Extracellular Vesicle-Derived miR-124 Resolves Radiation-Induced Brain Injury

Abstract

Radiation-induced cognitive dysfunction (RICD) is a progressive and debilitating health issue facing patients following cranial radiotherapy to control central nervous system cancers. There has been some success treating RICD in rodents using human neural stem cell (hNSC) transplantation, but the procedure is invasive, requires immunosuppression, and could cause other complications such as teratoma formation. Extracellular vesicles (EV) are nanoscale membrane-bound structures that contain biological contents including mRNA, miRNA, proteins, and lipids that can be readily isolated from conditioned culture media. It has been previously shown that hNSC-derived EV resolves RICD following cranial irradiation using an immunocompromised rodent model. Here, we use immunocompetent wild-type mice to show that hNSC-derived EV treatment administered either intravenously via retro-orbital vein injection or via intracranial transplantation can ameliorate cognitive deficits following 9 Gy head-only irradiation. Cognitive function assessed on the novel place recognition, novel object recognition, and temporal order tasks was not only improved at early (5 weeks) but also at delayed (6 months) postirradiation times with just a single EV treatment. Improved behavioral outcomes were also associated with reduced neuroinflammation as measured by a reduction in activated microglia. To identify the mechanism of action, analysis of EV cargo implicated miRNA (miR-124) as a potential candidate in the mitigation of RICD. Furthermore, viral vector-mediated overexpression of miR-124 in the irradiated brain ameliorated RICD and reduced microglial activation. Our findings demonstrate for the first time that systemic administration of hNSC-derived EV abrogates RICD and neuroinflammation in cranially irradiated wild-type rodents through a mechanism involving miR-124. SIGNIFICANCE: Radiation-induced neurocognitive decrements in immunocompetent mice can be resolved by systemic delivery of hNSC-derived EVs involving a mechanism dependent on expression of miR-124.

Figure 1.. Stem cell-derived EV protect against radiation-induced cognitive dysfunction at five weeks and six months post irradiation.

The experimental design for the present studies is shown (A). Four-month-old male C57Bl/6J mice were immobilized and subjected to 9 Gy cranial irradiation using a ¹³⁷Cs γ irradiator at a dose rate of 2.07 Gy/min. Two days later mice were treated intracranially or retro-orbitally with EV or miR-124 AAV9 particles. At five weeks and six months post-irradiation, animals were administered spontaneous exploration tasks in the following order: NPR (B, E), NOR (C, F) and TO (D, G). The tendency to explore novelty (novel place or object) was calculated using the discrimination index [(novel location exploration time/total exploration time) – (familiar location exploration time/total exploration time)] × 100. All data are presented as mean ± SEM (N=10–14 mice per group). * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001 compared to the IRR group; P values are derived from one-way ANOVA and Dunnett’s test for multiple comparisons.

Figure 2.. Stem cell-derived EV tracked to the host hippocampus following retro-orbital or intracranial injections.

Fluorescently labeled hNSC-derived EV were transplanted using stereotactic intracranial (IC) or retro-orbital (RO) injections. Brain tissue was fixed at 48 hours post-surgery and brain sections were imaged using confocal microscopy. Analysis suggests that IC injected EV (A, C) and RO injected EV (B, D) are similarly effective in targeting the dentate gyrus (DG) (A, B) and CA1 (C, D) regions of the hippocampus. Fluorescently labeled EV membranes, red; DAPI nuclear counterstain, blue. Scale bars = 30 μm (retro-orbital method), 40 μm (intracranial method). dh, dentate hilus; gcl, granule cell layer; sr, striatum radiatum; pyr, pyramidal cell layer

Figure 3.. Stem cell-derived EV treatment reduces neuroinflammation in the hippocampus following irradiation.

Representative images of CD68⁺ activated microglia are shown from the dentate gyrus (DG) region of the hippocampus in all four groups for the five-week behavioral testing cohort. Relative to controls (A) the DG region of the hippocampus from irradiated mice show elevated levels of CD68 (B). EV treatment reduces CD68 levels in the irradiated brain (C and D; intracranial (IC) and retro-orbital (RO), respectively). Aggregate data from image processing with Imaris shows an increased volume of staining in the irradiated group compared to the control and EV-treated groups in the DG region in both the five-week (E) and six-month (G) cohorts. The same analysis showed similar trends in the CA1 region of the hippocampus in both the five-week (F) and six-month (H) cohorts. All data are presented as mean ± SEM (N = 4–6 mice per group). # P = 0.061, * P < 0.05, **** P < 0.0001 compared to the IRR group; P values are derived from ANOVA and Dunnett’s multiple comparisons test (all other groups compared to IRR group). CD68, red; DAPI nuclear counterstain, blue. Scale bars = 30 μm. dh, dentate hilus; gcl, granule cell layer

Figure 4.. hNSC-derived EV contain candidate miRNA that may mitigate RICD.

Total RNA extracted from therapeutic EV was analyzed by miRNA microarray. Select examples from array data (A) included four miRNAs with significant literature suggesting roles for them in synaptic function, dendrite outgrowth, and reduction in neuroinflammation. Of these four, three miRNAs (hsa-miR-124–3p, hsa-miR-125a-5p, and hsa-miR-125b-5p) could be validated using Taqman Advanced miRNA Assays (B, C, D).

Figure 5.. miR-124 overexpression in vivo shows functional mitigation of RICD.

To determine whether miR-124 was sufficient to mitigate RICD in wild type mice, the miRNA sequence was cloned into a vector designed to overexpress miR-124 and the construct packaged into AAV9 particles. The vector map (A, designed by Signagen) shows that miR-124 expression is driven by the U6 promoter and eGFP expression is driven by the CMV enhancer and promoter. These transgenes are flanked by AAV2 inverted terminal repeats (ITR) for efficient propagation of the AAV genome. Mice received stereotaxic IC injections of AAV9 particles containing this vector two days post-irradiation. At five weeks post-irradiation, animals were administered spontaneous exploration tasks in the following order: NPR (B), NOR (C), and TO (D). Tendency to explore novelty (novel place or object) was calculated using the discrimination index [(novel location exploration time/total exploration time) – (familiar location exploration time/total exploration time)] × 100. All data are presented as mean ± SEM (N=10–11 mice per group). # P = 0.0724, * P < 0.05, ** P < 0.01 compared to the IRR group. P values are derived from one-way ANOVA and Dunnett’s test for multiple comparisons (all other groups compared to the 9 Gy + miR-Scr group).

Figure 6.. Reporter gene confirmation and tracking of AAV9-miR-124 in vivo.

The AAV9 vector designed to express either intact or scrambled miR-124 carried the eGFP reporter gene to enable construct visualization in vivo. After completion of behavior (8–10 weeks post-surgery) coronal brain sections were imaged for the presence of eGFP signal. Widespread expression of vector (green) was found in the cortex (A, layers IV, V and VI), corpus callosum (B, CC) and hippocampus (B, CA1; pyr, pyramidal layer; sr, stratum radiatum). Vector expression was evident in the cells resembling neuronal (C, arrows) and glial morphologies (D, arrows). Scale bars: 100 μm, A-B and 20 μm, C-D.

Figure 7.. miR-124 overexpression in vivo following cranial irradiation reduces neuroinflammation in the hippocampus.

Representative images of Iba1⁺ microglia and CD68⁺ activated microglia immunohistochemistry are shown from the hippocampus and dentate gyrus (DG), respectively, for the 9 Gy miR-Scr and the 9 Gy miR-124 groups of the miR-124 cohort. Relative to miR-124-overexpressing group (B), the mice in the 9 Gy miR-Scr group show decreased numbers of Iba1⁺ cells in hippocampal subfields (A). miR-124 overexpression (D) resulted in a relative decrease in the CD68⁺ immunoreactivity in the DG region when compared to the 9 Gy miR-Scr group (C). Aggregate data from image processing with Imaris shows an increased Iba1-adjusted (E) volume of CD68 immunoreactivity in the irradiated group compared to the control and miR-124 overexpressing groups in the hippocampus (F). All data are presented as mean ± SEM (N = 4 mice per group). ** P < 0.01, *** P < 0.001 **** P < 0.0001 compared to the IRR group; P values are derived from ANOVA and Dunnett’s multiple comparisons test. Iba1, red; CD68, red; DAPI nuclear counterstain, blue. Scale bars = 150 μm (A, B) and 40 μm (C, D). dh, dentate hilus; gcl, granule cell layer.