目的

（1）通过系统比较并分析实验动物脂肪间充质干细胞（adipose-derived stem cells, ADSCs）与人脂肪间充质干细胞（hADSCs）的分离培养方法和生物学特性的差异，确定实验动物来源ADSCs与hADSCs生物学特性是否一致，为ADSCs移植治疗人类疾病的临床前研究提供实验数据和方法。

（2）通过研究来自不同年龄小鼠的脂肪干细胞（mADSCs）衰老表型的改变，明确机体衰老对ADSCs体外增殖及功能的影响，为后续ADSCs移植治疗帕金森病（Parkinson’s disease, PD）的供体年龄选择提供实验依据。

（3）探索脂肪间充质干细胞（adipose-derived stem cells, ADSCs）向功能性多巴胺能神经元（dopaminergic neurons, DNs）诱导分化的可能性，建立ADSCs来源DNs的分化体系，系统分析诱导分化过程中细胞基因表达和功能变化，进一步筛选并探究关键基因在ADSCs神经诱导分化中的作用。

（4）以人诱导多能干细胞（human induced pluripotent stem cells, hiPSCs）为阳性对照，比较并评价ADSCs及其诱导细胞移植治疗基于神经毒素建立的Sprague-Dawley （SD）大鼠和食蟹猴帕金森病模型的有效性，研究ADSCs及其诱导细胞的安全性，为PD治疗的细胞选择提供临床前实验基础。

方法

（1）将临床吸脂术患者、食蟹猴、C57BL/6（C57）小鼠和Sprague Dawley（SD）大鼠的皮下脂肪组织进行分离培养后获得hADSCs、食蟹猴脂肪间充质干细胞（cADSCs）、小鼠脂肪间充质干细胞（mADSCs）和大鼠脂肪间充质干细胞（rADSCs）。随后通过流式细胞术鉴定ADSCs表面标记物的表达，进行细胞成骨成脂诱导分化。

（2）分离1月龄（1M）和20月龄（20M）C57小鼠的脂肪组织并培养获得mADSCs。研究其细胞形态、超微结构、表面标记物、增殖、成脂成骨分化能力、细胞因子分泌能力、衰老相关基因表达水平、凋亡细胞占有比例、细胞周期、衰老相关-β-半乳糖苷酶（senescence-associated-β-galactosidase, SA-β-gal）染色和转录功能。

（3）采用N1、N2、N3三个改进的常规诱导培养方法进行ADSCs向DNs的定向分化。明场下细胞形态、电镜下超微结构、实时荧光定量多聚核苷酸链式反应（real-time quantitative polymerase chain reaction, RT-qPCR）检测神经发育相关基因表达确定最佳方案。而后通过免疫荧光染色、酶联免疫吸附试验（enzyme-linked immunosorbentAssays, ELISA）、高效液相色谱（high performance liquid chromatography，HPLC）和细胞膜片钳确定经最佳诱导方案分化成的细胞多巴胺能命运谱及功能。将诱导分化不同时间点的细胞进行转录组测序并进行后续分析，包括主成分分析（principal component analysis, PCA）、差异表达基因（differential expressed genes, DEGs）分析、功能富集分析及基因集分析等。确定细胞诱导分化后的多巴胺能命运，筛选关键差异基因并进行RT-qPCR表达验证。   

（4）6-OHDA大鼠PD模型进行人源细胞的移植，分组包括：ADSCs、神经诱导6d（NI 6d）ADSCs、NI 12d ADSCs、NI 24d ADSCs和hiPSCs神经诱导第18d（iPSC d18）。脑立体定位手术移植前，细胞均标记上带罗丹明B的超顺磁珠（molday ion rhodamine B, MIRB）。阿扑吗啡诱导旋转测试、旷场实验对大鼠进行行为学评价。磁共振成像（magnetic resonance imaging, MRI）示踪细胞在活体脑内的情况。免疫组织化学染色观察移植微环境小胶质细胞的活化和移植细胞的增殖、分化、迁移。蛋白质免疫印迹（Western blot, WB）和脱氧核苷酸末端转移酶（Terminal deoxynucleotidyl transferase, TdT）介导的dUTP缺口末端标记（TdT-mediated dUTP nick-end labeling, TUNEL）技术进行动物脑内蛋白改变的测定。食蟹猴采用下肢静脉注射1-甲基-4-苯基-1,2,3,6-四氢吡啶（1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine, MPTP）构建系统性PD模型。自体NI 12d ADSCs后进行脑立体定位移植。模型的筛选和移植有效性的评价均采用Bennazzouz量表评分和视频追踪系统。最后将ADSCs和NI 12d细胞注射到NOD-Prkdc^scidIL-2Rγ^null（NSG）小鼠的皮下和尾静脉，观察其是否具有致瘤性。取移植长达32周大鼠和16个月食蟹猴的主要脏器（心、肝、肾、肺）进行苏木素-伊红（hematoxylin-eosin, HE）染色，确定细胞的安全性。

结果

（1）hADSCs、cADSCs、mADSCs和rADSCs均采用胶原酶消化法进行分离，贴壁培养后呈成纤维细胞样生长。流式检测显示不同物种ADSCs的间充质干细胞标记物CD90、CD29为阳性，造血干细胞标记物CD34和血管内皮细胞标记物CD31为阴性，成脂和成骨分化能力检测显示hADSCs、cADSCs、mADSCs和rADSCs均具有成脂成骨分化能力，但其成脂成骨分化的时间略有差异。

（2）老年与年轻小鼠ADSCs在细胞超微结构存在不同，同时增殖、细胞因子分泌能力、衰老相关基因表达水平、凋亡细胞占有比例、细胞周期存在显著性差异（p<0.05）。而细胞形态、表面标记物、成脂成骨分化能力、SA-β-gal染色没有差异。转录组测序分析发现，随着ADSCs供体年龄的增加，Lef1表达显著增加，提示ADSCs的增龄性改变可能与Wnt信号通路有关。对年龄相关上调和下调DEGs分别进行富集分析，发现趋化因子信号通路和Hippo信号通路变化最明显。而对整体DEGs进行富集分析并结合后续的基因验证结果，发现趋化因子信号通路中的CCL7-CCL2-CCR2轴是不同年龄个体ADSCs差异的关键因素。

（3）确定建立了ADSCs向DNs诱导分化体系N2。采用N2方案的细胞诱导分化至6d处于DNs分化早期阶段，至第12d建立功能性DNs，其表达中脑DNs标记物，分泌多巴胺（dopamine, DA），具有接近DNs的静息膜电位，但不产生动作电位。ADSCs来源的DNs在细胞因子分泌方面具有其独特性。转录组测序筛选出许多神经诱导分化相关标记，如Ntsr1和Nptx1在诱导第6d细胞中高表达，而Foxs1阻碍DNs分化早期阶段细胞的形成，Itga8抑制ADSCs来源DNs的形成。轴突导向中的轴突延伸调节（regulation of axon extension involved in axon guidance）和间充质细胞分化（mesenchymal cell differentiation）是ADSCs来源DNs显著富集的基因本体论（gene ontology, GO）术语（terms）。而轴突导向（axon guidance）、神经素信号通路（neurite out-growth）和有助于神经突生长的神经细胞粘附分子信号（neural cell adhesion molecule signaling for neurite out-growth, NCAM signaling for neurite out-growth）是显著富集到ADSCs来源DNs的三条通路。

（4）大鼠细胞移植确定ADSCs、NI 12d细胞有效，且NI 12d细胞有效时长与iPSC d18一样，到达观察终点32周（p < 0.05）。与对照组相比，NI 12d细胞移植组旷场实验中总运动距离和时间增加，焦虑行为减少（p < 0.05）。说明NI 12d细胞是治疗PD的最佳移植细胞。成功构建的MPTP食蟹猴PD模型Bennazzouz量表评分稳定在8分以上，视频追踪系统显示动物存在明显的运动迟缓、震颤、肌强直和姿势步态障碍。经自体NI 12d细胞脑立体定位移植后，能有效改善行为，并且回调造模后上调的外周血小板数（p < 0.05）。MIRB标记的移植细胞能够在脑内生长，表达增殖细胞核抗原（proliferating cell nuclear antigen, PCNA）和趋化因子受体4（C-X-C chemokine receptor 4, CXCR4），激活移植部位的小胶质细胞，通过胼胝体向大脑对侧及后侧迁移。大鼠大脑转录组结果也显示移植处有免疫激活，同时炎症因子和抗炎因子均恢复到正常动物水平。后续WB检测显示，在动物模型中，移植后Akt和JNK通路被激活，最终抑制黑质多巴胺能神经元的凋亡。NSG小鼠静脉和皮下注射NI 12d细胞证实了细胞不具有成瘤性，且大鼠和食蟹猴移植后长期随访未发现明显的组织学病变，表明ADSCs来源的DNs可以安全得在PD临床治疗中应用。

结论 

（1） 实验动物来源的ADSCs与hADSCs的生物学特性具有一致性，是hADSCs临床前研究的良好模型，可用于评价临床前自体脂肪干细胞移植的有效性和安全性。

（2）老年供体来源的ADSCs存在年龄相关的改变。而CCL7-CCL2-CCR2轴是年老ADSCs改变的关键靶点。为了保证自体移植的效果，我们建议在青年期对ADSCs进行冻存，使用时尽量减少传代次数，或阻断CCL7-CCL2-CCR2轴，通过抑制趋化因子信号通路消除老年个体ADSCs的不利影响。

（3）首次证明ADSCs具有分化为功能性DNs的潜能，仅用时12天，且该细胞有其独特的细胞因子分泌特性，是临床治疗PD的种子细胞。

（4）ADSCs来源的DNs可有效改善6-OHDA大鼠的运动障碍。同样地，在与人类症状非常相似的MPTP非人灵长类动物中也观察到相同的治疗效果。替代、分化、迁移、旁分泌、免疫调节和抗凋亡在其中发挥重要作用。此外，在NSG小鼠和移植后动物中分别确定了ADSCs来源DNs的不存在短期和长期致瘤性。我们的研究结果表明，ADSCs来源的DNs能够有效治疗帕金森病动物模型，可以安全得扩展到帕金森病的临床治疗。

Objective

(1) To elucidate the consistency of adipose-derived stem cells (ADSCs) from laboratory animals and human (hADSCs), the isolation, culture and biological characteristics of ADSCs were systematic compared and analyzed, providing a foundation for the preclinical study of ADSCs transplantation.

(2) Senescence properties of ADSCs from different age donors and the influence of organismal aging on the proliferation and functions of ADSCs in vitro were accessed, providing the theoretical basis for the clinical application of autologous ADSCs transplantation.

(3) The possibility of ADSCs induced into midbrain dopaminergic neurons (mDNs) were accessed, and the differentiation protocol of ADSC-derived DNs were established. The gene expression and functions of cells during induction were systematically analyzed to further explore the key genes in ADSCs neuronal induction.

(4) The effects of ADSCs, induced cells, and human induced pluripotent stem cells (hiPSCs) on neurotoxin-based Sprague-Dawley (SD) rat and cynomolgus monkey models of Parkinson’s disease (PD) were systematically compared and evaluated. The safety of ADSCs and induced cells were also investigated, which provide experimental basis for the determination of the optimal cell types and appropriate induction time.

Methods

(1) hADSCs, mouse ADSCs (mADSCs), rat ADSCs (rADSCs), and cynomolgus monkey ADSCs (cADSCs) were identified by flow cytometry and differentiated towards adipogenic as well as osteogenic lineages.

(2) The ADSCs were obtained from 1-month-old and 20-month-old mice. The cells characteristics, functions, gene expression levels, apoptosis proportion, cell cycle, SA-β-gal staining, and transcription features were evaluated.

(3) Three conventional cell culture system N1, N2, and N3 were used for ADSCs neuronal induction. The optimal protocol was determined by cell morphology under bright field, ultrastructure under electron microscope, and expression of neurodevelopment-related genes detected via real-time quantitative polymerase chain reaction (RT-qPCR). Then, immunofluorescence staining, enzyme-linked immunosorbent assays (ELISA), high performance liquid chromatography (HPLC) and cell patch clamp were used to determine the stage at which cells are induced by an optimal protocol. mRNA transcriptome sequencing and bioinformatics analysis were performed on cells at different stages of induction, including gene expression analysis, differential expressed genes (DEGs) analysis, functional enrichment analysis, and gene set analysis. Key DEGs were screened and verified by RT-qPCR.

(4) The 6-OHDA rat PD model was transplanted with human ADSCs and induced ones. Grouping included: ADSCs, ADSCs at 6 days of neuronal induction (NI 6d), NI 12d cells, NI 24d cells, and hiPSCs at 18th day of neuronal induction (iPSC d18). Molday ion rhodamine B (MIRB)-labeled stem cells were transplanted by stereotactic brain surgery. Apomorphine induced rotation test and open field test were used to evaluate the behavior of rats. Magnetic resonance imaging (MRI) was used to trace the cells in the living brain. Immunohistochemical staining was used to observe the activation of microglia in the microenvironment. The proliferation, differentiation, and migration of transplanted cells were also detected. Western blot (WB) and terminal deoxynucleotidyl transferase (TdT) mediated dUTP nick-end labeling (TUNEL) were used to determine brain protein alterations. The systematical PD model was constructed by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) intravenously injected into the lower limbs of cynomolgus monkeys. Autologous ADSCs and NI 12d cells were brain stereotaxic transplanted. Bennazzouz scale and video tracking system were used to evaluate the modeling and effectiveness of transplantation. Lastly, ADSCs and NI 12d cells were injected into subcutaneous and caudal veins of NOD-Prkdc^scidIL-2Rγ^null (NSG) mice to observe whether they were tumorgenic. Hematoxylin - Eosin staining (HE staining) was performed on the vital organs (heart, liver, kidney, and lung) of 32-week-transplanted rats and 16-month-transplanted monkeys to determine the safety of the cells.

Results

(1) ADSCs of each species showed fibroblast-like morphology. Flow cytometry showed that mesenchymal stem cell surface markers CD90 and CD29 in all the ADSCs were positice, while the hemotopoietic stem cell marker CD34 and vascular endothelial cell marker CD31 were negative. Oil red O and alizarin red staining revealed that all ADSCs showed similar potential of osteogenic or adipogenic differentiation.

(2) Compared to ADSCs from 1-month-old mice, ADSCs from 20-month-old mice exhibited some senescence-associated changes, including inhibited abilities to proliferate (p < 0.05). Moreover, differentiation abilities, cell surface markers, and cytokines secreting differed between 1M and 20M ADSCs (p < 0.05). SA-β-Gal staining did not reveal differences between the two donor groups, while cells exhibited more remarkable age-related changes through continuous passages. Based on transcriptome analysis and further detection, the CCL7-CCL2-CCR2 axis is the most probable mechanism for the differences.

(3) The differentiated system N2 that ADSC-derived DNs was established. The cells induced by N2 protocol were at the early stage of DNs until day 6. Functional DNs were established on day 12, which time the cells expressed midbrain DNs markers and secreted dopamine (DA). Cells at day 12 had a resting membrane potential similar to DNs, which were lack of action potential. Besides, ADSC-derived DNs is unique in cytokine secretion. Transcriptomic sequencing showed that Ntsr1 and Nptx1 were highly expressed in cells on day 6. Foxs1 inhibited the early stage of DNs differentiation, and Itga8 inhibited the formation of DNs derived from ADSCs. Regulation of axon extension involved in axon guidance and mesenchymal cell differentiation were significantly enriched terms on day 12 of gene ontology (GO). Axon guidance, neurite out-growth, and neural cell adhesion molecule (NCAM) signaling for neurite out-of-growth were three pathway enriched on day 12.

(4) Cynomolgus monkeys lesioned by MPTP were successfully constructed that their Bennazzouz scale scores were stable over 8. Video tracking system showed monkeys that they had obvious bradykinesia, tremor, myotonia, and postural and gait disturbances. After NI 12d cell transplantation, the behavior was significantly improved (p<0.05). By transplanting NI 12d cells into the striatum of rats, which had been lesioned by 6-OHDA, the animals were able to recover from motor deficits. Differentiation, migration, paracrine functions, immunoregulation and anti-apoptosis play important roles in PD treatment using induced DNs derived from ADSCs. Furthermore, short-term tumorigenicity was detected in the xenograft model and additionally the long-term tumorigenicity of ADSC-derived neuronal cells was assessed for up to 32 weeks in rats and 16 months in monkeys. Collectively, our findings show the feasibility of ADSC-derived DA neurons for PD treatment in a preclinical context, marking safely expanded for the clinical treatment of PD.

Conclusion

(1) Laboratory animal ADSCs and hADSCs can be cultured using the same method and they all had similar surface marker expression, as well as adipogenic and osteogenic differentiation potential, indicating their universal applicability for the preclinical study of autologous ADSC transplantation.

(2) ADSCs from old donors have some age-related alterations. The CCL7-CCL2-CCR2 axis is a potential target for gene therapy to reduce the harmful effects of ADSCs from old donors. To improve on autologous transplantation, we would recommend that ADSCs should be cryopreserved in youth with a minimum number of passages or block CCL7-CCL2-CCR2 to abolish the effects of age-related alterations in ADSCs through the Chemokine signaling pathway.

  (3) ADSCs had the potential to be differentiated into DNs with unique cytokine secretion characteristics. They were proper cells for clinical treatment of PD.

  (4) ADSC-derived DNs corrected motor deficits in 6-OHDA rats. Notably, the therapeutic effects were also observed in MPTP-lesioned non-human primates, which closely mimics the features of human symptoms. Differentiation, migration, paracrine functions, immunoregulation and anti-apoptosis play important roles in PD treatment using ADSC-derived DNs. Furthermore, short-term tumorigenicity was detected in the xenograft model and additionally the long-term tumorigenicity of ADSC-derived neuronal cells was assessed for up to 32 weeks in rats and 16 months in monkeys. Collectively, our findings showed the feasibility of ADSC-derived DA neurons for PD treatment in a preclinical context, marking safely expanded for the clinical treatment of PD.