Targeting IL-17alpha to promote anti-PD-1 therapy effect by screening the tumor immune microenvironment in a mouse oral carcinogenesis model

Abstract

BACKGROUND:

Resistance to PD-1 blocking agents is not uncommon, limiting their wide clinical success. Certain tumor-infiltrating immune cells (e.g., TILs/CTLs) have emerged as biomarkers of response, and absence of such immune cells contributes to resistance.

OBJECTIVE:

We deconvoluted the dynamic immune microenvironment in a mouse model of oral carcinogenesis for augmenting the resistance to PD-1 blocking agents by combination.

METHODS:

Bioinformatics methods and routine biological experiments were adopted such as morphological analysis and ELISA in the 4NQO-treated mice model.

RESULTS:

Our findings revealed that dysplastic tongue tissues from 4NQO-treated mice were characterized by an immunosuppressive tumor microenvironment. Tongue tissues from mice treated with 4NQO for 12 weeks had higher levels of Th2 cells and Tregs compared to tissues taken from control mice or mice treated with 4NQO for 28 weeks; these results suggested a potential therapeutic benefit of anti-PD-1 in the oral cancer. The IL-17 pathway was significantly upregulated during progression from normal mucosa to hyperplasia and tumor formation in mice. Inhibition of IL-17 $\alpha$ combined with PD-1 blockade delayed the development of 4NQO-induced precancerous and cancerous lesions and prolonged the survival of 4NQO-treated mice.

CONCLUSIONS:

Our data suggested a strong rationale of IL-17 $\alpha$ blockade as a potential approach to augment the tumor-eliminating effects of anti-PD-1 therapy.

Keywords

Immunotherapy programmed death receptor 1 IL-17alpha resistance pre-cancerous lesions squamous cell carcinomas

Abbreviations

PD-1	Programmed cell death protein 1
PD-L1	Programmed death-ligand 1
CTLA-4	Cytotoxic T-lymphocyte-associated protein 4
OSCC	Oral squamous cell carcinoma
4NQO	4-nitroquinoline-1 oxide
ICIs	Immune checkpoint blockade inhibitors
IL-17 $\alpha$	Interleukin-17 $\alpha$
IL-19	Interleukin-19
IL-1	Interleukin-1 $\alpha$
IL-15	Interleukin-15
IL-22	Interleukin-22
CCL12	Chemokine (C-C motif) ligand 12
CCL28	Chemokine (C-C motif) ligand 28
CCR2	CC chemokine receptor 2
CCR5	CC chemokine receptor 5
SsGSEA	Single sample Gene Set Enrichment Analysis
Th17	T helper cell 17
Treg	Regulatory T cells
ELISA	Enzyme linked immunosorbent assay

1. Introduction

Cancer immunotherapies targeting immune checkpoint molecules, such as the cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed cell death protein 1 (PD-1), and programmed death-ligand 1 (PD-L1), have emerged as a promising therapeutic approach in several advanced cancers [1]. The blockade of immune checkpoint signaling pathways re-activates tumor-reactive T cells, leading to anti-tumor immune responses in the tumor microenvironment and thereby inhibiting tumor growth [2, 3].

Oral squamous cell carcinoma (OSCC) is an aggressive malignant neoplasm, with an overall survival rate of 50% that has remained unchanged for decades [4]. It has been previously shown that blockade of immune checkpoint receptors in mouse OSCC models is an effective cancer prevention strategy that reverses the carcinogenesis process prior to malignant transformation by blockade of immune checkpoint receptors in mouse model [5, 6].

However, primary resistance to anti-PD-1 therapy is not only observed in human tumors, resulting in tumor relapse and poor prognosis [7, 8, 9, 10], but also in precancerous lesions during progression to OSCC in mouse models [11, 12]. For example, the non-response rate was reported to be 30.43% in a 4NQO-induced mouse oral squamous carcinoma model after administration of anti-PD-1 antibodies, and this was associated with increased levels of regulatory T cells (Treg), which promoted resistance of the anti-PD-1 therapy.

Monotherapies are unlikely to overcome the major mechanisms that impede antitumor immunity in patients because the induction, potency, and persistence of host immune responses reflect the complex interplay between the different immune-cell populations and the progressing tumor [8]. Indeed, the generation of durable and clinically effective antitumor immune responses is probably dependent on the successful reprogramming of several immune processes. Herein, we employed the 4-nitroquinoline-1-oxide (4NQO)-induced carcinogenesis mouse model to unravel the changing immune microenvironment in different pathological stages of oral cancer, as well as to identify key signaling pathways as therapeutic targets to reverse immunosuppression. We achieved this by performing morphological, biological, and bioinformatics analyses in our in vivo mouse model. Additionally, the therapeutic potential of targeted inhibition of identified therapeutic targets in combination with anti-PD-1 blockade was investigated in our mouse tongue cancer model. We hypothesized that the identification of key players of resistance to immune checkpoint blockade would bring us one step forward to overcoming resistance to single-agent PD-1 therapy, expanding the application scope of immune checkpoint blockade inhibitors (ICIs), and ultimately achieving improved response rates and prolonged survival in patients with oral cancer.

2. Materials and methods

2.1 Mouse specimens

The animal protocol was approved by the Animal Care and Use Committee of the First Affiliated Hospital Harbin Medical University. BALB/c mice were used for morphological analysis, ELISA, and survival analysis. For bioinformatic analysis BALB/c mice were used in DATA1 and C57BL/6 mice were used in DATA2. 4NQO (Sigma-Aldrich) was administered to the mice in drinking water at the dose of 50 mg/L for 16 weeks, as described elsewhere [13]. The mice were randomly allocated into control IgG2a group, PBS group, anti-PD-1-treated group, anti-IL-17 $\alpha$ group, and anti-PD-1/anti-IL-17 $\alpha$ combination group. Monoclonal anti-mouse PD-1 (RMP1–14, #BE0146, BioXCell, West Lebanon, NH, USA), IgG2a isotype control (#BE0089, BioXCell, West Lebanon, NH, USA) and monoclonal anti-mouse IL-17 $\alpha$ (#MAB421, R&D Systems Minneapolis, MN, USA) were dissolved in PBS and stored at 4 ${}^{\circ}$ C until used. Mice were injected intraperitoneally with anti-PD-1 (200 $\mu$ g/mouse), anti-IL-17 $\alpha$ (200 $\mu$ g/mouse), combined anti-PD-1 (100 $\mu$ g/mouse) and anti-IL-17 $\alpha$ (200 $\mu$ g/mouse), PBS or IgG2a (200 $\mu$ g/mouse), once per week for a total of 4 weeks. All animals underwent a full oral cavity examination twice per week and were euthanized for tissue retrieval five weeks after initiation of the antibody administration.

2.2 Tissue dissection and sectioning for macroscopic and microcosmic analysis

To harvest oral tissue from above four groups, the mice were sacrificed. The tongues were excised and macroscopic lesions were counted after death. Quantified macroscopic lesions comprised those lesions that were visible, with a diameter of $>$ 1–2 mm, and presented with a whitish and papillary appearance. 4- $\mu$ m histological sections from each specimen were cut and the 10th serial slide was stained with hematoxylin and eosin (H&E) for histopathologic analysis. H&E sections were screened under the microscope for the presence of microscopic lesions. Individual lesions were quantified along the entire longitudinal section of the tongue [14].

2.3 Histological staining

Sections of tongues from 4NQO-treated mice were subjected to H&E to identify histologic changes associated with oral tumorigenesis, including normal mucosa in untreated mice tongue, and hyperplasia, dysplasia, and tumor. Formalin-fixed, paraffin-embedded tongue tissues were sectioned (4 $\mu$ m), placed in series on silanized glass slides, dewaxed, and rehydrated prior to immunofluorescence staining. Antigens were retrieved in EDTA buffer (pH 9.0) for 15 min in a pressure cooker. After rinsing with phosphate-buffered saline (PBS), the sections were immersed in 3% hydrogen peroxide solution for 10 min to block endogenous peroxidases. Non-specific binding was prevented by incubation in 5% normal goat serum for 20 min in a humidified chamber. For double staining of CD4, FOXP3, CD45, or ST2 antibody (1:200 dilution), samples were incubated at 4 ${}^{\circ}$ C overnight with the primary antibodies, washed, and incubated with goat anti-mouse IgG/FITC (catalog #bs-0296G-FITC, Biosynthesis Biotechnology) and goat anti-rabbit IgG/Cy3 (catalog #bs-0295G-Cy3, Biosynthesis Biotechnology) at 37 ${}^{\circ}$ C for 30 min. Rabbit polyclonal CD4 (catalog #bs-0647R, Biosynthesis Biotechnology); mouse monoclonal FOXP3 (catalog #sc-53876, Santa Cruz); rabbit polyclonal CD45 (catalog #bs-0522R, Biosynthesis Biotechnology); mouse monoclonal ST2 (catalog #60112-1-1g, Proteintech, Wuhan, China). Images were captured with a fluorescence microscope (Olympus BX51, Japan) equipped with Flash Point software.

2.4 Computational Immunophenotyping

The total immune infiltrates in tongue samples from 4NQO-treated mice were analyzed. The deconvolution was achieved using CIBERSORT Jar version 1.05 (https://cibersort.stanford.edu/) with the standard LM22 signature gene file, and 1000 permutations to calculate deconvolution $p$ -values.

2.5 Single-sample gene set enrichment analysis (ssGSEA)

Single-sample GSEA, a modified analysis of standard GSEA, was performed on RNA measurements for each sample using the GSVA package in R version 3.6.1. In order to identify significantly up and downregulated KEGG pathways, the $p$ -value was calculated for each pathway based on Student’s $t$ -test.

2.6 ELISA

IL-17 $\alpha$ , IL-1 $\alpha$ , and IL-22 concentrations of conditioned medium were assayed using ELISA kits (Catalog #M17AF0, MLA00, R&D systems, and Catalog #ab223857, Abcam), as described by the manufacturer. Optical density was measured at 450 nm.

2.7 Statistical analysis

All comparisons of the differences between two groups were performed using Student’s $t$ -test and Tukey’s multiple comparison test if more than two groups were compared. Survival analyses were conducted using the Kaplan-Meier method and compared using the Log-Rank test. All statistical tests were two-sided, and $p$ -values $\leqslant$ 0.05 were considered to be statistically significant. The statistical analysis was performed using GraphPad Prism version 6.00 (San Diego, SA), R version 3.6.1 or SPSS 20.0 (IBM Company, New York).

Figure 1.

Dynamic immune infiltration status during oral tumorigenesis in the 4NQO mouse model (A) Histologic changes during oral carcinogenesis in the mouse model a. Normal mucosa; b. Hyperplasia; c. Mild dysplasia; d. Moderate dysplasia; e. Severe dysplasia (carcinoma in situ); f. Invasive carcinoma. (B–C) Fractions of immune cell subsets in tongue tissue of BALB/c and C57BL/6 mice treated with 4NQO inferred from gene-expression data using CIBERSORT ${}^{13}$ . The width of bars is proportional to the -log10 $p$ -value of the deconvolution (Table S1 and Table S2). CIBERSORT empirical $p$ -value; * $p<$ 0.05.

3. Results

3.1 General information

We performed data mining of two publicly available RNA sequencing datasets acquired from Gene Expression Omnibus, GSE75421 and GSE101469, which for simplicity reasons will henceforth be referred to as Data 1 and Data 2, respectively, in this study. In Data 1, nine BALB/c mice were treated with 4NQO, which was added in drinking water at the dose of 100 $\mu$ g/ml for eight weeks. Three non-treated mice served as control. Total RNA was extracted from microdissected epithelial cells of tongue mucosa, obtained from three normal mucosa from control mice (week 12, 20, and 24), three hyperplastic lesions (one at week 16 and two at week 20), three dysplastic lesions (two at week16, one at week 24), and three tumor samples (at week 16, 20, and 24) from the 4NQO-treated mice In Data 2, C57BL/6 female mice were treated with 4NQO at the dose of 50 mg/L. Three independent lingual mucosa specimens were collected at 0, 12, and 28 weeks after initiation of the treatment. Dysregulation of immunerelated pathways during progression from normal tissue to hyperplasia, dysplasia and tumor, at the indicated time-points after 4NQO treatment, was assessed.

At the same time we established a 4NQO-induced carcinogenesis model in immunocompetent BALB/c mice to assess the therapeutic effect of targeted agents combined with immune checkpoint inhibitor anti-PD-1 Histologic changes during oral carcinogenesis in our BALB/c mouse model are shown in Fig. 1A.

Figure 2.

Representative images of immunofluorescence staining of dysplasia tissues obtained from mice treated with 4NQO for 12 weeks. (A) Tregs (CD4 ${}^{+}$ FOXP3 ${}^{+}$ ) and (B) Th2 (CD45 ${}^{+}$ ST2 ${}^{+}$ ).

3.2 Changes in the immune infiltration during 4NQO-induced oral cancer initiation and progression

To get more insight into the changes in immune infiltration levels by different immune cell subsets during oral carcinogenesis progression from normal tongue tissue to hyperplasia, dysplasia, and tumor, we performed computational analysis of the immune phenotype in gene expression data using CIBERSORT [13]. This analysis suggested higher levels of infiltration by Th2 cells in dysplastic samples compared to control and tumor samples and that dysplastic lesions were significantly more infiltrated by CD8+ T cells (Fig. 1B). The larger immune cell components corresponded to memory CD8+ T cells, Th17 cells, Tregs, and Th2 cells at 12 weeks than at 28 weeks and 0 weeks, although this reached statistical significance only for dysplastic lesions in BALB/c mice and at 12 weeks in C57BL/6 mice treated with 4NQO (Fig. 1C, Tables S1 and S2).

Following this analysis, samples from our BALB/c mouse model were subjected to immunofluorescence staining for T cell markers, including CD4, FOXP3, CD45, and ST2. Consistent with the virtual immunophenotyping analysis based on transcriptomics data, three BALB/C mice treated with 4NQO for 12 weeks were found to show high infiltration of Th2 cell and Treg cells in dysplastic and hyperplastic lesions in tongue mucosa (Fig. 2A and 2B) and normal mice were not sacrificed as control samples due to little inflammatory reaction in normal mucosa.

3.3 Changes in cytokine expression in the tumor microenvironment during 4NQO-induced oral cancer initiation and progression

Considering the relevance of tumor-associated inflammatory response in mice treated with 4NQO for 12 and 28 weeks, we also assessed for expression of interleukins, CC chemokines and their receptors, and inflammatory markers. Interestingly, we found an increase in IL-17 $\alpha$ , IL-19, IL-1 $\alpha$ , IL-15, and IL-22 (fold change $>$ 2) in the mice treated with 4NQO for 12 and 28 weeks compared to non-treated control mice (Fig. 3A). Additionally, the chemokines CCL12 and CCL28 (fold change $>$ 2) (Fig. 3C), as well as their receptors CCR2 (fold change $>$ 1.5) and CCR5 (fold change $>$ 2) (Fig. 3B) were also upregulated in mice treated with 4NQO.

Figure 3.

Expression of interleukins, CC chemokines, and their receptors after 4NQO treatment for 0, 12, and 28 weeks. The expression was compared between 0 and 12 weeks (in pink), 12 and 28 weeks (in green), and 0 and 28 weeks (in blue). Results were expressed as fold change relative to the control group. (A) Expression of interleukins. (B–C) Expression of CC chemokines and their receptors.

3.4 Expression of immune-related pathways during 4NQO-induced oral tumorigenesis in our mouse model

In order to assess the dynamics of immunerelated pathway expression in different tongue lesions and different durations of 4NQO treatment, we performed gene set enrichment analysis in Data 1 and Data 2 using single sample GSEA (ssGSEA). In Data 1, we identified complement and coagulation, as well as IL-17 signaling pathways to be highly associated with the immune state and were strengthened upregulated during the progression from normal mucosa to hyperplasia (Fig. 4A, Table S3 and Fig. S1). However, the platelet activation pathway was downregulated in tumors compared to dysplastic lesions (Fig. 4B, Table S3 and Fig. S2A). We also identified five pathways that were significantly differentially expressed between normal and tumor tissues: leukocyte transendothelial migration and platelet activation pathways were downregulated ( $p<$ 0.05) (Fig. 4C, Table S3 and Fig. S2B), while Th17 cell differentiation, IL-17signaling pathway, and chemokine signaling pathways were at higher level ( $p<$ 0.05).

Figure 4.

Screening for deregulated immune-related pathways depending on histological variation by single-sample GSEA (ssGSEA). Y axis represents the of ssGSEA score of various pathologic stages. (A) IL-17 signaling pathway, as well as complement and coagulation cascades were upregulated in hyperplasia compared to normal tissue ( $p<$ 0.05); (B) Platelet activation pathway was downregulated in tumor compared to dysplastic lesions ( $p<$ 0.05); (C) Chemokine signaling, IL-17 signaling, and Th17 cell differentiation pathways were expressed in higher levels ( $p<$ 0.05) in tumors compared to normal tongue tissues, while leukocyte transendothelial migration and platelet activation pathways were downregulated ( $p<$ 0.05).

Figure 5.

IL-1 $\alpha$ , IL-17 $\alpha$ , and IL-22 expression were assessed by ELISA in tongue tissue taken from mice treated with 4NQO for 0, 12, and 28 weeks. Data are represented as the mean of three replicates. IL-1 $\alpha$ /0wk (87.50 $\pm$ 4.30), IL-1 $\alpha$ /12wk (238.01 $\pm$ 17.10) and IL-1 $\alpha$ /28wk (282.80 $\pm$ 20.10); IL-17 $\alpha$ /0wk (80.51 $\pm$ 2.31), IL-17 $\alpha$ /12wk (246.88 $\pm$ 8.78) and IL-17 $\alpha$ /28wk (343.04 $\pm$ 15.78); IL-22/0wk (100.31 $\pm$ 5.32), IL-22/12wk(207.33 $\pm$ 10.78) and IL-22/28wk (225.00 $\pm$ 11.31).

In Data 2, we found natural killer-mediated cytotoxicity, Fc epsilon RI, Toll-like receptor, and Antigen processing and presentation signaling pathways to be overexpressed (fold change $>$ 1.5) in tissues from mice treated with 4NQO for 12 weeks compared to those treated for 0 weeks (Fig. S3). However, there were no immune-related differentially expressed pathways between tissues from mice treated with 4NQO for 12 and 28 weeks. Additionally, we found six differentially expressed pathways between mice treated with 4NQO for 0 and 28 weeks; there were the Toll-like receptor signaling, IL-17 signaling, Fc epsilon RI signaling, Th17 cell differentiation, natural killer-mediated cytotoxicity, and cytosolic DNA-sensing signaling pathways (Fig. S3).

Considering that the IL-17 signaling pathway was significantly upregulated during progression from normal mucosa to hyperplasia and tumor in BALB/c mice and that it was bioinformatically identified to be expressed at higher levels in mice treated with 4NQO for 28 weeks compared to 12 or 0 weeks, we reasoned to evaluate the expression of IL-17 $\alpha$ , IL-22, and IL-1a interleukins, which are involved in Th17, Treg, and Th2 cell differentiation, by ELISA. These cytokines were profoundly expressed in different levels between the control group and mice treated with 4NQO for 12 or 28 weeks (Fig. 5).Considering these findings, we hypothesized that inhibition of IL-17a could potentially delay the progression of carcinogenesis in oral lesions, augmenting the efficacy of anti-PD-1 therapy.

3.5 Anti-IL-17

\alpha

therapy combined with PD-1 blockade inhibits malignant progression of 4NQO-induced oral lesions

To morphologically assess the therapeutic potential of PD-1 immune checkpoint blockade combined with anti-IL-17 $\alpha$ in oral cancer development, we exposed 55 BALB/c mice to 4NQO (50 $\mu$ g/mL) in drinking water for 16 weeks. One week after completion of the 4NQO treatment, all mice showed white patches on their tongues, indicating the appearance of hyperkeratotic and/or dysplastic lesions.

The mice were randomly allocated into treatment and control groups (11 mice per group). Mice were injected with anti-PD-1 antibody (200 $\mu$ g/mouse), anti-IL-17 $\alpha$ antibody (200 $\mu$ g/mouse) or the combination of anti-PD-1 antibody (100 $\mu$ g/mouse) and anti-IL-17 $\alpha$ antibody (200 $\mu$ g/mouse), once per week for a total of 4 doses. An equal amount of isotype control IgG2a (200 $\mu$ g/mouse) or PBS buffer alone were injected in the 4NQO-treated mice as controls. The selection of the amount of antibody to be injected was based on literature and previous safety reports.

Figure 6.

PD-1, combined with IL-17 blockade prevents oral cancer development. (A) Quantification of the macroscopic lesions observed in control IgG2a, PBS, anti-PD-1, anti-IL-17 $\alpha$ and the combination of anti-PD-1 and anti-IL-17 groups, at the completion of the study, 16 weeks after initiation of the 4NQO treatment. (B) The number of microscopic lesions scored by counting the individual lesions detected in each mouse ( $p<$ 0.05). (C) The proportion of mild or moderate dysplasia and carcinoma in situ(pre-invasive carcinoma) by H&E staining.

We found that only when anti-PD-1 was combined with anti-IL-17 $\alpha$ mice developed a significantly lower number of macroscopic oral tongue lesions (Fig. 6A). In contrast, anti-PD-1 alone or in combination with anti-IL-17 $\alpha$ both delayed the development of microscopic lesions (Fig. 6B). These data indicated the combination of PD-1 and IL-17 $\alpha$ blockade to be a promising approach to prevent the development of early lesions.

More specifically, 72.72% (8/11) of the mice treated with IgG2 antibody and 90.09% (10/11) of those injected with PBS were found to have high-grade lesions (severe dysplasia or carcinoma), whereas only 9.09% (1/11) of the mice that had received anti-PD-1 in combination with anti-IL-17 $\alpha$ developed such lesions. Interestingly, although there was a favorable response in mice treated with anti-PD-1 alone (72.72%; 8/11), three mice developed high-grade disease and were considered to have developed anti-PD-1 resistance (Fig. 6C). The administration of anti-IL-17 $\alpha$ alone was not sufficient to inhibit the development of tongue cancer, as 36.36% rate of lesion development is considerably lower than the 72% and 90% that we observed in the control groups.

Figure 7.

Combination of anti-PD-1 and anti-IL-17 resulted in prolonged survival in BALB/c mice treated with 4NQO. BALB/c mice were treated with 4NQO for 16 weeks. The mice were administrated a total of four doses, at one-week intervals with PBS, control IgG2a, anti-IL-17 $\alpha$ antibody, and/or anti-PD-1 antibody at the indicated doses. Survival rate (A–D) were recorded.

3.6 Combined use of IL-17

\alpha

and PD-1-targeting antibodies improves survival in 4NQO-treated mice

For survival analysis, the mice were randomly allocated into treatment and control groups (10 mice per group). Treatments were carried out as described earlier. As shown in Fig. 7A, mice treated with anti-PD-1 had a favorable survival compared to mice treated with PBS or IgG2a control ( $p<$ 0.01). However, anti-IL-17 $\alpha$ alone was not sufficient to prolong the survival of 4NQO-treated mice (Fig. 7B). The combination of IL-17 $\alpha$ and PD-1 blockade significantly increased the survival rate compared to the control groups or to mice that had received anti-PD-1 or anti-IL17 $\alpha$ monotherapy ( $p<$ 0.01) (Fig. 7C and D). Notably, The general death rate in the process 4NQO treatment is 23.75% [15], and also anti-PD-1 combined with anti-IL-17 may associate with the occurrence of immune-mediated toxicities known as immune-related adverse events [16, 17]. Both the beneficial and harmful effects of ICI-targeted therapies appear to result from an over-reactive immune system. To exclude potential confounding factors and illustrate the effect of targeting IL-17 $\alpha$ , we added up to 30 mice arranged respectively in the anti-PD-1 group and anti-PD-1 combined with IL-17 group in Fig. 7D in this study. The result showed that targeting IL-17alpha can promote anti-PD-1therapy effect.

4. Discussion

In this study, we dissected the dynamic immune microenvironment in the 4NQO-induced mouse model of oral tumorigenesis using BALB/c and C57BL/6 mice. We found evidence of immune surveillance in the precancerous lesions (mild dysplasia and hyperplasia) of C57BL/6 mice treated with 4NQO for 12 weeks. Therefore, we treated mice having precancerous oral lesions and early-stage oral cancer with immune checkpoint inhibitors (ICIs). However, previous reports demonstrated a 33% non-response rate in mice bearing 4NQO-induced oral squamous carcinomas treated with anti-PD-1 therapy [12]. In addition, Levingston and his colleagues found an initial response to PD-1 treatment at early stages; however, lesions continued to progress when anti-PD-1 was administered at later stages [18].

To identify novel resistance mechanisms to anti-PD-1 at the early stages of tongue carcinogenesis, we analyzed changes in immune-related pathways during tongue carcinogenesis in BALB/c mice and and point-in-time of carcinogenesis in C57BL/6 respectively according to pathological classification and 4NQO exposure duration. We identified IL-17 signaling pathway to be a key pathway involved in the carcinogenesis, regardless of the stage in the pathologic process (tumor initiation, or progression from early stages to more advanced stages) or the period of exposure to the carcinogenic agent. In DATA 1, IL-17 signaling pathway was up-regulated during the transformation from normal tongue to hyperplasia and from normal tongue to tumor. Consistently, in DATA 2, the IL-17 signaling pathway was elevated in mice treated with 4NQO for varying periods. IL-17 is the founding member of a novel family of inflammatory cytokines. While the pro-inflammatory properties of IL-17 are key for its protective function, unrestrained IL-17 signaling is associated with immunopathology, autoimmune diseases, and cancer progression.

Interestingly, increased cytokine levels have been associated with resistance to targeted therapies [19, 20]. High IL-17 levels were associated with resistance to VEGF inhibitors [21]. Akbay et al. also suggested IL-17 to be a key molecule suppressing T cells and thereby promoting resistance to PD-1 immune checkpoint blockade [22]. Another research found that IL-17 could activate monocytes to express B7-H1 in a dose-dependent manner. These IL-17-exposed monocytes suppressed cytotoxic T cell immunity, which expressed high levels of the B7-H1 receptor programmed death 1 (PD-1) and exhibited an exhausted phenotype [23]. Therefore, IL-17-mediated immune tolerance should be considered as a therapeutic target for the design of effective immune-based anti-cancer therapies.

We demonstrated that targeting IL-17 $\alpha$ suppressed the oral carcinogenesis process to some extent, yet it was not sufficient to prolong the survival of 4NQO-treated mice. Therefore, we speculated that the higher levels of IL-17 $\alpha$ might be associated with resistance to PD-1 immune checkpoint blockade in the 4NQO mouse model. Noticeably, Wen et al. showed that the insufficient accumulation, activation, and effector function of T cells were associated with poor response to anti-PD-1 treatment in tongue cancer. Furthermore, Tregs and TIM-3 were found to be the potential cellular and molecular regulators, respectively, mediating resistance to anti-PD-1 therapy [12]. We speculated that these findings could be linked to the higher IL-17 expression in tongue lesions, which could impact the ration of Th17/Treg cell subsets. Some studies suggested that a fraction of Tregs probably have the capacity to differentiate into IL-17 producing cells, i.e. Th17 cells due to their cytokine milieu [24, 25, 26]. Mo et al. found that Tregs secreting IL-17 are involved in systemic sclerosis [27]. We cannot rule out the relevance of the above phenomenon in tongue lesions of 4NQO-treated mice.

Additionally, we believe that the ratio of Th17/Treg cells might be associated with the resistance anti-PD-1. It is certain that anti-IL-17 $\alpha$ alone had little effect on prolonging survival. In contrast, the combination of anti-IL-17 $\alpha$ and anti-PD-1 resulted in improved outcomes in 4NQO-treated mice. We speculate that anti-IL-17 $\alpha$ may play a key role in re-balance of Th17/Treg after anti-PD-1 treatment, the mechanism of which is yet to be unraveled. Another possible mechanism by which IL-17a blockade prevents resistance to anti-PD-1 is the inactivation of all three MAPK (ERK, p38, and JNK) and downstream transcription factors AP-1 and p65 NFkappaB [28].

One major limitation of this study lies in that we only investigated the role of IL-17 $\alpha$ in 4NQO-treated BALB/c mice, and our findings need to be validated in other strains and cancer models. Moreover, the treatment regimen and side effects of anti-IL-17 $\alpha$ and anti-PD-1 combination should be investigated more systematically.

5. Conclusions

Anti-PD-1 and PD-L1 are emerging as one of the routine methods of treating several cancers. Considering the prevalence of resistance to these therapies, research efforts aiming to identify combinational approaches are currently very active. In this study, we showed that IL-17 $\alpha$ blockade combined with anti-PD-1 therapy could impair the progression of tongue cancer and prevent the development of resistance to anti-PD-1 therapy to some extent, though IL-17 $\alpha$ blockade alone could not prolong the survival of 4NQO-treated mice. Further research needs to be conducted to elucidate whether IL-17 $\alpha$ potential regulation of Th17/Treg balance in the tumor microenvironment and the relevance of this phenomenon in preventing resistance to anti-PD-1 therapy.

Author contributions

Conception: S-W were responsible for conceptualization

Interpretation or analysis of data: S-W, HX-F, XR-Y, F-L and Z-H collected, interpreted the animal data, were responsible for funding, and methodology of the manuscript. XR-Y analyzed bioinformation data

Preparation of the manuscript: S-W and HX-F wrote the manuscript

Revision for important intellectual content: S-W and E-Z were responsible for revision for important intellectual

Supervision: S-W were responsible for supervision

Supplementary data

The supplementary files are available to download from https://dx-doi-org.web.bisu.edu.cn/10.3233/CBM-203092.

Footnotes

Acknowledgments

This work was supported by grants from Natural Science Foundation of Heilongjiang Province of China (No. YQ2019H020), and Foundation of First Affiliated Hospital of Harbin Medical University (No. 2019M20).

References

Sharma

and Allison

J.P.

, The future of immune checkpoint therapy, Science 348 (2015), 56–61.

Pardoll

D.M.

, The blockade of immune checkpoints in cancer immunotherapy, Nat Rev Cancer 12 (2012), 252–64.

Hanahan

and Weinberg

RA.

, Hallmarks of cancer, the next generation, Cell 144 (2011), 646–74.

Torre

L.A.

et al., Global cancer statistics, 2012 CA Cancer J Clin 65 (2015), 87–108.

Chen

et al., Blockade of PD-1 effectively inhibits in vivo malignant transformation of oral mucosa, Oncoimmunology 7 (2017), e1388484.

Wang

et al., PD-1 Blockade prevents the development and progression of carcinogen-induced oral premalignant lesions, Cancer Prev Res (Phila) 10 (2017), 684–693.

Jiang

et al., Molecular mechanisms and countermeasures of immunotherapy resistance in malignant tumor, J Cancer 10 (2019), 1764–1771.

Mooradian

M.J.

and Sullivan

R.J.

, What to do when Anti-PD-1 therapy fails in patients with melanoma, Oncology (Williston Park) 33 (2019), 141–148.

Weiss

S.A.

Wolchok

J.D.

and Sznol

, Immunotherapy of melanoma, facts and hopes, Clin Cancer Res 25 (2019), 5191–5201.

10.

Smyth

M.J.

et al., Combination cancer immunotherapies tailored to the tumour microenvironment, Nat Rev Clin Oncol 13 (2016), 143–58.

11.

Levingston

C.A.

and Young

M.R.I.

, Local Immune Responsiveness of Mice Bearing Premalignant Oral Lesions to PD-1 Antibody Treatment, Cancers (Basel) 9 (2017). pii, E62.

12.

Wen

et al., Contributions of T cell dysfunction to the resistance against anti-PD-1 therapy in oral carcinogenesis, J Exp Clin Cancer Res 38 (2019), 299.

13.

Wang

et al., Impact of immunodeficiency on lingual carcinogenesis and lymph node metastasis in mice, J Oral Pathol Med 48 (2019), 826–831.

14.

Wang

et al., PD-1 Blockade Prevents the Development and Progression of Carcinogen-Induced Oral Premalignant Lesions, Cancer Prev Res (Phila) 10,12 (2017), 684–693.

15.

Chen

. The animal model with tongue squamous cell carcinoma induced by 4NQO and its mechanism of molecular carcinogenesis, Zhonggguo, Shi Yong Kou Qqiang Yi Xue 3 (2010), 387-390. Chinese.

16.

Anderson

Theron

A.J.

and Rapoport

B.L.

, Immunopathogenesis of Immune Checkpoint Inhibitor-Related Adverse Events: Roles of the Intestinal Microbiome and Th17 Cells, Front Immunol 10 (2019), 2254.

17.

Saunte

D.M.

et al., Candida infections in patients with psoriasis and psoriatic arthritis treated with interleukin-17 inhibitors and their practical management, Br J Dermoatol 177 (2017), 47–62.

18.

Levingston

C.A.

and Young

M.I.

, Local Immune responsiveness of Mice Bearing Premalignant Oral Lesion to PD-1 Antibody Treatment, Cancer(Basel) 9 (2017), pii, E62.

19.

Keller

H.R.

et al., Overcoming resistance to targeted therapy with immunotherapy and combination therapy for metastatic melanoma, Oncotarget 8 (2017), 75675–75686.

20.

Smyth

M.J.

et al., Combination cancer immunotherapies tailored to the tumour microenvironment, Nat Rev Clin Oncol 13 (2016), 143–158.

21.

Chung

A.S.

et al., An interleukin-17-mediated paracrine network promotes tumor resistance to anti-angiogenic therapy, Nat Med 19 (2013), 1114–1123.

22.

Akbay

E.A.

et al., Interleukin-17A Promotes lung tumor progression through neutrophil attraction to tumor sites and mediating resistance to PD-1 blockade, J Thorac Oncol 12 (2017), 1268-1279.

23.

Zhao

et al., Interleukin-17-educated monocytes suppress cytotoxic T-cell function through B7-H1 in hepatocellular carcinoma patients, Eur J Immunol 41 (2011), 2314–2322.

24.

Abdulahad

W.H.

Boots

A.M.H.

and Kallenberg

C.G.M.

, FoxP3+ CD4+ T cells in systemic autoimmune diseases, the delicate balance between true regulatory T cells andeffector Th-17 cells, Rheumatology (Oxford) 50 (2011), 646–656.

25.

Liu

et al., Elevated levels of CD4(+)CD25(+)FoxP3(+) T cells in systemic sclerosis patients contribute to the secretion of IL-17and immunosuppression dysfunction, PLoS One 8 (2013), e64531.

26.

Koenen

H.J.P.M.

et al., Human CD25highFoxp3pos regulatory T cells differentiate into IL-17-producing cells, Blood 112 (2008), 2340–2352.

27.

et al., Imbalance between T helper 17 and regulatory T cell subsets plays a significant role in the pathogenesis of systemicsclerosis, Biomed Pharmacother 108 (2018), 177–183.

28.

Zhou

et al., IL-17A versus IL-17F induced intracellular signal transduction pathways and modulation by IL-17RA and IL-17RC RNA interference in AGS gastric adenocarcinoma cells, Cytokine 38 (2007), 157–164.