Our study demonstrated that canine MSCs can be effectively labeled by incubation with silica nanoparticles without impairing the cell morphology, viability or metabolic activity. In addition, silica nanoparticle-labeled MSCs exhibited a similar level of differentiation into adipocytes and osteocytes compared with that of unlabeled MSCs. Moreover, silica nanoparticle-labeled MSCs were uniformly visualized in the kidney at more than 3 weeks after transplantation.
These results demonstrate obvious benefits compared with other nanomaterials. Recently developed quantum dots show strong luminance and high photostability compared with those of conventional organic fluorophores . However, the toxic potential of cadmium nanoparticles has remained a concern . Boddington et al. reported  that the iron oxide nanoparticle ferucarbotran may impair MSC differentiation, which hinders their potential use as fluorescence tracers for stem cells. In contrast, the toxicological effects of amorphous silica have been characterized reasonably well, and this form has been found to be less toxic in animal experiments than that of other nanomaterials. Although some studies suggest that inhaling of materials such as colloidal silica by mice induces dose and size-dependent pulmonary effects [22, 23], silica nanoparticles demonstrate no apparent toxicity in mice when injected via the tail vein . Similarly, the safety of silica nanoparticles was confirmed in our study. The viability of unlabeled and silica nanoparticles-labeled cells, as measured by the WST-1 assay and trypan blue exclusion, was not significantly different, suggesting that silica nanoparticles are cytocompatible and do not induce cell death. Moreover, the differentiation capacity of MSCs toward multiple cell lineages was not affected by silica nanoparticle labeling. Labeled MSCs showed similar elaboration of mineralized matrix and accumulation of lipid droplets, compared with those of unlabeled MSCs during osteogenic and adipogenic differentiation, respectively.
We also demonstrated that silica nanoparticle-labeled MSCs can be histologically located in the kidney and tracked. Endocytosed silica nanoparticles persisted in expanded canine MSCs for up to 3 weeks. This 3 week period was sufficient for transplanted MSCs to migrate and differentiate in the host tissue. Furthermore, the fluorescence was not weakened until 3 weeks after transplantation. In contrast, superparamagnetic iron oxide nanoparticles show significant decreases in fluorescence after 1 month . Such long-term labeling of stem cells will be beneficial for elucidating the relative contributions of native and engineered tissues to morphogenesis. Furthermore, silica nanoparticles can be imaged in renal tissue without additional staining. The kidney is a highly autofluorescent tissue owing to its extremely high metabolic rate, accumulation of flavins and lipofuscins, and vascularization . Tissue autofluorescence has long posed a problem for studies of immunofluorescence labeling, and particularly for direct staining. Because it has been difficult to reduce autofluorescence in tissues, efficient labeling methods should be used to overcome this issue. We demonstrated that silica nanoparticles can be used as a reliable long-term tracking agent in autofluorescent tissues. Thus, we believe that this labeling method can track transplanted cells for extended periods in other autofluorescent tissues (e.g., liver, and pancreas).
We used an invasive method of injecting directly into the renal cortex after celiotomy to infuse stem cells into the kidney. After the abdominal incision surgery, the mice recovered immediately and gained weight normally, and no other complications were observed. Abdominal incisions can mobilize leukocytes and neutrophils to surgical areas, which is usually normalized within 7 days. In addition, we administrated broad-spectrum antibiotics to prevent infection from the surgery. Therefore, we believe that the effect of surgery on the mice and transplanted MSCs was insignificant. Renal tissues exhibited mechanical damage as shown by hemorrhaging and mild necrosis after the injection. The injury site may have been exposed to a number of inflammatory cytokines and chemokines owing to tissue inflammation. This microenvironment may be activated to enhance the homing of stem cells to these sites . In the case of systemic MSC infusion, although some transplanted cells are located in the pathogenic area, most cells are trapped in the liver or enter systemic circulation . In contrast, the detection rate of MSCs following intra-renal injection is much higher (20–50%) than that following systemic injection . Thus, we chose direct renal injection to avoid systemic circulation of labeled MSCs, and most MSCs were observed in the corticomedullary junction and the medullary region at 2 or 3 weeks after injection (Figure 6). The medullary region consists of the proximal and distal tubules and the loop of Henle, in which the main functions involve regulation of water and electrolyte reabsorption. We presume that MSCs gradually moved to the damaged tubular regions and participated in tissue reconstitution. This relationship between MSC migration and their therapeutic effect needs further investigation. In any case, we expect that silica nanoparticle labeling will help elucidate the distribution and fate of transplanted MSCs.
Frangioni et al.  reported the following criteria to show that an ideal agent for tracking stem cells must (1) be biocompatible and safe; (2) not require any genetic modification or perturbation of the stem cells; (3) permit single cell detection in any anatomical location; (4) allow quantification of cell number; (5) have minimal or no dilution following cell division; (6) have minimal or no transfer to non-stem cells; (7) permit non-invasive imaging in the living subject over months to years; and (8) require no injectable contrast agent for visualization. Our study demonstrated that silica nanoparticles satisfied most of these criteria. However, we used only one cell line each for UCBMSCs and ATMSCs. Multiple cell lines will need to be tested with silica nanoparticles because of possible differences among cell lines. Based on our study, we suggest that silica nanoparticles represent a viable approach for labeling and tracking of MSCs.