CXCL9 correlates with antitumor immunity and is predictive of a favorable prognosis in uterine corpus endometrial carcinoma

Abstract

Background:

The C-X-C motif chemokine ligand-9 (CXCL9) is related to the progression of multiple neoplasms. Yet, its biological functions in uterine corpus endometrioid carcinoma (UCEC) remain shrouded in confusion. Here, we assessed the prognostic significance and potential mechanism of CXCL9 in UCEC.

Methods:

Firstly, bioinformatics analysis of the public cancer database, including the Cancer Genome Atlas / the Genotype-Tissue Expression project (TCGA+ GTEx, n=552) and Gene Expression Omnibus (GEO): GSE63678 (n=7), were utilized for the CXCL9 expression-related analysis in UCEC. Then, the survival analysis of TCGA-UCEC was performed. Futher, the gene set enrichment analysis (GSEA) was carried out to reveal the potential molecular signaling pathway in UCEC associated with CXCL9 expression. Moreover, the immunohistochemistry (IHC) assay of our validation cohort (n=124) from human specimens were used to demonstrate the latent significance of CXCL9 in UCEC.

Results:

The bioinformatics analysis suggested that CXCL9 expression was significantly upregulated in UCEC patients; and hyper-expression of CXCL9 was related to prolonged survival. the GSEA enrichment analysis showed various immune response-related pathways, including T/NK cell, lymphocyte activation, cytokine-cytokine receptor interaction network, and chemokine signaling pathway, mediated by CXCL9. In addition, the cytotoxic molecules (IFNG, SLAMF7, JCHAIN, NKG7, GBP5, LYZ, GZMA, GZMB, and TNF3F9) and the immunosuppressive genes (including PD-L1) were positively related to the expression of CXCL9. Further, the IHC assay indicated that the CXCL9 protein expression was mainly located in intertumoral and significantly upregulated in the UCEC patients; UCEC with high intertumoral CXCL9 cell abundance harbored an improved prognosis; a higher ratio of anti-tumor immune cells (CD4⁺, CD8⁺, and CD56⁺ cell) and PD-L1 was found in UCEC with CXCL9 high expression.

Conclusion:

Overexpressed CXCL9 correlates with antitumor immunity and is predictive of a favorable prognosis in UCEC. It hinted that CXCL9 may serve as an independent prognostic biomarker or therapeutic target in UCEC patients, which augmented anti-tumor immune effects to furnish survival benefits.

1 Introduction

Uterine corpus endometrial carcinoma (UCEC) ranked as the third most popular gynecological malignant neoplasm worldwide. Merely in the United States, the estimated new cases and deaths of UCEC have reached 65,950 and 12,550, respectively, in 2022 (1). Approximately 75% of patients with UCEC are diagnosed at an early stage (International Federation of Gynecology and Obstetrics [FIGO] stages I or II), with an overall 5-year survival rate of approximately 80%, While UCEC patients diagnosed with advanced stage (stage III or IV) have an overall 5-year survival of about 60% or 25% (2). Currently, multiple therapeutic strategies, including surgery, chemotherapy, radiotherapy, and hormonal therapy, are still applied to UCEC therapy, but the incidence and mortality associated with the disease continue to increase annually (3, 4). Accurate assessment of prognosis is the cornerstone of effective treatment. However, patients with the same clinical stage may have different clinical features, suggesting that UCEC prognosis is not entirely accurate according to traditional clinicopathological staging (5). Immunotherapy has been an effective treatment for many cancers in recent years (6, 7), especially in non-small cell lung carcinoma (NSCLC) (8), melanoma (9), and colorectal cancer (10). The concept of UCEC as an immunogenic tumor has emerged (11), beginning with the observation that tumor-infiltrating lymphocyte subpopulations (TILs) are closely related to the prolongation of (OS) and progression-free survival (PFS) (12, 13). Therefore, the molecular mechanisms of the UCEC development should be further identified, and new biomarkers for prognostic assessment should be investigated. In addition to aberrant gene expression profiles, the tumor microenvironment (TME), particularly immune cells, is implicated in the occurrence and development of UCEC (14). Distinct roles for anti-tumor immunity depend on cytokine-cytokine interactions, which comprise the cytokine networks that commonly sustain homeostasis of the uterine corpus (15, 16). As a class of signaling cytokines, chemokines are prominent components in the interaction between tumor cells and the microenvironment, which interacts with receptors to regulate immune infiltration, tumor-associated angiogenesis, activation of the host immune response and tumor cell proliferation (17, 18). As a member of the chemokine superfamily that encodes secretory proteins implicated in the regulation of immune response and inflammatory process, the chemokine ligand C-X-C motif chemokine ligand-9 (CXCL9) is reported to be involved in T-cell trafficking, acting as a chemoattractant to lymphocytes but not neutrophils via the C-X-C motif 3 (CXCR3) pathway of the chemokine receptor (19). The CXCL9-CXCR3 signaling pathway involves in several physiological activities, including migration, differentiation, and activation of immune cells (20, 21). Overexpression of CXCL9 has been reported to associate with increased T-cell infiltration and prolonged OS in ovarian cancer (22). Based on the evidence of human samples and C57BL/6 mouse assay, hyper-expressed CXCL9 has been proven to stifle tumor cell growth and promote anti-PD-L1 therapy in ovarian cancer, especially the histological subtypes of clear-cell carcinomas (23). Chow et al. revealed that CXCL9 (derived from CD103⁺ dendritic cell) and CD8⁺ T cells (CXCR3⁺ are essential for the efficiency of anti-PD-1 therapy in melanoma, colon adenocarcinoma, and breast adenocarcinoma cell lines (21). However, the expression profile, masked prognosis significance, and molecular mechanism of CXCL9 in UCEC remain unclear. In addition, the CXCL9/10/11-CXCR3 signaling axis mainly regulates immune cell migration, differentiation, and activation. The immune response is triggered by the recruitment of immune cells, including cytotoxic lymphocytes (CTL), macrophages, natural killer cells (NK), and NK T cells. CXCL9 was crucial for the recruitment of NK and T cells, and facilitates interactions between DCs and T cells during immunotherapy (24). Tumor-infiltrating lymphocytes are the key to achieving a favorable prognosis and predicting response to existing checkpoint inhibitors. However, in vivo studies have demonstrated that the CXCL9-CXCR3 signaling axis also has tumorigenicity by enhancing tumor cell proliferation and metastasis (25). Therefore, more investigations on CXCL9 in the tumor microenvironment are required to learn whether it has any clinical relevance for immunotherapy or could serve as a predictive biomarker for UCEC.

2 Materials and methods

2.1 Patient UCEC specimens

In this study, three independent datasets were investigated, including the TCGA dataset (n=552), the GSE63678 dataset (n=7), and our validation cohort (n=124). All specimens in our validation cohort were gathered as formalin-fixed and paraffin-embedded (FFPE) specimens and diagnosed by two or three pathologists to evaluate pathologic characteristics. Finally, we collected 34 normal control endometrial tissues (corresponding normal tissues from leiomyoma or inflammation samples) and 90 UCEC samples from February 2019 to February 2022 for this study. All clinical samples and relevant information were gathered from patients or family members with informed consent (IFC). This study was supported and approved by the Ethics Committee of Taihe Hospital. Clinically relevant parameters of all selected patients were illustrated in Supplementary Table S1.

2.2 Immunohistochemistry (IHC) analysis

The IHC assay of all FFPE tissues was conducted according to the manufacturer’s protocol. Specifically, all 3 μm sections cut from FFPE were dewaxed with xylene and rehydrated with graded ethanol. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 10 min. Sections were incubated with primary antibody (Supplementary Table S2) at 4°C overnight and with HRP-labeled secondary antibody for 0.5 h at 37°C, and hematoxylin staining was performed for 30 s at 37°C.

The protein expression was assessed via a combined score based on the intensity and the extent of staining under 200× field microscopy. Three experienced pathologists scanned and scored all IHC staining results, and the final score was taken as the median value. The assessing criteria for the extent of staining of positive cells were denoted as follows: 0 (absent), 1 (1%-10%), 2 (11%-50%), 3 (51%-80%) and 4 (81%-100%). The staining intensity scores were grad as 1 (weak), 2 (moderate), and 3 (strong). The quantity and intensity scores were multiplied to yield an overall score of 0 to 12. Finally, cases with scores of 0-3 were seated as a low-expression subgroup. And, cases with scores of 4-12 were recorded as a high-expression subgroup.

2.3 Quantitative real-time PCR

RT-qPCR analyzed the relative expression of the CXCL9 according to the merchant’s protocol. Total RNA was isolated from FFPE blocks by RNeasy FFPE Kit (QIAGEN, Germany). Total RNA of 1 μg was reversely transcribed into cDNAs using Revert-Aid First Strand cDNA Synthesis Kit (Thermo Scientific, USA). CXCL9 was amplified in a 20-μL volume that contained 2 μL of cDNAs, 1 μl of forward and reverse primers (10 nmol/L) and 10 μL of PowerUp SYBR Green Master Mix (Thermo Scientific, USA) for 38 cycles (95˚C for 15 s, 57˚C for 15 s, 72˚C for 30 s) after an initial 120 s denaturation at 95˚C in an ABI Prism 7500 analyzer (Applied Biosystems, USA). GAPDH was utilized as an endogenous reference gene. All reactions were run in triplicate. The relative CXCL9 mRNA expression was assessed by using the 2^−ΔΔCt method. All primers were manufactured by Sangon Biotech (Shanghai, China), and the corresponding sequences were presented in Supplementary Table S3.

2.4 Immune cell infiltration signatures

Immune cell infiltration into tumors has emerged as a research hotspot as one of the crucial indicators for speculating on the effect of immunotherapy. The correlation between immune cell infiltration and CXCL9 mRNA expression was assessed by five algorithms, including ssGSEA, ESTIMATE, TIMER, TISIDB, and xCELL, based on TCGA-UCEC data. In addition, Spearman’s test was adopted to measure the correlation between CXCL9 expression and immune cell abundance.

2.5 Gene set enrichment analysis

For exploring the underlying enriched molecular pathways in CXCL9^high tumor samples, gene set enrichment analyses were investigated by the Molecular SignaturesDatabase of GSEA software (version 4.2.2; Broad Institute, USA). Then, three predefined gene sets, including ‘h.all.v7.2.symbols.gmt’, ‘c2.cp.go.v7.2.symbols.gmt’ and’c2.cp.kegg.v7.2.symbols.gmt’, was assessed to yield the enrichment score. The main statistical results of GSEA include two indicators (normalized enrichment score (NES) threshold: | NES |>1 and P-value< 0.05).

2.6 Differential expression analysis

The R package ‘limma’ was applied to analyze differentially expressed genes in UCEC with CXCL9^high versus CXCL9^low samples. The upregulated or down-regulated genes were defined with a fold change ≥1.5 and normalized P-value< 0.05.

2.7 Associations of CXCL9 expression with immunosuppressive-related genes

The associations of CXCL9 expression with several immunosuppressive-related genes, including CD274(PD-L1), CTLA4, HAVCR2, LAG3, PDCD1(PD-1), PDCD1LG2(PD-L2), TIGIT, and SIGLEC15 closely related to immunotherapy, were explored in UCEC. Further, the “ggstatsplot” package of R software was applied to analyze the correlation between CXCL9 expression and microsatellite instability (MSI) or tumor mutation burden (TMB) to predict the potential anti-tumor effect of immunosuppressants.

2.8 Statistical analysis

Statistical analyses were performed using GraphPad Prism 8.0 software (San Diego, CA) and SPSS 25.0 software (IBM SPSS Inc., Chicago, IL). Data were presented as the mean ± standard deviation (Mean ± SD). Comparisons between the 2 groups were performed by One-Way ANOVA or 2-tailed Student’s t-test. The Kruskal-Wallis test was used to assess the correlation of CXCL9 expression with clinical features. We deployed both uni-and multiple Cox regression assays to characterize the effect of CXCL9 expression on survival time and other clinical features. For prognostic analysis, the log-rank test was applied. The correlation between the two variables was tested using Pearson’s chi-square test.Statistical significance was established at a threshold of P <0.05 (ns=non-significant, * P <0.05, ** P <0.01, *** P <0.001).

3 Results

3.1 Aberrant expression of CXCL9 in UCEC, including various histomorphological subtypes and molecular subtypes

The bioinformatics analysis was carried out to state that the CXCL9 expression was remarkably upregulated in UCEC tumor, compared with control ones based on the TCGA+GTEx dataset (P< 0.0001, Figure 1A), TCGA dataset (Figure 1B, P=0.0003), GSE36389 (Figure 1C, P=0.014) and tissues of our validation cohort (Figure 1D, P= 0.011). However, no difference was found in different clinical stages (Figure 1E, P > 0.05), histological grades (Figure 1F, P> 0.05), and histological subtypes (Figure 1G, P > 0.05) based on the TCGA-UCEC data repository. Further, the analysis of classical molecular subtypes was performed to reveal that the expression of CXCL9 was significantly different among CN-high (high-level somatic copy number alterations), CN-low, MSI, and POLE mutation subgroups in UCEC patients (Figure 1H, P < 0.01). Specifically, the CXCL9 was significantly overexpressed in POLE^mutation (POLE mutation-type), compared with POLE^wt (POLE wild-type) (Figure 1I, P < 0.001), but not between TP53^mutation and TP53^wt (Figure 1J, P> 0.05). Additionally, a tight correlation was depicted between the expression of CXCL9 and MSI score in UCEC (Figure 1K, P< 0.001, r=0.27).

Figure 1

3.2 Correlation between CXCL9 expression and clinicopathological signatures and prognostic value in UCEC patients

All UCEC cases from TCGA were divided into two subgroups, including CXCL9^low (CXCL9 low expression, n=272) and CXCL9^high (CXCL9 high expression, n=273), according to the median cut-off to probe the correlation between CXCL9 expression and clinicopathological signatures in UCEC (Figure 2A). As illustrated in Table 1, CXCL9 expression was independent of age, BMI, histological subtype, histologic grade, clinical stage, etc. (P>0.05). However, UCEC patients with CXCL9^high undergo a favorable outcome, compared with CXCL9^low subgroups in the TCGA dataset (P=0.017) and the validation cohort (Supplementary Table S1, P=0.008). Then, the Sanguini diagram depicted the distribution of CXCL9 expression in age, TNM stages, grades, and survival status (Figure 2B). Further, the uni-cox and multi-cox regression assays were conducted to analyze the correlation between CXCL9 and clinical factors, such as age and TNM stage, on the OS of UCEC. In the uni-cox analysis, CXCL9 expression, age, clinical stage, histological subtype, histologic grade, tumor invasion, and residual tumor were closely related to OS in UCECC patients. In the multi-cox analysis, CXCL9 expression, age, and the clinical stage could be independent prognostic factors for the OS of UCEC patients (Table 2, all P<0.05). Then, the Nomogram was performed to evaluate prognosis at 1, 2, or 3 years in UCEC patients with CXCL9^high (Figures 2C, D). Compared with the CXCL9^low subgroup, CXCL9^high indicates a better OS (Figure 2E; HR = 0.48, log-rank P=0.0000836) and PFS (Figure 2F; HR =0.629, log-rank P= 0.0111) based on the predictive analysis of the TCGA-UCEC dataset.

Figure 2

3.3 Tumor immune microenvironment in UCEC associated with CXCL9 expression

To probe the masked mechanism of CXCL9 in UCEC, we investigated the relationship between CXCL9 and tumor immune microenvironment based on the TCGA dataset. Five algorithms, including ssGSEA, ESTIMATE, TIMER, TISIDB, and xCELL, were performed to study immune cell infiltration associated with differential CXCL9 expression in UCEC. The ssGSEA figured out that overt differences of multiple immune cell populations were shown between the CXCL9^high and CXCL9^low subgroups of UCEC. The expression of CXCL9 was positively correlated with T cells, B cells, NK cells, DC cells, macrophages, etc., but not mast cells and eosinophils (Figure 3A). The ESTIMATE suggested that the enrichment of stromal score, immune score, and ESTIMATE score were dramatically different in the UCEC with CXCL9^high compared to the CXCL9^low (Figure 3B, all P<0.001). In Figure 3C, the TIMER demonstrated that CXCL9 expression level was negatively related to tumor purity (P <0.001, r=-0.212) but positively associated with B cells (P<0.001, r=0.527), CD8⁺ T cells (P<0.001, r=0.435), CD4⁺ T cells (P<0.001, r=0.42), macrophages (P<0.001, r=0.211), neutrophil (P<0.001, r=0.433), and DC (P<0.001, r=0.57). As shown in Figure 3D, the TISIDB asserted that CXCL9 expression was positively related to the abundance of Act B (activated B cells; r = 0.57, P < 0.001), Act CD4 (activated CD4 T cells; r = 0.63, P< 0.001), Act CD8 (activated CD8 T cells; r = 0.72, P< 0.001), Act DC (activated dendritic cells; r = 0.442, P < 0.001), NK (natural killer cells; r = 0.49, P< 0.001), NKT (natural killer T cells; r = 0.51, P< 0.001), and Th1 (activated dendritic cells; r = 0.61, P< 0.001) in UCEC. The xCELL depicted that CXCL9 high expression was mainly enriched in CD4⁺/CD8⁺ T cells, B cells, NK cells, monocyte, M1/M2-like macrophages, and DC, as well as different stroma score, immune score, and microenvironment score (Figure 3E). As presented in Figure 3F, the GSEA enrichment analysis of the GO: BP items showed that UCEC patients with CXCL9^high mainly were enrolled in the lymphocyte activation involved in immune response (NES=2.58, P =0), lymphocyte-mediated immunity (NES=2.56, P =0), NK cell activation (NES=2.61, P=0), NK cell-mediated immunity (NES=2.52, P =0), and regulation of T cell activation (NES=2.57, P =0). The GSEA analysis of the HALLMARK description uncovered that the overexpression of CXCL9 was tightly related to the inflammatory response (NES=2.30, P=0) and interferon-γ response (NES=2.13, P=0.002). The GSEA analysis of KEGG pathways pointed out that UCEC patients with CXCL9^high were significantly correlated with the antigen processing and presentation (NES=2.38, P=0), chemokine signaling pathway (NES=2.47, P=0), cytokine-cytokine receptor interaction (NES=2.37, P=0), NK cell-mediated cytotoxicity (NES=2.59, P=0), and T cell receptor signaling pathway (NES=2.29, P=0), which are widely known to augment the anti-tumor immunity.

Figure 3

Then, the differentially expressed genes in UCEC patients with CXCL9^high versus CXCL9^low subgroups were studied to find that high CXCL9 mRNA expression was remarkably related to various immune-activated genes. As presented in Figure 4Avia the volcano plot, the CXCL9^high was enriched in the expression of IFNG, CXCL13, CXCL11, CXCL10, CCL4, CCL5, CCL18, CCL19, CD8A, and PD-L1. As illustrated in Figures 4B–F, CXCL11 was positively associated with the chemokine gene-set related to DC recruiting (CCL4: r=0.732, P<0.0001; CCL5: r=0.763, P<0.0001), T cell and NK cell recruiting (CXCL10: r=0.746, p<0.0001; CXCL11: r=0.73, P<0.0001), T cell chemoattractant (CCL18: r=0.525, P<0.0001), B cell chemoattractant (CXCL13: r=0.522, P<0.0001), and lymphocyte homing (CCL19: r=0.618, P<0.0001). To some extent, CXCL9 was identified to facilitate the antitumor immune microenvironment in UCEC. Additionally, the cytotoxicity molecules, including SLAMF7, JCHAIN, NKG7, GBP5, LYZ, GZMA, GZMB, TNF3F9, and IFNG, were positively related to the expression of CXCL9 mRNA expression (Figure 4G; Spearman’s r = 0.82, 0.63, 0.73, 0.79, 0.71, 0.7, 0.69, 0.28, and 0.83, respectively; all P< 0.01). Furthermore, several immunosuppressive-related genes, including CD274(PD-L1), CTLA4, HAVCR2, LAG3, PDCD1, PDCD1LG2, and TIGIT, were positively associated with CXCL9 expression (Figure 4H; Spearman’s r = 0.632, 0.751, 0.706, 0.725, 0.732, 0.711, and 0.873, respectively; all P< 0.001), but not SIGLEC15. Then, the OS analysis of TIMER suggested that high immune cell abundance, especially B cell (P=0.019) and CD8⁺ T cell (P=0.022), indicates a beneficial clinical outcome (Figure 4I). Lastly, it implied that CXCL9 expression was positively correlated with the TMB score in UCEC (Figure 4J; r=0.43, P<0.0001).

Figure 4

3.4 Aberrant expression and prognosis of CXCL9 in our validation cohort

To further investigate the role of CXCL9, the UCEC samples cohort of our hospital was collected. Firstly, the protein expression of CXCL9 in tumor tissue was upregulated, compared with control tissue (Figures 5A, B) by IHC assay. Then, all patients were divided into two subgroups, high-expression and low-expression, based on the overall IHC score, and the respective clinical prognosis was analyzed. It suggested that UCEC patients with high CXCL9 abundance in tumor interstitial tissue featured a significantly better cumulative survival within the validation cohort (Figure 5C, P=0.0023). Furthermore, based on the uni-cox and multi-cox regression analyses, interstitial CXCL9 expression was strongly associated with UCEC patients’ survival and could be used as an independent prognostic biomarker for UCEC patients (Figure 5D; HR = 0.208; 95% CI 0.062–0.705; P= 0.012. Figure 5E, HR = 0.454; 95% CI 0.128–0.607; P= 0.0221).

Figure 5

3.5 Validation of CXCL9⁺ cells for association with immune cells and PD-L1

Furthermore, the immunomodulatory function of CXCL9⁺ cell infiltration was assessed in the validation cohort. As presented in Figure 6A, tumors with significant infiltration of CXCL9+ cells showed significant intertumoral CD8⁺ T cells, CD4⁺ T cells, and CD56⁺ NK cells, which were related to antitumor immunity but not CD20⁺ B cells. Additionally, high levels of CXCL9⁺ cell infiltration were associated with the high level of PD-L1⁺ cells, which were expressed primarily at the junction between tumor cells and stroma. Thus, the expression of PD-L1, CD8A, CD4, and CD56, but not CD20, was higher in the CXCL9^high group than in the CXCL9^low group (Figure 6B). Within the validation cohort, it suggested that CXCL9 expression was positively associated with PD-L1, CD4, CD8A and CD56 (Figures 6C–F; r = 0.57, P< 0.001; r = 0.54, P< 0.001; r = 0.63, P< 0.001; r = 0.46, P< 0.001, respectively).

Figure 6

4 Discussion

UCEC is the most common gynecological neoplasm in high-income countries, and its incidence is rising worldwide (26). Based on clinical and endocrine characteristics (e.g., types I and II), histopathological signatures (e.g., endometrioid, clear-cell, or serous adenocarcinoma), or molecular subtypes (e.g., microsatellite instability, TP53 mutation, POLE mutation, or high-level somatic copy number alterations, HER2 amplification), it reveals the histological heterogeneity and complexity of this neoplasm from different perspectives (27, 28). Additionally, the homeostasis between tumor cells and tissue microenvironment (especially immune cells) played a crucial role in the tumorigenesis and progression of tumors (14). Cytokines are soluble proteins that arrange cell migration according to a defined concentration gradient. Cytokines form the immune landscape of the tumor microenvironment during the early stages of tumor development. CXCL9, also known as INF gamma (MIG) -induced monocytes, can be produced in inflammatory conditions by antigen-presenting cells (e.g., dendritic cells or macrophages) or tumor cells in the tumor microenvironment (29). It attracts cells expressing CXCR3 receptors, including activated T and NK cells, and has been proven closely related to the response to immune checkpoint therapy. Overexpressed CXCL9 has also been reported to impede tumor progression and metastasis by inhibiting angiogenesis (30). The unique roles of antitumor immunity are closely related to the cytokine-cytokine interactions, which form the cytokine networks to maintain the homeostasis of the uterine corpus (15, 16).

In this study, we found that the CXCL9 expression was dramatically upregulated in UCEC tumors, compared with control ones based on the TCGA dataset, GSE36389, and human tissues of our validation cohort. The overexpression of CXCL9 has independent prognostic value for UCEC. Furthermore, the results showed that upregulation of CXCL9 expression acts as a defense against the tumorigenesis or progression of UCEC.

In different cancers, CXCL9 serves as a tumor suppressor or promoter, as it may regulate cellular processes in a specific manner through different regulatory networks. The expression of CXCL9 could be induced by IL-27, IFN-γ, D-galactosamine, etc., via JAK/STAT1, NF-κB, Fra-1, and Eg-1 signaling pathways. Moreover, CXCL9, as one of the momentous chemoattractants for leukocytes, B cells, and T cells, can serve as a tumor suppressor (31). Zhang et al. reported that a combination of CXCL9 gene therapy with low-dose cisplatin improved therapeutic efficacy in colon carcinoma (CT26) and Lewis lung carcinoma (LL/2c) murine models via inhibiting angiogenesis, augmenting CTL infiltration, and showed thymus-dependent antitumor effects (32). Hoch et al. reported that the tumor microenvironment (TME) enriched by CXCL9 and CXCL10 contributes to the generation of a “hot” tumor microenvironment and predicts favorable OS in melanoma (33). Marcovecchio et al. showed that CXCL9-expressing tumor-associated macrophages enhance anti-PD(L)-1 response rates by regulating the recruitment of stem-like CD8 T cells, thus furnishing a novel clue to fight against cancer (34). Furthermore, Tokunaga et al. demonstrated that the CXCL9, -10, -11/CXCR3 axis regulates immune cell migration, differentiation, and activation through the paracrine axis (25). Thus, determining the type of this signaling axis is a potential target for cancer therapy in preclinical studies. Conversely, previous studies also identified that CXCL9 might act directly on multiple types of tumor cells expressing the CXCR3 receptor to improve epithelial-mesenchymal transition (EMT) and cell migration. Ding et al. suggested that CXCL9 boosted the migration and invasion of CD133⁺ liver cancer cells by activating the p-ERK1/2-MMP2/MMP9 pathway (35). Mir et al. demonstrated that elevated serum levels of CXCL9 indicated a shorter median event-free survival in follicular lymphoma patients (36). Amatschek et al. revealed that a low concentration of CXCL9 induced melanoma cell migration, while conversely at a high concentration (37).

Tumor-infiltrating immune cells are closely associated with tumorigenesis, tumor cell growth, and metastasis, which in turn regulate immune cell population and differentiation (38, 39). Evidence indicates that tumor progression may result from the escape of cancer cells from host immunosurveillance (40). Thus, clarifying the infiltrating immune cells in the TME may be able to enunciate the masking mechanism involving CXCL9 in UCEC. Based on the bioinformatics strategy, it showed that a higher ratio of anti-tumor immune cells characterized CXCL9 high expression. Further, the GSEA enrichment analysis showed multiple immune response-related pathways, including T/NK cell, lymphocyte activation, cytokine-cytokine receptor interaction network, and chemokine signaling pathway, mediated by CXCL9 high expression. Further, CXCL9 was positively related to CCL4, CCL5, CXCL10, and CXCL11 in UCEC, which are associated with DC, NK, and T cell recruitment and play an essential role in suppressing tumor growth and improving prognosis (41–43). Meanwhile, CXCL9 was also positively correlated with CXCL13, CCL18, and CCL19, which were closely related to the chemoattractant of B cell (44), T cell (45), and lymphocyte homing (46, 47). Then, positive correlations were found between CXCL9 and several cytotoxic molecules (SLAMF7, JCHAIN, NKG7, GBP5, LYZ, GZMA, GZMB, TNF3F9, and IFNG) that were shown by other immunocytes to promote cytotoxic function and contribute to immune-promoting (48). These results suggest that CXCL9 promotes anti-tumor immunity by mediating immune cell infiltration. Moreover, it suggested that CXCL9 expression positively correlated with PD-L1 in the TCGA and our validation cohort. CXCL9 expression was positively correlated with the TMB score in UCEC (Figure 4J). These discoveries were consistent to some extent with the finding that inhibition of CXCL9 expression or protein activity will weaken the therapeutic effect of anti-PD-1/PD-L1 therapy and lead to a dismal prognosis of melanoma patients (19). Since other immunosuppressants, including PD-1(PDCD1), CTLA4, HAVCR2, LAG3, PD-L2, and TIGIT, were positively correlated with CXCL9 mRNA expression, the related mechanism between CXCL9 and immune environment requires further study. Several pieces of literature have suggested that infiltration of large numbers of immune cells (e.g., CD8⁺T cells, NK cells, and DC cells) into tumor stroma can enhance the anti-tumor effect of immune checkpoint inhibitors (ICIs) (41, 49, 50). Nevertheless, anti-PD-1 therapy could not reduce the tumor growth in breast cancer and melanoma mouse models with CXCR3 knock-out treatment (51). Consequently, it is meaningful and promising to verify whether CXCL9 can infer the response of UCEC patients to ICIs. However, there are still some limitations of this paper. First, the specific cell population of CXCL9⁺ was not sorted out. Furthermore, the specific molecular mechanisms and exact functions of CXCL9 anti-tumor effects in UCEC need to be further explored and validated by in vivo and in vitro experiments.

5 Conclusion

In this study, we verified that overexpressed CXCL9 as an independent prognostic biomarker in UCEC patients augmented anti-tumor immunity and was related to significantly prolonging survival. Additionally, hyper-expression of PD-L1 associated with high-expression CXCL9 in UCEC may enhance the treatment response of patients to ICIs. Therefore, the above results can furnish new insights for clinicians to choose suitable treatment strategies for patients and improve the long-term outcomes of UCEC patients.

References

• 1

  SiegelRLMillerKDFuchs HEJemalA. Cancer statistics, 2022. CA Cancer J Clin (2022) 72:7–33. doi: 10.3322/caac.21708

  • CrossRef
  • Google Scholar

• 2

  MuraliRSoslow RAWeigeltB. Classification of endometrial carcinoma: more than two types. Lancet Oncol (2014) 15:e268–78. doi: 10.1016/s1470-2045(13)70591-6

  • CrossRef
  • Google Scholar

• 3

  XiaCDongXLiHCaoMSunDHeSet al. Cancer statistics in China and united states, 2022: profiles, trends, and determinants. Chin Med J (Engl). (2022) 135:584–90. doi: 10.1097/cm9.0000000000002108

  • CrossRef
  • Google Scholar

• 4

  IslamiFWardEMSungHCroninKATangkaFKLShermanRLet al. Annual report to the nation on the status of cancer, part 1: National cancer statistics. J Natl Cancer Inst (2021) 113:1648–69. doi: 10.1093/jnci/djab131

  • CrossRef
  • Google Scholar

• 5

  ZhouCLiCYanFZhengY. Identification of an immune gene signature for predicting the prognosis of patients with uterine corpus endometrial carcinoma. Cancer Cell Int (2020) 20:541. doi: 10.1186/s12935-020-01560-w

  • CrossRef
  • Google Scholar

• 6

  MillerKDNogueiraLDevasiaTMariottoABYabroffKRJemalAet al. Cancer treatment and survivorship statistics, 2022. CA Cancer J Clin (2022) 72:409–36. doi: 10.3322/caac.21731

  • CrossRef
  • Google Scholar

• 7

  ZhangYZhangZ. The history and advances in cancer immunotherapy: understanding the characteristics of tumor-infiltrating immune cells and their therapeutic implications. Cell Mol Immunol (2020) 17:807–21. doi: 10.1038/s41423-020-0488-6

  • CrossRef
  • Google Scholar

• 8

  CamidgeDRDoebele RCKerrKM. Comparing and contrasting predictive biomarkers for immunotherapy and targeted therapy of NSCLC. Nat Rev Clin Oncol (2019) 16:341–55. doi: 10.1038/s41571-019-0173-9

  • CrossRef
  • Google Scholar

• 9

  MaoLQiZZhangLGuoJSiL. Immunotherapy in acral and mucosal melanoma: Current status and future directions. Front Immunol (2021) 12:680407. doi: 10.3389/fimmu.2021.680407

  • CrossRef
  • Google Scholar

• 10

  GaneshKStadlerZKCercekAMendelsohnRBShiaJSegalNHet al. Immunotherapy in colorectal cancer: rationale, challenges and potential. Nat Rev Gastroenterol Hepatol (2019) 16:361–75. doi: 10.1038/s41575-019-0126-x

  • CrossRef
  • Google Scholar

• 11

  GlaireMARyanNAIjsselsteijnMEKedzierskaKObolenskiSAliRet al. Discordant prognosis of mismatch repair deficiency in colorectal and endometrial cancer reflects variation in antitumour immune response and immune escape. J Pathol (2022) 257:340–51. doi: 10.1002/path.5894

  • CrossRef
  • Google Scholar

• 12

  BrunoVCorradoGBaciDChiofaloBCarosiMARonchettiLet al. Endometrial cancer immune escape mechanisms: Let us learn from the fetal-maternal interface. Front Oncol (2020) 10:156. doi: 10.3389/fonc.2020.00156

  • CrossRef
  • Google Scholar

• 13

  MusacchioLBocciaSMCarusoGSantangeloGFischettiMTomaoFet al. Immune checkpoint inhibitors: A promising choice for endometrial cancer patients? J Clin Med (2020) 9:1721. doi: 10.3390/jcm9061721

  • CrossRef
  • Google Scholar

• 14

  GuoYELiYCaiBHeQChenGWangMet al. Phenotyping of immune and endometrial epithelial cells in endometrial carcinomas revealed by single-cell RNA sequencing. Aging (Albany NY). (2021) 13:6565–91. doi: 10.18632/aging.202288

  • CrossRef
  • Google Scholar

• 15

  WallaceAESalesKJCatalanoRDAndersonRAWilliamsARWilsonMRet al. Prostaglandin F2alpha-f-prostanoid receptor signaling promotes neutrophil chemotaxis via chemokine (C-X-C motif) ligand 1 in endometrial adenocarcinoma. Cancer Res (2009) 69:5726–33. doi: 10.1158/0008-5472.Can-09-0390

  • CrossRef
  • Google Scholar

• 16

  LuWHeFLinZLiuSTangLHuangYet al. Dysbiosis of the endometrial microbiota and its association with inflammatory cytokines in endometrial cancer. Int J Cancer. (2021) 148:1708–16. doi: 10.1002/ijc.33428

  • CrossRef
  • Google Scholar

• 17

  BulePAguiarSIAires-Da-SilvaFDiasJNR. Chemokine-directed tumor microenvironment modulation in cancer immunotherapy. Int J Mol Sci (2021) 22:9804. doi: 10.3390/ijms22189804

  • CrossRef
  • Google Scholar

• 18

  BhatAANisarSMaachaSCarneiro-LoboTCAkhtarSSiveenKSet al. Cytokine-chemokine network driven metastasis in esophageal cancer; promising avenue for targeted therapy. Mol Cancer. (2021) 20:2. doi: 10.1186/s12943-020-01294-3

  • CrossRef
  • Google Scholar

• 19

  HumblinEKamphorstAO. CXCR3-CXCL9: It's all in the tumor. Immunity. (2019) 50:1347–9. doi: 10.1016/j.immuni.2019.05.013

  • CrossRef
  • Google Scholar

• 20

  KarinN. CXCR3 ligands in cancer and autoimmunity, chemoattraction of effector T cells, and beyond. Front Immunol (2020) 11:976. doi: 10.3389/fimmu.2020.00976

  • CrossRef
  • Google Scholar

• 21

  ChowMTOzgaAJServisRLFrederickDTLoJAFisherDEet al. Intratumoral activity of the CXCR3 chemokine system is required for the efficacy of anti-PD-1 therapy. Immunity. (2019) 50:1498–512.e5. doi: 10.1016/j.immuni.2019.04.010

  • CrossRef
  • Google Scholar

• 22

  BrongerHSingerJWindmüllerCReuningUZechDDelbridgeCet al. CXCL9 and CXCL10 predict survival and are regulated by cyclooxygenase inhibition in advanced serous ovarian cancer. Br J Cancer. (2016) 115:553–63. doi: 10.1038/bjc.2016.172

  • CrossRef
  • Google Scholar

• 23

  SeitzSDreyerTFStangeCSteigerKBräuerRScheutzLet al. CXCL9 inhibits tumour growth and drives anti-PD-L1 therapy in ovarian cancer. Br J Cancer. (2022) 126:1470–80. doi: 10.1038/s41416-022-01763-0

  • CrossRef
  • Google Scholar

• 24

  OzgaAJChow MTLusterAD. Chemokines and the immune response to cancer. Immunity. (2021) 54:859–74. doi: 10.1016/j.immuni.2021.01.012

  • CrossRef
  • Google Scholar

• 25

  TokunagaRZhangWNaseemMPucciniABergerMDSoniSet al. CXCL9, CXCL10, CXCL11/CXCR3 axis for immune activation - a target for novel cancer therapy. Cancer Treat Rev (2018) 63:40–7. doi: 10.1016/j.ctrv.2017.11.007

  • CrossRef
  • Google Scholar

• 26

  CrosbieEJKitsonSJMcAlpineJNMukhopadhyayAPowell MESinghN. Endometrial cancer. Lancet. (2022) 399:1412–28. doi: 10.1016/s0140-6736(22)00323-3

  • CrossRef
  • Google Scholar

• 27

  HuvilaJPorsJThompson EFGilksCB. Endometrial carcinoma: molecular subtypes, precursors and the role of pathology in early diagnosis. J Pathol (2021) 253:355–65. doi: 10.1002/path.5608

  • CrossRef
  • Google Scholar

• 28

  Urick MEBellDW. Clinical actionability of molecular targets in endometrial cancer. Nat Rev Cancer. (2019) 19:510–21. doi: 10.1038/s41568-019-0177-x

  • CrossRef
  • Google Scholar

• 29

  ReschkeRGajewskiTF. CXCL9 and CXCL10 bring the heat to tumors. Sci Immunol (2022) 7:eabq6509. doi: 10.1126/sciimmunol.abq6509

  • CrossRef
  • Google Scholar

• 30

  Neo SYLundqvistA. The multifaceted roles of CXCL9 within the tumor microenvironment. Adv Exp Med Biol (2020) 1231:45–51. doi: 10.1007/978-3-030-36667-4_5

  • CrossRef
  • Google Scholar

• 31

  DingQLuPXiaYDingSFanYLiXet al. CXCL9: evidence and contradictions for its role in tumor progression. Cancer Med (2016) 5:3246–59. doi: 10.1002/cam4.934

  • CrossRef
  • Google Scholar

• 32

  ZhangRTianLChenLJXiaoFHouJMZhaoXet al. Combination of MIG (CXCL9) chemokine gene therapy with low-dose cisplatin improves therapeutic efficacy against murine carcinoma. Gene Ther (2006) 13:1263–71. doi: 10.1038/sj.gt.3302756

  • CrossRef
  • Google Scholar

• 33

  HochTSchulzDElingNGómezJMLevesque MPBodenmillerB. Multiplexed imaging mass cytometry of the chemokine milieus in melanoma characterizes features of the response to immunotherapy. Sci Immunol (2022) 7:eabk1692. doi: 10.1126/sciimmunol.abk1692

  • CrossRef
  • Google Scholar

• 34

  MarcovecchioPMThomasGSalek-ArdakaniS. CXCL9-expressing tumor-associated macrophages: new players in the fight against cancer. J Immunother Cancer. (2021) 9:e002045. doi: 10.1136/jitc-2020-002045

  • CrossRef
  • Google Scholar

• 35

  DingQXiaYDingSLuPSunLLiuM. An alternatively spliced variant of CXCR3 mediates the metastasis of CD133+ liver cancer cells induced by CXCL9. Oncotarget. (2016) 7:14405–14. doi: 10.18632/oncotarget.7360

  • CrossRef
  • Google Scholar

• 36

  MirMAMaurerMJZiesmerSCSlagerSLHabermannTMaconWRet al. Elevated serum levels of IL-2R, IL-1RA, and CXCL9 are associated with a poor prognosis in follicular lymphoma. Blood. (2015) 125:992–8. doi: 10.1182/blood-2014-06-583369

  • CrossRef
  • Google Scholar

• 37

  AmatschekSLucasREgerAPfluegerMHundsbergerHKnollCet al. CXCL9 induces chemotaxis, chemorepulsion and endothelial barrier disruption through CXCR3-mediated activation of melanoma cells. Br J Cancer. (2011) 104:469–79. doi: 10.1038/sj.bjc.6606056

  • CrossRef
  • Google Scholar

• 38

  Di CaroGMarchesiFLaghiLGrizziF. Immune cells: plastic players along colorectal cancer progression. J Cell Mol Med (2013) 17:1088–95. doi: 10.1111/jcmm.12117

  • CrossRef
  • Google Scholar

• 39

  LiXRamadoriPPfisterDSeehawerMZenderLHeikenwalderM. The immunological and metabolic landscape in primary and metastatic liver cancer. Nat Rev Cancer. (2021) 21:541–57. doi: 10.1038/s41568-021-00383-9

  • CrossRef
  • Google Scholar

• 40

  VinayDSRyanEPPawelecGTalibWHStaggJElkordEet al. Immune evasion in cancer: Mechanistic basis and therapeutic strategies. Semin Cancer Biol (2015) 35 Suppl:S185–s98. doi: 10.1016/j.semcancer.2015.03.004

  • CrossRef
  • Google Scholar

• 41

  ShimasakiNJainACampanaD. NK cells for cancer immunotherapy. Nat Rev Drug Discovery (2020) 19:200–18. doi: 10.1038/s41573-019-0052-1

  • CrossRef
  • Google Scholar

• 42

  WculekSKCuetoFJMujalAMMeleroIKrummel MFSanchoD. Dendritic cells in cancer immunology and immunotherapy. Nat Rev Immunol (2020) 20:7–24. doi: 10.1038/s41577-019-0210-z

  • CrossRef
  • Google Scholar

• 43

  Thommen DSSchumacherTN. T Cell dysfunction in cancer. Cancer Cell (2018) 33:547–62. doi: 10.1016/j.ccell.2018.03.012

  • CrossRef
  • Google Scholar

• 44

  RubioAJPorterTZhongX. Duality of b cell-CXCL13 axis in tumor immunology. Front Immunol (2020) 11:521110. doi: 10.3389/fimmu.2020.521110

  • CrossRef
  • Google Scholar

• 45

  CardosoAPPintoMLCastroFCostaÂMMarques-MagalhãesÂCanha-BorgesAet al. The immunosuppressive and pro-tumor functions of CCL18 at the tumor microenvironment. Cytokine Growth Factor Rev (2021) 60:107–19. doi: 10.1016/j.cytogfr.2021.03.005

  • CrossRef
  • Google Scholar

• 46

  ComerfordIHarata-LeeYBuntingMDGregorCKara EEMcCollSR. A myriad of functions and complex regulation of the CCR7/CCL19/CCL21 chemokine axis in the adaptive immune system. Cytokine Growth Factor Rev (2013) 24:269–83. doi: 10.1016/j.cytogfr.2013.03.001

  • CrossRef
  • Google Scholar

• 47

  YanYChenRWangXHuKHuangLLuMet al. CCL19 and CCR7 expression, signaling pathways, and adjuvant functions in viral infection and prevention. Front Cell Dev Biol (2019) 7:212. doi: 10.3389/fcell.2019.00212

  • CrossRef
  • Google Scholar

• 48

  Garzón-TituañaMAriasMASierra-MonzónJLMorte-RomeaESantiagoLRamirez-LabradaAet al. The multifaceted function of granzymes in sepsis: Some facts and a lot to discover. Front Immunol (2020) 11:1054. doi: 10.3389/fimmu.2020.01054

  • CrossRef
  • Google Scholar

• 49

  TumehPCHarviewCLYearleyJHShintakuIPTaylorEJRobertLet al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature. (2014) 515:568–71. doi: 10.1038/nature13954

  • CrossRef
  • Google Scholar

• 50

  SabadoRLBalanSBhardwajN. Dendritic cell-based immunotherapy. Cell Res (2017) 27:74–95. doi: 10.1038/cr.2016.157

  • CrossRef
  • Google Scholar

• 51

  ChhedaZSSharmaRKJalaVRLuster ADHaribabuB. Chemoattractant receptors BLT1 and CXCR3 regulate antitumor immunity by facilitating CD8⁺ T cell migration into tumors. J Immunol (2016) 197:2016–26. doi: 10.4049/jimmunol.1502376

  • CrossRef
  • Google Scholar