Stem cells offer an exciting new branch of therapy to treat a variety of conditions and diseases. It is therefore important to develop methods to monitor cell survival and location after transplantation. Due to its many advantages over conventional organic dyes, QDs serve as good candidates to monitor these parameters. In order to evaluate their in vivo ability, we delivered them by using commercially available QTracker. Strategies for ex vivo cell labeling by QDs include non-specific endocytosis, microinjection, liposome mediated uptake, electroporation, and peptide-based reagents. Previous studies have shown that the liposome-based reagent Lipofectamin 2000 had the highest delivery efficiency, but the QDs were delivered in aggregates . Electroporation also delivered QDs in aggregates , and may even cause cell death. Peptide-based QTracker  reagents (Invitrogen, CA) deliver QDs into the live cells, and have been shown to be an excellent and easy tool for studying live cell mobility  and cell fusion .
In this report, we evaluated ES cells labeled with QDs using commercially available Qtracker for non invasive in vivo imaging in living mice. Twenty-four hours after labeling ES cells with QDs, 72% of the cells were positive. However, by day 4 the percentage of positive cells dropped to 4%. This dramatic decrease could be due to the rapid division of ES cells (doubling time of 12 – 15 hours) or QD diffusion out of dividing cells over time thus causing a dilution of QD signal. The dramatic decrease in signal is consistent with a previous study that used QDs to label human cervical adenocarcinoma cells .
Another important question is whether QDs affect ES cell properties (i.e., pluripotency and self-renewal) that make them an attractive choice for regenerative therapy. Previous studies have shown that QD toxicity is dose dependent with increasing concentrations affecting cell growth and viability . However, we were interested in any toxicity caused at concentrations used for labeling cells for in vivo applications. Therefore, we examined ES cell proliferation and viability at one QD concentration (10 nM) and observed no significant changes between QD labeled ES cells and control unlabeled ES cells. This was true for all QDs tested: QD 525, 565, 605, 655, 705, and 800. These results concur with the study by Jaiswal et al. that also showed no adverse effects by QDs on the viability, morphology, function, and development of various other cells . Likewise, we confirmed that QDs also had no adverse affect on ES cell differentiation based on RT-PCR analysis of germ layer specific genes. Implanted ES cells are known to form teratoma tumors with a variety of differentiated tissues . In Figure 6, we found that the teratoma consisted of a variety of tissues including respiratory epithelium, osteochondroid, squamous cell, and immature brain-like neural cell based on histology. This confirmed that QD labeling did not affect in vivo differentiation as well. However, although ES cell-derived teratomas were retrieved from the animals, they were not shown to be QD labelled. We believe that the in vivo signal could be due to uptake of QD by neighboring host cells. Thus, the poor retention of QDs in targets cells may be a problem for long-term tracking, and more detailed analysis are needed to address this issue in the future.
Another advantage of QDs is their ability to do multiplex imaging of different QDs at the same time. However, in our study, ES cells labeled with different QDs were only capable of being imaged up to day 2 after subcutaneous implantation. A likely cause for this could be the loss of signal due to rapid cell division. Another possible cause could be serum instability of the QDs. Cai et al. reported that QD 705 lost 14% of its original intensity after 24 hours of incubation in mouse serum . Any loss of signal could hamper detection of QD labeled cells at later time points, especially those that are not within the near-infrared region since signals from these QDs will also be mostly absorbed by the skin. For those QDs that are in near-infrared region, QD 705 and QD 800, the difference in intensity could be due to transfection efficiency since these two QDs have similar extinction coefficients and quantum yield according to the manufacturer. However, extinction coefficients and quantum yield data were obtained in vitro and not in an animal setting. Moreover, the transfection efficiency was similar across all QDs. Therefore, we believe transfection efficiency is unlikely to be the cause of the difference in intensity observed in vivo. Due to its higher extinction coefficient and wider emission spectra within near-infrared region, only QD 800 signals were capable to be imaged in the animals for up to 14 days. We observed an increase in signal intensity when using a red shifted excitation laser (640 nm) to image QD 800 labeled ES cells. The normal excitation wavelength is 465 nm. This was somewhat surprising since the excitation coefficient of QD 800 is lower at 640 nm than it is at 465 nm. That is at 640 nm, QD 800 absorbs light with less efficiency than at 465 nm, so less QDs become excited and thus give off lower signal intensities. However, the tissue penetration is much greater at 640 nm. Therefore, labeled cells that would not have been excited at 465 nm could be excited at 640 nm. Thus, these newly excited cells could contribute to the greater signal intensity seen at the detection wavelength of 800 nm.