Identification of Sox17 as a Transcription Factor That Regulates Oligodendrocyte Development

Antibodies.

The following antibodies were used: NG2 (1:1000; Chemicon, Temecula, CA), A2B5, O4, and O1 (1:10 on cultures or 1:50 on sections; all from American Type Culture Collection, Manassas, VA), LB1 (1:10; Dr Giulio Levi, Istituto Superiore di Sanita, Rome, Italy) (Levi et al., 1986), anti-Nkx2.2 (NK2 transcription factor related, locus 2) (1:25; Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA), anti-GFAP (1:500; Sigma, St. Louis, MO), anti-neuronal-specific nuclear protein (NeuN) (1:500; Chemicon), anti-Olig2 (1:100,000; Dr. David Rowitch, Dana-Farber Cancer Institute, Boston, MA), anti-CNP (1:500; Covance/Sternberger Monoclonals, Lutherville, MD), anti-galactocerebroside (GalC) (1:50; Chemicon), anti-bromodeoxyuridine (BrdU) (1:20; DakoCytomation, Carpinteria, CA), anti-Ki67 (1:500; Novocastra Lab, Newcastle, UK), anti-active caspase-3 (1:500; Cell Signaling Technology, Beverly, MA), anti-p27 (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), anti-myelin basic protein (MBP) (1:500; Sternberger Monoclonals), and anti-GFP (1:1000; Chemicon). The anti-Sox17 antibody was an anti-recombinant Sox17 antiserum (Kanai et al., 1996) and was used at 1:6,000 (Kanai-Azuma et al., 2002). The specificity of this antiserum was confirmed by performing (1) immunoblotting on the protein extracts of COS cells transfected with pCDM/Sox17 and (2) immunostaining on COS cells and rodent adult testis tissue sections (Kanai et al., 1996). A band corresponding to Sox17 and nuclear localization were detected in Western blot and immunofluorescence analysis, respectively (Kanai et al., 1996). In immunofluorescence, the Sox17 signal was abolished by coincubation of antiserum with the immunogen (glutathione S-transferase–Sox17) (Kanai et al., 1996). In addition, Western blotting of COS7 cells transiently transfected with cytomegalovirus (CMV)–Sox17 or CMV5Sox10 expression plasmids showed no difference between our antibody and the R & D Systems (Minneapolis, MN) anti-human Sox17 antibody in specifically detecting a 60–70 kDa band (supplemental Fig. S1A, available at www.jneurosci.org as supplemental material). These antibodies did not cross-react with Sox10 (supplemental Fig. S1A, available at www.jneurosci.org as supplemental material), which is detected as a 50 kDa band. Furthermore, both anti-Sox17 antibodies showed very similar immunohistochemical staining in postnatal day 5 (P5) mouse spinal cord white matter (supplemental Fig. S1B,C, available at www.jneurosci.org as supplemental material).

Transgenic mice.

The CNP–EGFP transgenic mice have been described previously (Yuan et al., 2002; Belachew et al., 2003; Aguirre et al., 2004). All animal procedures were performed according to the Institutional Animal Care and Use Committee, Children's National Medical Center, and National Institutes of Health guidelines.

Fluorescence-activated cell sorting.

Brain tissue dissociation and FACS procedures were as described previously (Belachew et al., 2003; Aguirre et al., 2004).

Immunocytochemistry and cell counting in FACS-purified cells.

FACS-purified cells were plated on poly-lysine-coated (5 μg/ml) 12 mm glass coverslips at a density of 30,000 cells per coverslip and stained 1 h after plating with NG2, O4, and O1 antibodies as described previously (Yuan et al., 2002; Belachew et al., 2003). After staining, images were acquired using a 40× objective, and an average of 20–35 microscopic fields were counted. The number of total cells counted ranged between 420 and 1150.

Cell cultures.

Purified rat cortical OPC cultures were prepared as previously described from embryonic day 20 (E20) Sprague Dawley rats, using a standard experimental protocol (McCarthy and de Vellis, 1980) with slight modifications (Ghiani et al., 1999a,b). OP cells were plated at a density of 200,000 cells per dish on poly-ornithine-coated 30 mm plastic Petri dishes and cultured in Dulbecco's modified Eagle's DME–N1 supplemented with 10 ng/ml platelet-derived growth factor (PDGF) (human AB, heterodimer form; Upstate Biotechnology, Lake Placid, NY), unless otherwise stated.

mRNA expression profiling.

Approximately 1 million of FACS-isolated CNP–EGFP⁺ cells were homogenized in RNeasy lysis buffer (Qiagen, Valencia, CA). Total RNA was extracted following the standard protocol of RNeasy Mini kit (Qiagen). DNA was digested with DNase I (Ambion, Austin, TX), followed by clean-up using RNeasy Mini kit. Double-stranded cDNA was synthesized from 0.3 μg of total RNA with an oligo-dT primer containing T7 RNA polymerase promoter (Geneset, La Jolla, CA) and SuperScript Choice system (Invitrogen, Carlsbad, CA). After purification by ethanol precipitation, the double-stranded cDNA was resuspended in 8 μl of DEPC H₂O. First-round cRNA was synthesized from double-stranded cDNA by in vitro transcription using a MEGAscript T7 transcription kit (Ambion). A portion (0.2 μg) of purified cRNA was subjected to a second round of cDNA synthesis using a random pd(N6) primer (Roche, Indianapolis, IN) to synthesize the first-strand cDNA and a T7-(dT)₂₄ primer for the second-strand cDNA. The resulting double-stranded cDNA was purified via a phase lock gel (Brinkman Instrument, Westbury, NY), followed by ethanol precipitation. Biotinylated cRNA was produced by in vitro transcription using BioArray HighYield RNA Transcript Labeling Kit (Enzo, New York, NY). Second-round cRNA was purified and then fragmented. Biotinylated cRNA (15 μg) was hybridized to murine U74Av2 oligonucleotide microarray (12,488 probe sets; Affymetrix, Santa Clara, CA), and signal intensity was calculated using Affymetrix GeneChip software MAS 5.0 as described previously (Natale et al., 2003). Each microarray underwent a stringent quality control evaluation as reported previously (Natale et al., 2003). All values obtained from the microarrays in this experiment fell within the quality control range, including the following: cRNA fold change from two rounds of amplification ranged from 400- to 700-fold, scaling factors ranged from 0.4 to 0.9, percentage of probe sets reliably detected (“present”) ranged from 42 to 50%, mean signal value indicating the relative abundance of a probe set ranged from 1700 to 2200, and correlation coefficient of mean signal values for each transcript between microarrays at the same experimental time point ranged from 0.92 to 0.98.

Data filtering and statistical analysis.

As described previously, all analyses of mRNA expression is based on the subset of probe sets detected (present) on at least one of the eight total microarrays used in this experiment (Di Giovanni et al., 2003; Natale et al., 2004). The levels for signal sensitivity detection and signal-to-noise ratio are optimized for each project (Tumor Analysis Best Practices Working Group, 2004). Based on the known timing of oligodendrocyte development, we designed our study to examine oligodendrocyte mRNA expression at four key time points: P2, P8, P15, and P30. However, because developmentally regulated genes may be reliably expressed at only one of these selected stages, a less stringent data-filtering strategy was chosen, which provides the more important benefit of identifying genes that are significantly regulated at one stage despite increasing the number of falsely positive regulated genes. Experiment normalization and statistical analysis were performed using GeneSpring software, version 5.0 (Silicon Genetics, Redwood, CA). Signal intensity for each probe set was normalized to the median intensity across all chips in the experiment.

Immunohistochemistry and confocal microscopy.

Freshly cut, floating tissue sections (50 μm) from P15 CNP–EGFP transgenic mice were prepared as described previously (Yuan et al., 2002; Aguirre et al., 2004). Briefly, sections were incubated for 1 h at room temperature with blocking solution (10% normal goat serum, 1% bovine serum albumin, and 0.3% Tween 20 in 1× PBS, pH 7.4) and then incubated at 4°C overnight with primary antibody. Sections were washed with carrier solution (1% normal goat serum, 1% bovine serum albumin, and 0.3% Tween 20 in 1× PBS, pH 7.4) and then incubated for 1 h at room temperature with secondary antibody (goat anti-rabbit). Slices were washed with carrier solution before mounting in Mowiol. For analysis in tissue sections, a Bio-Rad (Hercules, CA) VMRC 1024 confocal laser-scanning microscope equipped with a krypton–argon laser and an Olympus Optical (Melville, NY) IX-70 inverted microscope were used to image localization of FITC (488 nm laser line excitation; 522/35 emission filter), Texas Red (568 nm excitation; 605/32 emission filter), or cyanine 5 (647 excitation; 680/32 emission filter). Optical sections (Z = 0.5 μm) of confocal epifluorescence images were sequentially acquired using a 40× [numerical aperture (NA) of 1.35], a 60× (NA of 1.40), or a 100× (NA of 1.35) oil objective with Bio-Rad LaserSharp version 3.2 software. Confocal Assistant 4.02 and NIH ImageJ1.26t (http://rsb.info.nih.gov) software were subsequently used to merge images and to perform orthogonal studies, respectively. Merged images were processed in Photoshop 7.0 (Adobe Systems, San Jose, CA) with minimal manipulation of contrast.

Fluorescence intensity or luminosity was determined by drawing a region (box) around an Nkx2.2⁺, O4⁺, or CNP⁺ cell body and measuring the mean luminosity value from the Histogram command in the Image pull-down menu of Adobe Photoshop or NIH ImageJ software. Cell processes were not included, and the whole cell body was taken, regardless of Sox17 distribution. The histogram feature calculates the average luminosity levels and SD of the pixels in the selected field. Recording and analysis of the digital images were performed with fixed settings. Background values taken from unlabeled areas were subtracted from Sox17 values for every section or field analyzed and averaged in each case to control for staining variation. Six separate sections were analyzed for each case, and SigmaStat software (Systat Software, Point Richmond, CA) was used for statistical analysis, using Student's t test to determine significance levels. Values were expressed relative to Sox17 readings in Nkx2.2⁺ cells. For cell counting of endogenous EGFP⁺ cells in P15 subcortical white matter (SCWM), we determined that 100% of the EGFP⁺ cells were Olig2⁺, confirming that cells being analyzed belonged to the OPC lineage. The analysis of the SCWM was performed on coronal sections. An average of four to five sections from three different brains (9–12 sections total) were counted for the SCWM to obtain an estimate of the total number of EGFP⁺ cells. The anatomical distribution of CNP–EGFP⁺ cells was analyzed in Z-series confocal scanning images [20–30 μm thickness; step size of 0.5 μm between successive images of the same field (228 μm²)]. Cell counting data in tissue sections was expressed as averages ± SEM. Statistical analysis was performed by paired t test.

For immunocytochemical analysis of Sox17 in spinal cord, embryos between days 8.5 and 16.5 (days of vaginal plug was designated as day 0.5) were dissected in cold PBS and fixed in 4% paraformaldehyde (PFA) overnight. For postnatal stages P0–P5, animals were perfused with 4% PFA. Spinal cords were dissected, cryoprotected in 20% sucrose overnight, and embedded in TissueTek OCT compound (Sakura, Tokyo, Japan) according to standard protocols. Coronal cryostat sections (10–12 μm thickness) were performed using a Reichert-Jung cryostat (Leica, Nussloch, Germany) and collected on Superfrost plus slides. For immunohistochemistry, slides were rehydrated in PBS with 0.1% Triton X-100 and 10% v/v normal goat serum for 1 h at room temperature before incubation with the following primary antibodies: rabbit polyclonal anti-EGFP (1:400; Invitrogen) for studies with the CNP–EGFP transgenic mice, mouse monoclonal anti-adenomatous polyposis coli (APC) (1:200, clone CC1, antibody 7, 1:200; Oncogene Sciences, Uniondale, NY), mouse monoclonal anti-Nkx2.2 (1:2; Developmental Studies Hybridoma Bank), and rabbit polyclonal anti-Sox17 (Kanai et al., 1996) (1:500) or goat polyclonal anti-Sox17 (R & D Systems). The sections were incubated with primary antibodies diluted in PBS, 10% v/v normal goat serum, and 0.1% Triton X-100 overnight at room temperature. Subsequently, the sections were washed three times for 5 min in PBS and incubated with adequate volumes of secondary antibodies diluted in PBS with 10% normal goat serum for 2 h at room temperature. During incubation with secondary antibodies, cell nuclei were labeled with bisbenzimide (diluted 1:10⁵; Sigma). After three additional washes in PBS, the sections were mounted in Fluoromount (Southern Biotechnology, Birmingham, AL).

In spinal cord, quantification of cells expressing Sox17 at various stages of the oligodendrocyte lineage was performed with NIH ImageJ software using 10 serial cord sections (100 μm apart) and at least three slides for each stage analyzed. Sox17 fluorescence intensity was measured at Nkx2.2, APC/CC1, and CNP stages of the oligodendrocyte lineage with NIH ImageJ software. For these experiments, double immunolabeling for Sox17 and Nkx2.2 or APC/CC1 was performed on coronal spinal cord sections from P5 CNP–EGFP transgenic mice. For each oligodendrocyte marker, Sox17 fluorescence intensity was expressed as the averaged value of the measurement performed on at least 100 labeled nuclei. Results are expressed as a percentage of Nkx2.2 fluorescence intensity.

Immunocytochemistry in cultured oligodendrocyte lineage cells.

Live staining for cell surface antigens with LB1, O4, and O1 antibodies was performed as described previously (Gallo and Armstrong, 1995; Gallo et al., 1996; Yuan et al., 1998; Belachew et al., 2003). Briefly, live cells were incubated at room temperature for 1 h with primary antibodies diluted 1:10 in DMEM, followed by fluorescein-conjugated goat anti-mouse IgM (1:200; Jackson ImmunoResearch, West Grove, PA) for 45 min. After three washes in PBS, cells were fixed in 4% paraformaldehyde (pH 7.3 in PBS) for 10 min at room temperature and washed. Cells were permeabilized with 0.1% Triton X-100 for 10 min at room temperature. For dual staining with cell surface antigens and Sox17, cells were blocked in 10% normal goat serum in DMEM for 30 min at room temperature before incubation with primary antibody to Sox17 (1:1000) for 2 h at room temperature. Anti-Sox17 primary antibody was diluted in 0.1 m phosphate buffer, pH 7.4, 0.1% gelatin, 1% BSA, 0.1% Triton X-100, 0.002% sodium azide, and 5% goat serum in PBS. This was followed by incubation with rhodamine-conjugated goat anti-rabbit IgG (1:200; Jackson ImmunoResearch) for 1 h at room temperature. The cells were then washed and mounted in Vectashield with 4′,6′-diamidino-2-phenylindole (DAPI). An Olympus Optical BX60 fluorescence inverted microscope was used to visualize immunofluorescence in cultured cells. Images were acquired using a 40× objective. For cell counting, 10 microscopic fields were counted for each coverslip, and two coverslips for each experiment were analyzed. At least three independent experiments (separate cell culture batches) were performed for each antibody staining. All data are presented as averages ± SEM. Luminosity measurements were performed with Adobe Photoshop 7.0. using LB1⁺, O4⁺, and O1⁺ cells as described above for tissue sections.

Real-time reverse-transcription PCR.

Total RNA (500 ng) was used to synthesize cDNA using oligo-dT primer (Invitrogen) in a 20 μl reaction. cDNA (1–2 μl) was then used for PCR in a 25 μl reaction. PCR was performed at 95°C for 15 s, 55°C for 30 s, and 72°C for 30 s for 40 cycles. For ex vivo experiments in Figure 3, multiplex real-time PCR was conducted in a 96-well spectrofluorometric thermal cycler (ABI PRISM 7700 Sequence detector system; Applied Biosystems, Foster City, CA). Fluorophore-labeled mouse Sox17 primer (forward, 5′ CGCAGCTACCAGGGACACGACTG[FAM]G 3′) and its unlabeled counterpart (reverse, 5′ GGACACCACGGAGGAAATGG 3′) were used for real-time quantitative PCR (qPCR). The primers were designed using the on-line LUX primer design site (Invitrogen). As a normalization control, β-actin primers (Invitrogen) were combined in each PCR mix. Fluorescence was monitored during every PCR cycle at the annealing or extension step and during the post-PCR temperature ramp.

For culture experiments in Figure 10, monoplex real-time PCR was conducted in a 96-well spectrofluorometric thermal cycler (ABI PRISM 7900 Sequence detector system; Applied Biosystems). Primers used for PCR are as follows: CNP I forward, 5′-CAAAGACAAGGGTTTGGTGCTT-3′; CNP I reverse, 5′-CGGGGAGGAGGAGATGTTAGCC[FAM]G-3′; myelin associated glycoprotein (MAG) forward, 5′-CGGAATGGGAGGCAAATACTATTTC[FAM]G-3′; MAG reverse, 5′- GCTCCGAGAAGGTGTACTGGTTG 3′; MBP forward, 5′-CGACCCTCCCCGTGTACCTTGGT[FAM]G- 3′; and MBP reverse, 5′-GGGTGAACTTGAAAGGGTGTCC-3′. β-Actin was also used as a normalization control. Fluorescence was monitored during every PCR cycle at the annealing or extension step and during the post-PCR temperature ramp. Fold changes were then measured according to instructions of the manufacturer (Invitrogen).

Western blot.

Cell lysates were prepared as described previously (Ghiani et al., 1999a,b). Sample separation on SDS polyacrylamide gels and Western blotting were performed as described previously (Ghiani et al., 1999a,b). Protein bands were detected using the Amersham Biosciences (Arlington Heights, IL) ECL kit with horseradish peroxidase-conjugated secondary antibodies (1:10,000; BD PharMingen, San Diego, CA).

Loss- and gain-of-function studies.

Double-stranded Sox17 small interfering RNA (siRNA) and siRNA-negative control (scrambled sequences; catalog #4615) were obtained from Ambion. A combination of three Sox17 siRNA sequences was used: ID101960 sense, GAGCUAAGCAAGAUGCUAGtt and antisense, CUAGCAUCUUGCUUAGCUCtg; ID 69339 sense, GGCUGUUCAAAAAUUUCGGtt and antisense, CCGAAAUUUUUGAACAGCCtC; ID 101869 sense, GAACCCAGAUCUGCACAACtt and antisense, GUUGUGCAGAUCUGGGUUCtg. Purified OPCs (cultured in 60 mm dishes; 6 × 10⁵ cells per dish) were treated with PDGF (10 ng/ml) for 36–40 h and then transfected in serum- and antibiotic-free DME–N1 medium with a 1:1:1 mixture of Sox17 siRNAs (final concentration, 70 nm) or the siRNA control (at equivalent concentration) using Lipofectamine 2000 (Invitrogen) for 6 h. The medium was then replaced with fresh DME–N1 supplemented with PDGF (10 ng/ml), and the cells were harvested at 24 and 48 h after transfection for Sox17 Western blot analysis. Under these conditions, transfection efficiency with a labeled siRNA control sequence (catalog #1027098; Qiagen) was estimated at 88 ± 2.4% after 12 h. For cell cycle exit and differentiation experiments, the siRNA-transfected cells (in chamber slides; Nunc, Naperville, IL) were treated with PDGF (10 ng/ml) in DME–N1 medium for 24 h, and the mitogen was then withdrawn. O4, O1, Ki67, BrdU, and caspase-3 immunostaining were performed at 2 and 4 d after mitogen withdrawal.

For gain-of-function experiments, 1–1.5 × 10⁵ cells were maintained in PDGF for 48 h in chamber slides before being transfected with 2.4 μl of Fugene and 0.8 μg of pCMV–Sox17–internal ribosome entry site 2 (IRES2)–Aequorea coerulescence (Ac)GFP or pCMV–IRES2–AcGFP for 8–12 h in DME–N1 plus PDGF. After transfection, medium was replaced and cells maintained in DME–N1 plus PDGF for an additional 24, 48, or 72 h before immunocytochemical or RNA analysis. We determined by qPCR analysis that this transfection procedure, which averaged an efficiency of 5%, resulted in 3.2-fold elevation of Sox17 RNA levels over vector-transfected controls (data not shown). For GFP and Ki67 double immunostaining, the transfected cells were fixed with 4% paraformaldehyde for 10 min and treated with 0.1% Triton X-100 for 8 min for permeabilization. GFP immunostaining was then performed, followed by Ki67 staining. For GFP and A2B5, O4, or O1 double immunostaining, the transfected cells were fixed and permeabilized as above and subjected to surface antigen staining. After the slides were blocked with 10% goat serum for 30 min, GFP staining was performed. For GFP and capase-3 double immunostaining, the transfected cells were fixed, permeabilized, and blocked as above. GFP staining was then performed, followed by caspase-3 staining.

Plasmid construction, transient transfection, and reporter assays.

SoxBSLuciferase reporter plasmid (Luc) contains four copies of the H4 site (two inverted AACAAT motifs separated by 4 bp) and was modified from pH4x4Luc reporter plasmid (Kanai et al., 1996) by cloning the sequence 5′ATCTAGACGAATTCAACAATCATCATTGTTGGGGCCAACAATCTACATTGTTCAGATCTAGAA 3′ into the BglII site of pGL3promoter vector (Promega, Madison, WI). This insert contains an additional EcoRI site close to the 5′ end. Simian virus 40–Luc (SV40Luc) refers to the parent pGL3 promoter vector that does not contain the insert (Promega). The 1.3 kb SphI–PstI fragment containing the mouse Sox17 cDNA from pGEMT-Easy Sox17 (Kanai et al., 1996) was subcloned into pLIT39 (New England Biolabs, Beverly, MA), and the SpeI fragment from this subclone was inserted into the NheI site of pCMV–IRES2–AcGFP (BD Biosciences Clontech, Palo Alto, CA). A separate Sox17 expression plasmid in pCMV–IRES2–discosoma red (dsRed) (BD Biosciences Clontech) was generated via inserting the SpeI fragment into pCMV–TnT (XbaI; Promega) and subsequently cloning into the SalI site of pCMV–IRES2–dsRed. Both AcGFP and dsRed plasmids produced similar results. The mouse MBP luciferase plasmid was generated by subcloning a 1.9 kb SacI–BamHI fragment from pMG2–MBP (Gow et al., 1992) between the SacI and BglII sites of pGL3Basic vector (Promega).

For transient transfection, COS7 cells were plated at 2.7 × 10⁵ per well in six-well tissue culture dishes 18 h before transfection. Transfection was performed with Lipofectamine 2000 (Invitrogen) according to the instructions of the manufacturer. Specifically, cells in 1.5 ml OptiMEM received 0.4 μg of luciferase reporter plasmid, 0.8 μg of expression vector or Sox17 plasmid, and 0.02 μg of SV40–β-galactosidase in 500 μl of OptiMEM with 2.4 μl of Lipofectamine 2000. After 5 h, the OptiMEM was replaced with DMEM plus 10% FBS, and cells were cultured in PDGF and harvested 48 h after the start of transfection. CG4 cells and purified OPCs were plated in N1 plus 30% B104-conditioned medium and in N1 plus PDGF, respectively, at 2.3 × 10⁵ cells per well in six-well poly-l-lysine-coated tissue culture dishes for 1–2 d before transfection. For reporter assays, transfected cells were collected in 200 μl of reporter lysis buffer (Promega) and 30 μl of lysate assayed for luciferase activity (Promega) on a Turner 20/20n Luminometer (Turner Biosystems, Sunnyvale, CA). Luciferase activity expressed as relative light units was normalized with total protein content by BCA assay (Pierce, Rockford, IL) and β-galactosidase activity, as described by Sambrook and Gething (1989).