MDSCs drive the process of endometriosis by enhancing angiogenesis and are a new potential therapeutic target

Abstract

Endometriosis affects women of reproductive age via unclear immunological mechanism(s). Myeloid-derived suppressor cells (MDSCs) are a heterogeneous group of myeloid cells with potent immunosuppressive and angiogenic properties. Here, we found MDSCs significantly increased in the peripheral blood of patients with endometriosis and in the peritoneal cavity of a mouse model of surgically induced endometriosis. Majority of MDSCs were granulocytic, produced ROS, and arginase, and suppressed T-cell proliferation. Depletion of MDSCs by antiGr-1 antibody dramatically suppressed development of endometrial lesions in mice. The chemokines CXCL1, 2, and 5 were expressed at sites of lesion while MDSCs expressed CXCR-2. These CXC-chemokines promoted MDSC migration toward endometriotic implants both in vitro and in vivo. Also, CXCR2-deficient mice show significantly decreased MDSC induction, endometrial lesions, and angiogenesis. Importantly, adoptive transfer of MDSCs into CXCR2-KO mice restored endometriotic growth and angiogenesis. Together, this study demonstrates that MDSCs play a role in the pathogenesis of endometriosis and identifies a novel CXC-chemokine and receptor for the recruitment of MDSCs, thereby providing a potential target for endometriosis treatment.

Keywords: Angiogenesis; CXCR2; Endometriosis; Immunosuppression; Myeloid-derived suppressor cells.

Fig 1.

Analysis of MDSCs and cytokines in endometriosis patients. (A) Gating strategy of human HLA-DR-CD11b+CD33+ MDSC and representative plots of flow cytometric analysis of MDSCs in PBMCs from (B) a normal female control and (C) a patient with endometriosis, respectively. (D) Comparison of the percentage of MDSCs in PBMCs between normal female controls and endometriosis female patients, 10 cases in each group. (E and F) Representative plots of flow cytometric analysis of MDSCs in PBMCs from the same patient with endometriosis before (pre-treatment) and 3 months after (post-treatment) laparoscopic removal of the endometrial lesions, respectively. (G) Comparison of the change of MDSCs in PBMCs from endometriosis patients before and after laparoscopic surgery (n=10). (H) Comparison of the concentrations of cytokines in the plasma between normal controls (n=15) and endometriosis patients (n=30). (H) Comparison of the concentrations of cytokines in the plasma from endometriosis patients between pre-treatment and post-treatment (n=30). All bar graphs show mean+SEM. Mann-Whitney U test was used to determine all the significant difference. * represents P < 0.05 ** represents P < 0.001 compared with control or pre-treatment.

Fig 2.

MDSCs in a mouse model of endometriosis. (A and B) Counts and exemplary FACS plot of peritoneal MDSCs within 24 hours after transplantation of endometrial fragments. (C) Variation in counts of peritoneal MDSCs from mice with endometriosis and from sham control mice within 7 days after transplantation. (D) The profile of different peritoneal immune cells in mouse model of endometriosis within 24 hours after transplantation. All bar graphs show mean+SEM. Student’s t test was used to determine the statistical changes (n=10 at each time point). * represents P <0.001 compared with sham controls † represents P<0.001 compared with baseline 0 day.

Fig 3.

Characterization of MDSC subsets in mouse model of endometriosis. (A) Gating strategy and exemplary FACS plot and morphology of MDSC subsets in peritoneum. G-MDSCs are CD11b+Ly-6C^loLy-6G+ and with polymorphonuclear morphology while M-MDSCs are CD11b+Ly-6C^hiLy-6G- with monocytic morphology. (B) Absolute cell number of peritoneal G-MDSC and M-MDSC within 24 hours after transplantation of endometrial fragments (n=10 at each time point). * represent P <0.001 compared with G-MDSCs † represents P<0.001 compared with baseline 0 hour. (C and D) Suppression of T cell proliferation assay. G-MDSCs and M-MDSCs were isolated from mice at 12 hours after tranplantation, 5 mice with transplantation per experiment. The experiments were repeated three times independently. * represents P<0.001 compared with naive T cells. (E and F) Arginase activity and reactive oxygen species production in MDSCs from sham mice (control), MCSCs, G-MDSCs and M-MDSCs from mice at 12 hours after transplantation, 3 sham mice and 2 mice with transplantation per experiment. The experiments were repeated three times independently.* represents P<0.001 compared with controls. All bar graphs show mean+SEM. Student’s t test (B, E and F) and Mann-Whitney U test (C and D) was used to determine the significant differences.

Fig 4.

Effects of MDSCs depletion on the growth and development of endometriosis. (A) Changes of peritoneal immune cells in mice at day 7 after transplantation of endometrial implants and antiGr-1 antibody treatment (n=5). (B and C) Changes of peritoneal G-MDSCs and M-MDSCs at days 3, 5 and 7 after transplantation with antiGr-1 antibody treatment (n=5 at each time point). (D and E) Comparision of the size and weight of endometrial lesions (yellow arrow) between IgG control and antiGr-1 antibody groups. All bar graphs show mean+SEM. Student’s t test, * represent P < 0.001 compared with control group † represents P<0.001 compared with baseline day 3. (F) Morphology of endometriotic lesions in two groups. The endometriotic glandular cells (g) and stromal cells (s) were marked. The magnification is 200X. (G) The endometrial lesions were stained with BrdU antibody for determination of proliferated cells as marked with black arrows. The magnification is 400X. (H) Micro-vessels in endometriotic lesions. The staining signals of DAPI were shown in blue, anti-CD31 mAb in red and anti-α-SMA in green. Premature vessels (white arrows) only express CD31. Mature vessels (yellow arrows) express both CD31 and α-SMA (magnification is 100X).

Fig 4.

Effects of MDSCs depletion on the growth and development of endometriosis. (A) Changes of peritoneal immune cells in mice at day 7 after transplantation of endometrial implants and antiGr-1 antibody treatment (n=5). (B and C) Changes of peritoneal G-MDSCs and M-MDSCs at days 3, 5 and 7 after transplantation with antiGr-1 antibody treatment (n=5 at each time point). (D and E) Comparision of the size and weight of endometrial lesions (yellow arrow) between IgG control and antiGr-1 antibody groups. All bar graphs show mean+SEM. Student’s t test, * represent P < 0.001 compared with control group † represents P<0.001 compared with baseline day 3. (F) Morphology of endometriotic lesions in two groups. The endometriotic glandular cells (g) and stromal cells (s) were marked. The magnification is 200X. (G) The endometrial lesions were stained with BrdU antibody for determination of proliferated cells as marked with black arrows. The magnification is 400X. (H) Micro-vessels in endometriotic lesions. The staining signals of DAPI were shown in blue, anti-CD31 mAb in red and anti-α-SMA in green. Premature vessels (white arrows) only express CD31. Mature vessels (yellow arrows) express both CD31 and α-SMA (magnification is 100X).

Fig 5.

Effect of cytokines and chemokines on MDSCs. (A and B) Analysis of peritoneal cytokines and chemokines in the peritoneal fluids from mice within 24h after transplantation of endometrial implants by cytokine (left) and chemokine (right) arrays. The sample of peritoneal fluid at different time points were pooled together from three mice. The intensities of cytokines and chemokines on the films were quantified using ImageQuant TL. The data analysis was permutation-based avoiding parametric assumptions about the distribution of individual cytokines or chemokines. Time course test was selected to detect the trend. (C) Migration of MDSCs in the presence of various cytokines/chemokines or serum and peritoneal fluid obtained from endometriosis recipient mice in intro (n=5) and in vivo ( n=5). * represents P<0.001 compared with control in vitro † represents P<0.001 compared with control in vivo. (D to F) Estimation of CXCL1, CXCL2 and CXCL5 by ELISA (n=7). * represent P< 0.001 as compared with the control † represents P<0.001 compared with baseline 0 hour. (G) Migration of MDSCs in the presence of peritoneal fluid obtained from endometriosis recipient mice and CXCR2 inhibitors including SB265610 and SB225002 in intro (n=5) and in vivo ( n=5). * represent P < 0.001 as compared with and without inhibitors in vitro † represents P<0.001 compared with and without inhibitors in vivo. All bar graphs show mean+SEM and data were analyzed by Student’s t test. (H) Exemplary FACS plot of CXCR2 expression in G-MDSCs and M-MDSCs. (I) Expression of CXCL1, CXCL2 and CXCL5 in the endometriotic lesions and control (magnification is 400X).

Fig 5.

Effect of cytokines and chemokines on MDSCs. (A and B) Analysis of peritoneal cytokines and chemokines in the peritoneal fluids from mice within 24h after transplantation of endometrial implants by cytokine (left) and chemokine (right) arrays. The sample of peritoneal fluid at different time points were pooled together from three mice. The intensities of cytokines and chemokines on the films were quantified using ImageQuant TL. The data analysis was permutation-based avoiding parametric assumptions about the distribution of individual cytokines or chemokines. Time course test was selected to detect the trend. (C) Migration of MDSCs in the presence of various cytokines/chemokines or serum and peritoneal fluid obtained from endometriosis recipient mice in intro (n=5) and in vivo ( n=5). * represents P<0.001 compared with control in vitro † represents P<0.001 compared with control in vivo. (D to F) Estimation of CXCL1, CXCL2 and CXCL5 by ELISA (n=7). * represent P< 0.001 as compared with the control † represents P<0.001 compared with baseline 0 hour. (G) Migration of MDSCs in the presence of peritoneal fluid obtained from endometriosis recipient mice and CXCR2 inhibitors including SB265610 and SB225002 in intro (n=5) and in vivo ( n=5). * represent P < 0.001 as compared with and without inhibitors in vitro † represents P<0.001 compared with and without inhibitors in vivo. All bar graphs show mean+SEM and data were analyzed by Student’s t test. (H) Exemplary FACS plot of CXCR2 expression in G-MDSCs and M-MDSCs. (I) Expression of CXCL1, CXCL2 and CXCL5 in the endometriotic lesions and control (magnification is 400X).

Fig 6.

Role of CXCR2 in the development of endometriosis. (A) Exemplary FACS plot and frequency of G-MDSCs and M-MDSCs in CXCR2 KO mice. (B) Development of endometrial lesions (yellow arrows) was investigated in CXCR2 KO mice and CXCR2 KO mice with exogenous G-MDSC transplantation. Endometrial lesions were collected at day7 after transplantation. (C) Comparision of the size and weight of endometrial lesions in wide type (WT) mice, CXCR2 KO mice and CXCR2 KO mice with exogenous G-MDSC transplantation. Data are mean±SEM (n=5). * represent P < 0.01 as compared with wild-type mice † represents P<0.001 compared with CXCR2KO mice by Student’s t test. (D) Morphology of endometriotic lesions. The endometriotic glandular cells (g) and stromal cells (s) were marked. The magnification is 200X.Angiogenesis of endometriotic leions was evaluated by IF staining of CD31 in green color and a-SMA in red color. Premature vessels (white arrows) only express CD31. Mature vessels (yellow arrows) express both CD31 and α-SMA. The magnification is 100X.