Intravenous miR-30c therapy confers dual neurovascular protection and improves long-term recovery after ischemic stroke

Abstract

Ischemia–reperfusion (I/R) injury is a major barrier to effective reperfusion therapy in acute ischemic stroke. Secondary microvascular thrombosis and neuronal endoplasmic reticulum (ER) stress are major contributors to poor outcomes, yet therapeutic strategies that simultaneously target both processes remain limited. Here, we investigated the efficacy of an intravenously administered PEGylated liposome–encapsulated miR-30c mimic in a transient middle cerebral artery occlusion (MCAO) mouse model. Endogenous miR-30c was highly expressed in cortical and hippocampal neurons under physiological conditions but was markedly downregulated after ischemia. Post-stroke intravenous delivery restored brain miR-30c levels, reduced infarct volume, and improved neurological function across age and sex groups. Therapeutic benefit was observed within a clinically relevant 3–4.5 h treatment window and was sustained for up to 35 days post-stroke. Mechanistically, miR-30c suppressed endothelial PAI-1, thereby attenuating intravascular fibrin and platelet deposition and preserving microvascular patency, while concurrently attenuating neuronal ER stress and mitochondrial apoptotic signaling to reduce neuronal apoptosis. Together, these findings identify miR-30c as a dual-action therapeutic that confers neurovascular protection and promotes long-term functional recovery after ischemic stroke. Targeting both vascular and neuronal injury pathways may represent a promising strategy to enhance the efficacy of reperfusion therapies in patients with acute ischemic stroke.

Graphical abstract

Intravenous PEG-liposome–encapsulated miR-30c confers dual vascular and neuronal protection by suppressing endothelial PAI-1–mediated microvascular thrombosis and attenuating neuronal ER stress, leading to acute neuroprotection and long-term functional recovery after ischemic stroke.

Introduction

Ischemic stroke remains a leading cause of death and disability worldwide. Current reperfusion therapies, including intravenous thrombolysis and endovascular thrombectomy, aim to rapidly restore cerebral blood flow, salvage the ischemic penumbra, and limit infarct expansion [1]. Clinical guidelines recommend intravenous thrombolysis with alteplase within 3–4.5 h of symptom onset and mechanical thrombectomy for large vessel occlusions within 12–24 h in selected patients[[2], [3], [4]]. Despite these advances, many patients experience neurological deterioration even after successful recanalization, largely due to ischemia/reperfusion (I/R) injury [[5], [6], [7]], for which no effective therapies are currently available. I/R injury involves multiple interrelated pathological processes, among which downstream microvascular thrombosis (DMT) and neuronal endoplasmic reticulum (ER) stress are increasingly recognized as critical drivers [8,9]. DMT obstructs microcirculatory flow, compromising perfusion in the penumbral tissue and exacerbating infarct progression [10]. ER stress, while initially adaptive, triggers mitochondrial dysfunction and neuronal apoptosis when prolonged [11]. Because these pathways converge to amplify both vascular and neuronal injury, concurrently targeting them could provide synergistic neurovascular protection.

MicroRNAs (miRNAs) are small noncoding RNAs that regulate gene expression post-transcriptionally and have emerged as both biomarkers and therapeutic candidates in ischemic stroke [12,13]. Reduced expression of miR-30c has been linked to major risk factors for stroke, including hyperlipidemia, atherosclerosis, and diabetes[[14], [15], [16], [17]]. Therapeutic administration of miR-30c mimics has shown protective effects in these models. Although two prior studies [18,19] investigated miR-30c in experimental stroke, both employed pre-stroke stereotactic delivery of miR-30c modulators and examined outcomes only within 24 h. Thus, these investigations modeled preconditioning rather than clinically relevant treatment. Accordingly, whether miR-30c provides therapeutic benefit when administered after stroke onset and via systemic, translationally feasible routes has not been addressed.

Here, we investigated the therapeutic efficacy and underlying mechanisms of intravenous miR-30c mimic therapy in a mouse model of focal cerebral I/R injury. We hypothesized that miR-30c confers neurovascular protection by attenuating endothelial plasminogen activator inhibitor-1 (PAI-1)-mediated microvascular thrombosis and alleviating neuronal ER stress–mitochondrial apoptotic signaling. Using PEGylated liposomes to enhance the stability and brain uptake of the miR-30c mimic, we evaluated acute and long-term outcomes across sex and age groups in a clinically relevant setting, positioning a miR-30c-based strategy as a promising therapy for ischemic stroke.

Materials and Methods

Animals

C57BL/6J wild type (WT) mice were purchased from Jackson lab (Jackson lab, Bar Harbor, Maine). Male and female mice across ages were used, as specified in each experiment. Animals were housed on a 12-h light/dark cycle with ad libitum access to water and standard chow. All procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at Penn State University College of Medicine. The study adhered to the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines [20].

Middle cerebral artery occlusion (MCAO) and reperfusion model

The mouse transient MCAO model was performed as previously described [[21], [22], [23], [24]]. Briefly, mice were anesthetized with isoflurane (3 % for induction, 1.5 % maintenance) in a mixture of 70 % N2O/30 % O2. After a midline incision, the right common, external, and internal carotid arteries were exposed, and a silicone-coated nylon monofilament (Doccol) was introduced through the external carotid artery stump and advanced into the internal carotid artery to occlude the origin of the MCA. The size of the monofilament (coating diameter and length) was selected according to mouse body weight and prior experience, the surgical setup and operator proficiency. After 45 min of MCA occlusion, mice were re-anesthetized, and the occluding filament was gently withdrawn to allow MCA reperfusion. Sham-operated mice underwent the same procedure except for the occlusion of the MCA. Rectal temperature was maintained at 37.0 ± 0.5 °C throughout the surgical procedure with a feedback-regulated heating pad. For analgesia, animals were treated subcutaneously with buprenorphine-SR (1.0 mg/kg body weight) and 4 % lidocaine gel applied on the wound margins. Regional cerebral blood flow (CBF) was monitored before ischemia, 5 min after MCAO, and 15 min after reperfusion using laser Doppler flowmetry (MSP300XP; AD Instruments Inc).

Animals were excluded if they met any of the following criteria: (1) less than a 75 % reduction in CBF during MCAO compared to the preischemic baseline, (2) Bederson score = 0 (indicating no neurological deficit) at 24 h post-MCAO, (3) body weight loss >20 % after 24 h of MCAO, or were moribund and unable to attend normal physiological needs such as eating, drinking, and grooming, which were euthanized, (4) unsuccessful stroke based on immunohistological evaluation, (5) animal death before completion of the respective study endpoint(s).

Experimental drugs and treatments

The hsa-miR-30c-5p (referred to as miR-30c) mimic and a non-targeting scramble control mimic (sc-miR) were purchased from Life Technologies (product type: mirVana® miRNA mimic). Both are chemically modified, double-stranded small RNAs with a molecular weight of approximately 15 kDa. The mature sequences of human hsa-miR-30c-5p and mouse mmu-miR-30c-5p are identical: UGUAAACAUCCUACACUCUCAGC. The scrambled sequence (Scr) was UCACAACCUCCUAGAAAGAGUAGA.

Lyophilized miR-30c mimic and scrambled (Scr) control were dissolved in sterile nuclease-free water to prepare 10 mg/ml stock solutions, which were aliquoted and stored at −80 °C until use. For in vivo administration, stock solutions were diluted to 1–2 mg/ml working concentrations, mixed with ½ volume of PEG-liposome In Vivo Transfection Reagent (Altogen Biosystems), and incubated at room temperature for 20 min. An enhancer solution (1/5 vol) was then added, followed by an additional 5 min incubation, according to the manufacturer's protocol. The resulting complexes were administered via tail vein injection at the indicated doses and time points in each experiment.

Vivo-Morpholinos, chemically modified Morpholino antisense oligonucleotides conjugated to an octaguanidinium dendrimer for in vivo uptake [25] were used to investigate the interaction between miR-30c and PAI-1. A Vivo-Morpholino targeting plasminogen activator inhibitor-1 (PAI-1-MO) was designed to block miR-30c binding to the 3′ untranslated region (3′UTR) of PAI-1 mRNA (Fig. 4C). By acting through steric blocking rather than inducing RNA degradation [25], PAI-1-MO does not directly modulate PAI-1 expression; instead, it specifically blocks the interaction between miR-30c and the PAI-1 3′UTR. PAI-1-MO and a scrambled control Morpholino (sc-MO), each with a molecular weight of approximately 10 kDa, were custom-synthesized by Gene Tools, LLC (Philomath, OR, USA). Lyophilized Morpholinos were dissolved in sterile saline to prepare 100 mg/ml stock solutions, stored at room temperature, and diluted with PEG-liposome In Vivo Transfection Reagent to 5 or 10 mg/ml for use. In selected experiments, PAI-1-MO or sc-MO was administered via tail vein injection at a 10:1 M ratio (PAI-1-MO:miR-30c mimic), 30 min before stroke surgery.

Fig. 4.

Intravenous miR-30c mimic inhibits MCAO-induced PAI-1 expression in ischemic brain tissue and cerebral endothelium. Young adult male C57BL/6 mice were subjected to 45 min of MCAO followed by reperfusion. Intravenous miR-30c mimic treatment (2.5 mg/kg) was initiated at 3 h after ischemia onset for 24-h endpoint assays, or with repeated doses at 24 and 48 h for 72-h endpoint assays. (A) RT-qPCR analysis of PAI-1 mRNA levels in the ischemic hemispheres at 24 and 72 h post-MCAO. Data are expressed as fold changes relative to normal (non-surgical) group (defined as 1.0). n = 5 per group. ∗∗p < 0.01; ∗∗∗p < 0.001. (B) ELISA quantification of active PAI-1 protein levels in the ischemic hemispheres at 72 h post-MCAO. Data are presented as fold changes relative to sham control. n = 4–6 per group. ∗p < 0.05, ∗∗p < 0.01. (C) Schematic illustration showing MO-PAI-1 blocking the interaction between miR-30c-5p and the 3′ untranslated region (3′UTR) of PAI-1 mRNA; the miR-30c-5p seed sequence is 5′→3′ GUAAACA. (D) Representative double immunofluorescence images showing colocalization (yellow) of PAI-1 (green) with CD31⁺ endothelial cells (red) in the peri-infarct cortex at 24 h post-stroke. (E) Quantification of PAI-1-positive microvessels per mm² in the peri-infarct cortex. n = 7 per group. Scale bars: 50 μm ∗p < 0.05, ∗∗p < 0.01 vs. vehicle or sc-miR; NS, not significant. Sham-operated mice and MCAO mice treated with vehicle or scrambled mimic (sc-miR) served as controls. Vivo-Morpholino antisense oligonucleotide targeting PAI-1 (MO-PAI-1) or scrambled control (sc-MO) was administered intravenously 30 min prior to MCAO. Data are means ± SD for all analyses. Statistical analysis: one-way ANOVA followed by post hoc Bonferroni test.

Brain infarction and atrophy measurements

Brain tissue infarction was assessed using either 2,3,5-triphenyltetrazolium chloride (TTC) staining or Nissl staining. Following transcardial perfusion with 30 mL of normal saline, with or without subsequent fixation using 4 % paraformaldehyde (PFA) in 0.1 M phosphate-buffered saline (PBS, pH 7.4), brains were removed. For TTC staining (performed on non-fixed tissue), 1-mm-thick coronal sections were stained with 1 % TTC (Sigma Aldrich, USA) dissolved in normal saline at 37 °C for 30 min, with slices flipped every 15 min to ensure even staining [22]. Normal brain tissue was stained red, while infarcted areas appeared pale or white. For Nissl staining (performed on fixed tissue), 40-μm-thick coronal sections (+1 to −3 mm relative to bregma, at 500-μm intervals) were processed and stained with cresyl violet [22,23]. Infarct volumes were quantified using NIH ImageJ software. In the long-term outcome experiment, brain atrophy was assessed by measuring the cortical width index [21].

Neurobehavioral tests

Behavioral tests were conducted at the time points specified in each experiment by an experimenter and a recorder who were blinded to the experimental design.

• (1)The modified Bederson score test [26] was used to assess global neurological function by a six-point scale as follows: 0, no deficit; 1, forelimb flexion; 2, decreased resistance to lateral push; 3, unidirectional circling; 4, longitudinal spinning; and 5, no movement. This test was routinely performed on all the mice subjected to MCAO or sham surgery, one day before MCAO (as a baseline) and on day 1 and day 3 after MCAO.
• (2)The foot-fault test (Grid-walk test) was performed to assess motor impairment as well as forelimb coordination and placement dysfunction [27]. Mice were placed on a metal wire grid with 2.5 cm square openings. A foot fault was recorded whenever the animal incorrectly placed its forelimb and fell through the opening (i.e., the forelimb was misplaced and thus fell through the grid). The total number of steps (movement of each forelimb) the animal takes to cross the grid and the number of foot faults in that period (2 min) were recorded. The percentage of foot-fault errors was calculated as follows: (number of errors for the left forelimb [paretic side] ÷ total number of steps taken by the left forelimb) × 100 %. Animals were subjected to three trials a day with an inter-trial interval of 5 min.
• (3)The grip strength test was performed to assess skeletal muscle function [28]. Briefly, the non-paretic forepaw (right) was wrapped with adhesive tape before the measurement. Then, the mouse was placed over a steel grid connected to Bio-GS4 Grip Strength System (Bioseb, Pinellas Park, FL, USA) by the tail so that its only left forelimb could grab a single steel bar. The animal was gently pulled backward until the grip was released. The maximal peak force applied to the grid just before it loses grip was recorded. Each mouse had five consecutive trials a day with a 1-min rest period between trials. The means were calculated for all five measurements.
• (4)The rotarod test was performed to evaluate the motor coordination of rodents after brain injury [29]. Briefly, the mouse was placed on a rotarod device (Columbus Instruments, Columbus, OH, USA), and the latency to fall of the accelerating rotarod was recorded by a photo beam circuit. Before MCAO or sham surgery, all mice are trained for three days to get used to the rotarod instrument. On the last day of training, any animals that are unable to stay on the rod with a speed of 5 rpm for 3 min will be excluded from the study. During the test, the mice were placed on a rotating rod at a constant speed of 5 rpm for 2 min. Then, the rod's speed was increased at a rate of 0.2 rpm/s (from 1 rpm to 50 rpm over 4 min). The time it took for the animals to fall after the rod started accelerating was recorded. Mice were tested for 3 trials a day with an inter-trial interval of 1 h.
• (5)The modified neurological severity score (mNSS) was used to assess a composite of motor, sensory, reflex, and balance tests [30]. Scoring was based on physical appearance and behavior using a scale previously developed to provide a precise assessment of focal neurologic damage. The mice were scored in each of the following seven categories: body symmetry, gait, climbing, circling behavior, forelimb symmetry, compulsory circling, and sensory response. Each category was scored on a scale from 0 (no deficit) to 4 (severe deficit), yielding a maximum total score of 28, as previously described [30].
• (6)The Barnes maze test was performed to evaluate the spatial learning and memory function as we previously described [31]. In brief, a small dark recessed chamber (target box) was placed under a circular platform (91 cm in diameter, Stoelting Co), which mice can access through an escape tunnel under the target hole. Mice were pre-trained for one trial with the use of the escape tunnel at 30 days after surgery, and then went through 4 trials per day with a 15 min inter-trial interval for 5 days. During each trial, the mice were placed in the center of the maze under the start chamber. After 10 s of habituation, the chamber was lifted, and both the buzzer (85 dB) and the red and green lights (80-W bulbs) were activated on opposite sides of the maze to motivate the mice to find the target box and allowed them to explore the maze for 3 min. The time spent to reach the target hole was recorded. Then, they were required to remain in the target box for 10 s before retrieval, regardless of success. Data from 4 trials per day for each animal were averaged. On the last day, a single trial lasting for 3 min was performed with the target box removed, and mice were allowed to explore the maze and locate the target hole. The time spent in the target quadrant was analyzed using any-maze software (version 7.16). The apparatus was cleansed between each trial.
• (7)The novel object recognition test (NOR) was performed to assess cognition, particularly recognition memory [32]. On the training day, a mouse was placed in a rectangular arena (30 × 30 cm) for 5 min to minimize any stress or anxiety from handling. 24 h after training, the animal was placed in the arena and allowed to explore two identical objects for 10 min (session 1) and then returned to its home cage. One hour later, the animal was exposed to a familiar object from session 1 and a novel object (same dimension but different in shape and color) for 10 min (session 2). The time spent exploring each object in session 2 was recorded. The animal directing its nose within 2 cm distance of the object, or sniffing or pawing the object was defined as object exploration. The data were expressed as a discrimination index using the following formula: Discrimination index = time spent in exploring novel object/(time exploring novel object + time exploring familiar object) x 100 %. The apparatus was cleansed between trials.

Tissue processing and imaging for Histological and immunostaining analyses

Mice were deeply anesthetized and transcardially perfused with ice-cold phosphate-buffered saline (PBS) followed by 4 % paraformaldehyde (PFA). Brains were carefully removed, post-fixed in 4 % PFA at 4 °C overnight, and cryoprotected in a graded sucrose series (15 % followed by 30 % in PBS) until fully sunk. Samples were then embedded in OCT compound, rapidly snap-frozen on dry ice or in pre-cooled isopentane, and stored at −80 °C until sectioning. Coronal sections (+2.0 to −3.0 mm relative to bregma) were cut at the desired thickness using a cryostat (Leica CM 1950), then either mounted on charged glass slides or collected as free-floating sections for subsequent assays. Images were acquired using a Nikon Eclipse Ti-E fluorescence microscope with NIS-Elements software.

For quantification, five representative coronal sections per brain were selected at ∼250 μm intervals within the range of +1.0 to −2.0 mm relative to bregma, encompassing the cortex, striatum, and hippocampus. Images from defined regions of interest (Suppl. Diagram I) were analyzed using ImageJ software (NIH). Data are presented as the number of positive cells per mm², per 200 μm length, or as vessel density per imaged area, depending on the specific analysis.

Fluorescence in situ hybridization (FISH)

FISH was used to assess the expression and spatial distribution of miR-30c-5p in the brain [33]. Brain samples were collected and processed under RNase-free conditions to preserve RNA integrity. Serial coronal sections (15 μm thick) were used for detection of miR-30c-5p, which was performed using the miRCURY LNA miRNA ISH Kit (Qiagen, catalog# 339,450) with a specific LNA probe for mmu-miR-30c-5p (catalog# 339,112). A scrambled LNA probe (provided in the kit) and omission of the probe served as negative controls.

For cell-type colocalization, FISH was combined with immunofluorescence staining [34] using primary antibodies against CD31 (endothelium, 1:500, Covance), GFAP (astrocytes, 1:200, Abcam), Iba1 (microglia, 1:1000, Wako), or NeuN (neurons, 1:1000, Millipore). Appropriate fluorophore-conjugated secondary antibodies were applied, and nuclei were counterstained with DAPI. Images were acquired from the cortex, striatum, and hippocampus (CA1, CA3).

Double immunofluorescence staining

PAI-1 expression and microvascular thrombosis

Coronal sections (30 μm thick) were co-stained with primary antibodies against PAI-1 (1:200, LSBio), fibrin/fibrinogen (1:200, Dako), or thrombocytes (1:100, LS-C348178, LSBio), together with CD31 (1:500, Covance) overnight at 4 °C. Sections were then incubated with appropriate fluorophore-conjugated secondary antibodies for 1 h at room temperature. Negative controls omitted the primary antibodies. The number of double-positive vessels in the peri-infarct cortex was quantified and expressed as vessel density relative to the imaged area (mm²) [35].

Neuronal ER stress markers

Coronal sections (10 μm thick) were co-stained with anti-NeuN (1:500, Millipore) and one of the following ER stress markers: anti-CHOP (1:100, Abcam), anti-ATF4 (1:200, Abcam), or anti-p-eIF2α (Ser51, 1:50, Cell Signaling Technology). The number of ER stress marker-NeuN double-positive cells was quantified in the peri-infarct cortex and hippocampal CA1 and CA3 regions. Results are expressed as cells/mm² in the peri-infarct cortex or per 200 μm length in CA1/CA3 [36].

Fluoro-Jade C (FJC) staining

Neuronal degeneration was assessed using the FJC staining kit (Bionesis, TR-160-FJC). Coronal sections (15 μm thick) were fixed in 4 % PFA, mounted on slides, rinsed in PBS, and incubated in 0.06 % potassium permanganate (KMnO₄) for 5 min to reduce tissue autofluorescence, followed by a 2-min rinse in distilled water. Sections were then stained with 0.0002 % FJC in 0.1 % acetic acid for 10 min at room temperature in the dark and rinsed three times in distilled water (1 min each). After staining, sections were air-dried, cleared with xylene, and coverslipped using DXP mounting medium. Co-staining with anti-NeuN confirmed that the majority of FJC-positive cells were neurons. FJC-positive cells were quantified in the peri-infarct cortex, striatum, and hippocampus [37].

TUNEL staining

Apoptotic neurons were detected using the In Situ Cell Death Detection Kit (Roche). Coronal sections (15 μm thick) were permeabilized with 0.2 % Triton X-100 for 10 min and washed with PBS, followed by TUNEL labeling according to the manufacturer's instructions. DNase I–treated sections served as positive controls, and omission of terminal deoxynucleotidyl transferase served as negative controls. After TUNEL labeling, sections were immunostained for NeuN (1:300) to identify neurons. The number of TUNEL/NeuN double-positive cells was quantified in the peri-infarct cortex and striatum (per 20 × field) and hippocampal CA1/CA3 regions (per 200 μm length) [36].

Measurement of microvascular patency

Microvascular patency was assessed as described previosuly [35]. Briefly, FITC-dextran (FD2000S, M.W. 2 × 10⁶, Sigma-Aldrich) was administered via tail vein injection (5.0 mg/mouse in 0.2 mL PBS) 5 min prior to sacrifice. Mice were euthanized by cervical dislocation, and brain tissues were harvested and fixed in 4 % paraformaldehyde for 48 h. Coronal brain sections (100-μm thick) were prepared using a vibratome. Five sections per brain, spanning from +1 mm to −1 mm relative to bregma, were imaged using a fluorescence microscope. FITC-dextran–labeled vessels were quantified using Image-Pro Plus software (version 5.1, Media Cybernetics). For each hemisphere, the number of FITC-positive pixels was divided by the total number of pixels within the region of interest. Vascular patency was expressed as the ratio of ipsilateral to contralateral fluorescence intensity and presented as a percentage.

Western blotting

Mice were transcardially perfused with 30 mL of ice-cold phosphate-buffered saline (PBS) to remove blood. Brains were collected and hemispheres (between +2 to −3 mm relative to bregma) were separated, snap-frozen in liquid nitrogen, and stored at −80 °C until use. Tissue samples were homogenized in 1 × RIPA buffer (Thermo Scientific) containing a protease inhibitor cocktail (cOmplete™, Sigma) using a Wheaton Potter-Elvehjem glass homogenizer with a Teflon pestle (Thermo Fisher). For subcellular fractionation, homogenates were processed using the Mitochondria Isolation Kit for Tissue (Abcam). Samples were centrifuged at 1000×g for 10 min at 4 °C to remove nuclei, followed by 12,000×g for 15 min to pellet mitochondria. The supernatant was collected as the cytosolic fraction, and the mitochondrial pellet was resuspended in isolation buffer. Protein concentrations were determined using the Bio-Rad Protein Assay (Bradford method). Protein samples were denatured in SDS loading buffer, separated by SDS-PAGE, and transferred onto PVDF membranes. Membranes were blocked with 5 % non-fat milk in TBS-T and incubated overnight at 4 °C with primary antibodies against Bcl-2 (1:1000), Bax (1:1000), and COX-IV (1:1000) (all from Cell Signaling). β-Actin and COX-IV were used as loading controls for the cytosolic and mitochondrial fractions, respectively. Horseradish peroxidase-conjugated secondary antibodies were detected using an ECL detection system (GE Healthcare), and immunoreactive bands were visualized with ECL Hyperfilm. Densitometric analysis was performed for semi-quantification.

Enzyme-linked immunosorbent assay (ELISA)

PAI-1 activity was measured using a mouse active PAI-1 ELISA kit (IMSPAI1KTA, Innovative Research, Inc.) according to the manufacturer's instructions. Supernatants from cytosolic fractions were used for analysis, and ELISA values were normalized to total protein concentrations determined by the Bradford assay.

TaqMan miRNA assay and quantitative RT-PCR

Following transcardial perfusion with 30 mL of ice-cold phosphate-buffered saline (PBS) to remove blood, brains were harvested, and the cortex and hippocampus from the ipsilateral hemispheres (+2 to −3 mm relative to bregma) were dissected under a surgical microscope [38]. Total RNA was extracted using the QIAsymphony RNA Kit (#931636, Qiagen) according to the manufacturer's instructions. RNA concentrations were measured using a NanoDrop ND-2000 spectrophotometer (Thermo Scientific). Reverse transcription quantitative PCR (RT-qPCR) was performed on a Bio-Rad CFX Connect thermocycler.

For miR-30c-5p detection, cDNA was synthesized using the TaqMan™ MicroRNA Reverse Transcription Kit (#4366596, Applied Biosystems). Quantitative PCR was conducted using the TaqMan™ miRNA Assay Kit under the following PCR conditions: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. Relative miRNA levels were normalized to U6 small nuclear RNA (snRNA) using the 2ˆ−ΔΔCt method. The primers (Thermo Fisher Scientific/Life Technologies) were as follows: miR-30c-5p: Forward 5′-AGCGTCGTATCCAGTGCAAT-3′, Reverse 5′-GTCGTATCCAGTGCGTGTCG-3′, and U6 snRNA: Forward 5′-CTCGCTTCGGCAGCACATATACT-3′, Reverse 5′-ACGCTTCACGAATTTGCGTGTC-3′.

For mRNA detection, cDNA was synthesized using the iScript™ cDNA Synthesis Kit (Bio-Rad). Quantitative PCR was performed using iTaq™ Universal SYBR® Green Supermix (Bio-Rad) under the following PCR conditions: 95 °C for 3 min, followed by 40 cycles of 95 °C for 35 s (denaturation), 59 °C for 30 s (annealing), and 72 °C for 30 s (extension). Relative mRNA levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA levels using the 2ˆ−ΔΔCt method. The primers were as follows: PAI-1: Forward 5′-GCATCACTCCACAAACCTGC-3′, Reverse 5′-GGTTGCCAAGCATCACCTTG-3′; and GAPDH: Forward 5′-TTGTCTCCTGCGACTTCA-3′, Reverse 5′-CACCACCCTGTTGCTGTA-3′.

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8 software. Data are presented as mean ± standard deviation (SD) or standard error of the mean (SEM). The D'Agostino-Pearson normality test was used to determine the distribution pattern of all data. The student's t-test (for normally distributed data) or Mann–Whitney U test (for non-normally distributed data) was used to compare two groups. Differences in means among multiple groups were analyzed using one- or two-way ANOVA, followed by Bonferroni-adjusted multiple comparisons. Data that do not follow a normal distribution were analyzed by Kruskal–Wallis test followed by the Dunn's test. Data from behavior tests, which consist of repeated testing on different days, were analyzed by two-way repeated measures ANOVA. Sample size calculation (power = 0.8, α = 0.05) was conducted using an online calculator (http://www.lasec.cuhk.edu.hk/sample-size-calculation.html). Based on our preliminary results and prior experience with the mouse tMCAO model[[21], [22], [23]]-8 mice (survived) per group were required to detect a 30 % reduction in infarct volume as functionally relevant. For comparison of survival data, the log-rank test was used. A p-value <0.05 was considered statistically significant.

Results

miR-30c expression in the normal brain, downregulation after focal cerebral ischemia, and upregulation by intravenous miR-30c mimic treatment

We first assessed miR-30c expression in young adult male mice using FISH, revealing abundant expression in the cortex, striatum, and hippocampus (Fig. 1A). Combined FISH and immunofluorescence staining revealed that miR-30c was highly expressed in NeuN-positive neurons (Fig. 1B), rarely detected in Iba1-positive microglia or GFAP-positive astrocytes, but readily detectable in CD31-positive endothelial cells (Suppl. Fig. 1).

Fig. 1.

miR-30c expression in the normal brain, downregulation following focal cerebral ischemia, and upregulation by intravenous miR-30c mimic treatment. Experiments were performed in young adult male C57BL/6 mice, either under normal conditions or subjected to 45 min of MCAO followed by reperfusion for the indicated durations. (A) Representative fluorescence in situ hybridization (FISH) showing miR-30c expression (green) in the cortex (Ctx), striatum (Str), and hippocampus (Hipp) of normal mice (n = 5). miR-30c was detected using an LNA antisense probe specific for mmu-miR-30c-5p. Negative control staining was performed on sections processed without the probe. Scale bar, 100 μm. (B) Representative images of FISH combined with immunohistochemistry showing miR-30c (green) colocalization with the neuronal marker NeuN (red) in the cortex, striatum, and hippocampal subregions (CA1, CA3) of normal mice (n = 5). Scale bar, 50 μm. Nuclei are counterstained with DAPI (blue). (C) RT-qPCR analysis of miR-30c levels in the peri-infarct cortex and ipsilateral hippocampus (ipsi Hipp) at 4, 24, and 72 h after MCAO. Normal mice served as baseline controls. Data are presented as fold-change relative to controls (mean ± SD; n = 5 per group). ∗p < 0.05 vs. normal, one-way ANOVA with Bonferroni post hoc test. (D) RT-qPCR analysis of miR-30c levels in the ipsilateral hemisphere and in non-surgical controls. Mice were subjected to MCAO and treated with miR-30c mimic (2.5 mg/kg), scramble mimic (sc-miR, 2.5 mg/kg), or vehicle via tail vein injection at 3 h after ischemia onset (n = 5 per group). Non-surgical mice receiving the same treatments served as controls (n = 3 per group). Brain tissues were collected 24 h after MCAO. Data are presented as fold-change relative to vehicle. ∗∗∗p < 0.01 vs. vehicle and sc-miR, one-way ANOVA with Bonferroni post hoc test.

We next assessed age- and sex-related differences in baseline miR-30c expression. FISH staining showed that the expression pattern of miR-30c in the cortex, striatum, and hippocampus of middle-aged male mice (Suppl. Fig. 2) was similar to that of young adult mice, suggesting that its expression is largely preserved with age. However, FISH is at best semi-quantitative and cannot reliably quantify miR-30c levels. Therefore, RT-qPCR analysis was performed and demonstrated that miR-30c levels were significantly lower in middle-aged mice compared with young adults (Suppl. Fig. 3), reflecting an age-associated decline. Young adult males exhibited lower miR-30c expression than females, whereas no significant sex differences were observed in middle-aged mice (Suppl. Fig. 3), indicating that sex-related differences are more evident in early adulthood.

Subsequently, we evaluated changes in brain miR-30c expression following transient middle cerebral artery occlusion (MCAO). RT-qPCR analysis showed that in the peri-infarct cortex, miR-30c levels were significantly reduced as early as 4 h post-MCAO and decreased further at 24 and 72 h (Fig. 1C). In the ipsilateral hippocampus, miR-30c expression was mildly decreased at 4 h and significantly reduced at 24 and 72 h post-MCAO (Fig. 1C). This delayed decrease in the hippocampus likely reflects secondary injury, consistent with its anatomical location distal to the primary infarct region.

Finally, we assessed whether intravenous (IV) administration of a miR-30c mimic could restore brain miR-30c levels after ischemia. RT-qPCR analysis showed that IV delivery of miR-30c mimic (2.5 mg/kg, 3 h post-ischemia) significantly increased miR-30c expression in the ischemic hemisphere at 24 h post-MCAO (Fig. 1D). No significant increase was observed in non-ischemic brains, consistent with the limited permeability of the intact blood–brain barrier (BBB) to large molecules such as the miR-30c mimic (∼15 kDa).

Post-stroke neuroprotection by intravenous miR-30c mimic in young and older mice

To assess the therapeutic potential of intravenous miR-30c mimic in ischemic stroke, we first conducted a dose–response study in young adult male mice (3–4 months old) subjected to 45 min of MCAO followed by reperfusion, with outcomes evaluated at 72 h miR-30c mimic treatment, initiated at 3 h after ischemia onset, significantly reduced infarct volumes at a dose of 2.5 or 5.0 mg/kg, which produced comparable effects, whereas the 1.5 mg/kg dose produced only mild reductions and 0.5 mg/kg had no significant impact (Fig. 2A). Both effective doses were also associated with comparable improvements in neurological function, as reflected by lower Bederson scores and increased grip strength (Fig. 2B).

Fig. 2.

Dose–response and therapeutic time window of intravenous miR-30c mimic treatment after stroke in young adult male mice. Young adult male C57BL/6 mice were subjected to 45 min of MCAO followed by reperfusion, and stroke outcomes were assessed at 72 h post-MCAO. (A–B) Dose–response study: Mice received intravenous miR-30c mimic at 0.5, 1.5, 2.5, or 5.0 mg/kg starting at 3 h after ischemia onset, with repeated doses at 24 and 48 h post-MCAO. Vehicle and scrambled mimic (sc-miR) served as controls. (A) Left: representative TTC-stained brain sections; Right: infarct volume quantification. (B) Bederson score and grip strength testing. (C–D) Therapeutic time window study: miR-30c mimic (2.5 mg/kg) was administered intravenously at 4.5, 6, or 9 h after ischemia onset, with repeated doses at 24 and 48 h post-MCAO. (C) Left: representative TTC-stained sections; Right: infarct volume quantification. (D) Bederson score and grip strength testing. Data are means ± SD (n = 7–9 per group). ∗p < 0.05. Statistical analysis: Infarct volumes and grip strength were analyzed by one-way ANOVA with Bonferroni post hoc test; Bederson scores were analyzed using Kruskal–Wallis test with Dunn's post hoc test.

To determine the therapeutic window, intravenous miR-30c mimic was initiated at 4.5, 6.0, or 9.0 h after ischemia onset, with outcomes assessed at 72 h. Treatment initiated at 4.5 h remained effective, reducing infarct volumes and improving neurological outcomes (Fig. 2C and D), whereas administration at 6.0 or 9.0 h provided no significant benefit. These results indicate that intravenous miR-30c mimic confers neuroprotection within a clinically relevant 3–4.5 h window after stroke onset.

Because age is a major determinant of stroke incidence and severity, we next tested the efficacy of intravenous miR-30c mimic in middle-aged male and female mice (12–13 months old), with outcomes assessed at 72 h post-stroke. Mice received miR-30c mimic at 0.5, 2.5, or 5.0 mg/kg, initiated at 3 h after ischemia onset. In both sexes, the 2.5 and 5.0 mg/kg doses, similar to those effective in young adult mice (Fig. 2), significantly reduced infarct volumes (Fig. 3A and C) and improved neurological function, as measured by the modified Bederson score, modified neurological severity score (mNSS), foot fault test, and grip strength test (Fig. 3B and D). The 0.5 mg/kg dose had no significant effect. Notably, treatment efficacy was comparable between male and female middle-aged mice.

Fig. 3.

Dose-response of intravenous miR-30c mimic treatment after stroke in middle-aged female and male mice. C57BL/6 mice (12–13 months old) were subjected to 45 min of MCAO followed by reperfusion, and stroke outcomes were assessed at 72 h post-MCAO. Mice received intravenois miR-30c mimic at 0.5, 2.5, or 5.0 mg/kg starting at 3 h after ischemia onset, with repeated doses at 24 and 48 h post-MCAO. Vehicle and scrambled mimic (sc-miR) served as controls. (A–B) Female mice: (A) Left: representative Cresyl Violet-stained brain sections; Right: infarct volume quantification. (B) Behavioral assessments, including Bederson score, modified Neurological Severity Score (mNSS), foot fault test, and grip strength testing. (C–D) Male mice: Panels correspond to (A–B), respectively. Data are means ± SD (n = 6–8 per group per sex). ∗p < 0.05. Statistical analysis: One-way ANOVA with Bonferroni post hoc test was used for all comparisons, except Bederson scores, which were analyzed using the Kruskal–Wallis test with Dunn's post hoc test.

Together, these results indicate that post-stroke intravenous administration of miR-30c mimic effectively reduces acute ischemic injury across age groups and sexes within a clinically relevant therapeutic window.

Intravenous miR-30c mimic attenuates secondary microvascular thrombosis via targeted inhibition of PAI-1

RT-qPCR analysis showed that plasminogen activator inhibitor-1 (PAI-1) mRNA levels were markedly elevated in the ischemic brain at 24 and 72 h after stroke, and these elevations were significantly attenuated by intravenous miR-30c mimic treatment (Fig. 4A). Consistent with these findings, ELISA confirmed increased levels of active PAI-1 protein in brain tissue at 72 h post-stroke, which were also significantly reduced by miR-30c treatment (Fig. 4B).

PAI-1 is a validated direct target of miR-30c, which binds to the 3′ untranslated region (UTR) of PAI-1 mRNA to inhibit its translation [[39], [40], [41]]. To determine whether this interaction is functionally relevant in vivo, we used PAI-1-MO, a Vivo-Morpholino antisense oligonucleotide designed to block the miR-30c binding site in the PAI-1 3′ UTR (Fig. 4C). Consistent with its mechanism of action, pretreatment with PAI-1-MO alone had no detectable effect on PAI-1 expression (Fig. 4D) or cerebral infarction after tMCAO (Suppl. Fig. 4), thereby ensuring that the effects observed in combination with the miR-30c mimic can be attributed specifically to blocking the miR-30c–PAI-1 interaction, rather than to independent effects on brain infarction.

Immunohistochemistry showed robust upregulation of PAI-1 expression in the ischemic hemisphere at 24h after tMCAO, predominantly localized to cerebral endothelial cells (Fig. 4D). Intravenous miR-30c mimic markedly suppressed tMCAO-induced endothelial PAI-1 expression, and this effect was abolished by PAI-1-MO pretreatment (Fig. 4D), supporting that miR-30c exerts its antithrombotic effect through direct inhibition of endothelial PAI-1.

Next, we investigated whether miR-30c reduces microvascular thrombosis and improves cerebral perfusion by inhibiting PAI-1. At 24 h post-stroke, substantial microvascular thrombosis was observed, as indicated by intravascular deposition of fibrin/fibrinogen (Fig. 5A) and platelets (Fig. 5B). This thrombotic burden was associated with a marked reduction in cerebral microvascular perfusion, as determined by the FITC-dextran assay(Fig. 5C and D). Intravenous miR-30c mimic significantly attenuated microvascular thrombosis and largely restored perfusion, whereas these beneficial effects were almost completely reversed by PAI-1-MO pretreatment. These findings further support that the therapeutic effects of miR-30c are mediated by targeted inhibition of endothelial PAI-1.

Fig. 5.

Intravenous miR-30c mimic treatment attenuates microvascular thrombosis and improves cerebral vascular patency by targeting PAI-1 in ischemic cerebral endothelium. Young adult male C57BL/6 mice were subjected to 45 min of MCAO followed by reperfusion. miR-30c mimic treatment (2.5 mg/kg) was administered intravenously at 3 h after ischemia onset. Vehicle and scrambled mimic (sc-miR) served as controls. A Vivo-Morpholino antisense oligonucleotide targeting PAI-1 (MO-PAI-1) or scrambled control (sc-MO) was intravenously administered 30 min prior to MCAO. All assessments were performed at 24 h post-stroke. (A, B) Left: representative immunofluorescence images showing intravascular fibrin/fibrinogen (green) and platelets (green) colocalizing with cerebral microvessels (CD31⁺, red) in the peri-infarct cortex. Right: quantification of fibrin/fibrinogen-positive and platelet-positive microvessels, expressed as the number of positive vessels per mm². Scale bars: 50 μm. n = 7 per group. (C) Representative images of FITC-dextran–perfused brain sections for assessing cerebral microvascular patency. Regions exhibiting green fluorescence indicate patent, perfused microvessels, whereas dark areas correspond to non-perfused or occluded vessels. (D) Quantification of cerebral microvascular patency, expressed as the percentage of FITC-dextran–perfused area in the ipsilateral hemisphere relative to the contralateral side. Scale bars: 2 mm. n = 3–4 per group. Data are means ± SD. Statistical analysis: one-way ANOVA followed by Tukey's multiple comparison test for (A) and (B); Kruskal-Wallis test followed by the Mann–Whitney U test was used for (D).

Together, these results provide the first in vivo evidence that endothelial PAI-1 upregulation is a critical mediator of secondary microvascular thrombosis following ischemic stroke, and that intravenous miR-30c mimic alleviates this process through targeted inhibition of endothelial PAI-1.

Intravenous miR-30c mimic attenuates neuronal injury via suppression of ER stress and the mitochondrial apoptotic pathway

Neuronal degeneration, assessed by Fluoro-Jade C (FJC) staining, was undetectable in sham brains but robustly induced after tMCAO, with signals more pronounced at 24 h than at 72 h across multiple regions, including the peri-infarct cortex, striatum, and hippocampal subfields (CA1, CA3) (Fig. 6A and B). Colocalization of FJC with NeuN confirmed neuronal degeneration (Fig. 6A). This pattern is consistent with previous reports showing that FJC selectively labels neurons in an acute degenerative state, which peaks within the first day following ischemia–reperfusion and declines thereafter, likely due to phagocytic clearance and progression of neuronal disintegration beyond FJC's detection window [42,43] Moreover, neuronal cell death assessed by TUNEL staining was absent in sham brains but markedly increased at 72 h after tMCAO in the same brain regions (Fig. 6C and D). Together, these findings suggest that early FJC-detectable degeneration evolves into apoptotic cell death at later time points after ischemic stroke.

Fig. 6.

Intravenous miR-30c mimic treatment attenuates neuronal degeneration and apoptosis after stroke. Young adult male C57BL/6 mice were subjected to 45 min of MCAO followed by reperfusion. Intravenous miR-30c mimic treatment (2.5 mg/kg) was initiated at 3 h after ischemia onset for 24-h endpoint assays, or with repeated doses at 24 and 48 h for 72-h endpoint assays. Sham surgery, vehicle, and scrambled mimic (sc-miR) groups served as controls. (A) Representative Fluoro-Jade C (FJC) staining showing degenerating cells in the cortex, striatum, and hippocampal subfields (CA1 and CA3) at 24 h post-stroke. Colocalization (indicated by yellow) of FJC (green) with NeuN (red) confirms neuronal degeneration. Scale bar: 100 μm. (B) Quantification of FJC-positive cells in the indicated regions at 24 h and 72 h post-stroke (∗p < 0.05; n = 7 per group). (C) Representative TUNEL staining images showing apoptotic cells in the cortex, striatum, and hippocampal subfields (CA1 and CA3) at 72 h post-stroke. Colocalization of TUNEL (green) with NeuN (red), indicated in yellow, confirms neuronal apoptosis. Scale bar: 100 μm. (D) Quantification of TUNEL/NeuN double-positive neurons in the indicated regions at 72 h post-stroke (∗p < 0.05, ∗∗p < 0.01; n = 3 per group). Data are presented as mean ± SD. Statistical analysis: one-way ANOVA with Dunnett's post hoc test.

Immunofluorescence staining further revealed significant upregulation of ER stress markers, including phosphorylated eIF2α (p-eIF2α), ATF4, and CHOP in the peri-infarct cortex and ipsilateral hippocampus at 72 h post-tMCAO (Fig. 7A). Double immunofluorescence confirmed that these markers were predominantly localized in NeuN⁺ neurons (Suppl. Fig. 5), indicating that ER stress is primarily induced in neurons following ischemic stroke.

Fig. 7.

Intravenous miR-30c mimic treatment attenuates ER stress and mitochondrial apoptosis after stroke. Young adult male C57BL/6 mice were subjected to 45 min of MCAO followed by reperfusion. miR-30c mimic (2.5 mg/kg) was administered intravenously starting at 3 h after ischemia onset, with repeated doses at 24 and 48 h post-stroke. Vehicle and sc-miR served as controls. All analyses were performed at 72 h post-stroke. (A) Immunofluorescence analysis of ER stress markers, including phospho-eIF2α, ATF4, and CHOP, in the peri-infarct cortex and hippocampal subfields (CA1 and CA3). n = 5 per group. ∗P < 0.05, ∗∗p < 0.01, ∗∗∗P < 0.001. Representative immunofluorescence images are shown in Suppl Fig. 5 (B) Representative immunoblot images showing Bcl-2 and Bax protein levels in mitochondrial and cytosolic fractions isolated from ischemic hemispheres (+2.0 to −3.0 mm relative to bregma). COX-IV and GAPDH were used as loading controls for mitochondrial fractions; β-actin for cytosolic fractions. (C) Densitometric analysis of Bcl-2 and Bax protein levels, normalized to COX-IV (mitochondrial) or β-actin (cytosolic). n = 4 for sham and 5 for each MCAO group. ∗p < 0.05. Data are presented as means ± SD. Statistical analysis: one-way ANOVA followed by post hoc Bonferroni test for immunofluoresence staining data. Kruskal–Wallis test followed by Mann–Whitney U test for immunoblotting data.

ER stress is known to contribute to neuronal apoptosis by disrupting mitochondrial homeostasis and modulating the balance of pro- and anti-apoptotic proteins, including downregulation of Bcl-2 and upregulation of Bax [9]. Western blot analysis of mitochondrial fractions from the ischemic hemisphere (+2 to −3 mm relative to bregma) at 72 h post-MCAO revealed a marked reduction of Bcl-2, which was profoundly restored by miR-30c mimic treatment (Fig. 7B and C, Suppl. Fig. 7). Conversely, MCAO induced robust elevation of mitochondrial Bax, which was significantly attenuated by miR-30c. As a result, miR-30c preserved the mitochondrial Bcl-2/Bax ratio, a critical determinant of apoptotic susceptibility. Interestingly, cytosolic Bax levels were also elevated after MCAO; however, miR-30c did not affect this increase, suggesting that its regulation of mitochondrial Bax accumulation may involve mechanisms beyond simple inhibition of cytosolic Bax translocation.

Taken together, these results indicate that intravenous miR-30c mimic mitigates ischemic neuronal injury by reducing early degeneration and suppressing ER stress-driven activation of the mitochondrial apoptotic pathway.

Intravenous miR-30c mimic improves long-term motor and cognitive recovery in older mice after stroke

Long-term neurological outcome assessments are critical in translational stroke research, as they capture persistent motor and cognitive deficits that may not be evident in short-term evaluations, thereby providing a more comprehensive measure of therapeutic efficacy. To examine the long-term benefits of miR-30c treatment, middle-aged female C57BL/6 mice (12–13 months old) were subjected to 45 min of transient MCAO followed by reperfusion. Mice received intravenous miR-30c mimic at an initial dose of 5.0 mg/kg at 3 h post-ischemia, followed by additional doses of 2.5 mg/kg at 24, 48, and 72 h post-stroke. Behavioral testing was conducted at the indicated time points (Fig. 8A).

Fig. 8.

Intravenous miR-30c mimic treatment improves long-term motor and cognitive recovery in middle-aged mice after stroke. Female C57BL/6 mice (12–13 months old) were subjected to 45 min of MCAO followed by reperfusion. miR-30c mimic treatment (5.0 mg/kg) was administered intravenously starting at 3 h after ischemia onset, followed by repeated doses (2.5 mg/kg) at 24, 48, and 72 h post-stroke. Scrambled miR-30c mimic (sc-miR) and vehicle were used as controls. (A) Experimental timeline outlining treatment and behavioral testing. (B) Motor function assessments, including rotarod performance and grip strength testing on days 7, 14, 21, and 28 post-stroke. (C) Cognitive assessments, including the Barnes maze (days 31–35) and novel object recognition (NOR) test (day 35). Mice were euthanized on day 35 for final analyses. (D) Kaplan–Meier survival curves for each group. Survival was compared using the log-rank test. Survived/total animals: Sham (n = 6/6), Vehicle (n = 8/13), sc-miR (n = 6/10), and miR-30c mimic (n = 7/9). Data are means ± SD. Statistical analysis: two-way repeated-measures ANOVA with Bonferroni's multiple comparisons test for Rorarod, grip stremgth, and Barnes maze tests; one-way ANOVA with Bonferroni's post hoc test for NOR test; log-rank test for survival (panel D). ∗p < 0.05 vs. sham, vehicle, or sc-miR, #p < 0.05, ##p < 0.01 vs.sham.

miR-30c mimic treatment significantly improved motor function from days 7–28 post-stroke, as evidenced by enhanced performance on the rotarod test and increased forelimb grip strength (Fig. 8B), indicating improved motor coordination and strength. Cognitive recovery was also enhanced, with miR-30c–treated mice exhibiting better spatial learning and memory in the Barnes maze (days 31–35) and improved recognition memory in the NOR test on day 35 (Fig. 8C).

Survival at 35 days post-stroke was 100 % in the sham group (6/6), 62 % in the vehicle group (8/13), 60 % in the sc-miR group (6/10), and 78 % in the miR-30c mimic group (7/9) (Fig. 8D), although these differences did not reach statistical significance, likely due to limited sample sizes. All deaths occurred within the first 7 days post-stroke, with no further mortality between days 7 and 35. Preliminary observations in age-matched male mice suggested a similar functional benefit with the same treatment protocol (Suppl. Fig. 6).

Discussion

This study demonstrates for the first time that post-stroke intravenous miR-30c mimic therapy attenuates acute cerebral ischemia–reperfusion injury and promotes long-term functional recovery in a clinically relevant setting. We performed comprehensive dose–response and therapeutic time window studies in both male and female mice across different age groups. Mechanistically, miR-30c exerts dual protective actions by reducing endothelial PAI-1–mediated microvascular thrombosis and alleviating ER stress-driven neuronal apoptosis and injury. Together, these findings establish miR-30c as a key regulator of ischemic stroke pathology and support its potential as a novel microRNA-based therapeutic strategy.

Previous studies have suggested a protective role for miR-30c in acute experimental stroke [18,19]. These reports showed that miR-30c expression in ischemic brain tissue is significantly reduced within hours to 24 after transient MCAO and that pre-stroke manipulation of brain miR-30c levels influences infarct size and neurological outcomes. Specifically, Zhang et al. [19] reported that stereotactic delivery of a miR-30c mimic three days before stroke reduced infarct size and improved neurological function, whereas Dong et al. [18] demonstrated that pre-stroke downregulation of miR-30c via stereotactic delivery of an antagomir 14 days before stroke exacerbated injury. While these findings provided important proof-of-concept, their translational relevance was limited: interventions were applied before stroke onset, outcomes were assessed only in the acute phase (<24 h), experiments included only young male rodents, delivery relied on invasive intracranial injection, and behavioral evaluations were minimal.

In the present study, we investigated post-stroke miR-30c treatment in a clinically relevant setting using multiple intravenous doses of a miR-30c mimic, with the first dose administered within 3–4.5 h after stroke onset, representing a feasible therapeutic route and time window. Neurological outcomes were assessed during both the acute (∼3 days) and long-term (∼35 days) phases across age and sex groups. Adult male C57BL/6 mice (3–4 months old) were used for initial dose–response and therapeutic window studies because they produce relatively consistent infarct sizes and neurological deficits under well-established conditions, whereas subsequent efficacy studies included middle-aged male and female mice (12–13 months old) to enhance translational relevance. Notably, human and mouse miR-30c share an identical mature sequence, and a human miR-30c-5p mimic was used in these mouse studies. The results demonstrate that post-stroke intravenous administration of the human miR-30c mimic provides robust neuroprotection and promotes functional recovery. Encouragingly, the treatment was consistently effective across age and sex groups, indicating that intravenous administration of miR-30c mimic (2.5 mg/kg), which increased miR-30c levels more than 2.5-fold above baseline in the ischemic brain, was sufficient to overcome age- and sex-related variability that often limits stroke therapy outcomes.

miRNA-based therapies have progressed to clinical trials for diseases such as cancer and viral infections [[44], [45], [46]]. Although they show promise in experimental stroke [12,47], their clinical translation remains limited by several key barriers, including rapid nuclease-mediated degradation in circulation, poor blood–brain barrier (BBB) penetration, and inefficient uptake by target brain cells. Chemically modified mimics (e.g. mirVana) improve molecular stability but still exhibit restricted BBB penetration, in part due to their relatively large molecular weight (∼15 kDa) [48]. In this study, encapsulation of the mirVana human miR-30c mimic in PEGylated liposomes—an FDA-approved delivery platform [49]—enhanced molecular stability, promoted cellular uptake, and increased miR-30c levels in ischemic brain tissue by more than twofold at 24 h post-injection, with minimal accumulation in non-ischemic regions. Together, these findings support PEGylated liposome-mediated delivery of the miR-30c mimic as a highly effective and clinically relevant strategy to overcome these major barriers, thereby enhancing the translational potential of miRNA-based therapies for ischemic stroke.

Mechanistically, our findings advance understanding of how miR-30c confers neurovascular protection. First, we demonstrate that intravenous administration of the miR-30c mimic reduced downstream microvascular thrombosis (DMT) by directly targeting endothelial PAI-1. DMT is increasingly recognized as a major contributor to reperfusion failure following successful large vessel recanalization [50,51] involves endothelial dysfunction, intravascular platelet adhesion, and fibrin deposition, which together compromise microcirculatory perfusion [[52], [53], [54]]. We demonstrated that miR-30c mimic suppressed PAI-1 expression at both the mRNA and protein levels, reduced intravascularr fibrin and platelet deposition, and restored microvascular patency. Importantly, these effects were abolished by pretreatment with a PAI-1-targeting Vivo-Morpholino that specifically disrupted the miR-30c–PAI-1 interaction, thereby establishing endothelial PAI-1 as a functional in vivo target of miR-30c. By selectively blocking miR-30c–mediated suppression of ischemia-induced PAI-1 upregulation, this Vivo-Morpholino-based strategy provides a unique approach to define miR-30c–dependent regulation of PAI-1 in vivo. Given that PAI-1 is a well-established marker of prothrombotic state and has been implicated in stroke pathology [55], these findings provide direct mechanistic evidence supporting for PAI-1 as a therapeutic target of miR-30c.

Second, we demonstrate that intravenous administration of the miR-30c mimic suppresses neuronal ER stress and mitochondrial apoptosis—two tightly interconnected processes central to ischemic neuronal death [11,56]. The elF2α-ATF4-CHOP signaling axis is a core pathway through which ER stress induces neuronal injury and contributes to neurodegenerative disorders. ER stress promotes apoptosis by altering the balance of pro- and anti-apoptotic proteins, including Bax and Bcl-2, with CHOP further amplifying this effect by downregulating Bcl-2 and upregulating Bax [56]. In the present study, MCAO robustly increased ER stress markers (phosphorylated eIF2α, ATF4, and CHOP) in cortical and hippocampal neurons, coinciding with neuronal injury. Treatment with miR-30c mimic attenuated these ER stress responses, restored mitochondrial Bcl-2 levels, and reduced mitochondrial Bax without affecting cytosolic Bax, indicating selective inhibition of the intrinsic mitochondrial apoptotic pathway. Bioinformatic analyses [57] and previous studies have identified Bcl-2 in glioma cells [58] and elF2α in microglia [59] and human cancer cells [60] as direct targets of miR-30c. In the present study, we provide the first in vivo evidence that miR-30c regulates these targets in the ischemic brain, although further investigation is warranted to explore additional upstream and downstream mechanisms underlying miR-30c–mediated regulation of ER stress and mitochondrial apoptotic signaling. Together, these results demonstrate that miR-30c modulates ER–mitochondria crosstalk, a process increasingly implicated in neuronal injury in both stroke and neurodegenerative disorders [61].

Although oxidative stress was not directly assessed in this study, extensive evidence supports a close interplay between oxidative stress and ER stress in ischemic brain injury. Ischemia-reperfusion-induced reactive oxygen species (ROS) disrupt ER protein folding and mitochondrial homeostasis, thereby triggering ER stress, which can further amplify oxidative stress through ER-mitochondria calcium dysfunction and mitochondrial ROS overproduction, thereby exacerbating secondary neuronal injury following cerebral I/R [62,63]. By attenuating ER stress, miR-30c may indirectly reduce oxidative stress in ischemic neurons. Whether miR-30c directly regulates oxidative pathways in vivo remains unknown. Future studies in our laboratory will investigate whether miR-30c modulates ROS generation from infiltrating leukocytes-particularly neutrophils- and activated microglia-the major sources of oxidative stress in the post-ischemic brain, thereby providing a more comprehensive understanding of its neuroprotective mechanisms.

Conclusion

This study establishes intravenous miR-30c mimic as a potent microRNA-based therapeutic strategy for ischemic stroke. Using PEGylated liposomes—an FDA-approved delivery platform—we achieved reliable post-stroke delivery, defined a clinically relevant therapeutic time window, and demonstrated consistent efficacy across age and sex groups, resulting in both acute neuroprotection and long-term functional recovery. Mechanistically, miR-30c confers dual neurovascular protection by attenuating endothelial PAI-1–mediated microvascular thrombosis and suppressing neuronal ER stress–mitochondrial apoptotic signaling. By simultaneously targeting vascular and neuronal injury pathways that limit reperfusion efficacy and promote secondary neuronal damage, these findings provide a strong rationale for advancing miR-30c–based therapies toward clinical translation for ischemic stroke.

Future Directions

Despite these promising findings, several important questions remain to be addressed to advance miR-30c-based therapy toward clinical translation. First, long-term safety, tolerability, and potential off-target effects will require systematic evaluation, particularly under repeated dosing paradigms relevant to clinical use. Second, given the high prevalence and clinical impact of post-stroke cognitive impairment, future studies should determine how therapeutic elevation of miR-30c in the brain influences hippocampal and white matter pathology, including neurodegeneration, demyelination, synaptic remodeling, and neuroinflammation, and whether these effects contribute to sustained cognitive recovery. Accordingly, future work will focus on detailed characterization of hippocampal and white matter injury to elucidate miR-30c–mediated mechanisms of long-term neuroprotection and repair.

In addition, it will be important to evaluate therapeutic efficacy in preclinical models incorporating common stroke-related comorbidities, such as diabetes, hypertension, and atherosclerosis, which frequently limit translational success. Finally, whether miR-30c-based therapy can be safely combined with standard reperfusion strategies, including thrombolysis with tissue plasminogen activator (tPA, alteplase) to mitigate ischemia-reperfusion injury and potentially extend the therapeutic window warrants further investigation. Addressing these questions will refine mechanistic understanding and accelerate the clinical translation of miR-30c-based therapies for ischemic stroke.

Data availability

All data will be made available by the corresponding author upon reasonable request.

Authors’ contributions

R.J. performed animal stroke model surgeries, Western blotting, ELISA, immunohistochemistry, and data analysis. M.S. conducted immunohistochemistry, RT-PCR, behavioral tests, and data analysis. W.Z. performed immunohistochemistry and RT-PCR. Y.H. performed immunohistochemistry. J.L. assisted with manuscript editing and data analysis, and M.W. assisted with behavioral tests. G.L. designed and supervised the study and wrote the manuscript. All authors reviewed and approved the final manuscript.

Declaration of competing interest

The authors declare no conflicts of interest in the manuscript.

Abbreviations

ATF4

Activating transcription factor 4

CHOP

C/EBP homologous protein

DMT

downstream microvascular thrombosis

eIF2α

the eukaryotic translation initiation factor 2 alpha

ER stress

endoplasmic reticulum stress

tMCAO

transient middle cerebral artery occlusion

PAI-1

Plasminogen activator inhibitor-1

WT

wild type mice.

Appendix A. Supplementary data

The following is the Supplementary data to this article.