Growth factors and their receptors in cell differentiation, Cancer and cancer therapy. 2011 Elsevier
Hepatocyte growth factor (HGF) is among a group of factors possessing angiogenic ability, which have often been described as heparin-binding growth factors. HGF, together with VEGF and FGFs and TGF-Я, bind to heparan sulphate proteoglycans, which affords a means of localising them on the cell surface and presenting them to their appropriate receptors in the most favourable conformation in order to facilitate the interaction of the growth factors with the receptors.
Molecular structure of HGF
HGF, initially known as the scatter factor or cytotoxic factor (Gherardi et al., 1989; Higashio et al., 1990; Rubin et al., 1991), is secreted by fibroblasts and is mitogenic for epithelial and endothelial cells as well as melanocytes, but does not affect fibroblasts (Rubin et al., 1991). It contains 728 amino acid residues. It is secreted as an inactive precursor which is processed into the active form. The active form of HGF is a heterodimer of disulphide-linked α and β chains (Nakamura et al., 1989; Hartmann et al., 1992; Kataoka et al., 2003). The a-chain is folded at the amino (N)-terminal domain, which is followed by four Kringle domains. Carboxy (C)-terminal fragments of the a-chain known as NK1, NK2 and NK4, generated by proteolytic cleavage, also bind the HGF receptor MET competitively but cannot activate the receptor and so inhibit signalling (Chan et al., 1991; Cioce et al., 1996; Date et al., 1997). As discussed below, NK4 has been deployed with some success to restrain tumour invasion and growth. The Я-chain begins with valine 495; it is also proteolytically processed (Perona and Craik, 1995; Hedstrom, 2002).
The MET gene encodes the HGF tyrosine kinase receptor. The mature receptor is formed by disulphide linkage between the extracellular a-subunit and the transmembrane β-subunit derived by post-translational proteolytic cleavage of a single-chain precursor molecule. Four domains have been identified in the extracellular region: the Sema domain (semaphorin domain), which encompasses the a-subunit and the N-terminal part of Я-subunit and comprises the PSI (plexin semophorin integrin) domain (cysteine-rich MET-related sequences), and three glycine–proline-rich repeats and four IPT (Ig) domains (Birchmeier et al., 2003; Trusolino and Comoglio, 2002). The Sema domain is required for ligand binding and receptor dimerisation (Kong-Beltran et al., 2004). The PSI domain is said to form a wedge between the ligand-binding Sema and the Ig domains and is believed to account for the correct positioning of the ligand-binding site of the receptor (Kozlov et al., 2004).
The MET receptor is activated by dimerisation upon ligand binding (Prat et al., 1998). Ligand binding activates the kinase activity of the receptor, leading to the phosphorylation of residues Tyr 1234 and 1235. Several signal transducers are recruited to the tyrosines. Transducers such as GRB2, SHC, SRC and p85 of PI3K directly interact with the docking sites of MET or through a scaffolding protein (Pelicci et al., 1995; Maina et al., 1996; Weidner et al., 1996). HGF/MET activates several signalling pathways, namely Ras, PI3K, MAPK/STAT and Я-catenin/Wnt, leading to the biological effects of cell scattering, proliferation, resistance to apoptosis, invasion and angiogenesis. These pathways are focused on here in the context of cancer cell invasion, angiogenesis and fo
Figure 5.1. The signalling pathways interacting with HGF. HGF activates MET and the MAPK/ERK pathway to induce cell proliferation. EGF can upregulate MET and in this way EGF ? HGF can combine efforts towards induction of cell proliferation. HGF can enhance cell population by inducing resistance to apoptosis and induce invasion through activation of PAX/ Sox transcription factors. HGF/MET activation of the MAPK/ERK pathway can also lead to the downregulation of TSP, which in turn again seems to induce invasive behaviour. As shown in Figure 5.2, the downregulation of TSP is also conducive to angiogenesis. So Figure 5.1 should be read in conjunction with Figure 5.2 below. The related references are cited in the text.
Figure 5.2. HGF interaction with other signalling systems to integrate the phenotypic effects of cell proliferation, invasion and angiogenesis, which form the essential ingredients of cancer cell dissemination and metastasis. References are provided and discussed in the text.
Rho et al. (2009) established PC-9 sublines resistant to EGFR inhibitors which displayed MET activation. They also found the T790M mutation in the EGFR gene, which has been linked with drug resistance compared with the wild-type gene. However, this mutation is also known to occur in disease progression quite independently of the receptor tyrosine kinase (RTK) inhibitors gefitinib or erlotinib (Kosaka et al., 2004). Furthermore, NSCLCs carrying activating mutations within the EGFR kinase domain may be more susceptible to the RTK inhibitor gefitinib. Mutant EGFRs have been found to activate Akt/STAT signalling and induce cell proliferation. In these cells, inhibition of mutant EGFR reversed the effect and induced apoptosis (Sordella et al., 2004). So the significance of the perceived link between MET and EGFR requires further investigation.
Another interacting factor is TSP (thrombospondin). HGF seems to be able to induce invasive behaviour in certain ovarian carcinoma cells through downregulation of TSP. In this process HGF seems to signal through the MAPK pathway. When this route is blocked, HGF-mediated increase of invasive behaviour is also blocked (Wei et al., 2010). TSP-1 is an inhibitor of angiogenesis, so its downregulation by HGF is compatible with the pro-angiogenic effects of HGF.
As noted in the preceding pages, VEGF165 and HGF activate the same signalling pathways and function synergistically. However, they do not activate each other’s receptors; this suggests differences in signalling downstream of the activation of their respective receptors (Sulpice et al., 2009). HGF is not only angiogenic, it also possesses the ability to promote cell proliferation, resist apoptosis and induce cell motility or invasion. In other words, there could be mechanisms that switch function specifically to target invasion. It was suggested some time ago that activation of PI3K/Akt signalling by angiogenic factors can lead to VEGF regulation and in turn to angiogenesis. The collusion between VEGF and HGF could augment cell proliferation, angiogenesis and invasion, with significant effects on cancer cell dissemination (Figures 5.1 and 5.2).
It should be noted here that a putative link between MET, VEGF165 and NRP1 (neuropilin-1), a VEGF co-receptor, has emerged recently. VEGF165 has been found to promote interaction between MET and NRP1 and promote MET phosphorylation in prostate cancer cells. Activation of MET seems to occur in parallel with the upregulation by VEGF of the Bcl-2 family anti-apoptotic Mcl-1, possibly through activation of Src kinase and STAT3 (Zhang et al., 2010). Angiogenesis is regulated by NO (nitric oxide), iNOS (inducible NO synthase) and COX-2. NO is known to enhance and NOS antagonists to inhibit VEGF synthesis. NO also upregulates COX-2, so a strong link has been established between NO, VEGF and angiogenesis. In this context it might be noted that HGF induces COX-2 through MET activation (Scarpino et al., 2009).
HGF is known to activate the zinc finger transcription factor egr-1 (early growth response-1) and induce the transcription of VEGF and PDGFA. In fact, egr-1 binding sites have been detected in the promoter regions of these genes. Here HGF seemed to activate MEK1/2 and PKC pathways to activate egr-1 (Worden et al., 2005). HGF upregulates both egr-1 and the transcription repressor Snail. Snail is said to repress the expression of E-cadherin and the tight junction protein claudin-3 of epithelial cells, which aids the processes of cell scattering and migration (Grotegut et al., 2006).
The transcription factors PAX (paired box) and Sox are said directly to interact with MET, induce cell proliferation and resistance to apoptosis, and increase cell migration. PAX6(5a), a variant form of PAX6, is expressed in pancreatic carcinoma cell lines at higher levels than PAX6 protein. These proteins bind to an enhancer element of MET promoter. Inhibition of PAX leads to an inhibition of their biological effects (Mascarenhas et al., 2009, 2010). HGF is known to activate PAX in the induction of cell proliferation, providing a clear activation signal to PAX/MET, leading to proliferation and invasion. General indications are that this could occur independently of ERK1/2 (Yablonka-Reuveni et al., 2008). Plexin B1 seems to block the activation of MET in melanoma and possibly inhibit invasion in this way (Stevens et al., 2010). This could be why Plexin is regarded as a metastasis suppressor in melanomas.
MET as a therapeutic target
MET is expressed in many forms of cancer. The gene is amplified or the protein is overexpressed with marked relationship to tumour grade, lymph node involvement, invasion and metastasis. The angiogenic ability of HGF was reported many years ago and expression levels correlated with breast tumour growth with attendant elevation of MET to correspond with tumour progression (Nagy et al., 1995, 1996). The involvement of HGF in angiogenesis was strengthened by the finding that MET expression was upregulated by bFGF and TGF-Я, which are powerful inducers of angiogenesis (Hiscox and Jiang, 1996). Subsequent to these early works, much evidence has emerged about the role of HGF as an active participant in differentiation and morphogenesis. HGF is a potent mitogen and an inhibitor of apoptosis, a promoter of cell invasion and motility, and an inducer of angio/lymphangiogenesis. With these attributes much effort has been expended on the role of HGF/MET in cancer invasion, metastasis and in predicting prognosis. Consistent with research and development in other areas of growth factors, MET has been viewed as a potential anticancer therapeutic target. Several types of MET inhibitor have been designed, namely monoclonal antibodies against MET or its ligand that prevent MET activation, and inhibitors of MET activation. Small interference RNAs have also been considered as potential inhibitors.
Among monoclonals is the humanized mAb IgG2, which is directed against HGF Rilotumumab (Amgen Inc.). Rilotumumab prevents interaction between HGF and MET, and inhibits MET phosphorylation and signalling. Phase I clinical trials are currently underway (see Giordano, 2009). Rilotumumab is said to be well tolerated by patients with solid tumours. Mild side effects have been encountered. Most patients showed favourable response to treatment and disease-free progress from 7 to 40 weeks (Gordon et al., 2010).
Small-molecule MET kinase inhibitors such as SU11274, K252a, PHA665752 and PF2341066 have been designed. Davis et al. (2010) tested SU11274 and Rilotumumab on clear-cell sarcoma cell lines. Both inhibited cell growth in culture. The antibody effectively suppressed xenografted cells. Kenessey et al. (2010) also reported that SU11274 blocked MET-mediated signalling in melanomas, decreased cell proliferation and increased apoptosis. In vitro, SU11274 at subapoptosis concentrations inhibited cell migration. Furthermore, when xenografted into SCID mice, the tumours showed reduced growth. Qian et al. (2009) have described another kinase inhibitor, EXEL-2880 (XL880, GSK1363089), which targets a range of growth factor receptors, notably MET and VEGFR. They reported loss of anchorage-dependent proliferation in vitro and in vivo; there was reduced tumour growth and inhibition of lung metastasis in experimental assays. The latter is probably a reflection of alterations in adhesive faculty of the tumour cells or changes in endothelial permeability.
Inhibition of angiogenesis by targeting HGF and MET has provided an attractive route to therapy of cancer progression. One approach has been to use NK4, the N-terminal hairpin domain and four kringle domains of HGF. NK4 competitively binds to MET, but it cannot activate the receptor and inhibits HGF/MET signalling. NK4 is said to function as an inhibitor of angiogenesis (Nakabayashi et al., 2003; Nakamura et al., 2010). Nakabayashi (2003) had shown earlier that NK4 inhibited VEGF-induced endothelial cell proliferation and migration, and effectively blocked angiogenesis in vivo. Besides, NK4 has been attributed with the ability to suppress HGF-induced cell proliferation and motility and to promote apoptosis (Yue et al., 2010).
Serine proteinases have also been deployed as inhibitors. These would prevent mature HGF being formed from its inactive precursor and in this way effectively inhibit HGF signalling (Parr et al., 2010). However, the NK4 fragment is also generated by proteolytic cleavage of HGF by serine and other peptidases as well as cathepsins, so the serine proteinase inhibitors might function in an antagonistic fashion in NK4-mediated suppression of biological effects.