DEV010801 Supplementary Material

Files in this Data Supplement:

• Supplemental Table S1 - Adobe PDF

• Supplemental Figure S1 -

  Fig. S1. Evaluation of anti-Sox2 antibodies by immunocytochemistry, immunohistochemistry and western blot analysis of wild-type and Sox2-null neural cells, and of recombinant Sox proteins by western blot. We evaluated the Sox2 specificity of two commercial antibodies (R&D, mouse monoclonal; Chemicon, rabbit polyclonal). Sox2-null neural cells, obtained by in vivo nestin-driven Cre-mediated deletion (R.F. et al., unpublished), were compared with wild-type cells. Both antibodies gave clear nuclear staining in most of the wild-type cells, but failed to show any reactivity with nuclei of Sox2-null cells. (A) Dissociated neurospheres allowed to attach to a slide were probed with the indicated antibodies at the beginning (day 1) or at the end (day 9) of the differentiation protocol described in Fig. 1. With both antibodies, a clear nuclear signal is visible in wild-type, but not in Sox2-null cells. Expression decreases with differentiation, but is still clearly detected in day 9 differentiated cells. A slight cytoplasmic staining can be seen with the rabbit antibody (Chemicon) at both day 1 and day 9, in wild type and null cells, thus likely representing a nonspecific background. Secondary antibodies only (bottom panels) yield no signal. (B) In vivo, neither antibody stains nuclei in brain sections of mutant null newborn mice. Immunohistochemistry with both mouse (left panels) and rabbit (right panels) anti-Sox2 antibodies detects abundant nuclear Sox2 expression in wild-type (wt), but not in Sox2-deleted (null) ventricular zone at P0. Some background staining seen in the null mouse sections does not localize to nuclei. (C) Western blot studies with the R&D antibody, confirming that it does not crossreact with any proteins in undifferentiated neurosphere lysates of Sox2-null cells, even in the presence of a large excess of protein and with long exposures. Proteins from neurosphere cultures of wild-type (+/+), Sox2 heterozygous (+/-) and Sox2-deleted (-/-) mice were probed with anti-Sox2 antibody. Positions of Sox2 and CP2 (ubiquitous nuclear protein, as loading control) are indicated. Left panels: two different exposures of a filter probed with anti-Sox2 and anti-CP2 antibodies. Genotypes are indicated above the lanes. The longer (top) exposure shows failure of the antibody to detect any non-specific signal in the −/− sample; the lower (shorter) exposure allows better comparison of the CP2 signal, demonstrating that equal amounts of extracts were loaded in all lanes. Middle panel: the same filter probed with the Sox2 antibody, prior to re-probing with the CP2 antibody. No signal is seen in the Sox2-null (−/−) extract, even with this long (1 minute) exposure. Asterisks indicate the expected position of the Sox1 (*) and Sox3 (**) transcription factors, which are expressed in the same cells at normal levels (see D). Right panels: progressive dilutions (1/10, 1/20) of the amount of extract (1 corresponds to the amount loaded in the +/+ lane of the upper left and middle panels) still yield a clearly visible Sox2 signal, even when the same filters exposed for only 6 seconds (lower panel), instead of 1 minute (top panel). Thus, a 10-fold overexposure of an amount of extract 20-fold in excess to that required for Sox2 detection, still does not yield any non-specific signal. (D) RT-PCR analysis of expression of SoxB family members Sox1 and Sox3 (co-expressed with Sox2 in neural precursors), in wild-type and Sox2-null neurosphere cultures. Samples shown were taken from the PCR reactions at 25, 30, 35 and 40 cycles for both wild-type and null. Expression levels of Sox1 and Sox3 are similar between wild-type and Sox2-null cells. −, control reaction with reverse transcriptase-negative null control (40 cycles); M, marker. (E,F) Lack of cross-reaction of the anti-Sox2 antibodies with recombinant Sox1, Sox3 and Sox6. NIH3T3 (E) or HeLa (F) cells were transfected with CMV promoter-driven expression vectors (pCDNA3) for Sox2, or Sox1, Sox3 and Sox6. Cell extracts were probed with R&D anti-Sox2 antibody. The Sox1, Sox3 (E) and Sox6 (F) positions are indicated beside the panels. Although Sox2 was easily detected, no reactivity was obtained with extracts from cells transfected with the other Sox expression vectors. In conclusion, anti-Sox2 antibodies do not significantly crossreact with protein present in neural cells at various differentiation stages. The staining experiments reported in the paper were always performed with both antibodies (as indicated in figures), with essentially identical results. When quantitation of the staining was required, the R&D antibody was used.

• Supplemental Figure S2 -

  Fig. S2. Evaluation of Sox2 immunofluorescence at the single-cell level. To evaluate Sox2 immunofluorescence at the single-cell level, digital images of Sox2 immunofluorescence-labeled nuclei were acquired, and individual nuclei were delimited and evaluated (on the monochromatic image taken on the appropriate fluorescence channel) with the image-processing algorithm of the Region Of Interest (ROI) program provided with the Leica TCS2 Confocal Microscope (Leica Microsystems), or the ImageJ.exe processing and analysis program (http://rsb.info.nih.gov/ij/), and expressed in arbitrary units as the sum of the background-subtracted pixel values within each ROI (nucleus). Background levels were established measuring nuclei of Sox2-null cells (see Fig. S1) or of cells treated with secondary antibody only (B), giving comparable values. The ratios between positive signals and internal background (measured on five different positions within each field) were plotted and statistical significances were assessed by nonparametric tests (heteroskedastic ANOVA, T-test; *P<0,05). (A) Examples of Sox2 immunofluorescence of normal and mutant cells at day 1 (left) or day 9 (right) of in vitro differentiation. In day 1 cells, a Sox2-bright cell population is seen in the normal, which is very reduced in the mutant. At day 9, fluorescence levels are very similar between wild type and mutant. (B) Evaluation of Sox2 immunofluorescence (R&D antibody) at the single-cell level in wild type (WT) and mutant (MUT) cells, on the overall population at days 1, 5 and 9 of in vitro differentiation (as indicated). Each dot represents the Sox2 fluorescence level of a single cell nucleus; each vertical dot series represents the values within an individual microscope field evaluated (see Materials and methods below). ‘II Ab’ indicates nuclear fluorescence values obtained with the secondary antibody only; the ‘0’ level was set just above the highest values obtained with this negative control, as shown in B (the same applies to C and D). Red dots identify the β-tubulin-positive cells within the samples shown (see also C). At least 500 nuclei per differentiation day per genotype were quantitated, within at least six different fields. The asterisk indicates a significant difference at day 1, but not at days 5 and 9, between wild-type and mutant Sox2 fluorescence distributions (one-way ANOVA, P<0.03; two-tailed t-test, P<0.001). (C,D) Evaluation of Sox2 immunofluorescence within the β-tubulin-positive cell population at day 9 of in vitro differentiation (C) or in in vivo differentiated P0 cortical cells (D), in normal (WT) and mutant (MUT). Fluorescence levels are indicated as explained in B. Examples of Sox2/β-tubulin-double-positive cells in differentiation day 9 cells and P0 cortical neurons are shown in Fig. 2A, Fig. S5B, respectively. In the in vitro-differentiated β-tubulin positive cells (C), the Sox2 level was slightly, but significantly, decreased in mutants (two-tailed t-test, P<0.01). This is at variance with the analysis reported in Fig. S2B for the overall population, where most cells are glia. A comparison between normal and mutant MAP2-positive cells for Sox2 expression was not performed, owing to the rarity of MAP2-positive cells in the mutant (see text). In D, the data document a slight (statistically non-significant) difference between the wild- type and the mutant (two-tailed t-test, P<0.34). At least 200 nuclei from β-tubulin-positive cells were analyzed in C and D, for n=2 wild type and n=2 mutants.

• Supplemental Figure S3 -

  Fig. S3. Expression of astrocytic markers S-100 and connexin 43 (CX43) (Nagy and Rash, 2000) in GFAP-positive in vitro differentiated astrocytes (untransduced, or day 1 transduced with Sox2-expressing lentivirus). (A) Double immunofluorescence for GFAP and S-100 (top panels) or CX43 (bottom panels) in differentiation day 9 cells, untransduced (left) or transduced with Sox2-GFP-expressing lentivirus (right). Virtually all cells positive for GFAP co-express S-100 (top panels) or CX43 (bottom panels) in untransduced cells. In Sox2-transduced cells, numerous cells can be seen which have low or absent GFAP expression (see Fig. 9) and are positive for S-100 (top) or for CX43 (bottom), confirming their astroglial identity (arrows indicate examples). (B) Double immunofluorescence for GFP (marking cells transduced with the Sox2-GFP-expressing lentivirus) and for S-100 (top) or CX43 (bottom). The vast majority of Sox2-transduced cells (where downregulation of endogenous GFAP is observed, see Fig. 8) express S-100 (top panels) and CX43 (bottom panels), consistent with an astrocytic identity. S-100 may be somewhat reduced in occasional Sox2-transduced cells. No fluorescence signal is observed in Sox2-GFP virus-transduced cells prior to antibody staining (lower right image, indicating that GFP endogenous green fluorescence is not detected in cells after fixation), nor with secondary antibodies only (not shown). Images are by non-confocal microscopy; see also Fig. 8 for confocal images of GFAP/S-100 and GFAP/CX43 immunofluorescence.

• Supplemental Figure S4 -

  Fig. S4. The block in neuronal maturation in Sox2 mutant cultures is not associated with apoptosis, nor with persistence of undifferentiated cells characteristics (nestin positivity). (A) Apoptosis between initial β-tubulin expression and MAP2/NeuN activation can be ruled out. In fact, between day 5 and 9, ∼15% of the cells show TUNEL positivity (green), both in normal and mutant; however, >98% of β-tubulin-positive cells (red) do not show TUNEL positivity. Shown are differentiation day 7 mutant cells. Furthermore, the total number of cells in mutant cultures at day 9, and the number of β-tubulin-positive cells were comparable between normal and mutant cells (see Table S1 in the supplementary material; data not shown), indicating that the maturation block is not associated with, or dependent on, apoptotic cell death. Numbers of Ki67-positive (dividing) cells were also similar (not shown). (B) Time course of nestin expression. The kinetics of decrease of the number of cells positive to nestin (a marker of the undifferentiated state) is very similar between wild-type and mutant cultures. Note that β-tubulin appeared at day 5 in mutant, as in normal cells (see Fig. 3C). Thus, initial differentiation steps are not significantly delayed in mutant cells.

• Supplemental Figure S5 -

  Fig. S5. Sox2 expression in the lateral ventricle (A), and in regions of neuronal differentiation (within the neonatal cortex, B,C, and in adult olfactory bulb, D), in normal and mutant mice. (A) Left: Sox2 (red) (Chemicon) and RC2 (green, a radial glia marker) (Merkle et al., 2004) immunofluorescence on sections of P0 lateral ventricle (P0 LV) of normal (wt) and mutant (mut) mice (confocal microscopy). Arrowheads: examples of Sox2/RC2 double-positive cells. Right: Sox2 (green) (Chemicon) and GFAP (red) immunofluorescence in adult lateral ventricle (LV) of wild type (wt) and mutant (mut). (B,C) Immunofluorescence of isolated P0 cortical neurons from normal (wt) and mutant (mut) brains with Sox2 (R&D) and β-tubulin (B) or MAP2 (C) antibodies (confocal microscopy). A large proportion of β-tubulin or MAP2-stained neurons are clearly Sox2-positive.Within the MAP2-positive population, the intensity of Sox2 staining inversely correlates with that of differentiated marker, and the most strongly MAP2-labeled cells are completely devoid of Sox2. Arrowheads: examples of Sox2/β-tubulin or Sox2/MAP2 double-positive cells. Sox2/MAP2 double-positive cells are generally weakly positive for both markers. Arrows indicate strongly MAP2-positive cells (generally Sox2-negative). Asterisks indicate strongly Sox2-positive cells (generally MAP2-weakly positive or negative). (D) Immunofluorescence analysis of Sox2 expression in the olfactory bulb. Top: Low-magnification image of an olfactory bulb section (DAPI nuclear staining); white boxes highlight the regions of the rostral migratory stream (RMS) and, more externally, sections of the peripheral layers where terminal neuronal differentiation is completed: the granule layer (GL) and periglomerular layer (PGL). Lower panels show higher magnifications of these regions (as indicated) analyzed in wild-type (wt) and mutant (mut), with the indicated antibodies In the RMS, Sox2 is expressed in numerous cells, many of which are positive for PSA-NCAM (Ferri et al., 2004), a marker of transit-amplifying progenitors (Doetsch, 2003; Lledo et al., 2006). In the differentiated peripheral layers, some weakly Sox2-positive cells are still visible; they are rare in the GL, but more numerous in the PGL, where calretinin-positive neurons differentiate 14-20 days after their birth (Lledo et al., 2006). Here, however, few if any calretinin or NeuN-positive cells show Sox2. In the mutant, the number of Sox2-positive cells is diminished, as expected on the basis of the observations on the SVZ. Arrowheads in GL indicate Sox2-positive NeuN-negative cells. Arrowhead in PGL indicates cell appearing weakly positive for Sox2 and calretinin.