Feasibility and Efficacy of Intra‐Arterial Administration of Mesenchymal Stem Cells in an Animal Model of Double Toxin‐Induced Multiple System Atrophy

Abstract Multiple system atrophy (MSA) is a sporadic neurodegenerative disease of the central and autonomic nervous system. Because no drug treatment consistently benefits MSA patients, neuroprotective strategy using mesenchymal stem cells (MSCs) has a lot of concern for the management of MSA. In this study, we investigated the safety and efficacy of intra‐arterial administration of MSCs via internal carotid artery (ICA) in an animal model of MSA. The study was composed of feasibility test using a ×10 and ×50 of a standard dose of MSCs (4 × 107 MSCs) and efficacy test using a ×0.2, ×2, and ×20 of the standard dose. An ultrasonic flow meter and magnetic resonance imaging (MRI) showed that no cerebral ischemic lesions with patent ICA blood flow was were observed in animals receiving a ×10 of the standard dose of MSCs. However, no MSA animals receiving a ×50 of the standard dose survived. In efficacy test, animals injected with a ×2 of the standard dose increased nigrostriatal neuronal survival relative to a ×0.2 or ×20 of the standard dose. MSA animals receiving MSCs at ×0.2 and ×2 concentrations of the standard dose exhibited a significant reduction in rotation behavior relative to ×20 of the standard dose of MSCs. Cerebral ischemic lesions on MRI were only observed in MSA animals receiving a ×20 of the standard dose. The present study revealed that if their concentration is appropriate, intra‐arterial injection of MSCs is safe and exerts a neuroprotective effect on striatal and nigral neurons with a coincidental improvement in motor behavior. Stem Cells Translational Medicine 2017;6:1424–1433


INTRODUCTION
Multiple system atrophy (MSA) is a sporadic neurodegenerative disease of the central and autonomic nervous system, and its clinical features include any combination of autonomic failure, parkinsonism, and pyramidal signs. Pathologically, MSA is regarded as a unique entity within the spectrum of oligodendrogliopathy, with extensive a-synuclein-positive glial cytoplasmic inclusions, and leading to striatonigral degeneration, olivopontocerebellar degeneration, astrogliosis, and microgliosis [1]. Because of grave prognosis, many clinical studies have been conducted to modify disease progression in MSA, such as trials involving growth hormone, immunoglobulin, monoamine oxidase type B (MOB-B) inhibitor, or rifampicine [2]. However, these trials have failed to achieve neuroprotective efficacy.
Mesenchymal stem cells (MSCs) are multipotent stem cells present in adult bone marrow that are capable of differentiating into various cell types under appropriate conditions. Additionally, MSCs secrete various cytotropic factors including neurotrophic growth factors, chemokines, cytokines, and extracellular matrix protein, which in turn, exert neuroprotective effects [3,4]. Our previous studies show that MSCs have potent neuroprotective effects in animal models of parkinsonian disorders through modulation of neuroinflammation, inhibition of apoptotic cell death, increases in neurogenesis and neuronal differentiation, and enhancement of autophagy [5][6][7][8]. In addition, MSC exert a neuroprotective effect in animal model of MSA [9]. Moreover, we recently demonstrated in investigator initiated trials that intra-arterial administration of MSCs delays progression of neurological deficits assessed clinically and radiologically in patients with MSA [10].
Of several potential delivery routes, intra-arterial administration of MSCs has received attention because this approach increases cell homing to the brain by circumventing the pulmonary circulation and thus leads to enhanced therapeutic outcomes [11][12][13]. However, cerebral ischemia related to intra-arterial cell infusion has been reported, although the complications related to intra-arterial administration of stem cells seem to depend on the infusion technique, cell dose, and infusion velocity [14,15]. Therefore, a careful optimization of the intra-arterial infusion procedures, especially with regard to cell dose is required before pharmacotherapeutic development of stem cell therapy. In the present study, we investigated the safety and efficacy of intraarterial administration of MSCs at different cell doses using a double-toxin-induced animal model of MSA.

Animal Study and Experimental Composition
To create a double toxin-induced animal model of MSA, adult male or female Sprague Dawley rats (200-220 g) were injected with 6-hydroydopamine (6-OHDA, Sigma, St. Louis, MO, USA) and quinolinic acid (QA, Sigma) [16]. 6-OHDA (8 lg/4 ll) was injected by stereotaxic injection into the left medial forebrain bundle (A 22.2, L 1 1.5 relative to bregma, V 28.0 relative to the dura, and with the tooth bar 4.5 mm above the interaural line). Three-four weeks after 6-OHDA injection, a total of 150 nmol/4 ml QA was stereotaxically injected into two sites of the left striatum (A 11.4/ 13.0, L 13.5/13.0 relative to bregma, V 24.0 relative to the dura, and with the tooth bar 4.5 mm above the interaural line). All injections were made over 1 minute, and the needle was left in place for a further 2 minutes before being slowly withdrawn. The rats were randomly divided into three groups (n 5 5; each group): (1) the control group; (2) the normal saline-treated group; (3) the MSC-treated group. The animal study was composed of two consecutive experiments; the first experiment was to test feasibility and distribution of intra-arterially injected MSCs and the second experiment was to exam efficacy of different doses of MSCs. The animal work was approved by the Institutional Animal Care and Use Committees of Yonsei University.

MSC Transplantation
One week after QA injection, MSCs were injected into the ICA or tail vein. Based on MSC concentration used by previous clinical trials [10], a concentration of 4 3 10 7 cells/60 kg was selected as a standard dose of MSCs. The cell dose of human MSCs (hMSCs) used in the first experiment was a 310 and 350 of the standard dose of MSCs. Then, a 30.2, 32, and 310 of the standard dose were used in the second experiment. For ICA injection, animals were anesthetized with 30% isoflurane. According to the previous experiment [14], the common carotid artery (CCA) was exposed and the external carotid artery and pterygopalatine artery were ligated. MSCs in 100 ll of solvent (Corestem, Inc., Seoul, Korea) were injected via a 33G microneedle into the CCA with preserved blood flow in the CCA and ICA for 30 seconds and bleeding was controlled by application of bioabsorbable Gelfoam (Pfizer, New York, NY, USA) at the injection site for 5 minutes. Pressure was maintained at the injection site until bleeding ceased and then fixed external carotid artery and pterygopalatine artery opened.

Isolation of hMSCs
Bone marrow aspirates (10 ml) were obtained from the iliac crests of human donors. The mononuclear cell layer was isolated by Corestem, Inc. At passage 6, hMSCs were injected into rats via the CCA or tail vein.

Fluorescence Activated Cell Sorting Analysis
The characterization of hMSCs was evaluated by fluorescence activated cell sorting (FACS) analysis. Briefly, cell suspensions were washed twice with phosphate buffered saline (PBS, Sigma) containing 0.1% bovine serum albumin (BSA, Sigma). For direct assays, 1 3 10 5 cells/ml were incubated with fluorescein isothiocyanate (FITC)-conjugated CD34 and CD45 (negative markers) as well as CD105 and anti-human CD73 (positive markers) at 48C for 30 minutes, and then washed twice with PBS containing 0.1% BSA. The cells were analyzed by cytometric analysis using an EPICS XL flow cytometer (Becton-Dickinson, San Diego, CA, USA) with EXPO32 software.

Behavioral Test
Three weeks after the 6-OHDA lesion, apomorphine-induced rotation was assessed (apomorphine hydrochloride, 0.5 mg/kg S.C. dissolved in sterile water; Sigma). Only animals showing a mean of 6 contralateral rotations/minutes over 30 minutes following the administration of apomorphine were retained. Behavioral test was repeated 2-3 weeks after striatal lesioning with QA as well as 1, 2, and 3 weeks after MSC transplantation.

Tissue Preparation
For immunohistochemistry, rats were perfused with a saline solution containing 0.5% sodium nitrate and heparin (10 U/ml) and fixed with 4% paraformaldehyde dissolved in 0.1 M phosphate buffer (PB) (50 ml/rat) at 8 weeks after the first injection. Brains were removed from the skulls, post-fixed overnight in buffered 4% paraformaldehyde at 48C, and stored in a 30% sucrose solution for 1-2 days at 48C until they sank. Coronal sections (30 lm) were obtained and stored in tissue stock solution (30% glycerol, 30% ethylene glycol, 30% three times distilled water, 10% 0.2 M PB) at 48C until use.

Immunohistochemistry
The coronal brain sections (4 lm thick) were rinsed twice in PBS and incubated in 0.2% Triton X-100 for 30 minutes at room temperature (RT). They were rinsed three times and blocked with 0.5% BSA in 31 PBS. After blocking, sections were incubated at 48C overnight with the following primary antibodies: mouse antityrosine hydroxylase (TH; 1:2,000; Pel-freez, Rogers, AR, USA), rabbit anti-dopamine-and cAMP-regulated neuronal phosphoprotein (DARPP-32) (1:7,500; Abcam, Cambridge, MA, USA), rabbit anti-Annexin V (1:500; Abcam, Cambridge, MA, USA), mouse anti-NeuN (1:500; Abcam, Cambridge, MA, USA), or mouse anti-nuclear matrix (NuMA; 1:100; Calbiochem, San Diego, CA, USA). Following overnight incubation, the brain sections were rinsed three times with 0.5% BSA in 31 PBS (10 minutes/rinse) and incubated with the appropriate biotinylated secondary antibody and avidin-biotin complex (Elite Kit; Vector Laboratories, Burlingame, CA, USA) for 1 hour at RT. Bound antibodies were visualized by incubating the sections with 0.05% diaminobenzidine-HCl (DAB, Dako, carpinteria, CA, USA) and 0.003% hydrogen peroxide in 0.1 M PB. The brain sections were rinsed with 0.1 M PB for DAB inhibition. Immunostained cells were analyzed by bright-field microscopy and viewed under a Zeiss LSM 780 confocal imaging system (Zeiss, Jena, Thuringia, Germany). To analyze the localizations of antigens in doublestained samples, immunofluorescence images were created from the same tissue sections and merged using the Zeiss ZEN software (Zeiss).

MSC Labeling for In Vivo Fluorescent Imaging
MSCs were labeled with NEO-LIVE TM -Magnoxide675 (Biterials, Seoul, South Korea), a fluorescent magnetic nanoparticle used for in vivo fluorescent imaging. NEO-LIVE TM -Magnoxide675 nanoparticles were prepared according to the manufacturer's protocol and MSCs were incubated for 24 hours with labeling solution containing NEO-LIVE TM -Magnoxide 675 at a concentration of 0.4 mg/ ml in cultured media. Before animal injection, labeled MSCs were analyzed by fluorescence microscopy.

Optical In Vivo Imaging System and Magnetic Resonance Imaging
For fluorescence images, animals were anesthetized with 30% isoflurane and images obtained using an Spectrum in vivo imaging system (IVIS; Caliper Lifesciences, Hopkinton, Massachusetts, USA) for data acquisition and analysis at 1 day, 3 days, 1 week, 2 weeks, 4 weeks, and 8 weeks after MSC transplantation. For magnetic resonance imaging (MRI), animals were anesthetized with 30% isoflurane, and images were obtained using a BioSpec 94/20 USR MRI (Bruker BioSpin, Ettlingen, Germany) for data acquisition and analysis 1 week after MSC transplantation. The in vivo imaging parameters were: (1) a three-dimensional T2*-weighted gradient echo sequence with repetition time (TR) 5 2,500 ms and echo time (TE) 5 24 ms for transverse sections of the brain, (2) a T2-weighted two-dimensional spin echo sequence with TR

Ultrasonic Blood Flow Measurement
For operative procedures, animals were anesthetized with 10% isoflurane in a mixture of 70% N 2 O and 30% O 2 . Anesthesia was maintained with 2% isoflurane. During operative procedures, body temperature was monitored continuously with a rectal probe and maintained at 37.0 6 0.28C using a homeothermic blanket control unit and a heating pad (Harvard Apparatus, Holliston, MA, USA). A midline cervical incision was made, and Instruments, Ithaca, NY, USA) was placed around midportion of the CCA. Carotid blood flow was obtained with a Transonic TS420 Blood Flow Meter (Transonic Instruments, Ithaca, NY, USA) and an iWorx IX-304T data acquisition system (iWorx Systems, Inc., Dover, NH). Computer-based analysis by iWorx Labscribe 2 software (version 2.045000) was performed to minimize any bias when assessing results. CCA baseline flow was measured for 3 minutes.

Stereological Cell Counts
The total number of cells in the substantia nigra (SN), striatum, and frontal cortex was estimated using an optical fractionator and unbiased stereology of stained cells, as previously described [17]. The sections used for counting covered the entire SN from the rostral tip of the pars compacta back to the caudal end of the pars reticulata. This generally yielded 8-9 sections in a series. Sampling was performed using the Olympus BX51 microscope (Olympus, Tokyo, Japan), and the SN was delineated at 31.25 objective. A counting frame (60%, 35 frames, 650 lm 2 ) was placed randomly on the first counting area and systematically moved though all counting areas until the entire delineated area was sampled. Actual counting was performed using a 340 oil objective lens. Guard volumes (4 lm from the top and 4-6 from the bottom of the section) were excluded from both surfaces to avoid the problem of lost cap, and only the profiles that came into focus within the counting volume (with a depth of 10 lm) were counted. The total number of stained cells was calculated according to the optical fractionator formula. To measure striatal density, images of DARPP immunostaining in the striatum were captured. The numbers of DARPP-positive cells per mm 2 were measured using ImageGauge. For quantitative measure of apoptosis in the frontal cortex, it was defined as a percentage of Annexin V-positive cells per total number of NeuN-positive cells.

Statistical Analysis
Comparisons between groups were made using Student's t tests (paired) or one-way analysis of variance followed by a Dunnett post-hoc tests. p Values less than .05 were considered statistically significant. Data were expressed as mean standard deviation. Statistical analyses were performed using commercially available software (version 10.0; SPSS, Inc., Chicago, IL).

The First Experiment
Evaluation of ICA Flow and Cerebral Ischemia Following Intra-Arterial Injection of MSCs. To evaluate ICA patency that might be influenced by intra-arterial administration of MSCs, an ultrasonic flow meter was used to measure blood flow through the ICA. The ICA blood flow was well preserved in animals receiving a 310 of the standard dose of MSCs ( Fig. 2A). In addition, to evaluate any delayed adverse effect of ICA occlusion, we performed MRI 4 weeks after MSC administration. No occlusive lesion of the ICA was observed at a 310 of the standard dose (Fig. 2B). Using T2-weighted MRI to detect ischemic brain lesions, no hyperintense signals were observed in the brain of animals receiving a 310 of the standard dose (Fig. 2C). At the 350 of the standard dose of MSCs, no animals survived more than 1 day after intraarterial MSC administration because MSCs were regurgitated from the injection site of the ICA with concomitant bleeding.
Histological and Behavioral Analysis. Immunohistochemical analysis revealed that MSA animals receiving double toxins showed a significant reduction of TH-positive and DARPP-32positive neurons in the SN and striatum compared with control animals (Fig. 2D). Compared with MSA animals receiving placebo, MSA animal receiving a 310 of the MSC standard dose in MSA animals led to a significant attenuation of TH-positive and DARPP-32-positive neurons in the SN and striatum (Fig. 2D). We did not perform immunohistochemical analysis in MSA animals receiving a 350 of the standard dose of MSCs because of subject death from intra-arterial MSC administration. On behavioral analysis, the number of rotations after apomorphine injection was significantly increased in double toxin-induced MSA animals compared with control animals (Fig. 2E). However, MSA animals receiving a 310 of the standard dose of MSCs exhibited a significant reduction in rotation behavior 1 week after MSC treatment compared with MSA animals receiving placebo (Fig. 2E).

The Second Experiment
Evaluation of ICA Flow and Cerebral Ischemia Following Intra-Arterial Injection of Different Doses of MSCs. To evaluate efficacy of intra-arterially injected MSCs, we chose 30.2, 32, or 320 of the standard dose that was individual dose below 310 of the standard dose or between 310 and 350 of the standard dose of MSCs. The ICA blood flow evaluated by an ultrasonic flow meter was well preserved in animals receiving the 30.2, 32, and 320 of the standard doses of MSCs (Fig. 3A). In addition, no occlusive lesion of the ICA was observed at any of the MSC standard doses on MR-angiography (Fig. 3B). On T2*-weighted MRI, no hyperintense signals were observed in the brain parenchymal of animals receiving a 30.2 and 32 of the standard doses. However, MSA animals receiving a 320 of the standard dose of MSCs exhibited high signal intensity in the middle cerebral artery territory (Fig.  3B) without hemorrhagic lesions.
Histological and Behavioral Analysis. Immunohistochemical analysis showed that MSC treatment with a 30.2 or 32 of the standard dose in MSA animals led to a significant increment of  TH-positive and DARPP-32-positive neurons in the SN and striatum compared with MSA animals receiving placebo (Fig. 4A, 4B). In direct comparison, the number of TH-positive and DARPP-32positive neurons was significantly higher in MSA animals receiving a 32 of the standard dose compared with those receiving a 30.2 of the standard dose (Fig. 4A, 4B). In MSA animals receiving a 320 of the standard dose of MSCs, the number of TH-positive and DARPP-32-positive neurons was significantly higher compared with those receiving placebo. However, the prosurvival effect on TH-positive and DARPP-32-positive neurons in MSA animals receiving a 320 of the standard dose was inferior to a 32 of the standard dose (Fig. 4A, 4B). In addition, MSC treatment with a 30.2, 32, or 320 of the standard dose in MSA animals decreased significantly Annexin V-positive cells in the frontal cortex compared with MSA animals receiving placebo. Anti-apoptotic effect was prominent in MSA animals receiving a 32 of the standard dose compared with those receiving a 30.2 or 320 of the standard dose (Supporting Information Fig. S1). Behavioral analysis showed that MSA animals receiving MSCs at 30.2 and 32 concentrations of the standard dose exhibited a significant reduction in rotation behavior 1 week after MSC treatment compared with MSA animals receiving placebo (Fig. 4C). In a direct comparison, there were no significant differences in rotation behavior between the 30.2 and 32 of the standard dose of MSCs at each posttreatment time (Fig. 4C). However, the number of rotations after apomorphine injection in MSA animals receiving a 320 of the standard dose did not differ significantly compared with placebotreated MSA animals (Fig. 4C).
Intracerebral Concentrations of MSCs. First, we examined how long MSCs would survive in the brain following intra-arterial injection of MSCs at a 310 standard dose. IVIS analysis showed that the distribution of MSCs in the brain gradually decreased over time, and 8 weeks after MSC injection, fluorescent signal intensity in the brain was markedly attenuated (Fig. 5A). Additionally, NuMA-positive MSCs comerged with nanoparticle were identified in the SN and striatum 8 weeks after intra-arterial administration (Fig. 5B). Next, we evaluated distribution patterns of MSCs in the brain depending on MSC dose. As expected, the distribution of MSCs in the brain 1 week after intra-arterial administration had an incremental tendency as the concentration of MSCs was increased (Fig. 5C). Finally, we used intravenous injection of MSCs to evaluate whether intracerebral concentrations of MSCs might differ depending on delivery routes. When MSCs were administrated at a 310 of the standard dose via the tail vein, fluorescent signal intensity density in the brain was even lower than intra-arterial injection of a 30.2 of the standard dose (Fig. 5C).

DISCUSSION
The present study revealed that if their concentration is appropriate, intra-arterial injection of MSCs is safe and exerts a neuroprotective effect on striatal and nigral neurons with a coincidental improvement in motor behavior. However, in addition to the fatal effect of an extremely high MSC dose, intra-arterial injection of a higher dose of MSCs led to the development of ischemic lesions and thus attenuated improvement of behavioral performance. Furthermore, IVIS analysis showed that intra-arterial injection of MSCs increases cellular density in the brain compared with intravenous injection. These data indicate that intra-arterial administration of MSCs is feasible and efficacious with respect to neuronal viability and behavioral performance in double toxininduced MSA models.
In terms of the pathogenesis of MSA, some clues to understanding the mechanisms of MSA have been suggested. Using MSA transgenic mice, studies have suggested that mitochondrial inhibitors or proteasome inhibitors might trigger or aggravate MSA pathologies [18,19]. Taken together, dysfunction in cellular clearing systems and neuroinflammation may cause further deterioration of these neurodegenerative processes [20]. Until now, many clinical trials have been performed to uncover the efficacy of various treatments in patients with MSA; however, these trials have failed to provide ample evidence of neuroprotection or had limited efficacy in MSA patients [21]. These results imply that if a certain molecular candidate is to achieve a neuroprotective effect in MSA patients, this candidate may need to fully prevent cell death cascades in every pathogenic step.
With the idea that MSCs may have pleiotropic effects against the neurodegenerative microenvironment, the present study used a double-toxin-induced animal model of MSA to examine the feasibility and efficacy of intra-arterial administration of various doses of MSCs to extend its clinical application to patients with MSA. In the present study, a concentration of MSCs that was used in previous clinical trials (4 3 10 7 MSCs) was selected as a standard concentration. We found that MSA animals only receiving 30.2 or 32 of the standard dose of MSCs exhibited behavioral improvement relative to the placebo-treated group, whereas prosurvival effects of MSCs in immunohistochemical analysis were evident in MSA animals receiving 30.2, 32, or 320 of the standard dose. In a comparative analysis among MSC treatment subgroups, behavioral analysis showed no significant difference in motor performance between MSA animals receiving 30.2 and 32 of the standard dose; however, immunohistochemical analysis demonstrated that prosurvival effects on striatonigral neurons were more prominent in MSA animals receiving the 32 of the standard dose compared with those receiving a 30.2 or 320 of the standard dose. Therefore, combining behavioral and histological analyses, the present data indicate that a 32 of the standard dose of MSCs may be the most efficacious dose of intra-arterial injection in MSA animals.
We also found that intra-arterial administration of higher doses of MSCs led to in situ occlusion of the injected artery or ischemic insults in the distal cerebral arterial circulation. Specifically, when MSA animals were injected with a 350 of the standard dose, the cells were regurgitated from the injection site in the ICA probably due to arterial occlusion, which led to higher mortality. In addition, intra-arterial injection with a 320 of the standard dose was associated with ischemic lesions in the distal MCA territory. This may explain why MSA animals receiving the 320 of the standard dose of MSCs did not show a significant improvement in behavioral performance and neuronal survival compared with those receiving a 30.2 or 32 of the standard dose. In terms of ischemic complication associated with intra-arterial injection of MSCs, cell dose, and patency of arterial flow are important factors in determining the development of ischemic insults, even though the complication rate appears to vary depending on the experimental design of intra-arterial injection [14,15,22]. Additionally, our previous clinical studies demonstrate that the frequency of ischemic insults is comparable between MSC-treated and placebo-treated groups, suggesting that ischemic lesions seem to be a consequence of the angiographic procedure rather than intra-arterial MSC injection. Accordingly, the present study Kim, Kim, Oh et al. 1431 www.StemCellsTM.com O c 2017 The Authors STEM CELLS TRANSLATIONAL MEDICINE published by provides further in vivo evidence that if arterial flow is patent during intra-arterial administration, the cellular density of MSCs around a standard dose is quite safe and free from ischemic insults in addition to providing neuroprotective efficacy. However, because we did not determine the infusion volume or infusion velocity as a contributing factor to the development of cerebral ischemia [23]. A further study would be required to resolve safety issues related with intra-arterial injection of MSCs.
IVIS analysis showed that intra-arterial injection of MSCs markedly increased migration of MSCs into the brain compared with intravenous injection. MSCs characteristically migrate toward injured brain areas in various animal models of ischemia and Parkinson's disease, possibly in response to signals that are upregulated under injury conditions [24,25]. SDF-1 is widely expressed in the brain, including the cortex, cerebellum, basal ganglia, and SN pars compacta [25]. Damage in the SN and striatum induced by 1methyl-4-phenyl-1,2,3,6-tetrahydropyridine and 3-NP may increase the expression of SDF-1 and CXCR4, leading to recruitment of MSCs to these regions. As expected, the present study showed that the distribution of intra-arterially injected MSCs in the brain had an incremental tendency as the concentration of MSCs was increased. Together, we directly compared the tendency of homing effects between intra-arterial and intravenous routes by analyzing IVIS. The number of homing MSCs after intravenous administration even at the 310 of the standard dose of MSCs was much lower than intra-arterial injection of a 30.2 of the standard dose. As these migrated cells may contribute to modulation of the microenvironmental cascade of the neurodegenerative process in the brain, the intra-arterial route would be by far superior to the intravenous route in terms of bioavailability of homing MSCs.
Finally, we found that when a 310 dose of MSCs was administrated via the intra-arterial route, the concentration of MSCs in the brain gradually decreased over time, and 8 weeks after MSC injection, the cellular signal intensity in the brain was markedly attenuated. Considering the survival time of MSCs, this may raise the possibility of the need for additional MSC administrations to bolster homing MSCs. Similarly, our previous clinical trial demonstrates that the neuroprotective effect of MSCs tend to fade away around day 240 after the first intra-arterial MSC injection, implying repeated administration of MSCs may be needed to maintain their neuroprotective properties [10]. However, further studies determining the most appropriate time interval of repetitive MSC administrations are required for clinical application in patients with MSA.

CONCLUSION
Our study has provided important details that intra-arterial administration of MSCs is feasible and efficient administration route, resulting in improvement of neuronal viability and behavioral performance in double toxin-induced MSA models. The present data may advance clinical applications of intra-arterial route of MSC administration in patients with MSA.