Culture Conditions for Generating Human Bone Marrow Stromal Cells Influence Cell Immunophenotype and In Vivo Biodistribution in Immune Deficient Mice.

Bone marrow stromal cells (MSCs) show promise for cell and gene therapies but their utility is currently limited by low levels of engraftment. A better understanding of factors underlying cell trafficking in vivo is likely to lead to enhanced MSC engraftment and tissue targeting. To address this issue, we studied the immunophenotype and transplantation potential of MSCs derived from two different culture systems, serially passaged adherent layers from standard long-term bone marrow cultures (LTBMC) and a novel stirred suspension bioreactor culture (SSBC) supplemented with IL-3 and SCF. MSCs were characterized as CD45 (−), VCAM-1 (+), CD44 (+), SH-3 (+), and CD49e (+) by flow cytometry. At earlier time points, the SSBC system generated 1.8x more MSCs compared with the LTBMC system when analyzing the CD45-negative fraction by flow cytometry. By the end of culture, however, the LTBMC system generated 7.3x more cells. Interestingly, 66% of the CD45 (−) cells from the bioreactor system were negative for both HLA Class I and II antigens. Furthermore, flow cytometry revealed that the bioreactor-derived cells expressed very low levels of the cell adhesion markers, VCAM-1 and CD44. These data suggest that the SSBC system generates cells that may have greater migratory freedom in vivo. Biodistribution patterns of the human MSCs derived from the two sources were examined in 11 completely unconditioned six-week old SCID mice. Six weeks after intravenous infusion of MSCs, femoral and tibial bone marrow, lung, liver, bone, spleen, brain, heart, and blood were analyzed for donor human cell engraftment by PCR and fluorescence in situ hybridization (FISH) against the murine background. Engraftment was determined by PCR with primers from the human alpha-satellite region (chromosome 17) that amplify a specific 850 bp fragment. One tibial bone marrow sample, two femoral bone marrow samples, one bone sample, and four lung samples from nine mice receiving LTBMC MSC were positive by PCR. In contrast, all of the same tissues from the mouse receiving SSBC MSC were negative for human donor cells, including the lung, with the intriguing exception of a positive PCR signal in heart tissue. The presence of donor MSC was confirmed in PCR+ specimens by FISH using a human Cy3- and a murine FITC-pan-centromeric probe. The frequencies of donor LTBMC MSC in the tibial bone marrow, femoral bone marrow, bone, and lung were 0.79%, 1.60%, 0.58%, and 1.99%, respectively. Donor bioreactor-derived MSCs accounted for 0.41% of the cardiomyocytes from one recipient, despite the absence of cardiac injury. Compared with LTBMC MSCs, SSBC-derived cells appear to display an unusual biodistribution pattern which may be attributable to the altered immunophenotype. The present study underscores the importance of culture conditions in influencing the immunophenotype of MSCs and holds promise for developing targeted cell therapy.