Salvaging the No-Reflow Zone: Intra-arterial Cranial Bone-Derived MSCs Restore Neurovascular Integrity After Stroke

Abstract

Successful macrovascular recanalization does not always translate into functional recovery, in part because early post-reperfusion injury can drive infarct expansion. Here, we investigated whether hyperacute intra-arterial transplantation of human cranial bone–derived mesenchymal stem cells (hcMSCs) immediately after reperfusion improves functional recovery after ischemic stroke. Transient middle cerebral artery occlusion was induced in adult rats, followed by intra-arterial administration of hcMSCs or vehicle immediately after reperfusion. Neurological and motor pathway recovery were assessed longitudinally using the modified neurological severity score and transcranial motor-evoked potentials. During the acute phase, molecular and histological analyses of peri-infarct tissue were performed and the transcriptomic profiles of hcMSCs compared with those of human bone marrow–derived mesenchymal stem cells (hbMSCs). Hyperacute intra-arterial hcMSC transplantation resulted in significantly faster neurological recovery compared with controls, accompanied by enhanced electrophysiological recovery of motor pathway function. These functional improvements were associated with preservation of neuronal structural integrity in the peri-infarct cortex, attenuation of inflammatory cytokines and apoptosis activity, and vascular endothelial growth factor upregulation. Further, transcriptomic analysis revealed that hcMSCs are enriched in neurotrophic, angiogenic, and immunoregulatory factors, providing a biological support for the observed therapeutic effects. Hyperacute intra-arterial hcMSC delivery immediately after reperfusion promotes functional and electrophysiological recovery in an experimental reperfusion model, likely through early modulation of inflammatory and apoptotic responses in the peri-infarct tissue. These findings support further investigation of hcMSC therapy as a biologically protective adjunctive strategy in the immediate post-recanalization setting.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s12975-026-01494-x.

Introduction

Despite major advances in acute ischemic stroke management, particularly the widespread adoption of endovascular thrombectomy, successful macrovascular recanalization does not necessarily translate into effective functional recovery. Persistent neurological deficits after reperfusion are increasingly attributed to post-recanalization microvascular dysfunction—commonly referred to as the “no-reflow” phenomenon—as well as secondary brain injury following reperfusion [1–3]. These phenomena are now recognized as mechanistically interrelated, whereby reperfusion injury and microvascular dysfunction mutually amplify neurovascular damage even after angiographic recanalization [4–6]. Recent clinical and imaging studies have indicated that incomplete tissue reperfusion despite successful thrombectomy is independently associated with infarct expansion and poor functional outcome, underscoring a critical therapeutic window in the immediate post-recanalization phase [7–9]. In response to this unmet need, intra-arterial thrombolytic strategies, including alteplase and tenecteplase, have been investigated as rescue approaches to improve distal microvascular perfusion; however, concerns remain regarding hemorrhagic transformation, particularly in tissue with established infarct core and blood-brain barrier (BBB) disruption. Thus, there is growing interest in biologically protective, non-thrombolytic strategies that can modulate early post-reperfusion injury and stabilize the neurovascular unit without increasing hemorrhagic risk.

Mesenchymal stem cells (MSCs), characterized by their multipotent nature, paracrine activity, and favorable safety profile, have been extensively explored as a cell-based therapeutic modality in regenerative medicine [10–13]. Beyond their regenerative potential, MSCs exert pleiotropic effects on inflammation, apoptosis, angiogenesis, and BBB integrity—mechanisms highly relevant to secondary injury processes after reperfusion. Furthermore, accumulating research suggests that MSC phenotypic characteristics and therapeutic effects vary depending on the tissue of origin, reflecting differences in developmental background and intrinsic cellular properties [14–16]. In our previous studies, we identified neural crest–derived cranial bone as a novel and biologically distinct source of MSCs in both rats and humans. These cranial bone–derived MSCs (cMSCs) satisfied the defining criteria of MSCs while exhibiting distinct immunophenotypic and paracrine characteristics compared with bone marrow–derived MSCs (bMSCs), appearing particularly suitable for neural repair [17–19]. Furthermore, in an experimental model of cerebral infarction, intravenous transplantation of human cranial bone–derived MSCs (hcMSCs) conferred greater therapeutic benefit than human bone marrow–derived MSCs (hbMSCs), as evidenced by enhanced neurotrophic factor expression and associated motor function improvements [18]. However, whether MSC-based therapy can effectively modulate early post-reperfusion tissue injury in the hyperacute phase remains unclear.

In this study, we investigated the therapeutic effects of hyperacute intra-arterial hcMSC transplantation immediately after reperfusion in a rat model of transient middle cerebral artery occlusion (MCAO). In addition, comparative transcriptomic analysis between hcMSCs and hbMSCs was conducted to provide molecular insights into the biological basis underlying their therapeutic potential. Through this integrated approach, we aimed to evaluate the therapeutic efficacy of hyperacute intra-arterial hcMSC delivery and explored its translational potential as a non-thrombolytic adjunctive strategy targeting the immediate post-recanalization setting.

Materials and Methods

Ethical Approval

All animal experiments were approved by the Animal Research Committee of Hiroshima University (approval number: A24-73) and conducted in strict accordance with institutional and national guidelines for the care and use of laboratory animals. This study is reported in compliance with the Animal Research: Reporting in Vivo Experiments (ARRIVE) guidelines (https://arriveguidelines.org) and the Hiroshima University Animal Experimentation Regulations. Written informed consent was obtained from patients for the use of surgically discarded cranial bone fragments under protocols approved by the Institutional Review Board of Hiroshima University (approval number: E2010-0379).

hcMSC isolation and culture

hcMSCs were isolated and cultured according to previously established protocols [17–21]. Cranial bone fragments from the sphenoid ridge of the frontotemporal bone were collected during routine craniotomy procedures and used for primary cell isolation. These bone fragments, otherwise discarded, were repurposed for primary cell isolation after obtaining written informed consent from patients. No additional invasive procedures were required.

The harvested bone fragments were seeded into 90-mm tissue culture dishes (Sumitomo Bakelite Co., Tokyo, Japan) containing low-glucose Dulbecco’s Modified Eagle Medium (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA), 100 U/mL penicillin, and 100 µg/mL streptomycin. Cultures were maintained in a humidified incubator at 37 °C with 5% CO₂. The medium was replaced every 2–3 days. The bone fragments were removed after 7–9 days of culture, and the adherent cells were further expanded for approximately 2 weeks. Once adherent cells reached 80% confluency, they were passaged and subsequently used as hcMSCs (Supplementary Fig. 1). Prior to transplantation, the hcMSC batch used in this study was validated by flow cytometric characterization and trilineage differentiation assays according to our previously established protocols (Supplementary Fig. 2, 3) [17, 21].

RNA Sequencing analysis

To compare the transcriptomic profiles of hcMSCs with those of hbMSCs, a conventional reference MSC source used in our prior studies, exploratory RNA sequencing analysis was performed [17, 18, 22]. Total RNA was extracted from cultured hcMSCs (n = 3) and hbMSCs (n = 3) using the NucleoSpin™ RNA mini kit (MACHEREY-NAGEL GmbH & Co. KG, Duren, Germany), following the manufacturer’s protocol. RNA concentration and integrity were assessed using the QuantiFluor RNA system (Promega) and Agilent Small RNA kit of the BioAnalyzer 2100 system (Agilent Technologies); only samples meeting quality criteria were used for subsequent analyses. RNA sequencing libraries were prepared using QIAseq mRNA library kit (Qiagen) and sequenced on an Illumina NextSeq 500 sequencer to generate 75-nucleotide single reads. Raw sequencing reads were subjected to quality control, including trimming of adapter sequences and removal of low-quality reads.

Differential gene expression analysis and GO analysis between hc and hb-MSCs were performed using DESeq2 (ver. 1.18.1) and clusterProfiler (ver. 3.6.0). Principal component analysis (PCA) was performed using variance-stabilized normalized count data generated by DESeq2, and an MA plot was generated to visualize global transcriptional differences between groups. The threshold for statistically significant differentially expressed genes was defined as an absolute log₂ fold change (|log₂FC|) > 1.5 with an adjusted p-value < 0.05. Genes with log₂FC > 1.5 were classified as upregulated, whereas those with log₂FC < − 1.5 were classified as downregulated.

Animals and inclusion/exclusion criteria

A total of 56 adult male Sprague–Dawley rats were used in this study. Separate cohorts were allocated for functional/electrophysiological assessment, histological analysis, qRT-PCR, western blotting, infarct validation, and cell-tracking experiments. Inclusion criteria for longitudinal outcome assessment were survival beyond the immediate postoperative period and a modified neurological severity score (mNSS) of 6–10 on day 1 after MCAO. Animals that died during the procedure or exhibited neurological deficits outside the predefined mNSS range were excluded. Details of animal allocation, exclusions, and final sample sizes are provided in Supplementary Table.

Surgical Procedure

Adult male Sprague–Dawley rats (300–350 g) were subjected to MCAO using a previously established intraluminal suture method [23]. Rats were placed in the supine position on a heating pad (Bio Research Center Co., Ltd., Aichi, Japan) to maintain a rectal temperature of 37 °C. General anesthesia was maintained with 3% isoflurane during surgery. The right common carotid artery (CCA), external carotid artery (ECA), and internal carotid artery (ICA) were exposed under a surgical microscope. Both the CCA and ICA were temporarily occluded with a microvascular clip, and the ECA was then ligated and severed. A 4 − 0 nylon monofilament with a silicone-coated tip (403745PK10; Doccol Co., Sharon, MA, USA) was introduced through the ECA into the ICA and advanced to the origin of the MCA to induce focal cerebral ischemia.

After 45 min of occlusion, the suture was withdrawn and an ultra-small-diameter microcatheter (Defrictor nano; Medicos Hirata, Osaka, Japan) was immediately inserted into the ECA. Following CCA declamping and restoration of cerebral blood flow under temporary occlusion of the pterygopalatine artery, either passage 3 hcMSCs (1 × 10⁴ cells suspended in 300 µL of PBS) or PBS alone were slowly infused through the microcatheter using a 1 mL syringe at a 100 µL/min rate. The infusion was initiated after a short preparation period required for catheter positioning and cell preparation, typically within approximately 5–10 min after reperfusion. After injection, the catheter was removed and the ECA ligated, ensuring the preservation of antegrade flow through the ICA. To prevent xenogeneic immune rejection, all rats received daily intraperitoneal injections of cyclosporine A (10 mg/kg), starting one day before MCAO. Postoperative analgesics were not routinely administered because of concerns regarding potential effects on neurological and behavioral assessments in the MCAO model. Animals were closely monitored throughout the postoperative period for signs of pain or distress.

Prior to transplantation analysis, brain infarction was confirmed in a separate group of MCAO rats by 2,3,5-triphenyl tetrazolium chloride (TTC) staining and diffusion-weighted MRI evaluation a day after MCAO. The 45-min occlusion duration was selected based on previous studies employing transient MCAO paradigms designed to induce moderate ischemic injury with partial cortical preservation, allowing investigation of post-reperfusion pathophysiology [24, 25]. TTC staining and diffusion-weighted MRI consistently demonstrated infarct formation with relative cortical sparing, confirming the reproducibility and appropriateness of this controlled reperfusion model (Supplementary Fig. 4).

Animals were randomly assigned to the treatment groups. All MCAO procedures were performed by the same surgeon to ensure consistency in infarct size and location across animals.

Safety Evaluation

The safety of hyperacute intra-arterial hcMSC administration was prospectively assessed. Animals were monitored for procedure-related mortality, intracranial hemorrhage, acute neurological deterioration, and longitudinal changes in body weight. Mortality was recorded for up to 35 days after transplantation. In cases of unexpected death, animals were scheduled to undergo postmortem examination to evaluate potential hemorrhagic or procedure-related complications. Acute neurological deterioration was defined as a sudden worsening in neurological status compared with the most recent mNSS. Body weight was measured longitudinally as an indicator of general health status.

Behavioral Testing

Neurological function was evaluated using the modified neurological severity score (mNSS), a composite scale that assesses motor function, sensory function, reflexes, and balance. The total score ranges from 0 to 18, with higher scores indicating more severe neurological dysfunction [26–28]. Behavioral assessments were conducted preoperatively and on days 1, 3, 7, 14, and 28 following MCAO. To minimize bias, all evaluations were performed by two independent observers blinded to treatment groups.

Electrophysiological Evaluation

Motor pathway function was quantitatively assessed using transcranial electrically stimulated motor-evoked potentials (tcMEPs), as previously described [29, 30]. Before tcMEP recording, bone-thinning method was performed to reduce impedance and achieve stable cortical stimulation [31].

Transcranial electrical stimulation and recordings were performed using a Nicolet Endeavor CR intraoperative monitoring system (Nicolet Biomedical, Madison, WI, USA). All tcMEP recordings were obtained under inhalational anesthesia with 1.5% isoflurane. Four stimulus trains over 0.5 ms duration were delivered with supramaximal stimulation. The stimulation intensity was determined during the experiment and maintained constant for each animal throughout the recording sessions. For tcMEP recordings, two needle electrodes were inserted into the contralateral forelimb extensor and hindlimb quadriceps muscles, respectively, with the ground electrode inserted into the hip. Recordings were obtained preoperatively and on days 1, 7, 14, 21, and 28 after MCAO. All recordings were performed by an observer blinded to group allocation.

The amplitude recovery rate was calculated as the ratio of the tcMEP amplitude at each time point to the preoperative baseline amplitude and expressed as a percentage. Onset latency (ms) and amplitude (µV) were separately analyzed for forelimbs and hindlimbs to detect differences in recovery profiles.

Fluorescent Labeling and Tracking of Transplanted hcMSCs

To qualitatively assess the early distribution of transplanted hcMSCs after intra-arterial administration, a separate set of animals was exclusively used for cell tracking analysis. Before transplantation, hcMSCs were labeled with a lipophilic fluorescent membrane dye (PKH26, Sigma-Aldrich) according to manufacturer’s instructions. Briefly, cultured hcMSCs were harvested, washed, and incubated with PKH26 dye, followed by termination of the labeling reaction and extensive washing to remove unbound dye. PKH-labeled hcMSCs were then resuspended in PBS and administered intra-arterially immediately after reperfusion in the MCAO model. The day after transplantation, brain tissue was harvested, fixed, cryosectioned, and counterstained with DAPI. To qualitatively assess the presence of PKH-positive signals within the host brain tissue, sections were examined using fluorescence microscopy.

Immunohistochemistry

An immunohistological evaluation was performed to visually assess neuronal preservation, astroglial responses, and apoptotic changes in the peri-infarct cortex after hcMSC transplantation. On day 3 after MCAO, rats were deeply anesthetized with isoflurane inhalation and transcardially perfused with 4% paraformaldehyde, after which the brains were harvested, embedded in paraffin, and cut into 5-µm-thick coronal sections using a microtome. After deparaffinization by immersion in xylene and ethanol, heat-induced antigen retrieval was performed using an antigen retrieval solution (Dako S2367; pH 9.0) at 98 °C. Sections were then blocked with PBS containing 5% bovine serum albumin and incubated overnight at 4 °C with the following primary antibodies: rabbit polyclonal anti–cleaved Caspase-3 (1:200; PA5114687, Invitrogen), goat polyclonal anti-GFAP (1:10,000; PA5143587, Invitrogen), and mouse monoclonal anti-NeuN (1:500; MA533103, Invitrogen). After washing, sections were incubated with the following species-appropriate highly cross-adsorbed secondary antibodies conjugated with Alexa Fluor dyes; donkey anti-rabbit Alexa Fluor 555(1:1,000; A32794, Invitrogen), donkey anti-goat Alexa Fluor 647 (1:500; A32849, Invitrogen), donkey anti-mouse Alexa Fluor 488 (1:1,000 ; A32766, Invitrogen). The stained brain sections were immersed in PBS in the dark, washed thoroughly, and sealed in VectaShield mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories).

Images were acquired using a point scanning confocal microscope (STELLARIS 5, Leica Microsystems) under identical acquisition settings for all experimental groups. Representative peri-infarct cortical images were used for qualitative comparison between the hcMSC and control groups. For exploratory quantitative analysis, five non-overlapping fields (0.018 mm² per field under a 40x objective) were assessed within the ipsilateral peri-infarct cortex of each animal (n = 3 per group). In accordance with previous reports, the number of NeuN-positive cells was counted in each field and averaged per animal [32–34]. In addition to raw NeuN counts, a NeuN/DAPI ratio was determined to normalize for potential differences in overall cell density [35]. Moreover, GFAP and cleaved Caspase-3 were quantified as immunoreactive area fractions because individual cell borders could not be reliably delineated. All analyses were performed in a blinded manner, and quantification was performed using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Gene Expression Analysis

Gene expression analysis was performed on peri-infarct brain tissue collected in the acute phase at 3 days after MCAO (control, n = 3; hcMSC, n = 4). Total RNA was extracted using the Isogen RNA extraction kit (Nippon Gene, Tokyo, Japan), and cDNA was synthesized by reverse transcription using the ReverTra Ace-α kit (Toyobo Co., Ltd., Osaka, Japan) according to manufacturer’s instructions. The quantitative real-time PCR (qRT-PCR) was conducted on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). The expression levels of selected neuroprotective (Bdnf, Gdnf, Ngf, and Vegf), inflammatory (Il-6, Il-1β, Tnf, Tgfb1), and apoptosis-related (Caspase-3, Caspase-9, Bax, and Bcl2) genes were analyzed using the TaqMan assay, using glyceraldehyde-3-phosphate dehydrogenase (Gapdh) as internal control.

Assessment of Apoptosis

To evaluate apoptosis, western blot analysis and terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining were performed using brain tissue collected 3 days after MCAO. For western blot analysis, peri-infarct brain tissue corresponding to the region used for gene expression analysis was collected from each group (n = 4 per group). Samples were homogenized in radioimmunoprecipitation assay (RIPA) buffer supplemented with protease and phosphatase inhibitors, and protein concentrations were determined using a bicinchoninic acid (BCA) protein assay. Equal amounts of protein (40 µg) were separated on 8–16% sodium dodecyl sulfate-polyacrylamide gels and transferred onto polyvinylidene difluoride (PVDF) membranes. After blocking for 1 h at room temperature, membranes were incubated overnight at 4 °C with an anti-Caspase-3 antibody (1:1000; #9662, Cell Signaling Technology), and an anti-GAPDH antibody (1:10000; 014-22383, FUJIFILM Wako). Membranes were subsequently incubated with horseradish peroxidase-conjugated secondary antibody (1:2000; A0545, Sigma-Aldrich) for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence reagents and quantified using ImageJ software. The cleaved/total Caspase-3 ratio was calculated to assess Caspase-3 activation, whereas total Caspase-3 expression was normalized to GAPDH.

TUNEL staining was also performed for qualitative assessment of apoptotic cell death on paraffin-embedded coronal brain sections collected 3 days after MCAO using an In Situ Apoptosis Detection Kit (Takara Bio). Sections were counterstained with DAPI and examined by fluorescence microscopy using a laser scanning confocal microscope (STELLARIS 5, Leica Microsystems).

Statistical Analysis

Statistical analyses were performed using JMP software, version 18.1 (SAS Institute Inc., Cary, NC, USA). Data are expressed as mean ± SD. Given the small sample size and potential non-normality of the data, between-group comparisons were performed using the Mann–Whitney U test. For repeated longitudinal comparisons in mNSS and tcMEP analyses, p-values were adjusted using the Holm–Bonferroni method to account for multiple comparisons across post-stroke time points. Comparisons between the hcMSC and control groups included mNSS, tcMEP amplitude recovery rates, NeuN-positive cell counts, NeuN/DAPI ratios, and qRT-PCR–based gene expression levels. A two-sided p-value < 0.05 was considered statistically significant. For RNA sequencing analysis, differential gene expression between hcMSCs and hbMSCs was assessed using DESeq2, with statistical significance defined as an absolute log₂ fold change > 1.5 and an adjusted p-value < 0.05.

Results

RNA Sequencing Analysis of hcMSCs and hbMSCs

RNA sequencing analysis demonstrated distinct transcriptomic profiles between hcMSCs and hbMSCs. Principal component analysis based on normalized gene expression data showed clear separation between the two groups, indicating substantial differences in their global transcriptional signatures (Fig. 1a). Consistent with this finding, an MA plot demonstrated widespread transcriptional differences between hcMSCs and hbMSCs, with numerous significantly upregulated and downregulated genes identified in hcMSCs (Fig. 1b).

Fig. 1.

Transcriptomic comparison between hcMSCs and hbMSCs. (a) Principal component analysis (PCA) based on normalized RNA sequencing data demonstrating clear separation between hcMSCs and hbMSCs (n = 3 per group). PC1 and PC2 explained 61.9% and 31.3% of the total variance, respectively. (b) MA plot showing differentially expressed genes between hcMSCs and hbMSCs. Red dots indicate genes with an adjusted p-value ≤ 0.05 and an absolute log₂ fold change (|log₂FC|) > 1.5, whereas blue dots indicate genes with an adjusted p-value > 0.05 or |log₂FC| ≤ 1.5. (c) Gene ontology (GO) enrichment analysis of differentially expressed genes. Upregulated genes in hcMSCs were enriched for biological processes related to cell cycle phase transition, inflammatory response, blood vessel morphogenesis, immune regulation, and neuronal and glial differentiation, whereas downregulated genes were enriched for pathways associated with cell adhesion, extracellular matrix organization, cell death, and chemotaxis

Representative genes included Ereg, Grem1, and Fgl2, which were significantly upregulated in hcMSCs, whereas Vcam1, Tnfsf15, and Dram1 were significantly downregulated compared with hbMSCs (Supplementary Fig. 5). Differential expression analysis further identified genes associated with immune regulation, including Fgl2, as being upregulated in hcMSCs, while genes involved in inflammatory activation and cell death, including Vcam1, Tnfsf15, and Dram1, were downregulated.

Gene ontology analysis revealed enrichment of biological processes related to neuronal and glial differentiation, axonal growth, synaptic signaling, blood vessel morphogenesis, and immune regulation among genes upregulated in hcMSCs. In contrast, genes downregulated in hcMSCs were enriched for pathways associated with cell adhesion, extracellular matrix organization, apoptosis-related signaling, and chemotaxis (Fig. 1c).

Neurological Recovery After hcMSC Transplantation

Longitudinal mNSS in hcMSC and control groups are shown in Fig. 2. On day 1 after MCAO, no significant difference in mNSS was observed between the groups (hcMSC group 8.4 ± 1.7 vs. control 8.3 ± 1.5, p = 0.90), confirming a comparable degree of neurological impairment before treatment. Subsequent behavioral assessments revealed progressive neurological improvement in the hcMSC group compared with the control group. After Holm–Bonferroni correction for multiple comparisons, significant between-group differences were observed on day 3 (adjusted p = 0.026), day 7 (adjusted p = 0.020), day 14 (adjusted p = 0.025), and day 28 (adjusted p = 0.022), indicating sustained neurological recovery in the hcMSC group.

Fig. 2.

Longitudinal assessment of neurological function after transient MCAO. Longitudinal changes in modified neurological severity score (mNSS) following transient MCAO in rats treated with intra-arterial hcMSC transplantation (n = 8) or vehicle control (n = 7). Data are presented as mean ± SD with individual data points. The hcMSC group showed significantly greater neurological recovery than the control group beginning on day 3 and continuing through days 14 and 28. *Adjusted p < 0.05 versus control

Electrophysiological Recovery After hcMSC Transplantation

Representative tcMEP waveforms and longitudinal amplitude changes recorded from the forelimb and hindlimb are shown in Fig.3. On day 1 after MCAO, tcMEP amplitudes in forelimb and hindlimb muscles were markedly reduced compared with baseline, without significant differences between the hcMSC and control groups (forelimb: hcMSC, 4,825 ± 978 µV; control, 5,364 ± 2,176 µV; p = 0.77; hindlimb: hcMSC, 4,047 ± 1,263 µV; control, 3,450 ± 696 µV; p = 0.27).

Fig. 3.

Longitudinal electrophysiological assessment of motor pathway function using transcranial motor-evoked potentials (tcMEPs). Representative tcMEP waveforms recorded from (a) forelimb and (b) hindlimb muscles before stroke and during follow-up after transient MCAO in rats treated with intra-arterial hcMSC transplantation or vehicle control. Longitudinal changes in tcMEP amplitude recovery rates in the (c) forelimb and (d) hindlimb. Data are presented as mean ± SD with individual data points. *Adjusted p < 0.05 versus control after Holm–Bonferroni correction

In both groups, tcMEP amplitudes gradually recovered over the study period. In the forelimb, the hcMSC group exhibited significantly greater tcMEP recovery than the control group from day 14 onward. After Holm–Bonferroni correction for multiple comparisons, significant between-group differences remained at day 14 (hcMSC: 93.3 ± 21.4% vs. control: 61.1 ± 17.6%; adjusted p = 0.032), day 21 (hcMSC: 104.8 ± 22.1% vs. control: 71.4 ± 13.9%; adjusted p = 0.010), and day 28 (hcMSC: 129.3 ± 23.7% vs. control: 78.0 ± 19.3%; adjusted p = 0.011), indicating sustained electrophysiological recovery in the hcMSC group. In contrast, in the hindlimb, the hcMSC group also demonstrated greater amplitude recovery compared with the control group at 2 weeks after MCAO (hcMSC: 131.5 ± 32.1% vs. control: 83.3 ± 23.7%; raw p = 0.003). This difference remained statistically significant after Holm–Bonferroni correction (adjusted p = 0.016). Thereafter, tcMEP amplitudes continued to improve in both groups, and no significant between-group differences were observed at later time points after correction.

tcMEP onset latencies remained stable throughout the observation period and did not differ significantly between groups, indicating preserved conduction velocity despite ischemic injury.

PKH Tracking and Immunofluorescence Analyses

Fluorescence microscopy confirmed robust PKH-positive membrane labeling of cultured hcMSCs, with clear cell morphology and colocalization with DAPI-stained nuclei. Brain sections obtained 24 h after intra-arterial administration of PKH-labeled hcMSCs were examined. PKH-positive signals were detected within the ipsilateral peri-infarct region, providing qualitative evidence of early cerebral delivery following intra-arterial hcMSC administration (Fig. 4). Given the exploratory nature of this analysis, no quantitative assessment of signal distribution was performed.

Fig. 4.

PKH-positive signals and immunohistochemical findings in the peri-infarct cortex. (a) Hematoxylin and eosin–stained coronal section showing the infarct core and peri-infarct region. The circled area indicates the region examined for PKH-positive signals, whereas the boxed area indicates the peri-infarct cortex used for immunofluorescence analyses. (b) PKH26-labeled hcMSCs in culture. (c) PKH-positive signals detected in the peri-infarct region 24 h after intra-arterial hcMSC administration. Representative immunofluorescence images of comparable peri-infarct regions demonstrate more structurally preserved NeuN-positive cells in the hcMSC group (d), with more uniform and intense staining than in the control group (e). Scale bar, 50 μm. (f) Representative image showing scattered cleaved Caspase-3–positive cells frequently observed in both groups, often accompanied by attenuated or diminished NeuN immunoreactivity. Scale bar, 20 μm

Representative immunofluorescence analyses are also shown in Fig. 4. The peri-infarct cortex showed structurally preserved NeuN-positive cells in the hcMSC group, with more uniform and intense staining than in the control group. In addition, characteristic scattered cleaved Caspase-3–positive cells were observed in the peri-infarct region in both groups, often accompanied by attenuated or diminished NeuN immunoreactivity. GFAP-positive astrocytic processes were also observed in both groups without apparent qualitative differences. Exploratory quantitative analyses of NeuN-positive cells, NeuN/DAPI ratios, GFAP-positive area fraction, and cleaved Caspase-3–positive area fraction are provided in Supplementary Fig. 6.

Gene Expression at the Cerebral Peri-Infarction Site

To assess molecular changes associated with hcMSC transplantation, mRNA expression levels in peri-infarcted brain tissue were analyzed by qRT-PCR on day 3 after MCAO (Fig. 5). Pro-inflammatory cytokines, including Il-6, Il-1β, and Tnf, as well as pro-apoptotic genes, such as Caspase-3, Caspase-9, and Bax, were significantly downregulated in the hcMSC group compared with the control group. Although the Bcl2/Bax ratio did not reach statistical significance, a decreasing trend was observed in the hcMSC group.

Fig. 5.

Quantitative real-time PCR analysis of gene expression in peri-infarct brain tissue on day 3 after MCAO. Expression levels of neurotrophic factors (Bdnf, Gdnf, Ngf, Vegf), inflammatory cytokines (Il-6, Il-1β, Tnf, Tgfb1), and apoptosis-related molecules (Caspase-3, Caspase-9, Bax) were normalized to Gapdh. Individual data points represent individual animals (control, n = 3; hcMSC, n = 4). Data are presented as mean ± SD. *P < 0.05 compared with the control group

In contrast, vascular endothelial growth factor (Vegf) expression was significantly upregulated in the hcMSC group. The expression of other neurotrophic factors, including nerve growth factor (Ngf), brain-derived neurotrophic factor (Bdnf), and glial cell line–derived neurotrophic factor (Gdnf), did not differ significantly between groups.

Apoptosis-Related Protein Expression and TUNEL Staining

Western blot analysis demonstrated a significantly reduced cleaved/total Caspase-3 ratio in the hcMSC group than in the control group (p = 0.03; Fig. 6), indicating reduced Caspase-3 activation. In contrast, total Caspase-3 expression normalized to GAPDH did not differ significantly between the two groups (p = 0.31). Full-length western blot images corresponding to the quantitative analyses are provided in Supplementary Fig. 8.

Fig. 6.

Apoptosis-related protein expression and representative TUNEL staining in the peri-infarct region. (a) Representative western blot images showing total Caspase-3 (32 kDa), cleaved Caspase-3 (17 kDa), and GAPDH (34 kDa) in peri-infarct tissue collected 3 days after MCAO. (b) Quantification of the cleaved/total Caspase-3 ratio demonstrated significantly lower Caspase-3 activation in the hcMSC group than in the control group. (c) Total Caspase-3 expression normalized to GAPDH did not differ significantly between groups. Data are presented as box-and-whisker plots with individual data points (n = 4 per group). *p < 0.05 (Mann–Whitney U test). (d) Representative TUNEL staining of the peri-infarct region in the control (upper panels) and hcMSC (lower panels) groups showing DAPI (left), TUNEL (middle), and merged images (right). Representative images suggest fewer TUNEL-positive signals in the hcMSC group than in the control group. Scale bars = 100 μm

Representative TUNEL staining of the peri-infarct region is also shown in Fig. 6. Consistent with the western blot findings, fewer TUNEL-positive signals were observed in the hcMSC group than in the control group, supporting a reduction in apoptotic cell death following hcMSC treatment.

Safety Outcomes

Hyperacute intra-arterial hcMSC administration was technically feasible and well tolerated. Procedure-related deaths occurred only during MCAO model induction before treatment allocation and were excluded according to the predefined criteria. During the 35-day follow-up period, no treatment-related mortality or overt neurological deterioration suggestive of hemorrhagic complications was observed. Body weight transiently decreased in both groups during the acute phase after MCAO and gradually recovered thereafter, with no significant differences between groups at any time point (Supplementary Fig. 7).

Discussion

This study positions hyperacute intra-arterial hcMSC transplantation as a biologically protective adjunctive strategy targeting early post-reperfusion injury. Administered immediately after reperfusion, hcMSCs conferred significant functional and electrophysiological benefits in a preclinical model of transient ischemia. The integration of behavioral, electrophysiological, histological, and molecular analyses provides converging evidence that hcMSC transplantation promotes neuroprotection and functional recovery during the critical early phase following reperfusion. These findings are particularly relevant given the increasing recognition that angiographic recanalization does not necessarily ensure optimal tissue recovery. Even after macrovascular recanalization, ongoing inflammatory and apoptotic processes within the peri-infarct territory contribute to infarct progression and neurological impairment. In this context, adjunctive strategies targeting early post-reperfusion tissue injury may extend therapeutic benefit beyond mechanical recanalization.

Post-recanalization tissue injury likely reflects the convergence of reperfusion injury and progressive microvascular dysfunction. Oxidative and inflammatory endothelial damage promotes capillary obstruction, while sustained hypoperfusion further exacerbates cellular stress and neurovascular disruption, forming a self-perpetuating cycle that limits tissue recovery despite angiographic success [4–6]. Whereas intra-arterial thrombolytic strategies primarily aim to dissolve residual distal thrombus and improve microvascular patency, hcMSC-based therapy may act through a fundamentally different mechanism [36–39]. Rather than focusing on clot dissolution alone, hyperacute hcMSC administration appears to modulate molecular processes associated with neurovascular unit dysfunction by stabilizing endothelial integrity, attenuating inflammatory signaling, and suppressing apoptosis within the peri-infarct region. Beyond neuronal protection, MSC-derived paracrine factors may also influence the vascular component of the neurovascular unit. Interactions within the neurovascular unit are increasingly recognized as critical determinants of secondary injury after ischemic stroke [40]. In the present study, suppression of inflammatory cytokines in the hcMSC group may indicate a microenvironment favorable for endothelial stability and preservation of microvascular integrity following reperfusion. This mechanistic distinction suggests the potential of MSC-based therapy as a biologically protective adjunctive strategy in the immediate post-recanalization phase.

A key strength of the present study is the incorporation of longitudinal tcMEP assessment, which provides electrophysiological evidence of functional recovery after stroke. Unlike behavioral scores such as mNSS, tcMEPs directly evaluate the integrity of corticospinal motor pathways. Previous experimental studies of MSC therapy for stroke have primarily relied on behavioral or histological endpoints, whereas electrophysiological assessments have been used less frequently because stable longitudinal tcMEP recording in rodents is technically demanding. Building on our previously established bone-thinning technique for repeated tcMEP monitoring, we were able to perform serial electrophysiological evaluation throughout the recovery period [31]. In the present study, hcMSC-treated animals demonstrated significantly greater recovery of tcMEP amplitudes, particularly in the forelimb, where differences persisted through the chronic phase. These results suggest that hcMSC therapy enhances recovery of motor pathway function beyond improvements detected by behavioral assessment alone and support the use of tcMEP as a translational biomarker for evaluating neurorestorative interventions after ischemic stroke.

Our previous studies established that hcMSCs fulfill the defining criteria of MSCs, including robust proliferative capacity and multilineage differentiation potential, while exhibiting distinctive immunophenotypic and paracrine characteristics enriched in neurotrophic and angiogenic factors [17, 18, 22]. In the present study, exploratory transcriptomic analysis demonstrated that hcMSCs possess a gene expression profile distinct from that of conventionally used hbMSCs. Differentially expressed pathways included those related to neuronal differentiation, axonal growth, synaptic signaling, angiogenesis, and immune regulation, whereas pathways associated with inflammatory cell migration and cell death were relatively suppressed. However, these transcriptomic findings were obtained under baseline culture conditions and should be interpreted as descriptive and hypothesis-generating rather than mechanistic evidence explaining the in vivo therapeutic effects observed in this study.

The therapeutic efficacy of hcMSCs in this context is closely linked to both the route and timing of administration. Although intravenous delivery has been shown to confer neuroprotection in experimental stroke models [18, 19], cerebral biodistribution is limited by first-pass pulmonary entrapment, and therapeutic benefit is considered to arise primarily from paracrine mechanisms rather than direct cell engraftment within the ischemic territory [41–43]. In contrast, intra-arterial administration represents an alternative strategy to enhance cerebral engraftment, with preclinical studies showing preferential MSC accumulation in the peri-infarct region [41, 44–46]. In the present study, PKH-positive signals were detected within brain Sect.  24 h after intra-arterial administration, providing qualitative evidence of early cerebral delivery following hcMSC transplantation. Beyond the route of administration, timing represents another critical determinant of therapeutic efficacy in the post-recanalization setting. Experimental studies indicate that administration within the early phase of reperfusion maximizes neuroprotection efficacy [46, 47]. The optimal therapeutic window likely reflects early paracrine modulation of reperfusion-related inflammatory and oxidative injury rather than delayed cellular migration associated with blood–brain barrier disruption [45, 46, 48]. In the present study, hcMSCs were administered immediately after reperfusion, allowing early localization and interaction with peri-infarct tissue during a phase characterized by active inflammatory and apoptotic signaling. This concept is supported by findings from our previous study demonstrating that hcMSCs exposed to oxidative and inflammatory stress adopt a secretory profile enriched in neuroprotective mediators [18]. In contrast to thrombolytic agents that primarily target residual clot, hcMSC therapy may attenuate oxidative-inflammatory amplification contributing to post-recanalization injury.

Mechanistic support for this proposed therapeutic window arises from molecular analyses at both the transcriptional and protein levels revealing significant suppression of pro-inflammatory cytokines such as Il-6, Il-1β, and Tnf in peri-infarct tissue during the acute phase. Given the central role of neuroinflammation in secondary ischemic injury, this attenuation likely contributed to the preservation of peri-infarct tissue integrity [26, 49, 50]. In parallel, proapoptotic mediators such as Caspase-3, Caspase-9, and Bax were downregulated, suggesting inhibition of mitochondria-dependent apoptotic cascades, consistent with previous reports on MSC-mediated neuroprotection [51–53]. Consistent with these transcriptional changes, western blot analysis demonstrated a significantly reduced cleaved/total Caspase-3 ratio without a change in total Caspase-3 expression, indicating attenuation of Caspase-3 activation. Although cleaved caspase-3 is widely used as a marker of apoptosis, its expression after experimental stroke may also reflect non-apoptotic processes, including reactive astrogliosis and tissue remodeling [54]. Therefore, the observed reduction in cleaved Caspase-3 should be interpreted together with the accompanying TUNEL findings, apoptosis-related gene expression changes, and neuronal preservation, which collectively strengthen the evidence that hcMSC treatment attenuates post-stroke apoptotic cell death. These findings underscore the neuroprotective potential of hcMSC-mediated suppression of inflammatory and apoptotic signaling in the peri-infarct region. Additionally, Vegf expression was significantly upregulated following hcMSC transplantation. While Vegf is recognized as a pleiotropic mediator in tissue repair and vascular responses in the postischemic brain [40], the contribution of elevated Vegf to the observed neuroprotection remains to be further elucidated. Together, these findings indicate that hcMSC transplantation immediately after reperfusion effectively modulates early inflammatory and apoptotic responses within the ischemic penumbra, promoting sustained functional recovery.

Safety considerations are essential for intra-arterial cell delivery. In the present study, hyperacute hcMSC infusion was well tolerated, with no procedure-related mortality or overt neurological deterioration observed. Animals were monitored throughout follow-up by serial neurological assessment, survival monitoring, and body weight measurements. The administration protocol was selected based on prior safety-oriented investigations of intra-arterial MSC delivery to minimize the risk of microvascular obstruction. Previous studies have demonstrated that the safety of intra-arterial MSC delivery critically depends on cell dose, infusion velocity, and maintenance of arterial blood flow [55]. Excessive cell doses may impair cerebral perfusion and increase the risk of microvascular obstruction, whereas low-dose administration with preserved arterial flow and controlled infusion speed substantially reduces these complications [46, 56]. In particular, Fukuda & Horie et al. reported functional recovery without adverse effects using low-dose intra-arterial MSC transplantation [57]. Based on these findings, we adopted a further reduced dose (1 × 10⁴ cells) with slow infusion (100 µL/min) under maintained arterial flow, constituting an optimized safety-oriented protocol.

Despite these mechanistic considerations, the principal finding of the present study is not direct evidence of no-reflow rescue or mitigation of reperfusion injury. Rather, our data demonstrate that hyperacute intra-arterial hcMSC administration is compatible with reperfusion and provides additional functional, electrophysiological, and molecular benefits beyond recanalization alone. Recent STAIR recommendations have emphasized the importance of therapies that preserve vulnerable tissue and augment the effects of reperfusion during the hyperacute stage of ischemic stroke, while emerging concepts in stroke therapy increasingly recognize the importance of interventions administered immediately before, during, or after reperfusion [58, 59]. The present study directly aligns with these priorities by targeting the earliest post-reperfusion period and demonstrates the feasibility and potential efficacy of hcMSC therapy as an adjunct to recanalization. In this context, hcMSC transplantation warrants further investigation as a translational cerebroprotective strategy for acute ischemic stroke.

Several limitations of the present study should be acknowledged. First, cerebral blood flow during MCAO and reperfusion was not directly monitored using laser Doppler flowmetry in the present study. Although this technique is commonly used to confirm the degree of ischemia, we instead applied predefined inclusion criteria based on neurological function (mNSS) to ensure a relatively uniform level of ischemic injury across animals. In addition, infarct reproducibility was confirmed using TTC staining and diffusion-weighted imaging in separate cohorts. Nevertheless, variability in cerebral blood flow among animals cannot be excluded and may have influenced the observed outcomes.

Second, although hyperacute intra-arterial hcMSC administration improved behavioral and electrophysiological outcomes and was associated with the attenuation of inflammatory and apoptotic signaling, the precise cellular targets and causal molecular mechanisms underlying these effects were not directly examined. In the present study, direct assessment of both microvascular perfusion (e.g., Laser Doppler flowmetry) and BBB integrity (e.g., Evans blue extravasation) was not performed; therefore, potential effects of hyperacute hcMSC delivery on microcirculation dysfunction and BBB disruption cannot be determined. In addition, reliable quantitative assessment, long-term tracking, and evaluation of terminal differentiation of transplanted cells were technically challenging because a relatively low cell dose was selected for intra-arterial administration based on safety considerations. Furthermore, PKH26 labeling provides only qualitative evidence of cell delivery and may be affected by dye leakage and uptake by host cells, limiting the specificity of cell-tracking results. Therefore, durable engraftment and lineage-specific differentiation of hcMSCs could not be determined in the present study. Nevertheless, the early functional and molecular changes observed in the early post-reperfusion phase may be more consistent with paracrine-mediated mechanisms than with direct structural replacement by transplanted cells.

Third, this study was designed as a proof-of-concept experiment using a fixed administration protocol in a transient MCAO model and did not address treatment optimization or long-term durability of recovery. Accordingly, the translational implications of hyperacute intra-arterial hcMSC transplantation should be interpreted with appropriate caution. Fourth, the present study did not include a direct in vivo comparison between hcMSCs and hbMSCs or other MSC sources. The primary aim of the present study was to evaluate the therapeutic potential of hyperacute intra-arterial hcMSC administration in a reperfusion model rather than to compare the relative efficacy among different MSC populations. In addition, transcriptomic analyses comparing hcMSCs and hbMSCs were performed only under baseline culture conditions and did not evaluate cellular responses under ischemic or inflammatory stress conditions such as oxygen-glucose deprivation. Therefore, the biological relevance of the observed molecular differences within the post-ischemic microenvironment remains uncertain. Further comparative studies incorporating other MSC sources and ischemic stress models will be necessary to clarify the significance of these in vitro findings in stroke pathology.

Finally, although no procedure-related mortality or overt neurological deterioration was observed, systematic imaging-based or histological assessment of infusion-related microembolic events or hemorrhagic transformation was not performed. A direct comparison with alternative adjunctive reperfusion strategies was also not conducted. Consequently, the safety profile should be interpreted within the constraints of the selected dose and experimental conditions.

The limited sample size represents an additional limitation (hcMSC group, n = 8; control group, n = 7) which may have reduced statistical power to detect differences in some endpoints. This increases the risk of type II error and limits generalizability to larger preclinical and clinical studies.

Conclusion

Hyperacute intra-arterial transplantation of hcMSCs after reperfusion significantly promoted neurological and electrophysiological recovery in a transient middle cerebral artery occlusion (MCAO) model. Unlike thrombolytic approaches that primarily promote clot dissolution, hcMSC therapy may represent a mechanistically distinct adjunctive strategy targeting neurovascular unit injury. Collectively, these findings support further investigation of hyperacute intra-arterial hcMSC transplantation as a reperfusion-compatible cerebroprotective strategy and a potential adjunct to endovascular therapy for acute ischemic stroke.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 ^{(10.1MB, pdf)}

Author Contributions

TH conceived and designed the study, performed experiments, analyzed data, and wrote the manuscript. MK contributed to study design and data analysis. MT, YM, HT, and SM contributed to data acquisition and analysis.TI, TK, and SH contributed to experimental methodology, data analysis, and interpretation. TM and NH supervised the study and critically revised the manuscript.All authors reviewed the manuscript.

Funding

Open Access funding provided by Hiroshima University. This research was partially supported by a Grant-in-Aid for Scientific Research from the Japanese Society for the Promotion of Science (Issue No. 23K085220A).

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Competing Interests

The authors declare no competing interests.