Growth factors and their receptors in cancer metastasis

Growth Factors and their Receptors in Cancer Metastasis

Cancer Metastasis – Biology and Treatment

VOLUME 2

Series Editors Richard J. Ablin, Innapharma, Inc., Parkridge, U.S.A. Wen G. Jiang, University of Wales College of Medicine, Cardiff, U.K.

Advisory Editorial Board Harold F. Dvorak Phil Gold Ian R. Hart Hiroshi Kobayashi Robert E. Mansel Marc Mareel

Growth Factors and their Receptors in Cancer Metastasis Edited by

Wen G. Jiang University of Wales College of Medicine, Cardiff, The United Kingdom

Kunio Matsumoto University of Osaka Medical School, Osaka, Japan

and

Toshikazu Nakamura University of Osaka Medical School, Osaka, Japan

KLUWER ACADEMIC PUBLISHERS NEW YORK, BOSTON, DORDRECHT, LONDON, MOSCOW

eBook ISBN: Print ISBN:

0-306-48399-8 0-7923-7141-0

©2004 Kluwer Academic Publishers New York, Boston, Dordrecht, London, Moscow Print ©2001 Kluwer Academic Publishers Dordrecht All rights reserved No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher Created in the United States of America Visit Kluwer Online at: and Kluwer's eBookstore at:

http://kluweronline.com http://ebooks.kluweronline.com

Table of Contents

Contributors The Role of Leukemia Inhibitory Factor in Cancer and Cancer Metastasis FARHAD RAVANDI AND ZEEV ESTROV Interleukin-2 and Its Receptors in Human Solid Tumours: Immunobiology and Clinical Significance THERESA L. WHITESIDE, TORSTEN E. REICHERT, AND QING PING DOU.

vii

1

27

Interleukin-8 and Angiogenesis TRACEY A. MARTIN.

51

The Role of Interleukin-11 in the Formation of Bone Metastases. NAOYA FUJITA AND TAKASHI TSURUO.

67

Therapeutic Potential of Adenovirus Mediated Interleukin-12 Gene Therapy for Prostate Cancer SHIN EBARA, YASUTOMO NASU, TAKEFUMI SATOH, SATORU SHIMURA, CHRIS H. BANGMA, GERALD W. HULL, MARK A. MCCURDY, JIANXIANG WANG, GUANG YANG, TERRY L. TIMME, AND TIMOTHY C. THOMPSON. Fibroblast Growth Factors and Their Receptors in Metastases of Prostate and Other Urological Cancers ZORAN CULIG, MARCUS V. CRONAUER, ALFRED HOBISCH, GEORG BARTSCH AND HELMUT KLOCKER.

79

93

vi

Table of Contents

Insulin-Like Growth Factor Axis Elements in Breast Cancer Progression EMILIA MIRA, ROSA ANA LACALLE, CARLOS MARTÍNEZ-A. AND SANTOS MAÑES. The Role of Platelet Derived Growth Factor (PDGF) and Its Receptors in Cancer and Metastasis SARA WEISS FEIGELSON, CHERYL FITZER-ATTAS, LEA EISENBACH.

107

167

TGF receptor Signaling in Cancer and Metastasis MARTIN OFT.

187

VEGF-C/VEGFRS and Cancer Metastasis YUTAKA YONEMURA, YOSHIO ENDOU, TAKUMA SASAKI, KAZUO SUGIYAMA, TETUMOURI YAMASHIMA, TAINA PARTANEN, KARI ALITALO.

223

HGF-c-met receptor Pathway in Tumor Invasion-Metastasis and Potential Cancer Treatment with NK4 KUNIO MATSUMOTO AND TOSHIKAZU NAKAMURA.

241

Growth Factor Receptors and Cell Adhesion Complexes in Cytoskeletal Assembly/Anchorage GAYNOR DAVIES, MALCOLM D. MASON AND WEN G. JIANG

277

Index

291

vii

Contributors

Alitalo, Kari. Molecular/Cancer Biology Laboratory, Haatman Institute, University of Helsinki, PL 21, 00014 Helsinki, Finland Bangma, Chris H. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Bartsch, Georg. Department of Urology, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria Cronauer, Marcus V. Department of Urology, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria Culig, Zoran. Department of Urology, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria Davies, Gaynor. Metastasis Research Group, University Department of Surgery, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XN.UK. Dou, Qing Ping. H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL 33612, USA Ebara, Shin. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Eisenbach, Lea. Weizmann Institute of Science, Department of Immunology, Rehovot, Israel Endou, Yoshio. Experimental Therapeutics Cancer Research Institute, Kanazawa University, Takara-Machi 13-1, Kanazawa 920, Japan Estrov, Zeev. Department of Bioimmunotherapy, The University of Texas, M. D. Anderson Cancer Center, Houston, Texas, U.S.A

viii

Contributors

Feigelson, Sara Weiss. Weizmann Institute of Science, Department of Immunology, Rehovot, Israel Fitzer-Attas, Cheryl. Weizmann Institute of Science, Department of Immunology, Rehovot, Israel Fujita, Naoya. Institute of Molecular and Cellular Biosciences, University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan Hobisch, Alfred. Department of Urology, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria Hull, Gerald W. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Jiang, Wen G. Metastasis Research Group, University Department of Surgery, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XN, UK. Klocker, Helmut. Department of Urology, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria Lacalle, Rosa Ana. Department of Immunology and Oncology, Centro Nacional de Biotecnología-CSIC, Campus de Cantoblanco, Universidad Autónoma de Madrid, E-28049 Madrid, Spain. Mañes, Santos. Department of Immunology and Oncology, Centro Nacional de Biotecnología-CSIC, Campus de Cantoblanco, Universidad Autónoma de Madrid, E-28049 Madrid, Spain. Martínez, Carlos. Department of Immunology and Oncology, Centro Nacional de Biotecnología-CSIC, Campus de Cantoblanco, Universidad Autónoma de Madrid, E-28049 Madrid, Spain. Mason, Malcolm D. Department of Medicine, Section of Clinical Oncology, Velindre Hospital, Cardiff, CF4 7XL. UK Martin, Tracey A. Metastasis Research Group, University Department of Surgery, University of Wales College of Medicine, Heath Park, Cardiff, CF14 4XN. UK. Matsumoto, Kunio. Division of Biochemistry, Department of Oncology, Biomedical Research Center, Osaka University Graduate School of Medicine, Suita, Osaka 565-0871, Japan McCurdy, Mark A. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA

Contributors

Mira, Emilia. Department of Immunology and Oncology, Centro Nacional de Biotecnología-CSIC, Campus de Cantoblanco, Universidad Autónoma de Madrid, E-28049 Madrid, Spam. Nakamura, Toshikazu. Division of Biochemistry, Department of Oncology, Biomedical Research Center, Osaka University Graduate School of Medicine, Suita, Osaka 565-0871, Japan Nasu, Yasutomo. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Oft, Martin. UCSF Cancer Center - Box 0875, University of California, San Francisco, 2340 Sutter Street, Room S271, San Francisco, CA 941430128, USA Partanen, Taina. Molecular/Cancer Biology Laboratory, Haatman Institute, University of Helsinki, PL 21, 00014 Helsinki, Finland Ravandi, Farhad. Department of Bioimmunotherapy, The University of Texas, M. D. Anderson Cancer Center, Houston, Texas, U.S. A Reichert, Torsten E. University of Mainz, Mainz, Germany Sasaki, Takuma. Experimental Therapeutics Cancer Research Institute, Kanazawa University, Takara-Machi 13-1, Kanazawa 920, Japan Satoh, Takefumi. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Shimura, Satoru. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Sugiyama, Kazuo. Virology Division, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuoo-Ku, Tokyo 104, Japan Thompson, Timothy C. Scott Department of Urology, Cell Biology and Radiology, Baylor College of Medicine, Houston, Texas, USA Timme, Terry L. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Tsuruo, Takashi. Institute of Molecular and Cellular Biosciences, University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan Wang, Jianxiang G. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA Wang Yang, Guang. Scott Department of Urology, Baylor College of Medicine, Houston, Texas, USA

ix

x

Contributors

Whiteside, Theresa L. University of Pittsburgh Cancer Institute, Pittsburgh, PA 15213, USA Yamashima, Tetumori. Department of Neurosurgery, School of Medicine, Kanazawa University, Takara-Machi 13-1, Kanazawa 920, Japan Yonemura, Yutaka. Second Department of Surgery, School of Medicine, Kanazawa University, Takara-Machi 13-1, Kanazawa 920, Japan

Chapter 1 THE ROLE OF LEUKEMIA INHIBITORY FACTOR IN CANCER AND CANCER METASTASIS

Farhad Ravandi and Zeev Estrov Department of Bioimmunotherapy, The University of Texas, M. D. Anderson Cancer Center, Houston, Texas, U.S.A

Key words:

Leukemia inhibitory factor (LIF), Cancer, Metastasis

Abstract:

Leukemia inhibitory factor (LIF) is a cytokine that exerts pleiotropic activities. LIF is a member of the interleukin-6 family of cytokines which share a similar receptor complex and signal through the gp 130 receptor subunit. Several neoplastic cells originating from various tissues express either LIF, its receptor, or both and respond to this cytokine. Data accumulated thus far provide a complex picture of LIF activities with LIF being stimulatory, inhibitory or having no effect, depending on the system in which it is studied. LIF appears to play an important role in stimulating the growth of certain tumours, and in affecting the surrounding tissue and the target organ of tumour metastases, particularly bone and skeletal tissue. Overproduction of LIF is likely to have significant constitutional effects. Studies using animal models have shown that LIF induces cachexia, metastatic-type bone calcifications, thrombocytosis, and an abnormal immune response. It is therefore possible that suppression of LIF activity might have a beneficial effect in some cancer patients.

induce proliferation, inhibit proliferation or cause apoptosis, depending on the system in which this cytokine is studied. Several studies have shown that LIF’s divergent physiological effects have been adopted by a variety of neoplastic cells and that LIF takes part in the pathophysiology of cancer. In many neoplasms LIF, produced by either normal tissue or tumour cells, provides the cancerous process with growth and survival advantage. LIF was initially characterized by its ability to induce differentiation of the

1. INTRODUCTION Leukemia inhibitory factor (LIF) is a pluripotent cytokine with pleiotropic activities. LIF is a member of a family of cytokines that includes the ciliary neurotrophic factor (CNTF), interleukin (IL)-6, IL-11, oncostatin-M (OSM), and cardiotropin-1 (1,2). These cytokines are grouped as a family because of their shared helical bundle structure (3-7), shared subunits of their receptor complexes, and in some cases, overlapping functions (8,9). As other members of this family, LIF can either 1

W.G. Jiang et al. (eds.), Growth Factors and their Receptors in Cancer Metastasis, 1–25. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

2

Ravandi and Estrov

murine myeloid leukemia cell line M1 (10-15) and was cloned from a murine Tcell library (12,16). Independently, a human molecule in the supernatant of Tcell clones was identified and termed human interleukin for DA cells (HILDA) (17-20). Once cloned, this molecule was found to be homologous to its murine counterpart (21-23). Subsequently, additional characteristics of LIF were described, and it was given several other names, including differentiation factor (Dfactor) (24-25), differentiation-inducing factor (DIF) (26), differentiation inhibitory activity (DIA) (27), differentiation-retarding factor (DRF) (27,28), hepatocyte-stimulating factor III (HSF III) (29), melanoma-derived lipoprotein lipase inhibitor I (MLPLI) (34), cholinergic neural differentiation factor (31), and osteoclast-activating factor (OAF) (26,32) (Table 1). However, because LIF exerts a broad spectrum of activities and despite its diverse and sometimes opposing effects on different leukemia cell lines (21,26,27,29,32,3335), LIF has become the official name of this cytokine (1). The effects of LIF on various tissues provide several clues to its possible role in cancer. For example, LIF stimulates embryonic stem cell proliferation (36-40). It affects blastocyst implantation (36-41) and influences the development of peripheral nerves from their precursors in the embryonic neural crest (32,42), which implies that LIF can stimulate immature cells and probably tumour cells with immature cell characteristics. In addition, LIF was shown to induce a catabolic state

Chapter 1 and cachexia in nude mice and in primates (43-45). It stimulated the release of acutephase proteins from hepatocytes, (45-47) and affected bone metabolism by inducing both osteoblastic and osteoclastic activities (48-52). These effects are characteristic clinical features of patients with neoplastic diseases likely to be induced by various cytokines including LIF. In this chapter we describe the physiological characteristics and the pathophysiological role of LIF in cancer and cancer metastasis. 2. MOLECULAR AND CELLULAR CHARACTERISTICS 2.1 LIF Distribution in Cells LIF is expressed in cells of different tissues, including osteoblasts, keratinocytes, thymic epithelium, T cells, monocytes, skin fibroblasts, embryonic stem cells, bone marrow stroma cells, central nervous system cells, hepatocytes, and a number of tumour cell lines that have become a source for this cytokine (1,20,21,26,37,53-57) (Table 2). 2.2 Structure and Genetics Naturally occurring LIF appears as a monomeric glycoprotein with a molecular weight is between 40 and 70 kDa despite a polypeptidic core of 22 kDa (3,21,58). This is due to the presence of several putative sites of N-glycosylation in the primary structure of the molecule allowing extensive post-translational modifications (29).

Abbreviations: LIF, leukemia inhibitory factor; LIFR, LIF receptor; TNF, tumour necrosis factor; AML, acute myeloid leukemia; MDS, myelodysplastic syndrome; OSM, oncostatin M; IFN, intrerferon; IL, interleukin; IFN, interferon, CNTF, ciliary neurotrophic factor; CFU, colony forming unit; MM, multiple myeloma; TGF, transforming growth factor; G-CSF, granulocyte colony-stimulating factor; HSF, hepatocyte stimulating factor; SP, neuropeptide substance P; CNS, central nervous system

1. LIF and cancer metastasis

3

4

Ravandi and Estrov

LIF is encoded by genes localized at chromosome 11A1 in mice and chromosome 22ql2 in humans (59,60). Although the location of the human gene and the high incidence of a translocation involving t(11;22)(q24q12) in Ewing’s sarcoma stimulated considerable interest, further analysis using somatic cell hybrids and pulse-field gel electrophoresis has shown that the gene is located distal to the breakpoint and is not involved in this translocation (61). The sequences of cloned LIF genes from four mammalian species are highly conserved in the coding regions (62,63). Murine and human LIF have the complete nucleotide sequence of 8.7 and 7.6 kilobase pairs, respectively (12,14,16, 64,65). Both genes consist of three exons, two introns, and an unusually large 3’untranslated region that is 3.2 kilobase pairs (65). The LIF transcript is 4.2 kilobases in length and predicts a sequence with 179 residues for the mature protein and a 79% homology between the murine and human products (12,14,64). This is the primary and biologically active form of LIF. The promoter region of the LIF gene contains four highly conserved TATA elements, with two identified start sites of transcription (62). Three regions within the 5’ flanking region have been identified as important to the function of the LIF promoter (21,62). The structure of LIF has been determined (66). The main chain fold comprises four α-helices linked by two loops. There are two regions of the LIF molecule involved in receptor interaction and biological function. The first is located within the D helix and comprises residues 161-180, and the second is located between residues 150 and 160 at the C-terminus of the CD loop.

Chapter 1 2.3 Biological Forms of LIF Two forms of LIF were detected: the “diffusible” (D) LIF glycoprotein and an “immobilized” (M) form incorporated into the extracellular matrix (67). Both D- and M-LIF forms are produced by the expression of alternative transcripts that diverge throughout the first exon and use different promoters. The two LIF forms are encoded by mRNAs that are spliced differently at the exon 1/exon 2 boundary. The transcript D encodes the diffusible form and the transcript M encodes the matrix-associated form. Splicing a 5’ exon to exons 2 and 3 of the LIF transcription unit produces the latter. The two transcripts co-migrate on agarose gel and therefore can be distinguished by ribonuclease protection analysis but not by Northern blot analysis. The molecular organization of the gene for LIF can explain the different localization of its two forms. Exons 2 and 3 produce the core hydrophobic secretory sequences, whereas the extracellular localization is determined by the first exon. Therefore, changes in the amino terminal of the translocation product direct the formation of a mature, functional LIF with extracellular matrix localization (reviewed in 1 and 2). Although the reported molecular weight of LIF ranges from 38 to 67 kDa, this heterogeneity can be explained by variable glycosylation of the protein (35,64). Recombinant forms of LIF displaying varying patterns of glycosylation (yeast-derived and Escherichia coli-derived) are active (12,64). 2.4 LIF Receptors and Their Signaling The IL-6 cytokine family members share common signaling components i.e. the LIF receptor (LIFR) and the receptor

1. LIF and cancer metastasis subunit gp130 (68). The LIFR was first isolated and found to be structurally related to the gp130 component of the IL6 receptor and the granulocyte colonystimulating factor (G-CSF) (69). This receptor is now termed LIFR It binds with low affinity to gp130, whereas LIF binds with high affinity to the LIFRß/gp130 complex, initiating its signal trasduction (70). Other components of this receptor complex, used by other members of the IL-6 cytokine family, have been identified. For example, the receptor component CTNFR is utilized by CTNF(71,72). LIF binds to a variety of cells from different tissues (24,33,56,73,74). Following receptor binding, signaling pathways involving both protein tyrosine and serine/threonine kinases are activated. Both the Janus-kinase-signal transducer and activator of transcription (JAKSTAT) and mitogen-activated protein kinase (MAPK) pathways are activated (39,72,75) (Figure 1). Activated STAT molecules dimerize and translocate to the nucleus. Although there are at least six STAT proteins, STAT3 tends to be the protein that is activated by LIF (76). The LIFRß is essential for motor neuron development as demonstrated in studies with the LIFRß knockout mouse model (77) (see below). 3. BIOLOGICAL EFFECTS OF LIF 3.1 Effects on the Reproductive System and Embryogenesis

The mammalian embryo develops from a quasi-stem cell system controlled by regulatory factors, one of which is LIF (13,32,78,79). LIF is expressed in both embryonic and maternal tissue. LIF transcripts were also detected in mouse blastocysts, implying its role as a regulator of embryonic stem cells and a

5 mediator of the trophoblast development (36,74). In the LIF knockout mouse model, homozygous and heterozygous null mice for a functional LIF gene enabled investigating the role of LIF in the reproductive system (80). Male -/- LIF mice were fertile, but female mice, although able to produce viable blastocytes, failed to implant and were therefore sterile. However, the injection of LIF into homozygous -/- female restored blastocyte implantation (81). Male mice engrafted with a LIFproducing cell line showed complete absence of spermatogenesis, whereas female mice had reduction or complete absence of corporae luteae (47). Recent data confirm the crucial role of LIF during implantation and pregnancy in primates such as monkeys (82) and western spotted skunk (83). In addition, it has been demonstrated that LIF has a crucial role in the maintenance of pregnancy in humans (84,85). 3.2 Effects on Bone Metabolism

Several studies have clearly demonstrated the role of LIF in bone remodeling (86). Both osteoclast and osteoblast activities are either stimulated or suppressed by LIF depending on the developmental stage of the respective cells. When the local effect of LIF was studied in mice by injecting the cytokine over one hemicalvaria, two major effects were observed: 1) increased osteoclastic activity and bone resorption in the injected right hemicalvaria and 2) increased total mineralization, including the periosteal area in the non-injected left hemicalvaria (87). Additional studies using an array of laboratory assays showed that LIF inhibited osteogenic calcification (88), affected osteoclast migration (89), increased osteoclast differentiation (90),

6 Ravandi and Estrov Chapter 1

1. LIF and cancer metastasis inhibited bone module formation (91), and reduced bone calcification (92). Engraftment of mice with LIF-producing cells yielded results similar to those described above (87). The engrafted mice had increased calcifications in both skeletal and extra-skeletal tissues such as the myocard (47). In vitro stimulation of bone resorption by LIF was accompanied by the release of calcium from prelabelled mouse calvaria. This effect, caused by an increase in the number of osteoclasts, could be inhibited by indomethacin, indicating that it is mediated through the activation of prostaglandins. The prostaglandindependent bone-resorptive effect of LIF is similar to that of other cytokines such as IL-1, tumour necrosis factor (TNF), and transforrming growth factor (TGF)-ß (93,94). 3.3 Effects on Lipid Metabolism LIF, like TNF, IL-1, and interferon (IFN)can inhibit the enzyme lipoprotein lipase, which is a key enzyme in triglyceride metabolism (95,96). High levels of LIF induced a fatal catabolic state with cachexia in mice and monkeys (12,33,43-45,97). It is likely that this effect of LIF is mediated by its ability to suppress adipogenic processes through its enzyme inhibitory effect. Thus, the inhibition of lipoprotein lipase is likely to reduce the intake of fatty acids by adipocytes and lead to cachexia. 3.4 Role of LIF in Inflammation and Tissue Injury A number of studies demonstrated the role of LIF in inflammation. LIF was found to have both a pro- and an antiinflammatory role in a variety of inflammatory disorders (98). LIF mRNA increased in various mouse tissues during systemic inflammation triggered by the

7 injection of either endotoxin or lipopolysaccharide (LPS) (99). Interestingly, passive immunisation against LIF prior to LPS injection protected the mice from the lethal effect of high-dose LPS (100), indicating that LIF is one of the agents associated with the lethality of septic shock. Surprisingly, LIF injection prior to a challenge with high dose LPS protected against the lethal dose of LIF (101,102). This dual effect of LIF was found in different diseases in humans, such as rheumatoid arthritis (103,104). LIF is highly elevated in the synovial tissue and fluids of patients with rheumatoid arthritis. In addition, human articular chondrocytes and synovial tissue produce LIF that in turn may upregulate proinflammatory cytokines (105-109). Injection of LIF-binding proteins into a goat joint atenuates the inflammatory reaction caused by a prior injection with LIF (110). LIF has also been detected in the pleural effusion of patients with tuberculosis (111) and in the bronchoalveolar lavage fluid of patients with acute respiratory distress syndrome (112). Local inflammatory processes have been shown to be mediated by LIF (113115). On the other hand, the response to injection of complete Freud’s adjuvant is significantly augmented in adult LIF knock-out mice (116). Some of the differences among these studies could be explained by dissimilar experimental designs, dose of LIF, and species and age of the studied animals. However, divergent effects of LIF on the thymus and on T and B lymphocytes (see below) may also contribute to dissimilar results in various experimental models. LIF also plays a role in tissue repair in cases such as stab wound injury and injury to the central and peripheral nervous systems (117-120). LIF mRNA was shown to be upregulated after muscle

8

Ravandi and Estrov

crash injury (121) consistent with LIF’s role as a stimulator of human muscle precursor-cell proliferation (122,123). In LIF knock-out mice, infiltration by neutrophils, macrophages, and mast cells is delayed in lesions of both the central and peripheral nervous systems (124), suggesting that LIF could be chemotactic for inflammatory cells. 3.5 Effects on Hepatic Function The hepatocyte-stimulating factor (HSF) III initially detected in cultured keratinocytes and squamous carcinoma cell lines was found to be identical to LIF (29,125). The liver secretes acute-phase proteins into the circulation upon various stimuli including those induced by several cytokines including LIF (126). Injection of LIF into rhesus monkeys strongly increased the levels of acute-phase proteins (127). This molecule is able to induce the production of a number of acute-phase proteins by the hepatocytes, an ability that it shares with other cytokines including IL-6 and TNF. Thus hepatocytes produce LIF which is capable of inducing the production of acute-phase proteins by the liver, suggesting an autocrine role for LIF. 3.6 Effects on the Nervous System Cholinergic neuronal differentiation factor, a protein acting on sympathetic neurons to induce acetylcholine synthesis and cholinergic function, is now known to be identical to LIF (32). LIF affects the development of peripheral neurons from their precursors in the embryonic neural crest (42). LIF also participates in the regulation of the neuropeptide substance P (SP) in sympathetic neurons, increasing SP in both neuronal cell cultures and cultures containing a mixture of neuronal and non-neuronal cells (126). LIF acts as a survival factor on mature sensory

Chapter 1 neurons (127). Neuronal differentiation of spinal-cord precursors is dependent on a functioning LIFRß (128). In the LIFRß knock-out mouse model, mice die shortly after birth, and they reveal a profound loss of astrocytes in the brain stem and spinal cord, and neurons with pycnotic nuclei and cytoplasmic vacuoles (77,129). These findings and the distribution of the LIFR mRNA in the brain and spinal cord suggests that LIF affects neuronal cells in the adult as well as during development (128). LIF can prevent the death of axotomised sensory and motor neurons (129,130). In the Wobbler mouse model, the animals develop lower-motor neuropathy. Injection of LIF has a sparing effect, improving the neuropathy (131), further demonstrating the complex effects of LIF in the nervous system. 3.7 Effects on the Hematopoietic System LIF has been characterized by its ability to induce differentiation and suppress the growth of M1 myeloid leukemia cells (12-15,48,132). However, in subsequent studies, LIF stimulated, inhibited or had no effect on leukemia cells, depending on the cell line or the system in which LIF’s activity was investigated (133-137). Similarly, LIF induces a divergent effect on normal hematopoietic progenitors. Exposure to LIF reduces the proliferative capability and survival of normal hematopoietic progenitors (138). Although LIF had no effect on CD34+ human bone marrow cells, it enhanced the stimulating effect of IL-3 (139). In another study, LIF stimulated the growth of colony-forming units granulocyte-erythroid-macrophagemegakaryocyte (CFU-GEMM) and CFUeosinophil (CFU-Eo), and burst-forming units-erithroid (BFU-E) colony-forming cells (140). Similar results were obtained

1. LIF and cancer metastasis with CD34+ cells stimulated with IL-3 and IL-6. LIF augmented the effect of megakaryocyte colony-forming cell stimulators and enhanced a chemoattractant effect on human and mouse eosinophils (19,20,73). Bone marrow stroma cells constitutively express LIF mRNA (53). Exposure of hematopoietic stroma to either , or increased the level of LIF mRNA (53,141). Stroma obtained from marrow cells of patients with chronic myelogenous leukemia who had high levels of expressed high levels of LIF (53,142). The first clue of the role LIF plays in normal hematopoiesis in vivo came from experiments carried out in mice (47,48). Mice tranfected with LIF-producing cells exhibited thymic atrophy and extramedullary hematopoiesis (47). Daily injection of LIF caused granulocytosis and an increase in megakaryocytes and platelets (46). Transgenic mice constitutively expressing diffusible LIF displayed B-cell hyperplasia, profound disorganization of the thymus, and loss of cortical CD4+ and CD8+ lymphocytes. Transplantation of transgenic bone marrow into wild-type mice recipients transferred the thymic and lymph node defects (143). Knock-out of the LIF gene significantly impaired the hematopoietic system (80). Both early and mature hematopoietic progenitors were dramatically reduced and a dose effect was seen because heterozygotes were less affected. However, mature hematopoietic elements in the marrow, spleen, and peripheral blood were normal, indicating that the defect was in the stem cell pool rather than in differentiation as found in other studies (144-149). Homozygous null mice for gp130 die mainly of cardiac defects due to the elimination of

9 cardiotrophin-1 signaling (3). In this mutant, the number of mononuclear cells in fetal liver was drastically reduced, as were the numbers of both early and mature CFUs. The thymuses were 50 percent smaller, consistent with other studies showing LIF’s role in hematopoiesis. 4. ROLE OF LIF IN CANCER Several tumour cell lines and neoplastic cells from various tissues produce LIF and express LIF receptors. However, the functional significance of either LIF or LIFR in human neoplasia is not fully understood. LIF can stimulate growth, induce differentiation, or trigger apoptotic cell death of various tumour cells (1,141,150,151) and data on the mechanisms controlling this diverse array of effects are scanty. Results of in vivo animal trials shed light on some of the possible roles of LIF in cancer and cancer metastasis. Cachexia (43,44), subcutaneous and abdominal fat loss, and elevated leukocyte and platelet counts often found in patients with metastatic cancer were induced by LIF in both mice and monkeys (46-48). In addition, at a high dose, LIF induced myelosclerosis whereas a low dose induced megakaryocytosis, reduced marrow cellularity and caused lymphopenia (48) suggesting a possible role for LIF in the pathogenesis of myeloproliferative disorders such as myelofibrosis and in marrow sclerosis. Furthermore, mice engrafted with FDS-P1 cells that produce high levels of LIF developed a fatal syndrome with cachexia, atrophy of liver and kidney, and excess bone formation with increased osteoblastic activity that resulted in metastatic-type calcifications (47) implying a role for LIF in bone tumours and neoplasms metastasizing to bone.

10

Ravandi and Estrov

Chapter 1

1. LIF and cancer metastasis

11

12

Ravandi and Estrov

Several in vitro studies were performed to delineate the effects of LIF on various tumours from different tissues. Though studies of cell lines often yielded conflicting results, experiments with fresh tissue confirmed LIF’s role in tumour growth, disease progression, and tumour metastasis (Table 3). 4.1 Hematological Malignancies LIF was originally characterized by virtue of its ability to induce differentiation in the murine myeloid leukemia cell line Ml, a property that it shares with IL-6 (13-15). However, LIF had no effect on the murine leukemia WEHI 3BD+ cell line that differentiates in response to IL-6 (150,192) whereas it stimulated the growth of the murine IL-3dependent DA-1 myeloid leukemia cell line (19,152). When injected into mice that had been implanted with T-22 cells, a subclone of the M1 cell line, it prolonged the animals’ survival by inducing differentiation (157). LIF is also produced by the THP-1 human monocytic leukemia cell line (58). LIF was found to be expressed in acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) cultured bone marrow stroma cells (155) and in human leukemia cell lines (140,154). Although LIF stimulated human normal marrow hematopoietic progenitor cell growth (139,140,193) and stroma-derived macrophage proliferation (194), it inhibited human leukemia cell growth (156,157). LIF also affects cells of the lymphoid lineage. T-cell clone (alloreactive) from lymphocytes rejecting kidney allografts and thymic epithelial cells (55) were found to produce LIF (reviewed in 1 & 2). Whereas normal human T lymphocytes did not bind radio-iodinated LIF (164), cells infected with human T-cell leukemia

Chapter 1 virus (HTLV)-I and –II expressed LIF (159) and proliferated in response to this cytokine (160). Similarly, various lymphoma cell lines were found to produce LIF (158), and LIF production was upregulated by IL-1 (66). Similar to IL-6, LIF plays a role in multiple myeloma (MM) cell proliferation. Human MM cell lines (161) and myeloma and plasmacytoma cells express LIF (162), LIFR, and the gp130 receptor subunit (163) and proliferate when exposed to LIF (163,164). Thus, similarly to IL-6 LIF may act as an autocrine growth factor for MM cells. The capability of LIF to induce both lytic and osteogenic effects in skeletal tissue, suggest that the osseous abnormalities typically found in MM are induced, among other factors, by LIF-producing myeloma cells. 4.2 Bone Tumours The effects of LIF on bone remodeling with LIF inducing both osteoclastic and osteoblastic activities suggest that LIFproducing tumour cells may significantly alter bone and skeletal tissue. Because the LIF gene was found to be mapped to chromosome 22q11-q12.2 (60), a question arose whether this site might be affected by chromosomal translocations that are related to tumours of neural-crest origin such as Ewing’s sarcoma and peripheral neuroepithelioma cytogenetically characterized by t(11;22)(q24;q12). It was found that the LIF gene is located far away from the Ewing’s sarcoma translocation (61,195). Nevertheless, bone tumours were found to produce high levels of LIF. Marusic et al. tested various rodent and human immortalized malignant bone tumour cell lines and found that LIF is constitutively expressed in several cell lines and is cytokine-inducible in others (165). LIF

1. LIF and cancer metastasis and LIFR were found in the cytoplasm of multinucleated giant tumour cells. Furthermore, LIF-stimulated giant tumour cells displayed osteoclast immunocytochemical features and resorbed large amounts of dentin (167,168). Additional indirect evidence for the role of LIF in bone tumours was provided by Gouin et al. who detected LIF in 34.7% of urine samples obtained from patients with a variety of bone tumours. They also found high LIF protein levels in supernatants of both neoplastic and benign bone tumour cells (166). Although LIF provides various bone tumours with a proliferation advantage and modulates their effects on bone tissue in either an autocrine or paracrine fashion, several studies showed that tumour cells that metastasize to bone may utilize similar mechanisms. 4.3 Breast Cancer Because LIF affects bone tissue and is produced by marrow stroma cells (86,155), several investigators asked whether LIF has a role in tumours such as breast cancer which metastasizes to this site (196). This was further emphasized by the study of Akatsu et al. who showed that the mouse mammary cell line MMT060562 produces LIF and supports osteoclast formation via a stroma celldependent pathway (197). Studies in breast cancer cell lines showed that some of these cells produce LIF, others express LIFR, and the cells may or may not respond to LIF. The diversity of cell lines and cell line clones that may have different features in different laboratories present a wide array of complex biological characteristics. For example, the estrogen-dependent breast cancer cell lines MCF-7 and T47-D do not produce LIF however their growth is stimulated by this cytokine (169-171).

13 MCF-7 cells bind LIF and, like several other breast cancer cell lines (172), express the gp130 subunit (169). In contrast, MDA-231 cells that express neither estrogen nor progesterone receptors produce LIF but their growth is not affected by this cytokine (170). Interestingly, progesterone treatment of MDA-231 cells co-transfected with both estrogen and progesterone induced the expression of LIF’s promoter (198). LIF also stimulated the estrogen-dependent T47D and the estrogen-independent SKBR3 and BT20 cell lines; inhibited, according to one study, MCF-7 cells (172), but had not effect on normal mammary epithelial cell growth (169,171). Interestingly, the SV40transformed mammary epithelium cell line HBL 100 was found to produce LIF (58). Breast cancer cells from 6 of 6 tumour samples expressed LIF transcripts (174) and widespread LIFR mRNA expression was found in primary breast tumours (172). Immunostaining of tumour samples obtained from 50 breast cancer patients detected LIF in 78% and LIFR in 80% of the samples. The presence of LIF correlated with a low S-phase fraction of the cell cycle and diploidy, whereas the presence of LIFR correlated with dipoidy, low S-phase fraction, and of estrogen receptor positivity. LIF and LIFR were also expressed in normal breast epithelium in 87% and 77% of the specimens, respectively (173). LIF stimulated colony formation of breast cancer cells obtained from five different patients in a dosedependent fashion (169) and the growth stimulation correlated with the presence of LIFR in these specimens (173). Taken together the data suggest a complex role of LIF and LIFR in breast cancer growth regulation. Because the bone marrow stroma produces LIF (155) and other cytokines such as stem cell

14

Ravandi and Estrov

factor that stimulate breast cell proliferation (169), cells that express LIFR and respond to these cytokines may have a growth advantage in the bone marrow microenvironment. 4.4 Kidney Cancer Renal carcinoma, like breast cancer, frequently metastasizes to bone. In addition, systemic symptoms, such as weight loss and fever, are common in kidney cancer and likely to result from overproduction of inflammatory cytokines. Moreover, the process of mouse nephrogenesis involves at least two distinct stages that can be blocked by LIF (199), and rat and human mesangial cells produce LIF and respond to this cytokine by transiently expressing the immediateearly genes c-fos, jun-B, and Egr-1 (200). These data suggest that LIF affects renal cell proliferation. Studies with cell lines have shown that both the primary kidney cancer line A-498 and the ACHN cell line established from pleural effusion of metastatic renal carcinoma produce LIF. Anti-LIF antibodies suppressed the cells’ growth and the inhibitory effect was reversed by exogenous LIF. These data suggest that the endogenously produced LIF stimulated kidney cancer cell line proliferation (170). 4.5 Prostate Cancer Prostate cancer cells selectively metastasize to the axial skeleton to produce osteolytic lesions. Laboratory data suggest that LIF plays a role in this disease. Paracrine-mediated growth factors may play a role in prostate cancer growth and development (201). In addition, IL-6, often expressed in parallel with LIF (202), was found to be expressed in prostate tissue (175) and might stimulate prostate cancer growth during

Chapter 1 disease progression (203). The hormoneindependent cancer cell lines TSU, PC-3 (204), and DU 145 (170) produce LIF and express gp130 (204). DU 145 cells did not proliferate in response to this cytokine (170,204) however, anti-LIF antibodies inhibited the cells’ growth (170). Thus, although only a few studies investigated the effect of LIF on prostate cancer cells and no data on binding of LIF to cellular LIFR are available, results from the above-described studies suggest that LIF plays a role in prostate cancer. 4.6 Malignant Melanoma In 1989, Mori et al. found that a factor produced by the melanoma cell line SEKI induced cachexia in tumour-bearing nude mice and inhibited lipoprotein lipase. This factor designated melanoma-derived lipoprotein lipase inhibitor was found to be identical to LIF (34,176). Subsequent studies found that LIF mRNA is expressed in various melanoma cell lines of which several produce the protein (58,177). Interestingly, oncostatin-M, another member of the IL-6 cytokine family, significantly increased LIF production by melanoma cells (205). LIF was detected in more than 60% of human melanoma samples and was found to enhance the expression of the intracellular adhesion molecule (ICAM)-1 in melanoma cells (177). Shedding of the soluble form of ICAM-1 from tumour cells impairs immune recognition and leads to tumour escape. Therefore, LIF may provide melanoma cells with a survival advantage. Furthermore, melanoma cells transfected with LIFR showed increased tumour growth suggesting that LIF may directly stimulate the growth of melanoma cells that express LIFR and provide them with a survival and growth advantage.

1. LIF and cancer metastasis 4.7 Hepatoma Only a few groups studied the effects of LIF in hepatoma. It was found that LIF is expressed in the HuH-7 and Hep-G2 hepatoma cell lines (179). LIF upregulated the expression of acute-phase proteins in the rat H-35 hepatoma cells (180) and activation of LIFR initiated signaling through the JAK pathway in Hep-G2 cells (181). 4.8 Gastrointestinal Malignancies The mRNA of LIF, LIFRß, and gp130 was detected in six stomach cancer, two colon cancer, one esophageal cancer, one gall bladder cancer, and seven pancreatic cancer cell lines (179). LIF induced apoptosis in the AZ-521 gastric and the GBK-1 gall bladder cancer cell lines and was detected in the MIA PACA pancreatic carcinoma cells (58). LIF did not affect the growth of either stomach or cancer cell lines; however, it stimulated the proliferation of two of seven pancreatic cancer cell lines (171). LIF is produced by the colon carcinoma cell lines SW948 and HRT18 (58). It has been shown to enhance human colon carcinoma HT24 cell proliferation suggesting that LIF facilitates the transition from ulcerative colitis to colon cancer (182). Because the results of cell line studies are inconsistent and since patient tumour tissue has not been studied yet, the biological significance of the cell line studies remains to be determined. 4.9 Central Nervous System Tumours Considering the variety of effects induced by LIF in the central nervous system (CNS), its involvement in CNS tumour growth is not surprising. LIF, LIFR, and the gp130 receptor subunit were detected in medulloblastoma tumour cells. Twelve of 12 tumour samples expressed LIF, and more than 90% of the

15 samples expressed LIFR and gp130. (185). In addition, LIF antisense inhibited medulloblastom cell proliferation (186). Taken together these data suggest that LIF acts as an autocrine growth factor in medulloblastoma. LIF was also studied in other CNS tumours. It either inhibited (183) or had no effect (184) on glioma cell lines. Meningioma cells expressed LIF transcripts; however, LIF did not affect the cells’ growth in vitro (206). 4.10 Other Neoplasms Several groups have reported LIF’s expression, production, and function in a variety of tumour cell lines. These studies implicate LIF’s role in the proliferation of neoplastic cells from several malignancies. Little is known about the role of LIF in tumours of the lung and the oral cavity. LIF is localized in the human airway mainly in fibroblasts, and IL-1ß can upregulate the expression of LIF’s mRNA and the release of LIF protein (207). LIF stimulated the growth of the metastatic human lung giant cell carcinoma PG cell line (187) and was found to be produced by the lung adenocarcinoma NCI-H23 cells (58) and the oral cavity carcinoma cell line OCC-1C (188). Because of LIF’s crucial role in the reproductive system, its effects on neoplasms originating from this system are of special interest. To our surprise, we were able to find only a limited number of studies addressing this issue. Bamberger et al. reported that LIF’s transcription is upregulated upon exposure of the SKUT1B uterine tumour cell line to a progesterone agonist (189). A soluble form of LIFR was detected in the supernatant of the choriocarcinoma cell line NJG, which also expressed LIF cDNA (190). Interestingly, human germ

16

Ravandi and Estrov

tumour cell lines express two forms of LIFR. Overexpression of LIFR of either form generated different levels of LIF protein activity, suggesting an autocrine role for LIF during germ cell tumourigenesis (191). LIF is also expressed in several other human tumour cell lines. Those include the bladder carcinoma line 5637, the epidermal carcinoma cell line HLFa (58) the squamous carcinoma line COLO-16 (29), and the SV40-transformed keratinocyte cell line SVK14 (58). The significance of these findings is yet to be determined.

5. CONCLUSION Similar to its diverse physiological effects, LIF exerts a broad spectrum of activities in various neoplastic cells, their surrounding tissues, and the cancer patient’s body as a whole. Although the accumulating data are incomplete and far from being conclusive, they indicate that LIF plays a major role in the pathophysiology of neoplasia. Tumours of bone, breast, kidney, the CNS, and other tissues, benefit from the presence of

Chapter 1 this cytokine. LIF, produced endogenously or by the tumours’ surrounding tissue, stimulates the cancer cells in an autocrine or paracrine fashion. In addition, LIF-producing tumour metastases, especially those metastasizing to bone, cause local distortion by inducing either blastic or lytic lessions. Moreover, overproduction of LIF is likely to be responsible for constitutional reactions such as an abnormal immune response; inflammatory and anti-inflammatory reactions, production of acute-phase proteins; abnormal responses of the hematopoietic system, including thrombocytosis; and neutrophilia and cachexia. Several groups worldwide have investigated the role LIF plays in normal physiology and in pathophysiology of cancer. These studies have revealed a wide array of complex effects that are not fully understood. Nevertheless, at least in a limited number of tumours, LIF appears to accelerate the cancerous process. Whether inhibition of LIF would be beneficial as an anticancer therapy remains to be seen.

References 1. Kurzrock R, Estrov Z, Wetzler M, Gutterman JU, Talpaz M. LIF: not just a leukemia inhibitory factor. Endocr Rev 1991;12(3):208-17 2. Taupin JL, Pitard V, Dechanet J, Miossec V, Gualde N, Moreau JF. Leukemia inhibitory factor: part of a large ingathering family. Int Rev Immunol 1998;16(3-4):397-26 3. Robinson RC, Grey LM, Staunton D. The crystal structure and biological function of leukemia inhibitory factor: implications for receptor binding. Cell 1994;77:1101-16 4. McDonald NQ, Panayotatos N, Hendrickson WA, Crystal structure of dimeric human ciliary neurotrophic factor determined by MAD phasing EMBO J 1995;14:2689-99 5. Somers W, Stahl M, Seehra JS. 1.9 A crystal structure of interleukin 6: implications for a novel mode of receptor dimerization and signaling. EMBO J 1997;16:989-97

6. Xu GY, Yu HA, Hong J. Solution structure of recombinant human interleukin-6. J Mol Biol 1997;268:468-81 7. Hinds MG, Maurer T, Zhang JG. Solution structure of leukemia inhibitory factor. J Biol Chem 1998;273:13738-45 8. Taga T. Gp130, a shared signal transducing receptor component for menatopoietic and neuropoietic cytokines. J. Neurochem 1996;67:1-10 9. Nakashima K, Taga T. gp130 and the IL-6 family of cytokines: signaling mechanisms and thrombopoirtic activities. Semin Hematol 1998;35:210-21 10. Ishikawa Y. Differentiation of a cell line of myeloid leukemia. J Cell Physiol 1969;74:22334 11. Tomida M, Yamamoto-Yamaguchi Y, Hozumi M. Purification of a factor inducing

1. LIF and cancer metastasis differentiation of mouse myeloid leukemic M1

cells from condititionedmedium of mouse fibro blast L929 cells. J Biol Chem 1984;259:10978-

82

12. Gearing DP, Gough NM, King JA, Hilton DJ, Nicola NA, Simpson RJ, Nice EC, Kelso A, Metcalf D. Molecular cloning and expression of a cDNA encoding a murine myeloid Leukemia Inhibitory Factor. EMBO J 1987;6:3995-02 13. Hilton DJ, Nicola NA, Metcalf D. Purification of a murine leukemia inhibitory factor from Krebs ascites cells. Anal Biochem 1988;173(2):359-67 14. Gough NM, Gearing DP, King JA, Willson TA, Hilton DJ, Nicola NA, et al. Molecular cloning and expression of the human homologue of the murine gene encoding myeloid leukemiainhibitory factor. Proc Natl Acad Sci U S A 1988;85(8):2623-27 15. Hilton DJ, Nicola NA, Gough NM, Metcalf D. Resolution and purification of three distinct factors produced by Krebs ascites cells which have differentiation-inducing activity on murine myeloid leukemic cell lines. J Biol Chem 1988;263(19):9238-43 16. Gearing DP, King JA, Gough NM. Complete sequence of murine leukemia inhibitory factor. Nuleic Acid Res 1988;16:9857 17. Ythier A, Abbud-Filho M, Williams JM, Loertscher R, Schuster MW, Nowill A, Hansen JA, Maltezos D, Strom T. Interleukin 2 dependent release of IL3 activity by T4+ human T-cell clones. Proc Natl Acad Sci USA 1985;82:7020-24 18. Moreau JF, Bonneville M, Peyrat MA, Jacques Y, Soulillou JP. Capacity of alloreactive human T cell clones to produce factor(s) inducing proliferation of the IL3-dependent DA-1 murine cell line. Ann Inst Pasteur Immunol 1986;137C:25-37 19. Moreau JF, Bonneville M, Godard A, Gascan H, Gruart V, Moore MA, et al. Characterization of a factor produced by human T cell clones exhibiting eosinophil-activating and burstpromoting activities. J Immunol 1987;138(11):3844-49 20. Godard A, Gascan H, Naulet J, Peyrat MA, Jacques Y, Soulillou JP, et al. Biochemical characterization and purification of HILDA, a human lymphokine active on eosinophils and bone marrow cells. Blood 1988;71(6):1618-23 21. Moreau JF, Donaldson DD, Bennett F, WitekGiannotti J, Clark SC, Wong GG. Leukaemia inhibitory factor is identical to the myeloid growth factor human interleukin for DA cells. Nature 1988;336(6200):690-92 22. Smith AG, Heath JK, Donaldson DD, Wong GG, Moreau JF, Stahl M, Rogers D. Inhibition of pluripotential embryonic stem cells

17 differentiation by purified polypeptides. Nature 1988;336:688-90 23. Gough NM, Gearing DP, King JA, Wilson TA, Hilton DJ, Nicola NA, Metcalf D. Molecular cloning and expression of the human homologue of the murine gene encoding myeloid Leukemia Inhibitory Factor. Proc Natl Acad Sci USA 1988;88:2623-27 24. Yamamoto-Yamaguchi Y, Tomida M, Hozumi M. Specific binding of a factor inducing differentiation to mouse myeloid leukemic M1 cells. Exp Cell Res 1986;164(1):97-02 25. Lowe DG, Nunes W, Bombara M, McCabe S, Ranges GE, Henzel W, et al. Genomic cloning and heterologous expression of human differentiation- stimulating factor. DNA 1989;8(5):351-59 26. Abe E, Tanaka H, Ishimi Y, Miyaura C, Hayashi T, Nagasawa H, et al. Differentiation-inducing factor purified from conditioned medium of mitogen-treated spleen cell cultures stimulates bone resorption. Proc Natl Acad Sci U S A 1986;83(16):5958-62 27. Smith AG, Hooper ML. Buffalo rat liver cells produce a diffusible activity which inhibits the differentiation of murine embryonal carcinoma and embryonic stem cells. Dev Biol 1987;121(1):1-9 28. Koopman P, Cotton RG. A factor produced by feeder cells which inhibits embryonal carcinoma cell differentiation. Characterization and partial purification. Exp Cell Res 1984;154(1):233-42 29. Baumann H, Wong GG. Hepatocyte-stimulating factor III shares structural and functional identity with leukemia-inhibitory factor. J Immunol 1989;143(4):1163-67 30. Mori M, Yamaguchi K, Abe K. Purification of a lipoprotein lipase-inhibitory protein produced by a melanoma cell line associated with cancer cachexia. Biochem Biophys Res Commun 1989;160:1085-92 31. Cornish J, Gallon K, King A, Edgar S, Reid IR. The effect of leukemia inhibitory factor on bone in vivo. Endocrinology 1993;132:1359-66 32. Yamamori T, Fukada K, Aebersold R, Korsching S, Fann MJ, Patterson PH. The cholinergic neuronal differentiation factor from heart cells is identical to leukemia inhibitory factor (published erratum appears in Science 1990 Jan 19;247(4940):271). Science 1989;246(4936):1412-16 33. Williams RL, Hilton DJ, Pease S, Willson TA, Stewart CL, Gearing DP, et al. Myeloid leukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature 1988;336(6200):684-87 34. Mori M, Yamaguchi K, Abe K. Purification of a lipoprotein lipase-inhibitory protein produced by

18

Ravandi and Estrov

a melanoma cell line associated with cancer cachexia. Biochem Biophys Res Commun 1989;160:1085-92 35. Gough NM, Williams RL. The pleiotropic actions of leukemia inhibitory factor. Cancer Cells 1989;1(3):77-80 36. Conquet F, Brulet P. Developmental expression of myeloid leukemia inhibitory factor gene in preimplantation blastocysts and in extraembryonic tissue of mouse embryos. Mol Cell Biol 1990;10(7):3801-05 37. Rathjen PD, Nichols J, Tom S, Edwards DR, Heath JK, Smith AG. Developmentally programmed induction of differentiation inhibiting activity and the control of stem cell populations. Genes Dev 1990;4(12B):2308-18 38. Evans MJ, Kaufman MH. Establishment in culture of pluripotential cells from mouse embryos. Nature 1981;292(5819): 154-56 39. Pease S, Williams RL. Formation of germ-line chimeras from embryonic stem cells maintained with recombinant leukemia inhibitory factor. Exp Cell Res 1990;190(2):209-11 40. Burgess AW. Life after LIF? Bioessays 1989;10(5):166-69 41. Shellard J, Perreau J, Brulet P. Role of leukemia inhibitory factor during mammalian development. Eur Cytokine Netw 1996;7(4):699-12 42. Murphy M, Reid K, Hilton DJ, Bartlett PF. Generation of sensory neurons is stimulated by leukemia inhibitory factor. Proc Natl Acad Sci U S A 1991;88(8):3498-01 43. Billingsley KG, Fraker DL, Strassmann G, Loeser C, Filot HM, Alexander HR. Macrophage-derived tumor necrosis factor and tumor-derived of leukemia inhibitory factor and interleukin-6: possible cellular mechanisms of cancer cachexia. Ann Surg Oncol 1996; 3:29-35 44. Mori M, Yamaguchi K, Honda S, Nagasaki K, Ueda M, Abe O, Abe K. Cancer cachexia syndrome developed in nude mice bearing melanoma cells producing leukemia-inhibitory factor. Cancer Res 1991; 51(24):6656-59 45. Ryffel B. Pathology induced by leukemia inhibitory factor. Int Rev Exp Pathol 1993; 34 PT B:69-72 46. Metcalf D, Nicola NA, Gearing DP. Effects of injected leukemia inhibitory factor on hematopoietic and other tissues in mice. Blood 1990;76(1):50-6 47. Metcalf D, Gearing DP. Fatal syndrome in mice engrafted with cells producing high levels of the leukemia inhibitory factor. Proc Natl Acad Sci U S A1989;86(15):5948-52 48. Metcalf D, Gearing DP. A myelosclerotic syndrome in mice engrafted with cells producing

Chapter 1 high levels of leukemia inhibitory factor (LIF). Leukemia 1989;3(12):847-52 49. Allan EH, Hilton DJ, Brown MA, Evely RS, Yumita S, Metcalf D, et al. Osteoblasts display receptors for and responses to leukemiainhibitory factor. J Cell Physiol 1990;145(1):110-29 50. Noda M, Vogel RL, Hasson DM, Rodan GA. Leukemia inhibitory factor suppresses proliferation, alkaline phosphatase activity, and type I collagen messenger ribonucleic acid level and enhances osteopontin mRNA level in murine osteoblast-like (MC3T3E1) cells. Endocrinology 1990;127(1):185-90 51. Rodan SB, Wesolowski G, Hilton DJ, Nicola NA, Rodan GA. Leukemia inhibitory factor binds with high affinity to preosteoblastic RCT1 cells and potentiates the retinoic acid induction of alkaline phosphatase. Endocrinology 1990;127(4):1602-08 52. Reid LR, Lowe C, Cornish J, Skinner SJ, Hilton DJ, Willson TA, et al. Leukemia inhibitory factor: a novel bone-active cytokine. Endocrinology 1990; 126(3): 1416-20 53. Wetzler M, Talpaz M, Lowe DG, Baiocchi G, Gutterman JU, Kurzrock R. Constitutive expression of leukemia inhibitory factor RNA by human bone marrow stromal cells and modulation by IL-1, TNF-alpha, and TGF-beta. Exp Hematol 1991;19(5):347-51 54. Anegon I, Moreau JF, Godard A, Jacques Y, Peyrat MA, Hallet MM, et al. Production of human interleukin for DA cells (HILDA)/leukemia inhibitory factor (LIF) by activated monocytes. Cell Immunol 1990;130(1):50-65 55. Le PT, Lazorick S, Whichard LP, Yang YC, Clark SC, Haynes BF, Singer KH. Human thymic epithelial cells produce IL-6, granulocyte-monocyte-CSF, and leukemia inhibitory factor. J Immunol 1990; 145:3310 56. Hilton DJ, Nicola NA, Metcalf D. Specific binding of murine leukemia inhibitory factor to normal and leukemic monocytic cells. Proc Natl Acad Sci USA 1988;85:5971-75 57. Baumann H, Jahreis GP, Saunder DN, Koj A. Human keratinocytes and monocytes release factors which regulate the synthesis of major acute phase plasma proteins in hepatic cells from man, rat, and mouse. J Biol Chem 1984;259:7331 58. Gascan H, Anegon I, Praloran V, Naulet J, Godard A, Soulillou JP, Jacques Y. Constitutive production of human interleukin for DA cells/leukemia inhibitory factor by human tumor cell lines derived from various tissues. J Immunol 1990;144:2592

1. LIF and cancer metastasis 59. Kola I, Davey A, Gough NM. Localization of the murine leukemia inhibitory factor gene near the centromere on chromosome 11. Growth Factors 1990;2(2-3):235-40 60. Sutherland GR, Baker E, Hyland VJ, Callen DF, Stahl J, Gough NM. The gene for human leukemia inhibitory factor (LIF) maps to 22q12. Leukemia 1989;3(1):9-13 61. Budarf M, Emanuel BS, Mohandas T, Goeddel DV, Lowe DG. Human differentiationstimulating factor (leukemia inhibitory factor, human interleukin DA) gene maps distal to the Ewing sarcoma breakpoint on 22q. Cytogenet Cell Genet 1989;52( 1-2): 19-22 62. Gough NM, Wilson TA, Stahl J, Brown MA. Molecular biology of the leukaemia inhibitory factor gene. Ciba Found Symp 1992;167:24-38 63. Gough NM. Molecular genetics of leukemia inhibitory factor (LIF) and its receptor. Growth Factors 1992;7(3):175-79 64. Tomida M, Yamamoto-Yamaguchi Y, Hozumi M. Purification of a factor inducing

differentiation of mouse myeloid leukemic M1

cells from conditioned medium of mouse

fibroblast L929 cells. J Biol Chem

1984;259(17):10978-82

65. Stahl J, Gearing DP, Willson TA, Brown MA, King JA, Gough NM. Structural organization of the genes for murine and human leukemia inhibitory factor. Evolutionary conservation of coding and non-coding regions. J Biol Chem 1990;265(15):8833-41 66. Gascan H, Godard A, Ferenz C, Naulet J, Praloran V, Peyrat MA, Hewick R, Jacques Y, Moreau JF, Soulillou JP. Characterization and amino-terminal aminoacid sequence of natural human interleukin of DA cells/leukemia inhibitory factor. J Biol Chem 1989;264:2150915 67. Rathjen PD, Toth S, Willis A, Heath JK, Smith AG. Differentiation inhibiting activity is produced in matrix-associated and diffusible forms that are generated by alternate promoter usage. Cell 1990;62(6):1105-14 68. Turnley AM, Bartlett PF. Cytokines that signal through the leukemia inhibitory factor receptor(beta) complex in the nervous system. J Neurochem, 2000;74:889-99 69. Gearing DP, Thut CJ, VandeBos T, Gimpel SD, Delaney PB, King J, Price V, Cosman D, Beckmann MP. Leukemia inhibitory factor receptor is structurally related to the IL-6 signal transducer, gp130. EMBO J 1991;10:2839-48 70. Gearing DP, Bruce AG. Oncostatin M binds the high-affinity leukemia inhibitory factor receptor. New Biol 1992;4:61-65 71. Davis S, Aldrich TH, Valenzuela DM, Wong VV, Furth ME, Squinto SP, Yancopoulos GD.

19 The receptor for ciliary neurotrophic factor. Science 1991;253:59-63 72. Stahl N, Yancopoulos GD. Thr tripartite CNTF receptor complex: activation and signalling involve components shared with other cytokines. J Neurobiol 1994;25:1454-66 73. Metcalf D, Hilton D, Nicola NA. Leukemia inhibitory factor can potentiate murine megakaryocyte production in vitro. Blood 1991;77(10):2150-53 74. Murray R, Lee F, Chiu CP. The genes for leukemia inhibitory factor and interleukin-6 are expressed in mouse blastocysts prior to the onset of hemopoiesis. Mol Cell Biol 1990;10(9):495356 75. Bartoe JL, Nathanson NM. Differential regulation of leukemia inhibitory factorstimulated neuronal gene expression by protein phosphatases SHP-1 and SHP-2 through mitogen-activated protein kinase-dependent and -independent pathways. J Neurochem 2000;74(5):2021-32 76. Zhong Z, Wen Z, Darnell JE Jr. Stat3: a STAT family member activated by tyrosine phosphorylation in response to epidermal growth factor and interleukin-6. Science 1994;264:95-98 77. Ware CB, Horowitz Md, Renshaw BR, Hunt JS, Liggitt D, Koblar SA, Gliniak BC, McKenna HJ, Papayannopoulou T, Thoma B, Cheng L, Donovan PJ, Peschon JJ, Bartlett PF, Willis CR, Wright BD, Carpenter MK, Davison BL, Gearing DP. Targeted disruption of the lowaffinity leukemia inhibitory factor receptor gene causes placental, skeletal, neural and metabolic defects and results in perinatal death. Development 1995;121:1283-99 78. Heath JK, Smith AG, Hsu LW, Rathjen PD. Growth and differentiation factors of pluripotential stem cells. J Cell Sci Suppl 1990;13:75-85 79. Vogiagis D, Salamonsen LA. Review: The role of leukaemia inhibitory factor in the establishment of pregnancy. J Endocrinol 1999; 160(2): 181-90 80. Escary JL, Perreau J, Dumenil D, Ezine S, Brulet P. Leukaemia inhibitory factor is necessary for maintenance of haematopoietic stem cells and thymocyte stimulation. Nature 1993; 363:361-64 81. Stewart CL, Kaspar P, Brunet LJ, Bhatt H, Gadi I, Kontgen F, Abbondanzo SJ. Blastocyst implantation depends on maternal expression of leukaemia inhibitory factor. Nature 1992;359:76-79 82. Yue ZP, Yang ZM, Wei P, Li SJ, Wang HB, Tan JH, Harper MJ. Leukemia inhibitory factor, leukemia inhibitory factor receptro, and glycoprotein 130 in rhesus monkey uterus during

20

Ravandi and Estrov

menstrual cycle and early pregnancy. Biol Reprod 2000;63(2);508-12 83. Passavant C, Zhao X, Das SK, Dey SK, Mead RA. Changes in uterine expression of leukemia inhibitory factor receptor gene during pregnancy and its Up-regulation by prolactin in the western spotted skunk. Biol Reprod 2000;63(1):301-07 84. Piccinni MP, Maggi E, Romagnani S. Role of hormone-controlled T-cell cytokines in the maintenance of pregnancy . Biochem Soc Trans 2000;28(2):212-15 85. Hsieh YY, Tsai HD, Chang CC, Hsu LW, Chang SC, Lo HY. Prolonged culture of human cryopreserved embryos with recombinant human leukemia inhibitory factor. J Assist Reprod Genet 2000;17(3):l31-34 86. Martin TJ, Allan EH, Evely RS, Reid IR. Leukaemia inhibitory factor and bone cell function. Ciba Foundation Symposium 1992;167:141-50 87. Cornish J, Gallon K, King A, Edgar S, Reid IR. The effect of leukemia inhibitory factor on bone in vivo. Endocrinology. 1993;132(3):1359-66 88. Malaval L, Gupta AK, Aubin JE. Leukemia inhibitory factor inhibits osteogenic differentiation in rat calvaria cell cultures. Endocrinology. 1995;136(4): 1411-18 89. Chandrasekhar S, Harvey AK. Modulation of PDGF mediated osteoblast chemmotaxis by leukemia inhibitory factor (LIF). J Cell Physiol 1996;169(3):481-90 90. Sabokbar A, Fujikawa Y, Brett J, Murray DW, Athanasou NA. Increased osteoclastic differentiation by PMMA particle-associated macrophages. Inhibitory effect by interleukin 4 and leukemia inhibitory factor. Acta Orthopaedica Scandinavica 1996;67(6): 593-98 91. Malaval L, Gupta AK, Liu F, Delmas PD, Aubin JE. LIF, but IL-6, regulates osteoprogenitor differentiation in rat calvaria cell cultures: modulation by dexamethasone. J Bone Miner Res 1998;13(2):175-84 92. Bohic S, Rohanizadeh R, Touchais S, Godard A, Daculsi G, Heymann D. Leukemia inhibitory factor and oncostatin M influence the mineral phases formed in a murine heterotopic calcification model: a Fourier transform-infrared microspectroscopic study. J Bone Miner Res 1998;13(10):1619-32 93. Thomson BM, Saklatvala J, Chambers TJ. Osteoblasts mediate interleukin 1 stimulation of bone resorption by rat osteoclasts. J Exp Med 1986;164(1):104-12 94. Thomson BM, Mundy GR, Chambers TJ. Tumor necrosis factors alpha and beta induce osteoblastic cells to stimulate osteoclastic bone resorption. J Immunol 1987;138(3):775-79

Chapter 1 95. Beutler B, Cerami A. Cachectin: more than a tumor necrosis factor. N Engl J Med 1987;316(7):379-85 96. Beutler BA, Cerami A. Recombinant interleukin 1 suppresses lipoprotein lipase activity in 3T3L1 cells. J Immunol 1985;135(6):3969-71 97. Metcalf D. Disease states induced by hemopoietic growth factor excess: their implications in medicine. Int J Cell Cloning 1990;8 Suppl 1:374-87; discussion 387-90 98. Gadient RA, Patterson PH. Leukemia inhibitory factor, Interleukin 6, and other cytokines using the GP130 transducing receptor: roles in inflammation and injury. Stem Cells 1999;17(3):127-37 99. Brown MA, Metcalf D, Gough NM. Leukaemia inhibitory factor and interleukin 6 are expressed at very low levels in the normal adult mouse and are induced by inflammation. Cytokine 1994:6:300-09 100. Block MI, Berg M, McNamara MJ. Passive immunization of mice against D factor blocks lethality and cytokine release during endotoxemia. J Exp Med 1993;178:1085-90 101. Alexander HR, Wong GG, Doherty GM. Differentiation factor/leukemia inhibitory factor protection against lethal endotoxemia inmice: synergistic effect with interleukin 1 and tumor necrosis factor. J Exp Med 1992; 175:1139-42 102. Waring PM, Waring LJ, Billington T. Leukemia inhibitory factor protects against experimental lethal Escherichia coli septic shock in mice. Proc Natl Acad Sci USA 1995;92:133741 103. Carroll G, Bell M, Wang H. Antagonism of the IL-6 cytokine subfamily—a potential strategy for more effective therapy in rheumatoid arthritis. Inflamm Res 1998;47:1-7 104. Hui W, Bell M, Carroll G. Oncostatin M (OSM) stimulates resorption and inhibits synthesis of proteoglycan in porcine articular cartilage explants. Cytokine 1996;8:495-00 105. Lotz M, Moats T, Villiger PM. Leukeamia inhibitory factor is expressed in cartilage and synovium and can contribute to the pathogenesis of arthritis. J Clin Invest 1992;90:888-96 106. Okamoto H, Yamamura M, Morita Y. The synovial expression and serum levels of interleukin-6, interleukin-11, leukemia inhibitory factor, and oncostatin M in rheumatoid arthritis. Arthritis Rheum 1997;40:1096-05 107. Waring PM, Carroll GJ, Kandiah DA. Increased levels of leukemia inhibitory factory in synovial fluid from patients with rheumatoid arthritis and other inflammatory arthritides. Arthritis Rheum 1993;36:911-15 108. Henrotin YE, De GD, Labasse AH. Effects of exogenous IL-1 beta, TNF alpha, IL-6, IL-8

1. LIF and cancer metastasis and LIF on cytokine production by human articular chondrocytes. Osteoarth Cart 1996;4:163-73 109. Chabaud M, Fossiez F, Taupin JL. Enhancing effect of IL-17 on IL-1-induced IL-6 and leukemia inhibitory factor production by rheumatoid arthritis synoviocytes and its regulation by Th2 cytokines. J Immunol 1998;161:409-14 Bell M, Carroll GJ, Chapman H. Leukemia 110. inhibitory factor (LIF) binding protein attenuates the phlogistic and abolishes the chondral effects of LIF in goat joints. J Rheumatol 1997;24:2394-02 111. Heymann D, Her E, Nguyen JM. Leukaemia inhibitory factor (LIF) production in pleural effusions: comparison with production of IL-4, IL-8, IL-10 and macrophage-colony stimulating factor (M-CSF). Cytokine 1996;410-16 112. Jorens PG, De JR, Bossaert LL. High levels of leukaemia inhibitory factor in ARDS. Cytokine 1996;8:873-76 Szepietowski JC, McKenzie RC, Keohane 113. SG. Leukaemia inhibitory factor: induction in the early phase of allergic contact dermatitis. Contact Dermatitis 1997; 36:21-25 114. McKenzie RC, Paglia D, Kondo S. A novel endogenous mediator or cutaneous inflammation: leukemia inhibitory factor. Acta Derm Venerol 1996;76:111-14 115. Thompson SW, Dray A, Urban L. Leukemia inhibitory factor induces mechanical allodynia but not thermal hyperalgesia in the juvenile rat. Neuroscience 1996;71:1091-94 116. Banner LR, Patterson pH, Allchorne A. Leukemia inhibitory factor is an antiinflammatory and analgesic cytokine. J Neurosci 1998;18:5456-62 117. Banner LR, Patterson PH. Major changes in the expression of the mRNAs for cholinergic differentiation factor/leukemia inhibitory factor and its receptor after injury to adult peripheral nerves and ganlia. Proc Natl Acad Sci USA 1994;91:7109-13 118. Banner LR, Moayeri NN, Patterson PH. Leukemia inhibitory factory is expressed in astrocytes following corticla brain injury. Exp Neurol 1997;147:1-9 119. Curtis R, Scherer SS, Somogyi R. Retrograde axonal trnsport of LIF is increased by peripheral nerve injury: correlation with increased LIF expression in distal nerve. Neuron 1994;12:191-04 120. Kurek JB, Ausin L, Cheema SS. Upregulation of leukaemia inhibitory factor and interleukin-6 in transected sciatic nerve and musce following denervation. Neuromuscul Disord 1996;6:105-14

21 121.

Kurek JB, Austin L, Cheema SS, Bartlett PF, Murphy M. Upregulation of leukaemia inhibitory factor and interleukin-6 in transected sciatic nerve and denervated muscle. Neuromusc Disor 1996;6:105-14 122. Austin L, Burgess AW. Stimulation of myoblast proliferation in culture by leukaemia inhibitory factor. J Neurol Sci 1991;101:193-97 123. Austin L, Bower J, Kurek J, Vakakis N. Effects of leukaemia inhibitory factor and other cytokines on murine and human myoblast proliferation. J Neurol Sci 1992;112:185-91 124. Patterson PH, Kou S-Y, Sugiura S. LIF coordinates neuronal and inflammatory response to nerve injury. Soc Neurosci Abstr 1997;23:393-33 125. Baumann H, Jahreis GP, Sauder DN, Koj A. Human keratinocytes and monocytes release factors which regulate the synthesis of major acute phase plasma proteins in hepatic cells from man, rat, and mouse. J Biol Chem 1984;259(11):7331-42 126. Kordula T, Rokita H, Koj A. Effects of interleukin-6 and leukemia inhibitory factor on the acute phase response and DNA synthesis in cultured rat hepatocytes. Lymphokine Cytokine Res 1991:10:23-26 127. Mayer P, Geissler K, Ward M. Recombinant human leukemia inhibitory factor induces acute phase proteins and raises the blood platelet counts in nonhuman primates. Blood 1993;81:3226-33 128. Richards LJ, Kilpatrick TJ, Dutton R, Tan SS, Gearing DP, Bartlett PF, Murphy M. Leukaemia inhibitory factor (LIF) and related factors promote the differentiation of neuronal and astrycytic precursors within the developing murine spinal cord. Eur J Neurosci 1996;8:29199 129. Li M, Sendtner M, Smith A. Essential function of LIF receptor in motor neurones. Nature 1995;378:724-27 130. Cheema SS, Richards LR, Murphy M, Bartlett PF. Leukemia inhibitory factor rescues motorneurones from axotomy-induced cell death. Neuroreport 1994;5:989-92 131. Ikeda K, Iwasaki Y, Shiojima T, Kinoshita M. Neuroprotective effect of cholinergic differentiation leukaemia inhibitory factor on Wobbler murine motor neurone disease. Muscle Nerve 1995:18:1344-47 132. Metcalf D. The induction and inhibition of differentiation in normal and leukaemic cells. Philos Trans R Soc Lond B Biol Sci 1990;327(1239):99-109 133. Shabo Y, Lotem J, Rubinstein M, Revel M, Clark SC, Wolf SF, et al. The myeloid blood cell

22

Ravandi and Estrov

differentiation-inducing protein MGI-2A is interleukin-6. Blood 1988;72(6):2070-73 134. Metcalf D. Suppression of myeloid leukemic cells by normal hemopoietic regulators. Recent Prog Cytokine Res 1989;10:11-19 135. Maekawa T, Metcalf D. Clonal suppression of HL60 and U937 cells by recombinant human leukemia inhibitory factor in combination with GM-CSF or G-CSF. Leukemia 1989;3(4):270-76 136. Maekawa T, Metcalf D, Gearing DP. Enhanced suppression of human myeloid leukemic cell lines by combinations of IL-6, LIF, GM-CSF and G-CSF. Int J Cancer 1990;45(2):353-58 137. Wang C, Lishner M, Minden MD, McCulloch EA. The effects of leukemia inhibitory factor (LIF) on the blast stem cells of acute myeloblastic leukemia. Leukemia 1990;4(8):548-52 138. Metcalf D, Hilton DJ, Nicola NA. Clonal analysis of the actions of the murine leukemia inhibitory factor on leukemic and normal murine hemopoietic cells. Leukemia 1988;2(4):216-21 139. Leary AG, Wong GG, Clark SC, Smith AG, Ogawa M. Leukemia inhibitory factor differentiation-inhibiting activity/human interleukin for DA cells augments proliferation of human hematopoietic stem cells. Blood 1990;75(10):1960-64 140. Verfaillie C, McGlave P. Leukemia inhibitory factor/human interleukin for DA cells: a growth factor that stimulates the in vitro development of multipotential human hematopoietic progenitors. Blood 1991;77(2):263-70 141. Estrov Z, Talpaz M, Wetzler M, Kurzrock R. The modulatory hematopoietic activities of leukemia inhibitory factor. Leuk Lymphoma 1992;8(1-2):1-7 142. Wetzler M, Kurzrock R, Lowe DG, Kantarjian H, Gutterman JU, Talpaz M. Alteration in bone marrow adherent layer growth factor expression: a novel mechanism of chronic myelogenous leukemia progression. Blood 1991;78(9):2400-06 143. Shen MM, Skoda RC, Cardiff RD, CamposTorres J, Leder P, Orntiz DM. Expression of LIF in transgenic mice results in altered thymic epitehlium and apparent interconversion of thymic and lymph node morphologies. EMBO J 1994;1375-85 144. Schaafsma MR, Falkenburg JH, Duinkerken N, Moreau JF, Soulillou JP, Willemze R, Fibbe WE. Human interleukin for DA cells (HILDA) does not affect the proliferation and differentiation of hematopoietic progenitor cells in human long-term bone marrow cultures. Exp Hematol 1992;20:6-10

Chapter 1 145.

Firkin FC, Birner R, Farag S. Differential action of diffusible molecules in long-term bone marrow culture on proliferation of leukaemic and normal haemopoietic cells. Br J Haematol 1993;84:8-15 146. Debili N, Masse JM, Katz A, Guichard J, Breton-Gorius J, Vainchenker W. Effects of the recombinant hematopoietic growth factors intermegakayocytic differentiation of CD34+ cells. Blood 1993;82:84-95 147. Gabutti V, Timeus F, Ramenghi U, Crescenzio N, Marranca D, Miniero R, Cornaglia G, Bagnara GP. Expansion of cord blood progenitors and use for hemopoietic reconstitution. Stem Cells 1993;11 Suppl: 105-12 148. Szilvassy SJ, Cory S. Efficient retroviral gene transfer to purified long-term repopulating hematopoietic stem cells. Blood 1994;84:74-83 149. Szilvassy SJ, Weller K.P, Kin W, Sharma AK, Ho AS, Tsukamoto A, Hoffman R, Leiby KR, Gearing DP. Leukemia inhibitory factor upregulates cytokine expression by a murine stromal cell line enabling the maintenance of highly enriched competitive repopulating stem cells. Blood 1996;87:4618-28 150. Metcalf D. Leukemia inhibitory factor – a puzzling poly-functional regulator. Growth Factors 1992; 7:169-73 151. Kamohara H, Sakamoto K, Ishiko T, Masuda Y, Abe T, Ogawa M. Leukemia inhibitory factor induces apoptosis and proliferation of human carcinoma cells through different oncogene pathways. Int J Cancer 1997;72(4):687-95 152. Moreau JF, Bonneville M, Peyart MA, Jacques Y, Soulillou JP. Capactiy of alloreactive human T clones to produce factor(s) inducing proliferation of the IL3-dependent DA-1 murine cell line. I. Evidence that this production is under IL2 control. Ann Inst Pasteur Immunol 1986; 137C(1):25-37 153. Abe T, Murakami M, Sato T, Kajiki M, Ohno M, Kodaira R. Macrophage differentiation inducing factor from human monocytic cells is equivalent to murine leukemia inhibitory factor. J Biol Chem 1989;264:8941 154. Xie P, Chan FS, Ip NY, Leung MF. Induction of gp130 and LIF by differentiation inducers in human myeloid leukemia K562 cells. Leuk Res 1999;23(12):1113-19 155. Wetzler M, Estrov Z, Talpaz M, Kim KJ, Alphonso M, Srinivasan R, Kurzrock R. Leukemia inhibitory factor in long-term adherent layer cultures: increased levels of bioactive protein in leukemia and modulation by IL-4, ILand Cancer Res 1994; 54:1837-42 156. Maekawa T, Metcalf D, Gearing DP. Enhanced suppression of human myeloid

1. LIF and cancer metastasis leukemic cell lines by combinations of IL-6,

LIF, GM-CSF and G-CSF. Int J Cancer

1990;45(2):353-58

Yamamoto-Yamaguchi Y, Tomida M, 157. Hozumi M. Prolongation by differentiationstimulating factor/leukemia inhibitory factor of the survival time of mice implanted with mouse myeloid leukemia cells. Leuk Res 1992;16(10):1025-29 Godard A, Heymann D, Raher S, Anegon I, 158. Peyrat MA, Le Mauff B, Mouray E, Gregoire M, Virdee K, Soulillou JP et al. High and low affinity receptors for human interleukin for DA cells/leukemia inhibitory factor on human cells. Molecular characterization and cellular distribution. J Biol Chem 1992;267(36):3214-22 Umemiya-Okada T, Natazuka T, Matsui T, 159. Ito M, Taniguchi T, Nakao Y. Expression and regulation of the leukemia inhibitory factor/D factor gene in human T-cell leukemia virus type 1 infected T-cell lines. Cancer Res 1992;52(24):6961-65 Lal RB, Rudolph D, Buckner C, Pardi D, 160. Hooper WC. Infection with human Tlymphotropic viruses leads to constitutive expression of leukemia inhibitory factor and interleukin-6. Blood 1993;81(7):1827-32 Gu ZJ, Zhang XG, Hallet MM, Lu ZY, 161. Wijdenes J, Rossi JF, Klein B. A ciliary neurotrophic factor-sensitive human myeloma cell line. Exp Hematol 1996; 24:1195-00 Portier M, Zhang XG, Ursule E, Lees D, 162. Jourdan M, Bataille R, Klein B. Cytokine gene expression in human multiple myeloma. Br J Haematol 1993; 85:514-20 Nishimoto N, Ogata A, Shima Y, Tani Y, 163. Ogawa H, Nakagawa M, Sugiyama H, Yoshizaki K, Kishimoto T. Oncostain M, leukemia inhibitory factor, and interleukin 6 induce the proliferation of human plasmacytoma cells via the common signal transducer, gp130. J Exp Med 1994; 179(4): 1343-47 Zhang XG, Gu JJ, Lu ZY, Yasukawa K, 164. Yancopoulos GD, Turner K, Shoyab M, Taga T, Kishimoto T, Bataille R. et al. Ciliary neurotropic factor, interleukin 11, leukemia inhibitory factor, and oncostatin M are growth factors for human myeloma cell lines using the interleukin 6 signal transducer gp130. Institute for Molecular Genetics, CNRS BP5051, Montepellier, France. Marusic A, Kalinowski JF, Jastrzebski S, 165. Lorenzo JA. Production of leukemia inhibitory factor mRNA and protein by malignant and immortalized bone cells. J Bone Miner Res 1993;8(5):617-24 Gouin F, Heymann D, Raher S, De Groote 166. D, Passuti N, Daculsi G, Godard A. Increased

23 levels of leukaemia inhibitory factor (LIF) in urine and tissue culture supernatant from human primary bone tumors. Cytokine 1998;10(2):11014 167. Soueidan A, Gan OL, Gouin F, Godard A, Heymann D, Jacques Y, Daculsi G. Culturing of cells from giant cell tumour of bone on natural and synthetic calcified substrata: the effect of leukaemia inhibitory factor and vitamin D3 on the resorbing activity of osteoclast-like cells. Virchows Archiv 1995;426(5):469-77 168. Gouin F, Couillaud S, Cottrel M, Godard A, Passuti N, Heymann D. Presence of leukaemia inhibitory factor (LIF) and LIF-receptor chain (gp190) in osteoclast-like cells cultured from human giant cell tumour of bone. Ultrastructural distribution. Cytokine 1999;11(4):282-89 169. Estrov Z, Samal B, Lapushin R, Kellokumpu-Lehitinen P, Sahin AA, Kurzrock R, Talpaz M, Aggarwal BB. Leukemia inhibitory factor binds to human breast cancer cells and stimulates their proliferation. J Interferon Cytokine Res 1995;15(10):905-13 170. Kellokumpu-Lehtinen P, Talpaz M, Harris D, Van Q, Kurzrock R, Estrov Z. Leukemiainhibitory factor stimulates breast, kidney and prostate cancer cell proliferation by paracrine and autocrine pathways. Int J Cancer 1996;66(4):515-19 Liu J, Hadjokas N, Mosley B, Estrov Z, 171. Spence MJ, Vestal RE. Oncostatin M-spccific receptor expression and function in regulating cell proliferation of normal and malignant mammary epithelial. Cytokine 1998;10(4):29502 172. Douglas AM, Goss GA, Sutherland RL, Hilton DJ, Berndt MC, Nicola NA, Begley CG. Expression and function of members of the cytokine receptor superfamily on breast cancer cells. Oncogene 1997;14(6):661-69 Dhingra K, Sahin A, Emami K, Hortobagyi 173. GN, Estrov Z. Expression of leukemia inhibitory factor and its receptor in breast cancer: a potential autocrine and paracrine growth regulatory mechanism. Breast Cancer Res Treat 1998;48(2):165-74 174. Crichton MB, Nichols JE, Zhao Y, Bulun SE, Simpson ER. Expression of transcripts of interleukin-6 and related cytokines by human breast tumors, breast cancer cells, and adipose stromal cells. Cell Endocrinol 1996;120(2):215 175. Tatoud R, Desgrandchamps F, Gegeorges A, Thomas F. Peptide growth factors in the prostate. Pathol Biol 1993 ;41:731-40 176. Mori M, Yamaguchi K, Honda S, Nagasaki K, Ueda M, Abe O, Abe K. Cancer cachexia syndrome developed in nude mice bearing

24

Ravandi and Estrov

melanoma cells producing leukemia-inhibitory factor. Cancer Res 1991;51(24):6656-59 177. Paglia D, Oran A, Lu C, Kerbel RS, Sauder DN, McKenzie RC. Expression of leukemia inhibitory factor and interleukin-11 by human melanoma cell lines: LIF, IL-6, and IL-11 are not coregulated. J Interferon Cytokine Res 1995;15(5):455-60 178. Heymann D, Godard A, Raher S, Ringeard S, Lassort D, Blanchard F, Harb J. Human interleukin for DA cells/leukemia inhibitory factor and oncostatin M enhance membrane expression of intercellular adhesion molecule-1 on melanoma cells but not the shedding of its soluble form. Cytokine 1995;7(2):111 179. Kamohara H, Sakamoto K, Ishiko T, Mita S, Masuda Y, Abe T, Ogawa M. Human carcinoma cell lines produce biologically active leukemia inhibitory factor (LIF). Res Commun Mol Pathol Pharmacol 1994 180. Baumann H, Ziegler SF, Mosley B, Morella KK, Pajovic S, Gearing DP. Reconstitution of the response to leukemia inhibitory factor, oncostatin M, and ciliary neurotrophic factor in hepatoma cells. J Biol Chem 1993;268(12):8414-17 181. Hermanns HM, Radtke S, Haan C, SchmitzVan de Leur H, Tavernier J, Heinrich PC, Behrmann I. Contributions of leukemia inhibitory factor receptor and oncostatin M receptor to signal transduction in heterodimeric complexes with glycoprotein 130. J Immunol 1999;163(12):6651-58 182. Guimbaud R, Abitbol V, Bertrand V, Quartier G, Chauvelot-Moachon L, Giroud J, Couturier D, Chaussade DC. Leukemia inhibitory factor involvement in human ulcerative colitis and its potential role in malignant course. Eur Cytokine Netw 1998;9(4):607-12 183. Halfter H, Kremerskothen J, Weber J, Hacker-Klom U, Barnekow A, Ringelstein EB, Stogbauer F. Growth inhibition of newly established human glioma cell lines by leukemia inhibitory factor. J Neuro Oncol 1998, 39(1):118 184. Halfter H, Lotfi R, Westermann R, Young P, Ringelstein EB, Stogbauer FT. Inhibition of growth and induction of differentiation of glioma cell lines by oncostatin M (OSM). Growth Factors 1998;15(2):135-47 185. Liu J, Li H, Moreau JF. Expression of LIF as autocrinal growth factor in human medulloblastomas. Chung Hua Ping Li Hsueh Tsa Chih 1996;25(l):24-26 186. Liu J, Li H, Hamou MF. Inhibitory effect of antisense LIF oligonucleotide on the outgrowth

Chapter 1 of human medulloblastoma cells. Chung Hua Ping Li Hsueh Tsa Chih 1996;25(3): 132-34 187. Fu J, Zheng J, Fang W, Wu B. Effect of interleukin-6 on the growth of human lung cancer cell line. Chin Med J 1998;111(3):265-68 188. Kajimura N, Iseki H, Tanaka R, Ohue C, Otsubo K, Gyoutoku M, Sasaki K, Akiyama Y, Yamaguchi K. Toxohormones responsible for cancer cachexia syndrome in nude mice bearing human cancer cell lines. Cancer Chemother Pharmacol 1996;38 Suppl:S48-52 189. Bamberger AM, Jenatschke S, Erdmann I, Schulte HM. Progestin-dependent stimulation of the human leukemia inhibitory factor promoter in SKUT-1B uterine tumor cells. J Reprod Immunol 1997;33(3):189-01 190. Tomida M. Presence of mRNAs encoding the soluble D-factor/LIF receptor in human choriocarcinoma cells and production of the soluble receptor. Biochem Biophys Res Commun 1997;232(2):427-31 191. Voyle RB, Haines BP, Pera MF, Forrest R, Rathjen PD. Human germ cell tumor cell lines express novel leukemia inhibitory factor transcripts encoding differentially localized proteins. Exp Cell Res 1999;249(2): 199-11 192. Shabo Y, Lotem J, Rubinstein M, Revel M, Clark SC, Wolf SF, Kamen R, Sachs L. The myeloid blood cell differentiation-inducing protein MG1-2A is interleukin-6. Blood 1988; 72:2070-73 193. Keller JR, Gooya JM, Ruscetti FW. Direct synergistic effects of leukemia inhibitory factor on hematopoietic progenitor cell growth: comparison with other hematopoietins that use the gp130 receptor subunit. Blood 1996;88(3):863-69 194. Heymann D, Gouin F, Guicheux J, Munevar JC, Godard A, Daculsi G. Upmodulation of multinucleated cell formation in long-term human bone marrow cultures by leukaemia inhibitory factor. Cytokine 1997;9(l):46-52 195. Zucman J, Delattre O, Desmaze C, Plougastel B, Joubert I, Melot T, Peter M, De Jong P, Rouleau G, Aurias A. et al. Cloning and characterization of the Ewing’s sarcoma and peripheral neuroepithelioma t(11;22) translocation breakpoints. Genes Chromosome Cancer 1992;5(4):271-77 196. Mansi JL, Berger U, McDonnell T, Pople A, Rayter Z, Gazet JC, Coombes RC. The fate of bone marrow micrometastases in patients with primary breast cancer. J Clin Oncol 1989;7:44549 197. Akatsu T, Ono K, Katayama Y, Tamura T, Nishikawa M, Kugai N, Yamamoto M, Nagata N. The mouse mammary tumor cell line, MMT060562, produces prostaglandin E2 and

1. LIF and cancer metastasis leukemia inhibitory factor and supports osteoclast formation in vitro via a stromal celldependent pathway. J Bone Miner Res 1998;13(3):400-08 198. Bamberger AM, Thuneke I, Schulte HM. Differential regulation of the human ‘leukemia inhibitory factor’ (LIF) promoter in T47D and MDA-MB231 breast cancer cells. Breast Cancer Res Treat 1998;47(2):153-61 Bard JB, Ross AS. LIF, the ES-cell 199. inhibition factor, reversibly blocks nephrogenesis in cultured mouse kidney rudiments. Development 1991;113(l):193-98 Hartner A, Sterzel BR, Reindi N, Hocke 200. GM, Fey GH, Gopplet-Struebe M. Cytokineinduced expression of leukemia inhibitory factor in renal mesangial cells. Kidney Int 1994;45:1562-71 Gleave ME, Hsieh JT, von Eschenbach AC, 201. Chung LW. Prostate and bone fibroblasts induce human prostate cancer growth in vivo: implications for bidirectional tumor-stromal cell implication in prostate carcinoma growth and metastasis. J Urol 1992;147:1151-59 202. Hirano T, Matsuda T, Nakajima K. Signal transduction through gp130 that is shared among the receptors for the interleukin 6 related subfamily. Stem Cells 1994;12:262-77 Chung TD, Yu JJ, Spiotto MY, Bartkowski 203. M, Simons JW. Characterization of the role of IL-6 in the progression of prostate cancer. Prostate 1999;38:199-207 Mori S, Murakami-Mori K, Bonavida B. 204. Oncostatin M (OM) promotes the growth of DU 145 human prostate cancer cells, but not PC-3 or LNCaP, through the signaling of the OM

25 specific receptor. Anticancer Res 1999;19:10115 205. Heymann D, Blanchard F, Raher S, De Groote D, Godard A. Modulation of LIF expression in human melanoma cells by oncostatin M. Immunol Lett 1995;46(3):245-51 206. Schrell UM, Koch HU, Marschalek R, Schrauzer T, Anders M, Adams E, Fahlbusch R. Formation of autocrine loops in human cerebral meningioma tissue by leukemia inhibitor factor, interleukin-6, and oncostatin M: inhibition of meningioma cell growth in vitro by recombinant oncostatin M. J Neurosurg 1998,88(3):541-48 207. Knight DA, Lydell CP, Zhou D, Weir TD, Schellenberg R, Bai TR. Leukemia inhibitory factor (LIF) and LIF receptor in human lung. Distribution and regulation of LIF release. Am J Respir Cell Mol Biol 1999;20(4);834-41 208. Freidin M, Kessler JA. Cytokine regulation of substance P expression in sympathetic neurons. Proc Natl Acad Sci U S A 1991;88(8):3200-03 209. Murphy M, Reid K, Hilton DJ, Bartlett PF. Generation of sensory neurones is stimulated by leukaemia inhibitory factor. Proc Natl Acad Sci USA 1991;88:3498-01 210. Yamakuni H, Minami M, Satoh M. Localisation of mRNA for leukaemia inhibitory factor receptor in the adult brain. J Neuroimmunol 1996;70:45-43 211. Cheema SS, Richards L, Murphy M, Bartlett PF. Leukemia inhibitory factor prevents the death of axotomised sensory neurones in the dorsal root ganglion of the neonatal rat. J Neurosci Res 1994;37:213-18

This page intentionally left blank

Chapter 2 INTERLEUKIN-2 AND ITS RECEPTORS IN HUMAN SOLID TUMOURS: IMMUNOBIOLOGY AND CLINICAL SIGNIFICANCE 1

Theresa L. Whiteside, 2Torsten E. Reichert, and 3Qing Ping Dou.

1

University of Pittsburgh Cancer Institute, Pittsburgh, 2 University of Mainz, 3 Mainz, Germany and H. Lee Moffitt Cancer Center & Research Institute, Tampa, USA

Key words:

Interleukin-2 (IL-2), IL-2R, tumour growth, carcinomas, cell cycle arrest

Abstract:

Human carcinomas were found to express IL-2 R and to produce, but not to secrete, IL-2. Intermediate affinity detected on the surface and in the cytoplasm of carcinoma cells binds exogenous IL-2 at the nanomolar or micromolar concentrations and mediates cell cycle arrest (CCA) possibly through the upregulation of the CDK inhibitor expression. In contrast, is modestly expressed on the cell surface, and it may be involved in the intracrine pathway of delivering endogenous IL-2 to the cell surface. IL-2 is a growth factor for human carcinomas, and as it binds to the high-affinity IL-2R, it promotes cellular proliferation by suppressing expression of It also protects tumour cells from apoptosis. The presence in carcinomas and in normal tissue cells of two IL-2/IL-2R pathways regulating cellular growth and survival is a novel finding. Both the endogenous and exogenous IL2/IL2R pathways could become therapeutic targets in the future and could be explored to obtain insights into the mechanisms of tumour growth control as well as to modulate its sensitivity to other therapies.

1. INTRODUCTION Human cancer cells are known to express receptors for hematopoietic growth factors and cytokines (1-3). Exposure of these cells to cognate ligands leads to positive or negative receptormediated regulation of cell growth. Several years ago, we reported the presence of functional receptors for IL-2 (IL-2R) on human carcinoma cell lines and tumours in situ (4). Since then, others

have described expression of IL-2R on a variety of human normal and malignant cells and have demonstrated that binding of IL-2 to these receptors has significant and varied consequences in several types of non-hematopoietic cells (5-11). The biological and physiologic significance of IL-2R expression on these tissue cells remains poorly understood and controversial. However, it is important to note that binding of exogenous IL-2 to 27

W.G. Jiang et al. (eds.), Growth Factors and their Receptors in Cancer Metastasis, 27–50. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

28

Whiteside et al

human non-lymphoid tumour cells has been reported to alter expression of cell surface molecules involved in cell-to-cell interactions, increase sensitivity of tumour cells to cytostatic effects of other cytokines and drugs or to cytotoxicity mediated by immune effector cells. Furthermore, IL-2 is now known to either impair or promote growth of a variety of cells, including human tumour cells (1214). IL-2, also known as the T-cell growth factor (TCGF), is one of the first cytokines to be purified and cloned (15, 16). It is a 15 kDa glycoprotein encoded by a single gene on human chromosome 4 (17). It is considered to be a potent modulator of lymphocyte functions and to play a major role in the immune response by mediating activation, differentiation and growth of immune cells (17, 18). IL-2 acts through a receptor complex, containing at least three distinct chains, and which is found on all classes of lymphocytes and monocytes (18, 19). The IL-2R chain (p55) is the low-affinity while the IL-2R receptor chain function chain and the IL-2R together as the intermediate-affinity The non-covalent receptors association of the IL-2R a with the and chains on the cell surface results in expression of a high-affinity receptor IL-2-induced stimulation and proliferation of lymphocytes is mediated through the IL-2R and chains, since the α chain itself lacks the cytoplasmic domain necessary for signal transduction. The chain has been shown to be responsible for internalization of the IL2/IL-2R complex in addition to its signaltransducing function (20). Although IL-2 has been generally perceived as the key cytokine in the immune response, it is actually not essential to cellular immune functions, because "knock out" (KO)

Chapter 2 mice, lacking the ability to produce it, develop a normal immune system and are not immunodeficient, readily clear viruses such as LCMV or vaccinia and reject allotransplants (21). On the other hand, the chain is a critical element, as alterations in its gene cause X-linked severe combined immunodeficiency (22). IL-2 has been extensively used in cancer therapy, either as a locoregionally - or systemically - administered single agent or in combination with other biologicals (23-25). Antitumour effects of IL-2 have been universally attributed to its ability to up-regulate activities of immune cells (17). The published evidence that IL-2 can have direct antitumour effects (26, 27) or that it is produced by non-hematopoietic cells (5-11) has been largely ignored. More recently, the view of IL-2 as a "T-cell growth factor" has undergone a change, brought about by the realization that the IL-2/IL-2R pathway is involved in the regulation of both cell proliferation and cell death (14, 28). Experiments in mice have indicated that IL-2 can protect immune cells from apoptosis, but at the same time, it can potentiate death of these cells via, e.g., the Fas-mediated pathway (i.e., "activation-induced cell death" = AICD) under conditions of antigen-driven rapid proliferation (14). In light of these findings, the observation that the IL-2/IL2R pathway is ubiquitously expressed in human tissue cells and that it is upregulated in carcinoma cells take on a new significance. The possibility has to be considered that, similar to its effects in hematopoietic cells, the presence of the IL-2/IL-2R pathway in carcinomas might be linked to the susceptibility of these cells to apoptosis. Indeed, the growth of tumours is known to depend not only on cellular proliferation but also on the rate of cell death. This implies that effects of

2. IL-2 and IL-2R in solid tumours endogenous and/or exogenous IL-2 on ILtumour cells might influence tumour growth as well as tumour death. The therapeutic implication of this type of mechanism(s) is great, particularly since IL-2 has been established and approved as a therapeutic cytokine in some types of cancer, specifically, melanoma and renal cell carcinoma (29, 30). In the following narrative, we will first review evidence for the presence of functional IL-2R on human carcinomas. We will then consider experiments in support of the observation that endogenous, tumour-derived IL-2 functions as a growth factor and an apoptosis protection factor in tumour cells, while exogenously-delivered IL-2 has an entirely opposite effect, as it inhibits tumour cell growth and increases tumour cell susceptibility to immune cells or other cytokines. The mechanisms that might be responsible for these seemingly contradictory results will be considered. Finally, we will discuss the importance of the IL-2/IL-2R pathway for tumour therapy with IL-2. Overall, we expect to be able to convince the reader that the IL2/IL-2R pathway plays a significant role in the regulation of growth of human carcinomas and to provide some insights into the mechanisms responsible for growth or death of tumour cells as a result of perturbations of this pathway. 2. IL-2R ON CARCINOMAS 2.1 Expression of IL-2R on human tumours Receptors for various growth factors and cytokines are known to be expressed on human cancer cells. Ample evidence has accumulated for the crucial role of

29 these receptors, e.g., the epidermal growth factor receptor (EGFR) or insulin-like growth factor receptors (IGFR), in the development of cancer (1, 3, 31). Exposure of cells expressing these receptors to cognate factors leads to regulation of cell growth. The initial observation in our laboratory that recombinant IL-2 inhibited growth of tumour cells involved ex vivo exposure of head and neck cancer (HNC) cell lines and gastric carcinoma cell lines to various concentrations of the cytokine (Figure 1). To explain these results, we postulated that IL-2R were present on the tumour cells. Indeed, using monoclonal antibodies (Abs) specific for the IL-2R chain, we were able to confirm that nearly all of tumour cells in suspensions prepared from cellular monolayers expressed the chain of IL-2R on the cell surface when tested by flow cytometry (4, 26). The mean fluorescence intensity (MFI) for expression on the cell surface varied considerably among the tumour cell lines. Since human carcinomas grow as tight monolayers, it was possible that the dissociation solution or trypsin solution used for the preparation of tumour cell suspensions influenced the level of IL-2R expression. However, immunostaining of intact monolayers and of tumour biopsies confirmed the ubiquitous, albeit variable, expression of on the carcinoma cells/specimens examined. In contrast to the ubiquitous expression of the chain, the chain of IL-2R was either very weakly expressed or not detectable on the cell surface of chain expression carcinomas, and the was observed to be highly variable, ranging from 40 to 50% of tumour cells by flow cytometry (32).

30

Whiteside et al

Chapter 2

2. IL-2 and IL-2R in solid tumours To establish that IL-2R expressed on carcinomas were able to bind the ligand, IL-2, competitive binding studies with I25 I-labeled IL-2 were performed, using the PCI-1 cell line (4). Saturable binding of the radiolabeled ligand to the tumour cells was shown to be completely blocked by mAbs to the IL-2 binding site on the IL-2R chain but not by mAbs to other epitopes on this chain or by isotype control Abs (4, 33). By Scatchard analysis, PCI-1 tumour cells were found to express 13,000 of the intermediate affinity IL-2R and 300 of the high affinity IL-2R (14). Crosslinking of the receptors labeled with using 2 mM disuccinimidyl suberate (DSS) followed by SDSpolyacrylamide gel electrophoresis under reducing conditions, indicated the presence of a doublet corresponding to peptides with the molecular weights of 66 and 55 kDa (4). Thus, both and peptides were expressed on PCI-1 cells, although the chain had a smaller molecular weight than the expected 70-75 kDa, presumably due to the extent of its glycosylation. On the other hand, a possibility emerged that the on tumour cells was not identical to that expressed on hematopoietic cells, such as lymphocytes. The presence of mRNAs for all three IL-2R chains in carcinoma cells was confirmed by RT-PCR or RNase protection assays (32, 34) By flow cytometry performed with permeabilized cells, we compared intracytoplasmic expression of and chains in tumour cells and human lymphocytes (i.e., YT cell line or fresh NK cells). As shown in Table 1, the chain was detectable in nearly all carcinoma cells, the chain in about half of them and the chain in a very small proportion (l%-3%) of the cells.

31 The Scatchard analyses performed previously (4) indicated that these tumour cells expressed a small number (e.g., about 300) of chain-containing receptors on the cell surface. We, therefore, interpreted our combined intracytoplasmic and surface staining flow cytometry data to mean that a small number of chains were present on the surface of carcinoma cells. However, the level of expression of these chains was often at the borderline of the detection level for flow cytometry. On the other hand, abundant mRNA for the chain was always detectable in these cells (4, 26) and suggested post-transcriptional regulation of the α chain protein. Although our initial experiments were performed using squamous carcinoma of the head and neck (SCCHN) cell lines, subsequently obtained evidence indicated that expression of these IL-2R chains was detectable in human gastric and renal cell carcinoma cell lines (26). Surface expression of IL-2R was associated with significant inhibition of in vitro proliferation in these lines (26, 27). Normal epithelial cells, e.g., keratinocytes, in primary cultures were also shown to express IL-2R mRNA and protein. However, binding of exogenous IL-2 to IL-2R on keratinocytes did not result in a negative growth signal (26). These studies have led to a conclusion that functional IL-2Rs are ubiquitously expressed on human carcinomas as well as normal epithelial cells in culture, and that their function might be modified in cancer cells relative to that in normal lymphoid cells. Consistently observed expression of the intermediate-affinity IL-2R on HNC cell lines prompted us to examine frozen sections of tumour-involved as well as tumour- free adjacent oral mucosa of the oropharynx and of laryngeal tumours for the presence of IL-2R and localization of

32

Whiteside et al

IL-2R+ tumour or normal cells in tissues. A clear pattern emerged from these immunoperoxidase studies as follows: a) expression was documented in all tumour tissues examined (n=27) as well as in tumour-uninvolved oral mucosa ( n= 11) in the suprabasal location (see Figure 2a and b); b) the chain was not detectable in normal or tumour tissues by immunoperoxidase staining; c) expression of the chain was strongest in the basal epithelial layer both in normal oral

Chapter 2 mucosa and in tumour tissues. (Figure 2c and d). Furthermore, it appeared that poorly differentiated SCCHN tumours showed stronger staining for IL-2R than well differentiated tumours (33). The results of immunohistologic analyses for IL-2R in tissues appeared to correlate with those obtained for IL-2 expression (see below), although additional studies will be necessary to obtain statistically significant correlations.

2. IL-2 and IL-2R in solid tumours 2.2. Growth inhibition induced in carcinomas by exogenous IL-2 The detection of IL-2R in tumour and normal epithelial cells raised the possibility that a functional IL-2/IL-2R pathway might exist in tumour or tissue cells of the non-hematopoietic origin. Therefore, tumour cells were incubated in the presence of various concentrations of recombinant IL-2 to evaluate its effect on proliferation of these targets (Figure 3). Using 4-day colorimetric MTT assays (35), 4-day incorporation assays or cell counts (36), we were able to demonstrate that exogenous IL-2 at the concentrations consistently nM or inhibited growth of carcinoma cell lines. These effects were particularly pronounced in tumour cells cultured in the presence of low concentrations of fetal calf serum (i.e., 1% v/v). Furthermore, these inhibitory effects were IL-2 dosedependent (Figure 3). Growth inhibition induced by exogenous IL-2 in tumour cells could have resulted from the cell cycle arrest (CCA) or apoptosis. To distinguish between these two mechanisms, SCC cells inhibited in growth after 3-day culture in the presence of exogenous IL-2 (22nM) were compared with control tumour cells cultured in tissue culture medium (TCM) alone for the presence of DNA fragmentation, using flow cytometry TUNEL assays. No evidence for apoptosis was ever observed in SCC cells cultured in the presence of growth-inhibitory concentrations of IL-2. An alternative mechanism of CCA was, therefore, investigated by flow cytometry analyses of the cell cycle in tumour cells incubated in the presence of 22nM of IL-2 for 3 days. The tumour cells were labeled with propidium iodide (PI), and the proportion of cells in the S and G2/M phases of the cell cycle were quantitated. As shown in Table 2, the

33 fraction of tumour cells in the phase was significantly increased, while that in the S phase was decreased in cells incubated with the growth inhibitory doses of exogenous IL-2. It must be emphasized that although these experiments were not performed with synchronized cultures, the fraction of phase tumour cells arrested in the of the cell cycle was significant in cultures incubated with exogenous IL-2 at the concentrations >22nM. These ex vivo experiments, repeated with various carcinoma cell lines, indicated that exogenous IL-2 used at the relatively high concentration of 22nM or greater inhibited tumour cell growth by inducing the CCA. Furthermore, earlier immunotherapy experiments performed with IL-2 in nude mice bearing established human carcinoma xenografts demonstrated that either local or systemic (iv) delivery of this cytokine (e.g., 6000 IU/mL injected peritumourally once a day for 2 weeks) consistently resulted in a significant reduction of the tumour size (26, 27). In view of the fact that >22nM concentrations of exogenous IL-2 were necessary to induce the CCA in carcinoma cells, it was assumed that the effect was mediated via the intermediate IL-2R, i.e., . This hypothesis was consistent with considerable levels of expression on carcinomas and very low levels of IL2R expression, as detected by flow cytometry or immunostaining on frozen tissue sections. To investigate whether expression of the high affinity IL2R in carcinoma cells would result in greater sensitivity to IL-2 and more pronounced CCA, we transduced the gene into gastric carcinoma, RCC and SCCHN cell lines, using lipofection and the p55CDM 8Neo expression vector (34).

34

Whiteside et al

Chapter 2

2. IL-2 and IL-2R in solid tumours

The transduced and selected tumour cells expressed substantially higher levels on the cell surface than of IL-2R parental or LacZ control cells (Figure 4a). Growth of transduced PCI-1 cell clones (SCCHN) selected for high levels of the IL-2R chain expression and cultured in the presence of rIL-2 over a wide range of concentrations was comparable to that of the controls (Fig 4b;ref 32). Thus, increased expression of and the increased number of binding sites on the transduced HNC cells were not accompanied by their greater sensitivity to growth inhibitory effects of IL-2 (Figure 4b). On the other hand, when the IL-2R chain was transduced into PCI-1 cells, based on the observation that mRNA and was protein expression of demonstrated to be lower in these cells than that in control lymphoid cells (26), the transduced cell line became significantly more sensitive to growth inhibition mediated by exogenous IL-2 than parental or LacZ-transduced controls (Figure 4b). These experiments demonstrated that transfer of the chain gene into HNC cells increased the number of the IL-2 binding sites on the tumour cell surface (32) as well as tumour cell sensitivity to growth inhibition by exogenous IL-2 These observations were confirmed using another of our SCCHN cell lines (PCI-13) transduced with the IL2Ra or gene (data not shown). Taken together, the IL-2R gene transduction data indicated that in HNC cells, the complex (intermediate affinity receptor) was the functional receptor responsible for the observed growth inhibitory effects of exogenous IL-2.

35

In contrast to the results described above for SCCHN, both HR (gastric carcinoma) and a RCC tumour cell line chain gene transduced with the and found to express high-affinity IL-2R by flow cytometry (Fig 4a) became significantly more sensitive to growth inhibitory effects of exogenous IL-2 than parental or LacZ transduced cells (Figure 4b). When HR cells transduced with the IL-2R chain gene were compared to parental HR cells for growth in tissue culture medium not containing any exogenous IL-2, they were found to proliferate significantly better. These data suggested that expression of the IL-2R chain on HR was important for growth of these cells even in the absence of exogenous IL-2 (Figure 5). In experiments designed to confirm

the essential role of

in tumour cell growth, the HR cell line was transduced with pCEP4p70R, containing antisense IL-2R cDNA (34). Expression of ILon transduced and control (LacZ transduced) HR cells was assessed by flow cytometry and found to be absent (32). Growth of these tumour cells transduced with antisense was significantly inhibited (p50% reduction of tumour weight) compared to controls. In an effort to determine potential mechanisms for this response, specific immune cell populations were assayed either biochemically or directly through quantitative immunohistochemical staining. We observed extensive immune cell activation following injection of AdmIL-12 and ultimately concluded from this study that the local anti-tumour activities likely resulted from: 1) enhanced NK lysis during the first 7 days following virus injection, 2) enhanced macrophage activity such as NOS activation, and 3) support of cytokine production from and possible cytolytic activities of CD4-positive and/or CD8positive T cells within the local tumour tissue (13). The observation of multiple immunocyte activities that potentially could develop into a systemic anti-tumour immune response involving the generation of memory T cells was evident, and the results of analysis of distant antimetastatic activity in response to local injection of AdmIL-12 further supported this notion.

84

Ebara et al

Chapter 5

5. IL-12 gene therapy in cancer

3.2 In Situ IL-12 Gene Therapy Effects on Distant Metastases In this initial study, we further evaluated the potential for in situ IL-12 gene therapy to affect distant metastatic disease using two different approaches. As RM-9 cells metastasize spontaneously from the orthotopic site, we initially evaluated the effects of this gene therapy protocol on the extent of spontaneous metastasis to lymph nodes. The results indicated that localized gene therapy could significantly suppress the incidence of spontaneous lymph node metastasis. Further studies also demonstrated that in situ IL-12 gene therapy could suppress the formation of preestablished lung metastases following injection of RM-9 cells into the tail vein (13). These results clearly indicated that our therapeutic strategy was not only capable of generating localized cytotoxic response through specific effector cells but also of generating a systemic response was generated that impacted on metastatic disease (see Figure-1).

85

Specific NK depletion analysis demonstrated that NK cells were predominantly responsible for the antimetastatic effects of locally administered AdmIL-12 in preestablished RM-9 lung metastases. However, they also indicated that other cell types were likely involved. Recent studies concerning the relationship between IL-12 efficacy and NK cell versus T cells may explain in part the potent initial NK activities both in the local tumour as well as in distant metastases but also the later increase in CD8 infiltration that we observed in this study. IL-12 has been reported to induce an NK-mediated cytolytic phase, followed by a T cell phase that is characterized by CTL activities (15, 16, 45, 46). Although interrelationships between the NK phase and CTL phase are poorly understood, it is well established that the generation of a Th1 response is required for the CTL phase. We observed (13) maximal NK activity at day 1 after AdmIL-12 injection but increased

86

Ebara et al

infiltration of both CD4-positive and CD8-positive T cells 7 days following virus injection. The tumour-infiltrating CD4 and CD8 T cell populations subsequently decreased over time, and it is not clear whether the infiltrating CD4 cells were of the Th1 phenotype (13). However, the immune cells infiltrate that we observed in the local tumour following in situ IL-12 gene therapy may have relevance to the antimetastatic activities that it generated. Since NK depletion did not completely eliminate the IL-12 antimetastatic activities, we strongly suspect that CD4-/CD8-positive T cells are involved. We further suspect that the development of CD4 Th1 activities and potentially macrophage activities may serve to bridge the initial NK response to more potent and long-lasting antitumour effects that could significantly impact on metastatic disease (see Figure-2). Further studies are in progress to test these concepts and develop further therapeutic approaches, based on the results. 3.3 Gene Modified Cell-Based Vaccines The use of adenoviral vector mediated IL-12 gene therapy is not restricted to intratumoural vector injections, i.e., in situ gene therapy approaches. Indeed, the combination of an adenovirus delivery system together with remarkable immune inductive activities of IL-12 represents a highly versatile combination for use in various cell-based vaccine strategies. In particular for prostate cancer where multiple opportunities for neoadjuvant and adjuvant approaches exist vaccine therapies using adenovirus-mediated IL12 gene transfer may provide unique opportunities. For example, it is conceivable that tumour biopsies could be utilized for the recovery of viable cells into which IL-12 could be transduced using adenoviral vectors. In recent

Chapter 5 studies, we have demonstrated that adenoviral vectors could be used to effectively transfer IL-12 into tumour cells and following irradiation and subsequent subcutaneous implantation, this vaccine was capable of protecting a significant percentage of animals from later tumour challenge using RM-9 cells (44). Other studies have demonstrated the feasibility of using adenoviral vectors to transduce genes into dendritic cells (47, 48), and it was recently demonstrated that dendritic cells transduced with retroviral vectors expressing IL-12 have local and systemic antitumour activities when injected directly into various rodent tumours (49). Therefore, the potential of adenovirus-mediated IL-12 cell based vaccine strategies are numerous and various preclinical studies have demonstrated not only their feasibility but the therapeutic potential of this highly versatile gene therapy approach. 4. IL-12 TOXICITY Recent studies have indicated that the administration of recombinant IL-12 may present a difficult challenge in regard to its association with life-threatening toxicities. Indeed, because of the difficulties in controlling the systemic levels and the widespread systemic effects of recombinant IL-12 on various cell types, previous studies on IL-12 in humans have documented that lifethreatening myelosuppression and splenomegaly can occur (19, 50-52). Therefore, because of these considerations, we and others have considered local delivery of potent cytokines such as IL-12 by using gene therapy protocols to be preferable, in that they may provide sufficiently effective local concentrations of the cytokine without generating toxic systemic levels that are difficult to control.

5. IL-12 gene therapy in cancer In our previous study (13) we evaluated serum levels of IL-12 following orthotopic injection of AdmIL-12. The results indicated that serum concentrations of IL-12 were increased during a period of 10 days after initiation of the treatment and with maximal levels occurring the first day following vector injection. These levels led to enlargement of the spleen after a lag time of several days, as the maximum spleen size was observed on day 7 following treatment (13). This splenomegaly was reversible, since the gradual decrease in serum IL-12 strongly correlated with a return to normal spleen size. Further studies are indicated regarding the adenoviral vector delivery of IL-12 compared to systemic administration. Our studies indicated that adenoviral-vector-delivered IL-12 that had both local and antimetastatic effects was relatively safe and did not result in limiting systemic toxicities. 5. COMBINATION THERAPIES Although we are optimistic regarding the therapeutic potential of adenoviraldelivered in situ IL-12 gene therapy for prostate cancer, we consider this approach to be only another step in the ultimate generation of an optimized therapeutic protocol. The next step forward that we and others have considered is combining IL-12 gene transduction with other costimulatory molecules such as B7-1. This approach has been tested in various other systems and has indicated that the addition of this potent co-stimulatory molecule can modify the effector cell response to IL-12 and, in some cases, result in enhanced therapeutic activity (26, 44, 53). As discussed above IL-12 alone or together with co-stimulating genes could be used in various cell–based vaccine strategies that involve tumour cells and/or antigen presenting cells.

87 An additional approach would be to combine suicide gene therapy together with IL-12 or IL-12+B7. This therapeutic approach also may afford may specific advantages, as be effective at maximizing tumour antigen presentation to effector cells and in conjunction with IL-12, which could mature the effector cell responses and lead to highly synergistic activities (5). At our institution we now have the extensive experience with adenoviralvector-delivered gene therapy in ongoing clinical trials. The addition of adenoviral-vector-delivered IL-12 or IL-12+B7 could be integrated both into our preclinical and clinical programs in an effort to take the next step toward successfully treating both localized and systemic metastatic disease in prostate cancer. Hopefully, these new novel approaches together with currently available radical prostatectomy and radiation will make dramatic improvements in survival with this devastating disease. 6. SUMMARY The lack of effective therapy for locally invasive and metastatic prostate cancer dictates the necessity for intensive focus on the development of novel and effective therapeutic approaches for this important malignancy. The notion of initiating active and persistent antitumour immunity through molecular gene transfer approaches offers a viable avenue of investigation for addressing this problem. The cytokine IL-12 is remarkable and unique in its capacity to induce widespread multilevel cascades of cellular and gene activity that can contribute to the development of a specific antitumour immune response. The use of adenoviral vector gene transfer systems together with IL-12 offers a highly flexible and efficient

88

Ebara et al

avenue for initiating the potentially therapeutic activities yet limiting the toxicities associated with IL-12. For prostate cancer this strategy can involve in situ-based gene therapy approaches and various cell based vaccine protocols. With continued preclinical studies and clinical trials in this area it seems likely that immunomodulatory gene therapy will make a significant contribution to the

Chapter 5 effective treatment of prostate cancer in the future. Acknowledgements The work in the authors’ laboratory was supported by grants from CaP CURE and NIH (CA50588, CA68814 and SPORE P50-58204). We are grateful to Dr. Joann Trial for critical review of the manuscript.

References 1. Landis, S. H., Murray, T., Bolden, S., and Wingo, P. A. Cancer Statistics, 1999. CA Cancer J Clin 1999; 49:8-31 2. Smith, D. S. and Catalona, W. J. The nature of prostate cancer detected through prostate specific antigen based screening. J Urol 1994; 152:1732-6 3. Ohori, M., Wheeler, T. M., Kattan, M. W., Goto, Y., and Scardino, P. T. Prognostic significance of positive surgical margins in radical prostatectomy specimens. J Urol 1995; 154:1818-24 4. Zietman, A. L., Edelstein, R. A., Coen, J. J., Babayan, R. K., and Krane, R. J. Radical prostatectomy for adenocarcinoma of the prostate: the influence of preoperative and pathologic findings on biochemical disease-free outcome. Urology 1994; 43:828-33 5. Thompson, T. C. In situ gene therapy for prostate cancer. Oncol Res 1999; 11:1-8 6. Pham-Nguyen, K. B., Yang, W., Saxena, R., Thung, S. N., Woo, S. L., and Chen, S. H. Role of NK and T cells in IL-12-induced anti-tumour response against hepatic colon carcinoma. Int J Cancer 1999; 81:813-9 7. Gambotto, A., Tuting, T., McVey, D. L., Kovesdi, I., Tahara, H., Lotze, M. T., and Robbins, P. D. Induction of antitumour immunity by direct intratumoural injection of a recombinant adenovirus vector expressing interleukin-12. Cancer Gene Ther 1999; 6:45-53 8. Ogawa, M., Tsutsui, T., Zou, J. P., Mu, J., Wijesuriya, R., Yu, W. G., Herrmann, S., Kubo, T., Fujiwara, H., and Hamaoka, T. Enhanced induction of very late antigen 4/lymphocyte function-associated antigen 1-dependent T-cell migration to tumour sites following administration of interleukin 12. Cancer Res 1997; 57:2216-22 9. Ogawa, M., Umehara, K., Yu, W. G., Uekusa, Y., Nakajima, C., Tsujimura, T., Kubo, T., Fujiwara, H., and Hamaoka, T. A critical role for

a peritumoural stromal reaction in the induction of T-cell migration responsible for interleukin12-induced tumour regression. Cancer Res 1999; 59:1531-8 10. Boggio, K., Di Carlo, E., Rovero, S., Cavallo, F., Quaglino, E., Lollini, P. L., Nanni, P., Nicoletti, G., Wolf, S., Musiani, P., and Forni, G. Ability of systemic interleukin-12 to hamper progressive stages of mammary carcinogenesis in HER2/neu transgenic mice. Cancer Res 2000; 60:359-64 11. Nanni, P., Rossi, I., De Giovanni, C., Landuzzi, L., Nicoletti, G., Stoppacciaro, A., Parenza, M., Colombo, M. P., and Lollini, P. L. Interleukin 12 gene therapy of MHC-negative murine melanoma metastases. Cancer Res 1998; 58:1225-30 12. Colombo, M. P., Vagliani, M., Spreaflco, F., Parenza, M., Chiodoni, C., Melani, C., and Stoppacciaro, A. Amount of interleukin 12 available at the tumour site is critical for tumour regression. Cancer Res 1996; 56:2531-4 13. Nasu, Y., Bangma, C. H., Hull, G. W., Lee, H. M., Hu, J., Wang, J., McCurdy, M. A., Shimura, S., Yang, G., Timme, T. L., and Thompson, T. C. Adenovirus-mediated interleukin-12 gene therapy for prostate cancer: suppression of orthotopic tumour growth and pre-established lung metastases in an orthotopic model. Gene Ther 1999; 6:338-49 14. Gately, M. K., Renzetti, L. M., Magram, J., Stern, A. S., Adorini, L., Gubler, U., and Presky, D. H. The interleukin-12/interleukin-12-receptor system: role in normal and pathologic immune responses. Annu Rev Immunol 1998; 16:495521 15. Scott, P. IL-12: initiation cytokine for cellmediated immunity. Science 1993; 260:496-7 16. Gately, M. K. Interleukin-12: a recently discovered cytokine with potential for enhancing cell-mediated immune responses to tumours. Cancer Invest 1993; 11:500-6

5. IL-12 gene therapy in cancer 17. Wu, C. Y., Demeure, C., Kiniwa, M., Gately, M., and Delespesse, G. IL-12 induces the production of IFN-gamma by neonatal human CD4 T cells. J Immunol 1993; 151:1938-49 18. Stern, A. S., Podlaski, F. J., Hulmes, J. D., Pan, Y. C., Quinn, P. M., Wolitzky, A. G., Familletti, P. C., Stremlo, D. L., Truitt, T., Chizzonite, R., and Gately, M. K. Purification to homogeneity and partial characterization of cytotoxic lymphocyte maturation factor from human Blymphoblastoid cells. Proc Natl Acad Sci U S A 1990; 87:6808-12 19. Brunda, M. J., Luistro, L., Warrier, R. R., Wright, R. B., Hubbard, B. R., Murphy, M., Wolf, S. F., and Gately, M. K. Antitumour and antimetastatic activity of interleukin 12 against murine tumours. J Exp Med 1993; 178:1223-30 20. Zou, J. P., Yamamoto, N., Fujii, T., Takenaka, H., Kobayashi, M., Herrmann, S. H., Wolf, S. F., Fujiwara, H., and Hamaoka, T. Systemic administration of rIL-12 induces complete tumour regression and protective immunity: response is correlated with a striking reversal of suppressed IFN-gamma production by antitumour T cells. Int Immunol 1995; 7:1135-45 21. Boggio, K., Nicoletti, G., Di Carlo, E., Cavallo, F., Landuzzi, L., Melani, C., Giovarelli, M., Rossi, I., Nanni, P., De Giovanni, C., Bouchard, P., Wolf, S., Modesti, A., Musiani, P., Lollini, P. L., Colombo, M. P., and Forni, G. Interleukin 12-mediated prevention of spontaneous mammary adenocarcinomas in two lines of Her2/neu transgenic mice. J Exp Med 1998; 188:589-96 22. Szabo, S. J., Kim, S. T., Costa, G. L., Zhang, X., Fathman, C. G., and Glimcher, L. H. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 2000; 100:655-69 23. MacMicking, J., Xie, Q. W., and Nathan, C. Nitric oxide and macrophage function. Annu Rev Immunol 1997; 15:323-50 24. Michel, T. and Feron, O. Nitric oxide synthases: which, where, how, and why? J Clin Invest 1997; 100:2146-52 25. Cavallo, F., Di Carlo, E., Butera, M., Verrua, R., Colombo, M. P., Musiani, P., and Forni, G. Immune events associated with the cure of established tumours and spontaneous metastases by local and systemic interleukin 12. Cancer Res 1999; 59:414-21 26. Joki, T., Kikuchi, T., Akasaki, Y., Saitoh, S., Abe, T., and Ohno, T. Induction of effective antitumour immunity in a mouse brain tumour model using B7-1 (CD80) and intercellular adhesive molecule 1 (ICAM-1; CD54) transfection and recombinant interleukin 12. Int J Cancer 1999; 82:714-20

89 27. Colombo, M. P., Lombardi, L., Melani, C., Parenza, M., Baroni, C., Ruco, L., and Stoppacciaro, A. Hypoxic tumour cell death and modulation of endothelial adhesion molecules in the regression of granulocyte colony-stimulating factor-transduced tumours. Am J Pathol 1996; 148:473-83 28. Tannenbaum, C. S., Wicker, N., Armstrong, D., Tubbs, R., Finke, J., Bukowski, R. M., and Hamilton, T. A. Cytokine and chemokine expression in tumours of mice receiving systemic therapy with IL-12. J Immunol 1996; 156:693-9 29. Duda, D. G., Sunamura, M., Lozonschi, L., Kodama, T., Egawa, S., Matsumoto, G., Shimamura, H., Shibuya, K., Takeda, K., and Matsuno, S. Direct in vitro evidence and in vivo analysis of the antiangiogenesis effects of interleukin 12. Cancer Res 2000; 60:1111-6 30. Sgadari, C., Angiolillo, A. L., and Tosato, G. Inhibition of angiogenesis by interleukin-12 is mediated by the interferon-inducible protein 10. Blood 1996; 87:3877-82 31. Hobeika, A. C., Etienne, W., Cruz, P. E., Subramaniam, P. S., and Johnson, H. M. IFNgamma induction of p21WAF1 in prostate cancer cells: role in cell cycle, alteration of phenotype and invasive potential. Int J Cancer 1998; 77:138-45 32. Kominsky, S. L., Hobeika, A. C., Lake, F. A., Torres, B. A., and Johnson, H. M. Downregulation of neu/HER-2 by interferon-gamma in prostate cancer cells. Cancer Res 2000; 60:39048 33. Munoz-Fernandez, M. A., Fernandez, M. A., and Fresno, M. Synergism between tumour necrosis factor-alpha and interferon-gamma on macrophage activation for the killing of intracellular Trypanosoma cruzi through a nitric oxide-dependent mechanism. Eur J Immunol 1992; 22:301-7 34. Trinchieri, G. Proinflammatory and immunoregulatory functions of interleukin-12. Int Rev Immunol 1998; 16:365-96 35. Vassalli, P. The pathophysiology of tumour necrosis factors. Annu Rev Immunol 1992; 10:411-52 36. Fajardo, L. F., Kwan, H. H., Kowalski, J., Prionas, S. D., and Allison, A. C. Dual role of tumour necrosis factor-alpha in angiogenesis. Am J Pathol 1992; 140:539-44 37. Downey, G. P. Mechanisms of leukocyte motility and chemotaxis. Curr Opin Immunol 1994; 6:113-24 38. Wong, G. H. W., Vehar, G., and Kaspar, R. L. Apoptosis and Cancer. In: S. J. Martin (ed.), pp. 245-257: Karger Landes System, 1997.

90

Ebara et al

39. Sherwood, E. R., Ford, T. R., Lee, C., and Kozlowski, J. M. Therapeutic efficacy of recombinant tumour necrosis factor a in an experimental model of human prostatic carcinoma. J. Biol. Response Modif. 1998; 9:4452 40. Rokhlin, O. W., Gudkov, A. V., Kwek, S., Glover, R. A., Gewies, A. S., and Cohen, M. B. p53 is involved in tumour necrosis factor-alphainduced apoptosis in the human prostatic carcinoma cell line LNCaP. Oncogene 2000; 19:1959-68 41. Kimura, K., Bowen, C., Spiegel, S., and Gelmann, E. P. Tumour necrosis factor-alpha sensitizes prostate cancer cells to gammairradiation-induced apoptosis. Cancer Res 1999; 59:1606-14 42. Shimura, S., Yang, G., Wheeler, T. W., Frolov, A., and Thompson, T. C. Reduced Infiltration of Tumour-Associated Macrophages in Human Prostate Cancer: Association with Cancer Progression. Cancer Res 2000; 60: 5857-5861 43. Baley, P. A., Yoshida, K., Qian, W., Sehgal, I., and Thompson, T. C. Progression to androgen insensitivity in a novel in vitro mouse model for prostate cancer. J Steroid Biochem Mol Biol 1995; 52:403-13 44. Hull, G. W., McCurdy, M. A., Nasu, Y., Bangma, C. H., Shimura, S., Lee, H.-M., Wang, J., Albani, J., Ebara, S., Sato, T., Timme, T. L., and Thompson, T. C. Prostate cancer gene therapy: Comparison of adenovirus mediated expression of interleukin-12 with interleukin-12 plus B7-1 for in situ gene therapy and genemodified cell-based vaccines. Clin. Cancer Res. 2000; 6:4101-4109 45. Brunda, M. J. Interleukin-12. J Leukoc Biol 1994; 55:280-8 46. Banks, R. E., Patel, P. M., and Selby, P. J. Interleukin 12: a new clinical player in cytokine therapy [editorial]. Br J Cancer 1995; 71:655-9 47. Arthur, J. F., Butterfield, L. H., Roth, M. D., Bui, L. A., Kiertscher, S. M., Lau, R., Dubinett, S., Glaspy, J., McBride, W. H., and Economou, J. S. A comparison of gene transfer methods in human dendritic cells. Cancer Gene Ther 1997; 4:17-25 48. Zhong, L., Granelli-Piperno, A., Choi, Y., and Steinman, R. M. Recombinant adenovirus is an efficient and non-perturbing genetic vector for human dendritic cells. Eur J Immunol 1999; 29:964-72 49. Nishioka, Y., Hirao, M., Robbins, P. D., Lotze, M. T., and Tahara, H. Induction of systemic and therapeutic antitumour immunity using intratumoural injection of dendritic cells genetically modified to express interleukin 12. Cancer Res 1999; 59:4035-41

Chapter 5 50. Orange, J. S., Salazar-Mather, T. P., Opal, S. M., Spencer, R. L., Miller, A. H., McEwen, B. S., and Biron, C. A. Mechanism of interleukin 12mediated toxicities during experimental viral infections: role of tumour necrosis factor and glucocorticoids. J Exp Med 1995; 181:901-14 51. Lamont, A. G. and Adorini, L. IL-12: a key cytokine in immune regulation. Immunol Today 1996; 17:214-7 52. Leonard, J. P., Sherman, M. L., Fisher, G. L., Buchanan, L. J., Larsen, G., Atkins, M. B., Sosman, J. A., Dutcher, J. P., Vogelzang, N. J., and Ryan, J. L. Effects of single-dose interleukin-12 exposure on interleukin-12associated toxicity and interferon- gamma production. Blood. 1997; 90:2541-8 53. Putzer, B. M., Hitt, M., Muller, W. J., Emtage, P., Gauldie, J., and Graham, F. L. Interleukin 12 and B7-1 costimulatory molecule expressed by an adenovirus vector act synergistically to facilitate tumour regression. Proc Natl Acad Sci USA 1997; 94:10889-10894 54. Ferrone, S. and Marincola, F. M. Loss of HLA class I antigens by melanoma cells: molecular mechanisms, functional significance and clinical relevance. Immunol Today 1995; 16:487-94 55. Garrido, F., Ruiz-Cabello, F., Cabrera, T., PerezVillar, J. J., Lopez-Botet, M., Duggan-Keen, M., and Stern, P. L. Implications for immunosurveillance of altered HLA class I phenotypes in human tumours. Immunol Today 1997; 18:89-95 56. Nakazaki, Y., Tani, K., Lin, Z. T., Sumimoto, H., Hibino, H., Tanabe, T., Wu, M. S., Izawa, K., Hase, H., Takahashi, S., Tojo, A., Azuma, M., Hamada, H., Mori, S., and Asano, S. Vaccine effect of granulocyte-macrophage colony-stimulating factor or CD80 genetransduced murine hematopoietic tumour cells and their cooperative enhancement of antitumour immunity. Gene Ther 1998; 5:1355-62 57. Sehgal, I., Baley, P. A., and Thompson, T. C. Transforming growth factor betal stimulates contrasting responses in metastatic versus primary mouse prostate cancer-derived cell lines in vitro. Cancer Res 1996; 56:3359-65 58. Fakhrai, H., Dorigo, O., Shawler, D. L., Lin, H., Mercola, D., Black, K. L., Royston, I., and Sobol, R. E. Eradication of established intracranial rat gliomas by transforming growth factor beta antisense gene therapy. Proc Natl Acad Sci U S A 1996; 93:2909-14 59. Finke, J. H., Zea, A. H., Stanley, J., Longo, D. L., Mizoguchi, H., Tubbs, R. R., Wiltrout, R. H., O’Shea, J. J., Kudoh, S., Klein, E., Bukowski, R. M., and Ochoa, A. C. Loss of T-cell receptor zeta chain and p561ck in T-cells infiltrating

5. IL-12 gene therapy in cancer human renal cell carcinoma. Cancer Res 1993; 53:5613-6 60. Alexander, J. P., Kudoh, S., Melsop, K. A., Hamilton, T. A., Edinger, M G., Tubbs, R. R., Sica, D., Tuason, L., Klein, E,, Bukowski, R. M., and Finke, J. H. T-cells infiltrating renal cell carcinoma display a poor proliferative response even though they can produce interleukin 2 and express interleukin 2 receptors. Cancer Res 1993; 53:1380-7

91 61. Sturmhoefel, K., Lee, K., Gray, G. S., Thomas, J., Zollner, R., O’Toole, M., Swiniarski, H., Dorner, A., and Wolf, S. F. Potent activity of soluble B7-IgG fusion proteins in therapy of established tumours and as vaccine adjuvant. Cancer Res 1999; 59:4964-72 62. Lee, H. M., Timme, T. L., and Thompson, T. C. Resistance to lysis by cytotoxic T cells: a dominant effect in metastatic mouse prostate cancer cells. Cancer Res 2000; 60:1927-33

This page intentionally left blank

Chapter 6 FIBROBLAST GROWTH FACTORS AND THEIR RECEPTORS IN METASTASES OF PROSTATE AND OTHER UROLOGICAL CANCERS

Zoran Culig, Marcus V. Cronauer, Alfred Hobisch, Georg Bartsch, and Helmut Klocker Department of Urology, University of Innsbruck, Austria

Key words:

Acid and basic fibroblast growth factor, prostate cancer, bladder cancer, renal cell carcinoma, invasion, metastasis, angiogenesis

Abstract:

Theraputical options for advanced carcinoma of the prostate, bladder or kidney are limited. Therefore it is important to understand their invasion and metastasis, processes in which fibroblast growth factors play an important role. Basic fibroblast growth factor (bFGF) is expressed in androgen-insensitive prostate cancer cell lines PC-3 and DU-145 and in some clinical specimens. During progression of prostate cancer, the expression of the FGF receptor 2 isoform IIIb, which preferentially binds keratinocyte growth factor (KGF) decreases and the expression of the isoform IIIc, which preferentially binds bFGF increases. A similar phenomenon was observed in bladder cancer. Several FGFs are proposed to act as andromedins, proteins that mediate the effects of androgens in target tissues: FGF-7 (KGF), FGF-8 and FGF-10. In prostate and bladder cancer, FGFs regulate tumour metastases by induction of the matrix metalloproteinase promatrilysin. Matrix metalloproteinases degrade extracellular matrix proteins and their expression is elevated in prostate cancer cells. bFGF is strongly expressed in invasive bladder cancers in which it promotes angiogenesis. bFGF levels in bladder cancer are down-regulated by administration of interferon-alpha. bFGF and aFGF are frequently elevated in urine of bladder and renal cancer patients. There is a strong association between urinary bFGF and clinical parameters in bladder cancer. In renal cancer, it was shown that the transfection of bFGF cDNA leads to an increased invasiveness and formation of metastatic nodules.

93 W.G. Jiang et al. (eds.), Growth Factors and their Receptors in Cancer Metastasis, 93–106. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

94

Culig et al

1. FIBROBLAST GROWTH FACTOR FAMILY – STRUCTURE, FUNCTION AND SIGNAL TRANSDUCTION A common characteristics of members of the fibroblast growth factor (FGF) family is that they bind to heparin and this binding protects them from degradation by proteases, high temperature or low pH. Basic FGF (bFGF), a single-chain protein of 146 amino acids with a molecular mass of 16 500 Da, was purified from pituitary and brain extracts as a protein with strong mitogenic activity for fibroblasts (1,2). Bigger forms of bFGF which are translation products of alternative initiation sites have been isolated from various tissues. The existence of a second mitogen for fibroblasts distinct from bFGF was detected in the same tissues and the factor was named aFGF because of its acidic isoelectric point (3). aFGF is a 140-amino acid protein with a molecular mass of 15 500 Da. aFGF and bFGF are homologous proteins with about 55% sequence identity. The genomic organization of the human aFGF and bFGF is similar. The coding sequence consists of three exons, which are interrupted by two introns (4). The aFGF single-copy gene is located on chromosome 5 and the bFGF gene is situated on chromosome 4q26-27. Four members of the FGF family (FGFs 3-6) are oncogene products. FGF-7 was found to be a growth factor for keratinocytes and was therefore termed keratinocyte growth factor (KGF). Newly discovered FGFs are FGF-8, -9 and –10 (5-7). The FGF receptor (FGFR) family consists of four genes which exhibit structural heterogeneity; FGFR-1 (the fig gene product), FGFR-2 (the bek gene product), FGFR-3 and FGFR-4 (8). Both aFGF and bFGF lack a signal peptide consensus sequence and there are several hypotheses to explain their secretion from cells. The

Chapter 6 mechanism of growth factor secretion involves cell lysis or the damage of plasma membrane. Alternative release mechanisms may involve an ATP-driven peptide pump or a formation of a complex between FGF and and a carrier protein. FGFs are synthesized by many cell types – fibroblasts, endothelial, smooth muscle, granulosa and adrenocortical cells and astrocytes. Due to the presence of the respective receptors, autocrine proliferation loops exist. FGF production was observed in a number of tumours, such as central nervous system ones, leukemias and liver tumours. FGFs effects in target cells include mitogenesis, stimulation of motility and migration and synthesis of specific cellular proteins. Various signal transduction pathways are initiated after binding of FGFs to their receptors; induction of the protooncogenes fos and myc, stimulation of the adenylate cyclase, breakdown of phosphatidyl inositides and the generation of the second messenger diacylglycerol, induction of the protein kinase C or mitogen-activated protein kinase pathways and elevation of intracellular calcium were described (914). One of the most important functions of FGFs is induction of new blood vessel growth which occurs during early stages of tumour development. Formation of capillary blood vessels consists of endothelial cell proliferation, the sprouting of new capillaries, endothelial cell migration and the breakdown of extracellular matrix surrounding capillaries. FGFs stimulate these activities and also stimulate the proliferation of endothelial cells from blood vessels (15). Extracellular matrix molecules, such as laminin and collagens, send signals for capillary tubular formation, elongation and differentiation. Blood vessel formation is regulated by interplay between bFGF and transforming growth

6. FGF and cancer metastasis factor-ß (TGF-ß). The extracellular matrix molecule urokinase-type plasminogen activator (uPA), which is up-regulated by bFGF, induces TGF-ß activity. TGF-ß displays a biphasic effect on FGF-induced endothelial cell proliferation; at low concentrations it acts in cooperation with bFGF whereas at high doses it inhibits bFGF-induced proliferation (16). 2. THE ROLE OF FIBROBLAST GROWTH FACTORS IN PROSTATE CANCER Prostate cancer, which is the most commonly diagnosed malignant tumour in the Western world, could be cured with radical prostatectomy in its early stages. All other prostate tumours need to be treated with androgen ablation therapy. This therapy is only palliative and nearly all tumours progress to the therapyrefractory stage. Various experimental treatments for advanced disease have been tried with little success. Progression of prostate cancer involves alterations in expression and function of several positive and negative growth factors and their receptors, including those of the FGF family. The three peptides, aFGF, bFGF and, more recently, FGF-8 have been intensively investigated in prostate tumours with regard of their expression, interaction with the respective receptors, effects on growth in vivo and in vitro and regulation of angiogenesis and invasion. Importance of fibroblast-derived growth factors for prostate cancer was recognized in series of studies in which formation of LNCaP tumours in vivo was studied (17,18). The tumours were consistently induced when androgensensitive LNCaP cells were coinoculated with non-tumourigenic bone, rat

95 urogenital system mesenchymal and prostate fibroblasts. Coinoculation of LNCaP cells with a matrix absorbed with bFGF also resulted with induction of tumours (17). bFGF was shown to be a potent mitogen for LNCaP cells. However, the mitogenic effect of fibroblast-conditioned media was not eliminated by anti-bFGF antibodies thus indicating that other soluble factors contribute to interactions between fibroblasts and cancer cells. Early studies on FGFs in carcinoma of the prostate were focused on their expression in rat Dunning tumours. In the slow-growing and androgen-responsive cell line R3327 aFGF is predominantly produced whereas in the metastatic cell line AT-3 both aFGF and bFGF are expressed (19). Immunoreactivity for aFGF was not detected in stromal cells and was weak in both basal and luminal cells in the normal tissue. In normal prostate, strong staining for FGFR-1 was noted in stromal and endothelial cells of blood vesels as well as in basal cells (2022). Similarly, FGFR-2 was detected in endothelial and basal cells (22) whereas its appearance in normal epithelium is a matter of debate (22,23). In the Dunning tumour system progression is associated with changes in the expression of the FGFR-2 gene (24). In early stages, the predominant receptor form is the exon IIIb-isoform which preferentially binds stroma-derived keratinocyte growth factor (KGF) whereas during tumour progression the IIIc form which has a high affinity for bFGF is expressed. This change in receptor expression is associated with progression from a mixed stromalepithelial to a stromal independent

96

Culig et al

Chapter 6

6. FGF and cancer metastasis phenotype, a phenomenon that was observed in several studies on human prostate cancer (25,26). Tumour progression is further characterized with the development of the bFGF autocrine loop, activation of the FGFR-1 and enhanced expression of FGF-3 and -5. Splicing of the FGF-R2 was also investigated in the human cell lines LNCaP, PC-3 and DU-145. That study revealed a similar correlation between FGFR expression and malignant phenotype. DU-145 cells, which exhibit a more aggressive growth express the IIIc isoform whereas the IIIb isoform was found in LNCaP cells (27). Androgensensitive xenografts DUKAP-1 and –2 also displayed IIIb isoform expression and the IIIc isoform was found in the androgen-independent tumour DU 9479. Taken together, these results consistently show that the loss of FGFR-2 IIIb isoform might be a common event in prostate cancer progression. There was the lack of observable FGFR-2 protein in PC-3 cells (27). In one of initial studies on bFGF in prostate cancer cell lines high amounts of mRNA for the FGFR in PC-3 cells were reported. The cells did not respond to exogenous bFGF, most probably because of an autocrine production of the growth factor (28). An information as to functional significance of the two receptors in prostate, FGFR-1 and FGFR2, was obtained recently (29,30). Those researchers showed that nearly all rat prostate malignant cells express the R1 whereas the expression of the R2 was reduced. Following transfection into nonmalignant cells, FGFR-1 kinase accelerated progression to the malignant phenotype and promoted a mitogenic response. In contrast, reexpression of the FGFR-2 IIIb led to the re-establishement

97 of stromal dependency of a tumour and to epithelial differentiation. In clinical specimens derived from prostate cancer tissue both FGF receptors were detected with tendency towards a stronger expression in cancers with a higher Gleason grade (22) (Figure 1). These results were obtained by immunohistochemistry and, in some specimens, confirmed by Western blot. In this context, it is important to know that an increased expression of the FGFR-1 correlates with advanced tumour stage in head and neck squamous carcinomas and in non-small cell lung carcinoma (31,32). However, an information regarding expression of FGFR isoforms in clinical material is not available because of the lack of respective antibodies. In human prostate cancer cell lines, bFGF expression was consistently demonstrated in androgen-insensitive PC3 and DU-145 cells whereas no bFGF was detected in LNCaP cells (28,33,34). Since bFGF lacks its signaling sequence it is poorly secreted into the supernatants (33). There is a little information about the regulation of bFGF in prostate cancers available. In PC-3 cells down-regulation of bFGF is induced by interferons alpha and beta (35). The cell lines LNCaP and DU-145 respond to exogenous bFGF (28). Concentration of bFGF in carcinomatous tissue was measured by ELISA and it was found that the mean concentration in prostate cancers was markedly increased relative to controls (22). Immunohistochemical studies, however, revealed the expression of bFGF only in stromal cells. No immunoreactivity was seen in the neoplastic epithelium even in those samples in which high bFGF concentration was measured by ELISA.

98

Culig et al

Chapter 6

6. FGF and cancer metastasis These findings differ from those reported by Cronauer et al. and Dorkin et al. (33,36). Cronauer and associates found differences in bFGF-staining between healthy epithelium and cancer tissue; staining was more intense in the malignant tissue. Elevated bFGF values were measured in sera of men with prostate cancer but no correlation with clinical stage, Gleason score or prostate volume was found (33,37). On the basis of our immunohistochemical results, we have postulated that the source of elevated bFGF are tumour cells themselves. The importance of FGFs for prostate tumourigenesis has been studied in the PNT1a cell line transfected with bFGF cDNA (38). This transfection led to the acquisition of anchorage-independent growth that was not seen in the parental cells. bFGF injected orthotopically into the rat ventral prostate, preferentially stimulated growth of the epithelial compartment (39). In addition to aFGF and bFGF, the expression of FGF-8 in prostate cancer has become a subject of investigation. FGF-8 was first identified as a protein that mediates androgen-induced growth of the mouse mammary Shionogi carcinoma cell line SC-3 (40). It was shown that application of the neutralizing monoclonal antibody against FGF-8 abolishes effect of androgen on cell growth. FGF-8 is overexpressed in malignant prostatic epithelium and its expression correlated with Gleason score and clinical stage (36) (41). The mechanisms of regulation of FGF-8 in advanced carcinoma of the prostate have not been investigated but they may involve non-steroidal activation of the AR. It became clear that AR activity is modulated by a number of peptide hormones in ligand-independent and synergistic manner (for review see (42)). In addition to bFGF and FGF-8, a

99 high percentage of clinical cancers (86.1%) express aFGF (36,40). The correlation between expression of aFGF and FGF-8 was statistically significant. In the study by Wang and associates, a correlation between the expression of the androgen receptor and FGF-8 was found, consistent with findings previously reported by Tanaka et al. (40,43). The most important fibroblast growth factor produced by prostatic stromal cells is KGF which is, in contrast to bFGF, efficiently secreted. Based on tissue and organ culture studies, it was proposed that KGF and FGF-10 act as andromedins, mediators of of androgen-induced growth of epithelial cells (44,45). There is a consensus that KGF is expressed in stromal tissue whereas its receptor was detected exclusively in the epithelium (46). However, there is no supportive evidence that KGF expression decreases following castration and therefore its role in vivo is not clear at present (47). FGF-9, which was recently reported to be expressed in stromal cells, is mitogenic for both stromal and epithelial cells, but its alterations were not found in tumouradjacent stroma. In addition, tumour cells themselves do not express FGF-9 (48). An important mechanism by which FGFs regulate prostate tumour metastases is induction of promatrilysin, which belongs to matrix metalloproteinase (MMP) enzymes (Fig. 2) (49). MMPs are enzymes which degrade extracellular matrix proteins and their expression is elevated in prostate cancer cells (50). aFGF is a potent inducer of promatrilysin even at low concentrations. Similar effects were observed also with higher concentrations of bFGF, FGF-8 and -9 (49). A possible reason for the observed strong effect of aFGF is its high affinity for all FGF receptors. Induction of promatrilysin by FGFs was observed in

100

Culig et al

cancer cells but not in normal prostatic epithelial cells. A direct effect of bFGF on rat prostate cancer cell motility was demonstrated (51). bFGF utilizes different signaling pathways in various cell lines; these pathways include activation of mitogenactivated protein kinase or protein kinase C signal transduction cascades and elevation of intracellular calcium. Involvement of protein kinase C alpha and epsilon in bFGF signaling was demonstrated in rat prostate cancer cells (52). In those cells, application of aFGF and bFGF antisense oligonucleotides specifically inhibited proliferation (52). We have investigated bFGF signal transduction in connection with androgen receptor protein down-regulation by bFGF in LNCaP cells (53). It was investigated whether this down-regulation is mediated by elevation in intracellular calcium. In LNCaP cells, addition of calcium ionophore leads to reduction of expression of the androgen receptor (54). However, we have demonstrated that bFGF does not increase calcium levels in LNCaP cells (53). It is interesting to note that bFGF in LNCaP causes both stimulation of proliferation and inhibition of androgen receptor expression. 3. FIBROBLAST GROWTH FACTORS AND BLADDER CANCER Bladder cancer is the second most common malignancy of the genitourinary tract. If the metastases are present, the prognosis is poor. Therefore, there is a need to investigate the mechanisms of metastasis and invasion in that disease. In this context, expression and function of FGFs has been studied. In normal bladder, strong bFGF staining was observed in the basal lamina of the transitional epithelium and smooth

Chapter 6 muscle of the bladder wall. In a high percentage of superficial bladder cancer tissue bFGF could not be detected (55). In contrast, bFGF mRNA and protein expression are high in invasive tumours (56). In these cancers bFGF contributes to the degradation of extracellular matrix and angiogenesis and could therefore be considered therapeutical target. The proposed mechanism involves invasion of the basement membrane by tumoursecreted enzymes such as urokinase, cathepsin D and heparanase, involvement of the muscle layers of the bladder and release of bFGF from intracellular stores. bFGF that is released stimulates angiogenesis and is detected in blood vessels. In bladder cancer bFGF also upregulates the expression of MMPs (57) thus facilitating metastasis. There was no significant difference in proliferation and cell motility between parental and bFGF cDNA-transfected cells. However, the production and activity of MMP-2 and -9 were considerably higher in the cells expressing bFGF cDNA and were reduced in the cells treated with bFGF antisense oligonucleotides. In parallel, in vitro metastatic potential increased in the cells expressing bFGF. In concordance with those data, De Boer et al. reported only a minor effect of bFGF on growth of a human transitional cell cancer cell line and similar data were obtained with organoid-like primary cultures (58). A clinical study showed that the levels of MMP-2 and –9 are higher in invasive bladder tumours than in superficial ones (59). NBT-II cells transfected with aFGF cDNA were not more tumourigenic than control cells but these tumours were highly vascularized with a high density of enlarged vessels (60). Those findings also indicate that strong angiogenesis is not sufficient to accelerate tumourigenesis.

6. FGF and cancer metastasis Those aFGF expressing cells give rise to rapidly growing well-vascularized tumours (61). It also promoted motility of bladder cancer cells (62). It was recently postulated that nuclear bFGF confers metastatic properties on rat bladder carcinoma cells by a mechanism which does not involve FGF receptor (63). Systemic administration of interferonalpha inhibits the expression of bFGF mRNA and protein by human transitional cell cancer. This down-regulation is associated with inhibition of angiogenesis and tumour growth in vivo (64). Human umbilical vein endothelial cells, in medium containing interferon-alpha, showed a significant inhibition of branching. Daily application of interferonalpha in bladder cancers implanted in nude mice inhibited tumour growth, vascularization and expression of bFGF and MMP-9 (65). Bladder cancer cell lines strongly induced endothelial cell migration in the in vitro assay (66) whereas normal urethral cells were antiangiogenic. Antiangiogenic effect of bladder cell lines was abolished with neutralizing antibodies to vascular endothelial growth factor and bFGF. Thrombospondin-1 was identified as a compound mainly responsible for antiangiogenic effect of conditioned media from normal bladder cells. Secretion of thrombospondin-1 was downregulated by bladder cancer cells. KGF stimulated proliferation of bladder cancer cells and the expression of its receptor was reported (67). Progression of bladder tumours is associated with a decreased expression of the receptor IIIb, as revealed in clinical specimens (68). Thus, there is a similar trend in prostate and bladder cancer regarding expression of the FGFR-2 IIIb. In bladder cancer, there is a strong association between urinary bFGF and

101 clinical parameters (69,70). It was found that FGF-like activity measured in urine samples correlates with tumour volume (71). Urinary bFGF was elevated in 67% of locally active bladder cancer patients, a percentage that was higher than that in several other malignancies (kidney, prostate, breast) (72). The source of the elevated bFGF are most probably proliferating capillary endothelial cells. aFGF was also detected in urine of bladder cancer patients and its expression correlated with the stage of the disease (73). During embryogenesis and neoplastic transformation, bladder epithelium changes its state of differentiation. This epithelium-to-mesenchyme differentiation is mediated by aFGF and not by related FGFs (74). The changes in plasticity of bladder NBT-II cells are monitored by immunohistochemical staining of the desmosomal protein desmoglein. aFGF treatment causes disappearance of desmoglein. 4. EXPRESSION AND FUNCTION OF FIBROBLAST GROWTH FACTORS IN RENAL CANCER Renal cell carcinomas are characterized by hypervascularity, rapid metastasizing and poor prognosis. Therapeutical options for this malignancy are very limited. There was an interesting observation that human omental adipose tissue bFGF demonstrates greater angiogenic and mitogenic activity than either benign or cancerous tissue bFGF (75). It is known that renal cancer correlates with obesity, in particular in females (76). bFGF cDNA-transfected renal cancer cell lines were more invasive than controls (77). Those cell lines also show an increased MMP-2 production and formed more than 10 times as many metastatic nodules in lungs as non-transfected or

102

Culig et al

control cells. In that study, endogenous expression of bFGF fails to stimulate cell proliferation. bFGF was detected in conditioned media from the RC29 renal cancer cells which also respond to the exogenously added growth factor (78). Thus, at least some renal cancers might be stimulated by bFGF in an autocrine fashion. In human renal cancer cell cultures bFGF is inversely regulated to cell density (79). The expression levels of bFGF in renal cancer mRNA and protein were higher than those measured in normal tissue (80). High bFGF levels were measured in kidney tumour tissue and in urine. The expression of bFGF in renal cell carcinoma and urinary bFGF inversely

Chapter 6 correlated with patient survival (72,81). bFGF expression in the tumours is transient and dependent on the organ environment (82). Renal cell cancer metastatic cells were injected into the kidney or subcutis of nude mice. High vascularization was observed only in tumours injected into the kidney and urine of these animals contained higher bFGF levels. The studies summarized in the present review show that FGFs are important for progression of urological cancers. A number of experimental therapies targeting angiogenesis is being developed and it is expected that these treatments could be evaluated in near future.

References 1. Gospodarowicz D. Purification of a fibroblast

8. Partanen J, Makela TP, Eerola E, Korhonen J,

growth factor from bovine pituitary. J Biol Chem 1975;250:2515-2520. Böhlen P, Baird A, Esch F, Ling N, Gospodarowicz D. Isolation and partial molecular characterization of pituitary fibroblast growth factor. Proc Natl Acad Sci USA 1984;81:5364-5368. Thomas KA, Rios-Cadelore M, Fitzpatrick S. Purification and characterization of acidic fibroblast growth factor from bovine brain. Proc Natl Acad Sci USA 1984;81:357-361. Abraham JA, Whang JL, Tumolo A, Mergia A, Friedman J, Gospodarowicz D, Fiddes JC. Human basic fibroblast growth factor: nucleotide sequence and genomic organization. EMBO J 1986;5:2523-2528. Gemel J, Gorry M, Ehrlich GD, MacArthur CA: Structure and sequence of human FGF8. Genomics 1996;35:253-257. Miyamoto M, Naruo K, Seko C, Matsumoto S, Kondo T, Kurokawa T. Molecular cloning of a novel cytokine cDNA encoding the ninth member of the fibroblast growth factor family, which has a unique secretion property. Mol Cell Biol 1993; 13:4251-4259. Beer HD, Florence C, Dammeier J, McGuire L, Werner S, Duan DR. Mouse fibroblast growth factor 10: cDNA cloning, protein characterization, and regulation of mRNA expression. Oncogene 1997;15:2211-2218.

Hirvonen H, Claesson-Welsh L, Alitalo K. FGFR-4, a novel acidic fibroblast growth factor receptor with a distinct expression pattern. EMBO J 1991;10:1347-1354. Müller R, Bravo J, Burckhardt J. Induction of c-fos gene and protein by growth factors precedes activation of c-myc. Nature 1984;312:716-720. Tsuda T, Hamamori Y, Yamashita T, Fukumoto Y, Takai Y. Involvement of three intracellular messenger systems, protein kinase C, calcium ion and cyclic AMP, in the regulation of c-fos gene. FEBS Lett 1986;208:39-42. Tsuda T, Kaibuchi K, Kawahara Y, Fukuzaki H, Takai Y. Induction of protein kinase C activation and Ca2+ mobilization by fibroblast growth factor in Swiss 3T3 cells. FEBS Lett 1985;191:205-210. Takeyama Y, Tanimoto T, Hoshijima M, Kaibuchi K, Ohyanagi H, Saitoh, Y, Takai Y. Enhancement of fibroblast growth factorinduced diacylglycerol formation and protein kinase C activation by colon tumour-promoting bile acid in Swiss 3T3 cells. Different modes of action between bile and phorbol ester. FEBS Lett 1986;197:339-343. Abe K, Saito H. Neurotrophic effect of basic fibroblast growth factor is mediated by the p42/p44 mitogen-activated protein kinase

2.

3.

4.

5.

6.

7.

9.

10.

11.

12.

13.

6. FGF and cancer metastasis

14.

15.

16. 17.

18.

19.

20.

21.

22.

23.

24.

cascade in cultured rat cortical neurons. Brain Res Dev Brain Res 2000;122:81-85. Magnaldo I, L´Allemain G, Chambard JC, Moenner M, Barritault D, Pouyssegur J. The mitogenic signaling pathway of fibroblast growth factor is not mediated through phosphoinositide hydrolysis and protein kinase C activation in hamster fibroblasts. J Biol Chem 1986;261:16916-16922. Kanda S, Hodgin MN, Woodfield RJ, Wakelam MJ, Thomas G, Claesson-Welsh L. Phosphatidylinositol -kinase-independent S6 kinase activation by fibroblast growth

factor receptor-1 is important for proliferation

but not differentiation of endothelial cells. J

Biol Chem 1997;272:23347-23353.

Rifkin DB, Kojima S, Abe M, Harpel JG. TGFbeta: structure, function, and formation. Thromb Haemost 1993;70:177-179. Gleave ME, Hsieh JT, Gao CA, von Eschenbach AC, Chung LW. Acceleration of human prostate cancer growth in vivo by factors produced by prostate and bone fibroblasts. Cancer Res 1991;51:3753-3761. Gleave ME, Hsieh JT, von Eschenbach AC, Chung LWK. Prostate and bone fibroblasts induce human prostate cancer growth in vivo: implications for bidirectional tumour-stromal cell interaction in prostate carcinoma growth and metastasis. J Urol 1992;147:1151-1159. Mansson PE, Adams P, Kan M, McKeehan WL. Heparin-binding growth factor gene expression and receptor characteristics in normal rat prostate and two transplantable rat prostatic tumours. Cancer Res 1989;49:24852494. Sinowatz F, Amselgruber W, Lincoln D, Sasse J, Kolle S, Plendl J, Kayser K. Role of basic fibroblast growth factor in prostatic tumours. Nutrition 1995;11:619-621. Hamaguchi A, Tooyama I, Yoshiki T, Kimura H. Demonstration of fibroblast growth factor

receptor-I in human prostate by polymerase

chain reaction and immunohistochemistry.

Prostate 1995;27:141-147.

Giri D, Ropiquet F, Ittman M. Alterations in expression of basic fibroblast growth factor (FGF) 2 and its receptor FGFR-1 in human prostate cancer. Clin Cancer Res 1999;5:10631071. Story MT, Hopp KA, Molter M, Meier DA. Characteristics of FGF-receptors expressed by stromal and epithelial cells cultured from normal and hyperplastic prostates. Growth Factors 1994;10:269-280. Yan G, Fukabori Y, MacBride G, Nikolaropoulous S, McKeehan WL. Exon switching and activation of stromal and

103

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

embryonic fibroblast growth factor (FGF)-FGF receptor genes in prostate epithelial cells accompany stromal independence and malignancy. Mol Cell Biol 1993; 13:4513-4522. Gao J, Isaacs JT. Development of an androgen receptor-null model for identifying the initiation site for androgen stimulation of proliferation and suppression of programmed (apoptotic) death of PC-82 human prostate cancer cells. Cancer Res 1998;58:3299-3306. Olapade-Olaopa EO, MacKay EH, Taub NA, Sandhu DP, Terry TR, Habib FK. Malignant transformation of human prostatic epithelium is associated with the loss of androgen receptor immunoreactivity in the surrounding stroma. Clin Cancer Res 1999;5:569-576. Carstens RP, Eaton JV, Krigman HR, Walther PJ, Garcia-Blanco MA. Alternative splicing of fibroblast growth factor receptor 2 (FGF-R2) in human prostate cancer. Oncogene 1997;15:3059-3065. Nakamoto T, Chang CS, Li AK, Chodak GW. Basic fibroblast growth factor in human prostate cancer cells. Cancer Res 1992;52:571577. Feng S, Wang F, Matsubara A, Kan M, McKeehan WL. Fibroblast growth factor receptor 2 limits and receptor 1 accelerates tumourigenicity of prostate epithelial cells. Cancer Res 1997;57:5369-5378. Matsubara A, Kan M, Feng S, McKeehan WL. Inhibition of growth of malignant rat prostate tumour cells by restoration of fibroblast growth factor receptor 2. Cancer Res 1998;58:15091514. Dellacono F, Spiro J, Eisma R, Kreutzer D. Expression of basic fibroblast growth factor and its receptors by head and neck squamous carcinoma tumour and vascular endothelial cells. Am J Surg 1997;174:540-544. Volm M, Koomagi R, Mattern J, Stammler G. Prognostic value of basic fibroblast growth factor and its receptor (FGFR-1) in patients with non-small cell lung carcinomas. Eur J Cancer 1997;33:691-693. Cronauer MV, Hittmair A, Eder IE, Hobisch A, Culig Z, Ramoner R, Zhang J, Bartsch G, Reissigl A, Radmayr C, Thurnher M, Klocker H. Basic fibroblast growth factor levels in cancer cells and in sera of patients suffering from proliferative disorders of the prostate. Prostate 1997;31:223-33. Hepburn PJ, Griffiths K, Harper ME. Angiogenic factors expressed by human prostatic cell lines: effect on endothelial cell growth in vitro. Prostate 1997;33:123-132. Singh RK, Gutman M, Bucana CD, Sanchez R, Llansa N, Fidler IJ. Interferons alpha and beta

104

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

Culig et al down-regulate the expression of basic fibroblast growth factor in human carcinomas. Proc Natl Acad Sci USA 1995;92:4562-4566. Dorkin TJ, Robinson MC, Marsh C, Neal DE, Leung HY. aFGF immunoreactivity in prostate cancer and its co-localization with bFGF and FGF8. J Pathol 1999;189:564-569. Meyer GE, Yu E, Siegal JA, Petteway JC, Blumenstein BA, Brawer MK. Serum basic fibroblast growth factor in men with and without prostate carcinoma. Cancer 1995:76:2304-2311. Ropiquet F, Berthon P, Villette JM, Le Brun G, Maitland NJ, Cussenot O, Fiet J. Constitutive expression of FGF2/bFGF in non-tumourigenic human prostatic epithelial cells results in the acquisition of a partial neoplastic phenotype. Int J Cancer 1997;72:543-547. Marengo SR, Chung LW. An orthotopic model for the study of growth factors in the ventral prostate of the rat: effects of epidermal growth factor and basic fibroblast growth factor. J Androl 1994; 15:277-286. Tanaka A, Furuya A, Yamasaki M, Hanai N, Kuriki K, Kamiakito T, Kobayashi Y, Yoshida H, Koike M, Fukayama M. High frequency of fibroblast growth factor (FGF) 8 expression in clinical prostate cancers and breast tissues, immunohistochemically demonstrated by a newly established neutralizing monoclonal antibody against FGF 8. Cancer Res 1998;58:2052-2056. Dorkin TJ, Robinson MC, Marsh C, Bjartell A, Neal DE, Leung HY. FGF8 over-expression in prostate cancer is associated with decreased patient survival and persists in androgen independent disease. Oncogene 1999:18:27552761. Culig Z, Hobisch A, Hittmair A, Peterziel H, Cato ACB, Bartsch G, Klocker H. Expression, structure, and function of androgen receptor in advanced prostatic carcinoma. Prostate 1998:35:63-70. Wang Q, Stamp GW, Powell S, Abel P, Laniado M, Mahony C, Lalani EN, Waxman J. Correlation between androgen receptor expression and FGF8 mRNA levels in patients with prostate cancer and benign prostatic hypertrophy. J Clin Pathol 1999:52:29-34. Yan G, Fukabori Y, Nikolaropoulos S, Wang F, McKeehan WL. Heparin-binding keratinocyte growth factor is a candidate stromal-to-epithelial-cell andromedin. Mol Endocrinol 1992:6:2123-2128. Lu W, Luo Y, Kan M, McKeehan WL. Fibroblast growth factor-10. A second candidate to stromal to epithelial cell

Chapter 6

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

andromedin in prostate. J Biol Chem 1999;274:12827-12834. De Bellis A, Crescioli C, Grappone C, Milani S, Ghiandi P, Forti G, Serio M. Expression and cellular localization of keratinocyte growth factor and its receptor in human hyperplastic prostate tissue. J Clin Endocrinol Metab 1998;83:2186-2191. Nemeth JA, Zelner DJ, Lang S, Lee C. Keratinocyte growth factor in the rat ventral prostate: androgen-independent expression. J Endocrinol 1998;156:115-125. Giri D, Ropiquet F, Ittmann M. FGF9 is an autocrine and paracrine prostatic growth factor expressed by prostatic stromal cells. J Cell Physiol 1999:180:53-60. Klein RD, Maliner-Jongewaard MS, Udayakumar TS, Boyd JL, Nagle RB, Bowden GT. Promatrilysin expression is induced by fibroblast growth factors in the prostatic carcinoma cell line LNCaP but not in normal primary prostate epithelial cells. Prostate 1999:41:215-223. Pajouh MS, Nagle RB, Breathnach R, Finch JS, Brawer MK, Bowden GT. Expression of metalloproteinase genes in human prostate cancer. J Cancer Res Clin Oncol 1991; 117:144150. Pienta KJ, Isaacs WB, Vindivich D, Coffey DS. The effects of basic fibroblast growth factor and suramin on cell motility and growth of rat prostate cancer cells. J Urol 1991:145:199-202. Hrzenjak M, Shain SA. Fibroblast growth factor-2 and TPA enhance prostate cancer cell proliferation and activate members of the ras and PKC signal transduction pathways. Recept Signal Tranduct 1997:7:207-219. Cronauer MV, Nessler-Menardi C, Klocker H, Maly K, Hobisch A, Bartsch G, Culig Z. Androgen receptor protein is down-regulated by basic fibroblast growth factor in prostate cancer cells. Br J Cancer 2000:82:39-45. Gong Y, Blok LJ, Perry JE, Lindzey JK, Tindall DJ. Calcium regulation of androgen receptor expression in the human prostate cancer cell line LNCaP. Endocrinology 1995:136:2172-2178. O´Brien T, Cranston D, Fuggle S, Bicknell R, Harris AL. Two mechanisms of basic fibroblast growth factor-induced angiogenesis in bladder cancer. Cancer Res 1997:57:136-140. Allen LE, Maher PA. Expression of basic fibroblast growth factor and its receptor in an invasive bladder carcinoma cell line. J Cell Physiol 1993:155:368-375. Miyake H, Yoshimura K, Hara I, Eto H, Arakawa S, Kamidono S. Basic fibroblast

6. FGF and cancer metastasis

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

growth factor regulates matrix metalloproteinases production and in vitro invasiveness in human bladder cancer cell lines. J Urol 1997;157:2351-2355. De Boer WI, Vermeij M, Gil Diez de Medina S, Bindels E, Radvanyi F, van der Kwast T, Chopin D. Functions of fibroblast and transforming growth factors in primary organoid-like cultures of normal human urothelium. Lab Invest 1996;75:147-156. Davies B, Waxman J, Wasan H, Abel P, Williams G, Krausz T, Neal D, Thomas D, Hanby A, Balkwill F. Levels of matrix metalloproteases in bladder cancer correlate with tumour grade and invasion. Cancer Res 1993;53:5365-5369. Jouanneau J, Plouet J, Moens G, Thiery JP. FGF-2 and FGF-1 expressed in rat bladder carcinoma cells have similar angiogenic potential but different tumourigenic properties in vivo. Oncogene I997;14:671-676. Jouanneau J, Moens G, Montesano R, Thiery JP. FGF-1 but not FGF-4 secreted by carcinoma cells promotes in vitro and in vivo angiogenesis and rapid tumour proliferation. Growth Factors 1995;12:37-47. Jouanneau J, Moens G, Bourgeois Y, Poupon MF, Thiery JP. A minority of carcinoma cells producing acidic fibroblast growth factor induces a community effect for tumour progression. Proc Natl Acad Sci USA 1994;91:286-290. Okada-Ban M, Moens G, Thiery JP, Jouanneau J. Nuclear 24 kD fibroblast growth factor (FGF)-2 confers metastatic properties on rat bladder carcinoma cells. 1999. Dinney CPN, Bielenberg DR, Perrotte P, Reich R, Eve BY, Bucana CD, Fidler IJ. Inhibition of basic fibroblast growth factor expression, angiogenesis, and growth of human bladder carcinoma in mice by systemic interferon-alpha administration. Cancer Res 1998;58:808-814. Slaton JW, Perrotte P, Inoue K, Dinney CP, Fidler IJ. Interferon-alpha-mediated downregulation of angiogenesis-related genes and therapy of bladder cancer are dependent on optimization of biological dose and schedule. Clin Cancer Res 1999;5:2726-2734. Campbell SC, Volpert OV, Ivanovich M, Bouck NP. Molecular mediators of angiogenesis in bladder cancer. Cancer Res 1998;58:1298-1304. De Boer WI, Houtsmuller AB, Izadifar V, Muscatelli-Groux B, van der Kwast TH, Chopin DK.. Expression and functions of EGF, FGF and family members and their receptors in invasive human

105

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

transitional-cell-carcinoma cells. Int J Cancer 1997;71:284-291. Diez de Medina SG, Chopin D, El Marjou A, Delouvee A, LaRochelle WJ, Hoznek A, Abbou C, Aaronson SA, Thiery JP, Radvanyi F. Decreased expression of keratinocyte growth factor receptor in a subset of human transitional cell bladder carcinomas. Oncogene 1997;14:323-330. Chodak GW, Hospelhorn V, Judge SM, Mayforth R, Koeppen H, Sasse J. Increased levels of flbroblast growth factor-like activity in urine from patients with bladder or kidney cancer. Cancer Res 1988;48:2083-2088. O´Brien T, Smith K, Cranston D, Fuggle S, Bicknell R, Harris A. Urinary basic flbroblast growth factor is elevated in patients with bladder cancer and benign prostate hypertrophy. Br J Urol 1995;76:311-314. Nguyen M, Watanabe H, Budson AE, Richie JP, Folkman J. Elevated levels of the angiogenic peptide basic flbroblast growth factor in urine of bladder cancer patients. J Natl Cancer Inst 1993;85:241-242. Nguyen M, Watanabe H, Budson AE, Richie JP, Hayes DF, Folkman J. Elevated levels of an angiogenic peptide, basic flbroblast growth factor, in the urine of patients with a wide spectrum of cancers. J Natl Cancer Inst 1994;86:356-361. Chopin DK, Caruelle JP, Colombel M, Palcy S, Ravery V, Caruelle D, Abbou CC, Barritault D. Increased immunodetection of acidic flbroblast growth factor in bladder cancer, detectable in urine. J Urol 1993;150:1126-1130. Valles AM, Boyer B, Badet J, Tucker GC, Barritault D, Thiery JP. Acidic flbroblast growth factor is a modulator of epithelial plasticity in a rat bladder carcinoma cell line. Proc Natl Acad Sci USA 1990;87:1124-1128. Mydlo JH, Kral JG, Macchia RJ. Preliminary results comparing the recovery of basic flbroblast growth factor (FGF-2) in adipose tissue and benign and malignant renal tissue. J Urol 1998; 159:2159-2163. Chow WH, Mclaughlin JK, Mandel JS, Wacholder S, Niwa S, Fraumeni JFJ. Obesity and risk of renal cell cancer. Cacer Epidemiol Biomark Prevent 1996;5:17-21. Miyake H, Hara I, Yoshimura K, Eto H, Arakawa S, Wada S, Chihara K, Kamidono S. Introduction of basic flbroblast growth factor gene into mouse renal cell carcinoma enhances its metastatic potential. Cancer Res 1996;56:2440-2445. Mydlo JH, Zajac J, Macchia RJ. Conditioned media from a renal cell carcinoma cell line

106

Culig et al

demonstrates the presence of basic fibroblast growth factor. J Urol 1993; 150:997-1001. 79. Singh RK, Llansa N, Bucana CD, Sanchez R, Koura A, Fidler IJ. Cell density-dependent regulation of basic fibroblast growth factor expression in human renal cell carcinoma cells. Cell Growth Differ 1996;7:397-404. 80. Eguchi J, Nomata K, Kanda S, Igawa T, Taide M, Koga S, Matsuya F, Kanetake H, Saito Y. Gene expression and immunohistochemical localization of basic fibroblast growth factor in renal cell carcinoma. Biochem Biophys Res Commun 1992;183:937-944.

Chapter 6 81. Nanus DM, Schmitz-Drager BJ, Motzer RJ, Lee AC, Vlamis V, Cordon-Cardo C, Albino AP, Reuter VE. Expression of basic fibroblast growth factor in primary human renal tumours: correlation with poor survival. J Natl Cancer Inst 1994;85:1597-1599. 82. Singh RK, Bucana CD, Gutman M, Fan D, Wilson MR, Fidler IJ. Organ site-dependent expression of basic fibroblast growth factor in human renal cell carcinoma cells. Am J Pathology 1994;145:365-374.

Chapter 7 INSULIN-LIKE GROWTH FACTOR AXIS ELEMENTS IN BREAST CANCER PROGRESSION

Emilia Mira, Rosa Ana Lacalle, Carlos Martínez-A. and Santos Mañes Department of Immunology and Oncology, Centro National de BiotecnologíaCSIC, Campus de Cantoblanco, Universidad Autónoma de Madrid, Madrid, Spain

Key words: Insulin-like growth factor, chemotaxis, cell polarization, tumour invasion, breast cancer Abstract:

Insulin-like growth factors (IGF) exhibit very potent mitogenic activity, promote cell survival, and have insulin-like functions essential for embryogenesis and postnatal growth physiology. Attention has recently focused on the role of IGF in neoplastic transformation, growth, and dissemination in several cancer types, including human breast cancer. Neoplastic cells are characterized by relative growth autonomy, a consequence of the constitutive expression of IGF and their receptors involved in autocrine loops. This chapter will focus on the molecular mechanisms underlying IGF action in tumourigenesis, in particular its chemoattractant activity and its relevance in tumour motility, both of which lead to invasion. Several of the IGF-induced cellular changes will be highlighted, such as cell polarizatiom, adhesion and detachment, as well as proteolysis induction. Finally, we will summarize the significance of IGF system components as prognostic markers in human breast cancer, and discuss the possible therapeutic considerations encompassed by these factors.

active oncogenes and the loss of specific tumour suppressor genes. In addition to these genetic factors, however, tumour growth is controlled by hormonal cues that also regulate cell proliferation in nonpathological conditions. The steroid hormone estrogen, an important hormonal stimulus for breast development, is also one of the most potent mitogens for breast

1. OVERVIEW Progression from a benign, noninvasive in situ carcinoma to a malignant, invasive breast carcinoma is a complex multistage process, in which uncontrolled cell proliferation is a hallmark of the disease. This progression is believed to involve a number of genetic events, including activation of dominant107

W.G. Jiang et al. (eds.), Growth Factors and their Receptors in Cancer Metastasis, 107–166. © 2001 Kluwer Academic Publishers. Printed in the Netherlands.

108

Mira et al

cancer cells. Expression of the mediator of estrogen-induced proliferation, the estrogen receptor, is lost in advanced breast carcinomas, suggesting that other tumour-produced factors provide these cells with a hormone-independent proliferation status. Over the past two decades, considerable evidence has accumulated indicating that a family of growth factors, the insulin-like growth factors I and II (IGF-I and IGF-II), are critical elements in maintaining and supporting the progression of breast cancer cells. IGF-I and IGF-II are members of a multifunctional family of proteins that elicit diverse effects in a variety of biological processes in cell culture systems and have a broad range of functions in the embryo, fetus and adult. The name reflects their deep structural relationships with insulin and emphasizes their growth-promoting activity; nonetheless, IGF are more pleiotropic in their biological activities than insulin. In fact, the IGF were discovered as three entirely independent activities: the sulfation factor activity (SFA), later called somatomedin, as it mediates the message of growth hormone (GH) stimulating somatic growth in vivo (1), the nonsuppressible insulin-like activity (NSILA), since their hypoglycemic activity could not be abrogate with antiinsulin antibodies (2), and the multiplication-stimulating activity (MSA) as they promoted cell replication in vitro (3). Although initially IGF were considered as endocrine hormones, in the 70’s was observed that specific cell lines produced their own IGF and, hence, they may also act as autocrine/paracrine factors. The interplay between autocrine and endocrine roles of IGF may be of pivotal importance in the incidence and progression of different human cancers.

Chapter 7 All the IGF effects, both growthpromoting and metabolic, are mediated through interaction of these molecules with specific cellular receptors. Both IGFI and IGF-II bind to the insulin (IR), type I IGF (IGF-1R) and type II IGF receptors (IGF-2R) with differential affinity. The growth-promoting activities are mediated by the IGF-1R, a receptor tyrosine kinase that also binds to insulin, whereas both the IGF-1R and the IR probably mediate the metabolic effects. This promiscuity between ligands and receptors is only disrupted by the IGF-2R that binds specifically IGF-II. This receptor appears to function by sequestering IGF-II binding to the IGF-1R or the IR and, hence, must be considered an antagonist of the IGF-II function. Another relevant step in understanding IGF biological activity was the finding that these polypeptides do not exist in serum or in other body fluids in the free form; instead, IGF are bound to specific carrier proteins, termed IGFBP (IGF binding protein). At present, seven distinct IGFBP have been identified and cloned in humans, some of which are expressed by specific cell types or under determined circumstances. These IGFBP increase IGF half-life in the vasculature, but their precise function as regulators of IGF bioavailability at the cellular level remains unclear. Initial observations indicated that IGF forms bound to IGFBP do not readily interact with the IGF receptors, explaining why the high IGF concentrations in human serum do not cause hypoglycemia. Even though different experimental evidences argue to an inhibitory role of IGFBP, it is becoming evident that IGFBP are also required for the proper IGF biological activity at the cell level. Indeed, enhancement of the IGF-induced proliferative activity by IGFBP in different cell types has been reported.

7. IGF in breast cancer progression Different evidences suggest that this enhancing effect may be explained by the limited proteolysis of IGFBP at or in close proximity to the cell membrane. The proteolytic cleavage of IGFBP would lead to a decreased affinity for IGF and to the controlled release of the growth factor in the pericellular enviroment. We must thus consider the IGF axis composed of IGF ligands, IGF receptors, IGFBP, and IGFBP proteases. 2. INSULIN-LIKE GROWTH FACTOR AXIS COMPONENTS 2.1

Structure of IGF-I and IGF-II

Mature IGF-I and IGF-II are singlechain polypeptides of 70 and 67 amino acids, respectively, with 62% overall sequence identity (4-5). Due to structural identity with insulin, the IGF polypeptide chain has been divided into four domains arranged as B-C-A-D. IGF A- and Bdomains have 45% sequence identity with insulin A- and B-chain; however, the connecting peptide C is shorter than the proinsulin C-chain, and the carboxylterminal D-domain extension is exclusive to IGF. Another parallel between IGF and insulin structure is the presence of three intrachain disulphide bonds arranged in the same disposition as in insulin, i.e., two connecting B- and A-domains and one intra-A-domain (6). Moreover, the IGF are synthesized as preproproteins with signal peptides of about 25 amino acids at the N-terminus of the B-domain, and further extensions of 35-85 residues at the C-terminus of the D-domain, termed the E-peptide (7). Although the signal peptide and E-domain are deleted sequentially by post-translational processing before secretion, the presence of different IGF-II proforms has been reported in serum (8). Interestingly, these “large” IGF-II forms have the same mitogenic activity as the

109 processed species (8-10). In Northern hybridization studies, cDNA probes detect several IGF-I mRNA species, of which the predominant forms appear to be 7.0-8.0, 4.6-4.7, 1.7-2.1 and 1.0-1.2 kilobases (kb) in length. Similarly, IGF-II mRNA species of 3.4-4.0, 2.2, 1.61.75 and 1.1-1.2 kb have been described (11-12). These multiple mRNA arise from alternative splicing, since IGF genes have a discontinuous structure; IGF-I gene contains five exons that span at least 45 kb (13), whereas IGF-II consists of seven exons spanning more than 16 kb (14). The physiological significance of these different splicing forms nonetheless remains unknown. As mentioned above, IGF interact with four different molecular species: the IGF-1R, the IGF-2R, the IR and the IGFBP. The IGF domains involved in these interactions have been defined mostly by homologous scanning mutagenesis, replacing IGF-I domains with those corresponding to homologous areas of insulin (15). With this approach, it was found that IGF-I residues 1-3 and 49-51 are important for IGFBP and IGF2R binding, whereas amino acids 21, 23, 24, 44 and tyrosines 31 and 60 are required for binding to the IGF-1R (1522). Extrapolation of these amino acids on a partially-resolved three-dimensional structure of IGF-I (23) indicates that the IGFBP and the IGF-1R binding surfaces are on opposite sides of the IGF-I molecule, suggesting that IGFBP-IGF complexes may bind to the IGF-1R on the cell surface. However, these ternary complexes have not been demonstrated to date. Most studies indicate that IGFBP abrogates IGF-I biological activity; and this inhibition is not observed for IGF-I mutants with low affinity for IGFBP (24). Structure-function analysis with a panel of 28 monoclonal antibodies covering the

110

Mira et al

entire exposed IGF-I surface indicated an overlap of the IGF-I domains involved in IGF-1R and IGFBP binding (25). These results are compatible, since scanning mutagenesis provides direct information about the residues involved in ligandreceptor interaction, but not about the steric hindrance produced by each receptor. Even though IGF amino acids interacting with IGFBP and IGF-1R are on distinct sides of the molecule, the IGFBP and IGF-1R footprints thus overlap on the IGF-I surface. 2.2. IGF binding proteins (IGFBP) and IGFBP proteases Circulating IGF are tightly bound in serum and in the extracellular milieu to soluble receptors termed IGFBP. Seven members of the IGFBP family have been cloned so far (26). IGFBP-1 to -6 are structurally related proteins of 216-289 amino acids, with highly conserved cysteine-rich N- and C-terminal domains involved in ligand binding; the domains are linked by a central portion that is very dissimilar among each IGFBP. These six IGFBP bind with high affinity to both IGF-I and -II, but do not bind insulin (27). Conversely, IGFBP-7/mac25 binds to insulin with high affinity and with very low affinity to IGF-I or IGF-I (28). Most IGF-I and IGF-II is found in circulation as 50 and 150 kDa complexes with IGFBP (29-30). These complexes prolong IGF half-life in circulation; indeed, injection of radioactively-labeled IGF-I in rats showed that the 150 kDa form remained in the bloodstream for 3-4 h, compared with 10 minutes recorded for unbound IGF (31). The 150 kDa form, which is the main IGF reservoir in serum (32), is a ternary complex formed by IGF, IGFBP-3 and a glycoprotein termed acidlabile subunit (ALS) (33-34). The presence of ALS in the 150 kDa complex

Chapter 7 impedes its crossing the endothelial barrier, increasing IGF half-life in the bloodstream (35). Other IGFBP in serum, such as IGFBP-1, -2 and -4, are able to traverse the vasculature, and hence transport IGF from the circulation to peripheral tissues (27, 36). The endocrine effects of IGF on somatic growth or tumourigenesis may therefore be regulated by controlling its availability through the formation of IGF complexes with specific IGFBP. In addition to their role as IGF reservoirs and carrier proteins in the circulation, IGFBP act as inhibitors (3739) or enhancers (40-41) of IGF biological activity at cell level. The inhibitory effects of IGFBP have been attributed to competitive scavenging of IGF peptides away from the IGF-1R (4243). The enhancer mechanism is poorly understood, and usually involves binding of the proteolytic cleavage of membranebound IGF/IGFBP complex, which in turn results in a decrease in IGFBP binding affinity (44-45). Several authors have suggested that the IGFBP enhancing effect is mediated by the sequestering of IGF from the extracellular milieu, which prevents the negative IGF-1R feedback that occurs at high IGF concentrations (41, 46-47). Alternatively, IGFBP may facilitate IGF binding to IGF-1R by anchoring the ligands in close proximity to their cell receptors. In fact, binding of the IGF/IGFBP complex to the cell surface is critical in the IGFBP enhancing effect. IGFBP-3 blocks cell growth if added simultaneously with IGF-I, but increases IGF-I-induced mitogenesis if added prior to IGF-I stimulation (48). In addition to IGFBP binding to the cell membrane, other processes such as proteolytic cleavage of IGFBP, should account to enhance IGF activity. Several proteases, including serine proteases,

7. IGF in breast cancer progression cathepsins and matrix metalloproteinases (MMP), are reported to use IGFBP as substrates. In all cases, this proteolytic cleavage results in a drastic decrease in IGFBP affinity for IGF and in the controlled release of the growth factors (49-52). Furthermore, is reported that IGF-I upregulates MMP-2 synthesis in carcinoma cells (53), establishing a positive feedback loop between IGF activity and the IGFBP proteolytic activities. Strikingly, some of the proteases acting on IGFBP are also associated to the cell surface (54-56); this explains the observation that factors that increase IGFBP association to the plasma membrane also contribute to IGFBPenhanced IGF activity (57). Several recent lines of evidence in various cell systems have suggested that the IGFBP, especially IGFBP-3, may have more active, IGF-independent roles in cell growth regulation. In support of this hypothesis, high affinity IGFBP-3 binding to the surface of various cell types and IGF-independent direct inhibition of monolayer growth have been shown; both are presumably induced by specific interaction with cell membrane proteins that function as an IGFBP-3 receptor (5859). Interest in IGFBP has increased since it was found that IGFBP-3 could act as an antineoplastic agent. IGFBP-3 inhibits cell proliferation in fibroblasts that express or lack the IGF-1R (60-61), indicating that IGFBP-3-mediated growth suppression is independent of IGF-1R action, IGFBP-3 has also been identified as a p53-regulated target gene (62). Taken together, these data suggest that the IGFBP-3 gene may be a tumour suppressor. The newest member of the family, IGFBP-7, is also suggested to be a tumour suppressor binding protein. The IGFBP-7/mac25 cDNA was originally cloned from leptomeningial cells and subsequently

111 reisolated by differential display as a sequence expressed preferentially in senescent human mammary epithelial cells (63). mac25 mRNA is detectable in a wide range of normal human tissues, with decreased expression in breast, prostate, colon, and lung cancer cell lines, suggesting that IGFBP-7 may function as a growth-suppressing factor (26, 63). 2.3. The IGF receptors There are at least three receptors in the IGF axis, including the insulin receptor (IR), the type-1 IGF receptor (IGF-1R) and the type-2 IGF receptor (IGF-2R). The IGF-1R is the most active in terms of cellular proliferation, and crossreacts with all three ligands; it shares several functional and structural characteristics with the IR, but the IGF-1R subunit is ten-fold more mitogenic than the IR (64). Supraphysiological insulin concentrations also activate the IGF-1R, and the mitogenic effects of insulin at microgram concentrations are probably exerted through binding to the IGF-1R (65). The IGF-1R belongs to the tyrosine kinase receptor family (66) and its amino acid sequence is 70% identical to that of the IR (67). It is a heterotetrameric glycoprotein composed of two ligandbinding subunits, entirely extracellular, and two transmembrane subunits linked by disulfide bonds, which have the enzymatic activity (68-69). The receptor is synthesized as a single preprotein of 1,367 amino acids that is cleaved to generate two half receptors, which are joined by disulfide bonds between the a subunits to form the mature receptor. The IGF-1R binds to all three ligands with distinct affinities; IGF-I and IGF-II bind to the receptor at nanomolar concentrations, whereas insulin binds with 100-fold lower affinity. Ligand binding to subunits triggers autophosphorylathe

112

Mira et al

tion of subunits (70-71) by an intramolecular trans mechanism similar to that used by other receptors (72-73). The IGF-2R is a single chain, membrane-spanning glycoprotein that is identical to the cation-independent mannose-6-phosphate receptor (74). The mature human receptor consists of a large extracellular domain and a short intracytoplasmic domain (75). The extracellular domain contains the ligandbinding domain (mannose-6-phosphate and IGF-II) and comprises 15 repeated domains. The intracellular portion has no tyrosine kinase activity and has been implicated mainly in trafficking among different subcellular compartments. The function of IGF-2R remains puzzling. It does not stimulate cell proliferation; in fact, IGF-2R gene deletion causes body weight increase in mice, and blocking of IGF-II binding to the receptor does not alter its mitogenicity (76-77). This supports the argument that the IGF-2R is a specific downregulator of IGF-II (the receptor binds neither IGF-I nor insulin). IGF availability may thus be regulated in two ways, through IGF-2R, which controls IGF-II levels, and through IGFBP, which may serve as storage sites for both IGF-I and IGF-II ligands (insulin does not bind to IGFBP). An additional receptor in the IGF axis must be considered: the insulin-receptor related receptor (IRRR), which has substantial identity with the IGF-1R and the IR (78). In chimeric constructs with IR or IGF-1R subunits, the IRRR tyrosine kinase domain was shown to be mitogenic, but the ligand involved in the activation of this receptor has not yet been identified. In fact, serum does not activate the IRRR (79), suggesting that the IRRR may be activated by an unknown ligand through a strictly autocrine mechanism.

Chapter 7 3. ON THE PHYSIOLOGICAL ROLE OF IGF-I AND IGF-II

IGF-1R are present in a wide variety of cell types and mediate most in vitro IGF-I and IGF-II effects, as well as insulin effects when this hormone is present at sufficiently high concentrations. In general, the IGF effects are either acute anabolic effects on protein and carbohydrate metabolism, or longer-term effects on cell replication and differentiation. There are excellent reviews surveying the effects of IGF in multiple specific cell types, as well as its role as a mediator of GH-induced somatic growth (8, 27, 80). We will focus only on the mitogenic and anti-apoptotic activities of IGF, since the balance between these two processes regulates the growth of any tissue or tumour cell. 3.1.

Mitogenic effects

The IGF molecules appear to regulate cell proliferation in both epithelial and mesenchymal tissues. They act as a nontissue-specific permissive mitogen, required for optimal proliferative responses to highly tissue-specific trophic factors such as platelet-derived growth factor (PDGF), fibroblast growth factor (FGF) or thyroid-stimulating hormone (TSH). They consequently perform the function of a general “progression factor” for cells previously stimulated by a more specific “competent” factor (81). This has been formally demonstrated using cells derived from mice with a targeted disruption of the IGF-1R genes (IGFThese cells do not grow in serumfree medium supplemented with growth factors that sustain the same type of cells derived from wild type littermates (82). These cells do proliferate in medium containing 10% fetal calf serum, although the doubling time is much

7. IGF in breast cancer progression slower than for wild type fibroblasts (109 h and 44 h, respectively). IGF-I mitogenic activity has also been reported in vivo. Pygmies and Laron-type dwarfism are the consequence of low IGF-I levels (8). The correlation between IGF-I levels and body size also extends to dogs and mice. Transgenic mice overexpressing the IGF-I gene confirm the importance of the IGF axis in body growth. GH overexpression causes increased growth in transgenic mice (83). Although circulating GH levels at birth were ten-fold higher, accelerated growth was not evident until two weeks of age, at which time serum IGF-I was elevated compared with control mice. Transgenic mice expressing human IGF-I in liver and other tissues also showed enhanced growth, although to a lesser extent than that observed in GH transgenic mice (84). This may reflect the fact that GH also upregulates other components of the IGF axis, as IGFBP-3 (8). Further in vivo evidence indicating the essential role of the IGF axis in cell proliferation and survival has been elucidated using gene knockout (KO) techniques (Table I). Mice lacking IGF-II were only 60% the size of their wild-type littermates at birth, but a normal postnatal growth rate was observed (85). Equivalent impairment in fetal growth was also observed in homo- and heterozygote mice inheriting a paternal mutant IGF-II allele (86), indicating paternal imprinting of the IGF-II gene. Deletion of IGF-2R genes showed that, despite the puzzling signaling pattern, this receptor is critical

113 for fetal growth in rodents. In mice, IGF2R maps to an imprinted locus termed Tassociated maternal effect (Tme), which is essential for viability. Mice lacking IGF2R die of major cardiac anomalies in the perinatal period; they are 125-130% the size of their wild-type littermates (77, 87) and show up to a 2.6-fold increase in circulating IGF-II levels (77), indicating again that IGF-2R does not mediate IGFII-induced growth effects. Mice lacking IGF-I have markedly diminished postnatal growth as well as fetal retardation (88-90). IGF-I KO mice showed reduced viability, with perinatal death attributed to muscular hypoplasia and decreased lung maturation. The dwarfism observed in IGF-I KO mice is increased in the double IGF-I/IGF-IIdeficient mice, with 30% of wild-type size at birth. Deletion of the IGF-1R also results in postnatal death due to muscular hypoplasia (88-89). Deletion of early signaling molecules downstream of the IGF-1R also result in growth impairment. KO mice for the insulin receptor substrate-1 (IRS-1) show a 40% reduction in postnatal growth (91). Interestingly, deletion of IRS-2, an IRS-1 homologue, does not affect prenatal and postnatal growth (92), indicating that IRS-1 is the main element in IGF-1 R-mediated somatic growth. These results indicate that expression of both IGF ligands and receptors is required for normal embryonic growth, and highlight the potential of the IGF system in promoting cell proliferation and survival in many different tissues.

114

3.2.

Mira et al

Anti-apoptotic effects

The protective effect of the activated IGF-1R on cell survival has been known for some years, mainly in the central nervous system (CNS), where IGF-I inhibits low potassium-induced apoptosis in cerebellar granule neurons (93). Other reports have indicated the role of IGF-I as an anti-apoptotic agent. For example, IGF-I inhibits apoptosis induced following IL-3 withdrawal in pro-B cells (25), probably via a mechanism involving receptor-induced phosphatidyl inositol-3 kinase (PI-3K) activity (unpublished observations); it also protects cells from cmyc-induced apoptosis (94). All of these reports deal with the protective effects of the ligands, but it has become apparent that the receptor is the limiting factor. The

Chapter 7

protective effect of IGF-I following growth factor withdrawal, observed in IL3-dependent B cell lines (which usually express a high IGF-1R receptor number), is lost in other B cell types in which this receptor is absent or present at low levels (unpublished observations). An increase in IGF-1R receptor number may rescue cells from various apoptotic agents, including FAS (95). This is also true in vivo, where a decrease in IGF-1R using an antisense strategy or the use of a dominant negative receptor give rise to massive apoptosis of tumour cells (see below). IGF-1R apparently achieves its protective effect by interfering with ICE and ICElike proteins (96), which form the most commonly used apoptotic pathway. A specific role has recently been reported for the transcription factor nuclear

7. IGF in breast cancer progression in mediating IGF-I survival effects (9798). 4. IGF-1R-MEDIATED SIGNALING TRANSDUCTION PATHWAYS 4.1. Initiation of the IGF-1R-triggered signaling pathways

As mentioned above, IGF-2R activation appears not to mediate IGF-II biological activities. We will therefore assume that IGF-I and IGF-II mediate most of their cellular activities through activation of the IGF-1R. Activation by ligand binding causes rapid tyrosine phosphorylation of the IGF-1R and the subsequent recruitment and phosphorylation of insulin receptor substrates (IRS) (99). Four members of the IRS family have been cloned in mammals (100-102): IRS-1 and IRS-2 are widely expressed, IRS-3 is restricted to adipose tissue, and liver; and IRS-4 is expressed in thymus, brain and kidney. Targeted disruption of each IRS gene in mice suggests that IRS-1 and IRS-2 coordinate essential effects of IGF mitogenic and metabolic activities (92, 103-104). In contrast, disruption of IRS-3 and IRS-4 genes results only mild effects, suggesting that they have redundant roles in IGF action (105-106). IRS proteins have a conserved Nterminus composed of adjacent pleckstrin homology and phosphotyrosine binding domains that mediate IRS association to the activated IGF-1R. The C-terminus is tyrosine phosphorylated at multiple sites, creating docking motifs for proteins with src homology 2 (SH2) domains. A wide diversity of signaling proteins containing SH2 domains bind to distinct IRS docking sites, based both on recognition of the phosphorylated tyrosine and on the

115 sequence surrounding this modified tyrosine (100), thus triggering different cell responses (Fig. 1). Despite their high sequence similarity, IRS proteins are an important site of signal redundancy and diversity. In cell-based assays, IRS-1 and IRS-3 activate PI-3K more strongly than IRS-2, whereas IRS-4 barely activates PI3K (101, 107). IRS-2 nonetheless plays a major role in PI-3K activation in murine liver cells (92); IRS proteins may therefore coordinate distinct IGFsignaling pathways in a cell type-specific manner. Concurring with this, it has been reported that myoblast and adipocyte proliferation use mainly the p21 ras/mitogen-activated protein (MAP) kinase pathway, whereas breast tumours and brain capillary cells proliferate in response to signals by the pathway (108). 4.2 Cross-communication between the IGF-1R and other cell receptors

Although IRS proteins were originally described as specific signaling molecules for the IR and IGF-1R, recent reports have found them at the crossroads of several intracellular pathways. Indeed, IRS-1 appears essential for IL-4 stimulation of hematopoietic cells (109-110), for signaling by the GH receptor and interferon (111), interacts with the JAK family of transducing molecules (112), with certain integrins (113) and with G protein-coupled receptors (114). The central role of IRS proteins in various signaling pathways suggests intensive crosstalk between IGF and other transducing systems. This crosscommunication is essential for several IGF-I-induced cell activities, as discussed below.

116

Mira et al

Chapter 7

7. IGF in breast cancer progression 4.2.1. IGF-1R/estrogen receptor crosstalk: a mitogenic connection IGF function is regulated by other factors important in breast cancer development, such as estrogens. These hormones are required for the development of female secondary sexual characteristics, including mammary glands. Estrogens are potent mitogens for a number of breast cancer cells in vitro, and transcriptional upregulation of the estrogen receptor (ER) is one the earliest and most common defects associated with breast tumourigenesis (115). The IGF-1R and the ER are usually co-expressed, and the two signaling systems are engaged in complex functional crosstalk that controls cell proliferation (116). Estrogens and IGF synergistically stimulate proliferation in most ER-positive breast cancer cell lines; this synergism is achieved in part since ER upregulates IGF-1R and IRS-1 levels (117-118). Simultaneously, ER downregulates expression of IGF antagonists such as IGFBP-1 and IGF-2R (119-120). ER therefore increases IGF1R-mediated signaling by upregulating IGF-1R and IRS levels; this finally results in enhanced tyrosine phosphorylation and MAPK activation after ligand binding, leading to an increase in cell proliferation. IGF signaling may also regulate ERinduced cell proliferation through a mechanism involving the transcriptional activation of endogenous ER in an estrogen-independent manner (121-122). Inhibition of IGF-I signaling by addition of IGFBP produces a decrease in the expression of ER-inducible genes (123), indicating that a basal level of IGF-1R activation is required for maximal ER activity. This interplay between IGF and ER signaling has also been shown in vivo. Indeed, upregulation of IRS-1 expression was observed after estrogen stimulation of MCF-7 cells grown as xenografts in

117 athymic mice; conversely, estrogen withdrawal resulted in a dramatic decrease in tumour growth as well as much lower IRS-1 expression in the tumour cells (118). In ER-positive cells, therefore, both estrogen and IGF regulate proliferation in an interdependent manner, such that the blockage of one of these pathways results in inhibition of the other. This is of the utmost relevance, since the major treatment modality for ER-positive breast cancers is the use of estrogen antagonists such as tamoxifen. Antiestrogen inhibits IGF-I-stimulated cell proliferation (123). Its mechanism of interaction with IGF signaling has yet to be defined, but antiestrogens decrease circulating IGF-I levels and increase IGFBP-3 expression (123-124). It has been also suggested that antiestrogens may modulate the activity of phosphatases that shut down IGF-1R signaling (116). IGF/ER crosstalk in ER-positive cells may also sensitize the cells to the action of other growth factors such as transforming a growth growth factor inhibitor in normal and ER-positive tumour cells, that nonetheless contributes to tumour progression in ER-negative cells (125). The mechanism of this differential activity in ERpositive and -negative cells is not understood, although it is clear that it occurs in conjunction with changes in gene expression. Indeed, ER-positive cells do not produce IGF in an autocrine manner and activation of the IGF pathway hence depends on the exogenous growth factor supplement. Conversely, ERnegative cells usually show aberrant expression of IGF, which is the primary stimulus leading to their proliferation (discussed below). expression in ER-positive is tightly regulated by both IGF and ER, whereas this control is lost in

118

Mira et al

Chapter 7

ER-negative cells (125). Breast tumour progression thus depends on crosstalk between stimulatory and inhibitory factors that regulate cell proliferation. 4.2.2.IGF/chemokine invasive connection

crosstalk:

an

One important biological function of IGF-I as well as of other growth factors is to induce cell invasion. As will be discussed’ in detail later, achieving this invasive ability requires extreme rearrangements of cell structure that culminate in the acquisition of a motile, polarized cell phenotype. It has been established in different cell types that cell polarization requires trimeric protein signals. Activation of G proteins releases two potential downstream signaling molecules, a GTP-bound subunit that interacts directly with effectors, and a subunit that regulates an array of free targets, including ion channels and enzymes (126). dimers mediate cell polarization in mammalian cells (127, itself is not required in 128), whereas this process (129). In fact, we found that IGF-I-induced cell chemotaxis of the breast adenocarcinoma MCF-7 is inhibited by treatment of the cells with pertussis toxin, which specifically prevents activation and hence, release. The convergence of G-protein-coupled receptors (GPCR) and receptor tyrosine kinase (RTK) signaling pathways is supported by observations that the receptors of PDGF, PDGFR (130), epidermal growth factor, EGFR(131), and IGF-1R (132) are tyrosine phosphorylated after GPCR activation. Transactivation of GPCR signaling by IGF-I, although reported (133-135), remains a matter of controversy (136-138). Indeed, pertussis toxin is reported to inhibit both the IGF-Iinduced opening of a calcium-permeable

cation channel and MAPK activation in rat cerebellar granule neurons and NG108 neuronal cells (134-135). Furthermore, MAPK activation by IGF-I was inhibited in fibroblasts expressing subunit binding proteins (133). The question arises as to the mechanisms by which IGF-I, which signals through a receptor tyrosine kinase, triggers these G protein-mediated events. As mentioned earlier, an association has been suggested between IRS proteins with proteins following IGF-I stimulation (114). This interaction may occur through the pleckstrin homology domain in the subunits (139). It has likewise been to reported that activated IR recruits the signaling complex, in a manner similar to classical GPCR (140). Based on coprecipitation experiments, it has been suggested that heterotrimeric proteins may associate constitutively with the IGF1R (141); the authors proposed that IGF1R activation upon IGF-I binding leads to release of the subunit as consequence activation. Although these reports of point to an intracellular connection between IGF-1R and trimeric proteins, the physical association between these molecules does not explain how G proteins are activated. In fact, seems to be phosphorylation of independent of IGF-I stimulation (141), suggesting that activation does not occur through tyrosine phosphorylation. Moreover, there is no evidence for the mechanism by which the IGF-IR induces subunit release from the complex. We found that IGF-I stimulation of the ER-positive human breast adenocarcinoma MCF-7 cell line induces the specific transactivation of the coupled chemokine receptor CCR5, triggering its tyrosine phosphorylation and protein recruitment (142). This

7. IGF in breast cancer progression transactivation occurs via a mechanism involving upregulation of gene expression and secretion of RANTES, the natural CCR5 ligand. Neutralizing anti-RANTES antibodies or drugs that block protein secretion abrogate the IGF-I-induced CCR5 transactivation, indicating that an extracellular step is required to achieve chemokine receptor activation. Our data indicate that IGF-I-induced chemokine receptor activation involves two transmembrane signaling events: first, IGF-I-mediated IGF-1R activation induces synthesis of RANTES, which is secreted to the extracellular milieu, and second, extracellular RANTES binds to and activates CCR5, which recruits trimeric and promotes activation and release. A similar mechanism was recently demonstrated in RTK transactivation by GPCR receptors. Indeed, GPCR-induced EGFR transactivation is mediated by a paracrine loop involving EGF precursor release from the membrane (143), suggesting that this extracellular cross-communication between RTK and GPCR may operate in both directions. The chemokines belong to a superfamily of low molecular weight chemotactic proteins implicated mainly in leukocyte activation and chemotaxis (144). The IGF-I-induced CCR5 transactivation observed thus provides a new mechanism that links chronic inflammation and tumourigenesis (145). The constitutive IGF-I secretion observed in invasive ER-negative breast tumour cells may increase proinflammatory chemokine concentrations in the environment, providing the cells with a built-in invasive capacity. In fact, prevention of chemokine receptor activation by drugs such as pertussis toxin, by neutralizing anti-chemokine receptor antibodies, or by transdominant

119 negative chemokine receptor mutant overexpression, abolishes IGF-I-induced MCF-7 cell chemotaxis in vitro (142). Using the RANTES antagonist AOPRANTES (146-147) and the dominant negative CCR5 mutant (148150), we observed that chemokine signals do not promote cell motility but, rather are required for IGF-I-induced cell polarization (unpublished results). These results stress again the absolute requirement for heterotrimeric activation, probably through mediated signaling, to achieve the polarized motile cell phenotype. The cross-communication between IGF-1R and chemokine receptors is also important for in vivo tumour invasion; indeed, RANTES gene expression is upregulated in invasive human breast carcinomas (151-152). Our preliminary results from intravasation assays in chicken chorioallantoic membrane (CAM) (153) using highly invasive (ER-negative) and non-invasive (ER-positive) tumour cells concur with these data. The intravasation process constitutes an important event in tumour dissemination and is related to the invasive ability of the tumour (154). Overexpression of the transdominant negative mutant in ER-negative tumour cells, which produce large amounts of IGF-I in an autocrine fashion (155), significantly decreased intravasation capacity of the cells. Conversely, overexpression of RANTES in the low invasive MCF-7 cell line resulted in an increased intravasation capacity of these cells. Together, these results suggest that IGF-chemokine crosstalk may have a fundamental role in tumour progression, probably by modulating the invasiveness of tumour cells.

120

Mira et al

5. ROLE OF THE IGF AXIS IN TUMOUR FORMATION AND GROWTH It has been clear for many decades that the phenomenon called “cancer” is in some way related to defects in the mechanisms that control normal cell proliferation. The use of gene transfer (transfection) allowed the transmission of the cancerous state from one cell to another using DNA as the vehicle, indicating that distinct cell determinants of transformation, called oncogenes, resided in tumour cell DNA (156-158). In the ensuing years, evidence accumulated indicating that cell transformation paralleled genetic defects leading to the constitutive activation of oncogenes or the constitutive downregulation (or loss of function) of tumour suppressor genes. Nonetheless, cell transfection of oncogenes in vitro suggested that this relationship was not linear. For instance, ras-transformed 3T3 fibroblasts require progression factors found in serum to form colonies in soft agar. These progression factors were present in the serum of normal rats, but not in that derived from hypophysectomized rats; the factors were identified as IGF (159). Several retroviruses containing v-ras (vH-ras, v-K-ras) and v-mos oncogenes or growth factor receptor-derived oncogenes (v-erbB, v-fms) were shown to replace the epidermal keratinocyte proliferative requirement for EGF, a growth factor necessary for these cells in culture. Nevertheless, none of these oncogenes permitted keratinocyte growth in defined medium without IGF, indicating that IGF1R signals are important for cell proliferation independently of oncogene activation (160).

Chapter 7 Further evidence was obtained in vivo. Activation of H-ras is a frequent event for tumour initiation in thyroid follicular cells (161-162); however, in vivo induction of H-ras mutations by thyroid injection of retroviruses carrying a mutated ras, yields formation of few tumours. If the mitogenic activity of the gland is increased by stimulation with thyroidstimulating hormone besides with the introduction of ras mutations, the number of thyroid tumours is greatly enhanced (161, 163). These observations were explained by assuming that initial mutations leading to oncogene activation are not sufficient to induce cell proliferation at low primary mitogen levels. If clonal expansion of the oncogene-transformed cell occurs due to the influence of a physiological stimulus, however, new mutations may then take place, leading to a selective advantage for tumour growth. Later evidence indicated that growth factor signals, in particular those from the IGF-1R, are required for efficient tumour implantation and progression. The concept that IGF is a second signal in cell transformation has become stronger. This evidence will be discussed in this section, together with epidemiological data in humans that support this view. 5.1. IGF-1R signals are necessary for cell transformation 5.1.1.IGF-1R is required for cell transformation in vitro and in vivo

It has long been known that overexpression and/or constitutive activation of the IGF-1R leads to liganddependent neoplastic transformation (164), i.e., ability to form colonies in soft

7. IGF in breast cancer progression

agar and/or to produce tumours following injection into nude mice. Transformation can, however, be considered a common outcome of gene product overexpression; overexpression of many growth factor receptors, such as PDGF, EOF, FGF, insulin and CSF-1, as well as protooncogenes, activated oncogenes, signal transducing molecules, and glycolytic enzymes, results in cell transformation. Two findings distinguish IGF-1R. First, cells (derived from mice with a targeted IGF-1R gene disruption) are refractory to transformation by certain viral and cellular oncogenes that readily transform mouse embryo fibroblasts expressing IGF-1R. These oncogenes include the SV40 large T antigen (82), activated ras, or a combination of ras and T antigen (165), bovine papilloma virus E5 protein (166), or growth factor receptor overexpression (167). Second, the transformed phenotype of tumour cell lines can be reversed to a non-transformed

121

phenotype by decreasing IGF-1R number or by interfering with its function. Several approaches have been used, including antisense expression plasmids or oligonucleotides of IGF-II (168), IGF-I (169-170) or IGF-1R (171-177), antibodies to the IGF-1R (178-179), and dominant negative mutants of the IGF-1R (180-181). Cells from a highly metastatic human breast cancer line (182) or a murine carcinoma (183) carrying an antisense IGF-1R construct show a delay in tumour formation and a reduction in metastatic capacity. Inhibition of IGF-1R in metastatic breast cancer cells using a dominant negative mutant prevented metastasis to the lungs, liver and lymph nodes when cells were injected into the mammary fat pad (184); all these findings are summarized in Table II. The list is incomplete, but clearly shows that IGF-1R targeting can reverse the transformed phenotype in several tumour types of human and rodent origin.

122

Mira et al

5.1.2. IGF-1R domains involved cel l transformation The introduction of modified IGF-1R into cells has provided information on essential domains and critical amino acids involved in different aspects of receptor function, including cell transformation. Mutants of both the and have mutants been reported. The provided information about the IGF-1R residues involved in ligand specificity. This section will concentrate on IGF-1R which mutated in the intracellular furnish some information on the IGF-1R domains involved in the recruitment of cell signaling molecules and the engagement of the IGF-1R in cell transformation. The intracellular kinase domains of IGF-1R and IR are 84% identical and activate mostly the same signal intermediates (IRS-1, Shc, etc.). Despite this considerable structural identity, analysis of chimeric IGF-1R/IR receptors in fibroblasts revealed that the is ten times more potent IGF-1R than that of IR in promoting DNA synthesis (64). Conversely, metabolic effects such as glucose transport are exerted with the same potency. However, a chimeric IGF-1R containing the last 112 amino acids of the insulin receptor in place of the 121 homologous residues showed only a modest increase in glycogen synthesis, PI-3K activity, and MAP kinase activity compared with the wild type IGF1R; DNA synthesis was nevertheless unaltered (185). This experiment suggests that the mitogenic activity does not reside in the C-terminal part of the IGF-1R βchain. Analysis of deletion and substitution yielded new mutations within the information on IGF-1R function. A point mutation at lysine 1003 (the ATP binding

Chapter 7 site) results in an IGF-1R that has essentially lost its function (73). Replacement of tyrosines 1131, 1135 and 1136 (the tyrosine kinase domain) by phenylalanine results in a markedly reduced level of autophosphorylation and in the abolition of both mitogenicity and transforming activity (186). Mutation of tyrosine 1136 leads to reduced DNA synthesis, whereas cells expressing IGF1R mutated in tyrosines 1131 and 1135 can still replicate (187). In contrast, substitution of any of these residues blocks colony formation in soft agar. These results demonstrate that each tyrosine in this cluster is not equivalent, and indicate that a fully functional receptor is required for anchorage-independent growth, but not for mitogenesis. Other mutations affect only selected receptor functions. For instance, tyrosine 950 is essential for binding and phosphorylation of IRS-1 (99), although it does not influence autophosphorylation or ligand-dependent internalization. This mutation abolishes mitogenic and transforming activities, but still protects murine hematopoietic cells from apoptosis induced by IL-3 withdrawal (188). The conclusions drawn from the many mutants tested are that: (i) IGF-1R mitogenicity and transforming activity are located in different domains (189-191), indicating that there is at least one pathway for transformation that is additional to and distinct from the mitogenic pathway; (ii) the IGF-1R transforming domain is located between residues 1245 and 1310 (191), and (iii) tyrosine 1251 and the serines at 1280-1283 are extremely important for IGF-1R-induced cell transformation (190).

7. IGF in breast cancer progression 5.1.3. IGF-1R interaction with oncogenes and tumour suppressor genes The results described above provide compelling evidence that the IGF-1R or signals provided by this receptor are required for transformation by different oncogenes. It has also been demonstrated that IGF-1R can interact directly with oncogenes such as src. In cells transfected with temperature-sensitive v-src, the IGF1R is rapidly phosphorylated on tyrosine at permissive temperatures. Indeed, srcstimulated phosphorylation correlated with receptor activation, even in the absence of IGF-I (192). v-src is the only oncogene cells (165), able to transform probably due to its ability to interact with the IGF-1R signaling machinery, such as IRS proteins. A further corollary may be added, since oncogenes were observed to upregulate and tumour suppressors to downregulate several components of the IGF system (for an extensive discussion with examples of other receptors and oncogenes see ref. (193). Several reports have documented that transfection of active oncogenes such as H-ras (194-196), SV40 large T antigen (197) or c-myb (198) upregulate the autocrine or paracrine production of IGF-I and/or IGF-II through a transcriptional mechanism. Conversely, tumour suppressor genes such as WT1 (199-200) and (201) cause a decrease in receptor number at the transcriptional level. In addition to the control of IGF-1R levels, also regulates the expression of other IGF components. Wild-type p53 represses IGF-II and IR promoter transcription (202) and induces IGFBP-3 expression (62), which antagonizes the effects of IGF-I and -II. This is of utmost relevance, since about 30% of breast tumours show allelic loss in the gene

123 region, presumably exposing a single mutated allele that most often encodes a missense protein with altered or absent transactivating capacity (203). Loss of repression following mutations of may thus result IGF-1R and IGF-II overexpression (200). The fact that oncogenes and anti-oncogenes modify components of the IGF system is extremely interesting, as this suggests that they act through the regulation of growth factors and their receptors. Although signals from the IGF-1R are required for cell transformation, the IGF1R cannot be considered an oncogene. It is intriguing to ask whether other IGF components may be tumour suppressors. As discussed above, several lines of evidence in various cell systems suggest that IGFBP-3 and IGFBP-7/mac25 may be considered tumour suppressor factors (26, 58-59, 63). The case of IGFBP-3 is of special interest, since its growth suppressive activity appears to be independent of the IGF-1R activation by the IGF-I or IGF-II ligands (60-61). It has also been suggested that IGFBPassociated proteins, such as the ALS (see section 1.2), may have a fundamental role in controlling IGF activity in peripheral tissues. Certain IGF-II-secreting tumours, such as fibrosarcomas, rhabdosarcomas and leiomyosarcomas, produce severe tumour-associated hypoglycemia (204). Nonetheless, total IGF-I or IGF-II serum levels in these patients are not elevated and may even be reduced as compared with normal individuals (205). The hypoglycemic patients show marked abnormalities in serum IGFBP composition (205); the ternary complex ALS/IGFBP-3/IGF-I or IGF-II is virtually absent from the circulation of these individuals, even though IGFBP-3 is present (206). As mentioned earlier, this ternary complex may prevent the passage

124

Mira et al

of IGFBP/IGF complexes through the capillary membrane (27, 35-36). In the absence of ALS, IGFBP-3/IGF complexes may thus have greater access to peripheral targets. This mechanism to promote nonpancreatic tumour-induced hypoglycemia may also account for increased IGFinduced tumour cell proliferation. Several reports have indicated that IGF-2R may potentially be a tumour suppressor for liver cancer. IGF-2R expression is reduced in hepatocellular carcinomas (207), and loss of heterozygosity at the human IGF-2R locus has been found in approximately 70% of liver cancers (207). Tumours with inactivating mutations in the IGF-2R gene grew more rapidly than those with an intact IGF-2R allele, again suggesting that the IGF-2R may be implicated in the clearance and degradation of the IGF-II molecule. Recent observations indicate that IGF-2R levels are significantly lower in breast cancer cells than in normal epithelial cells (208), and downregulation of this receptor resulted in an increased tumour growth rate in vivo (209). All together, these results suggest that the IGF-2R may act as a putative tumour suppressor gene for liver and breast tumours, possibly by decreasing IGF-II biological activity through its binding to the IGF-1R (210). Nonetheless, these observations will require substantiation and development of experimental models to test the hypothesis directly. 5.2. IGF autocrine loops and tumour growth

Although the liver is the major synthesis site for the systemic circulating IGF-I pool in adulthood, there is also significant growth factor production by both epithelial and stromal cells (211). Many tissue-specific hormones, such as estrogen, adrenocorticotrophic hormone

Chapter 7 (ACTH), luteinizing hormone (LH) or thyroid-stimulating hormone (TSH), appear to induce or modulate local IGF-I expression in their target tissues (8). These findings promoted a shift in understanding of the role of IGF-I in growth regulation from an endocrine to a paracrine/autocrine mode. This autocrine/paracrine circuit is operative in many tissues of very diverse origin. In evolutionary terms, the nontissue-specific mitogenic activity of IGF constitutes a primitive mechanism providing independent regulation of many tissue types in higher organisms. This system has probably been maintained since it provides a mechanism that allows integration of multiple mitogenic signals to coordinate the final somatic growth of a tissue. Thus, for example, growing epithelial cells produce a non-specific secondary mitogen, such as IGF-I, which in turn stimulates the associated stromal elements that were not influenced by the tissue-specific primary mitogen. The IGF system is therefore justified in terms of cell-cell communication, which cannot be achieved by a convergence of intracellular signaling pathways. The IGF loop is thus a system for paracrine, rather than autocrine stimulation, that operates not only in a conventional heterotypic role coordinating stroma with epithelia, but also in a homotypic mode to coordinate the growth of the epithelium itself. Many tumour cell lines produce elevated levels of IGF-I and/or IGF-II (212). Proliferation of these cells is partially inhibited by interfering with the IGF/IGF receptor interaction (213-215), suggesting a role for the IGF autocrine loop in tumour cell growth. Using an antiIGF-I mAb that recognizes the IGF-I/IGF1R complex, we formally demonstrated the existence of this IGF autocrine loop, at least in human prostatic adenocarcinomas

7. IGF in breast cancer progression (155). Increased IGF secretion in tumours is not a consequence of a primary abnormality of the IGF genes themselves since IGF are not oncogenes, but represents constitutive activation as a result of inappropriate intracellular signal activation. In correlation with this, as indicated in the previous section, some oncogenes such as the SV40 large T antigen or c-myb modify IGF ligand expression at the transcriptional level. Interestingly, this relationship also works in the opposite direction, as IGF-I may also regulate the expression of some oncogenes (216). Many tumour cell lines are characterized by relative growth autonomy, hypothesized to be a consequence of the constitutive expression of growth factors and their receptors. During physiological growth, part of the normal cell response to the primary mitogen involves the autocrine induction of a secondary non-tissue-specific mitogen, such as IGF-I and/or IGF-II. In tumour cells, transformation by an oncogene produces constitutive activation of IGF secretion to the extracellular milieu. Expression of this autocrine loop in tumour cells is thus independent of, or independent of an increase in, primary mitogen stimulation. The obvious consequence of this abnormal activation is autonomous proliferation of the tumour. It must nonetheless be remembered that the IGF loop works not only in a homotypic mode, coordinating the growth of the tumour itself, but that it also operates in a heterotypic mode, coordinating tumour and stroma. In fact, IGF-I is also expressed in stromal zones of cancer tissues (217). Furthermore, breast cancers are infiltrated with IGF-IIexpressing stromal cells (218) and this stromal-derived growth factor production has been implicated in protecting pre-

125 neoplastic breast cancer cells from apoptosis (219). The IGF autocrine/paracrine loops may therefore coordinate different functions in tumour progression in a manner similar to which they coordinate the somatic growth of a tissue throughout fetal and childhood development. Another level of complexity is added, since IGF autocrine loops do not involve only one or two ligands (IGF-I and IGF-II) and one receptor (the IGF-1R), but are regulated by other components of the IGF axis. We will return to reveal the true complexity of this system. 5.3. Role of the IGF axis in the andtumour immune response

The ability of tumours to produce and respond to IGF-I may also modulate the immune response to the the tumour. The evasion of a host immune response to transformed cells is also pivotal in the development of established tumours. Several reports have described that the abolition of the IGF loop in tumour cells promotes a host anti-tumour immune response. This was first observed using rat C6 glioblastoma cells that express IGF-I and form rapidly-growing tumours in syngeneic animals. Subcutaneous injection of C6 cells transfected with an episomebased vector encoding anti-sense IGF-I cDNA caused the regression of established brain glioblastomas, although the antisense-bearing cells were injected in a distal site (169); wild type cells do not confer this resistance. These antitumour lioma-specific effects result from a gglioma-specific immune response involving lymphocytes (169), which is restricted to the cell type carrying the antisense cDNA (220). Similar results have been obtained using an antisense RNA for the IGF-1R (172). This is not particularly surprising, except that the animals become resistant to

126

Mira et al

challenge with wild type C6 cells even when they are pretreated with unrelated tumour cells, including tumour cells of other species (221). The mechanisms underlying this non-classical immune response are still unclear and require further investigation. 5.4. IGF as the second signal in tumourigenesis

Taking together all the observations discussed above, it becomes clear that the IGF axis is fundamental in the establishment and progression of tumours. This leads to the concept that IGF axis components may be a second signal in cell transformation (Fig. 2). As discussed above, oncogene-induced transformation is

Chapter 7 a highly inefficient process except when a simultaneous mitogenic stimulus is provided (163). Indeed, transgenic mice expressing the SV40 large T antigen in all Langerhans islets develop solid tumours in only 1-2% of them. If the SV40 large T antigen transgenic mice are crossed with IGF-II KO mice, the percentage of transformed islets as well as tumour malignancy is severely reduced (168). In addition, the tumours developed in SV40 large T antigen/ IGF-II KO mice showed a larger number of apoptotic cells than in SV40 large T antigen-expressing mice, indicating that IGF-II signals are required for the hyperproliferation and establishment of pancreatic tumours.

7. IGF in breast cancer progression In terms of autocrine/paracrine loops described in the previous section, the process leading to tumour formation is the final result of a homotypic cooperative loop (222). According to this, malignant transformation due to oncogene expression in a single cell would eventually result in the increased expression of a progression factor (IGF-I for instance); nevertheless, this mitogen will not reach the critical concentration required to promote proliferation of the transformed cell. If the oncogene transformation is followed by stimulation with a primary mitogen, however, both the oncogene-expressing cell and all neighboring untransformed cells will increase expression of the progression factor, which may now reach the threshold required to promote proliferation. At the end of this proliferation period for both cell types, the tumour would reach sufficient mass to maintain IGF-I above the minimal level required for cell proliferation; tumour growth would hence be independent of primary mitogen stimulation. Concurring with this hypothesis, growth of mammary tumour transplants is drastically reduced in the lit/lit mouse strain, which have extremely low circulating growth hormone (GH) and low IGF-I levels, compared with those transplanted in normal mice (223). Further, transgenic mice overexpressing IGF-II develop a diverse spectrum of tumours, including mammary cancer, at a higher frequency than control mice with normal IGF-II levels (224-225). Transgenic mice in which IGF-I expression was targeted specifically to the basal layer of epidermis likewise showed increased spontaneous tumour promotion (226). Epidemiological data indicate that cooperative homotypic IGF loops may

127 operate in promoting major human cancers. Acromegalic individuals have elevated circulating IGF-I levels and experience a high incidence of prostate and colorectal tumours (227). A weak, although consistent association has been reported between final adult height (indicated by elevated GH/IGF-I serum levels) and breast (228), prostate (229), colorectal (230), and hematopoietic (231) cancers. Two recent studies showed strong association between circulating IGF-I levels and the risk of breast and prostate cancers (232-233). The association between risk of prostate cancer and IGF-I was stronger in men over 60 years of age, when androgens may have less influence (233). This indicates that IGF-I acts not merely as a surrogate of steroid status but that it probably regulates steroid activity, as discussed above for estrogen. 5.5. Implications of the IGF axis in breast cancer incidence Experimental evidence and epidemiological data support the association of IGF in breast cancer development, and that intervention to reduce IGF levels may decrease proliferation of breast neoplasms. In breast cancer prevention strategies, IGF family members might therefore be considered potential therapeutic targets. Nonetheless, there is disagreement about the value of IGF as a prognostic marker, which will be summarized here. 5.5.1. Prognostic value of IGF and IGFBP The association between serum IGF-I levels and cancer risk has been suggested in several studies (234). Higher serum IGF-I levels are reported in breast cancer patients than in normal age-matched controls, and are associated with a poor prognosis (235). Stromal IGF-II levels are

128

Mira et al

above normal in 50% of invasive breast cancer cases, which correlates with increased cell proliferation (236). For breast cancer, IGF-I association was not apparent in postmenopausal women, but strong associations were observed for premenopausal cases (237) (reviewed in (238). Furthermore, in recent epidemiologic studies, relatively high plasma IGF-I and low IGFBP3 levels have been independently associated with greater risk of breast cancer among premenopausal women (232). According to this study, the relationship between IGF-I and risk of breast cancer may be greater than that of other established breast cancer risk factors, with the exception of a family history of breast cancer or a high density mammographic profile. IGF-I concentrations may be particularly relevant to breast cancer risk in premenopausal women, since estradiol enhances IGF-I action in the breast (239). Although an increase in serum IGF-I levels in breast cancer patients has been documented, causality has not been established. In a multivariate analysis, the conventional prognostic factors were superior to IGF-I expression in predicting disease outcome, although IGF-I expression in tumour stroma showed some independent prognostic significance in early phases of the disease (240). This may indicate that IGF-I in tumour stroma acts as a paracrine growth promoter for breast cancer cells, increasing tumour malignancy. These studies were nonetheless potentially limited by their retrospective design. Plasma IGF-I levels were evaluated after diagnosis in these cases, so that an effect of the tumour on IGF-I levels was not ruled out. In support of this possibility, surgical removal of the tumour lowers circulating IGF-I levels (241). When measured in tumour cytosol, IGF-I showed no correlation with other

Chapter 7 IGF family members (IGF-II, IGFBP-1, or IGFBP-3) or with other prognostic markers such as ER or nodal status. The Plasma IGF-I level did not correlate with better or worse short-term survival of patients at an advanced stage of breast cancer (242). In contrast, IGF-II expression correlated positively with IGFBP-1 and -3, and high IGF-II expression correlated weakly with poor prognostic indicators (high p53, low ER, and low cathepsin D). Immunohistochemical measurement of IGF-II showed a weak inverse relationship with poor prognostic features (tumour grade, S-phase, DNA ploidy) (243). Although in vitro data suggest that IGF expression would be a marker of poor prognosis in breast tumours, the very limited clinical evidence to date do not strongly support this. Breast tumours express IGFBP-2 through -6. In non-invasive breast cancer patients, ER-positive tumours expressed high IGFBP-4 and -5, whereas ERnegative tumours expressed high IGFBP-3 levels (244). IGFBP-4 was correlated with poor prognostic factors and was an independent prognostic marker when patients were sorted by tumour size; patients with large tumours (>2cm) and low IGFBP-4 expression had improved disease-free survival compared to patients with high IGFBP-4 expression (244). Contradictory results have been reported between IGFBP-3 expression and breast cancer prognosis. Serum IGFBP-3 levels have been found to correlate negatively with cancer risk (245). Although no significant association was found between IGFBP-3 and breast cancer recurrence, survival analysis indicated that risk of death was strikingly increased in patients with higher IGFBP-3 levels (246). IGFBP-3 was associated with poor prognostic features (low ER, high S-

7. IGF in breast cancer progression phase, tumour size and aneuploid tumours), although IGFBP-3 levels were not independent factors in multivariate analysis (247-248). Breast cancer patients with elevated levels of the IGFBP proteinase PSA have a better prognosis in multivariate analysis, and PSA levels fall in the transition from benign to malignant breast disease (249-250). The proteolytic activity of PSA on IGFBP-3 may account for a decrease in IGFBP-3 levels and thus an improved prognosis. 5.5.2. Prognostic value of IGF-1R

Despite the impressive evidence supporting a role for the IGF-1R in experimental carcinogenesis, there are only correlative data with regard to the importance of this receptor in human cancers. Early studies indicated that a percentage of primary breast tumours expressed IGF-1R (50-93%), and that expression correlated positively with ER expression (251-253). Nonetheless, IGF1R expression predicts a better prognosis both in relapse-free survival and overall survival of breast cancer patients (254255). Taken as an independent factor, the IGF-1R has been reported as a positive, a negative or an insignificant marker of disease-free survival (253, 256-257), indicating the contradictions in the literature. Indeed, long-term survival has been inversely correlated with IGF-1R levels (257), whereas more recently IGF-1R has been associated with a decline in shortterm survival but without predictive value for long-term survival (258). Although IGF-1R expression appears to be enhanced in breast cancers compared with adjacent normal tissue, no studies have yet been performed that directly implicate the IGF-1R in initiation or propagation of human cancers (80, 259-260). Elevated IGF-1R expression was also associated

129 with “in-breast” cancer recurrence after breast irradiation (258). IRS-1 expression did not correlate with other known prognostic factors, but elevated expression did forecast a shorter disease-free survival period in patients with small tumours (