microRNAs shuttled by extracellular vesicles derived from mesenchymal stem cells rescue glial activation in in-vitro models of amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a progressive, fatal, neurodegenerative disease characterized by the selective death of upper and lower motor neurons (MNs). The mechanism of MN damage and death has been ascribed to several cellular and molecular alterations, including neuroinflammation. ALS is also a non-cell-autonomous disease, due to the contribution of glial cells, such as astrocytes and microglia, that secrete neurotoxic factors and promote a noxious environment for MN. At present there is no effective cure for ALS. We previously demonstrated that intravenous administration of bone marrow mesenchymal stem cells (MSCs), prolonged survival probability, improved motor functions and ameliorated pathological features, including gliosis and neuroinflammation, in the spinal cord of the SOD1G93A mouse model of ALS. There is a consensus supporting that the beneficial outcomes of MSCs are unlikely due to trans-differentiation, but possibly to paracrine effects, thus we postulated that one of these mechanisms could be the transfer to target cells of microRNAs (miRNAs) shuttled by extracellular vesicles (EVs) isolated from the secretome of MSCs. To this aim we studied the activity of MSCs-derived EVs both on astrocytes isolated from the spinal cord of symptomatic SOD1G93A mice and human astrocytes (iAstrocytes) differentiated from inducible neural progenitor cells (iNPCs) of ALS patients. Mouse or human-derived ALS astrocytes were exposed to MSCs-derived EVs or transfected with single miRNA mimics. We determined protein expression by western blot, confocal microscopy and qPCR, while cytokine release was measured by ELISA assay. Co-cultures of spinal MNs with ALS astrocytes, exposed or not to EVs, were studied to assess MN viability. The astrocyte activation markers vimentin, GFAP, and S100β were overexpressed in SOD1G93A mice, and this over-expression was reduced by exposure to EVs. The same was true for the pro-inflammatory cytokines TNF-α, IL-1β, IL-6 and CCL2 that were aberrantly secreted from adult mouse SOD1G93A astrocytes, and which secretion was reduced in astrocytes treated with EVs. In human iAstrocytes, exposure to EVs increased the expression of the Nrf2 anti-oxidant factor and resulted in reduced accumulation of reactive oxygen species. Most importantly, the viability of MNs was significantly increased when co-cultured with mouse or human astrocytes previously exposed to MSCs-derived EVs. Aiming to identify possible factors that could be responsible for the neuroprotective effect of the MSC-derived EVs, we focused on microRNAs (miRNAs), which we found to be elevated in IFN-γ-primed MSCs. Of note, the transfection with specific mimics of miRNAs reverted the reactive phenotype of ALS astrocytes cultured from SOD1G93A mice. Similarly, in human iAstrocytes, transfection with the miR-29b-3p miRNA significantly upregulated the Nrf2 antioxidant pathway and rescued the viability of co-cultured MNs. Interestingly, MSCs-derived EVs also ameliorated the neuroinflammatory phenotype of microglia cells isolated from symptomatic SOD1G93A mice. Overall, our data suggest that EVs derived from MSCs represent a promising therapeutic strategy in ALS by releasing EV-shuttled anti-inflammatory and anti-oxidant miRNAs able to decrease glial cell toxicity towards motor neurons. In-vivo pre-clinical studies in SOD1G93A mice are ongoing aimed at gathering a crucial proof-of-concept to translate our in-vitro results into effective clinical trials.