MRI对自体骨髓间质干细胞移植于缺血心肌后动态监测价值的实验研究

Dynamic MRI of ferumoxide-labeled bone mesenchymal stem cells after transplantation in infarcted myocardium

目的评价MR／对干细胞移植于缺血心肌后的动态监测价值。方法中华小型猪6头，抽取髂骨骨髓并制备自体骨髓间质干细胞（mesenchymal stem cells，MSCs），以注射用超顺磁性氧化铁颗粒（SPIO）及4′，6-二脒基-2-苯基吲哚（DAPI）标记细胞并进行标记率检测。开胸结扎冠状动脉左前降支制备小型猪心肌梗死模型。模型建立后14d，MR延迟增强成像（DE—MRI）检测梗死面积，之后行第2次开胸，于每只动物心肌梗死周边区和正常心肌区直视下经心外膜各注射标记的MSCs 2个点，同时分别于各区注射培养基2个点作为对照点。细胞移植后24h及21d，快速梯度回波序列（FGRE）T2^＊像检测干细胞移植点低信号区的面积及信号强度。T2^＊信号减低区范围的测量由FGRE面积法实现，其信号减低的程度以正常心肌区和该区的T2^＊值之差与正常心肌区T2^＊值的百分比表示。病理组织学检查心肌细胞形态、瘢痕形成、毛细血管密度及干细胞分布情况。不同时间点MSCs注射点低信号区面积及T2^＊信号强度的比较采用重复测量方差分析及配对t检验。干细胞注射点与对照点毛细血管密度的比较、梗死周边区及正常心肌区MSCs注射点信号衰减幅度和DAPI阳性细胞密度的比较采用成组资料t检验。结果DAPI及SPIO对MSCs的标记率均达100％。心肌梗死模型建立后第14天，小型猪心肌梗死面积为（33．6±8．9）％。细胞移植后24h，标记MSCs注射点于MRI显示为边界清晰的卵圆形T2^＊低信号区，梗死周边区与正常心肌区MSCs注射点T2^＊低信号区的信号强度[分别为（67．00±5．48）％、（61．92±7．76）％，t=1．65，P=0．1158）]及面积[分别为（0．56±0．24）、（0．52±0．25）cm^2，t=0．39，P=0．7044）]差异均无统计学意义。移植后3周，T2^＊低信号区与正常心肌区的对比程度均较前下降，梗死周边区减低至（40．12±5．93）％（t=9．53，P〈0．01）；正常心肌区减低为（46．92±6．25）％（t=11．03，P〈0．01）。梗死边缘区T2^＊减低的程度大于正常心肌区MSCs注射点，且2组间信号强度减弱幅度分别为（26．88±7．27）％和（15．00±4．51）％，差异有统计学意义（F=20．08，P=0．0003）。正常心肌区MSCs注射点心肌组织中荧光标记的细胞核分布密度高于梗死边缘区MSCs注射点[分别为（106±25）和（143±31）个／高倍镜（t=-2．47，P=0．0293）]。梗死边缘区MSCs注射点组织中毛细血管分布密度高于该区对照点[分别为（13．4±4．0）和（9．4±3．1）个／高倍镜，t=2．49，P=0．0229]。MSCs移植于心肌组织中3周后，普鲁士蓝染色阳性的铁颗粒仅位于部分移植MSCs细胞核周围。结论MRI可在一定时间内实现对移植干细胞的示踪并反映移植MSCs在心肌局部的数量变化趋势，但是对于移植细胞的半定量监测存在局限性。移植干细胞所在的微环境是影响其生存时间的重要因素之一。

Objective To investigate the ability of magnetic resonance imaging （MRI） in tracking magnetically labeled mesenchymal stem cells （MR-MSCs） in a swine myocardial infarction （MI） model. Methods Adult Chinese mini-pigs （n = 6） were subjected to open-chest experimental MI operation. Their autogeneic bone marrow-derived mesenchymal stem cells （MSCs） was cultured and doubly labeled with ferumoxides and DAPI. On the 14 th day after MSCs transplantation, the size and location of the myocardial infarction were assessed by using delayed-enhancement MRI （DE-MRI）. Then the labeled MSCs were injected intramyocardially into peri-infarct zone and normal myocardium. At 24 hrs and 3 weeks after injection, the contrast and the volume of the MR-MSCs hypointense lesion from the MR images were acquired, and the contrast was determined using the difference in signal intensity between the hypointense and normal myocardium divided by signal intensity of the normal region. After humane euthanasia, the heart was excised and histology corresponding to MRI slices that demonstrated MR-MSCs lesions was performed. Repeated-measures ANOVA and a paired t test were used for comparison of the contrast and the volume of the MR-MSCs hypointense lesion at different time points. Comparisons between independent groups were performed with the standard Student t test. Results The labeling efficiency of ferumoxides and DAPI was 100% o On the 14 th day after the MI operation, the average percentage of infracted myocardial area was （33. 6±8. 9）% . Twenty- four hours after MSCs transplantation, MSCs injection sites appeared as ovoid hypointensive lesions with sharp border on T2^ ＊ images. At 24 h after injection, the signal contrast [（67.00±5.48）% vs （61.92 ±7.76）%,t =1.65,P=0.1158]and the size [（0.56±0.24）cm^2 vs （0. 52± 0. 25 ）cm^2, t = 0. 39, P = 0. 7044 ] of the lesions showed no statistical difference between the periinfarct zone and the normal myocardium. At 3 weeks after injection, the signal contrast decreased and the size diminished both in the peri-infarct zone and in the normal myocardium. Moreover, the contrast of the lesions in peri-infarct zone decreased more significantly than that in normal myocardium [ （26. 88 ± 7. 27） % vs （ 15.00 ± 4. 51 ） %, F = 20. 08, P = 0. 0003 ]. Post mortem analysis found the fluorescently labeled MSCs demonstrated on histological sections. There were much more dense fluorescently labled MSCs per high power fields at injection sites of normal myocardium than at injection sites of peri-infarct zone [ （ 106 ± 25）/HPF vs （143 ±31 ）/HPF, t = -2.47, P =0. 0293 ]. In MSCs injection sites of the peri-infarct zone, the capillary density was significantly higher than that in control sites [ （13.4 ± 4. 0）/HPF vs （9.4 ± 3.1 ）/HPF, t = 2.49, P = 0. 0229]. At 3 weeks after injection, ferumoxide was contained within partial original MSCs. Conclusion Magnetic resonance imaging of MSCs is a feasible method for the in vivo tracking of transplanted stem cells and could reflect the tendency of the local stem cell quantity, but there still has limitation for the semi-quantitation of the transplanted stem cells.