Fig 2: PHF14 silencing in glioma cell lines decelerates proliferation and clonogenic growth. a, b The knockdown efficacy of PHF14 examined in U251, U87MG and A172 cells by qRT-PCR (a) and Western blot (b). c, d Cellular DNA replication is mitigated after PHF14 knocked down. Ten thousand U87MG, U251 and A172 cells per well were seeded in 96-well plates overnight. EdU and Hoechest co-staining was performed on the second day to assess cellular DNA replication activities (c) (Scale bar = 100 µm). Values are counted as EdU + cells/total cells (d). e–g PHF14 silenced cells showed weaker proliferation and colony formation potential. 500 U251 and A172 cells per well were cultured in 6-well plates in Complete growth medium for 10 days and then stained by 1% crystal violet for colony formation assay (e) and single colony morphology (f) (Scale bar = 200 µm). The number of colonies containing over 50 cells were counted up for analysis (g). h Cell colonies were stained by Phalloidin-cy3 after 10 days of culture for assessment of actin filaments containment and cellular protrusions (Scale bar = 50 µm) (*P < 0.05; **P = 0.01; ***P < 0.001; ****P < 0.0001)

Fig 3: PHF14 is overexpressed in glioblastoma multiforme (GBM) and lower PHF14 expression correlates with better outcome in GBM and LGG. a The cellular location of PHF14 confirmed by immunofluorescence (Scale bar = 50 µm). b Representative stainings of PHF14 in normal brain, grade II, grade III and grade IV glioma (Scale bar = 50 µm). c Correlation of the staining intensity of PHF14 with glioma grade. d–f Correlation of PHF14 mRNA expression with increasing WHO glioma grade according to REMBRANDT, TCGA and FRENCH data. g, h The relationship between PHF14 mRNA expression and molecular pathology of glioma. i, j Upregulation of PHF14 is a dismal prognostic factor in LGG by TCGA database. k Upregulation of PHF14 is a dismal prognostic factor in GBM by REMBRANDT database (*P < 0.05; **P = 0.01; ***P < 0.001; ****P < 0.0001)

Fig 4: PHF14 gene silencing enhances apoptosis and inhibits migration and invasion of glioma cells in vitro. a, b Negative control and siPHF14 transfected U251 and A172 cells were stained with AnnexinV-PE and 7-AAD after cultured in serum-free medium for 7 days. Flow cytometric analysis was done afterwards for identification of live (both negative), early apoptosis (AnnexinV-PE-positive) and late apoptosis/dead cells (both positive). c–f Migration and invasion assay showed weaker migration and invasion potential of PHF14-silenced cells. 25 thousand U251, U87MG and A172 cells were seeded into the upper chamber of a transwell insert (c, e) (Scale bar = 100 µm). After 24 h of incubation, cells which had adhered to the lower membrane of the inserts were fixed, stained with 1% crystal violet and counted for analysis (d, f). For invasion assays, the upper inserts were pre-coated with Matrigel (e, f). g, h 1000 per well negative control and siPHF14 transfected U87 cells were placed in ultra-low attachment (ULA) 96-well round bottom plates with 10% Matrigel solution and cultured for 7 days. The invasion area outlined by ImageJ or the longest invasion distance (represented by red solid line) (g) were measured by quantitation afterwards (h) (Scale bar = 50 µm) (*P < 0.05; **P = 0.01; ***P < 0.001; ****P < 0.0001)

Fig 5: RNA sequencing analysis upon PHF14 depleted cells. a 137 upregulated and 120 downregulated genes (Log2 (Fold Change) = - 1/= 1) were differentially expressed upon PHF14 knockdown with 5% False Discovery Rate (FDR). b Validation of the RNA sequencing analysis was performed by quantitative real time PCR. Pearson correlation of RNA Seq and qPCR is 0.68. c, d Gene ontology (GO) and KEGG pathway analyses results summarized that functional categories of genes significantly enriched upon PHF14 knockdown