中文摘要：

目的探讨应用1．5T MRI活体示踪注入兔膝股骨髁软骨缺损关节腔内的超顺磁性氧化铁粒子（SPIO）标记的骨髓间充质干细胞（MSCs）的可行性。方法从兔骨髓中分离培养MSCs，经体外采用SPIO和BrdU双重标记后，与壳聚糖支架复合，然后注射到兔膝股骨髁自体软骨缺损关节腔中，术后1d及4、8、12周应用MR对膝关节腔内注入的磁标记MSCs进行扫描示踪，并与组织切片普鲁士染色及免疫组化BrdU对照。结果体外标记的MSCs普鲁士染色和电镜检查显示细胞胞质内含致密铁颗粒，标记细胞可正常成软骨分化。复合物注射后MRI GRET2^＊ WI序列检查显示关节腔内磁标记MSCs产生颗粒状低信号改变至少12周，不同时相信号改变在空间上呈不一致性。MRI信号改变区域与组织学检查结果基本一致。结论兔MSCs经SPIO标记后仍然具有成软骨细胞分化能力，利用1．5TMRI活体示踪注入自体软骨缺损膝关节腔内的兔磁标记MSCs分布和迁移是可行的。

英文摘要：

Objective To evaluate the feasibility of in vivo magnetic resonance imaging （MRI） with 1.5T system tracking of the survival, migration and differentiation of magnetically labeled seed cellsbone marrow-derived mesenchymal stem cells （MSCs） injected into the articular cavity. Methods Rabbit MSCs were isolated, purified, expanded, and then coincubated in vitro with supermagnetic iron oxide particles （SPIO） and 5-bromo-2-deoxyuridine （BrdU）. Prussian blue staining and transmission electron microscopy were performed to observe the intracellular iron. Some labeled MSCs were subjected to chondrogenic differentiation and the phenotype was examined to assess their chondrogenic differentiation capacity. MSCs colabeled with SPIO nanoparticles and （BrdU were suspended in chitosan and glycerophosphate （C-GP） gel. Eighteen rabbits underwent damage to the femoral trochlea to create cartilage defect models, and randomly divided into 3 groups 1 week later： Group A （ n = 6） undergoing injection of the MSC suspension in C-GP gel into the intra-articular space of knee joints, Group B （ n =6）, injected with un-labeled MSC suspension in C-GP gel, and Group C （ n = 6）, without injection. MRI of the knee was performed 1,4, 8, and 12 weeks after the injection respectively on a certain numbers of rabbits . and then the rabbits were killed with their knee joints taken out to undergo immunohistochemistry. The MR imaging findings were compared with the histological findings. Results Prussian blue staining and transmission electron microscopy showed intracytoplasmic nanoparticles in the SPIO-labeled cells. Safranin-O staining showed deposition of proteoglycan and type II collagen outside both the labeled and unlabeled MSCs, showing chondrogenesis. GRE T2-weighted MR image showed marked hypointense signal void areas, representing the implanted MSCs, in the intra-articular space after the MSC injection in Group A that persisted for 12 weeks at least; 2 week after the MSC injection hypointense signal could be