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

sc-12-0733_sm_SupplFigure1.tif716KFigure S1 (related to Figure 1). MELK protein is highly expressed in high-grade glioma Tissue microarray slides, containing 91 samples (12 adjacent normal brain, 13 Grade II glioma, 34 Grade III glioma, 32 GBM samples), were evaluated by MELK immunohistochemistry. Each dot represents the average of the percentage of MELK (+) cells in each sample. LGG: Low-grade glioma; HGG: High grade glioma.
sc-12-0733_sm_SupplFigure2.tif7556KFigure S2 (related to Figure 1). MELK is expressed in patient-derived GSCs. A) GBM146 and GBM528 spheres were subjected to immunocytochemistry by incubating their dissociated cells with anti-MELK, anti-Nestin, anti-SOX2, anti-NeuN, and anit-GFAP antibodies. Hoechst dye (blue) is used for nuclear staining. Representative images display the co-localization of MELK with Nestin and SOX2 but not with NeuN or GFAP. Bar: 5 μm. B) Immunocytochemistry with cryosectioned spheres derived from GBM patient samples (GBM157) demonstrates that MELK (+) cells are co-labeled with the immature neural cell markers Nestin, SOX2 and CD133 but not with the differentiation marker, GFAP. Bar: 50 μm.
sc-12-0733_sm_SupplFigure3.tif2431KFigure S3 (related to Figure 2): Knockdown of MELK decreases the neurosphere formation of GBM cells. A) Western blot of GBM1600 cells (cultured in serum-free media and serumcontaining media) and 528 spheres (cultured in serum-free media) displaying 3 different MELK shRNA constructs diminishes varying levels of decreased MELK expression compared to the control (non-targeting shRNA). B) Representative images of the effects of 2 shRNA constructs for MELK (shMELK #43 and #45) compared with non-targeting shRNA (control) on neurosphere formation with GBM 528 cells.
sc-12-0733_sm_SupplFigure4.tif2872KFigure S4 (related to Figure 2): MELK knockdown represses growth of GSCderived mouse intracranial tumors in vivo. A) Table displaying the difference in frequency of tumor formation between no target shRNA- or shMELK-infected GBM528 spheres in mouse brains. B) At 5 months after xenografting, Hematoxylin & Eosin (H&E) staining was used to determine the tumor volume in each mouse. The tumor volume analysis exhibits the statistically-significant difference between no target shRNA (NT) and shMELK (M) in 10,000 cell group where p value is ≤0.001. C) H&E staining of serial mouse brain slices demonstrates the differences of tumor size between shMELK-expressing tumors and the control. Control tumor (left panel) exhibits large necrotic area, whereas MELK shRNA-infected GBM528-derived tumors do not exhibit noticeable tumor necrosis or hemorrhage (right panel). Bar: 1mm.
sc-12-0733_sm_SupplFigure5.tif277KFigure S5 (related to Figure 2): Abolished effect of MELK silencing in tumors. Western blot demonstrates that there is no noticeable difference in MELK expression between no-target shRNA and MELK shRNA-infected mouse tumors at the time of death (at 4 months in the control mice and at 5.5 months in the shMELK expressing mice). GAPDH was used as a control.
sc-12-0733_sm_SupplFigure6.pdf51KFigure S6 (related to Figure 2). MELK silencing increases the survival periods of mice xenografted with U87MG cells intracranially. MELK-depleted U87MG cells are xenografted into immunocompromised mouse brains and the survival periods of tumor-burden mice are evaluated by Kaplan-Meier plots. Mice harboring MELK-silenced U87MG cells survived significantly longer than the control mice xenografted with U87MG cells harboring no-target shRNA.
sc-12-0733_sm_SupplFigure7.pdf126KFigure S7 (related to Figure 3). JNK inhibition reduces MELK expression in GBM spheres but not in normal neural progenitors. GBM146 (A), GBM157 (B), nomal neurospheres 1105 (C) and 1106 (D) were treated with JNK inhibitor II (Calbiochem) at indicated doses and were subjected to Western blot. JNK inhibition by JNK inhibitor II yields a dose-dependent reduction of MELK expression in GBM sphere (A and B) but not in normal spheres (C and D). DMSO is used as control vehicle. GAPDH is used as an internal control.
sc-12-0733_sm_SupplFigure8.tif858KFigure S8 (related to Figure 3): JNK inhibition decreases MELK expression in GBM but not normal cells Representative Western blot displaying the effects of 10 different pathway inhibitors on MELK expression in GBM sphere (GBM146) and normal spheres (16wf). Treatment with JNK inhibitor II (JNKiII), but not the other 9 inhibitors, selectively reduced MELK expression in GBM spheres without significantly affecting MELK expression in normal sphere sample (arrow head).
sc-12-0733_sm_SupplFigure9.tif565KFigure S9 (related to Figure 3): JNK has no noticeable effect on the MELK mRNA expression. JNK inhibitor II-treated GBM146 spheres were collected at indicated time points. RTPCR exhibits no noticeable change in the expression levels of MELK mRNA in GBM146 spheres by JNK inhibition. GAPDH is used as an internal control.
sc-12-0733_sm_SupplFigure10.tif1273KFigure S10 (related to Figure 3): Phosphorylated form of JNK2/3 is elevated in GSCs compared to normal progenitors. Western blot exhibits the expression of phosphorylated JNK2/3, but not phosphorylated JNK1, is elevated in GBM spheres (GBM146 and 157), compared to normal spheres (f16w). Total JNK expression serves as a control.
sc-12-0733_sm_SupplFigure11.pdf93KFigure S11 (related to Figure 4): Overexpression of the dominant-negative form of c-JUN diminishes MELK expression in GSCs. At 48 hours post-transfection of the dominant-negative (DN) form of c-JUN in GBM146 spheres, the expression levels of MELK are determined by Western blot analysis. GAPDH was used as an internal control.
sc-12-0733_sm_SupplFigure12.pdf68KFigure S12 (related to Figure 5): MELK negatively regulates p53 promoter activity. Left panel: Histogram representing the p53 promoter activity in U87 cells overexpressing MELK or GFP (control), determined by luciferase reporter assay. p value is ≤ 0.001. Right panel: U87 cells were co-transfected with p53-luciferase and shMELK+ MELK plasmids using Lipofectamine 2000 (Invitrogen). 48 hrs posttransfection, p53 promoter activity was measured by luciferase reporter assay (Promega). p value for shMELK with respect to non-target shRNA is ≤ 0.001. p value for shMELK+MELK with respect to shMELK is ≤ 0.05.
sc-12-0733_sm_SupplFigure13.tif9739KFigure S13 (related to Figure 5): p53 siRNA decreases mRNA level of p53. U87 cells were transfected with control or p53 siRNA using Lipofectamine RNAimax (Invitrogen). 72 hours post-transfection, mRNA level of p53 was measured by qRTPCR. p value ≤ 0.001. Experiment was done 3 times independently.
sc-12-0733_sm_SupplFigure14.tif2323KFigure S14 (related to Figure 6): Radiation increases MELK expression and MELK silencing increases p53 expression. Western blot shows that radiation treatment increases MELK expression in both GBM1600 (p53-intact) and GBM2313 (p53-mutant) cells. Note that depletion of MELK by MELK siRNA (siM) increases the level of p53 expression in p53-intact GBM1600 but not in p53-mutated GBM2313. GADPH was used as an internal control.
sc-12-0733_sm_SupplFigure15.pdf84KFigure S15 (related to Figure 6): Radiation does not increase CD133 expression in short duration. GBM 528 spheres were treated with 5 Gy of radiation, and the CD133 mRNA level was measured at indicated time points using qRTPCR. GAPDH was used as an internal control
sc-12-0733_sm_SupplData.pdf124KSupplementary Data

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.