Introduction

Compared to most normal cells and cancer cells, high-mannose-type (HM) N-glycans are enriched in human embryonic and mesenchymal stem cells (MSCs)¹. Due to the presence of five exchangeable hydroxyl groups on mannose, we hypothesized that MSCs transplanted in the brain may exhibit an enhanced, mannose-sensitive (man)CEST MRI signal compared to surrounding host cells. This would allow non-invasive tracking of stem cells in a “label-free” manner, avoiding three major issues with tracking of “labeled” cells: 1) Potential adverse effects of label on cell survival or function; 2) Limitations of long-term cell tracking due to cell division and dilution of label; and 3) Need for regulatory approval of label for clinical use and associated cost of GMP production.

Methods

In vitro studies: Galanthus nivalis lectin (GNL) was used as a mannose-specific lectin² to assess mannose expression on four cell lines: human MSCs (hMSCs), human astrocytes (hAstros), human glial-restricted progenitor cells (hGRPs), and normal human prostate cells (RWPE1). Cells were incubated with 20 µg/mL fluorescein (FITC)-GNL (FL-1241, Vector) at 4°C for 30 min. FACS analysis was performed using a Becton Dickinson LSR II flow cytometer. In vivo studies: 6 to 8-weeks old female immunodeficient rag2^−/− mice were anesthetized with isoflurane and placed on a stereotactic frame (BENCHmarkTM) with a burr hole drilled 0 mm caudal and 2 mm lateral to bregma. Using a Hamilton syringe, 2 μL of unlabeled cell suspension (1.5×10⁸ cells per mL in PBS) was injected into the striatum 3 mm at a rate of 0.2 μL/min. CEST MRI was performed 1 day after cell transplantation using an 11.7 T Bruker horizontal bore scanner. A modified RARE sequence (TR/TE=5,500/3.7 ms, RARE factor=16, 1 mm slice thickness, FOV=1.6×1.5 cm, matrix size=96×64, resolution=0.17×0.23, NA=2 and a saturation pulse B₁=2.4 µT) was used. Images were obtained with saturation frequencies from -5 ppm to +5 ppm.

Results

Flow cytometry analysis of FITC-GNL-labeled and unlabeled cells are shown in Fig. 1A. The highest mannose expression was found in hMSCs. One day after cell transplantation, hMSCs exhibited an in vivo CEST signal that was >2-fold higher than the mice transplanted with the other three cell lines (Fig. 1B). Serial CEST MRI was obtained over 14 days to track the transplantation of hMSCs in mice brain. As can be seen in Fig. 2A, the manCEST contrast of transplanted hMSCs reduced gradually from day 1 to day 14, which was validated with fluorescein-labeled GNL staining (Fig. 2B).

Conclusion

On the basis of the high abundance of high mannose (HM) N-glycans on hMSCs, we developed manCEST MRI as a novel mannose-sensitive technique for in vivo imaging of transplanted hMSCs. To the best of our knowledge, this is the first cell tracking study that uses native, unlabeled cells, without the need of any cellular modification or manipulation. While this new concept is still in its developmental proof-of-principle stage, we anticipate our label-free approach to be easily translatable for clinical stem cell therapy.

Acknowledgements

This project was supported by the Pearl and Yueh-Heng Yang Foundation.