CD206+ macrophage is an accelerator of endometriotic-like lesion via promoting angiogenesis in the endometriosis mouse model

In endometriosis, M2 MΦs are dominant in endometriotic lesions, but the actual role of M2 MΦ is unclear. CD206 positive (+) MΦ is classified in one of M2 type MΦs and are known to produce cytokines and chemokines. In the present study, we used CD206 diphtheria toxin receptor mice, which enable to deplete CD206+ cells with diphtheria toxin (DT) in an endometriosis mouse model. The depletion of CD206+ MΦ decreased the total weight of endometriotic-like lesions significantly (p < 0.05). In the endometriotic-like lesions in the DT group, a lower proliferation of endometriotic cells and the decrease of angiogenesis were observed. In the lesions, the mRNA levels of VEGFA and TGFβ1, angiogenic factors, in the DT group significantly decreased to approximately 50% and 30% of control, respectively. Immunohistochemical study revealed the expressions of VEGFA and an endothelial cell marker CD31 in lesions of the DT group, were dim compared to those in control. Also, the number of TGFβ1 expressing MΦ was significantly reduced compared to control. These data suggest that CD206+ MΦ promotes the formation of endometriotic-like lesions by inducing angiogenesis around the lesions.

www.nature.com/scientificreports/ endometriosis, it has been reported that M2 MΦs are dominant in endometriotic lesions 13,14 . Bacci et al. showed that intraperitoneal injection of M2 MΦ promoted the growth of endometriotic like lesions in a mouse model, suggesting M2 MΦs are responsible for promoting endometriotic lesions 13 . However, the actual role of MΦ in the pathogenesis of endometriosis is unclear.
To elucidate the role of MΦ in vivo, the use of neutralization antibody has been utilized in eliminating MΦs but is less effective in removing MΦs within lesions owing to differential concentrations of antibody 30 . Therefore, in the present study, we used CD206 diphtheria toxin receptor (DTR) mice, which enable to deplete CD206+ MΦs, one of a specific marker of M2 MΦ, with diphtheria toxin (DT) injection 31,32 to examine the role of CD206+ MΦ in progression of endometriosis.

Material and methods
Reagents and materials. Roswell Park Memorial Institute (RPMI)-1640 medium and Diphtheria Toxin (DT) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Estradiol valerate was from Asuka Pharmaceutical Co., Osaka, Japan. Fetal bovine serum (FBS) was from Life Technologies (Tokyo, Japan). Antibiotics (a mixture of penicillin, streptomycin, and amphotericin B) were from Wako Pure Chemical Industries (Osaka, Japan).
The intensity of the staining of VEGFA, CD31, and TGFβ1 was analyzed by a semi-quantitative method, H-scoring 33 . H-score was calculated by the following equation: H-score = ∑P i × i, where i is the intensity of staining with a value of 0, 1, 2, or 3 (negative, weak, moderate, or strong, respectively) and P i is the corresponding percentage of the cells. As for the Ki-67 evaluation, the percentage of Ki-67 positive cells per total cells was calculated.
Model of endometriosis using CD206 DTR mouse. All animal experiments were approved by the ethical committee of University of Toyama (G2015MED-38 and A2013MED-33), and performed in accordance with animal experiment guidelines and regulations in University of Toyama. Female, CD206 DTR mice, from 12 to 20-weeks-old mice were used. We housed in a specific pathogen-free (SPF) animal facility with a controlled environment, 22-24 °C and 60-70% relative humidity, and on a light/dark cycle (12 h light/12 h dark) with food and water ad libitum. CD206 DTR mice are genetically engineered transgenic (Tg) mice based on the transgenic expression of the diphtheria toxin receptor (DTR) under the control of the CD206 promoter to ablate CD206+ MΦs 31,32,34 specifically. In CD206 DTR mice, CD206+ MΦs are removed about more than 80% in various organs systemically, such as lung, adipose tissue, blood, spleen, and ovary 31,32,34 . Induction of endometriosis was studied as described previously 7 . Briefly, mice were injected s.c. with 100 μg/kg estradiol valerate in sesame oil once per week for three weeks. After three weeks, endometrium-rich fragments from donor mice were finely chopped using a razor blade. Fragments suspended in 0.6 ml phosphate buffered salts (PBS) were injected with an 18-gauge needle through the abdominal wall into the peritoneal cavity of recipient mice with the ratio of one donor to two recipients (designated day 0 when endometrial fragments were injected). We used wild type mice as donors and CD206 DTR mice as recipients. One week after endometrial inoculation, we checked the formation of endometriotic-like lesions in the peritoneal cavity by sacrificing a few mice. Then, the recipient mice underwent intraperitoneal injection of DT (DT group) or PBS for control group, every two days. DT of 20 ng/gram body weight was diluted with sterile PBS and injected intraperitoneally 31 . Two weeks after the inoculation of endometrial fragments, the recipient mice were sacrificed through cervical dislocation under anesthesia. Then, PBS (1 ml) was injected into the peritoneal cavity. After vigorous shaking, peritoneal fluid (PF) and PF cells were collected. Laparotomy was performed, and the numbers of endometriotic foci were counted. Each focus and uterus was excised to exclude as much normal surrounding tissues as possible. Immediately, the weight of the excised tissues was measured. During all the inspection procedures, examiners were blinded to the treatment given to each mouse.

Reverse transcription (RT) and quantitative real-time polymerase chain reaction (PCR) analysis.
Total RNA was extracted from mouse endometriotic-like lesions using the ISOGEN-II (NIPPON GENE, Tokyo, Japan) according to the manufacturer's instructions. About 1 μg of total RNA was reverse-transcribed using Rever Tra Ace qPCR RT Master Mix with gDNA Remover (TOYOBO, Tokyo, Japan). For the quantification of various mRNA levels, real-time PCR was performed using the Mx3000P Real-Time PCR System (Agilent Technologies, CA, USA) according to the manufacturer's instructions 31 . The PCR primers were selected from different exons of the corresponding genes to discriminate PCR products that might arise from possible chromosomal DNA contaminants. The SYBR Green thermal cycling conditions were one cycle of 95 °C for 30 s,

Human peritoneal fluids MΦ (PF MΦ) and culture of human endometriotic stromal cells (ESCs).
The experimental procedures were approved by the institutional review board of Kitasato University (approved number: B18-265), and University of Toyama (Approved Number: 25-44) and signed informed consent for the use of samples was obtained from each patient. PFMΦs were purified as previously described 35 . Culture of primary ESCs was described elsewhere 7 . Human samples were taken and used under each patient's consent. All experimental methods were carried out in accordance with relevant guidelines and regulations of Declaration of Helsinki.
Statistical analysis. Data were evaluated by Mann Whitney using Jump version 10. P less than 0.05 was accepted as statistically significant.

Results
Endometriosis model using CD206 DTR mice. As shown in Fig. 1, we induced endometriotic-like lesions using CD206 DTR mice. After one week of inoculation of endometrial fragments into the peritoneal cavity, we started to deplete CD206+ MΦs for a week and measured the total weight and the number of endometriotic-like lesions per mouse. As shown in Fig. 2, in CD206 DTR, the depletion treatment of CD206+ MΦ significantly decreased the total weight of endometriotic-like lesions (p < 0.05), but the number of lesions per mouse was not changed compared to control (1.4 ± 0.5 vs 1.2 ± 1.1). We checked the depletion of CD206+ MΦ in peritoneal micro-environment by quantitative-PCR and found that in peritoneal fluid cells, about 80% of CD206+ MΦ were reduced in CD206DTR mice with DT administration (supplemental Fig. 1). The mRNA expression of CD11b, which is a pan MΦ marker, was not decreased significantly with DT stimuli, suggesting that the amount of total MΦs was not changed (Supplemental Fig. 1). As for other types of MΦs, the ratio of iNOS, a classical M1 MΦ marker, to CD11b was increased significantly, but the ratio of inflammatory MΦs producing TNFα to CD11b was not increased significantly (Supplemental Fig. 1).
Depletion of CD206+ MΦ lead to a histological change of the endometriotic-like lesion. Endometriotic lesion is composed by glandular epithelial cells and stromal cells. In the DT group, the appearance of endometriotic glandular epithelial cells was thinner, (Fig. 3b,d) compared to the appearance of epithelial cells in control (Fig. 3a,c). These surface cells covering lesions were confirmed to be epithelial cells by cytokeratin staining (Fig. 3e,f). Figure 3g,h show the negative control of Fig. 3e,f, respectively. www.nature.com/scientificreports/ The positive ratio per total cells of proliferation marker, Ki-67 + cells, was significantly decreased in both the glandular epithelial cells and stromal cells of endometriotic-like lesions of DT group, receptively (1.6 ± 1.8 and 2.3 ± 1.2%, mean ± SD) compared to control (9.7 ± 13.4% and 3.3 ± 1.8, p < 0.001) (Fig. 4a-d), suggesting that lower proliferation ratio of endometriotic-like cells resulted in a decrease of endometriotic-like lesions in the absence of CD206+ MΦ.   Fig. 5a, the depletion of CD206+ MΦ significantly decreased the expression of CD206 mRNA by more than 90% compared to control (p < 0.01) in the lesions. Although the levels of TNFα and IL-1β were comparable in control and DT group (,c), the mRNA expression of CD11b, a pan-MΦ marker, was not changed (Fig. 5d). The proportion of CD206/CD11b was significantly decreased in DT group, while that of TNFα/CD11b was not changed in both groups (data not shown). The expression of VEGFA (Fig. 5e) and TGFβ1 (Fig. 5f) in DT group were significantly decreased to the levels of approximately 50% and 30% of control, respectively (p < 0.05, p < 0.05).

Discussion
In endometriosis, it has been reported that M2 MΦs are dominant in endometriotic lesions 13 36 . Therefore, in the endometriosis mouse model, M1 to M2 MΦ shift occurs on day 7 after the inoculation of endometrial tissue. This is consistent with a transition from classical M1 MΦ activity to an alternate M2 profile, which correlates to the findings of initially acute inflammation followed by tissue remodeling in the process of development of endometirosis 36 . Given that some M2 MΦs have a role of suppressing inflammation 37 , the contribution of M2 MΦ to the endometriosis-like lesions is uncertain, suppressive, or progressive. To address this question, we used CD206 DTR mice and depleted CD206+ MΦs exclusively in mice from day 7 after induction of endometriotic-like www.nature.com/scientificreports/ lesions, when the shift from M1 to M2 MΦ could occur in endometriotic lesions 36 . CD206 is known to be a type 1 mannose receptor and one of M2a subtype markers in M2 MΦs. These MΦs secrete cytokines such as TGFβ1 and IL-10 and chemokines such as CCLs 37 . In CD206 DTR mice model, CD206+ MΦs were depleted by more than 80% in endometriotic-like lesion and PF cells ( Fig. 5a and Supplemental Fig. 1). At mRNA levels, although the ratio of iNOS/CD11b was increased in PF cells, the mRNA expression of CD11b was unchanged, and there was no increase of TNFα/CD11b, suggesting that inflammatory MΦs which exacerbate endometriotic lesions 6,38 did not increase in both PF cells and lesions. We found that the depletion of CD206+ MΦ resulted in the decrease of the size of endometriotic-like lesions without affecting the number of lesions, implying that the role of CD206+ MΦ is an accelerator of endometriotic-like disease in mice model. Duan et al. used CD11b DTR mice in which pan MΦs could be depleted with DT treatment, and showed the reduction of lesion weight in an endometriosis mice model. Moreover, they also presented that the adoptive transfer of M2a, a subtype of M2 MΦ, systemically after the MΦ depletion significantly increased the weight of endometriotic lesions 39 , suggesting that M2a MΦ play a role in the progression of the endometriotic lesion. In consistent with the notion, as CD206 is one of M2a subtype markers in M2 MΦs 37 , our present results also suggest that CD206+ MΦs are involved in the exacerbation of endometriosis. In the DT group, endometriotic lesions exhibited thinner appearance, and lower proliferation rate of epithelial and stromal cells compared to control (Fig. 4), suggesting that CD206+ MΦs influence the growth of endometriotic cells. Among endometriosis-related factors, in the DT group, the levels of TGFβ1 and VEGFA mRNA, known as angiogenic factors 40,41 , were decreased in endometriotic-like lesions.
In accordance with the notion, CD31, an endothelial cell marker, was almost disappeared in the endometrioticlike lesions with the depletion of CD206+ MΦ. It has been reported that the administration of bevacizumab, a VEGF-A antibody, resulted in reduced lesion formation in mouse endometriosis model studies 41 , suggesting neovascularization of ectopic endometrial tissue is crucial in the development of the endometriotic-like lesion.
In the present study, immunohistochemical analysis showed that CD206+ MΦs expressed TGFβ1 (Fig. 7) and produced high level of VEGFA in the endometriotic-like lesions (Fig. 6). Therefore, we performed an in vitro study to examine these relationships. It is well known that IL-33, an alarmin, deviates macrophage to classical M2 type [42][43][44][45] . We also have reported that IL-33 induced CD206+ MΦs in human peritoneal MΦs 35 . In the present study, we found that induced CD206+ MΦs with recombinant IL-33 stimuli (100 ng/ml, 8hrs) increased the expression of TGFβ1 mRNA (Supplemental Fig. 2). In addition, to investigate the relationship between TGFβ1 derived from CD206+ MΦ and VEGFA of the endometriotic lesion, we checked the mRNA expression of VEGFA in human endometriotic stromal cells (ESCs) with TGFβ1 stimulation and found the significant increase of VEGFA mRNA expression (Supplemental Fig. 2) suggesting that increased expression of TGFβ1 in www.nature.com/scientificreports/ CD206-skewed-MΦ may induce angiogenesis via increasing VEGFA expression in ESCs. TGFβ1 is related to angiogenesis 40 and is also reported to induce VEGFA in other cells 46 . Therefore, the reduction in TGFβ1 expression by the depletion of CD206+ MΦ might have led to the decrease in angiogenesis, resulting in the reduction of endometriotic-like lesion.
In endometriosis patients, it is reported that Natural Killer (NK) cells impair their function in the intraperitoneal environment, which makes endometriotic cells more likely to survive and proliferate in the abdominal cavity. TGFβ1 is known to be increased in the peritoneal fluids of patients with endometriosis 47 and to exacerbate endometriosis by decreasing NK cell activity 48 . Therefore, decreasing of TGFβ1 by the depletion of CD206+ MΦ may contribute to the recovery of NK activity and result in the reduction of endometriosis lesions. Further studies are necessary to clary the role of CD206+ MΦ in the pathogenesis of endometriosis.
In the present study, we proved that CD206+ MΦ played a vital role in the promotion of endometriosis via inducing angiogenesis using CD206 DTR mice. Therefore, CD206+ MΦ might be a target for a new therapy for  (c and d), and a marker of the endothelium, CD31 (e and f), were examined in the endometriotic like lesions of control mice (control) and CD206+ MΦ depletion mice (DT). H-scores of VEGFA (g), TGFβ1 (h), and CD31 (i) immunostaining in endometriotic-like lesions were calculated and compared in control and DT mice. Data were evaluated by Mann Whitney U. P less than 0.05 was accepted as statistically significant.  CD206 (b), a marker of macrophage (MΦ), were used. As a second antibody, the gout anti-rabbit antibody was used. 4′,6-diamidino-2-phenylindole (DAPI) was used to detect nuclei. In the serial section, each cell marked in the same shape was identical. Rabbit IgG was used instead of the primary antibody for negative control (c).