Isolation of the Stromal/Vascular Fraction of Murine Bone Marrow Dramatically Enhances MSC Yield, Allows Isolation of Marrow Endothelial Cells and Reveals Multiple Subpopulations of Stromal Cells.

Abstract 3861

The bone marrow is the organ of residence of a population of multipotent progenitor cells most commonly referred to as mesenchymal stem cells (MSC) based upon their multilineage differentiation potential into bone, cartilage and adipose tissue. The capacity for MSC to contribute to tissue repair demonstrated by numerous previous reports has engendered considerable interest in their application to a broad range of cellular therapies. It follows that a robust reproducible methodology for obtaining high yields of MSC from preclinical animal models, such as rodents, would greatly facilitate the development of these various MSC-based cellular therapies. The plastic-adherent, clonogenic progenitors termed colony forming unit-fibroblast (CFU-F) originally identified by Freidenstein and colleagues that initiate MSC cultures are a rare population in the marrow of all mammalian species so far examined. This is particularly so in the case of the mouse where reported incidences of CFU-F are typically in the range of 1/200,000 bone marrow BM cells. The low incidence of CFU-F significantly complicates the isolation of homogeneous populations of MSC from mouse BM, a common problem being contamination with hematopoietic cells. Seeking to develop an improved methodology to harvest MSC from mouse using methods based on plastic adherent bone marrow, we took advantage of burgeoning evidence demonstrating the perivascular location of MSC not only in the bone marrow, but also in multiple tissues. We hypothesized that a potential reason for the low yield of mMSC from mBM is the flushing of the marrow used to remove single cells suspensions and the consequent destruction of the marrow vasculature, which may adversely affect recovery of MSC physically associated with the abluminal surface of blood vessels. Herein, we describe a simple methodology based on preparation of intact marrow plugs that yields distinct populations of both stromal and endothelial cells. BM plugs are subjected to 3 sequential rounds of digestion in collagenase/dispase and each fraction assayed for content of CFU-F. The recovery of CFU-F obtained by pooling the product of each digestion (1643+199) reproducibly exceeds that obtained using the standard BM flushing technique (13.3+1.9) by at least 2 orders of magnitude (P=<0.001; N = 8) with an accompanying 196-fold enrichment of CFU-F frequency. Purified BM stromal cell populations devoid of hematopoietic contamination are readily obtained by FACS at P0 and these demonstrate robust multilineage differentiation into bone, adipose and chondrogenic progeny using standard in vitro bioassays. A detailed immunophenotypic analysis of the P0 cultures demonstrated the existence of multiple stromal cell subpopulations many of the markers analyzed, including Sca-1, CD90, CD105, CD146 and PDGFRa, which was progressively lost with serial passaging. Discrete subpopulations of stromal cells identified at P0, in many cases had phenotypically identical counterparts in the BM cell suspensions prepared by serial digestion and we are in the process of quantitatively analyzing the evolution of selected phenotypes in vitro to provide clues as to the identity of the founder population of stromal cells that gives rise to ‘MSC' in vitro. Finally, the phenotypic analysis of P0 cultures also revealed a discrete population of CD105^BrightPDGFRa^Neg cells representing a mean of 26.7% of hematopoietic lineage-negative cells. Upon isolation and serial propagation, the cells maintain expression of all of the vascular endothelial markers examined including CD31, CD105, VCAM-1, CD144 and MECA32 and also demonstrate inducible expression of E-selectin upon treatment with TNF-a. In conclusion, we describe a simple and robust methodology that, for the first time, allows the simultaneous isolation of both the stromal and vascular components of mouse BM. Secondly, the yield of ‘MSC' afforded by this technique far exceeds that reported in any previous study. Thirdly, this technique reveals a level of stromal cell heterogeneity not apparent in previous analyses of mouse BM-derived MSC that more realistically reflects the likely complexity of stromal cell populations in vivo and represents a platform for the eventual prospective isolation of specific subpopulations. These studies will greatly enhance experimental strategies designed to analyze not only MSC identity but also the function of the vascular hematopoietic niche.