Introduction
Extranodal natural killer (NK)/T cell lymphoma (ENKL) has distinct epidemiological, clinical, histological and etiological features. Clinically, ENKL predominantly occurs in the nasal–paranasal area, skin, gastrointestinal tract or other extranodal sites, and it has a poor prognosis caused by rapid lesion progression [
1]. Among the Epstein–Barr virus (EBV)-related lymphomas including Hodgkin lymphoma and Burkitt lymphoma, ENKL is the one most closely associated with EBV infection. EBV latent type II antigens, including latent membrane protein-1 and protein-2 (LMP1 and LMP2) and EBV nuclear antigen 1 (EBNA1), are present in ENKL tumor cells. Immune imbalance has been shown to be an important feature of ENKL patients [
2,
3]. However, the role of immune cells during ENKL progression remains largely unclear.
Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of bone marrow-derived myeloid progenitors including macrophages, granulocytes, dendritic cells and immature myeloid cells [
4,
5]. Studies in recent years have revealed that MDSCs expand dramatically during tumor growth and are a cause of immune evasion of many types of tumors, including multiple myeloma [
6,
7]. MDSCs enhance tumor growth by inhibiting immune responses and T cell proliferation as well as facilitating tumor metastasis and angiogenesis [
8‐
12]. MDSCs can inhibit anti-tumor immunity by suppressing T cell and NK cell functions by increasing the production of arginine, reactive oxygen species (ROS) and nitric oxide (NO) as well as by inducing Treg cells and TGF-β secretion to mediate T cell suppression [
13‐
15]. To our knowledge, the role of MDSCs, a novel immune-suppressive cell subset, during ENKL tumor progression has not previously been reported. In this study, we detected the frequency of MDSCs in the peripheral blood of ENKL patients to characterize the phenotypic and functional features of MDSCs in ENKL, and we further assessed its clinical significance and prognostic value.
Materials and methods
Patients
Peripheral blood mononuclear cells (PBMCs) were collected from 32 age-matched healthy donors and 32 patients with ENKL at the first time of diagnosis at Sun Yat-Sen University Cancer Center (Guangzhou, China) from July 2010 to December 2012. The clinical details of the patients are shown in Supplementary Table 1. All patients were diagnosed with ENKL, and the lymphoma involved nasal and paranasal lesions in 25 cases (upper aerodigestive tract NK/T cell lymphoma, UNKTL; 84.4 %). The median age was 40.5 years old, and the age range was from 17 to 70 years. There were 19 patients in stage I, 3 patients in stage II, 3 patients in stage III and 7 patients in stage IV. Nine patients had elevated serum lactate dehydrogenase (LDH) levels, and 20 patients had B symptoms. The International Prognostic Index (IPI) was high-intermediate/high (2–5) in eight patients. For the Korean Prognostic Index (KPI) model, 17 patients (53.1 %) had none or one adverse factor, and 15 patients (46.9 %) had two to four adverse factors. In the Peripheral T cell lymphoma Prognostic Index (PIT) model, the majority of the patients (20 cases, 62.5 %) had none or one adverse factor, and the other 12 cases (37.5 %) had at least two adverse factors. Nine of the 32 patients were deceased, and the 5-year overall survival was 71.9 % with a median follow-up of 52 months.
All patients and healthy donors provided informed consent prior to the blood sampling. The study was approved by the Research Ethics Committee of the Sun Yat-Sen University Cancer Center.
Flow cytometry analysis
Human monoclonal antibodies against HLA-DR, CD33, CD11b, CD14, CD15, CD66b, iNOS, Arg-1, IL-10, IL-17 and TGFβ conjugated to different fluorescent dyes were obtained from BD Pharmingen (San Jose, CA, USA) or eBioscience (San Diego, CA, USA), and they were used to measure the frequency and phenotype of the MDSCs via surface staining or intracellular staining (Supplementary Table 2). PBMCs were isolated via Ficoll-Hypaque gradient centrifugation to measure the proportion and phenotype of MDSCs. For surface staining, the cells were washed twice and stained for 1 h on ice with mixtures of fluorescence-conjugated surface mAbs or isotype-matched controls. The cells were then washed twice and resuspended in PBS buffer for flow cytometry analysis. The intracellular staining of IL-17 and the other cytokines was performed on PBMCs stimulated with lipopolysaccharide (LPS, 1 μg/ml) for 4 h in RPMI 1640 medium, and the cytokine secretion was blocked by the addition of brefeldin A (10 µg/ml, eBioscience). After washing, the cells were stained with anti-CD33, anti-CD11b and anti-HLA-DR. The cells were then fixed, permeabilized with Perm/Fix solution (eBiosciences) and stained intracellularly with anti-IL-17 or fluorescence-conjugated antibodies for other cytokines. The samples were evaluated on a FC500 flow cytometer (Beckman Coulter) and analyzed with CXP Software (Beckman Coulter, Inc., Fullerton, CA, USA).
T cell suppression assay
CD33+ cells were isolated from the PBMCs from the healthy donors or ENKL patients using human CD33 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. The PBMCs from healthy donors were labeled with 5 μM carboxyfluorescein succinimidyl ester (CFSE; Molecular Probes, Eugene, Oregon, USA) in 1 ml of PBS for 15 min at 37 °C. The labeling was halted by adding an excess of FCS, and the samples were washed twice with RPMI 1640 (Gibco, Life Technologies, China) supplemented with 10 % fetal bovine serum (FBS; ExCell Biology, South America). The CSFE-labeled cells were cultured in an anti-CD3 antibody (OKT3)-coated 96-well plate with or without sorting the CD33+ cells from the ENKL patients or healthy donors at different ratios for 3 days, and N-hydroxy-nor-l-arginine (NOHA; 1 mM), l-NG-monomethylarginine (l-NMMA, 100 μM) or N-acetylcysteine (NAC 1, mM) was added to a portion of the samples. The CFSE fluorescence intensity was analyzed by flow cytometry after 7 days of co-culture and proliferation.
Statistical analyses
The numerical data are shown as the mean ± standard error (SEM). The statistical analysis was performed with the SPSS 13.0 software (SPSS, Chicago, IL, USA) or GraphPad Prism analysis tools (La Jolla, CA, USA). Two group comparisons were tested using Student’s t test, and the association of the density of the MDSCs with the clinical pathological features was examined using Pearson’s chi-square test. The overall survival (OS) was measured from the date of the diagnosis to the date of death from any cause or to the date of the last follow-up visit. The disease-free survival (DFS) was defined as the time from the diagnosis to the first occurrence of progression, relapse after a response, death from any cause, or to the date of the last follow-up of the surviving patients. The survival curves were determined by the Kaplan–Meier method and the log-rank test. A Cox proportional hazards regression analysis was performed to identify the independent prognostic factors for the OS or DFS. The cutoff value was the median of all variants. The statistical tests were based on a level of significance at P < 0.05.
Discussion
It has been suggested that tumor pathogenesis is linked to immune imbalance and immune cell dysfunction. In this regard, tumors are found to affect myelopoiesis and induce the expansion of myeloid cells with immunosuppressive activity in tumor-bearing hosts, including animal models and human patients [
18‐
22]. In this study, we found an expansion of HLA-DR
−CD33
+CD11b
+ and HLA-DR
−CD33
−CD11b
+ cells in the peripheral blood of ENKL patients. However, only the density of HLA-DR
−CD33
+CD11b
+ MDSCs and not that of HLA-DR
−CD33
−CD11b
+ cells was a significant and independent predictor for ENKL patient survival. This result was in line with our study on nasopharyngeal carcinoma (NPC) (unpublished data) and indicated that CD33 expression is an important marker for the MDSC population in cancer patients. Although the HLA-DR
−CD33
−CD11b
+ cell population was expanded in ENKL patients, no clinical relevance and prognostic value was found in this cell population, and this cell population lacked the phenotypic features of MDSCs (Supplementary Figure 1). Our observations indicated that the immune-suppressive cell subset of HLA-DR
−CD33
+CD11b
+ MDSCs has a prognostic value similar to that of Treg cells and other clinical parameters, including TNM stage, IPI score, and LDH level, in ENKL [
23,
24].
Human MDSCs constitute a heterogeneous group. The definitive identification of human MDSCs is complicated by a lack of a specific marker and by the absence of a human homolog of mouse Gr-1 [
12,
25,
26]. Human MDSCs include the Mo-MDSC and the PMN-MDSC subsets and, according to recent data, the myeloid subset, which has suppressive activity. The MDSC phenotypes are commonly evaluated using a single multicolor staining protocol for MDSC1–MDSC6 as follows: MDSC1 (CD14
+IL-4Rα
+); MDSC2 (CD15
+ IL-4Rα
+); MDSC3 (Lineage
− HLA-DR
− CD33
+); MDSC4 (CD14
+HLA-DR
low/−); MDSC5 (CD11b
+CD14
−CD15
+); and MDSC6 (CD15
+ FSC
low SSC
high) [
27]. The MDSC phenotype varies by differentiation status and function in response to the environmental conditions of different cancers, and the MDSC phenotype has been defined as the HLA-DR
−CD33
+CD11b
+ cell population, including PMN- and Mo-MDSCs, in many human cancers, including multiple myeloma [
12,
27]. Based on our observations and those of others, ENKL-MDSCs were immunophenotyped as an HLA-DR
−CD33
+CD11b
+ cell population in this study.
The ENKL-MDSC population consisted predominantly of CD14+ Mo-MDSCs with a minority of CD15+ PMN-MDSCs. Compared to healthy controls, however, the proportion of Mo-MDSCs in ENKL-MDSCs was decreased, and the proportion of PMN-MDSCs in ENKL-MDSCs was increased. The ENKL-MDSC population highly expressed immune mediator molecules, including Arg-1 and iNOS, and it expressed a low level of CD66b. Furthermore, these ENKL-MDSCs secreted moderate levels of suppressive cytokines, including IL-17, IL-10 and TGFβ, and they did not secrete the IFNγ inflammatory cytokine (data not shown). Compared with MDSCs from healthy donors, the ENKL-MDSCs expressed significant higher level of Arg-1 and iNOS, and they secreted higher levels of IL-17 (P < 0.05).
MDSCs can suppress T cell activation and proliferation in tumor-bearing hosts [
28]. Our previous study and other studies have identified that human MDSCs from solid tumors or multiple myeloma can suppress anti-CD3-induced autologous or allogeneic T cell proliferation, including CD4
+ and CD8
+ T cells. There have been reports indicating that MDSC suppression requires antigen presentation through major histocompatibility complex (MHC) class I molecules [
25,
29‐
33]. However, some studies have suggested that the MDSC suppression is dependent on innate immune sensing and that the MDSC-mediated T cell inhibition is a result of the activation of iNOS, leading to increased production of NO and ROS. Thus, the activated antigen-specific CD4
+ T cells interact with MDSCs loaded with specific antigens, converting these cells to non-specific suppressors in cancers [
16,
34]. In this study, we observed that ENKL-MDSCs strongly suppressed the OKT3-stimulated allogeneic or autologous CD4 T cell proliferation but that they only slightly suppressed the OKT3-stimulated allogeneic and autologous CD8 T cell proliferation. These results indicated that the suppression of T cell proliferation by ENKL-MDSCs is both antigen specific and non-antigen specific, especially for CD4 T cell proliferation. Furthermore, our data were in line with the suggestion that MDSCs from tumor-bearing hosts, as characterized by a high level of iNOS/NOS2 and Arg-1, are potent inhibitors of Ag-specific T cell functions that are able to suppress T cells in an Ag-independent manner [
5,
13,
20,
35‐
40]. Furthermore, our results showed that blockage of iNOS, Arg-1 and ROS recovered the MDSC-mediated inhibition of anti-CD3-induced allogeneic and autologous PBMC proliferation. Interestingly, our observations suggested that the inhibition of T cell proliferation by ENKL-MDSCs also correlated with suppressed cytokine secretion, including IL-10 and TGFβ, as well as induction of Treg cells, which was in line with other reports in solid cancers [
41,
42]. Our observations suggested that multiplex mechanisms that include NO production, ROS production, cytokine induction (IL-10 and TGFβ), and Treg cell induction are involved in ENKL-MDSC-mediated suppression.
The percentage or the frequency of MDSC population is always correlated with poor survivals of cancer patients [
43,
44]. Here, our study demonstrated that the HLA-DR
−CD33
+CD11b
+ MDSC population was an independent poor prognostic indicator for DFS and OS of ENKL patients. Our data further showed that the Mo-MDSC population, but not the PMN-MDSC population, is an independent predictor for DSF and OS in ENKL patients. These observations may explain why the Mo-MDSCs were the main component of the MDSC population in ENKL patients. Our results are in line with reports by others. Some studies have indicated that the CD14
+ MDSC population is associated with disease progression in cancers [
29,
45,
46].
In addition, the number of circulating IL-17-producing MDSCs correlated with patient DFS and OS. IL-17 is an inflammatory cytokine typically secreted by CD4 Th17 and CD8 Tc17 cells [
47]. Recent findings have indicated that the role of IL-17 in tumor development is controversial, and IL-17 could promote the induction of MDSCs at a tumor site and enhance the suppressive function of MDSCs on T cell proliferation [
48‐
52]. Our observations for the first time indicate that ENKL-MDSCs can secrete higher levels of IL-17 compared to healthy donors and that the number of IL-17-producing MDSCs is correlated with ENKL patient prognosis (Supplementary Figure 2). A functional investigation of IL-17-producing MDSCs should be performed in future studies.