研究背景：

进行性半侧颜面萎缩(Progressive Hemifacial Atrophy，PHA)是一种发病率较低的颅颌面疾病，主要特征是涉及半侧面部软组织，包括皮肤、皮下组织、肌肉及软骨出现不同程度的萎缩，儿童时期发病还会造成颧骨、颞骨及上、下颌骨组织发育不良，以皮下组织和结缔组织为甚。也有部分患者会出现神经系统和眼相关的并发症，以及斑秃和发色改变等。此病的病因目前尚不明确，无法针对其病因治疗，目前整复外科的治疗都是针对该疾病遗留的半侧颜面的萎缩凹陷畸形，以恢复和重建患侧颜面的形态和丰满度，力求患侧与正常侧对称。带血管蒂的游离皮瓣移植和正颌外科或骨性填充是修复面部不对称的常选用的方法，但由于游离皮瓣移植和颅颌面正颌手术创伤较大，且移植皮瓣存活与否的不确定性等因素而受到患者一定程度的排斥。自体脂肪移植技术因脂肪组织来源广泛、手术方法相对简便、自体来源无免疫原性、术后恢复快等优势，近年来在治疗轻中度半侧颜面萎缩领域越来越受到推崇。然而，半侧颜面萎缩患者进行自体脂肪移植后的脂肪存活率和健康求美者进行自体脂肪移植后的脂肪存活率是否存在显著差异，这类患者的脂肪组织、脂肪细胞、细胞衍生物和健康求美者是否存在显著差异，细胞辅助脂肪移植技术（Cell-assisted Lipotransfer，CAL）和外泌体辅助脂肪移植技术（Exosome-assisted Lipotransfer，EAL)对此类患者的临床应用是否存在价值，国内外尚缺乏详细的实验研究和数据支持。

研究目的：

1. 对比进行性半侧颜面萎缩患者和健康求美者吸脂区脂肪来源的干细胞及外泌体的体外生物学特性：两组脂肪来源干细胞（ADSCs）的三系分化（成脂、成骨和成软骨）、脂肪来源干细胞的表型、CCK-8生长增殖曲线、在糖氧剥夺（Oxygen and Glucose Deprivation，OGD）条件下的增殖差异以及自噬、凋亡、成脂、衰老方面的基因表达差异和检测两组外泌体的电镜下形态（TEM)、纳米颗粒跟踪分析（Nanoparticle Tracking Analysis，NTA）及Western blot（WB)检测表面蛋白表达的异同。

2. 建立裸鼠的颗粒脂肪移植模型，脂肪移植物分四组：第一组为半侧颜面萎缩患者腹部吸脂区脂肪来源干细胞（ADSCs）进行辅助的颗粒脂肪转移组；第二组为健康求美者腹部吸脂区脂肪来源干细胞（ADSCs）进行辅助的颗粒脂肪转移组；第三组为半侧颜面萎缩患者腹部吸脂区脂肪来源干细胞的上清液提取的外泌体进行辅助的颗粒脂肪转移组；第四组为对照组，为单纯的半侧颜面萎缩患者腹部吸脂获得的颗粒脂肪转移组。对比四组脂肪移植物在移植后2周，4周，8周及12周时移植物的体积和质量保留率的差异。

3. 将从裸鼠皮下取出的四组脂肪移植物分别进行苏木精-伊红(H&E)和免疫组化染色：CD31染色检测血管形成，CD68染色检测巨噬细胞浸润，Perilipin-1染色检测脂肪细胞形成。采用RT-qPCR分析对比它们在自噬、凋亡、成脂、衰老和血管生成等方面的mRNA表达水平异同，为脂肪来源干细胞和外泌体辅助脂肪移植治疗半侧颜面萎缩患者提供理论支持。

研究方法：

    第一部分：本部分为体外细胞学实验，分别从健康求美者和半侧颜面萎缩患者腹部抽吸的脂肪按照标准流程胶原酶消化法提取脂肪来源干细胞（ADSCs），分别标记为PHA-ADSCs和NORM-ADSCs。两组ADSCs的上清液分别提取外泌体后标记为PHA-ADSCs-Exos和NORM-ADSCs-Exos，并分别用NTA粒径、TEM透射电镜和WB检测CD63和TSG101的异同。

第二部分：本部分为BALB/c裸鼠动物模型的建立。

脂肪移植物分四组：

第1组：为半侧颜面萎缩患者腹部吸脂区脂肪来源干细胞（ADSCs）进行辅助的颗粒脂肪转移组，包括半侧颜面萎缩患者的脂肪来源干细胞, 即PHA-ADSCs(约10⁵细胞/100μl)混合400μl的相应半侧颜面萎缩患者的颗粒脂肪；

第2组：为健康求美者腹部吸脂区脂肪来源干细胞（ADSCs）进行辅助的颗粒脂肪转移组，包括健康求美者的脂肪来源干细胞，即NORM-ADSCs(约10⁵细胞/100μl)混合400μl相应健康求美者的颗粒脂肪；

第3组：为外泌体辅助脂肪转移(100μl)组，包含半侧颜面萎缩患者来源的外泌体(约50μg/100μl)混合400μl半侧颜面萎缩患者的颗粒脂肪；

第4组：为对照组，由400μl半侧颜面萎缩患者腹部吸脂获得的颗粒脂肪混合100μl磷酸盐缓冲盐(PBS)。

将四组不同的脂肪移植物分别移植到裸鼠体内（双侧肩部和背部），每只裸鼠体内注射的4个位点采用顺时针注射法，以消除因注射位置不同可能造成的差异。分别在2w、4w、8w、12w时解剖出脂肪移植物（每个时间点处死的裸鼠数量n=4)，称重。并计算四组脂肪移植物的重量和体积保留率。

第三部分：本部分为动物体内实验，将不同时间点取出的脂肪移植物进行APRC5、ATG5、ATG7、ATG12、BAX、PPARG、CDKN1A、CDKN2A基因和蛋白表达水平分析，比较PHA-ADSCs组和NORM-ADSCs组的脂肪移植物的差异。同时进行H&E、CD31、CD68和Perilipin-1免疫组化染色，评估各组脂肪移植物的组织学差异。

研究结果：

第一部分：细胞计数试剂盒（CCK-8）检测显示半侧颜面萎缩患者的脂肪来源干细胞（PHA-ADSCs）的细胞增殖能力在第6天开始显著不及健康求美者的脂肪来源干细胞（NORM-ADSCs）。两组ADSCs均具有典型的脂肪来源干细胞的表型特征和三系分化能力。半侧颜面萎缩患者的脂肪来源干细胞表现出比健康求美者的脂肪来源干细胞较弱的脂滴形成能力。Transwell法检测出半侧颜面萎缩患者的脂肪来源干细胞的细胞迁移能力较健康求美者的脂肪来源干细胞弱。两组细胞Calcein/PI双染后，ImageJ计算活/死细胞的比例显示：经过糖氧剥夺（OGD）处理后，半侧颜面萎缩患者的脂肪来源干细胞的活细胞比例为46.11%，而健康求美者的脂肪来源干细胞的活细胞比例为54.21%。半侧颜面萎缩患者的脂肪来源干细胞和健康求美者的脂肪来源干细胞进行相关检测比较, 得出结果：半侧颜面萎缩组的脂肪来源干细胞的凋亡率较高，ATG7和ATG12表达显著较低，BAX的表达水平显著较高，ARPC5的表达水平显著较低，CDKN1A和CDKN2A的表达显著较高。蛋白质印迹（WB)分析显示：两组的ADSCs的外泌体表面标记物CD63和TSG101均有阳性表达，但表达强度的有显著差异，半侧颜面萎缩组明显低于健康对照组。

第二部分：四组脂肪移植物，均呈黄色脂肪组织样外观。与对照组（单纯的颗粒脂肪移植组）相比，半侧颜面萎缩患者的脂肪来源干细胞（PHA-ADSCs）辅助脂肪移植组、健康求美者的脂肪来源干细胞（NORM-ADSCs）辅助脂肪移植组和外泌体辅助脂肪移植组的脂肪移植物的体积和重量更大。12周时，PHA-ADSCs辅助组的脂肪移植物的体积保留率为40.66%，NORM-ADSCs组脂肪移植物的体积保留率为48%，外泌体组的体积保留率为38.24%，而对照组的体积保留率仅为27.44%。此外，12周时的重量保留率与12周时的体积保留率相似。对照组的重量保留率为24.67%，PHA-ADSCs辅助组的重量保留率为35.50%，NORM-ADSCs辅助组的重量保留率为42.47%，外泌体辅助组的重量保留率为31.08%。

第三部分：H&E染色全景扫描显示在移植后2周和12周后四组脂肪移植物中，除了对照组，脂肪移植物的外周区和中心区都有大量的新生脂肪细胞，特别是健康求美者的脂肪来源干细胞（NORM-ADSCs）辅助组。此外，半侧颜面萎缩患者的脂肪来源干细胞（PHA-ADSCs）辅助组和外泌体辅助组之间在新生的脂肪细胞方面没有显著差异。对脂肪细胞数量的定量分析显示：从2周到12周，细胞辅助和外泌体辅助均使移植物的中央区和外周区的脂肪细胞数量增加。CD31染色显示细胞辅助组和外泌体辅助组的脂肪移植物在12周时的血管网络重建较对照组更好。通过VEGF的mRNA表达水平分析，证实了健康求美者的脂肪来源干细胞（NORM-ADSCs）辅助组的脂肪细胞活性和血管重建能力均优于半侧颜面萎缩患者的脂肪来源干细胞（PHA-ADSCs）辅助组。12周时，脂周蛋白染色显示细胞辅助组和外泌体组的移植物的脂周蛋白阳性脂肪细胞的数量显著高于对照组，其中NORM-ADSCs组高于PHA-ADSCs组，这与12周时脂肪生成相关基因PPARG的表达水平对比结果完全吻合。CD68染色显示：在12周时，其他三组的巨噬细胞浸润数量明显比对照组少。PHA-ADSCs辅助组和NORM-ADSCs辅助组的RT-qPCR检测结果相比：自噬下降，凋亡增加，细胞更新和再生能力下降。

结论：

半侧颜面萎缩患者的脂肪来源干细胞（PHA-ADSCs）较健康求美者的脂肪来源干细胞（NORM-ADSCs）具有相对较弱的增殖能力，凋亡率较高，脂滴形成较少，细胞迁移能力较弱，对糖氧剥夺（OGD）环境的耐受性较差，容易衰老，自我更新和可修复性较弱。半侧颜面萎缩患者的脂肪来源干细胞（PHA-ADSCs）辅助颗粒脂肪移植和外泌体辅助颗粒脂肪移植可显著提高半侧颜面萎缩患者脂肪移植后的脂肪留存率。细胞辅助和外泌体辅助脂肪移植是治疗进行性半侧颜面萎缩的一种行之有效的方法。

Background：

Progressive Hemifacial Atrophy (PHA) is a rare craniomaxillofacial disease， which involves the hemifacial soft tissues, including skin，subcutaneous tissue, muscle，and  cartilage, especially subcutaneous and connective tissues. Those who develop the disease in childhood also suffer from dysplasia of the zygomatic temporal bone, maxillary and mandibular. Some patients also develop neurological and ocular complications, alopecia areata, and hair color changes. The etiology of this disease is not clear at present, so it cannot be treated according to its etiology. At present, plastic surgery treatment is aimed at the sunken deformity of the hemifacial left by the disease to restore and rebuild the morphology and fullness of the face and strive to make the affected side and the normal side symmetrical. Vascularized free flap transplantation and orthognathic surgery or bone filling are the most commonly used methods for repairing facial asymmetry. However, due to the enormous trauma of free flap transplantation and craniomaxillofacial orthognathic surgery and the uncertainty of the survival of the transplanted flap, patients have been rejected to a certain extent. Autologous fat transplantation has been promoted more and more for the treatment of mild to moderate PHA in recent years due to its advantages, such as the relatively simple operation and the wide range of sources of adipose tissue, no immunogenicity, and fast recovery. However, whether the fat survival rate of patients with hemifacial atrophy who underwent autologous fat transplantation was significantly different from that of regular patients, whether adipose tissue cells and cell derivatives are significantly different from those of healthy patients, and whether cell-assisted lipotransfer (CAL) and exosome-assisted lipotransfer (EAL) is valuable for these patients are still lacking in detailed experimental studies and data support at home and abroad.

Objectives：

1. To compare the in vitro biological characteristics of adipose-derived adipogenic stem cells and exosomes in liposuction area between patients with progressive hemifacial atrophy and healthy patients, including the differentiation of ADSCs between the two groups, the phenotypes, the growth increment curve of CCK-8, the increment under OGD condition, and the gene expression differences in autophagy, apoptosis, adipogenesis, and senescence. The morphology of exosomes by electron microscopy, NTA particle size, and the expression of surface proteins by WB were detected.

2. To establish granular fat transplantation model in nude mice.To compare the differences in fat graft volume and mass retention at 2, 4, 8, and 12 weeks after transplantation among two groups of cell-assisted fat transplantation (PHA-ADSC-assisted group and NORM-ADSCs-assisted group), one experimental group of exosome-assisted fat transplantation and the control group (granulated fat transplantation alone).

3. To perform hematoxylin-eosin (H&E) and immunohistochemical staining, including CD31 staining for angiogenesis, CD68 staining for macrophage infiltration, and Perilipin staining for adipogenesis, on the four groups of fat grafts removed from the subcutaneous skin of nude mice.RT-qPCR was used to analyze and compare the similarities and differences of autophagy, apoptosis, lipid senescence, and angiogenesis. This experiment aims to provide theoretical support for the treatment of PHA patients with adipogenic stem cells and exosomes assisted fat transplantation.

Methods:

Part I: This part was the in vitro cytological experiments part. ADSCs were extracted from abdominal fat aspirated from healthy young patients and patients with hemifacial atrophy according to the standard procedure of collagenase digestion and labeled PHA-ADSCs and Norm-ADSCs, respectively. Proliferation ability (CCK-8), stem cell phenotype (flow cytometry +BD kit), Triallel differentiation ability (oil red O staining, alizarin red staining, and alizarin blue staining after lipid, osteogenic and chondrogenic induction), migration ability (Transwell), apoptosis of ADSCs under OGD conditions, self-repair ability, apoptosis and autophagy (RT-qPCR for related gene expression) were detected respectively. The exosomes (PHA-ADSCs-Exos and Norm-ADSCs-Exos) were extracted from the supernatant of ADSCs in the two groups. NTA particle size, TEM, and WB were used to detect the similarities and differences of CD63 and TSG101.

Part II: This part is the establishment of the BALB/ C nude mouse model. Fat grafts were divided into four groups. Two groups were cell-assisted fat transfer (CAL) group, including PHA-ADSCs (about 10⁵ cells /100μL) mixed with 400μ L of granular fat from corresponding PHA patients and NORM-ADSCs (about 10⁵ cells /100μL) mixed with 400μL of granular fat from corresponding healthy patients. The third group was the exosome-assisted fat transfer group consisting of 100μL granular fat from PHA patients with exosomes (about 50μg/100μL) mixed with 400μl granular fat from PHA patients. The fourth group was the control group, including 400μl granular fat from PHA patients and 100μL phosphate buffer salt (PBS). Four different groups of fat grafts were transplanted into nude mice (bilateral shoulders and back). Each of the four injection sites in nude mice was injected clockwise to eliminate possible differences due to different injection sites. Four different groups of fat grafts were transplanted into nude mice (bilateral shoulders and back). Each of the four injection sites in nude mice was injected clockwise to eliminate possible differences due to different injection sites.

Part III: This part is the in vivo animal experiments part. APRC5, ATG5, ATG7 ATG12, BAX, PPARG, CDKN1A, and CDKN2A gene and protein expression levels were detected on the fat grafts removed at different time points by real-time (RT)-qPCR to compare the gene and protein expression differences of fat grafts between the PHA-ADSCs-assisted group and NORM-ADSCs-assisted group. H&E staining, CD31, CD68, and Perilipin-1 immunohistochemical staining were used to evaluate the specific histological changes of fat grafts in each group.

Results:

Part I: CCK-8 showed a statistically significant increase in cell proliferation on day 6, and the cell proliferation ability of NORM-ADSCs was significantly better than that of PHA-ADSCs from day six after transplantation. Both groups of ADSCs had typical adipogenic stem cell phenotype and differentiation ability. PHA-ADSCs showed weaker droplet formation ability than NORM-ADSCs.Transwell showed cell migration ability in PHA-ADSCs was weaker than that in NORM-ADSCs. After Calcein/PI double staining, ImageJ calculated the alive/dead ratio of PHA-ADSCs and NORM-ADSCs after OGD treatment was 46.11% and 54.21%, respectively. ATG7 and ATG12 were significantly down-regulated in PHA-ADSCs, and the apoptosis rate was higher in PHA-ADSCs. BAX expression was significantly up-regulated in PHA-ADSCs. In PHA-ADSCs, the expression of ARPC5 was significantly down-regulated. The expression of CDKN1A and CDKN2A was significantly up-regulated in PHA-ADSCs. WB analysis confirmed that CD63 and TSG101 were the positive expressions of ADSCs exosome surface markers in the two groups, but the difference in expression intensity was significant.

Part II: The fat grafts of the four groups all showed a yellow adipose tissue appearance. Compared with the control group, PHA-ADSCs, NORM-ADSCs and exosome groups had larger fat grafts in volume and weight. At 12 weeks, the volume retention rate of fat grafts was 40.66% in the PHA-ADSCs group, 48% in the NORM-ADSCs group, and 38.24% in the exosome group. The volume retention rate in the control group was 27.44%. In addition, the weight retention rate was similar to that at 12 weeks. The weight retention rate was 24.67% in the control group, 35.50% in the PHA-ADSCs group, 42.47% in the NORM-ADSCs group, and 31.08% in the exosome group.

Part III: HE staining panoramas showed better preservation of new fat cells in the peripheral and central areas of fat grafts, especially in the NORM-ADSCs group, among the four groups of fat grafts except the control group at 2 and 12 weeks after transplantation. In addition, there was no significant difference in adipocyte formation between the PHA-ADSCs and exosome groups. Quantitative analysis of the number of adipocytes showed that both cell-assisted and exosomal assistance increased the number of adipocytes in the central and peripheral regions from 2 weeks to 12 weeks. CD31 staining showed that adipose grafts in the ADSCs and exosomes group had better vascular network reconstruction at 12 weeks than the control group. The mRNA expression analysis of VEGF confirmed that the adipocyte activity and vascular remodeling ability of the norm-Adscs group were better than that of the PHA-ADSCs group. At week 12, the perilipin-1 staining showed that the number of perilipin-1 adipose-positive adipocytes in the ADSCs group and exosome group was significantly higher than that in the control group, and the NORM-ADSCs group was higher than the PHA-ADSCs group, which was entirely consistent with the expression result of adipogenesis gene PPARG at 12 weeks. CD68 staining showed that at 12 weeks, macrophage infiltration in the other three groups was significantly reduced compared with the control group. RT-PCR results showed that autophagy was down-regulated, apoptosis was up-regulated, and cell regeneration ability was down-regulated in the PHA-ADSCs group.

Conclusions:

Compared with NORM-ADSCs, PHA-ADSCs have relatively weak proliferation ability, higher apoptosis rate, less lipid droplet formation, weak cell migration ability, poor tolerance to OGD, easy aging, and weak self-renewal and repairability. PHA-ADSCs-assisted fat transplantation and exosome-assisted fat transplantation can significantly improve fat retention after fat transplantation in hemifacial atrophy patients. Cell-assisted and exosome-assisted fat transfer is an effective method to treat progressive hemifacial atrophy.