Post-Stroke Environmental Enrichment Improves Neurogenesis and Cognitive Function and Reduces the Generation of Aberrant Neurons in the Mouse Hippocampus

Abstract

Ischemic lesions stimulate adult neurogenesis in the dentate gyrus, however, this is not associated with better cognitive function. Furthermore, increased neurogenesis is associated with the formation of aberrant neurons. In a previous study, we showed that a running task after a stroke not only increases neurogenesis but also the number of aberrant neurons without improving general performance. Here, we determined whether stimulation in an enriched environment after a lesion could increase neurogenesis and cognitive function without enhancing the number of aberrant neurons. After an ischemic stroke induced by MCAO, animals were transferred to an enriched environment containing a running wheel, tunnels and nest materials. A GFP-retroviral vector was delivered on day 3 post-stroke and a modified water maze test was performed 6 weeks after the lesion. We found that the enriched environment significantly increased the number of new neurons compared with the unstimulated stroke group but not the number of aberrant cells after a lesion. Increased neurogenesis after environmental enrichment was associated with improved cognitive function. Our study showed that early placement in an enriched environment after a stroke lesion markedly increased neurogenesis and flexible learning but not the formation of aberrant neurons, indicating that rehabilitative training, as a combination of running wheel training and enriched environment housing, improved functional and structural outcomes after a stroke.

Keywords: MCAO; aberrant neurogenesis; enriched environment.

Figure 2

Impact of MCAO and EE on (A) hippocampal volume, (B) brain volume and (C) infarct volume. Significant differences were found in the hippocampal volume between sham and MCAO under standard conditions, both ipsi- and contralateral, 7 weeks post-surgery. The boxplots show the median, upper and lower quartile, and minimum and maximum values. Statistical analyses were performed using the Mann–Whitney U test.

Figure 1

Experimental design. All animals were randomly assigned to 4 groups: 2 groups underwent a sham surgery and 2 groups an MCAO with survival period of 2 or 7 weeks. Subsequently, animals were transferred to a standard cage (Sham-SD; MCAO-SD) or to an EE housing (Sham-EE; MCAO-EE). Mice with a survival period of 7 weeks were injected with the proliferation marker EdU over a period of 3 to 15 days post-surgery and, on day 4, with the retroviral vector. At six weeks post-surgery, the spatial learning abilities were tested with the Morris water maze task over a period of 5 days. This was followed by transcardial perfusion, tissue collection and analysis. Created with BioRender.com.

Figure 3

Proliferation of MCM2⁺ cells after stroke and EE housing. (A) Peroxidase-stained images of MCM2⁺ cells in the dentate gyrus. Dark brownish cells were identified as MCM2⁺ cells. (B) Significant changes in the endogenous proliferation in the dentate gyrus were detected under standard housing conditions. Statistical analysis was performed using the Mann–Whitney U test, scale bar = 100 µm, higher magnification 10 µm.

Figure 4

Impact of stroke and EE on DCX⁺ cell proliferation. (A) Peroxidase-stained brain slices for DCX 2 and 7 weeks after stroke. (B) Quantification of DCX⁺ cells in the dentate gyrus. The absolute number of DCX⁺ cells significantly decreased after stroke in the standard but not in the EE groups. Statistical analysis was performed using the ANOVA and post-hoc Bonferroni, MW ± SEM; scale bar = 100 µm. (C) Exemplary immunofluorescence staining for MCM2 (red) and DCX (green) after MCAO in standard and EE housing groups. (D) Percentage of proliferating DCX⁺ cells from all proliferating cells in the dentate gyrus. Statistical analysis was performed using the Mann–Whitney U test, scale bar = 100 µm, higher magnification 10 µm.

Figure 5

Stroke and EE-induced neurogenesis. (A,B) MCAO induction significantly increased new neuron formation 7 weeks later. EE significantly increases the number of newly generated neurons in both sham and MCAO mice. (C) This increase occured ipsi- and contralaterally in all groups. The boxplots show the median, upper and lower quartile, and minimum and maximum values. Statistical analysis was performed using the Mann-Whitney U test, scale bar = 50 µm. (D) Aberrant GFP-labelled new neurons with two dendritic trees (white arrows), one dendritic tree extending into the hilus (GFP green, NeuN blue). Scale bar = 10 µm. For the morphological analyses of the newly born neurons, we determined the following numbers of cells: MCAO-SD n = 20; MCAO-sham n = 28; MCAO-EE n = 24; sham-EE n = 9. Scale bar = 10 µm.

Figure 6

Morphological analysis of newly generated neurons in the dentate gyrus. (A) Overview images GFP-labelled neurons in the dentate gyrus of an EE-MCAO animal (GFP green, NeuN purple). Scale bar = 100 µm (B) Representative confocal images of 42 days GFP-positive neurons in the different groups post-surgery. Scale bar = 50 µm. (C–E) Dendritic length and complexity of newly generated GFP-positive neurons in the different groups. No significant differences were found between sham and MCAO groups. Statistical analysis was performed using linear mixed model for dendritic length and One-way ANOVA with post-hoc Bonferroni for dendritic complexity.

Figure 7

Learning and re-learning in dependence of stroke and EE housing. The graphs show (A–D) latency and (E–H) distance of navigation to the hidden platform in the Morris water maze. (A,E) Impairments in learning were evident in the standard housing group when comparing sham and MCAO. (B,F) By re-learning the new platform position, sham-EE animals, in particular, showed an improvement compared with the MCAO-EE mice. (C,G) There were no differences between the sham groups. (D,H) Between the MCAO groups in the different housing conditions, there was a significant impairment in learning on day 1 in the MCAO-SD group. Graphs show mean ± SEM using 2-way ANOVA with repeated measurements and post-hoc Bonferroni (per group and per day), * p < 0.05.

Figure 8

The trial was conducted on (A) day 4 (before the platform change) and (B) day 5 (after the platform change). The platform was removed from the pool for the probe trail, and the time spent in the different quadrants was determined. (A) All animals learned the first platform position in the quadrant SE without significant differences between the groups. (B) On day 5, all groups showed a higher preference for the new platform position. There was no impairment in spatial learning during the probe trial. The boxplots show the median, the upper and lower quartiles and the minimum and maximum values. The analysis was performed with the Mann-Whitney U test, all p- and n-values. SD = standard housing; EE = enriched environment housing.

Figure 9

Analysis of hippocampus-dependent and independent search strategies in the MWM test. (A,C,E) Representation of hippocampus-dependent strategies over time. The data show an improved learning behavior in the EE sham group compared with the EE MCAO. Furthermore, there was more frequent use of the hippocampus-dependent strategies on day 1 in the MCAO-EE group compared with the MCAO-SD group. (B) Within the hippocampus-independent strategies, the MCAO-SD animals showed increased use of the scanning strategy on day 5. For analysis of the different hippocampal-dependent and -independent search strategies on each day, an exploratory data analysis by means of an algorithm based on the generalized estimating equations method was performed. Significant differences in strategies usage are represented by colorful squares for each day, * p < 0.05.