Abstract
The tumor associated microenvironment is known to play a vital role during the development and progression of different malignant tumors. As a part of tumor microenvironment, tumor associated macrophages (TAMs) are crucial for the genesis, proliferation, metastasis, and survival of tumor cells. Recently, more and more studies showed that TAMs were related with poor clinical status and survival in patients with diffuse large B cell lymphoma (DLBCL). Considering the complex roles which TAMs play in the tumor microenvironment of DLBCL, the aim of this study was to review the biological mechanisms between TAMs and DLBCL cells, including extracellular matrix remodeling and angiogenesis promotion, tumor promotion, immune suppression, and phagocytosis inhibition. This review will help us to further understand the comprehensive impact of TAMs on DLBCL and explore possible prognostic markers and therapeutic targets.
Introduction
D
Macrophages, as one of the major compositions of the tumor microenvironment (Yang and Zhang, 2017), were divided into two subgroups as follows: M1 and M2. M1 is a suppressive type of macrophages, which expresses both CD68 and HLA-DR (Baj-Krzyworzeka et al., 2007). It inhibits the growth of tumor by producing reactive oxygen species (ROS), active nitrogen intermediates, and tumor necrosis factor. M2, known as tumor associated macrophages (TAMs) and expressing both CD68 and CD163, promotes the growth and drug resistance of tumor cells through promoting angiogenesis, metastasis, and immunosuppression (Ruffell et al., 2012). By double staining of CD68+ and CD163+, more and more studies showed that TAMs were highly associated with the adverse outcome of patients with DLBCL, even in the era of rituximab (Wada et al., 2012; Marchesi et al., 2015).
Herein, we summarized the biological mechanism models of TAMs on the progression and drug resistance of DLBCL cells, including extracellular matrix (ECM) remodeling and angiogenesis promotion, tumor promotion, immune suppression, and phagocytosis inhibition. The understanding of these complex mechanisms would help us to explain the poor outcome of DLBCL and provide new prognostic markers and therapeutic targets in the future.
ECM Remodeling and Angiogenesis Promotion
Matrix metalloprotein-9
TAMs, as a special subtype of macrophages, secrete lots of active factors, including adhesion molecules, chemokines, and cytokines, to regulate the biological features of DLBCL. Matrix metalloprotein-9 (MMP-9), one of the matrix metalloproteinases, is secreted by TAMs in DLBCL (Lenz et al., 2008; Yoshida et al., 2013). MMP-9 can disrupt the ECM and stimulate tumor metastasis (Liu et al., 2012) (Fig. 1). MMP-9 is also involved in the process of angiogenesis by cleaving fibrillar type I collagen, enhances growth factor-induced angiogenesis (Asha Nair et al., 2003), recruits precursor cells from the circulation for vasculogenesis (Ahn and Brown, 2008), and promotes the releasing of vascular endothelial growth factor (VEGF) and the growth of endothelial cells (Bergers et al., 2000). What's more, MMP-9 could increase the level of soluble interleukin-2 receptor α (sIL-2Rα) in the serum by cleaving IL-2Rα of chains and then inhibits the antitumor immunity by blocking the activation of T cells (El Houda Agueznay et al., 2007; De Paiva et al., 2009; Yoshida et al., 2013; Sakai and Yoshida, 2014). That's why the serum IL-2Rα can indirectly reveal the antitumor activity and imply the outcome in the DLBCL patients (Ennishi et al., 2009).
The biology of TAMs in DLBCL. The biological mechanisms include ECM remodeling and angiogenesis promotion (red), tumor promotion (blue), immune suppression (yellow), and phagocytosis inhibition (purple). (1) TAMs remodel ECM remodeling and promote angiogenesis through producing MMP-9, legumain, and VEGF. (2) pre-HGF from TAMs is activated by activator from DLBCL, which can then promote angiogenesis and tumor progression through HGF/c-MET pathway. ROS from TAMs contributes to DLBCL progression and drug resistance by increasing the expression of CD44. (3) RCAS1 expressed by TAMs is cleaved by MMP-9, and soluble RCAS1 will induce the apoptosis of natural killer cells and normal B cells. IL-27 heterodimer, formed by macrophage derived IL-27p28 and IL-27B produced by lymphoma cells, contributed to the overexpression of PD-L1 on macrophages in lymphoma through an IL-27/Stat3 axis. TAMs express PD-L1, which induces the apoptosis of T cells. (4) DLBCL cells can avoid phagocytosis through CD47/SIRPα pathway. Death of DLBCL cells turns M1 into TAMs through self-produced galectin-3. c-MET, c-mesenchymal–epithelial transition; DLBCL, diffuse large B cell lymphoma; ECM, extracellular matrix; HGF, hepatocyte growth factor; IL, interleukin; MMP-9, matrix metalloprotein-9; PD-L1; programmed death-1 ligand; ROS, reactive oxygen species; RCAS1, receptor-binding cancer antigen expressed on SiSo cells; SIRPα, signal regulatory protein α; TAM, tumor associated macrophage; VEGF, vascular endothelial growth factor. Color images available online at
Legumain
Legumain is an aspartate endonuclease and can hydrolyze asparagine in protein and small molecule substrates. Legumain is highly expressed in the tumor microenvironment (Mai et al., 2016) and involves in invasion and metastasis in a variety of tumors (Liu et al., 2003). TAMs can catabolize fibronectin and collagen I and stimulate angiogenesis by overexpressing legumain, which contributes to the progression of the DLBCL (Shen et al., 2016) (Fig. 1). The overexpression of legumain is relevant to advanced clinical stages and poor prognosis in many tumors (Zhen et al., 2015). A cell-impermeable doxorubicin-based prodrug (leg-3) can selectively kill legumain-expressing TAMs and inhibit the tumor growth and metastasis in murine breast carcinoma models (Lin et al., 2013). Inhibition of legumain is also proved to suppress the progression of DLBCL in a murine model (Shen et al., 2016). Legumain can be a novel potential biomarker to predict the prognosis and may guide the treatment.
Vascular endothelial growth factors
Macrophages synthesize and secrete a high level of VEGF during inflammatory reaction (Cho et al., 2001), wound healing (van der Plas et al., 2009), and hypoxia (Wu et al., 2010). TAMs can help tumor cells to increase the production of VEGF (Torisu et al., 2000), which boosts the growth of vascular endothelial cells, increases microvascular density (MVD) in the tumor microenvironment (Fig. 1), and eventually promotes the progression of DLBCL. The elevation of serum VEGF and MVD at diagnosis predicts poor response and survival in DLBCL patients treated with chemoimmunotherapy (Duletic-Nacinovic et al., 2016). CD31 and CD34 can be used as MVD markers to evaluate the clinical and prognostic significance of MVD in different tumors (Nico et al., 2008; Cardesa-Salzmann et al., 2011; Perry et al., 2012; Gomez-Gelvez et al., 2016). The inhibitor of VEGF and endothelial growth factor receptor tyrosine kinase inhibitor are increasingly used to treat different tumors. They may be also the potential targets for the treatment of DLBCL in the future.
Tumor Promotion
Hepatocyte growth factor/c-mesenchymal–epithelial transition pathway
Hepatocyte growth factor (HGF) is a protein tyrosine kinase. Pre-HGF is produced by TAMs and stimulated by DLBCL cells into active form (Tjin et al., 2006). c-Mesenchymal–epithelial transition (c-MET), the receptor of HGF, is mainly expressed on DLBCL and vascular endothelial cells. An uncontrolled activation of HGF/c-MET pathway has been confirmed related to adhesion, proliferation, survival, and metastasis in many kinds of lymphomas (van der Voort et al., 2000; Teofili et al., 2001; Trusolino and Comoglio, 2002) (Fig. 1). HGF can also upregulate the expression of VEGF, which stimulates angiogenesis in DLBCL (Jiang et al., 1999; Xin et al., 2001; Lin et al., 2012). The inhibition of c-Met prevents cell proliferation and induces mitochondrial and caspase-dependent apoptosis in DLBCL cell lines (Uddin et al., 2010). Clinical studies have shown that MET gene copy number and serum HGF level are related with treatment response and outcome of DLBCL (Hsiao et al., 2003; Giles et al., 2004; Huang and Chuang, 2013).
Reactive oxygen species
Both M1 and M2 polarization macrophages produce ROS during activation (Brune et al., 2013; Mendoza-Coronel and Ortega, 2017). An excessive accumulation of ROS induces the apoptosis of DLBCL through DNA damage response (Xu et al., 2016). However, moderate level of ROS supports stem cell proliferation, differentiation, and mobilization, which contribute to the progression of stem-associated diseases (Chaudhari et al., 2014). Macrophage-derived ROS has been proved to increase the expression of CD44 in glioblastoma and gastrointestinal cancer (Ishimoto et al., 2014) (Fig. 1), which promotes tumor growth and chemoresistance in tumor cells by stabilizing the xCT subunit of system xc(−) and then enhances the synthesis of glutathione (GSH), an antioxidant capable of preventing damage caused by ROS (Singer et al., 2015). Our previous study also showed that high expression of CD44 was associated with a poor prognosis in patients with DLBCL (Wei et al., 2014).
Immune Suppression
Programmed death-1/programmed death-1 ligand pathway and IL-27
The pathway of programmed death-1 (PD-1)/PD-1 ligand (PD-L1) is important for the balance of self-tolerance and the control of excessive immune responses, including inhibiting antitumor immunity (Keir et al., 2008; Pardoll, 2012). The high expression of PD-L1 in tumor microenvironment has been proved to be an adverse prognosis in brain glioma (Wang et al., 2016), squamous cell carcinoma, and DLBCL (Hirai et al., 2017). Blockade of PD-L1 efficiently inhibits tumor growth and enhances survival in squamous cell carcinoma, HL, melanoma, nonsmall-cell lung cancer, and renal-cell cancer (Tsushima et al., 2006; Brahmer et al., 2012; Mahoney et al., 2015; Villasboas and Ansell, 2016; Hirai et al., 2017). IL-27 heterodimer, formed by macrophage derived IL-27p28 and IL-27B produced by lymphoma cells, contributes to the immune-suppressive environment by overexpression of PD-L1 on macrophages in lymphoma microenvironment through an IL-27/Stat3 axis (Karakhanova et al., 2011; Matta et al., 2012; Horlad et al., 2016) (Fig. 1). Double immunostaining of PD-L1 and CD68 in the specimen of DLBCL and classical HL (cHL) demonstrates that PD-L1 is mainly expressed in infiltrating macrophages in the microenvironment and little in tumor cells (Chen et al., 2013; Kwon et al., 2016). Although the proportion of DLBCL patients expressing PD-L1 is low, overexpression of PD-L1 or PD-1 is related to a poor outcome (Kiyasu et al., 2015; Ansell, 2016). The first approved anti-PD-1 agent Nivolumab has showed a promising clinical outcome in relapsed/refractory cHL (Juarez-Salcedo et al., 2017). The monoclonal antibody of anti-PD-1 or anti-PD-L1 would also be a potential choice for the treatment of DLBCL.
Receptor-binding cancer antigen expressed on SiSo cells
Receptor-binding cancer antigen expressed on SiSo cells (RCAS1) is a tumor associated antigen, which is expressed in many tumors. RCAS1-positive macrophages are higher in the DLBCL tissue than in the stroma and they contribute to tumor growth by acting against the immune activity (Nagamatsu and Schust, 2010; Kazmierczak et al., 2015). MMP-9 that we mentioned above is involved in the ectodomain shedding of RCAS1 (Sonoda and Kato, 2014). Soluble RCAS1 is mainly responsible for the tumor escaping from host immunological surveillance by inducing the apoptosis of peripheral lymphocytes and natural killer cells and creating a suppressive tumor microenvironment (Sonoda et al., 2010) (Fig. 1). The RCAS1-positive macrophages are correlated with poor prognosis in the cervical cancer nest and hydatidiform mole, and the high level of RCAS-1 is related to the adverse outcome in different tumors (Basta et al., 2011; Sonoda, 2011; Galazka et al., 2012).
Phagocytosis Inhibition
CD47/signal regulatory protein α pathway
Phagocytosis is one of the major antitumor functions of macrophages. Normal hematopoietic cells evade phagocytosis by expressing CD47, which binds to the signal regulatory protein α (SIRPα) on macrophages (Weiskopf et al., 2013). However, DLBCL cells can weaken the antitumor function of macrophages by increasing the expression of CD47 (Fig. 1). High CD47 expression is significantly associated with high risk of death in DLBCL patients. The anti-CD47 monoclonal antibody synergizes with rituximab to promote phagocytosis and eradicate tumor cells in murine DLBCL models (Chao et al., 2010; Goto et al., 2014; Weiskopf et al., 2016).
Galectin-3
Galectin-3 is a glycoprotein and is highly expressed on the surface of macrophages. Plenty of cells die during the tumor growth process because of the imbalance between proliferation and death (Gregory and Pound, 2011). Galectin-3 participates in the clearance of apoptotic cells (Sano et al., 2003). But galectin-3 overexpression switches macrophages from an antitumor subtype to a TAM phenotype in aggressive B cell lymphoma (Ford et al., 2015; Voss et al., 2017) (Fig. 1). Galectin-3 also plays an important part on tumor progression, metastasis as well as resistance to therapy by inducing secretion of metastasis-promoting cytokines from vascular endothelial cells (Colomb et al., 2017), and impairs the tumor suppressive effects of miR-128 (Lu et al., 2017). A high level of circulating galectin-3 is a predictor of poor prognosis of many tumors (De Iuliis et al., 2017; Li et al., 2017).
Conclusion
More and more clinical studies showed that high TAMs in microenvironment were associated with poor outcome in patients with DLBCL. Given the profound impacts of TAMs on tumor progression and chemoresistance, there has been increasing interest in the biology of TAMs in DLBCL and the development of direct or indirect therapeutic inhibitors. The underlying mechanisms of TAMs impact on DLBCL were involved in ECM remodeling and angiogenesis promotion, tumor promotion, immune suppression, and phagocytosis inhibition. Currently antitumor therapies directed at TAMs are mostly based on the following strategies: macrophage recruitment inhibition, TAM depletion, M1 polarization promotion, and M2 polarization and activity inhibition. The colony stimulating factor 1 receptor inhibitors preventing macrophage recruitment and eliminating TAMs have shown encouraging preliminary results in refractory/relapse cHL and other tumors. Future studies would be directed at specific drugs targeting TAMs for better understanding of the role of TAMs on DLBCL. With the rapid advancement of molecular biology and emerging strategies for disrupting TAMs and DLBCL interactions, it will be promising that selective TAM inhibitors can be developed to improve the outcome of DLBCL in the future.
Footnotes
Acknowledgments
This work was supported by the National Natural Science Foundation of China (Grant Nos. 81600165, 81670183) and the Medical Science and Technology Project of Guangdong Province (Grant No. A2018120).
Disclosure Statement
No competing financial interests exist.