PRODUCTION, CHARACTERIZATION AND EFFICIENT TRANSFECTION OF HIGHLY PURE OLIGODENDROCYTE PRECURSOR CULTURES FROM MOUSE EMBRYONIC NEURAL PROGENITORS

Cell cultures

Neurospheres (Nphs) were prepared from embryos at 14.5 days of gestation obtained from timed pregnant females of either wild type mice or the Plp-EGFP transgenic mouse (Mallon, et al., 2002), as described elsewhere (Kornblum, et al., 2000; Zheng, et al., 2007). In brief, after separation of meninges and cerebellum, cerebrum tissue was mechanically triturated with a 1 ml Gilson pipette until total tissue disaggregation was achieved, filtered through a 70 μm cell strainer (Fisher Scientific, Pittsburg, PA) and plated in 25 cm² plastic culture flasks (2 brains/flask). Nph proliferation media (NPM) was DMEM/F12, B27 neuronal supplement (Gibco, Baltimore, MD) and 10 ng/ml EGF (Sigma, St. Louis, MO). After 48–72 hr, floating Nphs were passaged at a 1:3 ratio in the same medium every 3–4 days.

To produce oligospheres (Olphs), passage 2–6 EGF-Nphs were mechanically dissociated into single cell suspensions and re-suspended in NPM, supplemented with 10 ng/ml platelet derived growth factor (PDGF, Peprotech, Rocky Hill, NJ) and 10 ng/ml bFGF (Peprotech), instead of EGF. After 72 hr, cell aggregates were passaged at a 1:2 ratio every 4–7 days, as with Nphs. Nphs and Olphs at passage 2–6 were used for subsequent experiments.

Oligodendrocyte precursors from either Nphs or Olphs were generated by either mechanical dissociation of spheres with a 1 ml pipette or by chemical disgregation with the mouse NeuroCult dissociation kit (StemCell Technologies, Alberta, Canada). The single cell suspension was filtered through a 70 μm cell strainer and plated on 6 or 24-well poly-D-Lysine (PDL)-coated plates at a cell density of 1.5×10⁴ cells/cm². Cells were maintained in NPM supplemented with 10 ng/ml PDGF/bFGF. Bipolar cells were observed 24 hours after plating, which were highly proliferative when maintained in this medium.

Immunocytochemistry

Nphs, Olphs and oligodendrocyte precursors plated on PDL-coated dishes were initially fixed in 2 % paraformaldehyde (PFA) for 10 min at room temperature (RT) and then 4 % PFA for 15 min, washed three times with PBS and permeabilized with 0.1% Triton X100 for 10 min at RT. After blocking with 3% BSA in PBS for 60 min at RT, the spheres/cells were incubated with primary antibodies at the indicated concentrations overnight (ON) at 4°C. The cells were washed three times with cold PBS incubated with secondary antibodies for 45 min at RT, and washed again before being visualized on a Leica DMR (Wetzlar, Germany) or a Nikon Eclipse TE300 (Melville, NY) microscope. Primary antibodies used in this study included rabbit anti-NG2 proteoglycan (1:300, a kind gift from Dr. William Stallcup, Burnham Institute, San Diego, CA), mouse anti-glial acidic fibrillary protein (GFAP, 1:450, Sigma), rat anti-PLP/DM20 (1:50, clone AA3), mouse anti-β-III-tubulin (1:300, Chemicon, Temecula, CA), rabbit anti-PDGFR-α (1:300, Santa Cruz, CA), rat anti-PDGFR-α (1;300, Upstate, Lake Placid, NY). Anti-rat, anti-rabbit or anti-mouse AlexaFluor secondary antibodies (Gibco) were used at 1:700.

RNA extraction and real-time PCR (qRT-PCR)

Total RNA was isolated from spheres or single cell cultures using Tri-Zol reagent (Gibco). cDNA was generated from 1–3 μg total RNA by Retrotranscriptase II (Gibco) reaction and the selected transcripts were detected and amplified by qRT-PCR in a LightCycler machine (Roche, Indianalpolis, IN) following the manufacturer’s instructions. The specific primers used in this study to detect mRNA expression of indicated genes are listed in Table 1. The results are presented as specific mRNA expression relative to glyceraldehydes-3-phosphatase dehydrogenase (GAPDH), used as internal control. RT-PCR conditions were: first denaturation: 95 °C, 30 seconds, 1 cycle; annealing and amplification: 59°C, 5 seconds; 72°C, 20 seconds respectively, 40 cycles; melting curve: 65°C, 15 seconds, 1 cycle.

siRNA transfection of Nphs

Passage 2–6 EGF-Nphs were dissociated to single cell suspensions and transfected by electroporation with the Amaxa system according to the manufacturer’s instructions for mouse neural precursors (Amaxa, Gaithersburg, MD). Cells (3×10⁶) were transfected in each experiment with a transfection efficiency of 55–70% as assessed by GFP expression. To achieve effective inhibition of PDGFRα expression, a combination of three commercially available PDGFRα siRNAs (0.2 nm, based on PDGFRα NCBI GenBank accession number NM_011058) were obtained from Ambion (Austin, TX), whose sequences are based on their design algorithm. Sense 1: GGU ACA GCUUAUGGA UUA ATT; antisense 1: UUA AUC CAU AAG CUG UAC CTT; sense 2: GCG AGU UUA AUGUUU AUG CTT; antisense 2: GCA UAA ACA UUA AAC UCG CTG; sense 3: CCG AUU CCU UAG UCA UCG UTT; antisense 3: ACG AUG ACU AAG GAA UCG GTC. Control transfections were performed with an unrelated siRNA negative control (Ambion). Immediately after transfection, cells were plated in PDGF/bFGF to generate oligodendrocyte precursors.

Statistical analysis

For statistical analysis, paired Student t-test was conducted. The results are shown as mean ± SEM and mean ± SD as indicated.

Differential impact of growth factors on cells generated from embryonic neural precursors

In order to generate oligodendrocyte lineage cells from Nphs, we analyzed the effect of different growth factors known to impact the development of oligodendrocyte precursor cells, separately or in combination. Thus, after mechanical dissociation of the EGF-Nphs, the EGF-containing culture medium was replaced with medium supplemented with either basic fibroblast growth factor (bFGF) or platelet derived growth factor (PDGF) or a combination of both, and the de novo formation of spheres was evaluated by phase contrast microscopy.

After 48 hours in media supplemented with bFGF, we observed cell aggregation similar to the EGF-Nphs, forming rounded, yellow spheres that proliferated rapidly and could be passaged and maintained for long periods (Figure 1C). PDGF alone on the other hand, was unable to sustain either cell aggregation or survival. Only a few spheres survived after 48 h in media supplemented with this growth factor (Figure 1E). The addition of PDGF to the bFGF medium was able to induce de novo formation of spheres that were passaged several times and maintained in vitro for long periods (Fig 1G). Interestingly, after attachment of the PDGF/bFGF spheres to poly-D-lysine (PDL)-coated dishes, bipolar, elongated cells with an oligodendrocyte progenitor-like morphology were produced, which migrated from the spheres in greater numbers than those from bFGF-grown spheres (Figure 1 H vs. 1 D). Thus, the cells from bFGF-treated spheres showed some migratory behavior (Figure 1D), but this growth factor appeared insufficient for longterm survival of plated cells since a great number of dead, floating cells was observed (not shown). The PDGF-supplemented cells neither attached nor survived to any significant extent (Figure 1F). Since PDGF/bFGF-treated spheres had greater oligodendrocyte progenitor-like cells with increased migratory capabilities (see below), we called them oligospheres (Olph).

We took advantage of our Plp-EGFP transgenic mouse (Mallon, et al., 2002), which facilitated the identification of oligodendrocyte lineage cells generated by the neural progenitors. We found that a number of cells forming the primary EGF-Nphs as well as subsequent passaged EGF-Nphs showed green fluorescence indicative of Plp promoter activity (Figure 1I, J). Nevertheless, the Plp promoter activity was stronger in Olphs as early as 48 hours in PDGF/bFGF (not shown), and was greatly increased after 72 hr (Figure 1K, L). The few cells migrating from the EGF-Nphs and most of the cells migrating from the Olphs also showed Plp-EGFP expression (data not shown). These results suggest that the combination of PDGF/bFGF induced the commitment of the neural progenitors to the oligodendrocyte lineage, and that these oligodendrocyte lineage progenitors were highly motile.

Cells in oligospheres increased glial gene expression

We confirmed the glial and oligodendrocyte precursor commitment of cells in Olphs by analyzing mRNA expression patterns of specific astrocyte and oligodendrocyte lineage cell proteins, myelin proteins and transcription factors. Samples from Olphs in suspension were compared to EGF-Nphs by qRT-PCR. We first analyzed the expression of cell membrane receptors that might be implicated in the production of Olphs. As expected, we found a great increase in PDGFRα mRNA in Olphs (6 fold), while FGFR1 and 3 did not show significant variation as compared with EGF-Nph. Interestingly, FGRF2 expression appeared to be downregulated in Olphs, as was EGF receptor expression (EGFR, Figure 2A).

More importantly, expression of several cell-specific protein markers helped to identify the cell types in these samples. While the intermediate filament protein nestin, which is specific for undifferentiated progenitor cells (Jori, et al., 2003), showed little variation between Olphs and EGF-Nphs, the expression of the astroglial protein GFAP increased (3.5 fold). There was also a significant increase in expression in the mRNAs for the oligodendrocyte precursor cell marker NG2 proteoglycan (2.3 fold) and for the myelin protein PLP (2 fold). These increases were highly statistically significant.

The transcription factors described to actively regulate the generation, migration and maturation of oligodendrocyte precursors during development showed an interesting expression pattern in Olphs. The increased expression of Olig1 (3-fold) and Olig2 (1.8-fold) along with Sox9 (3-fold) confirmed the oligodendrocyte lineage of cells in Olphs (Wu, et al., 2006; Jakovcevsky and Zecevic, 2005; Ligon, et al., 2006; Liu, et al., 2007; Stolt, et al., 2003). Additionally, the transcription factors of the homeobox family implicated in the regulation of differentiation of progenitors into oligodendrocyte precursors (Nkx2.1, Tekki-Kessaris, et al., 2001) and the maturation of oligodendrocyte precursors (Nkx2.2, Fancy, et al., 2004; Kitada and Rowitch., 2006) showed a significant increase (1.25- and 1.80-fold respectively; p<0.05) (Figure 2C). Other transcription factors involved in oligodendrocyte production during development were also altered in expression (Corbin, et al., 2003; Wei, et al., 2005; Wang, et al., 2001; Liang, et al., 2005). The transcription factors GHS2, SP1, ID2 and ID4 showed significant increases in Olph (1.7-, p<0.01; 1.5-, p<0.05; 2.5-, p<0.01; and 1.4-fold, p<0.05, respectively), while EMX1 and SP3 expression had no statistically significant change relative to EGF-Nphs (Figure 2D). EMX1 has been shown to be expressed in oligodendrocyte progenitors arising in postnatal cortex (Kessaris et al., 2006). Presumably the progenitors derived in the current studies are not in the same pool as the EMX1-expressing cells that arise later in development. Interestingly, mRNA concentration for bone morphogenic protein 2 (BMP2), which drives cells to astrocyte differentiation at the expense of oligodendrocytes (Mabie, et al., 1999; Grinspan, et al., 2000; Gossrau, et al., 2007), did not change, although ID4 and particularly ID2 mRNAs were upregulated (1.4-, p<0.05 and 2.6-fold, p<0.01, respectively).

Oligodendrocyte precursor cells were highly migratory in spheres grown in PDGF/bFGF

In order to evaluate the glial lineage of the cells produced in EGF-Nphs and Olphs, we analyzed expression of PDGFR-α and GFAP as specific markers for early oligodendrocyte precursors and astrocytes respectively in plated spheres. After 48 hours of plating, EGF-Nphs plated in EGF produced for the most part non-migratory cells that accumulated around the sphere and a few migratory PDGFR-α positive cells (Figure 3A–C). It was noticeable that most of the cells within and immediately surrounding the sphere were GFAP-labeled (Figure 3B, C). On the other hand, the cells generated by Olphs actively migrated long distances from the sphere and, as judged by the generalized PDGFR-α staining, these cells were oligodendrocyte precursors (Figure 3G–I). The shift to highly migratory oligodendrocyte progenitor cells occurred quickly in the spheres. We evaluated the impact of treating EGF-Nphs with PDGF/bFGF immediately after plating on PDL-coated dishes. As shown in Figure 3(D–F), a large number of migratory cells were observed 48 hours after plating EGF-Nphs in PDGF/bFGF, which had oligodendrocyte precursor morphology and PDGFR-α positive staining (Figure 3D). We also detected a dense core of GFAP-positive cells indicating the conserved multipotentiality of these spheres regardless of the elevated number of oligodendrocyte precursors generated in response to PDGF/bFGF (Figure 3E, F). In all samples, the GFAP-positive cells remained in the dense core, while the oligodendrocyte precursors migrated away.

To characterize the migratory capabilities of cells originating from spheres, we analyzed the migration by time lapse for 25 hours. Migration from EGF-Nphs plated and maintained in EGF was not imaged, because migration was minimal, but cells from EGF-Nphs plated in PDGF/bFGF or cells from Olphs were highly migratory and were imaged. This allowed us to observe behavioral differences between the cells produced under both conditions (Figure 4). EGF-Nph-derived cells, which were being shifted from EGF to PDGF/bFGF medium at plating, showed random motility, with cells migrating erratically back and forth around the sphere. After movement away from the sphere, we observed some cells undergoing cell division followed by continued unpredictably migration (Figure 4A, cell 2). On the other hand, the cells produced by Olphs, which had already been exposed for several days to PDGF/bFGF and already had increased expression of oligodendrocyte markers before plating (Figure 3), migrated in a slow, organized fashion heading directly out of the spheres in a radial movement (Figure 4B).

Dissociation of Oligospheres and EGF-Neurospheres into single cells cultured in PDGF/bFGF resulted in a highly pure oligodendrocyte precursor culture

In order to assess how efficient the conversion of multipotential progenitors to oligodendrocyte progenitors was, the relative number of oligodendrocyte progenitors in single cell preparations from dissociated Olphs or EGF-Nphs was quantified. Passage 2–6 Olphs or EGF-Nphs from our Plp-EGFP transgenic mouse were mechanically or chemically dissociated, filtered and plated in media supplemented with PDGF/bFGF. In dispersed cell suspensions from both types of spheres we observed attached cells of rounded to elongated morphology immediately after plating. Twenty four hours later, the cell population increased, forming a homogenous cell culture composed of bipolar, elongated cells that were already extending processes and showed Plp-EGFP expression (Figure 5A, B). The expression of PLP was further confirmed by immunocytochemistry with an antibody directed against PLP/DM20 showing colocalization with Plp-EGFP (Figure 5C, D). After 4 days, the cells approached confluence and a few cells showed multiprocess morphology typical of spontaneously maturing oligodendrocytes (Figure 5E, F).

Most cells derived from Olphs (Figure 5G, I) or EGF-Nphs plated in PDGF/bFGF (Figure 5H, J) were positive for PDGFR-α and NG2 (Figure 5G–J), confirming their oligodendrocyte lineage. We also detected some GFAP-positive cells, which appeared to be more abundant in EGF-Nph-derived oligodendrocyte precursor cultures (Figure 5H), relative to cells generated from Olphs (Figure 5G). By contrast, no cells expressing the neuronal markers β-III tubulin or phosphorylated neurofilament were detected in cultures derived from Olphs, although a few such cells were occasionally seen in cultures derived from EGF-Nph (data not shown). We quantified the purity of Olph- and EGF-Nph-derived oligodendrocyte progenitors, by counting NG2-positive cells as a percentage of the total number of cells. EGF-Nph-derived cultures contained 77.9 ± 11.2 % NG2+ cells, and this percentage increased to 96.8 ± 1.6 % in the oligodendrocyte progenitor cultures obtained after Olph dissociation. It should be noted that the current studies were done with Nphs obtained from E14.5 embryos. This was the optimal time point for highly pure oligodendrocyte production. For example, Nphs obtained from E15.5 embryos and treated in the same manner as these generated far more astrocytes (not shown). In general, while the purity of oligodendrocyte progenitor cells was higher in cells obtained from Olphs, the total number of cells tended to be fewer. This correlated with a slower sphere growth rate in Olph which required 4–7 days in culture before every passage while EGF-Nph needed to be subcultured every 3–4 days (data not shown). Thus, for high level production of oligodendrocyte progenitor cells, plating of dissociated EGF-Nph cells in PDGF/bFGF was preferable where a small contamination of astrocytes was not a problem.

To assess gene expression in our single cell cultures, we performed qRT-PCR using specific primers for the cell membrane receptors PDGFR-α, EGFR and FGFR1, along with the transcription factors Olig1, Olig2, Sox9 and Nkx2.1. Total RNA was isolated from single cell preparations generated from EGF-Nphs plated in EGF, or EGF-Nphs and Olphs plated in PDGF/bFGF. We found that the cells obtained from EGF-Nph showed low, yet detectable mRNA expression levels of the genes studied, suggesting the presence of a small number of such cells in EGF-Nphs. Notably, Olph-derived oligodendrocyte precursors showed a remarkable increase in the expression of PDGFR-α and a clear increase in FGFR1 as compared to the cells from EGF-Nph plated in PDGF/bFGF. As expected, EGFR showed decreased expression as a consequence of the growth factor change from EGF to PDGF/FGF (Figure 6A).

The expression of the oligodendrocyte-specification transcription factor Sox9 was clearly upregulated in cells obtained from Olphs as well as the early oligodendrocyte precursor specific genes Olig1, Olig2 and Nkx2.1 (Figure 6B).

Effective RNA silencing of PDGFRα in oligodendrocyte precursors generated from neural progenitors

In order to assess the relative importance of the dramatic increase in PDGFRα after our cells were exposed to PDGF/bFGF, we studied the effect of siRNA-mediated knock-down of PDGFRα on the generation, proliferation and oligodendrocyte protein expression of oligodendrocyte precursors generated from Nphs.

Dissociated Nphs were transfected with a combination of three PDGFRα siRNAs, plated and grown for 7 days in PDGF/bFGF. PDGFRα knock-down was demonstrated by immunocytochemistry (Figure 7A, B). There was a decrease of 29.1 % in total cell numbers as compared to control siRNA transfected cells (Figure 7E). This resulted from reduced cell proliferation, since there was reduced BrdU incorporation in PDGFRα siRNA-transfected cells (61.8 ± 2.4 control siRNA-transfected versus 52.2 ± 3.2 PDGFRα siRNA-transfected BrdU⁺ cells as a percentage of total cells, a reduction of approximately 15% in proliferative rate, Figure 7F). The PDGFRα knock-down appeared to impact cell proliferation but not oligodendrocyte precursor specification. As observed in Figure 7F, the percentage of total cells that were NG2-positive and Olig2–positive cells was comparable for both control siRNA- and PDGFRα siRNA-transfected cells (Figure 7F). On the other hand, the morphology of the NG2-positive cells in PDGFRα siRNA-transfected cells appeared flattened, relative to control siRNA-transfected cells (Figure 7C, D arrows), suggesting that in addition to impacting oligodendrocyte progenitor cell proliferation, this growth factor receptor also impacted interaction of the cell with the substrate.

Oligodendrocyte progenitor cells obtained from embryonic Olphs and EGF-Nphs mature in response to ciliary neurotrophic factor

To further confirm the functionality of our highly pure mouse oligodendrocyte precursor cultures, we tested their ability to mature when exposed to ciliary neurotrophic factor (CNTF), a well characterized inducer of oligodendrocyte maturation (Mayer et al, 1994; Stankoff et al., 2002). After 4 days in PDGF/bFGF, EGF-Nph- or Olph-derived oligodendrocyte precursors were treated with 10 ng/ml CNTF for 48 hours. In Plp-EGFP cells, clear morphological changes were induced by CNTF, which were characteristic of oligodendrocyte precursor maturation. These included the formation of thick, extended cell processes and the production of membranous structures characteristic of in vitro maturing oligodendrocytes. The morphological changes were accompanied by an increase in Plp-EGFP fluorescence, suggesting increased Plp gene expression induced by CNTF (Figure 8A, B). Immunocytochemistry after 7 days of CNTF treatments for the oligodendrocyte antigens O4 (Figure 8C, D) and myelin basic protein (MBP) (Figure 8E, F) further confirmed the maturation effect of CNTF on these oligodendrocyte progenitor cells.