Dissociation of In Vivo and in Vitro Differentiation Capacity of Human Mesenchymal Stem Cells Is Reflected by a Tissue Specific Epigenetic Memory

Abstract 2386

Human mesenchymal stem and progenitor cells (MSPCs) from various tissues are currently evaluated in clinical trials for bone and marrow regeneration and their immune modulation potential. MSPCs from virtually all tissues appear indistinguishable regarding immune phenotype and their multipotent differentiation capacity in vitro. Improvement of so far limited clinical efficiency is hampered by a lack of understanding MSPC functionality in vivo. Here we demonstrate that the capacity of in vivo endochondral bone formation followed by establishment of a hematopoietic niche through infiltration of blood producing hematopoietic components function as a surrogate to determine in vivo multipotentiality (differentiation into more than 3 mesodermal lineages) of isolated MSPCs.

MSPCs from bone marrow (BM), adipose tissue (AT) and umbilical cord (UC) have been isolated by plastic adherence and were propagated under humanized culture conditions using pooled human platelet lysate (pHPL) as previously described. Human skin fibroblast (Fibs) derived under identical culture conditions where used as control throughout the study. Comparative analyses of surface immune phenotype, adipo-, chondro- and osteogenic differentiation potential in vitro as well as expression analysis of key mesenchymal lineage genes were performed. Epigenetic profiling of MSPCs from different tissues was done using a methylation array including CpG-islands in- and outside of coding regions, CpG-shores and non CpG sides (450K array; Illumina). In vivo differentiation capacity was tested by using two million of MSPCs transplanted subcutaneously into immune-deficient NSG mice. The developmental sequence of chondro- and osteogenic as compared to perivascular mesenchymal tissue formation was analyzed using histology and immune histochemistry. In vivo near infrared (NIR) fluorescence imaging and micro computed tomography (microCT) was used to study bone development. Formation of a human MSPC-derived marrow niche with establishment of the complete host hematopoiesis was studied in situ and by polychromatic flow cytometry. Secondary transplants of MSPCs isolated from primary marrow organs were performed and analyzed equally.

MSPCs from all tissues analyzed and control Fibs show an almost identical immune phenotype using a classical MSPC marker profile. Osteo- and adipogenic differentiation potential in vitro as well as gene expression did not distinguish tissue-specific MSPCs. Using a stringent 3D chondrogenesis assay and appropriate histological stainings of synthesized chondrogenic matrix proteins (Safranin O, Alcian Blue, Toluidin Blue) MSPCs from all tissues except BM failed to form cartilage in vitro. In vivo mouse studies could further strengthen these findings, because BM-derived MSPCs were the only cell type capable of generating ectopic bone through an endochondral ossification process. Bone formation was followed by mouse marrow infiltration including megakaryocytes as well as lineage negative, sca-1 positive, c-kit positive (LSK) hematopoietic stem and progenitor cells (HSPCs). These results correlated with the epigenetic status of the cells. Comparing their methylation profile using principal component analysis (PCA), BM-MSPCs cluster separately, whereas MSPCs from all other tissues cluster together.

In this study BM was the only tissue containing MSPCs with multipotent differentiation capacity including chondrogenesis, osteogenesis and hematopoietic niche formation. This is reflected by a BM specific epigenetic profile that differs from that of all other tissues analyzed. Since cartilage formation is one initial developmental process important for bone generation and hematopoiesis attraction, an epigenetic predisposition of BM MSPCs to undergo endochondral ossification seems to make these cells unique for bone and marrow regeneration purposes. More stringent test systems including consequent in vivo potency assays should be mandatory in context of clinical studies. Our data may argue against the application of non BM-MSPCs for bone and marrow regeneration in patients in the absence of experimental in vivo evidence.