【作者】 黄国锋；

【导师】 夏照帆； Martin Yarmush；

【作者基本信息】 第二军医大学 ， 外科学（专业学位）， 2013， 博士

【摘要】 研究背景：真皮替代物的研究一直是皮肤组织工程的重点和难点，也是复合型皮肤替代物发展的基础。真皮替代物作为创面修复过程中的支架结构，不仅能促进表皮细胞的增殖、迁移和分化，调节基底膜的形成，而且能引导成纤维细胞和血管内皮细胞的浸润增殖，沉积新的胶原和形成新的血管，从而实现真皮结构重建。迄今为止一系列的真皮替代物已经被成功研制出来，包括自然来源的材料，如异体或异种脱细胞真皮、胶原和透明质酸等，以及人工合成材料，包括一些聚合体生物材料和其他纳米材料等。另外的一些新型真皮替代物材料也正在被研制之中。在这些真皮替代物中，Integra和Tegaderm等已经被商品化并成功运用于临床治疗。大部分的真皮替代物都能很好地模拟正常真皮的组织结构特点，但是除了脱细胞真皮外，普遍都缺乏基底膜结构成分。而且即使是脱细胞真皮，为了去除表皮细胞层和真皮中的成纤维细胞和血管内皮细胞，一般需要很强的细胞清除剂处理，而后者会对真皮的基底膜成分产生严重的破坏。基底膜成分是正常皮肤中重要的成分，位于是表皮和真皮连接处，对于维持皮肤的正常功能起到重要的作用。正常皮肤中的表皮干细胞主要位于基底层，通过半桥粒结构紧密连接在基底膜表面；后者可以调节表皮干细胞的增殖，迁移和分化。体外分离培养的表皮干细胞由于失去了基底膜的支撑和调节作用，逐渐丧失了增殖的能力，分化成角质形成细胞。大量的研究表明：在人工合成的真皮支架表面添加自然来源或者人工合成的基底膜成分如IV型胶原和纤维弹性蛋白等，可以有效地促进表皮细胞的增殖，改善所形成的表皮结构的形态和功能。人羊膜基质(amniotic membrane, AM)可能是一种良好的真皮替代物材料。它来源于胎盘的最内层，可在分娩时获得，主要由三部分结构组成：单层立方状表皮细胞，富含细胞生长因子的厚基底膜，以及成纤维细胞散在其中的疏松网状纤维基质。羊膜基质含有I型胶原、IV型胶原、VII型胶原、弹性蛋白和糖胺聚糖等成分，类似于人体的真皮。羊膜的基底膜主要成分为层粘连蛋白、IV型和VII型胶原，与皮肤和角膜的基底膜成分相似，是人体内最厚的基底膜。羊膜还具有促进上皮化、抑制瘢痕增生、抗炎症、抗血管生成、抗菌和抗病毒等生物特性，且免疫原性极低，因而被广泛用作外科手术材料和创面覆盖物，尤其是用于眼科角膜重建治疗。近年来更有研究者更是将羊膜作为角膜干细胞、间充质干细胞、以及其他体细胞扩增和移植的载体，用于修复各种组织缺损。羊膜使用时既可以保留失活的表皮细胞，也可以完全去除表皮细胞。完整羊膜含有大量的细胞生长因子，有利于细胞培育过程中干细胞特性的维持；而脱细胞羊膜(accelullar amniotic membrane, AAM)借助于良好的基质和基底膜成分，能促进培育细胞的增殖、迁移和分化；部分体细胞在脱细胞羊膜上的扩增速度要明显快于完整羊膜。在本课题中，我们利用反复冻融和DNase消化的方法消化处理人新鲜羊膜，获得保留完整基底膜结构的羊膜基质；随后又利用水溶性碳二亚胺对脱细胞羊膜进行适当的交联处理，以改善脱细胞羊膜的机械强度和生物稳定性。得到的交联脱细胞羊膜具有良好的可操作性和抗酶降解能力，作为培养基质可以促进表皮细胞的体外增殖速度，而且保持了良好的生物组织相容性。将表皮细胞培养在交联脱细胞羊膜表面形成复合型表皮-真皮替代物，并将其用于移植修复裸鼠全层皮肤缺损，结果发现这种复合皮肤替代物可以明显促进创面愈合速度，改善新生表皮的厚度和功能，促进真皮结构重建，减轻创面收缩。实验方法：一、将新鲜羊膜根据脱细胞方法的不同分为以下几组：Dispase II消化+细胞刷处理组、Freeze-thaw+DNase消化组和完整羊膜对照组。通过下面的几种检测方法观察不同组的脱细胞方法效果。（一）H&E表面和切片染色；（二）Hoechst细胞核染色；（三）免疫组织化学抗IV型胶原、VI型胶原、VII型胶原、层粘连蛋白（Laminin）、主要组织相容性抗原-I（MHC-I）、MHC-II和波形蛋白（Vimentin）染色；（四）扫描电镜和透射电镜检查。二、利用水溶性碳二亚胺（EDC,0.05mmol/mgAAM）分别交联脱细胞羊膜5min，30min和6h，得到不同交联程度的脱细胞羊膜；然后通过人表皮细胞经浸提液培养和直接接触培养的方法观察交联脱细胞羊膜的细胞毒性效应；最后将交联脱细胞羊膜皮下埋置于具有正常免疫功能的大鼠以观察其体内组织相容性。通过以下的方法检测交联脱细胞羊膜的机械强度和生物稳定性，以及细胞毒性效应和体内外生物组织相容性。（一）茚三酮法测定脱细胞羊膜的交联程度--交联指数；（二）单轴张力计测定脱细胞羊膜交联前后的最大张力和最大拉伸长度；（三）交联前后脱细胞羊膜的体外胶原酶降解速度测定；（四）CCK-8法和Live/Dead染色观察表皮细胞的增殖活性；（五）H&E染色，Masson三色三胶原染色观察交联脱细胞羊膜体内埋置后的免疫炎症反应和支架降解情况；（六）免疫组织化学抗CD31, CD11b, CD4, CD8, CD68和Vimenin染色观察支架埋置后浸润细胞的类型，评估炎症反应。三、以交联脱细胞羊膜为载体体外扩增人表皮细胞，并与常规细胞培养皿相比较；构建表皮细胞-交联脱细胞羊膜的复合型皮肤替代物，并将其用于移植修复裸鼠全层皮肤缺损创面，单纯表皮膜片移植组和空白组作为对照。通过以下的方法检测表皮细胞的体外扩增速度和复合皮肤替代物移植后的创面愈合及真皮重建情况。（一）CCK-8法和Hoechst细胞核染色观察两组表皮细胞的扩增速度；（二）免疫组织化学抗P63染色观察表皮细胞的增殖活性；（三）H&E染色观察复合皮肤替代物的形态；（四）观察移植创面愈合过程的大体形态；（五）H&E染色和免疫组织化学抗laminin染色观察愈合创面的组织结构。实验结果：一、反复液氮冻融+DNase消化处理可以完全清除羊膜的上皮细胞层和间质细胞，而Dispase II消化+细胞刷法处理后的脱细胞羊膜仍有少量上皮细胞残留，间质细胞则基本不能清除。两种方法处理后的脱细胞羊膜DNA总量均显著降低，组间无明显差异；但反复液氮冻融+DNase消化处理的脱细胞羊膜保留的总蛋白量明显高于DispaseII消化+细胞刷处理的脱细胞羊膜（77.2±4.72%VS48.5±4.16%，p＜0.05）。反复液氮冻融+DNase消化处理的脱细胞羊膜的基底膜成分（层粘连蛋白、IV型胶原、VI型胶原和VII型胶原）基本保留，且对基质纤维结构无明显影响；而Dispase II消化+细胞刷法处理的脱细胞羊膜基底膜被严重破坏，蛋白成分丢失明显，且基质胶原纤维变得疏松，排列紊乱。正常羊膜中的表皮细胞和间质细胞均表达MHC-I抗原，不表达MHC-II抗原，其中间质细胞还表达Vimentin，而在反复液氮冻融+DNase消化处理的脱细胞羊膜中几乎未检测到MHC-I抗原和Vimentin的存在。二、脱细胞羊膜的形态极其柔软和光滑，经过EDC交联5min后依然平整，并具有一定的硬度，而交联30min和6h后逐渐变得卷曲和僵硬。随交联时间的增加，脱细胞羊膜的机械强度显著增强，体外胶原酶完全降解时间也明显延长，并与交联程度存在正相关关系。不同交联程度的脱细胞羊膜的浸提液培养7天对人表皮细胞的增殖活性无不良影响。交联5min的脱细胞羊膜直接负载人表皮细胞培养7天后细胞增殖活性良好，与未交联脱细胞羊膜组无明显统计学差异（p＞0.05）；而交联30min和6h的脱细胞羊膜组的细胞增殖活性明显受损，两组细胞在培养7天后的凋亡或死亡细胞比例分别为1.27±0.30%和10.02±1.43%，明显高于未交联和交联5min脱细胞羊膜组的0.42±0.14%和0.44±0.18%（p均＜0.05）。皮下埋置后的组织相容性检测结果显示交联5min的脱细胞羊膜在体内的完全降解时间约为4个月，降解后形成一层厚的皮下组织，胶原沉积和血管化良好，并无明显的急慢性炎症反应发生。三、表皮细胞在交联5min的脱细胞羊膜表面种植培养7和14天后的细胞相对增殖率分别为367±33%和631±43%，显著高于同一时间点的常规培养皿组（294±30%和503±41%, p均<0.05)。培养14天后表皮细胞在交联脱细胞羊膜表面形成一个2-3层的复层表皮结构。免疫组化染色结果显示交联脱细胞羊膜上P63阳性表皮细胞的比例明显高于对照的常规培养皿组（54.32±4.27%VS33.32±3.18%, p<0.05）。将培养形成的表皮细胞-交联脱细胞羊膜复合皮肤替代物移植于裸鼠全层皮肤缺损创面后，表皮细胞成活良好并完全封闭创面，形成类似正常皮肤的表皮。复合皮肤替代物移植组的创面修复效果要明显优于单纯表皮膜片组和空白对照组，创面收缩明显改善。创面组织学观察结果提示复合皮肤替代物移植后创面的真皮结构重建良好，新生基底膜厚而完整。实验结论：一、反复液氮冻融+DNase消化的脱细胞方法能有效去除羊膜的上皮细胞和间质细胞，优于传统的Dispase II消化+细胞刷处理方法；更重要的是该方法能有效保留羊膜的基质成分，尤其是基底膜结构，且脱细胞后羊膜的免疫原性极低。二、EDC（0.05mmol/mg AAM）交联5min的脱细胞羊膜，不仅具备改善了的机械强度和抗酶降解能力，而且无明显细胞毒性效应，能有效负载表皮细胞的粘附和增殖。皮下埋置实验结果显示该交联脱细胞羊膜具有良好的体内生物组织相容性和降解特性。三、交联脱细胞羊膜作为真皮替代物在体外可以负载并促进表皮细胞的快速扩增，有利于维持细胞的增殖能力。用交联脱细胞羊膜构建复合型皮肤替代物并移植可以促进全层皮肤缺损创面的真皮结构和基底膜重建，改善创面愈合质量，因而是理想的真皮支架材料。更多还原

【Abstract】 Background: The development of favorable dermal substitute has always been amain focus in skin tissue engineering research and the basis of composite skin equivalents.Dermal substitutes can serve as the structural template for wound healing by introducingdermal reconstruction, regulating the proliferation and differentiation of keratinocytes, andpromoting the formation of an intact and functional basement membrane (BM). Largenumbers of dermal substitutes have already been derived either from natural materials suchas allogeneic/xenogeneic acellular dermis, collagen, and hyaluronic acid, or from syntheticmaterials such as polymers and electrospun nanomaterials, and development of many othernew dermal substitutes is also in progress. Among all these currently available dermalsubstitutes, Integra and Tegaderm have already been successfully used in clinic. Althoughmost dermal substitutes can mimic the structural and fuctional characteristics of normaldermis, nearly all of them lack the component of basement membrane, except foraccellular dermis. Even in accellular dermis, since the removal of epidermal layer andfibroblasts and vascular endothelial cells in the matrix requires strong cell removaldetergents, severe damage can be caused to the basement membrane.Basement membrane is a very important structure in normal skin, located at theepidermal-dermal junction. It plays a vital role in maintaining the fine function of skin. Innormal skin, the epidermal stem cells attached closely to the basement membrane byhemidesmosomes, which can regulate the proliferation, migration and differentiation ofepidermal stem cells. Epidermal stem cells are deprived of the attachment and support ofbasement membrane when cultured in vitro, and gradually lose their ability of continousproliferation, and finally differentiate into epidermal keratinocytes. There are abundantstudies showing that by adding natural or artificial basement membrane components ontothe dermal substitutes, the in vitro proliferation rate of epidermal keratinocytes wassignificantly enhanced, and the function of the epidermal layer was also improved.Human amniotic membrane (AM) is the innerest layer of the placeta. It comprises asingle layer of epithelial cells, a dense BM and an avascular stromal matrix rich in collagen. The amniotic matrix is composed of type I, type IV and type VII collagen, fibronectin andglycosaminoglycans, similar with human dermis. There are fibroblasts in the matrixwithout containing any capillaries or vascular endothelial cells. The basement membranecompsists of laminin, type IV collagen and type VII collagen, similar with the basementmembrane of skin and cornea, and is the thickest basement membrane in the body. Theamniotic membrane also possesses properties of promoting epithelialization, alleviatinginflammation and inhibiting scarring. With very low immunogenicity, it has been widelyused in the treatment of various corneal diseases, and tissue banks of amniotic membranehave also been established in some countries. Many studies reported that the amnioticmembrane had good histocompatibility for xenografting/allografting and did not inducesignificant immune rejection response. The amniotic membrane matrix with a basementmembrane structure has been proved to be a favorable substrate for in vitro expansion ofhuman limbal stem cells, mesenchymal stem cells and epidermal stem cells, preventing celldifferentiation and maintaining the properties of stem cells, thus it could be a good carriercandidate for quick proliferation and grafting of epidermal keratinocytes. The intactamniotic membrane has plenty of growth factors and is very helpful in maintaining theproliferating ability of stem cells, while the acelluar amniotic membrane (AAM) is moresuitable for the adhesion, proliferation and migration of stem cells.In the present study, we used repeated freeze-thaw cycles and DNase digestion toprepare an acellular amniotic membrane with an intact BM structure, and cross-linked itwith soluble1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) to improve itsmechanical strength and biostability. The appropriately cross-linked acellular amnioticmembrane thus obtained had improved mechanical strength and enhanced biostability, andit was able to promote the adhesion and proliferation of epidermal keratinocytesremarkably, while retaining its good in vivo histocompatibility. Epidermal keratinocyteswere cultured on the cross-linked acellular amniotic membrane for2weeks to form acomposite skin equivalent, and transplanted onto full-thickness skin defects in nude mice.The results showed that the composite skin equivalent survived well and covered thewound in3weeks. New epidermis formation and dermal reconstruction during woundhealing in the composite skin equivalent transplantation group is apparently better than inthe epidermal sheet transplantation and blank control groups.Materials and Methods:(1) Fresh amniotic membranes obtained from15different donors were divided into three groups: one was treated with repeatd freeze-thaw and DNase digestion, one wastreated with Dispase II digestion and scraping, and the untreated one served as control. Thedecellularization effect in different groups was compared by the following examinations.1) H&E surface and section staining.2) Hoechst DNA staining.3) Immunohistochemical staining against laminin, type IV, type VI, and type VIIcollage, MHC-I, MHC-II, and vimentin.4) Scanning and Transmission Eelectronic Microscopy.(2) EDC cross-linking was used to improve the the mechanical strength andbiostability of acellular amniotic membrane, and then the histocompatibility of thecross-linked acellular amniotic membrane was examined. Briefly, the acellular amnioticmembrane was corss-linked with EDC at a concentration of0.05mmol per mg AAM for5min,30min and6h, resulting in acellular amniotic membranes with different cross-linkingdegrees. Epidermal keratinocytes were cultured either in extraction medium fromcross-linked AAM or cultured directly on cross-linked AAM to test the cytotoxicity of thecross-linked AAM. Biocompatibility of the cross-linked AAM was further evaluated in asubcutaneous model in immunocompetent rats. The following examinations wereperformed.1) The ninhydrin method was used to determine the cross-linking index.2) The uniaxial tension meter was used to measure the maximum tension andmaximum stretch length of acellular amniotic membrane.3) The in vitro degradation rate was measured by type I collagenase digestion method.4) CCK-8assay and live/dead staining were used to observe the viability of cells.5) H&E and Masson’s trichrosome stainings were used to observe the inflammatoryresponse and in vivo degradation condition of the scaffold after subcutaneous implantation.6) Immunohistochemical staining against CD31, CD11b, CD4, CD8, CD68andvimentin was used to detect the types of cells infiltrating the scaffold.(3) The EDC cross-linked AAM was used as a substrate for expansion of epidermalkeratinocytes, and compared with the conventional cell culture dish. Compostite skinequivalents constructed with keratinocytes and the cross-linked AAM were transplantedonto full-thickness skin defects in nude mice, and transplantation of epidermal sheet andsham transplantation groups served as control. The following tests were performed todetermine the proliferation rate of epidermal keratinocytes and to evaluate the effect of wound healing after transplantation.1) CCK-8assay and Hochest nucleus staining were used to determine the proliferationrate of cultured epidermal keratinocytes.2) H&E staining was performed to observe the structure of composite skinequivalents.3) Immunohistochemical staining against P63was performed to dertermine theproliferating status of epidermal keratinocytes.4) Gross appearance of the wound was recorded daily.5) H&E staining and immunohistochemical staining against laminin were used toexplore the histological structure of the healed wounds.Results:(1) The method of repeated freeze-thaw+DNase digestion was capable of removingboth the epithelial cells and mesenchymal cells in amniotic membranes clearly, while afterDispase II digestion and scraping there were still some epithelial cells residual, and themesenchymal cells were rarely removed. The cleavage rate of DNA amount between twodecellularization groups was not significantly different (p＞0.05), but the remain of totalprotein amount in the repeated freeze-thaw+DNase digestion group was significantlyhigher than that in the Dispase II+scraping group, as77.2±4.72%VS48.5±4.16%(p＜0.05). Laminin, type IV collagen, type VI collagen and type VII collagen were seen presentat the basement membrane area of the intact amniotic membrane, and after repeatedfreeze-thaw+DNase digestion treatment they were still there, similar with the intactamniotic membrane. In the Dispase II+scraping treated group, nearly all these basementmembrane components were gone. TEM results showed that the fibrous collagen structureof the matrix in the repeated freeze-thaw+DNase digestion group was barely influenced,while the collagen fibers in the Dispase II+scraping group became apparently loose. Boththe epithelial cells and mesenchymal cells in the aminotic membrane expressed MHC-Iantigen and did not express MHC-II antigen, and only the mesenchymal cells expressedvimentin, but after repeated freeze-thaw+DNase digestion, the staining of MHC-I,MHC-II and vimentin in the acellular amniotic membrane were all negative.(2) The fresh AAM appeared soft and smooth, and it was difficult to handle it as itwas easy to coil up. Appropriate EDC cross-linking made the AAM become a bit harderwith a flat and smooth surface, and it was easy for handling. SEM showed that EDCcross-linking made the originally uniform and dense mesh collagen fibers in the matrix of fresh AAM associate with each other into thicker fibrous bundles or even cords. Themechanical strength of AAM increased and its stretch length decreased with the extent ofcross-linking increasing. EDC cross-linking also significantly increased the biostability ofAAM. Epidermal keratinocytes cultured for7days using the extracted culture mediumfrom cross-linked AAM did not show any difference from normal culture medium group.Direct cytotoxicity test showed that there was no significant difference in the cell viabilityof epidermal keratinocytes between the fresh AAM and5min-AAM groups when culturedfor7days (p＞0.05), but the OD value of the30min-AAM and6h-AAM groups wassignificantly lower than that of the fresh AAM and5min-AAM groups at day7(1.27±0.30%and10.02±1.43%VS0.42±0.14%and0.44±0.18%, p<0.05). The5min-AAMretained the flat and smooth morphology and the ability of supporting the growth ofepidermal keratinocytes as did fresh AAM. In vivo histocompatibility observation resultsshowed that the5min-AAM degraded completely in4months, forming a thicksubcutaneous tissue, and without obvious acute and chronic inflammatory response,indicating that the5min-AAM possessed good biocompatibility.(3) CCK-8assay results showed that the relative cell viability of the5min-AAM groupat day7and14was367±33%and631±43%, respectively, significantly higher than294±30%and503±41%of the conventional cell culture dish (CCD) group (p<0.05). Theepidermal keratinocytes on the5min-AAM formed a2-3layer epidermis structure aftercultured for14days. On day7, the percentage of P63positive cells was54.32±4.27%,significantly higher than33.32±3.18%of the CCD group (p<0.05). When composite skinequivalents constructed with epidermal keratinocytes and5min-AAM were transplanted tofull-thickness skin defects in nude mice, the cells survived well and covered the wounds byforming a new epidermis in3weeks, similar to normal skin. The wound healing effect inthe composite skin equivalent transplantation group was apparently better than in theepidermal sheet transplantation and control group. Histological observation results showedthat dermal reconstruction in the healed wounds was good, with an intact and thickbasement membrane.Conclusions:(1) Repeated freeze-thaw+DNase digestion method is capable of removing theepithelial and mysenchymal cells clearly in aminotic membrane, better than theconventional Dispase II digestion+scraping method. More importantly, the basementmembrane of the amniotic membrane can be intactly retained after decellularization by repeated freeze-thaw+DNase digestion.(2) Appropriate cross-linking by EDC can improve the mechanical strength andanti-collagenase degradation ability of the acellular amniotic membrane, and maintain itsgood in vitro and in vivo biocompatibility.(3) As a culture substrate, the cross-linked acellular amniotic membrane can increasethe expansion rate of human epidermal keratinocytes by retaining their proliferating ability.It can also improve the healing of full-thickness skin defects by promoting dermalreconstruction and new basement membrane regeneration after grafted, thus is an idealdermal substitute.更多还原