Tumor-induced CD14⁺HLA-DR^−/low myeloid-derived suppressor cells correlate with tumor progression and outcome of therapy in multiple myeloma patients

Abstract

Myeloid-derived suppressor cells (MDSCs) are heterogeneous, immature, myeloid progenitor cells, which suppress immune responses against tumors. CD14⁺HLA-DR^−/low monocytic MDSCs (M-MDSC) are increased in patients suffering from multiple myeloma (MM). However, the frequency and function of M-MDSCs with the relationship between the tumor development and outcome of therapy in MM remain unclear. In this study, we analyzed the changes in M-MDSCs in newly diagnosed, relapsed and remission MM patients. In addition, we also assessed the response of M-MDSCs in MM patients treated with a bortezomib-based therapy as well as the impact of bortezomib on the modulation of M-MDSCs in vitro. The levels of M-MDSCs in newly diagnosed and relapsed MM patients were significantly increased compared with those in remission MM patients and healthy donors. Moreover, the levels of M-MDSCs were shown to correlate with tumor progression. The decrease in M-MDSCs after proteasome inhibitory therapy suggested that M-MDSCs could be considered as an indicator for the efficacy of therapy. Finally, we found the plasma from newly diagnosed MM patients, and MM cells were able to induce the accumulation of M-MDSCs in vitro. These results indicated that M-MDSCs could be considered as a prognostic predictor and an important cell type contributing to immune suppressive microenvironment in MM patients. Treatments targeting for M-MDSCs may improve therapeutic outcomes for MM patients.

Introduction

Multiple myeloma (MM) is a hematologic cancer with malignant plasma cells abnormally accumulated within the bone marrow (BM). The introduction of three novel agents (lenalidomide, thalidomide and bortezomib) and autologous stem cell transplantation have improved therapeutic outcomes in MM, but it still remains incurable and often relapses due to minimal residual disease [1–3]. Impaired immune function is a feature of MM patients and may contribute to tumorigenesis and disease progression [4]. Tumor cells have evolved various mechanisms to evade the immune surveillance and generated an immunosuppressive microenvironment. Therefore, the novel therapies to improve anti-tumor immunity could be a novel approach to improve disease-free progression in MM patients.

In previous studies, myeloid-derived suppressor cells (MDSCs) have been reported to promote tumor development and evade the immune surveillance by suppressing anti-tumor immune response [5, 6]. Therefore, therapeutic approaches targeting for MDSCs may improve anti-tumor reactions [7]. MDSCs are a group of cells, which expand during cancer, inflammation and infection, and have a remarkable ability to suppress the function of T cells [8, 9]. In mice, the phenotype of MDSCs could be specified as CD11b⁺Gr-1⁺Ly6C⁺ monocytic MDSCs and CD11b⁺Gr-1⁺ Ly6G⁺ granulocytic MDSCs [8, 10]. However, human MDSCs are highly heterogeneous due to the lack of cell-specific markers (unlike mice) and are often characterized by multiple surface antigens: CD11b⁺, CD33⁺, CD14⁺, CD15⁺, S100A9⁺ and HLA-DR^−/low [11–13]. In recent studies, MDSCs were reported to be increased and correlated with disease progression in nonsmall cell lung cancer, renal carcinoma, hepatocellular carcinoma and breast cancer [14–17]. MDSCs inhibited the function of T cells by secreting immunosuppressive cytokines, including IL-6 and IL-10, and inducing regulatory T cells (Tregs). However, MDSCs have not been well characterized in MM patients. In the majority of previous studies, MDSCs have been examined in the peripheral blood (PB) of patients with many solid tumors, but not in the BM microenvironment of MM patients.

The aim of this study is to examine the levels of M-MDSCs in both PB and BM from patients with MM, and to investigate the interaction between MM and MDSCs. We found the levels of M-MDSCs were increased in patients with MM and were associated with tumor progression and outcome of therapy. MM cells promoted the accumulation of MDSCs directly and indirectly via plasma from MM patients.

Materials and methods

Patients and samples

A total of 93 patients diagnosed with MM from the First Affiliated Hospital (n = 37) and Second Affiliated Hospital (n = 56), Anhui Medical University, China, were enrolled in this study. Patients with coexisting medical illnesses that were likely to affect their immune status including inflammatory or autoimmune diseases were excluded from this study. Blood and/or aspirates of BM were collected from newly diagnosed (n = 41), relapsed (n = 12) and remission (n = 40) [including complete remission (CR, n = 9) and very good partial remission (VGPR, n = 31)] MM patients. All newly diagnosed MM patients were classified according to the International Staging System (ISS, β-microglobulin level and serum albumin level) [18]. In the 41 newly diagnosed patients, 20 patients received at least two cycles of bortezomib-based therapy (BD: bortezomib and dexamethasone). Therapy regimen consisted of bortezomib 1.3 mg/m², intravenously days 1, 4, 8 and 11; and dexamethasone 40 mg days 1–4, 8–11. Responses were assessed by the international uniform response criteria for MM [19]. Table 1 showed the clinical characteristics of these MM patients in this study. Thirty age- and sex-matched healthy donors (HDs) were recruited as normal controls. The study protocol was approved by the Ethics Committee of Anhui Medical University. Written informed consents were obtained from all patients and volunteers. Samples were examined within 6 h of collection.

Table 1.

Characteristics of HDs and MM patients

State of disease at sample draw	No. of patients	Average age (range)	Gender (M/F)
Newly diagnosed	41	63.2 (30–96)	28/13
Type of myeloma
IgG	25	61.7 (30–80)	17/8
IgA	9	67.1 (50–96)	7/2
IgM	1	71	1/0
light chain	6	62 (54–79)	3/3
ISS stage
I	1	61	0/1
II	12	63.6 (44–80)	9/3
III	28	63.0 (30–96)	19/9
Relapsed	12	62.2 (43–79)	7/5
Remission	40	62.4 (41–83)	22/18
CR	8	62.6 (41–83)	5/3
VGPR	32	62.4 (46–76)	17/15
HD	30	61.2 (32–88)	19/11

IgA immunoglobulin A, IgG immunoglobulin G, IgM immunoglobulin M, ISS international staging system, CR complete remission, VGPR very good partial remission

Sample collection

Samples from all newly diagnosed MM patients were collected prior to any treatments. Specifically, both blood samples and aspirates of BM were collected on the same day from the same patient in 11 newly diagnosed MM patients. In the 20 newly diagnosed MM patients who received the bortezomib-based therapy, blood samples were collected immediately prior to receiving the second and third cycles of therapy. The samples from relapsed and remission patients were also collected when the patients received at least four cycles of therapy of BD or VAD (vincristine, doxorubicin and dexamethasone). In those patients, samples were taken 4–8 weeks (rest) after the therapies.

Flow cytometric analysis

The following monoclonal antibodies (mAbs) specific for human surface antigens were purchased from Beckman Coulter–Immunotech: FITC-conjugated CD14; PE-conjugated HLA-DR; three color reagent kit for Tregs containing of FITC-conjugated anti-CD25, PE-conjugated anti-CD4 and Percp-Cy5.5-conjugated anti-CD127; APC-conjugated CD14; and ECD-conjugated HLA-DR.

PBMCs and bone marrow mononuclear cells (BMMCs) were layered over Ficoll-Hypaque (Amersham Biosciences, Sweden) and centrifuged at 500×g for 25 min. After density gradient centrifugation, mononuclear cells were collected and washed with RPMI 1640 media (Gibco, USA) containing 5 % fetal bovine serum (FBS, Hyclone, USA) and 1 % penicillin/streptomycin (Sigma-Aldrich, USA). PBMCs and BMMCs were collected, washed, incubated with CD antibodies and analyzed by FACS immediately. Data acquisition and analysis were performed using a flow cytometer (FC500 MPL, Beckman Coulter) and EXPO 32 MultiComp software (Beckman Coulter, Miami, FL, USA). Isotype-matched antibodies were used with all the samples as controls.

Cytokine assay

Plasma from MM patients and healthy controls was collected at the same time of PBMCs and BMMCs isolated. After centrifugation at 3,000 rpm for 10 min at 4 °C, plasma fractions were obtained and stored at −80 °C until used. Plasma concentration of IL-6, IL-10, transforming growth factor-β (TGF-β) and arginase I (Arg-1) were measured by ELISA Kit (R&D System; ESM) according to the manufacturer’s instructions.

Cell culture

Plasma (500 μl) from newly diagnosed MM patients was cultured with PBMCs (1 × 10⁶) from HDs in the absence or presence of bortezomib (IC50, 5 nM; Ben Venue Laboratories Inc, Bedford, USA) in 24-well plates (Wuxi Nest Biotechnology Co, Ltd, Wuxi, China) for 60 h in vitro. Moreover, IM9 MM cell line was kindly provided by Prof. Weihua Xiao (University of Science and Technology of China, Hefei, China). IM9 MM cells (1 × 10⁵) and HDs PBMCs (1 × 10⁶) were cultured in RPMI-1640 containing 10 % FBS with or without bortezomib (5 nM) in 24-well plates for 60 h in vitro. Carboxy fluorescein succinimidyl ester (CFSE)-based (0.5 μmol/L) proliferation assays were performed according to manufacturer’s instructions (Invitrogen, Carlsbad, CA). All cells were incubated at 37 °C in a humidified 5 % CO2 atmosphere incubator. All in vitro experiments were performed in triplicate and repeated three times.

Statistical analysis

Statistical analysis was performed using the SPSS 17.0 software (SPSS Inc, Chicago, IL, USA). Statistical significance was determined by the nonparametric unpaired Mann–Whitney U test and the parametric Student’s t test for paired or unpaired samples when appropriate. Levels of M-MDSCs in different groups of patient and between various time points during bortezomib-based treatments were evaluated by one-way ANOVA followed by LSD t tests for pairwise comparisons. Correlations were evaluated using the Pearson’s coefficient test. Differences were considered significant for a level of P values less than 0.05.

Result

M-MDSCs are increased in newly diagnosed MM patients

To determine whether MDSCs played a role in the development of MM, we examined the expression of M-MDSCs in PB (n = 41) and BM (n = 11) from 41 newly diagnosed MM patients, as well as in PB from 30 age- and sex-matched HDs. A higher proportion of CD14⁺ monocytes were detected in newly diagnosed MM patients compared with HDs (31.7 ± 14.8 % vs 19.8 ± 6.1 %, P < 0.01) in PB (Fig. 1b). The levels of M-MDSCs were also increased significantly in both PB (33.8 ± 19.1 %) and BM (24.7 ± 14.4 %) from newly diagnosed MM patients compared with healthy controls in PB (6.8 ± 4.1 %, P < 0.01, Fig. 1c). The M-MDSCs represented 6.1–80.9 % in the CD14⁺ monocytes in PBMCs, and 13.2–59.1 % in BMMCs. In the 11 patients whose PB samples and aspirates of BM were collected on the same day, we found no significant difference in the ratio of M-MDSCs between the PB (30.9 ± 18.5 %) and BM (24.7 ± 13.7 %, P > 0.05) (Fig. 1d). When MM patients were grouped by types of monoclonal immunoglobulin,the levels of M-MDSCs in all groups were increased compared with HDs in PB, but no significant difference was observed among the groups defined by monoclonal immunoglobulin (Fig. 1e).

Fig. 1.

Comparison of M-MDSCs in newly diagnosed MM patients and HDs. a Representative flow cytometry dot plots demonstrate the expression of M-MDSCs in PB and BM from a newly diagnosed patient and a HD. b The levels of CD14⁺ monocytes in newly diagnosed MM patients and HDs. c The levels of M-MDSCs in newly diagnosed MM patients and HDs. d The frequency of M-MDSCs in PB and matched BM. e The frequency of M-MDSCs in different types of monoclonal immunoglobulin in PB from newly diagnosed MM patients. Each dot represents one individual. Horizontal bars indicate mean values. *P < 0.05; **P < 0.01

The levels of M-MDSCs are associated with MM progression and outcome of therapy

To assess the frequency of M-MDSCs in different disease statuses, we found the levels of PB CD14⁺ monocytes were 31.7 ± 14.8, 30.6 ± 10.1 and 31.2 ± 14.6 % in the newly diagnosed, relapsed and remission MM patients, respectively, compared with only 19.8 ± 6.1 % in HDs (Fig. 2a, P < 0.01). However, this difference between MM patients and healthy controls was not observed in BMMCs (Fig. 2b). There were increased levels of M-MDSCs in both blood and BM in newly diagnosed (33.8 ± 19.1 % in PB and 24.7 ± 14.4 % in BM), relapsed (29.8 ± 17.7 % in PB and 43.7 ± 14.1 % in BM), MM patients compared with remission (10.1 ± 7.1 % in PB and 8.3 ± 1.9 % in BM), MM patients and HDs (6.8 ± 4.1 % in PB, Fig. 2c, d).

Fig. 2.

Clinical relevance of M-MDSCs in MM patients. a, b. The frequency of CD14⁺ monocytes was significantly increased in PB in newly diagnosed (ND), relapsed (Rel) and remission (Rem) MM patients. However, this difference was not observed in BM. c, d. The frequency of M-MDSCs was significantly increased in PB and BM in newly diagnosed, relapsed MM patients compared with remission MM patients and HDs. e The frequency of M-MDSCs was associated with MM clinical stages. f The frequency of M-MDSCs in MM patients with renal inadequacy was significantly increased compared with the patients with normal renal function. g In the 20 newly diagnosed MM patients who received the bortezomib-based therapy, the frequency of M-MDSCs was significantly decreased after treatments. Each dot represents one individual. Horizontal bars indicate mean values. *P < 0.05; **P < 0.01

Further research demonstrated that the frequency of M-MDSCs was associated with disease progression. Patients had a higher ratio of peripheral blood M-MDSCs in disseminated disease (stage III: 37.3 ± 20.7 %) than that in limited disease (stage I + II: 26.3 ± 10.6 %, P < 0.05, Fig. 2e). However, the frequency of M-MDSCs in patients with renal inadequacy (Group B, 43.7 ± 20.3 %, serum creatinine ≥177 μmol/L) was significantly increased compared with the patients with normal renal function (Group A, 29.3 ± 14.9 %, P < 0.05, Fig. 2f). To evaluate the frequency of M-MDSCs in relation to the outcome of therapy, 20 newly diagnosed MM patients who received at least two cycles of bortezomib-based therapy were examined. Blood samples were collected prior to each cycle of treatment. The levels of M-MDSCs were significantly decreased after the bortezomib-based therapy (29.3 ± 17.3 vs 15.3 ± 7.5 vs 9.4 ± 8.1 %, prior to vs after treatments, P < 0.01, Fig. 2g). Interestingly, the levels of M-MDSCs remained at the same level or even increased after treatments in several patients as shown in Fig. 2g.

Correlation between M-MDSCs and Tregs

To investigate the relationship between Tregs and M-MDSCs, the levels of circulating Tregs, which were defined as CD4⁺CD25⁺CD127^−/low cells, were measured in newly diagnosed MM patients (n = 20) and HDs (n = 15) in PB. The proportion of CD4⁺ T cells was not different between MM patients and HDs (Fig. 3a). However, the levels of CD4⁺CD25⁺CD127^−/low Tregs were significantly increased in MM patients compared with HDs (9.6 ± 2.7 vs 5.5 ± 1.1 %, P < 0.01, Fig. 3b). There was no significant correlation between the frequency of M-MDSCs and CD4⁺CD25⁺CD127^−/low Tregs (r = 0.27, P > 0.05, Fig. 3c).

Fig. 3.

Relationship between the elevated frequency of Tregs and M-MDSCs. a There was no significantly higher ratio of CD4⁺ lymphocytes in newly diagnosed MM patients compared with HDs. b The frequency of Tregs was significantly increased in newly diagnosed MM patients compared with HDs. c There was no significant correlation between the frequency of M-MDSCs and Tregs (r = 0.27, P > 0.05). Each dot represents one individual. Horizontal bars indicate mean values. *P < 0.05; **P < 0.01

Correlation between M-MDSCs and plasma concentrations of cytokines

The plasma concentrations of IL-6, IL-10, TGF-β and Arg-1 in PB (n = 22) and BM (n = 11) were measured in newly diagnosed MM patients and HDs (n = 15). The plasma concentration of IL-6 was significantly increased (56.8 ± 45.0 pg/ml in PB and 61.0 ± 63.2 pg/ml in BM) compared with HDs (10.1 ± 4.3 pg/ml in PB, P < 0.01, Fig. 4a). However, the increased level of IL-6 did not correlate with the frequency of M-MDSCs (data not shown). There was no significant increase in the levels of IL-10 and TGF-β in both PB and BM in MM patients compared with healthy controls (P < 0.05, Fig. 4b, c). A higher concentration of Arg-1 was present in PB (8.1 ± 15.8 ng/ml) and BM (11.4 ± 24.7 ng/ml) in MM patients compared with HDs (3.9 ± 1.3 ng/ml), but there was no statistical difference existed (P > 0.05) (Fig. 4d).

Fig. 4.

Plasma concentration of IL-6, IL-10, TGF-β and Arg-1 was measured by ELISA kit in newly diagnosed MM patients. a The concentration of IL-6 was significantly increased in newly diagnosed MM patients but not correlated to the frequency of M-MDSCs. b, c Plasma levels of IL-10 and TGF-β were not significantly elevated in MM patients compared with HDs. d Plasma level of Arg-1 was increased, but no significant difference existed. *P < 0.05; **P < 0.01

Interaction between M-MDSCs and tumor cells

After plasma from newly diagnosed (n = 3) MM patients and IM9 MM cells were cultured with PBMCs from HDs for 60 h in vitro, the levels of M-MDSCs were significantly increased with plasma (5.6 vs 24.8 %, pre- vs after culture, P < 0.01, Fig. 5a) and MM cells (5.6 vs 25.4 %, pre- vs after culture, P < 0.01, Fig. 5b). CFSE-based proliferation assays indicated that the M-MDSCs proliferation induced by the tumors could be one of the ways for the accumulation of M-MDSCs (Fig. 5c).

Fig. 5.

Tumor-induced M-MDSCs were increased and could be modulated by bortezomib in PBMCs from HDs in vitro. PBMCs from HDs were cultured with plasma from newly diagnosed MM patients and MM cells in the absence or presence of bortezomib for 60 h in vitro. MDSCs were determined by flow cytometric analysis. Data shown are representative of 3 different experiments. a, b The frequency of M-MDSCs was significantly increased after culturing, and this effect could be modulated by bortezomib when cultured with MM cells (pre-cul, pre-culture). c CFSE-based proliferation assays indicated the cell proliferation was one of the ways for the accumulation of M-MDSCs. The untreated cells were cultured without plasma and MM cells. d Representative flow cytometry dot plots demonstrate the accumulation of M-MDSCs cultured with plasma and MM cells. *P < 0.05; **P < 0.01

To investigate whether the accumulation of M-MDSCs induced by plasma from MM patients or MM cells could be inhibited by novel therapeutic agent bortezomib, PBMCs from HDs were cultured with plasma from MM patients or with IM9 MM cells in presence or absence of bortezomib (5 nM) for 60 h in vitro. A significant decrease of M-MDSCs was observed in the co-culture of MM cells and PBMCs in presence of bortezomib compared with that in absence of bortezomib (15.4 vs 25.4 %, P < 0.05). However, there was no change in M-MDSCs in co-culture of PBMCs and plasma in presence of bortezomib (Fig. 5d).

Discussion

Immune suppressor cells, including MDSCs, play an important role in impaired immune system [12]. It has previously been demonstrated that expansion of MDSCs contributed to tumor immune tolerance and promoted tumor progression in many solid malignancies [13, 20]. However, the frequency and phenotype of MDSCs and their relationship between tumor development and outcome of therapy in MM remain unclear. In this study, we assessed the presence of M-MDSCs in MM patients and their associations with disease statuses and responses to bortezomib-based therapy.

MDSCs are a heterogeneous population of BM-derived myeloid progenitors and have defects in differentiation into mature myeloid cells [8]. MDSCs have been studied in many malignancies, including nonsmall cell lung cancer, nonHodgkin lymphoma, melanoma, hepatocellular carcinoma and breast cancer, in which MDSCs increased predominantly in late stage rather than in early stage of cancer [21–26]. MDSCs promote tumor growth through a variety of ways, including the production of Arg-1, reactive species oxygen, inducible nitric oxide synthase and immunosuppressive cytokines (IL-6, IL-10) [27–29].

We found the frequency of CD14⁺ monocytes was increased in a cohort of MM patients. The elevated monocytes in PB were reported to be increased and associated with poor survival in many solid tumors [30, 31]. Previous studies found that a type of MDSCs, CD11b⁺CD14⁻HLA-DR^−/lowCD33⁺CD15⁺cells, was significantly elevated in the PB and the BM from patients with active (newly diagnosed and relapsed) MM patients compared with HDs [32, 33]. Also, the levels of CD11b⁺CD14⁻HLA-DR^−/lowCD33⁺CD15⁺MDSCs were significantly increased in PB than that in their BM of MM patients. In agreement with their findings, we found that M-MDSCs increased in the PB and BM from patients with active MM compared with patients with stable MM and HDs. In contrast to the previous report, we found there was no significant difference in the levels of M-MDSCs between PB and their BM in MM patients. The difference of MDSCs levels reported in our study compared with that of Favaloro’s [33] may be related to the different group of MDSCs assessed, and also related to the relatively small size of patients (only ten) examined. The frequency of M-MDSCs has also been reported to correlate with tumor progression in patients with many solid tumors, including nonsmall cell lung cancer, hepatocellular carcinoma and bladder cancer [21, 24, 34]. In present study, we found for the first time that the frequency of M-MDSCs in PB was increased and correlated with the progression of disease in hematological malignancy—MM. However, there was no significant difference in the subtypes of monoclonal immunoglobulin in MM patients. This finding indicates that the increase in M-MDSCs in MM may be related to tumorigenesis, but not to the subtype of tumor.

A significantly higher ratio of Tregs was found in newly diagnosed MM patients. In a previous study [33], MDSCs were reported to suppress the function of T cells through the induction of Tregs in vitro. But no significant correlation was found between the frequency of MDSCs and Tregs in present study, suggesting that immunosuppressive functions of MDSCs were not dependent on their ability to generate Tregs only.

Tumor cells and tumor-derived factors such as IL-6 and G-CSF have the ability to induce the accumulation of MDSCs in PBMCs from HDs in vitro [32, 35, 36]. IL-6 is an important pro-inflammatory cytokine and has also been recognized as a key regulator of immunosuppression in advanced cancer [37]. We found that the plasma concentration of IL-6 was significantly increased in PB and BM in newly diagnosed MM patients. TGF-β has been shown to be critical for tumor progression and evasion from immune surveillance [38]. MDSCs were reported as an abundant source of TGF-β production [39, 40]. In this study, the concentration of TGF-β was increased in MM patients compared with HDs. However, there was no significant difference in MM patients compared with HDs. A limitation of the present study is that only serum levels of TGF-β were assessed and it is possible there may be changes in the membrane-bound TGF-β.

Another plasma concentration of immunosuppressive cytokine Arg-1 was increased. Several studies have demonstrated that Arg-1 metabolism induced the suppression of T cell proliferation in vitro [15, 41, 42]. However, no correlation between increased concentration of Arg-1 and elevated levels of M-MDSCs were found.

In this study, we assessed whether the increased frequency of M-MDSCs caused by tumor via its derived factors or by direct interaction with cells in MM. Both the plasma from newly diagnosed MM patients and IM9 MM cells were found to induce the accumulation of M-MDSCs in PBMCs from HDs in vitro. This finding suggested that the microenvironment in MM patients was favorable for the accumulation of M-MDSCs and induced further immunosuppression. It has been reported that bortezomib-based induction regimens were able to achieve higher overall response rates in MM patients [43]. Bortezomib had effects on both MM cells and BM microenvironment. A significant decrease of M-MDSCs was observed in co-culture of MM cells and PBMCs with bortezomib, but not in culture of PBMCs with plasma and bortezomib. This result suggested that bortezomib-mediated reduction of M-MDSCs was due to its effect on MM cells rather than its direct effect on M-MDSCs.

In present study, a close correlation was found between the frequency of M-MDSCs and outcome of bortezomib-based therapy. The frequency of M-MDSCs was decreased in most patients after bortezomib-based therapy. The decrease in proportion of M-MDSCs reflected the efficiency of bortezomib-based therapy. Arihara [24] reported the frequency of circulating MDSCs after treatment was an independent prognostic factor for recurrence in hepatocellular carcinoma patients.

In this study, bortezomib has shown to inhibit MM cells development, but no effect was exerted directly on M-MDSCs. Therefore, the increased frequency of M-MDSCs was due to MM cells and their secreted cytokines. The decrease in frequency of M-MDSCs after bortezomib-based therapy resulted from the reduction of tumor burden. This was well demonstrated in bladder carcinoma, in which the frequency of circulating MDSCs was significantly decreased after surgery [34]. In addition, a bidirectional interaction with MDSCs and tumor also has been demonstrated in which MDSCs promoted tumor growth and, conversely, tumor cells induced MDSCs development [32].

In summary, our results demonstrate, for the first time, that the increased frequency of M-MDSCs is associated with tumor progression and therapeutic responses to the bortezomib-based treatment in MM patients. M-MDSCs may be considered as a prognostic predictor and a new target to improve cancer immunotherapy in MM patients.

Abbreviations

Arg-1

Arginase I

BM

Bone marrow

BMMCs

Bone marrow mononuclear cells

CD

Cluster of differentiation

CFSE

Carboxy fluorescein succinimidyl ester

CR

Complete remission

ELISA

Enzyme-linked immunosorbent assay

FACS

Fluorescence-activated cell sorter

FBS

Fetal bovine serum

G-CSF

Granulocyte colony stimulating factor

HD

Healthy donor

IC50

50 % inhibiting concentration

Ig

Immunoglobulin

iNOS

Inducible nitric oxide synthase

ISS

International staging system

MDSCs

Myeloid-derived suppressor cells

MM

Multiple myeloma

ND

Newly diagnosed

PB

Peripheral blood

PBMCs

Peripheral blood mononuclear cells

Pre-cul

Pre-culture

Rel

Relapsed

Rem

Remission

ROS

Reactive species oxygen

TGF-β

Transforming growth factor-β

Tregs

Regulatory T cells

VGPR

Very good partial remission