3.1 Clinical characteristics and the impact of SLC7A11 expression on prognosis

We examined RNA sequencing and clinical data from 293 cervical cancer patients. The initial datasets are illustrated in Fig. 1A, and the datasets with confounding factors removed are depicted in Fig. 1B. Using the survminer package, the cut-off value for SLC7A11 expression was established at 0.64. Patients were then divided into high-expression and low-expression groups according to the determined cut-off value.

Fig. 1

Data preprocessing and the effect of SLC7A11 on the cervical cancer prognosis. Barplot before (A) and after (B) normalization. C The Kaplan–Meier analysis with log-test of the overall survival of patients between the SLC7A11 expression group. D The forest plot of univariate Cox regression analysis. E The forest plot of multivariate Cox regression analysis

Kaplan–Meier analysis revealed that patients with high SLC7A11 expression had significantly lower overall survival (OS) (p = 0.0059) (Fig. 1C). We thoroughly examined the influence of SLC7A11 expression levels alongside clinical factors such as age, FIGO stage, BMI, cancer status, ethnicity, histology, lymph nodes examined and positive, lymphovascular invasion, radiation therapy, and tumor grade on patient prognosis. This study utilized both univariate and multivariate Cox proportional hazards analyses. We used Schoenfeld residuals to check for proportional hazards before the multivariate Cox proportional hazards. And the result showed the means of survival satisfy the Proportional Hazard assumption (Figure S2).

Univariate Cox regression analysis revealed significant associations between prognosis in cervical cancer patients and the expression level of SLC7A11, FIGO stage, BMI, cancer status, lymphovascular invasion, and the number of positive lymph nodes (Fig. 1D, p < 0.05). However, several clinical variables exhibited a substantial amount of missing data (sample size N < 200). Consequently, these variables were excluded from subsequent multivariate analyses. Variables that demonstrated significant differences in the univariate analysis and had fewer missing values (BMI, FIGO stage, and SLC7A11 expression level) were incorporated into the multivariate analysis.

The findings demonstrated that both SLC7A11 expression levels and FIGO stage were significant (Fig. 1E , p < 0.05). Therefore, the two variables were chosen for additional analysis. The analysis indicated that SLC7A11 expression levels may act as a prognostic biomarker for cervical cancer patients.

3.2 In vitro assays for validation

A validation experiment was conducted in vitro. We transfected si-SLC7A11 and si-NC into HeLa cells. Post-transfection, SLC7A11 expression was downregulated in HeLa cells, as illustrated in Fig. 2A. Figure 2B demonstrates that transfection with si-SLC7A11 inhibited cervical cell proliferation, as shown by the CCK-8 assays. Si-SLC7A11 markedly reduced migration and invasion in HeLa cells relative to the control group (Fig. 2C-D). These findings establish SLC7A11 as a key regulator that enhances the proliferation, migration, and invasion of cervical cancer cells.

Fig. 2

SLC7A11 promote the Hela cells proliferation, migration and invasion. A SLC7A11 expression was downregulated in Hela cells after transfection. B The CCK-8 assays result between si-SLC7A11 and si-NC. C The migration of Hela cells was inhibited in si-SLC7A11 group. D The invasion of Hela cells was inhibited in si-SLC7A11 group

3.3 Nomogram construction

Utilizing multivariate Cox analyses, we constructed a nomogram integrating SLC7A11 expression levels and FIGO stage to forecast cervical cancer patient survival outcomes (Fig. 3A).

Fig. 3

The nomogram constructed and validated. A The nomogram of the gene signature for predicting patient survival. B The ROC curves of the nomogram on 1, 3, and 5-year OS in TCGA database. C–E The calibrate curve of the nomogram on 1, 3, and 5-year OS. F The DCA curve of the nomogram. G The ability to differentiate cervical cancer patients of the nomogram. OS, overall survival

The nomogram's AUC values for predicting survival rates at 1, 3, and 5 years were 0.684, 0.649, and 0.646, respectively (Fig. 3B). And the c-index was 0.693. The calibration curve indicated that the predicted values closely corresponded with the actual patient survival rates. Additionally, nearly perfect calibration curves were observed (Fig. 3C–E). Decision curve analysis was conducted to assess the nomogram model's clinical applicability. The results demonstrated that the nomogram exhibits good performance (Fig. 3F). Furthermore, the nomogram effectively differentiated the prognosis of patients with cervical cancer (Fig. 3G). The above results indicate that nomogram has good predictive performance and can be used as a biomarker to predict the prognosis of cervical cancer patients.

3.4 Identified differential expression genes between high and low SLC7A11 expression groups

The differential expression analysis revealed 113 DEGs between the two subgroups, with 56 genes upregulated and 57 downregulated in the high expression group (Fig. 4A). Furthermore, the heatmap depicts the expression levels of DEGs across the two groups (Fig. 4B).

Fig. 4

Differential expression analysis. A The volcano plot of these DEGs, with green dots representing down-regulated genes and red dots representing up-regulated genes. B The heatmap of DEGs between SLC7A11 expression group

3.5 Identification of SLC7A11-related gene by WGCNA analysis

A weighted co-expression network was constructed based on the expression profiles of the 113 DEGs from 293 samples. To achieve a scale-free network, the soft threshold power value was set to 9 in this study (Fig. 5A).

Fig. 5

WGCNA analysis. A Soft power in WGCNA. B Clustering and merging of the co-expression modules. C Association heatmap of module genes and clinical features. Red means positive association, and green refers to negative association

After establishing the adjacency matrix and Topological Overlap Matrix (TOM), Fig. 5B depicts the module characteristic genes, which represent the overall gene expression levels of each module. Two distinct modules, each identified by unique colors, were determined based on gene correlations (Fig. 5C). Subsequently, we examined the correlation between each trait gene and the phenotypic variables, specifically overall survival time and SLC7A11 expression levels. The analysis revealed that the grey modules exhibited a significant correlation with SLC7A11 expression levels (Correlation = 0.42, p = 9e−14) and survival status (Correlation = 0.2, p = 7e−04) (Fig. 5C). The genes in grey modules were list in Table 1.

Table 1 The genes in grey modules3.6 Functional enrichment analysis

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were employed using the ClusterProfiler tool to elucidate the potential functions of differentially expressed genes (DEGs) between the subgroups. The top 20 results from the GO enrichment analysis are illustrated in Fig. 6A. BP analyses showed that these genes were enriched in chemokine related pathways (Fig. 6B). CC analysis showed that the apical plasma membrane, intermediate filament, basement membrane were enriched (Fig. 6C). CC results showed that most of these genes were located on the membrane. MF analysis indicated that these DEGs were majorly located in receptor-ligand activity, cytokine activity, chemokine activity, CXCR chemokine receptor binding (Fig. 6D). Chemokines are a large family of small (typically 8–14 kDa), chemoattractant proteins that play a central role in controlling leukocyte migration during development, homeostasis, and inflammation. While chemokines play an essential role in directing cell migration, chemokines and chemokine receptors are also associated with a diverse array of pathological conditions. Due to their abilities to control leukocyte trafficking and inflammatory responses, high chemokine expression levels or unregulated chemokine signaling has the potential to drive excessive or persistent inflammation. Additionally, chemokines are involved in regulating anti-tumor immunity not only by regulating the immune cell composition of the tumor microenvironment, but also by regulating tumor cell proliferation and metastasis.

Fig. 6

Function enrichment analysis of the DEGs. A The top 20 GO enrichment analysis results. B The result of biological process analyses. C The results of cellular component analysis. D The results of molecular function analysis. E The results of KEGG enrichment analysis. F The results of GSEA analysis in the SLC7A11 high expression group. G The results of GSEA analysis in the SLC7A11 low expression group

Based on KEGG pathway analysis, these genes were primarily involved in IL-17 signaling pathway, Cytokine-cytokine receptor interaction, NF-kappa B signaling pathway, TNF signaling pathway, Chemokine signaling pathway and so on (Fig. 6E). IL-17 plays a key role in host immunology. The nuclear factor-κB (NF-κB)/Rel proteins include NF-κB2 p52/p100, NF-κB1 p50/p105, c-Rel, RelA/p65, and RelB. These proteins act as dimerizing transcription factors and regulate gene expression, affecting various biological processes, including innate and adaptive immunity, inflammation, stress response, B cell development, and lymphoid organ development. Tumor necrosis factor (TNF), also known as TNF-α, is a cytokine that can directly kill tumor cells without obvious cytotoxicity to normal cells. It is involved in systemic inflammatory response and is one of the cytokines that make up the acute phase response.

We also performed GSEA analysis and the results showed that metabolism was mainly enriched in the SLC7A11 high-expression group (Fig. 6F). Meanwhile, genes related to the innate immune system, and developmental biology were involved in the SLC7A11 low expression group (Fig. 6G). The high expression group has strong metabolism and can provide more energy for cancer cells to promote their proliferation, invasion and metastasis. In the low expression group, the enrichment of immune-related pathways indicated that the immunosuppressive state in the low expression group may be higher than that in the expression group, which may be one of the reasons for the better prognosis.

3.7 Tumor microenvironment

Additionally, we utilized the CIBERSORT and MCPcounter algorithm to assess whether the expression of SLC7A11 could affect the distribution of infiltrating immune cells. As the results showed, NK cells resting, Neutrophils, Macrophages M0, and Mast cells activated were more enriched in the SLC7A11 high expression group, while Dendritic cells resting, Mast cells resting, T cells follicular helper, were more enriched in the SLC7A11 low expression group (Fig. 7A , p < 0.05).

Fig. 7

Tumor microenvironment. A The immune cell infiltration between two groups using CIBERSORT algorithm. B The correlation between SL7A11 expression and immune infiltrating cells. C The correlation between SL7A11 expression and immune genes. D The TMB score between the high and low SLC7A11 expression groups

Utilizing the CIBERSORT algorithm, we conducted an in-depth analysis of the correlation between SLC7A11 expression and immune infiltrating cells in cervical cancer. The results (Fig. 7B) showed that SL7A11 was positively related with Mast cells activated (r = 0.20), Neutrophils (r = 0.19), Macrophages M0 (r = 0.14), NK cells resting (r = 0.25) and negatively correlated with T cells follicular helper(r = − 0.15), Mast cells resting (r = − 0.17), T cells regulatory (Tregs) (r = − 0.17). The above results were statistically significant (p < 0.05). Utilizing the MCPcounter algorithm, the result showed that the NK cells were more enriched in the SLC7A11 high expression group (p < 0.05) (Figure S3).

Utilizing the ImmPort Portal database, we identified a total of 2,483 genes.After deinterlacing the differentially expressed genes (DEGs), 29 immune-related genes were identified.SLC7A11 expression showed a positive correlation with STC2, AREG, CXCL8, EREG, CXCL3, CXCL5, STC1, CXCL2, PTHLH, CCL20, IFITM1, CXCL1, SEMA3B, and IL1A.The expression of SLC7A11 was negative correlation with IGHV1-69-2, FAM3B, PTN, CXCL17, RBP1, BMP7, RBP7, ACKR1, CCL19, IL34 (Fig. 7C).

Tumor mutational burden (TMB) has been shown to predict the response to immune checkpoint blockade [28]. Consequently, we compared the TMB scores between groups with high and low SLC7A11 expression. Our findings indicate that the group with high SLC7A11 expression exhibited significantly higher TMB scores (Fig. 7D), which means these patients may have a positive effect on immunotherapy.

3.8 Drug sensitivity analysis

To enhance the therapeutic outcomes for cervical cancer patients, we conducted an in-depth investigation into the differential sensitivity to commonly utilized chemotherapeutic agents and targeted drugs between the two patient groups. Patients with low SLC7A11 expression had higher IC50 for AZ6102, AZD1332, AZD2014, BMS-536924, Dabrafenib, ERK, PAK, RO-3306, Staurosporine, VSP34, and WIKI4.Patients exhibiting elevated SLC7A11 expression could potentially respond well to these drugs (Fig. 8). In contrast, the IC50 values of Acetalax, AZD5991, BIBR-1532, BMS-754807, Dihydrorotenone, Doramapimod, Elephantin, GSK1904529A, Ibrutinib, JQ1, Lapatinib, LGK974, MK-2206, ML323, OF-1, OSI-027, Pyridostatin, Sabutoclax, Sepantronium bromide, Sorafenib, TAF1, Topotecan, Uprosertib, Venetoclax, Wnt-C59, ZM447439 were lower in the low expression group (Fig. 8). These drugs may be more beneficial for patients with low SLC7A11 expression.

Fig. 8

The sensitivity of the drug by analyzing the half-maximal inhibitory concentration (IC50) to treat cervical cancer