Mesenchymal stem cells (MSC) are currently being tested as treatment for a variety of diseases. We have little information about the changes that occur during isolation and expansion, and the ultimate disposition of MSC in the body. We have shown that MSC isolated from the bone marrow of mice and expanded in culture undergo spontaneous mutations within oncogenes yet retain tri-lineage differentiation potential. Here we define the surface marker expression of these cells and compare it to the marker expression of previously isolated and characterized cultured pluripotent cells. Culture expanded spontaneously transformed MSC (stMSC) infused into immune competent mice can be recovered from all peripheral tissue as endoderm, ectoderm and mesoderm lineages. Both direct transdifferentiation of cells to organ appropriate lineages and fusion with host cells occurs and cellular function is maintained. stMSC have a competitive advantage over native MSC and freshly isolated MSC resulting in higher levels of engraftment with stMSC. stMSC injected into blastocysts contribute to all cellular lineages including the hematopoietic system, however long term hematopoietic function is not achieved: aplastic anemia and marrow failure occur after 14 weeks. These findings suggest that culture expansion of cells has the potential to introduce genetic mutations which have long term consequences. Infused MSC may be long lived in the body, and reside in a wide variety of tissues however long term function may not persist. Use of these cells for clinical therapy must be done with caution as the long term consequences in humans is not yet known.
Mesenchymal stem cells (MSC) are a subset of cells within the mesenchymal pool which when subcultured as single cell clones, are capable of differentiating into bone, cartilage, and fat. A precise molecular definition for the pluripotent cell is lacking, and cells are usually defined retrospectively based on their behavior in culture . Nevertheless, these cells have been touted as therapeutic breakthroughs for a wide variety of clinical conditions (reviewed ). Interestingly, it has recently been shown that the MSC phenotype may not be stable, and cells may alter their molecular programming when in culture to become more or less pluripotent, depending upon culture conditions and cell density . In addition, several laboratories have reported that MSC isolated from the bone marrow and expanded in culture are capable of multipotent differentiation beyond the confines of mesenchymal lineage restrictions [4,5]. These findings have been challenged, however, and remain controversial. The ability to differentiate across multiple lineage barriers calls into question the usual dogma of adult stem cell differentiation capacity, whereby cells of mesoderm, endoderm and ectoderm lineages arise from unique independent stem cells in the adult. Additionally, conventional dogma holds that stem cells within each of these groups are only capable of limited differentiation potential. It is unclear if the cultured cells reported to have this multipotent phenotype are found normally in the marrow or if this is a culture induced property. However, several laboratories have reported this phenomenon with cells cultured under different conditions and these cell populations possess unique patterns of surface markers [4,5] suggesting that this is not an isolated occurrence, and these findings may represent different populations of cells with similar potential or the effects of different isolation and culture conditions.
The implications for a circulating or culture derived multipotent MSC are broad. The presence of a multipotent stem cell in circulation raises the notion of mobilizing and exploiting these cells for therapies aimed at tissue repair [6-8]. Culture-based isolation or culture expansion of cells could provide additional sources of cells for therapeutics while avoiding the present method of iPS creation using genetic manipulation. Widespread (though low level) contribution of these cells to peripheral tissues, however, raises the important concern of their contribution to cancer if genetic damage occurs to the multipotent cells in vivo, or during culture in vitro. Our laboratory [9,10] and others [11-17] have shown that bone marrow-derived cells are found as either a major component of epithelial cancer, or as a supporting component of both the tumor and the tumor stroma in various malignancies, thus substantiating a role for these cells in human malignancy and supporting the notion of multilineage differentiation of MSC in vivo. It is not known if the circulating stem cells acquire genetic mutations within the peripheral tissue they ultimately reside in, or if they arrive at the tissue with mutations already in place.
We have isolated multiple MSC cell lines from the bone marrow mesenchymal cells of mice and expanded them as single cell clones. These cells are spontaneously transformed (MSC) ) and form fibrosarcomas if injected subcutaneously in mice, however, when they are infused intravenously, they do not adversely affect the mouse and they can be recovered from the bone marrow in a viable state up to two years later . Here we show infused cultured stMSCs are found within the marrow cavity, and also circulate freely, and contribute to all the tissues of the body during the life of the mouse. The level of engraftment into peripheral tissue varies widely between organs, between mice and increases with time, suggesting to us that marrow-derived circulating cells continually populate peripheral organs. In the absence of tissue injury or inflammation, there is no adverse effect on the mouse due to the presence of these cells, suggesting that transformed MSC with multipotent ability can be found widespread in the body, yet they do not form tumors or cause disease in the absence of additional factors.
To further assess the differentiation potential of the cells, we tested their ability to contribute to the developing embryo and found they are able to contribute to all tissue types- and sustain tissue function through gestation to at least several weeks of age. Additionally, germline transmission was obtained, and though not viable, late-term embryos were recovered and verified to have tissues which were stMSC derived. These studies confirm the multipotent potential of culture expanded marrow-derived stem cell, and raise caution for culture expansion of MSC for therapeutic use.
C57BL/6 MSC were isolated and cultured as previously described . Selected spontaneously transformed MSC lines (stMSC) were stably transfected with the plasmid pDS-Redmonomer-hyg-C1 or infection with lentivirus FUGW vector containing a EGFP sequence driven by the ubiquitin promoter. RFP or GFP expression was verified by RT-PCR and FACS, and single cell clones isolated. Cells were at 70% confluence at the time of harvest for injection.
IV Cell Injection
All animal work was performed at the University of Massachusetts Medical School under IACUC approval. stMSC from culture were collected, washed and re-suspended in PBS. 4-6 week old C57BL/6 mice were restrained in an approved lucite box, tails warmed under heat lamps and injected with 1 x 106 RFP labeled stMSC via tail vein. Mice were returned to the cage, allowed free access to food and water and euthanized at predetermined time points.
For identification of macrophages/Kupffer cells, mice were given 0.1 ml of a 1:5 dilution of India ink (speedball USA) in PBS via tail vein and euthanized 1 hour later
Peripheral Organ Examination
At 12 months of age, mice were euthanized and necropsy performed. Organs were collected and processed for sectioning. H&E, direct fluorescence microscopy and IHC for tissue-specific markers combined with antibody targeting RFP or GFP protein were performed according to standard protocol. Rabbit polyclonal antibody directed against Troponin I(H-170), IQGAP1(H-109), (Santa Cruz Biotechnology), F4/80, Amylase , RFP (Biotin conjugated) (Abcam), GFP(Biotin conjugated) Insulin, (Cell Signaling Technology) RFP (Rockland) GFP (Invitrogen) and detected using goat anti-rabbit –IgG—alkaline phosphatase (Sigma) or Goat anti-Biotin HRP conjugated antiTwenty albino B6 blastocysts were injected with 10 cultured GFP labeled MSC each and green fluorescence confirmed under the dissecting microscope prior to injection into pseudopregnant dams according to standard protocol of the UMASS mouse transgenic core facility. All pregnant mice delivered the expected number of pups, which were phenotypically normal at birth. Mice were euthanized at 7 or 14 weeks of age or mated to C57BL/6 wild type mice. Blood was drawn at 7 weeks of age, and prior to euthanasia at 14 weeks for the older group and analyzed by FACS for GFP expression. Those samples positive for GFP expression were further analyzed with fluorescent antibodies directed against lineage-specific surface markers as follows: anti-CD3-PE or APC, anti-CD8-PE or APC, anti-CD4- APC.
At necropsy, all organs were collected, prepared and analyzed by direct fluorescence microscopy, H & E staining, IHC directed against GFP and for GFP expression by PCR. PCR was performed using standard protocols and the following forward and reverse primers on 0.1 ugs of genomic DNA : GAPDH: 5’-ACCCAGAAGACTGTGGATGG-3’, 5’-ACACATTGGGGGTAGGAACA-3’ and GFP: 5’-TACCCCGACCACATGAAGCA-3’, 5’-TGCTTGTCGGCCATGATATAGA-3’ and 5’-CACATGAAGCAGCACGACTT-3’, 5’-TGCTCAGGTAGTGGTTGTCG-3’. Two unique GFP primer sets were used for each sample. GFP labeled MSC (termed A3G) used for injection served as a positive control at a dilution of 1:10 through 1:10,000.
Two chimeric male mice were mated to wild type C57BL/6 female mice, and one pregnancy resulted. The mouse was euthanized due to dystocia, pups removed, tissue blocked and analyzed by direct fluorescence microscopy and IHC directed against GFP and lineage-specific targets.
Cultured, Genetically Transformed Circulating MSC Contribute to all Adult Tissue in the Mouse
We have reported on single cell clones isolated from long term cultures of bone marrow-derived MSC . Cultured MSC transforms at a high rate [18-21] and acquire a pattern of gene expression and gene mutations which are similar to those found in human cancers . Our stMSC clonal cell lines have clinically relevant p53 mutations and form fibrosarcoma when injected subcutaneously, but not when injected intravenously . Here we define the surface marker expression and compare it to the published receptor expression found on the marrow-isolated adult multi-lineage inducible cells (MIAMI)  and multipotent adult progenitor cells (MAPC) [22,23] cells, and find our cells carry a unique pattern of receptors (Table 1) yet retain their ability to differentiate down the classic MSC pathways of bone, fat, and cartilage . The surface marker expression of our cells was similar in several clonal cell lines tested.
Table1. Comparison of Expression Signature of Distinct Multipotent MSC Populations
We injected 1 x 106 stMSC labeled with ds-Red into non-irradiated C57BL/6 mice via tail vein. FACS analysis of peripheral blood samples shows that stMSC circulates in the peripheral blood of mice 7 weeks (Figure 1), 6 months and one year after injection (0.19% in control, and 0.45- 1.54% in mice receiving iv injected cells; eight different mice shown).
The percent of circulating cells remained consistent over the year time (repeat data not shown). The circulating MSC did not express any leukocyte-specific markers and lacked epithelial markers (data not shown).
To test the extent that stMSC contribute to peripheral organs in adult mice, we injected 1 x 106 red fluorescently labeled cells into adult C57BL/6 or GFP- C57BL/6 mice. At 4 months, wide tissue distribution of RFP was seen using the IVIS animal imaging system in live animals (Fig 2A- control mouse; left, RFP-stMSC injected mouse; right). Mice were allowed free access to food and water and received no other intervention until they were euthanized at 12 months post stMSC infusion. Comparison between experimental and control mice showed no difference in weight, activity or general well being. AST, ALT, albumin, creatinine, BUN and peripheral glucose levels were normal in all mice confirming the preservation of organ function (data not shown).
Examination of peripheral organs in the C57BL/6 mice injected with RFP-stMSC revealed grossly fluorescent kidney, liver (shown in Fig 2B) and heart. Fluorescence in other organs was not seen on the gross level but was identified within peripheral tissues (Table 2) by a combination of anti-RFP IHC, direct fluorescence and PCR (tabulated in Table 2).
The level of engraftment varied widely between organs, and between mice, and not all mice had engraftment into all tissue. The level of engraftment seen with culture transformed cells however, was significantly higher than seen with freshly isolated MSC or short term-culture expanded MSC ( [24-30] and personal observation).
Next, we verified the identity of the fluorescent cells within individual organs using tissue-specific immunostaining combined with anti-RFP IHC. Additionally, we evaluated organs from GFP positive mice infused with RFP labeled stMSC in order to determine if the fusion of MSC with the peripheral host tissue had occurred (Figure 2C-L).
Kidney tubules were either completely derived from stMSC or devoid of RFP expressing cells (Fig 2 C, brown stain- RFP). We rarely saw tubules which were of a mixed population of RFP (+) and RFP (–) cells. Aquaporin (AQP-1) is an integral membrane protein expressed on subsets of renal tubule cells . Co-staining with anti AQP1 clearly labels RFP positive cells. (Figure 2C). In contrast, few renal glomeruli contained RFP expressing cells (table 1 and Figure 2C). Examination of kidneys from the GFP host, demonstrates RFP positive endothelial cells within rare glomeruli, juxtaposed with GFP expressing host cells (Figure 2H). Examination of tissue sections, as well as enzymatically disaggregated cells and cell clusters from the kidney, show renal tubule cells often expressed both GFP and RFP and when they did, contained two nuclei. Manually disaggregated kidneys formed a monolayer of binucleate cells in culture which expressed both RFP and GFP, confirming fusion (Figure 2M,N. Arrow highlighting two nuclei). In contrast, endothelial cells within the kidney which expressed RFP did not express GFP and contained one nucleus supporting direct differentiation of stMSC to endothelial cells. These findings in the kidney support both direct transdifferentiation and fusion with host cells.
Skeletal muscle contained RFP positive cells detected by anti-RFP antibody (Figure 2D- brown stain) which co-stained with anti-tropomyosin-1 antibody (Figure 2D, blue stain). Red fluorescence was seen by direct fluorescence microscopy of frozen tissue sections (Figure 2I) Muscle from GFP positive mice infused with RFP labeled cells fluoresced red or green, without coexpression suggesting fusion with host cells did not occur in skeletal muscle (data not shown)
Entire pancreatic islet cell clusters were derived from RFP expressing cells. Co-staining with an antibody directed against insulin confirmed insulin production by these cells (Figure 2 E). These clusters were present in all mice examined, and constituted up to 10% of islets. Pancreatic acinar cells were only rarely found to express RFP protein. A single isolated acinar cell is shown in a GFP host mouse, with the surrounding pancreatic cells expressing abundant GFP (Figure 2J).
Figure 1. stMSC are found in the peripheral circulation of mice after iv injection. Intravenous injection of 1 x 106 stMSC expressing RFP are found in the peripheral circulation at 7 weeks post-injection. A.) ungated cells, B) wild type control blood, C)- J) experimental mice injected with RFP expressing cells.
The gastrointestinal tract had few, rare epithelial cells which expressed RFP, however numerous RFP expressing cells were found in the lamina propria and muscularis mucosa within the fibroblast and smooth muscle populations of all portions of the luminal GI tract (data not shown). Lung tissue contained RFP positive endothelial cells, adipocytes and muscle cells, and rare alveolar cells. The liver contained either rarely isolated hepatocytes or patches of RFP positive hepatocytes which expressed albumin (Figure 2F, RFP- brown, Albumin- Blue). The quantity of RFP expressing cells within the liver varied widely between mice without any obvious reason for the difference.
Examination of livers from GFP positive mice infused with RFP labeled cells support that the majority of RFP expressing cells were fused with GFP expressing hepatocytes, and the cells within the sinusoids expressed RFP alone (Figure 2K and L). Interestingly, Kupffer cells within the sinusoids (identified by avid uptake of carbon particles) were almost entirely derived from stMSC, which were clearly seen on IHC (not shown) and fluoresced vividly on direct fluorescent microscopy (Figure 2G, L).
Table 2. RFP Expressing Cultured MSC is Found in Peripheral Organs of Mice 12 Months After iv Injection.
stMSC were found in circulation at low levels however the phenotype of these cells could not be determined due to the low numbers of cells. In order to determine if stMSC could repopulate the hematopoietic system in adult mice, we took two approaches. Mice were lethally irradiated and transplanted with a mix of GFP+ wild type marrow and an equal number of RFP expressing MSC, or sublethally irradiated and transplanted with only RFP expressing stMSC. In both cases, wild type native marrow, or GFP transplanted wild type marrow cells repopulated the peripheral blood of irradiated mice. RFP positive stMSC did not participate in reconstituting the marrow or the peripheral blood in these experiments (data not shown).
These findings suggest that culture transformed MSC can substantially contribute to peripheral organs as epithelial, endothelial, and mesenchymal cells with the function of the peripheral organs maintained during the life of the mouse, but they do not contribute to the hematopoietic system of the adult mouse.
We next addressed if the MSC could contribute to the formation of viable embryos by injecting 10 GFP labeled stMSC into blastocysts followed by implantation into pseudopregnant mice. Twenty live pups were genotyped by tail clip PCR for GFP sequences and 8 were found to be positive. GFP sequences were confirmed using 2 distinct primer sets and results from two liters are shown (Figure 3). At 14 weeks, mice were euthanized and peripheral organs evaluated. We were able to detect as few as 1:1,000 cells within the tissue (Figure 3B) and wild type control tissue failed to amplify when the reaction was run up to 40 cycles (Figure 3C).
Figure 2. RFP positive stMSC are found in peripheral tissues. C57BL/6 mice infused with MSC expressing RFP examined at 4 and 12 months. A. IVIS whole-body imaging of mice at 4 months. Control mouse- left, injected mouse - right. B. excised livers examined under the fluorescent dissecting scope as labeled. First panel- white light, second panel RFP filter, third panel merged images. IHC of RFP (brown) and tissue-specific proteins (blue) in C. kidney (anti-AQP-1), D. skeletal muscle (anti-tropomyosin-1) E. pancreas (anti-insulin), F. liver (anti albumin). G. uptake of IV injected carbon particles by RFP positive sinusoidal cells of the liver. GFP-C57BL/6 mice infused with cMSC expressing RFP examined at 12 months. H. kidney, I. skeletal muscle, J. pancreas (higher power panel of isolated RFP+ acinar cell). K, L Liver. M, N. Culture of kidney tubule cells. Coexpression of RFP and GFP, and two nuclei in all cells (arrows).
Examination of peripheral organs by PCR for GFP confirmed stMSC contribute to multiple organs including heart, lung, kidney, stomach/small intestine, colon, skeletal muscle, brain, spinal cord, and the ovary/testis, though not all mice had an expression in all of these organs.
Figure 3. GFP expressing stMSC incorporate into developing mouse embryos.
PCR results are shown from two representative mice (Figure 3D, arrows depict the positive results). GFP expression within organs was confirmed using direct fluorescence microscopy and in tissue sections using antibody directed against GFP. Skeletal and heart muscle was co-stained with an antibody directed against tropomyosin. In addition, GFP-expressing sperm was recovered from 3 male mice (Figure 3F). Thirty percent of the mice (6/20) had to circulate GFP positive cells identified in the peripheral blood at 7 (Figure 4A) and 14 weeks (data not shown).
Figure 4. GFP expressing stMSC function as hematopoietic stem cells during embryogenesis.
Despite the continued recovery of GFP labeled circulating nucleated cells during the 14 weeks of the experiment, total RBC and WBC count declined significantly and at the time of necropsy, the marrow within the femur and tibia was white, aplastic and contained few cells. FACS analysis of the GFP labeled cells within the thymus demonstrate these cells are CD3 positive and carry both CD4 and CD8 markers (Figure 4C). Developed for the gestational age of 20 days. By fluorescent dissecting microscope, the pup was GFP positive throughout. Frozen whole mount sections confirmed abundant GFP expression (Figure 5A.). The tissue structure contained bone, muscle and fat and a malformed heart. There was no identifiable GI tract or lung structures. Staining of the whole mount section was positive for muscle (Figure 5B), bone and fibroblasts and negative for epithelial lineages.
The degree of plasticity of MSC both in vivo and in culture has been a hotly debated topic. Safe, reproducible and predictable plasticity is central to the success of efforts to exploit MSC therapy for tissue repair and organ remodeling. MSC are known to undergo mutations both in vivo and in culture systems [18-21] under conditions similar to what has been proposed to expand cells for clinical use, however, the ramifications of these mutations in vivo are not known. Here we show that cultured MSC has a wide degree of differentiation potential in peripheral organs, remain in the tissue for the life of the animal and function as cells appropriate for the organ they are found in. Interestingly, although cells carrying clinically relevant genetic mutations have a significant population advantage over normal MSC- they behave in a benign fashion in vivo. It is not yet known if the addition of pro-carcinogenic environmental stimuli will allow these cells to become frankly tumorigenic.
Several different laboratories have shown that multipotent stem cells can be cultured from the bone marrow mesenchymal stem cell pool in both mouse and humans [1-3]. The validity of these experimental findings and the reproducibility has been questioned, however, several different laboratories have cultured similarly behaving cells, albeit with different surface receptor patterns [4,5], offering support for the existence of this cell type. Here we detail the behavior of culture-derived cells which are pluripotent, long-lived and carry a unique set of surface markers when compared to the previously reported MAPC or MIAMI cells [4,5]. Our findings suggest that the surface marker profile may be due more to culture conditions and/or isolation techniques than to the inherent pluripotent nature.
Stem cells are increasingly touted as safe and effective therapeutics for a variety of diseases , however long term safety has not been adequately assessed, stressing caution for widespread clinical application of culture expanded cells in clinical medicine. Our laboratory and others have shown cells that reside in the MSC population circulate and populate peripheral organs, and contribute clonally to a wide variety of tissue. MSC are at high risk of spontaneous genetic mutation [18-20] and previously we reported that cultured MSC which are forced to replicate, acquire gene mutations commonly found in human malignant cells . Here we show that mutated cMSC repopulate the peripheral organs at a much higher level than reported for non transformed cultured MSC suggesting these cells may have an advantage over normal cells, and thus may place peripheral tissue at risk for malignancy even at a low circulating number of cells.
There is a growing body of evidence from human studies that bone marrow-derived stem cells, specifically MSC can contribute to solid tumors through a variety of mechanisms including transdifferentiation and initiation of tumors, incorporation into existing tumors as cancer cells, or via integration into stromal structures of the tumor. It is not known if the MSC acquire genetic mutations once in peripheral tissue due to environmental exposure or chronic inflammation, or if cells undergo mutations prior to engrafting into tissues. There is evidence of clonal origins in a subset of synchronous and metachronous malignancies supporting a common cell of origin [31,32]. Genetic analysis of synchronous lesions clearly demonstrates that a large number of these lesions are a non-metastatic, yet share a clonal origin; a common circulating precursor cell could explain this.
There are significant implications from these studies for cell-based regenerative medicine. Most adults derived stem cells for therapeutic use are cultured, at least briefly, and some are culture expanded, potentially introducing genetic alterations. We show that MSCs introduced into the circulation are able to engraft in normal, non-injured tissues, persist in the peripheral organs and under normal conditions function as organ-specific cells. Great care needs to be applied when expanding cells for clinical use; the apparent “benign” behavior of these cells in early patient follow up may be misleading, as our studies demonstrate that genetically compromised MSC are found throughout the body without apparent ill effect on the mouse. The potential for further mutations or epigenetic modification of the cells once in the peripheral tissue may increase cancer risk and should be considered a potential risk of any cell-based therapy.
Grant support: NIH R01- CA119061 to JH, NIH R01-CA077735 to SJ
1. Deskins DL, Bastakoty D, Saraswati S, Shinar A, Holt GE et al. Human mesenchymal stromal cells: identifying assays to predict potency for therapeutic selection. Stem Cells Transl Med. 2013, 2(2): 151-158.
3. Shoshani O, Ravid O, Massalha H, Aharonov A, Ovadya Y et al. Cell isolation induces fate changes of bone marrow mesenchymal cells leading to loss or alternatively to acquisition of new differentiation potentials. Stem Cells. 2014, 32(8): 2008-2020.
4. D’Ippolito G, Diabira S, Howard GA, Menei P, Roos BA et al. Marrow-isolated adult multilineage inducible (MIAMI) cells, a unique population of postnatal young and old human cells with extensive expansion and differentiation potential. J Cell Sci. 2004, 117(Pt 14): 2971-2981.
7. Shin L, Peterson DA. Human mesenchymal stem cell grafts enhance normal and impaired wound healing by recruiting existing endogenous tissue stem/progenitor cells. Stem Cells Transl Med. 2013, 2(1): 33-42.
10. Hutchinson L, Stenstrom B, Chen D, Piperdi B, Levey S et al. Human Barrett’s adenocarcinoma of the esophagus, associated myofibroblasts, and endothelium can arise from bone marrow-derived cells after allogeneic stem cell transplant. Stem Cells Dev. 2011, 20(1): 11-17.
16. Jodele S, Chantrain CF, Blavier L, Lutzko C, Crooks GM et al. The contribution of bone marrow-derived cells to the tumor vasculature in neuroblastoma is matrix metalloproteinase-9 dependent. Cancer Res. 2005, 65(8): 3200-3208.
18. Li H, Fan X, Kovi RC, Jo Y, Moquin B et al. Spontaneous expression of embryonic factors and p53 point mutations in aged mesenchymal stem cells: a model of age-related tumorigenesis in mice. Cancer Res. 2007, 67(22): 10889-10898.
21. Miura M, Miura Y, Padilla-Nash HM, Molinolo AA, Fu B et al. Accumulated chromosomal instability in murine bone marrow mesenchymal stem cells leads to malignant transformation. Stem Cells. 2006, 24(4): 1095-1103.
23. Subramanian K, Geraerts M, Pauwelyn KA, Park Y, Owens DJ et al. Isolation procedure and characterization of multipotent adult progenitor cells from rat bone marrow. Methods Mol Biol. 2010, 636: 55-78.
24. Sasaki M, Abe R, Fujita Y, Ando S, Inokuma D et al. Mesenchymal stem cells are recruited into wounded skin and contribute to wound repair by transdifferentiation into multiple skin cell type. J Immunol. 2008, 180(4): 2581-2587.
25. Phinney DG, Prockop DJ. Concise review: mesenchymal stem/multipotent stromal cells: the state of transdifferentiation and modes of tissue repair--current views. Stem Cells. 2007, 25(11): 2896-2902.
27. Deng J, Petersen BE, Steindler DA, Jorgensen ML, Laywell ED. Mesenchymal stem cells spontaneously express neural proteins in culture and are neurogenic after transplantation. Stem Cells. 2006, 24(4): 1054-1064.
28. Lim JY, Yi T, Choi JS, Jang YH, Lee S et al. Intraglandular transplantation of bone marrow-derived clonal mesenchymal stem cells for amelioration of post-irradiation salivary gland damage. Oral Oncol. 2013, 49(2): 136-143.
31. Li M, Cannizzaro LA. Identical clonal origin of synchronous and metachronous low-grade, noninvasive papillary transitional cell carcinomas of the urinary tract. Hum Pathol. 1999, 30(10): 1197-1200.