Endometrial trauma often causes intrauterine adhesion (IUA) (Deans and Abbott, 2010), also known as Asherman syndrome, which may leads to infertility. IUAs reduce endometrial receptivity, which is crucial for successful implantation and pregnancy. Therefore, improving endometrial receptivity is the ultimate goal of the treatment of IUAs (Heger et al., 2012). At present, the main treatment for IUAs is hysteroscopic adhesion decomposition, but it has a high recurrence rate and poor prognosis (Valle and Sciarra, 1988). To improve the effectiveness of IUA therapy and reduce the recurrence rate, diverse methods have emerged in recent years, such as estrogen therapy (Roy et al., 2014), physical barriers (intrauterine device or Foley catheter) after hysteroscopy (Pabuccu et al., 2008), biological barriers (sodium hyaluronate gel (Acunzo et al., 2003), amniotic membrane (Zheng et al., 2018)) and stem cell therapy (Santamaria et al., 2016). In addition, studies have reported that granulocyte colony stimulating factor (G-CSF) (Gleicher et al., 2013) has a beneficial effect on endometrial growth, but little is known about the underlying mechanism.

Macrophages are important components of the innate immune system (Chambers et al., 2020), acting as regulators of inflammation, tissue repair, regeneration, and fibrosis (Wynn and Vannella, 2016). They not only play an important role in periodic regeneration of the endometrium during the non-gestation period but are irreplaceable during the gestation process, including early embryo implantation, placenta formation and fetal delivery (Chambers et al., 2020). Macrophages are heterogeneous and plastic and can exhibit different phenotypes and functions in different tissue microenvironments (Galli et al., 2011). Most studies have classified macrophages as classically activated macrophages (M1) and alternatively activated macrophages (M2). M1 macrophages is induced by LPS and IFN-r, express the cell surface marker cluster of differentiation 86 (CD86) and secrete proinflammatory cytokines, such as inducible nitric oxide synthase (iNOS), TNF, interleukin-6 (IL-6) and interleukin 1 (IL-1) (Ferrante et al., 2013), which are essential for the clearance of bacterial, viral and fungal infections (Takebayashi et al., 2015). They also increase myofibroblast production and promote extracellular matrix deposition and fibrosis (Smigiel and Parks, 2018). M2 macrophages is induced by IL-4 and/ or IL-13, M-CSF (Lacey et al., 2012) and other factors, express cell surface markers such as macrophage mannose receptor (CD206) and CD163 and secrete anti-inflammatory cytokines such as Fizz1, dectin-1, Arg1 (Ferrante et al., 2013) and interleukin 10 (IL-10) (Qi et al., 2016), which play an important role in alleviating inflammation, tissue remodeling and angiogenesis (Smigiel and Parks, 2018). Nevertheless, it has been pointed out that M1 and M2 polarized macrophages are extremes of a continuum of activation states in a universe of adaptive responses (Locati et al., 2020). In vivo, macrophages can be divided into tissue-resident macrophages and monocyte-derived macrophages; tissue-resident macrophages are mainly of the M2 phenotype (Geissmann et al., 2010, Sica and Mantovani, 2012, Mantovani et al., 2013); these macrophage populations are critical for maintaining physiological homeostasis, and monocyte-derived macrophage populations are recruited in large numbers from bone marrow after tissue injury (Smigiel and Parks, 2018). During the process of tissue repair, macrophages with different phenotypes secrete different cytokines to perform their respective functions, and these effects are in dynamic balance with each other. If this balance is broken, the tissue microenvironment will be changed, leading to abnormal tissue healing, scar hyperplasia or fibrosis. Studies have shown that macrophages are involved in the injury, repair, regeneration and fibrosis of various organs and tissues, such as the lung (Zhang et al., 2018), liver (Melino et al., 2016), kidney (Cao et al., 2015) and skin (Shook et al., 2018). Recent studies reported that the proportion of M2-like macrophages was reduced in endometrial tissues of patients with uterine adhesions, and the supernatant of M2-like macrophages could promote the proliferation and migration of epithelial cells (Liu et al., 2020a). These results suggest that macrophages may play an important role in the repair of endometrial injury, and chemotaxis and regulation of local macrophages may promote this process.

GM-CSF, also known as CSF-2, is a member of the colony-stimulating factor family and was originally identified as an inducer of differentiation and proliferation in granulocytes and macrophages derived from hematopoietic progenitor cells (Wicks and Roberts, 2016). GM-CSF is a multifunctional cytokine that is present in a variety of tissue cells and can promote wound healing and tissue repair by enhancing the functions of cells, such as activating monocytes/macrophages, regulating inflammation, promoting keratinocyte proliferation and regulating fibroblast differentiation (Fang et al., 2007, Hu et al., 2020). In assisted reproduction, GM-CSF participates in many key processes, such as maturation of oocytes and sperm, endometrial receptivity, embryo implantation, and development of embryos and fetuses, possibly extending to birth (Coughlan et al., 2014). Recently, it was reported that intrauterine infusion of GM-CSF gel improved the endometrial thickness and clinical pregnancy rate of embryo transplantation in patients with a thin endometrium (Mao et al., 2020). However, the biological mechanism remains unclear. In addition, the effect of GM-CSF on the polarization and function of macrophages is still controversial. GM-CSF acts as an M1 inducer in most contexts (Fleetwood et al., 2007), but some studies have shown that GM-CSF-induced macrophages exhibit some M2 characteristics, with anti-inflammatory effects (Weinhage et al., 2015).

In this study, a mouse model of endometrial injury was established to observe the changes in local uterine macrophages and the repair of the injured endometrium after treatment with GM-CSF to explore the mechanism by which GM-CSF promotes repair of the injured endometrium by regulating macrophages. In addition, we confirmed the effect of GM-CSF on macrophages through in vitro cell experiments to clarify the mechanism of action of GM-CSF.