Evaluation of electroacupuncture as a non-pharmacological therapy for astrocytic structural aberrations and behavioral deficits in a post-ischemic depression model in mice

Abstract

Background: Ascending clinical evidence supports that electroacupuncture (EA) is effective in treating post-ischemic depression (PID), but little is known about how it works at the cellular level. Astrocytes are exquisitely sensitive to their extracellular environment, and under stressful conditions, they may experience aberrant structural remodeling that can potentially cause neuroplastic disturbances and contribute to subsequent changes in mood or behavior.

Objectives: This study aimed to investigate the effect of EA on behavioral deficits associated with PID in mice and verify the hypothesis that astrocytic morphology may be involved in this impact.

Methods: We established a PID animal model induced by transient bilateral common carotid artery occlusion (BCCAO, 20 min) and chronic restraint stress (CRS, 21 days). EA treatment (GV20 + ST36) was performed for 3 weeks, from Monday to Friday each week. Depressive- and anxiety-like behaviors and sociability were evaluated using SPT, FST, EPM, and SIT. Immunohistochemistry combined with Sholl and cell morphological analysis was utilized to assess the process morphology of GFAP+ astrocytes in mood-related regions. The potential relationship between morphological changes in astrocytes and behavioral output was detected by correlation analysis.

Results: Behavioral assays demonstrated that EA treatment induced an overall reduction in behavioral deficits, as measured by the behavioral Z-score. Sholl and morphological analyses revealed that EA prevented the decline in cell complexity of astrocytes in the prefrontal cortex (PFC) and the CA1 region of the hippocampus, where astrocytes displayed evident deramification and atrophy of the branches. Eventually, the correlation analysis showed there was a relationship between behavioral emotionality and morphological changes.

Conclusion: Our findings imply that EA prevents both behavioral deficits and structural abnormalities in astrocytes in the PID model. The strong correlation between behavioral Z-scores and the observed morphological changes confirms the notion that the weakening of astrocytic processes may play a crucial role in depressive symptoms, and astrocytes could be a potential target of EA in the treatment of PID.

Keywords: BCCAO; astrocyte; bilateral common carotid artery occlusion; branch morphology; depressive-like behavior; electroacupuncture; post-ischemic depression; stress.

FIGURE 1

Schematic illustration of the experimental timeline and design. Male C57BL/6J mice were randomly assigned to three groups after 1 week of acclimation. Mice in each group have SPT adaption, SPT baseline measurement, and a preference test before the operation. The sham group was treated with a sham operation and sham EA (NA) treatment. The PID group was treated with BCCAO + CRS and sham EA (NA) treatment. The EA group was treated with BCCAO + CRS and EA treatment. The NS was applied 1 day before and 2 days after the operation to assess neurological deficits. Behavioral measures, including the SPT, SIT, EPM test, and FST, were conducted 21 days after CRS. SPT was recorded weekly for the entire experiment. After the behavioral tests, brain tissues were collected for further research. BCCAO, bilateral common carotid artery occlusion; CRS, chronic restraint stress; NS, neurological scoring; SPT, sucrose preference test; SIT, social interaction test; EPMT, elevated plus maze test; FST, forced swim test.

FIGURE 2

Behavioral performance and Z-scores. Dot-plot graphs show the individual values for the preference index of sucrose consumption (A), social interaction ratio (B), time spent in the open arms of the EPM (C), and time spent immobile in the FST (D). (E–H) Z-scores for the SPT (E), SIT (F), EPMT (G), and FST (H). (I) The overall Z-score of behavioral emotionality. All data are presented as mean ± SD. n = 9–12 mice per group. *p < 0.05, **p < 0.01, ***p < 0.001, and ns p > 0.05. This analysis was conducted using a one-way ANOVA.

FIGURE 3

The branching complexity of astrocytes in the PFC and hippocampal CA1 regions. The mean number of total Sholl intersections of astrocytes and a representative GFAP-positive cell from each group in the PFC (A) and the hippocampal CA1 (C). Those analyses were conducted using a one-way ANOVA. (B,D) Sholl distributive analyses of astrocytes in the PFC (B) and hippocampal CA1 region (D) in the sham group (black), PID group (red), and EA group (green). All data are presented as mean ± SD. *p < 0.05, **p < 0.01, and ***p < 0.001 compared to the sham group. #p < 0.05, ##p < 0.01, and ###p < 0.001 compared to the PID group. This analysis was conducted using a two-way ANOVA.

FIGURE 4

Astrocytic process morphology in the PFC and hippocampal CA1 regions. (A) Scenario illustration of an astrocyte. (B–D) Schematic representation of an astrocyte having similar branching complexity but smaller size (B), less branching complexity but similar size (C), or less branching complexity combined with smaller size (D), compared to the astrocyte in (A). (E–I) Schematic drawing of the astrocytic morphology for branch points and branch processes (E), endpoints (F), total length of processes (G), CHP (H), and CHA (I) per cell. (J,O) The total length of processes of astrocytes in the PFC (J) and the hippocampal CA1 region (O). (K,P) The number of endpoints (K) of astrocytes in the PFC (K) and the hippocampal CA1 region (P). (L,Q) The number of branch processes of astrocytes in the PFC (L) and the hippocampal CA1 region (Q). (M,R) CHP of astrocytes in the PFC (M) and the hippocampal CA1 region (R). (N,S) CHA of astrocytes in the PFC (N) and the hippocampal CA1 region (S). Data presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, and ns p > 0.05. This analysis was conducted using a one-way ANOVA.

FIGURE 5

The correlation between astrocyte morphology and behavioral performances. (A–C) Graphical representation of the correlation between the total length of processes (A), branches (B), and endpoints (C) of astrocytes in the PFC with behavioral Z-scores. (F) Graphical representation of the correlation between the total length of processes of astrocytes in the hippocampal CA1 with behavioral Z-scores. r and p-values were obtained following a Pearson correlation analysis. (D,G) The sum of astrocyte intersections at distal radiuses from 35 to 65 μm in the PFC (D) and at distal radiuses from 40 to 65 μm in the hippocampus CA1 (G). Data presented as mean ± SD. ***p < 0.001, one-way ANOVA. (E,H) The correlation between distal intersections of astrocytes in the PFC (E) and the hippocampal CA1 region (H) with behavioral Z-scores. r and p-values were obtained following a Pearson correlation analysis.