CXC ligand 13 orchestrates an immunoactive microenvironment and enhances immunotherapy response in head and neck squamous cell carcinoma

Abstract

Objectives: This study aims to systematically explore the role of chemokine CXC ligand 13 (CXCL13) in head and neck squamous cell carcinoma (HNSCC). Methods: The Genotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) databases provided the RNA-seq data for cancer and normal tissues, respectively. Gene set enrichment analysis was applied to search the cancer hallmarks associated with CXCL13 expression. TIMER2.0 was the main platform used to investigate the immune cell infiltration related to CXCL13. Immunohistochemistry was applied to explore the relationship between CXCL13 and patients’ prognosis and the relationship between CXCL13 and tertiary lymphoid structures (TLSs). Results: The expression of CXCL13 was upregulated in most tumors, including HNSCC. The higher expression of CXCL13 was closely related to the positive prognosis of HNSCC. CXCL13 was mainly expressed in B cells and CD8 + T cells, revealing the relationship between its expression and immune activation in the tumor microenvironment. Furthermore, immunohistochemistry and multiple fluorescence staining analysis of HNSCC samples showed a powerful correlation between CXCL13 expression, TLSs formation, and positive prognosis. Finally, CXCL13 significantly increased the response to cancer immunotherapy. Conclusions: CXCL13 may function as a potential biomarker for predicting prognosis and immunotherapy response and associate with TLSs in HNSCC.

Keywords

CXC ligand 13 head and neck squamous cell carcinoma prognosis tertiary lymphoid structures immunotherapy

Introduction

Head and neck squamous cell carcinoma (HNSCC) ranks as the sixth most prevalent malignant neoplasm worldwide, with an annual incidence of 1,000,000 newly diagnosed cases.¹ Despite the active development of comprehensive and sequential therapeutic strategies for HNSCC, more than half of HNSCC patients still suffer local recurrence or lymph node/distant metastasis.² Immunotherapy blocking PD-1/PD-L1 axis has demonstrated the ability to induce long-term tumor control and improve the overall survival of patients with HNSCC.³ Recent clinical trials have shown promising results with the use of pembrolizumab (PD-1 antibody) in combination with chemotherapy as a first-line treatment for recurrent or metastatic HNSCC. However, the effectiveness of immunotherapy varies among patients due to the heterogeneity heterogeneous nature of the of tumor microenvironment.^4,5 Therefore, it is crucial to identify the factors that determine response and resistance in order to enhance survival outcomes and explore novel therapeutic targets.

Chemokines, which are small secreted proteins, play a crucial role in forming concentration gradients within the tumor microenvironment. They are released by immune cells, tumor cells, and other cells associated with the tumor.⁶ These chemokines can undergo dynamic changes over time and attract a diverse array of that can either promote or inhibit tumor growth. The migration of immune cells in response to these chemokines not only impacts the progress and metastasis of tumor but also directly influences the immune response.^7,8 The CXCL family is a group of small chemokine proteins that play crucial roles in the immune system. Chemokines are signaling molecules involved in cell migration and recruitment during inflammatory responses, immune surveillance, as well as various physiological and pathological processes. The CXCL family is named based on the presence of the C-X-C motif in their protein structure, which indicates the arrangement of cysteine amino acids separated by a single amino acid (represented as “X"). These chemokines are produced by different cell types, including immune cells, endothelial cells, and fibroblasts, in response to various stimuli such as infection, inflammation, and tissue damage. They bind to specific receptors called chemokine receptors on the surface of target cells, guiding these cells towards sites of inflammation or infection. One specific member of the CXCL family is chemokine CXC ligand 13 (CXCL13), which plays a critical role in the homing of B lymphocytes to lymph follicles. The activated germinal center helper T cells showed high expression.^9,10 Elevated expression of CXCL13 can promote B cell recruitment and induce the formation of ectopic germinal centers in a distinct population of lung fibroblasts.¹¹ Besides, CXCL13 influences local anti-tumor immunity by recruiting immune cells and facilitating the forming of tertiary lymphoid structures (TLSs). Previous studies have revealed that TLSs, driven by the expression of CXCL13, could serve as predictive or prognostic biomarker for immunotherapy in various cancers, such as colorectal cancer, melanoma, and breast cancer.^6,12–14 These significant studies encourage a comprehensive understanding of CXCL13 in HNSCC.

In this study, we comprehensively valued the potential functions of chemokines in HNSC patients through prognostic analysis and immunity analysis. Our results verified that higher expression of CXCL13 was significantly related to the positive prognosis of HNSCC. In addition, CXCL13 was showed correlations with immune-infiltration, particularly with immune cells related to tertiary lymphoid structures (TLSs). Patients with higher expression of CXCL13 exhibited higher response rate to anti-CTLA-4 or anti-PD-1 treatment and improved overall survival. Furthermore, immunohistochemistry and multiple fluorescence staining analysis further supported that CXCL13 expression was significantly associated with higher TLSs formation and a favorable prognosis in HNSCC patients. These discoveries suggested that CXCL13 may function as a potential biomarker for predicting prognosis and response to immunotherapy for HNSCC.

Methods

Data sources

This study lasted for a total of 12 months, from May 2022 to May 2023. In this study, we obtained the mRNA expression and clinical data of patients or samples from the TCGA pan-cancer cohort and Genotype-Tissue Expression (GTEx) datasets. These data were downloaded from the UCSC Xena database (https://xenabrowser.net/datapages/). To assess the genomic change frequency of CXCL13 in 32 different cancer types, we utilized the web application cBioPortal for Cancer Genomics (https://cbioportal.org/). Furthermore, to investigate the protein interactions of CXCL13, we employed the GeneMANIA web tool (GeneMANIA). The overall analysis process is depicted in flow-process diagram (supplementary Figure 1). The study employed a randomized controlled trial design, including a tumor tissue group and a peritumor tissue group.

Patients and specimens

The study was approved by the Ethics Committee of Shanghai Ninth People’s Hospital affiliated with Shanghai Jiao Tong University, School of Medicine. HNSCC tissues were obtained from the Shanghai Sharing Platform for the Tissue and Bioinformatics Database of Oral Maxillofacial Tumors (https://mdl.shsmu.edu.cn/OMNDB/page/home/home_en.jsp), which was established by the Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, and the Shanghai Institute of Stomatology (Shanghai, China) (2018-86-T77). All tissue samples used for the Sharing Platform were collected from the Department of Oral and Maxillofacial-Head and Neck Oncology, Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine. Written informed consent was obtained from legally authorized representatives before the study because the study subjects were unable to make informed decisions due to their survival status.

Prognostic analysis of CXCL13 across cancers

To assess the prognostic role of CXCL13 in each cancer type, we obtained overall survival (OS) data from the UCSC Xena database (https://xenabrowser.net/datapages/). We then performed univariate Cox regression and the Kaplan–Meier model using the continuous variable of CXCL13 expression data to evaluate its prognostic significance for a specific prognosis type in each cancer. Furthermore, we used bivariate CXCL13 expression levels to perform Kaplan–Meier curve analysis. The log-rank p value of the K-M method and the hazard ratio (HR) along with the 95% confidence interval (95% CI) were calculated, and the results were presented as a heatmap.

Single-cell analysis of CXCL13

We conducted a related single-cell analysis using the Tumor Immune Single-cell Hub (TISCH) web tool (https://tisch.comp-genomics.org/documentation/).¹⁵ The analysis focused on the expression of CXCL13 (gene), major lineage (cell-type annotation), and all cancers (cancer type). These visualizations provided valuable insights into the distribution and variation of CXCL13 expression across different cell types within the tumor microenvironment.

Gene set enrichment analysis

We downloaded the “gmt” file of the hallmark gene set (h.all.v7.4.symbols.gmt) from the Molecular Signatures Database (MSigDB) website (https://www.gsea-msigdb.org/gsea/index.jsp). This file contains 50 hallmark gene sets that represent different biological processes. To assess the enrichment of differentially expressed genes (DEGs) between the low- and high-CXCL13 expression cancer groups in the context of biological processes, we utilized the R package “clusterProfiler.” We calculated the normalized enrichment score (NES) and false discovery rate (FDR) for each biological process using the hallmark gene set. Using the results obtained from the GSEA analysis, we created a bubble plot summarizing the findings. The bubble plot was generated using the R package “ggplot2” and visually represents the NES and FDR for each biological process associated with DEGs between the low- and high-CXCL13 expression cancer groups in HNSCC.

Immune cell infiltration analysis in TIMER2

To assess the correlations between CXCL13 mRNA expression and various immune cell populations in HNSCC, we utilized the Tumor Immune Estimation Resource (TIMER) (https://timer.cistrome.org/). Spearman correlation provides a measure of the strength and direction of the monotonic relationship between two variables, in this case, CXCL13 expression and the abundance of immune cell populations.

Immunotherapy prediction analysis

To explore the relationship between CXCL13 and well-known immunotherapy biomarkers, such as tumor mutation burden (TMB), microsatellite instability (MSI), and other well-known immune checkpoint genes, we performed Spearman correlation analysis. Using this statistical analysis method, we quantified the strength and direction of the monotonic relationships between CXCL13 expression and these immunotherapy biomarkers across various cancer types. The Spearman correlation coefficient provides a measure of the statistical association between two variables.

Immunohistochemistry (IHC)

A total of 50 HNSCC paraffin-embedded tissue samples were randomly selected. Immunohistochemistry was performed as described previously.¹⁶ Briefly, we sliced paraffin-embedded tissue blocks into 3 μm sections, deparaffinized and rehydrated them. Pressure cooking with EDTA buffer (pH 8.0) for 15 min was performed for antigen retrieval. We blocked endogenous peroxidase with 0.3% H2O2. Subsequent to the blocking with 5% bovine serum albumin for 45 min, sections were incubated with CXCL13 (Proteintech, 10,927-1-AP, 1:250), CD3 (Proteintech, 17,617-1-AP, 1:1000), CD20 (Proteintech, 60,271-1-lg, 1:3000) and secondary antibody. We covered the resulting sections with diaminobenzidine to visualize staining and counterstained them with hematoxylin. The staining intensity was classified into four categories: none (0), weak brown (1+), moderate brown (2+), and strong brown (3+). The proportion of cells with positive staining was divided into five categories: 0 (0%), 1 (1%–25%), 2 (26%–50%), 3 (51%–75%), and 4 (76%–100%). The staining score was calculated by multiplying the staining intensity and the percentage of positive cells. Inclusion criteria: Aged 18 years and above; received a confirmed pathological diagnosis of HNSCC; first occurrence of HNSCC; absence of significant metastatic malignant tumors; general good health status suitable for undergoing surgical treatment. Exclusion criteria: Pregnant or lactating women; history of other malignant tumors; severe cardiac, pulmonary, hepatic, renal, or other organ diseases; mental illness or cognitive impairment.

Multiple fluorescence staining

The multiplex fluorescence staining procedure using the Absin kit (item number: abs50012) was conducted as follows: paraffin sections were heated at 60°C for 1 h and dewaxed in xylene, gradient alcohol dehydration, 10% neutral formalin immersion for 10 min; antigen was repaired in EDTA solution using microwave repair and cooled to room temperature, after 10 min of blocking, CXCL13 (Proteintech, 10,927-1-AP, 1:100) was incubated overnight as an antibody; the secondary antibody was incubated for 10 min then incubated with fluorescent dye for 10 min, After the microwave repair was cooled to room temperature, the above steps were repeated again to complete the staining of CD3 (Proteintech, 17,617-1-AP, 1:100) and CD20 (Proteintech, 60,271-1-Ig, 1:400), and finally incubated with DAPI for 5 min to stain the nucleus and anti-fluorescence quencher seal. The expression was observed under fluorescence microscope.

Statistical analysis

To compare the expression levels of CXCL13 between normal tissues and tumor tissues, we utilized the Wilcoxon rank-sum test. Univariate Cox regression analysis and the Kaplan–Meier method were employed to assess the prognostic value of CXCL13 expression in each cancer. Spearman correlation analysis was performed to assess the statistical relationships between CXCL13 and other factors. All hypothetical tests were two-sided, and p values <0.05 were considered significant.

Results

Basic information of CXCL13

We first assessed the basic information of chemokines in HNSCC. There have been reports on 48 different chemokines in this context, which can be categorized into four main classes based on the location of the first two cysteine (C) residues in their primary protein structure, namely, the C, the CC, the CXC and the CX3C chemokines. We employed the TCGA database to evaluate the expression levels of CCL1-CCL28, CXCL1-CXCL17, XCL1-XCL2 and CX3CL1 in HNSCC. Through prognostic analysis and immunity analysis, we found three predictably genes with high expression and prognosis matches immune promotion or immunosuppression function, including CCL20, CCL25, and CXCL13 (Figure 1(a)). Eventually, we analyzed the tumor immune-infiltration score of these three genes (Figures 1(b)–(d)), and CXCL13 got the highest score. We therefore focus on CXCL13 for further research.

Figure 1.

Basic information of chemokines. (a) A heatmap containing the expression level, the prognostic role and tumor immune-infiltration score of 48 chemokines in HNSCC. (b) Tumor immune-infiltration score of CXCL13 in HNSCC. (c) Tumor immune-infiltration score of CCL25 in HNSCC. (d) Tumor immune-infiltration score of CCL20 in HNSCC.

In order to gain a better understanding of the role of CXCL13 in cancer, we conducted an analysis utilizing the TCGA and GTEx databases to evaluate the expression level of CXCL13 in cancers compared with normal tissues (Figure 2(a)). The results revealed high expression levels of CXCL13 were identified in the majority of TCGA cancers, including HNSCC (Figure 2(b) and (c)). Furthermore, we performed genomic alteration analysis of CXCL13 across different cancer types. The analysis indicated that alterations in the CXCL13 gene were not universally observed across all cancers, and the most frequently altered cancer type was Bladder Urothelial Carcinoma (BLCA); approximately 2.2% of BLCA patients showed CXCL13 alterations, mostly mutation and amplification (Figure 2(d)). Figure 2(e) showed the Methylation of CXCL13. However, no CpG Island was found. Besides, we found ZNF680 through the transcription factor prediction web tool UCSC, which got the highest score. Next, by analyzing the JASPAR database, we predicted the potential ZNF680 binding sites in the promoter regions of CXCL13 (Figure 2(f)). Furthermore, a protein–protein interaction (PPI) network was constructed based on the interaction data obtained from the GeneMANIA website, showing that the protein closely related to CXCL13 contributed most to chemokines and related receptors (Figure 2(g)). Finally, the functions of co-expression in HNSCC patients were then predicted by Gene Ontology (GO) and KEGG enrichment analyses. The top GO terms in the biological process (BP), molecular function (MF), and cellular component (CC) groups were cell chemotaxis, external side of plasma membrane, and G protein-coupled receptor binding, respectively. KEGG analysis revealed that the cytokine-cytokine receptor interaction was enriched (Figure 2(h)).

Figure 2.

Basic information of CXCL13. (a) The expression level of CXCL13 in tumor and normal tissues across all TCGA tumors. (b, c) The expression level of CXCL13 in tumor and normal tissues in HNSCC. (d) CXCL13 alteration frequency analysis in the pan-cancer study according to the cBioPortal database. (e) Distribution of CXCL13 methylation island in promoter region. (f) The recognition sequences of ZNF680 from the JASPAR database and potential ZNF680 binding sites were found in the promoter sequence of CXCL13. (g) The protein-protein interaction (PPI) network presents the proteins interacting with CXCL13. (h)The GO, KEGG enrichment analysis of CXCL13 and proteins interacting with CXCL13.

Prognostic analysis of CXCL13 in HNSCC

The diagnostic ROC curve analysis demonstrated that CXCL13 had superior diagnostic efficacy for HNSCC (Figure 3(a)). Additionally, Kaplan–Meier curve analysis indicated that lower CXCL13 expression was associated with poor survival outcomes (Figure 3(b)). The Sankey diagram illustrated the distribution trends of high and low CXCL13 expression in different stages, age groups, and other clinical characteristics of HNSCC samples along with patient survival (Figure 3(c)). Univariate Cox regression analysis revealed that tumor stage (N, M), radiation therapy, lymphovascular invasion, and CXCL13 expression were prognostic factors for overall survival (OS). Moreover, multivariate Cox regression confirmed that CXCL13 expression was an independent risk factor in OSCC patients (Figure 3(d) and (e)). Additionally, Table 1 illustrated the relationship between CXCL13 and clinically relevant variables in TCGA-HNSC. Considering that CXCL13 emerged as an independent prognostic factor in HNSCC, we attempted to establish a prediction model of OS by incorporating the expression of CXCL13 with the clinicopathological parameters. A nomogram was constructed by integrating CXCL13 expression and other prognostic factors, including the tumor stage (N), radiation therapy, and lymphovascular invasion (Figure 3(f)). Higher points on the nomogram indicated a worse prognosis. The performance of the nomogram, including CXCL13 expression, was evaluated using a calibration curve (Figure 3(g)), suggesting that this nomogram might outperform individual prognostic factors in predicting survival among HNSCC patients. HNSCC is a highly heterogeneous type of cancer that is influenced by two main factors: etiology and tumor localization. Figure 3(h) demonstrated a high level of CXCL13 expression in tumors located in the tonsil. Additionally, Figure 3(i) revealed that patients who tested positive for HPV exhibited significantly elevated levels of CXCL13 expression in their tumors. Furthermore, a heatmap displaying the normalized coefficient of the CXCL13 in the Cox model demonstrated that lower expression of CXCL13was associated with poor survival outcomes across various cancers (Figure 3(J)), including BRCA, HNSC, OV, SKCM, and UCEC. This highlights the potential relevance of CXCL13 as a prognostic marker in multiple cancer types.

Figure 3.

Prognostic analysis of CXCL13 in HNSCC. (a) The diagnostic ROC curve of CXCL13 in HNSCC. (b) Kaplan‒Meier overall survival curves of CXCL13 in HNSCC. (c) The Sankey diagram showed the distribution trend of high and low CXCL13 expression in different stages, age and other clinical characteristics of HNSCC samples and the survival of patients. (d) The forest plot shows the prognostic role of CXCL13 in HNSCC by the univariate Cox regression method. (e) The forest plot shows the prognostic role of CXCL13 in HNSCC by the multivariate Cox regression method. (f) The nomogram to integrate CXCL13 and other prognostic factors, including the tumor (N) stage, radiation therapy, and lymphovascular invasion. (g) The calibration curve using to evaluate nomogram performance with CXCL13. (h) The expression level of CXCL13 in various tumor location. (i) The relationship between HPV status and the levels of CXCL13 expression in HNSCC patients. (j)A heatmap showing the prognostic role of CXCL13 expression across cancers.

Table 1.

Relationship between CXCL13 and the indicated clinicologic features in TCGA-HNSC.

Characteristics	Number of cases	CXCL13 expression		p value
Characteristics	Number of cases	Low (n = 252)	High (n = 252)	p value
Age (years)				0.26
≤60	247	130	117
༞60	256	122	134
Gender				0.84
Female	134	66	68
Male	370	186	184
Pathology grade				0.07
Well	95	38	57
Moderate	79	44	35
Poor	262	136	126
Tumor size				0.03
T1, T2	180	78	102
T3, T4	268	144	124
Lymph node status				0.27
pN0	171	79	92
pN+	240	124	116
Tumor stage				0.02
I, II	95	38	57
III, IV	341	180	161

Function analysis of CXCL13 in HNSCC immune microenvironment

The differentially expressed genes (DEGs) between the low- and high-CXCL13 subgroups of each cancer were used to perform GSEA to discern the CXCL13-associated cancer hallmarks. We found that CXCL13 expression was remarkably related to immune-related pathways, such as the inflammatory-response, IL6-JAK-STAT3-signaling, IL2-STAT5-signaling, interferon-α-response, and interferon-γ-response (Figure 4(a)). These suggested a potential connection between CXCL13 expression and immune activation in the tumor microenvironment (TME). To illustrate the relationships between CXCL13 and cancer immunity, we further examined the correlations between CXCL13 expression and immune cell infiltration. Spearman correlation analyses were conducted utilizing pan-cancer immune cell infiltration data from the TIMER2 database. The outcomes revealed the infiltration levels of B cells, CD8 + T cells, CD4 + T cells, macrophages, neutrophils, and dendritic cells in HNSCC (Figure 4(b)). The results indicated that CXCL13 was positively associated with the infiltration levels of B cells, CD8 + T cells, CD4 + T cells, macrophages, neutrophils, and dendritic cells in HNSCC. To understand the main cell types that express CXCL13 in cancer microenvironments, we performed a single-cell analysis of CXCL13 in single-cell datasets of cancer samples. In the GSE103322 HNSC dataset, we analyzed 5902 cells from 18 HNSCC patients; CXCL13 was highly expressed in CD8 + T cells and B cells in the HNSCC microenvironment (Figure 4(c)). Consistently, the same results also appear in other datasets, including GSE180268 HNSC dataset (Figure 4(d)), GSE139324 HNSC dataset (Figure 4(e)), GSE115978 SKCM dataset (Figure 4(f)), EMTAB8107 BRCA dataset (Figure 4(g)), GSE127465 NSCLC dataset (Figure 4(h)), GSE148673 THCA dataset (Figure 4(i)) and GSE147082 OV dataset (Figure 4(j)). And CXCL13 receptor CXCR5 was mainly expressed in B cells, CD8 + T cells and CD4 + T cells (Figure 4(k)). These findings indicated that higher CXCL13 expression is associated with the immune-activation status of cancers and might provide valuable insights for investigating of the functions and roles of CXCL13 in cancer progression and immunotherapy.

Figure 4.

Function analysis of CXCL13 in HNSCC immune microenvironment. (a) The enrichment of genes in immune-related pathways by GSEA in HNSCC. (b) The correlations of CXCL13 expression and the infiltration levels of CD4 + T cells, CD8 + T cells, B cells, neutrophils, monocytes, macrophages, dendritic cells in HNSCC. (c–j) Scatter plot showing the distributions of different cell types in the GSE103322 HNSC dataset, GSE180268 HNSC dataset, GSE139324 HNSC dataset, GSE115978 SKCM dataset, EMTAB8107 BRCA dataset, GSE127465 NSCLC dataset, GSE148673 THCA dataset, and GSE147082 OV dataset. (k) The level of CXCR5 expression in different cell types in the above datasets.

Relationships between CXCL13 and tumor-infiltrating lymphocytes, immune regulators, TMB, and MSI in pan-cancer

We then further explored the role of CXCL13 in tumor immunity across cancers. The heatmaps showed the associations between CXCL13 expression and tumor-infiltrating lymphocytes (TILs) (Figure 5(a)), copy number (CNA) of CXCL13 and TILs, methylation of CXCL13 and TILs, CXCL13 expression and immunoinhibitors, CNA of CXCL13 and immunoinhibitors, methylation of CXCL13 and immunoinhibitors (supplementary Figures 2A–2(e)) across human cancers. Furthermore, we performed a single-cell analysis of CXCL13 in single-cell datasets of cancer samples to understand the main cell types that express CXCL13 in cancer microenvironments. The heatmap depicted in supplementary Figure 5(b) represents the expression levels of CXCL13 in 20 cell types in 47 datasets using the TISCH web tool. The results indicated that CXCL13 was mainly expressed in immune cells, especially CD8 + T cells. Besides, the expression levels of CXCL13 were closely related to chemokine, chemokine receptor genes, and MHC, indicating its importance in immune modulation (Figure 5(c)). CXCL13 had a strong positive relationship with tumor-infiltrating lymphocytes and most immune regulators across various cancers (Figure 5(d)). Furthermore, positive correlations between CXCL13 expression and TMB were identified in BLCA, BRCA, COAD, THYM, and UCEC (supplementary Figure 2(f)). Moreover, for the correlation between CXCL13 expression and MSI, positive associations were discovered in COAD and UCEC, and negative correlations were discovered in DLBC, HNSC and LUSC (supplementary Figure 2(g)). These findings provide valuable insights into the role of CXCL13 in immune modulation and its potential as a predictive biomarker for ICIs.

Figure 5.

Relationships between CXCL13 and tumor-infiltrating lymphocytes, immune regulators, TMB, and MSI across cancers. (a) Spearman correlations between expression of CXCL13 and TILs across human cancers. (b) A Spearman correlation heatmap showing the correlations between CXCL13 expression and chemokine genes, chemokine receptor genes and MHC genes. (c) A summary of CXCL13 expression in 20 cell types in 47 single-cell datasets. (d) A Spearman correlation heatmap showing the correlations between CXCL13 expression and 60 types of immune regulators across cancers.

Relationships between CXCL13 expression and TLSs in TCGA-HNSC

To investigate the relationship between CXCL13 and TLSs, we analysis the relationship of CXCL13 and CD3D, CD3E, CD3G (markers of T cells in TLSs), MS4A1 (markers of B cells in TLSs) and 9 marker genes of TLSs ¹⁷ in HNSCC, the result showed a strong positive relationship between CXCL13 and TLSs (Figure 6(a)). To comprehend the basic information of TLSs in HNSCC, we employed the TCGA and GTEx databases to evaluate the expression level of 9 marker genes (Figure 6(b)) and 12 chemokine genes¹⁸ (Figure 6(c)) in HNSCC compared with normal tissues. Most genes were highly expressed in HNSCC tumors. We developed a novel gene signature consisting of four genes to specifically assess HNSCC TLSs at the transcriptomic level from using transcriptomic data the dataset of TCGA (HNSCC) (Figures 6(d)–(h)). This gene signature demonstrated the significant impact of TLSs on the prognosis of HNSCC.

Figure 6.

Relationships between CXCL13 expression and TLSs in TCGA-HNSC. (a) The relationship of CXCL13 and CD3D, CD3E, CD3G, MS4A1 and 9 marker genes of TLSs in HNSCC. (b) The expression level of 9 marker genes of TLSs in tumor and normal tissues in HNSCC. (c) The expression level of 12 chemokine genes in tumor and normal tissues in HNSCC. (d–h) The novel gene signature composed of four genes to specifically evaluate HNSCC TLSs at the transcriptomic level from the dataset of TCGA (HNSCC).

Relationships between CXCL13 expression and TLSs in HNSCC samples

To verify the prognostic role of CXCL13 in HNSCC, we further performed immunohistochemistry and multiple fluorescence staining on 50 HNSCC samples (Figure 7(a) and (b)) tissue. The expression of CXCL13 in the HNSCC tissue was labeled “low” or “high expression” based on average values. The TLSs was assessed on HE sections and verified via CD3 (T cells) and CD20 (B cells) staining. As expected, the expression level of CXCL13 was positively correlated with the expression status of TLSs markers CD3 and CD20 (Figure 7(a) and (b)). Kaplan–Meier survival analysis validated that the higher CXCL13 expression in the HNSCC tissue was correlated with great OS prognosis (Figure 7(c)). The formation of TLSs also correlated with positive OS prognosis (Figure 7(d)). In addition, the TLSs were positive correlated with the expression of CXCL13 (Figure 7(e)). And time-dependent ROC curve showed the positive prognosis of TLSs in HNSCC samples (Figure 7(f)). Moreover, Table 2 illustrated the relationship between CXCL13 and clinically relevant variables in HNSCC samples. Our results again confirm the strong correlation between CXCL13 and TLSs, which predicts positive prognosis.

Figure 7.

Relationships between CXCL13 expression and TLSs in HNSCC samples. (a, b) HE staining, IHC and IF of HNSCC samples, including CXCL13, CD3, and CD20. (c) Kaplan‒Meier overall survival curves of CXCL13 in HNSCC samples (n = 50). (d) Kaplan‒Meier overall survival curves of TLSs in HNSCC samples (n = 50). (e) The correlation between CXCL13 expression and TLSs formation in HNSCC samples (n = 50). (f) Time-dependent ROC curve showed the prognosis prediction effect of TLSs in HNSCC samples (n = 50).

Table 2.

Relationship between CXCL13 and TLS expression levels and the indicated clinicologic features in HNSCC samples.

Characteristics	Number of cases	CXCL13 expression		p value	TLSs formation		p value
Characteristics	Number of cases	Low (n = 27)	High (n = 23)	p value	Low (n = 24)	High (n = 26)	p value
Age (years)				0.24			0.9
≤60	15	10	5		7	8
>60	35	17	18		17	18
Gender				0.2			0.13
Female	20	13	7		7	13
Male	30	14	16		17	13
Pathology grade				0.08			0.85
Well	10	6	4		5	5
Moderate	33	20	13		15	18
Poor	7	1	6		4	3
Tumor size				0.61			0.01
T1, T2	28	16	12		9	19
T3, T4	22	11	11		15	7
Lymph node status				0.7			0.27
pN0	29	15	14		12	17
pN+	21	12	9		12	9
Tumor stage				0.13			0.08
I, II	21	14	7		7	14
III, IV	29	13	16		17	12

Analysis of immunotherapeutic potential of CXCL13

To explore whether CXCL13 plays a role in immunotherapy response, we analysis the GSE93157 HNSCC dataset, patients were treated with anti-PD1 treatment. Differentially expressed genes (DEGs) in GSE93157 were plot in volcano plot (Figure 8(a)) and heatmap (Figure 8(b)), including CXCL13. KEGG analysis revealed that the viral protein interaction with cytokine and cytokine receptor was enriched (Figure 8(c)). Furthermore, we analysis the PRJEB23709 melanoma anti-CTLA-4 and anti-PD-1 treatment dataset. The associations between the response to immunotherapy and the expression of CXCL13 was showed in Figure 8(d), high expression levels of CXCL13 were identified in the group of responder and the KM curve suggested that lower CXCL13 expression was associated with poor survival outcomes. Due to the correlation between CXCL13 and TLSs, we further analyzed CD3D, CD3E, CD3G, and MS4A1 (Figures 8(e)–(h)), the results are consistent with CXCL13, which further proved the close relationship between CXCL13 and TLSs. Similar results also appeared in PRJEB25780 stomach adenocarcinoma anti-PD-1 treatment dataset, further confirming the immunotherapeutic potential of CXCL13.

Figure 8.

Analysis of immunotherapeutic potential of CXCL13. (a) A volcano plot showed the differentially expressed genes (DEGs) in GSE93157 HNSCC dataset. (b) A heatmap showed the DEGs in GSE93157 HNSCC dataset. (c) KEGG analysis the DEGs in GSE93157 HNSCC dataset. (d–h) The associations between the response to immunotherapy and the expression of genes and the KM curve, including CXCL13, CD3D, CD3E, CD3G, and MS4A1 in PRJEB23709 melanoma anti-CTLA-4 and anti-PD-1 treatment dataset. (i) The associations between the response to immunotherapy and the expression of CXCL13 in PRJEB25780 stomach adenocarcinoma anti-PD-1 treatment dataset.

Discussion

In this study, we found that CXCL13 was an effective biomarker for HNSCC prognosis prediction and may significantly enhance immunotherapy response. Besides, transcriptional upregulation of CXCL13 was regarded as a surrogate marker of TLSs. It is plausible to assume that excessive CXCL13 expression could affect the loco-regional anti-tumor immunity of tumors given the compelling data linking CXCL13 to the development of TLSs. Our findings can offer some suggestions for additional investigation into CXCL13’s potential function in cancer immunity and immunotherapy.

Chemotaxis of different cell types and chemokine signal transduction are key factors in the development of the tumor microenvironment in cancer. After activation, effector chemicals such as granzyme B and perforin are released by CD8 + T cells and NK cells, which can move to tumors and promote anti-tumor immunity. They mostly expressed CXCR3, and it has been demonstrated that CXCL9 and CXCL10, its ligands, can attract these immune cells into tumors.¹⁹ High levels of CCL2 have been linked to poor survival in breast cancer patients as well as increased macrophage infiltration in gastric cancer, according to studies.^20–22 CXCL8 can also encourage tumor development, epithelial-mesenchymal transition (EMT), and the attraction of MDSCs.^23,24 In this investigation, we found that CXCL13 expression increased noticeably in the majority of cancer forms, including HNSCC. Another evidence that CXCL13 plays a substantial role in prognosis prediction and has the potential to develop into a promising, robust prognostic biomarker for HNSCC patients comes from the significant association between high expression of CXCL13 and positive prognosis in HNSCC patients. Our research suggests that CXCL13 plays a crucial role in shaping the tumor microenvironment and influencing cancer development and prognosis through its association with immune-activated processes and immune cell infiltration in HNSCC. Previous studies have demonstrated the importance of immune-related pathways, such as the IFN-response, IFN-response, TNFA signaling-via-NFKB pathways, and immune-infiltration, in predicting prognosis and determining immunotherapy response in LGG patients with epilepsy.²⁵ Our findings provide further insights into the potential impact of immune modulation on cancer progression and treatment.TLSs refer to lymphoid structures found in non-lymphoid tissues. TLSs detection is currently primarily carried out by IHC analysis of pathological sections, which is useful for conclusively identifying the presence of TLSs. TLSs is found in many kinds of inflamed tissues and can drive the activation of immune cells.²⁶ Its formation is related to many chronic diseases such as chronic inflammatory diseases, autoimmune diseases and cancer.²⁷ In tumor environment, TLSs can promote immune cells to immerse into solid tumors, so the development of TLSs is significantly related to the survival rate of untreated patients.^28,29 Similarly, the development of TLSs is usually associated with improved treatment response in patients treated with immune checkpoint inhibitors.^29,30 This suggests that TLSs is the conjecture of producing anti-tumor immune sites. IHC-based TLSs section analysis does have some limitations, though, including the need for excessive clinical sample consumption, inconvenience with quantitative analysis, complexity with combination analysis with TME, and difficulty with elucidating underlying mechanisms. Therefore, there is a strong demand for using transcriptome sequencing data to establish TLSs and immunotherapy of tumors, which will be more suitable for clinical and basic research related to TLSs.

Former data revealed that the main source of CXCL13 are CD4 + T cells. Certainly, the discrepancy between previous data suggesting CD4 + T cells as the main source of CXCL13 and the new finding that CD8 + T cells also produce CXCL13 merits discussion. This unexpected observation may have several explanations that could be addressed in the discussion section of a research article. Here are a few possible points to consider:

Context-dependent expression: It is possible that the production of CXCL13 by CD8 + T cells occurs in specific contexts or microenvironments that were not previously investigated. The differential expression of CXCL13 by CD4+ and CD8 + T cells might depend on the stage of immune response, tissue location, or the presence of specific stimuli or antigens. Exploring these contextual factors could shed light on the contrasting findings.

Functional differences: Although CD4 + T cells are generally recognized as the primary producers of CXCL13, it is important to consider the functional significance of CXCL13 production by CD8 + T cells. Investigating whether CXCL13 produced by CD8 + T cells serves a distinct purpose or has a different impact on the tumor microenvironment compared to CD4 + T cell-derived CXCL13 could provide insights into their respective roles in anti-tumor immune responses.

Heterogeneity within T cell populations: T cell populations, including both CD4+ and CD8 + T cells, comprise diverse subsets with different functional properties. It is possible that specific subsets of CD8 + T cells, which were not previously examined, are responsible for CXCL13 production. Subpopulation analysis may reveal subsets of CD8 + T cells specialized in producing CXCL13 and further explain the observed discrepancy.

Methodological considerations: Differences in experimental techniques, sample preparation, and detection sensitivity between different studies could contribute to conflicting results. Assessing potential methodological variations and limitations may help reconcile the discrepancies.

Impact on immune response: The discovery of CXCL13 production by CD8 + T cells raises questions about the interplay and coordination between different immune cell subsets in the tumor microenvironment. Discussing possible implications of CD8 + T cell-derived CXCL13 on immune cell recruitment, organization of TLSs, and overall anti-tumor immune responses could provide valuable insights.

Conclusions

Our comprehensive analysis of CXCL13 in HNSCC not only identified its potential role as a prognostic biomarker but also unveiled its role in enhancing the efficacy of immunotherapy. Our findings confirmed a robust correlation between CXCL13 and tumor-infiltrating lymphocytes (TLSs), which in turn predicted a positive prognosis. This suggests that CXCL13 plays a pivotal role in regulating the growth of an immunoactive tumor microenvironment, ultimately leading to improved response rates to immunotherapy in HNSCC. These results highlight the potential utility of CXCL13 as a therapeutic target and emphasize the importance of immune modulation in treating HNSCC.

Chemokine-based immunotherapies fall into two categories: those that target tumor chemokines and those that boost anti-tumor chemokine levels. Both can be used separately as therapies or in conjunction with other forms of treatment. According to preclinical research, HuMax-IL8 can prevent CXCL8, which can lessen the recruitment of MDSC and EMT to tumor locations. The findings of the phase I clinical investigation demonstrated that, despite the absence of an objective tumor response, the serum CXCL8 level may be decreased.³¹ In this study, we discovered a favorable correlation between TLSs and greater levels of the anti-tumor chemokine CXCL13 in HNSCC tissue samples. Immunotherapy dataset also demonstrated the relationship between CXCL13 and prognosis and the relationship between CXCL13 and TLSs. Therefore, increasing CXCL13 concentration can be used as a new immunotherapy strategy in HNSCC.

Supplemental Material

Supplemental Material - CXC ligand 13 orchestrates an immunoactive microenvironment and enhances immunotherapy response in head and neck squamous cell carcinoma

Supplemental Material for CXC ligand 13 orchestrates an immunoactive microenvironment and enhances immunotherapy response in head and neck squamous cell carcinoma by Xiaohu Lin, Xiaomei Zhao, Yiming Chen, Rong Yang, Zhenlin Dai, Wei Li, Chengzhong Lin, and Wei Cao in International Journal of Immunopathology and Pharmacology

Footnotes

Authors’ contributions

Xiaohu Lin and Xiaomei Zhao: the acquisition, analysis, or interpretation of data for the work. Zhenlin Dai, Yiming Chen and Rong Yang: the analysis of data for the work. Wei Li, Chengzhong Lin and Wei Cao: substantial contributions to the conception or design of the work and final approval of the version to be published. All authors read and approved the final manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (82272816, 81972589); and the Innovation research team of high-level local universities in Shanghai (SHSMU-ZDCX20212802).

Ethical statement

ORCID iD

Xiaohu Lin

Supplemental Material

Supplemental material for this article is available online

Appendix

References

Bray

Ferlay

Soerjomataram

, et al. (2018) Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for Clinicians 68(6): 394–424.

Fasano

D'Onofrio

Belfiore

, et al. (2022) Head and neck squamous cell carcinoma in elderly patients: role of radiotherapy and chemotherapy. Cancers 14(3): 472.

Sánchez-Magraner

Gumuzio

Miles

, et al. (2023) Functional engagement of the PD-1/PD-L1 complex but not PD-L1 expression is highly predictive of patient response to immunotherapy in non-small-cell lung cancer. Journal of Clinical Oncology 41: Jco2201748.

Chakraborty

Darido

Liu

, et al. (2023) Head and neck cancer immunotherapy: molecular biological aspects of preclinical and clinical research. Cancers 15(3): 852.

Bhatia

Burtness

(2023) Treating head and neck cancer in the age of immunotherapy: a 2023 update. Drugs 83: 217–248.

Payne

Cornelius

(2002) The role of chemokines in melanoma tumor growth and metastasis. The Journal of Investigative Dermatology 118(6): 915–922.

Yang

Sun

, et al. (2022) The role of CXC chemokines in cancer progression. Cancers 15(1): 167.

Israr

DeVoti

Papayannakos

, et al. (2022) Role of chemokines in HPV-induced cancers. Seminars in Cancer Biology 87: 170–183.

Wang

, et al. (2022) CXCL13-CXCR5 axis: regulation in inflammatory diseases and cancer. Biochimica et Biophysica Acta. Reviews on Cancer 1877(5): 188799.

10.

Pan

Zhu

Liu

, et al. (2022) Role of the CXCL13/CXCR5 Axis in autoimmune diseases. Frontiers in Immunology 13: 850998.

11.

Denton

Innocentin

Carr

, et al. (2019) Type I interferon induces CXCL13 to support ectopic germinal center formation. The Journal of Experimental Medicine 216(3): 621–637.

12.

Bule

Aguiar

Aires-Da-Silva

, et al. (2021) Chemokine-directed tumor microenvironment modulation in cancer immunotherapy. International Journal of Molecular Sciences 22(18): 9804.

13.

Gao

Liu

Wang

, et al. (2021) CXCL13 in cancer and other diseases: biological functions, clinical significance, and therapeutic opportunities. Life (Basel) 11(12): 1282.

14.

Zhang

Zhu

Wang

, et al. (2017) Yin-yang effect of tumor infiltrating B cells in breast cancer: from mechanism to immunotherapy. Cancer letters 393: 1–7.

15.

Sun

Wang

Han

, et al. (2020) TISCH: a comprehensive web resource enabling interactive single-cell transcriptome visualization of tumor microenvironment. Nucleic Acids Research 49(D1): D1420–d1430.

16.

Lin

Cao

Ren

, et al. (2017) GDNF secreted by nerves enhances PD-L1 expression via JAK2-STAT1 signaling activation in HNSCC. Oncoimmunology 6(11): e1353860.

17.

Cabrita

Lauss

Sanna

, et al. (2020) Tertiary lymphoid structures improve immunotherapy and survival in melanoma. Nature 577(7791): 561–565.

18.

Prabhakaran

Rizk

, et al. (2017) Evaluation of invasive breast cancer samples using a 12-chemokine gene expression score: correlation with clinical outcomes. Breast cancer Research: BCR 19(1): 71.

19.

Tokunaga

Zhang

Naseem

, et al. (2018) CXCL9, CXCL10, CXCL11/CXCR3 axis for immune activation - a target for novel cancer therapy. Cancer Treatment Reviews 63: 40–47.

20.

Wang

, et al. (2022) Regulation of CCL2 by EZH2 affects tumor-associated macrophages polarization and infiltration in breast cancer. Cell Death and Disease 13(8): 748.

21.

Jin

Guo

Mao

, et al. (2022) Total glucosides of paeony inhibit breast cancer growth by inhibiting TAMs infiltration through NF-κB/CCL2 signaling. Phytomedicine: International Journal of Phytotherapy and Phytopharmacology 104: 154307.

22.

Liu

Song

Pan

, et al. (2020) Wnt5a/CaMKII/ERK/CCL2 axis is required for tumor-associated macrophages to promote colorectal cancer progression. International Journal of Biological Sciences 16(6): 1023–1034.

23.

Xiong

Liao

Qiu

, et al. (2022) CXCL8 in tumor biology and its implications for clinical translation. Frontiers in Molecular Biosciences 9: 723846.

24.

Lopez-Bujanda

Haffner

Chaimowitz

, et al. (2021) Castration-mediated IL-8 promotes myeloid infiltration and prostate cancer progression. Nature Cancer 2(8): 803–818.

25.

Cheng

Duan

, et al. (2021) Multi-omics data integration analysis of an immune-related gene signature in LGG patients with epilepsy. Frontiers in Cell and Developmental biology 9: 686909.

26.

Fridman

Sibéril

Pupier

, et al. (2023) Activation of B cells in tertiary lymphoid structures in cancer: anti-tumor or anti-self? Seminars in Immunology 65: 101703.

27.

Vaghjiani

Skitzki

(2022) Tertiary lymphoid structures as mediators of immunotherapy response. Cancers 14(15): 3748.

28.

Zhang

(2022) Tertiary lymphoid structures are critical for cancer prognosis and therapeutic response. Frontiers in Immunology 13: 1063711.

29.

Zou

Zhang

Zeng

, et al. (2022) Tertiary lymphoid structures: a potential biomarker for anti-cancer therapy. Cancers 14(23): 5968.

30.

Bery

Shepherd

, et al. (2022) Role of tertiary lymphoid organs in the regulation of immune responses in the periphery. Cellular and Molecular Life Sciences: CMLS 79(7): 359.

31.

Bilusic

Heery

Collins

, et al. (2019) Phase I trial of HuMax-IL8 (BMS-986253), an anti-IL-8 monoclonal antibody, in patients with metastatic or unresectable solid tumors. Journal for Immunotherapy of Cancer 7(1): 240.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB