Files in this Data Supplement:
Table S1. RT-PCR primer sequences
Table S2. Genes identified by microarray analysis for which mRNA was upregulated >1.5-fold in E12.5 NS/P that lack REST. Tissue expression profiles were determined using GNF Mouse Gene Atlas V3 data (Lattin et al., 2008) and were divided into transcripts showing neural expression (expression in at least one neural tissue, no restriction on expression in other tissues) or neural enrichment (expression in at least one neural tissue, expression in three or fewer other tissues). In some instances, information for a transcript was either unavailable or ambiguous; GNF Gene Expression Atlas 2 data was then used for these instances (Su et al., 2004). Presence of RE1 sequences within 50 kB of transcript boundaries was analyzed by SACO-based computational prediction (Otto et al., 2007), comparison with sequences reported by Johnson et al. (Johnson et al., 2007) or by visualizing mapped RE1 sites accessed via the UCSC genome browser (NCBI37/mm9 assembly) (Johnson et al., 2008; Kent et al., 2002; Waterston et al., 2002). Class was determined by literature searches for published functional characterization studies and or homology/family characteristics. Transcripts belonging to more than one class were designated to the class with their most predominant function.
Fig. S1. The lack of REST or CoREST in ES cells does not compromise ES cell pluripotency based on the presence of alkaline phosphatase (AP). (A) Western blot analysis of whole cell extract after transduction of ES cells with lenti virus carrying the indicated shRNAs shows an efficient knockdown of REST or CoREST. Sin3A served as a loading control. (B) Alkaline phosphotase (AP) assay showing similar staining in ES cells transduced with lenti virus carrying vector alone or scrambled shRNA as a control and ES cells transduced with lenti virus carrying the indicated shRNAs.
Fig. S2. The lack of REST or CoREST in ES cells does not compromise ES cell pluripotency based on the presence of the pluripotency markers SSEA1 and OCT4. Representative images of immunostaining of ES cells after transduction with lenti virus carrying Rest shRNA, CoRest shRNA or scrambled shRNA. SSEA1 (blue) and OCT4 (red) represent pluripotency markers. GF (green) served as a marker for shRNA expressing cells. Scale bars: 50 µm.
Fig. S3. ES cells that lack REST are able to generate secondary and tertiary neurospheres similar to wild-type ES cells although the primary neurospheres lack uniform neurosphere morphology. (A) Representative images of primary neurospheres derived from Rest+/+, Rest+/− and Rest−/− ES cells. Note the lack of uniformity in Rest−/− ES cell-derived primary neurospheres. (B) Bar graph showing the volume of the indicated secondary neurospheres (SNS) generated from the primary neurospheres with same genotypes. There is no significant difference between the volume of SNS generated from Rest−/− and Rest+/+ primary neurospheres. The Rest+/− secondary neurospheres were significantly larger compared with Rest+/+ or Rest+/− SNS (*P<0.05). (C,D) Bar graphs showing the number (C) and the volume (D) of the different tertiary neurospheres generated from the secondary neurospheres with same genotypes. There is no significant difference between the number and volume of tertiary neurospheres generated from Rest−/− and Rest+/+ secondary neurospheres. However, the number and volume of Rest+/− tertiary neurospheres were significantly larger compared with Rest+/+ or Rest+/− SNS (*P<0.05). All error bars are mean ± s.e.m. based on three independent experiments.
Fig. S4. ES cells that lack or have reduced amount of REST are able to generate mitotically active neurospheres. (A) Bar graph indicates that there is no significant difference in the number of Ki67-positive cells in the dissociated cells of secondary neurospheres generated from Rest+/+ (wild type), Rest+/− and Rest−/− ES cells. (B) Bar graph indicates that there is no significantly difference in the number of Ki67-positive cells in 10 µm sections of secondary neurospheres generated from Rest+/+ (wild type), Rest+/− and Rest−/− ES cells. (C) The mitotic index (fraction of Phospho H3+/GFP+ cells) of dissociated cells of secondary neurospheres generated from Rest−/− ES cells is significantly reduced compared with Rest+/− ES cells. (D) Bar graph indicates that there is a significant decrease in the number of phospho H3-positive cells in secondary neurospheres generated from Rest+/+ (wild type) and Rest−/− ES cells compared with those generated from Rest+/−ES cells. All error bars are mean ± s.e.m. based on two independent experiments and counting 500-600 cells for each condition (*P<0.05, ** P<0.05, +P<0.05).
Fig. S5. ES cells that lack or have reduced amount of REST are able to generate a uniform nestin-positive neurospheres. Representative images of 10-µm sections cut from neurospheres generated from Rest+/+ (wild type), Rest+/− and Rest−/− ES cells, showing that they consist largely of nestin-positive NS/P cells, which are uniformly distributed throughout the spheres. (A) Sections of ES cell-derived primary neurospheres. Note that although the morphology of sectioned Rest−/− primary neurospheres lack of uniformity the cells within the sectioned neurospheres are largely nestin positive. (B) Sections of ES cell-derived secondary neurospheres. Nestin (red) represents the NS/P marker; DAPI (blue) represents nuclear staining. Scale bars: 50 µm.
Fig. S6. REST and CoREST are efficiently knocked down in E12.5 NS/P cells transduced with lenti virus carrying the specific shRNAs. (A) Quantitative RT-PCR analysis shows a significant reduction in the levels of Rest or CoRest mRNA in NS/P cells expressing Rest shRNA#2 (left graph) or CoRest shRNA#1 (right graph), respectively, compared with NS/P cells expressing scrambled shRNA. Error bars are mean ± s.e.m. based on three independent experiments. (*P<0.05). (B) Western blot indicating a significant and specific reduction of REST or CoREST proteins in E12.5 NS/P cells expressing Rest shRNA or CoRest shRNA, respectively. Actin served as loading control.
Fig. S7. Lack of REST in E12.5 NS/P cells results in reduced mitotic activity and precocious neuronal differentiation. Analysis of E12.5 NS/P transduced with lenti virus, carrying either scrambled shRNA or a Rest shRNA, after the second passage. (A) Representative images of immunostaining showing fewer NS/P cells expressing the mitotic marker phosphohistone H3 (phospho H3) (red) when Rest shRNA vs scrambled shRNA is expressed (green). Arrows point to GFP-positive cells expressing phospho H3. GFP (green) represents a marker for cells expressing the indicated shRNAs. (B) The mitotic index (fraction of phospho H3/GFP+ cells) of NS/P cells expressing Rest shRNA is significantly reduced relative to NS/P cells expressing scrambled shRNA. (C) Representative images of immunostaining show some precocious neuronal differentiation, indicated by the presence of TUJ1 (red)+ cells, in NS/P cells expressing Rest shRNA. Arrows point to GFP-positive cells expressing the neuronal marker TUJ1. GFP (green) represents a marker for cells expressing the indicated shRNAs. (D) Bar graph indicating a low but significant number of Rest shRNA-expressing NS/P cells that are TUJ1 positive. (E) Quantitative RT-PCR of analysis of the indicated cell cycle genes shows a significant reduction in the levels of mRNA in NS/P cells expressing Rest shRNA compared with NS/P cells expressing scrambled shRNA. Scale bars: 50 µm. All error bars are mean ± s.e.m. based on three independent experiments (*P<0.05, ** P<0.05, +P<0.05).
Fig. S8. E12.5 and E17.5 NS/P cells differentiate into neurons and glia but with different proportion. E12.5 NS/P cells after 3 days of differentiation produced significantly more neurons than astrocytes. (A) Representative images of immunostaining of E12.5 and E17.5 NS/P cells. TUJ1 (red), GFAP (green), and DAPI (blue) represent neuronal, astrocytic and nuclear markers, respectively. Scale bar: 50 µm. (B) Bar graph showing that E12.5 NS/P produced significantly more neurons than astrocytes E12.5 NS/P cells after 3 days of differentiation (*P<0.05), whereas the number of neurons and astrocytes differentiated from E17.5 NS/P cells was not significantly different. The number of neurons generated from E12.5 NS/P cells was also higher than the number of neurons generated from E17.5 NS/P cells (+P<0.05). Error bars are mean ± s.e.m. based on three independent experiments.
