Page 1

10

Scatter Factor as a Tumor

Angiogenesis Factor

Eliot M. Rosen

Long Island Jewish Medical Center, New Hyde Park, and Albert Einstein

College of Medicine, Bronx, New York

Katrin Lamszus, Saijun Fan, and Itzhak D. Goldberg

Long Island Jewish Medical Center, New Hyde Park, New York

John Laterra

The Johns Hopkins School of Medicine and Kennedy-Krieger Research

Institute, Baltimore, Maryland

Peter J. Polverini

University of Michigan School of Dentistry, Ann Arbor, Michigan

I. INTRODUCTION

Scatter factor (SF) is a heparin-binding glycoprotein that dissociates (“ scatters” ) sheets

of epithelium (Stoker et al., 1987; Rosen et al., 1989). SF is identical to hepatocyte growth

factor (HGF) (Weidner et al., 1991a; Bhargava et al., 1992), a serum mitogen that stimu­

lates the proliferation of primary hepatocyte cell cultures and functions as an hepatotrophic

factor for liver repair (Miyazawa et al., 1989; Nakamura et al., 1989). SF is a heterodimeric

protein consisting of a 60-kDa a-chain containing four kringle domains and a 30-kDa

trypsin-like p-chain (Gherardi et al., 1989; Rosen et al., 1990b; Weidner et al., 1990). SF

has 38% amino acid sequence identity to the proenzyme plasminogen (Nakamura et al.,

1989) but lacks protease activity (Rosen et al., 1990b) due to replacement of two essential

amino acids at the catalytic triad of the p-chain. SF is synthesized as a 728-amino acid

precursor (preproSF); intracellular cleavage of a 31-amino acid signal peptide results in

its secreted single-chain form (proSF), which is biologically inactive (Lokker et al., 1992).

Extracellular cleavage of proSF at 494arg-495val yields active two-chain SF. HGF activator,

a novel serine protease homologous to coagulation factor XII (Hagemann factor), may be

a physiological cleavage enzyme for proSF (Miyazawa et al., 1993). This enzyme is pro­

duced in zymogen form; it may be activated by a proteolytic cascade initiated by tissue

injury (Miyazawa et al., 1994). ProSF is also cleaved by the plasminogen activators [uroki­

nase (uPA) and tissue plasminogen activator (tPA)], but at stoichiometric rather than enzy­

matic quantities. The SF receptor is the protein product of the c-Met proto-oncogene (Bot-

taro et al., 1991), a transmembrane tyrosine kinase (TK) expressed predominantly by

epithelia (Gonzatti-Haces et al., 1988).

145

146

Rosen et al.

SF transduces three major classes of cellular actions in vitro: motility, growth, and

morphogenesis (reviewed in Rosen et al., 1994b). In addition to cell dissociation, SF in­

duces random movement of isolated epithelial cells, chemotactic (gradient-directed) mi­

gration, migration from carrier beads to flat surfaces, and invasion through extracellular

matrix proteins (Rosen et al., 1990a-c, 1991a; Weidner et al., 1990; Bhargava et al., 1992;

Li et al., 1994). SF stimulates the mRNA and protein expression of both urokinase (uPA)

and uPA receptor (uPAR) (Pepper et al., 1992; Grant et al., 1993; Rosen et al., 1994a).

The net effect is to increase the amount of uPA bound to uPAR on the cell surface.

Receptor-bound uPA on the cell surface is catalytically active and is thought to mediate

focal degradation of the extracellular matrix necessary to clear a path for invading cells

(Saksela and Rifkin, 1988). Thus, SF appears to be able to ‘‘switch on” a program of

cell activities for invasion. SF is mitogenic for various normal cell types, including epithe­

lial cells, vascular endothelial cells, and melanocytes (Kan et al., 1991; Rubin et al., 1991;

Halaban et al., 1992). SF is also a potent morphogen. SF induces Madin-Darby canine

kidney (MDCK) epithelial cells incubated in collagen I gels to organize into a network

of branching tubules with the proper apical-basolateral polarity (Montesano et al., 1991;

Santos et al., 1993). Similarly, SF induces mammary epithelial cells to form ductlike

structures (Tsarfaty et al., 1992). Thus, SF can activate specific programs of cell differenti­

ation depending upon the cell type and environment.

II. BLOOD VESSEL WALL CELLS AS TARGETS FOR

SCATTER FACTOR: IN VITRO ANGIOGENIC

ACTIVITIES

A. Vascular Endothelial Cells (ECs)

During the early stages of angiogenesis in vivo, ECs from pre-existing small vessels (usu­

ally venules that lack a smooth muscle covering) focally degrade the subendothelial base­

ment membrane, migrate into the interstitium toward an angiogenic stimulus, and form

capillary sprouts (Folkman, 1985). Sprouting ECs proximal to the migrating tip proliferate;

and the EC sprouts organize into an anastamosing network of capillary tubes. Finally,

these ECs synthesize new basement membrane. Adhesion of smooth muscle cells (SMCs)

and pericytes and synthesis of new basement membrane are associated with the termination

of the angiogenic response (Antonelli-Orlidge et al., 1989). Stimulation of EC motility,

proliferation, and capillary-like tube formation in vitro correlate with the induction of

angiogenesis in vivo (Folkman, 1985).

Both large-vessel and microvessel ECs express the c-Met receptor and are biologi­

cally responsive to SF (Rosen et al., 1990b,c, 1991b; Bussolino et al., 1992; Grant et al.,

1993; Naidu et al., 1994). SF is chemotactic to ECs and stimulates random motility, as

demonstrated in assays using micro well modified Boy den chambers (Rosen et al., 1990b,

1991b). In addition, SF induces the migration of ECs cultured on microcarrier beads from

the beads to flat culture surfaces (Rosen et al., 1990b,c). In chemoinvasion assays, SF

induces penetration of ECs through porous filters coated with Matrigel, a reconstituted

matrix of basement membrane (Rosen et al., 1991b). Maximal stimulation of EC motility

and invasion is typically observed at SF concentrations of 2-20 ng/mL.

In addition to motility, SF stimulates DNA synthesis and proliferation of some EC

types, including HUVEC and human omental micro vessel ECs (Rubin et al., 1991; Mori-

moto et al., 1991). Capillary-like tube formation is an independent property of ECs, not

directly related to motility or proliferation (Grant et al., 1989). When ECs are plated onto

Scatter Factor in Tumor Angiogenesis

147

a surface of Matrigel, they cease proliferation, extend long cytoplasmic processes, and

begin to organize into capillary-like tubes. SF stimulates capillary-like tube formation in

various EC lines by up to five to tenfold, as determined by computerized digital image

analysis of stained cultures (Rosen et al., 1991b; Grant et al., 1993). SF also induces

increased expression of uPA activity by EC cultures (Rosen et al., 1991b; Grant et al.,

1993). Most of the SF-induced urokinase activity is cell-associated rather than secreted.

The majority of cell-associated urokinase is bound to uPAR on the cell surface, where it

is well positioned to mediate focal degradation of extracellular matrix proteins, a prerequi­

site for invasion. Taken together, these findings indicate that SF can induce most or all

of the phenotypic characteristics expected of ECs undergoing angiogenesis.

B. Vascular Smooth Muscle Cells (SMCs) and Pericytes

In vitro, various types of SMCs produce SF at rates comparable to those of high-producer

human lung fibroblasts (Rosen et al., 1989, 1990a). The biological and chemical properties

of SMC-derived SF are very similar to those of fibroblast-derived SF, and it is likely that

these molecules are identical (Rosen et al., 1990a,b).

In addition to producing SF, SMCs also express the c-Met receptor and are respon­

sive to SF. Thus, SMCs within microvessel walls in psoriatic plaques and in tumor micro­

vasculature stained positively for c-Met protein (Grant et al., 1993; Jin et al., 1997; Joseph

et al., 1995), suggesting that they are potential target cells for SF. Recruitment of SMCs

and pericytes (which are generally regarded as microvascular SMCs) is an essential com­

ponent of angiogenesis. These cells stabilize newly formed vessels, thus contributing to

termination of the angiogenic process (Antonelli-Orlidge et al., 1989). Cultured pericytes

from bovine retina express c-Met mRNA, as determined by reverse-transcriptase PCR

analysis (unpublished results), consistent with the finding that pericytes express c-Met

protein in vivo. Moreover, we found that SF stimulates the proliferation of bovine aortic

SMC and bovine retinal pericytes in vitro (Rosen and Goldberg, 1995), consistent with

the putative role of SMCs in angiogenesis and the presence of SMCs in new microvessels

induced by SF (see below).

C. Cytoprotective Activity of Scatter Factor

We have reported that SF protects various lines of epithelial and carcinoma cells against

apoptosis (a form of programmed cell death) induced by cytotoxic agents, including DNA-

damaging agents doxoruxicin (adriamycin), ionizing radiation (6 MeV therapeutic x-rays),

and ultraviolet (UV-C) radiation (Fan et al., 1998). The cytoprotective activity of SF for

these cell types is due in part to upregulation of cell-survival pathways, including enhanced

expression of Bcl-XL, an antiapoptotic protein closely related to Bcl-2. In agreement with

these findings, SF also protected bovine aortic endothelial cells (BAECs) against cytotox­

icity caused by Adriamycin (a DNA topoisomerase Ila inhibitor) and nitrogen mustard

(a DNA alkylating agent) (see Table 1A); and SF blocked apoptosis induced by adriamycin

(data not shown). In contrast to these two DNA-damaging agents, SF gave little or no

protection of ECs against taxol and vincristine, microtubule poisons that arrest cells in

mitosis (Table 1A).

In another set of experiments, we examined the ability of SF to protect BAECs

against Adriamycin in the present of thrombospondin-1 (TSP-1), an endothelial cell inhibi­

tor and anti-angiogenic protein. At a dose of 4 nM, TSP-1 alone did not affect cell viability,

although TSP-1 slightly enhanced the toxicity of Adriamycin. However, TSP1 markedly

148

Rosen et al

Table 1 Effect of Scatter Factor (SF) on Cell Viability of Bovine Aortic

Endothelial Cells Treated with Cytotoxic Agents Using the MTT Assay3

A. SF protects endothelial cells against DNA-damaging agents but not mitotic

spindle poisons

Cell viability (%)

Agent

Dose (pM)

-SF

+ SF

DNA-damaging agents

Adriamycin (ADR)

5

40 ± 2

95 ± 2

Nitrogen mustard

Mitotic spindle poisons

15

12 ± 3

60 ± 2

Taxol

10

43 ± 2

44 ± 3

Vincristine

10

51 ± 2

57 ± 1

B. Thrombospondin-1 (TSP1) reverses SF-mediated protection of endothelial

cells

Additions

SF (ng/mL)

TSP-1 (ng/mL)

ADR (pM)

Cell viability (%)

0

0

0

100

100

0

0

102 ± 3

0

4

0

96 ± 3

0

0

5

40 ± 2

0

4

5

33 ± 1

100

0

5

90 ± 2

100

4

5

60 ± 3

Subconfluent proliferating cells in 96-well dishes were preincubated without (—) or with

(+ ) SF (100 ng/mL X 48 hr) and ± TSP-1 in cell culture medium (DMEM plus 5% fetal

calf serum). Cells were then sham-treated or treated with agent (2 hr for Adriamycin and

nitrogen mustard, 24 hr for Taxol and vincristine); washed twice to remove drug; incubated

for another 72 hr in fresh drug-free culture medium; and assayed for MTT dye conversion

by spectrophotometry (ref). Values represent cell viability relative to sham-treated control

cells and are means ± SEMs of N = 10 replicate wells.

inhibited SF-mediated protection against ADR, resulting in a large decrease in cell viabil­

ity of cells treated with (SF + Adriamycin + TSP-1) as compared with (SF + Adriamycin)

(Table IB). At > 4 nM, TSP1 alone caused decreases in cell viability of BAECs. Corre­

sponding results were obtained in assays of apoptosis using agarose gel electrophoresis

(data not shown).

These findings are interesting for two reasons. First, they suggest another mechanism

by which SF may promote angiogenesis: i.e., by blocking endothelial cell apoptosis, a

mechanism by which microvessels within granulation tissue regress during the later phases

of wound healing (ref). Second, the findings may have therapeutic implications for cancer

treatment. Thus, SF present in tumors (see below) may protect not only tumor cells but

also the tumor neovasculature against the cytotoxic actions of some types of chemotherapy

and radiation therapy. If our findings hold true for ECs in general, then mitotic spindle

poisons may be more effective in targeting the vasculature of tumors with a high endoge­

nous SF content.

Scatter Factor in Tumor Angiogenesis

149

III. IN VIVO ANGIOGENIC ACTIVITY OF SCATTER

FACTOR

We used two different assays, the mouse Matrigel assay and the rat cornea assay, to

demonstrate the ability of SF to induce new blood vessel formation in vivo (Grant et al.,

1993; Naidu et al., 1994). In the former, different doses of SF were mixed with 0.5 mL

of Matrigel in the liquid state at 4°C. The Matrigel was injected subcutaneously into either

XID nude beige mice or C57/BL mice, allowing the formation of a solid gel that retained

SF and permitted prolonged exposure of the surrounding tissues to it. Animals were sacri­

ficed after 10 days, and the ingrowth of blood vessels into the Matrigel plugs was quanti­

tated by computerized image analysis of tissue sections stained with Masson’s trichrome.

Angiogenesis assessed at day 10 increased in a dose-dependent manner from 2-200 ng/

mL of SF, up to four to five times control values. Responses were similar in nude mice

Table 2 Inhibition of Scatter Factor (SF)-Induced

Angiogenesis in the Rat Corneal Neovascularization Assay by

SF a-Chain Peptides NK1 and NK2 and by c-Met IgG

Agent(s) added

Positive assays (%)

A. Inhibition of SF-mediated angiogenesis by NK1 and NK2

peptides

Control assays

Hydron + PBS

0/2 (0)

bFGF (50 ng)

2/2 (100)

rhSF (50 ng)

3/3 (100)

NK1 (25 ng)

0/2 (0)

NK2 (25 ng)

0/2 (0)

rhSF ± peptide

bFGF (50 ng) + NK1 (25 ng)

2/2 (100)

bFGF (50 ng) + NK2 (25 ng)

2/2 (100)

rhSF (50 ng) + NK1 (25 ng)

0/2 (0)

rhSF (50 ng) + NK2 (25 ng)

0/2 (0)

B. Inhibition of SF-mediated angiogenesis by c-Met IgG

Control assays

Hydron + PBS

0/2 (0)

bFGF (100 ng)

2/2 (100)

rhSF (100 ng)

3/3 (100)

c-Met IgG (1 fig)

0/4 (0)

rhSF ± c-Met IgG

rhSF (100 ng)

3/3 (100)

rhSF (100 ng) + c-Met IgG (250 ng)

3/6 (50)

rhSF (100 ng) + c-Met IgG (500 ng)

1/5 (20)

rhSF (100 ng) + c-Met IgG (1 |ig)

0/5 (0)

Abbreviations: bFGF, basic fibroblast growth factor; PBS, phosphate-buf­

fered saline; rhSF, recombinant human scatter factor. c-Met IgG is a chime­

ric antibody consisting of the ligand-binding domain of the c-Met receptor

fused to a human IgG heavy chain backbone. Assays were performed and

responses scored as described before (Tolsma et al., 1993).

150

Rosen et al.

and C57/BL mice. Inflammatory responses were not observed in nude mice and were

found only at supramaximal SF doses (2000 ng/mL) in C57/BL mice.

In the second assay, SF was dissolved in Hydron polymer and dried Hydron pellets

were placed in surgically created pockets about 1.5 mm from the limbus of the avascular

rat cornea. Animals were perfused with colloidal carbon and sacrified after 7 days; the

growth of new vessels from the limbus toward the pellet was them assessed. SF induced

dose-dependent corneal neovascularization similar to that observed in the murine assay

(Grant et al., 1993). The maximal SF-induced responses were similar in intensity to those

induced by basic fibroblast growth factor (bFGF) or vascular endothelial growth factor

(VEGF). Antibodies against SF blocked SF-induced angiogenesis but did not affect bFGF-

induced angiogenesis. Inflammatory responses (monocyte/macrophage infiltration) were

observed only at supramaximal doses of SF. Therefore, SF appears to be at least as potent

an inducer of angiogenesis as bFGF or VEGF.

Our findings indicate that recruitment of inflammatory cells does not play a major

role in SF-induced angiogenesis. However, in the mouse Matrigel assay, histologic sec­

tions prepared at early times (days 2-3), revealed many SMC/pericyte-like cells present

in the Matrigel. Moreover, at higher doses of SF, histological sections prepared on day

10 revealed SMCs in some of the newly formed vessels in the Matrigel (Grant et al.,

Table 3 Inhibition of Chemotaxis of Vascular Endothelial Cells Toward

Scatter Factor (SF) by Recombinant Human SF a-Chain Peptides NK1 and

NK2a

Agent(s) added (ng/ml)

Cells per ten 40 grids (N)

% Inhibition

None

18

3(4)

—

rhSF (20)

136

12(4)

0

NK1 (20)

21 -i-

4(3)

—

NK1 (50)

26 -i-

11 (3)

—

NK1 (200)

12

4(3)

—

rhSF (20) + NK1 (20)

130

10 (3)

8

rhSF (20) + NK1 (50)

101

4(3)

36

rhSF (20) + NK1 (200)

56

6(3)

63

NK2 (20)

15 + 3(3)

—

NK2 (50)

9 H -

3 (3)

—

NK2 (200)

17 + 10 (3)

—

rhSF (20) + NK2 (20)

94 + 1 (3)

33

rhSF (20) + NK2 (50)

100 4 -

10(3)

23

rhSF (20) + NK2 (200)

54

5(3)

68

aChemotactic migration of human dermal microvascular endothelial cells toward different

combinations of rhSF and NK1 or NK2 was assayed using 48-well micro well-modified Boy-

den chambers, as described before (see Lamszus et al., 1997). Migration values (means ±

standard errors) represent the number of cells that had migrated across porous collagen-

coated filters over a4-hr assay incubation period. Values in parentheses represent the number

of replicate wells tested.

Scatter Factor in Tumor Angiogenesis

151

1993). Thus, SF-induced angiogenesis appears to be mediated by direct effects on ECs

and, in addition, by direct or indirect effects on SMCs.

Several peptide fragments of the SF a-chain, designated NK1 and NK2, are gener­

ated by alternative mRNA processing. These peptides consist of the N-terminal prekringle

region plus the first kringle domain (NK1) or the first two kringle domains (NK2) of the

SF a-chain (Chan et al., 1991; Lokker et al., 1992; Cioce et al., 1996). Recombinant

human NK1 and NK2 peptides have been found to bind with high affinity to the c-Met

receptor and to exert partial agonistic and partial antagonistic activity compared with full-

length heterodimeric SF (Cioce et al., 1996; Lokker et al., 1992). Utilizing the rat corneal

neovascularization assay, we found that neither NK1 nor NK2 were capable of inducing

angiogenesis, but each of these peptides inhibited angiogenesis induced by full-length SF

(Table 2A). A chimeric antibody consisting of a human IgG heavy chain fused to bivalent

c-Met receptor extracellular binding domains [designated “c-Met IgG” (Mark et al.,

1992)] also blocked SF-mediated angiogenesis in the rat cornea (Table 2B). One micro­

gram of c-Met IgG ablated the angiogenic action of 100 ng/mL of rhSF, a dose sufficient

to induce a maximal angiogenic response. NK1 and NK2 also inhibited chemotactic migra­

tion of human microvascular endothelial cells toward SF (Table 3), indicating that these

may be useful reagents for specifically blocking SF-induced angiogenesis.

IV. ROLE OF SCATTER FACTOR AS A TUMOR

ANGIOGENESIS FACTOR

A. Clinical Studies

Recent studies suggest that tumor angiogenesis, indicated by large numbers of microves­

sels in tumor stroma, is an independent indicator of poor prognosis in patients with inva­

sive breast cancer and other tumor types (Weidner et al., 1991b, 1992; Bosari et al., 1992).

Consistent with the idea that SF may function as a tumor angiogenesis factor, we have

found that SF and c-Met expression are both upregulated during the transition from low-

grade noninvasive to high-grade invasive tumors in breast cancer, urothelial bladder can­

cer, and gliomas (Jin et al., 1997a, b; Joseph et al., 1995; Lamszus et al., 1998); Rosen

et al., 1996; Rosen et al., 1997; Yao et al., 1996; Ole Schmidt et al., in press). In addition,

other investigators have reported that a high SF content of invasive breast cancers is a

powerful and independent predictor of tumor relapse and death (Yamashita et al., 1994).

To investigate the potential role of SF as a tumor angiogen, we measured the SF

content and the content of von Willebrand’s factor (VWF), an endothelial-specific marker

protein, in extracts of invasive human breast cancers. Studies of over 200 tumor specimens

revealed a strong and highly significant correlation between the SF content and the VWF

content of the tumor extract (Jin et al., 1997a, b) (Table 4). These results suggest that the

overall level of tumor vascularity, as indicated by the VWF content, is related to the SF

content in biopsy specimens of breast cancers. It was initially though that tumor-derived

SF was produced by tumor-associated stromal cells (e.g., fibroblasts, SMCs, macro­

phages), since the vast majority of carcinoma cell lines, including a number of human

breast cancer cell lines, do not produce any detectable SF in vitro. However, more recent

studies indicate that breast carcinoma cells express both SF protein and SF mRNA in

vivo, suggesting that loss of the ability to produce SF may be an artifact of cell culture

(Jin et al., 1997; Tuck et al., 1996; Wang et al., 1994).

152

Rosen et al

Table 4 Relationship Between Context of von Willebrand

factor (VWF) and Scatter Factor (SF) in Extracts of Invasive

Human Breast Cancers3

SF content range (ng/mg)

VWF content (ng/mg protein)

I. 0.00-0.50

26 ± 7 (50)

II. 0.51-1.00

49 ± 8 (62)

III. 1.01-2.00

93 ± 27 (57)

IV. 2.01-up

203 ± 51 (55)

aHuman breast cancer tissue specimens frozen in OCT compound were

thawed; and tissue extracts were prepared as described before (ref). Values

of SF and VWF content were determined by two-antibody ELISAs and

were normalized to the protein content of the extract (ref). Values repre­

sent means ± standard errors (number of individual tumors assayed). Sta­

tistical comparisons of VWF vs. SF content range were made using

two-tailed Student’s t-tests: 0-1.00 vs. 1.01-up, p < 0.001; 0-0.50 vs.

0.51-1.00, p = 0.040; 0-0.50 vs. 2.01-up, p < 0.001; 1.01-2.00 vs.

2.01-up, p = 0.061.

B. Kaposi’s Sarcoma (KS)

KS is a multicellular neoplasm characterized by a major component of EC proliferation

and angiogenesis. The epidemic form of KS occurs frequently in homosexual males with

acquired immunodeficiency syndrome (AIDS). AIDS-KS is usually regarded as a multifo­

cal cytokine-dependent tumor rather than a metastasizing cancer. Recently, SF was found

to mediate several functions that may be related to the pathogenesis of AIDS-related KS:

(a) conversion of normal human ECs into a KS tumor cell-like phenotype; (b) autocrine

stimulation of KS tumor cell proliferation; and (c) angiogenic activity of conditioned me­

dium from retrovirally infected human T lymphocyte cultures (Naidu et al., 1994). Further­

more, immunoreactive SF and c-Met receptor were detected in a variety of cell types

within KS tumor sections. These findings suggest that SF may function as a mediator of

EC proliferation and angiogenesis in AIDS-related KS lesions.

C. Experimental Animal Models

We have examined the ability of SF to stimulate angiogenesis in an animal model of

human breast cancer. We transfected human SF cDNA into MDAMB231 human breast

cancer cells so that they overexpressed SF. Tumors generated from SF-transfected (SF+)

cell clones grew much more rapidly in the mammary fat pads of nude mice than did

tumors generated control-transfected (SF—) clones (Lamszus et al., 1997). Extracts from

the SF+ tumors contained 50-fold a higher SF content than did extracts prepared from

SF— control tumors. In contrast to the in vivo results, in vitro proliferation rates and

cloning efficiencies were no greater in SF4- vs. SF—clones. We investigated the discrep­

ancy between the in vivo and in vitro growth results and found that (a) microvessel densi­

ties were significantly higher in SF4- vs. S F- tumors; (b) chemotactic activity for capillary

endothelial cells was greater in conditioned media and primary tumor extracts from SF+

vs SF—clones; (c) angiogenic activity in the rat cornea assay was much higher in SF+

vs. S F- tumor extracts; (d) chemotactic and angiogenic activity in SF+ tumor extracts

was not explained by secondary changes in levels of other regulators; and (e) those activi­

ties were neutralized by an anti-SF monoclonal (Lamszus et al., 1997). Thus, the increased

Scatter Factor in Tumor Angiogenesis

153

tumor growth rates of SF+ MDAMB231 cells appears to be due in part to an angiogenic

response induced by SF.

In addition to the breast cancer model, we have developed several rodent models

for the orthotopic growth of glial tumors. We found that 9L rat glioma cells transfected

with human SF cDNA grew much more rapidly than control-transfected cells both as brain

tumors and as subcutaneous tumors in syngeneic host rats (Laterra et al., 1997). Analysis

of brain tumor sections using a digital image analysis system revealed significantly higher

microvessel densities in SF+ tumors than in SF— control tumors. SF+ tumors also

showed significantly higher expression of urokinase, an enzyme required for proteolytic

degradation of extracellular matrix and invasion of migrating ECs during the angiogenic

process. In parallel studies, SF also induced increased growth and micro vessel formation

in brain tumors from SF-transfected human glioblastoma cells (U373MG) grown as brain

tumors in immunocompromised (SCID) mice, as compared with control U373MG cells

(Laterra et al., 1997). These findings suggest that overexpression of SF stimulates in vivo

tumor growth, in part by conferring an angiogenic tumor phenotype.

V. CONCLUSIONS

The cytokine scatter factor is potent inducer of endothelial cell activation and of angiogen­

esis. Studies of clinical tumor specimens and experimental animal models of human tu­

mors suggest that the expression of SF confers a more invasive and angiogenic phenotype.

ACKNOWLEDGMENTS

This work was supported in part by grants from the U.S. Public Health Service (CA-

64869 and CA-64416). We are indebted to Dr. Ralph Schwall, Genentech, Inc. (South

San Francisco, CA) for providing sheep antiserum against SF, recombinant human two-

chain human SF, and c-Met IgG. We are grateful to Drs. Jeffrey Rubin, Donald Bottaro,

Stephen Stahl, and Paul Wingfield, Laboratory of Cellular and Molecular Biology, Na­

tional Cancer Institute (Bethesda, MD) for providing recombinant human truncated SF

peptides (NK1 and NK2).

REFERENCES

Antonelli-Orlidge A, Smith SR, D’Amore PA. Influence of pericytes on capillary endothelial cell

growth. Am Rev Resp Dis 140:1129-1131, 1989.

Bhargava M, Joseph A, Knesel J, Halaban R, Li Y, Pang S, Goldberg I, Setter E, Donovan MA,

Zamegar R, Michalopoulos GA, Nakamura T, Faletto D, Rosen EM. Scatter factor and hépa­

tocyte growth factor: activities, properties, and mechanism. Cell Growth Differ 3:11-20,

1992.

Bosari S, Lee AK, DeLellis RA, Wiley BD, Heatley GJ, Silverman ML. Microvessel quantitation

and prognosis in invasive breast carcinoma. Hum Pathol 23:755-761, 1992.

Bottaro DP, Rubin JS, Faletto DL, Chan AM-L, Kmiecik TE, Vande Woude GF, Aaronson SA.

Identification of the hepatocyte growth factor receptor as the c-met proto-oncogene product.

Science 251:802-804, 1991.

154

Rosen et al.

Bussolino F, DiRenzo MF, Ziche M, Bocchieto E, Olivero M, Naldini L, Gaudino G, Tamagnone

L, Coffer A, Comoglio PM. Hepatocyte growth factor is a potent angiogenic factor which

stimulates endothelial cell motility and growth. J Cell Biol 119:629-641, 1992.

Chan AM-L Rubin JS, Bottaro DP, Hirschfield DW, Chedid M, Aaronson SA: Identification of a

competitive antagonist encoded by an alternative transcript. Science 254:1382-1385,

1991.

Cioce V, Csaky KG, Chan AM-L, Bottaro DP, Taylor WG, Jemsen R, Aaronson SA, Rubin JS.

HGF/NK1 is a naturally occurring HGF/SF variant with partial agonist/antagonist activity.

J Biol Chem 271:13110-13115, 1996.

Fan S, Wang J-A, Yuan R-Q, Rockwell S, Andres J, Zlatapolskiy A, Goldberg ID, Rosen EM.

Scatter factor protects epithelial and carcinoma cells against apoptosis induced by DNA-

damaging agents. Oncogene 17:131-141, 1998.

Folkman J. Tumor angiogenesis. Adv Cancer Res 43:175-203, 1985.

Gherardi E, Gray J, Stoker M, Perryman M, Furlong R. Purification of scatter factor, a fibroblast-

derived basic protein which modulates epithelial interactions and movement. Proc Natl Acad

Sci USA 86:5844-5848, 1989.

Gohda E, Matsunaga T, Kataoka H, Yamamoto I. TGF-p is a potent inhibitor of hepatocyte growth

factor secretion by human fibroblasts. Cell Biol Int Reports 16:917-926, 1992.

Gonzatti-Haces M, Seth A, Park M, Copeland T, Oroszlan S, Vande Woude GF. Characterization

of the TPR-MET oncogene p65 and the MET protooncogene pi40 protein tyrosine kinases.

Proc Natl Acad Sci USA 85:21-25, 1988.

Grant DS, Tashiro KI, Segui-Real B, Yamada Y, Martin GR, Kleinman HK. Two different laminin

domains mediate the differentiation of human endothelial cells into capillary-like structures

in vitro. Cell 58:933-943, 1989.

Grant DS, Kleinman HK, Goldberg ID, Bhargava M, Nickoloff BJ, Polverini P, Rosen EM. Scatter

factor induces blood vessel formation in vivo. Proc Natl Acad Sci USA 90:1937-1941, 1993.

Halaban R, Rubin J, Funasaka Y, Cobb M, Boulton T, Faletto D, Rosen E, Chan A, Yoko K,

White W, Cook C, Moellmann G. Met and hepatocyte growth factor/scatter factor signal

transduction in normal melanocytes and melanoma cells. Oncogene 7:2195-2206, 1992.

Jin L, Fuchs A, Schnitt SS, Yao Y, Joseph A, Lamszus K, Park M, Goldberg ID, Rosen EM. Expres­

sion of SF and c-met receptor in benign and malignant breast tissue. Cancer 79:749-760,

1997a.

Jin L, Yuan R-Q, Fuchs A, Yao Y, Joseph A, Schwall R, Guida A, Schnitt SJ, Hastings HM, Andres

J, Turkel G, Polverini PJ, Goldberg ID, Rosen EM. Expression of IL-lp in human breast

cancer. Cancer 80:421-434, 1997b.

Joseph A, Weiss GH, Jin L, Fuchs A, Chowdhury S, O’Shaughnessy P, Goldberg ID, Rosen EM.

Expression of scatter factor in human bladder carcinoma. J Natl Cancer Inst 87:372-377,

1995.

Kan M, Zhang GH, Zamegar R, Michalapoulos G, Myoken Y, McKeehan WL, Stevens JL. Hepato­

cyte growth factor-hepatopoietin A stimulates the growth of rat proximal tubule epithelial

cells (rpte), rat non-parenchymal liver cells, human melanoma cells, mouse keratinocytes,

and stimulates anchorage-independent growth of SV40-transformed rpte. Biochem Biophys

Res Commun 174:331-337, 1991.

Lamszus K, Jin L, Fuchs A, Shi YE, Chowdhury S, Yao Y, Polverini PJ, Goldberg ID, Rosen EM.

Scatter factor stimulates tumor growth and tumor angiogenesis in human breast cancers in

the mammary fat pads of nude mice. Lab Invest 76:339-353, 1997.

Lamszus K, Schmidt NO, Jin L, Laterra J, Zagzag D, Way D, Witte M, Weinand M, Goldberg ID,

Westfal M, Rosen EM. Scatter factor promotes motility of human glioma and neuromicrovas-

cular endothelial cells. Int J Cancer 75:19-28, 1998.

Laterra J, Rosen E, Nam M, Rangathan S, Fielding K, Johnston P. Scatter factor/hepatocyte growth

factor expression enhances human glioblastoma tumorigenicity and growth. Biochem Biophys

Res Commun 235:743-747, 1997.

Scatter Factor in Tumor Angiogenesis

155

Laterra J, Nam M, Rosen EM, Rao JS, Lamszus K, Johnston P. Scatter factor/hepatocyte growth

factor gene transfer enhances glioma growth and angiogenesis in vivo. Lab Invest 76:565-

577, 1997.

Li Y, Bhargava MM, Joseph A, Jin L, Rosen EM, Goldberg ID. The effect of scatter factor and

hepatocyte growth factor on motility and morphology of non-tumorigenic and tumor cells.

In Vitro Cell Dev Biol 30A: 105-110, 1994.

Lokker NA, Mark MR, Luis EA, Bennett GL, Robbins KA, Baker JB, Godowski PJ. Structure-

function analysis of hepatocyte growth factor: identification of variants that lack mitogenic

activity yet retain high affinity receptor binding. EMBO J 11:2403-2410, 1992.

Mark MR, Lokker NA, Zioncheck TF, Luis EA, Godowski PJ. Expression and characterization of

hepatocyte growth factor receptor-IgG fusion proteins: effects of mutations in the potential

proteolytic cleavage sites on processing and ligand binding. J Biol Chem 267:26166-26171,

1992.

Miyazawa K, Tsubouchi H, Naka D, Takahashi K, Okigaki M, Arakaki N, Nakayama S, Hirono

S, Sakiyama O, Gohda E, Daikuhara Y, Kitamura N. Molecular cloning and sequence analysis

of cDNA for human hepatocyte growth factor. Biochem Biophys Res Commun 163:967-

973, 1989.

Miyazawa K, Shimomura T, Kitamura A, Kondo J, Morimoto Y, Kitamura N. Molecular cloning

and sequence analysis of the cDNA for a human serine protease responsible for activation

of hepatocyte growth factor: structural similarity of protease precursor to blood coagulation

factor XII. J Biol Chem 268:10024-10028, 1993.

Miyazawa K, Shimomura T, Naka D, Kitamura N. Proteolytic activation of hepatocyte growth factor

in response to tissue injury. J Biol Chem 269:8966-8970, 1994.

Montesano R, Matsumoto K, Nakamura T, Orci L. Identification of a fibroblast-derived epithelial

morphogen as hepatocyte growth factor. Cell 67:901-908, 1991.

Morimoto, A., Okamura, K., Hamanaka, R., Sato, Y., Shima, N., Higashio, K., Kuwano, M. Hepato­

cyte growth factor modulates migration and proliferation of microvascular endothelial cells

in culture. Biochem Biophys Res Commun 179:1042-1049, 1991.

Naidu YM, Rosen EM, Zitnik R, Goldberg I, Park M, Naujokas M, Polverini PJ, Nickoloff BJ.

Role of scatter factor in the pathogenesis of AIDS-related Kaposi’s sarcoma. Proc Natl Acad

Sci USA 91:5281-5285, 1994.

Nakamura T, Nishizawa T, Hagiya M, Seki T, Shimonishi M, Sugimura A, Shimizu S. Molecular

cloning and expression of human hepatocyte growth factor. Nature 342:440-443, 1989.

Ole Schmidt N, Westfal M, Hagel C, Stavrou D, Ergun S, Rosen EM, Lamszus K. Expression of

vascular endothelial growth factor, hepatocyte growth factor/scatter factor, and basic fibro­

blast growth factor in human gliomas and their relation to angiogenesis. Int J Cancer. In press.

Pepper MS, Matsumoto K, Nakamura T, Orci L, Montesano R. Hepatocyte growth factor increases

urokinase-type plasminogen activator (u-PA) and u-PA receptor expression in Madin-Darby

canine kidney epithelial cells. J Biol Chem 267:20493-20496, 1992.

Rosen EM, Goldberg ID, Kacinski BM, Buckholz T, Vinter DW. Smooth muscle releases an epithe­

lial cell scatter factor which binds to heparin. In Vitro Cell Dev Biol 25:163-173, 1989.

Rosen EM, Laterra J, Joseph A, Jin L, Fuchs A, Way D, Witte M, Weinand M, Goldberg ID. Scatter

factor expression and regulation in human glial tumors. Int J Cancer 67:248-255, 1996.

Rosen EM, Meromsky L, Setter E, Vinter DW, Goldberg ID. Smooth muscle-derived factor stimu­

lates mobility of human tumor cells. Invasion Metastasis 10:49-64, 1990a.

Rosen, EM, Meromsky L, Setter E, Vinter DW, Goldberg ID. Purification and migration-stimulating

activities of scatter factor. Proc Soc Exp Biol Med 195:34-43, 1990b.

Rosen EM, Meromsky L, Setter E, Vinter DW, Goldberg ID. Quantitation of cytokine-stimulated

migration of endothelium and epithelium by a new assay using microcarrier beads. Exp Cell

Res 186:22-31, 1990c.

Rosen EM, Goldberg ID: Scatter factor and angiogenesis. Adv Cancer Res 67:257-279. 1995.

Rosen EM, Goldberg ID, Liu D, Setter E, Donovan MA, Bhargava M, Reiss M, Kacinski BM.

156

Rosen et al.

Tumor necrosis factor stimulates epithelial tumor cell motility. Cancer Res 57:5315-5321,

1991a.

Rosen EM, Grant D, Kleinman H, Jaken S, Donovan MA, Setter E, Luckett PM, Carley W. Scatter

factor stimulates migration of vascular endothelium and capillary-like tube formation. In:

Goldberg ID, Rosen EM, eds. Cell Motility Factors. Basel: Birkhauser-Verlag, 1991b, pp. 76-

88.

Rosen EM, Jaken S, Carley W, Setter E, Bhargava M, Goldberg ID. Regulation of motility in bovine

brain endothelial cells. J Cell Physiol 146:325-335, 1991c.

Rosen EM, Joseph A, Jin L, Yao Y, Chau M-H, Fuchs A, Gomella L, Hasings H, Goldberg ID,

Weiss GH. Urinary and tissue levels of scatter factor in transitional cell carcinoma of bladder.

J Urol 157:72-78, 1997.

Rosen EM, Knesel J, Goldberg ID, Bhargava M, Joseph A, Zitnik R, Wines J, Kelley M, Rockwell

S. Scatter factor modulates the metastatic phenotype of the EMT6 mouse mammary tumor.

Int J Cancer 57:706-714, 1994a.

Rosen EM, Nigam SK, Goldberg ID. Mini-Review: Scatter factor and the c-met receptor: a paradigm

for mesenchymal: epithelial interaction. J Cell Biol 127:1783-1787, 1994b.

Rubin JS, Chan AM-L, Bottaro DP, Burgess WH, Taylor WG, Cech AC, Hirschfield DW, Wong

J, Miki T, Finch PW, Aaronson SA. A broad spectrum human lung fibroblast-derived mitogen

is a variant of hepatocyte growth factor. Proc Natl Acad Sci USA 88:415-419, 1991.

Saksela O, Rifkin DB. Cell-associated plasminogen activation: regulation and physiologic functions.

Annu Rev Cell Biol 4:93-126, 1988.

Santos OFP, Moira LA, Rosen EM, Nigam SK. Modulation of HGF-induced tubulogenesis and

branching by multiple phosphorylation mechanisms. Dev Biol 159:538-545, 1993.

Stoker M, Gherardi E, Perryman M, Gray J. Scatter factor is a fibroblast-derived modulator of

epithelial cell mobility. Nature 327:238-242, 1987.

Tolsma SS, Volpert OV, Good DJ, Frazier WA, Polverini PJ, Bouck N. Peptides derived from two

separate domains of the matrix protein thrombospondin-1have antiangiogenic activity. J Cell

Biol 122:497-511, 1993.

Tsarfaty I, Resau JH, Rulong S, Keydar I, Faletto D, Vande Woude GF. The met proto-oncogene

receptor and lumen formation. Science 257:1258-1261, 1992.

Tuck AB, Park M, Stems EE, Boag A, Elliott BE. Coexpression of hepatocyte growth factor and

receptor (Met) in human breast carcinoma. Am J Pathol 148:225-232, 1996.

Wang Y, Selden AC, Morgan N, Stamp GW, Hodgson HJ. Hepatocyte growth factor/scatter factor

expression in human mammary epithelium. Am J Pathol 144:675-682, 1994.

Weidner KM, Behrens J, Vandekerckhove J, Birchmeier W. Scatter factor: molecular characteristics

and effect on invasiveness of epithelial cells. J Cell Biol 111:2097-2108, 1990.

Weidner KM, Arakaki N, Vandekereckhove J, Weingart S, Hartmann G, Rieder H, Fonatsch C,

Tsubouchi H, Hishida T, Daikuhara Y, Birchmeier W. Evidence for the identity of human

scatter factor and human hepatocyte growth factor. Proc Natl Acad Sci USA 88:7001-7005,

1991a.

Weidner N, Semple JP, Welch WR, Folkman J. Tumor angiogenesis and metastasis—correlation

in invasive breast carcinoma. New Engl J Med 324:1-8, 1991b.

Weidner N, Folkman J, Pozza F, Bevilaqua P, Allred EN, Moore DH, Meli S, Gasparini G. Tumor

angiogenesis: a new significant and independent prognostic indicator in early stage breast

carcinoma. J Natl Cancer Inst 84:1875-1887, 1992.

Yamashita J, Ogawa M, Yamashita S, Nomura K, Kuramoto M, Saishoji T, Sadahito S. Immunoreac-

tive hepatocyte growth factor is a strong and independent predictor of recurrence and survival

in human breast cancer. Cancer Res 54:1630-1633, 1994.

Yao Y, Jin L, Fuchs A, Joseph A, Hastings H, Goldberg ID, Rosen EM. Scatter factor protein levels

in human breast cancers: Clinicopathologic and biologic correlations. Am J Pathol 149:1707-

1717, 1996.