miR-9-5p is involved in the rescue of stress-dependent dendritic shortening of hippocampal pyramidal neurons induced by acute antidepressant treatment with ketamine

Abstract

Converging clinical and preclinical evidence demonstrates that depressive phenotypes are associated with synaptic dysfunction and dendritic simplification in cortico-limbic glutamatergic areas. On the other hand, the rapid antidepressant effect of acute ketamine is consistently reported to occur together with the rescue of dendritic atrophy and reduction of spine number induced by chronic stress in the hippocampus and prefrontal cortex of animal models of depression. Nevertheless, the molecular mechanisms underlying these morphological alterations remain largely unknown. Here, we found that miR-9-5p levels were selectively reduced in the hippocampus of rats vulnerable to Chronic Mild Stress (CMS), while acute subanesthetic ketamine restored its levels to basal condition in just 24h; miR-9-5p expression inversely correlated with the anhedonic phenotype. A decrease of miR-9-5p was reproduced in an in vitro model of stress, based on primary hippocampal neurons incubated with the stress hormone corticosterone. In both CMS animals and primary neurons, decreased miR-9-5p levels were associated with dendritic simplification, while treatment with ketamine completely rescued the changes. In vitro modulation of miR-9-5p expression showed a direct role of miR-9-5p in regulating dendritic length and spine density in mature primary hippocampal neurons. Among the putative target genes tested, Rest and Sirt1 were validated as biological targets in primary neuronal cultures. Moreover, in line with miR-9-5p changes, REST protein expression levels were remarkably increased in both CMS vulnerable animals and corticosterone-treated neurons, while ketamine completely abolished this alteration. Finally, the shortening of dendritic length in corticosterone-treated neurons was shown to be partly rescued by miR-9-5p overexpression and dependent on REST protein expression. Overall, our data unveiled the functional role of miR-9-5p in the remodeling of dendritic arbor induced by stress/corticosterone in vulnerable animals and its rescue by acute antidepressant treatment with ketamine.

Keywords: CORT; Dendrite morphology; Ketamine; REST; Stress; miR-9-5p.

Fig. 1

miR-9-5p expression and sucrose preference measure in CMS rats. (A) qPCR analysis of miR-9-5p expression in the HPC. Tukey's post-hoc test: *p < 0.05 vs CNT; ^#p < 0.05 vs CMS-V; n = 7–10. Data are reported as means ± S.E.M. (B) Sucrose preference test of CNT and CMS rats at Day +35 of CMS, 24 h after KET/VEH treatment n = CNT 10; CMS-R 10; CMS-V 10; CMS-V + KET 8. Tukey's post-hoc test: ***p < 0.001 vs CNT, ^##p < 0.01 vs CMS-V, ^§§§p < 0.001 vs CMS-R. Data are reported as means ± S.E.M. (C) Correlation between miR-9-5p HPC expression and the sucrose preference in CMS rats. Pearson's r, r = 0.3677, p < 0.05.

Fig. 2

miR-9-5p expression changes induced by CMS and ketamine in the sub-regions of the HPC of rats. (A) Representative images of miR-9-5p in situ hybridization signal in CA3, CA1 and DG regions of HPC in CNT, CMS-V, CMS-R and CMS-V + KET rats. Scramble probe was used as a negative control. Scale bar 100 μm. (B–D) Expression of miR-9-5p measured as optical density in the soma, proximal (20 μm) and distal (40 μm) dendrites of neurons in CA3. (E–G) Expression of miR-9-5p measured as optical density in the soma, proximal (20 μm) and distal (40 μm) dendrites of neurons in CA1. (H–J) Expression of miR-9-5p measured as optical density in the soma, proximal (20 μm) and distal (40 μm) dendrites of neurons in DG. n = 3–4/group, 2-3 slices/rat. Tukey's post-hoc test: *p < 0.05, **p < 0.01, ***p < 0.001 vs CNT; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs CMS-V; ^§p < 0.05, ^§§§p < 0.001 vs CMS-R. Data are reported as means ± S.E.M.

Fig. 3

Analysis of dendritic morphology and miR-9-5p expression in primary hippocampal cultures treated with corticosterone and ketamine. (A) Timing of primary hippocampal cultures in vitro treatment with CORT and KET. (B) Representative image of a hippocampal neuron used for morphological study and reconstruction of the dendritic tree performed using plug-in Fiji at T0 (blue line) and T1 (red line). (C) The difference between T1 and T0 measurements within the same cell is reported as Δ dendritic length T1-T0 (μm), for vehicle, CORT and CORT plus KET treatment. Tukey's post-hoc test: *p < 0.05 vs DMSO, ^#p < 0.05 vs CORT; n = 10 (DMSO), n = 22 (CORT), n = 10 (CORT + KET). (D) miR-9-5p expression in primary hippocampal neurons treated with CORT and KET. Tukey's post-hoc test: **p < 0.01 vs DMSO, ^#p < 0.05 vs CORT; n = 9–10. Scale bar 50 μm.. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

miR-9-5p modulation in primary neurons changes dendritic morphology and spine density. (A–B) Representative reconstruction using plug-in Fiji of the dendritic tree and spines of primary hippocampal neurons transfected with Twl-GFP/RFP and Twl-sponge-miR-9-5p, respectively. Scale bar 50 μm. Morphological analysis of primary hippocampal neurons after 72 h of miR-9-5p downregulation: (C) total dendritic length; n = 15, (D) total number of branches; n = 15, (E) dendritic spine density; n = 20. Unpaired Student's t-test with Welch's correction: *p < 0.05, ***p < 0.001 vs Twl-GFP/RFP. (F–G) Representative reconstruction using plug-in Fiji of the dendritic tree and spines of primary hippocampal neurons transfected with Twl-H1 and Twl-H1-miR-9-5p, respectively. Scale bar 50 μm. Morphological analysis of primary hippocampal neurons after 72 h of miR-9-5p overexpression: (H) total dendritic length; n = 30, (I) total number of branches; n = 30, (J) dendritic spine density; n = 40. Unpaired Student's t-test with Welch's correction: *p < 0.05, ***p < 0.001 vs Twl-H1. Dendritic length, number of branches and spine density were measured using ImageJ or the plug-in Fiji.

Fig. 5

Analysis of miR-9-5p interaction with predicted target genes using Luciferase assay. Relative Luciferase Activity was measured in HEK293T cells co-transfected with specific dual-reporter vectors containing an approximately 70 bp-long sequence of the 3′UTR of (A) DRD2, (B) ELAVL1, (C) GSK-3β, (D) NTRK3, (E) REST, (F) SIRT1, (G) SORT1 and with specific mimics for miR-9-5p or negative control. Unpaired Student's t-test with Welch's correction: *p < 0.05, **p < 0.01 vs NEG control. n = 3; all experiments were repeated 3 times.

Fig. 6

Validation of miR-9-5p target genes. (A,B,C) Relative Luciferase Activity for GSK-3β, REST and SIRT1, long 3′ UTR fragments. Unpaired Student's t-test: *p < 0.05, ***p < 0.001 vs NEG control. n = 3; all experiments were repeated 3 times. (D,E,F) Analysis of GSK-3β, REST and SIRT1 expression in primary hippocampal neurons treated with miR-9-5p mimics or negative control. Unpaired Student's t-test: ***p < 0.001 vs NEG control. n = 9–14. (G,H,I) Analysis of GSK-3β, REST and SIRT1 expression in primary hippocampal neurons treated with corticosterone and ketamine. One-way ANOVA followed by Tukey's post-hoc test: *p < 0.05 vs DMSO; ^#p < 0.05 vs CORT. n = 6–12. (J,K,L) Analysis of GSK-3β, REST and SIRT1 protein expression in the HPC of CMS rats. One-way ANOVA followed by Tukey's post-hoc test: *p < 0.05, **p < 0.01 vs CNT; ^#p < 0.05 vs CMS-V; ^§§p < 0.01 vs CMS-R. n = 4–8.

Fig. 7

Analysis of preventive effect of miR-9-5p overexpression and REST downregulation on dendritic morphology changes induced by CORT treatment. (A) Total dendritic length analysis of primary hippocampal neurons transfected with Twl-H1 or Twl-H1-miR-9 and treated with repeated CORT (or DMSO). one-way ANOVA followed by Tukey's post-hoc test: *p < 0.05 vs Twl-H1+DMSO. n = 12–14. (B) Total dendritic length analysis of primary hippocampal neurons transfected with Twl-H1, Twl-H1-shREST-1 or Twl-H1-shREST-2 and treated with repeated CORT (or DMSO). one-way ANOVA followed by Tukey's post-hoc test: **p < 0.01 vs Twl-H1+DMSO. n = 12–14. Dendritic length was measured using plug-in Fiji.