Background:B cell acute lymphoblastic leukemia (B-ALL) is a clonal hematopoietic stem cell neoplasm. Despite advances in its therapeutic strategies, including induction chemotherapy followed by allogeneic hematopoietic stem cell transplantation (allo-HSCT), relapse remains the major barrier to the improvement of survival. Unfortunately, the underlying mechanism of relapse in B-ALL is largely unknown. Myeloid-derived suppressor cell (MDSC), a relatively novel population of immunomodulatory cells, has been identified to be associated with the onset of solid tumor and poor prognosis. However, the role of MDSCs played in prompting B-ALL relapse is vague. Aims:The present study was performed to explore the association between the percentage of MDSCs in bone marrow (BM) and B-ALL relapse and to investigate the effects of MDSCs on relapse of B-ALL. Methods:Firstly, twenty relapsed B-ALL patients and 10 non-relapsed patients after allo-HSCT as well as 14 healthy donors were enrolled. The percentage of CD33+HLA-DR-CD11b+ MDSCs in BM was analyzed by flow cytometry (FCM). Secondly, we co-cultured GFP+ Nalm-6 cells (a B cell precursor leukemia cell line initiated from an adolescent male) with MDSCs depleted bone marrow mononuclear cells (BMMCs) or replete of MDSCs from relapsed B-ALL patients post-transplantation. Then, we tested the direct effects of MDSCs on leukemia cells proliferation by co-culturing Nalm-6 cells with MDSCs. In addition, the effects of MDSCs on the functions of T cells and natural killer (NK) cells were also investigated. Briefly, CD3+ T cells or CD3-CD56+ NK cells purified from BMMCs via FCM were co-cultured with the same donor's CD11b+HLA-DR-CD33+ MDSCs using anti-CD3/CD28 beads, PMA or K562 cells as activators. The functional indexes of T cells or NK cells, including proliferation, IFN-γ, CD107a, Granzyme B, and perforin were detected by FCM. Furthermore, we analyzed the expression levels of immune response regulating proteins, including Foxp3 and PD-1 in T cells co-cultured with MDSCs. Finally, we detected the cytokines regulating MDSCs proliferation and activation in the sera of BM from relapsed B-ALL patients after allo-HSCT . Results:We demonstrated an increased percentage of MDSCs in BM of relapsed patients compared with paired non-relapsed cases (n=10 vs. 10, 28.76% vs. 5.18%, P=0.0068) and healthy donors (n=20 vs. 14, 22.76% vs. 6.57%, P=0.0007). In vitro experiments showed that the MDSCs sorted from relapsed patients' BM enhanced the proliferation of tumor cells (n=3, the mean number of GFP+Nalm-6 cells in co-culture with MDSCs group 2.2×104, vs. without MDSCs group 1.2×104, P=0.0447) in a BMMCs condition rather than a co-culture of tumor cells with MDSCs directly. MDSCs obtained from relapsed B-ALL patients' BM could suppress T cell proliferation and IFN-γ production in a cell dose-dependent manner. The percentages of proliferated CD4+T cells were 89.5%, 61.2%, 49.7%, 7.1% at a ratio of T: MDSC 1:0, 1:0.5, 1:1, 1:2 respectively (n=3, P<0.05); The percentages of proliferated CD8+T cells were 74.7%, 57.6%, 37.8%, 10.6% at a ratio of T: MDSC 1:0, 1:0.5, 1:1, 1:2 respectively (n=3, P<0.05). MDSCs from relapsed patients suppressed the cytotoxicity of NK cells. The percentages of IFN-γ+,CD107a+ andGranzyme B+ cells among NK cells were significantly lower in co-culture with MDSCs group than in without MDSCs group (n=4, IFN-γ+ cells, mean 0.6 fold, P=0.0237; CD107a+ cells, mean 0.8 fold, P=0.0242; Granzyme B+ cells, mean 0.8 fold, P=0.0326). We revealed that MDSCs sorted from relapsed patients' BMs induced PD-1 expression on T cells. The percentages of PD-1+ cells among CD4+ cells and CD8+ cells were both significantly higher in co-culture with MDSCs group than in without MDSCs group (n=6, among CD4+ cells, mean 1.8 fold, P=0.0342; among CD8+ cells, mean 1.4 fold, P=0.0041), which may contribute to the dysfunction of T cells. In addition, we found that the mean levels (ng/L) of IL-17A, IL-21, and IL-22 were significantly higher in the sera of BM from relapsed patients than in non-relapsed patients (n=5 vs. 4, IL-17A, 46.9 vs. 6.1, P=0.0077; IL-21, 31.8 vs. 5.6, P=0.0078;IL-22, 132.7 vs. 20.7, P=0.0050). Conclusions:Our results suggest the increased MDSCs might contribute to the relapse of B-ALL via promoting leukemia proliferation, suppressing the function of T cells and NK cells. Our findings suggest that MDSC could be served as a potential therapeutic target for this disease. Disclosures No relevant conflicts of interest to declare.