*1
Rig

1, .2 a:
V

 

W383

[moms/w STATE UNNER

I l! I!

l! j/I/l/IZI Ill/l!!!

/ Ill/IIII/l/l//I//l’/I//I////I/Il

93 01015 4270

This is to certify that the
dissertation entitled

Analysis by Cell Fusion of the Loss of
Tumor Suppressor Functions at Specific Stages
in the Malignant Transformation of
Human Fibroblasts in Culture

presented by

P. Ann Ryan

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Microbiology

 

/fl(< ~

Major professor

Date June l5, l994

 

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

 

 

LIBRARY
Mlchlgan State
Unlverslty

 

 

 

PLACE DI RETURN BOX to remove thle checkoum your record.
To AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1

 

ANALYSIS BY CELL FUSION OF THE LOSS OF
TUMOR SUPPRESSOR FUNCTIONS AT SPECIFIC STAGES
IN THE MALIGNANT TRANSFORMATION OF
HUMAN FIBROBLASTS IN CULTURE

By
P. Ann Ryan

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Microbiology
1994

ABSTRACT
ANALYSIS BY CELL FUSION OF THE LOSS OF
TUMOR SUPPRESSOR FUNCTIONS AT SPECIFIC STAGES

IN THE NALIGNANT TRANSFORMATION OF
NUNAN FIBROBLASTS IN CULTURE

By
P. Ann Ryan

The observation that fusions of infinite life span cells with finite life span
cells produce hybrid cells with finite life spans led to the conclusion that an
infinite life span in culture is a recessive trait resulting from loss of the
function of suppressor genes. Furthermore, finding that certain pairs of
infinite life span cells can complement each other to yield finite life span
hybrids allowed 19 infinite life span cell lines to be assigned to four
imnortality complementation groups (1). Ne fused a near diploid, morphologically
normal, infinite life span cell strain, designated MSU-1.1, with its finite life
span precursor cell strain and obtained finite life span hybrids, as expected if
infinite life span in culture is a recessive trait. However, 14 of the 14
hybrids from our fusions of MSU-1.1 cells with representative cell lines from
each of the four immortality complementation groups, and 38 of the 39 hybrids
from our fusions of infinite life span cells that have been reported to
complement each other, failed to exhibit finite life spans. This result suggests
that infinite life span cells cannot complement each other to yield finite life
span hybrids. He obtained evidence that long-term dual drug selection can be
deleterious to hybrid cells, indicating that the cell death of such hybrids
observed in other studies may have resulted from the cytotoxic effect of long-
term drug selection, rather than from senescence.

Cell fusion studies designed to demonstrate suppression of res-induced

tumorigenicity must control for suppression of infinite life span since loss of

infinite life span will necessarily result in loss of tumorigenicity. Such
controlled studies have previously been performed with rodent cells but not with
human cells. We fused infinite life span MSU-1.1 cells with ras-transformed
malignant cells that were derivatives of MSU-1.1 cells. The degree of
suppression of tumorigenicity varied among the hybrid strains and did not
correlate’with the levels of expression of ras protein, which increased when the
cells were grown in athymic mice and decreased when they were grown in tissue

culture dishes.

1. Pereira-Smith, 0.M., and Smith, J.R. (1988). Genetic analysis of indefinite
division in human cells: identification of four complementation groups.

Proc. Natl. Acad. Sci. USA 8526042, 1988

This work is dedicated to...

Will, my son, who throughout my eight years in the Medical Scientist
Training Program has helped to keep me in touch with the bright side of
life, from its simplest delights to its deepest Joys.

Mary, whose love and companionship have brought me a renewed sense of hope

and direction.

Linda, whose nurturing has made this accomplishment possible for me.

iv

ACKNONLEOGNENTS

My deepest thanks go to Dr. Justin McCormick for his teachings,
guidance, encouragement and support. The opportunities and challenges
that he consistently provided for me enriched my educational experience
beyond my highest expectations. Dr. Veronica Maher also contributed
imensely to my graduate education. Her keen mind, sharp skills,
enthusiastic spirit, and countless hours of assistance helped greatly to
guide me through the challenging process of scientific publishing.
Furthermore, I want to express my heartfelt thanks to the other members of
my committee, Dr. Richard Schwartz, Dr. Maria Patterson, and Dr. David
Kaufman, for their time and helpful advice.

My co-workers at the Carcinogenesis Laboratory have felt like a
family to me. They have always been personally, as well as
professionally, supportive of me. So many have helped me in such a
variety of ways that it is not possible to list them all here. A few to
whom I wish to give special mention are Suzanne Kohler, Lonnie Milam, Dr.
Jeanette Scheid, Dr. Dennis Fry, Dr. Chia Miao Mah, Connie Williams,
Bethany Heinlen, Cindy Wilson, Ena Zaccagnini, and Alicia Holzhausen.

Thank you. I will miss you all.

TABLE OF CONTENTS

LIST OF TABLES .................................................... ix
LIST OF FIGURES ................................................... x
INTRODUCTION ...................................................... 1
References .................................................. 6
CHAPTER I. LITERATURE REVIEW ..................................... 7
A. Dominantly Acting Dncogenes - a Brief Overview .......... 7
1. Dncogenes in acutely transforming retroviruses.... 7
2. Protooncogenes .................................... 8
3. A prototypic oncogene: ras ........................ 11
4. Mechanisms of activation .......................... 13
8. Tumor Suppressor Genes .................................. 14
1. Early evidence .................................... 14
2. A prototypic human tumor suppressor gene: RBI ..... 16
3. Identification through the study of familial
cancers ........................................... 19
3.1 Hilms’ tumor ............................... 20
3.2 Familial adenomatous polyposis coli ........ 24
4. Identification by loss of heterozygosity in
tumor cells ....................................... 27
4.1 Colon carcinomas ........................... 27
a. The DCC gene .......................... 28
b. The p53 gene .......................... 29
4.2 Other malignancies ......................... 32
C. Contributions from Cell Hybrid Studies .................. 32

1. Chromosomal assignment of tumor suppressor genes.. 33

1.1 flamster_gell_x_numan_gg11 hybrids .......... 34
a. gun 9g]] X HQE hybrids ................ 34

vi

b. Bfl§_§§ll_x_flflfi hybrids ................ 35

1.2 flgmag_ga11_x_ngmag_§all hybrids ............ 36
a. flaLa gal 1 X HOE hybrids ............... 36
b. HIIQQQ gall X HOE hybrids ............. 38

c. Other malignant gal ] X garmal gall
hybrids ............................... 38

2. Suppression of specific transformed phenotypes.... 39

2.1 Anchorage independence ..................... 39

2.2 Growth factor independence ................. 40

2.3 Infinite life span ......................... 40

2.4 Other transformed phenotypes ............... 41

3. Complementation analysis .......................... 42

3.1 Complementation of tumorigenicity .......... 42

3.2 Complementation of infinite life span ...... 43

References .................................................. 45

CHAPTER 11 Failure of Infinite Life Span Human Cells
from Different Immortality Complementation

Groups to Yield Finite Life Span Hybrids .............. 62
Abstract .................................................... 63
Introduction ................................................ 64
Materials and Methods ....................................... 67

Cells used ............................................ 67

Cell culture .......................................... 67

Transfection .......................................... 70

Selective media ....................................... 70

Cell fusion, selection, and initial expansion ......... 72

Determination of the life span of the hybrids ......... 72

DNA flow cytometry .................................... 72
Results ..................................................... 73

Life span analysis of hybrids formed by fusions
with MSU-1.I cell strains ....................... 73
Determining the ability of representatives of the
established complementation groups to

complement each other ........................... 76
Testing the hypothesis that parental cells
survived selection .............................. 77

Testing the hypothesis that our infinite life
span populations represented overgrowth
of senescing populations by revertants .......... 79

vii

Testing the effects of long term drug selection

on senescence ......................................... 84
Discussion .................................................. 89
Acknowledgments ............................................. 97
Literature cited ............................................ 98

Chapter III Suppression of the Tumorigenicity of H-, N-, or
v-K-ras Oncogene-Transformed Malignant Human
Fibroblast Strains by Fusion with their Infinite
Life Span, Non-Tumorigenic Parental Strain, MSU-1.1.. 101

Introduction ................................................ 102
Materials and Methods ....................................... 106
Cells and cell culture ................................ 106
Cell fusion ........................................... 106
Assay for tumorigenicity .............................. 106
Immunoprecipitation of p21 ras protein ................ 106
Results ..................................................... 108

Fusion of malignant MSU-1.1-ras cells with the

finite life span, non-tumorigenic MSU-1.1

precursor cells ................................. 108
Expression levels of ras protein in parental cells,

hybrid cells, and tumor-derived hybrid cells.... 111

Discussion .................................................. 116

References .................................................. 118

viii

LIST OF TABLES

TABLE Page
Chapter II
1. Human cells used in fusion expiriments .......................... 68

2. Proliferative potential of hybrids formed by fusion
of various cell strains with MSU-1.1 CHAT’Hgr cells ............. 74

3. Proliferative potential of cell hybrids formed by fusions
among cells from the four immortality complementation groups.... 78

4. Effects of short-term versus long-term drug
selection on the life span of the hybrid cells ................... 88

Chapter III

1. Cell strains .................................................... 107
2. Tumorigenicity of cell strains .................................. 109
3. Suppression of tumorigenicity in hybrid strains ................. 110
4. The ras protein expression of MSU-1.1-N-ras-8T

cells measured at various cell passages ......................... 115

ix

LIST OF FIGURES
FIGURE Page

Chapter II

1. Growth curves of hybrids formed by fusion of
HTIOBO-CHAT’Ourr cells with cells reprepresentative
of immortality complementation groups A, 8, and C ............... 80

2. Growth curves of hybrids formed by fusion of
A1698-CHAT‘Ourr cells with cells reprepresentative
of immortality complementation groups A, 8, C, and D ............ 82

3. Growth curves of hybrids formed by fusion of
HTIOBO-CHAT‘Ourr cells with HeLa cells .......................... 85

Chapter III

1. H-ras protein expression, relative to that of
MSU-1.1-Hg'CHATs cells, for H-ras-transformed
MSU-1.1 parental and hybrid cells ............................... 112

2. N-ras protein expression, relative to that of
MSU-1.1-Hg'CHATs cells, for N-ras-transformed
MSU-1.1 parental and hybrid cells ............................... 113

INTRODUCTION

Epidemiologic analysis of the frequency and age of occurrence of various
cancers, genetic analysis of animal and human tumors, and experimental induction
of malignant cells both in vivo and in vitro have all contributed to our
understanding of the genetic basis of carcinogenesis. The results of numerous
and varied studies indicate that carcinogenesis is a multistep process in which
a cell accumulates, in a stepwise fashion, at least five or six genetic changes
that confer upon the cell the various altered growth properties that together
allow it to fornla malignant tumor. This process is thought to occur by a series
of sequential mutations and clonal selections. According to this theory; a normal
cell acquires a mutation that gives it a proliferative advantage over surrounding
cells. Within the resulting population of altered cells, a single cell incurs
a second mutation that further enhances its proliferative potential. Among the
progeny of this doubly mutated cell, a third advantageous transforming mutation
occurs, and so on. As a result of this process, a cell that has gained the
combination of altered growth characteristics necessary for it to form a
malignant tumor eventually arises.

In discussions of the multistep process by which cells become malignant,
it is necessary to carefully define the terms referring to the different stages.
In this thesis, the term "transformed" will refer to cells that have taken on one
or more characteristics of cancer cells. "Transformation" will refer to a change
from the normal to the transformed phenotype. "Tumorigenic" or "malignant" will

refer to the ability of cells to form progressively growing tumors that typically

1

are found to invade normal tissues.

Two main genetic mechanisms contribute to the process of carcinogenesis.
One is the activation of dominantly acting oncogenes, which are genes that
promote cell proliferation; the other is the inactivation of tumor suppressor
genes, which are genes that normally act to limit cell proliferation. Numerous
genes of both types have been identified and characterized. The relative
contributions to human carcinogenesis of these two classes of genes and the
interactions among these genes are currently the subject of intense study.

One method for assessing the relative contributions of these two classes
of genes is illustrated by the work of Vogelstein and his colleagues on the
genetic alterations that occur during colon carcinogenesis (Vogelstein et al.,
1988). Colorectal tumors progress through a series of five easily recognizable
clinical stages: from hyperplastic epithelial cells; through three stages of
benign adenomas of increasing size, dysplasia, and villous content; to malignant
carcinomas. Using tissue taken from human hosts, Vogelstein and his colleagues
analyzed the genetic alterations of cells at each of these stages. They found
that the ras oncogene was activated in a large percentage of colorectal
neoplasms, and that this change usually occurred at an early stage of
transformation. They also found evidence for the inactivation of tumor
suppressor genes on chromosomes 5, 18, and 17, with the losses occurring,
respectively, at early, intermediate, and late stages of transformation.

Another way to study the genetic alterations that contribute to
carcinogenesis is by inducing this process in cells in culture and analyzing the
phenotypic and genetic changes that occur during individual steps of
transformation. The study of cells that have been transformed in culture has
many advantages over the study of cells that have been obtained from tumors.

Transformation of cells in culture allows the direct observation of the temporal

3

sequence of the transformation stages, whereas the sequence of the changes in
vivo can only be inferred. Transformation of cells in culture also allows the
comparison of an altered cell with its inmediate precursor cell. When one
examines cells that have become transformed in an animal, however, one cannot
know for sure which cell gave rise to a particular p0pulation of cells.
Furthermore, transformation of cells in culture has the obvious value of allowing
introduction of a specific altered gene, e.g., transfection of a known oncogene,
in order to study the transforming effects of the gene.

In our laboratory we are studying the phenotypic and genetic changes that
occur during the transformation, in culture, of normal foreskin-derived human
fibroblasts, through at least two partially transformed intermediates, to the
fully malignant state (McCormick and Maher, 1991). The normal fibroblast LGI was
transfected with the myc oncogene. One of the myc-transfected strains gave rise
to the morphologically normal, diploid, immortal strain designated MSU-1.0. This
strain spontaneously gave rise to a partially growth factor independent , near
diploid variant harboring two marker chromosomes. This second intermediate,
designated MSU-1.1, has given rise to numerous malignant derivatives. Some of
these have arisen spontaneously. Others have been induced by transfection with
oncogenes or have arisen following carcinogen treatment.

The goal of my doctoral studies was to determine if inactivation of tumor
suppressor genes contributed to transformation at specific steps within the MSU-I
lineage. I used the method of cell fusion between cells from separate stages of
the lineage to make this determination. When a cell having a particular
transformed characteristic is fused with a cell lacking this characteristic, the
phenotype of the hybrid cell provides information about the genetic alteration
that induced the transformed phenotype. If the hybrid cell exhibits the

transformed phenotype, one assumes that the activation of a dominantly acting

4
oncogene brought about the transformed phenotype. The oncogene is supplied to
the hybrid by the transformed parental cell and still acts in a dominant fashion
to maintain the transformed phenotype in the hybrid. If the hybrid cell exhibits
the non-transformed phenotype, one assumes that transformation occurred as a
result of inactivation of both copies of a tumor suppressor gene. A functional
tumor suppressor gene is supplied to the hybrid by the non-transformed parent
cell causing the suppression of the transformed phenotype in the hybrid.

Chapter I of this thesis gives a brief overview of the discovery of the
dominantly acting oncogenes. The mechanisms by which these genes become
activated are discussed and are illustrated by an example of a prototypic gene
from this class. The primary focus of this chapter, however, is on the second
class of genes - tumor suppressor genes. The early evidence for the existence
of these genes is presented, and the identification and characterization of
several specific suppressor genes is discussed in detail. Particular attention
is devoted to contributions to this area of cancer research that have come from
cell fusion studies, since this methodology served as the basis of many of the
experimental studies I carried out during the research described.

Chapter II is a manuscript by'P. Ann Ryan, Veronica M. Maher, and J. Justin
McCormick which was published in Journal of Cellular Physiology 159, 151-160
(1994). The manuscript describes the results of my cell fusion studies showing
that the infinite life span of MSU-1.1 is a recessive trait, as has been found
for all other infinite life span cells studied (Bunn and Tarrant, 1980;
Muggleton-Harris and DeSimone, 1980; Pereira-Smith and Smith, 1983), but that,
contrary to earlier reports (Pereira-Smith and Smith, 1988), infinite life span
cells are not able to complement each other to yield finite life span hybrids.
The discussion section of Chapter II is a somewhat longer version than that

included in the published manuscript.

5
Chapter III describes the results of cell fusion studies suggesting that
ras oncogene-induced transformation of MSU-1.1 to the malignant state requires
loss of a suppressor function that is present in MSU-1.1 cells. Examination of
the ras protein expression of the hybrid cells and the malignant parent cells
suggested that ras expression is increased by the growth of cells in athymic mice
and decreased by the growth of cells in tissue culture dishes. Further studies

will be undertaken by others to confirm these preliminary results.

REFERENCES

Bunn, C.L., and Tarrant, G.M. (1980). Limited lifespan in somatic cell hybrids
and cybrids. Experimental Cell Res., 127, 385-396.

McCormick, J.J., and Maher, V.M. (1991). Malignant transformation of human
fibroblasts in vitro. In Neoplastic Transformation in Human Cell Culture, J.S.
Rhim and A. Dritschilo, ed. (Totowa, NJ: The Humana Press Inc.), pp. 347-357.

Muggleton-Harris, A.L., and DeSimone, D.W. (1980). Replicative potentials of
various fusion products between WI-38 and SV40 transformed WI-38 cells and their
components. Somatic Cell Genetics, 6, 689-698.

Pereira-Smith, 0.M., and Smith, J.R. (1983). Evidence for the recessive nature
of cellular immortality. Science, 221, 964-966.

Vogelstein, 8., Fearon, E.R., Hamilton, S.R., Kern, S.E., Preisinger, A.C.,
Leppart, M., Nakamura, Y., White, R., Smits, A.M.M., Bos, J.L. (1988). Genetic
alterations during colorectal-tumor development. New'England Journal of Medicine
319, 525-532.

CHAPTER I

LITERATURE REVIEW

A. Doeinantly Acting Oncogenes - a Brief Overview

1. Oncogene: in acutely transforeing retroviruses

A major contribution to our understanding of the genetic basis of
cancer comes from studies of acutely transforming retroviruses. Theselare
RNA viruses capable of rapidly inducing tumors in infected animals.
Examples are lavian leukemia viruses, murine leukemia viruses, mouse
mammary tumor viruses, feline leukemia viruses and feline sarcoma viruses
(reviewed by Weiss et al., 1982). The RNA tumor virus first identified
was the Rous sarcoma virus (RSV). Rous (1911) found that a cell-free,
bacteria-free extract from a spontaneous chicken sarcoma was capable of
inducing tumors when inoculated into chickens. This finding established
a virus as the etiological agent of the tumors. The infecting particle
was identified as a virus that contained RNA (Claude et al., 1947;
Gaylord, 1955; Crawford & Crawford, 1961). Martin (1970) isolated a
temperature sensitive mutant RSV that replicated normally at both
restrictive and permissive temperatures but did not transform infected
cells at the nonpermissive temperature. This showed that a transforming
protein was encoded by the mutated RSV gene, which was named "src", for
sarcoma. A c-DNA fragment bearing src alone, and no other RSV genes, was
found to be capable~of transforming fibroblasts in culture (Martin, 1970).

In a similar manner, the genes responsible for the transforming properties

7

8

of other acutely transforming RNA tumor viruses were identified. Examples
of such genes are fps, yes, and ros from avian sarcoma viruses; myc, erb,
and myb from avian leukemia viruses; and mos, ras, fes, rms, and sis from
various mammalian sarcoma viruses (reviewed in Bishop and Varmus, 1982).

Subsequently, homologs of these viral cancer genes were found in the
genomes of various animals, including species as diverse as fish, birds,
mammals, and flies (Stehelen et al., 1976; Spector et al., 1978; Shilo and
Weinburg, 1981). Convincing evidence suggested that these transforming
genes were endogenous cellular genes that had been acquired by viruses
from the animal hosts, rather than being viral genes that had become
integrated into the animal genome. For example, if a homolog of a
particular viral transforming gene was found in an animal genome, it was
present in all members of the species, whereas the transforming gene was
usually restricted to a single viral strain. Each cellular homolog was
found to reside at a constant genetic locus within an animal species,
whereas retroviruses were found to integrate at diverse positions within
a host genome. The cellular homologs were found to contain introns,
whereas the viral transforming genes lack such material and, therefore,
probably represent cellular cDNA that has become integrated into the viral

genome.

2. Protooncogenes

As more viral cancer genes were identified and the cellular origins
of each gene was shown, it became clear that normal cells contain numerous
genes that can contribute to cellular transformationlwhen they are altered
in specific ways. The normal cellular forms of these genes became known

as protooncogenes, and are designated by the prefix "c-" for cellular,

9
e.g., c-src. The altered viral forms capable of transforming cells are
called oncogenes and are designated by the prefix "v-" for viral, e.g., v-
src.

Although retroviral transformation is not a common mechanism of
human carcinogenesis, the identification of virally-activated oncogenes
gave researchers a valuable tool with which to study human cancer because,
with few exceptions, sequences homologous to the viral oncogenes were also
found in the human genome. Moreover, cells from many human tumors were
found to contain activated forms of these genes. For example, Eva et al.
(1982) used the technique of Northern Blotting to demonstrate that a human
homolog to the chicken v-myc oncogene is overexpressed in a human sarcoma
cell line and in two human carcinomas, and that a human homolog to the
wolly monkey v-sis oncogene is overexpressed in certain human sarcomas and
glioblastomas. Additional oncogenes were found to reside at common
chromosomal translocations seen in specific types of tumors, e.g., bcl-I
at the translocation site in a lymphocytic leukemia cell line (Tsujimoto
et al., 1984), or to be amplified in the genomes of tumor cells, e.g., N-
myc in neuroblastomas and retinoblastomas (Schwab et al., 1983).

Another widely used means of identifying activated oncogenes
utilizes transfection of tumor cell DNA into established rodent
fibroblasts, most commonly mouse NTH/3T3 cells. Neoplastic transformation
of the recipient cell occurs upon acquisition of a DNA fragment containing
an activated oncogene, allowing isolation and analysis of the transfected
gene (Shih et al., 1979a). The human N-ras oncogene was identified in
this manner (Marshall at al., 1982; Hall et al., 1983) As was true for
the oncogenes that were first identified in viruses, human oncogenes

discovered by this method were also found to be altered forms of normal

10
human genes.

The discovery of protooncogenes prompted intense study of their
function. Why would cells carry genes predisposing them to cancer? As
the genes were cloned and their protein products were analyzed, it became
clear that, rather than representing hapless sequences waiting for the
chance to become destructive, protooncogenes code for-essential proteins
that have specific cellular functions (reviewed in Hunter, 1991). For
example, the oncogene v-srC' was found to code for a protein that
phosphorylates tyrosine (Brugg and Erickson, 1977; Hunter and Sefton,
1980). It is now known that this is only one in a family of more than a
dozen oncogenes, including neu and erbB (Downward et al., 1984), that code
for receptor and nonrecepter membrane associated protein-tyrosine kinases.
These oncogenes, along with a family of oncogenes coding for membrane
associated G-proteins [e.g., ras (Hurley et al., 1984)], and another group
coding for cytoplasmic protein-serine kinases. function as signal
transducers in growth factor signalling pathways. Other oncogenes code
for protein growth factors [e.g., sis, which produces an altered form of
the B-chain of platelet-derived growth factor (Doolittle et al., 1983)].
Finally, another group of oncogenes encodes nuclear proteins. Many of
these (e.g., jun, fos, and myb) act as transcription factors that are
induced when resting cells are treated with mitogens. They have been
shown to initiate expression of genes involved in cell replication
(reviewed in Seemayer and Cavenee, 1989; Hunter, 1991).

Each of these groups of oncogenes (i.e., growth factors, signal
transducers, and transcription factors) plays a critical role in the
complex process of cell replication. Therefore, controlled expression of

the normal forms of these genes results in normally regulated cell

11
proliferation. However, when these genes are expressed in excessive
amounts (see discussion of v-myc and v-sis, p.9) or in mutated forms (see

discussion of ras, pp.12-13), transformation can result.

3. A prototypic oncogene: ras

A description of the discovery of a prototypic oncogene, the
isolation of its protein product, and the determination of its function
will illustrate these concepts. In the 1960’s, two closely related
defective viruses, which became known as murine sarcoma viruses, were
isolated from sarcomas that developed in rats following inoculation with
mouse leukemia viruses (Harvey, 1964; Kirstin & Mayer, 1967). The sarcoma
viruses, although clearly derived from the leukemia viruses that had been
injected into the rats, differed substantially from the leukemia viruses.
The only protein found to be produced by these viruses was a 21,000 dalton
phospho-protein (Shih et al., 1979b).

The proteins from these two viruses were antigenically related
(Scheinberg and Strand, 1980) and had guanine nucleotide-binding
activities (Scolnick et al., 1979). A temperature dependent mutation in
the gene encoding for the protein demonstrated that it was responsible for
the transforming properties of the viruses (Shih et al., 1979c).
Fluorescent and ferritin-tagged antisera.were used to localize the protein
to the inner surfaces of plasma membranes in cells transformed by the
virus (Willingham et al., 1980).

The oncogenes encoding the transforming proteins were named Ha-ras
and Ki-ras, after the scientists who isolated the viruses from the [at
‘sarcomas. DNA probes specific for the ras genes hybridized to single-copy

DNA from uninfected rat cells (Langbeheim, 1980). This led to the

12
discovery that Ha-ras is derived from one of two closely related normal
rat cellular genes, and that Ki-ras is derived from a third gene that is
partially homologous to the other two (DeFeo et al., 1981; Ellis et al.,
1981).

The NIH/3T3 cell assay identified transforming genes from human
tumors that shared homology with the viral H- and K-ras oncogenes
(Krontris and Cooper, 1981; Shih et al., 1981; Pulciani et al., 1982; Der
et al., 1982). A third transforming ras gene, named N-ras, was also
identified by this method (Hall et al., 1983; Shimizu et al.,
1983a,1983b). As expected, these three human oncogenes were found to be
mutated forms of normal cellular genes (Chang et al., 1982; Davis et al.,
1983). The oncogenic mutations occurred only at a few restricted sites,
most commonly in codons 12, 13, 59, 60, 61, and 117 (reviewed in Varmus,
1984; Barbacid, 1987).

In humans, as in rats, the ras protein (p21-ras) was found to be
located on the inner plasma membrane surface and to have the ability to
bind guanine nucleotides. This suggested that p21-ras participates in
signal transduction from the cell surfaces. Investigations along these
lines revealed that p21-ras belongs to a family of G-proteins that
transduce signals from activated receptors of extracellular growth factors
to secondary intracellular messenger systems (reviewed in Haubruck and
McCormick, 1991; Valencia et al., 1991). GTP-bound p21-ras is the active
form of the protein, which is subsequently transformed into the inactive
GDP-bound form by hydrolysis of GTP.

The transforming mutated forms of ras yield proteins that are
insensitive to GTP-ase activating regulator proteins. Because the GTP is

not hydrolyzed, the oncogenic ras proteins remain in a permanently active

13
conformation (Trahey and McCormick, 1987). As a result, a cell containing
an oncogenic ras continues proliferating even in the absence of exogenous

growth factors.

4. Mechanises of activation

The ras gene is only one of many protooncogenes for which the normal
gene function and oncogenic activation have been explored in detail. The
protooncogenes are of two types that differ in their general mechanisms of
activation (reviewed in Weinburg, 1989). Protooncogenes of one type,
exemplified by ras, are expressed in constant amounts. The protein
products of these protooncogenes, however, oscillate between an active and
an inactive state. The transitions between the active and inactive states
are induced by other biochemical regulators, e.g., phosphatases and
kinases. The oncogenic forms of these genes generally produce proteins
that are locked in the active state. Because these proteins participate
in the growth factor signalling pathway, their constitutive activity
causes cell replication to proceed unchecked.

Protooncogenes of the other type, on the other hand, generally are
regulated by modulatable expression. Oncogenic mutations in these genes,
e.g., sis and erbB, usually uncouple them from their normal regulators,
leading to constitutive expression of the gene. Elevated amounts of their
protein products, usually transcription factors, enhance the expression of
genes whose products are critical to cell growth and differentiation.
Increased production of these critical proteins leads to excessive cell
replication.

Despite the differences in activation between these two types of

oncogenes, they have a very important feature in common. For both, the

14
oncogenic form of the gene acts in a dominant fashion over the wild type
alleles. Both constitutively active and constitutively expressed
oncoproteins promote unbridled cellular proliferation even when
accompanied by their normal counterparts, except in certain instances of
abnormally high expression of the wild type allele. Therefore, these

genes are called dominantly acting oncogenes.

8. Tuner Suppressor Genes

1. Early evidence

In the excitement over the discovery of dominantly acting oncogenes
and the fervor with which the identities and functions of their protein
products were pursued, relatively little attention was paid to indications
that an entirely different kind of cancer gene existed. In 1969, a year
before the first dominantly acting oncogene was identified, Henry Harris
and George Klein published results from cell fusion experiments that
suggested that some cancer genes were recessive. These scientists fused
highly malignant mouse cells with mouse cells that were low in malignant
potential and discovered that the resultant hybrids were suppressed in
their ability to form tumors (Harris et al., 1969; Wiener et al., 1974).
The suppressed hybrids, however, rapidly reverted to the malignant state,
a phenomenon that made it difficult to observe suppression in earlier
investigations (Barski and Cornefert, 1962). It was noted that this
reversion was accompanied by loss of substantial numbers of chromosomes.

The results of these cell fusion experiments and of similar

15

corroborating studies (Silagi, 1967; Kao and Hartz, 1977; Sager and Kovac,
1978) led to the following conclusions: The mutant genes responsible for
the tumorigenic phenotype in the malignant cells are recessive.
Therefore, mutation of both of the normal alleles of these genes is
required for malignant transformation. Fusion of a malignant cell with
one that still possesses a normal copy of the gene results in suppression
of the tumorigenic phenotype. The chromosomal instability of the rodent
cell hybrids allows rapid loss of the chromosomes bearing the normal
suppressing alleles. When this happens, tumorigenicity is again
expressed.

Because expression of tumorigenicity was overridden by the presence
of the normal genome in the tetraploid hybrids, the mutant genes
responsible for the tumorigenicity in the malignant parent were labeled
"recessive oncogenes". "Recessive" as used here is similar but not
identical to the term "recessive" as used in Mendalian genetics. The
normal alleles, having the ability to suppress the malignant phenotype,
were dubbed "tumor suppressor genes" or "anti-oncogenes".

The idea that some genes act to suppress tumorigenicity was not new.
In 1964, Stoker reported that the growth of polyoma-transformed cells was
suppressed by surrounding normal cells, suggesting that normal cells
produce tumor suppressor molecules that they transfer to neighboring
cells. This same phenomenon was later reported for rodent fibroblasts
transformed by transfection with both the myc and the ras oncogenes (Land
et al., 1986). In studies aimed at identifying these molecules, growth
regulatory polypeptides were found that inhibited the replication of
certain cells. examples include TGF-B, tumor necrosis factor, tumor

inhibitory factors 1 and 2, interferons, interleukins, and oncostatin

16

(Resnitzky et al., 1986; Yardin and Kimchi, 1986; Zarling et al., 1986;
Newmark, 1987; Takehare et al., 1987; Sporn and Roberts, 1988). These
polypeptides could be thought of as products of tumor suppressor genes.

While scientists were struggling to understand the phenomenon of
tumor suppression in vertebrates, rapid progress was being made in
identifying specific tumor suppressor genes in a less complex organism,
i.e., the fruit fly Dr050phila melanogaster. By 1982, 25 recessive
Drosophila genes had been implicated in cancer of flies (Gateff, 1982).
In 1985, the most extensively studied of these, the lethal(2) giant larvae
gene [l(2)gl], was cloned. Recessive mutations in this gene resulted in
uncontrolled cell proliferation and death of the animal (Melcher et al.,
1985). Introduction of the cloned gene into cells that were deficient at
this locus suppressed malignancy, confirming the hypothesis that l(2)gl is

a tumor suppressor gene (Opper, 1987).

2. A prototypic huaan tulor suppressor gene: RBI

Closely following the>cloning of the l(2)gl gene came the cloning of
the most extensively studied human tumor suppressor gene, the
retinoblastoma gene. Retinoblastoma is a pediatric ocular cancer. As is
characteristic of several other kinds of human cancers, retinoblastoma
occurs both sporadically and as an inherited disease. In individuals with
the hereditary form of the disease, multiple tumors form, commonly in
both eyes, and the tumors usually arise prior to birth. The sporadic form
of retinoblastoma differs from the hereditary form in that affected
individuals usually develop only'one tumor, and the tumors are not present
at birth, but appear during the first few years of life.

To explain the basis for this difference, Knudson (1971) proposed

17

that two mutations are required for transformation of the proliferating
retinoblast to the malignant state. In hereditary retinoblastoma, one
mutation occurs in the germline, making every cell in the body a candidate
for malignant transformation by the second hit. Retinoblasts proliferate
during the first few years of life until there are approximately 107 cells
in the retina. Because the somatic mutation rate is approximately one per
105 to 10‘ cell divisions, retinal development provides ample chance for
inheritants of a germline mutation to incur a second mutation. In
sporadic retinoblastoma, however, both mutations are somatic, with the
second occurring in one of the progeny of the cell that acquired the first
mutation. This accounts for the relative rarity of sporadic tumors, their
occurrence in only one eye, and the longer length of time required for
their development.

When scientists examined the genetic alterations of retinoblastoma
cells, they found strong evidence in support of Knudson’s hypothesis.
Karyotypic analysis revealed frequent deletions in band 14 of the long arm
of chromosome'13 in retinoblastoma cells from both sporadic and congenital
tumors (Franke, 1978; Balaban, 1982; Benedict, 1983). This implicated
13q14 as the location of one of the disease-associated mutations and
raised the possibility that the mutation inactivates the gene and leads to
loss of some critical growth suppressing function. Restriction fragment
length polymorphisnistudies showed that the second mutation leading to the
development of retinoblastomas is loss of the remaining wild type allele
at the 13q14 locus (Cavenee, 1983; Sparkes, 1983). This verified that the
locus contains a tumor suppressor gene, and that loss of the function of
both alleles leads to tumorigenicity. Chromosome walking from a known

cloned, tightly linked gene allowed cloning and sequencing of the

18

retinoblastoma susceptibility gene, R81 (Friend et al., 1986; Fung et al.,
1987; Lee et al., 1987a).

The»RBl genelencodes a 105 kiloDalton nuclear phospho-protein (p105-
RBl) that appears to be expressed in all human tissues (Friend et al.,
1986; Lee et al., 1987b). Consistent with the homozygous alteration of
the gene in retinoblastoma cells is the fact that all retinoblastoma cell
lines and short term cultures lack production of p105-RBI (Horowitz et
al., 1990). Introduction of a functional R81 gene into retinoblastoma
cells suppressed tumorigenicity, providing confirmation that the gene
functions as a tumor suppressor (Huang, 1988; Bookstein, 1990).

Investigations into the phosphorylation of p105-RBI led to clues
regarding its function. The phosphorylation status of the protein
fluctuates regularly during the cell cycle (Chen et al., 1989; Buchkovich
et al., 1989; DeCaprio et al., 1989; Ludlow et al., 1990). The p105-Rb of
cells in the 60- and GI-phase is unphosphorylated. It becomes
phosphorylated, possibly by cdc-2 kinase; when the cells are stimulated to
divide, remains phosphorylated throughout S-phase, and is dephosphorylated
during mitosis (M-phase). This suggests that the unphosphorylated form of
p105-Rb inhibits cell division, and that phosphorylation of the protein
inactivates it, thereby allowing cell division. The unphosphorylated
protein apparently blocks passage from the Gl-phase to the S-phase of the
cell cycle by complexing with and inhibiting the function of the
transcription factor E2F, which normally can activate important
proliferation genes (Chellappan et al., 1991; Bagchi et al., 1991;
Chittenden et al., 1991; Hamel et al., 1992; Hiebert et al., 1992).

Evidence indicates that inactivation of the RBI gene also

participates in the genesis of cancers other than retinoblastoma.

19
Patients with a germline alteration in an RBI allele are also predisposed
to developing osteosarcomas and other soft tissue sarcomas in their adult
years (Derkinderen et all, 1988). Examination of these tumors and a wide
variety of tumors from patients that do not have a constitutional defect
in RBI showed that homozygous alterations in this gene are found with a
high frequency in osteosarcomas, soft tissue sarcomas, small cell lung
carcinomas, and breast cancers. They also are found in bladder cancers,
glioblastomas, leukemias, and squamous cell carcinomas, but with a lower
frequency (reviewed in Cavenee et al., 1989). This list by no means
encompasses the full range of tissues in which the R81 gene is expressed.
It is not yet understood why the RBI gene’s oncogenic potential is limited
to certain cell types, whereas its expression is ubiquitous, and this
question is the subject of intense studyu More precise elucidation of the

function of p105-Rb may help answer this question.

3. Identification through the study of falilial cancers

In the 1980’s, using Knudson’s hypothesis as a paradigm land
encouraged by the rapid progress made in identifying and cloning the
retinoblastoma susceptibility gene, cancer researchers began searching for
tumor suppressor genes involved in the genesis of other familial cancers.
For common hereditary cancers, linkage analysis and occasional large
constitutional deletions enabled assignment of the susceptibility genes to
specific chromosomal locations. In rarer cancers with more subtle genetic
alterations, loss of heterozygosity at specific genetic loci in tumor
tissues, which was detected by restriction fragment length polymorphism,
pinpointed the locations of candidate suppressor genes.

Careful systematic application of these techniques in the genetic

20
analysis of numerous hereditary cancers allowed rapid identification of
several tumor suppressor genes. By 1989, chromosomal sites of putative
suppressor genes were identified in Wilms’ tumors (11p), familial
adenomatous polyposis coli (5q, 17p, 18q), multiple endocrine neoplasia
type 2 (1p), renal cell carcinoma and von Hippel-Lindau disease (3p), von
Recklinghausen neurofibromatosis (17q), and acoustic neurofibromatosis
(22q) (reviewed in Friend et al., 1988; Marx, 1989; Sager, 1989; Weinburg,
1989). Tumor suppressor genes have been identified at most of these sites
and at a few additional ones. To date, eight tumor suppressor genes,
including R81, have been cloned through study of familial cancers, and
germ-line mutations have been identified for all but one of these genes
(reviewed in Knudson, 1993). Two extensively studied familial cancers and

their associated susceptibility genes are discussed below.

3.1 Wilms’ tumor

Wilms’ tumor, a pediatric renal tumor, is similar to retinoblastoma
in that it occurs both sporadically and as part of a congenital condition.
Congenitally affected individuals usually develop multiple tumors in both
kidneys during early childhood. In contrast, individuals with the
sporadic form of the disease develop a single renal tumor and these
sporadic tumors arise two or three years later in childhood than those in
congenital cases. This suggests that Wilms’ tumor, like retinoblastoma,
occurs as the result of homozygous inactivation of a tumor suppressor gene
by two somatic mutations in sporadic cases, and by a germ-line mutation
followed by a somatic mutation in congenital cases.

The genetic analysis of Wilms’ tumors has revealed that this

prediction is generally accurate, but that the genetic basis of Wilms’

21

tumor is much more complex than that of retinoblastoma. Only 10% of
Wilms’ tumors are bilateral compared with 40% of retinoblastomas.
Furthermore, only 1% of Wilms’ tumors occur through familial transmission,
in contrast to the relatively frequent familial inheritance of
susceptibility to retinoblastoma (Matsunaga, 1981). This indicates that
new germline mutations are primarily responsible for the development of
bilateral Wilms’ tumor.

Another important difference between Wilms’ tumor and retinoblastoma
is that bilateral Wilms’ tumors are sometimes associated with congenital
abnormalities. These congenital abnormalities define two distinct
clinical syndromes. In one, which was given the acronym WAGR syndrome,
Wilms’ tumors occur along with aniridia (i.e. absence or malformation of
the iris), genitourinary abnormalities, and mental retardation (Miller et
al., 1964). The second, named the Beckwith-Wiedemann syndrome after the
authors who first described it in the literature, is characterized by
microglossia, gigantism, earlobe pits or creases, abdominal wall defects,
and an increased risk for the development of Wilms’ tumor,
rhabdomyosarcoma, hepatoblastoma, and abnormal carcinomas (Beckwith, 1969;
Sotelo-Avila and Coach, 1976).

Both syndromes are associated with bilateral (i.e. congenital)
Wilms’ tumor, but only the Beckwith-Wiedemann syndrome has been reported
to be passed by autosomal dominant inheritance in families (Best and
Hoekstra, 1981; Niikawa et al., 1986). Some familially transmitted Wilms'
tumors do not show the clinical symptomatology of either the WAGR or the
Beckwith-Wiedemann syndrome. Although it was not initially apparent, the
reason for the variety of presentations of Wilms’ tumors lies in the

multiplicity of genetic loci contributing to formation of these neoplasms.

22

The first clue to the location of a gene involved in Wilms’ tumor
came from karyotypic analysis of cells from WAGR syndrome patients. These
patients show a high frequency' of gross constitutional chromosomal
deletions on the short arm of chromosome 11, at band 13 (Riccardi et al.,
1978; Franke et al., 1979), suggesting that the tumor suppressor gene
responsible for Wilms’ tumor was located in this band. Further evidence
for this site as the location of a Wilms’ tumor suppressor gene came from
studies demonstrating loss of heterozygosity at polymorphic markers on
chromosome 11p13 in sporadic Wilms’ tumors (Fearon et al., 1984; Koufos et
al., 1984; Orkin et al., 1984; Reeve et al., 1984).

Genetic analysis of patients with Beckwith-Wiedemann syndrome
pointed to a different locus for the Wilms’ tumor susceptibility gene.
These patients show constitutional chromosomal abnormalities, generally
duplications, of 11p15 (Waziri et al., 1983; Koufos et al., 1989).
Furthermore, 15-20% of sporadic Wilms' tumors show loss of heterozygosity
at this locus rather than at 11p13 (Reeve et al., 1989; Koufos et al.,
1989). It is possible that genes at these two loci may interact, because
some tumors arising in WAGR syndrome patients, who are constitutionally
affected at the 11p13 locus, have been shown to have loss of
heterozygosity within 11p15, but not 11p13 (Henry et al., 1989).

The genetic analysis of Wilms’ tumor became even more complex when
it was found that familial predisposition to Wilms’ tumor that is not
associated with WAGR or Beckwith-Wiedemann syndromes is linked to neither
11p13 nor 11p15 (Grundy, 1988). This indicates that a third gene, and
conceivably more, can contribute to the genesis of this neoplasm. No
location for this third Wilms’ tumor gene has yet been suggested.

Though multiple genes evidently contribute to the development of

23

Wilms’ tumor, only one has been cloned. At 11p13, within the 400
kilobases smallest overlapping region of a number of WAGR deletions (Rose
et al., 1990), a transcription unit was identified, cloned, and
characterized (Call et al., 1990; Gessler et al., 1990). The gene, which
was named WTI, encodes a 46-49 kilodalton protein. This protein contains
four zinc-finger domains that bind to the same recognition sequence as do
the early growth response (EGR) zinc-finger proteins (Rauscher et al.,
1990). Recently, the WTI protein has been shown to inhibit transcription
of several positive regulators of cell growth, including EGR-I, insulin-
like growth factor II, and the platelet-derived growth factor A-chain gene
(Madden et al., 1991; Drummond et al., 1992; Wang et al., 1992).
Conversely, the WT1 protein activates transcription of other genes,
including the tumor suppressor p53 (Wang et al., 1993; Maheswaran et al.,
1993). Both of these actions are consistent with the hypothesis that the
WTI gene functions as a tumor suppressor gene.

The expression pattern of the WTI gene in normal tissue and in
Wilms’ tumors is compatible with its reputed involvement in the WAGR
syndrome. In genetically normal individuals, expression of WT1 is limited
to the embryonic kidneys, the fetal gonads, and some hemopoietic cells
(Pritchard-Jones et al., 1990). This restricted range of expression
offers an excellent explanation for the participation of this gene in
renal neoplasms and genitourinary anomalies, and for the lack of effect of
this gene on other tissues. As predicted, Wilms’ tumors that have
homozygous deletions of 11p13 show no expression of WT1 (Gessler et al.,
1990). Most other Wilms’ tumors, however, show a high expression of an
inactive form of the gene (Cowell et al., 1991).

In summary, the epidemiology of Wilms’ tumor suggests that

24
homozygous inactivation of a tumor suppressor gene is the genetic basis
for this neoplasm. Although genetic analysis has suggested that three or
more chromosomal loci are involved in Wilms’ tumor formation, only one
candidate gene, WTI, which is associated with the WAGR syndrome, has been
cloned. This gene has a cellular function appropriate for a tumor
suppressor, and an expression pattern befitting a gene participating in

genitourinary abnormalities.

3.2 Familial adenomatous polyposis coli

One of the best known hereditary predispositions to cancer is
familial adenomatous polyposis (FAP) in which a susceptibility to
colorectal cancer is inherited as an autosomal dominant trait.
Individuals with this disease develop hundreds, sometimes thousands, of
benign adenomatous polyps on their colonic mucosa. The polyps begin to
appear as early as the first decade of life, and a small proportion of
them inevitably progress to carcinomas. Adenomas from FAP patients arise
independently from single stem cells, i.e., are monoclonal, in contrast to
normal colonic epithelium which arises from numerous stem cells and
therefore is polyclonal. Similar monoclonal precancerous polyps also
occur sporadically in individuals who do not have the disease.

Following Knudson’s idea on the origin of retinoblastoma,
researchers hypothesized that a germline defect in a tumor suppressor
gene, followed by a somatic mutation in the second allele, is responsible
for the formation of polyps in FAP patients, and that somatic mutations in
both alleles results in the formation of sporadic polyps. The study of
constitutional deletions in FAP patients, chromosomal linkage in FAP

family pedigrees, and loss of heterozygosity in colorectal tumors provided

25
the clues required for the localization of the FAP susceptibility gene,
but the pattern of inactivation of the gene in colorectal adenomas and
carcinomas suggested that the prevailing theories regarding recessive
oncogenes needed to be expanded.

The first definite clue to the location of the FAP susceptibility
gene was the discovery of a patient with a constitutional deletion of
chromosomal band 5q21 (Huerra et al., 1986). Analysis of FAP pedigrees
showed that markers on this band were tightly linked to the development of
adenomas (Bodmer et al., 1987; Leppert et al., 1987). Detection of loss
of heterozygosity of 5q21 markers in adenoma tissue also implicated this
locus in the susceptibility to this disease, but the results were puzzling
in that loss of heterozygosity was detected in sporadic adenomas but not
in adenomas from FAP patients (Solomon et al., 1987; Vogelstein et al.,
1988).

Yeast artificial chromosome vectors and chromosome walking
techniques enabled isolation of several candidate cDNAs within the
critical band, three of which fell within two newly discovered
constitutional FAP deletions (Joslyn et al., 1991; Kinzler et al., 1991).
One of these three genes was found to be mutated in the germline DNAs of
FAP patients and in DNAs from sporadic colon carcinomas, establishing it
as the FAP susceptibility gene (Nishisko et al., 1991). This gene was
named APC for adenomatous polyposis coli.

The APC gene encodes a cytoplasmic protein with features suggesting
a potential for interaction with other proteins. As expected, the
mutations in this gene that are found in FAP germlines and in somatic
colon carcinomas appear to inactivate the gene. The majority are

frameshift or point mutations that result in premature stop codons and

26

lead to the synthesis of truncated proteins. Others involve total
deletion of the gene. Recent evidence suggests that variations in the
positions of the truncations may account, in part, for the phenotypic
differences observed among FAP patients. For example, FAP patients with
ocular fundus lesions were found to have truncating mutations after exon
9 (Olschwang et al., 1993) and patients with an attenuated form of FAP
were found to have truncating mutations located in a discrete region of
the 5’ end of the APC gene (Spirio et al., 1993).

It is possible that some mutant APC genes are not true null alleles
and that there are important differences in the activities of the mutant
proteins produced. Su et al. (1993) reported that most truncated APC
peptides can associate with the wild-type APC in vivo, perhaps
inactivating it in a dominant negative manner. This could explain why
polyps from APC patients generally do not show evidence of a further
change on chromosome 5q21 (Solomon et al., 1987; Vogelstein et al., 1988).
Alternatively, the inactivation of only onelof the two alleles may lead to
polyp formation by reducing the production of the suppressor protein below
a critical threshold concentration.

Vogelstein (1988) hypothesized that the FAP susceptibility gene
normally acts as a negative regulator of colonic epithelium proliferation,
and that loss of a single FAP allele leads to ineffective control
resulting in hypertrophy of the colonic epithelium. Loss of the remaining
wild-type APC allele is not the event that promotes transition from the
hyperplastic epithelium to the adenomatous state. Loss of the second APC
allele does appear, however, to be instrumental in the progression of
adenomas to carcinomas. Support for this hypothesis comes from the

observation of frequent loss of heterozygosity at the APC locus in

27
colorectal carcinomas from FAP patients (Sasaki et al., 1989) and from the
finding that the cells of 81% of colorectal carcinomas are totally devoid
of normal, full-length APC protein (Smith et al., 1993).

These findings highlight the need for a clearer understanding of the
function of the APC geneland of the phenotypic effects of heterozygous and
homozygous alterations at this locus. Furthermore, they suggest that
theories regarding the oncogenic effect of inactivated tumor suppressor
genes need to be expanded to include the possible dominant negative and

dosage effects of these genes.

4. Identification by loss of heterozygosity in tulor cells
4.1 Colon carcinomas

As discussed in section A1.2, the progression of the colon cells
from hyperplastic epithelium, through several stages of adenomas, and
finally to invasive carcinomas, suggests that colon carcinogenesis
requires multiple genetic changes. Inactivation of one FAP allele
apparently results in hyperplasia of the colonic epithelium.
Proliferation of these cells then provides the opportunity for further
mutations to occur. Because tumors of all stages, from very small
adenomas to large metastatic carcinomas, can be obtained for study,
colorectal carcinogenesis is an excellent system in which to study the
number and kinds of somatic mutations required for malignant
transformation in human cells.

Although early studies of allelic losses in colon cancer patients
focused on chromosome 5q because it was known to carry the locus
segregating with this disease, it was also known from cytogenetic and

molecular studies that portions of chromosomes 17 and 18 are frequently

28

absent in colorectal carcinomas (Reichmann et al., 1981; Muleris et al.,
1985; Fearon et al., 1987). Vogelstein and his colleagues (1988) used
restriction fragment length polymorphisms to detect the frequency of loss
of heterozygosity on chromosomes 5, 17, and 18 in 172 colorectal tumor
specimens representing various stages of neoplastic development. They
found that specific regions of these three chromosomes were lost,
respectively, in 35, 73, and 75 percent of colon carcinomas.

They determined the prevalence of these chromosomal losses in
adenomas at various stages of neoplastic development, and found that
allelic deletions of 5q occur at an early stage, allelic deletions of 17p
occur at a late stage, and those of 18q occur at an intermediate stage.
The allelic losses at 17p and 18q were interpreted as evidence that these

regions encoded additional tumor suppressor genes.

a. The DCC gene

Vogelstein et al. (1988) determined that the colorectal tumor
suppressor genelon chromosome 18 resided between 18q21.3 and the telomere.
Within this region, they identified a large transcriptional unit which
they named DCC for Deleted in Colon Cancer (Fearon et al., 1990). The DCC
gene is expressed in most normal tissues, including colonic mucosa, but
its expression is greatly reduced or absent in most colorectal carcinomas.

The predicted amino acid sequence encoded by the DCC gene specifies
a protein with considerable homology to neural cell adhesion molecules,
suggesting that DCC may play a role in cell-cell interactions. The types
of mutations, mostly insertions and deletions, that have been found within
in the DCC genes of tumor cells usually produce termination codons (Fearon

et al, 1990). The proteins produced by these mutated genes are truncated

29
and, presumably, non-functional.

Confirmation of the DCC gene’s function as a tumor suppressor
requires analysis of the phenotypic effect of its introduction into tumor
cell lines lacking a functional DCC gene. Microcell transfer of a normal
chromosome 8 into colon carcinoma cells suppressed the tumor forming
ability of these cells (Tanaka et al., 1991; Goyette et al., 1992).
Introduction of the cloned DDC gene into such cells, an experiment that

would be more definitive, has not yet been done.

b. The p53 gene

After determining that loss of heterozygosity on the short arm of
chromosome 17 commonly occurs at a late stage in colon carcinogenesis,
Vogelstein and his colleagues used additional 17p markers to further
define the region of loss. The affected region contains a gene, p53, that
had already been implicated in cellular transformation (Baker et al.,
1989). The p53 protein was isolated in 1979 as a protein bound to SV40
large T antigen (Lane and Crawford, 1979; Linzer and Levine, 1979).
Later, mutant forms of the cloned p53 gene were shown to c00perate with
ras in transformation cells in culture (Hinds et al., 1989).

When the p53 coding regions of colorectal tumors with 17p allelic
losses were analyzed, the remaining p53 alleles were found to contain
mutations in highly conserved regions of the gene (Baker et al., 1989;
Nigro et al., 1989). This suggested that the wild-type allele acts as a
colorectal tumor suppressor gene. Transfection of the wild-type gene into
colorectal carcinoma cells suppressed cell growth, offering strong support
for this hypothesis (Baker et al., 1990). Furthermore, the wild-type p53

was found to inhibit raS' oncogene-mediated transformation to focus

30
formation of rat embryo fibroblasts (Finlay and Hinds, 1989; Eliyahu et
al., 1989).

The finding that p53 functions as a suppressor gene in colorectal
tumors prompted a search for p53 mutations in a wide assortment of other
human tumors, many of which were already known to have frequent losses of
chromosome 17p alleles. Over the past few years, p53 has been found to be
mutated, usually homozygously, in lung, breast, esophageal, liver,
bladder, ovarian, and brain tumors, and in lymphomas and leukemias
(reviewed in Hollstein et al., 1991). This list is still growing and the
accumulating data show p53 to be the most frequently involved gene in
human oncogenesis. The December 24, 1993, issue of SCience dubbed p53 the
"gene of the year" owing to the great frequency with which it appears in
the scientific literature and the enormous impact it has had on our
understanding of the role of tumor suppressor genes in carcinogenesis.

The p53 story presents a reverse chain of events relative to those
of the retinoblastoma paradigm. For retinoblastoma, the search for a
germline mutation that was associated with the familial cancer revealed
chromosome deletions that pinpointed the location of the susceptibility
gene. This, in turn, allowed cloning of the gene and characterization of
the protein product. For p53, discovery of the protein in a complex with
the SV40 large T antigen allowed cloning of the gene long before its role
as a tumor suppressor was revealed. After p53 was recognized as a tumor
suppressor gene, the gene was examined as a candidate for the
susceptibility gene in a rare familial cancer known as Li-Framneni
syndrome.

Li-Fran'meni syndrome (LFS), like retinoblastoma and Wilms’ tumor, is

inherited in an autosomal dominant fashion. Affected individuals are

31

predisposed to the development of diverse mesenchymal and epithelial
neoplasms at multiple sites, including breast carcinomas, soft tissue
sarcomas, brain tumors, osteosarcomas, leukemia and adrenocortical
carcinomas. The rarity and high mortality of LFS precluded genetic
analysis by the techniques used to find the retinoblastoma and Wilms’
tumor susceptibility genes. As an alternative approach, p53 was evaluated
as a candidate for the LFS susceptibility gene because the spectrum of
cancers in which it was known to be mutated included those malignancies
found in LFS patients. Malkin et al. (1990) and Srivastava et al. (1990)
detected germline p53 mutations in all six of the LFS families analyzed,
indicating that inherited p53 mutations are responsible for transmission
of this disease.

The p53 gene is expressed at low levels in all cells, and the
protein product has a brief half-life. The protein functions as a
transcriptional regulator (Fields and Jang, 1990) that activates
expression of an inhibitor of the cyclin dependent kinases (El-Oeiry et
al., 1993; Harper et al., 1993). Inhibition of cyclin dependent kinases
prevents passage through the cell cycle. By disruption of this cascade of
events, inactivation of p53 leads to uninhibited cell replication.

Most mutations in the p53 gene extend the half-life of the protein
and cause it to accumulate in the cell (Levine and Momand, 1990). Some
mutant proteins appear to exert a dominant negative effect (Herskowitz,
1987; Eliyahu et al., 1988) possibly by binding to the wild-type protein
in tetramers (Stenger et al., 1992). Evidence that some p53 mutant
proteins may have inherent transforming properties unrelated to their
ability to bind the normal p53 protein comes from the report that

introduction of a particular mutant p53 gene into a cell line lacking both

32
p53 alleles enhanced the cell’s tumorigenicity (Wolf et al., 1984). The
mechanisms by which these different mutations exert their various

phenotypic effects are not yet understood.

4.2 Other malignancies

Loss of heterozygosity has been used to study the role of tumor
suppressor gene inactivation in numerous malignancies other than
colorectal cancer. Among the malignancies studied are hepatic, breast,
and bladder canceru A large percentage'of human hepatocellular carcinomas
exhibit loss of the function of the p53 and R81 genes. Frequent loss of
heterozygosity on chromosomes 4 and 16 in these tumors indicates that
additional tumor suppressor genes may be inactivated (Buetow et al., 1989;
Fujimoto et al., 1994). Analyses of primary human breast tumors show loss
of heterozygosity on chromosome 11 and in multiple regions of chromosomes
1 and 17 (Ali et al., 1987; Bieche et al., 1993; Cropp et al., 1993).
Allelic losses in human bladder cancers give evidence for suppressor loci
on chromosomes 3, 11, and 17, and at two distinct loci on chromosome 9
(Tsai et al., 1990; Klingelhutz et al., 1992; Ruppert et al., 1993). The
cumulative data from these and numerous other studies indicate that the
inactivation of tumor suppressor genes may be a more common mechanism in
human carcinogenesis than is the activation of dominantly acting

oncogenes.

C. Contributions froa Cell Hybrid Studies

Although the methodology available for finding and characterizing

33
tumor suppressor genes is continually expanding, the technique of cell
hybridization, which initially brought these'genes to our attention, still
contributes significantly to the accomplishment of this task. This
technique is particularly well suited for obtaining certain kinds of
information. Examples of such applications, in particular those related

to carcinogenesis, are discussed below.

1. Chronosoaal assignaent of tulor suppressor'genes

Although the detection of loss of heterozygosity in tumor cells is
currently the most widely used method of locating tumor suppressor genes,
this is an arduous task that is eased greatly if a tumor suppressor gene
can first be assigned to a specific chromosome. Sometimes a particular
tumor type is associated with common gross deletions that point to the
location of an inactivated tumor suppressor gene. In the absence of
localizing deletions, cell hybridization experiments may enable
chromosomal assignment of the tumor suppressor gene.

As was briefly discussed in section 8.1., the hybrid cell resulting
from fusion of a tumor cell with a normal cell nearly always is non-
tumorigenic. In rodent cell hybrids, rapid chromosomal loss from the
tetraploid fusion product allows a high rate of reversion to the»malignant
phenotype, presumably by loss of the chromosome(s) bearing the suppressor
allele(s) (reviewed in Harris, 1988; Klein, 1988; Sager, 1989). When a
human cell is fused with another human cell (i.e., human X human cell
hybrids) and also in the case'of some rodent X human cell hybrids, greater
chromosomal stability allows determination of the chromosomes associated
with the suppressed phenotype. Preferential retention of certain

chromosomes in suppressed hybrids and non-random chromosomal loss in

34

malignant revertants are often observed.

1.1 W hybrids

Although very few tumorigenic mouse cell X normal human cell hybrids
are suppressed for tumorigenicity (Klinger et al., 1978; Kucherlapati and
Shin, 1979), a high proportion (30-50%) of tumorigenic hamster cell X
normal human cell hybrids are suppressed (Klinger et al., 1978; Stoler and
Bouck,1985; Wynford-Thomas et al., 1989). The hybrids are chromosomally
stable, allowing comparison of the>chromosome complement of the suppressed
hybrids with those of the non-suppressed hybrid populations. Furthermore,
the parental origin of the chromosomes can be determined by cytogenetic

and biochemical gene marker methods.

a. can call X flQF hybrids

Klinger and his colleagues fused tumorigenic Chinese hamster ovary
(CHO) cells with human diploid fibroblasts (HDF). They found that
retention of human chromosomes 2, 9, 10, 11, and 17 was associated with
suppression of tumorigenicity in the hybrids (Klinger et al., 1978;Klinger
and Shows, 1983). Chromosome 2 was never found in tumorigenic hybrid
cells, and the other four chromosomes listed were found only at very low
frequencies. The multiplicity of the chromosomes implicated in
suppression of this tumor type may reflect the multistep nature of
carcinogenesis. If the additive effect of several tumor suppressor gene
losses is required for the expression of tumorigenicity, then one would
expect that the retention of any one of these genes would suppress tumor
growth.

Additional human chromosomes are associated with tumor suppression

35

in the CHO cell X HDF hybrids if pairs of chromosomes are considered
(Klinger and Shows, 1983). For example, chromosomes 7 and 13 were found
together in only 1% of tumors. Considered singly, they were found,
respectively, in 17% and 24% of tumors. Evidently the combination of
these chromosomes is an effective tumor suppressor in CHO cells, whereas
one of them alone is not. In this manner, cell fusion studies can help to
identify tumor suppressor genes that require the cooperation of a second
tumor suppressor gene.

Most of the suppressor chromosomes identified by Klinger and Shows
(1983) are now known, or have been implicated by other studies, to contain
tumor suppressor genes. As discussed above, chromosome 17 contains p53,
chromosome 11 contains WTI, chromosome 13 contains RBI. Chromosome 10 has
been implicated through loss of heterozygosity as the location of a
suppressor gene involved in malignant melanomas (Rempel et al., 1993).
The chromosomal region 9p13-p22 is a frequent site of allelic loss in
leukemias, melanomas, malignant mesotheliomas, brain tumors, and lung and
bladder carcinomas (Diaz et al., 1990; James et al., 1991; Fountain et
al., 1992; Cairns et al., 1993; Center et al., 1993; Cheng et al., 1993;
Olopade et al., 1993). The strong suppressive effect of human chromosome
2 in the CHO cell hybrids suggests that it, too, is the location of a
tumor suppressor gene, although no role for this gene has yet been
suggested by the study of human tumors.

b. Bfl5_gall_x_flflfi hybrids

Cell fusion studies indicate that the genes responsible for
suppression may vary with different cell types. Stoler and Bouck (1985)
fused anchorage independent baby hamster kidney (BHK) fibroblasts with

normal human diploid fibroblasts. Human chromosome 1 was retained in all

36
hybrids that were suppressed for anchorage independence and was lost when
these hybrids reverted to the transformed phenotype. This suggests that
a gene or set of genes on chromosome 1 mediates suppression of anchorage
independence in BHK cells.

Chromosome 1 was not among the chromosomes associated singly or in
pairs with suppression of tumorigenicity in CHO cells (Klinger and Shows,
1983). Perhaps this discrepancy reflects the different embryological
origins of the two cell types. BHK cells are mesenchymal, whereas CHO
cells are epithelial. A second possibility is that the phenotypes of
anchorage independence and tumorigenicity, although closely associated,

are under separate genetic controls.

1.2 Human cell x human cell hybrids

The effectiveness of human chromosomes as suppressors of
transformation in hamster cells reflects the highly conserved nature of
many tumor suppressor genes and their fundamental role in basic cellular
processes. Some tumor suppressor genes, p53 for example, are ubiquitously
expressed in normal tissues and are inactivated in a wide variety of
tumors from many animals. Others, such as WTI, have a narrow range of
expression and‘limited oncogenic potential. Human genes of the latter
type may not be easily identified through interspecies cell hybridization
studies. For the identification of the chromosomes carrying these genes,

intraspecies human cell hybridization is likely to be more useful.

a. W hybrids

Successful identification of chromosomes associated with tumor

suppression in intraspecies human cell hybrids has been limited to one

37
hybrid cell system, namely human cervical carcinoma-derived HeLa cells
fused with human diploid fibroblasts. All hybrids from this cross failed
to form tumors when inoculated into nude mice (Stanbridge, 1976; Klinger
et al., 1978; and Klinger, 1980). Karyotypic analysis revealed that the
hybrids cells maintained nearly complete complements of chromosomes from
both parent cell lines during extensive massaging in culture.

Rare tumorigenic segregants arose within hybrid cell populations
only after prolonged periods of culture or through specific selection for
variants (Klinger, 1980; Stanbridge et al., 1981). These segregants
exhibited loss, primarily from the genetic material contributed by the
normal diploid parent, of 55% of their original chromosome complement.
Restriction fragment length polymorphisnianalysis revealed that every'non-
tumorigenic hybrid contained four copies of chromosome 11, two from each
parent cell. All but one of 57 tumorigenic segregants, however, lacked
one or both copies of chromosome 11 from the normal parent (Kaelbling and
Klinger, 1986; Klinger and Kaelbling, 1986; and Srivatsan et al., 1986).
This suggests that a suppressor gene or set of genes on the normal
chromosome 11 is involved in the control of tumor formation in these
hybrids.

Support for this hypothesis came from the detection of loss of
heterozygosity on the short arm of chromosome 11 in Hela cells, an
indication that HeLa cells may be deficient in a suppressor function at
this locus (Kaelbling et al., 1986). In the non-tumorigenic hybrids, the
normal chromosome 11 may complement this deficiency. Microcell transfer
of a normal chromosome 11 into HeLa cells, or into tumorigenic segregants
of HeLa cell X HDF hybrids, suppressed the tumorigenicity of these cells,

confirming the suppressive function of chromosome 11 in this cell type

38
(Saxon et al., 1986).

b. flllQflfl gall X flQF hybrids

Stanbridge and his colleagues attempted to analyze suppression of
malignancy in other human tumor cell X HDF hybrids (Stanbridge et al.,
1982). When they fused human fibrosarcoma-derived HT1080 cells with
normal human fibroblasts, most of the hybrids retained the tumorigenic
phenotype, a finding that agreed with that of Croce et al. (1979).
Karyotypic analysis revealed that all the tumorigenic hybrids were
hexaploid, containing a tetraploid complement of HT1080 chromosomes and a
diploid complement of HDF chromosomes.

Only two non-tumorigenic hybrids were obtained from these fusions.
Both were tetraploid with one complement of chromosomes from each parent
cell (Benedict et al., 1984). From one of these non-tumorigenic hybrids,
rare tumorigenic segregants arose. Reversion to tumorigenicity was
associated with loss of chromosomes 1 and 4, suggesting a role for these
chromosomes in the suppression of the tumor phenotype of HT1080 cells.
The difficulty in obtaining larger numbers of tetraploid, non-tumorigenic
HT1080 cell X HDF hybrids precluded further analysis of tumor suppression
in this system.

c. Other mallggagt gall X gggmal gall hybrids

From most fusions of malignant human cells with normal human
fibroblasts, Stanbridge et al. (1982) were unable to obtain long-term
hybrid cell populations. Although hybrid clones arose, the populations
senesced after a limited number of population doublings. Similar

observations were reported by Bunn and Tarrant (1980), Muggleton-Harris

39
and Desimone (1980), and Pereira-Smith and Smith (1981). On the basis of
these observations, it was postulated that the infinite life span of
transformed cells in is acquired through loss of a growth suppressor
function, and that the normal cell supplies this function in malignant
cell X normal cell hybrids, causing them to senesce. This hypothesis
altered the direction of suppressor gene research. Investigators began to
think in terms of suppression of specific transformed phenotypes, rather

than of malignancy per se.

2. Suppression of specific transfer-ed phenotypes

In CHO cell X HDF and HeLa cell X HDF hybrids, many specific
transformed phenotypes segregate independently of tumorigenicity
(Stanbridge and Wilkinson, 1978; Klinger, 1980; Stanbridge, et al., 1982;
Klinger and Shows, 1983). Although the hybrid cells are unable to form
tumors in nude mice, they retain several other transformed
characteristics, including infinite life span, growth in soft agar, growth
factor independence, and altered cellular morphology. This suggests that
the genetic mechanisms controlling tumorigenicity' differ from ‘those

controlling these other phenotypes.

2.1 Anchorage independence

Although anchorage independence'was not suppressed when CHO or HeLa
cells were fused with normal human cells, this phenotype was suppressed
when anchorage independent BHK cells were fused with normal BHK cells
(Bouck and di Mayorca, 1982) or with normal human fibroblasts (Stoler and
Bouck, 1985). As already mentioned, the human chromosome that was

associated with suppression of anchorage independence was not among those

40
reported by Klinger and Shows (1983) to suppress CHO cell tumorigenicity,
suggesting that these two phenotypes may be under separate genetic

control.

2.2 Growth factor independence

Strauss and Mohandas (1987) found that hybrids between mouse
melanoma cells and mouse embryo fibroblasts, or between mouse L cells and
normal human fibroblasts, are suppressed for growth factor independence.
In the mouse L cell X' HDF hybrids, suppression of growth factor
independence correlates with retention of human chromosomes 5, 22, and X.
This group of chromosomes differs from those reported to suppress either

agar growth or tumorigenicity.

2.3 Infinite life span

As described above (section C.1.2c, pp. 38-39), the phenotype of
infinite life span can also be suppressed in cell hybrids. The fusion of
infinite life span cells with finite life span cells yields finite life
span hybrids. Immortal hybrids are frequently obtained, however, when
rodent cells are used in such fusions, because the chromosomal instability
of rodent cell hybrids allows frequent reversion to infinite life span by
loss of the chromosomes that carry the suppressor genes.

Although Stanbridge and his colleagues and Klinger and his
colleagues reported that HeLa cell X HDF hybrids have infinite life spans
(Stanbridge, 1976; Klinger et al., 1978; Stanbridge and Wilkinson, 1978;
Klinger, 1980), Pereira-Smith et al. (1990) concluded otherwise after
closely charting the doubling rates of numerous hybrid populations. Out of

the 39 clonal HeLa cell X HDF hybrid populations that they examined, 28

41
senesced after about 25 doublings. In another six of the hybrid
populations, cell division slowed after about 25 population doublings.
remained slow for several weeks, then returned to theioriginal rapid rate.
Pereira-Smith and Smith interpreted this pattern as an indication that the
hybrid cells had an infinite life span and that infinite life span
revertants arose that outgrew the senescing hybrid populations. The five
remaining hybrid populations showed no decrease in their doubling rates
throughout 100 population doublingsc Pereira-Smith and Smith hypothesized
that infinite life span revertants arose in these populations soon after

cell fusion, before the cells of the hybrid population began to senesce.

2.4 Other transformed phenotypes

Hybrids from most fusions of infinite life span human cells with
finite life span human cells are chromosomally stable. Reversion to the
infinite life span phenotype in these hybrids is a rare occurrence
(Pereira-Smith and Smith, 1981, 1983). This explains why Stanbridge et
al. (1982) were unable to obtain long-term cultures of most human tumor
cell X HDF hybrids. To circumvent this problem, suppression of
tumorigenicity and other transformed phenotypes can be tested in hybrids
from fusions of tumor cells with infinite life span non-tumorigenic cells.
For example, Zajchowski et al. (1990) fused MCF-7 human breast cancer
cells with immortalized non-tumorigenic human mammary epithelial cells.
They successfully obtained long-termlcultures of hybrid cells and reported
that tumorigenicity, growth factor independence, tumor necrosis factor
sensitivity, and pS2 breast cancer marker expression were suppressed in
these hybrids.

Moroco et al. (1990) performed fusions among hamster buccal pouch

42
keratinocytes at various stages of transformation. These researchers
documented that the phenotypes of angiogenic activity, infinite life span,
anchorage independence, and tumorigenicity, which arose sequentially in
these cells and were therefore hypothesized to be under independent
genetic controls, are all linked to functional loss of suppressor genes.
Other examples of specific transformed characteristics that have been
shown to be suppressed in cell hybrids are the interleukin 3 expression of
mouse mastocytoma cells (Diamantis et al., 1989) and the high frequency'of

gene amplification of HT1080 cells (Tlsty et al., 1992).

3. Coapleaentation analysis
3.1 Complementation of tumorigenicity

Considering the evidence described above that various tumor types
may arise by the inactivation of separate tumor suppressor genes, it is
reasonable to predict that certain pairs of tumor cells may be able to
complement the genetic defects in one another to yield non-tumorigenic
hybrid cells. Wiener and Harris (1974) tested this hypothesis using
intraspecies mouse cell hybrids. Only one>of twelve crosses among various
tumor cells generated hybrid cells with reduced tumorigenicity. These
researchers concluded that the genetic lesions determining the malignant
phenotype, although recessive, are unable to complement each other.

Stanbridge et al. (1982) hypothesized that the apparent inability of
mouse tumor cells to complement each other resulted from the same
phenomenon that obscured tumor suppression when mouse tumor cells were
fused with normal mouse cells. Rapid loss of the suppressor chromosomes
may have allowed the hybrids to revert to the tumorigenic phenotype. To

eliminate this confounding variable, Stanbridge and his colleagues tested

43
the complementation hypothesis in human intraspecies hybrids. They
reported that tumorigenicity was suppressed when carcinoma cells were
fused with melanoma or sarcoma cells, but not when they were fused with
lymphoblastoid cells or other carcinoma cells (Stanbridge at al., 1982;
Weissman and Stanbridge, 1983). They concluded that complementation among
tumor cells can occur, that at least two complementation groups exist, and
that possibly a distinct gene or set of genes controls the expression of

tumorigenicity for each somatic cell type.

3.2 Complementation of infinite life span

Because multiple phenotypes contribute to a cell’s tumorigenic
potential, it is of interest to know which specific phenotypes were
suppressed in the non-tumorigenic hybrids from Stanbridge’s malignant cell
X malignant cell crosses. The studies referenced above did not report
this information. Since infinite life span is a recessive phenotype, it
may be that these hybrids lacked the ability to form tumors because
complementation of the genetic defects associated with infinite life span
yielded hybrids that had finite life spans that were too short to allow
tumor formation.

Pereira-Smith and Smith (1983) investigated complementation of the
infinite life span phenotype'by fusing several pairs of infinite life span
cells and assaying the proliferative potentials of the hybrids. They
reported that infinite life span was suppressed in the hybrids from some
fusions, but not in others. In extensions of these studies, they reported
that 30 different immortal human cell lines could be assigned to four
immortality complementation groups (designated A, B, C, and D), with no
cell line belonging to more than one group (Pereira-Smith and Smith, 1988;

44

Ning and Pereira-Smith, 1991). These results suggest that four different
genes or sets of genes contribute to the program for senescence in human
cells. Support for the validity'of the immortality complementation groups
came from the finding that chromosome 4 suppressed the immortal phenotype
when introduced into cells from complementation group B, but had no effect
when introduced into cells from the other three groups (Ning et al.,
1991).

Whitaker et al. (1992) fused the SV40-immortalized human cell line
BET-1 to representative cells fromleach of the four complementation groups
established by Pereira-Smith and Smith. They obtained immortal hybrids
only from fusions with cells in group 0, indicating that BET-I belongs to
group 0. When they later applied this method to three other SV40-
immortalized cell lines, however, each of the cell lines were found to
belong to more than one complementation group (Duncan et al., 1993).
These results indicate that complementation among immortal cells is a more
complicated phenomenon than the studies of Pereira-Smith and Smith

suggest.

REFERENCES

Ali, I.U., Lindereau, R., Theillet, C., Callahan, R. (1987). Reduction to
homozygosity'of genes on chromosome 11 in human breast neoplasia. Science,
238, 185-187.

Bagchi, S., Weinmann, R., Raychaudhuri, P. (1991). The retinoblastoma
protein copurifies with E2F-1, an ElA-regulated inhibitor of the
transcription factor E2F. Cell 65, 1063-1072.

Baker, S.J., Fearon, E.R., Nigro, J.M., Hamilton, S.R., Preisinger, A.C.,
Jessup, J.M., vanTuinen, P., Ledbetter, D.H., Barker, D.F., Nakamura, Y.,
White, R., Vogelstein, B. (1989). Chromosome 17 deletions and p53 gene
mutations in colorectal carcinomas.Science 244, 217-221.

Baker, S.J., Markowitz, S., Fearon, E.R., Willson, J.K.V., Vogelstein, B.
1990). Suppression of human colorectal carcinoma cell growth by wild-type
p53. Science 249, 912-915.

Balaban, 6., Gilbert, F., Nichols, W., Meadows, A.T., and Shields, J.
(1982). Abnormalities of chromosome 13 in retinoblastomas from individuals
with normal constitutional karyotypes. Cancer Genet. Cytogenet. 6, 213-
221.

Barbacid, M. (1987). ras genes. Ann. Rev. Biochem. 56, 779-827.

Barski, G. and Cornefert, F. (1962). Characteristics of "hybrid"-type
clonal cell lines obtained from mixed cultures in vitro. JNCI 28, 801-821

Beckwith, J.B. (1969). Macroglossia, omphalocele, adrenal cytomegaly,
gigantism and hyperplastic visceromegaly. Birth Defects 5, 188-196.

Benedict, W.F., Murphree, A.L., Banerjee, A., Spina, C.A., Sparkes, M.C.,
and Sparkes, R.S. (1983). Patient with 13 chromosome deletion: evidence
that the retinoblastoma gene is a recessive cancer gene. Science 219, 973-
975.

Benedict, W.F., Weissman, B.E., Mark, C., Stanbridge, E.J. (1984).
Tumorigenicity of human HT 1080 fibrosarcoma X normal fibroblast hybrids:
Chromosome dosage dependency. Cancer Research 44, 3471-3479.

Best, L.G., and Hoekstra, R.E. (I981). Wiedemann-Beckwith syndrome:
autosomal-dominant inheritance in a family. Am. J. Med. Genet. 9, 291-299.

Bieche, 1., Champéme, M.-H., Matifas, F., Cropp. c.s., Callahan, R., and
Lindereau, R. (1993). Two distinct regions involved in 1p deletion in
human primary breast cancer. Cancer Research, 53, 1990-1994.

Bishop, J.M. and Varmus, H. (1982). Functions and origins of retroviral

transforming genes. In RNA Tumor Viruses, Weiss, R., Teich, N., Varmus,
H., Coffin, J. eds. pp. 999-1108. New York: Cold Spring Harbor Laboratory.

45

46

Bodmer, W.F., Bailey, C., Bodmer, J., Bussey, H.J.R., Ellis, A., Gorman,
P., Lucibello, F.C., Murday, V.A., Rider, S.H., Scambler, P., Sheer, 0.,
Solomon, E., and Spurr, N. (1987). Localization of the gene for familial
adenomatous polyposis or chromosome 5. Nature 328, 614-616.

Bouck, N., and di Mayorca, G. (1982). Chemical carcinogens transform BHK
cells by inducing a recessive mutation. Molecular and Cellular Biology 2,
97-105

Bookstein, R., Shew, J-Y., Chen, P-L., Scully, P., Lee, W-H. (1990).
Suppression of tumorigenicity of human prostate carcinoma cells by
replacing a mutated R8 gene. Science 247, 712-715.

Brugge, J.S. and Erikson, R.L., (1977). Identification of a
transformation-specific antigen induced by an avian sarcoma virus. Nature
269, 346-348.

Buchkovich, K., Duffy, L.A., Harlow, E. (1989). The retinoblastoma protein
is phosphorylated during specific phases of the cell cycle. Cell 58, 1097-
1105.

8uetow, K.H., Murray, J.C., Israel, J.L., London, W.T., Smith, M., Kew,

M., Blanquet, V., Brechot, C., Redeker, A., and Govindarajah, S., (1989).
Loss of heterozygosity suggests tumor suppressor gene responsible for

grggary hepatocellular carcinoma. Proc. Natl. Acad. Sci., USA., 86, 8852-
8 .

Bunn, C.L., and Tarrant, G.M. (1980). Limited lifespan in somatic cell
hybrids and cybrids. Experimental Cell Res. 127, 385-396.

Cairns, P., Shaw, M.E., and Knowles, M.A. (1993). Preliminary mapping of
the deleted region of chromosome 9 in bladder cancer. Cancer Res., 53,
1230-1232.

Call, K.M., Glaser, T., Ito, C.Y., Buckler, A.J., Pelletier, J., Haber,
D.A., Rose, E.A., Kral, A., Yeger, N., Lewis, W.H., Jones, C., and
Housman, D.E. (1990). Isolation and characterization of a zinc finger
pglypeptide gene at the human chromosome 11 Wilms tumor locus. Cell 60,
5 9-520.

Cavenee, W.K., Dryja, T.P., Phillips, R.A., Benedict, W.F., Godbout, R.,
Gallie, B.L., Murphree, A.L., Strong, L.S., & White, R.L. (1983).
Expression of recessive» alleles by' chromosomal mechanisms in
retinoblastoma. Nature. 305, 779-784.

Cavenee, W., Hastie, N., Stanbridge, E. (1989). Recessive Oncogenes and
Tumor Suppression (New York: Cold Spring Harbor Laboratory Press), pp. 1-

Center, R., Lukeis, R., Dietzsch, E., Gillespie, M., and Garson, O.M.
(1993). Molecular deletion of 9p sequences in non-small cell lung cancer
and malignant mesothelioma. Genes Chromosomes Cancer 7, 47-53.

47

Chang, E.H., Gonda, M.A., Ellis, R.W., Scolnick, E.M., and Lowy, D.R.
(1982). Human genome contains four genes homologous to transforming genes
of Harbey and Kirsten murine sarcoma viruses. Proc. Natl. Acad. Sci. USA
79, 4848-4852.

Chellappan, S.P., Hiebert, S., Mudryj. M., Horowitz, J.M., Nevins, J.R.
(1991). The E2F transcription factor is a cellular target for the R8
protein. Cell 65, 1053-1061.

Chen, P-L., Scully, P., Shew, J-Y., Want, J.Y.J., Lee, W-H. (1989).
Phosphorylation of the retinoblastoma gene product is modulated during the
cell cycle and cellular differentiation. Cell 58, 1193-1198.

Cheng, J.O., Jhanwar, S.C., Lu, Y.Y., Testa, J.R. (1993). Homozygous
deletions within 9p21-p22 identify a small critical region of chromosomal
loss in human malignant mesotheliomas. Cancer Research 53, 4761-4763.

Chittenden, T., Livingston, 0.M., Kaelin, Jr., W.G. (1991). The T/EIA-
binding domain of the retinoblastoma product can interact selectively with
a sequence-specific DNA-binding protein. Cell 65, 1073-1082.

Claude, A., Porter, K.R., Pickels, E.G. (1947). Electron microscopy study
of chicken tumor cells. Cancer Res. 7, 421-430.

Cowell, J.K., Wadey, R.8., Haber, D.A., Call, K.H., Housman, D.E., and
Pritchard, J. (1991). Structural rearrangements of the WTl gene in Wilms’
tumour cells. Oncogene 6, 595-599.

Crawford, L.V. and Crawford, E.M. (1961). The properties of Rous sarcoma
virus purified by density gradient centrifugation. Virology 13, 227-232.

Croce, C.M., Barrick, J., Linnenbach, A., Koprowski, H. (1979). Expression
of nalignancy in hybrids between normal and nalignant cells. J. Cell.
Physiol. 99, 279-286.

CrOPP. C.S., Champeme, M.-H., Lindereau, R., Callahan, R. (1993).
Identification of three regions on chromosome 17q in primary human breast
carcinomas which are frequently deleted. Cancer Research 53, 5617-5619.

Davis, M., Malcom, S., Hall, A., and Marshal, C.J. (1983). Location of the
human N-ras oncogene to chromosome Icen-p21 by in situ hybridization. EMBO
J. 2, 2281-2283.

DeFeo, 0., Gonda, M.A., Young, H.A., Chang, E.H., Lowy, D.R., Scolnick,
E.M., Ellis, R.W. (1981). Analysis of two divergent rat genomic clones
homologous to the transforming gene of Harvey murine sarcoma virus. Proc.
Natl. Acad. Sci. USA 78, 3328-3332.

Der, C.J., Krontiris, T.G., Cooper, G.M. (1982). Transforming genes of
human bladder and lung carcinoma cell lines are homologous to the ras
genes of Harvey and Kirsten sarcoma viruses. Proc. Natl. Acad. Sci. USA
79, 3637-3640.

48

DerKinderen, D.J., Koten, J.W., Nagelkerke, N.J.D., Tan, K.E.W.P., Beemer,
F.A., and Den Otter, W. (1988). Non-ocular cancer in patients with
hereditary retinoblastoma and their relatives. International Journal of
Cancer 41, 499-504.

Diamantis, 1.0., Nair, A.P.K., Hirsch, H.H., Moroni, C. (1989). Tumor
suppression involves down-regulation of interleukin 3 expression in
hybrids between autocrine mastocytoma and interleukin 3-dependent parental
mast cells. Proc. Natl. Acad. Sci. USA 86, 9299-9302.

Diaz, M.O., Rubin, C.M., Harden, A., Ziemen, S., Larson, R., LeBeau, M.M.,
and Rowley, J.D. (1990). Deletions of interferon genes in acute
lumphoblastic leukemia. N. Engl. J. Med., 322, 77-82.

Doolittle, R.F., Hunkapiller, M.W., Hood, L.E., Devare, S.G., Robbins,
K.C., Aaronson, S.A., Antoniades, H.N. (1983). Simian Sarcoma Virus onc
Gene, v-sis, Is Derived from the Gene (or Genes) Encoding a Platelet-
Derived Growth Factor. Science 221, 275-277.

Downward, J., Yarden, Y., Mayes, E., Scrace, 6., Totty, N., Stockwell, P.,
Ullrich, A., Schlessinger, J., Waterfield, M.D. (1984). Close similarity
of epidermal growth factor receptor and v-erb-B oncogene protein
sequences. Nature 307, 521-527.

Drummond, I.A., Maddon, S.L., Rohwer-Nutter, P., Bell, G.I., Sukhatme,
V.P., and Rauscher III, F.J. (1992). Repression of the insulin-like growth
factor II gene by the Wilms tumor suppressor WTI. Science 257, 674-678.

Duncan, E.L., Whitaker, N.J., Moy, E.L., and Reddel, R.R. (1993).
Assignment of SV40-immortalized cells to more than one complementation
group for immortalization. Experimental Cell Res. 205, 337-344.

El-Deiry, W.A., Tokino, T., Velculescu, V.E., Levy, 0.8., Parsons, R.,
Trent, J.M., Lin, 0., Mercer, W.E., Kinzler, K.W., Vogelstein, B. (1993).
WAFI, a potential mediator of p53 tumor suppression. Cell 75, 817-825.

Eliyahu D., Goldfinger, N., Pinhasi-Kimki, O., Shaulsky, 6., Skurnik, Y.,
Arai, N., Rotter, V., and Oren, M. (1988). Meth A fibrosarcoma cells
express two transforming mutant p53 species. Oncogene 3, 313-321.

Eliyahu, D., Michalovitz, D., Eliyahu, S., Pinhasi-Kimhi, O., Oren, M.
(1989). Wild-typelp53 can inhibit.oncogene-mediated focus formation. Proc.
Natl. Acad. Sci. USA 86, 8763-8767.

Ellis, R.W., DeFeo, 0., Shih, T.Y., Gonda, M.A., Young, H.A., Tsuchida,
N., Lowry, D.R., and Scolnick, E.M. (1981). The p21 src genes of Harvey
and Kirsten sarcoma viruses originate from divergent members of a family
of normal vertebrate genes. Nature 292, 506-511.

Eva, A., Robbins, K.C., Andersen, P.R., Srinivasan, A., Tronick, S.R.,
Reddy, E.P., Ellmore, N.W., Galen,.A.T., Lautenberger, J.A., Pappas, T.S.,
Westin, E.H., Wong-Staal, F., Gallo, R.C., Aaronson, S.A. (1982).
Cellular genes analogous to retroviral onc genes are transcribed in human

49
tumour cells. Nature 295, 116-119.

Fearon, E.R., Cho, K.R., Nigro, J.M., Kern, S.E., Simons, J.W., Ruppert,
J.M., Hamilton, S.R., Preisinger, A.C., Thomas, G., Kinzler, K.W.,
Vogelstein, B. (1990). Identification of a chromosome 18q gene that is
altered in colorectal cancers. Science 247, 49-56.

Fearon, E.R., Hamilton, S.R., Vogelstein, 8. (1987) Clonal analysis of
human colorectal tumors. Science 238, 193-197.

Fearon, E., Vogelstein, 8., and Feinburg, A.P. (1984). Somatic deletion
and duplication of genes on chromosome 11 in Wilms’ tumors. Nature 309,
176-178.

Fields, S. and Jang, S.K. (1990). Presence of a potent transcription
activating sequence in the p53 protein. Science 249, 1046-1049.

Finlay, C.A., Hinds, P.W., Levine, A.J. (1989). The p53 proto-oncogene can
act as a suppressor of transformation. Cell 57, 1083-1093.

Fountain, J.W., Karayiorgou, M., Ernstoff, M.S., Kirkwood, J.M., Vlock,
D.R., Titus-Ernstoff, L. Bouchard, 8., Vijayasaradhi, S., Houghton, A.N.,
Lahti, J., Kidd, V.J., Housman, D.E., and Dracopoli, N.C. (1992).
Homozygous deletions within human chromosome band 9p21 in melanoma. Proc.
Natl. Acad. Sci. USA 89, 10557-10561.

Franke, U. (1978). Retinoblastoma and chromosome 13. Birth Defects 12,
131-137.

Franke, U., Holmes, L.8., Atkins, L., and Riccardi, V.M. (1979). Aniridia-
Wilms’ tumor association: evidence for specific deletion of 11p13.
Cytogenetics and Cell Genetics 24, 185-192.

Friend, S.H., Bernards, R., Rogelj, S., Weinberg, R.A., Rapaport, J.M.,
Albert, 0.M., & Dryja, T.P. (1986). A human DNA segment with properties of
the gene that predisposes to retinoblastoma and osteosarcoma. Nature 323,
643-646.

Friend, S.H., Dryja, T.P. Weinburg, R.A. (1988) Oncogenes and tumor-
suppressing genes. NEJM 318, 618-622.

Fujimoto, Y., Hampton L.L., Wirth, P.J., Wang, N.J., Xie, J.P., and
Thorgeirsson, 5.5. (1994). Alterations of tumor suppressor genes and
allelic losses in human hepatocellular carcinomas in China. Cancer
Research 54, 281-285.

Fung, Y.T., Murphree, L.A., T’Ang, A., Oian, J., Hinrichs, S.H., Benedict,
W.F. (1987). Structural evidence for the authenticity of the human
retinoblastoma gene. Science 236, 1657-1661.

Gateff, E. (1982). Cancer, genes, and development: the Drosophila case.
Adv. Cancer Res. 37, 33-73.

50

Gaylord, W.H., Jr. (1955). Virus-like particles associated with the Rous
sarcoma as seen in sections of the tumor. Cancer Res. 15, 80-83.

Gessler, M., Poustka, A., Cavenee, W., Neve, R.L., Orkin, S.H., and Bruns,
G.A.P. (1990). Homozygous deletion in Wilms tumors of a zinc-finger gene
identified by chromosome jumping. Nature 343, 774-778.

Goyette; M.C., Cho, K., Fasching, C.L., Levy, 0.8., Kinzler, K.W.,
Paraskeva, C., Vogelstein, 8., Stanbridge, E.J. (1992). Progression of
colorectal cancer is associated with multiple tumor suppressor gene
defects but inhibition of tumorigenicity is accomplished by correction of
any single defect via chromosome transfer. Molecular and Cellular Biology
12, 1387-1395.

Grundy, P., Koufos, A., Morgan, K., Li, F.P., Meadows, A.T., and Cavenee,
W.K., (1988). Familial predisposition to Wilms’ tumor does not map to
short arm of chromosome 11. Nature 336, 374-376.

Hall, A., Marshall, C.J., Spurr, N.K., Weiss, R.A. (1983). Identification
of transforming gene in two human sarcoma cell lines as a new member of
the ras gene family located on chromosome 1. Nature 303, 396-400.

Hamel, P.A., Gill, R.M., Phillips, R.A., Gallie, 8.L. (1992).
Transcriptional repression of the E2-containing promoters EI-IaE, C-myc
and R81 by the product of the RBI gene. Mol. Cell. Biol. 12, 3431-3438.

Harper, J.W., Adami, G.R., Wei, N., Keyomarsi, K., and Elledge, S.J.
(1993). The p21 cdk-interacting protein Cipl is a potent inhibitor of GI
cyclin-dependent kinases. Cell 75, 805-816.

Harris, H. (1988). The~analysis of malignancy by cell fusion: The position
in 1988. Cancer Research 48, 3302-3306.

Harris, H., Miller, D.J., Klein, 6., Worst, P. Tachibana, T. (1969).
Suppression of malignancy by cell fusion. Nature 223, 363-368.

Harvey, J.J. (1964). An unidentified virus which causes the rapid
production of tumours in mice. Nature 204, 1104-1105.

Haubruck, H., McCormick, F. (1991). Ras p21: effects and regulation.
Biochimica et Biophysica Acta 1072, 215-229.

Henry, I., Grandjouan, S., Couillin, P., Barichard, F., Huerra-Jean-
pierre, C., Glaser, T., Philip, T., Lenoir, G., Chaussain, J.L., and
Junien, C., (1989). Tumor-specific loss of 11p15.5 alleles in del11p13
Wilms tumor and in familial adenocortical carcinoma. Proc. Natl. Acad.
Sci. USA 86, 3247-3251.

Herskowitz, I. (1987) Functional inactivation of genes by dominant
negative mutations. Nature 239, 219-222.

Hiebert, S.H., Chellappan, S.P., Horowitz, J.M., and Nevins, J.R. (1992).
The interaction of R8 with E2F coincides with an inhibition of the

51
transcriptional activity of E2F. Genes Dev. 6, 177-185.

Hinds, P., Finlay, C. Levine, A.J. (1989). Mutation is required to
activate the p53 gene for cooperation with the ras oncogene and
transformation. Journal of Virology 63, 739-746.

Hollstein, M., Sidransky, 0., Vogelstein, 8., Harris, C.C. (1991). p53
mutations in human cancers. Science 253, 49-53.

Horowitz, J.M., Park, S-H, Bogenmann, E., Cheng, J-C., Yandell, D.W.,
Kaye, F.J., Minna, J.0., Dryja, T.P., Weinberg, R.A. (1990). Frequent
inactivation of the retinoblastoma anti-onCOgene is restricted to a subset
of human tumor cells. Proc. Natl. Acad. Sci. USA 87, 2775-2779.

Huang, H-J.S., Yee, J-K., Shew, J-Y., Chen, P-L., Bookstein, R.,
Friedmann, T., Lee, E.Y-H.P, Lee, W-H. (1988). Suppression of the
neoplastic phenotype by replacement of the R8 gene in human cancer cells.
Science 242, 1563-1566.

Huerra, L., Kakati, S., Gibas, L., Patrick, E., and Sandberg, A. (1986).
Brief clinical report: Gardner syndrome in a man with an interstitial
deletion of 5q. American Journal of Medical Genetics 25, 473-476.

Hunter, T. (1991). Cooperation between oncogenes. Cell 64, 249-270.

Hunter, T., Sefton, 8.M. (1980). Transforming gene product of Rous sarcoma
virus phosphorylates tyrosine. Proc. Natl. Acad. Sci. USA 77, 1311-1315.

Hurley, J.B., Simon, M.I., Teplow, 0.8., Robishaw, J.0., Gilman, A.G.
(1984). Homologies Between Signal Transducing G Proteins and ras Gene
Products. Science 226, 860-862.

James, C.0., He, J., Carlbom, E., Nordenskjold, M., Cavenee, W.K., and
Collins, V.P. (1991). Chromosome 9 deletion mapping reveals interferon a
and interferon 6-1 gene deletions in human glial tumors. Cancer Res. 51,
1684-1688.

Joslyn, G., Carlson, M., Thliveris, A., Albertson, H., Gelbert, L.,
Samowitz, W., Groden, J., Stevens, J., Spirio, L., Robertson, M.,
Sargeant, L., Krapcho, K., Wolff, E., Burt, R., Hughes, J.P., Harrington,
J., McPherson, J., Wasmuth, J., LePaslier, 0, Abderrahim, H., Cohen, 0.,
Leppert, M., White, R. (1991). Identification of deletion mutations and
three new genes at the familial polyposis locus. Cell 66, 610-613.

Kaelbling, M., and Klinger, H.P. (1986). Suppression of tumorigenicity in
somatic cell hybrids III. Cosegregation of human chromosome 11 of a normal
cell and suppression of tumorigenicity in intraspecies hybrids of normal
diploid X malignant cells. . Cytogenet Cell Genet 41, 65-70.

Kaelbling, M., Roginski, R.S., Klinger, H.P. (1986). DNA polymorphisms
indicate loss of heterozygosity for chromosome 11 of 098AH2 cells.
Cytogenet Cell Genet 41, 240-244.

52

Kao, F.-T. and Hartz, J.A. (1977). Genetic and tumorigenic characteristics
of cell hybrids formed in vivo between injected tumor cells and host cells
J. Natl. Cancer Inst. 59, 409-413.

Kinzler, K.W., Nilbert, M.C., Su, L-K., Vogelstein, 8., Bryan, T.M., Levy,
0.8., Smith, K.J., Preisinger, A.C., Hedge, P., McKechnie, 0., Finniear,
R., Markham, A., Groffen, J., Boguski, M.S., Altschul, S.F., Horii, A.,
Ando, H., Miyoshi, Y., Miki, Y., Nishisho, I., Nakamura, Y. (1991)
Identification of FAP locus genes from chromosome 5q21. Science 253, 661-
665.

Kirsten, W.H., and Mayer, L.A. (1967). Morphologic responses to a murine
erythroblastosis virus. Journal of the National Cancer Institute 39, 311-
319.

Klein, G. (1988) Tumor suppressor genes. J. Cell. Sci. Suppl. 10, 171-180.

Klingelhutz, A.J., Wu, S.-Q., Huang, J., and Reznikoff, C.A. (1992). Loss
of 3p13-+p21.2 in tumorigenic reversion of a hybrid between isogeneic
nontumorigenic and tumorigenic human uroepithelial cells. Cancer Research
52, 1631-1634.

Klinger, H.P. (1980). Suppression of tumorigenicity in somatic cell
hybrids I. Suppression and reexpression of tumorigenicity in diploid human
X 098AH2 hybrids and independent segregation of tumorigenicity from other
cell phenotypes. Cytogenet. Cell Genet. 27, 254-266.

Klinger, H.P. and Kaelbling, M. (1986). Suppression of tumorigenicity in
somatic cell hybrids 1V. Chromosomes of normal human cells associated with
suppression of tumorigenicity in hybrids with D98AH2 carcinoma cells.
Cytogenet Cell Genet 42, 225-235.

Klinger, H.P. and Shows, T.B. (1983). Suppression of tumorigenicity in
somatic cell hybrids 11. Human chromosomes implicated as suppressors of
tumorigenicity in hybrids with Chinese hamster ovary cells. JNCI 71, 559-
569.

Klinger, H.P., Baim, A.S., Eun, C.K., Shows, T.B., Ruddle, F.H. (1978).
Human chromosomes which affect tumorigenicity in hybrids of diploid human
with heteroploid human or rodent cells. Cytogenet. Cell Genet. 22, 245-
249.

Knudson, A.G. (1993). Antioncogenes and human cancer. Proc. Natl. Acad.
Sci. USA 90, 10914-10921.

Knudson, A. G. Jr. (1971). Mutation and cancer: statistical study of
retinoblastoma. Proc. Natl. Acad. Sci. USA 68, 820-823.

Koufos, A., Grundy, P., Morgan, K., Aleck, K.A., Hadro, T., Lampkin, 8.C.,
Kalbakji, A., and Cavenee, W.K. (1989). Familial Wiedemann-Beckwith
syndrome and a second Wilms’ tumor locus map to 11p15.5. Am. J. Hum.
Genet. 44, 711-719.

53

Koufos, A., Hansen, M.F., Lampkin, B.C., Workman, M.L., Capeland, 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 309, 170-172.

Krontiris, T.G., Cooper, G.M. (1981). Transforming activity'of human tumor
DNAs. Proc. Natl. Acad. Sci. USA 78, 1181-1184.

Kucherlapati, R. and Shin, S. (1979). Genetic controle of tumorigenicity
in interspecific mammalian cell hybrids. Cell 16, 639-638.

Land, H., Chen, A.C., Morgenstern, J.P., Parada, L.F., and Weinbert, R.A.
(1986). Behavior of myc and ras oncogenes in transformation of rat embryo
fibroblasts. Molecular and Cellular Biology 6, 1917-1925.

Lane, 0.P., and Crawford, L.V. (1979). T antigen is bound to a host
protein in SV40-transformed cells. Nature, 278, 261-263.

Langbehein, H., Shih, T.Y., and Scolnick, E.M. (1980) Identifacation of a
normal vertebate cell protein related to the p21 src of Harvy murine
sarcoma virus. Virology 106, 292-300.

Lee, W. H Bookstein, R, Hong, F, Young, L. J Shew, J. -Y., Lee, E. Y-
H. P. (1987a). Human retinoblastoma susceptibility gene: cloning,
identification and sequence. Science 235,1394-1399.

Lee, W.-H., Shew, J.-Y., Hong, F.0., Sery, T.W., Donoso, L.A., Young, L.-
J., Bookstein, R., Lee, E.Y.-H.P. (1987b). The retinoblastoma
susceptibility gene encodes a nuclear phosphoprotein associated with DNA
binding activity. Nature 329, 642-645.

Leppert, M., Dobbs, M., Scambler; P., O’Connell, P., Nakamura, Y.,
Stauffer, 0., Woodward, S., Burt, R., Hughes, J., Gardner, E., Lathrop,

M., Wasmuth, J., Lalouel, J.-M., and White, R. (1987). The genes for
familialapolyposis coli maps to the long arm of chromosome 5. Science 238,
4 - 4 .

Levine, A.J., and Momand, J. (1990). Tumor suppressor genes: The p53 and
retinoblastoma sensitivity genes and gene products. Biochem Biophys Acta
1032, 119-136.

Linzer, 0.1.H., and Levine, A.J. (1979). Characterization of a 54,000 MW
cellular SV40 tumor antigen present in SV40-transformed cells and
uninfected embryonal carcinoma cells. Cell 17, 43-52.

Ludlow, J.W., Shon, J., Pipas, J.M., Livingston, 0.M., DeCaprio, J.A.
(1990). The retinoblastoma susceptibility gene product undergoes cell
cycle-dependent dephosphorylation and binding to and release from SV40
large T. Cell 60, 387-396.

Madden, S.L., Cook, 0.M., Morris, J.F., Gashler, A., Sukhatme, V.P., and
Rauscher III, F.J. (1991). Transcriptional repression mediated by the WTI
Wilms Tumor gene product. Science 253, 1550-1553.

54

Maheswaran, S., Park, S., Bernard, A., Morris, J.A., Rauscher 111, F.J.,
Hill, D.E., and Haber, 0.A. (1993). Physical and functional interaction
between WTI and p53 proteins. Proc. Natl. Acad. Sci. USA 90, 5100-5104.

Malkin, 0., Li, F.P., Strong, L.C., Fraumeni, Jr., J.F., Nelson, C.E.,
Kim, D.H., Kassel, J., Gryka, M.A., Bischoff, F.Z., Tainsky, M.A., Friend,
S.H. (1990). Germ line p53 mutations in a familial syndrome of breast
cancer, sarcomas, and other neoplasms. Science 250, 1233-1238.

Marshall, C.J., Hall, A., Weiss, R.A. (1982). A transforming gene present
in human sarcoma cell lines. Nature 299, 171-173.

Martin, 6.5. (1970). Rous sarcoma virus: A function required for the
maintenance of the transformed phenotype. Nature 227, 1021-1023.

Marx, J. (1989) Many gene changes found in cancer. Science 246, 1386-1388.
Matsunaga, E. (1981). Genetics of Wilms’ tumor. Human Genetics 58, 231-
246.

Melcher, 8.M., McGinnis, W., and Gehring, W.J. (1985). Molecular cloning
of lethal (2) giant larvae, a recessive oncogene of Drosophila
melanogaster. EMBO J. 4, 1551 or 1337.

Miller, R.W., Fraumeni, J.F., and Manning, M.D. (1964). Association of
Wilms’ tumor with aniridia, hemihypertrophy, and other congenital
anomalies. N. Engl. J. Med. 270, 922.

Moroco, J.R., Solt, 0.8., Polverini, P.J. (1990) Sequential loss of
suppressor genes for three specific functions during in vivo
carcinogenesis. Laboratory Investigation 63, 298-306.

Muggleton-Harris, A.L., and DiSimone, D.W. (1980) Replicative potentials
of various fusion products between WI-38 and SV40 transformed WI-38 cells
and their components. Somatic Cell Genetics 6, 689-698.

Muleris, M., Salmon, R.J., Zafrani, 8., Girodet, J., Dutrillaux, B. (1985)
Consistent deficiencies of chromosome 18 and the short arm of chromosome
17 in eleven cases of human large bowel cancer: a possible recessive
determinism. Ann. Genet. 28, 206-213.

Newmark, P. (1987). Oncogenes and cell growth. Nature 327, 101-102.

Nigro, J.M., Baker, S.J., Preisinger, A.C., Jessup, J.M., Hostetter, R.,
Cleary, K., Bigner, S.H., Davidson, N., Baylin, S., Devilee, P., Glover,
T., Collins, F.S., Weston, A., Modali, R., Harris, C.C., and Vogelstein,
B. (1989). Mutations in the p53 gene occur in diverse human tumor types.
Nature 342, 705-708.

Ning, Y., and Pereira-Smith, O.M. (1991). Molecular genetic approaches to
the study of cellular senescence. Mutat. Res. 256, 303-310.

Ning, Y., Weber, J.L., Killary, A.N., Ledbetter, D.H., Smith, J.R., and

55

Pereira-Smith, O.M. (1991). Genetic analysis of indefinite division in
human cells: Evidence for a cell senescence-related gene(s) on human
chromosome 4. Proc. Natl. Acad. Sci. USA 88, 5653-5639.

Nishisho, I., Nakamura, Y., Miyoshi, Y., Miki, Y., Ando, H., Horii, A.,
Koyama, K., Utsunomiya, J., Baba, S., Hedge, P., Markham, A., Krush, A.J.,
Petersen, 6., Hamilton, S.R., Nilbert, M.C., Levy, 0.8., Bryan, T.M.,
Priesinger, A.C., Smith, K.J., Su, L-K., Kinzler, K.W., Vogelstein, B.
(1991). Mutations of chromosome 5q21 genes in FAP and colorectal cancer
patients. Science 253, 665-669.

NGkawa, N., Ishikiriyama, S., Takahashi, S., Inagawa, A., Tonoki, M.,
Ohta, Y., Hase, N., Kamei, T., and Kaju, T. (1986). ThelWiedemann-Beckwith
syndrome: pedigree studies on five families with evidence for autosomal
dominant inheritance with variable expressivity. Am. J. Med. Genet. 24,
41-55.

Olopade, O.I., Buchhagen, 0.L., Malik, K., Sherman, J., Nobori, T., Bader,
S., Nau, M.M., Gazdar, A.P., Minna, J.D. and Diaz, M.O. (1993). Homozygous
loss of the interferon genes defines the critical region on 9p that is
deleted in lung cancers. Cancer Res. 53, 2410-2415.

Olschwang, S., Tiret, A., Laurent-Puig, P., Muleris, M., Parc, R., Thomas,
G. (1993). Restriction of ocular fundus lesions to a specific subgroup of
APC mutations in adenomatous polyposis coli patients. Cell 75, 959-968.

Opper, M., Schuler, G., and Melcher, 8.M., (1987). Hereditary suppression
of lethal (2) giant larvae malignant tumor development in Drosophila by
gene transfer. Oncogene 1, 91-96.

Orkin, S.H., Goldman, 0.S., and Sallan, S.E. (1984). Development of
homozygosity for chromosome 11p markers in Wilms tumor. Nature 309, 172-

Pereira-Smith, 0.M., and Smith, J.R. (1981). Expression of SV40 T antigen
in finite life-span hybrids of normal and SV40-transformed fibroblasts.
Somatic Cell Genetics 7, 411-421.

Pereira-Smith, 0.M., and Smith, J.R. (1983). Evidence for the recessive
nature of cellular immortality. Science, 221, 964-966.

Pereira-Smith, 0.M., and Smith, J.R. (1988). Genetic analysis of
indefinite division in human cells: Identification of four complementation
groups. Proc. Natl. Acad. Sci. USA 85, 6042-6046.

Pereira-Smith, 0.M., Stein, G.H., Robetorye, S., and Meyer-Demarest, S.
(1990). Imnortal phenotype of the HeLa variant 098 is recessive in
hyubrids formed with normal human fibroblasts. J. of Cell. Physiol., 143,
222-225.

Pritchard-Jones, K., Fleming, S., Davidson, 0., Bickmore, W., Porteous,
0., Gosden, C., Bard, J., Buckler, A., Pelletier, J., Housman, 0., van
Heyningen, V., and Hastie, N. (1990). The candidate Wilms’ tumour gene is

56
involved in genitourinary development. Nature 346, 194-197.

Pulciani, S., Santos, E., Lauver, A.V., Long, L.K., Aaronson, S.A.,
Barbacid, M. (1982). Oncogenes in solid human tumours. Nature 300, 539-
542.

Rauscher III, F.J., Morris, J.F., Tournay, D.E., Cook, 0.M., and Curran,
T. (1990) Binding of the Wilms’ tumor locus zinc finger protein to the
EGR-I consensus sequence. Science 250, 1259-1262.

Reeve, A.E., Housiaux, P.J., Gardner, R.J.M., Chewings, W.E., Grindley,
R.M., and Millow, L.J. (1984). Loss of a Harvey ras allele in sporadic
Wilms’ tumor. Nature 309, 174-176.

Reeve, A.E., Sih, S.A., Raizis, A.N., and Feinburg, A.P. (1989). Loss of
allelic heterozygosity at a second locus on chromosome 11 in Wilms’ tumor
cells. Mol. Cell. Biol. 9, 1799-1803.

Reichmann, A., Martin, P., Levin, B. (1981) Chromosomal banding patterns
in human large bowel cancer. Int. J. Cancer 28, 431-440.

Rempel, S.A., Schwechheimer, K., Davis, R.L., Cavenee, W.K., Rosenblum,
M.L. (1993). Loss of heterozygosity for loci on chromosome 10 is
associated with morphologically malignant meningioma progression. Cancer
Research 53, 2386-2392.

Resnitzky, 0., Yarden, A., Zipori, 0., and Kimchi, A. (1986). Autocrine'B-
related interferon controls c-myc suppression and growth arrest during
hematopoietic cell differentiation. Cell 46, 31-40.

Riccardi, V.M., Sujansky, E., Smith, A.C., and Franke, U. (1978).
Chromosomal imbalance in the aniridia-Willms tumor association: 11p
interstitial deletion. Pediatrics 61, 604-610.

Rose, E.A., Glaser, T., Jones, C., Smith, C.L., Lewis, W.H., Call, K.H.,
Minden, M., Champagne, E., Bonetta, L., Yeger, H., and Housman, D.E.
(1900). Complete physical map of the WAGR region of 11p13 localizes a
candidate Wilms’ tumor gene. Cell 60, 495-508.

Rous, P. (1911). A sarcoma of the fowl transmissible by an agent separable
from the tumor cells. J. Exp. Med. 13, 397-411.

Ruppert, J.M., Tokino, K., and Sidransky, D. (1993). Evidence for two
bladder cancer suppressor loci on human chromosome 9. Cancer Research 53,
5093-5095.

Sager, R. (1989). Tumor suppressor genes: The puzzle and the promise.
Science 246, 1406-1412.

Sager, R. and Kovac, P.E. (1978). Genetic analysis of tumorigenesis: 1.
Expression of tumor-forming ability in hamster hybrid cell lines. Somat.
Cell. Genet. 4, 375-391.

57

Sasaki, M., Okamoto, N., Sato, C., Sugio, K., Soejima, J-i., Iwama, T.,
Ikeuchi, T., Tonomura, A., Miyaki, M., Sasazuki, T. (1989) Loss of
constitutional heterozygosity in colorectal tumors from patients with
familial polyposis coli and those with nonpolyposis colorectal carcinoma.
Cancer Research 49, 4402-4406.

Saxon, P.J., Srivatsan, E.S., Stanbridge, E.J. (1986). Introduction of
human chromosome 11 via microcell transfer controls tumorigenic expression
of HeLa cells. The EMBO Journal 5, 3461-3466.

Scheffner, M., Werness, 8.A., Huibregtse, J.M., Levine, A.J., Howley, P.M.
(1990). The E6 oncoprotein encoded by human papillomavirus types 16 and 18
promotes the degradation of p53. Cell 63, 1129-1136.

Scheinberg, D.A. and Strand, M. (1980). Transformation-related proteins
associated with Kirsten sarcoma virus. Virology 106, 335-348.

Schwab, M., Alitalo, K., Klempnauer, K.-H., Varmus, H.E., Bishop, J.M.,
Gilbert, F., Brodeur, G., Goldstein, M., and Trent, J. (1983). Amplified
DNA with limited homology to myc cellular oncogene is shared by human
neuroblastoma cell lines and a neuroblastoma tumour. Nature 305, 245-248.

Scolnick, E.M., Papageorge, A.G., and Shih, T.Y. (1979). Guanine
nucleotide-binding activity as an assay for src protein of rat-derived
murine sarcoma viruses. Proc. Natl. Acad. Sci. USA 76, 5355-5359.

Seemayer, T.A. and Cavenee, W.K. (1989). Biology of disease: molecular
mechanisms of oncogenesis. Laboratory Investigation 60, 585-599.

Shih, C., Padhy, L.C., Murray, M., Weinberg, R.A. (1981). Transforming
genes of carcinomas and neuroblastomas introduced into mouse fibroblasts
Nature 290, 261-264.

Shih, C., Shilo, B.-Z., Goldfarb, M.P., Dannenberg, A., Weinberg, R.A.
(1979a). Passage’ of phenotypes of chemically transformed cells via
transfection of DNA and chromatin. Proc. Natl. Acad. Sci. USA 76, 5714-
5718.

Shih, T.Y., Weeks, M.O., Young, H.A., Scolnick, E.M. (1979b).
Identification of a sarcoma virus-coded phosphoprotein in nonproducer
cells transformed by Kirsten or Harvey murine sarcoma virus. Virology 96,
64-79.

Shih, T.Y., Weeks, M.O., Young, H.A., Scolnick, E.M. (1979c). p21 of
Kirsten murine sarcoma virus is thermolabile in a viral mutant temperature
sensitive for the maintenance of transformation. Journal of Virology 31,
546-556.

Shilo, B.-Z., and Weinburg, R.A. (1981). DNA sequences homologous to
vertebrate oncogenes are conserved in Drosophila Melanogaster. Proc. Natl.
Acad. Sci. USA 78, 6789-6792.

Shimizu, K., Goldfarb, M., Perucho, M., Wigler, M. (1983a). Isolation and

58

preliminary characterization of the transforming gene of a human
neuroblastoma cell line. Proc. Natl. Acad. Sci. USA 80, 383-387.

Shimizu, K., Goldfarb, M., Suard, Y., Perucho, M., Li, Y., Kamata,
T.,Feramisco, J., Stavenzer, E., Fogh, J., Wiggler, M.H. (1983b). Three
human transforming genes are related to the viral ras oncogenes. Proc.
Natl. Acad. Sci. USA 80, 2112-16.

Silagi, S. (1967). Hybridization of a malignant melanoma cell line with l
cells in vitro. Cancer Res. 27, 1953-1957.

Smith, K.J., Johnson, K.A., Bryan, T.M., Hill, D.E., Markowitz, S.,
Willson, J.K.V., Paraskeva C., Petersen, G.M., Hamilton, S.R., Vogelstein,
8., Kinzler, K.W. (1993). The APC gene product in normal and tumor cells.
Proc. Natl. Acad. Sci. USA 90, 2846-2850.

Solomon, E., Voss, R., Hall, V., Bodmer, W.F., Jass, J.R., Jeffreys, A.J.,
Lucibello, F.C., Patel, 1., and Rider, S.H. (1987). Chromosome 5 allele
loss in human colorectal carcinomas. Nature 328, 616-619.

Sotelo-Avila, C., Gooch, H.N. (1976). Neoplasms associated with the
Beckwith-Wiedemann syndrome. Perspect. Pediatr. Pathol. 3, 255-272.

Sparkes, R.S., Murphree, A.L., Lingua, R.W., Sparkes, M.C., Field, L.L.,
Funderburk, S.J., Benedict, W.F. (1983). Gene for hereditary
retinoblastoma assigned to human chromosome 13 by linkage to esterase 0.
Science 219, 971-973.

Spector, D.H., Varmus, H.E., and Bishop, J.M. (1978). Nucleotide sequences
related to the transforming genelof avian sarcoma virus are present in the
DNA of uninfected vertebrates. Proc. Natl. Acad. Sci. 75, 4102-4106.

Spirio, L., Olschwang, S., Groden, J., Robertson, M., Samowitz, W.,
Joslyn, G., Gelbert, L., Thliveris, A., Carlson, M., Otterud, 8., Lynch,
H., Watson, P., Lynch, P., Laurent-Puig, P., Burt, R., Hughes, J.P.,
Thomas, G., Leppert, M., White, R. (1993). Alleles of the APC gene: An
attenuated form of familial polyposis. Cell 75, 951-957.

Sporn, M.B. & Roberts, A.B. (1988). Peptide growth factors are
multifunctional. Nature 332, 217-219.

Srivastava, S., Zou, Z., Pirollo, K., Blattner, W., Chang. E.H. (1990).
Germ-line transmission of a mutated p53 gene in a cancer-prone family with
Li-Fraumeni syndrome. Nature 348, 747-749.

Srivatsan, E.S., Benedict, W.F., Stanbridge, E.J. (1986). Implication of
chromosome 11 in the suppression of neoplastic expression in human cell
hybrids. Cancer Research 46, 6174-6179.

Stanbridge, E.J. (1976). Suppression of malignancy in human cells. Nature
260, 17-20.

Stanbridge, E.J., and Wilkinson, J. (1978). Analysis of nalignancy in

59

human cells: Malignant and transformed phenotypes are under separate
genetic control. Proc. Natl. Acad. Sci. USA 75, 1466-1469.

Stanbridge, E.J., Der, C.J., Doersen, C.-J., Nishimi, R.Y., Peehl, 0.M.,
Weissman, B.E., Wilkinson, J.E. (1982). Human cell hybrids: Analysis of
transformation and tumorigenicity. Science 215, 252-259.

Stanbridge, E.J., Flandermeyer, R.R., Daniels, D.W., Nelson-Rees, W.A.
(1981). Specific chromosome loss associated with the expression of
tumorigenicity in human cell hybrids. Somatic Cell Genetics 7, 699-712.

Stehelin, 0., Varmus, H.E., Bishop, J.M., and Vogt, P.K. (1976). DNA
related to the transforming gene(s) of avian sarcoma viruses is present in
normal avian DNA. Nature 260, 170-173.

Stenger, J.E., Mayr, G.A., Mann, K., Tegtmeyer, P. (1992). Formation of
stable p53 homotetramers and multiples of tetramers. Molecular
Carcinogenesis 5, 102-106.

Stoker, M. (1964). Regulation of growth and orientation in hamster cells
transformed by polyoma virus. Virology 24, 165-174.

Stoler, A. and Bouck, N. (1985). Identification of a single chromosome in
the normal human genome essential for suppression of hamster cell
transformation. Proc. Natl. Acad. Sci. USA 82, 570-574.

Straus, 0.S., Mohandas, T. (1987). Growth suppression of hybrids between
transformed cells and normal fibroblasts in serum-free medium: correlation
with retention of human chromosomes. Somatic Cell and Molecular Genetics
13, 587-596.

Su, L-K., Johnson, K.A., Smith, K.J., Hill, D.E., Vogelstein, 8., Kinzler,
K.W. (1993). Association between wild type and mutant APC gene products.
Cancer Research 53, 2728-2731.

Takehara, K., LeRoy, E. C., and Grotendorst, G.R. (1987). TGF-B inhibition
of endothelial cell proliferation: Alteration of EGF binding and EGF-
induced growth-regulatory (Competence) gene expression. Cell 49, 415-422.

Tanaka, K., Oshimura, M., Kikuchi, R., Seki, M., Hayashi, T, Miyaki, M.
(1991). Suppression of tumorigenicity in human colon carcinoma cells by
introduction of normal chromosome 5 or 18. Nature 349, 340-342.

Tlsty, T.D., White, A., Sanchez, J. (1992). Suppression of gene
amplification in human cell hybrids. Science 255, 1425-1427.

Trahey, M., McCormick, F. (1987). A cytoplasmic protein stimulates normal
N-ras p21 GTPase, but does not affect oncogenic mutants. Science 238, 542-
545.

Tsai, Y.C., Nichols, P.W., Hiti, A.L., Williams, 2., Skinner, 0.G., Jones,
P.A. (1990). Allelic losses of chromosomes 9, 11, and 17 in human bladder
cancer. Cancer Research 50, 44-47.

60

Tsujimoto, Y., Yunis, J., Onorato-Showe, L., Erickson, J., Nowell, P.C.,
Croce, C.M. (1984). Molecular cloning of the chromosomal breakpoint of 8-
cell lymphomas and leukemias with the t(11;14) chromosome translocation.
Science 224, 1403-1406.

Valencia, A., Chardin, P., Wittinghofer, A., Sander, C. (1991). The ras
protein family: Evolutionary tree and role of conserved amino acids.
Biochemistry 30, 4637-4647.

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

Vogelstein, 8., Fearon, E.R., Hamilton, S.R., Kern, S.E., Preisinger,
A.C., Leppart, M., Nakamura, Y., White, R., Smits, A.M.M., Bos, J.L.
(1988). Genetic alterations during colorectal-tumor development. New
England Journal of Medicine 319, 525-532.

Wang, Z.-Y., Madden, S.L., Deuel, T.F., and Rauseher, F.J., III (1992).
The Wilms’ tumor gene product, WTI, represses transcription of the
platlet-derived growth factor A-chain gene. J. Biol. Chem. 267, 21999-
22002.

Wang, Z.-Y., Oiu, Q.-Q., and Devel, T.F. (1993). The Wilms’ tumor gene
product WT1 activates or suppresses transcription through separate
functional domains. J. Biol. Chem. 268, 9172-9175.

Waziri, M., Patel, S.R., Hanson, J.W., and Bartley, J.A. (1983).
Abnormality of chromosome 11 in patients with features of Beckwith-
Wiedemann syndrome. Journal of Pediatrics 102, 873-876.

Weinberg, R.A. (1989). Positive and negative controls on cell growth.
Biochemistry 28, 8263-8269.

Weiss, R., Teich, N., Varmus, M., Coffin, J. eds. (1982). Origins of
contemporary RNA tumor virus research. In RNA Tumor Viruses. (New York:
Cold Spring Harbor Laboratory), pp. 1-24.

Weissman, B.E., Stanbridge, E.J. (1983). Complementation of the
tumorigenic phenotype in human cell hybrids. JNCI 70, 667-672.

Whitaker, N.J., Kidston, E.L., and Reddel, R.R. (1992). Finite life span
of hybrids formed by fusion of diferent simian virus 40-immortalized human
cell lines. J. Virol., 66, 1202-1206.

Wiener, F., Klein, 6., and Harris, H. (1974a). The analysis of malignancy
by cell fusion. V. Further evidence'of the ability of normal diploid cells
to suppress malignancy. J. Cell. Sci. 15, 177-183.

Wiener, F., Klein, 6., and Harris, H. (1974b). The analysis of malignancy
by cell fusion. V1. Hybrids between different tumour cells. J. Cell Sci.
16, 189-198.

Willingham, M.C., Pastan, 1., Shih, T.Y., Scolnick, E.M. (1980).

61

Localization of the src gene product of the Harvey strain of MSV to plasma
membrane of transformed cells by electron microsc0pic immunocytochemistry.
Cell 19, 1005-1014.

Wolf, 0., Harris, N., and Rotter, V. (1984). Reconstitution of p53
expression in a nonproducer Ab-MuLV-transformed cell line by transfection
of a functional p53 gene. Cell 38, 119-126.

Wynford-Thomas, 0., Bond, J.A., and Paterson, H. (1989). Suppression of
transformation and immortality in human/Chinese hamster fibroblast hybrids
- a model for suppressor gene isolation. Int. J. Cancer 42, 293-299.

Yarden, A., and Kimchi, A. (1986). Tumor necrosis factor reduces c-myc
expression and cooperates with interferon-y in HeLa cells. Science 234,
1419-1421.

Zajchowski, D.A., Band, V., Trask, 0.K., Kling, D., Connolly, J.L., Sager,
R. (1990). Suppression of tumor-forming ability and related traits in MCF-
7 human breast cancer cells by fusion with immortal mammary epithelial
cells. Proc. Natl. Acad. Sci. USA 87, 2314-2318.

Zarling, J.M., Shoyab, M., Marquardt, H., Hanson, M.B., Lioubin, H.N., and
Todaro, G.J. (1986). Oncostatin M: A growth regulator produced by
differentiated histiocytic lymphoma cells. Proc. Natl. Acad. Sci. USA 83,
9739-9743.

CHAPTER II

Failure of Infinite Life Span Huaan Cells frol Different Ialortality
Coapleaentation Groups to Yield Finite Life Span Hybrids

P. Ann Ryan, Veronica M. Maher, and J. Justin McCormick

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

Michigan State University, East Lansing, MI 48824-1316

62

The observation that fusion of infinite life span cells with finite
life span cells produces hybrid cells with finite life spans led to the
conclusion that an infinite life span in culture is a recessive trait
resulting from loss of the function of a gene or genes that contribute to
an active pragram for cellular senenscence. Furthermore, finding that
certain pairs of infinite life span cells, when fused to one another, can
complement each other to yield finite life span hybrids allowed 30
infinite life span cell lines to be assigned to four immortality
complementation groups (Pereira-Smith and Smith, 1988, Proc. Natl. Acad.
Sci. USA, 85:6042). In the present study, we fused a chromosomally
stable, near diploid, morphologically normal, infinite life span cell
strain, designated HSU-1.1, with its normal, finite life span, precursor
cell strain and obtained finite life span hybrids, as expected if infinite
life span in culture is a recessive trait. However, 14 of the 14 hybrids
from our fusions of MSU-1.1 cells with representative cell lines fromleach
of the four immortality complementation groups, and 38 of the 39 hybrids
from our fusions of infinite life span cells that have been reported to
complement each other, failed to exhibit finite life spans. This result
suggests that infinite life span cells cannot complement each other to
yield finite life span hybrids. In examining this unexpected result, we
obtained evidence that long-term dual drug selection can be deleterious to
hybrid cells even though they carry resistance markers for both drugs,
indicating that the cell death of such hybrids observed in other studies
may have resulted from the cytotoxic effect of long-term drug selection,

rather than from senescence.

63

INTRODUCTION

Normal diploid human fibroblasts in culture have a limited
proliferative potential (Hayflick and Moorhead, 1961; Hayflick, 1965).
However, many human tumor-derived cells can proliferate in culture
indefinitely. Studies by Bunn and Tarrant (1980), Muggleton-Harris and
DeSimone (1980), Pereira-Smith and Smith (1981, 1983), and Pereira-Smith
et al. (1990) indicate that cellular senescence is a genetically
programmed active process and that escape from cellular senescence is the
result of recessive genetic alterations in this program. These
investigators fused a variety of immortal human cells, including tumor-
derived and simian virus-40 (SV40)-transformed strains, to finite life
span human fibroblasts. The hybrids obtained from these fusions had a
limited life span. This phenomenon was interpreted as indicating that
immortality results from loss of function of one or more of the genes
responsible for senescence in normal cells and that, when an infinite life
span cell is fused with a finite life span cell, the hybrid cell senesces
because the latter parent supplies the missing function.

Further support for the theory that immortality results from loss of
function of the gene(s) responsible for senescence comes from the work of
Pereira-Smith and Smith and their colleagues. These investigators fused
various immortal human cell lines with one another and determined the life
span of the hybrid cells. Certain hybrids senesced, indicating that the
cells used in these fusions were able to complement the genetic defects in
each other (Pereira-Smith and Smith, 1983). Through life span analysis of
hybrids obtained from such cell fusions, 30 different immortal human cell

lines were assigned to four inmortality complementation groups (designated

64

65

A, B, C, and 0), with no cell line belonging to more than one group
(Pereira-Smith and Smith, 1988; Ning and Pereira-Smith, 1991). Ning et
al. (1991) reported that introduction of chromosome 4 into cell lines from
complementation group 8 resulted in loss of proliferation and reversal of
the immortal phenotype but had no effect on the proliferative potential of
representative cell lines from the other three complementation groups.
They attributed this result to complementation, by genes on chromosome 4,
of the genetic defect shared by the cell lines belonging to group B.

It has been difficult to determine the genetic and biochemical changes
responsible for acquisition of an infinite life span, because thelmajority
of the cell lines available for study not only exhibit an infinite life
span, but also have many other abnormal characteristics. Many are derived
from tumors, others arose following SV40 transformation (reviewed in Sack
1981), and two arose in populations of cells exposed repeatedly to
carcinogen treatment (Namba et al., 1981; McCormick and Maher 1988). The
majority of these cell lines are highly aneuploid and chromosomally
unstable. To study the genetic changes specific to the process of
immortalization, it is very useful to have infinite life span cells that
maintain a stable diploid or near-diploid karyotype and do not have
abnormal characteristics other than being immortal. Recently, we (Morgan
et al., 1991) successfully derived such cells by transfecting a foreskin-
derived, normal, diploid human fibroblast cell line, designated LGI, with
a plasmid carrying a v-myc oncogene and a drug resistance marker,
selecting for drug resistant transfectants, and expanding individual
clones to the end of their life span. An infinite life span diploid cell
strain, designated MSU-1.0, spontaneously arose in a v-myc-expressing

clonal population. The MSU-1.0 cell strain gave rise to a stable, near-

66
diploid cell strain, designated HSU-1.1. MSU-1.1 cells have a normal
morphology, are only partially growth factor independent, do not form
foci, fornlonly very small colonies at low frequency in 0.33% agarose, and
are not tumorigenic (Morgan et al., 1991).

Because'of their stable near-diploid karyotype'and their near normal
phenotype, we are using MSU-1.1 cells to study the genetic changes
involved in the process of inmortalization. We began by determining
whether their infinite life span resulted fromlactivation of a gene (i.e.,
was a dominant trait) or inactivation of dominant alleles (i.e., was a
recessive trait). We did so by fusing MSU-1.1 cells with finite life span
cells and determining the length of the life spans of the hybrids. The
results indicated that an infinite life span is a recessive trait. We
then fused MSU-1.1 cells with representative cell strains from<each of the
four immortality complementation groups established by Pereira-Smith and
Smith (1988) to determine whether the MSU-1.1 cells belong to one of the
four groups. All of these fusions yielded hybrids with infinite life
spans, indicating that MSU-1.1 cells belong to all four complementation
groups (i.e., had lost function of all the genes or sets of genes
represented by these groups). These results called into question the
existence of the four immortality complementation groups. To confirm the
existence of the four groups, we repeated a select series of fusions among
representatives of each of the four complementation groups, using cell
strains supplied by Pereira-Smith and Smith, and analyzed the life span of
the hybrids. These fusions consistently yielded immortal hybrids (i.e.,
they failed to confirm the existence of the four groups). The results of
additional studies of the effects of drug selection indicated that the
cell death reported by Pereira-Smith and Smith (1988) for the hybrids from

67
immortal cells fused with immortal cells probably resulted from the toxic

effect of long term drug selection.

MATERIALS AND NETHOOS

Cells used
The human cell lines and strains used and their origin and source are

listed in Table 1.

Cell culture

The cells were routinely cultured in Eagle’s minimal essential medium
(MEN) (Grand Island Biological Co., Grand Island, NY) supplemented with
0.2 mM L-aspartic acid, 0.2 mM L-serine, and 1.0 mM sodium pyruvate, or in
McM medium (Ryan et al., 1987) modified by substituting 0.884 mM monobasic
sodium phosphate for 3.0 mM dibasic sodium phosphate, adding 0.054 mM KCl,
and adjusting the NaCl content so that the osmolarity was 285 1 5 mOsm/kg
H20. The medium was supplemented with penicillin (100 units/ml),

streptomycin (100 ug/ml), hydrocortisone (10 pg/ml), 5% supplemented calf

serum (Sterile Systems, Logan, UT), and 5% fetal bovine serum to rake
"complete” medium. The cells were naintained at 37%: in a humidified
atmosphere containing 5% CO, and 95% air. Cells were subcultured by
removing the medium from the cells, washing with phosphate buffer saline,
dislodging them by brief exposure to 0.25% trypsin, and diluting them
appropriately in the designated medium. Cell counts were performed

electronically on a Coulter counter (Coulter Electronics, Inc., Hialeah,

68

 

 

TABLE 1. Human cells used in fusion experiments
Cell line Cell type - origin Source
or strain
WI-38 Finite life span diploid fibroblast ATCC'
LGI Finite life span, foreskin-derived --2
diploid fibroblasts
HSU-1.1 Infinite life span derivative of --2
LGl cells
MSU-1.1-CHATng'3 Clonally-derived from MSU-1.1 This study
VA13 SV40-importalized lung fibroblasts R. A. Schultz‘
T24 (EJ) Bladder carcinoma-derived cell line ATCC1
HT1080 Fibrosarcoma-derived cell line ATCC1
HT1080-CHATsOua'5 Clonally-derived from HT1080 Pereira-Smith

A1698
A1698-CHAT’Oua'S

Bladder carcinoma-derived cell line

Clonally derived from A1698

Pereira-Smith

Pereira-Smith

 

143BTK' Clonally derived from an osteo- NIGMS‘
sarcoma-derived cell line
HeLa Cervical carcinoma-derived cell line Ward Peterson7
CT-l Infinite life span cell derived from Namba
A0837 fibroblast ceiis8
'ATCC: American Type Culture Collection (Rockville, MD).

2See Introduction.

SCHAT’Hg': Cells lacking a functional hypoxanthine(guanine)phosphoribosyl-

transferase (HPRT) gene and, therefore, unable to survive in medium

containing 2 x 10" M deoxycytidine, 10" M hypoxanthine, 4 x 10'7 M

69
TABLE 1 (cont’d)

aminopterin, and 3 X 10‘5 M thymidine (CHAT) and resistant to 125
units/ml hygromycin B (Hg) (Calbiochem).

‘John Sealy Center for Molecular Science, University of Texas, Galveston,
TX

S'CHAT’Oua': HPRT’ cells unable to survive in medium containing 2 X 10'5 M
deoxycytidine, 10“ M hypoxanthine, 4 X 10‘7 H aminopterin, and
3 X 10'5 M thymidine (CHAT) and resistant to 10" M ouabain (Oua)
octahydrate (Sigma).

‘NIGHS: National Institute of General Medical Sciences Human Genetic
Mutant Cell Bank (Camden, NJ).

7Children’s Hospital of Michigan (Detroit, MI).

8Studies by McCormick and Maher (1988) showed that CT-I cells are identical
to SUSM-l cells and are very probably cells derived from A0837 cells, not
from WI-38 cells as originally reported by Namba et al. (1981).

70
FL). Frozen cultures were stored at -135°C in medium containing 15% serum

and 10% dimethyl sulfoxide.

Transfection
The plasmid pHT containing the bacterial hygromycin resistance gene was
provided by Dr. Norman Drinkwater (Univeristy of Wisconsin, Madison, WI).
MSU-1.1 cells were transfected using the dimethylsulfoxide/polybrene
method adapted for use with human fibroblasts (Morgan et al., 1986).
Cells were plated at 2 X 105 cells per 100 mm diameter dish and transfected
18 h later with 1 pg/ml plasmid DNA. Hygromycin (Calbiochem, LaJolla,

CA), at a concentration of 125 units/ml, was added to the medium 24 h
post-transfection. The cells were refed with selective media weekly, and

macroscopic drug-resistant colonies formed in 2-3 weeks.

Selective Iedia

For every cell fusion, one parent cell was a universal fuser, that is,
carried a dominant drug (hygromycin or ouabain) resistance marker, but
lacked a functional dominant gene coding for the enzyme hypoxanthine
(guanine) phosphoribosyl transferase (HPRT), and, therefore, was unable to
survive in CHAT medium. The other parent cell lacked the drug resistance
marker, but could survive in CHAT medium because it carried a functional
HPRT gene. Three universal fuser strains were used: MSU-1.1-CHAT‘Hg’,
HT1080-CHAT'0ua', and A1698-CHAT'0ua'.

For fusions with MSU-1.1-CHAT'Hgr cells and for selection, expansion,
and life span analysis of the hybrids, McM medium (with the modification
listed above) was used. Selection and expansion of hybrids took place in

complete medium to which were added CHAT and hygromycin. For fusions with

71
HT1080-CHAT'0ua' cells or A1698-CHAT'Oua' cells, and for selection,
expansion, and life span analysis of the hybrids, Eagles HEM (with the
modification listed above) was used. Selection and expansion of hybrids
took place in complete medium supplemented with CHAT and ouabain. When
CHAT selectioniwas discontinued, CHT (CHAT minus aminopterin) was added to

the complete medium for 1-2 weeks.

Cell fusion, selection, and initial expansion

Cell fusion and selection were performed according to the procedures
described by Pereira-Smith and Smith (1981, 1982) with the modifications
noted below. Cells (1 X 10‘) of each of two parental lines were seeded in
complete medium into a 35-mm diameter tissue culture dish. Eighteen to 24
h later the medium was replaced by medium lacking serum and containing 45%
polyethylene glycol (1,000) (Sigma, St. Louis, MO). After another 24 h,
the cells were harvested and reseeded into 100 mm diameter tissue culture
dishes at 103 and 10‘ cells per dish in selective medium. The cells were
incubated for 2-3 weeks, with weekly renewal of the selective medium. At
the end of the incubation period, the dishes were scanned to locate hybrid
clones. Each clone was transferred into a 35-mm diameter tissue culture
dish. When the cells had filled the dish, they were transferred to a 25
cm2 flask and then to a 75 cm? flask to expand the clonal population to
approximately 4 X 10‘ cells (i.e., about 22 population doublings ([pdls]).
During this expansion, selective medium was renewed twice weekly. At this
point, a portion of the progeny cells from each hybrid were seeded into

two 25 cm? flasks, and the remaining cells were stored frozen.

72
Deteraination of the life span of the hybrids

To determine the life span of the cells in a population derived from a
hybrid clone, the progeny cells were serially subcultured in complete
medium in two 25 cm? tissue culture flasks, with medium changes twice
weekly. Each time the cells reached confluence, they were subcultured at
a 1:4, 1:8, 1:16, or 1:32 cell dilution: the higher dilutions were used
for the cells that exhibited shorter doubling times. At the time of
subculture, the cells were recorded as having completed 2, 3, 4, or 5
population doublings, respectively, since the last subculture. Following
the criteria of Pereira-Smith et al. (1990), a hybrid was considered to be
at the end of its life span if the cells did not achieve at least one
population doubling within 3 weeks. Following the criteria of Pereira-
Smith and Smith (1988), a hybrid was considered to exhibit an infinite
life span if it achieved a total of 100 population doublings after fusion.

DNA flow cytoaetry
To confirm that isolated clones were hybrids, the DNA content of
parental cells and isolated hybrid clones was determined. Propidium
iodide stained cells were analyzed for DNA content using an Ortho
Diagnostics 50H Cytofluorograph with Acqcyte software supplied by Phoenix
Flow, Inc. (San Diego, CA).

73
RESULTS

Life span analysis of hybrids formed by fusions with MSU-1.1 cell
strains

To allow selection of hybrids between MSU-1.1 cells and various other
cell strains, we developed a clonal strain of MSU-1.1 which carries both
a dominant and a recessive drug resistant marker. First, MSU-1.1 cells
were transfected with plasmid pHT (which carries a bacterial gene for
hygromycin resistance), and a hygromycin resistant clone was isolated.
Secondly, this hygromycin resistant clone was expanded and treated with 6-
thioguanine to isolate an HPRT' CHAT sensitive subclone. This doubly-
marked (Hg' CHAT‘) subclone of MSU-1.1 was used in fusions with various
other human cell lines, as listed in Table 2.

To determine whether the infinite life span phenotype of the MSU-1.1
cell strain resulted from a dominant or a recessive genetic change, we
fused MSU-1.1 cells to the finite life span, diploid fibroblast line LGl
from which they were derived, and also to another finite life span,
diploid human fibroblast line, WI-38. As a control, we fused MSU-1.1
cells that carried both markers to MSU-1.1 cells lacking the markers (Hg‘
CHAT'). Representative hybrid clones (i.e., cells able to grow in medium
that selected against both sets of parental cells) were isolated,
expanded, and assayed for the length of their life span, as described in
Materials and Methods. In addition, LGI parental cells were plated at
cloning density in non-selective medium, and eight clones were isolated
and assayed for the length of their life span in that medium.

All of the eight LG1 clonal strains ceased division between 7 and 27
population doublings following cloning. Similarly, all five hybrids

74

TABLE 2. Proliferative potential of hybrids formed by fusion of
various cell strains with MSU-1.1 CHAT‘Hgr cells

 

 

Immortality Hybrids that
Finite complementation senesced per total

Cell strain life span group' hybrid clones
LGI + NA2 5/53
W1-38 + NA 4/43
MSU-1.1 - --‘ 0/6
VA13 - A 0/1
T24 (EJ) - A 0/2
HT1080 - A 0/2
HeLa - B 0/4

1438TK - C 0/4

CT-1 - n 0/1

 

‘Specified by Pereira-Smith and Smith (1988).

2NA, Not applicable.

’The progeny of these nine hybrid clones senesced at population
doubling 7-16 after fusion.

‘To be determined in these fusion experiments.

75

between MSU-1.1 and LGI cells and all four hybrids between MSU-1.1 cells
and WI-38 cells exhibited a finite life span, achieving only 7 to 16
population doublings before senescing. However, as expected, none of the
six MSU-1.1 self hybrids senesced, demonstrating that cell hybridization
itself does not cause cells to cease proliferation. These results
indicate that the infinite life span of MSU-1.1 cells resulted from a
recessive genetic change.

To determine to which of the four established immortality
complementation groups of Pereira-Smith and Smith (1988) theiMSU-1.1 cells
belonged, we fused them with representative cell strains from each of the
four groups and assayed the life span in culture of the resulting hybrid
clonal pepulations. The inability of MSU-1.1 cells to complement a
particular immortal cell line (i.e., to yield hybrids with finite life
spans) would indicate 'that. MSU-1.1 cells cannot supply ‘the genetic
information that is lacking in that particular cell line. In this case
MSU-1.1 cells would be assigned to the complementation group of that cell
line. We succeeded in obtaining hybrid clones from fusions with cell
lines from all four groups. Very unexpectedly, all 14 hybrids
proliferated to greater than 100 pdls, a life span that Pereira-Smith and
Smith (1988) use as their criterion for inmortality. These results
indicate that MSU-1.1 cells were not able to complement any of these
immortal cell lines (i.e., could not supply the missing genetic function
of any line). In other words, the MSU-1.1 cell strain would have to be

categorized as belonging to all four complementation groups.

76
Deteraining the ability of representatives of the established
coapleaentation groups to couple-ant each other

Before concluding that MSU-1.1 cells.were unable to complement immortal
cells from all four groups, we needed to confirm that fusions between
cells from different complementation groups would yield hybrids with
finite life spans. To do this, we used cell lines supplied by Dr.
Pereira-Smith to carry out some of the same fusions reported by Pereira-
Smith and Smith (1988) as yielding finite life span hybrids. Seven cell
fusions were carried out: two self-fusions (i.e., A1698 cells X A1698
cells and HT1080 cells X HTl080 cells) and five fusions between cell
strains reported to belong to separate>complementation groups (e.g., cells
from group A X cells from group B, A X C, etc). Table 3 lists our
resultsn 1n the experiments of Pereira-Smith and Smith (1988), self-
fusions yielded imnortal hybrids, whereas fusions between cells from
different groups yielded hybrids that senesced by population doubling
(pdl) 65 or earlier. In our experiments, self-fusions yielded immortal
hybrids as expected, but 38 out of 39 of the hybrids that were expected to
senesce failed to do so. Instead, such hybrids proliferated rapidly to
greater than 100 pdls.

When we did not observe the expected senescence in this set of hybrids,
we first considered the possibility that our culture conditions fostered
longer life spans of finite life span cell strains than expected and that
our hybrids would eventually senesce if grown for a longer time.
Therefore, we continued to passage nine hybrid clones beyond pdl 100. All
nine continued to replicate vigorously (data not shown) and were stored

frozen in liquid nitrogen after a total of 146-156 pdls.

77
Testing the hypothesis that parental cells survived selection

We next considered that, perhaps, our selection conditions were not
sufficiently rigorous and had, therefore, allowed infinite life span
parental cells to survive and overgrow any senescing cells in our life
span assay. Our selection system differed from that of Pereira-Smith and
Smith (1988) in that after progeny cells from a hybrid clone had been
expanded in selective medium from a small colony to more than 4 X 10‘ cells
(i.e., had undergone a 22 pdl), we discontinued drug selection. Pereira-
Smith and Smith maintained drug selection throughout the entire life span
assay. To verify that the cells growing vigorously in our immortal
populations were progeny of hybrid cells, we used flow cytometric analysis
to compare the amount of DNA per cell of representative hybrids with that
of their respective parental cells (data not shown). For 16 of the 17
hybrids analyzed at population doublings 25-30 post-fusion, the amount of
DNA per cell was virtually equal to the sum of that of the two parental
cell lines. The remaining hybrid had somewhat less DNA per cell than the
total amount from the two parental strains, but the amount of DNA per cell
was significantly higher than that of either parent. Six of these hybrids
were also analyzed at pdls greater than 100 post-fusion. Two of the six
had lost some DNA during this extensive passaging, but the amount of DNA
per cell for each of the six was still higher than that of either parental
cell, indicating that they were hybrid cells.

To obtain further evidence that our immortal populations were the
progeny of hybrid cells, rather than parental cells, we tested six of them
at population doublings greater than 100 for the ability to grow in the
selective medium. Five of the six doubled in selective medium at the same

rate as they did in parallel flasks in the absence of selective pressure.

78

TABLE 3. Proliferative potential of cell hybrids formed by fusions among

cells from the four immortality complementation groups

 

 

Hybrids Hybrids that
expected senesced per total
Cells fused' to senesce hybrid clones
HT1080 (A) x HT1080 (A) No 0/6
A1698 (D) X A1698 (0) No 0/5
HT1080 (A) X HeLa (8) Yes 1/13
HT1080 (A) X 1438TK' (C) Yes 0/15
A1698 (0) x HT1080 (A) Yes 0/5
A1698 (0) X HeLa (8) Yes 0/5
A1698 (0) x 1438TK" (C) Yes on

 

'For each cross listed, the parent cell line on the left was CHATs and
Oua'. The letters in parentheses indicate the complementation group to

which the cell line has been assigned by Pereira-Smith and Smith (1988).

 

79
The sixth doubled somewhat more slowly in selective medium than in its
absence, but the cells appeared perfectly healthy, indicating that they
were resistant to the selective agents and, therefore, were hybrid cells.
Therefore, we ruled out the explanation that parental cells had survived

selection.

Testing the hypothesis that our infinite life span populations
represented overgrowth of senescing populations by revertants

A third possible explanation for the infinite life spans of 38 of the
39 hybrids from putatively complementary parents is that the original
hybrid cells possessed a full complement of the genetic material that
confers a finite life span, but progeny cells subsequently lost some or
all of the critical genetic material and, by doing so, regained an
infinite life span. If a strain were to begin to senesce before such an
immortal progeny cell (revertant) has arisen, this would be reflected by
a temporary plateau in the growth curve. Such plateaus were reported by
Pereira-Smith et al. (1990) for hybrids of HeLa cells fused with human
diploid fibroblasts. We examined the growth curves of our hybrids (Fig.
1 and 2) for plateaus. Only one of our 38 immortal hybrids derived from
putatively complementing parents exhibited such a temporary decrease in
growth rate (Fig. 1C, closed diamonds). The decrease began at pdl 22, and
the cells returned to their previous rapid rate of growth at pdl 30.
However, we also noted that the rate of growth of several slow-growing
hybrids (e.g., Fig. 1C, closed triangles) increased at approximately pdl
22, the time at which drug selection was discontinued. We had originally
decided to discontinue drug selection at the end of the initial expansion

phase (i.e., when the clonal population had filled the flask [pdl 22])

 

80

Figure 1. Growth curves of hybrids formed by fusion of HT1080-CHAT‘Ouar
cells with cells representative of immortality complementation groups A,

B, and C

Drug selection was discontinued at pdl 22, as indicated by the arrows.
Where growth curves of several clones fall upon one another (e.g., panel
A, clones A-F), a single representative curve is depicted, and the number

of clones in the group is indicated (e.g., Panel A’, 6 clones).

 

 

 

 

 

 

 

 

81

 

 

 

 

 

 

 

 

 

. AK
A HT1080(Gr.A) x HT1080(GI’.A) HT1080(Gt.A) x HTIOBO(Gt.A)
iOO "' 100 '
m b as
g I g .
'5 3
3 b 3
O O
D D
c -0- Clone A C
.2 .9.
3 5° " '0- Clone B E 50 b
3 _ a
g + Clone C no.
Q ' -0- Clone D ’
—-)l -a- C.0.,. E '_)» "0- 6 Clones
. -A- Clone F >
o A L l A A A A l A L A 1 o L A A A l A A A A l A A A A l
O 50 100 150 O 50 100 150
Days Days
8. C.
"TI 08°(GI-A) x HoLa(Gr.B) HT1080(Gr.A) x 1438TK-(GLO)
ioo - 100 . . 1’;
in as /
O O
.5 5 .
1% 1% r ,f/ ’
O
c: :3 ’ f
e e (I
O 0
T3 5° 5 ":3 50 ‘0’ + B Clones
3 + 3 Clones 3 I
a. Q ~ (I / -0- 4Clones
0 -°- 5 Clones o ,
o. _ n. ’ [.0 f -l- 2 Clones
+ 3 Clones , A
). ‘ -—) -O- 1 Clone
/ + 1 Clone / -O- 1 Clone
.1. 1 Clone / -A- 1 Clone
o A A l A A A A 1 A A A A L o A l A A A A L A A A A l
O 50 100 150 O 50 100 150
Days Days
Figure 1. Growth curves of hybrids formed by fusion of HT1080-CHAT‘Oua'

cells with cells representative of immortality complementation groups A,

B, and C

82

Figure 2. Growth curves of hybrids formed by fusion of A1698-CHAT‘Oua'
cells with cells representative of immortality complementation groups A,

8, C, and 0

Drug selection was discontinued at pdl 22, as indicated by the arrows.

HAl‘Oue'

'0ups l.

TONS.

83

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A‘ A1698(Gt.D) x HT1080(Gr.A) 8' A1698(Gr.D) x HoLa(Gr.B)
TOO " / 100 .- f
as ' a ' ’
a a .
.5 » .s . /
g l- g ’ X
o o
a x o /
s ' r s I
'3 i- z '- + |
g 50 [A + Clone A g 50 C one A
:i :i
3' / _°_ Glam 8 3 . -°- Clone B
n' , ’f _0— 6.0". c a' , + Clone C
—,. + (non. D —)> .0. CID“. D
-a. man. 5 . --4- Clone E
o AAAAAAAAAAAAAA I o - _ A A l - A A A _1 ‘ A A ‘ l
O 50 100 150 0 50 100 150
Days Days
C- 0' A1698
A1698(Gr.D) x 1438TK-(Gt.C) (Gr-D) x A1698(Gr.0)
100 L 100 L I
a as ' .
3 . a . I
3 3
a . a
o o
D 0
c c i
.2 .2
E 50 " E 50 '-
3 _ a -0- Clone A
0 o A
a, b IL > [A -°- Clone B.
)
.__). . Clone A __)' /‘ + Clone C
-0- Clone D
-A- Clone E
o A A A A l A A A A L A A A A l o A A _ a L A A A A l A A A A g
0 50 100 150 O 50 100 150
Days Days

r

Figure 2. Growth curves of hybrids formed by fusion of A1698-CHAT'Oua

cells with cells representative of inmortality complementation groups A,

8, C, and D

84
because‘we noted that during expansion of the hybrids in selective medium,
the cells, although resistant to both drugs, gradually became granular in
appearance. Some populations doubled more slowly than the others, and the
cells in these populations tended to detach from the flask. When our
hybrid cells were no longer under selection, they lost their granularity.
The slow-growing hybrids became more firmly attached to the flask and

began to proliferate as fast as the others.

Testing the effects of long-tern drug selection on senescence

Since the hybrid cells appeared much healthier after discontinuation of
drug selection, we hypothesized that a longer period of drug selection can
result in cell death. If so, the difference between our drug selection
regimen and that of Pereira—Smith and Smith (1988) could account, at least
in part, for the difference between our results and theirs. To test
whether continuous propagation of hybrids in selective medium can result
in cessation of growth, we thawed cells from 22 of our hybrid clones that
had been frozen at pdl 22-26 and repeated the life span assay, passaging
each hybrid strain in parallel flasks in the presence and absence of drug
selection. Growth curves of representative hybrids are shown in Figure 3.
llhen a cell population was grown under non-selective conditions, its
growth curve closely matched that which we had obtained previously under
these conditions (Fig. 3, open circles). However, when cells were
propagated under continued drug selection (Fig. 3, closed circles), toxic
effects were observed. Eight of the 22 hybrids continued to proliferate
at a steady rate, but slightly slower than the parallel cultures grown in
the absence of drug selection (e.g., Fig. 3A). The other 14 hybrids, when

kept under selective conditions, entered a period of rapid cell death

 

85

Figure 3. Growth curves of hybrids formed by fusion of lillOBO-CHAT'Oua‘r

cells with HeLa cells

Closed symbols, continuous drug selection; open symbols, drug selection
discontinued at pdl 22. Arrows indicate the beginning and end of the

period of cell crisis.

86

 

7o
Clone A
eo -
so -
4o -

30'

20'

Population Doubllngs

 

 

 

O 20 4O 60 80 100 120 140

 

70

’ Clone B
so .

40’

30'

Population Doubllngs

10'

 

 

 

0““““““
O 20 40 60 80 100120140

 

7o
Clone C
so -
so »
40 -
30 T T

20‘

Population Doublings

10"

 

 

 

o 4 A l A I 4 A A A A A A

0 2o 40 so no mo 120 140
Days

Figure 3. Growth curves of hybrids formed by fusion of HT1080-CHAT‘0ua'
cells with HeLa cells

87

between pdl 22 and 40; that, is the cells became increasingly more
granular and then began detaching and disintegrating. Destruction of the
majority of the cells occurred over a period of 1-4 weeks. In some
populations this cell destruction was very rapid, and only a few cells
survived. These cells formed individual clones which, when dispersed,
slowly repopulated the flask over a period of 3 to 4 weeks. From the time
the cells began dying until the flask was repopulated, cell division was
occurring, but by definition, the cell population as a whole did not
double, so the growth curve was flat (e.g., the period in Figure 38
delineated by arrows). Once the flask was repopulated, the population
doubled in selective medium at the same rate as did the corresponding
population grown in the absence of drug selection. In other hybrid
populations cell destruction was slower. In these, a high rate of cell
death competed with cell division over a period of two to four weeks,
resulting in a slow rate of population doubling during the period of cell
destruction (e.g., the period in Figure 3C delineated by arrows).

We interpret the growth patterns illustrated in Figure 38,6 as
reflecting cell death caused by toxicity from long term exposure to CHAT
and ouabain, with subsequent repopulation of the flask by surviving
variant cells that have greater resistance to the drug. The eight hybrids
which did not exhibit this rapid cell death apparently'werehmore resistant
to the selection than were the other 14 tested. Table 4 summarizes our
results and compares them with those of Pereira-Smith and Smith (1988).
The incidence of rapid cell death that we observed with continuous drug
selection closely matches the incidence of "senescence" reported by
Pereira-Smith and Smith (1988) for clones assayed under the same

conditions (column 4 of Table 4). In contrast, when we discontinued drug

88

.Ammmfiv no.5m ecu cavemlucwmgma an umpcogma mupammmn
.auaum “comma; on» soc» mapsmmm~
.cmamn was» we mum mmpnoh cw umuucmrmmu mew amaze mmcrp ppmu on» .AmmmHV cupsm new zuvsm

lucrmsmm An xuaum one go» "N mpaup c? czogm mmco on» men vows» mcwmcum Ppmo on» .acaum acmmmgn on» com.

 

 

Asav mofi\mm Aoofiv H\H AoV H\o u x a
Aoofiv m~\m~ Aoov m\m on m\o m x a
Acesv HM\HM AooHv ~\~ on m\o < x a
Aoofiv -\N~ Acosv ~\~ AOV mH\o u x 4
Away mm\mm AmaV m\o Amv mH\H m x <
on vH\o on ~\o on m\o a x a
AOV mm\o on ~\o on o\o < x <

nAxv mmcopu ~Axv cmxmmmu maven»; ~Axv mmcopu .ummaw mppmu
urea»; Pepe» Pepe“ Log canon Fpmu upgn»; Page» mo museum
Lon cmummcmm urges umurnrcxm con umummcmm coruuucoEmpasou
pass maven»: sags maven»: sags muses»: umsaoamm

c CCF 2 c z c F P a; F c ?

 

mppmu even»; go cunm mwwp on» :0 coruumpmm mace Ecmu-mcop mamgm> ecmulusocm no muummmu .e m4m<p

89
selection at pdl 22, 98% of our hybrids showed no senescence or evidence
of rapid cell death (column 2 of Table 4). We interpret our results to
indicate that immortal cell lines, when fused, cannot complement each
other to yield hybrids with finite life spans. In addition, our results
strongly suggest that the cell death observed in hybrids from these cell
strains by Pereira-Smith and Smith resulted from the toxic effect of long

term drug selection.

DISCUSSION

Our results showing that fusion of infinite life span HSU-1.1 cells
with their own finite life span parental LGl cells, as well as with the
finite life span fibroblast cell line HI-38, yields finite life span
hybrid cells confirmlwhat has been reported by many investigators, namely
that immortalization of human cells is a recessive trait (i.e., results
from loss of the function of one or more of the genes responsible for
senescence). However, the results of our fusions of HSU-1.1 cells with
various other immortal cell lines failed to support the claim of Pereira-
Smith and Smith (1983, 1988) that immortal cells can complement one
another (i.e., supply the'missing function(s) to yield hybrids with finite
life spans). Hhen HSU-1.1 cells were fused to representative cell lines
from each of the four complementation groups established by these
investigators, none of the resulting hybrids senesced. Therefore, one
would have to assign MSU-1.1 cells to all four of the complementation

groups, rather than just one. This implies that the MSU-1.1 cells lack

 

90

each of the sets of senescence genes represented by the four groups.
Shortly after we obtained this result, another research group using the
same approach was able to assign the SV40-immortalized human cell line
BET-1 to a single complementation group, namely group D (Whitaker et al.,
1992). However, when they later applied the method to three other SV40-
immortalized human cell lines, each of these cell lines were found to
belong to more than one complementation group (Duncan et al., 1993).
These results, along with our own, indicate that complementation among
immortal cells is a more complicated phenomenon than the studies of
Pereira-Smith and Smith suggest. This conclusion is strengthened by the
results of our fusions of cells reported by Pereira-Smith and Smith (1988)
to be from different immortality complementation groups. Instead of
obtaining hybrid cells that senescenced, 38 out of 39 of our hybrids from
putatively complementary parents exhibited unlimited life spans. Flow
cytometry analysis of the immortal cells from these fusions verified that
they were hybrids.

Pereira-Smith and Smith proposed two two theories to explain why finite
life span hybrids might appear to have infinite life spans. In one, they
propose that very soon after fusion the tetraploid cell loses the gene or
genes that limit proliferation (rapid chromosomal segregation)(Pereira-
Smith and Smith, 1988; Pereira-Smith et al., 1990). Even if the combined
genomes of two parental cells confer on a hybrid cell the full complement
of the genes limiting cellular life span, immediate loss of the critical
genes would render the progeny of the hybrid immortal. Such immortal
hybrids comprised only 4-5% of the hybrids of complementary immortal cell
lines in the 1988 study of Pereira-Smith and Smith. Even among the
hybrids from HeLa D98 fused with human diploid fibroblasts, which Pereira-

91

Smith et al. (1990) reported have a high rate of reversion to immortality,
only 13% (5/39) failed to show at least some evidence of senescence (i.e.,
did not exhibit any slowing of the rate of population doubling). In
contrast, 95% of our hybrids (38/39) failed to show any evidence of
senescence. It seems highly unlikely that such a high proportion of our
hybrids should have undergone “rapid chromosomal segregation" and acquired
an infinite life very soon after fusion.

The second theory proposes that the loss of critical genes occurs in a
progeny cell of the expanded finite life span clonal population after the
rate of population doubling has slowed rather than soon after fusion
(Pereira-Smith and Smith, 1983; Pereira-Smith et al., 1990). The
resulting immortal cell proliferates and overgrows the culture as the
senescing population ceases cell division. One might interpret the slight
increase in growth rate which occurred for several of our immortal hybrid
clones when selection was discontinued (Fig. 1 and 2) as evidence that
this phenomenon*was occurring. However, we consider this highly unlikely,
since even the immortal control hybrids we obtained from self fusions of
immortal cell lines (e.g., A1698 cells X A1698 cells) exhibited such an
increase in growth rate when drug selection was discontinued (see Fig.
20). This provides evidence that the increase in growth rate represents
the cells’ recovery from the toxic effects of the drugs. For the hybrid
population shown in Fig. 1C, we interpret the temporary reduction in
growth rate (closed diamonds) following discontinuation of drug selection
as delayed recovery from the toxic effects of the drugs. In addition, it
could be that the clonal population that ceased dividing at population
doubling 22 (Fig. 18, closed triangles) was too damaged to recover.

In our study where hybrids were deliberately maintained on drug

92

selection throughout their life span assay, the incidence of rapid cell
death closely matched the incidence of senescence reported by Pereira-
Smith and Smith (1988)(Table 4). ‘The cell destruction1we observed was too
rapid to be attributed to what is normally referred to as senescence. In
populations that were rapidly dividing in selective medium, a majority of
the cells suddenly detached and disintegrated. The abruptness of the
cessation of growth and the rapidity of the onset of disintegration of the
cells suggest that instead of senescence we were observing necrosis or
apoptosis triggered by the cumulative cytotoxic effect of the drugs.
According to the description by Hyllie (1981) and Hyllie et al. (1980),
apoptosis involves nuclear condensation, membrane blebbing, and
cytoplasmic condensation, and it ends in cell fragmentation, with death of
an entire population of cells occurring over a period of several days. In
the case of necrosis, the dominant morphologic event is cellular swelling
from loss of control of cell volume, followed by rupture of the plasma
membrane, again with rapid destruction of whole populations of cells
(Nyllie, 1981; Hyllie et al., 1980). Both processes can be initiated by
a variety of cytotoxic drugs.

In contrast to apoptosis and necrosis, cell senescence is not a rapid
process. The population dynamics of normal finite life span cells include
first, a period of rapid proliferation; secondly, a gradual decrease in
proliferative rate over a period of many weeks, accompanied by
deteriorative changes in the cells; and finally, complete cessation of
growth (Hayflick and Moorhead, 1961; Hayflick, 1965; Bayreuther et al.,
1988). In an extensive review' on cellular senescence citing 575
references, Stanulis-Praeger (1987) states " 'True’ senescence ‘thus

comprises the prolonged period of proliferation sterility preceding the

93

death of the cells." The non-dividing senescent cells enlarge and
deteriorate but remain attached to the dish for weeks or months before
disintegrating or detaching. He observed just such a prolonged period of
proliferation sterility during our analysis of the life span of normal
fibroblastic LGl clones and of MSU-1.1 X LGl hybrid clones. However, the
cell death and detachment which occurred when the hybrids from our
immortal cell X immortal cell fusions were kept under continuous drug
selection was not preceded by this cessation of cell division.

It is important to know'whether the limited life spans exhibited by the
progeny of the hybrid clones reported by Pereira-Smith and Smith (1988)
resulted from cellular senescence or from drug cytotoxicity. However, to
determine this, one would need to know the appearance and behavior of
their cells as division ceased and the cells died. In their 1988
publication identifying the four immortality complementation groups,
Pereira-Smith and Smith do not provide the details of the cessation of
cell division by the progeny of cell hybrids formed by the fusion of
immortal cells with immortal cells. Instead, they refer readers to their
1981, 1982, and 1983 publications for the details of their culturing of
hybrid cells and their method of determining the proliferative potential
of the hybrids. Their 1981 article provides a detailed description of
hybrids derived from the fusion of immortal VA13 or GM 639 cells with
finite life span normal human diploid fibroblasts. They report that the
clones went through two identifiable phases of growth: a period of active
division, followed by a period of complete cessation of division. They
stated that the period of complete cessation of cell division
. .resembled the ’crisis’ phenomenon observed during viral transformation

of normal cells in vitro. During this period, a large number of cells

94

lysed. Micronucleation, nuclear lobulation, and nuclear blebbing were
observed in the cells." This growth pattern is not consistent with the
gradual decrease in growth rate of typical finite life span cell
populations. Sudden lysis of a large number of cells is not
characteristic of senescing cells. Instead, the sudden cessation of
division and the cell destruction described arelmore characteristic of the
necrosis or apoptosis process described above. Stein (1985) reports that
the cell crisis exhibited by SV40-transformed cells results from increased
cell death among cells that are still traversing the cell cycle because
they have lost their ability to enter the "normal G-1 arrested senescent
state". We have been unable to determine from the papers by Pereira-Smith
and Smith whether or not an abrupt crisis phenomenon was a common feature
of the division cessation of the majority of their hybrids.

Clearly, careful distinction among different types of cell death is
necessary for interpreting the results of studies on cellular immortality.
The ambiguities discussed above highlight the importance of finding
definiteland easily identifiable specific markers for cellular senescence,
necrosis, and apoptosis. Several markers for senescence have already been
proposed, but none of these provide a positive, quantitative marker that
is characteristic only of senescing cells as distinct from non-growing
quiescent cells. Necrosis can be detected by abnormal permeability to
certain molecular markers (Hoffstein et al., 1975). Apoptosis is
recognized by apoptotic body formation and by fragmentation of chromatin
into characteristic nucleosome ladders (Hyllie et al., 1980). Resolution
of the conflicting results obtained with hybrids from fusions of immortal
cells with immortal cells will require further research employing careful

microscopic and biochemical techniques to distinguish among these

95
phenomena.

If, as our data suggest, immortality is recessive in hybrids obtained
from fusion of finite life span cells with immortal cells, but immortal
cells cannot complement each other to yield hybrids with finite life
spans, what does this say about the genetic alterations responsible for
the immortal phenotype? One possible explanation is that all immortal
cells share a common defect in the gene(s) that control(s) life span.
This is highly unlikely since at least two different human chromosomes,
i.e., 1 and 4, have been implicated in cellular senescence (Ning et al.,
1991; Suguwara et al., 1990). It is more likely that several genes
control life span, and, therefore, defects in any'one of several different
genes result in an infinite life span. In this case, a possible
explanation for the failure of cells to complement each other comes from
the work of Harris (1988). He proposed a theory based on genetically
imprinted patterns of differentiation to explain the lack of
complementation among malignant cells. If one applies his theory to the
infinite life span phenotype, the essential genetic lesion determining
immortality would be the loss of the ability to complete a program of
terminal differentiation. According to that view, fusions between finite
life span cells and immortal cells yield hybrids with finite life spans
because the differentiation program of the finite life span cell can be
executed in the hybrid. The lack of complementation among immortal cell
strains could be attributed to the fact that a cell comitted to one
program of differentiation cannot simultaneously execute another, and that
the genetic imprints that determine major lineage of differentiation are
stably maintained in cell hybrids. Accordingly, Cell A, committed to
differentiation program A, but exhibiting immortality because of its

96

inability to complete that program, could not be expected to complement
the defect in an immortal cell 8 from another lineage because the genes
carried in cell A for differentiation program 8 have been inactivated by
stable imprinting. For the same reason, cell 8 cannot complement cell A.

This explanation could account for the inability of different types
of cells, (i.e., cells that are derived from independent differentiation
programs) to complement each other, e.g., HT1080, from.a fibrosarcoma, and
HeLa, from a cervical carcinoma. However, when cells of the same tissue
type, e.g., both fibroblastic in origin, fail to complement each other, as
we observed with MSU-1.1 cells and HT1080 cells, one would have to
conclude that the cells share a common defect in their program of
differentiation.

Even without knowledge of the underlying mechanisms involved, our data
support the hypothesis that the phenotype of infinite life span in human
cells results from recessive changes in a genetic program for cellular
senescence, but they suggest that immortal human cells cannot complement
one another to yield hybrid cells with finite life spans. Our data also
indicate that long-term dual drug selection is deleterious to hybrid
cells, even though the cells carry resistance markers for both drugs.
Therefore, investigation of senescence using cell fusion demands careful
discrimination among different types of cell death. The lack of
complementation that we observed in hybrids between immortal cells is not
easily explained by simple Mendelian genetics. More complex genetic
mechanisms may need to be invoked to explain the phenotypes of these

hybrids.

97
ACKNOWLEDGEMENTS

We thank Dr. 0. Pereira-Smith for graciously sharing her cell lines
with us. The excellent technical assistance of Alicia Holzhausen and John
Neimela is gratefully acknowledged. We thank Louis King of the
Biochemistry Department for his help and advice in analyzing the flow
cytometry data and Connie Williams for typing the manuscript. This
research was supported in part by DOE Grant ER60524 and DHHS Grant A611026
from the National Institute of Aging.

 

LITERATURE CITED

Bayreuther, K., Rodemann, H.P., Hommel, R., Dittmann, K., Albiez, M., and
Francz, P. I. (1988) Human skin fibroblasts in vitro differentiate
along a terminal cell lineage. Proc. Natl. Acad. Sci., U.S.A.,
85:5112-5116.

Bunn, C.L., and Tarrant, G.M. (1980) Limited lifespan in somatic cell
hybrids and cybrids. Exp. Cell Res., 127:385-396.

Duncan, E.L., Whitaker, N.J., Moy, E.L., and Reddel, R.R. (1993)
Assignment of SV40-immortalized cells to more than one
complementation group for imnortalization. Exp. Cell Res., 205:337-
344.

Hayflick, L., (1965) The limited in vitro lifetime of human diploid cell
strains. Exp. Cell Res., 37:614-636.

Hayflick, L., and Moorhead, P.S. (1961) The serial cultivation of human
diploid cell strains. Exp. Cell Res., 25:585-621.

Hoffstein, S., Gennaro, D.E., Fox, A.C., Hirsch, J., Streuli, F., and
Weissman, G. (1975) Colloidal lanthanum as a marker for impaired
plasma membrane permeability in ischemic dog myocardium. Am. J.
Pathol., 79:207-218.

McCormick, J.J., and Maher, V.M. (1988) Towards an understanding of
the malignant transformation of diploid human fibroblasts. Mutat.
Res., 199:273-291.

Morgan, T.L., Maher, V.M., and McCormick, J.J. (1986) Optimal parameters
for the Polybrene-induced DNA transfection of diploid human
fibroblasts. In Vitro Cell. Dev. 8101., 22:317-319.

98

 

99

Morgan, T.L., Yang, 0., Fry, 0.G., Hurlin, P.J., Kohler, S.K., Maher,
V.M., and McCormick, J.J. (1991) Characteristics of an infinite life
span diploid human fibroblast cell strain and a near diploid strain
arising from a clone of cells expressing a transfected v-myc
oncogene. Exp. Cell. Res., 197:125-136.

Muggleton-Harris, A.L., and Desimone, D.W. (1980) Replicative potentials
of various fusion products between WI-38 and SV40 transformed WI-38
cells and their components. Somat. Cell Mol. Genet., 6:689-698.

Namba, M., Nishitani, K., Fukushima, F., Kimoto, T., Utsunomiya, J., and
Hayflick, L. (1981) Neoplastic transformation of human diploid
fibroblasts treated with chemical carcinogens and Co-60 y-Rays. GANN
Monograph on Cancer Res., 27:221-230.

Ning, Y., and Pereira-Smith, 0.M. (1991) Molecular genetic approaches to
the study of cellular senescence. Mutat. Res., 256:303-310.

Ning. Y., Weber, J.L., Killary, A.M., Ledbetter, D.H., Smith, J.R., and
Pereira-Smith, 0.M. (1991) Genetic analysis of indefinite division
in human cells: Evidence for a cell senescence-related gene(s) on
human chromosome 4. Proc. Natl. Acad. Sci. U.S.A., 88:5635-5639.

Pereira-Smith, 0.M., and Smith, J.R. (1981) Expression of SV40 T antigen
in finite life-span hybrids of normal and SV40-transformed
fibroblasts. Somat. Cell Mol. Genet., 7:411—421.

Pereira-Smith, 0.M., and Smith, J.R. (1982) Phenotype of low proliferative
potential is dominant in hybrids of normal human fibroblasts.
Somat. Cell Mol. Genet., 8:731-742.

Pereira-Smith, 0.M., and Smith, J.R. (1983) Evidence for the recessive
nature of cellular immortality. Science, 22l:964-966.

 

100

Pereira-Smith, 0.M., and Smith, J.R. (1988) Genetic analysis of indefinite
division in human cells: Identification of four complementation
groups. Proc. Natl. Acad. Sci. U.S.A., 85:6042-6046.

Pereira-Smith, 0.M., Stein, G.H., Robetorye, S., and Meyer-Demarest, S.
(1990) Immortal phenotype of the HeLa variant D98 is recessive in
hybrids formed with normal human fibroblasts. J. Cell. Physiol.,
143:222-225.

Ryan, P.A., McCormick, J.J., and Maher, V.M. (1987) Modification of MCDB-
110 medium to support prolonged growth and consistent high cloning
efficiency of diploid human fibroblasts. Exp. Cell Res., 172:318-
328.

Sack, G.H., Jr. (1981) Human cell transformation by simian virus 40--A
review. In Vitro 17:1-19.

Stanulis-Praeger, R.M. (1987) Cellular senescence revistied: A review.
In: Mechanisms of Aging and Development. Elsevier Scientific
Publishers, Amsterdam, Vol. 38, pp. 1-48.

Stein, G.H. (1985) SV40-transformed human fibroblasts: Evidence) for
cellular aging in precrisis cells. J. of Cell. Physiol., 125:36-44.

Whitaker, N.J., Kidston, E.L., and Reddel, R.R. (1992) Finite life span of
hybrids formed by fusion of different simian virus 40-immortalized
human cell lines. J. Virol., 66:1202-1206.

Wylie, A. H. (1981) Cell death: A new classification separating apoptosis
from necrosis. In: Cell death in biology and pathology. I.D. Bowen
and R.A. Lockshin, eds. Chapman and Hall, London, New York, pp. 9-
34.

Wylie, A.H., Kerr, J.F.R., and Currie, A.R. (1980) Cell Death: The
significance of apoptosis. Int. Rev. Cytol., 68:251-305.

 

CHAPTER III

Suppression of the Tumorigenicity of ii-, ll-, or v-K-ras Oncogene-
Transfer-ed Malignant Huan Fibroblast Strains by Fusion with their
Infinite Life Span, lion-Tmrigenic Parental Strain, MSlJ-l.1

101

INTRODUCTION

Activated oncogenes of the ras family have been detected in a
variety of rodent and human tumors (reviewed in Barbacid, 1987). The
results of studies in which an activated ras oncogene has been transfected
into rodent or human cells in culture have shown that the expression of an
activated ras oncogene can contribute to the malignant transformation of
mammalian cells, but have indicated that one or more changes in addition
to the expression of an activated ras oncogene are required for a cell to
become malignant.

Malignant transformation by ras of rodent cells in culture requires
either that the cells first be immortalized or, if transformation is to be
accomplished in primary or early passage rodent cells, that a second
oncogene be transfected along with ras (Land et al., 1983; Ruley, 1983;
Newbold and Dverell, 1983; Parada et al., 1984; Oshimura et al., 1985).
The oncogenes, e.g., myc, that have been found to cooperate with ras in
the malignant transformation of rodent cells are ones that are considered
to be involved in causing cellular imnortalization. Similarly, ras-
induced malignant transformation of human fibroblasts in culture is
successful when infinite life span fibroblasts are used as the recipients
of the gene, but not when finite life span fibroblasts are used (Hurlin et
al., 1987; Hurlin et al., 1989; Wilson et al., 1989; Wilson et al., 1990;
Fry et al., 1990). One interpretation of this finding is that, in
addition to expression of the ras oncogene, other genetic alterations must
be acquired for a cell to become malignant, and that a cell must have an
infinite or greatly extended life span' to achieve the number of cell

divisions required for these changes to occur.

102

 

103

Other observations made during these studies support the hypothesis
that ras-induced malignant transformation of infinite life span human
fibroblasts requires changes in addition to the expression of a ras
oncogene. Hurlin et al. (1989) selected ras-expressing cells from
transfected populations of infinite life span fibroblasts by isolating
morphologically transformed colonies of cells (foci) that overgrew the
monolayer of treated fibroblasts. They reported that, although all six of
the clonal isolates expressed higher than normal levels of the activated
ras protein, only four were tumorigenic. This suggests that the four
tumorigenic isolates had acquired an additional change that is required to
produce the malignant state.

Cell fusion studies suggest that the loss of the function of a tumor
suppressor gene is required for ras-induced malignant transformation of
infinite life span cells. Sager and her colleagues used the H-ras
oncogene of the of the human bladder cancer cell line EJ, also known as
T24, to malignantly transform infinite life span Chinese hamster embryo
fibroblast (CHEF) cells (Smith et al., 1982). They later fused the
tumorigenic ras-transformed CHEF cells with the infinite life span, non-
tumorigenic CHEF cells from which the tumorigenic cells were derived
(Craig and Sager, 1985). Although the hybrid cells expressed p21—ras
protein at levels comparable with that of the malignant parental cells,
they had a decreased tumor-forming ability in nude mice. These
researchers concluded that the infinite life span non-tumorigenic CHEF
cells contain a suppressor gene capable of inhibiting the tumorigenic
effect of the activated EJ H-ras oncogene.

The results of other intraspecies rodent cell hybridization studies

also support the hypothesis that normal cells contain genes that suppress

 

104

ras-induced tumorigenicity. Willecke et al. (1987) reported that the EJ
H-ras-induced tumorigenicity of Rat-1 cells can be suppressed by fusion
of these cells with rat embryo fibroblasts. Because the tumorigenic cells
were fused to cells that were at least two steps removed from the
tumorigenic cells (i.e., not ras-transformed and not infinite life span),
and because three of the four hybrids expressed less p-21-ras protein than
did the tumorigenic parental cells, this study is less convincing than
that of Craig and Sager (1985).

Oshimura et al. (1988) showed that tumorigenicity of ras/myc-
transformed Syrian hamster cells can be suppressed by fusion of these
cells with Syrian hamster embryo cells. They reported that the majority
of cell hybrids senesced, as would be expected since infinite life span
has been shown to be a recessive characteristic (Bunn and Tarrant, 1980;
Muggleton-Harris and Desimone, 1980; Pereira-Smith and Smith, 1981, 1983),
but that some hybrids escaped senescence and grew indefinitely. Although
these non-senescent hybrid cells produced high levels of the mutant ras
protein, tumorigenicity was significantly suppressed.

Geiser et al. (1986) studied the suppression of the EJ H~ras in
human cells by fusing EJ carcinoma-derived cells with normal human
fibroblasts. They reported that all six of the hybrid clones that were
isolated proliferated beyond 200 population doublings (see below) and were
non-tumorigenic. After a period of time in culture, several tumorigenic
segregants arose within the non-tumorigenic hybrid populations. The
levels of expression of the mutant H-ras were similar in the EJ cell line,
the non—tumorigenic hybrid cells, and the tumorigenic segregants. Their
results suggest that ras-induced malignant transformation of human cells

requires inactivation of a tumor suppressor gene that is present in normal

 

105
human cells.

As did the rodent cell fusions of Willecke et al. (1987) and
Oshimura et al.(1988), the human cell fusion of Geiser et al. (1986)
spanned at least two steps of transformation, i.e., immortalization (EJ
cells have an infinite life span) and ras~transformation. In light of the
strong evidence from the studies referenced above that infinite life span
is a recessive characteristic, one‘would have expected most of the hybrids
obtained by the fusion of EJ cells with normal fibroblasts to have a
finite life span, which has been shown to be less than 100 population
doublings in normal cells. Although Geiser et al. (1986) did not indicate
any evidence of senescence in their report, it is likely that finite life
span hybrid cells were obtained, but senesced too soon to form colonies
that could be isolated.

In light of the cell fusion studies described above and the report
by Hurlin et al. (1989) that some ras oncogene-expressing MSU-1.1 cells
failed to produce tumors, I hypothesized that the malignant transformation
of MSU-1.1 by an activated ras oncogene requires, in addition to the
expression of the activated ras oncogene, the inactivation of a tumor
suppressor gene that is present in MSU-1.1 cells. I tested this
hypothesis by fusing cells from six independent malignant MSU-1.1-ras
strains with the infinite life span , non-tumorigenic, precursor MSU-1.1
cells and assaying the hybrids for tumorigenicity. Although
tumorigenicity was clearly suppressed in the majority of the hybrid cell
strains, the degreerof suppression varied among the hybrid strains and did
not correlate with the mutant ras protein expression levels in the hybrid
cells. The results of a comparison of the ras protein expression of

tumorigenic cells at various passages suggested that ras expression is

 

 

 

106
increased by growth of the cells in athymic mice and decreased by growth

of the cells in tissue culture dishes.

MATERIALS and METHODS

Cells and cell culture
The cell strains used and their sources are listed in Table 1. The
cells were routinely cultured as previously described (Chapter 11,

Materials and Methods, pp. 66 and 69).

Cell fusion
Cell fusions and the selection and initial expansion of hybrid
clones were performed as previously described (Chapter 11, Materials and

Methods, pp. 69-70).

Assay for tulorigenicity
8AL8/c athymic mice, ages 6 to 9 weeks, were injected subcutaneously
with 10‘ to 107 exponentially growing cells per site. Tumor growth was

monitored on a weekly basis.

Illunoprecipitation of p21 ras protein
The amount of p21 ras protein was assayed by the method described by
Wilson et al., 1990, with the following modifications. Cells were plated
at 10‘ cells/100 mm dish. The antibody, v13-259, which reacts with H-,K-,
and N-ras p21s was used. The ras protein migrates as a doublet band

(e.g., see autoradiographs depicted in Hurlin et al., 1989, and Wilson et

107

TABLE 1. Cell strains

 

Previously reported

 

Cell strain ras protein expression

designation relative to MSU-1.1 cells Reference

MSU-1.1 1.0 Morgan et al., 1991
MSU-1.1-Hg'CHATs -- Chapter II - this thesis
MSU-1.1-vK-ras-2T' 1.2-1.3 Fry et al., 1990

MSU-1.1-vK-ras-3T' --

MSU-1.1-H-ras-2T' 2.0-5.0
MSU-1.1-H-ras-3T' 2.0-5.0
MSU-1.1-N-ras-3T' 3.0-7.0
MSU-1.1-N-ras-8T' 3.0-7.0

unpublished studies
Hurlin et al., 1989
Hurlin et al., 1989
Wilson et al., 1990
Wilson et al., 1990

 

‘T: denotes that this is a strain derived from a tumor produced by the

respective ras-transformed MSU-1.1 strain.

 

108
al., 1990). The band densities were quantified using Visage 110 BioImage
System (Kodak Co., Ann Arbor, MI). For determination of H-ras expression
levels, only the upper band of the doublet (which contains only H-ras) was
quantified. For determination of N-ras expression levels, only the bottom
band of the doublet (which contains both N- and K-ras) was quantified.
The ras protein expression levels reported are relative to that of M50-

1.1-Hg'CHATs which was assigned a value of 1.

RESULTS

Fusion of malignant HSU-l.l-ras cells with the finite life
span, non-tulorigenic ISO-1.1 precursor cells.

MSU-1.1-Hg'CHAT‘ cells were fused with cells from six independently
derived malignant ras-transformed MSU-1.1 strains. The later are Hg'CHAT'.
From each of the cell fusions two or three independent hybrid clones were
isolated, and the clonal populations were expanded. The parental and
hybrid cells were injected into athymic mice to determine their
tumorigenic potential (Table 2). All six of the malignant parental cell
strains formed tumors at each of four injection sites within two months
post injection. In contrast, MSU-1.1-Hg'CHAT‘ cells failed to form tumors
at any of five injection sites over a period of 8.5 months. The
tumorigenic potential of the hybrid cell strains varied. Although some
were as tumorigenic as the malignant parental cells, most exhibited some
degree of suppression of tumorigenicity.

The degree'of tumor suppression in the hybrid strains, summarized in

Table 3, was scored as follows. If, for a particular hybrid strain,

109

TABLE 2. Tumorigenicity of cell strains

 

 

Injection sites Approximate time (mos)
Cell strain producing tumors until tumor size 1 cm3
MSU-1.1-Hg'CHATs 0/5 --
MSU-1.1-vK-ras-2T 4/4 2.0
MSU—1.1-vK-ras-3T 4/4 0.5
MSU-1.1-H-ras-2T 4/4 2.0
MSU-1.1-H-ras-3T 4/4 1.0
MSU-1.1-N-ras-3T 4/4 1.0
MSU-1.1-N-ras-8T 4/4 2.0
Hyb-vK-ras-2.1 0/4 --
Hyb-vK-ras-2.2 0/4 --
Hyb-vK-ras-2.3 O/4 --
Hyb-vK-ras-3.1 0/4 --
Hyb-vK-ras-3.2 0/3 --
Hyb-H-ras-2.1 0/4 --
Hyb-H-ras-2.2 2/4 5.0, 5.0
Hyb-H-ras-2.5 3 /4 2.5, 4.5, 4.5
Hyb-H-ras-3.1 0/2 --
Hyb-H-ras-3.4 2 /4 1.5, 4.0
Hyb-H-ras-3.9 0/4 --
Hyb-N-ras-3.1 2V4 2.0, 4.5
Hyb-N-ras-3.3 3V4 2.5, 2.5, 2.5
Hyb-N-ras-3.7 0/4 --
Hyb-N-ras-8.1 4/4 1.0, 1.0, 1.5, 1.5
Hyb-N-ras-8.2 4/4 1.0, 1.0, 1.0, 1.0
Hyb-N-ras-8.3 O/2 --

 

'Unable to adequately assess tumor formation at remaining sites because
mice were sacrificed prior to 8.5 months post injection. No tumors were

present at these sites at the time the mice were sacrificed.

110

tumors arose in all injection sites within two months, the hybrid strain
was scored as exhibiting no suppression of tumorigenicity. If no tumors
arose in any of the sites by 8.5 months post injection, the hybrid strain
was considered to be completely suppressed for tumorigenicity. If a
hybrid strain produced tumors but tumors did not arise at all sites, or if
some tumors formed after two months post injection, then the hybrid strain
was scored as being partially suppressed for tumorigenicity.

All five of the hybrids obtained by fusion of MSU-1.1-Hg'CHAT‘ cells
with MSU-1.1-vK-ras cells were completely suppressed for tumorigenicity.
Half of the six hybrids obtained from fusion of MSU-1.1-Hg'CHAT‘ cells with
MSU-1.1-H-ras cells were completely suppressed for tumorigenicity; the
other half were partially suppressed. The six hybrids obtained from
fusion of MSU-1.1-Hg'CHAT‘ cells with MSU-1.1-N-ras cells ranged from
complete, to partial, to no suppression, with two hybrids falling into

each category.

TABLE 3. Suppression of tumorigenicity in hybrid strains.

 

No. of hybrid strains

 

not partially completely
Cell fusion suppressed suppressed suppressed
MSU-1.1-vK-ras X MSU-1.1-Hg'CHATs 0 0 5
MSU-1.1-H-—ras X MSU-1.1-Hg"CHATs O 3 3

MSU-l . 1-N-ras X MSU-l . l-Hg'CHATs 2 2 2

 

 

111
Expression levels of ras protein in parental cells, hybrid cells,
and tuner-derived hybrid cells.

Because the tumorigenicity of the MSU-1.1-H-ras and MSU-1.1-N-ras
strains was reported to be dependent upon a high expression level of the
mutant ras protein (Hurlin et al.,1989; Wilson et al., 1990) and the
suppression of the ras-induced tumorigenicity' appeared to correlate
inversely with the reported expression levels of the ras protein in the
malignant parental cells (Table 1), I hypothesized that the suppression of
the ras-induced tumorigenicity in the hybrids cells may be secondary to a
dilution of the ras protein in the hybrid cells. To test this hypothesis,
I determined the ras protein expression of the H- and N-ras parental
cells, the hybrid cells obtained from fusion of these parents with MSU-
1.1-Hg'CHAT‘ cells, and cells that were derived from tumors that were
formed by these hybrids strains. Figures 1 and 2 depict the ras protein
expression of these cell strains relative to that of MSU-1.1-Hg'CHAT‘,
which was assigned a value of 1.

The ras protein expression levels of the MSU-1.1-H-ras-2T and MSU-
1.1-H-ras-3T cell strains were respectively 4.0 and 4.9, and that of the
hybrid strains obtained from fusion of MSU-1.1-Hg'CHAT’ with these two
malignant strains ranged from 1.4 (Hyb-H-ras-3.4) to 22.1 (Hyb-H-ras-2.5).
The ras protein expression levels of the MSU-1.1-N-ras-3T and MSU-1.1-N-
ras-8T cell strains were respectively 45.0 and 2.7, and that of the hybrid
strains obtained from fusion of MSU-1.1-Hg'CHAT‘ with these malignant
strains ranged from 7.8 (Hyb-N-ras-8.3) to 32.4 (Hyb-N-ras-3.7). The ras
protein expression levels in the various hybrid strains did not correlate
with their tumorigenic potentials. For example, the ras protein

expression of the non-tumorigenic hybrid strain Hyb-H-ras-3.2 was 10.4

112

 

1"!

30'
20"

‘IO"

 

RAS PROTEIN EXPRESSION

 

 

 

 

 

 

 

 

 

50

T11'

40

' I

RAS PROTEIN EXPRESSION

 

 

 

 

 

FIGURE 1. H-ras protein expression, relative to that of MSU-1.1-Hg'CHAT’
cells, for H-ras-transformed MSU-1.1 parental and hybrid cells.

113

 

A. 50

 

RAS PROTEIN EXPRESSION

 

 

 

 

 

 

 

 

 

 

RAS PROTEIN EXPRESSION

 

FIGURE 2. N-ras protein expression, relative to that of MSU-1.1-Hg'CHATs
cells, for N-ras-transformed MSU-1.1 parental and hybrid cells.

114

(2.1 times higher than that of the ualignant MSU-1.1-H-ras-3 parental
cells), whereas the ras expression of the tumorigenic hybrid strain Hyb-H-
ras—3.4 was only 1.4 (0.28 times that of the malignant MSU-1.1-H-ras-3
parental cells). I found, however, that the cells that were derived from
tumors that were formed by Hyb-H-ras-3.4 expressed much higher levels of
ras protein, i.e., 13.3 and 32.8, than did the pre-tumor population. This
appeared to be a general phenomenon, as seven of the nine tumor-derived
hybrid strains expressed higher levels of ras protein than did their
respective pre-tumor populations.

This latter finding, along with the observation that the ras
expression that I obtained for the MSU-1.1-N-ras-8T cell strain was lower
than that obtained previously for an earlier passage of this strain
(unpublished data), led me to hypothesize that growth of cells in athymic
cells increases the expression of ras, whereas growth of cells in tissue
culture dishes decreases the expression of ras. As a preliminary test of
this hypothesis, I compared the ras protein expression of late passage
MSU-1.1-N-ras-8T cells to that of early passage cells of this strain and
to that of very early passage cells that were derived from tumors formed
by early and late passage MSU-1.1-N-ras-8T cells. The results (shown in
Table 4) confirm that the late passage cells have a lower ras expression
than do the early passage cells and show, as predicted, that the very
early passage tumor-derived strains have a much higher ras expression than

do the pre-tumor populations.

115

TABLE 4. The ras protein expression of MSU-1.1-N-ras-8T cells measured at

various cell passages

 

Cell Passage Cell strain N-ras expression
late MSU-1.1-N-ras-BT 0.56
early MSU-1.1-N-ras-8T 2.59
very early Tumor strain derived from 20.11

late passage MSU—1.1-N-ras-8T

very early Tumor strain derived from 20.33

early passage MSU-1.1-N-ras-8T

 

116
DISCUSSION

The design of the human fibroblast cell fusion studies described
here closely parallels the studies conducted by Craig and Sager (1985)
with CHEF cells, and results similar to theirs were obtained. Our results
suggest, as did those of Craig and Sager, that normal cells contain genes
that are capable of suppressing ras-induced tumorigenicity, and indicate
that this suppression does not result from a reduction in the expression
of the ras protein. MSU-1.1 cells were found to suppress tumorigenicity
when they were fused with malignant ras-transformed MSU-1.1 derivative
cells, indicating that MSU-1.1 cells, although immortal, still possess a
functional ras-transformation suppressor gene.

Although both MSU-1.1 cells and normal human fibroblasts apparently
contain ras-transformation suppressor genes, MSU-1.1 cells can be
malignantly transformed by transfection with an activated ras oncogene
carried in a vector engineered to allow a high level of expression,
whereas normal human fibroblasts cannot. A plausible explanation for this
difference is that the infinite life span of MSU-1.1 cells allows them to
accumulate the minimum of three genetic alterations (acquisition of an
activated ras oncogene and inactivation of both copies of a ras-
transformation suppressor gene) required for ras-induced transformation,
whereas the life span of normal fibroblasts is too short to permit three
sequential genetic alterations and clonal selections.

A likely explanation for the isolation by Hurlin et al. (1989) of
MSU-1.1-H-ras strains that express high levels of mutant ras protein but
are non-tumorigenic is that these cells still possess a functional ras-

transformation suppressor gene. It is possible that many other non-

117
tumorigenic MSU-1.1-ras cells were present in the transfected populations
in the experiments of Hurlin et al. (1989), Wilson et al. (1990), and Fry
et al.,(1990), but that these were not detected because they did not form
foci.

Examination of the ras expression of the MSU-1.1-ras cell strains
revealed that the ras protein expression of a cell population is not
necessarily stable. The preliminary results presented here suggest that
growth of ras-transfected cells in athymic mice increases the ras protein
expression of the cell population, and growth of these cells in tissue
culture dishes decreases the ras expression of the cell population. This
modulation of ras protein expression could be achieved by selection for
cells with ras expression levels that are favorable to certain growth
conditions or by induction of altered ras expression in the cells. The
studies reported hereiwere not designed to differentiate between these two
possibilities. I observed, however, as haveiothers (Wilson et al., 1990),
that cells expressing a high level of mutant ras tend to round and detach
from tissue culture dishes. ‘This observation supports the hypothesis that
the growth of cells in tissue culture dishes decreases the ras protein
expression of the population by selecting for cells that have low ras
expression levels. Although further studies are needed to understand
these findings, the instability of the ras expression in these cells may
explain why the measured ras expression levels did not correlate with
tumorigenicity and indicate that caution must be exercised when
interpreting measurements of the ras expression levels of cell

populations.

REFERENCES

Barbacid, M. (1987). ras genes. Ann. Rev. Biochem. 56, 779-827.

Bunn, C.L., and Tarrant, G.M. (1980). Limited lifespan in somatic cell
hybrids and cybrids. Experimental Cell Res., 127, 385-396.

Craig, R.W., and Sager R. (1985). Supression of tumorigenicity in hybrids
of normal and oncogene-transformed CHEF cells. Proc. Natl. Acad. Sci. USA
82, 2062-2066.

Fry, 0.G., Milam, L.D., Dillberger,J.E., Maher, V.M., and McCormick, J.J.
(1990). Malignant transformation of an infinite life span human fibroblast
cell strain by transfection with v-Ki-ras. Oncogene 5, 1415-1418.

Geiser, A.G., Der, C.J., Marshall, C.J., and Stanbridge, E.J. (1986).
Supression of tumorigenicy with continued expresion of the c-Ha-ras
oncogene in EJ blader carcinoma-human fibroblast hybrid cells. Proc. Natl.
Acad. Sci. USA 83, 5209-5213.

Harris, H. (1988). The analysis of malignancy by cell fusion: the position
in 1988. Cancer Research 48, 3302-3306.

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-ras
oncogene. Cancer Research 47, 5752-5757.

Hurlin, P.J., Maher, V.M., and McCormick, J.J. (1989). Malignant
transformation of human fibroblasts caused by expression of a transfected
T24 HRAS oncogene. Proc. Natl. Acad. Sci. USA 86, 187-191.

Land, H., Parada, L.F., and Weinberg, R.A. (1983). Tumorigenic conversion
of primary embryo fibroblasts requires at least two cooperting oncogenes.
Nature 304, 596-602.

Morgan, T.L., Yang, 0., Fry, 0.G., Hurlin, P.J., Kohler, S.K., Maher,
V.M., and McCormick, J.J. (1991). Characteristics of an infinite life span
diploid human fibroblast cell strain and a near-diploid strain arising
from a clone of cells expressing a transfected v-myc oncogene.
Experimental Cell Reasearch 197, 125-136.

Muggleton-Harris, A.L., and DeSimone, D.W. (1980). Replicative potentials
of various fusion products between WI-38 and SV40 transformed WI-38 cells
and their components. Somatic Cell Genetics, 6, 689-698.

Newbold, R.F., amd Overell, R.W. (1983). Fibroblast inmortality is a
prerequisite for transformation by EJ c-Ha-ras oncogene. Nature 304, 648-
651.

Oshimura, M., Gilmer, T.M., and Barrett, J.C. (1985). Nonrandom loss of
cromosome 15 in Syrian hamster tumors induced by v-Ha-ras plus v-myc

118

119
oncogenes. Nature 316, 636-369.

Oshimura, M., Koi, M., Ozawa, N., Sugawara, 0., Lamb, P.W., and Barrett,
C.J. (1988). Role of chromosome loss in ras/myc-induced Syrian hamster
tumors. Cancer Research 48, 1623-1632.

Parada, L.F., Land, H., Weinberg, R.A., Wolf, 0., and Rotter, V. (1984).
Cooperation between gene encoding p53 tumour antegen and ras in cellular
transformation. Nature 312, 649-651.

Pereira-Smith, 0.M., and Smith, J.R. (1981). Expression of SV40T antigen
in finite life-span hybrids of normal and SV40-transformed fibroblasts.
Somatic Cell Genetics, 7, 411-421.

Pereira-Smith, 0.M., and Smith, J.R. (1983). Evidence for the recessive
nature of cellular immortality. Science, 221, 964-966.

Ruley, H.E. (1983). Adenovirus early region 1A enables viral and cellular
transforming genes to transform primary cells in culture. Nature 304, 602-
606.

Smith, B.L., Anisowicz, A., Chodosh, L.A., and Sager, R. (1982). DNA
transfer of focus- and tumor-forming ability into nontumorigenic CHEF
cells. Proc. Natl. Acad. Sci. USA 79, 1964-1968.

Sugawara, 0., Oshimura, M., Koi, M., Annab, L.A., and Barret, J.C. (1990).
Induction of cellular senescence in inmortalized cells by human chromosome
1. Science 247, 707-710.

Willecke, K., Griegel, S., Martin, W., Traub, 0., and Schifer, R. (1987).
The Ha-ras-induced transformed phenotype of Rat-1 cells can be supressed
in hybryds with rat embryonic fibroblasts. Journal of Cellular
Biochemistry 34, 23-30.

Wilson, 0.M., Fry, 0.G., Maher, V.M., and McCormick, J.J. (1989).
Transformation of diploid human fibroblasts by transfection of N-ras-
oncogenes. Carcinogenesis 10, 635-640.

Wilson, 0.M., Yang, 0., Oillberger, J.E., Dietrich, S.E., Maher, V.M.,
McCormick, J.J. (1990).Malignant transformation of human fibroblasts by a
transfected n-ras oncogene. Cancer Research 50, 5587-5593.

 

"11111111!"11111111“