(The FASEB Journal. 2006;20:1611-1621.)
© 2006 FASEB
The Met pathway: master switch and drug target in cancer progression
Massimiliano Mazzone1 and
Paolo M. Comoglio
Division of Molecular Oncology, Institute for Cancer Research and Treatment (IRCC), University of Torino Medical School, Candiolo (Torino), Italy
1Correspondence: Institute for Cancer Research and Treatment (IRCC), Division of Molecular Oncology, University of Torino Medical School, Strada Provinciale 142, km 3.95, I-10060 Candiolo (Torino), Italy. E-mail: massimiliano.mazzone{at}ircc.it
 |
ABSTRACT
|
|---|
It has been recognized for more than a century that most tumors tend to become more aggressive in clinical behavior over time, although this time course may be variable. This phenomenon has been termed "cancer progression," a process that appears to develop in a stepwise fashion through qualitatively different stages. Cancer progression relies on the ability of neoplastic cells to abandon their primary site of accretion, trespass tissue boundaries, and penetrate into the vasculature to colonize and repopulate distant sites. Among the various properties associated with cancer progression, the acquisition by neoplastic cells of the capacity to invade locally and to metastasize is of great clinical significance, and is still the fundamental definition of malignancy. This process represents the aberrant counterpart of a physiological morphogenetic program, known as invasive growth, occurring during embryo development and, in some instances, in adulthood for the generation and maintenance of normal organ complexity and architecture. Here we summarize some of the strategies adopted to inhibit cancer cell growth and spreading. We also review the current findings about cancer and metastasis inhibitors. As we suggest possible directions for drug development, we propose the receptor for the hepatocyte growth factor, Met, as an ideal target for tackling cancer progression.—Mazzone, M., Comoglio, P. M. The Met pathway: master switch and drug target in cancer progression.
Key Words: HGF • invasive growth • microenvironment • crosstalk • inhibitors
|
INTRODUCTION
|
|---|
CANCER PROGRESSION IS a well-characterized phenomenon that has been studied intensively over the past decade (1)
. Genetic events like inactivating mutations of tumor suppressor genes (e.g., p53, Rb) or activating alterations of dominant proto-oncogenes (e.g., Ras, Myc) are required for the beginning of neoplastic progression (2
, 3)
. These events generate genomic instability, release a cell from normal growth constraints, and enable primary tumor formation. To grow beyond a minimal size, the primary tumor needs to develop a blood supply that can support its metabolic needs. This process, called angiogenesis, is mediated by the expression of angiogenic factors (such as the vascular endothelial growth factor, or VEGF) that stimulate the formation of new vessels. Subsequently, individual tumor cells break off from the primary tumor mass, degrade extracellular matrix (ECM), invade the surrounding normal tissue, and enter the blood or lymphatic circulation. Intravasated cells have to survive in the systemic stream and overcome the apoptotic program normally recruited by cells in absence of anchorage ("anoikis"). This complex process requires the concerted action of numerous factors including proteolytic enzymes (e.g., matrix metalloproteases, urokinase-like plasminogen activator, catepsins), cell adhesion molecules, tumor growth factors, and suppressors of apoptotic signals (e.g., BclII). Circulating tumor cells should then arrest in a distant organ and extravasate. Since the new microenvironment of the metastatic site differs from that of primary tumors, cancer cells might die or survive. In case of survival, there are two possible outcomes: when proliferation is balanced by the apoptotic rate, cancer cells establish clinically undetectable dormant micrometastasis. Otherwise tumor cells can generate macrometastases that resume the previously described progression and represent—from a clinical viewpoint—the Achilles’ heel of cancer therapy (1
, 4)
. Indeed, when cancer is detected at an early stage, patients have a good chance of successful cure by conventional treatments such as surgery, radiotherapy, and chemotherapy. In contrast, spreading of cancer cells from the primary tumor toward distant sites correlates with poor prognosis. For most cancer types, the acquisition of metastatic ability leads to therapeutic failure and clinically untreatable disease. Metastasis—the final step of malignancy—is often the main cause of death for cancer patients.
|
INVASIVE GROWTH IS A PHYSIOLOGICAL PROGRAM
|
|---|
The above multistep program toward malignancy is conventionally termed invasive growth (5)
. Invasive growth does not occur only in cancer cells, but also takes place in physiological conditions such as embryo development and organ formation (Fig. 1
). During embryogenesis, invasive growth is involved in complex developmental events including morphogenesis of epithelia, angiogenesis, nerve sprouting, and myoblast migration. Indeed, these events occur when embryonic cells proliferate, survive, migrate, invade the matrix, and polarize, rearranging themselves into 3-dimensional structures. In adult life, epithelial cells also exploit the invasive growth program for wound healing, when cells at the wound edge start dividing and moving over the temporary matrix to regenerate tissue integrity. The mechanism on which these physiological events are based is named epithelial-mesenchymal transition (EMT). EMT is an orchestrated series of events in which cell-cell and cell-extracellular matrix interactions are altered to release epithelial cells from the surrounding tissue. The cytoskeleton is rearranged to confer the ability to move through the extracellular environment, and a new transcriptional program is induced to maintain the mesenchymal phenotype. This dramatic reshaping is accompanied by loss of E-cadherin-mediated intercellular junctions and enhanced production of matrix proteases, which digest basal lamina components and facilitate cell motility. During this phase, invading cells must induce a constant and dynamic remodeling of integrin-mediated adhesive contacts with the extracellular cell matrix, which provides a mechanical support for cell migration and prevents the induction of apoptosis (1
, 6)
. In mammals, EMT plays a role in many stages of development, including gastrulation, in which the embryonic epithelium gives rise to the mesoderm, and in delamination of the neural crest, which produces a population of highly mobile cells that migrate to and are incorporated into many different tissues (7
, 8)
. Once migrated to their target destinations, the cells may revert to their original epithelial phenotype through a mirror process known as mesenchymal-epithelial transition. Essential for development, EMT is nevertheless potentially destructive if aberrantly activated, and it is becoming increasingly clear that inappropriate utilization of EMT mechanisms is an integral component of the progression of many tumors of epithelial tissues (9
10
11)
. During tumor progression—from benign lesion to malignancy—cells that have acquired EMT can disseminate and form metastatic lesions. Strikingly, in the established metastases the EMT process could be reverted, recapitulating the differentiated phenotype of primary tumors (12
13
14)
. After this reversion, the metastasis—in a loop fashion—can undergo another EMT, giving rise to a lesion that is often significantly different from the primary tumor (Fig. 1)
.
View larger version (45K):
[in this window]
[in a new window]
|
Figure 1. Invasive growth is a physiological program aberrantly activated during cancer progression. Invasive growth is a physiological program that results from the integration of different biological activities, including cell proliferation, survival, cell-cell dissociation ("scattering"), migration, invasion, and morphogenesis. Normally occurring in development and adulthood for the generation and maintenance of organ complexity and architecture, it is aberrantly recruited by cancer cells during tumor progression. By acquisition of genetic alterations, a normal epithelial cell is transformed into a cancer cell. Through a mechanism known as epithelial-mesenchymal transition (E.M.T.), cancer cells acquire a metastatic phenotype. Metastatic cells reach a secondary site via blood or lymphatic vessels. Strikingly, in the established metastases, the EMT process could be reverted through a mesenchymal-epithelial transition (M.E.T.). After this reversion, the metastasis—in a loop fashion—can undergo another EMT, giving rise to a lesion, which is often significantly different from the primary tumor.
|
|
Here, we propose that tumor progression, and metastasis formation in particular, results from the deregulated activation of invasive growth.
|
INVASIVE GROWTH IS CONTROLLED BY GROWTH FACTORS
|
|---|
Several cytokines and growth factors can induce proliferation, differentiation, chemotaxis, migration, and protection from apoptosis. For example, epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), fibroblast growth factor (FGF), and ephrins (EphA1, EphB1) promote cell growth, motility, and survival in some cell types, although to different extents and under defined experimental conditions. However, spatial and chronological orchestration of the various steps of this program are optimally accomplished by a family of soluble growth factors, known as scatter factors (15
, 16)
. Hepatocyte growth factor (HGF/SF1) and macrophage stimulating protein (MSP/SF2) belong to this family. In the late 1980s, Stoker showed that Madin-Darby canine kidney (MDCK) cells—a polarized epithelial cell line—displayed a fibroblastic migratory phenotype when incubated with conditioned medium from cultured fibroblasts (17)
. The factor involved in this transition, first called scatter factor, was later identified as HGF (18)
. Under physiological conditions, HGF and its receptor, Met (19)
, are essential for embryo development. In fact, Met-dependent invasive growth program is implicated in several embryogenetic processes, such as muscle development, nervous system formation, bone remodeling, and angiogenesis (20)
. Mice that carry a homozygous mutation of either hgf or met die in utero because of severe defects during placental and liver development, and are devoid of muscles that derive from the migratory myogenic precursors that detach from the myotome (21
, 22)
. In adulthood, the epithelial-mesenchymal transition promoted by HGF/Met has a crucial role during acute injury repair inducing cell migration toward the wounded area (23
24
25)
. When aberrantly activated, Met signaling causes cellular transformation, whereas its ability to enhance motility and survival accounts for neoplastic invasion and metastasis (5)
. Overexpressed or mutated Met requires ligand stimulation in order to unleash its transforming potential (26
27
28)
. Recently, a link has been demonstrated between Met oncogene activation, cell transformation, and hemostasis. Oncogenic activation of Met triggers a genetic program that not only transforms cells but also creates a rudimentary scaffold of fibrin that is an absolute requirement for supporting the expansion of the tumor itself and the incoming new vessels (29
, 30)
. Many data provide strong evidence that HGF/Met signaling plays a crucial role in tumor development and malignant progression, particularly in tumor invasiveness and metastatic potential. First, cell lines that ectopically overexpress HGF or Met become tumorigenic and metastatic in nude mice (26)
. Second, the tumorigenic potential of endogenously Met expressing cancer cells is decreased when Met is down-regulated (31)
. Third, HGF or Met transgenic mice are characterized by the appearance of metastatic tumors (32)
. Notably, Met overexpression has been detected in many different solid tumors and correlates with a poor prognosis (33
34
35)
. Finally, Met germline missense mutations are responsible for hereditary papillary renal carcinoma (36)
, while somatic mutations have been reported in malignant pleural mesothelioma (37)
, lung cancer (38)
, and head and neck squamous cell carcinomas (39)
. Somatic mutations in met gene are selected during metastatic spread and have been identified in association with the progression to metastasis (39
40
41)
.
It would be too simplistic to consider Met the only regulator of invasive growth. For example, the Wnt pathway represents a well characterized and representative signal for which the aberrant activation—observed in a wide variety of cancer types, including those of the colon, prostate, ovary—reflects the same cascade physiologically turned on during embryo development (42)
. Through the binding to Frizzled receptors, Wnt promotes self-renewal of somatic stem cells, as well as neoplastic proliferation in the same tissue when deregulated (43)
. In colorectal cancer, nuclear localization of ß-catenin, a good marker of active Wnt signaling, has a very heterogeneous pattern in tumor cells reflecting the heterogeneity of tumor cell differentiation, both within an individual tumor and in different phases of progression from adenoma to carcinoma (13
, 44)
. High levels of nuclear ß-catenin are found prevalently in dissociated, dedifferentiated tumor cells that are at the tumor-host interface that have undergone EMT. Conversely, nuclear ß-catenin decreases to central and differentiated areas of the tumor. From a temporal viewpoint, nuclear ß-catenin increases during the progression from adenoma to carcinoma. It is noteworthy that ß-catenin nuclear accumulation, following Wnt deregulation, is not sufficient to initiate EMT and produce a metastatic tumor phenotype. However, Wnt and Met signaling can act in a synergistic manner to promote EMT, increasing cell transformation and invasive behavior. It is also known that Met is transcriptionally induced by ß-catenin (45)
and Met itself promotes Wnt-independent nuclear translocation of ß-catenin (46)
. On the contrary, it has been demonstrated that the activation of Notch pathway induces the transcriptional repression of Met gene, thus acting as a physiological inhibitor of Met-dependent invasive growth. Intriguingly, in an opposite but complementary fashion, Met activation results in stimulation of the Notch pathway through transcriptional induction of the Notch ligands Delta1 and Delta4 (47)
. In conclusion, a delicate balance between positive and negative signals is critical for normal cell homeostasis, and a prevalence of positive signals leading to excessive cell stimulation is commonly found in cancers.
If several signals are required to allow cell escape, other molecules that are present in specific organs can influence whether or not various types of cancer cells will grow there. The pattern of chemokine receptors expressed on tumor cells correlates with chemokines themselves specifically produced in organs to which these cancers commonly metastasize. For example, both breast cancer cell lines and primary breast tumors were found to express the chemokine receptors CXCR4 and CCR7 at high levels. The specific ligands for these receptors—CXCL12 and CCL21—are found at high levels in lymph nodes, lung, liver, and bone marrow, all organs to which breast tumors often metastasize. Furthermore, blocking CXCR4 inhibits breast cancer cell metastatization in tumor mouse models (48
, 49)
.
Chemokines and growth factors influence the environment where the tumor grows, and may drive tumor cell potency toward the multistep progression of cancer. In other words the quality of the "soil" is crucial for the growth of the "seed" (50)
.
|
CROSSTALK BETWEEN GROWTH FACTOR RECEPTOR AND ADHESIVE MOLECULES
|
|---|
The binding of a soluble ligand to its specific receptor is only one way to trigger the activation of the invasive program. During tumor progression, specific integrin signals enable cancer cells to detach from neighboring cells, reorientate their polarity during migration, and survive and proliferate in foreign microenvironments. Integrins transmit both mechanical and chemical signals that are essential during both normal and neoplastic cell migration. There is a body of evidence demonstrating that integrin-mediated cell migration signals require the focal adhesion kinase (FAK). It has been shown that cells from FAK-deficient mice display reduced cell motility (51)
. Notably, many metastatic human cancers have elevated levels of FAK (52)
. The role of FAK in cell migration is to recruit in an integrin-dependent manner Src family kinases and position them close to effectors that are crucial for cell migration, thereby contributing to the initial assembly of focal adhesion scaffolding (53)
. Moreover, FAK seems to integrate growth factor and integrin signals to promote cell migration (54)
. As well as imparting polarity to cells and organizing and remodelling their cytoskeleton during adhesion and migration, pathways activated by integrins exert a stringent control on cell survival and proliferation (55)
. For example, the expression pattern of integrin 
6ß4 implies that this molecule controls proliferation of normal epithelial cells in the basal cell compartment (56)
. In accordance, 6ß4 and its ligand, laminin-5, are required for Ras-mediated transformation of keratinocytes (57)
. Introduction of ß4 in ß4-negative breast carcinoma cells activates signaling from phosphatidylinositol 3-kinase (PI3K) to Rac and increases the invasive ability of these cells in vitro (58)
Moreover, there is increasing evidence that certain integrins associate with receptor tyrosine kinases (RTKs) to activate signaling pathways necessary for tumor and invasion. Recently it has been demonstrated that HGF increases cell adhesion on all substrates, irrespective of the specific integrins engaged, and that the strength of the adhesion correlates with cell scattering (59)
. In addition, it is known that ß4 tail functions as an adaptor and amplifier of proinvasive signals that are elicited by the HGF receptor. The effects of the oncogenic amplification of Met signaling in some carcinoma cells are independent of adhesion to laminin-5 (60)
. This implies that constitutive activation of 6ß4 associated to Met causes ECM-independent signaling that gives an invasive phenotype to nonmetastatic cells and unleashes Met-dependent tumorigenesis (60
, 61)
.
|
INTRACELLULAR TRANSDUCTION OF INVASIVE SIGNALS
|
|---|
Despite the existence of different receptor families for adhesive cues and cytokines, studies of signal transduction have revealed that the same pool of interactors can be activated by unrelated receptors. Ideally, the intrinsic features of the receptors and the cellular context in which they function should cooperate to generate an unequivocal biological response (62)
. How are all these extracellular signals converted into cellular responses? To simplify the complexity of signal transduction in invasive cells, we can distinguish an early and a late response, the first being transcription independent whereas the second requires transcription. Boyer et al. showed that dominant-negative Src or Ras abrogates EGF-induced EMT in NBT-II epithelial carcinoma cells, suggesting that Ras and Src are required for cell dispersion. However, inhibition of RNA synthesis represses the ability of the active Ras but not Src to induce epithelial cell scattering. In contrast to Ras, Src kinases may control epithelial cell dispersion in the absence of gene expression by directly regulating the organization of the cortical cytoskeleton (63)
. One of the main targets of the dual regulation (transcriptional and post-transcriptional) during cytoskeleton rearrangement and cell reshaping is E-cadherin, an essential molecule in cell-cell adhesion. Loss of E-cadherin-mediated adhesion is required for malignant conversion (64)
. It is known that cell-cell adhesion can be disrupted by two mechanisms. First, activated RTKs and Src induce tyrosine phosphorylation of the E-cadherin complex. After tyrosine phosphorylation, the E-cadherin is recognized by the Cbl-like E3ubiquitin protein ligase Hakai and consequently down-regulated by endocytosis (65)
. Second, cell signaling operates through transcription factors to suppress E-cadherin expression and thereby disrupt adherens junctions (66)
. In certain epithelial cells, this process seems to be mediated by integrin-linked kinases (67)
as well as RTKs (68)
.
The transcriptional response has been studied with controversial findings in recentyears thanks to microarray analysis in vitro on cell lines and on tumor samples collected from patients whose cancer stage was known. Gene expression profiling has provided new information on the biology of metastasis. Studies of breast cancer identified a gene expression signature strongly predictive of a short time interval to distant metastases in patients without tumor cells in local lymph nodes at diagnosis (69)
. Another recent paper has compared the gene expression profiles of 12 metastatic adenocarcinomas of diverse origin (lung, breast, prostate, colorectal, uterus, ovary) with those of 64 primary adenocarcinomas representing the same tumor types obtained from different individuals (70)
. This comparison identified an expression pattern of 128 genes that best distinguished primary and metastatic adenocarcinomas. None of the genes associated with metastasis represents an individual marker; rather, it is the total signature that seems to contain predictive information. These results argue for generic gene expression programs related to metastasis rather than distinct mechanism of metastases in different tumors. Many genes up-regulated in the signature are components of the protein transcriptional apparatus, consistent with reports of their amplification and overexpression in invasive tumors. Moreover, transfecting tumor cells with transcriptional factors, such as Snail, Slug, and Twist, induces a mesenchymal phenotype mainly due to the loss of E-cadherin (10
, 11
, 71
72
73)
. Conversely, NBT-II cells transfected with antisense Slug cDNA were refractory to EMT induction by HGF/SF (74)
, which not by chance we elected as the master switch of invasive growth. Moreover, silencing of Slug or Twist in cancer cells prevents metastases without affecting primary tumor growth, suggesting they are key components of the metastatic program (11
, 73
, 75)
. Even if the contribution of these transcriptional factors to metastasis is the promotion of EMT, the list of controlled genes contains disparate classes of targets: small GTPases (e.g., RhoC), tyrosine kinase receptors (e.g., Met), angiogenic and growth factors (e.g., VEGF), metalloproteinases (e.g., MMP3, MMP9) and matrix remodelling proteins (e.g., urokinase type plasminogen activator, uPA), cell-cell and cell-matrix adhesion proteins (E-cadherin and integrins), matrix proteins (e.g., osteopontin), chemokines, and chemokine receptor (e.g., CXCR3) and others (70
, 76
, 77)
. We think that transcriptional regulation of genes involved in the invasive growth signature provides an initial consensus for the onset of the biological program, a sort of "wake-up call." But the definition of the "program" imposes that the initial enabling transcriptional response must be accompanied by the spatial and temporal coordination of extracellular signals, receptors, and specific transducers to drive full invasive growth. A recent work shows that primary human melanocytes, transformed by a specific set of genes, form melanomas that frequently metastasize to multiple secondary sites, whereas human fibroblasts and epithelial cells transformed with the same genes generate primary tumors that rarely metastasize (74)
. Moreover, the expression profile of genes associated with neural crest migration—from where melanocytes origin—correlates with that of adult human nevi (78
, 79)
. These observations support the idea that part of the metastatic proclivity of cancer is attributable to lineage-specific factors present before neoplastic transformation, such as during physiological processes in embryogenesis.
|
THE ROLE OF THE MICROENVIRONMENT DURING INVASIVE GROWTH
|
|---|
Tumor mass is composed of more than just cancer cells. Endothelial cells, pericytes, stromal cells, and immune cells also surround cancer cells and influence each others’ development (Fig. 2
). Crosstalk between carcinoma cells and host stromal cells such as fibroblasts has a great influence on the invasive and metastatic behavior of cancer cells. A crucial role in this dialogue is played by the HGF/Met pathway. Indeed, several types of cancer cells secrete molecules that enhance HGF production in fibroblasts, while fibroblast-derived HGF in turn is a potent stimulator of the invasion of cancer cells. Fibroblast-specific genetic alterations leading to an overexpression of HGF are associated with the development of epithelial neoplasia and invasive carcinoma, demonstrating the prooncogenic action of fibroblasts (80
, 81)
. Besides fibroblasts, cells of the immune system represent an important fraction of the non-neoplastic component of the tumor mass. Infiltrating macrophages and lymphocytes may not only be antitumorigenic, but simultaneously support cancer progression by secreting EMT-inducing cytokines, angiogenic factor, and ECM proteases (82)
. One of the most important environmental elements within the tumor mass is represented by blood vessels (83)
. To obtain nutrients for their growth and to metastasize to distant organs, cancer cells recruit both endothelial cells from the bone marrow (vasculogenesis) and new vessels, sprouting from existing ones (angiogenesis). Although both processes are involved, vasculogenesis plays a very small part in tumor vascularization. Both events are primarily controlled by VEGF but additional proangiogenic factors, including HGF, FGF-1, FGF-2, platelet-derived growth factor (PDGF), placental-derived growth factor (PlGF); angiogenic inhibitors, such as thrombospondin-1, also exert an essential role. Recent data support the idea that the generation of new lymphatic vessels (lymphangiogenesis) within the tumor mass represents an alternative route for cell escaping (84)
. VEGF-C and VEGF-D have been shown to stimulate lymphangiogenesis specifically, and blocking their receptor, VEGFR-3,, inhibits lymphogenous metastases (85)
. During cancer progression, the balance between pro- and anti- angiogenic factors tilts in favor of the stimulators. The resulting vasculature is structurally and functionally abnormal. Vessels are leaky, tortuous, discontinuous, highly disorganized, and permeable. Endothelial cells have aberrant morphology, pericytes are loosely attached or absent, and the basement membrane is often abnormal or entirely absent. Collectively these vascular abnormalities lead to an abnormal tumor microenvironment characterized by interstitial hypertension, acidosis, and hypoxia (81)
. Among these consequences, hypoxia seems to be associated with higher tumor malignancy, with increased metastatic potential and resistance to therapies (86
87
88)
. The biological parameters of hypoxic tumor cells are closely related to altered gene expression under conditions of oxygen deficiency. Hypoxia-inducible factor 1 (HIF-1) is a master transcription factor that plays a central role in hypoxic expression of a variety of genes. Glycolytic enzymes, such as aldolase, hexokinase, and lactate dehydrogenase; glucose (Glc) transporters such as glc transporter (GLUT)-1 and 3; angiogenic molecules, such as VEGF and angiogenin; survival and growth factors such as PDGF-B and IGF-II; proteins involved in the ECM degradation such as MMP-2 and uPA receptor; and drug resistance proteins such as MDR-1 are an exemplificative panel of hypoxia-induced responsive genes (89)
. Recently it was demonstrated that hypoxia induces the expression of Met. It has been shown that Met is up-regulated in hypoxic regions of human tumor, sensitizing tumor cells to HGF stimulation (90)
. In other words, hypoxia synergizes with HGF in inducing invasion.
View larger version (41K):
[in this window]
[in a new window]
|
Figure 2. Three different strategies for tackling cancer progression. A tumor is composed of more than just cancer cells. Indeed, endothelial cells, pericytes, stromal cells and immune cells also surround cancer cells and influence each others’ development. Signals from tumor microenviroment and cancer cells are transformed into a biological response by public transducers through transmembrane receptors. Public transducers can be distinguished in early and late effectors. From a pharmacological point of view, therapeutic drugs can target tumor at one or more of these levels: the microenvironment, receptor-based signals, and signal transducers.
|
|
All these data suggest that tumor microenvironment, rather than being a passive tumor scaffold, plays an active role in cancer progression.
|
THERAPEUTIC IMPLICATIONS: CAN INVASIVE GROWTH BE TARGETED?
|
|---|
Knowledge of the basis of invasive growth and tumor progression has led us to significant gains in cancer therapeutics during the last decade (91
, 92)
. The most important therapeutic advances have come from agents targeting proteins encoded by genes that are altered in cancers. These include trastuzumab (HerceptinTM), an antibody (Ab) against the product of the erbB2 gene amplified in some breast cancers; imatinib (GleevecTM), an inhibitor of tyrosine kinases altered in chronic myelogenous leukemia, like Bcr/Abl fusion protein; and gastrointestinal tumors, like c-Kit, and gefitinib (IressaTM) or erlotinib (TarcevaTM), two EGFR kinase inhibitors used to treat efficaciously lung cancers in which EGFR is mutated. Though none of these therapeutic tools have resulted in the cure of patients with advanced disease, they can substantially improve prognosis and prolong lives. According to the previous excursion on cancer progression and invasive growth, therapeutic drugs that have been and will be developed in the future have one or more of these three possible targets: the microenvironment, receptor-based signals, and signal transducers (early and late effectors as mentioned above) (Fig. 2)
.
About the possibility to target the microenvironment, many attempts have been made to destroy tumor vascularization (93)
and several Phase I clinical trials that have tested different classes of angiogenic modulators, such as the small molecule tyrosine-kinase inhibitor of FLK1 (SU5416TM), the anti-VEGF Ab (AvastinTM), the MMP inhibitor (NeovastatTM) and the anti-endothelium drugs (combretastatins), have been completed. The latter are not antiangiogenic, but rather vascular disrupting agents aimed at causing the rapid and selective shutdown of the established, but disorganized, tumor vasculature (94)
. However, all these agents might induce an adaptive tumor remodeling that select more aggressive and therapy-resistant cells, with consequent failure of the therapeutical approach. The mirror therapeutical strategy is to target hypoxia (95
, 96)
. Its master regulator, HIF-1, induces the transcription of genes that are involved in crucial aspects of cancer biology. HIF-1 overexpression has been associated with increased patient mortality in several cancer types (97)
. In preclinical studies, inhibition of HIF-1 activity has marked effects on tumor growth. One of these studies has led to the discovery of a small molecule, chetomin, as a disrupter of HIF binding to the transcriptional coactivator p300 necessary for the transcriptional response (98)
. Systemic administration of chetomin inhibited hypoxia-inducible transcription within tumors and inhibited primary tumor growth.
Inhibitors of signal transducers or public intracellular signals seem to be as ineffective as unspecific. First, downstream molecules are active in normal tissues and are necessary to normal physiology. Second, specific biological cues are transduced by many downstream effectors. For example, Met recruits a wide spectrum of downstream signaling molecules including PI3K, the GRB2/SOS complex, Src, the transcription factor STAT3, and the adaptors Shc and Gab-1 that provide additional docking sites for many signaling molecules. Conversely, an effector can be downstream of many pathways (for instance, Src is activated by RTKs, G-protein-associated receptors, and integrins). The result is that the inhibition of downstream effectors does not necessarily result in the complete pathway arrest, but nevertheless can disturb vital biological cues. Clinical application of effectors’ inhibitors has to monitor toxicity, assuming that the therapy is suitable. Moreover, there is evidence supporting the idea that effective inhibition should involve multiple inhibitors. After many years of frustrating investigation, Src has recently become a target for drug therapy. Because it is commonly activated, but never mutated, in a large number of human cancers and because its mechanisms of activation are now better understood, drug development programs have been able to inhibit Src activity (99)
. Numerous inhibitors of Src kinase activity are entering early Phase I trials or are in preclinical trials, including compounds developed by Wyeth (SKI-606TM) and Sugen (SU6656TM) and by Ariad Pharmaceuticals (AP23464TM). Its effect on human cancer cells seems to be growth inhibitory. However, the fact that the noncatalytic domains of Src might cause integrin assembly indicates that Src inhibition, by targeting the kinase domain, could only be partially effective (100)
. Moreover, recent data indicate that simultaneous targeting of Src and FAK is even more effective in promoting apoptosis of cancer cells, and raises the potential for combined therapy strategies (101)
.
Our feeling is that receptor-based signal inhibition appears the most promising approach to develop cancer therapeutics. First, key regulators of invasive growth have been identified. Second, blocking receptor-based signals results in the inhibition of the entire downstream pathway. Third, invasive signals are specific by definition, in contrast to the more versatile downstream effectors. Accordingly, during much effort has been efficiently devoted to target receptor-based signals for cancer therapy, and several small molecules, antibodies and recombinant proteins have been developed and tested in preclinical and clinical studies (2
, 92
, 102)
.
|
MET IS AN IDEAL PHARMACOLOGICAL TARGET FOR CANCER PROGRESSION
|
|---|
As mentioned, clinical data have indicated that tumor RTK-targeted therapies (e.g., Herceptin, Gleevec, Iressa, Tarceva) have an effective response in patients that exhibit genetic alteration (mutation or amplification) of their targets. Conversely, invasive signaling such as that of HGF/Met seems to be an ideal target for cancer therapy even in the absence of genetic alteration (Fig. 3
). Indeed, the role of Met in EMT and cancer progression, its relevance in angiogenesis and in the crosstalk between tumor cells and microenviroment, the broad expression of both the ligand and receptor, the numerous mechanisms by which this target can be activated, also independently from HGF (103
104
105
106
, 60
, 61)
, and lack of redundancy within the subfamily support this idea. However, mutations or gene amplification of Met in selected clinical populations suggest that certain patients may be exquisitely sensitive to targeted therapy (107)
. In recent years multiple approaches have been attempted to inhibit Met (31
, 108)
. All the molecules tested can be divided into three classes: 1) kinase inhibitors, 2) HGF inhibitors, and 3) receptor competitors. Small molecules that prevent ATP binding to Met, such as SU11274TM, K52a, PHA-665752 TM (109
110
111)
, are more suitable from a pharmacological point of view, but their specificity represents a major problem. Moreover, they inhibit kinase-dependent Met activation only. Neutralizing antibodies anti-HGF (112)
, NK4 HGF (113)
, and an uncleavable form of HGF (114)
belong to the category of HGF inhibitors. Their limit is that they block HGF-dependent Met activation only. Recently it emerged that the best way to block HGF/Met-induced invasive program is the competition with Met receptor itself (27
, 115
, 116)
. Accordingly, RNA interference reveals that ligand-independent Met activity is required for tumor cell signaling and survival (117)
. Two works show that not only the full-sized extracellular portion of Met, but also the sole SEMA domain from the extracellular portion are sufficient to block receptor dimerization and activation (115
, 116)
. In our lab we engineered a soluble form of Met, Decoy Met, demonstrating that it inhibits Met activation mediated by both HGF and HGF-independent mechanisms, such as overexpression or oncogenic point mutation (116
, 103
104
105
106
, 60
, 61)
. In mice, Decoy Met impairs both tumor cell expansion and tumor angiogenesis, leading to tumor growth inhibition and metastasis prevention. It is noteworthy that the targeted tumor models do not require Met gene mutation or overexpression. A recent publication reinforces our data demonstrating that a monoclonal antibody (mAb), DN30, against the extracellular portion of Met induces Met shedding and hampers biological activity both in vitro and in vivo. The shed Met extracellular portion is responsible of preventing Met activation, acting as a decoy (118)
.
View larger version (32K):
[in this window]
[in a new window]
|
Figure 3. Met is an ideal pharmacological target for cancer therapy. Met, as a master regulator of invasive growth, elicits different biological activities including cell proliferation, survival, cell-cell dissociation (scattering), migration, invasion, and morphogenesis. Met pathway has been well studied and resumes the possibility to be blocked at each therapeutic level. Indeed, it is specifically activated by its soluble ligand, hepatocyte growth factor (HGF), even if several transmembrane molecules interact with it (receptors, like Ron and EGF receptor, B plexins, adhesive receptors like integrins, and CD44). Met docking site is responsible for the recruitment of a wide spectrum of downstream signal transducers including phosphatidylinositol 3-kinase (PI3K), the GRB2/SOS complex, the nonreceptor tyrosine kinase Src, the adapters Shc and Gab-1, and the transcriptional factor STAT3. Moreover, an environmental condition like hypoxia induces HIF-mediated Met gene transcription. At the same time, Met activation modifies tumor microenvironment since HGF is a potent angiogenic factor. A link between Met oncogene activation, cell transformation, and hemostasis was shown recently. Oncogenic activation of Met creates an extracellular environment that is fertile for tumor cell expansion and invasion by inducing the formation of a temporary scaffold of fibrin. Much effort has been undertaken to study inhibitors that block Met pathway in each point of this cascade (31
, 108)
.
|
|
If tumor progression arises from the activation of the same biological program that maintains normal organ complexity and architecture, will therapies directed against these pathways have toxic side effects? In our experiments mice expressing high plasmatic concentrations of Decoy Met display very few side effects, if any. Consistent with a role of the cytokine in hemopoiesis (119)
, HGF and its antagonist, Decoy Met, respectively decreased and increased apoptotic index in the bone marrow compartment, where tissue homeostasis and self-renewal are intensively active in adulthood. The differential effect of HGF/Met inhibition on tumor vs. normal tissues can easily be explained by—and is a reasonable proof of—the hypothesis that tumors are nothing but "miniorgans" attempting to organize themselves and expand (120)
. With regard to this concept, hgf or met knockout mice die in utero due to liver failure and placental defects resulting from impaired trophoblast invasion of uterine matrix and abnormal vascularization of corial villi (21
, 22
, 121)
. Conversely, hepato-specific conditional knockout mice show that Met is not required for liver physiology after birth (122
, 123)
. Thus, HGF-mediated invasive cues are essential during embryonic organ formation, but may be dispensable under normal circumstances in the adult unless a tumor "smartly" resumes a discontinued genetic program to establish itself and its progeny into the host organism. The observed effect on bone marrow apoptotic index—significant but compensable—suggests that the extent to which a normal cell population relies on particular signaling pathways depends on the mitotic activity of the cells, the amount of regenerative activity in the tissue and the stage of development, among other things. Therefore, tumors seem likely to be more dependent on some pathways than normal tissues, even if the pathways are active in both.
|
FUTURE PERSPECTIVE
|
|---|
Met is an exciting novel drug target due to the success observed in preclinical studies, its demonstrated role in experimental oncogenesis, and its deregulation and correlation with disease prognosis in numerous cancers. No wonder that in the last few years the interest of pharmaceutical companies in HGF/Met inhibitors has boosted dramatically. Upcoming challenges include the optimization of new and more specific drugs, the first clinical evaluation of a Met-targeted therapy, and the identification and selection of patient populations that will benefit mostly from treatment. Moreover, since HGF/Met signaling has been implicated in chemo- and radio-resistance of tumors (124)
, the combination of HGF/Met inhibitors with the current cancer drugs could result in a more effective but less toxic therapy.
From a clinical point of view, studies of the molecular mechanisms that drive invasive growth and tumor cell metastatization should provide a valuable direction in the development of targeted therapies. Met, as master switch of cancer progression, is surely an optimal pharmacological target, but in the future, enucleating all the steps of the neoplasia will enable the study of "magic bullets" that block cancerous tissue efficiently and specifically without affecting ideally normal tissue functions.
|
ACKNOWLEDGMENTS
|
|---|
Studies in the authors’ laboratory were supported in part by the Italian Association for Cancer Research (AIRC), the Giovanni Armenise-Harvard Foundation for Advanced Scientific Research (Harvard), the Compagnia San Paolo di Torino Foundation, and the Italian Ministry of University and Research (MIUR COFIN #2001068458). We thank Antonella Cignetto for secretarial assistance and Radhika Srinivasan, Miriam Martini and Daniela Dettori for help with the manuscript. We are grateful to Livio Trusolino, Paolo Michieli, Selma Pennacchietti, and Federica Di Nicolantonio for the helpful discussion.
Received for publication February 20, 2006.
Accepted for publication March 31, 2006.
|
REFERENCES
|
|---|
- Chambers, A. F., Groom, A. C., MacDonald, I. C. (2002) Dissemination and growth of cancer cells in metastatic sites. Nat. Rev. Cancer 2,563-572[CrossRef][Medline]
- Vogelstein, B., Krinzler, K. W. (2004) Cancer genes and pathways they control. Nat. Med. 10,789-799[CrossRef][Medline]
- Hanahan, D., Weinberg, R. A. (2000) The hallmarks of cancer. Cell 100,57-70[CrossRef][Medline]
- Steeg, P. S. (2003) Metastasis suppressors alter the signal transduction of cancer cells. Nat. Rev. Cancer 3,55-63[CrossRef][Medline]
- Trusolino, L., Comoglio, P. M. (2002) Scatter-factor and semaphorin receptors: cell signalling for invasive growth. Nat. Rev. Cancer 2,289-300[CrossRef][Medline]
- Comoglio, P. M., Boccaccio, C. (2001) Scatter factor and invasive growth. Semin. Cancer Biol. 11,153-165[CrossRef][Medline]
- Nieto, M. A. (2001) The early steps of neural crest development. Mech. Dev. 105,27-35[CrossRef][Medline]
- Shook, D., Keller, R. (2003) Mechanisms, mechanics and function or epithelial-mesenchymal transitions in early development. Mech. Dev. 120,1351-1383[CrossRef][Medline]
- Thiery, J. P. (2002) Epithelial mesenchymal transitions in tumour progression. Nat. Rev. Cancer 2,442-454[CrossRef][Medline]
- Kang, Y., Massague, J. (2004) Epithelial-mesenchymal transitions: twist in development and metastasis. Cell 118,277-279[CrossRef][Medline]
- Yang, J., Mani, S. A., Donaher, J. L., Ramaswamy, S., Itzykson, R. A., Come, C., Savagner, P., Gitelman, I., Richardson, A., Weinberg, R. A. (2004) Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell 117,927-939[CrossRef][Medline]
- Brabletz, T., Jung, A., Spaderna, S., Hlubek, F., Kirchner, T. (2005) Opinion: migrating cancer stem cells—an integrated concept of malignant tumour progression. Nat. Rev. Cancer 5,744-749[CrossRef][Medline]
- Brabletz, T., Jung, A., Reu, S., Porzner, M., Hlubek, F., Kunz-Schughart, L.A., Knuechel, R., Kirchner, T. (2001) Variable beta-catenin expression in colorectal cancers indicates tumor progression driven by the tumor environment. Proc. Natl. Acad. Sci. U. S. A. 98,10356-10361[Abstract/Free Full Text]
- Barker, N., Clevers, H. (2001) Tumor environment: a potent driving force in colorectal cancer?. Trends Mol. Med. 7,535-537[CrossRef][Medline]
- Nakamura, T., Nishizawa, T., Hagiya, M., Seki, T., Shimonishi, M., Sugimura, A., Tashiro, K., Shimizu, S. (1989) Molecular cloning and expression of human hepatocyte growth factor. Nature 342,440-443[CrossRef][Medline]
- Skeel, A., Yoshimura, T., Showalter, S. D., Skeel, A., Yoshimura, T., Showalter, S. D. (1991) Macrophage stimulation protein: purification, partial amino acid sequence, and cellular activity. J. Exp. Med. 173,1227-1234[Abstract/Free Full Text]
- Stoker, M., Gherardi, E., Perryman, M., Gray, J. (1987) Scatter factor is a fibroblast-derived modulator of epithelial cell mobility. Nature 327,239-242[CrossRef][Medline]
- Naldini, L., Weidner, K. M., Vigna, E., Gaudino, G., Bardelli, A., Ponzetto, C., Narsimhan, R. P., Hartmann, G., Zarnegar, R., Michalopoulos, G.K., Birchmeier, W., Comoglio, P. M. (1991) Scatter factor and hepatocyte growth factor are indistinguishable ligands for the MET receptor. EMBO J. 10,2867-2878[Medline]
- Giordano, S., Ponzetto, C., Di Renzo, M. F., Cooper, C. S., Comoglio, P. M. (1989) Tyrosine kinase receptor indistinguishable from the c-met protein. Nature 339,155-156[CrossRef][Medline]
- Birchmeier, C., Gherardi, E. (1998) Developmental roles of HGF/SF and its receptor, the c-Met tyrosine kinase. Trends Cell Biol. 8,404-410[CrossRef][Medline]
- Uehara, Y., Minowa, O., Mori, C., Shiota, K., Kuno, J., Noda, T., Kitamura, N. (1995) Placenta defect and embryonic lethality in mice lacking hepatocyte growth factor/scatter factor. Nature 373,702-705[CrossRef][Medline]
- Bladt, F., Riethmacher, D., Isenmann, S., Aguzzi, A., Birchmeier, C. (1995) Essential role for the c-met receptor in the migration of myogenic precursor cells into the limb bud. Nature 376,768-771[CrossRef][Medline]
- Yanagida, H., Kaibori, M., Hijikawa, T., Kwon, A. H., Kamiyama, Y., Okumura, T. (2006) Administration of rhHGF-activator via portal vein stimulates the regeneration of cirrhotic liver after partial hepatectomy in rats. J. Surg. Res. 130,38-44[CrossRef][Medline]
- Nakamura, T., Mizuno, S., Matsumoto, K., Sawa, Y., Matsuda, H., Nakamura, T. (2000) Myocardial protection from ischemia/reperfusion injury by endogenous and exogenous HGF. J. Clin. Invest. 106,1511-1519[Medline]
- Huh, C. G., Factor, V. M., Sanchez, A., Uchida, K., Conner, E. A., Thorgeirsson, S. S. (2004) Hepatocyte growth factor/c-met signaling pathway is required for efficient liver regeneration and repair. Proc. Natl. Acad. Sci. U. S. A. 101,4477-4482[Abstract/Free Full Text]
- Rong, S., Segal, S., Anver, M., Resau, J. H., Vande Woude, G. F. (1994) Invasiveness and metastasis of NIH 3T3 cells induced by Met-hepatocyte growth factor/scatter factor autocrine stimulation. Proc. Natl. Acad. Sci. U. S. A. 91,4731-4735[Abstract/Free Full Text]
- Michieli, P., Basilico, C., Pennacchietti, S., Maffe, A., Tamagnone, L., Giordano, S., Bardelli, A., Comoglio, P. M. (1999) Mutant Met-mediated transformation is ligand-dependent and can be inhibited by HGF antagonists. Oncogene 18,5221-5231[CrossRef][Medline]
- Chiara, F., Michieli, P., Pugliese, L., Comoglio, P. M. (2003) Mutations in the met oncogene unveil a "dual switch" mechanism controlling tyrosine kinase activity. J. Biol. Chem. 278,29352-29358[Abstract/Free Full Text]
- Boccaccio, C., Sabatino, G., Medico, E., Follenzi, A., Reato, G., Sottile, A., Naldini, L., Comoglio, P. M. (2005) The MET oncogene drivers a genetic programme linking cancer to haemostasis. Nature 434,396-400[CrossRef][Medline]
- Boccaccio, C., Comoglio, P. M. (2005) A functional role for hemostasis in early cancer development. Cancer Res. Rev. 65,8579-8582[CrossRef]
- Corso, S., Comoglio, P. M., Giordano, S. (2005) Cancer therapy: can the challenge be MET?. Trends Mol. Med. 11,284-292[CrossRef][Medline]
- Takayama, H., LaRochelle, W. J., Sharp, R., Otsuka, T., Kriebel, P., Anver, M., Aaronson, S. A., Merlino, G. (1997) Diverse tumorigenesis associated with aberrant development mice overexpressing hepatocyte growth factor/scatter factor. Proc. Natl. Acad. Sci. U. S. A. 94,701-706[Abstract/Free Full Text]
- Di Renzo, M. F., Olivero, M., Katsaros, D., Crepaldi, T., Gaglia, P., Zola, P., Sismondi, P., Comoglio, P. M. (1994) Overexpression of the Met/HGF receptor in ovarian cancer. Int. J. Cancer 58,658-662[Medline]
- Di Renzo, M. F., Olivero, M., Giacomini, A., Porte, H., Chastre, E., Mirossay, L., Nordlinger, B., Bretti, S., Bottardi, S., Giordano, S., Comoglio, P. M. (1995) Overexpression and amplification of the met/HGF receptor gene during the progression of colorectal cancer. Clin. Cancer Res. 1,147-154[Abstract]
- Morello, S., Olivero, M., Aimetti, M., Bernardi, M., Berrone, S., Di Renzo, M. F., Giordano, S. (2001) MET receptor is overexpressed but not mutated in oral squamous cell carcinomas. J. Cell. Physiol. 189,285-290[CrossRef][Medline]
- Schmidt, L., Junker, K., Nakaigawa, N., Kinjerski, T., Weirich, G., Miller, M., Lubensky, I., Neumann, H. P., Brauch, H., Decker, J., Vocke, C., Brown, J. A., Jenkins, R., Richard, S., Bergerheim, U., Gerrard, B., Dean, M., Linehan, W. M., Zbar, B. (1999) Novel mutations of the MET proto-oncogene in papillary renal carcinomas. Ocogene 18,2343-2350
- Jagadeeswaran, R., Ma, P. C., Seiwert, T. Y., Jagadeeswaran, S., Zumba, O., Nallasura, V., Ahmed, S., Filiberti, R., Paganuzzi, M., Puntoni, R., Kratzke, R. A., Gordon, G. J., Sugarbaker, D. J., Bueno, R., Janamanchi, V., Bindokas, V. P., Kindler, H. L., Salgia, R. (2006) Functional analysis of c-Met/hepatocyte growth factor pathway in malignant pleural mesothelioma. Cancer Res. 66,352-361[Abstract/Free Full Text]
- Kong-Beltran, M., Seshagiri, S., Zha, J., Zhu, W., Bhawe, K., Mendoza, N., Holcomb, T., Pujara, K., Stinson, J., Fu, L., Severin, C., Rangell, L., Schwall, R., Amler, L., Wickramasinghe, D., Yauch, R. (2006) Somatic mutations lead to an oncogenic deletion of met in lung cancer. Cancer Res. 66,283-289[Abstract/Free Full Text]
- Di Renzo, M. F., Olivero, M., Martone, T., Maffe, A., Maggiora, P., Stefani, A. D., Valente, G., Giordano, S., Cortesina, G., Comoglio, P. M. (2000) Somatic mutations of the MET oncogene are selected during metastatic spread of human HNSC carcinomas. Oncogene 19,1547-1555[CrossRef][Medline]
- Lorenzato, A., Olivero, M., Patane, S., Rosso, E., Oliaro, A., Comoglio, P. M., Di Renzo, M. F. (2002) Novel somatic mutations of the MET oncogene in human carcinoma metastases activating cell motility and invasion. Cancer Res. 62,7025-7030[Abstract/Free Full Text]
- Cortesina, G., Martone, T., Galeazzi, E., Olivero, M., De Stefani, A., Bussi, M., Valente, G., Comoglio, P. M., Di Renzo, M. F. (2000) Staging of head and neck squamous cell carcinoma using the MET oncogene product as marker of tumor cells in lymph node metastases. Int. J. Cancer 89,286-292[CrossRef][Medline]
- Reya, T., Duncan, A. W., Ailles, L., Domen, J., Scherer, D. C., Willert, K., Hintz, L., Nusse, R., Weissman, I. L. (2003) A role for Wnt signalling in self-renewal of haematopoietic stem cells. Nature 423,409-414[CrossRef][Medline]
- Polakis, P. (1999) The oncogenic activation of the beta-catenin. Curr. Opin. Genet. Dev. 9,15-21[CrossRef][Medline]
- Kirchner, T., Brabletz, T. (2000) Pattering and nuclear beta-catenin expression in the coloni adenoma-carcinoma sequence. Analogies with embryonic gastrulation. Am. J. Pathol. 157,1113-1121[Abstract/Free Full Text]
- Boon, E. M., van der Neut, R., van de Wetering, M., Clevers, H., Pals, S. T. (2002) Wnt signaling regulates expression of the receptor tyrosine kinase met in colorectal cancer. Cancer Res. 62,5126-5128[Abstract/Free Full Text]
- Monga, S. P., Mars, W. M., Pediaditakis, P., Bell, A., Mule, K., Bowen, W. C., Wang, X., Zarnegar, R., Michalopoulos, G. K. (2002) Hepatocyte growth factor induces Wnt-independent nuclear translocation of beta-catenin after Met-beta-catenin dissociation in hepatocytes. Cancer Res. 62,2064-2071[Abstract/Free Full Text]
- Stella, M. C., Trusolino, L., Pennacchietti, S., Comoglio, P. M. (2005) Negative feedback regulation of Met-dependent invasive growth by Notch. Mol. Cell. Biol. 25,3982-3996[Abstract/Free Full Text]
- Muller, A., Homey, B., Soto, H., Ge, N., Catron, D., Buchanan, M. E., McClanahan, T., Murphy, E., Yuan, W., Wagner, S. N., Barrera, J. L., Mohar, A., Verastegui, E., Zlotnik, A. (2001) Involvement of chemokine receptors in breast cancer metastasis. Nature 410,50-56[CrossRef][Medline]
- Tanaka, T., Bai, Z., Srinoulprasert, Y., Yang, B. G., Hayasaka, H., Miyasaka, M. (2005) Chemokines in tumor progression and metastasis. Cancer Sci. 96,317-322[CrossRef]
TOP
ABSTRACT
INTRODUCTION
