Biology of Blood and Marrow Transplantation
Volume 14, Issue 5 , Pages 546-555, May 2008

Neuropotency of Human Mesenchymal Stem Cell Cultures: Clonal Studies Reveal the Contribution of Cell Plasticity and Cell Contamination

Instituto de Ciencias, Facultad de Medicina, Clinica Alemana Universidad del Desarrollo, Santiago, Chile

Received 14 December 2007; accepted 24 February 2008.

Article Outline

Abstract 

Various studies have shown neuropotency of bone marrow-derived human mesenchymal stem cells (hMSC) based on the appearance of cells with neural phenotype before or after neural induction protocols. However, to date, it is unclear which mechanisms account for this observation. We hypothesized that neural phenotypes observed in hMSC cultures can be because of both intrinsic cell plasticity and contamination by cells of neural origin. Therefore, we characterized 38 clones from hMSC cultures by assessing their adipogenic/osteogenic potential with specific mesenchymal differentiation protocols, and their molecular neural phenotype by RT-PCR analysis before and after exposure to a defined neural stem cell (NSC) medium for 8 days (neural protocol). We found 33 clones with mesenchymal potential and 15 of them also showed a neural phenotype. As neural phenotypes were maintained during the neural protocol, this suggested neural cell plasticity in 39% of all clones through pluripotency. Importantly, we were able to induce neural phenotypes in 11 of mesenchymal clones applying the neural protocol, demonstrating neural cell plasticity in 29% of all clones through the mechanism of transdifferentiation. Finally, 2 of 5 nonmesenchymal clones (5% of all clones) displayed a neural phenotype indicating neural cell contamination of hMSC cultures. In conclusion, we found 2 different ways of neuropotency of hMSC cultures: cell plasticity and cell contamination.

Key Words: Mesenchymal stem cells, Multipotent stromal cells, Neuropotency, Pluripotency, Transdifferentiation, Neural cell plasticity, Neural cell contamination

 

Back to Article Outline

Introduction 

Bone marrow contains hematopoietic stem cells (HSC) and mesenchymal stem cells (MSC), also termed multipotent stromal cells, which can be separated in vitro by the property of MSC to adhere to plastic. They can be identified using standard MSC markers, as alpha-smooth muscle actin (ASMA) and surface marker CD105, and the absence of expression of the HSC markers CD45 and CD34 for their molecular characterization 1, 2. However, the “gold standard” to label these cells as MSC is the probe of their ability to differentiate into adipocytes, osteocytes, and chondrocytes 3, 4. On the other hand, various in vitro studies showed neuropotency of human MSC (hMSC) based on their expression of specific neural protein after in vitro stimulation with neural differentiation media 5, 6. In agreement with other definitions, “transdifferentiation” represents the activation of a new differentiation program. Therefore, when researchers show the appearance of neural phenotypes in hMSC cultures after applying neural induction media, they argue that cells are transdifferentiating toward neural lineage [7]. Recently, it has been shown that undifferentiated hMSC cultures isolated by plastic adhesion expressed low levels of neural-specific markers before any differentiation [8]. To date, it is unclear why some undifferentiated hMSC culture-derived cells have neural phenotype. We suggest that this might be because of pluripotency of hMSC culture-derived cells and also the presence of contaminated neural cells in the cultures.

To prove our hypothesis, we established clones from hMSC cultures and characterized them assessing mesenchymal differentiation potential and the neural gene expression patterns before and after exposure to the neural protocol. Data analysis allowed us to define the clones as mesenchymal versus nonmesenchymal and place them in a hierarchy relative to their molecular neural phenotype considering that neuroD1 expression revealed neural fate [9]; nestin expression, neural stem cell (NSC) phenotype [10]; medium chain neurofilament (NF-M) expression, neuronal phenotype [11] and, GFAP and GalC expression, glial phenotypes 12, 13.

Our clonal study demonstrates that the neuropotency of hMSC cultures depends on both mesenchymal and nonmesenchymal subpopulations of cells, including those of neural origin, which are present in bone marrow-derived cultures. Additionally, our neural protocol may serve to efficiently obtain neural-like cells from hMSC cultures.

Back to Article Outline

Material and Methods 

Isolation and Expansion of hMSC Cultures 

Leftover material was obtained from 4 heparinized aspirates of bone marrow from normal individuals undergoing bone marrow harvests for allogeneic transplantation, as part of a protocol approved by the Ethical Committee of the Hematology Department of Clinica Alemana (Santiago, Chile). Samples were diluted in 1/5 (v/v) phosphate-buffered saline (PBS) and centrifuged at 400 × g for 7 minutes. Total cells were seeded at a density of 1 × 106 nucleated cells/cm2 in α-10 medium composed by α-minimal essential medium (α-MEM, Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS) (v/v), 0.8 mg/L gentamicine (Biosano laboratory, Santiago, Chile). Cultures were incubated in a humidified atmosphere at 37°C in 5% CO2. After 24 hours, nonadherent cells were removed by replacing culture medium. When foci of fibroblastic-like cells were confluent, cells were detached with 0.25% (p/v) trypsin, 2.65 mM EDTA (Gibco), and subcultured at 7 × 103 cells/cm2 for further expansion. Isolated cells were characterized by real-time RT-PCR for CD34, CD105, and ASMA, together with the evaluation of their capacity to differentiate in vitro into adipocytes, chondrocytes, and osteocytes.

Isolation and Expansion of hMSC Culture-Derived Clones 

Clones were obtained from hMSC cultures by limiting dilution. Briefly, 100 cells/well were seeded in the first column of a 96-well plate, and immediately serial 1:2 dilutions were performed to reach 1 cell/well. To ensure single-cell clones, wells were examined daily under a phase-contrast microscope, and those containing more than 1 cell were excluded from the study. Single cells cultured in fresh medium were not able to proliferate. Single cells only grew when cultured with a mixed medium composed of 50% fresh medium and 50% conditioned medium collected from subconfluent hMSC cultures after 48 hours, centrifuged, and filtered through a 0.2 μm filter (Fisher Scientific International Inc., Pittsburgh, PA) to eliminate cellular components. Clones were expanded and further characterized according to their mesenchymal and neural phenotypes.

Mesenchymal Differentiation Protocols 

For adipogenic and osteogenic lineages, cells were seeded at a density of 2.5 × 104 cells/cm2 and later stimulated with adipogenic (α-10 medium supplemented with 100 μg/mL isomethylbutylxanthine [Calbiochem, La Jolla, CA], 1 μM dexamethasone, 0.2 U/mL insulin [Humalog], and 100 μM indomethacin [Sigma-Aldrich, St. Louis, MO], or osteogenic (α-MEM + 10% FBS with 0.1 μM dexamethasone, 50 μg/mL ascorbate-2-phosphate, and 10 mM β-glycerophosphate; Sigma-Aldrich) differentiation media during 10 and 21 days, respectively. Nonstimulated cultures were used as control and maintained in α-10 medium. Media were replaced twice a week. To assess adipogenic and osteogenic differentiation, intracellular lipid droplets were revealed by staining with Oil Red O (Merck, West Point, PA) and matrix mineralization by staining with Alizarin Red (Sigma-Aldrich), respectively [1].

For chondrogenic lineage, cells were cultured at a density of 5 × 103 cells/μL in 10 μL of α-10 medium to achieve the adequate tridimensional conditions for micromass formation. After 2 hours the α-10 medium was replaced with the chondrogenic differentiation medium modified from that used by Pittenger et al. [3]. After 7 days, the proteoglycans that compose the extracellular matrix of the micromass were revealed by staining with Safranin O (Merck).

Neural Protocol 

Cells were seeded in α-10 medium at a density of 4-6 × 103 cells/cm2. After 24 hours, the medium was replaced with NSC medium similar to that used in human NSC cultures [14], containing DMEM/F12 (1:1) (Gibco), 1% bovine serum albumin (BSA, Merck), 6 g/L D(+)-glucose (Merck), 0.8 mg/L gentamicin, supplemented with N2 and B27 supplements (Gibco), human epidermic growth factor (EGF), and human basic fibroblast growth factor (bFGF), both at 10 ng/mL (R&D Systems, Minnespolis, MN), in which the cells were cultured until the end of the experiment. Neural phenotype was evaluated according to the pattern of expression of neural genes, before (0 days) and after (8 days) exposure to NSC medium.

Isolation and Characterization of Rat NSC (rNSC) from Olfactory Bulbs 

Two olfactory bulbs were isolated from each postnatal (P0-P4) rat. Disaggregated cells were plated in suspension at cell density of 6 × 105 cells/mL in 7-10 mL of neural medium (DMEM/F12 (1:1), 1% bovine serum albumin, 6 g/L D(+)-glucose, 0.8 mg/L gentamicine) supplemented with 10% FBS. After 24 hours, cell suspension was centrifuged at 600 × g for 15 minutes and the neural growth medium was replaced with NSC medium, the same medium used for neural protocol. Cells growing in suspension formed cellular aggregates named neurospheres, and after 2-3 days were functionally characterized by their capacity to generate neurons and glial cells.

RT-PCR Analysis 

Total RNA was isolated from cells using TRIZOL® reagent (Invitrogen, Carlsbad, CA) according to the instruction of the manufacturer. RNA concentration was determined spectrophotometrically followed by treatment with RNase-free DNase (Invitrogen). Then, 1 μg of total RNA was used for reverse transcription of mRNA molecules to DNA copies. Real-time PCR was performed in a capillary containing 100 ng of cDNA, PCR LightCycler-DNA Master SYBR Green reaction mix (Roche, Indianapolis, IN), 3-4 mM MgCl2, and 0.5 μM of each specific primer, using a LightCycler® thermocycler (Roche). Total RNA from the U87 human glioblastoma cell line [15] was used as positive control for neuroD1, nestin, NF-M, and GFAP genes. To ensure that amplicons were derived from mRNA and not genomic DNA amplification, negative controls without reverse transcriptase (−RT) were perform. Negative PCR results were validated by amplification of the housekeeping gene GAPDH. To determinate the PCR sensitivity for each neural gene, serial dilutions of known amplicon concentration were reamplified, and the minimum quantity of cDNA (detection limit) that could be detected was calculated (Figure 1).

  • View full-size image.
  • Figure 1. 

    Real-time PCR analysis for mRNA detection. (A) Melting temperature (Tm) for amplicons was calculated from peaks obtained from –d[F]/dT versus T plots (F = SYBR Green fluorescence; T = °C) using a LightCycler® thermocycler (n = 3-5 reactions). Theorical amplicon size was verified by DNA-electrophoretic analysis (showed on the left side of each melting curve). Graphics at the botton of part A show representatives fluorescent-amplification profiles and a standard curve for GAPDH used to determinate the detection limit for PCR product. (B) Sequences of primers used and characteristic of the amplicons. n.d.: not determinated

Back to Article Outline

Results 

Characterization of hMSC Cultures 

hMSC from 4 bone marrow samples were selected by plastic adhesion, and cultures were characterized after passage 1 (Figure 2). Cells proliferated in α-10 medium and had a doubling time that ranged between 38 and 77 hours (data not shown). Protein and/or gene expression of markers described to be expressed on hMSC was detected in the cultures (ASMA and CD105; Figure 2B and F). hMSC culture-derived cells differentiated under specific induction media into adipocytes, osteocytes, and chondrocytes (Figure 2C-E), which demonstrated the functional properties of hMSC 1, 2, 3, 4. In addition, hMSC cultures expressed neural-lineage markers such as neuroD1, nestin, NF-M, GFAP, and GalC, but did not express hematopoietic markers as CD34 in undifferentiated culture conditions (Figure 2F).

  • View full-size image.
  • Figure 2. 

    Cellular and molecular characterization of hMSC cultures. (A) Cells showed a fibroblastic-like morphology and (B) immunolabeled using FITC-conjugated anti-ASMA monoclonal antibody. (C) hMSC culture-derived cells differentiate into adipocytes, (D) osteocytes, and (E) chondrocytes. (F) Gene expression profile of hMSC cultures using RT-PCR assay. +, detected; −, nondetected; +/−, detected in some samples. Scale bars: 100 μm.

Mesenchymal and Nonmesenchymal Phenotypes of hMSC Culture-Derived Clones 

hMSC culture-derived clones were generated from a total of 72 single cells obtained by limiting dilution method. Clonal efficiency ranged between 15% and 70%, obtaining 38 total clones that duplicated 18 times before characterization (Table 1). All 38 clones were induced by specific media (see Material and Methods) to either differentiate into adipogenic and osteogenic lineages to evaluate their mesenchymal potential. We found that the vast majority of clones differentiated into adipocytes and/or osteocytes. As previous clonal study demonstrated that adipogenic and osteogenic potentials are linked to the chondrogenic phenotype, clones that differentiated into both mesenchymal lineages were considered as MSC 4, 16. Thus, isolated clones represented phenotypes of MSC, adipogenic precursors (AdP), osteogenic precursors (OsP), and also nonmesenchymal phenotype (Figure 3). In summary, we obtained 33 mesenchymal clones (MSC, AdP, and OsP) and 5 nonmesenchymal clones (Table 2).

Table 1. Property Differences between hMSC Culture-Derived Clones
hMSC Culture∗Number of PassageNumber of Wells with a Single Cell†Number of ClonesClonal Efficiency (%)Expansion Time‡ (Days)
A415106730-35
B127197025-29
C32031538
D11066022-30

∗From 4 different donors (A-D).

†Obtained by limiting dilution method.

‡Up to ∼2 × 105 cells.

  • View full-size image.
  • Figure 3. 

    Mesenchymal potential of four representative hMSC culture-derived clones. Adipogenesis assessed by accumulation of lipid vacuoles (Oil Red O), osteogenesis indicated by extracellular calcium salts deposition (Alizarin Red). AdP: adipogenic precursors; OsP: osteogenic precursors. Scale bars: 100 μm.

Table 2. Mesenchymal∗ and Neural† Phenotypes of hMSC Culture-Derived Clones‡
MSCAdPOsPNonmesenchymal
NonneuralA10, A06, D05B27, A02, D08, B14, A07, B05, B13, B15B07D02, D07
NSC-likeA08B03, B04, A04, B01, B06, B11__
NeuronalB18, D03A03, B19, B16__
NeuroglialA09__C02, D06
GlialA05, B08B02, B09B10_
PseudoneuralA01C01, C03_B17

∗Thirty-eight clones analyzed by in vitro differentiation into adipocytes and osteocytes.

†Thirty-eight clones analyzed by RT-PCR in undifferentiated culture conditions.

‡Clone labels composed by respective donor letter (A-D) and clone number (01-19), for example, A08, B15, etc.

Neural Phenotypes Observed in Undifferentiated hMSC Culture-Derived Clones 

To investigate the neuropotency of hMSC culture-derived clones, we first analyzed the expression of neural genes in undifferentiated culture conditions. The majority of mesenchymal clones expressed the neural fate marker neuroD1. Additional expression of neuronal marker (NF-M) or glial markers (GFAP and/or GalC) was observed, indicating a neuronal or glial phenotype. We did not obtain any clone that had exclusively an oligodendrocytic phenotype. So, the expression of oligodendroglial marker GalC was always linked to GFAP expression. We also found 3 clones that coexpressed neuronal plus glial markers (neuroglial phenotype, Table 2). In addition, we classified as pseudoneural phenotype 4 clones that expressed neural genes except the transcription factor neuroD1. In summary, we found 15 mesenchymal clones that expressed neural genes in undifferentiated culture conditions (Table 2). On the other hand, only 2 of 5 nonmesenchymal clones showed a neural phenotype (Table 2). These 2 clones might be considered as a subpopulation of neural cells that are contaminated by the hMSC cultures. We also found 12 mesenchymal clones that did not express any of the neural genes analyzed (nonneural phenotype, Table 2).

Neural Phenotypes Observed in hMSC Culture-Derived Clones after Applying the Neural Protocol 

To demonstrate the neuropotency of the undifferentiated clones with neural phenotype (15 mesenchymal and 2 nonmesenchymal; Table 2) we exposed them to the neural protocol. We observed changes in the morphology and also in the pattern of neural gene expression in most clones (Figure 4). In some mesenchymal clones, cells began to form neurosphere-like structures and detached from the dish after exposure to the neural protocol (clone B10), whereas in others, these cellular aggregates kept attached until the end of the experiment (clone A08). Interestingly, these clones that showed an NSC-like morphology exclusively expressed neuroD1 gene after applying the neural protocol (Figure 4A). We found mesenchymal clones that maintained their patterns of neural gene expression, and other that upregulated or downregulated neural genes in response to the neural protocol (Figure 5A-C). On the other hand, 11 of 12 mesenchymal clones (>90%) with nonneural phenotype (Table 2) upregulated expression of neural genes after applying the neural protocol. These clones responded to the neural protocol changing their neural gene expression pattern in a different manner (Figure 5B). With regard to the nonmesenchymal clones with a neural phenotype, clone C02 showed a mature neural-like morphology together with a multineural gene expression pattern, after applying the neural protocol (Figure 4A).

  • View full-size image.
  • Figure 4. 

    Phenotypical changes in hMSC culture-derived clones after applying the neural protocol. (A) Photomicrographs showed no differences in the adherence to plastic and cell morphology between representative mesenchymal (B10 and A08) and nonmesenchymal (C02) clones before exposure to the neural protocol (d0), although there were differences in the neural gene expression pattern. Floating and adherent neurosphere-like structures and mature neural-like cells appear after applying the neural protocol (d8). (B) Typical morphology of rNSC culture-derived cells in different states of differentiation. Scale bars: 50 μm.

  • View full-size image.
  • Figure 5. 

    Mechanisms to explain neuropotency in hMSC culture-derived clones. (A) Distinct neural gene expression patterns found in clones before and after applying the neural protocol are represented in this chart. (B,C) Mesenchymal clones showed neural cell plasticity given by the mecanisms of transdifferentiation (B, 11 clones) and pluripotency (C, 15 clones). (D) The existence of 2 nonmesenchymal clones with neuropotency indicates neural cell contamination. Arrows represent the theoretical gene expression-routes of each clone. Beginning and end of arrows are related to neural gene expression patterns before and after the application of the neural protocol, respectively; dotted rectangles indicate nondifferentiated clones; black and white arrows indicate clones maturing and dedifferentiating, respectively.

Back to Article Outline

Discussion 

Our clonal analysis showed that the mesenchymal potential is maintained in 87% of isolated hMSC culture-derived clones (33 of 38, Table 2). The distinct mesenchymal phenotypes that we found (MSC, AdP, and OsP; Figure 3) confirmed that hMSC cultures are composed of heterogeneous populations of mesenchymal progenitors with different commitment, as Muraglia et al. [16] have demonstrated before. They also demonstrated that the mesenchymal potentials in hMSC clones were maintained for up to 22 cell doublings. Considering this evidence, we decided just to expand clones to no more than 18 cell doublings, assuming that the mesenchymal potential of our clones did not change during the expansion process. Thus, for each clone we obtained a limited number of cells (approximately 2 × 105 cells) that were used for mesenchymal and neural phenotyping. Because of the limited number of cells, neither immunolabeling nor electrophysiologic studies could be done. Instead, we characterized the neural phenotype based on the analysis of neural markers by RT-PCR.

According to other investigators, in the absence of functional data the analysis of the pattern of neural gene expression is more appropriate than immunolabeling to draw conclusions about the event of neural cell plasticity of hMSC 17, 18. This has been elucidated in an experiment, where apparent increases in expression of neuronal proteins, such as neuronal NSE (after applying neural induction media), also occurred when protein synthesis was inhibited [18]. Furthermore, to evaluate the event of neural cell plasticity of hMSC, it is important to emphasize that transdifferentiation involves genetic reprogramming, requiring the induction of some genes and the inhibition of others (with turning off and turning on of some genes) [19]. Therefore, if hMSC culture-derived clones could acquire a commitment toward a neural fate, an incipient cellular process might be detected at a transcriptional level. However, we did not expect to obtain mature neural cells as functional neurons using our protocol because (NSC medium is similar to other media used to expand undifferentiated neural cells, and 8 days is a short time frame for completing a neuronal differentiation process).

Finally, it has been described that neural-like cells derived from hMSC cultures can be an artefact because of the exposure to chemical stress 17, 18. Although all soluble factors present in the NSC medium were noncytotoxic (we also used this medium to expand rNSC obtained from olfactory bulbs; see Material and Methods), there was a change in morphology of clones A08 and C02 but not in their neural gene expression pattern. Therefore, the morphologic change may be given by the exposure to a serum-free medium (Figure 4A). Thus, the conclusions to characterize the distinct neural phenotypes of clones were based on their neural gene expression pattern rather than on their morphology. Specifically, the neuronal phenotype was defined by the expression of NF-M, the glial phenotype by the expression of GalC and/or GFAP, and neuroglial phenotype by coexpression of NF-M and the glial markers GFAP and GalC (see Table 2 and Figure 5A) to describe clones that had both neuronal and glial phenotype. NeuroD1 must be expressed in all neural cells, as it has been demonstrated that this transcription factor is necessary for development and maintenance of neural lineages 9, 20. Thus, in the present study, cells that neither express neuroD1 nor other neural marker were labeled as nonneural phenotypes, and those positive for any neural marker except neuroD1 as pseudoneural phenotypes (Table 2). On the other hand, it has been demonstrated that neuroD1 promotes the generation of progenitors localized in adult brain regions where there are immature and multipotent NSC [9], and the absence of neuroD1 expression in knockout mice affects to neurogenesis in the hippocampus [21]. Interestingly, some neuroD1-positive mesenchymal clones had the capacity to form cellular aggregates similar to neurospheres seen in rNSC cultures (Figure 4), were able to proliferate (data not shown), or to express neural genes of neurons and glial cells (Figure 5C), during the exposure to the NSC medium. According to other hMSC studies 5, 22, these evidence indicate that some mesenchymal clones can be committed toward an NSC phenotype. Therefore, in our study, all clones that exclusively expressed the neural gene neuroD1 were classified as an NSC-like phenotype. Although nestin protein is usually expressed in NSC from neural tissue [23] and has been used in various MSC studies as indicative of an NSC-like phenotype 5, 22, we did not find expression of nestin in our clones (discussed later). Thus, our data indicate that the expression of the neuroD1 gene but not the nestin gene is necessary to induce an NSC-like phenotype in clones derived from hMSC cultures. However, to confirm that these clones are NSC, it will be necessary to graft them into the mammalian nervous system and demonstrate their in vivo potential to differentiate into the 3 neural lineages by detecting neuronal, astrocytic, and oligodendrocytic cells.

The presence of clones with different grades of neural commitment before applying the neural protocol (18% NSC-like, 13% neuronal, 8% neuroglial, and 13% glial phenotypes; Table 2) could explain why hMSC cultures expressed basal levels of all markers of undifferentiated and mature neural cell states in undifferentiated culture conditions (Figure 2F). In addition, our data suggest that some neural phenotypes were linked to a specific mesenchymal phenotype: neuroglial phenotype was restricted to AdP clones, and glial phenotype was related exclusively to OsP clones (see Table 2).

 

Neural cell plasticity 

In agreement with other definitions, “transdifferentiation” is the mechanism by which the cell acquires phenotypes distinct from its original lineage [7]. In this way, 11 mesenchymal clones with a nonneural phenotype (Table 2) showed neuropotency upon the application of the neural protocol (Figure 5B). Therefore, we demonstrate neural cell plasticity in 11 of 33 mesenchymal clones through a transdifferentiation mechanism. On the other hand, Blondheim et al. [8] suggested that the neural phenotype of undifferentiated hMSC cultures could indicate a predisposition to differentiate into neural lineages and therefore an intrinsic neuropotency. Furthermore, it has been postulated that pluripotency showed in some somatic stem cells is because of the high accessibility of chromatin, allowing them to express many different lineage-specific genes [24]. Herein, we found 18 clones that have both mesenchymal and neural phenotypes (Table 2). Fifteen of them also have neural phenotypes after applying the neural protocol (Figure 5C), and therefore indicates that they had intrinsic neuropotency because of their pluripotent properties. So, “pluripotency” reveals another mechanism (different of “transdifferentiation”) by which 15 of 33 mesenchymal clones have neural cell plasticity. In summary, the clonal evidence of transdifferentiation and pluripotency confirm and demonstrate in vitro the neural cell plasticity of hMSC culture-derived cells with mesenchymal potential.

Neural cell contamination 

It has been suggested that bone marrow is a reservoir of stem cells and progenitors from different tissues [25], not only HSC and MSC. We found 5 nonmesenchymal clones in our study, and 2 of them had neural phenotype before and after the application of the neural protocol (Figure 5D). The other 3 nonmesenchymal clones did not express any neural gene. Previously, other investigators did not find any nonmesenchymal phenotype in their clonal study [16]. This discrepancy with our data could be explained by the different isolation methods used. Although Muraglia et al. [16] isolated clones directly from bone marrow samples, we obtained our clones from hMSC cultures after selecting them by adherence to plastic. Primary human NSC can also adhere to the dish and adopt a fibroblastic-like morphology when cultured in medium containing serum. Therefore, the culture conditions used to isolate bone marrow-derived hMSC might also serve to obtain neural progenitor cells that can be present in the bone marrow. The funding of 2 nonmesenchymal clones with neuropotency (Figure 5D) is in agreement with a recent study that demonstrates the presence of neural crest-derived stem cells in rodent adult bone marrow [26]. The 18 cell doublings before obtaining clones from a single cell, allowed us to characterize them as progenitor cells. However, what could explain the survival and expansion of “contaminated” neural progenitors in our clonal study? The answer could be found within the expansion medium we used (50% fresh medium and 50% conditioned medium). We detected BDNF protein by ELISA (data not shown) in the hMSC culture-derived conditioned medium used during the clonal expansion process. A previous work demonstrated the positive effect of hMSC-conditioned medium on the survival of neural cells because of the presence of neurotrophic factors such as BDNF and NGF [27]. Thus, BDNF present in our hMSC culture-derived conditioned medium could have allowed the survival and expansion of the 2 nonmesenchymal clones with neural phenotype found before applying the neural protocol (Table 2). On the other hand, a study evaluated the effect of exogenous BDNF added to hMSC cultures and found that this neurotrophic factor induced the expression of neuronal and glial markers and downregulated nestin gene expression [28]. Therefore, the BDNF of our hMSC-conditioned medium could explain the absence of nestin in the clones. In addition, we cannot exclude the fact that BDNF present during the expansion process induced the expression of neural genes in the mesenchymal clones that showed neural phenotype in undifferentiated culture conditions (Table 2), including the 15 pluripotent clones (Figure 5C).

Taken together, our results demonstrated that neural cell plasticity can be reached in 68% of all hMSC culture-derived clones, either by transdifferentiation or pluripotency mechanisms. In addition, we have evidence of cell contamination with neural progenitors in the hMSC cultures (40% of nonmesenchymal clones). In addition, our results using hMSC culture-derived clones obtained from adult human bone marrow suggest that hMSC cultures are heterogeneous with respect to the tissue-specific progenitors they contain, not only those derived from mesenchymal lineages but also from neural lineages. Finally, our neural protocol was able to induce neural phenotype in >90% (11 of 12) of the mesenchymal clones that previously did not show a neural phenotype. Thus, in the absence of standarized neural induction media, our neural protocol is a good alternative to enrich hMSC cultures derived from bone marrow with cells committed toward the neural lineage.

Back to Article Outline

Acknowledgments 

The authors thank Dr. Gonzalo Bustos and Jorge Abarca (Pontificia Universidad Catolica de Chile) for BDNF ELISA determinations. The authors also thank Mrs. Carolina Larrain and Dr. Tobias Manigold for critical reading, discussion, and helping writing the manuscript. This work was supported by Young Researchers Career Program, sponsored by Fundacion Andes (Grant C-14060/60), and the institutional grant 80.11.003 by Universidad del Desarrollo, Chile.

Back to Article Outline

References 

  1. Conget PA, Minguell J. Phenotypical and functional properties of human bone marrow mesenchymal progenitor cells. J Cell Physiol. 1999;181:67–73
  2. Galmiche MC, Koteliansky VE, Briere J, Herve P, Charbord P. Stromal cells from human long-term marrow cultures are mesenchymal cells that differentiate following a vascular smooth muscle differentiation pathway. Blood. 1993;82:66–76
  3. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science. 1999;284:143–147
  4. Horwitz EM, Le Blanc D, Dominici M, et al. Clarification of the nomenclature for MSC: the international society for cellular therapy position statement. Cytotherapy. 2005;7:393–395
  5. Hermann A, Gastl R, Liebau S, et al. Efficient generation of neural stem cell-like cells from adult human bone marrow stromal cells. J Cell Sci. 2004;117:4411–4422
  6. Sanchez-Ramos J, Cardozo-Pelaez F, Hazzi C, et al. Adult bone marrow stromal cells differentiate into neural cells in vitro. Exp Neurol. 2000;164:247–256
  7. Wagers A, Weissman I. Plasticity of adult stem cells. Cell. 2004;116:639–648
  8. Blondheim NR, Levy YS, Ben-Zur T, et al. Human mesenchymal stem cells express neural genes suggesting a neural predisposition. Stem Cells Dev. 2006;2:141–164
  9. Cho JH, Tsai MJ. The role of BETA2/NeuroD1 in the development of the nervous system. Mol Neurobiol. 2004;30:35–47
  10. Dahlstrand J, Zimmerman LB, McKay RD, Lendahl U. Characterization of the human nestin gene reveals a close evolutionary relationship to neurofilaments. J Cell Sci. 1992;2:589–597
  11. Steinert M, Liemt R. Intermediate filament dynamics. Cell. 1990;60:521–523
  12. Bogler O, Wren D, Barnett SC, Landt H, Noble M. Cooperation between two growth factors promotes extended self renewal and inhibits differentiation of oligodendrocyte-type-2 astrocyte (0-2A) progenitor cells. Proc Natl Acad Sci USA. 1990;87:6368–6372
  13. Rogister B, Ben-Hur T, Dubois-Dalcq M. From neural stem cells to myelinating oligodendrocytes. Mol Cell Neurosci. 1999;14:287–300
  14. Martínez-Serrano A, Rubio FJ, Navarro B, Bueno C, Villa A. Human neural stem and progenitor cells: in vitro and in vivo properties and potential for gene therapy and cell replacement in the CNS. Curr Gene Ther. 2001;1:279–299
  15. Carlsson J, Nilsson K, Westermark B. Formation and growth of multicellular spheroids of human origin. Int J Cancer. 1983;31:523–533
  16. Muraglia A, Cancedda R, Quarto R. Clonal mesenchymal progenitors from human bone marrow differentiate in vitro according to a hierarchical model. J Cell Sci. 2000;113:1161–1166
  17. Bertani N, Malatestá P, Volpi G, Sonego P, Perris R. Neurogenic potential of human mesenchymal stem cells revisited: analysis by immunostaining, timelapse video and microarray. J Cell Sci. 2005;118:3925–3936
  18. Lu P, Blesch A, Tuszynski M. Induction of bone marrow stromal cells to neurons: differentation, transdifferentation, or artifact?. J Neurosci Res. 2004;77:174–191
  19. Krabbe C, Zimmer J, Meyer M. Neural transdifferentation of mesenchymal stem cells—a critical review. APMIS. 2005;113:831–844
  20. Lee JK, Cho JH, Hwang WS, Lee YD, Reu DS, Suh-Kim H. Expression of neuroD/BETA2 in mitotic and postmitotic neuronal cells during the development of nervous system. Dev Dyn. 2000;217:361–367
  21. Miyata T, Maeda T, Lee JE. NeuroD1 is required for differentiation of the granule cells in the cerebellum and hippocampus. Genes Dev. 1999;13:1647–1652
  22. Suzukia H, Taguchia T, Tanakaa H, et al. Neurospheres induced from bone marrow stromal cells are multipotent for differentiation into neuron astrocyte and oligodendrocyte phenotypes. Biochem Biophys Res Commun. 2004;322:918–922
  23. Lendahl U, Zimmerman LB, McKay RD. CNS stem cells express a new class of intermediate filament protein. Cell. 1990;60:585–595
  24. Eckfeldt CE, Mendenhall EM, Verfaillie CM. The molecular repertoire of the “almighty” stem cell. Nat Rev Mol Cell Biol. 2005;6:726–737
  25. Ratajczak MZ, Kucia M, Majka M, Reca R, Ratajczak J. Heterogeneous populations of bone marrow stem cells—are we spotting on the same cells from the different angles?. Folia Histochem Cytobiol. 2004;42:139–146
  26. Nagoshi N, Shibata S, Nakamura M, Matsuzaki Y, Toyama Y, Okano H. Multipotent neural crest-derived stem cells in bone marrow, dorsal root ganglia, and facial skin in adult rodent. 37th Annual Meeting of Society for Neuroscience 2007; Abstract 237.2.
  27. Crigler L, Robey RC, Asawachaicharn A, Gaupp D, Phinney DG. Human mesenchymal stem cell subpopulations express a variety of neuron-regulatory molecules and promote neuronal cell survival and neuritogenesis. Exp. Neurol. 2006;198:54–64
  28. Choong PF, Mok PL, Cheong SK, Leong CF, Then KY. Generating neuron-like cells from BM-derived mesenchymal stromal cells in vitro. Cytotherapy. 2007;9:170–183

PII: S1083-8791(08)00093-1

doi:10.1016/j.bbmt.2008.02.017

Biology of Blood and Marrow Transplantation
Volume 14, Issue 5 , Pages 546-555, May 2008

Article Tools

* Image Usage & Resolution