Microvesicles from brain-extract-treated mesenchymal stem cells improve neurological functions in a rat model of ischemic stroke

Abstract

Transplantation of mesenchymal stem cells (MSCs) was reported to improve functional outcomes in a rat model of ischemic stroke, and subsequent studies suggest that MSC-derived microvesicles (MVs) can replace the beneficial effects of MSCs. Here, we evaluated three different MSC-derived MVs, including MVs from untreated MSCs (MSC-MVs), MVs from MSCs treated with normal rat brain extract (NBE-MSC-MVs), and MVs from MSCs treated with stroke-injured rat brain extract (SBE-MSC-MVs), and tested their effects on ischemic brain injury induced by permanent middle cerebral artery occlusion (pMCAO) in rats. NBE-MSC-MVs and SBE-MSC-MVs had significantly greater efficacy than MSC-MVs for ameliorating ischemic brain injury with improved functional recovery. We found similar profiles of key signalling proteins in NBE-MSC-MVs and SBE-MSC-MVs, which account for their similar therapeutic efficacies. Immunohistochemical analyses suggest that brain-extract-treated MSC-MVs reduce inflammation, enhance angiogenesis, and increase endogenous neurogenesis in the rat brain. We performed mass spectrometry proteomic analyses and found that the total proteomes of brain-extract-treated MSC-MVs are highly enriched for known vesicular proteins. Notably, MSC-MV proteins upregulated by brain extracts tend to be modular for tissue repair pathways. We suggest that MSC-MV proteins stimulated by the brain microenvironment are paracrine effectors that enhance MSC therapy for stroke injury.

Figure 1. Transplantation of NBE-MSC-MVs and SBE-MSC-MVs improves neurological outcomes.

(A) Analysis of body weight changes after injection of MSC-MVs, NBE-MSC-MVs, or SBE-MSC-MVs. Loss of body weight was significantly attenuated in rat groups treated with NBE-MSC-MVs (P < 0.011, P < 0.02) or SBE-MSC-MVs (P < 0.025, P < 0.05) compared with a group treated with PBS at day 3 and 7, respectively. (B) Torso-twisting test (left and right side). Score change from PBS group for right side was significant for NBE-MSC-MVs (P < 0.04) at day 3, and for both NBE-MSC-MVs (P < 0.025) and SBE-MSC-MVs (P < 0.03) at day 7. (C) Open field test (line cross score). Score change from PBS group was significant for NBE-MSC-MVs (P < 0.025) and SBE-MSC-MVs (P < 0.03) at day 3. (D) Beam balance test. All injection groups (MSC-MVs, NBE-MSC-MVs, and SBE-MSC-MVs) showed significant improvement compared with the PBS group at day 3 (P < 0.01, P < 0.001, P < 0.001, respectively) and day 7 (P < 0.012, P < 0.0001, P < 0.04, respectively). (E) Prehensile traction test. The score was significantly increased in animals treated with MSC-MVs (P < 0.0002) and NBE-MSC-MVs (P < 0.05) compared with that of the PBS control group at day 3, and in NBE-MSC-MVs (P < 0.005) at day 7. (F) Modified neurological severity scores (mNSS) indicate that animals treated with NBE-MSC-MVs and SBE-MSC-MVs displayed significant functional enhancement at day 3 (P < 0.0001 and P < 0.002, respectively) compared with the PBS control group, but animals treated with MSC-MVs did not. At 7 days after injection, animals treated with MSC-MVs, NBE-MSC-MVs, and SBE-MSC-MVs showed significant functional enhancement compared with that of the PBS group (P < 0.002, P < 0.0001, and P < 0.01, respectively). *P < 0.05; **P < 0.01, ***P < 0.001; n = 10 per group; Day −1, time of pMCAO surgery; Day 0, time of MV injection; Day 3 and 7, 3 and 7 days after MV injection, respectively. Sample sizes are n = 5 for sham-operated, n = 10 for NBE-MSC-MV and SBE-MSC-MV treatment groups, and n = 7 for MSC-MV treatment group.

Figure 2. NBE-MSC-MV and SBE-MSC-MV injection reduces infarct size in a rat stroke model.

(A) Schematic diagram depicting the experimental design of this study. Numerals in the black bar represent the days of experiment. Day −1, time of pMCAO surgery; Day 0, time of MV injection; Day 3 and 7, 3 and 7 days after MV injection. Downward arrows indicate the days for behavioural tests and weight measurement. Upward arrows indicate the days for treatments and brain tissue sampling. (B) Representative TTC-stained brain sections from rats treated with PBS (control), MSC-MVs, NBE-MSC-MVs, and SBE-MSC-MVs show ischemic lesions on day 7 after MV treatment. (C) Bar graph depicting infarct size in all animals. The infarct area in the ipsilateral hemisphere is expressed as a percentage of the contralateral hemisphere area. Values are indicated as means ± SEM. **P < 0.01 compared with controls. Sample size n = 5 for sham control, n = 10 for NBE-MSC-MV and SBE-MSC-MV treatment groups, and n = 7 for MSC-MV treatment group.

Figure 3. NBE-MSC-MV administration increases neurogenesis and reduces reactive astrogliosis at the ischemic boundary in the rat brain.

(A) Compared with PBS control treatment, the number of DCX-positive cells at the ischemic boundary significantly increased immediately after NBE-MSC-MV treatment. *P < 0.05, **P < 0.01, ***P < 0.001, n = 5 per group. (B) Representative micrographs of α-SMA-positive vessels in ipsilateral hippocampal vessels at day 7. Compared with PBS control treatment, the number of α-SMA-positive vessels increased in the ipsilateral hippocampal area after NBE-MSC-MV treatment. Data are presented as mean numbers of α-SMA-positive vessels per group. *P < 0.05, n = 5 per group. (C) The number of GFAP-positive cells significantly decreased in response to NBE-MSC-MV transplantation compared with that of PBS control injection at day 7. Data are presented as percent of GFAP-positive area/field. *P < 0.05, n = 5 per group. Scale bar = 200 μm. Values are indicated as means ± SD.

Figure 4. Anti-inflammatory effects of NBE-MSC-MVs in ischemic rat brain.

(A) Representative RT-PCR results. (B) Quantification of inflammatory and anti-inflammatory gene expression. NBE-MSC-MVs reduced inflammatory factors and enhanced anti-inflammatory factors in ischemic rat brain tissue. Values are indicated as means ± SEM. *P < 0.05, **P < 0.01, n = 7 per group.

Figure 5. Functional analysis of the NBE-MSC-MV proteome.

(A) The proportions of vesicular proteins in the NBE-MSC-MV proteome supported by more than 1, 2, or 3 experimental batches and random proteomes (e.g., 10 random samplings of 1,000 proteins), indicating that the NBE-MSC-MV proteome is enriched for vesicular proteins annotated in the Vesiclepedia database. The degree of enrichment for vesicular proteins was higher for proteins identified in multiple NBE-MSC-MV batches. (B) Plots from ‘within-group connectivity’ and ‘sharing network neighbour’ analysis of the NBE-MSC-MV proteome for a functional network (HumanNet) and a protein-protein interaction network (HPRD). Box-and-whisker plots represent ‘within-group edge count’ and ‘network neighbours overlap’ scores for 1,000 random gene sets of equal size, and red dots represent scores for the NBE-MSC-MV proteome.

Figure 6. Proposed network of 236 NBE-MSC-MV proteins belonging to four functional categories involved in tissue repair.

Each node and edge of the network represents a human gene and a functional link, respectively. Functional links were derived from a co-functional network of human genes (HumanNet). We found that 236 of 272 genes associated with GO terms for angiogenesis, neurogenesis, anti-inflammation, and apoptosis are interconnected by functional links in the largest component of the network. The circumference of each node represents the proportion of proteins from each of the four functional categories: green for angiogenesis, blue for anti-inflammation, yellow for neurogenesis, and red for apoptosis. Upregulated NBE-MSC-MV proteins are shown as purple nodes. The size of the purple node represents the fold change.