UTX inhibition suppresses proliferation and promotes apoptosis in patient‐derived glioblastoma stem cells by modulating periostin expression

Glioblastoma stem cells (GSCs) exert a crucial influence on glioblastoma (GBM) development, progression, resistance to therapy, and recurrence, making them an attractive target for drug discovery. UTX, a histone H3K27 demethylase, participates in regulating multiple cancer types. However, its functional role in GSCs remains insufficiently explored. This study aims to investigate the role and regulatory mechanism of UTX on GSCs. Analysis of TCGA data revealed heightened UTX expression in glioma, inversely correlating with overall survival. Inhibiting UTX suppressed GBM cell growth and induced apoptosis. Subsequently, we cultured primary GSCs from three patients, observing that UTX inhibition suppressed cell proliferation and induced apoptosis. RNA‐seq was performed to analyze the gene expression changes after silencing UTX in GSCs. The results indicated that UTX‐mediated genes were strongly correlated with GBM progression and regulatory tumor microenvironment. The transwell co‐cultured experiment showed that silencing UTX in the transwell chamber GSCs inhibited the well plate cell proliferation. Protein–protein interaction analysis revealed that periostin (POSTN) played a role in the UTX‐mediated transcriptional regulatory network. Replenishing POSTN reversed the effects of UTX inhibition on GSC proliferation and apoptosis. Our study demonstrated that UTX inhibition hindered POSTN expression by enhancing the H3K27me2/3 level, eventually resulting in inhibiting proliferation and promoting apoptosis of patient‐derived GSCs. Our findings may provide a novel and effective strategy for the treatment of GBM.

Glioblastoma (GBM) is the most common and aggressive primary malignant brain tumor, characterized by rapid progression, high metastasis, and fast relapse, leading to unfavorable treatment outcomes and a low 5-year survival rate of 3%-5% for patients (Davis, 2016;Shergalis et al., 2018).A significant challenge in treating GBM arises from its marked cellular heterogeneity, which includes a subpopulation of cancer stem cells known as glioblastoma stem cells (GSCs).GSCs have been shown to play a crucial role in tumor maintenance, recurrence, and resistance to therapy (Dirkse et al., 2019;Tang et al., 2007).Therefore, targeting GSCs has emerged as a promising therapeutic strategy for GBM.
The ubiquitously transcribed tetratricopeptide repeat on chromosome X (UTX, also called KDM6A) is a histone demethylase that catalyzes the removal of di-and tri-methyl marks at histone H3 lysine 27 (H3K27), which in turn, regulates several stem cell types, including embryonic stem cells (ESCs), induced pluripotent stem cells (iPSCs), and neural stem cells (NSCs) (Mansour et al., 2012;Shpargel et al., 2012Shpargel et al., , 2017)).Besides having a direct role in regulating normal stem cells, UTX also has a strong association with various cancer stem cells (CSCs), and significantly contributes to the development and progression of cancer (Tran et al., 2020;Van der Meulen et al., 2014).Aberrant expression of UTX has been linked to the enhancement of CSCs and the initiation of tumors in various cancers.
UTX deficiency promotes the generation of CSCs and contributes to the development of bladder cancer through cooperation with p53 haploinsufficiency (Kobatake et al., 2020).In contrast, overexpression of UTX has been identified to promote the proliferation and enrichment of colorectal CSCs (Ji et al., 2021).Although the exact role of UTX in GSCs is poorly understood at present, a recent study has shown that UTX could promote GBM growth, through interaction with HOXA3 (Yang et al., 2023).Therefore, it is reasonable to assume that UTX is involved in regulating GSCs.UTX modulates cell behavior by accurately regulating gene expression; nevertheless, its target genes in the context of GSCs remain unknown.
CSCs are localized in specialized microenvironments, referred to as niches, which are essential regulators of cell behaviors (Plaks et al., 2015).
Periostin (POSTN, also known as OSF-2), an extracellular matrix protein belonging to the fasciclin family, is secreted by various cell types within solid tumors and plays a pivotal role in mediating communication between cells and their extracellular microenvironment.By establishing and remodeling the tumor microenvironment (TME) and CSCs niche, POSTN plays a crucial role in the maintenance, metastasis, and tumorigenesis of CSCs in numerous cancers (Cui et al., 2017;Liu et al., 2019).For instance, in lung cancer, POSTN supports the survival and proliferation of CSCs, whereas its inhibition prevents metastasis (Malanchi et al., 2011).Likewise, the overexpression of POSTN promotes colorectal CSCs proliferation, and further leads to the progression and metastasis of colorectal cancer (Ma et al., 2020).In GBM, POSTN plays a key role in tumor progression, recurrence, and metastasis, with its expression level closely correlated with GBM malignancy and recurrence (Squadrito & De Palma, 2015).Besides, in GBM, overexpression of POSTN promotes adhesion and invasion of GBM cells (Mikheev et al., 2015).Instead, silencing POSTN markedly inhibits tumor growth and affects the survival of GSCs (Zhou et al., 2015).
These results suggested that POSTN might be a critical regulator of GSCs fates.Despite these findings, much is still unknown regarding how POSTN expression is controlled.
In this study, we report that UTX is highly expressed in glioma tissue and negatively correlates with survival rate.UTX inhibition increased the methylation of H3K27 at the POSTN gene locus, which results in repressed gene expression and consequent inhibition of tumor growth and tumorigenesis.These results may reveal new insight into the onset of gliomagenesis and progression, and provide a vigorous therapeutic strategy for GBM treatment.

| Human glioma transcriptome analysis
RNA-sequencing expression profiles and corresponding clinical information for GBM and low-grade gliomas (LGG) were downloaded from the Cancer Genome Atlas (TCGA) database and GETx database.

| Human glioma cell line culture
The LN229 and U251 MG cell lines were procured from American Type Culture Collection (ATCC) and Procell, respectively.3 × 10 5 cells were seeded in T25 flasks and incubated in an incubator (SANYO) with 5% CO 2 and 95% air at 37°C.The medium consisted of Dulbecco's modified Eagle's medium (DMEM), 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin (all from Gibco).Cells were trypsinized, mechanically dissociated into single-cell suspension, and cryopreserved with a liquid nitrogen tank when necessary.

| Patient-derive GSCs culture
Human GSCs were isolated from three postsurgical samples (GSC-02, GSC-05, and GSC-08).GSC-02 and GSC-05 were extracted from patients at the First Affiliated Hospital and Second Affiliated Hospital of Xi'an Jiaotong University, respectively, while GSC-08 was extracted from a patient at Shaanxi Provincial People's Hospital.
None of the patients were given any clinical treatment before their surgeries.This study was carried out following the principles of the Declaration of Helsinki.Clinical information and tumor samples were collected with informed consent.The detailed information is shown in Table S1.Single GSCs were isolated as previously described and with minor modifications (Azari et al., 2011).Briefly, small pieces from the GBM tissue were cut and washed three times with DMEM/F12 (1:1) medium (Gibco), before being incubated at 37°C for 12 min with a dissociation medium consisting of 200 μL Collagenase type I (10 mg/mL, Sigma), 200 μL Dispase II (20 mg/mL, Sigma), and 1.6 mL DMEM/F12 (1:1) medium.The tissue was then mechanically dissociated using a pipette, filtered using a 40-μm cell strainer (BD Falcon), and centrifuged at 1000 rpm for 3 min.Cells were seeded at a concentration of 4 × 10 5 cells/mL in T25 Cell culture flasks for suspension cells with 5 mL complete medium, which consisted of DMEM/F12 (1:1), 1% N2, 2% B27 (minus vitamin A), 20 ng/mL EGF, and 10 ng/mL bFGF.After culturing for 5-7 days, the formation of 90-280 μM sphere was observed.For single-cell adhesive culture, the sphere was dissociated into single cells using ACCUTASE™ (Stemcell Technologies) and plated in poly-D-lysine-coated 24-well plates.
The next day, the cells were selected using puromycin for a duration of 24 h.A concentration of 2.5 μg/mL of puromycin was utilized for the GBM cells and 1.6 μg/ml for GSCs.

| Immunostaining
Cells were fixed with 4% paraformaldehyde (PFA) at room temperature for 20 min followed by washing three times with phosphatebuffered saline (PBS).Cells were then permeabilized in 0.1% Triton X-100 for 15 min and blocked with blocking buffer (containing 5% bovine serum albumin and 5% horse serum in PBS) for 1 h.Cells were then incubated overnight at 4°C with the primary antibodies, followed by washing the cells three times with PBS and incubating them with the appropriate secondary antibodies.The information of the first and secondary antibodies is shown in Table S2.The negative control samples were incubated solely in the blocking buffer, instead of the primary antibody solution.The nuclei were visualized with a DAPI-containing mounting medium (Vector).Images were acquired using a fluorescence microscope equipped with a digital camera (BX51+DP71, Olympusn) and analyzed with ImageJ 3.5 software (NIH).The sphere-formation images were taken with a Leica SP8 confocal microscope equipped with a ×20 objective.
For fluorescence intensity measurements, the images of orthotopic glioma were acquired using a Leica SP8 confocal microscope equipped with a ×20 objective.A total of five images were captured from each sample, and the mean fluorescence intensity of the entire captured images was quantified using ImageJ 3.5 software.To ensure accuracy, image collection parameters, including neutral density filters, pinhole and detector gains, were fixed during acquisition for reliable comparisons between specimens.

| BrdU incorporation
Following the treatment, GBM cells and GSCs were subjected to BrdU treatment for 1 and 2 h, respectively, at a dosage of 10 μg/mL.
The BrdU-labeled cells were further detected by immunostaining.To identify the BrdU-labeled cells, cells were pretreated with 2 N HCl for 30 min at 37°C, followed by neutralizing with 0.1 M borate buffer (pH 8.5) for 15 min.The percentage of labeled cells was evaluated and normalized by the PI-stained nuclei or Sox2-positive cells.

| TUNEL assay
The TUNEL assay was used for detecting cell apoptosis according to the manufacturer's instructions (Roche Diagnostics).In brief, cells were fixed with 4% PFA for 30 min, followed by permeabilization using 0.1% Triton X-100 in 0.1% sodium citrate buffer for 2 min on ice.Cells were then incubated with 50 μL TUNEL reaction mixture at 37°C for 1 h.After washing with PBS, nuclei were stained with a DAPI-containing mounting medium (Vector).Images were acquired using a fluorescence microscope and analyzed with Image-Pro Plus 5.0 software (Media Cybernetics).

| Flow cytometry analysis
For cell cycle analysis, the treated cells were dissociated into single cells with trypsin and fixed with pre-cooling 75% ethanol overnight at 4°C.After washing twice with PBS, cells were stained with propidium iodide solution (100 μg/mL, Sigma-Aldrich) that contained 100 μg/mL RNase A (New England Biolabs) for 15 min at 37°C in the absence of light.The FACSCalibur system (BD Biosciences) was utilized to perform the cell cycle analysis with each sample containing 1 × 10 5 cells.The system was excited at 488 nm with 630 nm emission.The data were collected using the BD CellQuest™ software (BD Biosciences), and the DNA content and cell cycle distribution were determined using the Modfit LT software (BD Biosciences).The proliferation index was used to evaluate the changes in the cell cycle distribution with the following formula: proliferation index = (S + G2/ M)/(G0/G1 + S + G2/M).
Apoptosis analysis was performed utilizing the FITC Annexin V apoptosis detection kit (BD Biosciences).After treatment, cells were dissociated into single cells, washed twice with pre-cooled PBS, and resuspended in 1× binding buffer.Next, 200 μL of the cell suspension (greater than 1 × 10 5 cells) was transferred to a 5 mL FACS tube (BD Biosciences).Cells were then stained in duplicate for 15 min at room temperature in the absence of light with 10 μL of FITC Annexin V conjugate and 10 μL of propidium iodide (10 mg/mL).The analysis of apoptosis was carried out using a FACSCalibur (BD Biosciences) with 488 nm excitation for Annexin V (emission collected at 530 nm) and 488 nm excitation for PI (emission collected at 630 nm).FACSort Cellquect software (BD Biosciences) was used to gather the data, and the percent of apoptotic cells was calculated with the following formula: (LR + UR)/(UL + LL + LR + UR).

| Colony formation assay
Cells (500 cells/well) were seeded into a six-well plate and cultured in normal conditions for 14 days.After fixation with 4% PFA, cells were stained with a crystal violet solution.The number of colonies was counted for each sample and data were presented in samples from at least three independent experiments.

| Western blot analysis
After the treatment, cells or tissues were collected and lysed in RIPA lysis buffer supplemented with Protease Inhibitor Cocktail (Roche) for 15 min on ice, followed by sonication (Sonics) and centrifugation (Eppendorf).Then supernatants were collected, and the protein concentration of the samples was measured using the BCA assay (Pierce).Following boiling with loading buffer, proteins (20-40 μg depending on the target protein) were resolved by 10%-12% SDS-PAGE and transferred to polyvinylidene fluoride membranes (Bio-Rad).The membranes were blocked with 5% nonfat milk for 2 h at RT and then incubated overnight at 4°C with specific primary antibodies (Table S2).After washing three times with TBST, the membranes were incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG for 2 h at room temperature.The immunoreactive bands were visualized with an enhanced chemiluminescent substrate according to the manufacturer's protocol (Pierce).The bands were collected using GeneGnomeXRQ (Syngene) and analyzed using the ImageJ 3.5 software.The expression levels of target proteins were calculated and normalized to the housekeeping β-Actin.All western blot data were presented in samples from at least three independent experiments.

| Sphere-formation assay
Patient-derived GSCs were infected with lentivirus containing either shUTX (KD-UTX) or negative control vectors (shNC).An equal volume of culture medium was added to the blank control group (Ctrl).To perform sphere diameter measurement, single GSCs were seeded at a density of 3000 cells/well incubated in the 24-well plates.For sphere counting, cells were incubated at a seed density of 6000 cells/well in the 24-well plates.Sphere diameter and count were examined on the third, fifth, and seventh day, and the observations were recorded using an Olympus CKX41 inverted microscope.Inoculation in the 24-well plates was considered as 0 day.All the acquired images were analyzed using the ImageJ 3.5 software and included the spheres per well which had a diameter greater than or equal to 30 μm.

| Animal experiments
Pathogen-free male athymic BALB/c nude mice (5 weeks old) weighing 20-25 g were used for this study and all mouse experiments were approved by the Xi'an Jiaotong University Health Science Center Ethics Committee.The mice were purchased from the Xi'an Jiaotong University Laboratory Animal Center (Certificate No. 22-9601018).The staff at Xi'an Jiaotong University Laboratory Animal Center was responsible for housing and daily maintenance.Housing and environmental enrichment are according to standards.All efforts were undertaken to minimize the suffering of the mice.
Tumor growth was monitored on a daily basis, and tumor size was measured by using an electronic digital caliper.Mice were euthanized if the tumor size exceeds 2000 mm 3 or the diameter exceeds 15 mm.At 3, 7, 14, and 28 days postimplantation, five mice were killed, and their tumors were isolated, weighed, and measured using the formula: length × width 2 × 0.52 (Tomayko & Reynolds, 1989).Heterotopic tumor tissues were then collected and subjected to western blot analysis.
To establish the orthotopic glioma model, mice were anesthetized using 4% isoflurane for induction and 1.5% isoflurane for maintenance.The mice were then placed in a stereotactic frame (RWD LifeScience), and a longitudinal incision was made at the midline of the cranium to expose the bregma.After drilling a small hole, stably transfected patient-derived GSCs which express luciferase (1 × 10 5 /μL) were implanted into the right brain using the following coordinates: anterior-posterior, −1.3 mm posterior to the bregma; medial-lateral, 1.7 mm; dorsal-ventral, 2.2 mm below the skull surface.The Hamilton syringe was slowly administered over a period of 1 min at a speed of 5 µL/min.After the injection, the syringe was left in place for 5 min and the hole was sealed with bone wax.Mice were monitored daily for signs of ill health, and imaging was conducted using the IVIS ® Spectrum In Vivo Imaging System (Alameda) at Day 3, 7, and 14 postimplantation, respectively.

F I G U R E 1 (See caption on next page).
LUAN ET AL.
| 5 of 18 2.13 | Transwell co-culture analysis Patient-derived GSCs (8000 cells/well) with different treatments were seeded onto a 0.4-μm transwell chamber (Millipore).After forming a monolayer on the bottom of the inserts, untreated GSCs (3000 cells/well) were seeded in the lower compartment of the 24well plate.On the following day, the inserts were removed and cells in the plate were fixed in 4% PFA for 20 min followed by BrdU incorporation and TUNEL assay.
2.14 | RNA-sequencing (RNA-seq) Total RNA was extracted using a Trizolreagent kit (Invitrogen) according to the manufacturer's protocol.RNA quality was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies) and electrophoresis on RNase-free agarose gel.Eukaryotic mRNA was enriched by Oligo (dT) beads, while prokaryotic mRNA was enriched by removing rRNA by Ribo-Zero TM Magnetic Kit (Epicentre).Then the enriched mRNA was fragmented into short fragments using a fragmentation buffer and reverse transcripted into cDNA with random primers.Second-strand cDNA was synthesized by DNA polymerase I, RNase H, dNTP, and buffer.Then the cDNA fragments were purified with QiaQuick PCR extraction kit (Qiagen), end-repaired, poly(A) added, and ligated to Illumina sequencing adapters.The ligation products were size selected by agarose gel electrophoresis, amplified by PCR and sequenced using Illumina HiSeq.2500 by Gene Denovo Biotechnology Co. Bioinformatic analysis was performed using Omicsmart, a real-time interactive online platform for data analysis (http://www.omicsmart.com).

| Quantitative reverse-transcription PCR (qRT-PCR)
Total RNA was isolated from cells using Trizol reagent following the manufacturer's instructions, and 2 μg RNA was reverse-transcribed into cDNA using a RevertAid first-strand cDNA synthesis kit (ThermoFisher).
qRT-PCR was performed with GoTaq ® qPCR Master Mix (Promega) using an iQ5 Real-Time PCR Detection System (Bio-Rad).The primer pairs were synthesized by TaKaRa and displayed in Table S3.

| Chromatin immunoprecipitation (ChIP)-qPCR
ChIP was performed as previously described with minor modifications as detailed below (Lim et al., 2009).One hundred microliters single-cell suspensions (2 × 10 7 /mL) were crosslinked with 1% formaldehyde for 10 min and quenched with glycine (final concentration 125 mM) for 5 min.After washing twice with ice-cold PBS, the cells were lysed in 300 μL lysis buffer and sonicated for three cycles of 30 s using VCX 500 (SONICS).The lysate was diluted and incubated with Dynabeads™ Protein A (Invitrogen) pre-bound with 2 μg of antibody against H3K27me2/me3 (39435, Active Motif) or IgG (Abcam).The immune complexes were rotated at 4°C, 20 rpm on a Tube Revolver Rotator for 12 h.Chromatin-antibody complexes were washed four times with icecold RIPA buffer, rinsed with elution buffer for ChIP (containing 50 μg RNase A), and heated at 37°C, 1200 rpm on a Thermomixer (Eppendorf) for 1 h followed by incubating with 2 μL Proteinase K (New England Biolabs) at 65°C, 4 h.An input control was processed in parallel.ChIP DNA was extracted by phenol-chloroform extraction, precipitated with ethanol, and resuspended in 20 μL EB buffer (Qiagen).Analysis of DNA was performed using qRT-PCR with gene-specific primers (Table S4).
The percentage of input DNA was calculated as a percentage of the immunoprecipitated DNA relative to input DNA.

| The enzyme-linked immunosorbent assay (ELISA)
At the end of each time point, the cell culture medium was collected and analyzed using the POSTN, VCAM1, and Pro-Collagen I alpha 1 F I G U R E 1 UTX inhibition influences the proliferation and apoptosis of GBM cells.(a) The expression distribution of the UTX gene was evaluated among normal tissues (n = 1136), LGG (n = 509), and GBM (n = 153); ***p < 0.001 versus the normal group.(b) Kaplan-Meier survival analysis of the glioma patients with lower and higher levels of UTX from the TCGA data set and GETx database.LN229 cells were divided into three groups.The Blank group: cells were maintained without any treatment; the shNC group: cells infected with lentiviruses carrying negative control (shNC); the KD-UTX group: cells infected with lentiviruses carrying UTX-target shRNA (KD-UTX).(c, g) After culturing 3 days, BrdUpositive cells were determined by immunostaining, and the result was shown as percentages among PI-stained cells.Scale bar = 100 μm.Data are presented as the mean ± standard deviation of nine independent experiments (n = 9).**p < 0.01 versus the shNC group.(d, h) A colony formation assay was performed to assess cell colony-forming ability.Data are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05 versus the shNC group.After culturing 3 days, TUNEL staining (e) and flow cytometry analysis (f) were performed to evaluate cell apoptosis.Scale bar = 50 μm.(i) Quantitative data from nine independent experiments (n = 9) are presented as the percentage of TUNEL-positive cells in the total DAPI cells.**p < 0.01 versus the shNC group.(j) Data are presented as the mean ± standard deviation of three independent experiments (n = 3).***p < 0.001 versus the shNC group.(k) The UTX knockdown stable LN229 cells (1 × 10 6 per injection) were implanted subcutaneously into the nude mice to develop the xenograft model.The tumor volumes (l) and weights (m), and western blot analysis (n) of the tumors were determined at 3, 7, 14, and 28 days (the third day after cell inoculation was deemed as 0 day).Data are presented as the mean ± standard deviation of five independent experiments (n = 5).*p < 0.05, **p < 0.01 versus the shNC group.The ratio of Cleaved-caspase-3 to Pro-caspase-3 (o) and Cyclin D1 to β-Actin (p) were used to plot the band intensity.Data are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01 versus the shNC group.GBM, glioblastoma; LGG, Low-grade glioma.ELISA kit (R&D Systems) following the manufacturer's instructions.
The absorbance was measured at 450 nm using a multimicroplate spectrophotometer (BioTek).Triplicate parallel wells were examined in all the experiments and the data were presented as the average of at least three independent experiments.
Cell viability was assessed at the end of each treatment using the Cell Counting Kit-8 (CCK-8; 7sea).To each well, 20 μL/well of CCK-8 and then incubated for 2 h at 37°C.The absorbance was measured at 490 nm using a multimicroplate spectrophotometer (BioTek).Triplicate parallel wells were examined in all the experiments, and the data were collected as the average of at least three independent experiments.The results are presented as the absorbance value.

| Statistical analysis
Statistical analyses were performed using the GraphPad Prism 5.0 software (San Diego).Data were presented as mean ± standard deviation.Data were evaluated for normality and homogeneity of variance before comparison.One-way ANOVA was used for F I G U R E 3 UTX is robustly associated with the regulation of proliferation and extracellular matrix proteins expression.After culturing 3 days, whole-transcriptome sequencing (RNA-seq) was performed on patient-derived GSCs that had been stably transfected with UTX-target shRNA (KD-UTX) and cells that were transfected with scramble shRNA (shNC).(a) Heatmap of the interindividual correlation of all mRNA transcripts (RNA-seq).(b) Volcano plots showed global differences between the shNC and KD-UTX group.The RNA-seq results were further analyzed using gene-set enrichment analysis (GSEA) (c-h), Gene Ontology (GO) enrichment analysis (i), Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathway analysis (j), and Disease Ontology (DO) enrichment analysis (k).

F I G U R E 4 (See caption on next page).
LUAN ET AL.
| 9 of 18 significance testing.Comparisons between two groups were conducted by Fisher's PLSD test.The Kolmogorov-Smirnov test was used for normality and homogeneity.p < 0.05 was considered statistical significance.

| UTX inhibition contributes to GBM repression processes
To investigate the potential effect of UTX on GBM progression, we first evaluated the expression levels of UTX in normal brain tissues, LGG and GBM by analyzing RNA-seq data from TCGA and GETx database.UTX expression was significantly higher in GBM and LGG compared to normal brain tissues (Figure 1a).Moreover, the expression of UTX was negatively correlated with overall survival for glioma (including both LGG and GBM) patients, indicating that UTX plays a significant role in glioma progression (Figure 1b).To further elucidate the role of UTX in GBM, we used RNA interference (RNAi) to suppress UTX expression in two human GBM cell lines (LN229 and U251 MG).Three UTX-target siRNAs (si-UTX-1/2/3) were transfected into cells, and si-UTX-1 exhibited the most effective inhibition based on RT-PCR and western blot results (Figure S1A-E).Therefore, the si-UTX-1 sequence was used in the subsequent lentiviral packaging.Cells were transduced with the lentivirus for UTX inhibition (KD-UTX) and selected using puromycin.
The majority of cells expressed eGFP, and the expression of UTX was significantly inhibited (Figure S1F-J).This lentivirus for UTX inhibition (KD-UTX) was used for further experiments.
We then assessed the impact of UTX knockdown on GBM cell proliferation and apoptosis using BrdU incorporation, cell cycle analysis, colony-forming assay, and TUNEL assays.

| UTX inhibition reduced proliferation and increased apoptosis of patient-derived GSCs
GSCs are mainly responsible for tumor maintenance, recurrence, and therapeutic resistance.Due to the tumor's high heterogeneity, GSCs are likely to contribute to the bioinformatics analysis of GBM data from TCGA.To investigate the effect of UTX on GSCs, we first cultured patient-derived GSCs (GSC-02, GSC-05, and GSC-08) which were isolated from three postsurgical patients (Figure 2a).These GSCs formed spheres ranging from 90 to 280 μM after being cultured for 5-7 days, expressing CD133, CD15, CD44, and nestin (Figure 2b,d,e and Figure S3A,C,D).Additionally, single-GSC staining demonstrated that 96.89% ± 5.37% of cells were nestin-positive, with 96.65% ± 4.28% expressing SOX2, confirming their identity as human GSCs (Figure 2c and Figure S3B).
To measure the effect of UTX on patient-derived GSCs, we used lentivirus with shRNA to suppress UTX expression in patient-derived GSCs and confirmed knockdown efficacy through western blot analysis (Figure S4A,B).Sphere-formation and CCK-8 assay showed that UTX inhibition led to reductions in both the number and size of tumorspheres and cell viability (Figure 2f-h and Figure S3E-J F I G U R E 4 UTX inhibition suppresses the expression of extracellular matrix-related genes via enhancing the levels of H3K27 methylation.To harvest total protein, cells were cultured for 3 days and lysed.(a, b) The level of H3K27me2/3 and histone H3 were detected using western blot.Data from three independent experiments (n = 3) were presented as the ratio of H3K27me2/3 to H3.The values are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01 versus shNC group.The top 20 significantly downregulated genes in knockdown UTX GSCs compared to shNC GSCs are presented in the heat map (c), and these genes were validated by qRT-PC (d).(e) ChIP-qPCR was used to measure the H3K27me2/3 enrichment in the significantly downregulated genes.The values are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01, ***p < 0.001 versus shNC group.The intracellular protein expressions of COL1A1 (f), VCAM1 (h), and POSTN (j) were examined by western blot.The band intensity was quantified and plotted as a ratio of the target protein to β-Actin.(g, i, k) The values are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01, ***p < 0.001 versus shNC group.ELISA was used to detect the concentrations of COL1A1 (l), VCAM1 (m), and POSTN (n) in the culture medium.The values are presented as the mean ± standard deviation of nine independent experiments (n = 9).*p < 0.05, **p < 0.01, ***p < 0.001 versus shNC group.

F I G U R E 5 (See caption on next page).
LUAN ET AL.
| 11 of 18     suggested that UTX inhibition suppressed proliferation and tumor progression while promoting apoptosis in patient-derived GSCs.

| UTX-regulated gene is closely related to cell proliferation and extracellular microenvironment
To gain insight into the molecular mechanisms by which UTX exerts its effects, total GSCs-02 cell RNA was extracted and subjected it to differential RNA-sequencing after UTX inhibition treatment.The resulting data showed a high level of reproducibility, as indicated by the correlation heatmap (Figure 3a), and revealed that the expression of 568 genes was significantly altered by UTX withdrawal.Of these genes, 446 showed increased expression while 122 showed decreased expression (Figure 3b).Results for geneset enrichment analysis (GSEA) demonstrated that the UTX-regulated genes were strongly associated with DNA replication, cell cycle regulation, and extracellular matrix organization (Figure 3c-h).In addition, Gene Ontology (GO) enrichment (Figure 3i), KEGG databases (Figure 3j), and Disease Ontology (DO) analysis (Figure 3k) depicted the involvement of UTX-regulated genes in cancer, secretion, intercellular communication, and extracellular structure organization.
Taken together, these findings suggest that UTX plays a crucial role in GBM progression and the regulation of the TME.

| UTX inhibition suppresses many extracellular matrix proteins expression
UTX is associated with the demethylation of lysine residues on H3K27.
To verify that UTX regulates the level of H3K27me2/3 in GBM cells, western blot analysis was performed.As anticipated, the KD-UTX group exhibited a significant increase in the level of H3K27me2/3 (Figure 4a,b).
Because H3K27me2/3 is strongly associated with gene repression, the 122 downregulated genes might be directly regulated by UTX inhibition.
The top 20 downregulated genes were selected for validation through qRT-PCR (Figure 4c).Further, the H3K27me2/3 modification in these genes was examined via ChIP-qPCR, resulting in the identification of 17 significantly downregulated genes, 13 of which showed declined levels of H3K27me2/3 (Figure 4d,e).Notably, these downregulated genes included several major components of the extracellular matrix, such as POSTN, VCAM1, SCG5, COL1A1, and ITGBL1.To define the effect of UTX, the expression of extracellular matrix protein and intracellular protein in cultured GSCs were detected by western blot and ELISA, respectively.UTX inhibition significantly suppressed the expressions of intracellular proteins, including COL1A1, VCAM1, and POSTN (Figure 4f-k), and reduced their concentrations in the culture medium (Figure 4l-n).To further specify the effect of UTX on extracellular matrix proteins, immunostaining and western blot analysis were conducted to detect the expression of COL1A1, VCAM1, and POSTN in orthotopic glioma.Similar to the results obtained from cultured GSCs, UTX knockdown reduced the expression of these extracellular matrix proteins in orthotopic glioma (Figure 5a-f).Taken together, these findings suggested that UTX was involved in the regulation of the TME.

| UTX regulates the proliferation and apoptosis in patient-derived GSCs and GBM cells by changing the TME
Based on the results obtained, it is speculated that UTX could be regulating the cellular behavior through alteration of the TME.To mimic the alterations occurring in the TME, we conducted transwell co-culture experiments to observe variations in cell proliferation and apoptosis.The transwell chambers were seeded separately with normal, shNC, or UTX-inhibited GSCs.After completely filling the bottom of the transwell chamber with cells, the normal GSCs were seeded in the well plate (Figure 5g).One day later, the proliferation and apoptosis were evaluated for changes in the number of BrdU and TUNEL-positive cells in the well plate, respectively.The BrdU incorporation assay showed a significant reduction in BrdU-positive GSCs (Figure 5h,i) in the well plate, due to the silencing of UTX in the transwell chamber GSCs.However, TUNEL-positive cells in the well plate were not observably altered by UTX inhibition (Figure 5j,k).These phenomena suggested that the altered proliferation of GSC and GBM cells caused by UTX inhibition was mainly due to changes in the TME.

| UTX influences the proliferation and apoptosis of patient-derived GSCs by regulating POSTN expression
Although the above results demonstrated a significant association between UTX and the regulation of cellular processes such as cell

| DISCUSSION
In this study, we showed that UTX expression was much higher in the glioma, and negatively correlated with overall survival.Under normal conditions, patient-derived GSCs display high proliferative capability and a low rate of apoptosis due to adequate UTX expression.
Inhibition of UTX expression promotes H3K27 methylation in POSTN, resulting in the suppression of POSTN expression.The low expression level of POSTN inhibits the expression of several extracellular matrix proteins (including COL1A1, COL3A1, and VCAM1) that have an impact on the TME, leading to inhibits proliferation and promotes apoptosis (Figure 7f).Our results indicate that UTX promotes growth and tumorigenesis, and UTX inhibitors may be a potential therapeutic target for GBM drug development.
However, targeting UTX for cancer treatment remains controversial.
The significant difference in UTX expression across various cancer types may partially explain the variation in UTX's effects (Figure S5C).
Another reason may be that UTX acts as a subunit of MLL3 and MLL4, both of which belong to the COMPASS family of histone H3 lysine 4 (H3K4) methyltransferases (Sze & Shilatifard, 2016).H3K4 methylation has the opposite effect on gene expression compared to UTX (Zaidi et al., 2013), and it is tightly associated with transcriptional start sites of actively transcribed genes (Shilatifard, 2012).The biological role of UTX and MLL3/MLL4 in cancer pathogenesis is complex.The mechanistic relationship between UTX and MLL3/ MLL4 and their regulation of enhancer activity, whether local or global is still unclear.Our results showed that silencing UTX in GSCs affected the expression of only 568 genes, suggesting that UTX and MLL3/MLL4 primarily impact local genomic features.Further studies are required to fully comprehend the fundamental mechanism and precise functional impact of MLL3/MLL4 and UTX alterations.
The TME is a dynamic network structure that has recently been recognized as a key factor in regulating tumor occurrence, development, invasion, and recurrence (Ribeiro Franco et al., 2020;Wang et al., 2017).Anticancer research must, therefore, consider not only the features of cancer cells but also the effect of the TME (Panigrahi et al., 2020).In this study, RNA-seq, ELISA and western blot assay revealed that UTX inhibition selectively modulated the expression of genes related to extracellular structure organization and secretions.Specifically, we observed altered expression of POSTN, COL1A1, and VCAM1, which have all been critically linked to tumor progression and development (Kong et al., 2018;Li et al., 2022;Yamauchi et al., 2018).
Notably, the transwell co-cultured experiment demonstrated that silencing UTX inhibited cell proliferation primarily by altering the TME, F I G U R E 6 POSTN replenishment abolishes the effect of UTX inhibition on cell proliferation and apoptosis of patient-derived GSCs.UTXregulated proteins were screened for protein interactions (gray lines, the confidence score ≥0.4) by using the STRING database (a), and the interaction score between POSTN and other proteins was shown (b).Three patient-derived GSCs (GSC-02, GSC-05, and GSC-08) were divided into four groups: Blank group: cells were maintained without any treatment; shNC group: cells infected with lentiviruses carrying negative control (shNC); KD-UTX group: cells infected with lentiviruses carrying UTX-target shRNA (KD-UTX); KD-UTX + POSTN group: cells infected with KD-UTX followed by POSTN (0.2 μg/mL) treatment.After culturing 3 days, cell proliferation was identified by BrdU incorporation (c-e) and apoptotic cells were detected by TUNEL staining (f).(g) Quantitative data from nine independent experiments (n = 9) were shown as the percentage of BrdU-positive cells in total SOX2-positive cells.(h) Quantitative data from nine independent experiments (n = 9) are presented as the percentage of TUNEL-positive cells in the total DAPI cells.*p < 0.05, **p < 0.01 versus shNC group; # p < 0.05, ## p < 0.01 versus KD-UTX group.Scale bar, 50 μm (c-e), 100 μm (f).| 15 of 18 while no significant effect on apoptosis was observed.One potential explanation for this phenomenon is the presence of normal GSCs or GBM cells in the culture system, which together constitute the TME.
Although UTX inhibition reduced the expression and secretion of extracellular matrix proteins from GSCs in the transwell chamber, normal GSCs or GBM cells in the well plate may have partially compensated for the extracellular matrix protein deficiency, thereby offsetting the proapoptotic effect of UTX inhibition.Furthermore, GSEA analysis results showed that knockdown of UTX affected the expression of numerous extracellular matrix (ECM)-related genes, with the majority being activated.This finding suggests that the effects of UTX on gene expression are not isolated, as genes directly regulated by UTX can also affect downstream genes.For example, RNA-seq analysis revealed that knockdown of UTX affected the expression of genes in the COL1, ADAMTS, and ITGB families.The binding of COL1 to integrin receptors on the cell surface can regulate the expression of both ADAMTS and ITGB family genes (Chen et al., 2013;Leitinger, 2011;Lu et al., 2012).Based on the above discussion, we can infer that UTX-regulated genes form an intricate biochemical network that ultimately leads to various biological effects, including proliferation and apoptosis.
POSTN is one of the matricellular proteins mainly secreted by stromal cells in normal tissues.However, solid tumor cells, particularly cancer stem cells niche, can also secret POSTN in solid tumors (Liu et al., 2019).POSTN plays a crucial role in regulating tumor progression and tumorigenesis by remodeling various tumor microenvironments, such as CSCs niche, perivascular niche, and immunosuppressive microenvironment (Ghajar et al., 2013;González-González & Alonso, 2018;Masuoka et al., 2012;Zhou et al., 2015).Moreover, the K-M curve for POSTN displays a similar pattern to that of UTX.
POSTN expression showed a negative correlation with overall survival in both glioma patients and GBM patients (Figure S5A,B).In this study, PPI analysis revealed that POSTN protein interacts with many extracellular matrix proteins regulated by UTX and that POSTN replenishment could eliminate the effect of UTX inhibition on the proliferation and apoptosis in GSCs, restoring the previous level of COL1A1 and VCAM1.These results suggest that POSTN is a critical regulator of the antitumor effect of UTX inhibition.Furthermore, compared to the GSCs sorted from cell lines, patient-derived GSCs more closely recapitulate individual differences and provide a more realistic response (Nagle et al., 2018).In this study, we used patientderived GSCs to examine UTX's functions, which strengthens the trustworthiness of our findings.However, the effect of UTX on the inhibition of POSTN expression does not entirely account for its antitumor mechanism.UTX regulates many other tumor-associated proteins, such as COL1A1, ENC1, VCAM1, and ITGBL1, and their roles in mediating the antitumor mechanism remain to be discovered.
The specific mutations present in cancer cells significantly influence their abilities in terms of proliferation, migration, invasion, and metastatic potential (Jiang et al., 2015).Consequently, it is essential to take into account these variations when working with various GBM cell lines.In our current study, we have demonstrated that inhibiting UTX significantly reduces cell proliferation and enhances apoptosis in both the LN229 and U87 MG cell lines.
Significant disparities in gene mutations exist between the LN229 and U87 MG cell lines.LN229 cells exhibit a mutated p53 (TP53) genotype, along with potential homozygous deletions in the p16 and p14ARF tumor suppressor genes, while retaining a wild-type PTEN gene.In contrast, U87 cells possess a wild-type TP53 gene but harbor a mutated PTEN gene with homozygous deletions in both the p16 and p14ARF tumor suppressor genes (Demircan et al., 2021).These results strongly indicate the potential of UTX as a promising drug target for various subtypes of GBM.
The tumorigenesis and progression of GBM is a complex process involving both genetic and epigenetic alterations (Thompson et al., 2018).
As alterations of histone methylation are reversible, drugs targeting these processes have broad clinical applications for treating GBM (Bennett & Licht, 2018).In this study, we demonstrated that UTX inhibition could enhance the H3K27me2/3 level in the POSTN gene, leading to the suppression of many extracellular matrix proteins, including POSTN, and ultimately altering the proliferation and apoptosis of patient-derived GSCs.These results highlight UTX's role in the regulatory mechanism of tumorigenesis and development, and suggest a novel and effective strategy for the treatment of GBM.However, additional research, particularly into resistance, is required for an indepth understanding of the precise mechanisms of UTX in GSCs.

AUTHOR CONTRIBUTIONS
Luan Yan, Hanyue Zhang, and Yingfei Liu: Investigation and methodology.Jiangwen Xue, Ke Wang, and Bo Ma: Investigation.Kaige Ma: F I G U R E 7 POSTN replenishment eliminates the effect of UTX inhibition on the expression of COL1A1 and VCAM1.Three patient-derived GSCs (GSC-02, GSC-05, and GSC-08) were treated as above for 3 days.(a, b) ELISA assay was carried out to measure extracellular concentrations of COL1A1 and VCAM1.The values are presented as the mean ± standard deviation of nine independent experiments (n = 9).*p < 0.05, **p < 0.01, ***p < 0.001 versus shNC group; # p < 0.05 versus KD-UTX group.(c-e) The band intensity was quantified and expressed as a ratio of the target protein to β-Actin, with the values presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01 versus shNC group; # p < 0.05 versus KD-UTX group.(f) Illustration depicts the mechanism by which UTX inhibition regulates the proliferation and apoptosis of patient-derived GSCs.The expression level of UTX is significantly increased in GSCs (up panel), leading to low H3K27me2/3 levels in the POSTN gene, which ultimately activates transcription.The POSTN protein interacts with many secreted proteins, vital components of the tumor microenvironment, that regulate the proliferation and apoptosis of GSCs.Conversely, UTX inhibition (down panel) significantly increases the H3K27me2/3 level in the POSTN gene, resulting in the repression of POSTN transcription.This, in turn, leads to the secretion of fewer proteins, especially those that interact with POSTN, to the extracellular matrix and the alteration of the tumor microenvironment, thereby inhibiting proliferation and promoting apoptosis.

F
I G U R E 2 UTX inhibition suppresses cell proliferation and promotes apoptosis of patient-derived GSCs.(a) Flow diagram of isolation and culture of patient-derived GSCs.(b) Human GSCs were isolated from three postsurgical patients, and tumorspheres with a diameter of 90-280 μm were observed after culturing for 5-7 days.Scale bar = 200 μm.(C) SOX2 and nestin double-positive cells are detected in the single GSCs.Scale bar = 50 μm.The right panel represents magnified pictures of square frames in the corresponding left panel.The spheres were identified by double immunofluorescent labeling for specific markers CD133 and CD15 (d) or CD44 and nestin (e).The patient-derived GSCs were divided into three groups.Blank group: cells were maintained without any treatment; shNC group: cells infected with lentiviruses carrying negative control (shNC); KD-UTX group: cells infected with lentiviruses carrying UTX-target shRNA (KD-UTX).After treatment with the lentivirus, the sphere number (f), the mean diameter (g), and cell viability (h) were measured at different time points.Data are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01, ***p < 0.001 versus shNC group.The BrdU incorporation was used to assess cell proliferation (i), and TUNEL staining was used to detect apoptotic cells (k).(j, l) Data are presented as the mean ± standard deviation of nine independent experiments (n = 9).*p < 0.05, **p < 0.01 versus shNC group.Scale bar, 50 μm (I), 100 μm (k).(m) The luciferase-expressing shNC GSCs or KD-UTX GSCs were used to establish the orthotopic glioma mouse model.The in vivo imaging of shNC GSCs and KD-UTX GSCs in tumor-bearing mice at different time points after injection.(n) Data are presented as the mean ± standard deviation of five independent experiments (n = 5).*p < 0.05 versus the shNC group.
Figure S2I,K) and the rate of apoptosis after UTX silencing (Figure 1f,j and Figure S2J,L).To further validate these findings, the UTX knockdown stable LN229 cells were implanted into the subcutis of athymic nude mice to establish the heterotopic xenograft mode.We ). Subsequent BrdU incorporation and TUNEL staining showed that silencing UTX decreased the number of BrdU-positive cells (Figure 2i,j and Figure S3K-M), while increasing the number of TUNEL-positive cells (Figure 2k,l and Figure S3N-P).To validate these results, we carried out in vivo analyses using a patient-derived GSCs orthotopic glioma mouse model.Following the 3R principles, we just used GSC-02 (the best condition among the three patientderived GSCs) for the in vivo study.KD-UTX GSCs or shNC GSCs were implanted into the nude mice, and small animal imaging was used to detect fluorescence intensity in tumor tissues at different time points.Intravital imaging revealed suppression of tumor growth following UTX inhibition (Figure 2m,n).Collectively, these results

F
I G U R E 5 UTX affects cell proliferation by the tumor microenvironment.Immunofluorescent staining of COL1A1 (a), VCAM1 (b), and POSTN (c) in orthotopic glioma were shown.An enlarged image of the square frame is shown on the right-hand side.Scale bar = 200 μm.(d) The values of mean fluorescence intensity are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01 versus shNC group.(e, f) The expressions of COL1A1, VCAM1, and POSTN were examined by western blot.The intensity of each band was quantified and plotted as a ratio of the target protein to β-Actin.The values are presented as the mean ± standard deviation of three independent experiments (n = 3).*p < 0.05, **p < 0.01, ***p < 0.001 versus shNC group.(h) Diagram of the transwell coculture experiment.After transfection, GSCs with different treatments were seeded onto a transwell chamber.When the cells covered the bottom of the transwell chamber, normal GSCs were seeded in the 24-well plate.The proliferation of GSCs in the 24-well plate was measured by BrdU incorporation the day after seeding (h).Scale bar = 100 μm.The cell apoptosis in the 24-well plate was evaluated by TUNEL staining (j).Scale bar = 50 μm.(i, k) Data are presented as the mean ± standard deviation of nine independent experiments (n = 9).*p < 0.05 versus the shNC group.

F
I G U R E 6 (See caption on next page).
LUAN ET AL.|13 of 18      proliferation and apoptosis by altering the TME, it remains unclear what specific gene UTX directly regulates.To identify the master regulator of UTX, proteins that are regulated by H3K27me2/3 were analyzed using the STRING database to reveal the protein-protein interaction (PPI) networks and identify the hub proteins within the network.The result showed that the POSTN protein could interact with many extracellular matrix proteins (Figure6a,b).Furthermore, silencing UTX resulted in the inhibition of POSTN protein expression (FigureS4C,D) and mRNA (FigureS4E) in cultured patient-derived GSCs.Interestingly, there was an increase in the H3K27me2/3 levels in the POSTN gene (FigureS4F).Therefore, it is reasonable to conclude that POSTN plays a pivotal role in the UTX-mediated transcriptional regulatory network.To determine the involvement of POSTN in UTX-mediated regulation of cell proliferation and apoptosis, patient-derived GSCs' proliferation and apoptosis were measured through BrdU incorporation and TUNEL staining, respectively, after replenishing the KD-UTX cell culture medium with 0.2 μg/mL of recombinant human POSTN protein.Similar to our previous results, UTX inhibition significantly decreased the number of SOX2 BrdU double-positive cells.Interestingly, supplementing the KD-UTX cell culture medium with POSTN protein abolished the effect of UTX inhibition on cell proliferation (Figure 6c-e,g).Additionally, TUNEL staining revealed a heightened presence of positively stained cells within the KD-UTX group.Notably, the supplement of POSTN protein effectively alleviated the proapoptotic effect of UTX inhibition (Figure 6f,h).Moreover, we examined COL1A1 and VCAM1 expression changes since they both interact with POSTN.The ELISA assay showed that UTX inhibition decreased the extracellular concentrations of COL1A1 and VCAM1, and this difference was considerably reduced in the POSTN replenishment group (Figure 7a,b).Similar to the ELISA results, western blot analysis showed that the intracellular concentrations of both COL1A1 and VCAM1 decreased after inhibiting UTX, but this effect was blocked upon supplementing POSTN (Figure 7c-e).The above results indicate that POSTN plays a crucial role in the UTXmediated proliferation and apoptosis of patient-derived GSCs.

F
I G U R E 7 (See caption on next page).