CD19+IL-10+ regulatory B cells affect survival of tongue squamous cell carcinoma patients and induce resting CD4+ T cells to CD4+Foxp3+ regulatory T cells

Oral Oncology 53 (2016) 27–35

Contents lists available at ScienceDirect

Oral Oncology journal homepage: www.elsevier.com/locate/oraloncology

CD19+IL-10+ regulatory B cells affect survival of tongue squamous cell carcinoma patients and induce resting CD4+ T cells to CD4+Foxp3+ regulatory T cells Xi Zhou a, Yu-Xiong Su b, Xiao-Mei Lao a, Yu-Jie Liang a,⇑,1, Gui-Qing Liao a,⇑,1 a b

Department of Oral and Maxillofacial Surgery, Guanghua School of Stomatology, Guangdong Provincial Key Laboratory, Sun Yat-Sen University, Guangzhou, China Discipline of Oral & Maxillofacial Surgery, Faculty of Dentistry, The University of Hong Kong, Hong Kong, China

a r t i c l e

i n f o

Article history: Received 25 July 2015 Received in revised form 2 November 2015 Accepted 5 November 2015 Available online 26 November 2015 Keywords: Immunosuppression Tregs Bregs Tumor microenvironment IL-10 CD40L Tumor immunity Double staining Oral oncology Tongue squamous cell carcinoma

s u m m a r y Objectives: Increase of regulatory T cells (Tregs) in the tumor microenvironment predicts worse survival of patients with various types of cancer including tongue squamous cell carcinoma (TSCC). Recently, the cross-talk between Tregs and regulatory B cells (Bregs) has been shown in several tumor models. However the relevance of Bregs to tumor immunity in humans remains elusive. Our objective was to investigate the distribution and function of Bregs in TSCC microenvironment. Materials and Methods: Double staining (Bregs: IL10/CD19 and Tregs: Foxp3/CD4) was performed on tissue sections of 46 TSCC, 20 metastasis lymph nodes, and tumor adjacent normal tissue. Flow cytometry analysis was used to detect the Bregs from magnetic bead-sorted B cells after co-culture with TSCC cell lines, and Tregs from sorted CD4+CD25 T cells after co-culture with stimulated B cells. Results: The immunohistochemical (IHC) results showed that the frequency of Bregs/CD19+ B in TSCC (0.80 ± 0.08%) was significantly higher than adjacent normal tissue (0.52 ± 0.04% p < 0.01). And the increase of Bregs in TSCC microenvironment was related to Tregs and predicts worse survival in patients. Cytological experiments indicated that frequency of Bregs increased after co-culture with TSCC cell line and that the induced B cells converted CD4+CD25 T cells into Tregs. Conclusion: The increased expression of Bregs in the TSCC microenvironment plays a significant role in the differentiation of resting CD4+ T cells and influenced the prognosis of TSCC patients. Ó 2015 Elsevier Ltd. All rights reserved.

Introduction Cancer escape is an active process that regulates immune responses by employing immunosuppressive cells. Regulatory T cells (Tregs) as a major immunosuppressive cell mediate antitumor effector cells through inhibitory receptor cell contact (PD1/PDL1 interaction) or secreted factors such as IL-10, TGF-b, IL-27, and IL-35 [1,2]. Evidence linking Tregs in tumor-promoting was offered by the increased levels of Tregs in peripheral blood and the tumor microenvironment, and the association of Tregs to malignant tissue with poor clinical outcome was reported in various types of cancer [3–11]. However, the mechanisms of Tregs induction in the tumor microenvironment remain elusive. ⇑ Corresponding authors at: Department of Oral and Maxillofacial Surgery, Guanghua School of Stomatology, 56 West Lingyuan Road, Guangzhou, Guangdong 510055, China. Tel.: +86 20 83862531; fax: +86 20 83822807. E-mail addresses: [email protected] (Y.-J. Liang), [email protected] (G.-Q. Liao). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.oraloncology.2015.11.003 1368-8375/Ó 2015 Elsevier Ltd. All rights reserved.

The recent identification of regulatory B cells (Bregs) has attracted increasing attention as a new component in immunosuppression, much like Tregs and myeloid suppressive cells (MSC), and relevance to the former two [12–16]. Different B cell subsets have been ascribed with regulatory function, such as CD5+B-1a cells [17,18], CD19hiCD1dhiCD5+ B cells [19], marginal zone (MZ) B cells [20] or immature transitional-2 (T2-MZP) B cells [21,22] since 1974 when Katz et al. found that B suppress cells delayed hypersensitivity on adoptive transfer [23]. Immunosuppressive function of regulatory B cells has been shown in several murine models of chronic inflammation and tumors. IL-10 has been the most prevalent cytokine associated with Bregs in exacerbated autoimmune disease [24], adoptive transfer [25], and tumor models [26–28]. However, Zhang et al. [13] indicated that Bregs induced immunosuppression by cross-talk with Tregs cells in an IL-10 independent manner. Olkhanud et al. [7] clarified that tumor-evoked regulatory B cells (CD19+CD25hiCD69hi) promoted breast cancer metastasis by inducing Tregs dependent on TGF-b and cell contact rather than IL-10.

28

X. Zhou et al. / Oral Oncology 53 (2016) 27–35

Recently, Bregs have been identified in human healthy and diseased conditions, such as autoimmunity [3,29,30], long-term tolerance [31,32] and infection that utilize similar mechanisms of suppression to their murine counterparts. Iwata et al. validated IL-10+ B cells (B10), which represented 0.6% of blood B cells and predominantly found within the CD24hiCD27+B-cell subpopulation, was capable of negatively regulating monocyte cytokine production through IL-10-dependent pathways during in vitro functional assays [33]. Blair et al. demonstrated that CD19+CD24hiCD38hi B cells suppressed the differentiation of Th1, partially via the provision of IL-10 [30,34]. Although animal studies have illustrated the relevance of Bregs to cancer escape and human Bregs have been identified, the distribution and function of Bregs in tumor microenvironment in humans has not been elucidated. Therefore, we hypothesized that tumor microenvironment induces generation of Bregs, which promotes the generation of Tregs. Thus, the present study was aimed at illustrating the prevalence of CD19+IL-10+ Bregs expression in human TSCC microenvironment and identifying the association between Bregs with Tregs and the clinical relevance. In addition, cytological experiments were conducted to mimic the generation of Bregs in tumor microenvironment and illustrate the possible mechanisms.

Materials and methods Patients and tissue samples Informed consent was obtained from all subjects recruited for this study and approval from the Institutional Research Ethics Committee was obtained for the use of these clinical materials for research purposes. We obtained paraffin-embedded TSCC samples and paired tumor-adjacent non-neoplastic tongue epithelium (TANNTE) samples of 46 patients diagnosed with primary TSCC and; metastatic cervical lymph nodes (MLN) samples and paired normal cervical lymph nodes (NLN) samples from other 20 patients diagnosed with TSCC with cervical metastasis from the Department of Oral and Maxillofacial Surgery, Hospital of Stomatology, Sun Yat-sen University from July 2006 to December 2011. The mean followup was 43 months (range 37–79 months). Clinical information of the samples is summarized in Table 1. All patients underwent no chemotherapy or radiotherapy before surgical treatment. Tumor-node-metastasis (TNM) staging was determined according to the 2002 International Union against Cancer (UICC) TNM classification of malignant tumor. Tumor histology was classified according to the 2004 WHO classification.

Cell lines and culture The human TSCC cell lines UM-1, SCC-25, SCC-15, HSC-6 were obtained from the American Type Culture Collection (ATCC) and preserved in our laboratory. Normal cervical lymph nodes samples of 10 patients diagnosed with primary TSCC without cervical metastasis who underwent contralateral neck flap transfer were obtained from the Department of Oral and Maxillofacial Surgery, Hospital of Stomatology, Sun Yat-sen University from October 2013 to January 2015. NLN cell suspension was prepared by trituration and layered on to Ficoll-Hypaque medium before centrifugation. Mononuclear cells were then collected off the interface, washed twice in PBS and resuspended in RPMI-1640 medium supplemented with 50 IU/ml penicillin, 50 IU/ml streptomycin and 10% FBS.

Table 1 Clinicopathological features of TSCC patients and their association with Bregs and Tregs expression in TSCC lesion. Bregs expression

X2 text P

Tregs expression

X2 text P

Characteristics

n

High

Low

High

Low

Total Age <50 P50

46

15

31

28

18

17 29

6 9

11 20

0.766

12 16

5 13

0.301

Gender Female Male

22 24

7 8

15 16

0.913

13 15

9 9

0.813

N classification N0 N1–3

31 15

8 7

23 8

0.157

16 12

15 3

0.064

T classification T1–T2 T3–T4

41 5

12 3

29 2

0.166

24 4

17 1

0.353

Clinical stage I–II III–IV

28 18

6 9

22 9

0.044

14 14

14 4

0.06

Local recurrence Yes No

14 32

9 6

5 26

0.02

14 14

0 18

0.00

Regional recurrence Yes 16 No 30

9 6

7 24

0.012

16 12

0 18

0.00

Distant metastasis Yes 4 No 42

2 13

2 29

0.437

4 24

0 18

0.093

Human cell isolation, co-culture and analysis B cells were isolated by negative selection using the B cell isolation kit II (Miltenyi Biotech) by indirectly magnetically labeled antibodies against CD2, CD14, CD16, CD36, CD43, CD235a (Glycophorin A) to isolation of highly pure CD19+ B cells by depletion of magnetically labeled cells. Flow cytometry analysis showed the purity of isolated B cell was 94.5 ± 2.1%. Next, the isolated B cell suspension (1  106 cells/ml) was divided into seven groups: co-cultured with SCC-25, UM-1, HSC-6 and SCC-15 cell lines at a ratio of 100:1, two positive control stimulated with LPS (Sigma Escherichia coli 055:B5 5 lg/ml), CD40L (R&D 1 lg/ml) and one negative control with PBS for 72 h. LPS and CD40L are commonly used B cell stimulants, which have previously been demonstrated to induce Bregs [14,22,28,30,33]. An aliquot of B cells (1  106 cells) were stimulated with a stimulation cocktail including phorbol 12-myristate 13-acetate (PMA), ionomycin, brefeldin A, monensin (eBioscience) for 6 h and then stained with anti-human CD19 FITC, CD27 PE-CY7 antibody for 30 min. Fixation and permeabilization of cells was carried out by cytofix/cytoperm buffer kit(BD Pharmingen), and then cells were stained with anti-human IL-10 APC antibody followed by flow cytometry analysis. Anti-CD40L antibody (Santa Cruz Biotechnology) was added at a concentration of 20 lL/1  106 cells in the CD40L block group. The supernatant of the co-cultured cells were assayed for secreted IL10 and TGF-b by ELISA (Dakewei). Isolation of CD4+CD25 cells from lymph nodes cells was performed by negative selection using the Treg cell isolation kit II (Miltenyi Biotech), by indirectly magnetically labeled antibodies against CD8, CD14, CD15, CD16, CD19, CD36, CD56, CD123, TCRc/d, and CD235a (Glycophorin A) to isolate of highly pure CD4+ T cells by depletion of magnetically labeled cells, then microbeads conjugated anti-CD25 antibody used to deplete CD25+ cell. Flow cytometry analysis showed the purity of CD4+CD25 cell was 95.5 ± 2.3%.

X. Zhou et al. / Oral Oncology 53 (2016) 27–35

29

Fig. 1. Immunohistochemical double staining of IL10+CD19+ Bregs in TSCC, metastatic lymph nodes and normal tissue. Increased number of Bregs in TSCC (A) compared with adjacent normal tissue (B). Increased number of Bregs in metastatic lymph nodes (C) compared with normal lymph nodes of same patient (D). IL10 expressed in the cytoplasm stained black (green arrow), and CD19 stained orange (yellow arrow). Scale bar = large panels 100 lm (A, C, D); 200 lm (B); and 20 lm all small panels. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

An aliquot of CD4+CD25 cells (1  106 cells) were stained with anti-human CD4-APC-CY7, CD25-PE, CD45RA-FITC antibody for 30 min and followed by flow cytometry analysis. Finally, the co-cultured B cells were co-cultured with the CD4+CD25 T cells in CD3 antibody coated plate for 48 h. The B-T cell co-culture was stained with anti-human CD4 APC-CY7 antibody for 30 min, then used the Foxp3 staining buffer set kit; and used anti-human FOXP3-Alexa 488; (BD Pharmingen), followed by flow cytometry analysis. The anti-IL-10 antibody (5 lL/1  106 cells; Pepro Tech) was added to the IL-10 block group. Supernatant of the B-T cell co-culture was analyzed for IL-10, TGF-b, IFN-c, and IL17 by ELISA (Dakewei). Immunohistochemistry (IHC) double staining Two double staining of Bregs IL10+CD19+ and Tregs Foxp3+CD4+ was performed on the paraffin-embedded tissue sections of 46 TSCC, 20 MLN, which was compared with paired TANNTE and paired NLN as described [35,36] (Supplementary 1). Expression of Foxp3/CD4, IL-10/CD19 was evaluated independently by two pathologists specialized in head and neck pathology, both blinded to the clinic pathologic data. CD4+Foxp3+/CD4+ > 20% was defined as high expression. CD19+IL-10+/CD19+ > 0.6% was defined as high expression. Statistical analysis All statistical analyses were carried out using the SPSS18.0 statistical software package. The v2 tests were used to analyze the relationship between Tregs/Bregs expression and clinic pathologic characteristics. Survival curves were plotted by the Kaplan–Meier method and compared using the log-rank test. Survival data were evaluated using Cox regression analyses. Spearman rank correlation test to analyze the relationship between Tregs and Bregs. P < 0.05 in all cases was considered statistically significant.

Results Increased number of IL10+CD19+ Bregs in TSCC and metastatic lymph nodes tissue Histologically, lymphocytic infiltration is present in cancer stroma and usually seen in tumor infiltrating frontier. Therefore, we defined the tumor region as the area including the edge of the tumor, the 400 lm within the tumor, and 400 lm outside the tumor. Representative immunohistochemical features of TSCC, adjacent normal tissue, MLN, and NLN are shown in Figs. 1 and 2. The morphometric analysis revealed that the prevalence of Bregs in CD19+ B cells in TSCC (0.80 ± 0.09%) was significantly higher than that in non-neoplastic areas from patients with TSCC (0.52 ± 0.08%; p = 0.004). Bregs increased in metastatic nodes (0.81 ± 0.08%) compared with normal lymph nodes (0.26 ± 0.04%; p < 0.001) (Fig. 3). Association between Bregs and Tregs in TSCC Increased number of Tregs in TSCC and metastatic lymph nodes tissue was showed in Fig. 3. Prevalence of Tregs in CD4+ T cells in TSCC (20.10 ± 1.50%) was significantly higher than that in nonneoplastic inflammatory areas (8.45 ± 0.66%; p < 0.001). Tregs increased in metastatic lymph nodes (19.92 ± 1.45%) compared with normal lymph nodes (5.13 ± 0.56%; p < 0.001). Result of Spearman rank correlation test illustrated that distribution of Bregs and Tregs was correlative (p = 0.02, rho = 0.485). High expression of Bregs suggested the high expression of Tregs. Bregs is associated with TSCC clinicopathological features To investigate whether Bregs or Tregs was associated with clinicopathological characteristics of TSCC, immunohistochemical staining were further statistically analyzed. As shown in Table 1,

30

X. Zhou et al. / Oral Oncology 53 (2016) 27–35

Fig. 2. Immunohistochemical double staining of Foxp3+CD4+ Tregs in TSCC, metastatic lymph nodes and normal tissue. Increased number of Foxp3+CD4+ Tregs in TSCC (A) compared with adjacent normal tissue (B). Increased number of Foxp3+CD4+ Tregs in metastatic lymph nodes (C) compared with the normal lymph nodes of same patient (D). Foxp3 expressed in the nucleus stained black (green arrow), and CD4 stained orange (yellow arrow). Scale bar = large panels 100 lm (A, C, D); 200 lm (B); and 20 lm all small panels. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. Increased number of Bregs and Tregs in TSCC and metastatic lymph nodes tissue (A and B). Scatter diagram showed positive association between Bregs and Tregs (C). TANNTE = tumor-adjacent non-neoplastic inflammatory tongue epithelium; TSCC = tongue squamous cell carcinoma; NLN = normal lymph node; MLN = metastatic lymph node.

frequency of Bregs was significantly associated with clinical stage, local recurrence, and regional recurrence (p < 0.05). In contrast, there were no significant correlations between Bregs expression and age, gender, T classification or N classification (see Table 2). Association between Bregs and overall survival of TSCC patients Expression of Bregs and Tregs in TSCC lesions was inversely associated with overall survival of TSCC patient, as revealed by

the results of Kaplan–Meier analysis and log-rank test. As shown in Fig. 4, two survival curves were significantly different, and overall survival of Bregs low expression group was higher than that of Bregs high expression group in the TSCC microenvironment (p < 0.05). Multivariate analysis (Cox regression) was performed to determine the independent prognostic factor of patient outcomes, and Tregs and Bregs expression in TSCC was found to be relative (Table 3, Fig. 3). Tregs expression in TSCC microenvironment was an independent prognostic factor, while Bregs expression in

31

X. Zhou et al. / Oral Oncology 53 (2016) 27–35 Table 2 Log rank test of overall survival for clinical and pathologic factors.

Table 3 Cox regression of overall survival for clinical and pathologic factors.

Characteristic

X2

P

Characteristic

RR

P

Invasive depth Lymphoid infiltrate Clinical stages T Gender Tregs Bregs Pathologic diff Age

0.619 0.484 1.769 0.890 0.038 11.086 11.677 4.566 0.777

0.431 0.486 0.778 0.641 0.845 0.001 0.001 0.102 0.378

Tregs Bregs

0.066 0.635

0.021 0.528

TSCC microenvironment affected survival rate and was partially dependent on Tregs. Percent and phenotype of Bregs in the sorted B cells Staining of CD19+IL10+ Bregs was performed prior to cocultivation, and the percent of Bregs was (1.27 ± 0.12%). Phenotypic staining of B cells showed that the expression of CD27+B cells was 34.8 ± 1.3%, CD19+CD27+IL10+ B cells was 2.1 ± 0.6%, while CD19+CD27 IL10+ B cells was 0.1 ± 0.06%, and Bregs were rich in CD27+ B cell (p < 0.01) Fig. 5A. TSCC cells induce the generation of CD19+IL10+ B cells The magnetic bead-negative sorted B cells (CD19+) of 10 TSCC patients were divided into seven groups: co-cultured with SCC25, UM-1, HSC-6 and SCC-15 cell lines, stimulated with LPS, CD40L and PBS control group. As shown in Fig. 5B. After 72 h coculture, percentage of CD19+IL-10+ Bregs in all the TSCC cell line groups increased: HSC-6 (2.07 ± 0.31%, p < 0.001), SCC-15 (1.43 ± 0.22%, p = 0.008), SCC-25 (3.18 ± 0.87%, p < 0.001) except for UM-1 groups (1.24 ± 0.31%, p = 0.135) and was significantly higher than the PBS control group (0.73 ± 0.30%). However the addition of anti-CD40L antibody reduced the induction effect of Bregs (CD40L Block; Fig. 5D). ELISA assay demonstrated IL-10 expression and TGF-b expression in the supernatant of each group (Fig. 5E and F).

Bregs induce the generation of CD4+FoxP3+ Tregs Phenotypic staining of sorted T cell was performed before cocultivation, and naive T cells CD4+CD25 CD45+ was 42 ± 1.8% (Fig. 6A). Seven types of B cells were generated after co-culture with UM1, HSC-6, SCC-15, and SCC-25 cell lines, stimulated with LPS, CD40L and PBS 72 h. The B cells were then washed and co-cultured with CD4+CD25 T cells in the presence of plated-CD3 antibody for 48 h. As shown in Fig. 6B, the results of flow cytometry analysis showed the percentage of CD4+Foxp3+ Tregs of UM-1 (0.84 ± 0.12; p = 0.057), HSC-6 (1.27 ± 0.18; p = 0.002), SCC-15 (1.25 ± 0.29; p = 0.029), SCC-25 (1.41 ± 0.12; p < 0.001) stimulated B cells co-culture group, LPS and CD40L-stimulated B cells groups was higher than control B cell group (0.53 ± 0.08). UM-1 was significantly lower than CD40L groups (2.04 ± 0.30, p = 0.005) and LPS (1.51 ± 0.001, p = 0.001). However, the addition of anti-IL10 antibody or untouched co-culture in 4 lm transwell did not largely affect the Tregs induction (Fig. 6C and D) ELISA assay demonstrated IL-10, IL-17, TGF-b, and IFN-c expression in the supernatant of each group (Fig. 7).

Discussion At present, relatively little is known about the potential role of Bregs in the tumor microenvironment. This study demonstrates, for the first time, increased distribution of Bregs in TSCC microenvironment. The percentage of Bregs in the TSCC microenvironment was found to significantly correlate with percentage of Tregs and clinical stage, local recurrence, cervical metastatic, and was inversely associated with overall survival of TSCC patients. At the same time, we demonstrated that Bregs induced by TSCC cells could convert CD4+CD25 T cells into Tregs through secretion of IL-10, which

Fig. 4. Kaplan–Meier survival curves for TSCC patients according to Bregs and Tregs expression in tumor lesions. Log-rank test showed that the two curves were significantly different (p < 0.05).

32

X. Zhou et al. / Oral Oncology 53 (2016) 27–35

may be one of the mechanisms undertaken by Bregs in affecting TSCC development. A previous study reported that IL-10+ B cells (B10) represented 0.6% of human blood B cells. B10 cell subset represents 1–3% of spleen B cells in mice [33]. However, the existence and distribution of Bregs in human tumor microenvironment is still unknown. This study showed that CD19+IL-10+ Bregs existed in human TSCC microenvironment and the percentage increased in TSCC microenvironment as compared to adjacent tissue. Expression of Bregs in metastatic lymph nodes was also higher than that in normal lymph nodes. These findings suggest that TSCC microenvironment could induce the differentiation or recruitment of Bregs, which are

consistent with previous animal experiments in which breast cancer cells induced Bregs expression in mouse spleens [14]. It has been demonstrated that CD40L expression on tumor cells was associated with the generation of Bregs [34]. Blair et al. [22,28,30,34] regarded interaction of CD40L–CD40 as an initiating event for the proliferation of Bregs. In this study, Bregs increased in the CD40L-stimulated group as in TSCC cell line group, and Bregs of CD40L blockage group declined. These results not only imply that the CD40L–CD40 pathway is a potential target of immunotherapy, but also provides screening evidence of Bregs as a target for immunotherapy. However, the CD40L blocking did not completely inhibition induction, which indicates that there are likely other

Fig. 5. IL10 expression in sorted CD19+ B cells after 72 h co-culture with TSCC cell lines. Sorted B cells co-cultured with SCC-25, UM-1, HSC-6, and SCC-15 cell lines, stimulated with LPS, CD40L, and the control group for 72 h. For the last 6 h, a stimulation cocktail was added and then the surfaces of the cells were stained for CD19 expression and cytoplasmic IL-10. Background staining was similar to that observed in the presence isotype-matched control mAbs. (A) Cells were stained for CD19 before and after the sorting, and the relative number and phenotype of Bregs in the sorted B cells were determined before cocultivation. (B and C) Representative IL-10 production by B cells from a person. (D) The scatter diagram shows the mean (±SEM) B10-cell frequencies from 10 subjects. (B–D) However, the addition of anti-CD40L blocking antibody reduced the Bregs induction effect. Bar graphs show the mean (±SEM) in the ELISA assay, which demonstrated (E) IL10 expression and (F) TGF-b expression in the supernatant of each group after co-culture for 72 h.

X. Zhou et al. / Oral Oncology 53 (2016) 27–35

33

Fig. 6. Foxp3 expression in sorted CD4+ T cells after co-culture with stimulated B cells 48 h. (A) Relative number of naïve T cells CD4+CD25 CD45+ in sorted CD4+CD25 T cells before coculture. (B–D) Sorted CD4+CD25 T cells were co-cultured with B cells generated from SCC-25, UM-1, HSC-6, and SCC-15 and stimulated with LPS, CD40L, or PBS control stimulation, then stained for CD4 and Foxp3 expression, with background staining similar to that observed using isotype-matched control mAbs. (B) The scatter diagram represents the mean (±SEM) of Foxp3+CD4+ cell frequencies from 10 individuals, but the induction became unremarkable when added anti-IL10 antibody or untouched co-culture.

pathways involved. IL-21 is considered as a synergistic factor with CD40L in Bregs generation [37,38]. In addition, TNF-a secreted by tumor [27] and cancer-produced metabolites of 5-Lipoxygenase are regarded as Bregs-induced factors in tumor microenvironment [39]. Further studies on the mechanism of CD40L expression in the cross-talk between tumor TSCC cells and Bregs are currently underway in our laboratory. Another finding highlighted in this study was the illustration of the association between Bregs and Tregs. The continued presence of inflammation at sites of tumor can upset the homeostasis and cause local immunosuppression, potentially leading to tumor development. So far Tregs have been attributed to a crucial role in immunosuppression [40]. There are numerous reports of elevated numbers of Tregs cells in the tumor microenvironment and their role in promoting metastasis [1,5,7,9]. Our research indicated that distribution of Bregs was related to Tregs, which implied that Bregs may induce Tregs generation. Further co-culture experiments not only confirmed that Bregs induced Tregs, but also implied that the induction process may require the participation of cell contact and IL-10 as well. In previous studies, Blair et al. revealed that human Bregs suppression was dependent on IL-10, CD80, and CD86, but not TGF-b [34], which was consistent with our results. Nevertheless, an animal model of breast cancer showed that the conversion of resting CD4+ T cells to Tregs was dependent on high levels of TGF-b rather than IL-10 [14]. In addition, other

experiments reported Bregs induced immunosuppression by expressed granzyme B, lymphotoxin and contact molecules B7, CD80, CD86, CD40/CD40L to impact target cells such as CD4+ T cells, CTLs and NK, DC, and macrophages [41]. Therefore, the mechanisms of Bregs action may be varied, and are still controversial. This study showed that the number of Bregs increased after coculture with TSCC cell line, while IL-10 in the supernatant had no significant increase. There are two potential reasons: on one hand, although we could observe the enrichment of IL10 in the cytoplasm of co-culture induced-Bregs by flow cytometry, the IL10 secreted by induced-Bregs in the supernatant was insufficient to be detected by ELISA. The IL10 in supernatant was mainly secreted by the existing Bregs which declined after in vitro culture. On the other hand, a portion of induced-Bregs were likely still in precursor status, which had the potential to differentiate to Bregs but could not secret IL10 yet [33]. In preliminary experiments we found significant difference of IL-10 secretion after co-culture for 5 days, but viability of most B cell lost at 5–6 days which limited the time of co-culture. Similarly, although modes of Foxp3 Treg-dependent immune suppression mediators include the inhibition of proinflammatory cytokines such as IFN-c, Il-17 and secreted factors such as IL-10, TGF-b, existence of precursor Tregs [42] and limited quantity of cytokines led to no significant difference in ELISA results. What’s more, in the finite time, dominant immunosuppression effect of Tregs might be cell contact (PD1/PDL1 interaction).

34

X. Zhou et al. / Oral Oncology 53 (2016) 27–35

Fig. 7. ELISA assay showed (A) IL10, (B) TGF-b, (C) IL17, and (D) IFN-c expression in the supernatant of each group after 48 h of co-culture.

Limited time of lymphocytes in vitro culture is therefore one of the limitations of our study; however, this article to some extent provides the basis for further study. This study also demonstrated that Bregs in TSCC microenvironment was correlated with clinical stage, local recurrence, metastatic and overall survival, which suggested that Bregs play an important role in the progression of TSCC. Further studies in larger population are needed to confirm these findings. What’s more, further studies will elucidate the biological function of Bregs in the pathogenesis of TSCC in vivo. It remains to be further clarified the molecular targets of Bregs. Our study demonstrated that CD40L–CD40 played a role in Bregs induction by TSCC cells, which provides a pathway for Bregs target in immunotherapy. IL-10 was shown to be a critical medium of cross talk between Tregs and Bregs in this study, as blocking IL-10 could effectively reduce the generation of Tregs. However, both CD40L and IL-10 are important factors of the immune system. Hence, an effective and specific inhibition of Bregs without interfering with the balance of immune system is a key point that deserves further research. Lee-Chang et al. demonstrated the activation of signal transducer and activator of transcription-3 (STAT3) pathway in Bregs in mouse model, and Stattic, the STAT3 targeted drugs, reduced the generation of Tregs [43]. Whole-genome expression analysis showed that TLR9 stimulated B10 cells showed Janus kinase 1 (JAK1), STAT1, STAT3, STAT5A and PTPN11 upregulated in IL-10+ B cells in non-allergic beekeepers [31]. Still, genome-wide analysis of CD40L-associated pathways in TSCC and more specific targets in Bregs targeting therapy require further investigation. In conclusion, this study demonstrated that Bregs, which increased in the TSCC microenvironment, played a significant role in converting resting CD4+T cells to Tregs by IL-10 secretion and predicted worse prognosis of TSCC patients. However, further demonstration of exact mechanisms by which Bregs increase and subsequently function will allow us to manipulate Bregs-targeted therapy more specifically and effectively.

Conflict of interest We have no conflicts of interest to declare. Acknowledgements This study was supported by grants from National Natural Science Foundation of China (Nos. 81172566, 81372884, and 81302367), Specialized Research Fund for the Doctoral Program of Higher Education of China (No. 20130171120125), and Natural Science Foundation of Guangdong Province (No. S2013040015004). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.oraloncology. 2015.11.003. References [1] Barbi J, Pardoll D, Pan F. Treg functional stability and its responsiveness to the microenvironment. Immunol Rev 2014;259:115–39. [2] Lindau D, Gielen P, Kroesen M, Wesseling P, Adema GJ. The immunosuppressive tumour network: myeloid-derived suppressor cells, regulatory T cells and natural killer T cells. Immunology 2013;138:105–15. [3] Lim KP, Chun NA, Ismail SM, Abraham MT, Yusoff MN, Zain RB, et al. CD4 +CD25hiCD127low regulatory T cells are increased in oral squamous cell carcinoma patients. PLoS One 2014;9:e103975. [4] Liang Y, Liu H, Su Y, Zhang T, Chu M, Liang L, et al. Foxp3 expressed by tongue squamous cell carcinoma cells correlates with clinicopathologic features and overall survival in tongue squamous cell carcinoma patients. Oral Oncol 2011;47:566–70. [5] Boucek J, Mrkvan T, Chovanec M, Kuchar M, Betka J, Boucek V, et al. Regulatory T cells and their prognostic value for patients with squamous cell carcinoma of the head and neck. J Cell Mol Med 2010;14:426–33. [6] Leffers N, Gooden MJ, de Jong RA, Hoogeboom BN, Ten HK, Hollema H, et al. Prognostic significance of tumor-infiltrating T-lymphocytes in primary and metastatic lesions of advanced stage ovarian cancer. Cancer Immunol Immunother 2009;58:449–59.

X. Zhou et al. / Oral Oncology 53 (2016) 27–35 [7] Olkhanud PB, Baatar D, Bodogai M, Hakim F, Gress R, Anderson RL, et al. Breast cancer lung metastasis requires expression of chemokine receptor CCR4 and regulatory T cells. Cancer Res 2009;69:5996–6004. [8] Hiraoka N, Onozato K, Kosuge T, Hirohashi S. Prevalence of FOXP3+ regulatory T cells increases during the progression of pancreatic ductal adenocarcinoma and its premalignant lesions. Clin Cancer Res 2006;12:5423–34. [9] Bates GJ, Fox SB, Han C, Leek RD, Garcia JF, Harris AL, et al. Quantification of regulatory T cells enables the identification of high-risk breast cancer patients and those at risk of late relapse. J Clin Oncol 2006;24:5373–80. [10] Curiel TJ, Coukos G, Zou L, Alvarez X, Cheng P, Mottram P, et al. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat Med 2004;10:942–9. [11] Strauss L, Bergmann C, Gooding W, Johnson JT, Whiteside TL. The frequency and suppressor function of CD4+CD25highFoxp3+ T cells in the circulation of patients with squamous cell carcinoma of the head and neck. Clin Cancer Res 2007;13:6301–11. [12] Rosser EC, Blair PA, Mauri C. Cellular targets of regulatory B cell-mediated suppression. Mol Immunol 2014;296–304. [13] Zhang Y, Eliav Y, Shin SU, Schreiber TH, Podack ER, Tadmor T, et al. B lymphocyte inhibition of anti-tumor response depends on expansion of Treg but is independent of B-cell IL-10 secretion. Cancer Immunol Immunother 2013;62:87–99. [14] Olkhanud PB, Damdinsuren B, Bodogai M, Gress RE, Sen R, Wejksza K, et al. Tumor-evoked regulatory B cells promote breast cancer metastasis by converting resting CD4(+) T cells to T-regulatory cells. Cancer Res 2011;71:3505–15. [15] Weng JY, Du X, Geng SX, Peng YW, Wang Z, Lu ZS, et al. Mesenchymal stem cell as salvage treatment for refractory chronic GVHD. Bone Marrow Transplant 2010;45:1732–40. [16] Mauri C, Ehrenstein MR. The ’short’ history of regulatory B cells. Trends Immunol 2008;29:34–40. [17] Cantaert T, Doorenspleet ME, FrancoSalinas G, Paramarta JE, Klarenbeek PL, Tiersma Y, et al. Increased numbers of CD5+ B lymphocytes with a regulatory phenotype in spondylarthritis. Arthritis Rheumatism 2012;64:1859–68. [18] Silverman GJ, Srikrishnan R, Germar K, Goodyear CS, Andrews KA, Ginzler EM, et al. Genetic imprinting of autoantibody repertoires in systemic lupus erythematosus patients. Clin Exp Immunol 2008;153:102–16. [19] Yanaba K, Bouaziz JD, Haas KM, Poe JC, Fujimoto M, Tedder TF. A regulatory B cell subset with a unique CD1dhiCD5+ phenotype controls T cell-dependent inflammatory responses. Immunity 2008;28:639–50. [20] Lenert P, Brummel R, Field EH, Ashman RF. TLR-9 activation of marginal zone B cells in lupus mice regulates immunity through increased IL-10 production. J Clin Immunol 2005;25:29–40. [21] Blair PA, Ehrenstein MR, Mauri C. Immunoregulatory potential of T2-MZP B cells. Future Rheumatol 2008;3:79–84. [22] Blair PA, Chavez-Rueda KA, Evans JG, Shlomchik MJ, Eddaoudi A, Isenberg DA, et al. Selective targeting of B cells with agonistic anti-CD40 is an efficacious strategy for the generation of induced regulatory T2-like B cells and for the suppression of lupus in MRL/lpr mice. J Immunol 2009;182:3492–502. [23] Katz SI, Parker D, Turk JL. B-cell suppression of delayed hypersensitivity reactions. Nature 1974;251:550–1. [24] Fillatreau S, Sweenie CH, McGeachy MJ, Gray D, Anderton SM. B cells regulate autoimmunity by provision of IL-10. Nat Immunol 2002;3:944–50. [25] Mauri C. Regulation of immunity and autoimmunity by B cells. Curr Opin Immunol 2010;22:761–7. [26] de Visser KE, Korets LV, Coussens LM. De novo carcinogenesis promoted by chronic inflammation is B lymphocyte dependent. Cancer Cell 2005;7:411–23.

35

[27] Schioppa T, Moore R, Thompson RG, Rosser EC, Kulbe H, Nedospasov S, et al. B regulatory cells and the tumor-promoting actions of TNF-alpha during squamous carcinogenesis. Proc Natl Acad Sci U S A 2011;108:10662–7. [28] Inoue S, Leitner WW, Golding B, Scott D. Inhibitory effects of B cells on antitumor immunity. Cancer Res 2006;66:7741–7. [29] Bosma A, Abdel-Gadir A, Isenberg DA, Jury EC, Mauri C. Lipid-antigen presentation by CD1d+ B cells is essential for the maintenance of invariant natural killer T cells. Immunity 2012;36:477–90. [30] Flores-Borja F, Bosma A, Ng D, Reddy V, Ehrenstein MR, Isenberg DA, et al. CD19+CD24hiCD38hi B cells maintain regulatory T cells while limiting TH1 and TH17 differentiation. Sci Translational Med 2013;5(173):173ra23. http:// stm.sciencemag.org/content/5/173/173ra23.full-text.pdf+html. [31] van de Veen W, Stanic B, Yaman G, Wawrzyniak M, Söllner S, Akdis DG, et al. IgG4 production is confined to human IL-10 producing regulatory B cells that suppress antigen-specific immune responses. J Allergy Clin Immunol 2013;131:1204–12. [32] Rolle L, Memarzadeh Tehran M, Morell-García A, Raeva Y, Schumacher A, Hartig R, et al. Cutting edge: IL-10-producing regulatory B cells in early human pregnancy. Am J Reprod Immunol 2013;70:448–53. [33] Iwata Y, Matsushita T, Horikawa M, DiLillo DJ, Yanaba K, Venturi GM, et al. Characterization of a rare IL-10-competent B-cell subset in humans that parallels mouse regulatory B10 cells. Blood 2011;117:530–41. [34] Blair PA, Noreña LY, Flores-Borja F, Rawlings DJ, Isenberg DA, Ehrenstein MR, et al. CD19+CD24hiCD38hi B cells exhibit regulatory capacity in healthy individuals but are functionally impaired in systemic lupus erythematosus patients. Immunity 2010;32:129–40. [35] van der Loos CM. Multiple immuno enzyme staining: methods and visualizations for the observation with spectral imaging. J Histochem Cytochem 2007;56:313–28. [36] Schwarz S, Butz M, Morsczeck C, Reichert TE, Driemel O. Increased number of CD25+ FoxP3+ regulatory T cells in oral squamous cell carcinomas detected by chromogenic immunohistochemical double staining. J Oral Pathol Med 2008;37:485–9. [37] Lindner S, Dahlke K, Sontheimer K, Hagn M, Kaltenmeier C, Barth TF, et al. Interleukin 21-induced granzyme B-expressing B cells infiltrate tumors and regulate T cells. Cancer Res 2013;73:2468–79. [38] Yoshizaki A, Miyagaki T, DiLillo DJ, Matsushita T, Horikawa M, Kountikov EI, et al. Regulatory B cells control T-cell autoimmunity through IL-21-dependent cognate interactions. Nature 2012;491:264–8. [39] Wejksza K, Lee-Chang C, Bodogai M, Bonzo J, Gonzalez FJ, Lehrmann E, et al. Cancer-produced metabolites of 5-lipoxygenase induce tumor-evoked regulatory B cells via peroxisome proliferator-activated receptor alpha. J Immunol 2013;190:2575–84. [40] Maggio-Price L, Seamons A, Bielefeldt-Ohmann H, Zeng W, Brabb T, Ware C, et al. Lineage targeted MHC-II transgenic mice demonstrate the role of dendritic cells in bacterial-driven colitis. Inflamm Bowel Dis 2013;19:174–84. [41] Shao Y, Lo CM, Ling CC, Liu XB, Ng KT, Chu ACY, et al. Regulatory B cells accelerate hepatocellular carcinoma progression via CD40/CD154 signaling pathway. Cancer Lett 2014;355:264–72. [42] Zelenay S, Lopes-Carvalho T, Caramalho I, Moraes-Fontes MF, Rebelo M, Demengeot J. Foxp3+ CD25 CD4 T cells constitute a reservoir of committed regulatory cells that regain CD25 expression upon homeostatic expansion. Proc Natl Acad Sci U S A. 2005;102:4091–6. [43] Lee-Chang C, Bodogai M, Martin-Montalvo A, Wejksza K, Sanghvi M, Moaddel R, et al. Inhibition of breast cancer metastasis by resveratrol-mediated inactivation of tumor-evoked regulatory B cells. J Immunol 2013; 191:4141–51.