Mol. Cells 2015; 38(2): 95-104
Published online December 16, 2014
https://doi.org/10.14348/molcells.2015.2298
© The Korean Society for Molecular and Cellular Biology
Correspondence to : *Correspondence: salaha@lsuhsc.edu
The development of cancer has been an extensively researched topic over the past few decades. Although great strides have been made in cancer prevention, diagnosis, and treatment, there is still much to be learned about cancer’s micro-environmental mechanisms that contribute to cancer formation and aggressiveness. Macrophages, lymphocytes which originate from monocytes, are involved in the inflammatory response and often dispersed to areas of infection to fight harmful antigens and mutated cells in tissues. Macrophages have a plethora of roles including tissue development and repair, immune system functions, and inflammation. We discuss various pathways by which macrophages get activated, various approaches that can regulate the function of macrophages, and how these approaches can be helpful in developing new cancer therapies.
Keywords angiogenesis, cancer, invasion, macrophages, microenvironment, migration
Macrophages are important cells of the immune system. They are derived from monocytes which enter various tissues and differentiate into macrophages. In the immune response, they use phagocytosis, a process in which the macrophages engulf pathogens, to protect the human body from disease. In addition to macrophages’ importance to the general functions of the immune system, they also play various roles in cancer. These macrophages, or tumor-associated macrophages (TAMs), interact with cancer cells to cause phagocytosis and lysis of tumor cells. On the other hand, TAMs have also been discovered to be an integral part of the tumorigenic pathway. The balance of cytokines and other molecular signals determines the role of these TAMs and thus it is important to study them closely. However, it is important to note that approximately 80% of studies show an association between TAMs and pro-tumorigenic effects while only 10% of studies show the opposite effect.
Before delving into TAMs’ function in cancer, it is important to understand the general cancer process. Cancer can be broken into three distinct stages. In the first stage, cancerous cells invade an area’s surrounding tissues and blood vessels; this area is referred to as the primary site. In the second stage, the cancer cells begin to move to secondary sites through the circulatory system. In the third and final stage, the cancer cells invade the secondary site’s surrounding tissues and blood vessels. This cycle can then repeat yielding tertiary sites and so forth. Once the primary site is established, tumor development is spurred by angiogenesis. Hanahan and Weinberg (2000), created a model that lays out the six properties that a tumor acquires while growing. These properties include an ability to replicate endlessly, angiogenesis, evasion of apoptosis, creating its own growth signals, insensitivity to growth inhibitors, and tissue invasion and metastasis. A seventh property, tumor microenvironment inflammation, was later added to the model (Hanahan and Weinberg, 2000; Mantovani, 2009) (Fig. 1). Here we discuss the mechanisms by which macrophages get activated and their important role in tumor progression.
Two different types of phenotypes for macrophages have been described; these include: M1 and M2. The M1 phenotype is associated with active microbial killing while the M2 phenotype is associated with angiogenesis and wound repair. TAMs can typically be identified by low expression of tumor necrosis factors (TNF) and high expression of IL-1 and IL-6. Specifically M1 macrophage polarization is caused by interferon gamma (IFN-gamma) and TNF-α secretion by T helper 1 (Th1) lymphocytes or natural killer (NK) cells (Gordon, 2003; Martinez et al., 2009; Murray and Wynn, 2011). Interferon (IFNs) or toll-like receptor (TLR) signals can produce the M1 phenotype through the STAT1 signaling pathway. M1 macrophages also produce chemical attractants, such as pro-inflammatory cytokines, interleukin-12 (IL-12), IL-1β, and IL-6 (Heusinkveld and van der Burg, 2011; Porta et al., 2011). M2 macrophages, on the other hand, are activated by Th2. Th2 releases several cytokines, such as IL-4, IL-10, and IL-13, to activate the M2 macrophages (Gordon, 2003; Murray and Wynn, 2011). The M2 macrophage phenotype is caused by IL-4 and IL-13 via the STAT6 signaling pathways. In addition, Kruppel-like factor 4 (KFL4) works in conjunction with STAT6 to produce M2 genes, while inhibiting M1 genes through isolation of some NF-κB activators. Unlike the M1 phenotype, the M2 phenotype can be divided into 3 more categories: M2a, M2b, and M2c. M2a is differentiated by IL-4 and IL-13 while M2b by immune complexes. M2c is differentiated by IL-10 and also plays a role in immunosuppression and tissue remodeling (Hagemann et al., 2009). Furthermore, it has been shown that Th1 cytokines produce M1 macrophages while M2 macrophages play a role in Th1 adaptive immunity suppression, resolution of inflammation, parasite protection, wound healing promotion, angiogenesis, and tissue remodeling (Biswas and Mantovani, 2010).
Some of the most potent macrophage activators of metastasis involved in Lewis lung carcinoma lead to the production of IL-6 and TNF-α through the activation of essential-to-metastasis TLR2 and TLR6 (Chiodoni et al., 2010; Kim et al., 2009). These macrophages are recruited to tumor cells by the production of monocyte chemotactic protein (MCP), macrophage-colony stimulating factor (M-CSF), vascular endothelial growth factor (VEGF), angiopoietin 1 (ANG1), and ANG2. In addition to macrophage recruitment, VEGF, ANG1, and ANG2 play roles in promoting angiogenesis (Pollard, 2004). Anti-inflammatory molecules, such as IL-4, IL-10, transforming growth factor beta 1 (TGF-β1), and prostaglandin E2 (PGE2) produced by tumor cells cause macrophages to attain the M2 phenotype. Adding IL-10
Looking further into VEGF, it has been determined that it is an angiogenic factor that triggers angiogenesis and lymphangiogenesis. Angiogenic and lymphangiogenic factors include VEGF-A, VEGF-C, VEGF-D, and MMP that all bind to their respective coordinating receptors and play roles in enhancing endothelial cell proliferation, migration, and tube-like formation (Giraudo et al., 2004; Wu et al., 2012). MMP has also been shown to mediate VEGF-A levels in some studies. Specifically, MMPs are proteolytic enzymes tasked with the degradation of proteins within the extracellular matrix. They regulate many cell behaviors within the tumor microenvironment (Egeblad and Werb, 2002).
The importance of lymphangiogenesis must not be understated because it plays a large role in tumor stability and growth even though it occurs second to angiogenesis in a tumor. Lymphangiogenesis exposes the tumor to immune cells leading to lymphatic metastasis and without it, the tumor would likely stop growing and be destroyed (Kurahara et al., 2012). Lymphangiogenesis is typically induced by VEGF-A and VEGF C/D binding with VEGFR-2 and VEGFR-3 receptors. In addition, tissue inflammation in tumors has been shown to expand lymph nodes and lead to increased lymphangiogenesis and lymphoid hyperplasia. Inhibition of M-CSF has been shown to suppress tumor lymphangiogenesis since it plays a role in VEGF-A/C secretion. Interestingly, endothelial nitric oxide synthase (eNOS) and IL-6 also regulate lymphangiogenesis (Ji, 2011) (Fig. 3).
Research has shown that there is a positive correlation between the number of TAMs and advanced tumor progression and metastasis. Many tumors show overexpression of M-CSF and when down-regulated, there is a decrease in macrophage activity. Knocking out MCP has led to lower levels of macrophages and decreased tumor progression as well. These data demonstrate a clear relationship between TAMs and tumorigenesis. Surprisingly, further overexpression of MCP has shown that at a certain point, M1 macrophages are attracted to the site causing cytotoxic effects on the tumor cells (Shirabe et al., 2011).
In addition to the decrease in macrophage activity due to the downregulation of M-CSF, vascularization is also reduced. Without vascularization, larger tumors struggle to grow due to insufficient amounts of oxygen causing hypoxic regions. TAMs compensate for this effect by congregating in these areas causing upregulation of certain transcription factors that lead to the expression of various growth factors, cytokines, and other signaling molecules resulting in angiogenesis (Obeid et al., 2013). These transcription factors cause secretion of VEGF and basic fibroblast growth factor (bFGF) which promote the proliferation and maturation of endothelial cells (Lamagna et al., 2006; Siveen and Kuttan, 2009). TAM created proteases, such as MMPs, plasmin, and urokinase plasminogen activator, degrade and remodel the extracellular matrix leading to further angiogenesis (Siveen and Kuttan, 2009).
The majority of research indicates increased levels of TAM activity result in a negative prognosis. The M1 phenotype is usually present during early stages of tumor development while the M2 phenotype is associated with more advanced tumors. In addition, compounds such as VEGF and M-CSF allow for tumor development to progress by recruiting macrophages to the tumor site and initiating angiogenesis and lymphangiogenesis.
The tumor microenvironment is a niche that has been shown to affect tumor activity and outcome. It is created the moment the tumor becomes attached to a site and encompasses the extracellular matrix surrounding the tumor cell and any non-cancerous cells that happen to reside in the targeted organ (Fig. 2).
Colony stimulating factor 1 (CSF1) and chemokine ligand 2 (CCL2) in the blood recruit monocytes that become TAMs in the tumor microenvironment, a process called intraepithelial neoplasia. In addition to CSF-1, the cytokines VEGF and PDGF and chemokines CCL2, CCL5, and CCL8 all attract monocytes to the tumor microenvironment (Allavena et al., 2008; Goede et al., 1999; Mantovani et al., 2008; Pollard, 2004) (Fig. 3). It may be important to note that CCL2 upregulation is caused by a
It has been reported that some transcription factors also regulate TAM functions. Nuclear factor-κB (NF-κB) is activated by TNF-α, IL-1, and IKKβ. NF-κB refers to a family of transcription factors that plays a role in inflammation and immunity. NF-κB operates by causing the expression of pro-inflammatory cytokines, chemokines, and angiogenic factors (Porta et al., 2011). Thus NF-κB is an important part of the regulation of TAM function. Transcription factors for NF-κB include Re1A (p65), c-Rel, RE1B, p50, and p52. Out of those transcription factors, only Re1A, Re1B, and c-Rel can activate gene transcription (Hagemann et al., 2009). NF-κB is activated primarily by IKKβ (Maeda et al., 2005). Another activator of NF-κB is hypoxia, a common marker of tumors (Rius et al., 2008). Activating NF-κB in inflammatory macrophages releases cytokines, interleukins, and TNF-α. It has been shown in breast cancer cells that secretion of TNF-α, activation of NF-κB, and presence of JNK leads to a more invasion and tumor progression. Inhibition of NF-κB also directly leads to lower levels of VEGF, impeding angiogenesis. IκBα is a prominent inhibitor of NF-κB and results in decreased tumor formation when it is overexpressed (Pikarsky et al., 2004). Interestingly, it has also been shown that low levels of NF-κB activation results in macrophages acquiring the M2 phenotype. A theory explaining this discrepancy is that NF-κB may have a different role depending on the stage and location of the TAMs within a tumor. This was shown in epithelial cells when IKKβ was knocked out and inflammation decreased, while in myeloid cells, inflammation was increased (Fong et al., 2008; Lawrence et al., 2001). In contrast, some reports indicate that targeted NF-κB activation can induce apoptosis of hepatocytes while decreasing tumor size, suggesting multiple roles of NF-κB (Greten et al., 2004; Hagemann et al., 2008; Luedde et al., 2007; Maeda et al., 2005; Naugler et al., 2007; Swain and Arezzo, 2008).
There have been many studies examining the effect of hypoxia on TAMs within the tumor microenvironment (Fig. 4). Low levels of tissue oxygenation induce TAM differentiation of macrophages (Erler et al., 2009; Shen et al., 2012). HIF activation has been determined to promote tumorigenesis due to higher levels of angiogenesis, cell proliferation, and TAM differentiation. There are 3 types of HIFs: HIF-1α, -2α, and -3α. HIF-1α is an important part of TAM response to hypoxia due to its ubiquitous expression and its induction of many hypoxia-inducible genes (Rankin and Giaccia, 2008). VEGF and CXCL12 are dependent on HIF-1α (Knowles and Harris, 2007). HIF-1α also regulates CSF-1. NF-κB can be modulated to regulate HIF-1α (Murdoch et al., 2004). Furthermore, deletion of HIF-1α has shown reduced tumor growth (Shen et al., 2012). Hypoxia causes HIF-1α to upregulate genes such as VEGF, bFGF, IL-1β, IL-8, cyclooxygenase 2 (COX2), MMP-7, MMP-9, MMP-12, and ANG-1 which lead to macrophages assuming the M2 phenotype and formation of a vascular system and malignant transition (Murdoch and Lewis, 2005; Murdoch et al., 2008). Increased HIF-2α expression in macrophages leads to angiogenesis and worse survival (Leek et al., 2002). Accordingly, TAMs with the highest levels of HIF-2α indicate high-grade breast cancers (Leek et al., 2002). Without HIF-1α, the suppressive capabilities of TAMs were reduced while without HIF-2α, there was less TAM recruitment in inflammatory hepatocellular and colon carcinoma models (Imtiyaz et al., 2010).
TAMs in breast tumor microenvironments typically express four surface markers; receptors for Fc portion of IgG and C3, HLA-DR antigen, and a macrophage-associated antigen (Steele et al., 1985). CD68, the human homolog of macrosialin, has been widely used as a pan-macrophage marker. There are many antibodies that recognize CD68 including Ki-M6, Ki-M7, Y2/131 and Y1/82A, EBM11, KP1, Ki-M1P, and PG-M1 (Gottfried et al., 2008). Increased CD68 expression has been shown to correlate with high vascularity and nodal metastasis and reduced recurrence-free and survival in human breast cancer (Jubb et al., 2010; Leek et al., 1996). Mahmoud et al., showed that increased CD68 markers predicted worse breast cancer-specific survival and shorter disease-free interval (Mahmoud et al., 2011). Other markers such as estrogen receptors (ER), progesterone receptor (PR), and human epidermal growth factor 2 (HER2) give a positive prognosis and were specifically designed to aid in therapy. However, it is not known how TAMs are associated with ER/PR/HER2 and thus more research must be conducted on the relationship between them before suggesting it as a possible treatment option. TAMs have also been shown to interfere with chemotherapy for breast cancer by releasing chemo-resistance factors or regulating the drug-resistance of cancer stem cells.
Since CD68 staining stains both M2 and M1 macrophages, CD68 cannot be used by itself to predict overall survival (Jubb et al., 2010; Mahmoud et al., 2011; Tsutsui et al., 2005). Thus other TAM-associated markers such as CD163, VEGF, HIFs markers, proliferating cellular nuclear antigen (PCNA), ferritin light chain (FTL), and CCL-18 have been proposed as possible options to use for detection of TAMs along with CD68. CD163 has been shown to be primarily expressed by M2 macrophages and thus has been associated with worse characteristics (Medrek et al., 2012). VEGF and EGFR are involved in macrophage infiltration in human breast cancer (Leek et al., 2000). VEGF, as noted previously, recruits more macrophages to the tumor microenvironment. TAMs with PCNA have been associated with high-grade, HR-negative breast cancers with an increased risk of recurrence and decreased overall survival in human breast cancers (Campbell et al., 2010; Mukhtar et al., 2011a). FTL stored in M2 macrophages indicates an aggressive phenotype in node-negative breast cancer patients. CCL18 is abundantly produced by breast cancer TAMs and is associated with metastasis and reduced patient survival. The counts predicted disease stage, histological grade, and lymph node and distant metastasis (Chen et al., 2011). Recent studies have indicated that CCL18 is a new and promising biomarker for cancer diagnosis and treatment.
Tumor suppressors act in a variety of ways including inducing cell-cycle arrest, senescence, or apoptosis. Retinoblastoma protein (Rb) plays a role in cell cycle control, differentiation, and inhibition of oncogenic transformation in addition to activation of NF-κB pathway. Promyelocytic leukemia protein (PML) plays a role in tumorigenesis, DNA damage, senescence, apoptosis, and protein degradation. P53 plays a role in ribosome biogenesis, aging control, cell cycle arrest, and apoptosis. P53 also regulates MCP-1/CCL2. ARF, interestingly, is one of the most frequently mutated genes in human cancer (Sharpless, 2005). ARF activates p53 and inhibits ribosomal RNA processing and transcriptional factors that induce proliferation. ARF activates p53 by displacing Mdm2 in the nucleolus (Pomerantz et al., 1998; Stott et al., 1998), and inhibits ribosomal RNA processing and transcriptional factors that induce proliferation (Kuo et al., 2003; Martelli et al., 2001; Qi et al., 2004; Sugimoto et al., 2003). ARF has a protective effect against viral infection by interacting with cell proliferation nucleophosmin (NPM) and activating dsRNA-dependent protein kinase (PKR) (Garcia et al., 2006). ARF may even modulate M2 polarization of macrophages. ARF-deficient macrophages show less leukocyte recruitment and decreased ability to produce pro-inflammatory properties. E2F1 is increased in ARF-deficient mice which have been shown to be anti-inflammatory and immunosuppressive which impaired the signaling of the NF-κB pathway. On the other hand, it has been shown to be deleted in some cancers thus more research must be conducted to further elucidate ARF’s function and effectiveness as a target for possible gene therapy.
Other methods have been investigated in order to interrupt or decrease macrophage recruitment to tumor microenvironments. Leuschner et al. used lipid nanoparticles containing siRNA to interrupt the CCL2-CCR2 interaction in order to disrupt macrophage recruitment. Using this method, they showed that the TAM infiltration was decreased (Leavy, 2011; Leuschner et al., 2011; Shantsila et al., 2011). Deletion of the CSFR1 gene reduced macrophage recruitment as well but only in resident tissue macrophages under normal conditions (MacDonald et al., 2010). Denardo et al. showed that using the PlexxiKon inhibitor, which inhibits CSFR1, along with cKit and PDGFR improved overall survival in mouse mammary carcinoma models (Denardo et al., 2011; Hume and MacDonald, 2011). Dijkgraaf et al. showed that blocking NF-κB signaling stopped platinum chemotherapy’s effect of increased IL-6 and PGE2 causing more M2 differentiation. Thus anti-IL-6 therapy is another recommended route for therapies (Dijkgraaf et al., 2013). Other approaches include inhibition of GM-CSF
One way to directly target macrophage phenotype differentiation in tumor microenvironments is by targeting the NF-κB pathway. This was done by inhibiting IKKβ, the main activator of NF-κB in murine models. This led to reduced tumor volume (Hayakawa et al., 2009). Another method to reduce tumor survival is to reprogram the TME to impair tumor growth only. This can be accomplished by using antibodies such as ipilimumab which targets CTLA-4 activating T cells, nivolumab, lambrolizumab, and CD40 targeted antibodies. Nivolumab targets the PD1 receptor causing apoptosis, while lambrolizumab is a blocking antibody for the PD1 ligand (Hamid et al., 2013; Hodi et al., 2010; Hwu, 2010; Restifo et al., 2012; Sharma et al., 2011; Vonderheide and Glennie, 2013; Wolchok et al., 2013). Combining these strategies may reduce tumor cells’ evasion of the immune response and help destroy them. All of these methods have been somewhat effective and resulted in greater overall survival.
Therapeutic treatments for hepatocellular carcinoma (HCC) that target TAMs have focused on inhibition of recruitment and/or polarization of M2 macrophages, inhibition of angiogenic and tissue remodeling, and inhibition of TAMs’ immunosuppressive effects. Clodronate-encapsulated liposomes or aminobisphosphonates knock down macrophages
Bisphosphonates are stable inorganic analogs of pyrophosphonate in which the central oxygen atom is replaced with a carbon atom. They are common in the treatment of many bone diseases such as cancer-induced bone disease, Paget’s disease, and osteoporosis. There are two types of bisphosphonates: nitrogen-containing (N-BP) and non-nitrogen containing (non N-BP) which are characterized by the presence of a nitrogen atom on the R2 side chain (Rodan and Fleisch, 1996; Stresing et al., 2007; Winter and Coleman, 2009). Non N-BPs inhibit the conversion of ATP to ADP which causes mitochondrial dysfunction leading to apoptosis. N-BPs alter protein prenylation by inhibiting farnesyldiphosphonate (FPP) synthase which is important for the mevalonate pathway in eukaryotic cells (Green, 2002; Luckman et al., 1998; Monkkonen et al., 2006; Moreau et al., 2007; Stern, 2007).
To target TAMs in breast cancer, N-BPs are utilized. They interfere with the process of PTHrP, prostaglandin-E, and interleukins stimulating RANKL causing osteoclast progression which in turn causes bone resorption leading to release of TGF-β helping tumor cells proliferate and grow. Specifically, they interfere with the bone resorption step in this cycle (Mundy, 2002). N-BPs have also been shown to directly affect tumors by reducing tumor cell proliferation, adhesion, migration, invasion, and angiogenesis while increasing apoptosis (Boissier et al., 1997; 2000; Fromigue et al., 2000; Hiraga et al., 2004; Senaratne et al., 2000; 2002; van der Pluijm et al., 1996). Biphosphonates can also directly target macrophages due to their similarities with osteoclasts. Using them has been shown to inhibit macrophage proliferation, migration, invasion, and cause apoptosis. They can also inhibit MMP-9 as shown
Much research has been completed elucidating the effects of TAMs in cancer. It has been shown that TAMs have a generally pro-tumorigenic effect. With this knowledge, it is possible to determine new therapies targeting TAMs and using them against the tumor to reduce volume and stunt tumor growth. A few proteins such as Rb, PML, and p53 act as tumor suppressors allowing for modulation of the proteins in an attempt to curb tumor growth. A viable and promising option is to modulate ARF, one of the most commonly mutated genes in cancer, to activate p53. This method has been shown to be promising, but more research must be completed. Directly targeting TAMs by reducing monocyte recruitment is another feasible option. This may be accomplished by interrupting the CCL2-CCR2 interaction. Another method to accomplish this is to downregulate GM-CSF, a crucial component used to recruit monocytes to the tumor. Similarly, it may be more feasible to control M2 differentiation by upregulating HRG or CD-40 to cause macrophages to assume the M1 phenotype, associated with antitumorigenic effects. The NF-κB pathway can also be targeted by inhibiting it with IKKβ reducing M2 differentiation. N-BPs have also been introduced as a novel idea to decrease tumor growth. N-BPs were shown to interfere in tumor cell proliferation, adhesion, migration, invasion, and angiogenesis while increasing apoptosis. In addition, N-BPs can inhibit MMP-9 and VEGF resulting in less M2 polarization and decreased angiogenesis.
Research on each of the aforementioned methods seems to indicate that they are all viable methods in the fight against cancer. More research must be completed to confirm their effects and usefulness in fighting tumors by attacking TAMs. It is clear, however, that progress is being made and novel methods to fighting cancer are on the horizon.
Mol. Cells 2015; 38(2): 95-104
Published online February 28, 2015 https://doi.org/10.14348/molcells.2015.2298
Copyright © The Korean Society for Molecular and Cellular Biology.
Samthosh V Alahari1, Shengli Dong, and Suresh K Alahari*
Department of Biochemistry and Molecular Biology, LSU School of Medicine, New Orleans, LA 70112, USA, 1McAllister Heart Institute, University of North Carolina School of Medicine, Chapel Hill, NC 27599, USA
Correspondence to:*Correspondence: salaha@lsuhsc.edu
The development of cancer has been an extensively researched topic over the past few decades. Although great strides have been made in cancer prevention, diagnosis, and treatment, there is still much to be learned about cancer’s micro-environmental mechanisms that contribute to cancer formation and aggressiveness. Macrophages, lymphocytes which originate from monocytes, are involved in the inflammatory response and often dispersed to areas of infection to fight harmful antigens and mutated cells in tissues. Macrophages have a plethora of roles including tissue development and repair, immune system functions, and inflammation. We discuss various pathways by which macrophages get activated, various approaches that can regulate the function of macrophages, and how these approaches can be helpful in developing new cancer therapies.
Keywords: angiogenesis, cancer, invasion, macrophages, microenvironment, migration
Macrophages are important cells of the immune system. They are derived from monocytes which enter various tissues and differentiate into macrophages. In the immune response, they use phagocytosis, a process in which the macrophages engulf pathogens, to protect the human body from disease. In addition to macrophages’ importance to the general functions of the immune system, they also play various roles in cancer. These macrophages, or tumor-associated macrophages (TAMs), interact with cancer cells to cause phagocytosis and lysis of tumor cells. On the other hand, TAMs have also been discovered to be an integral part of the tumorigenic pathway. The balance of cytokines and other molecular signals determines the role of these TAMs and thus it is important to study them closely. However, it is important to note that approximately 80% of studies show an association between TAMs and pro-tumorigenic effects while only 10% of studies show the opposite effect.
Before delving into TAMs’ function in cancer, it is important to understand the general cancer process. Cancer can be broken into three distinct stages. In the first stage, cancerous cells invade an area’s surrounding tissues and blood vessels; this area is referred to as the primary site. In the second stage, the cancer cells begin to move to secondary sites through the circulatory system. In the third and final stage, the cancer cells invade the secondary site’s surrounding tissues and blood vessels. This cycle can then repeat yielding tertiary sites and so forth. Once the primary site is established, tumor development is spurred by angiogenesis. Hanahan and Weinberg (2000), created a model that lays out the six properties that a tumor acquires while growing. These properties include an ability to replicate endlessly, angiogenesis, evasion of apoptosis, creating its own growth signals, insensitivity to growth inhibitors, and tissue invasion and metastasis. A seventh property, tumor microenvironment inflammation, was later added to the model (Hanahan and Weinberg, 2000; Mantovani, 2009) (Fig. 1). Here we discuss the mechanisms by which macrophages get activated and their important role in tumor progression.
Two different types of phenotypes for macrophages have been described; these include: M1 and M2. The M1 phenotype is associated with active microbial killing while the M2 phenotype is associated with angiogenesis and wound repair. TAMs can typically be identified by low expression of tumor necrosis factors (TNF) and high expression of IL-1 and IL-6. Specifically M1 macrophage polarization is caused by interferon gamma (IFN-gamma) and TNF-α secretion by T helper 1 (Th1) lymphocytes or natural killer (NK) cells (Gordon, 2003; Martinez et al., 2009; Murray and Wynn, 2011). Interferon (IFNs) or toll-like receptor (TLR) signals can produce the M1 phenotype through the STAT1 signaling pathway. M1 macrophages also produce chemical attractants, such as pro-inflammatory cytokines, interleukin-12 (IL-12), IL-1β, and IL-6 (Heusinkveld and van der Burg, 2011; Porta et al., 2011). M2 macrophages, on the other hand, are activated by Th2. Th2 releases several cytokines, such as IL-4, IL-10, and IL-13, to activate the M2 macrophages (Gordon, 2003; Murray and Wynn, 2011). The M2 macrophage phenotype is caused by IL-4 and IL-13 via the STAT6 signaling pathways. In addition, Kruppel-like factor 4 (KFL4) works in conjunction with STAT6 to produce M2 genes, while inhibiting M1 genes through isolation of some NF-κB activators. Unlike the M1 phenotype, the M2 phenotype can be divided into 3 more categories: M2a, M2b, and M2c. M2a is differentiated by IL-4 and IL-13 while M2b by immune complexes. M2c is differentiated by IL-10 and also plays a role in immunosuppression and tissue remodeling (Hagemann et al., 2009). Furthermore, it has been shown that Th1 cytokines produce M1 macrophages while M2 macrophages play a role in Th1 adaptive immunity suppression, resolution of inflammation, parasite protection, wound healing promotion, angiogenesis, and tissue remodeling (Biswas and Mantovani, 2010).
Some of the most potent macrophage activators of metastasis involved in Lewis lung carcinoma lead to the production of IL-6 and TNF-α through the activation of essential-to-metastasis TLR2 and TLR6 (Chiodoni et al., 2010; Kim et al., 2009). These macrophages are recruited to tumor cells by the production of monocyte chemotactic protein (MCP), macrophage-colony stimulating factor (M-CSF), vascular endothelial growth factor (VEGF), angiopoietin 1 (ANG1), and ANG2. In addition to macrophage recruitment, VEGF, ANG1, and ANG2 play roles in promoting angiogenesis (Pollard, 2004). Anti-inflammatory molecules, such as IL-4, IL-10, transforming growth factor beta 1 (TGF-β1), and prostaglandin E2 (PGE2) produced by tumor cells cause macrophages to attain the M2 phenotype. Adding IL-10
Looking further into VEGF, it has been determined that it is an angiogenic factor that triggers angiogenesis and lymphangiogenesis. Angiogenic and lymphangiogenic factors include VEGF-A, VEGF-C, VEGF-D, and MMP that all bind to their respective coordinating receptors and play roles in enhancing endothelial cell proliferation, migration, and tube-like formation (Giraudo et al., 2004; Wu et al., 2012). MMP has also been shown to mediate VEGF-A levels in some studies. Specifically, MMPs are proteolytic enzymes tasked with the degradation of proteins within the extracellular matrix. They regulate many cell behaviors within the tumor microenvironment (Egeblad and Werb, 2002).
The importance of lymphangiogenesis must not be understated because it plays a large role in tumor stability and growth even though it occurs second to angiogenesis in a tumor. Lymphangiogenesis exposes the tumor to immune cells leading to lymphatic metastasis and without it, the tumor would likely stop growing and be destroyed (Kurahara et al., 2012). Lymphangiogenesis is typically induced by VEGF-A and VEGF C/D binding with VEGFR-2 and VEGFR-3 receptors. In addition, tissue inflammation in tumors has been shown to expand lymph nodes and lead to increased lymphangiogenesis and lymphoid hyperplasia. Inhibition of M-CSF has been shown to suppress tumor lymphangiogenesis since it plays a role in VEGF-A/C secretion. Interestingly, endothelial nitric oxide synthase (eNOS) and IL-6 also regulate lymphangiogenesis (Ji, 2011) (Fig. 3).
Research has shown that there is a positive correlation between the number of TAMs and advanced tumor progression and metastasis. Many tumors show overexpression of M-CSF and when down-regulated, there is a decrease in macrophage activity. Knocking out MCP has led to lower levels of macrophages and decreased tumor progression as well. These data demonstrate a clear relationship between TAMs and tumorigenesis. Surprisingly, further overexpression of MCP has shown that at a certain point, M1 macrophages are attracted to the site causing cytotoxic effects on the tumor cells (Shirabe et al., 2011).
In addition to the decrease in macrophage activity due to the downregulation of M-CSF, vascularization is also reduced. Without vascularization, larger tumors struggle to grow due to insufficient amounts of oxygen causing hypoxic regions. TAMs compensate for this effect by congregating in these areas causing upregulation of certain transcription factors that lead to the expression of various growth factors, cytokines, and other signaling molecules resulting in angiogenesis (Obeid et al., 2013). These transcription factors cause secretion of VEGF and basic fibroblast growth factor (bFGF) which promote the proliferation and maturation of endothelial cells (Lamagna et al., 2006; Siveen and Kuttan, 2009). TAM created proteases, such as MMPs, plasmin, and urokinase plasminogen activator, degrade and remodel the extracellular matrix leading to further angiogenesis (Siveen and Kuttan, 2009).
The majority of research indicates increased levels of TAM activity result in a negative prognosis. The M1 phenotype is usually present during early stages of tumor development while the M2 phenotype is associated with more advanced tumors. In addition, compounds such as VEGF and M-CSF allow for tumor development to progress by recruiting macrophages to the tumor site and initiating angiogenesis and lymphangiogenesis.
The tumor microenvironment is a niche that has been shown to affect tumor activity and outcome. It is created the moment the tumor becomes attached to a site and encompasses the extracellular matrix surrounding the tumor cell and any non-cancerous cells that happen to reside in the targeted organ (Fig. 2).
Colony stimulating factor 1 (CSF1) and chemokine ligand 2 (CCL2) in the blood recruit monocytes that become TAMs in the tumor microenvironment, a process called intraepithelial neoplasia. In addition to CSF-1, the cytokines VEGF and PDGF and chemokines CCL2, CCL5, and CCL8 all attract monocytes to the tumor microenvironment (Allavena et al., 2008; Goede et al., 1999; Mantovani et al., 2008; Pollard, 2004) (Fig. 3). It may be important to note that CCL2 upregulation is caused by a
It has been reported that some transcription factors also regulate TAM functions. Nuclear factor-κB (NF-κB) is activated by TNF-α, IL-1, and IKKβ. NF-κB refers to a family of transcription factors that plays a role in inflammation and immunity. NF-κB operates by causing the expression of pro-inflammatory cytokines, chemokines, and angiogenic factors (Porta et al., 2011). Thus NF-κB is an important part of the regulation of TAM function. Transcription factors for NF-κB include Re1A (p65), c-Rel, RE1B, p50, and p52. Out of those transcription factors, only Re1A, Re1B, and c-Rel can activate gene transcription (Hagemann et al., 2009). NF-κB is activated primarily by IKKβ (Maeda et al., 2005). Another activator of NF-κB is hypoxia, a common marker of tumors (Rius et al., 2008). Activating NF-κB in inflammatory macrophages releases cytokines, interleukins, and TNF-α. It has been shown in breast cancer cells that secretion of TNF-α, activation of NF-κB, and presence of JNK leads to a more invasion and tumor progression. Inhibition of NF-κB also directly leads to lower levels of VEGF, impeding angiogenesis. IκBα is a prominent inhibitor of NF-κB and results in decreased tumor formation when it is overexpressed (Pikarsky et al., 2004). Interestingly, it has also been shown that low levels of NF-κB activation results in macrophages acquiring the M2 phenotype. A theory explaining this discrepancy is that NF-κB may have a different role depending on the stage and location of the TAMs within a tumor. This was shown in epithelial cells when IKKβ was knocked out and inflammation decreased, while in myeloid cells, inflammation was increased (Fong et al., 2008; Lawrence et al., 2001). In contrast, some reports indicate that targeted NF-κB activation can induce apoptosis of hepatocytes while decreasing tumor size, suggesting multiple roles of NF-κB (Greten et al., 2004; Hagemann et al., 2008; Luedde et al., 2007; Maeda et al., 2005; Naugler et al., 2007; Swain and Arezzo, 2008).
There have been many studies examining the effect of hypoxia on TAMs within the tumor microenvironment (Fig. 4). Low levels of tissue oxygenation induce TAM differentiation of macrophages (Erler et al., 2009; Shen et al., 2012). HIF activation has been determined to promote tumorigenesis due to higher levels of angiogenesis, cell proliferation, and TAM differentiation. There are 3 types of HIFs: HIF-1α, -2α, and -3α. HIF-1α is an important part of TAM response to hypoxia due to its ubiquitous expression and its induction of many hypoxia-inducible genes (Rankin and Giaccia, 2008). VEGF and CXCL12 are dependent on HIF-1α (Knowles and Harris, 2007). HIF-1α also regulates CSF-1. NF-κB can be modulated to regulate HIF-1α (Murdoch et al., 2004). Furthermore, deletion of HIF-1α has shown reduced tumor growth (Shen et al., 2012). Hypoxia causes HIF-1α to upregulate genes such as VEGF, bFGF, IL-1β, IL-8, cyclooxygenase 2 (COX2), MMP-7, MMP-9, MMP-12, and ANG-1 which lead to macrophages assuming the M2 phenotype and formation of a vascular system and malignant transition (Murdoch and Lewis, 2005; Murdoch et al., 2008). Increased HIF-2α expression in macrophages leads to angiogenesis and worse survival (Leek et al., 2002). Accordingly, TAMs with the highest levels of HIF-2α indicate high-grade breast cancers (Leek et al., 2002). Without HIF-1α, the suppressive capabilities of TAMs were reduced while without HIF-2α, there was less TAM recruitment in inflammatory hepatocellular and colon carcinoma models (Imtiyaz et al., 2010).
TAMs in breast tumor microenvironments typically express four surface markers; receptors for Fc portion of IgG and C3, HLA-DR antigen, and a macrophage-associated antigen (Steele et al., 1985). CD68, the human homolog of macrosialin, has been widely used as a pan-macrophage marker. There are many antibodies that recognize CD68 including Ki-M6, Ki-M7, Y2/131 and Y1/82A, EBM11, KP1, Ki-M1P, and PG-M1 (Gottfried et al., 2008). Increased CD68 expression has been shown to correlate with high vascularity and nodal metastasis and reduced recurrence-free and survival in human breast cancer (Jubb et al., 2010; Leek et al., 1996). Mahmoud et al., showed that increased CD68 markers predicted worse breast cancer-specific survival and shorter disease-free interval (Mahmoud et al., 2011). Other markers such as estrogen receptors (ER), progesterone receptor (PR), and human epidermal growth factor 2 (HER2) give a positive prognosis and were specifically designed to aid in therapy. However, it is not known how TAMs are associated with ER/PR/HER2 and thus more research must be conducted on the relationship between them before suggesting it as a possible treatment option. TAMs have also been shown to interfere with chemotherapy for breast cancer by releasing chemo-resistance factors or regulating the drug-resistance of cancer stem cells.
Since CD68 staining stains both M2 and M1 macrophages, CD68 cannot be used by itself to predict overall survival (Jubb et al., 2010; Mahmoud et al., 2011; Tsutsui et al., 2005). Thus other TAM-associated markers such as CD163, VEGF, HIFs markers, proliferating cellular nuclear antigen (PCNA), ferritin light chain (FTL), and CCL-18 have been proposed as possible options to use for detection of TAMs along with CD68. CD163 has been shown to be primarily expressed by M2 macrophages and thus has been associated with worse characteristics (Medrek et al., 2012). VEGF and EGFR are involved in macrophage infiltration in human breast cancer (Leek et al., 2000). VEGF, as noted previously, recruits more macrophages to the tumor microenvironment. TAMs with PCNA have been associated with high-grade, HR-negative breast cancers with an increased risk of recurrence and decreased overall survival in human breast cancers (Campbell et al., 2010; Mukhtar et al., 2011a). FTL stored in M2 macrophages indicates an aggressive phenotype in node-negative breast cancer patients. CCL18 is abundantly produced by breast cancer TAMs and is associated with metastasis and reduced patient survival. The counts predicted disease stage, histological grade, and lymph node and distant metastasis (Chen et al., 2011). Recent studies have indicated that CCL18 is a new and promising biomarker for cancer diagnosis and treatment.
Tumor suppressors act in a variety of ways including inducing cell-cycle arrest, senescence, or apoptosis. Retinoblastoma protein (Rb) plays a role in cell cycle control, differentiation, and inhibition of oncogenic transformation in addition to activation of NF-κB pathway. Promyelocytic leukemia protein (PML) plays a role in tumorigenesis, DNA damage, senescence, apoptosis, and protein degradation. P53 plays a role in ribosome biogenesis, aging control, cell cycle arrest, and apoptosis. P53 also regulates MCP-1/CCL2. ARF, interestingly, is one of the most frequently mutated genes in human cancer (Sharpless, 2005). ARF activates p53 and inhibits ribosomal RNA processing and transcriptional factors that induce proliferation. ARF activates p53 by displacing Mdm2 in the nucleolus (Pomerantz et al., 1998; Stott et al., 1998), and inhibits ribosomal RNA processing and transcriptional factors that induce proliferation (Kuo et al., 2003; Martelli et al., 2001; Qi et al., 2004; Sugimoto et al., 2003). ARF has a protective effect against viral infection by interacting with cell proliferation nucleophosmin (NPM) and activating dsRNA-dependent protein kinase (PKR) (Garcia et al., 2006). ARF may even modulate M2 polarization of macrophages. ARF-deficient macrophages show less leukocyte recruitment and decreased ability to produce pro-inflammatory properties. E2F1 is increased in ARF-deficient mice which have been shown to be anti-inflammatory and immunosuppressive which impaired the signaling of the NF-κB pathway. On the other hand, it has been shown to be deleted in some cancers thus more research must be conducted to further elucidate ARF’s function and effectiveness as a target for possible gene therapy.
Other methods have been investigated in order to interrupt or decrease macrophage recruitment to tumor microenvironments. Leuschner et al. used lipid nanoparticles containing siRNA to interrupt the CCL2-CCR2 interaction in order to disrupt macrophage recruitment. Using this method, they showed that the TAM infiltration was decreased (Leavy, 2011; Leuschner et al., 2011; Shantsila et al., 2011). Deletion of the CSFR1 gene reduced macrophage recruitment as well but only in resident tissue macrophages under normal conditions (MacDonald et al., 2010). Denardo et al. showed that using the PlexxiKon inhibitor, which inhibits CSFR1, along with cKit and PDGFR improved overall survival in mouse mammary carcinoma models (Denardo et al., 2011; Hume and MacDonald, 2011). Dijkgraaf et al. showed that blocking NF-κB signaling stopped platinum chemotherapy’s effect of increased IL-6 and PGE2 causing more M2 differentiation. Thus anti-IL-6 therapy is another recommended route for therapies (Dijkgraaf et al., 2013). Other approaches include inhibition of GM-CSF
One way to directly target macrophage phenotype differentiation in tumor microenvironments is by targeting the NF-κB pathway. This was done by inhibiting IKKβ, the main activator of NF-κB in murine models. This led to reduced tumor volume (Hayakawa et al., 2009). Another method to reduce tumor survival is to reprogram the TME to impair tumor growth only. This can be accomplished by using antibodies such as ipilimumab which targets CTLA-4 activating T cells, nivolumab, lambrolizumab, and CD40 targeted antibodies. Nivolumab targets the PD1 receptor causing apoptosis, while lambrolizumab is a blocking antibody for the PD1 ligand (Hamid et al., 2013; Hodi et al., 2010; Hwu, 2010; Restifo et al., 2012; Sharma et al., 2011; Vonderheide and Glennie, 2013; Wolchok et al., 2013). Combining these strategies may reduce tumor cells’ evasion of the immune response and help destroy them. All of these methods have been somewhat effective and resulted in greater overall survival.
Therapeutic treatments for hepatocellular carcinoma (HCC) that target TAMs have focused on inhibition of recruitment and/or polarization of M2 macrophages, inhibition of angiogenic and tissue remodeling, and inhibition of TAMs’ immunosuppressive effects. Clodronate-encapsulated liposomes or aminobisphosphonates knock down macrophages
Bisphosphonates are stable inorganic analogs of pyrophosphonate in which the central oxygen atom is replaced with a carbon atom. They are common in the treatment of many bone diseases such as cancer-induced bone disease, Paget’s disease, and osteoporosis. There are two types of bisphosphonates: nitrogen-containing (N-BP) and non-nitrogen containing (non N-BP) which are characterized by the presence of a nitrogen atom on the R2 side chain (Rodan and Fleisch, 1996; Stresing et al., 2007; Winter and Coleman, 2009). Non N-BPs inhibit the conversion of ATP to ADP which causes mitochondrial dysfunction leading to apoptosis. N-BPs alter protein prenylation by inhibiting farnesyldiphosphonate (FPP) synthase which is important for the mevalonate pathway in eukaryotic cells (Green, 2002; Luckman et al., 1998; Monkkonen et al., 2006; Moreau et al., 2007; Stern, 2007).
To target TAMs in breast cancer, N-BPs are utilized. They interfere with the process of PTHrP, prostaglandin-E, and interleukins stimulating RANKL causing osteoclast progression which in turn causes bone resorption leading to release of TGF-β helping tumor cells proliferate and grow. Specifically, they interfere with the bone resorption step in this cycle (Mundy, 2002). N-BPs have also been shown to directly affect tumors by reducing tumor cell proliferation, adhesion, migration, invasion, and angiogenesis while increasing apoptosis (Boissier et al., 1997; 2000; Fromigue et al., 2000; Hiraga et al., 2004; Senaratne et al., 2000; 2002; van der Pluijm et al., 1996). Biphosphonates can also directly target macrophages due to their similarities with osteoclasts. Using them has been shown to inhibit macrophage proliferation, migration, invasion, and cause apoptosis. They can also inhibit MMP-9 as shown
Much research has been completed elucidating the effects of TAMs in cancer. It has been shown that TAMs have a generally pro-tumorigenic effect. With this knowledge, it is possible to determine new therapies targeting TAMs and using them against the tumor to reduce volume and stunt tumor growth. A few proteins such as Rb, PML, and p53 act as tumor suppressors allowing for modulation of the proteins in an attempt to curb tumor growth. A viable and promising option is to modulate ARF, one of the most commonly mutated genes in cancer, to activate p53. This method has been shown to be promising, but more research must be completed. Directly targeting TAMs by reducing monocyte recruitment is another feasible option. This may be accomplished by interrupting the CCL2-CCR2 interaction. Another method to accomplish this is to downregulate GM-CSF, a crucial component used to recruit monocytes to the tumor. Similarly, it may be more feasible to control M2 differentiation by upregulating HRG or CD-40 to cause macrophages to assume the M1 phenotype, associated with antitumorigenic effects. The NF-κB pathway can also be targeted by inhibiting it with IKKβ reducing M2 differentiation. N-BPs have also been introduced as a novel idea to decrease tumor growth. N-BPs were shown to interfere in tumor cell proliferation, adhesion, migration, invasion, and angiogenesis while increasing apoptosis. In addition, N-BPs can inhibit MMP-9 and VEGF resulting in less M2 polarization and decreased angiogenesis.
Research on each of the aforementioned methods seems to indicate that they are all viable methods in the fight against cancer. More research must be completed to confirm their effects and usefulness in fighting tumors by attacking TAMs. It is clear, however, that progress is being made and novel methods to fighting cancer are on the horizon.
Doo Jae Lee and Sang Won Kang
Mol. Cells 2013; 35(2): 93-98 https://doi.org/10.1007/s10059-013-0034-9Mingqi Qu, Ju Yu, Hongyuan Liu, Ying Ren, Chunxiao Ma, Xingyao Bu, and Qing Lan
Mol. Cells 2017; 40(10): 761-772 https://doi.org/10.14348/molcells.2017.0104Haejung Kim, Haein Hwang, Hansoo Lee, and Hyo Jeong Hong
Mol. Cells 2017; 40(5): 363-370 https://doi.org/10.14348/molcells.2017.2282