Aims:Systemically administered bone marrow (BM) mesenchymal stromal cells (MSCs) are potentially efficacious cell-therapies for the treatment of graft-vs-host disease (GVHD) or auto-immune complications. For infused MSCs to successfully resolve immune reactions, effective homing and biodistribution to target tissues are important pre-requisites and the delivered MSCs should also be biologically primed. However, retrospective studies have now demonstrated that MSCs, when ex vivoexpanded to reach numbers relevant for human use, are a heterogeneous cell population consisting of subpopulations with different functionalities. The overall non-uniform cellular properties result in unreliable biodistribution and immunosuppressive functions. A systematic study of expanded MSCs to discern useful subpopulations is hindered by the lack of unique surface markers. In lieu of immunophenotype, we hypothesized and verified that biophysical properties of MSCs can be a useful strategy for isolating unique subpopulations. Here, we will describe a clinically relevant biophysical strategy for deriving a primed MSC subpopulation that is equipped with the necessary properties for homing and immune suppression.

Methods:MSCs were derived from the healthy BM of patients aged 21-40 years and culture expanded for ~10-15 population doublings. To study putative MSC subtypes, we employed multivariate biophysical analyses of cell size, stiffness and nuclear fluctuations to identify subpopulations and a label-free microfluidic approach that takes advantage of unique biophysical cell traits was developed for cell sorting. This enabled correlation of these biophysical quantitative measures with biological attributes as well as in vitroand in vivofunctionality. For this study, the different biophysically derived MSC subpopulations were subjected to functional testing for tissue homing, engraftment as well as immune suppression.

Results:A multivariate analysis of MSCs using cell diameter, membrane stiffness and nuclear membrane fluctuations is predictive of at least four subpopulations in expanded MSCs: uni-potent osteogenic progenitors, bi-potent osteochondro progenitors, uni-potent adipogenic progenitors and tri-potent progenitors. These subpopulations are distinct in their transcriptome and differentiation potential but no surface marker reliably distinguishes them apart. Importantly, the tri-potent subpopulation possessed higher transcript and protein levels of alpha4, cx3cr1, cxcr4 and podxl, compared to the other subpopulations (~2X to ~4X). In functional assays for homing and engraftment, tri-potent progenitors showed increased chemotaxis towards SDF1a(31-43% increase) and secretome from commercial cancer cell lines (MDA-MB-231/435, 26-31% increase) as well as TNFastimulated HUVECs (27-55% increase). In MSC rolling in vitroassays on TNFastimulated HUVECs, the tri-potent progenitors exhibited reduced velocities (113±45 mm/s vs >350 mm/s compared to other subpopulations), indicating a greater level of interactions with the stimulated HUVECs, and finally leading to a greater number of cell arrest on the HUVEC layer (3-5 fold increase). In vivobiodistribution experiments against subcutaneously implanted MDA-MB-231/435 cancers as well as sub-lethally irradiated mice further demonstrated a greater level of homing and engraftment towards sites of inflammation (eg. implanted tumors, myeloablated BM, gut). In vitro, tri-potent progenitors were also stronger IDO, PGE2and TGFbproducers following stimulation with INFgand was able to suppress PHA stimulated CD8 T-cell proliferation most potently (76±18% suppression vs <48%).

Conclusions: These results demonstrate that heterogeneity in expanded MSCs subpopulations could lead to inconsistencies in their overall properties. Newer biophysical technologies have the potential for isolating useful subpopulations defined by key quality attributes for specific forms of MSC-therapy; these capabilities may advance MSC-based therapy for regenerative medicine. Specifically, we demonstrate a strategy for isolating the MSC subpopulation that is primed towards homing and immune suppression functions. Further validation of this subpopulation against other MSC subpopulations in animal models of GVHD or auto-immune diseases will be necessary for next translational steps.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution