SEARCH

SEARCH BY CITATION

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
sc-12-0876_sm_SupplFigure1.tiff2702KFigure S1: A) GB1 and GB3 cells were seeded out in 96−]well plates in appropriate dilutions to obtain at least 100 wells with a single cell. After one month each well containing a sphere was scored as positive. The spheres were then harvested and dissociated in order to repeat this experiment several times (7 rounds). Clonal proliferation efficiency, expressed in %, represents the average values from 4 independent experiments. B) Similar culture conditions were used to assess increased expression of GFAP, and the downregulation of stemness markers CD133, SOX1, SHH and A2B5 upon serum mediated differentiation of GB1 and GB3 cells. C) Brains were dissected from newborn mice, embedded in 4% agar−]agar Artificial CerebroSpinal Fluid then slices were cut to a thickness of 400 μm using a vibratome. Slices were deposited on a Millicell−]CM (0.4 μm) culture plate insert and maintained in air−]liquid interface conditions. Red fluorescent GB1 or GB3 cells were then seeded on the top surface of the slice for at least three weeks. The samples were then embedded in paraffin and sliced for visualization of fluorescent cells. White arrows indicate the infiltrating cells within the tissue.
sc-12-0876_sm_SupplFigure2.tiff2702KFigure S2: A) The number of SHH and NANOG expressing cells (TG1, TG6, GB1, GB3 lines) was evaluated in the indicated culture conditions (DM: defined medium; serum: 0.5% serum; serum/DM: 3 days serum medium and then switch to defined medium) by counting three separate fields from three independent immunostainings. B) Phase−]contrast images display changes in TG6 morphology after 8h, 15h, 24h and 48h of the dedifferentiation process. Oct4+ cells in undifferentiated, differentiated (4 days in serum) and dedifferentiated (48h of switch to defined medium) conditions were counted in three separate fields of three separate immunostainings and are expressed as percent of the total cell population. C) Histogram shows HSA−]miR−]18a* expression in TG6 cells either cultured in defined medium (DM), in serum (serum) or cultured in serum (4 days) prior to defined medium from 3h to 48h. HSA−]miR−]18a* expression was determined by qPCR using a Taqman probe as described in the Methods section. D) and E) Real time RT−]PCR analyses of mature miR18a and miR18a* expression in TG1 or TG6 constitutively expressing the precursor form miR18a/a* (TG1 miR−] 18a* or TG6 miR−]18a*). The results were compared to control GiCs expressing a non−] relevant construct (TG1 CTL). F) TG6 cells (50 000 cells) stably expressing miR−]18a* (miR−] 18a*) or a non−]relevant construct (ctl), were grafted in the brain of NOD/SCID mice. Both cell lines stably expressed a luciferase reporter gene. Tumor growth was assessed in living animals by measuring luciferase activity. Results represent the average of five mice for each group. G) 500 TG6 cells stably expressing, or not miR−]18A*, have been xenografted orthotopically. NOD/SCID mice were scarified 6 months later. Dissected brains were embedded in paraffin, sliced and subjected to immunohistochemistry using an antibody directed against human vimentin in order to stain the human tumor cells. Left panels show an example of two different mice grafted with TG6 cells stably expressing miR−]18a*. The right panels show their control counterparts. The histogram displays the ratio of tumour size between miR−]18a* and Ctl cells from three separate mice (p value = 0.01). Tumor size was assessed by determining both the surface and intensity of signal (image J software).
sc-12-0876_sm_SupplFigure3.tiff2702KFigure S3: A) Phospho−]ERK immunostaining in TG6 either cultured in serum for 4 days (serum), or placed in serum for 4 days then in defined medium for 2 days (serum / DM) in presence or not of the MEK1 inhibitor U0126 or its inactive form, U0124. The lower panel shows Dapi staining. B) qPCR analysis of ERK1 and ERK2 mRNA in TG1 and TG6 cells stably expressing, or not, both sh−]ERK1 and sh−]ERK2. C) Western blot analysis using a phospho−]ERK antibody to assess the efficiency of sh−]ERK1/2 or Parke−]Davis. D) Evaluation of the number of SHH and NANOG expressing cells in TG1, TG6, GB1, GB3 lines cultured in serum or following the switch to defined medium in presence or absence of U0126 or Parke Davis (PD) by counting three separate fields of three separate immunostainings. E) qPCR analysis of DLL3, miR−]18a*, SHH, Gli−]1 mRNA expression in TG6 cells cultured either in defined medium, serum, or following the switch to defined medium in presence or absence of Parke Davis. F) and G) qPCR analysis of SHH, Gli−]1 and HES1 mRNA expression in TG6 cells cultured either in serum, or following the switch to defined medium in presence or absence of Parke Davis or stable and simultaneous expression of sh−]ERK1 and sh−]ERK2. H) immunofluorescence showing NANOG expression in TG1 and TG6 cells stably expressing either control shRNA (sh−] ctl) or sh−]ERK. I) and J) GB1 and GB3 cells were cultured as single cells in defined medium, in serum, or switched from serum to defined medium in presence or absence of Parke Davis and U0126 in 96−]well plates for one month. One hundred wells were observed and each well containing a sphere was scored as positive. The results are the mean of three separate experiments. (*** p value < 0,001 and ** p value < 0,01)
sc-12-0876_sm_SupplFigure4.tiff2702KFigure S4: A) and B) Relative expression of miR18a* in TG6 transfected with anti−]miR18a* or a scrambled sequence. TG1 and TG6 cells were incubated in serum for 4 days, tranfected with anti−]miR18a* or control oligonucleotide sequences (scramble, siLuc) and then cultured in defined medium. C) TG6−]ctl or TG6−]miR−]18a* cells were cultured 4 days in serum then placed in defined medium for 24h and 48h. DLL3 protein expression in each condition was assessed by western blotting. An antibody specific for actin was used as loading control. D) Quantification of NANOG + cells in TG1, TG6, GB1 and GB3 cells following the switch to defined medium in presence or absence of anti miR−]18a* or anti miR scramble (scrb). Three separate fields from three independent experiments were counted. NANOG expression was assessed by immunofluorescence as described in the Method section. E) Similar experiments were performed to assess NANOG expression in TG1 miR−]18a*, TG1 ctl, TG6 miR−]18a*, TG6 ctl at 24h and 48h following the switch to defined medium. F) Immunoblot analyses of phospho−]ERK levels in TG6−]ctl or TG6−]miR−]18a*cells cultured in serum medium for 4 days then incubated in defined medium for 24h and 48h. Lower panel displays total ERK1 expression. G) Immunoblot analyses show DLL3 protein expression and phospho−]ERK levels in TG6 cells following DLL3 silencing (siDLL3), or rescue of DLL3 expression (siDLL3 + pDLL3 mut). siLuc and pcDNA were used as controls. (*** p value < 0,001)
sc-12-0876_sm_SupplFigure5.tiff2702KFigure S5: A) Real time RT−]PCR analysis of mature miR18a* in TG1 cells stably expressing miR−]18a* (TG1 miR−]18a*) and control TG1 cells (CTL) cultured in defined medium (DM), in serum for 4 days (serum), or in serum for 4 days prior to be incubated in defined medium (serum / DM) for 48h in presence of MEK1 inhibitor U0126 or its inactive form U0124. Error bars reflect results obtained in three separate experiments. B) Immunostaining showing cleaved NOTCH1 expression in TG1 cells either cultured in defined medium (DM), serum (serum) or serum for 4 days prior to defined medium for 48h in absence or presence of DAPT. C) qPCR showing HES mRNA relative expression in TG1 cells cultured for 4 days in serum prior to defined medium for 24h in the presence or absence of DAPT. Results are the average of three independent experiments. (*** p value < 0,001 and ** p value < 0,01)
sc-12-0876_sm_SupplFigure6.tiff2702KFigure S6: A) Immunostaining showing the effects of DLL3 overexpression and ERK inhibition on cleaved NOTCH−]1 expression. As indicated, TG1 cells, cultured in defined medium, were transfected with either pCDNA3 as control or pDLL3−]CDS, or treated with U0126 to inhibit ERK activity. B) qPCR shows the level of HES mRNA in TG1 cells undergoing the switch to defined medium for 24h and transfected with either siLuc, siDLL3, pcDNA3, synthetic miR−] 18a*, anti miR−]18a* and synthetic miR−]18a*+ pDLL3−]CDS. C) and D) qPCR analysis of miR−] 18a* and DLL3 expression in TG1 and TG6 cells undergoing the switch to defined medium in presence or absence of a gamma secretase inhibitor (DAPT) in order to block NOTCH1 activation. E) Quantification of NANOG + and SHH + expression in TG1, TG6, GB1 and GB3 cells undergoing the switch to defined medium in presence or absence of DAPT. Three separate fields from three independent experiments were counted. NANOG and SHH expression was assessed by immunofluorescence as described in the Method section. F) TG1 cells were cultured as single cells in defined medium (DM) in presence or absence of DAPT, to assess their clonal proliferation efficiency. Hundred wells were counted for each condition and the results are the average of 4 independent experiments.
sc-12-0876_sm_SupplFigure7.tiff2702KFigure S7: qPCR shows relative HES1, Gli−]1, NOTCH−]1, SHH, and miR−]18a* expression in TG1 A) or TG6 B) transfected either with siCTL or si−]NOTCH1. Results are the average of three independent experiments. Immunostaining showing activated NOTCH1 and NANOG expression in TG1 C) or TG6 D) cells cultured in defined medium (DM) and transfected with siCTL or si−]NOTCH1. E) Immunoblot analyses show phospho−]ERK levels in TG1 and TG6 cells cultured in defined medium and transfected with siCTL or si−]NOTCH1. Total ERK2 antibody has been used as loading control.
sc-12-0876_sm_SupplFigure8.tiff2702KFigure S8: Immunofluorescence analysis of 5 human glioblastomas (paraffin−]embedded tissue sections) shows DLL3 (red) and SHH (green) co−]expression (A) or DLL3 and SOX1 co−] expression (B). Nuclei were stained in blue (Dapi) to indicate the presence of cells. (C) Quantification of the SOX1+ DLL3−] cells, SOX1+ DLL3+ cells, SOX1−] DLL3+ and SOX1−] DLL3−] cells in 10 different GBM cases. Percentage represents the partitioning of each cell population within the tumor. Five independent fields per case have been counted.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.