In vitro analysis? ? We are isolating clonal human salivary cells and expanding them a variety of different media. In low calcium medium, they express both epithelial and mesenchymal markers by both mRNA and protein analyses, suggesting an epithelial to mesenchymal transformation (EMT). When grown in low calcium medium, cells did not express acinar cell-specific mRNAs (aquaporin), but in normal calcium conditions, they expressed both ductal (cytokeratin 19, claudin 1) and acinar mRNA markers. We further analyzed whether growth factors known to be expressed during salivary gland development (TGF-beta, FGF-1, FGF-7 and FGF-10) would further enhance the expression of both ductal and acinar markers in medium containing normal levels of calcium. Our results show that all of the growth factors enhanced expression of the markers compared to medium with normal calcium levels alone, but that there were no major differences between the different growth factors. However, a mesenchymal marker (vimentin), continued to be expressed under all conditions indicating that mesenchymal to epithelial transformation (MET) was not complete. ? ? To further enrich the clonal populations of cells for potential stem/progenitor cells, we isolated CD49f+ (alpha 6 integrin subunit)/CD29+ (beta 1 integrin subunit), which have been postulated to be present on epithelial stem/progenitor cells by FACS, and grew them under low calcium, normal calcium and normal calcium including FGF-10. Analysis of amylase expression by RT-PCR indicated a gradual increase in amylase mRNA with time in culture in medium containing normal levels of calcium, with or without FGF-10 compared to medium containing low levels of calcium. ELISA of amylase expression showed more significant differences between the low calcium and the normal calcium conditions. ? ? Clonal populations of cells grown in low calcium and subsequently plated in 50% Matrigel with normal calcium levels and FGF-10 form aggregates and begin to develop globular outgrowths, somewhat reminiscent of branching morphogenesis. We have now analyzed the expression of ductal and acinar markers by RT-PCR and immunohistochemistry. The aggregates expressed cytokeratin 19, claudin 1, claudin 3 and aquaporin 5, but unlike in the monolayer cultures, vimentin was not expressed, indicating a more complete mesenchymal to epithelial transformation. However, by histology, the aggregates lacked organization into distinct ductal and acinar regions, and amylase expression was not detectable, suggesting that while culture in Matrigel with FGF-10 induced MET, the culture conditions are not supportive of appropriate organization of the cells. ? ? In vivo analysis? ? In order to provide the cells with more conducive conditions to differentiate into appropriate structures, clonal cells were transplanted in vivo subcutaneously. Previously, we utilized Matrigel as a carrier for the cells, and found that small ductal structures were formed, but by immunohistochemistry, the expression of amylase was found to be interior to the ductal structures, indicating that the correct organization of the transplanted cells was lacking. We utilized poly glycolic acid polymer (PGA) in conjunction with Matrigel, and observed better organization of the cells, with ductal markers being interior to the clusters, and acinar markers (aquaporin) being exterior in the clusters. ? ? While the subcutaneous transplants provide the opportunity to study organization of the transplanted cells, they cannot be used to demonstrate functionality. For this reason, we have transplanted the cell/PGA/Matrigel constructs into damaged submandibular salivary glands. In this situation, structures similar to what was observed in the subcutaneous transplants were formed. The transplanted salivary glands are now being analyzed for the expression of ductal and acinar markers, and for the expression of human specific proteins. Interestingly, preliminary results suggest that human cells are found not only in the cell/PGA/Matrigel portion of the gland, but also within the murine portion of the gland. Furthermore, we are trying to determine if histatin, a protein that is not present in murine saliva, but is present in human saliva, is formed by the transplanted salivary gland. This would be proof of function. ? ? Identification of salivary stem cells in situ? ? To label a putative stem cell (slowly-dividing, label retaining cell) in vivo, we injected BrdU into young-adult mice intraperitoneally for 18 days. This extended series of injections is needed to increase the chance that the stem cell, which rarely divides, will be labeled. On the last day of injection, and then at 0.5, 1, 2, 4, 6, and 7 months later, submandibular salivary glands were harvested in order to watch the disappearance of labeled cells (differentiated cells and transiently dividing progenitor cells) and to identify the location of the label-retaining cell, which would be a putative salivary stem cell.? ? We found that from 7 days to 100 days, label-retaining cells range from 70% to 30% of the total number of cells respectively. But after 152 days, the label-retaining cells are less than 3% of the total cells. Label-retaining cells are not only ductal cells but also acinar cells, and a few label-retaining cells are non-ductal, non-acinar, and non-myoepithelial in nature. After 186 days, the label-retaining cells have disappeared from the salivary gland. These results suggest that a subset of acinar cells, striated and granular duct cells, and a cell type of unknown identity are slowly dividing, label-retaining cells that serve as reserve cell pools, perhaps by self-renewal. We are currently trying to determine if the label-retaining cells express CD49f and CD29. ? ? After injury, it is thought that a stem cell will divide immediately in order to initiate regeneration. In current studies, we are creating injury by a small incision into the gland, by ligation, and by cannulation of the gland with injection of ethionine, followed by immediate injection of BrdU. From these studies, we hope to determine which of the label-retaining cells identified in the previous study will begin to divide and initiate repair.
