Plasticity of Cultured Mesenchymal Stem Cells: Switch from Nestin-Positive to Excitable Neuron-Like Phenotype
http://www.100md.com
《干细胞学杂志》
a Center for Cellular and Molecular Neurobiology and
b Department of Neurology, University of Liège, Liège, Belgium;
c Department of Physiology, Transnationale Universiteit Limburg / Limburgs Universitair Centrum, Biomedisch Onderzoekinstituut, Diepenbeek, Belgium
Key Words. Action potential ? Differentiation ? Mesenchymal stem cells ? Neurons ? Neurotransmission
Correspondence: Sabine Wislet-Gendebien, Centre for Cellular and Molecular Neurobiology, University of Liège, 17 Place Delcour, B-4020 Liège, Belgium. Telephone: 32-4-366-5917; Fax: 32-4-366-5912; E-mail: Sabine.Wislet@utoronto.ca
ABSTRACT
A stem cell is an unspecialized cell with the ability to renew itself indefinitely; under appropriate conditions, stem cells can give rise to a wide range of mature cell types. Two types of stem cells can be distinguished according to their origin and their potential of differentiation: embryonic stem cells (ESCs) and somatic stem cells (SSCs) . ESCs are derived from the early blastocyst and the inner cell mass of the embryo and are able to differentiate into the three germ layer cell types (pluripotentiality) . SSCs are isolated from fetal (after gastrulation) or adult tissues, but classically they are programmed to produce only the cell types that belong to the tissue from which they originate. However, recent studies suggest that SSCs might be able to exhibit more plasticity than previously thought, as they seem able to differentiate into many cell types. These observations were mainly reported when SSCs were grafted in a damaged tissue . This phenomenon, also known as phenotypic plasticity of SSCs, has been described among others for mesenchymal stem cells (MSCs), which under appropriate in vivo or in vitro conditions can adopt a neural fate .
Over the past few years in parallel to stem cell research, there has been a growing interest in the molecular mechanisms involved in cell type determination and differentiation during development. The discovery of intrinsic and extrinsic factors leading to the differentiation into ectoderm, mesoderm, and endoderm during the early stages of development; and the identification of factors that are responsible for further differentiation into more restricted cell types later in development, would allow a better understanding of the mechanisms that control the phenotypic plasticity of stem cells.
Recently, we demonstrated that cultured MSCs from adult rat bone marrow are able to express nestin, an intermediate filament protein that is predominantly expressed during neural development . Two factors seem essential for this nestin expression to occur: the first one is the absence of serum-derived components in the culture medium, and the second one is an in vitro maturation of the MSCs. Indeed, a minimum of 25 population doublings are required for nest into be expressed in 75% of cultured MSCs. In this study, we demonstrated that nestin-positive MSCs, when co-cultured with mouse cerebellar granule (mCG) neurons, are able to differentiate into excitable neuron-like cells. By using quantitative reverse transcription polymerase chain reaction (RT-PCR), we demonstrated that sox and pax transcription factors, as well as the frizzled (wnt receptor) and erbB (neuregulin receptors) receptors, are upregulated in nestin-positive MSCs compared with nestin-negative MSCs (nnMSCs), suggesting a possible role for these genes in the acquisition of a responsiveness to neural fate–inducing signals in a nestin-positive MSC population. Electrophysiological recordings of npMSC-derived neuron-like (MDN) cells show that in co-culture with CG neurons, MDN cells first express functional voltage-gated K+ channels followed after 8 days of co-culture by voltage-gated Na+ channels, allowing firing of action potentials. Finally, several experiments strongly suggest that this neuronal fate in the nestin-positive MSCs is a consequence of a "transdifferentiation" process, although not completely ruling out that cell fusion processes could also contribute to the appearance of neuronal-like cells.
MATERIALS AND METHODS
Neuronal and Astroglial Marker Expression by Nestin-Positive MSCs
MSCs were isolated from femoral and tibial bones of adult rats, propagated in culture, and immunologically and functionally characterized . As we already demonstrated, a nestin-expressing stage is necessary to observe a GFAP+ cell differentiation of MSCs co-cultured with neural stem cells . To observe a possible neuronal differentiation of MSCs, we co-cultured nestin-positive MSCs for 5 days with GFP+ mCG neurons obtained from 3-day-old green mice. After 5 days of co-culture, immunological labelings revealed that MSCs began to express differentiation markers: 40.17% ± 2.95% expressed GFAP (Fig. 1A–C), 19.00% ± 1.58% expressed Tuj1 (Fig. 1D–F), and 19.30% ± 1.38% expressed NeuN (Fig. 1G) (n = 12 for each condition). These results were confirmed by FACS analyses (Fig. 1I,J) as we measured 44.57% ± 3.59% GFAP– and 18% ± 4.1% Tuj1+ cells, data which are not statistically different from that obtained by eye-counting immunolabeled cells. The FACS analyses were performed on GFP– cell populations corresponding to the first spike of cells in Figure 1K. Moreover, the absence of GFAP and NeuN double-labeled cells suggests that nestin-positive MSCs are specifically oriented toward either an astroglial or a neuronal fate (Fig. 1H). Two conditions are needed before MSCs begin to express neuronal markers: first, for the appearance of GFAP+ cells in co-cultures with neural stem cells, MSCs have to express nestin; second, a direct cell–cell contact between nestin-positive MSCs and CG neurons is needed. Indeed, when nnMSCs are co-cultivated with CG neurons or when nestin-positive MSCs and CG neurons are co-cultivated with a physical separation (millicell device), no neuronal markers could be detected (data not shown).
Figure 1. Neuronal and astroglial markers expression by nestin-positive mesenchymal stem cells (MSCs). Nestin-positive MSCs were co-cultured for 5 days with green fluorescent protein (GFP)–positive cerebellar granule (CG) neurons (green). Immunocytochemical labeling and fluorescence-activated cell sorting (FACS) analysis showed that some nestin-positive MSCs expressed neural markers. GFAP (red, A–C) was expressed by about 40% of the nestin-positive MSCs, and Tuj1 (red, D–F) was expressed by about 20% of the nestin-positive MSCs, as was the NeuN marker (red, G). Those results were confirmed by FACS analysis (GFAP, I; Tuj1, J) on a GFP– population (K). A double labeling directed against GFAP (red) and NeuN (blue) (H) showed that nestin-positive MSCs specifically oriented toward an astroglial-like or a neuronal-like fate. Nuclei were counter-stained with EtD1 (blue or red, depending on the color used for the secondary antibody). Arrowheads show neural expression markers by nestin-positive MSCs. Scale bars = 40 μm (A, D, and G–K) and 30 μm (B–C and E–F).
Characterization of Nestin-Positive MSCs
The nestin expression by MSCs coincides with the acquisition by these cells of the ability to respond to extrinsic signals and cues driving their neural differentiation . To get some clue into the molecular mechanism(s) involved in this differentiation, we decided to compare the expression in nestin-positive and nestin-negative MSCs of several factors known to have important roles during normal nervous system development. Using RT-PCR and quantitative RT-PCR, the nestin-positive MSC and nnMSC gene expression profiles of the following factors were compared, both qualitatively and quantitatively: the sox and pax transcription factors, as well as frizzled (wnt receptors), erbB (neuregulin receptors) and the notch-delta signaling pathway. Figure 2 shows that MSCs express sox2, sox9, sox10, sox11, pax6, notch1, notch2, delta1, erbB2, erbB4, and fzd and do not express sox1, sox3, pax3, delta2, and erbB3. In the analyses of fzd genes, we used degenerated primers that hybridize with all fzd genes . The resulting products were cloned using the TA cloning procedure, and random clones were analyzed by sequencing. A BLAST search allowed us to determine the identity of each clone: 80% of fzd receptors in nnMSCs were fzd2, and 20% were fzd1; likewise, in nestin-positive MSCs, we obtained 75% of fzd2, 20% of fzd1, and 5% of fzd5 (data not shown). Finally, quantitative RT-PCR demonstrated that nestin-positive MSCs overexpress sox2, sox10, pax6, fzd, erbB2, and erbB4 when compared with the nnMSCs. We also observed that nestin-positive MSCs overexpress the GFAP gene compared with nnMSCs, but none of them express the neurofilament-light gene.
Figure 2. Molecular characterization of nestin-positive mesenchymal stem cells (MSCs). Reverse transcription polymerase chain reaction (RT-PCR) and quantitative RT-PCR were performed to better characterize the molecular profile of nestin-positive MSCs compared with nestin-negative MSCs (nnMSCs) regarding some genes known to have a role during neuro-ontogenesis. In this case, we observed that nestin-positive MSCs overexpressed sox2, sox10, pax6, erbB2, erbB4, fzd, and GFAP. The results of quantitative RT-PCR are expressed as percentage of gene expression in nestin-positive MSCs compared with the same gene expressed in nnMSCs after normalization with the GAPDH housekeeping gene.
Electrophysiological Analysis of Mesenchymal-Derived Neuron-Like (MDN) Cells
Using fluorescent microscopy, we selected for cells that presented a neuron-like shape (rounded cell body with extended processes, hereafter termed MDN cells), that had incorporated DiD Vybrant (red) and that were clearly GFP– (see Methods). Whole-cell patch-clamp recordings were obtained from 234 MDN cells that had been co-cultured with CG neurons for 4–15 days. Based on the appearance of different voltage-gated current profiles (see next paragraph), three culture time periods were distinguished that, moreover, showed significantly (ANOVA; p < .001) different resting membrane potentials (Vrest): –37.6 ± 3.0 mV for MDNs co-cultured with CGs for 4–6 days in vitro (hereafter noted 5DIV; n = 61), –50.3 ± 2.0 mV for 7–9 days (8DIV; n = 76), and –55.7 ± 2.3 mV for 10–15 days (12.5DIV; n = 97).
Excitability ? The activation of voltage-gated channels at the three developmental stages (5, 8, and 12.5DIV) was assessed by applying to whole-cell patch-clamped MDN cells hyperpolarizing and depolarizing voltage steps from a holding potential (VH) of –80 mV (–80 to –125 mV and –80 to +30 mV in 15-mV increments; Fig. 3A–D). At 5DIV, depolarizing steps positive to –40 mV induced outward currents in 95% of MDN cells that showed little desensitization and were blocked by 10 μM TEA, indicating the activation of potassium voltage-gated channels. Potassium current densities were stable throughout the culture time period (7.25 ± 1.16 pA.pF–1). At 5DIV, no inward currents were observed. Conversely, depolarizing voltage steps positive to –30 mV induced inward currents in 85% of 8DIV MDN cells and in 95% of 12.5DIV MDN cells that rapidly activate and inactivate and that were abolished by a perfusion of 1μM TTX, showing that they are mediated by voltage-activated sodium channels. Between 8DIV and 12.5DIV, no significant differences were found regarding the sodium current densities (8.29 ± 0.87 pA.pF–1 (n = 51) at 8DIV compared with 10.15 ± 1.04 pA.pF–1(n = 59) at 12.5DIV). However—and interestingly—the sodium current densities in MDN cells were two times lower (18.63 ± 2.45 pA.pF–1; n = 21) than the one recorded in co-cultured CG neurons. Consistent with the occurrence of sodium and potassium voltage-gated currents, an injection of 600 nA current into whole-cell patch-clamped and current-clamped MDN cells elicited action potentials in 21.5% of 8DIV MDN cells (n = 34) (Fig. 4A) and in 25% of 12.5DIV MDN cells. These action potentials were always present as single spikes and were reversibly inhibited by a 1-μM TTX application (Fig. 4).
Figure 3. Voltage-gated channel activation in mesenchymal stem cell (MSC)–derived neuron-like (MDN) cells. (A): Typical recordings of single whole-cell patch-clamped and voltage-clamped MDN cells at three culture time periods (5, 8, and 12.5DIV). Hyperpolarizing and depolarizing steps were applied from a holding potential of –80 mV (–80 to –120 mV and –80 to +30 mV in 15-mV increments). The first and third lines show recordings in control conditions, and the second in the presence of 1 μM tetrodotoxin (TTX). The arrows indicate the time where the measures were performed (filled arrows for outward currents, and open arrows for inward currents). (B–D): Current-voltage relationships for outward (triangle; recordings obtained in the presence of 1 μM TTX) and inward (square; recording obtained in the presence of 10 μM tetraethyl ammonium ).
Figure 4. Excitability as shown in a typical recording of a single whole-cell patch-clamped and current-clamped mesenchymal stem cell (MSC) neuron-like (MDN) cell at 8DIV. A depolarizing 600 nA current injection induced the firing of a single action potential that could be reversibly blocked by 1 μM tetrodotoxin (TTX).
Neurotransmitter Effects ? Responses of whole-cell voltage-clamped MDNs (VH = –80 mV) to microperfusion-applied neurotransmitters (1 mM GABA, 100 μM glycine, and 100 μM glutamate) were subsequently measured. As soon as at 5DIV, neurotransmitters elicited inward currents in 98% of MDN cells (n = 78, Fig. 5A). Since no significant differences in current densities were observed between the different culture time periods, measures of current densities were pooled for each neurotransmitter, yielding means of 28.2 ± 2.7 pA.pF–1 for GABA, 61.4 ± 10.6 pA.pF–1 for glycine, and 7.5 ± 2.6 pA.pF–1 for glutamate. The responses elicited by GABA, glycine, and glutamate were mediated by their respective ionotropic receptors—that is, GABAA receptors, glycinereceptors, and AMPA/KA/NMDA receptors—since transmitter-evoked currents were reversibly blocked by specific antagonists of these receptors: namely, by gabazine (10 μM), strychnine (1 μM), and a mix of CNQX and AP-5 (1 μM each), respectively. Before the co-culture of nestin-positive MSCs with CG neurons, inward currents elicited by those neurotransmitters were already present, but only in 35% of nestin-positive MSCs (n = 15, Fig. 5B) and with much lower mean current densities: 0.06 ± 0.02 pA.pF–1 for GABA, 0.07 ± 0.03 pA.pF–1 for glycine, and 0.13 ± 0.05 pA.pF–1 for glutamate. Current densities of neurotransmitter-evoked responses were also measured in CG neurons with which nestin-positive MSCs were co-cultured. In CG neurons, GABA, glycine, and glutamate elicited currents in 100% of the cells (n = 15), with mean current densities of 19.71 ± 3.96 pA.pF–1 for GABA, 26.02 ± 5.55 pA.pF–1 for glycine, and 12.68 ± 4.54 pA.pF–1 for glutamate. Current density comparison between MDN cells and CG neurons yielded a significant difference for glycine responses only (ANOVA; p < .01). Finally, we looked for a possible synaptic activation of the neurotransmitter receptors present in MDN cells. Although MDN cells expressed axonal (MAP2ab, Fig. 6A–C), dendritic (SMI31, Fig. 6D–F), and synaptic markers (synaptophysin, Fig. 6G–I), we never observed spontaneous synaptic events in cultured MDN cells.
Figure 5. Neurotransmitter sensitivities. (A): Typical recording of single whole-cell patch-clamped and voltage-clamped mesenchymal stem cell (MSC) neuron-like (MDN) cell at 8DIV. Either 1 mM GABA, 100 μM glycine, or 100 μM glutamate was applied for 10 seconds every 30 seconds, and the resulting current was recorded in the absence or presence of specific antagonists—namely, 10 μM gabazine, 1 μM strychnine, and a mix of 1 μM CNQX and 1 μM D-AP5, respectively. Antagonists were first applied alone for 10 seconds before being applied for 10 seconds with neurotransmitters. (B): Responses to GABA, glycine, and glutamate in mouse cerebellar granule (mCG) neurons, MSC-derived neuron-like (MDN) cells, and nestin-positive MSCs (before co-culturing with mCG neurons) are expressed as current densities—that is, mean peak amplitudes corrected for cell capacitance (pA.pF–1). **p < .01 (one-way ANOVA followed by Dunnett’s post-tests)
Figure 6. (A–I): Expression of axonal (MAP2ab, blue, A–C), dendritic (SMI31, blue, D–F), and synaptic (synaptophysin, blue, G–I) markers by 8DIV cultured MDN cells.
Mechanisms of Neural Differentiation of Nestin-Positive MSCs ? Recent studies have indicated that MSCs are able to fuse with mature neurons . In this study, three types of experiments were performed which strongly suggest that MSCs differentiate into neuron-like cells rather than fuse with existing CG neurons. The first experiment involved the analysis of the DNA content of nestin-positive MSCs and CG neurons before and after the co-culture. In all cases, we did not observe a significant difference in ploidy, which could explain the 60% of neural differentiation (40% into astrocytes and 20% into neurons) in the MSC population (Fig. 7A). Although the sensitivity of this technique does not allow us to completely rule out that some rare polyploid cells could be present, the majority of the observed cells were diploid. In a second set of experiments, we used the M2 and M6 markers, which specifically recognize respectively astrocytes and neurons from mouse but not from rat . Rat npMSCs were co-cultivated for 5 days with mouse CG neurons. Thereafter, double labeling with GFAP and M2 antibodies, on one hand, and Tuj1 and M6, on the other hand, were performed. We could observe (1) that all GFAP+/ GFP– and Tuj1+/GFP– cells are also negative for M2 and M6, respectively, allowing us to conclude that those cells are of rat origin and thus likely derive from the MSC population (Fig. 7B–G); and (2) that all the cells recognized either by the M2 or by the M6 antibodies are also GFP+, ruling out a downregulation of GFP expression during the co-culture period. Finally, nestin-positive MSCs were cultured for 5 days on paraformaldehyde-fixed GFP+ CG neurons in the presence of CG-conditioned medium. In such conditions, we observed that 16.1% ± 2.6% (n = 8) nestin-positive MSC-derived cells were NeuN+ (Fig. 7H–J, and 23.1% ± 2.1% (n = 7) nestin-positive MSC-derived cells were GFAP+ (Fig. 7K–M). Note that the level of GFP fluorescence in the fixed cells maintained for 5 days in culture remains stable. Although some rare fusion events cannot be excluded in these experiments, we can conclude that most of the MDN cells result from a true differentiation process of nestin-positive MSCs.
Figure 7. Differentiation of nestin-positive mesenchymal stem cells (MSCs) into neural cells. (A): Nestin-positive MSCs (red) and mouse cerebellar granule (mCG) neurons (yellow) before and after co-culture (blue) were stained with propidium iodide and subjected to fluorescence-activated cell sorting (FACS) analysis to determine their ploidy. (B–D): A double-labeling with M2 (red) and GFAP (blue) antibodies allowed the confirmation of the mesenchymal origin of some GFAP+ cells. Nestin-positive MSC-derived GFAP+ cells (C) are not recognized by the M2 antibody (B), but the GFP+ astrocytes (green astrocytes) are recognized by the M2 antibody (D). (E–G): Likewise, a double-labeling with M6 (blue) and Tuj1 (red) antibodies allows the confirmation of the mesenchymal origin of some Tuj1+ cells. Nestin-positive MSC-derived Tuj1+ cells (E) are not recognized by the M6 antibody (F), but the GFP+ neurons (green neurons) are recognized by the M6 antibody (G). (H–M): Finally, we cultivated nestin-positive MSCs onto paraformaldehyde-fixed green fluorescent protein (GFP)–positive mCG neurons in centrifuged and filtrated mCG-conditioned medium for 5 days. GFAP+ cells (red, H–J) and NeuN+ cells (red, K–M) derived from nestin-positive MSCs were observed. Arrowheads indicate the nestin-positive MSC-derived cells that express neural markers. Scale bars (B–C, E–F, H–I, and K–L) 20 μm and (D, G, J, and M) 40 μm.
DISCUSSION
This work was supported by a grant of the Fonds National de la Recherche Scientifique (FNRS) of Belgium, by the Fondation Médical Reine Elisabeth (FMRE), by the Fonds Charcot, and by the Belgian League against Multiple Sclerosis. S.W. is a research fellow of the Télévie (FNRS). B.R. is senior research associate, P.L. a research associate, and G.H. a research fellow of the FNRS. We thank Bernard Coomans (CNCM, University of Liège, Liège, Belgium) for his help with the FACS analysis and Patricia Ernst for her broad competence in cerebellar granule culture.
REFERENCES
Stanford WL, Caruana G, Vallis KA et al. Expression trapping: identification of novel genes expressed in hematopoietic and endothelial lineages by gene trapping in ES cells. Blood 1998;92:4622–4631.
Amit M, Carpenter MK, Inokuma MS et al. Clonally derived human embryonic stem cell lines maintain pluri-potency and proliferative potential for prolonged periods of culture. Dev Biol 2000;227:271–278.
Itskovitz-Eldor J, Schuldiner M, Karsenti D et al. Differentiation of human embryonic stem cells into embryoid bodies compromising the three embryonic germ layers. Mol Med 2000;6:88–95.
Schuldiner M, Yanuka O, Itskovitz-Eldor J et al. Effects of eight growth factors on the differentiation of cells derived from human embryonic stem cells. Proc Natl Acad Sci U S A 2000;97:11307–11312.
Raff M. Adult stem cell plasticity: fact or artifact? Annu Rev Cell Dev Biol 2003;19:1–22.
Kopen GC, Prockop DJ, Phinney DG. Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 1999;96:10711–10716.
Brazelton TR, Rossi FM, Keshet GI et al. From marrow to brain: expression of neuronal phenotypes in adult mice. Science 2000;290:1775–1779.
Alvarez-Dolado M, Pardal R, Garcia-Verdugo JM et al. Fusion of bone-marrow-derived cells with Purkinje neurons, cardiomyocytes and hepatocytes. Nature 2003;425:968–973.
Wislet-Gendebien S, Leprince P, Moonen G et al. Regulation of neural markers nestin and GFAP expression by cultivated bone marrow stromal cells. J Cell Sci 2003;116:3295–3302.
Azizi SA, Stokes D, Augelli BJ et al. Engraftment and migration of human bone marrow stromal cells implanted in the brains of albino rats: similarities to astrocyte grafts. Proc Natl Acad Sci U S A 1998;95:3908–3913.
Lefebvre PP, Rogister B, Delree P et al. Potassium-induced release of neuronotoxic activity by astrocytes. Brain Res 1987;413:120–128.
Okabe M, Ikawa M, Kominami K et al. "Green mice" as a source of ubiquitous green cells. FEBS Lett 1997;407:313–319.
Shibata T, Yamada K, Watanabe M et al. Glutamate transporter GLAST is expressed in the radial glia-astrocyte lineage of developing mouse spinal cord. J Neurosci 1997;17:9212–9219.
Helmbrecht H, Kispert A, von Wasielewski R et al. Identification of Wnt/?-Catenin signalling pathway in human thyroid cells. Endocrinology 2001;142: 5261–5266.
Weimann JM, Johansson CB, Trejo A et al. Stable reprogrammed heterokaryons form spontaneously in Purkinje neurons after bone marrow transplant. Nat Cell Biol 2003;5:959–966.
Weimann JM, Charlton CA, Brazelton TR et al. Contribution of transplanted bone marrow cells to Purkinje neurons in human adult brains. Proc Natl Acad Sci U S A 2003;100:2088–2093.
Lagenaur C, Schachner M. Monoclonal antibody (M2) to glial and neuronal cell surfaces. J Supramol Struct Cell Biochem 1981;15:335–346.
Isacson O, van Horne C, Schumacher JM et al. Improved surgical cell therapy in Parkinson’s disease: physiological basis and new transplantation methodology. Adv Neurol 2001;86:447–454.
Dunnett SB. Functional analysis of fronto-striatal reconstruction by striatal grafts. Novartis Found Symp 2000;231:21–41.
Hall ED. Pharmacological treatment of acute spinal cord injury: how do we build on past success? J Spinal Cord Med 2001;24:142–146.
Wagers AJ, Weissman IL. Plasticity of adult stem cells. Cell 2004;116:639–648.
Graham V, Khudyakov J, Ellis P et al. SOX2 functions to maintain neural progenitor identity. Neuron 2003;39:749–765.
Wakamatsu Y, Endo Y, Osumi N et al. Multiple roles of Sox2, an HMG-box transcription factor in avian neural crest development. Dev Dyn 2004;229:74–86.
Britsch S, Goerich DE, Riethmacher D et al. The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev 2001;15:66–78.
Stolt CC, Rehberg S, Ader M et al. Terminal differentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev 2002;16:165–170.
Kocsis JD, Akiyama Y, Lankford KL et al. Cell transplantation of peripheral-myelin-forming cells to repair the injured spinal cord. J Rehabil Res Dev 2002;39:287–298.
Tyas DA, Pearson H, Rashbass P et al. Pax6 regulates cell adhesion during cortical development. Cereb Cortex 2003;13:612–619.
Talamillo A, Quinn JC, Collinson JM et al. Pax6 regulates regional development and neuronal migration in the cerebral cortex. Dev Biol 2003;255:151–163.
Calaora V, Rogister B, Bismuth K et al. Neuregulin signaling regulates neural precursor growth and the generation of oligodendrocytes in vitro. J Neurosci 2001;21:4740–4751.
Canoll PD, Kraemer R, Teng KK et al. GGF/neuregulin induces a phenotypic reversion of oligodendrocytes. Mol Cell Neurosci 1999;13:79–94.
BaoJ, WolpowitzD, RoleLWetal. Back signaling by the Nrg-1 intracellular domain. J Cell Biol 2003;161:1133–1141.
Lemke G. Glial control of neuronal development. Annu Rev Neurosci 2001;24:87–105.
Carleton A, Petreanu LT, Lansford R et al. Becoming a new neuron in the adult olfactory bulb. Nat Neurosci 2003;6:507–518.
Herzog EL, Chai L, Krause DS. Plasticity of marrow-derived stem cells. Blood 2003;102:3483–3493.(Sabine Wislet-Gendebiena,)
b Department of Neurology, University of Liège, Liège, Belgium;
c Department of Physiology, Transnationale Universiteit Limburg / Limburgs Universitair Centrum, Biomedisch Onderzoekinstituut, Diepenbeek, Belgium
Key Words. Action potential ? Differentiation ? Mesenchymal stem cells ? Neurons ? Neurotransmission
Correspondence: Sabine Wislet-Gendebien, Centre for Cellular and Molecular Neurobiology, University of Liège, 17 Place Delcour, B-4020 Liège, Belgium. Telephone: 32-4-366-5917; Fax: 32-4-366-5912; E-mail: Sabine.Wislet@utoronto.ca
ABSTRACT
A stem cell is an unspecialized cell with the ability to renew itself indefinitely; under appropriate conditions, stem cells can give rise to a wide range of mature cell types. Two types of stem cells can be distinguished according to their origin and their potential of differentiation: embryonic stem cells (ESCs) and somatic stem cells (SSCs) . ESCs are derived from the early blastocyst and the inner cell mass of the embryo and are able to differentiate into the three germ layer cell types (pluripotentiality) . SSCs are isolated from fetal (after gastrulation) or adult tissues, but classically they are programmed to produce only the cell types that belong to the tissue from which they originate. However, recent studies suggest that SSCs might be able to exhibit more plasticity than previously thought, as they seem able to differentiate into many cell types. These observations were mainly reported when SSCs were grafted in a damaged tissue . This phenomenon, also known as phenotypic plasticity of SSCs, has been described among others for mesenchymal stem cells (MSCs), which under appropriate in vivo or in vitro conditions can adopt a neural fate .
Over the past few years in parallel to stem cell research, there has been a growing interest in the molecular mechanisms involved in cell type determination and differentiation during development. The discovery of intrinsic and extrinsic factors leading to the differentiation into ectoderm, mesoderm, and endoderm during the early stages of development; and the identification of factors that are responsible for further differentiation into more restricted cell types later in development, would allow a better understanding of the mechanisms that control the phenotypic plasticity of stem cells.
Recently, we demonstrated that cultured MSCs from adult rat bone marrow are able to express nestin, an intermediate filament protein that is predominantly expressed during neural development . Two factors seem essential for this nestin expression to occur: the first one is the absence of serum-derived components in the culture medium, and the second one is an in vitro maturation of the MSCs. Indeed, a minimum of 25 population doublings are required for nest into be expressed in 75% of cultured MSCs. In this study, we demonstrated that nestin-positive MSCs, when co-cultured with mouse cerebellar granule (mCG) neurons, are able to differentiate into excitable neuron-like cells. By using quantitative reverse transcription polymerase chain reaction (RT-PCR), we demonstrated that sox and pax transcription factors, as well as the frizzled (wnt receptor) and erbB (neuregulin receptors) receptors, are upregulated in nestin-positive MSCs compared with nestin-negative MSCs (nnMSCs), suggesting a possible role for these genes in the acquisition of a responsiveness to neural fate–inducing signals in a nestin-positive MSC population. Electrophysiological recordings of npMSC-derived neuron-like (MDN) cells show that in co-culture with CG neurons, MDN cells first express functional voltage-gated K+ channels followed after 8 days of co-culture by voltage-gated Na+ channels, allowing firing of action potentials. Finally, several experiments strongly suggest that this neuronal fate in the nestin-positive MSCs is a consequence of a "transdifferentiation" process, although not completely ruling out that cell fusion processes could also contribute to the appearance of neuronal-like cells.
MATERIALS AND METHODS
Neuronal and Astroglial Marker Expression by Nestin-Positive MSCs
MSCs were isolated from femoral and tibial bones of adult rats, propagated in culture, and immunologically and functionally characterized . As we already demonstrated, a nestin-expressing stage is necessary to observe a GFAP+ cell differentiation of MSCs co-cultured with neural stem cells . To observe a possible neuronal differentiation of MSCs, we co-cultured nestin-positive MSCs for 5 days with GFP+ mCG neurons obtained from 3-day-old green mice. After 5 days of co-culture, immunological labelings revealed that MSCs began to express differentiation markers: 40.17% ± 2.95% expressed GFAP (Fig. 1A–C), 19.00% ± 1.58% expressed Tuj1 (Fig. 1D–F), and 19.30% ± 1.38% expressed NeuN (Fig. 1G) (n = 12 for each condition). These results were confirmed by FACS analyses (Fig. 1I,J) as we measured 44.57% ± 3.59% GFAP– and 18% ± 4.1% Tuj1+ cells, data which are not statistically different from that obtained by eye-counting immunolabeled cells. The FACS analyses were performed on GFP– cell populations corresponding to the first spike of cells in Figure 1K. Moreover, the absence of GFAP and NeuN double-labeled cells suggests that nestin-positive MSCs are specifically oriented toward either an astroglial or a neuronal fate (Fig. 1H). Two conditions are needed before MSCs begin to express neuronal markers: first, for the appearance of GFAP+ cells in co-cultures with neural stem cells, MSCs have to express nestin; second, a direct cell–cell contact between nestin-positive MSCs and CG neurons is needed. Indeed, when nnMSCs are co-cultivated with CG neurons or when nestin-positive MSCs and CG neurons are co-cultivated with a physical separation (millicell device), no neuronal markers could be detected (data not shown).
Figure 1. Neuronal and astroglial markers expression by nestin-positive mesenchymal stem cells (MSCs). Nestin-positive MSCs were co-cultured for 5 days with green fluorescent protein (GFP)–positive cerebellar granule (CG) neurons (green). Immunocytochemical labeling and fluorescence-activated cell sorting (FACS) analysis showed that some nestin-positive MSCs expressed neural markers. GFAP (red, A–C) was expressed by about 40% of the nestin-positive MSCs, and Tuj1 (red, D–F) was expressed by about 20% of the nestin-positive MSCs, as was the NeuN marker (red, G). Those results were confirmed by FACS analysis (GFAP, I; Tuj1, J) on a GFP– population (K). A double labeling directed against GFAP (red) and NeuN (blue) (H) showed that nestin-positive MSCs specifically oriented toward an astroglial-like or a neuronal-like fate. Nuclei were counter-stained with EtD1 (blue or red, depending on the color used for the secondary antibody). Arrowheads show neural expression markers by nestin-positive MSCs. Scale bars = 40 μm (A, D, and G–K) and 30 μm (B–C and E–F).
Characterization of Nestin-Positive MSCs
The nestin expression by MSCs coincides with the acquisition by these cells of the ability to respond to extrinsic signals and cues driving their neural differentiation . To get some clue into the molecular mechanism(s) involved in this differentiation, we decided to compare the expression in nestin-positive and nestin-negative MSCs of several factors known to have important roles during normal nervous system development. Using RT-PCR and quantitative RT-PCR, the nestin-positive MSC and nnMSC gene expression profiles of the following factors were compared, both qualitatively and quantitatively: the sox and pax transcription factors, as well as frizzled (wnt receptors), erbB (neuregulin receptors) and the notch-delta signaling pathway. Figure 2 shows that MSCs express sox2, sox9, sox10, sox11, pax6, notch1, notch2, delta1, erbB2, erbB4, and fzd and do not express sox1, sox3, pax3, delta2, and erbB3. In the analyses of fzd genes, we used degenerated primers that hybridize with all fzd genes . The resulting products were cloned using the TA cloning procedure, and random clones were analyzed by sequencing. A BLAST search allowed us to determine the identity of each clone: 80% of fzd receptors in nnMSCs were fzd2, and 20% were fzd1; likewise, in nestin-positive MSCs, we obtained 75% of fzd2, 20% of fzd1, and 5% of fzd5 (data not shown). Finally, quantitative RT-PCR demonstrated that nestin-positive MSCs overexpress sox2, sox10, pax6, fzd, erbB2, and erbB4 when compared with the nnMSCs. We also observed that nestin-positive MSCs overexpress the GFAP gene compared with nnMSCs, but none of them express the neurofilament-light gene.
Figure 2. Molecular characterization of nestin-positive mesenchymal stem cells (MSCs). Reverse transcription polymerase chain reaction (RT-PCR) and quantitative RT-PCR were performed to better characterize the molecular profile of nestin-positive MSCs compared with nestin-negative MSCs (nnMSCs) regarding some genes known to have a role during neuro-ontogenesis. In this case, we observed that nestin-positive MSCs overexpressed sox2, sox10, pax6, erbB2, erbB4, fzd, and GFAP. The results of quantitative RT-PCR are expressed as percentage of gene expression in nestin-positive MSCs compared with the same gene expressed in nnMSCs after normalization with the GAPDH housekeeping gene.
Electrophysiological Analysis of Mesenchymal-Derived Neuron-Like (MDN) Cells
Using fluorescent microscopy, we selected for cells that presented a neuron-like shape (rounded cell body with extended processes, hereafter termed MDN cells), that had incorporated DiD Vybrant (red) and that were clearly GFP– (see Methods). Whole-cell patch-clamp recordings were obtained from 234 MDN cells that had been co-cultured with CG neurons for 4–15 days. Based on the appearance of different voltage-gated current profiles (see next paragraph), three culture time periods were distinguished that, moreover, showed significantly (ANOVA; p < .001) different resting membrane potentials (Vrest): –37.6 ± 3.0 mV for MDNs co-cultured with CGs for 4–6 days in vitro (hereafter noted 5DIV; n = 61), –50.3 ± 2.0 mV for 7–9 days (8DIV; n = 76), and –55.7 ± 2.3 mV for 10–15 days (12.5DIV; n = 97).
Excitability ? The activation of voltage-gated channels at the three developmental stages (5, 8, and 12.5DIV) was assessed by applying to whole-cell patch-clamped MDN cells hyperpolarizing and depolarizing voltage steps from a holding potential (VH) of –80 mV (–80 to –125 mV and –80 to +30 mV in 15-mV increments; Fig. 3A–D). At 5DIV, depolarizing steps positive to –40 mV induced outward currents in 95% of MDN cells that showed little desensitization and were blocked by 10 μM TEA, indicating the activation of potassium voltage-gated channels. Potassium current densities were stable throughout the culture time period (7.25 ± 1.16 pA.pF–1). At 5DIV, no inward currents were observed. Conversely, depolarizing voltage steps positive to –30 mV induced inward currents in 85% of 8DIV MDN cells and in 95% of 12.5DIV MDN cells that rapidly activate and inactivate and that were abolished by a perfusion of 1μM TTX, showing that they are mediated by voltage-activated sodium channels. Between 8DIV and 12.5DIV, no significant differences were found regarding the sodium current densities (8.29 ± 0.87 pA.pF–1 (n = 51) at 8DIV compared with 10.15 ± 1.04 pA.pF–1(n = 59) at 12.5DIV). However—and interestingly—the sodium current densities in MDN cells were two times lower (18.63 ± 2.45 pA.pF–1; n = 21) than the one recorded in co-cultured CG neurons. Consistent with the occurrence of sodium and potassium voltage-gated currents, an injection of 600 nA current into whole-cell patch-clamped and current-clamped MDN cells elicited action potentials in 21.5% of 8DIV MDN cells (n = 34) (Fig. 4A) and in 25% of 12.5DIV MDN cells. These action potentials were always present as single spikes and were reversibly inhibited by a 1-μM TTX application (Fig. 4).
Figure 3. Voltage-gated channel activation in mesenchymal stem cell (MSC)–derived neuron-like (MDN) cells. (A): Typical recordings of single whole-cell patch-clamped and voltage-clamped MDN cells at three culture time periods (5, 8, and 12.5DIV). Hyperpolarizing and depolarizing steps were applied from a holding potential of –80 mV (–80 to –120 mV and –80 to +30 mV in 15-mV increments). The first and third lines show recordings in control conditions, and the second in the presence of 1 μM tetrodotoxin (TTX). The arrows indicate the time where the measures were performed (filled arrows for outward currents, and open arrows for inward currents). (B–D): Current-voltage relationships for outward (triangle; recordings obtained in the presence of 1 μM TTX) and inward (square; recording obtained in the presence of 10 μM tetraethyl ammonium ).
Figure 4. Excitability as shown in a typical recording of a single whole-cell patch-clamped and current-clamped mesenchymal stem cell (MSC) neuron-like (MDN) cell at 8DIV. A depolarizing 600 nA current injection induced the firing of a single action potential that could be reversibly blocked by 1 μM tetrodotoxin (TTX).
Neurotransmitter Effects ? Responses of whole-cell voltage-clamped MDNs (VH = –80 mV) to microperfusion-applied neurotransmitters (1 mM GABA, 100 μM glycine, and 100 μM glutamate) were subsequently measured. As soon as at 5DIV, neurotransmitters elicited inward currents in 98% of MDN cells (n = 78, Fig. 5A). Since no significant differences in current densities were observed between the different culture time periods, measures of current densities were pooled for each neurotransmitter, yielding means of 28.2 ± 2.7 pA.pF–1 for GABA, 61.4 ± 10.6 pA.pF–1 for glycine, and 7.5 ± 2.6 pA.pF–1 for glutamate. The responses elicited by GABA, glycine, and glutamate were mediated by their respective ionotropic receptors—that is, GABAA receptors, glycinereceptors, and AMPA/KA/NMDA receptors—since transmitter-evoked currents were reversibly blocked by specific antagonists of these receptors: namely, by gabazine (10 μM), strychnine (1 μM), and a mix of CNQX and AP-5 (1 μM each), respectively. Before the co-culture of nestin-positive MSCs with CG neurons, inward currents elicited by those neurotransmitters were already present, but only in 35% of nestin-positive MSCs (n = 15, Fig. 5B) and with much lower mean current densities: 0.06 ± 0.02 pA.pF–1 for GABA, 0.07 ± 0.03 pA.pF–1 for glycine, and 0.13 ± 0.05 pA.pF–1 for glutamate. Current densities of neurotransmitter-evoked responses were also measured in CG neurons with which nestin-positive MSCs were co-cultured. In CG neurons, GABA, glycine, and glutamate elicited currents in 100% of the cells (n = 15), with mean current densities of 19.71 ± 3.96 pA.pF–1 for GABA, 26.02 ± 5.55 pA.pF–1 for glycine, and 12.68 ± 4.54 pA.pF–1 for glutamate. Current density comparison between MDN cells and CG neurons yielded a significant difference for glycine responses only (ANOVA; p < .01). Finally, we looked for a possible synaptic activation of the neurotransmitter receptors present in MDN cells. Although MDN cells expressed axonal (MAP2ab, Fig. 6A–C), dendritic (SMI31, Fig. 6D–F), and synaptic markers (synaptophysin, Fig. 6G–I), we never observed spontaneous synaptic events in cultured MDN cells.
Figure 5. Neurotransmitter sensitivities. (A): Typical recording of single whole-cell patch-clamped and voltage-clamped mesenchymal stem cell (MSC) neuron-like (MDN) cell at 8DIV. Either 1 mM GABA, 100 μM glycine, or 100 μM glutamate was applied for 10 seconds every 30 seconds, and the resulting current was recorded in the absence or presence of specific antagonists—namely, 10 μM gabazine, 1 μM strychnine, and a mix of 1 μM CNQX and 1 μM D-AP5, respectively. Antagonists were first applied alone for 10 seconds before being applied for 10 seconds with neurotransmitters. (B): Responses to GABA, glycine, and glutamate in mouse cerebellar granule (mCG) neurons, MSC-derived neuron-like (MDN) cells, and nestin-positive MSCs (before co-culturing with mCG neurons) are expressed as current densities—that is, mean peak amplitudes corrected for cell capacitance (pA.pF–1). **p < .01 (one-way ANOVA followed by Dunnett’s post-tests)
Figure 6. (A–I): Expression of axonal (MAP2ab, blue, A–C), dendritic (SMI31, blue, D–F), and synaptic (synaptophysin, blue, G–I) markers by 8DIV cultured MDN cells.
Mechanisms of Neural Differentiation of Nestin-Positive MSCs ? Recent studies have indicated that MSCs are able to fuse with mature neurons . In this study, three types of experiments were performed which strongly suggest that MSCs differentiate into neuron-like cells rather than fuse with existing CG neurons. The first experiment involved the analysis of the DNA content of nestin-positive MSCs and CG neurons before and after the co-culture. In all cases, we did not observe a significant difference in ploidy, which could explain the 60% of neural differentiation (40% into astrocytes and 20% into neurons) in the MSC population (Fig. 7A). Although the sensitivity of this technique does not allow us to completely rule out that some rare polyploid cells could be present, the majority of the observed cells were diploid. In a second set of experiments, we used the M2 and M6 markers, which specifically recognize respectively astrocytes and neurons from mouse but not from rat . Rat npMSCs were co-cultivated for 5 days with mouse CG neurons. Thereafter, double labeling with GFAP and M2 antibodies, on one hand, and Tuj1 and M6, on the other hand, were performed. We could observe (1) that all GFAP+/ GFP– and Tuj1+/GFP– cells are also negative for M2 and M6, respectively, allowing us to conclude that those cells are of rat origin and thus likely derive from the MSC population (Fig. 7B–G); and (2) that all the cells recognized either by the M2 or by the M6 antibodies are also GFP+, ruling out a downregulation of GFP expression during the co-culture period. Finally, nestin-positive MSCs were cultured for 5 days on paraformaldehyde-fixed GFP+ CG neurons in the presence of CG-conditioned medium. In such conditions, we observed that 16.1% ± 2.6% (n = 8) nestin-positive MSC-derived cells were NeuN+ (Fig. 7H–J, and 23.1% ± 2.1% (n = 7) nestin-positive MSC-derived cells were GFAP+ (Fig. 7K–M). Note that the level of GFP fluorescence in the fixed cells maintained for 5 days in culture remains stable. Although some rare fusion events cannot be excluded in these experiments, we can conclude that most of the MDN cells result from a true differentiation process of nestin-positive MSCs.
Figure 7. Differentiation of nestin-positive mesenchymal stem cells (MSCs) into neural cells. (A): Nestin-positive MSCs (red) and mouse cerebellar granule (mCG) neurons (yellow) before and after co-culture (blue) were stained with propidium iodide and subjected to fluorescence-activated cell sorting (FACS) analysis to determine their ploidy. (B–D): A double-labeling with M2 (red) and GFAP (blue) antibodies allowed the confirmation of the mesenchymal origin of some GFAP+ cells. Nestin-positive MSC-derived GFAP+ cells (C) are not recognized by the M2 antibody (B), but the GFP+ astrocytes (green astrocytes) are recognized by the M2 antibody (D). (E–G): Likewise, a double-labeling with M6 (blue) and Tuj1 (red) antibodies allows the confirmation of the mesenchymal origin of some Tuj1+ cells. Nestin-positive MSC-derived Tuj1+ cells (E) are not recognized by the M6 antibody (F), but the GFP+ neurons (green neurons) are recognized by the M6 antibody (G). (H–M): Finally, we cultivated nestin-positive MSCs onto paraformaldehyde-fixed green fluorescent protein (GFP)–positive mCG neurons in centrifuged and filtrated mCG-conditioned medium for 5 days. GFAP+ cells (red, H–J) and NeuN+ cells (red, K–M) derived from nestin-positive MSCs were observed. Arrowheads indicate the nestin-positive MSC-derived cells that express neural markers. Scale bars (B–C, E–F, H–I, and K–L) 20 μm and (D, G, J, and M) 40 μm.
DISCUSSION
This work was supported by a grant of the Fonds National de la Recherche Scientifique (FNRS) of Belgium, by the Fondation Médical Reine Elisabeth (FMRE), by the Fonds Charcot, and by the Belgian League against Multiple Sclerosis. S.W. is a research fellow of the Télévie (FNRS). B.R. is senior research associate, P.L. a research associate, and G.H. a research fellow of the FNRS. We thank Bernard Coomans (CNCM, University of Liège, Liège, Belgium) for his help with the FACS analysis and Patricia Ernst for her broad competence in cerebellar granule culture.
REFERENCES
Stanford WL, Caruana G, Vallis KA et al. Expression trapping: identification of novel genes expressed in hematopoietic and endothelial lineages by gene trapping in ES cells. Blood 1998;92:4622–4631.
Amit M, Carpenter MK, Inokuma MS et al. Clonally derived human embryonic stem cell lines maintain pluri-potency and proliferative potential for prolonged periods of culture. Dev Biol 2000;227:271–278.
Itskovitz-Eldor J, Schuldiner M, Karsenti D et al. Differentiation of human embryonic stem cells into embryoid bodies compromising the three embryonic germ layers. Mol Med 2000;6:88–95.
Schuldiner M, Yanuka O, Itskovitz-Eldor J et al. Effects of eight growth factors on the differentiation of cells derived from human embryonic stem cells. Proc Natl Acad Sci U S A 2000;97:11307–11312.
Raff M. Adult stem cell plasticity: fact or artifact? Annu Rev Cell Dev Biol 2003;19:1–22.
Kopen GC, Prockop DJ, Phinney DG. Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci U S A 1999;96:10711–10716.
Brazelton TR, Rossi FM, Keshet GI et al. From marrow to brain: expression of neuronal phenotypes in adult mice. Science 2000;290:1775–1779.
Alvarez-Dolado M, Pardal R, Garcia-Verdugo JM et al. Fusion of bone-marrow-derived cells with Purkinje neurons, cardiomyocytes and hepatocytes. Nature 2003;425:968–973.
Wislet-Gendebien S, Leprince P, Moonen G et al. Regulation of neural markers nestin and GFAP expression by cultivated bone marrow stromal cells. J Cell Sci 2003;116:3295–3302.
Azizi SA, Stokes D, Augelli BJ et al. Engraftment and migration of human bone marrow stromal cells implanted in the brains of albino rats: similarities to astrocyte grafts. Proc Natl Acad Sci U S A 1998;95:3908–3913.
Lefebvre PP, Rogister B, Delree P et al. Potassium-induced release of neuronotoxic activity by astrocytes. Brain Res 1987;413:120–128.
Okabe M, Ikawa M, Kominami K et al. "Green mice" as a source of ubiquitous green cells. FEBS Lett 1997;407:313–319.
Shibata T, Yamada K, Watanabe M et al. Glutamate transporter GLAST is expressed in the radial glia-astrocyte lineage of developing mouse spinal cord. J Neurosci 1997;17:9212–9219.
Helmbrecht H, Kispert A, von Wasielewski R et al. Identification of Wnt/?-Catenin signalling pathway in human thyroid cells. Endocrinology 2001;142: 5261–5266.
Weimann JM, Johansson CB, Trejo A et al. Stable reprogrammed heterokaryons form spontaneously in Purkinje neurons after bone marrow transplant. Nat Cell Biol 2003;5:959–966.
Weimann JM, Charlton CA, Brazelton TR et al. Contribution of transplanted bone marrow cells to Purkinje neurons in human adult brains. Proc Natl Acad Sci U S A 2003;100:2088–2093.
Lagenaur C, Schachner M. Monoclonal antibody (M2) to glial and neuronal cell surfaces. J Supramol Struct Cell Biochem 1981;15:335–346.
Isacson O, van Horne C, Schumacher JM et al. Improved surgical cell therapy in Parkinson’s disease: physiological basis and new transplantation methodology. Adv Neurol 2001;86:447–454.
Dunnett SB. Functional analysis of fronto-striatal reconstruction by striatal grafts. Novartis Found Symp 2000;231:21–41.
Hall ED. Pharmacological treatment of acute spinal cord injury: how do we build on past success? J Spinal Cord Med 2001;24:142–146.
Wagers AJ, Weissman IL. Plasticity of adult stem cells. Cell 2004;116:639–648.
Graham V, Khudyakov J, Ellis P et al. SOX2 functions to maintain neural progenitor identity. Neuron 2003;39:749–765.
Wakamatsu Y, Endo Y, Osumi N et al. Multiple roles of Sox2, an HMG-box transcription factor in avian neural crest development. Dev Dyn 2004;229:74–86.
Britsch S, Goerich DE, Riethmacher D et al. The transcription factor Sox10 is a key regulator of peripheral glial development. Genes Dev 2001;15:66–78.
Stolt CC, Rehberg S, Ader M et al. Terminal differentiation of myelin-forming oligodendrocytes depends on the transcription factor Sox10. Genes Dev 2002;16:165–170.
Kocsis JD, Akiyama Y, Lankford KL et al. Cell transplantation of peripheral-myelin-forming cells to repair the injured spinal cord. J Rehabil Res Dev 2002;39:287–298.
Tyas DA, Pearson H, Rashbass P et al. Pax6 regulates cell adhesion during cortical development. Cereb Cortex 2003;13:612–619.
Talamillo A, Quinn JC, Collinson JM et al. Pax6 regulates regional development and neuronal migration in the cerebral cortex. Dev Biol 2003;255:151–163.
Calaora V, Rogister B, Bismuth K et al. Neuregulin signaling regulates neural precursor growth and the generation of oligodendrocytes in vitro. J Neurosci 2001;21:4740–4751.
Canoll PD, Kraemer R, Teng KK et al. GGF/neuregulin induces a phenotypic reversion of oligodendrocytes. Mol Cell Neurosci 1999;13:79–94.
BaoJ, WolpowitzD, RoleLWetal. Back signaling by the Nrg-1 intracellular domain. J Cell Biol 2003;161:1133–1141.
Lemke G. Glial control of neuronal development. Annu Rev Neurosci 2001;24:87–105.
Carleton A, Petreanu LT, Lansford R et al. Becoming a new neuron in the adult olfactory bulb. Nat Neurosci 2003;6:507–518.
Herzog EL, Chai L, Krause DS. Plasticity of marrow-derived stem cells. Blood 2003;102:3483–3493.(Sabine Wislet-Gendebiena,)