Study On Multipotential Differentiation Of Human Mesenchymal Stem Cells From Bone Marrow

Human mesenchymal stem cells (hMSCs) are non-hematopoietic progenitor cells, capable of differentiating into mature cells of multiple mesenchymal tissues including fat, bone, cartilage, muscle. The hMSCs isolated from bone marrow are also called marrow stromal cells. In recent years, the multiple differentiation potentials of hMSCs coupled with the ease of in vitro culturing has spurred considerable interests in better understanding the biology of hMSCs and their potential therapeutic applications.Despite these increasing interests, compared with hematopoietic stem cells (HSCs), our understanding of hMSCs biology remains rudimentary. In fact, hMSCs maintained in different laboratories or MSCs isolated from different species are various in their characteristics. There are many factors confounding the understanding of the characteristics of hMSCs and their differentiation potentials. First, during the individual development, hMSCs have different phenotypes and functions. Second, there are subpopulations of MSCs in bone marrow, differing by the stage of their differentiation. Third, MSCs at different sites in the bone marrow may be different. Fourth, different isolating methods and culture conditions may affect the characteristics of the MSCs maintained. Finally, the induction paradigm used in vitro and in vivo may affect their differentiation potentials. To date, there is no specific marker or combination of markers to identify hMSCs either in vivo or in vitro, and there is no standardized procedure to isolate hMSCs. Friedenste AJ, 1970 and Gronthos S, 2003 observed that single-bone marrow stromal cell adherent to plastic culture dish can form colony of cells morphologically resembling fibroblasts in vitro, called clonogenic progenitor cells or colony-forming fibroblastic units (CFU-F). They further demonstrated that the in vitro growth of different adherent colonies are heterogeneous and only cells from some colonies may be multi-potential hMSCs. However, like all the other current isolation procedure, this method of isolation yields a heterogeneous cell population. These plastic adherent cells exhibit widely varying growth kinetics, cell markers and differentiation capabilities.Under these premises, we initiated our study by establishing clonal hMSCs celllines from these CFU-F, and uncovering culture conditions to keep hMSCs at undifferentiation state. We then went to characterize these hMSCs cell lines and identify clones for further investigation. Finally, we evaluated the multipotential differentiation capability of this hMSCs into hepatic and neural lineage in vivo and in vitro.hMSCs were isolated from healthy human bone marrow by first taking advantage of their preferential adherence to the plastic surface of the culture dish. Then, individual clones of clonogenic cells which are rapidly adherent and larger in size were picked with sterile pipette tips under microscope. Clonal cell lines were established after expansion in vitro. These cell lines were systematically characterized and their differentiation capabilities were evaluated. One cell line that has the classical characteristics of hMSCs and was able to differentiate into heptocyte-like cells and neuron-like cells in vitro was further analyzed by evaluating their fates in injured liver and developmental brain. The main results are as followed:(1) Some isolated clonal cell lines have basic features of mesenchymal stem cells. Under phase-contrast microscope, these cells are homogeneously small spindle-shaped and little broader cells were seen among them. Growth curve was drawed at passages 8 and the population doubling( PD ) time was 29.5+2.3h by calculation. Several lines of evidence suggest these cells are hMSCs. First, even after in vitro culture over 60PD, these cells have the capacity for self-renewal. Second, they maintain the ability to give rise to three different types of differentiated mesenchymal lineage progeny (osteogenic cells, adiopogenic cells and chondrogenic cells). In addition, these cells have normal karyotype, diploid DNA content. No tumors were formed for 3 months after they were injected subcutaneously into SCID mice. Therefore, we will call them hMSCs thereafter.(2) hMSCs expressed various markers for different lineage cells including mesenchymal cells, but not hemopoietic stem cells. By using RT-PCR, we show that these cells express the following set of gene markers: OCT-4, SDF-1 a , CD49a, CK19, and c-met. By using Fluorescence Activated Cell Sorting (FACS), we show that these cells express the following cell surface antigens: CD44, CD29, CD90, HLA-1, but not CD34, CD45, CD14, and HLA-DR. These results showed that hMSCs do not expressmarkers of hemopoietic stem cells(CD34, CD45) and HLA-1, either, but do express express markers of mesenchymal stem cells(CD44> CD29> CD90> CD49a ^ SDF-1 a ) and OCT-4.(3) hMSCs have obvious proteome changes after they differentiate into adipocytes. Proteome of the hMSCs before and after differentiation were analyzed by 2D SDS-PAGE. Comparative analysis of the protein spots pattern revealed substantial changes of the proteome composition during the differentiation process. Because of limited amount of protein used for the analysis, subsequent MALDI-TOF analysis only identified the actin and tubulin, which decrease sharply during differentiation. However, this preliminary study demonstrates the feasibility of using the hMSCs to characterize the mesenchymal stem cell proteome and as a model to investigate the mechanisms of differentiation.(4) In vitro differentiation study showed that hMSCs can differentiate into neural-like cells and hepatic-like cells. To induce hMSCs into neural differentiation, we use 5rr,M P 2 mercaptoethanol, or 0.2mM butylated hydroxyanisole/2% Dimethyl Sulfoxide or forskolin. After 5 hours of induction, fibroblast-like hMSCs start to exhibit neuron-like morphology, extending long processes terminating in typical growth cones and filopodia. Further immunocytochemical analysis revealed that most of differentiated cells express many neuron makers: neuron-specific enolase, neurofilament-M, and tau. However, it is difficult to maintain the viability of the cells after induction. Majority of cells were dead after 3-5d induction, no longer than 7days. To induce hMSCs into hepatic differentiation, we use FGF-4, EGF, HGF, OSM, 2%FBS in IMDM/F12 medium. After 21 days induction, the shape of hMSCs changes from spindle-shape to cuboid-shape with smaller nuclear to cytoplasm ratio. Binuclear cells could be observed. Immunocytochemistry and RT-PCR analysis showed that cells express heptic cell specific markers: while small percentage of differentiated cells expresses CK19 and CK18, some of differentiated cells express albumin, a1-antitrypsin, Periodic acidishiff (PAS) staining further revealed abundant glycogen stores in cytoplasm of some of the differentiated cells, which is a characteristic of hepatic cells. In conclusion, the clonal cell line of hMSCs candifferentiate into neuron-like and hepatic-like cells in vitro.(5) In vivo differentiation study showed that hMSCs can differentiate into neuron cells in neonatal mouse brain. We further analyzed the hMSCs differentiation potentials in vivo. hMSCs were marked genetically with enhanced green fluorescent protein(EGFP) by retroviral methods and were subsequently injected into the lateral ventricle of neonatal mice brain. At 0 day, 9 days and 14d post-injection, mice were killed and their brains were fixed with 4% paraformaldehyde. Examination of cryostat section by fluorescent microscopy revealed that hMSCs can engraft into brain, and migrate from injection site to striatum, hippocampus, the cortex, even areas of brain contralateral to the side of injection. Neuron-specific differentiation and engraftment was confirmed by examining the expression of neuron-specific protein markers of these EGFP-expressing hMSCs. By using confocal microscopy, we were able to co-localize transplanted EGFP-labled hMSCs have ï¿¡ -Tublin IIK MAP2^ Tau staining although some have GFAP staining. These results suggest that these hMSCs can respond to brain microenvironment and differentiate into neurons and astrocytes.(6) In vivo differentiation study showed that hMSCs can differentiate into hepatocytes in CCI4 injuried SCID mouse liver. hMSCs labeled with EGFP were also surgically injected into the spleen of recipient SCID mice treated by carbon tetrachloride. Cryostat sections of recipient liver were observed under fluorescent and confocal microscopy at different time point. Transplanted cell clusters were detected by EGFP fluorescence at 2, 5 and 8 weeks after cell transplantation. The shape of cells changed from fibroblast-like into hepatocyte-like. Some of them become bi-nucleated. Clusters were observed in both central vein and portal areas. Statistical results indicated that clusters with 2 cells or more constituted the majority (79.4%) of cell repopulation at 2 weeks post-transplantation. These data demonstrated that transplanted cells can engraft long time (observed 8 weeks after cell transplantation) and most of survival cells have proliferated in the injuried liver 2 weeks posttransplantation. Furthermore, immuno-colocalization studies demonstrate that the transplanted cells express hepatocyte specific protein markers: albumin and a1-antitrypsin in injuried SCID mice...