KLF15-POU3F4 axis activates PI3K/AKT signaling to promote glioma progression

Introduction

Gliomas are common brain tumors (1,2). They originate from glial cells and exhibit high heterogeneity, invasiveness, and treatment resistance. Glioblastoma [GBM, World Health Organization (WHO) Grade 4] is the most malignant subtype with the worst prognosis (3,4). Despite continuous advances in surgical techniques, radiotherapy, and chemotherapy strategies in recent years, and the introduction of novel methods, such as tumor electric field therapy, the clinical outcomes of patients with GBM remain poor. The primary challenges in glioma treatment include diffuse infiltration and growth, blood-brain barrier limitations, significant intratumor heterogeneity, and a complex immune microenvironment (5). Therefore, to further reveal the underlying mechanisms of glioma occurrence and development, novel targets for early diagnosis, prognostic judgment, and precise treatment are urgently required.

As core regulators of gene expression networks, transcription factors have been reported to participate in the pathogenesis of different tumors (6). The Pit-Oct-Unc (POU) family of transcription factors has attracted significant attention in tumor development (7). POU class 3 homeobox 4 (POU3F4) is a member of this family, with its gene located on xq21.1 (8,9). POU3F4 was first reported to participate in auditory system development, and mutations can result in X-linked recessive deafness (10). POU family transcription factors [POU5F1/octamer-binding transcription factor 4 (OCT4)] may regulate tumor stemness, facilitate epithelial-mesenchymal transition (EMT), and enhance tumor cell invasion and metastasis (11,12). Research has found that POU3F4 is usually expressed at a low level in most cancers, but is upregulated in invasive breast cancer, GBM multiforme, hepatocellular carcinoma, and thyroid cancer, with the highest expression level in GBM multiforme. It is also involved in lung adenocarcinoma and neuroendocrine carcinoma (7,13). However, compared with other family members, studies on POU3F4 in gliomas remain limited. A deep understanding of the expression regulation mechanism of POU3F4 is crucial for revealing its role in glioma.

Given that POU3F4 itself is also a transcription factor, its expression is inevitably regulated by other upstream transcription factors. A deep understanding of the expression regulation mechanism of POU3F4 is crucial for revealing its role in glioma. Recent studies suggest that KLF15 (Kr ü ppel like factor 15), as an important transcription factor, plays a role in various cellular physiological processes, and its role in tumors is gradually gaining attention. Among various solid tumors, KLF15 mainly exerts tumor suppressive function. For example, in lung adenocarcinoma, KLF15 can partially upregulate the expression of cell cycle inhibitory proteins CDKN1A/p21 and CDKN2A/p15, effectively inhibiting the proliferation activity of tumor cells (14). In gastric cancer, KLF15 inhibits tumor growth and metastasis by regulating the TFAP2A-AS1/Nisch signaling axis (15). However, existing research suggests that KLF15 is highly expressed in gliomas and may exert a pro cancer effect opposite to the aforementioned tumor types. Upregulation of its expression is closely related to the malignant progression of tumors (16). Meanwhile, knocking down KLF15 can enhance the sensitivity of gliomas to treatment with temozolomide (TMZ). KLF15 can upregulate its expression by directly binding to the MGMT promoter region, playing an important role in the TMZ resistance process of gliomas (17). This phenomenon suggests that it may participate in the occurrence and development of gliomas through unique molecular mechanisms in the microenvironment.

The phosphatidylinositol 3-kinases/protein kinase B (PI3K/AKT) signaling pathway is one of the core signaling pathways that regulate cell survival, proliferation, metabolism, and apoptosis within cells. In gliomas, abnormal activation of this pathway is very common, with over 80% of GBMs exhibiting excessive activation of the PI3K/AKT pathway. Mutations or amplifications of upstream receptor tyrosine kinases (such as EGFR, PDGFR), as well as deletions or mutations of PTEN tumor suppressor genes, can all lead to sustained activation of the PI3K/AKT pathway, thereby promoting proliferation, migration, and invasion of glioma cells, inhibiting cell apoptosis, and inducing therapeutic resistance (18,19). Therefore, the PI3K/AKT pathway is considered an important target for targeted therapy of gliomas.

Based on the above background, this study was designed to clarify the potential pathogenic role of POU3F4 in gliomas using a multi-level research strategy. First, bioinformatics analysis was performed using large public databases, such as The Cancer Genome Atlas (TCGA) and The Cancer Imaging Archive Atlas (ATLAS) to assess the role of POU3F4 in glioma. Second, POU3F4 expression in glioma tissues and plasma samples was verified and its potential diagnostic value was examined. At the functional level, we knocked down POU3F4 expression in glioma cells using RNA interference technology and assessed its regulatory role in glioma cell growth and metastasis in vitro. In addition, we identified the upstream transcription factor KLF15 of POU3F4 through JASPAR database prediction combined with ChIP experiments, and verified the function of the KLF15-POU3F4 axis through response experiments. Finally, we explored its downstream molecular mechanisms and confirmed that the KLF15-POU3F4 axis promotes glioma progression by activating the PI3K/AKT signaling pathway. We present this article in accordance with the MDAR reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1-0326/rc).

Methods

Bioinformatic analysis

TCGA database

The expression and prognostic analysis of POU3F4 in gliomas were based on transcriptome sequencing data from the TCGA database. All data is accessed through the official TCGA data portal (https://portal.gdc.cancer.gov). The samples obtained and included in the analysis are histologically diagnosed gliomas (including GBM and low-grade gliomas), with complete RNA seq data and survival follow-up records.

ATLAS database

The expression data of POU3F4 in normal brain tissue and other tumor tissues were obtained from the Human Protein Atlas database (version v21.1, https://www.proteinatlas.org). The screening criteria for differentially expressed genes are: |log₂ fold difference| >1 and corrected P<0.05.

Samples

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Affiliated Brain Hospital of Nanjing Medical University (No. 20210304). All patients provided informed consent before surgery. Glioma tumor tissues (n=48) and paired normal tissues adjacent to the tumor (>5 cm from the tumor edge, n=48) were isolated from patients with pathologically confirmed gliomas. Tissue and plasma samples were obtained from patients with glioma and healthy candidates from Department of Neurosurgery, The Affiliated Brain Hospital of Nanjing Medical University, between September 2022 and September 2024. The samples were stored at −80 ℃. Plasma samples were collected from patients with glioma (n=48) and healthy volunteers (n=48) who were examined at the Physical Examination Center of our hospital. Blood samples were collected using ethylenediaminetetraacetic acid anticoagulant tubes and stored at −80 ℃.

Reverse transcription quantitative polymerase chain reaction (qPCR)

Total RNA was isolated using TRIzol reagent (Invitrogen, USA). RNA purity was assessed using a Nanodrop 2000, and 1 µg of RNA was used for genomic DNA removal and reverse transcription into complementary DNA using the PrimeScript-RT kit (TaKaRa, Dalian, China). Subsequently, the SYBR Premix Ex Taq II kit (Takara) was used for amplification on ABI 7500 PCR system (Applied Biosystems, USA). The PCR conditions were as follows: 95 ℃ for 30 s, followed by 40 cycles of 95 ℃ for 5 s and 60 ℃ for 34 s. The primer sequences were as follows: POU3F4 upstream: 5'-GGGAGTCCTTTCCGCAACC-3', downstream: 5'-AGCGGTGATGGATGATCGC-3', and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) upstream: 5'-TGTGGGCATCAATGGATTTGG-3', downstream: 5'-ACACCATGTATTCCGGGTCAAT-3'. Gene expression was calculated using the 2^-ΔΔCT method.

Cell culture and transfection

Human glioma cells (U87MG, U251MG, T98G, A172, and LN229) and normal human astrocytes (NHAs) were obtained from the American Typical Culture Collection. All cells were maintained at 37 ℃ and 5% CO₂. Glioma cells were maintained using Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS, GIBCO) and 1 % penicillin/streptomycin, whereas the NHA cells were maintained using a special astrocyte medium (Science cell, USA). Two pairs of small interfering RNAs and negative controls (synthesized by Shanghai Jima Pharmaceutical Technology Co., Ltd.) were designed and synthesized targeting the human POU3F4 gene. Two pairs of small interfering RNAs (si-KLF15-1 # and si-KLF15-2 #) and a negative control (synthesized by Shanghai Jima Pharmaceutical Technology Co., Ltd.) were designed and synthesized targeting the human KLF15 gene. The KLF15 OE plasmid (pcDNA3.1-KLF15) and the empty vector control (pcDNA3.1-Empty) were synthesized by the same company. When the cell fusion degree reaches 70–80%, according to the manufacturer’s protocol, siRNA or plasmid is transfected into U87MG and U251MG cells using Lipofectamine 3000 (Thermo Fisher Scientific). For the response experiment, cells were transfected simultaneously with si-KLF15 and pcDNA3.1-POU3F4 (or corresponding controls).

Cell Counting Kit-8 (CCK-8)

The CCK-8 method indirectly reflects cell viability by detecting the ability of intracellular dehydrogenases to reduce WST-8 to formate products, and the amount of formate produced is proportional to the number of live cells. Inoculate the transfected HS683 and U251MG cells into a 96-well plate at a density of 3,000 cells per well, with 5 replicates per well. After culturing for 24, 48, and 72 hours, 10 µL of CCK-8 reagent (Beyotime) was added to each well and incubated for 2 hours at 37 ℃. Subsequently, the absorbance value at 450 nm wavelength was detected using a BioTek (USA) enzyme-linked immunosorbent assay (ELISA) reader. Cell viability is expressed as the percentage of absorbance values between the experimental group and the control group, with the control group being cells transfected with negative control siRNA. Independently repeat the experiment 3 times at each time point.

Clone formation

The clone formation experiment evaluates the proliferation ability of cells by counting the cell colonies formed by the division and proliferation of individual cells. Inoculate the transfected cells at a density of 500 cells per dish into a 6-cm culture dish, with 3 replicates per group. Cultivate for 14 days in a 37 ℃, 5% CO₂ incubator, with the culture medium changed every 3 days. After cultivation, fix the cells with 4% paraformaldehyde for 30 minutes, stain with 0.1% crystal violet for 30 minutes, wash and dry, take photos, and count cell colonies with a diameter greater than 1 mm. The colony formation rate is expressed as the percentage of the number of experimental group colonies to the number of control group colonies, and the experiment is independently repeated for 3 times.

5-Ethynyl-2’-deoxyuridine (EdU) assay

The EdU method utilizes the principle of thymidine analogue EdU being incorporated into replication chains during DNA synthesis, and fluorescently labels proliferating cells through click chemistry reactions. Inoculate the transfected cells into a 24 well plate at a density of 5×10⁴ cells per well, with 3 replicates per group. After 24 hours of cultivation, EdU working solution was added to the final concentration of 10 µM, and incubation was continued for 2 hours. Subsequently, the cells were fixed with 4% paraformaldehyde for 30 minutes, permeabilized with 0.5% Triton X-100 for 20 minutes, and incubated with click reaction solution at room temperature in the dark for 30 minutes. Finally, the nuclei were stained with Hoechst 33342 for 10 minutes. Randomly select 5 fields of view under a fluorescence microscope for imaging, and the proliferation rate is expressed as the percentage of EdU positive cell nuclei to the total Hoechst stained cell nuclei. The experiment is independently repeated for 3 times.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay

The TUNEL method utilizes terminal deoxynucleotide transferase to connect fluorescein labeled dUTP to the 3’-OH end of the fragmented DNA in apoptotic cells, thereby achieving in situ detection of apoptotic cells. The experiment used TUNEL cell apoptosis detection kit (Roche, Basel, Switzerland). Inoculate the transfected cells into a 24-well plate at a density of 5×10⁴ cells per well, with 3 replicates per group. After 48 hours of cultivation, fix with 4% paraformaldehyde for 30 minutes, permeabilize with 0.2% Triton X-100 for 20 minutes, add TUNEL reaction mixture and incubate at 37 ℃ in the dark for 1 hour. Wash with PBS and stain the nuclei with DAPI for 10 minutes. Randomly select 5 fields of view under a fluorescence microscope to take pictures, and the apoptosis rate is expressed as the percentage of TUNEL positive cell nuclei to the total DAPI stained cell nuclei. The experiment is independently repeated for 3 times.

Wound healing assay

Scratch experiments evaluate cell migration by creating cell-free areas on a monolayer of fused cells and observing their ability to migrate to blank areas. Inoculate the transfected cells into a 6-well plate at a density of 2×10⁵ cells per well, with 3 replicates per group. When the cell fusion degree reaches 100%, use a 200-µL sterile pipette tip to draw a straight line scratch vertically, wash with PBS three times to remove detached cells, and replace with a maintenance medium containing 2% fetal bovine serum to continue culturing. Take photos with an inverted microscope at 0 and 24 hours after scratching, and randomly select 3 different positions for each hole. Measure the scratch width using ImageJ software, and calculate the migration rate as (scratch width of 0 h − scratch width of 24 h)/scratch width of 0 h multiplied by 100%. The experiment was independently repeated 3 times.

Transwell assay

The Transwell invasion experiment uses a small chamber coated with Matrigel gel to simulate the extracellular matrix barrier, and evaluates the invasion ability of cells by counting the number of cells that penetrate to the lower surface of the chamber. Dilute Matrigel gel with pre cooled serum-free medium at a ratio of 1:8, take 50 µL and evenly spread it on the bottom of the Transwell chamber. Place it at 37 ℃ for 2 hours to solidify. The transfected cells were resuspended in serum-free medium to 5×10⁴ cells/200 µL and added to the upper chamber. 600 µL of medium containing 20% fetal bovine serum was added to the lower chamber as a chemotactic agent. After 24 hours of cultivation, wipe off the surface cells of the upper chamber with a cotton swab, fix the chamber with 4% paraformaldehyde for 30 minutes, and stain with 0.1% crystal violet for 30 minutes. Randomly select 5 fields of view (200×) under an inverted microscope to take pictures and count the number of penetrating cells. The average number of cells in each group’s 5 fields of view is used to represent the invasion ability. The experiment is independently repeated for 3 times.

Western blotting (WB)

Forty-eight hours after transfection, radioimmunoprecipitation assay buffer (containing 1% phenylmethylsulfonyl fluoride and phosphatase inhibitor) was used for extracting total proteins, and protein concentration was assessed using the bicinchoninic acid assay method. Subsequently, 30 µg of protein samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes. The membranes were blocked with skimmed milk (5%) for 1 h and incubated overnight at 4 ℃ with the following primary antibodies: anti-POU3F4 (1:1,000; Abcam), anti-PI3K (1:1,000; CST), anti-p-PI3K (1:1,000; CST), anti AKT (1:2,000; CST), anti-p-AKT (ser473) (1:2,000; CST), and anti GAPDH (1:5,000; Abcam). On day 2, the samples were washed with Tris-buffered saline with Tween 20 and treated with horseradish peroxidase-labeled secondary antibodies (1:5,000, room temperature, 1 h). Finally, the signals were visualized using an ECL Chemiluminescence Kit (Millipore, USA) and a chemiluminescence imaging system (Tanon, China).

Prediction of transcription factor binding sites

Using JASPAR database (https://jaspar.genereg.net) Predict transcription factors that may bind to the POU3F4 promoter region (2,000 bp upstream of the transcription start site). The screening criteria are relative scores >0.80 and P<0.05.

Chromatin immunoprecipitation (ChIP) experiment

The ChIP experiment was conducted according to the instructions of the reagent kit (Millipore, USA). HS683 and U251MG cells were crosslinked with 1% formaldehyde at room temperature for 10 minutes, and the reaction was terminated by adding glycine. After cell lysis, the chromatin was sonicated to fragment sizes of 200–600 bp. 5% of the lysate was taken as the input control, while the remaining lysate was added with anti-KLF15 antibody (Abcam, 1:50) and negative control IgG, and incubated overnight at 4 ℃. Add Protein A/G magnetic beads to enrich the complex, and purify the DNA after washing, elution, and reverse crosslinking. Use POU3F4 promoter region specific primers for qPCR detection. The results are expressed as a percentage of Input.

Statistical analysis

Data are presented as the mean ± standard deviation. Statistical analyses were performed using GraphPad Prism software (version 9.0). Student’s t-test was used for two-group comparisons, whereas one-way analysis of variance was conducted for multiple-group comparisons. Statistical significance was set at P<0.05.

Results

Bioinformatic analysis for POU3F4 expression in glioma

First, POU3F4 expression in gliomas was examined using bioinformatics methods. Data from TGCA database revealed that POU3F4 was upregulated in gliomas (Figure 1A), and the results of survival analysis indicated that patients with gliomas with downregulated POU3F4 had better prognosis (Figure 1B). Subsequently, POU3F4 expression in brain tissue was examined using the ATLAS database. The POU3F4 structure is illustrated in Figure 1C. Across cancer types, POU3F4 levels were very low, except in brain cancer (Figure 1D). However, single-cell analysis revealed low POU3F4 levels in normal brain tissue (Figure 1E).

Figure 1 Bioinformatic analysis for POU3F4 expression in glioma. (A) POU3F4 expression in glioma in TCGA. (B) Results of survival analysis. (C) Structure of POU3F4 protein. (D) Expression of POU3F4 in all type of cancers. (E) Results of single cell analysis. *, P<0.05. GBM, glioblastoma; TCGA, The Cancer Genome Atlas; TPM.

Upregulated POU3F4 in glioma

Based on the bioinformatics results, we collected glioma and adjacent normal tissue samples and compared POU3F4 expression levels. POU3F4 was significantly upregulated in glioma tumors than that in adjacent normal tissue (Figure 2A). Additionally, plasma samples were collected from patients with glioma and healthy controls, and POU3F4 levels were significantly higher in patients with glioma than that in the controls (Figure 2B). For clinicopathological correlation analysis, POU3F4 expression levels were analyzed using median value. High POU3F4 expression in tumor tissues was significantly associated with higher WHO grade (P=0.02), increased tumor size (P=0.04) and lower Karnofsky Performance Status (KPS) score (P=0.009) (Table 1). High POU3F4 expression in plasma of the patients was significantly associated with higher WHO grade (P=0.04), male (P=0.04) and lower KPS score (P=0.004) (Table 2).

Figure 2 POU3F4 was up-regulated in glioma. (A) Expressions of POU3F4 in glioma tumor tissue and the adjacent normal tissue. (B) Expressions of POU3F4 in plasma of glioma patients and healthy controls. (C) ROC curve for tissue expression of POU3F4. (D) ROC curve for plasma expression of POU3F4. (E) Correlation between expressions of POU3F4 in glioma tumor tissue and plasma of glioma patients. ***, P<0.001. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.

Subsequently, ROC analysis revealed that the area under the curve (AUC) for tissue POU3F4 was 0.9115 [95% confidence interval (CI): 0.8439–0.9791] (Figure 2C). Using the optimal cut-off value of 1.663, tissue POU3F4 distinguished glioma tissue from adjacent normal tissue with a sensitivity of 79.17% and specificity of 93.75%. Whereas the AUC for plasma POU3F4 was 0.7675 (95% CI: 0.6716–0.8596) (Figure 2D). Using the optimal cut-off value of 1.072, plasma POU3F4 distinguished glioma patients from healthy controls with a sensitivity of 72.92% and specificity of 70.83%. These results indicate that POU3F4 levels in both tissue and plasma may serve as suitable biomarkers for glioma diagnosis. Additionally, correlation analysis indicated that POU3F4 expression in glioma tissues and plasma samples was positively correlated (Figure 2E).

POU3F4 knockdown reduces glioma cell growth

We assessed the effects of POU3F4 expression on glioma cell growth and apoptosis. POU3F4 expression in different glioma cells, including HS683, U251MG, T98G, A172, and LN229, and NHAs was examined. POU3F4 expression levels were higher in all glioma cell lines than those in NHAs, specifically in HS683 and U251MG cells (Figure 3A). Therefore, these cell lines were used for subsequent experiments. Subsequently, the two cell lines were treated with POU3F4 siRNA-1 and POU3F4 siRNA-2, with siRNA-2 exhibiting more significant inhibitory effects. Therefore, siRNA-2 was used for the cell experiments (Figure 3B,3C).

Figure 3 POU3F4 expression in glioma cells. (A) Expressions of POU3F4 in glioma cell lines and normal human astrocytes NHA. (B) mRNA expression of POU3F4 in HS683 and U251MG cells after transfection of siRNA. (C) Protein expression of POU3F4 in HS683 and U251MG cells after transfection of siRNA. ***, P<0.001. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; NC, negative control; NHA, normal human astrocyte.

After transfection with POU3F4 siRNA, HS683 and U251MG cell viability was assessed using the CCK-8 method. POU3F4 siRNA significantly reduced the viability of HS683 and U251MG cells (Figure 4A). HS683 and U251MG cell proliferation was assessed using colony formation and EdU assays. POU3F4 siRNA inhibited the proliferation of both HS683 and U251MG cells (Figure 4B,4C). Finally, HS683 and U251MG cell apoptosis was assessed through TUNEL staining, and the results in Figure 4D indicate that POU3F4 siRNA increased HS683 and U251MG cell apoptosis.

Figure 4 Knockdown of POU3F4 decrease the proliferation and apoptosis of glioma cells. (A) Results of CCK-8 assay. (B) Results of colony formation assay. Staining method: ; magnification: . (C) Results of EdU assay. Scale bar =. (D) Results of TUNEL staining assay. Scale bar =. **, P<0.01; ***, P<0.001. CCK-8, Cell Counting Kit-8; DAPI, ; EdU, 5-ethynyl-2’-deoxyuridine; NC, negative control; OD, optical density; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.

POU3F4 knockdown reduces glioma cell metastasis

We assessed whether POU3F4 affects glioma cell metastasis using scratch and Transwell assays. The results of scratch wound healing assay indicated that POU3F4 knockdown reduced HS683 and U251MG cell migration (Figure 5A). Transwell assay results indicated that POU3F4 knockdown decreased HS683 and U251MG cell invasion (Figure 5B).

Figure 5 Knockdown of POU3F4 decrease the migration and invasion ability of glioma cells. (A) Results of scratch wound-healing assay. Staining method: ; magnification: . (B) Results of transwell assay. Staining method: ; magnification: .**, P<0.01; ***, P<0.001. NC, negative control.

POU3F4 knockdown inhibited PI3K/AKT signaling in glioma cells

Finally, the effect of POU3F4 expression on PI3K/AKT signaling in glioma cells was assessed. POU3F4 siRNA did not affect PI3K or AKT mRNA (Figure 6A) or protein expression (Figure 6B). In contrast, POU3F4 siRNA downregulated p-PI3K and p-AKT (Figure 6B).

Figure 6 Knockdown of POU3F4 inhibited PI3K/AKT signaling in glioma cells. (A) Results of RT-qPCR analysis. (B) Results of WB analysis. NS, not significant (P≥0.05); **, P<0.01; ***, P<0.001. NC, negative control; RT-qPCR, reverse transcription quantitative polymerase chain reaction; WB, western blotting.

KLF15 as an upstream transcription factor regulates POU3F4 expression

To explore the upstream regulatory mechanism of POU3F4, JASPAR database was used to predict transcription factors that may bind to the promoter region of POU3F4. The results showed that the DNA binding site motif of KLF15 was located at a specific position in the POU3F4 promoter region (Figure 7A,7B). The ChIP experiment further validated the in vivo binding ability of KLF15 in the POU3F4 promoter region. The results showed that compared with the IgG control group, the DNA enrichment of the POU3F4 promoter region was significantly increased in the KLF15 antibody group (Figure 7C). Subsequently, the regulatory effect of KLF15 on POU3F4 expression was investigated through overexpression and knockdown experiments. Overexpression of KLF15 significantly upregulated the mRNA and protein expression levels of POU3F4 in HS683 and U251MG cells (Figure 7D,7E); Knocking down KLF15 significantly downregulated the expression level of POU3F4 (Figure 7C,7D, P<0.05). These results indicate that KLF15, as a transcription factor, positively regulates the expression of POU3F4.

Figure 7 JASPAR database prediction of KLF15 binding to the POU3F4 promoter region and ChIP validation. (A,B) JASPAR database predicted the presence of KLF15 binding motifs in the POU3F4 promoter region. (C) ChIP assay validated the binding of KLF15 to the POU3F4 promoter region. (D,E) Overexpression of KLF15 upregulated the mRNA and protein expression levels of POU3F4 in HS683 and U251MG cells. (F,G) Knockdown of KLF15 downregulated the mRNA and protein expression levels of POU3F4 in HS683 and U251MG cells. NS, not significant (P≥0.05); **, P<0.01; ***, P<0.001. ChIP, GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

KLF15-POU3F4 axis regulates glioma cell proliferation, migration, and invasion

To verify whether KLF15 affects the malignant phenotype of glioma cells by regulating POU3F4, we conducted a response experiment. The western blot results showed that the knockdown efficiency of siRNA-KLF15 2# was better than that of 1# (Figure 8A), so siRNA2 was selected for subsequent experiments. Knocking down KLF15 significantly reduced the protein expression of POU3F4, while overexpression of POU3F4 partially reversed this effect (Figure 8B). The results of CCK-8, clone formation, EdU and TUNEL staining experiments showed that knocking down KLF15 significantly inhibited the proliferation and increase apoptosis ability of HS683 and U251MG cells, while the combined overexpression of POU3F4 partially offset the decrease in proliferation ability and increase in apoptosis ability caused by knocking down KLF15 alone (Figure 8C-8F). The scratch experiment and Transwell experiment results showed that knocking down KLF15 significantly inhibited the migration and invasion ability of glioma cells, while co overexpression of POU3F4 partially reversed this effect (Figure 9A,9B).

Figure 8 The KLF15-POU3F4 axis regulates growth of glioma cells. (A) Western blot validation of si-KLF15 knockdown efficiency. (B) Knockdown of KLF15 decreased POU3F4 protein expression, which was partially reversed by POU3F4 overexpression. (C) CCK-8 assay detected cell viability. (D) Colony formation assay detected cell proliferation ability. Staining method: ; magnification: . (E) Results of EdU assay. Magnification: . (F) Results of TUNEL staining assay. Magnification: . **, P<0.01. CCK-8, Cell Counting Kit-8; DAPI; EdU, 5-ethynyl-2’-deoxyuridine; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; OD, optical density; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.

Figure 9 The KLF15-POU3F4 axis regulates migration and invasion of glioma cells. (A) Wound healing assay detected cell migration ability. Staining method: ; magnification:. (B) Transwell assay detected cell invasion ability. Staining method:; magnification:. **, P<0.01.

The KLF15-POU3F4 axis functions through the PI3K/AKT signaling pathway

Finally, the impact of KLF15-POU3F4 axis on the PI3K/AKT signaling pathway was evaluated. Western blot results showed that knocking down KLF15 did not affect the total protein expression levels of PI3K and AKT, but significantly reduced the expression levels of p-PI3K and p-AKT; However, co overexpression of POU3F4 partially reversed the decrease in phosphorylation levels caused by KLF15 knockdown (Figure 10). This suggests that KLF15 activates the PI3K/AKT signaling pathway by regulating the expression of POU3F4, thereby affecting the occurrence and development of gliomas.

Figure 10 Effect of the KLF15-POU3F4 axis on the PI3K/AKT signaling pathway. Results of western blot analysis. **, P<0.01. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Discussion

This study systematically discussed the expression, clinical significance, and biological function of POU3F4 in gliomas using comprehensive bioinformatics analysis, clinical sample verification, and in-depth in vitro functional experiments. The primary findings of this study were as follows: (I) POU3F4 was upregulated in glioma and associated with an undesired prognosis; (II) POU3F4 in patient plasma was significantly upregulated and positively associated with tissue expression, indicating its great potential as a novel noninvasive diagnostic biomarker; (III) in vitro, POU3F4 knockdown effectively inhibited the growth and metastasis of glioma cells, while increasing cell apoptosis; (IV) KLF15 has been identified as an upstream transcription factor of POU3F4, which can directly bind to the POU3F4 promoter region and positively regulate its expression; and (V) the KLF15-POU3F4 axis promotes glioma progression by activating the PI3K/AKT signaling pathway. These data indicated that POU3F4 plays a primary role in the occurrence and development of gliomas.

First, our bioinformatics analysis based on the TCGA and ATLAS databases confirmed that POU3F4 was specifically overexpressed in gliomas, in contrast to its low expression in normal brain tissue. This indicated biological significance. POU3F4 plays a central role in embryonic development, cell fate determination, and differentiation (9,10). Notably, other POU family transcription factors, such as POU3F2/BRN-2 and OCT4 are the primary carcinogenic drivers in glioma (20,21) and melanoma (22,23) and facilitate tumor progression by regulating genes involved in stem maintenance, cell cycle, and EMT. Our study identified POU3F4 as a carcinogenic transcription factor for the first time, revealing that its abnormal activation in gliomas may reprogram cellular transcriptional networks and drive the malignant phenotype of gliomas.

In contrast, POU3F4 was upregulated at the tissue level and in the plasma of patients, with a significant positive association between them. This indicates that POU3F4 may be actively secreted or released into the blood circulation by other means, such as exosomes or necrotic cells. ROC curve analysis supported its potential as a diagnostic biomarker. Although tissue biopsy remains the gold standard for glioma diagnosis, liquid biopsy technology has been a major research focus because of its invasiveness and sampling bias (24). This study identified a novel blood-based candidate molecule for the noninvasive auxiliary diagnosis and postoperative monitoring of glioma. In the future, we will need to verify its sensitivity and specificity in larger cohorts.

POU3F4 maintained the malignant phenotype of glioma through a series of in vitro experiments. POU3F4 knockdown suppressed the growth and metastasis of U87MG and U251MG cells while inducing apoptosis. These phenotypes are consistent with the inhibitory effects of numerous known oncogenes, such as myc and epidermal growth factor receptor (EGFR) (25,26). Notably, POU3F4 knockdown affected glioma cell metastasis that aligned with the functions of other members of the POU family (7,27).

This study further explored the upstream regulatory mechanism of POU3F4. Through bioinformatics prediction and ChIP qPCR experiments, we confirmed that KLF15 can specifically bind to the promoter region of POU3F4, and this binding also exists in vivo. KLF15 is a member of the Kr ü ppel like transcription factor family and plays an important regulatory role in various tumors. Previous studies have shown that the function of KLF15 in gliomas is tissue-specific. Some studies have reported it as a tumor suppressor, while others have found it to have a pro cancer effect (28-30). This study found that KLF15 activates POU3F4 expression through transcription, thereby promoting the proliferation, migration, and invasion of glioma cells, and activating the PI3K/AKT signaling pathway. The response experiment further confirmed that overexpression of POU3F4 can partially reverse the tumor suppressive effect caused by KLF15 knockdown, indicating that the KLF15-POU3F4 axis is an important regulatory pathway in glioma progression. These findings fill the research gap in the upstream regulatory mechanism of POU3F4 and provide new potential targets for targeted therapy of gliomas.

Conclusions

Our study made significant progress in mechanistic exploration. We identified KLF15 as an upstream transcription factor of POU3F4, which can directly bind to its promoter region and positively regulate its expression. Functional recovery experiments have shown that the KLF15-POU3F4 axis promotes the proliferation, migration, and invasion of glioma cells by activating the PI3K/AKT signaling pathway. In addition, knocking down POU3F4 does not affect the total protein levels of PI3K and AKT, but reduces their phosphorylation levels, indicating that POU3F4 does not regulate the expression of the main components of this pathway, but promotes the activation of the PI3K/AKT signaling pathway. The PI3K/AKT pathway is a major signaling pathway that facilitates cell survival and proliferation. It is abnormally activated in over 80% of GBMs and is a crucial target for tumor therapy (31,32). Although the specific mechanisms by which KLF15 regulates POU3F4 and how POU3F4 activates the PI3K/AKT pathway still need further clarification, the findings of this study provide new potential targets for targeted therapy of gliomas. Future research should identify the downstream target gene network of POU3F4 through techniques such as RNA seq and ChIP seq, and further explore the specific molecular mechanisms of the KLF15-POU3F4 axis in the occurrence and development of gliomas.

Acknowledgments

We would like to thank Editage (http://www.editage.com) for editing and reviewing this manuscript for English language.

Footnote

Reporting Checklist: The authors have completed the MDAR reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1-0326/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1-0326/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1-0326/prf

Funding: This study was supported by Naning Medical University (No. NMUB20240211).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2026-1-0326/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The Affiliated Brain Hospital of Nanjing Medical University (No. 20210304). All patients provided informed consent before surgery.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wang LM, Englander ZK, Miller ML, et al. Malignant Glioma. Adv Exp Med Biol 2023;1405:1-30. [Crossref] [PubMed]
2. Xu S, Tang L, Li X, et al. Immunotherapy for glioma: Current management and future application. Cancer Lett 2020;476:1-12. [Crossref] [PubMed]
3. Berger TR, Wen PY, Lang-Orsini M, et al. World Health Organization 2021 Classification of Central Nervous System Tumors and Implications for Therapy for Adult-Type Gliomas: A Review. JAMA Oncol 2022;8:1493-501. [Crossref] [PubMed]
4. Salvalaggio A, Pini L, Bertoldo A, et al. Glioblastoma and brain connectivity: the need for a paradigm shift. Lancet Neurol 2024;23:740-8. [Crossref] [PubMed]
5. Yousaf S, Ahmed S, Arshad M, et al. Glioma-Immune crosstalk in the tumor microenvironment: mechanistic insights and therapeutic translation. Med Oncol 2026;43:172. [Crossref] [PubMed]
6. Rao SV, Young L, Cheeseman D, et al. Transcription factor switching drives subtype-specific pancreatic cancer. Nat Genet 2025;57:3016-26. [Crossref] [PubMed]
7. Chen W, Ji Y, Wang R, et al. Investigating POU3F4 in cancer: Expression patterns, prognostic implications, and functional roles in tumor immunity. Heliyon 2025;11:e41587. [Crossref] [PubMed]
8. Bernardinelli E, Huber F, Roesch S, et al. Clinical and Molecular Aspects Associated with Defects in the Transcription Factor POU3F4: A Review. Biomedicines 2023;11:1695. [Crossref] [PubMed]
9. Zhang L, Wang J, Xu N, et al. POU3F4 up-regulates Gli1 expression and promotes neuronal differentiation and synaptic development of hippocampal neural stem cells. Stem Cell Res Ther 2024;15:440. [Crossref] [PubMed]
10. Dang J, Bian P, Chen C, et al. Impact of POU3F4 mutation on cochlear development and auditory function. Cell Commun Signal 2025;23:121. [Crossref] [PubMed]
11. Wang Y, Jin Y, Shen X, et al. POU2F3: A Sensitive and Specific Diagnostic Marker for Neuroendocrine-low/negative Small Cell Lung Cancer. Am J Surg Pathol 2023;47:1059-66. [Crossref] [PubMed]
12. Jung SN, Kang YE, Lee GH, et al. Brn3a/Pou4f1 Functions as a Tumor Suppressor by Targeting c-MET/STAT3 Signaling in Thyroid Cancer. J Clin Endocrinol Metab 2020;105:dgaa316. [Crossref] [PubMed]
13. Masawa M, Sato-Yazawa H, Kashiwagi K, et al. REST Inactivation and Coexpression of ASCL1 and POU3F4 Are Necessary for the Complete Transformation of RB1/TP53-Inactivated Lung Adenocarcinoma into Neuroendocrine Carcinoma. Am J Pathol 2022;192:847-61. [Crossref] [PubMed]
14. Wang X, He M, Li J, et al. KLF15 suppresses cell growth and predicts prognosis in lung adenocarcinoma. Biomed Pharmacother 2018;106:672-7. [Crossref] [PubMed]
15. Zhao X, Chen L, Wu J, et al. Transcription factor KLF15 inhibits the proliferation and migration of gastric cancer cells via regulating the TFAP2A-AS1/NISCH axis. Biol Direct 2021;16:21. [Crossref] [PubMed]
16. Chen L, Jing SY, Liu N, et al. MiR-376a-3p alleviates the development of glioma through negatively regulating KLF15. Eur Rev Med Pharmacol Sci 2020;24:11666-74. [Crossref] [PubMed]
17. Qu X, Liu X, Zhang Y, et al. Depletion of Kruppel-like factor 15 sensitized gliomas to temozolomide cytotoxicity through O (6)-methylguanine-DNA methyl-transferase. Biochem Biophys Rep 2021;27:101058. [Crossref] [PubMed]
18. Shi Z, Shen D, Wu J, et al. TNKS1 mediates the PTEN-PI3K/AKT pathway to regulate glycolysis and proliferation in gliomas. IBRO Neurosci Rep 2026;20:186-95. [Crossref] [PubMed]
19. Li S, Liu K, Li X, et al. TFAP2C Affects PI3K/AKT/mTOR Pathway-Mediated EMT for Glioblastoma Development Through Transcriptional Regulation. Cell Biol Int 2025;49:1411-24. [Crossref] [PubMed]
20. Yao J, Wang L. Integrin α3 Mediates Stemness and Invasion of Glioblastoma by Regulating POU3F2. Curr Protein Pept Sci 2023;24:247-56. [Crossref] [PubMed]
21. Dai Y, Yu T, Yu C, et al. ISG15 enhances glioma cell stemness by promoting Oct4 protein stability. Environ Toxicol 2022;37:2133-42. [Crossref] [PubMed]
22. Cesar MCM, de Sant’ana AKC, Mortara RA, et al. Peptides derived from the POU domain of BRN2 show antitumor activity against murine melanoma model cells in vitro and in vivo. Cancer Chemother Pharmacol 2025;95:67. [Crossref] [PubMed]
23. Zhang Y, Urquijo MA, Zitnay RG, et al. A BRN2:MYC transcriptional axis regulates interconversion between therapy-resistant and tumorigenic phenotypes in melanoma. Cell Rep 2025;44:116675. [Crossref] [PubMed]
24. Nayak L, Bettegowda C, Scherer F, et al. Liquid biopsy for improving diagnosis and monitoring of CNS lymphomas: A RANO review. Neuro Oncol 2024;26:993-1011. [Crossref] [PubMed]
25. Zhao K, Wang Q, Wang Y, et al. EGFR/c-myc axis regulates TGFβ/Hippo/Notch pathway via epigenetic silencing miR-524 in gliomas. Cancer Lett 2017;406:12-21. [Crossref] [PubMed]
26. Sabbah DA, Hajjo R, Sweidan K. Review on Epidermal Growth Factor Receptor (EGFR) Structure, Signaling Pathways, Interactions, and Recent Updates of EGFR Inhibitors. Curr Top Med Chem 2020;20:815-34. [Crossref] [PubMed]
27. Lekkos K, Bhuiyan AA, Albloshi AMK, et al. Validation of positional candidates Rps6ka6 and Pou3f4 for a locus associated with skeletal muscle mass variability. G3 (Bethesda) 2024;14:jkae046. [Crossref] [PubMed]
28. Kanyomse Q, Le X, Tang J, et al. KLF15 suppresses tumor growth and metastasis in Triple-Negative Breast Cancer by downregulating CCL2 and CCL7. Sci Rep 2022;12:19026. [Crossref] [PubMed]
29. Wang N, Cao QX, Tian J, et al. Circular RNA MTO1 Inhibits the Proliferation and Invasion of Ovarian Cancer Cells Through the miR-182-5p/KLF15 Axis. Cell Transplant 2020;29:963689720943613. [Crossref] [PubMed]
30. Ghouse J, Gellert-Kristensen H, O’Rourke CJ, et al. Genome-wide meta-analysis identifies nine loci associated with higher risk of hepatocellular carcinoma development. JHEP Rep 2025;7:101485. [Crossref] [PubMed]
31. Barzegar Behrooz A, Talaie Z, Jusheghani F, et al. Wnt and PI3K/Akt/mTOR Survival Pathways as Therapeutic Targets in Glioblastoma. Int J Mol Sci 2022;23:1353. [Crossref] [PubMed]
32. Kundu P, Jain R, Kanuri NN, et al. DNA Methylation in Recurrent Glioblastomas: Increased TEM8 Expression Activates the Src/PI3K/AKT/GSK-3β/B-Catenin Pathway. Cancer Genomics Proteomics 2024;21:485-501. [Crossref] [PubMed]