Non-invasive longitudinal tracking of human amniotic fluid stem cells in the mouse heart

Abstract

Human stem cells from various sources have potential therapeutic applications. The clinical implementation of these therapies introduces the need for methods of noninvasive tracking of cells. The purpose of this study was to evaluate a high resolution magnetic resonance imaging (MRI) technique for in vivo detection and tracking of superparamagnetic micron sized iron oxide particle (MPIO)-labeled human amniotic fluid stem (hAFS) cells injected in the mouse heart. Because of the small subject size, MR signal and resolution of the in vivo MRI were increased using strong gradients, a 7.0 Tesla magnet, and an ECG and respiratory gated gradient echo sequence. MRI images of mouse heart were acquired during a 4 week course of this longitudinal study. At the end of the study, histological analysis was used to correlate cell localization with the MRI results. Introduction of MPIOs into hAFS had no significant effect upon cell proliferation and differentiation. Results of flow cytometry analysis indicated that hAFS cells remained labeled for up to 4 weeks. MRI of MPIO-labeled hAFS cells injected in agarose gels resulted in significant hypointense regions. Labeled hAFS cells injected into mouse hearts produced hypointense regions in the MR images that could be detected 24 hours and 7, 14, 21 and 28 days post injection. The co-localization of labeled cells within the hypointense regions was confirmed by histological analysis. These results indicate that high resolution MRI can be used successfully for noninvasive longitudinal tracking of hAFS cells injected in the mouse heart. The potential utility of this finding is that injected stem cells can be tracked in vivo and might serve to monitor cell survival, proliferation and integration into myocardial tissue.

FIG. 1.

MPIO labeling of human amniotic fluid stem cells. Cells were imaged by fluorescence microscopy to confirm successful incorporation of the MPIOs in the cells (A and B). Agarose phantoms were made by injecting 1.5 × 10⁶ MPIO labeled stem cells into 1% agarose in conical tubes and imaged by MRI. A clear hypointense region where the labeled stem cells were injected was imaged successfully by MRI (C). The control phantom demonstrated no hypointense region (D).

FIG. 2.

Proliferation analysis of MPIOs-labeled hAFS cells. 8,000 hAFS cells were plated in each well of a 24 well plate. Cells were counted in triplicates using a Coulter Counter after 1, 2, 3, 4, and 5 days. Counts were plotted as a function of percent change in growth relative to Day 1. Results demonstrated that the labeled stem cells do not differ in proliferation from the unlabeled stem cells.

FIG. 3.

Time course analysis of MPIO-labeled hAFS cells. The percentage of labeled hAFS cells was determined by flow cytometry at several time points up to 28 days after labeling (A). There are statistically significant differences at 7 and 14 days (*p = 0.0213 and **p = 0.0095, respectively), but at 21 and 28 days there are no statistical differences (p > 0.05) between the two used doses of MPIOs. Histograms (B and C) represent the distribution of labeled hAFS cells with 20 μL of MPIOs compared with unlabeled AFS cells after 24 hours and 28 days, respectively.

FIG. 4.

Lineage differentiation of MPIO-labeled hAFS cells. The multipotentiality of MPIO-labeled hAFS cells was confirmed after adipogenic and osteogenic differentiation. After 16 days of adipogenic differentiation, MPIO-labeled hAFS cells were analyzed by Oil-Red-O assay demonstrating presence of lipid vacuoles in the differentiated MPIO-labeled hAFS cells. After 23 days of osteogenic differentiation, cells were analyzed by Alizarin Red Assay demonstrating presence of calcium production in the differentiated MPIO-labeled hAFS cells.

FIG. 5.

MPIO-labeled cells are detectd in the live mouse heart by MRI. One million cells were injected in one location in the mouse heart. Multiple mice (n = 6) were imaged by MRI at the short and long axis images after 1 day and after 1, 2, and 4 weeks, as indicated. In all scans, significant hypointense regions are seen in the location of the MPIO-labeled hAFS cell injection sites. Size and intensity remains essentially unchanged.

FIG. 6.

Quantification of the hypointense region. (A) Depicts the average of the normalized hypointense region volume over time. The average normalized hypointense region volume remained unchanged for the duration of the study. (B) Depicts the average signal intensity of the hypointense region/tissue intensity over time. The average signal intensity of the hypointense regions remained unchanged for the duration of the study.

FIG. 7.

Localization of MPIO labeled hAFS cells in the live mouse heart by MRI and histology. One million cells were injected in two different locations in the left ventricle of the mouse heart. The mice were scanned by MRI after 14 days (A). Cell integration was further confirmed histologically by visualizing MPIOs using fluorescence microscopy (B and C) and by Prussian blue Iron staining indicating that the labeled cells colocalize with the hypointense region seen by MRI (D and E). Further, it was confirmed by immunostaining for a human specific nuclear matrix antibody (Anti-NuMA) to prove colocalization of MPIOs with injected AFS cells (scale bar = 10 μm) (F).