Fig. 2

HOXB5 was overexpressed in glioma and correlated with poor patient survival, and regulated the proliferation of GSCs. a HOXB5 was expressed at higher levels in glioma tissues, compared with adjacent brain tissues as detected by qRT-PCR. (Adjacent vs. Tumor: n = 10, p < 0.0001, Student’s t-test). b-d The expression level of HOXB5 in different glioma tissues as detected by qRT-PCR (b), immunohistochemistry (c), and western blotting (d). Scale bar = 50 μm. (grade II, n = 20; grade III, n = 25; grade IV, n = 25; qRT-PCR: III vs. II: p < 0.01, IV vs. III: p < 0.01, IV vs. II: p < 0.0001, One-Way ANOVA; immunohistochemistry: III vs. II: p = 0.0042, IV vs. III: p = 0.0074, IV vs. II: p < 0.0001, One-Way ANOVA). e-g Kaplan-Meier analysis showed the prognostic significance of the 70 glioma patients, 45 LGG patients, and 25 GBM patients with high versus low HOXB5 expressions detected by qRT-PCR. (LGG: p = 0.0258; GBM: p = 0.0226; 70 cases: p = 0.0056; Log-rank test). h MTS assays showed that HOXB5 knockdown or overexpression affected the cell viability of GSCs. (GSC406: p < 0.0001; GSC201: p < 0.0001; One-Way ANOVA). i The EDU assays showed that HOXB5 knockdown or overexpression affected the proliferation of GSCs. Scale bar = 100 μm. (GSC406: p < 0.01; GSC201: p < 0.001; One-Way ANOVA). k The neurospheres formation assays showed that HOXB5 knockdown or overexpression affected the relative size of the neurospheres of GSCs. Scale bar = 20 μm. (GSC406: p < 0.01; GSC201: p < 0.01; One-Way ANOVA). j and l Limiting dilution assays showed that HOXB5 knockdown or overexpression affected the neurosphere-forming capacity of GSCs (GSC406: p < 0.001; GSC201: p < 0.001; ELDA analysis; circles represent corresponding points, triangles mean the point is outside of the log fraction number wells). EV: empty vector, OE: overexpression, NC: negative control, KD: knockdown. All data were expressed as the mean ± SD (three independent experiments). *p < 0.05; **p < 0.01; ***p < 0.001