired-samples t-tests for object- vs. social: d t₍₅₎ = 8.767, p < 0.001; i t₍₄₎ = 5.301, p = 0.006. Wilcoxon test for object- vs. social: f W = −39.00, p = 0.019; k W = −16.00, p = 0.219). g Experimental design. After a pause of 7 days, the mice that performed the experiments described in (a) were intraperitoneally injected with LPS and 24 h later were subjected to a second three chambers test. Abbreviations: OF optic fiber, Jv juvenile, O object. Error bars report SEM.

Fig. 6. Trpv4 antagonist infused in the NAc improves social deficits in Shank3^+/− mice challenged with LPS.

a Experimental design. Adult Shank3^+/− mice were intraperitoneally injected with LPS and 24 h later were subjected to the behavioral task. 30 min before the test, mice were infused (in the NAc) either with Trpv4 antagonist (HC-067047) or vehicle. Left: Time spent around the target for Shank3^+/ mice after LPS challenge and vehicle or HC-067047 infusion in the NAc (paired-samples t-tests for object- vs. social: b t₍₆₎ = 0.408, p = 0.697; c t₍₆₎ = 2.787, p = 0.032). Right: juvenile preference index (one-sample t-tests against chance level = 0.5: b t₍₆₎ = 0.629, p = 0.439; c t₍₆₎ = 2.852, p = 0.029). Error bars report SEM.