It is known that in the early stages of tumor colonization or growth, immune cells and related matrix components form a tumor suppressive inflammatory microenvironment, which hinders tumor progression. However, prolonged exposure to tumor antigens and immune activation can lead to exhaustion or remodeling of the relevant effector cells, creating an immunosuppressive microenvironment (Nakamura and Smyth 2020). TAMs are key players in this process, linking inflammation and tumor progression. By targeting TAMs with immunotherapy strategies, it is possible to reshape the tumor immune microenvironment and elicit a comprehensive antitumor response (Pitt et al. 2016). In breast cancer microenvironment, particularly in TNBC, TAMs, especially M2-like macrophages, have been broadly studied for their effect in tumor growth, metastasis and evasion of immune surveillance (Jin et al. 2022; Mehta et al. 2021). Our study revealed that M2 macrophages and their secreted factor, CXCL1, induce PD-L1 expression in TNBC cells by the CXCR2/PI3K/AKT/NF-κB signaling pathway. The use of LY294002 and Bay11-7082, which inhibit this pathway, reversed the increased PD-L1 expression induced by M2 macrophages or CXCL1. These findings highlight the significant role of CXCL1-producing M2 macrophages in TNBC progression and their contribution to a poorer prognosis by creating an immunosuppressive microenvironment. Targeting this pathway may represent a potential therapeutic strategy for TNBC. It is important to further explore and develop treatments that specifically target TAMs and their associated signaling pathways to improve outcomes for TNBC patients.
TAMs play a significant role in suppressing effective adaptive immunity. Extensive evidence has highlighted their crucial involvement in immune evasion in several cancers, such as gastric cancer (Lin et al. 2019), colorectal cancer (Wei et al. 2021) and melanoma (Jarosz-Biej et al. 2018). TAMs are primarily of the M2 macrophage subtype, which actively promotes tumor development. M2 macrophages produce immune-suppressive factors such as IL-10, transforming growth factor-β (TGF-β), and various chemokines like CCL2, CCL5, CCL20 and CCL22. These factors directly inhibit anti-tumor immune responses and facilitate regulatory T cells recruitment into the TME (Li et al. 2021; Liu et al. 2020). Additionally, TAMs express immunoregulatory proteins like PD-L1, leading to dysfunction of T cells. For instance, in gastric cancer, TAMs promote their own high expression of PD-L1 through secreting CXCL8, which subsequently inhibits the activity and reduces infiltration of CD8+ T cells (Lin et al. 2019). In glioblastoma, bioinformatic analysis have also confirmed that PD-L1-mediated immunosuppression is related to the infiltration and M2 polarization of TAMs (Zhu et al. 2020). In our study, we discovered that M2 macrophages actively promote the high expression of PD-L1 on the tumor cells surface, facilitating their immune escape and ultimately leading to poorer prognosis. These findings further emphasize the crucial link between TAMs and the modulation of PD-L1 expression and function in the TME.
M2-like TAMs exhibit various functions that hinder anti-tumor immune responses through multiple mechanisms, including the cytokine-cytokine receptor interaction pathways. These TAMs can be recruited by various cytokines like CCL2, CCL5, CSF-1 and VEGF, which are abundantly present in the TME. Once in the TME, TAMs predominantly display the M2 phenotype and secrete a range of cytokines, chemokines, and proteases that activate signaling pathways, promote tumor inflammatory responses, and exacerbate tumor progression and metastasis (Noy and Pollard 2014). In our bioinformatics analysis, we observed up-regulation and down-regulation of numerous cytokines in monocyte-derived macrophages polarized by TNBC cell conditioned medium, including CXCL1, CXCL2, IL1B, CCL3, IL5 and CXCL9. Further qPCR and ELISA experiments confirmed higher CXCL1 expression in M2 macrophages compared to M0 macrophages or TNBC cells. CXCL1 is known to promote angiogenesis in various solid tumors, including bladder (Miyake et al. 2016) and gastric cancers (Z. Wang et al. 2017). Primary colorectal cancer cells can generate VEGF-A, stimulating TAMs to secrete CXCL1, which promotes tumor angiogenesis, revascularization, and recurrence after chemotherapy in mouse tumor models (D. Wang et al. 2017). Our study identified M2 macrophages as the dominant source of CXCL1 in TNBC, which is in line with previous research. They isolated TAMs from breast tumors in mice with MMTV- PyMT+/−. Cytokine arrays revealed the highest expression of CXCL1 in the supernatants of TAMs (Wang et al. 2018). It was also found that CSF-2 facilitates macrophage-derived CXCL1 secretion, although CSF-2 was previously known to facilitate the differentiation of infiltrating monocytes into M1 macrophages (Noy and Pollard 2014). The phenotype and function of TAMs are heterogeneous and not solely limited to M1 or M2 phenotypes. The CSF-2 positive feedback loop is a complex network that induces an inflammatory chain reaction involving other pro-inflammatory factors such as TNF, IL-1, IL-6, IL-12 and IL-23, which may also contribute to CXCL1 expression. While our study did not experimentally verify the role of CSF-1 in macrophage polarization, Filardi et al analyzed gene expression profiling in CSF-2- and CSF-1-polarized macrophages revealed that a high CCL2 expression characterizes macrophages generated under the influence of CSF-1, whereas CCR2 was expressed only by CSF-2-polarized macrophages (Sierra-Filardi et al. 2014). This may explain the high expression of the CSF-2 receptor CSF2RB in breast cancer tissues with high CXCL1 expression, as observed in our GSEA analysis. Overall, our findings illustrate that M2 macrophages play a vital role in promoting the expression of CXCL1 in TNBC, and this mechanism may contribute to tumor progression and lead to poorer prognosis by inducing an immunosuppressive microenvironment.
Based on the fact that M2 macrophages are the main source of CXCL1 in TNBC, we investigated whether CXCL1 derived from M2 macrophages contributes to the high expression of PD-L1 in TNBC. Previous research have demonstrated that CXCL1 and its receptor CXCR2 are involved in the activation and trafficking of granulocytes and neutrophils to sites of inflammation. In cancer, CXCL1 binding to CXCR2 primarily promotes neutrophil recruitment, tumor cell invasion and metastasis. CXCL1 plays an important part in the complex tumor microenvironment, and its high expression is significantly correlated with breast cancer lymph node metastasis, poor overall survival and the molecular subtype being TNBC (D. Wang et al. 2017; Wang et al. 2018). CXCL1 expression is the highest in TNBC, followed by normal tissues and the lowest in luminal subtypes (Li et al. 2020). Parallel networks of CXCL1 and Mincle signaling induced by necropsies promote macrophage-induced adaptive immunosuppression, thereby enabling pancreatic ductal adenocarcinoma progression (Seifert et al. 2016). Meanwhile, CXCL1 could also play a role in recruiting immunosuppressive cells, attracting myeloid-derived suppressor cells toward gastric cancer (Zhou et al. 2022) and induce the aggregation of naive CD4+ T cells in breast cancer (Li et al. 2021). Inflammatory cytokines can regulate PD-L1 expression and function in tumor cells through multiple levels and pathways. These cytokines include interferon (IFN), transforming growth factor (TGF), tumor necrosis factor (TNF) and interleukin (IL) family molecules. However, there have been few studies on the relationship between chemokines and PD-L1 in cancer. One study revealed that patients with high CXCL scores expressed higher levels of PD-L1 and had better survival outcomes compared to patients with low CXCL scores who received anti-PD-1 therapy (Li et al. 2023). In our study, we observed that M2 macrophages secreted CXCL1, which bound to CXCR2 on the surface of tumor cells. The existence of CXCL1 resulted in a concentration-dependent increase in PD-L1 expression in MDA-MB-231 and SUM159 cells. Furthermore, the upregulation of PD-L1 induced by M2 macrophages was reversed when CXCL1 or CXCR2 was silenced. The signaling pathways associated with the CXCL1/CXCR2 axis in tumor cells are not fully determined, but tumor cells generally upregulate PD-L1 expression through various pathways, including EGFR, MAPK, PI3K-AKT or NF-κB pathways (Yi et al. 2021). Multiple oncogenic signals could impair immune surveillance by activating the NF-κB-PD-L1 axis. Additionally, activation of the PI3K-AKT pathway in gastric cancer cells increased PD-L1 abundance, while PI3K inhibitors reduced PD-L1 levels (Kim et al. 2019). In our study, co-culturing with M2 macrophages or treating cells with rhCXCL1 led to enhancive expression of p-AKT, p-p65 and nucleus p65 protein in MDA-MB-231 and SUM159 cells. When the PI3K-AKT signaling pathway was inhibited using LY294002, PD-L1, p-AKT, p-p65 and nucleus p65 protein expression was markedly decrease, suggesting that the PI3K/AKT/NF-κB signaling pathway is involved in CXCL1/CXCR2-mediated regulation of PD-L1 expression in TNBC cells. Overall, our findings highlight the significance of CXCL1 derived from M2 macrophages in promoting PD-L1 expression in TNBC cells through the CXCL1/CXCR2 axis, and the involvement of the PI3K/AKT/NF-κB signaling pathway in this process.
Several studies have highlighted the potential benefits of targeting the CXCR2 pathway in various cancers. Huang and his team investigated a drug called navarixin, a CXCR2 inhibitor, and demonstrated that when combined with the drug enzalutamide, it effectively killed hormone-resistant prostate tumors, whereas enzalutamide alone had no significant effect (Li et al. 2019). This suggests the potential of using small molecule inhibitors that target the CXCL1/CXCR2 axis for clinical applications. While our study confirmed cancer-promoting role of M2 macrophages through the excitation of the CXCL1/CXCR2 axis using gene manipulation and animal experiments, we did not develop or apply specific drug targets in this study. Further research is warranted to explore potential treatment strategies in this context. The development of therapies targeting M2 macrophages and CXCL1/CXCR2 axis could hold promise for TNBC immunotherapy and may open new avenues for improving treatment outcomes. Additional studies are required to advance the understanding and application of these therapeutic approaches in the future.