**5. Drug discovery and development**

Close involvement of aberrant HGF-Met signaling in tumorigenesis and progression to malignant disease has facilitated drug discovery and development. Several distinct lines of approach to the inhibition of the HGF-Met pathway have been demonstrated, including small synthetic inhibitors of Met tyrosine kinase, ribozymes, small-interfering RNA (siRNA), neutralizing monoclonal antibodies (mAbs), soluble forms of Met, antagonists composed of selected domains in HGF, and uncleavable single-chain HGF (Fig. 5). Among recombinant protein-based inhibitors, conventionally called biologics in drug development, mAbs targeting HGF or Met have been in clinical development earlier than the other biological inhibitors, predominantly because of their availability due to established technologies for manufacturing of recombinant mAbs.

Fig. 5. Outline for different approaches to targeting HGF and Met.

### **5.1 Biologicals**

320 Advances in Cancer Therapy

surviving cancer more aggressive with more metastatic behavior (Casanovas et al., 2005). Pathological and clinical studies indicate that the presence of hypoxic regions within neoplastic lesions correlates with poor prognosis and an increased risk of the development of distant metastases (Höckel & Vaupel, 2001). Importantly, a hypoxic condition induced the transcriptional activation of the Met receptor gene and subsequent amplification of HGF-Met signaling, thereby increasing the invasiveness of cancer cells (Penancchietti et al., 2004). A connection between hypoxia and the Met receptor seems to explain why hypoxia often correlates with invasive and metastatic behavior. Angiogenesis inhibition retards tumor growth by oxygen deprivation, at least in part. However, hypoxia caused by the inhibition of angiogenesis enhances HGF-Met signaling, thereby promoting tumor invasion and metastasis. The involvement of the HGF-Met pathway in the aggressive characteristics in the hypoxic regions of cancers, which includes tumors treated with antiangiogenic drugs, is considerable.

Close involvement of aberrant HGF-Met signaling in tumorigenesis and progression to malignant disease has facilitated drug discovery and development. Several distinct lines of approach to the inhibition of the HGF-Met pathway have been demonstrated, including small synthetic inhibitors of Met tyrosine kinase, ribozymes, small-interfering RNA (siRNA), neutralizing monoclonal antibodies (mAbs), soluble forms of Met, antagonists composed of selected domains in HGF, and uncleavable single-chain HGF (Fig. 5). Among recombinant protein-based inhibitors, conventionally called biologics in drug development, mAbs targeting HGF or Met have been in clinical development earlier than the other biological inhibitors, predominantly because of their availability due to established

**5. Drug discovery and development** 

technologies for manufacturing of recombinant mAbs.

Fig. 5. Outline for different approaches to targeting HGF and Met.

Biological inhibitors against HGF-Met include the following: 1) selected domains in HGF (NK4 and engineered NK1); 2) engineered single-chain HGF forms that are resistant to proteolytic processing; 3) truncated soluble forms of the Met extracellular region; and, 4) humanized monoclonal antibodies (mAbs) against HGF or Met.

Among the -chain of HGF, NK2 (the N-terminal, 1st kringle, and 2nd kringle domains), an alternative splicing variant, was first shown to competitively antagonize the growth stimulation by HGF (Chan et al., 1991). However, NK2 was later shown to stimulate cell motility and enhance HGF-driven metastasis in a mouse model (Stahl et al., 1997; Yu & Merlino, 2002).

NK4 is the first identified HGF-Met inhibitor devoid of biological activity through its Met binding. NK4 is composed of the N terminal and 4 kringle domains (Date et al., 1997; Matsumoto et al., 1998; Matsumoto et al., 2008). NK4 inhibits biological responses triggered by activation of HGF-Met signaling, including the spreading and invasion of cancer cells (Fig. 6). It should be emphasized that NK4 inhibits angiogenesis in addition to its antagonistic action against HGF, and this angioinhibitory action of NK4 is independent of its antagonist action against HGF. NK4 inhibited proliferation, migration, and tube formation of vascular endothelial cells induced by basic fibroblast growth factor and VEGF as well as by HGF (Kuba et al., 2001; Sakai et al., 2009). NK4 binds to perlecan and inhibits the cell-associated assembly of fibronectin, and the impaired fibronectin assembly suppresses integrin-dependent endothelial cell proliferation and migration. Having two different biological activities through completely different mechanisms is unique to NK4. Combination therapy of NK4 with antiangiogenic drugs is expected.

Fig. 6. Inhibition of tumor invasion by NK4. Invasion of human gallbladder cancer cells through the Matrigel basement membrane was induced by co-culture with stromal fibroblasts, and this aggressive invasion was inhibited by NK4 (**A**). 3-D invasion of human pleural malignant mesothelioma in collagen gel was enhanced by HGF, and was inhibited by NK4 (**B**).

The therapeutic effect of NK4 has been demonstrated in a variety of cancer models (Matsumoto et al., 2008). The inhibition of tumor growth by NK4 treatment was observed in a variety of tumors, and this inhibitory effect was associated with a reduction in blood vessels in tumor tissues. NK4 treatment inhibited *in situ* Met tyrosine phosphorylation, and

Significance, Mechanisms, and Progress of Anticancer Drugs Targeting HGF-Met 323

Table 1. HGF-Met inhibitors in clinical development.

mAbs against Met with different characteristics have been developed (Martens et al., 2006; Jin et al., 2008; van der Horst et al., 2009; Pacchiana et al., 2010). Anti-Met mAb, MetMab (formerly OA5D5), is a monovalent mAb that blocked binding of HGF to the Met (Martens

this was associated with an inhibition of the spread and invasive growth of cancer cells (Matsumoto et al., 2008). NK4 treatment inhibited metastasis in different types of models, including breast, colon, gastric, lung, ovarian, and pancreatic carcinomas, and malignant melanoma. These results provided the base for proof-of-concept; inhibition of the HGF-Met pathway is a way to inhibit the invasive and metastatic growth of cancer. Moreover, the combination therapy of NK4 and gefitinib overcame HGF-induced gefitinib resistance of lung cancer in a mouse model (Wang et al., 2009).

NK1 was identified as a product of an alternative spliced variant of HGF, similar to NK2, that consists of the N-terminal and first kringle (K1) domains (Cioce et al., 1996). NK1 binds Met and acts as a partial agonist in cell-based assays and transgenic mice (Lokker et al., 1994; Schwall et al., 1996). Biochemical and structural analysis indicated the following two points: 1) NK1 is responsible for the high affinity binding of HGF to the Met-Sema domain; and, 2) Met dimerization may be mediated by the NK1-NK1 dimer interface (Watanabe et al., 2002; Gherardi et al., 2006). Based on structural analysis, mutations designed to alter the NK1 dimer interface (Y124A, K85A, K85A/D123A, and K85A/N127A) abolish its ability to promote Met dimerization, but these mutated NK1s retain Met-binding activity (Rubin et al., 2001; Tolbert et al., 2007). These NK1 mutants act as Met antagonists by inhibiting HGF-mediated cell scattering, proliferation, and invasion (Gherardi et al., 2006; Rubin et al., 2001; Tolbert et al., 2007). Although it is yet to be determined if NK1 acts as an angiogenesis inhibitor, NK1 can be expected to exert anti-cancer action by inhibiting the HGF-Met pathway.

Single-chain HGF variants that are resistant to proteolytic activation exploit the requirement for processing machinery that converts pro-HGF to mature HGF. Indeed, uncleavable forms of HGF have been engineered by substituting single amino acids in the proteolytic site, and these engineered uncleavable HGFs suppress Met-driven tumor growth, metastasis, and angiogenesis in murine tumor models (Mazzone et al., 2004). Related antagonists consisting of two-chain HGF mutants exploit the mechanism by which proteolytic conversion allosterically stabilizes HGF–Met binding to promote kinase activation (Kirchhofer et al., 2007). Insertion of the newly formed N-terminus of the HGF chain into the activation pocket stabilizes the interaction between the HGF chain and Met. Full-length 2-chain HGF mutants engineered to interrupt these interactions efficiently inhibited HGF-mediated Met activation. Studies on the structure-function relationship of Met extracellular domains provided the development of Met-based biological HGF-Met inhibitor. A soluble Met-Sema domain is not only necessary for Met receptor association but is also essential for HGF binding, whereby the Sema domain inhibits HGF-dependent and -independent receptor phosphorylation and functional receptor activation (Kong-Beltran et al., 2004). In a mouse model, soluble Met Sema domain suppressed tumor growth and metastasis (Michieli et al., 2004).

Among different types of mAbs against HGF or Met, one anti-Met mAb decreases Met activation by inducing ectodomain shedding and degradation (Petrelli et al., 2006), while the others inhibit the binding of HGF to Met. Neutralizing mAb against human HGF, such as L2G7, AMG102 and SCH900105 (formerly AV299) inhibited HGF-dependent Met activation and the growth of tumor xenografts in mice (Petrelli et al., 2006; Kim et al., 2006; Jun et al., 2007). AMG102 is currently in phase I and II clinical trials (HYPERLINK "http://www.clinicaltrials.gov" www.clinicaltrials.gov) (Table 1). AMG 102 was well tolerated in humans and adverse events were predominantly low grade (Cecchi et al., 2011). SCH900105 was also well tolerated by patients in phase I trials. In its first completed trial, SCH900105 treatment was associated with a stabilizing of disease in half of the patients (Cecchi et al., 2011).

this was associated with an inhibition of the spread and invasive growth of cancer cells (Matsumoto et al., 2008). NK4 treatment inhibited metastasis in different types of models, including breast, colon, gastric, lung, ovarian, and pancreatic carcinomas, and malignant melanoma. These results provided the base for proof-of-concept; inhibition of the HGF-Met pathway is a way to inhibit the invasive and metastatic growth of cancer. Moreover, the combination therapy of NK4 and gefitinib overcame HGF-induced gefitinib resistance of

NK1 was identified as a product of an alternative spliced variant of HGF, similar to NK2, that consists of the N-terminal and first kringle (K1) domains (Cioce et al., 1996). NK1 binds Met and acts as a partial agonist in cell-based assays and transgenic mice (Lokker et al., 1994; Schwall et al., 1996). Biochemical and structural analysis indicated the following two points: 1) NK1 is responsible for the high affinity binding of HGF to the Met-Sema domain; and, 2) Met dimerization may be mediated by the NK1-NK1 dimer interface (Watanabe et al., 2002; Gherardi et al., 2006). Based on structural analysis, mutations designed to alter the NK1 dimer interface (Y124A, K85A, K85A/D123A, and K85A/N127A) abolish its ability to promote Met dimerization, but these mutated NK1s retain Met-binding activity (Rubin et al., 2001; Tolbert et al., 2007). These NK1 mutants act as Met antagonists by inhibiting HGF-mediated cell scattering, proliferation, and invasion (Gherardi et al., 2006; Rubin et al., 2001; Tolbert et al., 2007). Although it is yet to be determined if NK1 acts as an angiogenesis inhibitor, NK1 can be

Single-chain HGF variants that are resistant to proteolytic activation exploit the requirement for processing machinery that converts pro-HGF to mature HGF. Indeed, uncleavable forms of HGF have been engineered by substituting single amino acids in the proteolytic site, and these engineered uncleavable HGFs suppress Met-driven tumor growth, metastasis, and angiogenesis in murine tumor models (Mazzone et al., 2004). Related antagonists consisting of two-chain HGF mutants exploit the mechanism by which proteolytic conversion allosterically stabilizes HGF–Met binding to promote kinase activation (Kirchhofer et al., 2007). Insertion of the newly formed N-terminus of the HGF chain into the activation pocket stabilizes the interaction between the HGF chain and Met. Full-length 2-chain HGF mutants engineered to interrupt these interactions efficiently inhibited HGF-mediated Met activation. Studies on the structure-function relationship of Met extracellular domains provided the development of Met-based biological HGF-Met inhibitor. A soluble Met-Sema domain is not only necessary for Met receptor association but is also essential for HGF binding, whereby the Sema domain inhibits HGF-dependent and -independent receptor phosphorylation and functional receptor activation (Kong-Beltran et al., 2004). In a mouse model, soluble Met Sema domain suppressed

Among different types of mAbs against HGF or Met, one anti-Met mAb decreases Met activation by inducing ectodomain shedding and degradation (Petrelli et al., 2006), while the others inhibit the binding of HGF to Met. Neutralizing mAb against human HGF, such as L2G7, AMG102 and SCH900105 (formerly AV299) inhibited HGF-dependent Met activation and the growth of tumor xenografts in mice (Petrelli et al., 2006; Kim et al., 2006; Jun et al., 2007). AMG102 is currently in phase I and II clinical trials (HYPERLINK "http://www.clinicaltrials.gov" www.clinicaltrials.gov) (Table 1). AMG 102 was well tolerated in humans and adverse events were predominantly low grade (Cecchi et al., 2011). SCH900105 was also well tolerated by patients in phase I trials. In its first completed trial, SCH900105 treatment was associated with a stabilizing of disease in half of the patients

expected to exert anti-cancer action by inhibiting the HGF-Met pathway.

lung cancer in a mouse model (Wang et al., 2009).

tumor growth and metastasis (Michieli et al., 2004).

(Cecchi et al., 2011).


Table 1. HGF-Met inhibitors in clinical development.

mAbs against Met with different characteristics have been developed (Martens et al., 2006; Jin et al., 2008; van der Horst et al., 2009; Pacchiana et al., 2010). Anti-Met mAb, MetMab (formerly OA5D5), is a monovalent mAb that blocked binding of HGF to the Met (Martens

Significance, Mechanisms, and Progress of Anticancer Drugs Targeting HGF-Met 325

for the treatment of non-neoplastic diseases with tissue damage and for malignant diseases, respectively. HGF exhibits therapeutic effects for the protection and healing of tissues against tissue damage and pathology. Clinical trials using recombinant HGF or HGF gene

Based on the basic knowledge of the significance of the HGF-Met pathway in tumor biology and pathology, during the last several years the one-to-one relationship between HGF and Met has facilitated the discovery and development of drug candidates that selectively inhibit HGF-Met in different ways. Preclinical and clinical development of drugs targeting HGF-Met will move into practice in the near future as new anticancer drugs. However, although drug discoveries in molecular-targeted cancer therapy have been beneficial for patients with malignancies, the appearance of persistent characteristics of malignant tumors in regard to resistance to anticancer therapies and drugs remains an obstacle to disease-free survival. The choice of the better, or best, way to inhibit HGF-Met signaling, i.e., ligand inhibition, receptor inhibition, biologics, mAb, or small synthetic, would gradually become

The studies from the authors' laboratories were supported by Grants from the Ministry of Education, Culture, Science, Sports, and Technology of Japan, the Program for Promotion of Fundamental Studies in Health Sciences of the National Institute of Biomedical Innovation,

Bean, J., Brennan, C., Shih, J.Y., Riely, G., Viale, A., Wang, L., Chitale, D., Motoi, N., Szoke, J.,

Basilico, C., Arnesano, A., Galluzzo, M., Comoglio, P.M. & Michieli, P. (2008) A high affinity

Berthou, S., Aebersold, D.M., Schmidt, L.S., Stroka, D., Heigl, C., Streit, B., Stalder, D.,

Bladt, F., Riethmacher, D., Isenmann, S., Aguzzi, A. & Birchmeier, C. (1995) Essential role for

Borowiak, M., Garratt, A.N., Wustefeld, T., Strehle, M., Trautwein, C. & Birchmeier, C.

motility and more. *Nature Review Molecular Cell Biology*, 4: 915-925

Broderick, S., Balak, M., Chang, W.C., Yu, C.J., Gazdar, A., Pass, H., Rusch, V., Gerald, W., Huang, S.F., Yang, P.C., Miller, V., Ladanyi, M., Yang, C.H., & Pao, W. (2007) MET amplification occurs with or without T790M mutations in EGFR mutant lung tumors with acquired resistance to gefitinib or erlotinib. *Proceeding National* 

hepatocyte growth factor-binding site in the immunoglobulin-like region of Met.

Gruber, G., Liang, C., Howlett, A.R., Candinas, D., Greiner, R.H., Lipson, K.E. & Zimmer, Y. (2003) The Met kinase inhibitor SU11274 exhibits a selective inhibition pattern toward different receptor mutated variants. *Oncogene*, 23: 5387-5393, 2004. Birchmeier, C., Birchmeier, W., Gherardi, E. & Vande Woude, G.F. (2003) Met, metastasis,

the c-met receptor in the migration of myogenic precursor cells into the limb bud.

(2004) Met provides essential signals for liver regeneration. *Proceeding National* 

drugs have been approved for the treatment of diseases with unmet needs.

and the Hokkou Foundation for the Promotion of Cancer Research.

*Academy of Science USA*, 104: 20932-20937

*Academy of Science USA,* 101: 10608-10613

*Journal of Biological Chemistry*, 283: 21267-21277

clearer following clinical experiences.

*Nature,* 376: 768-771

**7. Acknowledgement** 

**8. References** 

et al., 2006). Anti-Met mAb R13 and R28 synergistically inhibited HGF binding to MET and elicited antibody-dependent cellular cytotoxicity (van der Horst et al., 2009). The combination of R13/28 inhibited tumor growth in various colon tumor xenograft models. MetMab reduced Met phosphorylation, and this was associated with inhibition of orthotopic tumor growth and improvement of survival in a pancreatic xenograft model (Jin et al., 2008). MetMab is currently in phase I/II human clinical trials in comparison with erlotinib for patients with NSCLC (HYPERLINK "http://www.clinicaltrials.gov" www.clinicaltrials.gov).

### **5.2 Small synthetic kinase inhibitors**

As small synthetic Met tyrosine kinase inhibitors, SU11274 and PHA665752 provided the basic notion that small synthetic Met tyrosine kinase inhibitors selectively inhibit Met activation and suppress tumor growth (Christensen ; 2003; Sattler et al., 2003; Berthou et al; 2004; Ma et al., 2005; Smolen et al., 2006). Subsequent research and development led to the discovery of various types of synthetic tyrosine kinase inhibitors with different structures, chemical properties, and target specificity. Based on the wealth of accumulated knowledge gained from the success of preclinical and clinical development of small synthetic tyrosine kinase inhibitors, more than 10 small synthetic Met tyrosine kinase inhibitors have been entered into clinical trials (Table 1).

PF-02341066 targets Met as well as anaplastic lymphoma kinase (ALK) (Sattler & Salgia, 2009). MP470 inhibits PDGFR, Kit, and Met tyrosine kinases. In combination with erlotinib, MP470 inhibited prostate cancer cell proliferation and tumor xenograft growth (Qi et al., 2009). E7050 targets both Met and VEGFR2 (Nakagawa et al., 2009). JNJ-38877605 shows a >1,000-fold selectivity for the Met kinase, compared to a >200-fold selectivity for related receptor tyrosine kinases (Eder et al., 2009). AMG 208 selectively inhibits both ligand-dependent and ligand-independent Met activation. BMS777607 has completed a phase I/II study in metastatic cancer patients (Schroeder et al., 2009). Phase I clinical trials were discontinued for SGX523 after renal toxicity was observed in patients receiving relatively low doses ( HYPERLINK "http://www.sgxpharma.com" www.sgxpharma.com). PF02341066 and XL184 have progressed the furthest of all Met inhibitors in clinical development. PF-02341066 has greater Met selectivity compared with PF-04217903 (Timofeevski et al., 2009). Preclinical studies indicate PF-02341066 is highly effective against the product of the EML4-ALK translocation found in a subset of NSCLC patients (Shaw et al., 2009). PF-02341066t is currently in phase I, II, and III clinical trials ( HYPERLINK "http://www.clinicaltrials.gov" www.clinicaltrials.gov). XL184 targets Met, VEGFR2, and Ret. A current phase III trial is investigating the efficacy of XL184 as a first-line treatment, compared to a placebo, in patients with medullary thyroid cancer ( HYPERLINK "http://www.clinicaltrials.gov" www.clinicaltrials.gov).
