Aberrant expression and/or activation of c-kit has been implicated in a variety of tumors. Although many tumor cells express SCF, which may indirectly stimulate tumor cell growth by inducing cytokine release from mast cells or other cells that normally respond to SCF, the discussion in this chapter will be limited to c-kit activation that has been directly implicated in tumor cell growth. In this regard, the strongest evidence for a contribution of c-kit to neoplastic pathology is associated with leukemias and mast cell tumors, small cell lung cancer, testicular cancer, and some cancers of the gastrointestinal tract and central nervous system (see below). In addition, c-kit has been implicated in playing a role in carcinogenesis of the female genital tract (Inoue et al., 1994), sarcomas of neuroectodermal origin (Ricotti et al., 1998), and Schwann cell neoplasia associated with neurofibromatosis (Ryan et al., 1994). It is of interest that c-kit has not been found to be associated with melanoma, given that activation of c-kit is essential for normal melanocyte development, and mice with mutations of either c-kit or SCF have abnormal pigmentation. Although SCF appears to induce proliferation of normal melanocytes, activation of c-kit appears to inhibit the proliferation of, and induce apoptosis in melanoma cells (Zakut et al., 1993; Huang et al., 1996). This may explain why c-kit is not expressed in the majority of melanoma cells or tumors (Lassam and Bickford, 1992; Natali et al., 1992b; Zakut et al., 1993; Ohashi et al., 1996).
SCF binding to the c-kit RTK protects hematopoietic stem and progenitor cells from apoptosis (Lee et al., 1997), thereby contributing to colony formation and hematopoiesis. Expression of c-kit is frequently observed in acute myelocytic leukemia (AML), but is less common in acute lymphocytic leukemia (ALL)
(for reviews, see Sperling et al., 1997, and Escribano et al., 1998). Although c-kit is expressed in the majority of AML cells, its expression does not appear to be prognostic of disease progression (Sperling et al., 1997). However, SCF protected AML cells from apoptosis induced by chemotherapeutic agents (Hassan and Zander, 1996), suggesting that inhibition of c-kit would enhance the efficacy of these agents. The clonal growth of cells from patients with myelodysplastic syndrome (Sawada et al., 1996) or chronic myelogenous leukemia (CML) (Sawai et al., 1996) was found to be significantly enhanced by SCF in combination with other cytokines. CML is characterized by expansion of Philadelphia chromosome-positive cells of the marrow (Verfaillie, 1998), which appears to primarily result from inhibition of apoptotic death (Jones, 1997). The product of the Philadelphia chromosome, p210BCR-ABL, has been reported to mediate inhibition of apoptosis (Bedi et al., 1995). Since p210BCR-ABL and the c-kit RTK both inhibit apoptosis and p62dok has been suggested as a substrate (Carpino et al., 1997), it is possible that clonal expansion mediated by these kinases occurs through a common signaling pathway. However, c-kit has also been reported to interact directly with p210BCR-ABL (Hallek et al., 1996), which suggests that c-kit may have a more causative role in CML pathology.
SCF is expressed in lung epithelial cells (Pietsch et al., 1998) and lung cancer cells (Hibi et al., 1991; Turner et al., 1992). In contrast, c-kit is not expressed in normal lung tissues (Hida et al., 1994; Pietsch et al., 1998), but it becomes aberrantly expressed in a subset of non-small cell lung cancers (Pietsch et al., 1998) and in the majority of small cell lung cancers (SCLC) (Hibi et al., 1991, Sekido et al. 1991). In fact, more than 70% of SCLC appear to express both SCF and c-kit, suggesting that this autocrine loop may contribute to the malignant features of SCLC (Hibi et al., 1991; Turner et al., 1992; Rygaard et al., 1993). Although it has been reported that addition of SCF to SCLC cells exerts, at best, a modest stimulation of cell growth (Papadimitriou et al., 1995; Shui et al., 1995) and no significant enhancement of survival after irradiation (Shui et al., 1995), there is evidence that the SCF/c-kit autocrine loop that is often observed in SCLC cells may contribute to growth factor-independent proliferation and protection of the tumor cells from apoptosis. When an autocrine loop was introduced into a SCLC cell line, it conferred a significant growth advantage in serum-free medium and enabled the cells to grow well in the presence of other growth factors (Krystal et al., 1996). Conversely, introduction of a dominant-negative c-kit receptor into a SCLC auto-crine cell line significantly inhibited the ability of the cells to grow in serum-
free medium (Krystal et al., 1996). The growth of SCLC cells has also been shown to be inhibited by c-kit antisense RNA transcripts introduced with an adenoviral vector (Yamanish et al., 1996), recombinant SCF-exotoxin (Ni-shida et al., 1997), or the synthetic tyrosine kinase inhibitor AG1296 (Krystal et al., 1997). In the latter case, the inhibition of c-kit RTK induced apoptosis in SCLC cells, suggesting that inhibition of c-kit kinase may have anticancer utility for SCLC (Krystal et al., 1997). Since SCLC is characterized by an aggressive primary disease that often relapses after chemotherapy (see Ihde et al., 1997 for a review), treatment of relapsed SCLC with an agent that targets c-kit may provide an additional means to treat recurrent disease in these patients.
Normal colorectal mucosa does not express c-kit (Bellone et al., 1997). However, c-kit is frequently expressed in colorectal carcinoma (Bellone et al., 1997), and autocrine loops of SCF and c-kit have been observed in several colon carcinoma cell lines (Toyota et al., 1993; Lahm et al., 1995; Bellone et al., 1997). Furthermore, disruption of the autocrine loop by the use of neutralizing antibodies (Lahm et al., 1995) and downregulation of c-kit and/or SCF significantly inhibits cell proliferation (Lahm et al., 1995; Bellone et al., 1997). SCF/c-kit autocrine loops have been observed in gastric carcinoma cell lines (Turner et al., 1992; Hassan et al., 1998), and constitutive c-kit activation also appears to be important for gastrointestinal stromal tumors (GIST). GIST are the most common mesenchymal tumors of the digestive system. More than 90% of GIST express c-kit, which is consistent with the putative origin of these tumor cells from interstitial cells of Cajal (ICC) (Hirota et al., 1998). ICC are thought to regulate contraction of the gastrointestinal tract, and patients lacking c-kit in their ICC exhibited a myopathic form of chronic idio-pathic intestinal pseudo-obstruction (Isozaki et al., 1997). The c-kit expressed in GIST from several different patients was observed to have mutations in the intracellular juxtamembrane domain leading to constitutive activation of this RTK (Hirota et al., 1998).
Male germ cell tumors have been histologically categorized into seminomas, which retain germ cell characteristics and nonseminomas which can display characteristics of embryonal differentiation. Both seminomas and nonsemino-mas are thought to initiate from a preinvasive stage designated carcinoma in situ (CIS) (Murty and Chaganti, 1998). Both c-kit and SCF are essential for normal gonadal development during embryogenesis (Loveland and Schlatt,
1997). Loss of either the receptor or the ligand resulted in animals devoid of germ cells. In postnatal testes, c-kit has been found to be expressed in Leydig cells and spermatogonia, whereas SCF was expressed in Sertoli cells (Love-land and Schlatt, 1997). Testicular tumors develop from Leydig cells with high frequency in transgenic mice expressing human papilloma virus 16 (HPV16) E6 and E7 oncogenes (Kondoh et al. 1991, 1994). These tumors express both c-kit and SCF, suggesting that an autocrine loop may contribute to the tumorigenesis (Kondoh et al., 1995) associated with cellular loss of functional p53 and the retinoblastoma gene product by association with E6 and E7 (Dyson et al., 1989 Scheffner et al., 1990 Werness et al., 1990). The observation that defective signaling mutants of SCF (Kondoh et al., 1995) or c-kit (Li et al., 1996) inhibited formation of testicular tumors in mice expressing HPV16 E6 and E7 indicates that c-kit activation is pivotal to tumorigenesis in these animals. Expression of c-kit on germ cell tumors has been examined by several groups with similar results. The receptor is expressed by the majority of carcinomas in situ and seminomas, but c-kit is expressed in only a minority of nonseminomas (Strohmeyer et al., 1991; Rajpert-de Meyts and Skakke-baek, 1994; Izquierdo et al., 1995; Strohmeyer et al., 1995; Bokemeyer et al., 1996; Sandlow et al., 1996). There is only one report of coexpression of c-kit and SCF (Bokemeyer et al., 1996) and no reports of activation mutations of c-kit in testicular tumors.
SCF and c-kit are expressed throughout the CNS of developing rodents, and the pattern of expression suggests a role in growth, migration, and differentiation of neuroectodermal cells. Expression of both receptor and ligand have also been reported in the adult brain (Hamel and Westphal, 1997). Expression of c-kit has also been observed in normal human brain tissue (Tada et al., 1994). Glioblastoma and astrocytoma, which define the majority of intracranial tumors, arise from neoplastic transformation of astrocytes (Levin et al., 1997). Expression of c-kit has been observed in glioblastoma cell lines and tissues (Berdel et al., 1992; Tada et al., 1994; Stanulla et al., 1995). However, exogenous addition of SCF to glioblastoma cell lines appears to be mitogenic in only a minority of cases (Berdel et al., 1992; Stanulla et al., 1995), and antibodies that block the interaction of SCF with c-kit did not inhibit the proliferation of cells with an autocrine loop (Stanulla et al., 1995). The association of c-kit with astrocytoma pathology is less clear. Reports of expression of c-kit in normal astrocytes have been made (Natali, et al., 1992a; Tada et al., 1994), whereas others report it is not expressed (Kristt et al., 1993). In the latter case, high levels of c-kit expression in high-grade tumors were observed
(Kristt et al., 1993), whereas the former groups were unable to detect any expression in astrocytomas. In addition, contradictory reports of c-kit and SCF expression in neuroblastomas also exist. One study found that neuroblastoma cell lines often express SCF but rarely express c-kit. In primary tumors, c-kit was only detected in about 8% of neuroblastomas, whereas SCF was found in only 18% of tumors (Beck et al., 1995). In contrast, other studies (Cohen et al., 1994) have reported that all 14 neuroblastoma cell lines examined contained c-kit/SCF autocrine loops, and expression of both the receptor and li-gand were observed in 45% of tumor samples examined. In two cell lines, anti-c-kit antibodies inhibited cell proliferation, suggesting that the SCF/c-kit autocrine loop contributed to growth (Cohen et al., 1994).
Mastocytosis As mentioned above, SCF (also known as mast cell growth factor) stimulation of c-kit has been shown to be essential for the growth and development of mast cells (Galli and Hammel, 1994; Kitamura et al., 1995). Mice with mutations of c-kit that attenuate its signaling activity have exhibited significantly fewer mast cells in their skin (Tsujimura et al., 1996). Therefore, it is not surprising that excessive activation of c-kit might be associated with diseases resulting from an over abundance of mast cells. Mastocytosis is the term used to describe a heterogeneous series of disorders characterized by excessive mast cell proliferation (Metcalfe, 1991; Valent, 1996; Golkar and Bernhard, 1997). Mastocytosis is limited to the skin in the majority of patients, but it can involve other organs in 15-20% of patients (Valent, 1996; Golkar and Bernhard, 1997). Even among patients with systemic mastocytosis, the disease can range from having a relatively benign prognosis to aggressive mastocytosis and mast cell leukemia (Valent, 1996; Golkar and Bernhard, 1997). c-kit Has been observed on malignant mast cells from canine mast cell tumors (London et al., 1996), as well as on mast cells from patients with aggressive systemic mastocytosis (Baghestanian et al., 1996; Castells, et al., 1996). Elevated c-kit expression was reported on mast cells from patients with aggressive mastocytosis but not on mast cells from patients with indolent mastocytosis (Nagata et al., 1998), suggesting that overexpression may contribute to the pathology associated with more aggressive forms of the disease in some patients.
SCF has been shown to be expressed on stromal cells as a membrane-bound protein, and its expression can be induced by fibrogenic growth factors such as PDGF (Hiragun et al., 1998). It has also been shown to be expressed on keratinocytes as a membrane-bound protein in normal skin. However, in the skin of patients with mastocytosis, an increased amount of soluble SCF
has been observed (Longley et al., 1993). Mast cell chymase has been reported to cleave membrane-associated SCF to a soluble and biologically active form. This mast cell-mediated process could serve to generate a feedback loop to enhance mast cell proliferation and function (Longley et al., 1997), and may be important for the etiology of mastocytosis. Transgenic mice overexpressing a form of SCF that could not be proteolytically released from keratinocytes did not develop mastocytosis, whereas similar animals expressing normal SCF in keratinocytes exhibited a phenotype resembling human cutaneous masto-cytosis (Kunisada et al., 1998). This observation suggested that formation of large amounts of soluble SCF can contribute to the pathology associated with mastocytosis in some patients. Several different mutations of the c-kit RTK that resulted in constitutive kinase activity have been found in human and rodent mast cell tumor cell lines (Furitsu et al., 1993; Tsujumura et al., 1994, 1995, 1996). In addition, activating mutations of the c-kit gene have been observed in peripheral mononuclear cells isolated from patients with masto-cytosis and associated hematological disorders (Nagata et al., 1998) and in mast cells from a patient with urticaria pigmentosa and aggressive masto-cytosis (Longley et al., 1996). These reports indicate that, in some patients, activating mutations of the c-kit RTK may be responsible for the pathogenesis of the disease. SCF activation of c-kit has been shown to prevent mast cell apoptosis, which may be critical for maintaining cutaneous mast cell homeo-stasis (Iemura et al., 1994; Mekori and Metcalfe, 1994,1995; Yee et al., 1994). Inhibition of mast cell apoptosis could lead to the mast cell accumulation associated with mastocytosis. Thus, observation of c-kit activation resulting from overexpression of the receptor, excessive formation of soluble SCF, or mutations of the c-kit gene that constitutively activate its kinase provides a rationale that inhibition of the kinase activity of c-kit may decrease the number of mast cells, and this aspect may provide benefit for patients with masto-cytosis.
Asthma and Allergy Mast cells and eosinophils represent key cells in parasitic infection, allergy, inflammation, and asthma (Thomas and Warner, 1996; Costa et al., 1997; Holgate, 1997; Metcalfe et al., 1997; Naclerio and Solomon, 1997). SCF has been shown to be essential for mast cell development, survival, and growth (Kitamura et al., 1995; Metcalfe et al., 1997). In addition, SCF cooperates with the eosinophil-specific regulator, interleukin-5 (IL-5) to increase the development of eosinophilic progenitors (Metcalf, 1998). SCF has also been shown to induce mast cells to secrete factors (Oka-yama et al., 1997, 1998) that promote the survival of eosinophils (Kay et al., 1997), which may contribute to chronic, eosinophil-mediated inflammation (Okayama et al., 1997, 1998). In this regard, SCF directly and indirectly regulates activation of both mast cells and eosinophils. SCF induces mediator release from mast cells, as well as priming these cells for IgE-induced degranulation (Columbo et al., 1992) and sensitizing their responsiveness to eosino-phil-derived granular major basic protein (Furuta et al., 1998). Among the factors released by activated mast cells are IL-5, granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor-a (TNF-a), which influence eosinophilic protein secretion (Okayama et al., 1997, 1998). In addition to inducing histamine release from mast cells (Luckacs et al., 1996; Hogaboam et al., 1998), SCF promotes the mast cell production of the eosinophilic chemotactic factor eotaxin (Hogaboam et al., 1998) and eosinophilic infiltration (Luckacs et al., 1996). SCF also directly influences the adhesion of both mast cells (Dastych and Metcalfe, 1994; Kinashi and Springer, 1994) and eosinophils (Yuan et al., 1997), which in turn regulates tissue infiltration. Thus, SCF can influence the primary cells involved in allergy and asthma through multiple mechanisms. Currently, corticosteroids are the most effective treatment for chronic rhinitis and inflammation associated with allergy (Mel-tzer, 1997; Naclerio and Solomon, 1997). These agents work through multiple mechanisms, including reduction of circulating and infiltrating mast cells and eosinophils and diminished survival of eosinophils associated with inhibition of cytokine production (Meltzer, 1997). Steroids have also been reported to inhibit the expression of SCF by fibroblasts and resident connective tissue cells, which leads to diminished mast cell survival (Finotto et al., 1997). Because of the mutual regulation of mast cell and eosinophilic function, and the role that SCF can play in this regulation, inhibition of c-kit kinase may provide a means to treat allergy-associated chronic rhinitis, inflammation, and asthma.
Was this article helpful?