Introduction

Lung cancer is the leading cause of cancer death and affects public health in China and throughout the world [1, 2]. Furthermore, the morbidity and mortality of lung cancer are rising every year. More than 80 % of all lung cancer cases are diagnosed as non-small-cell lung cancer (NSCLC). Local and systemic immune defects in patients with NSCLC often occur and are associated with poor outcomes in patients [3, 4]; thus, it is critical to understand the contribution of genetic and environmental interactions in the development of NSCLC.

T-cell-related immune responses protect the host from tumorigenesis and tumor progression and have become the core of tumor immunotherapy [5]. Regulatory T cells (Tregs) generate down-regulating signals to inhibit immune response and play an important role in the formation and maintenance of immune tolerance. CD25 is the interleukin (IL)-2 receptor α-chain and a phenotypic marker of CD4+ Tregs [6]. CD4+CD25+ Tregs have been shown to infiltrate tumor tissues and hinder immune response against tumor cells [7]. The transcription factor forkhead box protein P3 (Foxp3) is an important intracellular marker of CD4+CD25+ Tregs [9, 10]. Foxp3 gene is located on the X chromosome at Xp11.23 and is essential for differentiation processes of Tregs [8]. Furthermore, it is reported that naïve murine CD4+CD25+ T cells can convert to the phenotype of Tregs after Foxp3 gene transfer [11]. CD127, α-chain of the IL-7 receptor, is another important maker of Tregs and plays a significant role in the maintenance of immune self-tolerance [12]. Tregs accumulate in tumor microenvironment, effectively suppress immune responses and generally contribute to the poor outcomes seen in NSCLC patients [13, 14]. Previous studies have critically discussed the roles of a complex network of cytokine interactions in the development and progression of primary malignant tumors [15, 16]. Among the cytokines secreted by Tregs, the immune-modulating cytokine transforming growth factor-β1 (TGF-β1) or IL-10 regulate immune functions in the host [17] by down-regulating the immune reactivity of effector T cells [18]. Immunosuppression induced by Tregs has become one of the important mechanisms of tumor immune escape and the main difficulty of successful tumor immunotherapy [19]. In the tumor immunotherapy field, it is necessary to identify specific and functionally relevant surface molecules associated with Tregs to facilitate the development of effective therapeutic strategies against tumor-promoting Tregs.

Epigenetic modification, such as DNA methylation, often occurs in CpG islands of many gene promoters and is involved in numerous pathological processes in tumors [20–22]. DNA methyltransferases (DNMTs) catalyze the methylation process by using S-adenosylmethionine (SAM) to add a methyl molecule to cytosine [23]. DNMT1 is involved in maintaining methylation patterns and prefers hemi-methylated substrates [24, 25]. DNMT3a and DNMT3b function as de novo methyltransferases, as they mainly methylate completely unmethylated DNA [26, 27]. When promoter regions contain 5-methylcytosine in their CpG islands, transcription factors are unable to bind DNA, preventing gene transcription and leading to gene silencing [28, 29]. Nevertheless, DNA demethylation is very important in regulating Foxp3 expression. Studies have shown a Treg-specific demethylated region (TSDR) in the Foxp3 promoter is closely related with Foxp3 expression. Furthermore, cytosine demethylation in the Foxp3 promoter within CpG islands would likely result in the binding of transcription factors and lead to enhanced expression of Foxp3 [30, 31].

In the present study, we identified an important subpopulation, CD4+CD25+Foxp3+ Tregs in the peripheral blood of NSCLC patients and analyzed the demethylation status of the Foxp3 promoter in CD4+ T cells. The correlation between Foxp3 promoter demethylation and gene expression was further established. We also investigated the effect of NSCLC tumor cell line on the phenotype and immunosuppressive function of CD4+ Tregs. The link between Foxp3 gene promoter demethylation and Foxp3 expression was further confirmed in the coculture environment using NSCLC tumor cell line.

Materials and Methods

Patients and samples

The use of blood from patients and healthy donors and the clinical data was approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University (Nanjing, China). Twenty-eight cases with newly diagnosed NSCLC (14 males and 14 females, mean age of 63.3 ± 7.7 years) who received no prior treatment, 9 new cases with benign lung tumor (5 males and 4 females, mean age of 51 ± 10.4 years) with no other known medical conditions and 26 age-matched healthy donors (14 males and 12 females, mean age of 52.8 ± 9.3 years) with no family history of autoimmune diseases or tumors were enrolled in this study from the First Affiliated Hospital of Nanjing Medical University (Nanjing, China).

Of the 28 NSCLC samples, 20 samples were lung adenocarcinoma and 8 samples were lung squamous carcinoma. The 9 benign lung tumor samples comprised 3 lung hematomas, 3 benign nodules, 2 adenomatous hyperplasias and 1 lung spindle cell tumor.

Blood sample collection and CD4+ T cells isolation

Peripheral blood (10 ml) was collected in EDTA-K2 anticoagulant tubes from lung cancer patients, benign lung tumor patients and healthy donors. Serum was stored at −70 °C. Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll–Hypaque density gradient centrifugation (GE Health Care Life Sciences, Piscataway, NJ, USA). Subsequently, CD4+ T cells were separated using Dynabeads Untouched Human CD4 T Cells Kit and the manufacturer’s instructions (Dynal, Oslo, Norway) [3]. PE-conjugated anti-CD4 and PC5-conjugated anti-CD3 monoclonal antibodies (Beckman, Marseille, France) were used to determine the purity of isolated CD4+ T cells by flow cytometry.

Flow cytometry analysis

Foxp3 was labeled using Human Foxp3 Buffer Set (BD Biosciences, New York, NY, USA) according to the manufacturer’s instructions. Cells (105) were labeled with fluorochrome-conjugated monoclonal antibodies for Tregs (CD4+CD25+Foxp3+ T and CD4+CD127− T). The antibodies were FITC-conjugated anti-CD4, APC-conjugated anti-CD25, PE-conjugated anti-Foxp3 and relevant isotype control, PE-conjugated anti-CD4 and FITC-conjugated anti-CD127 and relevant isotype control (all obtained from eBioscience, San Diego, CA, USA). Samples were detected on a Beckman-Coulter Gallios flow cytometer (Beckman-Coulter, Brea, CA, USA) and analyzed using Kaluza 1.3 software (Beckman-Coulter).

Enzyme-linked immunosorbent assay (ELISA)

After 5 days of coculture, CD4+ T cells were collected and cultured in 96-well plates (5 × 104 cells per well). About 1 mg/ml of anti-CD3 (eBioscience, San Diego, CA, USA) and anti-CD28 (eBioscience, San Diego, CA, USA) was added for 12 or 24 h. After stimulation, the levels of IL-10 and TGF-β1 in the culture supernatants were tested using an ELISA Kit (eBioscience, San Diego, CA, USA) and instructions recommended by the manufacturer.

DNA isolation and methylation analysis

Bisulfite sequencing was used to conduct methylation analysis. Genomic DNA from CD4+ T cells was extracted and modified according to the instructions of QIAamp Mini Kit (Qiagen, Hilden, Germany) and CpGenome DNA modification Kit (Millipore Corporation, Billerica, MA, USA). Foxp3 promoter was amplified in ABI 2720 Thermal Cycler (Applied Biosystems/Life Technologies, Grand Island, NY, USA) by using Takara Ex Taq Hot Start (Takara, Shigaken, Japan) at the concentration recommended. The primers used to amplify the Foxp3 promoter were 5′-TGGTGAAGTGGATTGATAGAAAAGG-3′ and 5′-TATAAAAACCCCCCCCCACC-3′. Subsequently, the DNA was treated with TA cloning using DNA A-Tailing (Takara) and sequenced on an ABI 3730 (Applied Biosystems/Life Technologies, Grand Island, NY, USA) in GeneScript Corporation (a Sino-America joint venture, Nanjing, China). Fifteen clones were detected and sequenced for each sample. Sequences were analyzed using the bioinformatics software Chromas version 1.45 (Technelysium, South Brisbane, Australia) and QUantification tool for Methylation Analysis (QUMA; http://quma.cdb.riken.jp/) online analysis system.

Cell lines and culture conditions

The human NSCLC cell lines SPC-A1 and A549, and the human fetal lung fibroblast cell line HFL-1 were purchased from the Chinese Academy of Sciences in Shanghai. They were cultured at 37 °C, with 5 % CO2 in RPMI 1640 media (Gibco, Gaithersburg, MD, USA) containing 10 % fetal bovine serum (FBS; Invitrogen, Carlsbad, CA, USA), penicillin G (50 U/ml) and streptomycin (50 mg/ml).

Coculture of NSCLC cell line and CD4+ T cells

Transwell coculture experiments were carried out in 24-well plates with a 0.4-μm pore inner well (Greiner, Frickenhausen, Germany) to physically separate CD4+ T cells and tumor cells. Tumor cells were prior added in the outer wells of the plates at 2 × 105 cells per well. Isolated human CD4+ T cells were seeded into the inner wells at 6 × 105 cells per well. After 5 days of incubation, CD4+ T cells were collected for further analysis.

Real-time PCR for mRNA expression

Total RNA was prepared and subjected to first-strand cDNA synthesis using RNeasy Micro Kit (Qiagen, Hilden, German) and PrimeScript RT Master Mix (Takara). Gene expression levels were measured by an ABI 7500 Real-time PCR system (Applied Biosystems/Life Technologies, Grand Island, NY, USA) using SYBR Premix DimerEraser (Takara). The sequences of the primers were Foxp3: 5′-CAGCACATTCCCAGAGTTCCTC-3′ and 5′-GCGTGTGAACCAGTGGTAGATC-3′; DNMT1: 5′-GATGTGGCGTCTGTGAGGT-3′ and 5′-CCTTGCAGGCTTTACATTTCC-3′; DNMT3a: 5′-CTCCTGTGGGAGCCTCAATGTTACC-3′ and 5′-CAGTTCTTGCAGTTTTGGCACATTCC-3′; IL-10: 5′-GCTGGAGGACTTTAAGGGTTACCT-3′ and 5′-CTTGATGTCTGGGTCTTGGTTCT-3′; TGF-β1: 5′-CAGCAACAATTCCTGGCGATAC-3′ and 5′-TCAACCACTGCCGCACAACT-3′; β-actin: 5′-GAGCTACGAGCTGCCTGACG-3′ and 5′-GTAGTTTCGTGGATGCCACAG-3′. Cycle threshold (CT) values were estimated by normalizing these values of detected samples against CT values of β-action and converted to be relative expression levels to show the fold change.

Western blot analysis

Cytoplasmic and nuclear protein was extracted from cells using NE-PER nuclear and cytoplasmic extraction reagents (Thermo scientific, Waltham, MA, USA). Total protein was separated by SDS-PAGE and transferred onto PVDF membranes (Bio-Rad, Hercules, CA, USA). Appropriate primary antibody (rabbit anti-human Foxp3, 1:1000 dilution, Cell signaling technology, Danvers, MA, USA; mouse anti-human GAPDH, 1:1000 dilution, ZSGB, Beijing, China) and horseradish peroxidase (HRP)-conjugated secondary antibody (ZSGB) were used to detect specific protein. Protein expression was determined using the enhanced chemiluminescence reagents (Millipore, Billerica, MA, USA) on X-ray film.

Proliferation assay of naïve CD4+ T cells

Naïve CD4+ T cells were isolated from healthy donors by using Dynabeads Untouched Human CD4 T Cells Kit (Dynal, Oslo, Norway) and Dynabeads FlowComp Human CD45RA Kit (Dynal). Different numbers of CD4+ T cells collected from the transwell system were cocultured with naïve CD4+ T cells (1 × 105) and irradiated autologous PBMCs (5 × 104) and stimulated with 500 ng/ml anti-CD3 antibody (eBioscience, San Diego, CA, USA) for 56 h. The proliferative activity of naïve CD4+ T cells was detected by measuring [3H]-thymidine [32].

Nuclear DNMTs activity assay

Nuclear protein from CD4+ cells was extracted using NE-PER nuclear and cytoplasmic extraction reagents (Thermo scientific, Waltham, MA, USA) after 5 days of incubation. Nuclear protein (20 µg) was used in the total DNMTs activity assay, which was performed using the total DNMTs activity detection Kit (Genmed Scientifics, USA) according to the manufacturer’s instructions.

Statistics

SPSS 16.0 software (IBM, Chicago, IL, USA) was used to analyze all data. Student’s two-tailed t test was performed for statistical analysis of group differences. Results were shown as mean ± SD of at least three independent experiments. Pearson correlation analysis was performed to test the correlation. Statistical significance was set as p < 0.05.

Results

Elevated level of Tregs in patients with lung cancer

The level of Tregs in peripheral blood of patients was analyzed by flow cytometry. We found that the percentage of CD4+CD25+ T cells, CD4+CD25+Foxp3+ T cells and CD4+CD127− T cells was significantly higher in peripheral blood samples of patients with lung cancer than those with benign tumor or healthy control (Fig. 1a). The proportion of total CD4+CD25+ T cells in patients with lung cancer (n = 28) was increased compared with benign tumor (n = 9) or healthy control (n = 26) [(19.79 ± 7.10 %) vs. (6.58 ± 4.67 %) (mean ± SD), p < 0.05 and 19.79 ± 7.10 vs. 7.90 ± 5.51 %, p < 0.05, respectively] (Fig. 1b). A statistically significant increase in CD4+CD25+Foxp3+ T cells in lung cancer was also found in comparison with benign tumor or healthy control (6.04 ± 5.12 vs. 1.7 ± 1.21 %, p < 0.05 and 6.04 ± 5.12 vs. 1.35 ± 0.90 %, p < 0.05, respectively) (Fig. 1b). Additionally, the percentage of CD4+CD127− T cells in patients with lung cancer was higher than in patients with benign tumor or healthy control (1.82 ± 1.19 vs. 1.12 ± 0.59 %, p < 0.05 and 1.82 ± 1.19 vs. 1.18 ± 0.77 %, p < 0.05, respectively) (Fig. 1b). There was no obvious difference in the percentage of CD4+CD25+ T cells, CD4+CD25+Foxp3+ T cells and CD4+CD127− T cells between benign tumor and healthy control groups.

Fig. 1
figure 1

High percentage of Tregs in patients with lung cancer. a Flow charts of Foxp3+ T cells from the CD4+CD25+ T cells gate and CD127− T cells from CD4+ T cells gate in lung cancer, benign tumor and healthy control group. The horizontal axis and vertical axis represent the corresponding fluorescence degree of phenotype expressions. Gray and black histograms represent flow cytometry results of isotype control and clinical samples, respectively. b Percentage of CD4+CD25+ T cells, CD4+CD25+Foxp3+ T cells and CD4+CD127− T cells in lung cancer (n = 28), benign lung tumor (n = 9) and healthy control (n = 26) by flow cytometry. Percentage of CD4+CD25+Foxp3+ T cells is obtained by multiplying percentage of CD4+CD25+ T cells by Foxp3+ T cells. Percentage of CD4+CD127− T cells is obtained by multiplying percentage of CD4+ T cells by CD127− T cells. *p < 0.05 compared with benign lung tumor or healthy control group. SS side scatter

Full size image

High levels of IL-10 and TGF-β1 in patients with lung cancer

ELISA was used to detect cytokines levels in the serum of tumor-bearing patients (Fig. 2). The IL-10 and TGF-β1 levels in lung cancer patients (n = 21) were significantly higher than in healthy control (n = 23) and benign tumor (n = 8) (26.3 ± 10.32 vs. 14.04 ± 2.83 pg/ml, p < 0.05 and 26.3 ± 10.32 vs. 17.42 ± 9.57 pg/ml, p < 0.05 for IL-10; 88.98 ± 49.82 vs. 55.81 ± 22.26 pg/ml, p < 0.05 and 88.98 ± 49.82 vs. 61.26 ± 22.23 pg/ml, p < 0.05 for TGF-β1).

Fig. 2
figure 2

High levels of IL-10 and TGF-β1 in patients with lung cancer. Levels of IL-10 and TGF-β1 in serums of lung cancer (n = 23), benign lung tumor (n = 8) and healthy control (n = 21) by ELISA. *p < 0.05 compared with benign lung tumor or healthy control group

Full size image

Correlation between demethylation status of Foxp3 promoter and gene expression in patients with lung cancer

The Foxp3 gene has an associated CpG island that incorporates the promoter and transcriptional start site (TSS) important for Foxp3 gene expression. This CpG island contains eight CpG sites (−138, −126, −113, −77, −65, −58, −43 and −15), TATA box and other transcription-enhancing factors, such as nuclear factor of activated T cells (NF-AT) and activator protein-1 (AP-1) regions (Fig. 3a). CD4+ T cells DNA was isolated from PBMCs from 17 cases of healthy control, 8 cases of benign tumor and 17 cases of lung cancer. These samples underwent bisulfite-sequencing PCR. Foxp3 promoter of CD4+ T cells from lung cancer patients overall demonstrated a low methylation ratio at the CpG sites investigated compared with benign tumor or healthy control samples (Fig. 3b).

Fig. 3
figure 3

Demethylation status of the Foxp3 promoter in patients with lung cancer. a Schematic representation of the conserved region upstream of the transcription start site (TSS) indicating the location of promoter elements and CpG sites. The CpG island contains eight CpG sites, TATA box, transcriptional start site and other transcription-enhancing factors such as NF-AT and AP-1 region. The arrow shows the position of TSS. b Gene map of the methylation status of the Foxp3 promoter, which contains eight CpG sites in CD4+ T cells from lung cancer (n = 17), benign lung tumor (n = 8) and healthy control (n = 17). The methylation degree at each CpG site is depicted by the strength of shading. c Statistical analysis of the methylation ratio at each CpG site among lung cancer (n = 17), benign lung tumor (n = 8) and healthy control (n = 17) groups. *p < 0.05 compared with the healthy control group. # p < 0.05 compared with the benign lung tumor group. d Pearson correlation analysis between the percentage of CD4+CD25+Foxp3+ T cells and methylation ratios. Levels of CD4+CD25+Foxp3+ T cells in patients with NSCLC were negatively correlated with the methylation ratios of the Foxp3 promoter (CpG site −126, −113, −77, −43 and total methylation ratios)

Full size image

The methylation ratio of total CpG sites in lung cancer was significantly lower than benign tumor and healthy control samples (50.22 ± 14.43 vs. 67.10 ± 8.29 %, p < 0.05 and 50.22 ± 14.43 vs. 70.64 ± 8.71 %, p < 0.05, respectively) (Fig. 3c). Foxp3 promoter in CD4+ T cells from lung cancer displayed a low methylation ratio and differed significantly from healthy controls at CpG positions −138, −126, −113, −77, −65, −58 and −43 (Fig. 3c). The results also showed that all the methylation sites in lung cancer, except sites −77 and −15, decreased relative to benign tumors. Pearson correlation analysis showed that the percentage of CD4+CD25+Foxp3+ T cells negatively correlated with the total methylation ratios (r = −0.721, p < 0.05). CD4+CD25+Foxp3+ T cells also negatively correlated with methylation ratios at CpG positions −126, −113, −77 and −43 (r = −0.724, p < 0.05; r = −0.692, p < 0.05; r = −0.623, p < 0.05; r = −0.755, p < 0.05, respectively) (Fig. 3d).

Enhanced Foxp3 expression in CD4+ T cells from the coculture system

A transwell coculture system was used to investigate the effect of NSCLC cell line on Foxp3 level in CD4+ T cells (Fig. 4a). As shown in Fig. 4b, we observed significantly increased levels of Foxp3 mRNA in CD4+ T cells after being cocultured with SPC-A1 for 5 days when compared with HFL-1 coculture (1.89-fold, p < 0.05) or CD4+ T cells cultured alone (2.11-fold, p < 0.05). Foxp3 protein levels in cytoplasmic and nuclear fractions were increased in CD4+ T cells after being cocultured with SPC-A1 (Fig. 4c). Gray-level analysis also indicated Foxp3 protein was increased in SPC-A1 coculture compared to HFL-1 coculture group (2.27-fold, p < 0.05 and 1.67-fold, p < 0.05) or CD4+ T cells cultured alone (2.5-fold, p < 0.05 and 1.5-fold, p < 0.05).

Fig. 4
figure 4

Enhanced Foxp3 expression in CD4+ T cell from the coculture system. a CD4+ T cells were cocultured with SPC-A1 or HFL-1 cells. CD4+ T cells (6 × 105 cells/well) and SPC-A1 or HFL-1 cells (2 × 105 cells/well) were grown separated by a 0.4-μm pore insert in 24-well culture plates. SPC-A1 or HFL-1 cells were grown in the outer wells. CD4+ T cells were grown in suspension in the inner wells. b Expression of Foxp3 mRNA in CD4+ T cells from SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. *p < 0.05 compared with HFL-1 coculture and CD4+ T cells cultured alone. c Cytoplasm and nuclear protein expression and gray-level analysis of Foxp3 in CD4+ T cells from SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. *p < 0.05 compared with HFL-1 coculture and CD4+ T cells cultured alone group. d Flow charts of Foxp3+ T cells from CD4+CD25+ T cells gate and CD127− T cells from CD3+CD4+ T cells gate in SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. The horizontal axis and vertical axis represent the corresponding fluorescence degree of phenotype expressions. Gray and black histograms represent flow cytometry results of isotype control and clinical samples

Full size image

To identify the amount of CD4+ T cells with the Tregs’ phenotype, we evaluated the levels of CD4+CD25+Foxp3+ T cells and CD4+CD127− T cells in the coculture system by flow cytometry. The percentage of CD4+CD25+ T cells was increased after being cocultured with SPC-A1 relative to HFL-1 coculture or CD4+ T cells cultured alone (14.20 ± 2.42 vs. 7.12 ± 1.66 and 14.20 ± 2.42 vs. 4.98 ± 3.20 %, respectively) (Fig. 4d). The numbers of CD4+CD25+Foxp3+ T cells were also increased after being cocultured with SPC-A1 relative to HFL-1 coculture or CD4+ T cells cultured alone (3.72 ± 1.55 vs. 0.78 ± 1.02 %, p < 0.05 and 3.72 ± 1.55 vs. 0.32 ± 0.95 %, p < 0.05, respectively). Coculture with SPC-A1 increased the prevalence of CD4+CD127− T cells more than coculture with HFL-1 or CD4+ T cells cultured alone (17.50 ± 3.56 vs. 7.15 ± 2.05 %, p < 0.05 and 17.50 ± 3.56 vs. 5.13 ± 1.87 %, p < 0.05, respectively).

Furthermore, we also observed Foxp3 mRNA in CD4+ T cells was significantly increased in A549 coculture group relative to CD4+ T cells cultured alone (2.81-fold, p < 0.05) (Suppl. Fig. 1a). The percentages of CD4+CD25+ T cells, CD4+CD25+Foxp3+ T cells and CD4+CD127− T cells were increased after being cocultured with A549 cells compared with CD4+ T cells cultured alone (Suppl. Fig. 1b).

Immunosuppressive function of CD4+ T cells from the coculture system

To study the expression of immunosuppressive cytokines in the coculture system, we analyzed IL-10 and TGF-β1 mRNA and supernatant protein after anti-CD3 and anti-CD28 stimulation for 12 and 24 h. Interestingly, CD4+ T cells showed higher IL-10 and TGF-β1 responses after being cocultured with SPC-A1. The results demonstrated that mRNA levels of IL-10 and TGF-β1 in SPC-A1-cocultured CD4+ T cells were significantly increased compared with HFL-1-cocultured CD4+ T cells (4.68-fold, p < 0.05 and 2.82-fold, p < 0.05 for 12 h; 5.11-fold, p < 0.05 and 5.14-fold, p < 0.05 for 24 h) or CD4+ T cells cultured alone (5.85-fold, p < 0.05 and 3.95-fold, p < 0.05 for 12 h; 7.15-fold, p < 0.05 and 5.40-fold, p < 0.05 for 24 h) (Fig. 5a). Additionally, IL-10 and TGF-β1 protein found in supernatants was also significantly enhanced in SPC-A1-cocultured CD4+ T cells compared with HFL-1-cocultured CD4+ T cells (p < 0.05 for 12 and 24 h) and CD4+ T cells cultured alone (p < 0.05 for 12 and 24 h) (Fig. 5b).

Fig. 5
figure 5

Immunosuppressive function of CD4+ T cells from the coculture system. a mRNA expression of IL-10 and TGF-β1 in CD4+ T cells isolated from the coculture system after co-stimulation with anti-CD3 and anti-CD28 for 12 or 24 h. *p < 0.05 compared with HFL-1 coculture and CD4+ T cells cultured alone. b Supernatant levels of IL-10 and TGF-β1 in CD4+ T cells isolated from the coculture system after co-stimulation with anti-CD3 and anti-CD28 for 12 or 24 h. *p < 0.05 compared with HFL-1 coculture and CD4+ T cells cultured alone. c Suppressive effect of CD4+ Tregs on naïve CD4+ T cells. Naïve CD4+ T cells were cultured at a ratio of 10:1 with either naive CD4+ T cells or SPC-A1 CD4+ T cells in the presence of anti-CD3 antibody. SPC-A1 CD4+ T cells possess significant suppressive activity for naïve CD4+ T cells. *p < 0.05 compared to the 1:0 group

Full size image

We next determined whether SPC-A1-cocultured CD4+ T cells had suppressive effects on naïve CD4+ T cells. Figure 5c showed the proliferative activity of naïve CD4+ T cells cultured at different ratios with naïve CD4+ T cells or SPC-A1-cocultured CD4+ T cells. 10:1 represented that the ratio of naïve CD4+ T cells and cocultured CD4+ T cells. We found that SPC-A1-cocultured CD4+ T cells showed low proliferative activity in response to anti-CD3 antibody stimulation. When cocultured with purified naïve CD4+ T cells, these cells dramatically inhibited the proliferative activity of naïve CD4+ T cells following the addition of anti-CD3 antibody, in a dose-dependent manner. In contrast, neither HFL-1-cocultured CD4+ T cells or CD4+ T cells cultured alone showed any suppressive effects on naive CD4+ T cells, with both groups enhancing the proliferative activity of naïve CD4+ T cells.

Demethylation status of Foxp3 promoter in CD4+ T cells from the coculture system

To evaluate the underlying mechanism of methylation regulation of Foxp3 gene expression, we detected the methylation status of the Foxp3 promoter in CD4+ T cells after coculture with tumor cells. We found that the Foxp3 promoter displayed demethylation changes after coculture with SPC-A1 compared with HFL-1-cocultured CD4+ T cells or CD4+ T cells cultured alone in the repeat tests of three times (Fig. 6a). The percentage of total CpG sites in SPC-A1-cocultured CD4+ T cells was significantly lower than HFL-1-cocultured CD4+ T cells (29.23 ± 2.06 vs. 54.7 ± 5.22 %, p < 0.05) or CD4+ T cells cultured alone (29.23 ± 2.06 vs. 62.13 ± 5.32 %, p < 0.05) (Fig. 6b). Additionally, Fig. 6b showed that all the CpG sites displayed obvious differences between SPC-A1 coculture group and CD4+ T control group. In SPC-A1 coculture group, the methylation ratio of the Foxp3 promoter showed a remarkable decrease in CpG positions −126, −113, −77, −65, −58 and −43 relative to HFL-1 coculture group (Fig. 6b). Furthermore, all the CpG sites of the Foxp3 promoter displayed demethylation changes in A549 coculture group relative to CD4+ T cells cultured alone (Suppl. Fig. 2a, b).

Fig. 6
figure 6

Demethylation status of the Foxp3 promoter in CD4+ T cells from the coculture system. a A gene map of methylation status of the Foxp3 promoter, which contains eight CpG sites in CD4+ T cells from SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. The methylation degree at each CpG site is depicted by the strength of shading. b Statistical analysis of the methylation ratio of each CpG site in SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. *p < 0.05 compared with CD4+ T cells cultured alone group. # p < 0.05 compared with HFL-1 coculture group. c mRNA expression of DNMT1 and DNMT3a in SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. *p < 0.05 compared with HFL-1 coculture or CD4+ T cells cultured alone. d Nuclear DNMTs activity levels of CD4+ T cells from SPC-A1 coculture, HFL-1 coculture and CD4+ T cells cultured alone. *p < 0.05 compared with HFL-1 coculture or CD4+ T cells cultured alone

Full size image

Additionally, we detected that mRNA expression of DNMT1 and DNMT3a was decreased in CD4+ T cells after being cocultured with SPC-A1 relative to HFL-1 cocultured CD4+ T cells (0.30-fold, p < 0.05 and 0.29-fold, p < 0.05, respectively) or CD4+ T cells cultured alone (0.35-fold, p < 0.05 and 0.25-fold, p < 0.05, respectively) (Fig. 6c). Nuclear DNMTs activity in CD4+ T cells was also decreased after being cocultured with SPC-A1 relative to HFL-1 coculture (0.22-fold, p < 0.05) or CD4+ T cells cultured alone (0.23-fold, p < 0.05) (Fig. 6d).

Discussion

CD4+CD25+ T cells play a vital role in tumor immunity and peripheral tolerance and have become a significant research focus [33, 34]. Tumor-induced Tregs maintain immune self-tolerance in tumor patients and enable tumor cells to develop special mechanisms to escape immune surveillance and anti-tumor immune response [35–37].

It was previously thought that the inhibitory state in lung cancer patients was partially due to the increase in Tregs observed in our study. Ju et al. [38] identified the percentage of CD13+CD4+CD25hi Tregs was significantly increased and was related to pathological stages in NSCLC. Research by Schneider et al. [39] showed a high density of Foxp3+ Tregs was found in the tumor center of adenocarcinomas. We hypothesized that the effect of the increased CD4+CD25+Foxp3+ Tregs and CD4+CD127− Tregs in our experiments were important for NSCLC tumor cells escaping the immune system. Our study also demonstrated that the levels of IL-10 and TGF-β1 were significantly higher in serums of NSCLC patients. Yang et al. also demonstrated a high level of IL-10 and TGF-β1 in malignant pleural effusion, which predicted a poor prognosis in patients with lung cancer [40]. Indirect suppression via the secretion of immunosuppressive mediators, such as IL-10 and TGF-β1, plays a particularly significant role for Tregs-mediated immunosuppression in the tumor microenvironment [41].

Considering the importance of Foxp3 in the immunosuppressive microenvironment mainly established by CD4+CD25+ T cells [42], the factors that control and regulate Foxp3 gene expression needed to be determined. Epigenetic mechanisms have some relationships with the regulation of various genes in the process of cell differentiation and development [43, 44]; thus, it was hypothesized that the methylation status of the Foxp3 promoter may lead to the differences in Foxp3 expression level. In this study, the methylation status of Foxp3 gene promoter region (198 bp) of CD4+ T cells in human peripheral blood was determined. The methylation ratio of the Foxp3 promoter in CD4+ T cells from lung cancer displayed was low and significantly differed from healthy control and benign tumor samples. Interestingly, levels of CD4+CD25+Foxp3+ T cells negatively correlated with the methylation ratios. Our data suggested that the enhanced expression of Foxp3 in CD4+ T cells from NSCLC patients maybe a direct consequence of the demethylation at the Foxp3 promoter region.

To elucidate the effects of tumor cells on Foxp3 promoter demethyation, we cocultured CD4+ T cells and NSCLC cell line. It was found that the expression of Foxp3 in CD4+ T cells was increased in NSCLC cell line coculture group. Additionally, the levels of CD4+CD25+ T cells, CD4+CD25+Foxp3+ T cells and CD4+CD127− T cells were remarkably increased in the coculture system with the NSCLC cell line. However, HFL-1 coculture or CD4+ T culture alone did not display these changes. These results suggest that the phenotype of T cells can be influenced and changed by NSCLC tumor cells. Additionally, indirect inhibition by Tregs through the secretion of anti-inflammatory cytokines, such as IL-10 or TGF-β1, would be in effect. Moreover, NSCLC cell line coculture induced T cells, which could significantly suppress the proliferative activity of naïve CD4+ T cells in a dose-dependent manner. It was possible that the presence of Tregs may improve the survival of NSCLC tumor cells.

The demethylation status of Foxp3 promoter was further analyzed to understand the regulation mechanism of Foxp3 gene expression. Demethylation of Foxp3 promoter appeared to be influenced by SPC-A1 cells. Nuclear DNMTs activity of CD4+ T cells was decreased after being cocultured with NSCLC cell line. It is possible that DNMT1 and DNMT3a participate in Foxp3 promoter demethylation process, as they can catalyze DNA methylation [45]. Elevated expression of Foxp3 in CD4+ T cells is thought to be regulated by demethylation status changes in the CpG island of the Foxp3 promoter. This provided a strict mechanism for the regulation of Tregs’ phenotypic marker expression. The tumor-related Foxp3+ Tregs, in combination with cytokine secretion, showed immunosuppressive potential and suggested a vital role in the formation and development of an immunosuppressive tumor microenvironment.

In this study, we demonstrated that NSCLC tumor cells constitutively influence function of CD4+ T cells from healthy donors. By down-regulating the activation of DNMTs, NSCLC tumor cells demethylated the Foxp3 gene promoter and enhanced the transcription and expression of Foxp3 gene. These Foxp3+ T cells were likely responsible for the secretion of immunosuppressive cytokines, IL-10 and TGF-β1, to increase immunosuppression in tumor-bearing patients. Tregs showed a crucial role in establishing and maintaining the immunosuppressive microenvironment of NSCLC. NSCLC tumor cells have the potential to generate different regulation mechanisms to evade anti-tumor immunity and defeat conventional tumor immunotherapy. Therefore, specific activity inhibition of Tregs may become an effective therapeutic focus to develop successful strategies for NSCLC immunotherapy.