Increased expression of RAGE has been documented in a variety of acute and chronic inflammatory diseases, such as septicaemia, rheumatoid arthritis, osteoarthritis, arteriosclerosis, chronic renal disease, inflammatory bowel disease, vasculitis and late diabetic complications (see references in 18). In the past, numerous findings have advanced our knowledge on the molecular mechanisms through which RAGE could contribute to cellular processes of adaptive and innate immune responses 1415. First, RAGE expression has been found on most key cell types linked integrally to an immune response, i.e. monocytes/macrophages, neutrophils, dendritic cells, as well as T and B lymphocytes 192021 (Figure 1). Second, RAGE is expressed on endothelial cells and acts as an adhesion receptor for leukocytes by physical interaction with β2 integrin Mac-1. Interestingly, RAGE-Mac-1 interaction is induced by pro-inflammatory RAGE ligands in vitro and promotes leukocyte recruitment in mouse models of inflammation 2223. Third, among the broad spectrum of extracellular ligands that trigger RAGE signalling, many have been functionally linked with acute and chronic immune responses 242526. Fourth, engagement of RAGE induces activation of the transcription factor NF-κB and some of its downstream target genes that are well-known regulators of the adaptive and innate immune system 10. It is worthwhile to note that RAGE also exhibits a functional NF-κB binding site in its proximal promoter and has been shown to be a direct target gene of NF-κB signalling 27. This might explain, at least in part, why RAGE expression is absent or very low in normal adult tissues, except for the lung, and increases under stress and pathological conditions 282930. Furthermore, ligands of RAGE have been shown to accumulate and trigger intracellular activation of NF-κB at sites of tissue damage and inflammation 31. This interconnected signalling enables a cycle of prolonged cellular response supporting the establishment of chronic tissue alterations 10. In addition, sustained NF-κB signalling in response to RAGE activation is also mediated by de novo RelA (p65) mRNA synthesis, which results in a growing pool of transcriptionally active NF-κB, overriding endogeneous negative feedback mechanisms 32.

Important insights regarding the specific function of RAGE in pathological conditions affected by innate and adaptive immune responses have been obtained from mouse models in which Rage signalling was inhibited. This was achieved by (i) administration of blocking antibodies or recombinant soluble RAGE (sRAGE), a truncated form that acts as a potent decoy-receptor for ligands, (ii) expression of a dominant-negative Rage (DN-Rage) transgene in genetically modified mice, and (iii) complete or tissue-specific deletion of Rage (Rage^-/-). Treatment with sRAGE and expression of DN-Rage interfered with inflammatory conditions and revealed therapeutic effects in mouse models of diabetic atherosclerosis, delayed-type hypersensitivity (DTH), neuroinflammatory disorder, colitis, periodontitis, and others 1833. However, it is important to note that sRAGE affects immune responses even in the absence of functional Rage, leading to the assumption that its beneficial effects in mouse disease models are not solely caused by preventing ligand engagement of Rage but also of other receptors 34. Studies with Rage-deficient mice provided further in vivo evidence for the central role of RAGE in initiation and perpetuation of immune responses 18. Although Rage^-/- mice display a mild pro-inflammatory phenotype characterized by a moderately increased basal NF-κB activation and cytokine expression, they are protected from lethal septic shock in a model of cecal ligation and puncture (CLP), and survive in a Listeria monocytogenes infection model 3435.

While an important function of RAGE signalling is widely accepted in myeloid cells affecting primarily the innate immune response, opposing data were published in the context of the adaptive immune response. On the one hand, Liliensiek and colleagues reported that Rage^-/- mice showed unchanged onset and progression of disease in models of DTH and experimental autoimmune encephalomyelitis (EAE), suggesting a normal T cell-dependent immune response 34. On the other hand, specific loss of Rage function in T lymphocytes impaired priming of adaptive immune responses in vivo 2136.

RAGE expression has been detected in a variety of human tumours, including brain, breast, colon, colorectal, lung, prostate, oral squamous cell, and ovarian cancer, as well as lymphoma and melanoma 37. Although our knowledge on the molecular function of RAGE during neoplastic transformation and malignant progression is limited, recent experimental data ranging from in vitro analyses to studies in mouse models and clinical data support a direct link between RAGE activation and proliferation, survival, migration, and invasion of tumour cells 3738. Moreover, gene expression studies with samples of human patients and mouse tumour models revealed overexpression of RAGE ligands in most types of solid tumours 2439. Taguchi and colleagues found that blockade of the Hmgb1-Rage axis suppressed tumour growth in two independent mouse models, and thereby provided for the first time experimental data for an in vivo function of Rage during cancer development 40. Yet, in contrast to the concept of RAGE as a potent inducer of tumour growth and malignant conversion, few reports raised the perception that it also has tumour-suppressive functions in distinct cell types. As an example, RAGE is highly expressed in lung tissue, especially at the site of alveolar epithelium, and its expression is significantly reduced in lung carcinomas 4142. This suggests that lung cancer progression may be enhanced by loss of RAGE function and indeed re-expression of RAGE in lung tumour cell lines reduced their proliferation and revealed diminished tumour growth in athymic mice 434445. In addition, functional inactivation of RAGE in myoblasts resulted in reduced myogenesis and enhanced proliferation and invasion in vitro as well as increased tumour growth in vivo 4647. Nevertheless, tissue-specific differences in levels of RAGE expression, its splice variants, and ligands give room for the speculation that RAGE might exhibit tumour-suppressive functions in tissues characterized by constitutive RAGE expression, while it fulfils a tumour-promoting role in tissues with inducible RAGE expression.

Given the evidence from a large body of experimental data that signals downstream of RAGE can fuel chronic inflammatory conditions (see above), one could speculate on its potential impact on creating a microenvironment that is ideal for neoplastic transformation and malignant conversion (Figure 1). Indeed, the usage of Rage-deficient mice (Rage^-/-) in well-established mouse models of inflammation-associated carcinogenesis, such as chemically induced skin carcinogenesis and colitis-associated cancer, provided direct genetic evidence for a novel role of Rage signalling in linking chronic inflammation and cancer 2048. Chemically induced skin carcinogenesis is one of the best-established in vivo models to study the multistage nature of tumour development. Moreover, it represents a classical inflammation-associated tumour model, as its promotion phase solely depends on repeated topical treatments with the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA), a potent inducer of dermal inflammation 49. We could show that Rage^-/- mice were almost resistant to 9,10-dimethyl-1,2-benzanthracene (DMBA)/TPA-induced skin carcinogenesis and that the few tumours that developed in the absence of Rage were smaller in size, less progressed, highly differentiated, and hyperkeratotic 20. Importantly, dramatically decreased levels of pro-inflammatory mediators (e.g. Ptgs2, S100a8, S100a9, and macrophage inflammatory proteins) and reduced numbers of infiltrating immune cells accompanied impaired tumour formation in Rage^-/- animals. These data suggest a severe defect in sustaining inflammatory conditions during the promotion phase. One intriguing finding of the study was the ability of Rage to induce its own ligands, as we found strongly induced S100a8 and S100a9 expression in epithelial cells only in the presence of Rage. Hence, the S100-Rage axis might generate a feed-forward loop that further aggravates an inflammatory environment 20. In addition, S100A8/S100A9 could directly induce proliferation of tumour cells via RAGE ligation, as was shown recently in a cell culture model 50. Bearing in mind the importance of the pro-inflammatory microenvironment in tumour development, it is interesting to note, that bone marrow chimera experiments revealed Rage expression on immune cells, but not keratinocytes or endothelial cells, to be essential for innate immune cell recruitment and induction of epidermal hyperplasia in vivo 20. This is in line with published data on a model of HMGB1-induced peritonitis 23. Nevertheless, Rage expression on keratinocytes might also contribute to the transformation and malignant progression of epithelial cells, and it will be a challenge for the future to investigate cell type-specific Rage knockout mice using the DMBA/TPA tumour model.

In line with the data obtained from the model of inflammation-associated skin carcinogenesis, Turovskaya and colleagues demonstrated a crucial role for Rage in a mouse model of colitis-induced cancer (CAC) 48. Although Rage was dispensable for initial acute inflammation, Rage^-/- mice showed defects in the transition to chronic inflammatory conditions and revealed significantly less tumours. Interestingly, impaired CAC was not only observed in Rage^-/- mice, but also in wt mice after administration of mAbGB3.1, a monoclonal anti-carboxylate glycan antibody 48. Glycosylation of the extracellular part of RAGE significantly increases receptor affinity for HMGB1 and S100 proteins. In addition, carboxylated glycans were found on a subset of RAGE on colon cancer cells 5152.

A common feature of both mouse models, chemically induced skin carcinogenesis and CAC, was a reduced number of Gr1⁺CD11b⁺ myeloid precursor cells in the absence of Rage function, implying that RAGE signalling also contributes to the development and recruitment of myeloid-derived suppressor cells (MDSCs). Accumulation of MDSCs is one of the major immunological abnormalities in cancer resulting in T cell tolerance and suppression of anti-tumour immune response 53 (Figure 1). The molecular mechanisms of this phenomenon remain largely undefined, however, a recent study demonstrated that tumour-induced and Stat3-dependent up-regulation of S100a8 and S100a9 proteins in myeloid precursor cells are necessary for the accumulation of Gr1⁺CD11b⁺MDSCs 54. Moreover, Sinha and colleagues found that S100a8/S100a9 produced and secreted by tumour cells binds to Rage on MDSCs and promotes their migration and accumulation through NF-κB signalling pathways 55.