Both inflammation and angiogenesis are exacerbated by increased production of chemokines/cytokines, growth factors, proteolytic enzymes, proteoglycans, lipid mediators and prostaglandins. It has been reported that approximately 15–20% of all malignancies are initiated or exacerbated by inflammation. Initiation and progression of cancer are also closely linked to angiogenesis. Infiltration of macrophages is a dramatic and common feature of inflammation, angiogenesis and cancer, and has been recently highlighted in an attempt to develop novel strategies for treating cancer. The recruitment and infiltration of macrophages in the tumor microenvironment activates them to support the malignant progression of cancer cells, and these macrophages are called tumor-associated macrophages. In a model of experimental angiogenesis using mouse corneas, macrophages infiltrated tissue in response to inflammatory cytokines and produced chemokines and angiogenesis-promoting factors, such as vascular endothelial growth factor-A, interleukin-8, matrix metalloproteinases, prostanoids and reactive oxygen species. Moreover, in a cancer xenograft model, inflammatory stimuli by a representative inflammatory cytokine, interleukin-1β, enhanced tumor growth and angiogenesis with infiltration and activation of macrophages. Co-culture of cancer cells with macrophages synergistically stimulated production of various angiogenesis-related factors when stimulated by the inflammatory cytokine. This inflammatory angiogenesis in both mouse cornea and a tumor model was mediated, in part, by activation of nuclear factor κB and activator protein 1 (Jun/Fos). Administration of either nuclear factor κB-targeting drugs or cyclooxygenase 2 inhibitors or depletion of macrophages could block both inflammatory angiogenesis and tumor angiogenesis. Thus, both inflammatory and angiogenic responses in tumor stroma could be targets for development of anticancer therapeutic drugs. (Cancer Sci 2008; 99: 1501–1506)
Inflammation and angiogenesis often share common pathways
The process of inflammation involves a complex network of chemical signals and cell interactions in response to cell or tissue damage. This process is the result of a balance between pro-inflammatory and anti-inflammatory factors (Fig. 1). Of various inflammatory cell types, leukocytes (neutrophils and eosinophils) are first recruited upon stimulation of chemotactic cytokines. Mast cells are then recruited, thus stimulating a network of pro-inflammatory cytokines and chemokines such as the interleukin (IL) family, tumor necrosis factor (TNF)-α and interferons.(1–3) In response to such chemotactic signals, monocytes migrate into the inflamed area and differentiate into macrophages. Macrophages with other inflammatory cell types provide growth factors, cytokines, proteolytic enzymes, proteoglycan, lipid mediators and prostaglandins. All of these factors cause marked changes in the inflammatory loci by interacting with epithelial, mesenchymal and vascular endothelial cells.
Angiogenesis is also a complex multistep process of growth and remodeling involving degradation of the extracellular matrix (ECM), cell migration and proliferation, and tube formation.(2,4) Under normal conditions, this process requires a balance between pro-angiogenic factors and anti-angiogenic factors. Angiogenesis also requires the activation of many receptors by their each cognate ligands. These ligands include vascular endothelial growth factors (VEGF), placental growth factor (PIGF), fibroblast growth factors (FGF-1 and -2), platelet-derived growth factor (PDGF), hepatocyte growth factor (HGF), angiopoietins (Ang-1 and -2), epidermal growth factor/transforming growth factor-α (EGF/TGF-α) and others. VEGF are known to play the most important role in angiogenesis.
The above two processes, angiogenesis and inflammation, are closely linked in the following ways: (i) they are coupled with some chronic inflammatory diseases including Crohn disease, diabetes, psoriasis, rheumatoid arthritis, osteoarthritis, obesity, ocular diseases as well as cancer; (ii) inflammatory cells interact with endothelial cells, fibroblasts and ECM in the inflamed loci; and (iii) the same molecular events trigger both inflammation and angiogenesis(5) (Fig. 1).
Almost 150 years ago, Rudolf Virchow first indicated the concept that lymphoreticular infiltration reflects the origin of cancer at sites of chronic inflammation, suggesting a close connection between inflammation and the development of cancer. Nowadays, the features shared by cancer and inflammation have been highlighted by proposing novel therapeutic strategies targeting the inflammatory responses in tumor stroma. A recent study demonstrated that stromal interactions between malignant cells and inflammatory cells could be closely associated with angiogenesis and the progression of cancer (Fig. 2).(6) In this article, we will focus on the progression of cancer in association with both inflammation and angiogenesis.
Macrophages have at least two faces of friend and foe for cancer progression
Inflammatory stimuli include chemicals and foreign bodies (asbestos fiber, silica particles, catheters, alcohol, bile acids, gastric acids, gall bladder stones and ultraviolet light) and infectious organisms (Hellicobacter pylori, hepatitis B and C viruses, Epstein–Barr virus, herpes virus, papilloma virus, HIV). These agents often promote the progression of malignancies, indicating the close link between inflammation and cancer. In response to such inflammatory stimuli, monocytes and macrophages are thought to infiltrate the inflammatory sites and the tumor area. Monocytes are differentiated into two types of cells with different functions: tumor-suppressive and tumor-supportive macrophages. Tumor-supportive macrophages are active in matrix remodeling, tissue repair and angiogenesis, whereas tumor-suppressive macrophages have microbicidal activity, immunostimulatory functions and tumor cytotoxicity. During the accumulation of somatic mutations of oncogenes and oncosuppressor genes, various immunoreactive cell types, including tumor-suppressive macrophages, are thought to suppress the progression of cancer. However, when such macrophages fail to suppress the progression of tumors, other macrophages, possibly tumor-supportive or tumor-associated macrophage (TAM), may support tumor growth and suppress the immune response against cancer.(7–9) One may then ask how such macrophages acquire their functions as friends to cancer in the context of inflammation and angiogenesis.
Macrophages are educated to promote tumor growth, angiogenesis and metastasis through cross-talks with cancer cells and other stromal cells
Many resident cell types, such as fibroblasts and adipocytes, and migrating hematopoietic cell types, such as macrophages, neutrophils and mast cells, populate the tumor microenvironment.(1,10,11) Among these cells, macrophages play a pivotal role in the initiation and promotion of tumorigenesis and the progression of malignant cells.(8,9,12) TAM are thought to acquire their ability to enhance production of various factors of angiogenesis, cell growth, cell motility, ECM degradation as well as inflammation-related chemokines (Table 1). Clinical studies have demonstrated a close association between an abundance of TAM and poor prognosis in breast, prostate, colon and cervical cancer, however, this association was not found in prostate and lung cancer.(13–15) In contrast, an abundance of TAM is associated with a favorable prognosis in stomach cancer. We found a significant correlation between the number of infiltrating macrophages and the microvascular density or tumor progression levels in glioblastomas and melanoma.(16,17) Taken together, increased numbers of TAM are significantly associated with poor prognosis, suggesting their role in the malignant progression of most solid tumors.
Table 1. Wide variety of growth factors and chemokines, proteases and other factors that are produced from tumor-associated macrophages
The mutual interaction of many stromal cell types with cancer cells further enhances production of inflammatory cytokines, chemokines, proteases, prostanoids, growth factors and angiogenesis-related factors. These substances transform the tumor microenvironment so that it favors the survival, growth and motility of cancer cells (Fig. 2). In various tumor types, TAM, growth factors and chemokines affect the clinical outcome of malignant tumors (Table 1).(18,19) Among the factors produced by TAM, the most important cytokines in the recruitment of macrophages are monocyte chemoattractant protein-1 (MCP-1), which is also produced by fibroblasts, endothelial cells and human cancer cells.(15) MCP-1 is positively associated with TAM accumulation in tumors and, in a form of feedback inhibition, may limit the number of macrophages in tumors.(20) High levels of macrophage-colony stimulating growth factor/colony stimulating growth factor-1 (M-CSF/CSF-1) and its receptor are also closely associated with TAM accumulation in various tumors, including breast cancer.(21) Moreover, CSF-1 stimulated the differentiation of monocytes into macrophages.(18,19) As shown in Fig. 2, monocytes are recruited to tumor areas by VEGF, MCP-1, CSF-1 and other chemokines, and this interaction of monocytes/macrophages with cancer cells might promote further recruitment and activation of TAM. This population of TAM promote production of growth factors, proteases, angiogenic factors and reactive oxygen species, which support the progression of cancer. In particular, TAM that are recruited to the tumor have a key role in the angiogenic switch and malignant transition of cancer (Fig. 2).(22)
Inflammatory cytokine-induced angiogenesis is accompanied by macrophage infiltration and activation of both cyclooxygenase 2 and nuclear factor κB
In the tissue injury model in cornea of mice, we observed recruitment of neutrophils first and then that of monocytes/macrophages.(23) IL-1α and IL-1β specifically bind to type I receptors of IL-l and induce inflammatory stimuli. IL-1α/β signaling, in the context of angiogenesis, is associated with pathological conditions, such as rheumatoid arthritis, septic shock, graft-versus-host disease, arteriosclerosis, asthma, adult T-cell leukemia, multiple myeloma and other diseases. Moreover, in vitro treatment of both vascular endothelial cells and cancer cells with IL-1α/β, TNF-α and reactive oxygen species results in a marked induction of VEGF-A, FGF-2, IL-8 and plasminogen activators through transcriptional activation of nuclear factor (NF)-κB, specificity protein 1 (Sp-1), activator protein 1 (AP-1) (Jun/Fos) and hypoxia response elements.(24–26) IL-1α and IL-1β are also required for angiogenesis and tumor growth, and they promote invasion and metastasis of cancer cells in animal models.(27–30) In addition, the angiogenesis stimulated by these inflammatory cytokine expressing cancer cells in a xenograft model is augmented by upregulation of various angiogenesis-related factors and matrix metallopeptidase (MMP), but is abolished by IL-1α- and IL-1β-knockdowns.
Interleukin-1β can induce angiogenesis in mouse corneas on days 4 and 6 after implantation (Fig. 3). Cyclooxygenase 2 (COX2) is a representative mediator of inflammatory responses, and its expression is highly susceptible to inflammatory stimuli, and COX2 also plays a key role in tumor angiogenesis by catalyzing the production of prostanoids, including prostaglandin E2 (PGF2) and thromboxane A2 (TXA2).(31,32) In mouse corneal angiogenesis induced by IL-1β, either administration of COX2 inhibitors or COX2 knockdown almost completely attenuated angiogenesis (Fig. 3).(33) Two PGE2 receptor (EP2 and EP4) agonists and a TXA2 receptor agonist induce angiogenesis in vitro and also in mouse cornea, and IL-1β-induced angiogenesis is inhibited by a PGE2 receptor, EP4, antagonist and an TXA2 antagonist.(33) This finding suggests that prostanoids activated by COX2 are involved in inflammatory angiogenesis. COX expression is upregulated in monocytes/macrophages during their recruitment by IL-1β to neovascular areas in the cornea.(34)
The IL-1β-induced angiogenesis is almost completely blocked by knockdown of MCP-1 in mouse corneas, suggesting the pivotal role of MCP-1 in this process.(35) Nakao et al. further demonstrated a marked inhibition of IL-1β-induced angiogenesis by dexamethasone, a synthetic corticosteroid. This inhibition of the inflammatory angiogenesis is also attenuated by an inhibitor of NF-κB signaling. This finding suggests that NF-κB activation in the corneal stromal cells is important for angiogenesis through the upregulation of VEGF-A and PGE2. Thus, IL-1β-induced angiogenesis in corneas is attributable to upregulation of various angiogenesis-related factors in monocytes/macrophages, other corneal stromal cells and endothelial cells. This process is mediated through activation of inflammatory signaling, COX2, chemokines, and NF-κB and other relevant transcriptional regulators.
Tumor angiogenesis and TAM under inflammatory stimuli as a target for anticancer therapy
Infiltration and activation of inflammatory cells, including TAM, confer favorable conditions for the progression of cancer cells. One can thus expect development of novel anticancer therapies that target stromal tumors cells such as TAM.(11,12) Van-Rooijen developed a novel macrophages-targeting drug, clodronate, one of the first generation of bisphosphonates, encapsulated by a liposome, C12MDP-LIP.(36) The C12MDP-LIP ingested by macrophages is then destroyed, following phospholipase-mediated disruption of the liposomal bilayers, and release of clodronate. The number of macrophages in peripheral blood and corneas can be markedly decreased by this clodronate/liposome-mediated macrophage suicide. This macrophage depletion by clodronate/liposome in turn reduces both tumor neovascularization and tumor growth in lung cancers that express IL-1β (Fig. 4). Co-cultivation of cancer cells and macrophages with inflammatory stimuli synergistically enhances migration of vascular endothelial cells in vitro, angiogenesis in vivo, and the expression of MCP-1, VEGF and IL-8.(37) This IL-1β-dependent stimulatory effect on the expression of angiogenesis-related factors and MMP-9 can be blocked by inhibitors of NF-κB and AP-1,(37) and by COX2 inhibitors(33) as well. Consistent with these findings, administration of clodronate/liposomes resulted in marked decrease of bone metastasis by lung cancer cells with concomitant depletion of both macrophages in peripheral blood and osteoclasts in foci of bone metastases.(38) Thus, therapies that target macrophages with clodronate/liposome could be useful in preventing tumor angiogenesis, growth and metastasis. Therapeutic agents that target inflammatory signals, such as NF-κB and COX2, may also be useful in suppressing tumor angiogenesis.
Angiogenesis and inflammation often share common pathways, and these two biological processes are also closely coupled with cancer. Of various inflammatory cell types infiltrating the tumor area in response to inflammatory stimuli, tumor-supporting macrophages, TAM, are thought to play key roles in further production of various growth factors, angiogenic factors, proteinases, chemokines and cytokines, through cross-talks with cancer cells and other tumor stromal cells (Fig. 5). These factors in turn stimulate cell migration/motility, proliferation, survival, angiogenesis and metastasis, resulting in a dynamic environment that favors the progression of cancer. Administration of a macrophage-targeting drug as well as inhibitors of inflammation mediators, such as COX2 and NF-κB, can markedly block both angiogenesis and tumor growth in experimental therapeutic models (Fig. 5). To develop any anticancer drug that can specifically target tumor-supporting macrophages and tumor-specific inflammatory responses, further research will be required to understand the molecular mechanisms of how tumor stroma responds to inflammatory stimuli in human cancer, and what specific characteristics are acquired in tumor-supporting TAM. A more complete understanding of these issues could lead to the development of new anticancer drugs targeting both angiogenesis and inflammation.
I thank Dr M. Kuwano (Kyushu University) for fruitful discussions during preparation of this article, and also thank our colleagues, K. Watari, S. Nakao, T. Kuwano, F. Hosoi, Y. Basaki, Y. Nishiyama (Kyushu University), N. Kimura, A. Fotovati and K. Hiraoka (Kurume University) for their collaborative study. This study was supported by a Grant-in-Aid for Scientific Research on Priority Areas, Cancer, from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and also by the Formation of Special Coordination Funds for Promoting Science and Technology (SCF), Japan.