Loss of transient receptor potential channel 5 causes obesity and postpartum depression

Abstract

Hypothalamic neural circuits regulate instinctive behaviors such as food seeking, the fight/flight response, socialization, and maternal care. Here, we identified microdeletions on chromosome Xq23 disrupting the brain-expressed transient receptor potential (TRP) channel 5 (TRPC5). This family of channels detects sensory stimuli and converts them into electrical signals interpretable by the brain. Male TRPC5 deletion carriers exhibited food seeking, obesity, anxiety, and autism, which were recapitulated in knockin male mice harboring a human loss-of-function TRPC5 mutation. Women carrying TRPC5 deletions had severe postpartum depression. As mothers, female knockin mice exhibited anhedonia and depression-like behavior with impaired care of offspring. Deletion of Trpc5 from oxytocin neurons in the hypothalamic paraventricular nucleus caused obesity in both sexes and postpartum depressive behavior in females, while Trpc5 overexpression in oxytocin neurons in knock-in mice reversed these phenotypes. We demonstrate that TRPC5 plays a pivotal role in mediating innate human behaviors fundamental to survival, including food seeking and maternal care.

Keywords: TRP channels; anxiety; autism; hypothalamus; innate behavior; obesity; postpartum depression; sensory signaling.

Figure 1.. TRPC5 variants identified in people with severe obesity

(A) Deletions on the X chromosome disrupting TRPC5 found in Cases 1 and 2; genes shown using a color code representing their tolerance to loss-of-function variants, pLI (loss intolerance probability). Coding and untranslated regions are shown in Figure S1A. (B) Co-segregation of deletions with obesity (blue), anxiety (orange), and postpartum depression (green) in families of Cases 1 and 2; males (squares), females (circles), and probands (arrows); age, BMI, and BMI standard deviation score (sds) for children under 18 years are indicated. (C) Rare TRPC5 variants found in people with severe obesity shown on a schematic of the TRPC5 protein: NH2, amino-terminal domain; ANK, ankyrin domains 1–4; bHLH, basic helix loop helix; S1–S6, transmembrane helical domains; TRP, transient receptor potential domain; CIRB, calmodulin/inositol 1,4,5-triphosphate receptor binding domain; CBII, second calmodulin-binding domain; COOH, carboxy-terminal domain. (D–G) Functional characterization of TRPC5 variants. (D) Expression of WT/mutant TRPC5 in cells by western blotting (normalized to glyceraldehyde-3-phosphate dehydrogenase, (GAPDH) as a loading control). (E) Stability of WT/mutant TRPC5 in the cycloheximide (CHX) chase analysis. (F) Cell membrane localization of TRPC5; eGFP-tagged TRPC5 (green), DiI-stained cell membrane (red), and Hoechst33342-stained nucleus (blue). Scale bars, 20 μm. (G) TRPC5-mediated currents in cells transfected with empty vector, WT/mutant TRPC5 stimulated with the acetylcholine receptor agonist, carbachol (CCh); AC1903, TRPC5 inhibitor. Representative inward current traces and current density measured as ratio of peak current amplitude to cell membrane capacitance (pA/pF). Data expressed as % WT in (D) and (F); bars represent standard error of mean. p values were determined by unpaired t test with Welch’s correction; *p < 0.05, **p < 0.01, and ***p < 0.001. See also Tables S1–S3 and Figure S1.

Figure 2.. Metabolic and behavioral phenotype of male Trpc5^K34del/Y hemizygous mice

(A–C) Experiments in male WT and Trpc5^K34del/Y hemizygous mice on a high-fat diet (HFD). Body weight (A), body composition (B), and weekly food intake (C) (n = 8–9 per group). (D) Locomotor activity (xy axis) during 24 h, dark and light cycles (n = 5–8 per group, 13 weeks of age). (E) Energy expenditure (EE) during 24 h, dark and light cycles, regression of EE with body mass, and predicted EE over 24 h based on 26 g body mass/mouse (n = 4–6 per group, 13 weeks of age). (F) Food-hoarding test. (G) Amount of food (grams) hoarded at 24°C or 28°C (n = 6–21 per group, 20 weeks of age). (H) Open field test: heatmap of movement, distance traveled in the center area, number (#) of center entries, time spent in the center, duration of rearing (n = 8–17 per group, 16 weeks of age). (I) Awake time during 24 h in fed, fasted, and refed conditions (n = 8–16 per group, 16 weeks of age). (J) Three-chamber test used to study social behavior. (K) Interaction time with object and mouse in chamber and preference ratio (mouse vs. object) (n = 10–12 per group, 12 weeks of age). (L) Resident-intruder assay. (M) % of mice that attacked intruders (n = 8–16 per group, 13 weeks of age). (N) Latency to attack, number (#) of attacks, and duration of attack (n = 8–16 per group, 13 weeks of age). Data presented as mean ± SEM, p value determined using 2-way ANOVA (A and C), unpaired t tests (B, D, E, G–I, K, and N), paired t tests (G), or chi-squared test (M). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Overall difference between groups is indicated in the panel legend as appropriate. See also Figure S2.

Figure 3.. Metabolic and behavioral phenotype of female Trpc5^K34del/+ heterozygous mice

(A–H) Experiments in female WT and Trpc5^K34del/+ heterozygous mice on a high-fat diet (HFD). Body weight (A), body composition (B), and food intake at 7 and 33 weeks (C) (n = 5–7 per group). (D) Food-hoarding test. (E) Amount of food (grams) hoarded at 24°C or 28°C (n = 7 per group, 20 weeks of age). Energy expenditure (F), locomotor activity (xy axis) (G), and respiratory exchange ratio (RER) (H) over 2 days (left) and during 24 h (right) at 24°C or 28°C (n = 5–7 per group, 20 weeks of age). (I) Protocol and timeline of tests of maternal behavior; PPD, postpartum day; FST, forced swim test; SPT, sucrose preference test. (J) Abandon behavior was observed on PPD1. Left: Trpc5^K34del/+ dams failed to make a proper nest; pups were scattered randomly among the bedding; middle: % of dams that ignored and abandoned pups; right: % of dams gathered with pups in the nest (n = 7–9 per group, 14 weeks of age). (K) Left: maternal behavior assay in the home cage; right: sample behavior raster plot of WT and Trpc5^K34del/+ dams. (L) Latency to retrieve pups, % of pups retrieved, number (#) of pups retrieved (n = 9–12 per group, 14 weeks of age). (M) Duration of crouching above pups, pup grooming, and nest-building behavior exhibited by dams, total time spent in nest, and duration of maternal care (n = 9–11 per group, 14 weeks of age). (N) Representative images at the beginning (t = 0 min) and end (t = 10 min) of the pup retrieval test (14 weeks of age). (O) % of pups gathered in the nest, distance of pups from the nest (n = 9–12 per group, 14 weeks of age). (P) Retrieval behavior in the open field arena. (Q) Latency to retrieve pups during 5 trials, number (#) of successfully retrieved trials among 5 trials, number of encounters with the pup in each trial (n = 7–9 per group, 14 weeks of age). (R) Serum prolactin at PPD 12 before and after suckling (n = 3–5 per group, 16 weeks of age). (S) Forced swim test, immobile time in forced swim test (n = 16–19 per group, 18 weeks of age). (T) Sucrose preference test, sucrose preference ratio (n = 15–18 per group, 18 weeks of age). Data presented as mean ± SEM, p value determined using 2-way ANOVA (H, L, and Q), unpaired t tests (A–C, E, F–H, and R–T), Mann-Whitney test (J, L, M, O, and Q), paired t tests (E–H and R), Kolmogorov-Smirnov test (L), or chi-squared test (J). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Overall difference between groups is indicated in the panel legend as appropriate. See also Figure S2 and Video S1.

Figure 4.. Trpc5 deficiency impairs anorectic effects mediated by Pomc neurons

(A–C) (A) Representative images (left) and percentage (right) of ARH Pomc neurons (green) that express Trpc5 (red) in male WT mice (20 weeks of age). Scale bars, 25 μm; DAPI (4′, 6-diamidino-2-phenylindole) staining of nuclei; 3V is third ventricle. Chow intake (B) and HFD intake (C) within 2 h after injection of vehicle or a Trpc5 activator, benzothiadiazine derivative (BTD) in male WT mice (chow intake, n = 5–6 per group, 18 weeks of age; HFD intake, n = 7 per group, 28 weeks of age). (D) Immunoreactivity of β-endorphin (β-END; green), c-Fos (magenta), and merged images (yellow) and % of Pomc neurons (labeled by β-END) expressing c-Fos in vehicle or BTD-injected male WT mice (n = 3 per group, 16 weeks of age). Scale bars, 50 μm. (E) Upper: expression of hM4D(Gi) virus in Pomc neurons; lower: immunofluorescence images showing Pomc neurons expressing hM4D(Gi). Scale bars, 25 μm. (F) Effects of CNO co-injected with vehicle or BTD on food intake in male Pomc-Cre mice receiving inhibitory AAV8-hSyn-DIO-hM4D(Gi)-mCherry infection in the ARH (n = 7 per group, 20 weeks of age). (G) Immunoreactivity of β-END (green), c-Fos (magenta), and merged images (yellow) after injection of vehicle or BTD in male WT and Trpc5^K34del/Y mice (n = 3–4 per group, 12 weeks of age) and % Pomc neurons (labeled by β-END) expressing c-Fos (n = 3–4 per group, 12 weeks of age). Scale bars, 25 μm. (H) Food intake within 2 h after injection of vehicle or BTD in male WT and Trpc5^K34del/Y mice (n = 11–14 per group, 8 weeks of age). (I–K) Chronic injection of BTD in HFD-fed female WT and Trpc5^{K34del/K34del} homozygous females. Body weight change (I), daily food intake (J), and cumulative food intake (K) (n = 3–5 per group, 20 weeks of age). Data presented as mean ± SEM, p value determined using 2-way ANOVA (I–K), unpaired t tests (B–D, G, and H), or paired t tests (F). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Overall difference between groups is indicated in the panel legend as appropriate. See also Figures S3 and S4.

Figure 5.. Metabolic effects of Trpc5 deletion in OXT neurons

(A–E) (A) Left: representative image showing Trpc5 (red) and OXT (green) expression in the PVH in male WT mice. Right: quantification of Trpc5 expression within PVH OXT neurons (n = 4 mice, 16 weeks of age). Scale bars, 50 μm. DAPI staining of nuclei; 3V is third ventricle. Experiments in male control and Trpc5^f/Y/OXT-Cre mice on chow diet. Body weight (B), body composition (C), and weekly food intake (D) (n = 11 per group). (E) Cumulative food intake during 3 days (clock time shown) and during 24 h, dark and light cycles (n = 6–7 per group, 10 weeks of age). (F) Locomotor activity (xy axis) during 24 h, dark and light cycles (n = 7 per group, 10 weeks of age). (G) Food intake and body weight change after saline or OXT injection (n = 6 per group, 20 weeks of age). Data presented as mean ± SEM, p value determined using 2-way ANOVA (B, D, and E), unpaired t tests (C and E–G), or paired t tests (G). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001). See also Figures S5 and S6.

Figure 6.. Behavioral effects of Trpc5 deletion in OXT neurons

(A–C) Experiments in male control and Trpc5^f/Y/OXT-Cre mice. (A and B) Open field test: heatmap of movement, distance traveled in the center area, number (#) of center entries, time spent in the center, and duration of rearing (n = 11 per group, 11 weeks of age). (C) Three-chamber test used to study social behavior, interaction time with object and mouse in chamber, and preference ratio (mouse vs. object) (n = 8 per group, 16 weeks of age). (D) Maternal behavior assay in home cage, sample behavior raster plot of female control, and Trpc5^f/+/OXT-Cre dams. (E–G) Maternal behavior assay in home cage. Latency to retrieve pups (E), percentage of pups retrieved (F), and duration of crouching above pups, pup grooming and nest-building behavior, total time spent in nest, and duration of maternal care (G) (n = 9–14 per group, 11 weeks of age). (H) Retrieval behavior in the open field arena and latency to retrieve pups during 5 trials (n = 9–13 per group, 11 weeks of age). (I) Forced swim test, immobile time in forced swim test (n = 10–16 per group, 15 weeks of age). (J) Sucrose preference test, sucrose preference (n = 9–10 per group, 15 weeks of age). Data presented as mean ± SEM, p value determined using 2-way ANOVA (E and H), unpaired t tests (B, C, I, and J), Mann-Whitney test (G), or Kolmogorov-Smirnov test (F). *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Overall difference between groups is indicated in the panel legend as appropriate. See also Figure S6.

Figure 7.. Restoration of Trpc5 in PVH OXT neurons reverses features of Trpc5 deficiency

(A–G) (A) Viral overexpression of Trpc5 selectively in PVH OXT neurons in male Trpc5^K34del/Y/OXT-Cre mice. Body weight gain (B), body composition (C), cumulative food intake when fed chow and then high-fat diet (HFD) (D), food intake over 3 days and during 24 h, dark and light cycles (E) in male Trpc5^K34del/Y/OXT-Cre mice receiving control adeno-associated virus (AAV) and AAV-DIO-Trpc5 (n = 6–8 per group, 22 weeks of age). Male Trpc5^K34del/Y/OXT-Cre mice receiving control AAV and AAV-DIO-Trpc5 were studied (F–I). Open field test: heatmap of movement (F), distance traveled in the center area, number (#) of center entries, time spent in the center, duration of rearing (G) (n = 8 per group, 13 weeks of age). (H and I) Three-chamber social interaction test (H), interaction time with object and mouse in chamber and preference ratio (mouse vs. object) (I) (n = 8 per group, 13 weeks of age). (J) Experimental design to overexpress Trpc5 or GCaMP6m (control) selectively in PVH OXT neurons of virgin female Trpc5^K34del/+/OXT-Cre mice, followed by the pup retrieval test, forced swim test (FST), and sucrose preference test (SPT). (K) Maternal behavior assay in home cage and sample behavior raster plot of female Trpc5^K34del/+/OXT-Cre mice receiving control AAV and AAV-DIO-Trpc5. (L–N) Maternal behavior assay in home cage. Latency to retrieve pups (L), percentage of pups retrieved (M), duration of crouching above pups, pup grooming and nest-building behavior, total time spent in nest, and duration of maternal care (N) (n = 9–11 per group, 24 weeks of age). (O) Retrieval behavior in the open field arena and latency to retrieve pups during 5 trials (n = 7 per group, 24 weeks of age). (P) Forced swim test, immobile time in forced swim test (n = 11 per group, 28 weeks of age). (Q) Sucrose preference test, sucrose preference ratio (n = 9–10 per group, 28 weeks of age). Data presented as mean ± SEM, p value determined using 2-way ANOVA (B, D, E, L, and O), unpaired t tests (C, E, G, I, P, and Q), Mann-Whitney test (N) or Kolmogorov-Smirnov test (M). *p < 0.05 and **p < 0.01. Overall difference between groups is indicated in the panel legend as appropriate. See also Figure S7.