Skip to main content
NCBI home page
Clinical and Experimental Immunology logoLink to Clinical and Experimental Immunology
. 2023 Feb 8;213(1):102–113. doi: 10.1093/cei/uxad019

Tumor gene signatures that correlate with release of extracellular vesicles shape the immune landscape in head and neck squamous cell carcinoma

Isabella Kallinger 1,#, Dominique S Rubenich 2,3,#, Alicja Głuszko 4,#, Aditi Kulkarni 5,6,7, Gerrit Spanier 8, Steffen Spoerl 9, Juergen Taxis 10, Hendrik Poeck 11, Mirosław J Szczepański 12, Tobias Ettl 13, Torsten E Reichert 14, Johannes K Meier 15, Elizandra Braganhol 16,17, Robert L Ferris 18,19,20,21, Theresa L Whiteside 22,23,24,25, Nils Ludwig 26,✉
PMCID: PMC10324554  PMID: 36752300

Abstract

Head and neck squamous cell carcinomas (HNSCCs) evade immune responses through multiple resistance mechanisms. Extracellular vesicles (EVs) released by the tumor and interacting with immune cells induce immune dysfunction and contribute to tumor progression. This study evaluates the clinical relevance and impact on anti-tumor immune responses of gene signatures expressed in HNSCC and associated with EV production/release. Expression levels of two recently described gene sets were determined in The Cancer Genome Atlas Head and Neck Cancer cohort (n = 522) and validated in the GSE65858 dataset (n = 250) as well as a recently published single-cell RNA sequencing dataset (n = 18). Clustering into HPV(+) and HPV(−) patients was performed in all cohorts for further analysis. Potential associations between gene expression levels, immune cell infiltration, and patient overall survival were analyzed using GEPIA2, TISIDB, TIMER, and the UCSC Xena browser. Compared to normal control tissues, vesiculation-related genes were upregulated in HNSCC cells. Elevated gene expression levels positively correlated (P < 0.01) with increased abundance of CD4(+) T cells, macrophages, neutrophils, and dendritic cells infiltrating tumor tissues but were negatively associated (P < 0.01) with the presence of B cells and CD8(+) T cells in the tumor. Expression levels of immunosuppressive factors NT5E and TGFB1 correlated with the vesiculation-related genes and might explain the alterations of the anti-tumor immune response. Enhanced expression levels of vesiculation-related genes in tumor tissues associates with the immunosuppressive tumor milieu and the reduced infiltration of B cells and CD8(+) T cells into the tumor.

Keywords: extracellular vesicles, exosomes, HNSCC, tumor-infiltrating immune cells, tumor microenvironment

The expression of vesiculation-related genes (EVsig) is upregulated in HNSCC and is associated with a specific immune cell landscape. Hereby, the EVsig correlates with a decreased abundance of tumor-infiltrating effector lymphocytes and with the expression of immunosuppressive markers NT5E and TGBF1.

Graphical Abstract

Graphical Abstract.

Graphical Abstract

Open in a new tab

Introduction

Head and neck squamous cell carcinomas (HNSCCs) are a heterogeneous group of carcinomas that affect multiple anatomical sites including the oral cavity, pharynx, and larynx. They represent the largest proportion of all malignant tumors in the head and neck region and are considered the sixth most common malignancy worldwide [1]. The 5-year survival rate has improved only marginally in recent decades, which emphasizes the need for advancements in the understanding of the tumor biology of HNSCC which will consequently translate into improvements of current treatment protocols [2].

Recently, release by tumor cells of extracellular vesicles (EVs) and their characterization has been one of the most intensively studied aspects of the tumor biology in various malignancies, including HNSCC [3]. These studies revealed that the release of different subsets of EVs by tumor cells is a major contributor to intercellular communication and induces functional alterations in the tumor microenvironment (TME) and at distant sites [4]. The EV subset of small EVs sized 30–150 nm that are derived from the endocytic pathway in tumor cells appears to play an important role in the TME [5]. They are actively produced by HNSCC cells, are released into extracellular space, and can be detected in all body fluids of HNSCC patients [6]. They carry a complex cargo of proteins, lipids, nucleic acids (microRNAs and other RNA-species), including soluble factors such as cytokines, chemokines, and numerous enzymes [5]. These cargo components are biologically active and induce functional changes in recipient cells [7], which can either be mediated by receptor/ligand interactions on the cell surface, fusion of EVs with the plasma membrane of the recipient cell, or by internalization of EVs via a variety of uptake mechanisms, including phagocytosis and endocytosis. Internalization of EVs has been shown to result in the transfer of factors which induce transcriptional changes in recipient cells. Recent studies emphasize that EVs released by HNSCC cells interact with different cell types leading to functional reprogramming of cells in the TME, including endothelial cells [7, 8], macrophages [9, 10], dendritic cells [11], CD4(+) and CD8(+) T cells [12, 13], B cells [14], T regulatory cells (Tregs) [14], and neurons [15]. Uniformly, these studies indicate that HNSCC-derived EVs reprogram recipient cells leading to promotion of tumor growth and ultimately to the establishment of a strongly immunosuppressive TME.

While the effects of HNSCC-derived EVs on individual cellular components of the TME are well described based on functional in vitro, ex vivo, or in vivo studies, the mechanisms underlying the effects of HNSCC-derived EVs on the TME composition remain unclear. In this study, we used a bioinformatics approach to better define the role EVs play in shaping the cellular composition of the TME. We have focused on two recently and independently published gene signatures whose expression has been linked to increased EV release from cancer cells [16, 17].

Materials and methods

Gene signatures

To test the correlation between clinical patient characteristics and the transcriptional signatures, we used two previously described sets of genes known to be involved in EV secretion (Table 1). The first signature (Fathi-EVsig) was published by Fathi et al. [16] and consists of 41 genes known to be involved in EV secretion based on previous findings in the literature. The second signature (Hurwitz-EVsig) was published by Hurwitz et al. [17] and it is based on the proteomic profiling of EVs isolated from 60 cell lines from the National Cancer Institute (NCI-60). The authors performed a subsequent network analysis of vesicle quantity and proteomes to identify EV components associated with vesicle secretion and created the gene signature shown in Table 1 based on their results. The signatures by Fathi et al. and Hurwitz et al. mostly consist of distinct genes, with RAB7A and RALB being the only genes which are included in both gene sets.

Table 1:

Gene signatures associated with EV secretiona

Fathi et al. [16] “Fathi-EVsig” Hurwitz et al. [17] “Hurwitz-EVsig”
ARF6 ANXA2
ATG3 ARF4
ATG12 BSG
BST2 CD59
CD9 CD81
CD63 HLA-A
CD82 HRAS
CHMP4C HSPA1B
CIT HSPB1
CTTN ITGB1
DGKA LGALS3BP
HGS MYL3
PDCD6IP MYL6
PKM2 NME2
PLD2 NRAS
RAB2B RAB5C
RAB5A RAB7A
RAB7A RAC1
RAB9A RALB
RAB11A RAP1B
RAB27A RHOA
RAB27B SEC22B
RAB35 TAGLN2
RALA UBE2D3
RALB VAMP3
SDC1
SDC2
SDC3
SDC4
SDCBP
SMPD3
SNAP23
STAM
STX1A
SYT7
TSG101
TSPAN8
VAMP7
VPS4A
VTA1
YKT6
Open in a new tab

aSymbols of genes used in this study.

Additionally, gene signatures representing hypoxia, matrix metalloproteinases, cancer stem cells, and EMT were composed of following genes: HIF1A; MMP2, MMP9, and MMP13; CD44, PROM1, and ALDH1A1; TWIST1, SNAI1, and SNAI2, respectively. The gene set HALLMARK_ANGIOGENESIS (systematic name: M5944) was downloaded from the open access molecular signature database (https://www.gsea-msigdb.org/gsea/msigdb/genesets.jsp) provided by UC San Diego Broad Institute. Dataset consisted of 36 genes known to be upregulated during formation of blood vessels and was used for correlation analysis.

RNA sequencing data acquisition

Gene expression levels based on RNA-seq data from HNSCC patients and clinical characteristics were obtained from the Cancer Genome Atlas (TCGA; https://gdc.nci.nih.gov) using the University of California, Santa Cruz (UCSC) Xena Browser [18]. Expression profiles of Fathi-EVsig and Hurwitz-EVsig were compared between a total of 522 cases of primary HNSCC and 44 normal control samples. Ninety-eight cases of HPV(+) HNSCC and 422 cases of HPV(−) HNSCC were used for further analysis. For data validation, the GSE65858 dataset was used with 250 cases of primary HNSCC which were further separated into 54 cases of HPV(+) HNSCC and 196 cases of HPV(−) HNSCC.

Single-cell RNA sequencing data acquisition

Transcriptomic profiles of peripheral blood lymphocytes (PBL) and tumor-infiltrating lymphocytes (TIL; n = 18 patients), as well as non-immune cells from patient tumors (n = 15 patients) were analyzed based on a previously published dataset [19, 20]. Using the scanpy.tl.score_genes() function, a gene signature score from both published EV signatures was calculated for each cell [21]. Cells were binned into appropriate groups, for calculating differences in the EV signature scores.

GEPIA2 analysis

To measure the expression levels of individual candidate genes as well as gene signatures in HNSCC and normal samples, the Gene Expression Profiling Interactive Analysis2 (GEPIA2; http://gepia2.cancer-pku.cn/) was used [22]. Moreover, the tool was used to generate overall survival and disease specific survival Kaplan–Meier plots, compare between subtypes, and perform correlation analysis.

TISIDB analysis

Tumor–immune system interactions and drug bank database (TISIDB; http://cis.hku.hk/TISIDB/) was used to determine the potential association between candidate genes and immunomodulators and chemokines in HNSCC patients [23]. Gene expression levels of immunomodulators and vesiculation-related genes were analyzed using Spearman correlation analysis. Data are presented as Fisher Z-transformation of correlation coefficient.

TIMER analysis

Tumor IMmune Estimation Resource (TIMER; https://cistrome.shinyapps.io/timer/), a web server for comprehensive analysis of tumor-infiltrating immune cells, was used to estimate the potential association between candidate genes, immune cell infiltration, and clinical parameters [24]. In this study, gene module was used to select genes of interest and visualize the correlation of its expression with immune infiltration levels in HNSCC patients. For further analysis, patients were separated into HPV(+) and HPV(−) HNSCC cases.

Statistical analysis

Statistical analysis was performed using GraphPad Prism version 9.4.0. Data are presented with P values determined in t tests or fold changes. For survival analyses, the Kaplan–Meier method was used to analyze the correlation between gene expression levels and survival, and the log-rank test to compare the survival curves. Correlations were assessed using Spearman’s correlation analysis. Fisher Z-Transformation was used to transform the sampling distribution of correlation coefficients. A P value < 0.05 was defined as a statistically significant difference.

Results

Expression of vesiculation-related genes differs in HNSCC samples and normal tissues and has a negative impact on patient survival

Gene expression levels of vesiculation-related genes were increased in HNSCC tissues for both Fathi-EVsig (Fig. 1A) and Hurwitz-EVsig (Fig. 1B) compared to those in solid normal tissues, although not reaching statistically significant differences. Analysis of the individual gene expression levels of the vesiculation-related genes in tumors comparing to normal control tissues revealed that 19 of the 41 (Fathi-EVsig) and 12 of the 25 (Hurwitz-EVsig) vesiculation-related genes were upregulated in HNSCC, whereas the expression levels of 20 and 6 genes, respectively, were downregulated in HNSCCs compared to solid normal tissues (Fig. 1C and D). For both gene signatures no significant impact on overall survival was detectable (Fig. 1E and G); however, there was a trend that increased expression levels were associated with poor disease-free survival (Fig. 1F and H). For Hurwitz-EVsig, disease-free survival was significantly worse at high gene expression levels compared to patients with low expression levels (P = 0.031; Fig. 1H). Analysis of individual gene expression levels showed that nearly all genes of the Hurwitz-EVsig contributed to poor survival (Fig. 1J), while approximately half of the genes belonging to the Fathi-EVsig were associated with poor survival (Fig. 1I).

Figure 1:

Figure 1:

Open in a new tab

Analysis of gene expression levels of Fathi-EVsig and Hurwitz-EVsig in HNSCC and impact on patient survival. (A) Comparison of Fathi-EVsig gene expression in normal solid tissue (n = 44) and in primary HNSCC tumors (n = 519). (B) Comparison of Hurwitz-EVsig gene expression in normal solid tissue (n = 44) and in primary HNSCC tumors (n = 519). (C) Comparison of gene expression levels of the individual genes of the Fathi-EVsig in normal solid tissue (n = 44) and in primary HNSCC tumors (n = 519). (D) Comparison of gene expression levels of the individual genes of the Hurwitz-EVsig in normal solid tissue (n = 44) and in primary HNSCC tumors (n = 519). (E) Kaplan–Meier overall survival curves of Fathi-EVsig high expression versus low expression in patients diagnosed with HNSCC. (F) Kaplan–Meier disease free survival curves of Fathi-EVsig high expression versus low expression in patients diagnosed with HNSCC. (G) Kaplan–Meier overall survival curves of Hurwitz-EVsig high expression versus low expression in patients diagnosed with HNSCC. (H) Kaplan–Meier disease free survival curves of Hurwitz-EVsig high expression versus low expression in patients diagnosed with HNSCC. (I) Overall survival based on the expression of the individual genes of the Fathi-EVsig. Red: increased risk; Blue: decreased risk. (J) Overall survival based on the expression of the individual genes of the Hurwitz-EVsig. Red: increased risk; Blue: decreased risk (for colour figure refer to online version)

To validate these findings in an independent cohort and further investigate the expression patterns of the two EV signatures in HNSCC, a previously published single-cell sequencing dataset which included 18 HNSCC patients was analyzed. The expression levels of both gene signatures were significantly higher in non-immune cells, compared to expression levels in PBLs and TILs (Fig. 2A and D). Both EV signatures were equally expressed in pericytes, endothelial cells, fibroblasts, and epithelial cells; however, gene expression levels were significantly lower in immune cells. Dendritic cells, macrophages, and mast cells were the immune cell subsets with the highest expression levels of Fathi-EVsig and Hurwitz-EVsig (Fig. 2B, C, E, and F). Further, we compared the gene expression levels of the two EV signatures in PBLs and TILs and levels of Fathi-EVsig were significantly elevated in all studied immune cell subsets with the exception of monocytes and macrophages (Fig. 2G). CD8(+) T cells, dendritic cells, and Tregs were characterized by higher levels of Hurwitz-EVsig when comparing tumor-infiltrating and circulating cells (Fig. 2H). These results indicate that the EV release is accelerated in immune cells infiltrating the tumor tissue.

Figure 2:

Figure 2:

Open in a new tab

Analysis of gene expression levels of Fathi-EVsig and Hurwitz-EVsig in the HNSCC single-cell RNA sequencing cohort. (A) Expression of the Fathi-EVsig in PBL, CD45(+) tumor-infiltrating lymphocytes and CD45(−) cells. (B) Expression of the Fathi-EVsig by individual PBL subsets. (C) Expression of the Fathi-EVsig by individual immune and non-immune cell subsets. (D) Expression of the Hurwitz-EVsig in PBL, CD45(+) and CD45(−) cells. (E) Expression of the Hurwitz-EVsig by individual PBL subsets. (F) Expression of the Hurwitz-EVsig by individual immune and non-immune cell subsets. (G) Comparison of Fathi-EVsig gene expression levels in PBL and TILs. (H) Comparison of Hurwitz-EVsig gene expression levels in PBL and TILs. All values represent means ± SEM; **P < 0.01 versus PBL; ***P < 0.001 versus PBL; ****P < 0.0001 versus PBL

Associations of vesiculation-related genes with clinicopathologic data of HNSCC patients

To evaluate the clinical relevance of vesiculation-related genes, individual expression patterns were compared between patients with distinct clinical characteristics such as: tumor stage, tumor grade, perineural or lymphovascular invasion, remission/progression of disease, treatment approach (radiation or molecular targeted therapy), and habitual risk factors (alcohol or tobacco abuse). We were not able to detect significantly different values for these parameters; however, several trends were observable. Progressive disease stage was associated with decreased expression levels of vesiculation-related genes, especially for the Hurwitz-EVsig in stage III-IV patients and for the Fathi-EVsig in stage IVB and IVC patients (Fig. 3A). Levels of Fathi-EVsig were upregulated in patients with grade G4 tumors in comparison to G1, G2, and G3 tumors (Fig. 3B). For Hurwitz-EVsig, the gene expression levels decreased with higher grade of the tumors (Fig. 3B). The expression levels of Fathi-EVsig were similar in patients treated with or without radiation therapy or targeted molecular therapy (Fig. 3C and D). However, patients treated with radiation therapy or targeted molecular therapy showed downregulated gene expression levels of the Hurwitz-EVsig (Fig. 3C and D). Lymphovascular and perineural invasion were not associated with altered gene expression levels of Fathi-EVsig; however, expression levels of Hurwitz-EVsig were increased in patients with lymphovascular and perineural invasion (Fig. 3E and F). Alcohol or tobacco abuse, which are typical risk factors for HNSCC, was not associated with altered gene expression levels of both signatures (Fig. 3G and H). Gene expression levels of Fathi-EVsig were upregulated in patients showing partial remission/response and patients with persistent or stable disease (Fig. 3I). Gene expression levels of Hurwitz-EVsig were upregulated in patients showing partial remission/response, persistent disease, progressive disease, and stable disease in comparison to patients with complete remission/response (Fig. 3I).

Figure 3:

Figure 3:

Open in a new tab

Correlation of Fathi-EVsig and Hurwitz-EVsig relative gene expression levels with clinical parameters of HNSCC patients. Clinical stage (A), neoplasm histologic grade (B), radiation therapy (C), targeted molecular therapy (D), lymphovascular invasion (E), perineural invasion (F), alcohol history (G), smokeless tobacco use (H), and clinical course (I). All values represent means ± SEM

mRNA levels of vesiculation-related genes positively correlate with malignant gene expression levels in HNSCC

To further characterize the molecular phenotype which is associated with the two EV gene signatures, correlations between gene expression levels of Fathi-EVsig and Hurwitz-EVsig with pro-malignant genes were analyzed. The analysis focused on hallmarks of HNSCC progression [1] and included HIF1A as a gene representing hypoxic conditions (Fig. 4A and F); the HALLMARK_ANGIOGENESIS gene set, representing tumor angiogenesis (Fig. 4B and G); MMP2, MMP9, and MMP13 to represent matrix metalloproteinases (Fig. 4C and H); CD44, PROM1, and ALDH1A1 to represent cancer stem cell properties (Fig. 4D and I); and TWIST1, SNAI1, and SNAI2 to represent epithelial to mesenchymal transition (Fig. 4E and J). Fathi-EVsig and Hurwitz-EVsig both significantly correlated with all the above-mentioned malignant gene expression levels in HNSCC (P < 0.001; Fig. 4).

Figure 4:

Figure 4:

Open in a new tab

Correlation of Fathi-EVsig expression levels with HIF1A (A), HALLMARK_ANGIOGENESIS (GSEA, Molecular Signatures Database M5944) (B), MMP2, MMP9, and MMP13 (C), CD44, PROM1, and ALDH1A1 (D), TWIST1, SNAI1, and SNAI2 (E). Correlation of Hurwitz-EVsig expression levels with HIF1A (F), HALLMARK_ANGIOGENESIS (GSEA, Molecular Signatures Database M5944) (G), MMP2, MMP9, and MMP13 (H), CD44, PROM1, and ALDH1A1 (I), TWIST1, SNAI1, and SNAI2 (J). Correlation analysis (descriptive and inferential statistics) was performed using GEPIA2

Correlation of vesiculation-related genes with immunomodulatory factors in HNSCC tissues

To evaluate the impact of vesiculation-related genes on the immune phenotype of HNSCC patients, gene expression levels of Fathi-EVsig and Hurwitz-EVsig were correlated with expression levels of genes known to play an immune-suppressive or immune-stimulatory role as well as with chemokines. While the Fathi-EVsig negatively correlated with most immunostimulatory factors (Fig. 5C), there was a highly significant positive correlation with NT5E (encoding for CD73) and TGFB1 (P < 0.0001; Fig. 5A), suggesting that tumors with elevated expression levels of vesiculation-related genes have a potential impact on the immune landscape by adenosinergic and TGFβ signaling. There was also a significantly positive correlation with CD274 (encoding for PD-L1) further indicating that vesiculation-related genes are associated with an immunosuppressive TME (P < 0.001; Fig. 5A). Interestingly, these results were identical when looking at the Hurwitz-EVsig, with mostly negative correlations with immunostimulatory factors (Fig. 5D) and highly significant correlations with NT5E, TGFB1, and CD274 (P < 0.0001; Fig. 5B). These results indicate that the EV genetic profiles associate with a distinct immunosuppressive profile in HNSCC. Among chemokines, which exert a double-edged sword in cancer by their pro- or anti-tumor effects, both signatures correlated mostly with chemoattractant factors associated with tumor inflammation, reprogramming, EMT, and neovascularization. The most significant correlations were CXCL8 in Fathi-EVsig, CXCL11 in Hurwitz-EVsig, and CCL27 in both (P < 0.001; Fig. 5E and F). CXCL8 is known to be a neutrophil chemoattractant [25], CXCL11 is considered as an inducer of PD-L1 expression [26], and CCL27 is considered as a mediator of lymphatic endothelial cell migration during tumor-associated lymphangiogenesis [27]. CXCL14, also highly correlating with both EVsig datasets (P < 0.001; Fig. 5E and F), is known to be associated with reduced tumor growth and increased tumor infiltrating lymphocytes [28]. These results were validated in the non-TCGA bulk sequencing cohort and the overall correlation profiles of both EV signatures with immunosuppressive and immunostimulatory factors as well as with chemokines were found to be congruent. TGFB1, NT5E, and CD274 were among the factors correlating best with both EV signatures (Fig. S1A and B). The same was observed for the chemokines CXCL14, CXCL11, CCL27, and CXCL8 (Fig. S1E and F).

Figure 5:

Figure 5:

Open in a new tab

Correlation of vesiculation-related genes with immunomodulatory factors in HNSCC tissues. (A) Correlation between immunoinhibitory factors and the Fathi-EVsig. (B) Correlation between immunoinhibitory factors and the Hurwitz-EVsig. (C) Correlation between immunostimulatory factors and the Fathi-EVsig. (D) Correlation between immunostimulatory factors and the Hurwitz-EVsig. (E) Correlation between chemokines and the Fathi-EVsig. (F) Correlation between chemokines and the Hurwitz-EVsig. All values represent means ± SEM

mRNA levels of vesiculation-related genes were associated with a characteristic landscape of infiltrating immune cells

High gene expression levels of Fathi-EVsig correlated with elevated abundance of tumor-infiltrating CD4(+) T cells, macrophages, neutrophils, and dendritic cells into tumor tissues (P < 0.01; Fig. 6A). Elevated expression of Fathi-EVsig was also associated with an exclusion of B cells and CD8(+) T cells in HNSCC tissues (P < 0.01; Fig. 6A). Interestingly, the same correlations with immune cell abundance were also observed for the expression levels of Hurwitz-EVsig confirming our analysis of the Fathi-EVsig. Thus, elevated expression levels of Hurwitz-EVsig were associated with increased abundance of tumor-infiltrating CD4(+) T cells, macrophages, neutrophils, and dendritic cells (P < 0.01) and exclusion of B cells and CD8(+) T cells (P < 0.01; Fig. 6B). In the non-TCGA bulk sequencing validation cohort, we also observed an exclusion of B cells and CD8(+) T cells and increased infiltration abundance of neutrophils and dendritic cells. However, in contrast to the TCGA cohort the infiltration abundance of CD4(+) T cells and macrophages negatively correlated with expression levels of Fathi-EVsig and Hurwitz-EVsig (Fig. S2). In aggregate, these results indicate that the increased expression of vesiculation-related genes in the tumor is associated with a characteristic landscape of tumor-infiltrating immune cells, where immune effector cells are depleted.

Figure 6:

Figure 6:

Open in a new tab

Correlation of vesiculation-related genes with immune cell infiltration. (A) Heat map visualizing the correlation of the expression levels of the Fathi-EVsig and the immune cell infiltration levels of B cells, CD8(+) T cells, CD4(+) T cells, macrophages, neutrophils, and dendritic cells. Each cell represents the Spearman’s rho value of each gene of the Fathi-EVsig. (B) Heat map visualizing the correlation of the expression levels of the Hurwitz-EVsig and the immune cell infiltration levels of B cells, CD8(+) T cells, CD4(+) T cells, macrophages, neutrophils, and dendritic cells. Each cell represents the Spearman’s rho value of each gene of the Hurwitz-EVsig

Expression levels of vesiculation-related genes in HPV(+) and HPV(−) HNSCCs translate into distinct immunological profiles

HPV(+) and HPV(−) HNSCCs are characterized by the distinct composition of immune cells infiltrating in the tumor tissues [29]. HPV(+) tumors display much more strongly immune infiltration compared with HPV(−) tumors, with higher levels of T cell infiltration and CD8(+) T cell activation [30]. Therefore, the HNSCC cohorts were subdivided into HPV(+) and HPV(−) patients, and the gene expression levels of vesiculation-related genes and their impact on immune cell infiltration were analyzed separately. No differences were observed for the expression levels of Fathi-EVsig and Hurwitz-EVsig with regard to HPV status in the TCGA as well as in the validation bulk sequencing cohort (Figs. 7A; Supplementary S3A), and overall expression levels of vesiculation-related genes were associated with a similar immunological profile in HPV(+) and HPV(−) patients (Fig. 7B and C; Supplementary Fig. S3B and C). This profile was characterized by increased infiltration of CD4(+) T cells, macrophages, neutrophils, and dendritic cells and the exclusion of B cells and CD8(+) T cells (Fig. 7B and C). However, high gene expression levels of Fathi-EVsig in HPV(+) tumors correlated with an increased abundance of Tregs and a decreased abundance of CAFs and MDSCs when compared to HPV(−) tumors (P < 0.05; Fig. 7B). These significant differences were not detected in the non-TCGA bulk sequencing validation cohort, although we observed a similar trend for Tregs and CAFs (Fig. S3B). High gene expression levels of Hurwitz-EVsig in HPV(+) tumors correlated with a decreased abundance of CD4(+) T cells when compared to HPV(−) tumors (P < 0.05; Fig. 7C). The significant correlation of gene expression levels of the Hurwitz-EVsig with decreased abundance of CD4(+) T cells in HPV(+) tumors was also observed in the validation cohort (P < 0.05; Fig. S3C). Additionally, the levels of Hurwitz-EVsig expression in the validation cohort significantly correlated with decreased infiltration abundance of B cells and dendritic cells and increased abundance of CAFs (P < 0.05; Supplementary Fig. S3C). These results indicate that the landscape of infiltrating immune cells which is associated with vesiculation-related genes differs depending on the HPV status of the tumors.

Figure 7:

Figure 7:

Open in a new tab

Comparison of vesiculation-related genes in HPV(+) and HPV(−) HNSCCs. (A) Comparison of Fathi-EVsig and Hurwitz-EVsig gene expression in HPV(+) and HPV(−) HNSCC patients. (B) Infiltration abundance of immune cells based on Fathi-EVsig expression in HPV(+) and HPV(−) patients. (C) Infiltration abundance of immune cells based on Hurwitz-EVsig expression in HPV(+) and HPV(−) patients. (D) Fathi-EVsig score in non-immune cells of individual HPV(+) and HPV(−) patients. (E) Hurwitz-EVsig score in non-immune cells of individual HPV(+) and HPV(−) patients. (F) Fathi-EVsig score in individual PBL, TIL, and non-immune cell subsets in HPV(+) and HPV(−) patients. (G) Hurwitz-EVsig score in individual PBL, TIL, and non-immune cell subsets in HPV(+) and HPV(−) patients. Values in F and G represent means ± SEM

To further define the expression patterns of the two EV signatures in HPV(+) and HPV(−) HNSCC, we separated the single-cell sequencing cohort based on the patients viral status into six HPV(+) and nine HPV(−) patients. While the gene expression levels appeared similar in the bulk sequencing cohorts (Figs. 7A; Supplementary S3A), clustering of the patients into EVsiglow and EVsighigh patients based on expression in only non-immune cells revealed that most HPV(+) patients had increased levels of Fathi-EVsig (Fig. 7D). Hurwitz-EVsig gene expression levels almost separated HPV(+) from HPV(−) patients and were elevated in HPV(−) patients (Fig. 7E). This also reflected in increased or decreased expression in HPV(+) epithelial cells of Fathi-EVsig and Hurwitz-EVsig, respectively (P < 0.05; Fig. 7F and G). Significantly higher levels of Hurwitz-EVsig were also found in HPV(−) fibroblasts compared to their HPV(+) counterparts (P < 0.05; Fig. 7G). Interestingly, the expression patterns in immune cells were distinct in HPV(+) and HPV(−) HNSCCs and both EV signatures were expressed significantly higher in HPV(−) B cells, CD4(+) T cells, plasmacytoid dendritic cells, and Tregs compared to their HPV(+) counterparts (P < 0.05; Fig. 7F and G). While the PBLs of HPV(+) and HPV(−) HNSCCs expressed similar levels of Fathi-EVsig (Fig. 7F), CD4(+) T cells, CD8(+) T cells, and monocytes expressed significantly higher levels of Hurwitz-EVsig in HPV(+) HNSCC (P < 0.05; Fig. 7G).

Discussion

The progression of HNSCC is dependent on the formation of an immunosuppressive niche which favors tumor growth and metastasis. Understanding the mechanisms underlying the initiation and maintenance of this immunosuppressive microenvironment is of great current interest. Attention has been centered on communication between malignant and non-malignant cells in the TME, which leads to phenotypic and functional reprogramming of tumor resident and infiltrating cells and ultimately promotes disease progression. Tumor-derived EVs are reported to contribute to intercellular communication locally in the tumor and at distant sites [31]. We and others have demonstrated that HNSCC-derived EVs carry a cargo of biologically active components which promote tumor progression [7–10]. Although HNSCC-derived EVs are implicated in mediating cancer progression, their actual contribution to the formation of the tumor-promoting TME remains elusive. Here, we first searched for gene datasets related to EV production by tumor cells, and then used the available data for comparison of HNSCCs with high and low production rates of EVs.

Our bioinformatics analysis demonstrated that increased expression levels of vesiculation-related genes correlated with a profile of TGFβ and adenosinergic signaling in the tumor. Both signaling pathways are known to have major immunosuppressive effects in HNSCC and promote tumor progression [32–34]. The two main pathway components, TGFβ and CD73, were previously identified as functionally active cargo components in HNSCC-derived EVs [10, 35]. Recent literature emphasizes the role of EVs as carriers and producers of adenosine [36–38]. Our bioinformatics analysis based on gene expression levels of the adenosinergic pathway components has confirmed that this pathway is one of the major immunosuppressive mechanisms mediated by HNSCC-derived EVs. Our data demonstrated that—along with a TGFβ/CD73 profile—a collection of cytokine genes correlated with the expression of vesiculation-related genes in HNSCC patients. This observation supports the presence of a tumor-promoting environment in the TME. Expression levels of the vesiculation-related genes in the datasets we used, showed significant correlation with IL-8 (P < 0.0001), responsible for immune cell recruitment [39] and with IL-6 and IL-10 (P < 0.01), which mediate pro-tumor functions of neutrophils and macrophages [40, 41]. Concerning immunosuppressive function, IL-10 impairs CD8(+) T cells [42], which is consistent with our findings, since abundance of infiltrating CD8(+) T cells in tumor tissues was negatively associated with expression levels of the vesiculation-related genes in both EV signatures. Moreover, both EV signatures have positive correlation with immune checkpoints, such as CD274, CD276, PVR, and c10orf54 (Fig. 5A–D), which have become a hallmark of the immunosuppressive TME [43].

Tumor-derived EVs carry both immunosuppressive and immunostimulatory receptor/ligands and thus may be involved in various immunomodulatory pathways [44]. The cargo of tumor-derived EVs contains costimulatory molecules, major histocompatibility complex class I and class II molecules, tumor-associated antigens, and intraluminal growth-promoting cytokines in addition to a plethora of immunoinhibitory molecules [3, 45]. Analogous with these reports, we demonstrated that the EV signatures correlate with the expression levels of a variety of immunostimulatory factors as well as tumor-suppressive factors, such as CXCL14. Although this seems counterintuitive when looking at the immunosuppressive effects of tumor-derived EVs in the TME, it appears that stimulatory and inhibitory signals of tumor-derived EVs are delivered simultaneously. It remains unclear how this mode of delivery of multiple signals is able to stimulate or inhibit immune responses translates into specific functional alterations in recipient cells [3].

It is under current investigation, whether tumor-derived EVs play a distinct role in HPV(+) and HPV(−) HNSCC. Recent findings indicate that EVs from HPV(+) and HPV(−) tumors have differential protein contents that might translate to distinct functional effects and, thus, contribute to the disparity in immune responses that characterize HPV(+) and HPV(−) HNSCCs [46]. In direct comparison of EVs isolated from HPV(+) and HPV(−) HNSCCs, HPV(+) EVs were capable of sustaining dendritic cell functions and, therefore, may play a role in promoting anti-tumor immune responses in patients with HPV(+) HNSCC [11]. Our bioinformatics analysis validated these findings since we also observed distinct immune landscapes depending on the HPV status of the primary tumors. In HPV(+) tumors vesiculation-related genes correlated with an increased abundance of Tregs and a decreased abundance of CD4(+) T cells, CAFs, and MDSCs. These findings need further validation in clinical tumor samples as well as functional in vitro and in vivo studies and might accelerate the understanding of the differences in the immune landscape of viral- and carcinogen-driven HNSCC.

One of the current challenges in the field is the heterogeneity of EV subpopulations that tumors produce. Various subtypes of tumor-derived EVs may be released by the tumor, carrying different cargos and mediating different functions in recipient cells [47]. Capturing these subtypes selectively using protein markers or gene signatures remains a challenge. Therefore, the two gene signatures we used in this study may not reflect the complexity of EV biogenesis and their heterogeneity. Also, these gene signatures need to be validated in vivo by parallel analysis of tumor gene expression levels and quantification of local and circulating tumor-derived EVs. However, although the Fathi-EVsig and Hurwitz-EVsig share only two common genes and are, therefore, mostly independent signatures, they gave similar results, suggesting that the specific shaping of the immune landscape may be an overall mechanism which can be attributed to tumor-derived EVs.

Conclusions

The presented bioinformatic data indicate that tumor cell-derived EVs carrying a complex cargo of immunosuppressive proteins are likely to shape immune cell infiltration into the tumor by interaction with immune cells in the periphery or locally in the tumor. The biological effects of EVs—orchestrated by a variety of factors enriched in EVs—may shape and modulate the immune landscape in a very specific manner, with exclusion of effector lymphocytes and promotion of the abundant infiltration of CD4(+) T cells, macrophages, neutrophils, and dendritic cells. Also, gene signatures related to vesiculation are enriched in tumor cells and are potential immune biomarkers in HPV(+) and HPV(−) HNSCC. Our in silico results provide a baseline for future prospective studies to be conducted with a well-defined HNSCC patient cohort to compare expression levels of vesiculation-related genes with plasma levels of EVs, immunopathology of the tumor, and clinical parameters. Overall, our analysis confirmed a large volume of data on tumor-derived EVs as key components of the reprogramming that characterizes the TME.

Supplementary Material

uxad019_suppl_Supplementary_Material
Click here for additional data file. (463.5KB, docx)

Acknowledgements

N/A

Glossary

Abbreviations

CAF

cancer-associated fibroblast

CD

cluster of differentiation

EVs

extracellular vesicles

EVsig

vesiculation-related genes

GEPIA2

gene expression profiling interactive analysis2

HNSCC

head and neck squamous cell carcinoma

HPV

human papilloma virus

MDSC

myeloid-derived suppressor cell

OS

overall survival

PBL

peripheral blood lymphocytes

PD-L1

programmed death-ligand 1

SEM

standard error of the mean

TCGA

the cancer genome atlas

TGFβ

transforming growth factor β

TIL

tumor-infiltrating lymphocytes

TIMER

tumor immune estimation resource

TISIDB

tumor–immune system interactions and drug bank database

TME

tumor microenvironment

Tregs

T regulatory cells

UCSC

University of California, Santa Cruz

Contributor Information

Isabella Kallinger, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Dominique S Rubenich, Programa de Pós-Graduação em Biociências, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Porto Alegre, RS, Brazil; Instituto de Cardiologia do Rio Grande do Sul/Fundação Universitária do Instituto de Cardiologia (IC-FUC), Porto Alegre, RS, Brazil.

Alicja Głuszko, Chair and Department of Biochemistry, Faculty of Medicine, Medical University of Warsaw, Warsaw, Poland.

Aditi Kulkarni, UPMC Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA, USA; Tumor Microenvironment Center, UPMC Hillman Cancer Center, Pittsburgh, PA, USA; Department of Otolaryngology, University of Pittsburgh, Pittsburgh, PA, USA.

Gerrit Spanier, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Steffen Spoerl, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Juergen Taxis, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Hendrik Poeck, Clinic and Polyclinic for Internal Medicine III, University Hospital Regensburg and Leibniz Institute for Immunotherapy (LIT), Regensburg, Germany.

Mirosław J Szczepański, Chair and Department of Biochemistry, Faculty of Medicine, Medical University of Warsaw, Warsaw, Poland.

Tobias Ettl, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Torsten E Reichert, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Johannes K Meier, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Elizandra Braganhol, Programa de Pós-Graduação em Biociências, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Porto Alegre, RS, Brazil; Instituto de Cardiologia do Rio Grande do Sul/Fundação Universitária do Instituto de Cardiologia (IC-FUC), Porto Alegre, RS, Brazil.

Robert L Ferris, UPMC Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA, USA; Tumor Microenvironment Center, UPMC Hillman Cancer Center, Pittsburgh, PA, USA; Department of Otolaryngology, University of Pittsburgh, Pittsburgh, PA, USA; Department of Immunology, University of Pittsburgh, Pittsburgh, PA, USA.

Theresa L Whiteside, UPMC Hillman Cancer Center, University of Pittsburgh, Pittsburgh, PA, USA; Department of Otolaryngology, University of Pittsburgh, Pittsburgh, PA, USA; Department of Immunology, University of Pittsburgh, Pittsburgh, PA, USA; Department of Pathology, University of Pittsburgh, Pittsburgh, PA, USA.

Nils Ludwig, Department of Oral and Maxillofacial Surgery, University Hospital Regensburg, Regensburg, Germany.

Funding

I.K. was supported by the Verein zur Förderung der wissenschaftlichen Zahnheilkunde (VFwZ). D.S.R. was supported by the German Academic Exchange Service (DAAD) research grant #57588369. A.G. was supported by Medical University of Warsaw MB/M/48(79)#. M.J.S. was supported by National Science Centre, Poland UMO-2017/26/M/NZ5/00877#. T.L.W. was partially supported by grants U01-DE-029759 and R01-CA256068 from the National Institute of Health, USA. N.L. was supported by the Walter Schulz Foundation.

Conflict of interests

The authors declare that they have no conflict of interest.

Author contributions

Conceptualization: N.L.; Data curation: I.K., D.S.R., A.G., A.K., N.L.; Formal Analysis: I.K., D.S.R., A.G., A.K., N.L.; Funding acquisition: I.K., D.S.R., A.G., M.J.S., T.L.W., N.L., T.E., T.E.R.; Investigation: I.K., D.S.R., A.G., A.K., N.L.; Supervision: T.E.R., E.B., M.J.S., J.M., R.L.F., T.L.W., N.L.; Visualization: I.K., D.S.R., A.G., N.L.; Writing—original draft: I.K., D.S.R., A.G., N.L.; Writing—review and editing: G.S., S.S., J.T., H.P., M.J.S., T.E., T.E.R., J.M., E.B., R.L.F., T.L.W., N.L.

The animal research adheres to the ARRIVE guidelines (https://arriveguidelines.org/arrive-guidelines)

N/A

Permission to reproduce (for relevant content)

N/A

Clinical trial registration

N/A

Data Availability

The data presented in this study are available on request from the corresponding author.

References

  • 1. Johnson DE, Burtness B, Leemans CR, Lui VWY, Bauman JE, Grandis JR. Head and neck squamous cell carcinoma. Nat Rev Dis Prim 2020, 6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Coordes A, Lenz K, Qian X, Lenarz M, Kaufmann A, Albers A. Meta-analysis of survival in patients with HNSCC discriminates risk depending on combined HPV and p16 status. Eur Arch Oto-Rhino-Laryngol 2015. [DOI] [PubMed] [Google Scholar]
  • 3. Whiteside T. The effect of tumor-derived exosomes on immune regulation and cancer immunotherapy. Futur Oncol 2017, 13, 2583–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Ko SY, Lee WJ, Kenny HA, Dang LH, Ellis LM, Jonasch E, et al. Cancer-derived small extracellular vesicles promote angiogenesis by heparin-bound, bevacizumab-insensitive VEGF, independent of vesicle uptake. Commun Biol 2019, 2, 1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Mashouri L, Yousefi H, Aref AR, Ahadi AM, Molaei F, Alahari SK. Exosomes: composition, biogenesis, and mechanisms in cancer metastasis and drug resistance. Mol Cancer 2019, 18, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Ludwig N, Gillespie D, Reichert T, Jackson E, Whiteside TL. Purine metabolites in tumor-derived exosomes may facilitate immune escape of head and neck squamous cell carcinoma. Cancers 2020, 12, 1602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Sato S, Vasaikar S, Eskaros A, Kim Y, Lewis JS, Zhang B, et al. EPHB2 carried on small extracellular vesicles induces tumor angiogenesis via activation of ephrin reverse signaling. JCI Insight 2019, 4, e132447. doi: 10.1172/jci.insight.132447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Ludwig N, Yerneni SS, Razzo BM, Whiteside TL.. Exosomes from HNSCC promote angiogenesis through reprogramming of endothelial cells. Mol Cancer Res 2018, 16, 1798–1808. doi: 10.1158/1541-7786.MCR-18-0358. [DOI] [PubMed] [Google Scholar]
  • 9. Ono K, Sogawa C, Kawai H, Tran MT, Taha EA, Lu Y, et al. Triple knockdown of CDC37, HSP90-alpha and HSP90-beta diminishes extracellular vesicles-driven malignancy events and macrophage M2 polarization in oral cancer. J Extracell Vesicles 2020, 9, 2–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Ludwig N, Yerneni S, Azambuja J, Gillespie D, Menshikova E, Jackson E, et al. Tumor-derived exosomes promote angiogenesis via adenosine A2B receptor signaling. Angiogenesis 2020, 23, 599–610. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Ludwig S, Sharma P, Theodoraki M-N, Pietrowska M, Yerneni SS, Lang S, et al. Molecular and functional profiles of exosomes from HPV(+) and HPV(−) head and neck cancer cell lines. Front Oncol 2018, 12, 445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Razzo BM, Ludwig N, Hong C, Sharma P, Fabian KP, Fecek RJ, et al. Tumor-derived exosomes promote carcinogenesis of murine oral squamous cell carcinoma. Carcinogenesis 2019, 1, 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Ludwig S, Floros T, Theodoraki M-N, Hong C-S, Jackson EK, Lang S, et al. Suppression of lymphocyte functions by plasma exosomes correlates with disease activity in patients with head and neck cancer. Clin Cancer Res 2017, 23, 4843–54. doi: 10.1158/1078-0432.CCR-16-2819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Schuler PJ, Saze Z, Hong C-S, Muller L, Gillespie DG, Cheng D, et al. Human CD4+ CD39+ regulatory T cells produce adenosine upon co-expression of surface CD73 or contact with CD73+ exosomes or CD73+ cells. Clin Exp Immunol 2014, 177, 531–43. doi: 10.1111/cei.12354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Madeo M, Colbert PL, Vermeer DW, Lucido CT, Cain JT, Vichaya EG, et al. Cancer exosomes induce tumor innervation. Nat Commun 2018, 9, 4284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Fathi M, Joseph R, Adolacion JRT, Martinez-Paniagua M, An X, Gabrusiewicz K, et al. Single-cell cloning of breast cancer cells secreting specific subsets of extracellular vesicles. Cancers 2021, 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Hurwitz SN, Rider MA, Bundy JL, Liu X, Singh RK, Meckes DG.. Proteomic profiling of NCI-60 extracellular vesicles uncovers common protein cargo and cancer type-specific biomarkers. Oncotarget 2016, 7, 86999–7015. doi: 10.18632/oncotarget.13569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Goldman M, Craft B, Kamath A, Brooks A, Zhu J, Haussler D. The UCSC Xena platform for cancer genomics data visualization and interpretation. bioRxiv 2018, 326470. [Google Scholar]
  • 19. Cillo AR, Kürten CHL, Tabib T, Qi Z, Onkar S, Wang T, et al. Immune landscape of viral- and carcinogen-driven head and neck cancer. Immunity 2020, 52, 183-99.e9. doi: 10.1016/j.immuni.2019.11.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Kürten CHL, Kulkarni A, Cillo AR, Santos PM, Roble AK, Onkar S, et al. Investigating immune and non-immune cell interactions in head and neck tumors by single-cell RNA sequencing. Nat Commun 2021, 12, 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Wolf F, Angerer P, Scanpy TF.. large-scale single-cell gene expression data analysis. Genome Biol 2018, 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Tang Z, Kang B, Li C, Chen T, Zhang Z.. GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res 2019, 47, W556–60. doi: 10.1093/nar/gkz430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Ru B, Wong CN, Tong Y, Zhong JY, Zhong SSW, Wu WC, et al. TISIDB: an integrated repository portal for tumor-immune system interactions. Bioinformatics 2019, 35, 4200-2. doi: 10.1093/bioinformatics/btz210. [DOI] [PubMed] [Google Scholar]
  • 24. Li T, Fan J, Wang B, Traugh N, Chen Q, Liu JS, et al. TIMER: a web server for comprehensive analysis of tumor-infiltrating immune cells. Cancer Res 2017, 77, e108-10. doi: 10.1158/0008-5472.CAN-17-0307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Asokan S, Bandapalli OR.. CXCL8 signaling in the tumor microenvironment. Adv Exp Med Biol 2021, 1302, 25–39. doi: 10.1007/978-3-030-62658-7_3. [DOI] [PubMed] [Google Scholar]
  • 26. Puchert M, Obst J, Koch C, Zieger K, Engele J.. CXCL11 promotes tumor progression by the biased use of the chemokine receptors CXCR3 and CXCR7. Cytokine 2020, 125, 154809. doi: 10.1016/j.cyto.2019.154809. [DOI] [PubMed] [Google Scholar]
  • 27. Karnezis T, Farnsworth RH, Harris NC, Williams SP, Caesar C, Byrne DJ, et al. CCL27/CCL28-CCR10 chemokine signaling mediates migration of lymphatic endothelial cells. Cancer Res 2019, 79, 1558–72. doi: 10.1158/0008-5472.CAN-18-1858. [DOI] [PubMed] [Google Scholar]
  • 28. Parikh A, Shin J, Faquin W, Lin DT, Tirosh I, Sunwoo JB, et al. Malignant cell-specific CXCL14 promotes tumor lymphocyte infiltration in oral cavity squamous cell carcinoma. J ImmunoTher Cancer 2020, 8, e001048. doi: 10.1136/jitc-2020-001048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Zhou D, Wang J, Wang J, Liu X.. Profiles of immune cell infiltration and immune-related genes in the tumor microenvironment of HNSCC with or without HPV infection. Am J Transl Res 2021, 13, 2163–80. [PMC free article] [PubMed] [Google Scholar]
  • 30. Mandal R, Şenbabaoğlu Y, Desrichard A, Havel JJ, Dalin MG, Riaz N, et al. The head and neck cancer immune landscape and its immunotherapeutic implications. JCI Insight 2016, 1, 1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Ludwig N, Rao A, Sandlesh P, Yerneni SS, Swain AD, Bullock KM, et al. Characterization of systemic immunosuppression by IDH mutant glioma small extracellular vesicles. Neuro Oncol 2021;24:197–209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Deng WW, Li YC, Ma SR, Mao L, Yu G-T, Bu L-L, et al. Specific blockade CD73 alters the “exhausted” phenotype of T cells in head and neck squamous cell carcinoma. Int J Cancer 2018, 143, 1494–504. doi: 10.1002/ijc.31534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Ma SR, Deng WW, Liu JF, Mao L, Yu GT, Bu LL, et al. Blockade of adenosine A2A receptor enhances CD8+ T cells response and decreases regulatory T cells in head and neck squamous cell carcinoma. Mol Cancer 2017, 16, 1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Ludwig N, Wieteska L, Hinck CS, Yerneni SS, Azambuja JH, Bauer RJ, et al. Novel TGFβ inhibitors ameliorate oral squamous cell carcinoma progression and improve the antitumor immune response of Anti–PD-L1 immunotherapy. Mol Cancer Ther 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Jablonska J, Rist M, Spyra I, Tengler L, Domnich M, Kansy B, et al. Evaluation of immunoregulatory biomarkers on plasma small extracellular vesicles for disease progression and early therapeutic response in head and neck cancer. Cells 2022, 11, 902. doi: 10.3390/cells11050902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Ploeg EM, Ke X, Britsch I, Hendriks MAJM, Van der Zant FA, Kruijff S, et al. Bispecific antibody CD73xEpCAM selectively inhibits the adenosine-mediated immunosuppressive activity of carcinoma-derived extracellular vesicles. Cancer Lett 2021, 521, 109–18. doi: 10.1016/j.canlet.2021.08.037. [DOI] [PubMed] [Google Scholar]
  • 37. Wang M, Jia J, Cui Y, Peng Y, Jiang Y. CD73-positive extracellular vesicles promote glioblastoma immunosuppression by inhibiting T-cell clonal expansion. Cell Death Dis 2021, 12, 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Turiello R, Capone M, Morretta E, Monti MC, Madonna G, Azzaro R, et al. Exosomal CD73 from serum of patients with melanoma suppresses lymphocyte functions and is associated with therapy resistance to anti-PD-1 agents. J ImmunoTher Cancer 2022, 10, e0040431–11. doi: 10.1136/jitc-2021-004043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Gauglitz GG, Finnerty CC, Herndon DN, Mlcak RP, Jeschke MG. Are serum cytokines early predictors for the outcome of burn patients with inhalation injuries who do not survive? Crit Care 2008, 12, 1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Codony VL, Tavassoli M.. Hypoxia-induced therapy resistance: available hypoxia-targeting strategies and current advances in head and neck cancer. Transl Oncol 2021, 14, 101017. doi: 10.1016/j.tranon.2021.101017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Zhang X, Zhang W, Yuan X, Fu M, Qian H, Xu W.. Neutrophils in cancer development and progression: roles, mechanisms, and implications. Int J Oncol 2016, 49, 857–67. doi: 10.3892/ijo.2016.3616. [DOI] [PubMed] [Google Scholar]
  • 42. Wang C, Li Y, Jia L, Kim JK, Li J, Deng P, et al. CD276 expression enables squamous cell carcinoma stem cells to evade immune surveillance. Cell Stem Cell 2021, 28, 1597–613.e7. doi: 10.1016/j.stem.2021.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. He X, Xu C.. Immune checkpoint signaling and cancer immunotherapy. Cell Res 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Wu J, Li S, Zhang P.. Tumor-derived exosomes: immune properties and clinical application in lung cancer. Cancer Drug Resist 2022, 5, 102–13. doi: 10.20517/cdr.2021.99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Labani-Motlagh A, Naseri S, Wenthe J, Eriksson E, Loskog A. Systemic immunity upon local oncolytic virotherapy armed with immunostimulatory genes may be supported by tumor-derived exosomes. Mol Ther Oncolytics 2021, 20, 508–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Ludwig S, Marczak L, Sharma P, Abramowicz A, Gawin M, Widlak P, et al. Proteomes of exosomes from HPV(+) or HPV(-) head and neck cancer cells: differential enrichment in immunoregulatory proteins. Oncoimmunology 2019, 8, e15938081–11. doi: 10.1080/2162402x.2019.1593808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Couch Y, Buzàs EI, Vizio DD, Gho YS, Harrison P, Hill AF, et al. A brief history of nearly EV-erything—the rise and rise of extracellular vesicles. J Extracell Vesicles 2021, 10. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

uxad019_suppl_Supplementary_Material
Click here for additional data file. (463.5KB, docx)

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology

RESOURCES