Oncogene (1997) 14, 2417 ± 2423
 1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00
Activated ras and ret oncogenes induce over-expression of c-met
(hepatocyte growth factor receptor) in human thyroid epithelial cells
Mircea Ivan1, Jane A Bond1, Maria Prat2, Paolo Maria Comoglio2 and David Wynford-Thomas1
1
Cancer Research Campaign Thyroid Tumor Biology Research Group, Department of Pathology, University of Wales College of
Medicine, Heath Park, Cardi€ CF4 4XN, UK; 2Department of Biomedical Science and Oncology, University of Torino Medical
School, Torino, Italy
Hepatocyte Growth Factor (HGF) receptor, encoded by
the protooncogene c-met, is overexpressed in many
human tumours, including those of thyroid epithelium.
The absence in most cases of any primary structural
abnormality of the met gene suggests that overexpression
is secondary to mutation of other gene(s). To test this
hypothesis we investigated the e€ect on met expression
of two activated oncogenes known to play a major role in
thyroid oncogenesis, ras and ret. To minimize the
possibility of unknown co-operating events, we introduced these genes directly into normal human thyrocytes
in primary culture using amphotropic retroviral vectors
and assessed met expression as early as possible in the
resulting epithelial colonies. Double immuno¯uorescence
revealed expression of met protein, strictly localized to
cells expressing the mutant ras and ret vectors,
expression in background normal cells being barely
detectable. In contrast, colonies induced to proliferate
at a comparable rate by a vector expressing SV40 T
showed no increase in met expression. To permit
quantitation by Western blotting we also extended these
®ndings to a thyroid cell line (R18) containing a zincinducible mutant ras gene. Induction of the oncogene led
to a fourfold increase in met protein expression. We
conclude that overexpression of met is induced by
activation of the ras or ret signalling pathway and not
simply by deregulation of the cell cycle per se. The data
suggest that the proliferative advantage conferred by
these oncogenes may be in part due to the resulting
sensitization of tumour epithelium to paracrine HGF
secreted by stromal cells.
Keywords: met; ras; ret; thyroid; epithelium; paracrine
Introduction
The c-met proto-oncogene encodes a transmembrane
tyrosine kinase of 190 kDa, identi®ed as the receptor
for hepatocyte growth factor (HGF) (Bottaro et al.,
1991; Naldini et al., 1991). Although c-met activation
was originally associated with rearrangements or gene
ampli®cation in some cell lines (Park et al., 1986;
Ponzetto et al., 1991), in primary human tumours this
is very rare, abnormalities in expression of c-met being
usually found in the apparent absence of any underlying structural gene alteration (Houldsworth et al.,
1990; Di Renzo et al., 1991). Met protein has now been
Correspondence: D Wynford-Thomas
Received 16 August 1996; revised 4 February 1997; accepted 5
February 1997
reported to be overexpressed in comparison to the
corresponding normal epithelium in a wide range of
epithelial tumours, including pancreatic (Di Renzo et
al., 1995a), colorectal (Di Renzo et al., 1991; Liu et al.,
1992), ovarian (Di Renzo et al., 1994), gastric and
renal carcinomas (Di Renzo et al., 1991). Thyroid is a
particularly good example with met protein being
overexpressed up to 100-fold in the vast majority of
papillary carcinomas, and also in a signi®cant
proportion of follicular tumours (Di Renzo et al.,
1991, 1992; Prat et al., 1991).
The ligand for the receptor encoded by c-met, HGF,
is produced by mesenchymal cells from various organs
(Stoker et al., 1987; Rubin et al., 1991; Sonnenberg et
al., 1993) and has been shown to be a potent mitogen
for a variety of epithelial cells in vitro, including
hepatocytes (Strain et al., 1991), renal tubular cells
(Igawa et al., 1991), melanocytes (Matsumoto et al.,
1991a), keratinocytes (Matsumoto et al., 1991b),
normal and oncogene-transformed thyroid cells
(Eccles et al., 1996). Overexpression of the c-met
protein by human epithelium may therefore confer
increased sensitivity to HGF produced by surrounding
stromal cells, thereby representing a signi®cant
paracrine contribution to tumour growth.
The absence in most cases of structural abnormality of
the met gene in human tumours suggests that overexpression is secondary to mutation of other genes, an
obvious possibility being induction by activation of
signal pathways containing an activated oncoprotein.
The proportion of cases exhibiting met overexpression
among some tumour types greatly exceeds that for any
individual oncogene activation event so far identi®ed (Di
Renzo et al., 1991, 1995b; Prat et al., 1991), suggesting
that multiple oncogenic signals converge on met.
To test this hypothesis, we have investigated the e€ect
on met expression of activating the two principal
oncogenes known to play a major role in early thyroid
oncogenesis: ras, which is activated by point mutation in
up to 50% of follicular tumours (Lemoine et al., 1989a;
Namba et al., 1990; Suarez et al., 1990) and ret, which is
activated by rearrangement in at least 25% of papillary
tumours (Grieco et al., 1990; Santoro et al., 1992).
To minimize the possibility of unknown cooperating genetic abnormalities, we ®rst tested the
response to mutant ras or ret genes in the context of
the normal human thyroid follicular cell. We have
previously demonstrated the feasibility of obtaining
stable expression of these genes in such cells in primary
culture by the use of amphotropic retroviral vectors
(Lemoine et al., 1990; Bond et al., 1994). In both cases
expression of the mutant gene induces rapid growth of
epithelial colonies which are easily distinguishable from
Induction of c-met by activated oncogenes
M Ivan et al
2418
the quiescent normal background monolayer, allowing
us to investigate the early e€ect of oncogene activation
on c-met protein expression by immunocytochemistry.
To permit more accurate quantitation by Western
analysis we also investigated the expression of met
protein in a thyroid cell line (R18) containing a zincinducible mutant ras expression vector which has been
shown to have no e€ect on cell proliferation (Dawson
et al., 1995).
Results
Mutant ret and ras oncogenes induce overexpression of
met in human thyroid cells in primary culture
Infection of primary cultures of normal human thyroid
follicular cells with retroviral vectors encoding mutant
H-ras or ret oncogenes, led as expected (Lemoine et al.,
1990; Bond et al., 1994) to the appearance of rapidly
growing epithelial colonies which were readily distinguishable from the surrounding uninfected quiescent
monolayer by 1 ± 2 weeks. (Normal thyroid cells show
very limited proliferative capacity in monolayer
(Dawson et al., 1991)). Many of the ret colonies
displayed the distinctive `fenestrated' morphology
reported earlier (Bond et al., 1994).
Expression of the appropriate oncogene in these
colonies was con®rmed by immunocytochemistry using
antibodies to ret and ras antigen (Figure 1a and d). Ret
is not expressed in normal thyroid cells; in the case of
ras, endogenous expression is much lower than that of
the vector-encoded mutant (Lemoine et al., 1990;
Dawson et al., 1995).
Expression of met in these two types of colonies was
also investigated by immunocytochemistry using a
monoclonal antibody against the extracellular domain
(DO24) (Prat et al., 1991). Immunoblot analysis was
not performed because we wished to analyse cells as
soon as possible after introduction of the mutant
oncogene, to minimise selection for additional abnormalities. Furthermore, only a minority of colonies would
be expected to generate sucient cell numbers for
Western analysis before undergoing senescence (Bond
et al., 1994). An additional advantage o€ered by
immunocytochemistry here was that in cultures not
selected in G418 we were able to directly compare the
staining in oncogene-generated colonies with that of
the surrounding normal cells (Figure 1b, e and g).
Strong immunostaining with anti-met antibody was
observed in a large proportion of ras- and ret-induced
colonies (25/36 and 22/30 respectively) analysed 12 days
after introduction of the activated oncogene. Within
these clones, the majority of cells were positive. By
comparison, in surrounding normal cells staining was
either very faint or undetectable (Figure 1b, c and e).
The above analyses were necessarily performed on
di€erent examples of each type of colony. In order to
directly con®rm the association between overexpression
of c-met/HGF receptor and expression of the activated
oncogene, double immuno¯uorescence was used to
permit simultaneous analysis of ret and met proteins in
appropriate colonies (Figure 2). Met was seen to be
expressed on the plasma membrane and/or in the
cytoplasm of 68% (15 out of 22) colonies positive for
ret.
Met overexpression is not correlated with proliferation
rate
We next addressed the question of whether the
increased expression of met was speci®cally due to
oncogene activation or whether it was simply a
secondary consequence of the resulting stimulation of
cell proliferation.
A set of 14 randomly-selected colonies induced by
mutant ras were subjected to simultaneous analysis of
met protein expression (by immuno¯uorescence) and
DNA synthesis (by nuclear bromodeoxyuridine incorporation). Twelve colonies were positive for met
and showed a BrdU labelling index of 57.7%+2.8
(mean+s.e.m.); two were negative for met and had a
labelling index of 69.5%+9.5 (means+s.e.m.). Figure
3 illustrates the two extremes of met positivity in a ras
colony with low proliferation rate (Figure 3a)
contrasting with met negativity in a rapidly proliferating clone (Figure 3b). Overall, there was therefore no
correlation between the level of met expression and
proliferation rate.
We also induced colonies by expression of SV40
large T antigen as reported previously (Lemoine et al.,
1989b; Bond et al., 1996 in press). In contrast with ras
and ret, 12 SV40 T-induced colonies investigated by
immunoperoxidase never showed met immunostaining
exceeding that of the surrounding normal cells (Figure
1g). Nine colonies were also analysed by simultaneous
immuno¯uorescence and bromodeoxyuridine labelling.
Again all were met-negative despite a BrdU index
(66.3+3.7) similar to that of the met-positive ras
clones (57.7%+9.5). An example of an SV40 T colony
double labelled for bromodeoxyuridine and met is
shown in Figure 3c.
Induction of met by mutant ras in a thyroid cell line
(R18)
To provide more quantitative support for our ®ndings
in primary cultures, we also used a thyroid epithelial
cell line previously developed in our laboratory by
infecting an SV40 transformed line with a retroviral
vector encoding a mutant H-ras under the control of a
metallothionein promoter (Dawson et al., 1995). The
expression of the oncogene in this cell line (R18) is
inducible by zinc ions and has been shown to not
signi®cantly a€ect cell proliferation (Dawson et al.,
1995). After 4 days induction with Zn2+ immunoprecipitation followed by Western analysis showed a
fourfold increase in met expression in R18 cells,
assessed by densitometry (Figure 4).
Uninduced R18 showed higher expression than the
parental HT-ori3 line owing to vector leakiness, as
reported previously for this cell line (Dawson et al.,
1995).
Discussion
Investigation of the direct phenotypic consequences of
activating a single oncogene is ideally performed in the
context of a normal cell derived from the appropriate
human tissue, cell lines always being liable to
misleading results due to pre-existing, additional
genetic abnormalities. Primary epithelial cells have
Induction of c-met by activated oncogenes
M Ivan et al
however been infrequently used as recipients for
activated oncogenes because of diculties of isolation, and/or refractoriness to growth in culture and to
gene transfection (Wynford-Thomas, 1993). The
thyroid follicular cell is an exception. Viable monolayer cultures of epithelium can readily be isolated
which are near-quiescent in simple culture conditions
but show a dramatic proliferative response following
Figure 1 Expression of met protein in thyroid epithelial cells induced by mutant H-ras and ret oncogenes but not by SV40 T.
Colonies were induced in primary culture by infection with retroviral vectors encoding mutant ret (a ± c), ras (d, e) or SV40 T (f, g)
oncogenes and analysed by immunocytochemistry without prior selection in G418 in order to permit comparison with the
surrounding normal epithelial monolayer. High level expression, restricted to the colony, is seen for ret (a), ras (d) and SV40 T (f)
proteins. Matched colonies from the same experiment show a greatly increased expression of met in colonies induced by ret (b) and
ras (e) but not SV40 T (g). Localization of met was predominantly to the plasma membrane as shown by high magni®cation of part
of ret-induced colony (c). Immunoperoxidase, no counter-stain. Magni®cations: a, b, f and g: 680; d, e: 640; c: 6250
2419
Induction of c-met by activated oncogenes
M Ivan et al
2420
retrovirally-mediated transduction of activated oncogenes (ras or ret) known to be associated with tumour
initiation in vivo (Lemoine et al., 1990; Bond et al.,
1994). The resulting colonies represent a good in vitro
a
b
d
e
analogue of the two common types of thyroid
tumour ± follicular and papillary ± associated respectively with these two oncogenes and provide an
excellent `isogeneic' model for studying the speci®c
c
Figure 2 Co-expression of ret and met proteins in ret-induced colonies of thyroid epithelium. Photomicrographs of the same
colony (with surrounding normal cells) (a) showing expression of ret by immuno¯uorescence using a TRITC-labelled second
antibody; (b) expression of met using an FITC-linked antibody; (c) phase contrast (680). (d) and (e): higher magni®cation of a
di€erent colony showing co-localization of ret (d) and met (e) in the same cells (6150)
a
b
c
Figure 3 Simultaneous analysis of met protein expression by immuno¯uorescence using a TRITC-labelled second antibody (red)
and DNA synthesis by nuclear bromodeoxyuridine incorporation detected by an FITC-labelled antibody (green). Example of a
slowly proliferating ras colony with detectable met expression (a) contrasting with a rapidly proliferating ras colony with very low
level of met (b). SV40 T colonies were all negative for met expression (c). Magni®cation: 6150
Induction of c-met by activated oncogenes
M Ivan et al
2
3
▲
1
Figure 4 Western blot analysis showing increased expression of
met protein in R18 cells following induction of the mutant ras
gene. The p145 b chain (closed arrowhead) and the p170
precursor (open arrowhead) of met were detected using a rabbit
polyclonal against its C-terminal region. Lane 1: R18 cells 4 days
after induction of mutant ras by Zn2+; lane 2: uninduced R18
cells; lane 3: the parental cell line HT-ori 3
phenotypic changes induced by each gene. Here we
have used this system to test the hypothesis that one
such consequence of ras or ret activation is c-met
overexpression.
The results show clearly that met protein expression,
as demonstrated by immunoperoxidase or immuno¯uorescence analysis was greatly elevated in the
majority of epithelial colonies expressing activated ras
or ret oncogenes in comparison to its barely-detectable
level in the background normal cells. (The reason for
its absence in a minority of colonies is presently
unclear.) Met was detected both at the plasma
membrane and in the cytoplasm of transformed cells,
which is in agreement with the localization described in
thyroid tumours in vivo (Prat et al., 1991; Thomas et
al., 1995). Quantitation of the magnitude of met
protein overexpression by Western blot analysis was
not practicable using early oncogene-induced colonies
due to their limited size. Instead this was achieved
using a cell line model (R18) in which the induction of
mutant ras expression was shown to lead to a fourfold
increase in met protein expression.
Three independent results suggest that met overexpression is not simply a passive consequence of any
deregulation of the cell cycle, but is related more
speci®cally to `upstream' signals generated by ras, ret
(and possibly other) oncogenes. Firstly, undetectable or
very low levels of met protein (similar to uninfected
quiescent cells) were observed in all SV40 T-induced
colonies despite similar proliferation rate to ras/ret
colonies. Secondly, in ras colonies, where the labelling
index showed considerable clone to clone variability,
this was not correlated with the intensity of met
staining. Thirdly, in R18 cells met expression was
increased by the induction of the ras oncogene without
any corresponding e€ect on cell proliferation.
The mechanism of oncogene-dependent met overexpression has not been addressed here and will form
the subject of a future study. Western blot and
immunocytochemical studies of clinical thyroid samples have led to the suggestion of a post-transcriptional
level of control (Di Renzo et al., 1991). However, this
has not been supported, as far as we are aware, by any
direct comparison of mRNA and protein levels
between normal and neoplastic thyroid tissue. In
other tumour systems, for example colon and pancreas
(also noted for their high frequency of ras oncogene
activation) met mRNA is clearly increased in the
tumors (Liu et al., 1992; Di Renzo et al., 1995a)
suggesting a possible transcriptional level of control.
Furthermore, met transcription in cell lines has been
shown to be stimulatable by serum or TPA (Boccaccio
et al., 1994), which are likely to share some signalling
pathways with the activated oncogenes employed here.
Analysis of the met promoter region has so far
revealed the presence of AP2 sites, which could
mediate such signals (Gambarotta et al., 1994).
Taken together, our data demonstrate that activated
oncogenes can induce met overexpression in thyroid
epithelium. Although our experimental model does not
reveal any resulting proliferative advantage, this is
readily explained by the growth-factor-rich in vitro
environment rendering any met-mediated growth
signals redundant. In the in vivo tumour situation,
however, increased met expression may well confer a
signi®cant growth advantage by sensitizing tumour
epithelium to HGF derived from adjacent stromal cells,
thereby contributing to the proliferogenic action of
oncogenes such as ras and ret. It is interesting in this
regard that stroma is particularly abundant not only in
papillary thyroid (Williams et al., 1988) but also in
pancreatic cancer (Scarpelli, 1994), which also exhibits
high levels of met expression (Di Renzo et al., 1995a).
We have recently demonstrated (Eccles et al., 1996)
that HGF can stimulate the growth of ret-induced
thyrocyte colonies, giving further support to such a
paracrine mechanism.
Materials and methods
Cell cultures
Monolayer primary cultures were prepared from surgical
samples of histologically- normal human thyroid tissue, by
protease digestion and mechanical disaggregation as
described previously (Williams and Williams, 1988).
Cultures were maintained in a 2 : 1 : 1 (by volume) mixture
of Dulbecco's modi®ed Eagles' medium, Ham's F-12 and
MCDB 104 (Bond et al., 1992), supplemented with 10%
fetal calf serum (Imperial Laboratories, London, UK). R18
cells and the parent cell line HT-ori3 were maintained in
RPMI medium supplemented with 10% fetal calf serum
and Geneticin (G418, Gibco BRL) 400 mg/ml. Ras gene
expression was induced in R18 using 100 mM ZnSO4 for
48 h followed by 50 mM in serum free RPMI containing
0.01% BSA (Dawson et al., 1995). As controls, R18 and
HT-ori3 were grown in similar conditions but in the
absence of Zn2+.
Retroviral vectors
Monolayer primary cultures were infected with the
following defective amphotropic vectors:
(i) psi-CRIP-DOEJ coding for the val12 mutant of human
H-ras, described previously (Lemoine et al., 1990);
(ii) psi-CRIP-PTC (Bond et al., 1994) encoding an activated
human ret protein from a rearranged cDNA (Santoro et
al., 1993);
(iii) psi-CRIP-SVU19 (Bond et al., in press) expressing the
non-origin-binding U19 mutant of SV40 large T (Jat
and Sharp, 1986).
All the vectors also express the neo gene, allowing selection
of infected cells using Geneticin, 400 mg/ml.
2421
Induction of c-met by activated oncogenes
M Ivan et al
2422
Retroviral gene transfer
Primary cultures were plated at 5610 cells per 60 mm dish
and infected 2 days later with retrovirus-containing
medium from near-con¯uent cells containing 8 mg/ml
polybrene as described previously (Bond et al., 1994).
Four days later, cells were passaged with or without G418
into single-well `permanox' chamber-slides (Life Technologies Ltd, Paisley, UK).
5
Immunocytochemistry
Twelve days after infection with vectors encoding ras or ret
(24 days for SV40 large T), monolayers were ®xed in
methanol : acetone (1 : 1) and analysed by immunocytochemistry using the following primary antibodies diluted as
appropriate in PBS/1% BSA:
(i) Y 13-259 rat anti-ras p21 monoclonal antibody (Furth
et al., 1982);
(ii) rabbit polyclonal anti-ret, directed against a C-terminal
20-mer of p57 ret (Ishizaka et al., 1992) (kindly provided
by T Ushijima, NCCRI, Tokyo);
(iii) mouse monoclonal anti-large T antibody pAb 419
(Harlow et al., 1981);
(iv) mouse monoclonal anti-met antibody DO24 (Prat et al.,
1991), directed against the extracellular domain of
human c-met oncogene product.
For immunoperoxidase, a standard indirect procedure was
followed, using HRP-conjugated rabbit anti-mouse IgG, swine
anti-rabbit IgG or sheep anti-rat IgG (Dako) (1 : 100 dilution) as
appropriate. For double immuno¯uorescence, ret was detected
by the rabbit polyclonal followed by TRITC-conjugated goat
anti- rabbit Ig (Southern Biotech, Alabama); met was detected
by mouse monoclonal D024 followed by FITC-conjugated goat
anti-mouse IgG (Southern Biotech, Alabama). A pilot study
using mouse monoclonal anti-cytokeratin LE61 and rabbit
polyclonal anti-thyroglobulin as primary antibodies on human
thyroid cells was carried out to check the species-speci®city of
the second antibodies. No cross reaction was detected.
A special protocol was used to investigate any potential
correlation between met expression and cell proliferation.
Brie¯y, colonies were incubated with 10 mM 5-bromo-2'deoxyuridine (Boehringer) for 24 h followed by 15 min
®xation in 4% formaldehyde. Fixed cells were washed in
PBS for 30 min, permeabilized in 0.2% Triton6100 for
15 min, `blocked' in 10% fetal calf serum for 30 min,
incubated for 1 h with 1 : 400 dilution of DO24 ascites
¯uid, followed by TRITC-conjugated goat anti mouse IgG
(Southern Biotech, Alabama). Finally, FITC-labelled mouse
monoclonal antibody to bromodeoxyuridine (Boehringer)
was added (following the manufacturer's protocol).
Immunoprecipitation and Western blotting
Cells were lysed in NET bu€er (150 mM NaCl, 50 mM Tris,
1%NP40, 10 mM aprotinin), centrifuged at 20 000 g for
30 min to remove high molecular weight DNA. Protein
concentration in the lysate was measured using a
Coomassie Plus Protein Assay kit (Pierce) and 200 mg of
protein were used from each sample. 1 ml DO24 ascites
¯uid was added to immunoprecipitate met protein (overnight incubation) followed by extraction with Protein G
Sepharose (Pharmacia) for 40 min. Bound proteins were
resolved on a 7.5% polyacrylamide gel alongside high
molecular weight markers (Biorad) and electroblotted onto
a PVDF membrane. This was probed with 1 mg/ml rabbit
polyclonal antibody recognising the carboxyterminal
region of met protein (C-28, Santa Cruz). Met protein
was visualised with an ECL system using HRP-coupled
anti rabbit IgG (Amersham).
Acknowledgements
We are grateful to the Cancer Research Campaign for grant
support, to Dr Ushijima for supply of antibody and to Dr
Bharat Jasani for assistance with immunocytochemistry
References
Boccaccio C, Gaudino G, Gambarotta G, Galimi F and
Comoglio PM. (1994). J. Biol. Chem., 269, 12846 ± 12851.
Bond JA, Dawson T, Lemoine N and Wynford-Thomas D.
(1992). Mol. Carcinogen., 5, 129 ± 135.
Bond JA, Wyllie FS, Rowson J, Radulescu A and WynfordThomas D. (1994). Oncogene, 9, 281 ± 290.
Bond JA, White D, Ness GO, Rowson J, Ivan M and
Wynford-Thomas D. Int. J. Cancer, in press.
Bottaro DP, Rubin JS, Faletto DL, Chan AML, Kmiecik
TE, Vande Woode GF and Aaronson SA. (1991). Science,
251, 802 ± 804.
Dawson TP, Radulescu A and Wynford-Thomas D. (1995).
Cancer Res., 55, 915 ± 920.
Dawson TP, Wyllie FS and Wynford-Thomas D. (1991). Br.
J. Cancer, 63, 897 ± 900.
Di Renzo MF, Narsimhan RP, Olivero M, Bretti S,
Giordano S, Medico E, Gaglia P, Zara P and Comoglio
PM. (1991). Oncogene, 6, 1997 ± 2003.
Di Renzo MF, Olivero M, Ferro S, Prat M, Bongarzone I,
Pilotti S, Bel®ore A, Costantino A, Vigneri R, Pierotti MA
and Comoglio PM. (1992). Oncogene, 7, 2549 ± 2553.
Di Renzo MF, Poulsom R, Olivero M, Comoglio PM and
Lemoine NR. (1995a). Cancer Res., 55, 1129 ± 1138.
Di Renzo MF, Olivero M, Giacomini A, Porte H, Chastre E,
Mirossay L, Nordlinger B, Bretti S, Bottardi S, Giordano
S, Plebani M, Gespach C and Comoglio PM. (1995b).
Clin. Cancer Res., 1, 147 ± 154.
Di Renzo MF, Olivero M, Katsaros D, Crepaldi T, Gaglia P,
Zola P, Sismondi P and Comoglio PM. (1994). Int. J.
Cancer, 58, 658 ± 662.
Eccles N, Ivan M and Wynford-Thomas D. (1996). Mol.
Cell. Endocrinol., 117, 247 ± 251.
Furth ME, Davis LJ, Fleurdelys B and Scolnick EM. (1982).
J. Virol., 43, 294 ± 304.
Gambarotta G, Pistoi S, Giordano S, Comoglio PM and
Santoro C. (1994). J. Biol. Chem., 269, 12852 ± 12857.
Grieco M, Santoro M, Berlingieri MT, Melillo RM, Donghi
R, Bongarzone I, Pierotti MA, Della Porta G, Fusco A
and Vecchio G. (1990). Cell, 60, 557 ± 563.
Harlow ED, Crawford LV, Pim DL and Williamson NM.
(1981). J. Virol., 39, 861 ± 869.
Houldsworth J, Cordon-Cardo C, Ladanyi M, Kelsen DP
and Chaganti RSK. (1990). Cancer Research, 50, 6417 ±
6422.
Igawa T, Kanda S, Kanetake H, Saitoh Y, Ichihara A,
Tomita Y and Nakamura T. (1991). Biochem Biophys. Res.
Commun., 174, 831 ± 838.
Ishizaka Y, Shima H, Sugimura T and Nagao M. (1992).
Oncogene, 7, 1441 ± 1444.
Jat PS and Sharp PA. (1986). J. Virol., 59, 746 ± 750.
Lemoine NR, Mayall ES, Wyllie FS, Williams ED, Goyns
M, Stringer B and Wynford-Thomas D. (1989a).
Oncogene, 4, 159 ± 164.
Lemoine NR, Mayall ES, Jones T, Scheer D, McDermitt S,
Kendall-Taylor P and Wynford-Thomas D. (1989b). Br. J.
Cancer, 60, 897 ± 903.
Lemoine NR, Staddon S, Bond JA, Wyllie FS, Shaw JJ and
Wynford-Thomas D. (1990). Oncogene, 5, 1833 ± 1837.
Induction of c-met by activated oncogenes
M Ivan et al
Liu C, Park M and Tsao MS. (1992). Oncogene, 7, 181 ± 185.
Matsumoto K, Tajima H and Nakamura T. (1991a).
Biochem. Biophys. Res. Commun., 176, 45 ± 51.
Matsumoto K, Hashimoto K, Yoshikawa K and Nakamura
T. (1991b). Exp. Cell Res., 19, 114 ± 120.
Naldini L, Vigna E, Narsimhan RP, Gaudino G, Zarnegar
R, Michalopoulos GK and Comoglio PM. (1991).
Oncogene, 6, 501 ± 504.
Namba H, Rubin SA and Fagin JA. (1990). Mol. Endocrin.,
4, 1474 ± 1479.
Park M, Dean M, Cooper CS, Schmidt M, O'Brien SJ, Blair
DG and Vande Woude GF. (1986). Cell, 45, 895 ± 904.
Ponzetto C, Giordano S, Peverali F, Della Valle G, Abate
ML, Vaula G and Comoglio PM. (1991). Oncogene, 6,
553 ± 559.
Prat M, Narsimhan RP, Crepaldi T, Nicotra MR, Natali PG
and Comoglio PM. (1991). Int. J. Cancer, 49, 323 ± 328.
Rubin J, Chan AML, Bottaro DP, Burgess WH, Taylor WG,
Chech AC, Hirsch®eld DW, Wong J, Miki T, Finch PW
and Aaronson SA. (1991). Proc. Natl. Acad. Sci. USA, 88,
415 ± 419.
Santoro M, Carlomagno F, Hay ID, Herrmann MA, Grieco
M, Melillo R, Pierotti MA, Bongarzone I, Della Porta G,
Berger N, Peix JL, Paulin C, Fabien N, Vecchio G, Jenkins
RB and Fusco A. (1992). J. Clin. Invest., 89, 1517 ± 1522.
Santoro M, Melillo RM, Grieco M, Berlingieri MT, Vecchio
G and Fusco A. (1993). Cell Growth Di€., 4, 77 ± 84.
Scarpelli DG. (1994). Pathology. Rubin E and Farber JL
(eds). JB Lippincott Company: Philadelphia, pp. 794 ±
797.
Sonnenberg E, Meyer D, Weidner KM and Birchmeier C.
(1993). J. Cell Biol., 123, 223 ± 235.
Stoker M, Gherardi E, Perryman M and Gray J. (1987).
Nature, 327, 239 ± 242.
Strain AJ, Ismail T, Tsubouchi H, Arakaki N, Hishida T,
Kitamura N, Daikuhara Y and McMaster P. (1991). J.
Clin. Invest., 87, 1853 ± 1857.
Suarez HG, du Villard JA, Severino M, Caillou B,
Schlumberger M, Tubiana M, Parmentier C and Monier
R. (1990). Oncogene, 5, 565 ± 570.
Thomas GA, Takahashi MH, Davies HG, Hooton N and
Williams ED. (1995). J. Endocrinology, 144S, RC3.
Williams DW and Williams ED. (1988). Thyroid Tumours.
Wynford-Thomas D and Williams ED (eds). Churchill
Livingstone: Edinburgh, pp. 57 ± 65.
Williams DW, Williams ED and Wynford-Thomas D.
(1988). Br. J. Cancer, 57, 535 ± 539.
Wynford-Thomas D. (1993). Cancer Surveys, 16, 115 ± 134.
2423