.2...‘

.3. . L25

‘ '_-.c‘~_...;.

.l' 4
v—v

-I ’.'L;

_4- ,‘JR. . ,1

 

 

.. . “I
7W". W

up 4):."
k. L ,n
"'»‘g (hr/‘9'“
J‘I

,J..
TW‘

”"I‘

”(:53 pt! f5:
W ,

a". 1;" 317.,“
$7,973,432;

1'
‘1‘: ‘1'
. " “3.3.53

 

.2,

4:44

V I
453:4}? 55““; "its? 2:,-
> “7:44 “-V " .'
.m 4334+ n r ,f'” ff-f-fs’u
W Aw+493;
- ’{ fig? ‘4'.

’ {‘1' (Egrfiw C"‘-‘-.»“3?.\" ‘I-V. ”a
. Ignx'lufi ,‘fl, V>WI"'J”:' " ,ch‘;:. &}
”my ,1\é~3;,_ {‘g. .
pita-39. ' 34%;};

921: N 1‘ > x u
‘..' ‘ §
\ {:7}... ' I. \‘Jhw mkfi
A. ‘9‘." a

. ..‘; - .. 7“ 7" , '

w- 54-433., I“ 7) v 1-“ . _, 9y, ’2‘. P;%§‘f,~%~
‘ ."i‘Ja'yy 2€l ' . if} v..\ 5:}, i’k'i ) ‘ -
kh'y'fi' .4451"; ;_. ‘3 J}; '3‘; fl! hf}, )‘l ‘1‘»; ‘

. a. . .

' a c a fi

' x '4' v. .' W “ .

$343??? ’t 'I f: V‘iggéhsi I
'R'e'nfl‘? K.—

I " I\’\~
. :‘.~:"A::.“‘$_:<=~ x
c K 5' 5C).
,4 gék‘fl‘ I.’>:';§£g
45],.

~ ‘3’,” '41. 3
.. .1.’ .. 4.! , a. .
‘ "‘é’i’éauzzs :4 a
I, "0.11%?‘Agxfi‘
. wing}. ‘34?
"4 Wk»- YQM “‘
‘ 3&1. hfin:
V ' ‘2"

{Qt-gm:

' , YE”:

'F/q'" 1. "'55.. 9 H : A" I

:1 3:5; 'v4' v'. +91%. 1“: :4" ‘31:“
‘ v‘. V“ VA

“‘4 :63- 3'
‘ z 0:3,) . , 4 :. "=3: ,4 33,,“
fl 4‘5": ' i ‘, '2. ,4 w? 4""? 1:5 'a'fikglii
41': 4 ,. ,(1: ‘Viflt'a.

{Mn-vi“): " ‘,
.“figgfi, ‘ .
22*‘54W‘ ' '-" 5%“ ~' - «f 3.4%

1,4,3” ‘.;l’(‘i "t 3 ‘ 1%“) .7 ,j'< 9A. {.31.
“3} 1' ' i H .‘ f? .
“:5” 0.).“ I“ {'7
“ {AS-141’

' 355‘
r‘ 45:55:; ., "‘4.
u
, REI‘VJ

4:

1'34»; 3.4; $54
‘1 g, f I Pi» if
in? 3:13;- 3:!
-A‘"2‘

> g ~35
, ~
‘2} ”’23,?! .;-
..
i!
«a

“figs“ '{5 'f‘

l o
‘ A o.( .
4 r. y _ .
V 6;) "
$1,213“ Ln F
‘12:,
; £317,
44‘3", 4%
,. 4
, -

V

. l‘

This is to certify that the

dissertation entitled

Expression of Growth Factor Genes in Transformed Human
Fibroblasts and Ceiis Derived From Human Fibrosarcomas:
A Possible Mechanism for Replication in the Absence
of Exogenous Growth Factors.

presented by

Robert J. Schiiz

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Biochemistry

 

 

EVA (11 {Mg T

Major professor

Date 10/21/88

MS U is an Affirmative Action/Equal Opportunity Institution 0 12771

_..__F__———v___

 

_J

LIBRARY
Mi‘l’hiflen State

‘Pvniversity J

a

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
was return on or betore date due.

T
DATE DUE DATE DUE DATE DUE l'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EXPRESSION OF GROHTH FACTOR GENES IN TRANSFORMED HUMAN
FIBROBLASTS AND CELLS DERIVED FROM HUMAN FIBROSARCOMAS:
A POSSIBLE MECHANISM FOR REPLICATION IN THE ABSENCE
OF EXOGENOUS GROWTH FACTORS.

By

Robert J. Schiiz

A DISSERTATION

Submitted to
Michigan State University
in partiai fulfiTTment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1988

XIV

ABSTRACT
EXPRESSION OF GROWTH FACTOR GENES IN TRANSFORMED HUMAN FIBROBLASTS AND

CELLS DERIVED FROM HUMAN FIBROSARCOMAS: A POSSIBLE MECHANISM FOR
REPLICATION IN THE ABSENCE OF EXOGENOUS GROWTH FACTORS.

by
Robert J. Schilz

This study was undertaken to identify characteristics of malignantly
transformed human fibroblasts which reflect the cellular changes required
for malignant transformation. When tested in a serum-free medium developed
in this laboratory, cell lines derived from human fibrosarcomas (the type
of tumor expected from the malignant transformation of human fibroblasts)
or transformed human fibroblast cell lines grew in serum-free medium
lacking exogenously added growth factors and containing reduced calcium
(0.1 mM). Normal fibroblasts remained viable under these conditions but
did not replicate. These growth factor independent cells responded in one
of two ways to the addition of various growth factors to the medium.
Transformed cells that grew rapidly in the absence of exogenously added
growth factors did not increase their rate of replication in response to
such growth factors. Transformed fibroblasts which grew slowly under these
conditions increased their rate of replication in response to exogenously
added growth factors. The rate of replication of these transformed cell
lines in culture correlated with their rate of tumor formation in athymic
mice.

One explanation for the observed growth factor independence is that
transformed human fibroblasts synthesize growth factors and are able to

replicate by autocrine stimulation. To test this hypothesis, cellular RNA

from the transformed human fibroblasts was analyzed for the expression of
genes coding fer specific growth factors and factors which have been
implicated in the process of growth factor signal transduction. Those cell
lines which replicated rapidly in the absence of exogenously added growth
factors, showed increased expression of more than one growth factor.
Those cell lines which grew slowly showed increased expression of only one
or none of the growth factor genes. The expression of such genes could
explain the observed growth factor independence and may reflect the types
of cellular changes required for the malignant transformation of human

cells.

To my parents

Eugene and Helen Schilz

iv

ACKNOWLEDGMENTS

I am grateful to my graduate advisor and research director, Dr. J.
Justin McCormick for his guidance and support throughout my graduate
studies. I also owe special thanks to Dr. Veronica Haher for her insights,
support and critical review and to all members of the Carcinogenesis
Laboratory past and present that have influenced this thesis. Thanks also
to Carol Howland for assistance in typing, editing and surviving.

I am thankful for the love and support of my parents and family

during my endeavors as well as the unfailing loyalty of many special

friends.

TABLE OF CONTENTS

Page

LIST OF TABLES .................................................. ix
LIST OF FIGURES ................................................. x
INTRODUCTION .................................................... 1
CHAPTER 1. LITERATURE REVIEW ................................... 5
A. Multistep Nature of Carcinogenesis ...................... 5

1. Epidemiological Evidence ............................. 5

2. Genetic Evidence ..................................... 5

3. Evidence from Dysplastic Histology of Premalignant
Conditions in Humans ................................. 8
B. Experimental Investigation of Multistep Carcinogenesis

Using Fibroblasts in Culture ............................ 10

1. Studies Utilizing Rodent Fibroblasts ................. 10

a. Spontaneous Transformation ........................ 11

b. Carcinogen-Induced Transformation ................. 12

c. Oncogene-Mediated Transformation .................. 14

2. Studies Utilizing Human Fibroblasts .................. 16

a. Spontaneous Transformation ........................ 16

b. Carcinogen—Induced Transformation ................. 18

c. Oncogene-Mediated Transformation .................. 21

C. Properties Associated with the Transformed Phenotype.... 23

l. Tumorigenicity ....................................... 23

2. Anchorage Independence ............................... 25

3. Focus Formation ...................................... 27

4. Growth in Medium Containing Reduced Calcium .......... 28

5. Growth Factor Independence ........................... 29

6. Infinite Lifespan .................................... 31

7. Other Correlates ..................................... 32

D. Growth Requirements of Fibroblasts in Culture ........... 33

1. Rodent Fibroblasts ................................... 33

2. Human Fibroblasts .................................... 35

E. Mechanisms of Growth Factor Independence ................ 37

l) Autocrine Stimulation of Cell Growth ................. 37

a. B-Chain of Platelet Derived Growth Factor (PDGF-B) 37
b. A-Chain of Platelet Derived Growth Factor (PDGF—A) 39

vi

c. Transforming Growth Factor Alpha (TGF-a) .......... 40

d. Epidermal Growth Factor (EGF) ..................... 42

e. Transforming Growth Factor Beta (TGF-b) ........... 43

f. Fibroblast Growth Factors (FGF’s) ................. 44

g. Insulin-Like Growth Factors (IGF’s) ............... 45

2) Internal Mechanisms .................................. 47
a. Early Gene Products ............................... 47

1. My; Gene ....................................... 48

2. figs Gene ....................................... 50

3. p53 Gene ....................................... 52

b. Genes Involved in Signal Transduction - Ra; Genes. 53

LIST OF REFERENCES .............................................. 57

CHAPTER II. MODIFICATION OF SERUM FREE MEDIUM TO ALLOW
DETECTION OF THE EFFECTS OF EXOGENOUSLY ADDED PROTEIN

GROWTH FACTORS .................................................. 82
Introduction .................................................... 82
Materials and Methods ........................................... 82
Cells and culture medium ..................................... 82
Growth factors ............................................... 83
Growth factor assay medium ................................... 83
Assay of cells for growth requirements ....................... 83
Results and Discussion .......................................... 84
References ...................................................... 88
Tables .......................................................... 89
Figures ......................................................... 91

CHAPTER III. OVEREXPRESSION OF MULTIPLE GROWTH FACTOR
GENES BY FIBROSARCOMA-DERIVED AND OTHER TRANSFORMED

HUMAN FIBROBLASTS IN CULTURE .................................... 99
Abbreviations ................................................... 100
Abstract ........................................................ 101
Introduction .................................................... 103
Materials and Methods ........................................... 105

Cells and culture conditions ................................. 105

vii

Growth factors ............................................... 105

Growth factor assay medium ................................... 105
Assay of cells for growth factor requirements ................ 106
Assay for tumorigenicity ..................................... 107
Assay of cells for expression of mRNA ........................ 107
Gel electrophoresis .......................................... 108
Capillary gel blotting ....................................... 108
Probe labelling .............................................. 109
Probes used .................................................. 110
Hybridization and washing .................................... 111
Radiography .................................................. 111
Results ......................................................... 112

Analysis of the type of tumor formed by transformed
cell lines ............................................. 112

Analysis of the growth factor requirements of normal

fibroblasts and their response to exogenously

added growth factors ................................... 112
Analysis of the growth factor requirements of

transformed fibroblasts and their response to

exogenously added growth factors ....................... 113
Expression of genes coding for protein growth factors ........ 115
Expression of factors putatively involved in mitogenic
signal transduction .......................................... 117

Expression of N-ras .................................... 118

Expression of K-[as .................................... 118

Expression of c-myg and c-fgs .......................... 119

Discussion ...................................................... 120
References ...................................................... 125
Tables .......................................................... 131
Figures ......................................................... 135

viii

LIST OF TABLES

CHAPTER 11. page

Table 1. Effect of EGF and insulin on the growth of normal
human fibroblast cell line SL 68 in serum-free medium ................. 89

Table 2. Effect of dexamethasone and prostaglandin E] on the growth
of normal human fibroblast cell line SL 68 ............................ 90

CHAPTER III.
Table 1. Fibroblastic cell lines studied ............................... 131

Table 2. Tumorigenicity of fibroblasts transformed in culture or
fibrosarcoma-derivedcelllines ........................................ 132

Table 3. Tumorigenicity and growth factor independence and
responsiveness of the human fibroblast cell lines ...................... 133

Table 4. Expression of genes coding for peptide growth factors and

some proto-oncogenes in transformed and non-transformed human
fibroblasts ............................................................ 134

ix

LIST OF FIGURES

CHAPTER 11. page

Figure 1. Effect of medium calcium concentration on the
replication of diploid human fibroblast cell line SL 68 .............. 92

Figure 2. Effect of medium EGF concentration on the replication
of diploid human fibroblast cell line SL 68 .......................... 94

Figure 3. Effect of medium PDGF concentration on the replication
of diploid human fibroblast cell line SL 68 .......................... 96

Figure 4. Effect of medium insulin concentration on the
replication of diploid human fibroblast cell line SL 68 .............. 98

CHAPTER III.

Figure 1. Effect of growth factors, serum and calcium on the
replication of human fibroblast cell lines ........................... 136

Figure 2. Northern blot analysis of the expression of
TGF-a in non-transformed, tumor-derived, or transformed human
fibroblasts ......................................................... 138

Figure 3. Northern blot analysis of the expression of
TGF-b in non-transformed, tumor derived or transformed human
fibroblasts ......................................................... 140

Figure 4. Expression of PDGF-B in non-transformed, tumor
derived or transformed human fibroblasts ............................. 142

Figure 5. Northern blot analysis of the expression of
PDGF-A in non-transformed, tumor derived or transformed human
fibroblasts ......................................................... 144

Figure 6. Overexpression of basic FGF by the spontaneously
transformed cell line VIPzFT ......................................... 146

Figure 7. Northern blot analysis of the expression of
N-[as in non- transformed, tumor derived or transformed human
fibroblasts .......................................................... 148

X

page

Figure 8. Northern blot analysis of the expression of IGF-II and
c-myg in non—transformed, tumor derived or transformed human
fibroblasts .......................................................... 150

Figure 9. Results of hybridization with a probe complementary to
the K-ras gene ....................................................... 152

xi

INTRODUCTION

Epidemiological evidence suggests that the development of cancer in
humans is a multistep process (Doll, 1978; Peto, 1977). Among the main
goals of cancer research are; identifying these steps or cellular changes
and understanding how they lead to the malignant state. It is widely
believed that some of these steps are caused by exposure to environmental
carcinogens (Health and Welfare Department, 1977). Early experiments
designed to identify such carcinogens, consisted of treating experimental
animals with suspected carcinogens and examining the animals for the
appearance of tumors. In these studies, tumors typically arose in animals
only after' multiple» exposures of’ carcinogens and after long latent
periods. The tumors were clonal in origin. These observations are
consistent with the hypothesis that multiple changes are required for
malignant transformation. Such animal experiments, however, do not
ordinarily allow the analysis of the specific changes which cells undergo
as they become malignantly transformed. For this reason, many
investigators have utilized mammalian cells in culture to study the
stepwise acquisition of transformed phenotypes. For example, Barrett and
Ts’o (1978) reported that primary cultures of Syrian hamster embryo
fibroblasts, treated with carcinogen, become capable of forming tumors
only after the acquisition of multiple transformed phenotypes and that
these phenotypes occur after extended passage in culture. Because of the

long period between the exposure of the cells to carcinogen and the

2
isolation of malignantly transformed cells it is difficult to determine
the mmlecular mechanism(s) which are responsible for malignant
transformation of these cells. The identification of activated cellular
oncogenes which are homologous to retroviral transforming genes (Varmus,
1984). suggested molecular mechanisms which could be involved in malignant
transformation. Several groups of investigators introduced such oncogenes
into rodent cells and found that multiple, cooperating oncogenes were
required to transform primary culture rodent fibroblasts. (Land gt gl.,
1983; Ruley gt gl., 1983). The tumors formed by such transformed cells,
however, were not malignant as judged by their inability to grow
progressively in the test animal, invade adjacent tissue, or metastasize.
Clearly, additional transformed characteristics are required to produce
such a malignantly transformed cell.

Similar in vitro transformation studies have been performed using

 

diploid human fibroblasts, but few have yielded malignantly transformed
cells. Two such studies are those of Namba gt g1. (1986, 1988) and Hurlin
gt gl. (1988). These workers obtained malignantly transformed fibroblast
cell lines following the introduction of the H-tgg oncogene into non-
tumorigenic, infinite lifespan human fibroblasts. These infinite lifespan
cells were isolated after repeated carcinogen treatment of diploid human
fibroblasts (Namba gt gl., 1981) or after transfection with the v-myg
oncogene (Hurlin gt 11., 1988). Since greater than 80% of human cancer is
of epithelial origin (Cairns, 1975), one might suspect that the inability
to routinely transform human fibroblasts may be due to some inherent
resistance to transformation which cells of mesenchymal origin possess.

However human epithelial cells in culture appear similarly resistant to

3

transformation by carcinogen treatment or by oncogene transfection
(Chang, 1986).

This lack of success may reflect the need to treat or transfect
populations of human cells which have acquired phenotypes intermediate
between the normal and the malignant state which investigators have been
unable to recognize and isolate. For this reason, I have chosen to study
human fibrosarcoma-derived cell lines and other transformed human
fibroblasts in an attempt to identify changes common to these transformed
human cells which represent important intermediates in the multistep
process of malignant transformation. Human fibrosarcomas are tumors which
arise from the in XLLO transformation of human fibroblasts and thus

represent the kind of cell which most lg vitro transformation experiments

 

have failed to produce.

One property which is common to all tumors is the ability to
proliferate in the body while normal cells in the same tissue do not. A
number of in ytttg selection techniques commonly used in transformation
experiments utilize growth conditions which allow the selective
proliferation of transformed fibroblasts (See literature review). Since
protein growth factors have been shown to control the replication of non-
transformed fibroblasts in culture (Scher gt gl., 1978; Westermark and

Heldin, 1985), I assayed the i vitro growth characteristics of

 

transformed human fibroblasts to determine whether their growth
requirements differed from those of non-transformed human fibroblasts. In
addition, I assayed these transformed cell lines for their level of mRNA
coding for known growth factors and for factors putatively involved in the

transduction of mitogenic signals.

4

Chapter 1 reviews background literature relevant to these questions.
Chapter 11 describes fundamental work which I carried out to modify a
serum-free medium in order to detect the effects of protein growth factors
on normal human fibroblasts. Chapter III consists of a manuscript to be
submitted to ancer Research which contains the bulk of my doctoral
research. J. Dillberger, a second author of the manuscript, performed the
microscopic evaluation of the tumors which arose after injection of some
of the transformed cell lines into athymic mice. The results show that
growth factor independence is a phenotype characteristic of each of the
transformed human fibroblast cell lines tested and that these same
transformed cell lines show overexpress messenger RNA corresponding to the
genes assayed. Increased expression of these genes could cause the

observed growth factor independence.

CHAPTER I. LITERATURE REVIEW

A. Multistep Nature of Carcinogenesis.

l. Epidemiological Evidence.

The development of human cancer is widely understood to be the result
of a multistep process (Doll, 1978; Knudson, 1977; Peto, 1977). Support
for this hypothesis comes from several different observations and areas
of investigation. For example, treatment of animals with carcinogens
typically results in tumor formation only after multiple exposures over
an extended period of time (Berenblum, 1940; Peraina, 1975). Analysis of
such data using mathematical models to predict cancer incidence as a
function of time, leads to various multistep models (Peto, 1977). When
used to analyze the data on human cancer incidence, which for most types
of cancer peaks in the fifth or sixth decade of life (Silverberg, 1984),
these same multistep models predict that 4—7 discrete steps (e.g.,
cellular changes) are required. If, however, one of the steps confers a
selective growth advantage to the affected cell, the number of postulated
steps may be as few as two (Peto, 1977). Such models assume that each
change in the cells is a stable, heritable alteration consistent with a
genotypic change, but these models do not provide further insight into the

nature of the changes necessary for malignant transformation.

2. Genetic Evidence.
As noted above, most human tumors are found in the fifth or sixth

decade of life. However, a few rare tumors occur in the neonate or very

6

young children (Putnam gt gl., 1978). The predisposition to these kinds
of tumors has been shown to be inherited. The identification and
characterization of genes associated with such hereditary cancers have
provided important supporting evidence for the multistep process of
carcinogenesis in humans. It is now recognized that these types of tumors
occur early in life because one of the changes required for tumorigenicity
is inherited.

Hereditary retinoblastoma is a malignant tumor of the retinal cells
of the infant, often occurring bilaterally. The disease is inherited in
an autosomal dominant manner (Sorsby, 1972). Chromosomal analysis of this
disease reveals that deletion of the short arm of chromosome 13 is
frequently associated with the occurrence of the disease (Knudson, 1985).
The gene whose loss or inactivation is associated with the development of
hereditary retinoblastoma has been identified (Friend gt gl., 1986; Lee
gt gl., 1987a) and the protein isolated (Lee gt gl., 1987b). This p110-
114 protein has recently been shown to form a complex with either the SV-
40 large T protein (Decaprio gt fl“ 1988) or the Ela protein of
adenovirus (Whyte gt _l., 1988), but the function of p110-114 remains
unknown. Both copies of this gene must be lost or inactivated for
retinoblastoma to occur (Cavenee gt g1., 1983; Dryja gt gl., 1986). If one
assumes that the loss of each copy of the retinoblastoma gene is an
independent event, at least two genomic changes are necessary for the jg
ytyg transformation of retinal cells. One might expect that the
inheritance of cellular changes which appear to transform retinal cells
may also result in the increased incidence of other types of cancer in

these affected individuals. In fact, affected children have an increased

7

incidence of osteosarcomas compared to children that do not have
hereditary retinoblastoma (Abramson gt _1. 1984). The incidence of
osteosarcomas among these hereditary retinoblastoma patients is low,
however, estimated at l-20%. Clearly the changes necessary for
transformation of retinal cells are not sufficient for the production of
osteosarcomas at a high frequency. The latter tumors may result from
additional changes.

Wilm’s tumor is another example of a form of cancer which occurs
early in the life of affected individuals, usually within the first
decade. The autosomal dominant heritable form is often associated with
other congenital malformations, such as aniridia, genitourinary
abnormalities and mental retardation (Pendergrass, 1976). It is
associated with a deletion of chromosome 11 (Yunis, 1983; Knudson, 1972;
Cavenee gt gl., 1983), but has not been characterized as extensively as
hereditary retinoblastoma. Wilm’s tumor, however, is generally considered
to result from inactivation of cancer suppressing genes similar to the
gene inactivation described above for hereditary retinoblastoma (Knudson,
1985). Studies by Koufos gt g1., (1984), Fearon gt gl., (1984) and Orkin
gt gl.,(l984) use restriction fragment length polymorphisms to demonstrate
that the specific deletion on chromosome 11, often seen in the hereditary
from of Wilm’s tumor, is associated with duplication of the remaining
locus leading to a homozygousity of the tumor cells. If the gene
associated with Wilm’s tumor were recessive, one might expect this
duplication to allow expression of the recessive gene leading to the

development of the tumor. Alternately, if the deleted sequences contained

a dominant suppressor gene, the loss of both alleles for this gene would

8
allow tumor formation. The gene responsible for Wilm’s tumor has not yet
been accomplished because familial cases are rare and because there is no
marker analogous to esterase D (Benedict gt g1., 1983), which aided in the

identification of the retinoblastoma gene.

3. Evidence from Dysplastic Histology of Premalignant Conditions in
Humans.

If one or more of the cellular changes necessary for full malignant
transformation of human cells in 1119 manifest themselves as phenotypic
changes, one would expect to find cells within the body which have some,
but not all, of the characteristics of malignant tumor cells. Indeed, the
appearance of a number of dysplastic cell populations in humans suggests
the existence of such preneoplastic states. These dysplastic cells share
some characteristics with malignant cells but are not able to form
malignant tumors. An example of such a dysplastic condition is cervical
intraepithelial neoplasia (CIN), also called "carcinoma in situ", has been
described by several authors (Christopherson, 1977; Ferency, 1982; Scully,
1981 and Buckley gt gl., 1982). Normal cervical epithelium consists of
orderly layers of cells arranged with the most differentiated cell types
occurring near the surface and less differentiated types located near the
basement membrane, an arrangement known as polarity. Microscopic
examination of normal epithelium reveals few nfitotic figures, with an
orderly, uniform cell morphology and polarity. This is in contrast to the
appearance of frank cervical carcinoma, which exhibits multiple mitotic
figures, total loss of polarity, invasion of the cells through the

basement membrane, bizarre, irregular cell morphology and atypical nuclei.

9

The microscopic appearance of "carcinoma in situ" is intermediate between
these two extremes. The cells do not invade the basement membrane, and do
not appear as disordered or as bizarre as cervical carcinoma cells, but
there is obvious loss of polarity and a predominance of less
differentiated cell types which are aneuploid. Such dysplasia is
classified as CIN grades I, II, or 111, depending upon the degree of
regular cellular morphology, loss of polarity and the number of mitotic
figures seen. Grade 111 is also known as "carcinoma in situ".

The incidence of carcinoma in situ peaks at age 25-35, whereas
cervical cancer’s peak incidence occurs about ten years later (Silverberg,
1984). This suggests that dysplastic cervical cells progress through
intermediate phenotypes prior to forming frank carcinoma. This conclusion
is supported by clinical studies showing that early identification and
removal of the dysplastic cells dramatically decreases the incidence of
cervical cancer compared to that found in subjects receiving no
intervention (Cristopherson gt at, 1976). These clinical and
epidemiologic observations strongly suggest that cervical intraepithelial
neoplasia represents a premalignant state, with cells comprising this
lesion possessing some, but not all, of the characteristics of frank
malignant cells.

Another example of preneoplastic populations of cells is the
dysplastic nevus commonly seen on the skin, which consists of aggregates
of melanocytes which have a slightly rounded morphology compared to
normal melanocytes (Murphy and Mihm, 1983). These aggreagtes often extend
beyond the epidermis into the dermis. Most nevi are well circumscribed

with uniform pigment distribution, but on occasion, some are observed to

10
increase in size and become irregularly shaped and pigmented (Elder,
1980). The microscopic appearance of the cells comprising these latter
lesions is less regular, with an increase in mitotic figures. In
contrast, malignant melanoma, the fully malignant tumor produced by these
melanin-producing epithelial cells, consists of cells exhibiting altered
morphology, loss of polarity and a high mitotic index (Clark gt g1., 1969;
Mihm gt gl., 1971). As noted with CIN above, removal of dysplastic nevi
lowers the frequency of malignant carcinoma (Day gt gl., 1982), which is
consistent with the hypothesis that such nevi are composed of premalignant
cells. Not all malignant melanoma lesions develop from nevi, but the fact
that a significant proportion of them (20-40%) (Allen and Spitz, 1954)
arise from an existing nevus suggests that the disordered growth of cells

in a nevus is indicative of the preneoplastic state.

8. Experimental Investigation of Multistep Carcinogenesis using

Fibroblasts in Culture.

1. Studies Utilizing Rodent Fibroblasts.
Recent experimental data that support the hypothesis that
carcinogenesis is a multistep process have been obtained using rodent

fibroblasts to study the process of malignant transformation in vitro. The

 

use of fibroblasts in culture allows examination of large populations of
cells to analyze the differences between normal and tumor cells. These
systems allow selection of cells which have acquired a few, but not all
changes necessary for full malignant transformation. The aim of this kind

of approach is to elucidate intermediate genotypic and/or phenotypic

11

changes which contribute to the fully transformed state.

a. Spontaneous Transformation.

An example of such studies has been described by Pouyssegur and
associates using the Chinese hamster lung fibroblast line CCL39. This
fibroblast line is an infinite lifespan, near-diploid cell line which
cannot form colonies in soft agar and which requires a high concentration
of serum for growth on plastic (Van Obberghen-Schilling gt gl., 1983).
Subcutaneous injection of these cells into athymic mice occasionally
results in a tumor at the site of injection after 4-6 weeks (Perez-
Rodriguez gt g1., 1982). These investigators found that, unlike the
parent CCL39 cells, the cells derived from such tumors showed high cloning
efficiency in agar and grew well on plastic with a low concentration of
serum or without serum. To determine if cells which had acquired only a
single tumor-cell related phenotype could produce tumors, Perez-Rodriguez
and associates (1982) selected clones of CCL39 cells able to grow in a
serum-free medium lacking exogenously added protein growth factors (growth
factor independence). The cells that had acquired this phenotype exhibited
it as a heritable, stable trait as evidenced by its persistence during
continued passage under non-selective conditions. When cell populations
resulting from such growth factor independent isolates were injected into
athymic mice whose residual immunological response had been suppressed by
either cyclophosphamide or x-irradiation treatment, they formed nodules.
These nodules eventually regressed, but no such nodules appeared in mice
that did not receive such immunosuppression. In some instances,

progressively-growing tumors subsequently appeared (40-60 days later) at

12
the site of injection in both immunosuppressed and non—immunosuppressed
mice. When populations of cells derived from such late appearing tumors
were injected into mice, they formed progressively growing tumors with no
latency period. The investigators’ conclusion was that the acquisition
of the growth factor independent phenotype was not sufficient to cause
CCL39 cells to become malignant. However, these cells possess the ability
to form neoplastic growths, as shown by the appearance of regressing
nodular growths. Presumably, an additional change was required for
malignant transformation and this was acquired spontaneously by the cells
that made the non-regressing tumor: ‘The genotypic changes responsible for

these phenotypes are not known.

b. Carcinogen-Induced Transformation.

A stepwise carcinogen—induced acquisition of transformed phenotypes
by rodent cells in culture was observed by Barrett and Ts’o (1978) during
the investigation of the malignant transformation of Syrian hamster embryo
(SHE) cells. They observed that altered morphology, enhanced fibrinolytic
activity, and the ability to grow in soft agar, phenotypes frequently
associated with tumors of fibroblastic and epithelial origin (Weinstein
gt g1., 1976), arose independently and at different times after treatment
of 'the target cell population with the chemical carcinogen,
benzo(a)pyrene. These same characteristics were observed at very low
frequency in untreated Syrian hamster embryo cells (Barrett and Ts’o,
1978).

The experiments involved repeated subculture of treated cells in

order to determine the time of appearance of the various transformed

13

characteristics. Assays for morphological alteration, fibrinolytic
activity, soft agar growth, and tumorigenicity were performed at selected
culture passages after carcinogen treatment. Colonies resulting from early
passage cells displayed morphological alteration within one»week following
exposure of the cells to benzo(a)pyrene and, thereafter, the proportion
of such altered cells in the population remained steady, i.e., about 3%
of the colonies scored. Cells exhibiting enhanced fibrinolytic activity
could be detected 1-2 weeks after carcinogen treatment, at frequencies of
8.6% - 17% of the tested colonies. The ability of cells from the
population to form colonies in soft agar was not seen until the ninth
subculturing. By that time, the cells had undergone approximately 46
population doublings. At the tenth and all subsequent subcultures, the
frequency of cells able to form colonies in soft agar progressively
increased. By this time, greater than 90% of the cells were shown to have
morphologic alteration and enhanced fibrinolytic activity. Populations of
cells which displayed growth in soft agar formed tumors in 100% of the
animals injected. Cells which displayed only altered morphology were not
tumorigenic.

These results are consistent with the hypothesis that malignant
transformation of rodent fibroblasts in culture, a model for i_n v_ivg
carcinogenesis, is a multistep process. An alternate theory which could
explain these data is that early populations of SHE cells contained a few
malignantly transformed cells which were subsequently selected by the
extended passage in culture. However, if this were the case, one should
be able to detect such cells at early passage with an assay of sufficient

sensitivity. Barrett and Ts’o (1978) estimated that their soft agar assay

14
could detect as few as 1 anchorage independent cell per 105-106 cells
plated as evidenced by the average cloning efficiency of cells derived
from tumors experimentally produced by their treatments. Even with this
sensitivity, no agar' colonies were observed until passage 9, i.e.,
approximately 46 population doublings after treatment. This strongly
suggests that transformation did not occur in a single step at low
frequency, but rather populations of cells acquired multiple cellular
changes in a stepwise fashion which eventually resulted in a tumorigenic

cell.

c. Oncogene-Mediated Transformation.

The discovery of oncogenes as the transforming genes of
retroviruses (see Varmus, 1984 for review), and the realization that these
genes were captured by these viruses during infection of vertebrate cells,
has led to the identification of cellular genes causally involved in
tumorigenesis. As the techniques became available to routinely introduce
such cloned oncogenes into cells, investigators have been able to
determine what role they have in the malignant transformation of cells.
The results of such studies demonstrate that the expression of two
complementary oncogenes are ordinarily required for the transformation of
primary rodent fibroblasts in culture.

An example of one such study is that of Land gt g1. (1983), who
transfected primary rat embryo fibroblasts with the H-tgg oncogene
isolated from a cell line (EJ) derived from a human bladder carcinoma
(Santos gt gl., 1982). The resultant populations of transfectants

expressed the H-rgg gene, formed colonies with increased frequency in soft

l5

agar, and were morphologically altered. But these cells had a finite
lifespan and produced only cartilagenous nodules at the site of injection
into animals, indicating that the cells were not tumorigenic. The
investigators then co-transfected the H-ta_s and myg oncogenes and analyzed
these co-transfected populations of cells their ability to form a focus
on a confluent monolayer. Cells from such foci were isolated, grown to
large populations, and injected into athymic mice. Tumors formed at the
site of injection and grew to a diameter of 1 cm and then remained
stationary. Clearly, these cells had acquired many of the properties
necessary for tumor formation, but were not malignant as judged by their
inability to form progressively growing, invasive tumors. In a later
study, Land gt_gl. (1983b) reported that the functions of the myg oncogene
could be replaced by either the polyoma large T gene or the adenovirus Ela
gene, and that the H-tgg function could be replaced by either N-tgg or
polyoma middle T.

Thomasson and co-workers (1985), in a similar set of experiments,
transfected normal SHE cells with plasmids containing Harvey murine
sarcoma pro-viral DNA (containing the v-H-[gg gene) and/or the pro-viral
MC-29 DNA (the former carries the v-Harvey tgg gene, the latter the v-myg
gene). These investigators observed full malignant transformation of
cells expressing both genes, with tumors arising 3-5 weeks after
injection. Cells transfected with the plasmid containing the v-myg
sequences alone were not tumorigenic in nude mice, but were not further
characterized. SHE cells transfected only with the plasmid containing v-
H-tgg sequences were morphologically transformed, had a finite lifespan,

and were not tumorigenic. Clearly, increased expression of a ras oncogene

16
alone produced some, but not all, of the cellular changes characteristic
of a fully malignant cell. Full malignant transformation of SHE cells in
culture required at least two cellular changes represented by the
increased expression of a mutated H-tgg gene and the unregulated

expression of the my; gene.

2. Studies Utilizing Human Fibroblasts.

While it seems reasonable to assume that the mechanisms causing the
malignant transformation of human cells are the same as those operating
in rodent cells, there may well be some species specific differences. A
more appropriate model for studying the origins of human cancer is the
transformation of human cells in culture. Since it has been estimated that
85% of human tumors are carcinomas, i.e., tumors of epithelial origin

(Cairns, 1975), studying the lg vitro transformation of normal epithelial

 

cells would be of primary interest in human carcinogenesis. For example,
treatment of bronchial epithelial cells with carcinogenic agents present
in cigarette smoke might be a reasonable model to study the production of
bronchogenic carcinoma, but until recently, normal human epithelial cells
generally have proven difficult to maintain in culture (Harris, 1987).
Human fibroblasts, in contrast, are easily maintained in culture.
Therefore, most human cell transformation experiments have utilized
fibroblasts in experiments similar in design to rodent fibroblast cell

transformation experiments.

a. Spontaneous Transformation.

The malignant transformation of human fibroblasts in culture has

17
proven extraordinarily difficult compared to rodent fibroblast
transformation (McCormick and Maher, 1988). Normal diploid human
fibroblasts have a finite lifespan in culture (Hayflick, 1965), are
karyotypically stable, and rarely, if ever, transform spontaneously in
culture (Thompson and Holliday, 1975). Only three cell lines resulting
from the spontaneous transformation of human fibroblasts in culture have
been described (Azzarone and Pedulla 1976; Thielmann gt gl,, 1983 and
Mukherji gt g1. 1984). Recent analysis of these cell lines indicates that
VIP-F:T cells isolated by Mukherji and associates may represent a
”legitimate, spontaneous, malignantly transformed fibroblast cell line."
(McCormick and Maher, 1988). The cells isolated by Azzarone and Pedulla
(1976) were reported to form tumors after four weeks incubation in athymic
mouse host, and the tumors were classified as progressively growing
fibrosarcomas. However, these tumorigenic cells were subsequently lost
in a freezer accident and cannot be further characterized. The original
experiment was reported in 1976 and has never been reproduced. Thielmann
gt g1. (1983) reported the isolation of the malignantly transformed cell
line XP29MA-MAL from a parental cell line, XP29MA, derived from a male
xeroderma pigmentosum patient. However, the cell line XP29MA-MAL has been
shown by isozyme analysis and by restriction fragment length polymorphism
analysis using M13 bacteriophage DNA as probe, to differ markedly from the
parental cell line (McCormick and Maher, 1988), indicating that the line
is not derived from the putative parent line, XP29MA. Thus, the VIP—F:T
cell line may represent the only malignant human fibroblast cell line to
have arisen spontaneously in yittg. In contrast, spontaneous malignant

transformation of primary rodent cells of some species has been observed

18
in culture with a low but regularly detectable frequency (Van Obberghen-
Schilling, 1983; Barrett and Ts’o, 1978; and Kraemer gt g1., 1986). The
reason(s) for the observed difference in the rates of spontaneous
transformation of human fibroblasts compared to rodent fibroblasts is not

known.

b. Carcinogen-Induced Transformation.

The first report of the malignant transformation of diploid human
fibroblasts was by Kakunaga (1977). K0 cells, an adult fibroblast cell
line, were reported to give rise to foci after treatment with 4—
nitroquinoline-I-oxide (4-NQO) (H’ N-methyl-N’-nitro-N-nitrosoguanidine
(MNNG). Cell lines derived from the foci were designated HuT 1 through
15. However, HuT-Il, Hut—12 and Hut-14, representative members of the Hut
series of cell lines, have recently been shown not to be derived from KD
cells. In fact, analysis of Hut-11, Hut-12 and Hut-l4 cell lines reveals
HLA types, RFLP patterns and isozymes identical to cell line 8387, a
fibrosarcoma-derived human cell line (McCormick gt gl., 1988).

Cells derived from human fibrosarcomas display an infinite lifespan
in culture and form progressively growing, invasive fibroblastic tumors
in athymic mice (McCormick and Maher, 1988). Other reports of malignant
transformation of human fibroblasts (Min gt gl., 1980; Ming gt g1., 1986)
have appeared in the literature, but these investigators have not
demonstrated that the derived cells were malignant as evidenced by an
ability to form progressively-growing invasive tumors in an animal host.
Therefore, insufficient evidence exists to accept claim that these cells

have become malignantly transformed.

19

There are several reports of carcinogen-induced transformation of
human fibroblasts in culture, with transformation most often assayed by
an increased frequency of growth in soft agar (Freedman and Shin, 1977;
Milo and DiPaolo, 1978, Silinskas gt .gl. 1981, Wang ,gt ._1. 1986,
Biedermann and Landolph, 1987). The cells isolated from agar in these
cases had a finite lifespan, were morphologically similar to normal
fibroblasts, and either did not form tumors or formed small nodular
growths which later regressed when injected into athymic mice. These
cells were neoplastic, that is, they at times formed a new growth, but
were not malignantly transformed. DeMars and Jackson (1977) also observed
the appearance of such nodules at the site of injection of focus-derived
human cells. These cells had been obtained by a single treatment of
normal human fibroblasts with MNNG and the subsequent isolation of cells
from a morphologically altered focus on a uniform monolayer of confluent
cells.

Because of their limited lifespan, human fibroblasts derived from
such selection experiments cannot be used to identify the additional
cellular changes required to complete the malignant transformation of
these cells in culture. The limited lifespan of these cells makes further
treatment or biochemical analysis difficult, because such clonally-
isolated cells typically senesce before they can be expanded into the
large populations needed for many assays.

A series of human cell transformation experiments of particular
interest is that reported by Namba and associates (1978, 1980 and 1981).

They reported the isolation of 6 cell lines which were morphologically

altered and had acquired an infinite lifespan in culture. These cell lines

20

were generated by repeatedly treating the same population of fibroblasts
with either 4-NQO or x-rays. These lines proved not to be tumorigenic in
athymic mice, but the cells are morphologically altered, immortal, and
aneuploid, and at least one cell line exhibits growth factor independent
replication in a serum-free medium (Namba gt g1., 1984). Such
characteristics have been associated‘with malignant transformation of some
rodent cells in culture (VanObberghen-Schilling gt gl., 1983). The cell
lines isolated by Namba and associates after repeated carcinogen treatment
clearly exhibit many characteristics of tumorigenic cells in culture; and
the fact that they were detected only after multiple carcinogen exposures
carried out over an extended period of time, is support for the multistep
nature of malignant transformation.

At least one example exists where human cells which were clearly not
tumorigenic were treated with carcinogen and gave rise to fully malignant
cells. Rhim gt g1. (1975) reported the transformation of a non-
tumorigenic, immortal cell line derived from a human osteosarcoma.
Osteoblasts, like fibroblasts, are cells of mesenchymal origin. In this
experiment, immortal, aneuploid cells were treated with MNNG and then
repeatedly subcultured. Cell populations were periodically assayed and
first displayed the ability to form foci approximately 52-59 days after
carcinogen treatment. Cells from these foci were isolated and grown to
large populations. These focus-derived cells grew in soft agar and formed
progressively growing tumors in athymic mice. The implication of this
study is that some cell lines exist that have already acquired some of the
cellular changes necessary for production of a tumor and that the process

has been completed by carcinogen treatment and subsequent subculture as

21
reported by Rhim gt g1. (1975).

c. Oncogene-Mediated Transformation.

The earliest experiments involving the introduction of oncogenes
into cells were viral infection studies, where human fibroblasts were
infected with the oncogenic DNA virus, SV-40 (Shein and Enders, 1962).
Since that time, numerous groups have studied various aspects of SV-40
transformation and these have been extensively reviewed by Sack (1981).
As a sumary, SV—40 infected human fibroblasts form morphologically
altered foci, and cells recovered from these foci form anchorage
independent colonies. When injected into an appropriate animal host, cells
derived from such colonies occasionally give rise to a nodular growth. On
rare occasions, SV-40 transformed cells which have acquired an infinite
lifespan have been isolated from normally senescent populations of
infected cells (Sack, 1981). When characterized, these infinite lifespan
cells have been reported to grow in medium containing reduced levels of
calcium or serum (McKeehan and Ham, 1978).

Fry gt g1. (1986) reported the transfection of v-gig oncogene into
normal human fibroblasts. This gene is the viral homolog of the B-chain
of platelet derived growth factor (Devare gt gl., 1983). The cells were
selected by their ability to form foci on a monolayer. The cells had a
finite lifespan and were non-tumorigenic, but they demonstrated the
ability to proliferate in a serum-free medium more rapidly than control
cells and grow in soft agar. Infection of a non-human primate cell line
(marmoset) with simian sarcoma virus which carries the v-gig oncogene

yielded cell lines with similar characteristics (Robbins gt gl., 1985).

22

Hurlin gt g1. (1987) transfected normal diploid human fibroblsts with
a mutated H-rgg oncogene originally isolated from the T24 human bladder
carcinoma cell line (Santos ,gt,.g1., 1982) and cloned into a high
expression vector (Spandidos and Wilke, 1984). The transfectants
exhibited focus forming ability, anchorage independence and morphological
alteration, but had a finite lifespan and were non-tumorigenic.
Similarly, Wilson gt g1. (1987) introduced the N-ng oncogene cloned into
a high expression vector and obtained foci composed of morphologically
altered cells which had the ability to form colonies in soft agar. These
cells also had retained their finite lifespan and were non—tumorigenic.

There are several studies which describe the introduction of
oncogenes into human fibroblasts which have acquired an infinite lifespan
in culture. The results of such experiments, in contrast to those of
transformation studies in diploid human fibroblasts, show that tumorigenic
cells are produced. For example, O’Brien _e_t_ g1. (1986) infected Va2
cells, a human fibroblast cell line that has acquired an infinite lifespan
after SV-40 infection, with Kirsten murine sarcoma virus and selected
foci. Cells from such foci were grown to appropriate numbers and injected
into athymic mice. Static tumors composed of human cells occurred in all
animals. Recently, an immortal cell line designated MSU-1.l was isolated
in this laboratory following transfection of normal diploid fibroblasts
with the v-myg oncogene. These cells have a normal morphology, are near-
diploid, do not form foci, do not proliferate in the absence of
exogenously added growth factors, and are non-tumorigenic in athymic mice.
Transfection of this cell line with an H-ng oncogene expressed at a high

level resulted in tumorigenic cell lines exhibiting growth factor

23
independence in culture (Hurlin gt gl., 1988). Namba and associates also
reported the successful malignant transformation of an infinite lifespan
human cell line, KMST-6, by transfection with c-Ha-tgg oncogene (Namba gt
g1., 1986) or infection with MSV (Namba gt gl., 1988). The result of
these studies support the hypothesis that transformation of human
fibroblasts requires at least three steps, i.e., the acquisition of
immortality, a mutation of the [g5 gene, and its expression at a high
level. There is, as yet, no method for the reproducible transformation
of diploid fibroblastic (McCormick and Maher, 1988) or epithelial cells
(Chang, 1986) intoimalignant cells, but epithelial and fibroblastic tumors

occur in humans. This failure of workers to transform such cells 1g vitro

 

may reflect an inability to detect cells that have acquired the

intermediate phenotypic changes.
C. Properties Associated with the Transformed Phenotype.

An alternative approach to defining steps required for malignant
transformation of human fibroblasts is to study properties of tumor-
derived cells in an attempt to find phenotypes not expressed in non-
transformed cells of the same type. Ideally, knowledge of the phenotypes
could suggest mechanisms by which the changes involved lead to the
development of tumors 1g ytyg. The following section reviews reported
characteristics of tumor cells which have been commonly studied and are

used as indicators of la vitro transformation.

 

1. Tumorigenicity.

24

The tumorigenic phenotype is the ultimate indicator of full malignant
transformation. Experimental injection of cells into a genetically
identical host is, of course, the ideal assay for their tumorigenic
potential, but this is impossible when dealing with the human species. The
athymic mouse is considered the most suitable animal host in which to
assess the tumorigenic potential of a wide variety of human cells ig_ytyg
(Fogh, 1982). Inbred mice with this deficiency are genetically uniform
and relatively easy to breed and maintain. The congenital lack of the
thymus renders the animals unable to produce normal levels of T-
lymphocytes which function in the immune destruction of genetically
incompatible cells. Even so, Fogh (1982) has shown that a wide variety of
human tumors are not transplantable into athymic mice. Given this
restriction, negative results are less useful than positive results in
assessing the tumorigenic potential of cells. Notwithstanding, a number
of‘ the rodent transformation systems using fibroblasts have ,yielded
progressively-growing tumors in athymic mice. In addition, cell lines
derived from several human fibrosarcomas and the spontaneously transformed
line VIP-F:T have been shown to produce tumors in athymic mice at a high
frequency (Paterson gt gl., 1987; Fogh, personal communication; Azzaronne
gt g1., 1976; Mukherji gt l., 1984), which indicates that tumorigenicity
in the athymic mouse is a reasonable indicator of the malignant potential
of both rodent and human fibroblasts. Although the formation of a
malignant tumor does not offer insight into the cellular changes which
have occurred in the process of making the cells tumorigenic, the fact
that tumor cells grow in the body when normal cells do not, strongly

suggests that the abnormal production of, or response to, growth factors

25

may be involved in the production of tumors.

2. Anchorage Independence.

A variety of cells derived from human and rodent tumors have the
ability to grow in semisolid medium (Stevens gt g1., 1988). Freedman and
Shin, 1977) using SV-40 transformed rodent fibroblasts, demonstrated that
this phenotype‘was very highly correlated with tumor formation. For these
reasons, the ability of human fibroblasts to grow suspended in soft agar
or methyl cellulose has been used as an endpoint of transformation by many
workers, including Freedman and Shin (1974), Milo and DiPaolo (1977),
Sutherland gt gfl, (1980), Silinskas gt g1. (1981) and others. A very
important question in the evaluation of this phenotype is whether human
fibroblasts which are transformed to anchorage independence are
malignantly transformed. A recent review by McCormick and Maher (1988)
summarized the findings of several workers who obtained anchorage
independent colonies from carcinogen-treated human fibroblasts, grew these
colonies to large numbers and injected them into athymic mice. Tumors are
sometimes formed at the site of injection but these tumors are non-
invasive, have a variety of histological types, do not grow progressively
or invade, and usually regress. The cell lines derived from these tumors
have a finite lifespan. This is in contrast to cells derived from human
fibrosarcomas which grow progressively, invade, and have an infinite
lifespan in culture (McCormick and Maher, 1988).

The changes in fibroblasts that are responsible for increased ability
to form colonies in soft agar are unknown at present, but are thought to

be the result of a mutation. At least two observations are consistent with

26

this hypothesis. First, the frequency of anchorage independence induced
by carcinogen treatment is higher in cells that cannot repair damage in
their DNA than in repair-proficient cells, indicating that events which
cause anchorage independent growth, occur as a result of potentially
mutagenic DNAldamage. Secondly, this phenotype is stably inherited similar
to a germ-line mutation (Mahgr__et al., 1982). 988).mechanism by
which cells acquire the ability to grow in soft agar is by a mutation in
a [1; gene. Hurlin gt g1. (1987) and Wilson gt g1. (1987) have shown that
increased expression of either mutated H or N-La_§ oncogenes following
their transfection into diploid human fibroblasts, caused an increase in
the frequency of growth in soft agar compared to that of control-
transfected cells. Evidence consistent with these studies comes from the
recent reports by Stevens gt gt. (1988) who recovered mutant H-tgg genes
from several anchorage-independent cell populations obtained by treating
normal diploid human fibroblasts with the mutagen benzo(a]pryene i anti-
7,8-dihydrodiol 9,10-epoxide.

Recent evidence suggests that one method to increase the frequency
of colony-formation in soft agar is to add specific protein growth
factors. Palmer gt g1. (1988) has shown that platelet-derived growth
factor or basic fibroblast growth factor is able to stimulate growth in
soft agar of normal diploid human fibroblasts. Similarly, Peehl and
Stanbridge (1981) have shown transient anchorage independent growth of
normal human fibroblasts in the presence of high concentrations of serum
and hydrocortisone. This predicts that a cell which is capable of
synthesizing and secreting such growth stimulatory proteins would also

show an increased ability to grow in soft agar.

27

3. Focus Formation.

A dense multi-layered, criss-crossed array of fibroblastic cells on
a monolayer of fibroblastic cells has become known as a "focus".
Reznikoff gt g1. (1973) studied the malignant transformation of rodent
fibroblasts utilizing focus formation as a criterion of transformation.
The foci formed after carcinogen treatment of C3H/10T1/2 mouse fibroblasts
were graded from Type I to Type 111 depending on the density of staining
and abnormality of morphology. Cells from the most advanced foci (Grade
111) were shown to form tumors with high frequency when injected into

appropriate hosts (Reznikoff gt l., 1973). This characteristic is

commonly used to identify transformed cells after experimental treatments
with carcinogens. The first description of this phenotype by Rubin and
Temin was as a "loss of contact inhibition” (Rubin, 1988). More recent
evidence suggests that focus formation may be related to increased
production of, or response to, growth promoting proteins. The v-s_i_s;
oncogene is the transforming gene of the simian sarcoma virus and is
homologous to the B-chain of platelet-derived growth factor (PDGF)
(Johnsson gt gl., 1984). PDGF is a a.nfitogen for human fibrobasts in
culture (Westermark and Heldin, 1985; Clemmons and Van Wyk, 1981). As
mentioned above, when Fry gt, _1. (1986) transfected diploid human
fibroblasts with the v-gtg oncogene, they obtained cells which formed
distinct foci. Cells resulting from such foci produce increased levels of
v-stg RNA and were able to grow in a serum-free medium in the absence of

exogenously added growth factors, while normal fibroblasts were not (Fry

gt g1., 1986). This is what would be expected from a cell which produces

28
its own growth factor and suggests that focus formation, like the ability
to grow in soft agar, is the result of the expression of protein growth

factors.

4. Growth in Medium Containing Reduced Calcium.

Parsons (1978) described the propagation of cells derived from human
tumors in medium containing reduced levels of calcium. Normal cells in
this same medium did not replicate. Fibroblasts are routinely cultured in
medium containing 1.0-2.0 mM calcium. McKeehan and McKeehan (1979) found
that an infinite lifespan cell line resulting from SV—40 infection of
normal human fibroblasts had a greatly reduced requirement for calcium
compared to the parental cells. Boynton gt_gl., (1977) studied the calcium
requirements of normal, partially transformed and malignantly transformed
rodent fibroblasts, and noted an inverse relationship between degree of
transformation and calcium requirement for rapid growth in the presence
of serum.

Calcium alone in the form of hydroxyappatite was shown to cause the
replication of normal human fibroblasts in culture by Cheung gt _a_l_.
(1986), who showed that normal production of somatomedin C was required
for this effect. Recently, Praeger gt g1. (1986) utilized a serum-free
medium to test the effect of calcium in the medium on the rate of
replication of diploid human fibroblasts in culture. They showed that
calcium alone at a concentration of 2.0 mM is able to stimulate the growth
of normal human fibroblasts in the absence of exogenously added protein
growth factors.

Since calcium acts like a growth factor, one might expect that a

29

protein growth factor could substitute for calcium in causing replication
of cells in culture. Just such results have been reported by McKeehan and
McKeehan (1979). They concluded that epidermal growth factor (EGF)
"replaced" the extracellular calcium required for replication of a human
fibroblast cell line. Earlier studies by McKeehan and Ham (1978) showed
that human fibroblasts in medium containing reduced levels of calcium, but
given additional growth factors in the form of serum, replicated at a rate
comparable to that rate observed with lower concentrations of serum but
at a normal concentration of calcium. These studies support the
hypothesis that calcium can act like a growth factor. However, the
mechanism by which calcium exerts these mitogenic effects is unknown.
Boynton gt g1. (1983) has reported that malignantly transformed rat liver
cells, which are able to proliferate in medium containing calcium levels
too low to support the growth of non-transformed cells, have increased
levels of protein kinase C activity compared to non-transformed cells.
It is not known, however, if this increase in protein kinase C activity
is a finding common to a variety of tumor-derived cell lines grown in such
a medium.

One important implication of the hypothesis that calcium acts like
a growth factor is that any cell capable of producing its own growth
factor could replicate in media which have reduced concentrations of

calcium.

5. Growth Factor Independence.
The growth of cells in culture has been shown to be dependent on a

variety of mitogenic factors which are typically provided by serum in the

30

culture medium. Cells derived from a variety of human tumors frequently
display the ability to replicate in medium containing concentrations of
serum too low for normal human cells to replicate (Holley, 1975; Tubo and
Rhinewald, 1988; van Zoelen gt gl., 1985), indicating that cells which
comprise these tumors either synthesize their own growth factors or are
abnormally responsive to low concentrations of mitogens. Kaplan gt g1.
(1982) has shown that Kirsten murine sarcoma virus-infected rodent
fibroblasts replicate in medium containing low concentrations of serum.
Van Obberghen-Schilling and associates (1983) studied the stepwise
transformation of Chinese hamster fibroblasts and noted a loss of growth
factor requirement as cells became malignantly transformed. Zhan and
Goldfarb (1986) have shown that transfection of NIH 3T3 cells with some
oncogenes is accompanied by the loss of growth factor requirement. They
used a serum free medium to show that transfection of non-tumorigenic NIH
3T3 mouse cells with the tgg, f , gtg, but not :9; oncogenes, caused
cells which expressed these genes to grow‘without exogenously added growth
factors. These genes are the transforming sequences of various oncogenic
retroviruses. These results are typical of transformed rodent fibroblasts
which frequently display diminished or absent growth factor requirements
compared to non-transformed cells (Liboi gt g1., 1986).

Similar results have been reported for transformed human fibroblasts.
McKeehan and Ham (1978) have shown that infinite lifespan SV-40 infected
human fibroblasts grow rapidly in medium containing concentrations of
serum too low to cause growth of normal human fibroblasts. Similarly,
Namba gt g1. (1984) showed that an immortal human fibroblast cell line,

CT-l, which he isolated after multiple carcinogen treatments, was able to

31
grow in a serum-free medium'without exogenously added growth factors. Most
recently, members of our laboratory have reported that human fibroblasts
transfected with gt; are capable of growth in serum free medium without
added growth factors (Fry gt g1. 1987). Similarly, cells which became
immortal after transfection with the my; gene (Hurlin gt g1. 1988) also
display'growth factor independence. These findings are consistent with the
reported loss of growth factor dependence which accompanies the

transformation of rodent fibroblasts.

6. Infinite Lifespan in Culture.

Cells isolated from normal human tissues or primary rodent cultures
replicate for a finite number of population doublings before the cells
senesce (Hayflick, 1965; Thompson and Holliday, 1975). Tumor cells often
grow continuously in culture without senescing and are said to have an
infinite lifespan. Cells derived from human fibrosarcomas have infinite
lifespans (McCormick and Maher, 1988) and human fibroblast-derived cell
lines with an infinite lifespan have been isolated on rare occasions after
carcinogen treatment (Namba, 1977), SV-40 infection (McKeehan and Ham,
1978), or v-myg transfection (Hurlin gt gl., 1988). The relationship of
the acquisition of infinite lifespan to the original treatment is obscure
in these experiments since each protocol involved extended periods of
culturing during which multiple changes occurred in the cells as evidenced
by the finding of aneuploidy, morphological alteration and decreased
growth factor requirements of the resultant cell lines.

The role of the infinite lifespan phenotype in malignant

transformation is not known, but this phenotype is probably necessary for

32

malignant transformation both 1g ytyg and in yittg (Newbold and Overell,
1982, Land gt gl., 1983, Ruley gt g1., 1983, McCormick and Maher, 1988).
Lack of this necessary infinite lifespan would eXplain the inability to
routinely obtain malignant transformation of diploid human fibroblasts or
primary culture rodent fibroblasts.

Dissecting the process by which diploid human cells acquire an
infinite lifespan will be important to ultimately understanding
how cancer is produced in humans. However, the lack of a reproducible
method to obtain such human cells in a defined, stepwise manner makes
direct examination of this phenotype difficult at best in cultured human

cells.

7. Other Correlates of the Transformed Phenotype.

A number of other phenotypes have also been associated with
transformation. Some of these properties include increased fibrinolytic
capability (Barrett and Ts’o, 1978), actin disorganization (Pollack gt
g1., 1975), morphological changes (Sanford ,gt,._1., 1974), decreased
fibronectin synthesis (Yamada and Olden, 1978), decreased cAMP levels
(Pastan and Willingham, 1978), altered nutrient requirements (Martin and
Stein, 1976), and ganglioside alterations (Hakamori, 1975). These
phenotypes have been inconsistently associated with transformation of many
different fibroblast lines (Vasiliev and Gelfand, 1981), consequently;
their function as a causative cellular change must be questioned.

A common theme in the analysis of the various transformed cellular
characteristics is the involvement of growth factors or growth factor

signalling pathways in the production of that transformed phenotype. Since

33
cancer is clearly a disease of unregulated growth, identifying those
cellular changes which affect the production of, or the response to,
growth factors is critical to dissect the steps in the process of

malignant transformation.

D. Growth Requirements of Non-Transformed Fibroblasts in Culture.
Although a number of changes in the cell are likely to occur in the
process of malignant transformation, only a few are likely to be
causative. Of the phenotypes reviewed, anchorage independence, growth
factor independence, and focus formation are most often used as indicators
of transformation. Since these phenotypes are indicative of the same
process, cells selected fer these phenotypes should share some common
cellular alterations which are likely to be essential in the process.
Acquisition of these phenotypes has been induced by experimental
treatments which result in the growth of fibroblasts under conditions in
which untreated cells do not proliferate. These treatments include:
addition of exogenous growth factors (Palmerggtug1., 1988), the expression
of exogenous oncogenes coding for growth factors (Fry gt gl., 1986) or
the overexpression of transfected 3g; genes (Hurlin gt _l., 1987; Wilson
fl g1., 19878; Stevens et g_l_., 1988). Clearly these treatments have
allowed the transformed cells to circumvent the pathways which control

normal cell replication.

1. Rodent Fibroblasts.
Much of the knowledge about growth factor requirements of mammalian

cells in culture has been obtained from investigation of the requirements

34
of rodent cells in culture. The most widely studied system describes
growth factor requirements of quiescent Balb/c-3T3 cells. Prior to the
identification and purification of protein growth factors, growth of cells
in culture could be maintained by whole serum added to nutrient medium at
concentrations from 10-20% (Eagle, 1955). In addition to substances which
mediate attachment and supply essential lipids and perhaps other
nutrients, serum contains a growth factor released by platelets during
clot formation (Ross gt gl., 1974; Busch gt gl., 1976; Hayashi and Sato,
1976). This observation led to the identification, isolation and eventual
purification of platelet derived growth factor (reviewed later), which has
been shown to be a major mitogen in serum, responsible for the growth of
many cell types of mesodermal origin in culture. (Ross and Vogel, 1978;
Scher gt g1., 1978). Pledger and associates (1977) investigated the role
of platelet-derived growth factor in the stimulation of quiescent Balb/c-
3T3 fibroblasts. In a series of experiments, they found that PDGF allowed
the cells to enter the cell cycle, but alone it was not sufficient to
stimulate DNA synthesis. The completion of the traverse of 3T3 cells from
Go/Glphase into S phase has been shown to require either platelet-poor

plasma (Pledger et l., 1977), or the sequential addition of EGF and

Somatomedin C (Stiles et al., 1979) to cells rendered competent by PDGF.
These observations formed the basis of a two step model of the
prereplicative phase of Balb/c-3T3 cells stimulated to synthesize DNA by
growth factors. In this model, PDGF and EGF have separate roles. PDGF
represents a group of molecules termed "competence" factors, which allow
cells to be acted upon by "progression” factors such as EGF and insulin.

The complementary action of growth factors in these classes cause the cell

35
to traverse Go/Glphase of the cell cycle into S phase. Subsequent studies
by Singh gt g1. (1983) utilized PDGF neutralizing antibody to show that
exposure to competence factors, i.e., PDGF, for as little as 5 minutes,
was sufficient to render quiescent 3T3 cells "competent"; whereas, EGF
was required continuously for the traverse of cells from 60/61 into S
phase. Somatomedin C or alternately insulin, was shown to be needed for
cells to traverse the portion of the cell cycle immediately preceeding S
(Leof'gt_gl. 1983). This has been the prevailing model used to explain the

growth requirements for non-transformed rodent fibroblasts in culture.

2. Human Fibroblasts.

Similar studies have been conducted with human fibroblasts in
attempts to define the growth requirements of normal human fibroblasts.
The results differ from those reported for the rodent system.

Balk (1971) was the first to demonstrate that human glial cells had
an obligatory requirement for serum and remained stationary in plasma.
However, workers using human fetal fibroblasts noted that a high
percentage of cells entered the cell cycle after exposure to plasma alone
(Bright and Gaffney, 1982). Similarly, Scher gt l. (1978) noted that
embryonic fibroblasts maintained in a medium containing 1.8imM calcium and
plasma were able to divide. The calcium concentration in the medium needed
to be lowered to 0.2 mM before the cells would stop cycling. At this
reduced calcium level, PDGF alone, could drive replication. More recently,
workers have investigated the growth factor requirements of neonatal and
adult fibroblasts using serum-free media (Westermark and Heldin, 1985;

Praeger and Cristofalo, 1987).

36

Westermark gt g1. (1985) studied the ability of EGF or PDGF to
stimulate DNA replication of quiescent human fibroblasts in a series of
experiments similar to those described for Balb/c-3T3 mouse fibroblasts.
In contrast to the findings of Stiles gt g1. (1979), however, Westermark
and Heldin showed that either EGF or PDGF alone was sufficient to cause
the quiescent cells to synthesize DNA. Insulin was not required to
initiate DNA synthesis under these conditions and it is generally thought
that the ability of normal human fibroblasts to synthesize IGF-l, also
called somatomedin C, obviates the need for this growth factor which
Balb/c 3T3 fibroblasts are unable to produce (Clemmons gt gl., 1983). It
is interesting to note that Westermark and Heldin reduced the level of
calcium to 0.5 mM in order to detect the mitogenic effect of either PDGF
or EGF, suggesting that calcium itself might be mitogenic, a finding later
reported by Cristofalo and associates. Praeger and Cristofalo (1987)
showed that WI-38 fibroblasts, a diploid human cell line derived from
embryonic lung, replicate in a serum-free medium in the absence of
exogenously added growth factors, if the concentration of calcium in the
medium is at least 2.0 mM. This replication of normal human fibroblasts
in calcium containg medium lacking exogenously added growth factors, is
not well understood at this time. Human fibroblasts maintained in such a
medium have been shown to have increased metabolism of ADP-ribose (Duncan
gt g1., 1988), but is not clear what role this plays in mitogenic effect
of calcium on human fibroblasts in culture. It is clear, however, that
human fibroblasts maintained in medium containing reduced levels of
calcium require only a single exogenously added protein growth factor such

as PDGF or EGF to cause them to replicate.

37

E. Mechanisms of Growth Factor Independence.

It is clear that rodent cell lines, such as Balb-c/3T3, require
different exogenously added growth factors for growth in culture than
diploid human fibroblasts do. However, transformed cells of either type
frequently lose some or all of these requirements and become growth factor
independent to some degree. The following review discusses some of the
mechanisms by which cells may acquire the growth factor independent

phenotype.

l. Autocrine Stimulation of Cell Growth.

The autocrine hypothesis has frequently been invoked to explain the
fact that tumor cells grown in the body under conditions where normal
cells do not (Sporn and Todaro, 1980). Briefly, the hypothesis is that
tumor cells which produce their own growth factors can stimulate their own
division and obtain a selective growth advantage. Such autocrine systems
have been identified in certain normal as well as transformed cells (Sporn
and Roberts, 1985; Heldin gt gl., 1986), and several oncogenes have been
implicated in these loops either by coding for a growth factor for which
the cell has an intact, functional receptor or by encoding a protein which
regulates the expression of such a growth factor. Some of these known

autocrine loops are discussed below.

a. B-Chain of Platelet Derived Growth Factor (PDGF-B).

PDGF has been identified as a major mitogenic growth factor in serum

38

and has been purified from platelets by several investigators (Antoniades
gt 31., 1981; Deuel gt gl., 1981; Heldin gt gl., 1981). This 30 kd
glycoprotein is a heterodimer in humans, consisting of an A and a 8 chain
(Johnsson et gl., 1984). These chains share approximately 60% homology
(Betsholtz gt gl., 1986; Johnsson gt 1., 1982). The v-_s_i_s oncogene
(Devare gt g1. 1983) and the transforming gene of Parodi-Irgens feline
sarcoma virus (Besmer gt gl., 1983) are derived from the same B-chain PDGF
proto-oncogene (Doolittle gt gl., 1983). In v-gtg-transformed cells, the
expressed protein is a 24 kd dimer which is homologous to a B-chain
homodimer (Robbins gt g1. 1983), and which is glycosylated to yield p28Sis
(Ross gt gl., 1986). Human cells transfected with the v-gtg oncogene form
foci and show an increased rate of growth in serum-free medium compared
to control cells (Fry gt g1., 1986). This transforming protein acts at the
PDGF receptor as evidenced by the use of either PDGF antibodies (Johnsson
gt g1., 1985) or suramin (Betsholz gt g1., 1986), to decrease growth and
revert the transformed morphology of transformed SSV-transformed
fibroblasts. The c-gtg gene is developmentally regulated and found to be
expressed in the developing trophoblasts during placenta formation in
humans (Goustin gt g1., 1985). A number of transformed rodent cells have
been shown to produce PDGF like molecules (Bowen-Pope gt g1., 1984). The
c-gtg gene has been shown to be expressed in many human tumor cell lines
including fibrosarcoma-derived cell lines 8387 (Eva gt gl., 1982) and HT
1080 (Pantazis gt gl., 1985), osteosarcoma-derived lines (Betsholtz gt
g1., 1984), neuroblastoma-derived cell lines (van Zoelen gt g1., 1985) and
other human tumor-derived cell lines (Bowen-Pope gt gl., 1984).

The events following PDGF binding, which transduce the mitogenic

39
signal are not completely understood. The first event following PDGF
binding is the activation of the receptor kinase (Ek and Heldin, 1982;
Frackelton gt g1., 1984). After this, a number of cellular changes have
been shown, including increased phosphotidyl inositol turnover (Berridge,
1984), activation of phospholipase A2 (Shier, 1980), an increase in
intracellular calcium (Ives and Daniel, 1987), shape change (Bockus and
Stiles, 1984) and induction of myg and fig; cellular protooncogenes (Kelley
gt gl., 1983; Greenburg and Ziff, 1984). All these events precede cell
replication stimulated by PDGF, but have not been completely defined. It
is not known if all of these same events occur in a human fibroblasts as

well.

b. A-Chain of Platelet Derived Growth Factor (PDGF-A).

The A-chain of PDGF has been less extensively studied than the 8-
chain of PDGF. Human PDGF has been shown to be a heterodimer of an A- and
a B-chain (Johnsson gt g1., I984). The A-chain mRNA has been shown to be
expressed in human tumors independently of the B-chain (Versnel gt g1.,
1988), forming homodimers which share some biological activities with
PDGF. Recent evidence suggests that the A-chain of PDGF is involved in an
autocrine feedback loop in normal human fibroblasts stimulated by either
PDGF or EGF (Paulsson gt g1., 1987). Quiescent cells treated with either
EGF or PDGF transiently express A-chain mRNA. A-chain homodimers have also
been shown to activate the tyrosine kinase activity of the PDGF receptor,
and stimulate growth in Chinese hamster ovary (CHO) cells transfected with
the cDNA coding for the PDGF receptor. (CHO cells normally do not have

PDGF receptors.) This activity is comparable to either the A,B heterodimer

40
or the BB homodimer (Escobedo gt g1., 1988). Transfection of the A—chain
into NIH 3T3 cells has a transforming effect similar to the transfection
of PDGF-B gene (Beckman gt g1., 1988), but is much less efficient. Beckman
gt g1. (1988) have shown that 3T3 cells transfected with the PDGF A-chain
gene under the control of an inducible metallothionein promoter, secrete
more PDGF-like activity into the medium than 3T3 cells transfected with
the PDGF-B construct. This suggests that a subtle difference in the range
of action of the two kinds of molecules. Another difference in the action
of PDGF-B gene compared to the A-chain is suggested by experiments by Hart
gt g1. (1988) showing two classes of PDGF receptors in cells. Type "A"
receptors bind all three PDGF dimers (AA,AB and 88); whereas, type "B"
receptors bind only B-chain containing dimers. It is not known whether
these two classes of receptors mediate different cellular responses to the

different forms of PDGF.

c. Transforming Growth Factor Alpha (TGF-a).

Transforming growth factor alpha was first identified in the
conditioned medium of Kirsten murine sarcoma virus-infected cells and
designated sarcoma growth factor. This growth factor was shown to
stimulate the growth of NRK rat cells in soft agar, a functional assay
used to identify transforming growth factors (Todaro and DeLarco, 1980).
Subsequently, this sarcoma growth factor was found to be a mixture of both
TGF-a and TGF-b (Anzano gt _l., 1983). Human TGF-a is a 50 amino acid
polypeptide with a 40% sequence homology'with EGF (Marquart gt gl., 1983).
The mitogenic effect of TGF-a is mediated through the EGF receptor
(Carpenter _et _al., 1983; Massague, 1983; Reynolds gt _l., 1981) and

41

binding of this growth factor to EGF receptors activates the receptor-
associated tyrosine kinase (Reynolds gt gl., 1981). TGF-a has been shown
to be a potent angiogenic factor (Schreiber gt g1 ., 1986) and may
stimulate the neovascularization seen during wound healing and observed
in some tumors. It is known that TGF-a is expressed in rat embryos during
early fetal development (Twardzick gt gl., 1982; Matrisan gt _1., 1982)
and may also function in human embryogenesis.

This growth factor is produced by cells transformed by retroviruses
(Todaro and DeLarco, 1978) and by cell lines derived from human tumors
(Nickell gt g1., 1983). Expression of TGF-a has been demonstrated in 11

out of 16 human glioma or sarcoma cell lines investigated (Heldin t al.,

1987), and a variety of tumors cell lines of epithelial and mesothelial
origin (Derynk gt gl., 1987; Coffey gt gl., 1987). At least two studies
have shown that TGF-a is produced it 2119 by human tumors. Derynk gt g1.,
(1987) showed increased expression of TGF-a mRNA in a surgical specimen
of melanoma. Ellis gt g1. (1987). found that the inappropriate secretion
of TGF-a by tumor cells appeared to cause the paraneoplastic syndromes of
Lesser-Trelat, ancanthosis nigricans and multiple acrochordons,
dermatologic conditions characterized by abnormal growth of skin. In that
study, serum and urine from a patient having a large malignant melanoma
and the above syndromes showed elevated levels of TGF-a. These levels
decreased to normal after surgical removal of the malignancy, and the
dermatologic lesions regressed. This suggests that the tumoral production
of TGF-a caused the aberrant growth syndromes in addition to stimulating

the growth of the tumor itself. An earlier report that is consistent with

the hypothesis that TGF-a is produced by tumors, was described by Kim gt

42
gl. (1985). They reported the partial isolation of a transforming growth
factor from the urine of melanoma patients that competes with EGF in
receptor binding. This factor was probably TGF-a and the report is
consistent with the studies of Derynk gt g1. (1987) and Ellis gt g1.
(1987).

The wide distribution of TGF-a expression in human tumor cells
suggests that inappropriate expression of this growth factor contributes
to the malignant state of a variety of human cell types, presumably by
stimulating neoplastic growth. This hypothesis is supported by the fact
that introduction of TGF-a cDNA into RAT-1 fibroblasts, a rat cell line
which has an infinite lifespan in culture but is non-tumorigenic, resulted

in autocrine growth stimulation jg vitro and weak tumor forming potential

 

in ijg (Rosenthal gt gl., 1986). A comparable study has not been reported

for human cells.

d. Epidermal Growth Factor (EGF).

EGF was first isolated from the submaxillary gland of mice (Cohen,
1962) as the substance which promoted early incisor eruption and eye
opening in mouse embryos. Human epidermal growth factor, also known as
urogastrone, (Gregory, 1975) is a 53 amino acid polypeptide containing
three disulfide bridges (Taylor gt g1., 1972). The receptor for EGF is a
185 kD transmembrane protein with tyrosine kinase activity which is
stimulated by EGF binding (Ushiro and Cohen, 1980). Like other growth
factors, EGF has been shown to cause increases in cytoplasmic calcium,
phosphotidyl inositol turnover (Hepler gt _a_l_., 1987), and early gene

transcription (Muller gt gl., 1984), which are some of the mechanisms

43

which have been suggested to be involved in the transduction of the
mitogenic signal of various peptide growth factors. The v-erb-B oncogene
(transforming gene of the avian erythroblastosis virus) has been shown to
be homologous to the EGF receptor. v-ggg;t is a truncated molecule lacking
the extracellular binding domain (Downward gt gl., 1984) while retaining
tyrosine kinase activity (Gilmore gt g1., 1985). Although there is no
direct proof, it is widely proposed that this oncogene functions as a
constitutively activated receptor (Downward gt gl., 1984).

EGF has been shown to be mitogenic for a variety of epithelial and
fibroblastic cells in culture (Cohen, 1983). In spite of the mitogenic
action on a number of cell types, epidermal growth factor itself has
rarely been shown to be produced by tumor cells. For example, one
submaxillary tumor line has been shown to produce EGF and has been
suggested to grow in an autocrine fashion (Sato gt g1., 1985). HuT cells,
which have recently been shown to be the fibrosarcoma-derived cell line
8387 (McCormick and Maher, 1988), have been reported to produce a large
peptide which is related to EGF. This molecule has not been further
characterized (Burbeck gt gl., 1984). Normal human fibroblasts in
culture, however, have been reported to produce small amounts of EGF
(Kurobe gt g1., 1985).

In addition to its mitogenic actions described above, EGF has been
shown to inhibit gastric acid production in the stomach (Gregory, 1975)
and stimulate uterine contraction (Gardner gt g1., 1987). This indicates

that this peptide has a number of different actions on target tissues.

e. Transforming Growth Factor Beta (TGF-b).

44

This transforming growth factor was first discovered not as a
transforming growth factor, but as an inhibitory growth factor (Holley gt
_a_l_., 1980). It was subsequently shown to be homologous to a peptide
secreted by transformed cells (Tucker gt g1., 1984), which was capable of
stimulating anchorage independent growth of rat NRK cells (Roberts gt g1.,
1982) and AKR-ZB mouse cells (Moses gt _1., 1981). The molecule is a 25
kd homodimer with disulfide bridges (Derynk gt g1., 1984) that has been
detected in a wide variety of normal cells (Roberts gt gl., 1981; Frolik
gt 31., 1983) and tumor cells (Derynk gt g1. 1985; Nickell gt l., 1983).
TGF-b exerts its effects through binding to its own receptor (Frolik gt
gl., 1984; Massague and Like, 1985). Although the factor has been shown
to be mitogenic for a variety of rodent fibroblasts in culture (Roberts
gt gl., 1985; Massague, 1984; Shipley gt fl., 1985), it is growth
inhibitory for human fibroblasts in culture (Paulsson gt g1., 1987). In
the case of AKR-ZB cells, the growth stimulatory effect of TGF-b is
indirect, and reflects its ability to induce the expression of c-gtg which
functions in an autocrine fashion (Leof g _l_., 1986). TGF-b has been
shown to be produced by the fibrosarcoma-derived cell line HT 1080 (Derynk
gt gl., 1985) but the effect of this factor on HT 1080 cell are not known.
It seems unlikely that this factor would cause growth in the fibrosarcoma
cells, given that its action in normal human fibroblasts is growth
inhibitory. The occurrence of TGF-b in many different tumor cell lines
suggests that this protein may play some role in tumor formation or

maintenance, but this role has not been identified.

f. Fibroblast Growth Factors (FGF’s).

45

Acidic and basic FGF are related growth factors and were originally
isolated from bovine brain (Thomas gt g1., 1984; Bohlen gt g1., 1984). Two
members of this growth factor family have been characterized in some
detail, acidic and basic FGF. Additionally, a recent report suggests
another related growth factor, k-FGF, which also may represent a
fibroblast growth factor (Delli-Bovi gt g1.,1988). The amino acid
sequences for acidic and basic FGF are 55% identical (Thomas, 1987). FGF’s
are mitogenic for fibroblasts (Gospodarowicz gt fl“ 1976), vascular
endothelial cells (Gimenez-Gallego gt gt, 1986), and rat neuroblasts
(Gensburger gt 31.,1987). These growth factors have been implicated in
embryogenesis because of reports of FGF-like molecules which have been
detected in developing chick brain (Risau, 1986), and differentiating
mouse kidney'mesenchyme (Risau and Ekblom, 1986). Acidic FGF (Lobb gt g1.,
1986) and basic FGF (Klagsbrun gt g1., 1986) have been identified in human
tumor cells in culture where they may serve to drive growth in an
autocrine fashion.

Although there is no direct evidence that FGF’s induce malignant

transformation, two recently described oncogenes, i.e. , int-2 derived from

 

mouse mamary tumor (Dickson and Peters, 1987) and fit identified in a
human stomach tumor (Yoshida gt gl., 1987), have been shown to have
closely related primary structures resembling the structure of the

fibroblast growth factors. In addition, Palmer t al . (1988) has recently

shown that the growth of normal human fibroblasts in soft agar is

stimulated by basic FGF.

g. Insulin-Like Growth Factors (IGF’s).

46

Members of ‘this group of’ related growth factors include IGF-I
(Somatomedin C), IGF-II (Somatomedin A), and factors which have been
described as having multiplication stimulating activity (MSA). MSA has not
been fully characterized but is generally thought to be homologous to IGF-
II (Froesch gt ,g1., 1987). IGF-I and IGF-II have been chemically
characterized and their structures have been elucidated (Rinderimecht and
Humbel, 1978). They are structurally homologous to insulin with
approximately 45% homology within the A and 8 regions (Blundell gt gl.,
1978). The C-peptide usually cleaved in the processing of proinsulin,
however, is present in the IGF’s and an additional D region extends from
the C-terminus of the A-chain. These molecules have molecular weights of
approximately 7 kd. They do not exist free in serum, but rather are bound
to carrier proteins in complexes from approximately 50,000-150,000 Kd
(Zapf gt gl., 1981).

The physiological roles of these molecules include regeneration of
tissues of mesodermal origin, wound healing and growth of cartilage and
bone (Froesch gt g1., 1987). The growth stimulatory effects of growth
hormone have been shown to be due to its ability to induce IGF-I synthesis
(Clemmons gt 11., 1983). IGF-I and IGF-II have been shown to stimulate DNA
synthesis and growth in human fibroblasts (Flier gt_gl., 1986; Froesch gt
g1., 1979; Clemmons and Van Wyk, 1981). The cells, however, proliferate
at a rate considerably less than the rate observed when serum or PDGF is
used as a mitogen (Schmidt gt g1., 1984). Most biological effects of IGF’s
can be mimicked by high doses of insulin (Froesch gt g1. 1979; Schmidt gt
g1., 1984), an observation consistent with the fact that IGF’s cross-react

with the insulin receptor (Rechler gt _l., 1983). The IGF-I and IGF-II

47
molecules interact preferentially'with type I and 11 receptors, which have
been characterized by chemical cross-linking studies and are found to be
distinct (Gameltoft, gt gl., 1986). Binding of IGF-I to its cognate
receptor activates the tyrosine kinase located on the cytoplasmic side of
the cellular membrane (Kasuga gt g1., 1983).

IGF-I has been detected in a variety of human tumor cells (Pavelick
gt 31,, 1986), and normal cells (Clemmons, 1983). IGF-I has been shown to
be increased in surgical biopsies of lung carcinomas (Minuto gt _a_l_.,
1986), and MSA factors have been detected in human fibrosarcoma-derived
8387 cell line (Marquart gt gl., 1980) and breast cancer cell lines (Huff
_t _1., 1986).

2. Internal Mechanisms.

The pathways leading to cell replication after growth factor
stimulation of mammalian cells are not completely elucidated. A number of
second messages, gene inductions and signals are reversibly activated in
response to mitogens. It is widely supposed that the constitutive
activation of some of these signal transduction pathways can lead to
uncontrolled cellular division, a hallmark of the malignant phenotype. For
these reasons, many investigators are studying the various steps in the
pathways that mediate growth factor action. Examples of major areas of
research include early gene transcription following mitogenic stimulation

and coupling proteins of the 1g; gene family.

a. Early Gene Transcription.

The addition of“ growth factors to quiescent cells induces the

48
expression of’a number of genes detectable within the first hour following
mitogenic stimulation. Genes such as my; (Kelley et.g1., 1983) fig; (Muller
et gl., 1984), b, JE and KC (Cochran et gl., 1983), p53 (Reich and
Levine, 1984), egr (Sukahamte et g1., 1987) and others (Lau and Nathans,
1987), have been shown to be induced in quiescent mammalian cells after
addition of serum or purified growth factors. A number of these genes were
first identified as the transforming genes of retroviruses. For example,
fig; is the transforming gene of the FBR murine sarcoma virus, my; is
homologous to the transforming gene of the MC-29 virus, and c-myg is the
transforming gene of the avian myeloblastosis virus (Varmus, 1984). The
ability of these viral gene products to cause malignant transformation in
appropriate animal hosts suggests that inappropriate expression or
increased expression of such genes is an important step in the production

of human cancer.

1) My; Gene.

The my; family of genes is a multigene family that share homology
with the v-_y; oncogene of the MC-29 virus (DePinho gt gl., 1987). C-my;
was first identified as the cellular homolog of v-myg, the transforming
gene of the MC-29 retrovirus (Varmus, 1984). N and L my; were identified
because of cross hybridization of conserved sequences to c-my; (Kohl gt
gl., 1986; Nau gt gl., 1985).

The my; protein is a 62,000 MW nuclear—associated phosphoprotein that
has a strong affinity in vitro for nucleic acids (Donner gt gl.,1982;
Persson and Leder, 1984, Stone gt gl., 1987). Although the function of

this protein is still unknown, ample evidence exists to suggest that c-

49

my; is involved in the pathways leading to cellular proliferation. C—my;
is expressed in fibroblasts after stimulation by a variety of mitogens
including PDGF (Kelley gt gl., 1983), EGF (Paulsson gt _l., 1987), FGF
(Mueller gt g1., 1984). Its peak expression in 3T3 cells occurs in 1-2
hours after stimulation, declining rapidly thereafter (Kelley gt gl.,
1983). Inappropriate overexpression of'my;_under an inducible promoter has
been show to stimulate cell growth (Armelin gt gl., 1984) and enhance
anchorage—independent growth stimulated by various growth factors
(Sorrentino gt gl., 1986).

There exist numerous examples of the association of my; expression
with tumor formation. C-my; may be activated in various classes of tumors
by a number of different mechanisms including viral transduction,
amplification, and promoter or enhancer insertion (Varmus, 1984). Tissue
specific deregulation of my; in transgenic mice leads to tumor appearance
with targeted organs (Stewart gt gl., 1984; Adams gt g1., 1985). Burkitt’s
lymphoma, a human malignancy of the B-cells, is frequently associated with
translocation of the my; gene to transcriptionally active areas near
immunogloblin genes (Taub gt gl., 1982; Erickson gt gl., 1983; Klein and
Klein, 1985). Amplification of c-my; has been detected in leukemias
(Collins and Groudine, 1982), and carcinomas (Alitalo gt gl., 1983). C-
my; has recently been shown to be overexpressed in SHAC cells, a human
fibrosarcoma-derived cell line (Andeol gt gl., 1988; Suarez gt gl., 1987).
Amplification of' N-my; is frequently associated with neuroblastomas
(Seeger, et g1., 1985). In experiments showing transformation of rodent
fibroblasts with my; and Lg; genes, my; is associated with the acquisition

of an infinite lifespan and mg; with the transformed phenotypes (Ruley gt

50
11., 1983; Land gt 111., 1983). Recently, an infinite lifespan human
fibroblast cell line has been isolated following transfection with the v—

my; gene (Hurlin gt gl., 1988).

2) fig; Gene.
The fig; gene was first identified as the transforming gene in two

mouse osteosarcoma viruses, FBJ-MSV (Finkel gt l., 1966) and FBR-MSV

(Finkel gt gl., 1973). The human cellular c-fig; and mouse c—fig; are 94%
homologous (Van Beveren gt g1., 1983). The protein product is a 380 amino
acid product with extensive post-translational modification, giving an
apparent molecular weight on SOS/PAGE gels of 62,000 Kd (Curran gt gl.,
1984) C-fig; is rapidly induced after stimulation by a variety of growth
factors and other agents (Kelley gt gfl,, 1983; Kruijer gt gl,, 1984;
Morgan and Curran, 1986; Ran gt gl., 1986). Peak expression occurs 30 to
45 minutes following stimulation and returns to basal levels after 1 to
2 hours (Kruijer gt _l_., 1984). fig; protein has been shown to form
complexes with other cellular proteins that bind DNA (Sambucetti and
Curran, 1986).

It is not known if fig; expression is an obligatory step in
stimulation of cellular proliferation, but expression of fig; antisense RNA
is able to partially inhibit the mitogenic response of 3T3 fibroblasts
stimulated by serum (Holt gt g_l_., 1986). This suggests that fig; is
involved with the transduction of the mitogenic signal from serum. C-fig;
expression in growing NIH 3T3 cells is slightly elevated compared to
quiescent cells (Muller gt g1., 1984). However, it is clear that continued

cellular division of exponentially growing cells does not require the

51
continued high level of fig; expression which is seen immediately after
mitogenic stimulation (Bravo gt gl., 1985). Several studies suggest that
the expression of fig; alone is not sufficient to drive replication. For
example, Zhan and Goldfarb (1986) transfected fig; into non-transformed 3T3
cells, isolated the transformants and assayed these cells in a serum free
medium which did not support the growth of control 3T3 cells. fig;-
transformants did not grow in the medium in the absence of exogenously

added growth factors, unlike ras, si

or ;y; transformants. Similarly,
Mueller gt g1. (1986) transfected 3T3 cells with fig; and assayed the
ability of the transfected cells to grow in medium supplemented with
platelet-poor plasma. The transformants did not replicate in this medium,
even though presumably it contained "progression" factors similar to
insulin and EGF as described by Stiles gt g1. (1979), but lacked growth
factors which could initiate replication of quiescent cells. In this study
the expression of fig; could not replace growth factors in stimulating the
replication of 3T3 fibroblasts. Recently, fig; has been shown to form a
complex with the c-jgg proto-oncogene and to enhance transcription of
genes containing AP-l transcriptional sites (Chiu gt _1., 1988; Sasoni-
Corsi gt g1., 1988). If this is the normal function of fig;, the inability
of increased expression of fig; alone to stimulate replication of cells in
culture, probably reflects the necessity to synergize with another gene
product to exert a mitogenic effect. Increased expression of c-fig; has
occasionally been reported in human tumor cells (Rijinders gt g1., 1985).

In summary, these studies suggest that fig; is frequently associated
with growth factor stimulation and conditions under which cells begin

rapid proliferation. However c-fig; alone is unable to replace mitogens

52
which initiate cellular division in fibroblasts and probably acts in
conjunction with other genes products to modulate cell growth growth in

response to mitogens.

3) p53 Gene.

The protein coded by the p53 gene is a nuclear phosphoprotein
(Crawford gt gl., 1980) present in high amounts in mammalian cells
transformed with DNA viruses (Lane and Crawford, 1979; Linzer gt g1.,
1980), RNA viruses (Rotter gt l. 1985) or by carcinogen treatment (Rotter
gt g1., 1985; DeLeo gt _l., 1979). Expression of high levels of p53 can
cause primary rodent cells to acquire an infinte lifespan in culture
(Jenkins gt_g1., 1984), and has been shown to cooperate with mg; genes in
the full malignant transformation of primary rodent cells similar to other
'immortalizing genes" such as fitg or my; (Parada gt g1., 1984).

In addition to its suggested role in transformation, p53 is
implicated in the control of normal cellular proliferation as shown by the
following studies. Expression of p53 mRNA, like other early transcripts,
is increased prior to DNA synthesis in quiescent 3T3 cells stimulated to
divide by various mitogens (Reich and Levine, 1984). Mercer gt g1. (1982)
microinjected p53 antibody into serum stimulated 3T3 mouse fibroblasts.
The normal entry of these cells into S phase was blocked by this
procedure. Although the mechanism of action and occurrence of p53 has not
been studied as extensively as that of some other early genes, it is
implicated in the control of normal cellular division in response to

certain mitogens. The overexpression of p53 leads to initiation of DNA

synthesis, and 'immortalization" of some rodent fibroblast cell lines,

53
suggesting that this protein could cause autonomous cell growth by

activation of various intracellular mechanisms.

b. Genes Involved in Signal Transduction- Rg; Genes.

The mamalian Lg; gene family consists of a group of homologous
guanine nucleotide binding proteins (Shih gt gl., 1982; Papageorge gt g1.
1982) that have an apparent molecular weight of 21,000 (Shih gt g1,,
1982). fig; proteins contain regions homologous to regulatory "G" proteins
which couple retinal light receptors to the optic nerve and that regulate
adenylate cyclase in response to binding of ligands at the B-adrenergic
receptor (Lochrie gt g1., 1985; Hurley gt _l., 1984). Therefore it is
postulated that cellular Lg; genes function in mitogenic signal
transduction. Early evidence that fig; proteins may be involved in normal
cell growth and development was obtained from experiments using the yeast,
S; Qervisiae. Deletion mutants which inactivate both fig;t and fig;; genes
render the yeast cells unable to enter a vegetative state (Tatchell gt
g1., 1984). Toda gt g1. (1985) utilized missense mutations to demonstrate
that some yg; function was needed for normal growth and that the fig; genes
in S; Cerjvjsige controlled adenylate cyclase. The fig;t and fig;g genes of
yeast exhibit a high degree of homology to the mammalian cellular mg;
(Tatchell gt gl., 1984).

Investigations of the role of tg;_genes in mammalian cell replication
have yielded evidence that Lg; may cause cell replication by a number of
different mechanisms. Direct microinjection of the protein product of the
c:fl;fig;_protooncogene or the T24 c-flgyg; oncogene has been shown to induce

DNA synthesis in normal human fibroblasts which were density inhibited and

54

quiescent (Lumpkin gt gl., 1986). This finding suggests that the
expression of increased fig; alone is mitogenic without exogenously added
growth factors. In agreement with this finding, anti-p21 antibody
microinjected into a normal diploid human fibroblast cell line, decreased
the fraction of replicating cells stimulated by serum (Mulcahy gt g1.,
1985). Consistent with these data are results of transfection studies in
NIH/3T3 cell by Zhan and Goldfarb (1986), who showed that increased fig;
expression allowed these transfected cells to grow in a serum-free medium
lacking peptide growth factors.

One mechanism by which fig; genes could stimulate cell growth is by
stimulating the production of growth factors. There is evidence that
transforming ES proteins can cause the production of certain growth
factors by Lg;-transformed cells. Spandidos (1985) described an experiment
where conditioned medium taken from Chinese hamster fibroblasts
transformed by either H-fl; or N-g;L was tested for the ability to
stimulate anchorage independent growth of NRK cells. High expression of
the ,gg; genes was correlated with high activity in the bioassay,
suggesting that transforming growth factors were secreted. Dickson gt_gl.
(1987) reported that transfection of MCF-7, a human breast cancer-derived
cell line, with v-H-fig; caused production of TGF-a, TGF-b and insulin like
growth factor. Similarly, Durkin and Whitfield (1987) studied normal rat
kidney fibroblasts transformed by a temperature-sensitive mutant of v-K;
fig;. These investigators found that protamine sulfate, a compound which
has been reported to interfere specifically with PDGF action, inhibited
cellular proliferation at the permissive temperature. The investigators

concluded that v-K-yg; caused production and secretion of PDGF.

55

Another mechanism by which fig; proteins cause cellular replication
may be through internal activation of biochemical cascades which are
triggered by growth factors. One of the earliest events known to occur
upon binding of some growth factors to the membrane is phospholipase-C-
catalyzed hydrolysis of phosphatidylinositol 4,5-bis-phosphate (PIPz), a
minor component of certain lipids in the plasma membrane (Berridge, 1984).
The two major hydrolysis products of PIPz are 1,2-diacylglycerol (DAG) and
inositol 1,4,5-triphosphate (1P3). DAG synergizes with calcium ions to
activate protein kinase-C (PKC) (Kikkawa gt gl., 1983), and 1P3 causes
release of calcium from intracellular stores (Batty gt g1., 1985). PKC is
considered to be a key regulatory enzyme, the activation of which has been
shown to cause phosphorylation of many cellular macromolecules (Nishizuka,
1980; Connolly gt _1., 1986; Gould gt _l., 1986) and the secretion of
peptides (Nishizuka, 1986). The events necessary for the initiation of DNA
synthesis subsequent to PKC activation are not clear. In many cell types,
the cellular my; and fig; genes are transiently expressed within a few
hours following growth factor binding (McCaffrey' gt ,g1., 1987). In
addition, a membrane bound Na+/H+ antiporter is activated within a few
minutes which causes cytoplasmic alkalinization, an event considered
critical for DNA replication (Hesketh gt gl., 1985).

Several studies suggest cellular fig; genes may participate in this
cascade. For example, Fleischman g g. (1986) compared the DAG/PIPz ratio
in normal and fig; transformed rodent fibroblasts. They found that
transformed cells had a 2.5 fold higher ratio and suggested that the fig;
p21 affected the levels of PIP2 through regulation of phospholipase-C in

these transformed cells. The experiments by Wakelam gt g1. (1986) support

56

the theory that fig; p21 proteins were regulating phospholipase-C. They
used a dexamethasone-inducible promoter linked to N-fig; to show that the
non-transforming N-r_a_; p21 enhanced PIPz hydrolysis in response to growth
factor stimulation. Further support for the hypothesis that fig; p21
regulates phospholipase-C was given by Lacal gt g1. (1987), who
demonstrated DAG production in Xenopus oocytes stimulated by
microinjection of H-fig; p21.

Rg;_genes are implicated in both the transduction of normal cellular
growth signals and the production of cells capable of autonomous growth
when the fig; gene is mutated or inappropriately expressed. While the
expression of a mutated fig; gene appears to be a main mechanism by which
fig; causes malignant transformation (Spandidos, 1985), there is evidence
to suggest that overexpression of a normal r s can also cause

transformation of cells.

LIST OF REFERENCES

57

LIST OF REFERENCES

Abramson, D. H., Ellsworth, R. M., Kitchin, F. D., and Tung, G. (1984).
Second nonocular tumors in retinoblastoma survivors - are they radiation-
induced? Opthalmology, 91, 1351-1355.

Adams, J. M., Harris, A. W., Pinkert, C. A., Corcoran, L. M., Alexander,
S., Cory, S., Palmiter, R. D. and Brinster, R. L. (1985). The c-_y;
oncogene driven by immunoglobulin enhancers induces lymphoid malignancy
in transgenic mice. Nature, 118, 533.

Alitalo, K., Ramsay, 6., Bishop, J. M., Pfeifer, S. 0., Colby, W. W., and
Levinson, A. D. (1983). Identification of nuclear proteins encoded by
viral and cellular my; oncogenes. Nature 19;, 274.

Allen, A. C., and Spitz, S. (1954). Histogenesis and clinicopathologic
correlation of nevi and malignant melanoma. .Arch. Dermatol. Syphilol. g;,
150.

Andeol, Y., Nardeux, P. C., Daya-Grosjean, L., Brisbon, 0., Cebrian, J.,
and Suarez, H. (1988). Both N-fig; and c-my; are activated in the shac
human stomach fibrosarcoma cell line. Int. J. Cancer. 51, 732-737.

Antoniades, H. N. (1981). Human platelet-derived growth factor (PDGF):
Purification of PDGF-I and PDGF-II and separation of ‘their reduced
subunits. Proc. Natl. Acad. Sci. USA 18, 7314-7317.

Anzano, M. R., Roberts, A. 8., Smith, J. M., Sporn, M. 8., and De Larco,
J. E. (1983). Sarcoma growth factor from conditioned medium of virally
transformed cells is composed of both type and type transforming
growth factors. Proc. Natl. Acad. Sci. USA 80, 6264-6268.

Armelin, H. A., Armelin, M. C. S., Kelly, K., Stewart, T., Leder, P.,
Cochran, B. H., and Stiles, C. D. (1984). Functional role of c-my; in

mitogenic response to platelet-derived growth factor. Nature 319, 655-
660.

Azzarone, 8., and Pedulla, D. (1976). Spontaneous transformation of
human fibroblast cultures derived from bronchial carcinomata. Eur. J.
Cancer. 12, 557-561.

Balk, S. D. (1971). Calcium as a regulator of the proliferation of normal,
but not of transformed, chicken fibroblasts in a plasma-containing medium.
Proc. Natl. Acad. Sci. USA gg, 271-275.

Barrett, J. C., and Ts’o, P. O. P. (1978). Evidence for the progressive

58

nature of neoplastic transformation in vitro. Proc. Natl. Acad. Sci. USA
15, 3761-2765.

 

Batty, I. R., Shorski, S. R., and Irvine, R. F. (1985). Rapid formation
of inositol-1,3,4,5-tetrakisphosphate following muscarinic receptor
stimulation of rat cerebral cortical slices. Biochem. J. ggg, 211-215.

Beckman, M. P., Betscholtz, C., Heldin, C.-H., Westermark, 8., Di Marco,
E., Di Fiore, P. P., Robbins, K. C., and Aaronson, S. A. (1988).
Comparison of biological properties and transforming potential of human
PDGF-A and PDGF-8 chains. Science 251, 1346-1349.

Benedict, W.F., Banerji, A., Mark, C. and Murphee, A.L. (1983).
Nonrandom chromosomal changes in untreated retinobl astomas. Cancer Genet.
Cytogenet. 19, 311-333.

Berenblum, I. (1940). The cocarcinogenic action of croton resin. Cancer

Res. 55-18.

Berridge, M. J. (1984). Inositol triphospahte and diacylglycerol as
second messengers. Biochem. J. zgg, 345-360.

Besmer, P., Snyder, H. W., Murphy, J. E., Hardy, W. D., and Parodi, A.
(1983). The Parodi-Irgens feline sarcoma virus and simian sarcoma virus
have homologous oncogenes, but in different contexts of the viral genomes.
J. Virol. 5;, 606-613.

Betscholtz, C., Westermark, 8., Ek, 8., and Heldin, C.-H. (1984).
Coexpression of a PDGF-like factor and PDGF receptors in a human

osteosarcoma cell line: implications for autocrine receptor activation.
Cell. 32, 447-457.

Betsholtz, C., Johnnson, A., Heldin, C., and Westermark, B. (1986).
Efficient reversion of Simian sarcoma virus-transformation and inhibition
of growth factor-induced mitogenesis by suramin. Proc. Natl. Acad. Sci.
USA, 83, 6440-6444.

Biederman, K. A., and Landolph, J. R. (1987). Induction of anchroage
independence in human diploid foreskin fibroblasts by carcinogenic meteal
salts. Cancer Res., 51, 3815-3823.

Blundell, T. L., Bedarkar, S., Rinderknecht, E., and Humbel, R. E.
(1978). Insulin-like growth factors: A model for tertiary structure
accounting for immunoreactivity and receptor binding. Proc. Natl. Acad.
Sci. USA, 15, 180.

Bockus, B. J. and Stiles, C. D. (1984). Regulation of cytoskeletal
architecture by platelet-derived growth factor, insulin, and epidermal
growth factor. Exp. Cell Res., 151, 186-197.

Bohlen, P., Baird, A., Esch, F., Ling, N., and Gospodarowicz, D. (1984).
Isolation and partial molecular characterization of pituitary fibroblast

59
growth factor. Proc. Natl. Acad. Sci. USA 81, 5364-5368.

Bowen-Pope, D. F., Vogel, A., and Ross, R. (1984). Production of
platelet-derived growth factor-like molecules and reduced expression of
platelet-derived growth factor receptors accompany transformation by a
wide spectrum of agents. Proc. Natl. Acad. Sci. USA 81, 2396-240.

Boyton, A., Whitfield, J. F., and Tremblay, R. G. (1977). Different
extracellular requirements for proliferation of nonneoplastic,
preneoplastic, and neoplastic mouse cells. Cancer Res. 81, 2657-2661.

Boyton, A., Whitfield, J. F., and Kleine, L. P. (1983).
Ca++/phospholipid-dependent protein kinase activity correlates to the
ability of transformed liver cells to proliferate in Ca++-deficient
medium. Biochem. Biophys. Res. Commun. 118, 383-390.

Bravo, R., Burckhardt, J., Curran, T., and Muller, R. (1985).
Stimulation and inhibition of growth by EGF in different A431 cell clones
is accompanied by the rapid induction of c-fig; and c-my; proto-oncogenes.
EMBO J. t, 1153.

Bright, M. D. and Gaffney, E. D. (1982). Demonstration of competence and
progression activities for human fibroblasts. Exp. Cell Res. 181, 309-316.

Buckley, C. H., Butler, E. B., and Fox, H. (1982). Cervical
intraepithelial neoplasia. J. Clin. Pathol. 88, 1-13.

Burbeck, S., Latter, G., Metz, E., and Leavitt, J. (1984). Neoplastic
human fibroblast proteins are related to epidermal growth factor
precursor. Proc. Natl. Acad. Sci. USA 81, 5360-5363.

Busch, C., Wasteson, A., and Westermark, B. (1976). Release of a cell
growth promoting factor from human platelets. Thromb. Res. 8, 493-500.

Cairns, J. (1975). The cancer problem. Sci. Amer. 888, 64.

Carpenter, G., Stoscheck, C. M., Preston, Y. A., and De Larco, J. E.
(1983). Antibodies to the epidermal growth factor receptor block the
biological activities of sarcoma growth factor. Proc. Natl. Acad. Sci.
USA 88, 5627.

Cavanee, W. E., Dryja, T. P., Phillips, R. A., Benedict, W. F., Godbout,
R., Gallie, B. L., Murphee, A. L., Strong, L. C., and White, R. I.
(1983). Expression of recessive alleles by chromosomal mechanisms
inretrinoblastoma. Nature 888, 779-784.

Chang, S. E. (1986). In vitro transformation of human epithelial cells.
Biochim. Biophys. Acta. 888, 161-194.

Cheung, H. S., Van Wyk, J. J., Russell, W. E., and McCarty, D. J. (1986).
Mitogenic activity of hydroxyapatite: requirement for Somatomedin C. J.
Cell. Physiol. 188, 143-148.

60

Chiu, R., Boyle, W. J., Meek, J., Smeal, T., Hunter, T., and Karin, M.
(1988). The c-fos protein interacts with c-Jun-AP-l to stimulate
transcription of AP-l responsive genes. Cell 88, 541-552.

Christopherson, W. M. et al. (1976). Cervical cancer control: a study of
morbidity and mortality trends over a twenty-one-year period. Cancer 88,
1357-1366.

Christopherson, W. M. (1977). Dysplasia, carcinoma-in-situ and
microinvasive carcinoma of the uterine cervix. Hum. Pathol. 8, 481.

Clark, Jr., W. J., From, L., Bernardino, E. A., and Mihm, M. C. (1969).
The histogenesis and biologic behavior of primary human malignant melanoma
of the skin. Cancer Res. 88, 705-727.

Clemmons, D. R. and Van Wyk, J. J. (1981). Somatomedin-C and platelet-
derived growth factor stimulate human fibroblast replication. J. Cell.
Physi8ol. 188, 361-367.

Clemmons, D. R., Underwood, L. E., and Van Wyk, J. J. (1981). Hormonal
control of immunoreactive somatomedin production by cultured human
fibroblasts. J. Clin. Invest. 81, 10-19.

Clemmons, D. R. (1983). Age dependent production of a competence factor
by human fibroblasts. J. Cell. Physiol. 111, 61-67.

Cochran, 8. H., Reffle, A. C., and Stiles, C. D. (1983). Molecular

cloning of gene sequences regulated by platelet-derived growth factor.
Cell 88, 939-947.

Coffey, Jr., R. J., Goustin, A. S., Soderquist, A. M., Shipley, G. D.,
Wolfshohl, J., Carpenter, G., and Moses, H. L. (1987). Transforming
growth factor and expression in human colon cancer lines:
Implications for an autocrine model. Cancer Res. 11, 4590-4594.

Cohen, H. (1962). Isolation of a mouse submaxillary gland protein
accelerating incisor eruption and eyelid opening in the newborn animal.
J. Biol. Chem. 881, 1555-1562.

Cohen, S. (1983). The epidermal gorwth factor. Cancer 81, 1787.

Collins, S. and Groudine, M. (1982). Amplification of endogenous my;-
related DNA seuqences in a human myeloid leukaemia cell line. Nature 888,
679-681.

Connolly, T. M., Lawing, W. J., and Majerus, P. W. (1986). Protein kinase
C phosphorylates human platelet inositol trisphosphate 5’-
phosphomonoesterase, increasing the phosphatase activity. Cell 18, 951-
958.

Crawford, L. V., Lane, D. P., Denhardt, D. T., Harlow, E. E., Nicklin, P.

61

M., Osborn, K., and Pim, D. C. (1980). Characterization of the complex
between SV40 large T antigen and the 53 k host protein in transformed
mouse cells. Cold Spring Harbor Symp. Quant. Biol. 11, 179.

Day. Jr., C. L., Mihm, Jr., M. C., Lew, R. A., Kopf, A. W., Sober, A. J.,
and Fitzpatrick, T. B. (1982). Cutaneous malignant melanoma: Prognostic
guidelines for physicians and patients. CA 88, 113-122.

DeCaprico, J. A., Ludlow, J. W., Figge, J., Shew, J.-Y.,Huang, C.-M., Lee,
W.-H., Marsilio, E., Paucha, E., and Livingston, D. M. (1988). SV40
large tumor antigen forms a specific complex with the product of the
retinoblastoma susceptibility gene. Cell 88, 275-283.

DeLarco, J. E. and Todaro, G. J. (1978). A human fibrosarcoma cell line
producing multiplication stimulating activity (MS)-related peptides.
Nature 818, 356-358.

DeLeo, A. 8., Jay, G., Appella, E., Dubois, G. C., Law, L. W., and Old,
L. J. (1979). Detection of a transformation-related antigen in chemically
induced sarcomas and other transformed cells of the mouse. Proc. Natl.
Acad. Sci. USA 18, 2420.

Delli-Bovi, P., Curatola, A. M., Newman, K. M., Sato, Y., Moscatelli, D.,
Hewick, R. M., Rifkin, D. 8., and Basilico, C. (1988). Processing,
secretion, and biological properties of a novel growth factor of the
fibroblast growth factor family'with oncogenic potential. Mol. Cell. Biol.
8, 2933-2941.

Demars, R. and Jackson, J. L. (1977). Mutagenicity detection with human
cells. J. Environ. Pathol. Toxicol. 1, 55-77.

DePinho, R. A., Mitsock, L., Hatton, K., Ferrier, P., Zimmerman, K.,
Legouy, E., Tesfaye, A., Collum, R., Yancopoulos, G., Nisen, P., Kriz, R.,
and Alt, F. W. (1987). Myc family of cellular oncogenes. J. Cell.
Biochem. 88, 257-266.

Derynck, R., Roberts, A. 8., Winkler, M. E., Chen, E. Y., and Goeddel, D.
V. (1984). Human transforming growth factor-a: Precursor structure and
expression in E. coli. Cell 88, 287-297.

Derynck, R., Goeddel, D. V., Ullrich, A., Gutterman, J. U., Williams, R.
0., Bringman, T. S., and Berger, W. H. (1987). Synthesis of messenger
RNAs for transforming growth factors a and b and the epidermal growth
factor receptor by human tumors. Cancer Res. 81, 707-712.

Deuel, T. F., Huang, J.S., Proffitt, R. T., Baenziger, J. U., Chang, D.,
and Kennedy, 8. B. (1981). Human platelet-derived growth factor.
Pufication and resolution into two active protein factions. .1. Biol.
Chem. 888, 8896-8899.

Devare, S. G., Reddy, E. P., Law, D. J., Robbins, K. C., and Aarsonson,
S. A. (1983). Nucleotide sequence of the simian sarcoma virus genome:

62

demonstration thgts its acquired cellular sequences encode the transforming
gene product p28 .Proc. Natl. Acad. Sci. USA 80, 731- 735.

Dickson, C., and Peters, G. (1987). Potential oncogene product related
to growth factors. Nature (London) 888, 833.

Doll, R. (1978). An epidemiological perspective of the biology of
cancer. Cancer Res. 88, 3573-3583.

Donner, P., Greiser-Wilke, 1., and Moelling, K. (1982). Nuclear
localization and DNA binding of the transforming gene product of avian
myelocytomatosis virus. Nature 888, 262.

Doolittle, R. F., Hunkapiller, M. W., Hood, L. E., Devare, S. G., Robbins,
K. C., Aaronson, S. A., and Antoniades, H. N. (1983). Simian sarcoma
virus one gene, v-;t;, is derived from the gene (or genes) encoding a
platelet-derived growth factor. Science 881, 275-277.

Downward, J., Yarden, E., Mayes, E., Scrace, G., Totty, N., Stockwell, P.,
Ullrich, A., Schlessinger, J., and Waterfield, M. (1984). Close
similarity of the epidermal growth factor receptor and v-grg-B oncogene
protein sequences. Nature 881, 521-527.

Dryja, T.P., Rapport, J.M., Joyce, J.M>, Petersen, R.A. (1986). Molecular
detection of deletions involving the ql4 band of chromosomne 13 in
retinoblastoma. Proc. Natl. Acad. Sci. 88: 7391-7394.

Duncan, M. R., Rankin, P. R., Kind, R. L., Jacobson, M. K., and Dell’Orco,
R. T. (1988). Stimulation of mono(ADP-Ribosyl)ation by reduced
extracellular calcium levels in human fibroblasts. J. Cell. Physiol. 181,
161-165.

Durkin, J. P. and Whitfield, fls (1987). Evidence that the viral Ki-
ras protein, but not the pp60_ rdrotein of ASV, stimulates proliferation
through the PDGF receptor. Biochem. Biophys. Res. Comm. 148, 376- 383.

Eagle, H. (1955). Nutrition needs of mammalian cells in tissue culture.
Science 188, 501-504.

Ek, B. and Heldin, C. H. (1982). Characterization of a tyrosine specific
kinase activity in human fibroblast membranes stimulated by platelet
derived growth factor. J. Biol. Chem. 881, 10486.

Elder, D. E., Goldman, L. I., Goldman, S. C., Greene, M. H., and Clark,
Jr., W. J. (1980). Dysplastic nevus syndrome. Cancer 88, 1787.

Ellis, 0. L., Kafka, S. P., Chow, J. C., Nanney, L. 8., Inman, W. H.,
McCadden, M. E., and King, Jr. , L. E. (1987). Melanoma, growth factors,
acanthosis nigricans, the sight of ester- thelat, and multiple
acrochordons. N. Engl. J. Med. 811, 1582- 1587.

Erickson, J. ar-Rushdi, A., Drwinga, H. L., Nowell, P. C., and Croce, C.

63

M. (1983). Transcriptional activation of the translocated c-my; oncogene
in Burkitt lymphoma. Proc. Natl. Acad. Sci. USA 88, 820-824.

Escobedo, J. A., Navankasatussas, S., Cousens, L. S., Coughlin, S. R.,
Bell, G. 1., and Williams, L. T. (1988). A conlnon PDGF receptor is
activated by homodimeric A and 8 forms of PDGF. Science 888, 1532-1534.

Eva, A., Robbins, K. C., Andersen, P. R., Srinivasan, A., Tronick, S. R.,
Reddy, E. P., Ellomore, N. W., Galen, A. T., Lautenberger, J. A., Papas,
T. S., Westin, E. H., Wong-Staal, F., Gallo, R. C., and Aaronson, S. A.
(1982). Cellular genes analogous to retroviral gg; genes are transcribed
in human tumour cells. Nature 888, 116-119.

Fearon, E. R., Vogelstein, 8., and Feinberg, A. P. (1984). Somatic
deletion and duplication of genes on chromosome 11 in Wilms’ tumours.
Nature 888, 176-178.

Ferency, A. (1982). Anatomy and histology of the cervis; benign lesions
of the cervix., cervical intraepithelial neoplastia; carcinoma and other
malignant tumors of the cervix. In: A. Blaustein (Ed.), Pathology of the
Female Genital Tract, 2nd Ed., Springer-Verlag, NY, pp. 119-135.

Finkel, M. P., Biskis, 8. 0., and Jinkins, P. 8. (1966). Virus induction
of osteosarcomas in mice. Science 181, 698-701.

Finkel, M.P., Reilly, C.A., Biskis, 8.0. and Greco, I.L. (1973). Bone
tumor viruses. In: Colston Papers, Proc. 24th Symp. Colston Res. Soc.,
Butterwoths, London Vol. 24, 353-366.

Fleischman, L. F., Chahwala, S. 8., and Cantley, L. (1986). Ras-
transformed cells: Altered levels of phosphatidylinositol-r,5k-
bisphosphate and catabolites. Science 881, 407-410.

Flier, J. S., Usher,P., and Moses, A. C. (1986). Monoclonal antibody to
the type I insulin-like growth factor (IGF-I) receptor blocks IGF-I
receptor-mediated DNA synthesis: Clarification of the mitogenic
mechanisms of IGF-I and insulin in human skin fibroblasts. Proc. Natl.
Acad. Sci. USA 88, 664-668.

Fogh, J. and Giovanella, B. C. (Eds.) (1982). The Nude~ Mouse in
Experimental and Clinical Research, Vol. 2. Academic Press, NY.

Frackelton, Jr., A. E., Tremble, P. J., and Williams, L. T. (1984).
Evidence for the platelet-derived growth factor-stimulated tyrosine
phosphorylation of the platelet-derived growth factor receptor ig vivo.
J. Biol. Chem. 888, 7909.

Freedman, V. H. and Shin, S. I. (1974). Cellular tumorigenicity in nude
mice: correlation with cell growth in semi-solid medium. Cell 8, 355-
359.

Freedman, V. H. and Shin, S. (1977). Isolation of human diploid cell

64

variants with enhanced colony-forming efficiency in cell variants with
enhanced colony-forming efficiency in semisolid medium after a single-step
chemical mutagenesis. J. Natl. Cancer Inst. 88, 1873-1875.

Friend, S.H., Bernards, S., Rogelj, S., Weinberg, R.A., Rapaport, J.M.,
Albert, D.M., Dryja, T.P. (1986). A human DNA segment with the properties
of the gene that prediposes to retinoblastoma and osteosarcoma. Nature
(London) 88: 643-646.

Froesch, E. R., Zapf, J., Walter, H., Schoenle, E., Rinderknecht, E., and
Humbel, R. E. (1979). Serum NSILA and IGF: News and riddles. Horm. Cell
Regul. 3, I49.

Frolik, C. A., Wakefield, L. M., Smith, D. M., and Sporn, M. B. (1984).
Characterization of a membrane receptor for transforming growth factor-
b in normal rat kidney cells. J. Biol. Chem. 888, 10995-11000.

Frolik, C. A., Dart, L. L., Meyers, C. A., Smith, D. M., and Sporn, M. 8.
(1983). Purification and initial characterization of a type-b
transforming growth factor from human placenta. Proc. Natl. Acad. Sci.
USA 88, 3676-3689.

Fry, 0. G., Milam, L. 0., Maher, V. M., and McCormick, J. J. (1986).
Transformation of diploid human fibroblasts by DNA transfection with v-
;1; oncogene. J. Cell. Physiol. 188, 313-321.

Gammeltoft, S., Haselbacher, G. K., Humbel, R. E., Fehlmann, M., and Van
Obberghen, E. (1986). Two types of receptor for insulin-like growth
factors in mammalian brain. EMBO J. 1, 3407-3412.

Gardner, R. M., Lingham, R. 8., and Stancel, G. M. (1987). Contractions
of the isolated uterus stimulated by epidermal growth factor. FASEB, 224-
228.

Gensburger, G., Labourdette, G., and Sensenbrenner, M. (1987). Brain
basic fibroblast growth factor stimulates the proliferation of rat
neuronal precursor cells in vitro. FEBS Lett. 811, 1-5.

Gilmore, T., DeClue, J. E., and Martin, G. S. (1985). Tyrosine kinase
activity associated with the v-erb-B gene product. In: Cancer Cells
3/Growth Factors and Transformation. J. Fermisco, B. Ozzane, and C.
Stiles (Eds.), Cold Spring Harbor Press, Cold Spring Harbor, NY, pp. 25-
32.

Gimenez-Gallego, G., Conn, G., Hatcher, V. M., and Thomas, K. A. (1986).
The complete amino acid sequence of human brain-derived acidic fibroblast
growth factor. Biochem. Biophys. Res. Commun. 188, 611-617. FASEB, 224-
228.

Gospodarowicz, D., Moran, M., Braun, D., and Birdwell, C. R. (1976).
Clonal growth of bovine endotghelial cells in tissue culture:
fibroblastic growth factor as a survival agent. Proc. Natl. Acad. Sci.

65
USA 18, 4120-4124.

Gould, K. L., Woodgett, J. R., Isacke, C. M., and Hunter, T. (1986). The
protein-tyrosine kinase substrate p36 is also a substrate for protein
kinase C in vitro and in vivo. Mol. Cell. Biol. 8, 2378-2744.

Goustin, A. S., Betscholtz, C., Pfeifer-Ohisson, S., Persson, H., Rydnert,
J., Bywater, M., Holmgren, G., Heldin, C.-H., Westermark, 8., and Ohlsson,
R. (1985). Coexpression of the ;t; and my; proto-oncogenes in developing
human placenta suggests autocrine control of trophoblast growth. Cell 81,
301-312.

Greenberg, M. E., and Ziff, E. G. (1984). Stimulation of 3T3 cells
induces transcription of the c-fos proto-oncogene. Nature 811, 433-438.

Gregory, H. (1975). Isolation and structure of urogastrone and its
relationship to epidermal growth factor. Nature 881, 325-327.

Harris, C. C. (1987). Human tissues and cells in carcinogenesis
research. Cancer Res. 81, 1-10.

Hart, C. E., Forstrom, J. W., Kelly, J. D., Seifert, R. A., Smith, R. A.,
Ross, R., Murray, M. J., and Bowen-Pope, D. F. (1988). Two classes of
PDGF receptors recognize different isoforms of PDGF. Science 888, 1529-
1531.

Hayashi, I. and Sato, G. (1976). Replacement of serum by hormones permits
growth of cells in a defined medium. Nature 888, 132-134.

Hayflick, L. (1965). The limited 1_ vitro lifetime of human diploid
strains. Exp. Cell. Res. 88, 614.

 

Health and Welfare Department, Canada. New Release, March 9, 1977.

Heldin, C-H., Westermark, 8., and Wasteson, A. (1981). Platelet-derived
growth factor: isolation by a large-scale procedure and analysis of
subunit composition. Biochem. J. 188, 907-913.

Heldin, C.-H., Johnsson, A., Wennergren, S., Wernstedt, C., Betscholtz,
C., and Westermark, B. (1986). A human osteosarcoma cell line secretes
a growth factor structurally related to a homodimer of PDGF A-chains.
Nature 818, 511-514.

Heldin, C.-H., Betsholtz, C., Claesson-Welsh, C., and Westermark, B.
(1987). Subversion of growth regulatory pathways in malignant
transformation. Biochim. Biophys. Acta 881, 219-244.

Hepler, J. R., Nakahata, N., Lovenberg, T. W., DiGuiseppi, J., Herman, 8.,
Earp, H. S., and Harden, T. K. (1987). Epidermal growth factor
stimulates the rapid accumulation of inositol (1,4,5)-trisphosphate and
a rise in cytosolic calcium mobilized from intracellular stores in A431
cells. J. Biol. Chem. 1, 2951-2956.

66

Hesketh, T. R., Moore, H. P., Morris, J. D. H., Taylor, M. V., Rogers, J.,
Smith, G. A., and Metcalfe, J. C. (1985). A common sequence of calcium
and pH signals in the mitogenic stimulation of eukaryotic cells. Nature
818, 481-484.

Holley, R. W., Boylen, R. W., Fava, R., Baldwin, J. H., Kleeman, G., and
Armour, R. (1980). Purification of kidney epithelial cell growth
inhibitors. Proc. Natl. Acad. Sci. USA 11, 5989.

Holley, R. W. (1975). Control of growth of mammalian cells in cell
culture. Nature 888, 487-490.

Holt, J. T., Vankat, Gopal, T., Moulton, A. D., and Nienhuis, A. W.
(1986). Inducible production of c-fig; antisense RNA inhibits 3T3 cell
proliferation. Proc. Natl. Acad. Sci. USA 88, 4794-4798.

Huff, K. K., Kaufman, 0., Gabbay, K. H., Spender, E. M., Lippman, M. E.,
and Dickson, R. B. (1986). Secretion of an insulin-like growth factor-
I-related protein by human breast cancer cells. Cancer Res. 18, 4613-
4619.

Hurley, J. 8., Simon, M. 1., and Teplow, D. B. (1984). Homologies

between signal transducing G proteins and fig; gene products. Science 888,
860-862.

Hurlin, P. J., Fry, 0. G., Maher, V. M., and McCormick, J. J. (1987).
Morphological transformation, focus formation, and anchorage independence
induced in diploid human fibroblasts by expression of a transfected H-fig;
oncogene. Cancer Res. 11, 5252-5257.

Hurlin, P., Maher, V. M., and McCormick, J. J. (1988). Malignant
transformation of a human fibroblast cell strain MSU-1.l, by expression
of a transfected T24 H-fig; oncogene. Proc. Natl. Acad. Sci. USA, in
press.

Ives, H. E. and Daniel, T. 0. (1987). Interrelationship between growth
factor-induced pH changes and intracellular Ca++. Proc. Natl. Acad. Sci.
USA 81, 1950-1954.

Jenkins, J. R., Rudge, K., and Currie, G. A. (1984). Cellular
immortalization by a cDNA clone encoding the transformation-associated
phosphoprotein p53. Nature 818, 651.

Johnsson, A., Heldin, C.-H., Westermark, 8., and Wasteson, A. (1982).
Platelet-derived growth factor: identification of constituent
polypeptides. Biochem. Biophys. Res. Commun. 188, 66-74.

Johnsson, A., Heldin, C.-H., Wasteson, A., Westermark, 8., Deuel, T. F.,
Huang, J. S., Seeburg, D. H., Gray, E., Ullrich, A., Scrace, G.,
Stroobant, P., and Waterfield, M. D. (1984). The c-;t; gene encodes a
precursor of the 8 chain of platelet-derived growth factor. EMBO J. 8,

67
921-928.

Johnsson, A. C., Betscholtz, C., Heldin, C.-H., and Westermark, B.
(1985). Antibodies against platelet-derived growth factor inhibit acute
transformation by simian sarcoma virus. Nature 811, 438-440.

Kakunaga, T. (1977). Neoplastic transformation of human diploid
fibroblasts by chemical carcinogens. Proc. Natl. Acad. Sci. USA 18, 1334-
1338.

Kaplan, P. L., Anderson, M., and Ozanne, B. (1982). Transforming growth
factor(s) production enables cells to grow in the absence of serum: An
autocrine system. Proc. Natl. Acad. Sci. USA 18, 485-489.

Kasuga, M., Zick, Y., Blithe, D. L., Karlsson, F. A., Haring, H. U., and
Kahn, C. R. (1983). Insulin stimulation of phosphorylation of the beta
subunit of the insulin receptor. Formation of both phosphoserine and
phosphotyrosine. J. Biol. Chem. 881, 9891-9894.

Kelley, K., Cochran, B. A., Stiles, C. D., and Leder, P. (1983). Cell
specific regulationof the c-my; gene by lymphocyte mitogens and platelet-
derived growth factor. Cell 88, 305.

Kikkawa, U., Takai, Y., Tanaka, Y., Miyake, R., and Nishizuka, Y. (1983).
Protein kinase C as a possible receptor protein of tumor-promoting phorbol
esters. J. Biol. Chem. 888, 11442-11445.

Kim, M. K., Warren, T. C., and Kimball, E. S. (1985). Purification and
characterization of a low'molecular'weight transforming growth factor from
the urine of melanoma patients. J. Biol. Chem. 888, 9237-9243.

Klagsbrun, M., Sasse, J., Sullivan, R., and Smith, J. A. (1986). Human
tumor cells synthesize an endothelial cell growth factor that is
structurally related to basic fibroblast growth factor. Proc. Natl. Acad.
Sci. USA 88, 2448-2452.

Klein, G. and Klein, E. (1985). Myc-lg juxtaposition by chromosomal
translocations: Some new insights, puzzles and paradoxes. Immunol. Today
8, 208.

Knudson, A. G. (1977). Genetic predisposition to cancer, In: H. H.
Hiatt, JJ. 0. Watson, and J. A. Winsten (Eds.), The Origins of Human
Cancer, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp.
45-54.

Knudson, A. G. and Strong, L. C. (1972). Mutation and cancer: a model
for Wilms’ tumor of the kidney. J. Natl. Can. Inst. 18, 313-324.

Knudson, Jr., A. G. (1985). Hereditary cancer, oncogenes, and
antioncogenes. Cancer Res. 18, 1437-1443.

Kohl, N. E., Legouy, E.., DePinho, R. A., Nisen, P. D., Smith, R. K., Gee,

68

C. E., and Alt, F. W. (1986). Human N-my; is closely related in
organization and nucleotide sequence to c-myg. Nature 818, 73-77.

Koufos, A., Hansen, M. F., Lampkin, 8. C., Workman, M. L., Copeland, N.
G., Jenkins, N. A., and Cavenee, W. K. (1984). Loss of alleles at loci
on human chromosome 11 during genesis of Wilms’ tumor. Nature 888, 170-
178.

Kraemer, P. M., Ray, F. A., Brothman, A. R., Bartholdi, M. F., and Cram,
L. C. (1986). Spontaneous imortalization rate of cultured Chinese
hamster cells. J. Natl. Cancer Inst. 18, 703-709.

Kruijer, W., Cooper, J. A., Hunter, T., and Verma, I. M. (1984).
Platelet-derived growth factor induces rapid but transient expression of
the c-fig; gene and protein. Nature 818, 711-720.

Kurobe, M., Furukawa, S., and Hayashi, K. (1985). Synthesis and
secretion of an epidermal growth factor (EGF) by human fibroblast cells
in culture. Biochem. Biophys. Res. Comm. 181, 1080-1085.

Lacal, J. C., de la Pena, P., Moscat, J., Garcia-Barreno, P., Anderson,
P. S., and Aaronson, S. A. (1987). Rapid stimulation of diacylglycerol
production of lenogus oocytes by microinjection of H-fig; p21. Science
888, 533-536.

Land, H., Parada, L. F., and Weinberg, R. A. (1983). Cellular multistep
carcinogensis. Science 888, 771-778.

Land, H., Parada, L. F., and Weinberg, R. A. (1983b). Tumorigenic
conversion of primary embryo fibroblasts requires at least two cooperating
oncogenes. Nature 888, 596-602.

Lane, 0. P. and Crawford, L. V. (1979). T antigen is bound to a host
protein in SV40-transformed cells. Nature 818, 261.

Lau, L. F. and Nathans, D. (1987). Expression of a set of growth-related
immediate early genes in Balb/c 3T3 cells: Coordinate regulation with c-
fig; and c-my;. Proc. Natl. Acad. Sci. USA 88, 1182-1186.

Lee, W.H., Bookstein, R., Hong, F., Young, L.J., Shew, J.Y., Lee, E.Y.
(1987a). Human retinoblastoma susceptibility'gene: cloning, identification
and sequence. Science 888: 1394-1399.

Lee, W.H., Shew, J.Y., Hong, F.D., Sery, T.W., Donoso, L.A., Young, L.J.,
Bookstein, R., Lee, E.Y. (1987b). The retinoblastoma susceptibility gene
encodes a nuclear phosophoprotein associated with DNA binding activity.
Nature (London) 888: 642-645.

Leof, E. 8., Van Wyc, J. J., O’Keefe, E. J., Pledger, W. J. (1983).
Epidermal growth factor (EGF) is required only during the traverse of
early Gl in PDGF stimulated density-arrested BALB/c-3T3 cells. Exp. Cell.
Res. 181, 202-208.

69

Leaf, E. 8., Proper, J. A., Goustin, A. S., Shipley, G. D., DiCorleto, P.,
and Moses, H. L. (1986). Induction of c-;_; mRNA and platelet derived
growth-like activity by transforming growth factor, type- : a proposed
model for indirect mitogenesis involving autocrine activity. Proc. Natl.
Acad. Sci. USA 88, 2453-2457.

Liboi, E., Pelosi, E., Testa, U., Peschle, C., and Rossi, G. B. (1986).
Proliferative response and oncogene expression induced by epidermal growth
factor in EL2 rat fibroblasts. Mol. Cell. Biol. 8, 2275-2278.

Linzer, D. I. H., Maltzman, W., and Levine, A. J. (1980).
Characterization of a murine cellular SV40 T antigen in SV40-transformed

cells and uninfected embryonal carcinoma. Cold Spring Harbor Symp. Quant.
Biol. 15, 215.

Lobb, R. R., Rybak, S. M., St. Clair, D. K., and Fett, J. W. (1986).
Lysates of two established human tumor lines contain heparin-binding
growth factors related to bovine acidic fibroblast growth factor.
Biochem. Biophys. Res. Commun. 188, 861-867.

Lochrie, M. A., Hurley, J. 8., and Simon, M. I. (1985). Sequence of the
alpha subunit of photoreceptor G protein: Homologies between transducin,
fig;, and elongation factors. Science 888, 96-99.

Lumpkin, C. K., Knepper, J. E., Butel, J. S., Smith, J. R., and Pereira-
Smith, 0. M. (1986). Mitogenic effects of the proto-oncogene and oncogene
forms of c-H-ras DNA in human diploid fibroblasts. Mol. Cell. Biol. 8,
2990-2993.

Maher; V. M., Rowan, L. A., Silinskas, K. C., Kateley; S. A., and
McCormick, J. J. (1982). Frequency of UV-induced neoplastic
transformation of’ diploid human fibroblasts is higher in xeroderma
pigmentosum cells than in normal cells. Proc. Natl. Acad. Sci. USA 18,
2613-1617.

Marquardt, H., Wilson, G. L., and Todaro, G. J. (1980). Isolation and
characterization of a multiplication-stimulating activity (MSA)-like
polypeptide produced by a human fibrosarcoma cell line. J. Biolog. Chem.
888, 9177-9181.

Marquardt, H., Hunkapiller, M. W., Hood, L. E., Twardzik, D. R., De Larco,
J. E., Stephenson, J. R., and Todaro, G. J. (1983). Transforming growth
factors produced by retrovirus-transformed rodent fibroblasts and human
melanoma cells: amino acid sequence homology with epidermal growth
factor. Proc. Natl. Acad. Sci. USA 88, 4684-4688.

Martin, R. G. and Stein, S. (1976). Resting state in normal and Simian
virus 40 trnsformed Chinese hamster lung cells. Proc. Natl. Acad. Sci.
USA 18, 1655-1659.

Massague, J. and Like, 8. (1985). Cellular receptors for type beta

70

transforming growth factor ligand binding and affinity labeling in human
and rodent cell lines. J. Biol. Chem. 888, 3636-3645.

Massague, J. (1985). Subunit structure of a high-affinity receptor for
type-a transforming growth factor: evidence for a disulfide-linked
glycosylated receptor complex. J. 8i80l. Chem. 888, 7059-7066.

Massague, J. (1983). Epidermal growth factor-like transforming growth
factor. J. Biol. Chem. 888, 13614-13620.

Massague, J. (1984). Type a transforming growth factor from feline
sarcoma virus-transformed rat cells, isolation and biological properties.
J. Biol. Chem. 888, 9756-9761.

Matrisian, L. M., Pathak, M., and Magun, B. C. (1982). Identification
of an epidermal growth factor-related transforming growth factor from rat
fetuses. Biochem. Biophys. Res. Commun. 181, 761-769.

McCaffrey, P., Ran, W., Campisi, J., and Rosner, M. R. (1987). Two
independent growth factor-generated signals regulate c-fig; and c-my; mRNA
levels in Swiss 3T3 cells. J. Biol. Chem. 888, 1442-1445.

McCormick, J. J. and Maher, V. M. (1988). Towards an understanding of
the malignantg transformation of diploid human fibroblasts. Mutat. Res.
188, 273-292.

McCormick, J. J., Yang, 0., Maher, V. M., Farber, R. A., Neuman, W.,
Peterson, Jr., W. D., and Pollack, M. S. (1988). The HuT series of
”carcinogen-transformed" human fibroblast cell lines are derived from the
human fibrosarcoma cell line 8387. Carcinogenesis, in press.

McKeehan, W. L., and Ham, R. G. (1978). Calcium and magnesium ions in
the regulation of multiplication in normal and transformed cells. Nature

228), 756-758.

McKeehan, W. L. and Mchehan, K. A. (1979). Epidermal growth factor
modules extracellular Ca'Irequirement for multiplication of normal human
skin fibroblasts. Exp. Cell Res. 188, 397-400.

Mercer, W. 0., Nelson, 0., DeLeo, A. 8., Old, L. J., and Baserga, R.
(1982). Microinjection of monoclonal antibody to protein p53 inhibits
serum-induced DNA synthesis in 3T3 cells. Proc. Natl. Acadl. Sci. USA 18,
6309.

Mihm, M. C. et al. (1971). The clinical diagnosis, classification, and
hisotgenetic concepts of the early stages of cutaneous malignant
melanomas. N. Engl. J. Med. 881, 989.

Milo, G. E. and DiPaolo, J. (1977). 18 yitro transformation of diploid
human cells with chemical carcinogens. In Vitro, 18, 193.

Milo, G. E. and DiPaolo, J. (1978). Neoplastic transformation of human

71

diploid fibroblasts 1g_vitro after chemical carcinogen treatment. Nature,
818, 130-132.

 

Min, W., Xiuquin, W., and Shende, L. (1980). Chemically-induced in vitro
malignant transformation of human esophagus fibroblasts. Sci. Sin 88,
658-662.

Ming, H., Zhihua, W., Xiuqin, W., and Min, W. (1986). Malignant
transformation of human fetal lung fibroblasts induced by nitrosamine
compounds in vitro. Sci. Sin. 88, 1192-1198.

Minuto, F., Del Monte, P., Barreca, A., Fortini, P., Cariola, G.
Catrambone, G., and Giordano, G. (1986). Evidence for an increased
somatomedin-C/insulin-like growth factor 1 content an primary human lung
tumors. Cancer Res. 18, 985-988.

Morgan, J. I. and Curran, T. (1986). Role of ion flux in the control of
c-fig; expression. Nature 888, 552-555,

Moses, H. L., Branum, E. L., Proper, J. A., and Robinson, R. A. (1981).
Transforming growth factor production by chemically transformed cells.
Cancer Res. 11, 2842-2848.

Mukherji, B.,MacAlister, T. J., Guha, A., Gillies, C. G., Jeffers, D. C.,
and Slocum, S. K. (1984). Spontaneous ig vitro transformation of human
fibroblasts. J. Natl. Cancer Inst. 18, 583-587.

Mulcahy, L. S., Smith, M. R., and Stacey, D. M. (1985). Requirement for
fig; proto-oncogene function during serum-stimulated growth of NIH 3T3
cells. Nature 818, 241-243.

Muller, R., Bravo, R., Burckhardt, J., Curran, T. (1984). Induction of
c-fig; and protein preceeds activation of c-myg. Nature 818, 716-720.

Muller, R. (1986). Cellular and viral fig; genes: structure, regulation
of expression and biological properties of their encoded products.
Biochim. Biophys. Acta 888, 207-225.

Murphy, G. F. and Mihm, M. C. (1983). Origin and fate ofpigmented nevi.
In: H. B. Williams (Ed.), Symposium on Vascular Malformations and
Melanocytic Lesions. C. V. Mosby Co., St. Louis, MO, pp. 268.

Namba, M., Nishitani, K., and Kimoto, K. (1978). Carcinogenesis in
tissue culture 29: Neoplastic transformation of a normal human diploid
cell strain, WI-38, with Co-60 gamma rays. Japan J. Exp. Med. 18, 303-
311.

Namba, M., Nishitani, K., and Kimoto, T. (1980). Characteristics of WI-
38 cells (WI-38 CT-I) transformed by treatment with C0-60 gamma rays.
Gann 11, 300-307.

Namba, M., Nishitani, K., Fukushima, F., Kimoto, T., Utsunomiya, J., and

72

Hayflick, L. (1981). Neoplastic transformation of human diploid
fibroblasts treated with chemical carcinogens and Co-60 Y-rays. Gann 81,
221-230.

Namba, M., Kusaka, T., Fukushima, F., and Kimoto, T. (1984).
Differential growth response of normal human diploid fibroblasts and in
vitro transformed human fibroblasts in serum-free defined culture medium.
Int. J. Cancer 88, 645-649.

Namba, M., Nishitani, K., Fukushima, F., Kimoto, T., and Nose, K. (1986).
Multbstep process of neoplastic transformation of normal human fibroblasts
by Cogamma rays and Harvey sarcoma viruses. Int. J. Cancer 81, 419-
423.

Namba, M., Nishitani, K., Fukushima, F., Kimoto, T., and Yuasa, Y.
(1988). Multi-step neoplastic transformation of normal human fibroblasts
by Co-60 gamma rays and H-fig; oncogenes. Mutat. Res. 188, 427-435.

Nau, M. M., Brooks, G. J., Battey, J., Sausville, E., Gazdar, A. F.,
Kirsch, I. R., McBride, 0. W., Bertness, V., Hollis, G. F., and Minna, J.
D. (1985). L-myg, a new myg-related gene amplified and expressed in human
small cell lung cancer. Nature 818, 689-73.

Newbold, R. R., Overell, R. W., and Connell, J. R. (1982). Induction of
immortality is an early event in malignant transformation of mammalian
cells by carcinogens. Nature 888, 633-635.

Nickell, K. A., Halper, J., and Moses, H. L.. (1983). Transforming growth
factors in solid human malignant neoplasms. Cancer Res. 88, 1966-1971.

Nishizuka, Y. (1980). Three multifunctional protein kinase systems in
transmembrane control. Molec. Biol. Biochem. Biophys. 88, 113-135.

Nishizuka, Y. (1986). Studies and perspectives of protein kinase C.
Science 888, 305-312.

O’Brien, W., Stenman, G., and Sager, R. (1986). Suppression of tumor
growth by senescence in virally transformed human fibroblasts. Proc.
Natl. Acad. Sci. USA 88, 8659-8663.

Orkin, S. H., Goldman, D. S., and Sallan, S. E. (1984). Development of
homozygosity for chromosome 11p markers in Wilms’ tumour. Nature 888,
172-175.

Palmer, H., Maher, Y. M., and McCormick, J. J. (1988). Platelet-derived
growth factor or basic fibroblast growth factor induce anchorage-
independent growth of human fibroblasts. J. Cell. Physiol. (in press).

Pantazis, P., Pellicci, P. G., Dalla-Favera, R., and Antoniades, H. N.
(1985). Synthesis and secretion of protein resembling platelet-derived
growth factor by human glioblastoma and fibrosarcoma cells in culture.
Proc. Natl. Acad. Sci. USA 88, 2404-2408.

73

Papageorge, A., Lowy, D., and Scolnick, E. M. (1982). Comparative
biochemical properties of p21 fig; molecules coded for viral and clelular
fig; genes. J. Virol. 11, 509-519.

Parada, L. F., Land, H., Weinberg, R. A., Wolf, 0., and Rotter, V.
(1984). Cooperation between gene encoding p53 tumor antigen and fig; in
cellular transformation. Nature 818, 649.

Parsons, P. G. (1978). Selective proliferation of human tumor cells in
calcium depleted medium. Aust. J. Exp. Biol. and Med. Sci. 88, 297-300.

Pastan, I. and Willingham, M. (1978). Cellular transformation and the
’morphologic phenotype’ of transformed cells. Nature 818, 645-650.

Paterson, H., Reeves, 8., Brown,R., Hall, A., Furth, M., 805, J., Jones,
P., and Marshall, C. (1987). Activated N-fig; controls the transformed
phenotype of HT1080 human fibrosarcoma cells. Cell 81, 803-812.

Paulsson, Y., Hammacher, A., Heldin, C.-H., and Westermark, B. (1987).
Possible positive autocrine feedback in the prereplicative phase of human
fibroblasts. Nature 888, 715-717.

Pavlic, K., Vrbanec, D., MArusic, S., Levanat, S., and Cabrijan, T.
(1986). Autocrine tumour growth regulation by somatomedin C: an in-vitro
model. J. Endocrinol. 188, 233-238.

Peehl, D. M. and Stanbridge, E. J. (1981). Anchorage-independent growth
of normal human fibroblasts. Proc. Natl. Acad. Sci. USA 18, 3053-3057.

Pendergrass, T. W. (1976). Congenital anomalies of children with Wilm’s
tumor. Cancer 81, 403.

Peraina, C., Fry, R. J. M., Staffeld, E., and Christopher, J. P. (1975).
Comparative chancing effects of phenobarbital, aminobarbitol, diphenyl
bydantoin and dichlorodiphenyltrichloro-

ethane on 2-aminofluorene-induced hepatic tumorigenesis in the rat.
Cancer Res. 88, 2884-2890.

Perez-Rodriguez, R., Franchi, A., Deys, B. F., and Pouyssegur, J. (1982).
Evidence that hamster fibroblast tumors emerge in nude mice through the
process of two tg,ijg selections leading to growth factor "relaxation"
and to immune resistance. Int. J. Cancer 88, 309-314.

Persson, H. and Leder, P. (1984). Nuclear localization and DNA binding

properties of a protein expressed by human c-my; oncogene. Science 888,
718.

Peto, R. (1977). Epidemiology, multistage models and short-term
mutagenicity tests. In: The Origins of Human Cancer, Book C. H. H.
Hiatt, J. 0. Watson, and J. A. Winsted (Eds.), Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY, pp. 1403-1437.

74

Pledger, W. J., Stiles, C. D., Antoniades, H. N., and Scher, C. D.
(1977), Induction of DNA synthesis in Balb/c 3T3 cells by serum
components: reevaluation of the commitment process. Proc. Natl. Acad.
Sci. USA, 11, 4481-4485.

Pollack, R., Osborn, M., and Weber, K. (1975). Patterns of organization
of actin and myosin in normal and transformed muscle cells. Proc. Natl.
Acad. Sci. USA 18, 994.

Praeger, F. C. and Cristofalo, V. J. (1986). The growth of WI-38 cells
in a serum-free, growth factor-free, medium with elevated calcium
concentrations. In Vitro Cell. Develop. Biol. 88, 355-359.

Putnam, T. C., Cohen, H. J., and Constine, L. S. (1978). The growth of
WI-38 cells in a serum-free, growth factor-free, medium with elevated
calcium concentrations. In Vitro Cell. Develop. Biol. 88, 355-359.

Ran, W., Dean, M., Levine, R. A., Henkle, C., and Campisi, J. (1986).
Induction of c-fig; and c-my; mRNA by epidermal growth factor or calcium
ionophore is cAMP dependent. Proc. Natl. Acad. Sci. USA 88, 8216-8220.

Rechler, M. M., Podskalny, J. M., Goldfine, I. D., and Wells, C. A.
(1974). J. Clin. Endocrinol. Metab. 88, 512-521.

Rechler, M. M., Zapf, J., Nissley, S. P., Froesch, E. R., Moses, A. C.,
Podskalny, J. M., Schilling, E. E., and Humbel, R. E. (1980).
Interactions of insulin-like growth factors I and II and multiplication
stimulating activity with receptors and serum carrier proteins.
Endocrinology, 181, 1451.

Reich, N. C. and Levine, A. J. (1984). Growth regulation of a cellular
tumor antigen, p53, in non-transformed cells. Nature 888, 199.

Reynolds, Jr., F. H., Todaro, G. J., Fryling, C., and Stephenson, J. R.
(1981). Human transforming growth factors induce tyrosine phosphorylation
of EGF receptors. Nature 888, 259.

Reznikoff, C. A., Bertram, J.S., Brankow, D. W., and Heidelberger, C.
(1973). Quantitative and qualitative studies of chemical transformation
of clones of C3H mouse embryo cells sensitive to post confluence
inhibition of cell division. Cancer Res. 88, 3239-3249.

Rhim, J. S., Park, 0. K., Arnstein, P., and Nelson-Rees, W. A. (1975).
Transformation of human cells in culture by N-methyl-N’-nitro-N-
nitrosoguanidine. Nature 888, 751-753.

Rijnders, A. W. M., van der Korput, J. A. G. M., van Steenbrugge. G. J.,
Romijn, J. C., and Trapman, J. (1985). Expression of cellular oncogenes
in human prostatic carcinoma cell lines. Biochem. Biophys. Res. Comm.
188, 548-554.

7S

Rinderknecht, E. and Humbel, R. E. (1978a). The amino acid sequence of
human insulin-like growth factor I and its structural homology with
proinsulin. J. Biol. Chem., 888, 2769.

Rinderknecht, E. and Humbel, R. E. (1978b). Primary structure of human
insulin-like growth factor II. FEBS Lett. 88, 283.

Risau, W. and Ekblom, P. (1986). Production of a heparin-binding
angiogenesis factor by the embryonic kidney. J. Cell Biol., 188, 1101-
1107.

Risau, W. (1986). Developing brain produces an angiogenesis factor.
Proc. Natl. Acad. Sci. USA, 88, 3855-3859.

Robbins, K. D., Leal, F., Pierce, J. H., and Aaronson, S. A. (1985). The
v-;_;/PDGF-2 transforming gene product localizes to cell membranes but is
not a secretory protein. EMBO J. 8, 1783-1792.

Robbins, K. C., Antoniades, H. N., Devare, S. G., Hunkapiller, M. W., and
Aaronson, S. A. (1983). Structural and imunological similarities
between simian sarcoma virus gene product(s) and human platelet-derived
growth factor. Nature 888, 605-608.

Roberts, A. B., Anzano, M. A., Lamb, L. C., Smith, J. M., and Sporn, M.
B. (1981). New class of transforming growth factors potentiated by
epidermal growth factor: isolation from non-neoplastic tissues. Proc.
Natl. Acad. Sci. USA 18, 5339-5343.

Roberts, A. 8., Anzano, M.A., Lamb, L. C., Smith, J. M., Frolik, C. A.,
Marquardt, H., Todaro, G. J., and Sporn, M. B. (1982). Isolation from
murine sarcoma cells of a new class of transforming growth factors
potentiated by epidermal factor. Nature 888, 417-419.

Roberts, A. 8., Anzano, M. A., Wakefield, L. M., Roche, N. S., Stern, D.
F., and Sporn, M. B. (1985). Type b transforming growth factor: a
bifunctional regulator of cellular growth. Proc. Natl. Acad. Sci. USAl88,
119-123.

Rosenthal, A., Lindquist, P. 8., Bringman, T. S., Goeddel, D. V., and
Derynck, R. (1986). Expression in rat fibroblasts of a human
transforming growth factor-a cDNA results in transformation. Cell 88,
301-309.

Ross, R. and Vogel, A. (1978). The platelet-derived growth factor. Cell
18 203-210.

Ross, R., Glomset, J., Kariya, 8., and Harker, L. (1974). A platelet-
dependent serum factor that stimulates the proliferation of arterial
smooth muscle cells in vitro. Proc. Natl. Acad. Sci. USA, 11, 1207-1210.

Ross, R., Raines, E. W., and Bowen-Pope, D. F. (1986). The biology of
platelet-derived growth factor. Cell 88, 155-169.

76

Rotter, V. and Wolf, 0. (1985). Biological and molecular analysis of p53
cellular-encoded tumor antigen. Adv. Cancer Res. 88, 113.

Rubin, H. (1988). Personal communication to J. J. McCormick.

Ruley, H. E. (1983). Adenovirus early region 1A enables viral and
cellular transforming genes to transform primary cells in culture. Nature
888, 596-602.

Sack, Jr., G. H., (1981). Human cell transformation by simian virus 40:
A review. In Vitro 11, 1-19.

Sambucetti, LC. and Curran, T. (1986). The fos protein complex is
associated with DNA in isolated nuclei and binds to DNA cellulose.
Science 881, 1417-1419.

Sanford, K. K., Handelman, S. L., and Fox, C. H. (1974). Effects of
chemical carcinogens on neoplastic transformation and morphology of cells
in culture. J. Natl. Cancer Inst. 88, 1647-1659.

Santos, E., Tronick, S. R., Aaronson, S. A., Pulciani, S., and Barbacid,
M. (1982). T24 human bladder carcinoma oncogene is an activated form of
the normal human homologue of BALB-c and Harvey-MSV transforming genes.
Nature 888, 343-347.

Sassone-Corsi, P., Lamph, W. W., Kamps, M., and Verma, I.M. (1988). fig;-
associated cellular p39 is related to nuclear transcription factor AP-l.
Cell 81, 553-560.

Sato, M., Yoshida, H., Hayashi, Y., Miyakami, K., Bando, T., Yanagawa, T.,
Yuva, Y., Azuna, M., and Veno, A. (1985). Expression of epidermal growth
factor and transforming growth factor-b in a human salivary gland
adenocarcinoma cell line. Cancer Res. 18, 6160-6167.

Scher, C. D., Pledger, W. J., Martin, P., Antoniades, H., and Stiles, C.
D. (1978). Transforming viruses directly reduce the cellular growth
requirement for a platelet derived growth factor. J. Cell. Physiol. 81,
371-380.

Schmidt, C., Steiner, T., and Froesch, E. R. (1984). Insulin-like growth
factor 1 supports the differentiation of cultured osteoblast-like cells.
FEBS Lett. 118, 48.

Schreiber, A. 8., Winkler, M. E., and Derynk, R. (1986). Transforming
growth factor a is a more potent angiogenic mediator than epidermal
growth factor. Science 888, 1250-1253.

Scully, R. E. (1981). Definitions of precursors in gynecologic cancer.
Cancer 18, 531.

Seeger, R., Brodeur, G., Sather, H., Dalton, A., Siegel, S., Wong, K., and

77

Hammond, 0. (1985). Association of multiple copies of the N-my; oncogene
with rapid progression of neuroblastomas. N. Engl. J. Med. 818, 1111-
1116.

Shein, H. M., and Enders, J. F. (1962). Transformation induced by SV40
in human renal cell cultures. Proc. Natl. Acad. Sci. USA 88, 1164-1172.

Shier, W. T. (1980). Serum stimulation of phospholipase A2 (and
prostaglandin release in 3T3 cells is associated with platelet-derived
growth factor-promoting activity. Proc. Natl. Acad. Sci. USA, 11, 137-
141.

Shih, T., Weeks, M., Gruss, P., Dhar, R., Oroszian, S., and Scolnick, E.
(1982). Identification of a precursor in the biosynthesis of the p21
transforming protein. J. Virol. 18, 253-261.

Shipley, G. 0., Tucker, R. F., and Moses, H. L. (1985). Type-b
transforming growth factor/growth inhibitor stimulates entry of'monolayer
cultures of AKR-2B cells into S phase after a prolonged prereplicative
interval. Proc. Natl. Acad. Sci. USA 88, 4147-4151.

Silinskas, K. C., Kateley, S. A., Tower, J. T., Maher, V. M., and
McCormick, J. J. (1981). Induction of anchorage-independent growth in
human fibroblasts by propane sultone. Cancer Res. 81, 1617-1620.

Silverberg, S. G. (1984). Cancer statistics. CA 88, 7.

Singh, J. P., Chaikin, M. A., Pledger, W. J., Scher, C. 0., and Stiles,
C. D. (1983). Persistence of the mitogenic response to platelet-derived
growth factor (competence) does not reflect a ‘long-term interaction
between the growth factor and the target cell. J. Cell Biol. 88, 1497-
1502.

Sorrentino, V., Drozdoff, V., McKinney, M. D., Zeitz, L., and Fleissner,
E. (1986). Potentiation of growth factor activity by exogenous c-my;
expression. Proc. Natl. Acad. Sci. USA 88, 8167-8171.

Sorsby, A. (1972). Bilateral retinoblastoma: a dominantly inherited
affection. Br. Med. J. 8, 580-583.

Spandidos, D. A. and Wilke, N. M. (1984). Malignant transformation of
early passage rodent cells by a single mutated human oncogene. Nature
(London), 818, 469-475.

Spandidos, D. A. (1985). Mechanism of carcinogenesis: The role of
oncogenes, transcriptional enhancers and growth factors. Anticancer Res.
8, 485-498.

Sporn, M. 8., and Todaro, G. J. (1980). Autocrine secretion and
malignant transformation of cells. N. Engl. J. Med. 888, 878-880.

Sporn, M. B. and Roberts, A. B. (1985). Autocrine growth factors and

78
cancer. Nature 818, 745-747.

Stevens, C. W., Manorahan, T. H., and Fahl, W. E. (1988).
Characterization of mutagen-activated oncogenes that confer anchorage
independence to human fibroblasts and tumorigenicity to NIH 3T3 cells:
Sequence analysis of an enzymatically amplified mutant H-fig; allele.
Proc. Natl. Acad. Sci. USA 88, 3875-3879.

Stewart, T. P., Pattengale, P., and Leder, P. (1984). Spontaneous
mammary adenocarcinomas in transgeneic:mice that carry and express MTV/my;
fusion genes. Cell 88, 627.

Stiles, C. D., Capone, G. T., Scher, C. D., Antoniades, H. N., Van Wyk,
J. J., and Pledger, W. J. (1979). Proc. Natl. Acad. Sci. USA, 18, 1279-
1283.

Stone, J., Lang, T., Ramsay, G., Jakobovits, E., Bishop, J. M., Varmus,
H., and Lee, W. (1987). Definitions of regions in human c-my; that are
involved in transformation and nuclear localization. Mol. Cell. Biol. 1,
1697-1709.

Suarez, H.G., Nardeux, P. C., Andeol, Y., and Sarasin, A. (1987).
Multiple activated oncogenes in human tumors. Oncogene Res. 1, 210-207.

Sukhatme, V. P., Kartha, S., Toback, F. G., Taub, R., Hoover, R. G., and
Tsai-morris, C. H. (1987). A novel early growth response gene rapidly
induced by fibroblast, epithelial cell and lympocyte mitogens. Oncogene
Res. 1, 343-355.

Sutherland, B. M., Cimino, J., Delihas, N., Shih, A., and Oliver, R. P.
(1980). Ultraviolet light-induced transformation of human cells to
anchorage-independent growth. Cancer Res. 88, 1934-1939.

Tatchell, K., Chaleff, D. T., DeFeo-Jones, D., and Scolnick, E. M.
(1984). Requirement of either of a pair of fig;-related genes of
8g;;8gfigmyces cerevisiae for spore viability. 888, 523-527.

Taub, R., Kirsch, I., Morton, C., Lenoir, G., Swan, 0., Tronick, S.,
Aaronsom, S., and Leder, P. (1982). Translocation of the c-my; gene into
the immunoglobulin heavy chain locus in human Burkitt lymphoma and murine
plasmacytoma cells. Proc. Natl. Acad. Sci. USA, 18, 7837-7841.

Taylor, J. M., Mitchell, W.H., and Cohen, S. (1972). Epidermal growth
factor: physical and chemical properties. J. Biol.Chem. 881, 5928-5934.

Thielmann, H. W., Fischer, E., Dzarlieva, R. T., Komitowski, D., Popanda,
J., and Edler, L. (1983). Spontaneous in vitro malignant transformation
in a xeroderma pigmentosum fibroblast line. Int. J. Cancer 81, 687-700.

Thomas, K. C., Rios-Candelore, M., and Fitzpatrick, S. (1984).
Purification and characterization of acidic fibroblast growth factor from
bovine brain. Proc. Natl. Acad. Sci. USA 81, 357-361.

79

Thomas, K. A. (1987). Fibroblast growth factors. FASEB J. 1, 434-440.

Thomassen, D. G., Gilmer, T. M., Annab, L. A., and Barrett, J. C. (1985).
Evidence for multiple steps in neoplastic transformation of normal and
preneoplastic Syrian hamster embryo cells following transfection of Harvey
murine sarcoma virus oncogene (v-Ha-fig;). Cancer Res. 88, 726-732.

Thompson, K. V. A. and Holliday, R. (1975). Chromosome changes during
the in yttfig ageing of MRC-5 human fibroblasts. Exp. Cell. Res. 88, 1-6.

Toda, T., Uno, 1., Ishikawa, T., Powers, S., Kataoka, T., and Broek, 0.,
Cameron, 5., Broach, J., Matsumoto, K., and Wigler, M. (1985). Cell 18,
27-36.

Todaro, G. J. and De Larco, J. E. (1978). Growth factors produced by
sarcoma virus-transformed cells. Cancer Res. 88, 4147-4154.

Todaro, G. J., Fryling, C., and DeLarco, J. E. (1980). Transforming
growth factors produced by certain human tumor cells: polypeptides that
interact with epidermal growth factor receptors. Proc. Natl. Acad. Sci.
USA 11, 5258-5262.

Tubo, R. A., and Rhinewald, J. G. (1987). Normal mesothelial cells and
fibroblasts transfected with the EJ-fig; oncogene become EGF-independent
but are not malignantly transformed. Oncogene Res. 88, 4147-4154.

Tucker, R. F., Shipley, G. D., Moses, H. L., and Holley, R. W. (1984).
Growth inhibitor from BSC-I cells closely related to the platelet type
transforming growth factor. Science 888, 705-707.

Twardzik, D. R., Ranchalis, J. E., and Todaro, G. H. (1982). Mouse
embryonic transforming growth factors related to those isolated from tumor
cells. Cancer Res. 18, 590-593.

Ushiro, H. and Cohen, S. (1980). Identification of phosphotyrosine as
a product of epidermal growth factor-activated protein kinase in A431 cell
membranes. J. Biol. Chem. 888, 8363.

Van Beveren, C., Van Straaten, F., Curran, T., Muller, R., and Verma, I.
M. (1983). Analysis of FBJ-MuSV provirus and c-fig; (mouse) gene reveals
that viral and cellular fig; gene products have different carboxy termini.
Cell 88, 1241-1245.

Van Obberghen-Schilling, E., Perez-Rodriguez, R., Franchi, A., Chambard,
J. C., and Pouyssegur, J. (1983). Analysis of growth factor "relaxation
in Chinese hamster lung fibroblasts required for tumoral expression. J.
Cell. Physiol. 118, 123-130.

Van Zoelen, E. J. J., van de Ven, W. J. M., Franssen, H. J., van
Oosterward, T. M. J., van der Saag, P. T., Heldin, C.-H., and de Laat, S.
W. (1985). Neuroblastoma cells express c-;;; and produce a transforming

80

growth factor antigenically related to the platelet-derived growth factor.
Mol. Cell. Biol. 8, 2289-2297.

Van Zoelen, E. J. J., Van Oostwaard, T. M. J., vander Saag, P. T., and de
Laat, S. W. (1985). Phenotypic transformation of normal rat kidney cells
in a growth factor defined medium: induction by a neuroblastoma-derived
transforming growth factor independent of the EGF receptor. J. Cell.
Physiol. 188, 151-160.

Varmus, H. E. (1984). The molecular genetics of cellular oncogenes. Ann.
Rev. Genet. 18, 553-612.

Vasiliew, J. M. and Gelfand, I. M. (1981). Neoplastic and normal cells.
Cambridge University Press, New York, NY, pp. 160-272.

Versnel, M. A., Hagemeijer, A., Bouts, M. J., van der Kwast, T. H.,
Hoogsteden, H. C. (1988). Expression of c-;1; (PDGF B-chain) and PDGF A-
chain genes in ten human malignant mesothelioma cell lines derived from
primary and metastatic tumors. Oncogene 8, 601-605.

Wakelam, M., Davies,S.,ik11w_l'a{%M., McKay, 1., Marshall, C. J., and
Hall, A. (1986). Normal p21uples bombesin and other growth factor
receptors to inositol phosphate production. Nature 888, 173-176.

gang,Y ,Kateley- -Kohler, S, Maher, V. M. , and McCormick, J. J. (1986).

C0 radiation- induced transformation to anchorage independence of
fibroblasts from normal persons and patients with inherited predisposition
to retinoblastoma. Carcinogenesis 1, 1927-1929.

Weinstein, I. 8., Wigler, M., and Stadler, U. (1976). Analysis of the
mechanism of chemical carcinogenesis of epithelial cell cultures. In:
IHR/WHO Scientific Publication No. 12, Screening tests in chemical
carcinogenesis. R. Montesano, H. Bartsch, and L. Tomatis (Eds.).

Westermark, B. and Heldin, C.-H. (1985). Similar action of platelet-
derived growth factor and epidermal growth factor in the prereplicative
phase of human fibroblasts suggests a common intracellular pathway. J.
Cell. Physiol. 188, 43-48.

Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H., Raybuck, M.,
Weinberg, R. A., and Harlow, E. (1988). Association between an oncogene
and an anti-oncogene: The adenovirus E1a proteins bind to the
retinoblastoma gene product. Nature 888, 124-129.

Wilson, 0. M., Maher, V. M., and McCormick, J. J. (1987). Transformation
of normal diploid human fibroblasts by a transfected N-fig; oncogene.
Proc. Amer. Assoc. Cancer Res. 88, 152 (Abstract).

Yamada, K. M. and Olden, K. (1978). Fibronectin-adhesive glycoproteins
of cell surface and blood. Nature 818, 179-184.

Yoshida, T., Miyagawa, K., Odagiri, H., Sakamoto, H., Little, P. F. R.,

81

Terada, M., and Sugimura, T. (1987). Genomic sequence of m, a
transforming gene encoding a protein homologous to fibroblast growth
factors and the tgt-Z-encoded protein. Proc. Natl. Acad. Sci. USA 81,
7305-7309.

Yunis, J. J. (1983). The chromosomal basis of human neoplasia. Science

881, 227.

Zapf, J., Waldvogel, M., and Froesch, E. R. (1975). Binding of non-
suppressible insulin-like activity to human serum: evidence for a carrier
protein. Arch. Biochem. Biophys. 188, 638.

Zhan, X. and Goldfarb, M. (1986). Growth factor requirements of
oncogene-transformed NIH 3T3 and BALB/c 3T3 cells cultured in defined
media. Mol. Cell. Biol. 8, 3541-3544.

82

CHAPTER II - MODIFICATION OF A SERUM-FREE MEDIUM TO ASSAY THE MITOGENIC
EFFECTS OF GROWTH FACTORS OF DIPLOID HUMAN FIBROBLASTS IN CULTURE.

Introduction

Serum contains a variety of mitogenic substances which have been
shown to affect the growth of cells in culture (Campisi et al., 1984). For
this reason, a medium which can support the long term growth of
fibroblasts in the absence of serum is required to examine the action of
single purified growth factors which may be present in serum in nanogram
amounts. A variety of such media have been designed (Barnes and Sato,
1980), but often they fail to support the long term growth of cells or
colony formation by cells plated at low density. Failure to satisfy these
more stringent conditions implies that such a medium is inadequate
compared to serum-supplemented medium. Ryan et al. (1987) described a
serum-free medium which is able to support both long term growth and a
high efficiency of colony formation by diploid human fibroblasts. Such a
medium is ideal for testing the growth promoting action of purified growth

factors.

Materials and Methods

Cells and Culture Medium. Prior to assay, normal diploid human
fibroblast cell line SL 68 was routinely maintained in Eagle’s minimal
essential medium supplemented with 0.2 mM serine, 0.2 mM aspartate and 1.0

mM pyruvate, 10% fetal bovine serum and antibiotics (100 U/ml penicillin

83
and 100 ug/ml streptomycin) at 37°C in a humidified atmosphere of 5% C02
and air.

Growth Factors. Epidermal growth factor (EGF) was obtained from
Bethesda Research Laboratories, Gaithersburg, MD and platelet-derived
growth factor (PDGF) was purchased from PDGF, Inc., Boston, MA. These
peptide growth factors were dissolved in an aqueous solution of purified
human serum albumin (5 mg/ml) (Sigma Chemical Co., St. Louis, M0) to make
stock solutions of EGF and PDGF with a concentration of 10 mg/ml, which
were prepared fresh weekly and stored at 4°C.

Growth Factor Assay Medium. Cells were tested for growth factor and
calcium requirements using as a serum-free base medium, McM medium, the
modification of the MCDBIIO base medium of Bettger gt g1. (1981), which
was adapted by Ryan et al. (1987) for use with various serum replacement
supplements. For ‘the initial series of' experiments, EGF, insulin,
dexamethasone, or prostaglandin E1 were omitted from the standard
supplements described by Ryan et al. (1987). For the rest of the
experiments, the test medium, designated growth factor assay medium,
consisted of the base medium McM containing 0.1 mM calcium instead of the
usual 1.0 mM and the serum replacement supplements of Ryan gt g1. (1987)
with the omission of EGF and insulin.

Assay of Cells for Growth Requirements. Prior to being tested in
this growth factor assay medium, the cells were adapted to the base
medium, McM, by being grown for at least one week in that base medium
supplemented with 10% fetal bovine serum and antibiotics. The cells were
then suspended in the serum-free growth factor assay medium and plated

into a series of 60-mm diameter dishes at a density of 0.7 to 2.2 x 103

84
cells/cmz. After 12-16 hours the medium was removed and freshly prepared
assay medium with the designated modifications was added to the cells.
Every three to four days, the medium was replaced with freshly prepared
test medium. The increase in cell number in duplicate dishes was

monitored by electronic cell counting (Coulter Electronics, Hialeah, FL).

Results and Discussion

The serum-free medium described by Ryan et al. (1987) contains both
EGF and insulin. Since both of these peptides are reported to stimulate
the replication of human fibroblasts (Westermark.and Heldin, 1985; Praeger
and Cristofalo et al. 1986), they were omitted from the growth factor
assay medium and cells were assayed for the ability to replicate in this
modified medium. As shown in Table 1, there was a reduction in final cell
number for the cells maintained in this modified medium compared to that
for cells maintained in the original medium of Ryan et al. (1987), but the
cells still increased in number. Since Ryan et al. (1987) had demonstrated
that the various components of the medium, other than EGF or insulin were
free of endogenous growth factor activity, contamination of peptide growth
factor was not likely to be the cause of the observed replication. A
second possibility was that dexamethasone or prostaglandin E1 was causing
the observed replication in the absence of exogenously added growth
factors. As shown in Table 2, omission of these substances from the
medium did not affect the increase in cell number during the six day
period assayed.

A 1985 report by Heldin et al. showed that the concentration of

calcium in medium needed to be reduced from 1.0 mM to 0.5 mM if one were

85

to be able to detect the ability of either EGF or PDGF to stimulate DNA
synthesis by diploid human fibroblasts. Therefore, the level of calcium
in the McM base medium was reduced and the medium was tested for its
ability to support growth. As shown in Figure 1, a ten-fold decrease in
calcium concentration to 0.1 It“ was required before replication was
prevented. Levels of calcium between 0.25 and 1.1 it“ showed a dose
dependent increase in the rate of replication. The maximal rate of
replication occurred in cells maintained in 1.1 mM calcium. Doubling this
concentration to 2.1 mM did not increase the rate of replication.

It was always possible that a concentration of calcium as low as 0.1
mM would also prevent cells from responding to the mitogenic effects of
growth factors. Therefore, purified growth factors were added to the
serum-free medium containing 0.1 mM calcium. Figures 2 and 3 show the
dose»dependent stimulation of replication of diploid human fibroblast cell
line SL 68 in this low calcium medium supplemented with either EGF or
PDGF. EGF at a final concentration of 30 ng/ml did not increase the rate
of replication of cells over that seen with 3 ng/ml of EGF.
Supplementation of the medium with PDGF caused a dose dependent increase
in rate of replication with no evidence of saturation even at the highest
dose of 20 ng/ml. Higher levels of PDGF were not tested because of the
prohibitive cost of purified growth factor, but there is evidence from
Westermark and Heldin (1985) that a maximum level of mitogenic activity
can be reached. Addition of insulin to the base medium at final
concentrations as high as 100 ug/ml did not stimulate cell replication
(Figure 4).

A ten-fold reduction of the concentration of calcium in the serum-

86

free medium, from 1.0 nfl to 0.1 M, was necessary in order to stop
replication of diploid human fibroblasts. This level of calcium, however,
did not prevent stimulation of growth by the peptide growth factors, EGF
and PDGF. These results are consistent with those reported by Westermark
and Heldin (1985) who observed stimulation of DNA synthesis in quiescent
diploid human fibroblasts by either EGF or PDGF. These investigators also
reported that the presence of insulin did not affect DNA synthesis.
However, in contrast to the present results, Westermark and Heldin (1985)
reported that a two-fold reduction of calcium from 1.0 mM to 0.5 mM was
required to detect the effect of PDGF and EGF. In our study, the calcium
concentration had to be reduced ten-fold. One explanation for this
disparity is the difference in the culture conditions and the assay
conditions. Westermark and Heldin (1985) used confluent fibroblasts to
examine the ability of growth factors to stimulate DNA synthesis over a
48 hour period. The present study measured the increase in cell number
by sparsely plated fibroblasts maintained for in 6-8 days in medium
containing various growth factors. It may be that calcium has less effect
on contact-inhibited human cells than on sparsely plated cells.

An alternate explanation for the difference in the decrease in level
of calcium required to prevent cell replication is that the medium used
by Westermark and Heldin (1985) was sub-optimal and therefore could not
support the long term growth of human fibroblasts. If this were the case,
one might expect that a mitogen such as calcium would be less effective
than it would be in an optimal medium and, therefore, a two-fold reduction
in calcium might be sufficient to prevent cell replication.

The mitogenic effect of calcium shown in Figure 1 was subsequently

87

reported by Praeger and Cristofalo (1986). They used a serum-free medium
to show that calcium alone, in the absence of exogenously added growth
factors, was able to cause the replication of diploid human fibroblasts
in culture. Calcium concentrations from 1.0 mM to 5.0 mM were shown to
cause replication of sparsely plated cells. The nfitogenic effect of
calcium reported by Praeger and Cristofalo is qualitatively similar to
what was observed in the present study, but somewhat higher concentrations
of calcium were required. Praeger and Cristofalo (1986) did not show that
the serum-free medium which they used could support the long term growth
of cells. Since their medium lacked the attachment factors or lipids
which have been shown to be necessary to support optimal long term growth
(Ryan et al., 1987), conditions may have been suboptimal and, therefore,
higher concentrations of calcium were required to stimulate cell growth.

In summary, modifying the medium of Ryan et al. (1987) by omitting
insulin and EGF and decreasing the concentration of calcium ten-fold
allows one to assay diploid human fibroblasts for the effects of various
protein growth factors. Such fibroblasts remained viable in this medium
for at least three weeks (data not shown) and were able to resume
replication at that time in response to EGF and replicate at the same rate

as cells freshly plated into such medium.

88
References.

Barnes, 0. and Sato, G. (1980). Methods for growth of cultured cells in
serum-free medium. Anal. Biochem., 188: 255-270.

Bettger, W.J., Boyce, S.T., Walthall, B.J., and Ham, R.G. (1981). Rapid
clonal growth and serial passage of human fibroblasts in a lipid-enriched
synthetic medium supplemented with epidermal growth factor, insulin, and
dexamethasone. Proc. Natl. Acad. Sci. USA, 18: 5588-5592.

Campisi, J., Morreo, G. and Pardee, A.8. (1984). Kinetics of G1 transit
following brief starvation for serum factors. Exp. Cell Res., 188: 459-
466.

Praeger, F.C. and Cristofalo, V.J. (1986).The growth of WI-38 cells in
a serum-free, growth factor free medium with elevated calcium
concentrations. In Vitro Cell. and Dev. Biol., 82(6): 355-359.

Ryan, P.A., Maher, V.M. and McCormick, J.J. (1987). Modification of MCDB
110 medium to support prolonged growth and consistent high cloning
efficiency of diploid human fibroblasts. Exp. Cell Res., 118: 318-328.

Westermark, B. and Heldin, C. (1985). Similar action of platelet derived
growth factor and epidermal growth factor in the prereplicative phase of

human fibroblasts suggests a cannon intracellular pathway. J.
Cell.Physiol., 181: 43-48.

89

Table 1 - Effect of EGF and insulin on the growth of normal human
fibroblast cell line SL 68 in serum-free medium.

 

Cells per dish

 

Medium composition Day 1 Dgy 3 Day 5
McM + snll 21,552 106,870 538,440
McM + SR1, 41152 20,096 129,930 332,592
McM + SR1, -EGF3 21,168 44,760 101,024
McM + SR1, -INS, -EGF 18,912 30,660 53,213

 

l'McM + SR] is the serum free medium as described by Ryan gt g1, (1987).

2 INS is insulin, normally present at final concentration of 1
ug/ml in McM + SR1.

3 EGF is epidermal growth factor, normally present at a final
concentration of 3 ng/ml in McM + SR].

90

Table 2 - Effect of dexamethasone and prostaglandin E]_on the growth
of normal human fibroblast cell line SL 68 in serum-
free medium.

 

Cells per dish

 

Medium composition Day 3 Day 5
McM + SR11,-INS2, —EGF3 440,657 635,528
McM + SR1, -INS, -EGF, on“ 459,442 642,736
McM + SR1, -INS, -EGF, 10:5 421,345 580,040
McM + SR1, -INS, 4507:, on, -PGE 408,068 596,207

 

1McM + SR] is the serum free medium as described by Ryan gt g1.
(1987).

2 INS is insulin, normally present at a final concentration of l
ug/ml in McM + SR].

3 EGF is epidermal growth factor, normally present at a final
concentration of 3 ng/ml in McM + SR1.

“DEX is dexamethasone, normally present at a final concentration
of 0.2 ug/ml in McM + SR1.

5 PGE is prostaglandin E1, normally present at a final concentration of
10 ng/ml in McM + SR1.

91

Figure 1. Effect of medium calcium concentration on the replication of
diploid human fibroblast cell line SL 68. Cells were plated into 60 mm
culture dishes in 12 - 16 hours prior to determining the initial cell
numbers on Day 0. Plating medium was then replaced with growth factor
assay medium containing calcium at a final concentration of 0.1 mM (e),
0.15mM (A), 0.25 11114 (a ), 0.48 mM (A), 0.60 mM(<7),1.1mM(o), or 2.1
mM he). The media were replaced with freshly prepared test media every 3-

4 days. Each point represents the average of duplicate dishes.

92

 

 

 

11 111111111.—

 

0 L0

(tom HSIG 838 STIBO

 

DAYS

93

Figure 2. Effect of medium EGF concentration on the replication of
diploid human fibroblast cell line SL 68. Cells were plated into 60 mm
culture dishes 12 - 16 hours prior to determining the initial cell numbers
on Day 0. Plating medium was then replaced with growth factor assay medium
containing EGF at a final concentration of 0 ng/ml (O), 0.03 ng/ml (A),
0.30 ng/ml (D), 3.0 ng/ml (O), or 30.0 ng/ml (o). The media were
replaced with freshly prepared test media every 3-4 days. Each point

represents the average of duplicate dishes.

94

 

 

 

 

II I I‘lllll I 1
OD <1 0 -co
0 E1 <le
\\ \I .
<3 -v
- \\ H
\C'\ if
\C€'\<\J’ _N
(Bil
ll 1 C|>HllKleI l _O

 

0
L0

(tom HSICI 833 STE-1:)

95

Figure 3. Effect of medium PDGF concentration on the replication of
diploid human fibroblast cell line SL 68. Cells were plated into 60 mm
culture dishes 12 - 16 hours prior to determining the initial cell numbers
on Day 0. Plating medium was then replaced with growth factor assay medium
containing PDGF at a final concentration of 0 ng/ml (er), 2 ng/ml on), 5
ng/ml (:3), 10 ng/ml (Q), or 20 ng/ml (o). The media were replaced with
freshly prepared test media every 3-4 days. Each point represents the

average of duplicate dishes.

96

 

 

 

 

liilllll

 

 

O _

0
L0

(toe) HSlCl 83d S'l'lBO

97

Figure 4. Effect of medium insulin concentration on the replication of
diploid human fibroblast cell line SL 68. Cells were plated into 60 mm
culture dishes 12 - 16 hours prior to determining the initial cell numbers
on Day 0. Plating medium was then replaced with growth factor assay medium
containing insulin at a final concentration of 0 ug/ml (O), 0.1 ug/ml

(A), 1 ug/ml (Ci), 10 ug/ml (I), or 100 ug/ml (A). The media were
replaced with freshly prepared test media every 3-4 days. Each point

represents the average of duplicate dishes.

98

 

Fill ”Iii:

 

 

Ill lLlllll

l

(93

 

IO

 

 

O 0
LC

07.000 HSIG 838 S'l‘lElO

99

CHAPTER III

OVEREXPRESSION OF MULTIPLE GROWTH FACTOR GENES
BY FIBROSARCOMA-DERIVED AND OTHER TRANSFORMED HUMAN
FIBROBLASTS IN CULTURE1

Robert J. Schilz, John E. Dillberger, Veronica M. Maher,

and J. Justin McCormick2

Carcinogenesis Laboratory - Fee Hall
Department of Microbiology and Department of Biochemistry

Michigan State University, East Lansing, MI 48824

100

1This research was supported in part by Department of Energy Grant DE-
FG02-76-ER60524 and Contract NOI-ES-65152, by Department of Health and
Human Services Grant CA21289 from the National Cancer Institute and by a
grant from the Michigan Osteopathic College Foundation.
2T0 whom all correspondence should be addressed:

Dr. J. Justin McCormick

Carcinogenesis Laboratory-Fee Hall

Michigan State University

East Lansing, MI 48824-1316

517/353-7785
3The abbreviations used are: EGF, epidermal growth factor; HEPES, N—2-
Hydroxyethylpiperazine-N’-2-ethanesulf0nic acid; McM, a base nutrient
medium (21); MOPS, 3-[N-morpholino]pr0panesulfonicacid; PBS, phosphate
buffered saline; PDGF, platelet-derived growth factor; SOS, sodium'dodecyl
sulfate; SV40, Simian virus 40; SSC, 1X SSC is 0.15 M NaCl, 0.015 M sodium

citrate; FGF, fibroblast growth factor; IGF, insulin-like growth factor.

101
Abstract

Acquisition of the growth factor independent phenotype has been
reported to accompany the transformation of several rodent fibroblast cell
lines. We studied the growth factor requirements of normal, tumor-derived,
and other transformed human fibroblasts, as well as their response to
added growth factors, to determine whether this phenotype is
characteristic of such transformed cells. Analysis of the growth of these
cell lines in a serum-free medium showed that all transformed human
fibroblasts were able to grow in the absence of exogenously added protein
growth factors. The rate of growth of these cell lines in the growth
factor assay medium was correlated with their rate of tumor formation in
athymic mice. Northern analysis of cellular RNA from normal and
transformed cell lines for the expression of the A-chain and B-chain of
platelet derived growth factor, transforming growth factors a and b,
acidic and basic fibroblast. growth factor, epidermal growth factor,
insulin-like growth factor II, N-fig;, K-fig;, my;, and fig; genes was
carried out. The results showed that the malignant cell lines which
replicated rapidly jg yitfig and jg_yiyg, and failed to increase their rate
of replication in response to exogenously added growth factors,
overexpressed transcripts of several protein growth factors. The set of
genes which was overexpressed, however, was not identical for each cell
line. Our results support the hypothesis that the acquisition of growth
factor independence can account for the abnormal proliferation of

tumorigenic fibroblasts and indicate that the specific changes involved

in achieving this characteristic can differ. Our data also show that

102
acquisition of growth factor independence, per se, was not sufficient to

cause the malignant transformation of human fibroblasts.

103

Introduction

The development of cancer in humans is generally recognized to
result from a multistep process (1-3). One approach to elucidating the
steps involved is to study the process of malignant transformation of
human cells in culture since the changes necessary for malignant
transformation can be expected to occur by similar mechanisms 1_ v v0 and

i vitro. For historical reasons fibroblasts have been the cell type most

 

commonly employed for H1 ill—127:8 transformation experiments. Studies
involving diploid human fibroblasts have demonstrated that a single
exposure of cells to carcinogen does not result in malignant
transformation (for review, see ref. 4), and even repeated exposure of
cultures to carcinogen has failed to yield tumorigenic cells (5). Recent
experiments involving the transfection of oncogenes into fibroblasts in
culture, such as the study by Thomassen gt g1. (6) or Land gt g1. (7),
demonstrate that multiple cellular changes must occur before a cell
becomes malignant. The implication of such studies is that successful
dissection of the steps involved in the malignant transformation may
require introducing a detectable cellular change, isolating cells which
have acquired that. change, and then introducing additional cellular
changes in a stepwise fashion until a malignant cell is obtained.

One of these critical changes may be acquisition of growth factor
independence. Both benign and malignant tumors have been shown to arise
in the body' as clonal cell populations that can proliferate under
conditions where non-transformed cells of the same type do not (8).

Furthermore, many of the cell lines derived from tumors can proliferate

104
in medium containing a lower concentration of serum than is required for
normal cells in culture (9,10). Since one of the important roles of serum
in cell culture is to supply protein growth factors that drive replication
(11), it is likely that the ability of tumor-derived cells to replicate
with reduced levels of serum reflects their decreased requirement for
exogenous protein growth factors. If tumorigenic cells can grow without
exogenously added protein growth factors or with reduced levels of these

factors, this could account for their continued proliferation i viv

O

This hypothesis is supported by the data of'Van Obberghen-Schilling gt g1.
(12) who showed that each of the cell lines derived from the tumors that
developed spontaneously in host animals, following injection of an
infinite lifespan Chinese hamster cell line into host animals exhibited
a decreased requirement for protein growth factors compared with the
parent cell line.

To see if growth factor independence was characteristic of
malignantly transformed human fibroblasts, we tested a series of cell
lines derived from human fibrosarcomas or fibroblasts transformed in
culture for the ability to replicate in the absence of exogenously-added
growth factors. Each of the cell lines tested grew under these
conditions. To determine if the transformed cells produced their own
growth factors enabling them to replicate autonomously, we analyzed
cellular RNA from each cell line for the presence of transcripts of genes
coding for various relevant growth factors. The results showed that the
majority of the cell lines expressed high levels of mRNA for one or more

of these genes.

105

Materials and Methods

Cells and Culture Conditions. The cell lines used are identified
in Table 1. They were routinely maintained in Eagle’s minimal essential
medium supplemented with 0.2 1184 serine, 0.2 nM aspartate and 1.0 mM
pyruvate, 10-15% fetal bovine serum and antibiotics (100 U/ml penicillin
and 100 ug/ml streptomycin) at 37°C in a humidified atmosphere of 5% C02
and air.

Growth Factors. Epidermal growth factor (EGF)3 was obtained from
Bethesda Research Laboratories, Gaithersburg, MD and PDGF from PDGF, Inc.,
Boston, MA. They were dissolved in an aqueous solution of purified human
serum albumin (5 mg/ml) (Sigma Chemical Co. St. Louis, M0) to make stock
solutions of 3 ug/ml EGF and 0.75-1.5 ug/ml PDGF. Plastic containers were
used to prevent adsorption of the growth factors to glass and aliquots of
the stock solutions were stored at -20°C until use. Insulin (Sigma, St.
Louis, MO) was dissolved in a minimal volume of 4 mM HCl and diluted with
triple distilled water to make a stock solution of 10 mg/ml, which was
prepared fresh weekly and stored at 4°C.

Growth Factor Assay Medium. Cells were tested for growth factor and
calcium requirements using as base medium, McM, the modification of the
MCDBllO base medium of Bettger g g1. (22) which was adapted in this
laboratory by Ryan gt g1. (21) for use with serum replacement supplements.
However, for these studies, the McM medium was prepared with a calcium
concentration of 0.1 mM instead of the usual 1.0 mM, and EGF and insulin

were omitted from ”Supplement C”. The calcium concentration of the base

106
medium was confirmed to be 0.1 mM by atomic absorption flame
spectrophotometry.

Assay of Cells for Growth Factor Requirements. Prior to being
tested in this growth factor assay medium, the cells were adapted to the
base medium, McM, by being grown for at least one week in that medium
supplemented with 10% fetal bovine serum and antibiotics. The cells were
then suspended in growth factor assay medium and plated into a series of
60-mm-diameter dishes at a density of 0.7 - 2.2 x 103<cells/cm2. Three
cell lines (SW-594, SW-684, and VIP:FT) did not attach well in this growth
factor assay medium. Therefore, they were suspended and plated in McM
base medium supplemented with 5% or 10% growth factor-inactivated serum.
The latter was prepared by treatment of serum with dithiothreitol and
iodoacetamide as described by van Zoelen gt g1. (23). After 12-16 h the
medium was removed and the cells were rinsed with calcium- and
magnesium-free PBS. Freshly prepared growth factor assay medium,
supplemented with EGF, insulin, or calcium, was added to the cells as
indicated. For the tests involving PDGF, the cells were rinsed and given
freshly prepared growth factor assay medium, and then PDGF was added from
a stock solution using plastic pipette tips to reduce adsorption of this
highly charged molecule to glass. For the studies determining the rate
of cell replication in the presence of serum, the growth factor assay
medium was replaced by the base medium, McM, supplemented with 10% fetal
bovine serum and antibiotics. The concentration of calcium in that McM
medium was 1 mM (21). Every three to four days, the medium was replaced
with freshly prepared test medium. The increase in cell number in

duplicate dishes was monitored by electronic cell counting (Coulter

107
Electronics, Hialeah, FL).

Assay for Tumorigenicity. The ability of cells to form tumors in
athymic mice was assayed by injecting 1-10 x 106 exponentially growing
cells into each of two or three BALB/c athymic mice. The three-week-old
mice were irradiated with 300 rads from a 60C0 source 24 h prior to
injection. As a control, 10 x 106 non-transformed human fibroblasts, taken
from cultures initiated in this laboratory from neonatal foreskins, were
injected into similarly irradiated mice. The mice were examined weekly
for tumors. The resultant tumors were excised, fixed in formalin,
embedded in paraffin, sectioned, and stained with hematoxylin and eosin
prior to histological evaluation.

Assay of Cells for Expression of mRNA. Cells that were analyzed for
growth factor mRNA expression were seeded into lSO-mm-diameter dishes in
McM, medium supplemented with 5% growth factor-inactivated serum and
antibiotics, and were allowed to attach for at least 12 h. The medium was
exchanged for growth factor assay medium, and the cells were harvested for
RNA when they neared confluence. Non-transformed fibroblasts were plated
at higher densities than transformed cells because they did not increase
in number under these culture conditions. RNA was extracted using the
procedure described by Chomczynski and Sacchi (24). Briefly, cells were
washed two to three times in cold PBS and lysed in 3 ml lysing solution
containing 4M guanidinium thiocyanate, 25 mM sodium citrate, 0.5% SDS and
0.1M mercaptoethanol. This solution was acidified with 0.3 ml Na acetate
(2N) and the RNA was extracted by adding 3 ml of water-saturated phenol,
followed by 0.3 ml of a 24:1 mixture of chloroform and isoamyl alcohol,

mixing, incubating on ice for 15 min, and centrifuging the mixture at

108
10,000 x g at 4°C for 25 min. RNA was precipitated from aqueous phase with
isopropanol, pelleted by centrifuging at 10,000 xg for 25 min, washed with
75% ethanol, resuspended in 0.5% SDS at 65°C for 10 min and stored at -
20°C. The concentration of'RNA.was determined from the optical density at
260 nm and its integrity was qualitatively estimated on a mini-gel by
appearance of ribosomal RNA bands.

Gel Electrophoresis. 0.9% agarose gels were prepared by dissolving
low melting agar (Bethesda Research Laboratories, Gaithersburg, MD) in 3X
distilled H20 and heating this mixture to dissolve the agar. The solution
was partially cooled and 10X MOPS buffer (50 ml of 1 M MOPS, 5 ml of 0.5
M EDTA, 5 ml of 3M sodium acetate and 190 ml H20) and formaldehyde were
added to give final concentrations of 6% formaldehyde, 0.02M MOPS, and 1
mM EDTA. Electrophoresis was carried out in 1X MOPS buffer. RNA samples
were prepared by mixing sample with sample buffer (consisting of 100 ul
deionized formamide, 40 ul formaldehyde and 30 ul of 10X MOPS buffer), in
a ratio of 1:3, warming the solution at 65 Co for 5 min. The samples were
then placed on ice and 2 ul tracking dye was added prior to
electrophoresis. Approximately 25 ug of cellular RNA was loaded per lane.

Capillary Gel Blotting. Transfer of the gel to a Zeta-Blot membrane
(Biorad Inc., Rockville Centre, NY) was accomplished as recommended by the
company. Briefly, membrane transfer paper was cut to the size of the gel,
pre-moistened with water, and placed in contact with the gel. Two pieces
of Whatman #3 mm paper were placed under the gel, six pieces were placed
over the gel, and dry paper towers were placed on top of the stack to
provide an upper reservoir for the transfer buffer. The entire apparatus

was placed on a filter paper wick leading to a reservoir containing 10X

109

SSC. Transfer was carried out for 16-24 h, after which the membrane was
separated from the gel, washed in 2X SSC to remove agar, dried at room
temperature, and baked at 80°C for 45 min to reverse the formaldehyde
reaction.

Probe Labeling. 3zP-labeledprobes for Northern analysis of cellular
RNA were prepared by a modification of the random hexamer labeling
technique of Feinberg and Vogelstein (25). Briefly, probe DNA (100-150
ng) was mixed with 2 ul random hexamer solution and triple distilled H20
to give a final volume of 33 ul in a 0.5 ml polypropylene microfuge tube.
This probe/hexamer mixture was denatured at 95°C for 10 min and then cooled
immediately by placing the reaction vial on ice. The following reagents
were added to this solution: 10 ul of oligolabeling buffer, 1.5 ul of a
BSA solution (10 mg/ml of H20, Sigma Chemical Co., St. Louis, MO, nucleic
grade), and 5 ul of a mixture of 32P-labeled nucleotides (labeling
mixture). The oligolabeling buffer consisted of 500 mM Tris, 50 mM MgCl2
and 1M HEPES (pH 7.8). This solution was stored frozen at -20°C1u1til use.
Immediately before use, 2-mercaptoethanol was added to give a final
concentration of'0.7%.(v/v). The 32P-labelednucleotides mixture consisted
of dATP, dCTP, dGTP, and dTTP (New England Nuclear, Boston, MA) (specific
activity 600-800 Ci/mmole) combined to give 12.5 uCi of each nucleotide
in a final volume of 5 ul. ‘These 5 ul aliquots were stored at -20°C1unfil
use. Stock solutions of random hexamers (Pharmacia Co., Piscataway, NJ)
were prepared by diluting the manufacturer’s solution with triple
distilled H20 to give a solution with an optical density of 90 units per
ml, as determined by absorbance at 260 nm. This labeling mixture was

incubated at room temperature for 4-16 h. Unincorporated nucleotides were

110

removed using G-25 spin-columns (181, New Haven, CT) according to the
manufacturer’s recommendations. Hybridization was performed with a
minimum of 0.5 x 106 cpm/ml of hybridization buffer.

Probes Used. The following plasmids were obtained from the American
Type Culture Collection unless otherwise specified: TGF-a, a 1.4 kb
fragment from the EcoRI digestion of the p565 CAN3 plasmid (26) provided
by R. Derynk; TGF-b ., a 30 bp synthetic oligonucleotide commercially
available (Oncogene Science, Manhasset, NY); PDGF-B, a 1.0 kb fragment
from the digest of the pSSVneo plasmid described by Fry gt g1. (27); PDGF-
A, a 1.3 kb fragment from the digest of the p01 plasmid described by
Betsholz gt g1. (28); bFGF, a 0.8 kb fragment from the 8;gRI digest of the
pHFL1-7 plasmid (29), provided by J. Fiddes; IGF-II, a 1.0 kb fragment
from the EggRI digestion of the pIGF2/8-1 plasmid, described by Dull gt
g1. (30); aFGF, a 2.24 kb fragment from the fltngII digest of the pCBIlOZ
plasmid, described by Casadaban g g1. (31); EGF, a 0.4 kb fragment
resulting from the 8;t1 digest of the pMEGF-26F12 plasmid, described by
Gray gt g1. (32); c-_y;, a 4.8 kb fragment from the 18gI/8gmHI digest of
the pSVc-myc-I, described by Land gt g1. (33); c-fig;, a 1.0 kb fragment
resulting from the 8;t I digest of p-fos-I plasmid, described by Curran
gt g1. (34); N-fig;, a 0.65 kb fragment from the 8gmHI/8;gRI digest of the
p6AI plasmid (35), provided by D. Spandidos; K-fig;, a 1.36 kb fragment
from the 18g! digestion of the pk4E plasmid, described by Toshida and
Ryder (36); 82 microglobulin, the entire plasmid containing the 82
microglobulin gene was labeled for use as a probe. This plasmid was a
gift of S. Conrad. The probe was denatured by heating at 95°C with 0.4 ml
of a stock solution of salmon sperm DNA (10 mg/ml of H20) and 100 ul of 0.2

111
N NaOH. The denatured solution was quenched on ice and the solution was
added to the hybridization buffer.

Hybridization and Washing. Hybridization with 150 ul of buffer
containing 1% SDS, 0.7 M NaCl, 10% dextran sulfate, 50% formamide, 0.04M
NazHP04,and 0.5% Blotto (Biorad, Inc.) per cm2 of transfer membrane was
performed overnight with the membrane placed between two sheets of filter
paper (Schleicher and Schuell Inc., Keene, NH).

Membranes were washed twice in a minimum of 100 ml 2X SSC for five
min with agitation, then with the following solutions and conditions: 2
X SSC/0.1% SDS at room temperature for 15 min, 0.5 X SSC/1% SDS at room
temperature for 15 min, and 0.1 X SSC/0.1% SDS room temperature for 15
min. Background hybridization was assessed by autoradiography after 1-2
d. In case of high background, hybridization blots were rewashed under
more stringent conditions, consisting of one wash in 0.5 X SSC/1% SDS at
65°C for 10 min and one wash in 0.1 X SSC/0.1% SDS at room temperature for
20 min.

Blots were prepared for rehybridization by removing bound probe as
suggested by the membrane manufacturer. Briefly, the membranes were washed
twice with 0.1X SSC/0.5% SDS for 20 min at 95 Co. Removal of the probe was
verified by autoradiography.

Radiography. Membranes were blotted to remove excess liquid after
washing and wrapped in a single layer of plastic wrap and radiographed at
-80°C using Kodak XAR-5 film (Eastman Kodak Co., Rochester, NY) with

intensifier screens.

112

Results

Analysis of the Type of Tumor Formed by the Transformed Cell Lines.
Prior to analyzing the transformed or fibrosarcoma-derived cell lines
listed in Table 1 for growth factor requirements, we wished to verify that
they were fibroblastic in origin and determine their tumorigenicity.
Since there are no biochemical or antigenic markers specific for
fibroblasts (37,38), and since the morphology of the majority of these
transformed cell lines in culture varied considerably from the spindle
shape which is characteristic of non—transformed fibroblasts, we injected
them into athymic mice and examined the histology of the tumors that
developed. As shown in Table 2, eight of the cell lines produced
fibrosarcomas and one cell line, SW-594, produced a tumor with both
mesenchymal and epithelial elements. The average latent period for these
tumors ranged from 2 to 8 wks. The CT-l cell line did not produce tumors.
Cell line SW-982, even though originally derived from a fibrosarcoma, has
not given rise to tumors in the 52 wks since injection. The rate of growth
18 yiyg, as determined by the length of time required to produce a 1 cm-
diameter tumor differed among the cell lines. We determined the growth
factor requirements of the various transformed and non-transformed cell
lines to determine whether their degree of growth factor independence
correlated with their rate of tumor-formation, and compared these
requirements with those of the four non-transformed cell lines listed.

Growth Factor Requirements of Normal Fibroblasts and their Response
to Exogenously-Added Growth Factors. To carry out these measurements, we

utilized McM medium (21). This medium with serum replacement supplements

113

supports vigorous long-term growth of diploid human fibroblasts and allows
them to clone with high efficiency (21). Therefore, it permits analysis
of growth factor requirements by eliminating the need for serum. The cell
lines were assayed for ability to replicate in McM medium (which contains
1 mM CaTT), supplemented with all the serum replacements described by Ryan
gt g1. (21) except the protein growth factors, i.e., EGF and insulin. All
four cell lines could replicate in this medium. Therefore, we modified
the medium by reducing the calcium concentration to 0.1 mM, as suggested
by the studies of Westermark gt g1. (39). None of the normal cell lines
were able to replicate in this modified medium, although they remained
viable for as long as tested (>3 wk). Therefore, this modified McM medium
with 0.1 mM Ca'H' was used for assaying the ability of cells to respond to
exogenously added growth factors. Representative data from two of the
normal cell lines, SL68 and K0, are shown in Fig. 1, panels A and B. All
four cell lines were stimulated to replicate by the protein growth
factors. Insulin did not stimulate replication, even when present at a
level of 100 ug/ml (data not shown). However, the average cell doubling
addition of PDGF or EGF was not as great as when the cells were grown in
McM base medium containing 1 mM Ca” or medium supplemented with 10% fetal
bovine serum.

Growth Factor Requirements of Transformed Fibroblasts and their
Response to Exogenously Added Growth Factors. All of the transformed or
tumor-derived cell lines listed in Table 1 were tested for their ability
to replicate in the growth factor assay medium. In contrast to the normal
cell lines, each of themlwas able to replicate. Cell lines HT1080, VIP:FT,
HuT-12, HuT-14, and SHAC replicated rapidly, doubling every 24 h just as

114

they did in medium supplemented with 10% serum. Cell lines CT-l and N01
replicated in the growth factor assay medium at a rate of 50% that seen
in serum supplemented medium. The other four cell lines grew only very
slowly in the growth factor assay medium. Representative growth curves are
shown in Fig. 1, panels C-F.

Eight of these transformed or tumor-derived cell lines were tested
for the effect of increasing the calcium concentration from 0.1 mM to 1
mM. Their responses fell into the two groups. Increasing calcium did not
increase the rate of replication of the five cell lines that grew as
rapidly in growth factor assay medium as in the medium supplemented with
10% serum (HT1080, VIP:FT, HuT-12, HuT-14, and SHAC). The other three,
CT-l, NCI, and SW-982, responded with an increased rate of replication.
See Fig. 1, panels C-F for representative data. When tested for the effect
of adding PDGF, EGF, or insulin to the growth factor assay medium, the
responses of the eight cell lines fell into the same two groups. The five
rapidly growing cell lines cited above did not multiply any faster in the
presence of these added growth factors than in their absence, indicating
complete independence from added growth factors. The other three cell
lines, CT-l, NCI, and SW-982, responded with an increase in rate of
replication but SW-982 cells still did not replicate as rapidly as they
did in medium supplemented with 10% serum. (See Fig. 1 for representative
data.) The characteristics of each of the cell lines tested are summarized
in Table 3. Cell lines 8387, SW-684, and SW-594 could not be examined
further for their response to exogenously added growth factors because of
their inability to remain stably attached to the surface of the dish over
the period of the assay.

115

For the most part, the rate of growth of the eleven cell lines in
growth factor assay medium was highly correlated with their rate of tumor
formation. Four of the five rapidly growing cell lines formed a I cmr
diameter tumor in less than 14 d. The fifth, HuT-12, took 45 d. Four of
the cell lines which grew slowly in the growth factor assay medium
required 30 to 60 d to form a 1 cm-diameter tumor. The other two, CT-l and
SW-982, failed to form tumors within the assay period of one year. No
tumor formation has ever been observed with any non-transformed diploid
human fibroblast cell line in numerous tests carried out in this
laboratory.

Expression of Genes Coding for Protein Growth Factors. One possible
explanation for the observed growth factor independence of the transformed
cell lines is that they produce their own protein growth factor(s) and are
able to replicate by autocrine stimulation as shown by Sporn and Todaro
(40) for certain viral-infected rodent cell lines. In fact, cell lines
HT1080 and 8387 have been shown to express ;1;, the gene coding for the
PDGF B-chain (41,42), and in addition, HT1080 cells have been shown to
express both TGF-a and TGF-b (43). To test this hypothesis, we screened
normal and transformed cell lines for* expression of several growth
factors. The results of the Northern blot analyses of cellular RNA from
human fibroblasts hybridized with 32P-labelechNA from the genes coding
for the PDGF-B, PDGF-A, TGF-a, TGF-b , basic FGF, acidic FGF, EGF, and
IGF-II are summarized in Table 4. All experiments were performed at least
twice using freshly prepared RNA samples from different cultures of the
cell lines indicated and normalization for the amount of RNA loaded was

performed by hybridization with the probe to 82 microglobulin.

116
Representative blots are shown in Figs. 2-6.

TGF-a was overexpressed in four of the five rapidly growing
tumorigenic cell lines, i.e., Hut-12, Hut-14, HT1080, VIP:FT, and in the
non-tumorigenic cell line SW-982. The predicted 4.8 kb transcripts (43)
are shown in Fig. 2. No transcripts were detected in normal cells under
these hybridization conditions. In contrast to TGF-a, expression of the
2.4 kb transcript corresponding to TGF-b (43) was detected in normal cell
lines NF812 and SL68, as well as in the SL68 cells which were grown in
medium supplemented with 10% fetal bovine serum. Cell lines HT1080, SW-
982, and CT-l showed increased expression of this transcript (Fig. 3) and
the remaining transformed cell lines demonstrated expression levels
comparable to the normal cell lines as seen by overexposure of another
such blot (Not shown).

Overexpression of a 4.0 kb transcript, corresponding to PDGF-B (44)
was highest in non-tumorigenic cell line CT-l (Fig. 4). PDGF-B is the only
growth factor known to be mitogenic for human fibroblasts in culture that
was expressed by CT-l cell line. Overexpression of the PDGF-B transcript
was seen in each of the rapidly growing transformed cell lines; HT1080,
HuT-12, HuT-14, VIP:FT, and SHAC. Autoradiography showed faint
hybridization with the c-;1; probe in the SW-684 cell line, but the lane
was over-loaded. In the rapidly growing cell lines HuT-12, HuT-14, SHAC,
and HT1080, hybridization with the PDGF-A probe showed overexpression of
2.8, 2.3 and 1.9 kb, sizes predicted by the work of Betsholtz gt g1. (45).
Cell line SW-684 also appeared to overexpress PDGF-A but as seen from the
82 microglobulin hybridization at lane clearly contained more RNA than

the other lanes. Overexposure of the blot showed that normal cell lines

117
and the remaining transformed cell lines expressed PDGF-A transcripts at
a reduced level under the test conditions (Fig. 5).

Increased expression of bFGF was seen only in the VIP:FT cell line
(Fig. 6). We observed the predicted transcripts of 3.7 and 7.0 kb (46).
Overexposure of this blot did not reveal any evidence of cell lines
expressing even small amounts of b-FGF. Increased expression of IGF-II was
seen only in one cell line, SHAC. The expected 6.0 kb transcript was seen
above the 28 S ribosomal RNA (Fig. 8).

Similar analysis was performed for EGF and a-FGF. We did not detect
expression of either of these genes. The RNA on the blots, however, was
able to hybridize with other probes and the amount of RNA loaded was equal
to that loaded for the other cell lines as judged by hybridization with
82 microglobulin. The probes used to detect EGF and a-FGF contained the
predicted DNA sequences as judged by restriction analysis of the digested
plasmids containing the probe.

Expression of’ Factors Putatively’ Involved in Mitogenic Signal
Transduction. Our results showed that the transformed cells which grew
rapidly in culture expressed multiple growth factors. However, cell lines
NCI, CT-l, and SW-982, which grew slowly in the growth factor assay
medium, did not overexpress any of the growth factors. Since proto-
oncogenes have been implicated in transduction of the mitogenic signal,
and their constitutive activation has been shown to increase the rate of
growth of some cell lines (47,48), we analyzed the various cell lines for
the overexpression of N-fig;, K-fig;, my;, and fig; to determine if the
overexpression of these genes could explain the observed growth factor

independence of some of the transformed cell lines.

118

Expression of N-_r_g;. N-fig; has been shown to have two major
transcripts in human cells, one just less than 5.0 kb and one
approximately 2.0 kb (49). Fig. 7 shows the results of Northern analysis
of cellular RNA extracted from normal and transformed human fibroblasts
maintained in the growth factor assay medium. The amount of RNA loaded in
each lane was approximately 25 ug and normalization for loading error was
carried out by hybridization to 82 - microglobulin. We detected transcripts
of 3.0 and 2.0 kb as determined by their relationship to 288 and 18S
ribosomal RNA bands. Cell lines CT-l, SHAC, and 8387 were shown to
overexpress N-fig; compared to normal cell lines or to the other
transformed cell lines. Cell line SW-684 appeared to overexpress the N-
fig; gene, but that lane was slightly overloaded as seen from 82
microglobulin transcript. SL68 cells (Lane 2) appeared to overexpress N-
fig; compared to the other non-transformed cell lines tested, but this lane
was also overloaded. The VIP:FT cell line did not overexpress N-fig; (data
not shown). The increased amounts of N-fig; transcripts observed could
represent overexpression of a normal or a mutated N-fig; gene since the
probe used was not specific for mutated N-fig;.

Expression of K-fig_. The results of hybridization of cellular RNA
with a probe derived from the v-K-Q; gene are shown in Fig. 9. The
presence of' a 3.8 kb band is seen in Northern analysis of K-fig;
expression, however, we do not observe a 5.5 kb transcript which usually
accompanies the 3.8 kb transcript. Two additional bands are seen
corresponding to sizes of approximately 2.5 and 1.7 kb. Increased levels
of these transcripts are seen in serum-stimulated normal fibroblast cell

line, SL68 and transformed cell lines, SW-684 and VIP:FT. We are currently

119

obtaining alternate probes for K-fig; to determine whether the transcripts
observed, represent the overexpression of c-K-fig; in these cell lines.

Expression of c- c and c-fig;. Figure 8 shows the results of the
Northern analysis of RNA for the expression of c-myg. The single expected
transcript of 2.3 kb is indicated. Cell lines CT-l, NCI and SHAC are shown
to overexpress c-myg. These lanes are not overloaded as seen by the
results of hybridizaton with anficroglobulin. Overexpression of c-my; is
not due to the fact that transformed cells were replicating in the growth
factor assay medium while normal cells did not. Lane 13 contains RNA
extracted from exponentially growing SL 68 cells, maintained in growth
factor assay medium containing 10% fetal bovine serum. These cells do not
express increased c-myc even though they are replicating at a rate
comparable to the other transformed cell lines. A single c-fig; transcript
of 2.2 kb was detected in all cell lines but was not overexpressed

(results not shown).

120

Discussion

One of the changes common to all tumor cells is the ability to
proliferate in the body under conditions where normal cells do not. In
spite of this common characteristic, the cellular changes which allow this
proliferation of tumor are not completely understood. We show here that
all of the transformed cell lines tested for growth factor independence
exhibited the common phenotype of being able to replicate in serum-free
medium in the absence of exogenously-added growth factors. These data
support the hypothesis that the reason tumorigenic cells continue to
proliferate in the body under conditions in which non-transformed cells
cease to divide is they have acquired a degree of growth factor
independence. Our finding that, with the exception of the HuT-12 cell
line, the degree of growth factor independence of cell lines in culture
was highly correlated with the rate at which they developed 1 cm diameter
tumors in athymic mice also supports this hypothesis.

Our data on the level of expression of mRNA of genes coding for
various growth factors and factors putatively involved in mitogenic signal
transduction indicated that the series of cell lines that replicated at
the maximum rate even in the absence of added growth factors (Table 3)
were expressing multiple growth factors (Table 4).

Several cell lines expressed multiple growth factor genes, but the
set of genes expressed and/or the level of expression was not identical
for each. Furthermore, two malignant cell lines, NCI and SW-684, exhibited
the ability to grow in the absence of exogenous growth factors (Table 2)

and, yet, did not overexpress any of the series of growth factor genes

121

assayed (Table 4). It is always possible that these two cell lines
overexpress one or more growth factors that were not assayed, but if they
do not, they provide evidence that acquisition of growth factor
independence can be achieved by mechanisms other than autocrine
stimulation. Taken all together, our data suggest that in the course of
becoming tumorigenic, cells acquire a conmon phenotype, viz., growth
factor independence, but the specific changes involved in achieving this
characteristic can differ.

Two cell lines, CT-I and SW-982, were growth factor independent, and
yet did not form tumors, It is always possible, however, that the rate of
replication of these was too slow to produce a palpable tumor within the
assay period of' one .year. But if ‘these cell lines are truly non-
tumorigenic, our results indicate that the acquisition of growth factor
independence p_efi gg, even by a cell which has acquired an infinite
lifespan in culture, is not sufficient to cause malignancy. This agrees
with the finding of Perez-Rodriguez gt g1. (50) that growth factor
independence in Chinese hamster cell line CCL 39 was highly correlated
with malignant transformation, but its acquisition was not sufficient to
cause malignant growth.

Recently, it has been shown that HuT-12 and HuT-14 are, in fact,
derived from cell line 8387 (51), and consequently represent clonal
isolates of that original fibrosarcoma-derived cell line. The fact that
these clonal isolates are more growth factor independent and tumorigenic
than the parental cell line is not surprising. The characteristics of
clonal isolates can differ from those of original cell lines. For example,

non-tumorigenic cell lines can be isolated from tumorigenic cell

122
populations (52).

We did not detect the expression of EGF transcripts in any of the
cell lines tested even under conditions of low stringency hybridization
even though Kurobe gt g1. (53) have reported that normal human fibroblasts
synthesize small amounts of EGF (urogastrone). It is possible that we were
unable to detect such low levels of expression under the conditions of the
assay. The fact that EGF has rarely been reported to be overexpressed by
a tumor cell line is consistent with our data, but it is somewhat
surprising that this growth factor is not expressed in at least some cell
lines given the variety of growth factors that are mitogenic for human
fibroblasts which we do see overexpressed.

The fact that we observed overexpression of multiple growth factor
genes in the five rapidly growing , malignantly transformed cell lines
(Table 4) raises the question whether each growth factor is activated
separately or whether some common mechanism exists which activates more
than one. One possible explanation for the overexpression of multiple
growth factors is that stimulation by one growth factor may cause the
induction of another. Heldin and associates (54) recently demonstrated
that density-inhibited, normal human fibroblasts, transiently express
PDGF-Alwhen stimulated to proliferate with either EGF or PDGF, implicating
a positive feedback loop in the prereplicative phase of normal human
fibroblasts. However, unlike our studies, they did not investigate PDGF-
A expression in cycling cells. If a cell is able to produce its own PDGF-
B, one might see expression of both PDGF-B and PDGF-A. Since TGF-a
functions mitogenically through the EGF receptor (55), it is possible that

overexpression of TGF-a could cause the expression of PDGF-A in

123

fibroblasts. We observed such coordinate expression in several of the cell
lines. Another recent report by Rappolee gt g1. (56) demonstrated the
expression of TGF-a, TGF-b, PDGF-A, and IGF-1 in adherent cells
(predominantly macrophages) isolated from wound cylinders in mice. This
coordinate expression supports the hypothesis that mechanisms exist which
direct the expression of multiple growth factors in normal cells under
some conditions. Perhaps these mechanisms are altered in some malignantly
transformed fibroblasts, causing the expression of multiple growth
factors.

There are relatively few examples of studies of the growth factor
requirements of transformed human fibroblasts in serum-free medium. Namba
gt g1. (57) showed, as we did, that the infinite lifespan, non-tumorigenic
fibroblast cell line CT-l, that was isolated after repeated exposure to
carcinogen was able to grow in a serum-free medium which did not allow the
growth of normal fibroblasts. Similarly, Fry gt g1. (27) and Hurlin gt g1.
(58) showed that human fibroblasts, transformed by the ;t; oncogene or the
H-fig; oncogene , were able to grow in the absence of exogenously added
protein growth factors. A variety of transformed rodent cell lines have
also been shown to be able to grow in such media (59,60), indicating that
these media select for growth factor independent cells regardless of the
mechanisms which cause the phenotype. One implication of both studies is
that selection for growth factor independence in serum-free medium may be
a useful technique for isolating cells which have intermediate phenotypes
between normal and malignantly transformed. Such an approach may be
important, given the inability to routinely'obtain malignantly transformed

human fibroblasts in culture, even with carcinogen treatment (4).

124
Acknowledgements
We wish to express our sincere thanks to the scientists who provided
us with nine transformed fibroblast cell lines. The excellent technical

assistance of Lonnie Milam and Karen Hawes is gratefully acknowledged.

125

References

1.

10.

II.

12.

Doll, R. An epidemiological perspective of the biology of cancer.
Cancer Res., 88: 3573-3583, 1978.

Peto, R. "Epidemiology, multistage models and short-term
mutagenicity tests", 1;: J. D. Wilson, and J. A. Winsten (eds.),
Origins of Human Cancer. Book C, pp. 1403-1428. Cold Spring
Harbor: Cold Spring Harbor Laboratory, 1977.

Knudson, A. G. "Genetic predisposition to cancer”, 18: H. H. Hiatt,
J. 0. Watson, and J. A. Winsten (Eds.), The Origins of Human Cancer.
pp. 45-54. Cold Spring Harbor: Cold Spring Harbor Laboratory, 1977.

McCormick, J. J. and Maher, V. M. Towards an understanding
of the malignant transformation of diploid human fibroblasts.
Mutat. Res., 188: 273-292, 1988.

Namba, M., Nishitani, K., Fukushima, F., Kimoto, T., Utsunomiya, J.,
and Hayflick, L. Neoplastic transformation of human diploid
fibroblasts treated with chemical carcinogens and Co-60 Y-rays.
Gann 81: 221-230, 1981.

Thomassen, D. G., Gilmer, T. M., Annab, L. A., and Barrett, J. C.
Evidence for multiple steps in neoplastic transformation of normal
and preneoplastic Syrian hamster'embryo cells following transfection
of Harvey murine sarcoma virus oncogene (v-Ha-fig;). Cancer Res.,
18: 726-732, 1985.

Land, H., Parada, L. F., and Weinberg, R. A. Tumorigenic conversion
of primary embryo fibroblasts requires at least two cooperating
oncogenes. Nature (London) 881: 596-602, 1983.

Innacone, P. M., Weinberg, W. C., and Deamant, F. D. On the
clonal origin of tumors: A review of experimental models.
Int. J. Cancer, 88: 778-784, 1987.

Holley, R. W. Control of growth of mammalian cells in cell culture.
Nature (London) 888: 487-490, 1975.

Tubo, R. A., and Rheinwald, J. G. Normal human mesothelial cells
and fibroblasts transfected with the E113; oncogene become EGF-
independent, but are not malignantly transformed. Oncogene Res. 1:
407-421, 1987.

Eagle, H. Nutrition needs of mammalian cells in tissue culture.
Science, 188: 501-504, 1955.

Van Obberghen-Schilling, E., Perez-Rodriguez, R., Franchi, A.,
Chambard, J. C., and Pouyssegur, J. Analysis of growth factor
”relaxation" in Chinese hamster lung fibroblasts required for
tumoral expression. J. Cell. Physiol. 118: 123-130, 1983.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

126

McCormick, J.J. and Maher, V.M. ”Measurement of colony-forming
ability and mutagenesis in diploid human cells." 1;: E.C.
Friedberg and P.A. Hanwalt (eds). DNA Repair: A Laboratory
Manual of Research Procedures, pp. 501-522. New York: Marcel
Dekker, 1981.

Day, R.S. Studies on repair of adenovirus 2 by human
fibroblasts using normal , xeroderma pigmentosum, and
xeroderma pigmentosum heterozygous strains. Cancer Res., 88:
1965-1970, 1974.

Namba, M., Nishitani, K., Fukushima, F., Kimoto, T., and Nose,
K. Multistep process oEopeoplastic transformation of normal
human fibroblasts by Co gamma rays and Harvey sarcoma
viruses. Int. J. Cancer, 81, 419-423, 1986.

Kakunaga, T. Neoplastic transformation of human diploid
fibroblasts by chemical carcinogens. Proc. Natl. Acad. Sci.
USA 18, 1334-1338, 1977.

Mukherji, B.,MacAlister, T. J., Guha, A., Gillies, C. G.,
Jeffers, D. C., and Slocum, S. K. Spontaneous in vitro
transformation of human fibroblasts. J. Natl. Cancer Inst.
18, 583-587, 1984.

Fogh, J., Wright, W.C., and Loveless, J.D. Absence of HeLa
cell contamination in 169 cell lines. derived from human
tumors. J.Natl. Cancer Inst., 88: 209-214, 1977.

Rasheed, S., Nelson-Rees, W.A., Toth, E.M., Arnstein, P., and
Gardner, M.B. Characterization of a newly-derived human
sarcoma cell line (HT1080). Cancer, 88: 1027-1032, 1974.

Azzarone, 8., and Pedulla, D. Spontaneous transformation of
human fibroblast cultures derived from bronchial carcinomata.
Eur. J. Cancer. 18, 557-561, 1976.

Ryan, P. A., McCormick, J. J., and Maher, V. M. Modification of
MCDB-IIO medium to support prolonged growth prolonged growth and
consistent high cloning efficiency of human diploid fibroblasts.

Exp. Cell. Res. 2, 318-328, 1987.
Bettger, W. J., Boyce, S. T., Walthall, B. J., and Ham, R. G.

clonal grwoth and serial passage of human diploid fibroblasts in a
lipid-enriched synthetic medium supplemented with epidermal growth
factor, insulin, and dexamethasone. Proc. Natl. Acad. Sci. USA, 18:

5588-5592, 1981.

Van Zoelen, E. J. J., Van Oostwaard, T. M. J., vander Saag,
P. T., and de Laat, S. W. Phenotypic transformation of normal
rat kidney cells in a growth factor defined medium: induction

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

127

by a neuroblastoma-derived transforming growth factor
independent of the EGF receptor. J. Cell. Physiol. 188, 151-
160, 1985.

Chomczynski, P., and Sacchi, N. Single-step method RNA
isolation by acid guanidinium thiocyanate-phenol-chloroform
extraction. Anal. Biochem. 188, 156-159, 1987.

Feinberg, A. P., and Vogelstein, B. A technique for radiolabeling
DNA restriction endonuclease fragments to high specific activity.
Anal. Biochem. 188: 6-13, 1983.

Derynck, R., Roberts, A. 8., Winkler, M. E., Chen, E. Y., and
Goeddel, D. Y. Human transforming growth factor-a : Precursor
structure and expression in E. coli. Cell 88, 287-297, 1984.

Fry, 0. G., Hurlin, P. J., Maher, V. M., and McCormick, J. J.
Transformation of diploid human fibroblasts by transfection with the
v-;1;, EDGF8/c-;;;, or T24 H-fig; genes. Mutat. Res. 188, 341-352,
1988.

Betsholz, C., Johnsson, A., Heldin, C.-H., Westermark, 8., Lind, P.,
Urdea, M. S., Eddy, R., Shows, T. 8., Philpott, K., Mellor, A. L.,
Knott, T. J., and Scott, J. cDNA sequence and chromosomal
localization of human platelet-derived growth factor'A-chain and its
expression in tumour cell lines. Nature (London), 888: 695-699,
1986.

Schweigerer, L., Neufeld, G., Mergia, A., Abraham, J. A., Fiddes,
J. C., and Gospodarowicz, 0. Basic fibroblast growth factor in
human rhabdomyosarcoma cells: Implications for the proliferation
and neovascularization of myoblast-derived tumors. Proc. Natl.
Acad. Sci. USA, 88: 842-846, 1987.

Dull, T. J., Gray, A., Hayflick, J. S., and Ullrich, A.
Insulin-like growth factor 11 precursor gene organization in
relation to insulin gene family. Nature (London), 818: 777-
781, 1984.

Casadaban, M. J., and Cohen, S. N. Analysis of gene control signals
by DNA fusion and cloning in Escherichia coli. J. Mol. Biol. 188:
179-207, 1980.

Gray, A., Dull, T. J., and Ullrich, A. Nucleotide sequence of
epidermal growth factor cDNA predits a 128,000-molecular weight
protein precursor. Nature (London), 888: 722-725, 1983.

Land, H., Parada, L. F., and Weinberg, R. A. Tumorigenic conversion
of primary embryo fibroblasts requires at least two cooperating
oncogenes. Nature (London), 888: 596-602, 1983.

Curran, T., Miller, A. D., Zokas, L., and Verma, I.M. Viral and

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

128

cellular fig; proteins: A comparative analysis. Cell. 88: 259-268,

1984.

Wilson, 0. M., Maher, V. M., and McCormick, J. J. Transformation
of normal diploid human fibroblasts by a transfected N-fig; oncogene.

Proc. Amer. Assoc. Cancer Res. 88: 152 (Abstract), 1984.

Tsuchida, N. Nucleotide sequence of the oncogene encoding the p21
transforming protein of Kirsten murine sarcoma virus. Science, 811:

937-938, 1982.

Kindblom, L., Jacobsen, G.K. and Jacobsen, M.
Immunohistochemical investigation of tumors of supposed
fibroblastic-histiocytic origin. Human Pathol., 8: 834-840,
1982.

Swanson, P.E., Manivel, J.C., and Wick, M.R. Immunoreactivity
for leu-7 in neurofibrosarcoma and and other spindle cell
sarcomas of soft tissue. Am. J. Pathol., 188: 546-560, 1987.

Westermark, B. and Heldin, C.-H. Similar action of platelet-
derived growth factor and epidermal growth factor in the
prereplicative phase of human fibroblasts suggests a common
intracellular pathway. J. Cell. Physiol. 188, 43-48, 1985.

Sporn, M. 8., and Todaro, G. J. Autocrine secretion and
malignant transformation of cells. N. Engl. J. Med. 888, 878-
880, 1982.

Eva, A., Robbins, K. C., Andersen, P. R., Srinivasan, A.,
Tronick, S. R., Reddy, E. P., Ellomore, N. W., Galen, A. T.,
Lautenberger, J. A., Papas, T. S., Westin, E. H., Wong-Staal,
F., Gallo, R. C., and Aaronson, S. A. Cellular genes
analogous to retroviral gg; genes are transcribed in human
tumour cells. Nature 888, 116-119, 1982.

Pantazis, P., Pellicci, P. G., Dalla-Favera, R., and
Antoniades, H. N. Synthesis and secretion of protein
resembling platelet-derived growth factor by human
glioblastoma and fibrosarcoma cells in culture. Proc. Natl.
Acad. Sci. USA 88, 2404-2408, 1985.

Derynck, R., Goeddel, D. V., Ullrich, A., Gutterman, J. U.,
Williams, R. D., Bringman, T. S., and Berger, W. H.
Synthesis of messenger RNAs for transforming growth factors
a and b and the epidermal growth factor receptor by human
tumors. Cancer Res. 81, 707-712, 1987.

Versnel, M. A., Hagemeijer, A., Bouts, M. J., van der Kwast,
T. H., Hoogsteden, H. C. Expression of c-;;; (PDGF B-chain)
and PDGF A-chain genes in ten human malignant mesothelioma
cell lines derived from primary and metastatic tumors.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

129
Oncogene 8, 601-605, 1988.

Bywater, M., Rorsman, F., Bongcam-Rudloff, E., Mark, G.,
Hammacher, A., Heldin, C.-H., Westermark, 8., Betscholtz, C.
Expression of recombinant platelet-derived growth factor A-
and B-chain homodimers in rat-1 cells and human fibroblasts
reveals differences in protein processing and autocrine
effects. Mol. Cell. Biol. 8, 2753-2762, 1988.

Klagsbrun, M., Sasse, J., Sullivan, R., and Smith, J. A.
Human tumor'cells synthesize an endothelial cell growth factor
that is structurally related to basic fibroblast growth
factor. Proc. Natl. Acad. Sci. USA 88, 2448-2452, 1986.

Lumpkin, C. K., Knepper, J. E., Butel, J. S., Smith, J. R.,

Pereira-Smith, 0. M. Mitogenic effects of the proto- oncogene and

oncogene forms of c-H-ras DNA in human diploid fibroblasts.
Cell. Biol. 8, 2990-2993, 1986.

Sorrentino, V., Drozdoff, V., McKinney, M. 0., Zeitz, L., and
Fleissner, E. Potentiation of growth factor activity by
exogenous c-my; expression. Proc. Natl. Acad. Sci. USA 88,
8167-8171, 1986.

Hall, A. and Brown, R. Human N-fig;: cDNA cloning and gene
structure. Nucleic Acids Res. 18, 5255-5268, 1985.

Perez-Rodriguez, R., Franchi, A., Deys, B. F., and
Pouyssegur,J. Evidence that hamster fibroblast tumors emerge
in nude mice through the process of two in vjvo selections
leading to growth factor "relaxation“ and to immune
resistance. Int. J. Cancer 88, 309-314, 1982.

McCormick, J. J., Yang, 0., Maher, V. M., Farber, R. A.,
Neuman, W., Peterson, Jr., W. D., and Pollack, M. S. The HuT
series of "carcinogen-transformed” human fibroblast cell lines
are derived from the human fibrosarcoma cell line 8387.
Carcinogenesis, in press, 1988.

Paterson, H., Reeves, 8., Brown,R., Hall, A., Furth, M., 805,
J., Jones, P., and Marshall, C. Activated N-fig; controls the
transformed phenotype of HT1080 human fibrosarcoma cells.
Cell 81, 803-812, 1987.

Kurobe, M., Furukawa, S., and Hayashi, K. Synthesis and
secretion of an epidermal growth factor (EGF) by human
fibroblast cells in culture. Biochem. Biophys. Res. Comm.
181, 1080-1085, 1985.

Heldin, C.-H., Johnsson, A., Wennergren, S., Wernstedt, C.,
Betscholtz, C., and Westermark, B. A human osteosarcoma
cell line secretes a growth factor structurally related to

55.

56.

57.

58.

59.

60.

130
a homodimer of PDGF A-chains. Nature 818, 511-514, 1986.

Carpenter, G., Stoscheck, C. M., Preston, Y. A., and De Larco,
J. E. Antibodies tothe epidermal growth factor receptor block
the biological activities of sarcoma growth factor. Proc.
Natl. Acad. Sci. USA 88, 5627, 1983.

Rappolee, D. A., Mark, D., Banda, M. J., and Werb, Z. Wound
macrophages express TGF-a and other growth factors in vivo:
Analysis by mRNA phenotyping. Science 881, 708-712, 1988.

Namba, M., Kusaka, T., Fukushima, F., and Kimoto, T.
Differential growth response of normal human diploid
fibroblasts and jg vitrg transformed human fibroblasts in
serum-free defined culture medium. Int. J. Cancer 88, 645-
649, 1984.

Hurlin, P., Maher, V. M., and McCormick, J. J. Malignant
transformation of a human fibroblast cell strain MSU-1.l, by
expression of a transfected T24 H-fig; oncogene. Proc. Natl.
Acad. Sci. USA, 1988, in press.

Zhan, X. and Goldfarb, M. Growth factor requirements of
oncogene-transformed NIH 3T3 and BALB/c 3T3 cells cultured in
defined media. Mol. Cell. Biol. 8, 3541-3544, 1986.

Chang, L. C., Silnutzer, J., Pipas, J. M., and Barnes,

Selection of transformed cells in serum-free media. In Vitro Cell.

Develop. Biol. 81, 707-712, 1985.

131
Table 1 Fibroblastic cell lines studied

 

 

Cell Original
line Description of origin Supplier reference
SL68 Neonatal foreskin Initiated here 13
SL84 Neonatal foreskin Initiated here 13
NF812 Neonatal foreskin Initiated here 13
KD Adult skin biopsy J. Leavitta 14
CT-l Carcinogen—transformed M. Nambab 15
HuT-12 Fibrosarcoma J. Leavitta 16
HuT-14 Fibrosarcoma J. Leavitta 16
VIP:FT Spontaneously-transformed B. Mukerjic 17
NCI Fibrosarcoma C. Cooperd --
SW-982 Fibrosarcoma J. Foghe --
SW-684 Fibrosarcoma J. Foghe 18
SW-594 Fibrosarcoma J. Foghe --
HT1080 Fibrosarcoma A.T.C.C.f 19
SHAC Fibrosarcoma B. Azzaroneg 20
8387 Fibrosarcoma W. Petersonh --

 

. Leavitt, Armand Hammer Cancer Research Center, Palo Alto, CA.

. Namba, Kawasaki Medical School, Kurashiki, Japan.

. Mukerji, University of Connecticut Health Center, New Haven, CT.
. Cooper, NCI, Bethesda, MD.

Fogh, formerly of Sloan-Kettering Institute, New York, deceased.

n
D
1
unwzc-n

fAmericanType Culture Collection.
9Dr. B. Azzarone, Villijuif, France.
'th W. Peterson, Jr., Children’s Hospital of Michigan, Detroit, MI.

 

I32

.u.=o=.;=§=a-=== ou=_m o_aau__::m 69::

.=o_auan=. mus—m one» age =. cu>uomao use: muosad ozu

.a:¢a.;z be as.“ 8:“ us So m.a ago: abuzz. omm;_a

.Lmdosa_c :. so _ :oaoc a. meas:d on. mac: .3 Lassa: «:8 no 18:..83 m. 56.28; Auzodaaa

 

=5.

«.5.

aaoucamoca_u
«Laden. _a__o:d_ao ecu _caa;o:mm9=
ado: a=_>a= Lesa“ d=a=m__a:
asouuamoaa.u

asoucamoLa_. coda.acocobb.v A.Loom
aaoocomoca.u

assucomoLa_m

aeoucmmoaa.u

aaouccmoca_m

esoucumaua_u

en

en

an

L..")
V

 

xao_odm_:

Loszp

a.§e.: so _

some. a“
Lease Lo.

man: .32

 

2.3. sm>_ca=-eEoucamoLa.u ~ca-:m
a.o~ awaccbm:aca :oac:.ucou _-_u
c.o_ um>.go=-asoucamoca_u Nanc
3.:— =o>.;o=-a§=ucamcaa_g cam-:m
3.9. =o>.ao=-asoucamcaa_u _uz
=.o. =o>_aa=-asaucamcaa_5 emo-:m
c.3— Nmmm .6 o=o_u:=m ~.-_::
o.c_ um>.amt-~scucomoaa_u u<:n
3.. =m>.cm=-asoucomoca.u ace—_z
N.“ smaaobmzmcu x_m:ow=uucoam _mua_>
3.3— Nana .0 e::_u:=m e_-_::
Ao-a_ xv =.m_La 8:.—
omaoa Lo: _.ou

aasooq=_ m__ou

 

«_peu 58>.Loeteseocamoca_. Le chad—:9 :_ uoscobm=aca mama—aoca_u be >~.u_=ea.cosap ~ o_aa—

133

Table 3 Tumorigenicity and growth factor independence and responsiveness of

the human fibroblast cell lines

 

 

Ability to form Responsiveness
Cell Origin of tumors in Growth factor to exogenous
line cell line athymic micea independence growth factors
SL68 Neonatal foreskin -- -- +
SL84 Neonatal foreskin -- -- +
NF812 Neonatal foreskin -- -- +
KD Adult skin biopsy -- -- +
SW982 Fibrosarcoma-derived -- + +
CT-I Carcinogen transformed -- + +
NCI Fibrosarcoma-derived + + +
SW-594 Fibrosarcoma-derived + + NDb
SW-684 Fibrosarcoma-derived + + nub
8387 Fibrosarcoma-derived + + NDb
HT1080 Fibrosarcoma-derived + + --
VIP:FT Spontaneously transformed + + --
HuT-12 Fibrosarcoma-derived + + --
HuT-14 Fibrosarcoma-derived + + --
SHAC Fibrosarcoma-derived + + --

 

3The symbol. --, indicates no tumors were formed during a period of at least
one year following injection of 107 cells per mouse.

bNot determined.

134

Table 4 Expression of genes coding for peptide growth factors and sole

proto-oncogenes in transformed and non-transforued human fibroblasts

 

 

O N '3 F- <' N E

N 00 H H u. on CD 3

H 00 CD 1 1 U so Oi --0 1x QS-

m to e-O p— p— < a u—e 1 1 I 00 £00)

1.1. _J O i— 3 3 I .—. U 3 3 I— M _Jin

2 v1 )4 I I I m > z w W U Q WV
PDGF-Ba 0 0 0 1d 1 1 1 1 0 0/1 0 2 N0c 0
PDGF-Ab 1 1 1 2 4 4 4 0 1 1 1 1 NDC 1
TGF<x 0 o 0 3 3 3 0 3 0 0 1 0 N0C 0
b-FGF o 0 0 0 0 0 0 1 0 0 o 0 0 0
IGF-II 0 0 o o 0 0 1 0 0 0 0 0 0 0
16F78 1 1 1 3 1 1 1 1 1 1 1 3 1 1
a-FGF 0 0 0 0 0 0 0 0 0 o 0 o 0 0
EGF 0 0 0 0 0 0 0 0 0 0 0 0 0 0
N-fig§_ 1 1 1 1 1 1 3 1 1 1 1 3 3 1
K-[g§_ 0 NDC N0c o 0 0 0 1 0 1 0 0 0 1
Eng; 1 1 ’ 1 1 1 1 3 1 3 1 1 2 1 1
fos 1 1 1 1 1 1 1 1 1 1 1 1 1 1

 

aPDGF-B is the c-gj; gene.

bPDGF-A is the A-chain of PDGF.

cNot determined.

d1 represents the intensity of the signal from the lowest level of

transcript detectable. All other intensities are estimated relative
to this.

135

Figure 1. Effect of growth factors, serum and calcium on the replication
of human fibroblast cell lines. Normal (panels A and B) and transformed
human fibroblast cell lines (panels C through F) are described in Table
1. Cells were plated into culture dishes at a density of 0.7 - 2.2 x 103
cell/cm2 12-16 h prior to determining the initial cell numbers on Day 0.
Plating medium was then replaced by McM medium containing 1 mM calcium
supplemented with 10% fetal bovine serum (0) or growth factor assay
medium containing no growth factors (A), 1.5 ng/ml PDGF (A), 3 ng/ml EGF
(O), 10 ug/ml insulin ([11), or 1 11M calcium (I), as described in the
Materials and Methodslsection. The media were replaced with freshly
prepared test media after 3-4 d as indicated by arrows (1 ). Each point

represents the average of duplicate dishes.

 

CELLS PER DISH
6
U

 

 

 

 

    

  

 

 

Io4

 

106

CELLS PER DISH
%

CELLS PER DISH

  

   

 

 

 

 

O 2 4 6 8 O
: I 1 T 1 1 T I T 1: : 1 1 1 1 1 l5
: 55 B :
I - - KO -
I— - L.. —:
A 1 :' “.1
i E ::
- ‘1 A~‘ _ .. -
l i i
I I l l I I I L I t 1 I l I
I- I I I I I I I I I: 1:
E C I 1
1'. . -
:' /°‘5 ‘1:
: /§/‘: .1
i i l
I l I I l I I I L I I I I I
I T I I I I I I I I I I i I I
" E 5/3 ‘1 :- F '1
E HuT-l4 I ; r; VIP:FT é/ ‘ .3
: if 3 E g/g/ :
l 1 e E g
\ / I I I I I
I I P I I I I I I I i I I
O 2 4 6 8 O 2 4 6
DAYS DAYS

 

 

 

 

137

Figure 2. Northern blot analysis of the expression of TGF-a in non-
transformed, tumor derived, or transformed human fibroblasts. 3zP-labelled
probe DNA was hybridized to Northern blots of cellular RNA extracted from
the various cell lines grown in growth factor assay medium as described.
The 4.8 kb transcripts observed represent overexpression of this gene. The
position of the ribosomal RNA bands are indicated on the right for size

comparison.

138

Eacmmumogm Amav

u<zm Away.

ea 82> AHSV
www-3m Aofiv

vwouzm
Huz
fiuho
omouhz
«Huhaz
Nfiuhax
ox
mmdm
Nfimmz

Amy
Amy 8
ARV
Amy
Amv
Aev
Amv
ANV
AHV

i-ZBS

4.8 kb-

139

Figure 3. Northern blot analysis of the expression of TGF-b in non-
transformed, tumor derived, or transformed human fibroblasts. 32P-labelled
probe DNA was hybridized to Northern blots of cellular RNA extracted from
the various cell lines grown in growth factor assay medium as described.
The 2.4 kb transcripts observed represent expression of this gene. The
position of the ribosomal RNA bands are indicated on the right for size

comparison.

140

Nfimmz Amflv
meaams
mg AHHV

NH-P== Aofiv.

SJSIAmV
emcee: Amy
H-»u ARV
_oz Amy

888-3m Amy
Nma-3m “av

u<=m Amv
seem-3.5 3V

Emma RHV

 

n
b
k
4
2

141

Figure 4. Expression of PDGF-B in non-transformed, tumor derived, or
transformed human fibroblasts. 32P-labelledprobe DNA was hybridized to
Northern blots of cellular RNA extracted from the various cell lines grown
in growth factor assay medium as described. The 4.0 kb transcripts
observed represent overexpression of this gene. The position of the
ribosomal RNA bands are indicated on the right for size comparison. Cell
line CT-l is shown to express the highest level of this transcript of all

cell lines tested.

142

53288-883m Away
u<=m .mfiv
huuaH> Amfiv
Nmo-=m Aofiv
emo-=m Amy
Luz va
H-hu ARV
cmoflez Amv
8H-»== Amy
~m-h=: A38
ax Amy

meow ANV
NHmmz A_v

4.0 kb-

 

"I

p
0
r
c
4|
m

lobul

”3

143

Figure 6. Overexpression of basic FGF by the spontaneously transformed
cell line VIP:FT. 32P-labelledprobe DNA was hybridized to Northern blots
of cellular RNA extracted from the various cell lines grown in growth
factor assay medium as described. The 7.0 kb and 3.7 kb transcripts
observed represent overexpression of this gene. The position of the
' ribosomal RNA bands are indicated on the right for size comparison.
Overexposure of this blot did not reveal expression of this growth factor

by the other cell lines tested.

144

Nfimaz AHV
meow ANV

ax any
NH-»== A88
8H-»== Amy
omofiez A88
H-Hu Any
Huz Amv
388-:m Amy
www-3m Acme
em 82> AHHV
u<=m ANAV
Ezeam-moom AMHV

7.0 kb-

-285

3.7 kb-

145

Figure 5. Northern blot analysis of the expression of PDGF-A in non-

transformed, tumor derived, or transformed human fibroblasts. 32P-labelled
probe DNA was hybridized to Northern blots of cellular RNA extracted from

the various cell lines grown in growth factor assay medium as described.

The 2.8, 2.3. and 1.9 kb transcripts observed represent the expression of

this gene. The position of the ribosomal RNA bands are indicated on the

right for size comparison.

Rapidly growing, malignant cell lines, HuT-12, HuT-14, SHAC, and HT1080

overexpress this gene. Cell line SW 684 shows increased signal but the

lane is overloaded. Normal cell lines are shown to express this transcript

at low levels as seen by the overexposure of the blot.

146

Ezamm-meom Amfiv
u<=m ANHV
Ha 8H> RHHV
Nam-=m Aoflv
«no-3m Amy
Luz va
A-LU ARV
swamp: on
eH-e=I Amv
NH-P=: A38
ax Amy

meow “my
Nmmaz “my

2.8 kb -

2.3 kb -

-185

1.9 kb -

 

 

e
r
U
S
0
p
X
e
P.
e
V
0

 

1n-

lobul

m1crog

m.'ee

a

147

Figure 7. Northern blot analysis of the expression of N-fig; in non-
transformed, tumor derived, or transformed human fibroblasts. 32P—labelled
probe DNA was hybridized to Northern blots of cellular RNA extracted from
the various cell lines grown in growth factor assay medium as described.
The 5.0 kb and 2.0 kb transcripts are diagnostic for the expression of N-
fig;. The position of the ribosomal RNA bands are indicated on the right

for size comparison. N-fig; is overexpressed by cell lines CT-l, SHAC, and

8387.

148

nwmm Angv
EzcmmummAm ANHV

u<=m AHHV

www-3m Aoflv
888-:m Amy
882 Amy
”-88 ARV
omofihz Amy
¢~-e== Amy
N72: 2:
ax Amy

meow Amy
Nommz Amy

1
_-

2.0 kb

 

   

.

B2

m1cro-
globulin

149

Figure 8. Northern blot analysis of the expression of IGF-II and c-my; in
non-transformed, tumor derived, or transformed human fibroblasts. 32P-
labelled probe DNA was hybridized to Northern blots of cellular RNA
extracted from the various cell lines grown in growth factor assay medium
as described. Expression of the 6.0 kb transcript corresponding to IGF-II
was seen only in cell line, SHAC. Rehybridization of this blot with probe
complementary to c-my; is also shown. The position of the ribosomal RNA

bands are indicated on the right for size comparison.

150

Rama Ammv

Encomnwmam ANAV

u<zm fififiv

Nam-zm Aofiv

emmnzm
Huz
“the
owouh:
vunhzx
Nfiuhzz
ax

mmgm
Nfimuz

Amy
Amy
ARV
Amv
Amy
Aev
Amy

ANV
AHV

IGF-II

6.0 kb-

 

HUT

 

 

B
m1cro-

n
.11
1

u
b

O
1

g

151

Figure 9. Results of hybridization with a probe complementary to K-fig;
gene. 32P-labelledprobe DNA was hybridized to Northern blots of cellular
RNA extracted from the various cell lines grown in growth factor assay
medium as described. The 3.8 kb transcripts observed is consistent with
expression of the K-fig; gene however we did not see 5.5 kb transcripts
which have been reported. In addition we also observe bands corresponding

to approximately 2.5 and 1.7 kb. The position of the ribosomal RNA bands

are indicated on the right for size comparison.

152

S

8

1
.

        

Nam» Amfiv
Ezeom-mmom ANHV
u<=m “HHV

Hm 82> Aoav

Nmanzm
cam-3w
Huz
“the
oonh:
enuhzz

Nfinhzx

Eaemmummmm
Nfiwuz

_
8b
uk
3

 

MICHIGAN STATE UNIV. LIBRARIES
llHI”WIWIIWWHI‘llIllWillWIWIIHWI
31293007904398