Identification of a novel disulfideptosis-related gene signature for prognostic implication in lower-grade gliomas

Data collection and procession

As shown in Figure 1, 696 GBMLGG patients’ mRNA-sequencing data and associated clinical information were obtained from the Cancer Genome Atlas (TCGA) data set. (https://www.cancer.gov/about-nci/association/ccg/research/underlyinggenomics/tcga, (which was accessed on 22 April 2023)). The specific information about TCGA-GBMLGG is presented in Table 1. What’s more, the microarray information profiles of GSE15824, GSE34152, and GSE50161 in view of a similar stage GPL570 and comparing clinical data were gotten from the Gene Expression Omnibus (GEO) data set (https://www.ncbi.nlm.nih.gov/geo/, (which was accessed on 22 April 2023)). After that, batch effects were removed and the two datasets were combined using the “sva” R package.

Figure 1. Flow chart of this study.

Development and validation of a prognostic disulfidptosis-related gene signature

After downloading the GEO query package from the GEO database, the probes that corresponded to different compounds were eliminated. Only the probes with the highest signal value were kept when multiple probes relating to the same molecule were found. Then the clustering among sample groups was checked through PCA maps, and the difference analysis between the two groups was carried out by the limma package. In the TCGA cohort, we employed Kaplan-Meier analysis to examine the variations in OS between the high-risk and low-risk groups. A significance threshold was set at p < 0.05. The expression of genes linked to disulfidptosis was compared in normal and tumor samples from the GEO cohort using the “limma” R program. When the p-value was less than 0.05, genes were deemed to be significantly different. We use the KEGG rest API (https://www.kegg.jp/kegg/rest/keggapi.html) to retrieve the most recent KEGG Pathway gene annotation for the purpose of enriching gene set function analysis, as background, the difference of gene mapping to the background in the collection, the R software package clusterProfiler (version 3.14.3) was used to conduct enrichment analysis and produce the gene set enrichment results. Put five as the minimum gene and five thousand as the maximum gene, a P-value of less than 0.05 was deemed statistically noteworthy. For the functional enrichment analysis of gene set, we used the GO annotation of genes in the R software package org.Hs.eg.db (version 3.1.0) as the background to map the differential genes into the background set. The R software package clusterProfiler (version 3.14.3) was used to conduct enrichment analysis and produce the gene set enrichment results. Put five as the minimum gene and five thousand as the maximum gene, a P-value of less than 0.05 was deemed statistically noteworthy. For Gene set enrichment analysis (GSEA), we separated the sample into two groups based on glioma and normal tissue, using the GSEA software (version 3.0) from GSEA (https://doi.org/10.1073/pnas. 0506580102, http://software.broadinstitute.org/gsea/index.jsp), and derived Molecular Signatures Database (https://doi.org/10.1093/bioinformatics/btr260, http://www.gsea-msigdb.org/gsea/downloads.jsp) to download the c2. Cp. Kegg. V7.4. Symbols. The GMT Set, to evaluate the relevant pathways and molecular mechanisms, based on gene expression profiles and phenotypic grouping, put five as the minimum gene and five thousand as the maximum gene, 1000 resampling, a P-value of less than 0.05 was deemed statistically noteworthy. Utilizing the TCGA and GEO datasets, we distinguished the convergence of qualities with genuinely tremendous contrasts. In addition, a multivariate COX analysis of OS revealed that the aforementioned genes had prognostic values (p-values below 0.05).

Clinical features and establishment of a prognostic nomogram

We used the “rms” R package to make a nomogram for predicting OS based on the clinical information and gene expression profiles of GBMLGG patients in the TCGA cohort. Time-subordinate alignment bends and AUC bends were attracted to check the legitimacy of the nomogram. We further performed both univariate and multivariate Cox regression analyses to investigate whether this prognostic model could on its own predict OS in GBMLGG patients.

Protein expression analysis

We reviewed cancer Omics data using the University of Alabama Cancer Database (UALCAN) portal [24]. Using the UALCAN portal, protein expression was also looked at in the dataset from the Clinical proteomic tumor analysis consortium (CPTAC). Our examination group assessed the degrees of HNRNPC (NP 001070910.1) all out protein and phosphoprotein articulation in essential cancer and typical tissues subsequent to entering the expression “HNRNPC” into the pursuit box.

DNA methylation

TCGA-GBMLGG in the Illumina human methylation 450 methylation data and the project level 3 HTSeq - RNAseq FPKM format data. RNAseq information in FPKM design was changed over into TPM design and log2 changed. In this review, the ggplot2 bundle was mostly utilized for perception.

Patient samples and immunohistochemistry (IHC)

3 paired paraffin-embedded GBMLGG and counterpart normal adjacent tissues were collected from the second hospital of Dalian Medical University. Using IHC, the protein expression levels of IQGAP1 were found in 4-μm tissue slices. Sections were rehydrated in varying alcohols after being deparaffinized in xylene. Sections underwent a 20-minute, 95°C pretreatment in citrate buffer (0.01 mol/L citric acid, pH 6.0). They were then submerged in PBS containing 3% H₂O₂ for ten minutes at room temperature. Following treatment with 10% normal goat serum in PBS for 30 minutes at room temperature, the tissue slices were incubated with rabbit polyclonal anti-IQGAP1 antibodies (1:1000 dilution) (ab86064, Abcam) for an entire night at 4°C. Following a PBS washing, the sections were treated with 3,30-diaminobenzidine chromogen for five minutes at room temperature and incubated for twenty minutes with biotinylated goat anti-rabbit IgG. Sections were then counterstained for six minutes using hematoxylin. The second hospital of Dalian Medical University’s Ethics Committee accepted the portion of the study that involved experimentation on human tissues. We acquired informed consent from each and every participant.

Cell culture and transfections

The American Type Culture Collection (ATCC) provided the human glioma cell lines U87, which were then cultured in DMEM media (Gibco, USA) supplemented with 10% fetal bovine serum (Hyclone, USA), Penicillin-Streptomycin (100 U/ml, Hyclone), and glutamine (2 mM, Hyclone) at 37°C in a humidified environment with 5% CO₂. After being separated by 0.25% trypsin and 0.02% EDTA solution, the cells were subcultured once every two to three days.

The shRNA targeting IQCAP1 (sh-IQGAP1: 5′-CAACGACATTGCCAGGGATAT-3′) and the control non-silencing RNA (PLV-Ctrl: 5′-GGAATCTCATTCGATGCATAC-3′) were presented by Dong et al. [25]. The shRNA-containing lentiviruses were packaged by co-transfecting the pLVTHM-GFP vector together with psPAX2.0 and pMD2.G into 293T cells. U87 cell suspensions of 5 × 104 cells were cultured in 6-well plates for 24 h at 30% to 40% confluence, and then infected with lentiviruses at 10 multiplicity of infection (MOI), which were estimated by a range of 5, 7.5, 10, 12.5, 15 and 20. After 8 hours, the medium was switched, and if more than 80% of the cells were GFP-fluorescent when seen under an Olympus IX71 fluorescent microscope 72 hours after infection, the cells were screened with puromycin (2 μg/ml). Quantitative real-time PCR analysis verified the transfection effectiveness of IQGAP1.

Flow cytometry analysis of the cell cycle

Cells were reaped and cultivated in 6-well plates. Cells were digested and collected after 24 hours of incubation, then centrifuged for 5 minutes at 4°C at 1000 rpm. After being washed and suspended in ice-cold PBS, the deposited cells were then resuspended in 70% ethanol for 30 minutes at 4°C. After that, the cells were cleaned and suspended once more for 30 minutes at 4°C in 100 mL of PBS containing the final concentration of 50 mg/mL of RNase A (Sigma, US) and 0.25% Triton X-100. Afterwards, the cells were stained with 10 mg/mL propidium iodide (PI) (Life Technologies, USA) for thirty minutes. After that, a FACScan flow cytometer (Becton Dickinson, USA) was used to immediately analyze the cells. The examples were all examined multiple times, and the small part of every cell cycle stage was estimated.

Cell apoptosis analysis

Six-well plates were used to seed the cells at a density of 1 × 10⁵ cells per well. There was 2 ml of culture medium in each well. After being incubated for 24 hours, the cells were given 20 milligrams of cisplatin, rinsed with PBS, and then resuspended in 300 milliliters of binding buffer. Subsequently, the cell suspensions were incubated for 15 minutes in complete darkness with 5 μl of PI added. The next step involved using a BD FACSVerse stream cytometer to identify cell apoptosis.

Quantitative real-time PCR

As directed by the manufacturer, RNA was extracted from U87 cells using the RNA isoPlus^® Reagent Kit (Takara Biotechnology, Japan). To convert RNA into cDNA, the PrimeScript^® RT Reagent Pack (Takara, Shiga, Japan) was used. Using the SYBR^® Premix Ex Taq™ Unit and the 7500 Ongoing PCR Framework (Applied Biosystems, 7500 Continuous PCR Framework, Thermo, USA), the cDNA was enhanced. The conditions for cycling were as follows: The results were analyzed using the comparative Ct technique, with GAPDH acting as the loading control for the target genes during the forty cycles of 30 s at 95°C and 34 s at 60°C. The preliminaries were as per the following: IQGAP1 (Forward: 5′-ACCGTGGACCCAAAGAAC-3′, turn around: 5′-CTTCCCGTAGAACTTTTTGTTG-3′; GAPDH (Following: 5′-GCACCGTCAAGGCTGAGAAC-3′, switch: 5′-TGGTGAAGACGCCAGTGGA-3′).

Cell proliferation assay

96 well plates were utilized to seed cells (2 × 10⁵ cells/well). The cells were incubated for three hours at 37°C and 5% CO₂ using the CCK-8 kit from Tiangen (Hangzhou, China), which was mixed at a volume of 10 L per well. At long last, we read the absorbance at 450 nm on the microplate peruser (Thermo Fisher Logical, Inc., USA).

Experimental glioma xenograft model

Ten male BALB/c nude mice, weighing 20 g and aged 4–6 weeks, were obtained from the Experimental Animal Center of Dalian Medical University (Dalian, China). The animals were kept in housing with a 12-hour light/dark cycle and free access to food and water. The research center circumstances were kept up with at 22°C ± 2°C and an overall dampness of 60% ± 5%. The regulations and guidelines established by the National Institutes of Health were strictly adhered to throughout the execution of each experimental protocol. The Dalian Medical University Experimental Animal Committee (Dalian, China) approved this study.

1 × 10⁷ U87 cells expressing sh-IQGAP1 or sh-Ctrl were implanted subcutaneously in the left axilla after being suspended in 75 μl of DMEM and 75 μl of Matrigel (Collaborative Biomedical Products, USA). The tumor volume was calculated using Super Nova^® PET/CT (SNPC-103, Pingsheng Medical Technology (Kunshan Co., Ltd., China) during the 14-day termly observation of the xenograft tumor development.

Statistical analysis

ANOVA, or one-way analysis of variance, was used to examine group differences. The statistical analyses were performed using the SPSS 25.0 software. All examinations were performed freely and rehashed multiple times. P < 0.05 was viewed as genuinely critical.

Identification of prognostic disulfidptosis-related DEGs in the TCGA cohort

A log-rank examination uncovered a sum of 14 qualities connected with visualization with a p-value under 0.05 (Figure 2). Among them, a worse prognosis was associated with high expression of 11 genes related to disulfideptosis. Patients with a better prognosis may have higher levels of expression of three genes related to disulfidptosis.

Figure 2. The Kaplan–Meier OS curves for patients in the high- and low-risk groups in the TCGA cohort (log-rank test). (A) ACTB, (B) ACTN4, (C) CD2AP, (D) CAPZB, (E) DSTN, (F) FLNA, (G) FLNB, (H) INF2, (I) IQGAP1, (J) MYH9, (K) MYH10, (L) MYL6, (M) PDLIM1, (N) SLC7A11, (O) TLN1.

Identification of differentially expressed disulfidptosis-related genes clusters in the GEO cohort

We gathered 209 samples (166 GBMs and 43 normal ones) after combining the GSE15824, GSE34152, and GSE35493 (platform GPL570) datasets. Figure 3A shows that the samples from the two groups are separated in the PCA diagram, demonstrating that the two groups’ differences are clear and that the outcomes of the difference analysis that follows will be more insightful. The criterion for |logFC|>1 and p-value 0.05 for genes that are differently expressed in tumor and normal tissue are shown in the volcano figure (Figure 3B). The expression of each of the top 20 genes in the expression profile was visualized using a heat map (Figure 3C). The ring heat map shows the four differentially communicated disulfidptosis-related qualities in the GEO dataset (Figure 3D). In KEGG analysis, the expression of MAPK pathway was the most different (Figure 3E). In the GO analysis, the expression of system development was most different in glioma and normal tissue (Figure 3F). In GSEA, the expression of CARDIAC_MUSCLE_CONTRACTION is the most diverse (Figure 3G).

Figure 3. Screening differential genes in GEO datasets. (A) Principal component analysis showed the clustering between tumor group and normal group of GEO combined dataset. (B) The volcano plot shows the differential expression between tumor group and normal group of GEO combined dataset. (C) The heat map showed the differential expression between tumor group and normal group of GEO combined data set. (D) The ring heat map showed the common differentially expressed genes between tumor group and normal group of TCGA dataset and GEO combined dataset. (E) KEGG analysis of differential genes. (F) GO analysis of differential genes. (G) Gene set enrichment analysis of GEO dataset.

Development of a nomogram for OS prediction

An UpSet plot (Figure 4A) was used to screen the intersection of differentially expressed genes from the TCGA and GEO datasets. Through multivariate investigation, we screened the variables related to the operating system (Figure 4B). A list with age, gender, WHO grade, IDH status, IQGAP1, primary therapy outcome, and risk scores was made to predict OS in GBMLGG patients from the TCGA cohort (Figure 4C). The nomogram’s predicted survival probability is represented by the C-index (Figure 4D). Using the nomogram, AUC demonstrated certainty of prediction for prognosis, reaching 0.965 at one year, 0.914 at three years, and 0.766 at five years (Figure 4E).

Figure 4. Establishing a nomogram model in GBMLGG. (A) UpSet diagram shows the common differentially expressed genes between tumor group and normal group of TCGA dataset and GEO pooled dataset. (B) Multivariate regression analysis was used to analyze the clinical information and differentially expressed genes of GBMLGG. (C) Nomogram showed the factors related to the prognosis of GBMLGG. (D) C-index and (E) ROC curve were used to calculate the reliability and accuracy of the Nomogram model.

DNA methylation and immune cell infiltration

The CPTAC dataset uncovered that the IQGAP1 absolute protein articulation was essentially more noteworthy in essential growth tissues than in ordinary tissues (Supplementary Figure 1A). We investigated the impact of the tumor microenvironment (TME) on patient outcomes and responses to therapy, with a particular focus on tumor infiltrating immune cells (TIICs), which have a major effect on growth movement and the effectiveness of therapy (Supplementary Figure 1B). The typical degree of the IQGAP1 bunch was 4.507 ± 1.063, and the typical degree of cg17891123 bunch was 0.349 ± 0.176. Relationship examination of IQGAP1 and cg17891123: There is a negative correlation between IQGAP1 and cg17891123, with a Pearson correlation coefficient of r = −0.467 and a P-value of 0.001. Connection examination of IQGAP1 and cg06327621: Pearson connection coefficient r = −0.583, P < 0.001, showing a negative relationship somewhere in the range of IQGAP1 and cg06327621 (Supplementary Figure 1C).

The oncogenic roles of IQGAP1 in glioma U87 cells

Immunohistochemistry analysis demonstrated that the protein expression level of IQGAP1 in GBMLGG were higher than those in the adjacent tissues, which indicated the oncogenic roles of IQGAP1 in glioma (Figure 5A). To further assess the roles of IQGAP1 in glioma, the shRNA vector was performed to steadily lower IQGAP1 expression in U87 cells. Utilizing quantitative continuous PCR, articulation levels were evaluated (Figure 5B). Decreased IQGAP1 expression significantly reduced cell viability (Figure 5C). The results of flow cytometry showed that sh-IQGAP1 cells contained 40% G0/G1 phase cells, while sh-Ctrl cells contained 26.5 percent G0/G1 phase cells, indicating that IQGAP1 knockdown halted the G0/G1 phase of the cell cycle (Figure 5D). Furthermore, IQGAP1 knockdown significantly enhanced the sensitivity to cisplatin treatment (Figure 5E). In addition, we used the xenografts that were injected into BALB/c mice to investigate the possibility that IQGAP1 plays a role in tumor growth (Figure 6A). When compared to the vector control, we discovered that IQGAP1 knockdown significantly inhibits tumor growth (Figure 6B), and there was no significant difference in mice’s body weight between groups (Figure 6C).

Figure 5. Relationship between IQGAP1 and viability of U87 cells. (A) Immunohistochemistry of IQGAP1 protein expression in GBMLGG and counterpart normal adjacent tissues. (B) qRT-PCR assay shows the transcriptional levels of the IQGAP1 gene with GAPDH used as the loading control in U87 cells. (C) Effect of sh-IQGAP1 on the proliferation of U87 cells was detected by CCK-8 assays. (D) The cell cycle distribution of U87 cells were measured using flow cytometry analyses. (E) Apoptosis of U87 cells induced by cisplatin (20 μM) were measured by flow cytometry analyses. Data are presented as the mean ± SD for three independent experiments (^*P < 0.05).

Figure 6. Effect of IQGAP1 knockdown on tumor growth of human glioma U87 xenografts. (A) Representative tumor images by Super Nova^® PET/CT at 14th day after tumor implantation. (B) Body weight of mice. (C) Tumor weight. ^*P < 0.05, compared with the control group.