Small RNA shields the blood–brain barrier by silencing DDX58 after injury

A microRNA carried by stem-cell-derived exosomes protects the blood–brain barrier after ischemic injury by targeting and suppressing DDX58, a cytoplasmic pattern recognition receptor, according to a study published in Molecular Biology Reports. The research team demonstrated through bioinformatics prediction and dual luciferase reporter assays that miR-21a-5p binds directly to DDX58 sequences, and that this interaction is associated with reduced activation of the Toll-like receptor 2/nuclear factor kappa B signaling pathway in brain microvascular endothelial cells subjected to oxygen-glucose deprivation and reoxygenation. The findings identify DDX58 as a key intermediary target through which miR-21a-5p preserves tight junction integrity and limits matrix metalloproteinase activity in cellular models of barrier disruption.

The blood–brain barrier is a highly selective interface within the brain, composed of endothelial cells, pericytes, and glial cells, that functions as a crucial filter protecting the brain from harmful substances in the bloodstream while maintaining neural homeostasis. The endothelial cells that form the barrier’s structural and functional core are central to its selective permeability, and early markers of barrier dysfunction include degradation of the tight junction glycocalyx on endothelial cells. These tight junctions play an essential role in maintaining barrier integrity, and their disruption is often an early sign of breakdown. Endothelial cells also maintain interconnectivity through direct interactions with astrocytes, which facilitate the formation of large multiprotein complexes. Among these, the transmembrane tight junction proteins claudin-5 and occludin, as well as the auxiliary protein ZO-1, which links to the actin cytoskeleton, are critical for maintaining the barrier’s integrity and selective filtration function. Matrix metalloproteinases, particularly MMP-2 and MMP-9, contribute to barrier damage by degrading tight junction proteins, exacerbating disruption.

Inflammation-driven mechanisms of barrier injury are central to the progression of various conditions, including ischemic injury. Both peripheral inflammatory mediators and inflammation within brain endothelial cells contribute to barrier disruption, highlighting inflammation and tight junction degradation as key risk factors for damage. Impaired barrier function is implicated in numerous diseases, such as aging, neuroimmunological disorders, and cerebrovascular diseases. Developing targeted strategies to repair and maintain barrier integrity is therefore of critical importance. The research team’s previous work established that exosomes derived from bone marrow mesenchymal stem cells carrying miR-21a-5p can mitigate barrier disruption and hemorrhagic transformation, but the specific molecular targets through which this microRNA exerts protective effects remained unclear.

MicroRNA-21a-5p is a multifunctional microRNA implicated in a variety of diseases, including traumatic spinal cord injury inflammation, periodontitis, acute lung injury, diabetes, and renal fibrosis. DDX58 encodes retinoic acid-inducible gene I, a cytoplasmic pattern recognition receptor that plays a pivotal role in immune responses by recognizing viral RNA. DDX58 has been implicated in various diseases, including colorectal cancer, lupus nephritis, tumors, and hepatocellular carcinoma, though its role in blood–brain barrier function had not been previously explored. To identify whether DDX58 serves as a direct target of miR-21a-5p, the researchers employed bioinformatics analysis using the Starbase database to predict targeted binding between the microRNA and DDX58 sequences.

The predicted interaction was validated through dual luciferase reporter assays performed in human renal embryonic kidney 293T cells. The researchers constructed plasmids containing either wild-type or mutant DDX58 sequences cloned downstream of a luciferase reporter gene. When miR-21a-5p mimics were co-transfected with the wild-type DDX58 plasmid, luciferase activity was specifically suppressed, confirming direct binding. In contrast, co-transfection with the mutant DDX58 plasmid produced no significant reduction in luciferase activity, demonstrating that the microRNA requires the intact binding site to exert its regulatory effect. Three independent biological transfections were performed, each with three technical replicate wells, and the results were considered statistically significant at p < 0.05 following one-way ANOVA and Tukey HSD post hoc comparisons.

Expression analysis of DDX58 in mouse models of middle cerebral artery occlusion followed by reperfusion revealed significant upregulation of DDX58 in ischemia/reperfusion animals compared with sham controls, which was markedly reduced following treatment with bone marrow mesenchymal stem cell exosomes. The in vivo samples used in this study were archived specimens from the team’s previous research and were subjected to secondary molecular analysis. In an in vitro blood–brain barrier injury model induced by oxygen-glucose deprivation and reoxygenation, exosome treatment significantly decreased both mRNA and protein levels of DDX58. Inhibition of miR-21a-5p expression led to an increase in DDX58 expression, whereas miR-21a-5p overexpression resulted in the opposite effect. These findings suggest that exosomes transport miR-21a-5p into brain microvascular endothelial cells, targeting and regulating DDX58 in vitro to modulate cellular phenotypes.

To explore how DDX58 affects endothelial cell behavior and contributes to barrier injury, the researchers performed overexpression and interference experiments in an immortalized mouse brain microvascular endothelial cell line co-cultured with primary astrocytes in Transwell filters. The team used short hairpin RNA sequences to silence DDX58, with sh-DDX58#2 demonstrating the highest knockdown efficiency. Overexpression of DDX58 significantly impaired endothelial cell metabolic activity as measured by CCK-8 assay, whereas DDX58 silencing enhanced metabolic activity. Mechanistically, although DDX58 did not significantly affect total NF-κB protein levels, overexpression of DDX58 led to increased expression of TLR2 and phosphorylated p65, while DDX58 knockdown inhibited these effects. Overexpression of DDX58 also significantly downregulated the tight junction proteins claudin-5, occludin, and ZO-1, while knockdown reversed this reduction.

Overexpression of DDX58 significantly increased MMP-2 and MMP-9 expression and enzymatic activity, whereas knockdown of DDX58 inhibited both expression and activity of these matrix metalloproteinases. Overexpression of DDX58 also significantly reduced the transepithelial electrical resistance of the co-culture model and increased permeability to fluorescent yellow tracer, indicating compromised barrier function. Immunofluorescence analysis of DDX58 and ZO-1 expression corroborated these protein-level findings. These results indicate that DDX58 manipulation is associated with changes in TLR2/NF-κB-related markers, increased matrix metalloproteinase expression, and reduced tight junction protein expression, suggesting a promoting role in barrier damage within the cellular model.

To validate the mechanism by which miR-21a-5p targets DDX58 and improves barrier integrity, the researchers performed rescue experiments in the in vitro injury model. The results confirmed that DDX58 overexpression did not significantly affect miR-21a-5p RNA levels, establishing that DDX58 is downstream of the microRNA. Treatment with miR-21a-5p mimics significantly suppressed DDX58 expression and promoted endothelial cell metabolic activity, while simultaneous DDX58 overexpression reversed these effects. miR-21a-5p treatment significantly inhibited the TLR2/NF-κB pathway, whereas co-overexpression of DDX58 during miR-21a-5p treatment reactivated the pathway. miR-21a-5p treatment upregulated the expression of claudin-5, occludin, and ZO-1, while DDX58 overexpression suppressed this upregulation. miR-21a-5p treatment also significantly decreased MMP-2 and MMP-9 expression and enzymatic activity, increased transepithelial electrical resistance, and reduced permeability, with DDX58 overexpression reversing each of these protective changes.

The researchers note several important limitations. Although the effects of miR-21a-5p on barrier injury were validated both in vitro and in vivo, the specific mechanism by which miR-21a-5p targets DDX58 was primarily explored in cellular models, and direct in vivo manipulation of DDX58 was not performed. The causal mediation of barrier protection by DDX58 in vivo cannot be concluded from the present data, and the in vivo validation was limited to expression analysis. The relationship between DDX58 and TLR2 is not a simple linear upstream-downstream hierarchy, as previous studies have demonstrated that DDX58 and TLR2 form a complex bidirectional regulatory crosstalk network. In the oxygen-glucose deprivation and reoxygenation inflammatory microenvironment, DDX58 may indirectly upregulate TLR2 transcription through NF-κB positive feedback loops, or alternatively achieve pathway crosstalk via inflammasome and cytokine-mediated paracrine effects. These possibilities warrant further mechanistic investigation to delineate the precise mode of DDX58-TLR2 interaction in barrier endothelial cells under ischemic stress. The uptake process of the exosomes was not independently and repeatedly verified in this study, and the relevant conclusions rely on characterization data from previous research and functional consistency inference. The study lays groundwork for developing targeted therapeutic strategies, including precision delivery systems featuring a triple modification of miRNA, exosome, and targeted peptides, though multi-center and multi-species studies will be crucial for advancing clinical translation of the miR-21a-5p-DDX58-TLR2/NF-κB regulatory axis.

Subject of Research: The role of miR-21a-5p and its target DDX58 in blood–brain barrier repair after ischemic injury

Article Title: miR-21a-5p targets DDX58 to suppress TLR2/NF-κB signaling and protect the blood–brain barrier after injury

Article References: Yang, J., Tang, N., Dai, R., Fu, H., Li, S., & Meng, Q. (2026). miR-21a-5p targets DDX58 to suppress TLR2/NF-κB signaling and protect the blood–brain barrier after injury. Molecular Biology Reports, 53(1), Article 1812. https://doi.org/10.1007/s11033-026-13010-3

Image Credits: AI Generated

DOI: 10.1007/s11033-026-13010-3

Keywords: miR-21a-5p, DDX58, blood-brain barrier, TLR2, NF-κB, tight junctions, claudin-5, occludin, ZO-1, MMP-2, MMP-9, exosomes