miR-6780-5p-Enriched Exosomes Derived From Butylidenephthalide-Pre-Conditioned Human Olfactory Ensheathing Cells Via Autophagy Improve Motor Coordination and Balance in a SCA3/MJD Mouse Model

27 28 Background: The development of acellular products is a new trend for regeneration 29 medicine. To provide an acellular product exhibiting characteristics of cells and usefulness 30 as a therapeutic agent, exosomes were employed in the current studies. Method and Result: 31 The therapeutic agent hsa-miRNA-6780-5p was enriched up to 98 folds in exosomes derived 32 from butylidenephthalide (bdph)-pre-conditioned human olfactory ensheathing cells (hOECs) 33 compared to naïve hOECs exosomes. The particle size of exosomes derived hOECs and 34 exosomes derived hOECs pre-conditioned bdph were around 124.17 nm and 117.47 nm, 35 respectively. The role of hsa-miRNA-6780-5p was first demonstrated in our studies using a 36 liposome system, showing that it enhances autophagy and inhibits spinocerebellar ataxia type 37 3 (SCA3) disease proteins of polyglutamine (polyQ) tract expression. At the same time, the 38 exosomes with enriched hsa-miRNA-6780-5p were further applied to HEK-293-84Q, thus 39 resulting in decreased expressions of polyQ and increased autophagy in the cells. In contrast, 40 the results were reversed when the autophagy inhibitor, 3MA, was added to the cells treated 41 with hsa-miRNA-6780-5p enriched exosomes, indicating that the decreased polyQ 42 expression was modulated via autophagy. The SCA3 mice showed improved motor 43 coordination behavior when they received intracranially injected exosomes enriched with 44 hsa-miRNA-6780-5p. The SCA3 mouse cerebellum tissue having received hsa-miRNA- 45 6780-5p enriched exosomes also showed a decreased expression of polyQ and increased expression of autophagy marker. Conclusions: Together, our findings provide an alternative 47 therapeutic strategy for SCA3 disease treatment, using miRNA enriched exosomes derived 48 from chemically pre-conditioned cells. 49 These indicate that the exosome of BPPexo in

autosomal dominant cerebellar ataxia. The disease is rare, with prevalence around 1-56 5/100,000 individuals. Mainland China has the highest prevalence of SCA3 with 62.6%, 57 followed by Brazil with 59.6%, Japan with 43%, and Germany with 42% [1]. The highest 58 prevalence of spinocerebellar ataxia (SCA) in Taiwan is type 3 with the level of 47.3% [2]. 59 In SCA3, an expanded CAG repeat by more than 45 is found in the exon 10 of ATXN3, which 60 is translated into a disease mutant ataxin-3 with expanded polyglutamine (polyQ) tract. A 61 negative correlation between expanded polyQ and disease onset has been demonstrated [3]. of studies indicating that exosomes are an acellular product which may reduce tumorigenesis 87 and display high potential for regeneration [11][12][13] Enhancement of autophagy may improve the disease progress. Beclin-1, an initiator of 103 autophagy involved in an mTOR-dependent pathway, present lower expression in the 104 patient's fibroblast [25,26]. In addition, the ratio of LC3II/ LC3I is also lower in patients 105 [25]. The chemical agent Butylidenephthalide (Bdph) has been demonstrated to induce 106 autophagy in cells [27]. Therefore, a new strategy using Bdph to trigger glial cell production 107 of exosomes stimulating autophagy was investigated in the current study. And moreover, 108 Bdph-magnified therapeutic hsa-miRNA in exosomes were as a delivering therapeutic agent 109 for SCA3 treatment. 110 111

Exosome characterization 113
To confirm exosome production quality, the total protein and particle size were analyzed for 114 each batch. We performed three times the same procedure for exosome production in the 115 hOECs, with or without Bdph treatment. The average concentration of exosomes derived 116 from hOECs without Bdph (OECexo) or with Bdph (BPPexo) treatment were 9.03 and 8.78 117 µg/ml using the BCA kit analysis. The average size of exosome from the hOECs without 118 Bdph and with Bdph treatment performed were 124.17 ± 11.91 nm and 117.47 ± 10.09 nm 119 by NTA measurement, respectively. (Figure 1a). The two exosome populations were well-120 distributed as shown in Figure 1b and c. 121 122 Bdph increases the autophagy element production in exosomes and their treatment in 123

HEK-293-84 Q cells 124
To confirm whether autophagy markers can be secreted from hOECs cells pre-conditioned 125 with Bdph, the expression of two markers, beclin1 and LC3I/LC3II, was analyzed. The total 126 proteins served as an internal control for each group of exosomes. The exosomes derived 127 from the glial cells, hOEC, treated with Bdph (BPPexo) had increased expression of the 128 autophagy markers, Beclin1 and LC3I/LC3II, compared with exosomes derived from naïve 129 hOECs without Bdph (OECexo) (Figure 2a). These results indicate that the exosome of BPPexo play the role in autophagy pathway and 143 thus BPPexo did not have any effect on HEK-293-84Q cells when the autophagy pathway 144 was inhibited. These findings imply that the exosomes derived from Bdph treated hOECs 145 were able to encapsulate a therapeutic agent related to autophagy. 146 The miRNA microarray of hOEC cells treated with Bdph or without Bdph was analyzed to 147 screen for candidate miRNA. The changes above three-fold intensity were included in the 148 panel ( Figure 2c). In addition, the panel of miRNA in the Figure 2c was further narrowed to 149 key candidate miRNAs related to the mTOR pathway. According to the online tool of Introducing an autophagy enhancer, hsa-miRNA 6780-5P, via a liposome system into 163

HEK-293-84Q cells 164
Before screening the potential miRNA candidates for polyQ decrease and autophagy, we 165 used a homemade liposome system to confirm miRNA delivery efficiency. The miRNA-Fan 166 acted as model miRNA based on its fluorescence, with excitation at 493 nm and emission at 167 517 nm, reporting a successful miRNA delivery with fluorescence in the cells. The liposomal 168 system has DiL dye fluorescence with red color upon successful delivery in the cells. The 169 results showed that the cells presented with red color fluorescence after treatment for one 170 hour, meaning that the liposomes were uptake by the cells (Figure 3a). At the same time, the 171 cells also showed green color fluorescence, meaning that the miRNA-Fan were in the cells 172 Bdph were applied into SCA3 mice to confirm the efficacy of exosome treatment. The 208 exosomes from hOECs with (BPPexo) or without Bdph (OEC exo) priming were injected 209 into the SCA3 animal. The low and high dose of exosomes derived from hOECs with Bdph 210 priming were also examined in the study. The behavior of the animals on the rotarod was 211 assessed to evaluate the balance and coordination after treatment. To determine the treatment 212 efficacy, untreated SCA3 and wild-type animals were compared in the study. We found that 213 the latency to fall in the wild-type mice was around 150 seconds in the rotarod test, whereas 214 it was below 100 seconds in untreated SCA3 mice. The results further showed that SCA3 215 animals receiving a high dose of BPPexo stayed longer on the accelerating rotating rod for 216 an average above 120 seconds compared to below 100 seconds for the untreated group 217 ( Figure 5a). The group receiving a low dose of BPPexo showed they can stay above 120 218 seconds on the accelerating rotating rod in the first three weeks, which was then decreased 219 after three weeks of injection (Figure 5a). The SCA3 animals received OEC exosomes (OEC 220 exo) displayed an increased time on the accelerating rotating rod in the first two weeks of 221 treatment, with 122 seconds on the rod, but their latency to fall was decreased like the 222 untreated group (Figure 5a). The results indicate that both exosomes were effective, with the 223 higher dose of BPPexo resulting in a longer period of therapy. The same dose of two 224 exosomes resulted in different efficacy, suggesting different therapeutic mechanisms for the 225 two exosomes. Lastly, we further to confirmed the expression of polyQ and autophagy using 226 brain tissue. Brain tissue was collected from the animals after 6 weeks of treatment, at the 227 age of 20 weeks, to confirm the expression of polyQ and autophagy. The wild-type animals 228 showed less extended polyQ expression at 63 kDa, whereas the untreated SCA3 mice showed 229 overexpression of polyQ ( Figure 5b). The expression of polyQ at 63 kDa was decreased when 230 the SCA3 mice received a high dose of BPPexo compared with to low dose of BPPexo or 231 high dose of OEC (Figure 5b). To confirm whether the treatment efficacy resulted from the 232 autophagy in mice, the autophagy marker, LC3B, was analyzed from the brain tissue. The 233 results showed that wild-type animals had more LC3B expression compared to the untreated 234 SCA3 group (Figure 5c). The expression of LC3B following the high dose of BPPexo showed 235 higher expression in the brain than those of untreated group and slightly higher that of 236 OECexo ( Figure 5c). Therefore, these results reveal that BPPexo has a potential in treating 237 SCA3 disease via a therapeutic mechanism involving autophagy. 238 239

Discussion 240
The strategy of using miRNAs, such as miR-25 [28], miR-9, miR-181a, and miR-494 mimics 241 [29] for SCA3 therapy has been widely studied. These miRNAs were shown to directly target the mRNA of ATXN3. However, these strategies also result in inhibition of normal ATXN3, 243 indicating that some patient with heterozygous ATXN3-SCA3 may be not suitable to receive 244 this treatment strategy. The downregulated autophagy has been measured in SCA3 patients 245 [25,26,30]. According to previous studies on current therapeutic targets for SCA3 therapy, 246 the destination of those strategies and related mechanisms for SCA3 treatment would go to 247 autophagy enhancement [31]. 248 The miR-6780-5p plays a role in promoting ovarian cancer metastasis in the previously 249 studies [32]. The role of miR-6780-5p in autophagy has not been discussed yet. We found 250 that miR-6780-5p inhibits the expression of PIK3R5 and increases the expression of LC3I, 251 LC3II, and Beclin1 ( Figure 4). The PI3KR5 inhibitor has been used in previous studies as an 252 enhancer of autophagy. For example, the miR-210-5p, PI3KR5 inhibitor has been 253 demonstrated to activate oncogenic autophagy in osteosarcoma cells [33]. Therefore, using 254 miRNA related to AKT/mTOR signaling pathway provides an alternative strategy for 255 autophagy enhancement. 256 We found that exosomes carry autophagy induction elements, in which the exosomes were 257 derived from hOECs pretreated with Bdph ( Figure 2). The hsa-miR-6780-5p, a PIK3R5 therapeutic mechanism is different. Hsieh and colleagues found increased levels of 276 tryptophan hydroxylase 2 and ryanodine receptor (RYR) in the brain after hOECs treatment 277 [9]. In our studies, we found elevated levels of autophagy markers in the SCA3 mice after 278 receiving exosomes derived from Bdph primed hOECs. Therefore, exosomes are considered 279 as potential drug carriers, hence providing another strategy using drugs to enhance 280 therapeutic agent expression and act as a biomarker to evaluate the disease treatment efficacy. 281 282

Conclusion 283
In conclusion, the proposed therapeutic mechanism for the treatment of SCA3 via exosomes 284 derived pre-condition medium of hOECs is summarized in Figure 6. The has-miR-6780-5p 285 was firstly demonstrated when evaluating autophagy in the disease cell model HEK-293-84Q 286 and showing efficacy in decreasing polyQ expression in our study. We also found that Bdph, 287 an autophagy enhancement drug, elevates the production of has-miR-6780-5p in exosomes. 288 Most importantly, the elevated has-miR-6780 in the exosomes was further employed to treat 289 SCA3 mice. The results showed an improved motor coordination in the SCA3 mice received 290 exosomes with elevated has-miR-6780 compared to naïve exosomes. GIBCO). The exosome production started with the cell seeding density reached 1.2 × 10 4 /cm 2 300 followed by replacement of the Hycolne FBS with a commercial exosome depletion FBS 301 medium (GIBCO). For priming studies, 10 and 50 μg /ml Bdph were mixed with medium 302 after cell seeded for 24 hours. The conditioned medium was harvested for immediate EV 303 purification after another 48 hours in culture. The harvested condition medium was filtered 304 with 0.2 μm filter (regenerated cellulose) to remove cell debris followed by a 10 kDa 305 ultrafiltration membrane to concentrate the medium. The exosome purification kit-306 ExoQuick TM was used by adding 1/5 volume of concentrated conditioned medium. Finally, 307 the exosomes were resuspended in PBS and stored at −80 °C for further analysis. 308 309

Liposome preparation 310
The cholesterol and lipid of DPPC, TAP18, DSPCPEG2000, or dye of DiL were dissolved in 311 chloroform (10 mg/ml) in a round bottom flask. The flask was then connected to a rotary 312 evaporator (N-1300, EYELA) and maintained with temperature at 25 °C using a water bath. 313 When the chloroform was removed, a homogeneous lipid film was formed on the walls of 314 the flask. The residual chloroform was then to evaporate using 5 × 10 −4 Ttorr vacuum system 315 for at least one hour. The lipid film was then hydrated with PBS by sonication in a water bath 316 at 24 °C. The liposome dispersion was then extruded by a Mini-Extruder (Whatman, Inc., 317 Clifton, NJ, USA) using 0.2 μm polycarbonate filters at room temperature. The extruded 318 liposomes were kept at 4 °C for at least one hour and then stored at −20 °C for further 319 experiments. 320 321

Microarray miRNA gene profile from hOEC cells 322
To obtain miRNA of cells with or without Bdph treatment, the total Exosome RNA & Protein 323 Isolation Kit (Thermo Fisher Scientific Inc.) was used. Briefly, the cell lysate was added one 324 volume of 2X denaturing solution and mixed thoroughly. One volume of acid-phenol: 325 chloroform was added to each sample followed by a glass-fiber filter to obtain the total 326 miRNA. The analysis of miRNA gene profile was contracted to GeneDireX Inc. 327 328

Characterization of exosome 353
The purified exosomes were precipitated by ExoQuick TM to obtain the exosome pellet and 354 then resuspended with RIPA lysis buffer. The concentration of protein was assessed using the 355 bicinchoninic acid (BCA) Protein Assay Kit (Merck-Millipore). The 10 μg of protein was 356 loaded into 12% TGX FastCast Acrylamide Solutions (Bio-Rad Laboratories, Inc.). The TGX 357 FastCast gels were imaged with Fusion Solo S (Vilber) and then transferred to a PVDF 358 membrane. The membranes were blocked with 5% low-fat milk powder suspended in 0.05% 359 tween 20 in PBS for one hour at room temperature. Membranes were then probed for 360 exosome proteins with LC3B and BECN1. These primary antibodies were subsequently 361 probed with appropriate secondary antibodies conjugated to horseradish peroxidase. 362 Enhanced chemiluminescent (ECL) HRP substrate was added and chemiluminescence was 363 detected using the Fusion Solo S image analyzer (Vilber). 364 365

Nanoparticle tracking analysis (NTA) 366
Nanoparticle tracking analysis was used to determine the size and particle concentration of 367 exosomes or liposomes using a Nanosight LM10 (Malvern Instruments; Malvern, United 368 Kingdom). The exosomes or liposomes were diluted to 1-2 μg /mL or 10-20 ng/mL in PBS, 369 and 3 videos of 30 s were acquired for each triplicate. The HEK 293-84Q-GFP cells were seeded to 1 × 10 6 /well into 6 well cm dish for overnight 395 cultivation. The exosome (30 μg) or liposome/miRNA (above described method) was added 396 to the cells within the medium. The exosome or liposome treated cells were analyzed via 397 microscopy, Western blot or RT-PCR. 398 399

Statistical analysis 484
The data were analyzed using two tailed student's t-test. A statistically significant difference 485 was reported when the p value was below 0.05. Data are presented as mean ± standard 486 deviation (SD).