Color Key PCA. mir- 15a let-7c 106b let-7b let-7a 16 10b 99a 26a 20b 374b 19b 135b 125b a-5p 199b-5p 93 92b MES PN.

Transcription

1 Comp Comp2 a b Color Key MES PN PCA Comp mi- 15a let-7c 16b let-7b let-7a 16 1b 99a 26a 2b 374b 19b 135b 125b a-5p 199b-5p 93 92b MES PN c PN MES Supplementary igure 1 (elated to igure 1) (a) Principal component analysis (PCA) using Cluster 3. software shows PN mina expression profile clusters glioma PN spheres 84, 17, 2, and MES clusters 83, (b) The top 2 minas that are enriched in PN glioma spheres compared to MES glioma spheres. Heat maps generated using Java Tree view using Cluster 3. software as shown in igure 1A. (c) Analysis of mina expression in GBM of TCGA dataset. 2 minas with top-ratios between expression averages of PN and MES glioma spheres also cluster in PN and MES clinical GBM.

2 SOX2 MES1123 CD44 Percentage of positive cells SOX2 CD44 Percentage of positive cells elative expression level elative expression level elative expression level a b 8 2b 4 b b 125b c MES 2b 125b ALDH1A3 CD44 4 HMGA2 ALDH1A3 CD44 HMGA2 MES d SOX2 OLIG2 SOX2 OLIG2 MES b 2b AXIN2 c-myc AXIN2 c-myc MES b 2b 8 4 e 125b 2b 12 CD44 SOX2 125b 2b 8 4 CD44 SOX2

3 Supplementary igure 2 (elated to igure 2). Overexpression of mi-125b and mi-2b suppressed MES-associated genes expression and increased Wnt-related genes in glioma spheres. qt-pc analyses were performed to determine the expression levels of mi-125b and mi-2b (a), MES-associated genes ALDH1A3, CD44, and HMGA2 (b), and PN-associate genes SOX2 and OLIG2, as well as Wnt-related genes Axin2 and c-myc (c), in MES 83 and 1123 glioma spheres that separately overexpress mi-125b, mi-2b, or a control mina. (d) and (e), immunofluorescent IHC analyses of protein expression of CD44 and SOX2 in brain tumor xenografts established at 16 days post-implantation by MES 83 (d) and 1123 (e) spheres that separately overexpress mi-125b, mi-2b, or a control mina. Scale bar: 2 µm. Yellow arrows indicate positively stained cells. ight bar graphs, quantitation of positively stained cells. Data were collected from 3 separate tumors per group and 5 images per tumor. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.5, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

4 SOX2 PN157 CD44 Percentage of positive cells SOX2 PN84 CD44 Percentage of positive cells elative expression level elative expression level PN84 PN157 elative expression level a c 125b ZIP 2b ZIP b d b ZIP 125b ZIP ALDH1A3 CD44 PN157 ALDH1A3 CD44 PN b ZIP 125b ZIP SOX2 OLIG2 SOX2 OLIG2 PN157 PN84 2b ZIP 125b ZIP AXIN2 c-myc AXIN2 c-myc PN84 PN157 e 125b ZIP 2b ZIP b ZIP 2b ZIP 8 4 CD44 SOX2 f 125b ZIP 2b ZIP b ZIP 2b ZIP 8 4 CD44 SOX2

5 Supplementary igure 3 (elated to igure 2) Inhibition of mi-125b and mi-2b induced MES-associated gene expression, but suppressed Wnt-related genes in PN glioma spheres. (a) epresentative images of neurospheres of PN 157 and PN 84 that separately express mi-125 ZIP, mi-2b ZIP, or a control mizip. Scale bar: 5 µm. (b) MES-associated genes ALDH1A3 and CD44, (c) PN-associated genes SOX2 and OLIG2, and (d) Wnt-related genes AXIN2 and c-myc were analyzed using qt-pc analyses in PN 157 and 84 that express mi-2b ZIP, mi-125 ZIP or a control mina ZIP. was used as an internal control. (e), (f), immunofluorescent IHC analyses of protein expression of CD44 and SOX2 in brain tumor xenografts established by PN 84 (e) and 157 (f) spheres that separately overexpress mi-125 ZIP, mi-2b ZIP, or a control mizip. Both control PN 84 and 157 glioma spheres failed to form sizable brain tumor xenografs on 12 (PN 84) or 15 (PN 157) days post-implantation. Thus it was difficult to detect PN-associated Sox2 protein expression in these brain sections (e and f, lower panels and bar graphs). Scale bar: 2 µm. Yellow arrows indicate positively stained cells. ight bar graphs, quantitation of positively stained cells. Data were collected from 3 separate tumors per group and 5 images per tumor. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.5, p <.1, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

6 elative expression level a 8 125b-1 125b b shtc4 TC4 M r(k) -7-4 Supplementary igure 4 (elated to igure 3) (a) The pre-mi-125b-1 on chromosome hsa11 is a major pre-mi that contributes to mature mi-125b expression in glioma spheres. qt-pc analyses were performed to determine the expression levels of pre-mi-125b-1 and pre-mi-125b-2 in indicated glioma spheres. (b) IB analysis of TC4 protein on PN expressing a control shna and two TC4 specific shnas. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.1, paired two-way Student t-test. Data are representative from two independent experiments with similar results.

7

8 elative expression level Wnt1 Wnt2 Wnt2B Wnt3 Wnt3A Wnt4 Wnt5A Wnt5B Wnt6 Wnt7A Wnt7B Wnt8A Wnt8B Wnt9A Wnt9B Wnt1A Wnt1B Wnt11 Wnt16 Wnt1 Wnt2 Wnt2B Wnt3 Wnt3A Wnt4 Wnt5A Wnt5B Wnt6 Wnt7A Wnt7B Wnt8A Wnt8B Wnt9A Wnt9B Wnt1A Wnt1B Wnt11 Wnt16 elative expression level elative expression level elative expression level Wnt1 Wnt2 Wnt2B Wnt3 Wnt3A Wnt4 Wnt5A Wnt5B Wnt6 Wnt7A Wnt7B Wnt8A Wnt8B Wnt9A Wnt9B Wnt1A Wnt1B Wnt11 Wnt16 elative expression level Wnt1 Wnt2 Wnt2B Wnt3 Wnt3A Wnt4 Wnt5A Wnt5B Wnt6 Wnt7A Wnt7B Wnt8A Wnt8B Wnt9A Wnt9B Wnt1A Wnt1B Wnt11 Wnt16 elative expression level Wnt1 Wnt2 Wnt2B Wnt3 Wnt3A Wnt4 Wnt5A Wnt5B Wnt6 Wnt7A Wnt7B Wnt8A Wnt8B Wnt9A Wnt9B Wnt1A Wnt1B Wnt11 Wnt16 elative expression level Wnt1 Wnt2 Wnt2B Wnt3 Wnt3A Wnt4 Wnt5A Wnt5B Wnt6 Wnt7A Wnt7B Wnt8A Wnt8B Wnt9A Wnt9B Wnt1A Wnt1B Wnt11 Wnt16 f g PN MES WNT WNT WNT2B WNT WNT3A WNT WNT5A WNT5B WNT WNT7A WNT7B WNT8A WNT8B WNT9A WNT9B WNT1A WNT1B WNT WNT PN MES ZD ZD ZD ZD ZD ZD ZD ZD ZD ZD LP O YK MES PN157 3 PN PN23 1 PN h 6 MES1123 PN84 PN157 PN23 PN ZD1 ZD2 ZD3 ZD4 ZD5 ZD6 ZD7 ZD8 ZD9 ZD1 LP6 YK O2

9 Sphere # per well elative growth rate elative expression level i shwnt4 Wnt4 Wnt4 PN84 PN157 M r (K) PN84 PN157 PN84 PN157 shwnt3a shwnt5a j 8 4 PN157 shwnt3a shwnt4 shwnt5a C #1 #2 C #1 #2 C #1 # shwnt3a shwnt4 shwnt5a Supplementary igure 5 (elated to igure 4) Wnt signaling activity is active in PN but not MES glioma spheres. (a) Gene Set Enrichment Analysis (GSEA) of data sets of Nakano (GEO# GSE6789) 1 for enrichment of the canonical Wnt pathway between PN and MES GBM. (b) Analysis of TCGA dataset for enrichment of Wnt signaling pathway (KEGG) clustered group 1 (Grp1, most are MES) and group 2 (Grp2, most are PN) GBM tumors. (c) Kaplan-Meier survival

10 analysis of TCGA data (Wnt signaling pathway enrichment) of two patient subgroups (Grp1- MES vs Grp2-PN) obtained from (b). (d,e) qt-pc analyses were performed to validate expression of LE1, ZD6 and APC, and canonical Wnt signaling target genes SOX2, ID2 and JAG1 in indicated PN and MES glioma spheres. (f) cdna array data of Wnt ligands, ZDs, LP6, O2, and YK (GEO# GSE6789) 1. (g,h) qt-pc analyses were performed to validate the expression of 19 WNT ligands, 1 ZDs, LP6, O2, and YK in PN and MES spheres. (i) IB analysis of WNT4 protein on PN 84 and 157 expressing a control shna and WNT4 specific shnas. ight panels, due to incapability of detecting WNT3a and WNT5a proteins using IB with commercial available antibodies, qt-pc analyses were used to validate mna expression of WNT3a and WNT5a in PN 84 and 157 expressing a control shna vector and specific shnas targeting WNT3a or WNT5a, respectively. (j) Cell proliferation (upper panels) and neural sphere formation assays (lower panels) for PN157 expressing a control shna or specific shnas targeting WNT3a, WNT4 or WNT5a, respectively. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.5, p <.1, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

11 % of positive cells elative growth rate Sphere # per well ZD6 -V elative growth rate % of positive cells Sox2 Sox2 CD44 CD44 Sphere # per well 125b 125b ZIP 2b 2b ZIP a MES1123 PN157 β-catenin Hoechst Merge β-catenin Hoechst Merge MES1123-mi 125b 2b C N C N C N β-catenin GAPDH Histone H3 PN157-ZIP 125b 2b C N C N C N M r (K) b d sh-c shzd6#1 shzd6# shc + -V shc + ZD6 shzd6#3 + -V MES1123 shzd6#3 + -V shzd6#3 + ZD6 shc shzd6#1 shzd6#3 + ZD6 3 shzd6# MES1123 shc shzd6# V ZD ZD shc shzd6#3 M (K) r e Tumor size (mm 3 ) 5 12 shc + -V shc + ZD sh shzd Tumor size (mm 3 ) 2 1 c sh-c shzd6# V ZD sh shzd6 Tumor size (mm 3 ) 4 2 MES shc shzd6# V ZD MES1123 sh-c shzd6# V ZD f sh shzd6 MES1123 sh shzd6 g PN84 2 2b ZIP + shc 2b ZIP + shzd6#1 125b ZIP + shc 125b ZIP + shzd6# sh shzd6 4 CD44 Sox2 8 4 sh shzd6 CD44 Sox shc shzd6#1 shc shzd6#1 2b ZIP 125b ZIP

12 Supplementary igure 6 (elated to igure 5) mi-2b/mi-125b affected the cellular localization of β-catenin and inhibition of ZD6-impeded growth, self-renewal, and tumor growth of MES glioma spheres, (a) Left panels, IB analyses. Protein levels of β-catenin in cytoplasmic (C) and nucleus (N) were examined by IB using the indicated antibodies in mi-transduced MES 1123, or mizip-transduced PN 157. GAPDH and histone H3 were used as cytoplasmic and nuclear protein markers, respectively. ight panels, immunofluorescence staining of subcellular β-catenin localization in the indicated cells using indicated antibodies. Scale bar, 2 µm. (b) Cell proliferation of and 1123 expressing shzd6s or shc, with or without ZD6 rescued expression (#1/#2 targeting ZD6 O, while #3 targeting ZD6 3 UT). (c) Neurosphere formation of MES 1123 expressing shzd6s or shc, with or without ZD6 rescued expression. (d) Knockdown of ZD6 #3 inhibited tumor growth of MES 1123 GBM xenografts in the brain, whereas re-expression of ZD6, but not a vector control, rescues shzd6 #3-inhibited tumor growth (n = 5). (e) Tumor areas of indicated group in (d) and in ig. 5g were quantified. Scale bar: 1. mm. (f) Immunofluorescent IHC analyses of protein expression of CD44 and Sox2 in brain tumor xenografts established by MES 83 and 1123 glioma spheres that separately express shna for ZD6 (shzd6#1), or a control shna (sh). Scale bar: 2 µm. Yellow arrows indicate positively stained cells. Bar graphs below show quantitation of positively stained cells. Data were collected from three separate tumors per group and five images per tumor. (g) Cell proliferation (left) and neurosphere formation (right) of PN 84 expressing mi-2b ZIP, mi-125b ZIP, or a control mizip with or without ZD6 knockdown. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.1, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

13 elative growth rate sphere # per well sphere # per well elative growth rate a c APC mna mi-2b APC mna mi-125b '... AGUGAAGUAUAAUCAGCACUUUG...3' GAUGGACGUGAUACUCGUGAAAC 5'...CUGACUAAACUUACAUCAGGGAA...3' AGUGUUCAAUCCCAGAGUCCCU MES1123 sh sh shapc shapc b d shc shapc#1 ctrl ZIP 2b ZIP 125b ZIP APC TC4 ABC shapc ctrl #1#2 APC shc shapc#1 shapc#2 M r (K) PN MES1123 shapc PN APC shc shapc#1 shapc#2 ctrl #1 # M r (K) M r (K) e PN157 2b ZIP + shc 2b ZIP + shapc#1 125b ZIP + shc 125b ZIP + shapc# PN84 2b ZIP + shc 2b ZIP + shapc#1 125b ZIP + shc 125b ZIP + shapc# Supplementary igure 7 (elated to igure 5) (a) Targeting sites of mi-2b and mi-125b in 3 -UT downstream of the coding sequences of APC gene. Nucleotides complementary to the seed sequence of mi-2b and mi-125b are underlined. (b,c) Inhibition of APC exhibited negligible effects on cell growth and self-renewal of glioma spheres. Cell proliferation (b) and neurosphere formation assay (c) were performed in MES 83 and 1123 expressing a control shna or APC specific shnas. Inserts: IB of shna knockdown of APC in glioma spheres using the indicated antibodies. (d) IB analysis. Effects of shna knockdown of APC with or without expression of mi-2b ZIP, mi-125b-zip, or a control Zip on expression of TC4 and active -catenin (ABC) in PN 157 or 84. (e,f), Effects of inhibition of APC (shna knockdown) or mis (mizip inhibition) on cell proliferation (e) and neurosphere formation (f) of PN 157 and 84. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results. f shc shapc#1 2b ZIP PN157 shc shapc#1 125b ZIP PN84 shc shapc#1 shc shapc#1 2b ZIP 125b ZIP

14 elative luciferase activity elative expression level elative expression level Sox2 CD44 Percentage of positive cells elative expresion level elative growth rate Sphere # per well a PN84 shc shzd6# ZIP b ZIP b ZIP ZD6 p-p65 p65 p-stat3 Stat3 p-tak1 TAK1 NLK TC4 ABC PN M r (K) b Olig2 Sox2 PN ZD6 ALDH1A3 CD44 c ZD6 PN ZD6 d PN23 ZD6 12 ZD6 e f NLK ZD6 KN93 PN17 KN93+ZD6 shnlk PN23 M (K) r z-7-oxo+ZD6 shnlk+zd6 g shnlk C #1 C #1 NLK ZD6 TC4 PN17 5 2b 125b 2.5 Vec ZD6 PN23 M r (K) ZD6 KN93+ZD6 5z-7-oxo+ZD6 shnlk shnlk+zd6 Vec ZD6 C #1 C #1 NLK ZD6 TC4 h CD44 PN ZD6 SOX2 PN23 M r (K) ID2 KN93 KN93+ZD6 5z-7-oxo 5z-7-oxo+ZD6 sh-nlk sh-nlk+zd6

15 Supplementary igure 8, (elated to igure 6) ZD6 regulates properties of glioma spheres and mi-125b and mi-2b expression through the CaMKII-TAK1-NLK signaling. (a) IB analysis using indicated antibodies on PN 84 and 157 that express mi-2b ZIP or mi-125b ZIP with or without knockdown of ZD6. shc, a control scrambled shna. : control. (b) qt-pc analysis of PN-associated markers (Olig2 and Sox2) and MES-associated markers (ALDH1A3 and CD44). (c) Cell proliferation (left) and sphere formation assays (right) of PN 17 spheres with ZD6 overexpression. (d) Immunofluorescent IHC analyses of protein expression of CD44 and Sox2 in brain tumor xenografts established by PN 23 spheres that express ZD6, or a control vector (). Scale bar: 2 µm. Yellow arrows indicate positively stained cells. ight bar graph is the quantitation of positively stained cells. Data were collected from 3 separate tumors per group and 5 images per tumor. (e) IB analysis using indicated antibodies on PN 23 and 17 spheres with or without NLK knockdown and ZD6 overexpression. (f) Luciferase reporter assay using a lentivirus TOP lash/op lash to quantify relative Wnt signaling activity on PN 17 spheres that overexpress ZD6 or a control vector with or without treatments of CaMKII Inhibitor KN93 (5 μm), TAK1 inhibitor 5Z-7-oxo (5 μm), or shnlk lentivirus. (g,h) qt-pc analysis of the relative expression of mi-2b, mi-125b and a Wnt target gene ID2 in PN 17 (g) or 23 (h) spheres that overexpress ZD6 or a control vector, with or without various treatments as indicated. Error bars (s.d.) represent data of triplicate samples for each set of experiments.. p <.5, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

16 Sphere # per well elative growth rate a 83 Vec ZD b b ZD6 p-tak1 TAK1 p-p65 p65 p-stat3 Stat3 MES M r (K) b MES b + 125b + ZD6 2b + 2b + ZD6 c MES ZD6 ZD6 2b 125b ZD6 ZD6 2b 125b Supplementary igure 9 (elated to igure 6) mi-2b, mi-12b and ZD6 regulate downstream effectors of DZ6, TAK1, N- B and STAT3 and self-renewal, and cell growth of MES glioma spheres. (a) IB analysis using indicated antibodies. (b) Cell proliferation and (c) neurosphere formation assay of MES 83 and 1123 that expressed mi-2b, mi-125b or a control mina with or without ZD6 overexpression. Error bars (s.d.) represent data of triplicate samples for each set of experiments.. p <.5, p <.1, p <.1, paired twoway Student t-test. Data are representative from three independent experiments with similar results.

17 elative mna level of Wnt5a elative mna level of Wnt3a a PN23 PN17 PN23 PN17 PN23 PN ZD M ZD ZD M (K) r (K) M r r (K) ZD1 7 ZD3 7 ZD5-7 TC4 7 TC4 7 TC4 7 ABC - 1 ABC - 1 ABC c b PN23 PN17 PN23 PN17 PN23 PN17 ZD1 ZD3 ZD ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 shzd3 M r (K) shlp6 M r (K) sho2 M r (K) ZD3 LP O2-1 ZD6 7 ZD6-7 ZD6-7 TC4-7 TC4 7 TC4 7 ABC ABC ABC p-p65-7 p-p65-7 p-p65-7 p p65 p65-7 p-stat3-7 p-stat3-7 p-stat3-7 Stat3 7 Stat3-7 Stat3-7 p-tak1-7 p-tak1-7 p-tak1-7 TAK1-7 TAK1-7 TAK1 7 NLK - 5 NLK - 5 NLK PN23 PN17 PN23 PN17 PN23 PN shwnt3a ZD6 TC4 ABC p-p65 p65 p-stat3 Stat3 p-tak1 TAK1 NLK ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 shwnt4 M r (K) shwnt5a M M (K) r (K) Wnt4-4 r ZD6 ZD TC4 TC ABC ABC p-p65 p-p p65 p p-stat3 - p-stat Stat3 Stat p-tak1 - p-tak TAK1 TAK NLK 5 NLK PN23 PN17 PN23 PN17 PN23 PN17-1 shyk YK ZD6 TC4 ABC p-p65 p65 p-stat3 Stat3 p-tak1 TAK1 NLK ZD6 ZD6 M r (K) PN23 PN17 d ZD ZD #1 #2 #1 #2 #1 #2 #1 #2 shwnt3a ZD ZD #1 #2 #1 #2 #1 #2 #1 #2 shwnt5a

18 elative expression level old change old change Sphere # per well Sphere # per well e PN23 shzd3 shlp6 sho2 shyk shwnt3a shwnt4 shwnt5a ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 PN17 shzd3 shlp6 sho2 shyk shwnt3a shwnt4 shwnt5a f ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 shzd3 shlp6 sho2 shyk shwnt3a shwnt4 shwnt5a PN g ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 shzd3 shlp6 sho2 shyk shwnt3a shwnt4 shwnt5a PN PN23 PN17 PN23 PN17 PN23 PN17 PN23 PN17 2b 125b ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 ZD6 shlp6 shlp6 shzd3 shzd3 sho2 sho2 shwnt5a shwnt5a Supplementary igure 1. (elated to igure 6) ZD3, O2 and Wnt5a are involved in ZD6 overexpression-induced phenotypes in glioma PN spheres. (a) IB analysis on PN 23 and 17 spheres with or without ZD1, ZD3 or ZD5 overexpression using indicated antibodies. (b) Cell proliferation assays on PN 23 and 17 spheres with or without ZD1, ZD3 or ZD5 overexpression. (c) IB analysis using indicated antibodies on PN 23 and 17 expressing ZD6

19 or a control vector, with or without knockdown of ZD3, LP6, O2, YK, WNT3a, WNT4 or WNT5a. C: a control shna, shna #1/#2 are the specific shnas of indicated gene. (d) Due to incapability of detecting WNT3a and WNT5a using IB with commercial available antibodies, qt-pc analyses were used to validate mna expression of WNT3a and WNT5a in PN 23 and 17 expressing ZD6 or a control vector with or without knocking down of WNT3a or WNT5a. (e) Neurosphere formation assays, and (f) cell proliferation were analyzed in indicated glioma spheres. (g) qt-pc analysis of the relative expression of mi-2b and mi-125b in indicated PN spheres. Error bars (s.d.) represent data of triplicate samples for each set of experiments.. p <.5, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

20 PN17 PN23 MES2 Sphere # per well elative growth rate Sphere # per well elative growth rate a b KN93 5z-7-oxo BAY JSI-124 KN93 MES2 PN17 LGK974 ICG-1 5z-7-oxo BAY JSI-124 LGK974 ICG-1 Indo LGK974 +Wnt3a KN93 5Z-7-oxo Bay JSI-124 LGK974 ICG-1 Indo LGK974+Wnt3a KN93 5Z-7-oxo Bay JSI-124 PN17 MES2 LGK974 ICG shtak1 shnκb shstat3 shcamk II M r (K) M r (K) M r (K) shctnnb1 shnlk p-tak1-7 p-p65-7 p-stat3 M r (K) - 7 M r (K) M r (K) CaMK II - 5 TAK1-7 p65 Stat3 ABC NLK CaMK II - 5 p-tak1 p-p65-7 p-stat3-7 ABC NLK TAK1 p65-7 Stat shtak1 shstat3 shnκb shctnnb1 shtc4 shcamk II M r (K) shnlk M r (K) M r (K) M #3 M r (K) #3 p-tak1 p-stat3 p-p65-7 r (K) M r (K) ABC - 1 TC4-7 CaMK II TAK1 Stat3 p65 NLK C #2 #3 ABC - 1 TC4-7 CaMK II - 5 p-tak1-7 p-stat3-7 p-p65-7 NLK TAK1-7 Stat3-7 p

21 Sphere # per well Sphere # per well Sphere # per well Sphere # per well old change old change c MES2 d shcamk II 1 shctnnb1 4 shtak1 shn-kb shstat3 shctnnb1 shnlk PN23 #3 #3 shtc4 shcamk II shtak1 shstat3 shn-kb shcamk II shnlk 5 shctnnb1 shtak1 shn-kb shstat3 shctnnb1 shnlk PN17 C #2 #3 shtc4 shcamk II shtak1 shstat3 shn-kb shnlk 2 shcamk II shtak1 shn-kb shstat3 shctnnb1 shnlk 4 MES shcamk II shctnnb1 shtak1 shn-kb shstat3 shctnnb1 shnlk #3 #3 shtc4 shcamk II PN23 stak1 PN17 shstat3 shn-kb shnlk 4 2 shctnnb1 C #2 #3 shtc4 scamk II shtak1 shstat3 shn-kb shnlk

22 Supplementary igure 11 (elated to igure 7) Targeting key components of differentially activated signaling pathways effectively inhibited distinct GBM subtypes. (a) Neurosphere formation (left) and cell proliferation (right) of indicated glioma spheres treated with DMSO (), KN93 (5 µm), 5Z-7-oxo (5 µm), BAY (5 µm), JSI-124 (.1 µm), LGK974 (5 µm), ICG-1 (5 µm) or LGK974 (5 µm) + WNT3a (2 ng/ml), respectively. (b) IB analysis using indicated antibodies on MES 83 and 2, PN 23 and 17 glioma spheres expressing control or specific shnas targeting CaMKII, -catenin (CTNNB1), TC4, TAK1, N- B, STAT3, and NLK, respectively. (c) Cell proliferation and (d) neurosphere formation assays were analyzed on MES 83 and 2, PN 23 and 17 glioma spheres expressing control or specific shnas targeting CaMKII, CTNNB1, TC4, TAK1, N- B, STAT3 and NLK, respectively. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.5, p <.1, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results.

23 LGK974 BAY elative expression level Ki-67-positive cell number (% of ctrl) BAY LGK974 elative expression level Ki-67-positive cell number (% of ctrl) a PN23 #1 #2 #3 b LGK974 BAY PN23 p-p65 p65 ABC BAY- LGK & M r (K) LE1 ID2 JAG1 4 CD44 IL6 IL & LGK974 BAY c #1 #2 #3 d BAY LGK974 Supplementary igure 12 (elated to igure 7) Pathway-specific inhibitors selectively inhibited growth of PN and MES glioma sphere tumor xenografts established by PN 23 (a,b) and MES 83 (c,d) glioma spheres and Wnt- and N- B signaling in vivo. (a,c), IHC images of Ki-67 stained individual PN 23 or MES 83 glioma sphere tumor xenografts. Scale bar, 5 µm. (b,d), Upper left, quantification of Ki-67-positive cells up right, IB of p-p65, p65 and activated β- catenin (ABC) in tumor lysates, lower graphs, qt-pc analysis of Wnt pathway target genes, LE1, ID2 and JAG1 and N- B target genes CD44, IL6 and IL8, respectively. In PN GBM tumors, Wnt inhibitor LGK974, but not N- B inhibitor BAY-11782, suppressed tumor growth and Wnt signaling. Conversely, in MES GBM tumors, N- B inhibitor BAY-11782, but not Wnt inhibitor LGK974, suppressed tumor growth and N- B signaling. Data in bar graphs in b and d were collected from three separate tumors per group and five images per tumor. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p <.5, p <.1, p <.1, paired two-way Student t-test. Data are representative from three independent experiments with similar results BAY LGK p-p65 p65 ABC LE1 ID2 JAG1 4 CD44 IL6 IL8 3 2 M r (K) LGK974 BAY-11782

24 Percent survival Percent survival Percent survival Percent survival elative expression level T4 T41 T42 T43 T44 T45 T46 T47 T48 T49 T5 T51 T52 T53 T54 T55 T56 T57 T58 T59 T6 T61 Percent survival Percent survival elative expression level T1 T2 T3 T4 T5 T6 T7 T8 T9 T1 T11 T12 T13 T14 T15 T16 T17 T18 T19 T2 T21 T22 T23 T24 T25 T26 T27 T28 T29 T3 T31 T32 T33 T34 T35 T36 T37 T38 T39 a b 125b c TC4 ZD TC4 ZD b high- TC4 high (n=2) 2b low- TC4 low (n=16) p < b high- ZD6 low (n=15) 2b low- ZD6 high (n=24) p < b 125b 125b high- TC4 high (n=27) 125b low- TC4 low (n=16) p < b high- ZD6 low (n=15) 125b low- ZD6 high (n=18) p <.5 Supplementary igure 13 (elated to igure 8) Clinical correlations of expression levels of mi-2b, mi-125b, and TC4 or ZD6 in clinical GBM tumors. (a) qpc (upper bar graphs) for mi-2b and mi-125b, and IB analysis of 61 snap-frozen clinical GBM tumor specimens using indicated antibodies for expression of TC4 and ZD6 (lower panels). (b, c,) Kaplan- Meier analyses of 61 snap-frozen GBM samples for expression of mi-2b or mi-125b (b), in combination of mi-2b or mi-125b with TC4 or ZD6 (c). (d) Kaplan-Meier analyses of 76 paraffin-embedded GBM samples analyzed by IHC (see igure 8c,d) for expression of TC4 or ZD6. Error bars (s.d.) represent data of triplicate samples for each set of experiments. p values were calculated by using log-rank and Gehan-Breslow-Wilcoxon tests. b M r (K) d 1 5 2b high (n=3) 2b low (n=31) p < b high (n=36) 125b low (n=25) 5 p < TC4 high (n=25) TC4 low (n=51) p < ZD6 high (n=36) ZD6 low (n=4) 5 p < M r (K)

25 Percent Survival Percent Survival Percent Survival Percent Survival Percent Survival a 1 5 m i 2 b high ( = 5 8 ) m i 2 b ow (n=51) p =.146 b 1 m i 12 5 b high ( = 4 3 ) m i 12 5 b ow (n=37) p = c Z D 6 high ( = 6 2 ) ZD6 low (n=7) p =.137 d m i 2 b high ZD6 low (n=19) m i 2 b o Z D 6 high ( = 16 ) p =.353 e m i 12 5 b high ZD6 low (n=13) m i 12 5 b o Z D 6 high ( = 15 ) p = Supplementary igure 14 (elated to igure 8e, 8f) Clinical correlations of expression levels of mi-2b, mi-125b and ZD6 in PN and MES GBM tumors available in TCGA data sets. (a to e) Kaplan-Meier curves of PN and MES GBM tumors available in TCGA data sets were analyzed individually with mi-2b, mi-125b, ZD6, or in combination of mi-2b or mi- 125b with ZD6. p values were calculated by using log-rank and Gehan-Breslow-Wilcoxon tests.

26 TC4 TC4 Percent of specimens Percent Survival Percent Survival Percent Survival Percent Survival a 1 5 TC4 high (n=64) TC4 low (n=57) p <.5 b 1 5 TC4high ZD6 low (n=26) TC4 lo ZD6 high (n=14) p =.228 c mi 2bhigh TC4 high (n=2) mi 2b o TC4 low (n=24) p <.1 d mi 125bhigh TC4 high (n=12) mi 125b o TC4 low (n=14) p = e. TC4 high TC4 low p <.5 p < High Low 2 b High Low 12 5 b 2 b high low low 6 24 high b high low low 3 14 high Supplementary igure 15 (elated to igure 8e, 8f) Clinical correlations of expression levels of mi-2b, mi-125b and TC4 in in PN and MES GBM tumors available in TCGA data sets. (a to d) Kaplan-Meier curves of PN and MES GBM tumors available in TCGA data sets were analyzed individually with mi-2b, mi-125b, TC4, or in combination of mi-2b or mi- 125b with TC4. (e) Percentages of specimens showing high or low mi-2b or mi-125b expression relative to level of TC4 in TCGA data sets. p values were calculated by using logrank and Gehan-Breslow-Wilcoxon tests.

27 igure 3d igure 3g TC4 7 5 C-MYC NC NC c-myc 5 4 NC 125b-TBS b-TBS4-5 c-myc 125b-TBS 4-5 2b 2b 125b-TBS b-TBS b-TBS 4-5 2b

28

29 Supplementary igure 16: Uncropped scans of the IB data presented in the main text (ig. 3 to 6).

30 Supplementary Table 1 The list of primers used in this study S.No. Gene Primer Primer sequence Quantitative everse Transcription (qt) primers 1 ALDH1A3 2 APC 3 APC (3 UT) 4 ZD6 5 ZD6 (UT) 6 HMGA2 7 ID2 8 JAG1 9 LE1 1 Olig2 11 pri-mi-125b 12 pri-mi-125b 13 SOX2 14 TC4 15 AXIN2 GCATGAGCCCATTGGTGTCT CGCAGGCTTCAGGACCAT AAGCATGAAACCGGCTCACAT CATTCGTGTAGTTGAACCCTGA AAGAGAGGAAGAATGAAACTAAGAAAATTC GCAAAGTGCTGATTATACTTCACTGCTG AGAGGTGAAAGCGGACGGA AGAGAGTCTGGAGATGGATGCT AAACTCGAGTGGGAGGACAGAGTTAGAGGAA AAAGCGGCCGCTCAGTGTAAAAGCTACCAAAGTGC TCCCTCTAAAGCAGCTCAAAA ACTTGTTGTGGCCATTTCCT GCTATACAACATGAACGACTGCT AATAGTGGGATGCGAGTCCAG GGGGCAACACCTTCAACCTC CCAGGCGAAACTGAAAGGC ATGTCAACTCCAAACAAGGCA CCCGGAGACAAGGGATAAAAAGT CTCCTCAAATCGCATCCAGA AGAAAAAGGTCATCGGGCTC AGAAAAAGGTCATCGGGCTC 1 CCATACCACCTGTTTGTTGCATCT 1 CTGAGAGGAGCGCAACAATGT 2 GAAGAATTCTACCGCATCAAACCA 2 CTGCAGACAATCAATAAGGTCCAA CCCTGCTGAGAATAGGACAT CCCTGCAGTACAACTCTATG TCCCACCACATCATACGCTACAC TCGCTTGCTCTTCTCTGGACAG TACACTCCTTATTGGGCGATCA TTGGCTACTCGTAAAGTTTTGGT 16 WNT1 CGGCGTTTATCTTCGCTATCA

31 17 WNT1A 18 WNT1B 19 WNT11 2 WNT16 21 WNT2 22 WNT2B 23 WNT3 24 WNT3A 25 WNT4 26 WNT5A 27 WNT5B 28 WNT6 29 WNT7A 3 WNT7B 31 WNT8A 32 WNT8B GCAGGATTCGATGGAACCTTCT GGAGACTCGCAACAAGATCCC CGATGGCGTAGGCAAAAGC GTGAGCGAGACCCCACTATG CACTCTGTAACCTTGCACTCATC GACCTCAAGACCCGATACCTG TAGACGAGTTCCGAGTCCTTC GCAGAGAATGCAACCGTACAT CACATGGGTGTTGTAACCTCG GCCTTTGTTTATGCCATCTCCT CTTGGCGCTTCCCATCTTCTT CGGGACCACACCGTCTTTG GCGAGTAATAGCGTGGACTAC GGAGAGGGACCTGGTCTACTA CTTGTGCCAAAGGAACCCGT AGCTACCCGATCTGGTGGTC CAAACTCGATGTCCTCGCTAC GTACGCCATCTCTTCGGCAG GCGATGTTGTCAGAGCATCCT TCGACTATGGCTACCGCTTTG CACTCTCGTAGGAGCCCTTG CGCTTCGCCAAGGAGTTTG TGCCATCTTATACACAGCCCT GGTGCGAGAGTGCCAGTTC CGTCTCCCGAATGTCCTGTT CTGTGGCTGCGACAAAGAGAA GCCGTGGCACTTACATTCC CGCAGCTATCAGAAGCCCAT CAGGTGTTGCACTTGACGA GAACTGCCCTGAAAATGCTCT TCGAAGTCACCCATGCTACAG AAGGCCGAGAGTGCCTAAG CTGCGCGGCTACAGAAGTA

32 33 WNT9A 34 WNT9B 35 ZD1 36 ZD2 37 ZD3 38 ZD4 39 ZD5 4 ZD7 41 ZD8 42 ZD9 43 ZD1 44 LP6 45 O2 46 YK 47 pri-mi-125b 48 SOX2 49 TC CCACCGTGAGAAGAACTGC GCCTGCACTCCACATAGCA TGTGCGGTGACAACCTCAAG ACAGGAGCCTGATACGCCAT GGGGCTTAACAACGTGGAC CAGAAAGGACGTGCCGATAAA GTGCCATCCTATCTCAGCTACA CTGCATGTCTACCAAGTACGTG AATATGGACGTGTCACACTTCC GGATATGGCTCATCACAATCTGG GTGTCACTCTGTGGGAACCAA GGCTGTATAAGCCAGCATCAT CCGTTCGTGTGCAAGTGTC GAAGCGTTCCATGTCGATGAG CAGACGTGCAAGAGCTATGC ACGATCATGGTCATCAGGTACT ATCTTGTCGCTCACATGGTTC CATGGTGCCGATGAAGAGGTA TGCGAGAACCCCGAGAAGT GGGACCAGAACACCTCGAC GCTCATGGTGCGTATCGGG GAGGCGTTCGTAAAAGTAGCA ACGATTGTAGTTGGAGGCTTG ATGGCTTCTTCGCTGACATCA ATGGTTCACGACTGCGAATCC AATGGTCTTCATCCCGTTGGT CCCAGGTCAACATTTCTGTTCA TGCCAGTACAGGAAAGCTCTAC CTGAGAGGAGCGCAACAATGT GAAGAATTCTACCGCATCAAACCA CTGCAGACAATCAATAAGGTCCAA CCCTGCTGAGAATAGGACAT CCCTGCAGTACAACTCTATG TCCCACCACATCATACGCTACAC TCGCTTGCTCTTCTCTGGACAG

33 1 ZD6 (O) MI-2b 2 promoter 3 NLK 4 TAK1 5 ZD1 (O) 6 ZD3 (O) 7 ZD5 (O) mi-2b 1 promoter mi-125b 2 promoter,1-3 mi-125b 3 promoter, 4-5 mi-myc promoter EcoI, NotI, Eco1 BamH1 Eco1 BamH1 Xho I, BamH I, Eco1, Eco1, Eco1, BamH I, Cloning primers GTTGTTGAATTCGCCACCATGGAGATGTTTACATTTTTGTTGAC GTGTA GTTGTTGCGGCCGCTCAAGTATCTGAATGACAACCACCTCCC GTTTTCGCTTTGGCGGGGTGGG GCCCCAACGAAGGGCTCCCTTCTTGCC GTTGTTGAATTCGCCACCATGGCGGCTTACAATGGCGGTACAT C GTTGTTGGATCCTCACTCCCACACCAGAGGAGATGGG GTTGTTGAATTCGCCACCATGTCTACAGCCTCTGCCGCC GTTGTTGGATCCTCATGAAGTGCCTTGTCGTTTCTGC GATCTCGAGATGGCTGAGGAGGAGGCGCCTAAGAAGTC ACGGGATCCTCAGACTGTAGTCTCCCCTTGTTTG GTTGTTGAATTCGCCACCATGGCTATGACTTGGATTGTCTTCTC TC GTTGTTGAATTCGCCACCTTAAGCACTGGTTCCATCTTCTTC AAAAGAATTCATGGCTCGGCCTGACCCATCCGCG ACGGGATCCCTACACGTGCGACAGGGACACCTGCT CHIP-sequencing primers GACAGCGCTCTGTAGAATAAAATG GTATATAGCCCGGTTGTCTCTCAT AGAAGAACAAGAAGAAGAAAG TCTCGAGACTGTAACTCTGTAGCTT TCGAATGGGTGAGTTCAGAACGC CGCATATACAATCACGCACATACAC GCACGGAAGTAATACTCCTCTCCTC CAGAAGAGACAAATCCCCTTTGCGC Supplementary eference 1. Mao P, et al. Mesenchymal glioma stem cells are maintained by activated glycolytic metabolism involving aldehyde dehydrogenase 1A3. Proc Natl Acad Sci U S A 11, (213).