Induced Differentiation of Neural Stem Cells of Astrocytic Origin to Motor Neurons in the Rat

Abstract

Destruction of the motor neurons will lead to loss of innervation of the somatic muscle, which has long been considered an illness with no remedy. The only possible treatment is to substitute the injured motor neurons by neurons differentiated from stem cells. It has been recently reported that embryonic stems cells can be induced to differentiate to motor neurons. However, the use of embryonic stem cells has innate problems. The ideal source of motor neurons should be the cells from the patients themselves, which have the potential to be induced to motor neurons. Our previous study demonstrated that mature astrocyte has the potential of being dedifferentiated to neural stem cell. The present study was aimed to investigate if the neural stem cells of astrocytic origin can be induced to motor neurons. The results demonstrated that neural stem cells of astrocytic origin could be induced to differentiate into motor neurons and their progenitor cells with rich harvest. Further, it has been reported that astrocytes can be readily obtained via biopsy from the cerebral cortex of the patient, rendering autologous transplantation possible. In conclusion, matured astrocytes can be induced to motor neurons and be autologously transplanted to patients suffering from motor neuron destruction.

Introduction

T ransplantation of stem cells has been frequently used in animal experiments and has shown benefits for functional recovery of spinal cord injury [1]. Loss of the anterior horn motor neurons in spinal cord injury is an illness with no remedy. The only hope lies on the substitution of the injured motor neurons by neurons differentiated from stem cells. It has been reported that motor neurons can be differentiated from a variety of stem cells [2 –7]. Thus, Wichterle et al. (2002) have successfully induced mouse embryonic stem (ES) cells to differentiate into motor neurons by retinoic acid (RA) and sonic hedgehog (Shh) with a yield of 20%–30% [2]. Li et al. (2008) had showed that the Shh could be replaced by purmorphamine with a higher yield of motor neurons (∼50%) from human ES (hES) cells [8]. A major drawback of using cells from mouse or hES cells is immune rejection, requiring lifelong suppression of immunity [9]. Autologous transplantation could be the better solution to this issue. Astrocytes can be readily obtained via biopsy from the cerebral cortex of the patient [10], and there are studies demonstrating that mature astrocytes can be induced by different types of injuries to dedifferentiate into neural stem cells [10 –12]. It has not yet been studied whether the neural stem cells of astrocytic origin can be induced to form motor neurons by RA and Shh, an issue well deserved to be investigated, and hence, the aim of this study.

In this study, we confirmed that astrocytes from day 15–20 postnatal rat spinal cord, after being cultured in a defined culture system, could dedifferentiate and generate neural stem cells. More importantly, we found that the neural stem cells could be further induced by RA and Shh to motor neuron progenitors and motor neurons with high yields.

Materials and Methods

Cell culture

Postnatal (15–20 days) Sprague Dawley rats were used. The investigation conforms to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996), and the protocol was approved by the Animal Research Committee of Shanghai Jiao Tong University. A longitudinal cut was made along the dorsal root entry zone of the spinal cord and the tissue lateral to the cut was dissected out, thereby excluding the ependyma and periependymal zone, where stem cells and radial cell body reside. The spinal cord tissues were pooled into D-Hank's balanced sodium salts without Ca²⁺ and Mg²⁺ (Gibco) and the meninges and blood vessels were quickly peeled off. After the spinal cord tissue was mechanically minced and further dissociated in D-Hank's containing 0.125% trypsin (Sigma-Aldrich), 0.05% collagenase (Sigma-Aldrich), 30 U/mL papain (Sigma-Aldrich), and 0.03% dispase (Sigma-Aldrich) for 20 min, the pieces were mechanically dissociated by gentle trituration for 40–60 times through a pipette. The cell suspension was then passed through a sterile nylon gauze (80 μm pore size). These dissociated cells were plated onto 60- or 35-mm dishes coated with poly-l-lysine (Sigma-Aldrich) at a density of 2–3 × 10⁵ cells/cm². The cultures were grown in Dulbecco's modified Eagle's medium (DMEM) containing 20% fetal bovine serum (FBS; Gibco), supplemented with 2 mM glutamine, at 37°C in 5% CO₂. Four days later, the cells were cultured following the protocol described by Yang et al. [13] with slight modification, omitting glucose oxidase and glutamate, because our pilot experiments showed that they caused more cell death in longer cultures. Modified chemically defined medium (MCM) was used, which consisted of DMEM without glucose and l-Val (Beijing Tsinghua Skywing Bio Tech) containing 25 mM sorbitol, 94 ng/mL d-Val, and 25 ng/mL epidermal growth factor (EGF; Sigma-Aldrich). The cultures were fed thrice a week by replacement of 1/3 volume of the medium with fresh MCM for 14–18 days. The cells were then resuspended and cultured in a medium for growing neurospheres for 9–12 days. The medium contained DF12 supplemented with 1% N2 and 1% B27, basic fibroblast growth factor (bFGF; Gibco) 20 ng/mL, and EGF 20 ng/mL. The neurospheres were passaged once every week. Neurospheres of passages 2–6 (P2–P6) were used for differentiation. The neurospheres were dissociated into single cells and plated on laminin– (Sigma-Aldrich) and poly-l-lysine–coated coverslips at a density of 5 × 10⁴ cells/cm². For characterization of neurospheres, the cells were incubated with 1% FBS for 7–10 days. For motor neuron differentiation, the cells were incubated for 5 days with recombinant mouse Shh (200 ng/mL; R&D Systems) and RA (1 μM; Sigma-Aldrich). To help the differentiated cells to mature and survive, the cells were further incubated for 14 or more days with 4 neurotrophic factors from R&D Systems: brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor (CNTF, 10 ng/mL), glia-derived neurotrophic factor (GDNF, 10 ng/mL), and neurotrophin-3 (NT-3, 10 ng/mL) in addition to 2% FBS and the concentration of Shh was reduced to 50 ng/mL.

Immunostaining

The cells were fixed in 4% paraformaldehyde for 15 min at room temperature, rinsed with phosphate-buffered saline (PBS), and fixed in methanol for 15 min at −20°C. Nonspecific antibody binding was blocked by incubation with 10% donkey serum in PBS. The cells were then stained overnight at 4°C with the following primary antibodies for different purposes: mouse monoclonal antibodies against Nestin (1: 1,000, MAB353; Chemicon), βIII-tubulin (1:1,000, T8660; Sigma-Aldrich), Map2 (1:1,000, clone HM-2, M4403; Sigma-Aldrich), and O4 (1:80, IgM, MAB345; Chemicon); rabbit antibodies against glial fibrillary acidic protein (GFAP, 1:3,000, G9269; Sigma-Aldrich) and Olig2 (1:500, AB9610; Chemicon); and goat antibodies against HB9 (1:50, sc-22542; Santa Cruz) and choline acetyl transferase (ChAT, 1:200, AB144P; Chemicon). The cells were stained with the corresponding secondary antibodies for 50 min at room temperature: Alexa Fluor 594/488-conjugated donkey anti-mouse (1:800; Molecular Probes), Alexa Fluor 594/488-conjugated donkey anti-goat (1:800; Molecular Probes), and Alexa Fluor 594/488-conjugated donkey anti-rabbit (1:800; Molecular Probes). When necessary, Hoechst 33342 was used at a concentration of 10 μg/mL for 6 min to label the nuclei of cells. The labeled cells were observed under an Olympus confocal microscope (FV1000) or fluorescence microscope (BX-61).

Because of relatively large size of the neurospheres, they had to be sectioned. The neurospheres were collected and fixed in 4% paraformaldehyde overnight, washed with PBS, and put into 25% sucrose overnight. The neurospheres were then cut to 16 μm in a Leica CM3050 cryostat. Staining of the slices followed the procedures described earlier.

Western blot

The protocol of Jin et al. was followed [14]. Cells from the spinal cord of postnatal (15–20 days) SD rats were collected and proteins were extracted by KCl lysis buffer with a complete protease inhibitor cocktail (Roche Applied Science). Thirty micrograms of protein was loaded onto 8% sodium dodecyl sulfate–polyacrylamide gel and resolved by standard sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The proteins were electrophoretically transferred onto PVDF membranes (Roche Applied Science). The membranes were blocked with 1% blocking solution in Tris buffer solution for 60 min and incubated overnight at 4°C with the primary antibodies against GFAP at 1:50,000 dilution and Nestin at 1:500 dilution. Subsequently, the membranes were incubated with POD-labeled secondary antibody (1:12,500; Roche Applied Science). HRP-labeled anti–glyceraldehyde-3-phosphate dehydrogenase at dilution 1:5,000 (Kang Chen) was used as loading control. The signals were revealed by BM Chemiluminescence Western Blotting substrate (Roche Applied Science).

Flow cytometric analysis

Cultured cells were detached with 0.125% trypsin, rinsed, and triturated to single cell suspension in PBS. The cells were then fixed in 2% paraformaldehyde for 15 min on ice, washed, and incubated in PBS–0.1% saponin with 5% donkey serum for 15 min on ice. After washing with PBS–1% FBS, the cells were incubated for 30 min with primary antibodies against rabbit GFAP (1:3,000) and mouse Nestin (1:1,000). Rabbit IgG (Sigma-Aldrich) was used as isotype control. After 2 washes, the cells were incubated for 30 min with Alexa fluor 488 anti-rabbit IgG and R-phycoerythrin–conjugated anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA). The cells were then washed and analyzed using a Becton Dickinson FACSCalibur instrument and CellQuest Pro software. Photomultiplier tube voltage and spectral compensation were established using cells single-stained with Alexa Fluor 488 alone or phycoerythrin alone. Data were collected for at least 10,000 cells per sample.

Transfection

Astrocytes were cultured 2–3 days in MCM on glass-bottomed dishes and then Nestin-eGFP plasmid (pNH163, kind donation of Dr. Naihe Jing [15]) was transfected into the culture by Fugene HD (Roche Applied Science) for 24 h according to the manufacturer's protocol. Two days later, time-lapse images were picked up under a Leica DMI6000B microscope.

Reverse transcript-polymerase chain reaction assay

Total RNA was extracted from motor neuron differentiation cultures using Trizol Reagent (Invitrogen) and then treated with RNase-free DNase (Roche Applied Science). The cDNA was synthesized using the BcaBEST^TM RNA PCR Kit (Takara) according to the manufacturer's instructions. PCR was performed in a final volume of 25 μL containing resulting cDNAs. The following primers were used: rat Olig2, 5′-ACGCCAGCCTGGTGTCTAGT-3′, 5′-TCGGGCTCAGTCATCTGCT T-3′, 304 bp; rat Pax6, 5′-CTTGGTGGTGTCTTTGTCA-3′, 5′-GTCCGT TCAGCATCCTTAGTTT-3′, 424 bp; rat Nkx6.1, 5′-GCGGACCAAGTGGAGAAAGA-3′, 5′-CTCCGCTGGATTTGTGCTTT-3′, 191 bp; rat glyceraldehyde-3-phosphate dehydrogenase, 5′-TCCCTCAAG ATTGTCAGCAA-3′, 5′-AGATCCACAACGGATACATT-3′, 308 bp.

Cell counts and statistical analysis

Data for the percentage of Olig2, βIII-tubulin, ChAT, Map2, and HB9/Hoechst cells were collected from a total of 3 independent experiments. Twenty fields were chosen in a uniform random fashion for quantification. Each field was scored for Hoechst-positive nuclei as the total cells. For each independent experiment, an average of >300 cells were selected. Data are presented as means ± standard error of the mean.

Results

Dedifferentiation of astrocytes in vitro

Our previous study showed that culture of primary astrocytes in MCM is optimal for astrocytes proliferation, and also it inhibits or reduces the population of fibroblasts, oligodendrocytes, neurons, and microglia [13]. In this study, the isolated cells were cultured in DMEM +20% FBS for 4 days before they were transferred into MCM. The cells were immunostained with GFAP and Nestin. Figure 1 shows the process of dedifferentiation of astrocytes from purely GFAP-immunoreactive (ir) to purely Nestin-ir with different degree of coexistence in between. At day 1, only a few double-ir cells can be identified (Fig. 1A). They increased in number at days 4 and 7, concomitant with an increase in the intensity of Nestin immunoreactivity (Fig. 1B, C). The number of double labeled cells and the intensity of GFAP immunoreactivity subsided at day 14 (Fig. 1D). The strongly Nestin-ir bipolar cells with long processes but little GFAP immunoreactivity, which first appear at day 4 (Fig. 1B), are indicative of a more advanced stage of dedifferentiation of astrocytes [16,17]. They increased in number at days 7 and 14 (Fig. 1C, D). Western blotting showed that the spinal cord tissue in postnatal (15–20 days) rats did not express Nestin; the culture cells expressed Nestin at day 3, and higher levels were demonstrated at days 7 and 14 (Fig. 1E). Fluorescence activated cell sorting analysis showed that 95.8% ± 0.8% of the cells expressed GFAP, 71.2% ± 3.8% of the cells expressed Nestin, and 69.3% ± 4.0% of the cells coexpressed Nestin and GFAP (Fig. 1F–H). A challenge to the above study is whether there was contamination of stem cells in the starting material. The frozen cross-sections of the lateral part of the spinal cord of a mature rat (15–20 days postnatal) were GFAP and Nestin immunostained. There was no Nestin-ir cell (Fig. 2A). The fresh dissociated cells were immediately plated with a cytospin device onto a glass slide and double stained with Hoechst and anti-Nestin. Again, no Nestin-ir cells could be identified (Fig. 2B, C). Further, the cells were dissociated before transferred into MCM and then cultured in DF-12 medium with 20 ng/mL EGF and 20 ng/mL bFGF. No formation of neurosphere was observed after 14 days (Fig. 2D), which provided additional evidence of the purity of the starting material.

FIG. 1.

Dedifferentiation of astrocytes in vitro. (A–D) Double immunostaining of GFAP and Nestin. Confocal images of astrocytes cultured in MCM for 1 day (A), 4 days (B), 7 days (C), and 14 days (D). The GFAP/Nestin double-immunoreactivity appears in different shades of yellow. In A, only a few double-ir cells can be identified. They increase in number in B and C, concomitant with an increase in the intensity of Nestin immunoreactivity. In D, the number of double labeled cells and the intensity of GFAP immunoreactivity subside. The bipolar shape of Nestin-ir cells, which appear in B and increase in number in C and D, is indicative of a more advanced stage of dedifferentiation of astrocytes. Scale bars = 20 μm. (E) Western blotting shows Nestin and GFAP expression in the spinal cord tissue of 15–20-day-old postnatal rat and astrocytes cultured in MCM for 3 days (3D), 7 days (3D), and 14 days. (F–H) Fluorescence-activated cell sorting: astrocytes cultured in MCM for 14 days. (F) Isotype control. (G) GFAP/Nestin expression. (H) Fluorescence-activated cell sorting analysis showing the percentages of GFAP, Nestin, and GFAP/Nestin–coexpressing cells. Data are presented as mean ± standard error of the mean; n = 3. GAPDH was used as internal control. GFAP, glial fibrillary acidic protein; MCM, modified chemically defined medium; ir, immunoreactive; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

FIG. 2.

(A) Cross section of the lateral part of the postnatal (15–20 days) spinal cord triple stained with antibodies against GFAP, Nestin, and Hochest. No Nestin-positive cell can be identified. (B, C) Dissociated cells double stained with Hoechst and anti-Nestin. (B) There is no Nestin-ir cell. (C) Fourteen days after being cultured in vitro, many Nestin-positive cells appeared. (D) No formation of neurospheres after being cultured in DF-12 with 20 ng/mL epidermal growth factor and 20 ng/mL basic fibroblast growth factor at 2 weeks. Scale bars: A–C = 20 μm; D = 100 μm.

To prove the genuineness of Nestin immunoreactivity in the above experiments, the cells were transfected with Nestin-enhanced green fluorescence protein (eGFP) vector, which can specially express eGFP in Nestin-expressing cells [15]. Two days after culture, the eGFP-expressing cells could be identified (Fig. 3A). The number of the eGFP-expressing cells increased successively at days 6 and 15, and most of the cells became bipolar with long and thin processes (Fig. 3B, C). For tracing transformation of the cells, 2 days after transfection, time-lapse microphotos were taken from 0 h onward every 30 min till 25 h and 30 min. Figure 3D–F shows 3 photos highlighting a single cell that gained increasing intensity of eGFP expression and changed from polygonal to bipolar.

FIG. 3.

Time-lapse microscopy images showing the transformation of astrocytes transfected with Nestin-eGFP. (A) Day 2; (B) day 6; (C) day 15. (A) A flat polygonal astrocyte expressed enhanced green fluorescence protein (eGFP) at day 2. (B, C) There is successive increase of the number of the eGFP fluorescent cells at days 6 and 15. (D–F″) Morphological transformation and eGFP expression of astrocytes, highlighting single cells that gain increasing intensity of Nestin fluorescence and change from polygonal to bipolar. Scale bar = 50 μm.

Formation and characterization of neurospheres

The cells cultured in MCM for 14–18 days were dissociated and cultured with 20 ng/mL EGF and 20 ng/mL bFGF. Primary neurospheres were found generated at days 9–12. The neurospheres were triturated into single cells, which were then cultured in the same medium. After 5–7 days, secondary (P2) neurospheres were generated and grew robustly (Fig. 4A). Further, the neurospheres were dissociated into single cells and plated onto a 96-well plate at a density of 10 cells/well. The stem cells in each well were checked every day thereafter. None of the stem cells aggregated, and each stem cell formed a small neurosphere by 1 week. After additional 3 weeks, the size of the neurosphere could reach >100 μm in diameter (Fig. 4B). The cells from the neurospheres were successfully passaged once per week to at least 16 passages. After the neurosphere cells were plated at a density of 100 cells/well in 48 wells of the 96-well plate, 22.7 ± 9.9 new neurospheres were formed in each well after 3 weeks, with sizes ranging from 50 to 150 μm. To test the proliferation potential of the neurosphere cells, when 1 × 10⁶ neurosphere cells were plated, (6.6 ± 0.4) × 10⁶ (n = 3) cells were obtained in 7 days. A similar number of cells were generated when neurospheres from passages 2 to 16 were examined. The neurospheres were frozen sectioned for immunostaining. They were full of Nestin-ir cells (Fig. 4C). To examine the potential of differentiation of the neurosphere cells, the P2–P6 neurospheres were triturated into single cells, then seeded onto poly-l-lysine–coated glass coverslips, and cultured for 7–10 days in 1% FBS-containing medium. Under this condition, βIII-tubulin-ir neurons, GFAP-ir astrocytes, and O4-ir oligodendrocytes could all be differentiated, among which the βIII-tubulin-ir neurons occupied a great majority of the population (Fig. 4D, E). Thus, the neurosphere cells met the criteria of neural stem cell.

FIG. 4.

Dedifferentiation of astrocytes to neural stem cells. (A) Robust proliferation of neurospheres from P2 neural stem cell. (B) A neurosphere developed from a neural stem cell in 4 weeks. (C) Nestin immunoreactivity of a neurosphere. (D) βIII-Tubulin-ir neurons and GFAP-ir astrocytes with Hoechst-stained nuclei. (E) O4-ir oligodendrocyte. Scale bars: A = 100 μm; B = 50 μm; C–E = 20 μm.

Differentiation of the neural stem cells into motor neuron progenitors and motor neurons

It has been reported that spinal motor neurons can be successfully induced from ES cells by RA and Shh [2,3,18]. We followed the method of Wichterle et al. [2] with slight modification as described in the Materials and Methods section. The expression of the basic helix–loop–helix transcription factors Olig2, a specific marker for motor neuron progenitor [19], and βIII-tubulin was examined. After 5 days of induction, as many as 97.7% ± 0.5% of the nuclei of the cells were Olig2-ir, and 97.6% ± 0.5% cells were βIII-tubulin-ir (Fig. 5A, B). The merged figure (Fig. 5C) clearly demonstrated that almost all of the Olig2-ir cells were βIII-tubulin-ir. These data showed that these astrocyte-derived neural stem cells can be highly and efficiently induced into Olig2/βIII-tubulin–coexpressing motor neuron progenitors.

FIG. 5.

Differentiation of neural stem cells to motor neuron progenitors. (A–C) The P2–P6 neurospheres were cultured for 5 days with RA and Shh. The cells were stained for βIII-tubulin and Olig2. (A) βIII-Tubulin-ir neurons; (B) Olig2, a motor neuron progenitor marker that is expressed in nucleus; (C) a merge of A and B. Almost all Olig2-ir cells coexpress βIII-tubulin. Scale bar = 20 μm. (D) Reverse transcript-polymerase chain reaction (RT-PCR) displays the expressions of Pax6, Nkx6.1, and Olig2 during the process of motor neuron differentiation. For differentiation of motor neurons, neurosphere cells were cultured for 5 days in the presence of RA and Shh and then the 4 neurotrophic factors (brain-derived neurotrophic factor, ciliary neurotrophic factor, glial-derived neurotrophic factor, and neurotrophin-3), 2% fetal bovine serum, and 50 ng/mL Shh were added and cultured for additional 14 days. Lanes 1, 3, and 5: without RA and Shh treatment for 5, 12, and 19 days; lanes 2, 4, and 6: with RA and Shh treatment for 5, 12, and 19 days, respectively. GAPDH is an internal control. (E) Statistical analysis of RT-PCR results. Number of the lanes are similar to that in D. Data are presented as mean ± standard error of the mean; n = 3. RA, retinoic acid; Shh, sonic hedgehog.

RT-PCR analysis of Olig2 and the homeodomain transcription factors Pax6 and Nkx6.1, which are involved in the specification of motor neurons [20,21], showed that, after treating RA and Shh for 5 days, the Pax6, NKx6.1, and Olig2 expressions were all upregulated, and from 12 to 19 days the expressions decreased successively (Fig. 5D, E). This proves that the neural stem cells are responsive to RA and Shh factors to regulate the Pax6, NKx6.1, and Olig2 expressions, which are essential for motor neuron differentiation.

For differentiation of motor neurons, the neurospheres were first cultured for 5 days in the presence of RA and Shh, and then, for mature differentiation, BDNF, GDNF, CNTF, NT-3, 2% FBS, and Shh reduced to 50 ng/mL were added and cultured for additional 14 days. Most of the cells coexpressed βIII-tubulin and ChAT (Fig. 6A–A”). Immunostaining with a specific motor neuron marker, HB9, which is an homeodomain transcription factor that participates in motor neuron specification and is expressed in postmitotic motor neurons [22,23], demonstrated that all of them coexpressed βIII-tubulin immunoreactivity (Fig. 6B–B”). Figure 6C–C” shows colocalization of ChAT and Map2, a marker for dendrites. Statistical analysis of immunostaining revealed that 93.3% ± 1.7% of the cells were βIII-tubulin-ir, 87.2% ± 1.4% of the cells were ChAT-ir, 27.6% ± 4.6% cells were MAP2-ir, and 24.4% ± 1.1% cells were HB9-ir at 2 weeks (Fig. 6D). The cells without RA and Shh treatment generated significantly fewer HB9-ir motor neurons (4.0% ± 0.4%; P < 0.001; Mann–Whitney test). These data demonstrated that RA and Shh can significantly stimulate the neural stem cells to produce motor neurons. Also, if differentiation time is increased, the yield of motor neuron could be further increased (34.1% ± 2.9% at 4 weeks). Statistical analysis demonstrated that the percentage of ChAT-ir cells is more than that of HB9-ir cells (Fig. 6D), indicating that the ChAT-ir neurons involves populations other than motor neurons.

FIG. 6.

Differentiation of neural stem cells to motor neurons. After being treated with RA and Shh for 5 days, the cells were cultured for additional 14 days with Shh plus brain-derived neurotrophic factor, glial-derived neurotrophic factor, ciliary neurotrophic factor, neurotrophin-3, 2% fetal bovine serum, and 50 ng/mL Shh to promote maturation of motor neurons. (A–A’’) Coexpression of βIII-tubulin and ChAT; (B–B’’) coexpression of βIII-tubulin and HB9; (C–C’’) coexpression of ChAT and Map2. (A, B) Confocal images. Scale bars: A–C = 20 μm. (D) Percentages of βIII-tubulin, ChAT, Map2, and HB9 stained cells at 2 weeks. ChAT, choline acetyl transferase.

Discussion

The motor neurons are the final common path of the somatic nervous system. Destruction of the motor neurons will lead to loss of innervation of the somatic muscle, which has long been considered an illness with no remedy. The advance of stem cell study ignites a hope to substitute the injured motor neurons by neurons differentiated from stem cells. It has been recently reported that ES cells of mouse [2] or human [8] can be induced to differentiate to motor neurons. However, there are drawbacks in using ES cell and heterologous transplantation. An ideal solution could be to dedifferentiate certain type of mature neural cells from the patients themselves to neural stem cells, which can then be induced to differentiate to motor neurons and transplanted autologously. Our previous study has demonstrated that astrocytes from mature rats have the potential to be dedifferentiated into neural stem cells [11]. The present study further proved that astrocytes for mature rats can be successfully induced to motor neurons and progenitor cells, and astrocytes can be readily obtained via biopsy from the cerebral cortex of the patient [10]. There are different cell sources for autologous transplantation. Dimos et al. demonstrated that induced pluripotent stem (iPS) cells generated from patients with amyotrophic lateral sclerosis could differentiate into motor neurons [6]. Takahashi et al. successfully generated iPS cells from adult human dermal fibroblasts with Oct3/4, Sox2, Klf4, and c-Myc [24]. However, the low efficacy of reprogramming primary human cells and the high risk of tumor formation limit its therapeutic application [25]. Huangfu et al. reprogrammed the primary human fibroblasts with only 2 factors, Oct4 and Sox2, omitting the oncogenes c-Myc or Klf4 to reduce the risk of oncogenesis [25]. Olfactory neuroepithelium may be another source of stem cells to differentiate into motor neurons for autologous transplantation in spinal cord injury. Zhang et al. found that simultaneous transfection of Ngn2 and HB9 cDNA into neurosphere-forming cells from cultures of human adult olfactory neuroepithelium generated motor neurons, when the cells were treated with RA, forskolin, and Shh [5]. However, both iPS and olfactory neuroepithelium cells aforementioned as sources of autologous cells all face the safety problem. Takahashi et al. [24] and Huangfu et al. [25] applied gene engineering with retrovirus infection. Zhang et al. [5] introduced cDNA with plasmid. The foreign genes or proteins introduced and the unpredictable changes that may happen in the genome will arouse hesitation in accepting this therapy. But none of these procedures is needed for generating motor neurons from astrocytes in our studies.

All the so far reported transplantation of stem cell-derived motor neurons [26 –28] actually had transplanted a mixture of cells, including at least progenitor cells. Thus, strictly speaking, the final outcome should not be due to motor neuron alone. It is not known what role the progenitor cells may play when transplanted, good or bad by themselves or their interactions with the motor neurons. Also, can the transplanted progenitor cells differentiate into motor neurons? Can the transplanted motor neuron survive in their new environment? Further studies are needed to clarify these issues. Moreover, the motor neurons are generated with many manipulations in vitro. The possibility of changes in phenotype should not be overlooked and should be screened and evaluated before transplantation into the spinal cord can be applied. In summary, this study confirmed our previous studies that mature astrocytes can dedifferentiate to neural stem cells and, for the first time, demonstrated that they can be induced to differentiate to 97.7% of motor neuron progenitors and at least 30% of motor neurons. It offers a valuable alternative for substitution of the lost motor neurons in spinal cord injuries.

Footnotes

Acknowledgments

This study was supported by the National Basic Research Program of China (2003CB515301). Advices from Drs. S.T. Li, W.L. Jin, and L.Z. Gao are highly appreciated.

Author Disclosure Statement

The authors declare no competing financial or commercial interests.

References

Bareyre

. 2008. Neuronal repair and replacement in spinal cord injury. J Neurol Sci, 265:63–72.

Wichterle

, Lieberam

, Porter

, Jessell

. 2002. Directed differentiation of embryonic stem cells into motor neurons. Cell, 110:385–397.

, Du

, Zarnowska

, Pankratz

, Hansen

, Pearce

, Zhang

. 2005. Specification of motoneurons from human embryonic stem cells. Nat Biotechnol, 23:215–221.

Wada

, Honda

, Minami

, Tooi

, Amagai

, Nakatsuji

, Aiba

. 2009. Highly efficient differentiation and enrichment of spinal motor neurons derived from human and monkey embryonic stem cells. Plos One, 4:e6722.

Zhang

, Cai

, Klueber

, Guo

, Lu

, Winstead

, Qiu

, Roisen

. 2006. Role of transcription factors in motoneuron differentiation of adult human olfactory neuroepithelial-derived progenitors. Stem Cells, 24:434–442.

Dimos

, Rodolfa

, Niakan

, Weisenthal

, Mitsumoto

, Chung

, Croft

, Saphier

, Leibel

, Goland

, Wichterle

, Henderson

, Eggan

. 2008. Induced pluripotent stem cells generated from patients with ALS can be differentiated into motor neurons. Science, 321:1218–1221.

Moreno-Manzano

, Rodriguez-Jimenez

, Garcia-Rosello

, Lainez

, Erceg

, Calvo

, Ronaghi

, Lloret

, Planells-Cases

, Sanchez-Puelles

, Stojkovic

. 2009. Activated spinal cord ependymal stem cells rescue neurological function. Stem Cells, 27:733–743.

, Hu

, Jones

, Zhang

, Vaute

, Du

, Zhang

. 2008. Directed differentiation of ventral spinal progenitors and motor neurons from human embryonic stem cells by small molecules. Stem Cells, 26:886–893.

Hentze

, Graichen

, Colman

. 2007. Cell therapy and the safety of embryonic stem cell-derived grafts. Trends Biotechnol, 25:24–32.

10.

Buffo

, Rite

, Tripathi

, Lepier

, Colak

, Horn

, Mori

, Götz

. 2008. Origin and progeny of reactive gliosis: a source of multipotent cells in the injured brain. Proc Natl Acad Sci U S A, 105:3581–3586.

11.

Lang

, Liu

, Feng

, Jiao

, Ju

. 2004. Astrocytes in injured adult rat spinal cord may acquire the potential of neural stem cells. Neuroscience, 128:775–783.

12.

Chen

, Leonq

, Schachner

. 2005. Differential expression of cell fate determinants in neurons and glial cells of adult mouse spinal cord after compression injury. Eur J Neurosci, 22:1895–1906.

13.

Yang

, Liang

, Li

, Cheng

, Luo

, Ju

. 2006. Optimized and efficient preparation of astrocyte cultures from rat spinal cord. Cytotechnology, 52:87–97.

14.

Jin

, Liu

, Yang

, Wang

, Jiao

, Ju

. 2003. Intraneuronal localization of Nogo-A in the rat. J Comp Neurol, 458:1–10.

15.

Jin

, Liu

, Bian

, Chen

, Xu

, Cheng

, Jing

. 2009. Different transcription factors regulate nestin gene expression during P19 cell neural differentiation and central nervous system development. J Biol Chem, 284:8160–8173.

16.

, Cao

, Feng

. 2006. Low temperature induced de-differentiation of astrocytes. J Cell Biochem, 99:1096–1107.

17.

Sharif

, Legendre

, Pre′vot

, Allet

, Romao

, Studler

, Chneiweiss

, Junier

. 2007. Transforming growth factor a promotes sequential conversion of mature astrocytes into neural progenitors and stem cells. Oncogene, 26:2695–2706.

18.

Novitch

, Wichterle

, Jessell

, Sockanathan

. 2003. A requirement for retinoic acid-mediated transcriptional activation in ventral neural patterning and motor neuron specification. Neuron, 40:81–95.

19.

Novitch

, Chen

, Jessell

. 2001. Coordinate regulation of motor neuron subtype identity and pan-neuronal properties by the bHLH repressor Olig2. Neuron, 31:773–789.

20.

Jessell

. 2000. Neuronal specification in the spinal cord; inductive signals and transcriptional codes. Nat Rev Genet, 1:20–29.

21.

Lee

, Pfaff

. 2001. Transcriptional networks regulating neuronal identity in the developing spinal cord. Nat Neurosci, 4,Suppl:1183–1191.

22.

Arber

, Han

, Mendelsohn

, Smith

, Jessell

, Sockanathan

. 1999. Requirement for the homeobox gene Hb9 in the consolidation of motor neuron identity. Neuron, 23:659–674.

23.

Thaler

, Harrison

, Sharma

, Lettieri

, Kehrl

, Pfaff

. 1999. Active suppression of interneuron programs within developing motor neurons revealed by analysis of homeodomain factor HB9. Neuron, 23:675–687.

24.

Takahashi

, Tanabe

, Ohnuki

, Narita

, Ichisaka

, Tomoda

, Yamanaka

. 2007. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 131:861–872.

25.

Huangfu

, Osafune

, Maehr

, Guo

, Eijkelenboom

, Chen

, Muhlestein

, Melton

. 2008. Induction of pluripotent stem cells from primary human fibroblasts with only Oct4 and Sox2. Nat biotechnol, 26:1269–1275.

26.

Lee

, Shamy

, Elkabetz

, Schofield

, Harrsion

, Panagiotakos

, Socci

, Tabar

, Studer

. 2007. Directed differentiation and transplantation of human embryonic stem cell-derived motoneurons. Stem Cells, 25:1931–1939.

27.

Yohn

, Miles

, Rafuse

, Brownstone

. 2008. Transplanted mouse embryonic stem cell-derived motoneurons form functional motor units and reduce muscle atrophy. J Neurosci, 28:12409–12418.

28.

Harper

, Krishnan

, Darman

, Deshpande

, Peck

, Shats

, Backovic

, Rothstein

, Kerr

. 2004. Axonal growth of embryonic stem cell-derived motoneurons in vitro and in motoneuron-injured rats. Proc Natl Acad Sci U S A, 101:7123–7128.