HMGB 1 , S 100 proteins and other RAGE ligands in cancer-markers , mediators and putative therapeutic targets

Background and Aims. Activation of RAGE due to its increased expression in cancer cells or its stimulation by multiple ligands (AGEs, HMGB1, S100 proteins, etc.) may contribute to the proliferation, invasiveness of tumor cells and formation of distant metastases and also to the resistance of cancer to treatment. RAGE ligands could thus become both useful markers of disease severity and its outcome and, a potential therapeutic target. Conclusions. Better understanding of the role of RAGE activation in different types of cancer may help to define the role of ligand/RAGE antagonists as promising cancer treatment.


INTRODUCTION
Mechanisms of innate immunity mediate reactions to exogenous pathogens.They can also mediate the response to dying or modified cells and may play an important role in the pathogenesis of cancer.Damaged necrotic cells release different damage-associated molecular patterns (DAMPs, alarmins) which are then recognized by different pattern-recognition receptors (PRRs), activate innate immunity and induce subsequent inflammation 1 .
The prototypic DAMP is the high mobility group box 1 (HMGB1), a DNA-binding nuclear protein released from the cells either passively during cell death, or actively on cytokine stimulation and activating PRRs (namely TLRs and RAGE).HMGB1 may associate with Toll-like receptor (TLR) ligands and activate cells through either TLR2 and TLR4, or the multiligand receptor called RAGE (receptor for advanced glycation end-products).Whereas TLRs are more often involved in detecting PAMPs (pathogen associated molecular patterns), they are also involved in recognizing endogenous molecules associated with tissue damage 2 .Some ligands (e.g.HMGB1, S100A8/A9 and LPS) and signaling pathways may be shared by TLRs and RAGE and they may also cooperate in the innate immune response 3,4 .
RAGE is a single-transmembrane, multiligand receptor, a member of the immunoglobulin superfamily and the gene for RAGE is located in the Class III region of the major histocompatibility complex 5,6 .Apart from membrane-bound RAGE, there is also a soluble RAGE which may partly function as its natural antagonist 7,8 .RAGE is constitutively expressed only in the lungs, but its activation may be involved in tissue homeostasis, resolution of inflammation and tissue repair after acute injury in inflammation but under pathological condition, the activation of RAGE has also been demonstrated in diabetes, atherosclerosis, nephropathy, neurodegeneration and cancer 9,10 .The effects of RAGE activation greatly depend on cell type and overall context but in the setting of limited nutrients or oxygenation, RAGE activation results in enhanced autophagy, diminished apoptosis, and (in the case of ATP depletion) necrosis 6 with a potential role not only in inflammation but also in cancerogenesis and cancer progression 10 .

Major RAGE ligands -AGEs, high mobility group B1 (HMGB1) and S100 proteins
RAGE may be bound by many ligands which include advanced glycation endproducts (AGEs), certain members of the S100/calgranulin family, extracellular HMGB1 -amphoterin, the integrin Mac-1, amyloid beta-peptide and amyloid fibrils 2,10 .Serving as a counter-receptor for leukocyte integrins (β-2 integrins) RAGE may also play an important role in cell adhesion and clustering as well as recruitment of inflammatory cells 11 .Other important ligands for RAGE may be glycosaminoglycans (including chondroitin sulfate, dermatan sulfate and heparan sulfate) which are frequently attached to proteoglycans on the surface of cancer cells and play an important role in the malignant transformation of the tumor and metastasis 12 .Lysophosphatidic acid regulates proliferation, survival, motility and invasion of cells.Lysophosphatidic acid avidly binds to RAGE and RAGE is required for its signaling 13 and phosphorylation of Akt and cyclin D, which may result in promoting carcinogenesis.Although these ligands are chemically very different they all share a negative charges on their surface and have a tendency to oligomerize 2 .
AGEs are formed by nonenzymatic glycation and during oxidative and carbonyl stress and, they represent a heterogeneous group of compounds 14 .Highly reactive cabonyl compounds (e.g.methyglyoxal, glyoxal, 3-deoxyglucoson) whose increase is characteristic for carbonyl stress are generated via oxidative and non-oxidative pathways and are precursors of AGEs.Pentosidine, carboymethyllysine (CML) are methylglyoxallysine dimer (MOLD) and are the best known AGE-products.AGEs modify proteins.Some of them cause crosslinking, and have toxic effects via binding to specific receptors, of which RAGE is the best known.Food and tobacco smoke represent exogenous sources of AGEs.AGEs may be degraded on their binding to RAGE (ref. 15) and some of their deleterious vasuclar effects may be also mediated by RAGE.
The high mobility group box 1 (HMGB1) protein is an abundant non-histone component of chromatin wellknown for its two DNA binding domains, HMG box A and HMG box B. The main characteristics of the HMGB1 protein as an "architectural" factor, are its ability to recognize and bind with high affinity to distorted DNA and its ability to induce kinks in linear DNA fragments 11,16,17 .HMGB1 plays important intranuclear, cytosolic and extracellular roles in the regulation of autophagy 18 , a degradation of namely dysfunctional organelles and proteins to generate metabolic fuels during starvation.
Although loosely bound to chromatin, HMGB1 can be either passively released from necrotic, but not apoptotic cells, or actively secreted by activated macrophages in a partially tumor necrosis factor-dependent manner, or in response to inflammatory and angiogenic signals.HMGB1 interacts with RAGE on endothelial cells causing activation and leukocyte recruitment 16 .HMGB1 correlates with the inflammatory response and may also act as an endogenous pyrogen 11 and may function as a cytokine, differentiation 16 and proangiogenetic factor 19 .
HMGB1 can activate both TLRs (TLR2 and TLR4) and RAGE resulting in increased production of inflammatory mediators, but its main signaling pathway is activated through the interaction with RAGE (ref. 20) resulting in the activation of NF-κB further promoting inflammation 19 in a positive feedback loop and contributing thus to sustaining inflammation and angiogenesis under different pathological conditions.
HMGB1 has been implicated in different disease states, including Alzheimer's disease, sepsis, ischemia-reperfusion, arthritis, and cancer and targeting the HMGB1 ligand or its receptor may have important potential application in the treatment of these diverse pathological conditions.
The S100 protein family is the largest subgroup within the superfamily of proteins carrying the Ca2+-binding EFhand motif expressed only in vertebrates 21 with a plethora of tissue-specific intra-and extracellular functions.
The S100 protein family consists of 24 members involved within the cells in the regulation of proliferation, differentiation, apoptosis, calcium homeostasis, energy metabolism, inflammation and migration/invasion interacting with a variety of target proteins, e.g.enzymes, cytoskeletal subunits, transcription factors and nucleic ac-ids 22 .Some S100 proteins may be released or secreted and regulate cell functions in an autocrine and paracrine manner via different cell surface receptors (including RAGE and TLR4, or G-protein-coupled receptors) on many cell types (e.g.macrophages, endothelial and epithelial and vascular smooth muscle cells) and interact with heparan sulfate proteoglycans and N-glycans.S100A4 and S100B interact also with epidermal growth factor (EGF) and basic fibroblast growth factor (FGF2) enhancing the activity of their receptors.S100 proteins participate in regulating both innate and adaptive immunity, chemotaxis and cell migration, tissue development and repair and also leukocyte and tumor cell invasion.S100 proteins are thus implicated in the pathogenesis of many diseases, including inflammatory and neurodegenerative diseases and cancer.

AGEs, their metabolism and cancer
Activation of RAGE by AGEs was shown to stimulate growth and/or migration of pancreatic cancer, melanoma and breast cancer cells 23 .In the case of breast cancer cells, this effect canbe completely prevented by metformin 23 (Table 1).
Elevated levels of Nε-carboxymethyllysine (CML) were asociated with increased risk of pancreatic cancer but this relation was attenuated after adjustment for body mass index and smoking 24 .In a large study in a prospectively followed Finnish population CML-AGE and soluble RAGE (sRAGE) concentrations were inversely associated with liver cancer 25 .Unexpected negative association between CML-AGE and the risk of liver cancer is difficult to explain (CML-AGE may not be an optimal indicator of overall AGE production.Other RAGE ligands, e.g.HMGB1 may be more important in activating RAGE in this setting).In another large prospective cohort of the Finnish male smokers, low sRAGE, but not high CML-AGE was associated with increased risk of colorectal cancer 26 .
Serum levels of AGEs and also AOPP (advanced oxidation protein products) were also increased in patients with breast cancer compared to controls 27,28 even in the early stages (stage I and II) of the disease.Patients with advanced breast cancer (stage III and IV) had higher se- rum levels of both AGEs and AOPP not only compared to controls, but also to patients in early stages of the disease suggesting a role of carbonyl and oxidative stress in malignant transformation and progression of breast cancer.
In conclusion, although AGEs/RAGE interaction was shown to stimulate cancer cell proliferation available clinical data do not demonstrate AGEs (at least CML-AGEs) to be a good marker of the risk of either colorectal or liver cancer, although they may be related to the progression of breast cancer.The role of AGEs (and their activation of RAGE) remains to be established and the putative role of confounding factors (diabetes, impaired renal function) needs to be taken into consideration.
Impaired metabolim of AGEs may also play an important role in cancer.Methylglyoxal, a precursor of AGEs and potent inducer of apoptosis, may be degraded by the system of glutathione-dependent glyoxalases composed of glyoxalase I (Glo I) and glyoxalase II (Glo II) enzymes.Glyoxal and methylglyoxal bind to DNA and induce multibase deletions and base-pair substitutions, mostly occurring at G:C sites 29 (Table 1).Glyoxalase may be activated in chronic inflammatory diseases and in diabetes and uremia to counteract the increased carbonyl and oxidative stress with the overproduction of AGEs, ALEs and AOPP (ref. 14).S100A12, a RAGE ligand may decrease the expression of glyoxalase I and impair the degradation of AGEs 30 .
Mutations arising from DNA glycation could explain the link of carbohydrate intake to incidence of colorectal cancer and increased risk of cancer in patients with diabetes 29 .Suppression of nucleotide glycation by glyoxalase I protects DNA not only in normal, but also in cancer cells from damage and contributes to its recovery and experimental overexpression of glyoxalase I confers the cancer cells with resistance to drug-induced apoptosis 31 .Small cell-permeable glyoxalase inhibitors were shown to counteact drug resistance in lung and prostate cancer 31 .
Alterations in the expression of the glyoxalase genes have been reported in several human cancers.Amplification of glyoxalase 1 gene was demonstrated in some tumours (e.g. in invasive ovarian and breast cancer 29 , especially drug-resistant tumour cell lines and increased glyoxalase 1 expression may be induced by malignant transformation and antitumour treatment, possibly supporting the viability of cancer cells with high glycolytic rates.In breast cancer, glyoxalase gene expression may be induced by 17β-estradiol (probably due to the presence of estrogen response element in the glyoxalase gene) resulting in much higher activity of both glyoxalases I an II in cancer compared to normal breast tissue 32 .Glyoxalase I was also overexpressed in the majority of patients with breast cancer and its upregulation correlated with advanced tumor grade 33 .
The activity of glyoxalase I may be modulated by various polyphenols, e.g.curcumin is a strong competitive inhibitor of glyoxalase I with concomitant antiinflammatory activity 34 .Curcumin was shown to inhibit the growth of breast and prostate cancer and brain astrocytoma and could be a promising anticancer compound.Resistence to doxorubicine in leukemia cells associated with increased glyoxalase I activity may be counteracted by the administration of thiazolidinedione troglitazone which downregulates glyoxalase I expression 35 .
Some cancers are associated with high glyoxalase activity which may contribute to the progression and resistance to chemotherapy.Some small cell permeable glyoxalase I inhibitors (e.g.polyphenols, PPARγ agonists) could be a good adjunct treatment in different types of cancer.
Polymorphism of glyoxalase I was studied in several types of cancer.We studied the Glu111Ala polymorphism of glyoxalase I in patients with breast cancer 36 and were able to show that the higher frequency of the mutated C allele was found in patients with negative estrogen receptors and in patients and more advanced disease (clinical stage III) compared to controls (P < 0.05) suggesting that the presence of the C allele could be a negative prognostic factor in breast cancer.In a large Italian study of patients with breast cancer cancer 37 polymorphism of glyoxalase I was associated with breast cancer in univariate analysis but a number of confounding factors obfuscate the results.In another large study in Malaysian patients with breast cancer 38 the genotype and allele frequencies of Ala111Glu glyoxalase I polymorphism were not significantly different for patients and controls.The Glu allele genotype, was, however, associated with the absence of progesterone receptor.No difference in allelic and genotype frequencies of glyoxalase I gene was found in patients with pancreatic cancer 39 .Very recently in a large study, Ala111Glu glyoxalase polymorphism was shown to be associated with clear cell renal cancer and RAGE polymorphisms (-429T/C and 2184A/G) with its aggressiveness 40 .
The role of glyoxalase gene polymorphisms in cancer thus remain uncertain, but impaired glyoxalase activity may be associated with clear cell renal cancer and with outcome in breast cancer.Further studies are to elucidate the role of glyoxalase in cancer are therefore warranted.

HMGB1 and cancer
HMGB1 released from necrotic cancer cells (e.g.due to chemotherapy) may stimulate (through RAGE) the proliferation of the remnant cancer cells and metastasis contributing to the resistance to cancer therapy.In the mouse colon cancer model, lung and liver metastasis after doxorubicin treatment were abrogated by anti-HMGB1 treatment 41 (Table 1).Tumor cells require ATP to support their proliferation.HMGB1-RAGE enhance tumor cell mitochondrial complex I activity, ATP production, tumor cell proliferation and migration.In the experimental setting, blockade of RAGE, or inhibition of HMGB1 release diminish ATP production and retard tumor growth 42 .CpG oligodeoxynucleotides enhance the growth and invasive potential of lung cancer cells.Cpg oligodeoxynucleotides stimulate the secretion of HMGB1 and blockade of extracellular HMGB1 abrogated the CpG oligodeoxynucleotides-induced progression of cancer cells.Activation of both TL4 and RAGE was critical for the response to HMGB1 in this mode 43 .
Table 1.Types of cancer associated with RAGE ligands and their putative effects.

Type of cancer RAGE ligand
Lung cancer S100A4, S100A6 and RAGE expressed in lung cancer, S100P detected in lung cancer associated with malignant phenotype, hormone independence and resistance to chemotherapy, progression and metastasis possibly through autocrine RAGE-mediated signaling (ref. 52,57)

Colorectal cancer
Highest quintile of CML-AGE not associated with an increased risk of colorectal cancer, HMGB1 and its receptors overexpressed in colon cancer and associated with the proliferation and metastasis, S100P detected in colon cancer associated with malignant phenotype, resistance to chemotherapy, progression and metastasis possibly through autocrine RAGE-mediated signaling, S100P may promote the development of colon cancer, S100A4 prognostic biomarker of the formation of distant metastases and reduced overall and metastasis-free survival (ref. 19,26,57,54,60) Hepatocellular cancer CML-AGE (and sRAGE) concentrations inversely associated with liver cancer, HMGB1 levels increase after transarterial chemoembolization of liver cancer (ref. 25,49)

Pancreatic cancer
AGEs shown to stimulate through RAGE growth and/or migration of pancreatic cancer, elevated prediagnostic levels of CML asociated with increased pancreatic cancer risk, no difference in polymorphisms of GLO genes in pancreatic cancer compared to controls, S100P detected in pancreas cancer associated with malignant phenotype, hormone independence and resistance to chemotherapy, progression and metastasis possibly through autocrine RAGE-mediated signaling, S100 proteins (namely S100A11 and S100P) involved in the progression and metastases of pancreatic cancer and associated with poor outcome after surgical resection, expression of S100A4 and S100P associated with drug resistence, differentiation, metastasis and clinical outcome (ref. 23,24,39,57,61) Kidney cancer EN-RAGE related to obesity status in renal cancer, upregulated in tumor tissue of clear cell renal cancer, especially in pts with poorer overall survival and related to obesity status, Ala111Glu polymorphism linked to clear cell renal cancer (ref. 40,71) Prostate cancer HMBG1 co-expressed with RAGE in the prostatectomy specimens from a majority of pts with metastatic prostate cancer, S100P detected in breast, prostate, pancreas, lung and colon cancer associated with malignant phenotype, hormone independence and resistance to chemotherapy, progression and metastasis possibly through autocrine RAGE-mediated signaling, S100A4 expression related to progreesion of prostatic cancer in the mouse model, S100A8 and S100A9 (and RAGE) expresssion enhanced in human prostate cancer and both proteins secreted by prostate cancer cells, serum levels of S100A9 increased in prostate cancer compared to benign prostatic hyperplasia (ref. 46,55,57,65,66)

Melanoma
AGEs shown to stimulate through RAGE growth and/or migration of melanoma cells, HMGB1 and its receptors overexpressed in melanoma and associated with the proliferation and metastasis , S100A4 secreted by tumorassociated macrophages may contribute to the progression of melanoma and increase metastatic potential of melanoma cells, S100A8/S100A9 (calprotectin) may stimulate migration of melanoma cells by the RAGE-independent mechanism, S100B overexpressed in melanoma and reliable prognostic biomarker (ref. 19,23,56,67,70)

Type of cancer RAGE ligand
Brain tumors HMGB1 released from necrotic cells induces proliferation and migration in human malignant glioma cells, S100B overexpressed by gliomas, its expression related to the infiltration with tumor-associated macrophages, inflammatory response and increased vascularity, these effect may not be mediated by RAGE, but possibly through CCL2 (ref. 45,69)

Chronic lymphatic leukemia
Drug resistance in leukemia cells asociated with upregulation of glyoxalase I, HMGB1 levels increased in pts with CLL and HMGB1 concentration associated with absolute lymphocyte cell count, CLL cells passively release HMGB1, release of HMGB1 related to the differentiation of nurse-like cells (NLC), S100A8 promotes (through RAGE activation) autophagy of leukemia cells and contributes to the drug resistance of leukemia cells (ref. 35,47) Lymphoma S100A2 expression observed in lymphoma biopsies (ref. 52) Table 1.continued HMGB1 and its receptors are widely overexpressed (and its protein levels increased) in virtually every examined type of cancer 19 and its overexpression was (usually in parallel with the overexpression of RAGE) associated with the proliferation and metastasis of many tumor types, including breast, colon, melanoma, and others.On the other hand, data on the relation of HMGB1 to the histological grade of the tumor are very limited 44 .In human malignant glioma cells, HMGB1 induced a dosedependent increase in cell proliferation and cell migration.In glioma cells (as in other types of cancer) HMGB1 is predominantly bound in the nucleus and is released only by necrotic glioma cells (e.g. after chemotherapy or radiotherapy 45 ).
The HMGB1 protein has been correlated to cancer progression, especially invasion and metastasis in melanoma, colon, prostate, pancreatic and breast cancer 17 .HMBG1 is co-expressed with RAGE in the prostatectomy specimens from a majority of patients with metastatic prostate cancer, but in only less than a quarter of nonmetastatic prostate cancers.The invasive capacity of prostate cancer cells could have been suppressed in vitro with HMGB1 antisense S-oligodeoxynucleotide, but, on the other hand, HMGB1 secretion was induced by androgen deprivation 46 .
HMGB1 may be an important survival factor for malignat B cells in chronic lymphocytic leukemia (CLL).HMGB1 levels are increased in patients with CLL compared to controls and are associated with absolute lymphocyte cell count.CLL cells are able to passively release HMGB1 and this release is related to the differentiation of nurse-like cells (promoting growth of malignant lymphocytes) which can be blocked by inhibiting RAGE and TLR9 (ref. 47).
In another study, serum levels of HMGB1 were lower (in parallel with sRAGE) in patients with metastatic breast cancer compared to controls 48 , but patients with no response to neodajuvant chemotherapy tended to have, compared to patients achieving complete and partial remission, higher HMGB1 and lower sRAGE levels before therapy suggesting that both parameters could serve as prognostic markers of therapeutic response.
HMGB1 levels increase early after transarterial chemoembolization of liver cancer, but the response to this treatment is not associated with high HMGB1, but low sRAGE levels 49 .
HMGB1 thus seems to play an important role in the stimulation of the proliferation of cancer cells, especially after previous chemotherapy.Except for competitive antagonists of HMGB1, HMGB1 may be also bound by a range of small natural or synthetic molecules, e.g.glycyrrhizin, or gabexate mesilate and neutralized by HMGB1-specific antibodies 50 .The putative role of these interventions in the treatment of different types of cancer remains to be established.

S100 proteins and cancer
S100 proteins may play a role in different stages of tumorigenesis including cell differentiation, cell cycle regulation, cell growth, apoptosis, cell motility, migration, invasiveness and metastasis formation and also tumor microenvironment 51 .
Expression of different S100 proteins was demonstrated in tumour cells and may be specific for different types of cancer.Using microarray technology to study S100 protein expression in tumor samples, S100A2 expression was observed in lymphoma biopsies, S100A4 and S100A6 (in parallel with RAGE) was abundant in breast and lung tumours 52 (Table 1).
S100A4 was secreted by both tumor and stromal cells in melanoma xenograft model and supported (via RAGE) tumorigenesis and angiogenesis synergizing with vascular endothelial growth factor (VEGF) and also promoting endothelial cell migration by increasing MMP-9 activity.Endothelial cell migration, tumor growth and angiogenesis could have been abolished in this mode by the administration of anti-S100A4 monoclonal antibody 53 .S100A4 increased cell migration and invasion also in colon cancer cells and this effect may have been counteracted by soluble RAGE and anti-RAGE antibodies 54 .S100A4 also progresssively increased in prostatic tissue with the progression of the disease in the mouse model of prostate cancer and S100A4 cancer cells grew more quickly (via RAGE and NFκB activation) than S100A4 negative cells 55 .As heterozygously deleted S1004 mice exhibited an increased tumor latency period, reduced prostatic weights and no metastases S100A4 inhibition could be a promising therapeutic option to be tested in prostate cancer.S100A4 may be synthesized not only by tumor, but also by stroma cells.In melanoma, infiltrating tumor-associated macrophages may also secrete S100A4 and in this paracrine manner increased the metastatic potential of melanoma cells 56 .Expression of another member of the S100 family, S100P, was detected in a spectrum of human tumor cell lines and tissues derived from breast, prostate, pancreas, lung and colon, where it was connected with malignant phenotype, hormone independence,resistance to chemotherapy and metastasis 57 .S100P was shown to stimulate RAGE and increase proliferation and prolong survival of cancer cells in an autocrine manner 58 Its effects can be blocked by anti-RAGE antibodies.S100-P derived small antagonistic peptides could be potentially used to block tumor proliferation 59 .In colon cancer cells elevated S100P stimulated RAGE, AP-1 and induced oncogenic miR-155 and S100P-induced proliferation, motility and invasion may have been blocked either by anti-RAGE antibodies, or miR-155 knockdown 60 .
In pancreatic cancer, the expression of S100A4 and S100P was associated with drug resistence, differentiation, metastasis and clinical outcome and S100A11 and S100P, and possibly also S100A2 and S100A6 were related to the unfavorable outcome of the patients after surgical resection 61 .
S100A7/psoriasin was highly expressed in high-grade ductal breast carcinoma in situ 62 , high-grade comedo ductal carcinoma in situ with higher risk of local recurrence 63 and in invasive estrogen receptor negative breast cancer 64 .Its expression was related to ductal hyperplasia, recruitment of tumor-associated macrophages, tumor growth, angiogenesis and metastasis.Tumor growth was inhib-ited by the downregulation of psoriasin by short hairpin RNA (shRNA) resulting in decreased expression of vascular endothelial growth factor (VEGF) (ref. 62).Invasion of breast cancer may be also stimulated by S100A8/A9 via RAGE-mediated epithelial-mesenchymal transition through NF-κB mediated stabilization of Snail.In invasive ductal carcinoma S100A8/A9 binding is associated with lymph node involvement and distant (lung) metastases 65 .
Deregulated expression of different S100 proteins, e.g.S100A8 and S100A9, is associated with different neoplastic disorders.S100A8 and S100A9 expresssion is enhanced in human prostate cancer and are secreted by prostate cancer cells 66 .Extracellular S100A8 and S100A9 activate RAGE, induce the activation of NF-κB and increased phosphorylation of p38 and p44/42 MAP kinases and stimulate migration of benign prostatic cells in vitro.S100A8, S100A9 and RAGE are co-expressed and upregulated in prostatic intraepithelial neoplasia and preferentially in high-grade adenocarcinomas but not in benign prostatic tissue.Serum levels of S100A9 may help to distinguish between prostate cancer and benign prostatic hyperplasia 66 .
S100A8 and S100A9 may dimerize and form calprotectin. Cell surface glycoprotein EMMPRIN/BASIGIN (CD147/BSG) may serve as a receptor for calprotectin and was demonstrated to specifically bind S100A9 and induction of cytokines, matrix metalloproteinases and cell migration by S100A9 may be downregulated in melanoma cells by attenuation of EMMPRIN independently of RAGE 67 .
S100A8 are elevated in drug-resistant leukemia cell lines and may play an important role in the drug resistance of leukemia cells by promoting autophagy.Adriamycine and vincristine increase S100A8 in human leukemia cells in parallel with the upregulation of autophagy 68 .Knockdown of S100A8 induced by RNA interference restored the chemosensitivity of leukemia cells.S100A8 could be thus a novel therapeutic target for improved drug sensitivity in leukemia therapy.
S100B is overexpressed by gliomas and its downregulation of S100B abrogates tumor growth in vivo and is related to higher infiltration with tumor-associated macrophages, stronger inflammatory response and increased vascularity.As RAGE ablation had no effect on the infiltration of gliomas with tumor-associated macrophages, other pathways (possibly CCL2 expression) may be involved and could be targeted 69 .S100B expression may also be a prognostic marker in malignant melanoma 70 .S100A12 (ENRAGE) is related to obesity status in clear cell renal cancer (which is a risk factor of this type of cancer) and is overexpressed in tumor tissue, especially in patients with poorer overall survival.S10012 may serve through the activation of RAGE as an autocrine stimulator of the tumor growth 71 .

Targeting RAGE and its ligands -can it contribute to the treatment of breast cancer?
The activity of RAGE and its ligands may be suppressed in many possible way using both small molecules, monoclonal antibodies, or siRNAs (Table 2).
Both statins 72 and angiotensin receptor blockers 73 inhibit RAGE signaling in diabetic nephropathy, but no data on their putative effect on RAGE activation in cancer are available.Metformin use may decrease the incidence and mortality in breast cancer by inhibition of AGEs induced proliferation of breast cancer cells 23 .Peroxisome proliferator-activated receptor-gamma (PPARgamma) agonists were also shown to interfere with the AGE-RAGE system 74,75 .Antiallergic drug cromolyn and its derivatives may block the interaction of S100P with RAGE and high concentrations of cromolyn were shown to improve gemcitabine effectiveness in pancreatic ductal adenocarcinoma by inhibiting S100P-induced increase of NF-κB, cell growth and apoptosis and cromolyn derivatives may be promising drugs to block S100P in different types of cancer 76 .
Heparin and its low anti-coagulant derivative 2-0,3-0-desulfated heparin disrupt CD11b/CD18 mediated leukocyte adhesion to RAGE and inhibited activation of RAGE by AGEs, HMGB1 and S100 proteins.These heparins with low anticoagulant activity were also able to prevent metastases 77 .Altered expression of chondroitin sulfate (CS -with higher proportion of E-disaccharide units) on the surface of tumor cells may contribute to malignant transformation and metastasis which canbe inhibited in lung cancer cells by pre-administration of CS-E from squid cartilage rich in E units or antibodies against CS-E interfering with CS-RAGE signaling suggesting a putative role of these approaches in the treatment of pulmonary metastasis 78 .
AGEs induced proliferation of breast cancer cells could be completely prevented by anti-RAGE antibodies 23 .Antibodies against RAGE were also shown to inhibit metastasis of experimental lung cancer and melanoma cells 12 .Administration of recombinant soluble RAGE was shown to block RAGE signaling pathway in animal models, suggesting that the circulating sRAGE could protect the tissue againgst RAGE-induced tissue damage 79 and sRAGE could serve also in cancer as a putative biomarker.
HMGB-1 can be blocked in several different ways 80 : by anti-HMGB-1 antibodies, by the inhibition of HMGB-1 release from the nucleus into the extracellular space, by HMGB-A box, a competitive antagonist of HMGB-1, by blockage of RAGE-HMGB-1 signaling using RAGE antagonists, by blockage of TLR-HMGB-1 signaling using anti-TLR2 antibodies and by other molecules that modulate HMGB-1 activity using e.g.human soluble thrombomodulin.
The recently developed small S100P-derived RAGE antagonist peptide (RAP) blocking activation of RAGE by multiple ligands was shown to inhibit the interaction of S100P, S100A4, and HMGB-1 with RAGE at micromolar concentrations.Systemic in vivo administration of RAP reduced the growth and metastasis of pancreatic tumors and also inhibited glioma tumor growth 76 .
Decrease in the proliferation of different types of experimental breast cancer with increased RAGE expression (correlating with the severity of breast cancer) could have been induced by small interfering RNA against RAGE (RAGE siRNA) (ref. 81).
Overall specific modes of RAGE inhibition and its ligands could be promising not only in diabetes and other inflammatory and neurodegenerative disorders but also in cancer.

CONCLUSIONS
The activity of the RAGE ligand(s)/RAGE system appears to be involved in a number of cancers contributing to the proliferation of cancer cells, their invasiveness, metastasis and resistence to treatment.Serum levels and tissue expression of RAGE and RAGE ligands may thus be useful biomarkers of diseaase severity and disease outcome and potential therapeutic targets.We believe that the role RAGE and its ligands in cancer will be a fruitful area for further research and will help our understanding of the pathogenesis of cancer progression and metastasis.

Fig. 1 .
Fig. 1.Activation of RAGE in cancer and its putative consequences.

Table 2 .
Putative interference with RAGE ligands and RAGE-mediated effect.