Alteration by p11 of mGluR5 localization regulates depression-like behaviors

Abstract

Mood disorders and antidepressant therapy involve alterations of monoaminergic and glutamatergic transmission. The protein S100A10 (p11) was identified as a regulator of serotonin receptors, and it has been implicated in the etiology of depression and in mediating the antidepressant actions of selective serotonin reuptake inhibitors. Here we report that p11 can also regulate depression-like behaviors via regulation of a glutamatergic receptor in mice. p11 directly binds to the cytoplasmic tail of metabotropic glutamate receptor 5 (mGluR5). p11 and mGluR5 mutually facilitate their accumulation at the plasma membrane, and p11 increases cell surface availability of the receptor. Whereas p11 overexpression potentiates mGluR5 agonist-induced calcium responses, overexpression of mGluR5 mutant, which does not interact with p11, diminishes the calcium responses in cultured cells. Knockout of mGluR5 or p11 specifically in glutamatergic neurons in mice causes depression-like behaviors. Conversely, knockout of mGluR5 or p11 in GABAergic neurons causes antidepressant-like behaviors. Inhibition of mGluR5 with an antagonist, 2-methyl-6-(phenylethynyl)pyridine (MPEP), induces antidepressant-like behaviors in a p11-dependent manner. Notably, the antidepressant-like action of MPEP is mediated by parvalbumin-positive GABAergic interneurons, resulting in a decrease of inhibitory neuronal firing with a resultant increase of excitatory neuronal firing. These results identify a molecular and cellular basis by which mGluR5 antagonism achieves its antidepressant-like activity.

Figure 1

p11 interacts with cytoplasmic tail of mGluR5 and regulates mGluR5 signaling. (a) HEK293 cells were co-transfected with full-length mGluR5 plus an empty plasmid (CTR), a plasmid for flag-tagged WT p11, or a plasmid for flag-tagged C83Q mutant p11 which is not able to bind to AnxA2. p11 was immunoprecipitated with anti-flag antibody, and co-precipitated samples (AnxA2 and mGluR5) were detected by immunoblotting. (b) Detergent-solubilized synaptosomal fractions prepared from cortex and hippocampus were subjected to GST-pull down assay. Immobilized GST and GST-p11 fused to the N-terminal region of AnxA2 (GST-p11-AnxA2 fusion protein) were used to measure co-precipitated mGluR5. (c) Schematic diagram showing p11 binding motif (aa 836–844) in the cytoplasmic tail (aa 826–1171) of rat mGluR5 (WT) and alanine substitution of 5 critical amino acids in the motif (MT). (d) In vitro translated cytoplasmic tails of WT or MT mGluR5 were subjected to GST-pull down assay. Immobilized GST, GST-p11, and GST-p11-AnxA2 fusion protein were used to measure co-precipitated cytoplasmic tails of mGluR5. Data represent three independent experiments. (e–h) HEK293 cells were co-transfected with a plasmid for WT mGluR5 plus either a control plasmid (CTR) or a plasmid for p11 (e, f), or transfected with a plasmid for WT or MT mGluR5 (g, h). Calcium oscillations were measured from Fura-2 AM loaded cells (e) and Oregon Green Bapta 488-1 AM loaded cells (g) during 10 min exposure to 10 µM DHPG. Dashed lines indicate mean frequencies. Normalized mean frequencies of calcium peaks are shown (f, h). Wilcoxon rank sum test (f, Z = −3.7191, ***p = 1.9995e-04, n = 179, 225 cells; h, Z = −9.1242, ****p = 7.2294e-20, n = 223, 87 cells). Bar graphs are means ± SEM.

Figure 2

p11 colocalizes with mGluR5 at the plasma membrane and increases cell surface localization of mGluR5. (a–c) Confocal images of HEK293 cells transfected with control vector (a), WT mGluR5 (b), or MT mGluR5 (c). Bright field imaging was performed to display cell membrane and nucleus without labeling (1^st column). Color coding is as follows: red, immunofluorescence labeling of mGluR5 (2^nd column); green, immunofluorescence labeling of p11 (3^rd column). Merge of images of 1^st, 2^nd, and 3^rd columns are displayed in the 4^th column. Specific pixel-by-pixel colocalization of mGluR5 and p11 is represented with an intensity heat-map in the 5^th column. Scale bars, 10 µm. Note: Additional imaging data sets from two independent experiments are provided in Supplementary Figure S3. (d–g) COS7 cells were co-transfected with a plasmid expressing mGluR5 plus (d, e) control RNA duplex (CTR) or p11 siRNA, or (f, g) an empty vector (CTR) or a plasmid expressing p11. (h, i) COS7 cells were transfected with a plasmid expressing WT or MT mGluR5. Cell surface proteins were biotinylated and precipitated with streptavidin-coupled beads. p11 and total mGluR5 in the lysates and the precipitated surface mGluR5 were detected by immunoblotting (d, f, h). Quantification of the surface level of mGluR5 normalized to total mGluR5. Unpaired two tailed t-test (e, t = 3.193, *p = 0.0188, n = 4 per group; g, t = 2.797, *p = 0.0189, n = 6 per group; i, t = 2.873, *p = 0.0166, n = 6 per group). Data represent three independent experiments. (j–l) Immunoblotting of synaptosomal fractions and total lysates prepared from wild-type (+/+) and p11 null (−/−) mice (j). Quantification of the level of mGluR5 normalized to PSD95 and GAPDH in synaptosomal fractions (k, unpaired two tailed t-test, t = 3.762, *p = 0.0197, n = 3 per group) and in total lysates (l, unpaired two tailed t-test, t = 0.2128, p = 0.8419, n = 3 per group), respectively. All bar graphs are means ± SEM.

Figure 3

Effects of mGluR5 KO and p11 KO in glutamatergic and in GABAeric neurons on depression-like behaviors. Deletion of mGluR5 (a–f) or p11 (g–l) in forebrain excitatory neurons or in pan-GABAergic neurons was achieved by breeding floxed mGluR5 or floxed p11 mice with EMX-Cre or GAD-Cre mice. Depression-like behaviors were examined by forced swim test (FST) (a, d, g, j) and tail suspension test (TST) (b, e, h, k) to measure immobility, and by sucrose preference test (SPT) to measure anhedonia (c, f, i, l). Unpaired two tailed t-test (a, t = 2.138, *p = 0.0445, n = 9, 14; b, t = 2.306, *p = 0.0358, n = 8, 9; c, t = 2.644, *p = 0.0238, n = 10 per group; d, t = 2.164, *p = 0.0428, n = 12, 10; e, t = 2.191, *p = 0.0459, n = 8 per group; f, t = 2.794, *p = 0.0112, n = 12, 10; g, t = 3.789, **p = 0.002, n = 8 per group; h, t = 2.751, *p = 0.0156, n = 8 per group; i, t = 2.185, *p = 0.0398, n = 12 per group; j, t = 2.278, *p = 0.0389, n = 11, 8; k, t = 2.259, *p = 0.0373, n = 11, 8; l, t = 2.150, *p = 0.0418, n = 14, 12). Bar graphs are means ± SEM.

Figure 4

p11 in Parvalbumin (PV)-positive GABAergic interneurons is required for antidepressant-like behavioral effect of mGluR5 antagonism. (a) Wild-type (WT, +/+) and p11 null (KO, −/−) mice were intraperitoneally injected (IP) with vehicle (Veh) or 3 mg/kg of MPEP at 24 hr prior to novelty suppressed feeding test (NSF) (two-way ANOVA: interaction between genotype (WT, KO) and drug (Veh, MPEP), F (1, 36) = 5.364, p = 0.0264, n = 10 per group). Unpaired two tailed t-test, WT-Veh vs WT-MPEP (t = 2.613, * p = 0.0126). (b) WT and KO mice were IP injected with Veh or 1 mg/kg of mGluR2/3 antagonist, LY341495, at 24 hr prior to NSF test (two-way ANOVA: interaction between genotype (WT, KO) and drug (Veh, LY341495), F (1, 28) = 0.01109, P =0.9169, n = 8 per group). Unpaired two tailed t-test (WT-Veh vs WT- LY341495 (t = 2.160, *p = 0.0486), KO-Veh vs KO- LY341495 (t = 2.258, *p = 0.0404). (c–h) EMX-, GAD-, and PV- p11 (c–e) or mGluR5 (f–h) KO mice were administered Veh or MPEP (3mg/kg) at 24 hr prior to NSF test. Two-way ANOVA: interaction between genotype (WT, KO) and drug (Veh, MPEP) (c, F (1, 36) = 0.02865, p = 0.8665, n = 10 per group; d, F (1, 26) = 5.023, p = 0.0338, n = 8, 8, 7, 7; e, F (1, 28) = 4.853, p = 0.0360, n = 8 per group; f, F (1, 28) = 0.001575, P = 0.9686, n = 8 per group; g, F (1, 33) = 4.965, p = 0.0328, n = 10, 9, 10, 8; h, F (1, 40) = 4.204, p = 0.0469, n = 14, 13, 8, 9). Unpaired two tailed t-test (c, WT-Veh vs WT-MPEP (t = 2.528, *p = 0.021), KO-Veh vs KO-MPEP (t = 2.705, *p = 0.0145); d, WT-Veh vs WT-MPEP (t = 2.820, *p = 0.0136), WT-Veh vs KO-Veh (t = 2.799, *p = 0.015); e, WT-Veh vs WT-MPEP (t = 2.963, *p = 0.0103); f, WT-Veh vs WT-MPEP (t = 2.703, *p = 0.0172), KO-Veh vs KO-MPEP (t = 2.964, *p = 0.0103); g, WT-Veh vs WT-MPEP (t = 2.703, *p = 0.0151), WT-Veh vs KO-Veh ( t= 2.660, *p = 0.016); h, WT-Veh vs WT-MPEP (t = 2.706, *p = 0.0121), WT-Veh vs KO-Veh (t = 2.391, *p = 0.0268)). Bar graphs are means ± SEM.

Figure 5

MPEP inhibits the firing rates of inhibitory neurons and consequently enhances the firing rates of excitatory neurons. (a) Population of inhibitory or excitatory neurons in mPFC showing increase (>120%, red), no change (80–120%, grey) or decrease in firing rate (<80%, blue) at 20 min after administration of MPEP. The actual number of neurons (from 8 mice) in each category is shown inside bar graphs. (b, c) Decreased (blue) and increased (red) firing rates in subpopulations of inhibitory neurons (b) and excitatory neurons (c) were quantified over the indicated time period before (solid bar) and after (open bars) administration of MPEP. Paired t test (upper graph in b, n = 6 from 8 mice; 0–20 min, t = 2.355, p = 0.065; 20–40 min, t = 2.859, *p = 0.035; 40–60 min, t = 3.656, *p = 0.015). Wilcoxon signed rank test (upper graph in c, n = 24 from 8 mice; 0–20 min, Z = 3.943, **p < 0.001; 20–40 min, Z = 4.286, **p < 0.001; 40–60 min, Z = 4.257, **p < 0.001). All bar graphs represent means ± SEM. (d) A proposed model for mGluR5 as a regulator of neuronal activity in both glutamatergic and GABAergic neurons, and a dominant role for decreased GABAergic inhibitory input onto glutamatergic neurons in antidepressant-like behavior.