The effects of prolonged physical training on memory performance and underlying presynaptic mechanisms were investigated in old C57BL/6 mice. Training via voluntary running wheels was initiated at 16 months of age and continued for 5 months (1 h per day, 5 days per week), followed by testing of learning and memory functions and counting of presynaptic puncta and cholinergic inputs in the hippocampus. Trained old mice were compared to their age-matched sedentary controls and adult controls. This training strategy improved hippocampal-dependent spatial memory function tested via a novel location task, and enhanced memory was accompanied by restored presynaptic puncta and cholinergic fibers in area CA1 and DG of the hippocampus in old mice. Particularly, the training selectively affected presynaptic vesicle protein synaptophysin but not growth associated protein GAP-43, and the increased number of synaptophysin puncta positively correlates with improved memory performance. To better understand the neurochemical mechanisms by which prolonged physical training protects against aging-related memory deficits, the cholinergic inputs to the hippocampus were compared among the three groups of mice and correlated with memory performance. While the running prevented age-related loss of cholinergic inputs, it has limited impact on the projection source cells in the medial septum-diagonal band (MS-DB). Importantly, cholinergic fibers in area CA1 and DG positively correlated with spatial memory function. These data suggest that the preservation of presynaptic inputs, particularly those involved in the integrity of memory performance, contributes critically to the beneficial effects of physical running initiated at an older age.

Netrin-1 functions largely via combined receptors and downstream effectors. Evidence has shown that astrocytes express netrin-1 receptors, including DCC and UNC5H2. However, whether netrin-1 influences the function of astrocytes was previously unknown. Lipopolysaccharide was used to stimulate the primary cultured astrocytes; interleukin release was used to track astrocyte activation. In vivo, shRNA and netrin-1 protein were injected in the mouse brain. Infarct volume, astrocyte activation, and interleukin release were used to observe the function of netrin-1 in neuroinflammation and brain injury after middle cerebral artery occlusion. Our results demonstrated that netrin-1 reduced lipopolysaccharide-induced interleukin-1β and interleukin-12β release in cultured astrocytes, and blockade of the UNC5H2 receptor with an antibody reversed this effect. Additionally, netrin-1 increased p-AKT and PPAR-γ expression in primary cultured astrocytes. In vivo studies showed that knockdown of netrin-1 increased astrocyte activation in the mouse brain after middle cerebral artery occlusion (p < 0.05). Moreover, injection of netrin-1 attenuated GFAP expression (netrin-1 0.27 ± 0.06 vs. BSA 0.62 ± 0.04, p < 0.001) and the release of interleukins and reduced infarct volume after brain ischemia (netrin-1 0.27 ± 0.06 vs. BSA 0.62 ± 0.04 mm3, p < 0.05). Our results indicate that netrin-1 is an important molecule in regulating astrocyte activation and neuroinflammation in cerebral ischemia and provides a potential target for ischemic stroke therapy.

Memory loss is the key symptom of Alzheimer's disease (AD). As successful drug treatments have not yet been identified, non-pharmaceutical interventions such as physical exercise and training have been employed to improve the memory function of people with dementia. We investigated the effect of prolonged physical running on hippocampal-dependent spatial memory and its underlying mechanisms using a well-established rodent model of AD. 3xTg-AD transgenic mice and non-transgenic mice were subjected to voluntary wheel running for 5 months (1 hour per day, 5 days per week), followed by spatial memory testing. After the behavioral testing, dendritic spines, synapses, and synaptic proteins as well as amyloid-beta (Aβ) pathology were analyzed in the dorsal hippocampi. Running improved hippocampal-dependent spatial memory in 3xTg-AD mice. This running strategy prevented both thin and mushroom-type spines on CA1 pyramidal cells in 3xTg-AD mice, whereas the effects of running in non-transgenic mice were limited to thin spines. The enormous effects of running on spines were accompanied by an increased number of synapses and upregulated expression of synaptic proteins. Notably, running downregulated the processing of amyloid precursor protein, decreasing intracellular APP expression and extracellular Aβ accumulation, and spatial memory performance correlated with levels of Aβ peptides Aβ_1-40 and Aβ_1-42. These data suggest that prolonged running may improve memory in preclinical AD via slowing down the amyloid pathology and preventing the loss of synaptic contacts.

Adverse experiences in early life can shape neuronal structures and synaptic function in multiple brain regions, leading to deficits of distinct cognitive functions later in life. Focusing on the pyramidal cells of the prelimbic cortex (PrL), a main subregion of the medial prefrontal cortex, the impact of early-life adversity (ELA) was investigated in a well-established animal model generated by changing the rearing environment during postnatal days 2 to 9 (P2-P9), a sensitive developmental period. ELA has enduring detrimental impacts on the dendritic spines of PrL pyramidal cells, which is most apparent in a spatially circumscribed region. Specifically, ELA affects both thin and mushroom-type spines, and ELA-provoked loss of spines is observed on selective dendritic segments of PrL pyramidal cells in layers II-III and V-VI. Reduced postsynaptic puncta represented by postsynaptic density protein-95 (PSD-95), but not synaptophysin-labelled presynaptic puncta, in ELA mice supports the selective loss of spines in the PrL. Correlation analysis indicates that loss of spines and postsynaptic puncta in the PrL contributes to the poor spatial working memory of ELA mice, and thin spines may play a major role in working memory performance. To further understand whether loss of spines affects glutamatergic transmission, AMPA- and NMDA-receptor-mediated synaptic currents (EPSCs) were recorded in a group of Thy1-expressing PrL pyramidal cells. ELA mice exhibited a depressed glutamatergic transmission, which is accompanied with a decreased expression of GluR1 and NR1 subunits in the PrL. Finally, upregulating the activation of Thy1-expressing PrL pyramidal cells via excitatory DREADDs can efficiently improve the working memory performance of ELA mice in a T-maze-based task, indicating the potential of a chemogenetic approach in restoring ELA-provoked memory deficits.

Pluripotent stem cells (PSCs) are regarded as potential sources that provide specific neural cells for cell therapy in some nervous system diseases. However, the mechanisms underlying the neural differentiation of PSCs remain largely unknown. MicroRNAs (miRNAs or miRs) are a class of small non‑pro­tein-coding RNAs that act as critical regulatory molecules in many cellular processes. In this study, we found that miR‑146b‑5p expression was markedly increased following the neural induction of mouse embryonic stem cells (ESCs) or induced PSCs (iPSCs). In this study, to further identify the role of miR‑146b‑5p, we generated stable miR‑146b‑5p-overexpressing ESC and iPSC cell lines, and induced the differentiation of these cells by the adherent monolayer culture method. In the miR‑146b‑5p-overexpressing ESC- or iPSC-derived cultures, RT-qPCR analysis revealed that the mRNA expression levels of neuroectoderm markers, such as Sox1, Nestin and Pax6, were markedly increased, and flow cytometric analysis verified that the number of Nestin‑positive cells was higher in the miR‑146b‑5p-overexpressing compared with the control cells. Mechanistically, the miR‑146b‑5p-overexpressing ESCs or iPSCs exhibited a significant reduction in Oct4 expression, which may be an explanation for these cells having a tendency to differentiate towards the neural lineage. Moreover, we confirmed that miR‑146b‑5p directly targeted Smad4 and negatively regulated the transforming growth factor (TGF)-β signaling pathway, which contributed to the neural commitment of PSCs. Collectively, our findings uncover the essential role of miR‑146b‑5p in the neural conversion of PSCs.