Effect of TO901317 on promoting the differentiation of human bone marrow mesenchymal stem cells into dopamine neurons on Parkinson's disease


 Background: Stem cells from different sources could differentiate into dopamine-producing cells and ameliorate behavioral deficits in Parkinson's disease (PD) models. Especially, human bone marrow mesenchymal stem cells (hBMSCs) have many advantages without ethical dispute. Liver X receptors (LXRs) are involved in the maintenance of the normal function of the central nervous system myelin. The previous work of our team have reported the induction of cocktail-induced da phenotypes from adult rat BMSCs by using sonic hedgehog (SHH), fibroblast growth factor 8 (FGF8), basic fibroblast growth factor (bFGF) and TO901317 (agonist of LXRs) with 87.42% of efficiency in 6 days of period of induction. But it did not verify whether the induced cells had the corresponding neural function. Methods: The present study was to investigate the effect of TO901317 on the differentiation of hBMSCs into dopaminergic (DA) neurons. In vitro, the cells were divided into control group, GF group and LXR+GF group. Neuronal markers (Tuj1, Neun and Nestin) and DA neuron markers (tyrosine hydroxylase, TH) were detected to explore the optimal concentration, optimal addition time and optimal induction time of TO901317 to promote hBMSCs differentiation. Determine induced cells’ DA neuron properties by detecting dopamine release. To clarify its underlying mechanisms, the expressions of LXRa, LXRb and ABCA1 were detected. In vivo, male Sprague Dawley (SD) rats were divided into control group, 6-OHDA group (model group) and 6-OHDA+Cells group (cell transplantation group) to observe whether cell transplantation had a therapeutic effect on PD. Results: TO901317 significantly enhanced the differentiation of hBMSCs into DA neurons. Compared to control group and GF group, only the LXR+GF group released dopamine by the result of enzyme linked immunosorbent assay (ELISA). Compared with the control group and GF group, the optimal time for differentiation of hBMSCs treated by 0.5mM TO901317 combined with GF was six days in LXR+GF group. And the maximum induction efficiency of differentiation of hBMSCs into DA neurons was 91.67% in LXR+GF group. After the long-term use of TO901317, LXRα and LXRβ decreased significantly, and ABCA1mRNA increased significantly in LXR+GF group. After induced-cells were transplanted into PD rats, Compared with 6-OHDA group, the expression of TH in the striatum increased significantly, and the behavior of PD rats induced by apomorphine was significantly improved in the 6-OHDA+Cells group.Conclusions: TO901317 promoted differentiation of hBMSCs into DA neurons may be related to activation of LXR-ABCA1 signaling pathway. These data suggested that TO901317 may serve as a potential therapeutic methods for PD.


Introduction
Parkinson's disease (PD) is a complex, age-related neurodegenerative disease with early prominent death and loss of dopaminergic (DA) neurons in substantia nigra (SN) 1,2 . The main clinical manifestations of PD are static tremor, increased muscle tone and unstable posture. PD is also closely related to various non-motor symptoms, such as cognitive dysfunction, mood and psychotic disorder 3 . A considerable number of patients have cognitive impairment in late stage of PD. Non-motor symptoms might be present up to 20 years before manifestation of the characteristic motor symptoms 4 .
PD is associated with cholinergic and monoaminergic neurons, and oligodendrocytes play an important role in PD 5 . Traumatic brain injury, post-traumatic stress disorder 6 , history of melanoma and exposure to pesticides are increased risk of PD 7 . In the degenerative diseases of the nervous system, the incidence of PD in the 65-year-old population is more than 1%, second only to Alzheimer's disease, and it escalates burden on economic terms and quality of life to these patients' families and society.
The main managements of PD are drug therapy, surgical treatment and stem cell replacement therapy. Drug therapy mainly improves the symptoms of patients by increasing the concentration of dopamine in the brain. But it is important to consider the side-effects and tolerance levels of the patient 8 . Surgical treatments include thalamicectomy and deep brain stimulation (DBS). Thalamotomy is an intrusive technique. Although it can relieve the tremor of PD patients well, it seems to alter the physiological regulation of PD patients and cause motor dysfunction 9 . When PD patients' symptoms do not respond to medication adjustments, DBS treatment needs to be started. It can signi cantly control dyskinesia induced by levodopa in the treatment of PD. However, for this invasive treatment, PD patients must experience surgery again and again. It increases the risk of infection of them 10 . Currently available drugs or surgery are merely symptomatic treatments and do not slow down or prevent the progress of the disease. Some data suggested that supplementing brain-lost DA neurons by cell transplantation may be the most promising therapy for PD 11 . A more immediate and reachable goal of cell transplantation may be neuronal protection 12 . Currently, embryonic stem cells (ESCs), neural stem cells (NSCs), induced pluripotent stem cells (iPSCs) and bone marrow mesenchymal stem cells (BMSCs) are available for stem cell replacement therapy 13 . The use of ESCs and NSCs existed ethical problems inherent. So much hope is transferred to iPSCs which induced human broblasts into a source of patient-speci c and diseasespeci c neurons, especially as, in theory, this approach would avoid many of the ethical issues associated with using ESCs 14,15 . Cell replacement therapy is derived from the patient's own cells without immune rejection. In autologous transplantation, it is necessary to establish iPSCs from each patient, and current technical operations take much time and high cost, so it is di cult to spread to general treatment.
In addition, the iPSCs own the patient's own genetic factors, the sensitivity of the disease may be high. It also has high risk of introducing cancer cells 16 . Speci cally, iPSCs often obtained chromosomal abnormalities, with gains or losses of whole chromosomes 17 . Different from these stem cells, BMSCs come from patients themselves without ethics dispute. Especially, BMSCs have multi-directional differentiation potential 18 , low risk of tumorigenesis 19 and rich in source. Many studies in vitro and preclinical strongly proved the therapeutic potential of BMSCs when be applied as a treatment for different pathological conditions 20,21 .
There have been different methods involved in the differentiation of BMSCs into DA neurons, including cell growth factors 22,23 , chemical inducer 24 and lentiviral transduction 25 . Currently, in vitro induction of stem cells using growth factors with sonic hedgehog (SHH) and broblast growth factors (FGFs), succeeded in inducing adult human BMSCs into DA neurons with 67% of e ciency in 12 days 26,27 . This is the maximum induction e ciency and the shortest induction time.
Liver X receptors (LXRs) include LXRα and LXRβ. It are the member of the nuclear receptor supergene family of ligand-activated transcription factors and is a major regulator of lipid metabolism 28 . It play a key role in the regulation of cholesterol and fatty acid homeostasis 29 . LXRs is also essential for central nervous system (CNS) 30,31 . The loss of LXRβ in mice affected the formation of progenitor cells and granule cell differentiation, and then leading to hypoplasia of the dentate gyrus 32 . It has been found that LXRs played crucial roles in regulation of genes related to CSF production and structural integrity of choroid plexus 33 . LXRα and LXRβ are involved in the processes of myelination and remyelination 34 .
Activation of LXRα and LXRβ can promote the regeneration and survival of motor neurons 35 . Meanwhile, LXR agonists activate LXR target genes, plays therapeutic role in different neurodegeneration animal models [36][37][38] . Therefore, we speculate whether LXR agonists have a therapeutic effect on PD and whether they can promote the formation of DA neurons.
Our previous work have found the induction of cocktail-induced da phenotypes in adult rat BMSCs by using SHH, broblast growth factor 8 (FGF8), basic broblast growth factor (bFGF) and TO901317 with 87.42% of e ciency in 6 days of period of induction 39 . LXR agonists signi cantly shortened induction time and improved induction e ciency compared to methods which had been reported. But these previous work of our team did not investigate whether induced-cells released dopamine. We did not sure if the induced cells have a therapeutic effect on Parkinson's disease.
In this study we investigated the effect of TO901317 on the differentiation of hBMSCs into DA neurons. We also explored whether induced-cells had dopamine neuronal function and the possible mechanism. Finally, we transplanted the induced-cells into PD rats to observe the therapeutic effect. Parkinson's disease rats All male SD rats were weighted 220g-250g. To establish the rat model of PD, All SD rats received intraperitonal injection of apomorphine (0.5mg/kg). Rats were selected without rotation behavior to receive a 6-OHDA lesion of the medial forebrain bundle on the right side (MFB, AP: 1.8mm, ML: 2.0 mm, DV: 8.0/7.8mm). 30 male rats received a 6-OHDA lesion of the MFB on the right side and 10 rats were as control group only injected with solvent which was used to dissolve 6-OHDA 40,41 . Brie y, rats were anesthetized with chloral hydrate (4% chloral hydrate and 96% saline solution, 1ml/100g) by intraperitoneal injection. The MFB was targeted with an injection of 4μl 0.02% L-ascorbic acid saline solution containing a total amount of 16μg of 6-OHDA. The lesion to stereotaxic coordinates were adjusted to the age and weight of the animals with the help of brain stereotaxic instrument (RWD, Shenzhen, China). After two weeks to one month, to determine whether the models were successful, we used apomorphine to induced rotation and behaviour was recorded over a period of 30 min. Judging that the model is successful is that the number of rotations to the healthy side is more than 7 rotations per minute.

Differentiation of hBMSCs
Human BMSCs were divided into 3 groups and plated onto 24 well plates where each plate contained 2.0×10 4 cells and cells were cultured at 37°C with 5% CO2.
Explore the best time to add TO901317 during growth factor induction period: On the basis of GF, TO901317 were added on the rst day, the third day, the sixth day, and the ninth day, respectively, to induce differentiation for 12 days.
Explore the induction time of the TO901317 in combination with GF: According to the result of (a) (b), 0.5 μM TO901317 was added to the culture medium, and the cell morphology was observed every 3 days.
Cell Counting kit-8 assay A Cell Counting Kit-8 (CCK-8; Dojindo, Japan) assay was used to test the growth rate by following the manufacturer's instructions. 3000 cells/well of hBMSCs were cultured into a 96-well plate in every group. At 3-day intervals, the growth rates of cells in each group were measured by application of CCK-8 kit and optical density (OD) was determined at 450 nm using a microplate reader (Thermo Scienti c, USA).

Immuno uorescence
In brief, cells of each group were xed in 4% paraformaldehyde and permeabilized with 0.3% Triton-X100 and blocked in phosphate buffer solution (PBS) containing 5% normal goat serum (Beijing Dingguo Changsheng Biotechnology, China). Cells were incubated with monoclonal antibodies overnight with 4℃. Antibodies' dilutions were as follows: β III tubulin Enzyme-Linked Immunosorbent Assay (ELISA) To investigate whether cells in each group release dopamine. Culture supernatant and cells of control group, GF group and LXR+GF group were collected. Cells were added with PBS, then grinded on ice. The mixture was centrifuged at 3000g for 20min at 4°C and collected the supernatants. Dopamine was detected by ELISA kits (n=6) (Mei biao, Jiangsu, China). The samples and standards were tested according to the instructions.

Western Blotting Test
Cells were plated on six-well plate for 24 hours with 10% FBS, then grown in induction medium for six days to be used for preparation of whole cell extract. After removing the media, cells were washed three times with PBS. Then cells in each well were added 150ml lysis buffer containing 1% phenylmethanesulfonyl uoride (PMSF) and cracked on ice for 30 minutes. The mixture was centrifuged at 12,000 g for 20 min at 4°C and collected the supernatants. The protein concentration was determined by BCA Protein Assay Kit (Beyotime, China). The protein was separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membranes (Millipore, United States). The membranes were blocked with 5% bovine serum albumin (BSA) for 2 h at room temperature and then incubated with speci c primar antibodies, TH, 1:500 (Abcam, Cambridge, UK); LXR α receptor, 1:500 (Abcam, Cambridge, UK); LXR β receptor, 1:500 (GeneTex, United States) were included and overnight at 4℃. The membranes were rinsed three times in TBST and incubated with HRP conjugated secondary antibodies at room temperature for 1h. Then after washed three times in TBST, protein signals were visualized by ECL (Bio-Rad, United States).

Real-Time Polymerase Chain Reaction (qPCR)
Total RNA was isolated from the cells in control group, GF group and LXR+GF group by Trizol reagent (Vazyme, Nanjing, China) according to the manufacturer's protocol. Then mRNA was subjected to reverse transcription using HiScript Q Select RT SuperMix (Vazyme, Nanjing, China). SYBR Green II (Biomake, USA) incorporation method was used to detect the amount of mRNA. Negative controls were used as no template cDNA reactions and melting curves were used to con rm the results. The results were normalized using GAPDH concentration of each sample. The primer sequences are reported in Table 1.

Histopathological Examination
Hematoxylin and eosin (HE) staining was performed to show pathological histological damage in the substantia nigra pars compacta (SNc) and striatum. The rats of control group, 6-OHDA group and 6-OHDA+Cells group were anesthetized with sodium chloral hydrate (4%, 1ml/100g) and perfused with PBS, and then perfusion with 4% paraformaldehyde. After that, the brains were dehydrated in a graded series of alcohols and embedded in para n. A series of 5-μm-thick sections were cut from the brain.
Finally, the sections were stained with HE reagents for histopathological examination.

Statistical analysis
All results were expressed as the means ± standard deviation (SD) and the statistical signi cance of differences was analyzed by GraphPad Prism (GraphPad Software, La Jolla, CA, USA). For the comparison of multiple groups, statistical analysis was performed using one-way analysis of variance (ANOVA) with Bonferroni's posthoc test. Probability values less than 0.05 (p < 0.05) were considered to be statistically signi cant.

Result
Morphological features of hBMSCs hBMSCs were purchased in Cyagen Biosciences. Cells exhibited a long fusiform and vortex arrangement (FIGURE 1).

Effect of different concentrations of TO901317 on cell survival
CCK8 kit was used to detect the survival rate of cells in control, GF and LXR+GF groups. In addition, the concentration of TO901317 in the LXR+GF group was given to 0.125mM, 0.25mM, 0.5mM, 1mM and 2mM respectively. After three days, the survival rate of cells in each group did not show signi cant difference (FIGURE 2A). The survival rate signi cantly increased in GF group and LXR+GF group compared to control group when cells were cultured for six days, nine days and twelve days. Interestingly, there was no signi cant difference in cell survival rate when cells were cultured with different concentrations of TO901317 (FIGURE 2).
The optimal concentration of TO901317 Immuno uorescence was used to detect the expression of neuronal markers (Neun, Nestin and Tuj1) and dopamine neuron markers (TH).
When GF combined with 0.5mM TO901317, the number of TH + positive cells reached the maximum The best time to add TO901317 into GF In this study, 0.5mM TO901317 was used in combination with GF to induce hBMSCs into DA neurons. On the premise of GF addition during 12 days, TO901317 was added on the rst day, third day, sixth day, and ninth day. Expression of TH (Cy3, red) and Tuj1 (Alexa Fluor 488, green) were determined by immuno uorescence. The results showed that the expression of TH was the highest when TO901317 was added on the rst day (FIGURE 4).
The optimal period induced by TO901317 Immuno uorescence was used to detect the expressions of Tuj1 and TH after three days, six days, nine days and twelve days of induction period. Cells in control group only expressed Tuj1 without TH. While cells in GF group and LXR+GF group expressed Tuj1 and TH simultaneously (FIGURE 5A). In the LXR+GF group, TH + cells reached maximum after six days of induction (FIGURE 5B). Compared with control group and GF groups, the expression of TH was signi cantly increased in cells of LXR+GF group (FIGURE 5D).
Cells showed typical neuronal morphology with extended long cell processes and enlarged cell bodies in GF group and LXR group under the bright eld (FIGURE 5C). In particular, in the LXR+GF group, the morphology of the neurons expressed by the cells was more obvious (FIGURE 5E).

Nestin, Neun expression and dopamine release
Nestin, a neuroectodermal marker, seems to be a prerequisite for acquisition of the aptness to progress towards the neural lineage 42,43 . The cells in the control group probed with nestin and neun antibodies were negative and just revealed nuclear staining. Both GF and LXR+GF cells stained with neun and nestin (FIGURE 6A, 6B).
A dopamine kit was used to study whether differentiated hBMSCs release dopamine. The results showed that cells in control group and GF group did not secrete dopamine. Although the GF group's cells showed positive staining of neuronal markers, dopamine levels were extremely low and could not be detected in this experiment. Only the cells in the LXR + GF group had the characteristics of dopaminergic neurons, and the content of dopamine is high (FIGURE 6C).

The role of LXRs in cell differentiation and its possible mechanism
The expressions of LXRα and LXRβ did not show signi cant difference among control, GF and LXR+GF groups (FIGURE 7A, B). Result of western blot showed that expressions of LXRα and LXRβ were decreased signi cantly in LXR+GF group compared to control and GF groups (FIGURE 7C, D).
ABCA1mRNA in LXR+GF group was signi cantly elevated when compared to control and GF groups (FIGURE 7E).

Establishment of Parkinson's disease model
Compared with the control group, the number of contralateral rotations per min was scored. Rats with stable lesions (>7 rpm/min) were selected as PD models. The expression of TH in the model group decreased signi cantly. The results suggested that the PD model was successfully established (FIGURE 8).

Effect of cell transplantation on PD rats
Four weeks after injection of 6-OHDA, the apomorphine-induced contralateral rotation test was performed. the 6-OHDA+Cells group already appeared signi cantly improved behavioral performance compared to the rats of the 6-OHDA group (FIGURE 9F). Compared with control group, cells in 6-OHDA group presented signi cant nuclear pyknosis, vacuolization and nuclear deep staining. After cell transplantation, the nuclear deep staining of the nucleus and the vacuolization of cells in 6-OHDA+Cells were reduced (FIGURE 9A). Results revealed that TH-positive signals were almost absent at the striatum in the 6-OHDA-lesioned rats that received no grafts. In contrast, the rats in the 6-OHDA+Cells group, THpositive signals were greatly recovered in the striatum (FIGURE 9C). TH-positive signals also signi cantly increased at SNc in 6-OHDA+Cells group compared to 6-OHDA group (FIGURE 9B). Western blot detection also showed the same results (FIGURE 9D, E).

Discussion
PD is a complex neurodegenerative disease which is characterized by motor dysfunctions, and it also accompany non-motor symptoms. These symptoms are tightly associated with the loss and death of DA neurons. In the current research, cell transplantation has become an important topic, and it has very important signi cance in clinical application. Neural stem cells, embryonic stem cells, induced pluripotent stem cells and bone marrow mesenchymal stem cells have been used to the treatment of PD. Neural stem cells and embryonic stem cells exist a serious ethical dispute. As for induced pluripotent stem cells, there is a possibility of carcinogenesis. Therefore, more and more researches have been taken on BMSCs.
Some studies have shown that BMSCs expressed genes and proteins related with the neural lineage, and have been displayed to hold neurogenic differentiation potential the proper conditions in vitro [44][45][46] . BMSCs transplantation restrained multiple parameters of spinal neuroin ammation found in diabetic mice 47 . BMSCs can release neurotrophic factors, including GDNF and BDNF to protect neurons 48,49 .
In the reported studies, many methods were adopted to drive BMSCs to differentiate into DA neurons. And induction time is generally 12 days or more 50 . The vital method was to induce directional differentiation of cells by using cytokines. Astrocyte-derived bFGF is required for regulation of DA differentiation of the stem cells and promotes growth and survival of midbrain DA cells [51][52][53][54] . SHH participates in a broad array of neurodevelopmental processes in the vertebrate embryo, including morphogenesis, cell proliferation and speci cation, and axon path nding 55,56 . SHH exists in the postnatal and adult CNS, such as modulation of neuronal activity, progenitor cells and astrocytes as well as in differentiated neurons [57][58][59] .
FGF8 is essential for the development of multiple brain regions, such as suprachiasmatic nuclei (SCN) and hypothalamo-pituitary 60,61 . The development and survive of DA neurons is associated with GF-SHH and FGF8 62,63 . ABCA1 affects cognitive function, leading to amyloid-beta (Aβ) production and apoptosis and promote neurorestoration 64-66 . In the present study, our aim was to explore the effect of TO901317 on the differentiation of hBMSCs into DA neurons. We found that the growth rate of cells in GF and LXR+GF groups was initially increased and later decreased with the extension of induction time compared to control group. But the growth rate of cells in LXR+GF group was not different with GF group. We hypothesized that TO901317 had little or no effect on cell proliferation, and TO901317 may have a special effect on cell differentiation. Some studies have explored the effect of GF on inducing BMSCs differentiation into DA neurons, but the result were designated DA neuronal progenitors without neuronal function 67,68 . Our results showed that hBMSCs treated with GF alone or TO901317 combination with GF leaded to directed neuronal differentiation. TO901317 could promote differentiation of hBMSCs into DA neurons in cooperation with GF. The maximum induction e ciency was low, and the shortest induction time was 12 days when hBMSCs treated with GF alone. While the maximum induction e ciency was 91%, and the shortest induction time was 6 days when hBMSCs treated with 0.5μM TO901317 in cooperation with GF. This greatly improved the e ciency of induction. Cells in GF group and LXR+GF group expressed neun and nestin. The lack of ABCA1 leads to transport disorders of central nervous system cholesterol, which in turn leads to defects in neuronal structure and function 72 . Our study found that ABCA1 mRNA signi cantly elevated and LXRs obviously decreased in LXR+GF group compared to GF group. Differentiation of hBMSCs into DA neurons may be related to LXR-ABCA1 pathway.
The results of the studies showed that short-term xenotransplantation has less immune rejection 73,74 .
Moreover, BMSCs have immunosuppressive effects in vivo and vitro 75, 76 . The damage caused by 6-OHDA could be reduced after induced-cells transplantation in PD rats. Compared with the 6-OHDA group, the DA neurons in the 6-OHDA+Cells group had a signi cant increase. The results indicated that TO901317 can promote the differentiation of hBMSCs into DA neurons by activating the LXRs-ABCA1 signal pathway.

Conclusions
Collectively, based on GF to culture cells, 0.5μM TO901317 could promote the differentiation of hBMSCs into DA neurons. Induced-cells had therapeutic effect of on PD. The mechanism of TO901317 promoting the differentiation of hBMSCs into DA neurons may be related to the activation of the LXR-ABCA1  Figure 1 hBMSCs were observed at inverted microscope and showed a broblastoid cell pro le and produced cell colonies with unique vortex shape.